{ "best_metric": null, "best_model_checkpoint": null, "epoch": 4.0, "eval_steps": 3190, "global_step": 51040, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 7.836990595611286e-05, "grad_norm": 2.000732183456421, "learning_rate": 1.0000000000000001e-07, "loss": 1.5363, "step": 1 }, { "epoch": 7.836990595611286e-05, "eval_loss": 2.6673760414123535, "eval_runtime": 107.7821, "eval_samples_per_second": 24.93, "eval_steps_per_second": 6.235, "step": 1 }, { "epoch": 0.00015673981191222572, "grad_norm": 2.0449366569519043, "learning_rate": 2.0000000000000002e-07, "loss": 1.9456, "step": 2 }, { "epoch": 0.00023510971786833856, "grad_norm": 2.5398354530334473, "learning_rate": 3.0000000000000004e-07, "loss": 1.5086, "step": 3 }, { "epoch": 0.00031347962382445143, "grad_norm": 3.4343698024749756, "learning_rate": 4.0000000000000003e-07, "loss": 1.72, "step": 4 }, { "epoch": 0.00039184952978056425, "grad_norm": 2.3693788051605225, "learning_rate": 5.000000000000001e-07, "loss": 2.4189, "step": 5 }, { "epoch": 0.0004702194357366771, "grad_norm": 2.37947154045105, "learning_rate": 6.000000000000001e-07, "loss": 1.7211, "step": 6 }, { "epoch": 0.00054858934169279, "grad_norm": 1.8710862398147583, "learning_rate": 7.000000000000001e-07, "loss": 1.7106, "step": 7 }, { "epoch": 0.0006269592476489029, "grad_norm": 2.6463980674743652, "learning_rate": 8.000000000000001e-07, "loss": 2.0555, "step": 8 }, { "epoch": 0.0007053291536050156, "grad_norm": 1.7873148918151855, "learning_rate": 9.000000000000001e-07, "loss": 1.8861, "step": 9 }, { "epoch": 0.0007836990595611285, "grad_norm": 2.225167751312256, "learning_rate": 1.0000000000000002e-06, "loss": 1.9058, "step": 10 }, { "epoch": 0.0008620689655172414, "grad_norm": 4.56266450881958, "learning_rate": 1.1e-06, "loss": 1.8828, "step": 11 }, { "epoch": 0.0009404388714733542, "grad_norm": 1.672350525856018, "learning_rate": 1.2000000000000002e-06, "loss": 1.8847, "step": 12 }, { "epoch": 0.0010188087774294671, "grad_norm": 1.429488182067871, "learning_rate": 1.3e-06, "loss": 1.7607, "step": 13 }, { "epoch": 0.00109717868338558, "grad_norm": 1.3268245458602905, "learning_rate": 1.4000000000000001e-06, "loss": 1.6577, "step": 14 }, { "epoch": 0.0011755485893416929, "grad_norm": 1.994361162185669, "learning_rate": 1.5e-06, "loss": 1.8638, "step": 15 }, { "epoch": 0.0012539184952978057, "grad_norm": 1.6489429473876953, "learning_rate": 1.6000000000000001e-06, "loss": 1.4157, "step": 16 }, { "epoch": 0.0013322884012539186, "grad_norm": 1.6975210905075073, "learning_rate": 1.7000000000000002e-06, "loss": 1.732, "step": 17 }, { "epoch": 0.0014106583072100312, "grad_norm": 1.9607741832733154, "learning_rate": 1.8000000000000001e-06, "loss": 1.9753, "step": 18 }, { "epoch": 0.0014890282131661441, "grad_norm": 1.9672473669052124, "learning_rate": 1.9000000000000002e-06, "loss": 1.9246, "step": 19 }, { "epoch": 0.001567398119122257, "grad_norm": 1.8094316720962524, "learning_rate": 2.0000000000000003e-06, "loss": 1.2836, "step": 20 }, { "epoch": 0.0016457680250783699, "grad_norm": 1.8976380825042725, "learning_rate": 2.1000000000000002e-06, "loss": 1.7655, "step": 21 }, { "epoch": 0.0017241379310344827, "grad_norm": 3.6831774711608887, "learning_rate": 2.2e-06, "loss": 2.3351, "step": 22 }, { "epoch": 0.0018025078369905956, "grad_norm": 3.2898950576782227, "learning_rate": 2.3000000000000004e-06, "loss": 1.3007, "step": 23 }, { "epoch": 0.0018808777429467085, "grad_norm": 1.5813645124435425, "learning_rate": 2.4000000000000003e-06, "loss": 1.6024, "step": 24 }, { "epoch": 0.001959247648902821, "grad_norm": 2.538123607635498, "learning_rate": 2.5e-06, "loss": 1.9458, "step": 25 }, { "epoch": 0.0020376175548589342, "grad_norm": 2.9492053985595703, "learning_rate": 2.6e-06, "loss": 2.5114, "step": 26 }, { "epoch": 0.002115987460815047, "grad_norm": 1.724377155303955, "learning_rate": 2.7000000000000004e-06, "loss": 1.923, "step": 27 }, { "epoch": 0.00219435736677116, "grad_norm": 2.4704678058624268, "learning_rate": 2.8000000000000003e-06, "loss": 2.3528, "step": 28 }, { "epoch": 0.0022727272727272726, "grad_norm": 1.4286386966705322, "learning_rate": 2.9e-06, "loss": 1.4729, "step": 29 }, { "epoch": 0.0023510971786833857, "grad_norm": 2.0248165130615234, "learning_rate": 3e-06, "loss": 1.8362, "step": 30 }, { "epoch": 0.0024294670846394984, "grad_norm": 1.911228895187378, "learning_rate": 3.1000000000000004e-06, "loss": 2.1803, "step": 31 }, { "epoch": 0.0025078369905956114, "grad_norm": 2.7764575481414795, "learning_rate": 3.2000000000000003e-06, "loss": 2.2355, "step": 32 }, { "epoch": 0.002586206896551724, "grad_norm": 1.921965479850769, "learning_rate": 3.3000000000000006e-06, "loss": 1.7646, "step": 33 }, { "epoch": 0.002664576802507837, "grad_norm": 10.365942001342773, "learning_rate": 3.4000000000000005e-06, "loss": 3.2268, "step": 34 }, { "epoch": 0.00274294670846395, "grad_norm": 3.3987040519714355, "learning_rate": 3.5e-06, "loss": 2.3849, "step": 35 }, { "epoch": 0.0028213166144200625, "grad_norm": 2.0284228324890137, "learning_rate": 3.6000000000000003e-06, "loss": 1.8434, "step": 36 }, { "epoch": 0.0028996865203761756, "grad_norm": 1.9726064205169678, "learning_rate": 3.7e-06, "loss": 1.6637, "step": 37 }, { "epoch": 0.0029780564263322882, "grad_norm": 4.0307464599609375, "learning_rate": 3.8000000000000005e-06, "loss": 2.4971, "step": 38 }, { "epoch": 0.0030564263322884013, "grad_norm": 3.0770955085754395, "learning_rate": 3.900000000000001e-06, "loss": 1.9542, "step": 39 }, { "epoch": 0.003134796238244514, "grad_norm": 2.367420196533203, "learning_rate": 4.000000000000001e-06, "loss": 1.5655, "step": 40 }, { "epoch": 0.003213166144200627, "grad_norm": 2.6458802223205566, "learning_rate": 4.1e-06, "loss": 2.0123, "step": 41 }, { "epoch": 0.0032915360501567397, "grad_norm": 1.82381010055542, "learning_rate": 4.2000000000000004e-06, "loss": 1.7574, "step": 42 }, { "epoch": 0.003369905956112853, "grad_norm": 1.7684311866760254, "learning_rate": 4.3e-06, "loss": 1.7137, "step": 43 }, { "epoch": 0.0034482758620689655, "grad_norm": 2.959888458251953, "learning_rate": 4.4e-06, "loss": 1.8448, "step": 44 }, { "epoch": 0.0035266457680250786, "grad_norm": 2.4081685543060303, "learning_rate": 4.5e-06, "loss": 1.945, "step": 45 }, { "epoch": 0.003605015673981191, "grad_norm": 1.538400650024414, "learning_rate": 4.600000000000001e-06, "loss": 1.7192, "step": 46 }, { "epoch": 0.003683385579937304, "grad_norm": 2.014732837677002, "learning_rate": 4.7e-06, "loss": 1.9161, "step": 47 }, { "epoch": 0.003761755485893417, "grad_norm": 3.009460687637329, "learning_rate": 4.800000000000001e-06, "loss": 1.9414, "step": 48 }, { "epoch": 0.0038401253918495296, "grad_norm": 2.2501301765441895, "learning_rate": 4.9000000000000005e-06, "loss": 1.7802, "step": 49 }, { "epoch": 0.003918495297805642, "grad_norm": 6.4327802658081055, "learning_rate": 5e-06, "loss": 1.9543, "step": 50 }, { "epoch": 0.003996865203761755, "grad_norm": 2.0045437812805176, "learning_rate": 5.1e-06, "loss": 1.7196, "step": 51 }, { "epoch": 0.0040752351097178684, "grad_norm": 2.366637706756592, "learning_rate": 5.2e-06, "loss": 1.8601, "step": 52 }, { "epoch": 0.0041536050156739815, "grad_norm": 2.9116194248199463, "learning_rate": 5.300000000000001e-06, "loss": 2.2047, "step": 53 }, { "epoch": 0.004231974921630094, "grad_norm": 1.8513481616973877, "learning_rate": 5.400000000000001e-06, "loss": 2.0334, "step": 54 }, { "epoch": 0.004310344827586207, "grad_norm": 3.934149742126465, "learning_rate": 5.500000000000001e-06, "loss": 1.8477, "step": 55 }, { "epoch": 0.00438871473354232, "grad_norm": 2.2079784870147705, "learning_rate": 5.600000000000001e-06, "loss": 1.6278, "step": 56 }, { "epoch": 0.004467084639498433, "grad_norm": 2.5320563316345215, "learning_rate": 5.7e-06, "loss": 1.8694, "step": 57 }, { "epoch": 0.004545454545454545, "grad_norm": 2.9063713550567627, "learning_rate": 5.8e-06, "loss": 2.4397, "step": 58 }, { "epoch": 0.004623824451410658, "grad_norm": 2.527782440185547, "learning_rate": 5.9e-06, "loss": 1.9972, "step": 59 }, { "epoch": 0.004702194357366771, "grad_norm": 5.2695817947387695, "learning_rate": 6e-06, "loss": 2.379, "step": 60 }, { "epoch": 0.004780564263322884, "grad_norm": 1.702518343925476, "learning_rate": 6.1e-06, "loss": 1.6556, "step": 61 }, { "epoch": 0.004858934169278997, "grad_norm": 2.8417506217956543, "learning_rate": 6.200000000000001e-06, "loss": 1.784, "step": 62 }, { "epoch": 0.00493730407523511, "grad_norm": 1.8618627786636353, "learning_rate": 6.300000000000001e-06, "loss": 1.9011, "step": 63 }, { "epoch": 0.005015673981191223, "grad_norm": 1.5261634588241577, "learning_rate": 6.4000000000000006e-06, "loss": 1.678, "step": 64 }, { "epoch": 0.005094043887147335, "grad_norm": 2.456555128097534, "learning_rate": 6.5000000000000004e-06, "loss": 1.437, "step": 65 }, { "epoch": 0.005172413793103448, "grad_norm": 2.7490694522857666, "learning_rate": 6.600000000000001e-06, "loss": 1.7615, "step": 66 }, { "epoch": 0.005250783699059561, "grad_norm": 2.2953333854675293, "learning_rate": 6.700000000000001e-06, "loss": 1.7448, "step": 67 }, { "epoch": 0.005329153605015674, "grad_norm": 2.2576181888580322, "learning_rate": 6.800000000000001e-06, "loss": 1.3289, "step": 68 }, { "epoch": 0.005407523510971787, "grad_norm": 3.445425510406494, "learning_rate": 6.9e-06, "loss": 2.2761, "step": 69 }, { "epoch": 0.0054858934169279, "grad_norm": 3.0262675285339355, "learning_rate": 7e-06, "loss": 1.604, "step": 70 }, { "epoch": 0.005564263322884013, "grad_norm": 2.668276309967041, "learning_rate": 7.100000000000001e-06, "loss": 1.927, "step": 71 }, { "epoch": 0.005642633228840125, "grad_norm": 2.8227899074554443, "learning_rate": 7.2000000000000005e-06, "loss": 1.7573, "step": 72 }, { "epoch": 0.005721003134796238, "grad_norm": 1.7064225673675537, "learning_rate": 7.3e-06, "loss": 1.4902, "step": 73 }, { "epoch": 0.005799373040752351, "grad_norm": 1.816241979598999, "learning_rate": 7.4e-06, "loss": 1.8613, "step": 74 }, { "epoch": 0.005877742946708464, "grad_norm": 3.6941092014312744, "learning_rate": 7.500000000000001e-06, "loss": 1.4836, "step": 75 }, { "epoch": 0.0059561128526645765, "grad_norm": 1.616467833518982, "learning_rate": 7.600000000000001e-06, "loss": 1.6826, "step": 76 }, { "epoch": 0.00603448275862069, "grad_norm": 2.3458244800567627, "learning_rate": 7.7e-06, "loss": 1.8602, "step": 77 }, { "epoch": 0.006112852664576803, "grad_norm": 2.6882436275482178, "learning_rate": 7.800000000000002e-06, "loss": 1.0387, "step": 78 }, { "epoch": 0.006191222570532916, "grad_norm": 3.8694403171539307, "learning_rate": 7.9e-06, "loss": 2.0927, "step": 79 }, { "epoch": 0.006269592476489028, "grad_norm": 5.9689202308654785, "learning_rate": 8.000000000000001e-06, "loss": 1.6218, "step": 80 }, { "epoch": 0.006347962382445141, "grad_norm": 2.2513155937194824, "learning_rate": 8.1e-06, "loss": 1.5053, "step": 81 }, { "epoch": 0.006426332288401254, "grad_norm": 2.6841421127319336, "learning_rate": 8.2e-06, "loss": 1.7643, "step": 82 }, { "epoch": 0.006504702194357366, "grad_norm": 3.2995026111602783, "learning_rate": 8.3e-06, "loss": 1.8717, "step": 83 }, { "epoch": 0.0065830721003134795, "grad_norm": 2.6499273777008057, "learning_rate": 8.400000000000001e-06, "loss": 1.4626, "step": 84 }, { "epoch": 0.0066614420062695925, "grad_norm": 4.517232418060303, "learning_rate": 8.5e-06, "loss": 1.8598, "step": 85 }, { "epoch": 0.006739811912225706, "grad_norm": 2.690030813217163, "learning_rate": 8.6e-06, "loss": 1.612, "step": 86 }, { "epoch": 0.006818181818181818, "grad_norm": 2.5901801586151123, "learning_rate": 8.700000000000001e-06, "loss": 1.7486, "step": 87 }, { "epoch": 0.006896551724137931, "grad_norm": 9.752556800842285, "learning_rate": 8.8e-06, "loss": 2.9753, "step": 88 }, { "epoch": 0.006974921630094044, "grad_norm": 3.7411394119262695, "learning_rate": 8.900000000000001e-06, "loss": 1.7893, "step": 89 }, { "epoch": 0.007053291536050157, "grad_norm": 4.1588521003723145, "learning_rate": 9e-06, "loss": 2.0379, "step": 90 }, { "epoch": 0.007131661442006269, "grad_norm": 6.545043468475342, "learning_rate": 9.100000000000001e-06, "loss": 1.6656, "step": 91 }, { "epoch": 0.007210031347962382, "grad_norm": 2.168325901031494, "learning_rate": 9.200000000000002e-06, "loss": 1.9699, "step": 92 }, { "epoch": 0.0072884012539184955, "grad_norm": 1.800024390220642, "learning_rate": 9.3e-06, "loss": 1.6449, "step": 93 }, { "epoch": 0.007366771159874608, "grad_norm": 2.750563859939575, "learning_rate": 9.4e-06, "loss": 1.8265, "step": 94 }, { "epoch": 0.007445141065830721, "grad_norm": 2.593003511428833, "learning_rate": 9.5e-06, "loss": 1.7427, "step": 95 }, { "epoch": 0.007523510971786834, "grad_norm": 2.5204875469207764, "learning_rate": 9.600000000000001e-06, "loss": 1.9626, "step": 96 }, { "epoch": 0.007601880877742947, "grad_norm": 3.593327045440674, "learning_rate": 9.7e-06, "loss": 2.0768, "step": 97 }, { "epoch": 0.007680250783699059, "grad_norm": 2.332038164138794, "learning_rate": 9.800000000000001e-06, "loss": 1.7061, "step": 98 }, { "epoch": 0.007758620689655172, "grad_norm": 7.444894790649414, "learning_rate": 9.9e-06, "loss": 3.4639, "step": 99 }, { "epoch": 0.007836990595611285, "grad_norm": 2.3883562088012695, "learning_rate": 1e-05, "loss": 1.8937, "step": 100 }, { "epoch": 0.007915360501567398, "grad_norm": 3.40628719329834, "learning_rate": 9.999999990491285e-06, "loss": 1.7073, "step": 101 }, { "epoch": 0.00799373040752351, "grad_norm": 2.068798542022705, "learning_rate": 9.999999961965137e-06, "loss": 1.8256, "step": 102 }, { "epoch": 0.008072100313479625, "grad_norm": 3.372065544128418, "learning_rate": 9.999999914421558e-06, "loss": 2.0317, "step": 103 }, { "epoch": 0.008150470219435737, "grad_norm": 2.943406820297241, "learning_rate": 9.999999847860546e-06, "loss": 1.8469, "step": 104 }, { "epoch": 0.008228840125391849, "grad_norm": 4.211577415466309, "learning_rate": 9.999999762282105e-06, "loss": 1.5596, "step": 105 }, { "epoch": 0.008307210031347963, "grad_norm": 2.4062342643737793, "learning_rate": 9.999999657686231e-06, "loss": 1.6292, "step": 106 }, { "epoch": 0.008385579937304075, "grad_norm": 3.0066299438476562, "learning_rate": 9.999999534072928e-06, "loss": 2.0616, "step": 107 }, { "epoch": 0.008463949843260187, "grad_norm": 2.6405446529388428, "learning_rate": 9.999999391442195e-06, "loss": 1.3714, "step": 108 }, { "epoch": 0.008542319749216301, "grad_norm": 2.0420992374420166, "learning_rate": 9.999999229794031e-06, "loss": 1.6939, "step": 109 }, { "epoch": 0.008620689655172414, "grad_norm": 4.973214626312256, "learning_rate": 9.99999904912844e-06, "loss": 1.8013, "step": 110 }, { "epoch": 0.008699059561128526, "grad_norm": 2.7330503463745117, "learning_rate": 9.999998849445418e-06, "loss": 1.2381, "step": 111 }, { "epoch": 0.00877742946708464, "grad_norm": 3.702444314956665, "learning_rate": 9.99999863074497e-06, "loss": 2.1447, "step": 112 }, { "epoch": 0.008855799373040752, "grad_norm": 2.3001253604888916, "learning_rate": 9.999998393027096e-06, "loss": 1.5913, "step": 113 }, { "epoch": 0.008934169278996866, "grad_norm": 3.067291259765625, "learning_rate": 9.999998136291796e-06, "loss": 1.858, "step": 114 }, { "epoch": 0.009012539184952978, "grad_norm": 7.504228591918945, "learning_rate": 9.99999786053907e-06, "loss": 2.105, "step": 115 }, { "epoch": 0.00909090909090909, "grad_norm": 2.427489757537842, "learning_rate": 9.999997565768922e-06, "loss": 1.8571, "step": 116 }, { "epoch": 0.009169278996865204, "grad_norm": 2.1113977432250977, "learning_rate": 9.99999725198135e-06, "loss": 1.7114, "step": 117 }, { "epoch": 0.009247648902821317, "grad_norm": 2.0000805854797363, "learning_rate": 9.999996919176358e-06, "loss": 1.9736, "step": 118 }, { "epoch": 0.009326018808777429, "grad_norm": 1.7999093532562256, "learning_rate": 9.999996567353945e-06, "loss": 1.4964, "step": 119 }, { "epoch": 0.009404388714733543, "grad_norm": 3.566256523132324, "learning_rate": 9.999996196514115e-06, "loss": 1.8749, "step": 120 }, { "epoch": 0.009482758620689655, "grad_norm": 3.0884242057800293, "learning_rate": 9.999995806656866e-06, "loss": 1.6258, "step": 121 }, { "epoch": 0.009561128526645767, "grad_norm": 4.205544948577881, "learning_rate": 9.9999953977822e-06, "loss": 1.6567, "step": 122 }, { "epoch": 0.009639498432601881, "grad_norm": 4.235594749450684, "learning_rate": 9.99999496989012e-06, "loss": 2.1656, "step": 123 }, { "epoch": 0.009717868338557993, "grad_norm": 4.554929733276367, "learning_rate": 9.99999452298063e-06, "loss": 2.3383, "step": 124 }, { "epoch": 0.009796238244514107, "grad_norm": 3.0587403774261475, "learning_rate": 9.999994057053726e-06, "loss": 1.5004, "step": 125 }, { "epoch": 0.00987460815047022, "grad_norm": 3.047222137451172, "learning_rate": 9.999993572109414e-06, "loss": 1.8734, "step": 126 }, { "epoch": 0.009952978056426332, "grad_norm": 13.213157653808594, "learning_rate": 9.999993068147695e-06, "loss": 2.4945, "step": 127 }, { "epoch": 0.010031347962382446, "grad_norm": 2.3871679306030273, "learning_rate": 9.999992545168571e-06, "loss": 1.7867, "step": 128 }, { "epoch": 0.010109717868338558, "grad_norm": 11.418910026550293, "learning_rate": 9.999992003172042e-06, "loss": 2.1823, "step": 129 }, { "epoch": 0.01018808777429467, "grad_norm": 2.504373073577881, "learning_rate": 9.999991442158113e-06, "loss": 1.5334, "step": 130 }, { "epoch": 0.010266457680250784, "grad_norm": 3.6313090324401855, "learning_rate": 9.999990862126783e-06, "loss": 1.9961, "step": 131 }, { "epoch": 0.010344827586206896, "grad_norm": 1.9661351442337036, "learning_rate": 9.999990263078057e-06, "loss": 0.8363, "step": 132 }, { "epoch": 0.010423197492163009, "grad_norm": 2.472482204437256, "learning_rate": 9.999989645011936e-06, "loss": 2.1186, "step": 133 }, { "epoch": 0.010501567398119123, "grad_norm": 2.6644070148468018, "learning_rate": 9.999989007928424e-06, "loss": 1.655, "step": 134 }, { "epoch": 0.010579937304075235, "grad_norm": 3.118830680847168, "learning_rate": 9.999988351827518e-06, "loss": 1.5536, "step": 135 }, { "epoch": 0.010658307210031349, "grad_norm": 2.73866605758667, "learning_rate": 9.999987676709228e-06, "loss": 1.5982, "step": 136 }, { "epoch": 0.010736677115987461, "grad_norm": 2.1344823837280273, "learning_rate": 9.999986982573553e-06, "loss": 1.2754, "step": 137 }, { "epoch": 0.010815047021943573, "grad_norm": 5.359356880187988, "learning_rate": 9.999986269420493e-06, "loss": 1.3459, "step": 138 }, { "epoch": 0.010893416927899687, "grad_norm": 4.063244819641113, "learning_rate": 9.999985537250057e-06, "loss": 2.0954, "step": 139 }, { "epoch": 0.0109717868338558, "grad_norm": 2.357769012451172, "learning_rate": 9.99998478606224e-06, "loss": 1.3829, "step": 140 }, { "epoch": 0.011050156739811912, "grad_norm": 10.983322143554688, "learning_rate": 9.999984015857053e-06, "loss": 2.1144, "step": 141 }, { "epoch": 0.011128526645768026, "grad_norm": 5.38508939743042, "learning_rate": 9.999983226634493e-06, "loss": 1.8356, "step": 142 }, { "epoch": 0.011206896551724138, "grad_norm": 2.7678067684173584, "learning_rate": 9.999982418394565e-06, "loss": 1.4708, "step": 143 }, { "epoch": 0.01128526645768025, "grad_norm": 3.517974853515625, "learning_rate": 9.999981591137272e-06, "loss": 1.9154, "step": 144 }, { "epoch": 0.011363636363636364, "grad_norm": 2.2770638465881348, "learning_rate": 9.999980744862618e-06, "loss": 1.517, "step": 145 }, { "epoch": 0.011442006269592476, "grad_norm": 2.865394353866577, "learning_rate": 9.999979879570605e-06, "loss": 1.448, "step": 146 }, { "epoch": 0.01152037617554859, "grad_norm": 2.9288253784179688, "learning_rate": 9.999978995261235e-06, "loss": 1.6494, "step": 147 }, { "epoch": 0.011598746081504702, "grad_norm": 4.021169185638428, "learning_rate": 9.999978091934516e-06, "loss": 1.5551, "step": 148 }, { "epoch": 0.011677115987460815, "grad_norm": 3.0731611251831055, "learning_rate": 9.999977169590448e-06, "loss": 1.5013, "step": 149 }, { "epoch": 0.011755485893416929, "grad_norm": 2.696424961090088, "learning_rate": 9.999976228229033e-06, "loss": 1.5734, "step": 150 }, { "epoch": 0.01183385579937304, "grad_norm": 3.94726824760437, "learning_rate": 9.999975267850277e-06, "loss": 1.5494, "step": 151 }, { "epoch": 0.011912225705329153, "grad_norm": 2.868079662322998, "learning_rate": 9.999974288454183e-06, "loss": 1.3224, "step": 152 }, { "epoch": 0.011990595611285267, "grad_norm": 3.8285090923309326, "learning_rate": 9.999973290040756e-06, "loss": 1.8648, "step": 153 }, { "epoch": 0.01206896551724138, "grad_norm": 2.2175371646881104, "learning_rate": 9.99997227261e-06, "loss": 1.2521, "step": 154 }, { "epoch": 0.012147335423197491, "grad_norm": 2.6777446269989014, "learning_rate": 9.999971236161917e-06, "loss": 1.523, "step": 155 }, { "epoch": 0.012225705329153605, "grad_norm": 11.39714527130127, "learning_rate": 9.99997018069651e-06, "loss": 1.6006, "step": 156 }, { "epoch": 0.012304075235109718, "grad_norm": 2.4597437381744385, "learning_rate": 9.999969106213786e-06, "loss": 1.6994, "step": 157 }, { "epoch": 0.012382445141065831, "grad_norm": 3.254561185836792, "learning_rate": 9.999968012713746e-06, "loss": 1.9339, "step": 158 }, { "epoch": 0.012460815047021944, "grad_norm": 4.502783298492432, "learning_rate": 9.999966900196395e-06, "loss": 2.1291, "step": 159 }, { "epoch": 0.012539184952978056, "grad_norm": 3.507673740386963, "learning_rate": 9.99996576866174e-06, "loss": 1.649, "step": 160 }, { "epoch": 0.01261755485893417, "grad_norm": 3.166006088256836, "learning_rate": 9.999964618109785e-06, "loss": 2.1624, "step": 161 }, { "epoch": 0.012695924764890282, "grad_norm": 2.622565746307373, "learning_rate": 9.99996344854053e-06, "loss": 1.9427, "step": 162 }, { "epoch": 0.012774294670846394, "grad_norm": 3.695037603378296, "learning_rate": 9.999962259953983e-06, "loss": 1.7139, "step": 163 }, { "epoch": 0.012852664576802508, "grad_norm": 3.588526725769043, "learning_rate": 9.999961052350149e-06, "loss": 2.2892, "step": 164 }, { "epoch": 0.01293103448275862, "grad_norm": 3.2208471298217773, "learning_rate": 9.99995982572903e-06, "loss": 1.2275, "step": 165 }, { "epoch": 0.013009404388714733, "grad_norm": 2.4216411113739014, "learning_rate": 9.999958580090632e-06, "loss": 1.2133, "step": 166 }, { "epoch": 0.013087774294670847, "grad_norm": 2.6174747943878174, "learning_rate": 9.999957315434961e-06, "loss": 1.4724, "step": 167 }, { "epoch": 0.013166144200626959, "grad_norm": 3.8765318393707275, "learning_rate": 9.99995603176202e-06, "loss": 1.7823, "step": 168 }, { "epoch": 0.013244514106583073, "grad_norm": 2.646132469177246, "learning_rate": 9.999954729071812e-06, "loss": 1.583, "step": 169 }, { "epoch": 0.013322884012539185, "grad_norm": 5.0111894607543945, "learning_rate": 9.999953407364347e-06, "loss": 1.7992, "step": 170 }, { "epoch": 0.013401253918495297, "grad_norm": 2.9633617401123047, "learning_rate": 9.999952066639627e-06, "loss": 1.9062, "step": 171 }, { "epoch": 0.013479623824451411, "grad_norm": 3.11181902885437, "learning_rate": 9.999950706897657e-06, "loss": 1.6072, "step": 172 }, { "epoch": 0.013557993730407523, "grad_norm": 3.698693037033081, "learning_rate": 9.99994932813844e-06, "loss": 1.4585, "step": 173 }, { "epoch": 0.013636363636363636, "grad_norm": 2.862200975418091, "learning_rate": 9.999947930361988e-06, "loss": 1.5318, "step": 174 }, { "epoch": 0.01371473354231975, "grad_norm": 2.5534281730651855, "learning_rate": 9.9999465135683e-06, "loss": 1.4798, "step": 175 }, { "epoch": 0.013793103448275862, "grad_norm": 5.496286392211914, "learning_rate": 9.999945077757384e-06, "loss": 2.2101, "step": 176 }, { "epoch": 0.013871473354231974, "grad_norm": 3.6966469287872314, "learning_rate": 9.999943622929247e-06, "loss": 1.5383, "step": 177 }, { "epoch": 0.013949843260188088, "grad_norm": 3.5947751998901367, "learning_rate": 9.99994214908389e-06, "loss": 1.5537, "step": 178 }, { "epoch": 0.0140282131661442, "grad_norm": 4.50016975402832, "learning_rate": 9.999940656221324e-06, "loss": 1.5883, "step": 179 }, { "epoch": 0.014106583072100314, "grad_norm": 2.840778112411499, "learning_rate": 9.999939144341549e-06, "loss": 1.7145, "step": 180 }, { "epoch": 0.014184952978056426, "grad_norm": 2.674307107925415, "learning_rate": 9.999937613444574e-06, "loss": 1.5999, "step": 181 }, { "epoch": 0.014263322884012539, "grad_norm": 2.655076026916504, "learning_rate": 9.999936063530404e-06, "loss": 1.5245, "step": 182 }, { "epoch": 0.014341692789968653, "grad_norm": 3.7354049682617188, "learning_rate": 9.999934494599047e-06, "loss": 1.6822, "step": 183 }, { "epoch": 0.014420062695924765, "grad_norm": 2.249037742614746, "learning_rate": 9.999932906650507e-06, "loss": 1.3365, "step": 184 }, { "epoch": 0.014498432601880877, "grad_norm": 3.3796589374542236, "learning_rate": 9.99993129968479e-06, "loss": 1.6986, "step": 185 }, { "epoch": 0.014576802507836991, "grad_norm": 5.090536594390869, "learning_rate": 9.999929673701903e-06, "loss": 1.4254, "step": 186 }, { "epoch": 0.014655172413793103, "grad_norm": 3.502622127532959, "learning_rate": 9.999928028701851e-06, "loss": 1.4192, "step": 187 }, { "epoch": 0.014733542319749215, "grad_norm": 3.7056591510772705, "learning_rate": 9.999926364684642e-06, "loss": 1.6948, "step": 188 }, { "epoch": 0.01481191222570533, "grad_norm": 8.842498779296875, "learning_rate": 9.999924681650279e-06, "loss": 1.8294, "step": 189 }, { "epoch": 0.014890282131661442, "grad_norm": 3.7301878929138184, "learning_rate": 9.999922979598773e-06, "loss": 1.8537, "step": 190 }, { "epoch": 0.014968652037617556, "grad_norm": 16.6197452545166, "learning_rate": 9.999921258530127e-06, "loss": 2.2627, "step": 191 }, { "epoch": 0.015047021943573668, "grad_norm": 3.0591611862182617, "learning_rate": 9.999919518444348e-06, "loss": 1.6981, "step": 192 }, { "epoch": 0.01512539184952978, "grad_norm": 4.088306427001953, "learning_rate": 9.999917759341444e-06, "loss": 1.4246, "step": 193 }, { "epoch": 0.015203761755485894, "grad_norm": 3.8525381088256836, "learning_rate": 9.999915981221421e-06, "loss": 1.7458, "step": 194 }, { "epoch": 0.015282131661442006, "grad_norm": 2.747241258621216, "learning_rate": 9.999914184084287e-06, "loss": 1.5587, "step": 195 }, { "epoch": 0.015360501567398118, "grad_norm": 2.7729110717773438, "learning_rate": 9.999912367930045e-06, "loss": 1.4263, "step": 196 }, { "epoch": 0.015438871473354232, "grad_norm": 3.0775208473205566, "learning_rate": 9.999910532758705e-06, "loss": 1.4207, "step": 197 }, { "epoch": 0.015517241379310345, "grad_norm": 3.807436466217041, "learning_rate": 9.999908678570275e-06, "loss": 1.7861, "step": 198 }, { "epoch": 0.015595611285266457, "grad_norm": 3.8611457347869873, "learning_rate": 9.999906805364761e-06, "loss": 2.04, "step": 199 }, { "epoch": 0.01567398119122257, "grad_norm": 4.101282596588135, "learning_rate": 9.999904913142166e-06, "loss": 1.9242, "step": 200 }, { "epoch": 0.015752351097178683, "grad_norm": 4.687671184539795, "learning_rate": 9.999903001902504e-06, "loss": 1.349, "step": 201 }, { "epoch": 0.015830721003134797, "grad_norm": 4.670878887176514, "learning_rate": 9.999901071645779e-06, "loss": 1.3217, "step": 202 }, { "epoch": 0.015909090909090907, "grad_norm": 3.8080220222473145, "learning_rate": 9.999899122371996e-06, "loss": 1.5172, "step": 203 }, { "epoch": 0.01598746081504702, "grad_norm": 8.507940292358398, "learning_rate": 9.999897154081169e-06, "loss": 2.3744, "step": 204 }, { "epoch": 0.016065830721003135, "grad_norm": 11.708200454711914, "learning_rate": 9.999895166773298e-06, "loss": 3.0678, "step": 205 }, { "epoch": 0.01614420062695925, "grad_norm": 2.711806297302246, "learning_rate": 9.999893160448396e-06, "loss": 1.6099, "step": 206 }, { "epoch": 0.01622257053291536, "grad_norm": 3.090500831604004, "learning_rate": 9.999891135106466e-06, "loss": 1.4628, "step": 207 }, { "epoch": 0.016300940438871474, "grad_norm": 8.678728103637695, "learning_rate": 9.999889090747519e-06, "loss": 2.6507, "step": 208 }, { "epoch": 0.016379310344827588, "grad_norm": 9.396039962768555, "learning_rate": 9.999887027371561e-06, "loss": 1.6558, "step": 209 }, { "epoch": 0.016457680250783698, "grad_norm": 7.16779088973999, "learning_rate": 9.999884944978604e-06, "loss": 1.2495, "step": 210 }, { "epoch": 0.016536050156739812, "grad_norm": 3.619818687438965, "learning_rate": 9.999882843568652e-06, "loss": 1.472, "step": 211 }, { "epoch": 0.016614420062695926, "grad_norm": 9.276970863342285, "learning_rate": 9.999880723141713e-06, "loss": 2.3742, "step": 212 }, { "epoch": 0.016692789968652037, "grad_norm": 3.4609856605529785, "learning_rate": 9.999878583697796e-06, "loss": 1.8788, "step": 213 }, { "epoch": 0.01677115987460815, "grad_norm": 6.4444780349731445, "learning_rate": 9.999876425236908e-06, "loss": 1.6415, "step": 214 }, { "epoch": 0.016849529780564265, "grad_norm": 2.784010887145996, "learning_rate": 9.999874247759058e-06, "loss": 1.6028, "step": 215 }, { "epoch": 0.016927899686520375, "grad_norm": 3.8186957836151123, "learning_rate": 9.999872051264256e-06, "loss": 1.4484, "step": 216 }, { "epoch": 0.01700626959247649, "grad_norm": 5.031325817108154, "learning_rate": 9.999869835752509e-06, "loss": 1.6286, "step": 217 }, { "epoch": 0.017084639498432603, "grad_norm": 3.44602108001709, "learning_rate": 9.999867601223822e-06, "loss": 1.2669, "step": 218 }, { "epoch": 0.017163009404388713, "grad_norm": 3.784182071685791, "learning_rate": 9.99986534767821e-06, "loss": 1.6107, "step": 219 }, { "epoch": 0.017241379310344827, "grad_norm": 4.527110576629639, "learning_rate": 9.999863075115677e-06, "loss": 1.4689, "step": 220 }, { "epoch": 0.01731974921630094, "grad_norm": 4.479491233825684, "learning_rate": 9.999860783536235e-06, "loss": 1.4936, "step": 221 }, { "epoch": 0.017398119122257052, "grad_norm": 3.199547052383423, "learning_rate": 9.999858472939888e-06, "loss": 1.5017, "step": 222 }, { "epoch": 0.017476489028213166, "grad_norm": 4.541609287261963, "learning_rate": 9.99985614332665e-06, "loss": 1.6925, "step": 223 }, { "epoch": 0.01755485893416928, "grad_norm": 11.585817337036133, "learning_rate": 9.999853794696525e-06, "loss": 2.5602, "step": 224 }, { "epoch": 0.01763322884012539, "grad_norm": 5.122155666351318, "learning_rate": 9.999851427049526e-06, "loss": 1.8001, "step": 225 }, { "epoch": 0.017711598746081504, "grad_norm": 4.302443504333496, "learning_rate": 9.999849040385658e-06, "loss": 1.5509, "step": 226 }, { "epoch": 0.017789968652037618, "grad_norm": 4.535111904144287, "learning_rate": 9.999846634704935e-06, "loss": 2.3529, "step": 227 }, { "epoch": 0.017868338557993732, "grad_norm": 7.2217116355896, "learning_rate": 9.999844210007361e-06, "loss": 1.8636, "step": 228 }, { "epoch": 0.017946708463949843, "grad_norm": 4.2245774269104, "learning_rate": 9.999841766292948e-06, "loss": 2.107, "step": 229 }, { "epoch": 0.018025078369905956, "grad_norm": 4.140704154968262, "learning_rate": 9.999839303561708e-06, "loss": 1.3263, "step": 230 }, { "epoch": 0.01810344827586207, "grad_norm": 2.6156952381134033, "learning_rate": 9.999836821813645e-06, "loss": 1.3204, "step": 231 }, { "epoch": 0.01818181818181818, "grad_norm": 3.2109546661376953, "learning_rate": 9.99983432104877e-06, "loss": 1.5273, "step": 232 }, { "epoch": 0.018260188087774295, "grad_norm": 2.3640003204345703, "learning_rate": 9.999831801267095e-06, "loss": 1.2872, "step": 233 }, { "epoch": 0.01833855799373041, "grad_norm": 3.754831552505493, "learning_rate": 9.999829262468627e-06, "loss": 1.3938, "step": 234 }, { "epoch": 0.01841692789968652, "grad_norm": 4.027071475982666, "learning_rate": 9.999826704653376e-06, "loss": 1.4211, "step": 235 }, { "epoch": 0.018495297805642633, "grad_norm": 11.884988784790039, "learning_rate": 9.999824127821354e-06, "loss": 2.4283, "step": 236 }, { "epoch": 0.018573667711598747, "grad_norm": 4.661321640014648, "learning_rate": 9.999821531972567e-06, "loss": 1.7751, "step": 237 }, { "epoch": 0.018652037617554858, "grad_norm": 9.281399726867676, "learning_rate": 9.999818917107029e-06, "loss": 1.3642, "step": 238 }, { "epoch": 0.01873040752351097, "grad_norm": 18.66299057006836, "learning_rate": 9.999816283224747e-06, "loss": 1.0698, "step": 239 }, { "epoch": 0.018808777429467086, "grad_norm": 3.1245715618133545, "learning_rate": 9.999813630325732e-06, "loss": 1.6592, "step": 240 }, { "epoch": 0.018887147335423196, "grad_norm": 6.348033428192139, "learning_rate": 9.999810958409995e-06, "loss": 1.4325, "step": 241 }, { "epoch": 0.01896551724137931, "grad_norm": 3.7534945011138916, "learning_rate": 9.999808267477544e-06, "loss": 1.7834, "step": 242 }, { "epoch": 0.019043887147335424, "grad_norm": 6.385669708251953, "learning_rate": 9.999805557528391e-06, "loss": 1.5649, "step": 243 }, { "epoch": 0.019122257053291535, "grad_norm": 3.9325568675994873, "learning_rate": 9.999802828562544e-06, "loss": 1.3224, "step": 244 }, { "epoch": 0.01920062695924765, "grad_norm": 7.395870685577393, "learning_rate": 9.999800080580017e-06, "loss": 1.8993, "step": 245 }, { "epoch": 0.019278996865203762, "grad_norm": 2.9509482383728027, "learning_rate": 9.999797313580818e-06, "loss": 1.5549, "step": 246 }, { "epoch": 0.019357366771159873, "grad_norm": 3.614684820175171, "learning_rate": 9.999794527564958e-06, "loss": 1.6557, "step": 247 }, { "epoch": 0.019435736677115987, "grad_norm": 4.4793267250061035, "learning_rate": 9.999791722532449e-06, "loss": 1.7388, "step": 248 }, { "epoch": 0.0195141065830721, "grad_norm": 4.869296073913574, "learning_rate": 9.999788898483298e-06, "loss": 1.5144, "step": 249 }, { "epoch": 0.019592476489028215, "grad_norm": 3.8741519451141357, "learning_rate": 9.999786055417519e-06, "loss": 1.9673, "step": 250 }, { "epoch": 0.019670846394984325, "grad_norm": 5.62731409072876, "learning_rate": 9.999783193335122e-06, "loss": 1.676, "step": 251 }, { "epoch": 0.01974921630094044, "grad_norm": 5.110932350158691, "learning_rate": 9.999780312236118e-06, "loss": 1.1761, "step": 252 }, { "epoch": 0.019827586206896553, "grad_norm": 6.495049476623535, "learning_rate": 9.999777412120517e-06, "loss": 1.9416, "step": 253 }, { "epoch": 0.019905956112852664, "grad_norm": 8.295595169067383, "learning_rate": 9.999774492988333e-06, "loss": 1.1063, "step": 254 }, { "epoch": 0.019984326018808778, "grad_norm": 4.20753288269043, "learning_rate": 9.999771554839572e-06, "loss": 1.1205, "step": 255 }, { "epoch": 0.02006269592476489, "grad_norm": 4.690052032470703, "learning_rate": 9.999768597674248e-06, "loss": 1.4645, "step": 256 }, { "epoch": 0.020141065830721002, "grad_norm": 6.071331977844238, "learning_rate": 9.999765621492374e-06, "loss": 1.3986, "step": 257 }, { "epoch": 0.020219435736677116, "grad_norm": 7.029843807220459, "learning_rate": 9.99976262629396e-06, "loss": 1.313, "step": 258 }, { "epoch": 0.02029780564263323, "grad_norm": 14.867573738098145, "learning_rate": 9.999759612079016e-06, "loss": 1.8718, "step": 259 }, { "epoch": 0.02037617554858934, "grad_norm": 5.976381301879883, "learning_rate": 9.999756578847554e-06, "loss": 1.2865, "step": 260 }, { "epoch": 0.020454545454545454, "grad_norm": 4.695678234100342, "learning_rate": 9.999753526599587e-06, "loss": 1.0509, "step": 261 }, { "epoch": 0.02053291536050157, "grad_norm": 4.324662208557129, "learning_rate": 9.999750455335126e-06, "loss": 1.5005, "step": 262 }, { "epoch": 0.02061128526645768, "grad_norm": 2.9913463592529297, "learning_rate": 9.999747365054182e-06, "loss": 1.1329, "step": 263 }, { "epoch": 0.020689655172413793, "grad_norm": 4.554172992706299, "learning_rate": 9.999744255756767e-06, "loss": 1.7773, "step": 264 }, { "epoch": 0.020768025078369907, "grad_norm": 14.82787036895752, "learning_rate": 9.999741127442892e-06, "loss": 1.9262, "step": 265 }, { "epoch": 0.020846394984326017, "grad_norm": 4.352891445159912, "learning_rate": 9.99973798011257e-06, "loss": 1.7219, "step": 266 }, { "epoch": 0.02092476489028213, "grad_norm": 8.423376083374023, "learning_rate": 9.999734813765813e-06, "loss": 1.8029, "step": 267 }, { "epoch": 0.021003134796238245, "grad_norm": 5.158044338226318, "learning_rate": 9.999731628402634e-06, "loss": 1.3717, "step": 268 }, { "epoch": 0.021081504702194356, "grad_norm": 12.061034202575684, "learning_rate": 9.999728424023043e-06, "loss": 1.8013, "step": 269 }, { "epoch": 0.02115987460815047, "grad_norm": 9.014357566833496, "learning_rate": 9.999725200627054e-06, "loss": 1.7209, "step": 270 }, { "epoch": 0.021238244514106584, "grad_norm": 5.041487216949463, "learning_rate": 9.999721958214676e-06, "loss": 1.7899, "step": 271 }, { "epoch": 0.021316614420062698, "grad_norm": 5.521505832672119, "learning_rate": 9.999718696785926e-06, "loss": 1.4135, "step": 272 }, { "epoch": 0.021394984326018808, "grad_norm": 4.162848949432373, "learning_rate": 9.999715416340812e-06, "loss": 1.716, "step": 273 }, { "epoch": 0.021473354231974922, "grad_norm": 5.177021503448486, "learning_rate": 9.99971211687935e-06, "loss": 1.7025, "step": 274 }, { "epoch": 0.021551724137931036, "grad_norm": 3.3822667598724365, "learning_rate": 9.999708798401549e-06, "loss": 1.4851, "step": 275 }, { "epoch": 0.021630094043887146, "grad_norm": 4.0692458152771, "learning_rate": 9.999705460907426e-06, "loss": 1.6627, "step": 276 }, { "epoch": 0.02170846394984326, "grad_norm": 4.147398471832275, "learning_rate": 9.999702104396991e-06, "loss": 1.7239, "step": 277 }, { "epoch": 0.021786833855799374, "grad_norm": 8.654167175292969, "learning_rate": 9.999698728870257e-06, "loss": 2.186, "step": 278 }, { "epoch": 0.021865203761755485, "grad_norm": 4.748649597167969, "learning_rate": 9.999695334327238e-06, "loss": 1.4476, "step": 279 }, { "epoch": 0.0219435736677116, "grad_norm": 7.759385108947754, "learning_rate": 9.999691920767945e-06, "loss": 1.4495, "step": 280 }, { "epoch": 0.022021943573667713, "grad_norm": 46.15164566040039, "learning_rate": 9.999688488192392e-06, "loss": 1.7367, "step": 281 }, { "epoch": 0.022100313479623823, "grad_norm": 4.146672248840332, "learning_rate": 9.999685036600591e-06, "loss": 1.8991, "step": 282 }, { "epoch": 0.022178683385579937, "grad_norm": 4.503333568572998, "learning_rate": 9.999681565992558e-06, "loss": 1.2665, "step": 283 }, { "epoch": 0.02225705329153605, "grad_norm": 4.726100444793701, "learning_rate": 9.999678076368302e-06, "loss": 1.3677, "step": 284 }, { "epoch": 0.02233542319749216, "grad_norm": 3.3651621341705322, "learning_rate": 9.999674567727842e-06, "loss": 1.3728, "step": 285 }, { "epoch": 0.022413793103448276, "grad_norm": 12.442410469055176, "learning_rate": 9.999671040071186e-06, "loss": 2.1138, "step": 286 }, { "epoch": 0.02249216300940439, "grad_norm": 7.02618408203125, "learning_rate": 9.999667493398348e-06, "loss": 2.1476, "step": 287 }, { "epoch": 0.0225705329153605, "grad_norm": 3.9460346698760986, "learning_rate": 9.999663927709345e-06, "loss": 1.544, "step": 288 }, { "epoch": 0.022648902821316614, "grad_norm": 4.950735569000244, "learning_rate": 9.999660343004187e-06, "loss": 1.2819, "step": 289 }, { "epoch": 0.022727272727272728, "grad_norm": 3.213573455810547, "learning_rate": 9.99965673928289e-06, "loss": 1.4916, "step": 290 }, { "epoch": 0.022805642633228842, "grad_norm": 3.3482704162597656, "learning_rate": 9.999653116545467e-06, "loss": 1.1632, "step": 291 }, { "epoch": 0.022884012539184952, "grad_norm": 5.800450801849365, "learning_rate": 9.99964947479193e-06, "loss": 2.2229, "step": 292 }, { "epoch": 0.022962382445141066, "grad_norm": 7.11964750289917, "learning_rate": 9.999645814022296e-06, "loss": 1.6962, "step": 293 }, { "epoch": 0.02304075235109718, "grad_norm": 4.511322498321533, "learning_rate": 9.999642134236577e-06, "loss": 1.8818, "step": 294 }, { "epoch": 0.02311912225705329, "grad_norm": 7.8351969718933105, "learning_rate": 9.999638435434787e-06, "loss": 0.9946, "step": 295 }, { "epoch": 0.023197492163009405, "grad_norm": 4.417308330535889, "learning_rate": 9.99963471761694e-06, "loss": 1.3289, "step": 296 }, { "epoch": 0.02327586206896552, "grad_norm": 5.865767955780029, "learning_rate": 9.99963098078305e-06, "loss": 1.8325, "step": 297 }, { "epoch": 0.02335423197492163, "grad_norm": 3.481109380722046, "learning_rate": 9.999627224933133e-06, "loss": 1.6831, "step": 298 }, { "epoch": 0.023432601880877743, "grad_norm": 5.764068126678467, "learning_rate": 9.999623450067202e-06, "loss": 1.811, "step": 299 }, { "epoch": 0.023510971786833857, "grad_norm": 3.7498934268951416, "learning_rate": 9.99961965618527e-06, "loss": 1.6086, "step": 300 }, { "epoch": 0.023589341692789968, "grad_norm": 4.913384914398193, "learning_rate": 9.999615843287355e-06, "loss": 1.5557, "step": 301 }, { "epoch": 0.02366771159874608, "grad_norm": 5.283233165740967, "learning_rate": 9.999612011373467e-06, "loss": 1.3393, "step": 302 }, { "epoch": 0.023746081504702195, "grad_norm": 4.168469429016113, "learning_rate": 9.999608160443624e-06, "loss": 1.4647, "step": 303 }, { "epoch": 0.023824451410658306, "grad_norm": 2.735356092453003, "learning_rate": 9.99960429049784e-06, "loss": 1.5776, "step": 304 }, { "epoch": 0.02390282131661442, "grad_norm": 6.41701602935791, "learning_rate": 9.99960040153613e-06, "loss": 1.2946, "step": 305 }, { "epoch": 0.023981191222570534, "grad_norm": 5.0400261878967285, "learning_rate": 9.999596493558506e-06, "loss": 1.6447, "step": 306 }, { "epoch": 0.024059561128526644, "grad_norm": 4.565143585205078, "learning_rate": 9.999592566564985e-06, "loss": 1.6538, "step": 307 }, { "epoch": 0.02413793103448276, "grad_norm": 4.682318210601807, "learning_rate": 9.999588620555582e-06, "loss": 1.5055, "step": 308 }, { "epoch": 0.024216300940438872, "grad_norm": 7.270197868347168, "learning_rate": 9.999584655530312e-06, "loss": 1.2827, "step": 309 }, { "epoch": 0.024294670846394983, "grad_norm": 3.64296293258667, "learning_rate": 9.999580671489191e-06, "loss": 1.5758, "step": 310 }, { "epoch": 0.024373040752351097, "grad_norm": 8.71967887878418, "learning_rate": 9.999576668432231e-06, "loss": 1.0128, "step": 311 }, { "epoch": 0.02445141065830721, "grad_norm": 3.4633703231811523, "learning_rate": 9.999572646359453e-06, "loss": 1.5332, "step": 312 }, { "epoch": 0.024529780564263325, "grad_norm": 3.3632795810699463, "learning_rate": 9.999568605270864e-06, "loss": 1.4568, "step": 313 }, { "epoch": 0.024608150470219435, "grad_norm": 4.397461891174316, "learning_rate": 9.999564545166486e-06, "loss": 1.625, "step": 314 }, { "epoch": 0.02468652037617555, "grad_norm": 5.416681289672852, "learning_rate": 9.999560466046332e-06, "loss": 1.1396, "step": 315 }, { "epoch": 0.024764890282131663, "grad_norm": 6.783762454986572, "learning_rate": 9.999556367910419e-06, "loss": 2.2871, "step": 316 }, { "epoch": 0.024843260188087773, "grad_norm": 3.330080032348633, "learning_rate": 9.999552250758761e-06, "loss": 1.7052, "step": 317 }, { "epoch": 0.024921630094043887, "grad_norm": 5.047704219818115, "learning_rate": 9.999548114591375e-06, "loss": 2.407, "step": 318 }, { "epoch": 0.025, "grad_norm": 5.002371311187744, "learning_rate": 9.999543959408274e-06, "loss": 1.2494, "step": 319 }, { "epoch": 0.025078369905956112, "grad_norm": 3.978026866912842, "learning_rate": 9.999539785209478e-06, "loss": 1.4891, "step": 320 }, { "epoch": 0.025156739811912226, "grad_norm": 4.040696620941162, "learning_rate": 9.999535591994998e-06, "loss": 1.6028, "step": 321 }, { "epoch": 0.02523510971786834, "grad_norm": 4.168571472167969, "learning_rate": 9.999531379764855e-06, "loss": 1.6864, "step": 322 }, { "epoch": 0.02531347962382445, "grad_norm": 3.1984856128692627, "learning_rate": 9.999527148519062e-06, "loss": 1.7217, "step": 323 }, { "epoch": 0.025391849529780564, "grad_norm": 2.95173978805542, "learning_rate": 9.999522898257634e-06, "loss": 1.5296, "step": 324 }, { "epoch": 0.025470219435736678, "grad_norm": 7.427578926086426, "learning_rate": 9.999518628980589e-06, "loss": 1.2201, "step": 325 }, { "epoch": 0.02554858934169279, "grad_norm": 8.229640007019043, "learning_rate": 9.999514340687945e-06, "loss": 1.9587, "step": 326 }, { "epoch": 0.025626959247648903, "grad_norm": 2.7177951335906982, "learning_rate": 9.999510033379714e-06, "loss": 1.5214, "step": 327 }, { "epoch": 0.025705329153605017, "grad_norm": 4.043224334716797, "learning_rate": 9.999505707055916e-06, "loss": 1.5227, "step": 328 }, { "epoch": 0.025783699059561127, "grad_norm": 4.021001815795898, "learning_rate": 9.999501361716568e-06, "loss": 1.5248, "step": 329 }, { "epoch": 0.02586206896551724, "grad_norm": 3.123143434524536, "learning_rate": 9.999496997361682e-06, "loss": 1.6984, "step": 330 }, { "epoch": 0.025940438871473355, "grad_norm": 2.9160823822021484, "learning_rate": 9.999492613991278e-06, "loss": 1.6452, "step": 331 }, { "epoch": 0.026018808777429465, "grad_norm": 3.979526996612549, "learning_rate": 9.999488211605373e-06, "loss": 1.4544, "step": 332 }, { "epoch": 0.02609717868338558, "grad_norm": 6.180127143859863, "learning_rate": 9.99948379020398e-06, "loss": 1.9064, "step": 333 }, { "epoch": 0.026175548589341693, "grad_norm": 3.7975432872772217, "learning_rate": 9.999479349787122e-06, "loss": 1.5911, "step": 334 }, { "epoch": 0.026253918495297807, "grad_norm": 3.2563462257385254, "learning_rate": 9.999474890354812e-06, "loss": 1.2168, "step": 335 }, { "epoch": 0.026332288401253918, "grad_norm": 3.3558497428894043, "learning_rate": 9.999470411907065e-06, "loss": 1.6826, "step": 336 }, { "epoch": 0.026410658307210032, "grad_norm": 2.990792751312256, "learning_rate": 9.999465914443902e-06, "loss": 1.4032, "step": 337 }, { "epoch": 0.026489028213166146, "grad_norm": 3.956791400909424, "learning_rate": 9.999461397965338e-06, "loss": 1.2471, "step": 338 }, { "epoch": 0.026567398119122256, "grad_norm": 12.50356674194336, "learning_rate": 9.999456862471391e-06, "loss": 1.6456, "step": 339 }, { "epoch": 0.02664576802507837, "grad_norm": 5.5708818435668945, "learning_rate": 9.999452307962079e-06, "loss": 1.6391, "step": 340 }, { "epoch": 0.026724137931034484, "grad_norm": 3.7792909145355225, "learning_rate": 9.999447734437418e-06, "loss": 1.4325, "step": 341 }, { "epoch": 0.026802507836990595, "grad_norm": 7.851444721221924, "learning_rate": 9.999443141897425e-06, "loss": 1.7917, "step": 342 }, { "epoch": 0.02688087774294671, "grad_norm": 2.742727518081665, "learning_rate": 9.999438530342118e-06, "loss": 1.3257, "step": 343 }, { "epoch": 0.026959247648902823, "grad_norm": 4.921871662139893, "learning_rate": 9.999433899771514e-06, "loss": 1.9316, "step": 344 }, { "epoch": 0.027037617554858933, "grad_norm": 3.074446201324463, "learning_rate": 9.999429250185632e-06, "loss": 1.4256, "step": 345 }, { "epoch": 0.027115987460815047, "grad_norm": 3.499037265777588, "learning_rate": 9.999424581584489e-06, "loss": 1.3432, "step": 346 }, { "epoch": 0.02719435736677116, "grad_norm": 13.086008071899414, "learning_rate": 9.999419893968105e-06, "loss": 1.6913, "step": 347 }, { "epoch": 0.02727272727272727, "grad_norm": 4.327911853790283, "learning_rate": 9.999415187336492e-06, "loss": 1.1277, "step": 348 }, { "epoch": 0.027351097178683385, "grad_norm": 4.0467963218688965, "learning_rate": 9.999410461689674e-06, "loss": 1.7468, "step": 349 }, { "epoch": 0.0274294670846395, "grad_norm": 5.53226900100708, "learning_rate": 9.999405717027666e-06, "loss": 1.8311, "step": 350 }, { "epoch": 0.02750783699059561, "grad_norm": 4.461264133453369, "learning_rate": 9.999400953350486e-06, "loss": 1.6548, "step": 351 }, { "epoch": 0.027586206896551724, "grad_norm": 5.602694034576416, "learning_rate": 9.999396170658152e-06, "loss": 2.0171, "step": 352 }, { "epoch": 0.027664576802507838, "grad_norm": 8.386537551879883, "learning_rate": 9.999391368950684e-06, "loss": 1.6682, "step": 353 }, { "epoch": 0.027742946708463948, "grad_norm": 3.38686203956604, "learning_rate": 9.999386548228099e-06, "loss": 1.2669, "step": 354 }, { "epoch": 0.027821316614420062, "grad_norm": 2.9905190467834473, "learning_rate": 9.999381708490415e-06, "loss": 1.6263, "step": 355 }, { "epoch": 0.027899686520376176, "grad_norm": 10.86367416381836, "learning_rate": 9.99937684973765e-06, "loss": 1.1596, "step": 356 }, { "epoch": 0.02797805642633229, "grad_norm": 3.9448513984680176, "learning_rate": 9.999371971969824e-06, "loss": 1.8707, "step": 357 }, { "epoch": 0.0280564263322884, "grad_norm": 8.249129295349121, "learning_rate": 9.999367075186955e-06, "loss": 1.6886, "step": 358 }, { "epoch": 0.028134796238244515, "grad_norm": 9.335741996765137, "learning_rate": 9.999362159389062e-06, "loss": 2.2865, "step": 359 }, { "epoch": 0.02821316614420063, "grad_norm": 3.7767581939697266, "learning_rate": 9.999357224576163e-06, "loss": 1.5599, "step": 360 }, { "epoch": 0.02829153605015674, "grad_norm": 8.07192611694336, "learning_rate": 9.999352270748278e-06, "loss": 1.3643, "step": 361 }, { "epoch": 0.028369905956112853, "grad_norm": 3.942678928375244, "learning_rate": 9.999347297905422e-06, "loss": 1.4192, "step": 362 }, { "epoch": 0.028448275862068967, "grad_norm": 3.563194513320923, "learning_rate": 9.999342306047618e-06, "loss": 1.7749, "step": 363 }, { "epoch": 0.028526645768025077, "grad_norm": 3.325456380844116, "learning_rate": 9.999337295174884e-06, "loss": 1.4499, "step": 364 }, { "epoch": 0.02860501567398119, "grad_norm": 3.293801784515381, "learning_rate": 9.999332265287238e-06, "loss": 1.306, "step": 365 }, { "epoch": 0.028683385579937305, "grad_norm": 4.384602069854736, "learning_rate": 9.9993272163847e-06, "loss": 1.4823, "step": 366 }, { "epoch": 0.028761755485893416, "grad_norm": 7.567427158355713, "learning_rate": 9.999322148467289e-06, "loss": 1.617, "step": 367 }, { "epoch": 0.02884012539184953, "grad_norm": 3.963580369949341, "learning_rate": 9.999317061535025e-06, "loss": 1.4834, "step": 368 }, { "epoch": 0.028918495297805644, "grad_norm": 4.242827415466309, "learning_rate": 9.999311955587927e-06, "loss": 1.2836, "step": 369 }, { "epoch": 0.028996865203761754, "grad_norm": 5.55364465713501, "learning_rate": 9.999306830626015e-06, "loss": 1.8078, "step": 370 }, { "epoch": 0.029075235109717868, "grad_norm": 5.6869916915893555, "learning_rate": 9.999301686649305e-06, "loss": 1.5479, "step": 371 }, { "epoch": 0.029153605015673982, "grad_norm": 3.845174551010132, "learning_rate": 9.999296523657821e-06, "loss": 1.6036, "step": 372 }, { "epoch": 0.029231974921630093, "grad_norm": 3.7438271045684814, "learning_rate": 9.999291341651581e-06, "loss": 1.3242, "step": 373 }, { "epoch": 0.029310344827586206, "grad_norm": 3.8681020736694336, "learning_rate": 9.999286140630604e-06, "loss": 1.6526, "step": 374 }, { "epoch": 0.02938871473354232, "grad_norm": 4.768889904022217, "learning_rate": 9.999280920594909e-06, "loss": 1.6265, "step": 375 }, { "epoch": 0.02946708463949843, "grad_norm": 6.085917949676514, "learning_rate": 9.99927568154452e-06, "loss": 1.5004, "step": 376 }, { "epoch": 0.029545454545454545, "grad_norm": 4.202488422393799, "learning_rate": 9.999270423479452e-06, "loss": 1.5, "step": 377 }, { "epoch": 0.02962382445141066, "grad_norm": 6.375894069671631, "learning_rate": 9.999265146399728e-06, "loss": 1.6515, "step": 378 }, { "epoch": 0.029702194357366773, "grad_norm": 5.684266090393066, "learning_rate": 9.999259850305366e-06, "loss": 1.921, "step": 379 }, { "epoch": 0.029780564263322883, "grad_norm": 5.179361343383789, "learning_rate": 9.999254535196387e-06, "loss": 1.7175, "step": 380 }, { "epoch": 0.029858934169278997, "grad_norm": 3.503124952316284, "learning_rate": 9.999249201072814e-06, "loss": 1.6001, "step": 381 }, { "epoch": 0.02993730407523511, "grad_norm": 4.641046047210693, "learning_rate": 9.999243847934663e-06, "loss": 1.8632, "step": 382 }, { "epoch": 0.03001567398119122, "grad_norm": 3.7010653018951416, "learning_rate": 9.999238475781957e-06, "loss": 1.5703, "step": 383 }, { "epoch": 0.030094043887147336, "grad_norm": 4.1754913330078125, "learning_rate": 9.999233084614715e-06, "loss": 1.3957, "step": 384 }, { "epoch": 0.03017241379310345, "grad_norm": 5.243168354034424, "learning_rate": 9.999227674432958e-06, "loss": 0.9345, "step": 385 }, { "epoch": 0.03025078369905956, "grad_norm": 3.774338960647583, "learning_rate": 9.999222245236708e-06, "loss": 1.4423, "step": 386 }, { "epoch": 0.030329153605015674, "grad_norm": 5.394372940063477, "learning_rate": 9.999216797025983e-06, "loss": 1.546, "step": 387 }, { "epoch": 0.030407523510971788, "grad_norm": 2.55607271194458, "learning_rate": 9.999211329800805e-06, "loss": 1.3631, "step": 388 }, { "epoch": 0.0304858934169279, "grad_norm": 4.705481052398682, "learning_rate": 9.999205843561196e-06, "loss": 1.6134, "step": 389 }, { "epoch": 0.030564263322884012, "grad_norm": 3.659888744354248, "learning_rate": 9.999200338307174e-06, "loss": 1.6005, "step": 390 }, { "epoch": 0.030642633228840126, "grad_norm": 5.19370698928833, "learning_rate": 9.999194814038764e-06, "loss": 1.2123, "step": 391 }, { "epoch": 0.030721003134796237, "grad_norm": 2.6598458290100098, "learning_rate": 9.999189270755983e-06, "loss": 1.2979, "step": 392 }, { "epoch": 0.03079937304075235, "grad_norm": 3.2173216342926025, "learning_rate": 9.999183708458854e-06, "loss": 1.3804, "step": 393 }, { "epoch": 0.030877742946708465, "grad_norm": 3.0825135707855225, "learning_rate": 9.999178127147398e-06, "loss": 1.5608, "step": 394 }, { "epoch": 0.030956112852664575, "grad_norm": 6.82411527633667, "learning_rate": 9.999172526821636e-06, "loss": 1.5975, "step": 395 }, { "epoch": 0.03103448275862069, "grad_norm": 3.559727430343628, "learning_rate": 9.99916690748159e-06, "loss": 1.4466, "step": 396 }, { "epoch": 0.031112852664576803, "grad_norm": 5.359574317932129, "learning_rate": 9.99916126912728e-06, "loss": 1.4269, "step": 397 }, { "epoch": 0.031191222570532914, "grad_norm": 7.388969898223877, "learning_rate": 9.999155611758729e-06, "loss": 1.4269, "step": 398 }, { "epoch": 0.03126959247648903, "grad_norm": 36.41108322143555, "learning_rate": 9.999149935375957e-06, "loss": 1.7026, "step": 399 }, { "epoch": 0.03134796238244514, "grad_norm": 4.060421466827393, "learning_rate": 9.999144239978987e-06, "loss": 1.2211, "step": 400 }, { "epoch": 0.031426332288401256, "grad_norm": 11.240849494934082, "learning_rate": 9.99913852556784e-06, "loss": 0.9835, "step": 401 }, { "epoch": 0.031504702194357366, "grad_norm": 2.946455478668213, "learning_rate": 9.999132792142537e-06, "loss": 1.4429, "step": 402 }, { "epoch": 0.031583072100313477, "grad_norm": 4.35193395614624, "learning_rate": 9.999127039703102e-06, "loss": 1.6879, "step": 403 }, { "epoch": 0.031661442006269594, "grad_norm": 6.126370429992676, "learning_rate": 9.999121268249554e-06, "loss": 1.291, "step": 404 }, { "epoch": 0.031739811912225704, "grad_norm": 6.408405303955078, "learning_rate": 9.999115477781917e-06, "loss": 1.4373, "step": 405 }, { "epoch": 0.031818181818181815, "grad_norm": 3.3850157260894775, "learning_rate": 9.999109668300213e-06, "loss": 1.4844, "step": 406 }, { "epoch": 0.03189655172413793, "grad_norm": 4.752655506134033, "learning_rate": 9.999103839804463e-06, "loss": 1.6332, "step": 407 }, { "epoch": 0.03197492163009404, "grad_norm": 3.959745407104492, "learning_rate": 9.99909799229469e-06, "loss": 1.5746, "step": 408 }, { "epoch": 0.03205329153605015, "grad_norm": 6.02347993850708, "learning_rate": 9.999092125770915e-06, "loss": 1.3847, "step": 409 }, { "epoch": 0.03213166144200627, "grad_norm": 4.508439064025879, "learning_rate": 9.999086240233163e-06, "loss": 1.3577, "step": 410 }, { "epoch": 0.03221003134796238, "grad_norm": 9.763604164123535, "learning_rate": 9.999080335681453e-06, "loss": 1.4165, "step": 411 }, { "epoch": 0.0322884012539185, "grad_norm": 4.911927700042725, "learning_rate": 9.99907441211581e-06, "loss": 1.0149, "step": 412 }, { "epoch": 0.03236677115987461, "grad_norm": 5.3709869384765625, "learning_rate": 9.999068469536254e-06, "loss": 1.5483, "step": 413 }, { "epoch": 0.03244514106583072, "grad_norm": 4.208844184875488, "learning_rate": 9.999062507942812e-06, "loss": 1.0924, "step": 414 }, { "epoch": 0.03252351097178684, "grad_norm": 2.9676120281219482, "learning_rate": 9.999056527335503e-06, "loss": 1.5388, "step": 415 }, { "epoch": 0.03260188087774295, "grad_norm": 3.134770154953003, "learning_rate": 9.999050527714349e-06, "loss": 1.5818, "step": 416 }, { "epoch": 0.03268025078369906, "grad_norm": 3.1707069873809814, "learning_rate": 9.999044509079375e-06, "loss": 1.2909, "step": 417 }, { "epoch": 0.032758620689655175, "grad_norm": 11.093718528747559, "learning_rate": 9.999038471430604e-06, "loss": 1.5068, "step": 418 }, { "epoch": 0.032836990595611286, "grad_norm": 3.4308958053588867, "learning_rate": 9.999032414768058e-06, "loss": 1.3773, "step": 419 }, { "epoch": 0.032915360501567396, "grad_norm": 6.816355228424072, "learning_rate": 9.999026339091761e-06, "loss": 1.6364, "step": 420 }, { "epoch": 0.032993730407523514, "grad_norm": 4.546479225158691, "learning_rate": 9.999020244401736e-06, "loss": 1.6794, "step": 421 }, { "epoch": 0.033072100313479624, "grad_norm": 4.446110248565674, "learning_rate": 9.999014130698004e-06, "loss": 1.767, "step": 422 }, { "epoch": 0.033150470219435735, "grad_norm": 3.569603443145752, "learning_rate": 9.99900799798059e-06, "loss": 1.5724, "step": 423 }, { "epoch": 0.03322884012539185, "grad_norm": 3.6784863471984863, "learning_rate": 9.99900184624952e-06, "loss": 1.6011, "step": 424 }, { "epoch": 0.03330721003134796, "grad_norm": 2.9470937252044678, "learning_rate": 9.99899567550481e-06, "loss": 1.6321, "step": 425 }, { "epoch": 0.03338557993730407, "grad_norm": 3.6330199241638184, "learning_rate": 9.99898948574649e-06, "loss": 1.2592, "step": 426 }, { "epoch": 0.03346394984326019, "grad_norm": 3.988162040710449, "learning_rate": 9.998983276974584e-06, "loss": 1.3866, "step": 427 }, { "epoch": 0.0335423197492163, "grad_norm": 4.025172710418701, "learning_rate": 9.998977049189112e-06, "loss": 1.4625, "step": 428 }, { "epoch": 0.03362068965517241, "grad_norm": 4.939523696899414, "learning_rate": 9.998970802390098e-06, "loss": 1.3329, "step": 429 }, { "epoch": 0.03369905956112853, "grad_norm": 5.121811389923096, "learning_rate": 9.998964536577566e-06, "loss": 1.5737, "step": 430 }, { "epoch": 0.03377742946708464, "grad_norm": 6.910219192504883, "learning_rate": 9.998958251751542e-06, "loss": 1.2334, "step": 431 }, { "epoch": 0.03385579937304075, "grad_norm": 6.975037097930908, "learning_rate": 9.998951947912048e-06, "loss": 1.1528, "step": 432 }, { "epoch": 0.03393416927899687, "grad_norm": 3.745504379272461, "learning_rate": 9.998945625059108e-06, "loss": 1.5886, "step": 433 }, { "epoch": 0.03401253918495298, "grad_norm": 3.248720645904541, "learning_rate": 9.998939283192749e-06, "loss": 1.5529, "step": 434 }, { "epoch": 0.03409090909090909, "grad_norm": 5.522452354431152, "learning_rate": 9.99893292231299e-06, "loss": 1.7149, "step": 435 }, { "epoch": 0.034169278996865206, "grad_norm": 11.876629829406738, "learning_rate": 9.998926542419856e-06, "loss": 1.0419, "step": 436 }, { "epoch": 0.034247648902821316, "grad_norm": 2.9213428497314453, "learning_rate": 9.998920143513377e-06, "loss": 1.2801, "step": 437 }, { "epoch": 0.03432601880877743, "grad_norm": 4.590330123901367, "learning_rate": 9.998913725593572e-06, "loss": 1.6133, "step": 438 }, { "epoch": 0.034404388714733544, "grad_norm": 3.5599071979522705, "learning_rate": 9.998907288660466e-06, "loss": 1.4166, "step": 439 }, { "epoch": 0.034482758620689655, "grad_norm": 4.595324993133545, "learning_rate": 9.998900832714084e-06, "loss": 1.2343, "step": 440 }, { "epoch": 0.034561128526645765, "grad_norm": 4.795053958892822, "learning_rate": 9.998894357754452e-06, "loss": 1.5119, "step": 441 }, { "epoch": 0.03463949843260188, "grad_norm": 3.6230521202087402, "learning_rate": 9.998887863781592e-06, "loss": 1.6873, "step": 442 }, { "epoch": 0.03471786833855799, "grad_norm": 3.685850143432617, "learning_rate": 9.998881350795532e-06, "loss": 1.5959, "step": 443 }, { "epoch": 0.034796238244514104, "grad_norm": 3.43145751953125, "learning_rate": 9.998874818796293e-06, "loss": 1.4751, "step": 444 }, { "epoch": 0.03487460815047022, "grad_norm": 3.8796658515930176, "learning_rate": 9.998868267783904e-06, "loss": 1.5257, "step": 445 }, { "epoch": 0.03495297805642633, "grad_norm": 2.598773717880249, "learning_rate": 9.998861697758386e-06, "loss": 0.4354, "step": 446 }, { "epoch": 0.03503134796238244, "grad_norm": 3.674429178237915, "learning_rate": 9.998855108719767e-06, "loss": 1.5618, "step": 447 }, { "epoch": 0.03510971786833856, "grad_norm": 4.18018913269043, "learning_rate": 9.998848500668071e-06, "loss": 1.5132, "step": 448 }, { "epoch": 0.03518808777429467, "grad_norm": 3.251185655593872, "learning_rate": 9.998841873603321e-06, "loss": 1.5307, "step": 449 }, { "epoch": 0.03526645768025078, "grad_norm": 3.3593478202819824, "learning_rate": 9.998835227525546e-06, "loss": 1.5303, "step": 450 }, { "epoch": 0.0353448275862069, "grad_norm": 4.038604259490967, "learning_rate": 9.998828562434768e-06, "loss": 1.4894, "step": 451 }, { "epoch": 0.03542319749216301, "grad_norm": 3.5434422492980957, "learning_rate": 9.998821878331014e-06, "loss": 1.5435, "step": 452 }, { "epoch": 0.035501567398119126, "grad_norm": 3.633449077606201, "learning_rate": 9.998815175214309e-06, "loss": 1.635, "step": 453 }, { "epoch": 0.035579937304075236, "grad_norm": 3.0259339809417725, "learning_rate": 9.998808453084678e-06, "loss": 1.1471, "step": 454 }, { "epoch": 0.03565830721003135, "grad_norm": 3.167663335800171, "learning_rate": 9.998801711942147e-06, "loss": 1.6059, "step": 455 }, { "epoch": 0.035736677115987464, "grad_norm": 3.854372024536133, "learning_rate": 9.998794951786745e-06, "loss": 1.2779, "step": 456 }, { "epoch": 0.035815047021943575, "grad_norm": 3.933387517929077, "learning_rate": 9.998788172618491e-06, "loss": 1.3875, "step": 457 }, { "epoch": 0.035893416927899685, "grad_norm": 3.6354541778564453, "learning_rate": 9.998781374437416e-06, "loss": 1.3998, "step": 458 }, { "epoch": 0.0359717868338558, "grad_norm": 3.5010342597961426, "learning_rate": 9.998774557243544e-06, "loss": 1.6066, "step": 459 }, { "epoch": 0.03605015673981191, "grad_norm": 2.4627065658569336, "learning_rate": 9.998767721036901e-06, "loss": 1.6409, "step": 460 }, { "epoch": 0.036128526645768023, "grad_norm": 12.363469123840332, "learning_rate": 9.998760865817514e-06, "loss": 1.3317, "step": 461 }, { "epoch": 0.03620689655172414, "grad_norm": 2.6148276329040527, "learning_rate": 9.998753991585406e-06, "loss": 1.5762, "step": 462 }, { "epoch": 0.03628526645768025, "grad_norm": 13.399910926818848, "learning_rate": 9.998747098340608e-06, "loss": 1.7076, "step": 463 }, { "epoch": 0.03636363636363636, "grad_norm": 7.8125786781311035, "learning_rate": 9.998740186083143e-06, "loss": 3.0239, "step": 464 }, { "epoch": 0.03644200626959248, "grad_norm": 4.7942938804626465, "learning_rate": 9.998733254813035e-06, "loss": 1.5283, "step": 465 }, { "epoch": 0.03652037617554859, "grad_norm": 3.2921600341796875, "learning_rate": 9.998726304530316e-06, "loss": 1.4873, "step": 466 }, { "epoch": 0.0365987460815047, "grad_norm": 6.1833391189575195, "learning_rate": 9.998719335235008e-06, "loss": 1.3757, "step": 467 }, { "epoch": 0.03667711598746082, "grad_norm": 4.53115177154541, "learning_rate": 9.99871234692714e-06, "loss": 1.5315, "step": 468 }, { "epoch": 0.03675548589341693, "grad_norm": 2.9986000061035156, "learning_rate": 9.998705339606739e-06, "loss": 1.6956, "step": 469 }, { "epoch": 0.03683385579937304, "grad_norm": 3.163743495941162, "learning_rate": 9.998698313273829e-06, "loss": 1.319, "step": 470 }, { "epoch": 0.036912225705329156, "grad_norm": 3.041780471801758, "learning_rate": 9.998691267928437e-06, "loss": 1.2725, "step": 471 }, { "epoch": 0.03699059561128527, "grad_norm": 17.14873504638672, "learning_rate": 9.998684203570594e-06, "loss": 1.4327, "step": 472 }, { "epoch": 0.03706896551724138, "grad_norm": 4.070298194885254, "learning_rate": 9.998677120200321e-06, "loss": 1.3367, "step": 473 }, { "epoch": 0.037147335423197494, "grad_norm": 4.100532531738281, "learning_rate": 9.998670017817649e-06, "loss": 1.3749, "step": 474 }, { "epoch": 0.037225705329153605, "grad_norm": 4.197404861450195, "learning_rate": 9.998662896422604e-06, "loss": 1.3902, "step": 475 }, { "epoch": 0.037304075235109715, "grad_norm": 4.082170486450195, "learning_rate": 9.998655756015212e-06, "loss": 1.6279, "step": 476 }, { "epoch": 0.03738244514106583, "grad_norm": 3.163956880569458, "learning_rate": 9.998648596595502e-06, "loss": 1.6769, "step": 477 }, { "epoch": 0.03746081504702194, "grad_norm": 3.801380157470703, "learning_rate": 9.9986414181635e-06, "loss": 1.4371, "step": 478 }, { "epoch": 0.037539184952978054, "grad_norm": 6.500656604766846, "learning_rate": 9.998634220719233e-06, "loss": 1.966, "step": 479 }, { "epoch": 0.03761755485893417, "grad_norm": 3.697519063949585, "learning_rate": 9.998627004262727e-06, "loss": 1.769, "step": 480 }, { "epoch": 0.03769592476489028, "grad_norm": 6.277353763580322, "learning_rate": 9.998619768794013e-06, "loss": 1.7781, "step": 481 }, { "epoch": 0.03777429467084639, "grad_norm": 3.556532382965088, "learning_rate": 9.998612514313117e-06, "loss": 1.6502, "step": 482 }, { "epoch": 0.03785266457680251, "grad_norm": 4.346358776092529, "learning_rate": 9.998605240820067e-06, "loss": 1.4513, "step": 483 }, { "epoch": 0.03793103448275862, "grad_norm": 3.228905200958252, "learning_rate": 9.99859794831489e-06, "loss": 1.5659, "step": 484 }, { "epoch": 0.03800940438871473, "grad_norm": 4.421380519866943, "learning_rate": 9.998590636797613e-06, "loss": 1.7449, "step": 485 }, { "epoch": 0.03808777429467085, "grad_norm": 5.963726997375488, "learning_rate": 9.998583306268265e-06, "loss": 1.2432, "step": 486 }, { "epoch": 0.03816614420062696, "grad_norm": 3.2585608959198, "learning_rate": 9.998575956726874e-06, "loss": 1.2746, "step": 487 }, { "epoch": 0.03824451410658307, "grad_norm": 4.329468727111816, "learning_rate": 9.998568588173467e-06, "loss": 1.4044, "step": 488 }, { "epoch": 0.038322884012539186, "grad_norm": 3.033015489578247, "learning_rate": 9.998561200608072e-06, "loss": 1.4509, "step": 489 }, { "epoch": 0.0384012539184953, "grad_norm": 3.4103827476501465, "learning_rate": 9.998553794030719e-06, "loss": 1.6194, "step": 490 }, { "epoch": 0.03847962382445141, "grad_norm": 3.266843795776367, "learning_rate": 9.998546368441434e-06, "loss": 1.3518, "step": 491 }, { "epoch": 0.038557993730407525, "grad_norm": 4.493947505950928, "learning_rate": 9.998538923840246e-06, "loss": 1.6216, "step": 492 }, { "epoch": 0.038636363636363635, "grad_norm": 3.1908295154571533, "learning_rate": 9.998531460227184e-06, "loss": 1.0531, "step": 493 }, { "epoch": 0.038714733542319746, "grad_norm": 2.633974075317383, "learning_rate": 9.998523977602275e-06, "loss": 1.2037, "step": 494 }, { "epoch": 0.03879310344827586, "grad_norm": 5.2708001136779785, "learning_rate": 9.998516475965548e-06, "loss": 1.5036, "step": 495 }, { "epoch": 0.038871473354231974, "grad_norm": 4.780086040496826, "learning_rate": 9.998508955317032e-06, "loss": 1.5741, "step": 496 }, { "epoch": 0.03894984326018809, "grad_norm": 18.12331199645996, "learning_rate": 9.998501415656755e-06, "loss": 0.6461, "step": 497 }, { "epoch": 0.0390282131661442, "grad_norm": 3.3362653255462646, "learning_rate": 9.998493856984746e-06, "loss": 1.4412, "step": 498 }, { "epoch": 0.03910658307210031, "grad_norm": 9.00538158416748, "learning_rate": 9.998486279301035e-06, "loss": 1.4486, "step": 499 }, { "epoch": 0.03918495297805643, "grad_norm": 6.226262092590332, "learning_rate": 9.998478682605648e-06, "loss": 1.9562, "step": 500 }, { "epoch": 0.03926332288401254, "grad_norm": 8.368409156799316, "learning_rate": 9.998471066898616e-06, "loss": 1.2466, "step": 501 }, { "epoch": 0.03934169278996865, "grad_norm": 3.6902077198028564, "learning_rate": 9.998463432179969e-06, "loss": 1.2005, "step": 502 }, { "epoch": 0.03942006269592477, "grad_norm": 2.6810519695281982, "learning_rate": 9.998455778449733e-06, "loss": 1.3631, "step": 503 }, { "epoch": 0.03949843260188088, "grad_norm": 5.7654924392700195, "learning_rate": 9.998448105707938e-06, "loss": 0.8972, "step": 504 }, { "epoch": 0.03957680250783699, "grad_norm": 5.023662090301514, "learning_rate": 9.998440413954615e-06, "loss": 1.5442, "step": 505 }, { "epoch": 0.039655172413793106, "grad_norm": 10.927292823791504, "learning_rate": 9.998432703189792e-06, "loss": 2.2003, "step": 506 }, { "epoch": 0.03973354231974922, "grad_norm": 4.692617893218994, "learning_rate": 9.998424973413497e-06, "loss": 1.6882, "step": 507 }, { "epoch": 0.03981191222570533, "grad_norm": 7.807013988494873, "learning_rate": 9.998417224625762e-06, "loss": 2.1695, "step": 508 }, { "epoch": 0.039890282131661445, "grad_norm": 4.8385491371154785, "learning_rate": 9.998409456826614e-06, "loss": 1.5982, "step": 509 }, { "epoch": 0.039968652037617555, "grad_norm": 3.2913992404937744, "learning_rate": 9.998401670016085e-06, "loss": 1.6895, "step": 510 }, { "epoch": 0.040047021943573666, "grad_norm": 2.9936323165893555, "learning_rate": 9.998393864194202e-06, "loss": 1.8212, "step": 511 }, { "epoch": 0.04012539184952978, "grad_norm": 3.6836442947387695, "learning_rate": 9.998386039360998e-06, "loss": 1.674, "step": 512 }, { "epoch": 0.040203761755485894, "grad_norm": 3.053448438644409, "learning_rate": 9.9983781955165e-06, "loss": 1.4201, "step": 513 }, { "epoch": 0.040282131661442004, "grad_norm": 4.070194244384766, "learning_rate": 9.998370332660738e-06, "loss": 1.5938, "step": 514 }, { "epoch": 0.04036050156739812, "grad_norm": 4.472990036010742, "learning_rate": 9.998362450793743e-06, "loss": 1.7051, "step": 515 }, { "epoch": 0.04043887147335423, "grad_norm": 3.6984713077545166, "learning_rate": 9.998354549915546e-06, "loss": 1.718, "step": 516 }, { "epoch": 0.04051724137931034, "grad_norm": 3.4712071418762207, "learning_rate": 9.998346630026174e-06, "loss": 1.4342, "step": 517 }, { "epoch": 0.04059561128526646, "grad_norm": 3.1042494773864746, "learning_rate": 9.99833869112566e-06, "loss": 1.3693, "step": 518 }, { "epoch": 0.04067398119122257, "grad_norm": 3.6265275478363037, "learning_rate": 9.998330733214031e-06, "loss": 1.6238, "step": 519 }, { "epoch": 0.04075235109717868, "grad_norm": 4.074193000793457, "learning_rate": 9.998322756291321e-06, "loss": 1.3103, "step": 520 }, { "epoch": 0.0408307210031348, "grad_norm": 4.826519012451172, "learning_rate": 9.998314760357557e-06, "loss": 2.2207, "step": 521 }, { "epoch": 0.04090909090909091, "grad_norm": 2.471278667449951, "learning_rate": 9.998306745412772e-06, "loss": 1.463, "step": 522 }, { "epoch": 0.04098746081504702, "grad_norm": 8.99332332611084, "learning_rate": 9.998298711456994e-06, "loss": 1.7436, "step": 523 }, { "epoch": 0.04106583072100314, "grad_norm": 4.057047367095947, "learning_rate": 9.998290658490257e-06, "loss": 1.2565, "step": 524 }, { "epoch": 0.04114420062695925, "grad_norm": 3.549862861633301, "learning_rate": 9.998282586512589e-06, "loss": 1.332, "step": 525 }, { "epoch": 0.04122257053291536, "grad_norm": 6.548694610595703, "learning_rate": 9.998274495524021e-06, "loss": 1.5565, "step": 526 }, { "epoch": 0.041300940438871475, "grad_norm": 3.904632568359375, "learning_rate": 9.998266385524584e-06, "loss": 0.8218, "step": 527 }, { "epoch": 0.041379310344827586, "grad_norm": 6.9812164306640625, "learning_rate": 9.998258256514308e-06, "loss": 0.8495, "step": 528 }, { "epoch": 0.041457680250783696, "grad_norm": 3.651665687561035, "learning_rate": 9.998250108493227e-06, "loss": 1.0919, "step": 529 }, { "epoch": 0.041536050156739814, "grad_norm": 4.472293853759766, "learning_rate": 9.998241941461369e-06, "loss": 1.602, "step": 530 }, { "epoch": 0.041614420062695924, "grad_norm": 3.668320655822754, "learning_rate": 9.998233755418765e-06, "loss": 0.9689, "step": 531 }, { "epoch": 0.041692789968652035, "grad_norm": 4.1797027587890625, "learning_rate": 9.998225550365448e-06, "loss": 1.5873, "step": 532 }, { "epoch": 0.04177115987460815, "grad_norm": 2.7384445667266846, "learning_rate": 9.998217326301449e-06, "loss": 1.4792, "step": 533 }, { "epoch": 0.04184952978056426, "grad_norm": 5.0417561531066895, "learning_rate": 9.998209083226798e-06, "loss": 1.471, "step": 534 }, { "epoch": 0.04192789968652037, "grad_norm": 2.4272749423980713, "learning_rate": 9.998200821141525e-06, "loss": 1.6336, "step": 535 }, { "epoch": 0.04200626959247649, "grad_norm": 3.288113832473755, "learning_rate": 9.998192540045664e-06, "loss": 1.6944, "step": 536 }, { "epoch": 0.0420846394984326, "grad_norm": 3.986438512802124, "learning_rate": 9.998184239939248e-06, "loss": 1.2475, "step": 537 }, { "epoch": 0.04216300940438871, "grad_norm": 3.323451042175293, "learning_rate": 9.998175920822304e-06, "loss": 1.3075, "step": 538 }, { "epoch": 0.04224137931034483, "grad_norm": 4.8423848152160645, "learning_rate": 9.998167582694867e-06, "loss": 1.5887, "step": 539 }, { "epoch": 0.04231974921630094, "grad_norm": 4.571223735809326, "learning_rate": 9.998159225556967e-06, "loss": 1.5334, "step": 540 }, { "epoch": 0.04239811912225706, "grad_norm": 3.626352310180664, "learning_rate": 9.99815084940864e-06, "loss": 1.6597, "step": 541 }, { "epoch": 0.04247648902821317, "grad_norm": 2.9825305938720703, "learning_rate": 9.99814245424991e-06, "loss": 1.6598, "step": 542 }, { "epoch": 0.04255485893416928, "grad_norm": 8.74788761138916, "learning_rate": 9.998134040080814e-06, "loss": 1.0124, "step": 543 }, { "epoch": 0.042633228840125395, "grad_norm": 2.1957778930664062, "learning_rate": 9.998125606901385e-06, "loss": 1.012, "step": 544 }, { "epoch": 0.042711598746081506, "grad_norm": 3.7527825832366943, "learning_rate": 9.998117154711654e-06, "loss": 1.5924, "step": 545 }, { "epoch": 0.042789968652037616, "grad_norm": 4.309939384460449, "learning_rate": 9.99810868351165e-06, "loss": 1.4602, "step": 546 }, { "epoch": 0.04286833855799373, "grad_norm": 4.934352397918701, "learning_rate": 9.99810019330141e-06, "loss": 1.6341, "step": 547 }, { "epoch": 0.042946708463949844, "grad_norm": 4.314807415008545, "learning_rate": 9.998091684080961e-06, "loss": 1.2699, "step": 548 }, { "epoch": 0.043025078369905954, "grad_norm": 4.640705108642578, "learning_rate": 9.99808315585034e-06, "loss": 1.3758, "step": 549 }, { "epoch": 0.04310344827586207, "grad_norm": 4.242049217224121, "learning_rate": 9.998074608609579e-06, "loss": 1.7413, "step": 550 }, { "epoch": 0.04318181818181818, "grad_norm": 6.263716220855713, "learning_rate": 9.99806604235871e-06, "loss": 1.0385, "step": 551 }, { "epoch": 0.04326018808777429, "grad_norm": 4.043020248413086, "learning_rate": 9.998057457097763e-06, "loss": 1.6614, "step": 552 }, { "epoch": 0.04333855799373041, "grad_norm": 8.352115631103516, "learning_rate": 9.998048852826772e-06, "loss": 1.2893, "step": 553 }, { "epoch": 0.04341692789968652, "grad_norm": 6.562094211578369, "learning_rate": 9.998040229545772e-06, "loss": 1.5674, "step": 554 }, { "epoch": 0.04349529780564263, "grad_norm": 3.672924041748047, "learning_rate": 9.998031587254794e-06, "loss": 1.3588, "step": 555 }, { "epoch": 0.04357366771159875, "grad_norm": 3.698286294937134, "learning_rate": 9.99802292595387e-06, "loss": 1.7285, "step": 556 }, { "epoch": 0.04365203761755486, "grad_norm": 3.734198808670044, "learning_rate": 9.998014245643036e-06, "loss": 1.5259, "step": 557 }, { "epoch": 0.04373040752351097, "grad_norm": 5.057974338531494, "learning_rate": 9.99800554632232e-06, "loss": 1.2394, "step": 558 }, { "epoch": 0.04380877742946709, "grad_norm": 4.077376842498779, "learning_rate": 9.99799682799176e-06, "loss": 0.9823, "step": 559 }, { "epoch": 0.0438871473354232, "grad_norm": 5.265910625457764, "learning_rate": 9.997988090651387e-06, "loss": 1.4037, "step": 560 }, { "epoch": 0.04396551724137931, "grad_norm": 4.001577854156494, "learning_rate": 9.997979334301233e-06, "loss": 1.7078, "step": 561 }, { "epoch": 0.044043887147335425, "grad_norm": 4.892011642456055, "learning_rate": 9.997970558941334e-06, "loss": 0.9195, "step": 562 }, { "epoch": 0.044122257053291536, "grad_norm": 3.0837690830230713, "learning_rate": 9.997961764571721e-06, "loss": 1.3046, "step": 563 }, { "epoch": 0.044200626959247646, "grad_norm": 5.372920513153076, "learning_rate": 9.997952951192431e-06, "loss": 2.0788, "step": 564 }, { "epoch": 0.044278996865203764, "grad_norm": 3.0779130458831787, "learning_rate": 9.997944118803493e-06, "loss": 1.6766, "step": 565 }, { "epoch": 0.044357366771159874, "grad_norm": 3.2894558906555176, "learning_rate": 9.997935267404944e-06, "loss": 1.6433, "step": 566 }, { "epoch": 0.044435736677115985, "grad_norm": 2.6519291400909424, "learning_rate": 9.997926396996815e-06, "loss": 1.4921, "step": 567 }, { "epoch": 0.0445141065830721, "grad_norm": 3.4774773120880127, "learning_rate": 9.997917507579142e-06, "loss": 1.3124, "step": 568 }, { "epoch": 0.04459247648902821, "grad_norm": 5.4589385986328125, "learning_rate": 9.997908599151957e-06, "loss": 1.3651, "step": 569 }, { "epoch": 0.04467084639498432, "grad_norm": 3.6628336906433105, "learning_rate": 9.997899671715296e-06, "loss": 1.6119, "step": 570 }, { "epoch": 0.04474921630094044, "grad_norm": 3.176048755645752, "learning_rate": 9.997890725269192e-06, "loss": 1.226, "step": 571 }, { "epoch": 0.04482758620689655, "grad_norm": 4.682283878326416, "learning_rate": 9.997881759813678e-06, "loss": 1.3537, "step": 572 }, { "epoch": 0.04490595611285266, "grad_norm": 4.634692192077637, "learning_rate": 9.99787277534879e-06, "loss": 1.3081, "step": 573 }, { "epoch": 0.04498432601880878, "grad_norm": 4.148625373840332, "learning_rate": 9.997863771874559e-06, "loss": 1.108, "step": 574 }, { "epoch": 0.04506269592476489, "grad_norm": 3.536374568939209, "learning_rate": 9.997854749391021e-06, "loss": 1.2597, "step": 575 }, { "epoch": 0.045141065830721, "grad_norm": 4.858344554901123, "learning_rate": 9.997845707898213e-06, "loss": 1.3667, "step": 576 }, { "epoch": 0.04521943573667712, "grad_norm": 10.70732307434082, "learning_rate": 9.997836647396165e-06, "loss": 2.2473, "step": 577 }, { "epoch": 0.04529780564263323, "grad_norm": 9.609424591064453, "learning_rate": 9.997827567884915e-06, "loss": 1.2604, "step": 578 }, { "epoch": 0.04537617554858934, "grad_norm": 4.808374404907227, "learning_rate": 9.997818469364494e-06, "loss": 1.4875, "step": 579 }, { "epoch": 0.045454545454545456, "grad_norm": 7.154182434082031, "learning_rate": 9.997809351834939e-06, "loss": 1.1616, "step": 580 }, { "epoch": 0.045532915360501566, "grad_norm": 4.944233417510986, "learning_rate": 9.997800215296284e-06, "loss": 2.001, "step": 581 }, { "epoch": 0.045611285266457684, "grad_norm": 3.5005335807800293, "learning_rate": 9.997791059748567e-06, "loss": 1.6042, "step": 582 }, { "epoch": 0.045689655172413794, "grad_norm": 4.1379780769348145, "learning_rate": 9.997781885191818e-06, "loss": 1.6096, "step": 583 }, { "epoch": 0.045768025078369905, "grad_norm": 4.396320819854736, "learning_rate": 9.997772691626074e-06, "loss": 1.5931, "step": 584 }, { "epoch": 0.04584639498432602, "grad_norm": 5.97050142288208, "learning_rate": 9.997763479051367e-06, "loss": 1.6605, "step": 585 }, { "epoch": 0.04592476489028213, "grad_norm": 2.878765344619751, "learning_rate": 9.997754247467738e-06, "loss": 1.2291, "step": 586 }, { "epoch": 0.04600313479623824, "grad_norm": 2.34733247756958, "learning_rate": 9.997744996875218e-06, "loss": 0.8201, "step": 587 }, { "epoch": 0.04608150470219436, "grad_norm": 3.6256091594696045, "learning_rate": 9.997735727273842e-06, "loss": 1.463, "step": 588 }, { "epoch": 0.04615987460815047, "grad_norm": 5.290642738342285, "learning_rate": 9.997726438663649e-06, "loss": 1.4508, "step": 589 }, { "epoch": 0.04623824451410658, "grad_norm": 4.899080753326416, "learning_rate": 9.997717131044668e-06, "loss": 1.4631, "step": 590 }, { "epoch": 0.0463166144200627, "grad_norm": 4.0700812339782715, "learning_rate": 9.997707804416941e-06, "loss": 2.2799, "step": 591 }, { "epoch": 0.04639498432601881, "grad_norm": 3.506967067718506, "learning_rate": 9.997698458780499e-06, "loss": 1.5487, "step": 592 }, { "epoch": 0.04647335423197492, "grad_norm": 2.6413657665252686, "learning_rate": 9.997689094135378e-06, "loss": 1.4318, "step": 593 }, { "epoch": 0.04655172413793104, "grad_norm": 3.0979135036468506, "learning_rate": 9.997679710481617e-06, "loss": 1.1923, "step": 594 }, { "epoch": 0.04663009404388715, "grad_norm": 2.9680864810943604, "learning_rate": 9.997670307819248e-06, "loss": 1.4215, "step": 595 }, { "epoch": 0.04670846394984326, "grad_norm": 5.8227314949035645, "learning_rate": 9.997660886148308e-06, "loss": 1.974, "step": 596 }, { "epoch": 0.046786833855799376, "grad_norm": 2.7013449668884277, "learning_rate": 9.997651445468835e-06, "loss": 1.2887, "step": 597 }, { "epoch": 0.046865203761755486, "grad_norm": 4.246476650238037, "learning_rate": 9.99764198578086e-06, "loss": 1.4239, "step": 598 }, { "epoch": 0.0469435736677116, "grad_norm": 2.436180830001831, "learning_rate": 9.997632507084422e-06, "loss": 1.4984, "step": 599 }, { "epoch": 0.047021943573667714, "grad_norm": 3.0507256984710693, "learning_rate": 9.99762300937956e-06, "loss": 1.1336, "step": 600 }, { "epoch": 0.047100313479623825, "grad_norm": 5.503668308258057, "learning_rate": 9.997613492666305e-06, "loss": 1.4694, "step": 601 }, { "epoch": 0.047178683385579935, "grad_norm": 5.7364044189453125, "learning_rate": 9.997603956944695e-06, "loss": 1.256, "step": 602 }, { "epoch": 0.04725705329153605, "grad_norm": 2.263286590576172, "learning_rate": 9.997594402214766e-06, "loss": 1.4975, "step": 603 }, { "epoch": 0.04733542319749216, "grad_norm": 3.2516870498657227, "learning_rate": 9.997584828476556e-06, "loss": 1.4719, "step": 604 }, { "epoch": 0.04741379310344827, "grad_norm": 5.195148468017578, "learning_rate": 9.9975752357301e-06, "loss": 2.0595, "step": 605 }, { "epoch": 0.04749216300940439, "grad_norm": 6.548011302947998, "learning_rate": 9.997565623975433e-06, "loss": 1.6986, "step": 606 }, { "epoch": 0.0475705329153605, "grad_norm": 2.387922763824463, "learning_rate": 9.997555993212595e-06, "loss": 1.2418, "step": 607 }, { "epoch": 0.04764890282131661, "grad_norm": 2.4284942150115967, "learning_rate": 9.997546343441622e-06, "loss": 1.4845, "step": 608 }, { "epoch": 0.04772727272727273, "grad_norm": 8.29259967803955, "learning_rate": 9.997536674662548e-06, "loss": 2.0138, "step": 609 }, { "epoch": 0.04780564263322884, "grad_norm": 3.3737266063690186, "learning_rate": 9.997526986875412e-06, "loss": 1.6622, "step": 610 }, { "epoch": 0.04788401253918495, "grad_norm": 3.621049404144287, "learning_rate": 9.997517280080248e-06, "loss": 1.4427, "step": 611 }, { "epoch": 0.04796238244514107, "grad_norm": 4.489663124084473, "learning_rate": 9.997507554277098e-06, "loss": 1.6459, "step": 612 }, { "epoch": 0.04804075235109718, "grad_norm": 3.175931692123413, "learning_rate": 9.997497809465995e-06, "loss": 1.5707, "step": 613 }, { "epoch": 0.04811912225705329, "grad_norm": 3.3833656311035156, "learning_rate": 9.997488045646977e-06, "loss": 1.2308, "step": 614 }, { "epoch": 0.048197492163009406, "grad_norm": 4.916474342346191, "learning_rate": 9.997478262820082e-06, "loss": 1.7713, "step": 615 }, { "epoch": 0.04827586206896552, "grad_norm": 2.715115785598755, "learning_rate": 9.997468460985347e-06, "loss": 1.5218, "step": 616 }, { "epoch": 0.04835423197492163, "grad_norm": 6.07381534576416, "learning_rate": 9.997458640142808e-06, "loss": 1.4921, "step": 617 }, { "epoch": 0.048432601880877744, "grad_norm": 3.8845810890197754, "learning_rate": 9.997448800292504e-06, "loss": 1.2634, "step": 618 }, { "epoch": 0.048510971786833855, "grad_norm": 3.3179099559783936, "learning_rate": 9.99743894143447e-06, "loss": 1.5686, "step": 619 }, { "epoch": 0.048589341692789965, "grad_norm": 3.7506234645843506, "learning_rate": 9.997429063568746e-06, "loss": 1.3937, "step": 620 }, { "epoch": 0.04866771159874608, "grad_norm": 3.438798666000366, "learning_rate": 9.99741916669537e-06, "loss": 1.1864, "step": 621 }, { "epoch": 0.04874608150470219, "grad_norm": 3.296687126159668, "learning_rate": 9.997409250814375e-06, "loss": 1.4872, "step": 622 }, { "epoch": 0.048824451410658304, "grad_norm": 3.317495107650757, "learning_rate": 9.997399315925806e-06, "loss": 1.526, "step": 623 }, { "epoch": 0.04890282131661442, "grad_norm": 5.1751298904418945, "learning_rate": 9.997389362029695e-06, "loss": 1.5098, "step": 624 }, { "epoch": 0.04898119122257053, "grad_norm": 12.079497337341309, "learning_rate": 9.99737938912608e-06, "loss": 1.4542, "step": 625 }, { "epoch": 0.04905956112852665, "grad_norm": 4.2278642654418945, "learning_rate": 9.997369397215002e-06, "loss": 1.2768, "step": 626 }, { "epoch": 0.04913793103448276, "grad_norm": 8.391400337219238, "learning_rate": 9.997359386296499e-06, "loss": 1.1772, "step": 627 }, { "epoch": 0.04921630094043887, "grad_norm": 6.636970043182373, "learning_rate": 9.997349356370607e-06, "loss": 1.2624, "step": 628 }, { "epoch": 0.04929467084639499, "grad_norm": 2.7216291427612305, "learning_rate": 9.997339307437364e-06, "loss": 1.3197, "step": 629 }, { "epoch": 0.0493730407523511, "grad_norm": 7.398859024047852, "learning_rate": 9.997329239496808e-06, "loss": 1.5938, "step": 630 }, { "epoch": 0.04945141065830721, "grad_norm": 4.074520587921143, "learning_rate": 9.99731915254898e-06, "loss": 1.5491, "step": 631 }, { "epoch": 0.049529780564263326, "grad_norm": 2.897634267807007, "learning_rate": 9.997309046593917e-06, "loss": 1.3743, "step": 632 }, { "epoch": 0.049608150470219436, "grad_norm": 5.633405685424805, "learning_rate": 9.997298921631657e-06, "loss": 1.8002, "step": 633 }, { "epoch": 0.04968652037617555, "grad_norm": 3.4483139514923096, "learning_rate": 9.997288777662239e-06, "loss": 1.4926, "step": 634 }, { "epoch": 0.049764890282131664, "grad_norm": 3.3149516582489014, "learning_rate": 9.9972786146857e-06, "loss": 1.3621, "step": 635 }, { "epoch": 0.049843260188087775, "grad_norm": 3.0590221881866455, "learning_rate": 9.997268432702081e-06, "loss": 1.1755, "step": 636 }, { "epoch": 0.049921630094043885, "grad_norm": 3.5727128982543945, "learning_rate": 9.99725823171142e-06, "loss": 2.1141, "step": 637 }, { "epoch": 0.05, "grad_norm": 4.366815567016602, "learning_rate": 9.997248011713754e-06, "loss": 1.7252, "step": 638 }, { "epoch": 0.05007836990595611, "grad_norm": 3.1103286743164062, "learning_rate": 9.997237772709124e-06, "loss": 1.5122, "step": 639 }, { "epoch": 0.050156739811912224, "grad_norm": 4.426663398742676, "learning_rate": 9.997227514697568e-06, "loss": 1.0239, "step": 640 }, { "epoch": 0.05023510971786834, "grad_norm": 22.96080780029297, "learning_rate": 9.997217237679125e-06, "loss": 1.2271, "step": 641 }, { "epoch": 0.05031347962382445, "grad_norm": 3.1722609996795654, "learning_rate": 9.997206941653835e-06, "loss": 1.822, "step": 642 }, { "epoch": 0.05039184952978056, "grad_norm": 11.811065673828125, "learning_rate": 9.997196626621737e-06, "loss": 1.1397, "step": 643 }, { "epoch": 0.05047021943573668, "grad_norm": 4.23255729675293, "learning_rate": 9.99718629258287e-06, "loss": 1.4672, "step": 644 }, { "epoch": 0.05054858934169279, "grad_norm": 4.23487663269043, "learning_rate": 9.997175939537271e-06, "loss": 1.4564, "step": 645 }, { "epoch": 0.0506269592476489, "grad_norm": 6.486722946166992, "learning_rate": 9.997165567484984e-06, "loss": 1.2424, "step": 646 }, { "epoch": 0.05070532915360502, "grad_norm": 3.9536588191986084, "learning_rate": 9.997155176426043e-06, "loss": 1.5121, "step": 647 }, { "epoch": 0.05078369905956113, "grad_norm": 3.7275784015655518, "learning_rate": 9.997144766360492e-06, "loss": 1.2354, "step": 648 }, { "epoch": 0.05086206896551724, "grad_norm": 5.077805519104004, "learning_rate": 9.99713433728837e-06, "loss": 1.5277, "step": 649 }, { "epoch": 0.050940438871473356, "grad_norm": 3.3976361751556396, "learning_rate": 9.997123889209714e-06, "loss": 1.7537, "step": 650 }, { "epoch": 0.05101880877742947, "grad_norm": 3.6526975631713867, "learning_rate": 9.997113422124567e-06, "loss": 1.001, "step": 651 }, { "epoch": 0.05109717868338558, "grad_norm": 3.0630481243133545, "learning_rate": 9.997102936032966e-06, "loss": 1.5898, "step": 652 }, { "epoch": 0.051175548589341695, "grad_norm": 4.101564407348633, "learning_rate": 9.997092430934952e-06, "loss": 1.6365, "step": 653 }, { "epoch": 0.051253918495297805, "grad_norm": 5.897068500518799, "learning_rate": 9.997081906830566e-06, "loss": 1.45, "step": 654 }, { "epoch": 0.051332288401253916, "grad_norm": 8.297818183898926, "learning_rate": 9.997071363719846e-06, "loss": 2.0392, "step": 655 }, { "epoch": 0.05141065830721003, "grad_norm": 4.230759620666504, "learning_rate": 9.997060801602833e-06, "loss": 1.6834, "step": 656 }, { "epoch": 0.051489028213166144, "grad_norm": 3.515939712524414, "learning_rate": 9.997050220479569e-06, "loss": 1.5174, "step": 657 }, { "epoch": 0.051567398119122254, "grad_norm": 6.56939172744751, "learning_rate": 9.997039620350091e-06, "loss": 1.3177, "step": 658 }, { "epoch": 0.05164576802507837, "grad_norm": 2.8037478923797607, "learning_rate": 9.997029001214442e-06, "loss": 1.4273, "step": 659 }, { "epoch": 0.05172413793103448, "grad_norm": 2.98742413520813, "learning_rate": 9.99701836307266e-06, "loss": 1.47, "step": 660 }, { "epoch": 0.05180250783699059, "grad_norm": 3.088625907897949, "learning_rate": 9.997007705924789e-06, "loss": 1.119, "step": 661 }, { "epoch": 0.05188087774294671, "grad_norm": 4.290188789367676, "learning_rate": 9.996997029770865e-06, "loss": 1.3418, "step": 662 }, { "epoch": 0.05195924764890282, "grad_norm": 2.9376535415649414, "learning_rate": 9.996986334610933e-06, "loss": 1.5219, "step": 663 }, { "epoch": 0.05203761755485893, "grad_norm": 3.8358278274536133, "learning_rate": 9.99697562044503e-06, "loss": 0.8421, "step": 664 }, { "epoch": 0.05211598746081505, "grad_norm": 3.6925854682922363, "learning_rate": 9.996964887273198e-06, "loss": 1.5247, "step": 665 }, { "epoch": 0.05219435736677116, "grad_norm": 6.326485633850098, "learning_rate": 9.99695413509548e-06, "loss": 1.3834, "step": 666 }, { "epoch": 0.05227272727272727, "grad_norm": 5.7541680335998535, "learning_rate": 9.996943363911912e-06, "loss": 1.6932, "step": 667 }, { "epoch": 0.05235109717868339, "grad_norm": 3.235422134399414, "learning_rate": 9.996932573722542e-06, "loss": 1.342, "step": 668 }, { "epoch": 0.0524294670846395, "grad_norm": 3.272825241088867, "learning_rate": 9.996921764527405e-06, "loss": 1.0092, "step": 669 }, { "epoch": 0.052507836990595615, "grad_norm": 4.2543230056762695, "learning_rate": 9.996910936326545e-06, "loss": 1.7984, "step": 670 }, { "epoch": 0.052586206896551725, "grad_norm": 3.3886642456054688, "learning_rate": 9.996900089120001e-06, "loss": 1.4673, "step": 671 }, { "epoch": 0.052664576802507836, "grad_norm": 4.442543029785156, "learning_rate": 9.996889222907817e-06, "loss": 1.612, "step": 672 }, { "epoch": 0.05274294670846395, "grad_norm": 3.8057379722595215, "learning_rate": 9.996878337690032e-06, "loss": 1.6227, "step": 673 }, { "epoch": 0.052821316614420064, "grad_norm": 3.304716110229492, "learning_rate": 9.996867433466689e-06, "loss": 1.1396, "step": 674 }, { "epoch": 0.052899686520376174, "grad_norm": 3.884275436401367, "learning_rate": 9.996856510237828e-06, "loss": 1.7856, "step": 675 }, { "epoch": 0.05297805642633229, "grad_norm": 7.488720893859863, "learning_rate": 9.996845568003493e-06, "loss": 1.56, "step": 676 }, { "epoch": 0.0530564263322884, "grad_norm": 9.740331649780273, "learning_rate": 9.996834606763722e-06, "loss": 1.8586, "step": 677 }, { "epoch": 0.05313479623824451, "grad_norm": 3.3284800052642822, "learning_rate": 9.99682362651856e-06, "loss": 1.4458, "step": 678 }, { "epoch": 0.05321316614420063, "grad_norm": 3.530440092086792, "learning_rate": 9.996812627268047e-06, "loss": 1.8346, "step": 679 }, { "epoch": 0.05329153605015674, "grad_norm": 5.112318515777588, "learning_rate": 9.996801609012225e-06, "loss": 1.2194, "step": 680 }, { "epoch": 0.05336990595611285, "grad_norm": 5.929282188415527, "learning_rate": 9.996790571751137e-06, "loss": 1.8528, "step": 681 }, { "epoch": 0.05344827586206897, "grad_norm": 4.024808883666992, "learning_rate": 9.996779515484824e-06, "loss": 1.5606, "step": 682 }, { "epoch": 0.05352664576802508, "grad_norm": 4.93729305267334, "learning_rate": 9.996768440213327e-06, "loss": 1.5601, "step": 683 }, { "epoch": 0.05360501567398119, "grad_norm": 7.021433353424072, "learning_rate": 9.996757345936692e-06, "loss": 1.9321, "step": 684 }, { "epoch": 0.05368338557993731, "grad_norm": 2.4045157432556152, "learning_rate": 9.996746232654956e-06, "loss": 1.052, "step": 685 }, { "epoch": 0.05376175548589342, "grad_norm": 3.1399247646331787, "learning_rate": 9.996735100368164e-06, "loss": 1.3257, "step": 686 }, { "epoch": 0.05384012539184953, "grad_norm": 5.939450263977051, "learning_rate": 9.996723949076358e-06, "loss": 1.5818, "step": 687 }, { "epoch": 0.053918495297805645, "grad_norm": 3.6414215564727783, "learning_rate": 9.996712778779583e-06, "loss": 1.3955, "step": 688 }, { "epoch": 0.053996865203761756, "grad_norm": 4.904694557189941, "learning_rate": 9.996701589477877e-06, "loss": 1.0288, "step": 689 }, { "epoch": 0.054075235109717866, "grad_norm": 2.2875192165374756, "learning_rate": 9.996690381171285e-06, "loss": 1.577, "step": 690 }, { "epoch": 0.05415360501567398, "grad_norm": 3.8440682888031006, "learning_rate": 9.99667915385985e-06, "loss": 1.5049, "step": 691 }, { "epoch": 0.054231974921630094, "grad_norm": 2.2218117713928223, "learning_rate": 9.996667907543614e-06, "loss": 1.571, "step": 692 }, { "epoch": 0.054310344827586204, "grad_norm": 3.0615200996398926, "learning_rate": 9.996656642222618e-06, "loss": 1.5388, "step": 693 }, { "epoch": 0.05438871473354232, "grad_norm": 4.097901344299316, "learning_rate": 9.996645357896907e-06, "loss": 1.3366, "step": 694 }, { "epoch": 0.05446708463949843, "grad_norm": 3.041779041290283, "learning_rate": 9.996634054566525e-06, "loss": 1.5327, "step": 695 }, { "epoch": 0.05454545454545454, "grad_norm": 3.387606382369995, "learning_rate": 9.996622732231513e-06, "loss": 1.2392, "step": 696 }, { "epoch": 0.05462382445141066, "grad_norm": 3.72232985496521, "learning_rate": 9.996611390891916e-06, "loss": 1.6082, "step": 697 }, { "epoch": 0.05470219435736677, "grad_norm": 3.4017882347106934, "learning_rate": 9.996600030547772e-06, "loss": 1.5377, "step": 698 }, { "epoch": 0.05478056426332288, "grad_norm": 5.143796443939209, "learning_rate": 9.99658865119913e-06, "loss": 1.3272, "step": 699 }, { "epoch": 0.054858934169279, "grad_norm": 3.3424394130706787, "learning_rate": 9.996577252846032e-06, "loss": 1.4204, "step": 700 }, { "epoch": 0.05493730407523511, "grad_norm": 6.280587196350098, "learning_rate": 9.99656583548852e-06, "loss": 1.2455, "step": 701 }, { "epoch": 0.05501567398119122, "grad_norm": 3.6109015941619873, "learning_rate": 9.996554399126637e-06, "loss": 1.8213, "step": 702 }, { "epoch": 0.05509404388714734, "grad_norm": 3.610240936279297, "learning_rate": 9.996542943760428e-06, "loss": 1.5453, "step": 703 }, { "epoch": 0.05517241379310345, "grad_norm": 2.861314296722412, "learning_rate": 9.996531469389935e-06, "loss": 1.3795, "step": 704 }, { "epoch": 0.05525078369905956, "grad_norm": 4.035175800323486, "learning_rate": 9.996519976015205e-06, "loss": 1.6455, "step": 705 }, { "epoch": 0.055329153605015675, "grad_norm": 5.151608943939209, "learning_rate": 9.996508463636276e-06, "loss": 1.4533, "step": 706 }, { "epoch": 0.055407523510971786, "grad_norm": 3.7661755084991455, "learning_rate": 9.996496932253199e-06, "loss": 1.5638, "step": 707 }, { "epoch": 0.055485893416927896, "grad_norm": 3.4815990924835205, "learning_rate": 9.996485381866011e-06, "loss": 1.5446, "step": 708 }, { "epoch": 0.055564263322884014, "grad_norm": 3.4322400093078613, "learning_rate": 9.996473812474761e-06, "loss": 1.3678, "step": 709 }, { "epoch": 0.055642633228840124, "grad_norm": 3.1311099529266357, "learning_rate": 9.996462224079489e-06, "loss": 1.5257, "step": 710 }, { "epoch": 0.055721003134796235, "grad_norm": 6.705576419830322, "learning_rate": 9.996450616680241e-06, "loss": 1.4548, "step": 711 }, { "epoch": 0.05579937304075235, "grad_norm": 8.150413513183594, "learning_rate": 9.996438990277064e-06, "loss": 2.088, "step": 712 }, { "epoch": 0.05587774294670846, "grad_norm": 3.3063156604766846, "learning_rate": 9.996427344869998e-06, "loss": 1.0171, "step": 713 }, { "epoch": 0.05595611285266458, "grad_norm": 2.9730746746063232, "learning_rate": 9.996415680459089e-06, "loss": 1.0743, "step": 714 }, { "epoch": 0.05603448275862069, "grad_norm": 3.809321403503418, "learning_rate": 9.996403997044379e-06, "loss": 1.2135, "step": 715 }, { "epoch": 0.0561128526645768, "grad_norm": 2.8480985164642334, "learning_rate": 9.996392294625917e-06, "loss": 1.4227, "step": 716 }, { "epoch": 0.05619122257053292, "grad_norm": 3.2118947505950928, "learning_rate": 9.996380573203745e-06, "loss": 1.4222, "step": 717 }, { "epoch": 0.05626959247648903, "grad_norm": 6.47639799118042, "learning_rate": 9.996368832777905e-06, "loss": 1.3118, "step": 718 }, { "epoch": 0.05634796238244514, "grad_norm": 3.496274948120117, "learning_rate": 9.996357073348447e-06, "loss": 1.4586, "step": 719 }, { "epoch": 0.05642633228840126, "grad_norm": 2.920452833175659, "learning_rate": 9.996345294915413e-06, "loss": 1.6131, "step": 720 }, { "epoch": 0.05650470219435737, "grad_norm": 4.361115455627441, "learning_rate": 9.996333497478845e-06, "loss": 1.3639, "step": 721 }, { "epoch": 0.05658307210031348, "grad_norm": 2.493709087371826, "learning_rate": 9.996321681038793e-06, "loss": 1.3592, "step": 722 }, { "epoch": 0.056661442006269595, "grad_norm": 3.8123629093170166, "learning_rate": 9.9963098455953e-06, "loss": 1.1412, "step": 723 }, { "epoch": 0.056739811912225706, "grad_norm": 2.6446049213409424, "learning_rate": 9.99629799114841e-06, "loss": 1.4477, "step": 724 }, { "epoch": 0.056818181818181816, "grad_norm": 3.9216067790985107, "learning_rate": 9.996286117698167e-06, "loss": 1.4403, "step": 725 }, { "epoch": 0.056896551724137934, "grad_norm": 3.868227243423462, "learning_rate": 9.99627422524462e-06, "loss": 1.9016, "step": 726 }, { "epoch": 0.056974921630094044, "grad_norm": 5.6113433837890625, "learning_rate": 9.99626231378781e-06, "loss": 1.4548, "step": 727 }, { "epoch": 0.057053291536050155, "grad_norm": 2.5805821418762207, "learning_rate": 9.996250383327785e-06, "loss": 1.3704, "step": 728 }, { "epoch": 0.05713166144200627, "grad_norm": 5.446245193481445, "learning_rate": 9.996238433864592e-06, "loss": 1.647, "step": 729 }, { "epoch": 0.05721003134796238, "grad_norm": 5.704201698303223, "learning_rate": 9.996226465398272e-06, "loss": 1.3047, "step": 730 }, { "epoch": 0.05728840125391849, "grad_norm": 2.2234559059143066, "learning_rate": 9.996214477928873e-06, "loss": 1.4103, "step": 731 }, { "epoch": 0.05736677115987461, "grad_norm": 4.102734088897705, "learning_rate": 9.996202471456443e-06, "loss": 1.5257, "step": 732 }, { "epoch": 0.05744514106583072, "grad_norm": 2.5847995281219482, "learning_rate": 9.996190445981023e-06, "loss": 1.5061, "step": 733 }, { "epoch": 0.05752351097178683, "grad_norm": 4.0973381996154785, "learning_rate": 9.996178401502659e-06, "loss": 1.4605, "step": 734 }, { "epoch": 0.05760188087774295, "grad_norm": 10.53648853302002, "learning_rate": 9.996166338021401e-06, "loss": 1.6041, "step": 735 }, { "epoch": 0.05768025078369906, "grad_norm": 3.723590850830078, "learning_rate": 9.99615425553729e-06, "loss": 1.7833, "step": 736 }, { "epoch": 0.05775862068965517, "grad_norm": 2.906801700592041, "learning_rate": 9.996142154050377e-06, "loss": 1.4878, "step": 737 }, { "epoch": 0.05783699059561129, "grad_norm": 2.4003677368164062, "learning_rate": 9.996130033560705e-06, "loss": 1.1476, "step": 738 }, { "epoch": 0.0579153605015674, "grad_norm": 2.3354651927948, "learning_rate": 9.99611789406832e-06, "loss": 1.7177, "step": 739 }, { "epoch": 0.05799373040752351, "grad_norm": 3.388631820678711, "learning_rate": 9.996105735573267e-06, "loss": 1.3415, "step": 740 }, { "epoch": 0.058072100313479626, "grad_norm": 8.03765869140625, "learning_rate": 9.996093558075596e-06, "loss": 1.5888, "step": 741 }, { "epoch": 0.058150470219435736, "grad_norm": 2.4162607192993164, "learning_rate": 9.996081361575352e-06, "loss": 1.2681, "step": 742 }, { "epoch": 0.05822884012539185, "grad_norm": 16.587989807128906, "learning_rate": 9.996069146072578e-06, "loss": 1.6138, "step": 743 }, { "epoch": 0.058307210031347964, "grad_norm": 4.234510898590088, "learning_rate": 9.996056911567324e-06, "loss": 1.4653, "step": 744 }, { "epoch": 0.058385579937304075, "grad_norm": 3.6059436798095703, "learning_rate": 9.996044658059637e-06, "loss": 1.4911, "step": 745 }, { "epoch": 0.058463949843260185, "grad_norm": 4.139786243438721, "learning_rate": 9.996032385549561e-06, "loss": 1.571, "step": 746 }, { "epoch": 0.0585423197492163, "grad_norm": 3.1962857246398926, "learning_rate": 9.996020094037144e-06, "loss": 1.5139, "step": 747 }, { "epoch": 0.05862068965517241, "grad_norm": 2.158923625946045, "learning_rate": 9.996007783522435e-06, "loss": 1.242, "step": 748 }, { "epoch": 0.05869905956112852, "grad_norm": 2.8615646362304688, "learning_rate": 9.995995454005476e-06, "loss": 1.3445, "step": 749 }, { "epoch": 0.05877742946708464, "grad_norm": 3.2810959815979004, "learning_rate": 9.995983105486316e-06, "loss": 1.5971, "step": 750 }, { "epoch": 0.05885579937304075, "grad_norm": 2.5148067474365234, "learning_rate": 9.995970737965004e-06, "loss": 1.5565, "step": 751 }, { "epoch": 0.05893416927899686, "grad_norm": 2.7377140522003174, "learning_rate": 9.995958351441585e-06, "loss": 1.2491, "step": 752 }, { "epoch": 0.05901253918495298, "grad_norm": 3.0022733211517334, "learning_rate": 9.995945945916106e-06, "loss": 1.3707, "step": 753 }, { "epoch": 0.05909090909090909, "grad_norm": 2.9923934936523438, "learning_rate": 9.995933521388614e-06, "loss": 1.5261, "step": 754 }, { "epoch": 0.05916927899686521, "grad_norm": 2.682640552520752, "learning_rate": 9.99592107785916e-06, "loss": 1.318, "step": 755 }, { "epoch": 0.05924764890282132, "grad_norm": 3.624115467071533, "learning_rate": 9.995908615327785e-06, "loss": 1.3246, "step": 756 }, { "epoch": 0.05932601880877743, "grad_norm": 2.813732862472534, "learning_rate": 9.99589613379454e-06, "loss": 1.1786, "step": 757 }, { "epoch": 0.059404388714733546, "grad_norm": 3.4927608966827393, "learning_rate": 9.995883633259474e-06, "loss": 1.866, "step": 758 }, { "epoch": 0.059482758620689656, "grad_norm": 3.744640588760376, "learning_rate": 9.995871113722631e-06, "loss": 1.4266, "step": 759 }, { "epoch": 0.05956112852664577, "grad_norm": 2.9959986209869385, "learning_rate": 9.995858575184062e-06, "loss": 1.5292, "step": 760 }, { "epoch": 0.059639498432601884, "grad_norm": 2.467294216156006, "learning_rate": 9.995846017643812e-06, "loss": 1.1413, "step": 761 }, { "epoch": 0.059717868338557994, "grad_norm": 3.268178701400757, "learning_rate": 9.99583344110193e-06, "loss": 1.4947, "step": 762 }, { "epoch": 0.059796238244514105, "grad_norm": 3.4098455905914307, "learning_rate": 9.995820845558464e-06, "loss": 1.9473, "step": 763 }, { "epoch": 0.05987460815047022, "grad_norm": 6.4519500732421875, "learning_rate": 9.99580823101346e-06, "loss": 1.3443, "step": 764 }, { "epoch": 0.05995297805642633, "grad_norm": 2.231196880340576, "learning_rate": 9.995795597466968e-06, "loss": 1.4038, "step": 765 }, { "epoch": 0.06003134796238244, "grad_norm": 4.044654846191406, "learning_rate": 9.995782944919037e-06, "loss": 1.3778, "step": 766 }, { "epoch": 0.06010971786833856, "grad_norm": 4.139191627502441, "learning_rate": 9.995770273369712e-06, "loss": 1.3477, "step": 767 }, { "epoch": 0.06018808777429467, "grad_norm": 6.074293613433838, "learning_rate": 9.995757582819045e-06, "loss": 1.6577, "step": 768 }, { "epoch": 0.06026645768025078, "grad_norm": 3.2110466957092285, "learning_rate": 9.995744873267079e-06, "loss": 1.0395, "step": 769 }, { "epoch": 0.0603448275862069, "grad_norm": 5.566353797912598, "learning_rate": 9.995732144713867e-06, "loss": 1.8051, "step": 770 }, { "epoch": 0.06042319749216301, "grad_norm": 3.32370662689209, "learning_rate": 9.995719397159458e-06, "loss": 1.637, "step": 771 }, { "epoch": 0.06050156739811912, "grad_norm": 3.0669472217559814, "learning_rate": 9.995706630603896e-06, "loss": 1.3332, "step": 772 }, { "epoch": 0.06057993730407524, "grad_norm": 6.5795512199401855, "learning_rate": 9.995693845047232e-06, "loss": 1.7593, "step": 773 }, { "epoch": 0.06065830721003135, "grad_norm": 3.0297632217407227, "learning_rate": 9.995681040489516e-06, "loss": 1.4883, "step": 774 }, { "epoch": 0.06073667711598746, "grad_norm": 6.644439220428467, "learning_rate": 9.995668216930793e-06, "loss": 1.8063, "step": 775 }, { "epoch": 0.060815047021943576, "grad_norm": 4.968495845794678, "learning_rate": 9.995655374371116e-06, "loss": 1.7751, "step": 776 }, { "epoch": 0.060893416927899686, "grad_norm": 5.758484363555908, "learning_rate": 9.99564251281053e-06, "loss": 1.6495, "step": 777 }, { "epoch": 0.0609717868338558, "grad_norm": 4.612893104553223, "learning_rate": 9.995629632249089e-06, "loss": 1.4354, "step": 778 }, { "epoch": 0.061050156739811914, "grad_norm": 4.73106050491333, "learning_rate": 9.995616732686835e-06, "loss": 1.5976, "step": 779 }, { "epoch": 0.061128526645768025, "grad_norm": 3.3993513584136963, "learning_rate": 9.995603814123823e-06, "loss": 1.3736, "step": 780 }, { "epoch": 0.061206896551724135, "grad_norm": 3.2123773097991943, "learning_rate": 9.9955908765601e-06, "loss": 1.6334, "step": 781 }, { "epoch": 0.06128526645768025, "grad_norm": 3.720738410949707, "learning_rate": 9.995577919995715e-06, "loss": 1.6071, "step": 782 }, { "epoch": 0.06136363636363636, "grad_norm": 3.4049885272979736, "learning_rate": 9.995564944430717e-06, "loss": 1.5791, "step": 783 }, { "epoch": 0.061442006269592474, "grad_norm": 2.61588978767395, "learning_rate": 9.995551949865157e-06, "loss": 1.4759, "step": 784 }, { "epoch": 0.06152037617554859, "grad_norm": 4.949117183685303, "learning_rate": 9.995538936299081e-06, "loss": 1.4194, "step": 785 }, { "epoch": 0.0615987460815047, "grad_norm": 3.179506301879883, "learning_rate": 9.995525903732543e-06, "loss": 1.4702, "step": 786 }, { "epoch": 0.06167711598746081, "grad_norm": 3.3377010822296143, "learning_rate": 9.99551285216559e-06, "loss": 1.5196, "step": 787 }, { "epoch": 0.06175548589341693, "grad_norm": 3.046560525894165, "learning_rate": 9.995499781598271e-06, "loss": 1.7439, "step": 788 }, { "epoch": 0.06183385579937304, "grad_norm": 3.408067464828491, "learning_rate": 9.995486692030637e-06, "loss": 1.2904, "step": 789 }, { "epoch": 0.06191222570532915, "grad_norm": 4.003661632537842, "learning_rate": 9.995473583462737e-06, "loss": 1.3, "step": 790 }, { "epoch": 0.06199059561128527, "grad_norm": 4.578888893127441, "learning_rate": 9.995460455894623e-06, "loss": 1.6741, "step": 791 }, { "epoch": 0.06206896551724138, "grad_norm": 3.0368645191192627, "learning_rate": 9.995447309326342e-06, "loss": 1.4924, "step": 792 }, { "epoch": 0.06214733542319749, "grad_norm": 7.498789310455322, "learning_rate": 9.995434143757944e-06, "loss": 1.4848, "step": 793 }, { "epoch": 0.062225705329153606, "grad_norm": 2.714195489883423, "learning_rate": 9.995420959189483e-06, "loss": 1.4899, "step": 794 }, { "epoch": 0.06230407523510972, "grad_norm": 5.059739112854004, "learning_rate": 9.995407755621004e-06, "loss": 1.7508, "step": 795 }, { "epoch": 0.06238244514106583, "grad_norm": 3.483281373977661, "learning_rate": 9.995394533052563e-06, "loss": 1.6154, "step": 796 }, { "epoch": 0.062460815047021945, "grad_norm": 3.7685558795928955, "learning_rate": 9.995381291484206e-06, "loss": 1.6238, "step": 797 }, { "epoch": 0.06253918495297806, "grad_norm": 3.4516243934631348, "learning_rate": 9.995368030915981e-06, "loss": 1.5153, "step": 798 }, { "epoch": 0.06261755485893417, "grad_norm": 2.2314629554748535, "learning_rate": 9.995354751347945e-06, "loss": 1.352, "step": 799 }, { "epoch": 0.06269592476489028, "grad_norm": 3.3955750465393066, "learning_rate": 9.995341452780145e-06, "loss": 1.5336, "step": 800 }, { "epoch": 0.0627742946708464, "grad_norm": 4.175930976867676, "learning_rate": 9.99532813521263e-06, "loss": 1.3857, "step": 801 }, { "epoch": 0.06285266457680251, "grad_norm": 4.824921131134033, "learning_rate": 9.995314798645454e-06, "loss": 1.5746, "step": 802 }, { "epoch": 0.06293103448275862, "grad_norm": 8.917034149169922, "learning_rate": 9.995301443078666e-06, "loss": 2.4325, "step": 803 }, { "epoch": 0.06300940438871473, "grad_norm": 5.980774879455566, "learning_rate": 9.995288068512317e-06, "loss": 1.4833, "step": 804 }, { "epoch": 0.06308777429467084, "grad_norm": 2.7619740962982178, "learning_rate": 9.995274674946459e-06, "loss": 1.3116, "step": 805 }, { "epoch": 0.06316614420062695, "grad_norm": 3.9906394481658936, "learning_rate": 9.995261262381141e-06, "loss": 1.5821, "step": 806 }, { "epoch": 0.06324451410658308, "grad_norm": 6.356921195983887, "learning_rate": 9.995247830816414e-06, "loss": 1.6192, "step": 807 }, { "epoch": 0.06332288401253919, "grad_norm": 4.726423263549805, "learning_rate": 9.995234380252331e-06, "loss": 1.1735, "step": 808 }, { "epoch": 0.0634012539184953, "grad_norm": 4.917217254638672, "learning_rate": 9.995220910688944e-06, "loss": 1.687, "step": 809 }, { "epoch": 0.06347962382445141, "grad_norm": 3.8944125175476074, "learning_rate": 9.995207422126299e-06, "loss": 1.283, "step": 810 }, { "epoch": 0.06355799373040752, "grad_norm": 3.4908344745635986, "learning_rate": 9.995193914564451e-06, "loss": 1.3714, "step": 811 }, { "epoch": 0.06363636363636363, "grad_norm": 2.315441846847534, "learning_rate": 9.995180388003454e-06, "loss": 1.1938, "step": 812 }, { "epoch": 0.06371473354231975, "grad_norm": 3.9528915882110596, "learning_rate": 9.995166842443354e-06, "loss": 1.4909, "step": 813 }, { "epoch": 0.06379310344827586, "grad_norm": 3.6708292961120605, "learning_rate": 9.995153277884204e-06, "loss": 0.9261, "step": 814 }, { "epoch": 0.06387147335423198, "grad_norm": 3.0794496536254883, "learning_rate": 9.99513969432606e-06, "loss": 1.328, "step": 815 }, { "epoch": 0.06394984326018809, "grad_norm": 3.9726402759552, "learning_rate": 9.995126091768967e-06, "loss": 1.2755, "step": 816 }, { "epoch": 0.0640282131661442, "grad_norm": 2.9693899154663086, "learning_rate": 9.995112470212981e-06, "loss": 1.5102, "step": 817 }, { "epoch": 0.0641065830721003, "grad_norm": 2.724940299987793, "learning_rate": 9.995098829658155e-06, "loss": 1.1043, "step": 818 }, { "epoch": 0.06418495297805643, "grad_norm": 6.3809051513671875, "learning_rate": 9.995085170104535e-06, "loss": 2.1077, "step": 819 }, { "epoch": 0.06426332288401254, "grad_norm": 2.816819190979004, "learning_rate": 9.99507149155218e-06, "loss": 1.3046, "step": 820 }, { "epoch": 0.06434169278996865, "grad_norm": 2.922861099243164, "learning_rate": 9.995057794001137e-06, "loss": 1.3974, "step": 821 }, { "epoch": 0.06442006269592476, "grad_norm": 6.1307830810546875, "learning_rate": 9.99504407745146e-06, "loss": 1.5735, "step": 822 }, { "epoch": 0.06449843260188087, "grad_norm": 8.233711242675781, "learning_rate": 9.9950303419032e-06, "loss": 1.3486, "step": 823 }, { "epoch": 0.064576802507837, "grad_norm": 3.5897624492645264, "learning_rate": 9.995016587356412e-06, "loss": 1.2839, "step": 824 }, { "epoch": 0.06465517241379311, "grad_norm": 8.798629760742188, "learning_rate": 9.995002813811145e-06, "loss": 2.2706, "step": 825 }, { "epoch": 0.06473354231974922, "grad_norm": 3.5818042755126953, "learning_rate": 9.994989021267454e-06, "loss": 1.4097, "step": 826 }, { "epoch": 0.06481191222570533, "grad_norm": 3.389334201812744, "learning_rate": 9.99497520972539e-06, "loss": 1.522, "step": 827 }, { "epoch": 0.06489028213166144, "grad_norm": 3.5448930263519287, "learning_rate": 9.994961379185007e-06, "loss": 1.195, "step": 828 }, { "epoch": 0.06496865203761755, "grad_norm": 3.5722641944885254, "learning_rate": 9.994947529646354e-06, "loss": 1.3208, "step": 829 }, { "epoch": 0.06504702194357367, "grad_norm": 8.050168991088867, "learning_rate": 9.994933661109486e-06, "loss": 0.9901, "step": 830 }, { "epoch": 0.06512539184952978, "grad_norm": 2.8615598678588867, "learning_rate": 9.994919773574459e-06, "loss": 1.4127, "step": 831 }, { "epoch": 0.0652037617554859, "grad_norm": 4.102141857147217, "learning_rate": 9.99490586704132e-06, "loss": 1.5258, "step": 832 }, { "epoch": 0.065282131661442, "grad_norm": 8.384621620178223, "learning_rate": 9.994891941510128e-06, "loss": 1.7726, "step": 833 }, { "epoch": 0.06536050156739812, "grad_norm": 3.3797333240509033, "learning_rate": 9.99487799698093e-06, "loss": 1.8838, "step": 834 }, { "epoch": 0.06543887147335423, "grad_norm": 3.7544167041778564, "learning_rate": 9.994864033453781e-06, "loss": 1.3152, "step": 835 }, { "epoch": 0.06551724137931035, "grad_norm": 4.191732406616211, "learning_rate": 9.994850050928737e-06, "loss": 1.7781, "step": 836 }, { "epoch": 0.06559561128526646, "grad_norm": 4.012787342071533, "learning_rate": 9.994836049405848e-06, "loss": 1.2734, "step": 837 }, { "epoch": 0.06567398119122257, "grad_norm": 4.341224193572998, "learning_rate": 9.994822028885167e-06, "loss": 1.2789, "step": 838 }, { "epoch": 0.06575235109717868, "grad_norm": 3.4440581798553467, "learning_rate": 9.994807989366748e-06, "loss": 1.3267, "step": 839 }, { "epoch": 0.06583072100313479, "grad_norm": 8.813708305358887, "learning_rate": 9.994793930850647e-06, "loss": 2.0795, "step": 840 }, { "epoch": 0.0659090909090909, "grad_norm": 3.5149269104003906, "learning_rate": 9.994779853336914e-06, "loss": 1.6396, "step": 841 }, { "epoch": 0.06598746081504703, "grad_norm": 6.324432373046875, "learning_rate": 9.994765756825606e-06, "loss": 1.5444, "step": 842 }, { "epoch": 0.06606583072100314, "grad_norm": 4.9064812660217285, "learning_rate": 9.994751641316771e-06, "loss": 1.4011, "step": 843 }, { "epoch": 0.06614420062695925, "grad_norm": 16.93113899230957, "learning_rate": 9.994737506810469e-06, "loss": 1.7583, "step": 844 }, { "epoch": 0.06622257053291536, "grad_norm": 6.3022990226745605, "learning_rate": 9.99472335330675e-06, "loss": 1.2915, "step": 845 }, { "epoch": 0.06630094043887147, "grad_norm": 3.674778699874878, "learning_rate": 9.99470918080567e-06, "loss": 1.3952, "step": 846 }, { "epoch": 0.06637931034482758, "grad_norm": 4.237362384796143, "learning_rate": 9.99469498930728e-06, "loss": 1.7638, "step": 847 }, { "epoch": 0.0664576802507837, "grad_norm": 5.834641933441162, "learning_rate": 9.994680778811636e-06, "loss": 1.5243, "step": 848 }, { "epoch": 0.06653605015673981, "grad_norm": 2.6394340991973877, "learning_rate": 9.994666549318793e-06, "loss": 1.1245, "step": 849 }, { "epoch": 0.06661442006269593, "grad_norm": 2.8129570484161377, "learning_rate": 9.994652300828803e-06, "loss": 1.5899, "step": 850 }, { "epoch": 0.06669278996865204, "grad_norm": 3.256989002227783, "learning_rate": 9.994638033341722e-06, "loss": 1.3703, "step": 851 }, { "epoch": 0.06677115987460815, "grad_norm": 3.4814774990081787, "learning_rate": 9.994623746857602e-06, "loss": 1.3616, "step": 852 }, { "epoch": 0.06684952978056426, "grad_norm": 3.1176133155822754, "learning_rate": 9.9946094413765e-06, "loss": 1.5859, "step": 853 }, { "epoch": 0.06692789968652038, "grad_norm": 3.0192315578460693, "learning_rate": 9.994595116898467e-06, "loss": 1.7388, "step": 854 }, { "epoch": 0.06700626959247649, "grad_norm": 4.456575393676758, "learning_rate": 9.994580773423561e-06, "loss": 1.4812, "step": 855 }, { "epoch": 0.0670846394984326, "grad_norm": 3.1650702953338623, "learning_rate": 9.994566410951835e-06, "loss": 1.3268, "step": 856 }, { "epoch": 0.06716300940438871, "grad_norm": 5.114577770233154, "learning_rate": 9.994552029483344e-06, "loss": 1.6218, "step": 857 }, { "epoch": 0.06724137931034482, "grad_norm": 2.6307711601257324, "learning_rate": 9.994537629018142e-06, "loss": 1.2517, "step": 858 }, { "epoch": 0.06731974921630093, "grad_norm": 5.124380588531494, "learning_rate": 9.994523209556283e-06, "loss": 1.6507, "step": 859 }, { "epoch": 0.06739811912225706, "grad_norm": 1.890252709388733, "learning_rate": 9.994508771097825e-06, "loss": 1.1919, "step": 860 }, { "epoch": 0.06747648902821317, "grad_norm": 5.499197006225586, "learning_rate": 9.994494313642822e-06, "loss": 1.2923, "step": 861 }, { "epoch": 0.06755485893416928, "grad_norm": 6.150635719299316, "learning_rate": 9.994479837191324e-06, "loss": 1.4417, "step": 862 }, { "epoch": 0.06763322884012539, "grad_norm": 11.192870140075684, "learning_rate": 9.994465341743393e-06, "loss": 1.2364, "step": 863 }, { "epoch": 0.0677115987460815, "grad_norm": 6.716517925262451, "learning_rate": 9.99445082729908e-06, "loss": 1.7099, "step": 864 }, { "epoch": 0.06778996865203762, "grad_norm": 3.225163459777832, "learning_rate": 9.994436293858442e-06, "loss": 1.7274, "step": 865 }, { "epoch": 0.06786833855799373, "grad_norm": 6.619974613189697, "learning_rate": 9.994421741421533e-06, "loss": 0.9808, "step": 866 }, { "epoch": 0.06794670846394985, "grad_norm": 4.382356643676758, "learning_rate": 9.99440716998841e-06, "loss": 1.8164, "step": 867 }, { "epoch": 0.06802507836990596, "grad_norm": 17.90068817138672, "learning_rate": 9.994392579559126e-06, "loss": 1.3542, "step": 868 }, { "epoch": 0.06810344827586207, "grad_norm": 2.636775255203247, "learning_rate": 9.994377970133739e-06, "loss": 1.3769, "step": 869 }, { "epoch": 0.06818181818181818, "grad_norm": 4.677873611450195, "learning_rate": 9.994363341712306e-06, "loss": 1.5453, "step": 870 }, { "epoch": 0.0682601880877743, "grad_norm": 5.773815631866455, "learning_rate": 9.994348694294875e-06, "loss": 1.3235, "step": 871 }, { "epoch": 0.06833855799373041, "grad_norm": 3.410998821258545, "learning_rate": 9.99433402788151e-06, "loss": 1.5102, "step": 872 }, { "epoch": 0.06841692789968652, "grad_norm": 3.22238826751709, "learning_rate": 9.994319342472263e-06, "loss": 1.3229, "step": 873 }, { "epoch": 0.06849529780564263, "grad_norm": 3.2888851165771484, "learning_rate": 9.994304638067191e-06, "loss": 1.4388, "step": 874 }, { "epoch": 0.06857366771159874, "grad_norm": 2.2366793155670166, "learning_rate": 9.994289914666348e-06, "loss": 1.3723, "step": 875 }, { "epoch": 0.06865203761755485, "grad_norm": 2.645061492919922, "learning_rate": 9.994275172269791e-06, "loss": 1.8685, "step": 876 }, { "epoch": 0.06873040752351098, "grad_norm": 4.858464241027832, "learning_rate": 9.994260410877579e-06, "loss": 2.1395, "step": 877 }, { "epoch": 0.06880877742946709, "grad_norm": 4.264606952667236, "learning_rate": 9.994245630489763e-06, "loss": 1.4947, "step": 878 }, { "epoch": 0.0688871473354232, "grad_norm": 7.33391809463501, "learning_rate": 9.994230831106404e-06, "loss": 1.4595, "step": 879 }, { "epoch": 0.06896551724137931, "grad_norm": 4.271832466125488, "learning_rate": 9.994216012727556e-06, "loss": 1.7546, "step": 880 }, { "epoch": 0.06904388714733542, "grad_norm": 12.092948913574219, "learning_rate": 9.994201175353273e-06, "loss": 1.5851, "step": 881 }, { "epoch": 0.06912225705329153, "grad_norm": 4.605420112609863, "learning_rate": 9.994186318983616e-06, "loss": 1.476, "step": 882 }, { "epoch": 0.06920062695924765, "grad_norm": 2.860887050628662, "learning_rate": 9.994171443618638e-06, "loss": 1.3351, "step": 883 }, { "epoch": 0.06927899686520377, "grad_norm": 5.503057479858398, "learning_rate": 9.994156549258399e-06, "loss": 1.787, "step": 884 }, { "epoch": 0.06935736677115988, "grad_norm": 2.975905179977417, "learning_rate": 9.994141635902952e-06, "loss": 1.4994, "step": 885 }, { "epoch": 0.06943573667711599, "grad_norm": 3.386075258255005, "learning_rate": 9.994126703552357e-06, "loss": 1.5374, "step": 886 }, { "epoch": 0.0695141065830721, "grad_norm": 5.305253505706787, "learning_rate": 9.994111752206668e-06, "loss": 1.5992, "step": 887 }, { "epoch": 0.06959247648902821, "grad_norm": 5.636916160583496, "learning_rate": 9.994096781865943e-06, "loss": 1.0585, "step": 888 }, { "epoch": 0.06967084639498433, "grad_norm": 3.5946853160858154, "learning_rate": 9.994081792530239e-06, "loss": 1.6455, "step": 889 }, { "epoch": 0.06974921630094044, "grad_norm": 4.973486423492432, "learning_rate": 9.994066784199613e-06, "loss": 1.6577, "step": 890 }, { "epoch": 0.06982758620689655, "grad_norm": 4.51138162612915, "learning_rate": 9.994051756874122e-06, "loss": 1.412, "step": 891 }, { "epoch": 0.06990595611285266, "grad_norm": 2.970527172088623, "learning_rate": 9.994036710553823e-06, "loss": 1.4634, "step": 892 }, { "epoch": 0.06998432601880877, "grad_norm": 4.1441216468811035, "learning_rate": 9.994021645238774e-06, "loss": 1.5638, "step": 893 }, { "epoch": 0.07006269592476488, "grad_norm": 3.3830714225769043, "learning_rate": 9.994006560929032e-06, "loss": 1.7426, "step": 894 }, { "epoch": 0.07014106583072101, "grad_norm": 9.927895545959473, "learning_rate": 9.993991457624654e-06, "loss": 0.8057, "step": 895 }, { "epoch": 0.07021943573667712, "grad_norm": 3.9676225185394287, "learning_rate": 9.993976335325697e-06, "loss": 1.508, "step": 896 }, { "epoch": 0.07029780564263323, "grad_norm": 2.9192423820495605, "learning_rate": 9.99396119403222e-06, "loss": 1.2385, "step": 897 }, { "epoch": 0.07037617554858934, "grad_norm": 3.1602540016174316, "learning_rate": 9.99394603374428e-06, "loss": 1.6786, "step": 898 }, { "epoch": 0.07045454545454545, "grad_norm": 2.5611462593078613, "learning_rate": 9.993930854461932e-06, "loss": 1.741, "step": 899 }, { "epoch": 0.07053291536050156, "grad_norm": 3.0633633136749268, "learning_rate": 9.993915656185239e-06, "loss": 1.4962, "step": 900 }, { "epoch": 0.07061128526645769, "grad_norm": 3.613264799118042, "learning_rate": 9.993900438914253e-06, "loss": 1.3756, "step": 901 }, { "epoch": 0.0706896551724138, "grad_norm": 4.591477394104004, "learning_rate": 9.993885202649036e-06, "loss": 1.6837, "step": 902 }, { "epoch": 0.0707680250783699, "grad_norm": 2.6184661388397217, "learning_rate": 9.993869947389646e-06, "loss": 1.5306, "step": 903 }, { "epoch": 0.07084639498432602, "grad_norm": 3.2204315662384033, "learning_rate": 9.993854673136138e-06, "loss": 1.6691, "step": 904 }, { "epoch": 0.07092476489028213, "grad_norm": 2.983339309692383, "learning_rate": 9.993839379888572e-06, "loss": 1.5974, "step": 905 }, { "epoch": 0.07100313479623825, "grad_norm": 2.030918836593628, "learning_rate": 9.993824067647007e-06, "loss": 1.1924, "step": 906 }, { "epoch": 0.07108150470219436, "grad_norm": 3.0143215656280518, "learning_rate": 9.993808736411499e-06, "loss": 1.3103, "step": 907 }, { "epoch": 0.07115987460815047, "grad_norm": 5.182043075561523, "learning_rate": 9.993793386182107e-06, "loss": 1.6286, "step": 908 }, { "epoch": 0.07123824451410658, "grad_norm": 5.133011817932129, "learning_rate": 9.993778016958893e-06, "loss": 1.2771, "step": 909 }, { "epoch": 0.0713166144200627, "grad_norm": 2.544740915298462, "learning_rate": 9.99376262874191e-06, "loss": 1.193, "step": 910 }, { "epoch": 0.0713949843260188, "grad_norm": 4.3515777587890625, "learning_rate": 9.993747221531219e-06, "loss": 1.8818, "step": 911 }, { "epoch": 0.07147335423197493, "grad_norm": 2.627758264541626, "learning_rate": 9.99373179532688e-06, "loss": 1.3239, "step": 912 }, { "epoch": 0.07155172413793104, "grad_norm": 2.8996434211730957, "learning_rate": 9.993716350128949e-06, "loss": 1.7851, "step": 913 }, { "epoch": 0.07163009404388715, "grad_norm": 4.040151119232178, "learning_rate": 9.993700885937484e-06, "loss": 1.4219, "step": 914 }, { "epoch": 0.07170846394984326, "grad_norm": 2.926701307296753, "learning_rate": 9.993685402752548e-06, "loss": 1.1535, "step": 915 }, { "epoch": 0.07178683385579937, "grad_norm": 2.668776035308838, "learning_rate": 9.993669900574199e-06, "loss": 1.6214, "step": 916 }, { "epoch": 0.07186520376175548, "grad_norm": 3.0097427368164062, "learning_rate": 9.993654379402493e-06, "loss": 1.5549, "step": 917 }, { "epoch": 0.0719435736677116, "grad_norm": 5.4617228507995605, "learning_rate": 9.99363883923749e-06, "loss": 1.3417, "step": 918 }, { "epoch": 0.07202194357366772, "grad_norm": 4.961665630340576, "learning_rate": 9.99362328007925e-06, "loss": 1.5684, "step": 919 }, { "epoch": 0.07210031347962383, "grad_norm": 4.076306343078613, "learning_rate": 9.993607701927832e-06, "loss": 1.5217, "step": 920 }, { "epoch": 0.07217868338557994, "grad_norm": 5.789394378662109, "learning_rate": 9.993592104783297e-06, "loss": 1.4174, "step": 921 }, { "epoch": 0.07225705329153605, "grad_norm": 4.611825466156006, "learning_rate": 9.9935764886457e-06, "loss": 1.3568, "step": 922 }, { "epoch": 0.07233542319749216, "grad_norm": 2.161283016204834, "learning_rate": 9.993560853515105e-06, "loss": 1.1351, "step": 923 }, { "epoch": 0.07241379310344828, "grad_norm": 2.8477346897125244, "learning_rate": 9.993545199391568e-06, "loss": 1.33, "step": 924 }, { "epoch": 0.07249216300940439, "grad_norm": 2.838064432144165, "learning_rate": 9.993529526275151e-06, "loss": 1.6268, "step": 925 }, { "epoch": 0.0725705329153605, "grad_norm": 7.426792621612549, "learning_rate": 9.993513834165911e-06, "loss": 1.6976, "step": 926 }, { "epoch": 0.07264890282131661, "grad_norm": 4.437683582305908, "learning_rate": 9.99349812306391e-06, "loss": 1.1023, "step": 927 }, { "epoch": 0.07272727272727272, "grad_norm": 4.760648250579834, "learning_rate": 9.993482392969207e-06, "loss": 1.4705, "step": 928 }, { "epoch": 0.07280564263322883, "grad_norm": 4.537635326385498, "learning_rate": 9.993466643881862e-06, "loss": 1.6409, "step": 929 }, { "epoch": 0.07288401253918496, "grad_norm": 3.34883975982666, "learning_rate": 9.993450875801935e-06, "loss": 1.4698, "step": 930 }, { "epoch": 0.07296238244514107, "grad_norm": 3.1179544925689697, "learning_rate": 9.993435088729484e-06, "loss": 1.3358, "step": 931 }, { "epoch": 0.07304075235109718, "grad_norm": 3.532888650894165, "learning_rate": 9.993419282664573e-06, "loss": 1.7021, "step": 932 }, { "epoch": 0.07311912225705329, "grad_norm": 3.7584099769592285, "learning_rate": 9.993403457607257e-06, "loss": 1.3469, "step": 933 }, { "epoch": 0.0731974921630094, "grad_norm": 9.424467086791992, "learning_rate": 9.993387613557602e-06, "loss": 1.8655, "step": 934 }, { "epoch": 0.07327586206896551, "grad_norm": 7.368868827819824, "learning_rate": 9.993371750515663e-06, "loss": 1.1126, "step": 935 }, { "epoch": 0.07335423197492164, "grad_norm": 3.4773311614990234, "learning_rate": 9.993355868481504e-06, "loss": 1.6338, "step": 936 }, { "epoch": 0.07343260188087775, "grad_norm": 5.5621137619018555, "learning_rate": 9.993339967455184e-06, "loss": 1.7095, "step": 937 }, { "epoch": 0.07351097178683386, "grad_norm": 3.7662954330444336, "learning_rate": 9.993324047436763e-06, "loss": 1.9666, "step": 938 }, { "epoch": 0.07358934169278997, "grad_norm": 2.6500093936920166, "learning_rate": 9.993308108426302e-06, "loss": 1.5896, "step": 939 }, { "epoch": 0.07366771159874608, "grad_norm": 3.3653862476348877, "learning_rate": 9.993292150423862e-06, "loss": 1.4009, "step": 940 }, { "epoch": 0.07374608150470219, "grad_norm": 4.112266540527344, "learning_rate": 9.993276173429504e-06, "loss": 1.5252, "step": 941 }, { "epoch": 0.07382445141065831, "grad_norm": 3.899223804473877, "learning_rate": 9.993260177443287e-06, "loss": 1.291, "step": 942 }, { "epoch": 0.07390282131661442, "grad_norm": 3.97039794921875, "learning_rate": 9.993244162465273e-06, "loss": 1.7218, "step": 943 }, { "epoch": 0.07398119122257053, "grad_norm": 7.834506988525391, "learning_rate": 9.993228128495524e-06, "loss": 1.4535, "step": 944 }, { "epoch": 0.07405956112852664, "grad_norm": 3.748962879180908, "learning_rate": 9.993212075534098e-06, "loss": 1.4503, "step": 945 }, { "epoch": 0.07413793103448275, "grad_norm": 4.727917671203613, "learning_rate": 9.99319600358106e-06, "loss": 1.3534, "step": 946 }, { "epoch": 0.07421630094043886, "grad_norm": 3.69296932220459, "learning_rate": 9.993179912636468e-06, "loss": 1.4518, "step": 947 }, { "epoch": 0.07429467084639499, "grad_norm": 3.946981906890869, "learning_rate": 9.993163802700385e-06, "loss": 1.6092, "step": 948 }, { "epoch": 0.0743730407523511, "grad_norm": 8.4278564453125, "learning_rate": 9.993147673772869e-06, "loss": 1.3688, "step": 949 }, { "epoch": 0.07445141065830721, "grad_norm": 3.734281539916992, "learning_rate": 9.993131525853987e-06, "loss": 1.1441, "step": 950 }, { "epoch": 0.07452978056426332, "grad_norm": 3.6313328742980957, "learning_rate": 9.993115358943796e-06, "loss": 1.9378, "step": 951 }, { "epoch": 0.07460815047021943, "grad_norm": 4.107640743255615, "learning_rate": 9.99309917304236e-06, "loss": 1.4408, "step": 952 }, { "epoch": 0.07468652037617556, "grad_norm": 4.187763214111328, "learning_rate": 9.993082968149738e-06, "loss": 1.2301, "step": 953 }, { "epoch": 0.07476489028213167, "grad_norm": 7.555318832397461, "learning_rate": 9.993066744265993e-06, "loss": 1.3184, "step": 954 }, { "epoch": 0.07484326018808778, "grad_norm": 3.1409804821014404, "learning_rate": 9.993050501391188e-06, "loss": 1.4632, "step": 955 }, { "epoch": 0.07492163009404389, "grad_norm": 4.370401859283447, "learning_rate": 9.993034239525382e-06, "loss": 1.497, "step": 956 }, { "epoch": 0.075, "grad_norm": 4.055842399597168, "learning_rate": 9.993017958668638e-06, "loss": 1.7251, "step": 957 }, { "epoch": 0.07507836990595611, "grad_norm": 11.247596740722656, "learning_rate": 9.99300165882102e-06, "loss": 1.1633, "step": 958 }, { "epoch": 0.07515673981191223, "grad_norm": 5.14241361618042, "learning_rate": 9.992985339982585e-06, "loss": 1.5467, "step": 959 }, { "epoch": 0.07523510971786834, "grad_norm": 4.846098899841309, "learning_rate": 9.992969002153403e-06, "loss": 2.0959, "step": 960 }, { "epoch": 0.07531347962382445, "grad_norm": 2.4357798099517822, "learning_rate": 9.99295264533353e-06, "loss": 1.3109, "step": 961 }, { "epoch": 0.07539184952978056, "grad_norm": 6.869080543518066, "learning_rate": 9.992936269523028e-06, "loss": 1.6269, "step": 962 }, { "epoch": 0.07547021943573667, "grad_norm": 7.091264247894287, "learning_rate": 9.992919874721962e-06, "loss": 1.5408, "step": 963 }, { "epoch": 0.07554858934169278, "grad_norm": 4.503604888916016, "learning_rate": 9.992903460930392e-06, "loss": 1.9025, "step": 964 }, { "epoch": 0.07562695924764891, "grad_norm": 4.227804660797119, "learning_rate": 9.992887028148384e-06, "loss": 1.7378, "step": 965 }, { "epoch": 0.07570532915360502, "grad_norm": 3.772351026535034, "learning_rate": 9.992870576375997e-06, "loss": 1.4643, "step": 966 }, { "epoch": 0.07578369905956113, "grad_norm": 11.706591606140137, "learning_rate": 9.992854105613294e-06, "loss": 1.1954, "step": 967 }, { "epoch": 0.07586206896551724, "grad_norm": 2.699357271194458, "learning_rate": 9.99283761586034e-06, "loss": 1.4383, "step": 968 }, { "epoch": 0.07594043887147335, "grad_norm": 4.0413923263549805, "learning_rate": 9.992821107117195e-06, "loss": 1.755, "step": 969 }, { "epoch": 0.07601880877742946, "grad_norm": 2.982168436050415, "learning_rate": 9.992804579383924e-06, "loss": 1.2445, "step": 970 }, { "epoch": 0.07609717868338559, "grad_norm": 5.444549083709717, "learning_rate": 9.992788032660588e-06, "loss": 1.592, "step": 971 }, { "epoch": 0.0761755485893417, "grad_norm": 2.4974184036254883, "learning_rate": 9.992771466947252e-06, "loss": 1.2588, "step": 972 }, { "epoch": 0.0762539184952978, "grad_norm": 3.739619016647339, "learning_rate": 9.992754882243978e-06, "loss": 1.164, "step": 973 }, { "epoch": 0.07633228840125392, "grad_norm": 3.4200618267059326, "learning_rate": 9.992738278550826e-06, "loss": 1.1751, "step": 974 }, { "epoch": 0.07641065830721003, "grad_norm": 3.9113662242889404, "learning_rate": 9.992721655867866e-06, "loss": 1.5848, "step": 975 }, { "epoch": 0.07648902821316614, "grad_norm": 2.4053380489349365, "learning_rate": 9.992705014195155e-06, "loss": 1.341, "step": 976 }, { "epoch": 0.07656739811912226, "grad_norm": 5.31331729888916, "learning_rate": 9.992688353532757e-06, "loss": 1.9655, "step": 977 }, { "epoch": 0.07664576802507837, "grad_norm": 2.6509006023406982, "learning_rate": 9.99267167388074e-06, "loss": 1.1956, "step": 978 }, { "epoch": 0.07672413793103448, "grad_norm": 4.553266525268555, "learning_rate": 9.992654975239162e-06, "loss": 1.5319, "step": 979 }, { "epoch": 0.0768025078369906, "grad_norm": 3.404928207397461, "learning_rate": 9.99263825760809e-06, "loss": 1.6596, "step": 980 }, { "epoch": 0.0768808777429467, "grad_norm": 2.698286294937134, "learning_rate": 9.992621520987586e-06, "loss": 1.4512, "step": 981 }, { "epoch": 0.07695924764890281, "grad_norm": 3.2910985946655273, "learning_rate": 9.992604765377715e-06, "loss": 1.7731, "step": 982 }, { "epoch": 0.07703761755485894, "grad_norm": 5.146914005279541, "learning_rate": 9.992587990778539e-06, "loss": 1.5919, "step": 983 }, { "epoch": 0.07711598746081505, "grad_norm": 3.063673734664917, "learning_rate": 9.992571197190123e-06, "loss": 1.4716, "step": 984 }, { "epoch": 0.07719435736677116, "grad_norm": 8.849552154541016, "learning_rate": 9.992554384612529e-06, "loss": 0.8863, "step": 985 }, { "epoch": 0.07727272727272727, "grad_norm": 2.552546739578247, "learning_rate": 9.992537553045825e-06, "loss": 1.3181, "step": 986 }, { "epoch": 0.07735109717868338, "grad_norm": 3.8068556785583496, "learning_rate": 9.992520702490072e-06, "loss": 1.2417, "step": 987 }, { "epoch": 0.07742946708463949, "grad_norm": 3.2639644145965576, "learning_rate": 9.992503832945332e-06, "loss": 1.6356, "step": 988 }, { "epoch": 0.07750783699059562, "grad_norm": 5.629429340362549, "learning_rate": 9.992486944411675e-06, "loss": 1.3158, "step": 989 }, { "epoch": 0.07758620689655173, "grad_norm": 5.843857288360596, "learning_rate": 9.99247003688916e-06, "loss": 1.5083, "step": 990 }, { "epoch": 0.07766457680250784, "grad_norm": 2.2947590351104736, "learning_rate": 9.992453110377854e-06, "loss": 1.478, "step": 991 }, { "epoch": 0.07774294670846395, "grad_norm": 4.805761337280273, "learning_rate": 9.99243616487782e-06, "loss": 1.5967, "step": 992 }, { "epoch": 0.07782131661442006, "grad_norm": 4.349510669708252, "learning_rate": 9.992419200389125e-06, "loss": 1.5007, "step": 993 }, { "epoch": 0.07789968652037618, "grad_norm": 5.282576560974121, "learning_rate": 9.992402216911829e-06, "loss": 1.4113, "step": 994 }, { "epoch": 0.07797805642633229, "grad_norm": 2.2954659461975098, "learning_rate": 9.992385214446001e-06, "loss": 1.2562, "step": 995 }, { "epoch": 0.0780564263322884, "grad_norm": 3.9144577980041504, "learning_rate": 9.992368192991704e-06, "loss": 1.642, "step": 996 }, { "epoch": 0.07813479623824451, "grad_norm": 3.24090576171875, "learning_rate": 9.992351152549003e-06, "loss": 1.2011, "step": 997 }, { "epoch": 0.07821316614420062, "grad_norm": 6.203169822692871, "learning_rate": 9.99233409311796e-06, "loss": 1.7484, "step": 998 }, { "epoch": 0.07829153605015673, "grad_norm": 3.91560435295105, "learning_rate": 9.992317014698643e-06, "loss": 1.1496, "step": 999 }, { "epoch": 0.07836990595611286, "grad_norm": 6.700224876403809, "learning_rate": 9.992299917291118e-06, "loss": 1.1632, "step": 1000 }, { "epoch": 0.07844827586206897, "grad_norm": 3.5670971870422363, "learning_rate": 9.992282800895447e-06, "loss": 1.5083, "step": 1001 }, { "epoch": 0.07852664576802508, "grad_norm": 2.5297915935516357, "learning_rate": 9.992265665511698e-06, "loss": 1.5829, "step": 1002 }, { "epoch": 0.07860501567398119, "grad_norm": 2.1178648471832275, "learning_rate": 9.992248511139933e-06, "loss": 1.1463, "step": 1003 }, { "epoch": 0.0786833855799373, "grad_norm": 8.441643714904785, "learning_rate": 9.992231337780218e-06, "loss": 1.4189, "step": 1004 }, { "epoch": 0.07876175548589341, "grad_norm": 2.0043976306915283, "learning_rate": 9.992214145432622e-06, "loss": 1.3894, "step": 1005 }, { "epoch": 0.07884012539184954, "grad_norm": 3.197603464126587, "learning_rate": 9.992196934097205e-06, "loss": 0.6682, "step": 1006 }, { "epoch": 0.07891849529780565, "grad_norm": 3.6968300342559814, "learning_rate": 9.992179703774036e-06, "loss": 1.523, "step": 1007 }, { "epoch": 0.07899686520376176, "grad_norm": 5.183149814605713, "learning_rate": 9.99216245446318e-06, "loss": 1.7786, "step": 1008 }, { "epoch": 0.07907523510971787, "grad_norm": 4.6505608558654785, "learning_rate": 9.992145186164701e-06, "loss": 1.1628, "step": 1009 }, { "epoch": 0.07915360501567398, "grad_norm": 3.8711698055267334, "learning_rate": 9.992127898878668e-06, "loss": 1.5244, "step": 1010 }, { "epoch": 0.07923197492163009, "grad_norm": 3.645777702331543, "learning_rate": 9.992110592605142e-06, "loss": 1.5705, "step": 1011 }, { "epoch": 0.07931034482758621, "grad_norm": 3.538872003555298, "learning_rate": 9.992093267344193e-06, "loss": 1.2535, "step": 1012 }, { "epoch": 0.07938871473354232, "grad_norm": 3.7113852500915527, "learning_rate": 9.992075923095885e-06, "loss": 1.3757, "step": 1013 }, { "epoch": 0.07946708463949843, "grad_norm": 4.169156551361084, "learning_rate": 9.992058559860284e-06, "loss": 1.4095, "step": 1014 }, { "epoch": 0.07954545454545454, "grad_norm": 2.950127363204956, "learning_rate": 9.992041177637457e-06, "loss": 1.5688, "step": 1015 }, { "epoch": 0.07962382445141065, "grad_norm": 2.943279266357422, "learning_rate": 9.992023776427469e-06, "loss": 1.6824, "step": 1016 }, { "epoch": 0.07970219435736677, "grad_norm": 11.036295890808105, "learning_rate": 9.992006356230385e-06, "loss": 1.1851, "step": 1017 }, { "epoch": 0.07978056426332289, "grad_norm": 2.5451133251190186, "learning_rate": 9.991988917046276e-06, "loss": 1.4589, "step": 1018 }, { "epoch": 0.079858934169279, "grad_norm": 3.117918014526367, "learning_rate": 9.991971458875203e-06, "loss": 1.3125, "step": 1019 }, { "epoch": 0.07993730407523511, "grad_norm": 3.4612488746643066, "learning_rate": 9.991953981717235e-06, "loss": 1.4206, "step": 1020 }, { "epoch": 0.08001567398119122, "grad_norm": 2.2174437046051025, "learning_rate": 9.991936485572439e-06, "loss": 1.3957, "step": 1021 }, { "epoch": 0.08009404388714733, "grad_norm": 4.263909816741943, "learning_rate": 9.99191897044088e-06, "loss": 1.4695, "step": 1022 }, { "epoch": 0.08017241379310344, "grad_norm": 3.4534618854522705, "learning_rate": 9.991901436322624e-06, "loss": 1.294, "step": 1023 }, { "epoch": 0.08025078369905957, "grad_norm": 3.6934359073638916, "learning_rate": 9.991883883217741e-06, "loss": 1.4143, "step": 1024 }, { "epoch": 0.08032915360501568, "grad_norm": 3.247445583343506, "learning_rate": 9.991866311126295e-06, "loss": 1.2504, "step": 1025 }, { "epoch": 0.08040752351097179, "grad_norm": 11.60682487487793, "learning_rate": 9.991848720048352e-06, "loss": 0.7028, "step": 1026 }, { "epoch": 0.0804858934169279, "grad_norm": 2.0627293586730957, "learning_rate": 9.991831109983982e-06, "loss": 1.2242, "step": 1027 }, { "epoch": 0.08056426332288401, "grad_norm": 4.924439907073975, "learning_rate": 9.991813480933251e-06, "loss": 1.627, "step": 1028 }, { "epoch": 0.08064263322884012, "grad_norm": 3.1315786838531494, "learning_rate": 9.991795832896224e-06, "loss": 1.6401, "step": 1029 }, { "epoch": 0.08072100313479624, "grad_norm": 2.9115099906921387, "learning_rate": 9.991778165872973e-06, "loss": 1.4129, "step": 1030 }, { "epoch": 0.08079937304075235, "grad_norm": 3.056459903717041, "learning_rate": 9.991760479863557e-06, "loss": 1.5612, "step": 1031 }, { "epoch": 0.08087774294670846, "grad_norm": 3.1248738765716553, "learning_rate": 9.99174277486805e-06, "loss": 1.2507, "step": 1032 }, { "epoch": 0.08095611285266457, "grad_norm": 3.879556655883789, "learning_rate": 9.991725050886519e-06, "loss": 1.5702, "step": 1033 }, { "epoch": 0.08103448275862069, "grad_norm": 2.4927968978881836, "learning_rate": 9.991707307919028e-06, "loss": 1.467, "step": 1034 }, { "epoch": 0.08111285266457681, "grad_norm": 2.313889741897583, "learning_rate": 9.991689545965646e-06, "loss": 1.4342, "step": 1035 }, { "epoch": 0.08119122257053292, "grad_norm": 12.842498779296875, "learning_rate": 9.991671765026441e-06, "loss": 1.5427, "step": 1036 }, { "epoch": 0.08126959247648903, "grad_norm": 3.067763566970825, "learning_rate": 9.991653965101482e-06, "loss": 1.4541, "step": 1037 }, { "epoch": 0.08134796238244514, "grad_norm": 4.493299961090088, "learning_rate": 9.991636146190834e-06, "loss": 1.1667, "step": 1038 }, { "epoch": 0.08142633228840125, "grad_norm": 2.834503173828125, "learning_rate": 9.991618308294567e-06, "loss": 1.3508, "step": 1039 }, { "epoch": 0.08150470219435736, "grad_norm": 4.13846492767334, "learning_rate": 9.991600451412746e-06, "loss": 1.5209, "step": 1040 }, { "epoch": 0.08158307210031349, "grad_norm": 5.078220367431641, "learning_rate": 9.991582575545442e-06, "loss": 1.5686, "step": 1041 }, { "epoch": 0.0816614420062696, "grad_norm": 3.6093153953552246, "learning_rate": 9.991564680692721e-06, "loss": 1.5807, "step": 1042 }, { "epoch": 0.08173981191222571, "grad_norm": 3.0131170749664307, "learning_rate": 9.991546766854653e-06, "loss": 1.2688, "step": 1043 }, { "epoch": 0.08181818181818182, "grad_norm": 2.8298001289367676, "learning_rate": 9.991528834031304e-06, "loss": 1.7436, "step": 1044 }, { "epoch": 0.08189655172413793, "grad_norm": 3.2754416465759277, "learning_rate": 9.991510882222743e-06, "loss": 1.4106, "step": 1045 }, { "epoch": 0.08197492163009404, "grad_norm": 4.077906131744385, "learning_rate": 9.99149291142904e-06, "loss": 1.4079, "step": 1046 }, { "epoch": 0.08205329153605016, "grad_norm": 7.455482482910156, "learning_rate": 9.99147492165026e-06, "loss": 1.9137, "step": 1047 }, { "epoch": 0.08213166144200627, "grad_norm": 3.8962132930755615, "learning_rate": 9.991456912886473e-06, "loss": 1.4455, "step": 1048 }, { "epoch": 0.08221003134796238, "grad_norm": 12.83207893371582, "learning_rate": 9.991438885137747e-06, "loss": 1.597, "step": 1049 }, { "epoch": 0.0822884012539185, "grad_norm": 4.0568528175354, "learning_rate": 9.991420838404153e-06, "loss": 2.007, "step": 1050 }, { "epoch": 0.0823667711598746, "grad_norm": 2.271040439605713, "learning_rate": 9.991402772685758e-06, "loss": 1.3668, "step": 1051 }, { "epoch": 0.08244514106583072, "grad_norm": 3.0501708984375, "learning_rate": 9.99138468798263e-06, "loss": 1.7326, "step": 1052 }, { "epoch": 0.08252351097178684, "grad_norm": 3.1162211894989014, "learning_rate": 9.991366584294839e-06, "loss": 1.178, "step": 1053 }, { "epoch": 0.08260188087774295, "grad_norm": 4.320234298706055, "learning_rate": 9.991348461622451e-06, "loss": 1.7578, "step": 1054 }, { "epoch": 0.08268025078369906, "grad_norm": 2.8827221393585205, "learning_rate": 9.99133031996554e-06, "loss": 1.4557, "step": 1055 }, { "epoch": 0.08275862068965517, "grad_norm": 3.7864768505096436, "learning_rate": 9.991312159324171e-06, "loss": 2.0932, "step": 1056 }, { "epoch": 0.08283699059561128, "grad_norm": 2.5161564350128174, "learning_rate": 9.991293979698415e-06, "loss": 1.7775, "step": 1057 }, { "epoch": 0.08291536050156739, "grad_norm": 3.238149404525757, "learning_rate": 9.99127578108834e-06, "loss": 0.9181, "step": 1058 }, { "epoch": 0.08299373040752352, "grad_norm": 3.2722299098968506, "learning_rate": 9.991257563494014e-06, "loss": 1.4825, "step": 1059 }, { "epoch": 0.08307210031347963, "grad_norm": 3.6327266693115234, "learning_rate": 9.991239326915509e-06, "loss": 1.8927, "step": 1060 }, { "epoch": 0.08315047021943574, "grad_norm": 6.546268463134766, "learning_rate": 9.991221071352894e-06, "loss": 1.6312, "step": 1061 }, { "epoch": 0.08322884012539185, "grad_norm": 2.771446943283081, "learning_rate": 9.991202796806237e-06, "loss": 1.4303, "step": 1062 }, { "epoch": 0.08330721003134796, "grad_norm": 3.457906723022461, "learning_rate": 9.991184503275608e-06, "loss": 1.2428, "step": 1063 }, { "epoch": 0.08338557993730407, "grad_norm": 14.115056037902832, "learning_rate": 9.991166190761078e-06, "loss": 0.7783, "step": 1064 }, { "epoch": 0.0834639498432602, "grad_norm": 2.9454126358032227, "learning_rate": 9.991147859262714e-06, "loss": 1.3405, "step": 1065 }, { "epoch": 0.0835423197492163, "grad_norm": 6.316283226013184, "learning_rate": 9.991129508780588e-06, "loss": 1.3056, "step": 1066 }, { "epoch": 0.08362068965517241, "grad_norm": 6.169098377227783, "learning_rate": 9.991111139314768e-06, "loss": 1.4791, "step": 1067 }, { "epoch": 0.08369905956112852, "grad_norm": 3.549154043197632, "learning_rate": 9.991092750865326e-06, "loss": 1.4558, "step": 1068 }, { "epoch": 0.08377742946708464, "grad_norm": 11.178854942321777, "learning_rate": 9.99107434343233e-06, "loss": 1.6203, "step": 1069 }, { "epoch": 0.08385579937304075, "grad_norm": 2.984226942062378, "learning_rate": 9.991055917015851e-06, "loss": 1.2733, "step": 1070 }, { "epoch": 0.08393416927899687, "grad_norm": 7.3797173500061035, "learning_rate": 9.99103747161596e-06, "loss": 1.5319, "step": 1071 }, { "epoch": 0.08401253918495298, "grad_norm": 7.410616397857666, "learning_rate": 9.991019007232725e-06, "loss": 1.5257, "step": 1072 }, { "epoch": 0.08409090909090909, "grad_norm": 4.127233982086182, "learning_rate": 9.991000523866217e-06, "loss": 1.9015, "step": 1073 }, { "epoch": 0.0841692789968652, "grad_norm": 7.815873622894287, "learning_rate": 9.990982021516506e-06, "loss": 1.2974, "step": 1074 }, { "epoch": 0.08424764890282131, "grad_norm": 2.5668652057647705, "learning_rate": 9.990963500183664e-06, "loss": 1.5071, "step": 1075 }, { "epoch": 0.08432601880877742, "grad_norm": 2.4939093589782715, "learning_rate": 9.99094495986776e-06, "loss": 1.7467, "step": 1076 }, { "epoch": 0.08440438871473355, "grad_norm": 3.1844167709350586, "learning_rate": 9.990926400568866e-06, "loss": 1.8163, "step": 1077 }, { "epoch": 0.08448275862068966, "grad_norm": 5.109973907470703, "learning_rate": 9.99090782228705e-06, "loss": 1.0502, "step": 1078 }, { "epoch": 0.08456112852664577, "grad_norm": 2.394277334213257, "learning_rate": 9.990889225022384e-06, "loss": 1.3342, "step": 1079 }, { "epoch": 0.08463949843260188, "grad_norm": 2.8680481910705566, "learning_rate": 9.990870608774938e-06, "loss": 1.3938, "step": 1080 }, { "epoch": 0.08471786833855799, "grad_norm": 4.142158031463623, "learning_rate": 9.990851973544786e-06, "loss": 1.7226, "step": 1081 }, { "epoch": 0.08479623824451411, "grad_norm": 8.125496864318848, "learning_rate": 9.990833319331997e-06, "loss": 1.3942, "step": 1082 }, { "epoch": 0.08487460815047022, "grad_norm": 4.447039604187012, "learning_rate": 9.990814646136638e-06, "loss": 1.5234, "step": 1083 }, { "epoch": 0.08495297805642633, "grad_norm": 2.702619791030884, "learning_rate": 9.990795953958786e-06, "loss": 1.3473, "step": 1084 }, { "epoch": 0.08503134796238244, "grad_norm": 4.264716148376465, "learning_rate": 9.99077724279851e-06, "loss": 1.4114, "step": 1085 }, { "epoch": 0.08510971786833856, "grad_norm": 3.313250780105591, "learning_rate": 9.990758512655881e-06, "loss": 1.3408, "step": 1086 }, { "epoch": 0.08518808777429467, "grad_norm": 6.259039878845215, "learning_rate": 9.990739763530968e-06, "loss": 0.4169, "step": 1087 }, { "epoch": 0.08526645768025079, "grad_norm": 3.3826093673706055, "learning_rate": 9.990720995423845e-06, "loss": 1.4644, "step": 1088 }, { "epoch": 0.0853448275862069, "grad_norm": 3.3442602157592773, "learning_rate": 9.990702208334583e-06, "loss": 1.4454, "step": 1089 }, { "epoch": 0.08542319749216301, "grad_norm": 4.289291858673096, "learning_rate": 9.990683402263254e-06, "loss": 0.9652, "step": 1090 }, { "epoch": 0.08550156739811912, "grad_norm": 3.6896140575408936, "learning_rate": 9.990664577209928e-06, "loss": 1.5576, "step": 1091 }, { "epoch": 0.08557993730407523, "grad_norm": 5.1214470863342285, "learning_rate": 9.990645733174677e-06, "loss": 1.5225, "step": 1092 }, { "epoch": 0.08565830721003134, "grad_norm": 7.19219446182251, "learning_rate": 9.990626870157574e-06, "loss": 1.7679, "step": 1093 }, { "epoch": 0.08573667711598747, "grad_norm": 3.313647747039795, "learning_rate": 9.990607988158689e-06, "loss": 1.2878, "step": 1094 }, { "epoch": 0.08581504702194358, "grad_norm": 3.2428677082061768, "learning_rate": 9.990589087178093e-06, "loss": 1.5816, "step": 1095 }, { "epoch": 0.08589341692789969, "grad_norm": 3.930399179458618, "learning_rate": 9.990570167215863e-06, "loss": 1.6651, "step": 1096 }, { "epoch": 0.0859717868338558, "grad_norm": 8.965780258178711, "learning_rate": 9.990551228272066e-06, "loss": 1.0707, "step": 1097 }, { "epoch": 0.08605015673981191, "grad_norm": 4.562810897827148, "learning_rate": 9.990532270346773e-06, "loss": 1.4366, "step": 1098 }, { "epoch": 0.08612852664576802, "grad_norm": 4.495123863220215, "learning_rate": 9.990513293440061e-06, "loss": 1.4396, "step": 1099 }, { "epoch": 0.08620689655172414, "grad_norm": 2.7299673557281494, "learning_rate": 9.990494297551999e-06, "loss": 1.6173, "step": 1100 }, { "epoch": 0.08628526645768025, "grad_norm": 5.043631553649902, "learning_rate": 9.99047528268266e-06, "loss": 1.6798, "step": 1101 }, { "epoch": 0.08636363636363636, "grad_norm": 2.228766918182373, "learning_rate": 9.990456248832117e-06, "loss": 1.3396, "step": 1102 }, { "epoch": 0.08644200626959248, "grad_norm": 2.7343146800994873, "learning_rate": 9.99043719600044e-06, "loss": 1.445, "step": 1103 }, { "epoch": 0.08652037617554859, "grad_norm": 5.281085014343262, "learning_rate": 9.990418124187704e-06, "loss": 1.5934, "step": 1104 }, { "epoch": 0.0865987460815047, "grad_norm": 3.993129253387451, "learning_rate": 9.990399033393983e-06, "loss": 0.9232, "step": 1105 }, { "epoch": 0.08667711598746082, "grad_norm": 3.5328471660614014, "learning_rate": 9.990379923619345e-06, "loss": 1.2862, "step": 1106 }, { "epoch": 0.08675548589341693, "grad_norm": 2.8833093643188477, "learning_rate": 9.990360794863865e-06, "loss": 1.6361, "step": 1107 }, { "epoch": 0.08683385579937304, "grad_norm": 3.113368511199951, "learning_rate": 9.990341647127616e-06, "loss": 0.9548, "step": 1108 }, { "epoch": 0.08691222570532915, "grad_norm": 5.589105606079102, "learning_rate": 9.99032248041067e-06, "loss": 1.3469, "step": 1109 }, { "epoch": 0.08699059561128526, "grad_norm": 2.7647786140441895, "learning_rate": 9.990303294713101e-06, "loss": 1.1837, "step": 1110 }, { "epoch": 0.08706896551724137, "grad_norm": 3.6072323322296143, "learning_rate": 9.990284090034984e-06, "loss": 1.5524, "step": 1111 }, { "epoch": 0.0871473354231975, "grad_norm": 2.581624984741211, "learning_rate": 9.990264866376384e-06, "loss": 1.6753, "step": 1112 }, { "epoch": 0.08722570532915361, "grad_norm": 5.42243766784668, "learning_rate": 9.990245623737384e-06, "loss": 1.7234, "step": 1113 }, { "epoch": 0.08730407523510972, "grad_norm": 6.927798271179199, "learning_rate": 9.990226362118051e-06, "loss": 1.4231, "step": 1114 }, { "epoch": 0.08738244514106583, "grad_norm": 3.0384521484375, "learning_rate": 9.990207081518461e-06, "loss": 1.4367, "step": 1115 }, { "epoch": 0.08746081504702194, "grad_norm": 3.6261539459228516, "learning_rate": 9.990187781938686e-06, "loss": 1.5553, "step": 1116 }, { "epoch": 0.08753918495297805, "grad_norm": 4.044138431549072, "learning_rate": 9.9901684633788e-06, "loss": 1.1645, "step": 1117 }, { "epoch": 0.08761755485893417, "grad_norm": 3.787003517150879, "learning_rate": 9.990149125838876e-06, "loss": 1.5994, "step": 1118 }, { "epoch": 0.08769592476489028, "grad_norm": 3.7199673652648926, "learning_rate": 9.990129769318988e-06, "loss": 1.532, "step": 1119 }, { "epoch": 0.0877742946708464, "grad_norm": 3.272542715072632, "learning_rate": 9.990110393819207e-06, "loss": 1.1017, "step": 1120 }, { "epoch": 0.0878526645768025, "grad_norm": 2.8597564697265625, "learning_rate": 9.990090999339612e-06, "loss": 1.6806, "step": 1121 }, { "epoch": 0.08793103448275862, "grad_norm": 3.725233793258667, "learning_rate": 9.990071585880274e-06, "loss": 1.3566, "step": 1122 }, { "epoch": 0.08800940438871474, "grad_norm": 3.348428726196289, "learning_rate": 9.990052153441267e-06, "loss": 0.9544, "step": 1123 }, { "epoch": 0.08808777429467085, "grad_norm": 3.2805287837982178, "learning_rate": 9.990032702022664e-06, "loss": 1.2836, "step": 1124 }, { "epoch": 0.08816614420062696, "grad_norm": 2.921738624572754, "learning_rate": 9.990013231624537e-06, "loss": 1.2231, "step": 1125 }, { "epoch": 0.08824451410658307, "grad_norm": 4.068424224853516, "learning_rate": 9.989993742246966e-06, "loss": 1.228, "step": 1126 }, { "epoch": 0.08832288401253918, "grad_norm": 3.264582633972168, "learning_rate": 9.98997423389002e-06, "loss": 1.5303, "step": 1127 }, { "epoch": 0.08840125391849529, "grad_norm": 2.5538129806518555, "learning_rate": 9.989954706553776e-06, "loss": 1.3214, "step": 1128 }, { "epoch": 0.08847962382445142, "grad_norm": 4.823193550109863, "learning_rate": 9.989935160238308e-06, "loss": 1.8026, "step": 1129 }, { "epoch": 0.08855799373040753, "grad_norm": 3.546360492706299, "learning_rate": 9.989915594943689e-06, "loss": 1.3902, "step": 1130 }, { "epoch": 0.08863636363636364, "grad_norm": 2.6425726413726807, "learning_rate": 9.989896010669994e-06, "loss": 1.3633, "step": 1131 }, { "epoch": 0.08871473354231975, "grad_norm": 3.7612900733947754, "learning_rate": 9.989876407417298e-06, "loss": 1.1893, "step": 1132 }, { "epoch": 0.08879310344827586, "grad_norm": 4.5097126960754395, "learning_rate": 9.989856785185675e-06, "loss": 1.7154, "step": 1133 }, { "epoch": 0.08887147335423197, "grad_norm": 2.7595059871673584, "learning_rate": 9.989837143975198e-06, "loss": 1.4675, "step": 1134 }, { "epoch": 0.0889498432601881, "grad_norm": 2.642768383026123, "learning_rate": 9.989817483785946e-06, "loss": 1.5753, "step": 1135 }, { "epoch": 0.0890282131661442, "grad_norm": 6.1398606300354, "learning_rate": 9.989797804617988e-06, "loss": 1.2161, "step": 1136 }, { "epoch": 0.08910658307210031, "grad_norm": 3.1034510135650635, "learning_rate": 9.989778106471406e-06, "loss": 1.7481, "step": 1137 }, { "epoch": 0.08918495297805643, "grad_norm": 3.5682222843170166, "learning_rate": 9.989758389346269e-06, "loss": 1.7134, "step": 1138 }, { "epoch": 0.08926332288401254, "grad_norm": 8.625456809997559, "learning_rate": 9.989738653242654e-06, "loss": 1.7654, "step": 1139 }, { "epoch": 0.08934169278996865, "grad_norm": 3.6200289726257324, "learning_rate": 9.989718898160636e-06, "loss": 1.4899, "step": 1140 }, { "epoch": 0.08942006269592477, "grad_norm": 3.4835586547851562, "learning_rate": 9.98969912410029e-06, "loss": 1.2466, "step": 1141 }, { "epoch": 0.08949843260188088, "grad_norm": 2.6823601722717285, "learning_rate": 9.989679331061692e-06, "loss": 1.5075, "step": 1142 }, { "epoch": 0.08957680250783699, "grad_norm": 2.278564691543579, "learning_rate": 9.989659519044917e-06, "loss": 1.5238, "step": 1143 }, { "epoch": 0.0896551724137931, "grad_norm": 4.624176502227783, "learning_rate": 9.989639688050038e-06, "loss": 1.2745, "step": 1144 }, { "epoch": 0.08973354231974921, "grad_norm": 4.517180442810059, "learning_rate": 9.989619838077134e-06, "loss": 1.638, "step": 1145 }, { "epoch": 0.08981191222570532, "grad_norm": 2.990661859512329, "learning_rate": 9.989599969126281e-06, "loss": 1.8617, "step": 1146 }, { "epoch": 0.08989028213166145, "grad_norm": 3.678753137588501, "learning_rate": 9.98958008119755e-06, "loss": 1.6558, "step": 1147 }, { "epoch": 0.08996865203761756, "grad_norm": 2.7664313316345215, "learning_rate": 9.989560174291019e-06, "loss": 1.3173, "step": 1148 }, { "epoch": 0.09004702194357367, "grad_norm": 3.6807219982147217, "learning_rate": 9.989540248406763e-06, "loss": 0.9231, "step": 1149 }, { "epoch": 0.09012539184952978, "grad_norm": 6.896298408508301, "learning_rate": 9.989520303544861e-06, "loss": 1.3309, "step": 1150 }, { "epoch": 0.09020376175548589, "grad_norm": 3.104125738143921, "learning_rate": 9.989500339705386e-06, "loss": 1.4379, "step": 1151 }, { "epoch": 0.090282131661442, "grad_norm": 3.790236234664917, "learning_rate": 9.989480356888415e-06, "loss": 1.47, "step": 1152 }, { "epoch": 0.09036050156739812, "grad_norm": 5.438091278076172, "learning_rate": 9.989460355094022e-06, "loss": 1.7786, "step": 1153 }, { "epoch": 0.09043887147335423, "grad_norm": 3.155688524246216, "learning_rate": 9.989440334322284e-06, "loss": 1.4143, "step": 1154 }, { "epoch": 0.09051724137931035, "grad_norm": 3.386768341064453, "learning_rate": 9.98942029457328e-06, "loss": 1.1679, "step": 1155 }, { "epoch": 0.09059561128526646, "grad_norm": 4.446099758148193, "learning_rate": 9.989400235847083e-06, "loss": 1.6189, "step": 1156 }, { "epoch": 0.09067398119122257, "grad_norm": 3.8436872959136963, "learning_rate": 9.98938015814377e-06, "loss": 1.8165, "step": 1157 }, { "epoch": 0.09075235109717868, "grad_norm": 9.486529350280762, "learning_rate": 9.989360061463418e-06, "loss": 1.7063, "step": 1158 }, { "epoch": 0.0908307210031348, "grad_norm": 3.1407573223114014, "learning_rate": 9.989339945806102e-06, "loss": 1.6593, "step": 1159 }, { "epoch": 0.09090909090909091, "grad_norm": 2.316798686981201, "learning_rate": 9.9893198111719e-06, "loss": 1.6866, "step": 1160 }, { "epoch": 0.09098746081504702, "grad_norm": 3.322309970855713, "learning_rate": 9.98929965756089e-06, "loss": 1.1757, "step": 1161 }, { "epoch": 0.09106583072100313, "grad_norm": 3.6657285690307617, "learning_rate": 9.989279484973144e-06, "loss": 1.1971, "step": 1162 }, { "epoch": 0.09114420062695924, "grad_norm": 3.9690754413604736, "learning_rate": 9.989259293408742e-06, "loss": 1.4353, "step": 1163 }, { "epoch": 0.09122257053291537, "grad_norm": 5.205148696899414, "learning_rate": 9.989239082867762e-06, "loss": 1.4917, "step": 1164 }, { "epoch": 0.09130094043887148, "grad_norm": 2.8213398456573486, "learning_rate": 9.989218853350278e-06, "loss": 1.5327, "step": 1165 }, { "epoch": 0.09137931034482759, "grad_norm": 6.3274126052856445, "learning_rate": 9.989198604856367e-06, "loss": 1.6395, "step": 1166 }, { "epoch": 0.0914576802507837, "grad_norm": 2.219956874847412, "learning_rate": 9.98917833738611e-06, "loss": 1.1947, "step": 1167 }, { "epoch": 0.09153605015673981, "grad_norm": 2.196936845779419, "learning_rate": 9.989158050939578e-06, "loss": 1.4351, "step": 1168 }, { "epoch": 0.09161442006269592, "grad_norm": 3.636730909347534, "learning_rate": 9.989137745516853e-06, "loss": 1.5747, "step": 1169 }, { "epoch": 0.09169278996865204, "grad_norm": 4.712217330932617, "learning_rate": 9.98911742111801e-06, "loss": 1.3786, "step": 1170 }, { "epoch": 0.09177115987460815, "grad_norm": 3.934715747833252, "learning_rate": 9.98909707774313e-06, "loss": 1.3464, "step": 1171 }, { "epoch": 0.09184952978056427, "grad_norm": 3.829592227935791, "learning_rate": 9.989076715392282e-06, "loss": 1.6837, "step": 1172 }, { "epoch": 0.09192789968652038, "grad_norm": 2.146516799926758, "learning_rate": 9.989056334065553e-06, "loss": 1.2839, "step": 1173 }, { "epoch": 0.09200626959247649, "grad_norm": 5.669703483581543, "learning_rate": 9.989035933763013e-06, "loss": 1.8395, "step": 1174 }, { "epoch": 0.0920846394984326, "grad_norm": 4.464505672454834, "learning_rate": 9.989015514484746e-06, "loss": 1.6573, "step": 1175 }, { "epoch": 0.09216300940438872, "grad_norm": 3.6126668453216553, "learning_rate": 9.988995076230824e-06, "loss": 1.721, "step": 1176 }, { "epoch": 0.09224137931034483, "grad_norm": 4.3076629638671875, "learning_rate": 9.988974619001328e-06, "loss": 1.4533, "step": 1177 }, { "epoch": 0.09231974921630094, "grad_norm": 7.969771862030029, "learning_rate": 9.988954142796336e-06, "loss": 1.7239, "step": 1178 }, { "epoch": 0.09239811912225705, "grad_norm": 5.011134147644043, "learning_rate": 9.988933647615923e-06, "loss": 1.6067, "step": 1179 }, { "epoch": 0.09247648902821316, "grad_norm": 3.797642707824707, "learning_rate": 9.98891313346017e-06, "loss": 1.5608, "step": 1180 }, { "epoch": 0.09255485893416927, "grad_norm": 4.617431163787842, "learning_rate": 9.988892600329155e-06, "loss": 1.2758, "step": 1181 }, { "epoch": 0.0926332288401254, "grad_norm": 4.029458522796631, "learning_rate": 9.988872048222955e-06, "loss": 1.4319, "step": 1182 }, { "epoch": 0.09271159874608151, "grad_norm": 2.399418830871582, "learning_rate": 9.988851477141646e-06, "loss": 1.2723, "step": 1183 }, { "epoch": 0.09278996865203762, "grad_norm": 2.897503614425659, "learning_rate": 9.98883088708531e-06, "loss": 1.2623, "step": 1184 }, { "epoch": 0.09286833855799373, "grad_norm": 5.1097493171691895, "learning_rate": 9.988810278054022e-06, "loss": 2.1551, "step": 1185 }, { "epoch": 0.09294670846394984, "grad_norm": 3.192941427230835, "learning_rate": 9.988789650047864e-06, "loss": 1.4331, "step": 1186 }, { "epoch": 0.09302507836990595, "grad_norm": 3.173779249191284, "learning_rate": 9.988769003066912e-06, "loss": 1.4562, "step": 1187 }, { "epoch": 0.09310344827586207, "grad_norm": 2.8843772411346436, "learning_rate": 9.988748337111245e-06, "loss": 1.3655, "step": 1188 }, { "epoch": 0.09318181818181819, "grad_norm": 3.2817392349243164, "learning_rate": 9.988727652180943e-06, "loss": 1.174, "step": 1189 }, { "epoch": 0.0932601880877743, "grad_norm": 4.368724822998047, "learning_rate": 9.988706948276081e-06, "loss": 1.6566, "step": 1190 }, { "epoch": 0.0933385579937304, "grad_norm": 5.998810768127441, "learning_rate": 9.988686225396743e-06, "loss": 1.3507, "step": 1191 }, { "epoch": 0.09341692789968652, "grad_norm": 2.485208511352539, "learning_rate": 9.988665483543002e-06, "loss": 1.3575, "step": 1192 }, { "epoch": 0.09349529780564263, "grad_norm": 2.4257733821868896, "learning_rate": 9.988644722714941e-06, "loss": 1.3087, "step": 1193 }, { "epoch": 0.09357366771159875, "grad_norm": 3.4506523609161377, "learning_rate": 9.98862394291264e-06, "loss": 1.2883, "step": 1194 }, { "epoch": 0.09365203761755486, "grad_norm": 4.397739410400391, "learning_rate": 9.988603144136172e-06, "loss": 1.2523, "step": 1195 }, { "epoch": 0.09373040752351097, "grad_norm": 3.734694719314575, "learning_rate": 9.98858232638562e-06, "loss": 1.4454, "step": 1196 }, { "epoch": 0.09380877742946708, "grad_norm": 2.182338237762451, "learning_rate": 9.988561489661066e-06, "loss": 1.1437, "step": 1197 }, { "epoch": 0.0938871473354232, "grad_norm": 3.038083791732788, "learning_rate": 9.988540633962586e-06, "loss": 1.3649, "step": 1198 }, { "epoch": 0.0939655172413793, "grad_norm": 6.3616461753845215, "learning_rate": 9.988519759290259e-06, "loss": 1.776, "step": 1199 }, { "epoch": 0.09404388714733543, "grad_norm": 2.9401910305023193, "learning_rate": 9.988498865644166e-06, "loss": 1.3676, "step": 1200 }, { "epoch": 0.09412225705329154, "grad_norm": 4.0268168449401855, "learning_rate": 9.988477953024382e-06, "loss": 1.7914, "step": 1201 }, { "epoch": 0.09420062695924765, "grad_norm": 11.336615562438965, "learning_rate": 9.988457021430994e-06, "loss": 1.6001, "step": 1202 }, { "epoch": 0.09427899686520376, "grad_norm": 3.2729740142822266, "learning_rate": 9.988436070864077e-06, "loss": 1.5128, "step": 1203 }, { "epoch": 0.09435736677115987, "grad_norm": 2.742685317993164, "learning_rate": 9.98841510132371e-06, "loss": 1.2057, "step": 1204 }, { "epoch": 0.09443573667711598, "grad_norm": 2.689631700515747, "learning_rate": 9.988394112809975e-06, "loss": 1.4983, "step": 1205 }, { "epoch": 0.0945141065830721, "grad_norm": 2.821439504623413, "learning_rate": 9.98837310532295e-06, "loss": 1.2289, "step": 1206 }, { "epoch": 0.09459247648902822, "grad_norm": 3.9998395442962646, "learning_rate": 9.988352078862717e-06, "loss": 1.491, "step": 1207 }, { "epoch": 0.09467084639498433, "grad_norm": 2.392380475997925, "learning_rate": 9.988331033429357e-06, "loss": 1.6749, "step": 1208 }, { "epoch": 0.09474921630094044, "grad_norm": 4.218817710876465, "learning_rate": 9.988309969022944e-06, "loss": 1.6362, "step": 1209 }, { "epoch": 0.09482758620689655, "grad_norm": 3.8356881141662598, "learning_rate": 9.988288885643565e-06, "loss": 1.1153, "step": 1210 }, { "epoch": 0.09490595611285267, "grad_norm": 3.3877923488616943, "learning_rate": 9.988267783291294e-06, "loss": 1.5101, "step": 1211 }, { "epoch": 0.09498432601880878, "grad_norm": 4.232354640960693, "learning_rate": 9.988246661966217e-06, "loss": 1.6497, "step": 1212 }, { "epoch": 0.09506269592476489, "grad_norm": 3.5964293479919434, "learning_rate": 9.98822552166841e-06, "loss": 1.2079, "step": 1213 }, { "epoch": 0.095141065830721, "grad_norm": 3.5325653553009033, "learning_rate": 9.988204362397957e-06, "loss": 1.4351, "step": 1214 }, { "epoch": 0.09521943573667711, "grad_norm": 3.847884178161621, "learning_rate": 9.988183184154938e-06, "loss": 1.2927, "step": 1215 }, { "epoch": 0.09529780564263322, "grad_norm": 3.0850188732147217, "learning_rate": 9.98816198693943e-06, "loss": 0.9143, "step": 1216 }, { "epoch": 0.09537617554858935, "grad_norm": 6.022005558013916, "learning_rate": 9.988140770751516e-06, "loss": 1.3901, "step": 1217 }, { "epoch": 0.09545454545454546, "grad_norm": 4.246365547180176, "learning_rate": 9.988119535591276e-06, "loss": 1.7816, "step": 1218 }, { "epoch": 0.09553291536050157, "grad_norm": 5.223204612731934, "learning_rate": 9.988098281458794e-06, "loss": 1.5028, "step": 1219 }, { "epoch": 0.09561128526645768, "grad_norm": 3.920497417449951, "learning_rate": 9.988077008354146e-06, "loss": 1.2982, "step": 1220 }, { "epoch": 0.09568965517241379, "grad_norm": 5.979863166809082, "learning_rate": 9.988055716277417e-06, "loss": 1.1881, "step": 1221 }, { "epoch": 0.0957680250783699, "grad_norm": 2.900714159011841, "learning_rate": 9.988034405228683e-06, "loss": 1.4338, "step": 1222 }, { "epoch": 0.09584639498432602, "grad_norm": 4.562904357910156, "learning_rate": 9.988013075208029e-06, "loss": 1.8457, "step": 1223 }, { "epoch": 0.09592476489028214, "grad_norm": 5.143942356109619, "learning_rate": 9.987991726215537e-06, "loss": 1.4007, "step": 1224 }, { "epoch": 0.09600313479623825, "grad_norm": 2.376072883605957, "learning_rate": 9.987970358251285e-06, "loss": 1.1789, "step": 1225 }, { "epoch": 0.09608150470219436, "grad_norm": 5.673533916473389, "learning_rate": 9.987948971315357e-06, "loss": 1.5767, "step": 1226 }, { "epoch": 0.09615987460815047, "grad_norm": 3.079735040664673, "learning_rate": 9.987927565407832e-06, "loss": 1.3855, "step": 1227 }, { "epoch": 0.09623824451410658, "grad_norm": 2.804150342941284, "learning_rate": 9.987906140528793e-06, "loss": 1.4367, "step": 1228 }, { "epoch": 0.0963166144200627, "grad_norm": 3.303755521774292, "learning_rate": 9.98788469667832e-06, "loss": 1.5384, "step": 1229 }, { "epoch": 0.09639498432601881, "grad_norm": 2.173527717590332, "learning_rate": 9.987863233856496e-06, "loss": 1.4847, "step": 1230 }, { "epoch": 0.09647335423197492, "grad_norm": 3.7438621520996094, "learning_rate": 9.987841752063403e-06, "loss": 1.5872, "step": 1231 }, { "epoch": 0.09655172413793103, "grad_norm": 2.9476659297943115, "learning_rate": 9.987820251299121e-06, "loss": 1.7563, "step": 1232 }, { "epoch": 0.09663009404388714, "grad_norm": 2.487187623977661, "learning_rate": 9.987798731563734e-06, "loss": 1.3048, "step": 1233 }, { "epoch": 0.09670846394984325, "grad_norm": 4.030777931213379, "learning_rate": 9.987777192857322e-06, "loss": 1.9593, "step": 1234 }, { "epoch": 0.09678683385579938, "grad_norm": 4.1750922203063965, "learning_rate": 9.987755635179968e-06, "loss": 1.5814, "step": 1235 }, { "epoch": 0.09686520376175549, "grad_norm": 3.559645652770996, "learning_rate": 9.987734058531751e-06, "loss": 1.7089, "step": 1236 }, { "epoch": 0.0969435736677116, "grad_norm": 3.638335704803467, "learning_rate": 9.987712462912757e-06, "loss": 1.2852, "step": 1237 }, { "epoch": 0.09702194357366771, "grad_norm": 4.174403190612793, "learning_rate": 9.987690848323067e-06, "loss": 2.0725, "step": 1238 }, { "epoch": 0.09710031347962382, "grad_norm": 2.606447219848633, "learning_rate": 9.987669214762764e-06, "loss": 1.2665, "step": 1239 }, { "epoch": 0.09717868338557993, "grad_norm": 3.228545904159546, "learning_rate": 9.987647562231926e-06, "loss": 1.8085, "step": 1240 }, { "epoch": 0.09725705329153606, "grad_norm": 4.63649320602417, "learning_rate": 9.987625890730641e-06, "loss": 1.799, "step": 1241 }, { "epoch": 0.09733542319749217, "grad_norm": 3.2913260459899902, "learning_rate": 9.987604200258987e-06, "loss": 1.5307, "step": 1242 }, { "epoch": 0.09741379310344828, "grad_norm": 2.5439462661743164, "learning_rate": 9.98758249081705e-06, "loss": 1.5599, "step": 1243 }, { "epoch": 0.09749216300940439, "grad_norm": 6.026932239532471, "learning_rate": 9.987560762404911e-06, "loss": 1.926, "step": 1244 }, { "epoch": 0.0975705329153605, "grad_norm": 2.466188669204712, "learning_rate": 9.987539015022652e-06, "loss": 1.3722, "step": 1245 }, { "epoch": 0.09764890282131661, "grad_norm": 3.0963032245635986, "learning_rate": 9.987517248670356e-06, "loss": 1.7943, "step": 1246 }, { "epoch": 0.09772727272727273, "grad_norm": 4.120593547821045, "learning_rate": 9.987495463348107e-06, "loss": 1.6153, "step": 1247 }, { "epoch": 0.09780564263322884, "grad_norm": 4.184610843658447, "learning_rate": 9.987473659055985e-06, "loss": 1.2618, "step": 1248 }, { "epoch": 0.09788401253918495, "grad_norm": 3.217583656311035, "learning_rate": 9.987451835794077e-06, "loss": 1.6808, "step": 1249 }, { "epoch": 0.09796238244514106, "grad_norm": 2.923982620239258, "learning_rate": 9.987429993562464e-06, "loss": 1.5495, "step": 1250 }, { "epoch": 0.09804075235109717, "grad_norm": 1.961083173751831, "learning_rate": 9.98740813236123e-06, "loss": 1.4955, "step": 1251 }, { "epoch": 0.0981191222570533, "grad_norm": 2.6731045246124268, "learning_rate": 9.987386252190454e-06, "loss": 1.3868, "step": 1252 }, { "epoch": 0.09819749216300941, "grad_norm": 2.780165910720825, "learning_rate": 9.987364353050224e-06, "loss": 1.1447, "step": 1253 }, { "epoch": 0.09827586206896552, "grad_norm": 2.9347784519195557, "learning_rate": 9.987342434940622e-06, "loss": 1.2475, "step": 1254 }, { "epoch": 0.09835423197492163, "grad_norm": 3.922088623046875, "learning_rate": 9.987320497861731e-06, "loss": 1.4572, "step": 1255 }, { "epoch": 0.09843260188087774, "grad_norm": 2.2987711429595947, "learning_rate": 9.987298541813635e-06, "loss": 1.5288, "step": 1256 }, { "epoch": 0.09851097178683385, "grad_norm": 3.220700979232788, "learning_rate": 9.987276566796416e-06, "loss": 1.3952, "step": 1257 }, { "epoch": 0.09858934169278998, "grad_norm": 3.8850626945495605, "learning_rate": 9.987254572810158e-06, "loss": 1.9142, "step": 1258 }, { "epoch": 0.09866771159874609, "grad_norm": 2.8368277549743652, "learning_rate": 9.987232559854945e-06, "loss": 1.2222, "step": 1259 }, { "epoch": 0.0987460815047022, "grad_norm": 3.7345941066741943, "learning_rate": 9.987210527930863e-06, "loss": 1.8262, "step": 1260 }, { "epoch": 0.0988244514106583, "grad_norm": 3.4880223274230957, "learning_rate": 9.987188477037992e-06, "loss": 1.2352, "step": 1261 }, { "epoch": 0.09890282131661442, "grad_norm": 8.150455474853516, "learning_rate": 9.987166407176419e-06, "loss": 1.7464, "step": 1262 }, { "epoch": 0.09898119122257053, "grad_norm": 5.058964252471924, "learning_rate": 9.987144318346225e-06, "loss": 1.6276, "step": 1263 }, { "epoch": 0.09905956112852665, "grad_norm": 3.6223409175872803, "learning_rate": 9.987122210547498e-06, "loss": 1.7745, "step": 1264 }, { "epoch": 0.09913793103448276, "grad_norm": 3.962021827697754, "learning_rate": 9.987100083780316e-06, "loss": 1.7318, "step": 1265 }, { "epoch": 0.09921630094043887, "grad_norm": 3.969942331314087, "learning_rate": 9.98707793804477e-06, "loss": 1.5258, "step": 1266 }, { "epoch": 0.09929467084639498, "grad_norm": 3.3815479278564453, "learning_rate": 9.98705577334094e-06, "loss": 1.6971, "step": 1267 }, { "epoch": 0.0993730407523511, "grad_norm": 9.047493934631348, "learning_rate": 9.987033589668911e-06, "loss": 1.8328, "step": 1268 }, { "epoch": 0.0994514106583072, "grad_norm": 4.576421737670898, "learning_rate": 9.987011387028769e-06, "loss": 1.3882, "step": 1269 }, { "epoch": 0.09952978056426333, "grad_norm": 2.447075366973877, "learning_rate": 9.986989165420596e-06, "loss": 1.1024, "step": 1270 }, { "epoch": 0.09960815047021944, "grad_norm": 6.459280490875244, "learning_rate": 9.986966924844477e-06, "loss": 1.3639, "step": 1271 }, { "epoch": 0.09968652037617555, "grad_norm": 3.849930763244629, "learning_rate": 9.986944665300498e-06, "loss": 1.9235, "step": 1272 }, { "epoch": 0.09976489028213166, "grad_norm": 4.272002220153809, "learning_rate": 9.986922386788745e-06, "loss": 1.2341, "step": 1273 }, { "epoch": 0.09984326018808777, "grad_norm": 3.6730916500091553, "learning_rate": 9.986900089309296e-06, "loss": 1.4154, "step": 1274 }, { "epoch": 0.09992163009404388, "grad_norm": 3.7175841331481934, "learning_rate": 9.986877772862244e-06, "loss": 1.474, "step": 1275 }, { "epoch": 0.1, "grad_norm": 2.4308409690856934, "learning_rate": 9.986855437447671e-06, "loss": 1.5298, "step": 1276 }, { "epoch": 0.10007836990595612, "grad_norm": 2.2574386596679688, "learning_rate": 9.986833083065659e-06, "loss": 1.3405, "step": 1277 }, { "epoch": 0.10015673981191223, "grad_norm": 3.628286361694336, "learning_rate": 9.986810709716297e-06, "loss": 1.9579, "step": 1278 }, { "epoch": 0.10023510971786834, "grad_norm": 3.661005735397339, "learning_rate": 9.986788317399667e-06, "loss": 1.0178, "step": 1279 }, { "epoch": 0.10031347962382445, "grad_norm": 5.348194599151611, "learning_rate": 9.986765906115857e-06, "loss": 1.9396, "step": 1280 }, { "epoch": 0.10039184952978056, "grad_norm": 4.751436710357666, "learning_rate": 9.986743475864951e-06, "loss": 1.5507, "step": 1281 }, { "epoch": 0.10047021943573668, "grad_norm": 2.900078296661377, "learning_rate": 9.986721026647033e-06, "loss": 1.3682, "step": 1282 }, { "epoch": 0.10054858934169279, "grad_norm": 4.033102989196777, "learning_rate": 9.986698558462188e-06, "loss": 1.4553, "step": 1283 }, { "epoch": 0.1006269592476489, "grad_norm": 6.625602722167969, "learning_rate": 9.986676071310506e-06, "loss": 1.8188, "step": 1284 }, { "epoch": 0.10070532915360501, "grad_norm": 3.665355682373047, "learning_rate": 9.986653565192069e-06, "loss": 1.5317, "step": 1285 }, { "epoch": 0.10078369905956112, "grad_norm": 3.5408098697662354, "learning_rate": 9.986631040106963e-06, "loss": 1.5628, "step": 1286 }, { "epoch": 0.10086206896551723, "grad_norm": 2.595802068710327, "learning_rate": 9.986608496055273e-06, "loss": 1.218, "step": 1287 }, { "epoch": 0.10094043887147336, "grad_norm": 2.7486114501953125, "learning_rate": 9.986585933037087e-06, "loss": 1.5466, "step": 1288 }, { "epoch": 0.10101880877742947, "grad_norm": 3.7369062900543213, "learning_rate": 9.986563351052487e-06, "loss": 1.6358, "step": 1289 }, { "epoch": 0.10109717868338558, "grad_norm": 2.0905416011810303, "learning_rate": 9.986540750101562e-06, "loss": 1.5327, "step": 1290 }, { "epoch": 0.10117554858934169, "grad_norm": 4.80623197555542, "learning_rate": 9.9865181301844e-06, "loss": 1.7484, "step": 1291 }, { "epoch": 0.1012539184952978, "grad_norm": 3.268281936645508, "learning_rate": 9.986495491301083e-06, "loss": 1.545, "step": 1292 }, { "epoch": 0.10133228840125391, "grad_norm": 4.377697467803955, "learning_rate": 9.986472833451697e-06, "loss": 1.5559, "step": 1293 }, { "epoch": 0.10141065830721004, "grad_norm": 12.733734130859375, "learning_rate": 9.98645015663633e-06, "loss": 1.31, "step": 1294 }, { "epoch": 0.10148902821316615, "grad_norm": 2.8974802494049072, "learning_rate": 9.986427460855069e-06, "loss": 1.7696, "step": 1295 }, { "epoch": 0.10156739811912226, "grad_norm": 6.476355075836182, "learning_rate": 9.986404746107997e-06, "loss": 0.7873, "step": 1296 }, { "epoch": 0.10164576802507837, "grad_norm": 2.5670523643493652, "learning_rate": 9.986382012395205e-06, "loss": 1.1053, "step": 1297 }, { "epoch": 0.10172413793103448, "grad_norm": 4.439074993133545, "learning_rate": 9.986359259716774e-06, "loss": 1.6297, "step": 1298 }, { "epoch": 0.1018025078369906, "grad_norm": 3.345081329345703, "learning_rate": 9.986336488072795e-06, "loss": 1.3969, "step": 1299 }, { "epoch": 0.10188087774294671, "grad_norm": 5.080874443054199, "learning_rate": 9.986313697463353e-06, "loss": 1.1211, "step": 1300 }, { "epoch": 0.10195924764890282, "grad_norm": 6.464119911193848, "learning_rate": 9.986290887888535e-06, "loss": 2.0106, "step": 1301 }, { "epoch": 0.10203761755485893, "grad_norm": 2.4051711559295654, "learning_rate": 9.986268059348428e-06, "loss": 1.4624, "step": 1302 }, { "epoch": 0.10211598746081504, "grad_norm": 2.6874406337738037, "learning_rate": 9.986245211843119e-06, "loss": 1.4489, "step": 1303 }, { "epoch": 0.10219435736677115, "grad_norm": 6.344262599945068, "learning_rate": 9.986222345372693e-06, "loss": 1.6851, "step": 1304 }, { "epoch": 0.10227272727272728, "grad_norm": 2.6556761264801025, "learning_rate": 9.986199459937239e-06, "loss": 0.9327, "step": 1305 }, { "epoch": 0.10235109717868339, "grad_norm": 18.0137882232666, "learning_rate": 9.986176555536844e-06, "loss": 0.5858, "step": 1306 }, { "epoch": 0.1024294670846395, "grad_norm": 6.424585819244385, "learning_rate": 9.986153632171592e-06, "loss": 1.8473, "step": 1307 }, { "epoch": 0.10250783699059561, "grad_norm": 4.267082691192627, "learning_rate": 9.986130689841574e-06, "loss": 1.8216, "step": 1308 }, { "epoch": 0.10258620689655172, "grad_norm": 2.82061505317688, "learning_rate": 9.986107728546877e-06, "loss": 1.6738, "step": 1309 }, { "epoch": 0.10266457680250783, "grad_norm": 3.5711560249328613, "learning_rate": 9.986084748287585e-06, "loss": 1.4947, "step": 1310 }, { "epoch": 0.10274294670846396, "grad_norm": 2.3853766918182373, "learning_rate": 9.98606174906379e-06, "loss": 1.4808, "step": 1311 }, { "epoch": 0.10282131661442007, "grad_norm": 3.1677980422973633, "learning_rate": 9.986038730875577e-06, "loss": 1.3966, "step": 1312 }, { "epoch": 0.10289968652037618, "grad_norm": 3.49666166305542, "learning_rate": 9.986015693723031e-06, "loss": 1.243, "step": 1313 }, { "epoch": 0.10297805642633229, "grad_norm": 3.482412099838257, "learning_rate": 9.985992637606244e-06, "loss": 2.0884, "step": 1314 }, { "epoch": 0.1030564263322884, "grad_norm": 3.0461697578430176, "learning_rate": 9.985969562525301e-06, "loss": 1.5581, "step": 1315 }, { "epoch": 0.10313479623824451, "grad_norm": 3.0834450721740723, "learning_rate": 9.985946468480294e-06, "loss": 1.2357, "step": 1316 }, { "epoch": 0.10321316614420063, "grad_norm": 9.384572982788086, "learning_rate": 9.985923355471305e-06, "loss": 2.5466, "step": 1317 }, { "epoch": 0.10329153605015674, "grad_norm": 2.838451385498047, "learning_rate": 9.985900223498424e-06, "loss": 1.5841, "step": 1318 }, { "epoch": 0.10336990595611285, "grad_norm": 3.9183990955352783, "learning_rate": 9.98587707256174e-06, "loss": 1.5876, "step": 1319 }, { "epoch": 0.10344827586206896, "grad_norm": 2.5397565364837646, "learning_rate": 9.98585390266134e-06, "loss": 1.2299, "step": 1320 }, { "epoch": 0.10352664576802507, "grad_norm": 3.4601247310638428, "learning_rate": 9.985830713797315e-06, "loss": 1.8321, "step": 1321 }, { "epoch": 0.10360501567398119, "grad_norm": 3.532384157180786, "learning_rate": 9.985807505969748e-06, "loss": 1.3673, "step": 1322 }, { "epoch": 0.10368338557993731, "grad_norm": 3.4208710193634033, "learning_rate": 9.985784279178732e-06, "loss": 1.2074, "step": 1323 }, { "epoch": 0.10376175548589342, "grad_norm": 3.4372262954711914, "learning_rate": 9.985761033424352e-06, "loss": 1.3352, "step": 1324 }, { "epoch": 0.10384012539184953, "grad_norm": 3.684277057647705, "learning_rate": 9.985737768706699e-06, "loss": 1.3461, "step": 1325 }, { "epoch": 0.10391849529780564, "grad_norm": 3.7102293968200684, "learning_rate": 9.98571448502586e-06, "loss": 1.5916, "step": 1326 }, { "epoch": 0.10399686520376175, "grad_norm": 3.218059539794922, "learning_rate": 9.985691182381925e-06, "loss": 1.6441, "step": 1327 }, { "epoch": 0.10407523510971786, "grad_norm": 6.01343297958374, "learning_rate": 9.985667860774979e-06, "loss": 1.7499, "step": 1328 }, { "epoch": 0.10415360501567399, "grad_norm": 2.5734760761260986, "learning_rate": 9.985644520205115e-06, "loss": 1.3945, "step": 1329 }, { "epoch": 0.1042319749216301, "grad_norm": 2.8465662002563477, "learning_rate": 9.98562116067242e-06, "loss": 1.3294, "step": 1330 }, { "epoch": 0.10431034482758621, "grad_norm": 2.727857828140259, "learning_rate": 9.985597782176983e-06, "loss": 1.3744, "step": 1331 }, { "epoch": 0.10438871473354232, "grad_norm": 2.7350573539733887, "learning_rate": 9.985574384718892e-06, "loss": 1.7713, "step": 1332 }, { "epoch": 0.10446708463949843, "grad_norm": 3.255059003829956, "learning_rate": 9.985550968298238e-06, "loss": 1.4763, "step": 1333 }, { "epoch": 0.10454545454545454, "grad_norm": 2.8807058334350586, "learning_rate": 9.985527532915109e-06, "loss": 1.2992, "step": 1334 }, { "epoch": 0.10462382445141066, "grad_norm": 4.291005611419678, "learning_rate": 9.985504078569593e-06, "loss": 1.3952, "step": 1335 }, { "epoch": 0.10470219435736677, "grad_norm": 4.401349067687988, "learning_rate": 9.98548060526178e-06, "loss": 1.2373, "step": 1336 }, { "epoch": 0.10478056426332288, "grad_norm": 4.239267349243164, "learning_rate": 9.98545711299176e-06, "loss": 1.9186, "step": 1337 }, { "epoch": 0.104858934169279, "grad_norm": 3.373670816421509, "learning_rate": 9.98543360175962e-06, "loss": 1.4212, "step": 1338 }, { "epoch": 0.1049373040752351, "grad_norm": 2.9461450576782227, "learning_rate": 9.985410071565453e-06, "loss": 1.4965, "step": 1339 }, { "epoch": 0.10501567398119123, "grad_norm": 3.7228662967681885, "learning_rate": 9.985386522409347e-06, "loss": 1.6906, "step": 1340 }, { "epoch": 0.10509404388714734, "grad_norm": 5.179684162139893, "learning_rate": 9.985362954291391e-06, "loss": 1.8038, "step": 1341 }, { "epoch": 0.10517241379310345, "grad_norm": 2.0478262901306152, "learning_rate": 9.985339367211675e-06, "loss": 1.1887, "step": 1342 }, { "epoch": 0.10525078369905956, "grad_norm": 6.581662178039551, "learning_rate": 9.985315761170288e-06, "loss": 1.3354, "step": 1343 }, { "epoch": 0.10532915360501567, "grad_norm": 3.101642370223999, "learning_rate": 9.985292136167321e-06, "loss": 1.4944, "step": 1344 }, { "epoch": 0.10540752351097178, "grad_norm": 4.3950419425964355, "learning_rate": 9.985268492202865e-06, "loss": 1.4835, "step": 1345 }, { "epoch": 0.1054858934169279, "grad_norm": 5.12552547454834, "learning_rate": 9.985244829277005e-06, "loss": 1.4825, "step": 1346 }, { "epoch": 0.10556426332288402, "grad_norm": 2.2404470443725586, "learning_rate": 9.985221147389836e-06, "loss": 1.3717, "step": 1347 }, { "epoch": 0.10564263322884013, "grad_norm": 3.1191093921661377, "learning_rate": 9.985197446541449e-06, "loss": 1.2696, "step": 1348 }, { "epoch": 0.10572100313479624, "grad_norm": 6.0314483642578125, "learning_rate": 9.985173726731927e-06, "loss": 1.726, "step": 1349 }, { "epoch": 0.10579937304075235, "grad_norm": 2.5733747482299805, "learning_rate": 9.985149987961367e-06, "loss": 1.61, "step": 1350 }, { "epoch": 0.10587774294670846, "grad_norm": 6.732783794403076, "learning_rate": 9.985126230229855e-06, "loss": 0.7639, "step": 1351 }, { "epoch": 0.10595611285266458, "grad_norm": 3.101213216781616, "learning_rate": 9.985102453537485e-06, "loss": 1.1875, "step": 1352 }, { "epoch": 0.1060344827586207, "grad_norm": 7.801120758056641, "learning_rate": 9.985078657884346e-06, "loss": 1.3162, "step": 1353 }, { "epoch": 0.1061128526645768, "grad_norm": 9.292405128479004, "learning_rate": 9.985054843270528e-06, "loss": 1.2156, "step": 1354 }, { "epoch": 0.10619122257053291, "grad_norm": 3.2666852474212646, "learning_rate": 9.985031009696122e-06, "loss": 1.5186, "step": 1355 }, { "epoch": 0.10626959247648902, "grad_norm": 3.5975570678710938, "learning_rate": 9.985007157161218e-06, "loss": 1.4966, "step": 1356 }, { "epoch": 0.10634796238244514, "grad_norm": 10.04698371887207, "learning_rate": 9.984983285665907e-06, "loss": 1.8981, "step": 1357 }, { "epoch": 0.10642633228840126, "grad_norm": 3.6700961589813232, "learning_rate": 9.98495939521028e-06, "loss": 1.7117, "step": 1358 }, { "epoch": 0.10650470219435737, "grad_norm": 3.628664255142212, "learning_rate": 9.98493548579443e-06, "loss": 1.5968, "step": 1359 }, { "epoch": 0.10658307210031348, "grad_norm": 2.7290618419647217, "learning_rate": 9.984911557418444e-06, "loss": 1.4574, "step": 1360 }, { "epoch": 0.10666144200626959, "grad_norm": 2.323507308959961, "learning_rate": 9.984887610082415e-06, "loss": 1.6618, "step": 1361 }, { "epoch": 0.1067398119122257, "grad_norm": 2.418248176574707, "learning_rate": 9.984863643786433e-06, "loss": 1.25, "step": 1362 }, { "epoch": 0.10681818181818181, "grad_norm": 3.7896978855133057, "learning_rate": 9.984839658530593e-06, "loss": 1.5718, "step": 1363 }, { "epoch": 0.10689655172413794, "grad_norm": 2.413041591644287, "learning_rate": 9.98481565431498e-06, "loss": 1.6321, "step": 1364 }, { "epoch": 0.10697492163009405, "grad_norm": 4.317252159118652, "learning_rate": 9.98479163113969e-06, "loss": 1.5678, "step": 1365 }, { "epoch": 0.10705329153605016, "grad_norm": 4.551120758056641, "learning_rate": 9.984767589004812e-06, "loss": 1.5426, "step": 1366 }, { "epoch": 0.10713166144200627, "grad_norm": 9.226299285888672, "learning_rate": 9.98474352791044e-06, "loss": 1.9655, "step": 1367 }, { "epoch": 0.10721003134796238, "grad_norm": 3.2684037685394287, "learning_rate": 9.984719447856663e-06, "loss": 1.3643, "step": 1368 }, { "epoch": 0.10728840125391849, "grad_norm": 2.2806897163391113, "learning_rate": 9.984695348843574e-06, "loss": 1.4825, "step": 1369 }, { "epoch": 0.10736677115987461, "grad_norm": 3.253774642944336, "learning_rate": 9.984671230871263e-06, "loss": 1.1113, "step": 1370 }, { "epoch": 0.10744514106583072, "grad_norm": 3.9240734577178955, "learning_rate": 9.984647093939826e-06, "loss": 1.3711, "step": 1371 }, { "epoch": 0.10752351097178683, "grad_norm": 3.052095413208008, "learning_rate": 9.984622938049349e-06, "loss": 1.62, "step": 1372 }, { "epoch": 0.10760188087774294, "grad_norm": 3.694463014602661, "learning_rate": 9.984598763199927e-06, "loss": 1.4505, "step": 1373 }, { "epoch": 0.10768025078369906, "grad_norm": 2.6655054092407227, "learning_rate": 9.984574569391652e-06, "loss": 1.2264, "step": 1374 }, { "epoch": 0.10775862068965517, "grad_norm": 2.768463373184204, "learning_rate": 9.984550356624616e-06, "loss": 1.2686, "step": 1375 }, { "epoch": 0.10783699059561129, "grad_norm": 3.110783338546753, "learning_rate": 9.98452612489891e-06, "loss": 1.7036, "step": 1376 }, { "epoch": 0.1079153605015674, "grad_norm": 3.008657455444336, "learning_rate": 9.984501874214628e-06, "loss": 1.3026, "step": 1377 }, { "epoch": 0.10799373040752351, "grad_norm": 2.93041729927063, "learning_rate": 9.984477604571858e-06, "loss": 1.464, "step": 1378 }, { "epoch": 0.10807210031347962, "grad_norm": 4.407774448394775, "learning_rate": 9.984453315970698e-06, "loss": 1.504, "step": 1379 }, { "epoch": 0.10815047021943573, "grad_norm": 4.3941755294799805, "learning_rate": 9.984429008411237e-06, "loss": 1.5582, "step": 1380 }, { "epoch": 0.10822884012539186, "grad_norm": 3.7349741458892822, "learning_rate": 9.984404681893569e-06, "loss": 1.5559, "step": 1381 }, { "epoch": 0.10830721003134797, "grad_norm": 4.282199382781982, "learning_rate": 9.984380336417786e-06, "loss": 1.9637, "step": 1382 }, { "epoch": 0.10838557993730408, "grad_norm": 4.052098274230957, "learning_rate": 9.984355971983979e-06, "loss": 1.3813, "step": 1383 }, { "epoch": 0.10846394984326019, "grad_norm": 3.4270694255828857, "learning_rate": 9.984331588592241e-06, "loss": 1.2075, "step": 1384 }, { "epoch": 0.1085423197492163, "grad_norm": 3.4472506046295166, "learning_rate": 9.984307186242668e-06, "loss": 1.4952, "step": 1385 }, { "epoch": 0.10862068965517241, "grad_norm": 2.5984065532684326, "learning_rate": 9.98428276493535e-06, "loss": 1.3802, "step": 1386 }, { "epoch": 0.10869905956112853, "grad_norm": 2.571802854537964, "learning_rate": 9.98425832467038e-06, "loss": 1.1977, "step": 1387 }, { "epoch": 0.10877742946708464, "grad_norm": 4.03612756729126, "learning_rate": 9.984233865447852e-06, "loss": 1.456, "step": 1388 }, { "epoch": 0.10885579937304075, "grad_norm": 3.4451241493225098, "learning_rate": 9.984209387267858e-06, "loss": 1.3768, "step": 1389 }, { "epoch": 0.10893416927899686, "grad_norm": 2.544605016708374, "learning_rate": 9.984184890130491e-06, "loss": 1.3684, "step": 1390 }, { "epoch": 0.10901253918495298, "grad_norm": 3.514357805252075, "learning_rate": 9.984160374035847e-06, "loss": 1.3495, "step": 1391 }, { "epoch": 0.10909090909090909, "grad_norm": 2.161104917526245, "learning_rate": 9.984135838984016e-06, "loss": 1.6937, "step": 1392 }, { "epoch": 0.10916927899686521, "grad_norm": 3.367159366607666, "learning_rate": 9.98411128497509e-06, "loss": 1.315, "step": 1393 }, { "epoch": 0.10924764890282132, "grad_norm": 2.659151554107666, "learning_rate": 9.984086712009167e-06, "loss": 1.2793, "step": 1394 }, { "epoch": 0.10932601880877743, "grad_norm": 3.6122546195983887, "learning_rate": 9.984062120086337e-06, "loss": 1.2398, "step": 1395 }, { "epoch": 0.10940438871473354, "grad_norm": 1.8135459423065186, "learning_rate": 9.984037509206695e-06, "loss": 1.3642, "step": 1396 }, { "epoch": 0.10948275862068965, "grad_norm": 2.8041727542877197, "learning_rate": 9.984012879370335e-06, "loss": 1.4261, "step": 1397 }, { "epoch": 0.10956112852664576, "grad_norm": 2.1876871585845947, "learning_rate": 9.983988230577349e-06, "loss": 1.3968, "step": 1398 }, { "epoch": 0.10963949843260189, "grad_norm": 2.9310173988342285, "learning_rate": 9.983963562827832e-06, "loss": 1.5165, "step": 1399 }, { "epoch": 0.109717868338558, "grad_norm": 9.018304824829102, "learning_rate": 9.983938876121876e-06, "loss": 1.7059, "step": 1400 }, { "epoch": 0.10979623824451411, "grad_norm": 3.3624703884124756, "learning_rate": 9.98391417045958e-06, "loss": 1.9486, "step": 1401 }, { "epoch": 0.10987460815047022, "grad_norm": 4.442452907562256, "learning_rate": 9.983889445841033e-06, "loss": 1.9667, "step": 1402 }, { "epoch": 0.10995297805642633, "grad_norm": 2.6633152961730957, "learning_rate": 9.983864702266329e-06, "loss": 1.4327, "step": 1403 }, { "epoch": 0.11003134796238244, "grad_norm": 3.49831223487854, "learning_rate": 9.983839939735564e-06, "loss": 1.607, "step": 1404 }, { "epoch": 0.11010971786833856, "grad_norm": 2.4681894779205322, "learning_rate": 9.983815158248831e-06, "loss": 1.6994, "step": 1405 }, { "epoch": 0.11018808777429467, "grad_norm": 3.8352622985839844, "learning_rate": 9.983790357806226e-06, "loss": 1.4315, "step": 1406 }, { "epoch": 0.11026645768025078, "grad_norm": 3.096067190170288, "learning_rate": 9.983765538407843e-06, "loss": 2.0365, "step": 1407 }, { "epoch": 0.1103448275862069, "grad_norm": 11.566673278808594, "learning_rate": 9.983740700053776e-06, "loss": 1.1626, "step": 1408 }, { "epoch": 0.110423197492163, "grad_norm": 2.4714200496673584, "learning_rate": 9.983715842744118e-06, "loss": 1.8019, "step": 1409 }, { "epoch": 0.11050156739811912, "grad_norm": 7.241783142089844, "learning_rate": 9.983690966478965e-06, "loss": 1.693, "step": 1410 }, { "epoch": 0.11057993730407524, "grad_norm": 25.633419036865234, "learning_rate": 9.983666071258411e-06, "loss": 1.7959, "step": 1411 }, { "epoch": 0.11065830721003135, "grad_norm": 5.431763648986816, "learning_rate": 9.983641157082552e-06, "loss": 1.1811, "step": 1412 }, { "epoch": 0.11073667711598746, "grad_norm": 2.969264507293701, "learning_rate": 9.983616223951482e-06, "loss": 1.4573, "step": 1413 }, { "epoch": 0.11081504702194357, "grad_norm": 3.557251214981079, "learning_rate": 9.983591271865293e-06, "loss": 2.0066, "step": 1414 }, { "epoch": 0.11089341692789968, "grad_norm": 5.7927141189575195, "learning_rate": 9.983566300824086e-06, "loss": 1.4966, "step": 1415 }, { "epoch": 0.11097178683385579, "grad_norm": 9.63031005859375, "learning_rate": 9.983541310827951e-06, "loss": 0.9652, "step": 1416 }, { "epoch": 0.11105015673981192, "grad_norm": 3.3896639347076416, "learning_rate": 9.983516301876983e-06, "loss": 1.3805, "step": 1417 }, { "epoch": 0.11112852664576803, "grad_norm": 3.2780730724334717, "learning_rate": 9.983491273971281e-06, "loss": 1.322, "step": 1418 }, { "epoch": 0.11120689655172414, "grad_norm": 2.843106746673584, "learning_rate": 9.983466227110937e-06, "loss": 1.3307, "step": 1419 }, { "epoch": 0.11128526645768025, "grad_norm": 2.8335888385772705, "learning_rate": 9.983441161296048e-06, "loss": 1.6538, "step": 1420 }, { "epoch": 0.11136363636363636, "grad_norm": 3.521028518676758, "learning_rate": 9.983416076526707e-06, "loss": 1.2155, "step": 1421 }, { "epoch": 0.11144200626959247, "grad_norm": 2.50395131111145, "learning_rate": 9.98339097280301e-06, "loss": 1.5003, "step": 1422 }, { "epoch": 0.1115203761755486, "grad_norm": 3.855677843093872, "learning_rate": 9.983365850125056e-06, "loss": 1.3661, "step": 1423 }, { "epoch": 0.1115987460815047, "grad_norm": 2.7695887088775635, "learning_rate": 9.983340708492937e-06, "loss": 1.5228, "step": 1424 }, { "epoch": 0.11167711598746081, "grad_norm": 12.377819061279297, "learning_rate": 9.983315547906749e-06, "loss": 1.4407, "step": 1425 }, { "epoch": 0.11175548589341693, "grad_norm": 2.808154821395874, "learning_rate": 9.98329036836659e-06, "loss": 1.4014, "step": 1426 }, { "epoch": 0.11183385579937304, "grad_norm": 6.351405620574951, "learning_rate": 9.983265169872551e-06, "loss": 1.5002, "step": 1427 }, { "epoch": 0.11191222570532916, "grad_norm": 3.0881028175354004, "learning_rate": 9.983239952424733e-06, "loss": 1.4689, "step": 1428 }, { "epoch": 0.11199059561128527, "grad_norm": 6.146876811981201, "learning_rate": 9.98321471602323e-06, "loss": 1.0561, "step": 1429 }, { "epoch": 0.11206896551724138, "grad_norm": 6.048740863800049, "learning_rate": 9.983189460668136e-06, "loss": 0.9314, "step": 1430 }, { "epoch": 0.11214733542319749, "grad_norm": 3.0211281776428223, "learning_rate": 9.98316418635955e-06, "loss": 1.2615, "step": 1431 }, { "epoch": 0.1122257053291536, "grad_norm": 2.1274588108062744, "learning_rate": 9.983138893097569e-06, "loss": 1.2173, "step": 1432 }, { "epoch": 0.11230407523510971, "grad_norm": 2.5367233753204346, "learning_rate": 9.983113580882285e-06, "loss": 1.4984, "step": 1433 }, { "epoch": 0.11238244514106584, "grad_norm": 3.6191585063934326, "learning_rate": 9.983088249713795e-06, "loss": 1.4524, "step": 1434 }, { "epoch": 0.11246081504702195, "grad_norm": 2.5541019439697266, "learning_rate": 9.983062899592198e-06, "loss": 1.7492, "step": 1435 }, { "epoch": 0.11253918495297806, "grad_norm": 3.0370755195617676, "learning_rate": 9.98303753051759e-06, "loss": 1.0229, "step": 1436 }, { "epoch": 0.11261755485893417, "grad_norm": 3.115790605545044, "learning_rate": 9.983012142490065e-06, "loss": 1.1747, "step": 1437 }, { "epoch": 0.11269592476489028, "grad_norm": 2.7803189754486084, "learning_rate": 9.982986735509724e-06, "loss": 1.6167, "step": 1438 }, { "epoch": 0.11277429467084639, "grad_norm": 2.302793025970459, "learning_rate": 9.982961309576657e-06, "loss": 1.2894, "step": 1439 }, { "epoch": 0.11285266457680251, "grad_norm": 3.9727869033813477, "learning_rate": 9.982935864690968e-06, "loss": 1.5011, "step": 1440 }, { "epoch": 0.11293103448275862, "grad_norm": 13.266225814819336, "learning_rate": 9.982910400852748e-06, "loss": 1.0506, "step": 1441 }, { "epoch": 0.11300940438871473, "grad_norm": 3.359929323196411, "learning_rate": 9.9828849180621e-06, "loss": 1.5056, "step": 1442 }, { "epoch": 0.11308777429467085, "grad_norm": 2.858736276626587, "learning_rate": 9.982859416319113e-06, "loss": 1.4486, "step": 1443 }, { "epoch": 0.11316614420062696, "grad_norm": 2.5531864166259766, "learning_rate": 9.982833895623889e-06, "loss": 1.5327, "step": 1444 }, { "epoch": 0.11324451410658307, "grad_norm": 3.9583280086517334, "learning_rate": 9.982808355976525e-06, "loss": 1.4975, "step": 1445 }, { "epoch": 0.11332288401253919, "grad_norm": 3.842825412750244, "learning_rate": 9.982782797377117e-06, "loss": 1.4112, "step": 1446 }, { "epoch": 0.1134012539184953, "grad_norm": 3.1194229125976562, "learning_rate": 9.982757219825763e-06, "loss": 1.5562, "step": 1447 }, { "epoch": 0.11347962382445141, "grad_norm": 3.716073751449585, "learning_rate": 9.98273162332256e-06, "loss": 1.0952, "step": 1448 }, { "epoch": 0.11355799373040752, "grad_norm": 5.152045726776123, "learning_rate": 9.982706007867604e-06, "loss": 1.6039, "step": 1449 }, { "epoch": 0.11363636363636363, "grad_norm": 2.2733066082000732, "learning_rate": 9.982680373460996e-06, "loss": 1.492, "step": 1450 }, { "epoch": 0.11371473354231974, "grad_norm": 2.86094069480896, "learning_rate": 9.982654720102828e-06, "loss": 1.6913, "step": 1451 }, { "epoch": 0.11379310344827587, "grad_norm": 3.5707473754882812, "learning_rate": 9.982629047793203e-06, "loss": 1.675, "step": 1452 }, { "epoch": 0.11387147335423198, "grad_norm": 7.139400959014893, "learning_rate": 9.982603356532215e-06, "loss": 0.7553, "step": 1453 }, { "epoch": 0.11394984326018809, "grad_norm": 2.1193697452545166, "learning_rate": 9.982577646319965e-06, "loss": 1.359, "step": 1454 }, { "epoch": 0.1140282131661442, "grad_norm": 2.3933603763580322, "learning_rate": 9.982551917156546e-06, "loss": 1.2244, "step": 1455 }, { "epoch": 0.11410658307210031, "grad_norm": 2.5984809398651123, "learning_rate": 9.982526169042061e-06, "loss": 1.5639, "step": 1456 }, { "epoch": 0.11418495297805642, "grad_norm": 3.0556514263153076, "learning_rate": 9.982500401976605e-06, "loss": 1.2801, "step": 1457 }, { "epoch": 0.11426332288401254, "grad_norm": 4.0801682472229, "learning_rate": 9.982474615960276e-06, "loss": 1.4742, "step": 1458 }, { "epoch": 0.11434169278996865, "grad_norm": 3.1154403686523438, "learning_rate": 9.982448810993173e-06, "loss": 1.6628, "step": 1459 }, { "epoch": 0.11442006269592477, "grad_norm": 2.395970582962036, "learning_rate": 9.982422987075396e-06, "loss": 1.5362, "step": 1460 }, { "epoch": 0.11449843260188088, "grad_norm": 2.746690034866333, "learning_rate": 9.982397144207037e-06, "loss": 1.3355, "step": 1461 }, { "epoch": 0.11457680250783699, "grad_norm": 4.991861343383789, "learning_rate": 9.982371282388201e-06, "loss": 1.3734, "step": 1462 }, { "epoch": 0.1146551724137931, "grad_norm": 3.1434574127197266, "learning_rate": 9.982345401618984e-06, "loss": 1.5319, "step": 1463 }, { "epoch": 0.11473354231974922, "grad_norm": 3.9758684635162354, "learning_rate": 9.982319501899483e-06, "loss": 1.4742, "step": 1464 }, { "epoch": 0.11481191222570533, "grad_norm": 3.4296517372131348, "learning_rate": 9.982293583229798e-06, "loss": 1.6712, "step": 1465 }, { "epoch": 0.11489028213166144, "grad_norm": 2.5148017406463623, "learning_rate": 9.982267645610028e-06, "loss": 1.3134, "step": 1466 }, { "epoch": 0.11496865203761755, "grad_norm": 3.361823558807373, "learning_rate": 9.98224168904027e-06, "loss": 1.89, "step": 1467 }, { "epoch": 0.11504702194357366, "grad_norm": 5.856967449188232, "learning_rate": 9.982215713520624e-06, "loss": 1.539, "step": 1468 }, { "epoch": 0.11512539184952979, "grad_norm": 3.43686580657959, "learning_rate": 9.982189719051189e-06, "loss": 1.3791, "step": 1469 }, { "epoch": 0.1152037617554859, "grad_norm": 3.011754274368286, "learning_rate": 9.982163705632062e-06, "loss": 1.5205, "step": 1470 }, { "epoch": 0.11528213166144201, "grad_norm": 6.164048671722412, "learning_rate": 9.982137673263345e-06, "loss": 1.7036, "step": 1471 }, { "epoch": 0.11536050156739812, "grad_norm": 4.092775821685791, "learning_rate": 9.982111621945135e-06, "loss": 1.1379, "step": 1472 }, { "epoch": 0.11543887147335423, "grad_norm": 9.093771934509277, "learning_rate": 9.98208555167753e-06, "loss": 1.7543, "step": 1473 }, { "epoch": 0.11551724137931034, "grad_norm": 3.1618638038635254, "learning_rate": 9.982059462460631e-06, "loss": 1.5311, "step": 1474 }, { "epoch": 0.11559561128526646, "grad_norm": 2.570397138595581, "learning_rate": 9.982033354294536e-06, "loss": 1.6212, "step": 1475 }, { "epoch": 0.11567398119122257, "grad_norm": 2.8112621307373047, "learning_rate": 9.982007227179347e-06, "loss": 1.3462, "step": 1476 }, { "epoch": 0.11575235109717869, "grad_norm": 5.692862510681152, "learning_rate": 9.981981081115162e-06, "loss": 1.8577, "step": 1477 }, { "epoch": 0.1158307210031348, "grad_norm": 3.0109763145446777, "learning_rate": 9.981954916102077e-06, "loss": 1.3332, "step": 1478 }, { "epoch": 0.1159090909090909, "grad_norm": 21.380558013916016, "learning_rate": 9.981928732140196e-06, "loss": 1.2706, "step": 1479 }, { "epoch": 0.11598746081504702, "grad_norm": 7.350719928741455, "learning_rate": 9.981902529229617e-06, "loss": 0.9679, "step": 1480 }, { "epoch": 0.11606583072100314, "grad_norm": 2.9110336303710938, "learning_rate": 9.98187630737044e-06, "loss": 0.9111, "step": 1481 }, { "epoch": 0.11614420062695925, "grad_norm": 3.113180160522461, "learning_rate": 9.981850066562763e-06, "loss": 1.4787, "step": 1482 }, { "epoch": 0.11622257053291536, "grad_norm": 3.9306089878082275, "learning_rate": 9.981823806806688e-06, "loss": 1.2653, "step": 1483 }, { "epoch": 0.11630094043887147, "grad_norm": 2.1635313034057617, "learning_rate": 9.981797528102313e-06, "loss": 1.0162, "step": 1484 }, { "epoch": 0.11637931034482758, "grad_norm": 2.7097039222717285, "learning_rate": 9.98177123044974e-06, "loss": 1.2238, "step": 1485 }, { "epoch": 0.1164576802507837, "grad_norm": 3.1930928230285645, "learning_rate": 9.981744913849068e-06, "loss": 1.6427, "step": 1486 }, { "epoch": 0.11653605015673982, "grad_norm": 3.5337741374969482, "learning_rate": 9.981718578300398e-06, "loss": 1.2734, "step": 1487 }, { "epoch": 0.11661442006269593, "grad_norm": 3.8529207706451416, "learning_rate": 9.981692223803829e-06, "loss": 1.3349, "step": 1488 }, { "epoch": 0.11669278996865204, "grad_norm": 3.1851277351379395, "learning_rate": 9.981665850359461e-06, "loss": 1.4364, "step": 1489 }, { "epoch": 0.11677115987460815, "grad_norm": 4.165377616882324, "learning_rate": 9.981639457967394e-06, "loss": 1.2691, "step": 1490 }, { "epoch": 0.11684952978056426, "grad_norm": 4.501339435577393, "learning_rate": 9.98161304662773e-06, "loss": 1.3778, "step": 1491 }, { "epoch": 0.11692789968652037, "grad_norm": 2.949492931365967, "learning_rate": 9.981586616340569e-06, "loss": 1.4003, "step": 1492 }, { "epoch": 0.1170062695924765, "grad_norm": 3.799280881881714, "learning_rate": 9.98156016710601e-06, "loss": 2.4234, "step": 1493 }, { "epoch": 0.1170846394984326, "grad_norm": 4.563580513000488, "learning_rate": 9.981533698924157e-06, "loss": 1.3676, "step": 1494 }, { "epoch": 0.11716300940438872, "grad_norm": 2.018812417984009, "learning_rate": 9.981507211795107e-06, "loss": 1.4538, "step": 1495 }, { "epoch": 0.11724137931034483, "grad_norm": 3.17291522026062, "learning_rate": 9.981480705718962e-06, "loss": 1.4537, "step": 1496 }, { "epoch": 0.11731974921630094, "grad_norm": 1.959571123123169, "learning_rate": 9.981454180695824e-06, "loss": 0.9738, "step": 1497 }, { "epoch": 0.11739811912225705, "grad_norm": 2.856323003768921, "learning_rate": 9.981427636725793e-06, "loss": 1.7237, "step": 1498 }, { "epoch": 0.11747648902821317, "grad_norm": 4.610827922821045, "learning_rate": 9.981401073808969e-06, "loss": 1.7185, "step": 1499 }, { "epoch": 0.11755485893416928, "grad_norm": 3.3605453968048096, "learning_rate": 9.981374491945455e-06, "loss": 1.2548, "step": 1500 }, { "epoch": 0.11763322884012539, "grad_norm": 2.3773961067199707, "learning_rate": 9.981347891135351e-06, "loss": 0.8837, "step": 1501 }, { "epoch": 0.1177115987460815, "grad_norm": 3.4002485275268555, "learning_rate": 9.981321271378756e-06, "loss": 1.6029, "step": 1502 }, { "epoch": 0.11778996865203761, "grad_norm": 5.144306659698486, "learning_rate": 9.981294632675775e-06, "loss": 1.4132, "step": 1503 }, { "epoch": 0.11786833855799372, "grad_norm": 2.4908971786499023, "learning_rate": 9.981267975026509e-06, "loss": 1.4681, "step": 1504 }, { "epoch": 0.11794670846394985, "grad_norm": 2.8793318271636963, "learning_rate": 9.981241298431057e-06, "loss": 1.8703, "step": 1505 }, { "epoch": 0.11802507836990596, "grad_norm": 2.7286887168884277, "learning_rate": 9.981214602889522e-06, "loss": 1.6649, "step": 1506 }, { "epoch": 0.11810344827586207, "grad_norm": 2.28155255317688, "learning_rate": 9.981187888402004e-06, "loss": 1.3812, "step": 1507 }, { "epoch": 0.11818181818181818, "grad_norm": 3.182689666748047, "learning_rate": 9.981161154968606e-06, "loss": 1.1209, "step": 1508 }, { "epoch": 0.11826018808777429, "grad_norm": 3.296691417694092, "learning_rate": 9.98113440258943e-06, "loss": 1.3042, "step": 1509 }, { "epoch": 0.11833855799373041, "grad_norm": 3.377288341522217, "learning_rate": 9.981107631264578e-06, "loss": 1.3634, "step": 1510 }, { "epoch": 0.11841692789968652, "grad_norm": 4.7656025886535645, "learning_rate": 9.98108084099415e-06, "loss": 1.5488, "step": 1511 }, { "epoch": 0.11849529780564264, "grad_norm": 2.3410000801086426, "learning_rate": 9.98105403177825e-06, "loss": 1.6402, "step": 1512 }, { "epoch": 0.11857366771159875, "grad_norm": 2.3829667568206787, "learning_rate": 9.981027203616978e-06, "loss": 1.2792, "step": 1513 }, { "epoch": 0.11865203761755486, "grad_norm": 2.644848585128784, "learning_rate": 9.981000356510436e-06, "loss": 1.181, "step": 1514 }, { "epoch": 0.11873040752351097, "grad_norm": 3.78999924659729, "learning_rate": 9.980973490458728e-06, "loss": 1.1236, "step": 1515 }, { "epoch": 0.11880877742946709, "grad_norm": 8.264822006225586, "learning_rate": 9.980946605461957e-06, "loss": 1.6372, "step": 1516 }, { "epoch": 0.1188871473354232, "grad_norm": 3.613678455352783, "learning_rate": 9.98091970152022e-06, "loss": 1.0923, "step": 1517 }, { "epoch": 0.11896551724137931, "grad_norm": 3.872220754623413, "learning_rate": 9.980892778633625e-06, "loss": 1.7547, "step": 1518 }, { "epoch": 0.11904388714733542, "grad_norm": 2.609035015106201, "learning_rate": 9.980865836802272e-06, "loss": 1.6618, "step": 1519 }, { "epoch": 0.11912225705329153, "grad_norm": 2.3058762550354004, "learning_rate": 9.980838876026263e-06, "loss": 1.1466, "step": 1520 }, { "epoch": 0.11920062695924764, "grad_norm": 5.821981906890869, "learning_rate": 9.980811896305702e-06, "loss": 1.5353, "step": 1521 }, { "epoch": 0.11927899686520377, "grad_norm": 6.652055263519287, "learning_rate": 9.980784897640691e-06, "loss": 1.5881, "step": 1522 }, { "epoch": 0.11935736677115988, "grad_norm": 2.7887816429138184, "learning_rate": 9.980757880031332e-06, "loss": 1.1151, "step": 1523 }, { "epoch": 0.11943573667711599, "grad_norm": 4.071245193481445, "learning_rate": 9.980730843477728e-06, "loss": 1.6047, "step": 1524 }, { "epoch": 0.1195141065830721, "grad_norm": 3.0548715591430664, "learning_rate": 9.980703787979983e-06, "loss": 1.4639, "step": 1525 }, { "epoch": 0.11959247648902821, "grad_norm": 3.964127540588379, "learning_rate": 9.980676713538199e-06, "loss": 1.2628, "step": 1526 }, { "epoch": 0.11967084639498432, "grad_norm": 2.54998517036438, "learning_rate": 9.980649620152478e-06, "loss": 1.338, "step": 1527 }, { "epoch": 0.11974921630094044, "grad_norm": 2.3490912914276123, "learning_rate": 9.980622507822924e-06, "loss": 1.3244, "step": 1528 }, { "epoch": 0.11982758620689656, "grad_norm": 2.4348044395446777, "learning_rate": 9.980595376549642e-06, "loss": 1.3515, "step": 1529 }, { "epoch": 0.11990595611285267, "grad_norm": 7.8188581466674805, "learning_rate": 9.980568226332733e-06, "loss": 1.5496, "step": 1530 }, { "epoch": 0.11998432601880878, "grad_norm": 3.221895694732666, "learning_rate": 9.980541057172299e-06, "loss": 1.1431, "step": 1531 }, { "epoch": 0.12006269592476489, "grad_norm": 9.136383056640625, "learning_rate": 9.980513869068446e-06, "loss": 1.1238, "step": 1532 }, { "epoch": 0.120141065830721, "grad_norm": 7.0938801765441895, "learning_rate": 9.980486662021275e-06, "loss": 1.6335, "step": 1533 }, { "epoch": 0.12021943573667712, "grad_norm": 3.471998691558838, "learning_rate": 9.980459436030894e-06, "loss": 1.631, "step": 1534 }, { "epoch": 0.12029780564263323, "grad_norm": 3.0134665966033936, "learning_rate": 9.9804321910974e-06, "loss": 1.4258, "step": 1535 }, { "epoch": 0.12037617554858934, "grad_norm": 3.369119882583618, "learning_rate": 9.980404927220902e-06, "loss": 1.7625, "step": 1536 }, { "epoch": 0.12045454545454545, "grad_norm": 2.216365337371826, "learning_rate": 9.980377644401503e-06, "loss": 1.3353, "step": 1537 }, { "epoch": 0.12053291536050156, "grad_norm": 5.412457466125488, "learning_rate": 9.980350342639303e-06, "loss": 1.5614, "step": 1538 }, { "epoch": 0.12061128526645767, "grad_norm": 2.666163444519043, "learning_rate": 9.980323021934409e-06, "loss": 1.6098, "step": 1539 }, { "epoch": 0.1206896551724138, "grad_norm": 3.047922372817993, "learning_rate": 9.980295682286924e-06, "loss": 1.3333, "step": 1540 }, { "epoch": 0.12076802507836991, "grad_norm": 3.044497013092041, "learning_rate": 9.980268323696953e-06, "loss": 1.3349, "step": 1541 }, { "epoch": 0.12084639498432602, "grad_norm": 2.7655813694000244, "learning_rate": 9.980240946164599e-06, "loss": 1.4745, "step": 1542 }, { "epoch": 0.12092476489028213, "grad_norm": 3.2695212364196777, "learning_rate": 9.980213549689968e-06, "loss": 1.5251, "step": 1543 }, { "epoch": 0.12100313479623824, "grad_norm": 10.929529190063477, "learning_rate": 9.98018613427316e-06, "loss": 1.5722, "step": 1544 }, { "epoch": 0.12108150470219435, "grad_norm": 2.46199107170105, "learning_rate": 9.980158699914285e-06, "loss": 1.322, "step": 1545 }, { "epoch": 0.12115987460815048, "grad_norm": 4.377476692199707, "learning_rate": 9.98013124661344e-06, "loss": 1.3257, "step": 1546 }, { "epoch": 0.12123824451410659, "grad_norm": 2.915031909942627, "learning_rate": 9.980103774370737e-06, "loss": 1.3557, "step": 1547 }, { "epoch": 0.1213166144200627, "grad_norm": 2.8089489936828613, "learning_rate": 9.980076283186278e-06, "loss": 1.4121, "step": 1548 }, { "epoch": 0.1213949843260188, "grad_norm": 3.200566291809082, "learning_rate": 9.980048773060163e-06, "loss": 1.1151, "step": 1549 }, { "epoch": 0.12147335423197492, "grad_norm": 4.9181294441223145, "learning_rate": 9.980021243992504e-06, "loss": 1.847, "step": 1550 }, { "epoch": 0.12155172413793103, "grad_norm": 2.7947914600372314, "learning_rate": 9.9799936959834e-06, "loss": 1.8653, "step": 1551 }, { "epoch": 0.12163009404388715, "grad_norm": 3.605782985687256, "learning_rate": 9.979966129032958e-06, "loss": 1.4519, "step": 1552 }, { "epoch": 0.12170846394984326, "grad_norm": 6.553066253662109, "learning_rate": 9.979938543141283e-06, "loss": 1.4061, "step": 1553 }, { "epoch": 0.12178683385579937, "grad_norm": 4.839369297027588, "learning_rate": 9.97991093830848e-06, "loss": 1.2859, "step": 1554 }, { "epoch": 0.12186520376175548, "grad_norm": 3.1438891887664795, "learning_rate": 9.979883314534652e-06, "loss": 1.669, "step": 1555 }, { "epoch": 0.1219435736677116, "grad_norm": 1.837239146232605, "learning_rate": 9.979855671819909e-06, "loss": 1.5371, "step": 1556 }, { "epoch": 0.12202194357366772, "grad_norm": 2.9273743629455566, "learning_rate": 9.97982801016435e-06, "loss": 1.3511, "step": 1557 }, { "epoch": 0.12210031347962383, "grad_norm": 2.5016632080078125, "learning_rate": 9.979800329568086e-06, "loss": 1.2696, "step": 1558 }, { "epoch": 0.12217868338557994, "grad_norm": 3.1072335243225098, "learning_rate": 9.979772630031216e-06, "loss": 1.5938, "step": 1559 }, { "epoch": 0.12225705329153605, "grad_norm": 3.9796042442321777, "learning_rate": 9.97974491155385e-06, "loss": 1.4779, "step": 1560 }, { "epoch": 0.12233542319749216, "grad_norm": 4.307261943817139, "learning_rate": 9.979717174136091e-06, "loss": 1.5614, "step": 1561 }, { "epoch": 0.12241379310344827, "grad_norm": 7.392599582672119, "learning_rate": 9.97968941777805e-06, "loss": 1.5151, "step": 1562 }, { "epoch": 0.1224921630094044, "grad_norm": 2.203206777572632, "learning_rate": 9.979661642479824e-06, "loss": 1.2268, "step": 1563 }, { "epoch": 0.1225705329153605, "grad_norm": 2.636850357055664, "learning_rate": 9.979633848241523e-06, "loss": 1.5522, "step": 1564 }, { "epoch": 0.12264890282131662, "grad_norm": 3.5631535053253174, "learning_rate": 9.979606035063254e-06, "loss": 1.502, "step": 1565 }, { "epoch": 0.12272727272727273, "grad_norm": 2.924654960632324, "learning_rate": 9.97957820294512e-06, "loss": 1.6117, "step": 1566 }, { "epoch": 0.12280564263322884, "grad_norm": 6.948179244995117, "learning_rate": 9.979550351887229e-06, "loss": 2.8268, "step": 1567 }, { "epoch": 0.12288401253918495, "grad_norm": 4.383874416351318, "learning_rate": 9.979522481889685e-06, "loss": 1.9069, "step": 1568 }, { "epoch": 0.12296238244514107, "grad_norm": 3.2853307723999023, "learning_rate": 9.979494592952596e-06, "loss": 1.3572, "step": 1569 }, { "epoch": 0.12304075235109718, "grad_norm": 3.1328542232513428, "learning_rate": 9.979466685076069e-06, "loss": 1.246, "step": 1570 }, { "epoch": 0.12311912225705329, "grad_norm": 3.6925158500671387, "learning_rate": 9.979438758260207e-06, "loss": 1.3226, "step": 1571 }, { "epoch": 0.1231974921630094, "grad_norm": 3.3027710914611816, "learning_rate": 9.979410812505117e-06, "loss": 1.2936, "step": 1572 }, { "epoch": 0.12327586206896551, "grad_norm": 3.3331847190856934, "learning_rate": 9.979382847810905e-06, "loss": 1.5662, "step": 1573 }, { "epoch": 0.12335423197492162, "grad_norm": 3.149618625640869, "learning_rate": 9.979354864177679e-06, "loss": 1.5151, "step": 1574 }, { "epoch": 0.12343260188087775, "grad_norm": 5.47764253616333, "learning_rate": 9.979326861605545e-06, "loss": 1.5717, "step": 1575 }, { "epoch": 0.12351097178683386, "grad_norm": 4.933157444000244, "learning_rate": 9.97929884009461e-06, "loss": 1.3332, "step": 1576 }, { "epoch": 0.12358934169278997, "grad_norm": 2.704098701477051, "learning_rate": 9.979270799644978e-06, "loss": 1.2204, "step": 1577 }, { "epoch": 0.12366771159874608, "grad_norm": 3.3472394943237305, "learning_rate": 9.979242740256757e-06, "loss": 1.2957, "step": 1578 }, { "epoch": 0.12374608150470219, "grad_norm": 4.502384185791016, "learning_rate": 9.979214661930057e-06, "loss": 1.6506, "step": 1579 }, { "epoch": 0.1238244514106583, "grad_norm": 3.180516004562378, "learning_rate": 9.97918656466498e-06, "loss": 1.6943, "step": 1580 }, { "epoch": 0.12390282131661443, "grad_norm": 2.7337706089019775, "learning_rate": 9.979158448461636e-06, "loss": 1.4178, "step": 1581 }, { "epoch": 0.12398119122257054, "grad_norm": 7.5288543701171875, "learning_rate": 9.97913031332013e-06, "loss": 1.1763, "step": 1582 }, { "epoch": 0.12405956112852665, "grad_norm": 6.273062705993652, "learning_rate": 9.979102159240567e-06, "loss": 1.5672, "step": 1583 }, { "epoch": 0.12413793103448276, "grad_norm": 2.875115156173706, "learning_rate": 9.979073986223061e-06, "loss": 1.0572, "step": 1584 }, { "epoch": 0.12421630094043887, "grad_norm": 5.879525184631348, "learning_rate": 9.979045794267713e-06, "loss": 1.8187, "step": 1585 }, { "epoch": 0.12429467084639498, "grad_norm": 3.2079033851623535, "learning_rate": 9.979017583374632e-06, "loss": 1.496, "step": 1586 }, { "epoch": 0.1243730407523511, "grad_norm": 5.991575717926025, "learning_rate": 9.978989353543926e-06, "loss": 1.9272, "step": 1587 }, { "epoch": 0.12445141065830721, "grad_norm": 2.680631160736084, "learning_rate": 9.978961104775703e-06, "loss": 1.3361, "step": 1588 }, { "epoch": 0.12452978056426332, "grad_norm": 4.668543815612793, "learning_rate": 9.978932837070067e-06, "loss": 0.9079, "step": 1589 }, { "epoch": 0.12460815047021943, "grad_norm": 3.492755651473999, "learning_rate": 9.97890455042713e-06, "loss": 1.6245, "step": 1590 }, { "epoch": 0.12468652037617554, "grad_norm": 4.746769428253174, "learning_rate": 9.978876244846996e-06, "loss": 1.4983, "step": 1591 }, { "epoch": 0.12476489028213165, "grad_norm": 7.731174945831299, "learning_rate": 9.978847920329775e-06, "loss": 0.7249, "step": 1592 }, { "epoch": 0.12484326018808778, "grad_norm": 2.9164211750030518, "learning_rate": 9.978819576875574e-06, "loss": 1.4847, "step": 1593 }, { "epoch": 0.12492163009404389, "grad_norm": 2.8336963653564453, "learning_rate": 9.9787912144845e-06, "loss": 1.6107, "step": 1594 }, { "epoch": 0.125, "grad_norm": 3.598443031311035, "learning_rate": 9.978762833156662e-06, "loss": 1.2726, "step": 1595 }, { "epoch": 0.1250783699059561, "grad_norm": 2.454798698425293, "learning_rate": 9.978734432892166e-06, "loss": 1.5077, "step": 1596 }, { "epoch": 0.12515673981191222, "grad_norm": 4.176780700683594, "learning_rate": 9.978706013691122e-06, "loss": 1.6408, "step": 1597 }, { "epoch": 0.12523510971786833, "grad_norm": 3.230109214782715, "learning_rate": 9.978677575553636e-06, "loss": 1.7101, "step": 1598 }, { "epoch": 0.12531347962382444, "grad_norm": 2.631211042404175, "learning_rate": 9.978649118479821e-06, "loss": 1.8614, "step": 1599 }, { "epoch": 0.12539184952978055, "grad_norm": 5.289422035217285, "learning_rate": 9.97862064246978e-06, "loss": 1.4893, "step": 1600 }, { "epoch": 0.1254702194357367, "grad_norm": 5.6266865730285645, "learning_rate": 9.978592147523625e-06, "loss": 1.5118, "step": 1601 }, { "epoch": 0.1255485893416928, "grad_norm": 4.523812770843506, "learning_rate": 9.97856363364146e-06, "loss": 2.1156, "step": 1602 }, { "epoch": 0.1256269592476489, "grad_norm": 2.559937000274658, "learning_rate": 9.978535100823397e-06, "loss": 1.645, "step": 1603 }, { "epoch": 0.12570532915360502, "grad_norm": 4.196362018585205, "learning_rate": 9.978506549069544e-06, "loss": 1.5645, "step": 1604 }, { "epoch": 0.12578369905956113, "grad_norm": 2.7557127475738525, "learning_rate": 9.97847797838001e-06, "loss": 1.4119, "step": 1605 }, { "epoch": 0.12586206896551724, "grad_norm": 4.24760103225708, "learning_rate": 9.978449388754902e-06, "loss": 1.4258, "step": 1606 }, { "epoch": 0.12594043887147335, "grad_norm": 3.5713040828704834, "learning_rate": 9.978420780194332e-06, "loss": 1.3652, "step": 1607 }, { "epoch": 0.12601880877742946, "grad_norm": 3.2361435890197754, "learning_rate": 9.978392152698401e-06, "loss": 1.1036, "step": 1608 }, { "epoch": 0.12609717868338557, "grad_norm": 3.015252113342285, "learning_rate": 9.978363506267228e-06, "loss": 1.1714, "step": 1609 }, { "epoch": 0.12617554858934169, "grad_norm": 5.040140151977539, "learning_rate": 9.978334840900917e-06, "loss": 1.429, "step": 1610 }, { "epoch": 0.1262539184952978, "grad_norm": 2.8428080081939697, "learning_rate": 9.978306156599575e-06, "loss": 1.1697, "step": 1611 }, { "epoch": 0.1263322884012539, "grad_norm": 2.224212884902954, "learning_rate": 9.978277453363314e-06, "loss": 1.4927, "step": 1612 }, { "epoch": 0.12641065830721004, "grad_norm": 2.5567755699157715, "learning_rate": 9.978248731192244e-06, "loss": 1.2635, "step": 1613 }, { "epoch": 0.12648902821316615, "grad_norm": 2.871568441390991, "learning_rate": 9.978219990086472e-06, "loss": 1.6095, "step": 1614 }, { "epoch": 0.12656739811912227, "grad_norm": 5.311619758605957, "learning_rate": 9.978191230046108e-06, "loss": 1.4119, "step": 1615 }, { "epoch": 0.12664576802507838, "grad_norm": 8.797713279724121, "learning_rate": 9.97816245107126e-06, "loss": 1.723, "step": 1616 }, { "epoch": 0.1267241379310345, "grad_norm": 2.2195918560028076, "learning_rate": 9.978133653162041e-06, "loss": 1.4052, "step": 1617 }, { "epoch": 0.1268025078369906, "grad_norm": 4.355201244354248, "learning_rate": 9.978104836318558e-06, "loss": 1.0806, "step": 1618 }, { "epoch": 0.1268808777429467, "grad_norm": 3.0675854682922363, "learning_rate": 9.978076000540922e-06, "loss": 1.5057, "step": 1619 }, { "epoch": 0.12695924764890282, "grad_norm": 3.95474910736084, "learning_rate": 9.978047145829241e-06, "loss": 1.4383, "step": 1620 }, { "epoch": 0.12703761755485893, "grad_norm": 2.887902021408081, "learning_rate": 9.978018272183625e-06, "loss": 1.346, "step": 1621 }, { "epoch": 0.12711598746081504, "grad_norm": 5.785144805908203, "learning_rate": 9.977989379604185e-06, "loss": 1.5874, "step": 1622 }, { "epoch": 0.12719435736677115, "grad_norm": 2.432250738143921, "learning_rate": 9.97796046809103e-06, "loss": 1.3899, "step": 1623 }, { "epoch": 0.12727272727272726, "grad_norm": 2.689493179321289, "learning_rate": 9.977931537644271e-06, "loss": 1.5875, "step": 1624 }, { "epoch": 0.1273510971786834, "grad_norm": 2.951904773712158, "learning_rate": 9.977902588264016e-06, "loss": 1.1513, "step": 1625 }, { "epoch": 0.1274294670846395, "grad_norm": 11.843594551086426, "learning_rate": 9.977873619950378e-06, "loss": 1.8042, "step": 1626 }, { "epoch": 0.12750783699059562, "grad_norm": 2.6109328269958496, "learning_rate": 9.977844632703464e-06, "loss": 1.3693, "step": 1627 }, { "epoch": 0.12758620689655173, "grad_norm": 3.071234941482544, "learning_rate": 9.977815626523387e-06, "loss": 1.4915, "step": 1628 }, { "epoch": 0.12766457680250784, "grad_norm": 2.6655006408691406, "learning_rate": 9.977786601410255e-06, "loss": 1.679, "step": 1629 }, { "epoch": 0.12774294670846395, "grad_norm": 4.565462112426758, "learning_rate": 9.97775755736418e-06, "loss": 1.5767, "step": 1630 }, { "epoch": 0.12782131661442006, "grad_norm": 3.0722854137420654, "learning_rate": 9.977728494385273e-06, "loss": 1.3186, "step": 1631 }, { "epoch": 0.12789968652037617, "grad_norm": 2.4646692276000977, "learning_rate": 9.97769941247364e-06, "loss": 1.4371, "step": 1632 }, { "epoch": 0.12797805642633228, "grad_norm": 9.303019523620605, "learning_rate": 9.977670311629399e-06, "loss": 1.5042, "step": 1633 }, { "epoch": 0.1280564263322884, "grad_norm": 9.638708114624023, "learning_rate": 9.977641191852656e-06, "loss": 1.6052, "step": 1634 }, { "epoch": 0.1281347962382445, "grad_norm": 2.9449474811553955, "learning_rate": 9.977612053143522e-06, "loss": 1.5872, "step": 1635 }, { "epoch": 0.1282131661442006, "grad_norm": 4.179955959320068, "learning_rate": 9.977582895502109e-06, "loss": 1.8335, "step": 1636 }, { "epoch": 0.12829153605015675, "grad_norm": 2.774400472640991, "learning_rate": 9.977553718928525e-06, "loss": 1.2561, "step": 1637 }, { "epoch": 0.12836990595611286, "grad_norm": 6.144996643066406, "learning_rate": 9.977524523422886e-06, "loss": 1.4684, "step": 1638 }, { "epoch": 0.12844827586206897, "grad_norm": 3.2578110694885254, "learning_rate": 9.977495308985298e-06, "loss": 1.7979, "step": 1639 }, { "epoch": 0.12852664576802508, "grad_norm": 3.0840344429016113, "learning_rate": 9.977466075615877e-06, "loss": 1.3703, "step": 1640 }, { "epoch": 0.1286050156739812, "grad_norm": 4.518738746643066, "learning_rate": 9.97743682331473e-06, "loss": 1.7769, "step": 1641 }, { "epoch": 0.1286833855799373, "grad_norm": 4.631345272064209, "learning_rate": 9.97740755208197e-06, "loss": 2.2226, "step": 1642 }, { "epoch": 0.12876175548589341, "grad_norm": 3.161003351211548, "learning_rate": 9.977378261917709e-06, "loss": 1.4052, "step": 1643 }, { "epoch": 0.12884012539184952, "grad_norm": 3.3923070430755615, "learning_rate": 9.977348952822058e-06, "loss": 1.4983, "step": 1644 }, { "epoch": 0.12891849529780564, "grad_norm": 2.7238526344299316, "learning_rate": 9.977319624795128e-06, "loss": 1.2829, "step": 1645 }, { "epoch": 0.12899686520376175, "grad_norm": 3.6871700286865234, "learning_rate": 9.977290277837028e-06, "loss": 1.4895, "step": 1646 }, { "epoch": 0.12907523510971786, "grad_norm": 3.626192569732666, "learning_rate": 9.977260911947874e-06, "loss": 1.5573, "step": 1647 }, { "epoch": 0.129153605015674, "grad_norm": 4.393764972686768, "learning_rate": 9.977231527127778e-06, "loss": 1.0777, "step": 1648 }, { "epoch": 0.1292319749216301, "grad_norm": 3.2258031368255615, "learning_rate": 9.977202123376848e-06, "loss": 1.9391, "step": 1649 }, { "epoch": 0.12931034482758622, "grad_norm": 3.6496946811676025, "learning_rate": 9.977172700695197e-06, "loss": 1.1181, "step": 1650 }, { "epoch": 0.12938871473354233, "grad_norm": 2.5823726654052734, "learning_rate": 9.977143259082939e-06, "loss": 1.6296, "step": 1651 }, { "epoch": 0.12946708463949844, "grad_norm": 4.230770111083984, "learning_rate": 9.977113798540183e-06, "loss": 1.5681, "step": 1652 }, { "epoch": 0.12954545454545455, "grad_norm": 4.757688522338867, "learning_rate": 9.977084319067043e-06, "loss": 1.8946, "step": 1653 }, { "epoch": 0.12962382445141066, "grad_norm": 2.6630935668945312, "learning_rate": 9.977054820663631e-06, "loss": 1.3321, "step": 1654 }, { "epoch": 0.12970219435736677, "grad_norm": 10.311514854431152, "learning_rate": 9.977025303330059e-06, "loss": 1.7755, "step": 1655 }, { "epoch": 0.12978056426332288, "grad_norm": 4.332208156585693, "learning_rate": 9.976995767066438e-06, "loss": 1.0455, "step": 1656 }, { "epoch": 0.129858934169279, "grad_norm": 4.744803428649902, "learning_rate": 9.976966211872883e-06, "loss": 1.6724, "step": 1657 }, { "epoch": 0.1299373040752351, "grad_norm": 7.637454032897949, "learning_rate": 9.976936637749503e-06, "loss": 0.7037, "step": 1658 }, { "epoch": 0.1300156739811912, "grad_norm": 3.429028034210205, "learning_rate": 9.976907044696413e-06, "loss": 1.4262, "step": 1659 }, { "epoch": 0.13009404388714735, "grad_norm": 2.7604541778564453, "learning_rate": 9.976877432713725e-06, "loss": 1.3827, "step": 1660 }, { "epoch": 0.13017241379310346, "grad_norm": 2.6920244693756104, "learning_rate": 9.976847801801553e-06, "loss": 1.4527, "step": 1661 }, { "epoch": 0.13025078369905957, "grad_norm": 4.33567476272583, "learning_rate": 9.976818151960006e-06, "loss": 1.0365, "step": 1662 }, { "epoch": 0.13032915360501568, "grad_norm": 2.8285231590270996, "learning_rate": 9.9767884831892e-06, "loss": 1.4134, "step": 1663 }, { "epoch": 0.1304075235109718, "grad_norm": 3.005160331726074, "learning_rate": 9.976758795489247e-06, "loss": 1.4974, "step": 1664 }, { "epoch": 0.1304858934169279, "grad_norm": 3.6190741062164307, "learning_rate": 9.97672908886026e-06, "loss": 1.602, "step": 1665 }, { "epoch": 0.130564263322884, "grad_norm": 3.210866928100586, "learning_rate": 9.976699363302351e-06, "loss": 1.2471, "step": 1666 }, { "epoch": 0.13064263322884012, "grad_norm": 2.755608320236206, "learning_rate": 9.976669618815635e-06, "loss": 1.3068, "step": 1667 }, { "epoch": 0.13072100313479623, "grad_norm": 2.1142964363098145, "learning_rate": 9.976639855400222e-06, "loss": 1.267, "step": 1668 }, { "epoch": 0.13079937304075234, "grad_norm": 2.7660632133483887, "learning_rate": 9.976610073056227e-06, "loss": 1.4527, "step": 1669 }, { "epoch": 0.13087774294670845, "grad_norm": 2.3563520908355713, "learning_rate": 9.976580271783766e-06, "loss": 1.528, "step": 1670 }, { "epoch": 0.13095611285266456, "grad_norm": 2.61811900138855, "learning_rate": 9.97655045158295e-06, "loss": 1.4183, "step": 1671 }, { "epoch": 0.1310344827586207, "grad_norm": 3.421889305114746, "learning_rate": 9.97652061245389e-06, "loss": 1.3931, "step": 1672 }, { "epoch": 0.1311128526645768, "grad_norm": 2.8892219066619873, "learning_rate": 9.976490754396702e-06, "loss": 1.3169, "step": 1673 }, { "epoch": 0.13119122257053292, "grad_norm": 2.6872832775115967, "learning_rate": 9.976460877411498e-06, "loss": 1.4887, "step": 1674 }, { "epoch": 0.13126959247648903, "grad_norm": 3.850919485092163, "learning_rate": 9.976430981498395e-06, "loss": 1.5701, "step": 1675 }, { "epoch": 0.13134796238244514, "grad_norm": 3.3904836177825928, "learning_rate": 9.976401066657504e-06, "loss": 1.4444, "step": 1676 }, { "epoch": 0.13142633228840125, "grad_norm": 3.82306170463562, "learning_rate": 9.97637113288894e-06, "loss": 1.2198, "step": 1677 }, { "epoch": 0.13150470219435736, "grad_norm": 4.764860153198242, "learning_rate": 9.976341180192817e-06, "loss": 1.2783, "step": 1678 }, { "epoch": 0.13158307210031348, "grad_norm": 6.963048458099365, "learning_rate": 9.976311208569246e-06, "loss": 1.6998, "step": 1679 }, { "epoch": 0.13166144200626959, "grad_norm": 3.1255745887756348, "learning_rate": 9.976281218018345e-06, "loss": 1.3076, "step": 1680 }, { "epoch": 0.1317398119122257, "grad_norm": 4.4072980880737305, "learning_rate": 9.976251208540224e-06, "loss": 1.7645, "step": 1681 }, { "epoch": 0.1318181818181818, "grad_norm": 5.518395900726318, "learning_rate": 9.976221180135002e-06, "loss": 1.401, "step": 1682 }, { "epoch": 0.13189655172413794, "grad_norm": 5.01060676574707, "learning_rate": 9.976191132802789e-06, "loss": 1.6782, "step": 1683 }, { "epoch": 0.13197492163009406, "grad_norm": 3.337641716003418, "learning_rate": 9.9761610665437e-06, "loss": 1.8737, "step": 1684 }, { "epoch": 0.13205329153605017, "grad_norm": 3.2720775604248047, "learning_rate": 9.976130981357852e-06, "loss": 1.6816, "step": 1685 }, { "epoch": 0.13213166144200628, "grad_norm": 3.451359748840332, "learning_rate": 9.976100877245356e-06, "loss": 1.2699, "step": 1686 }, { "epoch": 0.1322100313479624, "grad_norm": 2.8944008350372314, "learning_rate": 9.97607075420633e-06, "loss": 1.4064, "step": 1687 }, { "epoch": 0.1322884012539185, "grad_norm": 4.593804359436035, "learning_rate": 9.976040612240885e-06, "loss": 0.9998, "step": 1688 }, { "epoch": 0.1323667711598746, "grad_norm": 3.5950002670288086, "learning_rate": 9.976010451349139e-06, "loss": 1.4104, "step": 1689 }, { "epoch": 0.13244514106583072, "grad_norm": 4.015209674835205, "learning_rate": 9.975980271531205e-06, "loss": 1.353, "step": 1690 }, { "epoch": 0.13252351097178683, "grad_norm": 3.114668607711792, "learning_rate": 9.975950072787197e-06, "loss": 1.431, "step": 1691 }, { "epoch": 0.13260188087774294, "grad_norm": 4.219808101654053, "learning_rate": 9.97591985511723e-06, "loss": 1.4875, "step": 1692 }, { "epoch": 0.13268025078369905, "grad_norm": 3.655156373977661, "learning_rate": 9.97588961852142e-06, "loss": 1.6941, "step": 1693 }, { "epoch": 0.13275862068965516, "grad_norm": 2.8486037254333496, "learning_rate": 9.975859362999883e-06, "loss": 1.2619, "step": 1694 }, { "epoch": 0.1328369905956113, "grad_norm": 4.180792331695557, "learning_rate": 9.975829088552732e-06, "loss": 1.5715, "step": 1695 }, { "epoch": 0.1329153605015674, "grad_norm": 3.416055202484131, "learning_rate": 9.975798795180082e-06, "loss": 1.2166, "step": 1696 }, { "epoch": 0.13299373040752352, "grad_norm": 3.357898473739624, "learning_rate": 9.975768482882051e-06, "loss": 1.5605, "step": 1697 }, { "epoch": 0.13307210031347963, "grad_norm": 3.0318100452423096, "learning_rate": 9.975738151658751e-06, "loss": 1.5031, "step": 1698 }, { "epoch": 0.13315047021943574, "grad_norm": 2.5280356407165527, "learning_rate": 9.975707801510299e-06, "loss": 1.7224, "step": 1699 }, { "epoch": 0.13322884012539185, "grad_norm": 5.090761661529541, "learning_rate": 9.97567743243681e-06, "loss": 1.8654, "step": 1700 }, { "epoch": 0.13330721003134796, "grad_norm": 4.566983699798584, "learning_rate": 9.975647044438401e-06, "loss": 1.4415, "step": 1701 }, { "epoch": 0.13338557993730407, "grad_norm": 2.8746392726898193, "learning_rate": 9.975616637515186e-06, "loss": 1.5564, "step": 1702 }, { "epoch": 0.13346394984326018, "grad_norm": 3.5105185508728027, "learning_rate": 9.975586211667279e-06, "loss": 1.152, "step": 1703 }, { "epoch": 0.1335423197492163, "grad_norm": 4.072345733642578, "learning_rate": 9.9755557668948e-06, "loss": 1.5629, "step": 1704 }, { "epoch": 0.1336206896551724, "grad_norm": 2.6935434341430664, "learning_rate": 9.975525303197862e-06, "loss": 1.365, "step": 1705 }, { "epoch": 0.1336990595611285, "grad_norm": 5.67743444442749, "learning_rate": 9.97549482057658e-06, "loss": 1.3201, "step": 1706 }, { "epoch": 0.13377742946708465, "grad_norm": 11.194258689880371, "learning_rate": 9.975464319031071e-06, "loss": 0.8979, "step": 1707 }, { "epoch": 0.13385579937304076, "grad_norm": 4.840267658233643, "learning_rate": 9.975433798561454e-06, "loss": 1.344, "step": 1708 }, { "epoch": 0.13393416927899687, "grad_norm": 2.631603717803955, "learning_rate": 9.975403259167839e-06, "loss": 1.1133, "step": 1709 }, { "epoch": 0.13401253918495298, "grad_norm": 2.752401113510132, "learning_rate": 9.975372700850347e-06, "loss": 1.519, "step": 1710 }, { "epoch": 0.1340909090909091, "grad_norm": 3.8409218788146973, "learning_rate": 9.975342123609094e-06, "loss": 1.5913, "step": 1711 }, { "epoch": 0.1341692789968652, "grad_norm": 3.458341121673584, "learning_rate": 9.975311527444193e-06, "loss": 1.5015, "step": 1712 }, { "epoch": 0.13424764890282131, "grad_norm": 3.8928158283233643, "learning_rate": 9.975280912355765e-06, "loss": 1.3749, "step": 1713 }, { "epoch": 0.13432601880877743, "grad_norm": 3.704148054122925, "learning_rate": 9.975250278343923e-06, "loss": 1.2243, "step": 1714 }, { "epoch": 0.13440438871473354, "grad_norm": 3.525573253631592, "learning_rate": 9.975219625408783e-06, "loss": 1.1316, "step": 1715 }, { "epoch": 0.13448275862068965, "grad_norm": 3.086108446121216, "learning_rate": 9.975188953550464e-06, "loss": 1.4942, "step": 1716 }, { "epoch": 0.13456112852664576, "grad_norm": 2.2871720790863037, "learning_rate": 9.97515826276908e-06, "loss": 0.9843, "step": 1717 }, { "epoch": 0.13463949843260187, "grad_norm": 3.4875786304473877, "learning_rate": 9.975127553064751e-06, "loss": 1.3881, "step": 1718 }, { "epoch": 0.134717868338558, "grad_norm": 3.380866050720215, "learning_rate": 9.975096824437594e-06, "loss": 1.3389, "step": 1719 }, { "epoch": 0.13479623824451412, "grad_norm": 4.731466293334961, "learning_rate": 9.97506607688772e-06, "loss": 1.5872, "step": 1720 }, { "epoch": 0.13487460815047023, "grad_norm": 8.775562286376953, "learning_rate": 9.975035310415253e-06, "loss": 1.0078, "step": 1721 }, { "epoch": 0.13495297805642634, "grad_norm": 3.3407416343688965, "learning_rate": 9.975004525020307e-06, "loss": 1.646, "step": 1722 }, { "epoch": 0.13503134796238245, "grad_norm": 3.685396671295166, "learning_rate": 9.974973720702997e-06, "loss": 1.8565, "step": 1723 }, { "epoch": 0.13510971786833856, "grad_norm": 4.022880554199219, "learning_rate": 9.974942897463444e-06, "loss": 1.6364, "step": 1724 }, { "epoch": 0.13518808777429467, "grad_norm": 13.993475914001465, "learning_rate": 9.974912055301762e-06, "loss": 2.1574, "step": 1725 }, { "epoch": 0.13526645768025078, "grad_norm": 3.310593366622925, "learning_rate": 9.974881194218071e-06, "loss": 1.2093, "step": 1726 }, { "epoch": 0.1353448275862069, "grad_norm": 9.389638900756836, "learning_rate": 9.974850314212487e-06, "loss": 2.2157, "step": 1727 }, { "epoch": 0.135423197492163, "grad_norm": 3.7406888008117676, "learning_rate": 9.974819415285128e-06, "loss": 1.393, "step": 1728 }, { "epoch": 0.1355015673981191, "grad_norm": 3.0227510929107666, "learning_rate": 9.974788497436111e-06, "loss": 1.6204, "step": 1729 }, { "epoch": 0.13557993730407525, "grad_norm": 3.7903082370758057, "learning_rate": 9.974757560665554e-06, "loss": 1.5339, "step": 1730 }, { "epoch": 0.13565830721003136, "grad_norm": 3.1030263900756836, "learning_rate": 9.974726604973573e-06, "loss": 1.3591, "step": 1731 }, { "epoch": 0.13573667711598747, "grad_norm": 2.4516398906707764, "learning_rate": 9.974695630360289e-06, "loss": 1.2842, "step": 1732 }, { "epoch": 0.13581504702194358, "grad_norm": 7.328332424163818, "learning_rate": 9.974664636825817e-06, "loss": 2.4683, "step": 1733 }, { "epoch": 0.1358934169278997, "grad_norm": 3.742814302444458, "learning_rate": 9.974633624370276e-06, "loss": 1.4523, "step": 1734 }, { "epoch": 0.1359717868338558, "grad_norm": 2.3395509719848633, "learning_rate": 9.974602592993783e-06, "loss": 1.4885, "step": 1735 }, { "epoch": 0.1360501567398119, "grad_norm": 2.6180524826049805, "learning_rate": 9.974571542696459e-06, "loss": 1.3747, "step": 1736 }, { "epoch": 0.13612852664576802, "grad_norm": 2.2349419593811035, "learning_rate": 9.974540473478417e-06, "loss": 1.635, "step": 1737 }, { "epoch": 0.13620689655172413, "grad_norm": 8.760351181030273, "learning_rate": 9.974509385339781e-06, "loss": 1.593, "step": 1738 }, { "epoch": 0.13628526645768024, "grad_norm": 3.667074680328369, "learning_rate": 9.974478278280665e-06, "loss": 1.3052, "step": 1739 }, { "epoch": 0.13636363636363635, "grad_norm": 6.556699752807617, "learning_rate": 9.974447152301187e-06, "loss": 1.3928, "step": 1740 }, { "epoch": 0.13644200626959246, "grad_norm": 3.1611077785491943, "learning_rate": 9.974416007401471e-06, "loss": 1.1342, "step": 1741 }, { "epoch": 0.1365203761755486, "grad_norm": 2.656315803527832, "learning_rate": 9.974384843581629e-06, "loss": 1.7749, "step": 1742 }, { "epoch": 0.1365987460815047, "grad_norm": 4.2507195472717285, "learning_rate": 9.974353660841782e-06, "loss": 1.3101, "step": 1743 }, { "epoch": 0.13667711598746082, "grad_norm": 2.808974027633667, "learning_rate": 9.974322459182048e-06, "loss": 0.9466, "step": 1744 }, { "epoch": 0.13675548589341693, "grad_norm": 2.3591277599334717, "learning_rate": 9.97429123860255e-06, "loss": 1.4718, "step": 1745 }, { "epoch": 0.13683385579937304, "grad_norm": 4.075984001159668, "learning_rate": 9.9742599991034e-06, "loss": 1.7283, "step": 1746 }, { "epoch": 0.13691222570532915, "grad_norm": 4.618869781494141, "learning_rate": 9.97422874068472e-06, "loss": 1.1724, "step": 1747 }, { "epoch": 0.13699059561128527, "grad_norm": 4.37303352355957, "learning_rate": 9.97419746334663e-06, "loss": 1.5793, "step": 1748 }, { "epoch": 0.13706896551724138, "grad_norm": 2.9173049926757812, "learning_rate": 9.974166167089245e-06, "loss": 1.6536, "step": 1749 }, { "epoch": 0.1371473354231975, "grad_norm": 2.9214017391204834, "learning_rate": 9.974134851912688e-06, "loss": 1.497, "step": 1750 }, { "epoch": 0.1372257053291536, "grad_norm": 4.657212257385254, "learning_rate": 9.97410351781708e-06, "loss": 0.9242, "step": 1751 }, { "epoch": 0.1373040752351097, "grad_norm": 2.663430690765381, "learning_rate": 9.974072164802535e-06, "loss": 1.5511, "step": 1752 }, { "epoch": 0.13738244514106582, "grad_norm": 3.615481376647949, "learning_rate": 9.974040792869172e-06, "loss": 1.9546, "step": 1753 }, { "epoch": 0.13746081504702196, "grad_norm": 2.4577834606170654, "learning_rate": 9.974009402017117e-06, "loss": 1.42, "step": 1754 }, { "epoch": 0.13753918495297807, "grad_norm": 2.88028621673584, "learning_rate": 9.973977992246482e-06, "loss": 1.4541, "step": 1755 }, { "epoch": 0.13761755485893418, "grad_norm": 2.218517780303955, "learning_rate": 9.97394656355739e-06, "loss": 1.3956, "step": 1756 }, { "epoch": 0.1376959247648903, "grad_norm": 3.0479190349578857, "learning_rate": 9.973915115949961e-06, "loss": 1.4652, "step": 1757 }, { "epoch": 0.1377742946708464, "grad_norm": 11.067258834838867, "learning_rate": 9.973883649424312e-06, "loss": 0.7894, "step": 1758 }, { "epoch": 0.1378526645768025, "grad_norm": 4.018172740936279, "learning_rate": 9.973852163980566e-06, "loss": 1.4495, "step": 1759 }, { "epoch": 0.13793103448275862, "grad_norm": 2.6561827659606934, "learning_rate": 9.97382065961884e-06, "loss": 1.4285, "step": 1760 }, { "epoch": 0.13800940438871473, "grad_norm": 6.107227802276611, "learning_rate": 9.973789136339256e-06, "loss": 1.1718, "step": 1761 }, { "epoch": 0.13808777429467084, "grad_norm": 3.995917320251465, "learning_rate": 9.973757594141933e-06, "loss": 1.7053, "step": 1762 }, { "epoch": 0.13816614420062695, "grad_norm": 3.074949264526367, "learning_rate": 9.973726033026989e-06, "loss": 1.6454, "step": 1763 }, { "epoch": 0.13824451410658306, "grad_norm": 2.780745267868042, "learning_rate": 9.973694452994546e-06, "loss": 1.6121, "step": 1764 }, { "epoch": 0.13832288401253917, "grad_norm": 3.818877935409546, "learning_rate": 9.973662854044725e-06, "loss": 1.4199, "step": 1765 }, { "epoch": 0.1384012539184953, "grad_norm": 3.3719775676727295, "learning_rate": 9.973631236177645e-06, "loss": 1.4676, "step": 1766 }, { "epoch": 0.13847962382445142, "grad_norm": 3.49586820602417, "learning_rate": 9.973599599393427e-06, "loss": 1.3292, "step": 1767 }, { "epoch": 0.13855799373040753, "grad_norm": 5.324405193328857, "learning_rate": 9.97356794369219e-06, "loss": 1.8316, "step": 1768 }, { "epoch": 0.13863636363636364, "grad_norm": 2.696380376815796, "learning_rate": 9.973536269074055e-06, "loss": 1.4397, "step": 1769 }, { "epoch": 0.13871473354231975, "grad_norm": 2.053300619125366, "learning_rate": 9.973504575539143e-06, "loss": 1.1853, "step": 1770 }, { "epoch": 0.13879310344827586, "grad_norm": 3.055752992630005, "learning_rate": 9.973472863087575e-06, "loss": 1.2798, "step": 1771 }, { "epoch": 0.13887147335423197, "grad_norm": 4.157932281494141, "learning_rate": 9.97344113171947e-06, "loss": 1.9162, "step": 1772 }, { "epoch": 0.13894984326018808, "grad_norm": 4.143520355224609, "learning_rate": 9.973409381434948e-06, "loss": 1.7992, "step": 1773 }, { "epoch": 0.1390282131661442, "grad_norm": 3.553722381591797, "learning_rate": 9.973377612234132e-06, "loss": 1.3956, "step": 1774 }, { "epoch": 0.1391065830721003, "grad_norm": 3.1411635875701904, "learning_rate": 9.973345824117143e-06, "loss": 1.385, "step": 1775 }, { "epoch": 0.13918495297805641, "grad_norm": 3.260688304901123, "learning_rate": 9.973314017084102e-06, "loss": 1.5086, "step": 1776 }, { "epoch": 0.13926332288401255, "grad_norm": 2.327742338180542, "learning_rate": 9.973282191135126e-06, "loss": 1.6567, "step": 1777 }, { "epoch": 0.13934169278996866, "grad_norm": 3.092038631439209, "learning_rate": 9.97325034627034e-06, "loss": 1.6088, "step": 1778 }, { "epoch": 0.13942006269592477, "grad_norm": 5.166444778442383, "learning_rate": 9.973218482489866e-06, "loss": 1.5067, "step": 1779 }, { "epoch": 0.13949843260188088, "grad_norm": 6.159828186035156, "learning_rate": 9.97318659979382e-06, "loss": 1.6723, "step": 1780 }, { "epoch": 0.139576802507837, "grad_norm": 3.638559341430664, "learning_rate": 9.973154698182328e-06, "loss": 1.2134, "step": 1781 }, { "epoch": 0.1396551724137931, "grad_norm": 4.882579326629639, "learning_rate": 9.97312277765551e-06, "loss": 1.6384, "step": 1782 }, { "epoch": 0.13973354231974922, "grad_norm": 4.552928447723389, "learning_rate": 9.973090838213488e-06, "loss": 1.5683, "step": 1783 }, { "epoch": 0.13981191222570533, "grad_norm": 3.0576350688934326, "learning_rate": 9.973058879856383e-06, "loss": 1.314, "step": 1784 }, { "epoch": 0.13989028213166144, "grad_norm": 9.638943672180176, "learning_rate": 9.973026902584316e-06, "loss": 1.8015, "step": 1785 }, { "epoch": 0.13996865203761755, "grad_norm": 3.121521234512329, "learning_rate": 9.972994906397408e-06, "loss": 1.5657, "step": 1786 }, { "epoch": 0.14004702194357366, "grad_norm": 5.886914253234863, "learning_rate": 9.972962891295782e-06, "loss": 1.7292, "step": 1787 }, { "epoch": 0.14012539184952977, "grad_norm": 2.5360820293426514, "learning_rate": 9.972930857279562e-06, "loss": 1.2772, "step": 1788 }, { "epoch": 0.1402037617554859, "grad_norm": 5.798094272613525, "learning_rate": 9.972898804348863e-06, "loss": 1.1437, "step": 1789 }, { "epoch": 0.14028213166144202, "grad_norm": 3.195716142654419, "learning_rate": 9.972866732503814e-06, "loss": 1.4993, "step": 1790 }, { "epoch": 0.14036050156739813, "grad_norm": 2.5833892822265625, "learning_rate": 9.972834641744533e-06, "loss": 1.5533, "step": 1791 }, { "epoch": 0.14043887147335424, "grad_norm": 3.357452869415283, "learning_rate": 9.972802532071145e-06, "loss": 1.6564, "step": 1792 }, { "epoch": 0.14051724137931035, "grad_norm": 2.7330856323242188, "learning_rate": 9.972770403483767e-06, "loss": 1.4532, "step": 1793 }, { "epoch": 0.14059561128526646, "grad_norm": 2.7371907234191895, "learning_rate": 9.972738255982528e-06, "loss": 1.3208, "step": 1794 }, { "epoch": 0.14067398119122257, "grad_norm": 2.5383362770080566, "learning_rate": 9.972706089567546e-06, "loss": 1.1296, "step": 1795 }, { "epoch": 0.14075235109717868, "grad_norm": 2.915083169937134, "learning_rate": 9.972673904238942e-06, "loss": 1.3293, "step": 1796 }, { "epoch": 0.1408307210031348, "grad_norm": 3.660759687423706, "learning_rate": 9.972641699996842e-06, "loss": 1.135, "step": 1797 }, { "epoch": 0.1409090909090909, "grad_norm": 3.4988949298858643, "learning_rate": 9.972609476841368e-06, "loss": 1.7591, "step": 1798 }, { "epoch": 0.140987460815047, "grad_norm": 10.442649841308594, "learning_rate": 9.97257723477264e-06, "loss": 2.1733, "step": 1799 }, { "epoch": 0.14106583072100312, "grad_norm": 2.260491371154785, "learning_rate": 9.972544973790783e-06, "loss": 1.5538, "step": 1800 }, { "epoch": 0.14114420062695926, "grad_norm": 6.194129467010498, "learning_rate": 9.972512693895919e-06, "loss": 1.4435, "step": 1801 }, { "epoch": 0.14122257053291537, "grad_norm": 3.5760815143585205, "learning_rate": 9.972480395088171e-06, "loss": 1.654, "step": 1802 }, { "epoch": 0.14130094043887148, "grad_norm": 5.688802242279053, "learning_rate": 9.972448077367662e-06, "loss": 1.5174, "step": 1803 }, { "epoch": 0.1413793103448276, "grad_norm": 2.9021763801574707, "learning_rate": 9.972415740734514e-06, "loss": 1.6126, "step": 1804 }, { "epoch": 0.1414576802507837, "grad_norm": 2.767716407775879, "learning_rate": 9.972383385188851e-06, "loss": 1.5745, "step": 1805 }, { "epoch": 0.1415360501567398, "grad_norm": 3.1831815242767334, "learning_rate": 9.972351010730793e-06, "loss": 1.4053, "step": 1806 }, { "epoch": 0.14161442006269592, "grad_norm": 3.6612186431884766, "learning_rate": 9.97231861736047e-06, "loss": 1.4516, "step": 1807 }, { "epoch": 0.14169278996865203, "grad_norm": 6.448884963989258, "learning_rate": 9.972286205077999e-06, "loss": 1.2598, "step": 1808 }, { "epoch": 0.14177115987460814, "grad_norm": 10.409411430358887, "learning_rate": 9.972253773883505e-06, "loss": 1.3912, "step": 1809 }, { "epoch": 0.14184952978056425, "grad_norm": 3.203765392303467, "learning_rate": 9.97222132377711e-06, "loss": 1.7173, "step": 1810 }, { "epoch": 0.14192789968652036, "grad_norm": 4.561257839202881, "learning_rate": 9.972188854758941e-06, "loss": 1.2387, "step": 1811 }, { "epoch": 0.1420062695924765, "grad_norm": 3.18072772026062, "learning_rate": 9.97215636682912e-06, "loss": 1.0973, "step": 1812 }, { "epoch": 0.1420846394984326, "grad_norm": 3.1051368713378906, "learning_rate": 9.972123859987768e-06, "loss": 1.2775, "step": 1813 }, { "epoch": 0.14216300940438872, "grad_norm": 4.031684398651123, "learning_rate": 9.972091334235013e-06, "loss": 1.7015, "step": 1814 }, { "epoch": 0.14224137931034483, "grad_norm": 2.2371227741241455, "learning_rate": 9.972058789570973e-06, "loss": 1.2495, "step": 1815 }, { "epoch": 0.14231974921630094, "grad_norm": 2.324099540710449, "learning_rate": 9.972026225995778e-06, "loss": 1.4693, "step": 1816 }, { "epoch": 0.14239811912225706, "grad_norm": 3.1276533603668213, "learning_rate": 9.971993643509547e-06, "loss": 1.5533, "step": 1817 }, { "epoch": 0.14247648902821317, "grad_norm": 4.879640579223633, "learning_rate": 9.971961042112408e-06, "loss": 1.3778, "step": 1818 }, { "epoch": 0.14255485893416928, "grad_norm": 6.648894786834717, "learning_rate": 9.971928421804481e-06, "loss": 1.2074, "step": 1819 }, { "epoch": 0.1426332288401254, "grad_norm": 3.307549476623535, "learning_rate": 9.971895782585895e-06, "loss": 1.1944, "step": 1820 }, { "epoch": 0.1427115987460815, "grad_norm": 2.749541997909546, "learning_rate": 9.971863124456769e-06, "loss": 1.5383, "step": 1821 }, { "epoch": 0.1427899686520376, "grad_norm": 7.208172798156738, "learning_rate": 9.97183044741723e-06, "loss": 2.1159, "step": 1822 }, { "epoch": 0.14286833855799372, "grad_norm": 4.154603481292725, "learning_rate": 9.971797751467402e-06, "loss": 1.6879, "step": 1823 }, { "epoch": 0.14294670846394986, "grad_norm": 3.936985492706299, "learning_rate": 9.971765036607408e-06, "loss": 1.0174, "step": 1824 }, { "epoch": 0.14302507836990597, "grad_norm": 4.298600673675537, "learning_rate": 9.971732302837373e-06, "loss": 1.0775, "step": 1825 }, { "epoch": 0.14310344827586208, "grad_norm": 4.640676498413086, "learning_rate": 9.971699550157424e-06, "loss": 1.5865, "step": 1826 }, { "epoch": 0.1431818181818182, "grad_norm": 8.031147003173828, "learning_rate": 9.971666778567682e-06, "loss": 1.5667, "step": 1827 }, { "epoch": 0.1432601880877743, "grad_norm": 6.130635738372803, "learning_rate": 9.971633988068275e-06, "loss": 1.3784, "step": 1828 }, { "epoch": 0.1433385579937304, "grad_norm": 4.408290863037109, "learning_rate": 9.971601178659324e-06, "loss": 1.4993, "step": 1829 }, { "epoch": 0.14341692789968652, "grad_norm": 7.224981784820557, "learning_rate": 9.971568350340955e-06, "loss": 1.2048, "step": 1830 }, { "epoch": 0.14349529780564263, "grad_norm": 15.260326385498047, "learning_rate": 9.971535503113295e-06, "loss": 2.5597, "step": 1831 }, { "epoch": 0.14357366771159874, "grad_norm": 2.892456531524658, "learning_rate": 9.971502636976467e-06, "loss": 1.7017, "step": 1832 }, { "epoch": 0.14365203761755485, "grad_norm": 4.352650165557861, "learning_rate": 9.971469751930598e-06, "loss": 1.3754, "step": 1833 }, { "epoch": 0.14373040752351096, "grad_norm": 8.137859344482422, "learning_rate": 9.971436847975811e-06, "loss": 1.4224, "step": 1834 }, { "epoch": 0.14380877742946707, "grad_norm": 2.9430348873138428, "learning_rate": 9.97140392511223e-06, "loss": 0.8146, "step": 1835 }, { "epoch": 0.1438871473354232, "grad_norm": 4.195640563964844, "learning_rate": 9.971370983339983e-06, "loss": 1.2647, "step": 1836 }, { "epoch": 0.14396551724137932, "grad_norm": 4.052323341369629, "learning_rate": 9.971338022659193e-06, "loss": 1.2834, "step": 1837 }, { "epoch": 0.14404388714733543, "grad_norm": 4.8244733810424805, "learning_rate": 9.971305043069988e-06, "loss": 1.4212, "step": 1838 }, { "epoch": 0.14412225705329154, "grad_norm": 2.7448372840881348, "learning_rate": 9.971272044572494e-06, "loss": 1.5065, "step": 1839 }, { "epoch": 0.14420062695924765, "grad_norm": 2.8347649574279785, "learning_rate": 9.971239027166832e-06, "loss": 1.2466, "step": 1840 }, { "epoch": 0.14427899686520376, "grad_norm": 3.67577862739563, "learning_rate": 9.97120599085313e-06, "loss": 1.2915, "step": 1841 }, { "epoch": 0.14435736677115987, "grad_norm": 7.462957859039307, "learning_rate": 9.971172935631515e-06, "loss": 1.5441, "step": 1842 }, { "epoch": 0.14443573667711598, "grad_norm": 4.396081924438477, "learning_rate": 9.97113986150211e-06, "loss": 1.4398, "step": 1843 }, { "epoch": 0.1445141065830721, "grad_norm": 2.0170629024505615, "learning_rate": 9.971106768465042e-06, "loss": 1.1799, "step": 1844 }, { "epoch": 0.1445924764890282, "grad_norm": 2.8135159015655518, "learning_rate": 9.97107365652044e-06, "loss": 1.1888, "step": 1845 }, { "epoch": 0.14467084639498431, "grad_norm": 3.2568085193634033, "learning_rate": 9.971040525668424e-06, "loss": 1.3857, "step": 1846 }, { "epoch": 0.14474921630094043, "grad_norm": 2.883261203765869, "learning_rate": 9.971007375909125e-06, "loss": 1.3115, "step": 1847 }, { "epoch": 0.14482758620689656, "grad_norm": 4.241835594177246, "learning_rate": 9.970974207242667e-06, "loss": 1.5895, "step": 1848 }, { "epoch": 0.14490595611285267, "grad_norm": 5.312651634216309, "learning_rate": 9.970941019669176e-06, "loss": 2.2529, "step": 1849 }, { "epoch": 0.14498432601880878, "grad_norm": 2.9158878326416016, "learning_rate": 9.970907813188779e-06, "loss": 1.3443, "step": 1850 }, { "epoch": 0.1450626959247649, "grad_norm": 2.4045045375823975, "learning_rate": 9.970874587801602e-06, "loss": 1.2505, "step": 1851 }, { "epoch": 0.145141065830721, "grad_norm": 2.028628349304199, "learning_rate": 9.970841343507771e-06, "loss": 1.3373, "step": 1852 }, { "epoch": 0.14521943573667712, "grad_norm": 3.391632556915283, "learning_rate": 9.970808080307412e-06, "loss": 1.4785, "step": 1853 }, { "epoch": 0.14529780564263323, "grad_norm": 3.3227667808532715, "learning_rate": 9.970774798200653e-06, "loss": 1.1629, "step": 1854 }, { "epoch": 0.14537617554858934, "grad_norm": 4.596421718597412, "learning_rate": 9.97074149718762e-06, "loss": 0.9811, "step": 1855 }, { "epoch": 0.14545454545454545, "grad_norm": 4.104040145874023, "learning_rate": 9.970708177268439e-06, "loss": 1.4229, "step": 1856 }, { "epoch": 0.14553291536050156, "grad_norm": 3.634878635406494, "learning_rate": 9.970674838443237e-06, "loss": 1.602, "step": 1857 }, { "epoch": 0.14561128526645767, "grad_norm": 2.8559982776641846, "learning_rate": 9.970641480712143e-06, "loss": 1.2813, "step": 1858 }, { "epoch": 0.1456896551724138, "grad_norm": 2.5463171005249023, "learning_rate": 9.97060810407528e-06, "loss": 1.3604, "step": 1859 }, { "epoch": 0.14576802507836992, "grad_norm": 3.2006540298461914, "learning_rate": 9.970574708532777e-06, "loss": 1.2768, "step": 1860 }, { "epoch": 0.14584639498432603, "grad_norm": 2.8872008323669434, "learning_rate": 9.97054129408476e-06, "loss": 1.6656, "step": 1861 }, { "epoch": 0.14592476489028214, "grad_norm": 4.140438079833984, "learning_rate": 9.970507860731357e-06, "loss": 1.2261, "step": 1862 }, { "epoch": 0.14600313479623825, "grad_norm": 9.799880981445312, "learning_rate": 9.970474408472697e-06, "loss": 1.3432, "step": 1863 }, { "epoch": 0.14608150470219436, "grad_norm": 3.5664827823638916, "learning_rate": 9.970440937308904e-06, "loss": 1.7171, "step": 1864 }, { "epoch": 0.14615987460815047, "grad_norm": 2.9408652782440186, "learning_rate": 9.970407447240107e-06, "loss": 1.6193, "step": 1865 }, { "epoch": 0.14623824451410658, "grad_norm": 3.413858413696289, "learning_rate": 9.970373938266434e-06, "loss": 1.6013, "step": 1866 }, { "epoch": 0.1463166144200627, "grad_norm": 2.782902479171753, "learning_rate": 9.97034041038801e-06, "loss": 1.4088, "step": 1867 }, { "epoch": 0.1463949843260188, "grad_norm": 3.498619556427002, "learning_rate": 9.970306863604964e-06, "loss": 1.234, "step": 1868 }, { "epoch": 0.1464733542319749, "grad_norm": 5.77226448059082, "learning_rate": 9.970273297917425e-06, "loss": 1.3866, "step": 1869 }, { "epoch": 0.14655172413793102, "grad_norm": 3.3328285217285156, "learning_rate": 9.970239713325518e-06, "loss": 1.6057, "step": 1870 }, { "epoch": 0.14663009404388716, "grad_norm": 7.0485100746154785, "learning_rate": 9.970206109829373e-06, "loss": 0.9312, "step": 1871 }, { "epoch": 0.14670846394984327, "grad_norm": 2.9748058319091797, "learning_rate": 9.970172487429114e-06, "loss": 1.5079, "step": 1872 }, { "epoch": 0.14678683385579938, "grad_norm": 3.7146167755126953, "learning_rate": 9.970138846124876e-06, "loss": 1.9475, "step": 1873 }, { "epoch": 0.1468652037617555, "grad_norm": 2.437163829803467, "learning_rate": 9.970105185916781e-06, "loss": 1.5553, "step": 1874 }, { "epoch": 0.1469435736677116, "grad_norm": 3.0166540145874023, "learning_rate": 9.970071506804958e-06, "loss": 1.2049, "step": 1875 }, { "epoch": 0.1470219435736677, "grad_norm": 5.648652076721191, "learning_rate": 9.970037808789536e-06, "loss": 1.409, "step": 1876 }, { "epoch": 0.14710031347962382, "grad_norm": 3.7198972702026367, "learning_rate": 9.970004091870642e-06, "loss": 1.5394, "step": 1877 }, { "epoch": 0.14717868338557993, "grad_norm": 3.9857091903686523, "learning_rate": 9.969970356048406e-06, "loss": 1.689, "step": 1878 }, { "epoch": 0.14725705329153604, "grad_norm": 3.0974509716033936, "learning_rate": 9.969936601322956e-06, "loss": 1.2783, "step": 1879 }, { "epoch": 0.14733542319749215, "grad_norm": 3.27176833152771, "learning_rate": 9.969902827694419e-06, "loss": 1.8109, "step": 1880 }, { "epoch": 0.14741379310344827, "grad_norm": 2.8237392902374268, "learning_rate": 9.969869035162926e-06, "loss": 1.4374, "step": 1881 }, { "epoch": 0.14749216300940438, "grad_norm": 2.916233777999878, "learning_rate": 9.969835223728603e-06, "loss": 1.5079, "step": 1882 }, { "epoch": 0.1475705329153605, "grad_norm": 3.0367448329925537, "learning_rate": 9.969801393391578e-06, "loss": 1.3655, "step": 1883 }, { "epoch": 0.14764890282131662, "grad_norm": 2.8044281005859375, "learning_rate": 9.969767544151984e-06, "loss": 1.4716, "step": 1884 }, { "epoch": 0.14772727272727273, "grad_norm": 2.373551368713379, "learning_rate": 9.969733676009945e-06, "loss": 1.2411, "step": 1885 }, { "epoch": 0.14780564263322885, "grad_norm": 3.83211088180542, "learning_rate": 9.969699788965592e-06, "loss": 1.3717, "step": 1886 }, { "epoch": 0.14788401253918496, "grad_norm": 2.7472870349884033, "learning_rate": 9.969665883019054e-06, "loss": 1.3369, "step": 1887 }, { "epoch": 0.14796238244514107, "grad_norm": 6.646710395812988, "learning_rate": 9.969631958170458e-06, "loss": 1.7613, "step": 1888 }, { "epoch": 0.14804075235109718, "grad_norm": 6.151742458343506, "learning_rate": 9.969598014419936e-06, "loss": 1.6421, "step": 1889 }, { "epoch": 0.1481191222570533, "grad_norm": 3.0673270225524902, "learning_rate": 9.969564051767616e-06, "loss": 1.2521, "step": 1890 }, { "epoch": 0.1481974921630094, "grad_norm": 4.08449649810791, "learning_rate": 9.969530070213626e-06, "loss": 1.3917, "step": 1891 }, { "epoch": 0.1482758620689655, "grad_norm": 2.7912142276763916, "learning_rate": 9.969496069758097e-06, "loss": 1.3418, "step": 1892 }, { "epoch": 0.14835423197492162, "grad_norm": 2.329496145248413, "learning_rate": 9.969462050401157e-06, "loss": 1.2702, "step": 1893 }, { "epoch": 0.14843260188087773, "grad_norm": 2.3777525424957275, "learning_rate": 9.969428012142935e-06, "loss": 1.2102, "step": 1894 }, { "epoch": 0.14851097178683387, "grad_norm": 5.453620910644531, "learning_rate": 9.969393954983563e-06, "loss": 1.6618, "step": 1895 }, { "epoch": 0.14858934169278998, "grad_norm": 3.74977445602417, "learning_rate": 9.969359878923168e-06, "loss": 1.518, "step": 1896 }, { "epoch": 0.1486677115987461, "grad_norm": 2.927375078201294, "learning_rate": 9.969325783961881e-06, "loss": 1.2, "step": 1897 }, { "epoch": 0.1487460815047022, "grad_norm": 6.934253692626953, "learning_rate": 9.96929167009983e-06, "loss": 0.9709, "step": 1898 }, { "epoch": 0.1488244514106583, "grad_norm": 3.378031015396118, "learning_rate": 9.969257537337146e-06, "loss": 1.3686, "step": 1899 }, { "epoch": 0.14890282131661442, "grad_norm": 2.825441598892212, "learning_rate": 9.969223385673958e-06, "loss": 1.3882, "step": 1900 }, { "epoch": 0.14898119122257053, "grad_norm": 2.9002041816711426, "learning_rate": 9.9691892151104e-06, "loss": 1.1167, "step": 1901 }, { "epoch": 0.14905956112852664, "grad_norm": 2.618138074874878, "learning_rate": 9.969155025646595e-06, "loss": 1.2597, "step": 1902 }, { "epoch": 0.14913793103448275, "grad_norm": 14.28612995147705, "learning_rate": 9.969120817282677e-06, "loss": 1.9735, "step": 1903 }, { "epoch": 0.14921630094043886, "grad_norm": 2.5697081089019775, "learning_rate": 9.969086590018778e-06, "loss": 1.1456, "step": 1904 }, { "epoch": 0.14929467084639497, "grad_norm": 2.0360255241394043, "learning_rate": 9.969052343855023e-06, "loss": 1.4557, "step": 1905 }, { "epoch": 0.1493730407523511, "grad_norm": 2.142737627029419, "learning_rate": 9.969018078791548e-06, "loss": 1.6418, "step": 1906 }, { "epoch": 0.14945141065830722, "grad_norm": 2.1406168937683105, "learning_rate": 9.968983794828478e-06, "loss": 1.5154, "step": 1907 }, { "epoch": 0.14952978056426333, "grad_norm": 4.644115447998047, "learning_rate": 9.968949491965948e-06, "loss": 1.4396, "step": 1908 }, { "epoch": 0.14960815047021944, "grad_norm": 4.431135654449463, "learning_rate": 9.968915170204086e-06, "loss": 1.3214, "step": 1909 }, { "epoch": 0.14968652037617555, "grad_norm": 2.389084577560425, "learning_rate": 9.96888082954302e-06, "loss": 1.1767, "step": 1910 }, { "epoch": 0.14976489028213166, "grad_norm": 3.1379446983337402, "learning_rate": 9.968846469982886e-06, "loss": 1.1673, "step": 1911 }, { "epoch": 0.14984326018808777, "grad_norm": 4.36802864074707, "learning_rate": 9.968812091523812e-06, "loss": 1.2708, "step": 1912 }, { "epoch": 0.14992163009404388, "grad_norm": 5.5089826583862305, "learning_rate": 9.96877769416593e-06, "loss": 1.7802, "step": 1913 }, { "epoch": 0.15, "grad_norm": 14.25929069519043, "learning_rate": 9.968743277909368e-06, "loss": 0.8201, "step": 1914 }, { "epoch": 0.1500783699059561, "grad_norm": 2.692521333694458, "learning_rate": 9.968708842754259e-06, "loss": 1.2384, "step": 1915 }, { "epoch": 0.15015673981191222, "grad_norm": 3.131474256515503, "learning_rate": 9.968674388700735e-06, "loss": 1.3859, "step": 1916 }, { "epoch": 0.15023510971786833, "grad_norm": 2.8900532722473145, "learning_rate": 9.968639915748923e-06, "loss": 1.2141, "step": 1917 }, { "epoch": 0.15031347962382446, "grad_norm": 3.561978578567505, "learning_rate": 9.96860542389896e-06, "loss": 1.2537, "step": 1918 }, { "epoch": 0.15039184952978057, "grad_norm": 3.9946036338806152, "learning_rate": 9.968570913150973e-06, "loss": 1.9267, "step": 1919 }, { "epoch": 0.15047021943573669, "grad_norm": 8.150146484375, "learning_rate": 9.968536383505092e-06, "loss": 1.9585, "step": 1920 }, { "epoch": 0.1505485893416928, "grad_norm": 3.05491042137146, "learning_rate": 9.968501834961453e-06, "loss": 1.6173, "step": 1921 }, { "epoch": 0.1506269592476489, "grad_norm": 2.541835308074951, "learning_rate": 9.968467267520184e-06, "loss": 1.5462, "step": 1922 }, { "epoch": 0.15070532915360502, "grad_norm": 2.3422586917877197, "learning_rate": 9.968432681181416e-06, "loss": 1.3321, "step": 1923 }, { "epoch": 0.15078369905956113, "grad_norm": 2.7777295112609863, "learning_rate": 9.968398075945285e-06, "loss": 1.0508, "step": 1924 }, { "epoch": 0.15086206896551724, "grad_norm": 2.9095699787139893, "learning_rate": 9.968363451811919e-06, "loss": 1.2408, "step": 1925 }, { "epoch": 0.15094043887147335, "grad_norm": 2.826190710067749, "learning_rate": 9.96832880878145e-06, "loss": 1.4319, "step": 1926 }, { "epoch": 0.15101880877742946, "grad_norm": 4.633699893951416, "learning_rate": 9.96829414685401e-06, "loss": 1.6551, "step": 1927 }, { "epoch": 0.15109717868338557, "grad_norm": 4.255686283111572, "learning_rate": 9.96825946602973e-06, "loss": 1.4455, "step": 1928 }, { "epoch": 0.15117554858934168, "grad_norm": 4.642864227294922, "learning_rate": 9.968224766308745e-06, "loss": 1.3366, "step": 1929 }, { "epoch": 0.15125391849529782, "grad_norm": 3.8897979259490967, "learning_rate": 9.968190047691184e-06, "loss": 1.3433, "step": 1930 }, { "epoch": 0.15133228840125393, "grad_norm": 2.7856786251068115, "learning_rate": 9.96815531017718e-06, "loss": 1.6045, "step": 1931 }, { "epoch": 0.15141065830721004, "grad_norm": 7.350829124450684, "learning_rate": 9.968120553766863e-06, "loss": 0.2302, "step": 1932 }, { "epoch": 0.15148902821316615, "grad_norm": 3.0384905338287354, "learning_rate": 9.968085778460368e-06, "loss": 1.3204, "step": 1933 }, { "epoch": 0.15156739811912226, "grad_norm": 2.668107271194458, "learning_rate": 9.968050984257828e-06, "loss": 1.4304, "step": 1934 }, { "epoch": 0.15164576802507837, "grad_norm": 2.667289972305298, "learning_rate": 9.968016171159372e-06, "loss": 1.1564, "step": 1935 }, { "epoch": 0.15172413793103448, "grad_norm": 3.05548357963562, "learning_rate": 9.967981339165136e-06, "loss": 1.5517, "step": 1936 }, { "epoch": 0.1518025078369906, "grad_norm": 2.425184488296509, "learning_rate": 9.967946488275248e-06, "loss": 1.6603, "step": 1937 }, { "epoch": 0.1518808777429467, "grad_norm": 2.2820770740509033, "learning_rate": 9.967911618489848e-06, "loss": 1.4874, "step": 1938 }, { "epoch": 0.1519592476489028, "grad_norm": 2.80344295501709, "learning_rate": 9.96787672980906e-06, "loss": 1.4196, "step": 1939 }, { "epoch": 0.15203761755485892, "grad_norm": 3.9367222785949707, "learning_rate": 9.96784182223302e-06, "loss": 1.1421, "step": 1940 }, { "epoch": 0.15211598746081506, "grad_norm": 5.069917678833008, "learning_rate": 9.967806895761864e-06, "loss": 1.6518, "step": 1941 }, { "epoch": 0.15219435736677117, "grad_norm": 2.066244125366211, "learning_rate": 9.967771950395721e-06, "loss": 1.3219, "step": 1942 }, { "epoch": 0.15227272727272728, "grad_norm": 3.17935848236084, "learning_rate": 9.967736986134725e-06, "loss": 1.6511, "step": 1943 }, { "epoch": 0.1523510971786834, "grad_norm": 2.960317373275757, "learning_rate": 9.96770200297901e-06, "loss": 1.3422, "step": 1944 }, { "epoch": 0.1524294670846395, "grad_norm": 2.9469175338745117, "learning_rate": 9.967667000928706e-06, "loss": 1.41, "step": 1945 }, { "epoch": 0.1525078369905956, "grad_norm": 2.4640207290649414, "learning_rate": 9.967631979983951e-06, "loss": 1.4049, "step": 1946 }, { "epoch": 0.15258620689655172, "grad_norm": 4.42765474319458, "learning_rate": 9.967596940144875e-06, "loss": 2.0035, "step": 1947 }, { "epoch": 0.15266457680250783, "grad_norm": 2.7476704120635986, "learning_rate": 9.96756188141161e-06, "loss": 1.254, "step": 1948 }, { "epoch": 0.15274294670846394, "grad_norm": 2.994842529296875, "learning_rate": 9.967526803784292e-06, "loss": 1.1961, "step": 1949 }, { "epoch": 0.15282131661442006, "grad_norm": 3.897117853164673, "learning_rate": 9.967491707263054e-06, "loss": 1.6555, "step": 1950 }, { "epoch": 0.15289968652037617, "grad_norm": 2.5341618061065674, "learning_rate": 9.967456591848028e-06, "loss": 1.5439, "step": 1951 }, { "epoch": 0.15297805642633228, "grad_norm": 2.416581153869629, "learning_rate": 9.96742145753935e-06, "loss": 1.5824, "step": 1952 }, { "epoch": 0.15305642633228841, "grad_norm": 4.89817476272583, "learning_rate": 9.967386304337151e-06, "loss": 1.0878, "step": 1953 }, { "epoch": 0.15313479623824452, "grad_norm": 3.050717830657959, "learning_rate": 9.967351132241567e-06, "loss": 1.2032, "step": 1954 }, { "epoch": 0.15321316614420064, "grad_norm": 2.473306655883789, "learning_rate": 9.967315941252728e-06, "loss": 1.3183, "step": 1955 }, { "epoch": 0.15329153605015675, "grad_norm": 6.071987628936768, "learning_rate": 9.967280731370775e-06, "loss": 1.6155, "step": 1956 }, { "epoch": 0.15336990595611286, "grad_norm": 5.212542533874512, "learning_rate": 9.967245502595834e-06, "loss": 1.4229, "step": 1957 }, { "epoch": 0.15344827586206897, "grad_norm": 2.993032455444336, "learning_rate": 9.967210254928044e-06, "loss": 1.4434, "step": 1958 }, { "epoch": 0.15352664576802508, "grad_norm": 3.6061477661132812, "learning_rate": 9.967174988367536e-06, "loss": 1.6427, "step": 1959 }, { "epoch": 0.1536050156739812, "grad_norm": 3.9051103591918945, "learning_rate": 9.967139702914447e-06, "loss": 1.5654, "step": 1960 }, { "epoch": 0.1536833855799373, "grad_norm": 3.4337873458862305, "learning_rate": 9.96710439856891e-06, "loss": 1.2328, "step": 1961 }, { "epoch": 0.1537617554858934, "grad_norm": 2.7523043155670166, "learning_rate": 9.967069075331056e-06, "loss": 1.5702, "step": 1962 }, { "epoch": 0.15384012539184952, "grad_norm": 2.440046787261963, "learning_rate": 9.967033733201026e-06, "loss": 1.4949, "step": 1963 }, { "epoch": 0.15391849529780563, "grad_norm": 5.430251121520996, "learning_rate": 9.966998372178951e-06, "loss": 1.3756, "step": 1964 }, { "epoch": 0.15399686520376177, "grad_norm": 2.832385540008545, "learning_rate": 9.966962992264963e-06, "loss": 0.5785, "step": 1965 }, { "epoch": 0.15407523510971788, "grad_norm": 2.808274269104004, "learning_rate": 9.966927593459201e-06, "loss": 1.6268, "step": 1966 }, { "epoch": 0.154153605015674, "grad_norm": 5.607481002807617, "learning_rate": 9.966892175761797e-06, "loss": 1.7881, "step": 1967 }, { "epoch": 0.1542319749216301, "grad_norm": 3.5610616207122803, "learning_rate": 9.966856739172887e-06, "loss": 1.9344, "step": 1968 }, { "epoch": 0.1543103448275862, "grad_norm": 3.461355686187744, "learning_rate": 9.966821283692605e-06, "loss": 1.2881, "step": 1969 }, { "epoch": 0.15438871473354232, "grad_norm": 4.6827239990234375, "learning_rate": 9.966785809321085e-06, "loss": 1.7661, "step": 1970 }, { "epoch": 0.15446708463949843, "grad_norm": 2.569822311401367, "learning_rate": 9.966750316058463e-06, "loss": 1.6649, "step": 1971 }, { "epoch": 0.15454545454545454, "grad_norm": 3.1301257610321045, "learning_rate": 9.966714803904873e-06, "loss": 1.6966, "step": 1972 }, { "epoch": 0.15462382445141065, "grad_norm": 8.271649360656738, "learning_rate": 9.966679272860454e-06, "loss": 1.3791, "step": 1973 }, { "epoch": 0.15470219435736676, "grad_norm": 2.238295793533325, "learning_rate": 9.966643722925336e-06, "loss": 1.1844, "step": 1974 }, { "epoch": 0.15478056426332287, "grad_norm": 3.3161699771881104, "learning_rate": 9.966608154099654e-06, "loss": 1.4451, "step": 1975 }, { "epoch": 0.15485893416927898, "grad_norm": 2.780583620071411, "learning_rate": 9.96657256638355e-06, "loss": 1.3174, "step": 1976 }, { "epoch": 0.15493730407523512, "grad_norm": 2.4416563510894775, "learning_rate": 9.966536959777153e-06, "loss": 1.3595, "step": 1977 }, { "epoch": 0.15501567398119123, "grad_norm": 3.247391939163208, "learning_rate": 9.9665013342806e-06, "loss": 1.6077, "step": 1978 }, { "epoch": 0.15509404388714734, "grad_norm": 2.5967750549316406, "learning_rate": 9.966465689894027e-06, "loss": 1.3613, "step": 1979 }, { "epoch": 0.15517241379310345, "grad_norm": 2.6484193801879883, "learning_rate": 9.96643002661757e-06, "loss": 1.3917, "step": 1980 }, { "epoch": 0.15525078369905956, "grad_norm": 3.3409922122955322, "learning_rate": 9.966394344451364e-06, "loss": 1.045, "step": 1981 }, { "epoch": 0.15532915360501567, "grad_norm": 2.44873046875, "learning_rate": 9.966358643395544e-06, "loss": 1.0747, "step": 1982 }, { "epoch": 0.15540752351097178, "grad_norm": 3.6533076763153076, "learning_rate": 9.966322923450247e-06, "loss": 2.1044, "step": 1983 }, { "epoch": 0.1554858934169279, "grad_norm": 7.152013301849365, "learning_rate": 9.966287184615609e-06, "loss": 1.3021, "step": 1984 }, { "epoch": 0.155564263322884, "grad_norm": 27.50830078125, "learning_rate": 9.966251426891765e-06, "loss": 1.4815, "step": 1985 }, { "epoch": 0.15564263322884012, "grad_norm": 3.0099668502807617, "learning_rate": 9.966215650278853e-06, "loss": 1.3159, "step": 1986 }, { "epoch": 0.15572100313479623, "grad_norm": 3.2770285606384277, "learning_rate": 9.966179854777006e-06, "loss": 1.5145, "step": 1987 }, { "epoch": 0.15579937304075236, "grad_norm": 5.382201671600342, "learning_rate": 9.966144040386361e-06, "loss": 1.5416, "step": 1988 }, { "epoch": 0.15587774294670848, "grad_norm": 10.703737258911133, "learning_rate": 9.966108207107055e-06, "loss": 1.7448, "step": 1989 }, { "epoch": 0.15595611285266459, "grad_norm": 3.6348111629486084, "learning_rate": 9.966072354939225e-06, "loss": 1.3731, "step": 1990 }, { "epoch": 0.1560344827586207, "grad_norm": 2.9350383281707764, "learning_rate": 9.966036483883007e-06, "loss": 1.491, "step": 1991 }, { "epoch": 0.1561128526645768, "grad_norm": 2.577684164047241, "learning_rate": 9.966000593938534e-06, "loss": 1.5386, "step": 1992 }, { "epoch": 0.15619122257053292, "grad_norm": 2.8448715209960938, "learning_rate": 9.965964685105947e-06, "loss": 1.4656, "step": 1993 }, { "epoch": 0.15626959247648903, "grad_norm": 10.382515907287598, "learning_rate": 9.965928757385383e-06, "loss": 2.3499, "step": 1994 }, { "epoch": 0.15634796238244514, "grad_norm": 3.26242995262146, "learning_rate": 9.965892810776973e-06, "loss": 1.2642, "step": 1995 }, { "epoch": 0.15642633228840125, "grad_norm": 4.3275322914123535, "learning_rate": 9.965856845280862e-06, "loss": 1.5638, "step": 1996 }, { "epoch": 0.15650470219435736, "grad_norm": 3.1658341884613037, "learning_rate": 9.96582086089718e-06, "loss": 1.3888, "step": 1997 }, { "epoch": 0.15658307210031347, "grad_norm": 3.793200731277466, "learning_rate": 9.965784857626065e-06, "loss": 1.2493, "step": 1998 }, { "epoch": 0.15666144200626958, "grad_norm": 3.3511970043182373, "learning_rate": 9.965748835467654e-06, "loss": 1.6517, "step": 1999 }, { "epoch": 0.15673981191222572, "grad_norm": 3.2290103435516357, "learning_rate": 9.965712794422087e-06, "loss": 1.4126, "step": 2000 }, { "epoch": 0.15681818181818183, "grad_norm": 6.307653903961182, "learning_rate": 9.9656767344895e-06, "loss": 1.7002, "step": 2001 }, { "epoch": 0.15689655172413794, "grad_norm": 3.992858648300171, "learning_rate": 9.965640655670028e-06, "loss": 1.6831, "step": 2002 }, { "epoch": 0.15697492163009405, "grad_norm": 3.108978509902954, "learning_rate": 9.96560455796381e-06, "loss": 1.3533, "step": 2003 }, { "epoch": 0.15705329153605016, "grad_norm": 4.340388774871826, "learning_rate": 9.96556844137098e-06, "loss": 1.3656, "step": 2004 }, { "epoch": 0.15713166144200627, "grad_norm": 2.7450098991394043, "learning_rate": 9.965532305891682e-06, "loss": 1.4077, "step": 2005 }, { "epoch": 0.15721003134796238, "grad_norm": 2.9601340293884277, "learning_rate": 9.965496151526048e-06, "loss": 1.6187, "step": 2006 }, { "epoch": 0.1572884012539185, "grad_norm": 9.487305641174316, "learning_rate": 9.965459978274217e-06, "loss": 1.4101, "step": 2007 }, { "epoch": 0.1573667711598746, "grad_norm": 4.68189001083374, "learning_rate": 9.965423786136327e-06, "loss": 1.5045, "step": 2008 }, { "epoch": 0.1574451410658307, "grad_norm": 2.598500967025757, "learning_rate": 9.965387575112515e-06, "loss": 1.5413, "step": 2009 }, { "epoch": 0.15752351097178682, "grad_norm": 3.587636709213257, "learning_rate": 9.965351345202918e-06, "loss": 1.735, "step": 2010 }, { "epoch": 0.15760188087774293, "grad_norm": 5.210876941680908, "learning_rate": 9.965315096407678e-06, "loss": 1.857, "step": 2011 }, { "epoch": 0.15768025078369907, "grad_norm": 3.5148186683654785, "learning_rate": 9.965278828726926e-06, "loss": 1.3019, "step": 2012 }, { "epoch": 0.15775862068965518, "grad_norm": 6.185763835906982, "learning_rate": 9.965242542160806e-06, "loss": 1.2355, "step": 2013 }, { "epoch": 0.1578369905956113, "grad_norm": 3.6725411415100098, "learning_rate": 9.965206236709453e-06, "loss": 1.6981, "step": 2014 }, { "epoch": 0.1579153605015674, "grad_norm": 3.3071680068969727, "learning_rate": 9.965169912373005e-06, "loss": 1.6593, "step": 2015 }, { "epoch": 0.1579937304075235, "grad_norm": 2.982158899307251, "learning_rate": 9.965133569151601e-06, "loss": 1.6374, "step": 2016 }, { "epoch": 0.15807210031347962, "grad_norm": 4.698036193847656, "learning_rate": 9.965097207045382e-06, "loss": 1.4462, "step": 2017 }, { "epoch": 0.15815047021943573, "grad_norm": 3.7558298110961914, "learning_rate": 9.96506082605448e-06, "loss": 1.2744, "step": 2018 }, { "epoch": 0.15822884012539185, "grad_norm": 7.492794990539551, "learning_rate": 9.965024426179038e-06, "loss": 1.124, "step": 2019 }, { "epoch": 0.15830721003134796, "grad_norm": 9.00858211517334, "learning_rate": 9.964988007419195e-06, "loss": 1.5447, "step": 2020 }, { "epoch": 0.15838557993730407, "grad_norm": 3.340034008026123, "learning_rate": 9.964951569775085e-06, "loss": 1.3248, "step": 2021 }, { "epoch": 0.15846394984326018, "grad_norm": 7.815069198608398, "learning_rate": 9.964915113246851e-06, "loss": 1.1605, "step": 2022 }, { "epoch": 0.1585423197492163, "grad_norm": 3.4377269744873047, "learning_rate": 9.96487863783463e-06, "loss": 1.316, "step": 2023 }, { "epoch": 0.15862068965517243, "grad_norm": 4.341207504272461, "learning_rate": 9.96484214353856e-06, "loss": 1.6955, "step": 2024 }, { "epoch": 0.15869905956112854, "grad_norm": 3.364518404006958, "learning_rate": 9.96480563035878e-06, "loss": 0.7851, "step": 2025 }, { "epoch": 0.15877742946708465, "grad_norm": 2.6658823490142822, "learning_rate": 9.964769098295431e-06, "loss": 1.35, "step": 2026 }, { "epoch": 0.15885579937304076, "grad_norm": 2.7223808765411377, "learning_rate": 9.96473254734865e-06, "loss": 1.2735, "step": 2027 }, { "epoch": 0.15893416927899687, "grad_norm": 3.992013692855835, "learning_rate": 9.964695977518577e-06, "loss": 1.4475, "step": 2028 }, { "epoch": 0.15901253918495298, "grad_norm": 5.998990535736084, "learning_rate": 9.964659388805349e-06, "loss": 1.4752, "step": 2029 }, { "epoch": 0.1590909090909091, "grad_norm": 3.3395938873291016, "learning_rate": 9.964622781209109e-06, "loss": 1.0983, "step": 2030 }, { "epoch": 0.1591692789968652, "grad_norm": 3.4698400497436523, "learning_rate": 9.964586154729991e-06, "loss": 1.5476, "step": 2031 }, { "epoch": 0.1592476489028213, "grad_norm": 3.98482608795166, "learning_rate": 9.96454950936814e-06, "loss": 1.4159, "step": 2032 }, { "epoch": 0.15932601880877742, "grad_norm": 6.420323848724365, "learning_rate": 9.964512845123691e-06, "loss": 1.6996, "step": 2033 }, { "epoch": 0.15940438871473353, "grad_norm": 9.37367057800293, "learning_rate": 9.964476161996786e-06, "loss": 1.3765, "step": 2034 }, { "epoch": 0.15948275862068967, "grad_norm": 3.396373748779297, "learning_rate": 9.964439459987563e-06, "loss": 1.5048, "step": 2035 }, { "epoch": 0.15956112852664578, "grad_norm": 5.83917236328125, "learning_rate": 9.964402739096162e-06, "loss": 1.0924, "step": 2036 }, { "epoch": 0.1596394984326019, "grad_norm": 2.6738808155059814, "learning_rate": 9.964365999322722e-06, "loss": 1.0955, "step": 2037 }, { "epoch": 0.159717868338558, "grad_norm": 3.2467989921569824, "learning_rate": 9.964329240667385e-06, "loss": 1.206, "step": 2038 }, { "epoch": 0.1597962382445141, "grad_norm": 8.015402793884277, "learning_rate": 9.964292463130289e-06, "loss": 1.0165, "step": 2039 }, { "epoch": 0.15987460815047022, "grad_norm": 6.655613899230957, "learning_rate": 9.964255666711575e-06, "loss": 1.7277, "step": 2040 }, { "epoch": 0.15995297805642633, "grad_norm": 2.9713008403778076, "learning_rate": 9.964218851411382e-06, "loss": 1.3753, "step": 2041 }, { "epoch": 0.16003134796238244, "grad_norm": 3.5116562843322754, "learning_rate": 9.964182017229849e-06, "loss": 1.4939, "step": 2042 }, { "epoch": 0.16010971786833855, "grad_norm": 6.539211750030518, "learning_rate": 9.964145164167118e-06, "loss": 0.9094, "step": 2043 }, { "epoch": 0.16018808777429466, "grad_norm": 4.762463569641113, "learning_rate": 9.96410829222333e-06, "loss": 1.2593, "step": 2044 }, { "epoch": 0.16026645768025077, "grad_norm": 3.007042169570923, "learning_rate": 9.964071401398621e-06, "loss": 1.5765, "step": 2045 }, { "epoch": 0.16034482758620688, "grad_norm": 6.305665016174316, "learning_rate": 9.964034491693137e-06, "loss": 1.7865, "step": 2046 }, { "epoch": 0.16042319749216302, "grad_norm": 2.5448615550994873, "learning_rate": 9.963997563107014e-06, "loss": 1.3813, "step": 2047 }, { "epoch": 0.16050156739811913, "grad_norm": 3.0585453510284424, "learning_rate": 9.963960615640394e-06, "loss": 1.5858, "step": 2048 }, { "epoch": 0.16057993730407524, "grad_norm": 3.5513694286346436, "learning_rate": 9.963923649293418e-06, "loss": 1.36, "step": 2049 }, { "epoch": 0.16065830721003135, "grad_norm": 4.8230156898498535, "learning_rate": 9.963886664066224e-06, "loss": 1.0284, "step": 2050 }, { "epoch": 0.16073667711598746, "grad_norm": 4.446375370025635, "learning_rate": 9.963849659958957e-06, "loss": 1.4679, "step": 2051 }, { "epoch": 0.16081504702194357, "grad_norm": 4.437919616699219, "learning_rate": 9.963812636971755e-06, "loss": 1.7805, "step": 2052 }, { "epoch": 0.16089341692789969, "grad_norm": 3.3853797912597656, "learning_rate": 9.96377559510476e-06, "loss": 1.3165, "step": 2053 }, { "epoch": 0.1609717868338558, "grad_norm": 4.7122650146484375, "learning_rate": 9.963738534358112e-06, "loss": 1.5872, "step": 2054 }, { "epoch": 0.1610501567398119, "grad_norm": 4.3774213790893555, "learning_rate": 9.96370145473195e-06, "loss": 1.2532, "step": 2055 }, { "epoch": 0.16112852664576802, "grad_norm": 4.157581806182861, "learning_rate": 9.963664356226419e-06, "loss": 1.6881, "step": 2056 }, { "epoch": 0.16120689655172413, "grad_norm": 2.091235399246216, "learning_rate": 9.963627238841657e-06, "loss": 1.3396, "step": 2057 }, { "epoch": 0.16128526645768024, "grad_norm": 3.6295080184936523, "learning_rate": 9.963590102577809e-06, "loss": 1.0573, "step": 2058 }, { "epoch": 0.16136363636363638, "grad_norm": 3.3955767154693604, "learning_rate": 9.963552947435013e-06, "loss": 1.5738, "step": 2059 }, { "epoch": 0.1614420062695925, "grad_norm": 4.388021469116211, "learning_rate": 9.963515773413409e-06, "loss": 1.3432, "step": 2060 }, { "epoch": 0.1615203761755486, "grad_norm": 2.687870502471924, "learning_rate": 9.963478580513141e-06, "loss": 1.5169, "step": 2061 }, { "epoch": 0.1615987460815047, "grad_norm": 3.544248342514038, "learning_rate": 9.963441368734352e-06, "loss": 1.4013, "step": 2062 }, { "epoch": 0.16167711598746082, "grad_norm": 3.2833516597747803, "learning_rate": 9.96340413807718e-06, "loss": 1.0903, "step": 2063 }, { "epoch": 0.16175548589341693, "grad_norm": 3.295921802520752, "learning_rate": 9.963366888541769e-06, "loss": 1.4041, "step": 2064 }, { "epoch": 0.16183385579937304, "grad_norm": 3.7611188888549805, "learning_rate": 9.963329620128256e-06, "loss": 1.3004, "step": 2065 }, { "epoch": 0.16191222570532915, "grad_norm": 4.856956958770752, "learning_rate": 9.96329233283679e-06, "loss": 1.2669, "step": 2066 }, { "epoch": 0.16199059561128526, "grad_norm": 3.398118495941162, "learning_rate": 9.96325502666751e-06, "loss": 1.5631, "step": 2067 }, { "epoch": 0.16206896551724137, "grad_norm": 3.7717151641845703, "learning_rate": 9.963217701620554e-06, "loss": 1.7398, "step": 2068 }, { "epoch": 0.16214733542319748, "grad_norm": 3.980319023132324, "learning_rate": 9.96318035769607e-06, "loss": 1.7875, "step": 2069 }, { "epoch": 0.16222570532915362, "grad_norm": 3.6826529502868652, "learning_rate": 9.963142994894194e-06, "loss": 1.251, "step": 2070 }, { "epoch": 0.16230407523510973, "grad_norm": 4.8759589195251465, "learning_rate": 9.963105613215074e-06, "loss": 1.514, "step": 2071 }, { "epoch": 0.16238244514106584, "grad_norm": 3.436675548553467, "learning_rate": 9.963068212658848e-06, "loss": 1.358, "step": 2072 }, { "epoch": 0.16246081504702195, "grad_norm": 3.7804815769195557, "learning_rate": 9.96303079322566e-06, "loss": 1.7391, "step": 2073 }, { "epoch": 0.16253918495297806, "grad_norm": 2.473766565322876, "learning_rate": 9.962993354915653e-06, "loss": 1.32, "step": 2074 }, { "epoch": 0.16261755485893417, "grad_norm": 2.9705214500427246, "learning_rate": 9.962955897728968e-06, "loss": 1.5501, "step": 2075 }, { "epoch": 0.16269592476489028, "grad_norm": 3.0579025745391846, "learning_rate": 9.962918421665749e-06, "loss": 1.5495, "step": 2076 }, { "epoch": 0.1627742946708464, "grad_norm": 3.5319137573242188, "learning_rate": 9.962880926726135e-06, "loss": 1.7882, "step": 2077 }, { "epoch": 0.1628526645768025, "grad_norm": 5.151723384857178, "learning_rate": 9.962843412910272e-06, "loss": 1.5778, "step": 2078 }, { "epoch": 0.1629310344827586, "grad_norm": 2.356569766998291, "learning_rate": 9.962805880218303e-06, "loss": 1.2683, "step": 2079 }, { "epoch": 0.16300940438871472, "grad_norm": 3.101851463317871, "learning_rate": 9.962768328650367e-06, "loss": 1.3885, "step": 2080 }, { "epoch": 0.16308777429467083, "grad_norm": 9.899506568908691, "learning_rate": 9.962730758206612e-06, "loss": 2.044, "step": 2081 }, { "epoch": 0.16316614420062697, "grad_norm": 3.769655704498291, "learning_rate": 9.962693168887177e-06, "loss": 1.485, "step": 2082 }, { "epoch": 0.16324451410658308, "grad_norm": 3.1171367168426514, "learning_rate": 9.962655560692205e-06, "loss": 1.3068, "step": 2083 }, { "epoch": 0.1633228840125392, "grad_norm": 3.265183687210083, "learning_rate": 9.96261793362184e-06, "loss": 0.9974, "step": 2084 }, { "epoch": 0.1634012539184953, "grad_norm": 3.4954986572265625, "learning_rate": 9.962580287676228e-06, "loss": 1.4364, "step": 2085 }, { "epoch": 0.16347962382445141, "grad_norm": 5.682933330535889, "learning_rate": 9.962542622855508e-06, "loss": 1.4154, "step": 2086 }, { "epoch": 0.16355799373040752, "grad_norm": 4.006242752075195, "learning_rate": 9.962504939159824e-06, "loss": 1.6835, "step": 2087 }, { "epoch": 0.16363636363636364, "grad_norm": 3.4634909629821777, "learning_rate": 9.962467236589322e-06, "loss": 1.7609, "step": 2088 }, { "epoch": 0.16371473354231975, "grad_norm": 3.271531343460083, "learning_rate": 9.962429515144143e-06, "loss": 1.3639, "step": 2089 }, { "epoch": 0.16379310344827586, "grad_norm": 3.0415263175964355, "learning_rate": 9.962391774824432e-06, "loss": 1.5007, "step": 2090 }, { "epoch": 0.16387147335423197, "grad_norm": 2.861950159072876, "learning_rate": 9.962354015630328e-06, "loss": 1.3131, "step": 2091 }, { "epoch": 0.16394984326018808, "grad_norm": 6.003077507019043, "learning_rate": 9.962316237561982e-06, "loss": 1.8007, "step": 2092 }, { "epoch": 0.1640282131661442, "grad_norm": 2.432877779006958, "learning_rate": 9.96227844061953e-06, "loss": 1.5279, "step": 2093 }, { "epoch": 0.16410658307210033, "grad_norm": 5.512811660766602, "learning_rate": 9.962240624803124e-06, "loss": 1.6435, "step": 2094 }, { "epoch": 0.16418495297805644, "grad_norm": 3.2778635025024414, "learning_rate": 9.9622027901129e-06, "loss": 1.067, "step": 2095 }, { "epoch": 0.16426332288401255, "grad_norm": 2.8976778984069824, "learning_rate": 9.962164936549008e-06, "loss": 1.3863, "step": 2096 }, { "epoch": 0.16434169278996866, "grad_norm": 5.025100231170654, "learning_rate": 9.962127064111586e-06, "loss": 1.7516, "step": 2097 }, { "epoch": 0.16442006269592477, "grad_norm": 6.2759108543396, "learning_rate": 9.962089172800783e-06, "loss": 1.3947, "step": 2098 }, { "epoch": 0.16449843260188088, "grad_norm": 3.145476818084717, "learning_rate": 9.962051262616743e-06, "loss": 0.8222, "step": 2099 }, { "epoch": 0.164576802507837, "grad_norm": 2.3720550537109375, "learning_rate": 9.962013333559607e-06, "loss": 1.3508, "step": 2100 }, { "epoch": 0.1646551724137931, "grad_norm": 4.06635046005249, "learning_rate": 9.96197538562952e-06, "loss": 1.1891, "step": 2101 }, { "epoch": 0.1647335423197492, "grad_norm": 2.1825454235076904, "learning_rate": 9.96193741882663e-06, "loss": 1.4218, "step": 2102 }, { "epoch": 0.16481191222570532, "grad_norm": 3.0732154846191406, "learning_rate": 9.961899433151076e-06, "loss": 1.489, "step": 2103 }, { "epoch": 0.16489028213166143, "grad_norm": 4.550342082977295, "learning_rate": 9.961861428603007e-06, "loss": 1.8939, "step": 2104 }, { "epoch": 0.16496865203761754, "grad_norm": 2.493164300918579, "learning_rate": 9.961823405182565e-06, "loss": 1.0417, "step": 2105 }, { "epoch": 0.16504702194357368, "grad_norm": 3.480713129043579, "learning_rate": 9.961785362889895e-06, "loss": 1.3034, "step": 2106 }, { "epoch": 0.1651253918495298, "grad_norm": 2.641441822052002, "learning_rate": 9.961747301725142e-06, "loss": 1.2551, "step": 2107 }, { "epoch": 0.1652037617554859, "grad_norm": 2.664634943008423, "learning_rate": 9.96170922168845e-06, "loss": 1.5578, "step": 2108 }, { "epoch": 0.165282131661442, "grad_norm": 2.969672918319702, "learning_rate": 9.961671122779967e-06, "loss": 1.6187, "step": 2109 }, { "epoch": 0.16536050156739812, "grad_norm": 3.746957778930664, "learning_rate": 9.961633004999835e-06, "loss": 1.9067, "step": 2110 }, { "epoch": 0.16543887147335423, "grad_norm": 4.315760135650635, "learning_rate": 9.9615948683482e-06, "loss": 1.629, "step": 2111 }, { "epoch": 0.16551724137931034, "grad_norm": 2.639514923095703, "learning_rate": 9.961556712825204e-06, "loss": 1.3493, "step": 2112 }, { "epoch": 0.16559561128526645, "grad_norm": 2.8968639373779297, "learning_rate": 9.961518538430996e-06, "loss": 1.3121, "step": 2113 }, { "epoch": 0.16567398119122256, "grad_norm": 3.6175875663757324, "learning_rate": 9.961480345165719e-06, "loss": 0.7667, "step": 2114 }, { "epoch": 0.16575235109717867, "grad_norm": 5.247324466705322, "learning_rate": 9.961442133029522e-06, "loss": 1.7955, "step": 2115 }, { "epoch": 0.16583072100313478, "grad_norm": 3.2301790714263916, "learning_rate": 9.961403902022544e-06, "loss": 1.3359, "step": 2116 }, { "epoch": 0.16590909090909092, "grad_norm": 3.3711860179901123, "learning_rate": 9.961365652144937e-06, "loss": 1.4836, "step": 2117 }, { "epoch": 0.16598746081504703, "grad_norm": 4.37601900100708, "learning_rate": 9.961327383396842e-06, "loss": 1.3681, "step": 2118 }, { "epoch": 0.16606583072100314, "grad_norm": 2.4744837284088135, "learning_rate": 9.961289095778405e-06, "loss": 0.7081, "step": 2119 }, { "epoch": 0.16614420062695925, "grad_norm": 2.6645989418029785, "learning_rate": 9.961250789289773e-06, "loss": 1.3068, "step": 2120 }, { "epoch": 0.16622257053291536, "grad_norm": 2.6505508422851562, "learning_rate": 9.961212463931092e-06, "loss": 1.2451, "step": 2121 }, { "epoch": 0.16630094043887148, "grad_norm": 8.999702453613281, "learning_rate": 9.961174119702506e-06, "loss": 1.2265, "step": 2122 }, { "epoch": 0.16637931034482759, "grad_norm": 16.73430824279785, "learning_rate": 9.961135756604163e-06, "loss": 1.0115, "step": 2123 }, { "epoch": 0.1664576802507837, "grad_norm": 3.4925527572631836, "learning_rate": 9.961097374636207e-06, "loss": 1.505, "step": 2124 }, { "epoch": 0.1665360501567398, "grad_norm": 6.742713451385498, "learning_rate": 9.961058973798785e-06, "loss": 1.4039, "step": 2125 }, { "epoch": 0.16661442006269592, "grad_norm": 2.9743714332580566, "learning_rate": 9.961020554092042e-06, "loss": 1.2894, "step": 2126 }, { "epoch": 0.16669278996865203, "grad_norm": 3.6608216762542725, "learning_rate": 9.960982115516126e-06, "loss": 1.5917, "step": 2127 }, { "epoch": 0.16677115987460814, "grad_norm": 2.5194005966186523, "learning_rate": 9.960943658071183e-06, "loss": 1.5424, "step": 2128 }, { "epoch": 0.16684952978056428, "grad_norm": 2.500985860824585, "learning_rate": 9.960905181757357e-06, "loss": 1.5174, "step": 2129 }, { "epoch": 0.1669278996865204, "grad_norm": 4.822251796722412, "learning_rate": 9.960866686574797e-06, "loss": 1.5482, "step": 2130 }, { "epoch": 0.1670062695924765, "grad_norm": 3.6120429039001465, "learning_rate": 9.960828172523646e-06, "loss": 1.5647, "step": 2131 }, { "epoch": 0.1670846394984326, "grad_norm": 3.1081669330596924, "learning_rate": 9.960789639604055e-06, "loss": 1.5766, "step": 2132 }, { "epoch": 0.16716300940438872, "grad_norm": 3.8320329189300537, "learning_rate": 9.960751087816167e-06, "loss": 1.0662, "step": 2133 }, { "epoch": 0.16724137931034483, "grad_norm": 3.6924586296081543, "learning_rate": 9.960712517160129e-06, "loss": 1.032, "step": 2134 }, { "epoch": 0.16731974921630094, "grad_norm": 2.522697925567627, "learning_rate": 9.960673927636091e-06, "loss": 1.452, "step": 2135 }, { "epoch": 0.16739811912225705, "grad_norm": 3.882830858230591, "learning_rate": 9.960635319244195e-06, "loss": 1.6155, "step": 2136 }, { "epoch": 0.16747648902821316, "grad_norm": 4.75156831741333, "learning_rate": 9.960596691984593e-06, "loss": 1.4043, "step": 2137 }, { "epoch": 0.16755485893416927, "grad_norm": 2.967259645462036, "learning_rate": 9.960558045857426e-06, "loss": 1.1013, "step": 2138 }, { "epoch": 0.16763322884012538, "grad_norm": 2.50736665725708, "learning_rate": 9.960519380862845e-06, "loss": 1.3786, "step": 2139 }, { "epoch": 0.1677115987460815, "grad_norm": 10.567144393920898, "learning_rate": 9.960480697000996e-06, "loss": 0.9366, "step": 2140 }, { "epoch": 0.16778996865203763, "grad_norm": 2.9238367080688477, "learning_rate": 9.960441994272028e-06, "loss": 1.5605, "step": 2141 }, { "epoch": 0.16786833855799374, "grad_norm": 5.145018577575684, "learning_rate": 9.960403272676084e-06, "loss": 2.0428, "step": 2142 }, { "epoch": 0.16794670846394985, "grad_norm": 2.6611368656158447, "learning_rate": 9.960364532213315e-06, "loss": 1.5028, "step": 2143 }, { "epoch": 0.16802507836990596, "grad_norm": 4.503866672515869, "learning_rate": 9.960325772883868e-06, "loss": 1.7657, "step": 2144 }, { "epoch": 0.16810344827586207, "grad_norm": 5.957778453826904, "learning_rate": 9.960286994687888e-06, "loss": 1.6316, "step": 2145 }, { "epoch": 0.16818181818181818, "grad_norm": 2.6333606243133545, "learning_rate": 9.960248197625526e-06, "loss": 1.1975, "step": 2146 }, { "epoch": 0.1682601880877743, "grad_norm": 3.092292308807373, "learning_rate": 9.960209381696925e-06, "loss": 1.3665, "step": 2147 }, { "epoch": 0.1683385579937304, "grad_norm": 2.7578423023223877, "learning_rate": 9.960170546902238e-06, "loss": 1.1879, "step": 2148 }, { "epoch": 0.1684169278996865, "grad_norm": 2.8600058555603027, "learning_rate": 9.960131693241608e-06, "loss": 1.0823, "step": 2149 }, { "epoch": 0.16849529780564262, "grad_norm": 3.3838305473327637, "learning_rate": 9.960092820715186e-06, "loss": 1.6137, "step": 2150 }, { "epoch": 0.16857366771159873, "grad_norm": 7.908677101135254, "learning_rate": 9.960053929323117e-06, "loss": 2.2257, "step": 2151 }, { "epoch": 0.16865203761755485, "grad_norm": 3.3725779056549072, "learning_rate": 9.960015019065552e-06, "loss": 1.5927, "step": 2152 }, { "epoch": 0.16873040752351098, "grad_norm": 2.6094682216644287, "learning_rate": 9.959976089942636e-06, "loss": 1.2862, "step": 2153 }, { "epoch": 0.1688087774294671, "grad_norm": 6.787270545959473, "learning_rate": 9.959937141954522e-06, "loss": 1.2812, "step": 2154 }, { "epoch": 0.1688871473354232, "grad_norm": 4.4700469970703125, "learning_rate": 9.95989817510135e-06, "loss": 1.5203, "step": 2155 }, { "epoch": 0.16896551724137931, "grad_norm": 7.983309745788574, "learning_rate": 9.959859189383276e-06, "loss": 2.2971, "step": 2156 }, { "epoch": 0.16904388714733543, "grad_norm": 3.9180753231048584, "learning_rate": 9.959820184800443e-06, "loss": 1.4708, "step": 2157 }, { "epoch": 0.16912225705329154, "grad_norm": 3.4054689407348633, "learning_rate": 9.959781161353004e-06, "loss": 1.0764, "step": 2158 }, { "epoch": 0.16920062695924765, "grad_norm": 3.274292469024658, "learning_rate": 9.959742119041103e-06, "loss": 1.575, "step": 2159 }, { "epoch": 0.16927899686520376, "grad_norm": 3.0155704021453857, "learning_rate": 9.959703057864892e-06, "loss": 1.0259, "step": 2160 }, { "epoch": 0.16935736677115987, "grad_norm": 2.9210047721862793, "learning_rate": 9.959663977824517e-06, "loss": 1.645, "step": 2161 }, { "epoch": 0.16943573667711598, "grad_norm": 3.3917043209075928, "learning_rate": 9.959624878920127e-06, "loss": 1.2456, "step": 2162 }, { "epoch": 0.1695141065830721, "grad_norm": 3.222048044204712, "learning_rate": 9.959585761151873e-06, "loss": 1.6773, "step": 2163 }, { "epoch": 0.16959247648902823, "grad_norm": 3.35561203956604, "learning_rate": 9.959546624519902e-06, "loss": 1.7364, "step": 2164 }, { "epoch": 0.16967084639498434, "grad_norm": 4.3122406005859375, "learning_rate": 9.959507469024362e-06, "loss": 1.7626, "step": 2165 }, { "epoch": 0.16974921630094045, "grad_norm": 3.789891242980957, "learning_rate": 9.959468294665404e-06, "loss": 1.5476, "step": 2166 }, { "epoch": 0.16982758620689656, "grad_norm": 3.045879364013672, "learning_rate": 9.959429101443174e-06, "loss": 1.3926, "step": 2167 }, { "epoch": 0.16990595611285267, "grad_norm": 3.759153127670288, "learning_rate": 9.959389889357825e-06, "loss": 1.6049, "step": 2168 }, { "epoch": 0.16998432601880878, "grad_norm": 2.880356550216675, "learning_rate": 9.959350658409504e-06, "loss": 1.5823, "step": 2169 }, { "epoch": 0.1700626959247649, "grad_norm": 2.9399001598358154, "learning_rate": 9.95931140859836e-06, "loss": 1.3419, "step": 2170 }, { "epoch": 0.170141065830721, "grad_norm": 3.559110403060913, "learning_rate": 9.959272139924544e-06, "loss": 0.7656, "step": 2171 }, { "epoch": 0.1702194357366771, "grad_norm": 3.1525533199310303, "learning_rate": 9.959232852388202e-06, "loss": 1.2274, "step": 2172 }, { "epoch": 0.17029780564263322, "grad_norm": 3.2118546962738037, "learning_rate": 9.959193545989484e-06, "loss": 1.6498, "step": 2173 }, { "epoch": 0.17037617554858933, "grad_norm": 3.158548355102539, "learning_rate": 9.959154220728545e-06, "loss": 1.5699, "step": 2174 }, { "epoch": 0.17045454545454544, "grad_norm": 3.3337485790252686, "learning_rate": 9.959114876605528e-06, "loss": 1.31, "step": 2175 }, { "epoch": 0.17053291536050158, "grad_norm": 2.9062275886535645, "learning_rate": 9.959075513620587e-06, "loss": 1.4118, "step": 2176 }, { "epoch": 0.1706112852664577, "grad_norm": 4.633734226226807, "learning_rate": 9.959036131773867e-06, "loss": 1.8638, "step": 2177 }, { "epoch": 0.1706896551724138, "grad_norm": 2.929037094116211, "learning_rate": 9.958996731065522e-06, "loss": 1.3071, "step": 2178 }, { "epoch": 0.1707680250783699, "grad_norm": 2.365922451019287, "learning_rate": 9.958957311495702e-06, "loss": 1.5388, "step": 2179 }, { "epoch": 0.17084639498432602, "grad_norm": 2.8567278385162354, "learning_rate": 9.958917873064554e-06, "loss": 1.3711, "step": 2180 }, { "epoch": 0.17092476489028213, "grad_norm": 3.471113681793213, "learning_rate": 9.95887841577223e-06, "loss": 1.928, "step": 2181 }, { "epoch": 0.17100313479623824, "grad_norm": 3.031595468521118, "learning_rate": 9.958838939618879e-06, "loss": 1.2244, "step": 2182 }, { "epoch": 0.17108150470219435, "grad_norm": 4.443593978881836, "learning_rate": 9.958799444604652e-06, "loss": 1.5142, "step": 2183 }, { "epoch": 0.17115987460815046, "grad_norm": 3.5842344760894775, "learning_rate": 9.958759930729699e-06, "loss": 1.6588, "step": 2184 }, { "epoch": 0.17123824451410657, "grad_norm": 2.6065516471862793, "learning_rate": 9.958720397994171e-06, "loss": 1.5472, "step": 2185 }, { "epoch": 0.17131661442006269, "grad_norm": 3.578239917755127, "learning_rate": 9.958680846398217e-06, "loss": 1.2051, "step": 2186 }, { "epoch": 0.1713949843260188, "grad_norm": 2.6569974422454834, "learning_rate": 9.958641275941988e-06, "loss": 1.4571, "step": 2187 }, { "epoch": 0.17147335423197493, "grad_norm": 5.384851455688477, "learning_rate": 9.958601686625633e-06, "loss": 1.3184, "step": 2188 }, { "epoch": 0.17155172413793104, "grad_norm": 4.603341579437256, "learning_rate": 9.958562078449306e-06, "loss": 1.3971, "step": 2189 }, { "epoch": 0.17163009404388715, "grad_norm": 2.453235149383545, "learning_rate": 9.958522451413155e-06, "loss": 1.3644, "step": 2190 }, { "epoch": 0.17170846394984327, "grad_norm": 5.3487162590026855, "learning_rate": 9.95848280551733e-06, "loss": 0.8158, "step": 2191 }, { "epoch": 0.17178683385579938, "grad_norm": 3.8043134212493896, "learning_rate": 9.958443140761985e-06, "loss": 1.6811, "step": 2192 }, { "epoch": 0.1718652037617555, "grad_norm": 2.9445552825927734, "learning_rate": 9.95840345714727e-06, "loss": 1.2983, "step": 2193 }, { "epoch": 0.1719435736677116, "grad_norm": 5.612381935119629, "learning_rate": 9.958363754673332e-06, "loss": 1.5055, "step": 2194 }, { "epoch": 0.1720219435736677, "grad_norm": 3.079669952392578, "learning_rate": 9.958324033340328e-06, "loss": 1.5219, "step": 2195 }, { "epoch": 0.17210031347962382, "grad_norm": 2.4154107570648193, "learning_rate": 9.958284293148407e-06, "loss": 1.0751, "step": 2196 }, { "epoch": 0.17217868338557993, "grad_norm": 2.9597008228302, "learning_rate": 9.958244534097717e-06, "loss": 1.6231, "step": 2197 }, { "epoch": 0.17225705329153604, "grad_norm": 5.353300094604492, "learning_rate": 9.95820475618841e-06, "loss": 1.8563, "step": 2198 }, { "epoch": 0.17233542319749218, "grad_norm": 2.839172601699829, "learning_rate": 9.958164959420642e-06, "loss": 1.4465, "step": 2199 }, { "epoch": 0.1724137931034483, "grad_norm": 7.042070388793945, "learning_rate": 9.95812514379456e-06, "loss": 1.5831, "step": 2200 }, { "epoch": 0.1724921630094044, "grad_norm": 3.2036259174346924, "learning_rate": 9.958085309310316e-06, "loss": 1.4755, "step": 2201 }, { "epoch": 0.1725705329153605, "grad_norm": 5.258871078491211, "learning_rate": 9.958045455968063e-06, "loss": 1.4768, "step": 2202 }, { "epoch": 0.17264890282131662, "grad_norm": 2.798192262649536, "learning_rate": 9.95800558376795e-06, "loss": 1.3477, "step": 2203 }, { "epoch": 0.17272727272727273, "grad_norm": 4.532491683959961, "learning_rate": 9.957965692710132e-06, "loss": 1.0631, "step": 2204 }, { "epoch": 0.17280564263322884, "grad_norm": 3.8402891159057617, "learning_rate": 9.95792578279476e-06, "loss": 1.3136, "step": 2205 }, { "epoch": 0.17288401253918495, "grad_norm": 9.935574531555176, "learning_rate": 9.957885854021983e-06, "loss": 1.4827, "step": 2206 }, { "epoch": 0.17296238244514106, "grad_norm": 3.521669864654541, "learning_rate": 9.957845906391954e-06, "loss": 1.2813, "step": 2207 }, { "epoch": 0.17304075235109717, "grad_norm": 2.5763819217681885, "learning_rate": 9.957805939904828e-06, "loss": 1.4645, "step": 2208 }, { "epoch": 0.17311912225705328, "grad_norm": 2.6260406970977783, "learning_rate": 9.957765954560753e-06, "loss": 1.1072, "step": 2209 }, { "epoch": 0.1731974921630094, "grad_norm": 2.731055974960327, "learning_rate": 9.957725950359884e-06, "loss": 1.3078, "step": 2210 }, { "epoch": 0.17327586206896553, "grad_norm": 2.388174295425415, "learning_rate": 9.957685927302372e-06, "loss": 1.3445, "step": 2211 }, { "epoch": 0.17335423197492164, "grad_norm": 2.5592217445373535, "learning_rate": 9.957645885388367e-06, "loss": 1.2669, "step": 2212 }, { "epoch": 0.17343260188087775, "grad_norm": 2.9071333408355713, "learning_rate": 9.957605824618026e-06, "loss": 1.7198, "step": 2213 }, { "epoch": 0.17351097178683386, "grad_norm": 2.9144444465637207, "learning_rate": 9.957565744991498e-06, "loss": 1.3448, "step": 2214 }, { "epoch": 0.17358934169278997, "grad_norm": 3.6375555992126465, "learning_rate": 9.957525646508936e-06, "loss": 1.3192, "step": 2215 }, { "epoch": 0.17366771159874608, "grad_norm": 2.9738900661468506, "learning_rate": 9.957485529170493e-06, "loss": 1.3432, "step": 2216 }, { "epoch": 0.1737460815047022, "grad_norm": 3.0664398670196533, "learning_rate": 9.957445392976322e-06, "loss": 1.4904, "step": 2217 }, { "epoch": 0.1738244514106583, "grad_norm": 2.511121988296509, "learning_rate": 9.957405237926574e-06, "loss": 1.4705, "step": 2218 }, { "epoch": 0.17390282131661441, "grad_norm": 3.105257749557495, "learning_rate": 9.957365064021403e-06, "loss": 1.4362, "step": 2219 }, { "epoch": 0.17398119122257052, "grad_norm": 2.544989585876465, "learning_rate": 9.95732487126096e-06, "loss": 1.5805, "step": 2220 }, { "epoch": 0.17405956112852664, "grad_norm": 2.896688222885132, "learning_rate": 9.957284659645402e-06, "loss": 1.4243, "step": 2221 }, { "epoch": 0.17413793103448275, "grad_norm": 3.3004114627838135, "learning_rate": 9.95724442917488e-06, "loss": 1.1379, "step": 2222 }, { "epoch": 0.17421630094043888, "grad_norm": 5.635354518890381, "learning_rate": 9.957204179849543e-06, "loss": 1.5008, "step": 2223 }, { "epoch": 0.174294670846395, "grad_norm": 2.4977896213531494, "learning_rate": 9.95716391166955e-06, "loss": 1.1478, "step": 2224 }, { "epoch": 0.1743730407523511, "grad_norm": 3.7857608795166016, "learning_rate": 9.95712362463505e-06, "loss": 1.2409, "step": 2225 }, { "epoch": 0.17445141065830722, "grad_norm": 4.118441104888916, "learning_rate": 9.9570833187462e-06, "loss": 1.4638, "step": 2226 }, { "epoch": 0.17452978056426333, "grad_norm": 5.493500232696533, "learning_rate": 9.95704299400315e-06, "loss": 1.5355, "step": 2227 }, { "epoch": 0.17460815047021944, "grad_norm": 2.5959672927856445, "learning_rate": 9.957002650406053e-06, "loss": 1.3199, "step": 2228 }, { "epoch": 0.17468652037617555, "grad_norm": 3.372175455093384, "learning_rate": 9.956962287955067e-06, "loss": 1.3447, "step": 2229 }, { "epoch": 0.17476489028213166, "grad_norm": 6.438101291656494, "learning_rate": 9.956921906650342e-06, "loss": 1.5848, "step": 2230 }, { "epoch": 0.17484326018808777, "grad_norm": 3.0559208393096924, "learning_rate": 9.95688150649203e-06, "loss": 1.4717, "step": 2231 }, { "epoch": 0.17492163009404388, "grad_norm": 3.6048383712768555, "learning_rate": 9.956841087480289e-06, "loss": 1.4703, "step": 2232 }, { "epoch": 0.175, "grad_norm": 2.8252921104431152, "learning_rate": 9.956800649615269e-06, "loss": 1.534, "step": 2233 }, { "epoch": 0.1750783699059561, "grad_norm": 2.8125617504119873, "learning_rate": 9.956760192897125e-06, "loss": 1.5967, "step": 2234 }, { "epoch": 0.17515673981191224, "grad_norm": 17.0283260345459, "learning_rate": 9.956719717326012e-06, "loss": 1.0728, "step": 2235 }, { "epoch": 0.17523510971786835, "grad_norm": 3.1195285320281982, "learning_rate": 9.956679222902083e-06, "loss": 1.5383, "step": 2236 }, { "epoch": 0.17531347962382446, "grad_norm": 2.833773612976074, "learning_rate": 9.956638709625492e-06, "loss": 1.4211, "step": 2237 }, { "epoch": 0.17539184952978057, "grad_norm": 7.421668529510498, "learning_rate": 9.956598177496394e-06, "loss": 0.9928, "step": 2238 }, { "epoch": 0.17547021943573668, "grad_norm": 3.632671356201172, "learning_rate": 9.956557626514943e-06, "loss": 1.775, "step": 2239 }, { "epoch": 0.1755485893416928, "grad_norm": 3.61629319190979, "learning_rate": 9.95651705668129e-06, "loss": 0.872, "step": 2240 }, { "epoch": 0.1756269592476489, "grad_norm": 2.8099100589752197, "learning_rate": 9.956476467995593e-06, "loss": 1.3262, "step": 2241 }, { "epoch": 0.175705329153605, "grad_norm": 8.71871280670166, "learning_rate": 9.956435860458005e-06, "loss": 1.513, "step": 2242 }, { "epoch": 0.17578369905956112, "grad_norm": 2.11103892326355, "learning_rate": 9.956395234068683e-06, "loss": 1.5288, "step": 2243 }, { "epoch": 0.17586206896551723, "grad_norm": 3.2257611751556396, "learning_rate": 9.956354588827777e-06, "loss": 1.8147, "step": 2244 }, { "epoch": 0.17594043887147334, "grad_norm": 2.646285057067871, "learning_rate": 9.956313924735446e-06, "loss": 1.3798, "step": 2245 }, { "epoch": 0.17601880877742948, "grad_norm": 2.5874176025390625, "learning_rate": 9.95627324179184e-06, "loss": 1.3971, "step": 2246 }, { "epoch": 0.1760971786833856, "grad_norm": 2.644479513168335, "learning_rate": 9.956232539997118e-06, "loss": 1.4915, "step": 2247 }, { "epoch": 0.1761755485893417, "grad_norm": 3.4043402671813965, "learning_rate": 9.956191819351433e-06, "loss": 1.4919, "step": 2248 }, { "epoch": 0.1762539184952978, "grad_norm": 2.9724950790405273, "learning_rate": 9.956151079854938e-06, "loss": 1.4066, "step": 2249 }, { "epoch": 0.17633228840125392, "grad_norm": 2.747896671295166, "learning_rate": 9.956110321507793e-06, "loss": 1.2908, "step": 2250 }, { "epoch": 0.17641065830721003, "grad_norm": 4.705966472625732, "learning_rate": 9.956069544310148e-06, "loss": 1.4394, "step": 2251 }, { "epoch": 0.17648902821316614, "grad_norm": 3.1781418323516846, "learning_rate": 9.95602874826216e-06, "loss": 1.1858, "step": 2252 }, { "epoch": 0.17656739811912225, "grad_norm": 3.0171284675598145, "learning_rate": 9.955987933363986e-06, "loss": 1.3424, "step": 2253 }, { "epoch": 0.17664576802507836, "grad_norm": 6.2698845863342285, "learning_rate": 9.955947099615777e-06, "loss": 1.2541, "step": 2254 }, { "epoch": 0.17672413793103448, "grad_norm": 3.058931350708008, "learning_rate": 9.955906247017693e-06, "loss": 1.6484, "step": 2255 }, { "epoch": 0.17680250783699059, "grad_norm": 3.087780714035034, "learning_rate": 9.955865375569886e-06, "loss": 1.4696, "step": 2256 }, { "epoch": 0.1768808777429467, "grad_norm": 3.5487446784973145, "learning_rate": 9.955824485272514e-06, "loss": 1.6685, "step": 2257 }, { "epoch": 0.17695924764890283, "grad_norm": 4.40689754486084, "learning_rate": 9.955783576125728e-06, "loss": 1.8248, "step": 2258 }, { "epoch": 0.17703761755485894, "grad_norm": 2.7993061542510986, "learning_rate": 9.95574264812969e-06, "loss": 1.4935, "step": 2259 }, { "epoch": 0.17711598746081506, "grad_norm": 4.14443302154541, "learning_rate": 9.95570170128455e-06, "loss": 1.9441, "step": 2260 }, { "epoch": 0.17719435736677117, "grad_norm": 2.402714490890503, "learning_rate": 9.955660735590466e-06, "loss": 1.353, "step": 2261 }, { "epoch": 0.17727272727272728, "grad_norm": 4.190407752990723, "learning_rate": 9.955619751047596e-06, "loss": 1.318, "step": 2262 }, { "epoch": 0.1773510971786834, "grad_norm": 2.6464691162109375, "learning_rate": 9.955578747656094e-06, "loss": 1.5979, "step": 2263 }, { "epoch": 0.1774294670846395, "grad_norm": 7.114316463470459, "learning_rate": 9.955537725416114e-06, "loss": 1.6252, "step": 2264 }, { "epoch": 0.1775078369905956, "grad_norm": 2.987165689468384, "learning_rate": 9.955496684327814e-06, "loss": 1.1911, "step": 2265 }, { "epoch": 0.17758620689655172, "grad_norm": 2.475100517272949, "learning_rate": 9.95545562439135e-06, "loss": 0.9356, "step": 2266 }, { "epoch": 0.17766457680250783, "grad_norm": 3.108865261077881, "learning_rate": 9.955414545606879e-06, "loss": 1.5065, "step": 2267 }, { "epoch": 0.17774294670846394, "grad_norm": 2.8035457134246826, "learning_rate": 9.955373447974558e-06, "loss": 1.4794, "step": 2268 }, { "epoch": 0.17782131661442005, "grad_norm": 4.488780975341797, "learning_rate": 9.955332331494538e-06, "loss": 1.394, "step": 2269 }, { "epoch": 0.1778996865203762, "grad_norm": 3.472078323364258, "learning_rate": 9.95529119616698e-06, "loss": 1.3449, "step": 2270 }, { "epoch": 0.1779780564263323, "grad_norm": 3.4839131832122803, "learning_rate": 9.95525004199204e-06, "loss": 1.3231, "step": 2271 }, { "epoch": 0.1780564263322884, "grad_norm": 3.4766690731048584, "learning_rate": 9.955208868969875e-06, "loss": 1.5505, "step": 2272 }, { "epoch": 0.17813479623824452, "grad_norm": 3.358043909072876, "learning_rate": 9.955167677100639e-06, "loss": 1.5078, "step": 2273 }, { "epoch": 0.17821316614420063, "grad_norm": 3.717931032180786, "learning_rate": 9.955126466384491e-06, "loss": 1.7677, "step": 2274 }, { "epoch": 0.17829153605015674, "grad_norm": 2.7416906356811523, "learning_rate": 9.955085236821587e-06, "loss": 0.724, "step": 2275 }, { "epoch": 0.17836990595611285, "grad_norm": 3.9749867916107178, "learning_rate": 9.955043988412084e-06, "loss": 1.4408, "step": 2276 }, { "epoch": 0.17844827586206896, "grad_norm": 2.519770383834839, "learning_rate": 9.955002721156139e-06, "loss": 1.008, "step": 2277 }, { "epoch": 0.17852664576802507, "grad_norm": 7.819395542144775, "learning_rate": 9.954961435053908e-06, "loss": 1.4047, "step": 2278 }, { "epoch": 0.17860501567398118, "grad_norm": 3.860288381576538, "learning_rate": 9.95492013010555e-06, "loss": 1.221, "step": 2279 }, { "epoch": 0.1786833855799373, "grad_norm": 3.649620532989502, "learning_rate": 9.954878806311218e-06, "loss": 1.8128, "step": 2280 }, { "epoch": 0.1787617554858934, "grad_norm": 2.6152536869049072, "learning_rate": 9.954837463671075e-06, "loss": 1.178, "step": 2281 }, { "epoch": 0.17884012539184954, "grad_norm": 6.0249762535095215, "learning_rate": 9.954796102185273e-06, "loss": 1.3982, "step": 2282 }, { "epoch": 0.17891849529780565, "grad_norm": 2.6768836975097656, "learning_rate": 9.954754721853974e-06, "loss": 1.2402, "step": 2283 }, { "epoch": 0.17899686520376176, "grad_norm": 2.971730947494507, "learning_rate": 9.954713322677332e-06, "loss": 1.119, "step": 2284 }, { "epoch": 0.17907523510971787, "grad_norm": 3.287198543548584, "learning_rate": 9.954671904655505e-06, "loss": 1.2412, "step": 2285 }, { "epoch": 0.17915360501567398, "grad_norm": 3.577944040298462, "learning_rate": 9.954630467788651e-06, "loss": 1.4416, "step": 2286 }, { "epoch": 0.1792319749216301, "grad_norm": 3.875582695007324, "learning_rate": 9.954589012076927e-06, "loss": 1.5714, "step": 2287 }, { "epoch": 0.1793103448275862, "grad_norm": 3.4627532958984375, "learning_rate": 9.954547537520491e-06, "loss": 1.5774, "step": 2288 }, { "epoch": 0.17938871473354231, "grad_norm": 3.0868444442749023, "learning_rate": 9.954506044119502e-06, "loss": 1.7899, "step": 2289 }, { "epoch": 0.17946708463949843, "grad_norm": 4.667697906494141, "learning_rate": 9.954464531874118e-06, "loss": 0.8755, "step": 2290 }, { "epoch": 0.17954545454545454, "grad_norm": 4.056361198425293, "learning_rate": 9.954423000784493e-06, "loss": 1.6625, "step": 2291 }, { "epoch": 0.17962382445141065, "grad_norm": 3.063044309616089, "learning_rate": 9.95438145085079e-06, "loss": 1.5804, "step": 2292 }, { "epoch": 0.17970219435736678, "grad_norm": 4.089142322540283, "learning_rate": 9.954339882073163e-06, "loss": 1.4083, "step": 2293 }, { "epoch": 0.1797805642633229, "grad_norm": 3.288989543914795, "learning_rate": 9.954298294451771e-06, "loss": 1.5213, "step": 2294 }, { "epoch": 0.179858934169279, "grad_norm": 4.141533851623535, "learning_rate": 9.954256687986777e-06, "loss": 1.1041, "step": 2295 }, { "epoch": 0.17993730407523512, "grad_norm": 2.9818661212921143, "learning_rate": 9.954215062678332e-06, "loss": 1.68, "step": 2296 }, { "epoch": 0.18001567398119123, "grad_norm": 3.1064743995666504, "learning_rate": 9.954173418526599e-06, "loss": 1.8108, "step": 2297 }, { "epoch": 0.18009404388714734, "grad_norm": 2.2877025604248047, "learning_rate": 9.954131755531734e-06, "loss": 1.5715, "step": 2298 }, { "epoch": 0.18017241379310345, "grad_norm": 10.160141944885254, "learning_rate": 9.954090073693895e-06, "loss": 2.4554, "step": 2299 }, { "epoch": 0.18025078369905956, "grad_norm": 3.0197882652282715, "learning_rate": 9.954048373013245e-06, "loss": 1.352, "step": 2300 }, { "epoch": 0.18032915360501567, "grad_norm": 3.2683653831481934, "learning_rate": 9.954006653489937e-06, "loss": 1.0507, "step": 2301 }, { "epoch": 0.18040752351097178, "grad_norm": 4.268095970153809, "learning_rate": 9.953964915124132e-06, "loss": 1.6731, "step": 2302 }, { "epoch": 0.1804858934169279, "grad_norm": 2.660966634750366, "learning_rate": 9.953923157915992e-06, "loss": 1.8417, "step": 2303 }, { "epoch": 0.180564263322884, "grad_norm": 2.662682056427002, "learning_rate": 9.953881381865671e-06, "loss": 1.7918, "step": 2304 }, { "epoch": 0.18064263322884014, "grad_norm": 3.671569585800171, "learning_rate": 9.95383958697333e-06, "loss": 1.1677, "step": 2305 }, { "epoch": 0.18072100313479625, "grad_norm": 3.7490837574005127, "learning_rate": 9.953797773239126e-06, "loss": 1.1137, "step": 2306 }, { "epoch": 0.18079937304075236, "grad_norm": 3.4329872131347656, "learning_rate": 9.953755940663222e-06, "loss": 1.6441, "step": 2307 }, { "epoch": 0.18087774294670847, "grad_norm": 2.6720569133758545, "learning_rate": 9.953714089245774e-06, "loss": 1.6332, "step": 2308 }, { "epoch": 0.18095611285266458, "grad_norm": 5.053886890411377, "learning_rate": 9.953672218986942e-06, "loss": 1.5677, "step": 2309 }, { "epoch": 0.1810344827586207, "grad_norm": 3.1677684783935547, "learning_rate": 9.953630329886886e-06, "loss": 1.3306, "step": 2310 }, { "epoch": 0.1811128526645768, "grad_norm": 4.971002578735352, "learning_rate": 9.953588421945762e-06, "loss": 2.0283, "step": 2311 }, { "epoch": 0.1811912225705329, "grad_norm": 4.769565105438232, "learning_rate": 9.953546495163734e-06, "loss": 0.9552, "step": 2312 }, { "epoch": 0.18126959247648902, "grad_norm": 4.538699150085449, "learning_rate": 9.95350454954096e-06, "loss": 1.129, "step": 2313 }, { "epoch": 0.18134796238244513, "grad_norm": 3.0565149784088135, "learning_rate": 9.953462585077598e-06, "loss": 1.3084, "step": 2314 }, { "epoch": 0.18142633228840124, "grad_norm": 5.652162551879883, "learning_rate": 9.953420601773809e-06, "loss": 2.1223, "step": 2315 }, { "epoch": 0.18150470219435735, "grad_norm": 4.316612243652344, "learning_rate": 9.95337859962975e-06, "loss": 1.3411, "step": 2316 }, { "epoch": 0.1815830721003135, "grad_norm": 3.4100513458251953, "learning_rate": 9.953336578645586e-06, "loss": 1.312, "step": 2317 }, { "epoch": 0.1816614420062696, "grad_norm": 3.16292667388916, "learning_rate": 9.953294538821473e-06, "loss": 1.4026, "step": 2318 }, { "epoch": 0.1817398119122257, "grad_norm": 4.209228038787842, "learning_rate": 9.95325248015757e-06, "loss": 1.0638, "step": 2319 }, { "epoch": 0.18181818181818182, "grad_norm": 3.1274259090423584, "learning_rate": 9.953210402654043e-06, "loss": 1.1421, "step": 2320 }, { "epoch": 0.18189655172413793, "grad_norm": 5.310145854949951, "learning_rate": 9.953168306311045e-06, "loss": 0.9128, "step": 2321 }, { "epoch": 0.18197492163009404, "grad_norm": 5.403263568878174, "learning_rate": 9.953126191128738e-06, "loss": 1.2051, "step": 2322 }, { "epoch": 0.18205329153605015, "grad_norm": 3.5988340377807617, "learning_rate": 9.953084057107284e-06, "loss": 1.2833, "step": 2323 }, { "epoch": 0.18213166144200627, "grad_norm": 3.2255454063415527, "learning_rate": 9.953041904246843e-06, "loss": 1.5522, "step": 2324 }, { "epoch": 0.18221003134796238, "grad_norm": 11.599387168884277, "learning_rate": 9.952999732547573e-06, "loss": 1.2836, "step": 2325 }, { "epoch": 0.1822884012539185, "grad_norm": 3.05487060546875, "learning_rate": 9.952957542009637e-06, "loss": 1.1073, "step": 2326 }, { "epoch": 0.1823667711598746, "grad_norm": 4.663546562194824, "learning_rate": 9.952915332633195e-06, "loss": 1.3378, "step": 2327 }, { "epoch": 0.18244514106583073, "grad_norm": 3.8646092414855957, "learning_rate": 9.952873104418405e-06, "loss": 1.4191, "step": 2328 }, { "epoch": 0.18252351097178685, "grad_norm": 6.5130228996276855, "learning_rate": 9.952830857365432e-06, "loss": 1.5393, "step": 2329 }, { "epoch": 0.18260188087774296, "grad_norm": 2.8531956672668457, "learning_rate": 9.952788591474433e-06, "loss": 1.1481, "step": 2330 }, { "epoch": 0.18268025078369907, "grad_norm": 2.18894362449646, "learning_rate": 9.952746306745572e-06, "loss": 1.6401, "step": 2331 }, { "epoch": 0.18275862068965518, "grad_norm": 4.202295780181885, "learning_rate": 9.952704003179006e-06, "loss": 1.6708, "step": 2332 }, { "epoch": 0.1828369905956113, "grad_norm": 5.463100910186768, "learning_rate": 9.952661680774898e-06, "loss": 1.9951, "step": 2333 }, { "epoch": 0.1829153605015674, "grad_norm": 3.0956294536590576, "learning_rate": 9.95261933953341e-06, "loss": 1.7803, "step": 2334 }, { "epoch": 0.1829937304075235, "grad_norm": 7.903581142425537, "learning_rate": 9.952576979454702e-06, "loss": 0.6234, "step": 2335 }, { "epoch": 0.18307210031347962, "grad_norm": 6.223832130432129, "learning_rate": 9.952534600538932e-06, "loss": 1.7596, "step": 2336 }, { "epoch": 0.18315047021943573, "grad_norm": 2.648144245147705, "learning_rate": 9.952492202786269e-06, "loss": 1.5021, "step": 2337 }, { "epoch": 0.18322884012539184, "grad_norm": 4.101679801940918, "learning_rate": 9.952449786196865e-06, "loss": 1.5971, "step": 2338 }, { "epoch": 0.18330721003134795, "grad_norm": 2.5008206367492676, "learning_rate": 9.952407350770888e-06, "loss": 1.5189, "step": 2339 }, { "epoch": 0.1833855799373041, "grad_norm": 2.776618719100952, "learning_rate": 9.952364896508495e-06, "loss": 1.4855, "step": 2340 }, { "epoch": 0.1834639498432602, "grad_norm": 2.6414084434509277, "learning_rate": 9.952322423409852e-06, "loss": 1.4876, "step": 2341 }, { "epoch": 0.1835423197492163, "grad_norm": 2.5409905910491943, "learning_rate": 9.952279931475117e-06, "loss": 1.8619, "step": 2342 }, { "epoch": 0.18362068965517242, "grad_norm": 2.9439828395843506, "learning_rate": 9.952237420704454e-06, "loss": 1.4653, "step": 2343 }, { "epoch": 0.18369905956112853, "grad_norm": 3.1676135063171387, "learning_rate": 9.952194891098023e-06, "loss": 1.2644, "step": 2344 }, { "epoch": 0.18377742946708464, "grad_norm": 3.8374526500701904, "learning_rate": 9.952152342655986e-06, "loss": 1.6371, "step": 2345 }, { "epoch": 0.18385579937304075, "grad_norm": 3.460667371749878, "learning_rate": 9.952109775378504e-06, "loss": 1.4275, "step": 2346 }, { "epoch": 0.18393416927899686, "grad_norm": 3.3048501014709473, "learning_rate": 9.952067189265742e-06, "loss": 1.1652, "step": 2347 }, { "epoch": 0.18401253918495297, "grad_norm": 2.993039846420288, "learning_rate": 9.952024584317858e-06, "loss": 1.1741, "step": 2348 }, { "epoch": 0.18409090909090908, "grad_norm": 2.5381085872650146, "learning_rate": 9.951981960535017e-06, "loss": 1.4195, "step": 2349 }, { "epoch": 0.1841692789968652, "grad_norm": 6.469757080078125, "learning_rate": 9.951939317917381e-06, "loss": 1.0207, "step": 2350 }, { "epoch": 0.1842476489028213, "grad_norm": 2.1882436275482178, "learning_rate": 9.95189665646511e-06, "loss": 1.361, "step": 2351 }, { "epoch": 0.18432601880877744, "grad_norm": 3.8087387084960938, "learning_rate": 9.951853976178367e-06, "loss": 1.3188, "step": 2352 }, { "epoch": 0.18440438871473355, "grad_norm": 6.475164413452148, "learning_rate": 9.951811277057316e-06, "loss": 1.431, "step": 2353 }, { "epoch": 0.18448275862068966, "grad_norm": 3.129868268966675, "learning_rate": 9.95176855910212e-06, "loss": 1.3852, "step": 2354 }, { "epoch": 0.18456112852664577, "grad_norm": 3.052855968475342, "learning_rate": 9.951725822312936e-06, "loss": 1.4895, "step": 2355 }, { "epoch": 0.18463949843260188, "grad_norm": 7.019271373748779, "learning_rate": 9.951683066689933e-06, "loss": 1.5848, "step": 2356 }, { "epoch": 0.184717868338558, "grad_norm": 3.0744855403900146, "learning_rate": 9.95164029223327e-06, "loss": 1.9201, "step": 2357 }, { "epoch": 0.1847962382445141, "grad_norm": 2.9676549434661865, "learning_rate": 9.95159749894311e-06, "loss": 1.6366, "step": 2358 }, { "epoch": 0.18487460815047022, "grad_norm": 4.649144649505615, "learning_rate": 9.951554686819619e-06, "loss": 1.3072, "step": 2359 }, { "epoch": 0.18495297805642633, "grad_norm": 3.1635828018188477, "learning_rate": 9.951511855862954e-06, "loss": 1.4759, "step": 2360 }, { "epoch": 0.18503134796238244, "grad_norm": 2.7678163051605225, "learning_rate": 9.951469006073284e-06, "loss": 1.4685, "step": 2361 }, { "epoch": 0.18510971786833855, "grad_norm": 3.446629762649536, "learning_rate": 9.951426137450768e-06, "loss": 1.5299, "step": 2362 }, { "epoch": 0.18518808777429466, "grad_norm": 3.847965717315674, "learning_rate": 9.95138324999557e-06, "loss": 1.047, "step": 2363 }, { "epoch": 0.1852664576802508, "grad_norm": 3.0393900871276855, "learning_rate": 9.951340343707852e-06, "loss": 1.3733, "step": 2364 }, { "epoch": 0.1853448275862069, "grad_norm": 3.5325357913970947, "learning_rate": 9.95129741858778e-06, "loss": 1.4633, "step": 2365 }, { "epoch": 0.18542319749216302, "grad_norm": 3.507683753967285, "learning_rate": 9.951254474635517e-06, "loss": 1.7743, "step": 2366 }, { "epoch": 0.18550156739811913, "grad_norm": 2.825706720352173, "learning_rate": 9.951211511851223e-06, "loss": 1.6378, "step": 2367 }, { "epoch": 0.18557993730407524, "grad_norm": 3.7876133918762207, "learning_rate": 9.951168530235064e-06, "loss": 1.4609, "step": 2368 }, { "epoch": 0.18565830721003135, "grad_norm": 2.5506362915039062, "learning_rate": 9.951125529787203e-06, "loss": 1.4432, "step": 2369 }, { "epoch": 0.18573667711598746, "grad_norm": 2.729146718978882, "learning_rate": 9.951082510507804e-06, "loss": 1.7345, "step": 2370 }, { "epoch": 0.18581504702194357, "grad_norm": 3.7681527137756348, "learning_rate": 9.951039472397029e-06, "loss": 1.4066, "step": 2371 }, { "epoch": 0.18589341692789968, "grad_norm": 2.177928924560547, "learning_rate": 9.950996415455045e-06, "loss": 1.5568, "step": 2372 }, { "epoch": 0.1859717868338558, "grad_norm": 2.7673966884613037, "learning_rate": 9.950953339682013e-06, "loss": 1.3768, "step": 2373 }, { "epoch": 0.1860501567398119, "grad_norm": 2.8501741886138916, "learning_rate": 9.950910245078096e-06, "loss": 1.3117, "step": 2374 }, { "epoch": 0.18612852664576804, "grad_norm": 2.1615703105926514, "learning_rate": 9.95086713164346e-06, "loss": 1.0942, "step": 2375 }, { "epoch": 0.18620689655172415, "grad_norm": 3.552029609680176, "learning_rate": 9.950823999378268e-06, "loss": 1.1495, "step": 2376 }, { "epoch": 0.18628526645768026, "grad_norm": 9.385090827941895, "learning_rate": 9.950780848282687e-06, "loss": 1.5393, "step": 2377 }, { "epoch": 0.18636363636363637, "grad_norm": 5.238055229187012, "learning_rate": 9.950737678356874e-06, "loss": 2.0278, "step": 2378 }, { "epoch": 0.18644200626959248, "grad_norm": 2.9865825176239014, "learning_rate": 9.950694489601001e-06, "loss": 1.3558, "step": 2379 }, { "epoch": 0.1865203761755486, "grad_norm": 4.740077018737793, "learning_rate": 9.95065128201523e-06, "loss": 1.5728, "step": 2380 }, { "epoch": 0.1865987460815047, "grad_norm": 2.788538694381714, "learning_rate": 9.950608055599722e-06, "loss": 1.0794, "step": 2381 }, { "epoch": 0.1866771159874608, "grad_norm": 4.4743146896362305, "learning_rate": 9.950564810354644e-06, "loss": 1.3113, "step": 2382 }, { "epoch": 0.18675548589341692, "grad_norm": 6.602445125579834, "learning_rate": 9.95052154628016e-06, "loss": 1.3455, "step": 2383 }, { "epoch": 0.18683385579937303, "grad_norm": 3.988652229309082, "learning_rate": 9.950478263376435e-06, "loss": 1.7196, "step": 2384 }, { "epoch": 0.18691222570532914, "grad_norm": 2.9009270668029785, "learning_rate": 9.950434961643635e-06, "loss": 1.1892, "step": 2385 }, { "epoch": 0.18699059561128525, "grad_norm": 3.3853917121887207, "learning_rate": 9.950391641081922e-06, "loss": 1.6285, "step": 2386 }, { "epoch": 0.1870689655172414, "grad_norm": 12.12996768951416, "learning_rate": 9.950348301691462e-06, "loss": 0.8704, "step": 2387 }, { "epoch": 0.1871473354231975, "grad_norm": 2.3042311668395996, "learning_rate": 9.950304943472418e-06, "loss": 1.3292, "step": 2388 }, { "epoch": 0.1872257053291536, "grad_norm": 6.357119083404541, "learning_rate": 9.950261566424959e-06, "loss": 0.9867, "step": 2389 }, { "epoch": 0.18730407523510972, "grad_norm": 4.015512943267822, "learning_rate": 9.950218170549246e-06, "loss": 2.111, "step": 2390 }, { "epoch": 0.18738244514106583, "grad_norm": 3.6444334983825684, "learning_rate": 9.950174755845448e-06, "loss": 1.2232, "step": 2391 }, { "epoch": 0.18746081504702194, "grad_norm": 5.920401573181152, "learning_rate": 9.950131322313725e-06, "loss": 1.2672, "step": 2392 }, { "epoch": 0.18753918495297806, "grad_norm": 6.1970977783203125, "learning_rate": 9.950087869954246e-06, "loss": 0.8618, "step": 2393 }, { "epoch": 0.18761755485893417, "grad_norm": 2.731827974319458, "learning_rate": 9.950044398767175e-06, "loss": 1.3639, "step": 2394 }, { "epoch": 0.18769592476489028, "grad_norm": 3.834264039993286, "learning_rate": 9.95000090875268e-06, "loss": 1.2168, "step": 2395 }, { "epoch": 0.1877742946708464, "grad_norm": 4.664865493774414, "learning_rate": 9.949957399910921e-06, "loss": 1.5969, "step": 2396 }, { "epoch": 0.1878526645768025, "grad_norm": 3.4875705242156982, "learning_rate": 9.949913872242067e-06, "loss": 1.2403, "step": 2397 }, { "epoch": 0.1879310344827586, "grad_norm": 2.477871894836426, "learning_rate": 9.949870325746285e-06, "loss": 0.5566, "step": 2398 }, { "epoch": 0.18800940438871475, "grad_norm": 4.5431904792785645, "learning_rate": 9.949826760423737e-06, "loss": 1.4267, "step": 2399 }, { "epoch": 0.18808777429467086, "grad_norm": 3.0327250957489014, "learning_rate": 9.949783176274591e-06, "loss": 1.4567, "step": 2400 }, { "epoch": 0.18816614420062697, "grad_norm": 4.091623306274414, "learning_rate": 9.949739573299013e-06, "loss": 1.4017, "step": 2401 }, { "epoch": 0.18824451410658308, "grad_norm": 3.172805070877075, "learning_rate": 9.949695951497165e-06, "loss": 1.7613, "step": 2402 }, { "epoch": 0.1883228840125392, "grad_norm": 2.3512794971466064, "learning_rate": 9.949652310869219e-06, "loss": 1.3786, "step": 2403 }, { "epoch": 0.1884012539184953, "grad_norm": 3.372521162033081, "learning_rate": 9.949608651415337e-06, "loss": 1.3912, "step": 2404 }, { "epoch": 0.1884796238244514, "grad_norm": 2.70440936088562, "learning_rate": 9.949564973135686e-06, "loss": 1.2605, "step": 2405 }, { "epoch": 0.18855799373040752, "grad_norm": 6.787710666656494, "learning_rate": 9.949521276030432e-06, "loss": 1.9825, "step": 2406 }, { "epoch": 0.18863636363636363, "grad_norm": 3.8184523582458496, "learning_rate": 9.94947756009974e-06, "loss": 1.4226, "step": 2407 }, { "epoch": 0.18871473354231974, "grad_norm": 2.664942741394043, "learning_rate": 9.94943382534378e-06, "loss": 1.756, "step": 2408 }, { "epoch": 0.18879310344827585, "grad_norm": 3.133650779724121, "learning_rate": 9.949390071762714e-06, "loss": 1.2549, "step": 2409 }, { "epoch": 0.18887147335423196, "grad_norm": 3.266836166381836, "learning_rate": 9.949346299356711e-06, "loss": 1.2977, "step": 2410 }, { "epoch": 0.1889498432601881, "grad_norm": 3.412513256072998, "learning_rate": 9.949302508125936e-06, "loss": 1.4491, "step": 2411 }, { "epoch": 0.1890282131661442, "grad_norm": 4.3147196769714355, "learning_rate": 9.949258698070558e-06, "loss": 1.4675, "step": 2412 }, { "epoch": 0.18910658307210032, "grad_norm": 3.125903844833374, "learning_rate": 9.94921486919074e-06, "loss": 1.4255, "step": 2413 }, { "epoch": 0.18918495297805643, "grad_norm": 4.6250505447387695, "learning_rate": 9.949171021486653e-06, "loss": 2.3419, "step": 2414 }, { "epoch": 0.18926332288401254, "grad_norm": 3.089585542678833, "learning_rate": 9.94912715495846e-06, "loss": 1.5018, "step": 2415 }, { "epoch": 0.18934169278996865, "grad_norm": 5.114130020141602, "learning_rate": 9.94908326960633e-06, "loss": 1.2007, "step": 2416 }, { "epoch": 0.18942006269592476, "grad_norm": 9.34088134765625, "learning_rate": 9.949039365430428e-06, "loss": 2.2265, "step": 2417 }, { "epoch": 0.18949843260188087, "grad_norm": 2.5743820667266846, "learning_rate": 9.948995442430925e-06, "loss": 1.2604, "step": 2418 }, { "epoch": 0.18957680250783698, "grad_norm": 2.9115521907806396, "learning_rate": 9.948951500607983e-06, "loss": 0.9406, "step": 2419 }, { "epoch": 0.1896551724137931, "grad_norm": 3.2615106105804443, "learning_rate": 9.948907539961772e-06, "loss": 1.6812, "step": 2420 }, { "epoch": 0.1897335423197492, "grad_norm": 2.492877960205078, "learning_rate": 9.94886356049246e-06, "loss": 1.3614, "step": 2421 }, { "epoch": 0.18981191222570534, "grad_norm": 2.848690986633301, "learning_rate": 9.94881956220021e-06, "loss": 1.3163, "step": 2422 }, { "epoch": 0.18989028213166145, "grad_norm": 4.271764755249023, "learning_rate": 9.948775545085194e-06, "loss": 1.6956, "step": 2423 }, { "epoch": 0.18996865203761756, "grad_norm": 2.3592026233673096, "learning_rate": 9.948731509147578e-06, "loss": 1.3582, "step": 2424 }, { "epoch": 0.19004702194357367, "grad_norm": 7.283781051635742, "learning_rate": 9.94868745438753e-06, "loss": 1.514, "step": 2425 }, { "epoch": 0.19012539184952978, "grad_norm": 4.262030124664307, "learning_rate": 9.948643380805214e-06, "loss": 1.5268, "step": 2426 }, { "epoch": 0.1902037617554859, "grad_norm": 8.291853904724121, "learning_rate": 9.948599288400802e-06, "loss": 1.3818, "step": 2427 }, { "epoch": 0.190282131661442, "grad_norm": 3.1612114906311035, "learning_rate": 9.94855517717446e-06, "loss": 1.5408, "step": 2428 }, { "epoch": 0.19036050156739812, "grad_norm": 2.3953683376312256, "learning_rate": 9.948511047126355e-06, "loss": 1.4732, "step": 2429 }, { "epoch": 0.19043887147335423, "grad_norm": 2.6607861518859863, "learning_rate": 9.948466898256657e-06, "loss": 1.2366, "step": 2430 }, { "epoch": 0.19051724137931034, "grad_norm": 3.1765549182891846, "learning_rate": 9.94842273056553e-06, "loss": 1.4055, "step": 2431 }, { "epoch": 0.19059561128526645, "grad_norm": 4.308582782745361, "learning_rate": 9.948378544053148e-06, "loss": 1.064, "step": 2432 }, { "epoch": 0.19067398119122256, "grad_norm": 14.43879222869873, "learning_rate": 9.948334338719674e-06, "loss": 1.1986, "step": 2433 }, { "epoch": 0.1907523510971787, "grad_norm": 3.6365785598754883, "learning_rate": 9.948290114565277e-06, "loss": 1.4888, "step": 2434 }, { "epoch": 0.1908307210031348, "grad_norm": 3.3280222415924072, "learning_rate": 9.948245871590127e-06, "loss": 1.6649, "step": 2435 }, { "epoch": 0.19090909090909092, "grad_norm": 2.7486252784729004, "learning_rate": 9.948201609794393e-06, "loss": 1.2947, "step": 2436 }, { "epoch": 0.19098746081504703, "grad_norm": 3.048336982727051, "learning_rate": 9.948157329178238e-06, "loss": 1.7479, "step": 2437 }, { "epoch": 0.19106583072100314, "grad_norm": 2.926340103149414, "learning_rate": 9.948113029741837e-06, "loss": 1.6311, "step": 2438 }, { "epoch": 0.19114420062695925, "grad_norm": 2.375852108001709, "learning_rate": 9.948068711485354e-06, "loss": 1.5504, "step": 2439 }, { "epoch": 0.19122257053291536, "grad_norm": 2.5483436584472656, "learning_rate": 9.94802437440896e-06, "loss": 1.545, "step": 2440 }, { "epoch": 0.19130094043887147, "grad_norm": 4.563296318054199, "learning_rate": 9.947980018512822e-06, "loss": 1.575, "step": 2441 }, { "epoch": 0.19137931034482758, "grad_norm": 1.9909824132919312, "learning_rate": 9.947935643797111e-06, "loss": 1.0107, "step": 2442 }, { "epoch": 0.1914576802507837, "grad_norm": 4.498164653778076, "learning_rate": 9.947891250261994e-06, "loss": 1.1648, "step": 2443 }, { "epoch": 0.1915360501567398, "grad_norm": 2.8440208435058594, "learning_rate": 9.947846837907638e-06, "loss": 1.4239, "step": 2444 }, { "epoch": 0.1916144200626959, "grad_norm": 2.8497936725616455, "learning_rate": 9.947802406734215e-06, "loss": 1.4066, "step": 2445 }, { "epoch": 0.19169278996865205, "grad_norm": 3.2720682621002197, "learning_rate": 9.947757956741894e-06, "loss": 1.3102, "step": 2446 }, { "epoch": 0.19177115987460816, "grad_norm": 3.526698350906372, "learning_rate": 9.947713487930841e-06, "loss": 1.5384, "step": 2447 }, { "epoch": 0.19184952978056427, "grad_norm": 7.429240703582764, "learning_rate": 9.947669000301228e-06, "loss": 1.3417, "step": 2448 }, { "epoch": 0.19192789968652038, "grad_norm": 3.185976028442383, "learning_rate": 9.947624493853226e-06, "loss": 1.5896, "step": 2449 }, { "epoch": 0.1920062695924765, "grad_norm": 4.099184989929199, "learning_rate": 9.947579968586999e-06, "loss": 1.4187, "step": 2450 }, { "epoch": 0.1920846394984326, "grad_norm": 4.829122543334961, "learning_rate": 9.947535424502718e-06, "loss": 1.6521, "step": 2451 }, { "epoch": 0.1921630094043887, "grad_norm": 5.411080837249756, "learning_rate": 9.947490861600556e-06, "loss": 1.0998, "step": 2452 }, { "epoch": 0.19224137931034482, "grad_norm": 3.134310722351074, "learning_rate": 9.947446279880678e-06, "loss": 1.4514, "step": 2453 }, { "epoch": 0.19231974921630093, "grad_norm": 2.9988842010498047, "learning_rate": 9.947401679343258e-06, "loss": 1.629, "step": 2454 }, { "epoch": 0.19239811912225704, "grad_norm": 3.103614568710327, "learning_rate": 9.94735705998846e-06, "loss": 1.5932, "step": 2455 }, { "epoch": 0.19247648902821315, "grad_norm": 5.356962203979492, "learning_rate": 9.94731242181646e-06, "loss": 1.0059, "step": 2456 }, { "epoch": 0.19255485893416927, "grad_norm": 6.555475234985352, "learning_rate": 9.947267764827422e-06, "loss": 1.7935, "step": 2457 }, { "epoch": 0.1926332288401254, "grad_norm": 4.273386001586914, "learning_rate": 9.947223089021521e-06, "loss": 1.6211, "step": 2458 }, { "epoch": 0.1927115987460815, "grad_norm": 3.8306474685668945, "learning_rate": 9.947178394398922e-06, "loss": 1.4911, "step": 2459 }, { "epoch": 0.19278996865203762, "grad_norm": 3.061699628829956, "learning_rate": 9.9471336809598e-06, "loss": 1.4946, "step": 2460 }, { "epoch": 0.19286833855799373, "grad_norm": 2.465583086013794, "learning_rate": 9.94708894870432e-06, "loss": 1.4559, "step": 2461 }, { "epoch": 0.19294670846394985, "grad_norm": 2.878039598464966, "learning_rate": 9.947044197632656e-06, "loss": 1.1657, "step": 2462 }, { "epoch": 0.19302507836990596, "grad_norm": 4.240357398986816, "learning_rate": 9.946999427744976e-06, "loss": 1.4095, "step": 2463 }, { "epoch": 0.19310344827586207, "grad_norm": 7.152384281158447, "learning_rate": 9.946954639041451e-06, "loss": 1.1024, "step": 2464 }, { "epoch": 0.19318181818181818, "grad_norm": 3.537592887878418, "learning_rate": 9.946909831522254e-06, "loss": 1.3607, "step": 2465 }, { "epoch": 0.1932601880877743, "grad_norm": 4.127005577087402, "learning_rate": 9.946865005187549e-06, "loss": 1.3031, "step": 2466 }, { "epoch": 0.1933385579937304, "grad_norm": 2.4395105838775635, "learning_rate": 9.946820160037513e-06, "loss": 1.286, "step": 2467 }, { "epoch": 0.1934169278996865, "grad_norm": 3.1318747997283936, "learning_rate": 9.946775296072312e-06, "loss": 1.319, "step": 2468 }, { "epoch": 0.19349529780564265, "grad_norm": 3.0357859134674072, "learning_rate": 9.94673041329212e-06, "loss": 1.4209, "step": 2469 }, { "epoch": 0.19357366771159876, "grad_norm": 3.5393686294555664, "learning_rate": 9.946685511697108e-06, "loss": 1.3457, "step": 2470 }, { "epoch": 0.19365203761755487, "grad_norm": 2.652395009994507, "learning_rate": 9.946640591287442e-06, "loss": 1.7195, "step": 2471 }, { "epoch": 0.19373040752351098, "grad_norm": 3.0164895057678223, "learning_rate": 9.946595652063298e-06, "loss": 1.5161, "step": 2472 }, { "epoch": 0.1938087774294671, "grad_norm": 3.507675886154175, "learning_rate": 9.946550694024843e-06, "loss": 1.4892, "step": 2473 }, { "epoch": 0.1938871473354232, "grad_norm": 2.645660638809204, "learning_rate": 9.946505717172252e-06, "loss": 1.4685, "step": 2474 }, { "epoch": 0.1939655172413793, "grad_norm": 6.848975658416748, "learning_rate": 9.946460721505693e-06, "loss": 1.2071, "step": 2475 }, { "epoch": 0.19404388714733542, "grad_norm": 4.118106365203857, "learning_rate": 9.946415707025338e-06, "loss": 1.2802, "step": 2476 }, { "epoch": 0.19412225705329153, "grad_norm": 3.845430612564087, "learning_rate": 9.946370673731358e-06, "loss": 1.4348, "step": 2477 }, { "epoch": 0.19420062695924764, "grad_norm": 4.398134708404541, "learning_rate": 9.946325621623924e-06, "loss": 1.5363, "step": 2478 }, { "epoch": 0.19427899686520375, "grad_norm": 3.1522974967956543, "learning_rate": 9.946280550703211e-06, "loss": 1.6827, "step": 2479 }, { "epoch": 0.19435736677115986, "grad_norm": 3.70470929145813, "learning_rate": 9.946235460969384e-06, "loss": 1.3742, "step": 2480 }, { "epoch": 0.194435736677116, "grad_norm": 2.907158136367798, "learning_rate": 9.946190352422619e-06, "loss": 1.3252, "step": 2481 }, { "epoch": 0.1945141065830721, "grad_norm": 3.9626898765563965, "learning_rate": 9.946145225063086e-06, "loss": 1.0958, "step": 2482 }, { "epoch": 0.19459247648902822, "grad_norm": 5.130000114440918, "learning_rate": 9.946100078890958e-06, "loss": 1.7273, "step": 2483 }, { "epoch": 0.19467084639498433, "grad_norm": 2.9977450370788574, "learning_rate": 9.946054913906404e-06, "loss": 1.371, "step": 2484 }, { "epoch": 0.19474921630094044, "grad_norm": 3.9478752613067627, "learning_rate": 9.9460097301096e-06, "loss": 1.5448, "step": 2485 }, { "epoch": 0.19482758620689655, "grad_norm": 2.387575387954712, "learning_rate": 9.945964527500714e-06, "loss": 1.4657, "step": 2486 }, { "epoch": 0.19490595611285266, "grad_norm": 3.142817735671997, "learning_rate": 9.94591930607992e-06, "loss": 1.2971, "step": 2487 }, { "epoch": 0.19498432601880877, "grad_norm": 2.6853859424591064, "learning_rate": 9.945874065847389e-06, "loss": 1.5907, "step": 2488 }, { "epoch": 0.19506269592476488, "grad_norm": 2.8594024181365967, "learning_rate": 9.945828806803292e-06, "loss": 1.2396, "step": 2489 }, { "epoch": 0.195141065830721, "grad_norm": 8.206964492797852, "learning_rate": 9.945783528947805e-06, "loss": 1.3003, "step": 2490 }, { "epoch": 0.1952194357366771, "grad_norm": 2.6473541259765625, "learning_rate": 9.945738232281097e-06, "loss": 1.4801, "step": 2491 }, { "epoch": 0.19529780564263322, "grad_norm": 2.804976463317871, "learning_rate": 9.94569291680334e-06, "loss": 1.4187, "step": 2492 }, { "epoch": 0.19537617554858935, "grad_norm": 7.552343845367432, "learning_rate": 9.945647582514707e-06, "loss": 1.3224, "step": 2493 }, { "epoch": 0.19545454545454546, "grad_norm": 7.284636497497559, "learning_rate": 9.945602229415374e-06, "loss": 1.5404, "step": 2494 }, { "epoch": 0.19553291536050157, "grad_norm": 6.966006278991699, "learning_rate": 9.945556857505507e-06, "loss": 1.4539, "step": 2495 }, { "epoch": 0.19561128526645769, "grad_norm": 4.1185302734375, "learning_rate": 9.945511466785285e-06, "loss": 1.7589, "step": 2496 }, { "epoch": 0.1956896551724138, "grad_norm": 3.697145700454712, "learning_rate": 9.945466057254876e-06, "loss": 1.2453, "step": 2497 }, { "epoch": 0.1957680250783699, "grad_norm": 3.0235514640808105, "learning_rate": 9.945420628914453e-06, "loss": 1.831, "step": 2498 }, { "epoch": 0.19584639498432602, "grad_norm": 2.2258682250976562, "learning_rate": 9.945375181764192e-06, "loss": 1.4774, "step": 2499 }, { "epoch": 0.19592476489028213, "grad_norm": 2.3198320865631104, "learning_rate": 9.945329715804261e-06, "loss": 1.254, "step": 2500 }, { "epoch": 0.19600313479623824, "grad_norm": 3.2972958087921143, "learning_rate": 9.945284231034839e-06, "loss": 1.6893, "step": 2501 }, { "epoch": 0.19608150470219435, "grad_norm": 4.370995044708252, "learning_rate": 9.945238727456095e-06, "loss": 1.7609, "step": 2502 }, { "epoch": 0.19615987460815046, "grad_norm": 8.123254776000977, "learning_rate": 9.945193205068201e-06, "loss": 1.6389, "step": 2503 }, { "epoch": 0.1962382445141066, "grad_norm": 3.8990042209625244, "learning_rate": 9.945147663871335e-06, "loss": 1.6089, "step": 2504 }, { "epoch": 0.1963166144200627, "grad_norm": 3.9489123821258545, "learning_rate": 9.945102103865666e-06, "loss": 1.4334, "step": 2505 }, { "epoch": 0.19639498432601882, "grad_norm": 4.025939464569092, "learning_rate": 9.945056525051369e-06, "loss": 1.4216, "step": 2506 }, { "epoch": 0.19647335423197493, "grad_norm": 2.518717050552368, "learning_rate": 9.945010927428615e-06, "loss": 1.2736, "step": 2507 }, { "epoch": 0.19655172413793104, "grad_norm": 2.88553524017334, "learning_rate": 9.944965310997582e-06, "loss": 1.4939, "step": 2508 }, { "epoch": 0.19663009404388715, "grad_norm": 3.3734898567199707, "learning_rate": 9.94491967575844e-06, "loss": 1.7023, "step": 2509 }, { "epoch": 0.19670846394984326, "grad_norm": 5.082178592681885, "learning_rate": 9.944874021711363e-06, "loss": 1.4882, "step": 2510 }, { "epoch": 0.19678683385579937, "grad_norm": 4.7378458976745605, "learning_rate": 9.944828348856525e-06, "loss": 0.9804, "step": 2511 }, { "epoch": 0.19686520376175548, "grad_norm": 3.194429636001587, "learning_rate": 9.9447826571941e-06, "loss": 1.5643, "step": 2512 }, { "epoch": 0.1969435736677116, "grad_norm": 2.181349039077759, "learning_rate": 9.944736946724261e-06, "loss": 1.0992, "step": 2513 }, { "epoch": 0.1970219435736677, "grad_norm": 2.948251724243164, "learning_rate": 9.944691217447183e-06, "loss": 1.3985, "step": 2514 }, { "epoch": 0.1971003134796238, "grad_norm": 3.032769203186035, "learning_rate": 9.94464546936304e-06, "loss": 1.5324, "step": 2515 }, { "epoch": 0.19717868338557995, "grad_norm": 2.792555570602417, "learning_rate": 9.944599702472004e-06, "loss": 1.6647, "step": 2516 }, { "epoch": 0.19725705329153606, "grad_norm": 3.2863409519195557, "learning_rate": 9.944553916774251e-06, "loss": 1.4236, "step": 2517 }, { "epoch": 0.19733542319749217, "grad_norm": 2.8885695934295654, "learning_rate": 9.944508112269955e-06, "loss": 1.5496, "step": 2518 }, { "epoch": 0.19741379310344828, "grad_norm": 2.4126574993133545, "learning_rate": 9.944462288959288e-06, "loss": 1.2104, "step": 2519 }, { "epoch": 0.1974921630094044, "grad_norm": 3.0251293182373047, "learning_rate": 9.944416446842429e-06, "loss": 1.2489, "step": 2520 }, { "epoch": 0.1975705329153605, "grad_norm": 2.6189990043640137, "learning_rate": 9.944370585919548e-06, "loss": 1.981, "step": 2521 }, { "epoch": 0.1976489028213166, "grad_norm": 2.5486209392547607, "learning_rate": 9.94432470619082e-06, "loss": 0.6565, "step": 2522 }, { "epoch": 0.19772727272727272, "grad_norm": 5.199285984039307, "learning_rate": 9.944278807656422e-06, "loss": 1.5682, "step": 2523 }, { "epoch": 0.19780564263322883, "grad_norm": 4.079289436340332, "learning_rate": 9.944232890316527e-06, "loss": 1.4839, "step": 2524 }, { "epoch": 0.19788401253918494, "grad_norm": 4.43196964263916, "learning_rate": 9.94418695417131e-06, "loss": 1.5548, "step": 2525 }, { "epoch": 0.19796238244514106, "grad_norm": 2.872089147567749, "learning_rate": 9.944140999220942e-06, "loss": 1.4804, "step": 2526 }, { "epoch": 0.19804075235109717, "grad_norm": 4.921505928039551, "learning_rate": 9.944095025465603e-06, "loss": 1.9047, "step": 2527 }, { "epoch": 0.1981191222570533, "grad_norm": 3.338637351989746, "learning_rate": 9.944049032905467e-06, "loss": 1.5924, "step": 2528 }, { "epoch": 0.19819749216300941, "grad_norm": 3.153355598449707, "learning_rate": 9.944003021540707e-06, "loss": 1.6437, "step": 2529 }, { "epoch": 0.19827586206896552, "grad_norm": 2.872478723526001, "learning_rate": 9.9439569913715e-06, "loss": 1.2507, "step": 2530 }, { "epoch": 0.19835423197492164, "grad_norm": 2.637315273284912, "learning_rate": 9.943910942398019e-06, "loss": 1.7873, "step": 2531 }, { "epoch": 0.19843260188087775, "grad_norm": 2.350853681564331, "learning_rate": 9.943864874620439e-06, "loss": 0.97, "step": 2532 }, { "epoch": 0.19851097178683386, "grad_norm": 2.729522466659546, "learning_rate": 9.943818788038938e-06, "loss": 1.3769, "step": 2533 }, { "epoch": 0.19858934169278997, "grad_norm": 4.058058738708496, "learning_rate": 9.94377268265369e-06, "loss": 1.3029, "step": 2534 }, { "epoch": 0.19866771159874608, "grad_norm": 4.843713283538818, "learning_rate": 9.943726558464867e-06, "loss": 1.212, "step": 2535 }, { "epoch": 0.1987460815047022, "grad_norm": 4.022954940795898, "learning_rate": 9.94368041547265e-06, "loss": 1.3974, "step": 2536 }, { "epoch": 0.1988244514106583, "grad_norm": 3.4359192848205566, "learning_rate": 9.943634253677212e-06, "loss": 1.3382, "step": 2537 }, { "epoch": 0.1989028213166144, "grad_norm": 3.51267671585083, "learning_rate": 9.943588073078728e-06, "loss": 1.6395, "step": 2538 }, { "epoch": 0.19898119122257052, "grad_norm": 4.243436813354492, "learning_rate": 9.943541873677374e-06, "loss": 1.3527, "step": 2539 }, { "epoch": 0.19905956112852666, "grad_norm": 4.433006286621094, "learning_rate": 9.943495655473326e-06, "loss": 1.3257, "step": 2540 }, { "epoch": 0.19913793103448277, "grad_norm": 2.193601608276367, "learning_rate": 9.943449418466759e-06, "loss": 1.2861, "step": 2541 }, { "epoch": 0.19921630094043888, "grad_norm": 2.6406924724578857, "learning_rate": 9.943403162657851e-06, "loss": 1.5429, "step": 2542 }, { "epoch": 0.199294670846395, "grad_norm": 2.513537883758545, "learning_rate": 9.943356888046776e-06, "loss": 1.5072, "step": 2543 }, { "epoch": 0.1993730407523511, "grad_norm": 3.604571580886841, "learning_rate": 9.943310594633711e-06, "loss": 1.3653, "step": 2544 }, { "epoch": 0.1994514106583072, "grad_norm": 3.1163532733917236, "learning_rate": 9.94326428241883e-06, "loss": 2.1139, "step": 2545 }, { "epoch": 0.19952978056426332, "grad_norm": 2.960897445678711, "learning_rate": 9.94321795140231e-06, "loss": 1.1355, "step": 2546 }, { "epoch": 0.19960815047021943, "grad_norm": 3.2225379943847656, "learning_rate": 9.943171601584331e-06, "loss": 1.6679, "step": 2547 }, { "epoch": 0.19968652037617554, "grad_norm": 3.35160756111145, "learning_rate": 9.943125232965065e-06, "loss": 1.5663, "step": 2548 }, { "epoch": 0.19976489028213165, "grad_norm": 4.41306209564209, "learning_rate": 9.94307884554469e-06, "loss": 1.3959, "step": 2549 }, { "epoch": 0.19984326018808776, "grad_norm": 4.688736438751221, "learning_rate": 9.94303243932338e-06, "loss": 1.5041, "step": 2550 }, { "epoch": 0.1999216300940439, "grad_norm": 4.557742118835449, "learning_rate": 9.942986014301314e-06, "loss": 1.6907, "step": 2551 }, { "epoch": 0.2, "grad_norm": 8.127303123474121, "learning_rate": 9.94293957047867e-06, "loss": 1.2038, "step": 2552 }, { "epoch": 0.20007836990595612, "grad_norm": 3.761221170425415, "learning_rate": 9.94289310785562e-06, "loss": 1.2722, "step": 2553 }, { "epoch": 0.20015673981191223, "grad_norm": 4.3357648849487305, "learning_rate": 9.942846626432345e-06, "loss": 1.493, "step": 2554 }, { "epoch": 0.20023510971786834, "grad_norm": 3.130218267440796, "learning_rate": 9.94280012620902e-06, "loss": 1.6739, "step": 2555 }, { "epoch": 0.20031347962382445, "grad_norm": 2.8379697799682617, "learning_rate": 9.942753607185823e-06, "loss": 1.18, "step": 2556 }, { "epoch": 0.20039184952978056, "grad_norm": 17.20989418029785, "learning_rate": 9.94270706936293e-06, "loss": 2.2213, "step": 2557 }, { "epoch": 0.20047021943573667, "grad_norm": 2.4678196907043457, "learning_rate": 9.942660512740516e-06, "loss": 1.4448, "step": 2558 }, { "epoch": 0.20054858934169278, "grad_norm": 4.303750514984131, "learning_rate": 9.942613937318761e-06, "loss": 1.3695, "step": 2559 }, { "epoch": 0.2006269592476489, "grad_norm": 2.9323275089263916, "learning_rate": 9.942567343097843e-06, "loss": 1.4055, "step": 2560 }, { "epoch": 0.200705329153605, "grad_norm": 2.3260366916656494, "learning_rate": 9.942520730077934e-06, "loss": 1.5776, "step": 2561 }, { "epoch": 0.20078369905956112, "grad_norm": 4.155732154846191, "learning_rate": 9.942474098259217e-06, "loss": 1.4276, "step": 2562 }, { "epoch": 0.20086206896551725, "grad_norm": 4.084013938903809, "learning_rate": 9.942427447641868e-06, "loss": 0.5599, "step": 2563 }, { "epoch": 0.20094043887147336, "grad_norm": 5.148532867431641, "learning_rate": 9.942380778226062e-06, "loss": 1.4883, "step": 2564 }, { "epoch": 0.20101880877742948, "grad_norm": 5.870424747467041, "learning_rate": 9.94233409001198e-06, "loss": 1.7295, "step": 2565 }, { "epoch": 0.20109717868338559, "grad_norm": 3.1552510261535645, "learning_rate": 9.942287382999795e-06, "loss": 1.4062, "step": 2566 }, { "epoch": 0.2011755485893417, "grad_norm": 4.259579181671143, "learning_rate": 9.94224065718969e-06, "loss": 1.2134, "step": 2567 }, { "epoch": 0.2012539184952978, "grad_norm": 3.458047389984131, "learning_rate": 9.942193912581839e-06, "loss": 1.3742, "step": 2568 }, { "epoch": 0.20133228840125392, "grad_norm": 2.3436055183410645, "learning_rate": 9.94214714917642e-06, "loss": 1.2824, "step": 2569 }, { "epoch": 0.20141065830721003, "grad_norm": 2.430372476577759, "learning_rate": 9.942100366973613e-06, "loss": 1.4447, "step": 2570 }, { "epoch": 0.20148902821316614, "grad_norm": 5.486958026885986, "learning_rate": 9.942053565973592e-06, "loss": 1.7544, "step": 2571 }, { "epoch": 0.20156739811912225, "grad_norm": 2.9833147525787354, "learning_rate": 9.94200674617654e-06, "loss": 1.4589, "step": 2572 }, { "epoch": 0.20164576802507836, "grad_norm": 3.8286845684051514, "learning_rate": 9.941959907582632e-06, "loss": 1.4228, "step": 2573 }, { "epoch": 0.20172413793103447, "grad_norm": 3.798358917236328, "learning_rate": 9.941913050192048e-06, "loss": 0.9435, "step": 2574 }, { "epoch": 0.2018025078369906, "grad_norm": 3.0484120845794678, "learning_rate": 9.941866174004964e-06, "loss": 1.774, "step": 2575 }, { "epoch": 0.20188087774294672, "grad_norm": 6.656331539154053, "learning_rate": 9.941819279021559e-06, "loss": 1.807, "step": 2576 }, { "epoch": 0.20195924764890283, "grad_norm": 4.102455139160156, "learning_rate": 9.941772365242013e-06, "loss": 1.4923, "step": 2577 }, { "epoch": 0.20203761755485894, "grad_norm": 3.598609209060669, "learning_rate": 9.941725432666503e-06, "loss": 1.3032, "step": 2578 }, { "epoch": 0.20211598746081505, "grad_norm": 3.5682308673858643, "learning_rate": 9.941678481295206e-06, "loss": 1.5063, "step": 2579 }, { "epoch": 0.20219435736677116, "grad_norm": 3.693871259689331, "learning_rate": 9.941631511128305e-06, "loss": 1.2527, "step": 2580 }, { "epoch": 0.20227272727272727, "grad_norm": 5.440731525421143, "learning_rate": 9.941584522165974e-06, "loss": 1.8447, "step": 2581 }, { "epoch": 0.20235109717868338, "grad_norm": 2.6924383640289307, "learning_rate": 9.941537514408393e-06, "loss": 1.2086, "step": 2582 }, { "epoch": 0.2024294670846395, "grad_norm": 2.8839921951293945, "learning_rate": 9.941490487855743e-06, "loss": 1.8591, "step": 2583 }, { "epoch": 0.2025078369905956, "grad_norm": 4.067728042602539, "learning_rate": 9.941443442508201e-06, "loss": 1.116, "step": 2584 }, { "epoch": 0.2025862068965517, "grad_norm": 3.7277374267578125, "learning_rate": 9.941396378365947e-06, "loss": 1.7193, "step": 2585 }, { "epoch": 0.20266457680250782, "grad_norm": 2.6156249046325684, "learning_rate": 9.941349295429157e-06, "loss": 1.4088, "step": 2586 }, { "epoch": 0.20274294670846396, "grad_norm": 3.2906970977783203, "learning_rate": 9.941302193698012e-06, "loss": 1.1642, "step": 2587 }, { "epoch": 0.20282131661442007, "grad_norm": 2.603245258331299, "learning_rate": 9.941255073172694e-06, "loss": 1.4652, "step": 2588 }, { "epoch": 0.20289968652037618, "grad_norm": 2.7846481800079346, "learning_rate": 9.941207933853379e-06, "loss": 1.542, "step": 2589 }, { "epoch": 0.2029780564263323, "grad_norm": 2.621980667114258, "learning_rate": 9.941160775740247e-06, "loss": 1.5689, "step": 2590 }, { "epoch": 0.2030564263322884, "grad_norm": 5.020871162414551, "learning_rate": 9.941113598833475e-06, "loss": 1.7335, "step": 2591 }, { "epoch": 0.2031347962382445, "grad_norm": 2.664313793182373, "learning_rate": 9.941066403133248e-06, "loss": 1.5276, "step": 2592 }, { "epoch": 0.20321316614420062, "grad_norm": 3.863435983657837, "learning_rate": 9.941019188639742e-06, "loss": 1.4034, "step": 2593 }, { "epoch": 0.20329153605015673, "grad_norm": 3.4135515689849854, "learning_rate": 9.940971955353136e-06, "loss": 1.3489, "step": 2594 }, { "epoch": 0.20336990595611285, "grad_norm": 9.208494186401367, "learning_rate": 9.94092470327361e-06, "loss": 1.5578, "step": 2595 }, { "epoch": 0.20344827586206896, "grad_norm": 6.154026508331299, "learning_rate": 9.940877432401345e-06, "loss": 1.6729, "step": 2596 }, { "epoch": 0.20352664576802507, "grad_norm": 2.8962759971618652, "learning_rate": 9.940830142736519e-06, "loss": 1.6778, "step": 2597 }, { "epoch": 0.2036050156739812, "grad_norm": 2.2634668350219727, "learning_rate": 9.940782834279315e-06, "loss": 1.438, "step": 2598 }, { "epoch": 0.20368338557993731, "grad_norm": 2.2111692428588867, "learning_rate": 9.940735507029908e-06, "loss": 1.3463, "step": 2599 }, { "epoch": 0.20376175548589343, "grad_norm": 2.800065040588379, "learning_rate": 9.940688160988483e-06, "loss": 1.3542, "step": 2600 }, { "epoch": 0.20384012539184954, "grad_norm": 2.4313440322875977, "learning_rate": 9.940640796155216e-06, "loss": 1.3363, "step": 2601 }, { "epoch": 0.20391849529780565, "grad_norm": 4.04976224899292, "learning_rate": 9.940593412530292e-06, "loss": 1.7912, "step": 2602 }, { "epoch": 0.20399686520376176, "grad_norm": 14.532798767089844, "learning_rate": 9.940546010113887e-06, "loss": 1.6532, "step": 2603 }, { "epoch": 0.20407523510971787, "grad_norm": 2.7568626403808594, "learning_rate": 9.940498588906182e-06, "loss": 1.2292, "step": 2604 }, { "epoch": 0.20415360501567398, "grad_norm": 3.109308958053589, "learning_rate": 9.940451148907357e-06, "loss": 1.3123, "step": 2605 }, { "epoch": 0.2042319749216301, "grad_norm": 3.483182668685913, "learning_rate": 9.940403690117595e-06, "loss": 1.5611, "step": 2606 }, { "epoch": 0.2043103448275862, "grad_norm": 2.938535451889038, "learning_rate": 9.940356212537075e-06, "loss": 1.5067, "step": 2607 }, { "epoch": 0.2043887147335423, "grad_norm": 3.4337382316589355, "learning_rate": 9.940308716165978e-06, "loss": 0.8906, "step": 2608 }, { "epoch": 0.20446708463949842, "grad_norm": 9.297789573669434, "learning_rate": 9.940261201004481e-06, "loss": 1.4321, "step": 2609 }, { "epoch": 0.20454545454545456, "grad_norm": 3.415288209915161, "learning_rate": 9.940213667052772e-06, "loss": 1.4267, "step": 2610 }, { "epoch": 0.20462382445141067, "grad_norm": 18.447526931762695, "learning_rate": 9.940166114311025e-06, "loss": 2.0486, "step": 2611 }, { "epoch": 0.20470219435736678, "grad_norm": 3.849625825881958, "learning_rate": 9.940118542779425e-06, "loss": 1.5087, "step": 2612 }, { "epoch": 0.2047805642633229, "grad_norm": 3.2345173358917236, "learning_rate": 9.94007095245815e-06, "loss": 1.6909, "step": 2613 }, { "epoch": 0.204858934169279, "grad_norm": 3.0876734256744385, "learning_rate": 9.940023343347383e-06, "loss": 1.2541, "step": 2614 }, { "epoch": 0.2049373040752351, "grad_norm": 3.373870611190796, "learning_rate": 9.939975715447305e-06, "loss": 1.5365, "step": 2615 }, { "epoch": 0.20501567398119122, "grad_norm": 2.5033750534057617, "learning_rate": 9.939928068758097e-06, "loss": 1.2946, "step": 2616 }, { "epoch": 0.20509404388714733, "grad_norm": 2.771069049835205, "learning_rate": 9.93988040327994e-06, "loss": 1.3438, "step": 2617 }, { "epoch": 0.20517241379310344, "grad_norm": 3.595548152923584, "learning_rate": 9.939832719013015e-06, "loss": 1.3856, "step": 2618 }, { "epoch": 0.20525078369905955, "grad_norm": 2.9260175228118896, "learning_rate": 9.939785015957501e-06, "loss": 1.4996, "step": 2619 }, { "epoch": 0.20532915360501566, "grad_norm": 3.0371451377868652, "learning_rate": 9.939737294113585e-06, "loss": 1.1585, "step": 2620 }, { "epoch": 0.20540752351097177, "grad_norm": 6.954514503479004, "learning_rate": 9.939689553481445e-06, "loss": 2.3742, "step": 2621 }, { "epoch": 0.2054858934169279, "grad_norm": 5.54807710647583, "learning_rate": 9.939641794061263e-06, "loss": 1.3081, "step": 2622 }, { "epoch": 0.20556426332288402, "grad_norm": 3.4049811363220215, "learning_rate": 9.939594015853222e-06, "loss": 0.828, "step": 2623 }, { "epoch": 0.20564263322884013, "grad_norm": 3.718855381011963, "learning_rate": 9.9395462188575e-06, "loss": 0.9663, "step": 2624 }, { "epoch": 0.20572100313479624, "grad_norm": 9.787200927734375, "learning_rate": 9.939498403074282e-06, "loss": 0.7681, "step": 2625 }, { "epoch": 0.20579937304075235, "grad_norm": 2.5019662380218506, "learning_rate": 9.93945056850375e-06, "loss": 1.1454, "step": 2626 }, { "epoch": 0.20587774294670846, "grad_norm": 3.749397039413452, "learning_rate": 9.939402715146084e-06, "loss": 1.203, "step": 2627 }, { "epoch": 0.20595611285266457, "grad_norm": 5.000209331512451, "learning_rate": 9.93935484300147e-06, "loss": 1.0202, "step": 2628 }, { "epoch": 0.20603448275862069, "grad_norm": 2.982482671737671, "learning_rate": 9.939306952070084e-06, "loss": 1.423, "step": 2629 }, { "epoch": 0.2061128526645768, "grad_norm": 3.5217511653900146, "learning_rate": 9.939259042352113e-06, "loss": 1.4073, "step": 2630 }, { "epoch": 0.2061912225705329, "grad_norm": 6.325048446655273, "learning_rate": 9.939211113847736e-06, "loss": 1.4539, "step": 2631 }, { "epoch": 0.20626959247648902, "grad_norm": 4.419124126434326, "learning_rate": 9.939163166557139e-06, "loss": 1.3304, "step": 2632 }, { "epoch": 0.20634796238244515, "grad_norm": 6.137375354766846, "learning_rate": 9.9391152004805e-06, "loss": 1.435, "step": 2633 }, { "epoch": 0.20642633228840127, "grad_norm": 2.917560577392578, "learning_rate": 9.939067215618006e-06, "loss": 1.4406, "step": 2634 }, { "epoch": 0.20650470219435738, "grad_norm": 2.4865245819091797, "learning_rate": 9.939019211969835e-06, "loss": 1.4666, "step": 2635 }, { "epoch": 0.2065830721003135, "grad_norm": 2.9277310371398926, "learning_rate": 9.938971189536173e-06, "loss": 1.2616, "step": 2636 }, { "epoch": 0.2066614420062696, "grad_norm": 3.045776605606079, "learning_rate": 9.9389231483172e-06, "loss": 1.4913, "step": 2637 }, { "epoch": 0.2067398119122257, "grad_norm": 2.916487455368042, "learning_rate": 9.938875088313102e-06, "loss": 1.1938, "step": 2638 }, { "epoch": 0.20681818181818182, "grad_norm": 7.631406784057617, "learning_rate": 9.93882700952406e-06, "loss": 1.8164, "step": 2639 }, { "epoch": 0.20689655172413793, "grad_norm": 11.137042999267578, "learning_rate": 9.938778911950256e-06, "loss": 0.817, "step": 2640 }, { "epoch": 0.20697492163009404, "grad_norm": 6.9198222160339355, "learning_rate": 9.938730795591873e-06, "loss": 1.6604, "step": 2641 }, { "epoch": 0.20705329153605015, "grad_norm": 3.743529796600342, "learning_rate": 9.938682660449095e-06, "loss": 1.5779, "step": 2642 }, { "epoch": 0.20713166144200626, "grad_norm": 3.625756025314331, "learning_rate": 9.938634506522106e-06, "loss": 1.0986, "step": 2643 }, { "epoch": 0.20721003134796237, "grad_norm": 2.8759636878967285, "learning_rate": 9.938586333811087e-06, "loss": 1.5233, "step": 2644 }, { "epoch": 0.2072884012539185, "grad_norm": 3.296884298324585, "learning_rate": 9.938538142316223e-06, "loss": 1.4262, "step": 2645 }, { "epoch": 0.20736677115987462, "grad_norm": 4.729927062988281, "learning_rate": 9.938489932037696e-06, "loss": 1.7023, "step": 2646 }, { "epoch": 0.20744514106583073, "grad_norm": 3.5042617321014404, "learning_rate": 9.938441702975689e-06, "loss": 1.4866, "step": 2647 }, { "epoch": 0.20752351097178684, "grad_norm": 5.689444541931152, "learning_rate": 9.938393455130388e-06, "loss": 1.6714, "step": 2648 }, { "epoch": 0.20760188087774295, "grad_norm": 3.2485218048095703, "learning_rate": 9.938345188501974e-06, "loss": 1.8822, "step": 2649 }, { "epoch": 0.20768025078369906, "grad_norm": 8.2590913772583, "learning_rate": 9.938296903090631e-06, "loss": 1.275, "step": 2650 }, { "epoch": 0.20775862068965517, "grad_norm": 5.939743518829346, "learning_rate": 9.938248598896543e-06, "loss": 1.0187, "step": 2651 }, { "epoch": 0.20783699059561128, "grad_norm": 9.983951568603516, "learning_rate": 9.938200275919894e-06, "loss": 1.3874, "step": 2652 }, { "epoch": 0.2079153605015674, "grad_norm": 5.398952484130859, "learning_rate": 9.938151934160869e-06, "loss": 1.3148, "step": 2653 }, { "epoch": 0.2079937304075235, "grad_norm": 3.2851130962371826, "learning_rate": 9.938103573619649e-06, "loss": 1.4997, "step": 2654 }, { "epoch": 0.2080721003134796, "grad_norm": 3.3626341819763184, "learning_rate": 9.938055194296418e-06, "loss": 1.8655, "step": 2655 }, { "epoch": 0.20815047021943572, "grad_norm": 2.8398711681365967, "learning_rate": 9.938006796191363e-06, "loss": 1.6042, "step": 2656 }, { "epoch": 0.20822884012539186, "grad_norm": 2.891902446746826, "learning_rate": 9.937958379304668e-06, "loss": 1.5159, "step": 2657 }, { "epoch": 0.20830721003134797, "grad_norm": 8.845734596252441, "learning_rate": 9.937909943636513e-06, "loss": 1.5921, "step": 2658 }, { "epoch": 0.20838557993730408, "grad_norm": 4.212052345275879, "learning_rate": 9.937861489187086e-06, "loss": 1.3913, "step": 2659 }, { "epoch": 0.2084639498432602, "grad_norm": 2.4315080642700195, "learning_rate": 9.93781301595657e-06, "loss": 1.0322, "step": 2660 }, { "epoch": 0.2085423197492163, "grad_norm": 4.742528438568115, "learning_rate": 9.93776452394515e-06, "loss": 1.1356, "step": 2661 }, { "epoch": 0.20862068965517241, "grad_norm": 5.853255748748779, "learning_rate": 9.937716013153009e-06, "loss": 1.8673, "step": 2662 }, { "epoch": 0.20869905956112852, "grad_norm": 2.298882007598877, "learning_rate": 9.937667483580333e-06, "loss": 1.5313, "step": 2663 }, { "epoch": 0.20877742946708464, "grad_norm": 8.473748207092285, "learning_rate": 9.937618935227307e-06, "loss": 1.956, "step": 2664 }, { "epoch": 0.20885579937304075, "grad_norm": 5.017063617706299, "learning_rate": 9.937570368094113e-06, "loss": 1.5266, "step": 2665 }, { "epoch": 0.20893416927899686, "grad_norm": 2.843031167984009, "learning_rate": 9.937521782180937e-06, "loss": 1.2406, "step": 2666 }, { "epoch": 0.20901253918495297, "grad_norm": 3.02004337310791, "learning_rate": 9.937473177487967e-06, "loss": 1.5185, "step": 2667 }, { "epoch": 0.20909090909090908, "grad_norm": 3.276806354522705, "learning_rate": 9.937424554015382e-06, "loss": 1.1271, "step": 2668 }, { "epoch": 0.20916927899686522, "grad_norm": 3.4213340282440186, "learning_rate": 9.937375911763372e-06, "loss": 1.5159, "step": 2669 }, { "epoch": 0.20924764890282133, "grad_norm": 4.005698204040527, "learning_rate": 9.93732725073212e-06, "loss": 1.418, "step": 2670 }, { "epoch": 0.20932601880877744, "grad_norm": 17.790292739868164, "learning_rate": 9.93727857092181e-06, "loss": 1.4274, "step": 2671 }, { "epoch": 0.20940438871473355, "grad_norm": 2.566667079925537, "learning_rate": 9.937229872332628e-06, "loss": 1.5317, "step": 2672 }, { "epoch": 0.20948275862068966, "grad_norm": 2.9802260398864746, "learning_rate": 9.93718115496476e-06, "loss": 1.3755, "step": 2673 }, { "epoch": 0.20956112852664577, "grad_norm": 2.772625684738159, "learning_rate": 9.937132418818391e-06, "loss": 1.4403, "step": 2674 }, { "epoch": 0.20963949843260188, "grad_norm": 4.369579315185547, "learning_rate": 9.937083663893708e-06, "loss": 1.357, "step": 2675 }, { "epoch": 0.209717868338558, "grad_norm": 4.9235382080078125, "learning_rate": 9.937034890190892e-06, "loss": 1.7013, "step": 2676 }, { "epoch": 0.2097962382445141, "grad_norm": 8.073776245117188, "learning_rate": 9.936986097710131e-06, "loss": 1.8462, "step": 2677 }, { "epoch": 0.2098746081504702, "grad_norm": 2.37080717086792, "learning_rate": 9.936937286451612e-06, "loss": 1.1725, "step": 2678 }, { "epoch": 0.20995297805642632, "grad_norm": 6.962230682373047, "learning_rate": 9.936888456415518e-06, "loss": 0.8727, "step": 2679 }, { "epoch": 0.21003134796238246, "grad_norm": 3.393967628479004, "learning_rate": 9.936839607602038e-06, "loss": 1.3199, "step": 2680 }, { "epoch": 0.21010971786833857, "grad_norm": 2.656242847442627, "learning_rate": 9.936790740011355e-06, "loss": 1.2017, "step": 2681 }, { "epoch": 0.21018808777429468, "grad_norm": 2.81913423538208, "learning_rate": 9.936741853643656e-06, "loss": 1.3156, "step": 2682 }, { "epoch": 0.2102664576802508, "grad_norm": 3.257477283477783, "learning_rate": 9.936692948499126e-06, "loss": 1.5371, "step": 2683 }, { "epoch": 0.2103448275862069, "grad_norm": 2.5784833431243896, "learning_rate": 9.936644024577951e-06, "loss": 1.351, "step": 2684 }, { "epoch": 0.210423197492163, "grad_norm": 8.806841850280762, "learning_rate": 9.93659508188032e-06, "loss": 2.0001, "step": 2685 }, { "epoch": 0.21050156739811912, "grad_norm": 4.172999382019043, "learning_rate": 9.936546120406416e-06, "loss": 1.6416, "step": 2686 }, { "epoch": 0.21057993730407523, "grad_norm": 4.573471546173096, "learning_rate": 9.936497140156425e-06, "loss": 1.8198, "step": 2687 }, { "epoch": 0.21065830721003134, "grad_norm": 3.7249300479888916, "learning_rate": 9.936448141130534e-06, "loss": 1.445, "step": 2688 }, { "epoch": 0.21073667711598745, "grad_norm": 4.244632244110107, "learning_rate": 9.936399123328931e-06, "loss": 1.8495, "step": 2689 }, { "epoch": 0.21081504702194356, "grad_norm": 2.571373701095581, "learning_rate": 9.936350086751803e-06, "loss": 1.2819, "step": 2690 }, { "epoch": 0.21089341692789967, "grad_norm": 3.6694748401641846, "learning_rate": 9.936301031399331e-06, "loss": 1.2488, "step": 2691 }, { "epoch": 0.2109717868338558, "grad_norm": 3.4606611728668213, "learning_rate": 9.936251957271708e-06, "loss": 1.5611, "step": 2692 }, { "epoch": 0.21105015673981192, "grad_norm": 2.1456689834594727, "learning_rate": 9.936202864369118e-06, "loss": 1.3607, "step": 2693 }, { "epoch": 0.21112852664576803, "grad_norm": 3.337005615234375, "learning_rate": 9.936153752691747e-06, "loss": 1.3362, "step": 2694 }, { "epoch": 0.21120689655172414, "grad_norm": 3.505471706390381, "learning_rate": 9.936104622239782e-06, "loss": 1.3814, "step": 2695 }, { "epoch": 0.21128526645768025, "grad_norm": 3.3945188522338867, "learning_rate": 9.93605547301341e-06, "loss": 1.4278, "step": 2696 }, { "epoch": 0.21136363636363636, "grad_norm": 2.967769145965576, "learning_rate": 9.936006305012818e-06, "loss": 1.5746, "step": 2697 }, { "epoch": 0.21144200626959248, "grad_norm": 3.72493839263916, "learning_rate": 9.935957118238195e-06, "loss": 1.534, "step": 2698 }, { "epoch": 0.21152037617554859, "grad_norm": 5.522618770599365, "learning_rate": 9.935907912689727e-06, "loss": 0.6918, "step": 2699 }, { "epoch": 0.2115987460815047, "grad_norm": 2.7260634899139404, "learning_rate": 9.935858688367597e-06, "loss": 1.6691, "step": 2700 }, { "epoch": 0.2116771159874608, "grad_norm": 2.8712804317474365, "learning_rate": 9.935809445271998e-06, "loss": 1.5172, "step": 2701 }, { "epoch": 0.21175548589341692, "grad_norm": 4.383260726928711, "learning_rate": 9.935760183403116e-06, "loss": 1.3056, "step": 2702 }, { "epoch": 0.21183385579937303, "grad_norm": 4.737461566925049, "learning_rate": 9.935710902761137e-06, "loss": 1.3883, "step": 2703 }, { "epoch": 0.21191222570532917, "grad_norm": 3.491641044616699, "learning_rate": 9.935661603346248e-06, "loss": 1.4871, "step": 2704 }, { "epoch": 0.21199059561128528, "grad_norm": 3.3924367427825928, "learning_rate": 9.935612285158637e-06, "loss": 1.1727, "step": 2705 }, { "epoch": 0.2120689655172414, "grad_norm": 7.840830326080322, "learning_rate": 9.935562948198495e-06, "loss": 1.8293, "step": 2706 }, { "epoch": 0.2121473354231975, "grad_norm": 3.6174027919769287, "learning_rate": 9.935513592466001e-06, "loss": 1.3581, "step": 2707 }, { "epoch": 0.2122257053291536, "grad_norm": 2.394604444503784, "learning_rate": 9.935464217961352e-06, "loss": 1.3515, "step": 2708 }, { "epoch": 0.21230407523510972, "grad_norm": 3.828503370285034, "learning_rate": 9.935414824684733e-06, "loss": 1.2482, "step": 2709 }, { "epoch": 0.21238244514106583, "grad_norm": 6.484158515930176, "learning_rate": 9.93536541263633e-06, "loss": 1.5087, "step": 2710 }, { "epoch": 0.21246081504702194, "grad_norm": 3.4165830612182617, "learning_rate": 9.935315981816332e-06, "loss": 1.4475, "step": 2711 }, { "epoch": 0.21253918495297805, "grad_norm": 3.0261173248291016, "learning_rate": 9.935266532224926e-06, "loss": 1.3443, "step": 2712 }, { "epoch": 0.21261755485893416, "grad_norm": 6.408670425415039, "learning_rate": 9.9352170638623e-06, "loss": 1.8973, "step": 2713 }, { "epoch": 0.21269592476489027, "grad_norm": 4.190462112426758, "learning_rate": 9.935167576728646e-06, "loss": 1.7501, "step": 2714 }, { "epoch": 0.21277429467084638, "grad_norm": 4.836711883544922, "learning_rate": 9.935118070824148e-06, "loss": 1.6392, "step": 2715 }, { "epoch": 0.21285266457680252, "grad_norm": 4.656955242156982, "learning_rate": 9.935068546148996e-06, "loss": 1.8568, "step": 2716 }, { "epoch": 0.21293103448275863, "grad_norm": 3.4669978618621826, "learning_rate": 9.93501900270338e-06, "loss": 1.233, "step": 2717 }, { "epoch": 0.21300940438871474, "grad_norm": 4.102752685546875, "learning_rate": 9.934969440487483e-06, "loss": 1.2153, "step": 2718 }, { "epoch": 0.21308777429467085, "grad_norm": 3.3843319416046143, "learning_rate": 9.9349198595015e-06, "loss": 1.4008, "step": 2719 }, { "epoch": 0.21316614420062696, "grad_norm": 5.622435092926025, "learning_rate": 9.934870259745614e-06, "loss": 1.8296, "step": 2720 }, { "epoch": 0.21324451410658307, "grad_norm": 3.360924005508423, "learning_rate": 9.934820641220018e-06, "loss": 1.3979, "step": 2721 }, { "epoch": 0.21332288401253918, "grad_norm": 3.743847608566284, "learning_rate": 9.934771003924897e-06, "loss": 1.517, "step": 2722 }, { "epoch": 0.2134012539184953, "grad_norm": 3.3611109256744385, "learning_rate": 9.934721347860444e-06, "loss": 1.2301, "step": 2723 }, { "epoch": 0.2134796238244514, "grad_norm": 2.7799220085144043, "learning_rate": 9.934671673026844e-06, "loss": 1.0183, "step": 2724 }, { "epoch": 0.2135579937304075, "grad_norm": 7.823351860046387, "learning_rate": 9.934621979424289e-06, "loss": 1.5254, "step": 2725 }, { "epoch": 0.21363636363636362, "grad_norm": 3.0918586254119873, "learning_rate": 9.934572267052966e-06, "loss": 1.1538, "step": 2726 }, { "epoch": 0.21371473354231976, "grad_norm": 7.155656337738037, "learning_rate": 9.934522535913063e-06, "loss": 0.6949, "step": 2727 }, { "epoch": 0.21379310344827587, "grad_norm": 4.783837795257568, "learning_rate": 9.934472786004774e-06, "loss": 1.0973, "step": 2728 }, { "epoch": 0.21387147335423198, "grad_norm": 4.005302429199219, "learning_rate": 9.934423017328284e-06, "loss": 1.4767, "step": 2729 }, { "epoch": 0.2139498432601881, "grad_norm": 7.43612003326416, "learning_rate": 9.93437322988378e-06, "loss": 0.8855, "step": 2730 }, { "epoch": 0.2140282131661442, "grad_norm": 3.562615394592285, "learning_rate": 9.934323423671457e-06, "loss": 1.1349, "step": 2731 }, { "epoch": 0.21410658307210031, "grad_norm": 2.319118022918701, "learning_rate": 9.934273598691503e-06, "loss": 1.357, "step": 2732 }, { "epoch": 0.21418495297805643, "grad_norm": 2.7060909271240234, "learning_rate": 9.934223754944106e-06, "loss": 1.4101, "step": 2733 }, { "epoch": 0.21426332288401254, "grad_norm": 6.5574493408203125, "learning_rate": 9.934173892429458e-06, "loss": 1.5771, "step": 2734 }, { "epoch": 0.21434169278996865, "grad_norm": 2.753086566925049, "learning_rate": 9.934124011147744e-06, "loss": 0.9892, "step": 2735 }, { "epoch": 0.21442006269592476, "grad_norm": 6.054969310760498, "learning_rate": 9.934074111099158e-06, "loss": 1.9761, "step": 2736 }, { "epoch": 0.21449843260188087, "grad_norm": 3.3748738765716553, "learning_rate": 9.934024192283888e-06, "loss": 1.2647, "step": 2737 }, { "epoch": 0.21457680250783698, "grad_norm": 2.6002795696258545, "learning_rate": 9.933974254702125e-06, "loss": 1.0467, "step": 2738 }, { "epoch": 0.21465517241379312, "grad_norm": 3.588996648788452, "learning_rate": 9.933924298354056e-06, "loss": 2.0851, "step": 2739 }, { "epoch": 0.21473354231974923, "grad_norm": 2.4061901569366455, "learning_rate": 9.933874323239876e-06, "loss": 1.0403, "step": 2740 }, { "epoch": 0.21481191222570534, "grad_norm": 8.01933479309082, "learning_rate": 9.933824329359771e-06, "loss": 1.7676, "step": 2741 }, { "epoch": 0.21489028213166145, "grad_norm": 2.6320080757141113, "learning_rate": 9.933774316713932e-06, "loss": 1.191, "step": 2742 }, { "epoch": 0.21496865203761756, "grad_norm": 2.8946146965026855, "learning_rate": 9.93372428530255e-06, "loss": 1.1541, "step": 2743 }, { "epoch": 0.21504702194357367, "grad_norm": 2.7013115882873535, "learning_rate": 9.933674235125816e-06, "loss": 1.1042, "step": 2744 }, { "epoch": 0.21512539184952978, "grad_norm": 4.846853733062744, "learning_rate": 9.933624166183917e-06, "loss": 0.8706, "step": 2745 }, { "epoch": 0.2152037617554859, "grad_norm": 2.8471570014953613, "learning_rate": 9.933574078477048e-06, "loss": 1.3716, "step": 2746 }, { "epoch": 0.215282131661442, "grad_norm": 6.919156074523926, "learning_rate": 9.933523972005396e-06, "loss": 1.5672, "step": 2747 }, { "epoch": 0.2153605015673981, "grad_norm": 3.2943294048309326, "learning_rate": 9.933473846769153e-06, "loss": 1.666, "step": 2748 }, { "epoch": 0.21543887147335422, "grad_norm": 3.1145575046539307, "learning_rate": 9.93342370276851e-06, "loss": 1.4995, "step": 2749 }, { "epoch": 0.21551724137931033, "grad_norm": 3.44496488571167, "learning_rate": 9.933373540003656e-06, "loss": 1.4479, "step": 2750 }, { "epoch": 0.21559561128526647, "grad_norm": 2.6571552753448486, "learning_rate": 9.933323358474785e-06, "loss": 1.6244, "step": 2751 }, { "epoch": 0.21567398119122258, "grad_norm": 2.343350887298584, "learning_rate": 9.933273158182085e-06, "loss": 1.1314, "step": 2752 }, { "epoch": 0.2157523510971787, "grad_norm": 3.3194499015808105, "learning_rate": 9.933222939125746e-06, "loss": 1.5397, "step": 2753 }, { "epoch": 0.2158307210031348, "grad_norm": 3.7632663249969482, "learning_rate": 9.933172701305964e-06, "loss": 1.3791, "step": 2754 }, { "epoch": 0.2159090909090909, "grad_norm": 2.39975643157959, "learning_rate": 9.933122444722924e-06, "loss": 1.322, "step": 2755 }, { "epoch": 0.21598746081504702, "grad_norm": 2.994060516357422, "learning_rate": 9.933072169376821e-06, "loss": 1.5338, "step": 2756 }, { "epoch": 0.21606583072100313, "grad_norm": 4.298651218414307, "learning_rate": 9.933021875267846e-06, "loss": 1.3347, "step": 2757 }, { "epoch": 0.21614420062695924, "grad_norm": 2.6486475467681885, "learning_rate": 9.932971562396189e-06, "loss": 1.3994, "step": 2758 }, { "epoch": 0.21622257053291535, "grad_norm": 3.68770170211792, "learning_rate": 9.932921230762043e-06, "loss": 1.4908, "step": 2759 }, { "epoch": 0.21630094043887146, "grad_norm": 2.6630451679229736, "learning_rate": 9.932870880365597e-06, "loss": 1.3946, "step": 2760 }, { "epoch": 0.21637931034482757, "grad_norm": 3.7752678394317627, "learning_rate": 9.932820511207042e-06, "loss": 1.1109, "step": 2761 }, { "epoch": 0.2164576802507837, "grad_norm": 3.6041879653930664, "learning_rate": 9.932770123286574e-06, "loss": 1.43, "step": 2762 }, { "epoch": 0.21653605015673982, "grad_norm": 5.348921298980713, "learning_rate": 9.932719716604382e-06, "loss": 1.5912, "step": 2763 }, { "epoch": 0.21661442006269593, "grad_norm": 2.996549606323242, "learning_rate": 9.932669291160656e-06, "loss": 1.3965, "step": 2764 }, { "epoch": 0.21669278996865204, "grad_norm": 4.022462368011475, "learning_rate": 9.932618846955592e-06, "loss": 2.3974, "step": 2765 }, { "epoch": 0.21677115987460815, "grad_norm": 2.8560636043548584, "learning_rate": 9.932568383989379e-06, "loss": 1.6222, "step": 2766 }, { "epoch": 0.21684952978056427, "grad_norm": 3.1431996822357178, "learning_rate": 9.93251790226221e-06, "loss": 1.4103, "step": 2767 }, { "epoch": 0.21692789968652038, "grad_norm": 3.3962745666503906, "learning_rate": 9.932467401774275e-06, "loss": 1.6741, "step": 2768 }, { "epoch": 0.2170062695924765, "grad_norm": 3.472721576690674, "learning_rate": 9.932416882525768e-06, "loss": 1.1298, "step": 2769 }, { "epoch": 0.2170846394984326, "grad_norm": 2.584080934524536, "learning_rate": 9.932366344516879e-06, "loss": 1.5498, "step": 2770 }, { "epoch": 0.2171630094043887, "grad_norm": 7.031966209411621, "learning_rate": 9.932315787747804e-06, "loss": 1.7284, "step": 2771 }, { "epoch": 0.21724137931034482, "grad_norm": 2.4193167686462402, "learning_rate": 9.932265212218732e-06, "loss": 1.3207, "step": 2772 }, { "epoch": 0.21731974921630093, "grad_norm": 7.288213729858398, "learning_rate": 9.932214617929857e-06, "loss": 1.5556, "step": 2773 }, { "epoch": 0.21739811912225707, "grad_norm": 3.4501125812530518, "learning_rate": 9.93216400488137e-06, "loss": 1.4274, "step": 2774 }, { "epoch": 0.21747648902821318, "grad_norm": 6.98483943939209, "learning_rate": 9.932113373073466e-06, "loss": 0.8007, "step": 2775 }, { "epoch": 0.2175548589341693, "grad_norm": 2.9751179218292236, "learning_rate": 9.932062722506335e-06, "loss": 1.4889, "step": 2776 }, { "epoch": 0.2176332288401254, "grad_norm": 5.9476399421691895, "learning_rate": 9.932012053180171e-06, "loss": 1.0748, "step": 2777 }, { "epoch": 0.2177115987460815, "grad_norm": 2.912538766860962, "learning_rate": 9.931961365095167e-06, "loss": 1.4472, "step": 2778 }, { "epoch": 0.21778996865203762, "grad_norm": 2.145237684249878, "learning_rate": 9.931910658251514e-06, "loss": 1.5488, "step": 2779 }, { "epoch": 0.21786833855799373, "grad_norm": 13.115227699279785, "learning_rate": 9.931859932649406e-06, "loss": 1.6426, "step": 2780 }, { "epoch": 0.21794670846394984, "grad_norm": 2.993847370147705, "learning_rate": 9.931809188289036e-06, "loss": 1.6042, "step": 2781 }, { "epoch": 0.21802507836990595, "grad_norm": 5.453035831451416, "learning_rate": 9.931758425170599e-06, "loss": 1.567, "step": 2782 }, { "epoch": 0.21810344827586206, "grad_norm": 2.187833786010742, "learning_rate": 9.931707643294284e-06, "loss": 1.4274, "step": 2783 }, { "epoch": 0.21818181818181817, "grad_norm": 9.044525146484375, "learning_rate": 9.931656842660287e-06, "loss": 1.6407, "step": 2784 }, { "epoch": 0.21826018808777428, "grad_norm": 3.917888641357422, "learning_rate": 9.931606023268801e-06, "loss": 1.1784, "step": 2785 }, { "epoch": 0.21833855799373042, "grad_norm": 6.565915107727051, "learning_rate": 9.931555185120018e-06, "loss": 1.807, "step": 2786 }, { "epoch": 0.21841692789968653, "grad_norm": 4.3185625076293945, "learning_rate": 9.931504328214133e-06, "loss": 1.6672, "step": 2787 }, { "epoch": 0.21849529780564264, "grad_norm": 2.501380443572998, "learning_rate": 9.931453452551336e-06, "loss": 1.4684, "step": 2788 }, { "epoch": 0.21857366771159875, "grad_norm": 4.207957744598389, "learning_rate": 9.931402558131823e-06, "loss": 2.0267, "step": 2789 }, { "epoch": 0.21865203761755486, "grad_norm": 2.768026828765869, "learning_rate": 9.931351644955791e-06, "loss": 1.1112, "step": 2790 }, { "epoch": 0.21873040752351097, "grad_norm": 3.8805289268493652, "learning_rate": 9.931300713023429e-06, "loss": 1.5872, "step": 2791 }, { "epoch": 0.21880877742946708, "grad_norm": 5.848959922790527, "learning_rate": 9.931249762334931e-06, "loss": 1.3652, "step": 2792 }, { "epoch": 0.2188871473354232, "grad_norm": 4.990512847900391, "learning_rate": 9.931198792890491e-06, "loss": 1.3858, "step": 2793 }, { "epoch": 0.2189655172413793, "grad_norm": 3.039443016052246, "learning_rate": 9.931147804690306e-06, "loss": 1.5013, "step": 2794 }, { "epoch": 0.21904388714733541, "grad_norm": 6.098604202270508, "learning_rate": 9.931096797734565e-06, "loss": 1.9938, "step": 2795 }, { "epoch": 0.21912225705329152, "grad_norm": 5.293435573577881, "learning_rate": 9.931045772023467e-06, "loss": 1.8044, "step": 2796 }, { "epoch": 0.21920062695924764, "grad_norm": 2.9804012775421143, "learning_rate": 9.930994727557202e-06, "loss": 1.6566, "step": 2797 }, { "epoch": 0.21927899686520377, "grad_norm": 2.83992600440979, "learning_rate": 9.930943664335966e-06, "loss": 1.6423, "step": 2798 }, { "epoch": 0.21935736677115988, "grad_norm": 2.648862361907959, "learning_rate": 9.930892582359954e-06, "loss": 1.0013, "step": 2799 }, { "epoch": 0.219435736677116, "grad_norm": 2.532890558242798, "learning_rate": 9.930841481629358e-06, "loss": 1.3367, "step": 2800 }, { "epoch": 0.2195141065830721, "grad_norm": 2.3312010765075684, "learning_rate": 9.930790362144375e-06, "loss": 1.3747, "step": 2801 }, { "epoch": 0.21959247648902822, "grad_norm": 2.2962725162506104, "learning_rate": 9.930739223905197e-06, "loss": 1.7549, "step": 2802 }, { "epoch": 0.21967084639498433, "grad_norm": 5.197364330291748, "learning_rate": 9.93068806691202e-06, "loss": 1.6134, "step": 2803 }, { "epoch": 0.21974921630094044, "grad_norm": 4.618096828460693, "learning_rate": 9.930636891165037e-06, "loss": 1.4293, "step": 2804 }, { "epoch": 0.21982758620689655, "grad_norm": 6.548649311065674, "learning_rate": 9.930585696664445e-06, "loss": 0.8399, "step": 2805 }, { "epoch": 0.21990595611285266, "grad_norm": 3.8427276611328125, "learning_rate": 9.930534483410438e-06, "loss": 1.5576, "step": 2806 }, { "epoch": 0.21998432601880877, "grad_norm": 2.3601043224334717, "learning_rate": 9.930483251403208e-06, "loss": 1.3114, "step": 2807 }, { "epoch": 0.22006269592476488, "grad_norm": 1.9880081415176392, "learning_rate": 9.930432000642955e-06, "loss": 1.4173, "step": 2808 }, { "epoch": 0.22014106583072102, "grad_norm": 3.6095824241638184, "learning_rate": 9.930380731129869e-06, "loss": 0.8638, "step": 2809 }, { "epoch": 0.22021943573667713, "grad_norm": 4.2519025802612305, "learning_rate": 9.930329442864148e-06, "loss": 1.6437, "step": 2810 }, { "epoch": 0.22029780564263324, "grad_norm": 3.363693952560425, "learning_rate": 9.930278135845987e-06, "loss": 1.5882, "step": 2811 }, { "epoch": 0.22037617554858935, "grad_norm": 2.7604610919952393, "learning_rate": 9.93022681007558e-06, "loss": 1.1752, "step": 2812 }, { "epoch": 0.22045454545454546, "grad_norm": 8.34568977355957, "learning_rate": 9.930175465553123e-06, "loss": 1.1892, "step": 2813 }, { "epoch": 0.22053291536050157, "grad_norm": 6.489311695098877, "learning_rate": 9.930124102278809e-06, "loss": 1.6267, "step": 2814 }, { "epoch": 0.22061128526645768, "grad_norm": 3.0081849098205566, "learning_rate": 9.930072720252835e-06, "loss": 1.0452, "step": 2815 }, { "epoch": 0.2206896551724138, "grad_norm": 3.1043906211853027, "learning_rate": 9.930021319475399e-06, "loss": 1.2753, "step": 2816 }, { "epoch": 0.2207680250783699, "grad_norm": 2.4808266162872314, "learning_rate": 9.929969899946692e-06, "loss": 1.5304, "step": 2817 }, { "epoch": 0.220846394984326, "grad_norm": 6.612785339355469, "learning_rate": 9.929918461666913e-06, "loss": 1.8219, "step": 2818 }, { "epoch": 0.22092476489028212, "grad_norm": 5.624678134918213, "learning_rate": 9.929867004636256e-06, "loss": 1.4477, "step": 2819 }, { "epoch": 0.22100313479623823, "grad_norm": 3.1128368377685547, "learning_rate": 9.929815528854916e-06, "loss": 1.5778, "step": 2820 }, { "epoch": 0.22108150470219437, "grad_norm": 4.958602428436279, "learning_rate": 9.92976403432309e-06, "loss": 1.8233, "step": 2821 }, { "epoch": 0.22115987460815048, "grad_norm": 4.412398815155029, "learning_rate": 9.929712521040974e-06, "loss": 1.1768, "step": 2822 }, { "epoch": 0.2212382445141066, "grad_norm": 2.540684938430786, "learning_rate": 9.929660989008764e-06, "loss": 1.66, "step": 2823 }, { "epoch": 0.2213166144200627, "grad_norm": 5.26088285446167, "learning_rate": 9.929609438226658e-06, "loss": 1.4029, "step": 2824 }, { "epoch": 0.2213949843260188, "grad_norm": 3.3425307273864746, "learning_rate": 9.929557868694846e-06, "loss": 1.438, "step": 2825 }, { "epoch": 0.22147335423197492, "grad_norm": 2.264272451400757, "learning_rate": 9.92950628041353e-06, "loss": 1.4156, "step": 2826 }, { "epoch": 0.22155172413793103, "grad_norm": 3.207547664642334, "learning_rate": 9.929454673382904e-06, "loss": 1.5498, "step": 2827 }, { "epoch": 0.22163009404388714, "grad_norm": 9.339245796203613, "learning_rate": 9.929403047603164e-06, "loss": 1.8186, "step": 2828 }, { "epoch": 0.22170846394984325, "grad_norm": 3.2732856273651123, "learning_rate": 9.929351403074505e-06, "loss": 1.7405, "step": 2829 }, { "epoch": 0.22178683385579936, "grad_norm": 3.739483594894409, "learning_rate": 9.929299739797127e-06, "loss": 0.6937, "step": 2830 }, { "epoch": 0.22186520376175548, "grad_norm": 2.565507411956787, "learning_rate": 9.929248057771224e-06, "loss": 1.5367, "step": 2831 }, { "epoch": 0.22194357366771159, "grad_norm": 2.879319667816162, "learning_rate": 9.929196356996993e-06, "loss": 1.6581, "step": 2832 }, { "epoch": 0.22202194357366772, "grad_norm": 2.43908429145813, "learning_rate": 9.929144637474631e-06, "loss": 1.3657, "step": 2833 }, { "epoch": 0.22210031347962383, "grad_norm": 2.6282808780670166, "learning_rate": 9.929092899204335e-06, "loss": 1.4502, "step": 2834 }, { "epoch": 0.22217868338557994, "grad_norm": 3.4111411571502686, "learning_rate": 9.929041142186301e-06, "loss": 1.5147, "step": 2835 }, { "epoch": 0.22225705329153606, "grad_norm": 2.558011531829834, "learning_rate": 9.928989366420726e-06, "loss": 1.2841, "step": 2836 }, { "epoch": 0.22233542319749217, "grad_norm": 3.431570053100586, "learning_rate": 9.928937571907808e-06, "loss": 1.3869, "step": 2837 }, { "epoch": 0.22241379310344828, "grad_norm": 3.4057092666625977, "learning_rate": 9.928885758647742e-06, "loss": 0.924, "step": 2838 }, { "epoch": 0.2224921630094044, "grad_norm": 2.3164780139923096, "learning_rate": 9.928833926640726e-06, "loss": 1.6578, "step": 2839 }, { "epoch": 0.2225705329153605, "grad_norm": 5.247898578643799, "learning_rate": 9.928782075886959e-06, "loss": 1.6733, "step": 2840 }, { "epoch": 0.2226489028213166, "grad_norm": 4.671698093414307, "learning_rate": 9.928730206386636e-06, "loss": 1.2181, "step": 2841 }, { "epoch": 0.22272727272727272, "grad_norm": 3.1425459384918213, "learning_rate": 9.928678318139954e-06, "loss": 1.3504, "step": 2842 }, { "epoch": 0.22280564263322883, "grad_norm": 3.0158894062042236, "learning_rate": 9.928626411147114e-06, "loss": 1.0835, "step": 2843 }, { "epoch": 0.22288401253918494, "grad_norm": 2.82692813873291, "learning_rate": 9.928574485408306e-06, "loss": 1.3234, "step": 2844 }, { "epoch": 0.22296238244514108, "grad_norm": 5.884823799133301, "learning_rate": 9.928522540923736e-06, "loss": 1.7539, "step": 2845 }, { "epoch": 0.2230407523510972, "grad_norm": 3.6903209686279297, "learning_rate": 9.928470577693597e-06, "loss": 1.3874, "step": 2846 }, { "epoch": 0.2231191222570533, "grad_norm": 3.4373483657836914, "learning_rate": 9.928418595718086e-06, "loss": 1.5772, "step": 2847 }, { "epoch": 0.2231974921630094, "grad_norm": 3.31968092918396, "learning_rate": 9.928366594997401e-06, "loss": 0.5179, "step": 2848 }, { "epoch": 0.22327586206896552, "grad_norm": 2.488577365875244, "learning_rate": 9.928314575531743e-06, "loss": 1.4772, "step": 2849 }, { "epoch": 0.22335423197492163, "grad_norm": 2.601001024246216, "learning_rate": 9.928262537321307e-06, "loss": 1.5815, "step": 2850 }, { "epoch": 0.22343260188087774, "grad_norm": 3.3400723934173584, "learning_rate": 9.928210480366291e-06, "loss": 1.1829, "step": 2851 }, { "epoch": 0.22351097178683385, "grad_norm": 7.282707214355469, "learning_rate": 9.928158404666895e-06, "loss": 1.6206, "step": 2852 }, { "epoch": 0.22358934169278996, "grad_norm": 2.641103982925415, "learning_rate": 9.928106310223315e-06, "loss": 1.4104, "step": 2853 }, { "epoch": 0.22366771159874607, "grad_norm": 3.282647132873535, "learning_rate": 9.92805419703575e-06, "loss": 1.3708, "step": 2854 }, { "epoch": 0.22374608150470218, "grad_norm": 12.331071853637695, "learning_rate": 9.928002065104398e-06, "loss": 1.1396, "step": 2855 }, { "epoch": 0.22382445141065832, "grad_norm": 3.068387031555176, "learning_rate": 9.927949914429457e-06, "loss": 1.3048, "step": 2856 }, { "epoch": 0.22390282131661443, "grad_norm": 6.923477649688721, "learning_rate": 9.927897745011128e-06, "loss": 1.3409, "step": 2857 }, { "epoch": 0.22398119122257054, "grad_norm": 3.532014846801758, "learning_rate": 9.927845556849605e-06, "loss": 1.5994, "step": 2858 }, { "epoch": 0.22405956112852665, "grad_norm": 3.166306495666504, "learning_rate": 9.927793349945087e-06, "loss": 1.3813, "step": 2859 }, { "epoch": 0.22413793103448276, "grad_norm": 3.6062817573547363, "learning_rate": 9.927741124297776e-06, "loss": 1.2039, "step": 2860 }, { "epoch": 0.22421630094043887, "grad_norm": 4.150427341461182, "learning_rate": 9.92768887990787e-06, "loss": 1.2457, "step": 2861 }, { "epoch": 0.22429467084639498, "grad_norm": 4.03614616394043, "learning_rate": 9.927636616775565e-06, "loss": 1.5154, "step": 2862 }, { "epoch": 0.2243730407523511, "grad_norm": 3.3024377822875977, "learning_rate": 9.927584334901062e-06, "loss": 1.1917, "step": 2863 }, { "epoch": 0.2244514106583072, "grad_norm": 3.317608118057251, "learning_rate": 9.927532034284558e-06, "loss": 1.3679, "step": 2864 }, { "epoch": 0.22452978056426331, "grad_norm": 7.664342403411865, "learning_rate": 9.927479714926254e-06, "loss": 1.7595, "step": 2865 }, { "epoch": 0.22460815047021943, "grad_norm": 2.171962261199951, "learning_rate": 9.927427376826347e-06, "loss": 1.3187, "step": 2866 }, { "epoch": 0.22468652037617554, "grad_norm": 2.5600221157073975, "learning_rate": 9.92737501998504e-06, "loss": 1.5993, "step": 2867 }, { "epoch": 0.22476489028213167, "grad_norm": 3.4569990634918213, "learning_rate": 9.927322644402526e-06, "loss": 0.9282, "step": 2868 }, { "epoch": 0.22484326018808778, "grad_norm": 3.8771393299102783, "learning_rate": 9.927270250079009e-06, "loss": 1.3586, "step": 2869 }, { "epoch": 0.2249216300940439, "grad_norm": 3.411809206008911, "learning_rate": 9.927217837014687e-06, "loss": 1.8337, "step": 2870 }, { "epoch": 0.225, "grad_norm": 2.379484176635742, "learning_rate": 9.927165405209757e-06, "loss": 1.2217, "step": 2871 }, { "epoch": 0.22507836990595612, "grad_norm": 2.7023613452911377, "learning_rate": 9.927112954664422e-06, "loss": 1.6064, "step": 2872 }, { "epoch": 0.22515673981191223, "grad_norm": 3.270071029663086, "learning_rate": 9.92706048537888e-06, "loss": 1.0406, "step": 2873 }, { "epoch": 0.22523510971786834, "grad_norm": 2.535822868347168, "learning_rate": 9.927007997353332e-06, "loss": 1.413, "step": 2874 }, { "epoch": 0.22531347962382445, "grad_norm": 2.491856098175049, "learning_rate": 9.926955490587975e-06, "loss": 1.3395, "step": 2875 }, { "epoch": 0.22539184952978056, "grad_norm": 2.7039294242858887, "learning_rate": 9.926902965083011e-06, "loss": 1.573, "step": 2876 }, { "epoch": 0.22547021943573667, "grad_norm": 2.477440595626831, "learning_rate": 9.926850420838638e-06, "loss": 1.3211, "step": 2877 }, { "epoch": 0.22554858934169278, "grad_norm": 2.9278485774993896, "learning_rate": 9.926797857855055e-06, "loss": 1.3612, "step": 2878 }, { "epoch": 0.2256269592476489, "grad_norm": 2.092301607131958, "learning_rate": 9.926745276132465e-06, "loss": 1.187, "step": 2879 }, { "epoch": 0.22570532915360503, "grad_norm": 3.135500907897949, "learning_rate": 9.926692675671066e-06, "loss": 1.2945, "step": 2880 }, { "epoch": 0.22578369905956114, "grad_norm": 3.315674304962158, "learning_rate": 9.92664005647106e-06, "loss": 1.4909, "step": 2881 }, { "epoch": 0.22586206896551725, "grad_norm": 3.7310118675231934, "learning_rate": 9.926587418532646e-06, "loss": 1.1401, "step": 2882 }, { "epoch": 0.22594043887147336, "grad_norm": 5.141915798187256, "learning_rate": 9.92653476185602e-06, "loss": 1.1434, "step": 2883 }, { "epoch": 0.22601880877742947, "grad_norm": 4.93487548828125, "learning_rate": 9.92648208644139e-06, "loss": 1.9438, "step": 2884 }, { "epoch": 0.22609717868338558, "grad_norm": 7.382339000701904, "learning_rate": 9.926429392288952e-06, "loss": 0.9976, "step": 2885 }, { "epoch": 0.2261755485893417, "grad_norm": 3.7383382320404053, "learning_rate": 9.926376679398906e-06, "loss": 1.5854, "step": 2886 }, { "epoch": 0.2262539184952978, "grad_norm": 6.625704765319824, "learning_rate": 9.926323947771453e-06, "loss": 1.0793, "step": 2887 }, { "epoch": 0.2263322884012539, "grad_norm": 6.560940742492676, "learning_rate": 9.926271197406796e-06, "loss": 0.9385, "step": 2888 }, { "epoch": 0.22641065830721002, "grad_norm": 3.1947169303894043, "learning_rate": 9.926218428305133e-06, "loss": 1.5208, "step": 2889 }, { "epoch": 0.22648902821316613, "grad_norm": 3.081918954849243, "learning_rate": 9.926165640466664e-06, "loss": 1.2987, "step": 2890 }, { "epoch": 0.22656739811912227, "grad_norm": 3.4451968669891357, "learning_rate": 9.926112833891593e-06, "loss": 1.475, "step": 2891 }, { "epoch": 0.22664576802507838, "grad_norm": 3.1705126762390137, "learning_rate": 9.926060008580115e-06, "loss": 1.5835, "step": 2892 }, { "epoch": 0.2267241379310345, "grad_norm": 3.143958568572998, "learning_rate": 9.92600716453244e-06, "loss": 1.1621, "step": 2893 }, { "epoch": 0.2268025078369906, "grad_norm": 3.933919906616211, "learning_rate": 9.92595430174876e-06, "loss": 1.4264, "step": 2894 }, { "epoch": 0.2268808777429467, "grad_norm": 3.4503719806671143, "learning_rate": 9.925901420229283e-06, "loss": 1.6257, "step": 2895 }, { "epoch": 0.22695924764890282, "grad_norm": 3.3889546394348145, "learning_rate": 9.925848519974206e-06, "loss": 0.9694, "step": 2896 }, { "epoch": 0.22703761755485893, "grad_norm": 2.5469162464141846, "learning_rate": 9.92579560098373e-06, "loss": 1.2391, "step": 2897 }, { "epoch": 0.22711598746081504, "grad_norm": 5.447490215301514, "learning_rate": 9.92574266325806e-06, "loss": 1.9065, "step": 2898 }, { "epoch": 0.22719435736677115, "grad_norm": 7.315134048461914, "learning_rate": 9.925689706797394e-06, "loss": 1.7301, "step": 2899 }, { "epoch": 0.22727272727272727, "grad_norm": 3.6145846843719482, "learning_rate": 9.925636731601934e-06, "loss": 1.3087, "step": 2900 }, { "epoch": 0.22735109717868338, "grad_norm": 3.752547264099121, "learning_rate": 9.925583737671882e-06, "loss": 1.6277, "step": 2901 }, { "epoch": 0.2274294670846395, "grad_norm": 3.0148022174835205, "learning_rate": 9.925530725007438e-06, "loss": 1.6889, "step": 2902 }, { "epoch": 0.22750783699059562, "grad_norm": 2.1927542686462402, "learning_rate": 9.925477693608808e-06, "loss": 1.3988, "step": 2903 }, { "epoch": 0.22758620689655173, "grad_norm": 2.1497886180877686, "learning_rate": 9.925424643476189e-06, "loss": 1.4782, "step": 2904 }, { "epoch": 0.22766457680250785, "grad_norm": 8.103330612182617, "learning_rate": 9.925371574609783e-06, "loss": 1.3157, "step": 2905 }, { "epoch": 0.22774294670846396, "grad_norm": 3.13222336769104, "learning_rate": 9.925318487009795e-06, "loss": 1.4362, "step": 2906 }, { "epoch": 0.22782131661442007, "grad_norm": 2.3260040283203125, "learning_rate": 9.925265380676426e-06, "loss": 1.3559, "step": 2907 }, { "epoch": 0.22789968652037618, "grad_norm": 3.6463191509246826, "learning_rate": 9.925212255609875e-06, "loss": 1.7663, "step": 2908 }, { "epoch": 0.2279780564263323, "grad_norm": 2.9692723751068115, "learning_rate": 9.925159111810349e-06, "loss": 1.6422, "step": 2909 }, { "epoch": 0.2280564263322884, "grad_norm": 2.3746044635772705, "learning_rate": 9.925105949278047e-06, "loss": 1.292, "step": 2910 }, { "epoch": 0.2281347962382445, "grad_norm": 2.886396646499634, "learning_rate": 9.92505276801317e-06, "loss": 1.4353, "step": 2911 }, { "epoch": 0.22821316614420062, "grad_norm": 2.0217387676239014, "learning_rate": 9.924999568015923e-06, "loss": 1.3451, "step": 2912 }, { "epoch": 0.22829153605015673, "grad_norm": 3.4529056549072266, "learning_rate": 9.924946349286505e-06, "loss": 1.5957, "step": 2913 }, { "epoch": 0.22836990595611284, "grad_norm": 2.254528522491455, "learning_rate": 9.924893111825125e-06, "loss": 1.4022, "step": 2914 }, { "epoch": 0.22844827586206898, "grad_norm": 3.407294273376465, "learning_rate": 9.924839855631979e-06, "loss": 1.1132, "step": 2915 }, { "epoch": 0.2285266457680251, "grad_norm": 3.65615177154541, "learning_rate": 9.924786580707271e-06, "loss": 1.5624, "step": 2916 }, { "epoch": 0.2286050156739812, "grad_norm": 8.663552284240723, "learning_rate": 9.924733287051204e-06, "loss": 1.5241, "step": 2917 }, { "epoch": 0.2286833855799373, "grad_norm": 2.504277467727661, "learning_rate": 9.924679974663984e-06, "loss": 1.5836, "step": 2918 }, { "epoch": 0.22876175548589342, "grad_norm": 2.9955756664276123, "learning_rate": 9.92462664354581e-06, "loss": 1.1545, "step": 2919 }, { "epoch": 0.22884012539184953, "grad_norm": 3.4556376934051514, "learning_rate": 9.924573293696885e-06, "loss": 1.0159, "step": 2920 }, { "epoch": 0.22891849529780564, "grad_norm": 3.650765895843506, "learning_rate": 9.924519925117411e-06, "loss": 1.7881, "step": 2921 }, { "epoch": 0.22899686520376175, "grad_norm": 3.208102226257324, "learning_rate": 9.924466537807593e-06, "loss": 1.0768, "step": 2922 }, { "epoch": 0.22907523510971786, "grad_norm": 10.331052780151367, "learning_rate": 9.924413131767637e-06, "loss": 1.5866, "step": 2923 }, { "epoch": 0.22915360501567397, "grad_norm": 2.7858288288116455, "learning_rate": 9.924359706997739e-06, "loss": 1.4106, "step": 2924 }, { "epoch": 0.22923197492163008, "grad_norm": 3.453494071960449, "learning_rate": 9.924306263498107e-06, "loss": 1.5171, "step": 2925 }, { "epoch": 0.2293103448275862, "grad_norm": 7.2866435050964355, "learning_rate": 9.924252801268942e-06, "loss": 1.8261, "step": 2926 }, { "epoch": 0.22938871473354233, "grad_norm": 3.2823681831359863, "learning_rate": 9.924199320310452e-06, "loss": 1.546, "step": 2927 }, { "epoch": 0.22946708463949844, "grad_norm": 2.1665782928466797, "learning_rate": 9.924145820622835e-06, "loss": 1.2621, "step": 2928 }, { "epoch": 0.22954545454545455, "grad_norm": 3.622798442840576, "learning_rate": 9.924092302206297e-06, "loss": 1.179, "step": 2929 }, { "epoch": 0.22962382445141066, "grad_norm": 3.797879219055176, "learning_rate": 9.924038765061042e-06, "loss": 1.2376, "step": 2930 }, { "epoch": 0.22970219435736677, "grad_norm": 6.630140781402588, "learning_rate": 9.92398520918727e-06, "loss": 1.9348, "step": 2931 }, { "epoch": 0.22978056426332288, "grad_norm": 3.064176321029663, "learning_rate": 9.923931634585188e-06, "loss": 1.3338, "step": 2932 }, { "epoch": 0.229858934169279, "grad_norm": 1.8166773319244385, "learning_rate": 9.923878041255002e-06, "loss": 1.0499, "step": 2933 }, { "epoch": 0.2299373040752351, "grad_norm": 2.527076244354248, "learning_rate": 9.92382442919691e-06, "loss": 1.163, "step": 2934 }, { "epoch": 0.23001567398119122, "grad_norm": 2.5421340465545654, "learning_rate": 9.923770798411121e-06, "loss": 1.7086, "step": 2935 }, { "epoch": 0.23009404388714733, "grad_norm": 2.032867431640625, "learning_rate": 9.923717148897836e-06, "loss": 1.33, "step": 2936 }, { "epoch": 0.23017241379310344, "grad_norm": 3.90157151222229, "learning_rate": 9.92366348065726e-06, "loss": 1.7811, "step": 2937 }, { "epoch": 0.23025078369905957, "grad_norm": 3.5378615856170654, "learning_rate": 9.923609793689596e-06, "loss": 1.3026, "step": 2938 }, { "epoch": 0.23032915360501569, "grad_norm": 2.5377068519592285, "learning_rate": 9.92355608799505e-06, "loss": 1.5431, "step": 2939 }, { "epoch": 0.2304075235109718, "grad_norm": 3.0236520767211914, "learning_rate": 9.923502363573825e-06, "loss": 1.4515, "step": 2940 }, { "epoch": 0.2304858934169279, "grad_norm": 2.968407154083252, "learning_rate": 9.923448620426127e-06, "loss": 1.7719, "step": 2941 }, { "epoch": 0.23056426332288402, "grad_norm": 2.9483461380004883, "learning_rate": 9.923394858552159e-06, "loss": 1.3946, "step": 2942 }, { "epoch": 0.23064263322884013, "grad_norm": 3.2130610942840576, "learning_rate": 9.923341077952126e-06, "loss": 1.46, "step": 2943 }, { "epoch": 0.23072100313479624, "grad_norm": 2.618030309677124, "learning_rate": 9.923287278626233e-06, "loss": 1.3028, "step": 2944 }, { "epoch": 0.23079937304075235, "grad_norm": 4.03307580947876, "learning_rate": 9.923233460574682e-06, "loss": 1.5216, "step": 2945 }, { "epoch": 0.23087774294670846, "grad_norm": 3.124863386154175, "learning_rate": 9.923179623797681e-06, "loss": 1.6018, "step": 2946 }, { "epoch": 0.23095611285266457, "grad_norm": 2.666085720062256, "learning_rate": 9.923125768295433e-06, "loss": 1.267, "step": 2947 }, { "epoch": 0.23103448275862068, "grad_norm": 6.317298412322998, "learning_rate": 9.923071894068142e-06, "loss": 1.1662, "step": 2948 }, { "epoch": 0.2311128526645768, "grad_norm": 3.5064353942871094, "learning_rate": 9.923018001116016e-06, "loss": 1.4074, "step": 2949 }, { "epoch": 0.23119122257053293, "grad_norm": 4.063719749450684, "learning_rate": 9.922964089439257e-06, "loss": 1.8511, "step": 2950 }, { "epoch": 0.23126959247648904, "grad_norm": 4.429939270019531, "learning_rate": 9.922910159038073e-06, "loss": 1.7588, "step": 2951 }, { "epoch": 0.23134796238244515, "grad_norm": 3.66051983833313, "learning_rate": 9.922856209912666e-06, "loss": 1.1838, "step": 2952 }, { "epoch": 0.23142633228840126, "grad_norm": 3.9928956031799316, "learning_rate": 9.92280224206324e-06, "loss": 1.4567, "step": 2953 }, { "epoch": 0.23150470219435737, "grad_norm": 3.4716265201568604, "learning_rate": 9.922748255490006e-06, "loss": 1.3943, "step": 2954 }, { "epoch": 0.23158307210031348, "grad_norm": 2.4198126792907715, "learning_rate": 9.922694250193166e-06, "loss": 1.2713, "step": 2955 }, { "epoch": 0.2316614420062696, "grad_norm": 2.22967267036438, "learning_rate": 9.922640226172927e-06, "loss": 1.4163, "step": 2956 }, { "epoch": 0.2317398119122257, "grad_norm": 5.5635576248168945, "learning_rate": 9.92258618342949e-06, "loss": 1.882, "step": 2957 }, { "epoch": 0.2318181818181818, "grad_norm": 3.8599436283111572, "learning_rate": 9.922532121963064e-06, "loss": 1.8944, "step": 2958 }, { "epoch": 0.23189655172413792, "grad_norm": 3.815181255340576, "learning_rate": 9.922478041773856e-06, "loss": 1.6228, "step": 2959 }, { "epoch": 0.23197492163009403, "grad_norm": 7.187936305999756, "learning_rate": 9.922423942862067e-06, "loss": 1.2601, "step": 2960 }, { "epoch": 0.23205329153605014, "grad_norm": 2.4882915019989014, "learning_rate": 9.922369825227907e-06, "loss": 1.4155, "step": 2961 }, { "epoch": 0.23213166144200628, "grad_norm": 5.8725433349609375, "learning_rate": 9.922315688871582e-06, "loss": 1.4597, "step": 2962 }, { "epoch": 0.2322100313479624, "grad_norm": 3.1507174968719482, "learning_rate": 9.922261533793294e-06, "loss": 1.2744, "step": 2963 }, { "epoch": 0.2322884012539185, "grad_norm": 2.978886127471924, "learning_rate": 9.922207359993252e-06, "loss": 1.0779, "step": 2964 }, { "epoch": 0.2323667711598746, "grad_norm": 3.5459952354431152, "learning_rate": 9.922153167471662e-06, "loss": 1.4916, "step": 2965 }, { "epoch": 0.23244514106583072, "grad_norm": 2.8122520446777344, "learning_rate": 9.922098956228729e-06, "loss": 1.4069, "step": 2966 }, { "epoch": 0.23252351097178683, "grad_norm": 3.0333917140960693, "learning_rate": 9.922044726264661e-06, "loss": 1.1183, "step": 2967 }, { "epoch": 0.23260188087774294, "grad_norm": 2.7237775325775146, "learning_rate": 9.921990477579661e-06, "loss": 1.3139, "step": 2968 }, { "epoch": 0.23268025078369906, "grad_norm": 2.672102689743042, "learning_rate": 9.921936210173939e-06, "loss": 1.3859, "step": 2969 }, { "epoch": 0.23275862068965517, "grad_norm": 3.494407892227173, "learning_rate": 9.9218819240477e-06, "loss": 1.0973, "step": 2970 }, { "epoch": 0.23283699059561128, "grad_norm": 2.137795925140381, "learning_rate": 9.92182761920115e-06, "loss": 1.375, "step": 2971 }, { "epoch": 0.2329153605015674, "grad_norm": 2.567660093307495, "learning_rate": 9.921773295634494e-06, "loss": 1.5571, "step": 2972 }, { "epoch": 0.2329937304075235, "grad_norm": 7.082502365112305, "learning_rate": 9.921718953347941e-06, "loss": 1.2306, "step": 2973 }, { "epoch": 0.23307210031347964, "grad_norm": 2.7491402626037598, "learning_rate": 9.921664592341697e-06, "loss": 1.0424, "step": 2974 }, { "epoch": 0.23315047021943575, "grad_norm": 2.1592767238616943, "learning_rate": 9.92161021261597e-06, "loss": 0.9814, "step": 2975 }, { "epoch": 0.23322884012539186, "grad_norm": 2.1756949424743652, "learning_rate": 9.921555814170966e-06, "loss": 1.7323, "step": 2976 }, { "epoch": 0.23330721003134797, "grad_norm": 4.4373321533203125, "learning_rate": 9.92150139700689e-06, "loss": 1.2085, "step": 2977 }, { "epoch": 0.23338557993730408, "grad_norm": 2.1436357498168945, "learning_rate": 9.92144696112395e-06, "loss": 1.5314, "step": 2978 }, { "epoch": 0.2334639498432602, "grad_norm": 2.758821964263916, "learning_rate": 9.921392506522356e-06, "loss": 1.5592, "step": 2979 }, { "epoch": 0.2335423197492163, "grad_norm": 3.4573001861572266, "learning_rate": 9.92133803320231e-06, "loss": 1.4036, "step": 2980 }, { "epoch": 0.2336206896551724, "grad_norm": 3.775582790374756, "learning_rate": 9.921283541164023e-06, "loss": 1.5927, "step": 2981 }, { "epoch": 0.23369905956112852, "grad_norm": 4.11469030380249, "learning_rate": 9.921229030407702e-06, "loss": 1.4726, "step": 2982 }, { "epoch": 0.23377742946708463, "grad_norm": 2.9195175170898438, "learning_rate": 9.921174500933552e-06, "loss": 1.1545, "step": 2983 }, { "epoch": 0.23385579937304074, "grad_norm": 2.637301206588745, "learning_rate": 9.921119952741784e-06, "loss": 1.3155, "step": 2984 }, { "epoch": 0.23393416927899688, "grad_norm": 5.896999359130859, "learning_rate": 9.921065385832603e-06, "loss": 0.9581, "step": 2985 }, { "epoch": 0.234012539184953, "grad_norm": 3.6373021602630615, "learning_rate": 9.921010800206214e-06, "loss": 1.2998, "step": 2986 }, { "epoch": 0.2340909090909091, "grad_norm": 2.960383653640747, "learning_rate": 9.92095619586283e-06, "loss": 1.103, "step": 2987 }, { "epoch": 0.2341692789968652, "grad_norm": 7.585729122161865, "learning_rate": 9.920901572802656e-06, "loss": 0.9684, "step": 2988 }, { "epoch": 0.23424764890282132, "grad_norm": 2.938368082046509, "learning_rate": 9.920846931025899e-06, "loss": 1.7398, "step": 2989 }, { "epoch": 0.23432601880877743, "grad_norm": 1.8900566101074219, "learning_rate": 9.920792270532769e-06, "loss": 1.4594, "step": 2990 }, { "epoch": 0.23440438871473354, "grad_norm": 4.135896682739258, "learning_rate": 9.920737591323471e-06, "loss": 1.576, "step": 2991 }, { "epoch": 0.23448275862068965, "grad_norm": 3.324753761291504, "learning_rate": 9.920682893398217e-06, "loss": 1.044, "step": 2992 }, { "epoch": 0.23456112852664576, "grad_norm": 3.5332069396972656, "learning_rate": 9.92062817675721e-06, "loss": 1.3434, "step": 2993 }, { "epoch": 0.23463949843260187, "grad_norm": 2.86980938911438, "learning_rate": 9.920573441400662e-06, "loss": 0.8171, "step": 2994 }, { "epoch": 0.23471786833855798, "grad_norm": 2.8421738147735596, "learning_rate": 9.920518687328782e-06, "loss": 1.6276, "step": 2995 }, { "epoch": 0.2347962382445141, "grad_norm": 5.800183296203613, "learning_rate": 9.920463914541773e-06, "loss": 1.4158, "step": 2996 }, { "epoch": 0.23487460815047023, "grad_norm": 6.681927680969238, "learning_rate": 9.920409123039849e-06, "loss": 1.183, "step": 2997 }, { "epoch": 0.23495297805642634, "grad_norm": 2.6500942707061768, "learning_rate": 9.920354312823215e-06, "loss": 1.4915, "step": 2998 }, { "epoch": 0.23503134796238245, "grad_norm": 2.6855218410491943, "learning_rate": 9.92029948389208e-06, "loss": 1.4542, "step": 2999 }, { "epoch": 0.23510971786833856, "grad_norm": 2.50669002532959, "learning_rate": 9.920244636246654e-06, "loss": 1.4572, "step": 3000 }, { "epoch": 0.23518808777429467, "grad_norm": 5.921534061431885, "learning_rate": 9.920189769887143e-06, "loss": 1.4675, "step": 3001 }, { "epoch": 0.23526645768025078, "grad_norm": 4.654906272888184, "learning_rate": 9.920134884813759e-06, "loss": 1.0853, "step": 3002 }, { "epoch": 0.2353448275862069, "grad_norm": 3.1793553829193115, "learning_rate": 9.92007998102671e-06, "loss": 1.335, "step": 3003 }, { "epoch": 0.235423197492163, "grad_norm": 3.2996232509613037, "learning_rate": 9.9200250585262e-06, "loss": 1.4942, "step": 3004 }, { "epoch": 0.23550156739811912, "grad_norm": 2.835574150085449, "learning_rate": 9.919970117312445e-06, "loss": 1.6715, "step": 3005 }, { "epoch": 0.23557993730407523, "grad_norm": 2.866112232208252, "learning_rate": 9.91991515738565e-06, "loss": 1.6908, "step": 3006 }, { "epoch": 0.23565830721003134, "grad_norm": 4.159789562225342, "learning_rate": 9.919860178746024e-06, "loss": 1.3563, "step": 3007 }, { "epoch": 0.23573667711598745, "grad_norm": 2.2851216793060303, "learning_rate": 9.919805181393777e-06, "loss": 1.5525, "step": 3008 }, { "epoch": 0.23581504702194359, "grad_norm": 1.6767637729644775, "learning_rate": 9.919750165329118e-06, "loss": 1.1776, "step": 3009 }, { "epoch": 0.2358934169278997, "grad_norm": 2.4991559982299805, "learning_rate": 9.919695130552257e-06, "loss": 1.197, "step": 3010 }, { "epoch": 0.2359717868338558, "grad_norm": 2.4043374061584473, "learning_rate": 9.919640077063403e-06, "loss": 1.4285, "step": 3011 }, { "epoch": 0.23605015673981192, "grad_norm": 1.972780466079712, "learning_rate": 9.919585004862763e-06, "loss": 1.157, "step": 3012 }, { "epoch": 0.23612852664576803, "grad_norm": 2.9746007919311523, "learning_rate": 9.91952991395055e-06, "loss": 1.6073, "step": 3013 }, { "epoch": 0.23620689655172414, "grad_norm": 2.598954677581787, "learning_rate": 9.919474804326973e-06, "loss": 1.182, "step": 3014 }, { "epoch": 0.23628526645768025, "grad_norm": 8.90810775756836, "learning_rate": 9.919419675992239e-06, "loss": 1.0267, "step": 3015 }, { "epoch": 0.23636363636363636, "grad_norm": 2.089029550552368, "learning_rate": 9.91936452894656e-06, "loss": 1.365, "step": 3016 }, { "epoch": 0.23644200626959247, "grad_norm": 5.349265098571777, "learning_rate": 9.919309363190146e-06, "loss": 1.074, "step": 3017 }, { "epoch": 0.23652037617554858, "grad_norm": 2.2235379219055176, "learning_rate": 9.919254178723204e-06, "loss": 1.6776, "step": 3018 }, { "epoch": 0.2365987460815047, "grad_norm": 2.809387445449829, "learning_rate": 9.919198975545947e-06, "loss": 1.5618, "step": 3019 }, { "epoch": 0.23667711598746083, "grad_norm": 3.0445282459259033, "learning_rate": 9.919143753658584e-06, "loss": 1.4775, "step": 3020 }, { "epoch": 0.23675548589341694, "grad_norm": 2.524568557739258, "learning_rate": 9.919088513061324e-06, "loss": 1.4026, "step": 3021 }, { "epoch": 0.23683385579937305, "grad_norm": 2.9064717292785645, "learning_rate": 9.919033253754377e-06, "loss": 1.4574, "step": 3022 }, { "epoch": 0.23691222570532916, "grad_norm": 2.320200204849243, "learning_rate": 9.918977975737955e-06, "loss": 1.3371, "step": 3023 }, { "epoch": 0.23699059561128527, "grad_norm": 3.3360443115234375, "learning_rate": 9.918922679012269e-06, "loss": 1.2307, "step": 3024 }, { "epoch": 0.23706896551724138, "grad_norm": 3.1607179641723633, "learning_rate": 9.918867363577526e-06, "loss": 1.4168, "step": 3025 }, { "epoch": 0.2371473354231975, "grad_norm": 2.5068585872650146, "learning_rate": 9.918812029433938e-06, "loss": 1.4754, "step": 3026 }, { "epoch": 0.2372257053291536, "grad_norm": 2.582791328430176, "learning_rate": 9.918756676581716e-06, "loss": 1.0731, "step": 3027 }, { "epoch": 0.2373040752351097, "grad_norm": 3.0005862712860107, "learning_rate": 9.91870130502107e-06, "loss": 1.3933, "step": 3028 }, { "epoch": 0.23738244514106582, "grad_norm": 2.9741694927215576, "learning_rate": 9.91864591475221e-06, "loss": 1.3584, "step": 3029 }, { "epoch": 0.23746081504702193, "grad_norm": 1.744678020477295, "learning_rate": 9.918590505775348e-06, "loss": 1.4287, "step": 3030 }, { "epoch": 0.23753918495297804, "grad_norm": 5.827905178070068, "learning_rate": 9.918535078090694e-06, "loss": 1.509, "step": 3031 }, { "epoch": 0.23761755485893418, "grad_norm": 2.4051170349121094, "learning_rate": 9.918479631698459e-06, "loss": 1.4338, "step": 3032 }, { "epoch": 0.2376959247648903, "grad_norm": 7.828668594360352, "learning_rate": 9.918424166598852e-06, "loss": 1.2757, "step": 3033 }, { "epoch": 0.2377742946708464, "grad_norm": 2.8562707901000977, "learning_rate": 9.918368682792087e-06, "loss": 1.1173, "step": 3034 }, { "epoch": 0.2378526645768025, "grad_norm": 3.456869125366211, "learning_rate": 9.918313180278374e-06, "loss": 1.6754, "step": 3035 }, { "epoch": 0.23793103448275862, "grad_norm": 3.2349355220794678, "learning_rate": 9.918257659057925e-06, "loss": 1.7989, "step": 3036 }, { "epoch": 0.23800940438871473, "grad_norm": 2.6500682830810547, "learning_rate": 9.918202119130949e-06, "loss": 1.5831, "step": 3037 }, { "epoch": 0.23808777429467085, "grad_norm": 2.1677980422973633, "learning_rate": 9.918146560497658e-06, "loss": 1.232, "step": 3038 }, { "epoch": 0.23816614420062696, "grad_norm": 2.9511799812316895, "learning_rate": 9.918090983158264e-06, "loss": 1.3562, "step": 3039 }, { "epoch": 0.23824451410658307, "grad_norm": 4.930164813995361, "learning_rate": 9.918035387112976e-06, "loss": 1.5211, "step": 3040 }, { "epoch": 0.23832288401253918, "grad_norm": 3.775038957595825, "learning_rate": 9.91797977236201e-06, "loss": 1.4707, "step": 3041 }, { "epoch": 0.2384012539184953, "grad_norm": 2.390789747238159, "learning_rate": 9.917924138905574e-06, "loss": 1.174, "step": 3042 }, { "epoch": 0.2384796238244514, "grad_norm": 2.6327619552612305, "learning_rate": 9.917868486743881e-06, "loss": 1.0438, "step": 3043 }, { "epoch": 0.23855799373040754, "grad_norm": 2.100653886795044, "learning_rate": 9.917812815877142e-06, "loss": 1.4772, "step": 3044 }, { "epoch": 0.23863636363636365, "grad_norm": 2.917130708694458, "learning_rate": 9.91775712630557e-06, "loss": 1.2626, "step": 3045 }, { "epoch": 0.23871473354231976, "grad_norm": 3.6307830810546875, "learning_rate": 9.917701418029373e-06, "loss": 1.5369, "step": 3046 }, { "epoch": 0.23879310344827587, "grad_norm": 2.473808526992798, "learning_rate": 9.917645691048768e-06, "loss": 1.3028, "step": 3047 }, { "epoch": 0.23887147335423198, "grad_norm": 5.871737480163574, "learning_rate": 9.917589945363965e-06, "loss": 1.4249, "step": 3048 }, { "epoch": 0.2389498432601881, "grad_norm": 3.6156837940216064, "learning_rate": 9.917534180975175e-06, "loss": 1.3132, "step": 3049 }, { "epoch": 0.2390282131661442, "grad_norm": 2.318553924560547, "learning_rate": 9.91747839788261e-06, "loss": 1.0747, "step": 3050 }, { "epoch": 0.2391065830721003, "grad_norm": 4.704532623291016, "learning_rate": 9.917422596086483e-06, "loss": 1.164, "step": 3051 }, { "epoch": 0.23918495297805642, "grad_norm": 3.3886759281158447, "learning_rate": 9.917366775587007e-06, "loss": 1.2959, "step": 3052 }, { "epoch": 0.23926332288401253, "grad_norm": 2.3362927436828613, "learning_rate": 9.917310936384394e-06, "loss": 1.1621, "step": 3053 }, { "epoch": 0.23934169278996864, "grad_norm": 2.8650505542755127, "learning_rate": 9.917255078478855e-06, "loss": 1.3464, "step": 3054 }, { "epoch": 0.23942006269592475, "grad_norm": 2.7092208862304688, "learning_rate": 9.917199201870604e-06, "loss": 1.3009, "step": 3055 }, { "epoch": 0.2394984326018809, "grad_norm": 7.854827404022217, "learning_rate": 9.917143306559851e-06, "loss": 1.8183, "step": 3056 }, { "epoch": 0.239576802507837, "grad_norm": 2.7667157649993896, "learning_rate": 9.917087392546813e-06, "loss": 1.6269, "step": 3057 }, { "epoch": 0.2396551724137931, "grad_norm": 3.215191125869751, "learning_rate": 9.917031459831698e-06, "loss": 1.8412, "step": 3058 }, { "epoch": 0.23973354231974922, "grad_norm": 3.021509885787964, "learning_rate": 9.916975508414723e-06, "loss": 1.2677, "step": 3059 }, { "epoch": 0.23981191222570533, "grad_norm": 4.949398517608643, "learning_rate": 9.916919538296098e-06, "loss": 1.5571, "step": 3060 }, { "epoch": 0.23989028213166144, "grad_norm": 2.4402639865875244, "learning_rate": 9.916863549476035e-06, "loss": 1.372, "step": 3061 }, { "epoch": 0.23996865203761755, "grad_norm": 2.2883472442626953, "learning_rate": 9.91680754195475e-06, "loss": 1.503, "step": 3062 }, { "epoch": 0.24004702194357366, "grad_norm": 2.4196977615356445, "learning_rate": 9.916751515732455e-06, "loss": 1.1988, "step": 3063 }, { "epoch": 0.24012539184952977, "grad_norm": 5.373312950134277, "learning_rate": 9.916695470809361e-06, "loss": 1.7047, "step": 3064 }, { "epoch": 0.24020376175548588, "grad_norm": 2.7301197052001953, "learning_rate": 9.916639407185683e-06, "loss": 1.4504, "step": 3065 }, { "epoch": 0.240282131661442, "grad_norm": 3.0915017127990723, "learning_rate": 9.916583324861634e-06, "loss": 1.3423, "step": 3066 }, { "epoch": 0.24036050156739813, "grad_norm": 7.64423942565918, "learning_rate": 9.91652722383743e-06, "loss": 1.6579, "step": 3067 }, { "epoch": 0.24043887147335424, "grad_norm": 3.3582136631011963, "learning_rate": 9.916471104113278e-06, "loss": 1.6052, "step": 3068 }, { "epoch": 0.24051724137931035, "grad_norm": 2.8975212574005127, "learning_rate": 9.916414965689399e-06, "loss": 1.3679, "step": 3069 }, { "epoch": 0.24059561128526646, "grad_norm": 3.2936034202575684, "learning_rate": 9.916358808565999e-06, "loss": 1.2237, "step": 3070 }, { "epoch": 0.24067398119122257, "grad_norm": 6.7625203132629395, "learning_rate": 9.916302632743297e-06, "loss": 1.1248, "step": 3071 }, { "epoch": 0.24075235109717869, "grad_norm": 2.955226182937622, "learning_rate": 9.916246438221503e-06, "loss": 1.5252, "step": 3072 }, { "epoch": 0.2408307210031348, "grad_norm": 2.0707571506500244, "learning_rate": 9.916190225000837e-06, "loss": 1.2615, "step": 3073 }, { "epoch": 0.2409090909090909, "grad_norm": 3.3959760665893555, "learning_rate": 9.916133993081506e-06, "loss": 0.8222, "step": 3074 }, { "epoch": 0.24098746081504702, "grad_norm": 4.990645885467529, "learning_rate": 9.916077742463725e-06, "loss": 1.1427, "step": 3075 }, { "epoch": 0.24106583072100313, "grad_norm": 3.5622644424438477, "learning_rate": 9.916021473147712e-06, "loss": 1.8385, "step": 3076 }, { "epoch": 0.24114420062695924, "grad_norm": 5.993614673614502, "learning_rate": 9.915965185133676e-06, "loss": 1.3742, "step": 3077 }, { "epoch": 0.24122257053291535, "grad_norm": 4.081839084625244, "learning_rate": 9.915908878421835e-06, "loss": 1.4796, "step": 3078 }, { "epoch": 0.2413009404388715, "grad_norm": 2.5649476051330566, "learning_rate": 9.915852553012402e-06, "loss": 1.3383, "step": 3079 }, { "epoch": 0.2413793103448276, "grad_norm": 2.591797351837158, "learning_rate": 9.915796208905589e-06, "loss": 1.7048, "step": 3080 }, { "epoch": 0.2414576802507837, "grad_norm": 2.6211178302764893, "learning_rate": 9.915739846101613e-06, "loss": 1.1404, "step": 3081 }, { "epoch": 0.24153605015673982, "grad_norm": 2.993959426879883, "learning_rate": 9.915683464600688e-06, "loss": 1.312, "step": 3082 }, { "epoch": 0.24161442006269593, "grad_norm": 3.2042698860168457, "learning_rate": 9.915627064403028e-06, "loss": 1.251, "step": 3083 }, { "epoch": 0.24169278996865204, "grad_norm": 2.4581379890441895, "learning_rate": 9.915570645508847e-06, "loss": 1.6067, "step": 3084 }, { "epoch": 0.24177115987460815, "grad_norm": 2.1414592266082764, "learning_rate": 9.91551420791836e-06, "loss": 1.2645, "step": 3085 }, { "epoch": 0.24184952978056426, "grad_norm": 2.849107265472412, "learning_rate": 9.915457751631781e-06, "loss": 1.3687, "step": 3086 }, { "epoch": 0.24192789968652037, "grad_norm": 4.189999103546143, "learning_rate": 9.915401276649325e-06, "loss": 1.368, "step": 3087 }, { "epoch": 0.24200626959247648, "grad_norm": 4.556332111358643, "learning_rate": 9.91534478297121e-06, "loss": 1.512, "step": 3088 }, { "epoch": 0.2420846394984326, "grad_norm": 3.2719545364379883, "learning_rate": 9.915288270597647e-06, "loss": 1.1579, "step": 3089 }, { "epoch": 0.2421630094043887, "grad_norm": 3.05277943611145, "learning_rate": 9.915231739528851e-06, "loss": 1.4574, "step": 3090 }, { "epoch": 0.24224137931034484, "grad_norm": 7.050443172454834, "learning_rate": 9.915175189765038e-06, "loss": 1.9511, "step": 3091 }, { "epoch": 0.24231974921630095, "grad_norm": 9.208964347839355, "learning_rate": 9.915118621306426e-06, "loss": 1.0577, "step": 3092 }, { "epoch": 0.24239811912225706, "grad_norm": 5.446291923522949, "learning_rate": 9.915062034153224e-06, "loss": 0.9373, "step": 3093 }, { "epoch": 0.24247648902821317, "grad_norm": 3.8268609046936035, "learning_rate": 9.915005428305653e-06, "loss": 1.7148, "step": 3094 }, { "epoch": 0.24255485893416928, "grad_norm": 3.116547107696533, "learning_rate": 9.914948803763924e-06, "loss": 1.4202, "step": 3095 }, { "epoch": 0.2426332288401254, "grad_norm": 5.458043098449707, "learning_rate": 9.914892160528255e-06, "loss": 1.8606, "step": 3096 }, { "epoch": 0.2427115987460815, "grad_norm": 4.412992477416992, "learning_rate": 9.914835498598861e-06, "loss": 1.5061, "step": 3097 }, { "epoch": 0.2427899686520376, "grad_norm": 3.974112033843994, "learning_rate": 9.914778817975956e-06, "loss": 1.5123, "step": 3098 }, { "epoch": 0.24286833855799372, "grad_norm": 3.58120059967041, "learning_rate": 9.914722118659758e-06, "loss": 2.4765, "step": 3099 }, { "epoch": 0.24294670846394983, "grad_norm": 3.4518775939941406, "learning_rate": 9.91466540065048e-06, "loss": 1.0502, "step": 3100 }, { "epoch": 0.24302507836990594, "grad_norm": 6.243551254272461, "learning_rate": 9.914608663948341e-06, "loss": 0.3814, "step": 3101 }, { "epoch": 0.24310344827586206, "grad_norm": 3.4042506217956543, "learning_rate": 9.914551908553554e-06, "loss": 1.5083, "step": 3102 }, { "epoch": 0.2431818181818182, "grad_norm": 3.653801441192627, "learning_rate": 9.914495134466336e-06, "loss": 1.6997, "step": 3103 }, { "epoch": 0.2432601880877743, "grad_norm": 3.315600633621216, "learning_rate": 9.914438341686903e-06, "loss": 1.2646, "step": 3104 }, { "epoch": 0.24333855799373041, "grad_norm": 7.811936378479004, "learning_rate": 9.91438153021547e-06, "loss": 1.8675, "step": 3105 }, { "epoch": 0.24341692789968652, "grad_norm": 3.1923656463623047, "learning_rate": 9.914324700052253e-06, "loss": 1.5322, "step": 3106 }, { "epoch": 0.24349529780564264, "grad_norm": 2.5519163608551025, "learning_rate": 9.914267851197472e-06, "loss": 1.5715, "step": 3107 }, { "epoch": 0.24357366771159875, "grad_norm": 2.1613364219665527, "learning_rate": 9.914210983651338e-06, "loss": 1.299, "step": 3108 }, { "epoch": 0.24365203761755486, "grad_norm": 3.5305705070495605, "learning_rate": 9.914154097414071e-06, "loss": 1.2053, "step": 3109 }, { "epoch": 0.24373040752351097, "grad_norm": 11.784263610839844, "learning_rate": 9.914097192485883e-06, "loss": 1.2467, "step": 3110 }, { "epoch": 0.24380877742946708, "grad_norm": 2.0551798343658447, "learning_rate": 9.914040268866995e-06, "loss": 1.5855, "step": 3111 }, { "epoch": 0.2438871473354232, "grad_norm": 3.086782217025757, "learning_rate": 9.913983326557622e-06, "loss": 1.5201, "step": 3112 }, { "epoch": 0.2439655172413793, "grad_norm": 5.23169469833374, "learning_rate": 9.913926365557981e-06, "loss": 1.4936, "step": 3113 }, { "epoch": 0.24404388714733544, "grad_norm": 4.132696151733398, "learning_rate": 9.913869385868287e-06, "loss": 1.2205, "step": 3114 }, { "epoch": 0.24412225705329155, "grad_norm": 2.447291374206543, "learning_rate": 9.91381238748876e-06, "loss": 1.4646, "step": 3115 }, { "epoch": 0.24420062695924766, "grad_norm": 4.792943000793457, "learning_rate": 9.91375537041961e-06, "loss": 1.1879, "step": 3116 }, { "epoch": 0.24427899686520377, "grad_norm": 3.879836082458496, "learning_rate": 9.913698334661063e-06, "loss": 1.3999, "step": 3117 }, { "epoch": 0.24435736677115988, "grad_norm": 2.601487159729004, "learning_rate": 9.913641280213329e-06, "loss": 1.4466, "step": 3118 }, { "epoch": 0.244435736677116, "grad_norm": 3.6761553287506104, "learning_rate": 9.913584207076629e-06, "loss": 1.6432, "step": 3119 }, { "epoch": 0.2445141065830721, "grad_norm": 6.5079450607299805, "learning_rate": 9.913527115251177e-06, "loss": 1.4634, "step": 3120 }, { "epoch": 0.2445924764890282, "grad_norm": 2.1973419189453125, "learning_rate": 9.913470004737191e-06, "loss": 1.4306, "step": 3121 }, { "epoch": 0.24467084639498432, "grad_norm": 3.456726551055908, "learning_rate": 9.913412875534891e-06, "loss": 1.4849, "step": 3122 }, { "epoch": 0.24474921630094043, "grad_norm": 2.540046453475952, "learning_rate": 9.913355727644491e-06, "loss": 1.5037, "step": 3123 }, { "epoch": 0.24482758620689654, "grad_norm": 3.1310462951660156, "learning_rate": 9.91329856106621e-06, "loss": 1.3187, "step": 3124 }, { "epoch": 0.24490595611285265, "grad_norm": 4.579287528991699, "learning_rate": 9.913241375800262e-06, "loss": 1.3226, "step": 3125 }, { "epoch": 0.2449843260188088, "grad_norm": 2.4293556213378906, "learning_rate": 9.91318417184687e-06, "loss": 1.1992, "step": 3126 }, { "epoch": 0.2450626959247649, "grad_norm": 3.308403968811035, "learning_rate": 9.91312694920625e-06, "loss": 1.6709, "step": 3127 }, { "epoch": 0.245141065830721, "grad_norm": 4.241774559020996, "learning_rate": 9.913069707878617e-06, "loss": 1.3186, "step": 3128 }, { "epoch": 0.24521943573667712, "grad_norm": 13.958154678344727, "learning_rate": 9.913012447864191e-06, "loss": 0.7294, "step": 3129 }, { "epoch": 0.24529780564263323, "grad_norm": 3.711662769317627, "learning_rate": 9.91295516916319e-06, "loss": 1.5255, "step": 3130 }, { "epoch": 0.24537617554858934, "grad_norm": 2.1901540756225586, "learning_rate": 9.912897871775828e-06, "loss": 1.2368, "step": 3131 }, { "epoch": 0.24545454545454545, "grad_norm": 2.548283338546753, "learning_rate": 9.912840555702328e-06, "loss": 1.3975, "step": 3132 }, { "epoch": 0.24553291536050156, "grad_norm": 2.4132726192474365, "learning_rate": 9.912783220942906e-06, "loss": 1.5761, "step": 3133 }, { "epoch": 0.24561128526645767, "grad_norm": 2.079958438873291, "learning_rate": 9.91272586749778e-06, "loss": 1.4447, "step": 3134 }, { "epoch": 0.24568965517241378, "grad_norm": 1.9864671230316162, "learning_rate": 9.912668495367168e-06, "loss": 1.4372, "step": 3135 }, { "epoch": 0.2457680250783699, "grad_norm": 12.554539680480957, "learning_rate": 9.912611104551288e-06, "loss": 1.6176, "step": 3136 }, { "epoch": 0.245846394984326, "grad_norm": 2.760948896408081, "learning_rate": 9.91255369505036e-06, "loss": 1.6175, "step": 3137 }, { "epoch": 0.24592476489028214, "grad_norm": 3.449212074279785, "learning_rate": 9.912496266864599e-06, "loss": 1.8458, "step": 3138 }, { "epoch": 0.24600313479623825, "grad_norm": 2.065640926361084, "learning_rate": 9.912438819994225e-06, "loss": 1.2816, "step": 3139 }, { "epoch": 0.24608150470219436, "grad_norm": 3.065656900405884, "learning_rate": 9.91238135443946e-06, "loss": 1.5012, "step": 3140 }, { "epoch": 0.24615987460815048, "grad_norm": 6.469261169433594, "learning_rate": 9.912323870200516e-06, "loss": 2.0385, "step": 3141 }, { "epoch": 0.24623824451410659, "grad_norm": 2.9355359077453613, "learning_rate": 9.912266367277616e-06, "loss": 1.2213, "step": 3142 }, { "epoch": 0.2463166144200627, "grad_norm": 3.2783515453338623, "learning_rate": 9.91220884567098e-06, "loss": 1.6616, "step": 3143 }, { "epoch": 0.2463949843260188, "grad_norm": 3.5625202655792236, "learning_rate": 9.912151305380823e-06, "loss": 1.1037, "step": 3144 }, { "epoch": 0.24647335423197492, "grad_norm": 14.892630577087402, "learning_rate": 9.912093746407365e-06, "loss": 1.1155, "step": 3145 }, { "epoch": 0.24655172413793103, "grad_norm": 2.565621852874756, "learning_rate": 9.912036168750828e-06, "loss": 1.3801, "step": 3146 }, { "epoch": 0.24663009404388714, "grad_norm": 4.008517742156982, "learning_rate": 9.911978572411424e-06, "loss": 1.1414, "step": 3147 }, { "epoch": 0.24670846394984325, "grad_norm": 2.153994560241699, "learning_rate": 9.91192095738938e-06, "loss": 1.5721, "step": 3148 }, { "epoch": 0.2467868338557994, "grad_norm": 3.373049020767212, "learning_rate": 9.91186332368491e-06, "loss": 1.0659, "step": 3149 }, { "epoch": 0.2468652037617555, "grad_norm": 6.32871150970459, "learning_rate": 9.911805671298237e-06, "loss": 1.3182, "step": 3150 }, { "epoch": 0.2469435736677116, "grad_norm": 2.2644362449645996, "learning_rate": 9.911748000229575e-06, "loss": 1.3989, "step": 3151 }, { "epoch": 0.24702194357366772, "grad_norm": 3.6567225456237793, "learning_rate": 9.911690310479147e-06, "loss": 0.8065, "step": 3152 }, { "epoch": 0.24710031347962383, "grad_norm": 2.547914981842041, "learning_rate": 9.911632602047174e-06, "loss": 1.7678, "step": 3153 }, { "epoch": 0.24717868338557994, "grad_norm": 2.6417229175567627, "learning_rate": 9.911574874933871e-06, "loss": 1.7381, "step": 3154 }, { "epoch": 0.24725705329153605, "grad_norm": 4.015328407287598, "learning_rate": 9.911517129139461e-06, "loss": 1.361, "step": 3155 }, { "epoch": 0.24733542319749216, "grad_norm": 4.39141321182251, "learning_rate": 9.911459364664163e-06, "loss": 1.4624, "step": 3156 }, { "epoch": 0.24741379310344827, "grad_norm": 6.6511006355285645, "learning_rate": 9.911401581508195e-06, "loss": 1.2914, "step": 3157 }, { "epoch": 0.24749216300940438, "grad_norm": 2.3989455699920654, "learning_rate": 9.91134377967178e-06, "loss": 1.1754, "step": 3158 }, { "epoch": 0.2475705329153605, "grad_norm": 8.57794189453125, "learning_rate": 9.911285959155134e-06, "loss": 1.5662, "step": 3159 }, { "epoch": 0.2476489028213166, "grad_norm": 6.076840877532959, "learning_rate": 9.91122811995848e-06, "loss": 0.9171, "step": 3160 }, { "epoch": 0.24772727272727274, "grad_norm": 12.89234447479248, "learning_rate": 9.911170262082037e-06, "loss": 1.5793, "step": 3161 }, { "epoch": 0.24780564263322885, "grad_norm": 3.327042579650879, "learning_rate": 9.911112385526025e-06, "loss": 1.2833, "step": 3162 }, { "epoch": 0.24788401253918496, "grad_norm": 3.167271852493286, "learning_rate": 9.91105449029066e-06, "loss": 1.6553, "step": 3163 }, { "epoch": 0.24796238244514107, "grad_norm": 6.542513370513916, "learning_rate": 9.91099657637617e-06, "loss": 1.3227, "step": 3164 }, { "epoch": 0.24804075235109718, "grad_norm": 3.0651462078094482, "learning_rate": 9.910938643782773e-06, "loss": 1.2732, "step": 3165 }, { "epoch": 0.2481191222570533, "grad_norm": 2.652529001235962, "learning_rate": 9.910880692510686e-06, "loss": 2.0103, "step": 3166 }, { "epoch": 0.2481974921630094, "grad_norm": 5.721893787384033, "learning_rate": 9.91082272256013e-06, "loss": 0.8588, "step": 3167 }, { "epoch": 0.2482758620689655, "grad_norm": 5.919613361358643, "learning_rate": 9.910764733931328e-06, "loss": 1.284, "step": 3168 }, { "epoch": 0.24835423197492162, "grad_norm": 2.8910629749298096, "learning_rate": 9.910706726624498e-06, "loss": 1.6213, "step": 3169 }, { "epoch": 0.24843260188087773, "grad_norm": 2.3477792739868164, "learning_rate": 9.910648700639863e-06, "loss": 1.233, "step": 3170 }, { "epoch": 0.24851097178683385, "grad_norm": 2.593024730682373, "learning_rate": 9.910590655977641e-06, "loss": 1.4138, "step": 3171 }, { "epoch": 0.24858934169278996, "grad_norm": 3.0414581298828125, "learning_rate": 9.910532592638059e-06, "loss": 1.7197, "step": 3172 }, { "epoch": 0.2486677115987461, "grad_norm": 3.493349552154541, "learning_rate": 9.910474510621329e-06, "loss": 1.2351, "step": 3173 }, { "epoch": 0.2487460815047022, "grad_norm": 4.356367588043213, "learning_rate": 9.910416409927676e-06, "loss": 1.3636, "step": 3174 }, { "epoch": 0.24882445141065831, "grad_norm": 2.992575168609619, "learning_rate": 9.910358290557323e-06, "loss": 1.4458, "step": 3175 }, { "epoch": 0.24890282131661443, "grad_norm": 3.4292118549346924, "learning_rate": 9.910300152510488e-06, "loss": 1.2585, "step": 3176 }, { "epoch": 0.24898119122257054, "grad_norm": 4.013252258300781, "learning_rate": 9.910241995787394e-06, "loss": 1.4999, "step": 3177 }, { "epoch": 0.24905956112852665, "grad_norm": 3.6648428440093994, "learning_rate": 9.910183820388261e-06, "loss": 1.8754, "step": 3178 }, { "epoch": 0.24913793103448276, "grad_norm": 3.9117565155029297, "learning_rate": 9.910125626313312e-06, "loss": 1.6854, "step": 3179 }, { "epoch": 0.24921630094043887, "grad_norm": 3.440056324005127, "learning_rate": 9.910067413562766e-06, "loss": 1.628, "step": 3180 }, { "epoch": 0.24929467084639498, "grad_norm": 2.369858503341675, "learning_rate": 9.910009182136845e-06, "loss": 1.4003, "step": 3181 }, { "epoch": 0.2493730407523511, "grad_norm": 2.8642399311065674, "learning_rate": 9.909950932035773e-06, "loss": 1.4024, "step": 3182 }, { "epoch": 0.2494514106583072, "grad_norm": 3.0224876403808594, "learning_rate": 9.909892663259769e-06, "loss": 1.6614, "step": 3183 }, { "epoch": 0.2495297805642633, "grad_norm": 2.442396640777588, "learning_rate": 9.909834375809053e-06, "loss": 1.4675, "step": 3184 }, { "epoch": 0.24960815047021945, "grad_norm": 2.7759852409362793, "learning_rate": 9.90977606968385e-06, "loss": 1.6792, "step": 3185 }, { "epoch": 0.24968652037617556, "grad_norm": 2.3925223350524902, "learning_rate": 9.909717744884381e-06, "loss": 1.3, "step": 3186 }, { "epoch": 0.24976489028213167, "grad_norm": 2.5912296772003174, "learning_rate": 9.909659401410867e-06, "loss": 1.233, "step": 3187 }, { "epoch": 0.24984326018808778, "grad_norm": 2.8824870586395264, "learning_rate": 9.909601039263531e-06, "loss": 1.4169, "step": 3188 }, { "epoch": 0.2499216300940439, "grad_norm": 2.942199945449829, "learning_rate": 9.909542658442595e-06, "loss": 1.4679, "step": 3189 }, { "epoch": 0.25, "grad_norm": 3.0758674144744873, "learning_rate": 9.90948425894828e-06, "loss": 1.2474, "step": 3190 }, { "epoch": 0.25, "eval_loss": 1.416664719581604, "eval_runtime": 110.6475, "eval_samples_per_second": 24.284, "eval_steps_per_second": 6.073, "step": 3190 }, { "epoch": 0.2500783699059561, "grad_norm": 3.395426034927368, "learning_rate": 9.90942584078081e-06, "loss": 1.4185, "step": 3191 }, { "epoch": 0.2501567398119122, "grad_norm": 3.402768611907959, "learning_rate": 9.909367403940402e-06, "loss": 1.2466, "step": 3192 }, { "epoch": 0.25023510971786833, "grad_norm": 7.366020679473877, "learning_rate": 9.909308948427284e-06, "loss": 1.5868, "step": 3193 }, { "epoch": 0.25031347962382444, "grad_norm": 2.610994577407837, "learning_rate": 9.909250474241677e-06, "loss": 1.4676, "step": 3194 }, { "epoch": 0.25039184952978055, "grad_norm": 2.6993634700775146, "learning_rate": 9.909191981383801e-06, "loss": 1.1559, "step": 3195 }, { "epoch": 0.25047021943573666, "grad_norm": 2.4631636142730713, "learning_rate": 9.909133469853882e-06, "loss": 1.6155, "step": 3196 }, { "epoch": 0.2505485893416928, "grad_norm": 16.9494571685791, "learning_rate": 9.90907493965214e-06, "loss": 0.931, "step": 3197 }, { "epoch": 0.2506269592476489, "grad_norm": 3.4647433757781982, "learning_rate": 9.9090163907788e-06, "loss": 1.1309, "step": 3198 }, { "epoch": 0.250705329153605, "grad_norm": 3.3507251739501953, "learning_rate": 9.90895782323408e-06, "loss": 1.3535, "step": 3199 }, { "epoch": 0.2507836990595611, "grad_norm": 2.744288682937622, "learning_rate": 9.908899237018208e-06, "loss": 1.1581, "step": 3200 }, { "epoch": 0.2508620689655172, "grad_norm": 2.452540397644043, "learning_rate": 9.908840632131404e-06, "loss": 1.4765, "step": 3201 }, { "epoch": 0.2509404388714734, "grad_norm": 4.9568705558776855, "learning_rate": 9.908782008573891e-06, "loss": 1.4141, "step": 3202 }, { "epoch": 0.2510188087774295, "grad_norm": 5.048303127288818, "learning_rate": 9.908723366345893e-06, "loss": 1.549, "step": 3203 }, { "epoch": 0.2510971786833856, "grad_norm": 2.9679791927337646, "learning_rate": 9.908664705447633e-06, "loss": 1.4204, "step": 3204 }, { "epoch": 0.2511755485893417, "grad_norm": 4.482691287994385, "learning_rate": 9.908606025879333e-06, "loss": 1.1586, "step": 3205 }, { "epoch": 0.2512539184952978, "grad_norm": 3.3264448642730713, "learning_rate": 9.908547327641218e-06, "loss": 1.2641, "step": 3206 }, { "epoch": 0.25133228840125393, "grad_norm": 4.341264724731445, "learning_rate": 9.908488610733509e-06, "loss": 1.5473, "step": 3207 }, { "epoch": 0.25141065830721004, "grad_norm": 4.847893714904785, "learning_rate": 9.90842987515643e-06, "loss": 2.1171, "step": 3208 }, { "epoch": 0.25148902821316615, "grad_norm": 4.493956565856934, "learning_rate": 9.908371120910208e-06, "loss": 1.0166, "step": 3209 }, { "epoch": 0.25156739811912227, "grad_norm": 2.63785719871521, "learning_rate": 9.908312347995061e-06, "loss": 1.51, "step": 3210 }, { "epoch": 0.2516457680250784, "grad_norm": 4.09420108795166, "learning_rate": 9.908253556411214e-06, "loss": 1.491, "step": 3211 }, { "epoch": 0.2517241379310345, "grad_norm": 1.939839482307434, "learning_rate": 9.908194746158894e-06, "loss": 1.4905, "step": 3212 }, { "epoch": 0.2518025078369906, "grad_norm": 16.575843811035156, "learning_rate": 9.908135917238321e-06, "loss": 1.5081, "step": 3213 }, { "epoch": 0.2518808777429467, "grad_norm": 3.0876739025115967, "learning_rate": 9.90807706964972e-06, "loss": 1.6699, "step": 3214 }, { "epoch": 0.2519592476489028, "grad_norm": 2.4151408672332764, "learning_rate": 9.908018203393315e-06, "loss": 1.3469, "step": 3215 }, { "epoch": 0.25203761755485893, "grad_norm": 2.723161458969116, "learning_rate": 9.90795931846933e-06, "loss": 1.2261, "step": 3216 }, { "epoch": 0.25211598746081504, "grad_norm": 2.8666462898254395, "learning_rate": 9.907900414877987e-06, "loss": 0.9643, "step": 3217 }, { "epoch": 0.25219435736677115, "grad_norm": 5.59364128112793, "learning_rate": 9.907841492619513e-06, "loss": 1.9781, "step": 3218 }, { "epoch": 0.25227272727272726, "grad_norm": 3.9850738048553467, "learning_rate": 9.90778255169413e-06, "loss": 1.5412, "step": 3219 }, { "epoch": 0.25235109717868337, "grad_norm": 2.4131999015808105, "learning_rate": 9.907723592102062e-06, "loss": 1.0897, "step": 3220 }, { "epoch": 0.2524294670846395, "grad_norm": 6.987567901611328, "learning_rate": 9.907664613843537e-06, "loss": 2.1709, "step": 3221 }, { "epoch": 0.2525078369905956, "grad_norm": 3.098297595977783, "learning_rate": 9.907605616918775e-06, "loss": 1.3232, "step": 3222 }, { "epoch": 0.2525862068965517, "grad_norm": 3.291386127471924, "learning_rate": 9.907546601328002e-06, "loss": 1.5378, "step": 3223 }, { "epoch": 0.2526645768025078, "grad_norm": 3.7800283432006836, "learning_rate": 9.90748756707144e-06, "loss": 1.7326, "step": 3224 }, { "epoch": 0.2527429467084639, "grad_norm": 3.5469605922698975, "learning_rate": 9.90742851414932e-06, "loss": 1.7128, "step": 3225 }, { "epoch": 0.2528213166144201, "grad_norm": 4.45058536529541, "learning_rate": 9.90736944256186e-06, "loss": 1.4277, "step": 3226 }, { "epoch": 0.2528996865203762, "grad_norm": 2.63059139251709, "learning_rate": 9.90731035230929e-06, "loss": 1.2826, "step": 3227 }, { "epoch": 0.2529780564263323, "grad_norm": 6.339216232299805, "learning_rate": 9.90725124339183e-06, "loss": 1.4089, "step": 3228 }, { "epoch": 0.2530564263322884, "grad_norm": 7.075211524963379, "learning_rate": 9.907192115809706e-06, "loss": 0.7827, "step": 3229 }, { "epoch": 0.25313479623824453, "grad_norm": 3.4123265743255615, "learning_rate": 9.907132969563144e-06, "loss": 1.1828, "step": 3230 }, { "epoch": 0.25321316614420064, "grad_norm": 2.7739410400390625, "learning_rate": 9.90707380465237e-06, "loss": 1.5716, "step": 3231 }, { "epoch": 0.25329153605015675, "grad_norm": 3.7485787868499756, "learning_rate": 9.907014621077608e-06, "loss": 1.2674, "step": 3232 }, { "epoch": 0.25336990595611286, "grad_norm": 2.5088601112365723, "learning_rate": 9.906955418839082e-06, "loss": 1.5494, "step": 3233 }, { "epoch": 0.253448275862069, "grad_norm": 1.858393669128418, "learning_rate": 9.906896197937016e-06, "loss": 1.468, "step": 3234 }, { "epoch": 0.2535266457680251, "grad_norm": 2.504467725753784, "learning_rate": 9.90683695837164e-06, "loss": 1.4055, "step": 3235 }, { "epoch": 0.2536050156739812, "grad_norm": 6.564294338226318, "learning_rate": 9.906777700143175e-06, "loss": 1.5285, "step": 3236 }, { "epoch": 0.2536833855799373, "grad_norm": 3.351661205291748, "learning_rate": 9.906718423251849e-06, "loss": 1.4001, "step": 3237 }, { "epoch": 0.2537617554858934, "grad_norm": 2.9872779846191406, "learning_rate": 9.906659127697885e-06, "loss": 1.3747, "step": 3238 }, { "epoch": 0.2538401253918495, "grad_norm": 2.4503746032714844, "learning_rate": 9.90659981348151e-06, "loss": 1.6734, "step": 3239 }, { "epoch": 0.25391849529780564, "grad_norm": 4.755037784576416, "learning_rate": 9.90654048060295e-06, "loss": 1.27, "step": 3240 }, { "epoch": 0.25399686520376175, "grad_norm": 4.683676242828369, "learning_rate": 9.906481129062431e-06, "loss": 1.3645, "step": 3241 }, { "epoch": 0.25407523510971786, "grad_norm": 2.025205373764038, "learning_rate": 9.906421758860178e-06, "loss": 1.3178, "step": 3242 }, { "epoch": 0.25415360501567397, "grad_norm": 2.621358633041382, "learning_rate": 9.906362369996416e-06, "loss": 1.5995, "step": 3243 }, { "epoch": 0.2542319749216301, "grad_norm": 2.5608296394348145, "learning_rate": 9.90630296247137e-06, "loss": 1.5057, "step": 3244 }, { "epoch": 0.2543103448275862, "grad_norm": 5.274296760559082, "learning_rate": 9.90624353628527e-06, "loss": 1.3951, "step": 3245 }, { "epoch": 0.2543887147335423, "grad_norm": 2.6032297611236572, "learning_rate": 9.906184091438338e-06, "loss": 1.2716, "step": 3246 }, { "epoch": 0.2544670846394984, "grad_norm": 3.361764430999756, "learning_rate": 9.906124627930802e-06, "loss": 1.5027, "step": 3247 }, { "epoch": 0.2545454545454545, "grad_norm": 2.6770853996276855, "learning_rate": 9.906065145762889e-06, "loss": 1.5468, "step": 3248 }, { "epoch": 0.2546238244514107, "grad_norm": 4.168717861175537, "learning_rate": 9.906005644934825e-06, "loss": 1.7193, "step": 3249 }, { "epoch": 0.2547021943573668, "grad_norm": 4.314985275268555, "learning_rate": 9.905946125446832e-06, "loss": 1.5926, "step": 3250 }, { "epoch": 0.2547805642633229, "grad_norm": 3.8341686725616455, "learning_rate": 9.905886587299142e-06, "loss": 1.1986, "step": 3251 }, { "epoch": 0.254858934169279, "grad_norm": 2.314882278442383, "learning_rate": 9.90582703049198e-06, "loss": 1.6496, "step": 3252 }, { "epoch": 0.2549373040752351, "grad_norm": 3.18259859085083, "learning_rate": 9.90576745502557e-06, "loss": 1.464, "step": 3253 }, { "epoch": 0.25501567398119124, "grad_norm": 3.6495754718780518, "learning_rate": 9.90570786090014e-06, "loss": 1.5788, "step": 3254 }, { "epoch": 0.25509404388714735, "grad_norm": 4.057945728302002, "learning_rate": 9.905648248115918e-06, "loss": 1.2781, "step": 3255 }, { "epoch": 0.25517241379310346, "grad_norm": 2.8780739307403564, "learning_rate": 9.90558861667313e-06, "loss": 1.2929, "step": 3256 }, { "epoch": 0.25525078369905957, "grad_norm": 3.1204259395599365, "learning_rate": 9.905528966572002e-06, "loss": 1.3817, "step": 3257 }, { "epoch": 0.2553291536050157, "grad_norm": 8.455133438110352, "learning_rate": 9.905469297812761e-06, "loss": 0.9936, "step": 3258 }, { "epoch": 0.2554075235109718, "grad_norm": 2.4791784286499023, "learning_rate": 9.905409610395635e-06, "loss": 0.8856, "step": 3259 }, { "epoch": 0.2554858934169279, "grad_norm": 2.5248758792877197, "learning_rate": 9.90534990432085e-06, "loss": 1.3115, "step": 3260 }, { "epoch": 0.255564263322884, "grad_norm": 3.151407241821289, "learning_rate": 9.905290179588636e-06, "loss": 1.3375, "step": 3261 }, { "epoch": 0.2556426332288401, "grad_norm": 4.1619157791137695, "learning_rate": 9.905230436199215e-06, "loss": 1.3427, "step": 3262 }, { "epoch": 0.25572100313479623, "grad_norm": 2.5604450702667236, "learning_rate": 9.905170674152817e-06, "loss": 1.5063, "step": 3263 }, { "epoch": 0.25579937304075234, "grad_norm": 2.9146151542663574, "learning_rate": 9.90511089344967e-06, "loss": 1.199, "step": 3264 }, { "epoch": 0.25587774294670845, "grad_norm": 2.4416160583496094, "learning_rate": 9.90505109409e-06, "loss": 1.3895, "step": 3265 }, { "epoch": 0.25595611285266456, "grad_norm": 2.855092763900757, "learning_rate": 9.904991276074036e-06, "loss": 1.5536, "step": 3266 }, { "epoch": 0.2560344827586207, "grad_norm": 3.0252699851989746, "learning_rate": 9.904931439402003e-06, "loss": 1.6891, "step": 3267 }, { "epoch": 0.2561128526645768, "grad_norm": 3.0102853775024414, "learning_rate": 9.904871584074132e-06, "loss": 1.1752, "step": 3268 }, { "epoch": 0.2561912225705329, "grad_norm": 2.9453704357147217, "learning_rate": 9.904811710090648e-06, "loss": 1.2728, "step": 3269 }, { "epoch": 0.256269592476489, "grad_norm": 2.5279741287231445, "learning_rate": 9.90475181745178e-06, "loss": 1.7033, "step": 3270 }, { "epoch": 0.2563479623824451, "grad_norm": 5.235554218292236, "learning_rate": 9.904691906157755e-06, "loss": 1.8374, "step": 3271 }, { "epoch": 0.2564263322884012, "grad_norm": 2.4761104583740234, "learning_rate": 9.9046319762088e-06, "loss": 1.4884, "step": 3272 }, { "epoch": 0.2565047021943574, "grad_norm": 3.326552391052246, "learning_rate": 9.904572027605146e-06, "loss": 1.311, "step": 3273 }, { "epoch": 0.2565830721003135, "grad_norm": 3.484973907470703, "learning_rate": 9.904512060347017e-06, "loss": 1.4592, "step": 3274 }, { "epoch": 0.2566614420062696, "grad_norm": 2.4933581352233887, "learning_rate": 9.904452074434645e-06, "loss": 1.1462, "step": 3275 }, { "epoch": 0.2567398119122257, "grad_norm": 2.575950860977173, "learning_rate": 9.904392069868256e-06, "loss": 1.4242, "step": 3276 }, { "epoch": 0.25681818181818183, "grad_norm": 6.724659442901611, "learning_rate": 9.90433204664808e-06, "loss": 1.6208, "step": 3277 }, { "epoch": 0.25689655172413794, "grad_norm": 2.64272403717041, "learning_rate": 9.904272004774341e-06, "loss": 1.5286, "step": 3278 }, { "epoch": 0.25697492163009406, "grad_norm": 2.8964602947235107, "learning_rate": 9.904211944247274e-06, "loss": 1.5278, "step": 3279 }, { "epoch": 0.25705329153605017, "grad_norm": 2.9031124114990234, "learning_rate": 9.9041518650671e-06, "loss": 1.4582, "step": 3280 }, { "epoch": 0.2571316614420063, "grad_norm": 6.532564163208008, "learning_rate": 9.904091767234052e-06, "loss": 1.5587, "step": 3281 }, { "epoch": 0.2572100313479624, "grad_norm": 3.0756373405456543, "learning_rate": 9.904031650748358e-06, "loss": 1.2846, "step": 3282 }, { "epoch": 0.2572884012539185, "grad_norm": 2.741868495941162, "learning_rate": 9.903971515610247e-06, "loss": 1.6256, "step": 3283 }, { "epoch": 0.2573667711598746, "grad_norm": 6.112188339233398, "learning_rate": 9.903911361819948e-06, "loss": 1.3694, "step": 3284 }, { "epoch": 0.2574451410658307, "grad_norm": 4.195897579193115, "learning_rate": 9.903851189377688e-06, "loss": 2.0129, "step": 3285 }, { "epoch": 0.25752351097178683, "grad_norm": 2.4844110012054443, "learning_rate": 9.903790998283698e-06, "loss": 1.2853, "step": 3286 }, { "epoch": 0.25760188087774294, "grad_norm": 3.2525546550750732, "learning_rate": 9.903730788538204e-06, "loss": 1.9934, "step": 3287 }, { "epoch": 0.25768025078369905, "grad_norm": 3.0029079914093018, "learning_rate": 9.903670560141437e-06, "loss": 1.4983, "step": 3288 }, { "epoch": 0.25775862068965516, "grad_norm": 2.970547914505005, "learning_rate": 9.903610313093627e-06, "loss": 1.468, "step": 3289 }, { "epoch": 0.25783699059561127, "grad_norm": 3.5227932929992676, "learning_rate": 9.903550047395001e-06, "loss": 1.284, "step": 3290 }, { "epoch": 0.2579153605015674, "grad_norm": 5.148920059204102, "learning_rate": 9.903489763045791e-06, "loss": 1.3567, "step": 3291 }, { "epoch": 0.2579937304075235, "grad_norm": 2.8854174613952637, "learning_rate": 9.903429460046223e-06, "loss": 1.4833, "step": 3292 }, { "epoch": 0.2580721003134796, "grad_norm": 3.68147611618042, "learning_rate": 9.903369138396528e-06, "loss": 1.5899, "step": 3293 }, { "epoch": 0.2581504702194357, "grad_norm": 2.0593674182891846, "learning_rate": 9.903308798096936e-06, "loss": 1.3933, "step": 3294 }, { "epoch": 0.2582288401253918, "grad_norm": 4.455648899078369, "learning_rate": 9.903248439147674e-06, "loss": 1.528, "step": 3295 }, { "epoch": 0.258307210031348, "grad_norm": 3.408921957015991, "learning_rate": 9.903188061548976e-06, "loss": 1.1488, "step": 3296 }, { "epoch": 0.2583855799373041, "grad_norm": 2.4642434120178223, "learning_rate": 9.903127665301067e-06, "loss": 1.4148, "step": 3297 }, { "epoch": 0.2584639498432602, "grad_norm": 2.202465772628784, "learning_rate": 9.90306725040418e-06, "loss": 1.4355, "step": 3298 }, { "epoch": 0.2585423197492163, "grad_norm": 4.04816198348999, "learning_rate": 9.903006816858543e-06, "loss": 1.4567, "step": 3299 }, { "epoch": 0.25862068965517243, "grad_norm": 2.338773012161255, "learning_rate": 9.902946364664386e-06, "loss": 1.3538, "step": 3300 }, { "epoch": 0.25869905956112854, "grad_norm": 2.6377148628234863, "learning_rate": 9.902885893821941e-06, "loss": 1.521, "step": 3301 }, { "epoch": 0.25877742946708465, "grad_norm": 9.2626371383667, "learning_rate": 9.902825404331434e-06, "loss": 1.3343, "step": 3302 }, { "epoch": 0.25885579937304076, "grad_norm": 2.720400333404541, "learning_rate": 9.902764896193099e-06, "loss": 1.3724, "step": 3303 }, { "epoch": 0.2589341692789969, "grad_norm": 4.4656596183776855, "learning_rate": 9.902704369407164e-06, "loss": 1.0702, "step": 3304 }, { "epoch": 0.259012539184953, "grad_norm": 3.1772217750549316, "learning_rate": 9.902643823973861e-06, "loss": 1.553, "step": 3305 }, { "epoch": 0.2590909090909091, "grad_norm": 3.537886381149292, "learning_rate": 9.902583259893418e-06, "loss": 1.4664, "step": 3306 }, { "epoch": 0.2591692789968652, "grad_norm": 2.769599676132202, "learning_rate": 9.902522677166066e-06, "loss": 1.5301, "step": 3307 }, { "epoch": 0.2592476489028213, "grad_norm": 4.174251079559326, "learning_rate": 9.902462075792037e-06, "loss": 1.4796, "step": 3308 }, { "epoch": 0.2593260188087774, "grad_norm": 4.603775501251221, "learning_rate": 9.902401455771557e-06, "loss": 1.6035, "step": 3309 }, { "epoch": 0.25940438871473354, "grad_norm": 3.135509729385376, "learning_rate": 9.902340817104864e-06, "loss": 1.6422, "step": 3310 }, { "epoch": 0.25948275862068965, "grad_norm": 1.8728786706924438, "learning_rate": 9.902280159792183e-06, "loss": 1.3872, "step": 3311 }, { "epoch": 0.25956112852664576, "grad_norm": 2.315211057662964, "learning_rate": 9.902219483833746e-06, "loss": 1.4333, "step": 3312 }, { "epoch": 0.25963949843260187, "grad_norm": 4.051080703735352, "learning_rate": 9.902158789229784e-06, "loss": 1.7344, "step": 3313 }, { "epoch": 0.259717868338558, "grad_norm": 4.973834991455078, "learning_rate": 9.902098075980528e-06, "loss": 1.4941, "step": 3314 }, { "epoch": 0.2597962382445141, "grad_norm": 4.355926513671875, "learning_rate": 9.902037344086209e-06, "loss": 1.427, "step": 3315 }, { "epoch": 0.2598746081504702, "grad_norm": 4.703518390655518, "learning_rate": 9.901976593547057e-06, "loss": 1.4383, "step": 3316 }, { "epoch": 0.2599529780564263, "grad_norm": 3.4978187084198, "learning_rate": 9.901915824363307e-06, "loss": 1.6948, "step": 3317 }, { "epoch": 0.2600313479623824, "grad_norm": 3.2576286792755127, "learning_rate": 9.901855036535183e-06, "loss": 1.5129, "step": 3318 }, { "epoch": 0.26010971786833853, "grad_norm": 4.37445068359375, "learning_rate": 9.901794230062923e-06, "loss": 1.4998, "step": 3319 }, { "epoch": 0.2601880877742947, "grad_norm": 2.4524033069610596, "learning_rate": 9.901733404946752e-06, "loss": 1.7044, "step": 3320 }, { "epoch": 0.2602664576802508, "grad_norm": 3.5237233638763428, "learning_rate": 9.901672561186907e-06, "loss": 1.3219, "step": 3321 }, { "epoch": 0.2603448275862069, "grad_norm": 2.7226622104644775, "learning_rate": 9.901611698783618e-06, "loss": 1.288, "step": 3322 }, { "epoch": 0.260423197492163, "grad_norm": 5.705294609069824, "learning_rate": 9.901550817737115e-06, "loss": 1.4611, "step": 3323 }, { "epoch": 0.26050156739811914, "grad_norm": 4.226880073547363, "learning_rate": 9.90148991804763e-06, "loss": 1.5101, "step": 3324 }, { "epoch": 0.26057993730407525, "grad_norm": 3.252274751663208, "learning_rate": 9.901428999715396e-06, "loss": 1.7324, "step": 3325 }, { "epoch": 0.26065830721003136, "grad_norm": 4.1480607986450195, "learning_rate": 9.901368062740642e-06, "loss": 1.9737, "step": 3326 }, { "epoch": 0.26073667711598747, "grad_norm": 2.6681556701660156, "learning_rate": 9.901307107123603e-06, "loss": 1.4202, "step": 3327 }, { "epoch": 0.2608150470219436, "grad_norm": 4.0683183670043945, "learning_rate": 9.901246132864507e-06, "loss": 1.0233, "step": 3328 }, { "epoch": 0.2608934169278997, "grad_norm": 3.7232701778411865, "learning_rate": 9.90118513996359e-06, "loss": 1.721, "step": 3329 }, { "epoch": 0.2609717868338558, "grad_norm": 2.6184418201446533, "learning_rate": 9.901124128421082e-06, "loss": 1.3618, "step": 3330 }, { "epoch": 0.2610501567398119, "grad_norm": 2.800398826599121, "learning_rate": 9.901063098237216e-06, "loss": 1.3828, "step": 3331 }, { "epoch": 0.261128526645768, "grad_norm": 3.815915584564209, "learning_rate": 9.901002049412222e-06, "loss": 1.254, "step": 3332 }, { "epoch": 0.26120689655172413, "grad_norm": 2.445018768310547, "learning_rate": 9.900940981946334e-06, "loss": 1.4636, "step": 3333 }, { "epoch": 0.26128526645768024, "grad_norm": 4.497681140899658, "learning_rate": 9.900879895839783e-06, "loss": 1.0884, "step": 3334 }, { "epoch": 0.26136363636363635, "grad_norm": 2.552945375442505, "learning_rate": 9.900818791092802e-06, "loss": 1.4667, "step": 3335 }, { "epoch": 0.26144200626959246, "grad_norm": 2.4502313137054443, "learning_rate": 9.900757667705624e-06, "loss": 1.4496, "step": 3336 }, { "epoch": 0.2615203761755486, "grad_norm": 3.930328607559204, "learning_rate": 9.900696525678482e-06, "loss": 1.1362, "step": 3337 }, { "epoch": 0.2615987460815047, "grad_norm": 2.2915432453155518, "learning_rate": 9.900635365011607e-06, "loss": 1.4113, "step": 3338 }, { "epoch": 0.2616771159874608, "grad_norm": 2.761146306991577, "learning_rate": 9.900574185705232e-06, "loss": 1.5932, "step": 3339 }, { "epoch": 0.2617554858934169, "grad_norm": 2.851666212081909, "learning_rate": 9.90051298775959e-06, "loss": 1.2354, "step": 3340 }, { "epoch": 0.261833855799373, "grad_norm": 2.5798122882843018, "learning_rate": 9.900451771174913e-06, "loss": 1.2787, "step": 3341 }, { "epoch": 0.2619122257053291, "grad_norm": 3.4902565479278564, "learning_rate": 9.900390535951436e-06, "loss": 1.2042, "step": 3342 }, { "epoch": 0.2619905956112853, "grad_norm": 4.790990829467773, "learning_rate": 9.90032928208939e-06, "loss": 1.5802, "step": 3343 }, { "epoch": 0.2620689655172414, "grad_norm": 2.6733222007751465, "learning_rate": 9.900268009589007e-06, "loss": 1.1726, "step": 3344 }, { "epoch": 0.2621473354231975, "grad_norm": 5.444737434387207, "learning_rate": 9.900206718450522e-06, "loss": 1.7582, "step": 3345 }, { "epoch": 0.2622257053291536, "grad_norm": 4.423007965087891, "learning_rate": 9.900145408674166e-06, "loss": 1.7322, "step": 3346 }, { "epoch": 0.26230407523510973, "grad_norm": 3.0292539596557617, "learning_rate": 9.900084080260176e-06, "loss": 1.5168, "step": 3347 }, { "epoch": 0.26238244514106585, "grad_norm": 5.9396467208862305, "learning_rate": 9.900022733208782e-06, "loss": 1.6643, "step": 3348 }, { "epoch": 0.26246081504702196, "grad_norm": 3.3671164512634277, "learning_rate": 9.89996136752022e-06, "loss": 1.4113, "step": 3349 }, { "epoch": 0.26253918495297807, "grad_norm": 2.31318736076355, "learning_rate": 9.89989998319472e-06, "loss": 1.4554, "step": 3350 }, { "epoch": 0.2626175548589342, "grad_norm": 2.737607002258301, "learning_rate": 9.89983858023252e-06, "loss": 1.4156, "step": 3351 }, { "epoch": 0.2626959247648903, "grad_norm": 3.123948574066162, "learning_rate": 9.899777158633848e-06, "loss": 1.6765, "step": 3352 }, { "epoch": 0.2627742946708464, "grad_norm": 2.9213619232177734, "learning_rate": 9.89971571839894e-06, "loss": 1.7033, "step": 3353 }, { "epoch": 0.2628526645768025, "grad_norm": 2.288074016571045, "learning_rate": 9.89965425952803e-06, "loss": 1.3262, "step": 3354 }, { "epoch": 0.2629310344827586, "grad_norm": 2.6693994998931885, "learning_rate": 9.899592782021355e-06, "loss": 1.5055, "step": 3355 }, { "epoch": 0.26300940438871473, "grad_norm": 2.957813024520874, "learning_rate": 9.899531285879143e-06, "loss": 1.3859, "step": 3356 }, { "epoch": 0.26308777429467084, "grad_norm": 3.013759136199951, "learning_rate": 9.89946977110163e-06, "loss": 1.6761, "step": 3357 }, { "epoch": 0.26316614420062695, "grad_norm": 1.7784967422485352, "learning_rate": 9.899408237689052e-06, "loss": 0.8128, "step": 3358 }, { "epoch": 0.26324451410658306, "grad_norm": 3.6766655445098877, "learning_rate": 9.89934668564164e-06, "loss": 1.1899, "step": 3359 }, { "epoch": 0.26332288401253917, "grad_norm": 2.275850534439087, "learning_rate": 9.899285114959631e-06, "loss": 1.549, "step": 3360 }, { "epoch": 0.2634012539184953, "grad_norm": 4.552667140960693, "learning_rate": 9.899223525643258e-06, "loss": 1.8417, "step": 3361 }, { "epoch": 0.2634796238244514, "grad_norm": 2.3233225345611572, "learning_rate": 9.899161917692754e-06, "loss": 1.0139, "step": 3362 }, { "epoch": 0.2635579937304075, "grad_norm": 2.1937203407287598, "learning_rate": 9.899100291108356e-06, "loss": 1.2149, "step": 3363 }, { "epoch": 0.2636363636363636, "grad_norm": 4.947376251220703, "learning_rate": 9.899038645890295e-06, "loss": 1.7246, "step": 3364 }, { "epoch": 0.2637147335423197, "grad_norm": 5.376155376434326, "learning_rate": 9.898976982038809e-06, "loss": 1.1303, "step": 3365 }, { "epoch": 0.2637931034482759, "grad_norm": 2.475926160812378, "learning_rate": 9.89891529955413e-06, "loss": 1.5577, "step": 3366 }, { "epoch": 0.263871473354232, "grad_norm": 3.0147807598114014, "learning_rate": 9.898853598436495e-06, "loss": 1.5999, "step": 3367 }, { "epoch": 0.2639498432601881, "grad_norm": 5.87497091293335, "learning_rate": 9.898791878686134e-06, "loss": 1.5772, "step": 3368 }, { "epoch": 0.2640282131661442, "grad_norm": 4.567055702209473, "learning_rate": 9.898730140303287e-06, "loss": 0.8255, "step": 3369 }, { "epoch": 0.26410658307210033, "grad_norm": 8.044934272766113, "learning_rate": 9.898668383288185e-06, "loss": 1.4328, "step": 3370 }, { "epoch": 0.26418495297805644, "grad_norm": 3.2449421882629395, "learning_rate": 9.898606607641069e-06, "loss": 1.8203, "step": 3371 }, { "epoch": 0.26426332288401255, "grad_norm": 3.600090742111206, "learning_rate": 9.898544813362165e-06, "loss": 1.2091, "step": 3372 }, { "epoch": 0.26434169278996866, "grad_norm": 3.187681198120117, "learning_rate": 9.898483000451713e-06, "loss": 1.2702, "step": 3373 }, { "epoch": 0.2644200626959248, "grad_norm": 2.311276435852051, "learning_rate": 9.89842116890995e-06, "loss": 1.4632, "step": 3374 }, { "epoch": 0.2644984326018809, "grad_norm": 2.973095417022705, "learning_rate": 9.898359318737107e-06, "loss": 1.2526, "step": 3375 }, { "epoch": 0.264576802507837, "grad_norm": 3.4719200134277344, "learning_rate": 9.898297449933422e-06, "loss": 1.2915, "step": 3376 }, { "epoch": 0.2646551724137931, "grad_norm": 2.6591267585754395, "learning_rate": 9.898235562499127e-06, "loss": 1.3067, "step": 3377 }, { "epoch": 0.2647335423197492, "grad_norm": 4.504257678985596, "learning_rate": 9.898173656434463e-06, "loss": 1.5135, "step": 3378 }, { "epoch": 0.2648119122257053, "grad_norm": 2.1883931159973145, "learning_rate": 9.898111731739662e-06, "loss": 1.2634, "step": 3379 }, { "epoch": 0.26489028213166144, "grad_norm": 2.1714096069335938, "learning_rate": 9.898049788414959e-06, "loss": 1.3887, "step": 3380 }, { "epoch": 0.26496865203761755, "grad_norm": 1.785656452178955, "learning_rate": 9.89798782646059e-06, "loss": 1.2792, "step": 3381 }, { "epoch": 0.26504702194357366, "grad_norm": 2.509451150894165, "learning_rate": 9.897925845876792e-06, "loss": 1.408, "step": 3382 }, { "epoch": 0.26512539184952977, "grad_norm": 9.317475318908691, "learning_rate": 9.897863846663798e-06, "loss": 1.5581, "step": 3383 }, { "epoch": 0.2652037617554859, "grad_norm": 3.001281499862671, "learning_rate": 9.897801828821847e-06, "loss": 1.2033, "step": 3384 }, { "epoch": 0.265282131661442, "grad_norm": 2.7732009887695312, "learning_rate": 9.897739792351174e-06, "loss": 1.4464, "step": 3385 }, { "epoch": 0.2653605015673981, "grad_norm": 3.3380355834960938, "learning_rate": 9.897677737252014e-06, "loss": 1.0153, "step": 3386 }, { "epoch": 0.2654388714733542, "grad_norm": 2.477391004562378, "learning_rate": 9.897615663524602e-06, "loss": 1.4795, "step": 3387 }, { "epoch": 0.2655172413793103, "grad_norm": 2.660404920578003, "learning_rate": 9.897553571169178e-06, "loss": 1.09, "step": 3388 }, { "epoch": 0.26559561128526643, "grad_norm": 2.1671106815338135, "learning_rate": 9.897491460185975e-06, "loss": 1.5519, "step": 3389 }, { "epoch": 0.2656739811912226, "grad_norm": 2.595141649246216, "learning_rate": 9.897429330575228e-06, "loss": 1.4148, "step": 3390 }, { "epoch": 0.2657523510971787, "grad_norm": 2.184176206588745, "learning_rate": 9.897367182337178e-06, "loss": 1.5193, "step": 3391 }, { "epoch": 0.2658307210031348, "grad_norm": 7.242131233215332, "learning_rate": 9.897305015472058e-06, "loss": 1.1569, "step": 3392 }, { "epoch": 0.26590909090909093, "grad_norm": 4.494854927062988, "learning_rate": 9.897242829980104e-06, "loss": 1.5339, "step": 3393 }, { "epoch": 0.26598746081504704, "grad_norm": 2.8974766731262207, "learning_rate": 9.897180625861555e-06, "loss": 1.1085, "step": 3394 }, { "epoch": 0.26606583072100315, "grad_norm": 4.965991497039795, "learning_rate": 9.897118403116645e-06, "loss": 1.5597, "step": 3395 }, { "epoch": 0.26614420062695926, "grad_norm": 2.638104200363159, "learning_rate": 9.897056161745614e-06, "loss": 1.5339, "step": 3396 }, { "epoch": 0.26622257053291537, "grad_norm": 2.5544357299804688, "learning_rate": 9.896993901748695e-06, "loss": 1.5011, "step": 3397 }, { "epoch": 0.2663009404388715, "grad_norm": 3.1986427307128906, "learning_rate": 9.896931623126127e-06, "loss": 1.473, "step": 3398 }, { "epoch": 0.2663793103448276, "grad_norm": 3.2156870365142822, "learning_rate": 9.896869325878144e-06, "loss": 1.3233, "step": 3399 }, { "epoch": 0.2664576802507837, "grad_norm": 3.4388821125030518, "learning_rate": 9.896807010004988e-06, "loss": 0.8847, "step": 3400 }, { "epoch": 0.2665360501567398, "grad_norm": 2.7283568382263184, "learning_rate": 9.896744675506893e-06, "loss": 1.4762, "step": 3401 }, { "epoch": 0.2666144200626959, "grad_norm": 3.2712817192077637, "learning_rate": 9.896682322384096e-06, "loss": 1.372, "step": 3402 }, { "epoch": 0.26669278996865203, "grad_norm": 6.088237285614014, "learning_rate": 9.896619950636835e-06, "loss": 1.0547, "step": 3403 }, { "epoch": 0.26677115987460814, "grad_norm": 4.5713210105896, "learning_rate": 9.896557560265345e-06, "loss": 1.8097, "step": 3404 }, { "epoch": 0.26684952978056425, "grad_norm": 5.560515403747559, "learning_rate": 9.896495151269866e-06, "loss": 1.7148, "step": 3405 }, { "epoch": 0.26692789968652036, "grad_norm": 2.880826711654663, "learning_rate": 9.896432723650635e-06, "loss": 1.4635, "step": 3406 }, { "epoch": 0.2670062695924765, "grad_norm": 3.3962154388427734, "learning_rate": 9.89637027740789e-06, "loss": 1.5855, "step": 3407 }, { "epoch": 0.2670846394984326, "grad_norm": 8.932549476623535, "learning_rate": 9.896307812541865e-06, "loss": 1.1399, "step": 3408 }, { "epoch": 0.2671630094043887, "grad_norm": 4.182672023773193, "learning_rate": 9.896245329052803e-06, "loss": 1.6074, "step": 3409 }, { "epoch": 0.2672413793103448, "grad_norm": 2.9569876194000244, "learning_rate": 9.896182826940935e-06, "loss": 1.4248, "step": 3410 }, { "epoch": 0.2673197492163009, "grad_norm": 9.376962661743164, "learning_rate": 9.896120306206505e-06, "loss": 1.3384, "step": 3411 }, { "epoch": 0.267398119122257, "grad_norm": 3.147987127304077, "learning_rate": 9.896057766849746e-06, "loss": 1.1476, "step": 3412 }, { "epoch": 0.2674764890282132, "grad_norm": 2.476266860961914, "learning_rate": 9.8959952088709e-06, "loss": 1.2175, "step": 3413 }, { "epoch": 0.2675548589341693, "grad_norm": 3.642315149307251, "learning_rate": 9.895932632270202e-06, "loss": 1.4156, "step": 3414 }, { "epoch": 0.2676332288401254, "grad_norm": 2.9801766872406006, "learning_rate": 9.895870037047892e-06, "loss": 1.9081, "step": 3415 }, { "epoch": 0.2677115987460815, "grad_norm": 2.625866413116455, "learning_rate": 9.895807423204207e-06, "loss": 1.5116, "step": 3416 }, { "epoch": 0.26778996865203764, "grad_norm": 2.6587092876434326, "learning_rate": 9.895744790739383e-06, "loss": 1.0799, "step": 3417 }, { "epoch": 0.26786833855799375, "grad_norm": 2.5660476684570312, "learning_rate": 9.895682139653661e-06, "loss": 1.2148, "step": 3418 }, { "epoch": 0.26794670846394986, "grad_norm": 2.47714900970459, "learning_rate": 9.895619469947282e-06, "loss": 1.1398, "step": 3419 }, { "epoch": 0.26802507836990597, "grad_norm": 3.148404598236084, "learning_rate": 9.895556781620477e-06, "loss": 1.8139, "step": 3420 }, { "epoch": 0.2681034482758621, "grad_norm": 2.2776010036468506, "learning_rate": 9.89549407467349e-06, "loss": 1.2828, "step": 3421 }, { "epoch": 0.2681818181818182, "grad_norm": 3.539666175842285, "learning_rate": 9.895431349106559e-06, "loss": 1.5156, "step": 3422 }, { "epoch": 0.2682601880877743, "grad_norm": 3.0923397541046143, "learning_rate": 9.895368604919922e-06, "loss": 1.3383, "step": 3423 }, { "epoch": 0.2683385579937304, "grad_norm": 3.5518860816955566, "learning_rate": 9.895305842113816e-06, "loss": 1.0099, "step": 3424 }, { "epoch": 0.2684169278996865, "grad_norm": 3.056171417236328, "learning_rate": 9.89524306068848e-06, "loss": 1.084, "step": 3425 }, { "epoch": 0.26849529780564263, "grad_norm": 4.862053394317627, "learning_rate": 9.895180260644156e-06, "loss": 1.7079, "step": 3426 }, { "epoch": 0.26857366771159874, "grad_norm": 6.694140911102295, "learning_rate": 9.89511744198108e-06, "loss": 1.5389, "step": 3427 }, { "epoch": 0.26865203761755485, "grad_norm": 4.115976810455322, "learning_rate": 9.89505460469949e-06, "loss": 1.4597, "step": 3428 }, { "epoch": 0.26873040752351096, "grad_norm": 3.1336019039154053, "learning_rate": 9.894991748799627e-06, "loss": 1.3776, "step": 3429 }, { "epoch": 0.26880877742946707, "grad_norm": 4.869903087615967, "learning_rate": 9.89492887428173e-06, "loss": 1.4189, "step": 3430 }, { "epoch": 0.2688871473354232, "grad_norm": 2.3106324672698975, "learning_rate": 9.894865981146039e-06, "loss": 1.594, "step": 3431 }, { "epoch": 0.2689655172413793, "grad_norm": 5.621463775634766, "learning_rate": 9.89480306939279e-06, "loss": 1.3584, "step": 3432 }, { "epoch": 0.2690438871473354, "grad_norm": 3.502894163131714, "learning_rate": 9.894740139022225e-06, "loss": 1.4039, "step": 3433 }, { "epoch": 0.2691222570532915, "grad_norm": 5.081161022186279, "learning_rate": 9.894677190034583e-06, "loss": 1.5507, "step": 3434 }, { "epoch": 0.2692006269592476, "grad_norm": 2.6915860176086426, "learning_rate": 9.894614222430103e-06, "loss": 1.492, "step": 3435 }, { "epoch": 0.26927899686520373, "grad_norm": 4.442990779876709, "learning_rate": 9.894551236209025e-06, "loss": 1.3353, "step": 3436 }, { "epoch": 0.2693573667711599, "grad_norm": 2.749488353729248, "learning_rate": 9.894488231371587e-06, "loss": 1.6934, "step": 3437 }, { "epoch": 0.269435736677116, "grad_norm": 4.880468368530273, "learning_rate": 9.89442520791803e-06, "loss": 1.274, "step": 3438 }, { "epoch": 0.2695141065830721, "grad_norm": 4.64379358291626, "learning_rate": 9.894362165848593e-06, "loss": 1.4585, "step": 3439 }, { "epoch": 0.26959247648902823, "grad_norm": 2.8937108516693115, "learning_rate": 9.894299105163516e-06, "loss": 1.2412, "step": 3440 }, { "epoch": 0.26967084639498434, "grad_norm": 2.4958722591400146, "learning_rate": 9.89423602586304e-06, "loss": 1.3666, "step": 3441 }, { "epoch": 0.26974921630094045, "grad_norm": 2.907005786895752, "learning_rate": 9.894172927947405e-06, "loss": 1.4749, "step": 3442 }, { "epoch": 0.26982758620689656, "grad_norm": 2.7270588874816895, "learning_rate": 9.894109811416847e-06, "loss": 1.4613, "step": 3443 }, { "epoch": 0.2699059561128527, "grad_norm": 2.7021467685699463, "learning_rate": 9.894046676271613e-06, "loss": 1.6018, "step": 3444 }, { "epoch": 0.2699843260188088, "grad_norm": 9.74789047241211, "learning_rate": 9.893983522511935e-06, "loss": 1.2066, "step": 3445 }, { "epoch": 0.2700626959247649, "grad_norm": 4.564647674560547, "learning_rate": 9.89392035013806e-06, "loss": 1.8039, "step": 3446 }, { "epoch": 0.270141065830721, "grad_norm": 5.463686943054199, "learning_rate": 9.893857159150226e-06, "loss": 1.1899, "step": 3447 }, { "epoch": 0.2702194357366771, "grad_norm": 3.9323947429656982, "learning_rate": 9.893793949548672e-06, "loss": 1.3794, "step": 3448 }, { "epoch": 0.2702978056426332, "grad_norm": 6.9419169425964355, "learning_rate": 9.89373072133364e-06, "loss": 0.9535, "step": 3449 }, { "epoch": 0.27037617554858934, "grad_norm": 2.5403144359588623, "learning_rate": 9.89366747450537e-06, "loss": 1.5211, "step": 3450 }, { "epoch": 0.27045454545454545, "grad_norm": 2.7560079097747803, "learning_rate": 9.8936042090641e-06, "loss": 1.2529, "step": 3451 }, { "epoch": 0.27053291536050156, "grad_norm": 3.76340913772583, "learning_rate": 9.893540925010077e-06, "loss": 1.1857, "step": 3452 }, { "epoch": 0.27061128526645767, "grad_norm": 3.239929437637329, "learning_rate": 9.893477622343534e-06, "loss": 1.7785, "step": 3453 }, { "epoch": 0.2706896551724138, "grad_norm": 2.684561014175415, "learning_rate": 9.89341430106472e-06, "loss": 1.512, "step": 3454 }, { "epoch": 0.2707680250783699, "grad_norm": 6.084583282470703, "learning_rate": 9.893350961173868e-06, "loss": 1.0382, "step": 3455 }, { "epoch": 0.270846394984326, "grad_norm": 3.878706693649292, "learning_rate": 9.893287602671221e-06, "loss": 1.0614, "step": 3456 }, { "epoch": 0.2709247648902821, "grad_norm": 4.573005676269531, "learning_rate": 9.893224225557025e-06, "loss": 0.9496, "step": 3457 }, { "epoch": 0.2710031347962382, "grad_norm": 3.0258777141571045, "learning_rate": 9.893160829831517e-06, "loss": 1.5145, "step": 3458 }, { "epoch": 0.27108150470219433, "grad_norm": 3.3593685626983643, "learning_rate": 9.893097415494936e-06, "loss": 1.5979, "step": 3459 }, { "epoch": 0.2711598746081505, "grad_norm": 3.2942259311676025, "learning_rate": 9.893033982547528e-06, "loss": 1.3989, "step": 3460 }, { "epoch": 0.2712382445141066, "grad_norm": 2.9887359142303467, "learning_rate": 9.892970530989531e-06, "loss": 1.596, "step": 3461 }, { "epoch": 0.2713166144200627, "grad_norm": 8.3511962890625, "learning_rate": 9.892907060821186e-06, "loss": 0.9179, "step": 3462 }, { "epoch": 0.27139498432601883, "grad_norm": 2.776683807373047, "learning_rate": 9.89284357204274e-06, "loss": 1.4571, "step": 3463 }, { "epoch": 0.27147335423197494, "grad_norm": 4.966998100280762, "learning_rate": 9.892780064654424e-06, "loss": 1.4356, "step": 3464 }, { "epoch": 0.27155172413793105, "grad_norm": 2.081193447113037, "learning_rate": 9.892716538656491e-06, "loss": 1.2967, "step": 3465 }, { "epoch": 0.27163009404388716, "grad_norm": 11.669212341308594, "learning_rate": 9.892652994049174e-06, "loss": 1.6259, "step": 3466 }, { "epoch": 0.27170846394984327, "grad_norm": 3.0688066482543945, "learning_rate": 9.89258943083272e-06, "loss": 1.4707, "step": 3467 }, { "epoch": 0.2717868338557994, "grad_norm": 3.1663665771484375, "learning_rate": 9.892525849007366e-06, "loss": 1.4732, "step": 3468 }, { "epoch": 0.2718652037617555, "grad_norm": 2.3360118865966797, "learning_rate": 9.89246224857336e-06, "loss": 1.5644, "step": 3469 }, { "epoch": 0.2719435736677116, "grad_norm": 4.424076080322266, "learning_rate": 9.89239862953094e-06, "loss": 1.9062, "step": 3470 }, { "epoch": 0.2720219435736677, "grad_norm": 2.295304775238037, "learning_rate": 9.892334991880346e-06, "loss": 1.4088, "step": 3471 }, { "epoch": 0.2721003134796238, "grad_norm": 12.078376770019531, "learning_rate": 9.892271335621826e-06, "loss": 1.3357, "step": 3472 }, { "epoch": 0.27217868338557993, "grad_norm": 2.795818567276001, "learning_rate": 9.892207660755615e-06, "loss": 1.4001, "step": 3473 }, { "epoch": 0.27225705329153604, "grad_norm": 2.7092368602752686, "learning_rate": 9.892143967281962e-06, "loss": 1.8004, "step": 3474 }, { "epoch": 0.27233542319749215, "grad_norm": 4.312455654144287, "learning_rate": 9.892080255201105e-06, "loss": 1.0495, "step": 3475 }, { "epoch": 0.27241379310344827, "grad_norm": 2.8664486408233643, "learning_rate": 9.892016524513289e-06, "loss": 0.8511, "step": 3476 }, { "epoch": 0.2724921630094044, "grad_norm": 2.817767858505249, "learning_rate": 9.891952775218751e-06, "loss": 1.452, "step": 3477 }, { "epoch": 0.2725705329153605, "grad_norm": 4.45424222946167, "learning_rate": 9.89188900731774e-06, "loss": 1.4847, "step": 3478 }, { "epoch": 0.2726489028213166, "grad_norm": 3.2566256523132324, "learning_rate": 9.891825220810496e-06, "loss": 1.3541, "step": 3479 }, { "epoch": 0.2727272727272727, "grad_norm": 6.166380882263184, "learning_rate": 9.89176141569726e-06, "loss": 1.4337, "step": 3480 }, { "epoch": 0.2728056426332288, "grad_norm": 2.591137409210205, "learning_rate": 9.891697591978278e-06, "loss": 1.3566, "step": 3481 }, { "epoch": 0.27288401253918493, "grad_norm": 4.314634323120117, "learning_rate": 9.891633749653788e-06, "loss": 1.8915, "step": 3482 }, { "epoch": 0.27296238244514104, "grad_norm": 3.983762741088867, "learning_rate": 9.891569888724038e-06, "loss": 0.9509, "step": 3483 }, { "epoch": 0.2730407523510972, "grad_norm": 3.1043076515197754, "learning_rate": 9.891506009189269e-06, "loss": 1.4217, "step": 3484 }, { "epoch": 0.2731191222570533, "grad_norm": 3.5215656757354736, "learning_rate": 9.891442111049722e-06, "loss": 1.3628, "step": 3485 }, { "epoch": 0.2731974921630094, "grad_norm": 2.2578184604644775, "learning_rate": 9.891378194305643e-06, "loss": 1.4452, "step": 3486 }, { "epoch": 0.27327586206896554, "grad_norm": 2.1843907833099365, "learning_rate": 9.891314258957274e-06, "loss": 1.2582, "step": 3487 }, { "epoch": 0.27335423197492165, "grad_norm": 3.115318775177002, "learning_rate": 9.891250305004856e-06, "loss": 1.2754, "step": 3488 }, { "epoch": 0.27343260188087776, "grad_norm": 10.34497356414795, "learning_rate": 9.891186332448634e-06, "loss": 1.0911, "step": 3489 }, { "epoch": 0.27351097178683387, "grad_norm": 4.532251834869385, "learning_rate": 9.891122341288854e-06, "loss": 1.4968, "step": 3490 }, { "epoch": 0.27358934169279, "grad_norm": 3.8960695266723633, "learning_rate": 9.891058331525756e-06, "loss": 1.2787, "step": 3491 }, { "epoch": 0.2736677115987461, "grad_norm": 6.176685333251953, "learning_rate": 9.890994303159583e-06, "loss": 1.5629, "step": 3492 }, { "epoch": 0.2737460815047022, "grad_norm": 2.959531307220459, "learning_rate": 9.89093025619058e-06, "loss": 0.8351, "step": 3493 }, { "epoch": 0.2738244514106583, "grad_norm": 2.6809070110321045, "learning_rate": 9.890866190618992e-06, "loss": 1.401, "step": 3494 }, { "epoch": 0.2739028213166144, "grad_norm": 2.461322546005249, "learning_rate": 9.890802106445059e-06, "loss": 1.3427, "step": 3495 }, { "epoch": 0.27398119122257053, "grad_norm": 2.3076913356781006, "learning_rate": 9.890738003669029e-06, "loss": 1.1851, "step": 3496 }, { "epoch": 0.27405956112852664, "grad_norm": 2.1944267749786377, "learning_rate": 9.890673882291142e-06, "loss": 1.2285, "step": 3497 }, { "epoch": 0.27413793103448275, "grad_norm": 2.867896318435669, "learning_rate": 9.890609742311646e-06, "loss": 1.6516, "step": 3498 }, { "epoch": 0.27421630094043886, "grad_norm": 4.265921115875244, "learning_rate": 9.89054558373078e-06, "loss": 1.7031, "step": 3499 }, { "epoch": 0.274294670846395, "grad_norm": 3.7028751373291016, "learning_rate": 9.89048140654879e-06, "loss": 1.2838, "step": 3500 }, { "epoch": 0.2743730407523511, "grad_norm": 3.63551926612854, "learning_rate": 9.890417210765923e-06, "loss": 1.109, "step": 3501 }, { "epoch": 0.2744514106583072, "grad_norm": 2.6283059120178223, "learning_rate": 9.89035299638242e-06, "loss": 1.6297, "step": 3502 }, { "epoch": 0.2745297805642633, "grad_norm": 3.7216413021087646, "learning_rate": 9.890288763398526e-06, "loss": 1.6227, "step": 3503 }, { "epoch": 0.2746081504702194, "grad_norm": 2.2907071113586426, "learning_rate": 9.890224511814484e-06, "loss": 1.4866, "step": 3504 }, { "epoch": 0.2746865203761755, "grad_norm": 2.957890033721924, "learning_rate": 9.890160241630542e-06, "loss": 1.3735, "step": 3505 }, { "epoch": 0.27476489028213164, "grad_norm": 2.1462695598602295, "learning_rate": 9.89009595284694e-06, "loss": 1.3431, "step": 3506 }, { "epoch": 0.2748432601880878, "grad_norm": 2.625126600265503, "learning_rate": 9.890031645463927e-06, "loss": 1.64, "step": 3507 }, { "epoch": 0.2749216300940439, "grad_norm": 2.895481824874878, "learning_rate": 9.889967319481744e-06, "loss": 1.3993, "step": 3508 }, { "epoch": 0.275, "grad_norm": 5.248561859130859, "learning_rate": 9.889902974900637e-06, "loss": 1.6272, "step": 3509 }, { "epoch": 0.27507836990595613, "grad_norm": 4.0323100090026855, "learning_rate": 9.88983861172085e-06, "loss": 1.7803, "step": 3510 }, { "epoch": 0.27515673981191224, "grad_norm": 2.814037799835205, "learning_rate": 9.88977422994263e-06, "loss": 1.0676, "step": 3511 }, { "epoch": 0.27523510971786835, "grad_norm": 9.295085906982422, "learning_rate": 9.889709829566219e-06, "loss": 1.6241, "step": 3512 }, { "epoch": 0.27531347962382446, "grad_norm": 3.5007541179656982, "learning_rate": 9.889645410591865e-06, "loss": 1.6071, "step": 3513 }, { "epoch": 0.2753918495297806, "grad_norm": 8.096741676330566, "learning_rate": 9.88958097301981e-06, "loss": 1.1764, "step": 3514 }, { "epoch": 0.2754702194357367, "grad_norm": 2.8353042602539062, "learning_rate": 9.8895165168503e-06, "loss": 1.2119, "step": 3515 }, { "epoch": 0.2755485893416928, "grad_norm": 2.5151703357696533, "learning_rate": 9.889452042083583e-06, "loss": 1.4281, "step": 3516 }, { "epoch": 0.2756269592476489, "grad_norm": 3.117521047592163, "learning_rate": 9.8893875487199e-06, "loss": 1.6557, "step": 3517 }, { "epoch": 0.275705329153605, "grad_norm": 3.4221017360687256, "learning_rate": 9.889323036759497e-06, "loss": 1.4335, "step": 3518 }, { "epoch": 0.2757836990595611, "grad_norm": 14.63820743560791, "learning_rate": 9.88925850620262e-06, "loss": 1.9179, "step": 3519 }, { "epoch": 0.27586206896551724, "grad_norm": 3.130462884902954, "learning_rate": 9.88919395704952e-06, "loss": 1.4789, "step": 3520 }, { "epoch": 0.27594043887147335, "grad_norm": 2.8908133506774902, "learning_rate": 9.889129389300433e-06, "loss": 1.4162, "step": 3521 }, { "epoch": 0.27601880877742946, "grad_norm": 2.743499279022217, "learning_rate": 9.88906480295561e-06, "loss": 1.2572, "step": 3522 }, { "epoch": 0.27609717868338557, "grad_norm": 3.516669273376465, "learning_rate": 9.889000198015295e-06, "loss": 1.8546, "step": 3523 }, { "epoch": 0.2761755485893417, "grad_norm": 4.367371082305908, "learning_rate": 9.888935574479736e-06, "loss": 1.6002, "step": 3524 }, { "epoch": 0.2762539184952978, "grad_norm": 2.107501983642578, "learning_rate": 9.888870932349176e-06, "loss": 0.7822, "step": 3525 }, { "epoch": 0.2763322884012539, "grad_norm": 2.7368805408477783, "learning_rate": 9.888806271623863e-06, "loss": 1.0421, "step": 3526 }, { "epoch": 0.27641065830721, "grad_norm": 5.474120616912842, "learning_rate": 9.88874159230404e-06, "loss": 2.1095, "step": 3527 }, { "epoch": 0.2764890282131661, "grad_norm": 2.694664716720581, "learning_rate": 9.888676894389956e-06, "loss": 1.3258, "step": 3528 }, { "epoch": 0.27656739811912223, "grad_norm": 2.945117950439453, "learning_rate": 9.888612177881857e-06, "loss": 1.2146, "step": 3529 }, { "epoch": 0.27664576802507834, "grad_norm": 3.457632541656494, "learning_rate": 9.888547442779989e-06, "loss": 1.3729, "step": 3530 }, { "epoch": 0.2767241379310345, "grad_norm": 3.3756721019744873, "learning_rate": 9.888482689084595e-06, "loss": 1.4381, "step": 3531 }, { "epoch": 0.2768025078369906, "grad_norm": 3.7136476039886475, "learning_rate": 9.888417916795924e-06, "loss": 1.3305, "step": 3532 }, { "epoch": 0.27688087774294673, "grad_norm": 2.129915714263916, "learning_rate": 9.888353125914226e-06, "loss": 1.4435, "step": 3533 }, { "epoch": 0.27695924764890284, "grad_norm": 2.789994478225708, "learning_rate": 9.888288316439739e-06, "loss": 1.5667, "step": 3534 }, { "epoch": 0.27703761755485895, "grad_norm": 2.7349679470062256, "learning_rate": 9.888223488372715e-06, "loss": 1.5885, "step": 3535 }, { "epoch": 0.27711598746081506, "grad_norm": 2.761308193206787, "learning_rate": 9.8881586417134e-06, "loss": 1.4609, "step": 3536 }, { "epoch": 0.27719435736677117, "grad_norm": 3.6703550815582275, "learning_rate": 9.888093776462042e-06, "loss": 1.2917, "step": 3537 }, { "epoch": 0.2772727272727273, "grad_norm": 4.58017635345459, "learning_rate": 9.888028892618884e-06, "loss": 1.5591, "step": 3538 }, { "epoch": 0.2773510971786834, "grad_norm": 4.365350246429443, "learning_rate": 9.887963990184175e-06, "loss": 1.1992, "step": 3539 }, { "epoch": 0.2774294670846395, "grad_norm": 2.7527782917022705, "learning_rate": 9.887899069158161e-06, "loss": 1.3696, "step": 3540 }, { "epoch": 0.2775078369905956, "grad_norm": 2.3156397342681885, "learning_rate": 9.88783412954109e-06, "loss": 1.634, "step": 3541 }, { "epoch": 0.2775862068965517, "grad_norm": 2.655796766281128, "learning_rate": 9.887769171333209e-06, "loss": 1.5441, "step": 3542 }, { "epoch": 0.27766457680250783, "grad_norm": 3.128523588180542, "learning_rate": 9.887704194534763e-06, "loss": 1.4491, "step": 3543 }, { "epoch": 0.27774294670846394, "grad_norm": 3.6070597171783447, "learning_rate": 9.887639199146004e-06, "loss": 0.6478, "step": 3544 }, { "epoch": 0.27782131661442006, "grad_norm": 2.7072250843048096, "learning_rate": 9.887574185167173e-06, "loss": 1.0313, "step": 3545 }, { "epoch": 0.27789968652037617, "grad_norm": 2.7303998470306396, "learning_rate": 9.887509152598523e-06, "loss": 1.1583, "step": 3546 }, { "epoch": 0.2779780564263323, "grad_norm": 3.199185848236084, "learning_rate": 9.887444101440296e-06, "loss": 1.5168, "step": 3547 }, { "epoch": 0.2780564263322884, "grad_norm": 2.2864675521850586, "learning_rate": 9.887379031692743e-06, "loss": 1.4095, "step": 3548 }, { "epoch": 0.2781347962382445, "grad_norm": 2.1272170543670654, "learning_rate": 9.88731394335611e-06, "loss": 1.294, "step": 3549 }, { "epoch": 0.2782131661442006, "grad_norm": 2.81487774848938, "learning_rate": 9.887248836430645e-06, "loss": 1.4217, "step": 3550 }, { "epoch": 0.2782915360501567, "grad_norm": 4.611172199249268, "learning_rate": 9.887183710916595e-06, "loss": 1.5748, "step": 3551 }, { "epoch": 0.27836990595611283, "grad_norm": 3.0124878883361816, "learning_rate": 9.88711856681421e-06, "loss": 1.3712, "step": 3552 }, { "epoch": 0.27844827586206894, "grad_norm": 3.7838711738586426, "learning_rate": 9.887053404123736e-06, "loss": 1.513, "step": 3553 }, { "epoch": 0.2785266457680251, "grad_norm": 2.0820889472961426, "learning_rate": 9.88698822284542e-06, "loss": 1.5076, "step": 3554 }, { "epoch": 0.2786050156739812, "grad_norm": 2.984226942062378, "learning_rate": 9.88692302297951e-06, "loss": 1.3936, "step": 3555 }, { "epoch": 0.2786833855799373, "grad_norm": 2.5455660820007324, "learning_rate": 9.886857804526258e-06, "loss": 1.0274, "step": 3556 }, { "epoch": 0.27876175548589344, "grad_norm": 3.0578982830047607, "learning_rate": 9.886792567485908e-06, "loss": 1.1898, "step": 3557 }, { "epoch": 0.27884012539184955, "grad_norm": 3.0336880683898926, "learning_rate": 9.886727311858709e-06, "loss": 1.6328, "step": 3558 }, { "epoch": 0.27891849529780566, "grad_norm": 5.032535552978516, "learning_rate": 9.886662037644908e-06, "loss": 1.4477, "step": 3559 }, { "epoch": 0.27899686520376177, "grad_norm": 2.1578736305236816, "learning_rate": 9.886596744844755e-06, "loss": 1.4111, "step": 3560 }, { "epoch": 0.2790752351097179, "grad_norm": 3.8676750659942627, "learning_rate": 9.886531433458497e-06, "loss": 0.9382, "step": 3561 }, { "epoch": 0.279153605015674, "grad_norm": 5.9816083908081055, "learning_rate": 9.886466103486385e-06, "loss": 1.5632, "step": 3562 }, { "epoch": 0.2792319749216301, "grad_norm": 2.329354763031006, "learning_rate": 9.886400754928664e-06, "loss": 1.3455, "step": 3563 }, { "epoch": 0.2793103448275862, "grad_norm": 2.5975208282470703, "learning_rate": 9.886335387785586e-06, "loss": 1.3366, "step": 3564 }, { "epoch": 0.2793887147335423, "grad_norm": 4.875551223754883, "learning_rate": 9.886270002057399e-06, "loss": 1.3048, "step": 3565 }, { "epoch": 0.27946708463949843, "grad_norm": 2.346522331237793, "learning_rate": 9.886204597744349e-06, "loss": 1.479, "step": 3566 }, { "epoch": 0.27954545454545454, "grad_norm": 3.7712643146514893, "learning_rate": 9.886139174846686e-06, "loss": 1.4415, "step": 3567 }, { "epoch": 0.27962382445141065, "grad_norm": 11.657430648803711, "learning_rate": 9.886073733364659e-06, "loss": 2.1242, "step": 3568 }, { "epoch": 0.27970219435736676, "grad_norm": 3.572985887527466, "learning_rate": 9.886008273298518e-06, "loss": 1.0328, "step": 3569 }, { "epoch": 0.2797805642633229, "grad_norm": 2.843050956726074, "learning_rate": 9.885942794648511e-06, "loss": 1.7063, "step": 3570 }, { "epoch": 0.279858934169279, "grad_norm": 2.4131453037261963, "learning_rate": 9.885877297414888e-06, "loss": 1.4302, "step": 3571 }, { "epoch": 0.2799373040752351, "grad_norm": 7.264505386352539, "learning_rate": 9.885811781597897e-06, "loss": 1.6666, "step": 3572 }, { "epoch": 0.2800156739811912, "grad_norm": 2.579887628555298, "learning_rate": 9.885746247197787e-06, "loss": 1.1242, "step": 3573 }, { "epoch": 0.2800940438871473, "grad_norm": 2.0304954051971436, "learning_rate": 9.885680694214808e-06, "loss": 1.2657, "step": 3574 }, { "epoch": 0.2801724137931034, "grad_norm": 5.2248125076293945, "learning_rate": 9.885615122649209e-06, "loss": 1.4706, "step": 3575 }, { "epoch": 0.28025078369905954, "grad_norm": 2.7841928005218506, "learning_rate": 9.88554953250124e-06, "loss": 2.0195, "step": 3576 }, { "epoch": 0.28032915360501565, "grad_norm": 2.704968214035034, "learning_rate": 9.885483923771148e-06, "loss": 1.1302, "step": 3577 }, { "epoch": 0.2804075235109718, "grad_norm": 3.7984507083892822, "learning_rate": 9.885418296459186e-06, "loss": 1.1478, "step": 3578 }, { "epoch": 0.2804858934169279, "grad_norm": 4.84363317489624, "learning_rate": 9.885352650565602e-06, "loss": 1.4554, "step": 3579 }, { "epoch": 0.28056426332288403, "grad_norm": 8.610069274902344, "learning_rate": 9.885286986090646e-06, "loss": 0.8216, "step": 3580 }, { "epoch": 0.28064263322884014, "grad_norm": 2.348457098007202, "learning_rate": 9.885221303034567e-06, "loss": 1.3692, "step": 3581 }, { "epoch": 0.28072100313479625, "grad_norm": 5.4773101806640625, "learning_rate": 9.885155601397616e-06, "loss": 1.8932, "step": 3582 }, { "epoch": 0.28079937304075236, "grad_norm": 3.112748861312866, "learning_rate": 9.885089881180042e-06, "loss": 1.7589, "step": 3583 }, { "epoch": 0.2808777429467085, "grad_norm": 2.214329242706299, "learning_rate": 9.885024142382096e-06, "loss": 1.2556, "step": 3584 }, { "epoch": 0.2809561128526646, "grad_norm": 3.765813112258911, "learning_rate": 9.884958385004026e-06, "loss": 1.5018, "step": 3585 }, { "epoch": 0.2810344827586207, "grad_norm": 3.574202060699463, "learning_rate": 9.884892609046083e-06, "loss": 1.2008, "step": 3586 }, { "epoch": 0.2811128526645768, "grad_norm": 2.8150172233581543, "learning_rate": 9.884826814508519e-06, "loss": 1.3608, "step": 3587 }, { "epoch": 0.2811912225705329, "grad_norm": 3.1828455924987793, "learning_rate": 9.884761001391583e-06, "loss": 1.3315, "step": 3588 }, { "epoch": 0.281269592476489, "grad_norm": 3.5530121326446533, "learning_rate": 9.884695169695524e-06, "loss": 1.5275, "step": 3589 }, { "epoch": 0.28134796238244514, "grad_norm": 2.8434698581695557, "learning_rate": 9.884629319420593e-06, "loss": 1.5593, "step": 3590 }, { "epoch": 0.28142633228840125, "grad_norm": 4.678114414215088, "learning_rate": 9.884563450567042e-06, "loss": 1.5002, "step": 3591 }, { "epoch": 0.28150470219435736, "grad_norm": 2.579761505126953, "learning_rate": 9.88449756313512e-06, "loss": 1.5724, "step": 3592 }, { "epoch": 0.28158307210031347, "grad_norm": 4.504692554473877, "learning_rate": 9.88443165712508e-06, "loss": 1.424, "step": 3593 }, { "epoch": 0.2816614420062696, "grad_norm": 5.539649963378906, "learning_rate": 9.884365732537168e-06, "loss": 1.1935, "step": 3594 }, { "epoch": 0.2817398119122257, "grad_norm": 10.731175422668457, "learning_rate": 9.88429978937164e-06, "loss": 1.3753, "step": 3595 }, { "epoch": 0.2818181818181818, "grad_norm": 2.965749979019165, "learning_rate": 9.884233827628742e-06, "loss": 0.7461, "step": 3596 }, { "epoch": 0.2818965517241379, "grad_norm": 2.3668172359466553, "learning_rate": 9.884167847308728e-06, "loss": 1.4354, "step": 3597 }, { "epoch": 0.281974921630094, "grad_norm": 4.014881134033203, "learning_rate": 9.88410184841185e-06, "loss": 1.8144, "step": 3598 }, { "epoch": 0.28205329153605013, "grad_norm": 7.9822845458984375, "learning_rate": 9.884035830938354e-06, "loss": 0.6431, "step": 3599 }, { "epoch": 0.28213166144200624, "grad_norm": 2.765110731124878, "learning_rate": 9.883969794888497e-06, "loss": 1.3005, "step": 3600 }, { "epoch": 0.2822100313479624, "grad_norm": 2.1808736324310303, "learning_rate": 9.883903740262527e-06, "loss": 1.2193, "step": 3601 }, { "epoch": 0.2822884012539185, "grad_norm": 3.993992328643799, "learning_rate": 9.883837667060694e-06, "loss": 1.6158, "step": 3602 }, { "epoch": 0.28236677115987463, "grad_norm": 5.906666278839111, "learning_rate": 9.883771575283254e-06, "loss": 1.4653, "step": 3603 }, { "epoch": 0.28244514106583074, "grad_norm": 6.387022495269775, "learning_rate": 9.883705464930452e-06, "loss": 1.5264, "step": 3604 }, { "epoch": 0.28252351097178685, "grad_norm": 1.992735743522644, "learning_rate": 9.883639336002545e-06, "loss": 1.3088, "step": 3605 }, { "epoch": 0.28260188087774296, "grad_norm": 12.840409278869629, "learning_rate": 9.883573188499782e-06, "loss": 1.2338, "step": 3606 }, { "epoch": 0.28268025078369907, "grad_norm": 2.3594324588775635, "learning_rate": 9.883507022422414e-06, "loss": 1.3304, "step": 3607 }, { "epoch": 0.2827586206896552, "grad_norm": 4.345395565032959, "learning_rate": 9.883440837770695e-06, "loss": 1.0207, "step": 3608 }, { "epoch": 0.2828369905956113, "grad_norm": 2.661794662475586, "learning_rate": 9.883374634544874e-06, "loss": 1.5744, "step": 3609 }, { "epoch": 0.2829153605015674, "grad_norm": 2.2281782627105713, "learning_rate": 9.883308412745206e-06, "loss": 1.3385, "step": 3610 }, { "epoch": 0.2829937304075235, "grad_norm": 4.816803932189941, "learning_rate": 9.883242172371939e-06, "loss": 1.5051, "step": 3611 }, { "epoch": 0.2830721003134796, "grad_norm": 3.661299467086792, "learning_rate": 9.883175913425328e-06, "loss": 1.1979, "step": 3612 }, { "epoch": 0.28315047021943573, "grad_norm": 5.219409465789795, "learning_rate": 9.883109635905623e-06, "loss": 1.1696, "step": 3613 }, { "epoch": 0.28322884012539185, "grad_norm": 4.468589782714844, "learning_rate": 9.883043339813079e-06, "loss": 1.4743, "step": 3614 }, { "epoch": 0.28330721003134796, "grad_norm": 2.5906453132629395, "learning_rate": 9.882977025147945e-06, "loss": 1.4568, "step": 3615 }, { "epoch": 0.28338557993730407, "grad_norm": 2.9229509830474854, "learning_rate": 9.882910691910475e-06, "loss": 1.2634, "step": 3616 }, { "epoch": 0.2834639498432602, "grad_norm": 2.485565185546875, "learning_rate": 9.88284434010092e-06, "loss": 1.3461, "step": 3617 }, { "epoch": 0.2835423197492163, "grad_norm": 2.8404672145843506, "learning_rate": 9.882777969719532e-06, "loss": 1.1073, "step": 3618 }, { "epoch": 0.2836206896551724, "grad_norm": 2.392685651779175, "learning_rate": 9.882711580766566e-06, "loss": 1.3584, "step": 3619 }, { "epoch": 0.2836990595611285, "grad_norm": 2.6546342372894287, "learning_rate": 9.882645173242274e-06, "loss": 1.2356, "step": 3620 }, { "epoch": 0.2837774294670846, "grad_norm": 4.632045745849609, "learning_rate": 9.882578747146908e-06, "loss": 1.8753, "step": 3621 }, { "epoch": 0.28385579937304073, "grad_norm": 3.1682639122009277, "learning_rate": 9.882512302480717e-06, "loss": 1.6069, "step": 3622 }, { "epoch": 0.28393416927899684, "grad_norm": 3.4214937686920166, "learning_rate": 9.88244583924396e-06, "loss": 1.4494, "step": 3623 }, { "epoch": 0.284012539184953, "grad_norm": 2.6364023685455322, "learning_rate": 9.882379357436885e-06, "loss": 1.4189, "step": 3624 }, { "epoch": 0.2840909090909091, "grad_norm": 2.695908308029175, "learning_rate": 9.882312857059749e-06, "loss": 1.5354, "step": 3625 }, { "epoch": 0.2841692789968652, "grad_norm": 4.0749311447143555, "learning_rate": 9.8822463381128e-06, "loss": 1.582, "step": 3626 }, { "epoch": 0.28424764890282134, "grad_norm": 2.9766249656677246, "learning_rate": 9.882179800596295e-06, "loss": 1.307, "step": 3627 }, { "epoch": 0.28432601880877745, "grad_norm": 3.3829994201660156, "learning_rate": 9.882113244510486e-06, "loss": 1.647, "step": 3628 }, { "epoch": 0.28440438871473356, "grad_norm": 4.28725528717041, "learning_rate": 9.882046669855624e-06, "loss": 1.3882, "step": 3629 }, { "epoch": 0.28448275862068967, "grad_norm": 2.9649453163146973, "learning_rate": 9.881980076631965e-06, "loss": 1.8909, "step": 3630 }, { "epoch": 0.2845611285266458, "grad_norm": 7.2599029541015625, "learning_rate": 9.881913464839762e-06, "loss": 1.6566, "step": 3631 }, { "epoch": 0.2846394984326019, "grad_norm": 3.149113416671753, "learning_rate": 9.881846834479267e-06, "loss": 1.5527, "step": 3632 }, { "epoch": 0.284717868338558, "grad_norm": 2.2432446479797363, "learning_rate": 9.881780185550735e-06, "loss": 1.2663, "step": 3633 }, { "epoch": 0.2847962382445141, "grad_norm": 2.5200328826904297, "learning_rate": 9.881713518054418e-06, "loss": 1.6295, "step": 3634 }, { "epoch": 0.2848746081504702, "grad_norm": 2.605595827102661, "learning_rate": 9.88164683199057e-06, "loss": 1.2542, "step": 3635 }, { "epoch": 0.28495297805642633, "grad_norm": 5.461869239807129, "learning_rate": 9.881580127359444e-06, "loss": 1.8815, "step": 3636 }, { "epoch": 0.28503134796238244, "grad_norm": 3.014397144317627, "learning_rate": 9.881513404161295e-06, "loss": 1.5906, "step": 3637 }, { "epoch": 0.28510971786833855, "grad_norm": 4.649600028991699, "learning_rate": 9.88144666239638e-06, "loss": 1.2142, "step": 3638 }, { "epoch": 0.28518808777429466, "grad_norm": 2.50531005859375, "learning_rate": 9.881379902064945e-06, "loss": 1.4062, "step": 3639 }, { "epoch": 0.2852664576802508, "grad_norm": 3.14198899269104, "learning_rate": 9.88131312316725e-06, "loss": 1.204, "step": 3640 }, { "epoch": 0.2853448275862069, "grad_norm": 3.169656276702881, "learning_rate": 9.881246325703544e-06, "loss": 1.4577, "step": 3641 }, { "epoch": 0.285423197492163, "grad_norm": 3.6740801334381104, "learning_rate": 9.881179509674087e-06, "loss": 1.6885, "step": 3642 }, { "epoch": 0.2855015673981191, "grad_norm": 14.37088394165039, "learning_rate": 9.88111267507913e-06, "loss": 2.0106, "step": 3643 }, { "epoch": 0.2855799373040752, "grad_norm": 16.364044189453125, "learning_rate": 9.881045821918926e-06, "loss": 1.0121, "step": 3644 }, { "epoch": 0.2856583072100313, "grad_norm": 5.120818614959717, "learning_rate": 9.880978950193733e-06, "loss": 1.7022, "step": 3645 }, { "epoch": 0.28573667711598744, "grad_norm": 3.6874101161956787, "learning_rate": 9.880912059903801e-06, "loss": 1.7303, "step": 3646 }, { "epoch": 0.28581504702194355, "grad_norm": 3.5805108547210693, "learning_rate": 9.880845151049388e-06, "loss": 1.1037, "step": 3647 }, { "epoch": 0.2858934169278997, "grad_norm": 3.504401922225952, "learning_rate": 9.880778223630747e-06, "loss": 1.8772, "step": 3648 }, { "epoch": 0.2859717868338558, "grad_norm": 2.833179235458374, "learning_rate": 9.880711277648132e-06, "loss": 1.9068, "step": 3649 }, { "epoch": 0.28605015673981193, "grad_norm": 3.26249623298645, "learning_rate": 9.880644313101797e-06, "loss": 1.5012, "step": 3650 }, { "epoch": 0.28612852664576804, "grad_norm": 1.8364722728729248, "learning_rate": 9.880577329992e-06, "loss": 1.6763, "step": 3651 }, { "epoch": 0.28620689655172415, "grad_norm": 3.9784793853759766, "learning_rate": 9.880510328318993e-06, "loss": 1.2287, "step": 3652 }, { "epoch": 0.28628526645768027, "grad_norm": 3.1888816356658936, "learning_rate": 9.88044330808303e-06, "loss": 1.8668, "step": 3653 }, { "epoch": 0.2863636363636364, "grad_norm": 4.464120864868164, "learning_rate": 9.88037626928437e-06, "loss": 1.0915, "step": 3654 }, { "epoch": 0.2864420062695925, "grad_norm": 4.409801959991455, "learning_rate": 9.880309211923263e-06, "loss": 1.9724, "step": 3655 }, { "epoch": 0.2865203761755486, "grad_norm": 3.0651683807373047, "learning_rate": 9.88024213599997e-06, "loss": 1.1713, "step": 3656 }, { "epoch": 0.2865987460815047, "grad_norm": 3.3702521324157715, "learning_rate": 9.880175041514739e-06, "loss": 1.7004, "step": 3657 }, { "epoch": 0.2866771159874608, "grad_norm": 7.194216728210449, "learning_rate": 9.88010792846783e-06, "loss": 1.4234, "step": 3658 }, { "epoch": 0.28675548589341693, "grad_norm": 4.754070281982422, "learning_rate": 9.880040796859496e-06, "loss": 1.5864, "step": 3659 }, { "epoch": 0.28683385579937304, "grad_norm": 5.322964191436768, "learning_rate": 9.879973646689994e-06, "loss": 1.4007, "step": 3660 }, { "epoch": 0.28691222570532915, "grad_norm": 3.404700994491577, "learning_rate": 9.87990647795958e-06, "loss": 1.4287, "step": 3661 }, { "epoch": 0.28699059561128526, "grad_norm": 2.9362244606018066, "learning_rate": 9.879839290668507e-06, "loss": 1.0239, "step": 3662 }, { "epoch": 0.28706896551724137, "grad_norm": 2.730037212371826, "learning_rate": 9.879772084817033e-06, "loss": 1.6356, "step": 3663 }, { "epoch": 0.2871473354231975, "grad_norm": 3.716935157775879, "learning_rate": 9.879704860405411e-06, "loss": 1.137, "step": 3664 }, { "epoch": 0.2872257053291536, "grad_norm": 2.593998670578003, "learning_rate": 9.879637617433897e-06, "loss": 0.7018, "step": 3665 }, { "epoch": 0.2873040752351097, "grad_norm": 5.244908332824707, "learning_rate": 9.87957035590275e-06, "loss": 1.2417, "step": 3666 }, { "epoch": 0.2873824451410658, "grad_norm": 2.430298089981079, "learning_rate": 9.879503075812223e-06, "loss": 1.3752, "step": 3667 }, { "epoch": 0.2874608150470219, "grad_norm": 3.169769525527954, "learning_rate": 9.879435777162573e-06, "loss": 1.4886, "step": 3668 }, { "epoch": 0.28753918495297803, "grad_norm": 4.209900856018066, "learning_rate": 9.879368459954055e-06, "loss": 1.4133, "step": 3669 }, { "epoch": 0.28761755485893414, "grad_norm": 3.4945311546325684, "learning_rate": 9.879301124186926e-06, "loss": 1.591, "step": 3670 }, { "epoch": 0.2876959247648903, "grad_norm": 3.040902614593506, "learning_rate": 9.879233769861441e-06, "loss": 1.5816, "step": 3671 }, { "epoch": 0.2877742946708464, "grad_norm": 3.3518567085266113, "learning_rate": 9.879166396977858e-06, "loss": 1.7651, "step": 3672 }, { "epoch": 0.28785266457680253, "grad_norm": 3.2240946292877197, "learning_rate": 9.87909900553643e-06, "loss": 1.1324, "step": 3673 }, { "epoch": 0.28793103448275864, "grad_norm": 3.0175609588623047, "learning_rate": 9.879031595537418e-06, "loss": 1.6178, "step": 3674 }, { "epoch": 0.28800940438871475, "grad_norm": 1.9984076023101807, "learning_rate": 9.878964166981073e-06, "loss": 1.4094, "step": 3675 }, { "epoch": 0.28808777429467086, "grad_norm": 4.226181983947754, "learning_rate": 9.878896719867657e-06, "loss": 1.2325, "step": 3676 }, { "epoch": 0.288166144200627, "grad_norm": 4.076443195343018, "learning_rate": 9.878829254197422e-06, "loss": 1.298, "step": 3677 }, { "epoch": 0.2882445141065831, "grad_norm": 2.4352924823760986, "learning_rate": 9.878761769970628e-06, "loss": 1.2006, "step": 3678 }, { "epoch": 0.2883228840125392, "grad_norm": 1.9345741271972656, "learning_rate": 9.87869426718753e-06, "loss": 1.5569, "step": 3679 }, { "epoch": 0.2884012539184953, "grad_norm": 3.3599772453308105, "learning_rate": 9.878626745848384e-06, "loss": 1.026, "step": 3680 }, { "epoch": 0.2884796238244514, "grad_norm": 2.8415818214416504, "learning_rate": 9.878559205953447e-06, "loss": 1.4429, "step": 3681 }, { "epoch": 0.2885579937304075, "grad_norm": 2.5148024559020996, "learning_rate": 9.878491647502979e-06, "loss": 1.5551, "step": 3682 }, { "epoch": 0.28863636363636364, "grad_norm": 3.7165658473968506, "learning_rate": 9.878424070497232e-06, "loss": 1.3855, "step": 3683 }, { "epoch": 0.28871473354231975, "grad_norm": 3.2101407051086426, "learning_rate": 9.878356474936466e-06, "loss": 1.6488, "step": 3684 }, { "epoch": 0.28879310344827586, "grad_norm": 4.045097827911377, "learning_rate": 9.87828886082094e-06, "loss": 1.6931, "step": 3685 }, { "epoch": 0.28887147335423197, "grad_norm": 3.145533323287964, "learning_rate": 9.878221228150907e-06, "loss": 1.1309, "step": 3686 }, { "epoch": 0.2889498432601881, "grad_norm": 3.8809797763824463, "learning_rate": 9.878153576926625e-06, "loss": 1.9078, "step": 3687 }, { "epoch": 0.2890282131661442, "grad_norm": 2.807359218597412, "learning_rate": 9.878085907148353e-06, "loss": 1.3501, "step": 3688 }, { "epoch": 0.2891065830721003, "grad_norm": 2.1111910343170166, "learning_rate": 9.878018218816349e-06, "loss": 1.379, "step": 3689 }, { "epoch": 0.2891849529780564, "grad_norm": 3.067488670349121, "learning_rate": 9.877950511930868e-06, "loss": 1.4173, "step": 3690 }, { "epoch": 0.2892633228840125, "grad_norm": 3.501567840576172, "learning_rate": 9.877882786492168e-06, "loss": 1.3884, "step": 3691 }, { "epoch": 0.28934169278996863, "grad_norm": 2.466956853866577, "learning_rate": 9.87781504250051e-06, "loss": 1.3741, "step": 3692 }, { "epoch": 0.28942006269592474, "grad_norm": 2.171945810317993, "learning_rate": 9.877747279956146e-06, "loss": 1.3278, "step": 3693 }, { "epoch": 0.28949843260188085, "grad_norm": 9.79740047454834, "learning_rate": 9.877679498859339e-06, "loss": 1.0251, "step": 3694 }, { "epoch": 0.289576802507837, "grad_norm": 3.2443511486053467, "learning_rate": 9.877611699210343e-06, "loss": 1.488, "step": 3695 }, { "epoch": 0.2896551724137931, "grad_norm": 9.310868263244629, "learning_rate": 9.87754388100942e-06, "loss": 1.5056, "step": 3696 }, { "epoch": 0.28973354231974924, "grad_norm": 3.2739925384521484, "learning_rate": 9.877476044256822e-06, "loss": 1.3006, "step": 3697 }, { "epoch": 0.28981191222570535, "grad_norm": 6.942071914672852, "learning_rate": 9.877408188952813e-06, "loss": 1.3532, "step": 3698 }, { "epoch": 0.28989028213166146, "grad_norm": 2.512604236602783, "learning_rate": 9.877340315097646e-06, "loss": 1.4487, "step": 3699 }, { "epoch": 0.28996865203761757, "grad_norm": 3.4659464359283447, "learning_rate": 9.877272422691583e-06, "loss": 1.1103, "step": 3700 }, { "epoch": 0.2900470219435737, "grad_norm": 3.757681131362915, "learning_rate": 9.87720451173488e-06, "loss": 1.2984, "step": 3701 }, { "epoch": 0.2901253918495298, "grad_norm": 2.603938102722168, "learning_rate": 9.877136582227797e-06, "loss": 1.4812, "step": 3702 }, { "epoch": 0.2902037617554859, "grad_norm": 2.9488203525543213, "learning_rate": 9.877068634170593e-06, "loss": 1.0336, "step": 3703 }, { "epoch": 0.290282131661442, "grad_norm": 4.571272850036621, "learning_rate": 9.877000667563522e-06, "loss": 1.643, "step": 3704 }, { "epoch": 0.2903605015673981, "grad_norm": 3.028095245361328, "learning_rate": 9.876932682406847e-06, "loss": 1.1156, "step": 3705 }, { "epoch": 0.29043887147335423, "grad_norm": 3.567714214324951, "learning_rate": 9.876864678700824e-06, "loss": 1.5717, "step": 3706 }, { "epoch": 0.29051724137931034, "grad_norm": 2.750465154647827, "learning_rate": 9.876796656445714e-06, "loss": 1.6326, "step": 3707 }, { "epoch": 0.29059561128526645, "grad_norm": 3.235062599182129, "learning_rate": 9.876728615641772e-06, "loss": 1.5343, "step": 3708 }, { "epoch": 0.29067398119122256, "grad_norm": 10.303228378295898, "learning_rate": 9.876660556289263e-06, "loss": 1.3707, "step": 3709 }, { "epoch": 0.2907523510971787, "grad_norm": 2.452380895614624, "learning_rate": 9.87659247838844e-06, "loss": 1.0949, "step": 3710 }, { "epoch": 0.2908307210031348, "grad_norm": 3.842477560043335, "learning_rate": 9.876524381939563e-06, "loss": 0.8164, "step": 3711 }, { "epoch": 0.2909090909090909, "grad_norm": 2.5841774940490723, "learning_rate": 9.876456266942893e-06, "loss": 1.2274, "step": 3712 }, { "epoch": 0.290987460815047, "grad_norm": 2.346574306488037, "learning_rate": 9.876388133398688e-06, "loss": 1.3927, "step": 3713 }, { "epoch": 0.2910658307210031, "grad_norm": 3.2765583992004395, "learning_rate": 9.876319981307208e-06, "loss": 1.4482, "step": 3714 }, { "epoch": 0.2911442006269592, "grad_norm": 3.7400097846984863, "learning_rate": 9.876251810668711e-06, "loss": 1.528, "step": 3715 }, { "epoch": 0.29122257053291534, "grad_norm": 10.84585952758789, "learning_rate": 9.876183621483456e-06, "loss": 0.7659, "step": 3716 }, { "epoch": 0.29130094043887145, "grad_norm": 3.8477275371551514, "learning_rate": 9.876115413751703e-06, "loss": 1.4069, "step": 3717 }, { "epoch": 0.2913793103448276, "grad_norm": 2.5709228515625, "learning_rate": 9.876047187473713e-06, "loss": 0.9962, "step": 3718 }, { "epoch": 0.2914576802507837, "grad_norm": 3.3432910442352295, "learning_rate": 9.875978942649743e-06, "loss": 1.7826, "step": 3719 }, { "epoch": 0.29153605015673983, "grad_norm": 2.9118378162384033, "learning_rate": 9.875910679280053e-06, "loss": 1.4568, "step": 3720 }, { "epoch": 0.29161442006269594, "grad_norm": 2.5786402225494385, "learning_rate": 9.875842397364903e-06, "loss": 1.4801, "step": 3721 }, { "epoch": 0.29169278996865206, "grad_norm": 4.451971054077148, "learning_rate": 9.875774096904555e-06, "loss": 1.8377, "step": 3722 }, { "epoch": 0.29177115987460817, "grad_norm": 2.615623950958252, "learning_rate": 9.875705777899265e-06, "loss": 1.7841, "step": 3723 }, { "epoch": 0.2918495297805643, "grad_norm": 4.13932466506958, "learning_rate": 9.875637440349295e-06, "loss": 1.3521, "step": 3724 }, { "epoch": 0.2919278996865204, "grad_norm": 2.3236777782440186, "learning_rate": 9.875569084254904e-06, "loss": 1.5149, "step": 3725 }, { "epoch": 0.2920062695924765, "grad_norm": 6.602819919586182, "learning_rate": 9.875500709616352e-06, "loss": 1.349, "step": 3726 }, { "epoch": 0.2920846394984326, "grad_norm": 3.0060067176818848, "learning_rate": 9.875432316433902e-06, "loss": 1.4268, "step": 3727 }, { "epoch": 0.2921630094043887, "grad_norm": 2.942744493484497, "learning_rate": 9.87536390470781e-06, "loss": 1.5403, "step": 3728 }, { "epoch": 0.29224137931034483, "grad_norm": 3.922762870788574, "learning_rate": 9.875295474438337e-06, "loss": 1.5651, "step": 3729 }, { "epoch": 0.29231974921630094, "grad_norm": 2.505443811416626, "learning_rate": 9.875227025625744e-06, "loss": 1.3966, "step": 3730 }, { "epoch": 0.29239811912225705, "grad_norm": 2.783521890640259, "learning_rate": 9.875158558270293e-06, "loss": 1.3537, "step": 3731 }, { "epoch": 0.29247648902821316, "grad_norm": 3.794476270675659, "learning_rate": 9.875090072372242e-06, "loss": 0.9698, "step": 3732 }, { "epoch": 0.29255485893416927, "grad_norm": 2.286759614944458, "learning_rate": 9.875021567931853e-06, "loss": 1.3267, "step": 3733 }, { "epoch": 0.2926332288401254, "grad_norm": 3.167560338973999, "learning_rate": 9.874953044949387e-06, "loss": 1.4658, "step": 3734 }, { "epoch": 0.2927115987460815, "grad_norm": 3.009916067123413, "learning_rate": 9.874884503425102e-06, "loss": 1.6217, "step": 3735 }, { "epoch": 0.2927899686520376, "grad_norm": 2.2729923725128174, "learning_rate": 9.874815943359259e-06, "loss": 1.4248, "step": 3736 }, { "epoch": 0.2928683385579937, "grad_norm": 2.502378225326538, "learning_rate": 9.874747364752122e-06, "loss": 1.2529, "step": 3737 }, { "epoch": 0.2929467084639498, "grad_norm": 3.9648561477661133, "learning_rate": 9.874678767603949e-06, "loss": 2.0198, "step": 3738 }, { "epoch": 0.29302507836990593, "grad_norm": 4.077919006347656, "learning_rate": 9.874610151915003e-06, "loss": 1.5218, "step": 3739 }, { "epoch": 0.29310344827586204, "grad_norm": 2.0832831859588623, "learning_rate": 9.874541517685542e-06, "loss": 1.0658, "step": 3740 }, { "epoch": 0.29318181818181815, "grad_norm": 2.069944381713867, "learning_rate": 9.874472864915829e-06, "loss": 1.198, "step": 3741 }, { "epoch": 0.2932601880877743, "grad_norm": 2.422687530517578, "learning_rate": 9.874404193606127e-06, "loss": 1.6743, "step": 3742 }, { "epoch": 0.29333855799373043, "grad_norm": 3.10927414894104, "learning_rate": 9.874335503756693e-06, "loss": 1.516, "step": 3743 }, { "epoch": 0.29341692789968654, "grad_norm": 2.0978240966796875, "learning_rate": 9.874266795367791e-06, "loss": 1.3567, "step": 3744 }, { "epoch": 0.29349529780564265, "grad_norm": 7.243281364440918, "learning_rate": 9.874198068439682e-06, "loss": 0.9369, "step": 3745 }, { "epoch": 0.29357366771159876, "grad_norm": 3.235501527786255, "learning_rate": 9.874129322972626e-06, "loss": 1.2424, "step": 3746 }, { "epoch": 0.2936520376175549, "grad_norm": 2.564995527267456, "learning_rate": 9.874060558966888e-06, "loss": 1.6001, "step": 3747 }, { "epoch": 0.293730407523511, "grad_norm": 2.194822072982788, "learning_rate": 9.873991776422723e-06, "loss": 1.5575, "step": 3748 }, { "epoch": 0.2938087774294671, "grad_norm": 3.6452157497406006, "learning_rate": 9.8739229753404e-06, "loss": 1.1339, "step": 3749 }, { "epoch": 0.2938871473354232, "grad_norm": 3.0170536041259766, "learning_rate": 9.873854155720176e-06, "loss": 1.6466, "step": 3750 }, { "epoch": 0.2939655172413793, "grad_norm": 2.637744426727295, "learning_rate": 9.873785317562314e-06, "loss": 1.1749, "step": 3751 }, { "epoch": 0.2940438871473354, "grad_norm": 2.923717975616455, "learning_rate": 9.873716460867078e-06, "loss": 1.2668, "step": 3752 }, { "epoch": 0.29412225705329154, "grad_norm": 5.082257270812988, "learning_rate": 9.873647585634727e-06, "loss": 1.3731, "step": 3753 }, { "epoch": 0.29420062695924765, "grad_norm": 4.821718215942383, "learning_rate": 9.873578691865522e-06, "loss": 0.9142, "step": 3754 }, { "epoch": 0.29427899686520376, "grad_norm": 2.739267349243164, "learning_rate": 9.873509779559728e-06, "loss": 0.953, "step": 3755 }, { "epoch": 0.29435736677115987, "grad_norm": 10.82443618774414, "learning_rate": 9.873440848717606e-06, "loss": 1.8151, "step": 3756 }, { "epoch": 0.294435736677116, "grad_norm": 3.0360801219940186, "learning_rate": 9.873371899339419e-06, "loss": 0.9297, "step": 3757 }, { "epoch": 0.2945141065830721, "grad_norm": 11.17677116394043, "learning_rate": 9.873302931425428e-06, "loss": 0.6955, "step": 3758 }, { "epoch": 0.2945924764890282, "grad_norm": 2.4414548873901367, "learning_rate": 9.873233944975895e-06, "loss": 1.5534, "step": 3759 }, { "epoch": 0.2946708463949843, "grad_norm": 3.139979362487793, "learning_rate": 9.873164939991085e-06, "loss": 0.8843, "step": 3760 }, { "epoch": 0.2947492163009404, "grad_norm": 2.421962022781372, "learning_rate": 9.873095916471256e-06, "loss": 1.5245, "step": 3761 }, { "epoch": 0.29482758620689653, "grad_norm": 3.0694007873535156, "learning_rate": 9.873026874416675e-06, "loss": 1.3243, "step": 3762 }, { "epoch": 0.29490595611285264, "grad_norm": 3.941289186477661, "learning_rate": 9.872957813827602e-06, "loss": 1.3791, "step": 3763 }, { "epoch": 0.29498432601880875, "grad_norm": 3.471036195755005, "learning_rate": 9.872888734704302e-06, "loss": 1.5799, "step": 3764 }, { "epoch": 0.2950626959247649, "grad_norm": 3.198441982269287, "learning_rate": 9.872819637047034e-06, "loss": 1.6019, "step": 3765 }, { "epoch": 0.295141065830721, "grad_norm": 2.4619743824005127, "learning_rate": 9.872750520856063e-06, "loss": 1.3091, "step": 3766 }, { "epoch": 0.29521943573667714, "grad_norm": 2.313148260116577, "learning_rate": 9.872681386131652e-06, "loss": 1.6201, "step": 3767 }, { "epoch": 0.29529780564263325, "grad_norm": 3.4038727283477783, "learning_rate": 9.872612232874066e-06, "loss": 1.1797, "step": 3768 }, { "epoch": 0.29537617554858936, "grad_norm": 11.548521041870117, "learning_rate": 9.872543061083563e-06, "loss": 0.7437, "step": 3769 }, { "epoch": 0.29545454545454547, "grad_norm": 13.709024429321289, "learning_rate": 9.872473870760409e-06, "loss": 1.2522, "step": 3770 }, { "epoch": 0.2955329153605016, "grad_norm": 2.7005231380462646, "learning_rate": 9.872404661904869e-06, "loss": 1.6127, "step": 3771 }, { "epoch": 0.2956112852664577, "grad_norm": 4.933236122131348, "learning_rate": 9.872335434517203e-06, "loss": 1.3298, "step": 3772 }, { "epoch": 0.2956896551724138, "grad_norm": 3.0444998741149902, "learning_rate": 9.872266188597675e-06, "loss": 1.6175, "step": 3773 }, { "epoch": 0.2957680250783699, "grad_norm": 3.217466354370117, "learning_rate": 9.872196924146551e-06, "loss": 1.4694, "step": 3774 }, { "epoch": 0.295846394984326, "grad_norm": 4.7450103759765625, "learning_rate": 9.87212764116409e-06, "loss": 1.3847, "step": 3775 }, { "epoch": 0.29592476489028213, "grad_norm": 2.697896957397461, "learning_rate": 9.872058339650559e-06, "loss": 1.2207, "step": 3776 }, { "epoch": 0.29600313479623824, "grad_norm": 7.085564136505127, "learning_rate": 9.87198901960622e-06, "loss": 1.279, "step": 3777 }, { "epoch": 0.29608150470219435, "grad_norm": 2.695828914642334, "learning_rate": 9.871919681031339e-06, "loss": 0.8893, "step": 3778 }, { "epoch": 0.29615987460815046, "grad_norm": 2.983431100845337, "learning_rate": 9.871850323926178e-06, "loss": 1.4276, "step": 3779 }, { "epoch": 0.2962382445141066, "grad_norm": 3.861128091812134, "learning_rate": 9.871780948290999e-06, "loss": 1.2536, "step": 3780 }, { "epoch": 0.2963166144200627, "grad_norm": 3.1035516262054443, "learning_rate": 9.871711554126068e-06, "loss": 1.5193, "step": 3781 }, { "epoch": 0.2963949843260188, "grad_norm": 5.022363185882568, "learning_rate": 9.871642141431648e-06, "loss": 1.2255, "step": 3782 }, { "epoch": 0.2964733542319749, "grad_norm": 3.068668842315674, "learning_rate": 9.871572710208004e-06, "loss": 1.142, "step": 3783 }, { "epoch": 0.296551724137931, "grad_norm": 3.409482479095459, "learning_rate": 9.8715032604554e-06, "loss": 1.4758, "step": 3784 }, { "epoch": 0.2966300940438871, "grad_norm": 4.949272632598877, "learning_rate": 9.871433792174098e-06, "loss": 1.6305, "step": 3785 }, { "epoch": 0.29670846394984324, "grad_norm": 2.50893497467041, "learning_rate": 9.871364305364366e-06, "loss": 1.2293, "step": 3786 }, { "epoch": 0.29678683385579935, "grad_norm": 2.017371416091919, "learning_rate": 9.871294800026466e-06, "loss": 1.37, "step": 3787 }, { "epoch": 0.29686520376175546, "grad_norm": 3.108417510986328, "learning_rate": 9.871225276160664e-06, "loss": 1.3529, "step": 3788 }, { "epoch": 0.2969435736677116, "grad_norm": 3.2251780033111572, "learning_rate": 9.87115573376722e-06, "loss": 1.3128, "step": 3789 }, { "epoch": 0.29702194357366773, "grad_norm": 2.6215386390686035, "learning_rate": 9.871086172846403e-06, "loss": 1.2557, "step": 3790 }, { "epoch": 0.29710031347962385, "grad_norm": 2.8951385021209717, "learning_rate": 9.871016593398477e-06, "loss": 1.2897, "step": 3791 }, { "epoch": 0.29717868338557996, "grad_norm": 5.282125473022461, "learning_rate": 9.870946995423705e-06, "loss": 1.1282, "step": 3792 }, { "epoch": 0.29725705329153607, "grad_norm": 2.3631644248962402, "learning_rate": 9.870877378922352e-06, "loss": 1.1858, "step": 3793 }, { "epoch": 0.2973354231974922, "grad_norm": 4.297853946685791, "learning_rate": 9.870807743894683e-06, "loss": 1.2635, "step": 3794 }, { "epoch": 0.2974137931034483, "grad_norm": 2.064422845840454, "learning_rate": 9.870738090340964e-06, "loss": 1.1909, "step": 3795 }, { "epoch": 0.2974921630094044, "grad_norm": 2.510730504989624, "learning_rate": 9.87066841826146e-06, "loss": 1.0213, "step": 3796 }, { "epoch": 0.2975705329153605, "grad_norm": 6.154714107513428, "learning_rate": 9.870598727656433e-06, "loss": 0.9847, "step": 3797 }, { "epoch": 0.2976489028213166, "grad_norm": 5.393519401550293, "learning_rate": 9.870529018526151e-06, "loss": 1.4047, "step": 3798 }, { "epoch": 0.29772727272727273, "grad_norm": 2.234419107437134, "learning_rate": 9.870459290870878e-06, "loss": 0.7802, "step": 3799 }, { "epoch": 0.29780564263322884, "grad_norm": 2.973421573638916, "learning_rate": 9.87038954469088e-06, "loss": 1.752, "step": 3800 }, { "epoch": 0.29788401253918495, "grad_norm": 2.9373278617858887, "learning_rate": 9.870319779986422e-06, "loss": 1.6002, "step": 3801 }, { "epoch": 0.29796238244514106, "grad_norm": 3.080824613571167, "learning_rate": 9.870249996757769e-06, "loss": 1.5383, "step": 3802 }, { "epoch": 0.29804075235109717, "grad_norm": 3.9837944507598877, "learning_rate": 9.870180195005185e-06, "loss": 1.7039, "step": 3803 }, { "epoch": 0.2981191222570533, "grad_norm": 2.746018648147583, "learning_rate": 9.870110374728939e-06, "loss": 1.4112, "step": 3804 }, { "epoch": 0.2981974921630094, "grad_norm": 2.493690252304077, "learning_rate": 9.870040535929296e-06, "loss": 1.7017, "step": 3805 }, { "epoch": 0.2982758620689655, "grad_norm": 8.419581413269043, "learning_rate": 9.869970678606517e-06, "loss": 1.7877, "step": 3806 }, { "epoch": 0.2983542319749216, "grad_norm": 6.911056041717529, "learning_rate": 9.869900802760873e-06, "loss": 1.8626, "step": 3807 }, { "epoch": 0.2984326018808777, "grad_norm": 2.7836802005767822, "learning_rate": 9.869830908392626e-06, "loss": 1.3415, "step": 3808 }, { "epoch": 0.29851097178683383, "grad_norm": 3.579984188079834, "learning_rate": 9.869760995502043e-06, "loss": 1.0839, "step": 3809 }, { "epoch": 0.29858934169278994, "grad_norm": 2.5349276065826416, "learning_rate": 9.869691064089393e-06, "loss": 1.3925, "step": 3810 }, { "epoch": 0.29866771159874606, "grad_norm": 2.22464656829834, "learning_rate": 9.869621114154937e-06, "loss": 1.1698, "step": 3811 }, { "epoch": 0.2987460815047022, "grad_norm": 2.3763644695281982, "learning_rate": 9.869551145698946e-06, "loss": 1.1239, "step": 3812 }, { "epoch": 0.29882445141065833, "grad_norm": 3.084285020828247, "learning_rate": 9.869481158721681e-06, "loss": 1.4048, "step": 3813 }, { "epoch": 0.29890282131661444, "grad_norm": 2.321653127670288, "learning_rate": 9.869411153223411e-06, "loss": 1.2354, "step": 3814 }, { "epoch": 0.29898119122257055, "grad_norm": 7.261682033538818, "learning_rate": 9.869341129204404e-06, "loss": 1.484, "step": 3815 }, { "epoch": 0.29905956112852666, "grad_norm": 5.737380504608154, "learning_rate": 9.869271086664924e-06, "loss": 1.9552, "step": 3816 }, { "epoch": 0.2991379310344828, "grad_norm": 2.9508399963378906, "learning_rate": 9.869201025605236e-06, "loss": 1.4474, "step": 3817 }, { "epoch": 0.2992163009404389, "grad_norm": 2.4054415225982666, "learning_rate": 9.869130946025612e-06, "loss": 0.9362, "step": 3818 }, { "epoch": 0.299294670846395, "grad_norm": 3.3559730052948, "learning_rate": 9.86906084792631e-06, "loss": 1.6295, "step": 3819 }, { "epoch": 0.2993730407523511, "grad_norm": 2.7889041900634766, "learning_rate": 9.868990731307604e-06, "loss": 1.4782, "step": 3820 }, { "epoch": 0.2994514106583072, "grad_norm": 3.636091947555542, "learning_rate": 9.868920596169758e-06, "loss": 1.4617, "step": 3821 }, { "epoch": 0.2995297805642633, "grad_norm": 2.5948538780212402, "learning_rate": 9.868850442513039e-06, "loss": 1.4583, "step": 3822 }, { "epoch": 0.29960815047021944, "grad_norm": 2.7187602519989014, "learning_rate": 9.868780270337714e-06, "loss": 1.351, "step": 3823 }, { "epoch": 0.29968652037617555, "grad_norm": 3.548344373703003, "learning_rate": 9.868710079644048e-06, "loss": 1.2885, "step": 3824 }, { "epoch": 0.29976489028213166, "grad_norm": 2.6027984619140625, "learning_rate": 9.868639870432311e-06, "loss": 1.4649, "step": 3825 }, { "epoch": 0.29984326018808777, "grad_norm": 6.223665714263916, "learning_rate": 9.868569642702768e-06, "loss": 0.6963, "step": 3826 }, { "epoch": 0.2999216300940439, "grad_norm": 2.9060728549957275, "learning_rate": 9.868499396455688e-06, "loss": 1.3262, "step": 3827 }, { "epoch": 0.3, "grad_norm": 14.442137718200684, "learning_rate": 9.868429131691336e-06, "loss": 1.7382, "step": 3828 }, { "epoch": 0.3000783699059561, "grad_norm": 3.1819064617156982, "learning_rate": 9.868358848409981e-06, "loss": 0.8703, "step": 3829 }, { "epoch": 0.3001567398119122, "grad_norm": 2.939166307449341, "learning_rate": 9.868288546611888e-06, "loss": 1.504, "step": 3830 }, { "epoch": 0.3002351097178683, "grad_norm": 4.199913024902344, "learning_rate": 9.868218226297326e-06, "loss": 1.5858, "step": 3831 }, { "epoch": 0.30031347962382443, "grad_norm": 2.1774282455444336, "learning_rate": 9.868147887466563e-06, "loss": 1.4616, "step": 3832 }, { "epoch": 0.30039184952978054, "grad_norm": 3.0367021560668945, "learning_rate": 9.868077530119866e-06, "loss": 1.4147, "step": 3833 }, { "epoch": 0.30047021943573665, "grad_norm": 2.9760639667510986, "learning_rate": 9.868007154257502e-06, "loss": 1.6171, "step": 3834 }, { "epoch": 0.30054858934169276, "grad_norm": 3.2663848400115967, "learning_rate": 9.86793675987974e-06, "loss": 1.6349, "step": 3835 }, { "epoch": 0.30062695924764893, "grad_norm": 7.592523574829102, "learning_rate": 9.867866346986846e-06, "loss": 1.8753, "step": 3836 }, { "epoch": 0.30070532915360504, "grad_norm": 2.8953449726104736, "learning_rate": 9.867795915579089e-06, "loss": 1.7404, "step": 3837 }, { "epoch": 0.30078369905956115, "grad_norm": 4.462435722351074, "learning_rate": 9.867725465656736e-06, "loss": 0.9258, "step": 3838 }, { "epoch": 0.30086206896551726, "grad_norm": 2.8445639610290527, "learning_rate": 9.867654997220057e-06, "loss": 1.555, "step": 3839 }, { "epoch": 0.30094043887147337, "grad_norm": 2.5432212352752686, "learning_rate": 9.867584510269317e-06, "loss": 1.448, "step": 3840 }, { "epoch": 0.3010188087774295, "grad_norm": 3.763348340988159, "learning_rate": 9.867514004804788e-06, "loss": 1.3386, "step": 3841 }, { "epoch": 0.3010971786833856, "grad_norm": 2.3109524250030518, "learning_rate": 9.867443480826732e-06, "loss": 1.6285, "step": 3842 }, { "epoch": 0.3011755485893417, "grad_norm": 2.881124973297119, "learning_rate": 9.867372938335425e-06, "loss": 1.0685, "step": 3843 }, { "epoch": 0.3012539184952978, "grad_norm": 2.8054230213165283, "learning_rate": 9.86730237733113e-06, "loss": 1.1771, "step": 3844 }, { "epoch": 0.3013322884012539, "grad_norm": 5.687846660614014, "learning_rate": 9.867231797814115e-06, "loss": 2.0613, "step": 3845 }, { "epoch": 0.30141065830721003, "grad_norm": 2.6418564319610596, "learning_rate": 9.867161199784651e-06, "loss": 1.7362, "step": 3846 }, { "epoch": 0.30148902821316614, "grad_norm": 4.9819746017456055, "learning_rate": 9.867090583243008e-06, "loss": 1.3282, "step": 3847 }, { "epoch": 0.30156739811912225, "grad_norm": 3.588221788406372, "learning_rate": 9.86701994818945e-06, "loss": 1.4366, "step": 3848 }, { "epoch": 0.30164576802507836, "grad_norm": 4.505540370941162, "learning_rate": 9.866949294624249e-06, "loss": 1.4768, "step": 3849 }, { "epoch": 0.3017241379310345, "grad_norm": 3.4020628929138184, "learning_rate": 9.866878622547671e-06, "loss": 1.4067, "step": 3850 }, { "epoch": 0.3018025078369906, "grad_norm": 4.821435451507568, "learning_rate": 9.86680793195999e-06, "loss": 1.3309, "step": 3851 }, { "epoch": 0.3018808777429467, "grad_norm": 3.5542550086975098, "learning_rate": 9.86673722286147e-06, "loss": 1.4529, "step": 3852 }, { "epoch": 0.3019592476489028, "grad_norm": 2.9808897972106934, "learning_rate": 9.866666495252379e-06, "loss": 1.2385, "step": 3853 }, { "epoch": 0.3020376175548589, "grad_norm": 4.4823503494262695, "learning_rate": 9.86659574913299e-06, "loss": 1.4497, "step": 3854 }, { "epoch": 0.302115987460815, "grad_norm": 7.236866474151611, "learning_rate": 9.86652498450357e-06, "loss": 1.2584, "step": 3855 }, { "epoch": 0.30219435736677114, "grad_norm": 2.2432918548583984, "learning_rate": 9.866454201364389e-06, "loss": 1.2278, "step": 3856 }, { "epoch": 0.30227272727272725, "grad_norm": 2.200247287750244, "learning_rate": 9.866383399715716e-06, "loss": 1.4516, "step": 3857 }, { "epoch": 0.30235109717868336, "grad_norm": 2.168635368347168, "learning_rate": 9.86631257955782e-06, "loss": 1.5457, "step": 3858 }, { "epoch": 0.3024294670846395, "grad_norm": 2.894679546356201, "learning_rate": 9.86624174089097e-06, "loss": 1.5883, "step": 3859 }, { "epoch": 0.30250783699059564, "grad_norm": 7.417827606201172, "learning_rate": 9.866170883715435e-06, "loss": 1.2145, "step": 3860 }, { "epoch": 0.30258620689655175, "grad_norm": 4.328233242034912, "learning_rate": 9.866100008031487e-06, "loss": 0.9246, "step": 3861 }, { "epoch": 0.30266457680250786, "grad_norm": 3.2827796936035156, "learning_rate": 9.866029113839395e-06, "loss": 1.6069, "step": 3862 }, { "epoch": 0.30274294670846397, "grad_norm": 2.764063835144043, "learning_rate": 9.865958201139425e-06, "loss": 1.3714, "step": 3863 }, { "epoch": 0.3028213166144201, "grad_norm": 10.945127487182617, "learning_rate": 9.865887269931849e-06, "loss": 1.3939, "step": 3864 }, { "epoch": 0.3028996865203762, "grad_norm": 3.2401111125946045, "learning_rate": 9.865816320216938e-06, "loss": 1.5763, "step": 3865 }, { "epoch": 0.3029780564263323, "grad_norm": 3.76743745803833, "learning_rate": 9.865745351994962e-06, "loss": 1.1233, "step": 3866 }, { "epoch": 0.3030564263322884, "grad_norm": 4.972009181976318, "learning_rate": 9.86567436526619e-06, "loss": 1.8051, "step": 3867 }, { "epoch": 0.3031347962382445, "grad_norm": 4.984031677246094, "learning_rate": 9.86560336003089e-06, "loss": 1.0798, "step": 3868 }, { "epoch": 0.30321316614420063, "grad_norm": 3.639136791229248, "learning_rate": 9.865532336289336e-06, "loss": 1.6361, "step": 3869 }, { "epoch": 0.30329153605015674, "grad_norm": 3.050006866455078, "learning_rate": 9.865461294041796e-06, "loss": 1.5813, "step": 3870 }, { "epoch": 0.30336990595611285, "grad_norm": 8.546134948730469, "learning_rate": 9.865390233288538e-06, "loss": 1.2193, "step": 3871 }, { "epoch": 0.30344827586206896, "grad_norm": 3.6437759399414062, "learning_rate": 9.865319154029838e-06, "loss": 1.5026, "step": 3872 }, { "epoch": 0.30352664576802507, "grad_norm": 2.602052688598633, "learning_rate": 9.86524805626596e-06, "loss": 1.5138, "step": 3873 }, { "epoch": 0.3036050156739812, "grad_norm": 5.355443477630615, "learning_rate": 9.86517693999718e-06, "loss": 1.3959, "step": 3874 }, { "epoch": 0.3036833855799373, "grad_norm": 3.0957257747650146, "learning_rate": 9.865105805223762e-06, "loss": 1.5979, "step": 3875 }, { "epoch": 0.3037617554858934, "grad_norm": 20.52826690673828, "learning_rate": 9.865034651945983e-06, "loss": 1.9252, "step": 3876 }, { "epoch": 0.3038401253918495, "grad_norm": 2.9203238487243652, "learning_rate": 9.864963480164111e-06, "loss": 1.5068, "step": 3877 }, { "epoch": 0.3039184952978056, "grad_norm": 2.438504934310913, "learning_rate": 9.864892289878416e-06, "loss": 1.5775, "step": 3878 }, { "epoch": 0.30399686520376173, "grad_norm": 2.8822274208068848, "learning_rate": 9.86482108108917e-06, "loss": 0.9756, "step": 3879 }, { "epoch": 0.30407523510971785, "grad_norm": 3.9926722049713135, "learning_rate": 9.864749853796642e-06, "loss": 0.5877, "step": 3880 }, { "epoch": 0.30415360501567396, "grad_norm": 3.117114782333374, "learning_rate": 9.864678608001105e-06, "loss": 1.6809, "step": 3881 }, { "epoch": 0.3042319749216301, "grad_norm": 2.5361058712005615, "learning_rate": 9.864607343702832e-06, "loss": 1.575, "step": 3882 }, { "epoch": 0.30431034482758623, "grad_norm": 2.031792640686035, "learning_rate": 9.864536060902088e-06, "loss": 1.5784, "step": 3883 }, { "epoch": 0.30438871473354234, "grad_norm": 10.27558422088623, "learning_rate": 9.864464759599149e-06, "loss": 2.3033, "step": 3884 }, { "epoch": 0.30446708463949845, "grad_norm": 4.002869129180908, "learning_rate": 9.864393439794283e-06, "loss": 1.6023, "step": 3885 }, { "epoch": 0.30454545454545456, "grad_norm": 9.319286346435547, "learning_rate": 9.864322101487764e-06, "loss": 0.9945, "step": 3886 }, { "epoch": 0.3046238244514107, "grad_norm": 2.3921782970428467, "learning_rate": 9.864250744679862e-06, "loss": 1.5393, "step": 3887 }, { "epoch": 0.3047021943573668, "grad_norm": 2.2206034660339355, "learning_rate": 9.864179369370848e-06, "loss": 1.3521, "step": 3888 }, { "epoch": 0.3047805642633229, "grad_norm": 5.311975479125977, "learning_rate": 9.864107975560995e-06, "loss": 1.9988, "step": 3889 }, { "epoch": 0.304858934169279, "grad_norm": 3.804478168487549, "learning_rate": 9.864036563250575e-06, "loss": 1.2736, "step": 3890 }, { "epoch": 0.3049373040752351, "grad_norm": 2.8715991973876953, "learning_rate": 9.863965132439854e-06, "loss": 1.4237, "step": 3891 }, { "epoch": 0.3050156739811912, "grad_norm": 9.752155303955078, "learning_rate": 9.86389368312911e-06, "loss": 1.736, "step": 3892 }, { "epoch": 0.30509404388714734, "grad_norm": 6.4020304679870605, "learning_rate": 9.863822215318615e-06, "loss": 1.486, "step": 3893 }, { "epoch": 0.30517241379310345, "grad_norm": 2.4577765464782715, "learning_rate": 9.863750729008637e-06, "loss": 1.4595, "step": 3894 }, { "epoch": 0.30525078369905956, "grad_norm": 3.9279394149780273, "learning_rate": 9.863679224199448e-06, "loss": 1.445, "step": 3895 }, { "epoch": 0.30532915360501567, "grad_norm": 6.3319091796875, "learning_rate": 9.863607700891323e-06, "loss": 1.2864, "step": 3896 }, { "epoch": 0.3054075235109718, "grad_norm": 4.715821743011475, "learning_rate": 9.86353615908453e-06, "loss": 1.2102, "step": 3897 }, { "epoch": 0.3054858934169279, "grad_norm": 10.853287696838379, "learning_rate": 9.863464598779347e-06, "loss": 2.1016, "step": 3898 }, { "epoch": 0.305564263322884, "grad_norm": 2.636380195617676, "learning_rate": 9.86339301997604e-06, "loss": 1.2674, "step": 3899 }, { "epoch": 0.3056426332288401, "grad_norm": 2.79781174659729, "learning_rate": 9.863321422674885e-06, "loss": 1.2779, "step": 3900 }, { "epoch": 0.3057210031347962, "grad_norm": 4.033807277679443, "learning_rate": 9.863249806876153e-06, "loss": 0.7218, "step": 3901 }, { "epoch": 0.30579937304075233, "grad_norm": 5.336950778961182, "learning_rate": 9.86317817258012e-06, "loss": 1.8001, "step": 3902 }, { "epoch": 0.30587774294670844, "grad_norm": 2.841491222381592, "learning_rate": 9.86310651978705e-06, "loss": 1.7345, "step": 3903 }, { "epoch": 0.30595611285266455, "grad_norm": 2.562727928161621, "learning_rate": 9.863034848497223e-06, "loss": 1.3983, "step": 3904 }, { "epoch": 0.30603448275862066, "grad_norm": 3.635111093521118, "learning_rate": 9.86296315871091e-06, "loss": 1.5157, "step": 3905 }, { "epoch": 0.30611285266457683, "grad_norm": 3.272693395614624, "learning_rate": 9.86289145042838e-06, "loss": 1.5269, "step": 3906 }, { "epoch": 0.30619122257053294, "grad_norm": 3.377668857574463, "learning_rate": 9.862819723649912e-06, "loss": 1.348, "step": 3907 }, { "epoch": 0.30626959247648905, "grad_norm": 7.3221116065979, "learning_rate": 9.862747978375773e-06, "loss": 1.9278, "step": 3908 }, { "epoch": 0.30634796238244516, "grad_norm": 2.0616652965545654, "learning_rate": 9.86267621460624e-06, "loss": 0.907, "step": 3909 }, { "epoch": 0.30642633228840127, "grad_norm": 4.494427680969238, "learning_rate": 9.862604432341583e-06, "loss": 1.3739, "step": 3910 }, { "epoch": 0.3065047021943574, "grad_norm": 2.851440191268921, "learning_rate": 9.862532631582077e-06, "loss": 1.3156, "step": 3911 }, { "epoch": 0.3065830721003135, "grad_norm": 5.553094387054443, "learning_rate": 9.862460812327995e-06, "loss": 1.3364, "step": 3912 }, { "epoch": 0.3066614420062696, "grad_norm": 3.1940393447875977, "learning_rate": 9.862388974579608e-06, "loss": 1.138, "step": 3913 }, { "epoch": 0.3067398119122257, "grad_norm": 3.246802568435669, "learning_rate": 9.862317118337193e-06, "loss": 1.417, "step": 3914 }, { "epoch": 0.3068181818181818, "grad_norm": 2.814598798751831, "learning_rate": 9.862245243601019e-06, "loss": 1.2959, "step": 3915 }, { "epoch": 0.30689655172413793, "grad_norm": 2.6386797428131104, "learning_rate": 9.862173350371362e-06, "loss": 1.2416, "step": 3916 }, { "epoch": 0.30697492163009404, "grad_norm": 2.0305166244506836, "learning_rate": 9.862101438648496e-06, "loss": 1.2998, "step": 3917 }, { "epoch": 0.30705329153605015, "grad_norm": 2.759528875350952, "learning_rate": 9.862029508432693e-06, "loss": 1.5598, "step": 3918 }, { "epoch": 0.30713166144200627, "grad_norm": 1.9493683576583862, "learning_rate": 9.861957559724227e-06, "loss": 1.3172, "step": 3919 }, { "epoch": 0.3072100313479624, "grad_norm": 6.170773506164551, "learning_rate": 9.861885592523372e-06, "loss": 1.5145, "step": 3920 }, { "epoch": 0.3072884012539185, "grad_norm": 3.094058036804199, "learning_rate": 9.861813606830402e-06, "loss": 1.3398, "step": 3921 }, { "epoch": 0.3073667711598746, "grad_norm": 3.7196285724639893, "learning_rate": 9.861741602645588e-06, "loss": 1.5727, "step": 3922 }, { "epoch": 0.3074451410658307, "grad_norm": 3.4274251461029053, "learning_rate": 9.861669579969208e-06, "loss": 1.3652, "step": 3923 }, { "epoch": 0.3075235109717868, "grad_norm": 2.4808433055877686, "learning_rate": 9.861597538801533e-06, "loss": 1.6347, "step": 3924 }, { "epoch": 0.30760188087774293, "grad_norm": 2.572711944580078, "learning_rate": 9.861525479142839e-06, "loss": 1.3819, "step": 3925 }, { "epoch": 0.30768025078369904, "grad_norm": 14.317069053649902, "learning_rate": 9.861453400993398e-06, "loss": 1.6334, "step": 3926 }, { "epoch": 0.30775862068965515, "grad_norm": 3.141610860824585, "learning_rate": 9.861381304353486e-06, "loss": 1.3067, "step": 3927 }, { "epoch": 0.30783699059561126, "grad_norm": 5.4567413330078125, "learning_rate": 9.861309189223375e-06, "loss": 1.3884, "step": 3928 }, { "epoch": 0.3079153605015674, "grad_norm": 2.265070676803589, "learning_rate": 9.861237055603343e-06, "loss": 1.4676, "step": 3929 }, { "epoch": 0.30799373040752354, "grad_norm": 3.385174512863159, "learning_rate": 9.861164903493661e-06, "loss": 0.9152, "step": 3930 }, { "epoch": 0.30807210031347965, "grad_norm": 11.532037734985352, "learning_rate": 9.861092732894605e-06, "loss": 1.8917, "step": 3931 }, { "epoch": 0.30815047021943576, "grad_norm": 3.598130702972412, "learning_rate": 9.861020543806449e-06, "loss": 1.5227, "step": 3932 }, { "epoch": 0.30822884012539187, "grad_norm": 2.9250473976135254, "learning_rate": 9.860948336229468e-06, "loss": 1.6482, "step": 3933 }, { "epoch": 0.308307210031348, "grad_norm": 3.1398284435272217, "learning_rate": 9.860876110163935e-06, "loss": 1.4782, "step": 3934 }, { "epoch": 0.3083855799373041, "grad_norm": 3.9384617805480957, "learning_rate": 9.860803865610127e-06, "loss": 1.2561, "step": 3935 }, { "epoch": 0.3084639498432602, "grad_norm": 3.1284289360046387, "learning_rate": 9.860731602568317e-06, "loss": 1.2803, "step": 3936 }, { "epoch": 0.3085423197492163, "grad_norm": 5.457663536071777, "learning_rate": 9.860659321038782e-06, "loss": 1.5877, "step": 3937 }, { "epoch": 0.3086206896551724, "grad_norm": 3.6538848876953125, "learning_rate": 9.860587021021794e-06, "loss": 1.4926, "step": 3938 }, { "epoch": 0.30869905956112853, "grad_norm": 2.579326629638672, "learning_rate": 9.86051470251763e-06, "loss": 1.4929, "step": 3939 }, { "epoch": 0.30877742946708464, "grad_norm": 4.093516826629639, "learning_rate": 9.860442365526565e-06, "loss": 1.4606, "step": 3940 }, { "epoch": 0.30885579937304075, "grad_norm": 5.274941921234131, "learning_rate": 9.860370010048873e-06, "loss": 1.0106, "step": 3941 }, { "epoch": 0.30893416927899686, "grad_norm": 2.4214389324188232, "learning_rate": 9.860297636084831e-06, "loss": 1.5827, "step": 3942 }, { "epoch": 0.309012539184953, "grad_norm": 2.435465097427368, "learning_rate": 9.860225243634712e-06, "loss": 1.3366, "step": 3943 }, { "epoch": 0.3090909090909091, "grad_norm": 2.4884281158447266, "learning_rate": 9.860152832698793e-06, "loss": 1.4954, "step": 3944 }, { "epoch": 0.3091692789968652, "grad_norm": 2.9107372760772705, "learning_rate": 9.860080403277349e-06, "loss": 1.4133, "step": 3945 }, { "epoch": 0.3092476489028213, "grad_norm": 4.45574951171875, "learning_rate": 9.860007955370656e-06, "loss": 1.5789, "step": 3946 }, { "epoch": 0.3093260188087774, "grad_norm": 5.338067531585693, "learning_rate": 9.859935488978989e-06, "loss": 1.8274, "step": 3947 }, { "epoch": 0.3094043887147335, "grad_norm": 3.8476808071136475, "learning_rate": 9.859863004102622e-06, "loss": 1.4475, "step": 3948 }, { "epoch": 0.30948275862068964, "grad_norm": 2.3313722610473633, "learning_rate": 9.859790500741834e-06, "loss": 1.222, "step": 3949 }, { "epoch": 0.30956112852664575, "grad_norm": 3.576674461364746, "learning_rate": 9.859717978896898e-06, "loss": 1.2514, "step": 3950 }, { "epoch": 0.30963949843260186, "grad_norm": 4.133580684661865, "learning_rate": 9.859645438568093e-06, "loss": 2.0195, "step": 3951 }, { "epoch": 0.30971786833855797, "grad_norm": 5.168411731719971, "learning_rate": 9.85957287975569e-06, "loss": 1.0343, "step": 3952 }, { "epoch": 0.30979623824451413, "grad_norm": 3.2899951934814453, "learning_rate": 9.85950030245997e-06, "loss": 1.4997, "step": 3953 }, { "epoch": 0.30987460815047024, "grad_norm": 6.257352352142334, "learning_rate": 9.859427706681205e-06, "loss": 1.252, "step": 3954 }, { "epoch": 0.30995297805642635, "grad_norm": 2.85129451751709, "learning_rate": 9.859355092419673e-06, "loss": 1.5477, "step": 3955 }, { "epoch": 0.31003134796238246, "grad_norm": 3.1476593017578125, "learning_rate": 9.859282459675653e-06, "loss": 1.5194, "step": 3956 }, { "epoch": 0.3101097178683386, "grad_norm": 3.086371898651123, "learning_rate": 9.859209808449414e-06, "loss": 1.3443, "step": 3957 }, { "epoch": 0.3101880877742947, "grad_norm": 3.5776500701904297, "learning_rate": 9.85913713874124e-06, "loss": 1.4631, "step": 3958 }, { "epoch": 0.3102664576802508, "grad_norm": 2.85888671875, "learning_rate": 9.859064450551404e-06, "loss": 1.5014, "step": 3959 }, { "epoch": 0.3103448275862069, "grad_norm": 3.4476706981658936, "learning_rate": 9.858991743880183e-06, "loss": 1.2831, "step": 3960 }, { "epoch": 0.310423197492163, "grad_norm": 2.9533891677856445, "learning_rate": 9.858919018727851e-06, "loss": 1.6722, "step": 3961 }, { "epoch": 0.3105015673981191, "grad_norm": 4.875692367553711, "learning_rate": 9.858846275094687e-06, "loss": 1.4042, "step": 3962 }, { "epoch": 0.31057993730407524, "grad_norm": 3.633674383163452, "learning_rate": 9.858773512980968e-06, "loss": 1.4504, "step": 3963 }, { "epoch": 0.31065830721003135, "grad_norm": 2.8179266452789307, "learning_rate": 9.858700732386971e-06, "loss": 1.4906, "step": 3964 }, { "epoch": 0.31073667711598746, "grad_norm": 3.773101329803467, "learning_rate": 9.858627933312971e-06, "loss": 1.4213, "step": 3965 }, { "epoch": 0.31081504702194357, "grad_norm": 2.9722352027893066, "learning_rate": 9.858555115759247e-06, "loss": 1.7996, "step": 3966 }, { "epoch": 0.3108934169278997, "grad_norm": 3.066115617752075, "learning_rate": 9.858482279726074e-06, "loss": 1.7232, "step": 3967 }, { "epoch": 0.3109717868338558, "grad_norm": 2.22419810295105, "learning_rate": 9.85840942521373e-06, "loss": 1.4801, "step": 3968 }, { "epoch": 0.3110501567398119, "grad_norm": 2.674854040145874, "learning_rate": 9.858336552222492e-06, "loss": 1.387, "step": 3969 }, { "epoch": 0.311128526645768, "grad_norm": 2.4152066707611084, "learning_rate": 9.858263660752637e-06, "loss": 1.2539, "step": 3970 }, { "epoch": 0.3112068965517241, "grad_norm": 2.4839928150177, "learning_rate": 9.858190750804442e-06, "loss": 1.0835, "step": 3971 }, { "epoch": 0.31128526645768023, "grad_norm": 4.086140155792236, "learning_rate": 9.858117822378186e-06, "loss": 1.5243, "step": 3972 }, { "epoch": 0.31136363636363634, "grad_norm": 3.028449296951294, "learning_rate": 9.858044875474145e-06, "loss": 1.3159, "step": 3973 }, { "epoch": 0.31144200626959245, "grad_norm": 2.390076160430908, "learning_rate": 9.857971910092595e-06, "loss": 1.4228, "step": 3974 }, { "epoch": 0.31152037617554856, "grad_norm": 3.5069069862365723, "learning_rate": 9.857898926233815e-06, "loss": 1.9279, "step": 3975 }, { "epoch": 0.31159874608150473, "grad_norm": 2.7804207801818848, "learning_rate": 9.857825923898084e-06, "loss": 1.195, "step": 3976 }, { "epoch": 0.31167711598746084, "grad_norm": 2.655184268951416, "learning_rate": 9.857752903085679e-06, "loss": 1.5356, "step": 3977 }, { "epoch": 0.31175548589341695, "grad_norm": 2.2608182430267334, "learning_rate": 9.857679863796876e-06, "loss": 1.4135, "step": 3978 }, { "epoch": 0.31183385579937306, "grad_norm": 2.806579351425171, "learning_rate": 9.857606806031953e-06, "loss": 1.2709, "step": 3979 }, { "epoch": 0.31191222570532917, "grad_norm": 3.9527084827423096, "learning_rate": 9.85753372979119e-06, "loss": 1.7546, "step": 3980 }, { "epoch": 0.3119905956112853, "grad_norm": 2.3853402137756348, "learning_rate": 9.857460635074864e-06, "loss": 0.879, "step": 3981 }, { "epoch": 0.3120689655172414, "grad_norm": 2.6835877895355225, "learning_rate": 9.857387521883251e-06, "loss": 1.4283, "step": 3982 }, { "epoch": 0.3121473354231975, "grad_norm": 2.4081389904022217, "learning_rate": 9.857314390216632e-06, "loss": 1.3146, "step": 3983 }, { "epoch": 0.3122257053291536, "grad_norm": 5.047996520996094, "learning_rate": 9.857241240075283e-06, "loss": 1.3622, "step": 3984 }, { "epoch": 0.3123040752351097, "grad_norm": 6.658407211303711, "learning_rate": 9.857168071459484e-06, "loss": 0.911, "step": 3985 }, { "epoch": 0.31238244514106583, "grad_norm": 5.727260589599609, "learning_rate": 9.857094884369514e-06, "loss": 2.302, "step": 3986 }, { "epoch": 0.31246081504702194, "grad_norm": 3.142249822616577, "learning_rate": 9.857021678805648e-06, "loss": 1.1209, "step": 3987 }, { "epoch": 0.31253918495297806, "grad_norm": 4.278665542602539, "learning_rate": 9.856948454768164e-06, "loss": 1.6991, "step": 3988 }, { "epoch": 0.31261755485893417, "grad_norm": 2.6609368324279785, "learning_rate": 9.856875212257346e-06, "loss": 1.5431, "step": 3989 }, { "epoch": 0.3126959247648903, "grad_norm": 4.094048976898193, "learning_rate": 9.85680195127347e-06, "loss": 1.2895, "step": 3990 }, { "epoch": 0.3127742946708464, "grad_norm": 3.391983985900879, "learning_rate": 9.856728671816811e-06, "loss": 1.2753, "step": 3991 }, { "epoch": 0.3128526645768025, "grad_norm": 2.1785202026367188, "learning_rate": 9.856655373887653e-06, "loss": 1.4112, "step": 3992 }, { "epoch": 0.3129310344827586, "grad_norm": 4.846065044403076, "learning_rate": 9.856582057486272e-06, "loss": 1.4637, "step": 3993 }, { "epoch": 0.3130094043887147, "grad_norm": 3.408926486968994, "learning_rate": 9.856508722612947e-06, "loss": 1.8276, "step": 3994 }, { "epoch": 0.31308777429467083, "grad_norm": 4.6171770095825195, "learning_rate": 9.856435369267956e-06, "loss": 1.5851, "step": 3995 }, { "epoch": 0.31316614420062694, "grad_norm": 9.671489715576172, "learning_rate": 9.856361997451582e-06, "loss": 2.0345, "step": 3996 }, { "epoch": 0.31324451410658305, "grad_norm": 2.4779348373413086, "learning_rate": 9.856288607164099e-06, "loss": 1.1989, "step": 3997 }, { "epoch": 0.31332288401253916, "grad_norm": 2.6046745777130127, "learning_rate": 9.856215198405789e-06, "loss": 1.3647, "step": 3998 }, { "epoch": 0.31340125391849527, "grad_norm": 2.554966449737549, "learning_rate": 9.856141771176933e-06, "loss": 1.7433, "step": 3999 }, { "epoch": 0.31347962382445144, "grad_norm": 4.040160179138184, "learning_rate": 9.856068325477805e-06, "loss": 1.3022, "step": 4000 }, { "epoch": 0.31355799373040755, "grad_norm": 2.887350559234619, "learning_rate": 9.855994861308689e-06, "loss": 1.629, "step": 4001 }, { "epoch": 0.31363636363636366, "grad_norm": 2.4238357543945312, "learning_rate": 9.855921378669861e-06, "loss": 1.255, "step": 4002 }, { "epoch": 0.31371473354231977, "grad_norm": 4.326313495635986, "learning_rate": 9.855847877561604e-06, "loss": 1.3841, "step": 4003 }, { "epoch": 0.3137931034482759, "grad_norm": 2.075014114379883, "learning_rate": 9.855774357984196e-06, "loss": 1.3569, "step": 4004 }, { "epoch": 0.313871473354232, "grad_norm": 2.7974741458892822, "learning_rate": 9.855700819937915e-06, "loss": 1.3056, "step": 4005 }, { "epoch": 0.3139498432601881, "grad_norm": 5.290194988250732, "learning_rate": 9.855627263423042e-06, "loss": 2.4413, "step": 4006 }, { "epoch": 0.3140282131661442, "grad_norm": 2.3879024982452393, "learning_rate": 9.855553688439859e-06, "loss": 1.1965, "step": 4007 }, { "epoch": 0.3141065830721003, "grad_norm": 2.493727207183838, "learning_rate": 9.85548009498864e-06, "loss": 1.099, "step": 4008 }, { "epoch": 0.31418495297805643, "grad_norm": 2.2503082752227783, "learning_rate": 9.855406483069671e-06, "loss": 1.3171, "step": 4009 }, { "epoch": 0.31426332288401254, "grad_norm": 3.0243749618530273, "learning_rate": 9.85533285268323e-06, "loss": 1.9697, "step": 4010 }, { "epoch": 0.31434169278996865, "grad_norm": 3.7987453937530518, "learning_rate": 9.855259203829597e-06, "loss": 1.3307, "step": 4011 }, { "epoch": 0.31442006269592476, "grad_norm": 7.4469122886657715, "learning_rate": 9.855185536509052e-06, "loss": 1.7434, "step": 4012 }, { "epoch": 0.3144984326018809, "grad_norm": 3.83186936378479, "learning_rate": 9.855111850721874e-06, "loss": 1.6559, "step": 4013 }, { "epoch": 0.314576802507837, "grad_norm": 5.157718181610107, "learning_rate": 9.855038146468345e-06, "loss": 1.69, "step": 4014 }, { "epoch": 0.3146551724137931, "grad_norm": 5.45766544342041, "learning_rate": 9.854964423748743e-06, "loss": 1.7633, "step": 4015 }, { "epoch": 0.3147335423197492, "grad_norm": 2.0545740127563477, "learning_rate": 9.854890682563352e-06, "loss": 1.3452, "step": 4016 }, { "epoch": 0.3148119122257053, "grad_norm": 4.78878927230835, "learning_rate": 9.854816922912449e-06, "loss": 1.6565, "step": 4017 }, { "epoch": 0.3148902821316614, "grad_norm": 3.416396141052246, "learning_rate": 9.854743144796317e-06, "loss": 1.2142, "step": 4018 }, { "epoch": 0.31496865203761754, "grad_norm": 2.4637744426727295, "learning_rate": 9.854669348215234e-06, "loss": 1.2795, "step": 4019 }, { "epoch": 0.31504702194357365, "grad_norm": 3.1149744987487793, "learning_rate": 9.854595533169485e-06, "loss": 1.179, "step": 4020 }, { "epoch": 0.31512539184952976, "grad_norm": 4.292483806610107, "learning_rate": 9.854521699659344e-06, "loss": 2.3655, "step": 4021 }, { "epoch": 0.31520376175548587, "grad_norm": 3.410919427871704, "learning_rate": 9.8544478476851e-06, "loss": 1.2388, "step": 4022 }, { "epoch": 0.31528213166144203, "grad_norm": 4.919374942779541, "learning_rate": 9.854373977247027e-06, "loss": 1.433, "step": 4023 }, { "epoch": 0.31536050156739814, "grad_norm": 2.4513137340545654, "learning_rate": 9.854300088345411e-06, "loss": 1.2936, "step": 4024 }, { "epoch": 0.31543887147335425, "grad_norm": 2.8453006744384766, "learning_rate": 9.85422618098053e-06, "loss": 1.8411, "step": 4025 }, { "epoch": 0.31551724137931036, "grad_norm": 2.120997905731201, "learning_rate": 9.854152255152665e-06, "loss": 1.3211, "step": 4026 }, { "epoch": 0.3155956112852665, "grad_norm": 6.8868536949157715, "learning_rate": 9.854078310862099e-06, "loss": 1.0104, "step": 4027 }, { "epoch": 0.3156739811912226, "grad_norm": 2.821136713027954, "learning_rate": 9.854004348109112e-06, "loss": 1.2306, "step": 4028 }, { "epoch": 0.3157523510971787, "grad_norm": 2.5603764057159424, "learning_rate": 9.853930366893986e-06, "loss": 1.6665, "step": 4029 }, { "epoch": 0.3158307210031348, "grad_norm": 2.4348466396331787, "learning_rate": 9.853856367217001e-06, "loss": 1.2419, "step": 4030 }, { "epoch": 0.3159090909090909, "grad_norm": 1.7200771570205688, "learning_rate": 9.85378234907844e-06, "loss": 1.3832, "step": 4031 }, { "epoch": 0.315987460815047, "grad_norm": 5.492220401763916, "learning_rate": 9.853708312478584e-06, "loss": 1.5177, "step": 4032 }, { "epoch": 0.31606583072100314, "grad_norm": 8.531960487365723, "learning_rate": 9.853634257417715e-06, "loss": 1.7524, "step": 4033 }, { "epoch": 0.31614420062695925, "grad_norm": 5.238494873046875, "learning_rate": 9.853560183896113e-06, "loss": 1.348, "step": 4034 }, { "epoch": 0.31622257053291536, "grad_norm": 3.449906349182129, "learning_rate": 9.853486091914064e-06, "loss": 1.2324, "step": 4035 }, { "epoch": 0.31630094043887147, "grad_norm": 2.830535888671875, "learning_rate": 9.853411981471844e-06, "loss": 1.1964, "step": 4036 }, { "epoch": 0.3163793103448276, "grad_norm": 5.633446216583252, "learning_rate": 9.853337852569738e-06, "loss": 1.5328, "step": 4037 }, { "epoch": 0.3164576802507837, "grad_norm": 3.6921327114105225, "learning_rate": 9.853263705208029e-06, "loss": 1.9646, "step": 4038 }, { "epoch": 0.3165360501567398, "grad_norm": 3.731579303741455, "learning_rate": 9.853189539386997e-06, "loss": 1.5849, "step": 4039 }, { "epoch": 0.3166144200626959, "grad_norm": 3.368656635284424, "learning_rate": 9.853115355106923e-06, "loss": 1.5706, "step": 4040 }, { "epoch": 0.316692789968652, "grad_norm": 2.626666784286499, "learning_rate": 9.853041152368094e-06, "loss": 1.4, "step": 4041 }, { "epoch": 0.31677115987460813, "grad_norm": 3.644429922103882, "learning_rate": 9.852966931170787e-06, "loss": 0.8727, "step": 4042 }, { "epoch": 0.31684952978056424, "grad_norm": 2.825324058532715, "learning_rate": 9.852892691515287e-06, "loss": 1.7651, "step": 4043 }, { "epoch": 0.31692789968652035, "grad_norm": 1.9779161214828491, "learning_rate": 9.852818433401876e-06, "loss": 1.3719, "step": 4044 }, { "epoch": 0.31700626959247646, "grad_norm": 2.5162227153778076, "learning_rate": 9.852744156830835e-06, "loss": 1.6963, "step": 4045 }, { "epoch": 0.3170846394984326, "grad_norm": 3.229884147644043, "learning_rate": 9.85266986180245e-06, "loss": 1.3667, "step": 4046 }, { "epoch": 0.31716300940438874, "grad_norm": 3.345613718032837, "learning_rate": 9.852595548317001e-06, "loss": 1.1885, "step": 4047 }, { "epoch": 0.31724137931034485, "grad_norm": 3.9831104278564453, "learning_rate": 9.85252121637477e-06, "loss": 1.5346, "step": 4048 }, { "epoch": 0.31731974921630096, "grad_norm": 2.4079372882843018, "learning_rate": 9.852446865976041e-06, "loss": 1.5257, "step": 4049 }, { "epoch": 0.31739811912225707, "grad_norm": 2.8836917877197266, "learning_rate": 9.852372497121096e-06, "loss": 1.3958, "step": 4050 }, { "epoch": 0.3174764890282132, "grad_norm": 9.940301895141602, "learning_rate": 9.852298109810219e-06, "loss": 0.9016, "step": 4051 }, { "epoch": 0.3175548589341693, "grad_norm": 3.2599358558654785, "learning_rate": 9.852223704043692e-06, "loss": 1.0731, "step": 4052 }, { "epoch": 0.3176332288401254, "grad_norm": 3.0955188274383545, "learning_rate": 9.852149279821799e-06, "loss": 1.2833, "step": 4053 }, { "epoch": 0.3177115987460815, "grad_norm": 2.139582633972168, "learning_rate": 9.85207483714482e-06, "loss": 1.1597, "step": 4054 }, { "epoch": 0.3177899686520376, "grad_norm": 3.72491192817688, "learning_rate": 9.852000376013043e-06, "loss": 1.0548, "step": 4055 }, { "epoch": 0.31786833855799373, "grad_norm": 3.293142557144165, "learning_rate": 9.851925896426748e-06, "loss": 1.4777, "step": 4056 }, { "epoch": 0.31794670846394985, "grad_norm": 2.5364038944244385, "learning_rate": 9.851851398386218e-06, "loss": 1.0408, "step": 4057 }, { "epoch": 0.31802507836990596, "grad_norm": 2.4371533393859863, "learning_rate": 9.85177688189174e-06, "loss": 1.2518, "step": 4058 }, { "epoch": 0.31810344827586207, "grad_norm": 2.565612316131592, "learning_rate": 9.851702346943592e-06, "loss": 1.3051, "step": 4059 }, { "epoch": 0.3181818181818182, "grad_norm": 3.746142864227295, "learning_rate": 9.85162779354206e-06, "loss": 1.2072, "step": 4060 }, { "epoch": 0.3182601880877743, "grad_norm": 3.056126832962036, "learning_rate": 9.85155322168743e-06, "loss": 1.1402, "step": 4061 }, { "epoch": 0.3183385579937304, "grad_norm": 3.076442003250122, "learning_rate": 9.851478631379982e-06, "loss": 1.443, "step": 4062 }, { "epoch": 0.3184169278996865, "grad_norm": 2.9832091331481934, "learning_rate": 9.851404022620004e-06, "loss": 1.6521, "step": 4063 }, { "epoch": 0.3184952978056426, "grad_norm": 3.6703522205352783, "learning_rate": 9.851329395407774e-06, "loss": 1.6223, "step": 4064 }, { "epoch": 0.31857366771159873, "grad_norm": 2.4789462089538574, "learning_rate": 9.851254749743579e-06, "loss": 1.5498, "step": 4065 }, { "epoch": 0.31865203761755484, "grad_norm": 2.0333728790283203, "learning_rate": 9.851180085627703e-06, "loss": 1.4005, "step": 4066 }, { "epoch": 0.31873040752351095, "grad_norm": 2.615090847015381, "learning_rate": 9.85110540306043e-06, "loss": 1.393, "step": 4067 }, { "epoch": 0.31880877742946706, "grad_norm": 3.930274724960327, "learning_rate": 9.851030702042043e-06, "loss": 1.4624, "step": 4068 }, { "epoch": 0.31888714733542317, "grad_norm": 3.082285165786743, "learning_rate": 9.850955982572828e-06, "loss": 1.6513, "step": 4069 }, { "epoch": 0.31896551724137934, "grad_norm": 4.735537052154541, "learning_rate": 9.850881244653068e-06, "loss": 1.6029, "step": 4070 }, { "epoch": 0.31904388714733545, "grad_norm": 2.5986530780792236, "learning_rate": 9.850806488283048e-06, "loss": 1.5051, "step": 4071 }, { "epoch": 0.31912225705329156, "grad_norm": 2.905452251434326, "learning_rate": 9.850731713463048e-06, "loss": 1.2806, "step": 4072 }, { "epoch": 0.31920062695924767, "grad_norm": 5.79937219619751, "learning_rate": 9.85065692019336e-06, "loss": 1.0793, "step": 4073 }, { "epoch": 0.3192789968652038, "grad_norm": 2.941002607345581, "learning_rate": 9.850582108474262e-06, "loss": 1.237, "step": 4074 }, { "epoch": 0.3193573667711599, "grad_norm": 2.172274351119995, "learning_rate": 9.850507278306043e-06, "loss": 1.2484, "step": 4075 }, { "epoch": 0.319435736677116, "grad_norm": 3.842691421508789, "learning_rate": 9.850432429688985e-06, "loss": 1.6213, "step": 4076 }, { "epoch": 0.3195141065830721, "grad_norm": 2.8636136054992676, "learning_rate": 9.850357562623372e-06, "loss": 1.5817, "step": 4077 }, { "epoch": 0.3195924764890282, "grad_norm": 11.222128868103027, "learning_rate": 9.850282677109492e-06, "loss": 1.9709, "step": 4078 }, { "epoch": 0.31967084639498433, "grad_norm": 2.1612675189971924, "learning_rate": 9.850207773147627e-06, "loss": 1.1658, "step": 4079 }, { "epoch": 0.31974921630094044, "grad_norm": 2.0910961627960205, "learning_rate": 9.850132850738064e-06, "loss": 1.6114, "step": 4080 }, { "epoch": 0.31982758620689655, "grad_norm": 2.0903966426849365, "learning_rate": 9.850057909881086e-06, "loss": 1.4233, "step": 4081 }, { "epoch": 0.31990595611285266, "grad_norm": 3.983539581298828, "learning_rate": 9.849982950576977e-06, "loss": 1.6088, "step": 4082 }, { "epoch": 0.3199843260188088, "grad_norm": 1.9828747510910034, "learning_rate": 9.849907972826027e-06, "loss": 1.4047, "step": 4083 }, { "epoch": 0.3200626959247649, "grad_norm": 3.049884557723999, "learning_rate": 9.849832976628517e-06, "loss": 1.5611, "step": 4084 }, { "epoch": 0.320141065830721, "grad_norm": 3.054593801498413, "learning_rate": 9.849757961984731e-06, "loss": 1.7068, "step": 4085 }, { "epoch": 0.3202194357366771, "grad_norm": 4.074854373931885, "learning_rate": 9.84968292889496e-06, "loss": 1.322, "step": 4086 }, { "epoch": 0.3202978056426332, "grad_norm": 3.9651131629943848, "learning_rate": 9.849607877359487e-06, "loss": 1.6359, "step": 4087 }, { "epoch": 0.3203761755485893, "grad_norm": 5.276505947113037, "learning_rate": 9.849532807378594e-06, "loss": 1.1104, "step": 4088 }, { "epoch": 0.32045454545454544, "grad_norm": 5.765731334686279, "learning_rate": 9.849457718952569e-06, "loss": 1.6461, "step": 4089 }, { "epoch": 0.32053291536050155, "grad_norm": 4.05021333694458, "learning_rate": 9.849382612081698e-06, "loss": 1.7524, "step": 4090 }, { "epoch": 0.32061128526645766, "grad_norm": 3.924187421798706, "learning_rate": 9.849307486766267e-06, "loss": 1.4292, "step": 4091 }, { "epoch": 0.32068965517241377, "grad_norm": 4.951344013214111, "learning_rate": 9.849232343006561e-06, "loss": 1.2154, "step": 4092 }, { "epoch": 0.3207680250783699, "grad_norm": 3.0012195110321045, "learning_rate": 9.849157180802867e-06, "loss": 1.0217, "step": 4093 }, { "epoch": 0.32084639498432604, "grad_norm": 3.349520206451416, "learning_rate": 9.849082000155467e-06, "loss": 1.4684, "step": 4094 }, { "epoch": 0.32092476489028215, "grad_norm": 3.39080548286438, "learning_rate": 9.849006801064652e-06, "loss": 1.3474, "step": 4095 }, { "epoch": 0.32100313479623827, "grad_norm": 2.9854650497436523, "learning_rate": 9.848931583530704e-06, "loss": 1.5162, "step": 4096 }, { "epoch": 0.3210815047021944, "grad_norm": 2.6151976585388184, "learning_rate": 9.848856347553913e-06, "loss": 1.2317, "step": 4097 }, { "epoch": 0.3211598746081505, "grad_norm": 2.363701581954956, "learning_rate": 9.84878109313456e-06, "loss": 1.143, "step": 4098 }, { "epoch": 0.3212382445141066, "grad_norm": 4.647615432739258, "learning_rate": 9.848705820272937e-06, "loss": 1.5964, "step": 4099 }, { "epoch": 0.3213166144200627, "grad_norm": 3.4698023796081543, "learning_rate": 9.848630528969326e-06, "loss": 1.1957, "step": 4100 }, { "epoch": 0.3213949843260188, "grad_norm": 3.1264538764953613, "learning_rate": 9.848555219224015e-06, "loss": 1.2541, "step": 4101 }, { "epoch": 0.32147335423197493, "grad_norm": 2.9958243370056152, "learning_rate": 9.84847989103729e-06, "loss": 1.2002, "step": 4102 }, { "epoch": 0.32155172413793104, "grad_norm": 5.683789253234863, "learning_rate": 9.84840454440944e-06, "loss": 0.671, "step": 4103 }, { "epoch": 0.32163009404388715, "grad_norm": 2.6823277473449707, "learning_rate": 9.848329179340747e-06, "loss": 1.4089, "step": 4104 }, { "epoch": 0.32170846394984326, "grad_norm": 5.527092456817627, "learning_rate": 9.848253795831501e-06, "loss": 1.4643, "step": 4105 }, { "epoch": 0.32178683385579937, "grad_norm": 3.0232133865356445, "learning_rate": 9.848178393881989e-06, "loss": 1.6407, "step": 4106 }, { "epoch": 0.3218652037617555, "grad_norm": 2.1524887084960938, "learning_rate": 9.848102973492495e-06, "loss": 1.402, "step": 4107 }, { "epoch": 0.3219435736677116, "grad_norm": 2.585631847381592, "learning_rate": 9.848027534663307e-06, "loss": 1.23, "step": 4108 }, { "epoch": 0.3220219435736677, "grad_norm": 5.05243444442749, "learning_rate": 9.847952077394713e-06, "loss": 1.5248, "step": 4109 }, { "epoch": 0.3221003134796238, "grad_norm": 2.929755687713623, "learning_rate": 9.847876601687e-06, "loss": 1.6704, "step": 4110 }, { "epoch": 0.3221786833855799, "grad_norm": 2.7133026123046875, "learning_rate": 9.847801107540453e-06, "loss": 1.4195, "step": 4111 }, { "epoch": 0.32225705329153603, "grad_norm": 3.6069576740264893, "learning_rate": 9.847725594955362e-06, "loss": 1.6092, "step": 4112 }, { "epoch": 0.32233542319749214, "grad_norm": 6.624096393585205, "learning_rate": 9.847650063932013e-06, "loss": 1.5854, "step": 4113 }, { "epoch": 0.32241379310344825, "grad_norm": 2.5223143100738525, "learning_rate": 9.847574514470693e-06, "loss": 1.506, "step": 4114 }, { "epoch": 0.32249216300940436, "grad_norm": 3.2392396926879883, "learning_rate": 9.847498946571688e-06, "loss": 1.3189, "step": 4115 }, { "epoch": 0.3225705329153605, "grad_norm": 4.478021144866943, "learning_rate": 9.847423360235287e-06, "loss": 1.2449, "step": 4116 }, { "epoch": 0.32264890282131664, "grad_norm": 4.909348487854004, "learning_rate": 9.847347755461777e-06, "loss": 2.0295, "step": 4117 }, { "epoch": 0.32272727272727275, "grad_norm": 3.92494797706604, "learning_rate": 9.847272132251448e-06, "loss": 1.3532, "step": 4118 }, { "epoch": 0.32280564263322886, "grad_norm": 4.941641330718994, "learning_rate": 9.847196490604585e-06, "loss": 1.3047, "step": 4119 }, { "epoch": 0.322884012539185, "grad_norm": 4.046041965484619, "learning_rate": 9.847120830521476e-06, "loss": 1.4579, "step": 4120 }, { "epoch": 0.3229623824451411, "grad_norm": 3.737442970275879, "learning_rate": 9.847045152002407e-06, "loss": 1.6471, "step": 4121 }, { "epoch": 0.3230407523510972, "grad_norm": 4.090200424194336, "learning_rate": 9.84696945504767e-06, "loss": 1.3439, "step": 4122 }, { "epoch": 0.3231191222570533, "grad_norm": 3.806107759475708, "learning_rate": 9.84689373965755e-06, "loss": 1.664, "step": 4123 }, { "epoch": 0.3231974921630094, "grad_norm": 2.3374183177948, "learning_rate": 9.846818005832336e-06, "loss": 1.4839, "step": 4124 }, { "epoch": 0.3232758620689655, "grad_norm": 3.3055219650268555, "learning_rate": 9.846742253572316e-06, "loss": 1.5548, "step": 4125 }, { "epoch": 0.32335423197492164, "grad_norm": 4.111630916595459, "learning_rate": 9.846666482877776e-06, "loss": 1.5041, "step": 4126 }, { "epoch": 0.32343260188087775, "grad_norm": 2.826061248779297, "learning_rate": 9.846590693749009e-06, "loss": 1.6003, "step": 4127 }, { "epoch": 0.32351097178683386, "grad_norm": 5.659841537475586, "learning_rate": 9.8465148861863e-06, "loss": 1.4319, "step": 4128 }, { "epoch": 0.32358934169278997, "grad_norm": 3.1310529708862305, "learning_rate": 9.846439060189935e-06, "loss": 1.5535, "step": 4129 }, { "epoch": 0.3236677115987461, "grad_norm": 2.2463996410369873, "learning_rate": 9.846363215760208e-06, "loss": 1.249, "step": 4130 }, { "epoch": 0.3237460815047022, "grad_norm": 2.546825885772705, "learning_rate": 9.846287352897404e-06, "loss": 1.0288, "step": 4131 }, { "epoch": 0.3238244514106583, "grad_norm": 4.436508655548096, "learning_rate": 9.84621147160181e-06, "loss": 1.3686, "step": 4132 }, { "epoch": 0.3239028213166144, "grad_norm": 3.076550006866455, "learning_rate": 9.846135571873718e-06, "loss": 1.5347, "step": 4133 }, { "epoch": 0.3239811912225705, "grad_norm": 2.611111640930176, "learning_rate": 9.846059653713417e-06, "loss": 1.1652, "step": 4134 }, { "epoch": 0.32405956112852663, "grad_norm": 2.3011744022369385, "learning_rate": 9.845983717121192e-06, "loss": 1.2505, "step": 4135 }, { "epoch": 0.32413793103448274, "grad_norm": 3.336074113845825, "learning_rate": 9.845907762097336e-06, "loss": 1.3129, "step": 4136 }, { "epoch": 0.32421630094043885, "grad_norm": 2.5638210773468018, "learning_rate": 9.845831788642134e-06, "loss": 1.5567, "step": 4137 }, { "epoch": 0.32429467084639496, "grad_norm": 3.67258882522583, "learning_rate": 9.845755796755879e-06, "loss": 1.6346, "step": 4138 }, { "epoch": 0.32437304075235107, "grad_norm": 5.315969944000244, "learning_rate": 9.845679786438855e-06, "loss": 0.9572, "step": 4139 }, { "epoch": 0.32445141065830724, "grad_norm": 2.177947759628296, "learning_rate": 9.845603757691355e-06, "loss": 1.3246, "step": 4140 }, { "epoch": 0.32452978056426335, "grad_norm": 4.096035003662109, "learning_rate": 9.845527710513669e-06, "loss": 1.6504, "step": 4141 }, { "epoch": 0.32460815047021946, "grad_norm": 4.290928840637207, "learning_rate": 9.845451644906083e-06, "loss": 1.8172, "step": 4142 }, { "epoch": 0.32468652037617557, "grad_norm": 2.3565597534179688, "learning_rate": 9.845375560868887e-06, "loss": 1.3314, "step": 4143 }, { "epoch": 0.3247648902821317, "grad_norm": 2.681446075439453, "learning_rate": 9.845299458402374e-06, "loss": 1.0402, "step": 4144 }, { "epoch": 0.3248432601880878, "grad_norm": 3.175675630569458, "learning_rate": 9.845223337506827e-06, "loss": 1.3649, "step": 4145 }, { "epoch": 0.3249216300940439, "grad_norm": 2.849299192428589, "learning_rate": 9.84514719818254e-06, "loss": 1.169, "step": 4146 }, { "epoch": 0.325, "grad_norm": 3.0462958812713623, "learning_rate": 9.845071040429803e-06, "loss": 1.0364, "step": 4147 }, { "epoch": 0.3250783699059561, "grad_norm": 2.37495756149292, "learning_rate": 9.844994864248902e-06, "loss": 1.4432, "step": 4148 }, { "epoch": 0.32515673981191223, "grad_norm": 2.834230422973633, "learning_rate": 9.844918669640132e-06, "loss": 1.1651, "step": 4149 }, { "epoch": 0.32523510971786834, "grad_norm": 3.0914459228515625, "learning_rate": 9.844842456603779e-06, "loss": 1.3741, "step": 4150 }, { "epoch": 0.32531347962382445, "grad_norm": 2.7664365768432617, "learning_rate": 9.844766225140132e-06, "loss": 1.2953, "step": 4151 }, { "epoch": 0.32539184952978056, "grad_norm": 2.3221709728240967, "learning_rate": 9.844689975249484e-06, "loss": 1.5859, "step": 4152 }, { "epoch": 0.3254702194357367, "grad_norm": 2.7739851474761963, "learning_rate": 9.844613706932125e-06, "loss": 1.1355, "step": 4153 }, { "epoch": 0.3255485893416928, "grad_norm": 2.9490792751312256, "learning_rate": 9.844537420188342e-06, "loss": 1.6743, "step": 4154 }, { "epoch": 0.3256269592476489, "grad_norm": 2.529212236404419, "learning_rate": 9.844461115018428e-06, "loss": 1.7703, "step": 4155 }, { "epoch": 0.325705329153605, "grad_norm": 3.4671623706817627, "learning_rate": 9.844384791422673e-06, "loss": 0.7072, "step": 4156 }, { "epoch": 0.3257836990595611, "grad_norm": 2.9577205181121826, "learning_rate": 9.844308449401364e-06, "loss": 1.3494, "step": 4157 }, { "epoch": 0.3258620689655172, "grad_norm": 2.1805028915405273, "learning_rate": 9.844232088954794e-06, "loss": 1.3048, "step": 4158 }, { "epoch": 0.32594043887147334, "grad_norm": 3.7995378971099854, "learning_rate": 9.844155710083256e-06, "loss": 0.7015, "step": 4159 }, { "epoch": 0.32601880877742945, "grad_norm": 2.6810548305511475, "learning_rate": 9.844079312787037e-06, "loss": 1.4973, "step": 4160 }, { "epoch": 0.32609717868338556, "grad_norm": 3.2212579250335693, "learning_rate": 9.844002897066426e-06, "loss": 1.6243, "step": 4161 }, { "epoch": 0.32617554858934167, "grad_norm": 5.343084812164307, "learning_rate": 9.843926462921719e-06, "loss": 1.0807, "step": 4162 }, { "epoch": 0.3262539184952978, "grad_norm": 2.6493730545043945, "learning_rate": 9.8438500103532e-06, "loss": 0.4494, "step": 4163 }, { "epoch": 0.32633228840125394, "grad_norm": 2.5841779708862305, "learning_rate": 9.843773539361165e-06, "loss": 1.2255, "step": 4164 }, { "epoch": 0.32641065830721006, "grad_norm": 2.609677314758301, "learning_rate": 9.843697049945906e-06, "loss": 1.3691, "step": 4165 }, { "epoch": 0.32648902821316617, "grad_norm": 3.3253378868103027, "learning_rate": 9.84362054210771e-06, "loss": 1.6614, "step": 4166 }, { "epoch": 0.3265673981191223, "grad_norm": 5.105845928192139, "learning_rate": 9.843544015846867e-06, "loss": 1.3621, "step": 4167 }, { "epoch": 0.3266457680250784, "grad_norm": 3.4016599655151367, "learning_rate": 9.843467471163672e-06, "loss": 1.0977, "step": 4168 }, { "epoch": 0.3267241379310345, "grad_norm": 2.106165647506714, "learning_rate": 9.843390908058416e-06, "loss": 1.4805, "step": 4169 }, { "epoch": 0.3268025078369906, "grad_norm": 2.808058977127075, "learning_rate": 9.843314326531387e-06, "loss": 1.4136, "step": 4170 }, { "epoch": 0.3268808777429467, "grad_norm": 3.598938226699829, "learning_rate": 9.843237726582879e-06, "loss": 1.4978, "step": 4171 }, { "epoch": 0.32695924764890283, "grad_norm": 2.318612575531006, "learning_rate": 9.843161108213181e-06, "loss": 1.4208, "step": 4172 }, { "epoch": 0.32703761755485894, "grad_norm": 2.49871826171875, "learning_rate": 9.843084471422586e-06, "loss": 1.47, "step": 4173 }, { "epoch": 0.32711598746081505, "grad_norm": 4.498166084289551, "learning_rate": 9.843007816211385e-06, "loss": 2.0275, "step": 4174 }, { "epoch": 0.32719435736677116, "grad_norm": 2.114668846130371, "learning_rate": 9.842931142579872e-06, "loss": 1.3207, "step": 4175 }, { "epoch": 0.32727272727272727, "grad_norm": 3.399056911468506, "learning_rate": 9.842854450528333e-06, "loss": 1.7544, "step": 4176 }, { "epoch": 0.3273510971786834, "grad_norm": 4.370422840118408, "learning_rate": 9.842777740057066e-06, "loss": 1.7487, "step": 4177 }, { "epoch": 0.3274294670846395, "grad_norm": 3.1833667755126953, "learning_rate": 9.842701011166358e-06, "loss": 1.0631, "step": 4178 }, { "epoch": 0.3275078369905956, "grad_norm": 8.834128379821777, "learning_rate": 9.842624263856505e-06, "loss": 1.5664, "step": 4179 }, { "epoch": 0.3275862068965517, "grad_norm": 3.7297229766845703, "learning_rate": 9.842547498127794e-06, "loss": 1.2145, "step": 4180 }, { "epoch": 0.3276645768025078, "grad_norm": 3.158944606781006, "learning_rate": 9.842470713980521e-06, "loss": 1.5951, "step": 4181 }, { "epoch": 0.32774294670846393, "grad_norm": 2.994420051574707, "learning_rate": 9.842393911414976e-06, "loss": 1.5268, "step": 4182 }, { "epoch": 0.32782131661442004, "grad_norm": 3.6015231609344482, "learning_rate": 9.842317090431454e-06, "loss": 1.345, "step": 4183 }, { "epoch": 0.32789968652037615, "grad_norm": 3.1368651390075684, "learning_rate": 9.842240251030243e-06, "loss": 1.6856, "step": 4184 }, { "epoch": 0.32797805642633227, "grad_norm": 2.699827194213867, "learning_rate": 9.842163393211636e-06, "loss": 1.1547, "step": 4185 }, { "epoch": 0.3280564263322884, "grad_norm": 9.715398788452148, "learning_rate": 9.842086516975928e-06, "loss": 1.9369, "step": 4186 }, { "epoch": 0.32813479623824454, "grad_norm": 4.77770471572876, "learning_rate": 9.84200962232341e-06, "loss": 1.6677, "step": 4187 }, { "epoch": 0.32821316614420065, "grad_norm": 3.443274736404419, "learning_rate": 9.841932709254376e-06, "loss": 1.4597, "step": 4188 }, { "epoch": 0.32829153605015676, "grad_norm": 2.661630392074585, "learning_rate": 9.841855777769113e-06, "loss": 1.4528, "step": 4189 }, { "epoch": 0.3283699059561129, "grad_norm": 3.747865676879883, "learning_rate": 9.84177882786792e-06, "loss": 1.2521, "step": 4190 }, { "epoch": 0.328448275862069, "grad_norm": 4.077881336212158, "learning_rate": 9.841701859551088e-06, "loss": 1.6066, "step": 4191 }, { "epoch": 0.3285266457680251, "grad_norm": 6.267609596252441, "learning_rate": 9.841624872818908e-06, "loss": 1.7994, "step": 4192 }, { "epoch": 0.3286050156739812, "grad_norm": 2.472435235977173, "learning_rate": 9.841547867671675e-06, "loss": 1.424, "step": 4193 }, { "epoch": 0.3286833855799373, "grad_norm": 5.025116443634033, "learning_rate": 9.84147084410968e-06, "loss": 1.2692, "step": 4194 }, { "epoch": 0.3287617554858934, "grad_norm": 2.412902355194092, "learning_rate": 9.841393802133214e-06, "loss": 1.3342, "step": 4195 }, { "epoch": 0.32884012539184954, "grad_norm": 3.632664442062378, "learning_rate": 9.841316741742577e-06, "loss": 1.5885, "step": 4196 }, { "epoch": 0.32891849529780565, "grad_norm": 6.494612693786621, "learning_rate": 9.841239662938054e-06, "loss": 2.7029, "step": 4197 }, { "epoch": 0.32899686520376176, "grad_norm": 3.78413462638855, "learning_rate": 9.841162565719944e-06, "loss": 1.3761, "step": 4198 }, { "epoch": 0.32907523510971787, "grad_norm": 4.083627223968506, "learning_rate": 9.841085450088538e-06, "loss": 1.5723, "step": 4199 }, { "epoch": 0.329153605015674, "grad_norm": 5.785159587860107, "learning_rate": 9.84100831604413e-06, "loss": 1.1643, "step": 4200 }, { "epoch": 0.3292319749216301, "grad_norm": 4.287812232971191, "learning_rate": 9.840931163587011e-06, "loss": 0.9991, "step": 4201 }, { "epoch": 0.3293103448275862, "grad_norm": 4.757543563842773, "learning_rate": 9.840853992717477e-06, "loss": 2.2186, "step": 4202 }, { "epoch": 0.3293887147335423, "grad_norm": 3.801666736602783, "learning_rate": 9.840776803435821e-06, "loss": 1.3211, "step": 4203 }, { "epoch": 0.3294670846394984, "grad_norm": 2.640524387359619, "learning_rate": 9.840699595742337e-06, "loss": 1.5898, "step": 4204 }, { "epoch": 0.32954545454545453, "grad_norm": 3.4038641452789307, "learning_rate": 9.840622369637318e-06, "loss": 1.5896, "step": 4205 }, { "epoch": 0.32962382445141064, "grad_norm": 2.3172085285186768, "learning_rate": 9.840545125121057e-06, "loss": 1.2328, "step": 4206 }, { "epoch": 0.32970219435736675, "grad_norm": 2.4487593173980713, "learning_rate": 9.840467862193849e-06, "loss": 1.5694, "step": 4207 }, { "epoch": 0.32978056426332286, "grad_norm": 3.294912338256836, "learning_rate": 9.840390580855989e-06, "loss": 0.8779, "step": 4208 }, { "epoch": 0.329858934169279, "grad_norm": 3.0368571281433105, "learning_rate": 9.840313281107767e-06, "loss": 1.3614, "step": 4209 }, { "epoch": 0.3299373040752351, "grad_norm": 2.057537078857422, "learning_rate": 9.84023596294948e-06, "loss": 1.2519, "step": 4210 }, { "epoch": 0.33001567398119125, "grad_norm": 2.8062186241149902, "learning_rate": 9.840158626381421e-06, "loss": 1.2752, "step": 4211 }, { "epoch": 0.33009404388714736, "grad_norm": 2.048182487487793, "learning_rate": 9.840081271403887e-06, "loss": 1.5443, "step": 4212 }, { "epoch": 0.33017241379310347, "grad_norm": 3.865508556365967, "learning_rate": 9.840003898017169e-06, "loss": 1.6001, "step": 4213 }, { "epoch": 0.3302507836990596, "grad_norm": 3.9128923416137695, "learning_rate": 9.839926506221562e-06, "loss": 1.4613, "step": 4214 }, { "epoch": 0.3303291536050157, "grad_norm": 2.8857696056365967, "learning_rate": 9.83984909601736e-06, "loss": 1.6628, "step": 4215 }, { "epoch": 0.3304075235109718, "grad_norm": 3.2834572792053223, "learning_rate": 9.839771667404858e-06, "loss": 1.62, "step": 4216 }, { "epoch": 0.3304858934169279, "grad_norm": 4.139876365661621, "learning_rate": 9.839694220384351e-06, "loss": 1.2898, "step": 4217 }, { "epoch": 0.330564263322884, "grad_norm": 3.2775461673736572, "learning_rate": 9.839616754956131e-06, "loss": 1.3615, "step": 4218 }, { "epoch": 0.33064263322884013, "grad_norm": 1.7208483219146729, "learning_rate": 9.839539271120497e-06, "loss": 1.415, "step": 4219 }, { "epoch": 0.33072100313479624, "grad_norm": 2.884190082550049, "learning_rate": 9.83946176887774e-06, "loss": 1.2154, "step": 4220 }, { "epoch": 0.33079937304075235, "grad_norm": 2.5672457218170166, "learning_rate": 9.839384248228156e-06, "loss": 1.4596, "step": 4221 }, { "epoch": 0.33087774294670846, "grad_norm": 4.34293794631958, "learning_rate": 9.839306709172042e-06, "loss": 0.9472, "step": 4222 }, { "epoch": 0.3309561128526646, "grad_norm": 2.58943772315979, "learning_rate": 9.83922915170969e-06, "loss": 1.3595, "step": 4223 }, { "epoch": 0.3310344827586207, "grad_norm": 3.31882905960083, "learning_rate": 9.839151575841395e-06, "loss": 0.8459, "step": 4224 }, { "epoch": 0.3311128526645768, "grad_norm": 3.246044158935547, "learning_rate": 9.839073981567453e-06, "loss": 1.3526, "step": 4225 }, { "epoch": 0.3311912225705329, "grad_norm": 2.8780152797698975, "learning_rate": 9.83899636888816e-06, "loss": 1.6632, "step": 4226 }, { "epoch": 0.331269592476489, "grad_norm": 5.415517807006836, "learning_rate": 9.83891873780381e-06, "loss": 1.7064, "step": 4227 }, { "epoch": 0.3313479623824451, "grad_norm": 3.698549270629883, "learning_rate": 9.838841088314696e-06, "loss": 0.8871, "step": 4228 }, { "epoch": 0.33142633228840124, "grad_norm": 4.93413782119751, "learning_rate": 9.83876342042112e-06, "loss": 1.2271, "step": 4229 }, { "epoch": 0.33150470219435735, "grad_norm": 2.1269662380218506, "learning_rate": 9.83868573412337e-06, "loss": 1.4648, "step": 4230 }, { "epoch": 0.33158307210031346, "grad_norm": 2.423896551132202, "learning_rate": 9.838608029421748e-06, "loss": 1.4489, "step": 4231 }, { "epoch": 0.33166144200626957, "grad_norm": 6.502945899963379, "learning_rate": 9.838530306316544e-06, "loss": 1.3682, "step": 4232 }, { "epoch": 0.3317398119122257, "grad_norm": 4.156872272491455, "learning_rate": 9.838452564808057e-06, "loss": 1.6435, "step": 4233 }, { "epoch": 0.33181818181818185, "grad_norm": 3.2995376586914062, "learning_rate": 9.83837480489658e-06, "loss": 1.3029, "step": 4234 }, { "epoch": 0.33189655172413796, "grad_norm": 3.1630337238311768, "learning_rate": 9.838297026582413e-06, "loss": 1.3315, "step": 4235 }, { "epoch": 0.33197492163009407, "grad_norm": 13.904032707214355, "learning_rate": 9.838219229865847e-06, "loss": 2.4593, "step": 4236 }, { "epoch": 0.3320532915360502, "grad_norm": 2.1429710388183594, "learning_rate": 9.838141414747181e-06, "loss": 1.3373, "step": 4237 }, { "epoch": 0.3321316614420063, "grad_norm": 3.011254072189331, "learning_rate": 9.83806358122671e-06, "loss": 1.4312, "step": 4238 }, { "epoch": 0.3322100313479624, "grad_norm": 4.392909049987793, "learning_rate": 9.83798572930473e-06, "loss": 1.273, "step": 4239 }, { "epoch": 0.3322884012539185, "grad_norm": 3.314548969268799, "learning_rate": 9.837907858981536e-06, "loss": 1.226, "step": 4240 }, { "epoch": 0.3323667711598746, "grad_norm": 2.9152257442474365, "learning_rate": 9.837829970257429e-06, "loss": 1.299, "step": 4241 }, { "epoch": 0.33244514106583073, "grad_norm": 2.872999668121338, "learning_rate": 9.837752063132699e-06, "loss": 1.5222, "step": 4242 }, { "epoch": 0.33252351097178684, "grad_norm": 4.4019246101379395, "learning_rate": 9.837674137607644e-06, "loss": 1.1827, "step": 4243 }, { "epoch": 0.33260188087774295, "grad_norm": 5.252461910247803, "learning_rate": 9.837596193682563e-06, "loss": 1.4306, "step": 4244 }, { "epoch": 0.33268025078369906, "grad_norm": 3.330709457397461, "learning_rate": 9.83751823135775e-06, "loss": 1.2597, "step": 4245 }, { "epoch": 0.33275862068965517, "grad_norm": 5.849123001098633, "learning_rate": 9.837440250633503e-06, "loss": 1.7052, "step": 4246 }, { "epoch": 0.3328369905956113, "grad_norm": 3.172888994216919, "learning_rate": 9.83736225151012e-06, "loss": 1.7126, "step": 4247 }, { "epoch": 0.3329153605015674, "grad_norm": 5.3799333572387695, "learning_rate": 9.83728423398789e-06, "loss": 1.1264, "step": 4248 }, { "epoch": 0.3329937304075235, "grad_norm": 4.063713073730469, "learning_rate": 9.837206198067123e-06, "loss": 1.7374, "step": 4249 }, { "epoch": 0.3330721003134796, "grad_norm": 7.026551246643066, "learning_rate": 9.837128143748102e-06, "loss": 1.9418, "step": 4250 }, { "epoch": 0.3331504702194357, "grad_norm": 2.6930465698242188, "learning_rate": 9.837050071031132e-06, "loss": 1.2408, "step": 4251 }, { "epoch": 0.33322884012539183, "grad_norm": 4.256290435791016, "learning_rate": 9.836971979916508e-06, "loss": 1.5122, "step": 4252 }, { "epoch": 0.33330721003134794, "grad_norm": 2.9557926654815674, "learning_rate": 9.836893870404527e-06, "loss": 1.0933, "step": 4253 }, { "epoch": 0.33338557993730406, "grad_norm": 3.3287765979766846, "learning_rate": 9.836815742495486e-06, "loss": 1.5272, "step": 4254 }, { "epoch": 0.33346394984326017, "grad_norm": 6.889303684234619, "learning_rate": 9.836737596189683e-06, "loss": 0.9281, "step": 4255 }, { "epoch": 0.3335423197492163, "grad_norm": 2.8029723167419434, "learning_rate": 9.836659431487414e-06, "loss": 1.3951, "step": 4256 }, { "epoch": 0.3336206896551724, "grad_norm": 2.334280490875244, "learning_rate": 9.836581248388976e-06, "loss": 1.4535, "step": 4257 }, { "epoch": 0.33369905956112855, "grad_norm": 4.3227152824401855, "learning_rate": 9.836503046894668e-06, "loss": 1.3764, "step": 4258 }, { "epoch": 0.33377742946708466, "grad_norm": 6.215192794799805, "learning_rate": 9.836424827004787e-06, "loss": 1.1138, "step": 4259 }, { "epoch": 0.3338557993730408, "grad_norm": 2.721590280532837, "learning_rate": 9.83634658871963e-06, "loss": 1.3063, "step": 4260 }, { "epoch": 0.3339341692789969, "grad_norm": 2.3551995754241943, "learning_rate": 9.836268332039494e-06, "loss": 1.0107, "step": 4261 }, { "epoch": 0.334012539184953, "grad_norm": 2.7937121391296387, "learning_rate": 9.836190056964677e-06, "loss": 1.2758, "step": 4262 }, { "epoch": 0.3340909090909091, "grad_norm": 9.812662124633789, "learning_rate": 9.836111763495477e-06, "loss": 1.0369, "step": 4263 }, { "epoch": 0.3341692789968652, "grad_norm": 2.4387102127075195, "learning_rate": 9.836033451632193e-06, "loss": 1.6981, "step": 4264 }, { "epoch": 0.3342476489028213, "grad_norm": 4.375629425048828, "learning_rate": 9.835955121375121e-06, "loss": 1.6128, "step": 4265 }, { "epoch": 0.33432601880877744, "grad_norm": 2.8441498279571533, "learning_rate": 9.835876772724559e-06, "loss": 1.3757, "step": 4266 }, { "epoch": 0.33440438871473355, "grad_norm": 2.262005090713501, "learning_rate": 9.835798405680807e-06, "loss": 1.2482, "step": 4267 }, { "epoch": 0.33448275862068966, "grad_norm": 3.2584266662597656, "learning_rate": 9.835720020244161e-06, "loss": 1.4675, "step": 4268 }, { "epoch": 0.33456112852664577, "grad_norm": 2.9756596088409424, "learning_rate": 9.83564161641492e-06, "loss": 1.1439, "step": 4269 }, { "epoch": 0.3346394984326019, "grad_norm": 5.3268914222717285, "learning_rate": 9.835563194193382e-06, "loss": 1.3059, "step": 4270 }, { "epoch": 0.334717868338558, "grad_norm": 3.8878366947174072, "learning_rate": 9.835484753579845e-06, "loss": 1.2788, "step": 4271 }, { "epoch": 0.3347962382445141, "grad_norm": 5.860844135284424, "learning_rate": 9.835406294574609e-06, "loss": 1.4885, "step": 4272 }, { "epoch": 0.3348746081504702, "grad_norm": 3.595400094985962, "learning_rate": 9.83532781717797e-06, "loss": 1.7044, "step": 4273 }, { "epoch": 0.3349529780564263, "grad_norm": 4.82033109664917, "learning_rate": 9.835249321390227e-06, "loss": 1.7003, "step": 4274 }, { "epoch": 0.33503134796238243, "grad_norm": 2.3959109783172607, "learning_rate": 9.83517080721168e-06, "loss": 1.1262, "step": 4275 }, { "epoch": 0.33510971786833854, "grad_norm": 2.8631691932678223, "learning_rate": 9.835092274642625e-06, "loss": 1.5039, "step": 4276 }, { "epoch": 0.33518808777429465, "grad_norm": 4.424671173095703, "learning_rate": 9.835013723683365e-06, "loss": 1.3691, "step": 4277 }, { "epoch": 0.33526645768025076, "grad_norm": 6.812714576721191, "learning_rate": 9.834935154334196e-06, "loss": 1.9533, "step": 4278 }, { "epoch": 0.3353448275862069, "grad_norm": 4.3521504402160645, "learning_rate": 9.834856566595415e-06, "loss": 1.464, "step": 4279 }, { "epoch": 0.335423197492163, "grad_norm": 2.605865716934204, "learning_rate": 9.834777960467325e-06, "loss": 1.1994, "step": 4280 }, { "epoch": 0.33550156739811915, "grad_norm": 2.8089311122894287, "learning_rate": 9.834699335950223e-06, "loss": 1.553, "step": 4281 }, { "epoch": 0.33557993730407526, "grad_norm": 2.8525242805480957, "learning_rate": 9.834620693044406e-06, "loss": 1.4563, "step": 4282 }, { "epoch": 0.33565830721003137, "grad_norm": 7.96323823928833, "learning_rate": 9.834542031750177e-06, "loss": 1.6282, "step": 4283 }, { "epoch": 0.3357366771159875, "grad_norm": 3.1621458530426025, "learning_rate": 9.834463352067831e-06, "loss": 1.3229, "step": 4284 }, { "epoch": 0.3358150470219436, "grad_norm": 4.2433295249938965, "learning_rate": 9.83438465399767e-06, "loss": 1.6274, "step": 4285 }, { "epoch": 0.3358934169278997, "grad_norm": 3.4612789154052734, "learning_rate": 9.834305937539994e-06, "loss": 1.4997, "step": 4286 }, { "epoch": 0.3359717868338558, "grad_norm": 2.2138636112213135, "learning_rate": 9.834227202695101e-06, "loss": 1.1985, "step": 4287 }, { "epoch": 0.3360501567398119, "grad_norm": 2.8877670764923096, "learning_rate": 9.834148449463293e-06, "loss": 1.6071, "step": 4288 }, { "epoch": 0.33612852664576803, "grad_norm": 8.63598346710205, "learning_rate": 9.834069677844863e-06, "loss": 1.2046, "step": 4289 }, { "epoch": 0.33620689655172414, "grad_norm": 6.535299777984619, "learning_rate": 9.833990887840117e-06, "loss": 1.1651, "step": 4290 }, { "epoch": 0.33628526645768025, "grad_norm": 2.4655816555023193, "learning_rate": 9.833912079449353e-06, "loss": 1.6664, "step": 4291 }, { "epoch": 0.33636363636363636, "grad_norm": 2.700852155685425, "learning_rate": 9.833833252672872e-06, "loss": 1.7253, "step": 4292 }, { "epoch": 0.3364420062695925, "grad_norm": 3.208678960800171, "learning_rate": 9.833754407510968e-06, "loss": 1.2524, "step": 4293 }, { "epoch": 0.3365203761755486, "grad_norm": 2.6653788089752197, "learning_rate": 9.833675543963948e-06, "loss": 1.2994, "step": 4294 }, { "epoch": 0.3365987460815047, "grad_norm": 3.133504867553711, "learning_rate": 9.833596662032108e-06, "loss": 1.2667, "step": 4295 }, { "epoch": 0.3366771159874608, "grad_norm": 3.8713364601135254, "learning_rate": 9.833517761715747e-06, "loss": 1.1831, "step": 4296 }, { "epoch": 0.3367554858934169, "grad_norm": 2.951111078262329, "learning_rate": 9.83343884301517e-06, "loss": 1.3952, "step": 4297 }, { "epoch": 0.336833855799373, "grad_norm": 2.7976489067077637, "learning_rate": 9.833359905930673e-06, "loss": 1.1175, "step": 4298 }, { "epoch": 0.33691222570532914, "grad_norm": 2.7807364463806152, "learning_rate": 9.833280950462558e-06, "loss": 1.3665, "step": 4299 }, { "epoch": 0.33699059561128525, "grad_norm": 3.1350209712982178, "learning_rate": 9.833201976611125e-06, "loss": 1.1644, "step": 4300 }, { "epoch": 0.33706896551724136, "grad_norm": 2.876833915710449, "learning_rate": 9.833122984376672e-06, "loss": 1.3273, "step": 4301 }, { "epoch": 0.33714733542319747, "grad_norm": 2.617234230041504, "learning_rate": 9.833043973759505e-06, "loss": 1.0751, "step": 4302 }, { "epoch": 0.3372257053291536, "grad_norm": 2.098407745361328, "learning_rate": 9.83296494475992e-06, "loss": 1.4902, "step": 4303 }, { "epoch": 0.3373040752351097, "grad_norm": 2.8346168994903564, "learning_rate": 9.832885897378217e-06, "loss": 1.2696, "step": 4304 }, { "epoch": 0.33738244514106586, "grad_norm": 10.738700866699219, "learning_rate": 9.8328068316147e-06, "loss": 0.9702, "step": 4305 }, { "epoch": 0.33746081504702197, "grad_norm": 2.3100526332855225, "learning_rate": 9.832727747469668e-06, "loss": 1.3855, "step": 4306 }, { "epoch": 0.3375391849529781, "grad_norm": 2.354093551635742, "learning_rate": 9.832648644943421e-06, "loss": 0.9411, "step": 4307 }, { "epoch": 0.3376175548589342, "grad_norm": 7.33248233795166, "learning_rate": 9.832569524036262e-06, "loss": 1.1007, "step": 4308 }, { "epoch": 0.3376959247648903, "grad_norm": 4.306289196014404, "learning_rate": 9.83249038474849e-06, "loss": 1.1433, "step": 4309 }, { "epoch": 0.3377742946708464, "grad_norm": 2.631274461746216, "learning_rate": 9.832411227080407e-06, "loss": 1.392, "step": 4310 }, { "epoch": 0.3378526645768025, "grad_norm": 3.2535557746887207, "learning_rate": 9.832332051032312e-06, "loss": 1.4917, "step": 4311 }, { "epoch": 0.33793103448275863, "grad_norm": 2.343142509460449, "learning_rate": 9.832252856604509e-06, "loss": 1.3906, "step": 4312 }, { "epoch": 0.33800940438871474, "grad_norm": 4.894644260406494, "learning_rate": 9.832173643797298e-06, "loss": 1.999, "step": 4313 }, { "epoch": 0.33808777429467085, "grad_norm": 2.182237148284912, "learning_rate": 9.832094412610983e-06, "loss": 1.4867, "step": 4314 }, { "epoch": 0.33816614420062696, "grad_norm": 2.5045812129974365, "learning_rate": 9.832015163045861e-06, "loss": 1.3155, "step": 4315 }, { "epoch": 0.33824451410658307, "grad_norm": 11.68407917022705, "learning_rate": 9.831935895102234e-06, "loss": 1.0643, "step": 4316 }, { "epoch": 0.3383228840125392, "grad_norm": 3.071803569793701, "learning_rate": 9.831856608780405e-06, "loss": 1.1987, "step": 4317 }, { "epoch": 0.3384012539184953, "grad_norm": 1.9558606147766113, "learning_rate": 9.831777304080676e-06, "loss": 1.3815, "step": 4318 }, { "epoch": 0.3384796238244514, "grad_norm": 2.747013568878174, "learning_rate": 9.831697981003349e-06, "loss": 1.0512, "step": 4319 }, { "epoch": 0.3385579937304075, "grad_norm": 2.9138827323913574, "learning_rate": 9.831618639548724e-06, "loss": 1.1914, "step": 4320 }, { "epoch": 0.3386363636363636, "grad_norm": 3.3544585704803467, "learning_rate": 9.831539279717102e-06, "loss": 1.8932, "step": 4321 }, { "epoch": 0.33871473354231973, "grad_norm": 3.1183342933654785, "learning_rate": 9.831459901508787e-06, "loss": 1.4793, "step": 4322 }, { "epoch": 0.33879310344827585, "grad_norm": 4.381711483001709, "learning_rate": 9.831380504924079e-06, "loss": 1.652, "step": 4323 }, { "epoch": 0.33887147335423196, "grad_norm": 1.7635838985443115, "learning_rate": 9.831301089963283e-06, "loss": 1.4567, "step": 4324 }, { "epoch": 0.33894984326018807, "grad_norm": 7.139678955078125, "learning_rate": 9.831221656626697e-06, "loss": 1.4444, "step": 4325 }, { "epoch": 0.3390282131661442, "grad_norm": 4.719095230102539, "learning_rate": 9.831142204914627e-06, "loss": 1.5812, "step": 4326 }, { "epoch": 0.3391065830721003, "grad_norm": 3.9705049991607666, "learning_rate": 9.831062734827374e-06, "loss": 1.6535, "step": 4327 }, { "epoch": 0.33918495297805645, "grad_norm": 2.2876787185668945, "learning_rate": 9.830983246365238e-06, "loss": 1.2151, "step": 4328 }, { "epoch": 0.33926332288401256, "grad_norm": 2.83034086227417, "learning_rate": 9.830903739528526e-06, "loss": 1.3213, "step": 4329 }, { "epoch": 0.3393416927899687, "grad_norm": 2.7633330821990967, "learning_rate": 9.830824214317533e-06, "loss": 1.5891, "step": 4330 }, { "epoch": 0.3394200626959248, "grad_norm": 2.6668214797973633, "learning_rate": 9.830744670732571e-06, "loss": 1.289, "step": 4331 }, { "epoch": 0.3394984326018809, "grad_norm": 3.0676236152648926, "learning_rate": 9.830665108773933e-06, "loss": 1.3652, "step": 4332 }, { "epoch": 0.339576802507837, "grad_norm": 5.169258117675781, "learning_rate": 9.830585528441929e-06, "loss": 1.6913, "step": 4333 }, { "epoch": 0.3396551724137931, "grad_norm": 2.2554707527160645, "learning_rate": 9.830505929736858e-06, "loss": 1.5566, "step": 4334 }, { "epoch": 0.3397335423197492, "grad_norm": 2.5721054077148438, "learning_rate": 9.830426312659023e-06, "loss": 1.6039, "step": 4335 }, { "epoch": 0.33981191222570534, "grad_norm": 3.2997024059295654, "learning_rate": 9.830346677208727e-06, "loss": 1.5153, "step": 4336 }, { "epoch": 0.33989028213166145, "grad_norm": 2.8451919555664062, "learning_rate": 9.830267023386276e-06, "loss": 1.0882, "step": 4337 }, { "epoch": 0.33996865203761756, "grad_norm": 3.5064797401428223, "learning_rate": 9.830187351191966e-06, "loss": 0.9586, "step": 4338 }, { "epoch": 0.34004702194357367, "grad_norm": 2.379481792449951, "learning_rate": 9.830107660626107e-06, "loss": 1.5396, "step": 4339 }, { "epoch": 0.3401253918495298, "grad_norm": 7.189563751220703, "learning_rate": 9.830027951688999e-06, "loss": 1.8329, "step": 4340 }, { "epoch": 0.3402037617554859, "grad_norm": 1.7286487817764282, "learning_rate": 9.829948224380946e-06, "loss": 1.1866, "step": 4341 }, { "epoch": 0.340282131661442, "grad_norm": 4.27324914932251, "learning_rate": 9.829868478702249e-06, "loss": 1.4639, "step": 4342 }, { "epoch": 0.3403605015673981, "grad_norm": 3.866568088531494, "learning_rate": 9.829788714653215e-06, "loss": 1.2979, "step": 4343 }, { "epoch": 0.3404388714733542, "grad_norm": 2.584230422973633, "learning_rate": 9.829708932234145e-06, "loss": 1.4823, "step": 4344 }, { "epoch": 0.34051724137931033, "grad_norm": 3.7145614624023438, "learning_rate": 9.829629131445342e-06, "loss": 0.9996, "step": 4345 }, { "epoch": 0.34059561128526644, "grad_norm": 2.3802292346954346, "learning_rate": 9.829549312287111e-06, "loss": 1.4953, "step": 4346 }, { "epoch": 0.34067398119122255, "grad_norm": 2.9556069374084473, "learning_rate": 9.829469474759756e-06, "loss": 1.3556, "step": 4347 }, { "epoch": 0.34075235109717866, "grad_norm": 2.2718582153320312, "learning_rate": 9.829389618863579e-06, "loss": 1.4078, "step": 4348 }, { "epoch": 0.3408307210031348, "grad_norm": 2.5883712768554688, "learning_rate": 9.829309744598883e-06, "loss": 1.5674, "step": 4349 }, { "epoch": 0.3409090909090909, "grad_norm": 3.8168795108795166, "learning_rate": 9.829229851965974e-06, "loss": 1.3459, "step": 4350 }, { "epoch": 0.340987460815047, "grad_norm": 2.2725565433502197, "learning_rate": 9.829149940965157e-06, "loss": 1.5106, "step": 4351 }, { "epoch": 0.34106583072100316, "grad_norm": 3.2920193672180176, "learning_rate": 9.829070011596732e-06, "loss": 1.2374, "step": 4352 }, { "epoch": 0.34114420062695927, "grad_norm": 11.128900527954102, "learning_rate": 9.828990063861005e-06, "loss": 1.1956, "step": 4353 }, { "epoch": 0.3412225705329154, "grad_norm": 5.9755682945251465, "learning_rate": 9.828910097758282e-06, "loss": 0.731, "step": 4354 }, { "epoch": 0.3413009404388715, "grad_norm": 2.2666544914245605, "learning_rate": 9.828830113288863e-06, "loss": 1.2279, "step": 4355 }, { "epoch": 0.3413793103448276, "grad_norm": 3.2116169929504395, "learning_rate": 9.828750110453055e-06, "loss": 1.0867, "step": 4356 }, { "epoch": 0.3414576802507837, "grad_norm": 3.068028688430786, "learning_rate": 9.828670089251163e-06, "loss": 1.6198, "step": 4357 }, { "epoch": 0.3415360501567398, "grad_norm": 3.9662997722625732, "learning_rate": 9.828590049683489e-06, "loss": 1.2445, "step": 4358 }, { "epoch": 0.34161442006269593, "grad_norm": 3.5123326778411865, "learning_rate": 9.82850999175034e-06, "loss": 1.3049, "step": 4359 }, { "epoch": 0.34169278996865204, "grad_norm": 2.920476198196411, "learning_rate": 9.828429915452018e-06, "loss": 1.0371, "step": 4360 }, { "epoch": 0.34177115987460815, "grad_norm": 2.253389835357666, "learning_rate": 9.828349820788827e-06, "loss": 1.6765, "step": 4361 }, { "epoch": 0.34184952978056427, "grad_norm": 2.7688589096069336, "learning_rate": 9.828269707761075e-06, "loss": 1.696, "step": 4362 }, { "epoch": 0.3419278996865204, "grad_norm": 3.7607498168945312, "learning_rate": 9.828189576369066e-06, "loss": 1.1515, "step": 4363 }, { "epoch": 0.3420062695924765, "grad_norm": 3.5546696186065674, "learning_rate": 9.828109426613102e-06, "loss": 1.6772, "step": 4364 }, { "epoch": 0.3420846394984326, "grad_norm": 3.6960926055908203, "learning_rate": 9.82802925849349e-06, "loss": 1.6168, "step": 4365 }, { "epoch": 0.3421630094043887, "grad_norm": 3.4318323135375977, "learning_rate": 9.827949072010537e-06, "loss": 1.4456, "step": 4366 }, { "epoch": 0.3422413793103448, "grad_norm": 3.259472131729126, "learning_rate": 9.827868867164542e-06, "loss": 1.2752, "step": 4367 }, { "epoch": 0.34231974921630093, "grad_norm": 3.0611746311187744, "learning_rate": 9.827788643955815e-06, "loss": 0.8457, "step": 4368 }, { "epoch": 0.34239811912225704, "grad_norm": 3.2542645931243896, "learning_rate": 9.82770840238466e-06, "loss": 1.3178, "step": 4369 }, { "epoch": 0.34247648902821315, "grad_norm": 2.144754409790039, "learning_rate": 9.827628142451383e-06, "loss": 1.578, "step": 4370 }, { "epoch": 0.34255485893416926, "grad_norm": 4.359289646148682, "learning_rate": 9.827547864156285e-06, "loss": 1.5695, "step": 4371 }, { "epoch": 0.34263322884012537, "grad_norm": 2.60404109954834, "learning_rate": 9.827467567499676e-06, "loss": 1.5002, "step": 4372 }, { "epoch": 0.3427115987460815, "grad_norm": 2.0000569820404053, "learning_rate": 9.827387252481861e-06, "loss": 1.3762, "step": 4373 }, { "epoch": 0.3427899686520376, "grad_norm": 2.9479517936706543, "learning_rate": 9.827306919103143e-06, "loss": 1.2109, "step": 4374 }, { "epoch": 0.34286833855799376, "grad_norm": 2.12136173248291, "learning_rate": 9.82722656736383e-06, "loss": 1.3706, "step": 4375 }, { "epoch": 0.34294670846394987, "grad_norm": 2.812828779220581, "learning_rate": 9.827146197264226e-06, "loss": 1.577, "step": 4376 }, { "epoch": 0.343025078369906, "grad_norm": 3.661017894744873, "learning_rate": 9.827065808804636e-06, "loss": 1.1144, "step": 4377 }, { "epoch": 0.3431034482758621, "grad_norm": 7.031426429748535, "learning_rate": 9.826985401985369e-06, "loss": 1.1232, "step": 4378 }, { "epoch": 0.3431818181818182, "grad_norm": 2.283341407775879, "learning_rate": 9.826904976806728e-06, "loss": 1.384, "step": 4379 }, { "epoch": 0.3432601880877743, "grad_norm": 3.3034443855285645, "learning_rate": 9.82682453326902e-06, "loss": 1.5303, "step": 4380 }, { "epoch": 0.3433385579937304, "grad_norm": 3.1095986366271973, "learning_rate": 9.826744071372552e-06, "loss": 1.4748, "step": 4381 }, { "epoch": 0.34341692789968653, "grad_norm": 2.2873008251190186, "learning_rate": 9.826663591117627e-06, "loss": 1.1767, "step": 4382 }, { "epoch": 0.34349529780564264, "grad_norm": 5.289648532867432, "learning_rate": 9.826583092504553e-06, "loss": 1.7603, "step": 4383 }, { "epoch": 0.34357366771159875, "grad_norm": 2.2698922157287598, "learning_rate": 9.826502575533635e-06, "loss": 1.4148, "step": 4384 }, { "epoch": 0.34365203761755486, "grad_norm": 2.458230495452881, "learning_rate": 9.826422040205181e-06, "loss": 1.3689, "step": 4385 }, { "epoch": 0.343730407523511, "grad_norm": 3.349599599838257, "learning_rate": 9.826341486519497e-06, "loss": 1.6983, "step": 4386 }, { "epoch": 0.3438087774294671, "grad_norm": 4.581561088562012, "learning_rate": 9.82626091447689e-06, "loss": 0.8834, "step": 4387 }, { "epoch": 0.3438871473354232, "grad_norm": 4.824923992156982, "learning_rate": 9.826180324077663e-06, "loss": 1.055, "step": 4388 }, { "epoch": 0.3439655172413793, "grad_norm": 6.597009658813477, "learning_rate": 9.826099715322127e-06, "loss": 1.1699, "step": 4389 }, { "epoch": 0.3440438871473354, "grad_norm": 5.376677513122559, "learning_rate": 9.826019088210586e-06, "loss": 1.756, "step": 4390 }, { "epoch": 0.3441222570532915, "grad_norm": 3.057138204574585, "learning_rate": 9.825938442743347e-06, "loss": 1.3336, "step": 4391 }, { "epoch": 0.34420062695924764, "grad_norm": 4.4522247314453125, "learning_rate": 9.825857778920717e-06, "loss": 1.5041, "step": 4392 }, { "epoch": 0.34427899686520375, "grad_norm": 2.4496614933013916, "learning_rate": 9.825777096743003e-06, "loss": 1.1712, "step": 4393 }, { "epoch": 0.34435736677115986, "grad_norm": 2.3842616081237793, "learning_rate": 9.82569639621051e-06, "loss": 1.4993, "step": 4394 }, { "epoch": 0.34443573667711597, "grad_norm": 2.6332945823669434, "learning_rate": 9.825615677323548e-06, "loss": 1.1935, "step": 4395 }, { "epoch": 0.3445141065830721, "grad_norm": 2.30155348777771, "learning_rate": 9.825534940082423e-06, "loss": 1.3703, "step": 4396 }, { "epoch": 0.3445924764890282, "grad_norm": 2.1421561241149902, "learning_rate": 9.82545418448744e-06, "loss": 1.5201, "step": 4397 }, { "epoch": 0.34467084639498435, "grad_norm": 3.343414545059204, "learning_rate": 9.82537341053891e-06, "loss": 1.0747, "step": 4398 }, { "epoch": 0.34474921630094046, "grad_norm": 2.8700733184814453, "learning_rate": 9.825292618237136e-06, "loss": 1.2739, "step": 4399 }, { "epoch": 0.3448275862068966, "grad_norm": 2.6749908924102783, "learning_rate": 9.825211807582427e-06, "loss": 1.4034, "step": 4400 }, { "epoch": 0.3449059561128527, "grad_norm": 4.34603214263916, "learning_rate": 9.825130978575093e-06, "loss": 1.0996, "step": 4401 }, { "epoch": 0.3449843260188088, "grad_norm": 2.071242570877075, "learning_rate": 9.825050131215438e-06, "loss": 1.3174, "step": 4402 }, { "epoch": 0.3450626959247649, "grad_norm": 4.377833366394043, "learning_rate": 9.82496926550377e-06, "loss": 1.6869, "step": 4403 }, { "epoch": 0.345141065830721, "grad_norm": 2.916020154953003, "learning_rate": 9.824888381440397e-06, "loss": 1.0795, "step": 4404 }, { "epoch": 0.3452194357366771, "grad_norm": 2.578608512878418, "learning_rate": 9.824807479025628e-06, "loss": 0.9166, "step": 4405 }, { "epoch": 0.34529780564263324, "grad_norm": 3.0469565391540527, "learning_rate": 9.824726558259768e-06, "loss": 1.6744, "step": 4406 }, { "epoch": 0.34537617554858935, "grad_norm": 2.8641583919525146, "learning_rate": 9.824645619143128e-06, "loss": 1.6068, "step": 4407 }, { "epoch": 0.34545454545454546, "grad_norm": 2.1237895488739014, "learning_rate": 9.824564661676013e-06, "loss": 1.1021, "step": 4408 }, { "epoch": 0.34553291536050157, "grad_norm": 2.3339202404022217, "learning_rate": 9.824483685858733e-06, "loss": 1.3609, "step": 4409 }, { "epoch": 0.3456112852664577, "grad_norm": 2.872755527496338, "learning_rate": 9.824402691691595e-06, "loss": 1.3486, "step": 4410 }, { "epoch": 0.3456896551724138, "grad_norm": 4.013118743896484, "learning_rate": 9.824321679174906e-06, "loss": 1.3979, "step": 4411 }, { "epoch": 0.3457680250783699, "grad_norm": 2.8457462787628174, "learning_rate": 9.824240648308976e-06, "loss": 1.4235, "step": 4412 }, { "epoch": 0.345846394984326, "grad_norm": 3.916011333465576, "learning_rate": 9.824159599094113e-06, "loss": 0.762, "step": 4413 }, { "epoch": 0.3459247648902821, "grad_norm": 3.6940877437591553, "learning_rate": 9.824078531530623e-06, "loss": 1.2507, "step": 4414 }, { "epoch": 0.34600313479623823, "grad_norm": 3.3870840072631836, "learning_rate": 9.823997445618819e-06, "loss": 1.0498, "step": 4415 }, { "epoch": 0.34608150470219434, "grad_norm": 3.1406307220458984, "learning_rate": 9.823916341359004e-06, "loss": 1.2137, "step": 4416 }, { "epoch": 0.34615987460815045, "grad_norm": 9.512721061706543, "learning_rate": 9.82383521875149e-06, "loss": 1.7805, "step": 4417 }, { "epoch": 0.34623824451410656, "grad_norm": 3.4126393795013428, "learning_rate": 9.823754077796583e-06, "loss": 1.1844, "step": 4418 }, { "epoch": 0.3463166144200627, "grad_norm": 4.070539474487305, "learning_rate": 9.823672918494595e-06, "loss": 1.21, "step": 4419 }, { "epoch": 0.3463949843260188, "grad_norm": 2.5636796951293945, "learning_rate": 9.823591740845831e-06, "loss": 1.5033, "step": 4420 }, { "epoch": 0.3464733542319749, "grad_norm": 12.5818452835083, "learning_rate": 9.823510544850603e-06, "loss": 1.9268, "step": 4421 }, { "epoch": 0.34655172413793106, "grad_norm": 2.610877513885498, "learning_rate": 9.823429330509218e-06, "loss": 1.1142, "step": 4422 }, { "epoch": 0.34663009404388717, "grad_norm": 3.499335289001465, "learning_rate": 9.823348097821986e-06, "loss": 1.2889, "step": 4423 }, { "epoch": 0.3467084639498433, "grad_norm": 6.570889949798584, "learning_rate": 9.823266846789215e-06, "loss": 1.5372, "step": 4424 }, { "epoch": 0.3467868338557994, "grad_norm": 3.5825250148773193, "learning_rate": 9.823185577411213e-06, "loss": 1.5151, "step": 4425 }, { "epoch": 0.3468652037617555, "grad_norm": 2.49735426902771, "learning_rate": 9.82310428968829e-06, "loss": 1.3801, "step": 4426 }, { "epoch": 0.3469435736677116, "grad_norm": 7.933076858520508, "learning_rate": 9.823022983620758e-06, "loss": 1.0564, "step": 4427 }, { "epoch": 0.3470219435736677, "grad_norm": 3.2138214111328125, "learning_rate": 9.822941659208922e-06, "loss": 1.3599, "step": 4428 }, { "epoch": 0.34710031347962383, "grad_norm": 2.773406744003296, "learning_rate": 9.822860316453096e-06, "loss": 1.6246, "step": 4429 }, { "epoch": 0.34717868338557994, "grad_norm": 2.81518816947937, "learning_rate": 9.822778955353582e-06, "loss": 1.5369, "step": 4430 }, { "epoch": 0.34725705329153606, "grad_norm": 3.432668447494507, "learning_rate": 9.822697575910698e-06, "loss": 1.8723, "step": 4431 }, { "epoch": 0.34733542319749217, "grad_norm": 2.920900583267212, "learning_rate": 9.822616178124747e-06, "loss": 1.3495, "step": 4432 }, { "epoch": 0.3474137931034483, "grad_norm": 1.9400790929794312, "learning_rate": 9.822534761996043e-06, "loss": 1.1408, "step": 4433 }, { "epoch": 0.3474921630094044, "grad_norm": 3.0408101081848145, "learning_rate": 9.822453327524892e-06, "loss": 1.5158, "step": 4434 }, { "epoch": 0.3475705329153605, "grad_norm": 3.904550790786743, "learning_rate": 9.822371874711607e-06, "loss": 1.3018, "step": 4435 }, { "epoch": 0.3476489028213166, "grad_norm": 3.1498236656188965, "learning_rate": 9.822290403556496e-06, "loss": 0.9506, "step": 4436 }, { "epoch": 0.3477272727272727, "grad_norm": 7.919795036315918, "learning_rate": 9.822208914059868e-06, "loss": 1.7414, "step": 4437 }, { "epoch": 0.34780564263322883, "grad_norm": 3.7362492084503174, "learning_rate": 9.822127406222035e-06, "loss": 1.3089, "step": 4438 }, { "epoch": 0.34788401253918494, "grad_norm": 3.407703399658203, "learning_rate": 9.822045880043307e-06, "loss": 1.5828, "step": 4439 }, { "epoch": 0.34796238244514105, "grad_norm": 2.2294089794158936, "learning_rate": 9.82196433552399e-06, "loss": 1.1498, "step": 4440 }, { "epoch": 0.34804075235109716, "grad_norm": 3.029014825820923, "learning_rate": 9.821882772664402e-06, "loss": 1.4163, "step": 4441 }, { "epoch": 0.34811912225705327, "grad_norm": 4.050386428833008, "learning_rate": 9.821801191464846e-06, "loss": 1.1694, "step": 4442 }, { "epoch": 0.3481974921630094, "grad_norm": 2.55092716217041, "learning_rate": 9.821719591925634e-06, "loss": 1.3244, "step": 4443 }, { "epoch": 0.3482758620689655, "grad_norm": 2.8293538093566895, "learning_rate": 9.82163797404708e-06, "loss": 1.089, "step": 4444 }, { "epoch": 0.34835423197492166, "grad_norm": 2.446345090866089, "learning_rate": 9.82155633782949e-06, "loss": 1.2629, "step": 4445 }, { "epoch": 0.34843260188087777, "grad_norm": 2.203803539276123, "learning_rate": 9.821474683273175e-06, "loss": 1.3577, "step": 4446 }, { "epoch": 0.3485109717868339, "grad_norm": 4.357151508331299, "learning_rate": 9.821393010378449e-06, "loss": 1.0154, "step": 4447 }, { "epoch": 0.34858934169279, "grad_norm": 7.529298305511475, "learning_rate": 9.821311319145618e-06, "loss": 1.8876, "step": 4448 }, { "epoch": 0.3486677115987461, "grad_norm": 3.0164849758148193, "learning_rate": 9.821229609574998e-06, "loss": 1.1595, "step": 4449 }, { "epoch": 0.3487460815047022, "grad_norm": 2.782052993774414, "learning_rate": 9.821147881666896e-06, "loss": 1.1998, "step": 4450 }, { "epoch": 0.3488244514106583, "grad_norm": 3.130690813064575, "learning_rate": 9.821066135421624e-06, "loss": 1.0, "step": 4451 }, { "epoch": 0.34890282131661443, "grad_norm": 6.668652534484863, "learning_rate": 9.820984370839492e-06, "loss": 1.6531, "step": 4452 }, { "epoch": 0.34898119122257054, "grad_norm": 2.6301894187927246, "learning_rate": 9.820902587920812e-06, "loss": 1.343, "step": 4453 }, { "epoch": 0.34905956112852665, "grad_norm": 6.45269250869751, "learning_rate": 9.820820786665894e-06, "loss": 1.5026, "step": 4454 }, { "epoch": 0.34913793103448276, "grad_norm": 1.9861178398132324, "learning_rate": 9.820738967075049e-06, "loss": 0.9713, "step": 4455 }, { "epoch": 0.3492163009404389, "grad_norm": 2.9649696350097656, "learning_rate": 9.82065712914859e-06, "loss": 1.1614, "step": 4456 }, { "epoch": 0.349294670846395, "grad_norm": 5.267320156097412, "learning_rate": 9.820575272886828e-06, "loss": 1.8319, "step": 4457 }, { "epoch": 0.3493730407523511, "grad_norm": 2.3741862773895264, "learning_rate": 9.820493398290073e-06, "loss": 1.4612, "step": 4458 }, { "epoch": 0.3494514106583072, "grad_norm": 3.0204358100891113, "learning_rate": 9.820411505358639e-06, "loss": 0.8929, "step": 4459 }, { "epoch": 0.3495297805642633, "grad_norm": 3.0782439708709717, "learning_rate": 9.820329594092833e-06, "loss": 1.7238, "step": 4460 }, { "epoch": 0.3496081504702194, "grad_norm": 5.922768592834473, "learning_rate": 9.82024766449297e-06, "loss": 0.8503, "step": 4461 }, { "epoch": 0.34968652037617554, "grad_norm": 2.586599588394165, "learning_rate": 9.82016571655936e-06, "loss": 1.4329, "step": 4462 }, { "epoch": 0.34976489028213165, "grad_norm": 3.621654987335205, "learning_rate": 9.820083750292318e-06, "loss": 1.0768, "step": 4463 }, { "epoch": 0.34984326018808776, "grad_norm": 3.388201951980591, "learning_rate": 9.820001765692152e-06, "loss": 1.4839, "step": 4464 }, { "epoch": 0.34992163009404387, "grad_norm": 3.8700790405273438, "learning_rate": 9.819919762759174e-06, "loss": 1.3508, "step": 4465 }, { "epoch": 0.35, "grad_norm": 3.4107887744903564, "learning_rate": 9.819837741493696e-06, "loss": 1.7908, "step": 4466 }, { "epoch": 0.3500783699059561, "grad_norm": 7.25215482711792, "learning_rate": 9.819755701896033e-06, "loss": 2.2185, "step": 4467 }, { "epoch": 0.3501567398119122, "grad_norm": 2.5466558933258057, "learning_rate": 9.819673643966495e-06, "loss": 1.3822, "step": 4468 }, { "epoch": 0.35023510971786836, "grad_norm": 2.552159070968628, "learning_rate": 9.819591567705392e-06, "loss": 1.1559, "step": 4469 }, { "epoch": 0.3503134796238245, "grad_norm": 1.8787686824798584, "learning_rate": 9.819509473113039e-06, "loss": 1.0964, "step": 4470 }, { "epoch": 0.3503918495297806, "grad_norm": 4.129881381988525, "learning_rate": 9.819427360189748e-06, "loss": 1.4132, "step": 4471 }, { "epoch": 0.3504702194357367, "grad_norm": 3.0407824516296387, "learning_rate": 9.81934522893583e-06, "loss": 1.3503, "step": 4472 }, { "epoch": 0.3505485893416928, "grad_norm": 4.414443016052246, "learning_rate": 9.819263079351598e-06, "loss": 1.3722, "step": 4473 }, { "epoch": 0.3506269592476489, "grad_norm": 2.891690492630005, "learning_rate": 9.819180911437364e-06, "loss": 1.4135, "step": 4474 }, { "epoch": 0.350705329153605, "grad_norm": 1.9018912315368652, "learning_rate": 9.819098725193443e-06, "loss": 1.3467, "step": 4475 }, { "epoch": 0.35078369905956114, "grad_norm": 3.61955189704895, "learning_rate": 9.819016520620145e-06, "loss": 1.68, "step": 4476 }, { "epoch": 0.35086206896551725, "grad_norm": 6.890259265899658, "learning_rate": 9.818934297717785e-06, "loss": 1.2093, "step": 4477 }, { "epoch": 0.35094043887147336, "grad_norm": 3.3973190784454346, "learning_rate": 9.81885205648667e-06, "loss": 1.119, "step": 4478 }, { "epoch": 0.35101880877742947, "grad_norm": 2.8519785404205322, "learning_rate": 9.818769796927118e-06, "loss": 0.911, "step": 4479 }, { "epoch": 0.3510971786833856, "grad_norm": 2.310035467147827, "learning_rate": 9.818687519039444e-06, "loss": 1.3714, "step": 4480 }, { "epoch": 0.3511755485893417, "grad_norm": 10.71178150177002, "learning_rate": 9.818605222823956e-06, "loss": 1.2153, "step": 4481 }, { "epoch": 0.3512539184952978, "grad_norm": 3.3093090057373047, "learning_rate": 9.818522908280967e-06, "loss": 1.4291, "step": 4482 }, { "epoch": 0.3513322884012539, "grad_norm": 2.6050541400909424, "learning_rate": 9.818440575410794e-06, "loss": 1.2606, "step": 4483 }, { "epoch": 0.35141065830721, "grad_norm": 3.5445640087127686, "learning_rate": 9.818358224213747e-06, "loss": 1.8417, "step": 4484 }, { "epoch": 0.35148902821316613, "grad_norm": 3.087989568710327, "learning_rate": 9.81827585469014e-06, "loss": 1.3378, "step": 4485 }, { "epoch": 0.35156739811912224, "grad_norm": 2.881751537322998, "learning_rate": 9.818193466840287e-06, "loss": 1.1717, "step": 4486 }, { "epoch": 0.35164576802507835, "grad_norm": 3.1948964595794678, "learning_rate": 9.8181110606645e-06, "loss": 1.5539, "step": 4487 }, { "epoch": 0.35172413793103446, "grad_norm": 2.827277421951294, "learning_rate": 9.818028636163095e-06, "loss": 1.6206, "step": 4488 }, { "epoch": 0.3518025078369906, "grad_norm": 3.941783905029297, "learning_rate": 9.817946193336384e-06, "loss": 1.5014, "step": 4489 }, { "epoch": 0.3518808777429467, "grad_norm": 3.3257431983947754, "learning_rate": 9.817863732184677e-06, "loss": 1.7083, "step": 4490 }, { "epoch": 0.3519592476489028, "grad_norm": 3.9979798793792725, "learning_rate": 9.817781252708293e-06, "loss": 1.368, "step": 4491 }, { "epoch": 0.35203761755485896, "grad_norm": 4.79638671875, "learning_rate": 9.817698754907544e-06, "loss": 1.5952, "step": 4492 }, { "epoch": 0.35211598746081507, "grad_norm": 2.3909902572631836, "learning_rate": 9.817616238782743e-06, "loss": 1.5744, "step": 4493 }, { "epoch": 0.3521943573667712, "grad_norm": 3.1861133575439453, "learning_rate": 9.817533704334204e-06, "loss": 1.6951, "step": 4494 }, { "epoch": 0.3522727272727273, "grad_norm": 2.5056071281433105, "learning_rate": 9.817451151562241e-06, "loss": 0.9937, "step": 4495 }, { "epoch": 0.3523510971786834, "grad_norm": 3.0156736373901367, "learning_rate": 9.81736858046717e-06, "loss": 1.491, "step": 4496 }, { "epoch": 0.3524294670846395, "grad_norm": 2.390441417694092, "learning_rate": 9.817285991049301e-06, "loss": 1.0182, "step": 4497 }, { "epoch": 0.3525078369905956, "grad_norm": 2.6581180095672607, "learning_rate": 9.817203383308953e-06, "loss": 1.6289, "step": 4498 }, { "epoch": 0.35258620689655173, "grad_norm": 2.1797823905944824, "learning_rate": 9.817120757246436e-06, "loss": 1.3903, "step": 4499 }, { "epoch": 0.35266457680250785, "grad_norm": 3.293776512145996, "learning_rate": 9.817038112862065e-06, "loss": 1.5327, "step": 4500 }, { "epoch": 0.35274294670846396, "grad_norm": 3.015568971633911, "learning_rate": 9.816955450156156e-06, "loss": 1.6725, "step": 4501 }, { "epoch": 0.35282131661442007, "grad_norm": 6.304378986358643, "learning_rate": 9.816872769129024e-06, "loss": 1.4289, "step": 4502 }, { "epoch": 0.3528996865203762, "grad_norm": 3.637939453125, "learning_rate": 9.81679006978098e-06, "loss": 1.1364, "step": 4503 }, { "epoch": 0.3529780564263323, "grad_norm": 3.6035075187683105, "learning_rate": 9.816707352112343e-06, "loss": 1.2853, "step": 4504 }, { "epoch": 0.3530564263322884, "grad_norm": 2.645592451095581, "learning_rate": 9.816624616123422e-06, "loss": 1.2669, "step": 4505 }, { "epoch": 0.3531347962382445, "grad_norm": 2.754312038421631, "learning_rate": 9.816541861814537e-06, "loss": 1.5752, "step": 4506 }, { "epoch": 0.3532131661442006, "grad_norm": 2.853238105773926, "learning_rate": 9.816459089186001e-06, "loss": 1.3359, "step": 4507 }, { "epoch": 0.35329153605015673, "grad_norm": 3.22650408744812, "learning_rate": 9.816376298238128e-06, "loss": 1.5756, "step": 4508 }, { "epoch": 0.35336990595611284, "grad_norm": 3.090535879135132, "learning_rate": 9.816293488971233e-06, "loss": 0.9804, "step": 4509 }, { "epoch": 0.35344827586206895, "grad_norm": 4.752029895782471, "learning_rate": 9.816210661385633e-06, "loss": 1.7989, "step": 4510 }, { "epoch": 0.35352664576802506, "grad_norm": 3.5315918922424316, "learning_rate": 9.81612781548164e-06, "loss": 1.1463, "step": 4511 }, { "epoch": 0.35360501567398117, "grad_norm": 5.434285640716553, "learning_rate": 9.81604495125957e-06, "loss": 1.5197, "step": 4512 }, { "epoch": 0.3536833855799373, "grad_norm": 3.066652774810791, "learning_rate": 9.81596206871974e-06, "loss": 1.4518, "step": 4513 }, { "epoch": 0.3537617554858934, "grad_norm": 2.7389914989471436, "learning_rate": 9.815879167862461e-06, "loss": 1.4066, "step": 4514 }, { "epoch": 0.3538401253918495, "grad_norm": 2.3735556602478027, "learning_rate": 9.815796248688054e-06, "loss": 1.6223, "step": 4515 }, { "epoch": 0.35391849529780567, "grad_norm": 5.429184436798096, "learning_rate": 9.815713311196832e-06, "loss": 1.0989, "step": 4516 }, { "epoch": 0.3539968652037618, "grad_norm": 3.6599740982055664, "learning_rate": 9.815630355389109e-06, "loss": 1.6095, "step": 4517 }, { "epoch": 0.3540752351097179, "grad_norm": 4.636540412902832, "learning_rate": 9.815547381265203e-06, "loss": 1.4161, "step": 4518 }, { "epoch": 0.354153605015674, "grad_norm": 2.2305245399475098, "learning_rate": 9.815464388825427e-06, "loss": 1.3103, "step": 4519 }, { "epoch": 0.3542319749216301, "grad_norm": 2.943556785583496, "learning_rate": 9.815381378070097e-06, "loss": 1.6739, "step": 4520 }, { "epoch": 0.3543103448275862, "grad_norm": 3.0398614406585693, "learning_rate": 9.81529834899953e-06, "loss": 1.2772, "step": 4521 }, { "epoch": 0.35438871473354233, "grad_norm": 2.3463215827941895, "learning_rate": 9.815215301614041e-06, "loss": 1.5152, "step": 4522 }, { "epoch": 0.35446708463949844, "grad_norm": 4.082169532775879, "learning_rate": 9.815132235913947e-06, "loss": 1.3359, "step": 4523 }, { "epoch": 0.35454545454545455, "grad_norm": 2.963181972503662, "learning_rate": 9.815049151899564e-06, "loss": 1.0709, "step": 4524 }, { "epoch": 0.35462382445141066, "grad_norm": 3.8064582347869873, "learning_rate": 9.814966049571206e-06, "loss": 1.6808, "step": 4525 }, { "epoch": 0.3547021943573668, "grad_norm": 2.3754680156707764, "learning_rate": 9.81488292892919e-06, "loss": 1.6135, "step": 4526 }, { "epoch": 0.3547805642633229, "grad_norm": 2.306459665298462, "learning_rate": 9.814799789973835e-06, "loss": 1.4443, "step": 4527 }, { "epoch": 0.354858934169279, "grad_norm": 2.867459297180176, "learning_rate": 9.814716632705453e-06, "loss": 1.711, "step": 4528 }, { "epoch": 0.3549373040752351, "grad_norm": 3.8242688179016113, "learning_rate": 9.81463345712436e-06, "loss": 1.5216, "step": 4529 }, { "epoch": 0.3550156739811912, "grad_norm": 2.5200884342193604, "learning_rate": 9.814550263230875e-06, "loss": 1.566, "step": 4530 }, { "epoch": 0.3550940438871473, "grad_norm": 6.4847092628479, "learning_rate": 9.814467051025316e-06, "loss": 1.7659, "step": 4531 }, { "epoch": 0.35517241379310344, "grad_norm": 3.1476199626922607, "learning_rate": 9.814383820507996e-06, "loss": 1.3918, "step": 4532 }, { "epoch": 0.35525078369905955, "grad_norm": 3.309779405593872, "learning_rate": 9.814300571679231e-06, "loss": 1.821, "step": 4533 }, { "epoch": 0.35532915360501566, "grad_norm": 3.2340517044067383, "learning_rate": 9.814217304539342e-06, "loss": 1.2862, "step": 4534 }, { "epoch": 0.35540752351097177, "grad_norm": 7.76582145690918, "learning_rate": 9.81413401908864e-06, "loss": 1.5749, "step": 4535 }, { "epoch": 0.3554858934169279, "grad_norm": 15.750946044921875, "learning_rate": 9.814050715327448e-06, "loss": 2.1361, "step": 4536 }, { "epoch": 0.355564263322884, "grad_norm": 3.599743366241455, "learning_rate": 9.81396739325608e-06, "loss": 1.4596, "step": 4537 }, { "epoch": 0.3556426332288401, "grad_norm": 3.326774835586548, "learning_rate": 9.813884052874852e-06, "loss": 1.1662, "step": 4538 }, { "epoch": 0.35572100313479627, "grad_norm": 3.5887629985809326, "learning_rate": 9.81380069418408e-06, "loss": 1.7022, "step": 4539 }, { "epoch": 0.3557993730407524, "grad_norm": 3.292573928833008, "learning_rate": 9.813717317184085e-06, "loss": 1.2305, "step": 4540 }, { "epoch": 0.3558777429467085, "grad_norm": 2.965017557144165, "learning_rate": 9.813633921875178e-06, "loss": 1.5246, "step": 4541 }, { "epoch": 0.3559561128526646, "grad_norm": 4.441603183746338, "learning_rate": 9.813550508257683e-06, "loss": 1.2456, "step": 4542 }, { "epoch": 0.3560344827586207, "grad_norm": 2.1019787788391113, "learning_rate": 9.813467076331916e-06, "loss": 1.3355, "step": 4543 }, { "epoch": 0.3561128526645768, "grad_norm": 18.16289710998535, "learning_rate": 9.813383626098188e-06, "loss": 0.7751, "step": 4544 }, { "epoch": 0.35619122257053293, "grad_norm": 2.8392019271850586, "learning_rate": 9.813300157556824e-06, "loss": 1.0351, "step": 4545 }, { "epoch": 0.35626959247648904, "grad_norm": 4.087121963500977, "learning_rate": 9.813216670708138e-06, "loss": 2.6358, "step": 4546 }, { "epoch": 0.35634796238244515, "grad_norm": 2.841982841491699, "learning_rate": 9.813133165552447e-06, "loss": 1.2256, "step": 4547 }, { "epoch": 0.35642633228840126, "grad_norm": 3.7829012870788574, "learning_rate": 9.813049642090071e-06, "loss": 1.4505, "step": 4548 }, { "epoch": 0.35650470219435737, "grad_norm": 3.589946985244751, "learning_rate": 9.812966100321326e-06, "loss": 1.4333, "step": 4549 }, { "epoch": 0.3565830721003135, "grad_norm": 2.5476956367492676, "learning_rate": 9.81288254024653e-06, "loss": 1.3719, "step": 4550 }, { "epoch": 0.3566614420062696, "grad_norm": 3.003537654876709, "learning_rate": 9.812798961866e-06, "loss": 1.0656, "step": 4551 }, { "epoch": 0.3567398119122257, "grad_norm": 2.4743900299072266, "learning_rate": 9.812715365180055e-06, "loss": 1.5181, "step": 4552 }, { "epoch": 0.3568181818181818, "grad_norm": 3.04323148727417, "learning_rate": 9.812631750189014e-06, "loss": 1.4555, "step": 4553 }, { "epoch": 0.3568965517241379, "grad_norm": 3.112968683242798, "learning_rate": 9.812548116893192e-06, "loss": 1.6104, "step": 4554 }, { "epoch": 0.35697492163009403, "grad_norm": 2.9450807571411133, "learning_rate": 9.81246446529291e-06, "loss": 1.3702, "step": 4555 }, { "epoch": 0.35705329153605014, "grad_norm": 7.290327548980713, "learning_rate": 9.812380795388486e-06, "loss": 1.0481, "step": 4556 }, { "epoch": 0.35713166144200625, "grad_norm": 6.189059734344482, "learning_rate": 9.812297107180235e-06, "loss": 1.0193, "step": 4557 }, { "epoch": 0.35721003134796236, "grad_norm": 3.335024356842041, "learning_rate": 9.81221340066848e-06, "loss": 1.2495, "step": 4558 }, { "epoch": 0.3572884012539185, "grad_norm": 4.0760884284973145, "learning_rate": 9.812129675853535e-06, "loss": 1.8129, "step": 4559 }, { "epoch": 0.3573667711598746, "grad_norm": 2.1156792640686035, "learning_rate": 9.81204593273572e-06, "loss": 1.4506, "step": 4560 }, { "epoch": 0.3574451410658307, "grad_norm": 3.208350419998169, "learning_rate": 9.811962171315354e-06, "loss": 1.4208, "step": 4561 }, { "epoch": 0.3575235109717868, "grad_norm": 2.770267963409424, "learning_rate": 9.811878391592757e-06, "loss": 1.0993, "step": 4562 }, { "epoch": 0.357601880877743, "grad_norm": 2.383761167526245, "learning_rate": 9.811794593568245e-06, "loss": 1.1261, "step": 4563 }, { "epoch": 0.3576802507836991, "grad_norm": 3.079925298690796, "learning_rate": 9.811710777242139e-06, "loss": 1.4491, "step": 4564 }, { "epoch": 0.3577586206896552, "grad_norm": 3.0684757232666016, "learning_rate": 9.811626942614756e-06, "loss": 0.9451, "step": 4565 }, { "epoch": 0.3578369905956113, "grad_norm": 2.3303701877593994, "learning_rate": 9.811543089686415e-06, "loss": 1.3631, "step": 4566 }, { "epoch": 0.3579153605015674, "grad_norm": 2.839285135269165, "learning_rate": 9.811459218457436e-06, "loss": 1.6596, "step": 4567 }, { "epoch": 0.3579937304075235, "grad_norm": 3.629288911819458, "learning_rate": 9.811375328928137e-06, "loss": 1.2546, "step": 4568 }, { "epoch": 0.35807210031347964, "grad_norm": 3.1636528968811035, "learning_rate": 9.811291421098837e-06, "loss": 1.6241, "step": 4569 }, { "epoch": 0.35815047021943575, "grad_norm": 5.745759963989258, "learning_rate": 9.811207494969857e-06, "loss": 1.5442, "step": 4570 }, { "epoch": 0.35822884012539186, "grad_norm": 2.265394926071167, "learning_rate": 9.811123550541515e-06, "loss": 1.0202, "step": 4571 }, { "epoch": 0.35830721003134797, "grad_norm": 2.1633050441741943, "learning_rate": 9.811039587814128e-06, "loss": 1.3523, "step": 4572 }, { "epoch": 0.3583855799373041, "grad_norm": 3.1418161392211914, "learning_rate": 9.81095560678802e-06, "loss": 1.0287, "step": 4573 }, { "epoch": 0.3584639498432602, "grad_norm": 2.5579137802124023, "learning_rate": 9.810871607463507e-06, "loss": 1.5273, "step": 4574 }, { "epoch": 0.3585423197492163, "grad_norm": 3.1182610988616943, "learning_rate": 9.810787589840909e-06, "loss": 1.577, "step": 4575 }, { "epoch": 0.3586206896551724, "grad_norm": 2.3121066093444824, "learning_rate": 9.810703553920546e-06, "loss": 1.4422, "step": 4576 }, { "epoch": 0.3586990595611285, "grad_norm": 3.4487345218658447, "learning_rate": 9.810619499702738e-06, "loss": 0.9414, "step": 4577 }, { "epoch": 0.35877742946708463, "grad_norm": 2.411576747894287, "learning_rate": 9.810535427187805e-06, "loss": 1.1087, "step": 4578 }, { "epoch": 0.35885579937304074, "grad_norm": 2.9166901111602783, "learning_rate": 9.810451336376065e-06, "loss": 1.3062, "step": 4579 }, { "epoch": 0.35893416927899685, "grad_norm": 4.020886421203613, "learning_rate": 9.81036722726784e-06, "loss": 1.2864, "step": 4580 }, { "epoch": 0.35901253918495296, "grad_norm": 2.419901132583618, "learning_rate": 9.810283099863448e-06, "loss": 1.4363, "step": 4581 }, { "epoch": 0.35909090909090907, "grad_norm": 2.802475690841675, "learning_rate": 9.81019895416321e-06, "loss": 1.1531, "step": 4582 }, { "epoch": 0.3591692789968652, "grad_norm": 3.639155387878418, "learning_rate": 9.810114790167446e-06, "loss": 1.7481, "step": 4583 }, { "epoch": 0.3592476489028213, "grad_norm": 2.509702444076538, "learning_rate": 9.810030607876476e-06, "loss": 1.2405, "step": 4584 }, { "epoch": 0.3593260188087774, "grad_norm": 2.8600080013275146, "learning_rate": 9.809946407290619e-06, "loss": 1.2979, "step": 4585 }, { "epoch": 0.35940438871473357, "grad_norm": 2.862913131713867, "learning_rate": 9.809862188410197e-06, "loss": 1.7144, "step": 4586 }, { "epoch": 0.3594827586206897, "grad_norm": 2.707387685775757, "learning_rate": 9.809777951235531e-06, "loss": 1.398, "step": 4587 }, { "epoch": 0.3595611285266458, "grad_norm": 2.610830545425415, "learning_rate": 9.80969369576694e-06, "loss": 1.3731, "step": 4588 }, { "epoch": 0.3596394984326019, "grad_norm": 4.581230640411377, "learning_rate": 9.809609422004741e-06, "loss": 1.5764, "step": 4589 }, { "epoch": 0.359717868338558, "grad_norm": 2.8147757053375244, "learning_rate": 9.809525129949262e-06, "loss": 1.4366, "step": 4590 }, { "epoch": 0.3597962382445141, "grad_norm": 3.2594852447509766, "learning_rate": 9.809440819600819e-06, "loss": 1.3411, "step": 4591 }, { "epoch": 0.35987460815047023, "grad_norm": 3.6290781497955322, "learning_rate": 9.809356490959733e-06, "loss": 1.1687, "step": 4592 }, { "epoch": 0.35995297805642634, "grad_norm": 4.098678112030029, "learning_rate": 9.809272144026326e-06, "loss": 2.202, "step": 4593 }, { "epoch": 0.36003134796238245, "grad_norm": 4.096924304962158, "learning_rate": 9.809187778800917e-06, "loss": 1.3505, "step": 4594 }, { "epoch": 0.36010971786833856, "grad_norm": 2.1277236938476562, "learning_rate": 9.809103395283828e-06, "loss": 1.3112, "step": 4595 }, { "epoch": 0.3601880877742947, "grad_norm": 3.4891653060913086, "learning_rate": 9.80901899347538e-06, "loss": 1.1212, "step": 4596 }, { "epoch": 0.3602664576802508, "grad_norm": 2.2303526401519775, "learning_rate": 9.808934573375894e-06, "loss": 1.2301, "step": 4597 }, { "epoch": 0.3603448275862069, "grad_norm": 5.004639148712158, "learning_rate": 9.80885013498569e-06, "loss": 1.385, "step": 4598 }, { "epoch": 0.360423197492163, "grad_norm": 2.0281572341918945, "learning_rate": 9.80876567830509e-06, "loss": 1.5088, "step": 4599 }, { "epoch": 0.3605015673981191, "grad_norm": 3.052096128463745, "learning_rate": 9.808681203334416e-06, "loss": 1.5964, "step": 4600 }, { "epoch": 0.3605799373040752, "grad_norm": 2.943274974822998, "learning_rate": 9.80859671007399e-06, "loss": 1.396, "step": 4601 }, { "epoch": 0.36065830721003134, "grad_norm": 2.7843475341796875, "learning_rate": 9.80851219852413e-06, "loss": 1.5483, "step": 4602 }, { "epoch": 0.36073667711598745, "grad_norm": 2.772028923034668, "learning_rate": 9.808427668685159e-06, "loss": 1.3613, "step": 4603 }, { "epoch": 0.36081504702194356, "grad_norm": 3.3716137409210205, "learning_rate": 9.8083431205574e-06, "loss": 1.1449, "step": 4604 }, { "epoch": 0.36089341692789967, "grad_norm": 2.7452237606048584, "learning_rate": 9.808258554141174e-06, "loss": 1.6987, "step": 4605 }, { "epoch": 0.3609717868338558, "grad_norm": 3.2480194568634033, "learning_rate": 9.8081739694368e-06, "loss": 1.5784, "step": 4606 }, { "epoch": 0.3610501567398119, "grad_norm": 3.037344217300415, "learning_rate": 9.808089366444605e-06, "loss": 1.247, "step": 4607 }, { "epoch": 0.361128526645768, "grad_norm": 9.861132621765137, "learning_rate": 9.808004745164905e-06, "loss": 0.7483, "step": 4608 }, { "epoch": 0.3612068965517241, "grad_norm": 2.6099493503570557, "learning_rate": 9.807920105598025e-06, "loss": 1.1702, "step": 4609 }, { "epoch": 0.3612852664576803, "grad_norm": 2.27221417427063, "learning_rate": 9.807835447744288e-06, "loss": 1.4091, "step": 4610 }, { "epoch": 0.3613636363636364, "grad_norm": 3.7278456687927246, "learning_rate": 9.807750771604012e-06, "loss": 1.3187, "step": 4611 }, { "epoch": 0.3614420062695925, "grad_norm": 2.4468445777893066, "learning_rate": 9.807666077177523e-06, "loss": 1.5655, "step": 4612 }, { "epoch": 0.3615203761755486, "grad_norm": 3.3776357173919678, "learning_rate": 9.807581364465142e-06, "loss": 1.3295, "step": 4613 }, { "epoch": 0.3615987460815047, "grad_norm": 2.1540520191192627, "learning_rate": 9.80749663346719e-06, "loss": 1.1976, "step": 4614 }, { "epoch": 0.36167711598746083, "grad_norm": 2.4390103816986084, "learning_rate": 9.80741188418399e-06, "loss": 1.6645, "step": 4615 }, { "epoch": 0.36175548589341694, "grad_norm": 3.0446560382843018, "learning_rate": 9.807327116615866e-06, "loss": 1.6761, "step": 4616 }, { "epoch": 0.36183385579937305, "grad_norm": 3.3082120418548584, "learning_rate": 9.807242330763135e-06, "loss": 0.6967, "step": 4617 }, { "epoch": 0.36191222570532916, "grad_norm": 3.6265716552734375, "learning_rate": 9.807157526626127e-06, "loss": 1.2465, "step": 4618 }, { "epoch": 0.36199059561128527, "grad_norm": 4.586638450622559, "learning_rate": 9.80707270420516e-06, "loss": 1.4009, "step": 4619 }, { "epoch": 0.3620689655172414, "grad_norm": 5.016090393066406, "learning_rate": 9.806987863500558e-06, "loss": 1.6702, "step": 4620 }, { "epoch": 0.3621473354231975, "grad_norm": 5.214278221130371, "learning_rate": 9.806903004512642e-06, "loss": 1.2865, "step": 4621 }, { "epoch": 0.3622257053291536, "grad_norm": 3.002415657043457, "learning_rate": 9.806818127241736e-06, "loss": 1.1038, "step": 4622 }, { "epoch": 0.3623040752351097, "grad_norm": 3.166853189468384, "learning_rate": 9.806733231688164e-06, "loss": 1.5624, "step": 4623 }, { "epoch": 0.3623824451410658, "grad_norm": 6.285713195800781, "learning_rate": 9.806648317852246e-06, "loss": 1.6829, "step": 4624 }, { "epoch": 0.36246081504702193, "grad_norm": 2.5748848915100098, "learning_rate": 9.806563385734307e-06, "loss": 1.2866, "step": 4625 }, { "epoch": 0.36253918495297804, "grad_norm": 2.676830291748047, "learning_rate": 9.80647843533467e-06, "loss": 1.3037, "step": 4626 }, { "epoch": 0.36261755485893415, "grad_norm": 2.128506660461426, "learning_rate": 9.806393466653658e-06, "loss": 1.2966, "step": 4627 }, { "epoch": 0.36269592476489027, "grad_norm": 2.6985924243927, "learning_rate": 9.806308479691595e-06, "loss": 1.1411, "step": 4628 }, { "epoch": 0.3627742946708464, "grad_norm": 3.004911422729492, "learning_rate": 9.806223474448801e-06, "loss": 1.1567, "step": 4629 }, { "epoch": 0.3628526645768025, "grad_norm": 2.842942237854004, "learning_rate": 9.806138450925604e-06, "loss": 1.341, "step": 4630 }, { "epoch": 0.3629310344827586, "grad_norm": 5.68068790435791, "learning_rate": 9.806053409122323e-06, "loss": 1.4249, "step": 4631 }, { "epoch": 0.3630094043887147, "grad_norm": 3.512016534805298, "learning_rate": 9.805968349039285e-06, "loss": 1.4549, "step": 4632 }, { "epoch": 0.3630877742946709, "grad_norm": 2.3528614044189453, "learning_rate": 9.80588327067681e-06, "loss": 1.2739, "step": 4633 }, { "epoch": 0.363166144200627, "grad_norm": 3.990171432495117, "learning_rate": 9.805798174035226e-06, "loss": 1.5206, "step": 4634 }, { "epoch": 0.3632445141065831, "grad_norm": 2.2586867809295654, "learning_rate": 9.805713059114852e-06, "loss": 1.2883, "step": 4635 }, { "epoch": 0.3633228840125392, "grad_norm": 3.0904719829559326, "learning_rate": 9.805627925916015e-06, "loss": 1.3601, "step": 4636 }, { "epoch": 0.3634012539184953, "grad_norm": 5.628743648529053, "learning_rate": 9.805542774439037e-06, "loss": 1.4866, "step": 4637 }, { "epoch": 0.3634796238244514, "grad_norm": 5.148914813995361, "learning_rate": 9.805457604684241e-06, "loss": 1.0232, "step": 4638 }, { "epoch": 0.36355799373040754, "grad_norm": 3.808656692504883, "learning_rate": 9.805372416651955e-06, "loss": 0.9358, "step": 4639 }, { "epoch": 0.36363636363636365, "grad_norm": 5.814887046813965, "learning_rate": 9.8052872103425e-06, "loss": 1.5435, "step": 4640 }, { "epoch": 0.36371473354231976, "grad_norm": 2.4041554927825928, "learning_rate": 9.8052019857562e-06, "loss": 1.4545, "step": 4641 }, { "epoch": 0.36379310344827587, "grad_norm": 6.574066638946533, "learning_rate": 9.805116742893381e-06, "loss": 1.78, "step": 4642 }, { "epoch": 0.363871473354232, "grad_norm": 2.3154633045196533, "learning_rate": 9.805031481754364e-06, "loss": 1.3746, "step": 4643 }, { "epoch": 0.3639498432601881, "grad_norm": 2.4647092819213867, "learning_rate": 9.804946202339477e-06, "loss": 1.136, "step": 4644 }, { "epoch": 0.3640282131661442, "grad_norm": 5.394207000732422, "learning_rate": 9.804860904649044e-06, "loss": 1.0041, "step": 4645 }, { "epoch": 0.3641065830721003, "grad_norm": 5.1211442947387695, "learning_rate": 9.804775588683383e-06, "loss": 1.4512, "step": 4646 }, { "epoch": 0.3641849529780564, "grad_norm": 3.2988438606262207, "learning_rate": 9.804690254442828e-06, "loss": 0.9611, "step": 4647 }, { "epoch": 0.36426332288401253, "grad_norm": 3.281284809112549, "learning_rate": 9.804604901927698e-06, "loss": 1.617, "step": 4648 }, { "epoch": 0.36434169278996864, "grad_norm": 3.1898152828216553, "learning_rate": 9.804519531138318e-06, "loss": 1.4235, "step": 4649 }, { "epoch": 0.36442006269592475, "grad_norm": 2.8010029792785645, "learning_rate": 9.804434142075015e-06, "loss": 1.1869, "step": 4650 }, { "epoch": 0.36449843260188086, "grad_norm": 6.200675964355469, "learning_rate": 9.80434873473811e-06, "loss": 1.1325, "step": 4651 }, { "epoch": 0.364576802507837, "grad_norm": 6.132493019104004, "learning_rate": 9.804263309127931e-06, "loss": 1.5418, "step": 4652 }, { "epoch": 0.3646551724137931, "grad_norm": 2.8599071502685547, "learning_rate": 9.804177865244803e-06, "loss": 1.4028, "step": 4653 }, { "epoch": 0.3647335423197492, "grad_norm": 3.5849499702453613, "learning_rate": 9.804092403089049e-06, "loss": 1.1724, "step": 4654 }, { "epoch": 0.3648119122257053, "grad_norm": 5.134426593780518, "learning_rate": 9.804006922660997e-06, "loss": 1.6176, "step": 4655 }, { "epoch": 0.36489028213166147, "grad_norm": 6.09432315826416, "learning_rate": 9.803921423960966e-06, "loss": 1.3456, "step": 4656 }, { "epoch": 0.3649686520376176, "grad_norm": 2.632556438446045, "learning_rate": 9.803835906989287e-06, "loss": 0.7167, "step": 4657 }, { "epoch": 0.3650470219435737, "grad_norm": 2.6008472442626953, "learning_rate": 9.803750371746285e-06, "loss": 1.3807, "step": 4658 }, { "epoch": 0.3651253918495298, "grad_norm": 2.2011802196502686, "learning_rate": 9.803664818232284e-06, "loss": 1.4001, "step": 4659 }, { "epoch": 0.3652037617554859, "grad_norm": 2.70566987991333, "learning_rate": 9.803579246447609e-06, "loss": 1.579, "step": 4660 }, { "epoch": 0.365282131661442, "grad_norm": 3.4870948791503906, "learning_rate": 9.803493656392585e-06, "loss": 1.4881, "step": 4661 }, { "epoch": 0.36536050156739813, "grad_norm": 3.0123355388641357, "learning_rate": 9.80340804806754e-06, "loss": 1.4067, "step": 4662 }, { "epoch": 0.36543887147335424, "grad_norm": 3.9865691661834717, "learning_rate": 9.803322421472797e-06, "loss": 1.3064, "step": 4663 }, { "epoch": 0.36551724137931035, "grad_norm": 3.1868085861206055, "learning_rate": 9.803236776608683e-06, "loss": 1.5136, "step": 4664 }, { "epoch": 0.36559561128526646, "grad_norm": 3.7047722339630127, "learning_rate": 9.803151113475522e-06, "loss": 1.6977, "step": 4665 }, { "epoch": 0.3656739811912226, "grad_norm": 5.039989471435547, "learning_rate": 9.803065432073641e-06, "loss": 1.2895, "step": 4666 }, { "epoch": 0.3657523510971787, "grad_norm": 2.393467664718628, "learning_rate": 9.802979732403369e-06, "loss": 1.6428, "step": 4667 }, { "epoch": 0.3658307210031348, "grad_norm": 3.696099281311035, "learning_rate": 9.802894014465025e-06, "loss": 1.6987, "step": 4668 }, { "epoch": 0.3659090909090909, "grad_norm": 7.6246018409729, "learning_rate": 9.802808278258942e-06, "loss": 0.9114, "step": 4669 }, { "epoch": 0.365987460815047, "grad_norm": 2.676891326904297, "learning_rate": 9.802722523785442e-06, "loss": 1.4312, "step": 4670 }, { "epoch": 0.3660658307210031, "grad_norm": 2.787396192550659, "learning_rate": 9.802636751044855e-06, "loss": 1.3727, "step": 4671 }, { "epoch": 0.36614420062695924, "grad_norm": 4.1382646560668945, "learning_rate": 9.802550960037501e-06, "loss": 1.402, "step": 4672 }, { "epoch": 0.36622257053291535, "grad_norm": 2.612643003463745, "learning_rate": 9.802465150763712e-06, "loss": 1.4271, "step": 4673 }, { "epoch": 0.36630094043887146, "grad_norm": 7.325962066650391, "learning_rate": 9.802379323223812e-06, "loss": 1.3466, "step": 4674 }, { "epoch": 0.36637931034482757, "grad_norm": 3.7580344676971436, "learning_rate": 9.802293477418127e-06, "loss": 1.2907, "step": 4675 }, { "epoch": 0.3664576802507837, "grad_norm": 3.7988336086273193, "learning_rate": 9.802207613346984e-06, "loss": 1.5047, "step": 4676 }, { "epoch": 0.3665360501567398, "grad_norm": 2.702887535095215, "learning_rate": 9.80212173101071e-06, "loss": 1.5435, "step": 4677 }, { "epoch": 0.3666144200626959, "grad_norm": 3.244004726409912, "learning_rate": 9.802035830409634e-06, "loss": 1.212, "step": 4678 }, { "epoch": 0.366692789968652, "grad_norm": 2.4887657165527344, "learning_rate": 9.801949911544077e-06, "loss": 1.3928, "step": 4679 }, { "epoch": 0.3667711598746082, "grad_norm": 3.2789993286132812, "learning_rate": 9.80186397441437e-06, "loss": 1.4145, "step": 4680 }, { "epoch": 0.3668495297805643, "grad_norm": 7.070383071899414, "learning_rate": 9.80177801902084e-06, "loss": 1.149, "step": 4681 }, { "epoch": 0.3669278996865204, "grad_norm": 3.171694040298462, "learning_rate": 9.801692045363811e-06, "loss": 1.4749, "step": 4682 }, { "epoch": 0.3670062695924765, "grad_norm": 2.7446913719177246, "learning_rate": 9.80160605344361e-06, "loss": 1.2636, "step": 4683 }, { "epoch": 0.3670846394984326, "grad_norm": 2.5843794345855713, "learning_rate": 9.80152004326057e-06, "loss": 1.4829, "step": 4684 }, { "epoch": 0.36716300940438873, "grad_norm": 3.092254161834717, "learning_rate": 9.801434014815012e-06, "loss": 1.1168, "step": 4685 }, { "epoch": 0.36724137931034484, "grad_norm": 5.9976725578308105, "learning_rate": 9.801347968107265e-06, "loss": 1.7615, "step": 4686 }, { "epoch": 0.36731974921630095, "grad_norm": 3.200493812561035, "learning_rate": 9.801261903137657e-06, "loss": 1.6664, "step": 4687 }, { "epoch": 0.36739811912225706, "grad_norm": 5.335000038146973, "learning_rate": 9.801175819906514e-06, "loss": 1.0344, "step": 4688 }, { "epoch": 0.36747648902821317, "grad_norm": 3.1206369400024414, "learning_rate": 9.801089718414165e-06, "loss": 1.7145, "step": 4689 }, { "epoch": 0.3675548589341693, "grad_norm": 9.09692096710205, "learning_rate": 9.801003598660937e-06, "loss": 0.9805, "step": 4690 }, { "epoch": 0.3676332288401254, "grad_norm": 3.6254403591156006, "learning_rate": 9.800917460647157e-06, "loss": 1.4089, "step": 4691 }, { "epoch": 0.3677115987460815, "grad_norm": 3.252711772918701, "learning_rate": 9.800831304373152e-06, "loss": 1.255, "step": 4692 }, { "epoch": 0.3677899686520376, "grad_norm": 3.7067618370056152, "learning_rate": 9.800745129839254e-06, "loss": 1.1848, "step": 4693 }, { "epoch": 0.3678683385579937, "grad_norm": 3.262219190597534, "learning_rate": 9.800658937045783e-06, "loss": 1.1251, "step": 4694 }, { "epoch": 0.36794670846394983, "grad_norm": 5.206182956695557, "learning_rate": 9.800572725993073e-06, "loss": 1.3014, "step": 4695 }, { "epoch": 0.36802507836990594, "grad_norm": 2.5350844860076904, "learning_rate": 9.800486496681449e-06, "loss": 1.3152, "step": 4696 }, { "epoch": 0.36810344827586206, "grad_norm": 3.4942853450775146, "learning_rate": 9.800400249111241e-06, "loss": 1.3171, "step": 4697 }, { "epoch": 0.36818181818181817, "grad_norm": 3.0340912342071533, "learning_rate": 9.800313983282776e-06, "loss": 1.5423, "step": 4698 }, { "epoch": 0.3682601880877743, "grad_norm": 4.263535499572754, "learning_rate": 9.800227699196382e-06, "loss": 1.7921, "step": 4699 }, { "epoch": 0.3683385579937304, "grad_norm": 8.000631332397461, "learning_rate": 9.800141396852388e-06, "loss": 1.7548, "step": 4700 }, { "epoch": 0.3684169278996865, "grad_norm": 2.4784998893737793, "learning_rate": 9.800055076251122e-06, "loss": 1.4708, "step": 4701 }, { "epoch": 0.3684952978056426, "grad_norm": 2.827641010284424, "learning_rate": 9.799968737392911e-06, "loss": 1.539, "step": 4702 }, { "epoch": 0.3685736677115988, "grad_norm": 4.019493103027344, "learning_rate": 9.799882380278084e-06, "loss": 1.5552, "step": 4703 }, { "epoch": 0.3686520376175549, "grad_norm": 3.5495781898498535, "learning_rate": 9.79979600490697e-06, "loss": 1.3737, "step": 4704 }, { "epoch": 0.368730407523511, "grad_norm": 2.104749917984009, "learning_rate": 9.799709611279898e-06, "loss": 1.2799, "step": 4705 }, { "epoch": 0.3688087774294671, "grad_norm": 3.903061628341675, "learning_rate": 9.799623199397197e-06, "loss": 0.8531, "step": 4706 }, { "epoch": 0.3688871473354232, "grad_norm": 4.003563404083252, "learning_rate": 9.799536769259192e-06, "loss": 1.5701, "step": 4707 }, { "epoch": 0.3689655172413793, "grad_norm": 11.624016761779785, "learning_rate": 9.799450320866216e-06, "loss": 1.2375, "step": 4708 }, { "epoch": 0.36904388714733544, "grad_norm": 4.747986316680908, "learning_rate": 9.799363854218596e-06, "loss": 1.6886, "step": 4709 }, { "epoch": 0.36912225705329155, "grad_norm": 4.279014587402344, "learning_rate": 9.79927736931666e-06, "loss": 1.4214, "step": 4710 }, { "epoch": 0.36920062695924766, "grad_norm": 3.859055757522583, "learning_rate": 9.799190866160739e-06, "loss": 1.5029, "step": 4711 }, { "epoch": 0.36927899686520377, "grad_norm": 3.8672034740448, "learning_rate": 9.79910434475116e-06, "loss": 1.1569, "step": 4712 }, { "epoch": 0.3693573667711599, "grad_norm": 3.1179728507995605, "learning_rate": 9.799017805088254e-06, "loss": 1.373, "step": 4713 }, { "epoch": 0.369435736677116, "grad_norm": 3.160616397857666, "learning_rate": 9.798931247172349e-06, "loss": 1.5474, "step": 4714 }, { "epoch": 0.3695141065830721, "grad_norm": 2.881511688232422, "learning_rate": 9.798844671003774e-06, "loss": 1.3004, "step": 4715 }, { "epoch": 0.3695924764890282, "grad_norm": 2.8463757038116455, "learning_rate": 9.79875807658286e-06, "loss": 1.5564, "step": 4716 }, { "epoch": 0.3696708463949843, "grad_norm": 4.14139986038208, "learning_rate": 9.798671463909934e-06, "loss": 0.8052, "step": 4717 }, { "epoch": 0.36974921630094043, "grad_norm": 4.6869893074035645, "learning_rate": 9.798584832985327e-06, "loss": 1.3558, "step": 4718 }, { "epoch": 0.36982758620689654, "grad_norm": 2.4821088314056396, "learning_rate": 9.798498183809366e-06, "loss": 1.236, "step": 4719 }, { "epoch": 0.36990595611285265, "grad_norm": 4.811655521392822, "learning_rate": 9.798411516382385e-06, "loss": 1.8501, "step": 4720 }, { "epoch": 0.36998432601880876, "grad_norm": 4.6327104568481445, "learning_rate": 9.798324830704711e-06, "loss": 1.287, "step": 4721 }, { "epoch": 0.3700626959247649, "grad_norm": 2.5185649394989014, "learning_rate": 9.798238126776673e-06, "loss": 1.2045, "step": 4722 }, { "epoch": 0.370141065830721, "grad_norm": 3.597001791000366, "learning_rate": 9.7981514045986e-06, "loss": 1.3234, "step": 4723 }, { "epoch": 0.3702194357366771, "grad_norm": 4.243133544921875, "learning_rate": 9.798064664170828e-06, "loss": 1.3533, "step": 4724 }, { "epoch": 0.3702978056426332, "grad_norm": 4.948455810546875, "learning_rate": 9.797977905493678e-06, "loss": 1.2116, "step": 4725 }, { "epoch": 0.3703761755485893, "grad_norm": 3.6060357093811035, "learning_rate": 9.797891128567487e-06, "loss": 1.3174, "step": 4726 }, { "epoch": 0.3704545454545455, "grad_norm": 3.988436460494995, "learning_rate": 9.797804333392582e-06, "loss": 1.0783, "step": 4727 }, { "epoch": 0.3705329153605016, "grad_norm": 2.5284976959228516, "learning_rate": 9.797717519969294e-06, "loss": 1.5163, "step": 4728 }, { "epoch": 0.3706112852664577, "grad_norm": 4.344829082489014, "learning_rate": 9.797630688297953e-06, "loss": 1.2684, "step": 4729 }, { "epoch": 0.3706896551724138, "grad_norm": 3.264042377471924, "learning_rate": 9.797543838378886e-06, "loss": 1.749, "step": 4730 }, { "epoch": 0.3707680250783699, "grad_norm": 3.615788221359253, "learning_rate": 9.797456970212431e-06, "loss": 1.4523, "step": 4731 }, { "epoch": 0.37084639498432603, "grad_norm": 4.343363285064697, "learning_rate": 9.797370083798911e-06, "loss": 1.4722, "step": 4732 }, { "epoch": 0.37092476489028214, "grad_norm": 4.481517314910889, "learning_rate": 9.79728317913866e-06, "loss": 1.7126, "step": 4733 }, { "epoch": 0.37100313479623825, "grad_norm": 3.5535664558410645, "learning_rate": 9.797196256232007e-06, "loss": 1.5733, "step": 4734 }, { "epoch": 0.37108150470219436, "grad_norm": 2.41465425491333, "learning_rate": 9.797109315079285e-06, "loss": 1.4214, "step": 4735 }, { "epoch": 0.3711598746081505, "grad_norm": 4.25878381729126, "learning_rate": 9.797022355680823e-06, "loss": 1.3199, "step": 4736 }, { "epoch": 0.3712382445141066, "grad_norm": 2.9363601207733154, "learning_rate": 9.796935378036952e-06, "loss": 1.33, "step": 4737 }, { "epoch": 0.3713166144200627, "grad_norm": 2.6955959796905518, "learning_rate": 9.796848382148002e-06, "loss": 1.2186, "step": 4738 }, { "epoch": 0.3713949843260188, "grad_norm": 3.4828035831451416, "learning_rate": 9.796761368014302e-06, "loss": 1.2356, "step": 4739 }, { "epoch": 0.3714733542319749, "grad_norm": 3.285461187362671, "learning_rate": 9.79667433563619e-06, "loss": 1.5715, "step": 4740 }, { "epoch": 0.371551724137931, "grad_norm": 3.1869537830352783, "learning_rate": 9.79658728501399e-06, "loss": 1.6053, "step": 4741 }, { "epoch": 0.37163009404388714, "grad_norm": 2.552053689956665, "learning_rate": 9.796500216148034e-06, "loss": 1.1543, "step": 4742 }, { "epoch": 0.37170846394984325, "grad_norm": 4.493332386016846, "learning_rate": 9.796413129038658e-06, "loss": 1.1402, "step": 4743 }, { "epoch": 0.37178683385579936, "grad_norm": 3.386763572692871, "learning_rate": 9.79632602368619e-06, "loss": 1.5018, "step": 4744 }, { "epoch": 0.37186520376175547, "grad_norm": 3.222724437713623, "learning_rate": 9.796238900090959e-06, "loss": 1.3804, "step": 4745 }, { "epoch": 0.3719435736677116, "grad_norm": 4.004308700561523, "learning_rate": 9.7961517582533e-06, "loss": 1.5528, "step": 4746 }, { "epoch": 0.3720219435736677, "grad_norm": 2.868521213531494, "learning_rate": 9.796064598173545e-06, "loss": 1.5002, "step": 4747 }, { "epoch": 0.3721003134796238, "grad_norm": 3.8391647338867188, "learning_rate": 9.79597741985202e-06, "loss": 1.333, "step": 4748 }, { "epoch": 0.3721786833855799, "grad_norm": 4.330272197723389, "learning_rate": 9.795890223289064e-06, "loss": 1.2592, "step": 4749 }, { "epoch": 0.3722570532915361, "grad_norm": 2.3994574546813965, "learning_rate": 9.795803008485004e-06, "loss": 1.4074, "step": 4750 }, { "epoch": 0.3723354231974922, "grad_norm": 4.010336399078369, "learning_rate": 9.79571577544017e-06, "loss": 1.4423, "step": 4751 }, { "epoch": 0.3724137931034483, "grad_norm": 2.87308669090271, "learning_rate": 9.795628524154899e-06, "loss": 1.1923, "step": 4752 }, { "epoch": 0.3724921630094044, "grad_norm": 4.009082317352295, "learning_rate": 9.795541254629519e-06, "loss": 1.4027, "step": 4753 }, { "epoch": 0.3725705329153605, "grad_norm": 3.5633130073547363, "learning_rate": 9.795453966864365e-06, "loss": 1.5973, "step": 4754 }, { "epoch": 0.37264890282131663, "grad_norm": 4.096081256866455, "learning_rate": 9.795366660859766e-06, "loss": 1.5146, "step": 4755 }, { "epoch": 0.37272727272727274, "grad_norm": 2.6286654472351074, "learning_rate": 9.795279336616055e-06, "loss": 1.4383, "step": 4756 }, { "epoch": 0.37280564263322885, "grad_norm": 2.6776812076568604, "learning_rate": 9.795191994133564e-06, "loss": 1.3928, "step": 4757 }, { "epoch": 0.37288401253918496, "grad_norm": 2.630991220474243, "learning_rate": 9.795104633412627e-06, "loss": 1.5848, "step": 4758 }, { "epoch": 0.37296238244514107, "grad_norm": 4.128314018249512, "learning_rate": 9.795017254453573e-06, "loss": 0.9521, "step": 4759 }, { "epoch": 0.3730407523510972, "grad_norm": 14.93653392791748, "learning_rate": 9.79492985725674e-06, "loss": 1.6408, "step": 4760 }, { "epoch": 0.3731191222570533, "grad_norm": 14.913342475891113, "learning_rate": 9.794842441822451e-06, "loss": 1.0714, "step": 4761 }, { "epoch": 0.3731974921630094, "grad_norm": 3.4813218116760254, "learning_rate": 9.794755008151047e-06, "loss": 1.1284, "step": 4762 }, { "epoch": 0.3732758620689655, "grad_norm": 2.8254752159118652, "learning_rate": 9.794667556242858e-06, "loss": 1.5257, "step": 4763 }, { "epoch": 0.3733542319749216, "grad_norm": 4.5026350021362305, "learning_rate": 9.794580086098217e-06, "loss": 1.4014, "step": 4764 }, { "epoch": 0.37343260188087773, "grad_norm": 3.6794183254241943, "learning_rate": 9.794492597717455e-06, "loss": 1.4783, "step": 4765 }, { "epoch": 0.37351097178683385, "grad_norm": 5.411929130554199, "learning_rate": 9.794405091100907e-06, "loss": 1.5883, "step": 4766 }, { "epoch": 0.37358934169278996, "grad_norm": 3.4235363006591797, "learning_rate": 9.794317566248903e-06, "loss": 1.5982, "step": 4767 }, { "epoch": 0.37366771159874607, "grad_norm": 4.043653964996338, "learning_rate": 9.794230023161778e-06, "loss": 1.4209, "step": 4768 }, { "epoch": 0.3737460815047022, "grad_norm": 2.7677929401397705, "learning_rate": 9.794142461839864e-06, "loss": 1.2583, "step": 4769 }, { "epoch": 0.3738244514106583, "grad_norm": 2.9114232063293457, "learning_rate": 9.794054882283495e-06, "loss": 1.0716, "step": 4770 }, { "epoch": 0.3739028213166144, "grad_norm": 6.788397789001465, "learning_rate": 9.793967284493005e-06, "loss": 1.1977, "step": 4771 }, { "epoch": 0.3739811912225705, "grad_norm": 15.374281883239746, "learning_rate": 9.793879668468724e-06, "loss": 1.5986, "step": 4772 }, { "epoch": 0.3740595611285266, "grad_norm": 4.302168846130371, "learning_rate": 9.793792034210987e-06, "loss": 1.2018, "step": 4773 }, { "epoch": 0.3741379310344828, "grad_norm": 4.789059638977051, "learning_rate": 9.793704381720128e-06, "loss": 1.5854, "step": 4774 }, { "epoch": 0.3742163009404389, "grad_norm": 5.255116939544678, "learning_rate": 9.793616710996479e-06, "loss": 1.3516, "step": 4775 }, { "epoch": 0.374294670846395, "grad_norm": 2.7967476844787598, "learning_rate": 9.793529022040376e-06, "loss": 0.8332, "step": 4776 }, { "epoch": 0.3743730407523511, "grad_norm": 3.0811219215393066, "learning_rate": 9.793441314852147e-06, "loss": 1.5272, "step": 4777 }, { "epoch": 0.3744514106583072, "grad_norm": 5.570181846618652, "learning_rate": 9.793353589432134e-06, "loss": 2.0227, "step": 4778 }, { "epoch": 0.37452978056426334, "grad_norm": 7.411258697509766, "learning_rate": 9.793265845780662e-06, "loss": 1.3353, "step": 4779 }, { "epoch": 0.37460815047021945, "grad_norm": 2.907626152038574, "learning_rate": 9.793178083898073e-06, "loss": 1.1887, "step": 4780 }, { "epoch": 0.37468652037617556, "grad_norm": 3.5687334537506104, "learning_rate": 9.793090303784691e-06, "loss": 1.6428, "step": 4781 }, { "epoch": 0.37476489028213167, "grad_norm": 3.3248291015625, "learning_rate": 9.79300250544086e-06, "loss": 1.3785, "step": 4782 }, { "epoch": 0.3748432601880878, "grad_norm": 2.0832369327545166, "learning_rate": 9.792914688866908e-06, "loss": 1.5612, "step": 4783 }, { "epoch": 0.3749216300940439, "grad_norm": 3.033538579940796, "learning_rate": 9.792826854063168e-06, "loss": 1.3673, "step": 4784 }, { "epoch": 0.375, "grad_norm": 3.4502413272857666, "learning_rate": 9.792739001029979e-06, "loss": 1.4262, "step": 4785 }, { "epoch": 0.3750783699059561, "grad_norm": 2.341413974761963, "learning_rate": 9.792651129767671e-06, "loss": 1.4, "step": 4786 }, { "epoch": 0.3751567398119122, "grad_norm": 10.504347801208496, "learning_rate": 9.792563240276583e-06, "loss": 1.1516, "step": 4787 }, { "epoch": 0.37523510971786833, "grad_norm": 3.5854973793029785, "learning_rate": 9.792475332557042e-06, "loss": 1.5095, "step": 4788 }, { "epoch": 0.37531347962382444, "grad_norm": 2.606386423110962, "learning_rate": 9.792387406609388e-06, "loss": 1.4461, "step": 4789 }, { "epoch": 0.37539184952978055, "grad_norm": 3.965526580810547, "learning_rate": 9.792299462433955e-06, "loss": 1.88, "step": 4790 }, { "epoch": 0.37547021943573666, "grad_norm": 2.290419101715088, "learning_rate": 9.792211500031075e-06, "loss": 1.0456, "step": 4791 }, { "epoch": 0.3755485893416928, "grad_norm": 2.647615909576416, "learning_rate": 9.792123519401085e-06, "loss": 1.1955, "step": 4792 }, { "epoch": 0.3756269592476489, "grad_norm": 3.6093719005584717, "learning_rate": 9.792035520544316e-06, "loss": 1.3584, "step": 4793 }, { "epoch": 0.375705329153605, "grad_norm": 3.671600580215454, "learning_rate": 9.791947503461107e-06, "loss": 1.6051, "step": 4794 }, { "epoch": 0.3757836990595611, "grad_norm": 3.697812080383301, "learning_rate": 9.791859468151791e-06, "loss": 1.6996, "step": 4795 }, { "epoch": 0.3758620689655172, "grad_norm": 5.346830368041992, "learning_rate": 9.791771414616703e-06, "loss": 1.7853, "step": 4796 }, { "epoch": 0.3759404388714734, "grad_norm": 1.8998836278915405, "learning_rate": 9.791683342856177e-06, "loss": 1.3959, "step": 4797 }, { "epoch": 0.3760188087774295, "grad_norm": 4.623530864715576, "learning_rate": 9.791595252870549e-06, "loss": 1.3523, "step": 4798 }, { "epoch": 0.3760971786833856, "grad_norm": 2.9232397079467773, "learning_rate": 9.791507144660154e-06, "loss": 1.4183, "step": 4799 }, { "epoch": 0.3761755485893417, "grad_norm": 4.833846092224121, "learning_rate": 9.791419018225328e-06, "loss": 1.5866, "step": 4800 }, { "epoch": 0.3762539184952978, "grad_norm": 3.537188768386841, "learning_rate": 9.791330873566403e-06, "loss": 1.4152, "step": 4801 }, { "epoch": 0.37633228840125393, "grad_norm": 4.604305267333984, "learning_rate": 9.791242710683718e-06, "loss": 2.2422, "step": 4802 }, { "epoch": 0.37641065830721004, "grad_norm": 4.5691237449646, "learning_rate": 9.791154529577604e-06, "loss": 0.6947, "step": 4803 }, { "epoch": 0.37648902821316615, "grad_norm": 4.378882884979248, "learning_rate": 9.791066330248402e-06, "loss": 1.3124, "step": 4804 }, { "epoch": 0.37656739811912227, "grad_norm": 2.6090705394744873, "learning_rate": 9.790978112696443e-06, "loss": 1.4626, "step": 4805 }, { "epoch": 0.3766457680250784, "grad_norm": 2.7849881649017334, "learning_rate": 9.790889876922065e-06, "loss": 1.1612, "step": 4806 }, { "epoch": 0.3767241379310345, "grad_norm": 2.2135703563690186, "learning_rate": 9.790801622925603e-06, "loss": 1.5687, "step": 4807 }, { "epoch": 0.3768025078369906, "grad_norm": 1.9449456930160522, "learning_rate": 9.79071335070739e-06, "loss": 1.3047, "step": 4808 }, { "epoch": 0.3768808777429467, "grad_norm": 2.292100667953491, "learning_rate": 9.790625060267767e-06, "loss": 1.2904, "step": 4809 }, { "epoch": 0.3769592476489028, "grad_norm": 2.517068862915039, "learning_rate": 9.790536751607065e-06, "loss": 1.5062, "step": 4810 }, { "epoch": 0.37703761755485893, "grad_norm": 4.403970718383789, "learning_rate": 9.790448424725623e-06, "loss": 1.3876, "step": 4811 }, { "epoch": 0.37711598746081504, "grad_norm": 2.813303232192993, "learning_rate": 9.790360079623775e-06, "loss": 1.6011, "step": 4812 }, { "epoch": 0.37719435736677115, "grad_norm": 4.1433210372924805, "learning_rate": 9.790271716301858e-06, "loss": 1.5393, "step": 4813 }, { "epoch": 0.37727272727272726, "grad_norm": 2.7995216846466064, "learning_rate": 9.790183334760208e-06, "loss": 1.7951, "step": 4814 }, { "epoch": 0.37735109717868337, "grad_norm": 3.947786808013916, "learning_rate": 9.790094934999158e-06, "loss": 1.5564, "step": 4815 }, { "epoch": 0.3774294670846395, "grad_norm": 2.6428167819976807, "learning_rate": 9.790006517019052e-06, "loss": 1.225, "step": 4816 }, { "epoch": 0.3775078369905956, "grad_norm": 3.079228401184082, "learning_rate": 9.789918080820218e-06, "loss": 1.5434, "step": 4817 }, { "epoch": 0.3775862068965517, "grad_norm": 2.548208713531494, "learning_rate": 9.789829626402997e-06, "loss": 1.5266, "step": 4818 }, { "epoch": 0.3776645768025078, "grad_norm": 4.935700416564941, "learning_rate": 9.789741153767726e-06, "loss": 1.7009, "step": 4819 }, { "epoch": 0.3777429467084639, "grad_norm": 4.445859432220459, "learning_rate": 9.789652662914738e-06, "loss": 0.9777, "step": 4820 }, { "epoch": 0.3778213166144201, "grad_norm": 5.132225036621094, "learning_rate": 9.789564153844372e-06, "loss": 1.4205, "step": 4821 }, { "epoch": 0.3778996865203762, "grad_norm": 3.813424825668335, "learning_rate": 9.789475626556964e-06, "loss": 1.4654, "step": 4822 }, { "epoch": 0.3779780564263323, "grad_norm": 2.9724252223968506, "learning_rate": 9.78938708105285e-06, "loss": 1.1809, "step": 4823 }, { "epoch": 0.3780564263322884, "grad_norm": 3.555905818939209, "learning_rate": 9.789298517332368e-06, "loss": 1.004, "step": 4824 }, { "epoch": 0.37813479623824453, "grad_norm": 9.099664688110352, "learning_rate": 9.789209935395855e-06, "loss": 1.5693, "step": 4825 }, { "epoch": 0.37821316614420064, "grad_norm": 3.643279790878296, "learning_rate": 9.789121335243646e-06, "loss": 1.5865, "step": 4826 }, { "epoch": 0.37829153605015675, "grad_norm": 3.830173969268799, "learning_rate": 9.78903271687608e-06, "loss": 1.4011, "step": 4827 }, { "epoch": 0.37836990595611286, "grad_norm": 5.817699909210205, "learning_rate": 9.788944080293495e-06, "loss": 1.7573, "step": 4828 }, { "epoch": 0.378448275862069, "grad_norm": 3.255445957183838, "learning_rate": 9.788855425496224e-06, "loss": 1.5817, "step": 4829 }, { "epoch": 0.3785266457680251, "grad_norm": 4.407529830932617, "learning_rate": 9.788766752484606e-06, "loss": 1.1696, "step": 4830 }, { "epoch": 0.3786050156739812, "grad_norm": 5.667553424835205, "learning_rate": 9.788678061258981e-06, "loss": 1.3845, "step": 4831 }, { "epoch": 0.3786833855799373, "grad_norm": 2.3804829120635986, "learning_rate": 9.788589351819685e-06, "loss": 1.2323, "step": 4832 }, { "epoch": 0.3787617554858934, "grad_norm": 3.2791707515716553, "learning_rate": 9.788500624167052e-06, "loss": 1.1569, "step": 4833 }, { "epoch": 0.3788401253918495, "grad_norm": 9.614006996154785, "learning_rate": 9.788411878301425e-06, "loss": 1.1211, "step": 4834 }, { "epoch": 0.37891849529780564, "grad_norm": 3.137308120727539, "learning_rate": 9.788323114223137e-06, "loss": 1.513, "step": 4835 }, { "epoch": 0.37899686520376175, "grad_norm": 6.385653018951416, "learning_rate": 9.788234331932527e-06, "loss": 1.1238, "step": 4836 }, { "epoch": 0.37907523510971786, "grad_norm": 2.155571937561035, "learning_rate": 9.788145531429933e-06, "loss": 1.4445, "step": 4837 }, { "epoch": 0.37915360501567397, "grad_norm": 5.559189319610596, "learning_rate": 9.788056712715694e-06, "loss": 1.44, "step": 4838 }, { "epoch": 0.3792319749216301, "grad_norm": 3.186103105545044, "learning_rate": 9.787967875790146e-06, "loss": 1.3969, "step": 4839 }, { "epoch": 0.3793103448275862, "grad_norm": 2.579813003540039, "learning_rate": 9.787879020653627e-06, "loss": 1.1455, "step": 4840 }, { "epoch": 0.3793887147335423, "grad_norm": 2.2960245609283447, "learning_rate": 9.787790147306476e-06, "loss": 1.174, "step": 4841 }, { "epoch": 0.3794670846394984, "grad_norm": 3.012965202331543, "learning_rate": 9.78770125574903e-06, "loss": 1.1459, "step": 4842 }, { "epoch": 0.3795454545454545, "grad_norm": 3.6464290618896484, "learning_rate": 9.787612345981628e-06, "loss": 2.0518, "step": 4843 }, { "epoch": 0.3796238244514107, "grad_norm": 2.007181167602539, "learning_rate": 9.787523418004607e-06, "loss": 1.3802, "step": 4844 }, { "epoch": 0.3797021943573668, "grad_norm": 4.215270519256592, "learning_rate": 9.787434471818307e-06, "loss": 2.0117, "step": 4845 }, { "epoch": 0.3797805642633229, "grad_norm": 3.847459316253662, "learning_rate": 9.787345507423064e-06, "loss": 1.4404, "step": 4846 }, { "epoch": 0.379858934169279, "grad_norm": 11.259507179260254, "learning_rate": 9.787256524819219e-06, "loss": 1.1209, "step": 4847 }, { "epoch": 0.3799373040752351, "grad_norm": 4.032976150512695, "learning_rate": 9.787167524007109e-06, "loss": 1.2983, "step": 4848 }, { "epoch": 0.38001567398119124, "grad_norm": 3.338344097137451, "learning_rate": 9.78707850498707e-06, "loss": 1.4028, "step": 4849 }, { "epoch": 0.38009404388714735, "grad_norm": 3.673170566558838, "learning_rate": 9.786989467759445e-06, "loss": 1.0358, "step": 4850 }, { "epoch": 0.38017241379310346, "grad_norm": 2.823974132537842, "learning_rate": 9.78690041232457e-06, "loss": 1.4173, "step": 4851 }, { "epoch": 0.38025078369905957, "grad_norm": 2.2888376712799072, "learning_rate": 9.786811338682786e-06, "loss": 1.1357, "step": 4852 }, { "epoch": 0.3803291536050157, "grad_norm": 2.5585501194000244, "learning_rate": 9.786722246834431e-06, "loss": 1.2248, "step": 4853 }, { "epoch": 0.3804075235109718, "grad_norm": 4.596381664276123, "learning_rate": 9.786633136779842e-06, "loss": 1.9501, "step": 4854 }, { "epoch": 0.3804858934169279, "grad_norm": 4.258945465087891, "learning_rate": 9.786544008519358e-06, "loss": 1.9635, "step": 4855 }, { "epoch": 0.380564263322884, "grad_norm": 3.830554962158203, "learning_rate": 9.786454862053321e-06, "loss": 1.4082, "step": 4856 }, { "epoch": 0.3806426332288401, "grad_norm": 11.451960563659668, "learning_rate": 9.786365697382067e-06, "loss": 1.6183, "step": 4857 }, { "epoch": 0.38072100313479623, "grad_norm": 3.0807130336761475, "learning_rate": 9.786276514505936e-06, "loss": 1.2879, "step": 4858 }, { "epoch": 0.38079937304075234, "grad_norm": 2.3946168422698975, "learning_rate": 9.78618731342527e-06, "loss": 1.4756, "step": 4859 }, { "epoch": 0.38087774294670845, "grad_norm": 3.1665375232696533, "learning_rate": 9.786098094140404e-06, "loss": 1.4376, "step": 4860 }, { "epoch": 0.38095611285266456, "grad_norm": 3.154738664627075, "learning_rate": 9.786008856651678e-06, "loss": 1.1449, "step": 4861 }, { "epoch": 0.3810344827586207, "grad_norm": 3.222203493118286, "learning_rate": 9.785919600959435e-06, "loss": 1.5011, "step": 4862 }, { "epoch": 0.3811128526645768, "grad_norm": 2.463857889175415, "learning_rate": 9.78583032706401e-06, "loss": 1.0544, "step": 4863 }, { "epoch": 0.3811912225705329, "grad_norm": 7.012642860412598, "learning_rate": 9.785741034965746e-06, "loss": 1.3667, "step": 4864 }, { "epoch": 0.381269592476489, "grad_norm": 4.3894782066345215, "learning_rate": 9.78565172466498e-06, "loss": 1.0414, "step": 4865 }, { "epoch": 0.3813479623824451, "grad_norm": 2.2459521293640137, "learning_rate": 9.785562396162054e-06, "loss": 1.4165, "step": 4866 }, { "epoch": 0.3814263322884012, "grad_norm": 3.5849337577819824, "learning_rate": 9.785473049457306e-06, "loss": 1.7558, "step": 4867 }, { "epoch": 0.3815047021943574, "grad_norm": 3.181520700454712, "learning_rate": 9.785383684551077e-06, "loss": 1.9076, "step": 4868 }, { "epoch": 0.3815830721003135, "grad_norm": 2.5320634841918945, "learning_rate": 9.785294301443706e-06, "loss": 1.3228, "step": 4869 }, { "epoch": 0.3816614420062696, "grad_norm": 3.139341354370117, "learning_rate": 9.785204900135533e-06, "loss": 1.3612, "step": 4870 }, { "epoch": 0.3817398119122257, "grad_norm": 4.087210178375244, "learning_rate": 9.7851154806269e-06, "loss": 1.6157, "step": 4871 }, { "epoch": 0.38181818181818183, "grad_norm": 3.628643751144409, "learning_rate": 9.785026042918143e-06, "loss": 1.5454, "step": 4872 }, { "epoch": 0.38189655172413794, "grad_norm": 3.8824048042297363, "learning_rate": 9.784936587009606e-06, "loss": 1.7258, "step": 4873 }, { "epoch": 0.38197492163009406, "grad_norm": 2.792907953262329, "learning_rate": 9.784847112901628e-06, "loss": 1.2062, "step": 4874 }, { "epoch": 0.38205329153605017, "grad_norm": 3.1442861557006836, "learning_rate": 9.78475762059455e-06, "loss": 1.4339, "step": 4875 }, { "epoch": 0.3821316614420063, "grad_norm": 6.147545337677002, "learning_rate": 9.78466811008871e-06, "loss": 1.194, "step": 4876 }, { "epoch": 0.3822100313479624, "grad_norm": 3.2831125259399414, "learning_rate": 9.784578581384451e-06, "loss": 1.4705, "step": 4877 }, { "epoch": 0.3822884012539185, "grad_norm": 9.780365943908691, "learning_rate": 9.784489034482112e-06, "loss": 1.8475, "step": 4878 }, { "epoch": 0.3823667711598746, "grad_norm": 3.8540778160095215, "learning_rate": 9.784399469382035e-06, "loss": 1.5408, "step": 4879 }, { "epoch": 0.3824451410658307, "grad_norm": 2.9280741214752197, "learning_rate": 9.78430988608456e-06, "loss": 1.6262, "step": 4880 }, { "epoch": 0.38252351097178683, "grad_norm": 3.565258264541626, "learning_rate": 9.784220284590028e-06, "loss": 1.4785, "step": 4881 }, { "epoch": 0.38260188087774294, "grad_norm": 4.175281524658203, "learning_rate": 9.784130664898777e-06, "loss": 1.2961, "step": 4882 }, { "epoch": 0.38268025078369905, "grad_norm": 11.532373428344727, "learning_rate": 9.784041027011151e-06, "loss": 1.0661, "step": 4883 }, { "epoch": 0.38275862068965516, "grad_norm": 5.36152458190918, "learning_rate": 9.783951370927492e-06, "loss": 1.6305, "step": 4884 }, { "epoch": 0.38283699059561127, "grad_norm": 3.598067283630371, "learning_rate": 9.783861696648137e-06, "loss": 1.4603, "step": 4885 }, { "epoch": 0.3829153605015674, "grad_norm": 3.632312059402466, "learning_rate": 9.78377200417343e-06, "loss": 1.1517, "step": 4886 }, { "epoch": 0.3829937304075235, "grad_norm": 10.181333541870117, "learning_rate": 9.783682293503711e-06, "loss": 1.1721, "step": 4887 }, { "epoch": 0.3830721003134796, "grad_norm": 2.000673770904541, "learning_rate": 9.783592564639322e-06, "loss": 1.3209, "step": 4888 }, { "epoch": 0.3831504702194357, "grad_norm": 2.43749737739563, "learning_rate": 9.783502817580606e-06, "loss": 1.6644, "step": 4889 }, { "epoch": 0.3832288401253918, "grad_norm": 2.2036375999450684, "learning_rate": 9.783413052327899e-06, "loss": 1.0948, "step": 4890 }, { "epoch": 0.383307210031348, "grad_norm": 3.0891196727752686, "learning_rate": 9.783323268881547e-06, "loss": 1.5385, "step": 4891 }, { "epoch": 0.3833855799373041, "grad_norm": 3.8844850063323975, "learning_rate": 9.78323346724189e-06, "loss": 1.2507, "step": 4892 }, { "epoch": 0.3834639498432602, "grad_norm": 3.3274824619293213, "learning_rate": 9.78314364740927e-06, "loss": 1.4761, "step": 4893 }, { "epoch": 0.3835423197492163, "grad_norm": 4.240399360656738, "learning_rate": 9.78305380938403e-06, "loss": 1.455, "step": 4894 }, { "epoch": 0.38362068965517243, "grad_norm": 5.5157694816589355, "learning_rate": 9.782963953166507e-06, "loss": 1.0836, "step": 4895 }, { "epoch": 0.38369905956112854, "grad_norm": 4.615471363067627, "learning_rate": 9.782874078757047e-06, "loss": 1.1986, "step": 4896 }, { "epoch": 0.38377742946708465, "grad_norm": 4.468235492706299, "learning_rate": 9.782784186155992e-06, "loss": 1.3515, "step": 4897 }, { "epoch": 0.38385579937304076, "grad_norm": 2.2308387756347656, "learning_rate": 9.78269427536368e-06, "loss": 1.1415, "step": 4898 }, { "epoch": 0.3839341692789969, "grad_norm": 3.8566133975982666, "learning_rate": 9.782604346380457e-06, "loss": 1.3447, "step": 4899 }, { "epoch": 0.384012539184953, "grad_norm": 3.0400893688201904, "learning_rate": 9.782514399206664e-06, "loss": 1.4812, "step": 4900 }, { "epoch": 0.3840909090909091, "grad_norm": 3.5565335750579834, "learning_rate": 9.78242443384264e-06, "loss": 1.4161, "step": 4901 }, { "epoch": 0.3841692789968652, "grad_norm": 3.148381471633911, "learning_rate": 9.782334450288733e-06, "loss": 1.6227, "step": 4902 }, { "epoch": 0.3842476489028213, "grad_norm": 6.846405029296875, "learning_rate": 9.782244448545282e-06, "loss": 1.6389, "step": 4903 }, { "epoch": 0.3843260188087774, "grad_norm": 3.4165332317352295, "learning_rate": 9.782154428612627e-06, "loss": 1.1504, "step": 4904 }, { "epoch": 0.38440438871473354, "grad_norm": 2.3986284732818604, "learning_rate": 9.782064390491115e-06, "loss": 1.1201, "step": 4905 }, { "epoch": 0.38448275862068965, "grad_norm": 5.4434895515441895, "learning_rate": 9.781974334181085e-06, "loss": 1.237, "step": 4906 }, { "epoch": 0.38456112852664576, "grad_norm": 3.560786008834839, "learning_rate": 9.781884259682881e-06, "loss": 1.5466, "step": 4907 }, { "epoch": 0.38463949843260187, "grad_norm": 3.653294324874878, "learning_rate": 9.781794166996847e-06, "loss": 1.5427, "step": 4908 }, { "epoch": 0.384717868338558, "grad_norm": 3.975822925567627, "learning_rate": 9.781704056123323e-06, "loss": 1.5757, "step": 4909 }, { "epoch": 0.3847962382445141, "grad_norm": 6.395143032073975, "learning_rate": 9.781613927062652e-06, "loss": 0.7524, "step": 4910 }, { "epoch": 0.3848746081504702, "grad_norm": 2.3984267711639404, "learning_rate": 9.781523779815178e-06, "loss": 1.5845, "step": 4911 }, { "epoch": 0.3849529780564263, "grad_norm": 2.8493003845214844, "learning_rate": 9.781433614381243e-06, "loss": 1.4415, "step": 4912 }, { "epoch": 0.3850313479623824, "grad_norm": 3.629319667816162, "learning_rate": 9.781343430761192e-06, "loss": 1.5751, "step": 4913 }, { "epoch": 0.38510971786833853, "grad_norm": 3.0655252933502197, "learning_rate": 9.781253228955366e-06, "loss": 1.0289, "step": 4914 }, { "epoch": 0.3851880877742947, "grad_norm": 3.0395994186401367, "learning_rate": 9.781163008964108e-06, "loss": 1.5798, "step": 4915 }, { "epoch": 0.3852664576802508, "grad_norm": 2.861985206604004, "learning_rate": 9.781072770787761e-06, "loss": 1.7883, "step": 4916 }, { "epoch": 0.3853448275862069, "grad_norm": 3.5333337783813477, "learning_rate": 9.78098251442667e-06, "loss": 1.6658, "step": 4917 }, { "epoch": 0.385423197492163, "grad_norm": 5.849081516265869, "learning_rate": 9.780892239881177e-06, "loss": 1.4704, "step": 4918 }, { "epoch": 0.38550156739811914, "grad_norm": 4.106156826019287, "learning_rate": 9.780801947151626e-06, "loss": 0.9873, "step": 4919 }, { "epoch": 0.38557993730407525, "grad_norm": 3.9085233211517334, "learning_rate": 9.780711636238359e-06, "loss": 1.1857, "step": 4920 }, { "epoch": 0.38565830721003136, "grad_norm": 3.958399534225464, "learning_rate": 9.78062130714172e-06, "loss": 1.319, "step": 4921 }, { "epoch": 0.38573667711598747, "grad_norm": 2.0817341804504395, "learning_rate": 9.780530959862055e-06, "loss": 1.3689, "step": 4922 }, { "epoch": 0.3858150470219436, "grad_norm": 3.079315185546875, "learning_rate": 9.780440594399706e-06, "loss": 1.8302, "step": 4923 }, { "epoch": 0.3858934169278997, "grad_norm": 3.4120945930480957, "learning_rate": 9.780350210755015e-06, "loss": 1.1906, "step": 4924 }, { "epoch": 0.3859717868338558, "grad_norm": 2.54775071144104, "learning_rate": 9.780259808928324e-06, "loss": 1.3889, "step": 4925 }, { "epoch": 0.3860501567398119, "grad_norm": 3.033909559249878, "learning_rate": 9.780169388919984e-06, "loss": 1.1274, "step": 4926 }, { "epoch": 0.386128526645768, "grad_norm": 2.140746593475342, "learning_rate": 9.780078950730334e-06, "loss": 1.318, "step": 4927 }, { "epoch": 0.38620689655172413, "grad_norm": 2.4922397136688232, "learning_rate": 9.779988494359721e-06, "loss": 1.4907, "step": 4928 }, { "epoch": 0.38628526645768024, "grad_norm": 2.5665175914764404, "learning_rate": 9.779898019808483e-06, "loss": 1.3498, "step": 4929 }, { "epoch": 0.38636363636363635, "grad_norm": 3.3895211219787598, "learning_rate": 9.77980752707697e-06, "loss": 1.1239, "step": 4930 }, { "epoch": 0.38644200626959246, "grad_norm": 2.748248338699341, "learning_rate": 9.779717016165524e-06, "loss": 1.3485, "step": 4931 }, { "epoch": 0.3865203761755486, "grad_norm": 2.1507468223571777, "learning_rate": 9.77962648707449e-06, "loss": 1.5046, "step": 4932 }, { "epoch": 0.3865987460815047, "grad_norm": 3.354031562805176, "learning_rate": 9.779535939804211e-06, "loss": 1.6174, "step": 4933 }, { "epoch": 0.3866771159874608, "grad_norm": 5.965207576751709, "learning_rate": 9.779445374355032e-06, "loss": 1.2674, "step": 4934 }, { "epoch": 0.3867554858934169, "grad_norm": 2.8095591068267822, "learning_rate": 9.779354790727299e-06, "loss": 1.4801, "step": 4935 }, { "epoch": 0.386833855799373, "grad_norm": 4.721414089202881, "learning_rate": 9.779264188921354e-06, "loss": 1.1045, "step": 4936 }, { "epoch": 0.3869122257053291, "grad_norm": 2.2740955352783203, "learning_rate": 9.779173568937544e-06, "loss": 1.3793, "step": 4937 }, { "epoch": 0.3869905956112853, "grad_norm": 2.8183436393737793, "learning_rate": 9.779082930776211e-06, "loss": 1.5211, "step": 4938 }, { "epoch": 0.3870689655172414, "grad_norm": 3.104916572570801, "learning_rate": 9.778992274437701e-06, "loss": 1.2735, "step": 4939 }, { "epoch": 0.3871473354231975, "grad_norm": 2.5606765747070312, "learning_rate": 9.77890159992236e-06, "loss": 1.3893, "step": 4940 }, { "epoch": 0.3872257053291536, "grad_norm": 7.369415283203125, "learning_rate": 9.778810907230535e-06, "loss": 1.6183, "step": 4941 }, { "epoch": 0.38730407523510973, "grad_norm": 2.2937936782836914, "learning_rate": 9.778720196362564e-06, "loss": 1.3413, "step": 4942 }, { "epoch": 0.38738244514106585, "grad_norm": 2.1722779273986816, "learning_rate": 9.778629467318795e-06, "loss": 1.273, "step": 4943 }, { "epoch": 0.38746081504702196, "grad_norm": 3.1802103519439697, "learning_rate": 9.778538720099575e-06, "loss": 1.012, "step": 4944 }, { "epoch": 0.38753918495297807, "grad_norm": 2.4548962116241455, "learning_rate": 9.778447954705251e-06, "loss": 1.5242, "step": 4945 }, { "epoch": 0.3876175548589342, "grad_norm": 1.7063547372817993, "learning_rate": 9.778357171136161e-06, "loss": 1.1881, "step": 4946 }, { "epoch": 0.3876959247648903, "grad_norm": 3.282911539077759, "learning_rate": 9.778266369392657e-06, "loss": 1.2563, "step": 4947 }, { "epoch": 0.3877742946708464, "grad_norm": 3.624070167541504, "learning_rate": 9.778175549475081e-06, "loss": 1.3615, "step": 4948 }, { "epoch": 0.3878526645768025, "grad_norm": 6.851377010345459, "learning_rate": 9.778084711383782e-06, "loss": 1.127, "step": 4949 }, { "epoch": 0.3879310344827586, "grad_norm": 3.0104329586029053, "learning_rate": 9.777993855119101e-06, "loss": 1.3845, "step": 4950 }, { "epoch": 0.38800940438871473, "grad_norm": 2.36002516746521, "learning_rate": 9.777902980681386e-06, "loss": 1.3287, "step": 4951 }, { "epoch": 0.38808777429467084, "grad_norm": 3.2959957122802734, "learning_rate": 9.777812088070982e-06, "loss": 1.617, "step": 4952 }, { "epoch": 0.38816614420062695, "grad_norm": 3.409311532974243, "learning_rate": 9.777721177288233e-06, "loss": 1.3755, "step": 4953 }, { "epoch": 0.38824451410658306, "grad_norm": 2.4759252071380615, "learning_rate": 9.777630248333489e-06, "loss": 1.575, "step": 4954 }, { "epoch": 0.38832288401253917, "grad_norm": 3.2846713066101074, "learning_rate": 9.777539301207094e-06, "loss": 1.3467, "step": 4955 }, { "epoch": 0.3884012539184953, "grad_norm": 2.936387777328491, "learning_rate": 9.77744833590939e-06, "loss": 1.1085, "step": 4956 }, { "epoch": 0.3884796238244514, "grad_norm": 4.601265907287598, "learning_rate": 9.777357352440729e-06, "loss": 1.2989, "step": 4957 }, { "epoch": 0.3885579937304075, "grad_norm": 4.686232566833496, "learning_rate": 9.777266350801454e-06, "loss": 1.3948, "step": 4958 }, { "epoch": 0.3886363636363636, "grad_norm": 1.8275816440582275, "learning_rate": 9.777175330991912e-06, "loss": 1.4513, "step": 4959 }, { "epoch": 0.3887147335423197, "grad_norm": 3.7110977172851562, "learning_rate": 9.777084293012448e-06, "loss": 1.5823, "step": 4960 }, { "epoch": 0.3887931034482759, "grad_norm": 2.4725852012634277, "learning_rate": 9.77699323686341e-06, "loss": 1.2413, "step": 4961 }, { "epoch": 0.388871473354232, "grad_norm": 1.9848090410232544, "learning_rate": 9.776902162545141e-06, "loss": 1.3823, "step": 4962 }, { "epoch": 0.3889498432601881, "grad_norm": 7.498600482940674, "learning_rate": 9.776811070057993e-06, "loss": 1.3961, "step": 4963 }, { "epoch": 0.3890282131661442, "grad_norm": 2.9146716594696045, "learning_rate": 9.776719959402306e-06, "loss": 1.0607, "step": 4964 }, { "epoch": 0.38910658307210033, "grad_norm": 22.3254451751709, "learning_rate": 9.776628830578431e-06, "loss": 1.3179, "step": 4965 }, { "epoch": 0.38918495297805644, "grad_norm": 4.646812438964844, "learning_rate": 9.776537683586715e-06, "loss": 1.3116, "step": 4966 }, { "epoch": 0.38926332288401255, "grad_norm": 2.3647048473358154, "learning_rate": 9.7764465184275e-06, "loss": 1.3113, "step": 4967 }, { "epoch": 0.38934169278996866, "grad_norm": 2.2499210834503174, "learning_rate": 9.776355335101139e-06, "loss": 1.7434, "step": 4968 }, { "epoch": 0.3894200626959248, "grad_norm": 2.1550674438476562, "learning_rate": 9.776264133607973e-06, "loss": 1.254, "step": 4969 }, { "epoch": 0.3894984326018809, "grad_norm": 3.0725202560424805, "learning_rate": 9.776172913948353e-06, "loss": 1.5498, "step": 4970 }, { "epoch": 0.389576802507837, "grad_norm": 3.4682068824768066, "learning_rate": 9.776081676122624e-06, "loss": 1.3055, "step": 4971 }, { "epoch": 0.3896551724137931, "grad_norm": 1.7617831230163574, "learning_rate": 9.775990420131134e-06, "loss": 0.9642, "step": 4972 }, { "epoch": 0.3897335423197492, "grad_norm": 2.382284641265869, "learning_rate": 9.775899145974227e-06, "loss": 1.3543, "step": 4973 }, { "epoch": 0.3898119122257053, "grad_norm": 2.0058741569519043, "learning_rate": 9.775807853652255e-06, "loss": 1.5189, "step": 4974 }, { "epoch": 0.38989028213166144, "grad_norm": 3.055163621902466, "learning_rate": 9.775716543165563e-06, "loss": 1.592, "step": 4975 }, { "epoch": 0.38996865203761755, "grad_norm": 4.002503871917725, "learning_rate": 9.775625214514499e-06, "loss": 1.7511, "step": 4976 }, { "epoch": 0.39004702194357366, "grad_norm": 5.709103107452393, "learning_rate": 9.775533867699408e-06, "loss": 1.3439, "step": 4977 }, { "epoch": 0.39012539184952977, "grad_norm": 5.785693645477295, "learning_rate": 9.77544250272064e-06, "loss": 0.8702, "step": 4978 }, { "epoch": 0.3902037617554859, "grad_norm": 2.2662553787231445, "learning_rate": 9.77535111957854e-06, "loss": 1.5506, "step": 4979 }, { "epoch": 0.390282131661442, "grad_norm": 4.66221284866333, "learning_rate": 9.775259718273459e-06, "loss": 1.7196, "step": 4980 }, { "epoch": 0.3903605015673981, "grad_norm": 3.400705337524414, "learning_rate": 9.775168298805741e-06, "loss": 1.5654, "step": 4981 }, { "epoch": 0.3904388714733542, "grad_norm": 4.6524434089660645, "learning_rate": 9.775076861175738e-06, "loss": 1.6636, "step": 4982 }, { "epoch": 0.3905172413793103, "grad_norm": 3.607130289077759, "learning_rate": 9.774985405383793e-06, "loss": 1.2056, "step": 4983 }, { "epoch": 0.39059561128526643, "grad_norm": 5.052416801452637, "learning_rate": 9.774893931430256e-06, "loss": 1.1495, "step": 4984 }, { "epoch": 0.3906739811912226, "grad_norm": 2.547776699066162, "learning_rate": 9.774802439315477e-06, "loss": 1.5099, "step": 4985 }, { "epoch": 0.3907523510971787, "grad_norm": 2.951899290084839, "learning_rate": 9.774710929039801e-06, "loss": 1.5298, "step": 4986 }, { "epoch": 0.3908307210031348, "grad_norm": 2.752382755279541, "learning_rate": 9.774619400603575e-06, "loss": 1.4057, "step": 4987 }, { "epoch": 0.39090909090909093, "grad_norm": 2.9418351650238037, "learning_rate": 9.774527854007152e-06, "loss": 1.355, "step": 4988 }, { "epoch": 0.39098746081504704, "grad_norm": 2.594985008239746, "learning_rate": 9.774436289250877e-06, "loss": 1.297, "step": 4989 }, { "epoch": 0.39106583072100315, "grad_norm": 12.582456588745117, "learning_rate": 9.774344706335097e-06, "loss": 1.2026, "step": 4990 }, { "epoch": 0.39114420062695926, "grad_norm": 5.91680383682251, "learning_rate": 9.774253105260165e-06, "loss": 1.5844, "step": 4991 }, { "epoch": 0.39122257053291537, "grad_norm": 2.401405096054077, "learning_rate": 9.774161486026423e-06, "loss": 1.3509, "step": 4992 }, { "epoch": 0.3913009404388715, "grad_norm": 3.4274160861968994, "learning_rate": 9.774069848634224e-06, "loss": 1.6876, "step": 4993 }, { "epoch": 0.3913793103448276, "grad_norm": 8.069067001342773, "learning_rate": 9.773978193083918e-06, "loss": 1.7555, "step": 4994 }, { "epoch": 0.3914576802507837, "grad_norm": 2.4580342769622803, "learning_rate": 9.773886519375848e-06, "loss": 1.5019, "step": 4995 }, { "epoch": 0.3915360501567398, "grad_norm": 2.7509078979492188, "learning_rate": 9.773794827510366e-06, "loss": 1.1629, "step": 4996 }, { "epoch": 0.3916144200626959, "grad_norm": 1.766304612159729, "learning_rate": 9.773703117487821e-06, "loss": 1.3663, "step": 4997 }, { "epoch": 0.39169278996865203, "grad_norm": 4.9410810470581055, "learning_rate": 9.773611389308561e-06, "loss": 2.2552, "step": 4998 }, { "epoch": 0.39177115987460814, "grad_norm": 2.066284418106079, "learning_rate": 9.773519642972935e-06, "loss": 1.3243, "step": 4999 }, { "epoch": 0.39184952978056425, "grad_norm": 2.9409329891204834, "learning_rate": 9.773427878481294e-06, "loss": 1.5296, "step": 5000 }, { "epoch": 0.39192789968652036, "grad_norm": 2.0142087936401367, "learning_rate": 9.773336095833982e-06, "loss": 1.2719, "step": 5001 }, { "epoch": 0.3920062695924765, "grad_norm": 4.080066204071045, "learning_rate": 9.773244295031354e-06, "loss": 1.669, "step": 5002 }, { "epoch": 0.3920846394984326, "grad_norm": 2.563645839691162, "learning_rate": 9.773152476073755e-06, "loss": 1.407, "step": 5003 }, { "epoch": 0.3921630094043887, "grad_norm": 3.6590824127197266, "learning_rate": 9.773060638961535e-06, "loss": 1.565, "step": 5004 }, { "epoch": 0.3922413793103448, "grad_norm": 4.616818904876709, "learning_rate": 9.772968783695046e-06, "loss": 2.1451, "step": 5005 }, { "epoch": 0.3923197492163009, "grad_norm": 4.031745433807373, "learning_rate": 9.772876910274633e-06, "loss": 1.5778, "step": 5006 }, { "epoch": 0.392398119122257, "grad_norm": 2.2985336780548096, "learning_rate": 9.77278501870065e-06, "loss": 1.5573, "step": 5007 }, { "epoch": 0.3924764890282132, "grad_norm": 3.074640989303589, "learning_rate": 9.772693108973443e-06, "loss": 1.7737, "step": 5008 }, { "epoch": 0.3925548589341693, "grad_norm": 2.907623529434204, "learning_rate": 9.772601181093364e-06, "loss": 1.6004, "step": 5009 }, { "epoch": 0.3926332288401254, "grad_norm": 2.379081964492798, "learning_rate": 9.77250923506076e-06, "loss": 1.4945, "step": 5010 }, { "epoch": 0.3927115987460815, "grad_norm": 3.6945254802703857, "learning_rate": 9.772417270875983e-06, "loss": 1.4891, "step": 5011 }, { "epoch": 0.39278996865203764, "grad_norm": 1.942323923110962, "learning_rate": 9.772325288539381e-06, "loss": 1.376, "step": 5012 }, { "epoch": 0.39286833855799375, "grad_norm": 2.440976619720459, "learning_rate": 9.772233288051305e-06, "loss": 1.3436, "step": 5013 }, { "epoch": 0.39294670846394986, "grad_norm": 2.018739938735962, "learning_rate": 9.772141269412105e-06, "loss": 1.2398, "step": 5014 }, { "epoch": 0.39302507836990597, "grad_norm": 2.005235195159912, "learning_rate": 9.772049232622132e-06, "loss": 1.2377, "step": 5015 }, { "epoch": 0.3931034482758621, "grad_norm": 4.2368388175964355, "learning_rate": 9.771957177681732e-06, "loss": 1.3226, "step": 5016 }, { "epoch": 0.3931818181818182, "grad_norm": 2.459174871444702, "learning_rate": 9.77186510459126e-06, "loss": 1.4203, "step": 5017 }, { "epoch": 0.3932601880877743, "grad_norm": 10.233536720275879, "learning_rate": 9.771773013351065e-06, "loss": 0.9015, "step": 5018 }, { "epoch": 0.3933385579937304, "grad_norm": 3.879809856414795, "learning_rate": 9.771680903961495e-06, "loss": 0.9269, "step": 5019 }, { "epoch": 0.3934169278996865, "grad_norm": 3.53550386428833, "learning_rate": 9.7715887764229e-06, "loss": 0.7553, "step": 5020 }, { "epoch": 0.39349529780564263, "grad_norm": 2.7194113731384277, "learning_rate": 9.771496630735636e-06, "loss": 1.2528, "step": 5021 }, { "epoch": 0.39357366771159874, "grad_norm": 3.639160633087158, "learning_rate": 9.771404466900048e-06, "loss": 1.7445, "step": 5022 }, { "epoch": 0.39365203761755485, "grad_norm": 3.461174964904785, "learning_rate": 9.771312284916486e-06, "loss": 1.4785, "step": 5023 }, { "epoch": 0.39373040752351096, "grad_norm": 6.040081024169922, "learning_rate": 9.771220084785305e-06, "loss": 1.192, "step": 5024 }, { "epoch": 0.39380877742946707, "grad_norm": 2.5518476963043213, "learning_rate": 9.771127866506854e-06, "loss": 1.291, "step": 5025 }, { "epoch": 0.3938871473354232, "grad_norm": 2.8822247982025146, "learning_rate": 9.771035630081481e-06, "loss": 1.0902, "step": 5026 }, { "epoch": 0.3939655172413793, "grad_norm": 3.885582685470581, "learning_rate": 9.77094337550954e-06, "loss": 1.057, "step": 5027 }, { "epoch": 0.3940438871473354, "grad_norm": 3.014134168624878, "learning_rate": 9.770851102791381e-06, "loss": 1.5458, "step": 5028 }, { "epoch": 0.3941222570532915, "grad_norm": 1.8132708072662354, "learning_rate": 9.770758811927355e-06, "loss": 1.3713, "step": 5029 }, { "epoch": 0.3942006269592476, "grad_norm": 5.958470821380615, "learning_rate": 9.770666502917815e-06, "loss": 1.0559, "step": 5030 }, { "epoch": 0.39427899686520373, "grad_norm": 3.366276741027832, "learning_rate": 9.770574175763107e-06, "loss": 1.3584, "step": 5031 }, { "epoch": 0.3943573667711599, "grad_norm": 2.1884195804595947, "learning_rate": 9.770481830463586e-06, "loss": 1.2802, "step": 5032 }, { "epoch": 0.394435736677116, "grad_norm": 2.5792386531829834, "learning_rate": 9.770389467019602e-06, "loss": 1.7166, "step": 5033 }, { "epoch": 0.3945141065830721, "grad_norm": 2.723665475845337, "learning_rate": 9.770297085431506e-06, "loss": 1.2707, "step": 5034 }, { "epoch": 0.39459247648902823, "grad_norm": 2.2247848510742188, "learning_rate": 9.770204685699652e-06, "loss": 1.4417, "step": 5035 }, { "epoch": 0.39467084639498434, "grad_norm": 2.9004969596862793, "learning_rate": 9.77011226782439e-06, "loss": 1.1845, "step": 5036 }, { "epoch": 0.39474921630094045, "grad_norm": 2.1914584636688232, "learning_rate": 9.770019831806068e-06, "loss": 1.7381, "step": 5037 }, { "epoch": 0.39482758620689656, "grad_norm": 2.4732909202575684, "learning_rate": 9.769927377645043e-06, "loss": 1.5068, "step": 5038 }, { "epoch": 0.3949059561128527, "grad_norm": 3.6042518615722656, "learning_rate": 9.769834905341664e-06, "loss": 1.6791, "step": 5039 }, { "epoch": 0.3949843260188088, "grad_norm": 2.8074047565460205, "learning_rate": 9.769742414896281e-06, "loss": 1.4538, "step": 5040 }, { "epoch": 0.3950626959247649, "grad_norm": 2.82114839553833, "learning_rate": 9.769649906309251e-06, "loss": 1.7239, "step": 5041 }, { "epoch": 0.395141065830721, "grad_norm": 3.9493813514709473, "learning_rate": 9.76955737958092e-06, "loss": 1.3541, "step": 5042 }, { "epoch": 0.3952194357366771, "grad_norm": 5.65699577331543, "learning_rate": 9.769464834711644e-06, "loss": 1.4813, "step": 5043 }, { "epoch": 0.3952978056426332, "grad_norm": 1.939409852027893, "learning_rate": 9.769372271701773e-06, "loss": 1.3149, "step": 5044 }, { "epoch": 0.39537617554858934, "grad_norm": 2.188932180404663, "learning_rate": 9.769279690551658e-06, "loss": 1.3636, "step": 5045 }, { "epoch": 0.39545454545454545, "grad_norm": 2.112128734588623, "learning_rate": 9.769187091261656e-06, "loss": 1.5958, "step": 5046 }, { "epoch": 0.39553291536050156, "grad_norm": 2.9025373458862305, "learning_rate": 9.769094473832114e-06, "loss": 1.5993, "step": 5047 }, { "epoch": 0.39561128526645767, "grad_norm": 2.6677627563476562, "learning_rate": 9.769001838263385e-06, "loss": 0.6612, "step": 5048 }, { "epoch": 0.3956896551724138, "grad_norm": 5.414186477661133, "learning_rate": 9.768909184555825e-06, "loss": 1.4709, "step": 5049 }, { "epoch": 0.3957680250783699, "grad_norm": 3.0940403938293457, "learning_rate": 9.768816512709782e-06, "loss": 1.7216, "step": 5050 }, { "epoch": 0.395846394984326, "grad_norm": 2.4334874153137207, "learning_rate": 9.768723822725612e-06, "loss": 1.6199, "step": 5051 }, { "epoch": 0.3959247648902821, "grad_norm": 2.5433545112609863, "learning_rate": 9.768631114603664e-06, "loss": 1.4476, "step": 5052 }, { "epoch": 0.3960031347962382, "grad_norm": 3.0675082206726074, "learning_rate": 9.768538388344295e-06, "loss": 1.5542, "step": 5053 }, { "epoch": 0.39608150470219433, "grad_norm": 3.4310381412506104, "learning_rate": 9.768445643947853e-06, "loss": 1.4278, "step": 5054 }, { "epoch": 0.3961598746081505, "grad_norm": 3.0669658184051514, "learning_rate": 9.768352881414694e-06, "loss": 1.6294, "step": 5055 }, { "epoch": 0.3962382445141066, "grad_norm": 3.5431790351867676, "learning_rate": 9.76826010074517e-06, "loss": 1.4763, "step": 5056 }, { "epoch": 0.3963166144200627, "grad_norm": 3.88247013092041, "learning_rate": 9.768167301939633e-06, "loss": 1.3685, "step": 5057 }, { "epoch": 0.39639498432601883, "grad_norm": 2.693129777908325, "learning_rate": 9.768074484998437e-06, "loss": 1.034, "step": 5058 }, { "epoch": 0.39647335423197494, "grad_norm": 2.99407696723938, "learning_rate": 9.767981649921936e-06, "loss": 1.3951, "step": 5059 }, { "epoch": 0.39655172413793105, "grad_norm": 2.553938865661621, "learning_rate": 9.76788879671048e-06, "loss": 1.3544, "step": 5060 }, { "epoch": 0.39663009404388716, "grad_norm": 3.8212056159973145, "learning_rate": 9.767795925364424e-06, "loss": 1.3367, "step": 5061 }, { "epoch": 0.39670846394984327, "grad_norm": 3.2463266849517822, "learning_rate": 9.767703035884122e-06, "loss": 1.6713, "step": 5062 }, { "epoch": 0.3967868338557994, "grad_norm": 2.2746422290802, "learning_rate": 9.767610128269926e-06, "loss": 1.4861, "step": 5063 }, { "epoch": 0.3968652037617555, "grad_norm": 2.20910382270813, "learning_rate": 9.76751720252219e-06, "loss": 1.5817, "step": 5064 }, { "epoch": 0.3969435736677116, "grad_norm": 4.0343122482299805, "learning_rate": 9.767424258641268e-06, "loss": 1.246, "step": 5065 }, { "epoch": 0.3970219435736677, "grad_norm": 3.1960597038269043, "learning_rate": 9.767331296627513e-06, "loss": 1.5968, "step": 5066 }, { "epoch": 0.3971003134796238, "grad_norm": 2.1982359886169434, "learning_rate": 9.767238316481278e-06, "loss": 1.1528, "step": 5067 }, { "epoch": 0.39717868338557993, "grad_norm": 2.9419286251068115, "learning_rate": 9.767145318202916e-06, "loss": 2.289, "step": 5068 }, { "epoch": 0.39725705329153604, "grad_norm": 11.784679412841797, "learning_rate": 9.767052301792783e-06, "loss": 1.4879, "step": 5069 }, { "epoch": 0.39733542319749215, "grad_norm": 3.5396034717559814, "learning_rate": 9.76695926725123e-06, "loss": 2.1996, "step": 5070 }, { "epoch": 0.39741379310344827, "grad_norm": 2.03507661819458, "learning_rate": 9.766866214578614e-06, "loss": 1.1092, "step": 5071 }, { "epoch": 0.3974921630094044, "grad_norm": 2.8212966918945312, "learning_rate": 9.766773143775287e-06, "loss": 0.8493, "step": 5072 }, { "epoch": 0.3975705329153605, "grad_norm": 5.945255279541016, "learning_rate": 9.766680054841603e-06, "loss": 1.5448, "step": 5073 }, { "epoch": 0.3976489028213166, "grad_norm": 3.3476345539093018, "learning_rate": 9.766586947777917e-06, "loss": 1.867, "step": 5074 }, { "epoch": 0.3977272727272727, "grad_norm": 3.071580171585083, "learning_rate": 9.766493822584583e-06, "loss": 2.0583, "step": 5075 }, { "epoch": 0.3978056426332288, "grad_norm": 2.887596845626831, "learning_rate": 9.766400679261953e-06, "loss": 1.6884, "step": 5076 }, { "epoch": 0.39788401253918493, "grad_norm": 3.5669755935668945, "learning_rate": 9.766307517810383e-06, "loss": 1.457, "step": 5077 }, { "epoch": 0.39796238244514104, "grad_norm": 3.2372872829437256, "learning_rate": 9.766214338230229e-06, "loss": 1.3642, "step": 5078 }, { "epoch": 0.3980407523510972, "grad_norm": 2.330066680908203, "learning_rate": 9.766121140521843e-06, "loss": 0.7744, "step": 5079 }, { "epoch": 0.3981191222570533, "grad_norm": 2.611058473587036, "learning_rate": 9.766027924685579e-06, "loss": 1.3757, "step": 5080 }, { "epoch": 0.3981974921630094, "grad_norm": 3.266230344772339, "learning_rate": 9.765934690721794e-06, "loss": 1.5159, "step": 5081 }, { "epoch": 0.39827586206896554, "grad_norm": 3.2147011756896973, "learning_rate": 9.76584143863084e-06, "loss": 1.5224, "step": 5082 }, { "epoch": 0.39835423197492165, "grad_norm": 3.3601880073547363, "learning_rate": 9.765748168413076e-06, "loss": 1.8592, "step": 5083 }, { "epoch": 0.39843260188087776, "grad_norm": 12.144098281860352, "learning_rate": 9.765654880068851e-06, "loss": 1.5436, "step": 5084 }, { "epoch": 0.39851097178683387, "grad_norm": 5.097958564758301, "learning_rate": 9.765561573598525e-06, "loss": 1.4525, "step": 5085 }, { "epoch": 0.39858934169279, "grad_norm": 2.3619394302368164, "learning_rate": 9.765468249002449e-06, "loss": 1.3407, "step": 5086 }, { "epoch": 0.3986677115987461, "grad_norm": 3.3900723457336426, "learning_rate": 9.76537490628098e-06, "loss": 1.3882, "step": 5087 }, { "epoch": 0.3987460815047022, "grad_norm": 5.626278877258301, "learning_rate": 9.76528154543447e-06, "loss": 1.027, "step": 5088 }, { "epoch": 0.3988244514106583, "grad_norm": 4.673269748687744, "learning_rate": 9.76518816646328e-06, "loss": 1.5822, "step": 5089 }, { "epoch": 0.3989028213166144, "grad_norm": 3.7407872676849365, "learning_rate": 9.76509476936776e-06, "loss": 1.5247, "step": 5090 }, { "epoch": 0.39898119122257053, "grad_norm": 5.203621864318848, "learning_rate": 9.76500135414827e-06, "loss": 2.0091, "step": 5091 }, { "epoch": 0.39905956112852664, "grad_norm": 5.016223907470703, "learning_rate": 9.764907920805158e-06, "loss": 2.147, "step": 5092 }, { "epoch": 0.39913793103448275, "grad_norm": 1.585628867149353, "learning_rate": 9.764814469338786e-06, "loss": 1.0128, "step": 5093 }, { "epoch": 0.39921630094043886, "grad_norm": 2.822378158569336, "learning_rate": 9.764720999749506e-06, "loss": 1.2583, "step": 5094 }, { "epoch": 0.399294670846395, "grad_norm": 3.040325880050659, "learning_rate": 9.764627512037676e-06, "loss": 0.9055, "step": 5095 }, { "epoch": 0.3993730407523511, "grad_norm": 2.3212504386901855, "learning_rate": 9.76453400620365e-06, "loss": 1.2964, "step": 5096 }, { "epoch": 0.3994514106583072, "grad_norm": 4.315764427185059, "learning_rate": 9.764440482247784e-06, "loss": 2.0244, "step": 5097 }, { "epoch": 0.3995297805642633, "grad_norm": 2.3120317459106445, "learning_rate": 9.764346940170433e-06, "loss": 1.4838, "step": 5098 }, { "epoch": 0.3996081504702194, "grad_norm": 2.4271531105041504, "learning_rate": 9.764253379971954e-06, "loss": 1.3754, "step": 5099 }, { "epoch": 0.3996865203761755, "grad_norm": 6.9782891273498535, "learning_rate": 9.7641598016527e-06, "loss": 0.6999, "step": 5100 }, { "epoch": 0.39976489028213164, "grad_norm": 9.165565490722656, "learning_rate": 9.76406620521303e-06, "loss": 1.298, "step": 5101 }, { "epoch": 0.3998432601880878, "grad_norm": 1.7558575868606567, "learning_rate": 9.763972590653301e-06, "loss": 1.1199, "step": 5102 }, { "epoch": 0.3999216300940439, "grad_norm": 6.697197914123535, "learning_rate": 9.763878957973865e-06, "loss": 1.0082, "step": 5103 }, { "epoch": 0.4, "grad_norm": 3.277932643890381, "learning_rate": 9.763785307175082e-06, "loss": 1.4287, "step": 5104 }, { "epoch": 0.40007836990595613, "grad_norm": 3.700279712677002, "learning_rate": 9.763691638257306e-06, "loss": 1.3026, "step": 5105 }, { "epoch": 0.40015673981191224, "grad_norm": 8.085521697998047, "learning_rate": 9.763597951220894e-06, "loss": 1.6796, "step": 5106 }, { "epoch": 0.40023510971786835, "grad_norm": 5.451086521148682, "learning_rate": 9.763504246066202e-06, "loss": 1.3464, "step": 5107 }, { "epoch": 0.40031347962382446, "grad_norm": 6.762756824493408, "learning_rate": 9.763410522793586e-06, "loss": 0.8283, "step": 5108 }, { "epoch": 0.4003918495297806, "grad_norm": 3.13547420501709, "learning_rate": 9.763316781403402e-06, "loss": 1.0441, "step": 5109 }, { "epoch": 0.4004702194357367, "grad_norm": 3.5781733989715576, "learning_rate": 9.76322302189601e-06, "loss": 1.0866, "step": 5110 }, { "epoch": 0.4005485893416928, "grad_norm": 2.735300064086914, "learning_rate": 9.763129244271763e-06, "loss": 1.6161, "step": 5111 }, { "epoch": 0.4006269592476489, "grad_norm": 3.4657835960388184, "learning_rate": 9.76303544853102e-06, "loss": 2.0098, "step": 5112 }, { "epoch": 0.400705329153605, "grad_norm": 3.1373682022094727, "learning_rate": 9.762941634674135e-06, "loss": 1.7252, "step": 5113 }, { "epoch": 0.4007836990595611, "grad_norm": 2.0543127059936523, "learning_rate": 9.762847802701466e-06, "loss": 1.3764, "step": 5114 }, { "epoch": 0.40086206896551724, "grad_norm": 2.3010003566741943, "learning_rate": 9.762753952613372e-06, "loss": 1.368, "step": 5115 }, { "epoch": 0.40094043887147335, "grad_norm": 7.465642929077148, "learning_rate": 9.762660084410207e-06, "loss": 2.0917, "step": 5116 }, { "epoch": 0.40101880877742946, "grad_norm": 2.5171306133270264, "learning_rate": 9.76256619809233e-06, "loss": 1.3691, "step": 5117 }, { "epoch": 0.40109717868338557, "grad_norm": 2.113248586654663, "learning_rate": 9.762472293660097e-06, "loss": 1.1907, "step": 5118 }, { "epoch": 0.4011755485893417, "grad_norm": 4.5460638999938965, "learning_rate": 9.762378371113867e-06, "loss": 1.3144, "step": 5119 }, { "epoch": 0.4012539184952978, "grad_norm": 3.928450584411621, "learning_rate": 9.762284430453995e-06, "loss": 1.5428, "step": 5120 }, { "epoch": 0.4013322884012539, "grad_norm": 2.790627956390381, "learning_rate": 9.762190471680838e-06, "loss": 1.5671, "step": 5121 }, { "epoch": 0.40141065830721, "grad_norm": 2.3554325103759766, "learning_rate": 9.762096494794755e-06, "loss": 1.4253, "step": 5122 }, { "epoch": 0.4014890282131661, "grad_norm": 3.647740364074707, "learning_rate": 9.762002499796103e-06, "loss": 1.0659, "step": 5123 }, { "epoch": 0.40156739811912223, "grad_norm": 2.8602280616760254, "learning_rate": 9.761908486685239e-06, "loss": 1.7662, "step": 5124 }, { "epoch": 0.40164576802507834, "grad_norm": 5.345160007476807, "learning_rate": 9.761814455462521e-06, "loss": 1.3348, "step": 5125 }, { "epoch": 0.4017241379310345, "grad_norm": 2.6512906551361084, "learning_rate": 9.761720406128308e-06, "loss": 1.447, "step": 5126 }, { "epoch": 0.4018025078369906, "grad_norm": 2.897397518157959, "learning_rate": 9.761626338682955e-06, "loss": 1.4832, "step": 5127 }, { "epoch": 0.40188087774294673, "grad_norm": 4.642064571380615, "learning_rate": 9.761532253126822e-06, "loss": 1.4373, "step": 5128 }, { "epoch": 0.40195924764890284, "grad_norm": 11.27294635772705, "learning_rate": 9.761438149460265e-06, "loss": 2.6316, "step": 5129 }, { "epoch": 0.40203761755485895, "grad_norm": 3.86191725730896, "learning_rate": 9.761344027683643e-06, "loss": 1.418, "step": 5130 }, { "epoch": 0.40211598746081506, "grad_norm": 2.528520107269287, "learning_rate": 9.761249887797316e-06, "loss": 1.111, "step": 5131 }, { "epoch": 0.40219435736677117, "grad_norm": 2.0180580615997314, "learning_rate": 9.761155729801637e-06, "loss": 1.3045, "step": 5132 }, { "epoch": 0.4022727272727273, "grad_norm": 2.0199496746063232, "learning_rate": 9.76106155369697e-06, "loss": 1.4067, "step": 5133 }, { "epoch": 0.4023510971786834, "grad_norm": 3.180917263031006, "learning_rate": 9.760967359483668e-06, "loss": 1.2398, "step": 5134 }, { "epoch": 0.4024294670846395, "grad_norm": 2.9780519008636475, "learning_rate": 9.760873147162091e-06, "loss": 1.3875, "step": 5135 }, { "epoch": 0.4025078369905956, "grad_norm": 1.9688715934753418, "learning_rate": 9.7607789167326e-06, "loss": 1.3254, "step": 5136 }, { "epoch": 0.4025862068965517, "grad_norm": 1.8797268867492676, "learning_rate": 9.76068466819555e-06, "loss": 1.2099, "step": 5137 }, { "epoch": 0.40266457680250783, "grad_norm": 2.1530277729034424, "learning_rate": 9.760590401551302e-06, "loss": 1.4041, "step": 5138 }, { "epoch": 0.40274294670846394, "grad_norm": 3.43621563911438, "learning_rate": 9.76049611680021e-06, "loss": 1.6124, "step": 5139 }, { "epoch": 0.40282131661442006, "grad_norm": 3.1859354972839355, "learning_rate": 9.760401813942641e-06, "loss": 1.2734, "step": 5140 }, { "epoch": 0.40289968652037617, "grad_norm": 4.12294340133667, "learning_rate": 9.760307492978945e-06, "loss": 1.579, "step": 5141 }, { "epoch": 0.4029780564263323, "grad_norm": 4.176163673400879, "learning_rate": 9.760213153909485e-06, "loss": 1.7321, "step": 5142 }, { "epoch": 0.4030564263322884, "grad_norm": 3.447249174118042, "learning_rate": 9.76011879673462e-06, "loss": 1.6861, "step": 5143 }, { "epoch": 0.4031347962382445, "grad_norm": 2.3111114501953125, "learning_rate": 9.760024421454708e-06, "loss": 1.428, "step": 5144 }, { "epoch": 0.4032131661442006, "grad_norm": 2.2976369857788086, "learning_rate": 9.759930028070108e-06, "loss": 1.4172, "step": 5145 }, { "epoch": 0.4032915360501567, "grad_norm": 2.3826212882995605, "learning_rate": 9.759835616581176e-06, "loss": 1.7443, "step": 5146 }, { "epoch": 0.40336990595611283, "grad_norm": 3.0004289150238037, "learning_rate": 9.759741186988279e-06, "loss": 1.1591, "step": 5147 }, { "epoch": 0.40344827586206894, "grad_norm": 2.129748821258545, "learning_rate": 9.759646739291769e-06, "loss": 1.1268, "step": 5148 }, { "epoch": 0.4035266457680251, "grad_norm": 6.821484565734863, "learning_rate": 9.759552273492006e-06, "loss": 1.3981, "step": 5149 }, { "epoch": 0.4036050156739812, "grad_norm": 2.942192792892456, "learning_rate": 9.759457789589352e-06, "loss": 1.1062, "step": 5150 }, { "epoch": 0.4036833855799373, "grad_norm": 3.590940475463867, "learning_rate": 9.759363287584168e-06, "loss": 1.4621, "step": 5151 }, { "epoch": 0.40376175548589344, "grad_norm": 12.59056282043457, "learning_rate": 9.759268767476806e-06, "loss": 2.798, "step": 5152 }, { "epoch": 0.40384012539184955, "grad_norm": 4.174264907836914, "learning_rate": 9.759174229267633e-06, "loss": 1.8051, "step": 5153 }, { "epoch": 0.40391849529780566, "grad_norm": 5.6661176681518555, "learning_rate": 9.759079672957004e-06, "loss": 0.529, "step": 5154 }, { "epoch": 0.40399686520376177, "grad_norm": 2.2479677200317383, "learning_rate": 9.758985098545282e-06, "loss": 1.4201, "step": 5155 }, { "epoch": 0.4040752351097179, "grad_norm": 3.287404775619507, "learning_rate": 9.758890506032824e-06, "loss": 1.6692, "step": 5156 }, { "epoch": 0.404153605015674, "grad_norm": 2.332035779953003, "learning_rate": 9.758795895419992e-06, "loss": 1.5902, "step": 5157 }, { "epoch": 0.4042319749216301, "grad_norm": 2.370502471923828, "learning_rate": 9.758701266707144e-06, "loss": 1.2125, "step": 5158 }, { "epoch": 0.4043103448275862, "grad_norm": 2.5573947429656982, "learning_rate": 9.75860661989464e-06, "loss": 1.3869, "step": 5159 }, { "epoch": 0.4043887147335423, "grad_norm": 2.558368682861328, "learning_rate": 9.758511954982841e-06, "loss": 1.5328, "step": 5160 }, { "epoch": 0.40446708463949843, "grad_norm": 3.578423261642456, "learning_rate": 9.758417271972108e-06, "loss": 1.5413, "step": 5161 }, { "epoch": 0.40454545454545454, "grad_norm": 5.419931888580322, "learning_rate": 9.758322570862798e-06, "loss": 1.4146, "step": 5162 }, { "epoch": 0.40462382445141065, "grad_norm": 2.966808557510376, "learning_rate": 9.758227851655274e-06, "loss": 1.6876, "step": 5163 }, { "epoch": 0.40470219435736676, "grad_norm": 2.979511022567749, "learning_rate": 9.758133114349894e-06, "loss": 1.7015, "step": 5164 }, { "epoch": 0.4047805642633229, "grad_norm": 2.2031056880950928, "learning_rate": 9.75803835894702e-06, "loss": 1.2278, "step": 5165 }, { "epoch": 0.404858934169279, "grad_norm": 2.408111333847046, "learning_rate": 9.757943585447011e-06, "loss": 1.4324, "step": 5166 }, { "epoch": 0.4049373040752351, "grad_norm": 2.816220998764038, "learning_rate": 9.757848793850232e-06, "loss": 1.6252, "step": 5167 }, { "epoch": 0.4050156739811912, "grad_norm": 3.418802261352539, "learning_rate": 9.757753984157037e-06, "loss": 1.8799, "step": 5168 }, { "epoch": 0.4050940438871473, "grad_norm": 3.9070897102355957, "learning_rate": 9.757659156367789e-06, "loss": 1.7035, "step": 5169 }, { "epoch": 0.4051724137931034, "grad_norm": 3.0816097259521484, "learning_rate": 9.75756431048285e-06, "loss": 1.4355, "step": 5170 }, { "epoch": 0.40525078369905954, "grad_norm": 7.570591449737549, "learning_rate": 9.75746944650258e-06, "loss": 1.0146, "step": 5171 }, { "epoch": 0.40532915360501565, "grad_norm": 2.302841901779175, "learning_rate": 9.757374564427341e-06, "loss": 1.2594, "step": 5172 }, { "epoch": 0.4054075235109718, "grad_norm": 4.656459808349609, "learning_rate": 9.757279664257493e-06, "loss": 1.8411, "step": 5173 }, { "epoch": 0.4054858934169279, "grad_norm": 3.2585244178771973, "learning_rate": 9.757184745993396e-06, "loss": 1.3641, "step": 5174 }, { "epoch": 0.40556426332288403, "grad_norm": 2.4378604888916016, "learning_rate": 9.757089809635411e-06, "loss": 1.5681, "step": 5175 }, { "epoch": 0.40564263322884014, "grad_norm": 11.248680114746094, "learning_rate": 9.756994855183898e-06, "loss": 0.9868, "step": 5176 }, { "epoch": 0.40572100313479625, "grad_norm": 2.0079915523529053, "learning_rate": 9.756899882639224e-06, "loss": 1.3572, "step": 5177 }, { "epoch": 0.40579937304075236, "grad_norm": 2.6928114891052246, "learning_rate": 9.756804892001744e-06, "loss": 1.2998, "step": 5178 }, { "epoch": 0.4058777429467085, "grad_norm": 4.809455871582031, "learning_rate": 9.756709883271821e-06, "loss": 1.7251, "step": 5179 }, { "epoch": 0.4059561128526646, "grad_norm": 6.887524127960205, "learning_rate": 9.756614856449818e-06, "loss": 1.6154, "step": 5180 }, { "epoch": 0.4060344827586207, "grad_norm": 7.066320419311523, "learning_rate": 9.756519811536094e-06, "loss": 1.1006, "step": 5181 }, { "epoch": 0.4061128526645768, "grad_norm": 4.6763529777526855, "learning_rate": 9.756424748531012e-06, "loss": 1.291, "step": 5182 }, { "epoch": 0.4061912225705329, "grad_norm": 2.265231132507324, "learning_rate": 9.756329667434935e-06, "loss": 1.5754, "step": 5183 }, { "epoch": 0.406269592476489, "grad_norm": 2.593947410583496, "learning_rate": 9.75623456824822e-06, "loss": 1.1601, "step": 5184 }, { "epoch": 0.40634796238244514, "grad_norm": 2.1850006580352783, "learning_rate": 9.756139450971234e-06, "loss": 1.3822, "step": 5185 }, { "epoch": 0.40642633228840125, "grad_norm": 5.081132411956787, "learning_rate": 9.756044315604335e-06, "loss": 1.2793, "step": 5186 }, { "epoch": 0.40650470219435736, "grad_norm": 2.4550273418426514, "learning_rate": 9.755949162147888e-06, "loss": 1.2231, "step": 5187 }, { "epoch": 0.40658307210031347, "grad_norm": 2.0595500469207764, "learning_rate": 9.75585399060225e-06, "loss": 1.0598, "step": 5188 }, { "epoch": 0.4066614420062696, "grad_norm": 2.072937250137329, "learning_rate": 9.75575880096779e-06, "loss": 1.4871, "step": 5189 }, { "epoch": 0.4067398119122257, "grad_norm": 6.970439434051514, "learning_rate": 9.755663593244864e-06, "loss": 1.4736, "step": 5190 }, { "epoch": 0.4068181818181818, "grad_norm": 2.4111597537994385, "learning_rate": 9.755568367433837e-06, "loss": 1.1159, "step": 5191 }, { "epoch": 0.4068965517241379, "grad_norm": 2.4171950817108154, "learning_rate": 9.75547312353507e-06, "loss": 1.0881, "step": 5192 }, { "epoch": 0.406974921630094, "grad_norm": 2.764502763748169, "learning_rate": 9.755377861548927e-06, "loss": 1.6951, "step": 5193 }, { "epoch": 0.40705329153605013, "grad_norm": 2.1346309185028076, "learning_rate": 9.755282581475769e-06, "loss": 1.4809, "step": 5194 }, { "epoch": 0.40713166144200624, "grad_norm": 2.1992366313934326, "learning_rate": 9.755187283315958e-06, "loss": 1.4746, "step": 5195 }, { "epoch": 0.4072100313479624, "grad_norm": 16.52403450012207, "learning_rate": 9.755091967069857e-06, "loss": 1.9001, "step": 5196 }, { "epoch": 0.4072884012539185, "grad_norm": 2.6366817951202393, "learning_rate": 9.754996632737828e-06, "loss": 1.5382, "step": 5197 }, { "epoch": 0.40736677115987463, "grad_norm": 1.910776138305664, "learning_rate": 9.754901280320234e-06, "loss": 1.1495, "step": 5198 }, { "epoch": 0.40744514106583074, "grad_norm": 4.2737321853637695, "learning_rate": 9.75480590981744e-06, "loss": 0.8272, "step": 5199 }, { "epoch": 0.40752351097178685, "grad_norm": 4.568659782409668, "learning_rate": 9.754710521229804e-06, "loss": 1.5084, "step": 5200 }, { "epoch": 0.40760188087774296, "grad_norm": 2.6877152919769287, "learning_rate": 9.754615114557693e-06, "loss": 1.7457, "step": 5201 }, { "epoch": 0.40768025078369907, "grad_norm": 2.269656181335449, "learning_rate": 9.754519689801467e-06, "loss": 1.2854, "step": 5202 }, { "epoch": 0.4077586206896552, "grad_norm": 2.823310613632202, "learning_rate": 9.75442424696149e-06, "loss": 1.301, "step": 5203 }, { "epoch": 0.4078369905956113, "grad_norm": 3.9686973094940186, "learning_rate": 9.754328786038126e-06, "loss": 1.9577, "step": 5204 }, { "epoch": 0.4079153605015674, "grad_norm": 4.727001667022705, "learning_rate": 9.754233307031738e-06, "loss": 2.114, "step": 5205 }, { "epoch": 0.4079937304075235, "grad_norm": 3.5358211994171143, "learning_rate": 9.754137809942687e-06, "loss": 1.4228, "step": 5206 }, { "epoch": 0.4080721003134796, "grad_norm": 2.892258405685425, "learning_rate": 9.754042294771338e-06, "loss": 1.5121, "step": 5207 }, { "epoch": 0.40815047021943573, "grad_norm": 2.378575563430786, "learning_rate": 9.753946761518053e-06, "loss": 0.9698, "step": 5208 }, { "epoch": 0.40822884012539185, "grad_norm": 2.8288893699645996, "learning_rate": 9.753851210183199e-06, "loss": 1.761, "step": 5209 }, { "epoch": 0.40830721003134796, "grad_norm": 4.155208587646484, "learning_rate": 9.753755640767135e-06, "loss": 1.4036, "step": 5210 }, { "epoch": 0.40838557993730407, "grad_norm": 2.360625982284546, "learning_rate": 9.753660053270226e-06, "loss": 1.3861, "step": 5211 }, { "epoch": 0.4084639498432602, "grad_norm": 3.941662073135376, "learning_rate": 9.753564447692838e-06, "loss": 1.2098, "step": 5212 }, { "epoch": 0.4085423197492163, "grad_norm": 3.247288703918457, "learning_rate": 9.75346882403533e-06, "loss": 1.3989, "step": 5213 }, { "epoch": 0.4086206896551724, "grad_norm": 8.045462608337402, "learning_rate": 9.753373182298068e-06, "loss": 2.057, "step": 5214 }, { "epoch": 0.4086990595611285, "grad_norm": 2.5425467491149902, "learning_rate": 9.753277522481417e-06, "loss": 1.5754, "step": 5215 }, { "epoch": 0.4087774294670846, "grad_norm": 2.569671630859375, "learning_rate": 9.75318184458574e-06, "loss": 1.4622, "step": 5216 }, { "epoch": 0.40885579937304073, "grad_norm": 3.350767135620117, "learning_rate": 9.753086148611399e-06, "loss": 1.3755, "step": 5217 }, { "epoch": 0.40893416927899684, "grad_norm": 4.469066143035889, "learning_rate": 9.752990434558762e-06, "loss": 1.518, "step": 5218 }, { "epoch": 0.409012539184953, "grad_norm": 3.0174734592437744, "learning_rate": 9.752894702428189e-06, "loss": 1.4461, "step": 5219 }, { "epoch": 0.4090909090909091, "grad_norm": 4.3691277503967285, "learning_rate": 9.752798952220047e-06, "loss": 1.4917, "step": 5220 }, { "epoch": 0.4091692789968652, "grad_norm": 6.329367637634277, "learning_rate": 9.752703183934698e-06, "loss": 2.0962, "step": 5221 }, { "epoch": 0.40924764890282134, "grad_norm": 2.9990572929382324, "learning_rate": 9.752607397572506e-06, "loss": 1.808, "step": 5222 }, { "epoch": 0.40932601880877745, "grad_norm": 2.265533924102783, "learning_rate": 9.752511593133837e-06, "loss": 1.3697, "step": 5223 }, { "epoch": 0.40940438871473356, "grad_norm": 1.5848485231399536, "learning_rate": 9.752415770619057e-06, "loss": 1.4897, "step": 5224 }, { "epoch": 0.40948275862068967, "grad_norm": 3.3154149055480957, "learning_rate": 9.752319930028527e-06, "loss": 1.6085, "step": 5225 }, { "epoch": 0.4095611285266458, "grad_norm": 3.0264904499053955, "learning_rate": 9.752224071362614e-06, "loss": 1.4506, "step": 5226 }, { "epoch": 0.4096394984326019, "grad_norm": 3.504945993423462, "learning_rate": 9.75212819462168e-06, "loss": 1.375, "step": 5227 }, { "epoch": 0.409717868338558, "grad_norm": 2.3760807514190674, "learning_rate": 9.752032299806092e-06, "loss": 1.5475, "step": 5228 }, { "epoch": 0.4097962382445141, "grad_norm": 2.127505302429199, "learning_rate": 9.751936386916215e-06, "loss": 1.4533, "step": 5229 }, { "epoch": 0.4098746081504702, "grad_norm": 4.349267482757568, "learning_rate": 9.751840455952411e-06, "loss": 2.0931, "step": 5230 }, { "epoch": 0.40995297805642633, "grad_norm": 5.0799336433410645, "learning_rate": 9.751744506915047e-06, "loss": 1.0407, "step": 5231 }, { "epoch": 0.41003134796238244, "grad_norm": 2.8378283977508545, "learning_rate": 9.751648539804488e-06, "loss": 1.1884, "step": 5232 }, { "epoch": 0.41010971786833855, "grad_norm": 3.6163227558135986, "learning_rate": 9.751552554621097e-06, "loss": 1.5749, "step": 5233 }, { "epoch": 0.41018808777429466, "grad_norm": 2.7975382804870605, "learning_rate": 9.751456551365244e-06, "loss": 1.3327, "step": 5234 }, { "epoch": 0.4102664576802508, "grad_norm": 3.151174306869507, "learning_rate": 9.751360530037289e-06, "loss": 1.3433, "step": 5235 }, { "epoch": 0.4103448275862069, "grad_norm": 2.848844289779663, "learning_rate": 9.751264490637598e-06, "loss": 1.1421, "step": 5236 }, { "epoch": 0.410423197492163, "grad_norm": 3.5001280307769775, "learning_rate": 9.751168433166539e-06, "loss": 1.6219, "step": 5237 }, { "epoch": 0.4105015673981191, "grad_norm": 2.0429294109344482, "learning_rate": 9.751072357624473e-06, "loss": 1.3429, "step": 5238 }, { "epoch": 0.4105799373040752, "grad_norm": 3.1065731048583984, "learning_rate": 9.75097626401177e-06, "loss": 1.2616, "step": 5239 }, { "epoch": 0.4106583072100313, "grad_norm": 2.1801042556762695, "learning_rate": 9.750880152328793e-06, "loss": 1.6335, "step": 5240 }, { "epoch": 0.41073667711598744, "grad_norm": 3.261608839035034, "learning_rate": 9.750784022575908e-06, "loss": 1.415, "step": 5241 }, { "epoch": 0.41081504702194355, "grad_norm": 3.31945538520813, "learning_rate": 9.750687874753482e-06, "loss": 1.4876, "step": 5242 }, { "epoch": 0.4108934169278997, "grad_norm": 4.325932502746582, "learning_rate": 9.750591708861878e-06, "loss": 1.7656, "step": 5243 }, { "epoch": 0.4109717868338558, "grad_norm": 3.088284492492676, "learning_rate": 9.750495524901464e-06, "loss": 1.5813, "step": 5244 }, { "epoch": 0.41105015673981193, "grad_norm": 3.288649082183838, "learning_rate": 9.750399322872605e-06, "loss": 0.8209, "step": 5245 }, { "epoch": 0.41112852664576804, "grad_norm": 3.745201349258423, "learning_rate": 9.750303102775666e-06, "loss": 1.5287, "step": 5246 }, { "epoch": 0.41120689655172415, "grad_norm": 3.2701358795166016, "learning_rate": 9.750206864611015e-06, "loss": 1.4937, "step": 5247 }, { "epoch": 0.41128526645768027, "grad_norm": 5.758239269256592, "learning_rate": 9.750110608379016e-06, "loss": 1.1459, "step": 5248 }, { "epoch": 0.4113636363636364, "grad_norm": 3.3014755249023438, "learning_rate": 9.750014334080037e-06, "loss": 1.0117, "step": 5249 }, { "epoch": 0.4114420062695925, "grad_norm": 2.186162233352661, "learning_rate": 9.749918041714442e-06, "loss": 1.3882, "step": 5250 }, { "epoch": 0.4115203761755486, "grad_norm": 2.977389335632324, "learning_rate": 9.7498217312826e-06, "loss": 1.3256, "step": 5251 }, { "epoch": 0.4115987460815047, "grad_norm": 2.3936896324157715, "learning_rate": 9.749725402784876e-06, "loss": 1.6762, "step": 5252 }, { "epoch": 0.4116771159874608, "grad_norm": 3.031151294708252, "learning_rate": 9.749629056221634e-06, "loss": 1.2237, "step": 5253 }, { "epoch": 0.41175548589341693, "grad_norm": 7.724858283996582, "learning_rate": 9.749532691593245e-06, "loss": 1.3921, "step": 5254 }, { "epoch": 0.41183385579937304, "grad_norm": 2.578043222427368, "learning_rate": 9.749436308900072e-06, "loss": 1.4574, "step": 5255 }, { "epoch": 0.41191222570532915, "grad_norm": 2.481633186340332, "learning_rate": 9.749339908142481e-06, "loss": 1.3644, "step": 5256 }, { "epoch": 0.41199059561128526, "grad_norm": 4.110805511474609, "learning_rate": 9.749243489320843e-06, "loss": 1.7646, "step": 5257 }, { "epoch": 0.41206896551724137, "grad_norm": 2.470459461212158, "learning_rate": 9.749147052435522e-06, "loss": 1.2737, "step": 5258 }, { "epoch": 0.4121473354231975, "grad_norm": 2.1694159507751465, "learning_rate": 9.749050597486885e-06, "loss": 1.2651, "step": 5259 }, { "epoch": 0.4122257053291536, "grad_norm": 9.681133270263672, "learning_rate": 9.748954124475297e-06, "loss": 1.7655, "step": 5260 }, { "epoch": 0.4123040752351097, "grad_norm": 3.5091376304626465, "learning_rate": 9.748857633401128e-06, "loss": 1.3545, "step": 5261 }, { "epoch": 0.4123824451410658, "grad_norm": 2.234257698059082, "learning_rate": 9.748761124264743e-06, "loss": 1.3223, "step": 5262 }, { "epoch": 0.4124608150470219, "grad_norm": 6.651845455169678, "learning_rate": 9.748664597066511e-06, "loss": 1.4768, "step": 5263 }, { "epoch": 0.41253918495297803, "grad_norm": 3.1763484477996826, "learning_rate": 9.748568051806798e-06, "loss": 1.375, "step": 5264 }, { "epoch": 0.41261755485893414, "grad_norm": 3.3427340984344482, "learning_rate": 9.748471488485968e-06, "loss": 1.3832, "step": 5265 }, { "epoch": 0.4126959247648903, "grad_norm": 2.5489444732666016, "learning_rate": 9.748374907104395e-06, "loss": 1.3661, "step": 5266 }, { "epoch": 0.4127742946708464, "grad_norm": 4.038214206695557, "learning_rate": 9.748278307662442e-06, "loss": 1.4792, "step": 5267 }, { "epoch": 0.41285266457680253, "grad_norm": 7.656335830688477, "learning_rate": 9.748181690160476e-06, "loss": 1.7986, "step": 5268 }, { "epoch": 0.41293103448275864, "grad_norm": 2.6175003051757812, "learning_rate": 9.748085054598868e-06, "loss": 1.5238, "step": 5269 }, { "epoch": 0.41300940438871475, "grad_norm": 3.101376533508301, "learning_rate": 9.747988400977982e-06, "loss": 1.6438, "step": 5270 }, { "epoch": 0.41308777429467086, "grad_norm": 9.707697868347168, "learning_rate": 9.747891729298186e-06, "loss": 2.0058, "step": 5271 }, { "epoch": 0.413166144200627, "grad_norm": 2.0198283195495605, "learning_rate": 9.747795039559849e-06, "loss": 1.449, "step": 5272 }, { "epoch": 0.4132445141065831, "grad_norm": 10.61634349822998, "learning_rate": 9.747698331763338e-06, "loss": 1.3568, "step": 5273 }, { "epoch": 0.4133228840125392, "grad_norm": 2.1346914768218994, "learning_rate": 9.747601605909022e-06, "loss": 1.3233, "step": 5274 }, { "epoch": 0.4134012539184953, "grad_norm": 2.225874662399292, "learning_rate": 9.747504861997269e-06, "loss": 1.4032, "step": 5275 }, { "epoch": 0.4134796238244514, "grad_norm": 3.3585622310638428, "learning_rate": 9.747408100028443e-06, "loss": 1.1732, "step": 5276 }, { "epoch": 0.4135579937304075, "grad_norm": 2.199063301086426, "learning_rate": 9.747311320002916e-06, "loss": 1.3364, "step": 5277 }, { "epoch": 0.41363636363636364, "grad_norm": 4.2166643142700195, "learning_rate": 9.747214521921055e-06, "loss": 1.4352, "step": 5278 }, { "epoch": 0.41371473354231975, "grad_norm": 5.147922039031982, "learning_rate": 9.747117705783229e-06, "loss": 1.4653, "step": 5279 }, { "epoch": 0.41379310344827586, "grad_norm": 14.584259033203125, "learning_rate": 9.747020871589804e-06, "loss": 1.727, "step": 5280 }, { "epoch": 0.41387147335423197, "grad_norm": 5.015491008758545, "learning_rate": 9.746924019341152e-06, "loss": 1.6884, "step": 5281 }, { "epoch": 0.4139498432601881, "grad_norm": 4.171465873718262, "learning_rate": 9.74682714903764e-06, "loss": 1.2391, "step": 5282 }, { "epoch": 0.4140282131661442, "grad_norm": 3.0518908500671387, "learning_rate": 9.746730260679632e-06, "loss": 0.8633, "step": 5283 }, { "epoch": 0.4141065830721003, "grad_norm": 2.6263203620910645, "learning_rate": 9.746633354267505e-06, "loss": 1.2554, "step": 5284 }, { "epoch": 0.4141849529780564, "grad_norm": 2.3638875484466553, "learning_rate": 9.746536429801618e-06, "loss": 1.3046, "step": 5285 }, { "epoch": 0.4142633228840125, "grad_norm": 2.302841901779175, "learning_rate": 9.746439487282348e-06, "loss": 1.4648, "step": 5286 }, { "epoch": 0.41434169278996863, "grad_norm": 2.3051719665527344, "learning_rate": 9.746342526710057e-06, "loss": 1.4965, "step": 5287 }, { "epoch": 0.41442006269592474, "grad_norm": 2.5686707496643066, "learning_rate": 9.74624554808512e-06, "loss": 1.0774, "step": 5288 }, { "epoch": 0.41449843260188085, "grad_norm": 2.509566307067871, "learning_rate": 9.7461485514079e-06, "loss": 1.2923, "step": 5289 }, { "epoch": 0.414576802507837, "grad_norm": 2.207310438156128, "learning_rate": 9.74605153667877e-06, "loss": 1.5037, "step": 5290 }, { "epoch": 0.4146551724137931, "grad_norm": 2.5029807090759277, "learning_rate": 9.745954503898098e-06, "loss": 1.3325, "step": 5291 }, { "epoch": 0.41473354231974924, "grad_norm": 2.8969616889953613, "learning_rate": 9.745857453066251e-06, "loss": 1.2882, "step": 5292 }, { "epoch": 0.41481191222570535, "grad_norm": 3.532362937927246, "learning_rate": 9.745760384183602e-06, "loss": 1.617, "step": 5293 }, { "epoch": 0.41489028213166146, "grad_norm": 3.392996311187744, "learning_rate": 9.745663297250518e-06, "loss": 1.4961, "step": 5294 }, { "epoch": 0.41496865203761757, "grad_norm": 6.100953102111816, "learning_rate": 9.745566192267367e-06, "loss": 1.1341, "step": 5295 }, { "epoch": 0.4150470219435737, "grad_norm": 2.4026317596435547, "learning_rate": 9.745469069234522e-06, "loss": 1.3997, "step": 5296 }, { "epoch": 0.4151253918495298, "grad_norm": 4.917450428009033, "learning_rate": 9.745371928152348e-06, "loss": 1.4329, "step": 5297 }, { "epoch": 0.4152037617554859, "grad_norm": 3.086883068084717, "learning_rate": 9.745274769021216e-06, "loss": 1.5159, "step": 5298 }, { "epoch": 0.415282131661442, "grad_norm": 4.61680269241333, "learning_rate": 9.745177591841497e-06, "loss": 1.4648, "step": 5299 }, { "epoch": 0.4153605015673981, "grad_norm": 2.9919066429138184, "learning_rate": 9.74508039661356e-06, "loss": 1.3751, "step": 5300 }, { "epoch": 0.41543887147335423, "grad_norm": 2.74206280708313, "learning_rate": 9.744983183337773e-06, "loss": 1.3023, "step": 5301 }, { "epoch": 0.41551724137931034, "grad_norm": 3.6540415287017822, "learning_rate": 9.744885952014508e-06, "loss": 1.425, "step": 5302 }, { "epoch": 0.41559561128526645, "grad_norm": 2.0935981273651123, "learning_rate": 9.744788702644134e-06, "loss": 1.212, "step": 5303 }, { "epoch": 0.41567398119122256, "grad_norm": 2.250833034515381, "learning_rate": 9.744691435227023e-06, "loss": 0.8861, "step": 5304 }, { "epoch": 0.4157523510971787, "grad_norm": 2.6505913734436035, "learning_rate": 9.74459414976354e-06, "loss": 1.6189, "step": 5305 }, { "epoch": 0.4158307210031348, "grad_norm": 3.181563377380371, "learning_rate": 9.744496846254058e-06, "loss": 1.2687, "step": 5306 }, { "epoch": 0.4159090909090909, "grad_norm": 3.2023327350616455, "learning_rate": 9.744399524698949e-06, "loss": 1.4468, "step": 5307 }, { "epoch": 0.415987460815047, "grad_norm": 2.5395925045013428, "learning_rate": 9.74430218509858e-06, "loss": 1.4174, "step": 5308 }, { "epoch": 0.4160658307210031, "grad_norm": 2.2907278537750244, "learning_rate": 9.744204827453322e-06, "loss": 1.3678, "step": 5309 }, { "epoch": 0.4161442006269592, "grad_norm": 5.327610015869141, "learning_rate": 9.744107451763545e-06, "loss": 0.8441, "step": 5310 }, { "epoch": 0.41622257053291534, "grad_norm": 7.822359561920166, "learning_rate": 9.74401005802962e-06, "loss": 1.4997, "step": 5311 }, { "epoch": 0.41630094043887145, "grad_norm": 2.9755465984344482, "learning_rate": 9.743912646251918e-06, "loss": 1.4883, "step": 5312 }, { "epoch": 0.4163793103448276, "grad_norm": 2.779764175415039, "learning_rate": 9.74381521643081e-06, "loss": 1.2321, "step": 5313 }, { "epoch": 0.4164576802507837, "grad_norm": 2.4541592597961426, "learning_rate": 9.743717768566666e-06, "loss": 1.6108, "step": 5314 }, { "epoch": 0.41653605015673983, "grad_norm": 2.9217491149902344, "learning_rate": 9.743620302659853e-06, "loss": 1.3147, "step": 5315 }, { "epoch": 0.41661442006269594, "grad_norm": 5.35811710357666, "learning_rate": 9.743522818710747e-06, "loss": 1.3299, "step": 5316 }, { "epoch": 0.41669278996865206, "grad_norm": 6.145333766937256, "learning_rate": 9.743425316719716e-06, "loss": 1.6015, "step": 5317 }, { "epoch": 0.41677115987460817, "grad_norm": 3.124795436859131, "learning_rate": 9.743327796687131e-06, "loss": 1.7339, "step": 5318 }, { "epoch": 0.4168495297805643, "grad_norm": 3.1053922176361084, "learning_rate": 9.743230258613364e-06, "loss": 1.2133, "step": 5319 }, { "epoch": 0.4169278996865204, "grad_norm": 2.6621148586273193, "learning_rate": 9.743132702498785e-06, "loss": 1.1455, "step": 5320 }, { "epoch": 0.4170062695924765, "grad_norm": 4.695505619049072, "learning_rate": 9.743035128343766e-06, "loss": 1.1163, "step": 5321 }, { "epoch": 0.4170846394984326, "grad_norm": 3.5604825019836426, "learning_rate": 9.742937536148678e-06, "loss": 1.3367, "step": 5322 }, { "epoch": 0.4171630094043887, "grad_norm": 3.4038195610046387, "learning_rate": 9.742839925913891e-06, "loss": 1.2674, "step": 5323 }, { "epoch": 0.41724137931034483, "grad_norm": 3.5188374519348145, "learning_rate": 9.742742297639777e-06, "loss": 1.2287, "step": 5324 }, { "epoch": 0.41731974921630094, "grad_norm": 2.2473483085632324, "learning_rate": 9.742644651326707e-06, "loss": 1.4819, "step": 5325 }, { "epoch": 0.41739811912225705, "grad_norm": 3.4277420043945312, "learning_rate": 9.742546986975053e-06, "loss": 1.1138, "step": 5326 }, { "epoch": 0.41747648902821316, "grad_norm": 3.9024734497070312, "learning_rate": 9.742449304585186e-06, "loss": 1.337, "step": 5327 }, { "epoch": 0.41755485893416927, "grad_norm": 3.496291160583496, "learning_rate": 9.742351604157479e-06, "loss": 1.7089, "step": 5328 }, { "epoch": 0.4176332288401254, "grad_norm": 2.1543986797332764, "learning_rate": 9.742253885692301e-06, "loss": 1.4017, "step": 5329 }, { "epoch": 0.4177115987460815, "grad_norm": 4.996683120727539, "learning_rate": 9.742156149190025e-06, "loss": 1.3055, "step": 5330 }, { "epoch": 0.4177899686520376, "grad_norm": 2.44734787940979, "learning_rate": 9.742058394651024e-06, "loss": 1.2309, "step": 5331 }, { "epoch": 0.4178683385579937, "grad_norm": 6.721362590789795, "learning_rate": 9.741960622075667e-06, "loss": 1.3708, "step": 5332 }, { "epoch": 0.4179467084639498, "grad_norm": 3.286649227142334, "learning_rate": 9.741862831464328e-06, "loss": 1.1759, "step": 5333 }, { "epoch": 0.41802507836990593, "grad_norm": 3.2115697860717773, "learning_rate": 9.741765022817378e-06, "loss": 1.2595, "step": 5334 }, { "epoch": 0.41810344827586204, "grad_norm": 5.5924530029296875, "learning_rate": 9.74166719613519e-06, "loss": 1.9077, "step": 5335 }, { "epoch": 0.41818181818181815, "grad_norm": 3.202522039413452, "learning_rate": 9.741569351418136e-06, "loss": 1.3879, "step": 5336 }, { "epoch": 0.4182601880877743, "grad_norm": 4.1609787940979, "learning_rate": 9.741471488666586e-06, "loss": 1.6272, "step": 5337 }, { "epoch": 0.41833855799373043, "grad_norm": 1.9922133684158325, "learning_rate": 9.741373607880915e-06, "loss": 1.2981, "step": 5338 }, { "epoch": 0.41841692789968654, "grad_norm": 2.252124309539795, "learning_rate": 9.741275709061493e-06, "loss": 1.4605, "step": 5339 }, { "epoch": 0.41849529780564265, "grad_norm": 3.3151726722717285, "learning_rate": 9.741177792208695e-06, "loss": 1.197, "step": 5340 }, { "epoch": 0.41857366771159876, "grad_norm": 3.025068521499634, "learning_rate": 9.74107985732289e-06, "loss": 1.3816, "step": 5341 }, { "epoch": 0.4186520376175549, "grad_norm": 1.9649826288223267, "learning_rate": 9.740981904404454e-06, "loss": 1.3818, "step": 5342 }, { "epoch": 0.418730407523511, "grad_norm": 3.0985159873962402, "learning_rate": 9.740883933453757e-06, "loss": 1.2294, "step": 5343 }, { "epoch": 0.4188087774294671, "grad_norm": 2.5317234992980957, "learning_rate": 9.740785944471175e-06, "loss": 1.0804, "step": 5344 }, { "epoch": 0.4188871473354232, "grad_norm": 3.5214438438415527, "learning_rate": 9.740687937457075e-06, "loss": 1.3627, "step": 5345 }, { "epoch": 0.4189655172413793, "grad_norm": 3.28454852104187, "learning_rate": 9.740589912411835e-06, "loss": 1.6298, "step": 5346 }, { "epoch": 0.4190438871473354, "grad_norm": 1.862979769706726, "learning_rate": 9.740491869335826e-06, "loss": 1.2359, "step": 5347 }, { "epoch": 0.41912225705329154, "grad_norm": 3.559297561645508, "learning_rate": 9.74039380822942e-06, "loss": 1.494, "step": 5348 }, { "epoch": 0.41920062695924765, "grad_norm": 6.046116352081299, "learning_rate": 9.74029572909299e-06, "loss": 2.1238, "step": 5349 }, { "epoch": 0.41927899686520376, "grad_norm": 5.39988899230957, "learning_rate": 9.740197631926911e-06, "loss": 1.1694, "step": 5350 }, { "epoch": 0.41935736677115987, "grad_norm": 3.279301166534424, "learning_rate": 9.740099516731554e-06, "loss": 1.3778, "step": 5351 }, { "epoch": 0.419435736677116, "grad_norm": 2.6501336097717285, "learning_rate": 9.740001383507293e-06, "loss": 1.3865, "step": 5352 }, { "epoch": 0.4195141065830721, "grad_norm": 2.946350336074829, "learning_rate": 9.739903232254504e-06, "loss": 0.8764, "step": 5353 }, { "epoch": 0.4195924764890282, "grad_norm": 2.143954038619995, "learning_rate": 9.739805062973555e-06, "loss": 1.5624, "step": 5354 }, { "epoch": 0.4196708463949843, "grad_norm": 2.549201250076294, "learning_rate": 9.739706875664823e-06, "loss": 1.6326, "step": 5355 }, { "epoch": 0.4197492163009404, "grad_norm": 9.8991117477417, "learning_rate": 9.73960867032868e-06, "loss": 1.4966, "step": 5356 }, { "epoch": 0.41982758620689653, "grad_norm": 2.0420875549316406, "learning_rate": 9.739510446965501e-06, "loss": 1.6845, "step": 5357 }, { "epoch": 0.41990595611285264, "grad_norm": 9.02991771697998, "learning_rate": 9.739412205575658e-06, "loss": 1.364, "step": 5358 }, { "epoch": 0.41998432601880875, "grad_norm": 3.080770254135132, "learning_rate": 9.739313946159525e-06, "loss": 1.5712, "step": 5359 }, { "epoch": 0.4200626959247649, "grad_norm": 6.034005641937256, "learning_rate": 9.739215668717476e-06, "loss": 0.5756, "step": 5360 }, { "epoch": 0.420141065830721, "grad_norm": 4.045856475830078, "learning_rate": 9.739117373249886e-06, "loss": 1.2474, "step": 5361 }, { "epoch": 0.42021943573667714, "grad_norm": 2.0070114135742188, "learning_rate": 9.739019059757127e-06, "loss": 1.5824, "step": 5362 }, { "epoch": 0.42029780564263325, "grad_norm": 3.0702459812164307, "learning_rate": 9.738920728239574e-06, "loss": 1.8274, "step": 5363 }, { "epoch": 0.42037617554858936, "grad_norm": 3.0598175525665283, "learning_rate": 9.738822378697598e-06, "loss": 1.2972, "step": 5364 }, { "epoch": 0.42045454545454547, "grad_norm": 3.27669358253479, "learning_rate": 9.738724011131578e-06, "loss": 1.4302, "step": 5365 }, { "epoch": 0.4205329153605016, "grad_norm": 3.270824670791626, "learning_rate": 9.738625625541887e-06, "loss": 1.2415, "step": 5366 }, { "epoch": 0.4206112852664577, "grad_norm": 2.459228515625, "learning_rate": 9.738527221928896e-06, "loss": 1.4434, "step": 5367 }, { "epoch": 0.4206896551724138, "grad_norm": 3.1083602905273438, "learning_rate": 9.738428800292981e-06, "loss": 1.776, "step": 5368 }, { "epoch": 0.4207680250783699, "grad_norm": 5.145234107971191, "learning_rate": 9.738330360634518e-06, "loss": 1.8276, "step": 5369 }, { "epoch": 0.420846394984326, "grad_norm": 2.9901013374328613, "learning_rate": 9.73823190295388e-06, "loss": 1.5547, "step": 5370 }, { "epoch": 0.42092476489028213, "grad_norm": 2.213136911392212, "learning_rate": 9.738133427251441e-06, "loss": 1.0866, "step": 5371 }, { "epoch": 0.42100313479623824, "grad_norm": 2.599656581878662, "learning_rate": 9.738034933527576e-06, "loss": 1.1296, "step": 5372 }, { "epoch": 0.42108150470219435, "grad_norm": 2.7879698276519775, "learning_rate": 9.73793642178266e-06, "loss": 1.2489, "step": 5373 }, { "epoch": 0.42115987460815046, "grad_norm": 2.4620468616485596, "learning_rate": 9.737837892017068e-06, "loss": 1.5112, "step": 5374 }, { "epoch": 0.4212382445141066, "grad_norm": 3.0415613651275635, "learning_rate": 9.737739344231173e-06, "loss": 1.5913, "step": 5375 }, { "epoch": 0.4213166144200627, "grad_norm": 3.129023551940918, "learning_rate": 9.737640778425353e-06, "loss": 1.3797, "step": 5376 }, { "epoch": 0.4213949843260188, "grad_norm": 3.1628479957580566, "learning_rate": 9.737542194599978e-06, "loss": 1.3717, "step": 5377 }, { "epoch": 0.4214733542319749, "grad_norm": 3.749225616455078, "learning_rate": 9.737443592755427e-06, "loss": 1.4331, "step": 5378 }, { "epoch": 0.421551724137931, "grad_norm": 4.754988670349121, "learning_rate": 9.737344972892075e-06, "loss": 1.9102, "step": 5379 }, { "epoch": 0.4216300940438871, "grad_norm": 3.6903605461120605, "learning_rate": 9.737246335010295e-06, "loss": 1.2783, "step": 5380 }, { "epoch": 0.42170846394984324, "grad_norm": 2.491596221923828, "learning_rate": 9.737147679110463e-06, "loss": 1.5391, "step": 5381 }, { "epoch": 0.42178683385579935, "grad_norm": 2.7060585021972656, "learning_rate": 9.737049005192955e-06, "loss": 1.2554, "step": 5382 }, { "epoch": 0.42186520376175546, "grad_norm": 2.4199793338775635, "learning_rate": 9.736950313258144e-06, "loss": 1.6707, "step": 5383 }, { "epoch": 0.4219435736677116, "grad_norm": 2.72147274017334, "learning_rate": 9.73685160330641e-06, "loss": 1.4555, "step": 5384 }, { "epoch": 0.42202194357366773, "grad_norm": 3.010213613510132, "learning_rate": 9.736752875338124e-06, "loss": 1.3396, "step": 5385 }, { "epoch": 0.42210031347962385, "grad_norm": 3.329561710357666, "learning_rate": 9.736654129353663e-06, "loss": 1.3902, "step": 5386 }, { "epoch": 0.42217868338557996, "grad_norm": 3.409146547317505, "learning_rate": 9.736555365353402e-06, "loss": 1.238, "step": 5387 }, { "epoch": 0.42225705329153607, "grad_norm": 2.3756556510925293, "learning_rate": 9.736456583337717e-06, "loss": 1.238, "step": 5388 }, { "epoch": 0.4223354231974922, "grad_norm": 3.850276231765747, "learning_rate": 9.736357783306984e-06, "loss": 1.2669, "step": 5389 }, { "epoch": 0.4224137931034483, "grad_norm": 3.5062661170959473, "learning_rate": 9.73625896526158e-06, "loss": 1.2812, "step": 5390 }, { "epoch": 0.4224921630094044, "grad_norm": 2.1109585762023926, "learning_rate": 9.73616012920188e-06, "loss": 1.4077, "step": 5391 }, { "epoch": 0.4225705329153605, "grad_norm": 3.567884683609009, "learning_rate": 9.736061275128258e-06, "loss": 1.5098, "step": 5392 }, { "epoch": 0.4226489028213166, "grad_norm": 1.9213532209396362, "learning_rate": 9.735962403041093e-06, "loss": 1.4058, "step": 5393 }, { "epoch": 0.42272727272727273, "grad_norm": 3.1920080184936523, "learning_rate": 9.735863512940757e-06, "loss": 0.7616, "step": 5394 }, { "epoch": 0.42280564263322884, "grad_norm": 3.860090970993042, "learning_rate": 9.73576460482763e-06, "loss": 1.6178, "step": 5395 }, { "epoch": 0.42288401253918495, "grad_norm": 4.285135269165039, "learning_rate": 9.735665678702088e-06, "loss": 1.5329, "step": 5396 }, { "epoch": 0.42296238244514106, "grad_norm": 2.9959328174591064, "learning_rate": 9.735566734564506e-06, "loss": 1.2556, "step": 5397 }, { "epoch": 0.42304075235109717, "grad_norm": 3.39090895652771, "learning_rate": 9.735467772415259e-06, "loss": 1.4072, "step": 5398 }, { "epoch": 0.4231191222570533, "grad_norm": 3.4692764282226562, "learning_rate": 9.735368792254728e-06, "loss": 1.9576, "step": 5399 }, { "epoch": 0.4231974921630094, "grad_norm": 2.8921244144439697, "learning_rate": 9.735269794083281e-06, "loss": 1.7545, "step": 5400 }, { "epoch": 0.4232758620689655, "grad_norm": 2.2414588928222656, "learning_rate": 9.735170777901304e-06, "loss": 1.3351, "step": 5401 }, { "epoch": 0.4233542319749216, "grad_norm": 2.5994722843170166, "learning_rate": 9.735071743709168e-06, "loss": 1.3028, "step": 5402 }, { "epoch": 0.4234326018808777, "grad_norm": 2.220632791519165, "learning_rate": 9.734972691507252e-06, "loss": 1.2859, "step": 5403 }, { "epoch": 0.42351097178683383, "grad_norm": 6.395714282989502, "learning_rate": 9.734873621295933e-06, "loss": 1.1942, "step": 5404 }, { "epoch": 0.42358934169278994, "grad_norm": 2.6481070518493652, "learning_rate": 9.734774533075584e-06, "loss": 1.0991, "step": 5405 }, { "epoch": 0.42366771159874606, "grad_norm": 9.699251174926758, "learning_rate": 9.734675426846585e-06, "loss": 0.6762, "step": 5406 }, { "epoch": 0.4237460815047022, "grad_norm": 2.9473583698272705, "learning_rate": 9.734576302609315e-06, "loss": 1.005, "step": 5407 }, { "epoch": 0.42382445141065833, "grad_norm": 2.328772783279419, "learning_rate": 9.734477160364147e-06, "loss": 1.325, "step": 5408 }, { "epoch": 0.42390282131661444, "grad_norm": 2.4760642051696777, "learning_rate": 9.73437800011146e-06, "loss": 1.523, "step": 5409 }, { "epoch": 0.42398119122257055, "grad_norm": 6.817879676818848, "learning_rate": 9.734278821851631e-06, "loss": 1.6397, "step": 5410 }, { "epoch": 0.42405956112852666, "grad_norm": 3.1301748752593994, "learning_rate": 9.734179625585035e-06, "loss": 1.3043, "step": 5411 }, { "epoch": 0.4241379310344828, "grad_norm": 2.1423983573913574, "learning_rate": 9.734080411312053e-06, "loss": 1.3696, "step": 5412 }, { "epoch": 0.4242163009404389, "grad_norm": 3.3684349060058594, "learning_rate": 9.733981179033061e-06, "loss": 1.7646, "step": 5413 }, { "epoch": 0.424294670846395, "grad_norm": 2.673070192337036, "learning_rate": 9.733881928748435e-06, "loss": 1.3336, "step": 5414 }, { "epoch": 0.4243730407523511, "grad_norm": 3.11464524269104, "learning_rate": 9.733782660458556e-06, "loss": 1.5269, "step": 5415 }, { "epoch": 0.4244514106583072, "grad_norm": 2.6391353607177734, "learning_rate": 9.733683374163796e-06, "loss": 1.044, "step": 5416 }, { "epoch": 0.4245297805642633, "grad_norm": 2.987283945083618, "learning_rate": 9.733584069864537e-06, "loss": 1.412, "step": 5417 }, { "epoch": 0.42460815047021944, "grad_norm": 5.014761924743652, "learning_rate": 9.733484747561158e-06, "loss": 1.5392, "step": 5418 }, { "epoch": 0.42468652037617555, "grad_norm": 2.102168560028076, "learning_rate": 9.73338540725403e-06, "loss": 1.588, "step": 5419 }, { "epoch": 0.42476489028213166, "grad_norm": 2.1715478897094727, "learning_rate": 9.733286048943538e-06, "loss": 1.2693, "step": 5420 }, { "epoch": 0.42484326018808777, "grad_norm": 2.9536590576171875, "learning_rate": 9.733186672630056e-06, "loss": 1.4102, "step": 5421 }, { "epoch": 0.4249216300940439, "grad_norm": 2.902698040008545, "learning_rate": 9.733087278313964e-06, "loss": 1.7689, "step": 5422 }, { "epoch": 0.425, "grad_norm": 1.9063056707382202, "learning_rate": 9.732987865995637e-06, "loss": 1.0524, "step": 5423 }, { "epoch": 0.4250783699059561, "grad_norm": 2.439652442932129, "learning_rate": 9.732888435675457e-06, "loss": 1.2539, "step": 5424 }, { "epoch": 0.4251567398119122, "grad_norm": 2.57588529586792, "learning_rate": 9.7327889873538e-06, "loss": 1.1854, "step": 5425 }, { "epoch": 0.4252351097178683, "grad_norm": 2.5419974327087402, "learning_rate": 9.732689521031044e-06, "loss": 0.9241, "step": 5426 }, { "epoch": 0.42531347962382443, "grad_norm": 3.2313413619995117, "learning_rate": 9.73259003670757e-06, "loss": 1.4467, "step": 5427 }, { "epoch": 0.42539184952978054, "grad_norm": 2.7213926315307617, "learning_rate": 9.732490534383752e-06, "loss": 1.7865, "step": 5428 }, { "epoch": 0.42547021943573665, "grad_norm": 3.9272496700286865, "learning_rate": 9.732391014059972e-06, "loss": 1.1223, "step": 5429 }, { "epoch": 0.42554858934169276, "grad_norm": 2.0539872646331787, "learning_rate": 9.732291475736606e-06, "loss": 1.5636, "step": 5430 }, { "epoch": 0.42562695924764893, "grad_norm": 4.014435291290283, "learning_rate": 9.732191919414035e-06, "loss": 1.6779, "step": 5431 }, { "epoch": 0.42570532915360504, "grad_norm": 2.225858211517334, "learning_rate": 9.732092345092637e-06, "loss": 0.8498, "step": 5432 }, { "epoch": 0.42578369905956115, "grad_norm": 2.7014212608337402, "learning_rate": 9.731992752772789e-06, "loss": 1.4439, "step": 5433 }, { "epoch": 0.42586206896551726, "grad_norm": 1.996753215789795, "learning_rate": 9.731893142454872e-06, "loss": 0.8216, "step": 5434 }, { "epoch": 0.42594043887147337, "grad_norm": 2.6976044178009033, "learning_rate": 9.731793514139263e-06, "loss": 1.7471, "step": 5435 }, { "epoch": 0.4260188087774295, "grad_norm": 5.995304107666016, "learning_rate": 9.731693867826342e-06, "loss": 1.3061, "step": 5436 }, { "epoch": 0.4260971786833856, "grad_norm": 3.1830995082855225, "learning_rate": 9.731594203516488e-06, "loss": 1.3949, "step": 5437 }, { "epoch": 0.4261755485893417, "grad_norm": 2.201725482940674, "learning_rate": 9.731494521210081e-06, "loss": 1.5213, "step": 5438 }, { "epoch": 0.4262539184952978, "grad_norm": 9.19019889831543, "learning_rate": 9.731394820907498e-06, "loss": 1.1761, "step": 5439 }, { "epoch": 0.4263322884012539, "grad_norm": 2.526805877685547, "learning_rate": 9.73129510260912e-06, "loss": 0.9107, "step": 5440 }, { "epoch": 0.42641065830721003, "grad_norm": 3.333784580230713, "learning_rate": 9.731195366315325e-06, "loss": 1.6998, "step": 5441 }, { "epoch": 0.42648902821316614, "grad_norm": 3.2021279335021973, "learning_rate": 9.731095612026494e-06, "loss": 1.4429, "step": 5442 }, { "epoch": 0.42656739811912225, "grad_norm": 2.735560417175293, "learning_rate": 9.730995839743002e-06, "loss": 1.1174, "step": 5443 }, { "epoch": 0.42664576802507836, "grad_norm": 2.6770529747009277, "learning_rate": 9.730896049465234e-06, "loss": 1.2342, "step": 5444 }, { "epoch": 0.4267241379310345, "grad_norm": 2.852292060852051, "learning_rate": 9.73079624119357e-06, "loss": 1.3384, "step": 5445 }, { "epoch": 0.4268025078369906, "grad_norm": 2.2055158615112305, "learning_rate": 9.730696414928385e-06, "loss": 1.5874, "step": 5446 }, { "epoch": 0.4268808777429467, "grad_norm": 4.8112921714782715, "learning_rate": 9.73059657067006e-06, "loss": 1.3234, "step": 5447 }, { "epoch": 0.4269592476489028, "grad_norm": 5.472870349884033, "learning_rate": 9.730496708418975e-06, "loss": 0.8502, "step": 5448 }, { "epoch": 0.4270376175548589, "grad_norm": 3.6810827255249023, "learning_rate": 9.73039682817551e-06, "loss": 1.0458, "step": 5449 }, { "epoch": 0.427115987460815, "grad_norm": 2.621443271636963, "learning_rate": 9.730296929940046e-06, "loss": 1.1959, "step": 5450 }, { "epoch": 0.42719435736677114, "grad_norm": 2.5572009086608887, "learning_rate": 9.730197013712963e-06, "loss": 1.5194, "step": 5451 }, { "epoch": 0.42727272727272725, "grad_norm": 2.1671648025512695, "learning_rate": 9.73009707949464e-06, "loss": 1.5531, "step": 5452 }, { "epoch": 0.42735109717868336, "grad_norm": 2.056544065475464, "learning_rate": 9.729997127285457e-06, "loss": 1.2894, "step": 5453 }, { "epoch": 0.4274294670846395, "grad_norm": 4.207273006439209, "learning_rate": 9.729897157085794e-06, "loss": 1.5668, "step": 5454 }, { "epoch": 0.42750783699059564, "grad_norm": 3.231414556503296, "learning_rate": 9.729797168896032e-06, "loss": 0.7837, "step": 5455 }, { "epoch": 0.42758620689655175, "grad_norm": 3.03214693069458, "learning_rate": 9.72969716271655e-06, "loss": 1.7509, "step": 5456 }, { "epoch": 0.42766457680250786, "grad_norm": 2.629087448120117, "learning_rate": 9.72959713854773e-06, "loss": 1.4878, "step": 5457 }, { "epoch": 0.42774294670846397, "grad_norm": 2.828195810317993, "learning_rate": 9.729497096389951e-06, "loss": 1.3101, "step": 5458 }, { "epoch": 0.4278213166144201, "grad_norm": 2.735858917236328, "learning_rate": 9.729397036243595e-06, "loss": 1.3974, "step": 5459 }, { "epoch": 0.4278996865203762, "grad_norm": 9.574382781982422, "learning_rate": 9.729296958109042e-06, "loss": 1.6041, "step": 5460 }, { "epoch": 0.4279780564263323, "grad_norm": 5.1311445236206055, "learning_rate": 9.729196861986673e-06, "loss": 1.3263, "step": 5461 }, { "epoch": 0.4280564263322884, "grad_norm": 4.134424686431885, "learning_rate": 9.729096747876867e-06, "loss": 1.8048, "step": 5462 }, { "epoch": 0.4281347962382445, "grad_norm": 2.51320743560791, "learning_rate": 9.728996615780007e-06, "loss": 1.2796, "step": 5463 }, { "epoch": 0.42821316614420063, "grad_norm": 2.673602342605591, "learning_rate": 9.728896465696471e-06, "loss": 1.661, "step": 5464 }, { "epoch": 0.42829153605015674, "grad_norm": 2.200831174850464, "learning_rate": 9.728796297626644e-06, "loss": 1.5084, "step": 5465 }, { "epoch": 0.42836990595611285, "grad_norm": 4.400630950927734, "learning_rate": 9.728696111570903e-06, "loss": 1.7639, "step": 5466 }, { "epoch": 0.42844827586206896, "grad_norm": 2.800325632095337, "learning_rate": 9.728595907529633e-06, "loss": 1.2785, "step": 5467 }, { "epoch": 0.42852664576802507, "grad_norm": 4.371504306793213, "learning_rate": 9.728495685503212e-06, "loss": 1.5631, "step": 5468 }, { "epoch": 0.4286050156739812, "grad_norm": 17.621295928955078, "learning_rate": 9.72839544549202e-06, "loss": 0.9173, "step": 5469 }, { "epoch": 0.4286833855799373, "grad_norm": 8.793603897094727, "learning_rate": 9.728295187496444e-06, "loss": 2.3072, "step": 5470 }, { "epoch": 0.4287617554858934, "grad_norm": 2.643428087234497, "learning_rate": 9.728194911516859e-06, "loss": 1.7493, "step": 5471 }, { "epoch": 0.4288401253918495, "grad_norm": 2.6265811920166016, "learning_rate": 9.728094617553649e-06, "loss": 1.4272, "step": 5472 }, { "epoch": 0.4289184952978056, "grad_norm": 2.771238088607788, "learning_rate": 9.727994305607198e-06, "loss": 1.5653, "step": 5473 }, { "epoch": 0.42899686520376173, "grad_norm": 2.391080617904663, "learning_rate": 9.727893975677882e-06, "loss": 1.1276, "step": 5474 }, { "epoch": 0.42907523510971785, "grad_norm": 2.1544625759124756, "learning_rate": 9.72779362776609e-06, "loss": 1.2737, "step": 5475 }, { "epoch": 0.42915360501567396, "grad_norm": 2.6620171070098877, "learning_rate": 9.727693261872195e-06, "loss": 1.061, "step": 5476 }, { "epoch": 0.4292319749216301, "grad_norm": 2.2736685276031494, "learning_rate": 9.727592877996585e-06, "loss": 1.3082, "step": 5477 }, { "epoch": 0.42931034482758623, "grad_norm": 10.107622146606445, "learning_rate": 9.72749247613964e-06, "loss": 1.8819, "step": 5478 }, { "epoch": 0.42938871473354234, "grad_norm": 4.383995056152344, "learning_rate": 9.72739205630174e-06, "loss": 1.4334, "step": 5479 }, { "epoch": 0.42946708463949845, "grad_norm": 2.1586596965789795, "learning_rate": 9.727291618483273e-06, "loss": 1.1407, "step": 5480 }, { "epoch": 0.42954545454545456, "grad_norm": 3.7318191528320312, "learning_rate": 9.727191162684613e-06, "loss": 1.5251, "step": 5481 }, { "epoch": 0.4296238244514107, "grad_norm": 5.867218494415283, "learning_rate": 9.727090688906146e-06, "loss": 1.9301, "step": 5482 }, { "epoch": 0.4297021943573668, "grad_norm": 2.0723955631256104, "learning_rate": 9.726990197148256e-06, "loss": 1.3608, "step": 5483 }, { "epoch": 0.4297805642633229, "grad_norm": 2.29539155960083, "learning_rate": 9.726889687411323e-06, "loss": 1.5717, "step": 5484 }, { "epoch": 0.429858934169279, "grad_norm": 2.989818572998047, "learning_rate": 9.726789159695727e-06, "loss": 1.5495, "step": 5485 }, { "epoch": 0.4299373040752351, "grad_norm": 4.573372840881348, "learning_rate": 9.726688614001856e-06, "loss": 1.5847, "step": 5486 }, { "epoch": 0.4300156739811912, "grad_norm": 2.275520086288452, "learning_rate": 9.726588050330088e-06, "loss": 1.4095, "step": 5487 }, { "epoch": 0.43009404388714734, "grad_norm": 2.781947612762451, "learning_rate": 9.726487468680806e-06, "loss": 1.1222, "step": 5488 }, { "epoch": 0.43017241379310345, "grad_norm": 2.6682918071746826, "learning_rate": 9.726386869054393e-06, "loss": 1.4832, "step": 5489 }, { "epoch": 0.43025078369905956, "grad_norm": 2.160154342651367, "learning_rate": 9.726286251451233e-06, "loss": 1.2212, "step": 5490 }, { "epoch": 0.43032915360501567, "grad_norm": 4.648559093475342, "learning_rate": 9.726185615871709e-06, "loss": 1.711, "step": 5491 }, { "epoch": 0.4304075235109718, "grad_norm": 1.980804681777954, "learning_rate": 9.7260849623162e-06, "loss": 1.5002, "step": 5492 }, { "epoch": 0.4304858934169279, "grad_norm": 2.2802929878234863, "learning_rate": 9.725984290785094e-06, "loss": 1.4363, "step": 5493 }, { "epoch": 0.430564263322884, "grad_norm": 2.5195794105529785, "learning_rate": 9.72588360127877e-06, "loss": 1.2751, "step": 5494 }, { "epoch": 0.4306426332288401, "grad_norm": 3.2142441272735596, "learning_rate": 9.725782893797611e-06, "loss": 1.3441, "step": 5495 }, { "epoch": 0.4307210031347962, "grad_norm": 2.5047409534454346, "learning_rate": 9.725682168342001e-06, "loss": 1.4375, "step": 5496 }, { "epoch": 0.43079937304075233, "grad_norm": 11.830544471740723, "learning_rate": 9.725581424912325e-06, "loss": 2.7091, "step": 5497 }, { "epoch": 0.43087774294670844, "grad_norm": 6.097598075866699, "learning_rate": 9.725480663508964e-06, "loss": 1.4509, "step": 5498 }, { "epoch": 0.43095611285266455, "grad_norm": 2.6872007846832275, "learning_rate": 9.7253798841323e-06, "loss": 1.6394, "step": 5499 }, { "epoch": 0.43103448275862066, "grad_norm": 2.7367491722106934, "learning_rate": 9.725279086782719e-06, "loss": 1.4741, "step": 5500 }, { "epoch": 0.43111285266457683, "grad_norm": 3.0568065643310547, "learning_rate": 9.725178271460605e-06, "loss": 1.2119, "step": 5501 }, { "epoch": 0.43119122257053294, "grad_norm": 3.980952739715576, "learning_rate": 9.725077438166339e-06, "loss": 1.1425, "step": 5502 }, { "epoch": 0.43126959247648905, "grad_norm": 2.553257465362549, "learning_rate": 9.724976586900305e-06, "loss": 1.4501, "step": 5503 }, { "epoch": 0.43134796238244516, "grad_norm": 3.0273523330688477, "learning_rate": 9.724875717662887e-06, "loss": 0.8821, "step": 5504 }, { "epoch": 0.43142633228840127, "grad_norm": 3.100750684738159, "learning_rate": 9.724774830454469e-06, "loss": 0.9806, "step": 5505 }, { "epoch": 0.4315047021943574, "grad_norm": 2.825273036956787, "learning_rate": 9.724673925275434e-06, "loss": 1.0767, "step": 5506 }, { "epoch": 0.4315830721003135, "grad_norm": 2.811105251312256, "learning_rate": 9.724573002126166e-06, "loss": 1.6076, "step": 5507 }, { "epoch": 0.4316614420062696, "grad_norm": 3.2732996940612793, "learning_rate": 9.724472061007048e-06, "loss": 1.638, "step": 5508 }, { "epoch": 0.4317398119122257, "grad_norm": 2.6709251403808594, "learning_rate": 9.724371101918465e-06, "loss": 1.4591, "step": 5509 }, { "epoch": 0.4318181818181818, "grad_norm": 1.9457095861434937, "learning_rate": 9.724270124860804e-06, "loss": 1.1517, "step": 5510 }, { "epoch": 0.43189655172413793, "grad_norm": 7.052396774291992, "learning_rate": 9.724169129834444e-06, "loss": 0.7604, "step": 5511 }, { "epoch": 0.43197492163009404, "grad_norm": 2.1451237201690674, "learning_rate": 9.72406811683977e-06, "loss": 1.4137, "step": 5512 }, { "epoch": 0.43205329153605015, "grad_norm": 3.4646248817443848, "learning_rate": 9.723967085877169e-06, "loss": 1.2846, "step": 5513 }, { "epoch": 0.43213166144200627, "grad_norm": 2.867795705795288, "learning_rate": 9.723866036947022e-06, "loss": 1.4571, "step": 5514 }, { "epoch": 0.4322100313479624, "grad_norm": 1.765453815460205, "learning_rate": 9.723764970049715e-06, "loss": 1.6492, "step": 5515 }, { "epoch": 0.4322884012539185, "grad_norm": 1.8699429035186768, "learning_rate": 9.723663885185634e-06, "loss": 1.1833, "step": 5516 }, { "epoch": 0.4323667711598746, "grad_norm": 2.4640679359436035, "learning_rate": 9.723562782355162e-06, "loss": 1.3136, "step": 5517 }, { "epoch": 0.4324451410658307, "grad_norm": 2.215888023376465, "learning_rate": 9.723461661558682e-06, "loss": 1.306, "step": 5518 }, { "epoch": 0.4325235109717868, "grad_norm": 2.6541919708251953, "learning_rate": 9.723360522796582e-06, "loss": 1.1703, "step": 5519 }, { "epoch": 0.43260188087774293, "grad_norm": 5.885168075561523, "learning_rate": 9.723259366069242e-06, "loss": 1.5119, "step": 5520 }, { "epoch": 0.43268025078369904, "grad_norm": 3.6589064598083496, "learning_rate": 9.72315819137705e-06, "loss": 1.7478, "step": 5521 }, { "epoch": 0.43275862068965515, "grad_norm": 3.0724499225616455, "learning_rate": 9.72305699872039e-06, "loss": 1.5253, "step": 5522 }, { "epoch": 0.43283699059561126, "grad_norm": 2.102757692337036, "learning_rate": 9.722955788099648e-06, "loss": 1.3872, "step": 5523 }, { "epoch": 0.4329153605015674, "grad_norm": 2.3247597217559814, "learning_rate": 9.722854559515208e-06, "loss": 1.5192, "step": 5524 }, { "epoch": 0.43299373040752354, "grad_norm": 2.4630355834960938, "learning_rate": 9.722753312967455e-06, "loss": 1.239, "step": 5525 }, { "epoch": 0.43307210031347965, "grad_norm": 4.257775783538818, "learning_rate": 9.722652048456772e-06, "loss": 1.0316, "step": 5526 }, { "epoch": 0.43315047021943576, "grad_norm": 3.111936092376709, "learning_rate": 9.72255076598355e-06, "loss": 1.3727, "step": 5527 }, { "epoch": 0.43322884012539187, "grad_norm": 2.2500460147857666, "learning_rate": 9.722449465548169e-06, "loss": 1.5178, "step": 5528 }, { "epoch": 0.433307210031348, "grad_norm": 2.1737966537475586, "learning_rate": 9.722348147151015e-06, "loss": 1.4309, "step": 5529 }, { "epoch": 0.4333855799373041, "grad_norm": 2.8458144664764404, "learning_rate": 9.722246810792476e-06, "loss": 0.9752, "step": 5530 }, { "epoch": 0.4334639498432602, "grad_norm": 9.01622486114502, "learning_rate": 9.722145456472933e-06, "loss": 1.5053, "step": 5531 }, { "epoch": 0.4335423197492163, "grad_norm": 2.327800989151001, "learning_rate": 9.722044084192776e-06, "loss": 1.4064, "step": 5532 }, { "epoch": 0.4336206896551724, "grad_norm": 3.4011733531951904, "learning_rate": 9.721942693952388e-06, "loss": 1.478, "step": 5533 }, { "epoch": 0.43369905956112853, "grad_norm": 2.6134588718414307, "learning_rate": 9.721841285752156e-06, "loss": 1.1483, "step": 5534 }, { "epoch": 0.43377742946708464, "grad_norm": 3.3400638103485107, "learning_rate": 9.721739859592464e-06, "loss": 1.558, "step": 5535 }, { "epoch": 0.43385579937304075, "grad_norm": 2.4683964252471924, "learning_rate": 9.7216384154737e-06, "loss": 1.372, "step": 5536 }, { "epoch": 0.43393416927899686, "grad_norm": 4.891122817993164, "learning_rate": 9.721536953396248e-06, "loss": 1.3558, "step": 5537 }, { "epoch": 0.434012539184953, "grad_norm": 2.037259340286255, "learning_rate": 9.721435473360496e-06, "loss": 1.745, "step": 5538 }, { "epoch": 0.4340909090909091, "grad_norm": 2.3805084228515625, "learning_rate": 9.721333975366826e-06, "loss": 1.298, "step": 5539 }, { "epoch": 0.4341692789968652, "grad_norm": 4.379389762878418, "learning_rate": 9.721232459415628e-06, "loss": 1.4239, "step": 5540 }, { "epoch": 0.4342476489028213, "grad_norm": 4.10217809677124, "learning_rate": 9.721130925507287e-06, "loss": 1.1467, "step": 5541 }, { "epoch": 0.4343260188087774, "grad_norm": 2.3758292198181152, "learning_rate": 9.721029373642188e-06, "loss": 1.3574, "step": 5542 }, { "epoch": 0.4344043887147335, "grad_norm": 1.9551039934158325, "learning_rate": 9.720927803820719e-06, "loss": 1.3416, "step": 5543 }, { "epoch": 0.43448275862068964, "grad_norm": 2.6530771255493164, "learning_rate": 9.720826216043266e-06, "loss": 1.3575, "step": 5544 }, { "epoch": 0.43456112852664575, "grad_norm": 1.6451201438903809, "learning_rate": 9.720724610310212e-06, "loss": 1.2576, "step": 5545 }, { "epoch": 0.43463949843260186, "grad_norm": 2.795926570892334, "learning_rate": 9.720622986621949e-06, "loss": 1.4145, "step": 5546 }, { "epoch": 0.43471786833855797, "grad_norm": 2.576174736022949, "learning_rate": 9.720521344978859e-06, "loss": 1.3634, "step": 5547 }, { "epoch": 0.43479623824451413, "grad_norm": 5.0398993492126465, "learning_rate": 9.720419685381332e-06, "loss": 1.3779, "step": 5548 }, { "epoch": 0.43487460815047024, "grad_norm": 3.9859676361083984, "learning_rate": 9.720318007829754e-06, "loss": 1.3353, "step": 5549 }, { "epoch": 0.43495297805642635, "grad_norm": 2.023658037185669, "learning_rate": 9.720216312324508e-06, "loss": 1.2143, "step": 5550 }, { "epoch": 0.43503134796238246, "grad_norm": 6.802685260772705, "learning_rate": 9.720114598865985e-06, "loss": 1.2045, "step": 5551 }, { "epoch": 0.4351097178683386, "grad_norm": 5.593880653381348, "learning_rate": 9.72001286745457e-06, "loss": 1.1527, "step": 5552 }, { "epoch": 0.4351880877742947, "grad_norm": 2.2485642433166504, "learning_rate": 9.719911118090652e-06, "loss": 1.5096, "step": 5553 }, { "epoch": 0.4352664576802508, "grad_norm": 2.2758853435516357, "learning_rate": 9.719809350774613e-06, "loss": 1.6288, "step": 5554 }, { "epoch": 0.4353448275862069, "grad_norm": 4.992060661315918, "learning_rate": 9.719707565506848e-06, "loss": 1.3397, "step": 5555 }, { "epoch": 0.435423197492163, "grad_norm": 2.624269485473633, "learning_rate": 9.719605762287736e-06, "loss": 1.0532, "step": 5556 }, { "epoch": 0.4355015673981191, "grad_norm": 2.1167778968811035, "learning_rate": 9.71950394111767e-06, "loss": 1.4114, "step": 5557 }, { "epoch": 0.43557993730407524, "grad_norm": 3.1848883628845215, "learning_rate": 9.719402101997034e-06, "loss": 1.7538, "step": 5558 }, { "epoch": 0.43565830721003135, "grad_norm": 2.8156747817993164, "learning_rate": 9.719300244926216e-06, "loss": 1.3917, "step": 5559 }, { "epoch": 0.43573667711598746, "grad_norm": 2.4536778926849365, "learning_rate": 9.719198369905605e-06, "loss": 1.4729, "step": 5560 }, { "epoch": 0.43581504702194357, "grad_norm": 5.085902690887451, "learning_rate": 9.719096476935586e-06, "loss": 1.2199, "step": 5561 }, { "epoch": 0.4358934169278997, "grad_norm": 4.874725341796875, "learning_rate": 9.71899456601655e-06, "loss": 1.6521, "step": 5562 }, { "epoch": 0.4359717868338558, "grad_norm": 4.6530232429504395, "learning_rate": 9.71889263714888e-06, "loss": 1.2533, "step": 5563 }, { "epoch": 0.4360501567398119, "grad_norm": 2.374619245529175, "learning_rate": 9.718790690332969e-06, "loss": 1.3027, "step": 5564 }, { "epoch": 0.436128526645768, "grad_norm": 3.2195746898651123, "learning_rate": 9.718688725569198e-06, "loss": 1.7289, "step": 5565 }, { "epoch": 0.4362068965517241, "grad_norm": 5.767537593841553, "learning_rate": 9.718586742857963e-06, "loss": 1.2833, "step": 5566 }, { "epoch": 0.43628526645768023, "grad_norm": 2.8634018898010254, "learning_rate": 9.718484742199645e-06, "loss": 1.5977, "step": 5567 }, { "epoch": 0.43636363636363634, "grad_norm": 12.525891304016113, "learning_rate": 9.718382723594635e-06, "loss": 2.0899, "step": 5568 }, { "epoch": 0.43644200626959245, "grad_norm": 2.191786289215088, "learning_rate": 9.718280687043323e-06, "loss": 1.5076, "step": 5569 }, { "epoch": 0.43652037617554856, "grad_norm": 2.351219892501831, "learning_rate": 9.718178632546091e-06, "loss": 1.7466, "step": 5570 }, { "epoch": 0.43659874608150473, "grad_norm": 4.114883899688721, "learning_rate": 9.718076560103334e-06, "loss": 1.2677, "step": 5571 }, { "epoch": 0.43667711598746084, "grad_norm": 2.1010241508483887, "learning_rate": 9.717974469715434e-06, "loss": 1.4328, "step": 5572 }, { "epoch": 0.43675548589341695, "grad_norm": 3.195539951324463, "learning_rate": 9.717872361382786e-06, "loss": 1.7577, "step": 5573 }, { "epoch": 0.43683385579937306, "grad_norm": 3.0164999961853027, "learning_rate": 9.717770235105772e-06, "loss": 1.4634, "step": 5574 }, { "epoch": 0.43691222570532917, "grad_norm": 2.8228719234466553, "learning_rate": 9.717668090884785e-06, "loss": 1.5081, "step": 5575 }, { "epoch": 0.4369905956112853, "grad_norm": 2.61538028717041, "learning_rate": 9.71756592872021e-06, "loss": 1.4483, "step": 5576 }, { "epoch": 0.4370689655172414, "grad_norm": 2.0147039890289307, "learning_rate": 9.717463748612439e-06, "loss": 1.3957, "step": 5577 }, { "epoch": 0.4371473354231975, "grad_norm": 3.0008862018585205, "learning_rate": 9.717361550561857e-06, "loss": 1.5932, "step": 5578 }, { "epoch": 0.4372257053291536, "grad_norm": 3.2128350734710693, "learning_rate": 9.717259334568857e-06, "loss": 1.5345, "step": 5579 }, { "epoch": 0.4373040752351097, "grad_norm": 2.1432583332061768, "learning_rate": 9.717157100633823e-06, "loss": 1.0856, "step": 5580 }, { "epoch": 0.43738244514106583, "grad_norm": 2.5034875869750977, "learning_rate": 9.717054848757149e-06, "loss": 1.3773, "step": 5581 }, { "epoch": 0.43746081504702194, "grad_norm": 3.052197217941284, "learning_rate": 9.716952578939219e-06, "loss": 1.3171, "step": 5582 }, { "epoch": 0.43753918495297806, "grad_norm": 3.299100399017334, "learning_rate": 9.716850291180423e-06, "loss": 1.8742, "step": 5583 }, { "epoch": 0.43761755485893417, "grad_norm": 3.0231869220733643, "learning_rate": 9.716747985481155e-06, "loss": 1.0864, "step": 5584 }, { "epoch": 0.4376959247648903, "grad_norm": 2.514866828918457, "learning_rate": 9.716645661841796e-06, "loss": 1.4177, "step": 5585 }, { "epoch": 0.4377742946708464, "grad_norm": 2.207684278488159, "learning_rate": 9.716543320262744e-06, "loss": 1.447, "step": 5586 }, { "epoch": 0.4378526645768025, "grad_norm": 2.934260129928589, "learning_rate": 9.716440960744381e-06, "loss": 1.4676, "step": 5587 }, { "epoch": 0.4379310344827586, "grad_norm": 3.0045299530029297, "learning_rate": 9.716338583287099e-06, "loss": 1.447, "step": 5588 }, { "epoch": 0.4380094043887147, "grad_norm": 4.684900283813477, "learning_rate": 9.71623618789129e-06, "loss": 1.5128, "step": 5589 }, { "epoch": 0.43808777429467083, "grad_norm": 1.7034252882003784, "learning_rate": 9.716133774557337e-06, "loss": 1.1939, "step": 5590 }, { "epoch": 0.43816614420062694, "grad_norm": 8.55152416229248, "learning_rate": 9.716031343285636e-06, "loss": 0.439, "step": 5591 }, { "epoch": 0.43824451410658305, "grad_norm": 2.2792575359344482, "learning_rate": 9.715928894076576e-06, "loss": 1.0248, "step": 5592 }, { "epoch": 0.43832288401253916, "grad_norm": 4.20196008682251, "learning_rate": 9.715826426930542e-06, "loss": 0.9175, "step": 5593 }, { "epoch": 0.43840125391849527, "grad_norm": 5.898786544799805, "learning_rate": 9.715723941847927e-06, "loss": 1.3876, "step": 5594 }, { "epoch": 0.43847962382445144, "grad_norm": 2.8337225914001465, "learning_rate": 9.71562143882912e-06, "loss": 1.5787, "step": 5595 }, { "epoch": 0.43855799373040755, "grad_norm": 2.2201743125915527, "learning_rate": 9.715518917874512e-06, "loss": 1.3631, "step": 5596 }, { "epoch": 0.43863636363636366, "grad_norm": 2.373246431350708, "learning_rate": 9.71541637898449e-06, "loss": 1.4189, "step": 5597 }, { "epoch": 0.43871473354231977, "grad_norm": 4.787621974945068, "learning_rate": 9.71531382215945e-06, "loss": 1.5363, "step": 5598 }, { "epoch": 0.4387931034482759, "grad_norm": 2.343761920928955, "learning_rate": 9.715211247399774e-06, "loss": 1.8122, "step": 5599 }, { "epoch": 0.438871473354232, "grad_norm": 1.9373916387557983, "learning_rate": 9.715108654705858e-06, "loss": 1.2301, "step": 5600 }, { "epoch": 0.4389498432601881, "grad_norm": 2.900869607925415, "learning_rate": 9.715006044078091e-06, "loss": 1.2842, "step": 5601 }, { "epoch": 0.4390282131661442, "grad_norm": 2.4847140312194824, "learning_rate": 9.714903415516861e-06, "loss": 1.6372, "step": 5602 }, { "epoch": 0.4391065830721003, "grad_norm": 3.0355746746063232, "learning_rate": 9.714800769022564e-06, "loss": 1.6118, "step": 5603 }, { "epoch": 0.43918495297805643, "grad_norm": 3.246764659881592, "learning_rate": 9.714698104595584e-06, "loss": 1.4173, "step": 5604 }, { "epoch": 0.43926332288401254, "grad_norm": 7.012659549713135, "learning_rate": 9.714595422236312e-06, "loss": 1.9076, "step": 5605 }, { "epoch": 0.43934169278996865, "grad_norm": 3.968712091445923, "learning_rate": 9.714492721945144e-06, "loss": 1.2273, "step": 5606 }, { "epoch": 0.43942006269592476, "grad_norm": 2.8112363815307617, "learning_rate": 9.714390003722465e-06, "loss": 1.5585, "step": 5607 }, { "epoch": 0.4394984326018809, "grad_norm": 3.0436105728149414, "learning_rate": 9.714287267568669e-06, "loss": 1.2005, "step": 5608 }, { "epoch": 0.439576802507837, "grad_norm": 2.3830065727233887, "learning_rate": 9.714184513484146e-06, "loss": 1.3816, "step": 5609 }, { "epoch": 0.4396551724137931, "grad_norm": 5.755909442901611, "learning_rate": 9.714081741469286e-06, "loss": 1.5023, "step": 5610 }, { "epoch": 0.4397335423197492, "grad_norm": 3.3957085609436035, "learning_rate": 9.71397895152448e-06, "loss": 1.3521, "step": 5611 }, { "epoch": 0.4398119122257053, "grad_norm": 3.0750176906585693, "learning_rate": 9.713876143650118e-06, "loss": 1.4412, "step": 5612 }, { "epoch": 0.4398902821316614, "grad_norm": 10.199235916137695, "learning_rate": 9.713773317846596e-06, "loss": 1.5637, "step": 5613 }, { "epoch": 0.43996865203761754, "grad_norm": 4.597930431365967, "learning_rate": 9.713670474114298e-06, "loss": 1.6193, "step": 5614 }, { "epoch": 0.44004702194357365, "grad_norm": 2.0059661865234375, "learning_rate": 9.713567612453621e-06, "loss": 1.4205, "step": 5615 }, { "epoch": 0.44012539184952976, "grad_norm": 2.201960325241089, "learning_rate": 9.713464732864952e-06, "loss": 1.3709, "step": 5616 }, { "epoch": 0.44020376175548587, "grad_norm": 3.834315776824951, "learning_rate": 9.713361835348684e-06, "loss": 1.7414, "step": 5617 }, { "epoch": 0.44028213166144203, "grad_norm": 2.7712626457214355, "learning_rate": 9.71325891990521e-06, "loss": 1.5408, "step": 5618 }, { "epoch": 0.44036050156739814, "grad_norm": 2.5913493633270264, "learning_rate": 9.71315598653492e-06, "loss": 1.1981, "step": 5619 }, { "epoch": 0.44043887147335425, "grad_norm": 2.099203586578369, "learning_rate": 9.713053035238205e-06, "loss": 1.2008, "step": 5620 }, { "epoch": 0.44051724137931036, "grad_norm": 3.9945244789123535, "learning_rate": 9.712950066015456e-06, "loss": 0.9078, "step": 5621 }, { "epoch": 0.4405956112852665, "grad_norm": 5.815505027770996, "learning_rate": 9.712847078867066e-06, "loss": 1.3712, "step": 5622 }, { "epoch": 0.4406739811912226, "grad_norm": 3.7258265018463135, "learning_rate": 9.712744073793427e-06, "loss": 1.2948, "step": 5623 }, { "epoch": 0.4407523510971787, "grad_norm": 3.633713722229004, "learning_rate": 9.71264105079493e-06, "loss": 1.6574, "step": 5624 }, { "epoch": 0.4408307210031348, "grad_norm": 1.8395910263061523, "learning_rate": 9.712538009871969e-06, "loss": 1.3076, "step": 5625 }, { "epoch": 0.4409090909090909, "grad_norm": 2.0538291931152344, "learning_rate": 9.712434951024932e-06, "loss": 1.5233, "step": 5626 }, { "epoch": 0.440987460815047, "grad_norm": 5.962859630584717, "learning_rate": 9.712331874254213e-06, "loss": 1.4558, "step": 5627 }, { "epoch": 0.44106583072100314, "grad_norm": 1.8682414293289185, "learning_rate": 9.712228779560203e-06, "loss": 1.2173, "step": 5628 }, { "epoch": 0.44114420062695925, "grad_norm": 2.6910436153411865, "learning_rate": 9.712125666943297e-06, "loss": 1.3566, "step": 5629 }, { "epoch": 0.44122257053291536, "grad_norm": 3.201249122619629, "learning_rate": 9.712022536403885e-06, "loss": 1.5589, "step": 5630 }, { "epoch": 0.44130094043887147, "grad_norm": 2.4715185165405273, "learning_rate": 9.711919387942357e-06, "loss": 1.6143, "step": 5631 }, { "epoch": 0.4413793103448276, "grad_norm": 2.8958418369293213, "learning_rate": 9.711816221559109e-06, "loss": 1.3251, "step": 5632 }, { "epoch": 0.4414576802507837, "grad_norm": 3.1374635696411133, "learning_rate": 9.711713037254533e-06, "loss": 1.6122, "step": 5633 }, { "epoch": 0.4415360501567398, "grad_norm": 1.7421756982803345, "learning_rate": 9.711609835029022e-06, "loss": 0.541, "step": 5634 }, { "epoch": 0.4416144200626959, "grad_norm": 5.3964152336120605, "learning_rate": 9.711506614882965e-06, "loss": 1.1499, "step": 5635 }, { "epoch": 0.441692789968652, "grad_norm": 2.7589118480682373, "learning_rate": 9.711403376816757e-06, "loss": 1.4335, "step": 5636 }, { "epoch": 0.44177115987460813, "grad_norm": 2.9649786949157715, "learning_rate": 9.711300120830791e-06, "loss": 1.3392, "step": 5637 }, { "epoch": 0.44184952978056424, "grad_norm": 4.069084167480469, "learning_rate": 9.71119684692546e-06, "loss": 1.122, "step": 5638 }, { "epoch": 0.44192789968652035, "grad_norm": 2.763920545578003, "learning_rate": 9.711093555101155e-06, "loss": 1.1171, "step": 5639 }, { "epoch": 0.44200626959247646, "grad_norm": 2.6615734100341797, "learning_rate": 9.71099024535827e-06, "loss": 1.5508, "step": 5640 }, { "epoch": 0.4420846394984326, "grad_norm": 11.864046096801758, "learning_rate": 9.710886917697198e-06, "loss": 1.4354, "step": 5641 }, { "epoch": 0.44216300940438874, "grad_norm": 3.284196138381958, "learning_rate": 9.710783572118332e-06, "loss": 1.235, "step": 5642 }, { "epoch": 0.44224137931034485, "grad_norm": 4.4740519523620605, "learning_rate": 9.710680208622065e-06, "loss": 1.1939, "step": 5643 }, { "epoch": 0.44231974921630096, "grad_norm": 2.238133430480957, "learning_rate": 9.71057682720879e-06, "loss": 1.177, "step": 5644 }, { "epoch": 0.44239811912225707, "grad_norm": 6.30307674407959, "learning_rate": 9.7104734278789e-06, "loss": 1.4045, "step": 5645 }, { "epoch": 0.4424764890282132, "grad_norm": 2.79608416557312, "learning_rate": 9.710370010632788e-06, "loss": 1.3833, "step": 5646 }, { "epoch": 0.4425548589341693, "grad_norm": 3.4212486743927, "learning_rate": 9.71026657547085e-06, "loss": 1.2481, "step": 5647 }, { "epoch": 0.4426332288401254, "grad_norm": 2.2769615650177, "learning_rate": 9.710163122393475e-06, "loss": 1.406, "step": 5648 }, { "epoch": 0.4427115987460815, "grad_norm": 2.773547410964966, "learning_rate": 9.710059651401059e-06, "loss": 1.5909, "step": 5649 }, { "epoch": 0.4427899686520376, "grad_norm": 2.4123568534851074, "learning_rate": 9.709956162493996e-06, "loss": 1.0306, "step": 5650 }, { "epoch": 0.44286833855799373, "grad_norm": 6.930375099182129, "learning_rate": 9.709852655672679e-06, "loss": 1.1539, "step": 5651 }, { "epoch": 0.44294670846394985, "grad_norm": 15.280781745910645, "learning_rate": 9.7097491309375e-06, "loss": 2.259, "step": 5652 }, { "epoch": 0.44302507836990596, "grad_norm": 2.0690653324127197, "learning_rate": 9.709645588288857e-06, "loss": 1.1942, "step": 5653 }, { "epoch": 0.44310344827586207, "grad_norm": 10.41873836517334, "learning_rate": 9.70954202772714e-06, "loss": 1.6094, "step": 5654 }, { "epoch": 0.4431818181818182, "grad_norm": 2.3647196292877197, "learning_rate": 9.709438449252744e-06, "loss": 1.5049, "step": 5655 }, { "epoch": 0.4432601880877743, "grad_norm": 3.0413169860839844, "learning_rate": 9.709334852866063e-06, "loss": 1.602, "step": 5656 }, { "epoch": 0.4433385579937304, "grad_norm": 3.2450973987579346, "learning_rate": 9.709231238567492e-06, "loss": 1.0819, "step": 5657 }, { "epoch": 0.4434169278996865, "grad_norm": 4.123215675354004, "learning_rate": 9.709127606357423e-06, "loss": 1.6936, "step": 5658 }, { "epoch": 0.4434952978056426, "grad_norm": 3.2769854068756104, "learning_rate": 9.709023956236253e-06, "loss": 1.3634, "step": 5659 }, { "epoch": 0.44357366771159873, "grad_norm": 2.797913074493408, "learning_rate": 9.708920288204372e-06, "loss": 1.5079, "step": 5660 }, { "epoch": 0.44365203761755484, "grad_norm": 2.880924940109253, "learning_rate": 9.708816602262179e-06, "loss": 0.9317, "step": 5661 }, { "epoch": 0.44373040752351095, "grad_norm": 2.7890913486480713, "learning_rate": 9.708712898410065e-06, "loss": 1.4333, "step": 5662 }, { "epoch": 0.44380877742946706, "grad_norm": 2.4814724922180176, "learning_rate": 9.708609176648427e-06, "loss": 1.5288, "step": 5663 }, { "epoch": 0.44388714733542317, "grad_norm": 3.1342411041259766, "learning_rate": 9.708505436977657e-06, "loss": 1.5873, "step": 5664 }, { "epoch": 0.44396551724137934, "grad_norm": 3.225039482116699, "learning_rate": 9.708401679398152e-06, "loss": 1.7927, "step": 5665 }, { "epoch": 0.44404388714733545, "grad_norm": 3.3149476051330566, "learning_rate": 9.708297903910305e-06, "loss": 0.9828, "step": 5666 }, { "epoch": 0.44412225705329156, "grad_norm": 2.145612955093384, "learning_rate": 9.70819411051451e-06, "loss": 1.2319, "step": 5667 }, { "epoch": 0.44420062695924767, "grad_norm": 2.0686593055725098, "learning_rate": 9.708090299211162e-06, "loss": 1.6013, "step": 5668 }, { "epoch": 0.4442789968652038, "grad_norm": 2.435516834259033, "learning_rate": 9.707986470000658e-06, "loss": 1.5932, "step": 5669 }, { "epoch": 0.4443573667711599, "grad_norm": 4.6605000495910645, "learning_rate": 9.707882622883394e-06, "loss": 1.5452, "step": 5670 }, { "epoch": 0.444435736677116, "grad_norm": 2.738138437271118, "learning_rate": 9.70777875785976e-06, "loss": 1.5079, "step": 5671 }, { "epoch": 0.4445141065830721, "grad_norm": 1.9664170742034912, "learning_rate": 9.707674874930153e-06, "loss": 1.5544, "step": 5672 }, { "epoch": 0.4445924764890282, "grad_norm": 3.819599151611328, "learning_rate": 9.70757097409497e-06, "loss": 1.1985, "step": 5673 }, { "epoch": 0.44467084639498433, "grad_norm": 2.418004274368286, "learning_rate": 9.707467055354604e-06, "loss": 1.91, "step": 5674 }, { "epoch": 0.44474921630094044, "grad_norm": 5.097597122192383, "learning_rate": 9.707363118709453e-06, "loss": 1.3178, "step": 5675 }, { "epoch": 0.44482758620689655, "grad_norm": 1.9361625909805298, "learning_rate": 9.70725916415991e-06, "loss": 1.1926, "step": 5676 }, { "epoch": 0.44490595611285266, "grad_norm": 2.1104893684387207, "learning_rate": 9.707155191706371e-06, "loss": 1.1783, "step": 5677 }, { "epoch": 0.4449843260188088, "grad_norm": 2.827113628387451, "learning_rate": 9.707051201349232e-06, "loss": 1.9553, "step": 5678 }, { "epoch": 0.4450626959247649, "grad_norm": 2.8478102684020996, "learning_rate": 9.706947193088886e-06, "loss": 1.5891, "step": 5679 }, { "epoch": 0.445141065830721, "grad_norm": 2.741361379623413, "learning_rate": 9.706843166925733e-06, "loss": 1.0269, "step": 5680 }, { "epoch": 0.4452194357366771, "grad_norm": 2.4569501876831055, "learning_rate": 9.706739122860162e-06, "loss": 1.2732, "step": 5681 }, { "epoch": 0.4452978056426332, "grad_norm": 4.8726983070373535, "learning_rate": 9.706635060892577e-06, "loss": 1.7687, "step": 5682 }, { "epoch": 0.4453761755485893, "grad_norm": 2.626831531524658, "learning_rate": 9.706530981023368e-06, "loss": 1.5798, "step": 5683 }, { "epoch": 0.44545454545454544, "grad_norm": 3.7561874389648438, "learning_rate": 9.706426883252934e-06, "loss": 1.9202, "step": 5684 }, { "epoch": 0.44553291536050155, "grad_norm": 2.08947491645813, "learning_rate": 9.706322767581668e-06, "loss": 1.5986, "step": 5685 }, { "epoch": 0.44561128526645766, "grad_norm": 2.5847020149230957, "learning_rate": 9.706218634009968e-06, "loss": 1.0063, "step": 5686 }, { "epoch": 0.44568965517241377, "grad_norm": 3.1413533687591553, "learning_rate": 9.70611448253823e-06, "loss": 1.5063, "step": 5687 }, { "epoch": 0.4457680250783699, "grad_norm": 2.2463130950927734, "learning_rate": 9.706010313166848e-06, "loss": 1.1733, "step": 5688 }, { "epoch": 0.44584639498432604, "grad_norm": 2.612644672393799, "learning_rate": 9.70590612589622e-06, "loss": 1.4245, "step": 5689 }, { "epoch": 0.44592476489028215, "grad_norm": 6.760689735412598, "learning_rate": 9.705801920726745e-06, "loss": 0.9019, "step": 5690 }, { "epoch": 0.44600313479623827, "grad_norm": 2.9992690086364746, "learning_rate": 9.705697697658815e-06, "loss": 1.2046, "step": 5691 }, { "epoch": 0.4460815047021944, "grad_norm": 2.7704381942749023, "learning_rate": 9.705593456692828e-06, "loss": 1.1462, "step": 5692 }, { "epoch": 0.4461598746081505, "grad_norm": 3.818880081176758, "learning_rate": 9.70548919782918e-06, "loss": 1.762, "step": 5693 }, { "epoch": 0.4462382445141066, "grad_norm": 3.849412441253662, "learning_rate": 9.705384921068268e-06, "loss": 1.7522, "step": 5694 }, { "epoch": 0.4463166144200627, "grad_norm": 3.391329288482666, "learning_rate": 9.705280626410489e-06, "loss": 1.5161, "step": 5695 }, { "epoch": 0.4463949843260188, "grad_norm": 3.318554162979126, "learning_rate": 9.70517631385624e-06, "loss": 1.4277, "step": 5696 }, { "epoch": 0.44647335423197493, "grad_norm": 1.9680215120315552, "learning_rate": 9.705071983405916e-06, "loss": 1.2989, "step": 5697 }, { "epoch": 0.44655172413793104, "grad_norm": 3.941241979598999, "learning_rate": 9.704967635059914e-06, "loss": 0.9346, "step": 5698 }, { "epoch": 0.44663009404388715, "grad_norm": 2.225351095199585, "learning_rate": 9.704863268818633e-06, "loss": 1.4545, "step": 5699 }, { "epoch": 0.44670846394984326, "grad_norm": 2.5036652088165283, "learning_rate": 9.704758884682468e-06, "loss": 1.6252, "step": 5700 }, { "epoch": 0.44678683385579937, "grad_norm": 2.6568729877471924, "learning_rate": 9.70465448265182e-06, "loss": 1.4994, "step": 5701 }, { "epoch": 0.4468652037617555, "grad_norm": 3.4167821407318115, "learning_rate": 9.704550062727078e-06, "loss": 1.5055, "step": 5702 }, { "epoch": 0.4469435736677116, "grad_norm": 3.2001373767852783, "learning_rate": 9.704445624908646e-06, "loss": 1.3817, "step": 5703 }, { "epoch": 0.4470219435736677, "grad_norm": 3.006596565246582, "learning_rate": 9.704341169196918e-06, "loss": 1.2174, "step": 5704 }, { "epoch": 0.4471003134796238, "grad_norm": 3.298415422439575, "learning_rate": 9.704236695592295e-06, "loss": 0.5365, "step": 5705 }, { "epoch": 0.4471786833855799, "grad_norm": 2.7530786991119385, "learning_rate": 9.70413220409517e-06, "loss": 1.4358, "step": 5706 }, { "epoch": 0.44725705329153603, "grad_norm": 5.559868812561035, "learning_rate": 9.704027694705943e-06, "loss": 1.1655, "step": 5707 }, { "epoch": 0.44733542319749214, "grad_norm": 3.0667355060577393, "learning_rate": 9.703923167425012e-06, "loss": 1.4492, "step": 5708 }, { "epoch": 0.44741379310344825, "grad_norm": 2.7078640460968018, "learning_rate": 9.70381862225277e-06, "loss": 1.6315, "step": 5709 }, { "epoch": 0.44749216300940436, "grad_norm": 2.3143980503082275, "learning_rate": 9.70371405918962e-06, "loss": 1.1475, "step": 5710 }, { "epoch": 0.4475705329153605, "grad_norm": 3.0883193016052246, "learning_rate": 9.703609478235958e-06, "loss": 1.4149, "step": 5711 }, { "epoch": 0.44764890282131664, "grad_norm": 2.473043441772461, "learning_rate": 9.703504879392181e-06, "loss": 1.6114, "step": 5712 }, { "epoch": 0.44772727272727275, "grad_norm": 3.1764464378356934, "learning_rate": 9.703400262658687e-06, "loss": 0.8931, "step": 5713 }, { "epoch": 0.44780564263322886, "grad_norm": 1.958848237991333, "learning_rate": 9.703295628035876e-06, "loss": 1.4427, "step": 5714 }, { "epoch": 0.447884012539185, "grad_norm": 2.238342046737671, "learning_rate": 9.703190975524142e-06, "loss": 1.3404, "step": 5715 }, { "epoch": 0.4479623824451411, "grad_norm": 1.9165743589401245, "learning_rate": 9.703086305123887e-06, "loss": 1.2673, "step": 5716 }, { "epoch": 0.4480407523510972, "grad_norm": 2.591076135635376, "learning_rate": 9.702981616835504e-06, "loss": 1.3834, "step": 5717 }, { "epoch": 0.4481191222570533, "grad_norm": 3.679197311401367, "learning_rate": 9.702876910659398e-06, "loss": 1.4119, "step": 5718 }, { "epoch": 0.4481974921630094, "grad_norm": 2.889958620071411, "learning_rate": 9.702772186595964e-06, "loss": 1.6358, "step": 5719 }, { "epoch": 0.4482758620689655, "grad_norm": 3.398513078689575, "learning_rate": 9.7026674446456e-06, "loss": 1.5847, "step": 5720 }, { "epoch": 0.44835423197492164, "grad_norm": 2.543168306350708, "learning_rate": 9.702562684808701e-06, "loss": 1.4334, "step": 5721 }, { "epoch": 0.44843260188087775, "grad_norm": 4.878701210021973, "learning_rate": 9.702457907085672e-06, "loss": 1.3649, "step": 5722 }, { "epoch": 0.44851097178683386, "grad_norm": 2.3712241649627686, "learning_rate": 9.702353111476907e-06, "loss": 1.1785, "step": 5723 }, { "epoch": 0.44858934169278997, "grad_norm": 3.4736764430999756, "learning_rate": 9.702248297982807e-06, "loss": 1.0027, "step": 5724 }, { "epoch": 0.4486677115987461, "grad_norm": 3.4215645790100098, "learning_rate": 9.70214346660377e-06, "loss": 1.8528, "step": 5725 }, { "epoch": 0.4487460815047022, "grad_norm": 3.46659779548645, "learning_rate": 9.702038617340194e-06, "loss": 1.4459, "step": 5726 }, { "epoch": 0.4488244514106583, "grad_norm": 3.6586837768554688, "learning_rate": 9.701933750192478e-06, "loss": 1.4054, "step": 5727 }, { "epoch": 0.4489028213166144, "grad_norm": 3.04125714302063, "learning_rate": 9.701828865161022e-06, "loss": 1.6032, "step": 5728 }, { "epoch": 0.4489811912225705, "grad_norm": 2.539461374282837, "learning_rate": 9.701723962246223e-06, "loss": 1.549, "step": 5729 }, { "epoch": 0.44905956112852663, "grad_norm": 8.518688201904297, "learning_rate": 9.70161904144848e-06, "loss": 1.3288, "step": 5730 }, { "epoch": 0.44913793103448274, "grad_norm": 2.6679787635803223, "learning_rate": 9.701514102768195e-06, "loss": 1.5517, "step": 5731 }, { "epoch": 0.44921630094043885, "grad_norm": 6.009375095367432, "learning_rate": 9.701409146205765e-06, "loss": 1.2947, "step": 5732 }, { "epoch": 0.44929467084639496, "grad_norm": 2.8373091220855713, "learning_rate": 9.701304171761589e-06, "loss": 1.0254, "step": 5733 }, { "epoch": 0.44937304075235107, "grad_norm": 3.5192177295684814, "learning_rate": 9.701199179436068e-06, "loss": 1.3214, "step": 5734 }, { "epoch": 0.44945141065830724, "grad_norm": 3.4487674236297607, "learning_rate": 9.701094169229596e-06, "loss": 1.258, "step": 5735 }, { "epoch": 0.44952978056426335, "grad_norm": 3.0105905532836914, "learning_rate": 9.700989141142581e-06, "loss": 1.4619, "step": 5736 }, { "epoch": 0.44960815047021946, "grad_norm": 2.0756354331970215, "learning_rate": 9.700884095175414e-06, "loss": 1.3068, "step": 5737 }, { "epoch": 0.44968652037617557, "grad_norm": 2.243952989578247, "learning_rate": 9.7007790313285e-06, "loss": 1.4217, "step": 5738 }, { "epoch": 0.4497648902821317, "grad_norm": 2.1509082317352295, "learning_rate": 9.700673949602238e-06, "loss": 1.5118, "step": 5739 }, { "epoch": 0.4498432601880878, "grad_norm": 2.6677303314208984, "learning_rate": 9.700568849997026e-06, "loss": 1.5733, "step": 5740 }, { "epoch": 0.4499216300940439, "grad_norm": 3.2059054374694824, "learning_rate": 9.700463732513266e-06, "loss": 1.5016, "step": 5741 }, { "epoch": 0.45, "grad_norm": 2.3001043796539307, "learning_rate": 9.700358597151353e-06, "loss": 1.0284, "step": 5742 }, { "epoch": 0.4500783699059561, "grad_norm": 2.4395499229431152, "learning_rate": 9.700253443911693e-06, "loss": 1.2536, "step": 5743 }, { "epoch": 0.45015673981191223, "grad_norm": 2.368058919906616, "learning_rate": 9.700148272794682e-06, "loss": 1.2555, "step": 5744 }, { "epoch": 0.45023510971786834, "grad_norm": 7.875234603881836, "learning_rate": 9.700043083800721e-06, "loss": 1.7358, "step": 5745 }, { "epoch": 0.45031347962382445, "grad_norm": 3.7083356380462646, "learning_rate": 9.69993787693021e-06, "loss": 1.6266, "step": 5746 }, { "epoch": 0.45039184952978056, "grad_norm": 3.331923007965088, "learning_rate": 9.699832652183549e-06, "loss": 1.1336, "step": 5747 }, { "epoch": 0.4504702194357367, "grad_norm": 2.1128909587860107, "learning_rate": 9.69972740956114e-06, "loss": 1.262, "step": 5748 }, { "epoch": 0.4505485893416928, "grad_norm": 4.163225173950195, "learning_rate": 9.699622149063382e-06, "loss": 1.8996, "step": 5749 }, { "epoch": 0.4506269592476489, "grad_norm": 2.4597318172454834, "learning_rate": 9.699516870690674e-06, "loss": 1.6564, "step": 5750 }, { "epoch": 0.450705329153605, "grad_norm": 3.966616153717041, "learning_rate": 9.699411574443419e-06, "loss": 1.1618, "step": 5751 }, { "epoch": 0.4507836990595611, "grad_norm": 1.9765653610229492, "learning_rate": 9.699306260322014e-06, "loss": 1.5914, "step": 5752 }, { "epoch": 0.4508620689655172, "grad_norm": 3.543687105178833, "learning_rate": 9.699200928326863e-06, "loss": 1.4561, "step": 5753 }, { "epoch": 0.45094043887147334, "grad_norm": 6.9323554039001465, "learning_rate": 9.699095578458366e-06, "loss": 1.4172, "step": 5754 }, { "epoch": 0.45101880877742945, "grad_norm": 2.1412830352783203, "learning_rate": 9.698990210716922e-06, "loss": 1.5344, "step": 5755 }, { "epoch": 0.45109717868338556, "grad_norm": 4.40714168548584, "learning_rate": 9.698884825102931e-06, "loss": 1.3251, "step": 5756 }, { "epoch": 0.45117554858934167, "grad_norm": 2.582414150238037, "learning_rate": 9.6987794216168e-06, "loss": 1.2715, "step": 5757 }, { "epoch": 0.4512539184952978, "grad_norm": 3.0458967685699463, "learning_rate": 9.698674000258922e-06, "loss": 1.203, "step": 5758 }, { "epoch": 0.45133228840125394, "grad_norm": 4.8551788330078125, "learning_rate": 9.698568561029704e-06, "loss": 1.5883, "step": 5759 }, { "epoch": 0.45141065830721006, "grad_norm": 5.240562438964844, "learning_rate": 9.698463103929542e-06, "loss": 0.5969, "step": 5760 }, { "epoch": 0.45148902821316617, "grad_norm": 3.7694785594940186, "learning_rate": 9.69835762895884e-06, "loss": 1.2342, "step": 5761 }, { "epoch": 0.4515673981191223, "grad_norm": 9.915982246398926, "learning_rate": 9.698252136118002e-06, "loss": 1.5283, "step": 5762 }, { "epoch": 0.4516457680250784, "grad_norm": 5.493319034576416, "learning_rate": 9.698146625407422e-06, "loss": 1.6198, "step": 5763 }, { "epoch": 0.4517241379310345, "grad_norm": 5.130612373352051, "learning_rate": 9.698041096827509e-06, "loss": 0.941, "step": 5764 }, { "epoch": 0.4518025078369906, "grad_norm": 2.4372916221618652, "learning_rate": 9.697935550378658e-06, "loss": 1.2673, "step": 5765 }, { "epoch": 0.4518808777429467, "grad_norm": 4.568318843841553, "learning_rate": 9.697829986061274e-06, "loss": 1.9657, "step": 5766 }, { "epoch": 0.45195924764890283, "grad_norm": 1.9469536542892456, "learning_rate": 9.697724403875758e-06, "loss": 1.166, "step": 5767 }, { "epoch": 0.45203761755485894, "grad_norm": 2.673736810684204, "learning_rate": 9.697618803822511e-06, "loss": 1.455, "step": 5768 }, { "epoch": 0.45211598746081505, "grad_norm": 3.909097194671631, "learning_rate": 9.697513185901935e-06, "loss": 1.2126, "step": 5769 }, { "epoch": 0.45219435736677116, "grad_norm": 3.5082762241363525, "learning_rate": 9.69740755011443e-06, "loss": 1.4071, "step": 5770 }, { "epoch": 0.45227272727272727, "grad_norm": 2.821502923965454, "learning_rate": 9.697301896460403e-06, "loss": 1.6639, "step": 5771 }, { "epoch": 0.4523510971786834, "grad_norm": 2.1914308071136475, "learning_rate": 9.697196224940248e-06, "loss": 1.1411, "step": 5772 }, { "epoch": 0.4524294670846395, "grad_norm": 2.0363893508911133, "learning_rate": 9.697090535554374e-06, "loss": 1.3507, "step": 5773 }, { "epoch": 0.4525078369905956, "grad_norm": 1.9984722137451172, "learning_rate": 9.69698482830318e-06, "loss": 1.7612, "step": 5774 }, { "epoch": 0.4525862068965517, "grad_norm": 2.649327516555786, "learning_rate": 9.696879103187068e-06, "loss": 1.5286, "step": 5775 }, { "epoch": 0.4526645768025078, "grad_norm": 3.4482483863830566, "learning_rate": 9.696773360206439e-06, "loss": 1.2504, "step": 5776 }, { "epoch": 0.45274294670846393, "grad_norm": 2.943544387817383, "learning_rate": 9.696667599361699e-06, "loss": 1.7795, "step": 5777 }, { "epoch": 0.45282131661442004, "grad_norm": 2.8291077613830566, "learning_rate": 9.696561820653246e-06, "loss": 1.4927, "step": 5778 }, { "epoch": 0.45289968652037615, "grad_norm": 3.9097514152526855, "learning_rate": 9.696456024081485e-06, "loss": 1.3869, "step": 5779 }, { "epoch": 0.45297805642633227, "grad_norm": 7.408991813659668, "learning_rate": 9.696350209646817e-06, "loss": 0.8338, "step": 5780 }, { "epoch": 0.4530564263322884, "grad_norm": 3.0417098999023438, "learning_rate": 9.696244377349647e-06, "loss": 1.4104, "step": 5781 }, { "epoch": 0.45313479623824454, "grad_norm": 2.478566884994507, "learning_rate": 9.696138527190373e-06, "loss": 1.5633, "step": 5782 }, { "epoch": 0.45321316614420065, "grad_norm": 2.8170785903930664, "learning_rate": 9.6960326591694e-06, "loss": 1.4337, "step": 5783 }, { "epoch": 0.45329153605015676, "grad_norm": 2.5411722660064697, "learning_rate": 9.695926773287133e-06, "loss": 1.167, "step": 5784 }, { "epoch": 0.4533699059561129, "grad_norm": 2.650402069091797, "learning_rate": 9.69582086954397e-06, "loss": 1.0453, "step": 5785 }, { "epoch": 0.453448275862069, "grad_norm": 2.606121063232422, "learning_rate": 9.695714947940319e-06, "loss": 1.6921, "step": 5786 }, { "epoch": 0.4535266457680251, "grad_norm": 4.749131202697754, "learning_rate": 9.695609008476578e-06, "loss": 1.2502, "step": 5787 }, { "epoch": 0.4536050156739812, "grad_norm": 1.9326382875442505, "learning_rate": 9.695503051153152e-06, "loss": 1.4714, "step": 5788 }, { "epoch": 0.4536833855799373, "grad_norm": 6.908269882202148, "learning_rate": 9.695397075970445e-06, "loss": 1.0399, "step": 5789 }, { "epoch": 0.4537617554858934, "grad_norm": 3.302111864089966, "learning_rate": 9.69529108292886e-06, "loss": 1.7456, "step": 5790 }, { "epoch": 0.45384012539184954, "grad_norm": 3.0642905235290527, "learning_rate": 9.695185072028799e-06, "loss": 1.2403, "step": 5791 }, { "epoch": 0.45391849529780565, "grad_norm": 2.991595983505249, "learning_rate": 9.695079043270665e-06, "loss": 1.3834, "step": 5792 }, { "epoch": 0.45399686520376176, "grad_norm": 2.6335301399230957, "learning_rate": 9.694972996654861e-06, "loss": 1.2852, "step": 5793 }, { "epoch": 0.45407523510971787, "grad_norm": 6.174997329711914, "learning_rate": 9.694866932181793e-06, "loss": 1.6528, "step": 5794 }, { "epoch": 0.454153605015674, "grad_norm": 4.665185928344727, "learning_rate": 9.694760849851862e-06, "loss": 1.1815, "step": 5795 }, { "epoch": 0.4542319749216301, "grad_norm": 3.0143980979919434, "learning_rate": 9.694654749665471e-06, "loss": 1.3488, "step": 5796 }, { "epoch": 0.4543103448275862, "grad_norm": 5.51951789855957, "learning_rate": 9.694548631623026e-06, "loss": 1.6105, "step": 5797 }, { "epoch": 0.4543887147335423, "grad_norm": 2.227822780609131, "learning_rate": 9.69444249572493e-06, "loss": 1.167, "step": 5798 }, { "epoch": 0.4544670846394984, "grad_norm": 10.772110939025879, "learning_rate": 9.694336341971583e-06, "loss": 1.6699, "step": 5799 }, { "epoch": 0.45454545454545453, "grad_norm": 2.680109977722168, "learning_rate": 9.694230170363396e-06, "loss": 1.2048, "step": 5800 }, { "epoch": 0.45462382445141064, "grad_norm": 2.6296887397766113, "learning_rate": 9.694123980900766e-06, "loss": 1.6211, "step": 5801 }, { "epoch": 0.45470219435736675, "grad_norm": 4.666769504547119, "learning_rate": 9.694017773584099e-06, "loss": 1.5384, "step": 5802 }, { "epoch": 0.45478056426332286, "grad_norm": 7.791932106018066, "learning_rate": 9.6939115484138e-06, "loss": 0.5291, "step": 5803 }, { "epoch": 0.454858934169279, "grad_norm": 3.3566958904266357, "learning_rate": 9.693805305390274e-06, "loss": 0.8314, "step": 5804 }, { "epoch": 0.4549373040752351, "grad_norm": 1.9790246486663818, "learning_rate": 9.693699044513922e-06, "loss": 1.2072, "step": 5805 }, { "epoch": 0.45501567398119125, "grad_norm": 3.2008652687072754, "learning_rate": 9.693592765785148e-06, "loss": 1.7595, "step": 5806 }, { "epoch": 0.45509404388714736, "grad_norm": 3.3388428688049316, "learning_rate": 9.693486469204361e-06, "loss": 1.3319, "step": 5807 }, { "epoch": 0.45517241379310347, "grad_norm": 1.7865999937057495, "learning_rate": 9.69338015477196e-06, "loss": 0.6589, "step": 5808 }, { "epoch": 0.4552507836990596, "grad_norm": 4.128185749053955, "learning_rate": 9.693273822488352e-06, "loss": 1.5866, "step": 5809 }, { "epoch": 0.4553291536050157, "grad_norm": 4.203315734863281, "learning_rate": 9.69316747235394e-06, "loss": 1.9005, "step": 5810 }, { "epoch": 0.4554075235109718, "grad_norm": 2.3800535202026367, "learning_rate": 9.693061104369132e-06, "loss": 1.1902, "step": 5811 }, { "epoch": 0.4554858934169279, "grad_norm": 3.2756595611572266, "learning_rate": 9.692954718534328e-06, "loss": 1.3889, "step": 5812 }, { "epoch": 0.455564263322884, "grad_norm": 4.968656063079834, "learning_rate": 9.692848314849937e-06, "loss": 1.584, "step": 5813 }, { "epoch": 0.45564263322884013, "grad_norm": 2.544111490249634, "learning_rate": 9.692741893316359e-06, "loss": 1.6214, "step": 5814 }, { "epoch": 0.45572100313479624, "grad_norm": 19.441253662109375, "learning_rate": 9.692635453934003e-06, "loss": 0.9239, "step": 5815 }, { "epoch": 0.45579937304075235, "grad_norm": 7.486301422119141, "learning_rate": 9.692528996703272e-06, "loss": 1.11, "step": 5816 }, { "epoch": 0.45587774294670846, "grad_norm": 3.5999720096588135, "learning_rate": 9.692422521624569e-06, "loss": 1.5827, "step": 5817 }, { "epoch": 0.4559561128526646, "grad_norm": 5.194386005401611, "learning_rate": 9.692316028698302e-06, "loss": 1.8599, "step": 5818 }, { "epoch": 0.4560344827586207, "grad_norm": 1.7800672054290771, "learning_rate": 9.692209517924875e-06, "loss": 0.2361, "step": 5819 }, { "epoch": 0.4561128526645768, "grad_norm": 4.978375434875488, "learning_rate": 9.692102989304693e-06, "loss": 1.6366, "step": 5820 }, { "epoch": 0.4561912225705329, "grad_norm": 3.1198782920837402, "learning_rate": 9.691996442838161e-06, "loss": 1.2084, "step": 5821 }, { "epoch": 0.456269592476489, "grad_norm": 2.442242383956909, "learning_rate": 9.691889878525686e-06, "loss": 1.4443, "step": 5822 }, { "epoch": 0.4563479623824451, "grad_norm": 3.8898110389709473, "learning_rate": 9.69178329636767e-06, "loss": 1.4521, "step": 5823 }, { "epoch": 0.45642633228840124, "grad_norm": 3.9536731243133545, "learning_rate": 9.69167669636452e-06, "loss": 1.6308, "step": 5824 }, { "epoch": 0.45650470219435735, "grad_norm": 2.8021018505096436, "learning_rate": 9.691570078516644e-06, "loss": 1.4511, "step": 5825 }, { "epoch": 0.45658307210031346, "grad_norm": 5.237608909606934, "learning_rate": 9.691463442824443e-06, "loss": 1.6713, "step": 5826 }, { "epoch": 0.45666144200626957, "grad_norm": 2.7965891361236572, "learning_rate": 9.691356789288326e-06, "loss": 1.4471, "step": 5827 }, { "epoch": 0.4567398119122257, "grad_norm": 3.0764963626861572, "learning_rate": 9.691250117908695e-06, "loss": 1.4497, "step": 5828 }, { "epoch": 0.45681818181818185, "grad_norm": 4.189007759094238, "learning_rate": 9.69114342868596e-06, "loss": 1.0307, "step": 5829 }, { "epoch": 0.45689655172413796, "grad_norm": 2.742611885070801, "learning_rate": 9.691036721620525e-06, "loss": 1.4014, "step": 5830 }, { "epoch": 0.45697492163009407, "grad_norm": 1.9299060106277466, "learning_rate": 9.690929996712797e-06, "loss": 1.1949, "step": 5831 }, { "epoch": 0.4570532915360502, "grad_norm": 2.441704034805298, "learning_rate": 9.690823253963178e-06, "loss": 1.529, "step": 5832 }, { "epoch": 0.4571316614420063, "grad_norm": 2.450570583343506, "learning_rate": 9.69071649337208e-06, "loss": 1.5335, "step": 5833 }, { "epoch": 0.4572100313479624, "grad_norm": 3.2695648670196533, "learning_rate": 9.690609714939902e-06, "loss": 1.3761, "step": 5834 }, { "epoch": 0.4572884012539185, "grad_norm": 2.405203104019165, "learning_rate": 9.690502918667056e-06, "loss": 1.9399, "step": 5835 }, { "epoch": 0.4573667711598746, "grad_norm": 2.5397586822509766, "learning_rate": 9.690396104553946e-06, "loss": 1.6556, "step": 5836 }, { "epoch": 0.45744514106583073, "grad_norm": 3.866103410720825, "learning_rate": 9.690289272600979e-06, "loss": 1.1407, "step": 5837 }, { "epoch": 0.45752351097178684, "grad_norm": 2.8372325897216797, "learning_rate": 9.690182422808559e-06, "loss": 1.8284, "step": 5838 }, { "epoch": 0.45760188087774295, "grad_norm": 2.444647789001465, "learning_rate": 9.690075555177094e-06, "loss": 1.2952, "step": 5839 }, { "epoch": 0.45768025078369906, "grad_norm": 3.6022660732269287, "learning_rate": 9.689968669706992e-06, "loss": 1.6859, "step": 5840 }, { "epoch": 0.45775862068965517, "grad_norm": 5.485527038574219, "learning_rate": 9.689861766398658e-06, "loss": 1.7026, "step": 5841 }, { "epoch": 0.4578369905956113, "grad_norm": 3.9276766777038574, "learning_rate": 9.689754845252497e-06, "loss": 1.2242, "step": 5842 }, { "epoch": 0.4579153605015674, "grad_norm": 3.718975305557251, "learning_rate": 9.68964790626892e-06, "loss": 1.7202, "step": 5843 }, { "epoch": 0.4579937304075235, "grad_norm": 5.454185962677002, "learning_rate": 9.689540949448328e-06, "loss": 1.5193, "step": 5844 }, { "epoch": 0.4580721003134796, "grad_norm": 6.831442356109619, "learning_rate": 9.689433974791134e-06, "loss": 1.3121, "step": 5845 }, { "epoch": 0.4581504702194357, "grad_norm": 2.208883285522461, "learning_rate": 9.68932698229774e-06, "loss": 1.325, "step": 5846 }, { "epoch": 0.45822884012539183, "grad_norm": 5.758098125457764, "learning_rate": 9.689219971968556e-06, "loss": 1.0921, "step": 5847 }, { "epoch": 0.45830721003134794, "grad_norm": 2.5561366081237793, "learning_rate": 9.689112943803986e-06, "loss": 1.5268, "step": 5848 }, { "epoch": 0.45838557993730406, "grad_norm": 2.1882827281951904, "learning_rate": 9.68900589780444e-06, "loss": 1.1806, "step": 5849 }, { "epoch": 0.45846394984326017, "grad_norm": 2.5331218242645264, "learning_rate": 9.688898833970325e-06, "loss": 1.7795, "step": 5850 }, { "epoch": 0.4585423197492163, "grad_norm": 3.467938184738159, "learning_rate": 9.688791752302044e-06, "loss": 1.5564, "step": 5851 }, { "epoch": 0.4586206896551724, "grad_norm": 4.8007636070251465, "learning_rate": 9.68868465280001e-06, "loss": 1.5689, "step": 5852 }, { "epoch": 0.45869905956112855, "grad_norm": 2.1114089488983154, "learning_rate": 9.688577535464628e-06, "loss": 1.0525, "step": 5853 }, { "epoch": 0.45877742946708466, "grad_norm": 2.7557618618011475, "learning_rate": 9.688470400296306e-06, "loss": 1.213, "step": 5854 }, { "epoch": 0.4588557993730408, "grad_norm": 2.129241466522217, "learning_rate": 9.688363247295447e-06, "loss": 1.2443, "step": 5855 }, { "epoch": 0.4589341692789969, "grad_norm": 3.2075018882751465, "learning_rate": 9.688256076462465e-06, "loss": 1.5347, "step": 5856 }, { "epoch": 0.459012539184953, "grad_norm": 2.220027208328247, "learning_rate": 9.688148887797766e-06, "loss": 1.3712, "step": 5857 }, { "epoch": 0.4590909090909091, "grad_norm": 5.046515464782715, "learning_rate": 9.688041681301754e-06, "loss": 1.5004, "step": 5858 }, { "epoch": 0.4591692789968652, "grad_norm": 3.5097241401672363, "learning_rate": 9.687934456974841e-06, "loss": 1.5007, "step": 5859 }, { "epoch": 0.4592476489028213, "grad_norm": 4.316099166870117, "learning_rate": 9.687827214817433e-06, "loss": 1.5233, "step": 5860 }, { "epoch": 0.45932601880877744, "grad_norm": 5.341662883758545, "learning_rate": 9.687719954829939e-06, "loss": 1.3874, "step": 5861 }, { "epoch": 0.45940438871473355, "grad_norm": 7.136433124542236, "learning_rate": 9.687612677012765e-06, "loss": 1.4313, "step": 5862 }, { "epoch": 0.45948275862068966, "grad_norm": 2.27791690826416, "learning_rate": 9.687505381366319e-06, "loss": 1.5415, "step": 5863 }, { "epoch": 0.45956112852664577, "grad_norm": 2.272709846496582, "learning_rate": 9.687398067891011e-06, "loss": 1.4267, "step": 5864 }, { "epoch": 0.4596394984326019, "grad_norm": 1.7734366655349731, "learning_rate": 9.68729073658725e-06, "loss": 0.4442, "step": 5865 }, { "epoch": 0.459717868338558, "grad_norm": 2.7746167182922363, "learning_rate": 9.68718338745544e-06, "loss": 1.0841, "step": 5866 }, { "epoch": 0.4597962382445141, "grad_norm": 3.0071449279785156, "learning_rate": 9.687076020495993e-06, "loss": 1.3483, "step": 5867 }, { "epoch": 0.4598746081504702, "grad_norm": 3.7149171829223633, "learning_rate": 9.686968635709316e-06, "loss": 1.4422, "step": 5868 }, { "epoch": 0.4599529780564263, "grad_norm": 3.076258897781372, "learning_rate": 9.686861233095819e-06, "loss": 1.4872, "step": 5869 }, { "epoch": 0.46003134796238243, "grad_norm": 4.318869590759277, "learning_rate": 9.686753812655907e-06, "loss": 1.566, "step": 5870 }, { "epoch": 0.46010971786833854, "grad_norm": 2.5307345390319824, "learning_rate": 9.686646374389991e-06, "loss": 1.4187, "step": 5871 }, { "epoch": 0.46018808777429465, "grad_norm": 2.236599922180176, "learning_rate": 9.686538918298483e-06, "loss": 1.5753, "step": 5872 }, { "epoch": 0.46026645768025076, "grad_norm": 6.564812183380127, "learning_rate": 9.686431444381784e-06, "loss": 1.7525, "step": 5873 }, { "epoch": 0.4603448275862069, "grad_norm": 2.4897990226745605, "learning_rate": 9.686323952640309e-06, "loss": 1.3105, "step": 5874 }, { "epoch": 0.460423197492163, "grad_norm": 2.823951244354248, "learning_rate": 9.686216443074464e-06, "loss": 1.4301, "step": 5875 }, { "epoch": 0.46050156739811915, "grad_norm": 5.421823978424072, "learning_rate": 9.686108915684658e-06, "loss": 1.3008, "step": 5876 }, { "epoch": 0.46057993730407526, "grad_norm": 3.676377534866333, "learning_rate": 9.686001370471302e-06, "loss": 1.4502, "step": 5877 }, { "epoch": 0.46065830721003137, "grad_norm": 3.0482101440429688, "learning_rate": 9.685893807434803e-06, "loss": 1.4915, "step": 5878 }, { "epoch": 0.4607366771159875, "grad_norm": 4.296389579772949, "learning_rate": 9.68578622657557e-06, "loss": 1.7005, "step": 5879 }, { "epoch": 0.4608150470219436, "grad_norm": 3.708281993865967, "learning_rate": 9.685678627894015e-06, "loss": 1.3451, "step": 5880 }, { "epoch": 0.4608934169278997, "grad_norm": 1.9615166187286377, "learning_rate": 9.685571011390543e-06, "loss": 1.3899, "step": 5881 }, { "epoch": 0.4609717868338558, "grad_norm": 2.6278834342956543, "learning_rate": 9.685463377065567e-06, "loss": 1.7823, "step": 5882 }, { "epoch": 0.4610501567398119, "grad_norm": 4.2588934898376465, "learning_rate": 9.685355724919494e-06, "loss": 1.097, "step": 5883 }, { "epoch": 0.46112852664576803, "grad_norm": 3.025315523147583, "learning_rate": 9.685248054952734e-06, "loss": 1.8808, "step": 5884 }, { "epoch": 0.46120689655172414, "grad_norm": 4.584741115570068, "learning_rate": 9.685140367165699e-06, "loss": 1.5211, "step": 5885 }, { "epoch": 0.46128526645768025, "grad_norm": 2.731405019760132, "learning_rate": 9.685032661558795e-06, "loss": 1.2937, "step": 5886 }, { "epoch": 0.46136363636363636, "grad_norm": 2.6246068477630615, "learning_rate": 9.684924938132433e-06, "loss": 1.2482, "step": 5887 }, { "epoch": 0.4614420062695925, "grad_norm": 5.029036998748779, "learning_rate": 9.684817196887024e-06, "loss": 1.3674, "step": 5888 }, { "epoch": 0.4615203761755486, "grad_norm": 3.085573434829712, "learning_rate": 9.684709437822976e-06, "loss": 1.3955, "step": 5889 }, { "epoch": 0.4615987460815047, "grad_norm": 3.2853403091430664, "learning_rate": 9.6846016609407e-06, "loss": 1.799, "step": 5890 }, { "epoch": 0.4616771159874608, "grad_norm": 3.2454311847686768, "learning_rate": 9.684493866240603e-06, "loss": 1.2457, "step": 5891 }, { "epoch": 0.4617554858934169, "grad_norm": 2.566605567932129, "learning_rate": 9.6843860537231e-06, "loss": 1.5549, "step": 5892 }, { "epoch": 0.461833855799373, "grad_norm": 3.7444872856140137, "learning_rate": 9.684278223388598e-06, "loss": 1.6624, "step": 5893 }, { "epoch": 0.46191222570532914, "grad_norm": 2.1077895164489746, "learning_rate": 9.684170375237507e-06, "loss": 1.1089, "step": 5894 }, { "epoch": 0.46199059561128525, "grad_norm": 2.9868967533111572, "learning_rate": 9.684062509270239e-06, "loss": 1.1192, "step": 5895 }, { "epoch": 0.46206896551724136, "grad_norm": 2.7109975814819336, "learning_rate": 9.683954625487202e-06, "loss": 1.228, "step": 5896 }, { "epoch": 0.46214733542319747, "grad_norm": 3.4643750190734863, "learning_rate": 9.683846723888808e-06, "loss": 1.4517, "step": 5897 }, { "epoch": 0.4622257053291536, "grad_norm": 3.1464619636535645, "learning_rate": 9.683738804475467e-06, "loss": 1.2741, "step": 5898 }, { "epoch": 0.4623040752351097, "grad_norm": 2.7987558841705322, "learning_rate": 9.683630867247587e-06, "loss": 1.362, "step": 5899 }, { "epoch": 0.46238244514106586, "grad_norm": 2.865962266921997, "learning_rate": 9.683522912205583e-06, "loss": 1.3578, "step": 5900 }, { "epoch": 0.46246081504702197, "grad_norm": 5.11457633972168, "learning_rate": 9.683414939349865e-06, "loss": 1.7872, "step": 5901 }, { "epoch": 0.4625391849529781, "grad_norm": 4.691833972930908, "learning_rate": 9.68330694868084e-06, "loss": 1.9542, "step": 5902 }, { "epoch": 0.4626175548589342, "grad_norm": 6.067395210266113, "learning_rate": 9.68319894019892e-06, "loss": 1.4534, "step": 5903 }, { "epoch": 0.4626959247648903, "grad_norm": 2.7201876640319824, "learning_rate": 9.683090913904518e-06, "loss": 1.5014, "step": 5904 }, { "epoch": 0.4627742946708464, "grad_norm": 3.065351724624634, "learning_rate": 9.682982869798043e-06, "loss": 1.1585, "step": 5905 }, { "epoch": 0.4628526645768025, "grad_norm": 3.39687442779541, "learning_rate": 9.682874807879907e-06, "loss": 1.9257, "step": 5906 }, { "epoch": 0.46293103448275863, "grad_norm": 2.917111396789551, "learning_rate": 9.682766728150521e-06, "loss": 1.4511, "step": 5907 }, { "epoch": 0.46300940438871474, "grad_norm": 2.4506654739379883, "learning_rate": 9.682658630610296e-06, "loss": 1.4326, "step": 5908 }, { "epoch": 0.46308777429467085, "grad_norm": 8.339340209960938, "learning_rate": 9.682550515259641e-06, "loss": 1.9463, "step": 5909 }, { "epoch": 0.46316614420062696, "grad_norm": 8.776100158691406, "learning_rate": 9.682442382098969e-06, "loss": 1.1422, "step": 5910 }, { "epoch": 0.46324451410658307, "grad_norm": 3.061701536178589, "learning_rate": 9.682334231128691e-06, "loss": 1.7002, "step": 5911 }, { "epoch": 0.4633228840125392, "grad_norm": 2.609221935272217, "learning_rate": 9.68222606234922e-06, "loss": 1.5468, "step": 5912 }, { "epoch": 0.4634012539184953, "grad_norm": 2.7858664989471436, "learning_rate": 9.682117875760966e-06, "loss": 1.3068, "step": 5913 }, { "epoch": 0.4634796238244514, "grad_norm": 4.201392650604248, "learning_rate": 9.682009671364341e-06, "loss": 1.6752, "step": 5914 }, { "epoch": 0.4635579937304075, "grad_norm": 4.175846099853516, "learning_rate": 9.681901449159755e-06, "loss": 1.8318, "step": 5915 }, { "epoch": 0.4636363636363636, "grad_norm": 2.810511350631714, "learning_rate": 9.68179320914762e-06, "loss": 1.3142, "step": 5916 }, { "epoch": 0.46371473354231973, "grad_norm": 2.626868724822998, "learning_rate": 9.68168495132835e-06, "loss": 1.1109, "step": 5917 }, { "epoch": 0.46379310344827585, "grad_norm": 3.563502311706543, "learning_rate": 9.681576675702353e-06, "loss": 0.8543, "step": 5918 }, { "epoch": 0.46387147335423196, "grad_norm": 3.4087138175964355, "learning_rate": 9.681468382270045e-06, "loss": 1.7551, "step": 5919 }, { "epoch": 0.46394984326018807, "grad_norm": 2.7125086784362793, "learning_rate": 9.681360071031835e-06, "loss": 1.3941, "step": 5920 }, { "epoch": 0.4640282131661442, "grad_norm": 2.4125404357910156, "learning_rate": 9.681251741988135e-06, "loss": 1.6031, "step": 5921 }, { "epoch": 0.4641065830721003, "grad_norm": 4.942070484161377, "learning_rate": 9.68114339513936e-06, "loss": 1.5437, "step": 5922 }, { "epoch": 0.46418495297805645, "grad_norm": 2.7201950550079346, "learning_rate": 9.681035030485917e-06, "loss": 1.3983, "step": 5923 }, { "epoch": 0.46426332288401256, "grad_norm": 15.315901756286621, "learning_rate": 9.680926648028223e-06, "loss": 1.1237, "step": 5924 }, { "epoch": 0.4643416927899687, "grad_norm": 3.199871778488159, "learning_rate": 9.680818247766688e-06, "loss": 1.6374, "step": 5925 }, { "epoch": 0.4644200626959248, "grad_norm": 2.4436235427856445, "learning_rate": 9.680709829701725e-06, "loss": 1.6956, "step": 5926 }, { "epoch": 0.4644984326018809, "grad_norm": 3.663128137588501, "learning_rate": 9.680601393833743e-06, "loss": 1.4768, "step": 5927 }, { "epoch": 0.464576802507837, "grad_norm": 5.170372486114502, "learning_rate": 9.680492940163162e-06, "loss": 1.2139, "step": 5928 }, { "epoch": 0.4646551724137931, "grad_norm": 2.498614549636841, "learning_rate": 9.680384468690387e-06, "loss": 1.2683, "step": 5929 }, { "epoch": 0.4647335423197492, "grad_norm": 2.937535047531128, "learning_rate": 9.680275979415834e-06, "loss": 1.4799, "step": 5930 }, { "epoch": 0.46481191222570534, "grad_norm": 3.3301162719726562, "learning_rate": 9.680167472339916e-06, "loss": 1.4917, "step": 5931 }, { "epoch": 0.46489028213166145, "grad_norm": 2.608844757080078, "learning_rate": 9.680058947463044e-06, "loss": 1.5089, "step": 5932 }, { "epoch": 0.46496865203761756, "grad_norm": 3.0922024250030518, "learning_rate": 9.679950404785631e-06, "loss": 1.215, "step": 5933 }, { "epoch": 0.46504702194357367, "grad_norm": 3.213773488998413, "learning_rate": 9.67984184430809e-06, "loss": 1.7274, "step": 5934 }, { "epoch": 0.4651253918495298, "grad_norm": 2.6577115058898926, "learning_rate": 9.679733266030838e-06, "loss": 1.1179, "step": 5935 }, { "epoch": 0.4652037617554859, "grad_norm": 2.069702625274658, "learning_rate": 9.679624669954281e-06, "loss": 1.3777, "step": 5936 }, { "epoch": 0.465282131661442, "grad_norm": 2.5652613639831543, "learning_rate": 9.679516056078836e-06, "loss": 1.7507, "step": 5937 }, { "epoch": 0.4653605015673981, "grad_norm": 2.0707054138183594, "learning_rate": 9.679407424404916e-06, "loss": 1.5569, "step": 5938 }, { "epoch": 0.4654388714733542, "grad_norm": 5.3939290046691895, "learning_rate": 9.679298774932932e-06, "loss": 1.4117, "step": 5939 }, { "epoch": 0.46551724137931033, "grad_norm": 2.6989784240722656, "learning_rate": 9.679190107663302e-06, "loss": 1.426, "step": 5940 }, { "epoch": 0.46559561128526644, "grad_norm": 3.2018284797668457, "learning_rate": 9.679081422596433e-06, "loss": 1.6366, "step": 5941 }, { "epoch": 0.46567398119122255, "grad_norm": 3.303938627243042, "learning_rate": 9.678972719732745e-06, "loss": 1.4247, "step": 5942 }, { "epoch": 0.46575235109717866, "grad_norm": 2.6145551204681396, "learning_rate": 9.678863999072646e-06, "loss": 1.4375, "step": 5943 }, { "epoch": 0.4658307210031348, "grad_norm": 3.7844157218933105, "learning_rate": 9.67875526061655e-06, "loss": 1.5458, "step": 5944 }, { "epoch": 0.4659090909090909, "grad_norm": 2.128178834915161, "learning_rate": 9.678646504364874e-06, "loss": 0.9598, "step": 5945 }, { "epoch": 0.465987460815047, "grad_norm": 4.535811901092529, "learning_rate": 9.67853773031803e-06, "loss": 1.0965, "step": 5946 }, { "epoch": 0.46606583072100316, "grad_norm": 5.375667572021484, "learning_rate": 9.678428938476432e-06, "loss": 1.4677, "step": 5947 }, { "epoch": 0.46614420062695927, "grad_norm": 2.1646764278411865, "learning_rate": 9.67832012884049e-06, "loss": 1.4483, "step": 5948 }, { "epoch": 0.4662225705329154, "grad_norm": 2.1172945499420166, "learning_rate": 9.678211301410625e-06, "loss": 1.4398, "step": 5949 }, { "epoch": 0.4663009404388715, "grad_norm": 3.0283255577087402, "learning_rate": 9.678102456187246e-06, "loss": 1.1551, "step": 5950 }, { "epoch": 0.4663793103448276, "grad_norm": 3.922769784927368, "learning_rate": 9.677993593170766e-06, "loss": 1.0997, "step": 5951 }, { "epoch": 0.4664576802507837, "grad_norm": 2.8623569011688232, "learning_rate": 9.677884712361606e-06, "loss": 1.3723, "step": 5952 }, { "epoch": 0.4665360501567398, "grad_norm": 4.301238536834717, "learning_rate": 9.67777581376017e-06, "loss": 1.4336, "step": 5953 }, { "epoch": 0.46661442006269593, "grad_norm": 3.1050543785095215, "learning_rate": 9.67766689736688e-06, "loss": 1.4329, "step": 5954 }, { "epoch": 0.46669278996865204, "grad_norm": 4.143998146057129, "learning_rate": 9.677557963182146e-06, "loss": 1.4981, "step": 5955 }, { "epoch": 0.46677115987460815, "grad_norm": 2.05660343170166, "learning_rate": 9.677449011206387e-06, "loss": 1.3464, "step": 5956 }, { "epoch": 0.46684952978056427, "grad_norm": 4.181324481964111, "learning_rate": 9.677340041440014e-06, "loss": 1.578, "step": 5957 }, { "epoch": 0.4669278996865204, "grad_norm": 6.45393180847168, "learning_rate": 9.67723105388344e-06, "loss": 0.9953, "step": 5958 }, { "epoch": 0.4670062695924765, "grad_norm": 2.16477370262146, "learning_rate": 9.677122048537083e-06, "loss": 1.3249, "step": 5959 }, { "epoch": 0.4670846394984326, "grad_norm": 2.309847354888916, "learning_rate": 9.677013025401354e-06, "loss": 1.3642, "step": 5960 }, { "epoch": 0.4671630094043887, "grad_norm": 4.739735126495361, "learning_rate": 9.676903984476671e-06, "loss": 1.2142, "step": 5961 }, { "epoch": 0.4672413793103448, "grad_norm": 2.3033432960510254, "learning_rate": 9.676794925763449e-06, "loss": 0.9218, "step": 5962 }, { "epoch": 0.46731974921630093, "grad_norm": 3.6204147338867188, "learning_rate": 9.676685849262098e-06, "loss": 1.3113, "step": 5963 }, { "epoch": 0.46739811912225704, "grad_norm": 3.8065059185028076, "learning_rate": 9.676576754973038e-06, "loss": 1.2995, "step": 5964 }, { "epoch": 0.46747648902821315, "grad_norm": 2.9423954486846924, "learning_rate": 9.676467642896682e-06, "loss": 1.6833, "step": 5965 }, { "epoch": 0.46755485893416926, "grad_norm": 2.658259153366089, "learning_rate": 9.676358513033446e-06, "loss": 1.284, "step": 5966 }, { "epoch": 0.46763322884012537, "grad_norm": 3.203914165496826, "learning_rate": 9.676249365383743e-06, "loss": 1.4725, "step": 5967 }, { "epoch": 0.4677115987460815, "grad_norm": 2.691716194152832, "learning_rate": 9.676140199947988e-06, "loss": 1.4398, "step": 5968 }, { "epoch": 0.4677899686520376, "grad_norm": 1.850799798965454, "learning_rate": 9.676031016726598e-06, "loss": 1.397, "step": 5969 }, { "epoch": 0.46786833855799376, "grad_norm": 2.9101696014404297, "learning_rate": 9.675921815719992e-06, "loss": 1.3528, "step": 5970 }, { "epoch": 0.46794670846394987, "grad_norm": 2.2101693153381348, "learning_rate": 9.675812596928576e-06, "loss": 1.49, "step": 5971 }, { "epoch": 0.468025078369906, "grad_norm": 1.9874998331069946, "learning_rate": 9.675703360352772e-06, "loss": 1.1076, "step": 5972 }, { "epoch": 0.4681034482758621, "grad_norm": 2.6904053688049316, "learning_rate": 9.675594105992995e-06, "loss": 1.4729, "step": 5973 }, { "epoch": 0.4681818181818182, "grad_norm": 10.781261444091797, "learning_rate": 9.675484833849659e-06, "loss": 2.4359, "step": 5974 }, { "epoch": 0.4682601880877743, "grad_norm": 3.9010303020477295, "learning_rate": 9.67537554392318e-06, "loss": 1.4966, "step": 5975 }, { "epoch": 0.4683385579937304, "grad_norm": 2.4295458793640137, "learning_rate": 9.675266236213974e-06, "loss": 0.9664, "step": 5976 }, { "epoch": 0.46841692789968653, "grad_norm": 4.783323764801025, "learning_rate": 9.675156910722457e-06, "loss": 1.3744, "step": 5977 }, { "epoch": 0.46849529780564264, "grad_norm": 3.4256908893585205, "learning_rate": 9.675047567449043e-06, "loss": 1.4872, "step": 5978 }, { "epoch": 0.46857366771159875, "grad_norm": 2.332094669342041, "learning_rate": 9.674938206394152e-06, "loss": 1.4071, "step": 5979 }, { "epoch": 0.46865203761755486, "grad_norm": 2.636810779571533, "learning_rate": 9.674828827558194e-06, "loss": 1.6569, "step": 5980 }, { "epoch": 0.468730407523511, "grad_norm": 3.543907880783081, "learning_rate": 9.674719430941591e-06, "loss": 1.4682, "step": 5981 }, { "epoch": 0.4688087774294671, "grad_norm": 2.8498644828796387, "learning_rate": 9.674610016544754e-06, "loss": 1.5765, "step": 5982 }, { "epoch": 0.4688871473354232, "grad_norm": 3.1719183921813965, "learning_rate": 9.674500584368103e-06, "loss": 1.1894, "step": 5983 }, { "epoch": 0.4689655172413793, "grad_norm": 4.022519111633301, "learning_rate": 9.674391134412052e-06, "loss": 1.1521, "step": 5984 }, { "epoch": 0.4690438871473354, "grad_norm": 2.2418806552886963, "learning_rate": 9.674281666677018e-06, "loss": 1.1235, "step": 5985 }, { "epoch": 0.4691222570532915, "grad_norm": 3.680119037628174, "learning_rate": 9.67417218116342e-06, "loss": 1.5279, "step": 5986 }, { "epoch": 0.46920062695924764, "grad_norm": 2.82004714012146, "learning_rate": 9.67406267787167e-06, "loss": 1.3946, "step": 5987 }, { "epoch": 0.46927899686520375, "grad_norm": 9.052974700927734, "learning_rate": 9.673953156802184e-06, "loss": 1.4069, "step": 5988 }, { "epoch": 0.46935736677115986, "grad_norm": 5.227200031280518, "learning_rate": 9.673843617955383e-06, "loss": 1.4699, "step": 5989 }, { "epoch": 0.46943573667711597, "grad_norm": 3.0077998638153076, "learning_rate": 9.67373406133168e-06, "loss": 1.3892, "step": 5990 }, { "epoch": 0.4695141065830721, "grad_norm": 3.735565185546875, "learning_rate": 9.673624486931497e-06, "loss": 1.6674, "step": 5991 }, { "epoch": 0.4695924764890282, "grad_norm": 3.0479896068573, "learning_rate": 9.673514894755244e-06, "loss": 1.9629, "step": 5992 }, { "epoch": 0.46967084639498435, "grad_norm": 3.5799198150634766, "learning_rate": 9.673405284803341e-06, "loss": 1.5027, "step": 5993 }, { "epoch": 0.46974921630094046, "grad_norm": 2.717374563217163, "learning_rate": 9.673295657076204e-06, "loss": 1.4103, "step": 5994 }, { "epoch": 0.4698275862068966, "grad_norm": 2.908381938934326, "learning_rate": 9.673186011574254e-06, "loss": 1.6011, "step": 5995 }, { "epoch": 0.4699059561128527, "grad_norm": 4.196772575378418, "learning_rate": 9.673076348297901e-06, "loss": 1.7782, "step": 5996 }, { "epoch": 0.4699843260188088, "grad_norm": 4.242795467376709, "learning_rate": 9.672966667247569e-06, "loss": 1.3928, "step": 5997 }, { "epoch": 0.4700626959247649, "grad_norm": 2.401777982711792, "learning_rate": 9.67285696842367e-06, "loss": 1.3453, "step": 5998 }, { "epoch": 0.470141065830721, "grad_norm": 2.7357285022735596, "learning_rate": 9.672747251826624e-06, "loss": 1.7611, "step": 5999 }, { "epoch": 0.4702194357366771, "grad_norm": 4.085971355438232, "learning_rate": 9.672637517456846e-06, "loss": 1.3165, "step": 6000 }, { "epoch": 0.47029780564263324, "grad_norm": 2.933300495147705, "learning_rate": 9.672527765314756e-06, "loss": 1.3129, "step": 6001 }, { "epoch": 0.47037617554858935, "grad_norm": 2.996065139770508, "learning_rate": 9.672417995400772e-06, "loss": 1.2847, "step": 6002 }, { "epoch": 0.47045454545454546, "grad_norm": 3.6047332286834717, "learning_rate": 9.672308207715307e-06, "loss": 1.0854, "step": 6003 }, { "epoch": 0.47053291536050157, "grad_norm": 3.442213296890259, "learning_rate": 9.672198402258784e-06, "loss": 1.4691, "step": 6004 }, { "epoch": 0.4706112852664577, "grad_norm": 3.7218120098114014, "learning_rate": 9.672088579031615e-06, "loss": 1.6294, "step": 6005 }, { "epoch": 0.4706896551724138, "grad_norm": 11.30760383605957, "learning_rate": 9.671978738034221e-06, "loss": 1.6291, "step": 6006 }, { "epoch": 0.4707680250783699, "grad_norm": 2.3594300746917725, "learning_rate": 9.671868879267024e-06, "loss": 1.264, "step": 6007 }, { "epoch": 0.470846394984326, "grad_norm": 2.526789903640747, "learning_rate": 9.671759002730433e-06, "loss": 1.5686, "step": 6008 }, { "epoch": 0.4709247648902821, "grad_norm": 2.381828784942627, "learning_rate": 9.671649108424872e-06, "loss": 1.3055, "step": 6009 }, { "epoch": 0.47100313479623823, "grad_norm": 5.249819278717041, "learning_rate": 9.671539196350757e-06, "loss": 1.6184, "step": 6010 }, { "epoch": 0.47108150470219434, "grad_norm": 3.463315963745117, "learning_rate": 9.671429266508508e-06, "loss": 1.7528, "step": 6011 }, { "epoch": 0.47115987460815045, "grad_norm": 2.7500035762786865, "learning_rate": 9.671319318898539e-06, "loss": 1.5143, "step": 6012 }, { "epoch": 0.47123824451410656, "grad_norm": 4.038547992706299, "learning_rate": 9.671209353521272e-06, "loss": 1.6675, "step": 6013 }, { "epoch": 0.4713166144200627, "grad_norm": 2.9752581119537354, "learning_rate": 9.671099370377123e-06, "loss": 1.5642, "step": 6014 }, { "epoch": 0.4713949843260188, "grad_norm": 3.3788535594940186, "learning_rate": 9.670989369466511e-06, "loss": 1.6657, "step": 6015 }, { "epoch": 0.4714733542319749, "grad_norm": 2.4986798763275146, "learning_rate": 9.670879350789856e-06, "loss": 0.9173, "step": 6016 }, { "epoch": 0.47155172413793106, "grad_norm": 7.363916397094727, "learning_rate": 9.670769314347575e-06, "loss": 1.3618, "step": 6017 }, { "epoch": 0.47163009404388717, "grad_norm": 3.6105077266693115, "learning_rate": 9.670659260140084e-06, "loss": 1.9012, "step": 6018 }, { "epoch": 0.4717084639498433, "grad_norm": 2.271796226501465, "learning_rate": 9.670549188167808e-06, "loss": 1.4092, "step": 6019 }, { "epoch": 0.4717868338557994, "grad_norm": 2.88382887840271, "learning_rate": 9.67043909843116e-06, "loss": 1.4343, "step": 6020 }, { "epoch": 0.4718652037617555, "grad_norm": 2.4701616764068604, "learning_rate": 9.67032899093056e-06, "loss": 1.6518, "step": 6021 }, { "epoch": 0.4719435736677116, "grad_norm": 3.4590179920196533, "learning_rate": 9.670218865666427e-06, "loss": 1.5456, "step": 6022 }, { "epoch": 0.4720219435736677, "grad_norm": 3.485292673110962, "learning_rate": 9.670108722639182e-06, "loss": 1.532, "step": 6023 }, { "epoch": 0.47210031347962383, "grad_norm": 2.346900701522827, "learning_rate": 9.669998561849241e-06, "loss": 1.4794, "step": 6024 }, { "epoch": 0.47217868338557994, "grad_norm": 3.117260217666626, "learning_rate": 9.669888383297024e-06, "loss": 1.3937, "step": 6025 }, { "epoch": 0.47225705329153606, "grad_norm": 4.447678089141846, "learning_rate": 9.66977818698295e-06, "loss": 1.7528, "step": 6026 }, { "epoch": 0.47233542319749217, "grad_norm": 2.5443289279937744, "learning_rate": 9.66966797290744e-06, "loss": 1.6305, "step": 6027 }, { "epoch": 0.4724137931034483, "grad_norm": 3.793599843978882, "learning_rate": 9.66955774107091e-06, "loss": 1.832, "step": 6028 }, { "epoch": 0.4724921630094044, "grad_norm": 2.1616318225860596, "learning_rate": 9.66944749147378e-06, "loss": 1.4703, "step": 6029 }, { "epoch": 0.4725705329153605, "grad_norm": 3.004257917404175, "learning_rate": 9.669337224116473e-06, "loss": 1.6804, "step": 6030 }, { "epoch": 0.4726489028213166, "grad_norm": 2.4181690216064453, "learning_rate": 9.669226938999404e-06, "loss": 1.4294, "step": 6031 }, { "epoch": 0.4727272727272727, "grad_norm": 8.048446655273438, "learning_rate": 9.669116636122993e-06, "loss": 1.5171, "step": 6032 }, { "epoch": 0.47280564263322883, "grad_norm": 3.262220859527588, "learning_rate": 9.669006315487662e-06, "loss": 1.0803, "step": 6033 }, { "epoch": 0.47288401253918494, "grad_norm": 1.9931154251098633, "learning_rate": 9.668895977093828e-06, "loss": 1.4495, "step": 6034 }, { "epoch": 0.47296238244514105, "grad_norm": 3.184957265853882, "learning_rate": 9.668785620941912e-06, "loss": 1.403, "step": 6035 }, { "epoch": 0.47304075235109716, "grad_norm": 5.3210062980651855, "learning_rate": 9.668675247032334e-06, "loss": 2.0328, "step": 6036 }, { "epoch": 0.47311912225705327, "grad_norm": 3.6215624809265137, "learning_rate": 9.668564855365513e-06, "loss": 1.3574, "step": 6037 }, { "epoch": 0.4731974921630094, "grad_norm": 6.4771294593811035, "learning_rate": 9.668454445941869e-06, "loss": 1.3978, "step": 6038 }, { "epoch": 0.4732758620689655, "grad_norm": 3.776904582977295, "learning_rate": 9.668344018761823e-06, "loss": 1.5473, "step": 6039 }, { "epoch": 0.47335423197492166, "grad_norm": 3.4867372512817383, "learning_rate": 9.668233573825794e-06, "loss": 1.3389, "step": 6040 }, { "epoch": 0.47343260188087777, "grad_norm": 3.131692886352539, "learning_rate": 9.668123111134202e-06, "loss": 1.6183, "step": 6041 }, { "epoch": 0.4735109717868339, "grad_norm": 4.855672359466553, "learning_rate": 9.668012630687467e-06, "loss": 1.7452, "step": 6042 }, { "epoch": 0.47358934169279, "grad_norm": 5.142475605010986, "learning_rate": 9.667902132486009e-06, "loss": 1.6818, "step": 6043 }, { "epoch": 0.4736677115987461, "grad_norm": 2.8584513664245605, "learning_rate": 9.66779161653025e-06, "loss": 1.3281, "step": 6044 }, { "epoch": 0.4737460815047022, "grad_norm": 6.287015438079834, "learning_rate": 9.667681082820608e-06, "loss": 1.3775, "step": 6045 }, { "epoch": 0.4738244514106583, "grad_norm": 3.092085599899292, "learning_rate": 9.667570531357507e-06, "loss": 1.7039, "step": 6046 }, { "epoch": 0.47390282131661443, "grad_norm": 2.5031678676605225, "learning_rate": 9.667459962141363e-06, "loss": 1.375, "step": 6047 }, { "epoch": 0.47398119122257054, "grad_norm": 3.198296308517456, "learning_rate": 9.6673493751726e-06, "loss": 1.41, "step": 6048 }, { "epoch": 0.47405956112852665, "grad_norm": 2.9216439723968506, "learning_rate": 9.667238770451636e-06, "loss": 1.724, "step": 6049 }, { "epoch": 0.47413793103448276, "grad_norm": 4.110267639160156, "learning_rate": 9.667128147978895e-06, "loss": 1.6203, "step": 6050 }, { "epoch": 0.4742163009404389, "grad_norm": 3.674999237060547, "learning_rate": 9.667017507754795e-06, "loss": 1.0958, "step": 6051 }, { "epoch": 0.474294670846395, "grad_norm": 2.791855812072754, "learning_rate": 9.666906849779755e-06, "loss": 1.5726, "step": 6052 }, { "epoch": 0.4743730407523511, "grad_norm": 2.988194704055786, "learning_rate": 9.666796174054201e-06, "loss": 1.3461, "step": 6053 }, { "epoch": 0.4744514106583072, "grad_norm": 6.172515869140625, "learning_rate": 9.66668548057855e-06, "loss": 1.6658, "step": 6054 }, { "epoch": 0.4745297805642633, "grad_norm": 3.1631157398223877, "learning_rate": 9.666574769353225e-06, "loss": 1.3846, "step": 6055 }, { "epoch": 0.4746081504702194, "grad_norm": 2.898787260055542, "learning_rate": 9.666464040378646e-06, "loss": 1.6628, "step": 6056 }, { "epoch": 0.47468652037617554, "grad_norm": 4.67030143737793, "learning_rate": 9.666353293655238e-06, "loss": 1.3216, "step": 6057 }, { "epoch": 0.47476489028213165, "grad_norm": 3.209300994873047, "learning_rate": 9.666242529183414e-06, "loss": 1.3089, "step": 6058 }, { "epoch": 0.47484326018808776, "grad_norm": 3.4144351482391357, "learning_rate": 9.666131746963603e-06, "loss": 1.5988, "step": 6059 }, { "epoch": 0.47492163009404387, "grad_norm": 3.0492353439331055, "learning_rate": 9.666020946996223e-06, "loss": 1.3788, "step": 6060 }, { "epoch": 0.475, "grad_norm": 4.523740768432617, "learning_rate": 9.665910129281694e-06, "loss": 1.8753, "step": 6061 }, { "epoch": 0.4750783699059561, "grad_norm": 2.9667582511901855, "learning_rate": 9.665799293820443e-06, "loss": 1.6342, "step": 6062 }, { "epoch": 0.4751567398119122, "grad_norm": 5.009360313415527, "learning_rate": 9.665688440612885e-06, "loss": 1.2544, "step": 6063 }, { "epoch": 0.47523510971786836, "grad_norm": 2.8230903148651123, "learning_rate": 9.665577569659445e-06, "loss": 1.1406, "step": 6064 }, { "epoch": 0.4753134796238245, "grad_norm": 3.84295916557312, "learning_rate": 9.665466680960547e-06, "loss": 1.8827, "step": 6065 }, { "epoch": 0.4753918495297806, "grad_norm": 1.9058178663253784, "learning_rate": 9.665355774516608e-06, "loss": 1.4121, "step": 6066 }, { "epoch": 0.4754702194357367, "grad_norm": 3.776287794113159, "learning_rate": 9.665244850328052e-06, "loss": 1.5671, "step": 6067 }, { "epoch": 0.4755485893416928, "grad_norm": 2.583648920059204, "learning_rate": 9.6651339083953e-06, "loss": 1.2969, "step": 6068 }, { "epoch": 0.4756269592476489, "grad_norm": 2.8941259384155273, "learning_rate": 9.665022948718777e-06, "loss": 1.4118, "step": 6069 }, { "epoch": 0.475705329153605, "grad_norm": 3.4999799728393555, "learning_rate": 9.664911971298901e-06, "loss": 1.6448, "step": 6070 }, { "epoch": 0.47578369905956114, "grad_norm": 2.4261343479156494, "learning_rate": 9.664800976136096e-06, "loss": 1.2957, "step": 6071 }, { "epoch": 0.47586206896551725, "grad_norm": 3.440197706222534, "learning_rate": 9.664689963230784e-06, "loss": 1.7206, "step": 6072 }, { "epoch": 0.47594043887147336, "grad_norm": 2.7128844261169434, "learning_rate": 9.664578932583389e-06, "loss": 0.6473, "step": 6073 }, { "epoch": 0.47601880877742947, "grad_norm": 2.982503652572632, "learning_rate": 9.66446788419433e-06, "loss": 1.598, "step": 6074 }, { "epoch": 0.4760971786833856, "grad_norm": 2.671112060546875, "learning_rate": 9.66435681806403e-06, "loss": 1.4802, "step": 6075 }, { "epoch": 0.4761755485893417, "grad_norm": 3.4170031547546387, "learning_rate": 9.664245734192915e-06, "loss": 1.9046, "step": 6076 }, { "epoch": 0.4762539184952978, "grad_norm": 1.9360109567642212, "learning_rate": 9.664134632581404e-06, "loss": 1.2647, "step": 6077 }, { "epoch": 0.4763322884012539, "grad_norm": 2.0614922046661377, "learning_rate": 9.664023513229918e-06, "loss": 1.4664, "step": 6078 }, { "epoch": 0.47641065830721, "grad_norm": 2.465991258621216, "learning_rate": 9.663912376138885e-06, "loss": 1.1428, "step": 6079 }, { "epoch": 0.47648902821316613, "grad_norm": 2.572397232055664, "learning_rate": 9.663801221308725e-06, "loss": 1.4144, "step": 6080 }, { "epoch": 0.47656739811912224, "grad_norm": 2.4673240184783936, "learning_rate": 9.66369004873986e-06, "loss": 1.2812, "step": 6081 }, { "epoch": 0.47664576802507835, "grad_norm": 3.4785826206207275, "learning_rate": 9.663578858432713e-06, "loss": 1.5623, "step": 6082 }, { "epoch": 0.47672413793103446, "grad_norm": 3.0528180599212646, "learning_rate": 9.663467650387707e-06, "loss": 1.4622, "step": 6083 }, { "epoch": 0.4768025078369906, "grad_norm": 3.4016916751861572, "learning_rate": 9.663356424605266e-06, "loss": 1.4332, "step": 6084 }, { "epoch": 0.4768808777429467, "grad_norm": 2.325760841369629, "learning_rate": 9.663245181085812e-06, "loss": 1.474, "step": 6085 }, { "epoch": 0.4769592476489028, "grad_norm": 2.009296417236328, "learning_rate": 9.66313391982977e-06, "loss": 0.9222, "step": 6086 }, { "epoch": 0.47703761755485896, "grad_norm": 2.6653218269348145, "learning_rate": 9.663022640837561e-06, "loss": 1.5858, "step": 6087 }, { "epoch": 0.47711598746081507, "grad_norm": 5.952919960021973, "learning_rate": 9.662911344109609e-06, "loss": 2.0371, "step": 6088 }, { "epoch": 0.4771943573667712, "grad_norm": 2.4178640842437744, "learning_rate": 9.662800029646336e-06, "loss": 1.4179, "step": 6089 }, { "epoch": 0.4772727272727273, "grad_norm": 2.717711925506592, "learning_rate": 9.662688697448167e-06, "loss": 1.5021, "step": 6090 }, { "epoch": 0.4773510971786834, "grad_norm": 2.0387866497039795, "learning_rate": 9.662577347515528e-06, "loss": 1.4703, "step": 6091 }, { "epoch": 0.4774294670846395, "grad_norm": 4.4183735847473145, "learning_rate": 9.662465979848837e-06, "loss": 1.512, "step": 6092 }, { "epoch": 0.4775078369905956, "grad_norm": 2.8706209659576416, "learning_rate": 9.66235459444852e-06, "loss": 1.4571, "step": 6093 }, { "epoch": 0.47758620689655173, "grad_norm": 6.6806817054748535, "learning_rate": 9.662243191315001e-06, "loss": 1.6638, "step": 6094 }, { "epoch": 0.47766457680250785, "grad_norm": 3.2573890686035156, "learning_rate": 9.662131770448704e-06, "loss": 1.5123, "step": 6095 }, { "epoch": 0.47774294670846396, "grad_norm": 2.9429399967193604, "learning_rate": 9.662020331850053e-06, "loss": 1.2204, "step": 6096 }, { "epoch": 0.47782131661442007, "grad_norm": 3.30595064163208, "learning_rate": 9.66190887551947e-06, "loss": 1.402, "step": 6097 }, { "epoch": 0.4778996865203762, "grad_norm": 4.441981792449951, "learning_rate": 9.66179740145738e-06, "loss": 1.5845, "step": 6098 }, { "epoch": 0.4779780564263323, "grad_norm": 3.1477997303009033, "learning_rate": 9.661685909664208e-06, "loss": 1.0864, "step": 6099 }, { "epoch": 0.4780564263322884, "grad_norm": 3.119137763977051, "learning_rate": 9.661574400140378e-06, "loss": 1.784, "step": 6100 }, { "epoch": 0.4781347962382445, "grad_norm": 2.844358205795288, "learning_rate": 9.661462872886312e-06, "loss": 1.7693, "step": 6101 }, { "epoch": 0.4782131661442006, "grad_norm": 3.7021822929382324, "learning_rate": 9.661351327902435e-06, "loss": 1.5345, "step": 6102 }, { "epoch": 0.47829153605015673, "grad_norm": 5.566568374633789, "learning_rate": 9.661239765189173e-06, "loss": 1.4979, "step": 6103 }, { "epoch": 0.47836990595611284, "grad_norm": 7.543546199798584, "learning_rate": 9.661128184746949e-06, "loss": 1.3457, "step": 6104 }, { "epoch": 0.47844827586206895, "grad_norm": 3.0902884006500244, "learning_rate": 9.661016586576188e-06, "loss": 1.6252, "step": 6105 }, { "epoch": 0.47852664576802506, "grad_norm": 3.0317788124084473, "learning_rate": 9.660904970677311e-06, "loss": 1.4328, "step": 6106 }, { "epoch": 0.47860501567398117, "grad_norm": 2.797410488128662, "learning_rate": 9.660793337050749e-06, "loss": 1.6031, "step": 6107 }, { "epoch": 0.4786833855799373, "grad_norm": 3.2941102981567383, "learning_rate": 9.660681685696922e-06, "loss": 1.7485, "step": 6108 }, { "epoch": 0.4787617554858934, "grad_norm": 4.150946617126465, "learning_rate": 9.660570016616255e-06, "loss": 1.0539, "step": 6109 }, { "epoch": 0.4788401253918495, "grad_norm": 6.984428405761719, "learning_rate": 9.660458329809175e-06, "loss": 1.5918, "step": 6110 }, { "epoch": 0.47891849529780567, "grad_norm": 2.7065980434417725, "learning_rate": 9.660346625276104e-06, "loss": 1.2746, "step": 6111 }, { "epoch": 0.4789968652037618, "grad_norm": 2.946854591369629, "learning_rate": 9.660234903017469e-06, "loss": 1.4702, "step": 6112 }, { "epoch": 0.4790752351097179, "grad_norm": 2.4396352767944336, "learning_rate": 9.660123163033692e-06, "loss": 1.3908, "step": 6113 }, { "epoch": 0.479153605015674, "grad_norm": 2.8256688117980957, "learning_rate": 9.660011405325202e-06, "loss": 1.4959, "step": 6114 }, { "epoch": 0.4792319749216301, "grad_norm": 2.599560499191284, "learning_rate": 9.659899629892422e-06, "loss": 1.3618, "step": 6115 }, { "epoch": 0.4793103448275862, "grad_norm": 2.5134172439575195, "learning_rate": 9.659787836735777e-06, "loss": 1.5989, "step": 6116 }, { "epoch": 0.47938871473354233, "grad_norm": 2.88397216796875, "learning_rate": 9.659676025855692e-06, "loss": 1.5212, "step": 6117 }, { "epoch": 0.47946708463949844, "grad_norm": 4.22067928314209, "learning_rate": 9.659564197252594e-06, "loss": 1.4825, "step": 6118 }, { "epoch": 0.47954545454545455, "grad_norm": 5.137130260467529, "learning_rate": 9.659452350926905e-06, "loss": 1.7994, "step": 6119 }, { "epoch": 0.47962382445141066, "grad_norm": 4.070262908935547, "learning_rate": 9.659340486879055e-06, "loss": 1.3045, "step": 6120 }, { "epoch": 0.4797021943573668, "grad_norm": 2.281559467315674, "learning_rate": 9.659228605109466e-06, "loss": 1.3022, "step": 6121 }, { "epoch": 0.4797805642633229, "grad_norm": 4.173072814941406, "learning_rate": 9.659116705618563e-06, "loss": 1.596, "step": 6122 }, { "epoch": 0.479858934169279, "grad_norm": 2.7685184478759766, "learning_rate": 9.659004788406774e-06, "loss": 1.1662, "step": 6123 }, { "epoch": 0.4799373040752351, "grad_norm": 2.0156924724578857, "learning_rate": 9.658892853474526e-06, "loss": 1.3432, "step": 6124 }, { "epoch": 0.4800156739811912, "grad_norm": 2.27468204498291, "learning_rate": 9.65878090082224e-06, "loss": 1.3596, "step": 6125 }, { "epoch": 0.4800940438871473, "grad_norm": 6.675255298614502, "learning_rate": 9.658668930450346e-06, "loss": 0.6516, "step": 6126 }, { "epoch": 0.48017241379310344, "grad_norm": 2.5444376468658447, "learning_rate": 9.658556942359266e-06, "loss": 1.2446, "step": 6127 }, { "epoch": 0.48025078369905955, "grad_norm": 3.7142715454101562, "learning_rate": 9.65844493654943e-06, "loss": 0.7168, "step": 6128 }, { "epoch": 0.48032915360501566, "grad_norm": 2.1951494216918945, "learning_rate": 9.658332913021262e-06, "loss": 1.6008, "step": 6129 }, { "epoch": 0.48040752351097177, "grad_norm": 2.472118854522705, "learning_rate": 9.658220871775188e-06, "loss": 1.4731, "step": 6130 }, { "epoch": 0.4804858934169279, "grad_norm": 2.887146234512329, "learning_rate": 9.658108812811634e-06, "loss": 1.8952, "step": 6131 }, { "epoch": 0.480564263322884, "grad_norm": 4.54972505569458, "learning_rate": 9.657996736131027e-06, "loss": 1.4856, "step": 6132 }, { "epoch": 0.4806426332288401, "grad_norm": 2.709026336669922, "learning_rate": 9.657884641733793e-06, "loss": 1.3622, "step": 6133 }, { "epoch": 0.48072100313479627, "grad_norm": 3.12853741645813, "learning_rate": 9.657772529620358e-06, "loss": 1.4387, "step": 6134 }, { "epoch": 0.4807993730407524, "grad_norm": 3.6502530574798584, "learning_rate": 9.657660399791148e-06, "loss": 1.598, "step": 6135 }, { "epoch": 0.4808777429467085, "grad_norm": 4.145606517791748, "learning_rate": 9.657548252246593e-06, "loss": 1.338, "step": 6136 }, { "epoch": 0.4809561128526646, "grad_norm": 2.2226831912994385, "learning_rate": 9.657436086987114e-06, "loss": 1.531, "step": 6137 }, { "epoch": 0.4810344827586207, "grad_norm": 3.7513506412506104, "learning_rate": 9.657323904013141e-06, "loss": 1.0607, "step": 6138 }, { "epoch": 0.4811128526645768, "grad_norm": 2.4910380840301514, "learning_rate": 9.6572117033251e-06, "loss": 1.5151, "step": 6139 }, { "epoch": 0.48119122257053293, "grad_norm": 8.948182106018066, "learning_rate": 9.657099484923418e-06, "loss": 1.9785, "step": 6140 }, { "epoch": 0.48126959247648904, "grad_norm": 1.7659401893615723, "learning_rate": 9.656987248808522e-06, "loss": 1.244, "step": 6141 }, { "epoch": 0.48134796238244515, "grad_norm": 2.7587332725524902, "learning_rate": 9.656874994980837e-06, "loss": 1.5147, "step": 6142 }, { "epoch": 0.48142633228840126, "grad_norm": 3.6233856678009033, "learning_rate": 9.656762723440794e-06, "loss": 1.3654, "step": 6143 }, { "epoch": 0.48150470219435737, "grad_norm": 2.7798454761505127, "learning_rate": 9.656650434188814e-06, "loss": 1.4472, "step": 6144 }, { "epoch": 0.4815830721003135, "grad_norm": 5.23903226852417, "learning_rate": 9.65653812722533e-06, "loss": 1.022, "step": 6145 }, { "epoch": 0.4816614420062696, "grad_norm": 4.044146537780762, "learning_rate": 9.656425802550766e-06, "loss": 1.4994, "step": 6146 }, { "epoch": 0.4817398119122257, "grad_norm": 3.9023027420043945, "learning_rate": 9.656313460165551e-06, "loss": 1.4836, "step": 6147 }, { "epoch": 0.4818181818181818, "grad_norm": 2.4646058082580566, "learning_rate": 9.65620110007011e-06, "loss": 1.1816, "step": 6148 }, { "epoch": 0.4818965517241379, "grad_norm": 3.2420225143432617, "learning_rate": 9.656088722264873e-06, "loss": 1.6998, "step": 6149 }, { "epoch": 0.48197492163009403, "grad_norm": 3.2192935943603516, "learning_rate": 9.655976326750266e-06, "loss": 1.3973, "step": 6150 }, { "epoch": 0.48205329153605014, "grad_norm": 2.9274637699127197, "learning_rate": 9.655863913526713e-06, "loss": 1.54, "step": 6151 }, { "epoch": 0.48213166144200625, "grad_norm": 2.276344060897827, "learning_rate": 9.65575148259465e-06, "loss": 1.9756, "step": 6152 }, { "epoch": 0.48221003134796236, "grad_norm": 3.1920852661132812, "learning_rate": 9.655639033954496e-06, "loss": 1.4306, "step": 6153 }, { "epoch": 0.4822884012539185, "grad_norm": 2.8567240238189697, "learning_rate": 9.655526567606684e-06, "loss": 1.2193, "step": 6154 }, { "epoch": 0.4823667711598746, "grad_norm": 2.4051942825317383, "learning_rate": 9.655414083551639e-06, "loss": 1.279, "step": 6155 }, { "epoch": 0.4824451410658307, "grad_norm": 1.9530935287475586, "learning_rate": 9.65530158178979e-06, "loss": 1.4762, "step": 6156 }, { "epoch": 0.4825235109717868, "grad_norm": 4.861466407775879, "learning_rate": 9.655189062321566e-06, "loss": 1.3638, "step": 6157 }, { "epoch": 0.482601880877743, "grad_norm": 2.497542381286621, "learning_rate": 9.655076525147393e-06, "loss": 1.7017, "step": 6158 }, { "epoch": 0.4826802507836991, "grad_norm": 2.14408802986145, "learning_rate": 9.654963970267702e-06, "loss": 1.4059, "step": 6159 }, { "epoch": 0.4827586206896552, "grad_norm": 4.124880790710449, "learning_rate": 9.654851397682918e-06, "loss": 1.6833, "step": 6160 }, { "epoch": 0.4828369905956113, "grad_norm": 3.50007700920105, "learning_rate": 9.654738807393469e-06, "loss": 1.4481, "step": 6161 }, { "epoch": 0.4829153605015674, "grad_norm": 3.3048086166381836, "learning_rate": 9.654626199399784e-06, "loss": 1.778, "step": 6162 }, { "epoch": 0.4829937304075235, "grad_norm": 5.054560661315918, "learning_rate": 9.654513573702292e-06, "loss": 1.457, "step": 6163 }, { "epoch": 0.48307210031347964, "grad_norm": 2.3519527912139893, "learning_rate": 9.654400930301423e-06, "loss": 1.6299, "step": 6164 }, { "epoch": 0.48315047021943575, "grad_norm": 4.505644798278809, "learning_rate": 9.654288269197603e-06, "loss": 1.3562, "step": 6165 }, { "epoch": 0.48322884012539186, "grad_norm": 4.781591415405273, "learning_rate": 9.65417559039126e-06, "loss": 1.6805, "step": 6166 }, { "epoch": 0.48330721003134797, "grad_norm": 7.58634614944458, "learning_rate": 9.654062893882825e-06, "loss": 1.4968, "step": 6167 }, { "epoch": 0.4833855799373041, "grad_norm": 2.770528554916382, "learning_rate": 9.653950179672723e-06, "loss": 1.3915, "step": 6168 }, { "epoch": 0.4834639498432602, "grad_norm": 2.771131753921509, "learning_rate": 9.653837447761386e-06, "loss": 1.5654, "step": 6169 }, { "epoch": 0.4835423197492163, "grad_norm": 4.418568134307861, "learning_rate": 9.653724698149242e-06, "loss": 1.2118, "step": 6170 }, { "epoch": 0.4836206896551724, "grad_norm": 3.1229796409606934, "learning_rate": 9.653611930836719e-06, "loss": 1.3186, "step": 6171 }, { "epoch": 0.4836990595611285, "grad_norm": 6.795858383178711, "learning_rate": 9.653499145824248e-06, "loss": 1.9887, "step": 6172 }, { "epoch": 0.48377742946708463, "grad_norm": 2.867321252822876, "learning_rate": 9.653386343112254e-06, "loss": 1.2021, "step": 6173 }, { "epoch": 0.48385579937304074, "grad_norm": 2.11283016204834, "learning_rate": 9.65327352270117e-06, "loss": 1.3544, "step": 6174 }, { "epoch": 0.48393416927899685, "grad_norm": 2.037238597869873, "learning_rate": 9.653160684591423e-06, "loss": 1.1015, "step": 6175 }, { "epoch": 0.48401253918495296, "grad_norm": 3.8121368885040283, "learning_rate": 9.653047828783444e-06, "loss": 1.4228, "step": 6176 }, { "epoch": 0.48409090909090907, "grad_norm": 4.287092208862305, "learning_rate": 9.65293495527766e-06, "loss": 1.634, "step": 6177 }, { "epoch": 0.4841692789968652, "grad_norm": 2.853914260864258, "learning_rate": 9.6528220640745e-06, "loss": 1.5289, "step": 6178 }, { "epoch": 0.4842476489028213, "grad_norm": 2.785107135772705, "learning_rate": 9.652709155174398e-06, "loss": 1.4506, "step": 6179 }, { "epoch": 0.4843260188087774, "grad_norm": 1.8472974300384521, "learning_rate": 9.652596228577777e-06, "loss": 1.4423, "step": 6180 }, { "epoch": 0.48440438871473357, "grad_norm": 3.413743257522583, "learning_rate": 9.652483284285072e-06, "loss": 1.2734, "step": 6181 }, { "epoch": 0.4844827586206897, "grad_norm": 4.253451824188232, "learning_rate": 9.65237032229671e-06, "loss": 0.9769, "step": 6182 }, { "epoch": 0.4845611285266458, "grad_norm": 3.294386863708496, "learning_rate": 9.652257342613118e-06, "loss": 1.8682, "step": 6183 }, { "epoch": 0.4846394984326019, "grad_norm": 2.2923638820648193, "learning_rate": 9.652144345234731e-06, "loss": 1.2583, "step": 6184 }, { "epoch": 0.484717868338558, "grad_norm": 4.598265647888184, "learning_rate": 9.652031330161977e-06, "loss": 1.5662, "step": 6185 }, { "epoch": 0.4847962382445141, "grad_norm": 3.1475460529327393, "learning_rate": 9.651918297395285e-06, "loss": 1.4013, "step": 6186 }, { "epoch": 0.48487460815047023, "grad_norm": 2.746750831604004, "learning_rate": 9.651805246935085e-06, "loss": 1.0757, "step": 6187 }, { "epoch": 0.48495297805642634, "grad_norm": 5.824324131011963, "learning_rate": 9.651692178781807e-06, "loss": 1.1642, "step": 6188 }, { "epoch": 0.48503134796238245, "grad_norm": 2.8779430389404297, "learning_rate": 9.651579092935879e-06, "loss": 1.7318, "step": 6189 }, { "epoch": 0.48510971786833856, "grad_norm": 2.4265711307525635, "learning_rate": 9.651465989397735e-06, "loss": 1.1543, "step": 6190 }, { "epoch": 0.4851880877742947, "grad_norm": 3.002671241760254, "learning_rate": 9.651352868167804e-06, "loss": 1.3689, "step": 6191 }, { "epoch": 0.4852664576802508, "grad_norm": 3.568183422088623, "learning_rate": 9.651239729246516e-06, "loss": 0.4587, "step": 6192 }, { "epoch": 0.4853448275862069, "grad_norm": 4.203289031982422, "learning_rate": 9.651126572634302e-06, "loss": 1.1811, "step": 6193 }, { "epoch": 0.485423197492163, "grad_norm": 3.088108539581299, "learning_rate": 9.65101339833159e-06, "loss": 1.409, "step": 6194 }, { "epoch": 0.4855015673981191, "grad_norm": 2.9947094917297363, "learning_rate": 9.650900206338811e-06, "loss": 1.3708, "step": 6195 }, { "epoch": 0.4855799373040752, "grad_norm": 16.56060791015625, "learning_rate": 9.650786996656399e-06, "loss": 1.7493, "step": 6196 }, { "epoch": 0.48565830721003134, "grad_norm": 3.8184640407562256, "learning_rate": 9.650673769284779e-06, "loss": 1.3668, "step": 6197 }, { "epoch": 0.48573667711598745, "grad_norm": 2.860837697982788, "learning_rate": 9.650560524224387e-06, "loss": 1.6431, "step": 6198 }, { "epoch": 0.48581504702194356, "grad_norm": 2.448652505874634, "learning_rate": 9.65044726147565e-06, "loss": 1.6599, "step": 6199 }, { "epoch": 0.48589341692789967, "grad_norm": 6.674617290496826, "learning_rate": 9.650333981039003e-06, "loss": 1.6168, "step": 6200 }, { "epoch": 0.4859717868338558, "grad_norm": 3.158552646636963, "learning_rate": 9.65022068291487e-06, "loss": 0.9843, "step": 6201 }, { "epoch": 0.4860501567398119, "grad_norm": 4.890743732452393, "learning_rate": 9.65010736710369e-06, "loss": 1.396, "step": 6202 }, { "epoch": 0.486128526645768, "grad_norm": 4.25236177444458, "learning_rate": 9.649994033605888e-06, "loss": 1.2566, "step": 6203 }, { "epoch": 0.4862068965517241, "grad_norm": 1.8594321012496948, "learning_rate": 9.649880682421897e-06, "loss": 1.3234, "step": 6204 }, { "epoch": 0.4862852664576803, "grad_norm": 5.568452835083008, "learning_rate": 9.649767313552148e-06, "loss": 1.0419, "step": 6205 }, { "epoch": 0.4863636363636364, "grad_norm": 9.333484649658203, "learning_rate": 9.649653926997074e-06, "loss": 0.6542, "step": 6206 }, { "epoch": 0.4864420062695925, "grad_norm": 2.3351309299468994, "learning_rate": 9.649540522757103e-06, "loss": 1.3995, "step": 6207 }, { "epoch": 0.4865203761755486, "grad_norm": 7.3060150146484375, "learning_rate": 9.649427100832669e-06, "loss": 1.5625, "step": 6208 }, { "epoch": 0.4865987460815047, "grad_norm": 2.9979639053344727, "learning_rate": 9.649313661224202e-06, "loss": 1.2727, "step": 6209 }, { "epoch": 0.48667711598746083, "grad_norm": 2.551259756088257, "learning_rate": 9.649200203932133e-06, "loss": 1.3691, "step": 6210 }, { "epoch": 0.48675548589341694, "grad_norm": 2.8901479244232178, "learning_rate": 9.649086728956895e-06, "loss": 1.6319, "step": 6211 }, { "epoch": 0.48683385579937305, "grad_norm": 4.085253715515137, "learning_rate": 9.648973236298919e-06, "loss": 1.4321, "step": 6212 }, { "epoch": 0.48691222570532916, "grad_norm": 3.9695591926574707, "learning_rate": 9.648859725958637e-06, "loss": 1.4973, "step": 6213 }, { "epoch": 0.48699059561128527, "grad_norm": 2.6731507778167725, "learning_rate": 9.64874619793648e-06, "loss": 1.4427, "step": 6214 }, { "epoch": 0.4870689655172414, "grad_norm": 2.662510871887207, "learning_rate": 9.648632652232881e-06, "loss": 1.3898, "step": 6215 }, { "epoch": 0.4871473354231975, "grad_norm": 2.14998722076416, "learning_rate": 9.648519088848272e-06, "loss": 1.8077, "step": 6216 }, { "epoch": 0.4872257053291536, "grad_norm": 6.71682596206665, "learning_rate": 9.648405507783082e-06, "loss": 1.3975, "step": 6217 }, { "epoch": 0.4873040752351097, "grad_norm": 2.2306253910064697, "learning_rate": 9.648291909037745e-06, "loss": 1.2246, "step": 6218 }, { "epoch": 0.4873824451410658, "grad_norm": 2.9684643745422363, "learning_rate": 9.648178292612695e-06, "loss": 1.4912, "step": 6219 }, { "epoch": 0.48746081504702193, "grad_norm": 4.830972671508789, "learning_rate": 9.64806465850836e-06, "loss": 1.7231, "step": 6220 }, { "epoch": 0.48753918495297804, "grad_norm": 2.688460350036621, "learning_rate": 9.647951006725176e-06, "loss": 1.0957, "step": 6221 }, { "epoch": 0.48761755485893415, "grad_norm": 3.278261661529541, "learning_rate": 9.647837337263573e-06, "loss": 1.0837, "step": 6222 }, { "epoch": 0.48769592476489027, "grad_norm": 6.376089096069336, "learning_rate": 9.647723650123983e-06, "loss": 1.8483, "step": 6223 }, { "epoch": 0.4877742946708464, "grad_norm": 4.26053524017334, "learning_rate": 9.647609945306842e-06, "loss": 1.6745, "step": 6224 }, { "epoch": 0.4878526645768025, "grad_norm": 4.412795543670654, "learning_rate": 9.647496222812579e-06, "loss": 1.9908, "step": 6225 }, { "epoch": 0.4879310344827586, "grad_norm": 4.375613212585449, "learning_rate": 9.647382482641626e-06, "loss": 1.4527, "step": 6226 }, { "epoch": 0.4880094043887147, "grad_norm": 3.116077423095703, "learning_rate": 9.647268724794418e-06, "loss": 1.3256, "step": 6227 }, { "epoch": 0.4880877742946709, "grad_norm": 3.134599447250366, "learning_rate": 9.647154949271386e-06, "loss": 1.4507, "step": 6228 }, { "epoch": 0.488166144200627, "grad_norm": 3.4712657928466797, "learning_rate": 9.647041156072966e-06, "loss": 1.8519, "step": 6229 }, { "epoch": 0.4882445141065831, "grad_norm": 2.5159363746643066, "learning_rate": 9.646927345199587e-06, "loss": 1.2691, "step": 6230 }, { "epoch": 0.4883228840125392, "grad_norm": 2.913607120513916, "learning_rate": 9.646813516651683e-06, "loss": 1.8241, "step": 6231 }, { "epoch": 0.4884012539184953, "grad_norm": 2.77599835395813, "learning_rate": 9.646699670429687e-06, "loss": 1.3981, "step": 6232 }, { "epoch": 0.4884796238244514, "grad_norm": 16.358783721923828, "learning_rate": 9.646585806534033e-06, "loss": 1.426, "step": 6233 }, { "epoch": 0.48855799373040754, "grad_norm": 2.5911645889282227, "learning_rate": 9.646471924965151e-06, "loss": 1.7152, "step": 6234 }, { "epoch": 0.48863636363636365, "grad_norm": 4.352189540863037, "learning_rate": 9.646358025723477e-06, "loss": 0.2314, "step": 6235 }, { "epoch": 0.48871473354231976, "grad_norm": 11.333841323852539, "learning_rate": 9.646244108809445e-06, "loss": 1.0432, "step": 6236 }, { "epoch": 0.48879310344827587, "grad_norm": 3.261021137237549, "learning_rate": 9.646130174223487e-06, "loss": 1.4416, "step": 6237 }, { "epoch": 0.488871473354232, "grad_norm": 3.7513768672943115, "learning_rate": 9.646016221966035e-06, "loss": 1.3676, "step": 6238 }, { "epoch": 0.4889498432601881, "grad_norm": 4.201303005218506, "learning_rate": 9.645902252037524e-06, "loss": 1.415, "step": 6239 }, { "epoch": 0.4890282131661442, "grad_norm": 5.699334621429443, "learning_rate": 9.645788264438388e-06, "loss": 1.6314, "step": 6240 }, { "epoch": 0.4891065830721003, "grad_norm": 2.9552600383758545, "learning_rate": 9.645674259169058e-06, "loss": 1.3231, "step": 6241 }, { "epoch": 0.4891849529780564, "grad_norm": 4.644684791564941, "learning_rate": 9.645560236229971e-06, "loss": 1.7918, "step": 6242 }, { "epoch": 0.48926332288401253, "grad_norm": 3.4037134647369385, "learning_rate": 9.645446195621558e-06, "loss": 1.835, "step": 6243 }, { "epoch": 0.48934169278996864, "grad_norm": 2.2187089920043945, "learning_rate": 9.645332137344254e-06, "loss": 1.3831, "step": 6244 }, { "epoch": 0.48942006269592475, "grad_norm": 7.027005195617676, "learning_rate": 9.64521806139849e-06, "loss": 1.5902, "step": 6245 }, { "epoch": 0.48949843260188086, "grad_norm": 2.4103927612304688, "learning_rate": 9.645103967784707e-06, "loss": 1.6227, "step": 6246 }, { "epoch": 0.489576802507837, "grad_norm": 2.5170412063598633, "learning_rate": 9.644989856503333e-06, "loss": 1.3143, "step": 6247 }, { "epoch": 0.4896551724137931, "grad_norm": 4.260627269744873, "learning_rate": 9.644875727554801e-06, "loss": 1.8064, "step": 6248 }, { "epoch": 0.4897335423197492, "grad_norm": 2.8106706142425537, "learning_rate": 9.64476158093955e-06, "loss": 1.4442, "step": 6249 }, { "epoch": 0.4898119122257053, "grad_norm": 3.1199734210968018, "learning_rate": 9.64464741665801e-06, "loss": 1.2829, "step": 6250 }, { "epoch": 0.48989028213166147, "grad_norm": 8.81820011138916, "learning_rate": 9.644533234710618e-06, "loss": 0.984, "step": 6251 }, { "epoch": 0.4899686520376176, "grad_norm": 2.4261367321014404, "learning_rate": 9.644419035097807e-06, "loss": 1.3849, "step": 6252 }, { "epoch": 0.4900470219435737, "grad_norm": 4.920921802520752, "learning_rate": 9.644304817820012e-06, "loss": 1.2658, "step": 6253 }, { "epoch": 0.4901253918495298, "grad_norm": 5.2817463874816895, "learning_rate": 9.644190582877667e-06, "loss": 1.5011, "step": 6254 }, { "epoch": 0.4902037617554859, "grad_norm": 3.16513991355896, "learning_rate": 9.644076330271203e-06, "loss": 1.6986, "step": 6255 }, { "epoch": 0.490282131661442, "grad_norm": 2.7698259353637695, "learning_rate": 9.643962060001062e-06, "loss": 1.3081, "step": 6256 }, { "epoch": 0.49036050156739813, "grad_norm": 3.9840335845947266, "learning_rate": 9.643847772067675e-06, "loss": 1.2741, "step": 6257 }, { "epoch": 0.49043887147335424, "grad_norm": 4.5661749839782715, "learning_rate": 9.643733466471474e-06, "loss": 1.2545, "step": 6258 }, { "epoch": 0.49051724137931035, "grad_norm": 2.748598337173462, "learning_rate": 9.643619143212896e-06, "loss": 1.1578, "step": 6259 }, { "epoch": 0.49059561128526646, "grad_norm": 4.732070446014404, "learning_rate": 9.643504802292377e-06, "loss": 1.4269, "step": 6260 }, { "epoch": 0.4906739811912226, "grad_norm": 4.573231220245361, "learning_rate": 9.64339044371035e-06, "loss": 1.0158, "step": 6261 }, { "epoch": 0.4907523510971787, "grad_norm": 3.4213409423828125, "learning_rate": 9.64327606746725e-06, "loss": 1.5574, "step": 6262 }, { "epoch": 0.4908307210031348, "grad_norm": 13.182829856872559, "learning_rate": 9.643161673563515e-06, "loss": 1.0765, "step": 6263 }, { "epoch": 0.4909090909090909, "grad_norm": 6.316121578216553, "learning_rate": 9.643047261999577e-06, "loss": 1.5313, "step": 6264 }, { "epoch": 0.490987460815047, "grad_norm": 4.242210388183594, "learning_rate": 9.642932832775871e-06, "loss": 1.259, "step": 6265 }, { "epoch": 0.4910658307210031, "grad_norm": 3.191789150238037, "learning_rate": 9.642818385892833e-06, "loss": 1.4558, "step": 6266 }, { "epoch": 0.49114420062695924, "grad_norm": 1.9820061922073364, "learning_rate": 9.642703921350901e-06, "loss": 1.493, "step": 6267 }, { "epoch": 0.49122257053291535, "grad_norm": 2.7056097984313965, "learning_rate": 9.642589439150508e-06, "loss": 1.1341, "step": 6268 }, { "epoch": 0.49130094043887146, "grad_norm": 4.8985466957092285, "learning_rate": 9.642474939292086e-06, "loss": 1.689, "step": 6269 }, { "epoch": 0.49137931034482757, "grad_norm": 2.4670424461364746, "learning_rate": 9.642360421776075e-06, "loss": 1.2933, "step": 6270 }, { "epoch": 0.4914576802507837, "grad_norm": 18.31973648071289, "learning_rate": 9.642245886602912e-06, "loss": 1.8906, "step": 6271 }, { "epoch": 0.4915360501567398, "grad_norm": 5.714369297027588, "learning_rate": 9.642131333773026e-06, "loss": 1.398, "step": 6272 }, { "epoch": 0.4916144200626959, "grad_norm": 3.497753858566284, "learning_rate": 9.642016763286861e-06, "loss": 1.4229, "step": 6273 }, { "epoch": 0.491692789968652, "grad_norm": 2.574375867843628, "learning_rate": 9.641902175144847e-06, "loss": 1.1787, "step": 6274 }, { "epoch": 0.4917711598746082, "grad_norm": 2.191831111907959, "learning_rate": 9.64178756934742e-06, "loss": 1.1598, "step": 6275 }, { "epoch": 0.4918495297805643, "grad_norm": 3.580014228820801, "learning_rate": 9.641672945895017e-06, "loss": 0.9859, "step": 6276 }, { "epoch": 0.4919278996865204, "grad_norm": 2.3790202140808105, "learning_rate": 9.641558304788077e-06, "loss": 1.3966, "step": 6277 }, { "epoch": 0.4920062695924765, "grad_norm": 9.885272026062012, "learning_rate": 9.641443646027031e-06, "loss": 1.1587, "step": 6278 }, { "epoch": 0.4920846394984326, "grad_norm": 2.8010542392730713, "learning_rate": 9.641328969612318e-06, "loss": 1.327, "step": 6279 }, { "epoch": 0.49216300940438873, "grad_norm": 2.115095376968384, "learning_rate": 9.641214275544373e-06, "loss": 1.5489, "step": 6280 }, { "epoch": 0.49224137931034484, "grad_norm": 2.188063859939575, "learning_rate": 9.641099563823633e-06, "loss": 1.4694, "step": 6281 }, { "epoch": 0.49231974921630095, "grad_norm": 5.895400047302246, "learning_rate": 9.640984834450535e-06, "loss": 1.349, "step": 6282 }, { "epoch": 0.49239811912225706, "grad_norm": 1.978955626487732, "learning_rate": 9.640870087425514e-06, "loss": 1.4095, "step": 6283 }, { "epoch": 0.49247648902821317, "grad_norm": 2.5328404903411865, "learning_rate": 9.640755322749007e-06, "loss": 1.5169, "step": 6284 }, { "epoch": 0.4925548589341693, "grad_norm": 2.350728988647461, "learning_rate": 9.64064054042145e-06, "loss": 1.3598, "step": 6285 }, { "epoch": 0.4926332288401254, "grad_norm": 3.762345552444458, "learning_rate": 9.640525740443279e-06, "loss": 1.2048, "step": 6286 }, { "epoch": 0.4927115987460815, "grad_norm": 2.2931606769561768, "learning_rate": 9.640410922814933e-06, "loss": 1.1092, "step": 6287 }, { "epoch": 0.4927899686520376, "grad_norm": 5.178919792175293, "learning_rate": 9.640296087536848e-06, "loss": 1.0481, "step": 6288 }, { "epoch": 0.4928683385579937, "grad_norm": 3.789625644683838, "learning_rate": 9.640181234609459e-06, "loss": 1.5403, "step": 6289 }, { "epoch": 0.49294670846394983, "grad_norm": 2.6704013347625732, "learning_rate": 9.640066364033202e-06, "loss": 1.3068, "step": 6290 }, { "epoch": 0.49302507836990594, "grad_norm": 3.7779955863952637, "learning_rate": 9.63995147580852e-06, "loss": 1.1005, "step": 6291 }, { "epoch": 0.49310344827586206, "grad_norm": 3.4201037883758545, "learning_rate": 9.639836569935843e-06, "loss": 1.3755, "step": 6292 }, { "epoch": 0.49318181818181817, "grad_norm": 3.541266918182373, "learning_rate": 9.639721646415612e-06, "loss": 1.9068, "step": 6293 }, { "epoch": 0.4932601880877743, "grad_norm": 2.7789440155029297, "learning_rate": 9.639606705248263e-06, "loss": 1.441, "step": 6294 }, { "epoch": 0.4933385579937304, "grad_norm": 6.976071834564209, "learning_rate": 9.63949174643423e-06, "loss": 1.1015, "step": 6295 }, { "epoch": 0.4934169278996865, "grad_norm": 2.6099603176116943, "learning_rate": 9.639376769973958e-06, "loss": 1.3188, "step": 6296 }, { "epoch": 0.4934952978056426, "grad_norm": 2.32011342048645, "learning_rate": 9.639261775867878e-06, "loss": 1.2611, "step": 6297 }, { "epoch": 0.4935736677115988, "grad_norm": 3.7408273220062256, "learning_rate": 9.639146764116428e-06, "loss": 1.816, "step": 6298 }, { "epoch": 0.4936520376175549, "grad_norm": 2.696619987487793, "learning_rate": 9.639031734720048e-06, "loss": 1.4036, "step": 6299 }, { "epoch": 0.493730407523511, "grad_norm": 2.7071566581726074, "learning_rate": 9.638916687679172e-06, "loss": 1.5808, "step": 6300 }, { "epoch": 0.4938087774294671, "grad_norm": 3.315016984939575, "learning_rate": 9.638801622994243e-06, "loss": 1.8385, "step": 6301 }, { "epoch": 0.4938871473354232, "grad_norm": 4.096063613891602, "learning_rate": 9.638686540665692e-06, "loss": 1.6912, "step": 6302 }, { "epoch": 0.4939655172413793, "grad_norm": 2.6059608459472656, "learning_rate": 9.638571440693961e-06, "loss": 1.5402, "step": 6303 }, { "epoch": 0.49404388714733544, "grad_norm": 3.378099203109741, "learning_rate": 9.638456323079488e-06, "loss": 1.4868, "step": 6304 }, { "epoch": 0.49412225705329155, "grad_norm": 2.9771692752838135, "learning_rate": 9.638341187822708e-06, "loss": 1.4842, "step": 6305 }, { "epoch": 0.49420062695924766, "grad_norm": 2.0750770568847656, "learning_rate": 9.638226034924061e-06, "loss": 1.3889, "step": 6306 }, { "epoch": 0.49427899686520377, "grad_norm": 8.080730438232422, "learning_rate": 9.638110864383984e-06, "loss": 0.9486, "step": 6307 }, { "epoch": 0.4943573667711599, "grad_norm": 3.6716771125793457, "learning_rate": 9.637995676202917e-06, "loss": 1.1062, "step": 6308 }, { "epoch": 0.494435736677116, "grad_norm": 2.428865432739258, "learning_rate": 9.637880470381295e-06, "loss": 1.5098, "step": 6309 }, { "epoch": 0.4945141065830721, "grad_norm": 4.236795902252197, "learning_rate": 9.637765246919559e-06, "loss": 1.7897, "step": 6310 }, { "epoch": 0.4945924764890282, "grad_norm": 2.9999477863311768, "learning_rate": 9.637650005818146e-06, "loss": 1.2875, "step": 6311 }, { "epoch": 0.4946708463949843, "grad_norm": 4.754597187042236, "learning_rate": 9.637534747077495e-06, "loss": 1.5882, "step": 6312 }, { "epoch": 0.49474921630094043, "grad_norm": 2.7616777420043945, "learning_rate": 9.637419470698042e-06, "loss": 1.3789, "step": 6313 }, { "epoch": 0.49482758620689654, "grad_norm": 3.4657862186431885, "learning_rate": 9.637304176680228e-06, "loss": 1.7049, "step": 6314 }, { "epoch": 0.49490595611285265, "grad_norm": 2.04495906829834, "learning_rate": 9.637188865024492e-06, "loss": 1.1602, "step": 6315 }, { "epoch": 0.49498432601880876, "grad_norm": 4.668565273284912, "learning_rate": 9.63707353573127e-06, "loss": 1.7923, "step": 6316 }, { "epoch": 0.4950626959247649, "grad_norm": 2.593773126602173, "learning_rate": 9.636958188801004e-06, "loss": 1.334, "step": 6317 }, { "epoch": 0.495141065830721, "grad_norm": 2.9621059894561768, "learning_rate": 9.63684282423413e-06, "loss": 1.5897, "step": 6318 }, { "epoch": 0.4952194357366771, "grad_norm": 2.327479124069214, "learning_rate": 9.636727442031085e-06, "loss": 1.32, "step": 6319 }, { "epoch": 0.4952978056426332, "grad_norm": 3.3569397926330566, "learning_rate": 9.636612042192313e-06, "loss": 1.8409, "step": 6320 }, { "epoch": 0.4953761755485893, "grad_norm": 2.781733512878418, "learning_rate": 9.63649662471825e-06, "loss": 1.0467, "step": 6321 }, { "epoch": 0.4954545454545455, "grad_norm": 6.606746196746826, "learning_rate": 9.636381189609334e-06, "loss": 1.4827, "step": 6322 }, { "epoch": 0.4955329153605016, "grad_norm": 3.3139007091522217, "learning_rate": 9.636265736866006e-06, "loss": 1.0886, "step": 6323 }, { "epoch": 0.4956112852664577, "grad_norm": 3.1546778678894043, "learning_rate": 9.636150266488704e-06, "loss": 1.4522, "step": 6324 }, { "epoch": 0.4956896551724138, "grad_norm": 3.684536933898926, "learning_rate": 9.636034778477868e-06, "loss": 1.5213, "step": 6325 }, { "epoch": 0.4957680250783699, "grad_norm": 4.049106121063232, "learning_rate": 9.635919272833938e-06, "loss": 1.6513, "step": 6326 }, { "epoch": 0.49584639498432603, "grad_norm": 3.973766326904297, "learning_rate": 9.635803749557351e-06, "loss": 1.6929, "step": 6327 }, { "epoch": 0.49592476489028214, "grad_norm": 3.1333210468292236, "learning_rate": 9.635688208648549e-06, "loss": 1.657, "step": 6328 }, { "epoch": 0.49600313479623825, "grad_norm": 3.2685041427612305, "learning_rate": 9.635572650107968e-06, "loss": 1.6985, "step": 6329 }, { "epoch": 0.49608150470219436, "grad_norm": 2.1192667484283447, "learning_rate": 9.635457073936051e-06, "loss": 1.3275, "step": 6330 }, { "epoch": 0.4961598746081505, "grad_norm": 6.365508079528809, "learning_rate": 9.635341480133236e-06, "loss": 1.5387, "step": 6331 }, { "epoch": 0.4962382445141066, "grad_norm": 2.4333434104919434, "learning_rate": 9.635225868699963e-06, "loss": 1.3263, "step": 6332 }, { "epoch": 0.4963166144200627, "grad_norm": 5.294932842254639, "learning_rate": 9.635110239636672e-06, "loss": 1.4482, "step": 6333 }, { "epoch": 0.4963949843260188, "grad_norm": 2.1405420303344727, "learning_rate": 9.6349945929438e-06, "loss": 1.3051, "step": 6334 }, { "epoch": 0.4964733542319749, "grad_norm": 1.9898796081542969, "learning_rate": 9.634878928621793e-06, "loss": 1.1662, "step": 6335 }, { "epoch": 0.496551724137931, "grad_norm": 2.539492607116699, "learning_rate": 9.634763246671083e-06, "loss": 1.2831, "step": 6336 }, { "epoch": 0.49663009404388714, "grad_norm": 2.9510338306427, "learning_rate": 9.634647547092117e-06, "loss": 1.2943, "step": 6337 }, { "epoch": 0.49670846394984325, "grad_norm": 2.549626350402832, "learning_rate": 9.634531829885333e-06, "loss": 1.2225, "step": 6338 }, { "epoch": 0.49678683385579936, "grad_norm": 3.7581264972686768, "learning_rate": 9.634416095051168e-06, "loss": 1.2196, "step": 6339 }, { "epoch": 0.49686520376175547, "grad_norm": 2.431142568588257, "learning_rate": 9.634300342590067e-06, "loss": 1.311, "step": 6340 }, { "epoch": 0.4969435736677116, "grad_norm": 4.106764793395996, "learning_rate": 9.634184572502465e-06, "loss": 1.4673, "step": 6341 }, { "epoch": 0.4970219435736677, "grad_norm": 6.267800331115723, "learning_rate": 9.634068784788808e-06, "loss": 1.1261, "step": 6342 }, { "epoch": 0.4971003134796238, "grad_norm": 3.9273972511291504, "learning_rate": 9.633952979449532e-06, "loss": 1.2764, "step": 6343 }, { "epoch": 0.4971786833855799, "grad_norm": 5.462475776672363, "learning_rate": 9.633837156485078e-06, "loss": 1.0022, "step": 6344 }, { "epoch": 0.4972570532915361, "grad_norm": 3.1067235469818115, "learning_rate": 9.633721315895889e-06, "loss": 1.038, "step": 6345 }, { "epoch": 0.4973354231974922, "grad_norm": 4.743744373321533, "learning_rate": 9.633605457682404e-06, "loss": 1.5695, "step": 6346 }, { "epoch": 0.4974137931034483, "grad_norm": 2.662313938140869, "learning_rate": 9.633489581845063e-06, "loss": 1.1784, "step": 6347 }, { "epoch": 0.4974921630094044, "grad_norm": 3.9055144786834717, "learning_rate": 9.63337368838431e-06, "loss": 1.0193, "step": 6348 }, { "epoch": 0.4975705329153605, "grad_norm": 8.076315879821777, "learning_rate": 9.633257777300581e-06, "loss": 1.103, "step": 6349 }, { "epoch": 0.49764890282131663, "grad_norm": 4.482515811920166, "learning_rate": 9.63314184859432e-06, "loss": 1.8424, "step": 6350 }, { "epoch": 0.49772727272727274, "grad_norm": 2.539667844772339, "learning_rate": 9.633025902265966e-06, "loss": 1.4244, "step": 6351 }, { "epoch": 0.49780564263322885, "grad_norm": 4.283896446228027, "learning_rate": 9.632909938315962e-06, "loss": 1.3878, "step": 6352 }, { "epoch": 0.49788401253918496, "grad_norm": 2.6708273887634277, "learning_rate": 9.632793956744749e-06, "loss": 1.4501, "step": 6353 }, { "epoch": 0.49796238244514107, "grad_norm": 6.527881622314453, "learning_rate": 9.632677957552767e-06, "loss": 1.2869, "step": 6354 }, { "epoch": 0.4980407523510972, "grad_norm": 2.166782855987549, "learning_rate": 9.632561940740456e-06, "loss": 1.2229, "step": 6355 }, { "epoch": 0.4981191222570533, "grad_norm": 2.4661502838134766, "learning_rate": 9.632445906308261e-06, "loss": 1.3441, "step": 6356 }, { "epoch": 0.4981974921630094, "grad_norm": 2.4083590507507324, "learning_rate": 9.632329854256619e-06, "loss": 1.3851, "step": 6357 }, { "epoch": 0.4982758620689655, "grad_norm": 5.911746978759766, "learning_rate": 9.632213784585975e-06, "loss": 1.2911, "step": 6358 }, { "epoch": 0.4983542319749216, "grad_norm": 1.723502516746521, "learning_rate": 9.632097697296769e-06, "loss": 1.3799, "step": 6359 }, { "epoch": 0.49843260188087773, "grad_norm": 5.532501697540283, "learning_rate": 9.63198159238944e-06, "loss": 1.0694, "step": 6360 }, { "epoch": 0.49851097178683385, "grad_norm": 4.696610450744629, "learning_rate": 9.631865469864437e-06, "loss": 1.0702, "step": 6361 }, { "epoch": 0.49858934169278996, "grad_norm": 4.943333148956299, "learning_rate": 9.631749329722193e-06, "loss": 1.6333, "step": 6362 }, { "epoch": 0.49866771159874607, "grad_norm": 6.080482482910156, "learning_rate": 9.631633171963155e-06, "loss": 1.0694, "step": 6363 }, { "epoch": 0.4987460815047022, "grad_norm": 3.7008872032165527, "learning_rate": 9.631516996587762e-06, "loss": 1.5891, "step": 6364 }, { "epoch": 0.4988244514106583, "grad_norm": 2.151254892349243, "learning_rate": 9.63140080359646e-06, "loss": 1.1949, "step": 6365 }, { "epoch": 0.4989028213166144, "grad_norm": 2.952598810195923, "learning_rate": 9.631284592989686e-06, "loss": 1.4542, "step": 6366 }, { "epoch": 0.4989811912225705, "grad_norm": 2.276878833770752, "learning_rate": 9.631168364767884e-06, "loss": 1.2478, "step": 6367 }, { "epoch": 0.4990595611285266, "grad_norm": 3.3091046810150146, "learning_rate": 9.631052118931498e-06, "loss": 1.5093, "step": 6368 }, { "epoch": 0.4991379310344828, "grad_norm": 3.9607603549957275, "learning_rate": 9.630935855480968e-06, "loss": 1.2726, "step": 6369 }, { "epoch": 0.4992163009404389, "grad_norm": 3.38790225982666, "learning_rate": 9.630819574416735e-06, "loss": 1.2673, "step": 6370 }, { "epoch": 0.499294670846395, "grad_norm": 5.582522392272949, "learning_rate": 9.630703275739244e-06, "loss": 1.3337, "step": 6371 }, { "epoch": 0.4993730407523511, "grad_norm": 2.564025640487671, "learning_rate": 9.630586959448935e-06, "loss": 1.2712, "step": 6372 }, { "epoch": 0.4994514106583072, "grad_norm": 4.654569149017334, "learning_rate": 9.630470625546252e-06, "loss": 1.3662, "step": 6373 }, { "epoch": 0.49952978056426334, "grad_norm": 10.611857414245605, "learning_rate": 9.63035427403164e-06, "loss": 1.7231, "step": 6374 }, { "epoch": 0.49960815047021945, "grad_norm": 4.369502067565918, "learning_rate": 9.630237904905536e-06, "loss": 0.9156, "step": 6375 }, { "epoch": 0.49968652037617556, "grad_norm": 2.986171245574951, "learning_rate": 9.630121518168386e-06, "loss": 1.1603, "step": 6376 }, { "epoch": 0.49976489028213167, "grad_norm": 3.384472131729126, "learning_rate": 9.630005113820631e-06, "loss": 1.6889, "step": 6377 }, { "epoch": 0.4998432601880878, "grad_norm": 4.177973747253418, "learning_rate": 9.629888691862716e-06, "loss": 0.5864, "step": 6378 }, { "epoch": 0.4999216300940439, "grad_norm": 2.997004985809326, "learning_rate": 9.629772252295083e-06, "loss": 1.5087, "step": 6379 }, { "epoch": 0.5, "grad_norm": 2.634860038757324, "learning_rate": 9.629655795118171e-06, "loss": 1.3556, "step": 6380 }, { "epoch": 0.5, "eval_loss": 1.3585582971572876, "eval_runtime": 110.052, "eval_samples_per_second": 24.416, "eval_steps_per_second": 6.106, "step": 6380 }, { "epoch": 0.5000783699059561, "grad_norm": 2.539411783218384, "learning_rate": 9.629539320332428e-06, "loss": 1.0474, "step": 6381 }, { "epoch": 0.5001567398119122, "grad_norm": 2.352426052093506, "learning_rate": 9.629422827938297e-06, "loss": 1.4643, "step": 6382 }, { "epoch": 0.5002351097178683, "grad_norm": 3.110311269760132, "learning_rate": 9.629306317936218e-06, "loss": 1.0694, "step": 6383 }, { "epoch": 0.5003134796238244, "grad_norm": 2.736670970916748, "learning_rate": 9.629189790326635e-06, "loss": 1.7547, "step": 6384 }, { "epoch": 0.5003918495297806, "grad_norm": 2.1585161685943604, "learning_rate": 9.629073245109992e-06, "loss": 1.2496, "step": 6385 }, { "epoch": 0.5004702194357367, "grad_norm": 3.1686949729919434, "learning_rate": 9.628956682286733e-06, "loss": 1.2716, "step": 6386 }, { "epoch": 0.5005485893416928, "grad_norm": 1.8599673509597778, "learning_rate": 9.6288401018573e-06, "loss": 1.4605, "step": 6387 }, { "epoch": 0.5006269592476489, "grad_norm": 2.8015401363372803, "learning_rate": 9.628723503822137e-06, "loss": 1.184, "step": 6388 }, { "epoch": 0.500705329153605, "grad_norm": 2.908994436264038, "learning_rate": 9.628606888181687e-06, "loss": 1.4133, "step": 6389 }, { "epoch": 0.5007836990595611, "grad_norm": 4.0386433601379395, "learning_rate": 9.628490254936394e-06, "loss": 1.6882, "step": 6390 }, { "epoch": 0.5008620689655172, "grad_norm": 3.272028684616089, "learning_rate": 9.6283736040867e-06, "loss": 1.4252, "step": 6391 }, { "epoch": 0.5009404388714733, "grad_norm": 6.906998157501221, "learning_rate": 9.628256935633051e-06, "loss": 1.9985, "step": 6392 }, { "epoch": 0.5010188087774294, "grad_norm": 3.2501981258392334, "learning_rate": 9.628140249575891e-06, "loss": 1.4148, "step": 6393 }, { "epoch": 0.5010971786833855, "grad_norm": 2.5527098178863525, "learning_rate": 9.628023545915661e-06, "loss": 1.2695, "step": 6394 }, { "epoch": 0.5011755485893417, "grad_norm": 3.2543225288391113, "learning_rate": 9.627906824652807e-06, "loss": 1.2572, "step": 6395 }, { "epoch": 0.5012539184952978, "grad_norm": 4.645862102508545, "learning_rate": 9.627790085787772e-06, "loss": 1.4469, "step": 6396 }, { "epoch": 0.5013322884012539, "grad_norm": 2.6210155487060547, "learning_rate": 9.627673329321001e-06, "loss": 1.2947, "step": 6397 }, { "epoch": 0.50141065830721, "grad_norm": 2.9972288608551025, "learning_rate": 9.627556555252938e-06, "loss": 1.4997, "step": 6398 }, { "epoch": 0.5014890282131661, "grad_norm": 3.1336097717285156, "learning_rate": 9.627439763584027e-06, "loss": 1.2289, "step": 6399 }, { "epoch": 0.5015673981191222, "grad_norm": 4.05084753036499, "learning_rate": 9.62732295431471e-06, "loss": 1.2882, "step": 6400 }, { "epoch": 0.5016457680250783, "grad_norm": 3.039600133895874, "learning_rate": 9.627206127445436e-06, "loss": 1.3976, "step": 6401 }, { "epoch": 0.5017241379310344, "grad_norm": 5.506648063659668, "learning_rate": 9.627089282976645e-06, "loss": 1.0265, "step": 6402 }, { "epoch": 0.5018025078369905, "grad_norm": 7.410022735595703, "learning_rate": 9.626972420908782e-06, "loss": 1.7445, "step": 6403 }, { "epoch": 0.5018808777429468, "grad_norm": 3.3149309158325195, "learning_rate": 9.626855541242294e-06, "loss": 1.0642, "step": 6404 }, { "epoch": 0.5019592476489029, "grad_norm": 3.1999564170837402, "learning_rate": 9.626738643977624e-06, "loss": 1.2918, "step": 6405 }, { "epoch": 0.502037617554859, "grad_norm": 4.366936683654785, "learning_rate": 9.626621729115217e-06, "loss": 1.5689, "step": 6406 }, { "epoch": 0.5021159874608151, "grad_norm": 6.5102081298828125, "learning_rate": 9.626504796655516e-06, "loss": 0.8852, "step": 6407 }, { "epoch": 0.5021943573667712, "grad_norm": 2.952291965484619, "learning_rate": 9.626387846598968e-06, "loss": 1.5453, "step": 6408 }, { "epoch": 0.5022727272727273, "grad_norm": 2.7277960777282715, "learning_rate": 9.626270878946016e-06, "loss": 1.6255, "step": 6409 }, { "epoch": 0.5023510971786834, "grad_norm": 2.586576461791992, "learning_rate": 9.626153893697106e-06, "loss": 1.3651, "step": 6410 }, { "epoch": 0.5024294670846395, "grad_norm": 8.636916160583496, "learning_rate": 9.626036890852683e-06, "loss": 1.3704, "step": 6411 }, { "epoch": 0.5025078369905956, "grad_norm": 17.086170196533203, "learning_rate": 9.625919870413192e-06, "loss": 1.2426, "step": 6412 }, { "epoch": 0.5025862068965518, "grad_norm": 2.919243812561035, "learning_rate": 9.625802832379078e-06, "loss": 1.7835, "step": 6413 }, { "epoch": 0.5026645768025079, "grad_norm": 2.144585609436035, "learning_rate": 9.625685776750786e-06, "loss": 1.5295, "step": 6414 }, { "epoch": 0.502742946708464, "grad_norm": 2.8660366535186768, "learning_rate": 9.62556870352876e-06, "loss": 1.3783, "step": 6415 }, { "epoch": 0.5028213166144201, "grad_norm": 4.397817611694336, "learning_rate": 9.625451612713448e-06, "loss": 1.3218, "step": 6416 }, { "epoch": 0.5028996865203762, "grad_norm": 2.9775638580322266, "learning_rate": 9.625334504305292e-06, "loss": 1.7771, "step": 6417 }, { "epoch": 0.5029780564263323, "grad_norm": 4.58216667175293, "learning_rate": 9.625217378304742e-06, "loss": 1.7894, "step": 6418 }, { "epoch": 0.5030564263322884, "grad_norm": 2.478280782699585, "learning_rate": 9.625100234712239e-06, "loss": 1.5591, "step": 6419 }, { "epoch": 0.5031347962382445, "grad_norm": 2.876189708709717, "learning_rate": 9.624983073528232e-06, "loss": 1.0704, "step": 6420 }, { "epoch": 0.5032131661442006, "grad_norm": 2.714550256729126, "learning_rate": 9.624865894753161e-06, "loss": 1.3202, "step": 6421 }, { "epoch": 0.5032915360501568, "grad_norm": 3.1191799640655518, "learning_rate": 9.62474869838748e-06, "loss": 1.2891, "step": 6422 }, { "epoch": 0.5033699059561129, "grad_norm": 2.2715511322021484, "learning_rate": 9.624631484431628e-06, "loss": 1.4599, "step": 6423 }, { "epoch": 0.503448275862069, "grad_norm": 2.3162713050842285, "learning_rate": 9.624514252886053e-06, "loss": 1.4523, "step": 6424 }, { "epoch": 0.5035266457680251, "grad_norm": 2.096071720123291, "learning_rate": 9.624397003751203e-06, "loss": 1.4898, "step": 6425 }, { "epoch": 0.5036050156739812, "grad_norm": 2.663191556930542, "learning_rate": 9.62427973702752e-06, "loss": 1.4409, "step": 6426 }, { "epoch": 0.5036833855799373, "grad_norm": 4.006802558898926, "learning_rate": 9.624162452715454e-06, "loss": 1.074, "step": 6427 }, { "epoch": 0.5037617554858934, "grad_norm": 3.838121175765991, "learning_rate": 9.624045150815447e-06, "loss": 1.5751, "step": 6428 }, { "epoch": 0.5038401253918495, "grad_norm": 1.7602864503860474, "learning_rate": 9.623927831327948e-06, "loss": 1.3141, "step": 6429 }, { "epoch": 0.5039184952978056, "grad_norm": 5.374485015869141, "learning_rate": 9.623810494253403e-06, "loss": 0.9131, "step": 6430 }, { "epoch": 0.5039968652037617, "grad_norm": 2.548022747039795, "learning_rate": 9.623693139592258e-06, "loss": 1.3557, "step": 6431 }, { "epoch": 0.5040752351097179, "grad_norm": 3.047731876373291, "learning_rate": 9.623575767344959e-06, "loss": 1.4044, "step": 6432 }, { "epoch": 0.504153605015674, "grad_norm": 2.151996374130249, "learning_rate": 9.623458377511951e-06, "loss": 1.3117, "step": 6433 }, { "epoch": 0.5042319749216301, "grad_norm": 2.2458903789520264, "learning_rate": 9.623340970093684e-06, "loss": 1.2909, "step": 6434 }, { "epoch": 0.5043103448275862, "grad_norm": 3.928396463394165, "learning_rate": 9.623223545090604e-06, "loss": 1.4399, "step": 6435 }, { "epoch": 0.5043887147335423, "grad_norm": 3.4565391540527344, "learning_rate": 9.623106102503153e-06, "loss": 1.56, "step": 6436 }, { "epoch": 0.5044670846394984, "grad_norm": 2.222940683364868, "learning_rate": 9.622988642331782e-06, "loss": 1.3939, "step": 6437 }, { "epoch": 0.5045454545454545, "grad_norm": 4.315552234649658, "learning_rate": 9.622871164576936e-06, "loss": 1.7062, "step": 6438 }, { "epoch": 0.5046238244514106, "grad_norm": 2.4837863445281982, "learning_rate": 9.622753669239065e-06, "loss": 1.7586, "step": 6439 }, { "epoch": 0.5047021943573667, "grad_norm": 3.895890712738037, "learning_rate": 9.622636156318613e-06, "loss": 1.235, "step": 6440 }, { "epoch": 0.5047805642633229, "grad_norm": 2.6524250507354736, "learning_rate": 9.622518625816025e-06, "loss": 1.504, "step": 6441 }, { "epoch": 0.504858934169279, "grad_norm": 2.4500789642333984, "learning_rate": 9.622401077731752e-06, "loss": 1.3304, "step": 6442 }, { "epoch": 0.5049373040752351, "grad_norm": 1.9551602602005005, "learning_rate": 9.62228351206624e-06, "loss": 1.1873, "step": 6443 }, { "epoch": 0.5050156739811912, "grad_norm": 5.14886999130249, "learning_rate": 9.622165928819935e-06, "loss": 1.7233, "step": 6444 }, { "epoch": 0.5050940438871473, "grad_norm": 2.9140408039093018, "learning_rate": 9.622048327993285e-06, "loss": 1.44, "step": 6445 }, { "epoch": 0.5051724137931034, "grad_norm": 2.7631564140319824, "learning_rate": 9.621930709586738e-06, "loss": 2.1874, "step": 6446 }, { "epoch": 0.5052507836990595, "grad_norm": 2.5674426555633545, "learning_rate": 9.621813073600738e-06, "loss": 1.8681, "step": 6447 }, { "epoch": 0.5053291536050156, "grad_norm": 2.603236675262451, "learning_rate": 9.621695420035737e-06, "loss": 1.2457, "step": 6448 }, { "epoch": 0.5054075235109717, "grad_norm": 3.0971875190734863, "learning_rate": 9.621577748892181e-06, "loss": 1.5489, "step": 6449 }, { "epoch": 0.5054858934169278, "grad_norm": 2.6430182456970215, "learning_rate": 9.621460060170515e-06, "loss": 1.4994, "step": 6450 }, { "epoch": 0.5055642633228841, "grad_norm": 7.034477710723877, "learning_rate": 9.62134235387119e-06, "loss": 0.9802, "step": 6451 }, { "epoch": 0.5056426332288402, "grad_norm": 2.9279098510742188, "learning_rate": 9.621224629994654e-06, "loss": 1.7998, "step": 6452 }, { "epoch": 0.5057210031347963, "grad_norm": 2.0748417377471924, "learning_rate": 9.621106888541351e-06, "loss": 1.4828, "step": 6453 }, { "epoch": 0.5057993730407524, "grad_norm": 2.4101154804229736, "learning_rate": 9.62098912951173e-06, "loss": 1.7866, "step": 6454 }, { "epoch": 0.5058777429467085, "grad_norm": 3.2114882469177246, "learning_rate": 9.620871352906242e-06, "loss": 1.3947, "step": 6455 }, { "epoch": 0.5059561128526646, "grad_norm": 3.7175045013427734, "learning_rate": 9.620753558725334e-06, "loss": 1.2996, "step": 6456 }, { "epoch": 0.5060344827586207, "grad_norm": 2.9711806774139404, "learning_rate": 9.620635746969448e-06, "loss": 1.3319, "step": 6457 }, { "epoch": 0.5061128526645768, "grad_norm": 3.648529052734375, "learning_rate": 9.62051791763904e-06, "loss": 1.4353, "step": 6458 }, { "epoch": 0.506191222570533, "grad_norm": 2.398829698562622, "learning_rate": 9.620400070734556e-06, "loss": 1.5548, "step": 6459 }, { "epoch": 0.5062695924764891, "grad_norm": 1.9204894304275513, "learning_rate": 9.620282206256442e-06, "loss": 1.5143, "step": 6460 }, { "epoch": 0.5063479623824452, "grad_norm": 2.354590892791748, "learning_rate": 9.620164324205148e-06, "loss": 1.4652, "step": 6461 }, { "epoch": 0.5064263322884013, "grad_norm": 3.156611680984497, "learning_rate": 9.620046424581121e-06, "loss": 1.4689, "step": 6462 }, { "epoch": 0.5065047021943574, "grad_norm": 5.608380317687988, "learning_rate": 9.619928507384811e-06, "loss": 1.4956, "step": 6463 }, { "epoch": 0.5065830721003135, "grad_norm": 2.755686044692993, "learning_rate": 9.619810572616666e-06, "loss": 1.3904, "step": 6464 }, { "epoch": 0.5066614420062696, "grad_norm": 3.146233558654785, "learning_rate": 9.619692620277135e-06, "loss": 1.5353, "step": 6465 }, { "epoch": 0.5067398119122257, "grad_norm": 3.522246837615967, "learning_rate": 9.619574650366665e-06, "loss": 1.3409, "step": 6466 }, { "epoch": 0.5068181818181818, "grad_norm": 1.984312891960144, "learning_rate": 9.619456662885707e-06, "loss": 1.424, "step": 6467 }, { "epoch": 0.506896551724138, "grad_norm": 4.036223411560059, "learning_rate": 9.619338657834708e-06, "loss": 1.2997, "step": 6468 }, { "epoch": 0.506974921630094, "grad_norm": 2.0545854568481445, "learning_rate": 9.619220635214118e-06, "loss": 1.6331, "step": 6469 }, { "epoch": 0.5070532915360502, "grad_norm": 3.1608166694641113, "learning_rate": 9.619102595024384e-06, "loss": 1.4961, "step": 6470 }, { "epoch": 0.5071316614420063, "grad_norm": 2.8131308555603027, "learning_rate": 9.618984537265956e-06, "loss": 1.6724, "step": 6471 }, { "epoch": 0.5072100313479624, "grad_norm": 5.01046895980835, "learning_rate": 9.618866461939285e-06, "loss": 1.0653, "step": 6472 }, { "epoch": 0.5072884012539185, "grad_norm": 3.856846570968628, "learning_rate": 9.618748369044818e-06, "loss": 1.3566, "step": 6473 }, { "epoch": 0.5073667711598746, "grad_norm": 2.6718525886535645, "learning_rate": 9.618630258583004e-06, "loss": 1.2415, "step": 6474 }, { "epoch": 0.5074451410658307, "grad_norm": 7.488310813903809, "learning_rate": 9.618512130554293e-06, "loss": 0.914, "step": 6475 }, { "epoch": 0.5075235109717868, "grad_norm": 2.0513827800750732, "learning_rate": 9.618393984959134e-06, "loss": 1.6277, "step": 6476 }, { "epoch": 0.5076018808777429, "grad_norm": 7.000985145568848, "learning_rate": 9.618275821797975e-06, "loss": 2.1039, "step": 6477 }, { "epoch": 0.507680250783699, "grad_norm": 3.7369625568389893, "learning_rate": 9.618157641071268e-06, "loss": 1.2203, "step": 6478 }, { "epoch": 0.5077586206896552, "grad_norm": 5.143610000610352, "learning_rate": 9.618039442779463e-06, "loss": 1.1173, "step": 6479 }, { "epoch": 0.5078369905956113, "grad_norm": 3.186643600463867, "learning_rate": 9.617921226923005e-06, "loss": 1.2457, "step": 6480 }, { "epoch": 0.5079153605015674, "grad_norm": 2.6325550079345703, "learning_rate": 9.61780299350235e-06, "loss": 1.6787, "step": 6481 }, { "epoch": 0.5079937304075235, "grad_norm": 4.632503986358643, "learning_rate": 9.617684742517942e-06, "loss": 1.2171, "step": 6482 }, { "epoch": 0.5080721003134796, "grad_norm": 5.442326068878174, "learning_rate": 9.617566473970233e-06, "loss": 1.5777, "step": 6483 }, { "epoch": 0.5081504702194357, "grad_norm": 2.376007318496704, "learning_rate": 9.617448187859673e-06, "loss": 1.4213, "step": 6484 }, { "epoch": 0.5082288401253918, "grad_norm": 2.447232961654663, "learning_rate": 9.617329884186712e-06, "loss": 1.4145, "step": 6485 }, { "epoch": 0.5083072100313479, "grad_norm": 2.6966521739959717, "learning_rate": 9.6172115629518e-06, "loss": 1.4415, "step": 6486 }, { "epoch": 0.508385579937304, "grad_norm": 2.5408709049224854, "learning_rate": 9.617093224155388e-06, "loss": 1.3813, "step": 6487 }, { "epoch": 0.5084639498432602, "grad_norm": 3.4012882709503174, "learning_rate": 9.616974867797923e-06, "loss": 1.196, "step": 6488 }, { "epoch": 0.5085423197492163, "grad_norm": 1.8991163969039917, "learning_rate": 9.616856493879857e-06, "loss": 1.3834, "step": 6489 }, { "epoch": 0.5086206896551724, "grad_norm": 10.659693717956543, "learning_rate": 9.616738102401641e-06, "loss": 1.065, "step": 6490 }, { "epoch": 0.5086990595611285, "grad_norm": 2.365330934524536, "learning_rate": 9.616619693363724e-06, "loss": 1.3393, "step": 6491 }, { "epoch": 0.5087774294670846, "grad_norm": 1.9523929357528687, "learning_rate": 9.616501266766557e-06, "loss": 1.3765, "step": 6492 }, { "epoch": 0.5088557993730407, "grad_norm": 3.8497586250305176, "learning_rate": 9.616382822610591e-06, "loss": 1.6169, "step": 6493 }, { "epoch": 0.5089341692789968, "grad_norm": 4.502904415130615, "learning_rate": 9.616264360896274e-06, "loss": 0.9973, "step": 6494 }, { "epoch": 0.5090125391849529, "grad_norm": 2.3850109577178955, "learning_rate": 9.616145881624061e-06, "loss": 1.4031, "step": 6495 }, { "epoch": 0.509090909090909, "grad_norm": 3.2403528690338135, "learning_rate": 9.6160273847944e-06, "loss": 1.2192, "step": 6496 }, { "epoch": 0.5091692789968651, "grad_norm": 2.574131965637207, "learning_rate": 9.61590887040774e-06, "loss": 1.4623, "step": 6497 }, { "epoch": 0.5092476489028214, "grad_norm": 2.6669740676879883, "learning_rate": 9.615790338464533e-06, "loss": 1.9775, "step": 6498 }, { "epoch": 0.5093260188087775, "grad_norm": 3.188054323196411, "learning_rate": 9.615671788965233e-06, "loss": 1.4013, "step": 6499 }, { "epoch": 0.5094043887147336, "grad_norm": 1.8664038181304932, "learning_rate": 9.615553221910287e-06, "loss": 1.4061, "step": 6500 }, { "epoch": 0.5094827586206897, "grad_norm": 12.903779029846191, "learning_rate": 9.615434637300147e-06, "loss": 2.1484, "step": 6501 }, { "epoch": 0.5095611285266458, "grad_norm": 3.0346691608428955, "learning_rate": 9.615316035135263e-06, "loss": 1.5347, "step": 6502 }, { "epoch": 0.5096394984326019, "grad_norm": 8.158230781555176, "learning_rate": 9.615197415416088e-06, "loss": 1.7565, "step": 6503 }, { "epoch": 0.509717868338558, "grad_norm": 2.9325954914093018, "learning_rate": 9.615078778143074e-06, "loss": 1.3897, "step": 6504 }, { "epoch": 0.5097962382445141, "grad_norm": 2.7150981426239014, "learning_rate": 9.61496012331667e-06, "loss": 1.4554, "step": 6505 }, { "epoch": 0.5098746081504703, "grad_norm": 4.458243370056152, "learning_rate": 9.61484145093733e-06, "loss": 1.5845, "step": 6506 }, { "epoch": 0.5099529780564264, "grad_norm": 2.5367493629455566, "learning_rate": 9.614722761005501e-06, "loss": 1.3155, "step": 6507 }, { "epoch": 0.5100313479623825, "grad_norm": 3.980393171310425, "learning_rate": 9.614604053521637e-06, "loss": 1.5177, "step": 6508 }, { "epoch": 0.5101097178683386, "grad_norm": 4.073419570922852, "learning_rate": 9.614485328486189e-06, "loss": 1.6406, "step": 6509 }, { "epoch": 0.5101880877742947, "grad_norm": 3.388918876647949, "learning_rate": 9.614366585899611e-06, "loss": 1.6474, "step": 6510 }, { "epoch": 0.5102664576802508, "grad_norm": 4.353759765625, "learning_rate": 9.614247825762351e-06, "loss": 1.4178, "step": 6511 }, { "epoch": 0.5103448275862069, "grad_norm": 2.513268232345581, "learning_rate": 9.614129048074864e-06, "loss": 1.0992, "step": 6512 }, { "epoch": 0.510423197492163, "grad_norm": 2.6707332134246826, "learning_rate": 9.614010252837597e-06, "loss": 1.7164, "step": 6513 }, { "epoch": 0.5105015673981191, "grad_norm": 3.428508758544922, "learning_rate": 9.613891440051009e-06, "loss": 1.1618, "step": 6514 }, { "epoch": 0.5105799373040752, "grad_norm": 10.348128318786621, "learning_rate": 9.613772609715545e-06, "loss": 1.7582, "step": 6515 }, { "epoch": 0.5106583072100314, "grad_norm": 4.279936790466309, "learning_rate": 9.61365376183166e-06, "loss": 1.5685, "step": 6516 }, { "epoch": 0.5107366771159875, "grad_norm": 2.172837972640991, "learning_rate": 9.613534896399806e-06, "loss": 1.2355, "step": 6517 }, { "epoch": 0.5108150470219436, "grad_norm": 3.1385657787323, "learning_rate": 9.613416013420436e-06, "loss": 1.4419, "step": 6518 }, { "epoch": 0.5108934169278997, "grad_norm": 3.4457108974456787, "learning_rate": 9.613297112894e-06, "loss": 1.439, "step": 6519 }, { "epoch": 0.5109717868338558, "grad_norm": 4.034270763397217, "learning_rate": 9.613178194820952e-06, "loss": 1.4492, "step": 6520 }, { "epoch": 0.5110501567398119, "grad_norm": 2.701693296432495, "learning_rate": 9.613059259201742e-06, "loss": 1.2626, "step": 6521 }, { "epoch": 0.511128526645768, "grad_norm": 2.762712240219116, "learning_rate": 9.612940306036824e-06, "loss": 1.3598, "step": 6522 }, { "epoch": 0.5112068965517241, "grad_norm": 2.5869462490081787, "learning_rate": 9.612821335326652e-06, "loss": 1.3844, "step": 6523 }, { "epoch": 0.5112852664576802, "grad_norm": 1.8974480628967285, "learning_rate": 9.612702347071676e-06, "loss": 1.3793, "step": 6524 }, { "epoch": 0.5113636363636364, "grad_norm": 3.0773422718048096, "learning_rate": 9.61258334127235e-06, "loss": 1.3673, "step": 6525 }, { "epoch": 0.5114420062695925, "grad_norm": 2.7388358116149902, "learning_rate": 9.612464317929127e-06, "loss": 1.5695, "step": 6526 }, { "epoch": 0.5115203761755486, "grad_norm": 4.405027389526367, "learning_rate": 9.612345277042458e-06, "loss": 1.5555, "step": 6527 }, { "epoch": 0.5115987460815047, "grad_norm": 2.2660207748413086, "learning_rate": 9.612226218612797e-06, "loss": 1.3476, "step": 6528 }, { "epoch": 0.5116771159874608, "grad_norm": 4.564686298370361, "learning_rate": 9.612107142640595e-06, "loss": 1.7975, "step": 6529 }, { "epoch": 0.5117554858934169, "grad_norm": 6.14462423324585, "learning_rate": 9.611988049126308e-06, "loss": 2.2247, "step": 6530 }, { "epoch": 0.511833855799373, "grad_norm": 4.800806999206543, "learning_rate": 9.611868938070386e-06, "loss": 1.582, "step": 6531 }, { "epoch": 0.5119122257053291, "grad_norm": 2.5567986965179443, "learning_rate": 9.611749809473286e-06, "loss": 1.277, "step": 6532 }, { "epoch": 0.5119905956112852, "grad_norm": 2.679508924484253, "learning_rate": 9.611630663335455e-06, "loss": 1.1924, "step": 6533 }, { "epoch": 0.5120689655172413, "grad_norm": 2.8077056407928467, "learning_rate": 9.61151149965735e-06, "loss": 1.4829, "step": 6534 }, { "epoch": 0.5121473354231975, "grad_norm": 4.449097633361816, "learning_rate": 9.611392318439427e-06, "loss": 1.7572, "step": 6535 }, { "epoch": 0.5122257053291536, "grad_norm": 13.768218994140625, "learning_rate": 9.611273119682134e-06, "loss": 0.7343, "step": 6536 }, { "epoch": 0.5123040752351097, "grad_norm": 2.563828468322754, "learning_rate": 9.611153903385926e-06, "loss": 1.3157, "step": 6537 }, { "epoch": 0.5123824451410658, "grad_norm": 2.680893898010254, "learning_rate": 9.611034669551259e-06, "loss": 1.4886, "step": 6538 }, { "epoch": 0.5124608150470219, "grad_norm": 3.323664903640747, "learning_rate": 9.610915418178583e-06, "loss": 1.3727, "step": 6539 }, { "epoch": 0.512539184952978, "grad_norm": 2.4930076599121094, "learning_rate": 9.610796149268353e-06, "loss": 1.9238, "step": 6540 }, { "epoch": 0.5126175548589341, "grad_norm": 3.670570135116577, "learning_rate": 9.610676862821022e-06, "loss": 1.3981, "step": 6541 }, { "epoch": 0.5126959247648902, "grad_norm": 2.6625778675079346, "learning_rate": 9.610557558837046e-06, "loss": 1.5633, "step": 6542 }, { "epoch": 0.5127742946708463, "grad_norm": 2.500089645385742, "learning_rate": 9.610438237316878e-06, "loss": 1.3627, "step": 6543 }, { "epoch": 0.5128526645768025, "grad_norm": 2.0526084899902344, "learning_rate": 9.610318898260969e-06, "loss": 1.4557, "step": 6544 }, { "epoch": 0.5129310344827587, "grad_norm": 2.3129212856292725, "learning_rate": 9.610199541669776e-06, "loss": 1.6401, "step": 6545 }, { "epoch": 0.5130094043887148, "grad_norm": 2.7030370235443115, "learning_rate": 9.61008016754375e-06, "loss": 1.0346, "step": 6546 }, { "epoch": 0.5130877742946709, "grad_norm": 2.185011863708496, "learning_rate": 9.609960775883349e-06, "loss": 1.3016, "step": 6547 }, { "epoch": 0.513166144200627, "grad_norm": 3.9032676219940186, "learning_rate": 9.609841366689024e-06, "loss": 1.9842, "step": 6548 }, { "epoch": 0.5132445141065831, "grad_norm": 2.081059217453003, "learning_rate": 9.609721939961231e-06, "loss": 1.4257, "step": 6549 }, { "epoch": 0.5133228840125392, "grad_norm": 2.419982671737671, "learning_rate": 9.609602495700422e-06, "loss": 1.4229, "step": 6550 }, { "epoch": 0.5134012539184953, "grad_norm": 2.6842997074127197, "learning_rate": 9.609483033907054e-06, "loss": 1.7666, "step": 6551 }, { "epoch": 0.5134796238244514, "grad_norm": 6.219422340393066, "learning_rate": 9.609363554581582e-06, "loss": 1.024, "step": 6552 }, { "epoch": 0.5135579937304076, "grad_norm": 6.595817565917969, "learning_rate": 9.609244057724457e-06, "loss": 1.524, "step": 6553 }, { "epoch": 0.5136363636363637, "grad_norm": 2.874450206756592, "learning_rate": 9.609124543336133e-06, "loss": 1.3458, "step": 6554 }, { "epoch": 0.5137147335423198, "grad_norm": 4.581703186035156, "learning_rate": 9.609005011417067e-06, "loss": 1.0772, "step": 6555 }, { "epoch": 0.5137931034482759, "grad_norm": 4.539485454559326, "learning_rate": 9.608885461967717e-06, "loss": 1.5972, "step": 6556 }, { "epoch": 0.513871473354232, "grad_norm": 2.153818368911743, "learning_rate": 9.60876589498853e-06, "loss": 1.2809, "step": 6557 }, { "epoch": 0.5139498432601881, "grad_norm": 2.3208978176116943, "learning_rate": 9.608646310479966e-06, "loss": 1.4929, "step": 6558 }, { "epoch": 0.5140282131661442, "grad_norm": 4.780057430267334, "learning_rate": 9.608526708442478e-06, "loss": 1.4297, "step": 6559 }, { "epoch": 0.5141065830721003, "grad_norm": 2.548729419708252, "learning_rate": 9.608407088876523e-06, "loss": 1.4655, "step": 6560 }, { "epoch": 0.5141849529780564, "grad_norm": 3.8769850730895996, "learning_rate": 9.608287451782555e-06, "loss": 1.4968, "step": 6561 }, { "epoch": 0.5142633228840126, "grad_norm": 3.8716676235198975, "learning_rate": 9.608167797161025e-06, "loss": 1.6788, "step": 6562 }, { "epoch": 0.5143416927899687, "grad_norm": 4.175941467285156, "learning_rate": 9.608048125012395e-06, "loss": 1.5491, "step": 6563 }, { "epoch": 0.5144200626959248, "grad_norm": 3.750922203063965, "learning_rate": 9.607928435337115e-06, "loss": 1.2836, "step": 6564 }, { "epoch": 0.5144984326018809, "grad_norm": 3.2637672424316406, "learning_rate": 9.607808728135644e-06, "loss": 1.2346, "step": 6565 }, { "epoch": 0.514576802507837, "grad_norm": 4.414997100830078, "learning_rate": 9.607689003408433e-06, "loss": 1.2854, "step": 6566 }, { "epoch": 0.5146551724137931, "grad_norm": 2.4505319595336914, "learning_rate": 9.60756926115594e-06, "loss": 1.2232, "step": 6567 }, { "epoch": 0.5147335423197492, "grad_norm": 2.661797523498535, "learning_rate": 9.607449501378622e-06, "loss": 1.6293, "step": 6568 }, { "epoch": 0.5148119122257053, "grad_norm": 2.1888370513916016, "learning_rate": 9.607329724076931e-06, "loss": 1.3135, "step": 6569 }, { "epoch": 0.5148902821316614, "grad_norm": 4.461125373840332, "learning_rate": 9.607209929251325e-06, "loss": 1.0573, "step": 6570 }, { "epoch": 0.5149686520376175, "grad_norm": 8.560619354248047, "learning_rate": 9.607090116902259e-06, "loss": 0.9316, "step": 6571 }, { "epoch": 0.5150470219435737, "grad_norm": 2.4440035820007324, "learning_rate": 9.606970287030187e-06, "loss": 1.451, "step": 6572 }, { "epoch": 0.5151253918495298, "grad_norm": 3.4268195629119873, "learning_rate": 9.606850439635566e-06, "loss": 1.4553, "step": 6573 }, { "epoch": 0.5152037617554859, "grad_norm": 2.735799789428711, "learning_rate": 9.606730574718854e-06, "loss": 1.0631, "step": 6574 }, { "epoch": 0.515282131661442, "grad_norm": 3.104280471801758, "learning_rate": 9.606610692280505e-06, "loss": 1.5282, "step": 6575 }, { "epoch": 0.5153605015673981, "grad_norm": 2.5620269775390625, "learning_rate": 9.606490792320975e-06, "loss": 1.2066, "step": 6576 }, { "epoch": 0.5154388714733542, "grad_norm": 4.487237453460693, "learning_rate": 9.606370874840718e-06, "loss": 2.2385, "step": 6577 }, { "epoch": 0.5155172413793103, "grad_norm": 3.2519757747650146, "learning_rate": 9.606250939840196e-06, "loss": 1.1692, "step": 6578 }, { "epoch": 0.5155956112852664, "grad_norm": 3.1955888271331787, "learning_rate": 9.606130987319858e-06, "loss": 1.1567, "step": 6579 }, { "epoch": 0.5156739811912225, "grad_norm": 6.189169883728027, "learning_rate": 9.606011017280166e-06, "loss": 0.3272, "step": 6580 }, { "epoch": 0.5157523510971787, "grad_norm": 3.6872622966766357, "learning_rate": 9.60589102972157e-06, "loss": 0.7671, "step": 6581 }, { "epoch": 0.5158307210031348, "grad_norm": 3.1343202590942383, "learning_rate": 9.605771024644535e-06, "loss": 1.2394, "step": 6582 }, { "epoch": 0.5159090909090909, "grad_norm": 3.0299465656280518, "learning_rate": 9.60565100204951e-06, "loss": 1.524, "step": 6583 }, { "epoch": 0.515987460815047, "grad_norm": 2.915456771850586, "learning_rate": 9.605530961936955e-06, "loss": 1.3657, "step": 6584 }, { "epoch": 0.5160658307210031, "grad_norm": 2.159141778945923, "learning_rate": 9.605410904307325e-06, "loss": 1.1641, "step": 6585 }, { "epoch": 0.5161442006269592, "grad_norm": 4.889033794403076, "learning_rate": 9.605290829161078e-06, "loss": 1.5701, "step": 6586 }, { "epoch": 0.5162225705329153, "grad_norm": 2.068718671798706, "learning_rate": 9.60517073649867e-06, "loss": 1.6157, "step": 6587 }, { "epoch": 0.5163009404388714, "grad_norm": 3.0114433765411377, "learning_rate": 9.605050626320558e-06, "loss": 1.4999, "step": 6588 }, { "epoch": 0.5163793103448275, "grad_norm": 3.1318814754486084, "learning_rate": 9.604930498627199e-06, "loss": 1.1502, "step": 6589 }, { "epoch": 0.5164576802507836, "grad_norm": 3.0941522121429443, "learning_rate": 9.60481035341905e-06, "loss": 1.5012, "step": 6590 }, { "epoch": 0.5165360501567398, "grad_norm": 2.08372163772583, "learning_rate": 9.604690190696567e-06, "loss": 1.2654, "step": 6591 }, { "epoch": 0.516614420062696, "grad_norm": 2.221644639968872, "learning_rate": 9.604570010460205e-06, "loss": 1.2723, "step": 6592 }, { "epoch": 0.5166927899686521, "grad_norm": 4.456066131591797, "learning_rate": 9.604449812710428e-06, "loss": 1.4411, "step": 6593 }, { "epoch": 0.5167711598746082, "grad_norm": 3.530325174331665, "learning_rate": 9.604329597447688e-06, "loss": 1.5165, "step": 6594 }, { "epoch": 0.5168495297805643, "grad_norm": 9.442866325378418, "learning_rate": 9.60420936467244e-06, "loss": 1.6836, "step": 6595 }, { "epoch": 0.5169278996865204, "grad_norm": 1.9530311822891235, "learning_rate": 9.604089114385149e-06, "loss": 1.1647, "step": 6596 }, { "epoch": 0.5170062695924765, "grad_norm": 7.559389114379883, "learning_rate": 9.603968846586265e-06, "loss": 1.8804, "step": 6597 }, { "epoch": 0.5170846394984326, "grad_norm": 5.205963134765625, "learning_rate": 9.603848561276248e-06, "loss": 1.275, "step": 6598 }, { "epoch": 0.5171630094043888, "grad_norm": 3.181485652923584, "learning_rate": 9.603728258455557e-06, "loss": 1.3339, "step": 6599 }, { "epoch": 0.5172413793103449, "grad_norm": 3.540264368057251, "learning_rate": 9.603607938124649e-06, "loss": 1.8688, "step": 6600 }, { "epoch": 0.517319749216301, "grad_norm": 3.849738359451294, "learning_rate": 9.603487600283977e-06, "loss": 1.7174, "step": 6601 }, { "epoch": 0.5173981191222571, "grad_norm": 7.066091537475586, "learning_rate": 9.603367244934008e-06, "loss": 1.6409, "step": 6602 }, { "epoch": 0.5174764890282132, "grad_norm": 6.234677314758301, "learning_rate": 9.60324687207519e-06, "loss": 1.3694, "step": 6603 }, { "epoch": 0.5175548589341693, "grad_norm": 4.6699018478393555, "learning_rate": 9.603126481707989e-06, "loss": 1.6568, "step": 6604 }, { "epoch": 0.5176332288401254, "grad_norm": 1.6662596464157104, "learning_rate": 9.603006073832858e-06, "loss": 1.1872, "step": 6605 }, { "epoch": 0.5177115987460815, "grad_norm": 3.7492241859436035, "learning_rate": 9.602885648450254e-06, "loss": 1.5377, "step": 6606 }, { "epoch": 0.5177899686520376, "grad_norm": 2.6584689617156982, "learning_rate": 9.602765205560639e-06, "loss": 1.8728, "step": 6607 }, { "epoch": 0.5178683385579937, "grad_norm": 3.3716487884521484, "learning_rate": 9.60264474516447e-06, "loss": 1.9709, "step": 6608 }, { "epoch": 0.5179467084639499, "grad_norm": 7.215356826782227, "learning_rate": 9.602524267262202e-06, "loss": 1.7503, "step": 6609 }, { "epoch": 0.518025078369906, "grad_norm": 3.836815595626831, "learning_rate": 9.602403771854299e-06, "loss": 1.6017, "step": 6610 }, { "epoch": 0.5181034482758621, "grad_norm": 2.614529848098755, "learning_rate": 9.602283258941212e-06, "loss": 1.3415, "step": 6611 }, { "epoch": 0.5181818181818182, "grad_norm": 1.9393436908721924, "learning_rate": 9.602162728523406e-06, "loss": 1.1082, "step": 6612 }, { "epoch": 0.5182601880877743, "grad_norm": 3.448870897293091, "learning_rate": 9.602042180601336e-06, "loss": 1.2314, "step": 6613 }, { "epoch": 0.5183385579937304, "grad_norm": 1.9017523527145386, "learning_rate": 9.601921615175462e-06, "loss": 1.4549, "step": 6614 }, { "epoch": 0.5184169278996865, "grad_norm": 2.9956490993499756, "learning_rate": 9.601801032246242e-06, "loss": 1.1389, "step": 6615 }, { "epoch": 0.5184952978056426, "grad_norm": 8.2184476852417, "learning_rate": 9.601680431814134e-06, "loss": 1.5665, "step": 6616 }, { "epoch": 0.5185736677115987, "grad_norm": 2.771726131439209, "learning_rate": 9.601559813879597e-06, "loss": 1.364, "step": 6617 }, { "epoch": 0.5186520376175549, "grad_norm": 2.244663715362549, "learning_rate": 9.60143917844309e-06, "loss": 1.3277, "step": 6618 }, { "epoch": 0.518730407523511, "grad_norm": 4.004142761230469, "learning_rate": 9.601318525505073e-06, "loss": 1.0914, "step": 6619 }, { "epoch": 0.5188087774294671, "grad_norm": 3.126652479171753, "learning_rate": 9.601197855066002e-06, "loss": 1.4253, "step": 6620 }, { "epoch": 0.5188871473354232, "grad_norm": 2.9390311241149902, "learning_rate": 9.601077167126339e-06, "loss": 1.6615, "step": 6621 }, { "epoch": 0.5189655172413793, "grad_norm": 7.054262161254883, "learning_rate": 9.600956461686541e-06, "loss": 1.8039, "step": 6622 }, { "epoch": 0.5190438871473354, "grad_norm": 3.847543716430664, "learning_rate": 9.600835738747068e-06, "loss": 1.3189, "step": 6623 }, { "epoch": 0.5191222570532915, "grad_norm": 5.837996959686279, "learning_rate": 9.600714998308377e-06, "loss": 1.6054, "step": 6624 }, { "epoch": 0.5192006269592476, "grad_norm": 2.783889055252075, "learning_rate": 9.600594240370933e-06, "loss": 1.4573, "step": 6625 }, { "epoch": 0.5192789968652037, "grad_norm": 3.9298996925354004, "learning_rate": 9.600473464935189e-06, "loss": 1.4459, "step": 6626 }, { "epoch": 0.5193573667711598, "grad_norm": 2.571728467941284, "learning_rate": 9.600352672001608e-06, "loss": 1.5824, "step": 6627 }, { "epoch": 0.519435736677116, "grad_norm": 3.0851333141326904, "learning_rate": 9.600231861570648e-06, "loss": 1.3374, "step": 6628 }, { "epoch": 0.5195141065830721, "grad_norm": 2.533228635787964, "learning_rate": 9.60011103364277e-06, "loss": 1.4055, "step": 6629 }, { "epoch": 0.5195924764890282, "grad_norm": 1.799471378326416, "learning_rate": 9.59999018821843e-06, "loss": 1.4193, "step": 6630 }, { "epoch": 0.5196708463949843, "grad_norm": 4.2213358879089355, "learning_rate": 9.599869325298092e-06, "loss": 1.7047, "step": 6631 }, { "epoch": 0.5197492163009404, "grad_norm": 2.820324182510376, "learning_rate": 9.599748444882212e-06, "loss": 1.4821, "step": 6632 }, { "epoch": 0.5198275862068965, "grad_norm": 2.876645565032959, "learning_rate": 9.599627546971252e-06, "loss": 1.5541, "step": 6633 }, { "epoch": 0.5199059561128526, "grad_norm": 2.197545051574707, "learning_rate": 9.599506631565673e-06, "loss": 1.0864, "step": 6634 }, { "epoch": 0.5199843260188087, "grad_norm": 3.4875380992889404, "learning_rate": 9.59938569866593e-06, "loss": 1.2597, "step": 6635 }, { "epoch": 0.5200626959247648, "grad_norm": 4.227700233459473, "learning_rate": 9.59926474827249e-06, "loss": 1.0323, "step": 6636 }, { "epoch": 0.520141065830721, "grad_norm": 2.46632719039917, "learning_rate": 9.599143780385807e-06, "loss": 1.5351, "step": 6637 }, { "epoch": 0.5202194357366771, "grad_norm": 2.1499505043029785, "learning_rate": 9.599022795006343e-06, "loss": 1.635, "step": 6638 }, { "epoch": 0.5202978056426333, "grad_norm": 2.6058123111724854, "learning_rate": 9.59890179213456e-06, "loss": 1.3838, "step": 6639 }, { "epoch": 0.5203761755485894, "grad_norm": 3.2056119441986084, "learning_rate": 9.598780771770916e-06, "loss": 1.4374, "step": 6640 }, { "epoch": 0.5204545454545455, "grad_norm": 3.3369412422180176, "learning_rate": 9.598659733915871e-06, "loss": 1.2731, "step": 6641 }, { "epoch": 0.5205329153605016, "grad_norm": 2.6850547790527344, "learning_rate": 9.598538678569888e-06, "loss": 1.0288, "step": 6642 }, { "epoch": 0.5206112852664577, "grad_norm": 2.4608235359191895, "learning_rate": 9.598417605733426e-06, "loss": 1.2879, "step": 6643 }, { "epoch": 0.5206896551724138, "grad_norm": 3.841212749481201, "learning_rate": 9.598296515406945e-06, "loss": 1.2506, "step": 6644 }, { "epoch": 0.52076802507837, "grad_norm": 3.9010698795318604, "learning_rate": 9.598175407590905e-06, "loss": 1.352, "step": 6645 }, { "epoch": 0.520846394984326, "grad_norm": 2.6680309772491455, "learning_rate": 9.598054282285767e-06, "loss": 1.5359, "step": 6646 }, { "epoch": 0.5209247648902822, "grad_norm": 3.95768404006958, "learning_rate": 9.597933139491993e-06, "loss": 1.3869, "step": 6647 }, { "epoch": 0.5210031347962383, "grad_norm": 3.0498738288879395, "learning_rate": 9.597811979210043e-06, "loss": 1.6426, "step": 6648 }, { "epoch": 0.5210815047021944, "grad_norm": 2.8857340812683105, "learning_rate": 9.597690801440378e-06, "loss": 1.603, "step": 6649 }, { "epoch": 0.5211598746081505, "grad_norm": 2.4867990016937256, "learning_rate": 9.597569606183458e-06, "loss": 1.2766, "step": 6650 }, { "epoch": 0.5212382445141066, "grad_norm": 3.418487548828125, "learning_rate": 9.597448393439747e-06, "loss": 1.6963, "step": 6651 }, { "epoch": 0.5213166144200627, "grad_norm": 3.6811583042144775, "learning_rate": 9.597327163209702e-06, "loss": 1.3242, "step": 6652 }, { "epoch": 0.5213949843260188, "grad_norm": 2.265869617462158, "learning_rate": 9.597205915493786e-06, "loss": 1.56, "step": 6653 }, { "epoch": 0.5214733542319749, "grad_norm": 2.7922754287719727, "learning_rate": 9.597084650292458e-06, "loss": 1.3681, "step": 6654 }, { "epoch": 0.521551724137931, "grad_norm": 3.282390832901001, "learning_rate": 9.596963367606185e-06, "loss": 1.4766, "step": 6655 }, { "epoch": 0.5216300940438872, "grad_norm": 3.1631312370300293, "learning_rate": 9.596842067435421e-06, "loss": 1.2497, "step": 6656 }, { "epoch": 0.5217084639498433, "grad_norm": 4.01832389831543, "learning_rate": 9.596720749780633e-06, "loss": 1.5592, "step": 6657 }, { "epoch": 0.5217868338557994, "grad_norm": 12.960613250732422, "learning_rate": 9.59659941464228e-06, "loss": 1.4086, "step": 6658 }, { "epoch": 0.5218652037617555, "grad_norm": 2.3280489444732666, "learning_rate": 9.596478062020823e-06, "loss": 1.3778, "step": 6659 }, { "epoch": 0.5219435736677116, "grad_norm": 4.0884881019592285, "learning_rate": 9.596356691916725e-06, "loss": 1.4171, "step": 6660 }, { "epoch": 0.5220219435736677, "grad_norm": 5.8014912605285645, "learning_rate": 9.596235304330446e-06, "loss": 1.4663, "step": 6661 }, { "epoch": 0.5221003134796238, "grad_norm": 2.6512115001678467, "learning_rate": 9.59611389926245e-06, "loss": 1.3773, "step": 6662 }, { "epoch": 0.5221786833855799, "grad_norm": 2.317997455596924, "learning_rate": 9.595992476713198e-06, "loss": 1.561, "step": 6663 }, { "epoch": 0.522257053291536, "grad_norm": 2.83563232421875, "learning_rate": 9.59587103668315e-06, "loss": 1.3403, "step": 6664 }, { "epoch": 0.5223354231974922, "grad_norm": 9.029862403869629, "learning_rate": 9.595749579172768e-06, "loss": 0.9131, "step": 6665 }, { "epoch": 0.5224137931034483, "grad_norm": 8.467248916625977, "learning_rate": 9.595628104182517e-06, "loss": 0.7717, "step": 6666 }, { "epoch": 0.5224921630094044, "grad_norm": 3.268369436264038, "learning_rate": 9.595506611712857e-06, "loss": 1.1197, "step": 6667 }, { "epoch": 0.5225705329153605, "grad_norm": 6.752072811126709, "learning_rate": 9.59538510176425e-06, "loss": 1.2053, "step": 6668 }, { "epoch": 0.5226489028213166, "grad_norm": 3.243098020553589, "learning_rate": 9.59526357433716e-06, "loss": 1.4721, "step": 6669 }, { "epoch": 0.5227272727272727, "grad_norm": 2.101762056350708, "learning_rate": 9.595142029432044e-06, "loss": 1.2524, "step": 6670 }, { "epoch": 0.5228056426332288, "grad_norm": 3.2831289768218994, "learning_rate": 9.59502046704937e-06, "loss": 1.4352, "step": 6671 }, { "epoch": 0.5228840125391849, "grad_norm": 2.0546011924743652, "learning_rate": 9.594898887189598e-06, "loss": 1.3596, "step": 6672 }, { "epoch": 0.522962382445141, "grad_norm": 2.6027727127075195, "learning_rate": 9.594777289853191e-06, "loss": 1.4676, "step": 6673 }, { "epoch": 0.5230407523510971, "grad_norm": 2.3084819316864014, "learning_rate": 9.594655675040612e-06, "loss": 1.5513, "step": 6674 }, { "epoch": 0.5231191222570533, "grad_norm": 2.603501558303833, "learning_rate": 9.594534042752322e-06, "loss": 1.5314, "step": 6675 }, { "epoch": 0.5231974921630094, "grad_norm": 3.2433435916900635, "learning_rate": 9.594412392988783e-06, "loss": 1.4401, "step": 6676 }, { "epoch": 0.5232758620689655, "grad_norm": 3.9590704441070557, "learning_rate": 9.594290725750462e-06, "loss": 1.9792, "step": 6677 }, { "epoch": 0.5233542319749216, "grad_norm": 2.9475624561309814, "learning_rate": 9.594169041037817e-06, "loss": 1.4664, "step": 6678 }, { "epoch": 0.5234326018808777, "grad_norm": 2.1747405529022217, "learning_rate": 9.594047338851313e-06, "loss": 1.4064, "step": 6679 }, { "epoch": 0.5235109717868338, "grad_norm": 1.9427870512008667, "learning_rate": 9.593925619191411e-06, "loss": 1.2488, "step": 6680 }, { "epoch": 0.5235893416927899, "grad_norm": 2.7203564643859863, "learning_rate": 9.593803882058577e-06, "loss": 1.7025, "step": 6681 }, { "epoch": 0.523667711598746, "grad_norm": 2.7125463485717773, "learning_rate": 9.593682127453273e-06, "loss": 1.5658, "step": 6682 }, { "epoch": 0.5237460815047021, "grad_norm": 2.3362224102020264, "learning_rate": 9.593560355375961e-06, "loss": 1.6206, "step": 6683 }, { "epoch": 0.5238244514106583, "grad_norm": 2.4975380897521973, "learning_rate": 9.593438565827105e-06, "loss": 1.5343, "step": 6684 }, { "epoch": 0.5239028213166145, "grad_norm": 2.527897357940674, "learning_rate": 9.593316758807168e-06, "loss": 1.6662, "step": 6685 }, { "epoch": 0.5239811912225706, "grad_norm": 2.3649404048919678, "learning_rate": 9.593194934316615e-06, "loss": 1.3933, "step": 6686 }, { "epoch": 0.5240595611285267, "grad_norm": 2.432352066040039, "learning_rate": 9.593073092355905e-06, "loss": 1.3549, "step": 6687 }, { "epoch": 0.5241379310344828, "grad_norm": 2.9070167541503906, "learning_rate": 9.592951232925504e-06, "loss": 1.2985, "step": 6688 }, { "epoch": 0.5242163009404389, "grad_norm": 2.2728843688964844, "learning_rate": 9.59282935602588e-06, "loss": 1.3114, "step": 6689 }, { "epoch": 0.524294670846395, "grad_norm": 2.9415202140808105, "learning_rate": 9.592707461657486e-06, "loss": 1.2097, "step": 6690 }, { "epoch": 0.5243730407523511, "grad_norm": 2.801117420196533, "learning_rate": 9.592585549820796e-06, "loss": 1.4405, "step": 6691 }, { "epoch": 0.5244514106583072, "grad_norm": 3.305368661880493, "learning_rate": 9.592463620516268e-06, "loss": 1.4149, "step": 6692 }, { "epoch": 0.5245297805642634, "grad_norm": 2.802277088165283, "learning_rate": 9.592341673744367e-06, "loss": 1.654, "step": 6693 }, { "epoch": 0.5246081504702195, "grad_norm": 8.290311813354492, "learning_rate": 9.592219709505557e-06, "loss": 1.0025, "step": 6694 }, { "epoch": 0.5246865203761756, "grad_norm": 2.2734391689300537, "learning_rate": 9.592097727800303e-06, "loss": 1.2991, "step": 6695 }, { "epoch": 0.5247648902821317, "grad_norm": 1.7881171703338623, "learning_rate": 9.591975728629068e-06, "loss": 1.5076, "step": 6696 }, { "epoch": 0.5248432601880878, "grad_norm": 2.0270793437957764, "learning_rate": 9.591853711992314e-06, "loss": 1.4995, "step": 6697 }, { "epoch": 0.5249216300940439, "grad_norm": 5.436161041259766, "learning_rate": 9.591731677890508e-06, "loss": 1.506, "step": 6698 }, { "epoch": 0.525, "grad_norm": 2.3602070808410645, "learning_rate": 9.591609626324114e-06, "loss": 1.3383, "step": 6699 }, { "epoch": 0.5250783699059561, "grad_norm": 7.797458648681641, "learning_rate": 9.591487557293595e-06, "loss": 1.0284, "step": 6700 }, { "epoch": 0.5251567398119122, "grad_norm": 2.735011339187622, "learning_rate": 9.591365470799417e-06, "loss": 1.329, "step": 6701 }, { "epoch": 0.5252351097178684, "grad_norm": 2.013730764389038, "learning_rate": 9.59124336684204e-06, "loss": 1.3054, "step": 6702 }, { "epoch": 0.5253134796238245, "grad_norm": 1.538028597831726, "learning_rate": 9.591121245421933e-06, "loss": 1.1875, "step": 6703 }, { "epoch": 0.5253918495297806, "grad_norm": 3.0420193672180176, "learning_rate": 9.59099910653956e-06, "loss": 1.5129, "step": 6704 }, { "epoch": 0.5254702194357367, "grad_norm": 3.0808920860290527, "learning_rate": 9.590876950195383e-06, "loss": 1.5826, "step": 6705 }, { "epoch": 0.5255485893416928, "grad_norm": 6.144175052642822, "learning_rate": 9.590754776389868e-06, "loss": 1.597, "step": 6706 }, { "epoch": 0.5256269592476489, "grad_norm": 2.7382068634033203, "learning_rate": 9.59063258512348e-06, "loss": 1.4203, "step": 6707 }, { "epoch": 0.525705329153605, "grad_norm": 4.981502532958984, "learning_rate": 9.590510376396684e-06, "loss": 1.1899, "step": 6708 }, { "epoch": 0.5257836990595611, "grad_norm": 2.873053789138794, "learning_rate": 9.590388150209944e-06, "loss": 1.509, "step": 6709 }, { "epoch": 0.5258620689655172, "grad_norm": 3.9349827766418457, "learning_rate": 9.590265906563725e-06, "loss": 1.69, "step": 6710 }, { "epoch": 0.5259404388714733, "grad_norm": 4.016086101531982, "learning_rate": 9.590143645458494e-06, "loss": 1.5635, "step": 6711 }, { "epoch": 0.5260188087774295, "grad_norm": 2.0939836502075195, "learning_rate": 9.590021366894712e-06, "loss": 1.757, "step": 6712 }, { "epoch": 0.5260971786833856, "grad_norm": 3.1598246097564697, "learning_rate": 9.589899070872846e-06, "loss": 1.5731, "step": 6713 }, { "epoch": 0.5261755485893417, "grad_norm": 4.025397300720215, "learning_rate": 9.589776757393363e-06, "loss": 1.4101, "step": 6714 }, { "epoch": 0.5262539184952978, "grad_norm": 2.349742889404297, "learning_rate": 9.589654426456726e-06, "loss": 1.2837, "step": 6715 }, { "epoch": 0.5263322884012539, "grad_norm": 3.268872022628784, "learning_rate": 9.589532078063401e-06, "loss": 1.5757, "step": 6716 }, { "epoch": 0.52641065830721, "grad_norm": 4.514135360717773, "learning_rate": 9.589409712213855e-06, "loss": 1.7505, "step": 6717 }, { "epoch": 0.5264890282131661, "grad_norm": 5.01845121383667, "learning_rate": 9.58928732890855e-06, "loss": 1.315, "step": 6718 }, { "epoch": 0.5265673981191222, "grad_norm": 2.232023239135742, "learning_rate": 9.589164928147954e-06, "loss": 1.0125, "step": 6719 }, { "epoch": 0.5266457680250783, "grad_norm": 3.990600824356079, "learning_rate": 9.58904250993253e-06, "loss": 1.6864, "step": 6720 }, { "epoch": 0.5267241379310345, "grad_norm": 2.584372043609619, "learning_rate": 9.588920074262747e-06, "loss": 1.9949, "step": 6721 }, { "epoch": 0.5268025078369906, "grad_norm": 3.1282799243927, "learning_rate": 9.58879762113907e-06, "loss": 1.1623, "step": 6722 }, { "epoch": 0.5268808777429467, "grad_norm": 2.4928855895996094, "learning_rate": 9.588675150561962e-06, "loss": 1.6152, "step": 6723 }, { "epoch": 0.5269592476489028, "grad_norm": 2.483630657196045, "learning_rate": 9.58855266253189e-06, "loss": 1.5879, "step": 6724 }, { "epoch": 0.5270376175548589, "grad_norm": 5.3293657302856445, "learning_rate": 9.588430157049322e-06, "loss": 1.6474, "step": 6725 }, { "epoch": 0.527115987460815, "grad_norm": 2.809842109680176, "learning_rate": 9.588307634114723e-06, "loss": 1.418, "step": 6726 }, { "epoch": 0.5271943573667711, "grad_norm": 3.965764284133911, "learning_rate": 9.588185093728556e-06, "loss": 1.6123, "step": 6727 }, { "epoch": 0.5272727272727272, "grad_norm": 1.8628873825073242, "learning_rate": 9.588062535891291e-06, "loss": 1.362, "step": 6728 }, { "epoch": 0.5273510971786833, "grad_norm": 2.928373336791992, "learning_rate": 9.587939960603393e-06, "loss": 1.5856, "step": 6729 }, { "epoch": 0.5274294670846394, "grad_norm": 10.826461791992188, "learning_rate": 9.587817367865328e-06, "loss": 1.0922, "step": 6730 }, { "epoch": 0.5275078369905956, "grad_norm": 2.610924005508423, "learning_rate": 9.587694757677561e-06, "loss": 1.2373, "step": 6731 }, { "epoch": 0.5275862068965518, "grad_norm": 4.414702892303467, "learning_rate": 9.587572130040562e-06, "loss": 1.6303, "step": 6732 }, { "epoch": 0.5276645768025079, "grad_norm": 3.9596025943756104, "learning_rate": 9.587449484954792e-06, "loss": 1.7946, "step": 6733 }, { "epoch": 0.527742946708464, "grad_norm": 4.005970478057861, "learning_rate": 9.587326822420721e-06, "loss": 1.8288, "step": 6734 }, { "epoch": 0.5278213166144201, "grad_norm": 6.914369106292725, "learning_rate": 9.587204142438817e-06, "loss": 1.8603, "step": 6735 }, { "epoch": 0.5278996865203762, "grad_norm": 2.548035144805908, "learning_rate": 9.587081445009543e-06, "loss": 1.5188, "step": 6736 }, { "epoch": 0.5279780564263323, "grad_norm": 2.5217819213867188, "learning_rate": 9.586958730133366e-06, "loss": 1.2514, "step": 6737 }, { "epoch": 0.5280564263322884, "grad_norm": 5.262386322021484, "learning_rate": 9.586835997810757e-06, "loss": 1.0157, "step": 6738 }, { "epoch": 0.5281347962382446, "grad_norm": 2.512917995452881, "learning_rate": 9.586713248042177e-06, "loss": 1.2918, "step": 6739 }, { "epoch": 0.5282131661442007, "grad_norm": 2.492231607437134, "learning_rate": 9.586590480828097e-06, "loss": 1.3853, "step": 6740 }, { "epoch": 0.5282915360501568, "grad_norm": 10.508563995361328, "learning_rate": 9.586467696168983e-06, "loss": 1.6674, "step": 6741 }, { "epoch": 0.5283699059561129, "grad_norm": 2.009857654571533, "learning_rate": 9.5863448940653e-06, "loss": 1.2852, "step": 6742 }, { "epoch": 0.528448275862069, "grad_norm": 4.198917865753174, "learning_rate": 9.58622207451752e-06, "loss": 1.5599, "step": 6743 }, { "epoch": 0.5285266457680251, "grad_norm": 2.3172879219055176, "learning_rate": 9.586099237526104e-06, "loss": 1.0371, "step": 6744 }, { "epoch": 0.5286050156739812, "grad_norm": 1.9910764694213867, "learning_rate": 9.585976383091523e-06, "loss": 1.5155, "step": 6745 }, { "epoch": 0.5286833855799373, "grad_norm": 2.836223840713501, "learning_rate": 9.585853511214244e-06, "loss": 1.5567, "step": 6746 }, { "epoch": 0.5287617554858934, "grad_norm": 2.3983376026153564, "learning_rate": 9.585730621894733e-06, "loss": 1.4907, "step": 6747 }, { "epoch": 0.5288401253918495, "grad_norm": 4.3601555824279785, "learning_rate": 9.585607715133459e-06, "loss": 1.4345, "step": 6748 }, { "epoch": 0.5289184952978057, "grad_norm": 2.5872178077697754, "learning_rate": 9.585484790930885e-06, "loss": 1.2374, "step": 6749 }, { "epoch": 0.5289968652037618, "grad_norm": 2.7206032276153564, "learning_rate": 9.585361849287486e-06, "loss": 1.4814, "step": 6750 }, { "epoch": 0.5290752351097179, "grad_norm": 3.791307210922241, "learning_rate": 9.585238890203725e-06, "loss": 1.3348, "step": 6751 }, { "epoch": 0.529153605015674, "grad_norm": 3.437648296356201, "learning_rate": 9.58511591368007e-06, "loss": 1.5825, "step": 6752 }, { "epoch": 0.5292319749216301, "grad_norm": 2.3003530502319336, "learning_rate": 9.584992919716989e-06, "loss": 1.4167, "step": 6753 }, { "epoch": 0.5293103448275862, "grad_norm": 3.082984685897827, "learning_rate": 9.584869908314948e-06, "loss": 1.1282, "step": 6754 }, { "epoch": 0.5293887147335423, "grad_norm": 2.6355865001678467, "learning_rate": 9.58474687947442e-06, "loss": 1.4662, "step": 6755 }, { "epoch": 0.5294670846394984, "grad_norm": 2.228463888168335, "learning_rate": 9.584623833195867e-06, "loss": 1.4964, "step": 6756 }, { "epoch": 0.5295454545454545, "grad_norm": 3.3891847133636475, "learning_rate": 9.584500769479762e-06, "loss": 1.6183, "step": 6757 }, { "epoch": 0.5296238244514107, "grad_norm": 3.76955509185791, "learning_rate": 9.58437768832657e-06, "loss": 0.6916, "step": 6758 }, { "epoch": 0.5297021943573668, "grad_norm": 2.6377811431884766, "learning_rate": 9.58425458973676e-06, "loss": 1.0674, "step": 6759 }, { "epoch": 0.5297805642633229, "grad_norm": 3.2271084785461426, "learning_rate": 9.5841314737108e-06, "loss": 1.5525, "step": 6760 }, { "epoch": 0.529858934169279, "grad_norm": 3.577136278152466, "learning_rate": 9.58400834024916e-06, "loss": 1.4033, "step": 6761 }, { "epoch": 0.5299373040752351, "grad_norm": 2.1182706356048584, "learning_rate": 9.583885189352305e-06, "loss": 1.5736, "step": 6762 }, { "epoch": 0.5300156739811912, "grad_norm": 4.718636989593506, "learning_rate": 9.583762021020706e-06, "loss": 1.3804, "step": 6763 }, { "epoch": 0.5300940438871473, "grad_norm": 1.8765325546264648, "learning_rate": 9.583638835254829e-06, "loss": 1.5229, "step": 6764 }, { "epoch": 0.5301724137931034, "grad_norm": 2.4020354747772217, "learning_rate": 9.583515632055145e-06, "loss": 1.0142, "step": 6765 }, { "epoch": 0.5302507836990595, "grad_norm": 4.513117790222168, "learning_rate": 9.583392411422122e-06, "loss": 1.6766, "step": 6766 }, { "epoch": 0.5303291536050156, "grad_norm": 2.9893102645874023, "learning_rate": 9.583269173356228e-06, "loss": 1.7222, "step": 6767 }, { "epoch": 0.5304075235109718, "grad_norm": 3.0163819789886475, "learning_rate": 9.583145917857934e-06, "loss": 1.2601, "step": 6768 }, { "epoch": 0.5304858934169279, "grad_norm": 2.996730327606201, "learning_rate": 9.583022644927706e-06, "loss": 1.7814, "step": 6769 }, { "epoch": 0.530564263322884, "grad_norm": 3.2076380252838135, "learning_rate": 9.582899354566013e-06, "loss": 1.2506, "step": 6770 }, { "epoch": 0.5306426332288401, "grad_norm": 4.4687042236328125, "learning_rate": 9.582776046773326e-06, "loss": 1.2517, "step": 6771 }, { "epoch": 0.5307210031347962, "grad_norm": 2.2751853466033936, "learning_rate": 9.582652721550111e-06, "loss": 1.5609, "step": 6772 }, { "epoch": 0.5307993730407523, "grad_norm": 2.7280397415161133, "learning_rate": 9.58252937889684e-06, "loss": 1.5679, "step": 6773 }, { "epoch": 0.5308777429467084, "grad_norm": 3.202688694000244, "learning_rate": 9.58240601881398e-06, "loss": 1.2758, "step": 6774 }, { "epoch": 0.5309561128526645, "grad_norm": 2.4688942432403564, "learning_rate": 9.582282641302003e-06, "loss": 1.1963, "step": 6775 }, { "epoch": 0.5310344827586206, "grad_norm": 2.8023619651794434, "learning_rate": 9.582159246361375e-06, "loss": 1.5023, "step": 6776 }, { "epoch": 0.5311128526645768, "grad_norm": 8.269502639770508, "learning_rate": 9.582035833992568e-06, "loss": 1.207, "step": 6777 }, { "epoch": 0.5311912225705329, "grad_norm": 2.4031736850738525, "learning_rate": 9.58191240419605e-06, "loss": 1.3377, "step": 6778 }, { "epoch": 0.5312695924764891, "grad_norm": 1.7357538938522339, "learning_rate": 9.58178895697229e-06, "loss": 1.4398, "step": 6779 }, { "epoch": 0.5313479623824452, "grad_norm": 4.269893169403076, "learning_rate": 9.581665492321759e-06, "loss": 1.8318, "step": 6780 }, { "epoch": 0.5314263322884013, "grad_norm": 5.861775875091553, "learning_rate": 9.581542010244924e-06, "loss": 0.8814, "step": 6781 }, { "epoch": 0.5315047021943574, "grad_norm": 3.494621753692627, "learning_rate": 9.581418510742258e-06, "loss": 1.735, "step": 6782 }, { "epoch": 0.5315830721003135, "grad_norm": 5.447520732879639, "learning_rate": 9.581294993814229e-06, "loss": 1.0364, "step": 6783 }, { "epoch": 0.5316614420062696, "grad_norm": 3.3095474243164062, "learning_rate": 9.581171459461306e-06, "loss": 1.3232, "step": 6784 }, { "epoch": 0.5317398119122257, "grad_norm": 2.04953932762146, "learning_rate": 9.581047907683961e-06, "loss": 1.4844, "step": 6785 }, { "epoch": 0.5318181818181819, "grad_norm": 4.944747447967529, "learning_rate": 9.580924338482662e-06, "loss": 1.8034, "step": 6786 }, { "epoch": 0.531896551724138, "grad_norm": 2.974426507949829, "learning_rate": 9.580800751857879e-06, "loss": 1.7312, "step": 6787 }, { "epoch": 0.5319749216300941, "grad_norm": 3.842552900314331, "learning_rate": 9.580677147810085e-06, "loss": 1.877, "step": 6788 }, { "epoch": 0.5320532915360502, "grad_norm": 5.7846550941467285, "learning_rate": 9.580553526339747e-06, "loss": 1.5042, "step": 6789 }, { "epoch": 0.5321316614420063, "grad_norm": 5.141730785369873, "learning_rate": 9.580429887447334e-06, "loss": 0.8067, "step": 6790 }, { "epoch": 0.5322100313479624, "grad_norm": 3.5148050785064697, "learning_rate": 9.580306231133321e-06, "loss": 1.5426, "step": 6791 }, { "epoch": 0.5322884012539185, "grad_norm": 3.9356186389923096, "learning_rate": 9.580182557398174e-06, "loss": 1.3575, "step": 6792 }, { "epoch": 0.5323667711598746, "grad_norm": 2.7668211460113525, "learning_rate": 9.580058866242365e-06, "loss": 1.1109, "step": 6793 }, { "epoch": 0.5324451410658307, "grad_norm": 4.5068182945251465, "learning_rate": 9.579935157666368e-06, "loss": 1.0048, "step": 6794 }, { "epoch": 0.5325235109717869, "grad_norm": 2.629767894744873, "learning_rate": 9.579811431670646e-06, "loss": 1.7176, "step": 6795 }, { "epoch": 0.532601880877743, "grad_norm": 2.3097989559173584, "learning_rate": 9.579687688255675e-06, "loss": 1.3606, "step": 6796 }, { "epoch": 0.5326802507836991, "grad_norm": 3.6078662872314453, "learning_rate": 9.579563927421925e-06, "loss": 1.3639, "step": 6797 }, { "epoch": 0.5327586206896552, "grad_norm": 3.5913336277008057, "learning_rate": 9.579440149169864e-06, "loss": 1.537, "step": 6798 }, { "epoch": 0.5328369905956113, "grad_norm": 4.150689601898193, "learning_rate": 9.579316353499967e-06, "loss": 1.3818, "step": 6799 }, { "epoch": 0.5329153605015674, "grad_norm": 2.7962896823883057, "learning_rate": 9.579192540412702e-06, "loss": 1.3636, "step": 6800 }, { "epoch": 0.5329937304075235, "grad_norm": 4.207581043243408, "learning_rate": 9.57906870990854e-06, "loss": 1.3554, "step": 6801 }, { "epoch": 0.5330721003134796, "grad_norm": 2.2631032466888428, "learning_rate": 9.578944861987954e-06, "loss": 1.6108, "step": 6802 }, { "epoch": 0.5331504702194357, "grad_norm": 3.2517123222351074, "learning_rate": 9.578820996651412e-06, "loss": 1.5543, "step": 6803 }, { "epoch": 0.5332288401253918, "grad_norm": 4.0738654136657715, "learning_rate": 9.578697113899388e-06, "loss": 1.4382, "step": 6804 }, { "epoch": 0.533307210031348, "grad_norm": 3.1941747665405273, "learning_rate": 9.57857321373235e-06, "loss": 1.6274, "step": 6805 }, { "epoch": 0.5333855799373041, "grad_norm": 2.5833611488342285, "learning_rate": 9.578449296150774e-06, "loss": 1.4627, "step": 6806 }, { "epoch": 0.5334639498432602, "grad_norm": 3.434988498687744, "learning_rate": 9.578325361155126e-06, "loss": 1.2983, "step": 6807 }, { "epoch": 0.5335423197492163, "grad_norm": 3.554857015609741, "learning_rate": 9.578201408745881e-06, "loss": 1.1137, "step": 6808 }, { "epoch": 0.5336206896551724, "grad_norm": 3.463141441345215, "learning_rate": 9.57807743892351e-06, "loss": 1.3716, "step": 6809 }, { "epoch": 0.5336990595611285, "grad_norm": 2.7156200408935547, "learning_rate": 9.577953451688483e-06, "loss": 1.5404, "step": 6810 }, { "epoch": 0.5337774294670846, "grad_norm": 5.578649997711182, "learning_rate": 9.577829447041273e-06, "loss": 1.6117, "step": 6811 }, { "epoch": 0.5338557993730407, "grad_norm": 2.7123918533325195, "learning_rate": 9.57770542498235e-06, "loss": 1.2437, "step": 6812 }, { "epoch": 0.5339341692789968, "grad_norm": 2.017977714538574, "learning_rate": 9.577581385512188e-06, "loss": 1.4295, "step": 6813 }, { "epoch": 0.534012539184953, "grad_norm": 2.5214810371398926, "learning_rate": 9.577457328631258e-06, "loss": 1.6676, "step": 6814 }, { "epoch": 0.5340909090909091, "grad_norm": 4.075376510620117, "learning_rate": 9.57733325434003e-06, "loss": 1.3533, "step": 6815 }, { "epoch": 0.5341692789968652, "grad_norm": 2.9888014793395996, "learning_rate": 9.57720916263898e-06, "loss": 1.4084, "step": 6816 }, { "epoch": 0.5342476489028213, "grad_norm": 3.1947004795074463, "learning_rate": 9.577085053528575e-06, "loss": 1.6181, "step": 6817 }, { "epoch": 0.5343260188087774, "grad_norm": 2.3600876331329346, "learning_rate": 9.576960927009289e-06, "loss": 1.4085, "step": 6818 }, { "epoch": 0.5344043887147335, "grad_norm": 7.926484107971191, "learning_rate": 9.576836783081595e-06, "loss": 1.6732, "step": 6819 }, { "epoch": 0.5344827586206896, "grad_norm": 3.0577025413513184, "learning_rate": 9.576712621745965e-06, "loss": 1.5442, "step": 6820 }, { "epoch": 0.5345611285266457, "grad_norm": 2.351229667663574, "learning_rate": 9.576588443002871e-06, "loss": 1.5927, "step": 6821 }, { "epoch": 0.5346394984326018, "grad_norm": 2.742143154144287, "learning_rate": 9.576464246852786e-06, "loss": 1.605, "step": 6822 }, { "epoch": 0.534717868338558, "grad_norm": 2.1824142932891846, "learning_rate": 9.57634003329618e-06, "loss": 1.895, "step": 6823 }, { "epoch": 0.534796238244514, "grad_norm": 2.8476717472076416, "learning_rate": 9.576215802333529e-06, "loss": 1.3458, "step": 6824 }, { "epoch": 0.5348746081504702, "grad_norm": 2.4490418434143066, "learning_rate": 9.5760915539653e-06, "loss": 1.4748, "step": 6825 }, { "epoch": 0.5349529780564264, "grad_norm": 4.211207866668701, "learning_rate": 9.575967288191973e-06, "loss": 1.7973, "step": 6826 }, { "epoch": 0.5350313479623825, "grad_norm": 5.29326868057251, "learning_rate": 9.575843005014017e-06, "loss": 1.4411, "step": 6827 }, { "epoch": 0.5351097178683386, "grad_norm": 2.0519936084747314, "learning_rate": 9.575718704431901e-06, "loss": 1.4791, "step": 6828 }, { "epoch": 0.5351880877742947, "grad_norm": 7.141120910644531, "learning_rate": 9.575594386446103e-06, "loss": 1.4502, "step": 6829 }, { "epoch": 0.5352664576802508, "grad_norm": 3.5426957607269287, "learning_rate": 9.575470051057094e-06, "loss": 1.4765, "step": 6830 }, { "epoch": 0.5353448275862069, "grad_norm": 3.478571653366089, "learning_rate": 9.575345698265348e-06, "loss": 1.2825, "step": 6831 }, { "epoch": 0.535423197492163, "grad_norm": 1.9865195751190186, "learning_rate": 9.575221328071338e-06, "loss": 1.5825, "step": 6832 }, { "epoch": 0.5355015673981192, "grad_norm": 1.804303526878357, "learning_rate": 9.575096940475533e-06, "loss": 1.4993, "step": 6833 }, { "epoch": 0.5355799373040753, "grad_norm": 2.279508113861084, "learning_rate": 9.57497253547841e-06, "loss": 1.2391, "step": 6834 }, { "epoch": 0.5356583072100314, "grad_norm": 2.581144332885742, "learning_rate": 9.574848113080443e-06, "loss": 1.4077, "step": 6835 }, { "epoch": 0.5357366771159875, "grad_norm": 4.026242733001709, "learning_rate": 9.574723673282103e-06, "loss": 1.4115, "step": 6836 }, { "epoch": 0.5358150470219436, "grad_norm": 3.2797274589538574, "learning_rate": 9.574599216083862e-06, "loss": 1.3276, "step": 6837 }, { "epoch": 0.5358934169278997, "grad_norm": 2.5310821533203125, "learning_rate": 9.574474741486198e-06, "loss": 1.1265, "step": 6838 }, { "epoch": 0.5359717868338558, "grad_norm": 4.838089466094971, "learning_rate": 9.57435024948958e-06, "loss": 1.397, "step": 6839 }, { "epoch": 0.5360501567398119, "grad_norm": 3.116508722305298, "learning_rate": 9.574225740094483e-06, "loss": 1.3018, "step": 6840 }, { "epoch": 0.536128526645768, "grad_norm": 2.78192138671875, "learning_rate": 9.574101213301382e-06, "loss": 1.1233, "step": 6841 }, { "epoch": 0.5362068965517242, "grad_norm": 2.50933837890625, "learning_rate": 9.573976669110747e-06, "loss": 1.2066, "step": 6842 }, { "epoch": 0.5362852664576803, "grad_norm": 3.8864080905914307, "learning_rate": 9.573852107523056e-06, "loss": 1.1505, "step": 6843 }, { "epoch": 0.5363636363636364, "grad_norm": 2.7122960090637207, "learning_rate": 9.573727528538781e-06, "loss": 0.988, "step": 6844 }, { "epoch": 0.5364420062695925, "grad_norm": 5.168392658233643, "learning_rate": 9.573602932158396e-06, "loss": 1.3553, "step": 6845 }, { "epoch": 0.5365203761755486, "grad_norm": 7.216432571411133, "learning_rate": 9.573478318382374e-06, "loss": 1.5229, "step": 6846 }, { "epoch": 0.5365987460815047, "grad_norm": 2.5963401794433594, "learning_rate": 9.573353687211188e-06, "loss": 1.5932, "step": 6847 }, { "epoch": 0.5366771159874608, "grad_norm": 3.3339219093322754, "learning_rate": 9.573229038645315e-06, "loss": 1.5962, "step": 6848 }, { "epoch": 0.5367554858934169, "grad_norm": 1.9833918809890747, "learning_rate": 9.57310437268523e-06, "loss": 1.4387, "step": 6849 }, { "epoch": 0.536833855799373, "grad_norm": 3.2861037254333496, "learning_rate": 9.572979689331402e-06, "loss": 1.7674, "step": 6850 }, { "epoch": 0.5369122257053291, "grad_norm": 3.5574538707733154, "learning_rate": 9.572854988584308e-06, "loss": 1.1707, "step": 6851 }, { "epoch": 0.5369905956112853, "grad_norm": 2.939194917678833, "learning_rate": 9.572730270444424e-06, "loss": 1.3314, "step": 6852 }, { "epoch": 0.5370689655172414, "grad_norm": 2.5245726108551025, "learning_rate": 9.57260553491222e-06, "loss": 1.1151, "step": 6853 }, { "epoch": 0.5371473354231975, "grad_norm": 2.2483911514282227, "learning_rate": 9.572480781988176e-06, "loss": 1.6498, "step": 6854 }, { "epoch": 0.5372257053291536, "grad_norm": 3.003416061401367, "learning_rate": 9.572356011672763e-06, "loss": 1.8575, "step": 6855 }, { "epoch": 0.5373040752351097, "grad_norm": 4.531835079193115, "learning_rate": 9.572231223966454e-06, "loss": 1.7156, "step": 6856 }, { "epoch": 0.5373824451410658, "grad_norm": 3.0288002490997314, "learning_rate": 9.57210641886973e-06, "loss": 1.5564, "step": 6857 }, { "epoch": 0.5374608150470219, "grad_norm": 2.8705174922943115, "learning_rate": 9.57198159638306e-06, "loss": 1.615, "step": 6858 }, { "epoch": 0.537539184952978, "grad_norm": 3.4628491401672363, "learning_rate": 9.571856756506918e-06, "loss": 0.7897, "step": 6859 }, { "epoch": 0.5376175548589341, "grad_norm": 7.200858116149902, "learning_rate": 9.571731899241783e-06, "loss": 1.3377, "step": 6860 }, { "epoch": 0.5376959247648903, "grad_norm": 2.4643642902374268, "learning_rate": 9.571607024588127e-06, "loss": 1.2133, "step": 6861 }, { "epoch": 0.5377742946708464, "grad_norm": 2.678769111633301, "learning_rate": 9.571482132546428e-06, "loss": 1.5773, "step": 6862 }, { "epoch": 0.5378526645768025, "grad_norm": 2.3900721073150635, "learning_rate": 9.571357223117157e-06, "loss": 1.5825, "step": 6863 }, { "epoch": 0.5379310344827586, "grad_norm": 2.467525005340576, "learning_rate": 9.571232296300793e-06, "loss": 1.6583, "step": 6864 }, { "epoch": 0.5380094043887147, "grad_norm": 2.8277177810668945, "learning_rate": 9.571107352097808e-06, "loss": 1.167, "step": 6865 }, { "epoch": 0.5380877742946708, "grad_norm": 8.874598503112793, "learning_rate": 9.57098239050868e-06, "loss": 0.4961, "step": 6866 }, { "epoch": 0.5381661442006269, "grad_norm": 2.098104476928711, "learning_rate": 9.57085741153388e-06, "loss": 1.5647, "step": 6867 }, { "epoch": 0.538244514106583, "grad_norm": 3.4491078853607178, "learning_rate": 9.57073241517389e-06, "loss": 1.1159, "step": 6868 }, { "epoch": 0.5383228840125391, "grad_norm": 6.024547100067139, "learning_rate": 9.570607401429178e-06, "loss": 0.5094, "step": 6869 }, { "epoch": 0.5384012539184952, "grad_norm": 2.3807458877563477, "learning_rate": 9.570482370300224e-06, "loss": 1.3894, "step": 6870 }, { "epoch": 0.5384796238244514, "grad_norm": 10.50916576385498, "learning_rate": 9.570357321787503e-06, "loss": 0.9943, "step": 6871 }, { "epoch": 0.5385579937304075, "grad_norm": 2.6656525135040283, "learning_rate": 9.57023225589149e-06, "loss": 1.6961, "step": 6872 }, { "epoch": 0.5386363636363637, "grad_norm": 2.073054552078247, "learning_rate": 9.570107172612661e-06, "loss": 1.4115, "step": 6873 }, { "epoch": 0.5387147335423198, "grad_norm": 2.7285373210906982, "learning_rate": 9.569982071951494e-06, "loss": 1.3602, "step": 6874 }, { "epoch": 0.5387931034482759, "grad_norm": 2.4352896213531494, "learning_rate": 9.56985695390846e-06, "loss": 1.6605, "step": 6875 }, { "epoch": 0.538871473354232, "grad_norm": 4.992705345153809, "learning_rate": 9.569731818484038e-06, "loss": 1.5468, "step": 6876 }, { "epoch": 0.5389498432601881, "grad_norm": 3.3844614028930664, "learning_rate": 9.569606665678701e-06, "loss": 1.5362, "step": 6877 }, { "epoch": 0.5390282131661442, "grad_norm": 2.2099318504333496, "learning_rate": 9.56948149549293e-06, "loss": 1.3465, "step": 6878 }, { "epoch": 0.5391065830721004, "grad_norm": 3.5944154262542725, "learning_rate": 9.569356307927198e-06, "loss": 1.9276, "step": 6879 }, { "epoch": 0.5391849529780565, "grad_norm": 7.12248420715332, "learning_rate": 9.569231102981982e-06, "loss": 1.2145, "step": 6880 }, { "epoch": 0.5392633228840126, "grad_norm": 2.5245916843414307, "learning_rate": 9.569105880657756e-06, "loss": 1.3221, "step": 6881 }, { "epoch": 0.5393416927899687, "grad_norm": 2.6531758308410645, "learning_rate": 9.568980640955e-06, "loss": 1.4034, "step": 6882 }, { "epoch": 0.5394200626959248, "grad_norm": 3.9804019927978516, "learning_rate": 9.568855383874188e-06, "loss": 1.4991, "step": 6883 }, { "epoch": 0.5394984326018809, "grad_norm": 3.3974921703338623, "learning_rate": 9.568730109415795e-06, "loss": 1.4029, "step": 6884 }, { "epoch": 0.539576802507837, "grad_norm": 2.7805700302124023, "learning_rate": 9.568604817580303e-06, "loss": 1.295, "step": 6885 }, { "epoch": 0.5396551724137931, "grad_norm": 2.8825297355651855, "learning_rate": 9.568479508368181e-06, "loss": 1.2325, "step": 6886 }, { "epoch": 0.5397335423197492, "grad_norm": 4.62895393371582, "learning_rate": 9.568354181779912e-06, "loss": 1.3182, "step": 6887 }, { "epoch": 0.5398119122257053, "grad_norm": 2.34399676322937, "learning_rate": 9.56822883781597e-06, "loss": 1.391, "step": 6888 }, { "epoch": 0.5398902821316615, "grad_norm": 2.709808349609375, "learning_rate": 9.56810347647683e-06, "loss": 1.6132, "step": 6889 }, { "epoch": 0.5399686520376176, "grad_norm": 3.7804758548736572, "learning_rate": 9.567978097762973e-06, "loss": 1.2728, "step": 6890 }, { "epoch": 0.5400470219435737, "grad_norm": 4.122189044952393, "learning_rate": 9.56785270167487e-06, "loss": 0.9708, "step": 6891 }, { "epoch": 0.5401253918495298, "grad_norm": 3.851893186569214, "learning_rate": 9.567727288213005e-06, "loss": 1.394, "step": 6892 }, { "epoch": 0.5402037617554859, "grad_norm": 3.0718891620635986, "learning_rate": 9.567601857377851e-06, "loss": 1.3623, "step": 6893 }, { "epoch": 0.540282131661442, "grad_norm": 3.5688328742980957, "learning_rate": 9.567476409169884e-06, "loss": 1.5164, "step": 6894 }, { "epoch": 0.5403605015673981, "grad_norm": 4.2882843017578125, "learning_rate": 9.567350943589585e-06, "loss": 0.8592, "step": 6895 }, { "epoch": 0.5404388714733542, "grad_norm": 3.3148293495178223, "learning_rate": 9.567225460637429e-06, "loss": 2.0487, "step": 6896 }, { "epoch": 0.5405172413793103, "grad_norm": 3.3425285816192627, "learning_rate": 9.567099960313891e-06, "loss": 1.2595, "step": 6897 }, { "epoch": 0.5405956112852665, "grad_norm": 2.4902071952819824, "learning_rate": 9.56697444261945e-06, "loss": 1.5743, "step": 6898 }, { "epoch": 0.5406739811912226, "grad_norm": 3.0823538303375244, "learning_rate": 9.566848907554587e-06, "loss": 1.6645, "step": 6899 }, { "epoch": 0.5407523510971787, "grad_norm": 2.135594606399536, "learning_rate": 9.566723355119775e-06, "loss": 1.3088, "step": 6900 }, { "epoch": 0.5408307210031348, "grad_norm": 3.6105079650878906, "learning_rate": 9.566597785315492e-06, "loss": 1.5256, "step": 6901 }, { "epoch": 0.5409090909090909, "grad_norm": 3.598926544189453, "learning_rate": 9.566472198142218e-06, "loss": 1.5915, "step": 6902 }, { "epoch": 0.540987460815047, "grad_norm": 3.5385591983795166, "learning_rate": 9.566346593600428e-06, "loss": 1.7539, "step": 6903 }, { "epoch": 0.5410658307210031, "grad_norm": 2.0635251998901367, "learning_rate": 9.566220971690603e-06, "loss": 1.236, "step": 6904 }, { "epoch": 0.5411442006269592, "grad_norm": 3.7417819499969482, "learning_rate": 9.566095332413216e-06, "loss": 0.9872, "step": 6905 }, { "epoch": 0.5412225705329153, "grad_norm": 2.307451009750366, "learning_rate": 9.56596967576875e-06, "loss": 1.4574, "step": 6906 }, { "epoch": 0.5413009404388714, "grad_norm": 2.803276538848877, "learning_rate": 9.565844001757681e-06, "loss": 1.7261, "step": 6907 }, { "epoch": 0.5413793103448276, "grad_norm": 1.9714179039001465, "learning_rate": 9.565718310380485e-06, "loss": 1.2009, "step": 6908 }, { "epoch": 0.5414576802507837, "grad_norm": 3.430087089538574, "learning_rate": 9.565592601637641e-06, "loss": 0.7512, "step": 6909 }, { "epoch": 0.5415360501567398, "grad_norm": 3.215052604675293, "learning_rate": 9.56546687552963e-06, "loss": 1.5095, "step": 6910 }, { "epoch": 0.5416144200626959, "grad_norm": 2.8942294120788574, "learning_rate": 9.565341132056926e-06, "loss": 1.429, "step": 6911 }, { "epoch": 0.541692789968652, "grad_norm": 3.452195882797241, "learning_rate": 9.56521537122001e-06, "loss": 1.4605, "step": 6912 }, { "epoch": 0.5417711598746081, "grad_norm": 1.7727469205856323, "learning_rate": 9.565089593019358e-06, "loss": 1.3585, "step": 6913 }, { "epoch": 0.5418495297805642, "grad_norm": 3.780470609664917, "learning_rate": 9.564963797455453e-06, "loss": 1.2625, "step": 6914 }, { "epoch": 0.5419278996865203, "grad_norm": 3.142446756362915, "learning_rate": 9.56483798452877e-06, "loss": 1.5963, "step": 6915 }, { "epoch": 0.5420062695924764, "grad_norm": 2.312425136566162, "learning_rate": 9.564712154239786e-06, "loss": 1.3896, "step": 6916 }, { "epoch": 0.5420846394984326, "grad_norm": 5.737194061279297, "learning_rate": 9.564586306588983e-06, "loss": 1.0659, "step": 6917 }, { "epoch": 0.5421630094043887, "grad_norm": 2.2311370372772217, "learning_rate": 9.564460441576838e-06, "loss": 1.1914, "step": 6918 }, { "epoch": 0.5422413793103448, "grad_norm": 2.1269192695617676, "learning_rate": 9.564334559203829e-06, "loss": 1.3572, "step": 6919 }, { "epoch": 0.542319749216301, "grad_norm": 4.237939357757568, "learning_rate": 9.564208659470434e-06, "loss": 1.4324, "step": 6920 }, { "epoch": 0.5423981191222571, "grad_norm": 4.882114887237549, "learning_rate": 9.564082742377139e-06, "loss": 1.6406, "step": 6921 }, { "epoch": 0.5424764890282132, "grad_norm": 3.715916395187378, "learning_rate": 9.563956807924413e-06, "loss": 1.4611, "step": 6922 }, { "epoch": 0.5425548589341693, "grad_norm": 2.3594205379486084, "learning_rate": 9.563830856112741e-06, "loss": 1.203, "step": 6923 }, { "epoch": 0.5426332288401254, "grad_norm": 3.557976484298706, "learning_rate": 9.563704886942601e-06, "loss": 1.7141, "step": 6924 }, { "epoch": 0.5427115987460815, "grad_norm": 2.610599994659424, "learning_rate": 9.56357890041447e-06, "loss": 1.3627, "step": 6925 }, { "epoch": 0.5427899686520377, "grad_norm": 2.1234819889068604, "learning_rate": 9.563452896528831e-06, "loss": 1.1843, "step": 6926 }, { "epoch": 0.5428683385579938, "grad_norm": 3.1474740505218506, "learning_rate": 9.563326875286161e-06, "loss": 1.5203, "step": 6927 }, { "epoch": 0.5429467084639499, "grad_norm": 8.143006324768066, "learning_rate": 9.563200836686938e-06, "loss": 0.8695, "step": 6928 }, { "epoch": 0.543025078369906, "grad_norm": 2.7677383422851562, "learning_rate": 9.563074780731642e-06, "loss": 1.1878, "step": 6929 }, { "epoch": 0.5431034482758621, "grad_norm": 2.7829394340515137, "learning_rate": 9.562948707420756e-06, "loss": 1.3916, "step": 6930 }, { "epoch": 0.5431818181818182, "grad_norm": 2.553192377090454, "learning_rate": 9.562822616754756e-06, "loss": 1.4342, "step": 6931 }, { "epoch": 0.5432601880877743, "grad_norm": 4.753356456756592, "learning_rate": 9.562696508734122e-06, "loss": 1.5153, "step": 6932 }, { "epoch": 0.5433385579937304, "grad_norm": 2.9837138652801514, "learning_rate": 9.562570383359333e-06, "loss": 1.4247, "step": 6933 }, { "epoch": 0.5434169278996865, "grad_norm": 3.355069637298584, "learning_rate": 9.56244424063087e-06, "loss": 1.3258, "step": 6934 }, { "epoch": 0.5434952978056427, "grad_norm": 2.5281527042388916, "learning_rate": 9.562318080549214e-06, "loss": 2.0274, "step": 6935 }, { "epoch": 0.5435736677115988, "grad_norm": 2.9500885009765625, "learning_rate": 9.562191903114843e-06, "loss": 1.4138, "step": 6936 }, { "epoch": 0.5436520376175549, "grad_norm": 4.999754428863525, "learning_rate": 9.562065708328236e-06, "loss": 1.4349, "step": 6937 }, { "epoch": 0.543730407523511, "grad_norm": 8.055442810058594, "learning_rate": 9.561939496189875e-06, "loss": 1.9707, "step": 6938 }, { "epoch": 0.5438087774294671, "grad_norm": 2.6218464374542236, "learning_rate": 9.561813266700239e-06, "loss": 1.0975, "step": 6939 }, { "epoch": 0.5438871473354232, "grad_norm": 6.128293037414551, "learning_rate": 9.56168701985981e-06, "loss": 1.49, "step": 6940 }, { "epoch": 0.5439655172413793, "grad_norm": 2.10662579536438, "learning_rate": 9.561560755669066e-06, "loss": 1.2417, "step": 6941 }, { "epoch": 0.5440438871473354, "grad_norm": 3.9361460208892822, "learning_rate": 9.561434474128486e-06, "loss": 1.6447, "step": 6942 }, { "epoch": 0.5441222570532915, "grad_norm": 2.83111310005188, "learning_rate": 9.561308175238553e-06, "loss": 1.2405, "step": 6943 }, { "epoch": 0.5442006269592476, "grad_norm": 2.197024345397949, "learning_rate": 9.561181858999748e-06, "loss": 1.3575, "step": 6944 }, { "epoch": 0.5442789968652038, "grad_norm": 6.86292839050293, "learning_rate": 9.56105552541255e-06, "loss": 2.1959, "step": 6945 }, { "epoch": 0.5443573667711599, "grad_norm": 4.389736652374268, "learning_rate": 9.560929174477438e-06, "loss": 1.5868, "step": 6946 }, { "epoch": 0.544435736677116, "grad_norm": 3.3213727474212646, "learning_rate": 9.560802806194894e-06, "loss": 1.4751, "step": 6947 }, { "epoch": 0.5445141065830721, "grad_norm": 2.6372270584106445, "learning_rate": 9.5606764205654e-06, "loss": 1.3787, "step": 6948 }, { "epoch": 0.5445924764890282, "grad_norm": 2.52400803565979, "learning_rate": 9.560550017589433e-06, "loss": 1.383, "step": 6949 }, { "epoch": 0.5446708463949843, "grad_norm": 3.0289294719696045, "learning_rate": 9.560423597267479e-06, "loss": 1.6039, "step": 6950 }, { "epoch": 0.5447492163009404, "grad_norm": 2.927185535430908, "learning_rate": 9.560297159600014e-06, "loss": 1.3395, "step": 6951 }, { "epoch": 0.5448275862068965, "grad_norm": 2.3445446491241455, "learning_rate": 9.560170704587523e-06, "loss": 1.407, "step": 6952 }, { "epoch": 0.5449059561128526, "grad_norm": 3.065356731414795, "learning_rate": 9.560044232230483e-06, "loss": 1.5756, "step": 6953 }, { "epoch": 0.5449843260188088, "grad_norm": 4.216375350952148, "learning_rate": 9.559917742529378e-06, "loss": 1.7869, "step": 6954 }, { "epoch": 0.5450626959247649, "grad_norm": 2.704455614089966, "learning_rate": 9.559791235484688e-06, "loss": 1.6505, "step": 6955 }, { "epoch": 0.545141065830721, "grad_norm": 3.8826372623443604, "learning_rate": 9.559664711096893e-06, "loss": 1.2972, "step": 6956 }, { "epoch": 0.5452194357366771, "grad_norm": 3.9593112468719482, "learning_rate": 9.559538169366475e-06, "loss": 1.6596, "step": 6957 }, { "epoch": 0.5452978056426332, "grad_norm": 2.4903993606567383, "learning_rate": 9.559411610293919e-06, "loss": 1.0554, "step": 6958 }, { "epoch": 0.5453761755485893, "grad_norm": 3.072155475616455, "learning_rate": 9.5592850338797e-06, "loss": 1.2063, "step": 6959 }, { "epoch": 0.5454545454545454, "grad_norm": 5.298233985900879, "learning_rate": 9.559158440124305e-06, "loss": 0.992, "step": 6960 }, { "epoch": 0.5455329153605015, "grad_norm": 5.184647083282471, "learning_rate": 9.559031829028211e-06, "loss": 1.1917, "step": 6961 }, { "epoch": 0.5456112852664576, "grad_norm": 2.6773440837860107, "learning_rate": 9.558905200591901e-06, "loss": 1.3855, "step": 6962 }, { "epoch": 0.5456896551724137, "grad_norm": 1.8841110467910767, "learning_rate": 9.55877855481586e-06, "loss": 1.1492, "step": 6963 }, { "epoch": 0.5457680250783699, "grad_norm": 5.273311138153076, "learning_rate": 9.558651891700564e-06, "loss": 1.0453, "step": 6964 }, { "epoch": 0.545846394984326, "grad_norm": 2.4670298099517822, "learning_rate": 9.5585252112465e-06, "loss": 1.5274, "step": 6965 }, { "epoch": 0.5459247648902821, "grad_norm": 2.8485748767852783, "learning_rate": 9.558398513454145e-06, "loss": 1.6082, "step": 6966 }, { "epoch": 0.5460031347962383, "grad_norm": 3.6593716144561768, "learning_rate": 9.558271798323986e-06, "loss": 1.3802, "step": 6967 }, { "epoch": 0.5460815047021944, "grad_norm": 2.35341739654541, "learning_rate": 9.5581450658565e-06, "loss": 1.0989, "step": 6968 }, { "epoch": 0.5461598746081505, "grad_norm": 3.357577323913574, "learning_rate": 9.558018316052174e-06, "loss": 1.3428, "step": 6969 }, { "epoch": 0.5462382445141066, "grad_norm": 2.6670477390289307, "learning_rate": 9.557891548911486e-06, "loss": 1.7115, "step": 6970 }, { "epoch": 0.5463166144200627, "grad_norm": 4.0752854347229, "learning_rate": 9.55776476443492e-06, "loss": 1.163, "step": 6971 }, { "epoch": 0.5463949843260189, "grad_norm": 2.35559344291687, "learning_rate": 9.557637962622957e-06, "loss": 1.5018, "step": 6972 }, { "epoch": 0.546473354231975, "grad_norm": 2.5127151012420654, "learning_rate": 9.557511143476081e-06, "loss": 1.3849, "step": 6973 }, { "epoch": 0.5465517241379311, "grad_norm": 2.731563091278076, "learning_rate": 9.557384306994773e-06, "loss": 1.4645, "step": 6974 }, { "epoch": 0.5466300940438872, "grad_norm": 2.5737202167510986, "learning_rate": 9.557257453179516e-06, "loss": 0.9675, "step": 6975 }, { "epoch": 0.5467084639498433, "grad_norm": 1.9790791273117065, "learning_rate": 9.557130582030793e-06, "loss": 1.4234, "step": 6976 }, { "epoch": 0.5467868338557994, "grad_norm": 2.2908501625061035, "learning_rate": 9.557003693549085e-06, "loss": 1.4283, "step": 6977 }, { "epoch": 0.5468652037617555, "grad_norm": 5.263824939727783, "learning_rate": 9.556876787734877e-06, "loss": 1.1879, "step": 6978 }, { "epoch": 0.5469435736677116, "grad_norm": 3.6983461380004883, "learning_rate": 9.556749864588649e-06, "loss": 1.6362, "step": 6979 }, { "epoch": 0.5470219435736677, "grad_norm": 6.852066516876221, "learning_rate": 9.556622924110885e-06, "loss": 1.7577, "step": 6980 }, { "epoch": 0.5471003134796238, "grad_norm": 2.730858564376831, "learning_rate": 9.556495966302068e-06, "loss": 1.7153, "step": 6981 }, { "epoch": 0.54717868338558, "grad_norm": 2.7825276851654053, "learning_rate": 9.55636899116268e-06, "loss": 1.1825, "step": 6982 }, { "epoch": 0.5472570532915361, "grad_norm": 2.7046003341674805, "learning_rate": 9.556241998693208e-06, "loss": 1.4212, "step": 6983 }, { "epoch": 0.5473354231974922, "grad_norm": 3.349137306213379, "learning_rate": 9.556114988894128e-06, "loss": 1.3713, "step": 6984 }, { "epoch": 0.5474137931034483, "grad_norm": 2.623234987258911, "learning_rate": 9.555987961765929e-06, "loss": 1.5647, "step": 6985 }, { "epoch": 0.5474921630094044, "grad_norm": 2.303875684738159, "learning_rate": 9.55586091730909e-06, "loss": 1.216, "step": 6986 }, { "epoch": 0.5475705329153605, "grad_norm": 2.9486093521118164, "learning_rate": 9.555733855524098e-06, "loss": 1.5292, "step": 6987 }, { "epoch": 0.5476489028213166, "grad_norm": 5.977710723876953, "learning_rate": 9.555606776411433e-06, "loss": 1.0555, "step": 6988 }, { "epoch": 0.5477272727272727, "grad_norm": 4.160049915313721, "learning_rate": 9.55547967997158e-06, "loss": 1.3146, "step": 6989 }, { "epoch": 0.5478056426332288, "grad_norm": 12.515358924865723, "learning_rate": 9.555352566205024e-06, "loss": 1.088, "step": 6990 }, { "epoch": 0.547884012539185, "grad_norm": 4.672231197357178, "learning_rate": 9.555225435112243e-06, "loss": 1.4172, "step": 6991 }, { "epoch": 0.5479623824451411, "grad_norm": 3.6887166500091553, "learning_rate": 9.555098286693726e-06, "loss": 1.6156, "step": 6992 }, { "epoch": 0.5480407523510972, "grad_norm": 2.6462795734405518, "learning_rate": 9.554971120949954e-06, "loss": 1.2464, "step": 6993 }, { "epoch": 0.5481191222570533, "grad_norm": 2.6256518363952637, "learning_rate": 9.554843937881413e-06, "loss": 0.8032, "step": 6994 }, { "epoch": 0.5481974921630094, "grad_norm": 9.363532066345215, "learning_rate": 9.554716737488584e-06, "loss": 1.4897, "step": 6995 }, { "epoch": 0.5482758620689655, "grad_norm": 2.329517126083374, "learning_rate": 9.554589519771953e-06, "loss": 1.0907, "step": 6996 }, { "epoch": 0.5483542319749216, "grad_norm": 2.072382926940918, "learning_rate": 9.554462284732003e-06, "loss": 0.5642, "step": 6997 }, { "epoch": 0.5484326018808777, "grad_norm": 10.99841022491455, "learning_rate": 9.554335032369215e-06, "loss": 1.4139, "step": 6998 }, { "epoch": 0.5485109717868338, "grad_norm": 3.2268574237823486, "learning_rate": 9.55420776268408e-06, "loss": 2.078, "step": 6999 }, { "epoch": 0.54858934169279, "grad_norm": 3.0019419193267822, "learning_rate": 9.554080475677075e-06, "loss": 1.6784, "step": 7000 }, { "epoch": 0.548667711598746, "grad_norm": 3.168241262435913, "learning_rate": 9.553953171348687e-06, "loss": 1.4975, "step": 7001 }, { "epoch": 0.5487460815047022, "grad_norm": 2.3706130981445312, "learning_rate": 9.5538258496994e-06, "loss": 1.3653, "step": 7002 }, { "epoch": 0.5488244514106583, "grad_norm": 2.865928888320923, "learning_rate": 9.5536985107297e-06, "loss": 1.5183, "step": 7003 }, { "epoch": 0.5489028213166144, "grad_norm": 2.9847002029418945, "learning_rate": 9.55357115444007e-06, "loss": 1.5662, "step": 7004 }, { "epoch": 0.5489811912225705, "grad_norm": 2.5489048957824707, "learning_rate": 9.553443780830993e-06, "loss": 1.3884, "step": 7005 }, { "epoch": 0.5490595611285266, "grad_norm": 2.292708396911621, "learning_rate": 9.553316389902955e-06, "loss": 1.5893, "step": 7006 }, { "epoch": 0.5491379310344827, "grad_norm": 2.600877523422241, "learning_rate": 9.55318898165644e-06, "loss": 1.2082, "step": 7007 }, { "epoch": 0.5492163009404388, "grad_norm": 2.470515727996826, "learning_rate": 9.553061556091933e-06, "loss": 1.4877, "step": 7008 }, { "epoch": 0.5492946708463949, "grad_norm": 3.541558265686035, "learning_rate": 9.552934113209917e-06, "loss": 1.378, "step": 7009 }, { "epoch": 0.549373040752351, "grad_norm": 2.1925559043884277, "learning_rate": 9.55280665301088e-06, "loss": 1.4519, "step": 7010 }, { "epoch": 0.5494514106583072, "grad_norm": 2.7088329792022705, "learning_rate": 9.552679175495305e-06, "loss": 1.2144, "step": 7011 }, { "epoch": 0.5495297805642633, "grad_norm": 3.8212778568267822, "learning_rate": 9.552551680663676e-06, "loss": 1.9721, "step": 7012 }, { "epoch": 0.5496081504702194, "grad_norm": 3.712689161300659, "learning_rate": 9.552424168516479e-06, "loss": 1.3972, "step": 7013 }, { "epoch": 0.5496865203761756, "grad_norm": 2.6614952087402344, "learning_rate": 9.5522966390542e-06, "loss": 1.5609, "step": 7014 }, { "epoch": 0.5497648902821317, "grad_norm": 2.6022698879241943, "learning_rate": 9.552169092277321e-06, "loss": 1.2373, "step": 7015 }, { "epoch": 0.5498432601880878, "grad_norm": 3.272136688232422, "learning_rate": 9.552041528186329e-06, "loss": 1.4172, "step": 7016 }, { "epoch": 0.5499216300940439, "grad_norm": 2.550903558731079, "learning_rate": 9.55191394678171e-06, "loss": 1.2572, "step": 7017 }, { "epoch": 0.55, "grad_norm": 2.295450210571289, "learning_rate": 9.551786348063947e-06, "loss": 1.523, "step": 7018 }, { "epoch": 0.5500783699059562, "grad_norm": 2.434910535812378, "learning_rate": 9.551658732033529e-06, "loss": 1.3897, "step": 7019 }, { "epoch": 0.5501567398119123, "grad_norm": 2.8453149795532227, "learning_rate": 9.551531098690937e-06, "loss": 1.196, "step": 7020 }, { "epoch": 0.5502351097178684, "grad_norm": 3.3100783824920654, "learning_rate": 9.551403448036659e-06, "loss": 1.1966, "step": 7021 }, { "epoch": 0.5503134796238245, "grad_norm": 2.963001012802124, "learning_rate": 9.55127578007118e-06, "loss": 1.6928, "step": 7022 }, { "epoch": 0.5503918495297806, "grad_norm": 3.5957260131835938, "learning_rate": 9.551148094794986e-06, "loss": 1.5513, "step": 7023 }, { "epoch": 0.5504702194357367, "grad_norm": 2.7839608192443848, "learning_rate": 9.551020392208562e-06, "loss": 1.3835, "step": 7024 }, { "epoch": 0.5505485893416928, "grad_norm": 4.683842182159424, "learning_rate": 9.550892672312395e-06, "loss": 0.8014, "step": 7025 }, { "epoch": 0.5506269592476489, "grad_norm": 8.43674373626709, "learning_rate": 9.550764935106969e-06, "loss": 1.3241, "step": 7026 }, { "epoch": 0.550705329153605, "grad_norm": 5.050665378570557, "learning_rate": 9.55063718059277e-06, "loss": 1.3155, "step": 7027 }, { "epoch": 0.5507836990595611, "grad_norm": 2.1570611000061035, "learning_rate": 9.550509408770285e-06, "loss": 1.3012, "step": 7028 }, { "epoch": 0.5508620689655173, "grad_norm": 2.569244384765625, "learning_rate": 9.550381619639999e-06, "loss": 1.238, "step": 7029 }, { "epoch": 0.5509404388714734, "grad_norm": 2.529477119445801, "learning_rate": 9.5502538132024e-06, "loss": 1.6307, "step": 7030 }, { "epoch": 0.5510188087774295, "grad_norm": 2.783005952835083, "learning_rate": 9.550125989457972e-06, "loss": 1.4837, "step": 7031 }, { "epoch": 0.5510971786833856, "grad_norm": 11.352940559387207, "learning_rate": 9.5499981484072e-06, "loss": 1.0793, "step": 7032 }, { "epoch": 0.5511755485893417, "grad_norm": 2.8414626121520996, "learning_rate": 9.549870290050574e-06, "loss": 1.6439, "step": 7033 }, { "epoch": 0.5512539184952978, "grad_norm": 4.745594024658203, "learning_rate": 9.549742414388578e-06, "loss": 1.6524, "step": 7034 }, { "epoch": 0.5513322884012539, "grad_norm": 6.724367141723633, "learning_rate": 9.5496145214217e-06, "loss": 1.154, "step": 7035 }, { "epoch": 0.55141065830721, "grad_norm": 2.3030831813812256, "learning_rate": 9.549486611150423e-06, "loss": 1.5425, "step": 7036 }, { "epoch": 0.5514890282131661, "grad_norm": 3.2940590381622314, "learning_rate": 9.549358683575237e-06, "loss": 1.3449, "step": 7037 }, { "epoch": 0.5515673981191223, "grad_norm": 2.390505313873291, "learning_rate": 9.549230738696627e-06, "loss": 1.4973, "step": 7038 }, { "epoch": 0.5516457680250784, "grad_norm": 1.8742213249206543, "learning_rate": 9.54910277651508e-06, "loss": 1.1884, "step": 7039 }, { "epoch": 0.5517241379310345, "grad_norm": 3.7954776287078857, "learning_rate": 9.548974797031081e-06, "loss": 1.41, "step": 7040 }, { "epoch": 0.5518025078369906, "grad_norm": 4.9244513511657715, "learning_rate": 9.548846800245119e-06, "loss": 1.3887, "step": 7041 }, { "epoch": 0.5518808777429467, "grad_norm": 2.522937059402466, "learning_rate": 9.548718786157682e-06, "loss": 1.5479, "step": 7042 }, { "epoch": 0.5519592476489028, "grad_norm": 4.550972938537598, "learning_rate": 9.548590754769253e-06, "loss": 1.612, "step": 7043 }, { "epoch": 0.5520376175548589, "grad_norm": 2.8689379692077637, "learning_rate": 9.548462706080323e-06, "loss": 1.6116, "step": 7044 }, { "epoch": 0.552115987460815, "grad_norm": 1.9243844747543335, "learning_rate": 9.548334640091376e-06, "loss": 1.3828, "step": 7045 }, { "epoch": 0.5521943573667711, "grad_norm": 2.5871307849884033, "learning_rate": 9.5482065568029e-06, "loss": 1.4053, "step": 7046 }, { "epoch": 0.5522727272727272, "grad_norm": 3.258880853652954, "learning_rate": 9.548078456215383e-06, "loss": 1.6311, "step": 7047 }, { "epoch": 0.5523510971786834, "grad_norm": 2.3064169883728027, "learning_rate": 9.547950338329312e-06, "loss": 1.3705, "step": 7048 }, { "epoch": 0.5524294670846395, "grad_norm": 3.7328543663024902, "learning_rate": 9.547822203145173e-06, "loss": 1.7266, "step": 7049 }, { "epoch": 0.5525078369905956, "grad_norm": 3.227360963821411, "learning_rate": 9.547694050663455e-06, "loss": 1.4251, "step": 7050 }, { "epoch": 0.5525862068965517, "grad_norm": 3.3487133979797363, "learning_rate": 9.547565880884644e-06, "loss": 1.2095, "step": 7051 }, { "epoch": 0.5526645768025078, "grad_norm": 3.36967396736145, "learning_rate": 9.547437693809228e-06, "loss": 1.4458, "step": 7052 }, { "epoch": 0.5527429467084639, "grad_norm": 6.23745584487915, "learning_rate": 9.547309489437694e-06, "loss": 2.3684, "step": 7053 }, { "epoch": 0.55282131661442, "grad_norm": 6.496951580047607, "learning_rate": 9.547181267770534e-06, "loss": 2.1303, "step": 7054 }, { "epoch": 0.5528996865203761, "grad_norm": 2.4319112300872803, "learning_rate": 9.54705302880823e-06, "loss": 1.1465, "step": 7055 }, { "epoch": 0.5529780564263322, "grad_norm": 2.7969143390655518, "learning_rate": 9.546924772551269e-06, "loss": 1.0637, "step": 7056 }, { "epoch": 0.5530564263322884, "grad_norm": 2.378863573074341, "learning_rate": 9.546796499000144e-06, "loss": 1.3446, "step": 7057 }, { "epoch": 0.5531347962382445, "grad_norm": 2.2452893257141113, "learning_rate": 9.54666820815534e-06, "loss": 1.4236, "step": 7058 }, { "epoch": 0.5532131661442006, "grad_norm": 5.857650279998779, "learning_rate": 9.546539900017346e-06, "loss": 1.4546, "step": 7059 }, { "epoch": 0.5532915360501567, "grad_norm": 2.3554768562316895, "learning_rate": 9.546411574586649e-06, "loss": 1.4253, "step": 7060 }, { "epoch": 0.5533699059561129, "grad_norm": 2.520284652709961, "learning_rate": 9.546283231863738e-06, "loss": 1.4924, "step": 7061 }, { "epoch": 0.553448275862069, "grad_norm": 2.140106439590454, "learning_rate": 9.546154871849101e-06, "loss": 1.481, "step": 7062 }, { "epoch": 0.5535266457680251, "grad_norm": 3.7674992084503174, "learning_rate": 9.546026494543226e-06, "loss": 1.3628, "step": 7063 }, { "epoch": 0.5536050156739812, "grad_norm": 3.992494821548462, "learning_rate": 9.5458980999466e-06, "loss": 1.9386, "step": 7064 }, { "epoch": 0.5536833855799373, "grad_norm": 3.187920093536377, "learning_rate": 9.545769688059714e-06, "loss": 1.4615, "step": 7065 }, { "epoch": 0.5537617554858935, "grad_norm": 2.6700592041015625, "learning_rate": 9.545641258883053e-06, "loss": 1.596, "step": 7066 }, { "epoch": 0.5538401253918496, "grad_norm": 2.422481060028076, "learning_rate": 9.545512812417111e-06, "loss": 0.9369, "step": 7067 }, { "epoch": 0.5539184952978057, "grad_norm": 1.993808388710022, "learning_rate": 9.54538434866237e-06, "loss": 1.4612, "step": 7068 }, { "epoch": 0.5539968652037618, "grad_norm": 2.200850248336792, "learning_rate": 9.545255867619321e-06, "loss": 1.3795, "step": 7069 }, { "epoch": 0.5540752351097179, "grad_norm": 3.6391632556915283, "learning_rate": 9.545127369288455e-06, "loss": 1.4561, "step": 7070 }, { "epoch": 0.554153605015674, "grad_norm": 4.555868625640869, "learning_rate": 9.544998853670259e-06, "loss": 1.4701, "step": 7071 }, { "epoch": 0.5542319749216301, "grad_norm": 3.7009716033935547, "learning_rate": 9.544870320765221e-06, "loss": 1.1428, "step": 7072 }, { "epoch": 0.5543103448275862, "grad_norm": 3.664278030395508, "learning_rate": 9.544741770573832e-06, "loss": 1.3957, "step": 7073 }, { "epoch": 0.5543887147335423, "grad_norm": 3.863420248031616, "learning_rate": 9.544613203096578e-06, "loss": 1.1834, "step": 7074 }, { "epoch": 0.5544670846394985, "grad_norm": 1.9604138135910034, "learning_rate": 9.54448461833395e-06, "loss": 1.2443, "step": 7075 }, { "epoch": 0.5545454545454546, "grad_norm": 5.712419033050537, "learning_rate": 9.544356016286437e-06, "loss": 1.6391, "step": 7076 }, { "epoch": 0.5546238244514107, "grad_norm": 6.113937854766846, "learning_rate": 9.544227396954528e-06, "loss": 1.2045, "step": 7077 }, { "epoch": 0.5547021943573668, "grad_norm": 6.860296726226807, "learning_rate": 9.544098760338712e-06, "loss": 1.9602, "step": 7078 }, { "epoch": 0.5547805642633229, "grad_norm": 3.9384374618530273, "learning_rate": 9.543970106439478e-06, "loss": 1.3518, "step": 7079 }, { "epoch": 0.554858934169279, "grad_norm": 2.5282657146453857, "learning_rate": 9.543841435257315e-06, "loss": 1.2904, "step": 7080 }, { "epoch": 0.5549373040752351, "grad_norm": 2.76753306388855, "learning_rate": 9.543712746792714e-06, "loss": 1.5006, "step": 7081 }, { "epoch": 0.5550156739811912, "grad_norm": 2.8360118865966797, "learning_rate": 9.543584041046164e-06, "loss": 1.436, "step": 7082 }, { "epoch": 0.5550940438871473, "grad_norm": 2.9549639225006104, "learning_rate": 9.543455318018154e-06, "loss": 1.2767, "step": 7083 }, { "epoch": 0.5551724137931034, "grad_norm": 2.479602098464966, "learning_rate": 9.543326577709172e-06, "loss": 1.3845, "step": 7084 }, { "epoch": 0.5552507836990596, "grad_norm": 7.223255634307861, "learning_rate": 9.54319782011971e-06, "loss": 1.9731, "step": 7085 }, { "epoch": 0.5553291536050157, "grad_norm": 1.9399760961532593, "learning_rate": 9.543069045250257e-06, "loss": 1.1624, "step": 7086 }, { "epoch": 0.5554075235109718, "grad_norm": 2.1341311931610107, "learning_rate": 9.542940253101304e-06, "loss": 1.3882, "step": 7087 }, { "epoch": 0.5554858934169279, "grad_norm": 2.6336007118225098, "learning_rate": 9.542811443673338e-06, "loss": 1.7203, "step": 7088 }, { "epoch": 0.555564263322884, "grad_norm": 3.9176025390625, "learning_rate": 9.542682616966852e-06, "loss": 1.4046, "step": 7089 }, { "epoch": 0.5556426332288401, "grad_norm": 2.0075442790985107, "learning_rate": 9.542553772982334e-06, "loss": 1.2541, "step": 7090 }, { "epoch": 0.5557210031347962, "grad_norm": 7.223277568817139, "learning_rate": 9.542424911720275e-06, "loss": 1.4613, "step": 7091 }, { "epoch": 0.5557993730407523, "grad_norm": 3.420492172241211, "learning_rate": 9.542296033181165e-06, "loss": 1.4351, "step": 7092 }, { "epoch": 0.5558777429467084, "grad_norm": 2.1496925354003906, "learning_rate": 9.542167137365493e-06, "loss": 1.3575, "step": 7093 }, { "epoch": 0.5559561128526646, "grad_norm": 2.851710557937622, "learning_rate": 9.54203822427375e-06, "loss": 1.5227, "step": 7094 }, { "epoch": 0.5560344827586207, "grad_norm": 4.350184440612793, "learning_rate": 9.541909293906428e-06, "loss": 2.1172, "step": 7095 }, { "epoch": 0.5561128526645768, "grad_norm": 2.514467716217041, "learning_rate": 9.541780346264015e-06, "loss": 1.3503, "step": 7096 }, { "epoch": 0.5561912225705329, "grad_norm": 2.5045342445373535, "learning_rate": 9.541651381347004e-06, "loss": 1.1378, "step": 7097 }, { "epoch": 0.556269592476489, "grad_norm": 3.325321912765503, "learning_rate": 9.54152239915588e-06, "loss": 1.5016, "step": 7098 }, { "epoch": 0.5563479623824451, "grad_norm": 4.102065086364746, "learning_rate": 9.541393399691142e-06, "loss": 1.4159, "step": 7099 }, { "epoch": 0.5564263322884012, "grad_norm": 2.5325114727020264, "learning_rate": 9.541264382953273e-06, "loss": 1.2563, "step": 7100 }, { "epoch": 0.5565047021943573, "grad_norm": 2.2278647422790527, "learning_rate": 9.541135348942768e-06, "loss": 1.3136, "step": 7101 }, { "epoch": 0.5565830721003134, "grad_norm": 6.646543979644775, "learning_rate": 9.541006297660116e-06, "loss": 1.666, "step": 7102 }, { "epoch": 0.5566614420062695, "grad_norm": 2.410005569458008, "learning_rate": 9.54087722910581e-06, "loss": 1.5007, "step": 7103 }, { "epoch": 0.5567398119122257, "grad_norm": 2.489475727081299, "learning_rate": 9.540748143280337e-06, "loss": 1.3575, "step": 7104 }, { "epoch": 0.5568181818181818, "grad_norm": 3.1102583408355713, "learning_rate": 9.540619040184192e-06, "loss": 1.1456, "step": 7105 }, { "epoch": 0.5568965517241379, "grad_norm": 2.2431700229644775, "learning_rate": 9.540489919817863e-06, "loss": 1.3792, "step": 7106 }, { "epoch": 0.556974921630094, "grad_norm": 2.8097219467163086, "learning_rate": 9.540360782181843e-06, "loss": 1.4442, "step": 7107 }, { "epoch": 0.5570532915360502, "grad_norm": 1.7892701625823975, "learning_rate": 9.540231627276624e-06, "loss": 1.2156, "step": 7108 }, { "epoch": 0.5571316614420063, "grad_norm": 2.8114047050476074, "learning_rate": 9.540102455102696e-06, "loss": 1.3388, "step": 7109 }, { "epoch": 0.5572100313479624, "grad_norm": 2.6636059284210205, "learning_rate": 9.539973265660549e-06, "loss": 1.7418, "step": 7110 }, { "epoch": 0.5572884012539185, "grad_norm": 2.8567516803741455, "learning_rate": 9.539844058950676e-06, "loss": 1.0089, "step": 7111 }, { "epoch": 0.5573667711598747, "grad_norm": 2.612750291824341, "learning_rate": 9.539714834973568e-06, "loss": 1.4507, "step": 7112 }, { "epoch": 0.5574451410658308, "grad_norm": 3.1285297870635986, "learning_rate": 9.539585593729714e-06, "loss": 1.2795, "step": 7113 }, { "epoch": 0.5575235109717869, "grad_norm": 3.431558132171631, "learning_rate": 9.539456335219612e-06, "loss": 1.2653, "step": 7114 }, { "epoch": 0.557601880877743, "grad_norm": 3.346642017364502, "learning_rate": 9.539327059443748e-06, "loss": 1.2628, "step": 7115 }, { "epoch": 0.5576802507836991, "grad_norm": 3.4203968048095703, "learning_rate": 9.539197766402614e-06, "loss": 1.673, "step": 7116 }, { "epoch": 0.5577586206896552, "grad_norm": 3.320146083831787, "learning_rate": 9.539068456096706e-06, "loss": 1.6114, "step": 7117 }, { "epoch": 0.5578369905956113, "grad_norm": 2.6946206092834473, "learning_rate": 9.538939128526512e-06, "loss": 1.4391, "step": 7118 }, { "epoch": 0.5579153605015674, "grad_norm": 5.664988994598389, "learning_rate": 9.538809783692524e-06, "loss": 1.2742, "step": 7119 }, { "epoch": 0.5579937304075235, "grad_norm": 2.865995407104492, "learning_rate": 9.538680421595236e-06, "loss": 1.6559, "step": 7120 }, { "epoch": 0.5580721003134796, "grad_norm": 4.1148881912231445, "learning_rate": 9.538551042235138e-06, "loss": 1.4363, "step": 7121 }, { "epoch": 0.5581504702194358, "grad_norm": 3.9484236240386963, "learning_rate": 9.538421645612723e-06, "loss": 0.7524, "step": 7122 }, { "epoch": 0.5582288401253919, "grad_norm": 3.1010935306549072, "learning_rate": 9.538292231728483e-06, "loss": 1.219, "step": 7123 }, { "epoch": 0.558307210031348, "grad_norm": 8.775985717773438, "learning_rate": 9.538162800582911e-06, "loss": 1.3504, "step": 7124 }, { "epoch": 0.5583855799373041, "grad_norm": 6.653655529022217, "learning_rate": 9.538033352176499e-06, "loss": 1.1008, "step": 7125 }, { "epoch": 0.5584639498432602, "grad_norm": 2.4331629276275635, "learning_rate": 9.537903886509738e-06, "loss": 1.3412, "step": 7126 }, { "epoch": 0.5585423197492163, "grad_norm": 3.1767921447753906, "learning_rate": 9.537774403583123e-06, "loss": 1.408, "step": 7127 }, { "epoch": 0.5586206896551724, "grad_norm": 7.967082977294922, "learning_rate": 9.537644903397142e-06, "loss": 1.7541, "step": 7128 }, { "epoch": 0.5586990595611285, "grad_norm": 3.8048434257507324, "learning_rate": 9.537515385952295e-06, "loss": 1.4433, "step": 7129 }, { "epoch": 0.5587774294670846, "grad_norm": 2.1264429092407227, "learning_rate": 9.537385851249066e-06, "loss": 1.5293, "step": 7130 }, { "epoch": 0.5588557993730408, "grad_norm": 2.5420937538146973, "learning_rate": 9.537256299287954e-06, "loss": 1.7224, "step": 7131 }, { "epoch": 0.5589341692789969, "grad_norm": 2.29455828666687, "learning_rate": 9.537126730069447e-06, "loss": 1.3784, "step": 7132 }, { "epoch": 0.559012539184953, "grad_norm": 16.90359115600586, "learning_rate": 9.536997143594042e-06, "loss": 0.7436, "step": 7133 }, { "epoch": 0.5590909090909091, "grad_norm": 2.284303903579712, "learning_rate": 9.53686753986223e-06, "loss": 1.5979, "step": 7134 }, { "epoch": 0.5591692789968652, "grad_norm": 3.598487377166748, "learning_rate": 9.536737918874505e-06, "loss": 1.6345, "step": 7135 }, { "epoch": 0.5592476489028213, "grad_norm": 3.039438009262085, "learning_rate": 9.536608280631358e-06, "loss": 1.6843, "step": 7136 }, { "epoch": 0.5593260188087774, "grad_norm": 4.59898567199707, "learning_rate": 9.536478625133286e-06, "loss": 1.6643, "step": 7137 }, { "epoch": 0.5594043887147335, "grad_norm": 2.9081132411956787, "learning_rate": 9.536348952380776e-06, "loss": 1.48, "step": 7138 }, { "epoch": 0.5594827586206896, "grad_norm": 2.211777925491333, "learning_rate": 9.536219262374326e-06, "loss": 1.4551, "step": 7139 }, { "epoch": 0.5595611285266457, "grad_norm": 2.1745188236236572, "learning_rate": 9.536089555114428e-06, "loss": 1.4713, "step": 7140 }, { "epoch": 0.5596394984326019, "grad_norm": 3.785008192062378, "learning_rate": 9.535959830601573e-06, "loss": 1.3533, "step": 7141 }, { "epoch": 0.559717868338558, "grad_norm": 3.7071869373321533, "learning_rate": 9.53583008883626e-06, "loss": 1.7339, "step": 7142 }, { "epoch": 0.5597962382445141, "grad_norm": 2.8908467292785645, "learning_rate": 9.535700329818977e-06, "loss": 1.2484, "step": 7143 }, { "epoch": 0.5598746081504702, "grad_norm": 3.086883544921875, "learning_rate": 9.535570553550219e-06, "loss": 1.4754, "step": 7144 }, { "epoch": 0.5599529780564263, "grad_norm": 2.53379487991333, "learning_rate": 9.535440760030483e-06, "loss": 1.6052, "step": 7145 }, { "epoch": 0.5600313479623824, "grad_norm": 2.81658673286438, "learning_rate": 9.535310949260258e-06, "loss": 1.4942, "step": 7146 }, { "epoch": 0.5601097178683385, "grad_norm": 3.0099170207977295, "learning_rate": 9.53518112124004e-06, "loss": 1.2829, "step": 7147 }, { "epoch": 0.5601880877742946, "grad_norm": 5.236525535583496, "learning_rate": 9.535051275970323e-06, "loss": 1.234, "step": 7148 }, { "epoch": 0.5602664576802507, "grad_norm": 2.694957733154297, "learning_rate": 9.5349214134516e-06, "loss": 1.4452, "step": 7149 }, { "epoch": 0.5603448275862069, "grad_norm": 2.2507498264312744, "learning_rate": 9.534791533684365e-06, "loss": 1.1667, "step": 7150 }, { "epoch": 0.560423197492163, "grad_norm": 9.450895309448242, "learning_rate": 9.534661636669112e-06, "loss": 0.9266, "step": 7151 }, { "epoch": 0.5605015673981191, "grad_norm": 4.869312286376953, "learning_rate": 9.534531722406335e-06, "loss": 1.481, "step": 7152 }, { "epoch": 0.5605799373040752, "grad_norm": 4.060623645782471, "learning_rate": 9.534401790896527e-06, "loss": 0.9659, "step": 7153 }, { "epoch": 0.5606583072100313, "grad_norm": 3.7336127758026123, "learning_rate": 9.534271842140189e-06, "loss": 1.699, "step": 7154 }, { "epoch": 0.5607366771159875, "grad_norm": 4.06795072555542, "learning_rate": 9.534141876137805e-06, "loss": 0.6926, "step": 7155 }, { "epoch": 0.5608150470219436, "grad_norm": 2.411449909210205, "learning_rate": 9.534011892889875e-06, "loss": 1.3162, "step": 7156 }, { "epoch": 0.5608934169278997, "grad_norm": 3.288522243499756, "learning_rate": 9.533881892396893e-06, "loss": 1.3704, "step": 7157 }, { "epoch": 0.5609717868338558, "grad_norm": 2.285038948059082, "learning_rate": 9.533751874659353e-06, "loss": 1.5467, "step": 7158 }, { "epoch": 0.561050156739812, "grad_norm": 3.175434112548828, "learning_rate": 9.53362183967775e-06, "loss": 1.7044, "step": 7159 }, { "epoch": 0.5611285266457681, "grad_norm": 5.328977584838867, "learning_rate": 9.533491787452578e-06, "loss": 1.5688, "step": 7160 }, { "epoch": 0.5612068965517242, "grad_norm": 5.9406208992004395, "learning_rate": 9.533361717984331e-06, "loss": 1.7067, "step": 7161 }, { "epoch": 0.5612852664576803, "grad_norm": 2.6634180545806885, "learning_rate": 9.533231631273505e-06, "loss": 1.4637, "step": 7162 }, { "epoch": 0.5613636363636364, "grad_norm": 2.217308521270752, "learning_rate": 9.533101527320595e-06, "loss": 1.6732, "step": 7163 }, { "epoch": 0.5614420062695925, "grad_norm": 1.9280201196670532, "learning_rate": 9.532971406126095e-06, "loss": 1.4359, "step": 7164 }, { "epoch": 0.5615203761755486, "grad_norm": 2.548002243041992, "learning_rate": 9.532841267690499e-06, "loss": 1.4311, "step": 7165 }, { "epoch": 0.5615987460815047, "grad_norm": 2.154195547103882, "learning_rate": 9.532711112014303e-06, "loss": 1.4212, "step": 7166 }, { "epoch": 0.5616771159874608, "grad_norm": 4.696367263793945, "learning_rate": 9.532580939098005e-06, "loss": 0.9028, "step": 7167 }, { "epoch": 0.561755485893417, "grad_norm": 2.4805948734283447, "learning_rate": 9.532450748942092e-06, "loss": 1.366, "step": 7168 }, { "epoch": 0.5618338557993731, "grad_norm": 3.14508318901062, "learning_rate": 9.532320541547069e-06, "loss": 1.2921, "step": 7169 }, { "epoch": 0.5619122257053292, "grad_norm": 2.6809346675872803, "learning_rate": 9.532190316913424e-06, "loss": 1.4647, "step": 7170 }, { "epoch": 0.5619905956112853, "grad_norm": 2.535944700241089, "learning_rate": 9.532060075041656e-06, "loss": 1.1996, "step": 7171 }, { "epoch": 0.5620689655172414, "grad_norm": 1.871012806892395, "learning_rate": 9.53192981593226e-06, "loss": 1.6283, "step": 7172 }, { "epoch": 0.5621473354231975, "grad_norm": 2.4011783599853516, "learning_rate": 9.531799539585728e-06, "loss": 1.5571, "step": 7173 }, { "epoch": 0.5622257053291536, "grad_norm": 4.386277198791504, "learning_rate": 9.53166924600256e-06, "loss": 1.9804, "step": 7174 }, { "epoch": 0.5623040752351097, "grad_norm": 2.0022153854370117, "learning_rate": 9.531538935183252e-06, "loss": 0.9842, "step": 7175 }, { "epoch": 0.5623824451410658, "grad_norm": 2.9789223670959473, "learning_rate": 9.531408607128294e-06, "loss": 1.2144, "step": 7176 }, { "epoch": 0.562460815047022, "grad_norm": 4.890726089477539, "learning_rate": 9.531278261838187e-06, "loss": 1.1683, "step": 7177 }, { "epoch": 0.562539184952978, "grad_norm": 4.298180103302002, "learning_rate": 9.531147899313425e-06, "loss": 1.6099, "step": 7178 }, { "epoch": 0.5626175548589342, "grad_norm": 3.1446101665496826, "learning_rate": 9.531017519554501e-06, "loss": 1.5666, "step": 7179 }, { "epoch": 0.5626959247648903, "grad_norm": 2.0994513034820557, "learning_rate": 9.530887122561917e-06, "loss": 1.1653, "step": 7180 }, { "epoch": 0.5627742946708464, "grad_norm": 2.0023953914642334, "learning_rate": 9.530756708336164e-06, "loss": 1.3979, "step": 7181 }, { "epoch": 0.5628526645768025, "grad_norm": 2.36641263961792, "learning_rate": 9.53062627687774e-06, "loss": 1.4104, "step": 7182 }, { "epoch": 0.5629310344827586, "grad_norm": 6.571700096130371, "learning_rate": 9.530495828187141e-06, "loss": 1.3643, "step": 7183 }, { "epoch": 0.5630094043887147, "grad_norm": 3.0362751483917236, "learning_rate": 9.530365362264862e-06, "loss": 1.1242, "step": 7184 }, { "epoch": 0.5630877742946708, "grad_norm": 4.548489093780518, "learning_rate": 9.5302348791114e-06, "loss": 1.5107, "step": 7185 }, { "epoch": 0.5631661442006269, "grad_norm": 2.603046178817749, "learning_rate": 9.530104378727252e-06, "loss": 1.6464, "step": 7186 }, { "epoch": 0.563244514106583, "grad_norm": 10.905974388122559, "learning_rate": 9.529973861112913e-06, "loss": 1.6086, "step": 7187 }, { "epoch": 0.5633228840125392, "grad_norm": 4.001736164093018, "learning_rate": 9.529843326268881e-06, "loss": 1.6314, "step": 7188 }, { "epoch": 0.5634012539184953, "grad_norm": 2.511051654815674, "learning_rate": 9.529712774195651e-06, "loss": 1.5663, "step": 7189 }, { "epoch": 0.5634796238244514, "grad_norm": 5.584763050079346, "learning_rate": 9.529582204893721e-06, "loss": 1.2549, "step": 7190 }, { "epoch": 0.5635579937304075, "grad_norm": 3.3921375274658203, "learning_rate": 9.529451618363588e-06, "loss": 1.3798, "step": 7191 }, { "epoch": 0.5636363636363636, "grad_norm": 2.1521103382110596, "learning_rate": 9.529321014605746e-06, "loss": 1.4672, "step": 7192 }, { "epoch": 0.5637147335423197, "grad_norm": 3.652665853500366, "learning_rate": 9.529190393620693e-06, "loss": 1.27, "step": 7193 }, { "epoch": 0.5637931034482758, "grad_norm": 2.674931049346924, "learning_rate": 9.529059755408927e-06, "loss": 1.6007, "step": 7194 }, { "epoch": 0.5638714733542319, "grad_norm": 2.103407382965088, "learning_rate": 9.528929099970942e-06, "loss": 1.3338, "step": 7195 }, { "epoch": 0.563949843260188, "grad_norm": 3.077946186065674, "learning_rate": 9.52879842730724e-06, "loss": 1.5747, "step": 7196 }, { "epoch": 0.5640282131661442, "grad_norm": 5.175875186920166, "learning_rate": 9.528667737418314e-06, "loss": 1.3279, "step": 7197 }, { "epoch": 0.5641065830721003, "grad_norm": 2.6583969593048096, "learning_rate": 9.52853703030466e-06, "loss": 1.2718, "step": 7198 }, { "epoch": 0.5641849529780564, "grad_norm": 2.590484380722046, "learning_rate": 9.528406305966778e-06, "loss": 1.4672, "step": 7199 }, { "epoch": 0.5642633228840125, "grad_norm": 2.549276113510132, "learning_rate": 9.528275564405164e-06, "loss": 1.2925, "step": 7200 }, { "epoch": 0.5643416927899687, "grad_norm": 3.17397403717041, "learning_rate": 9.528144805620317e-06, "loss": 1.1601, "step": 7201 }, { "epoch": 0.5644200626959248, "grad_norm": 2.553319215774536, "learning_rate": 9.528014029612734e-06, "loss": 1.3431, "step": 7202 }, { "epoch": 0.5644984326018809, "grad_norm": 2.620222806930542, "learning_rate": 9.527883236382908e-06, "loss": 1.7299, "step": 7203 }, { "epoch": 0.564576802507837, "grad_norm": 3.959132432937622, "learning_rate": 9.527752425931342e-06, "loss": 1.4805, "step": 7204 }, { "epoch": 0.5646551724137931, "grad_norm": 2.998300313949585, "learning_rate": 9.527621598258531e-06, "loss": 1.3342, "step": 7205 }, { "epoch": 0.5647335423197493, "grad_norm": 2.8354451656341553, "learning_rate": 9.527490753364972e-06, "loss": 1.0822, "step": 7206 }, { "epoch": 0.5648119122257054, "grad_norm": 3.5517618656158447, "learning_rate": 9.527359891251166e-06, "loss": 1.7998, "step": 7207 }, { "epoch": 0.5648902821316615, "grad_norm": 2.784440040588379, "learning_rate": 9.527229011917606e-06, "loss": 1.192, "step": 7208 }, { "epoch": 0.5649686520376176, "grad_norm": 2.196901559829712, "learning_rate": 9.527098115364793e-06, "loss": 1.5704, "step": 7209 }, { "epoch": 0.5650470219435737, "grad_norm": 2.869556188583374, "learning_rate": 9.526967201593225e-06, "loss": 1.5486, "step": 7210 }, { "epoch": 0.5651253918495298, "grad_norm": 3.7996625900268555, "learning_rate": 9.526836270603396e-06, "loss": 1.5433, "step": 7211 }, { "epoch": 0.5652037617554859, "grad_norm": 2.9240710735321045, "learning_rate": 9.52670532239581e-06, "loss": 1.6796, "step": 7212 }, { "epoch": 0.565282131661442, "grad_norm": 2.9671919345855713, "learning_rate": 9.526574356970962e-06, "loss": 1.179, "step": 7213 }, { "epoch": 0.5653605015673981, "grad_norm": 9.639904022216797, "learning_rate": 9.526443374329348e-06, "loss": 1.4763, "step": 7214 }, { "epoch": 0.5654388714733543, "grad_norm": 2.489001989364624, "learning_rate": 9.52631237447147e-06, "loss": 0.9696, "step": 7215 }, { "epoch": 0.5655172413793104, "grad_norm": 2.5515031814575195, "learning_rate": 9.526181357397824e-06, "loss": 1.6268, "step": 7216 }, { "epoch": 0.5655956112852665, "grad_norm": 2.669713020324707, "learning_rate": 9.526050323108909e-06, "loss": 1.211, "step": 7217 }, { "epoch": 0.5656739811912226, "grad_norm": 2.38338303565979, "learning_rate": 9.525919271605224e-06, "loss": 1.1746, "step": 7218 }, { "epoch": 0.5657523510971787, "grad_norm": 2.124101161956787, "learning_rate": 9.525788202887266e-06, "loss": 1.1786, "step": 7219 }, { "epoch": 0.5658307210031348, "grad_norm": 3.012643814086914, "learning_rate": 9.525657116955533e-06, "loss": 1.3033, "step": 7220 }, { "epoch": 0.5659090909090909, "grad_norm": 2.436215400695801, "learning_rate": 9.525526013810527e-06, "loss": 1.2328, "step": 7221 }, { "epoch": 0.565987460815047, "grad_norm": 3.389294385910034, "learning_rate": 9.525394893452744e-06, "loss": 1.3067, "step": 7222 }, { "epoch": 0.5660658307210031, "grad_norm": 2.2082722187042236, "learning_rate": 9.525263755882684e-06, "loss": 1.4395, "step": 7223 }, { "epoch": 0.5661442006269592, "grad_norm": 2.3570809364318848, "learning_rate": 9.525132601100843e-06, "loss": 1.3145, "step": 7224 }, { "epoch": 0.5662225705329154, "grad_norm": 17.359243392944336, "learning_rate": 9.525001429107723e-06, "loss": 0.875, "step": 7225 }, { "epoch": 0.5663009404388715, "grad_norm": 2.6837010383605957, "learning_rate": 9.52487023990382e-06, "loss": 1.3469, "step": 7226 }, { "epoch": 0.5663793103448276, "grad_norm": 2.684234857559204, "learning_rate": 9.524739033489637e-06, "loss": 1.4028, "step": 7227 }, { "epoch": 0.5664576802507837, "grad_norm": 4.2849602699279785, "learning_rate": 9.52460780986567e-06, "loss": 1.0362, "step": 7228 }, { "epoch": 0.5665360501567398, "grad_norm": 2.3959648609161377, "learning_rate": 9.524476569032419e-06, "loss": 1.3792, "step": 7229 }, { "epoch": 0.5666144200626959, "grad_norm": 2.127002477645874, "learning_rate": 9.524345310990382e-06, "loss": 0.9638, "step": 7230 }, { "epoch": 0.566692789968652, "grad_norm": 2.1072537899017334, "learning_rate": 9.52421403574006e-06, "loss": 1.2346, "step": 7231 }, { "epoch": 0.5667711598746081, "grad_norm": 3.034285068511963, "learning_rate": 9.524082743281953e-06, "loss": 1.423, "step": 7232 }, { "epoch": 0.5668495297805642, "grad_norm": 3.174199104309082, "learning_rate": 9.523951433616557e-06, "loss": 1.7341, "step": 7233 }, { "epoch": 0.5669278996865204, "grad_norm": 2.440674304962158, "learning_rate": 9.523820106744374e-06, "loss": 1.2995, "step": 7234 }, { "epoch": 0.5670062695924765, "grad_norm": 2.7919180393218994, "learning_rate": 9.523688762665903e-06, "loss": 1.5925, "step": 7235 }, { "epoch": 0.5670846394984326, "grad_norm": 2.803511142730713, "learning_rate": 9.523557401381645e-06, "loss": 1.5609, "step": 7236 }, { "epoch": 0.5671630094043887, "grad_norm": 3.299100399017334, "learning_rate": 9.523426022892096e-06, "loss": 1.1953, "step": 7237 }, { "epoch": 0.5672413793103448, "grad_norm": 3.001555919647217, "learning_rate": 9.52329462719776e-06, "loss": 1.4801, "step": 7238 }, { "epoch": 0.5673197492163009, "grad_norm": 2.51121187210083, "learning_rate": 9.523163214299132e-06, "loss": 1.4004, "step": 7239 }, { "epoch": 0.567398119122257, "grad_norm": 2.2547919750213623, "learning_rate": 9.523031784196714e-06, "loss": 1.4081, "step": 7240 }, { "epoch": 0.5674764890282131, "grad_norm": 3.0673580169677734, "learning_rate": 9.522900336891007e-06, "loss": 1.6262, "step": 7241 }, { "epoch": 0.5675548589341692, "grad_norm": 2.3456499576568604, "learning_rate": 9.522768872382511e-06, "loss": 1.3787, "step": 7242 }, { "epoch": 0.5676332288401253, "grad_norm": 4.393799781799316, "learning_rate": 9.522637390671725e-06, "loss": 1.8067, "step": 7243 }, { "epoch": 0.5677115987460815, "grad_norm": 1.9207409620285034, "learning_rate": 9.522505891759151e-06, "loss": 1.225, "step": 7244 }, { "epoch": 0.5677899686520376, "grad_norm": 2.5092904567718506, "learning_rate": 9.522374375645285e-06, "loss": 1.3168, "step": 7245 }, { "epoch": 0.5678683385579937, "grad_norm": 6.58236026763916, "learning_rate": 9.522242842330632e-06, "loss": 1.6753, "step": 7246 }, { "epoch": 0.5679467084639498, "grad_norm": 2.3889591693878174, "learning_rate": 9.522111291815688e-06, "loss": 1.4174, "step": 7247 }, { "epoch": 0.568025078369906, "grad_norm": 2.2197532653808594, "learning_rate": 9.521979724100956e-06, "loss": 1.5524, "step": 7248 }, { "epoch": 0.5681034482758621, "grad_norm": 2.7369580268859863, "learning_rate": 9.521848139186936e-06, "loss": 1.7757, "step": 7249 }, { "epoch": 0.5681818181818182, "grad_norm": 6.6136274337768555, "learning_rate": 9.521716537074126e-06, "loss": 1.0415, "step": 7250 }, { "epoch": 0.5682601880877743, "grad_norm": 2.250059127807617, "learning_rate": 9.52158491776303e-06, "loss": 1.5163, "step": 7251 }, { "epoch": 0.5683385579937305, "grad_norm": 3.411775588989258, "learning_rate": 9.52145328125415e-06, "loss": 1.5412, "step": 7252 }, { "epoch": 0.5684169278996866, "grad_norm": 1.8794668912887573, "learning_rate": 9.521321627547982e-06, "loss": 1.539, "step": 7253 }, { "epoch": 0.5684952978056427, "grad_norm": 3.9639854431152344, "learning_rate": 9.521189956645029e-06, "loss": 1.1882, "step": 7254 }, { "epoch": 0.5685736677115988, "grad_norm": 3.780625820159912, "learning_rate": 9.521058268545792e-06, "loss": 1.0151, "step": 7255 }, { "epoch": 0.5686520376175549, "grad_norm": 2.2574169635772705, "learning_rate": 9.52092656325077e-06, "loss": 1.2509, "step": 7256 }, { "epoch": 0.568730407523511, "grad_norm": 4.002212047576904, "learning_rate": 9.520794840760467e-06, "loss": 1.6528, "step": 7257 }, { "epoch": 0.5688087774294671, "grad_norm": 2.327810287475586, "learning_rate": 9.520663101075382e-06, "loss": 1.2531, "step": 7258 }, { "epoch": 0.5688871473354232, "grad_norm": 2.8907811641693115, "learning_rate": 9.520531344196019e-06, "loss": 1.2468, "step": 7259 }, { "epoch": 0.5689655172413793, "grad_norm": 8.506978034973145, "learning_rate": 9.520399570122873e-06, "loss": 0.5895, "step": 7260 }, { "epoch": 0.5690438871473354, "grad_norm": 2.9502110481262207, "learning_rate": 9.520267778856449e-06, "loss": 1.1746, "step": 7261 }, { "epoch": 0.5691222570532916, "grad_norm": 4.195956230163574, "learning_rate": 9.52013597039725e-06, "loss": 1.3125, "step": 7262 }, { "epoch": 0.5692006269592477, "grad_norm": 2.62262225151062, "learning_rate": 9.520004144745775e-06, "loss": 1.4347, "step": 7263 }, { "epoch": 0.5692789968652038, "grad_norm": 2.7325775623321533, "learning_rate": 9.519872301902526e-06, "loss": 1.2687, "step": 7264 }, { "epoch": 0.5693573667711599, "grad_norm": 3.4990131855010986, "learning_rate": 9.519740441868002e-06, "loss": 1.364, "step": 7265 }, { "epoch": 0.569435736677116, "grad_norm": 2.4756693840026855, "learning_rate": 9.519608564642711e-06, "loss": 1.559, "step": 7266 }, { "epoch": 0.5695141065830721, "grad_norm": 2.4541163444519043, "learning_rate": 9.519476670227147e-06, "loss": 1.1294, "step": 7267 }, { "epoch": 0.5695924764890282, "grad_norm": 3.072387218475342, "learning_rate": 9.519344758621816e-06, "loss": 1.3296, "step": 7268 }, { "epoch": 0.5696708463949843, "grad_norm": 3.3479630947113037, "learning_rate": 9.519212829827221e-06, "loss": 1.4372, "step": 7269 }, { "epoch": 0.5697492163009404, "grad_norm": 2.7200279235839844, "learning_rate": 9.51908088384386e-06, "loss": 1.227, "step": 7270 }, { "epoch": 0.5698275862068966, "grad_norm": 2.5484819412231445, "learning_rate": 9.518948920672234e-06, "loss": 1.9394, "step": 7271 }, { "epoch": 0.5699059561128527, "grad_norm": 6.268827438354492, "learning_rate": 9.51881694031285e-06, "loss": 1.4535, "step": 7272 }, { "epoch": 0.5699843260188088, "grad_norm": 3.5674233436584473, "learning_rate": 9.518684942766207e-06, "loss": 1.2833, "step": 7273 }, { "epoch": 0.5700626959247649, "grad_norm": 2.226788282394409, "learning_rate": 9.518552928032807e-06, "loss": 1.2324, "step": 7274 }, { "epoch": 0.570141065830721, "grad_norm": 2.2122955322265625, "learning_rate": 9.518420896113153e-06, "loss": 1.4913, "step": 7275 }, { "epoch": 0.5702194357366771, "grad_norm": 1.8549672365188599, "learning_rate": 9.518288847007747e-06, "loss": 1.1143, "step": 7276 }, { "epoch": 0.5702978056426332, "grad_norm": 2.204277753829956, "learning_rate": 9.518156780717089e-06, "loss": 1.4946, "step": 7277 }, { "epoch": 0.5703761755485893, "grad_norm": 5.013751983642578, "learning_rate": 9.518024697241686e-06, "loss": 1.5008, "step": 7278 }, { "epoch": 0.5704545454545454, "grad_norm": 3.734341621398926, "learning_rate": 9.517892596582038e-06, "loss": 1.9789, "step": 7279 }, { "epoch": 0.5705329153605015, "grad_norm": 5.254560470581055, "learning_rate": 9.517760478738644e-06, "loss": 1.4956, "step": 7280 }, { "epoch": 0.5706112852664577, "grad_norm": 5.194239139556885, "learning_rate": 9.517628343712011e-06, "loss": 1.3432, "step": 7281 }, { "epoch": 0.5706896551724138, "grad_norm": 2.1437408924102783, "learning_rate": 9.51749619150264e-06, "loss": 1.3275, "step": 7282 }, { "epoch": 0.5707680250783699, "grad_norm": 7.001852512359619, "learning_rate": 9.517364022111036e-06, "loss": 1.7683, "step": 7283 }, { "epoch": 0.570846394984326, "grad_norm": 3.04301381111145, "learning_rate": 9.517231835537696e-06, "loss": 1.3113, "step": 7284 }, { "epoch": 0.5709247648902821, "grad_norm": 2.5238523483276367, "learning_rate": 9.517099631783128e-06, "loss": 1.1206, "step": 7285 }, { "epoch": 0.5710031347962382, "grad_norm": 3.030183792114258, "learning_rate": 9.516967410847834e-06, "loss": 1.5633, "step": 7286 }, { "epoch": 0.5710815047021943, "grad_norm": 2.2623367309570312, "learning_rate": 9.516835172732314e-06, "loss": 1.1355, "step": 7287 }, { "epoch": 0.5711598746081504, "grad_norm": 4.124265193939209, "learning_rate": 9.516702917437074e-06, "loss": 1.0941, "step": 7288 }, { "epoch": 0.5712382445141065, "grad_norm": 5.075222492218018, "learning_rate": 9.516570644962617e-06, "loss": 1.4114, "step": 7289 }, { "epoch": 0.5713166144200627, "grad_norm": 5.297740459442139, "learning_rate": 9.516438355309445e-06, "loss": 0.9757, "step": 7290 }, { "epoch": 0.5713949843260188, "grad_norm": 3.4992220401763916, "learning_rate": 9.51630604847806e-06, "loss": 1.6043, "step": 7291 }, { "epoch": 0.5714733542319749, "grad_norm": 2.1049513816833496, "learning_rate": 9.516173724468967e-06, "loss": 1.2554, "step": 7292 }, { "epoch": 0.571551724137931, "grad_norm": 2.399752378463745, "learning_rate": 9.51604138328267e-06, "loss": 1.2452, "step": 7293 }, { "epoch": 0.5716300940438871, "grad_norm": 4.592360019683838, "learning_rate": 9.51590902491967e-06, "loss": 1.9368, "step": 7294 }, { "epoch": 0.5717084639498433, "grad_norm": 2.739147901535034, "learning_rate": 9.515776649380471e-06, "loss": 1.2951, "step": 7295 }, { "epoch": 0.5717868338557994, "grad_norm": 3.172502279281616, "learning_rate": 9.515644256665577e-06, "loss": 1.4297, "step": 7296 }, { "epoch": 0.5718652037617555, "grad_norm": 2.6030232906341553, "learning_rate": 9.515511846775494e-06, "loss": 1.5241, "step": 7297 }, { "epoch": 0.5719435736677116, "grad_norm": 2.2136499881744385, "learning_rate": 9.51537941971072e-06, "loss": 1.285, "step": 7298 }, { "epoch": 0.5720219435736678, "grad_norm": 4.824967384338379, "learning_rate": 9.515246975471764e-06, "loss": 1.3649, "step": 7299 }, { "epoch": 0.5721003134796239, "grad_norm": 2.60050630569458, "learning_rate": 9.515114514059127e-06, "loss": 1.4788, "step": 7300 }, { "epoch": 0.57217868338558, "grad_norm": 3.554124116897583, "learning_rate": 9.514982035473315e-06, "loss": 1.2246, "step": 7301 }, { "epoch": 0.5722570532915361, "grad_norm": 4.140748500823975, "learning_rate": 9.514849539714828e-06, "loss": 1.5588, "step": 7302 }, { "epoch": 0.5723354231974922, "grad_norm": 2.8651161193847656, "learning_rate": 9.514717026784173e-06, "loss": 1.3245, "step": 7303 }, { "epoch": 0.5724137931034483, "grad_norm": 1.9682191610336304, "learning_rate": 9.514584496681855e-06, "loss": 1.1977, "step": 7304 }, { "epoch": 0.5724921630094044, "grad_norm": 3.0704469680786133, "learning_rate": 9.514451949408376e-06, "loss": 1.1889, "step": 7305 }, { "epoch": 0.5725705329153605, "grad_norm": 6.9222798347473145, "learning_rate": 9.514319384964238e-06, "loss": 1.228, "step": 7306 }, { "epoch": 0.5726489028213166, "grad_norm": 2.347167491912842, "learning_rate": 9.51418680334995e-06, "loss": 1.4626, "step": 7307 }, { "epoch": 0.5727272727272728, "grad_norm": 7.352041721343994, "learning_rate": 9.514054204566014e-06, "loss": 1.0386, "step": 7308 }, { "epoch": 0.5728056426332289, "grad_norm": 5.1545305252075195, "learning_rate": 9.513921588612934e-06, "loss": 1.4951, "step": 7309 }, { "epoch": 0.572884012539185, "grad_norm": 2.742701768875122, "learning_rate": 9.513788955491214e-06, "loss": 1.6411, "step": 7310 }, { "epoch": 0.5729623824451411, "grad_norm": 2.3607337474823, "learning_rate": 9.51365630520136e-06, "loss": 1.2042, "step": 7311 }, { "epoch": 0.5730407523510972, "grad_norm": 2.884586811065674, "learning_rate": 9.513523637743875e-06, "loss": 1.8594, "step": 7312 }, { "epoch": 0.5731191222570533, "grad_norm": 2.7700276374816895, "learning_rate": 9.513390953119266e-06, "loss": 1.4649, "step": 7313 }, { "epoch": 0.5731974921630094, "grad_norm": 3.179478883743286, "learning_rate": 9.513258251328034e-06, "loss": 1.4003, "step": 7314 }, { "epoch": 0.5732758620689655, "grad_norm": 1.8470147848129272, "learning_rate": 9.513125532370689e-06, "loss": 1.4559, "step": 7315 }, { "epoch": 0.5733542319749216, "grad_norm": 2.92630934715271, "learning_rate": 9.51299279624773e-06, "loss": 1.6165, "step": 7316 }, { "epoch": 0.5734326018808777, "grad_norm": 3.505417585372925, "learning_rate": 9.512860042959663e-06, "loss": 1.306, "step": 7317 }, { "epoch": 0.5735109717868339, "grad_norm": 2.916419744491577, "learning_rate": 9.512727272506998e-06, "loss": 1.4138, "step": 7318 }, { "epoch": 0.57358934169279, "grad_norm": 4.225955009460449, "learning_rate": 9.512594484890234e-06, "loss": 1.7211, "step": 7319 }, { "epoch": 0.5736677115987461, "grad_norm": 3.48496413230896, "learning_rate": 9.51246168010988e-06, "loss": 1.4005, "step": 7320 }, { "epoch": 0.5737460815047022, "grad_norm": 4.866702556610107, "learning_rate": 9.512328858166437e-06, "loss": 2.0787, "step": 7321 }, { "epoch": 0.5738244514106583, "grad_norm": 3.1905903816223145, "learning_rate": 9.512196019060416e-06, "loss": 1.5219, "step": 7322 }, { "epoch": 0.5739028213166144, "grad_norm": 5.462363243103027, "learning_rate": 9.512063162792316e-06, "loss": 2.3926, "step": 7323 }, { "epoch": 0.5739811912225705, "grad_norm": 3.475773334503174, "learning_rate": 9.511930289362647e-06, "loss": 1.6322, "step": 7324 }, { "epoch": 0.5740595611285266, "grad_norm": 2.5439765453338623, "learning_rate": 9.511797398771912e-06, "loss": 1.0228, "step": 7325 }, { "epoch": 0.5741379310344827, "grad_norm": 2.2530529499053955, "learning_rate": 9.511664491020618e-06, "loss": 1.2174, "step": 7326 }, { "epoch": 0.5742163009404389, "grad_norm": 2.6059517860412598, "learning_rate": 9.511531566109266e-06, "loss": 1.5635, "step": 7327 }, { "epoch": 0.574294670846395, "grad_norm": 6.54159688949585, "learning_rate": 9.511398624038369e-06, "loss": 0.6094, "step": 7328 }, { "epoch": 0.5743730407523511, "grad_norm": 3.927893877029419, "learning_rate": 9.511265664808428e-06, "loss": 0.8616, "step": 7329 }, { "epoch": 0.5744514106583072, "grad_norm": 2.7076735496520996, "learning_rate": 9.51113268841995e-06, "loss": 1.5293, "step": 7330 }, { "epoch": 0.5745297805642633, "grad_norm": 3.26871395111084, "learning_rate": 9.51099969487344e-06, "loss": 1.1866, "step": 7331 }, { "epoch": 0.5746081504702194, "grad_norm": 3.5533010959625244, "learning_rate": 9.510866684169404e-06, "loss": 1.3231, "step": 7332 }, { "epoch": 0.5746865203761755, "grad_norm": 6.519484043121338, "learning_rate": 9.510733656308347e-06, "loss": 1.524, "step": 7333 }, { "epoch": 0.5747648902821316, "grad_norm": 2.538599729537964, "learning_rate": 9.510600611290776e-06, "loss": 1.0589, "step": 7334 }, { "epoch": 0.5748432601880877, "grad_norm": 2.586735248565674, "learning_rate": 9.510467549117199e-06, "loss": 1.4783, "step": 7335 }, { "epoch": 0.5749216300940438, "grad_norm": 7.085366725921631, "learning_rate": 9.510334469788119e-06, "loss": 1.2519, "step": 7336 }, { "epoch": 0.575, "grad_norm": 2.3575878143310547, "learning_rate": 9.510201373304044e-06, "loss": 1.1726, "step": 7337 }, { "epoch": 0.5750783699059561, "grad_norm": 2.7239842414855957, "learning_rate": 9.510068259665478e-06, "loss": 1.4304, "step": 7338 }, { "epoch": 0.5751567398119122, "grad_norm": 9.033981323242188, "learning_rate": 9.50993512887293e-06, "loss": 1.1747, "step": 7339 }, { "epoch": 0.5752351097178683, "grad_norm": 2.4146389961242676, "learning_rate": 9.509801980926906e-06, "loss": 1.3529, "step": 7340 }, { "epoch": 0.5753134796238244, "grad_norm": 2.310668468475342, "learning_rate": 9.50966881582791e-06, "loss": 1.415, "step": 7341 }, { "epoch": 0.5753918495297806, "grad_norm": 2.5472664833068848, "learning_rate": 9.509535633576452e-06, "loss": 1.3649, "step": 7342 }, { "epoch": 0.5754702194357367, "grad_norm": 2.4696357250213623, "learning_rate": 9.509402434173034e-06, "loss": 1.3193, "step": 7343 }, { "epoch": 0.5755485893416928, "grad_norm": 3.16994047164917, "learning_rate": 9.509269217618167e-06, "loss": 1.5199, "step": 7344 }, { "epoch": 0.575626959247649, "grad_norm": 4.125771999359131, "learning_rate": 9.509135983912356e-06, "loss": 1.6808, "step": 7345 }, { "epoch": 0.5757053291536051, "grad_norm": 2.4924051761627197, "learning_rate": 9.509002733056108e-06, "loss": 1.3271, "step": 7346 }, { "epoch": 0.5757836990595612, "grad_norm": 2.1187007427215576, "learning_rate": 9.50886946504993e-06, "loss": 1.3387, "step": 7347 }, { "epoch": 0.5758620689655173, "grad_norm": 1.8226141929626465, "learning_rate": 9.508736179894327e-06, "loss": 1.4575, "step": 7348 }, { "epoch": 0.5759404388714734, "grad_norm": 3.49727725982666, "learning_rate": 9.508602877589808e-06, "loss": 1.899, "step": 7349 }, { "epoch": 0.5760188087774295, "grad_norm": 4.009121894836426, "learning_rate": 9.50846955813688e-06, "loss": 0.7131, "step": 7350 }, { "epoch": 0.5760971786833856, "grad_norm": 3.0256805419921875, "learning_rate": 9.508336221536048e-06, "loss": 1.6263, "step": 7351 }, { "epoch": 0.5761755485893417, "grad_norm": 2.313363552093506, "learning_rate": 9.508202867787821e-06, "loss": 1.4562, "step": 7352 }, { "epoch": 0.5762539184952978, "grad_norm": 2.002934694290161, "learning_rate": 9.508069496892707e-06, "loss": 1.4885, "step": 7353 }, { "epoch": 0.576332288401254, "grad_norm": 3.2867116928100586, "learning_rate": 9.507936108851212e-06, "loss": 1.5583, "step": 7354 }, { "epoch": 0.57641065830721, "grad_norm": 2.538330078125, "learning_rate": 9.507802703663843e-06, "loss": 1.3829, "step": 7355 }, { "epoch": 0.5764890282131662, "grad_norm": 2.6687543392181396, "learning_rate": 9.50766928133111e-06, "loss": 1.1576, "step": 7356 }, { "epoch": 0.5765673981191223, "grad_norm": 5.997373104095459, "learning_rate": 9.507535841853514e-06, "loss": 1.806, "step": 7357 }, { "epoch": 0.5766457680250784, "grad_norm": 7.67262077331543, "learning_rate": 9.50740238523157e-06, "loss": 0.7514, "step": 7358 }, { "epoch": 0.5767241379310345, "grad_norm": 2.8972373008728027, "learning_rate": 9.507268911465782e-06, "loss": 1.1968, "step": 7359 }, { "epoch": 0.5768025078369906, "grad_norm": 6.1098952293396, "learning_rate": 9.507135420556658e-06, "loss": 1.4619, "step": 7360 }, { "epoch": 0.5768808777429467, "grad_norm": 3.8512556552886963, "learning_rate": 9.507001912504705e-06, "loss": 1.7989, "step": 7361 }, { "epoch": 0.5769592476489028, "grad_norm": 7.003735065460205, "learning_rate": 9.506868387310433e-06, "loss": 1.4481, "step": 7362 }, { "epoch": 0.5770376175548589, "grad_norm": 4.585620880126953, "learning_rate": 9.506734844974347e-06, "loss": 1.4715, "step": 7363 }, { "epoch": 0.577115987460815, "grad_norm": 2.015845537185669, "learning_rate": 9.506601285496957e-06, "loss": 1.1888, "step": 7364 }, { "epoch": 0.5771943573667712, "grad_norm": 4.30645751953125, "learning_rate": 9.506467708878771e-06, "loss": 1.4458, "step": 7365 }, { "epoch": 0.5772727272727273, "grad_norm": 2.471698522567749, "learning_rate": 9.506334115120297e-06, "loss": 1.1675, "step": 7366 }, { "epoch": 0.5773510971786834, "grad_norm": 2.4204492568969727, "learning_rate": 9.506200504222041e-06, "loss": 1.0019, "step": 7367 }, { "epoch": 0.5774294670846395, "grad_norm": 3.6284780502319336, "learning_rate": 9.506066876184513e-06, "loss": 1.4121, "step": 7368 }, { "epoch": 0.5775078369905956, "grad_norm": 2.4990837574005127, "learning_rate": 9.505933231008221e-06, "loss": 1.2582, "step": 7369 }, { "epoch": 0.5775862068965517, "grad_norm": 2.337512254714966, "learning_rate": 9.505799568693676e-06, "loss": 1.1611, "step": 7370 }, { "epoch": 0.5776645768025078, "grad_norm": 2.601715564727783, "learning_rate": 9.505665889241382e-06, "loss": 1.47, "step": 7371 }, { "epoch": 0.5777429467084639, "grad_norm": 4.060576438903809, "learning_rate": 9.505532192651848e-06, "loss": 1.4401, "step": 7372 }, { "epoch": 0.57782131661442, "grad_norm": 2.608076333999634, "learning_rate": 9.505398478925585e-06, "loss": 1.4908, "step": 7373 }, { "epoch": 0.5778996865203762, "grad_norm": 2.19258451461792, "learning_rate": 9.5052647480631e-06, "loss": 1.2866, "step": 7374 }, { "epoch": 0.5779780564263323, "grad_norm": 2.1943747997283936, "learning_rate": 9.505131000064903e-06, "loss": 1.244, "step": 7375 }, { "epoch": 0.5780564263322884, "grad_norm": 3.33217191696167, "learning_rate": 9.504997234931501e-06, "loss": 1.4708, "step": 7376 }, { "epoch": 0.5781347962382445, "grad_norm": 4.304467678070068, "learning_rate": 9.504863452663402e-06, "loss": 1.3483, "step": 7377 }, { "epoch": 0.5782131661442006, "grad_norm": 2.6850428581237793, "learning_rate": 9.504729653261118e-06, "loss": 1.5345, "step": 7378 }, { "epoch": 0.5782915360501567, "grad_norm": 2.714993476867676, "learning_rate": 9.504595836725156e-06, "loss": 1.1132, "step": 7379 }, { "epoch": 0.5783699059561128, "grad_norm": 9.098724365234375, "learning_rate": 9.504462003056025e-06, "loss": 1.3918, "step": 7380 }, { "epoch": 0.5784482758620689, "grad_norm": 2.1620583534240723, "learning_rate": 9.504328152254235e-06, "loss": 1.2041, "step": 7381 }, { "epoch": 0.578526645768025, "grad_norm": 4.032106876373291, "learning_rate": 9.504194284320292e-06, "loss": 1.4759, "step": 7382 }, { "epoch": 0.5786050156739811, "grad_norm": 2.3961336612701416, "learning_rate": 9.504060399254709e-06, "loss": 1.2359, "step": 7383 }, { "epoch": 0.5786833855799373, "grad_norm": 12.08173942565918, "learning_rate": 9.503926497057993e-06, "loss": 0.9406, "step": 7384 }, { "epoch": 0.5787617554858934, "grad_norm": 2.645289421081543, "learning_rate": 9.503792577730655e-06, "loss": 1.2224, "step": 7385 }, { "epoch": 0.5788401253918495, "grad_norm": 3.3989429473876953, "learning_rate": 9.503658641273202e-06, "loss": 1.6001, "step": 7386 }, { "epoch": 0.5789184952978056, "grad_norm": 2.7863545417785645, "learning_rate": 9.503524687686145e-06, "loss": 1.2846, "step": 7387 }, { "epoch": 0.5789968652037617, "grad_norm": 3.2989046573638916, "learning_rate": 9.503390716969993e-06, "loss": 1.7511, "step": 7388 }, { "epoch": 0.5790752351097179, "grad_norm": 3.308096408843994, "learning_rate": 9.503256729125257e-06, "loss": 1.9243, "step": 7389 }, { "epoch": 0.579153605015674, "grad_norm": 3.4364376068115234, "learning_rate": 9.503122724152445e-06, "loss": 0.9548, "step": 7390 }, { "epoch": 0.5792319749216301, "grad_norm": 2.013950824737549, "learning_rate": 9.502988702052066e-06, "loss": 1.3013, "step": 7391 }, { "epoch": 0.5793103448275863, "grad_norm": 5.13709020614624, "learning_rate": 9.502854662824632e-06, "loss": 1.1491, "step": 7392 }, { "epoch": 0.5793887147335424, "grad_norm": 1.9147647619247437, "learning_rate": 9.502720606470653e-06, "loss": 1.3385, "step": 7393 }, { "epoch": 0.5794670846394985, "grad_norm": 2.3928849697113037, "learning_rate": 9.502586532990637e-06, "loss": 1.3254, "step": 7394 }, { "epoch": 0.5795454545454546, "grad_norm": 2.5174660682678223, "learning_rate": 9.502452442385093e-06, "loss": 1.7224, "step": 7395 }, { "epoch": 0.5796238244514107, "grad_norm": 2.1118714809417725, "learning_rate": 9.502318334654531e-06, "loss": 1.476, "step": 7396 }, { "epoch": 0.5797021943573668, "grad_norm": 2.5940680503845215, "learning_rate": 9.502184209799466e-06, "loss": 1.4735, "step": 7397 }, { "epoch": 0.5797805642633229, "grad_norm": 3.0562992095947266, "learning_rate": 9.502050067820403e-06, "loss": 1.58, "step": 7398 }, { "epoch": 0.579858934169279, "grad_norm": 2.1858277320861816, "learning_rate": 9.501915908717855e-06, "loss": 1.2025, "step": 7399 }, { "epoch": 0.5799373040752351, "grad_norm": 2.312100648880005, "learning_rate": 9.50178173249233e-06, "loss": 1.0985, "step": 7400 }, { "epoch": 0.5800156739811912, "grad_norm": 4.0440354347229, "learning_rate": 9.50164753914434e-06, "loss": 1.3135, "step": 7401 }, { "epoch": 0.5800940438871474, "grad_norm": 3.4511654376983643, "learning_rate": 9.501513328674395e-06, "loss": 1.3757, "step": 7402 }, { "epoch": 0.5801724137931035, "grad_norm": 3.0240633487701416, "learning_rate": 9.501379101083006e-06, "loss": 1.6272, "step": 7403 }, { "epoch": 0.5802507836990596, "grad_norm": 5.638311862945557, "learning_rate": 9.501244856370681e-06, "loss": 1.7479, "step": 7404 }, { "epoch": 0.5803291536050157, "grad_norm": 1.6604613065719604, "learning_rate": 9.501110594537935e-06, "loss": 1.1516, "step": 7405 }, { "epoch": 0.5804075235109718, "grad_norm": 1.842867374420166, "learning_rate": 9.500976315585276e-06, "loss": 0.8966, "step": 7406 }, { "epoch": 0.5804858934169279, "grad_norm": 2.25773024559021, "learning_rate": 9.500842019513214e-06, "loss": 1.2105, "step": 7407 }, { "epoch": 0.580564263322884, "grad_norm": 2.35795259475708, "learning_rate": 9.500707706322261e-06, "loss": 1.5703, "step": 7408 }, { "epoch": 0.5806426332288401, "grad_norm": 2.4196696281433105, "learning_rate": 9.500573376012928e-06, "loss": 1.364, "step": 7409 }, { "epoch": 0.5807210031347962, "grad_norm": 2.1143248081207275, "learning_rate": 9.500439028585724e-06, "loss": 1.384, "step": 7410 }, { "epoch": 0.5807993730407524, "grad_norm": 5.254983425140381, "learning_rate": 9.500304664041163e-06, "loss": 1.5471, "step": 7411 }, { "epoch": 0.5808777429467085, "grad_norm": 1.9175975322723389, "learning_rate": 9.500170282379753e-06, "loss": 1.3429, "step": 7412 }, { "epoch": 0.5809561128526646, "grad_norm": 2.507066249847412, "learning_rate": 9.500035883602009e-06, "loss": 1.4022, "step": 7413 }, { "epoch": 0.5810344827586207, "grad_norm": 2.3735923767089844, "learning_rate": 9.499901467708437e-06, "loss": 1.5731, "step": 7414 }, { "epoch": 0.5811128526645768, "grad_norm": 4.190301418304443, "learning_rate": 9.499767034699555e-06, "loss": 1.1811, "step": 7415 }, { "epoch": 0.5811912225705329, "grad_norm": 2.5558207035064697, "learning_rate": 9.499632584575866e-06, "loss": 1.3633, "step": 7416 }, { "epoch": 0.581269592476489, "grad_norm": 3.926596164703369, "learning_rate": 9.49949811733789e-06, "loss": 0.9096, "step": 7417 }, { "epoch": 0.5813479623824451, "grad_norm": 3.337716579437256, "learning_rate": 9.499363632986132e-06, "loss": 1.1913, "step": 7418 }, { "epoch": 0.5814263322884012, "grad_norm": 2.2998809814453125, "learning_rate": 9.499229131521104e-06, "loss": 1.3476, "step": 7419 }, { "epoch": 0.5815047021943573, "grad_norm": 3.623420476913452, "learning_rate": 9.499094612943323e-06, "loss": 1.6945, "step": 7420 }, { "epoch": 0.5815830721003135, "grad_norm": 2.0620343685150146, "learning_rate": 9.498960077253295e-06, "loss": 1.1654, "step": 7421 }, { "epoch": 0.5816614420062696, "grad_norm": 3.1849114894866943, "learning_rate": 9.498825524451533e-06, "loss": 1.3003, "step": 7422 }, { "epoch": 0.5817398119122257, "grad_norm": 2.8270182609558105, "learning_rate": 9.49869095453855e-06, "loss": 1.3023, "step": 7423 }, { "epoch": 0.5818181818181818, "grad_norm": 8.425070762634277, "learning_rate": 9.498556367514859e-06, "loss": 0.8373, "step": 7424 }, { "epoch": 0.5818965517241379, "grad_norm": 7.96156644821167, "learning_rate": 9.498421763380969e-06, "loss": 1.4295, "step": 7425 }, { "epoch": 0.581974921630094, "grad_norm": 3.154594898223877, "learning_rate": 9.498287142137392e-06, "loss": 1.2503, "step": 7426 }, { "epoch": 0.5820532915360501, "grad_norm": 3.23657488822937, "learning_rate": 9.498152503784641e-06, "loss": 1.4974, "step": 7427 }, { "epoch": 0.5821316614420062, "grad_norm": 3.2011733055114746, "learning_rate": 9.49801784832323e-06, "loss": 1.201, "step": 7428 }, { "epoch": 0.5822100313479623, "grad_norm": 4.334405899047852, "learning_rate": 9.497883175753671e-06, "loss": 1.4929, "step": 7429 }, { "epoch": 0.5822884012539185, "grad_norm": 2.884443998336792, "learning_rate": 9.497748486076472e-06, "loss": 1.4899, "step": 7430 }, { "epoch": 0.5823667711598746, "grad_norm": 2.83085560798645, "learning_rate": 9.497613779292147e-06, "loss": 0.923, "step": 7431 }, { "epoch": 0.5824451410658307, "grad_norm": 1.8962401151657104, "learning_rate": 9.497479055401212e-06, "loss": 1.1705, "step": 7432 }, { "epoch": 0.5825235109717868, "grad_norm": 2.795024871826172, "learning_rate": 9.497344314404176e-06, "loss": 1.5573, "step": 7433 }, { "epoch": 0.5826018808777429, "grad_norm": 2.867530584335327, "learning_rate": 9.49720955630155e-06, "loss": 1.4742, "step": 7434 }, { "epoch": 0.582680250783699, "grad_norm": 2.8558788299560547, "learning_rate": 9.497074781093851e-06, "loss": 1.3805, "step": 7435 }, { "epoch": 0.5827586206896552, "grad_norm": 3.4995789527893066, "learning_rate": 9.496939988781589e-06, "loss": 1.9401, "step": 7436 }, { "epoch": 0.5828369905956113, "grad_norm": 3.5744848251342773, "learning_rate": 9.496805179365277e-06, "loss": 1.5511, "step": 7437 }, { "epoch": 0.5829153605015674, "grad_norm": 3.6395180225372314, "learning_rate": 9.496670352845428e-06, "loss": 1.2136, "step": 7438 }, { "epoch": 0.5829937304075236, "grad_norm": 3.014683961868286, "learning_rate": 9.496535509222554e-06, "loss": 1.4721, "step": 7439 }, { "epoch": 0.5830721003134797, "grad_norm": 2.719707727432251, "learning_rate": 9.496400648497169e-06, "loss": 1.2186, "step": 7440 }, { "epoch": 0.5831504702194358, "grad_norm": 2.6530497074127197, "learning_rate": 9.496265770669786e-06, "loss": 1.7356, "step": 7441 }, { "epoch": 0.5832288401253919, "grad_norm": 3.3610143661499023, "learning_rate": 9.496130875740916e-06, "loss": 1.6893, "step": 7442 }, { "epoch": 0.583307210031348, "grad_norm": 3.5297698974609375, "learning_rate": 9.495995963711075e-06, "loss": 1.3249, "step": 7443 }, { "epoch": 0.5833855799373041, "grad_norm": 2.4306623935699463, "learning_rate": 9.495861034580774e-06, "loss": 1.2494, "step": 7444 }, { "epoch": 0.5834639498432602, "grad_norm": 5.202731132507324, "learning_rate": 9.495726088350526e-06, "loss": 1.9125, "step": 7445 }, { "epoch": 0.5835423197492163, "grad_norm": 3.3251943588256836, "learning_rate": 9.495591125020846e-06, "loss": 1.455, "step": 7446 }, { "epoch": 0.5836206896551724, "grad_norm": 2.454362392425537, "learning_rate": 9.495456144592245e-06, "loss": 1.4145, "step": 7447 }, { "epoch": 0.5836990595611286, "grad_norm": 3.5081946849823, "learning_rate": 9.49532114706524e-06, "loss": 1.6865, "step": 7448 }, { "epoch": 0.5837774294670847, "grad_norm": 5.126033782958984, "learning_rate": 9.495186132440342e-06, "loss": 1.4614, "step": 7449 }, { "epoch": 0.5838557993730408, "grad_norm": 2.8442251682281494, "learning_rate": 9.495051100718063e-06, "loss": 1.6739, "step": 7450 }, { "epoch": 0.5839341692789969, "grad_norm": 2.690324306488037, "learning_rate": 9.49491605189892e-06, "loss": 1.3876, "step": 7451 }, { "epoch": 0.584012539184953, "grad_norm": 3.134549617767334, "learning_rate": 9.494780985983425e-06, "loss": 1.3, "step": 7452 }, { "epoch": 0.5840909090909091, "grad_norm": 2.500976085662842, "learning_rate": 9.494645902972091e-06, "loss": 1.1239, "step": 7453 }, { "epoch": 0.5841692789968652, "grad_norm": 3.297304153442383, "learning_rate": 9.494510802865433e-06, "loss": 1.2303, "step": 7454 }, { "epoch": 0.5842476489028213, "grad_norm": 5.094430923461914, "learning_rate": 9.494375685663965e-06, "loss": 1.9088, "step": 7455 }, { "epoch": 0.5843260188087774, "grad_norm": 3.4647369384765625, "learning_rate": 9.4942405513682e-06, "loss": 1.5331, "step": 7456 }, { "epoch": 0.5844043887147335, "grad_norm": 10.575984954833984, "learning_rate": 9.494105399978653e-06, "loss": 1.2357, "step": 7457 }, { "epoch": 0.5844827586206897, "grad_norm": 3.2071850299835205, "learning_rate": 9.493970231495836e-06, "loss": 1.5398, "step": 7458 }, { "epoch": 0.5845611285266458, "grad_norm": 2.6102466583251953, "learning_rate": 9.493835045920266e-06, "loss": 1.346, "step": 7459 }, { "epoch": 0.5846394984326019, "grad_norm": 2.4466304779052734, "learning_rate": 9.493699843252454e-06, "loss": 1.6114, "step": 7460 }, { "epoch": 0.584717868338558, "grad_norm": 5.020155906677246, "learning_rate": 9.493564623492917e-06, "loss": 1.6557, "step": 7461 }, { "epoch": 0.5847962382445141, "grad_norm": 5.143722057342529, "learning_rate": 9.493429386642167e-06, "loss": 1.4066, "step": 7462 }, { "epoch": 0.5848746081504702, "grad_norm": 6.241345405578613, "learning_rate": 9.493294132700721e-06, "loss": 1.7055, "step": 7463 }, { "epoch": 0.5849529780564263, "grad_norm": 5.222882270812988, "learning_rate": 9.493158861669091e-06, "loss": 1.4716, "step": 7464 }, { "epoch": 0.5850313479623824, "grad_norm": 2.461073398590088, "learning_rate": 9.493023573547793e-06, "loss": 1.5271, "step": 7465 }, { "epoch": 0.5851097178683385, "grad_norm": 3.7398533821105957, "learning_rate": 9.492888268337341e-06, "loss": 1.0858, "step": 7466 }, { "epoch": 0.5851880877742947, "grad_norm": 4.470871448516846, "learning_rate": 9.49275294603825e-06, "loss": 1.3653, "step": 7467 }, { "epoch": 0.5852664576802508, "grad_norm": 9.014232635498047, "learning_rate": 9.492617606651034e-06, "loss": 0.6878, "step": 7468 }, { "epoch": 0.5853448275862069, "grad_norm": 2.5518717765808105, "learning_rate": 9.492482250176207e-06, "loss": 1.3272, "step": 7469 }, { "epoch": 0.585423197492163, "grad_norm": 2.762051820755005, "learning_rate": 9.492346876614287e-06, "loss": 1.289, "step": 7470 }, { "epoch": 0.5855015673981191, "grad_norm": 2.8978071212768555, "learning_rate": 9.492211485965787e-06, "loss": 1.7546, "step": 7471 }, { "epoch": 0.5855799373040752, "grad_norm": 5.123143672943115, "learning_rate": 9.49207607823122e-06, "loss": 1.3496, "step": 7472 }, { "epoch": 0.5856583072100313, "grad_norm": 2.3694326877593994, "learning_rate": 9.491940653411105e-06, "loss": 1.324, "step": 7473 }, { "epoch": 0.5857366771159874, "grad_norm": 2.518756628036499, "learning_rate": 9.491805211505953e-06, "loss": 1.223, "step": 7474 }, { "epoch": 0.5858150470219435, "grad_norm": 2.5387282371520996, "learning_rate": 9.491669752516284e-06, "loss": 1.3411, "step": 7475 }, { "epoch": 0.5858934169278996, "grad_norm": 5.0620551109313965, "learning_rate": 9.491534276442608e-06, "loss": 1.2551, "step": 7476 }, { "epoch": 0.5859717868338558, "grad_norm": 3.4985854625701904, "learning_rate": 9.491398783285442e-06, "loss": 1.295, "step": 7477 }, { "epoch": 0.5860501567398119, "grad_norm": 3.499065637588501, "learning_rate": 9.491263273045303e-06, "loss": 1.2985, "step": 7478 }, { "epoch": 0.586128526645768, "grad_norm": 1.749133586883545, "learning_rate": 9.491127745722706e-06, "loss": 1.1389, "step": 7479 }, { "epoch": 0.5862068965517241, "grad_norm": 3.159905195236206, "learning_rate": 9.490992201318165e-06, "loss": 1.2637, "step": 7480 }, { "epoch": 0.5862852664576802, "grad_norm": 6.111545562744141, "learning_rate": 9.490856639832197e-06, "loss": 2.2314, "step": 7481 }, { "epoch": 0.5863636363636363, "grad_norm": 3.124752998352051, "learning_rate": 9.490721061265315e-06, "loss": 1.386, "step": 7482 }, { "epoch": 0.5864420062695925, "grad_norm": 3.2693278789520264, "learning_rate": 9.490585465618039e-06, "loss": 1.0929, "step": 7483 }, { "epoch": 0.5865203761755486, "grad_norm": 2.4197499752044678, "learning_rate": 9.490449852890882e-06, "loss": 1.0093, "step": 7484 }, { "epoch": 0.5865987460815048, "grad_norm": 2.640650987625122, "learning_rate": 9.490314223084361e-06, "loss": 1.5076, "step": 7485 }, { "epoch": 0.5866771159874609, "grad_norm": 6.485011577606201, "learning_rate": 9.49017857619899e-06, "loss": 1.1933, "step": 7486 }, { "epoch": 0.586755485893417, "grad_norm": 4.68845272064209, "learning_rate": 9.490042912235284e-06, "loss": 1.1143, "step": 7487 }, { "epoch": 0.5868338557993731, "grad_norm": 2.4168014526367188, "learning_rate": 9.489907231193763e-06, "loss": 1.2097, "step": 7488 }, { "epoch": 0.5869122257053292, "grad_norm": 3.4347946643829346, "learning_rate": 9.489771533074941e-06, "loss": 1.8739, "step": 7489 }, { "epoch": 0.5869905956112853, "grad_norm": 3.164980411529541, "learning_rate": 9.489635817879334e-06, "loss": 1.2138, "step": 7490 }, { "epoch": 0.5870689655172414, "grad_norm": 4.159703254699707, "learning_rate": 9.48950008560746e-06, "loss": 1.4855, "step": 7491 }, { "epoch": 0.5871473354231975, "grad_norm": 2.9100418090820312, "learning_rate": 9.489364336259833e-06, "loss": 1.2718, "step": 7492 }, { "epoch": 0.5872257053291536, "grad_norm": 5.571642875671387, "learning_rate": 9.489228569836967e-06, "loss": 1.1978, "step": 7493 }, { "epoch": 0.5873040752351097, "grad_norm": 4.0714263916015625, "learning_rate": 9.489092786339383e-06, "loss": 1.6372, "step": 7494 }, { "epoch": 0.5873824451410659, "grad_norm": 2.0534915924072266, "learning_rate": 9.488956985767596e-06, "loss": 1.274, "step": 7495 }, { "epoch": 0.587460815047022, "grad_norm": 3.5961215496063232, "learning_rate": 9.488821168122124e-06, "loss": 1.3464, "step": 7496 }, { "epoch": 0.5875391849529781, "grad_norm": 6.076894283294678, "learning_rate": 9.48868533340348e-06, "loss": 1.4566, "step": 7497 }, { "epoch": 0.5876175548589342, "grad_norm": 3.7443106174468994, "learning_rate": 9.488549481612183e-06, "loss": 1.1595, "step": 7498 }, { "epoch": 0.5876959247648903, "grad_norm": 7.410421848297119, "learning_rate": 9.488413612748748e-06, "loss": 1.6042, "step": 7499 }, { "epoch": 0.5877742946708464, "grad_norm": 5.517515659332275, "learning_rate": 9.488277726813696e-06, "loss": 1.866, "step": 7500 }, { "epoch": 0.5878526645768025, "grad_norm": 2.517303943634033, "learning_rate": 9.488141823807538e-06, "loss": 1.3819, "step": 7501 }, { "epoch": 0.5879310344827586, "grad_norm": 2.3157756328582764, "learning_rate": 9.488005903730795e-06, "loss": 1.4208, "step": 7502 }, { "epoch": 0.5880094043887147, "grad_norm": 8.217096328735352, "learning_rate": 9.487869966583982e-06, "loss": 0.6574, "step": 7503 }, { "epoch": 0.5880877742946709, "grad_norm": 2.994131326675415, "learning_rate": 9.487734012367618e-06, "loss": 1.5948, "step": 7504 }, { "epoch": 0.588166144200627, "grad_norm": 3.5140092372894287, "learning_rate": 9.487598041082218e-06, "loss": 1.325, "step": 7505 }, { "epoch": 0.5882445141065831, "grad_norm": 3.2034096717834473, "learning_rate": 9.4874620527283e-06, "loss": 1.4034, "step": 7506 }, { "epoch": 0.5883228840125392, "grad_norm": 2.3341662883758545, "learning_rate": 9.487326047306381e-06, "loss": 1.5104, "step": 7507 }, { "epoch": 0.5884012539184953, "grad_norm": 2.3524842262268066, "learning_rate": 9.487190024816978e-06, "loss": 1.3343, "step": 7508 }, { "epoch": 0.5884796238244514, "grad_norm": 3.348177433013916, "learning_rate": 9.48705398526061e-06, "loss": 1.4464, "step": 7509 }, { "epoch": 0.5885579937304075, "grad_norm": 4.614846229553223, "learning_rate": 9.486917928637793e-06, "loss": 1.1859, "step": 7510 }, { "epoch": 0.5886363636363636, "grad_norm": 5.638894557952881, "learning_rate": 9.486781854949046e-06, "loss": 1.5575, "step": 7511 }, { "epoch": 0.5887147335423197, "grad_norm": 6.770296096801758, "learning_rate": 9.486645764194883e-06, "loss": 1.6342, "step": 7512 }, { "epoch": 0.5887931034482758, "grad_norm": 3.7652463912963867, "learning_rate": 9.486509656375827e-06, "loss": 1.5889, "step": 7513 }, { "epoch": 0.588871473354232, "grad_norm": 2.8733339309692383, "learning_rate": 9.48637353149239e-06, "loss": 1.2218, "step": 7514 }, { "epoch": 0.5889498432601881, "grad_norm": 4.008521556854248, "learning_rate": 9.486237389545094e-06, "loss": 1.5418, "step": 7515 }, { "epoch": 0.5890282131661442, "grad_norm": 3.3443055152893066, "learning_rate": 9.486101230534455e-06, "loss": 1.0549, "step": 7516 }, { "epoch": 0.5891065830721003, "grad_norm": 3.2124881744384766, "learning_rate": 9.485965054460991e-06, "loss": 1.7817, "step": 7517 }, { "epoch": 0.5891849529780564, "grad_norm": 3.252587080001831, "learning_rate": 9.485828861325219e-06, "loss": 1.7881, "step": 7518 }, { "epoch": 0.5892633228840125, "grad_norm": 6.143857955932617, "learning_rate": 9.48569265112766e-06, "loss": 2.1758, "step": 7519 }, { "epoch": 0.5893416927899686, "grad_norm": 3.7278850078582764, "learning_rate": 9.48555642386883e-06, "loss": 1.3069, "step": 7520 }, { "epoch": 0.5894200626959247, "grad_norm": 3.6638524532318115, "learning_rate": 9.485420179549245e-06, "loss": 1.7032, "step": 7521 }, { "epoch": 0.5894984326018808, "grad_norm": 5.416196823120117, "learning_rate": 9.485283918169428e-06, "loss": 1.1363, "step": 7522 }, { "epoch": 0.589576802507837, "grad_norm": 3.0451197624206543, "learning_rate": 9.485147639729892e-06, "loss": 1.4602, "step": 7523 }, { "epoch": 0.5896551724137931, "grad_norm": 3.6345295906066895, "learning_rate": 9.48501134423116e-06, "loss": 1.2948, "step": 7524 }, { "epoch": 0.5897335423197492, "grad_norm": 3.2140300273895264, "learning_rate": 9.48487503167375e-06, "loss": 0.9757, "step": 7525 }, { "epoch": 0.5898119122257053, "grad_norm": 2.7903032302856445, "learning_rate": 9.484738702058176e-06, "loss": 1.6298, "step": 7526 }, { "epoch": 0.5898902821316614, "grad_norm": 2.3385186195373535, "learning_rate": 9.484602355384961e-06, "loss": 1.6323, "step": 7527 }, { "epoch": 0.5899686520376175, "grad_norm": 3.0264177322387695, "learning_rate": 9.484465991654622e-06, "loss": 1.3603, "step": 7528 }, { "epoch": 0.5900470219435736, "grad_norm": 6.070839881896973, "learning_rate": 9.484329610867677e-06, "loss": 1.613, "step": 7529 }, { "epoch": 0.5901253918495298, "grad_norm": 4.424251556396484, "learning_rate": 9.484193213024647e-06, "loss": 1.8514, "step": 7530 }, { "epoch": 0.590203761755486, "grad_norm": 3.7231271266937256, "learning_rate": 9.484056798126048e-06, "loss": 1.6935, "step": 7531 }, { "epoch": 0.590282131661442, "grad_norm": 3.8700497150421143, "learning_rate": 9.483920366172401e-06, "loss": 1.4559, "step": 7532 }, { "epoch": 0.5903605015673982, "grad_norm": 3.907089948654175, "learning_rate": 9.483783917164223e-06, "loss": 0.9527, "step": 7533 }, { "epoch": 0.5904388714733543, "grad_norm": 3.7016284465789795, "learning_rate": 9.483647451102033e-06, "loss": 1.6424, "step": 7534 }, { "epoch": 0.5905172413793104, "grad_norm": 2.009277582168579, "learning_rate": 9.483510967986354e-06, "loss": 1.2929, "step": 7535 }, { "epoch": 0.5905956112852665, "grad_norm": 2.733552932739258, "learning_rate": 9.483374467817701e-06, "loss": 1.5657, "step": 7536 }, { "epoch": 0.5906739811912226, "grad_norm": 6.242856502532959, "learning_rate": 9.483237950596593e-06, "loss": 0.9239, "step": 7537 }, { "epoch": 0.5907523510971787, "grad_norm": 2.6170742511749268, "learning_rate": 9.483101416323553e-06, "loss": 1.4165, "step": 7538 }, { "epoch": 0.5908307210031348, "grad_norm": 1.9170663356781006, "learning_rate": 9.482964864999096e-06, "loss": 1.4986, "step": 7539 }, { "epoch": 0.5909090909090909, "grad_norm": 3.177211046218872, "learning_rate": 9.482828296623743e-06, "loss": 1.4417, "step": 7540 }, { "epoch": 0.590987460815047, "grad_norm": 3.941957712173462, "learning_rate": 9.482691711198013e-06, "loss": 1.6421, "step": 7541 }, { "epoch": 0.5910658307210032, "grad_norm": 2.487844705581665, "learning_rate": 9.482555108722428e-06, "loss": 1.642, "step": 7542 }, { "epoch": 0.5911442006269593, "grad_norm": 2.9490678310394287, "learning_rate": 9.482418489197506e-06, "loss": 1.3469, "step": 7543 }, { "epoch": 0.5912225705329154, "grad_norm": 2.085277557373047, "learning_rate": 9.482281852623765e-06, "loss": 1.279, "step": 7544 }, { "epoch": 0.5913009404388715, "grad_norm": 2.7345738410949707, "learning_rate": 9.482145199001726e-06, "loss": 1.1985, "step": 7545 }, { "epoch": 0.5913793103448276, "grad_norm": 2.05661678314209, "learning_rate": 9.48200852833191e-06, "loss": 1.0255, "step": 7546 }, { "epoch": 0.5914576802507837, "grad_norm": 2.8623082637786865, "learning_rate": 9.481871840614835e-06, "loss": 1.4557, "step": 7547 }, { "epoch": 0.5915360501567398, "grad_norm": 3.1278200149536133, "learning_rate": 9.481735135851022e-06, "loss": 1.2003, "step": 7548 }, { "epoch": 0.5916144200626959, "grad_norm": 3.4508190155029297, "learning_rate": 9.48159841404099e-06, "loss": 1.5497, "step": 7549 }, { "epoch": 0.591692789968652, "grad_norm": 2.9135921001434326, "learning_rate": 9.48146167518526e-06, "loss": 1.112, "step": 7550 }, { "epoch": 0.5917711598746082, "grad_norm": 2.518284797668457, "learning_rate": 9.481324919284352e-06, "loss": 1.3849, "step": 7551 }, { "epoch": 0.5918495297805643, "grad_norm": 4.8686113357543945, "learning_rate": 9.481188146338784e-06, "loss": 1.2241, "step": 7552 }, { "epoch": 0.5919278996865204, "grad_norm": 2.11932373046875, "learning_rate": 9.48105135634908e-06, "loss": 1.1756, "step": 7553 }, { "epoch": 0.5920062695924765, "grad_norm": 2.391270875930786, "learning_rate": 9.480914549315758e-06, "loss": 1.1906, "step": 7554 }, { "epoch": 0.5920846394984326, "grad_norm": 4.5249810218811035, "learning_rate": 9.480777725239338e-06, "loss": 1.6717, "step": 7555 }, { "epoch": 0.5921630094043887, "grad_norm": 3.570260524749756, "learning_rate": 9.48064088412034e-06, "loss": 1.4035, "step": 7556 }, { "epoch": 0.5922413793103448, "grad_norm": 4.461014270782471, "learning_rate": 9.480504025959287e-06, "loss": 1.0749, "step": 7557 }, { "epoch": 0.5923197492163009, "grad_norm": 2.33577036857605, "learning_rate": 9.480367150756697e-06, "loss": 1.47, "step": 7558 }, { "epoch": 0.592398119122257, "grad_norm": 3.277747869491577, "learning_rate": 9.480230258513092e-06, "loss": 1.3929, "step": 7559 }, { "epoch": 0.5924764890282131, "grad_norm": 1.8569384813308716, "learning_rate": 9.480093349228994e-06, "loss": 1.41, "step": 7560 }, { "epoch": 0.5925548589341693, "grad_norm": 3.107987403869629, "learning_rate": 9.47995642290492e-06, "loss": 0.9825, "step": 7561 }, { "epoch": 0.5926332288401254, "grad_norm": 2.0536746978759766, "learning_rate": 9.479819479541392e-06, "loss": 1.0493, "step": 7562 }, { "epoch": 0.5927115987460815, "grad_norm": 2.907123327255249, "learning_rate": 9.479682519138934e-06, "loss": 1.3268, "step": 7563 }, { "epoch": 0.5927899686520376, "grad_norm": 2.946704626083374, "learning_rate": 9.479545541698062e-06, "loss": 1.8539, "step": 7564 }, { "epoch": 0.5928683385579937, "grad_norm": 2.281991481781006, "learning_rate": 9.479408547219303e-06, "loss": 1.4131, "step": 7565 }, { "epoch": 0.5929467084639498, "grad_norm": 2.833592653274536, "learning_rate": 9.479271535703172e-06, "loss": 1.3256, "step": 7566 }, { "epoch": 0.5930250783699059, "grad_norm": 3.6709468364715576, "learning_rate": 9.479134507150195e-06, "loss": 1.6099, "step": 7567 }, { "epoch": 0.593103448275862, "grad_norm": 2.3076136112213135, "learning_rate": 9.478997461560888e-06, "loss": 1.2397, "step": 7568 }, { "epoch": 0.5931818181818181, "grad_norm": 2.198819875717163, "learning_rate": 9.478860398935776e-06, "loss": 1.5711, "step": 7569 }, { "epoch": 0.5932601880877743, "grad_norm": 5.205688953399658, "learning_rate": 9.47872331927538e-06, "loss": 1.4425, "step": 7570 }, { "epoch": 0.5933385579937304, "grad_norm": 2.7639660835266113, "learning_rate": 9.478586222580222e-06, "loss": 1.4254, "step": 7571 }, { "epoch": 0.5934169278996865, "grad_norm": 2.800189256668091, "learning_rate": 9.478449108850823e-06, "loss": 1.773, "step": 7572 }, { "epoch": 0.5934952978056426, "grad_norm": 4.509793758392334, "learning_rate": 9.478311978087701e-06, "loss": 1.2549, "step": 7573 }, { "epoch": 0.5935736677115987, "grad_norm": 6.396040916442871, "learning_rate": 9.478174830291382e-06, "loss": 1.7499, "step": 7574 }, { "epoch": 0.5936520376175548, "grad_norm": 2.6846108436584473, "learning_rate": 9.478037665462386e-06, "loss": 1.5187, "step": 7575 }, { "epoch": 0.5937304075235109, "grad_norm": 5.2540693283081055, "learning_rate": 9.477900483601235e-06, "loss": 1.5442, "step": 7576 }, { "epoch": 0.5938087774294671, "grad_norm": 3.2528891563415527, "learning_rate": 9.47776328470845e-06, "loss": 1.5856, "step": 7577 }, { "epoch": 0.5938871473354232, "grad_norm": 2.6989381313323975, "learning_rate": 9.477626068784554e-06, "loss": 1.3953, "step": 7578 }, { "epoch": 0.5939655172413794, "grad_norm": 3.364481210708618, "learning_rate": 9.477488835830068e-06, "loss": 1.4082, "step": 7579 }, { "epoch": 0.5940438871473355, "grad_norm": 3.1219003200531006, "learning_rate": 9.477351585845513e-06, "loss": 1.2982, "step": 7580 }, { "epoch": 0.5941222570532916, "grad_norm": 3.600250244140625, "learning_rate": 9.477214318831415e-06, "loss": 1.4914, "step": 7581 }, { "epoch": 0.5942006269592477, "grad_norm": 17.185501098632812, "learning_rate": 9.477077034788292e-06, "loss": 1.5247, "step": 7582 }, { "epoch": 0.5942789968652038, "grad_norm": 2.1306231021881104, "learning_rate": 9.476939733716667e-06, "loss": 1.4114, "step": 7583 }, { "epoch": 0.5943573667711599, "grad_norm": 2.570916175842285, "learning_rate": 9.476802415617064e-06, "loss": 1.6825, "step": 7584 }, { "epoch": 0.594435736677116, "grad_norm": 6.976287364959717, "learning_rate": 9.476665080490002e-06, "loss": 1.3518, "step": 7585 }, { "epoch": 0.5945141065830721, "grad_norm": 2.8626649379730225, "learning_rate": 9.476527728336007e-06, "loss": 1.6751, "step": 7586 }, { "epoch": 0.5945924764890282, "grad_norm": 3.6528780460357666, "learning_rate": 9.476390359155599e-06, "loss": 0.7247, "step": 7587 }, { "epoch": 0.5946708463949844, "grad_norm": 2.416673183441162, "learning_rate": 9.476252972949302e-06, "loss": 1.3486, "step": 7588 }, { "epoch": 0.5947492163009405, "grad_norm": 2.9778454303741455, "learning_rate": 9.476115569717637e-06, "loss": 1.3487, "step": 7589 }, { "epoch": 0.5948275862068966, "grad_norm": 2.5045502185821533, "learning_rate": 9.475978149461129e-06, "loss": 1.4691, "step": 7590 }, { "epoch": 0.5949059561128527, "grad_norm": 6.217997074127197, "learning_rate": 9.475840712180297e-06, "loss": 1.3045, "step": 7591 }, { "epoch": 0.5949843260188088, "grad_norm": 2.5505404472351074, "learning_rate": 9.475703257875669e-06, "loss": 1.504, "step": 7592 }, { "epoch": 0.5950626959247649, "grad_norm": 2.785900115966797, "learning_rate": 9.475565786547762e-06, "loss": 1.5287, "step": 7593 }, { "epoch": 0.595141065830721, "grad_norm": 2.649610757827759, "learning_rate": 9.475428298197101e-06, "loss": 1.4361, "step": 7594 }, { "epoch": 0.5952194357366771, "grad_norm": 2.841019868850708, "learning_rate": 9.47529079282421e-06, "loss": 1.1428, "step": 7595 }, { "epoch": 0.5952978056426332, "grad_norm": 2.5150089263916016, "learning_rate": 9.475153270429614e-06, "loss": 1.4284, "step": 7596 }, { "epoch": 0.5953761755485893, "grad_norm": 5.976132392883301, "learning_rate": 9.47501573101383e-06, "loss": 1.8815, "step": 7597 }, { "epoch": 0.5954545454545455, "grad_norm": 2.5446035861968994, "learning_rate": 9.474878174577386e-06, "loss": 1.2555, "step": 7598 }, { "epoch": 0.5955329153605016, "grad_norm": 3.163970708847046, "learning_rate": 9.474740601120806e-06, "loss": 1.6537, "step": 7599 }, { "epoch": 0.5956112852664577, "grad_norm": 6.4220757484436035, "learning_rate": 9.474603010644608e-06, "loss": 1.7313, "step": 7600 }, { "epoch": 0.5956896551724138, "grad_norm": 3.748570203781128, "learning_rate": 9.47446540314932e-06, "loss": 1.0212, "step": 7601 }, { "epoch": 0.5957680250783699, "grad_norm": 2.0437564849853516, "learning_rate": 9.474327778635464e-06, "loss": 1.4279, "step": 7602 }, { "epoch": 0.595846394984326, "grad_norm": 8.134843826293945, "learning_rate": 9.474190137103562e-06, "loss": 1.6733, "step": 7603 }, { "epoch": 0.5959247648902821, "grad_norm": 3.016840934753418, "learning_rate": 9.47405247855414e-06, "loss": 1.2022, "step": 7604 }, { "epoch": 0.5960031347962382, "grad_norm": 9.589198112487793, "learning_rate": 9.473914802987721e-06, "loss": 1.249, "step": 7605 }, { "epoch": 0.5960815047021943, "grad_norm": 2.551798105239868, "learning_rate": 9.473777110404827e-06, "loss": 1.3612, "step": 7606 }, { "epoch": 0.5961598746081505, "grad_norm": 2.212216377258301, "learning_rate": 9.473639400805982e-06, "loss": 1.5651, "step": 7607 }, { "epoch": 0.5962382445141066, "grad_norm": 6.016223430633545, "learning_rate": 9.473501674191711e-06, "loss": 1.5826, "step": 7608 }, { "epoch": 0.5963166144200627, "grad_norm": 6.2855706214904785, "learning_rate": 9.47336393056254e-06, "loss": 1.0989, "step": 7609 }, { "epoch": 0.5963949843260188, "grad_norm": 3.2515652179718018, "learning_rate": 9.473226169918987e-06, "loss": 1.3748, "step": 7610 }, { "epoch": 0.5964733542319749, "grad_norm": 2.0200490951538086, "learning_rate": 9.473088392261581e-06, "loss": 1.2004, "step": 7611 }, { "epoch": 0.596551724137931, "grad_norm": 3.2394320964813232, "learning_rate": 9.472950597590844e-06, "loss": 1.891, "step": 7612 }, { "epoch": 0.5966300940438871, "grad_norm": 8.5852689743042, "learning_rate": 9.472812785907299e-06, "loss": 0.8359, "step": 7613 }, { "epoch": 0.5967084639498432, "grad_norm": 3.5432119369506836, "learning_rate": 9.472674957211473e-06, "loss": 1.7472, "step": 7614 }, { "epoch": 0.5967868338557993, "grad_norm": 2.4960899353027344, "learning_rate": 9.472537111503887e-06, "loss": 1.4323, "step": 7615 }, { "epoch": 0.5968652037617554, "grad_norm": 1.8895890712738037, "learning_rate": 9.472399248785069e-06, "loss": 1.3459, "step": 7616 }, { "epoch": 0.5969435736677116, "grad_norm": 2.3069841861724854, "learning_rate": 9.47226136905554e-06, "loss": 1.3144, "step": 7617 }, { "epoch": 0.5970219435736677, "grad_norm": 8.566540718078613, "learning_rate": 9.472123472315829e-06, "loss": 1.4285, "step": 7618 }, { "epoch": 0.5971003134796238, "grad_norm": 4.404358386993408, "learning_rate": 9.471985558566454e-06, "loss": 1.6841, "step": 7619 }, { "epoch": 0.5971786833855799, "grad_norm": 4.757326602935791, "learning_rate": 9.471847627807944e-06, "loss": 1.6202, "step": 7620 }, { "epoch": 0.597257053291536, "grad_norm": 2.905043601989746, "learning_rate": 9.471709680040821e-06, "loss": 1.3184, "step": 7621 }, { "epoch": 0.5973354231974921, "grad_norm": 2.704852819442749, "learning_rate": 9.471571715265614e-06, "loss": 1.3451, "step": 7622 }, { "epoch": 0.5974137931034482, "grad_norm": 2.4786758422851562, "learning_rate": 9.471433733482843e-06, "loss": 1.3473, "step": 7623 }, { "epoch": 0.5974921630094044, "grad_norm": 3.2549500465393066, "learning_rate": 9.471295734693036e-06, "loss": 1.2537, "step": 7624 }, { "epoch": 0.5975705329153606, "grad_norm": 2.9823288917541504, "learning_rate": 9.471157718896716e-06, "loss": 1.4082, "step": 7625 }, { "epoch": 0.5976489028213167, "grad_norm": 4.76986837387085, "learning_rate": 9.471019686094407e-06, "loss": 1.4305, "step": 7626 }, { "epoch": 0.5977272727272728, "grad_norm": 4.118976593017578, "learning_rate": 9.470881636286638e-06, "loss": 1.4546, "step": 7627 }, { "epoch": 0.5978056426332289, "grad_norm": 9.806724548339844, "learning_rate": 9.47074356947393e-06, "loss": 0.918, "step": 7628 }, { "epoch": 0.597884012539185, "grad_norm": 3.1083574295043945, "learning_rate": 9.47060548565681e-06, "loss": 1.522, "step": 7629 }, { "epoch": 0.5979623824451411, "grad_norm": 2.861213445663452, "learning_rate": 9.470467384835804e-06, "loss": 1.444, "step": 7630 }, { "epoch": 0.5980407523510972, "grad_norm": 5.545220375061035, "learning_rate": 9.470329267011435e-06, "loss": 1.3356, "step": 7631 }, { "epoch": 0.5981191222570533, "grad_norm": 2.5929293632507324, "learning_rate": 9.470191132184231e-06, "loss": 1.3124, "step": 7632 }, { "epoch": 0.5981974921630094, "grad_norm": 6.6628737449646, "learning_rate": 9.470052980354717e-06, "loss": 1.0817, "step": 7633 }, { "epoch": 0.5982758620689655, "grad_norm": 2.2093913555145264, "learning_rate": 9.469914811523414e-06, "loss": 1.1924, "step": 7634 }, { "epoch": 0.5983542319749217, "grad_norm": 3.4180872440338135, "learning_rate": 9.469776625690853e-06, "loss": 1.9041, "step": 7635 }, { "epoch": 0.5984326018808778, "grad_norm": 2.7159292697906494, "learning_rate": 9.469638422857557e-06, "loss": 1.1856, "step": 7636 }, { "epoch": 0.5985109717868339, "grad_norm": 2.872246026992798, "learning_rate": 9.469500203024053e-06, "loss": 1.4429, "step": 7637 }, { "epoch": 0.59858934169279, "grad_norm": 3.065572738647461, "learning_rate": 9.469361966190863e-06, "loss": 1.6483, "step": 7638 }, { "epoch": 0.5986677115987461, "grad_norm": 2.191699743270874, "learning_rate": 9.469223712358519e-06, "loss": 1.5284, "step": 7639 }, { "epoch": 0.5987460815047022, "grad_norm": 2.094545364379883, "learning_rate": 9.469085441527543e-06, "loss": 1.0904, "step": 7640 }, { "epoch": 0.5988244514106583, "grad_norm": 2.3407816886901855, "learning_rate": 9.46894715369846e-06, "loss": 1.4987, "step": 7641 }, { "epoch": 0.5989028213166144, "grad_norm": 2.157870292663574, "learning_rate": 9.4688088488718e-06, "loss": 1.2998, "step": 7642 }, { "epoch": 0.5989811912225705, "grad_norm": 4.152471542358398, "learning_rate": 9.468670527048084e-06, "loss": 1.4875, "step": 7643 }, { "epoch": 0.5990595611285267, "grad_norm": 2.3929872512817383, "learning_rate": 9.468532188227841e-06, "loss": 1.0684, "step": 7644 }, { "epoch": 0.5991379310344828, "grad_norm": 3.0137064456939697, "learning_rate": 9.468393832411598e-06, "loss": 1.2302, "step": 7645 }, { "epoch": 0.5992163009404389, "grad_norm": 2.7093586921691895, "learning_rate": 9.468255459599878e-06, "loss": 1.3102, "step": 7646 }, { "epoch": 0.599294670846395, "grad_norm": 3.796586751937866, "learning_rate": 9.46811706979321e-06, "loss": 1.7693, "step": 7647 }, { "epoch": 0.5993730407523511, "grad_norm": 3.8422138690948486, "learning_rate": 9.467978662992122e-06, "loss": 1.3608, "step": 7648 }, { "epoch": 0.5994514106583072, "grad_norm": 2.2681140899658203, "learning_rate": 9.467840239197134e-06, "loss": 1.0474, "step": 7649 }, { "epoch": 0.5995297805642633, "grad_norm": 3.3277127742767334, "learning_rate": 9.467701798408779e-06, "loss": 1.545, "step": 7650 }, { "epoch": 0.5996081504702194, "grad_norm": 3.449404716491699, "learning_rate": 9.46756334062758e-06, "loss": 1.4852, "step": 7651 }, { "epoch": 0.5996865203761755, "grad_norm": 2.687037944793701, "learning_rate": 9.467424865854068e-06, "loss": 1.5745, "step": 7652 }, { "epoch": 0.5997648902821316, "grad_norm": 2.8259668350219727, "learning_rate": 9.467286374088762e-06, "loss": 1.2811, "step": 7653 }, { "epoch": 0.5998432601880878, "grad_norm": 3.097677230834961, "learning_rate": 9.467147865332196e-06, "loss": 1.5487, "step": 7654 }, { "epoch": 0.5999216300940439, "grad_norm": 2.3195202350616455, "learning_rate": 9.467009339584894e-06, "loss": 1.2594, "step": 7655 }, { "epoch": 0.6, "grad_norm": 2.2041232585906982, "learning_rate": 9.46687079684738e-06, "loss": 1.3611, "step": 7656 }, { "epoch": 0.6000783699059561, "grad_norm": 3.7042787075042725, "learning_rate": 9.466732237120187e-06, "loss": 1.6323, "step": 7657 }, { "epoch": 0.6001567398119122, "grad_norm": 2.6617681980133057, "learning_rate": 9.466593660403838e-06, "loss": 1.2475, "step": 7658 }, { "epoch": 0.6002351097178683, "grad_norm": 2.9313859939575195, "learning_rate": 9.466455066698859e-06, "loss": 1.3862, "step": 7659 }, { "epoch": 0.6003134796238244, "grad_norm": 2.821213960647583, "learning_rate": 9.466316456005783e-06, "loss": 1.7729, "step": 7660 }, { "epoch": 0.6003918495297805, "grad_norm": 4.896533966064453, "learning_rate": 9.46617782832513e-06, "loss": 1.2425, "step": 7661 }, { "epoch": 0.6004702194357366, "grad_norm": 7.114869594573975, "learning_rate": 9.46603918365743e-06, "loss": 1.3483, "step": 7662 }, { "epoch": 0.6005485893416928, "grad_norm": 3.523191213607788, "learning_rate": 9.465900522003212e-06, "loss": 1.7274, "step": 7663 }, { "epoch": 0.6006269592476489, "grad_norm": 2.100752830505371, "learning_rate": 9.465761843363003e-06, "loss": 1.543, "step": 7664 }, { "epoch": 0.600705329153605, "grad_norm": 2.614954710006714, "learning_rate": 9.465623147737328e-06, "loss": 1.5596, "step": 7665 }, { "epoch": 0.6007836990595611, "grad_norm": 4.3807220458984375, "learning_rate": 9.465484435126718e-06, "loss": 1.6554, "step": 7666 }, { "epoch": 0.6008620689655172, "grad_norm": 2.3172123432159424, "learning_rate": 9.4653457055317e-06, "loss": 1.5988, "step": 7667 }, { "epoch": 0.6009404388714733, "grad_norm": 2.643317222595215, "learning_rate": 9.465206958952797e-06, "loss": 1.6365, "step": 7668 }, { "epoch": 0.6010188087774294, "grad_norm": 2.7438063621520996, "learning_rate": 9.465068195390542e-06, "loss": 1.572, "step": 7669 }, { "epoch": 0.6010971786833855, "grad_norm": 3.3617215156555176, "learning_rate": 9.46492941484546e-06, "loss": 1.2337, "step": 7670 }, { "epoch": 0.6011755485893417, "grad_norm": 2.2494053840637207, "learning_rate": 9.464790617318082e-06, "loss": 1.2118, "step": 7671 }, { "epoch": 0.6012539184952979, "grad_norm": 3.3461291790008545, "learning_rate": 9.464651802808932e-06, "loss": 1.531, "step": 7672 }, { "epoch": 0.601332288401254, "grad_norm": 2.956815242767334, "learning_rate": 9.46451297131854e-06, "loss": 1.0973, "step": 7673 }, { "epoch": 0.6014106583072101, "grad_norm": 14.684349060058594, "learning_rate": 9.464374122847434e-06, "loss": 2.9859, "step": 7674 }, { "epoch": 0.6014890282131662, "grad_norm": 4.4229817390441895, "learning_rate": 9.464235257396142e-06, "loss": 1.7436, "step": 7675 }, { "epoch": 0.6015673981191223, "grad_norm": 1.8906512260437012, "learning_rate": 9.464096374965191e-06, "loss": 1.0172, "step": 7676 }, { "epoch": 0.6016457680250784, "grad_norm": 2.2309534549713135, "learning_rate": 9.463957475555112e-06, "loss": 1.8602, "step": 7677 }, { "epoch": 0.6017241379310345, "grad_norm": 4.596774101257324, "learning_rate": 9.463818559166432e-06, "loss": 1.4155, "step": 7678 }, { "epoch": 0.6018025078369906, "grad_norm": 1.8796011209487915, "learning_rate": 9.463679625799676e-06, "loss": 1.071, "step": 7679 }, { "epoch": 0.6018808777429467, "grad_norm": 2.4406697750091553, "learning_rate": 9.463540675455378e-06, "loss": 1.6553, "step": 7680 }, { "epoch": 0.6019592476489029, "grad_norm": 9.545746803283691, "learning_rate": 9.463401708134063e-06, "loss": 1.3919, "step": 7681 }, { "epoch": 0.602037617554859, "grad_norm": 2.824220657348633, "learning_rate": 9.46326272383626e-06, "loss": 1.532, "step": 7682 }, { "epoch": 0.6021159874608151, "grad_norm": 2.2147581577301025, "learning_rate": 9.4631237225625e-06, "loss": 1.4745, "step": 7683 }, { "epoch": 0.6021943573667712, "grad_norm": 3.4745168685913086, "learning_rate": 9.462984704313307e-06, "loss": 1.7893, "step": 7684 }, { "epoch": 0.6022727272727273, "grad_norm": 3.0044684410095215, "learning_rate": 9.462845669089213e-06, "loss": 1.1493, "step": 7685 }, { "epoch": 0.6023510971786834, "grad_norm": 2.5598886013031006, "learning_rate": 9.462706616890748e-06, "loss": 1.3461, "step": 7686 }, { "epoch": 0.6024294670846395, "grad_norm": 2.4530651569366455, "learning_rate": 9.462567547718437e-06, "loss": 1.664, "step": 7687 }, { "epoch": 0.6025078369905956, "grad_norm": 3.2428863048553467, "learning_rate": 9.462428461572815e-06, "loss": 1.0327, "step": 7688 }, { "epoch": 0.6025862068965517, "grad_norm": 3.685917854309082, "learning_rate": 9.462289358454403e-06, "loss": 1.6448, "step": 7689 }, { "epoch": 0.6026645768025078, "grad_norm": 2.87054443359375, "learning_rate": 9.462150238363737e-06, "loss": 1.5249, "step": 7690 }, { "epoch": 0.602742946708464, "grad_norm": 2.7670209407806396, "learning_rate": 9.462011101301343e-06, "loss": 1.1656, "step": 7691 }, { "epoch": 0.6028213166144201, "grad_norm": 2.0412070751190186, "learning_rate": 9.461871947267749e-06, "loss": 1.6381, "step": 7692 }, { "epoch": 0.6028996865203762, "grad_norm": 2.492558002471924, "learning_rate": 9.461732776263487e-06, "loss": 1.5503, "step": 7693 }, { "epoch": 0.6029780564263323, "grad_norm": 1.9282277822494507, "learning_rate": 9.461593588289085e-06, "loss": 1.4123, "step": 7694 }, { "epoch": 0.6030564263322884, "grad_norm": 3.780561923980713, "learning_rate": 9.461454383345072e-06, "loss": 1.2292, "step": 7695 }, { "epoch": 0.6031347962382445, "grad_norm": 2.7681899070739746, "learning_rate": 9.461315161431979e-06, "loss": 1.6019, "step": 7696 }, { "epoch": 0.6032131661442006, "grad_norm": 2.945770740509033, "learning_rate": 9.461175922550334e-06, "loss": 1.1648, "step": 7697 }, { "epoch": 0.6032915360501567, "grad_norm": 6.638734817504883, "learning_rate": 9.461036666700668e-06, "loss": 1.313, "step": 7698 }, { "epoch": 0.6033699059561128, "grad_norm": 5.314820766448975, "learning_rate": 9.46089739388351e-06, "loss": 1.069, "step": 7699 }, { "epoch": 0.603448275862069, "grad_norm": 2.254215955734253, "learning_rate": 9.460758104099388e-06, "loss": 1.2343, "step": 7700 }, { "epoch": 0.6035266457680251, "grad_norm": 5.592192649841309, "learning_rate": 9.460618797348835e-06, "loss": 2.1384, "step": 7701 }, { "epoch": 0.6036050156739812, "grad_norm": 2.498671293258667, "learning_rate": 9.460479473632379e-06, "loss": 1.4694, "step": 7702 }, { "epoch": 0.6036833855799373, "grad_norm": 2.8081562519073486, "learning_rate": 9.46034013295055e-06, "loss": 1.4932, "step": 7703 }, { "epoch": 0.6037617554858934, "grad_norm": 1.9013264179229736, "learning_rate": 9.460200775303878e-06, "loss": 1.5465, "step": 7704 }, { "epoch": 0.6038401253918495, "grad_norm": 3.073758363723755, "learning_rate": 9.460061400692892e-06, "loss": 1.3036, "step": 7705 }, { "epoch": 0.6039184952978056, "grad_norm": 2.922882318496704, "learning_rate": 9.459922009118124e-06, "loss": 1.4843, "step": 7706 }, { "epoch": 0.6039968652037617, "grad_norm": 2.4854001998901367, "learning_rate": 9.459782600580103e-06, "loss": 1.482, "step": 7707 }, { "epoch": 0.6040752351097178, "grad_norm": 4.276668548583984, "learning_rate": 9.45964317507936e-06, "loss": 1.4238, "step": 7708 }, { "epoch": 0.604153605015674, "grad_norm": 6.006348609924316, "learning_rate": 9.459503732616426e-06, "loss": 1.5088, "step": 7709 }, { "epoch": 0.60423197492163, "grad_norm": 2.475677967071533, "learning_rate": 9.45936427319183e-06, "loss": 0.8093, "step": 7710 }, { "epoch": 0.6043103448275862, "grad_norm": 3.0764241218566895, "learning_rate": 9.459224796806102e-06, "loss": 1.1929, "step": 7711 }, { "epoch": 0.6043887147335423, "grad_norm": 2.012571096420288, "learning_rate": 9.459085303459773e-06, "loss": 1.2176, "step": 7712 }, { "epoch": 0.6044670846394984, "grad_norm": 6.390148162841797, "learning_rate": 9.458945793153376e-06, "loss": 1.2743, "step": 7713 }, { "epoch": 0.6045454545454545, "grad_norm": 2.497741460800171, "learning_rate": 9.458806265887438e-06, "loss": 1.5159, "step": 7714 }, { "epoch": 0.6046238244514106, "grad_norm": 2.383204936981201, "learning_rate": 9.458666721662491e-06, "loss": 1.756, "step": 7715 }, { "epoch": 0.6047021943573667, "grad_norm": 2.1244850158691406, "learning_rate": 9.458527160479067e-06, "loss": 1.5189, "step": 7716 }, { "epoch": 0.6047805642633229, "grad_norm": 2.8779525756835938, "learning_rate": 9.458387582337694e-06, "loss": 1.1325, "step": 7717 }, { "epoch": 0.604858934169279, "grad_norm": 2.027491569519043, "learning_rate": 9.458247987238905e-06, "loss": 1.4893, "step": 7718 }, { "epoch": 0.6049373040752352, "grad_norm": 3.0372490882873535, "learning_rate": 9.45810837518323e-06, "loss": 1.4718, "step": 7719 }, { "epoch": 0.6050156739811913, "grad_norm": 4.700921535491943, "learning_rate": 9.457968746171202e-06, "loss": 1.1516, "step": 7720 }, { "epoch": 0.6050940438871474, "grad_norm": 4.413011074066162, "learning_rate": 9.457829100203351e-06, "loss": 1.8234, "step": 7721 }, { "epoch": 0.6051724137931035, "grad_norm": 4.4269700050354, "learning_rate": 9.457689437280206e-06, "loss": 1.7258, "step": 7722 }, { "epoch": 0.6052507836990596, "grad_norm": 3.6852941513061523, "learning_rate": 9.457549757402302e-06, "loss": 1.2864, "step": 7723 }, { "epoch": 0.6053291536050157, "grad_norm": 1.567409634590149, "learning_rate": 9.457410060570167e-06, "loss": 0.9499, "step": 7724 }, { "epoch": 0.6054075235109718, "grad_norm": 2.176527976989746, "learning_rate": 9.457270346784335e-06, "loss": 1.5359, "step": 7725 }, { "epoch": 0.6054858934169279, "grad_norm": 3.17619252204895, "learning_rate": 9.457130616045333e-06, "loss": 1.3006, "step": 7726 }, { "epoch": 0.605564263322884, "grad_norm": 2.370880365371704, "learning_rate": 9.456990868353699e-06, "loss": 1.5273, "step": 7727 }, { "epoch": 0.6056426332288402, "grad_norm": 3.4679880142211914, "learning_rate": 9.456851103709959e-06, "loss": 1.5099, "step": 7728 }, { "epoch": 0.6057210031347963, "grad_norm": 3.6937241554260254, "learning_rate": 9.456711322114647e-06, "loss": 1.4335, "step": 7729 }, { "epoch": 0.6057993730407524, "grad_norm": 2.541975975036621, "learning_rate": 9.456571523568293e-06, "loss": 1.4782, "step": 7730 }, { "epoch": 0.6058777429467085, "grad_norm": 1.9575332403182983, "learning_rate": 9.456431708071432e-06, "loss": 1.1608, "step": 7731 }, { "epoch": 0.6059561128526646, "grad_norm": 3.5360968112945557, "learning_rate": 9.456291875624592e-06, "loss": 1.6938, "step": 7732 }, { "epoch": 0.6060344827586207, "grad_norm": 3.195671558380127, "learning_rate": 9.456152026228307e-06, "loss": 1.4441, "step": 7733 }, { "epoch": 0.6061128526645768, "grad_norm": 2.444837808609009, "learning_rate": 9.45601215988311e-06, "loss": 1.6076, "step": 7734 }, { "epoch": 0.6061912225705329, "grad_norm": 7.120833873748779, "learning_rate": 9.455872276589531e-06, "loss": 1.0004, "step": 7735 }, { "epoch": 0.606269592476489, "grad_norm": 7.513619422912598, "learning_rate": 9.455732376348101e-06, "loss": 1.7266, "step": 7736 }, { "epoch": 0.6063479623824451, "grad_norm": 3.573802947998047, "learning_rate": 9.455592459159355e-06, "loss": 1.2799, "step": 7737 }, { "epoch": 0.6064263322884013, "grad_norm": 2.703625440597534, "learning_rate": 9.455452525023823e-06, "loss": 1.5523, "step": 7738 }, { "epoch": 0.6065047021943574, "grad_norm": 2.8807787895202637, "learning_rate": 9.45531257394204e-06, "loss": 1.7226, "step": 7739 }, { "epoch": 0.6065830721003135, "grad_norm": 3.391810655593872, "learning_rate": 9.455172605914533e-06, "loss": 1.4198, "step": 7740 }, { "epoch": 0.6066614420062696, "grad_norm": 2.802522897720337, "learning_rate": 9.45503262094184e-06, "loss": 1.6915, "step": 7741 }, { "epoch": 0.6067398119122257, "grad_norm": 1.8988372087478638, "learning_rate": 9.454892619024491e-06, "loss": 1.4714, "step": 7742 }, { "epoch": 0.6068181818181818, "grad_norm": 2.273338556289673, "learning_rate": 9.454752600163018e-06, "loss": 1.088, "step": 7743 }, { "epoch": 0.6068965517241379, "grad_norm": 2.3376126289367676, "learning_rate": 9.454612564357955e-06, "loss": 1.3167, "step": 7744 }, { "epoch": 0.606974921630094, "grad_norm": 2.3430838584899902, "learning_rate": 9.454472511609834e-06, "loss": 1.4879, "step": 7745 }, { "epoch": 0.6070532915360501, "grad_norm": 2.918407440185547, "learning_rate": 9.454332441919185e-06, "loss": 1.1592, "step": 7746 }, { "epoch": 0.6071316614420063, "grad_norm": 3.357489585876465, "learning_rate": 9.454192355286547e-06, "loss": 1.3012, "step": 7747 }, { "epoch": 0.6072100313479624, "grad_norm": 2.7207515239715576, "learning_rate": 9.454052251712447e-06, "loss": 1.4792, "step": 7748 }, { "epoch": 0.6072884012539185, "grad_norm": 3.440237522125244, "learning_rate": 9.45391213119742e-06, "loss": 1.4494, "step": 7749 }, { "epoch": 0.6073667711598746, "grad_norm": 4.095073223114014, "learning_rate": 9.453771993742e-06, "loss": 1.0498, "step": 7750 }, { "epoch": 0.6074451410658307, "grad_norm": 9.539093971252441, "learning_rate": 9.453631839346718e-06, "loss": 1.8794, "step": 7751 }, { "epoch": 0.6075235109717868, "grad_norm": 12.31799602508545, "learning_rate": 9.453491668012108e-06, "loss": 0.9531, "step": 7752 }, { "epoch": 0.6076018808777429, "grad_norm": 1.8792400360107422, "learning_rate": 9.453351479738703e-06, "loss": 1.3526, "step": 7753 }, { "epoch": 0.607680250783699, "grad_norm": 3.3013176918029785, "learning_rate": 9.453211274527037e-06, "loss": 1.4025, "step": 7754 }, { "epoch": 0.6077586206896551, "grad_norm": 2.4981915950775146, "learning_rate": 9.453071052377643e-06, "loss": 1.6967, "step": 7755 }, { "epoch": 0.6078369905956112, "grad_norm": 3.4050750732421875, "learning_rate": 9.452930813291052e-06, "loss": 1.1096, "step": 7756 }, { "epoch": 0.6079153605015674, "grad_norm": 2.977757215499878, "learning_rate": 9.452790557267802e-06, "loss": 1.3221, "step": 7757 }, { "epoch": 0.6079937304075235, "grad_norm": 2.3548693656921387, "learning_rate": 9.452650284308421e-06, "loss": 1.1771, "step": 7758 }, { "epoch": 0.6080721003134796, "grad_norm": 5.876055717468262, "learning_rate": 9.452509994413447e-06, "loss": 1.5168, "step": 7759 }, { "epoch": 0.6081504702194357, "grad_norm": 2.800309658050537, "learning_rate": 9.452369687583413e-06, "loss": 1.4477, "step": 7760 }, { "epoch": 0.6082288401253918, "grad_norm": 2.8125710487365723, "learning_rate": 9.45222936381885e-06, "loss": 0.9171, "step": 7761 }, { "epoch": 0.6083072100313479, "grad_norm": 2.9031267166137695, "learning_rate": 9.452089023120295e-06, "loss": 1.2868, "step": 7762 }, { "epoch": 0.608385579937304, "grad_norm": 2.772925853729248, "learning_rate": 9.451948665488277e-06, "loss": 1.453, "step": 7763 }, { "epoch": 0.6084639498432602, "grad_norm": 3.428412437438965, "learning_rate": 9.451808290923334e-06, "loss": 1.5447, "step": 7764 }, { "epoch": 0.6085423197492164, "grad_norm": 1.943709135055542, "learning_rate": 9.451667899426e-06, "loss": 1.1765, "step": 7765 }, { "epoch": 0.6086206896551725, "grad_norm": 4.525723934173584, "learning_rate": 9.451527490996808e-06, "loss": 1.5503, "step": 7766 }, { "epoch": 0.6086990595611286, "grad_norm": 2.9144904613494873, "learning_rate": 9.451387065636292e-06, "loss": 1.4844, "step": 7767 }, { "epoch": 0.6087774294670847, "grad_norm": 4.817169189453125, "learning_rate": 9.451246623344984e-06, "loss": 1.2646, "step": 7768 }, { "epoch": 0.6088557993730408, "grad_norm": 2.5418527126312256, "learning_rate": 9.451106164123423e-06, "loss": 1.4889, "step": 7769 }, { "epoch": 0.6089341692789969, "grad_norm": 2.2652668952941895, "learning_rate": 9.450965687972138e-06, "loss": 1.3943, "step": 7770 }, { "epoch": 0.609012539184953, "grad_norm": 6.415167331695557, "learning_rate": 9.450825194891666e-06, "loss": 1.5606, "step": 7771 }, { "epoch": 0.6090909090909091, "grad_norm": 2.89557147026062, "learning_rate": 9.450684684882542e-06, "loss": 1.7602, "step": 7772 }, { "epoch": 0.6091692789968652, "grad_norm": 2.32389497756958, "learning_rate": 9.450544157945298e-06, "loss": 1.4905, "step": 7773 }, { "epoch": 0.6092476489028213, "grad_norm": 3.2233874797821045, "learning_rate": 9.450403614080472e-06, "loss": 1.5855, "step": 7774 }, { "epoch": 0.6093260188087775, "grad_norm": 6.1515326499938965, "learning_rate": 9.450263053288595e-06, "loss": 1.5953, "step": 7775 }, { "epoch": 0.6094043887147336, "grad_norm": 2.6639726161956787, "learning_rate": 9.450122475570205e-06, "loss": 1.0214, "step": 7776 }, { "epoch": 0.6094827586206897, "grad_norm": 5.724981784820557, "learning_rate": 9.449981880925833e-06, "loss": 1.2383, "step": 7777 }, { "epoch": 0.6095611285266458, "grad_norm": 1.86972975730896, "learning_rate": 9.449841269356014e-06, "loss": 1.3772, "step": 7778 }, { "epoch": 0.6096394984326019, "grad_norm": 2.2258148193359375, "learning_rate": 9.449700640861288e-06, "loss": 1.5022, "step": 7779 }, { "epoch": 0.609717868338558, "grad_norm": 12.71562671661377, "learning_rate": 9.449559995442184e-06, "loss": 1.1936, "step": 7780 }, { "epoch": 0.6097962382445141, "grad_norm": 2.3842902183532715, "learning_rate": 9.449419333099242e-06, "loss": 1.5777, "step": 7781 }, { "epoch": 0.6098746081504702, "grad_norm": 4.363988876342773, "learning_rate": 9.44927865383299e-06, "loss": 1.1416, "step": 7782 }, { "epoch": 0.6099529780564263, "grad_norm": 2.595519781112671, "learning_rate": 9.44913795764397e-06, "loss": 1.2319, "step": 7783 }, { "epoch": 0.6100313479623825, "grad_norm": 12.554612159729004, "learning_rate": 9.448997244532714e-06, "loss": 0.9332, "step": 7784 }, { "epoch": 0.6101097178683386, "grad_norm": 2.51471209526062, "learning_rate": 9.448856514499759e-06, "loss": 1.3518, "step": 7785 }, { "epoch": 0.6101880877742947, "grad_norm": 1.9165476560592651, "learning_rate": 9.448715767545635e-06, "loss": 1.1601, "step": 7786 }, { "epoch": 0.6102664576802508, "grad_norm": 2.7806332111358643, "learning_rate": 9.448575003670885e-06, "loss": 1.5207, "step": 7787 }, { "epoch": 0.6103448275862069, "grad_norm": 2.6947014331817627, "learning_rate": 9.44843422287604e-06, "loss": 1.3418, "step": 7788 }, { "epoch": 0.610423197492163, "grad_norm": 2.7622478008270264, "learning_rate": 9.448293425161635e-06, "loss": 1.5662, "step": 7789 }, { "epoch": 0.6105015673981191, "grad_norm": 2.483208179473877, "learning_rate": 9.448152610528207e-06, "loss": 1.3088, "step": 7790 }, { "epoch": 0.6105799373040752, "grad_norm": 2.5130460262298584, "learning_rate": 9.44801177897629e-06, "loss": 1.3805, "step": 7791 }, { "epoch": 0.6106583072100313, "grad_norm": 3.225355386734009, "learning_rate": 9.447870930506422e-06, "loss": 0.9833, "step": 7792 }, { "epoch": 0.6107366771159874, "grad_norm": 3.576242446899414, "learning_rate": 9.447730065119139e-06, "loss": 1.6752, "step": 7793 }, { "epoch": 0.6108150470219436, "grad_norm": 2.1801884174346924, "learning_rate": 9.447589182814973e-06, "loss": 1.0395, "step": 7794 }, { "epoch": 0.6108934169278997, "grad_norm": 4.809851169586182, "learning_rate": 9.447448283594462e-06, "loss": 1.0828, "step": 7795 }, { "epoch": 0.6109717868338558, "grad_norm": 3.610158681869507, "learning_rate": 9.447307367458143e-06, "loss": 1.5781, "step": 7796 }, { "epoch": 0.6110501567398119, "grad_norm": 4.514618873596191, "learning_rate": 9.447166434406551e-06, "loss": 1.2531, "step": 7797 }, { "epoch": 0.611128526645768, "grad_norm": 3.7673532962799072, "learning_rate": 9.447025484440223e-06, "loss": 1.4256, "step": 7798 }, { "epoch": 0.6112068965517241, "grad_norm": 2.8834381103515625, "learning_rate": 9.446884517559693e-06, "loss": 1.4634, "step": 7799 }, { "epoch": 0.6112852664576802, "grad_norm": 4.572225570678711, "learning_rate": 9.4467435337655e-06, "loss": 1.2443, "step": 7800 }, { "epoch": 0.6113636363636363, "grad_norm": 6.683688640594482, "learning_rate": 9.446602533058175e-06, "loss": 1.8, "step": 7801 }, { "epoch": 0.6114420062695924, "grad_norm": 4.690990924835205, "learning_rate": 9.446461515438261e-06, "loss": 1.3134, "step": 7802 }, { "epoch": 0.6115203761755486, "grad_norm": 4.065093517303467, "learning_rate": 9.44632048090629e-06, "loss": 1.5839, "step": 7803 }, { "epoch": 0.6115987460815047, "grad_norm": 3.7494544982910156, "learning_rate": 9.446179429462801e-06, "loss": 1.4382, "step": 7804 }, { "epoch": 0.6116771159874608, "grad_norm": 2.7731597423553467, "learning_rate": 9.446038361108329e-06, "loss": 1.1645, "step": 7805 }, { "epoch": 0.6117554858934169, "grad_norm": 3.158355712890625, "learning_rate": 9.44589727584341e-06, "loss": 1.2093, "step": 7806 }, { "epoch": 0.611833855799373, "grad_norm": 3.575090169906616, "learning_rate": 9.445756173668581e-06, "loss": 1.9872, "step": 7807 }, { "epoch": 0.6119122257053291, "grad_norm": 4.250589370727539, "learning_rate": 9.44561505458438e-06, "loss": 1.2119, "step": 7808 }, { "epoch": 0.6119905956112852, "grad_norm": 3.860830545425415, "learning_rate": 9.445473918591342e-06, "loss": 1.5887, "step": 7809 }, { "epoch": 0.6120689655172413, "grad_norm": 2.5319957733154297, "learning_rate": 9.445332765690003e-06, "loss": 1.6828, "step": 7810 }, { "epoch": 0.6121473354231975, "grad_norm": 2.90828275680542, "learning_rate": 9.445191595880905e-06, "loss": 1.0871, "step": 7811 }, { "epoch": 0.6122257053291537, "grad_norm": 7.164095401763916, "learning_rate": 9.445050409164578e-06, "loss": 1.6627, "step": 7812 }, { "epoch": 0.6123040752351098, "grad_norm": 2.7476024627685547, "learning_rate": 9.444909205541563e-06, "loss": 1.2115, "step": 7813 }, { "epoch": 0.6123824451410659, "grad_norm": 3.7910797595977783, "learning_rate": 9.444767985012398e-06, "loss": 1.2542, "step": 7814 }, { "epoch": 0.612460815047022, "grad_norm": 3.3306143283843994, "learning_rate": 9.444626747577619e-06, "loss": 1.4379, "step": 7815 }, { "epoch": 0.6125391849529781, "grad_norm": 2.1484375, "learning_rate": 9.444485493237761e-06, "loss": 1.4091, "step": 7816 }, { "epoch": 0.6126175548589342, "grad_norm": 3.765958309173584, "learning_rate": 9.444344221993364e-06, "loss": 1.5087, "step": 7817 }, { "epoch": 0.6126959247648903, "grad_norm": 3.6369714736938477, "learning_rate": 9.444202933844963e-06, "loss": 1.2403, "step": 7818 }, { "epoch": 0.6127742946708464, "grad_norm": 2.5591790676116943, "learning_rate": 9.444061628793097e-06, "loss": 1.3139, "step": 7819 }, { "epoch": 0.6128526645768025, "grad_norm": 2.912715435028076, "learning_rate": 9.443920306838304e-06, "loss": 1.4533, "step": 7820 }, { "epoch": 0.6129310344827587, "grad_norm": 3.4066951274871826, "learning_rate": 9.443778967981121e-06, "loss": 0.9773, "step": 7821 }, { "epoch": 0.6130094043887148, "grad_norm": 6.0212507247924805, "learning_rate": 9.443637612222085e-06, "loss": 1.7277, "step": 7822 }, { "epoch": 0.6130877742946709, "grad_norm": 2.2964868545532227, "learning_rate": 9.443496239561734e-06, "loss": 1.4108, "step": 7823 }, { "epoch": 0.613166144200627, "grad_norm": 2.5083813667297363, "learning_rate": 9.443354850000606e-06, "loss": 1.2577, "step": 7824 }, { "epoch": 0.6132445141065831, "grad_norm": 5.107856750488281, "learning_rate": 9.443213443539237e-06, "loss": 1.2956, "step": 7825 }, { "epoch": 0.6133228840125392, "grad_norm": 3.7951183319091797, "learning_rate": 9.443072020178167e-06, "loss": 1.2563, "step": 7826 }, { "epoch": 0.6134012539184953, "grad_norm": 4.193093299865723, "learning_rate": 9.442930579917932e-06, "loss": 1.8798, "step": 7827 }, { "epoch": 0.6134796238244514, "grad_norm": 2.3372693061828613, "learning_rate": 9.442789122759073e-06, "loss": 1.6155, "step": 7828 }, { "epoch": 0.6135579937304075, "grad_norm": 2.3999485969543457, "learning_rate": 9.442647648702126e-06, "loss": 0.5789, "step": 7829 }, { "epoch": 0.6136363636363636, "grad_norm": 6.78156852722168, "learning_rate": 9.442506157747628e-06, "loss": 1.4103, "step": 7830 }, { "epoch": 0.6137147335423198, "grad_norm": 2.9323172569274902, "learning_rate": 9.442364649896118e-06, "loss": 1.2172, "step": 7831 }, { "epoch": 0.6137931034482759, "grad_norm": 4.957767486572266, "learning_rate": 9.442223125148136e-06, "loss": 1.6566, "step": 7832 }, { "epoch": 0.613871473354232, "grad_norm": 2.728311538696289, "learning_rate": 9.44208158350422e-06, "loss": 1.5293, "step": 7833 }, { "epoch": 0.6139498432601881, "grad_norm": 1.9929980039596558, "learning_rate": 9.441940024964903e-06, "loss": 1.2202, "step": 7834 }, { "epoch": 0.6140282131661442, "grad_norm": 2.25461745262146, "learning_rate": 9.441798449530732e-06, "loss": 1.5284, "step": 7835 }, { "epoch": 0.6141065830721003, "grad_norm": 1.989447832107544, "learning_rate": 9.44165685720224e-06, "loss": 1.2792, "step": 7836 }, { "epoch": 0.6141849529780564, "grad_norm": 3.776790142059326, "learning_rate": 9.441515247979966e-06, "loss": 1.1427, "step": 7837 }, { "epoch": 0.6142633228840125, "grad_norm": 3.224339008331299, "learning_rate": 9.44137362186445e-06, "loss": 1.4024, "step": 7838 }, { "epoch": 0.6143416927899686, "grad_norm": 3.192636013031006, "learning_rate": 9.44123197885623e-06, "loss": 1.6458, "step": 7839 }, { "epoch": 0.6144200626959248, "grad_norm": 2.80249285697937, "learning_rate": 9.441090318955843e-06, "loss": 1.1793, "step": 7840 }, { "epoch": 0.6144984326018809, "grad_norm": 4.602345943450928, "learning_rate": 9.44094864216383e-06, "loss": 1.3116, "step": 7841 }, { "epoch": 0.614576802507837, "grad_norm": 2.5889997482299805, "learning_rate": 9.44080694848073e-06, "loss": 1.4, "step": 7842 }, { "epoch": 0.6146551724137931, "grad_norm": 3.378385543823242, "learning_rate": 9.440665237907083e-06, "loss": 2.1256, "step": 7843 }, { "epoch": 0.6147335423197492, "grad_norm": 4.355015277862549, "learning_rate": 9.440523510443425e-06, "loss": 1.9257, "step": 7844 }, { "epoch": 0.6148119122257053, "grad_norm": 2.0235555171966553, "learning_rate": 9.440381766090296e-06, "loss": 1.3671, "step": 7845 }, { "epoch": 0.6148902821316614, "grad_norm": 3.336674213409424, "learning_rate": 9.440240004848236e-06, "loss": 1.5355, "step": 7846 }, { "epoch": 0.6149686520376175, "grad_norm": 7.195450782775879, "learning_rate": 9.440098226717781e-06, "loss": 1.0638, "step": 7847 }, { "epoch": 0.6150470219435736, "grad_norm": 7.49740743637085, "learning_rate": 9.439956431699477e-06, "loss": 1.075, "step": 7848 }, { "epoch": 0.6151253918495297, "grad_norm": 3.128491163253784, "learning_rate": 9.439814619793857e-06, "loss": 1.487, "step": 7849 }, { "epoch": 0.6152037617554859, "grad_norm": 2.292889356613159, "learning_rate": 9.439672791001463e-06, "loss": 1.1625, "step": 7850 }, { "epoch": 0.615282131661442, "grad_norm": 3.120918035507202, "learning_rate": 9.439530945322834e-06, "loss": 1.7741, "step": 7851 }, { "epoch": 0.6153605015673981, "grad_norm": 3.377450942993164, "learning_rate": 9.43938908275851e-06, "loss": 1.7807, "step": 7852 }, { "epoch": 0.6154388714733542, "grad_norm": 7.4548187255859375, "learning_rate": 9.43924720330903e-06, "loss": 1.7731, "step": 7853 }, { "epoch": 0.6155172413793103, "grad_norm": 2.1842081546783447, "learning_rate": 9.439105306974934e-06, "loss": 1.3073, "step": 7854 }, { "epoch": 0.6155956112852664, "grad_norm": 4.078190803527832, "learning_rate": 9.43896339375676e-06, "loss": 0.8185, "step": 7855 }, { "epoch": 0.6156739811912225, "grad_norm": 3.1979219913482666, "learning_rate": 9.43882146365505e-06, "loss": 1.776, "step": 7856 }, { "epoch": 0.6157523510971786, "grad_norm": 3.2169511318206787, "learning_rate": 9.438679516670344e-06, "loss": 1.6305, "step": 7857 }, { "epoch": 0.6158307210031349, "grad_norm": 3.6761398315429688, "learning_rate": 9.43853755280318e-06, "loss": 1.2181, "step": 7858 }, { "epoch": 0.615909090909091, "grad_norm": 2.8856937885284424, "learning_rate": 9.4383955720541e-06, "loss": 1.3034, "step": 7859 }, { "epoch": 0.6159874608150471, "grad_norm": 2.8379902839660645, "learning_rate": 9.438253574423642e-06, "loss": 1.3562, "step": 7860 }, { "epoch": 0.6160658307210032, "grad_norm": 2.7278337478637695, "learning_rate": 9.438111559912348e-06, "loss": 1.4424, "step": 7861 }, { "epoch": 0.6161442006269593, "grad_norm": 7.083784103393555, "learning_rate": 9.437969528520756e-06, "loss": 1.4196, "step": 7862 }, { "epoch": 0.6162225705329154, "grad_norm": 5.664133071899414, "learning_rate": 9.437827480249409e-06, "loss": 1.3677, "step": 7863 }, { "epoch": 0.6163009404388715, "grad_norm": 2.531946897506714, "learning_rate": 9.437685415098843e-06, "loss": 1.7378, "step": 7864 }, { "epoch": 0.6163793103448276, "grad_norm": 2.1542842388153076, "learning_rate": 9.437543333069603e-06, "loss": 1.0797, "step": 7865 }, { "epoch": 0.6164576802507837, "grad_norm": 6.7809600830078125, "learning_rate": 9.437401234162226e-06, "loss": 1.0631, "step": 7866 }, { "epoch": 0.6165360501567398, "grad_norm": 3.67140793800354, "learning_rate": 9.437259118377255e-06, "loss": 1.5518, "step": 7867 }, { "epoch": 0.616614420062696, "grad_norm": 3.5848777294158936, "learning_rate": 9.43711698571523e-06, "loss": 1.6683, "step": 7868 }, { "epoch": 0.6166927899686521, "grad_norm": 2.9463698863983154, "learning_rate": 9.43697483617669e-06, "loss": 1.2164, "step": 7869 }, { "epoch": 0.6167711598746082, "grad_norm": 2.18778657913208, "learning_rate": 9.436832669762177e-06, "loss": 1.1988, "step": 7870 }, { "epoch": 0.6168495297805643, "grad_norm": 3.2313034534454346, "learning_rate": 9.436690486472231e-06, "loss": 1.1577, "step": 7871 }, { "epoch": 0.6169278996865204, "grad_norm": 2.019585132598877, "learning_rate": 9.436548286307394e-06, "loss": 1.0436, "step": 7872 }, { "epoch": 0.6170062695924765, "grad_norm": 11.78153133392334, "learning_rate": 9.436406069268206e-06, "loss": 1.208, "step": 7873 }, { "epoch": 0.6170846394984326, "grad_norm": 4.407101154327393, "learning_rate": 9.436263835355207e-06, "loss": 1.7033, "step": 7874 }, { "epoch": 0.6171630094043887, "grad_norm": 4.438652515411377, "learning_rate": 9.43612158456894e-06, "loss": 2.308, "step": 7875 }, { "epoch": 0.6172413793103448, "grad_norm": 2.916707992553711, "learning_rate": 9.435979316909945e-06, "loss": 1.363, "step": 7876 }, { "epoch": 0.617319749216301, "grad_norm": 2.54927921295166, "learning_rate": 9.435837032378764e-06, "loss": 1.4694, "step": 7877 }, { "epoch": 0.6173981191222571, "grad_norm": 2.597604274749756, "learning_rate": 9.435694730975936e-06, "loss": 1.388, "step": 7878 }, { "epoch": 0.6174764890282132, "grad_norm": 4.201594352722168, "learning_rate": 9.435552412702004e-06, "loss": 1.1241, "step": 7879 }, { "epoch": 0.6175548589341693, "grad_norm": 2.4753060340881348, "learning_rate": 9.435410077557508e-06, "loss": 1.318, "step": 7880 }, { "epoch": 0.6176332288401254, "grad_norm": 2.8658642768859863, "learning_rate": 9.435267725542992e-06, "loss": 0.9783, "step": 7881 }, { "epoch": 0.6177115987460815, "grad_norm": 2.4452736377716064, "learning_rate": 9.435125356658995e-06, "loss": 1.2561, "step": 7882 }, { "epoch": 0.6177899686520376, "grad_norm": 2.5348527431488037, "learning_rate": 9.434982970906057e-06, "loss": 1.6624, "step": 7883 }, { "epoch": 0.6178683385579937, "grad_norm": 3.925887107849121, "learning_rate": 9.434840568284724e-06, "loss": 2.0756, "step": 7884 }, { "epoch": 0.6179467084639498, "grad_norm": 3.6740496158599854, "learning_rate": 9.434698148795534e-06, "loss": 0.8482, "step": 7885 }, { "epoch": 0.618025078369906, "grad_norm": 3.6040444374084473, "learning_rate": 9.434555712439032e-06, "loss": 1.3896, "step": 7886 }, { "epoch": 0.618103448275862, "grad_norm": 2.561084747314453, "learning_rate": 9.434413259215756e-06, "loss": 1.2206, "step": 7887 }, { "epoch": 0.6181818181818182, "grad_norm": 2.10255765914917, "learning_rate": 9.434270789126251e-06, "loss": 1.4798, "step": 7888 }, { "epoch": 0.6182601880877743, "grad_norm": 3.0057783126831055, "learning_rate": 9.434128302171057e-06, "loss": 1.0169, "step": 7889 }, { "epoch": 0.6183385579937304, "grad_norm": 3.863812208175659, "learning_rate": 9.433985798350715e-06, "loss": 1.8075, "step": 7890 }, { "epoch": 0.6184169278996865, "grad_norm": 2.413257598876953, "learning_rate": 9.43384327766577e-06, "loss": 1.4183, "step": 7891 }, { "epoch": 0.6184952978056426, "grad_norm": 2.0576460361480713, "learning_rate": 9.433700740116763e-06, "loss": 1.3633, "step": 7892 }, { "epoch": 0.6185736677115987, "grad_norm": 2.545194625854492, "learning_rate": 9.433558185704234e-06, "loss": 1.56, "step": 7893 }, { "epoch": 0.6186520376175548, "grad_norm": 6.070592403411865, "learning_rate": 9.43341561442873e-06, "loss": 1.8657, "step": 7894 }, { "epoch": 0.6187304075235109, "grad_norm": 3.32141375541687, "learning_rate": 9.433273026290787e-06, "loss": 1.4889, "step": 7895 }, { "epoch": 0.618808777429467, "grad_norm": 9.510733604431152, "learning_rate": 9.43313042129095e-06, "loss": 1.819, "step": 7896 }, { "epoch": 0.6188871473354232, "grad_norm": 2.6164917945861816, "learning_rate": 9.432987799429765e-06, "loss": 1.3166, "step": 7897 }, { "epoch": 0.6189655172413793, "grad_norm": 8.494526863098145, "learning_rate": 9.432845160707768e-06, "loss": 1.2223, "step": 7898 }, { "epoch": 0.6190438871473354, "grad_norm": 3.2695729732513428, "learning_rate": 9.432702505125507e-06, "loss": 1.2033, "step": 7899 }, { "epoch": 0.6191222570532915, "grad_norm": 6.708912372589111, "learning_rate": 9.432559832683523e-06, "loss": 1.4242, "step": 7900 }, { "epoch": 0.6192006269592476, "grad_norm": 2.0694937705993652, "learning_rate": 9.432417143382357e-06, "loss": 1.4492, "step": 7901 }, { "epoch": 0.6192789968652037, "grad_norm": 2.538027048110962, "learning_rate": 9.432274437222553e-06, "loss": 1.2875, "step": 7902 }, { "epoch": 0.6193573667711598, "grad_norm": 2.6431307792663574, "learning_rate": 9.432131714204653e-06, "loss": 1.2817, "step": 7903 }, { "epoch": 0.6194357366771159, "grad_norm": 3.0482118129730225, "learning_rate": 9.4319889743292e-06, "loss": 1.2023, "step": 7904 }, { "epoch": 0.6195141065830722, "grad_norm": 2.15883731842041, "learning_rate": 9.43184621759674e-06, "loss": 1.4769, "step": 7905 }, { "epoch": 0.6195924764890283, "grad_norm": 3.182595729827881, "learning_rate": 9.431703444007811e-06, "loss": 1.3191, "step": 7906 }, { "epoch": 0.6196708463949844, "grad_norm": 1.9094815254211426, "learning_rate": 9.431560653562958e-06, "loss": 1.2519, "step": 7907 }, { "epoch": 0.6197492163009405, "grad_norm": 3.2571961879730225, "learning_rate": 9.431417846262727e-06, "loss": 1.4061, "step": 7908 }, { "epoch": 0.6198275862068966, "grad_norm": 2.0652968883514404, "learning_rate": 9.431275022107657e-06, "loss": 1.1976, "step": 7909 }, { "epoch": 0.6199059561128527, "grad_norm": 2.623652696609497, "learning_rate": 9.431132181098293e-06, "loss": 1.3429, "step": 7910 }, { "epoch": 0.6199843260188088, "grad_norm": 3.0118532180786133, "learning_rate": 9.430989323235177e-06, "loss": 1.439, "step": 7911 }, { "epoch": 0.6200626959247649, "grad_norm": 2.2881767749786377, "learning_rate": 9.430846448518855e-06, "loss": 1.3588, "step": 7912 }, { "epoch": 0.620141065830721, "grad_norm": 3.817732572555542, "learning_rate": 9.430703556949868e-06, "loss": 1.5218, "step": 7913 }, { "epoch": 0.6202194357366771, "grad_norm": 6.3353471755981445, "learning_rate": 9.430560648528762e-06, "loss": 1.4706, "step": 7914 }, { "epoch": 0.6202978056426333, "grad_norm": 3.509847402572632, "learning_rate": 9.430417723256077e-06, "loss": 0.8818, "step": 7915 }, { "epoch": 0.6203761755485894, "grad_norm": 2.1753146648406982, "learning_rate": 9.43027478113236e-06, "loss": 1.3103, "step": 7916 }, { "epoch": 0.6204545454545455, "grad_norm": 2.7509944438934326, "learning_rate": 9.430131822158152e-06, "loss": 1.4985, "step": 7917 }, { "epoch": 0.6205329153605016, "grad_norm": 3.3398194313049316, "learning_rate": 9.429988846334e-06, "loss": 1.1369, "step": 7918 }, { "epoch": 0.6206112852664577, "grad_norm": 3.7419724464416504, "learning_rate": 9.429845853660444e-06, "loss": 1.4466, "step": 7919 }, { "epoch": 0.6206896551724138, "grad_norm": 3.6359615325927734, "learning_rate": 9.429702844138032e-06, "loss": 1.5438, "step": 7920 }, { "epoch": 0.6207680250783699, "grad_norm": 7.842044830322266, "learning_rate": 9.429559817767301e-06, "loss": 1.5826, "step": 7921 }, { "epoch": 0.620846394984326, "grad_norm": 2.259578227996826, "learning_rate": 9.429416774548804e-06, "loss": 1.4081, "step": 7922 }, { "epoch": 0.6209247648902821, "grad_norm": 7.171797752380371, "learning_rate": 9.42927371448308e-06, "loss": 1.1776, "step": 7923 }, { "epoch": 0.6210031347962383, "grad_norm": 5.080764293670654, "learning_rate": 9.429130637570672e-06, "loss": 1.1936, "step": 7924 }, { "epoch": 0.6210815047021944, "grad_norm": 4.880819797515869, "learning_rate": 9.428987543812127e-06, "loss": 1.8897, "step": 7925 }, { "epoch": 0.6211598746081505, "grad_norm": 3.00590181350708, "learning_rate": 9.428844433207988e-06, "loss": 1.5651, "step": 7926 }, { "epoch": 0.6212382445141066, "grad_norm": 3.9777028560638428, "learning_rate": 9.4287013057588e-06, "loss": 1.9277, "step": 7927 }, { "epoch": 0.6213166144200627, "grad_norm": 3.7537894248962402, "learning_rate": 9.428558161465107e-06, "loss": 0.863, "step": 7928 }, { "epoch": 0.6213949843260188, "grad_norm": 3.6453139781951904, "learning_rate": 9.428415000327453e-06, "loss": 0.9128, "step": 7929 }, { "epoch": 0.6214733542319749, "grad_norm": 3.3693885803222656, "learning_rate": 9.428271822346384e-06, "loss": 1.6107, "step": 7930 }, { "epoch": 0.621551724137931, "grad_norm": 5.115631580352783, "learning_rate": 9.428128627522443e-06, "loss": 1.5573, "step": 7931 }, { "epoch": 0.6216300940438871, "grad_norm": 2.634272336959839, "learning_rate": 9.427985415856173e-06, "loss": 1.4429, "step": 7932 }, { "epoch": 0.6217084639498432, "grad_norm": 2.775432825088501, "learning_rate": 9.427842187348124e-06, "loss": 1.6719, "step": 7933 }, { "epoch": 0.6217868338557994, "grad_norm": 2.9234514236450195, "learning_rate": 9.427698941998836e-06, "loss": 1.0003, "step": 7934 }, { "epoch": 0.6218652037617555, "grad_norm": 2.3321523666381836, "learning_rate": 9.427555679808855e-06, "loss": 1.5478, "step": 7935 }, { "epoch": 0.6219435736677116, "grad_norm": 2.2422537803649902, "learning_rate": 9.427412400778728e-06, "loss": 1.4038, "step": 7936 }, { "epoch": 0.6220219435736677, "grad_norm": 4.435381889343262, "learning_rate": 9.427269104908997e-06, "loss": 1.2098, "step": 7937 }, { "epoch": 0.6221003134796238, "grad_norm": 4.179271221160889, "learning_rate": 9.427125792200208e-06, "loss": 1.3257, "step": 7938 }, { "epoch": 0.6221786833855799, "grad_norm": 3.741518974304199, "learning_rate": 9.426982462652906e-06, "loss": 1.2934, "step": 7939 }, { "epoch": 0.622257053291536, "grad_norm": 2.966367483139038, "learning_rate": 9.426839116267638e-06, "loss": 1.3515, "step": 7940 }, { "epoch": 0.6223354231974921, "grad_norm": 5.605717658996582, "learning_rate": 9.426695753044947e-06, "loss": 1.2676, "step": 7941 }, { "epoch": 0.6224137931034482, "grad_norm": 5.585620403289795, "learning_rate": 9.42655237298538e-06, "loss": 1.2053, "step": 7942 }, { "epoch": 0.6224921630094044, "grad_norm": 2.8505139350891113, "learning_rate": 9.426408976089481e-06, "loss": 1.163, "step": 7943 }, { "epoch": 0.6225705329153605, "grad_norm": 3.378666400909424, "learning_rate": 9.426265562357795e-06, "loss": 1.5289, "step": 7944 }, { "epoch": 0.6226489028213166, "grad_norm": 3.1456387042999268, "learning_rate": 9.426122131790869e-06, "loss": 1.5456, "step": 7945 }, { "epoch": 0.6227272727272727, "grad_norm": 2.8276240825653076, "learning_rate": 9.425978684389247e-06, "loss": 1.3236, "step": 7946 }, { "epoch": 0.6228056426332288, "grad_norm": 4.7548980712890625, "learning_rate": 9.425835220153478e-06, "loss": 1.2877, "step": 7947 }, { "epoch": 0.6228840125391849, "grad_norm": 1.9103889465332031, "learning_rate": 9.425691739084103e-06, "loss": 1.4289, "step": 7948 }, { "epoch": 0.622962382445141, "grad_norm": 11.416051864624023, "learning_rate": 9.42554824118167e-06, "loss": 1.415, "step": 7949 }, { "epoch": 0.6230407523510971, "grad_norm": 8.133131980895996, "learning_rate": 9.425404726446723e-06, "loss": 1.2202, "step": 7950 }, { "epoch": 0.6231191222570532, "grad_norm": 7.280026912689209, "learning_rate": 9.425261194879812e-06, "loss": 1.0948, "step": 7951 }, { "epoch": 0.6231974921630095, "grad_norm": 3.382028102874756, "learning_rate": 9.425117646481479e-06, "loss": 1.438, "step": 7952 }, { "epoch": 0.6232758620689656, "grad_norm": 2.122044801712036, "learning_rate": 9.424974081252272e-06, "loss": 1.1301, "step": 7953 }, { "epoch": 0.6233542319749217, "grad_norm": 2.9053757190704346, "learning_rate": 9.424830499192737e-06, "loss": 1.6037, "step": 7954 }, { "epoch": 0.6234326018808778, "grad_norm": 2.120196580886841, "learning_rate": 9.424686900303418e-06, "loss": 1.4807, "step": 7955 }, { "epoch": 0.6235109717868339, "grad_norm": 4.315945148468018, "learning_rate": 9.424543284584864e-06, "loss": 0.3526, "step": 7956 }, { "epoch": 0.62358934169279, "grad_norm": 3.7869186401367188, "learning_rate": 9.42439965203762e-06, "loss": 1.8784, "step": 7957 }, { "epoch": 0.6236677115987461, "grad_norm": 5.290273666381836, "learning_rate": 9.42425600266223e-06, "loss": 1.6398, "step": 7958 }, { "epoch": 0.6237460815047022, "grad_norm": 1.8510693311691284, "learning_rate": 9.424112336459243e-06, "loss": 1.2286, "step": 7959 }, { "epoch": 0.6238244514106583, "grad_norm": 5.846332550048828, "learning_rate": 9.423968653429207e-06, "loss": 2.0504, "step": 7960 }, { "epoch": 0.6239028213166145, "grad_norm": 4.063852310180664, "learning_rate": 9.423824953572667e-06, "loss": 1.4708, "step": 7961 }, { "epoch": 0.6239811912225706, "grad_norm": 2.7065653800964355, "learning_rate": 9.423681236890167e-06, "loss": 1.1722, "step": 7962 }, { "epoch": 0.6240595611285267, "grad_norm": 2.3834896087646484, "learning_rate": 9.423537503382256e-06, "loss": 1.4123, "step": 7963 }, { "epoch": 0.6241379310344828, "grad_norm": 12.292506217956543, "learning_rate": 9.423393753049481e-06, "loss": 1.8617, "step": 7964 }, { "epoch": 0.6242163009404389, "grad_norm": 2.555596351623535, "learning_rate": 9.423249985892388e-06, "loss": 1.6219, "step": 7965 }, { "epoch": 0.624294670846395, "grad_norm": 2.856471538543701, "learning_rate": 9.423106201911526e-06, "loss": 1.6313, "step": 7966 }, { "epoch": 0.6243730407523511, "grad_norm": 2.520439863204956, "learning_rate": 9.422962401107437e-06, "loss": 1.1205, "step": 7967 }, { "epoch": 0.6244514106583072, "grad_norm": 3.9286720752716064, "learning_rate": 9.42281858348067e-06, "loss": 1.2442, "step": 7968 }, { "epoch": 0.6245297805642633, "grad_norm": 4.529133319854736, "learning_rate": 9.422674749031775e-06, "loss": 1.7797, "step": 7969 }, { "epoch": 0.6246081504702194, "grad_norm": 2.3872342109680176, "learning_rate": 9.422530897761297e-06, "loss": 1.4461, "step": 7970 }, { "epoch": 0.6246865203761756, "grad_norm": 2.828552484512329, "learning_rate": 9.422387029669781e-06, "loss": 1.6552, "step": 7971 }, { "epoch": 0.6247648902821317, "grad_norm": 2.4231276512145996, "learning_rate": 9.422243144757779e-06, "loss": 1.5955, "step": 7972 }, { "epoch": 0.6248432601880878, "grad_norm": 2.262071132659912, "learning_rate": 9.422099243025834e-06, "loss": 1.282, "step": 7973 }, { "epoch": 0.6249216300940439, "grad_norm": 2.5148396492004395, "learning_rate": 9.421955324474494e-06, "loss": 1.308, "step": 7974 }, { "epoch": 0.625, "grad_norm": 3.6428236961364746, "learning_rate": 9.42181138910431e-06, "loss": 1.5739, "step": 7975 }, { "epoch": 0.6250783699059561, "grad_norm": 3.607032299041748, "learning_rate": 9.421667436915825e-06, "loss": 1.7225, "step": 7976 }, { "epoch": 0.6251567398119122, "grad_norm": 3.843564748764038, "learning_rate": 9.421523467909586e-06, "loss": 1.4101, "step": 7977 }, { "epoch": 0.6252351097178683, "grad_norm": 2.120332717895508, "learning_rate": 9.421379482086146e-06, "loss": 1.5098, "step": 7978 }, { "epoch": 0.6253134796238244, "grad_norm": 4.939579486846924, "learning_rate": 9.421235479446047e-06, "loss": 1.6111, "step": 7979 }, { "epoch": 0.6253918495297806, "grad_norm": 2.5180747509002686, "learning_rate": 9.421091459989842e-06, "loss": 1.5026, "step": 7980 }, { "epoch": 0.6254702194357367, "grad_norm": 3.0772781372070312, "learning_rate": 9.420947423718073e-06, "loss": 1.33, "step": 7981 }, { "epoch": 0.6255485893416928, "grad_norm": 2.7339634895324707, "learning_rate": 9.420803370631293e-06, "loss": 1.2122, "step": 7982 }, { "epoch": 0.6256269592476489, "grad_norm": 2.4375741481781006, "learning_rate": 9.420659300730046e-06, "loss": 1.5071, "step": 7983 }, { "epoch": 0.625705329153605, "grad_norm": 2.4821698665618896, "learning_rate": 9.420515214014883e-06, "loss": 1.4826, "step": 7984 }, { "epoch": 0.6257836990595611, "grad_norm": 2.2755672931671143, "learning_rate": 9.420371110486349e-06, "loss": 1.3786, "step": 7985 }, { "epoch": 0.6258620689655172, "grad_norm": 5.892001628875732, "learning_rate": 9.420226990144994e-06, "loss": 0.8175, "step": 7986 }, { "epoch": 0.6259404388714733, "grad_norm": 2.6448910236358643, "learning_rate": 9.420082852991367e-06, "loss": 1.2181, "step": 7987 }, { "epoch": 0.6260188087774294, "grad_norm": 4.029330253601074, "learning_rate": 9.419938699026015e-06, "loss": 1.0874, "step": 7988 }, { "epoch": 0.6260971786833855, "grad_norm": 3.8540494441986084, "learning_rate": 9.419794528249486e-06, "loss": 1.616, "step": 7989 }, { "epoch": 0.6261755485893417, "grad_norm": 3.8135414123535156, "learning_rate": 9.419650340662329e-06, "loss": 1.471, "step": 7990 }, { "epoch": 0.6262539184952978, "grad_norm": 2.5868687629699707, "learning_rate": 9.41950613626509e-06, "loss": 1.1316, "step": 7991 }, { "epoch": 0.6263322884012539, "grad_norm": 3.6016476154327393, "learning_rate": 9.419361915058322e-06, "loss": 1.4306, "step": 7992 }, { "epoch": 0.62641065830721, "grad_norm": 8.797195434570312, "learning_rate": 9.419217677042572e-06, "loss": 1.14, "step": 7993 }, { "epoch": 0.6264890282131661, "grad_norm": 4.579136848449707, "learning_rate": 9.419073422218386e-06, "loss": 1.3232, "step": 7994 }, { "epoch": 0.6265673981191222, "grad_norm": 2.807396173477173, "learning_rate": 9.418929150586315e-06, "loss": 1.0065, "step": 7995 }, { "epoch": 0.6266457680250783, "grad_norm": 9.955937385559082, "learning_rate": 9.418784862146908e-06, "loss": 0.9842, "step": 7996 }, { "epoch": 0.6267241379310344, "grad_norm": 2.172651529312134, "learning_rate": 9.418640556900713e-06, "loss": 1.825, "step": 7997 }, { "epoch": 0.6268025078369905, "grad_norm": 4.626625061035156, "learning_rate": 9.418496234848277e-06, "loss": 1.3618, "step": 7998 }, { "epoch": 0.6268808777429468, "grad_norm": 7.0014424324035645, "learning_rate": 9.418351895990153e-06, "loss": 1.0128, "step": 7999 }, { "epoch": 0.6269592476489029, "grad_norm": 3.911738634109497, "learning_rate": 9.418207540326888e-06, "loss": 1.2808, "step": 8000 }, { "epoch": 0.627037617554859, "grad_norm": 5.064860820770264, "learning_rate": 9.418063167859029e-06, "loss": 0.9411, "step": 8001 }, { "epoch": 0.6271159874608151, "grad_norm": 8.6058988571167, "learning_rate": 9.417918778587128e-06, "loss": 1.6991, "step": 8002 }, { "epoch": 0.6271943573667712, "grad_norm": 2.5696866512298584, "learning_rate": 9.417774372511733e-06, "loss": 1.6777, "step": 8003 }, { "epoch": 0.6272727272727273, "grad_norm": 2.345630407333374, "learning_rate": 9.417629949633391e-06, "loss": 1.3593, "step": 8004 }, { "epoch": 0.6273510971786834, "grad_norm": 2.993189811706543, "learning_rate": 9.417485509952657e-06, "loss": 1.5372, "step": 8005 }, { "epoch": 0.6274294670846395, "grad_norm": 2.1881072521209717, "learning_rate": 9.417341053470076e-06, "loss": 1.5153, "step": 8006 }, { "epoch": 0.6275078369905956, "grad_norm": 2.049137830734253, "learning_rate": 9.417196580186199e-06, "loss": 0.9214, "step": 8007 }, { "epoch": 0.6275862068965518, "grad_norm": 5.610469341278076, "learning_rate": 9.417052090101573e-06, "loss": 1.4037, "step": 8008 }, { "epoch": 0.6276645768025079, "grad_norm": 5.003658294677734, "learning_rate": 9.416907583216752e-06, "loss": 0.8209, "step": 8009 }, { "epoch": 0.627742946708464, "grad_norm": 3.4532485008239746, "learning_rate": 9.416763059532282e-06, "loss": 1.7145, "step": 8010 }, { "epoch": 0.6278213166144201, "grad_norm": 9.697517395019531, "learning_rate": 9.416618519048715e-06, "loss": 1.5631, "step": 8011 }, { "epoch": 0.6278996865203762, "grad_norm": 3.0574347972869873, "learning_rate": 9.416473961766599e-06, "loss": 1.3779, "step": 8012 }, { "epoch": 0.6279780564263323, "grad_norm": 3.6862845420837402, "learning_rate": 9.416329387686485e-06, "loss": 1.6708, "step": 8013 }, { "epoch": 0.6280564263322884, "grad_norm": 3.4358136653900146, "learning_rate": 9.41618479680892e-06, "loss": 1.3907, "step": 8014 }, { "epoch": 0.6281347962382445, "grad_norm": 2.4745893478393555, "learning_rate": 9.416040189134458e-06, "loss": 1.4405, "step": 8015 }, { "epoch": 0.6282131661442006, "grad_norm": 2.080935478210449, "learning_rate": 9.415895564663649e-06, "loss": 1.6629, "step": 8016 }, { "epoch": 0.6282915360501568, "grad_norm": 2.4404327869415283, "learning_rate": 9.41575092339704e-06, "loss": 1.3449, "step": 8017 }, { "epoch": 0.6283699059561129, "grad_norm": 2.455709457397461, "learning_rate": 9.415606265335181e-06, "loss": 1.4723, "step": 8018 }, { "epoch": 0.628448275862069, "grad_norm": 2.0016562938690186, "learning_rate": 9.415461590478626e-06, "loss": 1.5056, "step": 8019 }, { "epoch": 0.6285266457680251, "grad_norm": 5.396523475646973, "learning_rate": 9.415316898827923e-06, "loss": 1.5408, "step": 8020 }, { "epoch": 0.6286050156739812, "grad_norm": 4.89795446395874, "learning_rate": 9.41517219038362e-06, "loss": 1.0545, "step": 8021 }, { "epoch": 0.6286833855799373, "grad_norm": 2.1247377395629883, "learning_rate": 9.415027465146272e-06, "loss": 1.5003, "step": 8022 }, { "epoch": 0.6287617554858934, "grad_norm": 2.096139907836914, "learning_rate": 9.414882723116428e-06, "loss": 1.3602, "step": 8023 }, { "epoch": 0.6288401253918495, "grad_norm": 2.940624237060547, "learning_rate": 9.414737964294636e-06, "loss": 1.4132, "step": 8024 }, { "epoch": 0.6289184952978056, "grad_norm": 3.8434271812438965, "learning_rate": 9.41459318868145e-06, "loss": 1.45, "step": 8025 }, { "epoch": 0.6289968652037617, "grad_norm": 6.481940746307373, "learning_rate": 9.414448396277416e-06, "loss": 0.9172, "step": 8026 }, { "epoch": 0.6290752351097179, "grad_norm": 2.155521869659424, "learning_rate": 9.414303587083088e-06, "loss": 1.1264, "step": 8027 }, { "epoch": 0.629153605015674, "grad_norm": 6.6006927490234375, "learning_rate": 9.41415876109902e-06, "loss": 1.5122, "step": 8028 }, { "epoch": 0.6292319749216301, "grad_norm": 2.466329336166382, "learning_rate": 9.414013918325756e-06, "loss": 1.5117, "step": 8029 }, { "epoch": 0.6293103448275862, "grad_norm": 2.288752794265747, "learning_rate": 9.413869058763851e-06, "loss": 1.3148, "step": 8030 }, { "epoch": 0.6293887147335423, "grad_norm": 2.427973747253418, "learning_rate": 9.413724182413858e-06, "loss": 1.3359, "step": 8031 }, { "epoch": 0.6294670846394984, "grad_norm": 5.003096580505371, "learning_rate": 9.413579289276322e-06, "loss": 1.8135, "step": 8032 }, { "epoch": 0.6295454545454545, "grad_norm": 1.9252055883407593, "learning_rate": 9.413434379351798e-06, "loss": 1.268, "step": 8033 }, { "epoch": 0.6296238244514106, "grad_norm": 6.979554176330566, "learning_rate": 9.413289452640838e-06, "loss": 1.8966, "step": 8034 }, { "epoch": 0.6297021943573667, "grad_norm": 2.469949960708618, "learning_rate": 9.413144509143988e-06, "loss": 1.091, "step": 8035 }, { "epoch": 0.6297805642633229, "grad_norm": 4.940969467163086, "learning_rate": 9.412999548861807e-06, "loss": 2.0106, "step": 8036 }, { "epoch": 0.629858934169279, "grad_norm": 2.4239180088043213, "learning_rate": 9.412854571794841e-06, "loss": 1.3295, "step": 8037 }, { "epoch": 0.6299373040752351, "grad_norm": 2.188950777053833, "learning_rate": 9.412709577943641e-06, "loss": 1.2527, "step": 8038 }, { "epoch": 0.6300156739811912, "grad_norm": 2.3077335357666016, "learning_rate": 9.412564567308763e-06, "loss": 1.3484, "step": 8039 }, { "epoch": 0.6300940438871473, "grad_norm": 3.9209508895874023, "learning_rate": 9.412419539890754e-06, "loss": 1.7989, "step": 8040 }, { "epoch": 0.6301724137931034, "grad_norm": 2.870793342590332, "learning_rate": 9.41227449569017e-06, "loss": 1.6344, "step": 8041 }, { "epoch": 0.6302507836990595, "grad_norm": 2.1344332695007324, "learning_rate": 9.412129434707558e-06, "loss": 1.2869, "step": 8042 }, { "epoch": 0.6303291536050156, "grad_norm": 4.425113201141357, "learning_rate": 9.41198435694347e-06, "loss": 1.4317, "step": 8043 }, { "epoch": 0.6304075235109717, "grad_norm": 3.384458303451538, "learning_rate": 9.411839262398463e-06, "loss": 1.499, "step": 8044 }, { "epoch": 0.6304858934169278, "grad_norm": 4.6798248291015625, "learning_rate": 9.411694151073085e-06, "loss": 1.237, "step": 8045 }, { "epoch": 0.6305642633228841, "grad_norm": 2.1597225666046143, "learning_rate": 9.411549022967888e-06, "loss": 1.2914, "step": 8046 }, { "epoch": 0.6306426332288402, "grad_norm": 3.331005096435547, "learning_rate": 9.411403878083424e-06, "loss": 1.2688, "step": 8047 }, { "epoch": 0.6307210031347963, "grad_norm": 4.44153356552124, "learning_rate": 9.411258716420246e-06, "loss": 1.4788, "step": 8048 }, { "epoch": 0.6307993730407524, "grad_norm": 2.716057300567627, "learning_rate": 9.411113537978905e-06, "loss": 1.3603, "step": 8049 }, { "epoch": 0.6308777429467085, "grad_norm": 2.527641773223877, "learning_rate": 9.410968342759954e-06, "loss": 1.1398, "step": 8050 }, { "epoch": 0.6309561128526646, "grad_norm": 10.035465240478516, "learning_rate": 9.410823130763945e-06, "loss": 0.5418, "step": 8051 }, { "epoch": 0.6310344827586207, "grad_norm": 3.522362470626831, "learning_rate": 9.41067790199143e-06, "loss": 1.5672, "step": 8052 }, { "epoch": 0.6311128526645768, "grad_norm": 2.889648914337158, "learning_rate": 9.410532656442962e-06, "loss": 1.4471, "step": 8053 }, { "epoch": 0.631191222570533, "grad_norm": 7.930133819580078, "learning_rate": 9.410387394119094e-06, "loss": 0.9475, "step": 8054 }, { "epoch": 0.6312695924764891, "grad_norm": 3.86039137840271, "learning_rate": 9.410242115020376e-06, "loss": 1.5404, "step": 8055 }, { "epoch": 0.6313479623824452, "grad_norm": 2.5172226428985596, "learning_rate": 9.410096819147365e-06, "loss": 1.5633, "step": 8056 }, { "epoch": 0.6314263322884013, "grad_norm": 3.3013272285461426, "learning_rate": 9.409951506500607e-06, "loss": 1.2803, "step": 8057 }, { "epoch": 0.6315047021943574, "grad_norm": 3.7675275802612305, "learning_rate": 9.409806177080663e-06, "loss": 1.8606, "step": 8058 }, { "epoch": 0.6315830721003135, "grad_norm": 2.692612409591675, "learning_rate": 9.409660830888077e-06, "loss": 1.156, "step": 8059 }, { "epoch": 0.6316614420062696, "grad_norm": 2.259097099304199, "learning_rate": 9.409515467923409e-06, "loss": 1.3639, "step": 8060 }, { "epoch": 0.6317398119122257, "grad_norm": 1.9855395555496216, "learning_rate": 9.409370088187209e-06, "loss": 0.7641, "step": 8061 }, { "epoch": 0.6318181818181818, "grad_norm": 3.887795925140381, "learning_rate": 9.409224691680027e-06, "loss": 1.3177, "step": 8062 }, { "epoch": 0.631896551724138, "grad_norm": 2.5282437801361084, "learning_rate": 9.409079278402422e-06, "loss": 1.485, "step": 8063 }, { "epoch": 0.631974921630094, "grad_norm": 3.8339388370513916, "learning_rate": 9.408933848354945e-06, "loss": 1.3979, "step": 8064 }, { "epoch": 0.6320532915360502, "grad_norm": 2.9359567165374756, "learning_rate": 9.408788401538145e-06, "loss": 1.308, "step": 8065 }, { "epoch": 0.6321316614420063, "grad_norm": 3.1742515563964844, "learning_rate": 9.40864293795258e-06, "loss": 1.6127, "step": 8066 }, { "epoch": 0.6322100313479624, "grad_norm": 2.1507599353790283, "learning_rate": 9.408497457598803e-06, "loss": 1.5776, "step": 8067 }, { "epoch": 0.6322884012539185, "grad_norm": 3.12729811668396, "learning_rate": 9.408351960477363e-06, "loss": 1.7026, "step": 8068 }, { "epoch": 0.6323667711598746, "grad_norm": 8.405387878417969, "learning_rate": 9.408206446588818e-06, "loss": 1.2779, "step": 8069 }, { "epoch": 0.6324451410658307, "grad_norm": 3.0117921829223633, "learning_rate": 9.40806091593372e-06, "loss": 1.7537, "step": 8070 }, { "epoch": 0.6325235109717868, "grad_norm": 2.2274277210235596, "learning_rate": 9.407915368512622e-06, "loss": 1.4312, "step": 8071 }, { "epoch": 0.6326018808777429, "grad_norm": 5.494700908660889, "learning_rate": 9.40776980432608e-06, "loss": 1.6913, "step": 8072 }, { "epoch": 0.632680250783699, "grad_norm": 2.978874921798706, "learning_rate": 9.407624223374643e-06, "loss": 1.277, "step": 8073 }, { "epoch": 0.6327586206896552, "grad_norm": 1.9170693159103394, "learning_rate": 9.407478625658868e-06, "loss": 1.3443, "step": 8074 }, { "epoch": 0.6328369905956113, "grad_norm": 14.161563873291016, "learning_rate": 9.407333011179308e-06, "loss": 0.8887, "step": 8075 }, { "epoch": 0.6329153605015674, "grad_norm": 3.2991220951080322, "learning_rate": 9.407187379936518e-06, "loss": 1.4028, "step": 8076 }, { "epoch": 0.6329937304075235, "grad_norm": 2.3486433029174805, "learning_rate": 9.40704173193105e-06, "loss": 0.997, "step": 8077 }, { "epoch": 0.6330721003134796, "grad_norm": 3.213757038116455, "learning_rate": 9.406896067163459e-06, "loss": 1.6363, "step": 8078 }, { "epoch": 0.6331504702194357, "grad_norm": 1.9306213855743408, "learning_rate": 9.406750385634298e-06, "loss": 1.1759, "step": 8079 }, { "epoch": 0.6332288401253918, "grad_norm": 2.7481939792633057, "learning_rate": 9.406604687344123e-06, "loss": 0.9531, "step": 8080 }, { "epoch": 0.6333072100313479, "grad_norm": 7.041552543640137, "learning_rate": 9.406458972293488e-06, "loss": 1.2098, "step": 8081 }, { "epoch": 0.633385579937304, "grad_norm": 5.388462543487549, "learning_rate": 9.406313240482944e-06, "loss": 2.0355, "step": 8082 }, { "epoch": 0.6334639498432602, "grad_norm": 6.5049543380737305, "learning_rate": 9.40616749191305e-06, "loss": 1.5906, "step": 8083 }, { "epoch": 0.6335423197492163, "grad_norm": 2.1902365684509277, "learning_rate": 9.406021726584358e-06, "loss": 1.5326, "step": 8084 }, { "epoch": 0.6336206896551724, "grad_norm": 1.8485373258590698, "learning_rate": 9.405875944497421e-06, "loss": 1.5627, "step": 8085 }, { "epoch": 0.6336990595611285, "grad_norm": 2.612159490585327, "learning_rate": 9.405730145652796e-06, "loss": 1.515, "step": 8086 }, { "epoch": 0.6337774294670846, "grad_norm": 6.259225845336914, "learning_rate": 9.405584330051036e-06, "loss": 1.5419, "step": 8087 }, { "epoch": 0.6338557993730407, "grad_norm": 3.3466544151306152, "learning_rate": 9.405438497692696e-06, "loss": 1.4649, "step": 8088 }, { "epoch": 0.6339341692789968, "grad_norm": 2.8549962043762207, "learning_rate": 9.40529264857833e-06, "loss": 1.4247, "step": 8089 }, { "epoch": 0.6340125391849529, "grad_norm": 3.942049264907837, "learning_rate": 9.405146782708494e-06, "loss": 1.6705, "step": 8090 }, { "epoch": 0.634090909090909, "grad_norm": 9.779869079589844, "learning_rate": 9.405000900083743e-06, "loss": 1.6187, "step": 8091 }, { "epoch": 0.6341692789968651, "grad_norm": 2.847407817840576, "learning_rate": 9.40485500070463e-06, "loss": 1.3681, "step": 8092 }, { "epoch": 0.6342476489028214, "grad_norm": 3.023401975631714, "learning_rate": 9.404709084571713e-06, "loss": 1.9968, "step": 8093 }, { "epoch": 0.6343260188087775, "grad_norm": 4.966886520385742, "learning_rate": 9.404563151685542e-06, "loss": 0.8091, "step": 8094 }, { "epoch": 0.6344043887147336, "grad_norm": 2.575242519378662, "learning_rate": 9.404417202046678e-06, "loss": 1.3064, "step": 8095 }, { "epoch": 0.6344827586206897, "grad_norm": 1.8665612936019897, "learning_rate": 9.40427123565567e-06, "loss": 1.3942, "step": 8096 }, { "epoch": 0.6345611285266458, "grad_norm": 6.681135654449463, "learning_rate": 9.404125252513078e-06, "loss": 1.2701, "step": 8097 }, { "epoch": 0.6346394984326019, "grad_norm": 2.9424285888671875, "learning_rate": 9.403979252619458e-06, "loss": 1.3309, "step": 8098 }, { "epoch": 0.634717868338558, "grad_norm": 2.920814275741577, "learning_rate": 9.403833235975361e-06, "loss": 1.2126, "step": 8099 }, { "epoch": 0.6347962382445141, "grad_norm": 5.831674098968506, "learning_rate": 9.403687202581344e-06, "loss": 1.6308, "step": 8100 }, { "epoch": 0.6348746081504703, "grad_norm": 2.309892416000366, "learning_rate": 9.403541152437963e-06, "loss": 1.1539, "step": 8101 }, { "epoch": 0.6349529780564264, "grad_norm": 4.570070743560791, "learning_rate": 9.403395085545774e-06, "loss": 1.3153, "step": 8102 }, { "epoch": 0.6350313479623825, "grad_norm": 3.3087069988250732, "learning_rate": 9.40324900190533e-06, "loss": 1.247, "step": 8103 }, { "epoch": 0.6351097178683386, "grad_norm": 9.205915451049805, "learning_rate": 9.40310290151719e-06, "loss": 1.5003, "step": 8104 }, { "epoch": 0.6351880877742947, "grad_norm": 2.764226198196411, "learning_rate": 9.402956784381907e-06, "loss": 0.9318, "step": 8105 }, { "epoch": 0.6352664576802508, "grad_norm": 3.941589593887329, "learning_rate": 9.40281065050004e-06, "loss": 1.5243, "step": 8106 }, { "epoch": 0.6353448275862069, "grad_norm": 2.1546177864074707, "learning_rate": 9.40266449987214e-06, "loss": 1.3504, "step": 8107 }, { "epoch": 0.635423197492163, "grad_norm": 2.9364163875579834, "learning_rate": 9.402518332498766e-06, "loss": 1.5315, "step": 8108 }, { "epoch": 0.6355015673981191, "grad_norm": 4.072878360748291, "learning_rate": 9.402372148380477e-06, "loss": 2.0098, "step": 8109 }, { "epoch": 0.6355799373040752, "grad_norm": 2.8283894062042236, "learning_rate": 9.402225947517822e-06, "loss": 1.4029, "step": 8110 }, { "epoch": 0.6356583072100314, "grad_norm": 3.4201767444610596, "learning_rate": 9.402079729911362e-06, "loss": 1.7364, "step": 8111 }, { "epoch": 0.6357366771159875, "grad_norm": 12.850128173828125, "learning_rate": 9.40193349556165e-06, "loss": 0.6039, "step": 8112 }, { "epoch": 0.6358150470219436, "grad_norm": 3.5207183361053467, "learning_rate": 9.401787244469246e-06, "loss": 1.3877, "step": 8113 }, { "epoch": 0.6358934169278997, "grad_norm": 4.181366920471191, "learning_rate": 9.401640976634703e-06, "loss": 1.4184, "step": 8114 }, { "epoch": 0.6359717868338558, "grad_norm": 3.294720411300659, "learning_rate": 9.401494692058581e-06, "loss": 1.5595, "step": 8115 }, { "epoch": 0.6360501567398119, "grad_norm": 4.2664995193481445, "learning_rate": 9.401348390741432e-06, "loss": 1.2831, "step": 8116 }, { "epoch": 0.636128526645768, "grad_norm": 1.7882364988327026, "learning_rate": 9.401202072683814e-06, "loss": 1.3942, "step": 8117 }, { "epoch": 0.6362068965517241, "grad_norm": 3.2187798023223877, "learning_rate": 9.401055737886285e-06, "loss": 1.2852, "step": 8118 }, { "epoch": 0.6362852664576802, "grad_norm": 2.223205089569092, "learning_rate": 9.4009093863494e-06, "loss": 1.2156, "step": 8119 }, { "epoch": 0.6363636363636364, "grad_norm": 4.454391956329346, "learning_rate": 9.400763018073715e-06, "loss": 1.2053, "step": 8120 }, { "epoch": 0.6364420062695925, "grad_norm": 2.0036544799804688, "learning_rate": 9.400616633059789e-06, "loss": 1.4362, "step": 8121 }, { "epoch": 0.6365203761755486, "grad_norm": 2.985903739929199, "learning_rate": 9.400470231308176e-06, "loss": 1.3421, "step": 8122 }, { "epoch": 0.6365987460815047, "grad_norm": 2.356219530105591, "learning_rate": 9.400323812819437e-06, "loss": 1.4224, "step": 8123 }, { "epoch": 0.6366771159874608, "grad_norm": 1.9720035791397095, "learning_rate": 9.400177377594126e-06, "loss": 1.2426, "step": 8124 }, { "epoch": 0.6367554858934169, "grad_norm": 2.2852776050567627, "learning_rate": 9.400030925632799e-06, "loss": 1.5811, "step": 8125 }, { "epoch": 0.636833855799373, "grad_norm": 5.342088222503662, "learning_rate": 9.399884456936015e-06, "loss": 1.3922, "step": 8126 }, { "epoch": 0.6369122257053291, "grad_norm": 3.386521577835083, "learning_rate": 9.39973797150433e-06, "loss": 1.384, "step": 8127 }, { "epoch": 0.6369905956112852, "grad_norm": 2.405517339706421, "learning_rate": 9.399591469338302e-06, "loss": 1.3434, "step": 8128 }, { "epoch": 0.6370689655172413, "grad_norm": 3.6227855682373047, "learning_rate": 9.39944495043849e-06, "loss": 1.4944, "step": 8129 }, { "epoch": 0.6371473354231975, "grad_norm": 3.0327630043029785, "learning_rate": 9.399298414805446e-06, "loss": 1.0966, "step": 8130 }, { "epoch": 0.6372257053291536, "grad_norm": 2.6708991527557373, "learning_rate": 9.39915186243973e-06, "loss": 1.2129, "step": 8131 }, { "epoch": 0.6373040752351097, "grad_norm": 9.629849433898926, "learning_rate": 9.399005293341904e-06, "loss": 2.0774, "step": 8132 }, { "epoch": 0.6373824451410658, "grad_norm": 10.27575397491455, "learning_rate": 9.398858707512517e-06, "loss": 1.1442, "step": 8133 }, { "epoch": 0.6374608150470219, "grad_norm": 2.5877749919891357, "learning_rate": 9.398712104952132e-06, "loss": 1.4847, "step": 8134 }, { "epoch": 0.637539184952978, "grad_norm": 7.6112589836120605, "learning_rate": 9.398565485661307e-06, "loss": 1.3534, "step": 8135 }, { "epoch": 0.6376175548589341, "grad_norm": 2.303546905517578, "learning_rate": 9.398418849640596e-06, "loss": 1.4195, "step": 8136 }, { "epoch": 0.6376959247648902, "grad_norm": 3.0341904163360596, "learning_rate": 9.398272196890562e-06, "loss": 1.4843, "step": 8137 }, { "epoch": 0.6377742946708463, "grad_norm": 2.642698049545288, "learning_rate": 9.398125527411757e-06, "loss": 1.2981, "step": 8138 }, { "epoch": 0.6378526645768025, "grad_norm": 2.1376583576202393, "learning_rate": 9.397978841204741e-06, "loss": 0.6428, "step": 8139 }, { "epoch": 0.6379310344827587, "grad_norm": 3.683222532272339, "learning_rate": 9.397832138270073e-06, "loss": 1.3665, "step": 8140 }, { "epoch": 0.6380094043887148, "grad_norm": 3.017353057861328, "learning_rate": 9.397685418608311e-06, "loss": 0.9068, "step": 8141 }, { "epoch": 0.6380877742946709, "grad_norm": 2.6852877140045166, "learning_rate": 9.397538682220014e-06, "loss": 1.4907, "step": 8142 }, { "epoch": 0.638166144200627, "grad_norm": 10.092913627624512, "learning_rate": 9.397391929105737e-06, "loss": 1.2492, "step": 8143 }, { "epoch": 0.6382445141065831, "grad_norm": 2.9626283645629883, "learning_rate": 9.39724515926604e-06, "loss": 1.6787, "step": 8144 }, { "epoch": 0.6383228840125392, "grad_norm": 3.793553113937378, "learning_rate": 9.397098372701481e-06, "loss": 0.923, "step": 8145 }, { "epoch": 0.6384012539184953, "grad_norm": 2.1106550693511963, "learning_rate": 9.396951569412618e-06, "loss": 1.1452, "step": 8146 }, { "epoch": 0.6384796238244514, "grad_norm": 2.932467460632324, "learning_rate": 9.396804749400008e-06, "loss": 1.3185, "step": 8147 }, { "epoch": 0.6385579937304076, "grad_norm": 2.8659884929656982, "learning_rate": 9.396657912664213e-06, "loss": 0.8949, "step": 8148 }, { "epoch": 0.6386363636363637, "grad_norm": 3.972625494003296, "learning_rate": 9.396511059205791e-06, "loss": 1.1194, "step": 8149 }, { "epoch": 0.6387147335423198, "grad_norm": 3.361199140548706, "learning_rate": 9.396364189025296e-06, "loss": 1.3036, "step": 8150 }, { "epoch": 0.6387931034482759, "grad_norm": 3.561187744140625, "learning_rate": 9.396217302123291e-06, "loss": 1.282, "step": 8151 }, { "epoch": 0.638871473354232, "grad_norm": 2.293611764907837, "learning_rate": 9.396070398500334e-06, "loss": 1.3424, "step": 8152 }, { "epoch": 0.6389498432601881, "grad_norm": 6.327779769897461, "learning_rate": 9.395923478156982e-06, "loss": 1.9972, "step": 8153 }, { "epoch": 0.6390282131661442, "grad_norm": 4.641220569610596, "learning_rate": 9.395776541093798e-06, "loss": 1.5028, "step": 8154 }, { "epoch": 0.6391065830721003, "grad_norm": 2.5978004932403564, "learning_rate": 9.395629587311334e-06, "loss": 1.2017, "step": 8155 }, { "epoch": 0.6391849529780564, "grad_norm": 3.5988385677337646, "learning_rate": 9.395482616810154e-06, "loss": 1.248, "step": 8156 }, { "epoch": 0.6392633228840126, "grad_norm": 2.632415533065796, "learning_rate": 9.395335629590817e-06, "loss": 1.0387, "step": 8157 }, { "epoch": 0.6393416927899687, "grad_norm": 2.2042958736419678, "learning_rate": 9.39518862565388e-06, "loss": 1.3545, "step": 8158 }, { "epoch": 0.6394200626959248, "grad_norm": 3.2163288593292236, "learning_rate": 9.395041604999904e-06, "loss": 1.2821, "step": 8159 }, { "epoch": 0.6394984326018809, "grad_norm": 2.748567819595337, "learning_rate": 9.394894567629445e-06, "loss": 1.4713, "step": 8160 }, { "epoch": 0.639576802507837, "grad_norm": 7.19326639175415, "learning_rate": 9.394747513543066e-06, "loss": 0.7385, "step": 8161 }, { "epoch": 0.6396551724137931, "grad_norm": 6.803121566772461, "learning_rate": 9.394600442741324e-06, "loss": 2.1201, "step": 8162 }, { "epoch": 0.6397335423197492, "grad_norm": 2.074648141860962, "learning_rate": 9.39445335522478e-06, "loss": 1.4695, "step": 8163 }, { "epoch": 0.6398119122257053, "grad_norm": 3.5691142082214355, "learning_rate": 9.394306250993993e-06, "loss": 1.5248, "step": 8164 }, { "epoch": 0.6398902821316614, "grad_norm": 5.666017532348633, "learning_rate": 9.39415913004952e-06, "loss": 0.8279, "step": 8165 }, { "epoch": 0.6399686520376175, "grad_norm": 8.72083854675293, "learning_rate": 9.394011992391924e-06, "loss": 1.9507, "step": 8166 }, { "epoch": 0.6400470219435737, "grad_norm": 3.5166947841644287, "learning_rate": 9.393864838021762e-06, "loss": 1.5905, "step": 8167 }, { "epoch": 0.6401253918495298, "grad_norm": 2.2932722568511963, "learning_rate": 9.393717666939597e-06, "loss": 1.6315, "step": 8168 }, { "epoch": 0.6402037617554859, "grad_norm": 3.704010009765625, "learning_rate": 9.393570479145986e-06, "loss": 0.9654, "step": 8169 }, { "epoch": 0.640282131661442, "grad_norm": 3.3861277103424072, "learning_rate": 9.393423274641489e-06, "loss": 1.1828, "step": 8170 }, { "epoch": 0.6403605015673981, "grad_norm": 4.021266460418701, "learning_rate": 9.393276053426667e-06, "loss": 1.5891, "step": 8171 }, { "epoch": 0.6404388714733542, "grad_norm": 3.6340436935424805, "learning_rate": 9.39312881550208e-06, "loss": 1.5908, "step": 8172 }, { "epoch": 0.6405172413793103, "grad_norm": 10.9508695602417, "learning_rate": 9.392981560868286e-06, "loss": 1.0871, "step": 8173 }, { "epoch": 0.6405956112852664, "grad_norm": 2.3229360580444336, "learning_rate": 9.392834289525847e-06, "loss": 1.6002, "step": 8174 }, { "epoch": 0.6406739811912225, "grad_norm": 5.021461009979248, "learning_rate": 9.392687001475323e-06, "loss": 1.4257, "step": 8175 }, { "epoch": 0.6407523510971787, "grad_norm": 3.7643704414367676, "learning_rate": 9.392539696717275e-06, "loss": 1.5143, "step": 8176 }, { "epoch": 0.6408307210031348, "grad_norm": 2.89089298248291, "learning_rate": 9.39239237525226e-06, "loss": 1.4394, "step": 8177 }, { "epoch": 0.6409090909090909, "grad_norm": 2.8407843112945557, "learning_rate": 9.392245037080842e-06, "loss": 1.562, "step": 8178 }, { "epoch": 0.640987460815047, "grad_norm": 3.063443183898926, "learning_rate": 9.392097682203578e-06, "loss": 1.6072, "step": 8179 }, { "epoch": 0.6410658307210031, "grad_norm": 13.12391471862793, "learning_rate": 9.391950310621031e-06, "loss": 2.0149, "step": 8180 }, { "epoch": 0.6411442006269592, "grad_norm": 6.862210273742676, "learning_rate": 9.391802922333764e-06, "loss": 1.231, "step": 8181 }, { "epoch": 0.6412225705329153, "grad_norm": 2.9846653938293457, "learning_rate": 9.39165551734233e-06, "loss": 1.3644, "step": 8182 }, { "epoch": 0.6413009404388714, "grad_norm": 2.2554192543029785, "learning_rate": 9.391508095647298e-06, "loss": 1.2846, "step": 8183 }, { "epoch": 0.6413793103448275, "grad_norm": 2.8929569721221924, "learning_rate": 9.391360657249222e-06, "loss": 1.8367, "step": 8184 }, { "epoch": 0.6414576802507836, "grad_norm": 2.7199618816375732, "learning_rate": 9.391213202148666e-06, "loss": 1.4921, "step": 8185 }, { "epoch": 0.6415360501567398, "grad_norm": 2.7722127437591553, "learning_rate": 9.391065730346191e-06, "loss": 1.7354, "step": 8186 }, { "epoch": 0.641614420062696, "grad_norm": 3.559535264968872, "learning_rate": 9.390918241842358e-06, "loss": 1.1722, "step": 8187 }, { "epoch": 0.6416927899686521, "grad_norm": 6.291553974151611, "learning_rate": 9.390770736637727e-06, "loss": 0.6336, "step": 8188 }, { "epoch": 0.6417711598746082, "grad_norm": 2.5723397731781006, "learning_rate": 9.39062321473286e-06, "loss": 1.4346, "step": 8189 }, { "epoch": 0.6418495297805643, "grad_norm": 7.4255146980285645, "learning_rate": 9.390475676128315e-06, "loss": 1.8676, "step": 8190 }, { "epoch": 0.6419278996865204, "grad_norm": 3.7696211338043213, "learning_rate": 9.390328120824656e-06, "loss": 1.278, "step": 8191 }, { "epoch": 0.6420062695924765, "grad_norm": 6.250247478485107, "learning_rate": 9.390180548822446e-06, "loss": 1.0538, "step": 8192 }, { "epoch": 0.6420846394984326, "grad_norm": 2.0184595584869385, "learning_rate": 9.390032960122243e-06, "loss": 1.675, "step": 8193 }, { "epoch": 0.6421630094043888, "grad_norm": 2.2776787281036377, "learning_rate": 9.38988535472461e-06, "loss": 1.5411, "step": 8194 }, { "epoch": 0.6422413793103449, "grad_norm": 2.817404270172119, "learning_rate": 9.389737732630106e-06, "loss": 1.2295, "step": 8195 }, { "epoch": 0.642319749216301, "grad_norm": 3.547086477279663, "learning_rate": 9.389590093839296e-06, "loss": 1.8295, "step": 8196 }, { "epoch": 0.6423981191222571, "grad_norm": 4.818434715270996, "learning_rate": 9.38944243835274e-06, "loss": 1.4925, "step": 8197 }, { "epoch": 0.6424764890282132, "grad_norm": 10.998289108276367, "learning_rate": 9.389294766171e-06, "loss": 1.3352, "step": 8198 }, { "epoch": 0.6425548589341693, "grad_norm": 2.1653125286102295, "learning_rate": 9.389147077294636e-06, "loss": 1.3141, "step": 8199 }, { "epoch": 0.6426332288401254, "grad_norm": 4.022658824920654, "learning_rate": 9.388999371724212e-06, "loss": 1.5025, "step": 8200 }, { "epoch": 0.6427115987460815, "grad_norm": 2.168225049972534, "learning_rate": 9.388851649460288e-06, "loss": 1.2405, "step": 8201 }, { "epoch": 0.6427899686520376, "grad_norm": 2.2117156982421875, "learning_rate": 9.388703910503427e-06, "loss": 1.2566, "step": 8202 }, { "epoch": 0.6428683385579937, "grad_norm": 3.076551675796509, "learning_rate": 9.388556154854189e-06, "loss": 0.7891, "step": 8203 }, { "epoch": 0.6429467084639499, "grad_norm": 2.4667303562164307, "learning_rate": 9.388408382513138e-06, "loss": 1.4608, "step": 8204 }, { "epoch": 0.643025078369906, "grad_norm": 4.533122539520264, "learning_rate": 9.388260593480838e-06, "loss": 2.0353, "step": 8205 }, { "epoch": 0.6431034482758621, "grad_norm": 2.1192896366119385, "learning_rate": 9.388112787757846e-06, "loss": 1.332, "step": 8206 }, { "epoch": 0.6431818181818182, "grad_norm": 2.1360177993774414, "learning_rate": 9.387964965344728e-06, "loss": 1.2071, "step": 8207 }, { "epoch": 0.6432601880877743, "grad_norm": 2.8900949954986572, "learning_rate": 9.387817126242045e-06, "loss": 1.7745, "step": 8208 }, { "epoch": 0.6433385579937304, "grad_norm": 2.5721049308776855, "learning_rate": 9.38766927045036e-06, "loss": 1.2175, "step": 8209 }, { "epoch": 0.6434169278996865, "grad_norm": 2.5368220806121826, "learning_rate": 9.387521397970233e-06, "loss": 1.6255, "step": 8210 }, { "epoch": 0.6434952978056426, "grad_norm": 3.693125009536743, "learning_rate": 9.38737350880223e-06, "loss": 1.4824, "step": 8211 }, { "epoch": 0.6435736677115987, "grad_norm": 3.6293423175811768, "learning_rate": 9.387225602946909e-06, "loss": 1.5185, "step": 8212 }, { "epoch": 0.6436520376175549, "grad_norm": 4.4073381423950195, "learning_rate": 9.387077680404837e-06, "loss": 1.7861, "step": 8213 }, { "epoch": 0.643730407523511, "grad_norm": 3.293037176132202, "learning_rate": 9.386929741176576e-06, "loss": 1.5329, "step": 8214 }, { "epoch": 0.6438087774294671, "grad_norm": 3.229536771774292, "learning_rate": 9.386781785262686e-06, "loss": 1.7009, "step": 8215 }, { "epoch": 0.6438871473354232, "grad_norm": 1.9817808866500854, "learning_rate": 9.386633812663731e-06, "loss": 1.0498, "step": 8216 }, { "epoch": 0.6439655172413793, "grad_norm": 3.0391523838043213, "learning_rate": 9.386485823380275e-06, "loss": 1.6878, "step": 8217 }, { "epoch": 0.6440438871473354, "grad_norm": 2.213111639022827, "learning_rate": 9.38633781741288e-06, "loss": 1.565, "step": 8218 }, { "epoch": 0.6441222570532915, "grad_norm": 2.7319929599761963, "learning_rate": 9.386189794762109e-06, "loss": 1.2812, "step": 8219 }, { "epoch": 0.6442006269592476, "grad_norm": 3.0665266513824463, "learning_rate": 9.386041755428523e-06, "loss": 1.5325, "step": 8220 }, { "epoch": 0.6442789968652037, "grad_norm": 2.323317766189575, "learning_rate": 9.38589369941269e-06, "loss": 1.2469, "step": 8221 }, { "epoch": 0.6443573667711598, "grad_norm": 2.863598346710205, "learning_rate": 9.385745626715168e-06, "loss": 1.6303, "step": 8222 }, { "epoch": 0.644435736677116, "grad_norm": 5.751726150512695, "learning_rate": 9.385597537336523e-06, "loss": 1.7222, "step": 8223 }, { "epoch": 0.6445141065830721, "grad_norm": 2.4588663578033447, "learning_rate": 9.385449431277317e-06, "loss": 1.3414, "step": 8224 }, { "epoch": 0.6445924764890282, "grad_norm": 4.981959342956543, "learning_rate": 9.385301308538115e-06, "loss": 1.7175, "step": 8225 }, { "epoch": 0.6446708463949843, "grad_norm": 2.2750258445739746, "learning_rate": 9.385153169119478e-06, "loss": 1.5377, "step": 8226 }, { "epoch": 0.6447492163009404, "grad_norm": 4.5192766189575195, "learning_rate": 9.385005013021972e-06, "loss": 1.2044, "step": 8227 }, { "epoch": 0.6448275862068965, "grad_norm": 4.361126899719238, "learning_rate": 9.384856840246157e-06, "loss": 1.4841, "step": 8228 }, { "epoch": 0.6449059561128526, "grad_norm": 3.1730382442474365, "learning_rate": 9.384708650792602e-06, "loss": 1.5492, "step": 8229 }, { "epoch": 0.6449843260188087, "grad_norm": 2.972694158554077, "learning_rate": 9.384560444661866e-06, "loss": 1.4947, "step": 8230 }, { "epoch": 0.6450626959247648, "grad_norm": 2.4927728176116943, "learning_rate": 9.384412221854513e-06, "loss": 1.5866, "step": 8231 }, { "epoch": 0.645141065830721, "grad_norm": 11.808757781982422, "learning_rate": 9.384263982371109e-06, "loss": 0.8219, "step": 8232 }, { "epoch": 0.6452194357366771, "grad_norm": 4.459304332733154, "learning_rate": 9.384115726212215e-06, "loss": 1.3787, "step": 8233 }, { "epoch": 0.6452978056426333, "grad_norm": 3.9125781059265137, "learning_rate": 9.383967453378398e-06, "loss": 1.7027, "step": 8234 }, { "epoch": 0.6453761755485894, "grad_norm": 2.933713674545288, "learning_rate": 9.383819163870221e-06, "loss": 1.5204, "step": 8235 }, { "epoch": 0.6454545454545455, "grad_norm": 3.225656270980835, "learning_rate": 9.383670857688247e-06, "loss": 1.6739, "step": 8236 }, { "epoch": 0.6455329153605016, "grad_norm": 5.184718132019043, "learning_rate": 9.38352253483304e-06, "loss": 0.7025, "step": 8237 }, { "epoch": 0.6456112852664577, "grad_norm": 6.5071821212768555, "learning_rate": 9.383374195305167e-06, "loss": 1.3232, "step": 8238 }, { "epoch": 0.6456896551724138, "grad_norm": 2.8115267753601074, "learning_rate": 9.383225839105188e-06, "loss": 1.2966, "step": 8239 }, { "epoch": 0.64576802507837, "grad_norm": 4.107856273651123, "learning_rate": 9.38307746623367e-06, "loss": 1.4689, "step": 8240 }, { "epoch": 0.645846394984326, "grad_norm": 2.6463658809661865, "learning_rate": 9.382929076691177e-06, "loss": 1.517, "step": 8241 }, { "epoch": 0.6459247648902822, "grad_norm": 2.3313331604003906, "learning_rate": 9.382780670478272e-06, "loss": 1.5699, "step": 8242 }, { "epoch": 0.6460031347962383, "grad_norm": 15.474045753479004, "learning_rate": 9.382632247595521e-06, "loss": 1.2013, "step": 8243 }, { "epoch": 0.6460815047021944, "grad_norm": 3.634809970855713, "learning_rate": 9.382483808043488e-06, "loss": 1.5149, "step": 8244 }, { "epoch": 0.6461598746081505, "grad_norm": 8.06813907623291, "learning_rate": 9.382335351822738e-06, "loss": 1.2284, "step": 8245 }, { "epoch": 0.6462382445141066, "grad_norm": 2.108003616333008, "learning_rate": 9.382186878933834e-06, "loss": 1.0347, "step": 8246 }, { "epoch": 0.6463166144200627, "grad_norm": 4.218008995056152, "learning_rate": 9.382038389377344e-06, "loss": 0.8824, "step": 8247 }, { "epoch": 0.6463949843260188, "grad_norm": 6.024203777313232, "learning_rate": 9.381889883153828e-06, "loss": 1.2658, "step": 8248 }, { "epoch": 0.6464733542319749, "grad_norm": 3.916343927383423, "learning_rate": 9.381741360263857e-06, "loss": 1.0352, "step": 8249 }, { "epoch": 0.646551724137931, "grad_norm": 3.2291412353515625, "learning_rate": 9.38159282070799e-06, "loss": 1.5867, "step": 8250 }, { "epoch": 0.6466300940438872, "grad_norm": 4.454059600830078, "learning_rate": 9.381444264486795e-06, "loss": 1.5893, "step": 8251 }, { "epoch": 0.6467084639498433, "grad_norm": 8.979894638061523, "learning_rate": 9.381295691600835e-06, "loss": 2.3105, "step": 8252 }, { "epoch": 0.6467868338557994, "grad_norm": 6.746976852416992, "learning_rate": 9.381147102050678e-06, "loss": 1.7438, "step": 8253 }, { "epoch": 0.6468652037617555, "grad_norm": 3.2528626918792725, "learning_rate": 9.380998495836889e-06, "loss": 1.0891, "step": 8254 }, { "epoch": 0.6469435736677116, "grad_norm": 2.3706135749816895, "learning_rate": 9.38084987296003e-06, "loss": 1.0886, "step": 8255 }, { "epoch": 0.6470219435736677, "grad_norm": 4.359561920166016, "learning_rate": 9.380701233420668e-06, "loss": 1.5453, "step": 8256 }, { "epoch": 0.6471003134796238, "grad_norm": 4.147063255310059, "learning_rate": 9.38055257721937e-06, "loss": 1.1935, "step": 8257 }, { "epoch": 0.6471786833855799, "grad_norm": 5.79437255859375, "learning_rate": 9.380403904356699e-06, "loss": 1.4346, "step": 8258 }, { "epoch": 0.647257053291536, "grad_norm": 2.612250566482544, "learning_rate": 9.380255214833223e-06, "loss": 1.2915, "step": 8259 }, { "epoch": 0.6473354231974922, "grad_norm": 2.462986707687378, "learning_rate": 9.380106508649504e-06, "loss": 1.2601, "step": 8260 }, { "epoch": 0.6474137931034483, "grad_norm": 2.639418601989746, "learning_rate": 9.379957785806112e-06, "loss": 1.3531, "step": 8261 }, { "epoch": 0.6474921630094044, "grad_norm": 2.1002187728881836, "learning_rate": 9.379809046303607e-06, "loss": 1.2227, "step": 8262 }, { "epoch": 0.6475705329153605, "grad_norm": 5.988256931304932, "learning_rate": 9.37966029014256e-06, "loss": 1.5692, "step": 8263 }, { "epoch": 0.6476489028213166, "grad_norm": 2.650960922241211, "learning_rate": 9.379511517323533e-06, "loss": 1.1843, "step": 8264 }, { "epoch": 0.6477272727272727, "grad_norm": 2.9301817417144775, "learning_rate": 9.379362727847097e-06, "loss": 0.9588, "step": 8265 }, { "epoch": 0.6478056426332288, "grad_norm": 4.313936233520508, "learning_rate": 9.379213921713812e-06, "loss": 1.3259, "step": 8266 }, { "epoch": 0.6478840125391849, "grad_norm": 4.170596122741699, "learning_rate": 9.379065098924247e-06, "loss": 1.1142, "step": 8267 }, { "epoch": 0.647962382445141, "grad_norm": 7.296196460723877, "learning_rate": 9.378916259478968e-06, "loss": 0.7809, "step": 8268 }, { "epoch": 0.6480407523510971, "grad_norm": 3.7039411067962646, "learning_rate": 9.378767403378542e-06, "loss": 1.1898, "step": 8269 }, { "epoch": 0.6481191222570533, "grad_norm": 7.1834540367126465, "learning_rate": 9.378618530623532e-06, "loss": 1.5277, "step": 8270 }, { "epoch": 0.6481974921630094, "grad_norm": 3.261009454727173, "learning_rate": 9.378469641214506e-06, "loss": 1.3034, "step": 8271 }, { "epoch": 0.6482758620689655, "grad_norm": 2.5464866161346436, "learning_rate": 9.37832073515203e-06, "loss": 1.745, "step": 8272 }, { "epoch": 0.6483542319749216, "grad_norm": 4.164957046508789, "learning_rate": 9.378171812436673e-06, "loss": 1.1602, "step": 8273 }, { "epoch": 0.6484326018808777, "grad_norm": 3.6382765769958496, "learning_rate": 9.378022873068998e-06, "loss": 1.7965, "step": 8274 }, { "epoch": 0.6485109717868338, "grad_norm": 3.039013147354126, "learning_rate": 9.377873917049573e-06, "loss": 1.2189, "step": 8275 }, { "epoch": 0.6485893416927899, "grad_norm": 4.273515224456787, "learning_rate": 9.377724944378963e-06, "loss": 1.2787, "step": 8276 }, { "epoch": 0.648667711598746, "grad_norm": 6.317056179046631, "learning_rate": 9.377575955057737e-06, "loss": 0.893, "step": 8277 }, { "epoch": 0.6487460815047021, "grad_norm": 6.113688945770264, "learning_rate": 9.37742694908646e-06, "loss": 0.9511, "step": 8278 }, { "epoch": 0.6488244514106583, "grad_norm": 3.9962117671966553, "learning_rate": 9.3772779264657e-06, "loss": 0.8627, "step": 8279 }, { "epoch": 0.6489028213166145, "grad_norm": 7.039605617523193, "learning_rate": 9.377128887196023e-06, "loss": 2.4821, "step": 8280 }, { "epoch": 0.6489811912225706, "grad_norm": 2.3954408168792725, "learning_rate": 9.376979831277995e-06, "loss": 1.4297, "step": 8281 }, { "epoch": 0.6490595611285267, "grad_norm": 2.084033489227295, "learning_rate": 9.376830758712185e-06, "loss": 1.32, "step": 8282 }, { "epoch": 0.6491379310344828, "grad_norm": 1.9596384763717651, "learning_rate": 9.376681669499157e-06, "loss": 0.991, "step": 8283 }, { "epoch": 0.6492163009404389, "grad_norm": 3.1245005130767822, "learning_rate": 9.376532563639482e-06, "loss": 1.3529, "step": 8284 }, { "epoch": 0.649294670846395, "grad_norm": 2.08978533744812, "learning_rate": 9.376383441133725e-06, "loss": 1.3365, "step": 8285 }, { "epoch": 0.6493730407523511, "grad_norm": 3.5260539054870605, "learning_rate": 9.376234301982451e-06, "loss": 0.8842, "step": 8286 }, { "epoch": 0.6494514106583072, "grad_norm": 4.09757137298584, "learning_rate": 9.376085146186232e-06, "loss": 1.5549, "step": 8287 }, { "epoch": 0.6495297805642634, "grad_norm": 1.921947717666626, "learning_rate": 9.37593597374563e-06, "loss": 1.3522, "step": 8288 }, { "epoch": 0.6496081504702195, "grad_norm": 3.051342725753784, "learning_rate": 9.375786784661218e-06, "loss": 1.47, "step": 8289 }, { "epoch": 0.6496865203761756, "grad_norm": 2.6313955783843994, "learning_rate": 9.37563757893356e-06, "loss": 1.6587, "step": 8290 }, { "epoch": 0.6497648902821317, "grad_norm": 2.462712287902832, "learning_rate": 9.375488356563224e-06, "loss": 1.0632, "step": 8291 }, { "epoch": 0.6498432601880878, "grad_norm": 6.834906578063965, "learning_rate": 9.375339117550778e-06, "loss": 1.6763, "step": 8292 }, { "epoch": 0.6499216300940439, "grad_norm": 6.527278423309326, "learning_rate": 9.375189861896788e-06, "loss": 1.1463, "step": 8293 }, { "epoch": 0.65, "grad_norm": 3.4068362712860107, "learning_rate": 9.375040589601825e-06, "loss": 1.095, "step": 8294 }, { "epoch": 0.6500783699059561, "grad_norm": 3.8031704425811768, "learning_rate": 9.374891300666453e-06, "loss": 1.7354, "step": 8295 }, { "epoch": 0.6501567398119122, "grad_norm": 10.807568550109863, "learning_rate": 9.374741995091243e-06, "loss": 1.5598, "step": 8296 }, { "epoch": 0.6502351097178684, "grad_norm": 2.4914021492004395, "learning_rate": 9.37459267287676e-06, "loss": 1.2684, "step": 8297 }, { "epoch": 0.6503134796238245, "grad_norm": 3.1372275352478027, "learning_rate": 9.374443334023572e-06, "loss": 1.2766, "step": 8298 }, { "epoch": 0.6503918495297806, "grad_norm": 3.6841793060302734, "learning_rate": 9.374293978532251e-06, "loss": 1.8896, "step": 8299 }, { "epoch": 0.6504702194357367, "grad_norm": 2.8204469680786133, "learning_rate": 9.374144606403363e-06, "loss": 1.5043, "step": 8300 }, { "epoch": 0.6505485893416928, "grad_norm": 2.6056015491485596, "learning_rate": 9.373995217637476e-06, "loss": 1.5667, "step": 8301 }, { "epoch": 0.6506269592476489, "grad_norm": 4.47748327255249, "learning_rate": 9.373845812235155e-06, "loss": 1.7216, "step": 8302 }, { "epoch": 0.650705329153605, "grad_norm": 4.97037410736084, "learning_rate": 9.373696390196971e-06, "loss": 1.8017, "step": 8303 }, { "epoch": 0.6507836990595611, "grad_norm": 4.987340927124023, "learning_rate": 9.373546951523496e-06, "loss": 1.5097, "step": 8304 }, { "epoch": 0.6508620689655172, "grad_norm": 3.5811257362365723, "learning_rate": 9.373397496215291e-06, "loss": 1.203, "step": 8305 }, { "epoch": 0.6509404388714733, "grad_norm": 2.7616138458251953, "learning_rate": 9.37324802427293e-06, "loss": 1.3622, "step": 8306 }, { "epoch": 0.6510188087774295, "grad_norm": 4.875091075897217, "learning_rate": 9.37309853569698e-06, "loss": 1.2518, "step": 8307 }, { "epoch": 0.6510971786833856, "grad_norm": 2.468815565109253, "learning_rate": 9.372949030488008e-06, "loss": 1.4717, "step": 8308 }, { "epoch": 0.6511755485893417, "grad_norm": 7.1369147300720215, "learning_rate": 9.372799508646587e-06, "loss": 1.1531, "step": 8309 }, { "epoch": 0.6512539184952978, "grad_norm": 2.6570448875427246, "learning_rate": 9.37264997017328e-06, "loss": 1.6642, "step": 8310 }, { "epoch": 0.6513322884012539, "grad_norm": 5.420291900634766, "learning_rate": 9.372500415068659e-06, "loss": 1.2357, "step": 8311 }, { "epoch": 0.65141065830721, "grad_norm": 2.19173264503479, "learning_rate": 9.37235084333329e-06, "loss": 1.0696, "step": 8312 }, { "epoch": 0.6514890282131661, "grad_norm": 14.109718322753906, "learning_rate": 9.372201254967748e-06, "loss": 1.3371, "step": 8313 }, { "epoch": 0.6515673981191222, "grad_norm": 2.439788818359375, "learning_rate": 9.372051649972596e-06, "loss": 1.5933, "step": 8314 }, { "epoch": 0.6516457680250783, "grad_norm": 2.127020835876465, "learning_rate": 9.371902028348407e-06, "loss": 1.7369, "step": 8315 }, { "epoch": 0.6517241379310345, "grad_norm": 7.5768232345581055, "learning_rate": 9.371752390095746e-06, "loss": 0.5105, "step": 8316 }, { "epoch": 0.6518025078369906, "grad_norm": 2.0805630683898926, "learning_rate": 9.371602735215187e-06, "loss": 1.3913, "step": 8317 }, { "epoch": 0.6518808777429467, "grad_norm": 3.0008740425109863, "learning_rate": 9.371453063707294e-06, "loss": 1.684, "step": 8318 }, { "epoch": 0.6519592476489028, "grad_norm": 3.3103368282318115, "learning_rate": 9.37130337557264e-06, "loss": 1.4677, "step": 8319 }, { "epoch": 0.6520376175548589, "grad_norm": 2.4214916229248047, "learning_rate": 9.371153670811792e-06, "loss": 1.902, "step": 8320 }, { "epoch": 0.652115987460815, "grad_norm": 3.3520541191101074, "learning_rate": 9.371003949425323e-06, "loss": 1.0005, "step": 8321 }, { "epoch": 0.6521943573667711, "grad_norm": 4.267852783203125, "learning_rate": 9.370854211413797e-06, "loss": 1.2786, "step": 8322 }, { "epoch": 0.6522727272727272, "grad_norm": 2.6933233737945557, "learning_rate": 9.37070445677779e-06, "loss": 1.3672, "step": 8323 }, { "epoch": 0.6523510971786833, "grad_norm": 2.803304433822632, "learning_rate": 9.370554685517864e-06, "loss": 1.5971, "step": 8324 }, { "epoch": 0.6524294670846394, "grad_norm": 3.2452948093414307, "learning_rate": 9.370404897634597e-06, "loss": 1.6251, "step": 8325 }, { "epoch": 0.6525078369905956, "grad_norm": 8.049468994140625, "learning_rate": 9.370255093128553e-06, "loss": 1.4853, "step": 8326 }, { "epoch": 0.6525862068965518, "grad_norm": 3.1393449306488037, "learning_rate": 9.370105272000303e-06, "loss": 1.445, "step": 8327 }, { "epoch": 0.6526645768025079, "grad_norm": 6.46574068069458, "learning_rate": 9.369955434250418e-06, "loss": 0.8394, "step": 8328 }, { "epoch": 0.652742946708464, "grad_norm": 3.6908557415008545, "learning_rate": 9.369805579879465e-06, "loss": 1.3836, "step": 8329 }, { "epoch": 0.6528213166144201, "grad_norm": 8.994416236877441, "learning_rate": 9.36965570888802e-06, "loss": 1.2937, "step": 8330 }, { "epoch": 0.6528996865203762, "grad_norm": 2.1821682453155518, "learning_rate": 9.369505821276645e-06, "loss": 1.1003, "step": 8331 }, { "epoch": 0.6529780564263323, "grad_norm": 4.078178405761719, "learning_rate": 9.369355917045916e-06, "loss": 1.4559, "step": 8332 }, { "epoch": 0.6530564263322884, "grad_norm": 2.7071566581726074, "learning_rate": 9.369205996196402e-06, "loss": 1.6684, "step": 8333 }, { "epoch": 0.6531347962382446, "grad_norm": 2.6972055435180664, "learning_rate": 9.36905605872867e-06, "loss": 1.5987, "step": 8334 }, { "epoch": 0.6532131661442007, "grad_norm": 2.162165880203247, "learning_rate": 9.368906104643295e-06, "loss": 1.4219, "step": 8335 }, { "epoch": 0.6532915360501568, "grad_norm": 4.971725940704346, "learning_rate": 9.368756133940844e-06, "loss": 2.0723, "step": 8336 }, { "epoch": 0.6533699059561129, "grad_norm": 2.2085440158843994, "learning_rate": 9.36860614662189e-06, "loss": 1.4361, "step": 8337 }, { "epoch": 0.653448275862069, "grad_norm": 1.9613680839538574, "learning_rate": 9.368456142687001e-06, "loss": 1.3159, "step": 8338 }, { "epoch": 0.6535266457680251, "grad_norm": 2.471397638320923, "learning_rate": 9.36830612213675e-06, "loss": 1.7242, "step": 8339 }, { "epoch": 0.6536050156739812, "grad_norm": 2.6116440296173096, "learning_rate": 9.368156084971705e-06, "loss": 1.2304, "step": 8340 }, { "epoch": 0.6536833855799373, "grad_norm": 3.1611218452453613, "learning_rate": 9.368006031192438e-06, "loss": 0.9669, "step": 8341 }, { "epoch": 0.6537617554858934, "grad_norm": 3.0016160011291504, "learning_rate": 9.367855960799519e-06, "loss": 1.8183, "step": 8342 }, { "epoch": 0.6538401253918495, "grad_norm": 3.3042445182800293, "learning_rate": 9.36770587379352e-06, "loss": 0.8955, "step": 8343 }, { "epoch": 0.6539184952978057, "grad_norm": 2.1782422065734863, "learning_rate": 9.367555770175012e-06, "loss": 1.2703, "step": 8344 }, { "epoch": 0.6539968652037618, "grad_norm": 2.9687230587005615, "learning_rate": 9.367405649944564e-06, "loss": 1.7812, "step": 8345 }, { "epoch": 0.6540752351097179, "grad_norm": 1.8871339559555054, "learning_rate": 9.367255513102748e-06, "loss": 1.0366, "step": 8346 }, { "epoch": 0.654153605015674, "grad_norm": 4.149489879608154, "learning_rate": 9.367105359650138e-06, "loss": 1.8604, "step": 8347 }, { "epoch": 0.6542319749216301, "grad_norm": 7.723112106323242, "learning_rate": 9.366955189587299e-06, "loss": 1.5851, "step": 8348 }, { "epoch": 0.6543103448275862, "grad_norm": 3.0627381801605225, "learning_rate": 9.366805002914807e-06, "loss": 1.3799, "step": 8349 }, { "epoch": 0.6543887147335423, "grad_norm": 4.2199273109436035, "learning_rate": 9.36665479963323e-06, "loss": 1.3802, "step": 8350 }, { "epoch": 0.6544670846394984, "grad_norm": 2.3027992248535156, "learning_rate": 9.366504579743142e-06, "loss": 1.3836, "step": 8351 }, { "epoch": 0.6545454545454545, "grad_norm": 2.505784749984741, "learning_rate": 9.366354343245114e-06, "loss": 1.0906, "step": 8352 }, { "epoch": 0.6546238244514107, "grad_norm": 3.126659631729126, "learning_rate": 9.366204090139716e-06, "loss": 1.7305, "step": 8353 }, { "epoch": 0.6547021943573668, "grad_norm": 14.52745246887207, "learning_rate": 9.366053820427522e-06, "loss": 0.7276, "step": 8354 }, { "epoch": 0.6547805642633229, "grad_norm": 4.09540319442749, "learning_rate": 9.3659035341091e-06, "loss": 1.8015, "step": 8355 }, { "epoch": 0.654858934169279, "grad_norm": 3.3156440258026123, "learning_rate": 9.365753231185025e-06, "loss": 1.3269, "step": 8356 }, { "epoch": 0.6549373040752351, "grad_norm": 11.545677185058594, "learning_rate": 9.365602911655865e-06, "loss": 1.6621, "step": 8357 }, { "epoch": 0.6550156739811912, "grad_norm": 2.226961612701416, "learning_rate": 9.365452575522197e-06, "loss": 1.2548, "step": 8358 }, { "epoch": 0.6550940438871473, "grad_norm": 2.5826973915100098, "learning_rate": 9.365302222784585e-06, "loss": 1.371, "step": 8359 }, { "epoch": 0.6551724137931034, "grad_norm": 5.754989147186279, "learning_rate": 9.365151853443608e-06, "loss": 1.4038, "step": 8360 }, { "epoch": 0.6552507836990595, "grad_norm": 4.468254566192627, "learning_rate": 9.365001467499837e-06, "loss": 1.1899, "step": 8361 }, { "epoch": 0.6553291536050156, "grad_norm": 4.0136613845825195, "learning_rate": 9.364851064953839e-06, "loss": 1.4712, "step": 8362 }, { "epoch": 0.6554075235109718, "grad_norm": 3.508148193359375, "learning_rate": 9.364700645806193e-06, "loss": 1.8476, "step": 8363 }, { "epoch": 0.6554858934169279, "grad_norm": 8.377963066101074, "learning_rate": 9.364550210057466e-06, "loss": 1.2791, "step": 8364 }, { "epoch": 0.655564263322884, "grad_norm": 4.889313220977783, "learning_rate": 9.364399757708231e-06, "loss": 1.0684, "step": 8365 }, { "epoch": 0.6556426332288401, "grad_norm": 3.003964424133301, "learning_rate": 9.364249288759061e-06, "loss": 1.6404, "step": 8366 }, { "epoch": 0.6557210031347962, "grad_norm": 4.89055871963501, "learning_rate": 9.36409880321053e-06, "loss": 1.2341, "step": 8367 }, { "epoch": 0.6557993730407523, "grad_norm": 2.8943662643432617, "learning_rate": 9.363948301063209e-06, "loss": 1.3848, "step": 8368 }, { "epoch": 0.6558777429467084, "grad_norm": 4.057961463928223, "learning_rate": 9.363797782317667e-06, "loss": 1.1832, "step": 8369 }, { "epoch": 0.6559561128526645, "grad_norm": 4.726708889007568, "learning_rate": 9.363647246974482e-06, "loss": 1.0964, "step": 8370 }, { "epoch": 0.6560344827586206, "grad_norm": 2.2751145362854004, "learning_rate": 9.363496695034225e-06, "loss": 1.567, "step": 8371 }, { "epoch": 0.6561128526645768, "grad_norm": 3.5795774459838867, "learning_rate": 9.363346126497465e-06, "loss": 1.246, "step": 8372 }, { "epoch": 0.6561912225705329, "grad_norm": 3.097264528274536, "learning_rate": 9.363195541364781e-06, "loss": 1.1304, "step": 8373 }, { "epoch": 0.6562695924764891, "grad_norm": 3.1721391677856445, "learning_rate": 9.363044939636739e-06, "loss": 1.6079, "step": 8374 }, { "epoch": 0.6563479623824452, "grad_norm": 2.9443345069885254, "learning_rate": 9.362894321313915e-06, "loss": 1.4243, "step": 8375 }, { "epoch": 0.6564263322884013, "grad_norm": 2.726618528366089, "learning_rate": 9.362743686396885e-06, "loss": 1.5002, "step": 8376 }, { "epoch": 0.6565047021943574, "grad_norm": 3.1228737831115723, "learning_rate": 9.362593034886216e-06, "loss": 1.6711, "step": 8377 }, { "epoch": 0.6565830721003135, "grad_norm": 2.418172836303711, "learning_rate": 9.362442366782485e-06, "loss": 1.4264, "step": 8378 }, { "epoch": 0.6566614420062696, "grad_norm": 7.690290927886963, "learning_rate": 9.362291682086265e-06, "loss": 2.3141, "step": 8379 }, { "epoch": 0.6567398119122257, "grad_norm": 3.6644599437713623, "learning_rate": 9.362140980798127e-06, "loss": 1.6171, "step": 8380 }, { "epoch": 0.6568181818181819, "grad_norm": 3.27241849899292, "learning_rate": 9.361990262918644e-06, "loss": 1.4137, "step": 8381 }, { "epoch": 0.656896551724138, "grad_norm": 3.672262668609619, "learning_rate": 9.361839528448391e-06, "loss": 1.3955, "step": 8382 }, { "epoch": 0.6569749216300941, "grad_norm": 5.949549198150635, "learning_rate": 9.361688777387943e-06, "loss": 1.4693, "step": 8383 }, { "epoch": 0.6570532915360502, "grad_norm": 2.730048656463623, "learning_rate": 9.36153800973787e-06, "loss": 0.9199, "step": 8384 }, { "epoch": 0.6571316614420063, "grad_norm": 9.12465763092041, "learning_rate": 9.361387225498745e-06, "loss": 0.7222, "step": 8385 }, { "epoch": 0.6572100313479624, "grad_norm": 2.6141233444213867, "learning_rate": 9.361236424671145e-06, "loss": 1.3141, "step": 8386 }, { "epoch": 0.6572884012539185, "grad_norm": 2.0059006214141846, "learning_rate": 9.361085607255641e-06, "loss": 1.1288, "step": 8387 }, { "epoch": 0.6573667711598746, "grad_norm": 2.3697402477264404, "learning_rate": 9.360934773252808e-06, "loss": 1.0113, "step": 8388 }, { "epoch": 0.6574451410658307, "grad_norm": 6.523565769195557, "learning_rate": 9.360783922663218e-06, "loss": 0.5421, "step": 8389 }, { "epoch": 0.6575235109717869, "grad_norm": 2.146705150604248, "learning_rate": 9.360633055487446e-06, "loss": 1.6015, "step": 8390 }, { "epoch": 0.657601880877743, "grad_norm": 2.568525791168213, "learning_rate": 9.360482171726066e-06, "loss": 1.3515, "step": 8391 }, { "epoch": 0.6576802507836991, "grad_norm": 5.348034858703613, "learning_rate": 9.360331271379652e-06, "loss": 1.1043, "step": 8392 }, { "epoch": 0.6577586206896552, "grad_norm": 2.6552798748016357, "learning_rate": 9.360180354448778e-06, "loss": 1.2607, "step": 8393 }, { "epoch": 0.6578369905956113, "grad_norm": 3.63014554977417, "learning_rate": 9.360029420934015e-06, "loss": 1.2967, "step": 8394 }, { "epoch": 0.6579153605015674, "grad_norm": 5.477138042449951, "learning_rate": 9.359878470835942e-06, "loss": 1.0761, "step": 8395 }, { "epoch": 0.6579937304075235, "grad_norm": 2.678262233734131, "learning_rate": 9.35972750415513e-06, "loss": 1.447, "step": 8396 }, { "epoch": 0.6580721003134796, "grad_norm": 3.940037250518799, "learning_rate": 9.359576520892152e-06, "loss": 1.6295, "step": 8397 }, { "epoch": 0.6581504702194357, "grad_norm": 9.021618843078613, "learning_rate": 9.359425521047587e-06, "loss": 2.8369, "step": 8398 }, { "epoch": 0.6582288401253918, "grad_norm": 2.646075963973999, "learning_rate": 9.359274504622005e-06, "loss": 1.4326, "step": 8399 }, { "epoch": 0.658307210031348, "grad_norm": 2.6339476108551025, "learning_rate": 9.359123471615981e-06, "loss": 1.7595, "step": 8400 }, { "epoch": 0.6583855799373041, "grad_norm": 7.575621128082275, "learning_rate": 9.358972422030091e-06, "loss": 0.5605, "step": 8401 }, { "epoch": 0.6584639498432602, "grad_norm": 2.347541332244873, "learning_rate": 9.35882135586491e-06, "loss": 1.1456, "step": 8402 }, { "epoch": 0.6585423197492163, "grad_norm": 3.426440954208374, "learning_rate": 9.35867027312101e-06, "loss": 1.2606, "step": 8403 }, { "epoch": 0.6586206896551724, "grad_norm": 2.9324731826782227, "learning_rate": 9.358519173798967e-06, "loss": 1.2719, "step": 8404 }, { "epoch": 0.6586990595611285, "grad_norm": 5.487076759338379, "learning_rate": 9.358368057899356e-06, "loss": 1.905, "step": 8405 }, { "epoch": 0.6587774294670846, "grad_norm": 2.968963384628296, "learning_rate": 9.358216925422752e-06, "loss": 1.5478, "step": 8406 }, { "epoch": 0.6588557993730407, "grad_norm": 3.012453556060791, "learning_rate": 9.358065776369729e-06, "loss": 1.428, "step": 8407 }, { "epoch": 0.6589341692789968, "grad_norm": 3.322026014328003, "learning_rate": 9.357914610740864e-06, "loss": 1.0061, "step": 8408 }, { "epoch": 0.659012539184953, "grad_norm": 1.9721932411193848, "learning_rate": 9.357763428536729e-06, "loss": 1.0729, "step": 8409 }, { "epoch": 0.6590909090909091, "grad_norm": 2.4803056716918945, "learning_rate": 9.357612229757898e-06, "loss": 1.081, "step": 8410 }, { "epoch": 0.6591692789968652, "grad_norm": 6.58884859085083, "learning_rate": 9.357461014404949e-06, "loss": 1.7878, "step": 8411 }, { "epoch": 0.6592476489028213, "grad_norm": 2.527688503265381, "learning_rate": 9.357309782478458e-06, "loss": 1.0834, "step": 8412 }, { "epoch": 0.6593260188087774, "grad_norm": 9.492834091186523, "learning_rate": 9.357158533978998e-06, "loss": 1.5685, "step": 8413 }, { "epoch": 0.6594043887147335, "grad_norm": 2.1471445560455322, "learning_rate": 9.357007268907146e-06, "loss": 1.2951, "step": 8414 }, { "epoch": 0.6594827586206896, "grad_norm": 2.565983772277832, "learning_rate": 9.356855987263475e-06, "loss": 1.3653, "step": 8415 }, { "epoch": 0.6595611285266457, "grad_norm": 4.019832611083984, "learning_rate": 9.35670468904856e-06, "loss": 1.5904, "step": 8416 }, { "epoch": 0.6596394984326018, "grad_norm": 3.009471893310547, "learning_rate": 9.35655337426298e-06, "loss": 1.6036, "step": 8417 }, { "epoch": 0.659717868338558, "grad_norm": 5.770840167999268, "learning_rate": 9.35640204290731e-06, "loss": 1.1301, "step": 8418 }, { "epoch": 0.659796238244514, "grad_norm": 2.2123420238494873, "learning_rate": 9.356250694982123e-06, "loss": 1.4246, "step": 8419 }, { "epoch": 0.6598746081504702, "grad_norm": 4.665791034698486, "learning_rate": 9.356099330487996e-06, "loss": 1.268, "step": 8420 }, { "epoch": 0.6599529780564264, "grad_norm": 2.425510883331299, "learning_rate": 9.355947949425503e-06, "loss": 1.1536, "step": 8421 }, { "epoch": 0.6600313479623825, "grad_norm": 2.8088057041168213, "learning_rate": 9.355796551795224e-06, "loss": 1.0735, "step": 8422 }, { "epoch": 0.6601097178683386, "grad_norm": 4.957390785217285, "learning_rate": 9.355645137597731e-06, "loss": 1.2203, "step": 8423 }, { "epoch": 0.6601880877742947, "grad_norm": 6.36531400680542, "learning_rate": 9.3554937068336e-06, "loss": 1.5831, "step": 8424 }, { "epoch": 0.6602664576802508, "grad_norm": 2.1459755897521973, "learning_rate": 9.355342259503409e-06, "loss": 1.2901, "step": 8425 }, { "epoch": 0.6603448275862069, "grad_norm": 2.8460536003112793, "learning_rate": 9.355190795607733e-06, "loss": 1.3479, "step": 8426 }, { "epoch": 0.660423197492163, "grad_norm": 2.339200973510742, "learning_rate": 9.355039315147149e-06, "loss": 1.525, "step": 8427 }, { "epoch": 0.6605015673981192, "grad_norm": 2.842833995819092, "learning_rate": 9.35488781812223e-06, "loss": 1.6464, "step": 8428 }, { "epoch": 0.6605799373040753, "grad_norm": 3.5177931785583496, "learning_rate": 9.354736304533558e-06, "loss": 1.3996, "step": 8429 }, { "epoch": 0.6606583072100314, "grad_norm": 3.7003822326660156, "learning_rate": 9.354584774381703e-06, "loss": 1.3551, "step": 8430 }, { "epoch": 0.6607366771159875, "grad_norm": 2.749445676803589, "learning_rate": 9.354433227667245e-06, "loss": 1.7764, "step": 8431 }, { "epoch": 0.6608150470219436, "grad_norm": 2.5831868648529053, "learning_rate": 9.35428166439076e-06, "loss": 1.356, "step": 8432 }, { "epoch": 0.6608934169278997, "grad_norm": 2.049240827560425, "learning_rate": 9.354130084552824e-06, "loss": 0.9492, "step": 8433 }, { "epoch": 0.6609717868338558, "grad_norm": 3.1280510425567627, "learning_rate": 9.353978488154012e-06, "loss": 1.5312, "step": 8434 }, { "epoch": 0.6610501567398119, "grad_norm": 5.737578868865967, "learning_rate": 9.353826875194905e-06, "loss": 1.2327, "step": 8435 }, { "epoch": 0.661128526645768, "grad_norm": 2.3474268913269043, "learning_rate": 9.353675245676075e-06, "loss": 1.3078, "step": 8436 }, { "epoch": 0.6612068965517242, "grad_norm": 3.3614485263824463, "learning_rate": 9.353523599598102e-06, "loss": 1.5641, "step": 8437 }, { "epoch": 0.6612852664576803, "grad_norm": 2.2436912059783936, "learning_rate": 9.35337193696156e-06, "loss": 1.3174, "step": 8438 }, { "epoch": 0.6613636363636364, "grad_norm": 2.6038460731506348, "learning_rate": 9.353220257767027e-06, "loss": 1.4651, "step": 8439 }, { "epoch": 0.6614420062695925, "grad_norm": 4.3741068840026855, "learning_rate": 9.353068562015081e-06, "loss": 1.1989, "step": 8440 }, { "epoch": 0.6615203761755486, "grad_norm": 2.176457405090332, "learning_rate": 9.352916849706297e-06, "loss": 1.2891, "step": 8441 }, { "epoch": 0.6615987460815047, "grad_norm": 2.726820707321167, "learning_rate": 9.352765120841255e-06, "loss": 1.6614, "step": 8442 }, { "epoch": 0.6616771159874608, "grad_norm": 2.712697744369507, "learning_rate": 9.352613375420528e-06, "loss": 1.0953, "step": 8443 }, { "epoch": 0.6617554858934169, "grad_norm": 3.0595710277557373, "learning_rate": 9.352461613444695e-06, "loss": 1.3908, "step": 8444 }, { "epoch": 0.661833855799373, "grad_norm": 4.695217132568359, "learning_rate": 9.352309834914337e-06, "loss": 1.5593, "step": 8445 }, { "epoch": 0.6619122257053291, "grad_norm": 3.059584856033325, "learning_rate": 9.352158039830024e-06, "loss": 1.24, "step": 8446 }, { "epoch": 0.6619905956112853, "grad_norm": 12.950100898742676, "learning_rate": 9.352006228192339e-06, "loss": 1.0534, "step": 8447 }, { "epoch": 0.6620689655172414, "grad_norm": 3.587655544281006, "learning_rate": 9.351854400001858e-06, "loss": 1.4857, "step": 8448 }, { "epoch": 0.6621473354231975, "grad_norm": 2.689150810241699, "learning_rate": 9.351702555259158e-06, "loss": 1.5404, "step": 8449 }, { "epoch": 0.6622257053291536, "grad_norm": 9.097716331481934, "learning_rate": 9.351550693964816e-06, "loss": 0.8767, "step": 8450 }, { "epoch": 0.6623040752351097, "grad_norm": 3.5737435817718506, "learning_rate": 9.35139881611941e-06, "loss": 1.2636, "step": 8451 }, { "epoch": 0.6623824451410658, "grad_norm": 3.438122272491455, "learning_rate": 9.35124692172352e-06, "loss": 1.571, "step": 8452 }, { "epoch": 0.6624608150470219, "grad_norm": 12.639951705932617, "learning_rate": 9.35109501077772e-06, "loss": 2.115, "step": 8453 }, { "epoch": 0.662539184952978, "grad_norm": 3.218979835510254, "learning_rate": 9.350943083282588e-06, "loss": 1.076, "step": 8454 }, { "epoch": 0.6626175548589341, "grad_norm": 4.986990928649902, "learning_rate": 9.350791139238707e-06, "loss": 1.8287, "step": 8455 }, { "epoch": 0.6626959247648903, "grad_norm": 2.8026740550994873, "learning_rate": 9.350639178646648e-06, "loss": 1.39, "step": 8456 }, { "epoch": 0.6627742946708464, "grad_norm": 4.853438377380371, "learning_rate": 9.350487201506994e-06, "loss": 2.1907, "step": 8457 }, { "epoch": 0.6628526645768025, "grad_norm": 1.9669874906539917, "learning_rate": 9.35033520782032e-06, "loss": 1.2612, "step": 8458 }, { "epoch": 0.6629310344827586, "grad_norm": 2.49918532371521, "learning_rate": 9.350183197587205e-06, "loss": 1.7251, "step": 8459 }, { "epoch": 0.6630094043887147, "grad_norm": 3.7129390239715576, "learning_rate": 9.350031170808229e-06, "loss": 1.3282, "step": 8460 }, { "epoch": 0.6630877742946708, "grad_norm": 3.9611382484436035, "learning_rate": 9.34987912748397e-06, "loss": 1.482, "step": 8461 }, { "epoch": 0.6631661442006269, "grad_norm": 2.446866273880005, "learning_rate": 9.349727067615001e-06, "loss": 1.4784, "step": 8462 }, { "epoch": 0.663244514106583, "grad_norm": 3.968104839324951, "learning_rate": 9.349574991201909e-06, "loss": 1.4011, "step": 8463 }, { "epoch": 0.6633228840125391, "grad_norm": 9.445345878601074, "learning_rate": 9.349422898245263e-06, "loss": 1.285, "step": 8464 }, { "epoch": 0.6634012539184952, "grad_norm": 2.4264252185821533, "learning_rate": 9.34927078874565e-06, "loss": 1.4943, "step": 8465 }, { "epoch": 0.6634796238244514, "grad_norm": 3.0381178855895996, "learning_rate": 9.349118662703644e-06, "loss": 1.3591, "step": 8466 }, { "epoch": 0.6635579937304075, "grad_norm": 2.293428659439087, "learning_rate": 9.348966520119824e-06, "loss": 1.6604, "step": 8467 }, { "epoch": 0.6636363636363637, "grad_norm": 2.0199472904205322, "learning_rate": 9.348814360994769e-06, "loss": 1.4379, "step": 8468 }, { "epoch": 0.6637147335423198, "grad_norm": 2.9354171752929688, "learning_rate": 9.348662185329059e-06, "loss": 1.5179, "step": 8469 }, { "epoch": 0.6637931034482759, "grad_norm": 3.5614635944366455, "learning_rate": 9.34850999312327e-06, "loss": 1.2413, "step": 8470 }, { "epoch": 0.663871473354232, "grad_norm": 2.9326534271240234, "learning_rate": 9.348357784377982e-06, "loss": 1.3476, "step": 8471 }, { "epoch": 0.6639498432601881, "grad_norm": 4.458063125610352, "learning_rate": 9.348205559093777e-06, "loss": 1.4166, "step": 8472 }, { "epoch": 0.6640282131661442, "grad_norm": 2.2652108669281006, "learning_rate": 9.34805331727123e-06, "loss": 1.5913, "step": 8473 }, { "epoch": 0.6641065830721004, "grad_norm": 4.852494239807129, "learning_rate": 9.347901058910921e-06, "loss": 1.2467, "step": 8474 }, { "epoch": 0.6641849529780565, "grad_norm": 2.5281178951263428, "learning_rate": 9.34774878401343e-06, "loss": 1.7751, "step": 8475 }, { "epoch": 0.6642633228840126, "grad_norm": 3.571298599243164, "learning_rate": 9.347596492579336e-06, "loss": 1.0709, "step": 8476 }, { "epoch": 0.6643416927899687, "grad_norm": 3.213083267211914, "learning_rate": 9.347444184609218e-06, "loss": 1.6601, "step": 8477 }, { "epoch": 0.6644200626959248, "grad_norm": 4.020646572113037, "learning_rate": 9.347291860103655e-06, "loss": 0.9766, "step": 8478 }, { "epoch": 0.6644984326018809, "grad_norm": 2.4038784503936768, "learning_rate": 9.347139519063227e-06, "loss": 1.6778, "step": 8479 }, { "epoch": 0.664576802507837, "grad_norm": 3.236438274383545, "learning_rate": 9.346987161488513e-06, "loss": 1.4288, "step": 8480 }, { "epoch": 0.6646551724137931, "grad_norm": 3.147237539291382, "learning_rate": 9.346834787380093e-06, "loss": 1.3405, "step": 8481 }, { "epoch": 0.6647335423197492, "grad_norm": 3.1354141235351562, "learning_rate": 9.346682396738546e-06, "loss": 1.7671, "step": 8482 }, { "epoch": 0.6648119122257053, "grad_norm": 2.582538604736328, "learning_rate": 9.34652998956445e-06, "loss": 1.3246, "step": 8483 }, { "epoch": 0.6648902821316615, "grad_norm": 2.480774402618408, "learning_rate": 9.346377565858388e-06, "loss": 1.3129, "step": 8484 }, { "epoch": 0.6649686520376176, "grad_norm": 4.753061771392822, "learning_rate": 9.34622512562094e-06, "loss": 1.701, "step": 8485 }, { "epoch": 0.6650470219435737, "grad_norm": 2.6089863777160645, "learning_rate": 9.34607266885268e-06, "loss": 1.3939, "step": 8486 }, { "epoch": 0.6651253918495298, "grad_norm": 7.411123275756836, "learning_rate": 9.345920195554196e-06, "loss": 1.1745, "step": 8487 }, { "epoch": 0.6652037617554859, "grad_norm": 4.8383026123046875, "learning_rate": 9.34576770572606e-06, "loss": 1.6875, "step": 8488 }, { "epoch": 0.665282131661442, "grad_norm": 2.8215887546539307, "learning_rate": 9.345615199368859e-06, "loss": 1.5258, "step": 8489 }, { "epoch": 0.6653605015673981, "grad_norm": 2.139343500137329, "learning_rate": 9.345462676483168e-06, "loss": 1.3577, "step": 8490 }, { "epoch": 0.6654388714733542, "grad_norm": 19.60417366027832, "learning_rate": 9.34531013706957e-06, "loss": 2.9549, "step": 8491 }, { "epoch": 0.6655172413793103, "grad_norm": 7.086748123168945, "learning_rate": 9.345157581128643e-06, "loss": 1.3373, "step": 8492 }, { "epoch": 0.6655956112852665, "grad_norm": 3.094994068145752, "learning_rate": 9.34500500866097e-06, "loss": 1.7169, "step": 8493 }, { "epoch": 0.6656739811912226, "grad_norm": 7.328402042388916, "learning_rate": 9.344852419667127e-06, "loss": 1.7352, "step": 8494 }, { "epoch": 0.6657523510971787, "grad_norm": 3.668408155441284, "learning_rate": 9.344699814147699e-06, "loss": 1.2779, "step": 8495 }, { "epoch": 0.6658307210031348, "grad_norm": 3.123265027999878, "learning_rate": 9.344547192103264e-06, "loss": 1.4266, "step": 8496 }, { "epoch": 0.6659090909090909, "grad_norm": 2.798854351043701, "learning_rate": 9.344394553534403e-06, "loss": 1.2858, "step": 8497 }, { "epoch": 0.665987460815047, "grad_norm": 3.6973652839660645, "learning_rate": 9.344241898441698e-06, "loss": 1.3198, "step": 8498 }, { "epoch": 0.6660658307210031, "grad_norm": 2.273134708404541, "learning_rate": 9.344089226825726e-06, "loss": 1.3217, "step": 8499 }, { "epoch": 0.6661442006269592, "grad_norm": 3.101087808609009, "learning_rate": 9.343936538687071e-06, "loss": 1.2056, "step": 8500 }, { "epoch": 0.6662225705329153, "grad_norm": 2.4616355895996094, "learning_rate": 9.343783834026311e-06, "loss": 0.8704, "step": 8501 }, { "epoch": 0.6663009404388714, "grad_norm": 4.187252998352051, "learning_rate": 9.34363111284403e-06, "loss": 1.3947, "step": 8502 }, { "epoch": 0.6663793103448276, "grad_norm": 2.5463309288024902, "learning_rate": 9.343478375140807e-06, "loss": 1.0422, "step": 8503 }, { "epoch": 0.6664576802507837, "grad_norm": 2.130005121231079, "learning_rate": 9.343325620917223e-06, "loss": 1.1987, "step": 8504 }, { "epoch": 0.6665360501567398, "grad_norm": 3.142510414123535, "learning_rate": 9.343172850173858e-06, "loss": 1.1872, "step": 8505 }, { "epoch": 0.6666144200626959, "grad_norm": 2.711332321166992, "learning_rate": 9.343020062911297e-06, "loss": 1.2385, "step": 8506 }, { "epoch": 0.666692789968652, "grad_norm": 3.2224645614624023, "learning_rate": 9.342867259130118e-06, "loss": 1.454, "step": 8507 }, { "epoch": 0.6667711598746081, "grad_norm": 4.0214385986328125, "learning_rate": 9.3427144388309e-06, "loss": 1.1518, "step": 8508 }, { "epoch": 0.6668495297805642, "grad_norm": 2.138873338699341, "learning_rate": 9.342561602014228e-06, "loss": 1.1746, "step": 8509 }, { "epoch": 0.6669278996865203, "grad_norm": 2.6518445014953613, "learning_rate": 9.342408748680683e-06, "loss": 1.8417, "step": 8510 }, { "epoch": 0.6670062695924764, "grad_norm": 3.626281261444092, "learning_rate": 9.342255878830844e-06, "loss": 1.081, "step": 8511 }, { "epoch": 0.6670846394984326, "grad_norm": 2.4288039207458496, "learning_rate": 9.342102992465296e-06, "loss": 1.0861, "step": 8512 }, { "epoch": 0.6671630094043887, "grad_norm": 2.8860270977020264, "learning_rate": 9.341950089584616e-06, "loss": 1.4339, "step": 8513 }, { "epoch": 0.6672413793103448, "grad_norm": 2.585181474685669, "learning_rate": 9.341797170189391e-06, "loss": 1.4475, "step": 8514 }, { "epoch": 0.667319749216301, "grad_norm": 2.728442907333374, "learning_rate": 9.341644234280198e-06, "loss": 1.3662, "step": 8515 }, { "epoch": 0.6673981191222571, "grad_norm": 4.136524200439453, "learning_rate": 9.34149128185762e-06, "loss": 1.3788, "step": 8516 }, { "epoch": 0.6674764890282132, "grad_norm": 2.516854763031006, "learning_rate": 9.34133831292224e-06, "loss": 1.3402, "step": 8517 }, { "epoch": 0.6675548589341693, "grad_norm": 6.07248067855835, "learning_rate": 9.341185327474639e-06, "loss": 1.1639, "step": 8518 }, { "epoch": 0.6676332288401254, "grad_norm": 2.4288454055786133, "learning_rate": 9.341032325515399e-06, "loss": 1.6506, "step": 8519 }, { "epoch": 0.6677115987460815, "grad_norm": 3.418391704559326, "learning_rate": 9.3408793070451e-06, "loss": 1.5274, "step": 8520 }, { "epoch": 0.6677899686520377, "grad_norm": 5.543583393096924, "learning_rate": 9.340726272064326e-06, "loss": 1.2634, "step": 8521 }, { "epoch": 0.6678683385579938, "grad_norm": 3.8418726921081543, "learning_rate": 9.34057322057366e-06, "loss": 1.4063, "step": 8522 }, { "epoch": 0.6679467084639499, "grad_norm": 2.8189573287963867, "learning_rate": 9.340420152573683e-06, "loss": 1.3937, "step": 8523 }, { "epoch": 0.668025078369906, "grad_norm": 2.657637357711792, "learning_rate": 9.340267068064976e-06, "loss": 1.7728, "step": 8524 }, { "epoch": 0.6681034482758621, "grad_norm": 8.935991287231445, "learning_rate": 9.340113967048123e-06, "loss": 1.4502, "step": 8525 }, { "epoch": 0.6681818181818182, "grad_norm": 3.2650089263916016, "learning_rate": 9.339960849523706e-06, "loss": 1.4775, "step": 8526 }, { "epoch": 0.6682601880877743, "grad_norm": 2.849839210510254, "learning_rate": 9.339807715492306e-06, "loss": 1.7724, "step": 8527 }, { "epoch": 0.6683385579937304, "grad_norm": 3.247316598892212, "learning_rate": 9.339654564954508e-06, "loss": 1.0473, "step": 8528 }, { "epoch": 0.6684169278996865, "grad_norm": 3.4829294681549072, "learning_rate": 9.339501397910892e-06, "loss": 1.6095, "step": 8529 }, { "epoch": 0.6684952978056427, "grad_norm": 4.032830238342285, "learning_rate": 9.339348214362042e-06, "loss": 1.4813, "step": 8530 }, { "epoch": 0.6685736677115988, "grad_norm": 2.597313642501831, "learning_rate": 9.33919501430854e-06, "loss": 1.3518, "step": 8531 }, { "epoch": 0.6686520376175549, "grad_norm": 2.3777215480804443, "learning_rate": 9.33904179775097e-06, "loss": 1.3651, "step": 8532 }, { "epoch": 0.668730407523511, "grad_norm": 4.487381935119629, "learning_rate": 9.338888564689912e-06, "loss": 1.72, "step": 8533 }, { "epoch": 0.6688087774294671, "grad_norm": 3.0062806606292725, "learning_rate": 9.338735315125952e-06, "loss": 1.5257, "step": 8534 }, { "epoch": 0.6688871473354232, "grad_norm": 5.111037254333496, "learning_rate": 9.33858204905967e-06, "loss": 1.5522, "step": 8535 }, { "epoch": 0.6689655172413793, "grad_norm": 2.313988447189331, "learning_rate": 9.338428766491649e-06, "loss": 1.2279, "step": 8536 }, { "epoch": 0.6690438871473354, "grad_norm": 2.3585634231567383, "learning_rate": 9.338275467422476e-06, "loss": 1.5203, "step": 8537 }, { "epoch": 0.6691222570532915, "grad_norm": 3.2725439071655273, "learning_rate": 9.338122151852731e-06, "loss": 1.4257, "step": 8538 }, { "epoch": 0.6692006269592476, "grad_norm": 2.61238694190979, "learning_rate": 9.337968819782996e-06, "loss": 1.0277, "step": 8539 }, { "epoch": 0.6692789968652038, "grad_norm": 2.0890133380889893, "learning_rate": 9.337815471213858e-06, "loss": 1.368, "step": 8540 }, { "epoch": 0.6693573667711599, "grad_norm": 2.2989771366119385, "learning_rate": 9.337662106145898e-06, "loss": 1.5043, "step": 8541 }, { "epoch": 0.669435736677116, "grad_norm": 3.939110279083252, "learning_rate": 9.337508724579696e-06, "loss": 0.8778, "step": 8542 }, { "epoch": 0.6695141065830721, "grad_norm": 2.017390012741089, "learning_rate": 9.337355326515843e-06, "loss": 1.1574, "step": 8543 }, { "epoch": 0.6695924764890282, "grad_norm": 7.23162841796875, "learning_rate": 9.337201911954914e-06, "loss": 1.7045, "step": 8544 }, { "epoch": 0.6696708463949843, "grad_norm": 3.156395196914673, "learning_rate": 9.3370484808975e-06, "loss": 1.1873, "step": 8545 }, { "epoch": 0.6697492163009404, "grad_norm": 5.340384006500244, "learning_rate": 9.33689503334418e-06, "loss": 1.4316, "step": 8546 }, { "epoch": 0.6698275862068965, "grad_norm": 2.0328493118286133, "learning_rate": 9.336741569295537e-06, "loss": 1.4641, "step": 8547 }, { "epoch": 0.6699059561128526, "grad_norm": 2.9499969482421875, "learning_rate": 9.336588088752158e-06, "loss": 1.6475, "step": 8548 }, { "epoch": 0.6699843260188088, "grad_norm": 3.681335926055908, "learning_rate": 9.336434591714627e-06, "loss": 1.847, "step": 8549 }, { "epoch": 0.6700626959247649, "grad_norm": 9.109598159790039, "learning_rate": 9.336281078183525e-06, "loss": 1.09, "step": 8550 }, { "epoch": 0.670141065830721, "grad_norm": 1.974523663520813, "learning_rate": 9.336127548159438e-06, "loss": 1.3177, "step": 8551 }, { "epoch": 0.6702194357366771, "grad_norm": 8.123501777648926, "learning_rate": 9.335974001642946e-06, "loss": 1.4096, "step": 8552 }, { "epoch": 0.6702978056426332, "grad_norm": 5.319675445556641, "learning_rate": 9.335820438634639e-06, "loss": 1.4376, "step": 8553 }, { "epoch": 0.6703761755485893, "grad_norm": 4.326093673706055, "learning_rate": 9.335666859135096e-06, "loss": 1.3971, "step": 8554 }, { "epoch": 0.6704545454545454, "grad_norm": 3.3832528591156006, "learning_rate": 9.335513263144905e-06, "loss": 1.2725, "step": 8555 }, { "epoch": 0.6705329153605015, "grad_norm": 3.153552770614624, "learning_rate": 9.335359650664647e-06, "loss": 1.3313, "step": 8556 }, { "epoch": 0.6706112852664576, "grad_norm": 2.761436700820923, "learning_rate": 9.335206021694907e-06, "loss": 1.2754, "step": 8557 }, { "epoch": 0.6706896551724137, "grad_norm": 2.3179872035980225, "learning_rate": 9.33505237623627e-06, "loss": 1.4945, "step": 8558 }, { "epoch": 0.6707680250783699, "grad_norm": 2.9731709957122803, "learning_rate": 9.334898714289324e-06, "loss": 0.9393, "step": 8559 }, { "epoch": 0.670846394984326, "grad_norm": 2.878469228744507, "learning_rate": 9.334745035854646e-06, "loss": 1.2864, "step": 8560 }, { "epoch": 0.6709247648902821, "grad_norm": 4.872025966644287, "learning_rate": 9.334591340932825e-06, "loss": 1.3561, "step": 8561 }, { "epoch": 0.6710031347962383, "grad_norm": 3.363492012023926, "learning_rate": 9.334437629524447e-06, "loss": 1.4508, "step": 8562 }, { "epoch": 0.6710815047021944, "grad_norm": 2.662184476852417, "learning_rate": 9.334283901630094e-06, "loss": 1.2662, "step": 8563 }, { "epoch": 0.6711598746081505, "grad_norm": 4.416682243347168, "learning_rate": 9.33413015725035e-06, "loss": 0.8968, "step": 8564 }, { "epoch": 0.6712382445141066, "grad_norm": 2.483684778213501, "learning_rate": 9.333976396385802e-06, "loss": 1.4099, "step": 8565 }, { "epoch": 0.6713166144200627, "grad_norm": 5.576457977294922, "learning_rate": 9.333822619037032e-06, "loss": 1.4427, "step": 8566 }, { "epoch": 0.6713949843260189, "grad_norm": 2.23557186126709, "learning_rate": 9.33366882520463e-06, "loss": 1.2903, "step": 8567 }, { "epoch": 0.671473354231975, "grad_norm": 4.650326251983643, "learning_rate": 9.333515014889175e-06, "loss": 1.5423, "step": 8568 }, { "epoch": 0.6715517241379311, "grad_norm": 3.883054256439209, "learning_rate": 9.333361188091255e-06, "loss": 1.6961, "step": 8569 }, { "epoch": 0.6716300940438872, "grad_norm": 3.5309548377990723, "learning_rate": 9.333207344811454e-06, "loss": 1.2671, "step": 8570 }, { "epoch": 0.6717084639498433, "grad_norm": 3.8850786685943604, "learning_rate": 9.333053485050361e-06, "loss": 1.4067, "step": 8571 }, { "epoch": 0.6717868338557994, "grad_norm": 4.607125282287598, "learning_rate": 9.332899608808555e-06, "loss": 1.6652, "step": 8572 }, { "epoch": 0.6718652037617555, "grad_norm": 1.808226466178894, "learning_rate": 9.332745716086627e-06, "loss": 1.4394, "step": 8573 }, { "epoch": 0.6719435736677116, "grad_norm": 4.3717217445373535, "learning_rate": 9.332591806885159e-06, "loss": 1.0761, "step": 8574 }, { "epoch": 0.6720219435736677, "grad_norm": 3.377617120742798, "learning_rate": 9.332437881204737e-06, "loss": 1.5493, "step": 8575 }, { "epoch": 0.6721003134796238, "grad_norm": 2.6670289039611816, "learning_rate": 9.332283939045945e-06, "loss": 1.5855, "step": 8576 }, { "epoch": 0.67217868338558, "grad_norm": 2.560600996017456, "learning_rate": 9.332129980409372e-06, "loss": 1.3855, "step": 8577 }, { "epoch": 0.6722570532915361, "grad_norm": 2.450747013092041, "learning_rate": 9.3319760052956e-06, "loss": 1.4915, "step": 8578 }, { "epoch": 0.6723354231974922, "grad_norm": 2.3322207927703857, "learning_rate": 9.331822013705218e-06, "loss": 1.3157, "step": 8579 }, { "epoch": 0.6724137931034483, "grad_norm": 2.3538074493408203, "learning_rate": 9.33166800563881e-06, "loss": 1.377, "step": 8580 }, { "epoch": 0.6724921630094044, "grad_norm": 2.720764398574829, "learning_rate": 9.331513981096961e-06, "loss": 1.0248, "step": 8581 }, { "epoch": 0.6725705329153605, "grad_norm": 1.9431260824203491, "learning_rate": 9.331359940080258e-06, "loss": 1.5145, "step": 8582 }, { "epoch": 0.6726489028213166, "grad_norm": 3.013847589492798, "learning_rate": 9.331205882589285e-06, "loss": 0.9093, "step": 8583 }, { "epoch": 0.6727272727272727, "grad_norm": 2.0883290767669678, "learning_rate": 9.331051808624632e-06, "loss": 1.3994, "step": 8584 }, { "epoch": 0.6728056426332288, "grad_norm": 6.292351722717285, "learning_rate": 9.33089771818688e-06, "loss": 1.4387, "step": 8585 }, { "epoch": 0.672884012539185, "grad_norm": 2.5481948852539062, "learning_rate": 9.33074361127662e-06, "loss": 1.3783, "step": 8586 }, { "epoch": 0.6729623824451411, "grad_norm": 2.0405654907226562, "learning_rate": 9.330589487894433e-06, "loss": 1.495, "step": 8587 }, { "epoch": 0.6730407523510972, "grad_norm": 8.417069435119629, "learning_rate": 9.33043534804091e-06, "loss": 1.3345, "step": 8588 }, { "epoch": 0.6731191222570533, "grad_norm": 2.386914014816284, "learning_rate": 9.330281191716635e-06, "loss": 1.3912, "step": 8589 }, { "epoch": 0.6731974921630094, "grad_norm": 3.1953392028808594, "learning_rate": 9.330127018922195e-06, "loss": 1.1248, "step": 8590 }, { "epoch": 0.6732758620689655, "grad_norm": 2.1437063217163086, "learning_rate": 9.329972829658174e-06, "loss": 1.2882, "step": 8591 }, { "epoch": 0.6733542319749216, "grad_norm": 3.6002309322357178, "learning_rate": 9.329818623925163e-06, "loss": 1.3071, "step": 8592 }, { "epoch": 0.6734326018808777, "grad_norm": 2.0596702098846436, "learning_rate": 9.329664401723742e-06, "loss": 1.3089, "step": 8593 }, { "epoch": 0.6735109717868338, "grad_norm": 5.944042205810547, "learning_rate": 9.329510163054505e-06, "loss": 1.1833, "step": 8594 }, { "epoch": 0.67358934169279, "grad_norm": 1.9288463592529297, "learning_rate": 9.329355907918034e-06, "loss": 1.4353, "step": 8595 }, { "epoch": 0.673667711598746, "grad_norm": 2.626322031021118, "learning_rate": 9.329201636314916e-06, "loss": 1.626, "step": 8596 }, { "epoch": 0.6737460815047022, "grad_norm": 2.9985952377319336, "learning_rate": 9.32904734824574e-06, "loss": 1.4229, "step": 8597 }, { "epoch": 0.6738244514106583, "grad_norm": 3.282703161239624, "learning_rate": 9.32889304371109e-06, "loss": 1.7797, "step": 8598 }, { "epoch": 0.6739028213166144, "grad_norm": 6.1988043785095215, "learning_rate": 9.328738722711555e-06, "loss": 1.0376, "step": 8599 }, { "epoch": 0.6739811912225705, "grad_norm": 5.329292297363281, "learning_rate": 9.328584385247722e-06, "loss": 1.8629, "step": 8600 }, { "epoch": 0.6740595611285266, "grad_norm": 2.8055126667022705, "learning_rate": 9.328430031320177e-06, "loss": 1.5339, "step": 8601 }, { "epoch": 0.6741379310344827, "grad_norm": 3.18129825592041, "learning_rate": 9.328275660929507e-06, "loss": 1.2804, "step": 8602 }, { "epoch": 0.6742163009404388, "grad_norm": 2.598649740219116, "learning_rate": 9.3281212740763e-06, "loss": 1.4063, "step": 8603 }, { "epoch": 0.6742946708463949, "grad_norm": 5.110921859741211, "learning_rate": 9.327966870761142e-06, "loss": 1.7245, "step": 8604 }, { "epoch": 0.674373040752351, "grad_norm": 10.854738235473633, "learning_rate": 9.32781245098462e-06, "loss": 0.6868, "step": 8605 }, { "epoch": 0.6744514106583072, "grad_norm": 3.7680983543395996, "learning_rate": 9.327658014747325e-06, "loss": 1.2881, "step": 8606 }, { "epoch": 0.6745297805642633, "grad_norm": 13.6677827835083, "learning_rate": 9.32750356204984e-06, "loss": 0.8395, "step": 8607 }, { "epoch": 0.6746081504702194, "grad_norm": 2.042217493057251, "learning_rate": 9.327349092892755e-06, "loss": 1.4695, "step": 8608 }, { "epoch": 0.6746865203761756, "grad_norm": 3.1446802616119385, "learning_rate": 9.327194607276656e-06, "loss": 1.2706, "step": 8609 }, { "epoch": 0.6747648902821317, "grad_norm": 2.1068944931030273, "learning_rate": 9.327040105202132e-06, "loss": 1.4479, "step": 8610 }, { "epoch": 0.6748432601880878, "grad_norm": 2.500649929046631, "learning_rate": 9.326885586669769e-06, "loss": 1.0061, "step": 8611 }, { "epoch": 0.6749216300940439, "grad_norm": 2.3304555416107178, "learning_rate": 9.326731051680157e-06, "loss": 1.2557, "step": 8612 }, { "epoch": 0.675, "grad_norm": 2.354114294052124, "learning_rate": 9.326576500233882e-06, "loss": 1.3456, "step": 8613 }, { "epoch": 0.6750783699059562, "grad_norm": 2.091109275817871, "learning_rate": 9.32642193233153e-06, "loss": 1.052, "step": 8614 }, { "epoch": 0.6751567398119123, "grad_norm": 2.9252195358276367, "learning_rate": 9.326267347973695e-06, "loss": 1.5311, "step": 8615 }, { "epoch": 0.6752351097178684, "grad_norm": 2.9840855598449707, "learning_rate": 9.326112747160958e-06, "loss": 1.0054, "step": 8616 }, { "epoch": 0.6753134796238245, "grad_norm": 3.4181604385375977, "learning_rate": 9.325958129893912e-06, "loss": 1.3201, "step": 8617 }, { "epoch": 0.6753918495297806, "grad_norm": 2.2763688564300537, "learning_rate": 9.325803496173142e-06, "loss": 1.4307, "step": 8618 }, { "epoch": 0.6754702194357367, "grad_norm": 4.602241516113281, "learning_rate": 9.325648845999237e-06, "loss": 1.8365, "step": 8619 }, { "epoch": 0.6755485893416928, "grad_norm": 2.907790422439575, "learning_rate": 9.325494179372787e-06, "loss": 1.6122, "step": 8620 }, { "epoch": 0.6756269592476489, "grad_norm": 7.773068904876709, "learning_rate": 9.325339496294378e-06, "loss": 0.6884, "step": 8621 }, { "epoch": 0.675705329153605, "grad_norm": 5.483340740203857, "learning_rate": 9.325184796764601e-06, "loss": 1.3029, "step": 8622 }, { "epoch": 0.6757836990595611, "grad_norm": 2.363452434539795, "learning_rate": 9.325030080784041e-06, "loss": 1.4101, "step": 8623 }, { "epoch": 0.6758620689655173, "grad_norm": 3.4250664710998535, "learning_rate": 9.324875348353288e-06, "loss": 1.4846, "step": 8624 }, { "epoch": 0.6759404388714734, "grad_norm": 3.743901491165161, "learning_rate": 9.32472059947293e-06, "loss": 1.5819, "step": 8625 }, { "epoch": 0.6760188087774295, "grad_norm": 2.5258634090423584, "learning_rate": 9.324565834143557e-06, "loss": 1.1765, "step": 8626 }, { "epoch": 0.6760971786833856, "grad_norm": 2.140353202819824, "learning_rate": 9.324411052365757e-06, "loss": 1.8071, "step": 8627 }, { "epoch": 0.6761755485893417, "grad_norm": 4.023594379425049, "learning_rate": 9.324256254140118e-06, "loss": 1.3781, "step": 8628 }, { "epoch": 0.6762539184952978, "grad_norm": 3.539212942123413, "learning_rate": 9.324101439467231e-06, "loss": 1.8196, "step": 8629 }, { "epoch": 0.6763322884012539, "grad_norm": 3.588451623916626, "learning_rate": 9.32394660834768e-06, "loss": 0.939, "step": 8630 }, { "epoch": 0.67641065830721, "grad_norm": 2.200963258743286, "learning_rate": 9.32379176078206e-06, "loss": 1.1249, "step": 8631 }, { "epoch": 0.6764890282131661, "grad_norm": 3.1664981842041016, "learning_rate": 9.323636896770956e-06, "loss": 1.7803, "step": 8632 }, { "epoch": 0.6765673981191223, "grad_norm": 3.037410259246826, "learning_rate": 9.323482016314956e-06, "loss": 1.2676, "step": 8633 }, { "epoch": 0.6766457680250784, "grad_norm": 2.305478096008301, "learning_rate": 9.323327119414654e-06, "loss": 1.492, "step": 8634 }, { "epoch": 0.6767241379310345, "grad_norm": 3.2507243156433105, "learning_rate": 9.323172206070635e-06, "loss": 1.5634, "step": 8635 }, { "epoch": 0.6768025078369906, "grad_norm": 3.8536956310272217, "learning_rate": 9.323017276283489e-06, "loss": 1.4671, "step": 8636 }, { "epoch": 0.6768808777429467, "grad_norm": 5.494091033935547, "learning_rate": 9.322862330053806e-06, "loss": 1.9176, "step": 8637 }, { "epoch": 0.6769592476489028, "grad_norm": 10.893982887268066, "learning_rate": 9.322707367382176e-06, "loss": 1.3147, "step": 8638 }, { "epoch": 0.6770376175548589, "grad_norm": 2.459627151489258, "learning_rate": 9.322552388269186e-06, "loss": 1.3528, "step": 8639 }, { "epoch": 0.677115987460815, "grad_norm": 2.5933310985565186, "learning_rate": 9.322397392715427e-06, "loss": 1.4771, "step": 8640 }, { "epoch": 0.6771943573667711, "grad_norm": 3.0048959255218506, "learning_rate": 9.32224238072149e-06, "loss": 1.2154, "step": 8641 }, { "epoch": 0.6772727272727272, "grad_norm": 7.768278121948242, "learning_rate": 9.322087352287961e-06, "loss": 1.8437, "step": 8642 }, { "epoch": 0.6773510971786834, "grad_norm": 1.9809097051620483, "learning_rate": 9.321932307415433e-06, "loss": 1.2517, "step": 8643 }, { "epoch": 0.6774294670846395, "grad_norm": 3.4758129119873047, "learning_rate": 9.321777246104496e-06, "loss": 1.6388, "step": 8644 }, { "epoch": 0.6775078369905956, "grad_norm": 3.5661723613739014, "learning_rate": 9.321622168355736e-06, "loss": 1.5628, "step": 8645 }, { "epoch": 0.6775862068965517, "grad_norm": 12.188340187072754, "learning_rate": 9.321467074169744e-06, "loss": 2.5749, "step": 8646 }, { "epoch": 0.6776645768025078, "grad_norm": 1.9430253505706787, "learning_rate": 9.321311963547113e-06, "loss": 1.1967, "step": 8647 }, { "epoch": 0.6777429467084639, "grad_norm": 5.775254726409912, "learning_rate": 9.32115683648843e-06, "loss": 1.4089, "step": 8648 }, { "epoch": 0.67782131661442, "grad_norm": 4.4414520263671875, "learning_rate": 9.321001692994285e-06, "loss": 1.3935, "step": 8649 }, { "epoch": 0.6778996865203761, "grad_norm": 2.8822641372680664, "learning_rate": 9.32084653306527e-06, "loss": 1.648, "step": 8650 }, { "epoch": 0.6779780564263322, "grad_norm": 1.8492724895477295, "learning_rate": 9.320691356701973e-06, "loss": 1.4262, "step": 8651 }, { "epoch": 0.6780564263322884, "grad_norm": 6.031779766082764, "learning_rate": 9.320536163904987e-06, "loss": 1.1974, "step": 8652 }, { "epoch": 0.6781347962382445, "grad_norm": 6.525443077087402, "learning_rate": 9.3203809546749e-06, "loss": 1.5291, "step": 8653 }, { "epoch": 0.6782131661442006, "grad_norm": 2.1048319339752197, "learning_rate": 9.320225729012304e-06, "loss": 1.2283, "step": 8654 }, { "epoch": 0.6782915360501567, "grad_norm": 3.753019094467163, "learning_rate": 9.320070486917788e-06, "loss": 1.6057, "step": 8655 }, { "epoch": 0.6783699059561129, "grad_norm": 3.461894989013672, "learning_rate": 9.319915228391941e-06, "loss": 1.5822, "step": 8656 }, { "epoch": 0.678448275862069, "grad_norm": 2.475670337677002, "learning_rate": 9.319759953435357e-06, "loss": 1.5037, "step": 8657 }, { "epoch": 0.6785266457680251, "grad_norm": 2.5552940368652344, "learning_rate": 9.319604662048623e-06, "loss": 1.5193, "step": 8658 }, { "epoch": 0.6786050156739812, "grad_norm": 2.9803202152252197, "learning_rate": 9.319449354232333e-06, "loss": 1.3606, "step": 8659 }, { "epoch": 0.6786833855799373, "grad_norm": 2.538666009902954, "learning_rate": 9.319294029987076e-06, "loss": 1.3243, "step": 8660 }, { "epoch": 0.6787617554858935, "grad_norm": 2.371464252471924, "learning_rate": 9.319138689313443e-06, "loss": 1.8186, "step": 8661 }, { "epoch": 0.6788401253918496, "grad_norm": 5.982067108154297, "learning_rate": 9.318983332212026e-06, "loss": 1.9755, "step": 8662 }, { "epoch": 0.6789184952978057, "grad_norm": 5.362741947174072, "learning_rate": 9.318827958683413e-06, "loss": 1.9227, "step": 8663 }, { "epoch": 0.6789968652037618, "grad_norm": 2.801361560821533, "learning_rate": 9.318672568728198e-06, "loss": 1.6351, "step": 8664 }, { "epoch": 0.6790752351097179, "grad_norm": 2.7871100902557373, "learning_rate": 9.31851716234697e-06, "loss": 1.307, "step": 8665 }, { "epoch": 0.679153605015674, "grad_norm": 5.279073715209961, "learning_rate": 9.31836173954032e-06, "loss": 1.1894, "step": 8666 }, { "epoch": 0.6792319749216301, "grad_norm": 2.141545295715332, "learning_rate": 9.318206300308841e-06, "loss": 1.5497, "step": 8667 }, { "epoch": 0.6793103448275862, "grad_norm": 2.0201075077056885, "learning_rate": 9.318050844653123e-06, "loss": 0.9118, "step": 8668 }, { "epoch": 0.6793887147335423, "grad_norm": 4.435454845428467, "learning_rate": 9.317895372573758e-06, "loss": 1.6052, "step": 8669 }, { "epoch": 0.6794670846394985, "grad_norm": 1.9347487688064575, "learning_rate": 9.317739884071337e-06, "loss": 1.3199, "step": 8670 }, { "epoch": 0.6795454545454546, "grad_norm": 2.7633562088012695, "learning_rate": 9.317584379146452e-06, "loss": 1.3608, "step": 8671 }, { "epoch": 0.6796238244514107, "grad_norm": 2.190086841583252, "learning_rate": 9.317428857799691e-06, "loss": 1.4846, "step": 8672 }, { "epoch": 0.6797021943573668, "grad_norm": 8.749131202697754, "learning_rate": 9.31727332003165e-06, "loss": 1.6133, "step": 8673 }, { "epoch": 0.6797805642633229, "grad_norm": 3.9844918251037598, "learning_rate": 9.317117765842917e-06, "loss": 1.7634, "step": 8674 }, { "epoch": 0.679858934169279, "grad_norm": 3.232428550720215, "learning_rate": 9.31696219523409e-06, "loss": 1.6526, "step": 8675 }, { "epoch": 0.6799373040752351, "grad_norm": 3.8856866359710693, "learning_rate": 9.316806608205752e-06, "loss": 1.2176, "step": 8676 }, { "epoch": 0.6800156739811912, "grad_norm": 2.89412260055542, "learning_rate": 9.316651004758498e-06, "loss": 1.4651, "step": 8677 }, { "epoch": 0.6800940438871473, "grad_norm": 4.0397233963012695, "learning_rate": 9.316495384892924e-06, "loss": 1.7528, "step": 8678 }, { "epoch": 0.6801724137931034, "grad_norm": 3.770794630050659, "learning_rate": 9.316339748609617e-06, "loss": 1.4981, "step": 8679 }, { "epoch": 0.6802507836990596, "grad_norm": 4.270389556884766, "learning_rate": 9.316184095909172e-06, "loss": 1.4583, "step": 8680 }, { "epoch": 0.6803291536050157, "grad_norm": 4.130189418792725, "learning_rate": 9.316028426792179e-06, "loss": 2.1272, "step": 8681 }, { "epoch": 0.6804075235109718, "grad_norm": 2.152146816253662, "learning_rate": 9.315872741259229e-06, "loss": 1.4533, "step": 8682 }, { "epoch": 0.6804858934169279, "grad_norm": 4.070977687835693, "learning_rate": 9.315717039310917e-06, "loss": 1.3319, "step": 8683 }, { "epoch": 0.680564263322884, "grad_norm": 2.680708169937134, "learning_rate": 9.315561320947834e-06, "loss": 1.4983, "step": 8684 }, { "epoch": 0.6806426332288401, "grad_norm": 2.7793309688568115, "learning_rate": 9.315405586170573e-06, "loss": 1.3509, "step": 8685 }, { "epoch": 0.6807210031347962, "grad_norm": 2.5240931510925293, "learning_rate": 9.315249834979725e-06, "loss": 1.5406, "step": 8686 }, { "epoch": 0.6807993730407523, "grad_norm": 2.8289709091186523, "learning_rate": 9.315094067375884e-06, "loss": 1.4361, "step": 8687 }, { "epoch": 0.6808777429467084, "grad_norm": 4.0863356590271, "learning_rate": 9.31493828335964e-06, "loss": 1.0682, "step": 8688 }, { "epoch": 0.6809561128526646, "grad_norm": 3.1313655376434326, "learning_rate": 9.314782482931588e-06, "loss": 0.9363, "step": 8689 }, { "epoch": 0.6810344827586207, "grad_norm": 8.06777286529541, "learning_rate": 9.31462666609232e-06, "loss": 1.688, "step": 8690 }, { "epoch": 0.6811128526645768, "grad_norm": 3.8549180030822754, "learning_rate": 9.314470832842428e-06, "loss": 1.6187, "step": 8691 }, { "epoch": 0.6811912225705329, "grad_norm": 2.74428653717041, "learning_rate": 9.314314983182504e-06, "loss": 1.3514, "step": 8692 }, { "epoch": 0.681269592476489, "grad_norm": 2.217615842819214, "learning_rate": 9.314159117113144e-06, "loss": 1.4353, "step": 8693 }, { "epoch": 0.6813479623824451, "grad_norm": 7.146295547485352, "learning_rate": 9.314003234634936e-06, "loss": 1.131, "step": 8694 }, { "epoch": 0.6814263322884012, "grad_norm": 4.124047756195068, "learning_rate": 9.313847335748479e-06, "loss": 1.0774, "step": 8695 }, { "epoch": 0.6815047021943573, "grad_norm": 9.510564804077148, "learning_rate": 9.313691420454359e-06, "loss": 2.6268, "step": 8696 }, { "epoch": 0.6815830721003134, "grad_norm": 3.3313260078430176, "learning_rate": 9.313535488753173e-06, "loss": 1.3598, "step": 8697 }, { "epoch": 0.6816614420062695, "grad_norm": 3.9852640628814697, "learning_rate": 9.313379540645515e-06, "loss": 1.7568, "step": 8698 }, { "epoch": 0.6817398119122257, "grad_norm": 2.8282082080841064, "learning_rate": 9.313223576131975e-06, "loss": 1.4585, "step": 8699 }, { "epoch": 0.6818181818181818, "grad_norm": 3.07132625579834, "learning_rate": 9.31306759521315e-06, "loss": 1.5515, "step": 8700 }, { "epoch": 0.6818965517241379, "grad_norm": 1.8615740537643433, "learning_rate": 9.31291159788963e-06, "loss": 1.4805, "step": 8701 }, { "epoch": 0.681974921630094, "grad_norm": 5.348463535308838, "learning_rate": 9.31275558416201e-06, "loss": 1.4097, "step": 8702 }, { "epoch": 0.6820532915360502, "grad_norm": 10.60466194152832, "learning_rate": 9.312599554030881e-06, "loss": 1.5148, "step": 8703 }, { "epoch": 0.6821316614420063, "grad_norm": 3.5024666786193848, "learning_rate": 9.312443507496841e-06, "loss": 1.4973, "step": 8704 }, { "epoch": 0.6822100313479624, "grad_norm": 2.9498281478881836, "learning_rate": 9.312287444560479e-06, "loss": 1.4505, "step": 8705 }, { "epoch": 0.6822884012539185, "grad_norm": 3.4658968448638916, "learning_rate": 9.312131365222392e-06, "loss": 1.3452, "step": 8706 }, { "epoch": 0.6823667711598747, "grad_norm": 2.1361372470855713, "learning_rate": 9.311975269483171e-06, "loss": 1.4789, "step": 8707 }, { "epoch": 0.6824451410658308, "grad_norm": 3.016744613647461, "learning_rate": 9.311819157343412e-06, "loss": 1.666, "step": 8708 }, { "epoch": 0.6825235109717869, "grad_norm": 10.672861099243164, "learning_rate": 9.311663028803707e-06, "loss": 1.267, "step": 8709 }, { "epoch": 0.682601880877743, "grad_norm": 3.884026288986206, "learning_rate": 9.311506883864652e-06, "loss": 1.5476, "step": 8710 }, { "epoch": 0.6826802507836991, "grad_norm": 2.2330965995788574, "learning_rate": 9.311350722526836e-06, "loss": 1.6295, "step": 8711 }, { "epoch": 0.6827586206896552, "grad_norm": 3.570540428161621, "learning_rate": 9.311194544790858e-06, "loss": 1.2892, "step": 8712 }, { "epoch": 0.6828369905956113, "grad_norm": 2.2003817558288574, "learning_rate": 9.311038350657312e-06, "loss": 1.1408, "step": 8713 }, { "epoch": 0.6829153605015674, "grad_norm": 3.0443015098571777, "learning_rate": 9.310882140126788e-06, "loss": 1.302, "step": 8714 }, { "epoch": 0.6829937304075235, "grad_norm": 2.6557915210723877, "learning_rate": 9.310725913199884e-06, "loss": 1.0151, "step": 8715 }, { "epoch": 0.6830721003134796, "grad_norm": 4.257128715515137, "learning_rate": 9.310569669877193e-06, "loss": 1.6707, "step": 8716 }, { "epoch": 0.6831504702194358, "grad_norm": 2.65277361869812, "learning_rate": 9.310413410159307e-06, "loss": 1.4107, "step": 8717 }, { "epoch": 0.6832288401253919, "grad_norm": 3.0517611503601074, "learning_rate": 9.310257134046823e-06, "loss": 1.5126, "step": 8718 }, { "epoch": 0.683307210031348, "grad_norm": 2.017651319503784, "learning_rate": 9.310100841540334e-06, "loss": 1.4388, "step": 8719 }, { "epoch": 0.6833855799373041, "grad_norm": 5.600471019744873, "learning_rate": 9.309944532640439e-06, "loss": 1.2854, "step": 8720 }, { "epoch": 0.6834639498432602, "grad_norm": 3.1637120246887207, "learning_rate": 9.309788207347725e-06, "loss": 1.0264, "step": 8721 }, { "epoch": 0.6835423197492163, "grad_norm": 2.392332077026367, "learning_rate": 9.309631865662791e-06, "loss": 1.3941, "step": 8722 }, { "epoch": 0.6836206896551724, "grad_norm": 3.3032619953155518, "learning_rate": 9.309475507586231e-06, "loss": 1.368, "step": 8723 }, { "epoch": 0.6836990595611285, "grad_norm": 2.9289097785949707, "learning_rate": 9.30931913311864e-06, "loss": 1.5126, "step": 8724 }, { "epoch": 0.6837774294670846, "grad_norm": 2.6728475093841553, "learning_rate": 9.309162742260613e-06, "loss": 1.2591, "step": 8725 }, { "epoch": 0.6838557993730408, "grad_norm": 3.9061973094940186, "learning_rate": 9.309006335012745e-06, "loss": 1.3272, "step": 8726 }, { "epoch": 0.6839341692789969, "grad_norm": 2.532183885574341, "learning_rate": 9.308849911375629e-06, "loss": 1.1981, "step": 8727 }, { "epoch": 0.684012539184953, "grad_norm": 4.463144302368164, "learning_rate": 9.30869347134986e-06, "loss": 1.6555, "step": 8728 }, { "epoch": 0.6840909090909091, "grad_norm": 1.8158572912216187, "learning_rate": 9.308537014936035e-06, "loss": 1.1, "step": 8729 }, { "epoch": 0.6841692789968652, "grad_norm": 2.048783779144287, "learning_rate": 9.30838054213475e-06, "loss": 1.4164, "step": 8730 }, { "epoch": 0.6842476489028213, "grad_norm": 3.6425817012786865, "learning_rate": 9.308224052946595e-06, "loss": 1.5204, "step": 8731 }, { "epoch": 0.6843260188087774, "grad_norm": 7.540510177612305, "learning_rate": 9.30806754737217e-06, "loss": 1.4897, "step": 8732 }, { "epoch": 0.6844043887147335, "grad_norm": 2.2728447914123535, "learning_rate": 9.30791102541207e-06, "loss": 1.5594, "step": 8733 }, { "epoch": 0.6844827586206896, "grad_norm": 9.277496337890625, "learning_rate": 9.307754487066887e-06, "loss": 1.348, "step": 8734 }, { "epoch": 0.6845611285266457, "grad_norm": 14.537263870239258, "learning_rate": 9.307597932337221e-06, "loss": 2.1464, "step": 8735 }, { "epoch": 0.6846394984326019, "grad_norm": 3.0820791721343994, "learning_rate": 9.307441361223664e-06, "loss": 1.4455, "step": 8736 }, { "epoch": 0.684717868338558, "grad_norm": 2.0821611881256104, "learning_rate": 9.30728477372681e-06, "loss": 1.1016, "step": 8737 }, { "epoch": 0.6847962382445141, "grad_norm": 2.075523853302002, "learning_rate": 9.307128169847261e-06, "loss": 1.42, "step": 8738 }, { "epoch": 0.6848746081504702, "grad_norm": 3.603604316711426, "learning_rate": 9.306971549585606e-06, "loss": 1.8389, "step": 8739 }, { "epoch": 0.6849529780564263, "grad_norm": 4.221465587615967, "learning_rate": 9.306814912942445e-06, "loss": 1.6692, "step": 8740 }, { "epoch": 0.6850313479623824, "grad_norm": 5.028193950653076, "learning_rate": 9.30665825991837e-06, "loss": 1.3653, "step": 8741 }, { "epoch": 0.6851097178683385, "grad_norm": 2.3932535648345947, "learning_rate": 9.306501590513982e-06, "loss": 1.5568, "step": 8742 }, { "epoch": 0.6851880877742946, "grad_norm": 3.1625969409942627, "learning_rate": 9.306344904729873e-06, "loss": 1.0402, "step": 8743 }, { "epoch": 0.6852664576802507, "grad_norm": 6.247427463531494, "learning_rate": 9.30618820256664e-06, "loss": 1.4288, "step": 8744 }, { "epoch": 0.6853448275862069, "grad_norm": 5.467637538909912, "learning_rate": 9.306031484024877e-06, "loss": 2.0264, "step": 8745 }, { "epoch": 0.685423197492163, "grad_norm": 3.6700520515441895, "learning_rate": 9.305874749105184e-06, "loss": 1.6673, "step": 8746 }, { "epoch": 0.6855015673981191, "grad_norm": 2.2726540565490723, "learning_rate": 9.305717997808155e-06, "loss": 1.1478, "step": 8747 }, { "epoch": 0.6855799373040752, "grad_norm": 2.9143006801605225, "learning_rate": 9.305561230134384e-06, "loss": 1.6958, "step": 8748 }, { "epoch": 0.6856583072100313, "grad_norm": 3.2957558631896973, "learning_rate": 9.305404446084472e-06, "loss": 1.3592, "step": 8749 }, { "epoch": 0.6857366771159875, "grad_norm": 2.9045517444610596, "learning_rate": 9.305247645659013e-06, "loss": 1.3105, "step": 8750 }, { "epoch": 0.6858150470219436, "grad_norm": 4.570237636566162, "learning_rate": 9.305090828858601e-06, "loss": 1.8676, "step": 8751 }, { "epoch": 0.6858934169278997, "grad_norm": 3.0502243041992188, "learning_rate": 9.304933995683836e-06, "loss": 1.1623, "step": 8752 }, { "epoch": 0.6859717868338558, "grad_norm": 2.3890068531036377, "learning_rate": 9.304777146135313e-06, "loss": 1.3647, "step": 8753 }, { "epoch": 0.686050156739812, "grad_norm": 8.57286262512207, "learning_rate": 9.304620280213629e-06, "loss": 1.4782, "step": 8754 }, { "epoch": 0.6861285266457681, "grad_norm": 5.691049098968506, "learning_rate": 9.304463397919381e-06, "loss": 1.2997, "step": 8755 }, { "epoch": 0.6862068965517242, "grad_norm": 9.191293716430664, "learning_rate": 9.304306499253165e-06, "loss": 2.034, "step": 8756 }, { "epoch": 0.6862852664576803, "grad_norm": 2.9499728679656982, "learning_rate": 9.304149584215578e-06, "loss": 1.1119, "step": 8757 }, { "epoch": 0.6863636363636364, "grad_norm": 2.534381151199341, "learning_rate": 9.303992652807216e-06, "loss": 1.4642, "step": 8758 }, { "epoch": 0.6864420062695925, "grad_norm": 4.01908540725708, "learning_rate": 9.303835705028676e-06, "loss": 1.5676, "step": 8759 }, { "epoch": 0.6865203761755486, "grad_norm": 2.317711591720581, "learning_rate": 9.303678740880558e-06, "loss": 1.0779, "step": 8760 }, { "epoch": 0.6865987460815047, "grad_norm": 3.042340040206909, "learning_rate": 9.303521760363454e-06, "loss": 1.5284, "step": 8761 }, { "epoch": 0.6866771159874608, "grad_norm": 4.866267204284668, "learning_rate": 9.303364763477965e-06, "loss": 1.6952, "step": 8762 }, { "epoch": 0.686755485893417, "grad_norm": 5.926168441772461, "learning_rate": 9.303207750224688e-06, "loss": 1.6291, "step": 8763 }, { "epoch": 0.6868338557993731, "grad_norm": 3.069312810897827, "learning_rate": 9.303050720604217e-06, "loss": 1.4548, "step": 8764 }, { "epoch": 0.6869122257053292, "grad_norm": 5.012956142425537, "learning_rate": 9.30289367461715e-06, "loss": 1.4969, "step": 8765 }, { "epoch": 0.6869905956112853, "grad_norm": 3.5945353507995605, "learning_rate": 9.302736612264088e-06, "loss": 1.1612, "step": 8766 }, { "epoch": 0.6870689655172414, "grad_norm": 4.980658054351807, "learning_rate": 9.302579533545625e-06, "loss": 1.9553, "step": 8767 }, { "epoch": 0.6871473354231975, "grad_norm": 2.7714827060699463, "learning_rate": 9.30242243846236e-06, "loss": 1.236, "step": 8768 }, { "epoch": 0.6872257053291536, "grad_norm": 2.094402551651001, "learning_rate": 9.302265327014889e-06, "loss": 1.2721, "step": 8769 }, { "epoch": 0.6873040752351097, "grad_norm": 4.4436259269714355, "learning_rate": 9.302108199203811e-06, "loss": 1.7999, "step": 8770 }, { "epoch": 0.6873824451410658, "grad_norm": 2.6754205226898193, "learning_rate": 9.301951055029724e-06, "loss": 1.5388, "step": 8771 }, { "epoch": 0.687460815047022, "grad_norm": 2.343773126602173, "learning_rate": 9.301793894493224e-06, "loss": 1.6381, "step": 8772 }, { "epoch": 0.687539184952978, "grad_norm": 7.020284652709961, "learning_rate": 9.30163671759491e-06, "loss": 1.342, "step": 8773 }, { "epoch": 0.6876175548589342, "grad_norm": 3.057659387588501, "learning_rate": 9.301479524335378e-06, "loss": 1.2981, "step": 8774 }, { "epoch": 0.6876959247648903, "grad_norm": 4.67368221282959, "learning_rate": 9.30132231471523e-06, "loss": 1.3352, "step": 8775 }, { "epoch": 0.6877742946708464, "grad_norm": 2.5374555587768555, "learning_rate": 9.301165088735057e-06, "loss": 1.568, "step": 8776 }, { "epoch": 0.6878526645768025, "grad_norm": 2.282716989517212, "learning_rate": 9.301007846395464e-06, "loss": 1.068, "step": 8777 }, { "epoch": 0.6879310344827586, "grad_norm": 2.3473358154296875, "learning_rate": 9.300850587697047e-06, "loss": 1.2756, "step": 8778 }, { "epoch": 0.6880094043887147, "grad_norm": 3.7551581859588623, "learning_rate": 9.300693312640403e-06, "loss": 1.339, "step": 8779 }, { "epoch": 0.6880877742946708, "grad_norm": 3.105156183242798, "learning_rate": 9.30053602122613e-06, "loss": 1.5948, "step": 8780 }, { "epoch": 0.6881661442006269, "grad_norm": 3.324373960494995, "learning_rate": 9.300378713454825e-06, "loss": 1.6842, "step": 8781 }, { "epoch": 0.688244514106583, "grad_norm": 2.4515321254730225, "learning_rate": 9.30022138932709e-06, "loss": 1.5476, "step": 8782 }, { "epoch": 0.6883228840125392, "grad_norm": 3.3993799686431885, "learning_rate": 9.300064048843522e-06, "loss": 1.5085, "step": 8783 }, { "epoch": 0.6884012539184953, "grad_norm": 4.531297206878662, "learning_rate": 9.299906692004719e-06, "loss": 1.3808, "step": 8784 }, { "epoch": 0.6884796238244514, "grad_norm": 3.9384756088256836, "learning_rate": 9.29974931881128e-06, "loss": 1.5875, "step": 8785 }, { "epoch": 0.6885579937304075, "grad_norm": 3.82079815864563, "learning_rate": 9.299591929263802e-06, "loss": 1.7464, "step": 8786 }, { "epoch": 0.6886363636363636, "grad_norm": 8.689525604248047, "learning_rate": 9.299434523362885e-06, "loss": 1.0308, "step": 8787 }, { "epoch": 0.6887147335423197, "grad_norm": 6.097297668457031, "learning_rate": 9.299277101109128e-06, "loss": 1.1789, "step": 8788 }, { "epoch": 0.6887931034482758, "grad_norm": 3.670776605606079, "learning_rate": 9.299119662503127e-06, "loss": 1.1287, "step": 8789 }, { "epoch": 0.6888714733542319, "grad_norm": 7.11730432510376, "learning_rate": 9.298962207545486e-06, "loss": 0.7262, "step": 8790 }, { "epoch": 0.688949843260188, "grad_norm": 5.841953754425049, "learning_rate": 9.298804736236798e-06, "loss": 1.3094, "step": 8791 }, { "epoch": 0.6890282131661442, "grad_norm": 3.0865304470062256, "learning_rate": 9.298647248577666e-06, "loss": 1.17, "step": 8792 }, { "epoch": 0.6891065830721003, "grad_norm": 3.1374289989471436, "learning_rate": 9.298489744568689e-06, "loss": 2.1639, "step": 8793 }, { "epoch": 0.6891849529780564, "grad_norm": 3.658726215362549, "learning_rate": 9.298332224210463e-06, "loss": 1.4362, "step": 8794 }, { "epoch": 0.6892633228840125, "grad_norm": 3.1076812744140625, "learning_rate": 9.29817468750359e-06, "loss": 1.5498, "step": 8795 }, { "epoch": 0.6893416927899687, "grad_norm": 6.0388898849487305, "learning_rate": 9.298017134448667e-06, "loss": 1.5754, "step": 8796 }, { "epoch": 0.6894200626959248, "grad_norm": 3.972409963607788, "learning_rate": 9.297859565046296e-06, "loss": 1.5109, "step": 8797 }, { "epoch": 0.6894984326018809, "grad_norm": 3.1213300228118896, "learning_rate": 9.297701979297073e-06, "loss": 1.3803, "step": 8798 }, { "epoch": 0.689576802507837, "grad_norm": 3.9693422317504883, "learning_rate": 9.297544377201602e-06, "loss": 1.0855, "step": 8799 }, { "epoch": 0.6896551724137931, "grad_norm": 3.2983744144439697, "learning_rate": 9.297386758760476e-06, "loss": 1.6315, "step": 8800 }, { "epoch": 0.6897335423197493, "grad_norm": 2.5307390689849854, "learning_rate": 9.297229123974301e-06, "loss": 1.7465, "step": 8801 }, { "epoch": 0.6898119122257054, "grad_norm": 1.8626480102539062, "learning_rate": 9.297071472843671e-06, "loss": 1.3275, "step": 8802 }, { "epoch": 0.6898902821316615, "grad_norm": 4.364910125732422, "learning_rate": 9.29691380536919e-06, "loss": 1.5112, "step": 8803 }, { "epoch": 0.6899686520376176, "grad_norm": 2.4941086769104004, "learning_rate": 9.296756121551456e-06, "loss": 1.5432, "step": 8804 }, { "epoch": 0.6900470219435737, "grad_norm": 3.03525972366333, "learning_rate": 9.296598421391068e-06, "loss": 1.2542, "step": 8805 }, { "epoch": 0.6901253918495298, "grad_norm": 2.7232275009155273, "learning_rate": 9.296440704888627e-06, "loss": 1.3388, "step": 8806 }, { "epoch": 0.6902037617554859, "grad_norm": 2.3994970321655273, "learning_rate": 9.296282972044733e-06, "loss": 1.4712, "step": 8807 }, { "epoch": 0.690282131661442, "grad_norm": 4.814522743225098, "learning_rate": 9.296125222859983e-06, "loss": 1.4519, "step": 8808 }, { "epoch": 0.6903605015673981, "grad_norm": 3.0898518562316895, "learning_rate": 9.29596745733498e-06, "loss": 1.6081, "step": 8809 }, { "epoch": 0.6904388714733543, "grad_norm": 2.476943254470825, "learning_rate": 9.295809675470325e-06, "loss": 1.4474, "step": 8810 }, { "epoch": 0.6905172413793104, "grad_norm": 4.1659345626831055, "learning_rate": 9.295651877266615e-06, "loss": 1.7577, "step": 8811 }, { "epoch": 0.6905956112852665, "grad_norm": 2.8877663612365723, "learning_rate": 9.295494062724453e-06, "loss": 1.329, "step": 8812 }, { "epoch": 0.6906739811912226, "grad_norm": 9.598562240600586, "learning_rate": 9.295336231844436e-06, "loss": 1.1428, "step": 8813 }, { "epoch": 0.6907523510971787, "grad_norm": 4.134608268737793, "learning_rate": 9.295178384627167e-06, "loss": 2.0699, "step": 8814 }, { "epoch": 0.6908307210031348, "grad_norm": 2.862783193588257, "learning_rate": 9.295020521073246e-06, "loss": 1.3743, "step": 8815 }, { "epoch": 0.6909090909090909, "grad_norm": 3.3466923236846924, "learning_rate": 9.294862641183273e-06, "loss": 1.277, "step": 8816 }, { "epoch": 0.690987460815047, "grad_norm": 5.868934154510498, "learning_rate": 9.294704744957847e-06, "loss": 1.4099, "step": 8817 }, { "epoch": 0.6910658307210031, "grad_norm": 2.980180025100708, "learning_rate": 9.294546832397571e-06, "loss": 1.6144, "step": 8818 }, { "epoch": 0.6911442006269592, "grad_norm": 5.109950542449951, "learning_rate": 9.294388903503046e-06, "loss": 1.6136, "step": 8819 }, { "epoch": 0.6912225705329154, "grad_norm": 3.740830421447754, "learning_rate": 9.29423095827487e-06, "loss": 1.4826, "step": 8820 }, { "epoch": 0.6913009404388715, "grad_norm": 3.2554404735565186, "learning_rate": 9.294072996713645e-06, "loss": 1.5203, "step": 8821 }, { "epoch": 0.6913793103448276, "grad_norm": 3.2521965503692627, "learning_rate": 9.293915018819973e-06, "loss": 1.3588, "step": 8822 }, { "epoch": 0.6914576802507837, "grad_norm": 3.1048758029937744, "learning_rate": 9.293757024594452e-06, "loss": 1.5791, "step": 8823 }, { "epoch": 0.6915360501567398, "grad_norm": 2.4395058155059814, "learning_rate": 9.293599014037686e-06, "loss": 1.2862, "step": 8824 }, { "epoch": 0.6916144200626959, "grad_norm": 5.327770233154297, "learning_rate": 9.293440987150275e-06, "loss": 2.108, "step": 8825 }, { "epoch": 0.691692789968652, "grad_norm": 2.6946959495544434, "learning_rate": 9.293282943932819e-06, "loss": 1.262, "step": 8826 }, { "epoch": 0.6917711598746081, "grad_norm": 2.54058837890625, "learning_rate": 9.29312488438592e-06, "loss": 1.1896, "step": 8827 }, { "epoch": 0.6918495297805642, "grad_norm": 2.9053289890289307, "learning_rate": 9.292966808510178e-06, "loss": 1.0869, "step": 8828 }, { "epoch": 0.6919278996865204, "grad_norm": 3.1853413581848145, "learning_rate": 9.292808716306196e-06, "loss": 1.1485, "step": 8829 }, { "epoch": 0.6920062695924765, "grad_norm": 2.461212158203125, "learning_rate": 9.292650607774576e-06, "loss": 1.618, "step": 8830 }, { "epoch": 0.6920846394984326, "grad_norm": 2.9748682975769043, "learning_rate": 9.292492482915915e-06, "loss": 1.2927, "step": 8831 }, { "epoch": 0.6921630094043887, "grad_norm": 7.684957027435303, "learning_rate": 9.29233434173082e-06, "loss": 1.4333, "step": 8832 }, { "epoch": 0.6922413793103448, "grad_norm": 3.6568737030029297, "learning_rate": 9.29217618421989e-06, "loss": 1.2604, "step": 8833 }, { "epoch": 0.6923197492163009, "grad_norm": 3.8119702339172363, "learning_rate": 9.292018010383724e-06, "loss": 1.3088, "step": 8834 }, { "epoch": 0.692398119122257, "grad_norm": 2.749877452850342, "learning_rate": 9.291859820222927e-06, "loss": 1.3551, "step": 8835 }, { "epoch": 0.6924764890282131, "grad_norm": 2.199324369430542, "learning_rate": 9.2917016137381e-06, "loss": 1.3698, "step": 8836 }, { "epoch": 0.6925548589341692, "grad_norm": 2.4425158500671387, "learning_rate": 9.291543390929844e-06, "loss": 1.3358, "step": 8837 }, { "epoch": 0.6926332288401253, "grad_norm": 2.8458361625671387, "learning_rate": 9.291385151798763e-06, "loss": 1.4233, "step": 8838 }, { "epoch": 0.6927115987460815, "grad_norm": 2.38069224357605, "learning_rate": 9.291226896345455e-06, "loss": 1.4557, "step": 8839 }, { "epoch": 0.6927899686520376, "grad_norm": 2.976609706878662, "learning_rate": 9.291068624570525e-06, "loss": 1.5446, "step": 8840 }, { "epoch": 0.6928683385579937, "grad_norm": 2.6546010971069336, "learning_rate": 9.290910336474575e-06, "loss": 1.2821, "step": 8841 }, { "epoch": 0.6929467084639498, "grad_norm": 3.3957738876342773, "learning_rate": 9.290752032058203e-06, "loss": 1.3678, "step": 8842 }, { "epoch": 0.693025078369906, "grad_norm": 3.94163179397583, "learning_rate": 9.290593711322018e-06, "loss": 1.2927, "step": 8843 }, { "epoch": 0.6931034482758621, "grad_norm": 4.458917617797852, "learning_rate": 9.290435374266616e-06, "loss": 0.9452, "step": 8844 }, { "epoch": 0.6931818181818182, "grad_norm": 3.6530351638793945, "learning_rate": 9.290277020892602e-06, "loss": 1.8816, "step": 8845 }, { "epoch": 0.6932601880877743, "grad_norm": 2.8857879638671875, "learning_rate": 9.290118651200577e-06, "loss": 1.569, "step": 8846 }, { "epoch": 0.6933385579937305, "grad_norm": 2.669560670852661, "learning_rate": 9.289960265191145e-06, "loss": 1.2275, "step": 8847 }, { "epoch": 0.6934169278996866, "grad_norm": 2.4098851680755615, "learning_rate": 9.289801862864906e-06, "loss": 1.3995, "step": 8848 }, { "epoch": 0.6934952978056427, "grad_norm": 2.3019983768463135, "learning_rate": 9.289643444222466e-06, "loss": 1.3087, "step": 8849 }, { "epoch": 0.6935736677115988, "grad_norm": 5.937759876251221, "learning_rate": 9.289485009264425e-06, "loss": 0.8812, "step": 8850 }, { "epoch": 0.6936520376175549, "grad_norm": 2.5524191856384277, "learning_rate": 9.289326557991388e-06, "loss": 1.2906, "step": 8851 }, { "epoch": 0.693730407523511, "grad_norm": 2.1466944217681885, "learning_rate": 9.289168090403953e-06, "loss": 1.2949, "step": 8852 }, { "epoch": 0.6938087774294671, "grad_norm": 2.489201545715332, "learning_rate": 9.289009606502727e-06, "loss": 1.2304, "step": 8853 }, { "epoch": 0.6938871473354232, "grad_norm": 3.199671745300293, "learning_rate": 9.288851106288311e-06, "loss": 1.597, "step": 8854 }, { "epoch": 0.6939655172413793, "grad_norm": 4.055056095123291, "learning_rate": 9.288692589761309e-06, "loss": 1.5218, "step": 8855 }, { "epoch": 0.6940438871473354, "grad_norm": 2.9995620250701904, "learning_rate": 9.288534056922322e-06, "loss": 1.4672, "step": 8856 }, { "epoch": 0.6941222570532916, "grad_norm": 2.8691658973693848, "learning_rate": 9.288375507771956e-06, "loss": 1.8258, "step": 8857 }, { "epoch": 0.6942006269592477, "grad_norm": 2.6708686351776123, "learning_rate": 9.28821694231081e-06, "loss": 1.4163, "step": 8858 }, { "epoch": 0.6942789968652038, "grad_norm": 6.0955681800842285, "learning_rate": 9.288058360539491e-06, "loss": 1.8777, "step": 8859 }, { "epoch": 0.6943573667711599, "grad_norm": 2.9406533241271973, "learning_rate": 9.287899762458602e-06, "loss": 1.2887, "step": 8860 }, { "epoch": 0.694435736677116, "grad_norm": 3.997138738632202, "learning_rate": 9.28774114806874e-06, "loss": 1.5994, "step": 8861 }, { "epoch": 0.6945141065830721, "grad_norm": 2.3097081184387207, "learning_rate": 9.287582517370517e-06, "loss": 1.3984, "step": 8862 }, { "epoch": 0.6945924764890282, "grad_norm": 5.547086238861084, "learning_rate": 9.28742387036453e-06, "loss": 2.3695, "step": 8863 }, { "epoch": 0.6946708463949843, "grad_norm": 4.504936695098877, "learning_rate": 9.287265207051387e-06, "loss": 1.7891, "step": 8864 }, { "epoch": 0.6947492163009404, "grad_norm": 5.603170871734619, "learning_rate": 9.287106527431688e-06, "loss": 1.1955, "step": 8865 }, { "epoch": 0.6948275862068966, "grad_norm": 1.9952223300933838, "learning_rate": 9.286947831506038e-06, "loss": 1.2552, "step": 8866 }, { "epoch": 0.6949059561128527, "grad_norm": 6.460237979888916, "learning_rate": 9.28678911927504e-06, "loss": 1.1163, "step": 8867 }, { "epoch": 0.6949843260188088, "grad_norm": 2.8688924312591553, "learning_rate": 9.2866303907393e-06, "loss": 1.1937, "step": 8868 }, { "epoch": 0.6950626959247649, "grad_norm": 3.9619858264923096, "learning_rate": 9.286471645899416e-06, "loss": 0.8693, "step": 8869 }, { "epoch": 0.695141065830721, "grad_norm": 3.138005256652832, "learning_rate": 9.286312884755997e-06, "loss": 1.1545, "step": 8870 }, { "epoch": 0.6952194357366771, "grad_norm": 2.8624331951141357, "learning_rate": 9.286154107309645e-06, "loss": 1.2153, "step": 8871 }, { "epoch": 0.6952978056426332, "grad_norm": 4.0385613441467285, "learning_rate": 9.285995313560967e-06, "loss": 1.5104, "step": 8872 }, { "epoch": 0.6953761755485893, "grad_norm": 3.2658462524414062, "learning_rate": 9.285836503510562e-06, "loss": 1.3722, "step": 8873 }, { "epoch": 0.6954545454545454, "grad_norm": 2.2730772495269775, "learning_rate": 9.285677677159038e-06, "loss": 1.1691, "step": 8874 }, { "epoch": 0.6955329153605015, "grad_norm": 2.927093267440796, "learning_rate": 9.285518834506994e-06, "loss": 1.4834, "step": 8875 }, { "epoch": 0.6956112852664577, "grad_norm": 2.804849624633789, "learning_rate": 9.28535997555504e-06, "loss": 1.4646, "step": 8876 }, { "epoch": 0.6956896551724138, "grad_norm": 2.3132612705230713, "learning_rate": 9.285201100303778e-06, "loss": 1.2705, "step": 8877 }, { "epoch": 0.6957680250783699, "grad_norm": 1.924919843673706, "learning_rate": 9.285042208753811e-06, "loss": 0.8825, "step": 8878 }, { "epoch": 0.695846394984326, "grad_norm": 7.4916486740112305, "learning_rate": 9.284883300905746e-06, "loss": 1.553, "step": 8879 }, { "epoch": 0.6959247648902821, "grad_norm": 3.4050629138946533, "learning_rate": 9.284724376760184e-06, "loss": 1.3025, "step": 8880 }, { "epoch": 0.6960031347962382, "grad_norm": 3.0147581100463867, "learning_rate": 9.284565436317732e-06, "loss": 1.1235, "step": 8881 }, { "epoch": 0.6960815047021943, "grad_norm": 2.5563697814941406, "learning_rate": 9.284406479578993e-06, "loss": 1.4242, "step": 8882 }, { "epoch": 0.6961598746081504, "grad_norm": 3.0423054695129395, "learning_rate": 9.284247506544573e-06, "loss": 1.6951, "step": 8883 }, { "epoch": 0.6962382445141065, "grad_norm": 6.012694358825684, "learning_rate": 9.284088517215076e-06, "loss": 1.9542, "step": 8884 }, { "epoch": 0.6963166144200627, "grad_norm": 3.1319971084594727, "learning_rate": 9.283929511591107e-06, "loss": 1.4659, "step": 8885 }, { "epoch": 0.6963949843260188, "grad_norm": 2.8561127185821533, "learning_rate": 9.28377048967327e-06, "loss": 1.2101, "step": 8886 }, { "epoch": 0.6964733542319749, "grad_norm": 3.0603039264678955, "learning_rate": 9.283611451462171e-06, "loss": 1.2939, "step": 8887 }, { "epoch": 0.696551724137931, "grad_norm": 2.503631353378296, "learning_rate": 9.283452396958413e-06, "loss": 1.2361, "step": 8888 }, { "epoch": 0.6966300940438871, "grad_norm": 3.6792514324188232, "learning_rate": 9.283293326162603e-06, "loss": 1.7915, "step": 8889 }, { "epoch": 0.6967084639498433, "grad_norm": 3.3085103034973145, "learning_rate": 9.283134239075345e-06, "loss": 0.7916, "step": 8890 }, { "epoch": 0.6967868338557994, "grad_norm": 2.5673928260803223, "learning_rate": 9.282975135697245e-06, "loss": 1.3098, "step": 8891 }, { "epoch": 0.6968652037617555, "grad_norm": 2.8987584114074707, "learning_rate": 9.282816016028907e-06, "loss": 1.3216, "step": 8892 }, { "epoch": 0.6969435736677116, "grad_norm": 3.703906297683716, "learning_rate": 9.282656880070937e-06, "loss": 1.0442, "step": 8893 }, { "epoch": 0.6970219435736678, "grad_norm": 3.220266819000244, "learning_rate": 9.282497727823937e-06, "loss": 1.5067, "step": 8894 }, { "epoch": 0.6971003134796239, "grad_norm": 2.1995644569396973, "learning_rate": 9.28233855928852e-06, "loss": 1.4394, "step": 8895 }, { "epoch": 0.69717868338558, "grad_norm": 2.8515915870666504, "learning_rate": 9.282179374465285e-06, "loss": 1.1652, "step": 8896 }, { "epoch": 0.6972570532915361, "grad_norm": 2.685908555984497, "learning_rate": 9.282020173354837e-06, "loss": 1.5945, "step": 8897 }, { "epoch": 0.6973354231974922, "grad_norm": 8.169974327087402, "learning_rate": 9.281860955957787e-06, "loss": 2.5982, "step": 8898 }, { "epoch": 0.6974137931034483, "grad_norm": 3.710214138031006, "learning_rate": 9.281701722274736e-06, "loss": 1.2368, "step": 8899 }, { "epoch": 0.6974921630094044, "grad_norm": 3.636380672454834, "learning_rate": 9.281542472306289e-06, "loss": 1.3304, "step": 8900 }, { "epoch": 0.6975705329153605, "grad_norm": 2.0463290214538574, "learning_rate": 9.281383206053056e-06, "loss": 1.4781, "step": 8901 }, { "epoch": 0.6976489028213166, "grad_norm": 4.406788349151611, "learning_rate": 9.281223923515638e-06, "loss": 1.8254, "step": 8902 }, { "epoch": 0.6977272727272728, "grad_norm": 3.3440816402435303, "learning_rate": 9.281064624694645e-06, "loss": 1.0519, "step": 8903 }, { "epoch": 0.6978056426332289, "grad_norm": 2.581500291824341, "learning_rate": 9.28090530959068e-06, "loss": 1.2947, "step": 8904 }, { "epoch": 0.697884012539185, "grad_norm": 5.142057418823242, "learning_rate": 9.280745978204351e-06, "loss": 1.3566, "step": 8905 }, { "epoch": 0.6979623824451411, "grad_norm": 2.542572021484375, "learning_rate": 9.280586630536263e-06, "loss": 1.5752, "step": 8906 }, { "epoch": 0.6980407523510972, "grad_norm": 3.1546828746795654, "learning_rate": 9.28042726658702e-06, "loss": 1.5843, "step": 8907 }, { "epoch": 0.6981191222570533, "grad_norm": 2.886012554168701, "learning_rate": 9.280267886357232e-06, "loss": 1.6037, "step": 8908 }, { "epoch": 0.6981974921630094, "grad_norm": 2.626474618911743, "learning_rate": 9.280108489847505e-06, "loss": 1.4157, "step": 8909 }, { "epoch": 0.6982758620689655, "grad_norm": 3.201022148132324, "learning_rate": 9.279949077058441e-06, "loss": 1.3021, "step": 8910 }, { "epoch": 0.6983542319749216, "grad_norm": 3.0652294158935547, "learning_rate": 9.27978964799065e-06, "loss": 1.5765, "step": 8911 }, { "epoch": 0.6984326018808777, "grad_norm": 2.8154356479644775, "learning_rate": 9.279630202644739e-06, "loss": 1.4217, "step": 8912 }, { "epoch": 0.6985109717868339, "grad_norm": 7.851816654205322, "learning_rate": 9.27947074102131e-06, "loss": 0.6022, "step": 8913 }, { "epoch": 0.69858934169279, "grad_norm": 8.880976676940918, "learning_rate": 9.279311263120974e-06, "loss": 1.6045, "step": 8914 }, { "epoch": 0.6986677115987461, "grad_norm": 4.050746917724609, "learning_rate": 9.279151768944336e-06, "loss": 1.5027, "step": 8915 }, { "epoch": 0.6987460815047022, "grad_norm": 2.142214298248291, "learning_rate": 9.278992258492003e-06, "loss": 1.6864, "step": 8916 }, { "epoch": 0.6988244514106583, "grad_norm": 2.469081401824951, "learning_rate": 9.27883273176458e-06, "loss": 1.3686, "step": 8917 }, { "epoch": 0.6989028213166144, "grad_norm": 5.277904510498047, "learning_rate": 9.278673188762676e-06, "loss": 1.229, "step": 8918 }, { "epoch": 0.6989811912225705, "grad_norm": 6.172658443450928, "learning_rate": 9.278513629486895e-06, "loss": 1.2836, "step": 8919 }, { "epoch": 0.6990595611285266, "grad_norm": 3.1620538234710693, "learning_rate": 9.278354053937848e-06, "loss": 1.1759, "step": 8920 }, { "epoch": 0.6991379310344827, "grad_norm": 3.2109789848327637, "learning_rate": 9.278194462116138e-06, "loss": 1.3598, "step": 8921 }, { "epoch": 0.6992163009404389, "grad_norm": 3.8070919513702393, "learning_rate": 9.278034854022375e-06, "loss": 0.6265, "step": 8922 }, { "epoch": 0.699294670846395, "grad_norm": 4.578334808349609, "learning_rate": 9.277875229657166e-06, "loss": 1.0209, "step": 8923 }, { "epoch": 0.6993730407523511, "grad_norm": 8.496167182922363, "learning_rate": 9.277715589021115e-06, "loss": 1.2828, "step": 8924 }, { "epoch": 0.6994514106583072, "grad_norm": 5.914509296417236, "learning_rate": 9.27755593211483e-06, "loss": 0.9416, "step": 8925 }, { "epoch": 0.6995297805642633, "grad_norm": 2.674062967300415, "learning_rate": 9.277396258938922e-06, "loss": 1.4351, "step": 8926 }, { "epoch": 0.6996081504702194, "grad_norm": 2.9058821201324463, "learning_rate": 9.277236569493993e-06, "loss": 1.6541, "step": 8927 }, { "epoch": 0.6996865203761755, "grad_norm": 2.964562177658081, "learning_rate": 9.277076863780654e-06, "loss": 1.2822, "step": 8928 }, { "epoch": 0.6997648902821316, "grad_norm": 2.6447646617889404, "learning_rate": 9.276917141799513e-06, "loss": 1.1255, "step": 8929 }, { "epoch": 0.6998432601880877, "grad_norm": 2.1083428859710693, "learning_rate": 9.276757403551173e-06, "loss": 1.2687, "step": 8930 }, { "epoch": 0.6999216300940438, "grad_norm": 3.056337356567383, "learning_rate": 9.276597649036247e-06, "loss": 1.4205, "step": 8931 }, { "epoch": 0.7, "grad_norm": 4.263062000274658, "learning_rate": 9.27643787825534e-06, "loss": 1.2307, "step": 8932 }, { "epoch": 0.7000783699059561, "grad_norm": 1.9178705215454102, "learning_rate": 9.27627809120906e-06, "loss": 1.2519, "step": 8933 }, { "epoch": 0.7001567398119122, "grad_norm": 4.07138729095459, "learning_rate": 9.276118287898013e-06, "loss": 1.7059, "step": 8934 }, { "epoch": 0.7002351097178683, "grad_norm": 3.742980718612671, "learning_rate": 9.275958468322808e-06, "loss": 1.7388, "step": 8935 }, { "epoch": 0.7003134796238244, "grad_norm": 6.131039619445801, "learning_rate": 9.275798632484055e-06, "loss": 1.5689, "step": 8936 }, { "epoch": 0.7003918495297806, "grad_norm": 2.3970930576324463, "learning_rate": 9.27563878038236e-06, "loss": 1.5583, "step": 8937 }, { "epoch": 0.7004702194357367, "grad_norm": 1.8212273120880127, "learning_rate": 9.275478912018329e-06, "loss": 1.2875, "step": 8938 }, { "epoch": 0.7005485893416928, "grad_norm": 3.210911750793457, "learning_rate": 9.275319027392576e-06, "loss": 1.4859, "step": 8939 }, { "epoch": 0.700626959247649, "grad_norm": 3.551119327545166, "learning_rate": 9.275159126505702e-06, "loss": 1.4725, "step": 8940 }, { "epoch": 0.7007053291536051, "grad_norm": 3.2855470180511475, "learning_rate": 9.27499920935832e-06, "loss": 1.6217, "step": 8941 }, { "epoch": 0.7007836990595612, "grad_norm": 2.265486001968384, "learning_rate": 9.274839275951037e-06, "loss": 1.2992, "step": 8942 }, { "epoch": 0.7008620689655173, "grad_norm": 2.946871042251587, "learning_rate": 9.274679326284461e-06, "loss": 1.658, "step": 8943 }, { "epoch": 0.7009404388714734, "grad_norm": 2.9882607460021973, "learning_rate": 9.2745193603592e-06, "loss": 2.0437, "step": 8944 }, { "epoch": 0.7010188087774295, "grad_norm": 3.3041350841522217, "learning_rate": 9.274359378175864e-06, "loss": 1.7201, "step": 8945 }, { "epoch": 0.7010971786833856, "grad_norm": 4.741800308227539, "learning_rate": 9.274199379735059e-06, "loss": 1.3722, "step": 8946 }, { "epoch": 0.7011755485893417, "grad_norm": 2.2329185009002686, "learning_rate": 9.274039365037395e-06, "loss": 1.513, "step": 8947 }, { "epoch": 0.7012539184952978, "grad_norm": 3.328378200531006, "learning_rate": 9.273879334083481e-06, "loss": 1.4534, "step": 8948 }, { "epoch": 0.701332288401254, "grad_norm": 2.4447033405303955, "learning_rate": 9.273719286873926e-06, "loss": 1.1857, "step": 8949 }, { "epoch": 0.70141065830721, "grad_norm": 4.631404399871826, "learning_rate": 9.273559223409336e-06, "loss": 1.4685, "step": 8950 }, { "epoch": 0.7014890282131662, "grad_norm": 2.733529806137085, "learning_rate": 9.273399143690324e-06, "loss": 1.2881, "step": 8951 }, { "epoch": 0.7015673981191223, "grad_norm": 2.6652848720550537, "learning_rate": 9.273239047717496e-06, "loss": 1.5696, "step": 8952 }, { "epoch": 0.7016457680250784, "grad_norm": 2.937742233276367, "learning_rate": 9.27307893549146e-06, "loss": 1.5437, "step": 8953 }, { "epoch": 0.7017241379310345, "grad_norm": 2.341468334197998, "learning_rate": 9.272918807012827e-06, "loss": 1.5072, "step": 8954 }, { "epoch": 0.7018025078369906, "grad_norm": 16.568204879760742, "learning_rate": 9.272758662282206e-06, "loss": 0.9428, "step": 8955 }, { "epoch": 0.7018808777429467, "grad_norm": 2.395822286605835, "learning_rate": 9.272598501300207e-06, "loss": 1.5441, "step": 8956 }, { "epoch": 0.7019592476489028, "grad_norm": 2.2825961112976074, "learning_rate": 9.272438324067437e-06, "loss": 1.2862, "step": 8957 }, { "epoch": 0.7020376175548589, "grad_norm": 5.0278425216674805, "learning_rate": 9.272278130584504e-06, "loss": 1.4734, "step": 8958 }, { "epoch": 0.702115987460815, "grad_norm": 7.919840335845947, "learning_rate": 9.272117920852021e-06, "loss": 1.0656, "step": 8959 }, { "epoch": 0.7021943573667712, "grad_norm": 4.113426208496094, "learning_rate": 9.271957694870596e-06, "loss": 1.2271, "step": 8960 }, { "epoch": 0.7022727272727273, "grad_norm": 5.181416034698486, "learning_rate": 9.271797452640838e-06, "loss": 1.0969, "step": 8961 }, { "epoch": 0.7023510971786834, "grad_norm": 2.795987367630005, "learning_rate": 9.271637194163356e-06, "loss": 1.5082, "step": 8962 }, { "epoch": 0.7024294670846395, "grad_norm": 5.080859184265137, "learning_rate": 9.271476919438762e-06, "loss": 1.9085, "step": 8963 }, { "epoch": 0.7025078369905956, "grad_norm": 3.010474920272827, "learning_rate": 9.271316628467663e-06, "loss": 1.0793, "step": 8964 }, { "epoch": 0.7025862068965517, "grad_norm": 6.756999969482422, "learning_rate": 9.271156321250669e-06, "loss": 1.0708, "step": 8965 }, { "epoch": 0.7026645768025078, "grad_norm": 3.4232146739959717, "learning_rate": 9.27099599778839e-06, "loss": 1.6995, "step": 8966 }, { "epoch": 0.7027429467084639, "grad_norm": 3.9009621143341064, "learning_rate": 9.270835658081437e-06, "loss": 1.1051, "step": 8967 }, { "epoch": 0.70282131661442, "grad_norm": 4.178445816040039, "learning_rate": 9.270675302130416e-06, "loss": 1.2413, "step": 8968 }, { "epoch": 0.7028996865203762, "grad_norm": 3.4975175857543945, "learning_rate": 9.270514929935942e-06, "loss": 1.3441, "step": 8969 }, { "epoch": 0.7029780564263323, "grad_norm": 2.813206195831299, "learning_rate": 9.270354541498623e-06, "loss": 1.4495, "step": 8970 }, { "epoch": 0.7030564263322884, "grad_norm": 3.011862277984619, "learning_rate": 9.270194136819067e-06, "loss": 1.4788, "step": 8971 }, { "epoch": 0.7031347962382445, "grad_norm": 7.294151306152344, "learning_rate": 9.270033715897888e-06, "loss": 0.776, "step": 8972 }, { "epoch": 0.7032131661442006, "grad_norm": 3.444742202758789, "learning_rate": 9.269873278735692e-06, "loss": 0.8831, "step": 8973 }, { "epoch": 0.7032915360501567, "grad_norm": 2.3875203132629395, "learning_rate": 9.269712825333093e-06, "loss": 1.2587, "step": 8974 }, { "epoch": 0.7033699059561128, "grad_norm": 2.46126389503479, "learning_rate": 9.269552355690697e-06, "loss": 1.6197, "step": 8975 }, { "epoch": 0.7034482758620689, "grad_norm": 4.0972771644592285, "learning_rate": 9.26939186980912e-06, "loss": 1.35, "step": 8976 }, { "epoch": 0.703526645768025, "grad_norm": 3.7883567810058594, "learning_rate": 9.269231367688967e-06, "loss": 1.5901, "step": 8977 }, { "epoch": 0.7036050156739811, "grad_norm": 5.341439247131348, "learning_rate": 9.26907084933085e-06, "loss": 0.6167, "step": 8978 }, { "epoch": 0.7036833855799373, "grad_norm": 3.6121904850006104, "learning_rate": 9.268910314735382e-06, "loss": 1.7633, "step": 8979 }, { "epoch": 0.7037617554858934, "grad_norm": 3.1203300952911377, "learning_rate": 9.268749763903171e-06, "loss": 1.1787, "step": 8980 }, { "epoch": 0.7038401253918495, "grad_norm": 3.7551562786102295, "learning_rate": 9.268589196834829e-06, "loss": 1.2282, "step": 8981 }, { "epoch": 0.7039184952978056, "grad_norm": 9.689659118652344, "learning_rate": 9.268428613530966e-06, "loss": 1.557, "step": 8982 }, { "epoch": 0.7039968652037617, "grad_norm": 8.25834846496582, "learning_rate": 9.268268013992191e-06, "loss": 0.8743, "step": 8983 }, { "epoch": 0.7040752351097179, "grad_norm": 6.99939489364624, "learning_rate": 9.268107398219119e-06, "loss": 1.4936, "step": 8984 }, { "epoch": 0.704153605015674, "grad_norm": 7.807664394378662, "learning_rate": 9.267946766212359e-06, "loss": 0.8501, "step": 8985 }, { "epoch": 0.7042319749216301, "grad_norm": 5.700026035308838, "learning_rate": 9.26778611797252e-06, "loss": 1.1618, "step": 8986 }, { "epoch": 0.7043103448275863, "grad_norm": 4.640040874481201, "learning_rate": 9.267625453500218e-06, "loss": 1.3144, "step": 8987 }, { "epoch": 0.7043887147335424, "grad_norm": 2.763016939163208, "learning_rate": 9.267464772796058e-06, "loss": 1.2609, "step": 8988 }, { "epoch": 0.7044670846394985, "grad_norm": 2.290472984313965, "learning_rate": 9.267304075860653e-06, "loss": 1.2183, "step": 8989 }, { "epoch": 0.7045454545454546, "grad_norm": 2.715841770172119, "learning_rate": 9.267143362694618e-06, "loss": 1.8747, "step": 8990 }, { "epoch": 0.7046238244514107, "grad_norm": 2.9856112003326416, "learning_rate": 9.266982633298557e-06, "loss": 1.2316, "step": 8991 }, { "epoch": 0.7047021943573668, "grad_norm": 3.7708840370178223, "learning_rate": 9.26682188767309e-06, "loss": 1.5246, "step": 8992 }, { "epoch": 0.7047805642633229, "grad_norm": 2.672853708267212, "learning_rate": 9.266661125818823e-06, "loss": 1.4901, "step": 8993 }, { "epoch": 0.704858934169279, "grad_norm": 2.094237804412842, "learning_rate": 9.266500347736368e-06, "loss": 0.9968, "step": 8994 }, { "epoch": 0.7049373040752351, "grad_norm": 4.609612941741943, "learning_rate": 9.266339553426338e-06, "loss": 1.7412, "step": 8995 }, { "epoch": 0.7050156739811912, "grad_norm": 2.3921444416046143, "learning_rate": 9.266178742889343e-06, "loss": 1.6697, "step": 8996 }, { "epoch": 0.7050940438871474, "grad_norm": 6.738348484039307, "learning_rate": 9.266017916125995e-06, "loss": 1.3814, "step": 8997 }, { "epoch": 0.7051724137931035, "grad_norm": 2.3422012329101562, "learning_rate": 9.265857073136905e-06, "loss": 1.4698, "step": 8998 }, { "epoch": 0.7052507836990596, "grad_norm": 4.007188320159912, "learning_rate": 9.265696213922688e-06, "loss": 1.3849, "step": 8999 }, { "epoch": 0.7053291536050157, "grad_norm": 4.135525703430176, "learning_rate": 9.265535338483953e-06, "loss": 1.7229, "step": 9000 }, { "epoch": 0.7054075235109718, "grad_norm": 3.076504707336426, "learning_rate": 9.265374446821311e-06, "loss": 1.4112, "step": 9001 }, { "epoch": 0.7054858934169279, "grad_norm": 5.306149959564209, "learning_rate": 9.265213538935378e-06, "loss": 1.1868, "step": 9002 }, { "epoch": 0.705564263322884, "grad_norm": 2.3149774074554443, "learning_rate": 9.265052614826762e-06, "loss": 1.1995, "step": 9003 }, { "epoch": 0.7056426332288401, "grad_norm": 1.9296839237213135, "learning_rate": 9.264891674496077e-06, "loss": 1.2576, "step": 9004 }, { "epoch": 0.7057210031347962, "grad_norm": 2.5836846828460693, "learning_rate": 9.264730717943933e-06, "loss": 1.6634, "step": 9005 }, { "epoch": 0.7057993730407524, "grad_norm": 2.4956212043762207, "learning_rate": 9.264569745170947e-06, "loss": 1.6859, "step": 9006 }, { "epoch": 0.7058777429467085, "grad_norm": 2.2896220684051514, "learning_rate": 9.264408756177726e-06, "loss": 1.3534, "step": 9007 }, { "epoch": 0.7059561128526646, "grad_norm": 2.080314874649048, "learning_rate": 9.264247750964883e-06, "loss": 1.0719, "step": 9008 }, { "epoch": 0.7060344827586207, "grad_norm": 5.590830326080322, "learning_rate": 9.264086729533033e-06, "loss": 1.6083, "step": 9009 }, { "epoch": 0.7061128526645768, "grad_norm": 3.222604513168335, "learning_rate": 9.26392569188279e-06, "loss": 1.635, "step": 9010 }, { "epoch": 0.7061912225705329, "grad_norm": 2.5271713733673096, "learning_rate": 9.263764638014761e-06, "loss": 1.0337, "step": 9011 }, { "epoch": 0.706269592476489, "grad_norm": 10.146517753601074, "learning_rate": 9.26360356792956e-06, "loss": 1.7625, "step": 9012 }, { "epoch": 0.7063479623824451, "grad_norm": 5.116921424865723, "learning_rate": 9.263442481627804e-06, "loss": 1.4951, "step": 9013 }, { "epoch": 0.7064263322884012, "grad_norm": 2.299502372741699, "learning_rate": 9.2632813791101e-06, "loss": 1.2568, "step": 9014 }, { "epoch": 0.7065047021943573, "grad_norm": 2.1460378170013428, "learning_rate": 9.263120260377065e-06, "loss": 1.4102, "step": 9015 }, { "epoch": 0.7065830721003135, "grad_norm": 6.41417121887207, "learning_rate": 9.262959125429312e-06, "loss": 0.8604, "step": 9016 }, { "epoch": 0.7066614420062696, "grad_norm": 1.9811625480651855, "learning_rate": 9.262797974267449e-06, "loss": 1.5113, "step": 9017 }, { "epoch": 0.7067398119122257, "grad_norm": 2.724194049835205, "learning_rate": 9.262636806892093e-06, "loss": 1.4978, "step": 9018 }, { "epoch": 0.7068181818181818, "grad_norm": 3.6374309062957764, "learning_rate": 9.262475623303858e-06, "loss": 1.0758, "step": 9019 }, { "epoch": 0.7068965517241379, "grad_norm": 3.330259323120117, "learning_rate": 9.26231442350335e-06, "loss": 1.6708, "step": 9020 }, { "epoch": 0.706974921630094, "grad_norm": 4.582838535308838, "learning_rate": 9.262153207491192e-06, "loss": 0.9355, "step": 9021 }, { "epoch": 0.7070532915360501, "grad_norm": 5.13906717300415, "learning_rate": 9.261991975267992e-06, "loss": 1.2105, "step": 9022 }, { "epoch": 0.7071316614420062, "grad_norm": 3.311583995819092, "learning_rate": 9.261830726834361e-06, "loss": 1.2498, "step": 9023 }, { "epoch": 0.7072100313479623, "grad_norm": 2.616669178009033, "learning_rate": 9.261669462190917e-06, "loss": 1.4003, "step": 9024 }, { "epoch": 0.7072884012539185, "grad_norm": 4.139252662658691, "learning_rate": 9.26150818133827e-06, "loss": 1.6251, "step": 9025 }, { "epoch": 0.7073667711598746, "grad_norm": 2.952313184738159, "learning_rate": 9.261346884277037e-06, "loss": 1.4551, "step": 9026 }, { "epoch": 0.7074451410658307, "grad_norm": 2.962972640991211, "learning_rate": 9.261185571007826e-06, "loss": 1.2557, "step": 9027 }, { "epoch": 0.7075235109717868, "grad_norm": 2.2741153240203857, "learning_rate": 9.261024241531256e-06, "loss": 1.3007, "step": 9028 }, { "epoch": 0.7076018808777429, "grad_norm": 1.8188117742538452, "learning_rate": 9.260862895847938e-06, "loss": 1.2779, "step": 9029 }, { "epoch": 0.707680250783699, "grad_norm": 6.241110324859619, "learning_rate": 9.260701533958484e-06, "loss": 1.4787, "step": 9030 }, { "epoch": 0.7077586206896552, "grad_norm": 4.466590404510498, "learning_rate": 9.260540155863512e-06, "loss": 1.6414, "step": 9031 }, { "epoch": 0.7078369905956113, "grad_norm": 3.569772243499756, "learning_rate": 9.260378761563633e-06, "loss": 1.4203, "step": 9032 }, { "epoch": 0.7079153605015674, "grad_norm": 3.9074645042419434, "learning_rate": 9.260217351059462e-06, "loss": 1.2348, "step": 9033 }, { "epoch": 0.7079937304075236, "grad_norm": 4.168699741363525, "learning_rate": 9.260055924351612e-06, "loss": 1.4799, "step": 9034 }, { "epoch": 0.7080721003134797, "grad_norm": 2.855705976486206, "learning_rate": 9.259894481440698e-06, "loss": 1.1245, "step": 9035 }, { "epoch": 0.7081504702194358, "grad_norm": 3.852353572845459, "learning_rate": 9.259733022327334e-06, "loss": 1.8575, "step": 9036 }, { "epoch": 0.7082288401253919, "grad_norm": 4.612443923950195, "learning_rate": 9.25957154701213e-06, "loss": 1.1137, "step": 9037 }, { "epoch": 0.708307210031348, "grad_norm": 4.458227157592773, "learning_rate": 9.259410055495707e-06, "loss": 2.1105, "step": 9038 }, { "epoch": 0.7083855799373041, "grad_norm": 2.123905897140503, "learning_rate": 9.259248547778674e-06, "loss": 1.3215, "step": 9039 }, { "epoch": 0.7084639498432602, "grad_norm": 2.6424431800842285, "learning_rate": 9.259087023861649e-06, "loss": 1.6855, "step": 9040 }, { "epoch": 0.7085423197492163, "grad_norm": 3.344851493835449, "learning_rate": 9.258925483745243e-06, "loss": 1.229, "step": 9041 }, { "epoch": 0.7086206896551724, "grad_norm": 2.679572105407715, "learning_rate": 9.258763927430073e-06, "loss": 1.3203, "step": 9042 }, { "epoch": 0.7086990595611286, "grad_norm": 2.0888302326202393, "learning_rate": 9.258602354916753e-06, "loss": 1.3703, "step": 9043 }, { "epoch": 0.7087774294670847, "grad_norm": 2.2316272258758545, "learning_rate": 9.258440766205894e-06, "loss": 1.4994, "step": 9044 }, { "epoch": 0.7088557993730408, "grad_norm": 7.2299580574035645, "learning_rate": 9.258279161298118e-06, "loss": 1.1143, "step": 9045 }, { "epoch": 0.7089341692789969, "grad_norm": 2.257347345352173, "learning_rate": 9.25811754019403e-06, "loss": 1.3819, "step": 9046 }, { "epoch": 0.709012539184953, "grad_norm": 2.463603973388672, "learning_rate": 9.257955902894255e-06, "loss": 1.1982, "step": 9047 }, { "epoch": 0.7090909090909091, "grad_norm": 2.914513111114502, "learning_rate": 9.257794249399402e-06, "loss": 1.4073, "step": 9048 }, { "epoch": 0.7091692789968652, "grad_norm": 2.170224905014038, "learning_rate": 9.257632579710084e-06, "loss": 1.3384, "step": 9049 }, { "epoch": 0.7092476489028213, "grad_norm": 8.334477424621582, "learning_rate": 9.25747089382692e-06, "loss": 1.7332, "step": 9050 }, { "epoch": 0.7093260188087774, "grad_norm": 3.094796895980835, "learning_rate": 9.257309191750522e-06, "loss": 1.5549, "step": 9051 }, { "epoch": 0.7094043887147335, "grad_norm": 1.7046668529510498, "learning_rate": 9.257147473481509e-06, "loss": 1.2751, "step": 9052 }, { "epoch": 0.7094827586206897, "grad_norm": 5.258345127105713, "learning_rate": 9.256985739020494e-06, "loss": 1.3413, "step": 9053 }, { "epoch": 0.7095611285266458, "grad_norm": 2.7094156742095947, "learning_rate": 9.25682398836809e-06, "loss": 1.2369, "step": 9054 }, { "epoch": 0.7096394984326019, "grad_norm": 3.8358137607574463, "learning_rate": 9.256662221524913e-06, "loss": 0.7859, "step": 9055 }, { "epoch": 0.709717868338558, "grad_norm": 3.521207571029663, "learning_rate": 9.256500438491581e-06, "loss": 1.5605, "step": 9056 }, { "epoch": 0.7097962382445141, "grad_norm": 2.788243532180786, "learning_rate": 9.256338639268707e-06, "loss": 1.4057, "step": 9057 }, { "epoch": 0.7098746081504702, "grad_norm": 2.7116899490356445, "learning_rate": 9.256176823856907e-06, "loss": 1.1412, "step": 9058 }, { "epoch": 0.7099529780564263, "grad_norm": 2.617156505584717, "learning_rate": 9.256014992256796e-06, "loss": 1.3338, "step": 9059 }, { "epoch": 0.7100313479623824, "grad_norm": 2.9546799659729004, "learning_rate": 9.255853144468991e-06, "loss": 1.7944, "step": 9060 }, { "epoch": 0.7101097178683385, "grad_norm": 2.4491772651672363, "learning_rate": 9.255691280494107e-06, "loss": 1.3644, "step": 9061 }, { "epoch": 0.7101880877742947, "grad_norm": 2.074627637863159, "learning_rate": 9.255529400332759e-06, "loss": 1.1429, "step": 9062 }, { "epoch": 0.7102664576802508, "grad_norm": 2.353682518005371, "learning_rate": 9.255367503985562e-06, "loss": 1.4246, "step": 9063 }, { "epoch": 0.7103448275862069, "grad_norm": 2.780787229537964, "learning_rate": 9.255205591453133e-06, "loss": 0.8109, "step": 9064 }, { "epoch": 0.710423197492163, "grad_norm": 3.406691551208496, "learning_rate": 9.255043662736087e-06, "loss": 1.1763, "step": 9065 }, { "epoch": 0.7105015673981191, "grad_norm": 3.2712149620056152, "learning_rate": 9.254881717835042e-06, "loss": 1.3197, "step": 9066 }, { "epoch": 0.7105799373040752, "grad_norm": 2.4454526901245117, "learning_rate": 9.254719756750612e-06, "loss": 1.4711, "step": 9067 }, { "epoch": 0.7106583072100313, "grad_norm": 3.1944186687469482, "learning_rate": 9.254557779483413e-06, "loss": 1.4006, "step": 9068 }, { "epoch": 0.7107366771159874, "grad_norm": 2.8085556030273438, "learning_rate": 9.254395786034061e-06, "loss": 1.581, "step": 9069 }, { "epoch": 0.7108150470219435, "grad_norm": 2.305105447769165, "learning_rate": 9.254233776403172e-06, "loss": 1.6305, "step": 9070 }, { "epoch": 0.7108934169278996, "grad_norm": 30.62647819519043, "learning_rate": 9.254071750591364e-06, "loss": 0.7415, "step": 9071 }, { "epoch": 0.7109717868338558, "grad_norm": 3.6765964031219482, "learning_rate": 9.253909708599252e-06, "loss": 1.271, "step": 9072 }, { "epoch": 0.7110501567398119, "grad_norm": 4.075085639953613, "learning_rate": 9.253747650427453e-06, "loss": 1.0521, "step": 9073 }, { "epoch": 0.711128526645768, "grad_norm": 2.3333685398101807, "learning_rate": 9.253585576076582e-06, "loss": 1.1313, "step": 9074 }, { "epoch": 0.7112068965517241, "grad_norm": 2.9896633625030518, "learning_rate": 9.253423485547258e-06, "loss": 1.533, "step": 9075 }, { "epoch": 0.7112852664576802, "grad_norm": 2.217052459716797, "learning_rate": 9.253261378840093e-06, "loss": 1.3077, "step": 9076 }, { "epoch": 0.7113636363636363, "grad_norm": 2.447892427444458, "learning_rate": 9.253099255955706e-06, "loss": 1.1866, "step": 9077 }, { "epoch": 0.7114420062695925, "grad_norm": 3.890218734741211, "learning_rate": 9.252937116894716e-06, "loss": 1.6004, "step": 9078 }, { "epoch": 0.7115203761755486, "grad_norm": 2.5844874382019043, "learning_rate": 9.252774961657738e-06, "loss": 1.2193, "step": 9079 }, { "epoch": 0.7115987460815048, "grad_norm": 3.2502169609069824, "learning_rate": 9.252612790245388e-06, "loss": 1.5394, "step": 9080 }, { "epoch": 0.7116771159874609, "grad_norm": 2.377185106277466, "learning_rate": 9.252450602658282e-06, "loss": 1.1112, "step": 9081 }, { "epoch": 0.711755485893417, "grad_norm": 2.0063350200653076, "learning_rate": 9.252288398897038e-06, "loss": 1.5267, "step": 9082 }, { "epoch": 0.7118338557993731, "grad_norm": 2.7935380935668945, "learning_rate": 9.252126178962275e-06, "loss": 1.5912, "step": 9083 }, { "epoch": 0.7119122257053292, "grad_norm": 2.65653395652771, "learning_rate": 9.251963942854607e-06, "loss": 1.563, "step": 9084 }, { "epoch": 0.7119905956112853, "grad_norm": 5.464169979095459, "learning_rate": 9.251801690574652e-06, "loss": 1.0277, "step": 9085 }, { "epoch": 0.7120689655172414, "grad_norm": 1.9924451112747192, "learning_rate": 9.251639422123026e-06, "loss": 1.2455, "step": 9086 }, { "epoch": 0.7121473354231975, "grad_norm": 2.499669313430786, "learning_rate": 9.251477137500349e-06, "loss": 1.3107, "step": 9087 }, { "epoch": 0.7122257053291536, "grad_norm": 1.8948769569396973, "learning_rate": 9.251314836707237e-06, "loss": 1.4892, "step": 9088 }, { "epoch": 0.7123040752351097, "grad_norm": 2.361531972885132, "learning_rate": 9.251152519744306e-06, "loss": 1.3219, "step": 9089 }, { "epoch": 0.7123824451410659, "grad_norm": 3.071012020111084, "learning_rate": 9.250990186612174e-06, "loss": 1.3869, "step": 9090 }, { "epoch": 0.712460815047022, "grad_norm": 4.275446891784668, "learning_rate": 9.25082783731146e-06, "loss": 1.0638, "step": 9091 }, { "epoch": 0.7125391849529781, "grad_norm": 2.8540921211242676, "learning_rate": 9.250665471842778e-06, "loss": 1.2807, "step": 9092 }, { "epoch": 0.7126175548589342, "grad_norm": 3.347043752670288, "learning_rate": 9.250503090206751e-06, "loss": 1.4492, "step": 9093 }, { "epoch": 0.7126959247648903, "grad_norm": 2.488344669342041, "learning_rate": 9.250340692403992e-06, "loss": 1.2783, "step": 9094 }, { "epoch": 0.7127742946708464, "grad_norm": 6.0329203605651855, "learning_rate": 9.25017827843512e-06, "loss": 1.3922, "step": 9095 }, { "epoch": 0.7128526645768025, "grad_norm": 2.6307051181793213, "learning_rate": 9.250015848300753e-06, "loss": 1.0021, "step": 9096 }, { "epoch": 0.7129310344827586, "grad_norm": 3.2232604026794434, "learning_rate": 9.249853402001509e-06, "loss": 1.3201, "step": 9097 }, { "epoch": 0.7130094043887147, "grad_norm": 6.220031261444092, "learning_rate": 9.249690939538005e-06, "loss": 1.4105, "step": 9098 }, { "epoch": 0.7130877742946709, "grad_norm": 2.282823324203491, "learning_rate": 9.249528460910859e-06, "loss": 1.1783, "step": 9099 }, { "epoch": 0.713166144200627, "grad_norm": 2.4254310131073, "learning_rate": 9.249365966120692e-06, "loss": 1.2875, "step": 9100 }, { "epoch": 0.7132445141065831, "grad_norm": 4.395481109619141, "learning_rate": 9.249203455168117e-06, "loss": 1.7808, "step": 9101 }, { "epoch": 0.7133228840125392, "grad_norm": 5.191954135894775, "learning_rate": 9.249040928053755e-06, "loss": 1.173, "step": 9102 }, { "epoch": 0.7134012539184953, "grad_norm": 5.81416130065918, "learning_rate": 9.248878384778223e-06, "loss": 0.7497, "step": 9103 }, { "epoch": 0.7134796238244514, "grad_norm": 2.5232439041137695, "learning_rate": 9.24871582534214e-06, "loss": 1.172, "step": 9104 }, { "epoch": 0.7135579937304075, "grad_norm": 3.696725368499756, "learning_rate": 9.248553249746126e-06, "loss": 1.4907, "step": 9105 }, { "epoch": 0.7136363636363636, "grad_norm": 5.221175670623779, "learning_rate": 9.248390657990796e-06, "loss": 1.878, "step": 9106 }, { "epoch": 0.7137147335423197, "grad_norm": 3.673306465148926, "learning_rate": 9.248228050076771e-06, "loss": 0.982, "step": 9107 }, { "epoch": 0.7137931034482758, "grad_norm": 4.420069217681885, "learning_rate": 9.248065426004668e-06, "loss": 1.397, "step": 9108 }, { "epoch": 0.713871473354232, "grad_norm": 2.7346580028533936, "learning_rate": 9.247902785775107e-06, "loss": 1.3978, "step": 9109 }, { "epoch": 0.7139498432601881, "grad_norm": 5.646728515625, "learning_rate": 9.247740129388705e-06, "loss": 1.8558, "step": 9110 }, { "epoch": 0.7140282131661442, "grad_norm": 3.125994920730591, "learning_rate": 9.247577456846081e-06, "loss": 1.2241, "step": 9111 }, { "epoch": 0.7141065830721003, "grad_norm": 8.61258602142334, "learning_rate": 9.247414768147855e-06, "loss": 1.2616, "step": 9112 }, { "epoch": 0.7141849529780564, "grad_norm": 2.944247007369995, "learning_rate": 9.247252063294644e-06, "loss": 1.3524, "step": 9113 }, { "epoch": 0.7142633228840125, "grad_norm": 3.9404726028442383, "learning_rate": 9.247089342287068e-06, "loss": 1.5818, "step": 9114 }, { "epoch": 0.7143416927899686, "grad_norm": 4.908975601196289, "learning_rate": 9.246926605125746e-06, "loss": 1.7305, "step": 9115 }, { "epoch": 0.7144200626959247, "grad_norm": 2.7286698818206787, "learning_rate": 9.246763851811295e-06, "loss": 0.9792, "step": 9116 }, { "epoch": 0.7144984326018808, "grad_norm": 1.955358624458313, "learning_rate": 9.246601082344337e-06, "loss": 1.3059, "step": 9117 }, { "epoch": 0.714576802507837, "grad_norm": 4.723377704620361, "learning_rate": 9.24643829672549e-06, "loss": 1.722, "step": 9118 }, { "epoch": 0.7146551724137931, "grad_norm": 3.002277374267578, "learning_rate": 9.246275494955373e-06, "loss": 1.0195, "step": 9119 }, { "epoch": 0.7147335423197492, "grad_norm": 3.5391786098480225, "learning_rate": 9.246112677034603e-06, "loss": 1.273, "step": 9120 }, { "epoch": 0.7148119122257053, "grad_norm": 4.315385341644287, "learning_rate": 9.245949842963802e-06, "loss": 1.5152, "step": 9121 }, { "epoch": 0.7148902821316614, "grad_norm": 3.101553201675415, "learning_rate": 9.24578699274359e-06, "loss": 1.274, "step": 9122 }, { "epoch": 0.7149686520376175, "grad_norm": 2.189728021621704, "learning_rate": 9.245624126374584e-06, "loss": 1.5402, "step": 9123 }, { "epoch": 0.7150470219435736, "grad_norm": 3.4246647357940674, "learning_rate": 9.245461243857406e-06, "loss": 1.4813, "step": 9124 }, { "epoch": 0.7151253918495298, "grad_norm": 2.5752975940704346, "learning_rate": 9.24529834519267e-06, "loss": 1.4418, "step": 9125 }, { "epoch": 0.715203761755486, "grad_norm": 1.9549944400787354, "learning_rate": 9.245135430381005e-06, "loss": 1.017, "step": 9126 }, { "epoch": 0.715282131661442, "grad_norm": 3.3827121257781982, "learning_rate": 9.244972499423022e-06, "loss": 1.6415, "step": 9127 }, { "epoch": 0.7153605015673982, "grad_norm": 2.437403917312622, "learning_rate": 9.244809552319344e-06, "loss": 1.3221, "step": 9128 }, { "epoch": 0.7154388714733543, "grad_norm": 2.9734420776367188, "learning_rate": 9.244646589070592e-06, "loss": 1.0583, "step": 9129 }, { "epoch": 0.7155172413793104, "grad_norm": 2.4505856037139893, "learning_rate": 9.244483609677384e-06, "loss": 1.2449, "step": 9130 }, { "epoch": 0.7155956112852665, "grad_norm": 2.0436089038848877, "learning_rate": 9.244320614140339e-06, "loss": 1.2231, "step": 9131 }, { "epoch": 0.7156739811912226, "grad_norm": 4.314119815826416, "learning_rate": 9.24415760246008e-06, "loss": 1.229, "step": 9132 }, { "epoch": 0.7157523510971787, "grad_norm": 2.268622636795044, "learning_rate": 9.243994574637225e-06, "loss": 1.74, "step": 9133 }, { "epoch": 0.7158307210031348, "grad_norm": 2.1376445293426514, "learning_rate": 9.243831530672395e-06, "loss": 1.322, "step": 9134 }, { "epoch": 0.7159090909090909, "grad_norm": 3.0539052486419678, "learning_rate": 9.24366847056621e-06, "loss": 1.0229, "step": 9135 }, { "epoch": 0.715987460815047, "grad_norm": 2.6001172065734863, "learning_rate": 9.24350539431929e-06, "loss": 1.0106, "step": 9136 }, { "epoch": 0.7160658307210032, "grad_norm": 3.3054752349853516, "learning_rate": 9.243342301932251e-06, "loss": 0.7884, "step": 9137 }, { "epoch": 0.7161442006269593, "grad_norm": 2.7023379802703857, "learning_rate": 9.243179193405722e-06, "loss": 1.4421, "step": 9138 }, { "epoch": 0.7162225705329154, "grad_norm": 3.6297762393951416, "learning_rate": 9.243016068740318e-06, "loss": 1.3711, "step": 9139 }, { "epoch": 0.7163009404388715, "grad_norm": 1.8910452127456665, "learning_rate": 9.24285292793666e-06, "loss": 1.4712, "step": 9140 }, { "epoch": 0.7163793103448276, "grad_norm": 2.5957818031311035, "learning_rate": 9.242689770995367e-06, "loss": 1.5804, "step": 9141 }, { "epoch": 0.7164576802507837, "grad_norm": 3.0943589210510254, "learning_rate": 9.242526597917064e-06, "loss": 1.3003, "step": 9142 }, { "epoch": 0.7165360501567398, "grad_norm": 3.6060678958892822, "learning_rate": 9.242363408702366e-06, "loss": 1.5939, "step": 9143 }, { "epoch": 0.7166144200626959, "grad_norm": 2.4141790866851807, "learning_rate": 9.242200203351898e-06, "loss": 1.4528, "step": 9144 }, { "epoch": 0.716692789968652, "grad_norm": 2.363086700439453, "learning_rate": 9.24203698186628e-06, "loss": 1.0353, "step": 9145 }, { "epoch": 0.7167711598746082, "grad_norm": 3.8575899600982666, "learning_rate": 9.241873744246133e-06, "loss": 1.4875, "step": 9146 }, { "epoch": 0.7168495297805643, "grad_norm": 3.2304577827453613, "learning_rate": 9.241710490492076e-06, "loss": 1.3102, "step": 9147 }, { "epoch": 0.7169278996865204, "grad_norm": 3.739332437515259, "learning_rate": 9.24154722060473e-06, "loss": 1.8036, "step": 9148 }, { "epoch": 0.7170062695924765, "grad_norm": 4.9688286781311035, "learning_rate": 9.241383934584718e-06, "loss": 1.4707, "step": 9149 }, { "epoch": 0.7170846394984326, "grad_norm": 3.564676284790039, "learning_rate": 9.24122063243266e-06, "loss": 1.5369, "step": 9150 }, { "epoch": 0.7171630094043887, "grad_norm": 2.6643619537353516, "learning_rate": 9.241057314149176e-06, "loss": 1.0262, "step": 9151 }, { "epoch": 0.7172413793103448, "grad_norm": 2.6977133750915527, "learning_rate": 9.240893979734888e-06, "loss": 1.261, "step": 9152 }, { "epoch": 0.7173197492163009, "grad_norm": 17.899106979370117, "learning_rate": 9.24073062919042e-06, "loss": 0.7, "step": 9153 }, { "epoch": 0.717398119122257, "grad_norm": 8.218905448913574, "learning_rate": 9.240567262516389e-06, "loss": 0.6588, "step": 9154 }, { "epoch": 0.7174764890282131, "grad_norm": 2.2050507068634033, "learning_rate": 9.24040387971342e-06, "loss": 1.1063, "step": 9155 }, { "epoch": 0.7175548589341693, "grad_norm": 3.7973906993865967, "learning_rate": 9.24024048078213e-06, "loss": 1.2726, "step": 9156 }, { "epoch": 0.7176332288401254, "grad_norm": 2.900503635406494, "learning_rate": 9.240077065723145e-06, "loss": 1.3024, "step": 9157 }, { "epoch": 0.7177115987460815, "grad_norm": 2.442305564880371, "learning_rate": 9.239913634537084e-06, "loss": 1.7835, "step": 9158 }, { "epoch": 0.7177899686520376, "grad_norm": 4.708406925201416, "learning_rate": 9.23975018722457e-06, "loss": 1.8695, "step": 9159 }, { "epoch": 0.7178683385579937, "grad_norm": 2.5433454513549805, "learning_rate": 9.239586723786223e-06, "loss": 1.5037, "step": 9160 }, { "epoch": 0.7179467084639498, "grad_norm": 3.083799123764038, "learning_rate": 9.239423244222668e-06, "loss": 1.2858, "step": 9161 }, { "epoch": 0.7180250783699059, "grad_norm": 2.6963624954223633, "learning_rate": 9.239259748534521e-06, "loss": 1.3483, "step": 9162 }, { "epoch": 0.718103448275862, "grad_norm": 2.987534761428833, "learning_rate": 9.239096236722409e-06, "loss": 1.4859, "step": 9163 }, { "epoch": 0.7181818181818181, "grad_norm": 2.5889289379119873, "learning_rate": 9.238932708786953e-06, "loss": 1.4794, "step": 9164 }, { "epoch": 0.7182601880877743, "grad_norm": 3.937117338180542, "learning_rate": 9.238769164728774e-06, "loss": 1.858, "step": 9165 }, { "epoch": 0.7183385579937304, "grad_norm": 2.4404296875, "learning_rate": 9.238605604548494e-06, "loss": 1.0949, "step": 9166 }, { "epoch": 0.7184169278996865, "grad_norm": 2.3174219131469727, "learning_rate": 9.238442028246734e-06, "loss": 1.0618, "step": 9167 }, { "epoch": 0.7184952978056426, "grad_norm": 3.0113308429718018, "learning_rate": 9.238278435824119e-06, "loss": 1.5022, "step": 9168 }, { "epoch": 0.7185736677115987, "grad_norm": 2.6261143684387207, "learning_rate": 9.238114827281271e-06, "loss": 1.359, "step": 9169 }, { "epoch": 0.7186520376175548, "grad_norm": 1.6707913875579834, "learning_rate": 9.237951202618809e-06, "loss": 1.3761, "step": 9170 }, { "epoch": 0.7187304075235109, "grad_norm": 4.88236141204834, "learning_rate": 9.237787561837356e-06, "loss": 0.7971, "step": 9171 }, { "epoch": 0.7188087774294671, "grad_norm": 2.5907208919525146, "learning_rate": 9.237623904937538e-06, "loss": 1.3613, "step": 9172 }, { "epoch": 0.7188871473354232, "grad_norm": 2.0328452587127686, "learning_rate": 9.237460231919974e-06, "loss": 1.3817, "step": 9173 }, { "epoch": 0.7189655172413794, "grad_norm": 3.7938597202301025, "learning_rate": 9.23729654278529e-06, "loss": 1.2031, "step": 9174 }, { "epoch": 0.7190438871473355, "grad_norm": 2.687807321548462, "learning_rate": 9.237132837534106e-06, "loss": 1.0512, "step": 9175 }, { "epoch": 0.7191222570532916, "grad_norm": 2.2924351692199707, "learning_rate": 9.236969116167043e-06, "loss": 1.1639, "step": 9176 }, { "epoch": 0.7192006269592477, "grad_norm": 3.0615382194519043, "learning_rate": 9.236805378684727e-06, "loss": 1.4565, "step": 9177 }, { "epoch": 0.7192789968652038, "grad_norm": 1.994323968887329, "learning_rate": 9.236641625087778e-06, "loss": 1.1761, "step": 9178 }, { "epoch": 0.7193573667711599, "grad_norm": 3.6821036338806152, "learning_rate": 9.236477855376822e-06, "loss": 1.7768, "step": 9179 }, { "epoch": 0.719435736677116, "grad_norm": 2.525775194168091, "learning_rate": 9.23631406955248e-06, "loss": 1.3782, "step": 9180 }, { "epoch": 0.7195141065830721, "grad_norm": 6.60976505279541, "learning_rate": 9.236150267615375e-06, "loss": 1.3115, "step": 9181 }, { "epoch": 0.7195924764890282, "grad_norm": 3.730879545211792, "learning_rate": 9.23598644956613e-06, "loss": 1.8312, "step": 9182 }, { "epoch": 0.7196708463949844, "grad_norm": 5.60256814956665, "learning_rate": 9.23582261540537e-06, "loss": 1.7587, "step": 9183 }, { "epoch": 0.7197492163009405, "grad_norm": 1.971343994140625, "learning_rate": 9.235658765133715e-06, "loss": 1.1011, "step": 9184 }, { "epoch": 0.7198275862068966, "grad_norm": 2.1580467224121094, "learning_rate": 9.235494898751788e-06, "loss": 1.418, "step": 9185 }, { "epoch": 0.7199059561128527, "grad_norm": 7.262831687927246, "learning_rate": 9.235331016260216e-06, "loss": 1.1759, "step": 9186 }, { "epoch": 0.7199843260188088, "grad_norm": 2.8830440044403076, "learning_rate": 9.23516711765962e-06, "loss": 1.8757, "step": 9187 }, { "epoch": 0.7200626959247649, "grad_norm": 1.9973596334457397, "learning_rate": 9.235003202950623e-06, "loss": 1.4925, "step": 9188 }, { "epoch": 0.720141065830721, "grad_norm": 6.809404373168945, "learning_rate": 9.23483927213385e-06, "loss": 1.5177, "step": 9189 }, { "epoch": 0.7202194357366771, "grad_norm": 5.480515956878662, "learning_rate": 9.234675325209923e-06, "loss": 2.0443, "step": 9190 }, { "epoch": 0.7202978056426332, "grad_norm": 2.4778175354003906, "learning_rate": 9.234511362179467e-06, "loss": 1.5117, "step": 9191 }, { "epoch": 0.7203761755485893, "grad_norm": 6.002614974975586, "learning_rate": 9.234347383043102e-06, "loss": 1.5954, "step": 9192 }, { "epoch": 0.7204545454545455, "grad_norm": 2.602964162826538, "learning_rate": 9.234183387801457e-06, "loss": 1.6187, "step": 9193 }, { "epoch": 0.7205329153605016, "grad_norm": 4.606175899505615, "learning_rate": 9.234019376455152e-06, "loss": 0.8954, "step": 9194 }, { "epoch": 0.7206112852664577, "grad_norm": 6.6400041580200195, "learning_rate": 9.233855349004812e-06, "loss": 1.4519, "step": 9195 }, { "epoch": 0.7206896551724138, "grad_norm": 2.475189685821533, "learning_rate": 9.233691305451061e-06, "loss": 1.6046, "step": 9196 }, { "epoch": 0.7207680250783699, "grad_norm": 3.005887269973755, "learning_rate": 9.233527245794524e-06, "loss": 1.2831, "step": 9197 }, { "epoch": 0.720846394984326, "grad_norm": 2.847843885421753, "learning_rate": 9.233363170035822e-06, "loss": 1.527, "step": 9198 }, { "epoch": 0.7209247648902821, "grad_norm": 2.3508107662200928, "learning_rate": 9.233199078175583e-06, "loss": 1.0623, "step": 9199 }, { "epoch": 0.7210031347962382, "grad_norm": 8.207636833190918, "learning_rate": 9.233034970214427e-06, "loss": 1.5386, "step": 9200 }, { "epoch": 0.7210815047021943, "grad_norm": 3.556337833404541, "learning_rate": 9.232870846152982e-06, "loss": 1.0677, "step": 9201 }, { "epoch": 0.7211598746081505, "grad_norm": 3.340618371963501, "learning_rate": 9.23270670599187e-06, "loss": 1.5767, "step": 9202 }, { "epoch": 0.7212382445141066, "grad_norm": 3.8726725578308105, "learning_rate": 9.232542549731715e-06, "loss": 1.3205, "step": 9203 }, { "epoch": 0.7213166144200627, "grad_norm": 3.2039570808410645, "learning_rate": 9.232378377373142e-06, "loss": 1.1711, "step": 9204 }, { "epoch": 0.7213949843260188, "grad_norm": 4.240915298461914, "learning_rate": 9.232214188916776e-06, "loss": 1.6099, "step": 9205 }, { "epoch": 0.7214733542319749, "grad_norm": 4.292571544647217, "learning_rate": 9.23204998436324e-06, "loss": 1.8033, "step": 9206 }, { "epoch": 0.721551724137931, "grad_norm": 3.405913829803467, "learning_rate": 9.231885763713162e-06, "loss": 1.335, "step": 9207 }, { "epoch": 0.7216300940438871, "grad_norm": 4.900636672973633, "learning_rate": 9.231721526967163e-06, "loss": 1.0985, "step": 9208 }, { "epoch": 0.7217084639498432, "grad_norm": 6.074000358581543, "learning_rate": 9.231557274125868e-06, "loss": 1.4614, "step": 9209 }, { "epoch": 0.7217868338557993, "grad_norm": 2.7278356552124023, "learning_rate": 9.231393005189903e-06, "loss": 1.7384, "step": 9210 }, { "epoch": 0.7218652037617554, "grad_norm": 4.552612781524658, "learning_rate": 9.231228720159893e-06, "loss": 1.6969, "step": 9211 }, { "epoch": 0.7219435736677116, "grad_norm": 3.3927037715911865, "learning_rate": 9.231064419036462e-06, "loss": 1.7519, "step": 9212 }, { "epoch": 0.7220219435736677, "grad_norm": 2.5497164726257324, "learning_rate": 9.230900101820235e-06, "loss": 1.1505, "step": 9213 }, { "epoch": 0.7221003134796238, "grad_norm": 2.146148920059204, "learning_rate": 9.230735768511836e-06, "loss": 1.5509, "step": 9214 }, { "epoch": 0.7221786833855799, "grad_norm": 2.9312071800231934, "learning_rate": 9.230571419111894e-06, "loss": 1.2255, "step": 9215 }, { "epoch": 0.722257053291536, "grad_norm": 2.933638572692871, "learning_rate": 9.230407053621029e-06, "loss": 1.1376, "step": 9216 }, { "epoch": 0.7223354231974921, "grad_norm": 2.8083128929138184, "learning_rate": 9.230242672039869e-06, "loss": 1.1924, "step": 9217 }, { "epoch": 0.7224137931034482, "grad_norm": 3.131563186645508, "learning_rate": 9.230078274369038e-06, "loss": 1.5338, "step": 9218 }, { "epoch": 0.7224921630094044, "grad_norm": 2.270034074783325, "learning_rate": 9.229913860609163e-06, "loss": 1.3682, "step": 9219 }, { "epoch": 0.7225705329153606, "grad_norm": 2.9183740615844727, "learning_rate": 9.229749430760868e-06, "loss": 1.4369, "step": 9220 }, { "epoch": 0.7226489028213167, "grad_norm": 5.507895469665527, "learning_rate": 9.229584984824779e-06, "loss": 1.1004, "step": 9221 }, { "epoch": 0.7227272727272728, "grad_norm": 3.097848892211914, "learning_rate": 9.22942052280152e-06, "loss": 1.1483, "step": 9222 }, { "epoch": 0.7228056426332289, "grad_norm": 3.0186681747436523, "learning_rate": 9.22925604469172e-06, "loss": 1.4362, "step": 9223 }, { "epoch": 0.722884012539185, "grad_norm": 3.2588050365448, "learning_rate": 9.229091550495999e-06, "loss": 1.6293, "step": 9224 }, { "epoch": 0.7229623824451411, "grad_norm": 2.306497573852539, "learning_rate": 9.22892704021499e-06, "loss": 1.3761, "step": 9225 }, { "epoch": 0.7230407523510972, "grad_norm": 2.1122446060180664, "learning_rate": 9.22876251384931e-06, "loss": 1.2455, "step": 9226 }, { "epoch": 0.7231191222570533, "grad_norm": 2.3730130195617676, "learning_rate": 9.228597971399593e-06, "loss": 1.6308, "step": 9227 }, { "epoch": 0.7231974921630094, "grad_norm": 4.206284046173096, "learning_rate": 9.22843341286646e-06, "loss": 1.4838, "step": 9228 }, { "epoch": 0.7232758620689655, "grad_norm": 3.063925266265869, "learning_rate": 9.228268838250539e-06, "loss": 1.327, "step": 9229 }, { "epoch": 0.7233542319749217, "grad_norm": 3.254897356033325, "learning_rate": 9.228104247552455e-06, "loss": 1.2463, "step": 9230 }, { "epoch": 0.7234326018808778, "grad_norm": 3.462955951690674, "learning_rate": 9.227939640772832e-06, "loss": 1.5456, "step": 9231 }, { "epoch": 0.7235109717868339, "grad_norm": 3.3880503177642822, "learning_rate": 9.227775017912299e-06, "loss": 1.5434, "step": 9232 }, { "epoch": 0.72358934169279, "grad_norm": 4.639939785003662, "learning_rate": 9.227610378971484e-06, "loss": 1.4002, "step": 9233 }, { "epoch": 0.7236677115987461, "grad_norm": 7.328334331512451, "learning_rate": 9.227445723951007e-06, "loss": 0.6885, "step": 9234 }, { "epoch": 0.7237460815047022, "grad_norm": 3.223276138305664, "learning_rate": 9.2272810528515e-06, "loss": 1.5694, "step": 9235 }, { "epoch": 0.7238244514106583, "grad_norm": 2.620905637741089, "learning_rate": 9.227116365673587e-06, "loss": 1.5097, "step": 9236 }, { "epoch": 0.7239028213166144, "grad_norm": 3.29343581199646, "learning_rate": 9.226951662417894e-06, "loss": 1.1875, "step": 9237 }, { "epoch": 0.7239811912225705, "grad_norm": 2.376037836074829, "learning_rate": 9.226786943085048e-06, "loss": 1.4706, "step": 9238 }, { "epoch": 0.7240595611285267, "grad_norm": 3.0333731174468994, "learning_rate": 9.226622207675674e-06, "loss": 1.6189, "step": 9239 }, { "epoch": 0.7241379310344828, "grad_norm": 1.9203200340270996, "learning_rate": 9.226457456190403e-06, "loss": 1.428, "step": 9240 }, { "epoch": 0.7242163009404389, "grad_norm": 2.4003937244415283, "learning_rate": 9.226292688629858e-06, "loss": 1.4139, "step": 9241 }, { "epoch": 0.724294670846395, "grad_norm": 1.8575901985168457, "learning_rate": 9.226127904994665e-06, "loss": 1.1288, "step": 9242 }, { "epoch": 0.7243730407523511, "grad_norm": 2.249119758605957, "learning_rate": 9.225963105285453e-06, "loss": 1.2796, "step": 9243 }, { "epoch": 0.7244514106583072, "grad_norm": 2.9313371181488037, "learning_rate": 9.225798289502848e-06, "loss": 1.3958, "step": 9244 }, { "epoch": 0.7245297805642633, "grad_norm": 2.790081024169922, "learning_rate": 9.225633457647475e-06, "loss": 1.5446, "step": 9245 }, { "epoch": 0.7246081504702194, "grad_norm": 1.9849205017089844, "learning_rate": 9.225468609719966e-06, "loss": 1.4116, "step": 9246 }, { "epoch": 0.7246865203761755, "grad_norm": 3.7808332443237305, "learning_rate": 9.225303745720942e-06, "loss": 0.9861, "step": 9247 }, { "epoch": 0.7247648902821316, "grad_norm": 2.394418239593506, "learning_rate": 9.225138865651034e-06, "loss": 1.4971, "step": 9248 }, { "epoch": 0.7248432601880878, "grad_norm": 2.9685375690460205, "learning_rate": 9.224973969510869e-06, "loss": 1.2966, "step": 9249 }, { "epoch": 0.7249216300940439, "grad_norm": 3.3019731044769287, "learning_rate": 9.224809057301072e-06, "loss": 1.6454, "step": 9250 }, { "epoch": 0.725, "grad_norm": 2.2301671504974365, "learning_rate": 9.224644129022271e-06, "loss": 1.7191, "step": 9251 }, { "epoch": 0.7250783699059561, "grad_norm": 2.9796254634857178, "learning_rate": 9.224479184675094e-06, "loss": 1.1967, "step": 9252 }, { "epoch": 0.7251567398119122, "grad_norm": 2.8885130882263184, "learning_rate": 9.224314224260166e-06, "loss": 1.7011, "step": 9253 }, { "epoch": 0.7252351097178683, "grad_norm": 2.5855040550231934, "learning_rate": 9.22414924777812e-06, "loss": 1.058, "step": 9254 }, { "epoch": 0.7253134796238244, "grad_norm": 17.68362045288086, "learning_rate": 9.223984255229578e-06, "loss": 0.8495, "step": 9255 }, { "epoch": 0.7253918495297805, "grad_norm": 3.1011083126068115, "learning_rate": 9.22381924661517e-06, "loss": 1.0075, "step": 9256 }, { "epoch": 0.7254702194357366, "grad_norm": 2.5147619247436523, "learning_rate": 9.223654221935523e-06, "loss": 1.3372, "step": 9257 }, { "epoch": 0.7255485893416928, "grad_norm": 2.6641690731048584, "learning_rate": 9.223489181191265e-06, "loss": 1.4571, "step": 9258 }, { "epoch": 0.7256269592476489, "grad_norm": 2.10772705078125, "learning_rate": 9.223324124383022e-06, "loss": 1.2785, "step": 9259 }, { "epoch": 0.725705329153605, "grad_norm": 6.158638954162598, "learning_rate": 9.223159051511424e-06, "loss": 1.1731, "step": 9260 }, { "epoch": 0.7257836990595611, "grad_norm": 4.233865737915039, "learning_rate": 9.222993962577098e-06, "loss": 1.6259, "step": 9261 }, { "epoch": 0.7258620689655172, "grad_norm": 6.5649518966674805, "learning_rate": 9.222828857580673e-06, "loss": 0.7244, "step": 9262 }, { "epoch": 0.7259404388714733, "grad_norm": 3.973682403564453, "learning_rate": 9.222663736522777e-06, "loss": 0.9664, "step": 9263 }, { "epoch": 0.7260188087774294, "grad_norm": 3.334932565689087, "learning_rate": 9.222498599404035e-06, "loss": 1.501, "step": 9264 }, { "epoch": 0.7260971786833855, "grad_norm": 4.609848976135254, "learning_rate": 9.222333446225076e-06, "loss": 1.4635, "step": 9265 }, { "epoch": 0.7261755485893417, "grad_norm": 2.9379756450653076, "learning_rate": 9.222168276986532e-06, "loss": 1.6626, "step": 9266 }, { "epoch": 0.7262539184952979, "grad_norm": 4.313635349273682, "learning_rate": 9.222003091689027e-06, "loss": 1.5554, "step": 9267 }, { "epoch": 0.726332288401254, "grad_norm": 3.931037425994873, "learning_rate": 9.221837890333192e-06, "loss": 1.5672, "step": 9268 }, { "epoch": 0.7264106583072101, "grad_norm": 2.452694892883301, "learning_rate": 9.221672672919653e-06, "loss": 1.5946, "step": 9269 }, { "epoch": 0.7264890282131662, "grad_norm": 2.491586208343506, "learning_rate": 9.22150743944904e-06, "loss": 1.2443, "step": 9270 }, { "epoch": 0.7265673981191223, "grad_norm": 6.594118595123291, "learning_rate": 9.22134218992198e-06, "loss": 1.6902, "step": 9271 }, { "epoch": 0.7266457680250784, "grad_norm": 3.876260995864868, "learning_rate": 9.221176924339103e-06, "loss": 0.9183, "step": 9272 }, { "epoch": 0.7267241379310345, "grad_norm": 2.163586378097534, "learning_rate": 9.221011642701038e-06, "loss": 1.1666, "step": 9273 }, { "epoch": 0.7268025078369906, "grad_norm": 2.6106109619140625, "learning_rate": 9.220846345008414e-06, "loss": 1.4525, "step": 9274 }, { "epoch": 0.7268808777429467, "grad_norm": 4.640730857849121, "learning_rate": 9.220681031261856e-06, "loss": 1.1416, "step": 9275 }, { "epoch": 0.7269592476489029, "grad_norm": 8.353708267211914, "learning_rate": 9.220515701461996e-06, "loss": 1.4426, "step": 9276 }, { "epoch": 0.727037617554859, "grad_norm": 3.275304079055786, "learning_rate": 9.220350355609462e-06, "loss": 1.2629, "step": 9277 }, { "epoch": 0.7271159874608151, "grad_norm": 2.3714590072631836, "learning_rate": 9.220184993704884e-06, "loss": 1.0804, "step": 9278 }, { "epoch": 0.7271943573667712, "grad_norm": 2.531139373779297, "learning_rate": 9.220019615748887e-06, "loss": 1.2632, "step": 9279 }, { "epoch": 0.7272727272727273, "grad_norm": 3.287501096725464, "learning_rate": 9.219854221742106e-06, "loss": 1.3005, "step": 9280 }, { "epoch": 0.7273510971786834, "grad_norm": 2.7866270542144775, "learning_rate": 9.219688811685166e-06, "loss": 1.0915, "step": 9281 }, { "epoch": 0.7274294670846395, "grad_norm": 2.5827512741088867, "learning_rate": 9.219523385578696e-06, "loss": 1.3158, "step": 9282 }, { "epoch": 0.7275078369905956, "grad_norm": 2.6445529460906982, "learning_rate": 9.219357943423329e-06, "loss": 1.489, "step": 9283 }, { "epoch": 0.7275862068965517, "grad_norm": 4.486658573150635, "learning_rate": 9.21919248521969e-06, "loss": 1.3624, "step": 9284 }, { "epoch": 0.7276645768025078, "grad_norm": 2.6560561656951904, "learning_rate": 9.219027010968409e-06, "loss": 1.57, "step": 9285 }, { "epoch": 0.727742946708464, "grad_norm": 5.467525005340576, "learning_rate": 9.218861520670117e-06, "loss": 1.082, "step": 9286 }, { "epoch": 0.7278213166144201, "grad_norm": 2.1949172019958496, "learning_rate": 9.218696014325444e-06, "loss": 1.1911, "step": 9287 }, { "epoch": 0.7278996865203762, "grad_norm": 3.1582038402557373, "learning_rate": 9.218530491935016e-06, "loss": 1.6621, "step": 9288 }, { "epoch": 0.7279780564263323, "grad_norm": 3.0496439933776855, "learning_rate": 9.218364953499466e-06, "loss": 1.6722, "step": 9289 }, { "epoch": 0.7280564263322884, "grad_norm": 7.761069297790527, "learning_rate": 9.218199399019423e-06, "loss": 0.5856, "step": 9290 }, { "epoch": 0.7281347962382445, "grad_norm": 6.571335315704346, "learning_rate": 9.218033828495513e-06, "loss": 1.7613, "step": 9291 }, { "epoch": 0.7282131661442006, "grad_norm": 2.706312417984009, "learning_rate": 9.217868241928372e-06, "loss": 1.2241, "step": 9292 }, { "epoch": 0.7282915360501567, "grad_norm": 2.1864936351776123, "learning_rate": 9.217702639318626e-06, "loss": 1.4745, "step": 9293 }, { "epoch": 0.7283699059561128, "grad_norm": 2.952968120574951, "learning_rate": 9.217537020666905e-06, "loss": 1.4151, "step": 9294 }, { "epoch": 0.728448275862069, "grad_norm": 6.2980732917785645, "learning_rate": 9.21737138597384e-06, "loss": 3.2931, "step": 9295 }, { "epoch": 0.7285266457680251, "grad_norm": 2.2252681255340576, "learning_rate": 9.21720573524006e-06, "loss": 1.1795, "step": 9296 }, { "epoch": 0.7286050156739812, "grad_norm": 12.5728178024292, "learning_rate": 9.217040068466196e-06, "loss": 1.638, "step": 9297 }, { "epoch": 0.7286833855799373, "grad_norm": 1.8391573429107666, "learning_rate": 9.216874385652878e-06, "loss": 1.2988, "step": 9298 }, { "epoch": 0.7287617554858934, "grad_norm": 3.1286251544952393, "learning_rate": 9.216708686800736e-06, "loss": 1.5725, "step": 9299 }, { "epoch": 0.7288401253918495, "grad_norm": 2.731966733932495, "learning_rate": 9.216542971910399e-06, "loss": 1.1886, "step": 9300 }, { "epoch": 0.7289184952978056, "grad_norm": 2.6333625316619873, "learning_rate": 9.216377240982497e-06, "loss": 1.232, "step": 9301 }, { "epoch": 0.7289968652037617, "grad_norm": 6.859716892242432, "learning_rate": 9.216211494017664e-06, "loss": 1.4894, "step": 9302 }, { "epoch": 0.7290752351097178, "grad_norm": 3.918440103530884, "learning_rate": 9.216045731016526e-06, "loss": 1.5904, "step": 9303 }, { "epoch": 0.729153605015674, "grad_norm": 3.3640103340148926, "learning_rate": 9.215879951979716e-06, "loss": 1.7612, "step": 9304 }, { "epoch": 0.72923197492163, "grad_norm": 2.9457664489746094, "learning_rate": 9.215714156907865e-06, "loss": 1.9376, "step": 9305 }, { "epoch": 0.7293103448275862, "grad_norm": 2.738556385040283, "learning_rate": 9.215548345801602e-06, "loss": 1.543, "step": 9306 }, { "epoch": 0.7293887147335423, "grad_norm": 4.966453552246094, "learning_rate": 9.21538251866156e-06, "loss": 1.2529, "step": 9307 }, { "epoch": 0.7294670846394984, "grad_norm": 2.473423957824707, "learning_rate": 9.215216675488365e-06, "loss": 1.458, "step": 9308 }, { "epoch": 0.7295454545454545, "grad_norm": 5.3578877449035645, "learning_rate": 9.215050816282654e-06, "loss": 1.6974, "step": 9309 }, { "epoch": 0.7296238244514106, "grad_norm": 2.6847057342529297, "learning_rate": 9.214884941045053e-06, "loss": 1.338, "step": 9310 }, { "epoch": 0.7297021943573667, "grad_norm": 1.9923174381256104, "learning_rate": 9.214719049776195e-06, "loss": 1.6592, "step": 9311 }, { "epoch": 0.7297805642633229, "grad_norm": 3.51393985748291, "learning_rate": 9.214553142476708e-06, "loss": 1.3735, "step": 9312 }, { "epoch": 0.729858934169279, "grad_norm": 3.3062336444854736, "learning_rate": 9.21438721914723e-06, "loss": 1.1186, "step": 9313 }, { "epoch": 0.7299373040752352, "grad_norm": 7.162853717803955, "learning_rate": 9.214221279788384e-06, "loss": 2.7377, "step": 9314 }, { "epoch": 0.7300156739811913, "grad_norm": 3.5322115421295166, "learning_rate": 9.214055324400807e-06, "loss": 1.2934, "step": 9315 }, { "epoch": 0.7300940438871474, "grad_norm": 2.986166000366211, "learning_rate": 9.213889352985128e-06, "loss": 1.3393, "step": 9316 }, { "epoch": 0.7301724137931035, "grad_norm": 2.4479384422302246, "learning_rate": 9.213723365541979e-06, "loss": 1.3143, "step": 9317 }, { "epoch": 0.7302507836990596, "grad_norm": 3.224639892578125, "learning_rate": 9.213557362071987e-06, "loss": 1.2473, "step": 9318 }, { "epoch": 0.7303291536050157, "grad_norm": 3.1700987815856934, "learning_rate": 9.21339134257579e-06, "loss": 1.7825, "step": 9319 }, { "epoch": 0.7304075235109718, "grad_norm": 1.6948405504226685, "learning_rate": 9.213225307054016e-06, "loss": 1.2283, "step": 9320 }, { "epoch": 0.7304858934169279, "grad_norm": 2.36334228515625, "learning_rate": 9.213059255507297e-06, "loss": 1.3948, "step": 9321 }, { "epoch": 0.730564263322884, "grad_norm": 2.555246353149414, "learning_rate": 9.212893187936262e-06, "loss": 1.1594, "step": 9322 }, { "epoch": 0.7306426332288402, "grad_norm": 4.769404888153076, "learning_rate": 9.212727104341548e-06, "loss": 1.2534, "step": 9323 }, { "epoch": 0.7307210031347963, "grad_norm": 2.2006118297576904, "learning_rate": 9.212561004723785e-06, "loss": 1.6289, "step": 9324 }, { "epoch": 0.7307993730407524, "grad_norm": 7.041990756988525, "learning_rate": 9.212394889083602e-06, "loss": 1.175, "step": 9325 }, { "epoch": 0.7308777429467085, "grad_norm": 2.2913904190063477, "learning_rate": 9.212228757421632e-06, "loss": 1.4059, "step": 9326 }, { "epoch": 0.7309561128526646, "grad_norm": 2.7714953422546387, "learning_rate": 9.212062609738507e-06, "loss": 1.372, "step": 9327 }, { "epoch": 0.7310344827586207, "grad_norm": 3.0192058086395264, "learning_rate": 9.21189644603486e-06, "loss": 1.5938, "step": 9328 }, { "epoch": 0.7311128526645768, "grad_norm": 6.773125648498535, "learning_rate": 9.211730266311323e-06, "loss": 1.2856, "step": 9329 }, { "epoch": 0.7311912225705329, "grad_norm": 2.1818017959594727, "learning_rate": 9.211564070568529e-06, "loss": 1.4603, "step": 9330 }, { "epoch": 0.731269592476489, "grad_norm": 2.7162399291992188, "learning_rate": 9.211397858807105e-06, "loss": 1.5738, "step": 9331 }, { "epoch": 0.7313479623824451, "grad_norm": 11.762201309204102, "learning_rate": 9.211231631027688e-06, "loss": 1.0747, "step": 9332 }, { "epoch": 0.7314263322884013, "grad_norm": 9.745194435119629, "learning_rate": 9.211065387230908e-06, "loss": 1.3991, "step": 9333 }, { "epoch": 0.7315047021943574, "grad_norm": 5.890563488006592, "learning_rate": 9.210899127417398e-06, "loss": 1.4476, "step": 9334 }, { "epoch": 0.7315830721003135, "grad_norm": 2.0408291816711426, "learning_rate": 9.210732851587793e-06, "loss": 1.4245, "step": 9335 }, { "epoch": 0.7316614420062696, "grad_norm": 2.7618355751037598, "learning_rate": 9.210566559742722e-06, "loss": 1.2575, "step": 9336 }, { "epoch": 0.7317398119122257, "grad_norm": 3.700732469558716, "learning_rate": 9.210400251882819e-06, "loss": 1.543, "step": 9337 }, { "epoch": 0.7318181818181818, "grad_norm": 2.740602731704712, "learning_rate": 9.210233928008713e-06, "loss": 1.5716, "step": 9338 }, { "epoch": 0.7318965517241379, "grad_norm": 6.54688835144043, "learning_rate": 9.210067588121042e-06, "loss": 1.9814, "step": 9339 }, { "epoch": 0.731974921630094, "grad_norm": 2.5430960655212402, "learning_rate": 9.209901232220436e-06, "loss": 1.2069, "step": 9340 }, { "epoch": 0.7320532915360501, "grad_norm": 3.3572804927825928, "learning_rate": 9.209734860307528e-06, "loss": 1.405, "step": 9341 }, { "epoch": 0.7321316614420063, "grad_norm": 3.9411981105804443, "learning_rate": 9.209568472382951e-06, "loss": 1.2424, "step": 9342 }, { "epoch": 0.7322100313479624, "grad_norm": 3.622990846633911, "learning_rate": 9.209402068447339e-06, "loss": 1.5391, "step": 9343 }, { "epoch": 0.7322884012539185, "grad_norm": 2.656639814376831, "learning_rate": 9.209235648501322e-06, "loss": 1.2461, "step": 9344 }, { "epoch": 0.7323667711598746, "grad_norm": 7.632052421569824, "learning_rate": 9.209069212545535e-06, "loss": 1.2709, "step": 9345 }, { "epoch": 0.7324451410658307, "grad_norm": 2.158356189727783, "learning_rate": 9.20890276058061e-06, "loss": 1.041, "step": 9346 }, { "epoch": 0.7325235109717868, "grad_norm": 6.112945079803467, "learning_rate": 9.208736292607182e-06, "loss": 1.9709, "step": 9347 }, { "epoch": 0.7326018808777429, "grad_norm": 2.9413528442382812, "learning_rate": 9.208569808625882e-06, "loss": 1.2347, "step": 9348 }, { "epoch": 0.732680250783699, "grad_norm": 4.024317264556885, "learning_rate": 9.208403308637344e-06, "loss": 1.476, "step": 9349 }, { "epoch": 0.7327586206896551, "grad_norm": 2.0738377571105957, "learning_rate": 9.208236792642202e-06, "loss": 1.4002, "step": 9350 }, { "epoch": 0.7328369905956112, "grad_norm": 6.983804702758789, "learning_rate": 9.208070260641088e-06, "loss": 1.3596, "step": 9351 }, { "epoch": 0.7329153605015674, "grad_norm": 3.9659600257873535, "learning_rate": 9.207903712634636e-06, "loss": 1.8158, "step": 9352 }, { "epoch": 0.7329937304075235, "grad_norm": 2.9149346351623535, "learning_rate": 9.20773714862348e-06, "loss": 1.3683, "step": 9353 }, { "epoch": 0.7330721003134796, "grad_norm": 4.42382287979126, "learning_rate": 9.207570568608253e-06, "loss": 1.6745, "step": 9354 }, { "epoch": 0.7331504702194357, "grad_norm": 3.724290132522583, "learning_rate": 9.20740397258959e-06, "loss": 1.0369, "step": 9355 }, { "epoch": 0.7332288401253918, "grad_norm": 10.61081314086914, "learning_rate": 9.207237360568122e-06, "loss": 1.7288, "step": 9356 }, { "epoch": 0.7333072100313479, "grad_norm": 6.663980007171631, "learning_rate": 9.207070732544486e-06, "loss": 1.9045, "step": 9357 }, { "epoch": 0.733385579937304, "grad_norm": 10.60007095336914, "learning_rate": 9.206904088519311e-06, "loss": 1.2101, "step": 9358 }, { "epoch": 0.7334639498432602, "grad_norm": 3.3027536869049072, "learning_rate": 9.206737428493235e-06, "loss": 1.6059, "step": 9359 }, { "epoch": 0.7335423197492164, "grad_norm": 2.4672696590423584, "learning_rate": 9.20657075246689e-06, "loss": 1.1849, "step": 9360 }, { "epoch": 0.7336206896551725, "grad_norm": 2.306347608566284, "learning_rate": 9.20640406044091e-06, "loss": 1.4834, "step": 9361 }, { "epoch": 0.7336990595611286, "grad_norm": 2.1777029037475586, "learning_rate": 9.20623735241593e-06, "loss": 1.1308, "step": 9362 }, { "epoch": 0.7337774294670847, "grad_norm": 3.524423360824585, "learning_rate": 9.206070628392584e-06, "loss": 1.1459, "step": 9363 }, { "epoch": 0.7338557993730408, "grad_norm": 2.5327885150909424, "learning_rate": 9.205903888371505e-06, "loss": 1.2296, "step": 9364 }, { "epoch": 0.7339341692789969, "grad_norm": 4.110353469848633, "learning_rate": 9.20573713235333e-06, "loss": 1.5388, "step": 9365 }, { "epoch": 0.734012539184953, "grad_norm": 7.322165489196777, "learning_rate": 9.20557036033869e-06, "loss": 1.4273, "step": 9366 }, { "epoch": 0.7340909090909091, "grad_norm": 3.315275192260742, "learning_rate": 9.20540357232822e-06, "loss": 1.214, "step": 9367 }, { "epoch": 0.7341692789968652, "grad_norm": 2.5644845962524414, "learning_rate": 9.205236768322554e-06, "loss": 1.4355, "step": 9368 }, { "epoch": 0.7342476489028213, "grad_norm": 3.6240038871765137, "learning_rate": 9.205069948322329e-06, "loss": 1.4834, "step": 9369 }, { "epoch": 0.7343260188087775, "grad_norm": 2.960841417312622, "learning_rate": 9.204903112328177e-06, "loss": 1.344, "step": 9370 }, { "epoch": 0.7344043887147336, "grad_norm": 2.7184886932373047, "learning_rate": 9.204736260340733e-06, "loss": 1.2863, "step": 9371 }, { "epoch": 0.7344827586206897, "grad_norm": 2.1398682594299316, "learning_rate": 9.204569392360632e-06, "loss": 1.2473, "step": 9372 }, { "epoch": 0.7345611285266458, "grad_norm": 4.511906147003174, "learning_rate": 9.204402508388509e-06, "loss": 1.8519, "step": 9373 }, { "epoch": 0.7346394984326019, "grad_norm": 3.0090394020080566, "learning_rate": 9.204235608424999e-06, "loss": 1.4784, "step": 9374 }, { "epoch": 0.734717868338558, "grad_norm": 5.463232517242432, "learning_rate": 9.204068692470735e-06, "loss": 2.0684, "step": 9375 }, { "epoch": 0.7347962382445141, "grad_norm": 2.338271379470825, "learning_rate": 9.203901760526353e-06, "loss": 1.1191, "step": 9376 }, { "epoch": 0.7348746081504702, "grad_norm": 3.589982271194458, "learning_rate": 9.203734812592489e-06, "loss": 1.4125, "step": 9377 }, { "epoch": 0.7349529780564263, "grad_norm": 2.4626269340515137, "learning_rate": 9.203567848669777e-06, "loss": 1.5202, "step": 9378 }, { "epoch": 0.7350313479623825, "grad_norm": 2.6698193550109863, "learning_rate": 9.20340086875885e-06, "loss": 1.7684, "step": 9379 }, { "epoch": 0.7351097178683386, "grad_norm": 2.761148452758789, "learning_rate": 9.203233872860346e-06, "loss": 1.1702, "step": 9380 }, { "epoch": 0.7351880877742947, "grad_norm": 2.614832639694214, "learning_rate": 9.203066860974902e-06, "loss": 1.5893, "step": 9381 }, { "epoch": 0.7352664576802508, "grad_norm": 3.213500499725342, "learning_rate": 9.202899833103147e-06, "loss": 1.4613, "step": 9382 }, { "epoch": 0.7353448275862069, "grad_norm": 2.072869062423706, "learning_rate": 9.20273278924572e-06, "loss": 1.0683, "step": 9383 }, { "epoch": 0.735423197492163, "grad_norm": 2.52652645111084, "learning_rate": 9.202565729403259e-06, "loss": 1.3554, "step": 9384 }, { "epoch": 0.7355015673981191, "grad_norm": 5.399397373199463, "learning_rate": 9.202398653576393e-06, "loss": 1.486, "step": 9385 }, { "epoch": 0.7355799373040752, "grad_norm": 2.4923503398895264, "learning_rate": 9.202231561765764e-06, "loss": 1.4313, "step": 9386 }, { "epoch": 0.7356583072100313, "grad_norm": 3.262322425842285, "learning_rate": 9.202064453972003e-06, "loss": 1.2761, "step": 9387 }, { "epoch": 0.7357366771159874, "grad_norm": 3.0736472606658936, "learning_rate": 9.201897330195748e-06, "loss": 1.7552, "step": 9388 }, { "epoch": 0.7358150470219436, "grad_norm": 2.8441412448883057, "learning_rate": 9.201730190437632e-06, "loss": 1.5967, "step": 9389 }, { "epoch": 0.7358934169278997, "grad_norm": 2.991210699081421, "learning_rate": 9.201563034698295e-06, "loss": 1.3876, "step": 9390 }, { "epoch": 0.7359717868338558, "grad_norm": 3.437208890914917, "learning_rate": 9.201395862978369e-06, "loss": 1.5176, "step": 9391 }, { "epoch": 0.7360501567398119, "grad_norm": 6.52949857711792, "learning_rate": 9.20122867527849e-06, "loss": 1.8601, "step": 9392 }, { "epoch": 0.736128526645768, "grad_norm": 2.42779278755188, "learning_rate": 9.201061471599297e-06, "loss": 1.6411, "step": 9393 }, { "epoch": 0.7362068965517241, "grad_norm": 3.2572410106658936, "learning_rate": 9.200894251941423e-06, "loss": 0.9196, "step": 9394 }, { "epoch": 0.7362852664576802, "grad_norm": 4.851722717285156, "learning_rate": 9.200727016305505e-06, "loss": 1.4945, "step": 9395 }, { "epoch": 0.7363636363636363, "grad_norm": 3.247523069381714, "learning_rate": 9.200559764692179e-06, "loss": 1.2731, "step": 9396 }, { "epoch": 0.7364420062695924, "grad_norm": 3.2490105628967285, "learning_rate": 9.200392497102082e-06, "loss": 1.3591, "step": 9397 }, { "epoch": 0.7365203761755486, "grad_norm": 2.505441427230835, "learning_rate": 9.200225213535848e-06, "loss": 1.0397, "step": 9398 }, { "epoch": 0.7365987460815047, "grad_norm": 2.3241374492645264, "learning_rate": 9.200057913994116e-06, "loss": 1.2221, "step": 9399 }, { "epoch": 0.7366771159874608, "grad_norm": 3.0855801105499268, "learning_rate": 9.19989059847752e-06, "loss": 1.3646, "step": 9400 }, { "epoch": 0.7367554858934169, "grad_norm": 1.9786826372146606, "learning_rate": 9.199723266986698e-06, "loss": 1.4895, "step": 9401 }, { "epoch": 0.736833855799373, "grad_norm": 2.4738028049468994, "learning_rate": 9.199555919522285e-06, "loss": 1.3453, "step": 9402 }, { "epoch": 0.7369122257053291, "grad_norm": 1.656833529472351, "learning_rate": 9.199388556084918e-06, "loss": 1.1626, "step": 9403 }, { "epoch": 0.7369905956112852, "grad_norm": 2.4389188289642334, "learning_rate": 9.199221176675236e-06, "loss": 1.3894, "step": 9404 }, { "epoch": 0.7370689655172413, "grad_norm": 4.9778947830200195, "learning_rate": 9.199053781293873e-06, "loss": 1.713, "step": 9405 }, { "epoch": 0.7371473354231975, "grad_norm": 2.5421969890594482, "learning_rate": 9.198886369941465e-06, "loss": 1.1741, "step": 9406 }, { "epoch": 0.7372257053291537, "grad_norm": 3.477656364440918, "learning_rate": 9.19871894261865e-06, "loss": 0.6682, "step": 9407 }, { "epoch": 0.7373040752351098, "grad_norm": 2.8079934120178223, "learning_rate": 9.198551499326064e-06, "loss": 1.0418, "step": 9408 }, { "epoch": 0.7373824451410659, "grad_norm": 2.287130117416382, "learning_rate": 9.198384040064346e-06, "loss": 1.4876, "step": 9409 }, { "epoch": 0.737460815047022, "grad_norm": 2.5103158950805664, "learning_rate": 9.19821656483413e-06, "loss": 1.5787, "step": 9410 }, { "epoch": 0.7375391849529781, "grad_norm": 2.4682865142822266, "learning_rate": 9.198049073636057e-06, "loss": 1.3933, "step": 9411 }, { "epoch": 0.7376175548589342, "grad_norm": 2.824657917022705, "learning_rate": 9.197881566470761e-06, "loss": 1.5593, "step": 9412 }, { "epoch": 0.7376959247648903, "grad_norm": 2.46091365814209, "learning_rate": 9.19771404333888e-06, "loss": 1.3805, "step": 9413 }, { "epoch": 0.7377742946708464, "grad_norm": 2.181159496307373, "learning_rate": 9.197546504241048e-06, "loss": 1.5073, "step": 9414 }, { "epoch": 0.7378526645768025, "grad_norm": 2.723475933074951, "learning_rate": 9.197378949177908e-06, "loss": 1.4259, "step": 9415 }, { "epoch": 0.7379310344827587, "grad_norm": 2.604835271835327, "learning_rate": 9.197211378150093e-06, "loss": 0.9662, "step": 9416 }, { "epoch": 0.7380094043887148, "grad_norm": 2.5198886394500732, "learning_rate": 9.197043791158242e-06, "loss": 1.0487, "step": 9417 }, { "epoch": 0.7380877742946709, "grad_norm": 3.5279147624969482, "learning_rate": 9.196876188202992e-06, "loss": 1.0658, "step": 9418 }, { "epoch": 0.738166144200627, "grad_norm": 4.148751735687256, "learning_rate": 9.19670856928498e-06, "loss": 1.4299, "step": 9419 }, { "epoch": 0.7382445141065831, "grad_norm": 1.9660640954971313, "learning_rate": 9.196540934404845e-06, "loss": 0.8921, "step": 9420 }, { "epoch": 0.7383228840125392, "grad_norm": 4.803493022918701, "learning_rate": 9.196373283563224e-06, "loss": 1.3708, "step": 9421 }, { "epoch": 0.7384012539184953, "grad_norm": 2.8804562091827393, "learning_rate": 9.196205616760754e-06, "loss": 1.8765, "step": 9422 }, { "epoch": 0.7384796238244514, "grad_norm": 4.583609580993652, "learning_rate": 9.196037933998075e-06, "loss": 1.8181, "step": 9423 }, { "epoch": 0.7385579937304075, "grad_norm": 8.732020378112793, "learning_rate": 9.195870235275821e-06, "loss": 1.3177, "step": 9424 }, { "epoch": 0.7386363636363636, "grad_norm": 4.350712299346924, "learning_rate": 9.195702520594631e-06, "loss": 1.1932, "step": 9425 }, { "epoch": 0.7387147335423198, "grad_norm": 3.481785535812378, "learning_rate": 9.195534789955146e-06, "loss": 1.6093, "step": 9426 }, { "epoch": 0.7387931034482759, "grad_norm": 3.7699663639068604, "learning_rate": 9.195367043358e-06, "loss": 1.4388, "step": 9427 }, { "epoch": 0.738871473354232, "grad_norm": 5.286339282989502, "learning_rate": 9.195199280803833e-06, "loss": 1.5484, "step": 9428 }, { "epoch": 0.7389498432601881, "grad_norm": 2.693319797515869, "learning_rate": 9.195031502293283e-06, "loss": 1.2946, "step": 9429 }, { "epoch": 0.7390282131661442, "grad_norm": 3.6171247959136963, "learning_rate": 9.194863707826987e-06, "loss": 1.7673, "step": 9430 }, { "epoch": 0.7391065830721003, "grad_norm": 2.292949914932251, "learning_rate": 9.194695897405585e-06, "loss": 1.2339, "step": 9431 }, { "epoch": 0.7391849529780564, "grad_norm": 3.8519740104675293, "learning_rate": 9.194528071029713e-06, "loss": 1.155, "step": 9432 }, { "epoch": 0.7392633228840125, "grad_norm": 4.09972620010376, "learning_rate": 9.194360228700012e-06, "loss": 1.5199, "step": 9433 }, { "epoch": 0.7393416927899686, "grad_norm": 3.329843044281006, "learning_rate": 9.194192370417119e-06, "loss": 1.3543, "step": 9434 }, { "epoch": 0.7394200626959248, "grad_norm": 2.8204762935638428, "learning_rate": 9.194024496181673e-06, "loss": 1.0245, "step": 9435 }, { "epoch": 0.7394984326018809, "grad_norm": 3.000123977661133, "learning_rate": 9.193856605994311e-06, "loss": 1.5759, "step": 9436 }, { "epoch": 0.739576802507837, "grad_norm": 2.4417312145233154, "learning_rate": 9.193688699855672e-06, "loss": 1.2273, "step": 9437 }, { "epoch": 0.7396551724137931, "grad_norm": 3.326016426086426, "learning_rate": 9.193520777766397e-06, "loss": 1.9736, "step": 9438 }, { "epoch": 0.7397335423197492, "grad_norm": 2.2462565898895264, "learning_rate": 9.193352839727122e-06, "loss": 1.5411, "step": 9439 }, { "epoch": 0.7398119122257053, "grad_norm": 2.1264402866363525, "learning_rate": 9.193184885738485e-06, "loss": 1.3514, "step": 9440 }, { "epoch": 0.7398902821316614, "grad_norm": 1.980741262435913, "learning_rate": 9.193016915801128e-06, "loss": 1.2289, "step": 9441 }, { "epoch": 0.7399686520376175, "grad_norm": 2.350710868835449, "learning_rate": 9.19284892991569e-06, "loss": 1.5752, "step": 9442 }, { "epoch": 0.7400470219435736, "grad_norm": 2.4599180221557617, "learning_rate": 9.192680928082804e-06, "loss": 1.3287, "step": 9443 }, { "epoch": 0.7401253918495297, "grad_norm": 2.376278877258301, "learning_rate": 9.192512910303117e-06, "loss": 0.9786, "step": 9444 }, { "epoch": 0.7402037617554859, "grad_norm": 2.871659755706787, "learning_rate": 9.192344876577264e-06, "loss": 1.2395, "step": 9445 }, { "epoch": 0.740282131661442, "grad_norm": 2.420328140258789, "learning_rate": 9.192176826905883e-06, "loss": 1.6695, "step": 9446 }, { "epoch": 0.7403605015673981, "grad_norm": 2.1430671215057373, "learning_rate": 9.192008761289614e-06, "loss": 1.4799, "step": 9447 }, { "epoch": 0.7404388714733542, "grad_norm": 3.965435028076172, "learning_rate": 9.191840679729099e-06, "loss": 1.1186, "step": 9448 }, { "epoch": 0.7405172413793103, "grad_norm": 4.249763011932373, "learning_rate": 9.191672582224973e-06, "loss": 1.5503, "step": 9449 }, { "epoch": 0.7405956112852664, "grad_norm": 2.298323154449463, "learning_rate": 9.191504468777878e-06, "loss": 1.1319, "step": 9450 }, { "epoch": 0.7406739811912225, "grad_norm": 16.41889762878418, "learning_rate": 9.191336339388455e-06, "loss": 1.2844, "step": 9451 }, { "epoch": 0.7407523510971786, "grad_norm": 2.630977153778076, "learning_rate": 9.191168194057339e-06, "loss": 1.0487, "step": 9452 }, { "epoch": 0.7408307210031349, "grad_norm": 4.035462856292725, "learning_rate": 9.191000032785173e-06, "loss": 1.523, "step": 9453 }, { "epoch": 0.740909090909091, "grad_norm": 2.010976791381836, "learning_rate": 9.190831855572594e-06, "loss": 1.4727, "step": 9454 }, { "epoch": 0.7409874608150471, "grad_norm": 1.7295632362365723, "learning_rate": 9.190663662420245e-06, "loss": 1.6392, "step": 9455 }, { "epoch": 0.7410658307210032, "grad_norm": 3.1558446884155273, "learning_rate": 9.190495453328763e-06, "loss": 1.3635, "step": 9456 }, { "epoch": 0.7411442006269593, "grad_norm": 2.9572808742523193, "learning_rate": 9.19032722829879e-06, "loss": 1.1423, "step": 9457 }, { "epoch": 0.7412225705329154, "grad_norm": 1.8700443506240845, "learning_rate": 9.190158987330964e-06, "loss": 1.2441, "step": 9458 }, { "epoch": 0.7413009404388715, "grad_norm": 2.1055338382720947, "learning_rate": 9.189990730425924e-06, "loss": 1.4441, "step": 9459 }, { "epoch": 0.7413793103448276, "grad_norm": 3.875260591506958, "learning_rate": 9.189822457584311e-06, "loss": 1.1693, "step": 9460 }, { "epoch": 0.7414576802507837, "grad_norm": 3.289090156555176, "learning_rate": 9.189654168806767e-06, "loss": 1.441, "step": 9461 }, { "epoch": 0.7415360501567398, "grad_norm": 3.256362199783325, "learning_rate": 9.18948586409393e-06, "loss": 1.1087, "step": 9462 }, { "epoch": 0.741614420062696, "grad_norm": 2.342308759689331, "learning_rate": 9.18931754344644e-06, "loss": 1.3806, "step": 9463 }, { "epoch": 0.7416927899686521, "grad_norm": 1.978055477142334, "learning_rate": 9.18914920686494e-06, "loss": 1.637, "step": 9464 }, { "epoch": 0.7417711598746082, "grad_norm": 4.0835065841674805, "learning_rate": 9.188980854350067e-06, "loss": 1.2682, "step": 9465 }, { "epoch": 0.7418495297805643, "grad_norm": 2.447756767272949, "learning_rate": 9.188812485902461e-06, "loss": 1.3632, "step": 9466 }, { "epoch": 0.7419278996865204, "grad_norm": 2.4238619804382324, "learning_rate": 9.188644101522766e-06, "loss": 1.4206, "step": 9467 }, { "epoch": 0.7420062695924765, "grad_norm": 4.69226598739624, "learning_rate": 9.18847570121162e-06, "loss": 1.2918, "step": 9468 }, { "epoch": 0.7420846394984326, "grad_norm": 2.9437077045440674, "learning_rate": 9.188307284969662e-06, "loss": 1.4053, "step": 9469 }, { "epoch": 0.7421630094043887, "grad_norm": 3.1931421756744385, "learning_rate": 9.188138852797535e-06, "loss": 1.7467, "step": 9470 }, { "epoch": 0.7422413793103448, "grad_norm": 1.807641863822937, "learning_rate": 9.18797040469588e-06, "loss": 0.6886, "step": 9471 }, { "epoch": 0.742319749216301, "grad_norm": 2.967853307723999, "learning_rate": 9.187801940665334e-06, "loss": 1.1606, "step": 9472 }, { "epoch": 0.7423981191222571, "grad_norm": 2.238003969192505, "learning_rate": 9.187633460706545e-06, "loss": 1.1201, "step": 9473 }, { "epoch": 0.7424764890282132, "grad_norm": 2.659815788269043, "learning_rate": 9.187464964820146e-06, "loss": 1.4714, "step": 9474 }, { "epoch": 0.7425548589341693, "grad_norm": 4.760280609130859, "learning_rate": 9.18729645300678e-06, "loss": 1.1393, "step": 9475 }, { "epoch": 0.7426332288401254, "grad_norm": 2.59417462348938, "learning_rate": 9.187127925267092e-06, "loss": 1.4231, "step": 9476 }, { "epoch": 0.7427115987460815, "grad_norm": 2.6160075664520264, "learning_rate": 9.186959381601717e-06, "loss": 1.2068, "step": 9477 }, { "epoch": 0.7427899686520376, "grad_norm": 2.5753090381622314, "learning_rate": 9.186790822011302e-06, "loss": 1.5611, "step": 9478 }, { "epoch": 0.7428683385579937, "grad_norm": 3.2503867149353027, "learning_rate": 9.186622246496484e-06, "loss": 1.4966, "step": 9479 }, { "epoch": 0.7429467084639498, "grad_norm": 2.2582931518554688, "learning_rate": 9.186453655057906e-06, "loss": 1.2055, "step": 9480 }, { "epoch": 0.743025078369906, "grad_norm": 5.216309070587158, "learning_rate": 9.186285047696206e-06, "loss": 0.8354, "step": 9481 }, { "epoch": 0.743103448275862, "grad_norm": 3.0215389728546143, "learning_rate": 9.18611642441203e-06, "loss": 1.4476, "step": 9482 }, { "epoch": 0.7431818181818182, "grad_norm": 3.2400596141815186, "learning_rate": 9.185947785206017e-06, "loss": 1.4243, "step": 9483 }, { "epoch": 0.7432601880877743, "grad_norm": 2.743276596069336, "learning_rate": 9.18577913007881e-06, "loss": 1.3715, "step": 9484 }, { "epoch": 0.7433385579937304, "grad_norm": 2.4889917373657227, "learning_rate": 9.185610459031047e-06, "loss": 1.281, "step": 9485 }, { "epoch": 0.7434169278996865, "grad_norm": 4.277037143707275, "learning_rate": 9.185441772063373e-06, "loss": 1.2186, "step": 9486 }, { "epoch": 0.7434952978056426, "grad_norm": 3.511334180831909, "learning_rate": 9.185273069176426e-06, "loss": 2.0042, "step": 9487 }, { "epoch": 0.7435736677115987, "grad_norm": 4.287774562835693, "learning_rate": 9.185104350370852e-06, "loss": 1.242, "step": 9488 }, { "epoch": 0.7436520376175548, "grad_norm": 3.0005717277526855, "learning_rate": 9.18493561564729e-06, "loss": 1.6203, "step": 9489 }, { "epoch": 0.7437304075235109, "grad_norm": 2.6687819957733154, "learning_rate": 9.184766865006384e-06, "loss": 1.3409, "step": 9490 }, { "epoch": 0.743808777429467, "grad_norm": 5.943172454833984, "learning_rate": 9.184598098448774e-06, "loss": 1.9795, "step": 9491 }, { "epoch": 0.7438871473354232, "grad_norm": 3.3641622066497803, "learning_rate": 9.184429315975101e-06, "loss": 1.8445, "step": 9492 }, { "epoch": 0.7439655172413793, "grad_norm": 2.164232015609741, "learning_rate": 9.184260517586008e-06, "loss": 1.3476, "step": 9493 }, { "epoch": 0.7440438871473354, "grad_norm": 3.072291612625122, "learning_rate": 9.184091703282136e-06, "loss": 1.4156, "step": 9494 }, { "epoch": 0.7441222570532915, "grad_norm": 2.2170026302337646, "learning_rate": 9.18392287306413e-06, "loss": 1.551, "step": 9495 }, { "epoch": 0.7442006269592476, "grad_norm": 2.7398698329925537, "learning_rate": 9.183754026932631e-06, "loss": 1.4411, "step": 9496 }, { "epoch": 0.7442789968652037, "grad_norm": 2.3156375885009766, "learning_rate": 9.18358516488828e-06, "loss": 1.5258, "step": 9497 }, { "epoch": 0.7443573667711598, "grad_norm": 2.7472493648529053, "learning_rate": 9.18341628693172e-06, "loss": 1.6207, "step": 9498 }, { "epoch": 0.7444357366771159, "grad_norm": 1.9477933645248413, "learning_rate": 9.183247393063592e-06, "loss": 1.1143, "step": 9499 }, { "epoch": 0.7445141065830722, "grad_norm": 3.008382558822632, "learning_rate": 9.183078483284541e-06, "loss": 1.1177, "step": 9500 }, { "epoch": 0.7445924764890283, "grad_norm": 2.534712314605713, "learning_rate": 9.182909557595207e-06, "loss": 1.6413, "step": 9501 }, { "epoch": 0.7446708463949844, "grad_norm": 4.573381423950195, "learning_rate": 9.182740615996234e-06, "loss": 1.8301, "step": 9502 }, { "epoch": 0.7447492163009405, "grad_norm": 2.0716919898986816, "learning_rate": 9.182571658488265e-06, "loss": 1.2889, "step": 9503 }, { "epoch": 0.7448275862068966, "grad_norm": 2.5399651527404785, "learning_rate": 9.18240268507194e-06, "loss": 1.6688, "step": 9504 }, { "epoch": 0.7449059561128527, "grad_norm": 2.376316547393799, "learning_rate": 9.182233695747905e-06, "loss": 1.4961, "step": 9505 }, { "epoch": 0.7449843260188088, "grad_norm": 2.43509840965271, "learning_rate": 9.1820646905168e-06, "loss": 1.4237, "step": 9506 }, { "epoch": 0.7450626959247649, "grad_norm": 2.378711223602295, "learning_rate": 9.181895669379268e-06, "loss": 1.4351, "step": 9507 }, { "epoch": 0.745141065830721, "grad_norm": 2.743197441101074, "learning_rate": 9.181726632335955e-06, "loss": 1.1599, "step": 9508 }, { "epoch": 0.7452194357366771, "grad_norm": 4.541423320770264, "learning_rate": 9.181557579387501e-06, "loss": 1.8936, "step": 9509 }, { "epoch": 0.7452978056426333, "grad_norm": 2.911797523498535, "learning_rate": 9.181388510534549e-06, "loss": 1.1966, "step": 9510 }, { "epoch": 0.7453761755485894, "grad_norm": 3.110736846923828, "learning_rate": 9.181219425777743e-06, "loss": 1.3241, "step": 9511 }, { "epoch": 0.7454545454545455, "grad_norm": 7.180263042449951, "learning_rate": 9.181050325117726e-06, "loss": 0.8664, "step": 9512 }, { "epoch": 0.7455329153605016, "grad_norm": 2.486713409423828, "learning_rate": 9.180881208555143e-06, "loss": 1.363, "step": 9513 }, { "epoch": 0.7456112852664577, "grad_norm": 3.6180801391601562, "learning_rate": 9.180712076090633e-06, "loss": 1.239, "step": 9514 }, { "epoch": 0.7456896551724138, "grad_norm": 2.471457004547119, "learning_rate": 9.180542927724843e-06, "loss": 1.0688, "step": 9515 }, { "epoch": 0.7457680250783699, "grad_norm": 1.9176173210144043, "learning_rate": 9.180373763458413e-06, "loss": 1.4121, "step": 9516 }, { "epoch": 0.745846394984326, "grad_norm": 1.9292718172073364, "learning_rate": 9.18020458329199e-06, "loss": 1.5452, "step": 9517 }, { "epoch": 0.7459247648902821, "grad_norm": 2.93866229057312, "learning_rate": 9.180035387226215e-06, "loss": 1.1625, "step": 9518 }, { "epoch": 0.7460031347962383, "grad_norm": 3.2248663902282715, "learning_rate": 9.179866175261734e-06, "loss": 1.3976, "step": 9519 }, { "epoch": 0.7460815047021944, "grad_norm": 2.410809278488159, "learning_rate": 9.179696947399188e-06, "loss": 1.0156, "step": 9520 }, { "epoch": 0.7461598746081505, "grad_norm": 4.006683826446533, "learning_rate": 9.179527703639222e-06, "loss": 1.1715, "step": 9521 }, { "epoch": 0.7462382445141066, "grad_norm": 3.116886854171753, "learning_rate": 9.179358443982478e-06, "loss": 1.6608, "step": 9522 }, { "epoch": 0.7463166144200627, "grad_norm": 4.713955402374268, "learning_rate": 9.179189168429601e-06, "loss": 1.3243, "step": 9523 }, { "epoch": 0.7463949843260188, "grad_norm": 3.012585401535034, "learning_rate": 9.179019876981237e-06, "loss": 1.3427, "step": 9524 }, { "epoch": 0.7464733542319749, "grad_norm": 2.400575876235962, "learning_rate": 9.178850569638027e-06, "loss": 1.3905, "step": 9525 }, { "epoch": 0.746551724137931, "grad_norm": 2.3448452949523926, "learning_rate": 9.178681246400616e-06, "loss": 1.337, "step": 9526 }, { "epoch": 0.7466300940438871, "grad_norm": 5.740418910980225, "learning_rate": 9.178511907269647e-06, "loss": 1.0352, "step": 9527 }, { "epoch": 0.7467084639498432, "grad_norm": 3.4990744590759277, "learning_rate": 9.178342552245764e-06, "loss": 1.8123, "step": 9528 }, { "epoch": 0.7467868338557994, "grad_norm": 4.71883487701416, "learning_rate": 9.178173181329614e-06, "loss": 1.1518, "step": 9529 }, { "epoch": 0.7468652037617555, "grad_norm": 1.6991872787475586, "learning_rate": 9.178003794521838e-06, "loss": 1.2195, "step": 9530 }, { "epoch": 0.7469435736677116, "grad_norm": 3.103607177734375, "learning_rate": 9.177834391823084e-06, "loss": 1.2759, "step": 9531 }, { "epoch": 0.7470219435736677, "grad_norm": 2.899688959121704, "learning_rate": 9.177664973233992e-06, "loss": 1.4516, "step": 9532 }, { "epoch": 0.7471003134796238, "grad_norm": 2.477745294570923, "learning_rate": 9.177495538755209e-06, "loss": 1.3094, "step": 9533 }, { "epoch": 0.7471786833855799, "grad_norm": 2.6383652687072754, "learning_rate": 9.177326088387377e-06, "loss": 1.4893, "step": 9534 }, { "epoch": 0.747257053291536, "grad_norm": 4.465756893157959, "learning_rate": 9.177156622131143e-06, "loss": 1.3338, "step": 9535 }, { "epoch": 0.7473354231974921, "grad_norm": 2.675731897354126, "learning_rate": 9.176987139987153e-06, "loss": 1.484, "step": 9536 }, { "epoch": 0.7474137931034482, "grad_norm": 3.2144827842712402, "learning_rate": 9.176817641956047e-06, "loss": 1.5315, "step": 9537 }, { "epoch": 0.7474921630094044, "grad_norm": 2.3217954635620117, "learning_rate": 9.176648128038472e-06, "loss": 1.4977, "step": 9538 }, { "epoch": 0.7475705329153605, "grad_norm": 3.226796865463257, "learning_rate": 9.176478598235072e-06, "loss": 1.5207, "step": 9539 }, { "epoch": 0.7476489028213166, "grad_norm": 2.1959643363952637, "learning_rate": 9.176309052546495e-06, "loss": 1.4761, "step": 9540 }, { "epoch": 0.7477272727272727, "grad_norm": 2.632652521133423, "learning_rate": 9.176139490973381e-06, "loss": 1.4418, "step": 9541 }, { "epoch": 0.7478056426332288, "grad_norm": 2.1968860626220703, "learning_rate": 9.175969913516379e-06, "loss": 1.3405, "step": 9542 }, { "epoch": 0.7478840125391849, "grad_norm": 2.8807621002197266, "learning_rate": 9.175800320176131e-06, "loss": 1.5485, "step": 9543 }, { "epoch": 0.747962382445141, "grad_norm": 3.080228090286255, "learning_rate": 9.175630710953285e-06, "loss": 1.6457, "step": 9544 }, { "epoch": 0.7480407523510971, "grad_norm": 3.680027961730957, "learning_rate": 9.175461085848485e-06, "loss": 1.4959, "step": 9545 }, { "epoch": 0.7481191222570532, "grad_norm": 2.4530749320983887, "learning_rate": 9.175291444862373e-06, "loss": 1.4551, "step": 9546 }, { "epoch": 0.7481974921630095, "grad_norm": 2.8719747066497803, "learning_rate": 9.175121787995598e-06, "loss": 0.8916, "step": 9547 }, { "epoch": 0.7482758620689656, "grad_norm": 3.9597508907318115, "learning_rate": 9.174952115248805e-06, "loss": 1.7979, "step": 9548 }, { "epoch": 0.7483542319749217, "grad_norm": 5.498205184936523, "learning_rate": 9.17478242662264e-06, "loss": 1.5713, "step": 9549 }, { "epoch": 0.7484326018808778, "grad_norm": 11.042363166809082, "learning_rate": 9.174612722117744e-06, "loss": 0.9041, "step": 9550 }, { "epoch": 0.7485109717868339, "grad_norm": 3.8407599925994873, "learning_rate": 9.174443001734768e-06, "loss": 0.9554, "step": 9551 }, { "epoch": 0.74858934169279, "grad_norm": 2.929736614227295, "learning_rate": 9.174273265474351e-06, "loss": 1.2279, "step": 9552 }, { "epoch": 0.7486677115987461, "grad_norm": 3.562371015548706, "learning_rate": 9.174103513337147e-06, "loss": 0.7455, "step": 9553 }, { "epoch": 0.7487460815047022, "grad_norm": 8.150503158569336, "learning_rate": 9.173933745323795e-06, "loss": 1.6599, "step": 9554 }, { "epoch": 0.7488244514106583, "grad_norm": 3.458763360977173, "learning_rate": 9.173763961434942e-06, "loss": 1.5904, "step": 9555 }, { "epoch": 0.7489028213166145, "grad_norm": 3.05407452583313, "learning_rate": 9.173594161671237e-06, "loss": 1.7708, "step": 9556 }, { "epoch": 0.7489811912225706, "grad_norm": 2.3831591606140137, "learning_rate": 9.173424346033322e-06, "loss": 1.0984, "step": 9557 }, { "epoch": 0.7490595611285267, "grad_norm": 2.5027682781219482, "learning_rate": 9.173254514521844e-06, "loss": 1.2938, "step": 9558 }, { "epoch": 0.7491379310344828, "grad_norm": 4.063108444213867, "learning_rate": 9.17308466713745e-06, "loss": 1.4869, "step": 9559 }, { "epoch": 0.7492163009404389, "grad_norm": 4.216082572937012, "learning_rate": 9.172914803880786e-06, "loss": 1.7915, "step": 9560 }, { "epoch": 0.749294670846395, "grad_norm": 3.702577829360962, "learning_rate": 9.172744924752497e-06, "loss": 1.4051, "step": 9561 }, { "epoch": 0.7493730407523511, "grad_norm": 2.390428066253662, "learning_rate": 9.172575029753229e-06, "loss": 1.1725, "step": 9562 }, { "epoch": 0.7494514106583072, "grad_norm": 6.339640140533447, "learning_rate": 9.17240511888363e-06, "loss": 1.3077, "step": 9563 }, { "epoch": 0.7495297805642633, "grad_norm": 5.891791820526123, "learning_rate": 9.172235192144343e-06, "loss": 1.3595, "step": 9564 }, { "epoch": 0.7496081504702194, "grad_norm": 3.0881946086883545, "learning_rate": 9.172065249536017e-06, "loss": 1.2819, "step": 9565 }, { "epoch": 0.7496865203761756, "grad_norm": 2.7884268760681152, "learning_rate": 9.171895291059298e-06, "loss": 1.5788, "step": 9566 }, { "epoch": 0.7497648902821317, "grad_norm": 2.4517695903778076, "learning_rate": 9.171725316714834e-06, "loss": 1.3787, "step": 9567 }, { "epoch": 0.7498432601880878, "grad_norm": 3.8441383838653564, "learning_rate": 9.171555326503267e-06, "loss": 1.3029, "step": 9568 }, { "epoch": 0.7499216300940439, "grad_norm": 13.299089431762695, "learning_rate": 9.171385320425248e-06, "loss": 2.953, "step": 9569 }, { "epoch": 0.75, "grad_norm": 2.994394540786743, "learning_rate": 9.17121529848142e-06, "loss": 1.2912, "step": 9570 }, { "epoch": 0.75, "eval_loss": 1.3302174806594849, "eval_runtime": 114.604, "eval_samples_per_second": 23.446, "eval_steps_per_second": 5.864, "step": 9570 }, { "epoch": 0.7500783699059561, "grad_norm": 2.4897267818450928, "learning_rate": 9.171045260672433e-06, "loss": 1.5097, "step": 9571 }, { "epoch": 0.7501567398119122, "grad_norm": 2.908703565597534, "learning_rate": 9.170875206998932e-06, "loss": 1.3753, "step": 9572 }, { "epoch": 0.7502351097178683, "grad_norm": 4.595193862915039, "learning_rate": 9.170705137461562e-06, "loss": 1.0397, "step": 9573 }, { "epoch": 0.7503134796238244, "grad_norm": 2.339905023574829, "learning_rate": 9.170535052060973e-06, "loss": 1.3468, "step": 9574 }, { "epoch": 0.7503918495297806, "grad_norm": 3.197401285171509, "learning_rate": 9.170364950797811e-06, "loss": 1.1833, "step": 9575 }, { "epoch": 0.7504702194357367, "grad_norm": 8.441263198852539, "learning_rate": 9.170194833672725e-06, "loss": 1.1196, "step": 9576 }, { "epoch": 0.7505485893416928, "grad_norm": 3.503783702850342, "learning_rate": 9.170024700686358e-06, "loss": 0.8819, "step": 9577 }, { "epoch": 0.7506269592476489, "grad_norm": 3.3224778175354004, "learning_rate": 9.169854551839359e-06, "loss": 1.7373, "step": 9578 }, { "epoch": 0.750705329153605, "grad_norm": 2.688682794570923, "learning_rate": 9.169684387132374e-06, "loss": 1.4034, "step": 9579 }, { "epoch": 0.7507836990595611, "grad_norm": 4.031322956085205, "learning_rate": 9.169514206566053e-06, "loss": 1.5208, "step": 9580 }, { "epoch": 0.7508620689655172, "grad_norm": 3.282315254211426, "learning_rate": 9.169344010141038e-06, "loss": 1.4149, "step": 9581 }, { "epoch": 0.7509404388714733, "grad_norm": 2.1417434215545654, "learning_rate": 9.169173797857985e-06, "loss": 1.1069, "step": 9582 }, { "epoch": 0.7510188087774294, "grad_norm": 2.2076261043548584, "learning_rate": 9.169003569717533e-06, "loss": 1.3654, "step": 9583 }, { "epoch": 0.7510971786833855, "grad_norm": 2.3865227699279785, "learning_rate": 9.168833325720334e-06, "loss": 1.1701, "step": 9584 }, { "epoch": 0.7511755485893417, "grad_norm": 2.7401540279388428, "learning_rate": 9.168663065867033e-06, "loss": 1.3504, "step": 9585 }, { "epoch": 0.7512539184952978, "grad_norm": 2.4839730262756348, "learning_rate": 9.168492790158282e-06, "loss": 1.1413, "step": 9586 }, { "epoch": 0.7513322884012539, "grad_norm": 5.7987799644470215, "learning_rate": 9.168322498594721e-06, "loss": 1.0743, "step": 9587 }, { "epoch": 0.75141065830721, "grad_norm": 10.706048965454102, "learning_rate": 9.168152191177005e-06, "loss": 1.4672, "step": 9588 }, { "epoch": 0.7514890282131661, "grad_norm": 3.726656675338745, "learning_rate": 9.167981867905778e-06, "loss": 1.4139, "step": 9589 }, { "epoch": 0.7515673981191222, "grad_norm": 3.2940831184387207, "learning_rate": 9.167811528781689e-06, "loss": 1.1254, "step": 9590 }, { "epoch": 0.7516457680250783, "grad_norm": 4.798154830932617, "learning_rate": 9.167641173805386e-06, "loss": 1.4792, "step": 9591 }, { "epoch": 0.7517241379310344, "grad_norm": 3.637974262237549, "learning_rate": 9.167470802977516e-06, "loss": 1.1237, "step": 9592 }, { "epoch": 0.7518025078369905, "grad_norm": 3.4989283084869385, "learning_rate": 9.16730041629873e-06, "loss": 1.6011, "step": 9593 }, { "epoch": 0.7518808777429468, "grad_norm": 5.989792823791504, "learning_rate": 9.16713001376967e-06, "loss": 1.2938, "step": 9594 }, { "epoch": 0.7519592476489029, "grad_norm": 5.8291916847229, "learning_rate": 9.16695959539099e-06, "loss": 0.9823, "step": 9595 }, { "epoch": 0.752037617554859, "grad_norm": 4.662905693054199, "learning_rate": 9.166789161163336e-06, "loss": 1.5333, "step": 9596 }, { "epoch": 0.7521159874608151, "grad_norm": 2.331334352493286, "learning_rate": 9.166618711087355e-06, "loss": 1.5007, "step": 9597 }, { "epoch": 0.7521943573667712, "grad_norm": 4.124197483062744, "learning_rate": 9.166448245163697e-06, "loss": 0.954, "step": 9598 }, { "epoch": 0.7522727272727273, "grad_norm": 2.2482495307922363, "learning_rate": 9.16627776339301e-06, "loss": 1.3877, "step": 9599 }, { "epoch": 0.7523510971786834, "grad_norm": 2.336388349533081, "learning_rate": 9.166107265775943e-06, "loss": 1.4587, "step": 9600 }, { "epoch": 0.7524294670846395, "grad_norm": 2.7556395530700684, "learning_rate": 9.165936752313142e-06, "loss": 1.2219, "step": 9601 }, { "epoch": 0.7525078369905956, "grad_norm": 3.2750089168548584, "learning_rate": 9.16576622300526e-06, "loss": 0.9891, "step": 9602 }, { "epoch": 0.7525862068965518, "grad_norm": 3.335815668106079, "learning_rate": 9.165595677852941e-06, "loss": 1.3347, "step": 9603 }, { "epoch": 0.7526645768025079, "grad_norm": 3.2207624912261963, "learning_rate": 9.165425116856837e-06, "loss": 1.6048, "step": 9604 }, { "epoch": 0.752742946708464, "grad_norm": 3.2561731338500977, "learning_rate": 9.165254540017595e-06, "loss": 1.6922, "step": 9605 }, { "epoch": 0.7528213166144201, "grad_norm": 2.294889450073242, "learning_rate": 9.165083947335863e-06, "loss": 1.6384, "step": 9606 }, { "epoch": 0.7528996865203762, "grad_norm": 3.7297487258911133, "learning_rate": 9.164913338812293e-06, "loss": 1.8558, "step": 9607 }, { "epoch": 0.7529780564263323, "grad_norm": 2.4039700031280518, "learning_rate": 9.16474271444753e-06, "loss": 1.2764, "step": 9608 }, { "epoch": 0.7530564263322884, "grad_norm": 3.387044906616211, "learning_rate": 9.164572074242225e-06, "loss": 1.5546, "step": 9609 }, { "epoch": 0.7531347962382445, "grad_norm": 4.1265058517456055, "learning_rate": 9.164401418197028e-06, "loss": 1.3655, "step": 9610 }, { "epoch": 0.7532131661442006, "grad_norm": 3.632798671722412, "learning_rate": 9.164230746312586e-06, "loss": 1.4227, "step": 9611 }, { "epoch": 0.7532915360501568, "grad_norm": 3.664698839187622, "learning_rate": 9.164060058589548e-06, "loss": 1.3016, "step": 9612 }, { "epoch": 0.7533699059561129, "grad_norm": 3.5424461364746094, "learning_rate": 9.163889355028568e-06, "loss": 1.5499, "step": 9613 }, { "epoch": 0.753448275862069, "grad_norm": 5.642236232757568, "learning_rate": 9.163718635630289e-06, "loss": 1.0987, "step": 9614 }, { "epoch": 0.7535266457680251, "grad_norm": 2.264943838119507, "learning_rate": 9.163547900395364e-06, "loss": 1.196, "step": 9615 }, { "epoch": 0.7536050156739812, "grad_norm": 2.7836809158325195, "learning_rate": 9.16337714932444e-06, "loss": 1.1322, "step": 9616 }, { "epoch": 0.7536833855799373, "grad_norm": 2.3222639560699463, "learning_rate": 9.163206382418168e-06, "loss": 1.36, "step": 9617 }, { "epoch": 0.7537617554858934, "grad_norm": 3.108370304107666, "learning_rate": 9.163035599677196e-06, "loss": 1.6985, "step": 9618 }, { "epoch": 0.7538401253918495, "grad_norm": 11.757229804992676, "learning_rate": 9.162864801102179e-06, "loss": 2.9457, "step": 9619 }, { "epoch": 0.7539184952978056, "grad_norm": 4.266202449798584, "learning_rate": 9.162693986693761e-06, "loss": 1.4863, "step": 9620 }, { "epoch": 0.7539968652037617, "grad_norm": 2.051140785217285, "learning_rate": 9.162523156452591e-06, "loss": 1.1107, "step": 9621 }, { "epoch": 0.7540752351097179, "grad_norm": 3.8020505905151367, "learning_rate": 9.16235231037932e-06, "loss": 1.5418, "step": 9622 }, { "epoch": 0.754153605015674, "grad_norm": 2.2408699989318848, "learning_rate": 9.162181448474602e-06, "loss": 1.3351, "step": 9623 }, { "epoch": 0.7542319749216301, "grad_norm": 10.695760726928711, "learning_rate": 9.162010570739083e-06, "loss": 1.2391, "step": 9624 }, { "epoch": 0.7543103448275862, "grad_norm": 3.288928508758545, "learning_rate": 9.161839677173413e-06, "loss": 1.1917, "step": 9625 }, { "epoch": 0.7543887147335423, "grad_norm": 3.5136935710906982, "learning_rate": 9.161668767778243e-06, "loss": 1.4267, "step": 9626 }, { "epoch": 0.7544670846394984, "grad_norm": 2.237865447998047, "learning_rate": 9.16149784255422e-06, "loss": 1.5624, "step": 9627 }, { "epoch": 0.7545454545454545, "grad_norm": 2.9072892665863037, "learning_rate": 9.161326901502e-06, "loss": 1.4175, "step": 9628 }, { "epoch": 0.7546238244514106, "grad_norm": 3.294420003890991, "learning_rate": 9.161155944622229e-06, "loss": 1.324, "step": 9629 }, { "epoch": 0.7547021943573667, "grad_norm": 3.9616968631744385, "learning_rate": 9.160984971915557e-06, "loss": 1.8184, "step": 9630 }, { "epoch": 0.7547805642633229, "grad_norm": 4.127350807189941, "learning_rate": 9.160813983382637e-06, "loss": 1.5674, "step": 9631 }, { "epoch": 0.754858934169279, "grad_norm": 2.0877227783203125, "learning_rate": 9.160642979024118e-06, "loss": 1.398, "step": 9632 }, { "epoch": 0.7549373040752351, "grad_norm": 4.180523872375488, "learning_rate": 9.160471958840648e-06, "loss": 1.7427, "step": 9633 }, { "epoch": 0.7550156739811912, "grad_norm": 1.9289805889129639, "learning_rate": 9.160300922832882e-06, "loss": 1.566, "step": 9634 }, { "epoch": 0.7550940438871473, "grad_norm": 2.2967023849487305, "learning_rate": 9.160129871001466e-06, "loss": 1.4501, "step": 9635 }, { "epoch": 0.7551724137931034, "grad_norm": 4.19975471496582, "learning_rate": 9.159958803347052e-06, "loss": 1.3404, "step": 9636 }, { "epoch": 0.7552507836990595, "grad_norm": 2.3154401779174805, "learning_rate": 9.159787719870294e-06, "loss": 1.3882, "step": 9637 }, { "epoch": 0.7553291536050156, "grad_norm": 5.38952112197876, "learning_rate": 9.159616620571838e-06, "loss": 1.7059, "step": 9638 }, { "epoch": 0.7554075235109717, "grad_norm": 2.9719066619873047, "learning_rate": 9.159445505452338e-06, "loss": 1.3639, "step": 9639 }, { "epoch": 0.7554858934169278, "grad_norm": 3.3730266094207764, "learning_rate": 9.159274374512444e-06, "loss": 1.6362, "step": 9640 }, { "epoch": 0.7555642633228841, "grad_norm": 2.8203301429748535, "learning_rate": 9.159103227752806e-06, "loss": 1.3676, "step": 9641 }, { "epoch": 0.7556426332288402, "grad_norm": 3.3370907306671143, "learning_rate": 9.158932065174075e-06, "loss": 1.6236, "step": 9642 }, { "epoch": 0.7557210031347963, "grad_norm": 4.580443859100342, "learning_rate": 9.158760886776905e-06, "loss": 1.6447, "step": 9643 }, { "epoch": 0.7557993730407524, "grad_norm": 3.65560245513916, "learning_rate": 9.158589692561943e-06, "loss": 1.2842, "step": 9644 }, { "epoch": 0.7558777429467085, "grad_norm": 2.7471375465393066, "learning_rate": 9.158418482529841e-06, "loss": 1.6247, "step": 9645 }, { "epoch": 0.7559561128526646, "grad_norm": 2.6801397800445557, "learning_rate": 9.158247256681253e-06, "loss": 1.4571, "step": 9646 }, { "epoch": 0.7560344827586207, "grad_norm": 1.9962855577468872, "learning_rate": 9.158076015016828e-06, "loss": 1.2034, "step": 9647 }, { "epoch": 0.7561128526645768, "grad_norm": 6.1604485511779785, "learning_rate": 9.157904757537215e-06, "loss": 1.6835, "step": 9648 }, { "epoch": 0.756191222570533, "grad_norm": 2.5166242122650146, "learning_rate": 9.15773348424307e-06, "loss": 1.4524, "step": 9649 }, { "epoch": 0.7562695924764891, "grad_norm": 2.7947940826416016, "learning_rate": 9.157562195135042e-06, "loss": 1.5782, "step": 9650 }, { "epoch": 0.7563479623824452, "grad_norm": 2.2424240112304688, "learning_rate": 9.157390890213781e-06, "loss": 1.1711, "step": 9651 }, { "epoch": 0.7564263322884013, "grad_norm": 1.8767367601394653, "learning_rate": 9.157219569479945e-06, "loss": 0.953, "step": 9652 }, { "epoch": 0.7565047021943574, "grad_norm": 6.535003662109375, "learning_rate": 9.157048232934178e-06, "loss": 1.3594, "step": 9653 }, { "epoch": 0.7565830721003135, "grad_norm": 2.7115883827209473, "learning_rate": 9.156876880577134e-06, "loss": 1.6384, "step": 9654 }, { "epoch": 0.7566614420062696, "grad_norm": 3.2475292682647705, "learning_rate": 9.156705512409468e-06, "loss": 1.2067, "step": 9655 }, { "epoch": 0.7567398119122257, "grad_norm": 2.5175163745880127, "learning_rate": 9.156534128431827e-06, "loss": 1.3961, "step": 9656 }, { "epoch": 0.7568181818181818, "grad_norm": 2.5817346572875977, "learning_rate": 9.156362728644865e-06, "loss": 1.1057, "step": 9657 }, { "epoch": 0.756896551724138, "grad_norm": 2.371966600418091, "learning_rate": 9.156191313049234e-06, "loss": 1.5698, "step": 9658 }, { "epoch": 0.756974921630094, "grad_norm": 11.14169979095459, "learning_rate": 9.156019881645589e-06, "loss": 2.3195, "step": 9659 }, { "epoch": 0.7570532915360502, "grad_norm": 6.620974063873291, "learning_rate": 9.155848434434575e-06, "loss": 0.9575, "step": 9660 }, { "epoch": 0.7571316614420063, "grad_norm": 2.3659777641296387, "learning_rate": 9.155676971416851e-06, "loss": 1.5044, "step": 9661 }, { "epoch": 0.7572100313479624, "grad_norm": 3.2395758628845215, "learning_rate": 9.155505492593067e-06, "loss": 1.5462, "step": 9662 }, { "epoch": 0.7572884012539185, "grad_norm": 1.9485880136489868, "learning_rate": 9.155333997963872e-06, "loss": 1.1209, "step": 9663 }, { "epoch": 0.7573667711598746, "grad_norm": 3.099475145339966, "learning_rate": 9.15516248752992e-06, "loss": 1.0372, "step": 9664 }, { "epoch": 0.7574451410658307, "grad_norm": 2.165163993835449, "learning_rate": 9.154990961291868e-06, "loss": 1.3941, "step": 9665 }, { "epoch": 0.7575235109717868, "grad_norm": 3.20829701423645, "learning_rate": 9.154819419250361e-06, "loss": 1.1467, "step": 9666 }, { "epoch": 0.7576018808777429, "grad_norm": 12.301413536071777, "learning_rate": 9.154647861406057e-06, "loss": 1.218, "step": 9667 }, { "epoch": 0.757680250783699, "grad_norm": 2.225292205810547, "learning_rate": 9.154476287759606e-06, "loss": 1.4338, "step": 9668 }, { "epoch": 0.7577586206896552, "grad_norm": 10.673547744750977, "learning_rate": 9.15430469831166e-06, "loss": 1.3817, "step": 9669 }, { "epoch": 0.7578369905956113, "grad_norm": 2.5223066806793213, "learning_rate": 9.154133093062874e-06, "loss": 1.6585, "step": 9670 }, { "epoch": 0.7579153605015674, "grad_norm": 3.443495750427246, "learning_rate": 9.1539614720139e-06, "loss": 1.467, "step": 9671 }, { "epoch": 0.7579937304075235, "grad_norm": 9.572517395019531, "learning_rate": 9.153789835165387e-06, "loss": 1.1143, "step": 9672 }, { "epoch": 0.7580721003134796, "grad_norm": 2.7523977756500244, "learning_rate": 9.153618182517995e-06, "loss": 1.3473, "step": 9673 }, { "epoch": 0.7581504702194357, "grad_norm": 2.9380674362182617, "learning_rate": 9.15344651407237e-06, "loss": 1.4507, "step": 9674 }, { "epoch": 0.7582288401253918, "grad_norm": 5.734396934509277, "learning_rate": 9.15327482982917e-06, "loss": 1.0691, "step": 9675 }, { "epoch": 0.7583072100313479, "grad_norm": 2.3586864471435547, "learning_rate": 9.153103129789043e-06, "loss": 1.0625, "step": 9676 }, { "epoch": 0.758385579937304, "grad_norm": 2.5223281383514404, "learning_rate": 9.152931413952647e-06, "loss": 1.4431, "step": 9677 }, { "epoch": 0.7584639498432602, "grad_norm": 2.4243569374084473, "learning_rate": 9.152759682320632e-06, "loss": 1.5378, "step": 9678 }, { "epoch": 0.7585423197492163, "grad_norm": 2.8367908000946045, "learning_rate": 9.152587934893653e-06, "loss": 1.4078, "step": 9679 }, { "epoch": 0.7586206896551724, "grad_norm": 4.225801467895508, "learning_rate": 9.152416171672361e-06, "loss": 1.5731, "step": 9680 }, { "epoch": 0.7586990595611285, "grad_norm": 3.4222073554992676, "learning_rate": 9.152244392657412e-06, "loss": 1.4494, "step": 9681 }, { "epoch": 0.7587774294670846, "grad_norm": 2.053709030151367, "learning_rate": 9.152072597849458e-06, "loss": 1.0727, "step": 9682 }, { "epoch": 0.7588557993730407, "grad_norm": 2.5605084896087646, "learning_rate": 9.151900787249153e-06, "loss": 1.1993, "step": 9683 }, { "epoch": 0.7589341692789968, "grad_norm": 1.979792594909668, "learning_rate": 9.151728960857149e-06, "loss": 1.4757, "step": 9684 }, { "epoch": 0.7590125391849529, "grad_norm": 4.161202430725098, "learning_rate": 9.1515571186741e-06, "loss": 1.3895, "step": 9685 }, { "epoch": 0.759090909090909, "grad_norm": 1.9765217304229736, "learning_rate": 9.15138526070066e-06, "loss": 1.4296, "step": 9686 }, { "epoch": 0.7591692789968651, "grad_norm": 4.727725505828857, "learning_rate": 9.151213386937485e-06, "loss": 1.4364, "step": 9687 }, { "epoch": 0.7592476489028214, "grad_norm": 2.6581671237945557, "learning_rate": 9.151041497385224e-06, "loss": 1.432, "step": 9688 }, { "epoch": 0.7593260188087775, "grad_norm": 3.447765827178955, "learning_rate": 9.150869592044535e-06, "loss": 1.5399, "step": 9689 }, { "epoch": 0.7594043887147336, "grad_norm": 4.039214134216309, "learning_rate": 9.150697670916068e-06, "loss": 1.5319, "step": 9690 }, { "epoch": 0.7594827586206897, "grad_norm": 2.041229486465454, "learning_rate": 9.150525734000481e-06, "loss": 1.4179, "step": 9691 }, { "epoch": 0.7595611285266458, "grad_norm": 2.602339029312134, "learning_rate": 9.150353781298426e-06, "loss": 1.2085, "step": 9692 }, { "epoch": 0.7596394984326019, "grad_norm": 5.559359073638916, "learning_rate": 9.150181812810555e-06, "loss": 1.2649, "step": 9693 }, { "epoch": 0.759717868338558, "grad_norm": 1.9289029836654663, "learning_rate": 9.150009828537527e-06, "loss": 1.3594, "step": 9694 }, { "epoch": 0.7597962382445141, "grad_norm": 6.348574161529541, "learning_rate": 9.149837828479991e-06, "loss": 1.8731, "step": 9695 }, { "epoch": 0.7598746081504703, "grad_norm": 4.334354400634766, "learning_rate": 9.149665812638604e-06, "loss": 1.0986, "step": 9696 }, { "epoch": 0.7599529780564264, "grad_norm": 2.2345545291900635, "learning_rate": 9.149493781014019e-06, "loss": 1.1294, "step": 9697 }, { "epoch": 0.7600313479623825, "grad_norm": 2.2978179454803467, "learning_rate": 9.149321733606892e-06, "loss": 1.5355, "step": 9698 }, { "epoch": 0.7601097178683386, "grad_norm": 2.9617912769317627, "learning_rate": 9.149149670417876e-06, "loss": 0.8829, "step": 9699 }, { "epoch": 0.7601880877742947, "grad_norm": 3.423391103744507, "learning_rate": 9.148977591447625e-06, "loss": 1.3935, "step": 9700 }, { "epoch": 0.7602664576802508, "grad_norm": 3.199568510055542, "learning_rate": 9.148805496696795e-06, "loss": 1.1592, "step": 9701 }, { "epoch": 0.7603448275862069, "grad_norm": 5.6300578117370605, "learning_rate": 9.14863338616604e-06, "loss": 1.1262, "step": 9702 }, { "epoch": 0.760423197492163, "grad_norm": 2.7872202396392822, "learning_rate": 9.148461259856014e-06, "loss": 1.1263, "step": 9703 }, { "epoch": 0.7605015673981191, "grad_norm": 2.322528839111328, "learning_rate": 9.148289117767373e-06, "loss": 1.4202, "step": 9704 }, { "epoch": 0.7605799373040752, "grad_norm": 2.880871295928955, "learning_rate": 9.14811695990077e-06, "loss": 1.4315, "step": 9705 }, { "epoch": 0.7606583072100314, "grad_norm": 2.85080623626709, "learning_rate": 9.14794478625686e-06, "loss": 1.6206, "step": 9706 }, { "epoch": 0.7607366771159875, "grad_norm": 2.4258100986480713, "learning_rate": 9.1477725968363e-06, "loss": 1.5403, "step": 9707 }, { "epoch": 0.7608150470219436, "grad_norm": 1.7321943044662476, "learning_rate": 9.147600391639744e-06, "loss": 1.4765, "step": 9708 }, { "epoch": 0.7608934169278997, "grad_norm": 4.19921875, "learning_rate": 9.147428170667845e-06, "loss": 1.3621, "step": 9709 }, { "epoch": 0.7609717868338558, "grad_norm": 4.795596122741699, "learning_rate": 9.14725593392126e-06, "loss": 1.3329, "step": 9710 }, { "epoch": 0.7610501567398119, "grad_norm": 2.281158208847046, "learning_rate": 9.147083681400643e-06, "loss": 1.5576, "step": 9711 }, { "epoch": 0.761128526645768, "grad_norm": 2.9636166095733643, "learning_rate": 9.146911413106651e-06, "loss": 0.7371, "step": 9712 }, { "epoch": 0.7612068965517241, "grad_norm": 2.764907121658325, "learning_rate": 9.146739129039938e-06, "loss": 1.3139, "step": 9713 }, { "epoch": 0.7612852664576802, "grad_norm": 4.477548122406006, "learning_rate": 9.146566829201157e-06, "loss": 1.8294, "step": 9714 }, { "epoch": 0.7613636363636364, "grad_norm": 2.5693447589874268, "learning_rate": 9.146394513590968e-06, "loss": 1.2659, "step": 9715 }, { "epoch": 0.7614420062695925, "grad_norm": 3.521448850631714, "learning_rate": 9.146222182210023e-06, "loss": 1.6168, "step": 9716 }, { "epoch": 0.7615203761755486, "grad_norm": 2.1054515838623047, "learning_rate": 9.146049835058979e-06, "loss": 1.2648, "step": 9717 }, { "epoch": 0.7615987460815047, "grad_norm": 2.5420753955841064, "learning_rate": 9.145877472138491e-06, "loss": 1.243, "step": 9718 }, { "epoch": 0.7616771159874608, "grad_norm": 2.1727452278137207, "learning_rate": 9.145705093449213e-06, "loss": 1.3439, "step": 9719 }, { "epoch": 0.7617554858934169, "grad_norm": 2.098878860473633, "learning_rate": 9.145532698991803e-06, "loss": 1.2236, "step": 9720 }, { "epoch": 0.761833855799373, "grad_norm": 2.6524417400360107, "learning_rate": 9.145360288766918e-06, "loss": 1.1682, "step": 9721 }, { "epoch": 0.7619122257053291, "grad_norm": 4.535770416259766, "learning_rate": 9.145187862775208e-06, "loss": 1.6622, "step": 9722 }, { "epoch": 0.7619905956112852, "grad_norm": 3.8167059421539307, "learning_rate": 9.145015421017336e-06, "loss": 1.5083, "step": 9723 }, { "epoch": 0.7620689655172413, "grad_norm": 3.0114831924438477, "learning_rate": 9.144842963493953e-06, "loss": 1.1353, "step": 9724 }, { "epoch": 0.7621473354231975, "grad_norm": 3.1392858028411865, "learning_rate": 9.144670490205716e-06, "loss": 1.3111, "step": 9725 }, { "epoch": 0.7622257053291536, "grad_norm": 5.159499168395996, "learning_rate": 9.144498001153281e-06, "loss": 1.3463, "step": 9726 }, { "epoch": 0.7623040752351097, "grad_norm": 3.3476455211639404, "learning_rate": 9.144325496337305e-06, "loss": 1.3298, "step": 9727 }, { "epoch": 0.7623824451410658, "grad_norm": 2.1950135231018066, "learning_rate": 9.144152975758442e-06, "loss": 1.3186, "step": 9728 }, { "epoch": 0.7624608150470219, "grad_norm": 2.4557723999023438, "learning_rate": 9.14398043941735e-06, "loss": 1.5326, "step": 9729 }, { "epoch": 0.762539184952978, "grad_norm": 2.6437559127807617, "learning_rate": 9.143807887314686e-06, "loss": 1.6005, "step": 9730 }, { "epoch": 0.7626175548589341, "grad_norm": 5.919617176055908, "learning_rate": 9.143635319451106e-06, "loss": 1.1549, "step": 9731 }, { "epoch": 0.7626959247648902, "grad_norm": 3.0588369369506836, "learning_rate": 9.143462735827265e-06, "loss": 1.2221, "step": 9732 }, { "epoch": 0.7627742946708463, "grad_norm": 2.4046099185943604, "learning_rate": 9.143290136443818e-06, "loss": 1.1562, "step": 9733 }, { "epoch": 0.7628526645768025, "grad_norm": 4.061573028564453, "learning_rate": 9.143117521301425e-06, "loss": 2.0054, "step": 9734 }, { "epoch": 0.7629310344827587, "grad_norm": 3.6571052074432373, "learning_rate": 9.142944890400742e-06, "loss": 1.7632, "step": 9735 }, { "epoch": 0.7630094043887148, "grad_norm": 5.484127521514893, "learning_rate": 9.142772243742423e-06, "loss": 1.4167, "step": 9736 }, { "epoch": 0.7630877742946709, "grad_norm": 4.1992974281311035, "learning_rate": 9.142599581327126e-06, "loss": 1.4531, "step": 9737 }, { "epoch": 0.763166144200627, "grad_norm": 2.2791357040405273, "learning_rate": 9.142426903155507e-06, "loss": 1.0898, "step": 9738 }, { "epoch": 0.7632445141065831, "grad_norm": 4.522120475769043, "learning_rate": 9.142254209228227e-06, "loss": 0.9807, "step": 9739 }, { "epoch": 0.7633228840125392, "grad_norm": 2.1541666984558105, "learning_rate": 9.142081499545939e-06, "loss": 1.0936, "step": 9740 }, { "epoch": 0.7634012539184953, "grad_norm": 2.3474626541137695, "learning_rate": 9.141908774109299e-06, "loss": 1.5229, "step": 9741 }, { "epoch": 0.7634796238244514, "grad_norm": 3.2080719470977783, "learning_rate": 9.141736032918966e-06, "loss": 1.6532, "step": 9742 }, { "epoch": 0.7635579937304076, "grad_norm": 3.5889532566070557, "learning_rate": 9.141563275975595e-06, "loss": 0.9584, "step": 9743 }, { "epoch": 0.7636363636363637, "grad_norm": 2.7844808101654053, "learning_rate": 9.141390503279846e-06, "loss": 1.5002, "step": 9744 }, { "epoch": 0.7637147335423198, "grad_norm": 2.8789870738983154, "learning_rate": 9.141217714832375e-06, "loss": 1.2928, "step": 9745 }, { "epoch": 0.7637931034482759, "grad_norm": 2.628783702850342, "learning_rate": 9.141044910633839e-06, "loss": 1.4419, "step": 9746 }, { "epoch": 0.763871473354232, "grad_norm": 3.4511983394622803, "learning_rate": 9.140872090684893e-06, "loss": 1.6273, "step": 9747 }, { "epoch": 0.7639498432601881, "grad_norm": 2.4992449283599854, "learning_rate": 9.140699254986198e-06, "loss": 1.241, "step": 9748 }, { "epoch": 0.7640282131661442, "grad_norm": 3.0591936111450195, "learning_rate": 9.14052640353841e-06, "loss": 1.5876, "step": 9749 }, { "epoch": 0.7641065830721003, "grad_norm": 3.830087661743164, "learning_rate": 9.140353536342187e-06, "loss": 1.6526, "step": 9750 }, { "epoch": 0.7641849529780564, "grad_norm": 2.958308219909668, "learning_rate": 9.140180653398184e-06, "loss": 1.4031, "step": 9751 }, { "epoch": 0.7642633228840126, "grad_norm": 4.68212366104126, "learning_rate": 9.14000775470706e-06, "loss": 2.1952, "step": 9752 }, { "epoch": 0.7643416927899687, "grad_norm": 2.2131104469299316, "learning_rate": 9.139834840269475e-06, "loss": 1.3958, "step": 9753 }, { "epoch": 0.7644200626959248, "grad_norm": 4.48800802230835, "learning_rate": 9.139661910086083e-06, "loss": 0.9332, "step": 9754 }, { "epoch": 0.7644984326018809, "grad_norm": 2.6631720066070557, "learning_rate": 9.139488964157546e-06, "loss": 1.3474, "step": 9755 }, { "epoch": 0.764576802507837, "grad_norm": 2.382746934890747, "learning_rate": 9.139316002484517e-06, "loss": 1.382, "step": 9756 }, { "epoch": 0.7646551724137931, "grad_norm": 3.2637808322906494, "learning_rate": 9.139143025067657e-06, "loss": 1.3261, "step": 9757 }, { "epoch": 0.7647335423197492, "grad_norm": 3.680051565170288, "learning_rate": 9.13897003190762e-06, "loss": 1.1838, "step": 9758 }, { "epoch": 0.7648119122257053, "grad_norm": 3.4343483448028564, "learning_rate": 9.13879702300507e-06, "loss": 1.4688, "step": 9759 }, { "epoch": 0.7648902821316614, "grad_norm": 5.40226411819458, "learning_rate": 9.138623998360662e-06, "loss": 1.8455, "step": 9760 }, { "epoch": 0.7649686520376175, "grad_norm": 5.254636287689209, "learning_rate": 9.138450957975053e-06, "loss": 1.4005, "step": 9761 }, { "epoch": 0.7650470219435737, "grad_norm": 2.3727617263793945, "learning_rate": 9.138277901848902e-06, "loss": 1.5171, "step": 9762 }, { "epoch": 0.7651253918495298, "grad_norm": 3.259303092956543, "learning_rate": 9.138104829982868e-06, "loss": 0.8244, "step": 9763 }, { "epoch": 0.7652037617554859, "grad_norm": 2.7003917694091797, "learning_rate": 9.137931742377608e-06, "loss": 1.8171, "step": 9764 }, { "epoch": 0.765282131661442, "grad_norm": 3.632668972015381, "learning_rate": 9.137758639033783e-06, "loss": 1.3384, "step": 9765 }, { "epoch": 0.7653605015673981, "grad_norm": 2.2410714626312256, "learning_rate": 9.137585519952046e-06, "loss": 1.1574, "step": 9766 }, { "epoch": 0.7654388714733542, "grad_norm": 3.6097748279571533, "learning_rate": 9.137412385133062e-06, "loss": 1.3827, "step": 9767 }, { "epoch": 0.7655172413793103, "grad_norm": 4.142628192901611, "learning_rate": 9.137239234577484e-06, "loss": 1.9387, "step": 9768 }, { "epoch": 0.7655956112852664, "grad_norm": 2.7203850746154785, "learning_rate": 9.137066068285975e-06, "loss": 1.202, "step": 9769 }, { "epoch": 0.7656739811912225, "grad_norm": 3.2793195247650146, "learning_rate": 9.136892886259189e-06, "loss": 1.5165, "step": 9770 }, { "epoch": 0.7657523510971787, "grad_norm": 2.845921277999878, "learning_rate": 9.13671968849779e-06, "loss": 1.6257, "step": 9771 }, { "epoch": 0.7658307210031348, "grad_norm": 3.454139471054077, "learning_rate": 9.136546475002433e-06, "loss": 1.5845, "step": 9772 }, { "epoch": 0.7659090909090909, "grad_norm": 1.987236738204956, "learning_rate": 9.136373245773778e-06, "loss": 1.1549, "step": 9773 }, { "epoch": 0.765987460815047, "grad_norm": 3.1992506980895996, "learning_rate": 9.136200000812483e-06, "loss": 1.4091, "step": 9774 }, { "epoch": 0.7660658307210031, "grad_norm": 2.034281015396118, "learning_rate": 9.13602674011921e-06, "loss": 1.5888, "step": 9775 }, { "epoch": 0.7661442006269592, "grad_norm": 4.436159610748291, "learning_rate": 9.135853463694614e-06, "loss": 1.2035, "step": 9776 }, { "epoch": 0.7662225705329153, "grad_norm": 2.4043500423431396, "learning_rate": 9.135680171539356e-06, "loss": 1.494, "step": 9777 }, { "epoch": 0.7663009404388714, "grad_norm": 2.742537021636963, "learning_rate": 9.135506863654094e-06, "loss": 1.6121, "step": 9778 }, { "epoch": 0.7663793103448275, "grad_norm": 4.396435260772705, "learning_rate": 9.135333540039488e-06, "loss": 1.2946, "step": 9779 }, { "epoch": 0.7664576802507836, "grad_norm": 5.085222244262695, "learning_rate": 9.135160200696197e-06, "loss": 1.4998, "step": 9780 }, { "epoch": 0.7665360501567398, "grad_norm": 2.178530693054199, "learning_rate": 9.134986845624882e-06, "loss": 1.3276, "step": 9781 }, { "epoch": 0.766614420062696, "grad_norm": 2.5859408378601074, "learning_rate": 9.134813474826202e-06, "loss": 1.3636, "step": 9782 }, { "epoch": 0.7666927899686521, "grad_norm": 2.3092353343963623, "learning_rate": 9.134640088300813e-06, "loss": 1.3334, "step": 9783 }, { "epoch": 0.7667711598746082, "grad_norm": 3.545591354370117, "learning_rate": 9.134466686049376e-06, "loss": 1.6375, "step": 9784 }, { "epoch": 0.7668495297805643, "grad_norm": 2.384298324584961, "learning_rate": 9.134293268072554e-06, "loss": 1.3603, "step": 9785 }, { "epoch": 0.7669278996865204, "grad_norm": 1.9561344385147095, "learning_rate": 9.134119834371002e-06, "loss": 1.624, "step": 9786 }, { "epoch": 0.7670062695924765, "grad_norm": 4.079212665557861, "learning_rate": 9.133946384945383e-06, "loss": 1.7292, "step": 9787 }, { "epoch": 0.7670846394984326, "grad_norm": 2.687469244003296, "learning_rate": 9.133772919796356e-06, "loss": 1.538, "step": 9788 }, { "epoch": 0.7671630094043888, "grad_norm": 3.0919277667999268, "learning_rate": 9.133599438924578e-06, "loss": 1.3451, "step": 9789 }, { "epoch": 0.7672413793103449, "grad_norm": 4.21734619140625, "learning_rate": 9.133425942330711e-06, "loss": 1.2881, "step": 9790 }, { "epoch": 0.767319749216301, "grad_norm": 3.1302614212036133, "learning_rate": 9.133252430015415e-06, "loss": 1.3551, "step": 9791 }, { "epoch": 0.7673981191222571, "grad_norm": 3.2710039615631104, "learning_rate": 9.133078901979352e-06, "loss": 1.665, "step": 9792 }, { "epoch": 0.7674764890282132, "grad_norm": 6.9011359214782715, "learning_rate": 9.132905358223177e-06, "loss": 1.9125, "step": 9793 }, { "epoch": 0.7675548589341693, "grad_norm": 2.446918487548828, "learning_rate": 9.132731798747552e-06, "loss": 0.9072, "step": 9794 }, { "epoch": 0.7676332288401254, "grad_norm": 2.965822696685791, "learning_rate": 9.132558223553142e-06, "loss": 0.9538, "step": 9795 }, { "epoch": 0.7677115987460815, "grad_norm": 3.6652309894561768, "learning_rate": 9.1323846326406e-06, "loss": 1.1141, "step": 9796 }, { "epoch": 0.7677899686520376, "grad_norm": 4.854565143585205, "learning_rate": 9.132211026010588e-06, "loss": 1.1962, "step": 9797 }, { "epoch": 0.7678683385579937, "grad_norm": 2.639680862426758, "learning_rate": 9.13203740366377e-06, "loss": 1.3322, "step": 9798 }, { "epoch": 0.7679467084639499, "grad_norm": 3.7201268672943115, "learning_rate": 9.131863765600804e-06, "loss": 1.792, "step": 9799 }, { "epoch": 0.768025078369906, "grad_norm": 3.3556430339813232, "learning_rate": 9.13169011182235e-06, "loss": 1.3745, "step": 9800 }, { "epoch": 0.7681034482758621, "grad_norm": 2.115455150604248, "learning_rate": 9.13151644232907e-06, "loss": 1.6556, "step": 9801 }, { "epoch": 0.7681818181818182, "grad_norm": 3.3932993412017822, "learning_rate": 9.131342757121623e-06, "loss": 1.451, "step": 9802 }, { "epoch": 0.7682601880877743, "grad_norm": 2.587374210357666, "learning_rate": 9.13116905620067e-06, "loss": 1.4553, "step": 9803 }, { "epoch": 0.7683385579937304, "grad_norm": 2.212240219116211, "learning_rate": 9.130995339566871e-06, "loss": 1.2975, "step": 9804 }, { "epoch": 0.7684169278996865, "grad_norm": 2.2106406688690186, "learning_rate": 9.13082160722089e-06, "loss": 1.3364, "step": 9805 }, { "epoch": 0.7684952978056426, "grad_norm": 2.4978280067443848, "learning_rate": 9.130647859163381e-06, "loss": 1.2995, "step": 9806 }, { "epoch": 0.7685736677115987, "grad_norm": 3.169147491455078, "learning_rate": 9.130474095395013e-06, "loss": 1.2275, "step": 9807 }, { "epoch": 0.7686520376175549, "grad_norm": 3.729851245880127, "learning_rate": 9.130300315916443e-06, "loss": 1.3914, "step": 9808 }, { "epoch": 0.768730407523511, "grad_norm": 2.680215835571289, "learning_rate": 9.130126520728331e-06, "loss": 1.2079, "step": 9809 }, { "epoch": 0.7688087774294671, "grad_norm": 3.882267475128174, "learning_rate": 9.129952709831338e-06, "loss": 1.0026, "step": 9810 }, { "epoch": 0.7688871473354232, "grad_norm": 3.320850372314453, "learning_rate": 9.129778883226128e-06, "loss": 1.3562, "step": 9811 }, { "epoch": 0.7689655172413793, "grad_norm": 2.6104087829589844, "learning_rate": 9.129605040913358e-06, "loss": 1.2958, "step": 9812 }, { "epoch": 0.7690438871473354, "grad_norm": 2.970881223678589, "learning_rate": 9.129431182893693e-06, "loss": 1.4865, "step": 9813 }, { "epoch": 0.7691222570532915, "grad_norm": 3.9561920166015625, "learning_rate": 9.129257309167793e-06, "loss": 1.1847, "step": 9814 }, { "epoch": 0.7692006269592476, "grad_norm": 1.7586898803710938, "learning_rate": 9.12908341973632e-06, "loss": 1.1761, "step": 9815 }, { "epoch": 0.7692789968652037, "grad_norm": 2.5599663257598877, "learning_rate": 9.128909514599933e-06, "loss": 1.531, "step": 9816 }, { "epoch": 0.7693573667711598, "grad_norm": 2.9313302040100098, "learning_rate": 9.128735593759297e-06, "loss": 1.6225, "step": 9817 }, { "epoch": 0.769435736677116, "grad_norm": 2.693913698196411, "learning_rate": 9.128561657215069e-06, "loss": 1.3495, "step": 9818 }, { "epoch": 0.7695141065830721, "grad_norm": 8.236421585083008, "learning_rate": 9.128387704967914e-06, "loss": 2.2531, "step": 9819 }, { "epoch": 0.7695924764890282, "grad_norm": 2.8309545516967773, "learning_rate": 9.128213737018493e-06, "loss": 1.2007, "step": 9820 }, { "epoch": 0.7696708463949843, "grad_norm": 2.4586410522460938, "learning_rate": 9.128039753367466e-06, "loss": 1.5372, "step": 9821 }, { "epoch": 0.7697492163009404, "grad_norm": 2.81182861328125, "learning_rate": 9.127865754015498e-06, "loss": 1.5376, "step": 9822 }, { "epoch": 0.7698275862068965, "grad_norm": 6.49778413772583, "learning_rate": 9.127691738963247e-06, "loss": 1.4552, "step": 9823 }, { "epoch": 0.7699059561128526, "grad_norm": 2.6680335998535156, "learning_rate": 9.127517708211378e-06, "loss": 1.4151, "step": 9824 }, { "epoch": 0.7699843260188087, "grad_norm": 3.7594704627990723, "learning_rate": 9.12734366176055e-06, "loss": 1.1699, "step": 9825 }, { "epoch": 0.7700626959247648, "grad_norm": 2.4623217582702637, "learning_rate": 9.127169599611428e-06, "loss": 1.8587, "step": 9826 }, { "epoch": 0.770141065830721, "grad_norm": 3.500129222869873, "learning_rate": 9.126995521764673e-06, "loss": 1.0119, "step": 9827 }, { "epoch": 0.7702194357366771, "grad_norm": 2.4196934700012207, "learning_rate": 9.126821428220947e-06, "loss": 1.2636, "step": 9828 }, { "epoch": 0.7702978056426333, "grad_norm": 3.2812047004699707, "learning_rate": 9.12664731898091e-06, "loss": 1.6801, "step": 9829 }, { "epoch": 0.7703761755485894, "grad_norm": 2.618001699447632, "learning_rate": 9.126473194045227e-06, "loss": 1.8724, "step": 9830 }, { "epoch": 0.7704545454545455, "grad_norm": 5.499317646026611, "learning_rate": 9.12629905341456e-06, "loss": 0.4137, "step": 9831 }, { "epoch": 0.7705329153605016, "grad_norm": 3.234898090362549, "learning_rate": 9.126124897089572e-06, "loss": 1.6145, "step": 9832 }, { "epoch": 0.7706112852664577, "grad_norm": 2.8992929458618164, "learning_rate": 9.125950725070921e-06, "loss": 1.3393, "step": 9833 }, { "epoch": 0.7706896551724138, "grad_norm": 2.227433443069458, "learning_rate": 9.125776537359276e-06, "loss": 1.202, "step": 9834 }, { "epoch": 0.77076802507837, "grad_norm": 5.738996982574463, "learning_rate": 9.125602333955294e-06, "loss": 1.271, "step": 9835 }, { "epoch": 0.770846394984326, "grad_norm": 2.682152032852173, "learning_rate": 9.125428114859642e-06, "loss": 1.1921, "step": 9836 }, { "epoch": 0.7709247648902822, "grad_norm": 4.243792533874512, "learning_rate": 9.125253880072979e-06, "loss": 1.8486, "step": 9837 }, { "epoch": 0.7710031347962383, "grad_norm": 3.3790535926818848, "learning_rate": 9.125079629595967e-06, "loss": 1.2535, "step": 9838 }, { "epoch": 0.7710815047021944, "grad_norm": 3.200166940689087, "learning_rate": 9.124905363429274e-06, "loss": 1.3448, "step": 9839 }, { "epoch": 0.7711598746081505, "grad_norm": 4.144827842712402, "learning_rate": 9.12473108157356e-06, "loss": 1.6286, "step": 9840 }, { "epoch": 0.7712382445141066, "grad_norm": 2.8590474128723145, "learning_rate": 9.124556784029485e-06, "loss": 1.2729, "step": 9841 }, { "epoch": 0.7713166144200627, "grad_norm": 3.0359394550323486, "learning_rate": 9.124382470797717e-06, "loss": 1.1296, "step": 9842 }, { "epoch": 0.7713949843260188, "grad_norm": 3.403090476989746, "learning_rate": 9.124208141878915e-06, "loss": 1.3149, "step": 9843 }, { "epoch": 0.7714733542319749, "grad_norm": 2.095397710800171, "learning_rate": 9.124033797273742e-06, "loss": 1.4288, "step": 9844 }, { "epoch": 0.771551724137931, "grad_norm": 6.109607696533203, "learning_rate": 9.123859436982865e-06, "loss": 2.4876, "step": 9845 }, { "epoch": 0.7716300940438872, "grad_norm": 2.9265191555023193, "learning_rate": 9.123685061006944e-06, "loss": 1.5477, "step": 9846 }, { "epoch": 0.7717084639498433, "grad_norm": 3.913520097732544, "learning_rate": 9.123510669346644e-06, "loss": 1.4882, "step": 9847 }, { "epoch": 0.7717868338557994, "grad_norm": 5.145763874053955, "learning_rate": 9.123336262002627e-06, "loss": 0.9148, "step": 9848 }, { "epoch": 0.7718652037617555, "grad_norm": 2.0488297939300537, "learning_rate": 9.123161838975556e-06, "loss": 1.2637, "step": 9849 }, { "epoch": 0.7719435736677116, "grad_norm": 3.146812915802002, "learning_rate": 9.122987400266095e-06, "loss": 1.1869, "step": 9850 }, { "epoch": 0.7720219435736677, "grad_norm": 2.24336576461792, "learning_rate": 9.122812945874909e-06, "loss": 1.4358, "step": 9851 }, { "epoch": 0.7721003134796238, "grad_norm": 3.1968607902526855, "learning_rate": 9.122638475802659e-06, "loss": 1.4484, "step": 9852 }, { "epoch": 0.7721786833855799, "grad_norm": 3.0329887866973877, "learning_rate": 9.12246399005001e-06, "loss": 1.826, "step": 9853 }, { "epoch": 0.772257053291536, "grad_norm": 3.9624409675598145, "learning_rate": 9.122289488617626e-06, "loss": 1.5072, "step": 9854 }, { "epoch": 0.7723354231974922, "grad_norm": 1.9898653030395508, "learning_rate": 9.122114971506169e-06, "loss": 1.1339, "step": 9855 }, { "epoch": 0.7724137931034483, "grad_norm": 2.8249623775482178, "learning_rate": 9.121940438716305e-06, "loss": 1.5152, "step": 9856 }, { "epoch": 0.7724921630094044, "grad_norm": 3.805832624435425, "learning_rate": 9.121765890248698e-06, "loss": 1.6745, "step": 9857 }, { "epoch": 0.7725705329153605, "grad_norm": 3.9207119941711426, "learning_rate": 9.121591326104006e-06, "loss": 1.626, "step": 9858 }, { "epoch": 0.7726489028213166, "grad_norm": 3.7268025875091553, "learning_rate": 9.121416746282902e-06, "loss": 0.9579, "step": 9859 }, { "epoch": 0.7727272727272727, "grad_norm": 9.05630874633789, "learning_rate": 9.121242150786044e-06, "loss": 1.8978, "step": 9860 }, { "epoch": 0.7728056426332288, "grad_norm": 2.4065053462982178, "learning_rate": 9.121067539614097e-06, "loss": 1.2034, "step": 9861 }, { "epoch": 0.7728840125391849, "grad_norm": 2.90413761138916, "learning_rate": 9.120892912767728e-06, "loss": 1.1038, "step": 9862 }, { "epoch": 0.772962382445141, "grad_norm": 3.5137641429901123, "learning_rate": 9.120718270247596e-06, "loss": 1.4496, "step": 9863 }, { "epoch": 0.7730407523510971, "grad_norm": 3.3931431770324707, "learning_rate": 9.12054361205437e-06, "loss": 1.3765, "step": 9864 }, { "epoch": 0.7731191222570533, "grad_norm": 6.4970269203186035, "learning_rate": 9.120368938188713e-06, "loss": 1.6639, "step": 9865 }, { "epoch": 0.7731974921630094, "grad_norm": 9.459409713745117, "learning_rate": 9.120194248651289e-06, "loss": 1.8297, "step": 9866 }, { "epoch": 0.7732758620689655, "grad_norm": 2.6417133808135986, "learning_rate": 9.12001954344276e-06, "loss": 1.6007, "step": 9867 }, { "epoch": 0.7733542319749216, "grad_norm": 2.3027615547180176, "learning_rate": 9.119844822563794e-06, "loss": 1.3255, "step": 9868 }, { "epoch": 0.7734326018808777, "grad_norm": 2.3008503913879395, "learning_rate": 9.119670086015055e-06, "loss": 1.3178, "step": 9869 }, { "epoch": 0.7735109717868338, "grad_norm": 1.8225370645523071, "learning_rate": 9.119495333797208e-06, "loss": 1.2292, "step": 9870 }, { "epoch": 0.7735893416927899, "grad_norm": 2.0663628578186035, "learning_rate": 9.119320565910917e-06, "loss": 1.4293, "step": 9871 }, { "epoch": 0.773667711598746, "grad_norm": 2.356821298599243, "learning_rate": 9.119145782356844e-06, "loss": 1.4513, "step": 9872 }, { "epoch": 0.7737460815047021, "grad_norm": 3.3006110191345215, "learning_rate": 9.118970983135655e-06, "loss": 1.3578, "step": 9873 }, { "epoch": 0.7738244514106583, "grad_norm": 1.9994945526123047, "learning_rate": 9.118796168248018e-06, "loss": 1.3236, "step": 9874 }, { "epoch": 0.7739028213166145, "grad_norm": 2.0594425201416016, "learning_rate": 9.118621337694598e-06, "loss": 1.0933, "step": 9875 }, { "epoch": 0.7739811912225706, "grad_norm": 2.4654769897460938, "learning_rate": 9.118446491476055e-06, "loss": 1.3962, "step": 9876 }, { "epoch": 0.7740595611285267, "grad_norm": 2.569281578063965, "learning_rate": 9.118271629593058e-06, "loss": 1.4372, "step": 9877 }, { "epoch": 0.7741379310344828, "grad_norm": 5.517215251922607, "learning_rate": 9.11809675204627e-06, "loss": 1.6969, "step": 9878 }, { "epoch": 0.7742163009404389, "grad_norm": 2.4053151607513428, "learning_rate": 9.117921858836358e-06, "loss": 1.3665, "step": 9879 }, { "epoch": 0.774294670846395, "grad_norm": 2.9686295986175537, "learning_rate": 9.117746949963986e-06, "loss": 1.3248, "step": 9880 }, { "epoch": 0.7743730407523511, "grad_norm": 2.926654100418091, "learning_rate": 9.11757202542982e-06, "loss": 1.3371, "step": 9881 }, { "epoch": 0.7744514106583072, "grad_norm": 4.449392795562744, "learning_rate": 9.117397085234525e-06, "loss": 1.3928, "step": 9882 }, { "epoch": 0.7745297805642634, "grad_norm": 2.9049344062805176, "learning_rate": 9.117222129378767e-06, "loss": 1.6172, "step": 9883 }, { "epoch": 0.7746081504702195, "grad_norm": 4.0739874839782715, "learning_rate": 9.117047157863209e-06, "loss": 1.4473, "step": 9884 }, { "epoch": 0.7746865203761756, "grad_norm": 7.577004909515381, "learning_rate": 9.11687217068852e-06, "loss": 1.9433, "step": 9885 }, { "epoch": 0.7747648902821317, "grad_norm": 9.24073600769043, "learning_rate": 9.116697167855363e-06, "loss": 1.5518, "step": 9886 }, { "epoch": 0.7748432601880878, "grad_norm": 6.745248794555664, "learning_rate": 9.116522149364404e-06, "loss": 1.5539, "step": 9887 }, { "epoch": 0.7749216300940439, "grad_norm": 2.306817054748535, "learning_rate": 9.116347115216308e-06, "loss": 1.3522, "step": 9888 }, { "epoch": 0.775, "grad_norm": 2.642367124557495, "learning_rate": 9.116172065411744e-06, "loss": 1.426, "step": 9889 }, { "epoch": 0.7750783699059561, "grad_norm": 2.785879135131836, "learning_rate": 9.115996999951376e-06, "loss": 1.4401, "step": 9890 }, { "epoch": 0.7751567398119122, "grad_norm": 3.136054277420044, "learning_rate": 9.115821918835868e-06, "loss": 1.2727, "step": 9891 }, { "epoch": 0.7752351097178684, "grad_norm": 14.344717979431152, "learning_rate": 9.11564682206589e-06, "loss": 2.4566, "step": 9892 }, { "epoch": 0.7753134796238245, "grad_norm": 2.285470962524414, "learning_rate": 9.115471709642103e-06, "loss": 1.2065, "step": 9893 }, { "epoch": 0.7753918495297806, "grad_norm": 5.813196182250977, "learning_rate": 9.115296581565174e-06, "loss": 1.5299, "step": 9894 }, { "epoch": 0.7754702194357367, "grad_norm": 2.0225932598114014, "learning_rate": 9.115121437835772e-06, "loss": 1.5278, "step": 9895 }, { "epoch": 0.7755485893416928, "grad_norm": 4.260652542114258, "learning_rate": 9.114946278454564e-06, "loss": 1.0041, "step": 9896 }, { "epoch": 0.7756269592476489, "grad_norm": 2.3726346492767334, "learning_rate": 9.114771103422211e-06, "loss": 1.1955, "step": 9897 }, { "epoch": 0.775705329153605, "grad_norm": 4.826798439025879, "learning_rate": 9.114595912739384e-06, "loss": 1.4619, "step": 9898 }, { "epoch": 0.7757836990595611, "grad_norm": 2.762636184692383, "learning_rate": 9.114420706406745e-06, "loss": 1.1424, "step": 9899 }, { "epoch": 0.7758620689655172, "grad_norm": 2.7060585021972656, "learning_rate": 9.114245484424966e-06, "loss": 1.0748, "step": 9900 }, { "epoch": 0.7759404388714733, "grad_norm": 3.296356201171875, "learning_rate": 9.114070246794709e-06, "loss": 1.3969, "step": 9901 }, { "epoch": 0.7760188087774295, "grad_norm": 4.0828752517700195, "learning_rate": 9.113894993516643e-06, "loss": 1.4859, "step": 9902 }, { "epoch": 0.7760971786833856, "grad_norm": 3.304049015045166, "learning_rate": 9.113719724591432e-06, "loss": 1.3166, "step": 9903 }, { "epoch": 0.7761755485893417, "grad_norm": 4.8048095703125, "learning_rate": 9.113544440019746e-06, "loss": 1.3913, "step": 9904 }, { "epoch": 0.7762539184952978, "grad_norm": 2.9652421474456787, "learning_rate": 9.113369139802248e-06, "loss": 1.506, "step": 9905 }, { "epoch": 0.7763322884012539, "grad_norm": 6.356359481811523, "learning_rate": 9.113193823939606e-06, "loss": 1.307, "step": 9906 }, { "epoch": 0.77641065830721, "grad_norm": 2.6379752159118652, "learning_rate": 9.113018492432489e-06, "loss": 1.2498, "step": 9907 }, { "epoch": 0.7764890282131661, "grad_norm": 2.672243356704712, "learning_rate": 9.112843145281561e-06, "loss": 1.1724, "step": 9908 }, { "epoch": 0.7765673981191222, "grad_norm": 2.24137282371521, "learning_rate": 9.11266778248749e-06, "loss": 1.6167, "step": 9909 }, { "epoch": 0.7766457680250783, "grad_norm": 2.744135618209839, "learning_rate": 9.112492404050944e-06, "loss": 1.6355, "step": 9910 }, { "epoch": 0.7767241379310345, "grad_norm": 3.091215133666992, "learning_rate": 9.112317009972589e-06, "loss": 1.3487, "step": 9911 }, { "epoch": 0.7768025078369906, "grad_norm": 3.031010150909424, "learning_rate": 9.112141600253092e-06, "loss": 1.3173, "step": 9912 }, { "epoch": 0.7768808777429467, "grad_norm": 2.866981029510498, "learning_rate": 9.111966174893122e-06, "loss": 1.6889, "step": 9913 }, { "epoch": 0.7769592476489028, "grad_norm": 2.89190673828125, "learning_rate": 9.111790733893344e-06, "loss": 1.2369, "step": 9914 }, { "epoch": 0.7770376175548589, "grad_norm": 2.774796724319458, "learning_rate": 9.111615277254425e-06, "loss": 1.3987, "step": 9915 }, { "epoch": 0.777115987460815, "grad_norm": 2.339169979095459, "learning_rate": 9.111439804977034e-06, "loss": 1.5795, "step": 9916 }, { "epoch": 0.7771943573667711, "grad_norm": 3.445079803466797, "learning_rate": 9.111264317061839e-06, "loss": 1.7272, "step": 9917 }, { "epoch": 0.7772727272727272, "grad_norm": 2.7551846504211426, "learning_rate": 9.111088813509503e-06, "loss": 1.5555, "step": 9918 }, { "epoch": 0.7773510971786833, "grad_norm": 2.198875904083252, "learning_rate": 9.1109132943207e-06, "loss": 1.3158, "step": 9919 }, { "epoch": 0.7774294670846394, "grad_norm": 3.3922417163848877, "learning_rate": 9.110737759496092e-06, "loss": 1.3144, "step": 9920 }, { "epoch": 0.7775078369905956, "grad_norm": 3.5634639263153076, "learning_rate": 9.11056220903635e-06, "loss": 1.3439, "step": 9921 }, { "epoch": 0.7775862068965518, "grad_norm": 3.8203303813934326, "learning_rate": 9.110386642942141e-06, "loss": 1.4214, "step": 9922 }, { "epoch": 0.7776645768025079, "grad_norm": 2.6637914180755615, "learning_rate": 9.110211061214132e-06, "loss": 1.4665, "step": 9923 }, { "epoch": 0.777742946708464, "grad_norm": 1.8912558555603027, "learning_rate": 9.11003546385299e-06, "loss": 1.4762, "step": 9924 }, { "epoch": 0.7778213166144201, "grad_norm": 2.876349449157715, "learning_rate": 9.109859850859387e-06, "loss": 1.2247, "step": 9925 }, { "epoch": 0.7778996865203762, "grad_norm": 4.785096168518066, "learning_rate": 9.109684222233986e-06, "loss": 1.7781, "step": 9926 }, { "epoch": 0.7779780564263323, "grad_norm": 2.552128791809082, "learning_rate": 9.109508577977456e-06, "loss": 1.6058, "step": 9927 }, { "epoch": 0.7780564263322884, "grad_norm": 2.4092423915863037, "learning_rate": 9.10933291809047e-06, "loss": 1.5144, "step": 9928 }, { "epoch": 0.7781347962382446, "grad_norm": 2.790106773376465, "learning_rate": 9.10915724257369e-06, "loss": 1.4964, "step": 9929 }, { "epoch": 0.7782131661442007, "grad_norm": 2.827085018157959, "learning_rate": 9.108981551427786e-06, "loss": 1.3719, "step": 9930 }, { "epoch": 0.7782915360501568, "grad_norm": 2.4432711601257324, "learning_rate": 9.108805844653427e-06, "loss": 1.3772, "step": 9931 }, { "epoch": 0.7783699059561129, "grad_norm": 3.8470067977905273, "learning_rate": 9.108630122251283e-06, "loss": 1.2749, "step": 9932 }, { "epoch": 0.778448275862069, "grad_norm": 2.149707794189453, "learning_rate": 9.108454384222016e-06, "loss": 1.4441, "step": 9933 }, { "epoch": 0.7785266457680251, "grad_norm": 2.613089084625244, "learning_rate": 9.108278630566304e-06, "loss": 1.3205, "step": 9934 }, { "epoch": 0.7786050156739812, "grad_norm": 2.5541558265686035, "learning_rate": 9.108102861284806e-06, "loss": 1.2303, "step": 9935 }, { "epoch": 0.7786833855799373, "grad_norm": 4.348318576812744, "learning_rate": 9.107927076378196e-06, "loss": 1.5405, "step": 9936 }, { "epoch": 0.7787617554858934, "grad_norm": 2.1283957958221436, "learning_rate": 9.10775127584714e-06, "loss": 1.4532, "step": 9937 }, { "epoch": 0.7788401253918495, "grad_norm": 2.5789031982421875, "learning_rate": 9.107575459692311e-06, "loss": 1.6858, "step": 9938 }, { "epoch": 0.7789184952978057, "grad_norm": 3.9821650981903076, "learning_rate": 9.107399627914373e-06, "loss": 1.2649, "step": 9939 }, { "epoch": 0.7789968652037618, "grad_norm": 3.3648948669433594, "learning_rate": 9.107223780513997e-06, "loss": 1.4971, "step": 9940 }, { "epoch": 0.7790752351097179, "grad_norm": 3.61843204498291, "learning_rate": 9.107047917491851e-06, "loss": 1.2626, "step": 9941 }, { "epoch": 0.779153605015674, "grad_norm": 2.1621625423431396, "learning_rate": 9.106872038848605e-06, "loss": 1.654, "step": 9942 }, { "epoch": 0.7792319749216301, "grad_norm": 3.555835485458374, "learning_rate": 9.106696144584927e-06, "loss": 1.0876, "step": 9943 }, { "epoch": 0.7793103448275862, "grad_norm": 2.176344156265259, "learning_rate": 9.106520234701486e-06, "loss": 1.4756, "step": 9944 }, { "epoch": 0.7793887147335423, "grad_norm": 3.3295483589172363, "learning_rate": 9.106344309198952e-06, "loss": 1.1357, "step": 9945 }, { "epoch": 0.7794670846394984, "grad_norm": 2.6003549098968506, "learning_rate": 9.106168368077993e-06, "loss": 1.5219, "step": 9946 }, { "epoch": 0.7795454545454545, "grad_norm": 7.377007484436035, "learning_rate": 9.105992411339278e-06, "loss": 1.4881, "step": 9947 }, { "epoch": 0.7796238244514107, "grad_norm": 2.58774471282959, "learning_rate": 9.105816438983475e-06, "loss": 1.2977, "step": 9948 }, { "epoch": 0.7797021943573668, "grad_norm": 4.523264408111572, "learning_rate": 9.105640451011257e-06, "loss": 1.4935, "step": 9949 }, { "epoch": 0.7797805642633229, "grad_norm": 3.4968271255493164, "learning_rate": 9.105464447423292e-06, "loss": 1.2788, "step": 9950 }, { "epoch": 0.779858934169279, "grad_norm": 4.341132164001465, "learning_rate": 9.105288428220249e-06, "loss": 1.1246, "step": 9951 }, { "epoch": 0.7799373040752351, "grad_norm": 2.1658098697662354, "learning_rate": 9.105112393402797e-06, "loss": 1.4962, "step": 9952 }, { "epoch": 0.7800156739811912, "grad_norm": 2.327314615249634, "learning_rate": 9.104936342971606e-06, "loss": 1.4797, "step": 9953 }, { "epoch": 0.7800940438871473, "grad_norm": 3.718939781188965, "learning_rate": 9.104760276927344e-06, "loss": 1.0414, "step": 9954 }, { "epoch": 0.7801724137931034, "grad_norm": 6.312605857849121, "learning_rate": 9.104584195270685e-06, "loss": 0.9174, "step": 9955 }, { "epoch": 0.7802507836990595, "grad_norm": 6.676341533660889, "learning_rate": 9.104408098002295e-06, "loss": 1.2339, "step": 9956 }, { "epoch": 0.7803291536050156, "grad_norm": 2.901672124862671, "learning_rate": 9.104231985122845e-06, "loss": 1.1828, "step": 9957 }, { "epoch": 0.7804075235109718, "grad_norm": 4.778061866760254, "learning_rate": 9.104055856633004e-06, "loss": 1.3294, "step": 9958 }, { "epoch": 0.7804858934169279, "grad_norm": 8.974276542663574, "learning_rate": 9.103879712533441e-06, "loss": 1.7341, "step": 9959 }, { "epoch": 0.780564263322884, "grad_norm": 3.722559690475464, "learning_rate": 9.10370355282483e-06, "loss": 1.5107, "step": 9960 }, { "epoch": 0.7806426332288401, "grad_norm": 11.442617416381836, "learning_rate": 9.103527377507837e-06, "loss": 1.0244, "step": 9961 }, { "epoch": 0.7807210031347962, "grad_norm": 2.686312675476074, "learning_rate": 9.103351186583134e-06, "loss": 1.2561, "step": 9962 }, { "epoch": 0.7807993730407523, "grad_norm": 2.2775261402130127, "learning_rate": 9.103174980051391e-06, "loss": 1.0162, "step": 9963 }, { "epoch": 0.7808777429467084, "grad_norm": 2.478422164916992, "learning_rate": 9.102998757913279e-06, "loss": 1.3407, "step": 9964 }, { "epoch": 0.7809561128526645, "grad_norm": 2.552175998687744, "learning_rate": 9.102822520169466e-06, "loss": 1.3195, "step": 9965 }, { "epoch": 0.7810344827586206, "grad_norm": 3.2815680503845215, "learning_rate": 9.102646266820622e-06, "loss": 1.7991, "step": 9966 }, { "epoch": 0.7811128526645768, "grad_norm": 2.2915189266204834, "learning_rate": 9.102469997867421e-06, "loss": 1.359, "step": 9967 }, { "epoch": 0.7811912225705329, "grad_norm": 3.1532604694366455, "learning_rate": 9.102293713310531e-06, "loss": 1.7853, "step": 9968 }, { "epoch": 0.7812695924764891, "grad_norm": 5.024583339691162, "learning_rate": 9.102117413150622e-06, "loss": 1.4365, "step": 9969 }, { "epoch": 0.7813479623824452, "grad_norm": 2.9328060150146484, "learning_rate": 9.101941097388364e-06, "loss": 1.65, "step": 9970 }, { "epoch": 0.7814263322884013, "grad_norm": 2.5919148921966553, "learning_rate": 9.101764766024432e-06, "loss": 1.3039, "step": 9971 }, { "epoch": 0.7815047021943574, "grad_norm": 2.3081183433532715, "learning_rate": 9.101588419059493e-06, "loss": 1.4549, "step": 9972 }, { "epoch": 0.7815830721003135, "grad_norm": 3.132497787475586, "learning_rate": 9.101412056494216e-06, "loss": 1.5524, "step": 9973 }, { "epoch": 0.7816614420062696, "grad_norm": 4.552556037902832, "learning_rate": 9.101235678329276e-06, "loss": 1.5326, "step": 9974 }, { "epoch": 0.7817398119122257, "grad_norm": 7.615969181060791, "learning_rate": 9.101059284565341e-06, "loss": 0.7621, "step": 9975 }, { "epoch": 0.7818181818181819, "grad_norm": 3.805755615234375, "learning_rate": 9.100882875203085e-06, "loss": 1.7341, "step": 9976 }, { "epoch": 0.781896551724138, "grad_norm": 2.495136260986328, "learning_rate": 9.100706450243176e-06, "loss": 1.6221, "step": 9977 }, { "epoch": 0.7819749216300941, "grad_norm": 2.221602439880371, "learning_rate": 9.100530009686285e-06, "loss": 1.1336, "step": 9978 }, { "epoch": 0.7820532915360502, "grad_norm": 6.583201885223389, "learning_rate": 9.100353553533083e-06, "loss": 1.2596, "step": 9979 }, { "epoch": 0.7821316614420063, "grad_norm": 12.401458740234375, "learning_rate": 9.100177081784245e-06, "loss": 2.4607, "step": 9980 }, { "epoch": 0.7822100313479624, "grad_norm": 2.5375263690948486, "learning_rate": 9.100000594440437e-06, "loss": 1.2428, "step": 9981 }, { "epoch": 0.7822884012539185, "grad_norm": 4.474145889282227, "learning_rate": 9.099824091502335e-06, "loss": 1.0895, "step": 9982 }, { "epoch": 0.7823667711598746, "grad_norm": 4.182112693786621, "learning_rate": 9.099647572970606e-06, "loss": 2.0111, "step": 9983 }, { "epoch": 0.7824451410658307, "grad_norm": 4.426580429077148, "learning_rate": 9.099471038845924e-06, "loss": 1.2412, "step": 9984 }, { "epoch": 0.7825235109717869, "grad_norm": 1.9018000364303589, "learning_rate": 9.09929448912896e-06, "loss": 1.602, "step": 9985 }, { "epoch": 0.782601880877743, "grad_norm": 2.265092611312866, "learning_rate": 9.099117923820384e-06, "loss": 1.5237, "step": 9986 }, { "epoch": 0.7826802507836991, "grad_norm": 1.8306934833526611, "learning_rate": 9.098941342920871e-06, "loss": 1.1546, "step": 9987 }, { "epoch": 0.7827586206896552, "grad_norm": 2.1161653995513916, "learning_rate": 9.098764746431088e-06, "loss": 1.4869, "step": 9988 }, { "epoch": 0.7828369905956113, "grad_norm": 3.1507205963134766, "learning_rate": 9.098588134351711e-06, "loss": 1.1544, "step": 9989 }, { "epoch": 0.7829153605015674, "grad_norm": 2.1354713439941406, "learning_rate": 9.098411506683409e-06, "loss": 1.5901, "step": 9990 }, { "epoch": 0.7829937304075235, "grad_norm": 2.990342617034912, "learning_rate": 9.098234863426855e-06, "loss": 1.2835, "step": 9991 }, { "epoch": 0.7830721003134796, "grad_norm": 2.5877742767333984, "learning_rate": 9.098058204582721e-06, "loss": 1.5422, "step": 9992 }, { "epoch": 0.7831504702194357, "grad_norm": 5.074970245361328, "learning_rate": 9.097881530151677e-06, "loss": 1.7539, "step": 9993 }, { "epoch": 0.7832288401253918, "grad_norm": 2.480634927749634, "learning_rate": 9.097704840134398e-06, "loss": 1.5052, "step": 9994 }, { "epoch": 0.783307210031348, "grad_norm": 2.2797977924346924, "learning_rate": 9.097528134531555e-06, "loss": 1.5341, "step": 9995 }, { "epoch": 0.7833855799373041, "grad_norm": 3.1028268337249756, "learning_rate": 9.097351413343818e-06, "loss": 1.3449, "step": 9996 }, { "epoch": 0.7834639498432602, "grad_norm": 2.07250714302063, "learning_rate": 9.097174676571863e-06, "loss": 1.2422, "step": 9997 }, { "epoch": 0.7835423197492163, "grad_norm": 5.937826156616211, "learning_rate": 9.096997924216358e-06, "loss": 1.5406, "step": 9998 }, { "epoch": 0.7836206896551724, "grad_norm": 2.976595401763916, "learning_rate": 9.096821156277977e-06, "loss": 1.2928, "step": 9999 }, { "epoch": 0.7836990595611285, "grad_norm": 3.0260910987854004, "learning_rate": 9.096644372757393e-06, "loss": 1.201, "step": 10000 }, { "epoch": 0.7837774294670846, "grad_norm": 2.8612234592437744, "learning_rate": 9.09646757365528e-06, "loss": 1.239, "step": 10001 }, { "epoch": 0.7838557993730407, "grad_norm": 3.382002115249634, "learning_rate": 9.096290758972306e-06, "loss": 1.0766, "step": 10002 }, { "epoch": 0.7839341692789968, "grad_norm": 3.8373963832855225, "learning_rate": 9.096113928709146e-06, "loss": 1.8305, "step": 10003 }, { "epoch": 0.784012539184953, "grad_norm": 2.658583164215088, "learning_rate": 9.095937082866473e-06, "loss": 1.4571, "step": 10004 }, { "epoch": 0.7840909090909091, "grad_norm": 1.9268039464950562, "learning_rate": 9.09576022144496e-06, "loss": 1.4701, "step": 10005 }, { "epoch": 0.7841692789968652, "grad_norm": 3.266495943069458, "learning_rate": 9.095583344445277e-06, "loss": 1.4944, "step": 10006 }, { "epoch": 0.7842476489028213, "grad_norm": 3.5928778648376465, "learning_rate": 9.0954064518681e-06, "loss": 1.2786, "step": 10007 }, { "epoch": 0.7843260188087774, "grad_norm": 3.834737777709961, "learning_rate": 9.0952295437141e-06, "loss": 1.5101, "step": 10008 }, { "epoch": 0.7844043887147335, "grad_norm": 3.5430264472961426, "learning_rate": 9.09505261998395e-06, "loss": 1.1912, "step": 10009 }, { "epoch": 0.7844827586206896, "grad_norm": 3.0504164695739746, "learning_rate": 9.094875680678324e-06, "loss": 1.0854, "step": 10010 }, { "epoch": 0.7845611285266457, "grad_norm": 2.3548476696014404, "learning_rate": 9.094698725797895e-06, "loss": 1.6709, "step": 10011 }, { "epoch": 0.7846394984326018, "grad_norm": 2.804151773452759, "learning_rate": 9.094521755343333e-06, "loss": 1.2973, "step": 10012 }, { "epoch": 0.784717868338558, "grad_norm": 3.3244457244873047, "learning_rate": 9.094344769315315e-06, "loss": 1.5938, "step": 10013 }, { "epoch": 0.784796238244514, "grad_norm": 2.614809513092041, "learning_rate": 9.094167767714512e-06, "loss": 1.5773, "step": 10014 }, { "epoch": 0.7848746081504702, "grad_norm": 3.226881504058838, "learning_rate": 9.093990750541597e-06, "loss": 1.4461, "step": 10015 }, { "epoch": 0.7849529780564264, "grad_norm": 5.567986488342285, "learning_rate": 9.093813717797244e-06, "loss": 1.5524, "step": 10016 }, { "epoch": 0.7850313479623825, "grad_norm": 3.925238609313965, "learning_rate": 9.093636669482126e-06, "loss": 1.3669, "step": 10017 }, { "epoch": 0.7851097178683386, "grad_norm": 2.3512017726898193, "learning_rate": 9.093459605596919e-06, "loss": 1.3256, "step": 10018 }, { "epoch": 0.7851880877742947, "grad_norm": 2.822148084640503, "learning_rate": 9.093282526142292e-06, "loss": 1.3288, "step": 10019 }, { "epoch": 0.7852664576802508, "grad_norm": 3.2369070053100586, "learning_rate": 9.093105431118921e-06, "loss": 1.7148, "step": 10020 }, { "epoch": 0.7853448275862069, "grad_norm": 2.217094659805298, "learning_rate": 9.09292832052748e-06, "loss": 1.3941, "step": 10021 }, { "epoch": 0.785423197492163, "grad_norm": 2.9204392433166504, "learning_rate": 9.092751194368642e-06, "loss": 1.4273, "step": 10022 }, { "epoch": 0.7855015673981192, "grad_norm": 3.030435562133789, "learning_rate": 9.09257405264308e-06, "loss": 1.3483, "step": 10023 }, { "epoch": 0.7855799373040753, "grad_norm": 2.3547351360321045, "learning_rate": 9.092396895351468e-06, "loss": 1.2443, "step": 10024 }, { "epoch": 0.7856583072100314, "grad_norm": 3.286247491836548, "learning_rate": 9.092219722494482e-06, "loss": 1.3691, "step": 10025 }, { "epoch": 0.7857366771159875, "grad_norm": 1.9807372093200684, "learning_rate": 9.092042534072793e-06, "loss": 1.3008, "step": 10026 }, { "epoch": 0.7858150470219436, "grad_norm": 2.019359588623047, "learning_rate": 9.091865330087075e-06, "loss": 1.4664, "step": 10027 }, { "epoch": 0.7858934169278997, "grad_norm": 2.9748716354370117, "learning_rate": 9.091688110538003e-06, "loss": 1.5586, "step": 10028 }, { "epoch": 0.7859717868338558, "grad_norm": 2.8252692222595215, "learning_rate": 9.091510875426252e-06, "loss": 1.2972, "step": 10029 }, { "epoch": 0.7860501567398119, "grad_norm": 4.060936450958252, "learning_rate": 9.091333624752497e-06, "loss": 1.4147, "step": 10030 }, { "epoch": 0.786128526645768, "grad_norm": 2.566636085510254, "learning_rate": 9.091156358517408e-06, "loss": 1.3736, "step": 10031 }, { "epoch": 0.7862068965517242, "grad_norm": 2.7916650772094727, "learning_rate": 9.090979076721662e-06, "loss": 1.5043, "step": 10032 }, { "epoch": 0.7862852664576803, "grad_norm": 2.751204490661621, "learning_rate": 9.090801779365932e-06, "loss": 1.2984, "step": 10033 }, { "epoch": 0.7863636363636364, "grad_norm": 2.1396989822387695, "learning_rate": 9.090624466450895e-06, "loss": 0.9828, "step": 10034 }, { "epoch": 0.7864420062695925, "grad_norm": 4.123531341552734, "learning_rate": 9.090447137977222e-06, "loss": 0.7756, "step": 10035 }, { "epoch": 0.7865203761755486, "grad_norm": 1.6395624876022339, "learning_rate": 9.09026979394559e-06, "loss": 1.2287, "step": 10036 }, { "epoch": 0.7865987460815047, "grad_norm": 2.7384915351867676, "learning_rate": 9.090092434356672e-06, "loss": 1.2917, "step": 10037 }, { "epoch": 0.7866771159874608, "grad_norm": 2.412588357925415, "learning_rate": 9.089915059211144e-06, "loss": 0.941, "step": 10038 }, { "epoch": 0.7867554858934169, "grad_norm": 2.8123626708984375, "learning_rate": 9.08973766850968e-06, "loss": 0.9515, "step": 10039 }, { "epoch": 0.786833855799373, "grad_norm": 3.2760138511657715, "learning_rate": 9.089560262252953e-06, "loss": 1.3468, "step": 10040 }, { "epoch": 0.7869122257053291, "grad_norm": 2.738508462905884, "learning_rate": 9.08938284044164e-06, "loss": 1.388, "step": 10041 }, { "epoch": 0.7869905956112853, "grad_norm": 3.9317452907562256, "learning_rate": 9.089205403076415e-06, "loss": 1.3586, "step": 10042 }, { "epoch": 0.7870689655172414, "grad_norm": 2.954756736755371, "learning_rate": 9.089027950157954e-06, "loss": 1.9704, "step": 10043 }, { "epoch": 0.7871473354231975, "grad_norm": 3.313169479370117, "learning_rate": 9.08885048168693e-06, "loss": 1.3906, "step": 10044 }, { "epoch": 0.7872257053291536, "grad_norm": 2.8872852325439453, "learning_rate": 9.08867299766402e-06, "loss": 1.2698, "step": 10045 }, { "epoch": 0.7873040752351097, "grad_norm": 2.2082788944244385, "learning_rate": 9.088495498089897e-06, "loss": 1.2764, "step": 10046 }, { "epoch": 0.7873824451410658, "grad_norm": 2.6918833255767822, "learning_rate": 9.088317982965237e-06, "loss": 1.2094, "step": 10047 }, { "epoch": 0.7874608150470219, "grad_norm": 3.0621657371520996, "learning_rate": 9.088140452290714e-06, "loss": 1.4136, "step": 10048 }, { "epoch": 0.787539184952978, "grad_norm": 4.6035566329956055, "learning_rate": 9.087962906067008e-06, "loss": 1.3509, "step": 10049 }, { "epoch": 0.7876175548589341, "grad_norm": 3.05625581741333, "learning_rate": 9.087785344294788e-06, "loss": 1.1793, "step": 10050 }, { "epoch": 0.7876959247648903, "grad_norm": 3.2163991928100586, "learning_rate": 9.087607766974734e-06, "loss": 1.8258, "step": 10051 }, { "epoch": 0.7877742946708464, "grad_norm": 5.35149621963501, "learning_rate": 9.087430174107518e-06, "loss": 1.8313, "step": 10052 }, { "epoch": 0.7878526645768025, "grad_norm": 15.244224548339844, "learning_rate": 9.087252565693817e-06, "loss": 1.2285, "step": 10053 }, { "epoch": 0.7879310344827586, "grad_norm": 4.141015529632568, "learning_rate": 9.08707494173431e-06, "loss": 1.6243, "step": 10054 }, { "epoch": 0.7880094043887147, "grad_norm": 2.315674304962158, "learning_rate": 9.086897302229664e-06, "loss": 1.315, "step": 10055 }, { "epoch": 0.7880877742946708, "grad_norm": 2.3914144039154053, "learning_rate": 9.086719647180562e-06, "loss": 1.4253, "step": 10056 }, { "epoch": 0.7881661442006269, "grad_norm": 2.69555926322937, "learning_rate": 9.086541976587679e-06, "loss": 1.1706, "step": 10057 }, { "epoch": 0.788244514106583, "grad_norm": 4.233980178833008, "learning_rate": 9.086364290451689e-06, "loss": 1.7292, "step": 10058 }, { "epoch": 0.7883228840125391, "grad_norm": 5.7460246086120605, "learning_rate": 9.086186588773267e-06, "loss": 1.5686, "step": 10059 }, { "epoch": 0.7884012539184952, "grad_norm": 2.6602578163146973, "learning_rate": 9.086008871553088e-06, "loss": 1.4563, "step": 10060 }, { "epoch": 0.7884796238244514, "grad_norm": 5.820210933685303, "learning_rate": 9.085831138791834e-06, "loss": 1.4461, "step": 10061 }, { "epoch": 0.7885579937304075, "grad_norm": 5.943119525909424, "learning_rate": 9.085653390490174e-06, "loss": 1.5637, "step": 10062 }, { "epoch": 0.7886363636363637, "grad_norm": 2.434054136276245, "learning_rate": 9.085475626648789e-06, "loss": 1.1724, "step": 10063 }, { "epoch": 0.7887147335423198, "grad_norm": 2.4446940422058105, "learning_rate": 9.085297847268352e-06, "loss": 1.3399, "step": 10064 }, { "epoch": 0.7887931034482759, "grad_norm": 2.6622438430786133, "learning_rate": 9.085120052349542e-06, "loss": 1.2933, "step": 10065 }, { "epoch": 0.788871473354232, "grad_norm": 10.26826000213623, "learning_rate": 9.084942241893031e-06, "loss": 1.0474, "step": 10066 }, { "epoch": 0.7889498432601881, "grad_norm": 3.436988353729248, "learning_rate": 9.084764415899499e-06, "loss": 1.4123, "step": 10067 }, { "epoch": 0.7890282131661442, "grad_norm": 2.9631409645080566, "learning_rate": 9.084586574369622e-06, "loss": 1.7031, "step": 10068 }, { "epoch": 0.7891065830721004, "grad_norm": 4.629853248596191, "learning_rate": 9.084408717304074e-06, "loss": 0.7009, "step": 10069 }, { "epoch": 0.7891849529780565, "grad_norm": 2.7907490730285645, "learning_rate": 9.084230844703534e-06, "loss": 1.3031, "step": 10070 }, { "epoch": 0.7892633228840126, "grad_norm": 4.635682106018066, "learning_rate": 9.084052956568678e-06, "loss": 1.335, "step": 10071 }, { "epoch": 0.7893416927899687, "grad_norm": 6.646115779876709, "learning_rate": 9.083875052900184e-06, "loss": 2.0636, "step": 10072 }, { "epoch": 0.7894200626959248, "grad_norm": 3.778568744659424, "learning_rate": 9.083697133698724e-06, "loss": 1.1486, "step": 10073 }, { "epoch": 0.7894984326018809, "grad_norm": 2.75044584274292, "learning_rate": 9.083519198964979e-06, "loss": 1.4188, "step": 10074 }, { "epoch": 0.789576802507837, "grad_norm": 2.6207945346832275, "learning_rate": 9.083341248699623e-06, "loss": 1.5285, "step": 10075 }, { "epoch": 0.7896551724137931, "grad_norm": 2.34260630607605, "learning_rate": 9.083163282903336e-06, "loss": 1.6518, "step": 10076 }, { "epoch": 0.7897335423197492, "grad_norm": 2.4654557704925537, "learning_rate": 9.082985301576792e-06, "loss": 1.0482, "step": 10077 }, { "epoch": 0.7898119122257053, "grad_norm": 2.2748255729675293, "learning_rate": 9.082807304720668e-06, "loss": 1.4205, "step": 10078 }, { "epoch": 0.7898902821316615, "grad_norm": 2.3194172382354736, "learning_rate": 9.082629292335645e-06, "loss": 1.2025, "step": 10079 }, { "epoch": 0.7899686520376176, "grad_norm": 2.3444430828094482, "learning_rate": 9.082451264422394e-06, "loss": 1.5222, "step": 10080 }, { "epoch": 0.7900470219435737, "grad_norm": 1.618489384651184, "learning_rate": 9.0822732209816e-06, "loss": 1.2389, "step": 10081 }, { "epoch": 0.7901253918495298, "grad_norm": 2.851309299468994, "learning_rate": 9.08209516201393e-06, "loss": 1.3394, "step": 10082 }, { "epoch": 0.7902037617554859, "grad_norm": 3.140143871307373, "learning_rate": 9.081917087520069e-06, "loss": 2.1789, "step": 10083 }, { "epoch": 0.790282131661442, "grad_norm": 2.0999717712402344, "learning_rate": 9.081738997500694e-06, "loss": 1.1954, "step": 10084 }, { "epoch": 0.7903605015673981, "grad_norm": 3.855546474456787, "learning_rate": 9.081560891956477e-06, "loss": 1.2287, "step": 10085 }, { "epoch": 0.7904388714733542, "grad_norm": 2.717150926589966, "learning_rate": 9.0813827708881e-06, "loss": 1.4656, "step": 10086 }, { "epoch": 0.7905172413793103, "grad_norm": 11.185187339782715, "learning_rate": 9.08120463429624e-06, "loss": 0.7407, "step": 10087 }, { "epoch": 0.7905956112852665, "grad_norm": 19.45210838317871, "learning_rate": 9.081026482181573e-06, "loss": 1.2015, "step": 10088 }, { "epoch": 0.7906739811912226, "grad_norm": 2.062729597091675, "learning_rate": 9.080848314544776e-06, "loss": 1.4171, "step": 10089 }, { "epoch": 0.7907523510971787, "grad_norm": 2.947598457336426, "learning_rate": 9.08067013138653e-06, "loss": 1.2986, "step": 10090 }, { "epoch": 0.7908307210031348, "grad_norm": 2.0355305671691895, "learning_rate": 9.080491932707512e-06, "loss": 1.4737, "step": 10091 }, { "epoch": 0.7909090909090909, "grad_norm": 3.162047863006592, "learning_rate": 9.080313718508399e-06, "loss": 1.669, "step": 10092 }, { "epoch": 0.790987460815047, "grad_norm": 3.3660471439361572, "learning_rate": 9.080135488789865e-06, "loss": 1.4671, "step": 10093 }, { "epoch": 0.7910658307210031, "grad_norm": 3.1000781059265137, "learning_rate": 9.079957243552593e-06, "loss": 1.4041, "step": 10094 }, { "epoch": 0.7911442006269592, "grad_norm": 2.1252920627593994, "learning_rate": 9.079778982797259e-06, "loss": 1.3124, "step": 10095 }, { "epoch": 0.7912225705329153, "grad_norm": 6.543288230895996, "learning_rate": 9.07960070652454e-06, "loss": 1.4812, "step": 10096 }, { "epoch": 0.7913009404388714, "grad_norm": 2.8230977058410645, "learning_rate": 9.079422414735117e-06, "loss": 1.3392, "step": 10097 }, { "epoch": 0.7913793103448276, "grad_norm": 2.6522979736328125, "learning_rate": 9.079244107429666e-06, "loss": 1.1096, "step": 10098 }, { "epoch": 0.7914576802507837, "grad_norm": 4.772490978240967, "learning_rate": 9.079065784608865e-06, "loss": 0.9773, "step": 10099 }, { "epoch": 0.7915360501567398, "grad_norm": 3.898777723312378, "learning_rate": 9.078887446273394e-06, "loss": 1.1744, "step": 10100 }, { "epoch": 0.7916144200626959, "grad_norm": 2.712195634841919, "learning_rate": 9.07870909242393e-06, "loss": 1.2246, "step": 10101 }, { "epoch": 0.791692789968652, "grad_norm": 2.4231362342834473, "learning_rate": 9.078530723061153e-06, "loss": 1.4361, "step": 10102 }, { "epoch": 0.7917711598746081, "grad_norm": 2.915219783782959, "learning_rate": 9.078352338185736e-06, "loss": 1.5694, "step": 10103 }, { "epoch": 0.7918495297805642, "grad_norm": 5.108213424682617, "learning_rate": 9.078173937798365e-06, "loss": 1.2091, "step": 10104 }, { "epoch": 0.7919278996865203, "grad_norm": 5.036059856414795, "learning_rate": 9.077995521899714e-06, "loss": 1.6573, "step": 10105 }, { "epoch": 0.7920062695924764, "grad_norm": 2.725290060043335, "learning_rate": 9.077817090490462e-06, "loss": 1.3292, "step": 10106 }, { "epoch": 0.7920846394984326, "grad_norm": 14.433834075927734, "learning_rate": 9.07763864357129e-06, "loss": 0.6725, "step": 10107 }, { "epoch": 0.7921630094043887, "grad_norm": 2.6831912994384766, "learning_rate": 9.077460181142873e-06, "loss": 1.4119, "step": 10108 }, { "epoch": 0.7922413793103448, "grad_norm": 2.2381067276000977, "learning_rate": 9.077281703205894e-06, "loss": 1.1798, "step": 10109 }, { "epoch": 0.792319749216301, "grad_norm": 3.6289098262786865, "learning_rate": 9.077103209761026e-06, "loss": 1.6527, "step": 10110 }, { "epoch": 0.7923981191222571, "grad_norm": 11.860089302062988, "learning_rate": 9.076924700808955e-06, "loss": 2.3439, "step": 10111 }, { "epoch": 0.7924764890282132, "grad_norm": 2.181079626083374, "learning_rate": 9.076746176350355e-06, "loss": 1.4852, "step": 10112 }, { "epoch": 0.7925548589341693, "grad_norm": 2.2147531509399414, "learning_rate": 9.076567636385906e-06, "loss": 1.2773, "step": 10113 }, { "epoch": 0.7926332288401254, "grad_norm": 16.389583587646484, "learning_rate": 9.076389080916288e-06, "loss": 1.5188, "step": 10114 }, { "epoch": 0.7927115987460815, "grad_norm": 2.688854217529297, "learning_rate": 9.07621050994218e-06, "loss": 1.4266, "step": 10115 }, { "epoch": 0.7927899686520377, "grad_norm": 2.19014835357666, "learning_rate": 9.076031923464261e-06, "loss": 1.1907, "step": 10116 }, { "epoch": 0.7928683385579938, "grad_norm": 2.790095329284668, "learning_rate": 9.07585332148321e-06, "loss": 0.7502, "step": 10117 }, { "epoch": 0.7929467084639499, "grad_norm": 2.702087163925171, "learning_rate": 9.075674703999707e-06, "loss": 1.4134, "step": 10118 }, { "epoch": 0.793025078369906, "grad_norm": 1.928894281387329, "learning_rate": 9.075496071014431e-06, "loss": 1.5575, "step": 10119 }, { "epoch": 0.7931034482758621, "grad_norm": 1.8470877408981323, "learning_rate": 9.07531742252806e-06, "loss": 1.2955, "step": 10120 }, { "epoch": 0.7931818181818182, "grad_norm": 4.917454719543457, "learning_rate": 9.075138758541275e-06, "loss": 1.3864, "step": 10121 }, { "epoch": 0.7932601880877743, "grad_norm": 2.637985944747925, "learning_rate": 9.074960079054755e-06, "loss": 1.406, "step": 10122 }, { "epoch": 0.7933385579937304, "grad_norm": 8.147107124328613, "learning_rate": 9.074781384069182e-06, "loss": 1.1709, "step": 10123 }, { "epoch": 0.7934169278996865, "grad_norm": 6.7854132652282715, "learning_rate": 9.074602673585231e-06, "loss": 0.7116, "step": 10124 }, { "epoch": 0.7934952978056427, "grad_norm": 2.3943710327148438, "learning_rate": 9.074423947603585e-06, "loss": 1.4834, "step": 10125 }, { "epoch": 0.7935736677115988, "grad_norm": 1.994634747505188, "learning_rate": 9.074245206124924e-06, "loss": 1.1374, "step": 10126 }, { "epoch": 0.7936520376175549, "grad_norm": 2.4284849166870117, "learning_rate": 9.074066449149927e-06, "loss": 1.4682, "step": 10127 }, { "epoch": 0.793730407523511, "grad_norm": 2.879398822784424, "learning_rate": 9.073887676679272e-06, "loss": 1.1564, "step": 10128 }, { "epoch": 0.7938087774294671, "grad_norm": 2.6621975898742676, "learning_rate": 9.073708888713642e-06, "loss": 1.3219, "step": 10129 }, { "epoch": 0.7938871473354232, "grad_norm": 2.683288335800171, "learning_rate": 9.073530085253716e-06, "loss": 1.473, "step": 10130 }, { "epoch": 0.7939655172413793, "grad_norm": 6.457763671875, "learning_rate": 9.073351266300174e-06, "loss": 1.62, "step": 10131 }, { "epoch": 0.7940438871473354, "grad_norm": 1.8020607233047485, "learning_rate": 9.073172431853693e-06, "loss": 1.4797, "step": 10132 }, { "epoch": 0.7941222570532915, "grad_norm": 1.9589204788208008, "learning_rate": 9.07299358191496e-06, "loss": 1.2773, "step": 10133 }, { "epoch": 0.7942006269592476, "grad_norm": 2.312981367111206, "learning_rate": 9.07281471648465e-06, "loss": 0.7976, "step": 10134 }, { "epoch": 0.7942789968652038, "grad_norm": 2.703408718109131, "learning_rate": 9.072635835563446e-06, "loss": 2.2029, "step": 10135 }, { "epoch": 0.7943573667711599, "grad_norm": 2.876505136489868, "learning_rate": 9.072456939152025e-06, "loss": 1.4278, "step": 10136 }, { "epoch": 0.794435736677116, "grad_norm": 8.20936107635498, "learning_rate": 9.072278027251072e-06, "loss": 1.429, "step": 10137 }, { "epoch": 0.7945141065830721, "grad_norm": 4.748106479644775, "learning_rate": 9.072099099861262e-06, "loss": 1.7138, "step": 10138 }, { "epoch": 0.7945924764890282, "grad_norm": 1.9681872129440308, "learning_rate": 9.071920156983281e-06, "loss": 1.5287, "step": 10139 }, { "epoch": 0.7946708463949843, "grad_norm": 2.1021924018859863, "learning_rate": 9.071741198617805e-06, "loss": 1.2305, "step": 10140 }, { "epoch": 0.7947492163009404, "grad_norm": 3.122781276702881, "learning_rate": 9.07156222476552e-06, "loss": 1.4725, "step": 10141 }, { "epoch": 0.7948275862068965, "grad_norm": 3.231226682662964, "learning_rate": 9.0713832354271e-06, "loss": 1.4276, "step": 10142 }, { "epoch": 0.7949059561128526, "grad_norm": 3.222094774246216, "learning_rate": 9.071204230603232e-06, "loss": 1.5178, "step": 10143 }, { "epoch": 0.7949843260188088, "grad_norm": 3.724289655685425, "learning_rate": 9.071025210294593e-06, "loss": 0.9982, "step": 10144 }, { "epoch": 0.7950626959247649, "grad_norm": 2.3957314491271973, "learning_rate": 9.070846174501866e-06, "loss": 1.3944, "step": 10145 }, { "epoch": 0.795141065830721, "grad_norm": 2.5830469131469727, "learning_rate": 9.070667123225729e-06, "loss": 1.5484, "step": 10146 }, { "epoch": 0.7952194357366771, "grad_norm": 4.405936241149902, "learning_rate": 9.070488056466868e-06, "loss": 1.5737, "step": 10147 }, { "epoch": 0.7952978056426332, "grad_norm": 3.5761983394622803, "learning_rate": 9.070308974225958e-06, "loss": 1.46, "step": 10148 }, { "epoch": 0.7953761755485893, "grad_norm": 3.5536134243011475, "learning_rate": 9.070129876503686e-06, "loss": 1.2535, "step": 10149 }, { "epoch": 0.7954545454545454, "grad_norm": 2.995511770248413, "learning_rate": 9.06995076330073e-06, "loss": 1.3286, "step": 10150 }, { "epoch": 0.7955329153605015, "grad_norm": 2.3116328716278076, "learning_rate": 9.069771634617772e-06, "loss": 1.5599, "step": 10151 }, { "epoch": 0.7956112852664576, "grad_norm": 2.007272481918335, "learning_rate": 9.069592490455492e-06, "loss": 1.5872, "step": 10152 }, { "epoch": 0.7956896551724137, "grad_norm": 1.92897367477417, "learning_rate": 9.069413330814573e-06, "loss": 1.0271, "step": 10153 }, { "epoch": 0.7957680250783699, "grad_norm": 2.3290181159973145, "learning_rate": 9.069234155695695e-06, "loss": 1.715, "step": 10154 }, { "epoch": 0.795846394984326, "grad_norm": 3.963702440261841, "learning_rate": 9.069054965099542e-06, "loss": 1.2199, "step": 10155 }, { "epoch": 0.7959247648902821, "grad_norm": 2.749833345413208, "learning_rate": 9.068875759026792e-06, "loss": 1.4399, "step": 10156 }, { "epoch": 0.7960031347962383, "grad_norm": 3.4617724418640137, "learning_rate": 9.06869653747813e-06, "loss": 1.9079, "step": 10157 }, { "epoch": 0.7960815047021944, "grad_norm": 2.1744749546051025, "learning_rate": 9.068517300454234e-06, "loss": 1.1775, "step": 10158 }, { "epoch": 0.7961598746081505, "grad_norm": 2.9273459911346436, "learning_rate": 9.068338047955792e-06, "loss": 1.5874, "step": 10159 }, { "epoch": 0.7962382445141066, "grad_norm": 6.745927333831787, "learning_rate": 9.068158779983478e-06, "loss": 0.9593, "step": 10160 }, { "epoch": 0.7963166144200627, "grad_norm": 3.528796911239624, "learning_rate": 9.067979496537979e-06, "loss": 0.9897, "step": 10161 }, { "epoch": 0.7963949843260189, "grad_norm": 2.9370713233947754, "learning_rate": 9.067800197619976e-06, "loss": 1.6609, "step": 10162 }, { "epoch": 0.796473354231975, "grad_norm": 2.5601048469543457, "learning_rate": 9.067620883230149e-06, "loss": 1.4321, "step": 10163 }, { "epoch": 0.7965517241379311, "grad_norm": 3.8255820274353027, "learning_rate": 9.067441553369184e-06, "loss": 1.5397, "step": 10164 }, { "epoch": 0.7966300940438872, "grad_norm": 2.4551024436950684, "learning_rate": 9.067262208037756e-06, "loss": 1.2583, "step": 10165 }, { "epoch": 0.7967084639498433, "grad_norm": 2.294783115386963, "learning_rate": 9.067082847236554e-06, "loss": 1.1424, "step": 10166 }, { "epoch": 0.7967868338557994, "grad_norm": 2.2096965312957764, "learning_rate": 9.06690347096626e-06, "loss": 1.2986, "step": 10167 }, { "epoch": 0.7968652037617555, "grad_norm": 2.499351978302002, "learning_rate": 9.066724079227552e-06, "loss": 1.2168, "step": 10168 }, { "epoch": 0.7969435736677116, "grad_norm": 12.145710945129395, "learning_rate": 9.066544672021113e-06, "loss": 1.4903, "step": 10169 }, { "epoch": 0.7970219435736677, "grad_norm": 3.2864937782287598, "learning_rate": 9.066365249347629e-06, "loss": 1.4551, "step": 10170 }, { "epoch": 0.7971003134796238, "grad_norm": 2.257375955581665, "learning_rate": 9.066185811207779e-06, "loss": 1.5627, "step": 10171 }, { "epoch": 0.79717868338558, "grad_norm": 3.0058999061584473, "learning_rate": 9.066006357602249e-06, "loss": 1.5705, "step": 10172 }, { "epoch": 0.7972570532915361, "grad_norm": 2.8905601501464844, "learning_rate": 9.065826888531717e-06, "loss": 1.2982, "step": 10173 }, { "epoch": 0.7973354231974922, "grad_norm": 2.4814839363098145, "learning_rate": 9.065647403996867e-06, "loss": 1.1178, "step": 10174 }, { "epoch": 0.7974137931034483, "grad_norm": 6.114429473876953, "learning_rate": 9.065467903998384e-06, "loss": 0.9433, "step": 10175 }, { "epoch": 0.7974921630094044, "grad_norm": 2.6727564334869385, "learning_rate": 9.065288388536949e-06, "loss": 1.1157, "step": 10176 }, { "epoch": 0.7975705329153605, "grad_norm": 2.4587037563323975, "learning_rate": 9.065108857613245e-06, "loss": 1.2901, "step": 10177 }, { "epoch": 0.7976489028213166, "grad_norm": 3.157905340194702, "learning_rate": 9.064929311227956e-06, "loss": 1.6626, "step": 10178 }, { "epoch": 0.7977272727272727, "grad_norm": 2.1479201316833496, "learning_rate": 9.064749749381763e-06, "loss": 1.5612, "step": 10179 }, { "epoch": 0.7978056426332288, "grad_norm": 2.295219898223877, "learning_rate": 9.064570172075349e-06, "loss": 1.1832, "step": 10180 }, { "epoch": 0.797884012539185, "grad_norm": 3.3835372924804688, "learning_rate": 9.064390579309399e-06, "loss": 1.6249, "step": 10181 }, { "epoch": 0.7979623824451411, "grad_norm": 2.797471284866333, "learning_rate": 9.064210971084595e-06, "loss": 1.1878, "step": 10182 }, { "epoch": 0.7980407523510972, "grad_norm": 2.0475902557373047, "learning_rate": 9.064031347401618e-06, "loss": 1.1647, "step": 10183 }, { "epoch": 0.7981191222570533, "grad_norm": 2.548055410385132, "learning_rate": 9.063851708261153e-06, "loss": 1.6314, "step": 10184 }, { "epoch": 0.7981974921630094, "grad_norm": 2.653486967086792, "learning_rate": 9.063672053663885e-06, "loss": 1.7306, "step": 10185 }, { "epoch": 0.7982758620689655, "grad_norm": 3.9422178268432617, "learning_rate": 9.063492383610497e-06, "loss": 1.6134, "step": 10186 }, { "epoch": 0.7983542319749216, "grad_norm": 2.449519634246826, "learning_rate": 9.06331269810167e-06, "loss": 1.0892, "step": 10187 }, { "epoch": 0.7984326018808777, "grad_norm": 2.729851484298706, "learning_rate": 9.063132997138087e-06, "loss": 1.1838, "step": 10188 }, { "epoch": 0.7985109717868338, "grad_norm": 4.067031383514404, "learning_rate": 9.062953280720435e-06, "loss": 1.6117, "step": 10189 }, { "epoch": 0.79858934169279, "grad_norm": 3.3603062629699707, "learning_rate": 9.062773548849393e-06, "loss": 1.1036, "step": 10190 }, { "epoch": 0.798667711598746, "grad_norm": 2.889784097671509, "learning_rate": 9.06259380152565e-06, "loss": 1.3, "step": 10191 }, { "epoch": 0.7987460815047022, "grad_norm": 3.605011463165283, "learning_rate": 9.062414038749885e-06, "loss": 1.4635, "step": 10192 }, { "epoch": 0.7988244514106583, "grad_norm": 6.670330047607422, "learning_rate": 9.062234260522785e-06, "loss": 1.7668, "step": 10193 }, { "epoch": 0.7989028213166144, "grad_norm": 4.16096305847168, "learning_rate": 9.062054466845031e-06, "loss": 1.3386, "step": 10194 }, { "epoch": 0.7989811912225705, "grad_norm": 3.6727969646453857, "learning_rate": 9.06187465771731e-06, "loss": 1.6513, "step": 10195 }, { "epoch": 0.7990595611285266, "grad_norm": 2.768169641494751, "learning_rate": 9.061694833140303e-06, "loss": 2.0495, "step": 10196 }, { "epoch": 0.7991379310344827, "grad_norm": 3.166924238204956, "learning_rate": 9.061514993114695e-06, "loss": 1.5093, "step": 10197 }, { "epoch": 0.7992163009404388, "grad_norm": 2.864637613296509, "learning_rate": 9.06133513764117e-06, "loss": 1.4976, "step": 10198 }, { "epoch": 0.7992946708463949, "grad_norm": 1.9320902824401855, "learning_rate": 9.061155266720412e-06, "loss": 1.3328, "step": 10199 }, { "epoch": 0.799373040752351, "grad_norm": 3.688570499420166, "learning_rate": 9.060975380353106e-06, "loss": 1.2881, "step": 10200 }, { "epoch": 0.7994514106583072, "grad_norm": 4.032062530517578, "learning_rate": 9.060795478539935e-06, "loss": 1.9546, "step": 10201 }, { "epoch": 0.7995297805642633, "grad_norm": 2.8538968563079834, "learning_rate": 9.060615561281585e-06, "loss": 1.3376, "step": 10202 }, { "epoch": 0.7996081504702194, "grad_norm": 3.5597033500671387, "learning_rate": 9.060435628578738e-06, "loss": 1.6914, "step": 10203 }, { "epoch": 0.7996865203761756, "grad_norm": 4.299002647399902, "learning_rate": 9.060255680432079e-06, "loss": 1.2722, "step": 10204 }, { "epoch": 0.7997648902821317, "grad_norm": 2.376490831375122, "learning_rate": 9.060075716842294e-06, "loss": 1.3386, "step": 10205 }, { "epoch": 0.7998432601880878, "grad_norm": 2.849820137023926, "learning_rate": 9.059895737810065e-06, "loss": 1.5452, "step": 10206 }, { "epoch": 0.7999216300940439, "grad_norm": 3.3094959259033203, "learning_rate": 9.059715743336079e-06, "loss": 1.4432, "step": 10207 }, { "epoch": 0.8, "grad_norm": 2.4247798919677734, "learning_rate": 9.059535733421019e-06, "loss": 1.4269, "step": 10208 }, { "epoch": 0.8000783699059562, "grad_norm": 2.5406670570373535, "learning_rate": 9.05935570806557e-06, "loss": 1.6259, "step": 10209 }, { "epoch": 0.8001567398119123, "grad_norm": 2.4831204414367676, "learning_rate": 9.059175667270417e-06, "loss": 1.3117, "step": 10210 }, { "epoch": 0.8002351097178684, "grad_norm": 3.3691155910491943, "learning_rate": 9.058995611036245e-06, "loss": 1.6232, "step": 10211 }, { "epoch": 0.8003134796238245, "grad_norm": 5.073750019073486, "learning_rate": 9.058815539363738e-06, "loss": 1.09, "step": 10212 }, { "epoch": 0.8003918495297806, "grad_norm": 5.625173091888428, "learning_rate": 9.058635452253582e-06, "loss": 1.7572, "step": 10213 }, { "epoch": 0.8004702194357367, "grad_norm": 3.194967269897461, "learning_rate": 9.058455349706461e-06, "loss": 1.4857, "step": 10214 }, { "epoch": 0.8005485893416928, "grad_norm": 1.7986701726913452, "learning_rate": 9.058275231723062e-06, "loss": 1.3498, "step": 10215 }, { "epoch": 0.8006269592476489, "grad_norm": 2.350384473800659, "learning_rate": 9.058095098304067e-06, "loss": 1.3531, "step": 10216 }, { "epoch": 0.800705329153605, "grad_norm": 2.746912956237793, "learning_rate": 9.057914949450162e-06, "loss": 1.3541, "step": 10217 }, { "epoch": 0.8007836990595611, "grad_norm": 3.634469747543335, "learning_rate": 9.057734785162033e-06, "loss": 1.6743, "step": 10218 }, { "epoch": 0.8008620689655173, "grad_norm": 3.1595866680145264, "learning_rate": 9.057554605440366e-06, "loss": 1.2255, "step": 10219 }, { "epoch": 0.8009404388714734, "grad_norm": 4.018957138061523, "learning_rate": 9.057374410285846e-06, "loss": 1.103, "step": 10220 }, { "epoch": 0.8010188087774295, "grad_norm": 2.7246363162994385, "learning_rate": 9.057194199699158e-06, "loss": 1.3709, "step": 10221 }, { "epoch": 0.8010971786833856, "grad_norm": 2.595215082168579, "learning_rate": 9.057013973680985e-06, "loss": 1.2971, "step": 10222 }, { "epoch": 0.8011755485893417, "grad_norm": 2.7640316486358643, "learning_rate": 9.056833732232015e-06, "loss": 1.534, "step": 10223 }, { "epoch": 0.8012539184952978, "grad_norm": 4.047426700592041, "learning_rate": 9.056653475352934e-06, "loss": 0.6772, "step": 10224 }, { "epoch": 0.8013322884012539, "grad_norm": 3.302980899810791, "learning_rate": 9.056473203044428e-06, "loss": 1.7003, "step": 10225 }, { "epoch": 0.80141065830721, "grad_norm": 3.102022409439087, "learning_rate": 9.056292915307179e-06, "loss": 1.3952, "step": 10226 }, { "epoch": 0.8014890282131661, "grad_norm": 4.054737091064453, "learning_rate": 9.056112612141875e-06, "loss": 1.3473, "step": 10227 }, { "epoch": 0.8015673981191223, "grad_norm": 7.92425012588501, "learning_rate": 9.055932293549204e-06, "loss": 1.9525, "step": 10228 }, { "epoch": 0.8016457680250784, "grad_norm": 2.200268507003784, "learning_rate": 9.055751959529848e-06, "loss": 1.5214, "step": 10229 }, { "epoch": 0.8017241379310345, "grad_norm": 2.0134241580963135, "learning_rate": 9.055571610084497e-06, "loss": 1.5632, "step": 10230 }, { "epoch": 0.8018025078369906, "grad_norm": 2.5977301597595215, "learning_rate": 9.055391245213833e-06, "loss": 1.4743, "step": 10231 }, { "epoch": 0.8018808777429467, "grad_norm": 3.0791499614715576, "learning_rate": 9.055210864918544e-06, "loss": 1.8289, "step": 10232 }, { "epoch": 0.8019592476489028, "grad_norm": 3.0416862964630127, "learning_rate": 9.055030469199316e-06, "loss": 1.5091, "step": 10233 }, { "epoch": 0.8020376175548589, "grad_norm": 2.641310453414917, "learning_rate": 9.054850058056834e-06, "loss": 1.5429, "step": 10234 }, { "epoch": 0.802115987460815, "grad_norm": 6.409017562866211, "learning_rate": 9.054669631491787e-06, "loss": 1.8149, "step": 10235 }, { "epoch": 0.8021943573667711, "grad_norm": 3.1653707027435303, "learning_rate": 9.054489189504858e-06, "loss": 1.6853, "step": 10236 }, { "epoch": 0.8022727272727272, "grad_norm": 3.273502826690674, "learning_rate": 9.054308732096735e-06, "loss": 1.5091, "step": 10237 }, { "epoch": 0.8023510971786834, "grad_norm": 2.3713111877441406, "learning_rate": 9.054128259268103e-06, "loss": 1.4452, "step": 10238 }, { "epoch": 0.8024294670846395, "grad_norm": 5.923417568206787, "learning_rate": 9.053947771019652e-06, "loss": 1.3522, "step": 10239 }, { "epoch": 0.8025078369905956, "grad_norm": 2.854177236557007, "learning_rate": 9.053767267352063e-06, "loss": 1.3625, "step": 10240 }, { "epoch": 0.8025862068965517, "grad_norm": 6.2187395095825195, "learning_rate": 9.053586748266027e-06, "loss": 1.1488, "step": 10241 }, { "epoch": 0.8026645768025078, "grad_norm": 2.6403989791870117, "learning_rate": 9.053406213762229e-06, "loss": 1.2672, "step": 10242 }, { "epoch": 0.8027429467084639, "grad_norm": 2.289078950881958, "learning_rate": 9.053225663841354e-06, "loss": 1.3348, "step": 10243 }, { "epoch": 0.80282131661442, "grad_norm": 6.680500030517578, "learning_rate": 9.053045098504092e-06, "loss": 1.7376, "step": 10244 }, { "epoch": 0.8028996865203761, "grad_norm": 7.733011722564697, "learning_rate": 9.052864517751128e-06, "loss": 1.5344, "step": 10245 }, { "epoch": 0.8029780564263322, "grad_norm": 6.903885841369629, "learning_rate": 9.05268392158315e-06, "loss": 1.4775, "step": 10246 }, { "epoch": 0.8030564263322884, "grad_norm": 6.634057998657227, "learning_rate": 9.052503310000842e-06, "loss": 1.0322, "step": 10247 }, { "epoch": 0.8031347962382445, "grad_norm": 3.047804355621338, "learning_rate": 9.052322683004895e-06, "loss": 1.6469, "step": 10248 }, { "epoch": 0.8032131661442006, "grad_norm": 3.9028120040893555, "learning_rate": 9.052142040595993e-06, "loss": 1.2639, "step": 10249 }, { "epoch": 0.8032915360501567, "grad_norm": 2.785979986190796, "learning_rate": 9.051961382774823e-06, "loss": 1.084, "step": 10250 }, { "epoch": 0.8033699059561129, "grad_norm": 2.22810435295105, "learning_rate": 9.051780709542075e-06, "loss": 1.1317, "step": 10251 }, { "epoch": 0.803448275862069, "grad_norm": 3.418729305267334, "learning_rate": 9.051600020898432e-06, "loss": 1.1367, "step": 10252 }, { "epoch": 0.8035266457680251, "grad_norm": 2.7478504180908203, "learning_rate": 9.051419316844585e-06, "loss": 1.3508, "step": 10253 }, { "epoch": 0.8036050156739812, "grad_norm": 3.5691726207733154, "learning_rate": 9.05123859738122e-06, "loss": 1.4919, "step": 10254 }, { "epoch": 0.8036833855799373, "grad_norm": 2.499776840209961, "learning_rate": 9.051057862509025e-06, "loss": 1.7326, "step": 10255 }, { "epoch": 0.8037617554858935, "grad_norm": 3.0975751876831055, "learning_rate": 9.050877112228685e-06, "loss": 1.3029, "step": 10256 }, { "epoch": 0.8038401253918496, "grad_norm": 2.1616835594177246, "learning_rate": 9.050696346540891e-06, "loss": 1.5436, "step": 10257 }, { "epoch": 0.8039184952978057, "grad_norm": 2.2917606830596924, "learning_rate": 9.050515565446326e-06, "loss": 1.1948, "step": 10258 }, { "epoch": 0.8039968652037618, "grad_norm": 2.4967522621154785, "learning_rate": 9.050334768945681e-06, "loss": 1.7698, "step": 10259 }, { "epoch": 0.8040752351097179, "grad_norm": 3.5493202209472656, "learning_rate": 9.050153957039645e-06, "loss": 1.2443, "step": 10260 }, { "epoch": 0.804153605015674, "grad_norm": 4.292097091674805, "learning_rate": 9.049973129728902e-06, "loss": 1.1284, "step": 10261 }, { "epoch": 0.8042319749216301, "grad_norm": 3.0442519187927246, "learning_rate": 9.04979228701414e-06, "loss": 1.1661, "step": 10262 }, { "epoch": 0.8043103448275862, "grad_norm": 2.1818764209747314, "learning_rate": 9.049611428896052e-06, "loss": 1.3421, "step": 10263 }, { "epoch": 0.8043887147335423, "grad_norm": 3.7246620655059814, "learning_rate": 9.04943055537532e-06, "loss": 1.2562, "step": 10264 }, { "epoch": 0.8044670846394985, "grad_norm": 2.9831864833831787, "learning_rate": 9.049249666452633e-06, "loss": 1.1282, "step": 10265 }, { "epoch": 0.8045454545454546, "grad_norm": 4.078615188598633, "learning_rate": 9.04906876212868e-06, "loss": 1.4982, "step": 10266 }, { "epoch": 0.8046238244514107, "grad_norm": 2.49006724357605, "learning_rate": 9.048887842404151e-06, "loss": 1.3215, "step": 10267 }, { "epoch": 0.8047021943573668, "grad_norm": 2.533419132232666, "learning_rate": 9.048706907279731e-06, "loss": 1.7007, "step": 10268 }, { "epoch": 0.8047805642633229, "grad_norm": 5.937045574188232, "learning_rate": 9.048525956756111e-06, "loss": 1.6302, "step": 10269 }, { "epoch": 0.804858934169279, "grad_norm": 3.476015329360962, "learning_rate": 9.048344990833978e-06, "loss": 1.2608, "step": 10270 }, { "epoch": 0.8049373040752351, "grad_norm": 3.240485668182373, "learning_rate": 9.048164009514017e-06, "loss": 1.5537, "step": 10271 }, { "epoch": 0.8050156739811912, "grad_norm": 2.366800546646118, "learning_rate": 9.047983012796922e-06, "loss": 1.4021, "step": 10272 }, { "epoch": 0.8050940438871473, "grad_norm": 2.3816254138946533, "learning_rate": 9.047802000683378e-06, "loss": 1.393, "step": 10273 }, { "epoch": 0.8051724137931034, "grad_norm": 3.0683140754699707, "learning_rate": 9.047620973174076e-06, "loss": 1.4123, "step": 10274 }, { "epoch": 0.8052507836990596, "grad_norm": 2.939784288406372, "learning_rate": 9.047439930269698e-06, "loss": 1.6775, "step": 10275 }, { "epoch": 0.8053291536050157, "grad_norm": 2.6611664295196533, "learning_rate": 9.047258871970943e-06, "loss": 0.8461, "step": 10276 }, { "epoch": 0.8054075235109718, "grad_norm": 2.9386677742004395, "learning_rate": 9.04707779827849e-06, "loss": 1.2526, "step": 10277 }, { "epoch": 0.8054858934169279, "grad_norm": 4.001298904418945, "learning_rate": 9.046896709193033e-06, "loss": 1.3925, "step": 10278 }, { "epoch": 0.805564263322884, "grad_norm": 3.5179543495178223, "learning_rate": 9.046715604715261e-06, "loss": 0.9533, "step": 10279 }, { "epoch": 0.8056426332288401, "grad_norm": 2.3524069786071777, "learning_rate": 9.04653448484586e-06, "loss": 1.5065, "step": 10280 }, { "epoch": 0.8057210031347962, "grad_norm": 3.060373544692993, "learning_rate": 9.046353349585522e-06, "loss": 1.7149, "step": 10281 }, { "epoch": 0.8057993730407523, "grad_norm": 2.969947099685669, "learning_rate": 9.046172198934932e-06, "loss": 1.0943, "step": 10282 }, { "epoch": 0.8058777429467084, "grad_norm": 2.65315580368042, "learning_rate": 9.04599103289478e-06, "loss": 1.2469, "step": 10283 }, { "epoch": 0.8059561128526646, "grad_norm": 2.1371169090270996, "learning_rate": 9.045809851465759e-06, "loss": 1.4268, "step": 10284 }, { "epoch": 0.8060344827586207, "grad_norm": 3.3824164867401123, "learning_rate": 9.045628654648555e-06, "loss": 0.9197, "step": 10285 }, { "epoch": 0.8061128526645768, "grad_norm": 2.698162317276001, "learning_rate": 9.045447442443857e-06, "loss": 1.3186, "step": 10286 }, { "epoch": 0.8061912225705329, "grad_norm": 2.660578489303589, "learning_rate": 9.045266214852355e-06, "loss": 1.4378, "step": 10287 }, { "epoch": 0.806269592476489, "grad_norm": 3.2001707553863525, "learning_rate": 9.045084971874738e-06, "loss": 1.3109, "step": 10288 }, { "epoch": 0.8063479623824451, "grad_norm": 2.3216629028320312, "learning_rate": 9.044903713511695e-06, "loss": 1.1752, "step": 10289 }, { "epoch": 0.8064263322884012, "grad_norm": 2.5481185913085938, "learning_rate": 9.044722439763916e-06, "loss": 1.366, "step": 10290 }, { "epoch": 0.8065047021943573, "grad_norm": 4.267971992492676, "learning_rate": 9.04454115063209e-06, "loss": 1.4395, "step": 10291 }, { "epoch": 0.8065830721003134, "grad_norm": 4.807520866394043, "learning_rate": 9.044359846116909e-06, "loss": 1.2136, "step": 10292 }, { "epoch": 0.8066614420062695, "grad_norm": 2.5813276767730713, "learning_rate": 9.044178526219058e-06, "loss": 0.8975, "step": 10293 }, { "epoch": 0.8067398119122257, "grad_norm": 2.375046730041504, "learning_rate": 9.043997190939231e-06, "loss": 1.4537, "step": 10294 }, { "epoch": 0.8068181818181818, "grad_norm": 2.717223882675171, "learning_rate": 9.043815840278114e-06, "loss": 1.3482, "step": 10295 }, { "epoch": 0.8068965517241379, "grad_norm": 2.317972183227539, "learning_rate": 9.0436344742364e-06, "loss": 1.7007, "step": 10296 }, { "epoch": 0.806974921630094, "grad_norm": 4.352264881134033, "learning_rate": 9.043453092814778e-06, "loss": 1.315, "step": 10297 }, { "epoch": 0.8070532915360502, "grad_norm": 4.667867183685303, "learning_rate": 9.043271696013935e-06, "loss": 1.6667, "step": 10298 }, { "epoch": 0.8071316614420063, "grad_norm": 2.916788101196289, "learning_rate": 9.043090283834565e-06, "loss": 1.9564, "step": 10299 }, { "epoch": 0.8072100313479624, "grad_norm": 2.563969612121582, "learning_rate": 9.042908856277354e-06, "loss": 1.4285, "step": 10300 }, { "epoch": 0.8072884012539185, "grad_norm": 6.7600603103637695, "learning_rate": 9.042727413342998e-06, "loss": 1.3898, "step": 10301 }, { "epoch": 0.8073667711598747, "grad_norm": 1.9246448278427124, "learning_rate": 9.04254595503218e-06, "loss": 1.4224, "step": 10302 }, { "epoch": 0.8074451410658308, "grad_norm": 4.278360366821289, "learning_rate": 9.042364481345596e-06, "loss": 1.0594, "step": 10303 }, { "epoch": 0.8075235109717869, "grad_norm": 2.834639310836792, "learning_rate": 9.042182992283933e-06, "loss": 1.5356, "step": 10304 }, { "epoch": 0.807601880877743, "grad_norm": 2.2925455570220947, "learning_rate": 9.04200148784788e-06, "loss": 1.3817, "step": 10305 }, { "epoch": 0.8076802507836991, "grad_norm": 2.928729772567749, "learning_rate": 9.041819968038132e-06, "loss": 1.3767, "step": 10306 }, { "epoch": 0.8077586206896552, "grad_norm": 1.9347387552261353, "learning_rate": 9.041638432855375e-06, "loss": 1.6244, "step": 10307 }, { "epoch": 0.8078369905956113, "grad_norm": 2.6710524559020996, "learning_rate": 9.041456882300303e-06, "loss": 1.1484, "step": 10308 }, { "epoch": 0.8079153605015674, "grad_norm": 3.636876344680786, "learning_rate": 9.041275316373605e-06, "loss": 1.3465, "step": 10309 }, { "epoch": 0.8079937304075235, "grad_norm": 2.9494760036468506, "learning_rate": 9.041093735075969e-06, "loss": 1.4373, "step": 10310 }, { "epoch": 0.8080721003134796, "grad_norm": 2.51884126663208, "learning_rate": 9.040912138408092e-06, "loss": 1.6329, "step": 10311 }, { "epoch": 0.8081504702194358, "grad_norm": 9.39806842803955, "learning_rate": 9.040730526370658e-06, "loss": 1.3723, "step": 10312 }, { "epoch": 0.8082288401253919, "grad_norm": 2.7259833812713623, "learning_rate": 9.040548898964362e-06, "loss": 0.9944, "step": 10313 }, { "epoch": 0.808307210031348, "grad_norm": 3.1071321964263916, "learning_rate": 9.040367256189891e-06, "loss": 1.2084, "step": 10314 }, { "epoch": 0.8083855799373041, "grad_norm": 3.407294273376465, "learning_rate": 9.040185598047939e-06, "loss": 1.4182, "step": 10315 }, { "epoch": 0.8084639498432602, "grad_norm": 2.7715721130371094, "learning_rate": 9.040003924539198e-06, "loss": 1.4015, "step": 10316 }, { "epoch": 0.8085423197492163, "grad_norm": 2.8928370475769043, "learning_rate": 9.039822235664356e-06, "loss": 1.465, "step": 10317 }, { "epoch": 0.8086206896551724, "grad_norm": 3.994981050491333, "learning_rate": 9.039640531424104e-06, "loss": 1.4056, "step": 10318 }, { "epoch": 0.8086990595611285, "grad_norm": 3.9885995388031006, "learning_rate": 9.039458811819135e-06, "loss": 1.1252, "step": 10319 }, { "epoch": 0.8087774294670846, "grad_norm": 3.048799514770508, "learning_rate": 9.039277076850141e-06, "loss": 1.4923, "step": 10320 }, { "epoch": 0.8088557993730408, "grad_norm": 4.127601146697998, "learning_rate": 9.039095326517809e-06, "loss": 1.4619, "step": 10321 }, { "epoch": 0.8089341692789969, "grad_norm": 2.869553327560425, "learning_rate": 9.038913560822835e-06, "loss": 1.6536, "step": 10322 }, { "epoch": 0.809012539184953, "grad_norm": 3.5248970985412598, "learning_rate": 9.038731779765907e-06, "loss": 1.3545, "step": 10323 }, { "epoch": 0.8090909090909091, "grad_norm": 1.8693650960922241, "learning_rate": 9.03854998334772e-06, "loss": 1.3849, "step": 10324 }, { "epoch": 0.8091692789968652, "grad_norm": 3.563220739364624, "learning_rate": 9.03836817156896e-06, "loss": 1.3875, "step": 10325 }, { "epoch": 0.8092476489028213, "grad_norm": 2.284858465194702, "learning_rate": 9.038186344430325e-06, "loss": 1.3339, "step": 10326 }, { "epoch": 0.8093260188087774, "grad_norm": 2.3979337215423584, "learning_rate": 9.038004501932502e-06, "loss": 1.6324, "step": 10327 }, { "epoch": 0.8094043887147335, "grad_norm": 2.5730807781219482, "learning_rate": 9.037822644076184e-06, "loss": 1.1914, "step": 10328 }, { "epoch": 0.8094827586206896, "grad_norm": 2.4042253494262695, "learning_rate": 9.037640770862062e-06, "loss": 1.3603, "step": 10329 }, { "epoch": 0.8095611285266457, "grad_norm": 2.5306992530822754, "learning_rate": 9.037458882290829e-06, "loss": 1.4303, "step": 10330 }, { "epoch": 0.8096394984326019, "grad_norm": 4.522964954376221, "learning_rate": 9.037276978363175e-06, "loss": 1.9415, "step": 10331 }, { "epoch": 0.809717868338558, "grad_norm": 4.400429725646973, "learning_rate": 9.037095059079794e-06, "loss": 1.3923, "step": 10332 }, { "epoch": 0.8097962382445141, "grad_norm": 2.5999915599823, "learning_rate": 9.036913124441378e-06, "loss": 1.5101, "step": 10333 }, { "epoch": 0.8098746081504702, "grad_norm": 2.097618579864502, "learning_rate": 9.036731174448617e-06, "loss": 0.972, "step": 10334 }, { "epoch": 0.8099529780564263, "grad_norm": 5.279187202453613, "learning_rate": 9.036549209102204e-06, "loss": 1.3191, "step": 10335 }, { "epoch": 0.8100313479623824, "grad_norm": 5.203166961669922, "learning_rate": 9.03636722840283e-06, "loss": 1.8293, "step": 10336 }, { "epoch": 0.8101097178683385, "grad_norm": 2.779393434524536, "learning_rate": 9.036185232351189e-06, "loss": 1.4009, "step": 10337 }, { "epoch": 0.8101880877742946, "grad_norm": 3.6107656955718994, "learning_rate": 9.036003220947973e-06, "loss": 1.0766, "step": 10338 }, { "epoch": 0.8102664576802507, "grad_norm": 2.1436712741851807, "learning_rate": 9.035821194193875e-06, "loss": 1.0517, "step": 10339 }, { "epoch": 0.8103448275862069, "grad_norm": 3.7738499641418457, "learning_rate": 9.035639152089584e-06, "loss": 1.4851, "step": 10340 }, { "epoch": 0.810423197492163, "grad_norm": 4.0495381355285645, "learning_rate": 9.035457094635794e-06, "loss": 1.4071, "step": 10341 }, { "epoch": 0.8105015673981191, "grad_norm": 2.350360870361328, "learning_rate": 9.0352750218332e-06, "loss": 1.3797, "step": 10342 }, { "epoch": 0.8105799373040752, "grad_norm": 3.2903876304626465, "learning_rate": 9.035092933682492e-06, "loss": 1.451, "step": 10343 }, { "epoch": 0.8106583072100313, "grad_norm": 2.7291219234466553, "learning_rate": 9.034910830184362e-06, "loss": 1.5056, "step": 10344 }, { "epoch": 0.8107366771159875, "grad_norm": 2.4220447540283203, "learning_rate": 9.034728711339504e-06, "loss": 1.628, "step": 10345 }, { "epoch": 0.8108150470219436, "grad_norm": 2.2548956871032715, "learning_rate": 9.034546577148611e-06, "loss": 1.2395, "step": 10346 }, { "epoch": 0.8108934169278997, "grad_norm": 3.386294364929199, "learning_rate": 9.034364427612374e-06, "loss": 1.3562, "step": 10347 }, { "epoch": 0.8109717868338558, "grad_norm": 9.693181037902832, "learning_rate": 9.034182262731488e-06, "loss": 0.8506, "step": 10348 }, { "epoch": 0.811050156739812, "grad_norm": 3.6170880794525146, "learning_rate": 9.034000082506644e-06, "loss": 1.1735, "step": 10349 }, { "epoch": 0.8111285266457681, "grad_norm": 6.0361104011535645, "learning_rate": 9.033817886938536e-06, "loss": 1.3051, "step": 10350 }, { "epoch": 0.8112068965517242, "grad_norm": 4.143431186676025, "learning_rate": 9.033635676027858e-06, "loss": 1.6806, "step": 10351 }, { "epoch": 0.8112852664576803, "grad_norm": 3.0709874629974365, "learning_rate": 9.0334534497753e-06, "loss": 1.2113, "step": 10352 }, { "epoch": 0.8113636363636364, "grad_norm": 2.608292818069458, "learning_rate": 9.033271208181558e-06, "loss": 1.4315, "step": 10353 }, { "epoch": 0.8114420062695925, "grad_norm": 2.3967199325561523, "learning_rate": 9.033088951247321e-06, "loss": 0.7088, "step": 10354 }, { "epoch": 0.8115203761755486, "grad_norm": 2.2372171878814697, "learning_rate": 9.032906678973289e-06, "loss": 1.4408, "step": 10355 }, { "epoch": 0.8115987460815047, "grad_norm": 2.109017848968506, "learning_rate": 9.03272439136015e-06, "loss": 1.372, "step": 10356 }, { "epoch": 0.8116771159874608, "grad_norm": 4.280178070068359, "learning_rate": 9.032542088408598e-06, "loss": 1.2731, "step": 10357 }, { "epoch": 0.811755485893417, "grad_norm": 2.8966941833496094, "learning_rate": 9.032359770119328e-06, "loss": 1.4709, "step": 10358 }, { "epoch": 0.8118338557993731, "grad_norm": 4.590433597564697, "learning_rate": 9.032177436493032e-06, "loss": 0.8949, "step": 10359 }, { "epoch": 0.8119122257053292, "grad_norm": 2.9591190814971924, "learning_rate": 9.031995087530403e-06, "loss": 1.2841, "step": 10360 }, { "epoch": 0.8119905956112853, "grad_norm": 1.8927547931671143, "learning_rate": 9.031812723232138e-06, "loss": 0.8405, "step": 10361 }, { "epoch": 0.8120689655172414, "grad_norm": 2.189817190170288, "learning_rate": 9.031630343598927e-06, "loss": 1.494, "step": 10362 }, { "epoch": 0.8121473354231975, "grad_norm": 2.4468560218811035, "learning_rate": 9.031447948631466e-06, "loss": 1.2931, "step": 10363 }, { "epoch": 0.8122257053291536, "grad_norm": 2.5934174060821533, "learning_rate": 9.031265538330447e-06, "loss": 1.2936, "step": 10364 }, { "epoch": 0.8123040752351097, "grad_norm": 2.402857542037964, "learning_rate": 9.031083112696564e-06, "loss": 1.3526, "step": 10365 }, { "epoch": 0.8123824451410658, "grad_norm": 3.503237724304199, "learning_rate": 9.03090067173051e-06, "loss": 1.273, "step": 10366 }, { "epoch": 0.812460815047022, "grad_norm": 2.1676156520843506, "learning_rate": 9.030718215432983e-06, "loss": 1.6128, "step": 10367 }, { "epoch": 0.812539184952978, "grad_norm": 1.9577653408050537, "learning_rate": 9.030535743804673e-06, "loss": 1.4927, "step": 10368 }, { "epoch": 0.8126175548589342, "grad_norm": 2.2407400608062744, "learning_rate": 9.030353256846276e-06, "loss": 1.3666, "step": 10369 }, { "epoch": 0.8126959247648903, "grad_norm": 2.8206944465637207, "learning_rate": 9.030170754558484e-06, "loss": 1.4179, "step": 10370 }, { "epoch": 0.8127742946708464, "grad_norm": 4.07908821105957, "learning_rate": 9.029988236941994e-06, "loss": 1.232, "step": 10371 }, { "epoch": 0.8128526645768025, "grad_norm": 3.0788466930389404, "learning_rate": 9.029805703997497e-06, "loss": 1.6015, "step": 10372 }, { "epoch": 0.8129310344827586, "grad_norm": 4.281591892242432, "learning_rate": 9.02962315572569e-06, "loss": 1.6511, "step": 10373 }, { "epoch": 0.8130094043887147, "grad_norm": 2.82177996635437, "learning_rate": 9.029440592127265e-06, "loss": 1.2784, "step": 10374 }, { "epoch": 0.8130877742946708, "grad_norm": 2.498976230621338, "learning_rate": 9.02925801320292e-06, "loss": 1.7396, "step": 10375 }, { "epoch": 0.8131661442006269, "grad_norm": 2.583813428878784, "learning_rate": 9.029075418953346e-06, "loss": 1.7164, "step": 10376 }, { "epoch": 0.813244514106583, "grad_norm": 3.735905408859253, "learning_rate": 9.028892809379237e-06, "loss": 1.0769, "step": 10377 }, { "epoch": 0.8133228840125392, "grad_norm": 2.905646562576294, "learning_rate": 9.02871018448129e-06, "loss": 1.497, "step": 10378 }, { "epoch": 0.8134012539184953, "grad_norm": 2.924777030944824, "learning_rate": 9.028527544260201e-06, "loss": 1.4355, "step": 10379 }, { "epoch": 0.8134796238244514, "grad_norm": 2.6510534286499023, "learning_rate": 9.02834488871666e-06, "loss": 1.0998, "step": 10380 }, { "epoch": 0.8135579937304075, "grad_norm": 3.8262290954589844, "learning_rate": 9.028162217851364e-06, "loss": 0.6833, "step": 10381 }, { "epoch": 0.8136363636363636, "grad_norm": 2.2162578105926514, "learning_rate": 9.02797953166501e-06, "loss": 0.9061, "step": 10382 }, { "epoch": 0.8137147335423197, "grad_norm": 9.875829696655273, "learning_rate": 9.027796830158288e-06, "loss": 1.398, "step": 10383 }, { "epoch": 0.8137931034482758, "grad_norm": 2.6922831535339355, "learning_rate": 9.027614113331897e-06, "loss": 1.5068, "step": 10384 }, { "epoch": 0.8138714733542319, "grad_norm": 2.2515485286712646, "learning_rate": 9.027431381186531e-06, "loss": 1.2047, "step": 10385 }, { "epoch": 0.813949843260188, "grad_norm": 2.515659809112549, "learning_rate": 9.027248633722885e-06, "loss": 1.4983, "step": 10386 }, { "epoch": 0.8140282131661442, "grad_norm": 4.382804870605469, "learning_rate": 9.027065870941653e-06, "loss": 1.3461, "step": 10387 }, { "epoch": 0.8141065830721003, "grad_norm": 2.3084311485290527, "learning_rate": 9.02688309284353e-06, "loss": 1.4321, "step": 10388 }, { "epoch": 0.8141849529780564, "grad_norm": 4.742498397827148, "learning_rate": 9.026700299429214e-06, "loss": 1.0688, "step": 10389 }, { "epoch": 0.8142633228840125, "grad_norm": 6.760594844818115, "learning_rate": 9.026517490699397e-06, "loss": 1.5858, "step": 10390 }, { "epoch": 0.8143416927899687, "grad_norm": 4.6599812507629395, "learning_rate": 9.026334666654777e-06, "loss": 1.562, "step": 10391 }, { "epoch": 0.8144200626959248, "grad_norm": 2.123358964920044, "learning_rate": 9.026151827296046e-06, "loss": 1.2949, "step": 10392 }, { "epoch": 0.8144984326018809, "grad_norm": 3.0314364433288574, "learning_rate": 9.025968972623902e-06, "loss": 1.2105, "step": 10393 }, { "epoch": 0.814576802507837, "grad_norm": 3.1666510105133057, "learning_rate": 9.02578610263904e-06, "loss": 1.5336, "step": 10394 }, { "epoch": 0.8146551724137931, "grad_norm": 4.228791236877441, "learning_rate": 9.025603217342156e-06, "loss": 1.3145, "step": 10395 }, { "epoch": 0.8147335423197493, "grad_norm": 2.8697338104248047, "learning_rate": 9.025420316733945e-06, "loss": 1.6717, "step": 10396 }, { "epoch": 0.8148119122257054, "grad_norm": 3.914459228515625, "learning_rate": 9.025237400815102e-06, "loss": 1.7644, "step": 10397 }, { "epoch": 0.8148902821316615, "grad_norm": 3.3793811798095703, "learning_rate": 9.025054469586325e-06, "loss": 1.2886, "step": 10398 }, { "epoch": 0.8149686520376176, "grad_norm": 2.1689093112945557, "learning_rate": 9.024871523048306e-06, "loss": 1.3358, "step": 10399 }, { "epoch": 0.8150470219435737, "grad_norm": 3.489424228668213, "learning_rate": 9.024688561201744e-06, "loss": 1.4799, "step": 10400 }, { "epoch": 0.8151253918495298, "grad_norm": 2.955430269241333, "learning_rate": 9.024505584047334e-06, "loss": 1.4731, "step": 10401 }, { "epoch": 0.8152037617554859, "grad_norm": 2.3058478832244873, "learning_rate": 9.024322591585772e-06, "loss": 1.4958, "step": 10402 }, { "epoch": 0.815282131661442, "grad_norm": 2.212125062942505, "learning_rate": 9.024139583817754e-06, "loss": 0.9603, "step": 10403 }, { "epoch": 0.8153605015673981, "grad_norm": 2.3033885955810547, "learning_rate": 9.023956560743976e-06, "loss": 1.3388, "step": 10404 }, { "epoch": 0.8154388714733543, "grad_norm": 3.0350475311279297, "learning_rate": 9.023773522365133e-06, "loss": 1.2979, "step": 10405 }, { "epoch": 0.8155172413793104, "grad_norm": 2.522761583328247, "learning_rate": 9.023590468681923e-06, "loss": 1.1219, "step": 10406 }, { "epoch": 0.8155956112852665, "grad_norm": 5.377599239349365, "learning_rate": 9.02340739969504e-06, "loss": 1.0785, "step": 10407 }, { "epoch": 0.8156739811912226, "grad_norm": 2.7762296199798584, "learning_rate": 9.023224315405184e-06, "loss": 1.3781, "step": 10408 }, { "epoch": 0.8157523510971787, "grad_norm": 3.117680549621582, "learning_rate": 9.023041215813048e-06, "loss": 0.9824, "step": 10409 }, { "epoch": 0.8158307210031348, "grad_norm": 2.604194164276123, "learning_rate": 9.022858100919332e-06, "loss": 1.4077, "step": 10410 }, { "epoch": 0.8159090909090909, "grad_norm": 2.351405382156372, "learning_rate": 9.022674970724728e-06, "loss": 1.4585, "step": 10411 }, { "epoch": 0.815987460815047, "grad_norm": 2.4191081523895264, "learning_rate": 9.022491825229934e-06, "loss": 1.1281, "step": 10412 }, { "epoch": 0.8160658307210031, "grad_norm": 2.916306257247925, "learning_rate": 9.022308664435648e-06, "loss": 1.0136, "step": 10413 }, { "epoch": 0.8161442006269592, "grad_norm": 2.5104453563690186, "learning_rate": 9.022125488342566e-06, "loss": 1.3671, "step": 10414 }, { "epoch": 0.8162225705329154, "grad_norm": 5.92319393157959, "learning_rate": 9.021942296951383e-06, "loss": 1.806, "step": 10415 }, { "epoch": 0.8163009404388715, "grad_norm": 2.635570526123047, "learning_rate": 9.021759090262799e-06, "loss": 1.3719, "step": 10416 }, { "epoch": 0.8163793103448276, "grad_norm": 3.007652997970581, "learning_rate": 9.02157586827751e-06, "loss": 0.9899, "step": 10417 }, { "epoch": 0.8164576802507837, "grad_norm": 2.348196029663086, "learning_rate": 9.02139263099621e-06, "loss": 1.2364, "step": 10418 }, { "epoch": 0.8165360501567398, "grad_norm": 4.2385711669921875, "learning_rate": 9.021209378419598e-06, "loss": 1.3351, "step": 10419 }, { "epoch": 0.8166144200626959, "grad_norm": 3.2876973152160645, "learning_rate": 9.021026110548372e-06, "loss": 1.57, "step": 10420 }, { "epoch": 0.816692789968652, "grad_norm": 2.534364700317383, "learning_rate": 9.020842827383228e-06, "loss": 1.5284, "step": 10421 }, { "epoch": 0.8167711598746081, "grad_norm": 5.792444229125977, "learning_rate": 9.020659528924863e-06, "loss": 1.504, "step": 10422 }, { "epoch": 0.8168495297805642, "grad_norm": 4.634458065032959, "learning_rate": 9.020476215173974e-06, "loss": 1.6537, "step": 10423 }, { "epoch": 0.8169278996865204, "grad_norm": 2.135356903076172, "learning_rate": 9.020292886131258e-06, "loss": 1.2838, "step": 10424 }, { "epoch": 0.8170062695924765, "grad_norm": 6.53789758682251, "learning_rate": 9.020109541797413e-06, "loss": 1.0848, "step": 10425 }, { "epoch": 0.8170846394984326, "grad_norm": 2.3752403259277344, "learning_rate": 9.019926182173137e-06, "loss": 0.9839, "step": 10426 }, { "epoch": 0.8171630094043887, "grad_norm": 8.616578102111816, "learning_rate": 9.019742807259125e-06, "loss": 0.9799, "step": 10427 }, { "epoch": 0.8172413793103448, "grad_norm": 3.9851388931274414, "learning_rate": 9.019559417056077e-06, "loss": 1.8384, "step": 10428 }, { "epoch": 0.8173197492163009, "grad_norm": 2.205104351043701, "learning_rate": 9.019376011564691e-06, "loss": 1.3359, "step": 10429 }, { "epoch": 0.817398119122257, "grad_norm": 2.628051519393921, "learning_rate": 9.019192590785662e-06, "loss": 1.2267, "step": 10430 }, { "epoch": 0.8174764890282131, "grad_norm": 4.243892192840576, "learning_rate": 9.019009154719688e-06, "loss": 1.3997, "step": 10431 }, { "epoch": 0.8175548589341692, "grad_norm": 2.4712307453155518, "learning_rate": 9.018825703367468e-06, "loss": 1.4207, "step": 10432 }, { "epoch": 0.8176332288401253, "grad_norm": 9.078531265258789, "learning_rate": 9.018642236729698e-06, "loss": 1.5059, "step": 10433 }, { "epoch": 0.8177115987460815, "grad_norm": 3.9633729457855225, "learning_rate": 9.01845875480708e-06, "loss": 1.624, "step": 10434 }, { "epoch": 0.8177899686520376, "grad_norm": 2.4072937965393066, "learning_rate": 9.018275257600307e-06, "loss": 1.4878, "step": 10435 }, { "epoch": 0.8178683385579937, "grad_norm": 2.5567915439605713, "learning_rate": 9.018091745110079e-06, "loss": 1.3878, "step": 10436 }, { "epoch": 0.8179467084639498, "grad_norm": 2.3902785778045654, "learning_rate": 9.017908217337093e-06, "loss": 1.3519, "step": 10437 }, { "epoch": 0.818025078369906, "grad_norm": 4.307888507843018, "learning_rate": 9.017724674282049e-06, "loss": 0.9816, "step": 10438 }, { "epoch": 0.8181034482758621, "grad_norm": 4.970592021942139, "learning_rate": 9.017541115945644e-06, "loss": 2.1237, "step": 10439 }, { "epoch": 0.8181818181818182, "grad_norm": 3.135586738586426, "learning_rate": 9.017357542328576e-06, "loss": 1.2656, "step": 10440 }, { "epoch": 0.8182601880877743, "grad_norm": 2.7384800910949707, "learning_rate": 9.017173953431543e-06, "loss": 1.5302, "step": 10441 }, { "epoch": 0.8183385579937305, "grad_norm": 3.2557506561279297, "learning_rate": 9.016990349255243e-06, "loss": 1.1577, "step": 10442 }, { "epoch": 0.8184169278996866, "grad_norm": 3.2722437381744385, "learning_rate": 9.016806729800377e-06, "loss": 1.4631, "step": 10443 }, { "epoch": 0.8184952978056427, "grad_norm": 2.606315851211548, "learning_rate": 9.01662309506764e-06, "loss": 1.4816, "step": 10444 }, { "epoch": 0.8185736677115988, "grad_norm": 2.263972759246826, "learning_rate": 9.016439445057732e-06, "loss": 1.3945, "step": 10445 }, { "epoch": 0.8186520376175549, "grad_norm": 2.3592593669891357, "learning_rate": 9.01625577977135e-06, "loss": 1.2229, "step": 10446 }, { "epoch": 0.818730407523511, "grad_norm": 2.0950872898101807, "learning_rate": 9.016072099209196e-06, "loss": 1.3624, "step": 10447 }, { "epoch": 0.8188087774294671, "grad_norm": 2.7890801429748535, "learning_rate": 9.015888403371967e-06, "loss": 1.3188, "step": 10448 }, { "epoch": 0.8188871473354232, "grad_norm": 3.7642362117767334, "learning_rate": 9.015704692260358e-06, "loss": 1.6287, "step": 10449 }, { "epoch": 0.8189655172413793, "grad_norm": 3.8929810523986816, "learning_rate": 9.015520965875073e-06, "loss": 1.8652, "step": 10450 }, { "epoch": 0.8190438871473354, "grad_norm": 1.7075889110565186, "learning_rate": 9.015337224216808e-06, "loss": 1.3303, "step": 10451 }, { "epoch": 0.8191222570532916, "grad_norm": 2.2796969413757324, "learning_rate": 9.015153467286264e-06, "loss": 1.0424, "step": 10452 }, { "epoch": 0.8192006269592477, "grad_norm": 3.239135980606079, "learning_rate": 9.014969695084138e-06, "loss": 1.0971, "step": 10453 }, { "epoch": 0.8192789968652038, "grad_norm": 2.4189112186431885, "learning_rate": 9.014785907611129e-06, "loss": 1.5608, "step": 10454 }, { "epoch": 0.8193573667711599, "grad_norm": 2.1512553691864014, "learning_rate": 9.014602104867936e-06, "loss": 1.5026, "step": 10455 }, { "epoch": 0.819435736677116, "grad_norm": 2.3052985668182373, "learning_rate": 9.01441828685526e-06, "loss": 0.9108, "step": 10456 }, { "epoch": 0.8195141065830721, "grad_norm": 2.767536163330078, "learning_rate": 9.014234453573797e-06, "loss": 0.9889, "step": 10457 }, { "epoch": 0.8195924764890282, "grad_norm": 2.539482355117798, "learning_rate": 9.014050605024252e-06, "loss": 1.2316, "step": 10458 }, { "epoch": 0.8196708463949843, "grad_norm": 9.681797981262207, "learning_rate": 9.013866741207316e-06, "loss": 1.462, "step": 10459 }, { "epoch": 0.8197492163009404, "grad_norm": 9.107319831848145, "learning_rate": 9.013682862123694e-06, "loss": 1.4709, "step": 10460 }, { "epoch": 0.8198275862068966, "grad_norm": 2.3463873863220215, "learning_rate": 9.013498967774084e-06, "loss": 1.0927, "step": 10461 }, { "epoch": 0.8199059561128527, "grad_norm": 3.1877291202545166, "learning_rate": 9.013315058159185e-06, "loss": 1.194, "step": 10462 }, { "epoch": 0.8199843260188088, "grad_norm": 2.2682130336761475, "learning_rate": 9.013131133279697e-06, "loss": 1.543, "step": 10463 }, { "epoch": 0.8200626959247649, "grad_norm": 2.7064855098724365, "learning_rate": 9.01294719313632e-06, "loss": 1.7843, "step": 10464 }, { "epoch": 0.820141065830721, "grad_norm": 3.1120331287384033, "learning_rate": 9.012763237729753e-06, "loss": 1.3956, "step": 10465 }, { "epoch": 0.8202194357366771, "grad_norm": 3.572465658187866, "learning_rate": 9.012579267060694e-06, "loss": 1.4371, "step": 10466 }, { "epoch": 0.8202978056426332, "grad_norm": 2.5062813758850098, "learning_rate": 9.012395281129847e-06, "loss": 1.4225, "step": 10467 }, { "epoch": 0.8203761755485893, "grad_norm": 2.35941743850708, "learning_rate": 9.012211279937908e-06, "loss": 1.3132, "step": 10468 }, { "epoch": 0.8204545454545454, "grad_norm": 2.35463809967041, "learning_rate": 9.012027263485578e-06, "loss": 1.4634, "step": 10469 }, { "epoch": 0.8205329153605015, "grad_norm": 2.065831422805786, "learning_rate": 9.011843231773558e-06, "loss": 1.4914, "step": 10470 }, { "epoch": 0.8206112852664577, "grad_norm": 2.458199977874756, "learning_rate": 9.011659184802547e-06, "loss": 1.2168, "step": 10471 }, { "epoch": 0.8206896551724138, "grad_norm": 2.8432958126068115, "learning_rate": 9.011475122573243e-06, "loss": 1.3708, "step": 10472 }, { "epoch": 0.8207680250783699, "grad_norm": 2.7517502307891846, "learning_rate": 9.011291045086349e-06, "loss": 0.6493, "step": 10473 }, { "epoch": 0.820846394984326, "grad_norm": 2.626875877380371, "learning_rate": 9.011106952342564e-06, "loss": 0.6392, "step": 10474 }, { "epoch": 0.8209247648902821, "grad_norm": 4.547032833099365, "learning_rate": 9.010922844342588e-06, "loss": 1.9912, "step": 10475 }, { "epoch": 0.8210031347962382, "grad_norm": 2.383573055267334, "learning_rate": 9.010738721087123e-06, "loss": 1.1931, "step": 10476 }, { "epoch": 0.8210815047021943, "grad_norm": 3.2398018836975098, "learning_rate": 9.010554582576868e-06, "loss": 1.5731, "step": 10477 }, { "epoch": 0.8211598746081504, "grad_norm": 3.303630828857422, "learning_rate": 9.010370428812522e-06, "loss": 1.4511, "step": 10478 }, { "epoch": 0.8212382445141065, "grad_norm": 3.8045454025268555, "learning_rate": 9.010186259794789e-06, "loss": 1.2735, "step": 10479 }, { "epoch": 0.8213166144200627, "grad_norm": 2.182866334915161, "learning_rate": 9.010002075524365e-06, "loss": 1.5067, "step": 10480 }, { "epoch": 0.8213949843260188, "grad_norm": 1.8413110971450806, "learning_rate": 9.009817876001953e-06, "loss": 1.2899, "step": 10481 }, { "epoch": 0.8214733542319749, "grad_norm": 4.127146244049072, "learning_rate": 9.009633661228253e-06, "loss": 1.2094, "step": 10482 }, { "epoch": 0.821551724137931, "grad_norm": 4.045842170715332, "learning_rate": 9.009449431203966e-06, "loss": 1.307, "step": 10483 }, { "epoch": 0.8216300940438871, "grad_norm": 3.942232370376587, "learning_rate": 9.009265185929795e-06, "loss": 1.1793, "step": 10484 }, { "epoch": 0.8217084639498433, "grad_norm": 4.840092182159424, "learning_rate": 9.009080925406438e-06, "loss": 1.4922, "step": 10485 }, { "epoch": 0.8217868338557994, "grad_norm": 10.775123596191406, "learning_rate": 9.008896649634596e-06, "loss": 2.2489, "step": 10486 }, { "epoch": 0.8218652037617555, "grad_norm": 6.66716194152832, "learning_rate": 9.008712358614968e-06, "loss": 1.4243, "step": 10487 }, { "epoch": 0.8219435736677116, "grad_norm": 4.570934295654297, "learning_rate": 9.00852805234826e-06, "loss": 1.6518, "step": 10488 }, { "epoch": 0.8220219435736678, "grad_norm": 2.611046552658081, "learning_rate": 9.00834373083517e-06, "loss": 1.559, "step": 10489 }, { "epoch": 0.8221003134796239, "grad_norm": 3.6543304920196533, "learning_rate": 9.008159394076396e-06, "loss": 1.1128, "step": 10490 }, { "epoch": 0.82217868338558, "grad_norm": 2.5659172534942627, "learning_rate": 9.007975042072646e-06, "loss": 1.4045, "step": 10491 }, { "epoch": 0.8222570532915361, "grad_norm": 2.403083086013794, "learning_rate": 9.007790674824616e-06, "loss": 1.7008, "step": 10492 }, { "epoch": 0.8223354231974922, "grad_norm": 3.0439469814300537, "learning_rate": 9.007606292333008e-06, "loss": 1.5325, "step": 10493 }, { "epoch": 0.8224137931034483, "grad_norm": 3.4260413646698, "learning_rate": 9.007421894598527e-06, "loss": 1.4035, "step": 10494 }, { "epoch": 0.8224921630094044, "grad_norm": 2.876657009124756, "learning_rate": 9.007237481621869e-06, "loss": 1.0822, "step": 10495 }, { "epoch": 0.8225705329153605, "grad_norm": 3.831003427505493, "learning_rate": 9.007053053403739e-06, "loss": 1.1672, "step": 10496 }, { "epoch": 0.8226489028213166, "grad_norm": 3.0510880947113037, "learning_rate": 9.006868609944837e-06, "loss": 1.9134, "step": 10497 }, { "epoch": 0.8227272727272728, "grad_norm": 5.648441791534424, "learning_rate": 9.006684151245863e-06, "loss": 1.3512, "step": 10498 }, { "epoch": 0.8228056426332289, "grad_norm": 4.612936019897461, "learning_rate": 9.006499677307523e-06, "loss": 1.234, "step": 10499 }, { "epoch": 0.822884012539185, "grad_norm": 2.9127540588378906, "learning_rate": 9.006315188130515e-06, "loss": 1.111, "step": 10500 }, { "epoch": 0.8229623824451411, "grad_norm": 3.092780113220215, "learning_rate": 9.006130683715543e-06, "loss": 1.396, "step": 10501 }, { "epoch": 0.8230407523510972, "grad_norm": 2.2206871509552, "learning_rate": 9.005946164063306e-06, "loss": 1.3918, "step": 10502 }, { "epoch": 0.8231191222570533, "grad_norm": 2.9851112365722656, "learning_rate": 9.005761629174507e-06, "loss": 1.1772, "step": 10503 }, { "epoch": 0.8231974921630094, "grad_norm": 8.424493789672852, "learning_rate": 9.00557707904985e-06, "loss": 1.5161, "step": 10504 }, { "epoch": 0.8232758620689655, "grad_norm": 2.503182888031006, "learning_rate": 9.005392513690034e-06, "loss": 1.1837, "step": 10505 }, { "epoch": 0.8233542319749216, "grad_norm": 4.301951885223389, "learning_rate": 9.005207933095763e-06, "loss": 1.7234, "step": 10506 }, { "epoch": 0.8234326018808777, "grad_norm": 2.091791868209839, "learning_rate": 9.005023337267737e-06, "loss": 1.1938, "step": 10507 }, { "epoch": 0.8235109717868339, "grad_norm": 4.5704026222229, "learning_rate": 9.004838726206661e-06, "loss": 1.7564, "step": 10508 }, { "epoch": 0.82358934169279, "grad_norm": 2.6144983768463135, "learning_rate": 9.004654099913234e-06, "loss": 1.5978, "step": 10509 }, { "epoch": 0.8236677115987461, "grad_norm": 5.77439546585083, "learning_rate": 9.004469458388161e-06, "loss": 1.8138, "step": 10510 }, { "epoch": 0.8237460815047022, "grad_norm": 3.000873327255249, "learning_rate": 9.004284801632143e-06, "loss": 1.5571, "step": 10511 }, { "epoch": 0.8238244514106583, "grad_norm": 11.20986557006836, "learning_rate": 9.004100129645881e-06, "loss": 0.8944, "step": 10512 }, { "epoch": 0.8239028213166144, "grad_norm": 5.415170192718506, "learning_rate": 9.00391544243008e-06, "loss": 1.4067, "step": 10513 }, { "epoch": 0.8239811912225705, "grad_norm": 4.500932216644287, "learning_rate": 9.003730739985442e-06, "loss": 1.24, "step": 10514 }, { "epoch": 0.8240595611285266, "grad_norm": 2.7755026817321777, "learning_rate": 9.003546022312666e-06, "loss": 1.3222, "step": 10515 }, { "epoch": 0.8241379310344827, "grad_norm": 4.2514753341674805, "learning_rate": 9.00336128941246e-06, "loss": 1.5112, "step": 10516 }, { "epoch": 0.8242163009404389, "grad_norm": 2.8673593997955322, "learning_rate": 9.00317654128552e-06, "loss": 1.2582, "step": 10517 }, { "epoch": 0.824294670846395, "grad_norm": 5.182305335998535, "learning_rate": 9.002991777932557e-06, "loss": 1.7785, "step": 10518 }, { "epoch": 0.8243730407523511, "grad_norm": 2.4845969676971436, "learning_rate": 9.002806999354268e-06, "loss": 0.5588, "step": 10519 }, { "epoch": 0.8244514106583072, "grad_norm": 2.52695894241333, "learning_rate": 9.002622205551357e-06, "loss": 1.5885, "step": 10520 }, { "epoch": 0.8245297805642633, "grad_norm": 4.114311695098877, "learning_rate": 9.002437396524525e-06, "loss": 1.189, "step": 10521 }, { "epoch": 0.8246081504702194, "grad_norm": 2.627004861831665, "learning_rate": 9.002252572274479e-06, "loss": 1.6, "step": 10522 }, { "epoch": 0.8246865203761755, "grad_norm": 2.204101800918579, "learning_rate": 9.00206773280192e-06, "loss": 1.1665, "step": 10523 }, { "epoch": 0.8247648902821316, "grad_norm": 3.4904446601867676, "learning_rate": 9.00188287810755e-06, "loss": 1.2844, "step": 10524 }, { "epoch": 0.8248432601880877, "grad_norm": 8.021377563476562, "learning_rate": 9.001698008192074e-06, "loss": 0.5259, "step": 10525 }, { "epoch": 0.8249216300940438, "grad_norm": 2.754939317703247, "learning_rate": 9.001513123056193e-06, "loss": 1.3332, "step": 10526 }, { "epoch": 0.825, "grad_norm": 2.0799524784088135, "learning_rate": 9.001328222700615e-06, "loss": 1.6119, "step": 10527 }, { "epoch": 0.8250783699059561, "grad_norm": 2.2957143783569336, "learning_rate": 9.001143307126034e-06, "loss": 0.8459, "step": 10528 }, { "epoch": 0.8251567398119122, "grad_norm": 2.622372627258301, "learning_rate": 9.000958376333161e-06, "loss": 1.4429, "step": 10529 }, { "epoch": 0.8252351097178683, "grad_norm": 3.839022159576416, "learning_rate": 9.0007734303227e-06, "loss": 0.9896, "step": 10530 }, { "epoch": 0.8253134796238244, "grad_norm": 5.55050802230835, "learning_rate": 9.000588469095349e-06, "loss": 1.2625, "step": 10531 }, { "epoch": 0.8253918495297806, "grad_norm": 1.9836297035217285, "learning_rate": 9.000403492651815e-06, "loss": 0.8582, "step": 10532 }, { "epoch": 0.8254702194357367, "grad_norm": 1.9245887994766235, "learning_rate": 9.000218500992802e-06, "loss": 1.6405, "step": 10533 }, { "epoch": 0.8255485893416928, "grad_norm": 2.146838426589966, "learning_rate": 9.000033494119011e-06, "loss": 1.4065, "step": 10534 }, { "epoch": 0.825626959247649, "grad_norm": 2.653829574584961, "learning_rate": 8.999848472031147e-06, "loss": 1.4799, "step": 10535 }, { "epoch": 0.8257053291536051, "grad_norm": 2.3176982402801514, "learning_rate": 8.999663434729914e-06, "loss": 1.3305, "step": 10536 }, { "epoch": 0.8257836990595612, "grad_norm": 4.1964802742004395, "learning_rate": 8.999478382216017e-06, "loss": 1.3765, "step": 10537 }, { "epoch": 0.8258620689655173, "grad_norm": 3.1638388633728027, "learning_rate": 8.999293314490158e-06, "loss": 1.3188, "step": 10538 }, { "epoch": 0.8259404388714734, "grad_norm": 2.5771472454071045, "learning_rate": 8.999108231553041e-06, "loss": 1.592, "step": 10539 }, { "epoch": 0.8260188087774295, "grad_norm": 2.9842116832733154, "learning_rate": 8.99892313340537e-06, "loss": 1.3419, "step": 10540 }, { "epoch": 0.8260971786833856, "grad_norm": 3.4041218757629395, "learning_rate": 8.99873802004785e-06, "loss": 1.0682, "step": 10541 }, { "epoch": 0.8261755485893417, "grad_norm": 4.335494518280029, "learning_rate": 8.998552891481185e-06, "loss": 1.6846, "step": 10542 }, { "epoch": 0.8262539184952978, "grad_norm": 2.418597936630249, "learning_rate": 8.998367747706077e-06, "loss": 1.4717, "step": 10543 }, { "epoch": 0.826332288401254, "grad_norm": 2.9139537811279297, "learning_rate": 8.998182588723232e-06, "loss": 1.7183, "step": 10544 }, { "epoch": 0.82641065830721, "grad_norm": 2.0651776790618896, "learning_rate": 8.997997414533355e-06, "loss": 1.6836, "step": 10545 }, { "epoch": 0.8264890282131662, "grad_norm": 2.328948497772217, "learning_rate": 8.99781222513715e-06, "loss": 1.1568, "step": 10546 }, { "epoch": 0.8265673981191223, "grad_norm": 3.0577638149261475, "learning_rate": 8.997627020535318e-06, "loss": 1.7509, "step": 10547 }, { "epoch": 0.8266457680250784, "grad_norm": 3.8347339630126953, "learning_rate": 8.997441800728568e-06, "loss": 1.6658, "step": 10548 }, { "epoch": 0.8267241379310345, "grad_norm": 2.226750135421753, "learning_rate": 8.997256565717602e-06, "loss": 1.319, "step": 10549 }, { "epoch": 0.8268025078369906, "grad_norm": 2.116285562515259, "learning_rate": 8.997071315503125e-06, "loss": 1.2753, "step": 10550 }, { "epoch": 0.8268808777429467, "grad_norm": 4.393030166625977, "learning_rate": 8.996886050085843e-06, "loss": 1.1531, "step": 10551 }, { "epoch": 0.8269592476489028, "grad_norm": 3.1571033000946045, "learning_rate": 8.99670076946646e-06, "loss": 1.3069, "step": 10552 }, { "epoch": 0.8270376175548589, "grad_norm": 2.5682895183563232, "learning_rate": 8.996515473645678e-06, "loss": 1.4578, "step": 10553 }, { "epoch": 0.827115987460815, "grad_norm": 4.68904972076416, "learning_rate": 8.996330162624205e-06, "loss": 1.4086, "step": 10554 }, { "epoch": 0.8271943573667712, "grad_norm": 2.9050891399383545, "learning_rate": 8.996144836402744e-06, "loss": 1.3711, "step": 10555 }, { "epoch": 0.8272727272727273, "grad_norm": 11.633191108703613, "learning_rate": 8.995959494982e-06, "loss": 1.6569, "step": 10556 }, { "epoch": 0.8273510971786834, "grad_norm": 3.3266358375549316, "learning_rate": 8.995774138362681e-06, "loss": 1.5612, "step": 10557 }, { "epoch": 0.8274294670846395, "grad_norm": 4.052304744720459, "learning_rate": 8.995588766545488e-06, "loss": 0.8903, "step": 10558 }, { "epoch": 0.8275078369905956, "grad_norm": 2.670954465866089, "learning_rate": 8.995403379531127e-06, "loss": 1.2402, "step": 10559 }, { "epoch": 0.8275862068965517, "grad_norm": 3.9993674755096436, "learning_rate": 8.995217977320304e-06, "loss": 1.715, "step": 10560 }, { "epoch": 0.8276645768025078, "grad_norm": 2.786273956298828, "learning_rate": 8.995032559913725e-06, "loss": 1.2463, "step": 10561 }, { "epoch": 0.8277429467084639, "grad_norm": 2.33266282081604, "learning_rate": 8.994847127312092e-06, "loss": 1.335, "step": 10562 }, { "epoch": 0.82782131661442, "grad_norm": 5.512632846832275, "learning_rate": 8.994661679516114e-06, "loss": 1.4035, "step": 10563 }, { "epoch": 0.8278996865203762, "grad_norm": 3.6507070064544678, "learning_rate": 8.994476216526495e-06, "loss": 1.4114, "step": 10564 }, { "epoch": 0.8279780564263323, "grad_norm": 4.282935619354248, "learning_rate": 8.99429073834394e-06, "loss": 1.6222, "step": 10565 }, { "epoch": 0.8280564263322884, "grad_norm": 3.490931272506714, "learning_rate": 8.994105244969153e-06, "loss": 1.5479, "step": 10566 }, { "epoch": 0.8281347962382445, "grad_norm": 7.6793904304504395, "learning_rate": 8.993919736402844e-06, "loss": 1.4066, "step": 10567 }, { "epoch": 0.8282131661442006, "grad_norm": 2.771544933319092, "learning_rate": 8.993734212645714e-06, "loss": 1.329, "step": 10568 }, { "epoch": 0.8282915360501567, "grad_norm": 6.4143805503845215, "learning_rate": 8.99354867369847e-06, "loss": 1.1243, "step": 10569 }, { "epoch": 0.8283699059561128, "grad_norm": 3.2520523071289062, "learning_rate": 8.993363119561819e-06, "loss": 1.4656, "step": 10570 }, { "epoch": 0.8284482758620689, "grad_norm": 3.410491466522217, "learning_rate": 8.993177550236464e-06, "loss": 1.4988, "step": 10571 }, { "epoch": 0.828526645768025, "grad_norm": 4.335739612579346, "learning_rate": 8.992991965723114e-06, "loss": 1.3605, "step": 10572 }, { "epoch": 0.8286050156739811, "grad_norm": 5.751286029815674, "learning_rate": 8.992806366022475e-06, "loss": 1.8818, "step": 10573 }, { "epoch": 0.8286833855799373, "grad_norm": 2.19356632232666, "learning_rate": 8.99262075113525e-06, "loss": 1.3487, "step": 10574 }, { "epoch": 0.8287617554858934, "grad_norm": 1.7799664735794067, "learning_rate": 8.992435121062146e-06, "loss": 1.5379, "step": 10575 }, { "epoch": 0.8288401253918495, "grad_norm": 1.9945522546768188, "learning_rate": 8.992249475803869e-06, "loss": 1.417, "step": 10576 }, { "epoch": 0.8289184952978056, "grad_norm": 3.961566209793091, "learning_rate": 8.992063815361127e-06, "loss": 1.6178, "step": 10577 }, { "epoch": 0.8289968652037617, "grad_norm": 2.7808425426483154, "learning_rate": 8.991878139734625e-06, "loss": 1.5405, "step": 10578 }, { "epoch": 0.8290752351097179, "grad_norm": 1.8446555137634277, "learning_rate": 8.991692448925067e-06, "loss": 0.9742, "step": 10579 }, { "epoch": 0.829153605015674, "grad_norm": 2.42116379737854, "learning_rate": 8.991506742933162e-06, "loss": 1.6874, "step": 10580 }, { "epoch": 0.8292319749216301, "grad_norm": 2.119027614593506, "learning_rate": 8.991321021759616e-06, "loss": 1.5247, "step": 10581 }, { "epoch": 0.8293103448275863, "grad_norm": 2.2634825706481934, "learning_rate": 8.991135285405134e-06, "loss": 1.0503, "step": 10582 }, { "epoch": 0.8293887147335424, "grad_norm": 3.288055896759033, "learning_rate": 8.990949533870424e-06, "loss": 1.3845, "step": 10583 }, { "epoch": 0.8294670846394985, "grad_norm": 2.926360845565796, "learning_rate": 8.990763767156193e-06, "loss": 1.5909, "step": 10584 }, { "epoch": 0.8295454545454546, "grad_norm": 3.414372444152832, "learning_rate": 8.990577985263144e-06, "loss": 1.6655, "step": 10585 }, { "epoch": 0.8296238244514107, "grad_norm": 3.7205495834350586, "learning_rate": 8.990392188191987e-06, "loss": 1.4864, "step": 10586 }, { "epoch": 0.8297021943573668, "grad_norm": 2.2160582542419434, "learning_rate": 8.99020637594343e-06, "loss": 1.143, "step": 10587 }, { "epoch": 0.8297805642633229, "grad_norm": 4.401430606842041, "learning_rate": 8.990020548518174e-06, "loss": 0.6545, "step": 10588 }, { "epoch": 0.829858934169279, "grad_norm": 5.733059406280518, "learning_rate": 8.989834705916929e-06, "loss": 1.2201, "step": 10589 }, { "epoch": 0.8299373040752351, "grad_norm": 2.000704526901245, "learning_rate": 8.989648848140405e-06, "loss": 1.1917, "step": 10590 }, { "epoch": 0.8300156739811912, "grad_norm": 2.6386055946350098, "learning_rate": 8.989462975189305e-06, "loss": 1.4519, "step": 10591 }, { "epoch": 0.8300940438871474, "grad_norm": 2.3982186317443848, "learning_rate": 8.989277087064336e-06, "loss": 1.3098, "step": 10592 }, { "epoch": 0.8301724137931035, "grad_norm": 2.914928674697876, "learning_rate": 8.989091183766207e-06, "loss": 1.6964, "step": 10593 }, { "epoch": 0.8302507836990596, "grad_norm": 2.6044349670410156, "learning_rate": 8.988905265295625e-06, "loss": 1.1855, "step": 10594 }, { "epoch": 0.8303291536050157, "grad_norm": 2.689669132232666, "learning_rate": 8.988719331653294e-06, "loss": 1.267, "step": 10595 }, { "epoch": 0.8304075235109718, "grad_norm": 2.903061866760254, "learning_rate": 8.988533382839923e-06, "loss": 1.4772, "step": 10596 }, { "epoch": 0.8304858934169279, "grad_norm": 4.418796062469482, "learning_rate": 8.988347418856222e-06, "loss": 1.6485, "step": 10597 }, { "epoch": 0.830564263322884, "grad_norm": 3.381948709487915, "learning_rate": 8.988161439702895e-06, "loss": 1.6205, "step": 10598 }, { "epoch": 0.8306426332288401, "grad_norm": 2.2695388793945312, "learning_rate": 8.98797544538065e-06, "loss": 0.9253, "step": 10599 }, { "epoch": 0.8307210031347962, "grad_norm": 2.225238800048828, "learning_rate": 8.987789435890196e-06, "loss": 1.3823, "step": 10600 }, { "epoch": 0.8307993730407524, "grad_norm": 4.655694961547852, "learning_rate": 8.987603411232238e-06, "loss": 1.8331, "step": 10601 }, { "epoch": 0.8308777429467085, "grad_norm": 3.849130630493164, "learning_rate": 8.987417371407484e-06, "loss": 1.5622, "step": 10602 }, { "epoch": 0.8309561128526646, "grad_norm": 2.903733253479004, "learning_rate": 8.987231316416644e-06, "loss": 1.3514, "step": 10603 }, { "epoch": 0.8310344827586207, "grad_norm": 2.630607843399048, "learning_rate": 8.987045246260423e-06, "loss": 1.3938, "step": 10604 }, { "epoch": 0.8311128526645768, "grad_norm": 3.75736665725708, "learning_rate": 8.986859160939531e-06, "loss": 0.9384, "step": 10605 }, { "epoch": 0.8311912225705329, "grad_norm": 3.3976709842681885, "learning_rate": 8.986673060454674e-06, "loss": 1.3947, "step": 10606 }, { "epoch": 0.831269592476489, "grad_norm": 2.677687168121338, "learning_rate": 8.986486944806558e-06, "loss": 1.5615, "step": 10607 }, { "epoch": 0.8313479623824451, "grad_norm": 2.5467567443847656, "learning_rate": 8.986300813995897e-06, "loss": 1.4755, "step": 10608 }, { "epoch": 0.8314263322884012, "grad_norm": 2.5440921783447266, "learning_rate": 8.986114668023393e-06, "loss": 1.5325, "step": 10609 }, { "epoch": 0.8315047021943573, "grad_norm": 2.027885913848877, "learning_rate": 8.985928506889757e-06, "loss": 1.1277, "step": 10610 }, { "epoch": 0.8315830721003135, "grad_norm": 3.291565179824829, "learning_rate": 8.985742330595696e-06, "loss": 1.8083, "step": 10611 }, { "epoch": 0.8316614420062696, "grad_norm": 2.6727375984191895, "learning_rate": 8.985556139141918e-06, "loss": 1.518, "step": 10612 }, { "epoch": 0.8317398119122257, "grad_norm": 4.496749401092529, "learning_rate": 8.985369932529133e-06, "loss": 1.5629, "step": 10613 }, { "epoch": 0.8318181818181818, "grad_norm": 4.529305934906006, "learning_rate": 8.985183710758047e-06, "loss": 1.3189, "step": 10614 }, { "epoch": 0.8318965517241379, "grad_norm": 3.459075450897217, "learning_rate": 8.984997473829368e-06, "loss": 1.5275, "step": 10615 }, { "epoch": 0.831974921630094, "grad_norm": 1.9560426473617554, "learning_rate": 8.984811221743806e-06, "loss": 1.2126, "step": 10616 }, { "epoch": 0.8320532915360501, "grad_norm": 6.907675266265869, "learning_rate": 8.98462495450207e-06, "loss": 0.427, "step": 10617 }, { "epoch": 0.8321316614420062, "grad_norm": 2.4991753101348877, "learning_rate": 8.984438672104866e-06, "loss": 1.0168, "step": 10618 }, { "epoch": 0.8322100313479623, "grad_norm": 3.2696690559387207, "learning_rate": 8.984252374552904e-06, "loss": 1.0932, "step": 10619 }, { "epoch": 0.8322884012539185, "grad_norm": 2.2837276458740234, "learning_rate": 8.984066061846893e-06, "loss": 0.7499, "step": 10620 }, { "epoch": 0.8323667711598746, "grad_norm": 6.959095001220703, "learning_rate": 8.98387973398754e-06, "loss": 1.3054, "step": 10621 }, { "epoch": 0.8324451410658307, "grad_norm": 2.3882734775543213, "learning_rate": 8.983693390975556e-06, "loss": 1.355, "step": 10622 }, { "epoch": 0.8325235109717868, "grad_norm": 5.1740007400512695, "learning_rate": 8.983507032811646e-06, "loss": 1.508, "step": 10623 }, { "epoch": 0.8326018808777429, "grad_norm": 3.450016736984253, "learning_rate": 8.983320659496524e-06, "loss": 1.6635, "step": 10624 }, { "epoch": 0.832680250783699, "grad_norm": 8.579347610473633, "learning_rate": 8.983134271030895e-06, "loss": 1.3245, "step": 10625 }, { "epoch": 0.8327586206896552, "grad_norm": 4.267877578735352, "learning_rate": 8.98294786741547e-06, "loss": 1.8587, "step": 10626 }, { "epoch": 0.8328369905956113, "grad_norm": 3.4462828636169434, "learning_rate": 8.982761448650958e-06, "loss": 1.5108, "step": 10627 }, { "epoch": 0.8329153605015674, "grad_norm": 2.0421743392944336, "learning_rate": 8.982575014738066e-06, "loss": 1.0512, "step": 10628 }, { "epoch": 0.8329937304075236, "grad_norm": 2.215951442718506, "learning_rate": 8.982388565677504e-06, "loss": 1.321, "step": 10629 }, { "epoch": 0.8330721003134797, "grad_norm": 1.8109030723571777, "learning_rate": 8.98220210146998e-06, "loss": 1.2648, "step": 10630 }, { "epoch": 0.8331504702194358, "grad_norm": 2.8654775619506836, "learning_rate": 8.982015622116207e-06, "loss": 1.6687, "step": 10631 }, { "epoch": 0.8332288401253919, "grad_norm": 2.4115397930145264, "learning_rate": 8.98182912761689e-06, "loss": 1.4176, "step": 10632 }, { "epoch": 0.833307210031348, "grad_norm": 4.952770709991455, "learning_rate": 8.981642617972742e-06, "loss": 1.3474, "step": 10633 }, { "epoch": 0.8333855799373041, "grad_norm": 3.302933931350708, "learning_rate": 8.981456093184468e-06, "loss": 1.1647, "step": 10634 }, { "epoch": 0.8334639498432602, "grad_norm": 6.853118419647217, "learning_rate": 8.981269553252783e-06, "loss": 1.3865, "step": 10635 }, { "epoch": 0.8335423197492163, "grad_norm": 3.134215831756592, "learning_rate": 8.981082998178393e-06, "loss": 1.3767, "step": 10636 }, { "epoch": 0.8336206896551724, "grad_norm": 2.114213466644287, "learning_rate": 8.980896427962006e-06, "loss": 1.4325, "step": 10637 }, { "epoch": 0.8336990595611286, "grad_norm": 3.971479892730713, "learning_rate": 8.980709842604335e-06, "loss": 1.384, "step": 10638 }, { "epoch": 0.8337774294670847, "grad_norm": 3.1412525177001953, "learning_rate": 8.980523242106088e-06, "loss": 1.0742, "step": 10639 }, { "epoch": 0.8338557993730408, "grad_norm": 2.1296749114990234, "learning_rate": 8.980336626467976e-06, "loss": 1.3137, "step": 10640 }, { "epoch": 0.8339341692789969, "grad_norm": 2.500654697418213, "learning_rate": 8.980149995690707e-06, "loss": 1.4979, "step": 10641 }, { "epoch": 0.834012539184953, "grad_norm": 4.395920276641846, "learning_rate": 8.979963349774992e-06, "loss": 0.775, "step": 10642 }, { "epoch": 0.8340909090909091, "grad_norm": 6.369048595428467, "learning_rate": 8.97977668872154e-06, "loss": 2.1538, "step": 10643 }, { "epoch": 0.8341692789968652, "grad_norm": 3.497606039047241, "learning_rate": 8.979590012531064e-06, "loss": 1.3554, "step": 10644 }, { "epoch": 0.8342476489028213, "grad_norm": 2.9782538414001465, "learning_rate": 8.97940332120427e-06, "loss": 1.7239, "step": 10645 }, { "epoch": 0.8343260188087774, "grad_norm": 2.3886966705322266, "learning_rate": 8.979216614741869e-06, "loss": 1.5219, "step": 10646 }, { "epoch": 0.8344043887147335, "grad_norm": 9.082096099853516, "learning_rate": 8.979029893144571e-06, "loss": 1.3006, "step": 10647 }, { "epoch": 0.8344827586206897, "grad_norm": 5.916952133178711, "learning_rate": 8.97884315641309e-06, "loss": 1.1048, "step": 10648 }, { "epoch": 0.8345611285266458, "grad_norm": 2.8196022510528564, "learning_rate": 8.978656404548132e-06, "loss": 1.3704, "step": 10649 }, { "epoch": 0.8346394984326019, "grad_norm": 2.246387004852295, "learning_rate": 8.978469637550407e-06, "loss": 1.1245, "step": 10650 }, { "epoch": 0.834717868338558, "grad_norm": 3.6423697471618652, "learning_rate": 8.97828285542063e-06, "loss": 1.1129, "step": 10651 }, { "epoch": 0.8347962382445141, "grad_norm": 2.945352554321289, "learning_rate": 8.978096058159507e-06, "loss": 1.4315, "step": 10652 }, { "epoch": 0.8348746081504702, "grad_norm": 2.3495090007781982, "learning_rate": 8.977909245767749e-06, "loss": 1.4023, "step": 10653 }, { "epoch": 0.8349529780564263, "grad_norm": 3.0173263549804688, "learning_rate": 8.97772241824607e-06, "loss": 1.2401, "step": 10654 }, { "epoch": 0.8350313479623824, "grad_norm": 1.770290732383728, "learning_rate": 8.977535575595174e-06, "loss": 1.2426, "step": 10655 }, { "epoch": 0.8351097178683385, "grad_norm": 4.013082504272461, "learning_rate": 8.977348717815778e-06, "loss": 1.4525, "step": 10656 }, { "epoch": 0.8351880877742947, "grad_norm": 2.054805278778076, "learning_rate": 8.977161844908591e-06, "loss": 1.4159, "step": 10657 }, { "epoch": 0.8352664576802508, "grad_norm": 2.096857786178589, "learning_rate": 8.976974956874321e-06, "loss": 1.1199, "step": 10658 }, { "epoch": 0.8353448275862069, "grad_norm": 3.418818473815918, "learning_rate": 8.976788053713684e-06, "loss": 1.8612, "step": 10659 }, { "epoch": 0.835423197492163, "grad_norm": 2.5167899131774902, "learning_rate": 8.976601135427386e-06, "loss": 1.5493, "step": 10660 }, { "epoch": 0.8355015673981191, "grad_norm": 8.221694946289062, "learning_rate": 8.976414202016139e-06, "loss": 0.8568, "step": 10661 }, { "epoch": 0.8355799373040752, "grad_norm": 2.549590587615967, "learning_rate": 8.976227253480657e-06, "loss": 1.4531, "step": 10662 }, { "epoch": 0.8356583072100313, "grad_norm": 2.8528385162353516, "learning_rate": 8.976040289821647e-06, "loss": 1.0426, "step": 10663 }, { "epoch": 0.8357366771159874, "grad_norm": 2.6524014472961426, "learning_rate": 8.975853311039822e-06, "loss": 1.7309, "step": 10664 }, { "epoch": 0.8358150470219435, "grad_norm": 3.2937991619110107, "learning_rate": 8.975666317135895e-06, "loss": 1.4135, "step": 10665 }, { "epoch": 0.8358934169278996, "grad_norm": 2.831479549407959, "learning_rate": 8.975479308110572e-06, "loss": 1.4275, "step": 10666 }, { "epoch": 0.8359717868338558, "grad_norm": 2.650987386703491, "learning_rate": 8.97529228396457e-06, "loss": 1.5878, "step": 10667 }, { "epoch": 0.8360501567398119, "grad_norm": 2.112812042236328, "learning_rate": 8.975105244698598e-06, "loss": 1.3685, "step": 10668 }, { "epoch": 0.836128526645768, "grad_norm": 3.532705545425415, "learning_rate": 8.974918190313367e-06, "loss": 1.4572, "step": 10669 }, { "epoch": 0.8362068965517241, "grad_norm": 7.2142157554626465, "learning_rate": 8.974731120809589e-06, "loss": 2.083, "step": 10670 }, { "epoch": 0.8362852664576802, "grad_norm": 1.7491319179534912, "learning_rate": 8.974544036187976e-06, "loss": 1.2446, "step": 10671 }, { "epoch": 0.8363636363636363, "grad_norm": 9.118948936462402, "learning_rate": 8.974356936449238e-06, "loss": 0.7204, "step": 10672 }, { "epoch": 0.8364420062695925, "grad_norm": 2.1856696605682373, "learning_rate": 8.974169821594088e-06, "loss": 1.3465, "step": 10673 }, { "epoch": 0.8365203761755486, "grad_norm": 2.1263787746429443, "learning_rate": 8.973982691623237e-06, "loss": 1.2192, "step": 10674 }, { "epoch": 0.8365987460815048, "grad_norm": 3.770512580871582, "learning_rate": 8.973795546537396e-06, "loss": 1.0676, "step": 10675 }, { "epoch": 0.8366771159874609, "grad_norm": 2.7011029720306396, "learning_rate": 8.973608386337279e-06, "loss": 1.6585, "step": 10676 }, { "epoch": 0.836755485893417, "grad_norm": 3.780123710632324, "learning_rate": 8.973421211023597e-06, "loss": 1.4654, "step": 10677 }, { "epoch": 0.8368338557993731, "grad_norm": 2.0902814865112305, "learning_rate": 8.973234020597062e-06, "loss": 1.6411, "step": 10678 }, { "epoch": 0.8369122257053292, "grad_norm": 1.7572057247161865, "learning_rate": 8.973046815058383e-06, "loss": 1.2955, "step": 10679 }, { "epoch": 0.8369905956112853, "grad_norm": 2.195288896560669, "learning_rate": 8.972859594408277e-06, "loss": 1.1942, "step": 10680 }, { "epoch": 0.8370689655172414, "grad_norm": 2.772745132446289, "learning_rate": 8.972672358647455e-06, "loss": 1.674, "step": 10681 }, { "epoch": 0.8371473354231975, "grad_norm": 2.8316779136657715, "learning_rate": 8.972485107776625e-06, "loss": 1.4727, "step": 10682 }, { "epoch": 0.8372257053291536, "grad_norm": 3.020209312438965, "learning_rate": 8.972297841796502e-06, "loss": 1.0247, "step": 10683 }, { "epoch": 0.8373040752351097, "grad_norm": 5.002416133880615, "learning_rate": 8.972110560707799e-06, "loss": 2.0111, "step": 10684 }, { "epoch": 0.8373824451410659, "grad_norm": 3.3462705612182617, "learning_rate": 8.971923264511228e-06, "loss": 1.4519, "step": 10685 }, { "epoch": 0.837460815047022, "grad_norm": 2.955430269241333, "learning_rate": 8.9717359532075e-06, "loss": 1.5067, "step": 10686 }, { "epoch": 0.8375391849529781, "grad_norm": 3.298922061920166, "learning_rate": 8.97154862679733e-06, "loss": 1.8329, "step": 10687 }, { "epoch": 0.8376175548589342, "grad_norm": 2.662851572036743, "learning_rate": 8.971361285281429e-06, "loss": 1.4971, "step": 10688 }, { "epoch": 0.8376959247648903, "grad_norm": 4.180102348327637, "learning_rate": 8.971173928660509e-06, "loss": 2.0267, "step": 10689 }, { "epoch": 0.8377742946708464, "grad_norm": 4.003176689147949, "learning_rate": 8.970986556935282e-06, "loss": 1.267, "step": 10690 }, { "epoch": 0.8378526645768025, "grad_norm": 3.123335361480713, "learning_rate": 8.970799170106462e-06, "loss": 1.2624, "step": 10691 }, { "epoch": 0.8379310344827586, "grad_norm": 2.9931888580322266, "learning_rate": 8.970611768174761e-06, "loss": 1.4573, "step": 10692 }, { "epoch": 0.8380094043887147, "grad_norm": 2.336271286010742, "learning_rate": 8.970424351140893e-06, "loss": 1.1602, "step": 10693 }, { "epoch": 0.8380877742946709, "grad_norm": 2.8000035285949707, "learning_rate": 8.970236919005573e-06, "loss": 1.3756, "step": 10694 }, { "epoch": 0.838166144200627, "grad_norm": 2.250389814376831, "learning_rate": 8.97004947176951e-06, "loss": 1.2804, "step": 10695 }, { "epoch": 0.8382445141065831, "grad_norm": 4.53618860244751, "learning_rate": 8.969862009433414e-06, "loss": 1.4891, "step": 10696 }, { "epoch": 0.8383228840125392, "grad_norm": 2.368995189666748, "learning_rate": 8.969674531998005e-06, "loss": 1.1726, "step": 10697 }, { "epoch": 0.8384012539184953, "grad_norm": 2.5692050457000732, "learning_rate": 8.96948703946399e-06, "loss": 1.5355, "step": 10698 }, { "epoch": 0.8384796238244514, "grad_norm": 2.998288154602051, "learning_rate": 8.96929953183209e-06, "loss": 1.6709, "step": 10699 }, { "epoch": 0.8385579937304075, "grad_norm": 2.5474789142608643, "learning_rate": 8.96911200910301e-06, "loss": 1.1486, "step": 10700 }, { "epoch": 0.8386363636363636, "grad_norm": 2.6364729404449463, "learning_rate": 8.968924471277467e-06, "loss": 1.4298, "step": 10701 }, { "epoch": 0.8387147335423197, "grad_norm": 4.215959072113037, "learning_rate": 8.968736918356175e-06, "loss": 1.629, "step": 10702 }, { "epoch": 0.8387931034482758, "grad_norm": 2.10624098777771, "learning_rate": 8.968549350339845e-06, "loss": 1.3012, "step": 10703 }, { "epoch": 0.838871473354232, "grad_norm": 2.2410199642181396, "learning_rate": 8.968361767229193e-06, "loss": 1.0347, "step": 10704 }, { "epoch": 0.8389498432601881, "grad_norm": 3.650984287261963, "learning_rate": 8.96817416902493e-06, "loss": 1.1222, "step": 10705 }, { "epoch": 0.8390282131661442, "grad_norm": 8.5905179977417, "learning_rate": 8.96798655572777e-06, "loss": 1.4765, "step": 10706 }, { "epoch": 0.8391065830721003, "grad_norm": 2.855923652648926, "learning_rate": 8.967798927338429e-06, "loss": 1.802, "step": 10707 }, { "epoch": 0.8391849529780564, "grad_norm": 2.589334011077881, "learning_rate": 8.967611283857619e-06, "loss": 1.4034, "step": 10708 }, { "epoch": 0.8392633228840125, "grad_norm": 3.1758344173431396, "learning_rate": 8.967423625286052e-06, "loss": 1.8707, "step": 10709 }, { "epoch": 0.8393416927899686, "grad_norm": 2.944366455078125, "learning_rate": 8.967235951624444e-06, "loss": 1.4209, "step": 10710 }, { "epoch": 0.8394200626959247, "grad_norm": 4.009138107299805, "learning_rate": 8.967048262873508e-06, "loss": 1.7823, "step": 10711 }, { "epoch": 0.8394984326018808, "grad_norm": 4.01072359085083, "learning_rate": 8.966860559033959e-06, "loss": 1.2906, "step": 10712 }, { "epoch": 0.839576802507837, "grad_norm": 5.090914249420166, "learning_rate": 8.96667284010651e-06, "loss": 1.2928, "step": 10713 }, { "epoch": 0.8396551724137931, "grad_norm": 2.6506242752075195, "learning_rate": 8.966485106091873e-06, "loss": 1.2406, "step": 10714 }, { "epoch": 0.8397335423197492, "grad_norm": 3.409419298171997, "learning_rate": 8.966297356990764e-06, "loss": 1.1162, "step": 10715 }, { "epoch": 0.8398119122257053, "grad_norm": 3.347745895385742, "learning_rate": 8.966109592803899e-06, "loss": 1.4389, "step": 10716 }, { "epoch": 0.8398902821316614, "grad_norm": 4.060476303100586, "learning_rate": 8.965921813531989e-06, "loss": 1.7726, "step": 10717 }, { "epoch": 0.8399686520376175, "grad_norm": 4.662229537963867, "learning_rate": 8.96573401917575e-06, "loss": 1.3942, "step": 10718 }, { "epoch": 0.8400470219435736, "grad_norm": 10.025208473205566, "learning_rate": 8.965546209735896e-06, "loss": 1.3944, "step": 10719 }, { "epoch": 0.8401253918495298, "grad_norm": 3.7900521755218506, "learning_rate": 8.96535838521314e-06, "loss": 1.2157, "step": 10720 }, { "epoch": 0.840203761755486, "grad_norm": 3.139432668685913, "learning_rate": 8.965170545608197e-06, "loss": 1.1729, "step": 10721 }, { "epoch": 0.840282131661442, "grad_norm": 3.34102201461792, "learning_rate": 8.964982690921784e-06, "loss": 1.2484, "step": 10722 }, { "epoch": 0.8403605015673982, "grad_norm": 7.107111930847168, "learning_rate": 8.964794821154611e-06, "loss": 0.7164, "step": 10723 }, { "epoch": 0.8404388714733543, "grad_norm": 4.3899359703063965, "learning_rate": 8.964606936307398e-06, "loss": 1.1249, "step": 10724 }, { "epoch": 0.8405172413793104, "grad_norm": 6.0734944343566895, "learning_rate": 8.964419036380853e-06, "loss": 0.3518, "step": 10725 }, { "epoch": 0.8405956112852665, "grad_norm": 2.0161986351013184, "learning_rate": 8.964231121375697e-06, "loss": 1.6054, "step": 10726 }, { "epoch": 0.8406739811912226, "grad_norm": 2.8053228855133057, "learning_rate": 8.964043191292642e-06, "loss": 1.2982, "step": 10727 }, { "epoch": 0.8407523510971787, "grad_norm": 3.0459799766540527, "learning_rate": 8.9638552461324e-06, "loss": 1.511, "step": 10728 }, { "epoch": 0.8408307210031348, "grad_norm": 2.6349246501922607, "learning_rate": 8.96366728589569e-06, "loss": 1.4792, "step": 10729 }, { "epoch": 0.8409090909090909, "grad_norm": 1.708835244178772, "learning_rate": 8.963479310583227e-06, "loss": 0.8798, "step": 10730 }, { "epoch": 0.840987460815047, "grad_norm": 2.6986801624298096, "learning_rate": 8.963291320195723e-06, "loss": 2.0831, "step": 10731 }, { "epoch": 0.8410658307210032, "grad_norm": 2.096109628677368, "learning_rate": 8.963103314733894e-06, "loss": 1.2525, "step": 10732 }, { "epoch": 0.8411442006269593, "grad_norm": 6.069739818572998, "learning_rate": 8.962915294198457e-06, "loss": 1.3155, "step": 10733 }, { "epoch": 0.8412225705329154, "grad_norm": 2.306386947631836, "learning_rate": 8.962727258590125e-06, "loss": 1.4726, "step": 10734 }, { "epoch": 0.8413009404388715, "grad_norm": 4.218695163726807, "learning_rate": 8.962539207909615e-06, "loss": 1.2264, "step": 10735 }, { "epoch": 0.8413793103448276, "grad_norm": 7.9132819175720215, "learning_rate": 8.96235114215764e-06, "loss": 1.7003, "step": 10736 }, { "epoch": 0.8414576802507837, "grad_norm": 2.753774404525757, "learning_rate": 8.962163061334917e-06, "loss": 1.6087, "step": 10737 }, { "epoch": 0.8415360501567398, "grad_norm": 2.193544626235962, "learning_rate": 8.96197496544216e-06, "loss": 1.5397, "step": 10738 }, { "epoch": 0.8416144200626959, "grad_norm": 3.709679365158081, "learning_rate": 8.961786854480086e-06, "loss": 1.3336, "step": 10739 }, { "epoch": 0.841692789968652, "grad_norm": 3.5636510848999023, "learning_rate": 8.961598728449409e-06, "loss": 1.8152, "step": 10740 }, { "epoch": 0.8417711598746082, "grad_norm": 1.9325352907180786, "learning_rate": 8.961410587350844e-06, "loss": 1.266, "step": 10741 }, { "epoch": 0.8418495297805643, "grad_norm": 1.830222487449646, "learning_rate": 8.96122243118511e-06, "loss": 1.558, "step": 10742 }, { "epoch": 0.8419278996865204, "grad_norm": 2.999713659286499, "learning_rate": 8.961034259952922e-06, "loss": 1.215, "step": 10743 }, { "epoch": 0.8420062695924765, "grad_norm": 2.0713586807250977, "learning_rate": 8.96084607365499e-06, "loss": 1.2437, "step": 10744 }, { "epoch": 0.8420846394984326, "grad_norm": 2.7742221355438232, "learning_rate": 8.960657872292035e-06, "loss": 1.4193, "step": 10745 }, { "epoch": 0.8421630094043887, "grad_norm": 2.5762877464294434, "learning_rate": 8.960469655864774e-06, "loss": 1.4213, "step": 10746 }, { "epoch": 0.8422413793103448, "grad_norm": 2.458789587020874, "learning_rate": 8.96028142437392e-06, "loss": 1.4205, "step": 10747 }, { "epoch": 0.8423197492163009, "grad_norm": 2.831669330596924, "learning_rate": 8.960093177820188e-06, "loss": 1.1583, "step": 10748 }, { "epoch": 0.842398119122257, "grad_norm": 4.614530086517334, "learning_rate": 8.959904916204297e-06, "loss": 1.4891, "step": 10749 }, { "epoch": 0.8424764890282131, "grad_norm": 3.028043508529663, "learning_rate": 8.959716639526962e-06, "loss": 1.6251, "step": 10750 }, { "epoch": 0.8425548589341693, "grad_norm": 3.3841960430145264, "learning_rate": 8.959528347788897e-06, "loss": 1.324, "step": 10751 }, { "epoch": 0.8426332288401254, "grad_norm": 6.104304790496826, "learning_rate": 8.959340040990821e-06, "loss": 1.6234, "step": 10752 }, { "epoch": 0.8427115987460815, "grad_norm": 2.721156120300293, "learning_rate": 8.95915171913345e-06, "loss": 1.3897, "step": 10753 }, { "epoch": 0.8427899686520376, "grad_norm": 3.904144525527954, "learning_rate": 8.958963382217497e-06, "loss": 1.323, "step": 10754 }, { "epoch": 0.8428683385579937, "grad_norm": 2.3219635486602783, "learning_rate": 8.958775030243682e-06, "loss": 1.1225, "step": 10755 }, { "epoch": 0.8429467084639498, "grad_norm": 2.203803777694702, "learning_rate": 8.958586663212721e-06, "loss": 1.1753, "step": 10756 }, { "epoch": 0.8430250783699059, "grad_norm": 3.0183212757110596, "learning_rate": 8.95839828112533e-06, "loss": 1.3907, "step": 10757 }, { "epoch": 0.843103448275862, "grad_norm": 2.285127878189087, "learning_rate": 8.958209883982223e-06, "loss": 1.3689, "step": 10758 }, { "epoch": 0.8431818181818181, "grad_norm": 2.0740485191345215, "learning_rate": 8.95802147178412e-06, "loss": 1.5529, "step": 10759 }, { "epoch": 0.8432601880877743, "grad_norm": 3.561025857925415, "learning_rate": 8.957833044531737e-06, "loss": 1.2909, "step": 10760 }, { "epoch": 0.8433385579937304, "grad_norm": 2.8540189266204834, "learning_rate": 8.957644602225788e-06, "loss": 1.6979, "step": 10761 }, { "epoch": 0.8434169278996865, "grad_norm": 2.064690589904785, "learning_rate": 8.957456144866993e-06, "loss": 0.9987, "step": 10762 }, { "epoch": 0.8434952978056426, "grad_norm": 2.596783399581909, "learning_rate": 8.957267672456066e-06, "loss": 1.37, "step": 10763 }, { "epoch": 0.8435736677115987, "grad_norm": 2.541794538497925, "learning_rate": 8.957079184993727e-06, "loss": 1.5756, "step": 10764 }, { "epoch": 0.8436520376175548, "grad_norm": 2.1215476989746094, "learning_rate": 8.956890682480692e-06, "loss": 1.1577, "step": 10765 }, { "epoch": 0.8437304075235109, "grad_norm": 4.358931541442871, "learning_rate": 8.956702164917676e-06, "loss": 1.2972, "step": 10766 }, { "epoch": 0.8438087774294671, "grad_norm": 2.137244701385498, "learning_rate": 8.956513632305398e-06, "loss": 1.2963, "step": 10767 }, { "epoch": 0.8438871473354232, "grad_norm": 2.261336326599121, "learning_rate": 8.956325084644573e-06, "loss": 1.3046, "step": 10768 }, { "epoch": 0.8439655172413794, "grad_norm": 1.8157238960266113, "learning_rate": 8.95613652193592e-06, "loss": 1.0922, "step": 10769 }, { "epoch": 0.8440438871473355, "grad_norm": 3.245910882949829, "learning_rate": 8.955947944180154e-06, "loss": 1.5554, "step": 10770 }, { "epoch": 0.8441222570532916, "grad_norm": 2.4408931732177734, "learning_rate": 8.955759351377997e-06, "loss": 1.3131, "step": 10771 }, { "epoch": 0.8442006269592477, "grad_norm": 4.610380172729492, "learning_rate": 8.955570743530162e-06, "loss": 1.4531, "step": 10772 }, { "epoch": 0.8442789968652038, "grad_norm": 3.9454708099365234, "learning_rate": 8.955382120637368e-06, "loss": 1.1561, "step": 10773 }, { "epoch": 0.8443573667711599, "grad_norm": 2.7698891162872314, "learning_rate": 8.95519348270033e-06, "loss": 1.4685, "step": 10774 }, { "epoch": 0.844435736677116, "grad_norm": 2.190035343170166, "learning_rate": 8.95500482971977e-06, "loss": 1.022, "step": 10775 }, { "epoch": 0.8445141065830721, "grad_norm": 8.57628345489502, "learning_rate": 8.954816161696402e-06, "loss": 0.7195, "step": 10776 }, { "epoch": 0.8445924764890282, "grad_norm": 3.6024930477142334, "learning_rate": 8.954627478630944e-06, "loss": 2.3505, "step": 10777 }, { "epoch": 0.8446708463949844, "grad_norm": 1.8097336292266846, "learning_rate": 8.954438780524116e-06, "loss": 1.2422, "step": 10778 }, { "epoch": 0.8447492163009405, "grad_norm": 3.3523993492126465, "learning_rate": 8.954250067376633e-06, "loss": 1.1861, "step": 10779 }, { "epoch": 0.8448275862068966, "grad_norm": 2.6214356422424316, "learning_rate": 8.954061339189214e-06, "loss": 1.2634, "step": 10780 }, { "epoch": 0.8449059561128527, "grad_norm": 2.2862932682037354, "learning_rate": 8.953872595962576e-06, "loss": 1.4326, "step": 10781 }, { "epoch": 0.8449843260188088, "grad_norm": 3.2627625465393066, "learning_rate": 8.953683837697439e-06, "loss": 1.2286, "step": 10782 }, { "epoch": 0.8450626959247649, "grad_norm": 3.0477302074432373, "learning_rate": 8.953495064394517e-06, "loss": 1.5736, "step": 10783 }, { "epoch": 0.845141065830721, "grad_norm": 13.61754035949707, "learning_rate": 8.953306276054532e-06, "loss": 0.8778, "step": 10784 }, { "epoch": 0.8452194357366771, "grad_norm": 3.174302577972412, "learning_rate": 8.9531174726782e-06, "loss": 1.5247, "step": 10785 }, { "epoch": 0.8452978056426332, "grad_norm": 2.7046360969543457, "learning_rate": 8.95292865426624e-06, "loss": 1.1828, "step": 10786 }, { "epoch": 0.8453761755485893, "grad_norm": 1.8674535751342773, "learning_rate": 8.95273982081937e-06, "loss": 0.9085, "step": 10787 }, { "epoch": 0.8454545454545455, "grad_norm": 5.773401260375977, "learning_rate": 8.952550972338307e-06, "loss": 0.9278, "step": 10788 }, { "epoch": 0.8455329153605016, "grad_norm": 2.6017684936523438, "learning_rate": 8.952362108823771e-06, "loss": 1.3487, "step": 10789 }, { "epoch": 0.8456112852664577, "grad_norm": 3.160658597946167, "learning_rate": 8.95217323027648e-06, "loss": 1.3969, "step": 10790 }, { "epoch": 0.8456896551724138, "grad_norm": 4.905806064605713, "learning_rate": 8.95198433669715e-06, "loss": 1.2598, "step": 10791 }, { "epoch": 0.8457680250783699, "grad_norm": 2.9273977279663086, "learning_rate": 8.951795428086502e-06, "loss": 1.1694, "step": 10792 }, { "epoch": 0.845846394984326, "grad_norm": 2.319146156311035, "learning_rate": 8.951606504445256e-06, "loss": 1.3279, "step": 10793 }, { "epoch": 0.8459247648902821, "grad_norm": 4.3843584060668945, "learning_rate": 8.951417565774128e-06, "loss": 1.0569, "step": 10794 }, { "epoch": 0.8460031347962382, "grad_norm": 2.8233699798583984, "learning_rate": 8.951228612073838e-06, "loss": 1.5135, "step": 10795 }, { "epoch": 0.8460815047021943, "grad_norm": 2.362107753753662, "learning_rate": 8.951039643345101e-06, "loss": 1.4031, "step": 10796 }, { "epoch": 0.8461598746081505, "grad_norm": 3.2923996448516846, "learning_rate": 8.95085065958864e-06, "loss": 0.7969, "step": 10797 }, { "epoch": 0.8462382445141066, "grad_norm": 2.72743821144104, "learning_rate": 8.950661660805172e-06, "loss": 1.3572, "step": 10798 }, { "epoch": 0.8463166144200627, "grad_norm": 4.37115478515625, "learning_rate": 8.950472646995416e-06, "loss": 1.3112, "step": 10799 }, { "epoch": 0.8463949843260188, "grad_norm": 3.48701548576355, "learning_rate": 8.950283618160091e-06, "loss": 1.7829, "step": 10800 }, { "epoch": 0.8464733542319749, "grad_norm": 2.585193157196045, "learning_rate": 8.950094574299917e-06, "loss": 1.0565, "step": 10801 }, { "epoch": 0.846551724137931, "grad_norm": 1.851006031036377, "learning_rate": 8.949905515415612e-06, "loss": 1.5426, "step": 10802 }, { "epoch": 0.8466300940438871, "grad_norm": 2.439418315887451, "learning_rate": 8.949716441507894e-06, "loss": 1.0338, "step": 10803 }, { "epoch": 0.8467084639498432, "grad_norm": 2.652500867843628, "learning_rate": 8.949527352577485e-06, "loss": 1.404, "step": 10804 }, { "epoch": 0.8467868338557993, "grad_norm": 12.616586685180664, "learning_rate": 8.949338248625102e-06, "loss": 1.8649, "step": 10805 }, { "epoch": 0.8468652037617554, "grad_norm": 2.338940143585205, "learning_rate": 8.949149129651467e-06, "loss": 1.3194, "step": 10806 }, { "epoch": 0.8469435736677116, "grad_norm": 3.40828800201416, "learning_rate": 8.948959995657293e-06, "loss": 1.4769, "step": 10807 }, { "epoch": 0.8470219435736677, "grad_norm": 2.544628858566284, "learning_rate": 8.948770846643306e-06, "loss": 1.235, "step": 10808 }, { "epoch": 0.8471003134796238, "grad_norm": 2.8053882122039795, "learning_rate": 8.948581682610222e-06, "loss": 1.4785, "step": 10809 }, { "epoch": 0.8471786833855799, "grad_norm": 1.947373628616333, "learning_rate": 8.948392503558763e-06, "loss": 1.2782, "step": 10810 }, { "epoch": 0.847257053291536, "grad_norm": 2.262592077255249, "learning_rate": 8.948203309489645e-06, "loss": 1.5439, "step": 10811 }, { "epoch": 0.8473354231974921, "grad_norm": 2.267838954925537, "learning_rate": 8.94801410040359e-06, "loss": 1.1562, "step": 10812 }, { "epoch": 0.8474137931034482, "grad_norm": 2.34344482421875, "learning_rate": 8.947824876301317e-06, "loss": 1.0633, "step": 10813 }, { "epoch": 0.8474921630094044, "grad_norm": 2.0270116329193115, "learning_rate": 8.947635637183548e-06, "loss": 1.5868, "step": 10814 }, { "epoch": 0.8475705329153606, "grad_norm": 4.336245059967041, "learning_rate": 8.947446383050997e-06, "loss": 1.5495, "step": 10815 }, { "epoch": 0.8476489028213167, "grad_norm": 9.48274040222168, "learning_rate": 8.94725711390439e-06, "loss": 1.8966, "step": 10816 }, { "epoch": 0.8477272727272728, "grad_norm": 19.931621551513672, "learning_rate": 8.947067829744445e-06, "loss": 1.1311, "step": 10817 }, { "epoch": 0.8478056426332289, "grad_norm": 2.8260183334350586, "learning_rate": 8.946878530571879e-06, "loss": 1.3871, "step": 10818 }, { "epoch": 0.847884012539185, "grad_norm": 2.255584478378296, "learning_rate": 8.946689216387416e-06, "loss": 1.4833, "step": 10819 }, { "epoch": 0.8479623824451411, "grad_norm": 2.0522899627685547, "learning_rate": 8.946499887191774e-06, "loss": 1.3388, "step": 10820 }, { "epoch": 0.8480407523510972, "grad_norm": 2.6873362064361572, "learning_rate": 8.946310542985673e-06, "loss": 1.0762, "step": 10821 }, { "epoch": 0.8481191222570533, "grad_norm": 3.8075222969055176, "learning_rate": 8.946121183769835e-06, "loss": 1.2025, "step": 10822 }, { "epoch": 0.8481974921630094, "grad_norm": 3.1681249141693115, "learning_rate": 8.945931809544977e-06, "loss": 1.0886, "step": 10823 }, { "epoch": 0.8482758620689655, "grad_norm": 4.178367614746094, "learning_rate": 8.945742420311822e-06, "loss": 1.5724, "step": 10824 }, { "epoch": 0.8483542319749217, "grad_norm": 8.054778099060059, "learning_rate": 8.945553016071089e-06, "loss": 1.5986, "step": 10825 }, { "epoch": 0.8484326018808778, "grad_norm": 4.683403491973877, "learning_rate": 8.9453635968235e-06, "loss": 1.5892, "step": 10826 }, { "epoch": 0.8485109717868339, "grad_norm": 2.295549154281616, "learning_rate": 8.945174162569774e-06, "loss": 1.2785, "step": 10827 }, { "epoch": 0.84858934169279, "grad_norm": 2.7957804203033447, "learning_rate": 8.944984713310632e-06, "loss": 1.3458, "step": 10828 }, { "epoch": 0.8486677115987461, "grad_norm": 3.434692859649658, "learning_rate": 8.944795249046793e-06, "loss": 1.4961, "step": 10829 }, { "epoch": 0.8487460815047022, "grad_norm": 4.521580219268799, "learning_rate": 8.94460576977898e-06, "loss": 1.5811, "step": 10830 }, { "epoch": 0.8488244514106583, "grad_norm": 3.138735055923462, "learning_rate": 8.944416275507914e-06, "loss": 1.1894, "step": 10831 }, { "epoch": 0.8489028213166144, "grad_norm": 3.2101006507873535, "learning_rate": 8.944226766234311e-06, "loss": 1.6992, "step": 10832 }, { "epoch": 0.8489811912225705, "grad_norm": 2.0101025104522705, "learning_rate": 8.944037241958899e-06, "loss": 1.438, "step": 10833 }, { "epoch": 0.8490595611285267, "grad_norm": 3.3057096004486084, "learning_rate": 8.943847702682393e-06, "loss": 1.2438, "step": 10834 }, { "epoch": 0.8491379310344828, "grad_norm": 6.089066505432129, "learning_rate": 8.943658148405518e-06, "loss": 2.0878, "step": 10835 }, { "epoch": 0.8492163009404389, "grad_norm": 3.012686252593994, "learning_rate": 8.943468579128992e-06, "loss": 1.4537, "step": 10836 }, { "epoch": 0.849294670846395, "grad_norm": 4.119509220123291, "learning_rate": 8.943278994853537e-06, "loss": 1.4443, "step": 10837 }, { "epoch": 0.8493730407523511, "grad_norm": 3.0366010665893555, "learning_rate": 8.943089395579874e-06, "loss": 1.5454, "step": 10838 }, { "epoch": 0.8494514106583072, "grad_norm": 3.076028347015381, "learning_rate": 8.942899781308723e-06, "loss": 1.4497, "step": 10839 }, { "epoch": 0.8495297805642633, "grad_norm": 3.2363760471343994, "learning_rate": 8.942710152040807e-06, "loss": 1.2583, "step": 10840 }, { "epoch": 0.8496081504702194, "grad_norm": 2.5605475902557373, "learning_rate": 8.942520507776849e-06, "loss": 0.7768, "step": 10841 }, { "epoch": 0.8496865203761755, "grad_norm": 1.8560764789581299, "learning_rate": 8.942330848517563e-06, "loss": 1.1387, "step": 10842 }, { "epoch": 0.8497648902821316, "grad_norm": 1.6922082901000977, "learning_rate": 8.94214117426368e-06, "loss": 1.2893, "step": 10843 }, { "epoch": 0.8498432601880878, "grad_norm": 2.3844995498657227, "learning_rate": 8.941951485015917e-06, "loss": 1.8687, "step": 10844 }, { "epoch": 0.8499216300940439, "grad_norm": 3.6489012241363525, "learning_rate": 8.94176178077499e-06, "loss": 1.5019, "step": 10845 }, { "epoch": 0.85, "grad_norm": 3.384115219116211, "learning_rate": 8.941572061541632e-06, "loss": 1.3963, "step": 10846 }, { "epoch": 0.8500783699059561, "grad_norm": 10.225573539733887, "learning_rate": 8.941382327316554e-06, "loss": 1.1528, "step": 10847 }, { "epoch": 0.8501567398119122, "grad_norm": 4.811491012573242, "learning_rate": 8.941192578100484e-06, "loss": 1.5633, "step": 10848 }, { "epoch": 0.8502351097178683, "grad_norm": 5.804539203643799, "learning_rate": 8.941002813894144e-06, "loss": 1.592, "step": 10849 }, { "epoch": 0.8503134796238244, "grad_norm": 2.8061282634735107, "learning_rate": 8.94081303469825e-06, "loss": 1.1643, "step": 10850 }, { "epoch": 0.8503918495297805, "grad_norm": 3.4971401691436768, "learning_rate": 8.940623240513528e-06, "loss": 1.4094, "step": 10851 }, { "epoch": 0.8504702194357366, "grad_norm": 2.4523372650146484, "learning_rate": 8.9404334313407e-06, "loss": 1.2537, "step": 10852 }, { "epoch": 0.8505485893416928, "grad_norm": 2.947481393814087, "learning_rate": 8.940243607180486e-06, "loss": 1.5153, "step": 10853 }, { "epoch": 0.8506269592476489, "grad_norm": 2.512183904647827, "learning_rate": 8.94005376803361e-06, "loss": 1.4773, "step": 10854 }, { "epoch": 0.850705329153605, "grad_norm": 1.846429467201233, "learning_rate": 8.939863913900794e-06, "loss": 1.1519, "step": 10855 }, { "epoch": 0.8507836990595611, "grad_norm": 4.114955902099609, "learning_rate": 8.939674044782758e-06, "loss": 1.4425, "step": 10856 }, { "epoch": 0.8508620689655172, "grad_norm": 3.0837700366973877, "learning_rate": 8.939484160680227e-06, "loss": 1.2409, "step": 10857 }, { "epoch": 0.8509404388714733, "grad_norm": 3.696439266204834, "learning_rate": 8.939294261593921e-06, "loss": 1.3441, "step": 10858 }, { "epoch": 0.8510188087774294, "grad_norm": 1.9718180894851685, "learning_rate": 8.93910434752456e-06, "loss": 1.4507, "step": 10859 }, { "epoch": 0.8510971786833855, "grad_norm": 2.1819138526916504, "learning_rate": 8.938914418472872e-06, "loss": 0.7422, "step": 10860 }, { "epoch": 0.8511755485893417, "grad_norm": 7.788959980010986, "learning_rate": 8.938724474439578e-06, "loss": 1.6429, "step": 10861 }, { "epoch": 0.8512539184952979, "grad_norm": 3.276313304901123, "learning_rate": 8.938534515425399e-06, "loss": 1.0538, "step": 10862 }, { "epoch": 0.851332288401254, "grad_norm": 2.737257242202759, "learning_rate": 8.938344541431057e-06, "loss": 1.3276, "step": 10863 }, { "epoch": 0.8514106583072101, "grad_norm": 2.2967686653137207, "learning_rate": 8.938154552457273e-06, "loss": 1.4606, "step": 10864 }, { "epoch": 0.8514890282131662, "grad_norm": 2.7317802906036377, "learning_rate": 8.937964548504774e-06, "loss": 1.5352, "step": 10865 }, { "epoch": 0.8515673981191223, "grad_norm": 3.1654324531555176, "learning_rate": 8.93777452957428e-06, "loss": 1.3991, "step": 10866 }, { "epoch": 0.8516457680250784, "grad_norm": 2.106182336807251, "learning_rate": 8.937584495666513e-06, "loss": 1.1025, "step": 10867 }, { "epoch": 0.8517241379310345, "grad_norm": 2.728203535079956, "learning_rate": 8.937394446782199e-06, "loss": 1.4375, "step": 10868 }, { "epoch": 0.8518025078369906, "grad_norm": 2.2950682640075684, "learning_rate": 8.93720438292206e-06, "loss": 1.3798, "step": 10869 }, { "epoch": 0.8518808777429467, "grad_norm": 4.792779445648193, "learning_rate": 8.937014304086814e-06, "loss": 1.403, "step": 10870 }, { "epoch": 0.8519592476489029, "grad_norm": 3.7078359127044678, "learning_rate": 8.93682421027719e-06, "loss": 1.6324, "step": 10871 }, { "epoch": 0.852037617554859, "grad_norm": 3.7624940872192383, "learning_rate": 8.93663410149391e-06, "loss": 1.24, "step": 10872 }, { "epoch": 0.8521159874608151, "grad_norm": 3.172133207321167, "learning_rate": 8.936443977737694e-06, "loss": 1.4609, "step": 10873 }, { "epoch": 0.8521943573667712, "grad_norm": 4.755229949951172, "learning_rate": 8.936253839009268e-06, "loss": 1.3782, "step": 10874 }, { "epoch": 0.8522727272727273, "grad_norm": 4.373086452484131, "learning_rate": 8.936063685309353e-06, "loss": 1.1281, "step": 10875 }, { "epoch": 0.8523510971786834, "grad_norm": 2.744224786758423, "learning_rate": 8.935873516638674e-06, "loss": 1.4071, "step": 10876 }, { "epoch": 0.8524294670846395, "grad_norm": 7.372638702392578, "learning_rate": 8.935683332997956e-06, "loss": 1.0961, "step": 10877 }, { "epoch": 0.8525078369905956, "grad_norm": 4.40244722366333, "learning_rate": 8.935493134387917e-06, "loss": 1.7262, "step": 10878 }, { "epoch": 0.8525862068965517, "grad_norm": 2.6628682613372803, "learning_rate": 8.935302920809283e-06, "loss": 1.163, "step": 10879 }, { "epoch": 0.8526645768025078, "grad_norm": 12.726116180419922, "learning_rate": 8.93511269226278e-06, "loss": 0.9167, "step": 10880 }, { "epoch": 0.852742946708464, "grad_norm": 2.6684093475341797, "learning_rate": 8.93492244874913e-06, "loss": 1.6746, "step": 10881 }, { "epoch": 0.8528213166144201, "grad_norm": 3.3984830379486084, "learning_rate": 8.934732190269054e-06, "loss": 1.075, "step": 10882 }, { "epoch": 0.8528996865203762, "grad_norm": 2.7402281761169434, "learning_rate": 8.93454191682328e-06, "loss": 1.4798, "step": 10883 }, { "epoch": 0.8529780564263323, "grad_norm": 2.8324270248413086, "learning_rate": 8.934351628412527e-06, "loss": 1.4574, "step": 10884 }, { "epoch": 0.8530564263322884, "grad_norm": 3.812678337097168, "learning_rate": 8.934161325037523e-06, "loss": 2.0953, "step": 10885 }, { "epoch": 0.8531347962382445, "grad_norm": 3.659043788909912, "learning_rate": 8.933971006698989e-06, "loss": 1.5344, "step": 10886 }, { "epoch": 0.8532131661442006, "grad_norm": 2.1766703128814697, "learning_rate": 8.93378067339765e-06, "loss": 1.0038, "step": 10887 }, { "epoch": 0.8532915360501567, "grad_norm": 2.091078042984009, "learning_rate": 8.933590325134232e-06, "loss": 1.3832, "step": 10888 }, { "epoch": 0.8533699059561128, "grad_norm": 2.63254714012146, "learning_rate": 8.933399961909455e-06, "loss": 1.7682, "step": 10889 }, { "epoch": 0.853448275862069, "grad_norm": 3.668330192565918, "learning_rate": 8.933209583724045e-06, "loss": 1.9038, "step": 10890 }, { "epoch": 0.8535266457680251, "grad_norm": 2.0767199993133545, "learning_rate": 8.933019190578726e-06, "loss": 1.4476, "step": 10891 }, { "epoch": 0.8536050156739812, "grad_norm": 2.5143415927886963, "learning_rate": 8.93282878247422e-06, "loss": 1.3679, "step": 10892 }, { "epoch": 0.8536833855799373, "grad_norm": 2.966535806655884, "learning_rate": 8.932638359411257e-06, "loss": 1.2422, "step": 10893 }, { "epoch": 0.8537617554858934, "grad_norm": 1.9312113523483276, "learning_rate": 8.932447921390556e-06, "loss": 1.6597, "step": 10894 }, { "epoch": 0.8538401253918495, "grad_norm": 2.884382963180542, "learning_rate": 8.932257468412843e-06, "loss": 1.3831, "step": 10895 }, { "epoch": 0.8539184952978056, "grad_norm": 1.9816875457763672, "learning_rate": 8.932067000478843e-06, "loss": 1.3631, "step": 10896 }, { "epoch": 0.8539968652037617, "grad_norm": 3.765277862548828, "learning_rate": 8.931876517589277e-06, "loss": 1.6265, "step": 10897 }, { "epoch": 0.8540752351097178, "grad_norm": 2.623579263687134, "learning_rate": 8.931686019744875e-06, "loss": 1.0942, "step": 10898 }, { "epoch": 0.854153605015674, "grad_norm": 3.678767204284668, "learning_rate": 8.931495506946358e-06, "loss": 1.6792, "step": 10899 }, { "epoch": 0.85423197492163, "grad_norm": 6.4811482429504395, "learning_rate": 8.931304979194452e-06, "loss": 1.0818, "step": 10900 }, { "epoch": 0.8543103448275862, "grad_norm": 3.0621025562286377, "learning_rate": 8.931114436489881e-06, "loss": 1.3037, "step": 10901 }, { "epoch": 0.8543887147335423, "grad_norm": 2.8627512454986572, "learning_rate": 8.93092387883337e-06, "loss": 1.125, "step": 10902 }, { "epoch": 0.8544670846394984, "grad_norm": 2.5113542079925537, "learning_rate": 8.930733306225642e-06, "loss": 0.9113, "step": 10903 }, { "epoch": 0.8545454545454545, "grad_norm": 2.583869218826294, "learning_rate": 8.930542718667425e-06, "loss": 1.2352, "step": 10904 }, { "epoch": 0.8546238244514106, "grad_norm": 2.2010605335235596, "learning_rate": 8.930352116159443e-06, "loss": 1.1958, "step": 10905 }, { "epoch": 0.8547021943573667, "grad_norm": 2.5716090202331543, "learning_rate": 8.93016149870242e-06, "loss": 1.2793, "step": 10906 }, { "epoch": 0.8547805642633229, "grad_norm": 2.9319605827331543, "learning_rate": 8.929970866297079e-06, "loss": 1.2641, "step": 10907 }, { "epoch": 0.854858934169279, "grad_norm": 4.613038063049316, "learning_rate": 8.929780218944148e-06, "loss": 1.3059, "step": 10908 }, { "epoch": 0.8549373040752352, "grad_norm": 5.253439426422119, "learning_rate": 8.929589556644354e-06, "loss": 2.0217, "step": 10909 }, { "epoch": 0.8550156739811913, "grad_norm": 7.429105758666992, "learning_rate": 8.929398879398417e-06, "loss": 1.7295, "step": 10910 }, { "epoch": 0.8550940438871474, "grad_norm": 3.044874906539917, "learning_rate": 8.929208187207066e-06, "loss": 1.3373, "step": 10911 }, { "epoch": 0.8551724137931035, "grad_norm": 4.320148468017578, "learning_rate": 8.929017480071026e-06, "loss": 1.3809, "step": 10912 }, { "epoch": 0.8552507836990596, "grad_norm": 3.7751872539520264, "learning_rate": 8.92882675799102e-06, "loss": 1.0928, "step": 10913 }, { "epoch": 0.8553291536050157, "grad_norm": 2.619058609008789, "learning_rate": 8.928636020967775e-06, "loss": 1.5648, "step": 10914 }, { "epoch": 0.8554075235109718, "grad_norm": 4.496121406555176, "learning_rate": 8.92844526900202e-06, "loss": 0.9052, "step": 10915 }, { "epoch": 0.8554858934169279, "grad_norm": 3.692650556564331, "learning_rate": 8.928254502094473e-06, "loss": 1.2035, "step": 10916 }, { "epoch": 0.855564263322884, "grad_norm": 4.608553409576416, "learning_rate": 8.928063720245865e-06, "loss": 1.367, "step": 10917 }, { "epoch": 0.8556426332288402, "grad_norm": 2.8802084922790527, "learning_rate": 8.92787292345692e-06, "loss": 1.4948, "step": 10918 }, { "epoch": 0.8557210031347963, "grad_norm": 3.3003437519073486, "learning_rate": 8.927682111728364e-06, "loss": 1.6067, "step": 10919 }, { "epoch": 0.8557993730407524, "grad_norm": 3.7815418243408203, "learning_rate": 8.927491285060923e-06, "loss": 1.3879, "step": 10920 }, { "epoch": 0.8558777429467085, "grad_norm": 2.8434226512908936, "learning_rate": 8.927300443455321e-06, "loss": 1.2979, "step": 10921 }, { "epoch": 0.8559561128526646, "grad_norm": 3.401484966278076, "learning_rate": 8.927109586912288e-06, "loss": 1.9505, "step": 10922 }, { "epoch": 0.8560344827586207, "grad_norm": 5.858608245849609, "learning_rate": 8.926918715432544e-06, "loss": 2.4511, "step": 10923 }, { "epoch": 0.8561128526645768, "grad_norm": 2.1115763187408447, "learning_rate": 8.92672782901682e-06, "loss": 1.2098, "step": 10924 }, { "epoch": 0.8561912225705329, "grad_norm": 3.341573476791382, "learning_rate": 8.92653692766584e-06, "loss": 1.8081, "step": 10925 }, { "epoch": 0.856269592476489, "grad_norm": 1.8466228246688843, "learning_rate": 8.92634601138033e-06, "loss": 1.3434, "step": 10926 }, { "epoch": 0.8563479623824451, "grad_norm": 2.447213649749756, "learning_rate": 8.926155080161016e-06, "loss": 1.2749, "step": 10927 }, { "epoch": 0.8564263322884013, "grad_norm": 2.0796070098876953, "learning_rate": 8.925964134008627e-06, "loss": 1.3531, "step": 10928 }, { "epoch": 0.8565047021943574, "grad_norm": 5.0432353019714355, "learning_rate": 8.925773172923883e-06, "loss": 1.8275, "step": 10929 }, { "epoch": 0.8565830721003135, "grad_norm": 3.1433212757110596, "learning_rate": 8.925582196907519e-06, "loss": 1.6555, "step": 10930 }, { "epoch": 0.8566614420062696, "grad_norm": 2.9207754135131836, "learning_rate": 8.925391205960254e-06, "loss": 1.0971, "step": 10931 }, { "epoch": 0.8567398119122257, "grad_norm": 2.264596939086914, "learning_rate": 8.925200200082818e-06, "loss": 1.2803, "step": 10932 }, { "epoch": 0.8568181818181818, "grad_norm": 2.0226056575775146, "learning_rate": 8.925009179275934e-06, "loss": 0.9202, "step": 10933 }, { "epoch": 0.8568965517241379, "grad_norm": 4.091192245483398, "learning_rate": 8.924818143540333e-06, "loss": 1.3252, "step": 10934 }, { "epoch": 0.856974921630094, "grad_norm": 3.9375483989715576, "learning_rate": 8.92462709287674e-06, "loss": 0.8119, "step": 10935 }, { "epoch": 0.8570532915360501, "grad_norm": 4.067739963531494, "learning_rate": 8.924436027285881e-06, "loss": 1.5664, "step": 10936 }, { "epoch": 0.8571316614420063, "grad_norm": 3.891019821166992, "learning_rate": 8.924244946768484e-06, "loss": 0.9962, "step": 10937 }, { "epoch": 0.8572100313479624, "grad_norm": 2.879884719848633, "learning_rate": 8.924053851325274e-06, "loss": 1.5604, "step": 10938 }, { "epoch": 0.8572884012539185, "grad_norm": 2.678835868835449, "learning_rate": 8.923862740956977e-06, "loss": 1.8296, "step": 10939 }, { "epoch": 0.8573667711598746, "grad_norm": 2.594545841217041, "learning_rate": 8.923671615664324e-06, "loss": 1.2524, "step": 10940 }, { "epoch": 0.8574451410658307, "grad_norm": 2.0763537883758545, "learning_rate": 8.923480475448039e-06, "loss": 1.4705, "step": 10941 }, { "epoch": 0.8575235109717868, "grad_norm": 2.609992027282715, "learning_rate": 8.92328932030885e-06, "loss": 1.543, "step": 10942 }, { "epoch": 0.8576018808777429, "grad_norm": 3.394232988357544, "learning_rate": 8.92309815024748e-06, "loss": 1.376, "step": 10943 }, { "epoch": 0.857680250783699, "grad_norm": 3.378844976425171, "learning_rate": 8.922906965264664e-06, "loss": 1.2925, "step": 10944 }, { "epoch": 0.8577586206896551, "grad_norm": 2.420346260070801, "learning_rate": 8.922715765361124e-06, "loss": 1.4839, "step": 10945 }, { "epoch": 0.8578369905956112, "grad_norm": 7.654642581939697, "learning_rate": 8.922524550537586e-06, "loss": 0.823, "step": 10946 }, { "epoch": 0.8579153605015674, "grad_norm": 3.9853782653808594, "learning_rate": 8.92233332079478e-06, "loss": 1.3567, "step": 10947 }, { "epoch": 0.8579937304075235, "grad_norm": 2.3213491439819336, "learning_rate": 8.922142076133433e-06, "loss": 1.5051, "step": 10948 }, { "epoch": 0.8580721003134796, "grad_norm": 2.7297685146331787, "learning_rate": 8.921950816554273e-06, "loss": 1.3918, "step": 10949 }, { "epoch": 0.8581504702194357, "grad_norm": 1.89362633228302, "learning_rate": 8.921759542058026e-06, "loss": 1.0325, "step": 10950 }, { "epoch": 0.8582288401253918, "grad_norm": 2.900799036026001, "learning_rate": 8.92156825264542e-06, "loss": 1.1988, "step": 10951 }, { "epoch": 0.8583072100313479, "grad_norm": 2.8944387435913086, "learning_rate": 8.921376948317181e-06, "loss": 1.1634, "step": 10952 }, { "epoch": 0.858385579937304, "grad_norm": 3.1838912963867188, "learning_rate": 8.92118562907404e-06, "loss": 1.7739, "step": 10953 }, { "epoch": 0.8584639498432602, "grad_norm": 6.853852272033691, "learning_rate": 8.920994294916723e-06, "loss": 1.6403, "step": 10954 }, { "epoch": 0.8585423197492164, "grad_norm": 2.1508078575134277, "learning_rate": 8.920802945845959e-06, "loss": 1.0082, "step": 10955 }, { "epoch": 0.8586206896551725, "grad_norm": 2.408656120300293, "learning_rate": 8.920611581862472e-06, "loss": 1.5362, "step": 10956 }, { "epoch": 0.8586990595611286, "grad_norm": 2.9849021434783936, "learning_rate": 8.920420202966994e-06, "loss": 1.4324, "step": 10957 }, { "epoch": 0.8587774294670847, "grad_norm": 3.49998140335083, "learning_rate": 8.920228809160249e-06, "loss": 1.2134, "step": 10958 }, { "epoch": 0.8588557993730408, "grad_norm": 2.5623576641082764, "learning_rate": 8.92003740044297e-06, "loss": 1.3829, "step": 10959 }, { "epoch": 0.8589341692789969, "grad_norm": 3.032240152359009, "learning_rate": 8.91984597681588e-06, "loss": 1.6839, "step": 10960 }, { "epoch": 0.859012539184953, "grad_norm": 2.914128065109253, "learning_rate": 8.91965453827971e-06, "loss": 1.4357, "step": 10961 }, { "epoch": 0.8590909090909091, "grad_norm": 4.394285202026367, "learning_rate": 8.919463084835187e-06, "loss": 1.2278, "step": 10962 }, { "epoch": 0.8591692789968652, "grad_norm": 2.953917980194092, "learning_rate": 8.91927161648304e-06, "loss": 1.0922, "step": 10963 }, { "epoch": 0.8592476489028213, "grad_norm": 2.505176067352295, "learning_rate": 8.919080133223996e-06, "loss": 1.2346, "step": 10964 }, { "epoch": 0.8593260188087775, "grad_norm": 9.271831512451172, "learning_rate": 8.918888635058786e-06, "loss": 1.1702, "step": 10965 }, { "epoch": 0.8594043887147336, "grad_norm": 3.177164316177368, "learning_rate": 8.918697121988135e-06, "loss": 1.1767, "step": 10966 }, { "epoch": 0.8594827586206897, "grad_norm": 6.450244903564453, "learning_rate": 8.918505594012774e-06, "loss": 1.1092, "step": 10967 }, { "epoch": 0.8595611285266458, "grad_norm": 2.8306548595428467, "learning_rate": 8.918314051133428e-06, "loss": 1.5253, "step": 10968 }, { "epoch": 0.8596394984326019, "grad_norm": 1.79981529712677, "learning_rate": 8.918122493350832e-06, "loss": 1.3169, "step": 10969 }, { "epoch": 0.859717868338558, "grad_norm": 3.2322781085968018, "learning_rate": 8.917930920665708e-06, "loss": 1.4154, "step": 10970 }, { "epoch": 0.8597962382445141, "grad_norm": 1.8969844579696655, "learning_rate": 8.917739333078786e-06, "loss": 1.3546, "step": 10971 }, { "epoch": 0.8598746081504702, "grad_norm": 4.148016929626465, "learning_rate": 8.917547730590797e-06, "loss": 1.1894, "step": 10972 }, { "epoch": 0.8599529780564263, "grad_norm": 2.8977465629577637, "learning_rate": 8.917356113202469e-06, "loss": 1.8533, "step": 10973 }, { "epoch": 0.8600313479623825, "grad_norm": 2.6065261363983154, "learning_rate": 8.917164480914531e-06, "loss": 1.5173, "step": 10974 }, { "epoch": 0.8601097178683386, "grad_norm": 2.270364284515381, "learning_rate": 8.91697283372771e-06, "loss": 1.5143, "step": 10975 }, { "epoch": 0.8601880877742947, "grad_norm": 2.3187899589538574, "learning_rate": 8.916781171642737e-06, "loss": 1.0898, "step": 10976 }, { "epoch": 0.8602664576802508, "grad_norm": 3.4176576137542725, "learning_rate": 8.91658949466034e-06, "loss": 1.5893, "step": 10977 }, { "epoch": 0.8603448275862069, "grad_norm": 1.8733824491500854, "learning_rate": 8.916397802781247e-06, "loss": 1.5006, "step": 10978 }, { "epoch": 0.860423197492163, "grad_norm": 2.6391971111297607, "learning_rate": 8.91620609600619e-06, "loss": 1.4118, "step": 10979 }, { "epoch": 0.8605015673981191, "grad_norm": 4.1775689125061035, "learning_rate": 8.916014374335895e-06, "loss": 1.7671, "step": 10980 }, { "epoch": 0.8605799373040752, "grad_norm": 4.84713888168335, "learning_rate": 8.915822637771095e-06, "loss": 1.5873, "step": 10981 }, { "epoch": 0.8606583072100313, "grad_norm": 3.0233254432678223, "learning_rate": 8.915630886312515e-06, "loss": 1.7251, "step": 10982 }, { "epoch": 0.8607366771159874, "grad_norm": 10.627058029174805, "learning_rate": 8.915439119960886e-06, "loss": 1.1226, "step": 10983 }, { "epoch": 0.8608150470219436, "grad_norm": 3.2708816528320312, "learning_rate": 8.91524733871694e-06, "loss": 1.164, "step": 10984 }, { "epoch": 0.8608934169278997, "grad_norm": 3.613494396209717, "learning_rate": 8.915055542581404e-06, "loss": 1.9013, "step": 10985 }, { "epoch": 0.8609717868338558, "grad_norm": 2.53774356842041, "learning_rate": 8.914863731555006e-06, "loss": 1.3875, "step": 10986 }, { "epoch": 0.8610501567398119, "grad_norm": 4.760189533233643, "learning_rate": 8.914671905638476e-06, "loss": 1.8046, "step": 10987 }, { "epoch": 0.861128526645768, "grad_norm": 2.524482250213623, "learning_rate": 8.914480064832547e-06, "loss": 1.1247, "step": 10988 }, { "epoch": 0.8612068965517241, "grad_norm": 2.020817995071411, "learning_rate": 8.914288209137946e-06, "loss": 0.7995, "step": 10989 }, { "epoch": 0.8612852664576802, "grad_norm": 2.1380257606506348, "learning_rate": 8.914096338555402e-06, "loss": 1.3495, "step": 10990 }, { "epoch": 0.8613636363636363, "grad_norm": 2.3099911212921143, "learning_rate": 8.913904453085647e-06, "loss": 1.2822, "step": 10991 }, { "epoch": 0.8614420062695924, "grad_norm": 1.4383211135864258, "learning_rate": 8.91371255272941e-06, "loss": 1.0547, "step": 10992 }, { "epoch": 0.8615203761755486, "grad_norm": 3.3636982440948486, "learning_rate": 8.913520637487421e-06, "loss": 1.5983, "step": 10993 }, { "epoch": 0.8615987460815047, "grad_norm": 2.7059202194213867, "learning_rate": 8.913328707360409e-06, "loss": 1.7476, "step": 10994 }, { "epoch": 0.8616771159874608, "grad_norm": 3.70198917388916, "learning_rate": 8.913136762349103e-06, "loss": 1.0811, "step": 10995 }, { "epoch": 0.8617554858934169, "grad_norm": 2.6860413551330566, "learning_rate": 8.912944802454238e-06, "loss": 1.2645, "step": 10996 }, { "epoch": 0.861833855799373, "grad_norm": 2.3059654235839844, "learning_rate": 8.91275282767654e-06, "loss": 1.1694, "step": 10997 }, { "epoch": 0.8619122257053291, "grad_norm": 5.081080913543701, "learning_rate": 8.912560838016738e-06, "loss": 1.1208, "step": 10998 }, { "epoch": 0.8619905956112852, "grad_norm": 2.962129592895508, "learning_rate": 8.912368833475564e-06, "loss": 1.6832, "step": 10999 }, { "epoch": 0.8620689655172413, "grad_norm": 2.770085334777832, "learning_rate": 8.91217681405375e-06, "loss": 1.278, "step": 11000 }, { "epoch": 0.8621473354231975, "grad_norm": 1.779097557067871, "learning_rate": 8.911984779752026e-06, "loss": 1.6541, "step": 11001 }, { "epoch": 0.8622257053291537, "grad_norm": 3.855358600616455, "learning_rate": 8.911792730571118e-06, "loss": 1.441, "step": 11002 }, { "epoch": 0.8623040752351098, "grad_norm": 3.6174168586730957, "learning_rate": 8.911600666511762e-06, "loss": 1.3769, "step": 11003 }, { "epoch": 0.8623824451410659, "grad_norm": 3.476865530014038, "learning_rate": 8.911408587574686e-06, "loss": 1.2545, "step": 11004 }, { "epoch": 0.862460815047022, "grad_norm": 9.597334861755371, "learning_rate": 8.911216493760619e-06, "loss": 1.743, "step": 11005 }, { "epoch": 0.8625391849529781, "grad_norm": 2.1424612998962402, "learning_rate": 8.911024385070294e-06, "loss": 1.4118, "step": 11006 }, { "epoch": 0.8626175548589342, "grad_norm": 2.613910675048828, "learning_rate": 8.910832261504443e-06, "loss": 1.3772, "step": 11007 }, { "epoch": 0.8626959247648903, "grad_norm": 6.748636245727539, "learning_rate": 8.910640123063792e-06, "loss": 1.5122, "step": 11008 }, { "epoch": 0.8627742946708464, "grad_norm": 2.7538530826568604, "learning_rate": 8.910447969749077e-06, "loss": 1.3035, "step": 11009 }, { "epoch": 0.8628526645768025, "grad_norm": 2.5984835624694824, "learning_rate": 8.910255801561026e-06, "loss": 1.368, "step": 11010 }, { "epoch": 0.8629310344827587, "grad_norm": 2.3081510066986084, "learning_rate": 8.91006361850037e-06, "loss": 1.4482, "step": 11011 }, { "epoch": 0.8630094043887148, "grad_norm": 3.026183843612671, "learning_rate": 8.90987142056784e-06, "loss": 1.1667, "step": 11012 }, { "epoch": 0.8630877742946709, "grad_norm": 4.68339729309082, "learning_rate": 8.909679207764168e-06, "loss": 1.3705, "step": 11013 }, { "epoch": 0.863166144200627, "grad_norm": 2.5040202140808105, "learning_rate": 8.909486980090083e-06, "loss": 1.3809, "step": 11014 }, { "epoch": 0.8632445141065831, "grad_norm": 2.9084222316741943, "learning_rate": 8.909294737546319e-06, "loss": 1.4427, "step": 11015 }, { "epoch": 0.8633228840125392, "grad_norm": 2.3022289276123047, "learning_rate": 8.909102480133607e-06, "loss": 0.8561, "step": 11016 }, { "epoch": 0.8634012539184953, "grad_norm": 2.621066093444824, "learning_rate": 8.908910207852674e-06, "loss": 1.4354, "step": 11017 }, { "epoch": 0.8634796238244514, "grad_norm": 4.058507919311523, "learning_rate": 8.908717920704256e-06, "loss": 1.3687, "step": 11018 }, { "epoch": 0.8635579937304075, "grad_norm": 2.8484442234039307, "learning_rate": 8.908525618689083e-06, "loss": 1.3898, "step": 11019 }, { "epoch": 0.8636363636363636, "grad_norm": 2.203483819961548, "learning_rate": 8.908333301807886e-06, "loss": 1.3128, "step": 11020 }, { "epoch": 0.8637147335423198, "grad_norm": 2.487551212310791, "learning_rate": 8.908140970061396e-06, "loss": 1.4288, "step": 11021 }, { "epoch": 0.8637931034482759, "grad_norm": 4.4435200691223145, "learning_rate": 8.907948623450343e-06, "loss": 1.5813, "step": 11022 }, { "epoch": 0.863871473354232, "grad_norm": 2.610281467437744, "learning_rate": 8.907756261975464e-06, "loss": 1.2674, "step": 11023 }, { "epoch": 0.8639498432601881, "grad_norm": 3.1218888759613037, "learning_rate": 8.907563885637487e-06, "loss": 1.2486, "step": 11024 }, { "epoch": 0.8640282131661442, "grad_norm": 7.330262184143066, "learning_rate": 8.907371494437142e-06, "loss": 1.8193, "step": 11025 }, { "epoch": 0.8641065830721003, "grad_norm": 12.28654670715332, "learning_rate": 8.907179088375165e-06, "loss": 0.9443, "step": 11026 }, { "epoch": 0.8641849529780564, "grad_norm": 3.542728900909424, "learning_rate": 8.906986667452284e-06, "loss": 1.3712, "step": 11027 }, { "epoch": 0.8642633228840125, "grad_norm": 4.645057678222656, "learning_rate": 8.906794231669233e-06, "loss": 2.0198, "step": 11028 }, { "epoch": 0.8643416927899686, "grad_norm": 2.3475875854492188, "learning_rate": 8.906601781026744e-06, "loss": 1.1407, "step": 11029 }, { "epoch": 0.8644200626959248, "grad_norm": 2.0641870498657227, "learning_rate": 8.906409315525548e-06, "loss": 1.4439, "step": 11030 }, { "epoch": 0.8644984326018809, "grad_norm": 2.5778610706329346, "learning_rate": 8.906216835166376e-06, "loss": 1.6121, "step": 11031 }, { "epoch": 0.864576802507837, "grad_norm": 6.319962024688721, "learning_rate": 8.906024339949963e-06, "loss": 1.1169, "step": 11032 }, { "epoch": 0.8646551724137931, "grad_norm": 5.628327369689941, "learning_rate": 8.90583182987704e-06, "loss": 1.801, "step": 11033 }, { "epoch": 0.8647335423197492, "grad_norm": 2.109067916870117, "learning_rate": 8.905639304948339e-06, "loss": 1.3146, "step": 11034 }, { "epoch": 0.8648119122257053, "grad_norm": 2.0772016048431396, "learning_rate": 8.90544676516459e-06, "loss": 1.4155, "step": 11035 }, { "epoch": 0.8648902821316614, "grad_norm": 3.499725580215454, "learning_rate": 8.90525421052653e-06, "loss": 1.8142, "step": 11036 }, { "epoch": 0.8649686520376175, "grad_norm": 2.9766902923583984, "learning_rate": 8.905061641034887e-06, "loss": 1.2563, "step": 11037 }, { "epoch": 0.8650470219435736, "grad_norm": 3.058265209197998, "learning_rate": 8.904869056690398e-06, "loss": 1.4215, "step": 11038 }, { "epoch": 0.8651253918495297, "grad_norm": 2.9794023036956787, "learning_rate": 8.90467645749379e-06, "loss": 1.1311, "step": 11039 }, { "epoch": 0.8652037617554859, "grad_norm": 2.0215096473693848, "learning_rate": 8.9044838434458e-06, "loss": 1.2958, "step": 11040 }, { "epoch": 0.865282131661442, "grad_norm": 3.9098241329193115, "learning_rate": 8.904291214547156e-06, "loss": 1.1724, "step": 11041 }, { "epoch": 0.8653605015673981, "grad_norm": 2.5083086490631104, "learning_rate": 8.904098570798596e-06, "loss": 1.5866, "step": 11042 }, { "epoch": 0.8654388714733542, "grad_norm": 2.238292694091797, "learning_rate": 8.903905912200852e-06, "loss": 1.8196, "step": 11043 }, { "epoch": 0.8655172413793103, "grad_norm": 4.7111053466796875, "learning_rate": 8.903713238754652e-06, "loss": 0.8931, "step": 11044 }, { "epoch": 0.8655956112852664, "grad_norm": 2.4848222732543945, "learning_rate": 8.903520550460734e-06, "loss": 1.446, "step": 11045 }, { "epoch": 0.8656739811912225, "grad_norm": 2.5122456550598145, "learning_rate": 8.903327847319828e-06, "loss": 1.3635, "step": 11046 }, { "epoch": 0.8657523510971786, "grad_norm": 4.017462730407715, "learning_rate": 8.903135129332666e-06, "loss": 1.31, "step": 11047 }, { "epoch": 0.8658307210031349, "grad_norm": 2.3860924243927, "learning_rate": 8.902942396499986e-06, "loss": 1.5053, "step": 11048 }, { "epoch": 0.865909090909091, "grad_norm": 3.2981011867523193, "learning_rate": 8.902749648822516e-06, "loss": 1.2559, "step": 11049 }, { "epoch": 0.8659874608150471, "grad_norm": 2.3616459369659424, "learning_rate": 8.90255688630099e-06, "loss": 1.267, "step": 11050 }, { "epoch": 0.8660658307210032, "grad_norm": 2.2464847564697266, "learning_rate": 8.902364108936143e-06, "loss": 1.1347, "step": 11051 }, { "epoch": 0.8661442006269593, "grad_norm": 5.603796005249023, "learning_rate": 8.902171316728707e-06, "loss": 1.6915, "step": 11052 }, { "epoch": 0.8662225705329154, "grad_norm": 2.8879828453063965, "learning_rate": 8.901978509679416e-06, "loss": 1.5482, "step": 11053 }, { "epoch": 0.8663009404388715, "grad_norm": 2.268131971359253, "learning_rate": 8.901785687789002e-06, "loss": 1.4071, "step": 11054 }, { "epoch": 0.8663793103448276, "grad_norm": 2.2570765018463135, "learning_rate": 8.901592851058201e-06, "loss": 1.1499, "step": 11055 }, { "epoch": 0.8664576802507837, "grad_norm": 2.882413625717163, "learning_rate": 8.901399999487744e-06, "loss": 1.1779, "step": 11056 }, { "epoch": 0.8665360501567398, "grad_norm": 4.700246334075928, "learning_rate": 8.901207133078366e-06, "loss": 1.322, "step": 11057 }, { "epoch": 0.866614420062696, "grad_norm": 2.3889551162719727, "learning_rate": 8.9010142518308e-06, "loss": 1.5239, "step": 11058 }, { "epoch": 0.8666927899686521, "grad_norm": 2.1039609909057617, "learning_rate": 8.900821355745778e-06, "loss": 1.5126, "step": 11059 }, { "epoch": 0.8667711598746082, "grad_norm": 2.1103203296661377, "learning_rate": 8.900628444824035e-06, "loss": 1.4174, "step": 11060 }, { "epoch": 0.8668495297805643, "grad_norm": 7.765743255615234, "learning_rate": 8.900435519066306e-06, "loss": 1.2996, "step": 11061 }, { "epoch": 0.8669278996865204, "grad_norm": 3.613236665725708, "learning_rate": 8.900242578473325e-06, "loss": 1.3662, "step": 11062 }, { "epoch": 0.8670062695924765, "grad_norm": 2.849116086959839, "learning_rate": 8.900049623045822e-06, "loss": 1.4706, "step": 11063 }, { "epoch": 0.8670846394984326, "grad_norm": 2.1809091567993164, "learning_rate": 8.899856652784534e-06, "loss": 1.5624, "step": 11064 }, { "epoch": 0.8671630094043887, "grad_norm": 2.927074432373047, "learning_rate": 8.899663667690195e-06, "loss": 1.8322, "step": 11065 }, { "epoch": 0.8672413793103448, "grad_norm": 2.775132656097412, "learning_rate": 8.899470667763538e-06, "loss": 1.2846, "step": 11066 }, { "epoch": 0.867319749216301, "grad_norm": 3.244680166244507, "learning_rate": 8.899277653005297e-06, "loss": 1.4578, "step": 11067 }, { "epoch": 0.8673981191222571, "grad_norm": 3.090378522872925, "learning_rate": 8.899084623416207e-06, "loss": 1.1932, "step": 11068 }, { "epoch": 0.8674764890282132, "grad_norm": 2.3265206813812256, "learning_rate": 8.898891578997002e-06, "loss": 1.3545, "step": 11069 }, { "epoch": 0.8675548589341693, "grad_norm": 4.118503093719482, "learning_rate": 8.898698519748417e-06, "loss": 1.474, "step": 11070 }, { "epoch": 0.8676332288401254, "grad_norm": 2.1087098121643066, "learning_rate": 8.898505445671184e-06, "loss": 1.3571, "step": 11071 }, { "epoch": 0.8677115987460815, "grad_norm": 6.007877349853516, "learning_rate": 8.89831235676604e-06, "loss": 1.7669, "step": 11072 }, { "epoch": 0.8677899686520376, "grad_norm": 3.261406421661377, "learning_rate": 8.898119253033716e-06, "loss": 1.8138, "step": 11073 }, { "epoch": 0.8678683385579937, "grad_norm": 2.9783177375793457, "learning_rate": 8.89792613447495e-06, "loss": 1.3156, "step": 11074 }, { "epoch": 0.8679467084639498, "grad_norm": 2.569646120071411, "learning_rate": 8.897733001090473e-06, "loss": 1.3323, "step": 11075 }, { "epoch": 0.868025078369906, "grad_norm": 2.4932913780212402, "learning_rate": 8.897539852881024e-06, "loss": 1.5361, "step": 11076 }, { "epoch": 0.868103448275862, "grad_norm": 4.51788330078125, "learning_rate": 8.897346689847337e-06, "loss": 1.055, "step": 11077 }, { "epoch": 0.8681818181818182, "grad_norm": 3.2921650409698486, "learning_rate": 8.897153511990142e-06, "loss": 1.1817, "step": 11078 }, { "epoch": 0.8682601880877743, "grad_norm": 2.0443115234375, "learning_rate": 8.896960319310178e-06, "loss": 1.5534, "step": 11079 }, { "epoch": 0.8683385579937304, "grad_norm": 2.132417917251587, "learning_rate": 8.896767111808177e-06, "loss": 1.0388, "step": 11080 }, { "epoch": 0.8684169278996865, "grad_norm": 2.136598587036133, "learning_rate": 8.896573889484877e-06, "loss": 1.4736, "step": 11081 }, { "epoch": 0.8684952978056426, "grad_norm": 2.505518913269043, "learning_rate": 8.89638065234101e-06, "loss": 1.5608, "step": 11082 }, { "epoch": 0.8685736677115987, "grad_norm": 2.049435615539551, "learning_rate": 8.896187400377313e-06, "loss": 1.3409, "step": 11083 }, { "epoch": 0.8686520376175548, "grad_norm": 3.4260170459747314, "learning_rate": 8.89599413359452e-06, "loss": 1.5595, "step": 11084 }, { "epoch": 0.8687304075235109, "grad_norm": 3.5063323974609375, "learning_rate": 8.895800851993367e-06, "loss": 1.4379, "step": 11085 }, { "epoch": 0.868808777429467, "grad_norm": 3.5900521278381348, "learning_rate": 8.89560755557459e-06, "loss": 1.4289, "step": 11086 }, { "epoch": 0.8688871473354232, "grad_norm": 2.25201416015625, "learning_rate": 8.89541424433892e-06, "loss": 1.3996, "step": 11087 }, { "epoch": 0.8689655172413793, "grad_norm": 2.3512520790100098, "learning_rate": 8.895220918287096e-06, "loss": 1.5514, "step": 11088 }, { "epoch": 0.8690438871473354, "grad_norm": 2.471968173980713, "learning_rate": 8.895027577419852e-06, "loss": 1.4957, "step": 11089 }, { "epoch": 0.8691222570532915, "grad_norm": 3.5407562255859375, "learning_rate": 8.894834221737924e-06, "loss": 1.3588, "step": 11090 }, { "epoch": 0.8692006269592476, "grad_norm": 2.26924204826355, "learning_rate": 8.89464085124205e-06, "loss": 1.394, "step": 11091 }, { "epoch": 0.8692789968652037, "grad_norm": 5.323990345001221, "learning_rate": 8.894447465932959e-06, "loss": 0.9657, "step": 11092 }, { "epoch": 0.8693573667711598, "grad_norm": 5.075180530548096, "learning_rate": 8.89425406581139e-06, "loss": 1.0373, "step": 11093 }, { "epoch": 0.8694357366771159, "grad_norm": 1.9081203937530518, "learning_rate": 8.89406065087808e-06, "loss": 1.397, "step": 11094 }, { "epoch": 0.8695141065830722, "grad_norm": 2.144009828567505, "learning_rate": 8.893867221133763e-06, "loss": 1.5274, "step": 11095 }, { "epoch": 0.8695924764890283, "grad_norm": 2.1835520267486572, "learning_rate": 8.893673776579176e-06, "loss": 1.2731, "step": 11096 }, { "epoch": 0.8696708463949844, "grad_norm": 2.252009630203247, "learning_rate": 8.893480317215052e-06, "loss": 1.4967, "step": 11097 }, { "epoch": 0.8697492163009405, "grad_norm": 5.843047618865967, "learning_rate": 8.893286843042131e-06, "loss": 2.0327, "step": 11098 }, { "epoch": 0.8698275862068966, "grad_norm": 2.2630186080932617, "learning_rate": 8.893093354061146e-06, "loss": 1.0693, "step": 11099 }, { "epoch": 0.8699059561128527, "grad_norm": 2.342071771621704, "learning_rate": 8.892899850272833e-06, "loss": 1.3718, "step": 11100 }, { "epoch": 0.8699843260188088, "grad_norm": 2.422159433364868, "learning_rate": 8.892706331677928e-06, "loss": 1.4682, "step": 11101 }, { "epoch": 0.8700626959247649, "grad_norm": 2.5772945880889893, "learning_rate": 8.892512798277168e-06, "loss": 2.1049, "step": 11102 }, { "epoch": 0.870141065830721, "grad_norm": 3.2083182334899902, "learning_rate": 8.892319250071289e-06, "loss": 1.3641, "step": 11103 }, { "epoch": 0.8702194357366771, "grad_norm": 4.3915510177612305, "learning_rate": 8.892125687061027e-06, "loss": 0.8917, "step": 11104 }, { "epoch": 0.8702978056426333, "grad_norm": 3.3020222187042236, "learning_rate": 8.891932109247117e-06, "loss": 1.6477, "step": 11105 }, { "epoch": 0.8703761755485894, "grad_norm": 1.9384750127792358, "learning_rate": 8.891738516630298e-06, "loss": 1.381, "step": 11106 }, { "epoch": 0.8704545454545455, "grad_norm": 8.237262725830078, "learning_rate": 8.891544909211303e-06, "loss": 2.0331, "step": 11107 }, { "epoch": 0.8705329153605016, "grad_norm": 4.718122482299805, "learning_rate": 8.89135128699087e-06, "loss": 1.3547, "step": 11108 }, { "epoch": 0.8706112852664577, "grad_norm": 2.6179263591766357, "learning_rate": 8.891157649969736e-06, "loss": 1.3653, "step": 11109 }, { "epoch": 0.8706896551724138, "grad_norm": 2.3386282920837402, "learning_rate": 8.890963998148637e-06, "loss": 1.3459, "step": 11110 }, { "epoch": 0.8707680250783699, "grad_norm": 7.941065788269043, "learning_rate": 8.890770331528309e-06, "loss": 1.9823, "step": 11111 }, { "epoch": 0.870846394984326, "grad_norm": 2.6273446083068848, "learning_rate": 8.890576650109488e-06, "loss": 1.4348, "step": 11112 }, { "epoch": 0.8709247648902821, "grad_norm": 4.130095958709717, "learning_rate": 8.890382953892914e-06, "loss": 1.4741, "step": 11113 }, { "epoch": 0.8710031347962383, "grad_norm": 1.9928418397903442, "learning_rate": 8.89018924287932e-06, "loss": 1.1731, "step": 11114 }, { "epoch": 0.8710815047021944, "grad_norm": 1.8025685548782349, "learning_rate": 8.889995517069446e-06, "loss": 1.4236, "step": 11115 }, { "epoch": 0.8711598746081505, "grad_norm": 2.9149351119995117, "learning_rate": 8.889801776464024e-06, "loss": 1.2887, "step": 11116 }, { "epoch": 0.8712382445141066, "grad_norm": 2.915126323699951, "learning_rate": 8.889608021063796e-06, "loss": 1.8215, "step": 11117 }, { "epoch": 0.8713166144200627, "grad_norm": 2.8510353565216064, "learning_rate": 8.889414250869497e-06, "loss": 1.5582, "step": 11118 }, { "epoch": 0.8713949843260188, "grad_norm": 2.2684812545776367, "learning_rate": 8.889220465881864e-06, "loss": 1.4574, "step": 11119 }, { "epoch": 0.8714733542319749, "grad_norm": 2.2311460971832275, "learning_rate": 8.889026666101635e-06, "loss": 1.2209, "step": 11120 }, { "epoch": 0.871551724137931, "grad_norm": 4.604244232177734, "learning_rate": 8.888832851529543e-06, "loss": 1.8998, "step": 11121 }, { "epoch": 0.8716300940438871, "grad_norm": 4.395787715911865, "learning_rate": 8.88863902216633e-06, "loss": 1.3918, "step": 11122 }, { "epoch": 0.8717084639498432, "grad_norm": 1.8365930318832397, "learning_rate": 8.888445178012733e-06, "loss": 1.395, "step": 11123 }, { "epoch": 0.8717868338557994, "grad_norm": 4.9543538093566895, "learning_rate": 8.888251319069485e-06, "loss": 1.4463, "step": 11124 }, { "epoch": 0.8718652037617555, "grad_norm": 3.6914875507354736, "learning_rate": 8.888057445337328e-06, "loss": 1.3378, "step": 11125 }, { "epoch": 0.8719435736677116, "grad_norm": 2.1485939025878906, "learning_rate": 8.887863556816995e-06, "loss": 1.1939, "step": 11126 }, { "epoch": 0.8720219435736677, "grad_norm": 2.893723726272583, "learning_rate": 8.88766965350923e-06, "loss": 1.5957, "step": 11127 }, { "epoch": 0.8721003134796238, "grad_norm": 2.3029887676239014, "learning_rate": 8.887475735414763e-06, "loss": 1.3037, "step": 11128 }, { "epoch": 0.8721786833855799, "grad_norm": 3.2460508346557617, "learning_rate": 8.887281802534337e-06, "loss": 1.4118, "step": 11129 }, { "epoch": 0.872257053291536, "grad_norm": 9.206500053405762, "learning_rate": 8.887087854868688e-06, "loss": 1.54, "step": 11130 }, { "epoch": 0.8723354231974921, "grad_norm": 3.115373373031616, "learning_rate": 8.886893892418552e-06, "loss": 1.3422, "step": 11131 }, { "epoch": 0.8724137931034482, "grad_norm": 3.1222405433654785, "learning_rate": 8.886699915184667e-06, "loss": 1.5617, "step": 11132 }, { "epoch": 0.8724921630094044, "grad_norm": 8.05886173248291, "learning_rate": 8.886505923167774e-06, "loss": 1.4779, "step": 11133 }, { "epoch": 0.8725705329153605, "grad_norm": 9.482497215270996, "learning_rate": 8.886311916368608e-06, "loss": 1.6987, "step": 11134 }, { "epoch": 0.8726489028213166, "grad_norm": 1.9391708374023438, "learning_rate": 8.886117894787907e-06, "loss": 1.2962, "step": 11135 }, { "epoch": 0.8727272727272727, "grad_norm": 3.2513906955718994, "learning_rate": 8.885923858426411e-06, "loss": 1.042, "step": 11136 }, { "epoch": 0.8728056426332288, "grad_norm": 2.128080368041992, "learning_rate": 8.885729807284855e-06, "loss": 1.3539, "step": 11137 }, { "epoch": 0.8728840125391849, "grad_norm": 3.03053879737854, "learning_rate": 8.88553574136398e-06, "loss": 1.4223, "step": 11138 }, { "epoch": 0.872962382445141, "grad_norm": 2.6397807598114014, "learning_rate": 8.885341660664521e-06, "loss": 1.3679, "step": 11139 }, { "epoch": 0.8730407523510971, "grad_norm": 2.2965993881225586, "learning_rate": 8.88514756518722e-06, "loss": 1.3845, "step": 11140 }, { "epoch": 0.8731191222570532, "grad_norm": 4.1668596267700195, "learning_rate": 8.884953454932812e-06, "loss": 1.8353, "step": 11141 }, { "epoch": 0.8731974921630095, "grad_norm": 2.5639147758483887, "learning_rate": 8.884759329902038e-06, "loss": 1.4139, "step": 11142 }, { "epoch": 0.8732758620689656, "grad_norm": 2.0578291416168213, "learning_rate": 8.884565190095632e-06, "loss": 1.3941, "step": 11143 }, { "epoch": 0.8733542319749217, "grad_norm": 6.901939868927002, "learning_rate": 8.884371035514338e-06, "loss": 1.8604, "step": 11144 }, { "epoch": 0.8734326018808778, "grad_norm": 1.846389651298523, "learning_rate": 8.88417686615889e-06, "loss": 1.3014, "step": 11145 }, { "epoch": 0.8735109717868339, "grad_norm": 2.19143009185791, "learning_rate": 8.883982682030029e-06, "loss": 1.33, "step": 11146 }, { "epoch": 0.87358934169279, "grad_norm": 2.189814805984497, "learning_rate": 8.883788483128493e-06, "loss": 1.3438, "step": 11147 }, { "epoch": 0.8736677115987461, "grad_norm": 1.8799822330474854, "learning_rate": 8.88359426945502e-06, "loss": 1.2973, "step": 11148 }, { "epoch": 0.8737460815047022, "grad_norm": 5.627345085144043, "learning_rate": 8.883400041010349e-06, "loss": 1.96, "step": 11149 }, { "epoch": 0.8738244514106583, "grad_norm": 2.309123992919922, "learning_rate": 8.883205797795219e-06, "loss": 1.5622, "step": 11150 }, { "epoch": 0.8739028213166145, "grad_norm": 2.485700845718384, "learning_rate": 8.883011539810368e-06, "loss": 1.3568, "step": 11151 }, { "epoch": 0.8739811912225706, "grad_norm": 2.8337604999542236, "learning_rate": 8.882817267056538e-06, "loss": 1.1399, "step": 11152 }, { "epoch": 0.8740595611285267, "grad_norm": 4.920188903808594, "learning_rate": 8.882622979534462e-06, "loss": 1.1584, "step": 11153 }, { "epoch": 0.8741379310344828, "grad_norm": 2.413578748703003, "learning_rate": 8.882428677244886e-06, "loss": 1.1559, "step": 11154 }, { "epoch": 0.8742163009404389, "grad_norm": 2.4391956329345703, "learning_rate": 8.882234360188544e-06, "loss": 1.016, "step": 11155 }, { "epoch": 0.874294670846395, "grad_norm": 3.4718215465545654, "learning_rate": 8.882040028366176e-06, "loss": 1.4128, "step": 11156 }, { "epoch": 0.8743730407523511, "grad_norm": 4.515082359313965, "learning_rate": 8.881845681778522e-06, "loss": 1.9659, "step": 11157 }, { "epoch": 0.8744514106583072, "grad_norm": 2.382340669631958, "learning_rate": 8.88165132042632e-06, "loss": 1.4534, "step": 11158 }, { "epoch": 0.8745297805642633, "grad_norm": 17.68781852722168, "learning_rate": 8.881456944310311e-06, "loss": 1.2093, "step": 11159 }, { "epoch": 0.8746081504702194, "grad_norm": 3.186600923538208, "learning_rate": 8.881262553431233e-06, "loss": 0.9041, "step": 11160 }, { "epoch": 0.8746865203761756, "grad_norm": 3.211449146270752, "learning_rate": 8.881068147789827e-06, "loss": 1.433, "step": 11161 }, { "epoch": 0.8747648902821317, "grad_norm": 4.473216533660889, "learning_rate": 8.880873727386829e-06, "loss": 1.5344, "step": 11162 }, { "epoch": 0.8748432601880878, "grad_norm": 3.473156213760376, "learning_rate": 8.880679292222982e-06, "loss": 1.0151, "step": 11163 }, { "epoch": 0.8749216300940439, "grad_norm": 2.832024574279785, "learning_rate": 8.880484842299024e-06, "loss": 1.375, "step": 11164 }, { "epoch": 0.875, "grad_norm": 1.9956886768341064, "learning_rate": 8.880290377615697e-06, "loss": 1.7063, "step": 11165 }, { "epoch": 0.8750783699059561, "grad_norm": 2.5563316345214844, "learning_rate": 8.880095898173735e-06, "loss": 1.4004, "step": 11166 }, { "epoch": 0.8751567398119122, "grad_norm": 3.072922468185425, "learning_rate": 8.879901403973882e-06, "loss": 1.4769, "step": 11167 }, { "epoch": 0.8752351097178683, "grad_norm": 2.5054662227630615, "learning_rate": 8.879706895016878e-06, "loss": 1.3262, "step": 11168 }, { "epoch": 0.8753134796238244, "grad_norm": 3.4625797271728516, "learning_rate": 8.879512371303462e-06, "loss": 1.4662, "step": 11169 }, { "epoch": 0.8753918495297806, "grad_norm": 2.381037950515747, "learning_rate": 8.879317832834372e-06, "loss": 1.2428, "step": 11170 }, { "epoch": 0.8754702194357367, "grad_norm": 2.642568588256836, "learning_rate": 8.87912327961035e-06, "loss": 1.2197, "step": 11171 }, { "epoch": 0.8755485893416928, "grad_norm": 2.3833274841308594, "learning_rate": 8.878928711632138e-06, "loss": 1.5217, "step": 11172 }, { "epoch": 0.8756269592476489, "grad_norm": 4.723454475402832, "learning_rate": 8.878734128900471e-06, "loss": 1.2881, "step": 11173 }, { "epoch": 0.875705329153605, "grad_norm": 2.8517544269561768, "learning_rate": 8.878539531416092e-06, "loss": 1.2996, "step": 11174 }, { "epoch": 0.8757836990595611, "grad_norm": 3.4553420543670654, "learning_rate": 8.87834491917974e-06, "loss": 1.3986, "step": 11175 }, { "epoch": 0.8758620689655172, "grad_norm": 2.067927360534668, "learning_rate": 8.878150292192158e-06, "loss": 1.6306, "step": 11176 }, { "epoch": 0.8759404388714733, "grad_norm": 2.130876302719116, "learning_rate": 8.877955650454082e-06, "loss": 1.0872, "step": 11177 }, { "epoch": 0.8760188087774294, "grad_norm": 2.003171920776367, "learning_rate": 8.877760993966256e-06, "loss": 1.3695, "step": 11178 }, { "epoch": 0.8760971786833855, "grad_norm": 15.317517280578613, "learning_rate": 8.877566322729419e-06, "loss": 1.7334, "step": 11179 }, { "epoch": 0.8761755485893417, "grad_norm": 2.8282032012939453, "learning_rate": 8.877371636744309e-06, "loss": 1.542, "step": 11180 }, { "epoch": 0.8762539184952978, "grad_norm": 2.3168671131134033, "learning_rate": 8.877176936011671e-06, "loss": 1.229, "step": 11181 }, { "epoch": 0.8763322884012539, "grad_norm": 3.486161470413208, "learning_rate": 8.876982220532242e-06, "loss": 1.5386, "step": 11182 }, { "epoch": 0.87641065830721, "grad_norm": 3.5457746982574463, "learning_rate": 8.876787490306767e-06, "loss": 1.131, "step": 11183 }, { "epoch": 0.8764890282131661, "grad_norm": 3.097165107727051, "learning_rate": 8.876592745335982e-06, "loss": 1.6524, "step": 11184 }, { "epoch": 0.8765673981191222, "grad_norm": 2.3452019691467285, "learning_rate": 8.876397985620628e-06, "loss": 1.4383, "step": 11185 }, { "epoch": 0.8766457680250783, "grad_norm": 2.225269079208374, "learning_rate": 8.876203211161448e-06, "loss": 1.5446, "step": 11186 }, { "epoch": 0.8767241379310344, "grad_norm": 3.5978150367736816, "learning_rate": 8.876008421959181e-06, "loss": 1.1563, "step": 11187 }, { "epoch": 0.8768025078369905, "grad_norm": 2.291342258453369, "learning_rate": 8.87581361801457e-06, "loss": 1.3274, "step": 11188 }, { "epoch": 0.8768808777429468, "grad_norm": 3.4178738594055176, "learning_rate": 8.875618799328354e-06, "loss": 0.7476, "step": 11189 }, { "epoch": 0.8769592476489029, "grad_norm": 2.527153253555298, "learning_rate": 8.875423965901275e-06, "loss": 1.2166, "step": 11190 }, { "epoch": 0.877037617554859, "grad_norm": 2.19871187210083, "learning_rate": 8.875229117734075e-06, "loss": 1.3392, "step": 11191 }, { "epoch": 0.8771159874608151, "grad_norm": 4.716139793395996, "learning_rate": 8.875034254827492e-06, "loss": 1.1994, "step": 11192 }, { "epoch": 0.8771943573667712, "grad_norm": 7.458566665649414, "learning_rate": 8.874839377182272e-06, "loss": 2.0781, "step": 11193 }, { "epoch": 0.8772727272727273, "grad_norm": 2.9882969856262207, "learning_rate": 8.87464448479915e-06, "loss": 1.3271, "step": 11194 }, { "epoch": 0.8773510971786834, "grad_norm": 8.92238998413086, "learning_rate": 8.874449577678872e-06, "loss": 1.6951, "step": 11195 }, { "epoch": 0.8774294670846395, "grad_norm": 2.2498507499694824, "learning_rate": 8.874254655822176e-06, "loss": 1.379, "step": 11196 }, { "epoch": 0.8775078369905956, "grad_norm": 3.164058208465576, "learning_rate": 8.874059719229809e-06, "loss": 1.1296, "step": 11197 }, { "epoch": 0.8775862068965518, "grad_norm": 3.753588914871216, "learning_rate": 8.873864767902504e-06, "loss": 0.9815, "step": 11198 }, { "epoch": 0.8776645768025079, "grad_norm": 2.8784284591674805, "learning_rate": 8.873669801841012e-06, "loss": 1.4214, "step": 11199 }, { "epoch": 0.877742946708464, "grad_norm": 2.2810914516448975, "learning_rate": 8.873474821046066e-06, "loss": 1.7737, "step": 11200 }, { "epoch": 0.8778213166144201, "grad_norm": 4.916818618774414, "learning_rate": 8.873279825518413e-06, "loss": 1.4083, "step": 11201 }, { "epoch": 0.8778996865203762, "grad_norm": 3.227552890777588, "learning_rate": 8.873084815258793e-06, "loss": 1.1354, "step": 11202 }, { "epoch": 0.8779780564263323, "grad_norm": 2.968836784362793, "learning_rate": 8.872889790267946e-06, "loss": 1.5704, "step": 11203 }, { "epoch": 0.8780564263322884, "grad_norm": 4.1769609451293945, "learning_rate": 8.872694750546617e-06, "loss": 1.2494, "step": 11204 }, { "epoch": 0.8781347962382445, "grad_norm": 32.99010467529297, "learning_rate": 8.872499696095546e-06, "loss": 1.2813, "step": 11205 }, { "epoch": 0.8782131661442006, "grad_norm": 3.1657512187957764, "learning_rate": 8.872304626915477e-06, "loss": 1.604, "step": 11206 }, { "epoch": 0.8782915360501568, "grad_norm": 2.7912356853485107, "learning_rate": 8.872109543007148e-06, "loss": 1.5042, "step": 11207 }, { "epoch": 0.8783699059561129, "grad_norm": 3.1447157859802246, "learning_rate": 8.871914444371303e-06, "loss": 1.6141, "step": 11208 }, { "epoch": 0.878448275862069, "grad_norm": 4.501244068145752, "learning_rate": 8.871719331008686e-06, "loss": 1.5382, "step": 11209 }, { "epoch": 0.8785266457680251, "grad_norm": 2.543377637863159, "learning_rate": 8.871524202920036e-06, "loss": 1.3623, "step": 11210 }, { "epoch": 0.8786050156739812, "grad_norm": 2.545119524002075, "learning_rate": 8.871329060106097e-06, "loss": 1.5473, "step": 11211 }, { "epoch": 0.8786833855799373, "grad_norm": 3.4171886444091797, "learning_rate": 8.87113390256761e-06, "loss": 0.7142, "step": 11212 }, { "epoch": 0.8787617554858934, "grad_norm": 1.920828104019165, "learning_rate": 8.870938730305318e-06, "loss": 1.3221, "step": 11213 }, { "epoch": 0.8788401253918495, "grad_norm": 3.6678977012634277, "learning_rate": 8.870743543319964e-06, "loss": 1.2304, "step": 11214 }, { "epoch": 0.8789184952978056, "grad_norm": 3.7914607524871826, "learning_rate": 8.870548341612288e-06, "loss": 1.2235, "step": 11215 }, { "epoch": 0.8789968652037617, "grad_norm": 2.2968909740448, "learning_rate": 8.870353125183035e-06, "loss": 1.4222, "step": 11216 }, { "epoch": 0.8790752351097179, "grad_norm": 14.705446243286133, "learning_rate": 8.870157894032948e-06, "loss": 1.309, "step": 11217 }, { "epoch": 0.879153605015674, "grad_norm": 1.788465142250061, "learning_rate": 8.869962648162767e-06, "loss": 1.1527, "step": 11218 }, { "epoch": 0.8792319749216301, "grad_norm": 4.747528076171875, "learning_rate": 8.869767387573236e-06, "loss": 1.8318, "step": 11219 }, { "epoch": 0.8793103448275862, "grad_norm": 4.370222568511963, "learning_rate": 8.869572112265099e-06, "loss": 1.2023, "step": 11220 }, { "epoch": 0.8793887147335423, "grad_norm": 2.9170570373535156, "learning_rate": 8.869376822239095e-06, "loss": 1.5039, "step": 11221 }, { "epoch": 0.8794670846394984, "grad_norm": 12.00589370727539, "learning_rate": 8.869181517495971e-06, "loss": 2.0317, "step": 11222 }, { "epoch": 0.8795454545454545, "grad_norm": 2.950519561767578, "learning_rate": 8.868986198036465e-06, "loss": 1.3118, "step": 11223 }, { "epoch": 0.8796238244514106, "grad_norm": 2.059549331665039, "learning_rate": 8.868790863861325e-06, "loss": 1.3759, "step": 11224 }, { "epoch": 0.8797021943573667, "grad_norm": 2.4244906902313232, "learning_rate": 8.868595514971292e-06, "loss": 1.2829, "step": 11225 }, { "epoch": 0.8797805642633229, "grad_norm": 1.9530096054077148, "learning_rate": 8.868400151367107e-06, "loss": 1.2742, "step": 11226 }, { "epoch": 0.879858934169279, "grad_norm": 2.939291000366211, "learning_rate": 8.868204773049515e-06, "loss": 0.8336, "step": 11227 }, { "epoch": 0.8799373040752351, "grad_norm": 4.796253204345703, "learning_rate": 8.86800938001926e-06, "loss": 0.9106, "step": 11228 }, { "epoch": 0.8800156739811912, "grad_norm": 4.350161075592041, "learning_rate": 8.867813972277084e-06, "loss": 1.1815, "step": 11229 }, { "epoch": 0.8800940438871473, "grad_norm": 3.6604766845703125, "learning_rate": 8.867618549823728e-06, "loss": 1.1235, "step": 11230 }, { "epoch": 0.8801724137931034, "grad_norm": 1.9841309785842896, "learning_rate": 8.867423112659941e-06, "loss": 1.4491, "step": 11231 }, { "epoch": 0.8802507836990595, "grad_norm": 4.331821918487549, "learning_rate": 8.867227660786461e-06, "loss": 1.6055, "step": 11232 }, { "epoch": 0.8803291536050156, "grad_norm": 2.923675537109375, "learning_rate": 8.867032194204034e-06, "loss": 1.5891, "step": 11233 }, { "epoch": 0.8804075235109717, "grad_norm": 2.575716972351074, "learning_rate": 8.866836712913403e-06, "loss": 1.5465, "step": 11234 }, { "epoch": 0.8804858934169278, "grad_norm": 4.405096054077148, "learning_rate": 8.86664121691531e-06, "loss": 1.1325, "step": 11235 }, { "epoch": 0.8805642633228841, "grad_norm": 2.144298553466797, "learning_rate": 8.8664457062105e-06, "loss": 1.3268, "step": 11236 }, { "epoch": 0.8806426332288402, "grad_norm": 2.711411237716675, "learning_rate": 8.866250180799716e-06, "loss": 1.5756, "step": 11237 }, { "epoch": 0.8807210031347963, "grad_norm": 2.1600775718688965, "learning_rate": 8.866054640683704e-06, "loss": 1.3778, "step": 11238 }, { "epoch": 0.8807993730407524, "grad_norm": 3.3163042068481445, "learning_rate": 8.865859085863205e-06, "loss": 1.4754, "step": 11239 }, { "epoch": 0.8808777429467085, "grad_norm": 2.318859577178955, "learning_rate": 8.865663516338963e-06, "loss": 1.281, "step": 11240 }, { "epoch": 0.8809561128526646, "grad_norm": 2.230191230773926, "learning_rate": 8.865467932111725e-06, "loss": 1.6892, "step": 11241 }, { "epoch": 0.8810344827586207, "grad_norm": 7.32847785949707, "learning_rate": 8.865272333182229e-06, "loss": 1.4861, "step": 11242 }, { "epoch": 0.8811128526645768, "grad_norm": 6.833795547485352, "learning_rate": 8.865076719551224e-06, "loss": 1.772, "step": 11243 }, { "epoch": 0.881191222570533, "grad_norm": 2.0179836750030518, "learning_rate": 8.864881091219453e-06, "loss": 1.2709, "step": 11244 }, { "epoch": 0.8812695924764891, "grad_norm": 2.1983821392059326, "learning_rate": 8.864685448187657e-06, "loss": 1.1532, "step": 11245 }, { "epoch": 0.8813479623824452, "grad_norm": 2.4820783138275146, "learning_rate": 8.864489790456586e-06, "loss": 1.4083, "step": 11246 }, { "epoch": 0.8814263322884013, "grad_norm": 4.444948673248291, "learning_rate": 8.86429411802698e-06, "loss": 1.2862, "step": 11247 }, { "epoch": 0.8815047021943574, "grad_norm": 2.292814016342163, "learning_rate": 8.864098430899582e-06, "loss": 1.4083, "step": 11248 }, { "epoch": 0.8815830721003135, "grad_norm": 4.059406757354736, "learning_rate": 8.86390272907514e-06, "loss": 1.2539, "step": 11249 }, { "epoch": 0.8816614420062696, "grad_norm": 2.8763351440429688, "learning_rate": 8.863707012554394e-06, "loss": 1.4671, "step": 11250 }, { "epoch": 0.8817398119122257, "grad_norm": 2.4045627117156982, "learning_rate": 8.863511281338095e-06, "loss": 1.1936, "step": 11251 }, { "epoch": 0.8818181818181818, "grad_norm": 2.529025077819824, "learning_rate": 8.863315535426981e-06, "loss": 1.4262, "step": 11252 }, { "epoch": 0.881896551724138, "grad_norm": 4.112610340118408, "learning_rate": 8.8631197748218e-06, "loss": 1.0559, "step": 11253 }, { "epoch": 0.881974921630094, "grad_norm": 7.414586067199707, "learning_rate": 8.862923999523295e-06, "loss": 2.1597, "step": 11254 }, { "epoch": 0.8820532915360502, "grad_norm": 2.047504186630249, "learning_rate": 8.862728209532213e-06, "loss": 1.585, "step": 11255 }, { "epoch": 0.8821316614420063, "grad_norm": 2.4575140476226807, "learning_rate": 8.862532404849293e-06, "loss": 1.456, "step": 11256 }, { "epoch": 0.8822100313479624, "grad_norm": 2.5992367267608643, "learning_rate": 8.862336585475287e-06, "loss": 1.3656, "step": 11257 }, { "epoch": 0.8822884012539185, "grad_norm": 3.508777141571045, "learning_rate": 8.862140751410934e-06, "loss": 1.1981, "step": 11258 }, { "epoch": 0.8823667711598746, "grad_norm": 2.41007924079895, "learning_rate": 8.861944902656985e-06, "loss": 1.5784, "step": 11259 }, { "epoch": 0.8824451410658307, "grad_norm": 3.1120848655700684, "learning_rate": 8.861749039214177e-06, "loss": 1.6614, "step": 11260 }, { "epoch": 0.8825235109717868, "grad_norm": 5.351693630218506, "learning_rate": 8.861553161083262e-06, "loss": 0.9869, "step": 11261 }, { "epoch": 0.8826018808777429, "grad_norm": 3.6083927154541016, "learning_rate": 8.861357268264982e-06, "loss": 1.2288, "step": 11262 }, { "epoch": 0.882680250783699, "grad_norm": 3.551354169845581, "learning_rate": 8.861161360760083e-06, "loss": 1.3573, "step": 11263 }, { "epoch": 0.8827586206896552, "grad_norm": 2.2372725009918213, "learning_rate": 8.860965438569308e-06, "loss": 1.3207, "step": 11264 }, { "epoch": 0.8828369905956113, "grad_norm": 2.1892781257629395, "learning_rate": 8.860769501693403e-06, "loss": 1.4385, "step": 11265 }, { "epoch": 0.8829153605015674, "grad_norm": 2.7899882793426514, "learning_rate": 8.860573550133114e-06, "loss": 0.9757, "step": 11266 }, { "epoch": 0.8829937304075235, "grad_norm": 3.7454259395599365, "learning_rate": 8.860377583889185e-06, "loss": 1.2579, "step": 11267 }, { "epoch": 0.8830721003134796, "grad_norm": 5.615606784820557, "learning_rate": 8.860181602962365e-06, "loss": 0.7753, "step": 11268 }, { "epoch": 0.8831504702194357, "grad_norm": 2.4402828216552734, "learning_rate": 8.859985607353396e-06, "loss": 1.4613, "step": 11269 }, { "epoch": 0.8832288401253918, "grad_norm": 4.031301021575928, "learning_rate": 8.859789597063024e-06, "loss": 1.2538, "step": 11270 }, { "epoch": 0.8833072100313479, "grad_norm": 2.205777645111084, "learning_rate": 8.859593572091995e-06, "loss": 1.1538, "step": 11271 }, { "epoch": 0.883385579937304, "grad_norm": 3.6121222972869873, "learning_rate": 8.859397532441055e-06, "loss": 1.3784, "step": 11272 }, { "epoch": 0.8834639498432602, "grad_norm": 2.9263525009155273, "learning_rate": 8.85920147811095e-06, "loss": 1.6376, "step": 11273 }, { "epoch": 0.8835423197492163, "grad_norm": 3.2478556632995605, "learning_rate": 8.859005409102424e-06, "loss": 1.6384, "step": 11274 }, { "epoch": 0.8836206896551724, "grad_norm": 2.634127140045166, "learning_rate": 8.858809325416221e-06, "loss": 1.4869, "step": 11275 }, { "epoch": 0.8836990595611285, "grad_norm": 3.191098928451538, "learning_rate": 8.858613227053093e-06, "loss": 1.265, "step": 11276 }, { "epoch": 0.8837774294670846, "grad_norm": 2.20343017578125, "learning_rate": 8.85841711401378e-06, "loss": 1.4398, "step": 11277 }, { "epoch": 0.8838557993730407, "grad_norm": 2.865067720413208, "learning_rate": 8.85822098629903e-06, "loss": 1.0925, "step": 11278 }, { "epoch": 0.8839341692789968, "grad_norm": 3.532452344894409, "learning_rate": 8.85802484390959e-06, "loss": 0.8918, "step": 11279 }, { "epoch": 0.8840125391849529, "grad_norm": 2.948462724685669, "learning_rate": 8.857828686846206e-06, "loss": 1.3918, "step": 11280 }, { "epoch": 0.884090909090909, "grad_norm": 2.093552827835083, "learning_rate": 8.857632515109621e-06, "loss": 1.5469, "step": 11281 }, { "epoch": 0.8841692789968651, "grad_norm": 3.2139031887054443, "learning_rate": 8.857436328700586e-06, "loss": 0.9493, "step": 11282 }, { "epoch": 0.8842476489028214, "grad_norm": 2.715249538421631, "learning_rate": 8.857240127619842e-06, "loss": 1.3595, "step": 11283 }, { "epoch": 0.8843260188087775, "grad_norm": 2.992283582687378, "learning_rate": 8.85704391186814e-06, "loss": 1.4702, "step": 11284 }, { "epoch": 0.8844043887147336, "grad_norm": 2.474269390106201, "learning_rate": 8.85684768144622e-06, "loss": 1.2929, "step": 11285 }, { "epoch": 0.8844827586206897, "grad_norm": 3.183201551437378, "learning_rate": 8.856651436354837e-06, "loss": 1.7649, "step": 11286 }, { "epoch": 0.8845611285266458, "grad_norm": 3.6682803630828857, "learning_rate": 8.856455176594732e-06, "loss": 1.4186, "step": 11287 }, { "epoch": 0.8846394984326019, "grad_norm": 2.4825103282928467, "learning_rate": 8.85625890216665e-06, "loss": 1.4855, "step": 11288 }, { "epoch": 0.884717868338558, "grad_norm": 2.66039776802063, "learning_rate": 8.856062613071342e-06, "loss": 0.7533, "step": 11289 }, { "epoch": 0.8847962382445141, "grad_norm": 2.311861515045166, "learning_rate": 8.85586630930955e-06, "loss": 1.05, "step": 11290 }, { "epoch": 0.8848746081504703, "grad_norm": 3.1438872814178467, "learning_rate": 8.855669990882026e-06, "loss": 1.1614, "step": 11291 }, { "epoch": 0.8849529780564264, "grad_norm": 3.1033623218536377, "learning_rate": 8.855473657789512e-06, "loss": 1.3142, "step": 11292 }, { "epoch": 0.8850313479623825, "grad_norm": 2.9368298053741455, "learning_rate": 8.855277310032758e-06, "loss": 1.5801, "step": 11293 }, { "epoch": 0.8851097178683386, "grad_norm": 1.8280885219573975, "learning_rate": 8.855080947612507e-06, "loss": 1.3898, "step": 11294 }, { "epoch": 0.8851880877742947, "grad_norm": 8.784198760986328, "learning_rate": 8.85488457052951e-06, "loss": 1.9776, "step": 11295 }, { "epoch": 0.8852664576802508, "grad_norm": 1.9938743114471436, "learning_rate": 8.854688178784512e-06, "loss": 1.4211, "step": 11296 }, { "epoch": 0.8853448275862069, "grad_norm": 3.3056252002716064, "learning_rate": 8.854491772378259e-06, "loss": 0.8391, "step": 11297 }, { "epoch": 0.885423197492163, "grad_norm": 2.77301025390625, "learning_rate": 8.8542953513115e-06, "loss": 0.9886, "step": 11298 }, { "epoch": 0.8855015673981191, "grad_norm": 2.9168660640716553, "learning_rate": 8.85409891558498e-06, "loss": 1.5101, "step": 11299 }, { "epoch": 0.8855799373040752, "grad_norm": 2.2025251388549805, "learning_rate": 8.853902465199448e-06, "loss": 1.2849, "step": 11300 }, { "epoch": 0.8856583072100314, "grad_norm": 1.8194406032562256, "learning_rate": 8.85370600015565e-06, "loss": 1.2096, "step": 11301 }, { "epoch": 0.8857366771159875, "grad_norm": 1.9413113594055176, "learning_rate": 8.853509520454335e-06, "loss": 1.0732, "step": 11302 }, { "epoch": 0.8858150470219436, "grad_norm": 2.5330159664154053, "learning_rate": 8.85331302609625e-06, "loss": 1.6169, "step": 11303 }, { "epoch": 0.8858934169278997, "grad_norm": 1.7282142639160156, "learning_rate": 8.853116517082139e-06, "loss": 1.255, "step": 11304 }, { "epoch": 0.8859717868338558, "grad_norm": 2.8890562057495117, "learning_rate": 8.852919993412751e-06, "loss": 1.2477, "step": 11305 }, { "epoch": 0.8860501567398119, "grad_norm": 2.627051830291748, "learning_rate": 8.852723455088838e-06, "loss": 2.012, "step": 11306 }, { "epoch": 0.886128526645768, "grad_norm": 3.3255372047424316, "learning_rate": 8.85252690211114e-06, "loss": 1.8446, "step": 11307 }, { "epoch": 0.8862068965517241, "grad_norm": 3.7471418380737305, "learning_rate": 8.852330334480413e-06, "loss": 1.9606, "step": 11308 }, { "epoch": 0.8862852664576802, "grad_norm": 3.333967447280884, "learning_rate": 8.852133752197397e-06, "loss": 1.5265, "step": 11309 }, { "epoch": 0.8863636363636364, "grad_norm": 2.140812873840332, "learning_rate": 8.851937155262842e-06, "loss": 1.2794, "step": 11310 }, { "epoch": 0.8864420062695925, "grad_norm": 2.3235998153686523, "learning_rate": 8.851740543677499e-06, "loss": 1.5715, "step": 11311 }, { "epoch": 0.8865203761755486, "grad_norm": 2.183833360671997, "learning_rate": 8.851543917442111e-06, "loss": 1.4971, "step": 11312 }, { "epoch": 0.8865987460815047, "grad_norm": 1.8224586248397827, "learning_rate": 8.85134727655743e-06, "loss": 1.4128, "step": 11313 }, { "epoch": 0.8866771159874608, "grad_norm": 5.004862308502197, "learning_rate": 8.8511506210242e-06, "loss": 1.4202, "step": 11314 }, { "epoch": 0.8867554858934169, "grad_norm": 3.174856662750244, "learning_rate": 8.850953950843173e-06, "loss": 1.4727, "step": 11315 }, { "epoch": 0.886833855799373, "grad_norm": 2.459592580795288, "learning_rate": 8.850757266015096e-06, "loss": 1.367, "step": 11316 }, { "epoch": 0.8869122257053291, "grad_norm": 2.258903741836548, "learning_rate": 8.850560566540714e-06, "loss": 1.4152, "step": 11317 }, { "epoch": 0.8869905956112852, "grad_norm": 2.3472232818603516, "learning_rate": 8.850363852420778e-06, "loss": 1.4872, "step": 11318 }, { "epoch": 0.8870689655172413, "grad_norm": 2.563051462173462, "learning_rate": 8.850167123656038e-06, "loss": 1.2203, "step": 11319 }, { "epoch": 0.8871473354231975, "grad_norm": 4.530460357666016, "learning_rate": 8.849970380247237e-06, "loss": 1.5999, "step": 11320 }, { "epoch": 0.8872257053291536, "grad_norm": 1.976378083229065, "learning_rate": 8.849773622195127e-06, "loss": 1.4833, "step": 11321 }, { "epoch": 0.8873040752351097, "grad_norm": 2.879502534866333, "learning_rate": 8.849576849500456e-06, "loss": 1.2134, "step": 11322 }, { "epoch": 0.8873824451410658, "grad_norm": 1.8811249732971191, "learning_rate": 8.84938006216397e-06, "loss": 1.312, "step": 11323 }, { "epoch": 0.8874608150470219, "grad_norm": 3.1190569400787354, "learning_rate": 8.84918326018642e-06, "loss": 1.4394, "step": 11324 }, { "epoch": 0.887539184952978, "grad_norm": 2.846961736679077, "learning_rate": 8.848986443568555e-06, "loss": 1.3982, "step": 11325 }, { "epoch": 0.8876175548589341, "grad_norm": 2.4518184661865234, "learning_rate": 8.848789612311123e-06, "loss": 1.43, "step": 11326 }, { "epoch": 0.8876959247648902, "grad_norm": 4.168776512145996, "learning_rate": 8.848592766414872e-06, "loss": 1.1213, "step": 11327 }, { "epoch": 0.8877742946708463, "grad_norm": 6.496700763702393, "learning_rate": 8.84839590588055e-06, "loss": 0.8423, "step": 11328 }, { "epoch": 0.8878526645768025, "grad_norm": 2.630591630935669, "learning_rate": 8.848199030708907e-06, "loss": 1.6461, "step": 11329 }, { "epoch": 0.8879310344827587, "grad_norm": 3.423886299133301, "learning_rate": 8.848002140900691e-06, "loss": 1.732, "step": 11330 }, { "epoch": 0.8880094043887148, "grad_norm": 3.718562126159668, "learning_rate": 8.847805236456653e-06, "loss": 1.2614, "step": 11331 }, { "epoch": 0.8880877742946709, "grad_norm": 6.076465606689453, "learning_rate": 8.847608317377538e-06, "loss": 0.4803, "step": 11332 }, { "epoch": 0.888166144200627, "grad_norm": 5.6502461433410645, "learning_rate": 8.847411383664098e-06, "loss": 1.4138, "step": 11333 }, { "epoch": 0.8882445141065831, "grad_norm": 2.2008309364318848, "learning_rate": 8.847214435317081e-06, "loss": 1.6048, "step": 11334 }, { "epoch": 0.8883228840125392, "grad_norm": 5.919304370880127, "learning_rate": 8.847017472337237e-06, "loss": 1.1658, "step": 11335 }, { "epoch": 0.8884012539184953, "grad_norm": 2.5745835304260254, "learning_rate": 8.846820494725315e-06, "loss": 1.013, "step": 11336 }, { "epoch": 0.8884796238244514, "grad_norm": 2.399095296859741, "learning_rate": 8.846623502482063e-06, "loss": 1.226, "step": 11337 }, { "epoch": 0.8885579937304076, "grad_norm": 3.3701393604278564, "learning_rate": 8.84642649560823e-06, "loss": 1.2644, "step": 11338 }, { "epoch": 0.8886363636363637, "grad_norm": 6.222731590270996, "learning_rate": 8.846229474104567e-06, "loss": 1.5484, "step": 11339 }, { "epoch": 0.8887147335423198, "grad_norm": 2.283353805541992, "learning_rate": 8.846032437971823e-06, "loss": 1.5063, "step": 11340 }, { "epoch": 0.8887931034482759, "grad_norm": 1.8365403413772583, "learning_rate": 8.845835387210748e-06, "loss": 1.3655, "step": 11341 }, { "epoch": 0.888871473354232, "grad_norm": 2.4767239093780518, "learning_rate": 8.845638321822087e-06, "loss": 1.3604, "step": 11342 }, { "epoch": 0.8889498432601881, "grad_norm": 3.9270224571228027, "learning_rate": 8.845441241806596e-06, "loss": 1.7018, "step": 11343 }, { "epoch": 0.8890282131661442, "grad_norm": 5.502744674682617, "learning_rate": 8.845244147165022e-06, "loss": 0.9965, "step": 11344 }, { "epoch": 0.8891065830721003, "grad_norm": 2.877218008041382, "learning_rate": 8.845047037898114e-06, "loss": 1.4368, "step": 11345 }, { "epoch": 0.8891849529780564, "grad_norm": 3.8527441024780273, "learning_rate": 8.84484991400662e-06, "loss": 1.4138, "step": 11346 }, { "epoch": 0.8892633228840126, "grad_norm": 2.787475347518921, "learning_rate": 8.844652775491293e-06, "loss": 1.3993, "step": 11347 }, { "epoch": 0.8893416927899687, "grad_norm": 2.6568238735198975, "learning_rate": 8.844455622352884e-06, "loss": 1.6835, "step": 11348 }, { "epoch": 0.8894200626959248, "grad_norm": 6.13340950012207, "learning_rate": 8.844258454592137e-06, "loss": 1.0984, "step": 11349 }, { "epoch": 0.8894984326018809, "grad_norm": 2.098238706588745, "learning_rate": 8.844061272209807e-06, "loss": 1.4441, "step": 11350 }, { "epoch": 0.889576802507837, "grad_norm": 4.305568695068359, "learning_rate": 8.84386407520664e-06, "loss": 1.2077, "step": 11351 }, { "epoch": 0.8896551724137931, "grad_norm": 1.966306209564209, "learning_rate": 8.843666863583392e-06, "loss": 1.4741, "step": 11352 }, { "epoch": 0.8897335423197492, "grad_norm": 1.9364537000656128, "learning_rate": 8.843469637340806e-06, "loss": 1.6001, "step": 11353 }, { "epoch": 0.8898119122257053, "grad_norm": 2.293316602706909, "learning_rate": 8.843272396479638e-06, "loss": 1.4645, "step": 11354 }, { "epoch": 0.8898902821316614, "grad_norm": 2.562069892883301, "learning_rate": 8.843075141000637e-06, "loss": 1.2571, "step": 11355 }, { "epoch": 0.8899686520376175, "grad_norm": 3.729079484939575, "learning_rate": 8.84287787090455e-06, "loss": 1.5706, "step": 11356 }, { "epoch": 0.8900470219435737, "grad_norm": 4.457377910614014, "learning_rate": 8.842680586192129e-06, "loss": 1.0157, "step": 11357 }, { "epoch": 0.8901253918495298, "grad_norm": 3.104574680328369, "learning_rate": 8.842483286864125e-06, "loss": 1.205, "step": 11358 }, { "epoch": 0.8902037617554859, "grad_norm": 2.9138545989990234, "learning_rate": 8.84228597292129e-06, "loss": 1.3661, "step": 11359 }, { "epoch": 0.890282131661442, "grad_norm": 4.703611373901367, "learning_rate": 8.84208864436437e-06, "loss": 1.4171, "step": 11360 }, { "epoch": 0.8903605015673981, "grad_norm": 2.35567045211792, "learning_rate": 8.841891301194121e-06, "loss": 1.5141, "step": 11361 }, { "epoch": 0.8904388714733542, "grad_norm": 2.007101058959961, "learning_rate": 8.84169394341129e-06, "loss": 1.1186, "step": 11362 }, { "epoch": 0.8905172413793103, "grad_norm": 2.099949836730957, "learning_rate": 8.841496571016627e-06, "loss": 1.5032, "step": 11363 }, { "epoch": 0.8905956112852664, "grad_norm": 1.8350468873977661, "learning_rate": 8.841299184010886e-06, "loss": 1.5503, "step": 11364 }, { "epoch": 0.8906739811912225, "grad_norm": 3.3489034175872803, "learning_rate": 8.841101782394815e-06, "loss": 1.3202, "step": 11365 }, { "epoch": 0.8907523510971787, "grad_norm": 2.401684522628784, "learning_rate": 8.840904366169165e-06, "loss": 1.7598, "step": 11366 }, { "epoch": 0.8908307210031348, "grad_norm": 3.8159916400909424, "learning_rate": 8.840706935334688e-06, "loss": 1.3466, "step": 11367 }, { "epoch": 0.8909090909090909, "grad_norm": 3.6186468601226807, "learning_rate": 8.840509489892134e-06, "loss": 0.664, "step": 11368 }, { "epoch": 0.890987460815047, "grad_norm": 2.5771124362945557, "learning_rate": 8.840312029842257e-06, "loss": 0.9648, "step": 11369 }, { "epoch": 0.8910658307210031, "grad_norm": 1.6979414224624634, "learning_rate": 8.840114555185804e-06, "loss": 1.0349, "step": 11370 }, { "epoch": 0.8911442006269592, "grad_norm": 1.9481847286224365, "learning_rate": 8.839917065923527e-06, "loss": 1.359, "step": 11371 }, { "epoch": 0.8912225705329153, "grad_norm": 4.842806816101074, "learning_rate": 8.839719562056178e-06, "loss": 1.4801, "step": 11372 }, { "epoch": 0.8913009404388714, "grad_norm": 3.074071168899536, "learning_rate": 8.839522043584508e-06, "loss": 1.3865, "step": 11373 }, { "epoch": 0.8913793103448275, "grad_norm": 4.6056294441223145, "learning_rate": 8.839324510509269e-06, "loss": 1.6559, "step": 11374 }, { "epoch": 0.8914576802507836, "grad_norm": 2.463930130004883, "learning_rate": 8.83912696283121e-06, "loss": 1.3057, "step": 11375 }, { "epoch": 0.8915360501567398, "grad_norm": 3.3866758346557617, "learning_rate": 8.838929400551087e-06, "loss": 1.0585, "step": 11376 }, { "epoch": 0.891614420062696, "grad_norm": 2.7909018993377686, "learning_rate": 8.838731823669646e-06, "loss": 1.7694, "step": 11377 }, { "epoch": 0.8916927899686521, "grad_norm": 3.4046144485473633, "learning_rate": 8.838534232187641e-06, "loss": 1.3774, "step": 11378 }, { "epoch": 0.8917711598746082, "grad_norm": 2.7158565521240234, "learning_rate": 8.838336626105823e-06, "loss": 1.3966, "step": 11379 }, { "epoch": 0.8918495297805643, "grad_norm": 3.679774761199951, "learning_rate": 8.838139005424945e-06, "loss": 0.841, "step": 11380 }, { "epoch": 0.8919278996865204, "grad_norm": 6.214791774749756, "learning_rate": 8.837941370145757e-06, "loss": 1.7579, "step": 11381 }, { "epoch": 0.8920062695924765, "grad_norm": 2.628429651260376, "learning_rate": 8.837743720269011e-06, "loss": 1.5693, "step": 11382 }, { "epoch": 0.8920846394984326, "grad_norm": 3.9331741333007812, "learning_rate": 8.83754605579546e-06, "loss": 1.5799, "step": 11383 }, { "epoch": 0.8921630094043888, "grad_norm": 2.5540058612823486, "learning_rate": 8.837348376725856e-06, "loss": 1.2496, "step": 11384 }, { "epoch": 0.8922413793103449, "grad_norm": 2.1126434803009033, "learning_rate": 8.837150683060947e-06, "loss": 1.3587, "step": 11385 }, { "epoch": 0.892319749216301, "grad_norm": 2.7272799015045166, "learning_rate": 8.83695297480149e-06, "loss": 1.4492, "step": 11386 }, { "epoch": 0.8923981191222571, "grad_norm": 2.0812339782714844, "learning_rate": 8.836755251948232e-06, "loss": 1.1526, "step": 11387 }, { "epoch": 0.8924764890282132, "grad_norm": 2.639214038848877, "learning_rate": 8.83655751450193e-06, "loss": 1.3142, "step": 11388 }, { "epoch": 0.8925548589341693, "grad_norm": 2.0618345737457275, "learning_rate": 8.836359762463333e-06, "loss": 1.1369, "step": 11389 }, { "epoch": 0.8926332288401254, "grad_norm": 4.637100696563721, "learning_rate": 8.836161995833194e-06, "loss": 1.6849, "step": 11390 }, { "epoch": 0.8927115987460815, "grad_norm": 3.5396580696105957, "learning_rate": 8.835964214612267e-06, "loss": 1.4543, "step": 11391 }, { "epoch": 0.8927899686520376, "grad_norm": 3.9237654209136963, "learning_rate": 8.8357664188013e-06, "loss": 1.0638, "step": 11392 }, { "epoch": 0.8928683385579937, "grad_norm": 2.527334690093994, "learning_rate": 8.835568608401048e-06, "loss": 1.1452, "step": 11393 }, { "epoch": 0.8929467084639499, "grad_norm": 2.0851452350616455, "learning_rate": 8.835370783412263e-06, "loss": 1.1926, "step": 11394 }, { "epoch": 0.893025078369906, "grad_norm": 2.5821590423583984, "learning_rate": 8.835172943835699e-06, "loss": 1.2992, "step": 11395 }, { "epoch": 0.8931034482758621, "grad_norm": 6.509283065795898, "learning_rate": 8.834975089672106e-06, "loss": 1.8867, "step": 11396 }, { "epoch": 0.8931818181818182, "grad_norm": 2.410200357437134, "learning_rate": 8.834777220922236e-06, "loss": 1.3523, "step": 11397 }, { "epoch": 0.8932601880877743, "grad_norm": 5.599639892578125, "learning_rate": 8.834579337586846e-06, "loss": 2.1419, "step": 11398 }, { "epoch": 0.8933385579937304, "grad_norm": 3.2768378257751465, "learning_rate": 8.834381439666684e-06, "loss": 1.5289, "step": 11399 }, { "epoch": 0.8934169278996865, "grad_norm": 6.731691837310791, "learning_rate": 8.834183527162504e-06, "loss": 1.3057, "step": 11400 }, { "epoch": 0.8934952978056426, "grad_norm": 2.2638955116271973, "learning_rate": 8.833985600075061e-06, "loss": 1.4732, "step": 11401 }, { "epoch": 0.8935736677115987, "grad_norm": 2.2873237133026123, "learning_rate": 8.833787658405104e-06, "loss": 1.23, "step": 11402 }, { "epoch": 0.8936520376175549, "grad_norm": 2.3518431186676025, "learning_rate": 8.833589702153389e-06, "loss": 1.4956, "step": 11403 }, { "epoch": 0.893730407523511, "grad_norm": 3.4410431385040283, "learning_rate": 8.833391731320669e-06, "loss": 2.0045, "step": 11404 }, { "epoch": 0.8938087774294671, "grad_norm": 5.614917278289795, "learning_rate": 8.833193745907694e-06, "loss": 0.5879, "step": 11405 }, { "epoch": 0.8938871473354232, "grad_norm": 3.9526619911193848, "learning_rate": 8.83299574591522e-06, "loss": 1.9581, "step": 11406 }, { "epoch": 0.8939655172413793, "grad_norm": 6.731198787689209, "learning_rate": 8.832797731343998e-06, "loss": 1.8543, "step": 11407 }, { "epoch": 0.8940438871473354, "grad_norm": 2.434342384338379, "learning_rate": 8.832599702194782e-06, "loss": 1.2464, "step": 11408 }, { "epoch": 0.8941222570532915, "grad_norm": 2.762779951095581, "learning_rate": 8.832401658468326e-06, "loss": 1.2135, "step": 11409 }, { "epoch": 0.8942006269592476, "grad_norm": 2.6103475093841553, "learning_rate": 8.832203600165383e-06, "loss": 1.2471, "step": 11410 }, { "epoch": 0.8942789968652037, "grad_norm": 4.867839336395264, "learning_rate": 8.832005527286705e-06, "loss": 1.4378, "step": 11411 }, { "epoch": 0.8943573667711598, "grad_norm": 2.802293300628662, "learning_rate": 8.831807439833047e-06, "loss": 1.3563, "step": 11412 }, { "epoch": 0.894435736677116, "grad_norm": 2.5001602172851562, "learning_rate": 8.83160933780516e-06, "loss": 1.3146, "step": 11413 }, { "epoch": 0.8945141065830721, "grad_norm": 2.284433364868164, "learning_rate": 8.831411221203801e-06, "loss": 1.5736, "step": 11414 }, { "epoch": 0.8945924764890282, "grad_norm": 2.384598731994629, "learning_rate": 8.83121309002972e-06, "loss": 1.271, "step": 11415 }, { "epoch": 0.8946708463949843, "grad_norm": 3.1201114654541016, "learning_rate": 8.831014944283674e-06, "loss": 1.2677, "step": 11416 }, { "epoch": 0.8947492163009404, "grad_norm": 12.799771308898926, "learning_rate": 8.830816783966414e-06, "loss": 1.6017, "step": 11417 }, { "epoch": 0.8948275862068965, "grad_norm": 2.2106244564056396, "learning_rate": 8.830618609078694e-06, "loss": 1.4165, "step": 11418 }, { "epoch": 0.8949059561128526, "grad_norm": 2.929769277572632, "learning_rate": 8.830420419621268e-06, "loss": 1.2103, "step": 11419 }, { "epoch": 0.8949843260188087, "grad_norm": 2.929635763168335, "learning_rate": 8.83022221559489e-06, "loss": 1.4448, "step": 11420 }, { "epoch": 0.8950626959247648, "grad_norm": 3.6314995288848877, "learning_rate": 8.830023997000317e-06, "loss": 1.4155, "step": 11421 }, { "epoch": 0.895141065830721, "grad_norm": 2.5129239559173584, "learning_rate": 8.829825763838298e-06, "loss": 1.1942, "step": 11422 }, { "epoch": 0.8952194357366771, "grad_norm": 4.527091026306152, "learning_rate": 8.829627516109589e-06, "loss": 1.3286, "step": 11423 }, { "epoch": 0.8952978056426333, "grad_norm": 10.8650484085083, "learning_rate": 8.829429253814942e-06, "loss": 0.4994, "step": 11424 }, { "epoch": 0.8953761755485894, "grad_norm": 2.5134551525115967, "learning_rate": 8.829230976955115e-06, "loss": 1.1974, "step": 11425 }, { "epoch": 0.8954545454545455, "grad_norm": 2.812875509262085, "learning_rate": 8.829032685530861e-06, "loss": 1.4807, "step": 11426 }, { "epoch": 0.8955329153605016, "grad_norm": 3.5430519580841064, "learning_rate": 8.828834379542932e-06, "loss": 1.2495, "step": 11427 }, { "epoch": 0.8956112852664577, "grad_norm": 12.202320098876953, "learning_rate": 8.828636058992085e-06, "loss": 2.0021, "step": 11428 }, { "epoch": 0.8956896551724138, "grad_norm": 2.242774248123169, "learning_rate": 8.828437723879071e-06, "loss": 1.3921, "step": 11429 }, { "epoch": 0.89576802507837, "grad_norm": 5.354654788970947, "learning_rate": 8.828239374204646e-06, "loss": 1.5216, "step": 11430 }, { "epoch": 0.895846394984326, "grad_norm": 4.6841301918029785, "learning_rate": 8.828041009969567e-06, "loss": 0.8243, "step": 11431 }, { "epoch": 0.8959247648902822, "grad_norm": 6.349275588989258, "learning_rate": 8.827842631174585e-06, "loss": 1.3306, "step": 11432 }, { "epoch": 0.8960031347962383, "grad_norm": 2.476228952407837, "learning_rate": 8.827644237820457e-06, "loss": 1.1508, "step": 11433 }, { "epoch": 0.8960815047021944, "grad_norm": 5.6137614250183105, "learning_rate": 8.827445829907934e-06, "loss": 1.666, "step": 11434 }, { "epoch": 0.8961598746081505, "grad_norm": 4.7746686935424805, "learning_rate": 8.827247407437775e-06, "loss": 1.3337, "step": 11435 }, { "epoch": 0.8962382445141066, "grad_norm": 3.1788694858551025, "learning_rate": 8.827048970410731e-06, "loss": 1.4594, "step": 11436 }, { "epoch": 0.8963166144200627, "grad_norm": 2.640671730041504, "learning_rate": 8.82685051882756e-06, "loss": 1.1832, "step": 11437 }, { "epoch": 0.8963949843260188, "grad_norm": 3.1761972904205322, "learning_rate": 8.826652052689015e-06, "loss": 1.4138, "step": 11438 }, { "epoch": 0.8964733542319749, "grad_norm": 2.574676990509033, "learning_rate": 8.82645357199585e-06, "loss": 1.11, "step": 11439 }, { "epoch": 0.896551724137931, "grad_norm": 2.13785982131958, "learning_rate": 8.826255076748823e-06, "loss": 1.6411, "step": 11440 }, { "epoch": 0.8966300940438872, "grad_norm": 2.8585968017578125, "learning_rate": 8.826056566948684e-06, "loss": 1.5015, "step": 11441 }, { "epoch": 0.8967084639498433, "grad_norm": 2.557173490524292, "learning_rate": 8.825858042596194e-06, "loss": 1.4581, "step": 11442 }, { "epoch": 0.8967868338557994, "grad_norm": 1.9138164520263672, "learning_rate": 8.825659503692103e-06, "loss": 1.0656, "step": 11443 }, { "epoch": 0.8968652037617555, "grad_norm": 4.095122814178467, "learning_rate": 8.825460950237167e-06, "loss": 1.2357, "step": 11444 }, { "epoch": 0.8969435736677116, "grad_norm": 3.4730873107910156, "learning_rate": 8.825262382232146e-06, "loss": 1.3453, "step": 11445 }, { "epoch": 0.8970219435736677, "grad_norm": 4.0156636238098145, "learning_rate": 8.82506379967779e-06, "loss": 1.6613, "step": 11446 }, { "epoch": 0.8971003134796238, "grad_norm": 4.2946553230285645, "learning_rate": 8.824865202574855e-06, "loss": 1.9059, "step": 11447 }, { "epoch": 0.8971786833855799, "grad_norm": 3.8891146183013916, "learning_rate": 8.8246665909241e-06, "loss": 1.3496, "step": 11448 }, { "epoch": 0.897257053291536, "grad_norm": 5.540561676025391, "learning_rate": 8.824467964726276e-06, "loss": 1.3616, "step": 11449 }, { "epoch": 0.8973354231974922, "grad_norm": 2.4676284790039062, "learning_rate": 8.82426932398214e-06, "loss": 1.4364, "step": 11450 }, { "epoch": 0.8974137931034483, "grad_norm": 2.3910791873931885, "learning_rate": 8.824070668692448e-06, "loss": 1.1654, "step": 11451 }, { "epoch": 0.8974921630094044, "grad_norm": 2.205876350402832, "learning_rate": 8.823871998857956e-06, "loss": 1.3201, "step": 11452 }, { "epoch": 0.8975705329153605, "grad_norm": 7.294281959533691, "learning_rate": 8.823673314479418e-06, "loss": 0.9532, "step": 11453 }, { "epoch": 0.8976489028213166, "grad_norm": 2.3070411682128906, "learning_rate": 8.82347461555759e-06, "loss": 1.2047, "step": 11454 }, { "epoch": 0.8977272727272727, "grad_norm": 2.2211334705352783, "learning_rate": 8.823275902093232e-06, "loss": 1.6787, "step": 11455 }, { "epoch": 0.8978056426332288, "grad_norm": 10.019740104675293, "learning_rate": 8.823077174087093e-06, "loss": 1.22, "step": 11456 }, { "epoch": 0.8978840125391849, "grad_norm": 12.347541809082031, "learning_rate": 8.822878431539932e-06, "loss": 1.7519, "step": 11457 }, { "epoch": 0.897962382445141, "grad_norm": 5.517566680908203, "learning_rate": 8.822679674452505e-06, "loss": 1.2407, "step": 11458 }, { "epoch": 0.8980407523510971, "grad_norm": 2.9120750427246094, "learning_rate": 8.82248090282557e-06, "loss": 1.3611, "step": 11459 }, { "epoch": 0.8981191222570533, "grad_norm": 10.256277084350586, "learning_rate": 8.822282116659879e-06, "loss": 1.3453, "step": 11460 }, { "epoch": 0.8981974921630094, "grad_norm": 3.0991575717926025, "learning_rate": 8.822083315956191e-06, "loss": 1.4054, "step": 11461 }, { "epoch": 0.8982758620689655, "grad_norm": 5.30595064163208, "learning_rate": 8.82188450071526e-06, "loss": 1.8018, "step": 11462 }, { "epoch": 0.8983542319749216, "grad_norm": 5.923596382141113, "learning_rate": 8.821685670937846e-06, "loss": 1.4248, "step": 11463 }, { "epoch": 0.8984326018808777, "grad_norm": 3.4337830543518066, "learning_rate": 8.821486826624702e-06, "loss": 1.3227, "step": 11464 }, { "epoch": 0.8985109717868338, "grad_norm": 3.048088788986206, "learning_rate": 8.821287967776583e-06, "loss": 1.5523, "step": 11465 }, { "epoch": 0.8985893416927899, "grad_norm": 2.952562093734741, "learning_rate": 8.821089094394249e-06, "loss": 1.6489, "step": 11466 }, { "epoch": 0.898667711598746, "grad_norm": 2.0206685066223145, "learning_rate": 8.820890206478453e-06, "loss": 1.404, "step": 11467 }, { "epoch": 0.8987460815047021, "grad_norm": 3.4224085807800293, "learning_rate": 8.820691304029954e-06, "loss": 1.4661, "step": 11468 }, { "epoch": 0.8988244514106583, "grad_norm": 2.77337646484375, "learning_rate": 8.82049238704951e-06, "loss": 0.7217, "step": 11469 }, { "epoch": 0.8989028213166145, "grad_norm": 3.044301748275757, "learning_rate": 8.820293455537872e-06, "loss": 0.9807, "step": 11470 }, { "epoch": 0.8989811912225706, "grad_norm": 5.150141716003418, "learning_rate": 8.820094509495802e-06, "loss": 1.6031, "step": 11471 }, { "epoch": 0.8990595611285267, "grad_norm": 2.1993930339813232, "learning_rate": 8.819895548924053e-06, "loss": 1.5162, "step": 11472 }, { "epoch": 0.8991379310344828, "grad_norm": 2.437694549560547, "learning_rate": 8.819696573823384e-06, "loss": 1.3264, "step": 11473 }, { "epoch": 0.8992163009404389, "grad_norm": 2.9187228679656982, "learning_rate": 8.819497584194552e-06, "loss": 1.6647, "step": 11474 }, { "epoch": 0.899294670846395, "grad_norm": 2.264050245285034, "learning_rate": 8.819298580038313e-06, "loss": 1.5174, "step": 11475 }, { "epoch": 0.8993730407523511, "grad_norm": 2.41640567779541, "learning_rate": 8.819099561355423e-06, "loss": 1.6403, "step": 11476 }, { "epoch": 0.8994514106583072, "grad_norm": 6.553354263305664, "learning_rate": 8.81890052814664e-06, "loss": 1.405, "step": 11477 }, { "epoch": 0.8995297805642634, "grad_norm": 2.5796265602111816, "learning_rate": 8.818701480412722e-06, "loss": 1.3401, "step": 11478 }, { "epoch": 0.8996081504702195, "grad_norm": 2.796123743057251, "learning_rate": 8.818502418154424e-06, "loss": 1.3426, "step": 11479 }, { "epoch": 0.8996865203761756, "grad_norm": 2.0334744453430176, "learning_rate": 8.818303341372502e-06, "loss": 1.3868, "step": 11480 }, { "epoch": 0.8997648902821317, "grad_norm": 2.0187535285949707, "learning_rate": 8.818104250067718e-06, "loss": 1.2465, "step": 11481 }, { "epoch": 0.8998432601880878, "grad_norm": 2.7012836933135986, "learning_rate": 8.817905144240827e-06, "loss": 1.399, "step": 11482 }, { "epoch": 0.8999216300940439, "grad_norm": 3.1860015392303467, "learning_rate": 8.817706023892585e-06, "loss": 1.7125, "step": 11483 }, { "epoch": 0.9, "grad_norm": 5.413716793060303, "learning_rate": 8.81750688902375e-06, "loss": 1.8222, "step": 11484 }, { "epoch": 0.9000783699059561, "grad_norm": 4.953280925750732, "learning_rate": 8.817307739635079e-06, "loss": 0.862, "step": 11485 }, { "epoch": 0.9001567398119122, "grad_norm": 2.0895283222198486, "learning_rate": 8.817108575727331e-06, "loss": 1.2792, "step": 11486 }, { "epoch": 0.9002351097178684, "grad_norm": 3.9614784717559814, "learning_rate": 8.816909397301261e-06, "loss": 1.6255, "step": 11487 }, { "epoch": 0.9003134796238245, "grad_norm": 1.715895175933838, "learning_rate": 8.81671020435763e-06, "loss": 1.354, "step": 11488 }, { "epoch": 0.9003918495297806, "grad_norm": 2.311980724334717, "learning_rate": 8.816510996897194e-06, "loss": 1.8815, "step": 11489 }, { "epoch": 0.9004702194357367, "grad_norm": 3.133244514465332, "learning_rate": 8.81631177492071e-06, "loss": 1.099, "step": 11490 }, { "epoch": 0.9005485893416928, "grad_norm": 2.8904812335968018, "learning_rate": 8.816112538428935e-06, "loss": 1.1171, "step": 11491 }, { "epoch": 0.9006269592476489, "grad_norm": 5.699036598205566, "learning_rate": 8.815913287422628e-06, "loss": 1.1325, "step": 11492 }, { "epoch": 0.900705329153605, "grad_norm": 3.0294017791748047, "learning_rate": 8.815714021902548e-06, "loss": 1.2505, "step": 11493 }, { "epoch": 0.9007836990595611, "grad_norm": 2.882457733154297, "learning_rate": 8.81551474186945e-06, "loss": 1.1507, "step": 11494 }, { "epoch": 0.9008620689655172, "grad_norm": 6.536410331726074, "learning_rate": 8.815315447324096e-06, "loss": 0.7338, "step": 11495 }, { "epoch": 0.9009404388714733, "grad_norm": 14.455535888671875, "learning_rate": 8.815116138267241e-06, "loss": 0.8764, "step": 11496 }, { "epoch": 0.9010188087774295, "grad_norm": 2.1136274337768555, "learning_rate": 8.814916814699643e-06, "loss": 1.308, "step": 11497 }, { "epoch": 0.9010971786833856, "grad_norm": 3.53605318069458, "learning_rate": 8.814717476622062e-06, "loss": 1.5518, "step": 11498 }, { "epoch": 0.9011755485893417, "grad_norm": 6.3714118003845215, "learning_rate": 8.814518124035254e-06, "loss": 1.3025, "step": 11499 }, { "epoch": 0.9012539184952978, "grad_norm": 2.2455832958221436, "learning_rate": 8.814318756939979e-06, "loss": 1.3322, "step": 11500 }, { "epoch": 0.9013322884012539, "grad_norm": 3.7967846393585205, "learning_rate": 8.814119375336993e-06, "loss": 1.5161, "step": 11501 }, { "epoch": 0.90141065830721, "grad_norm": 5.823702812194824, "learning_rate": 8.81391997922706e-06, "loss": 1.6742, "step": 11502 }, { "epoch": 0.9014890282131661, "grad_norm": 2.28515362739563, "learning_rate": 8.81372056861093e-06, "loss": 1.3256, "step": 11503 }, { "epoch": 0.9015673981191222, "grad_norm": 2.387172222137451, "learning_rate": 8.813521143489367e-06, "loss": 1.5046, "step": 11504 }, { "epoch": 0.9016457680250783, "grad_norm": 3.2037856578826904, "learning_rate": 8.813321703863129e-06, "loss": 1.1589, "step": 11505 }, { "epoch": 0.9017241379310345, "grad_norm": 2.549802303314209, "learning_rate": 8.813122249732974e-06, "loss": 1.3873, "step": 11506 }, { "epoch": 0.9018025078369906, "grad_norm": 3.3399996757507324, "learning_rate": 8.81292278109966e-06, "loss": 1.6992, "step": 11507 }, { "epoch": 0.9018808777429467, "grad_norm": 3.121830701828003, "learning_rate": 8.812723297963947e-06, "loss": 1.4898, "step": 11508 }, { "epoch": 0.9019592476489028, "grad_norm": 3.2934823036193848, "learning_rate": 8.812523800326594e-06, "loss": 1.6749, "step": 11509 }, { "epoch": 0.9020376175548589, "grad_norm": 2.27699875831604, "learning_rate": 8.812324288188355e-06, "loss": 1.433, "step": 11510 }, { "epoch": 0.902115987460815, "grad_norm": 4.069077491760254, "learning_rate": 8.812124761549995e-06, "loss": 1.096, "step": 11511 }, { "epoch": 0.9021943573667711, "grad_norm": 2.361450672149658, "learning_rate": 8.81192522041227e-06, "loss": 1.7841, "step": 11512 }, { "epoch": 0.9022727272727272, "grad_norm": 2.72875714302063, "learning_rate": 8.811725664775939e-06, "loss": 1.4242, "step": 11513 }, { "epoch": 0.9023510971786833, "grad_norm": 2.3565802574157715, "learning_rate": 8.811526094641762e-06, "loss": 1.4238, "step": 11514 }, { "epoch": 0.9024294670846394, "grad_norm": 3.789355516433716, "learning_rate": 8.811326510010497e-06, "loss": 1.4976, "step": 11515 }, { "epoch": 0.9025078369905956, "grad_norm": 2.5749287605285645, "learning_rate": 8.811126910882903e-06, "loss": 1.1079, "step": 11516 }, { "epoch": 0.9025862068965518, "grad_norm": 2.3672702312469482, "learning_rate": 8.810927297259741e-06, "loss": 1.4678, "step": 11517 }, { "epoch": 0.9026645768025079, "grad_norm": 6.028843402862549, "learning_rate": 8.81072766914177e-06, "loss": 1.0593, "step": 11518 }, { "epoch": 0.902742946708464, "grad_norm": 2.849627733230591, "learning_rate": 8.810528026529746e-06, "loss": 1.3688, "step": 11519 }, { "epoch": 0.9028213166144201, "grad_norm": 6.799354553222656, "learning_rate": 8.81032836942443e-06, "loss": 1.3779, "step": 11520 }, { "epoch": 0.9028996865203762, "grad_norm": 2.3820877075195312, "learning_rate": 8.810128697826585e-06, "loss": 1.1529, "step": 11521 }, { "epoch": 0.9029780564263323, "grad_norm": 2.745147943496704, "learning_rate": 8.809929011736966e-06, "loss": 1.4156, "step": 11522 }, { "epoch": 0.9030564263322884, "grad_norm": 3.120431423187256, "learning_rate": 8.809729311156334e-06, "loss": 1.4992, "step": 11523 }, { "epoch": 0.9031347962382446, "grad_norm": 3.462071180343628, "learning_rate": 8.809529596085447e-06, "loss": 1.2241, "step": 11524 }, { "epoch": 0.9032131661442007, "grad_norm": 4.788941383361816, "learning_rate": 8.809329866525067e-06, "loss": 1.6201, "step": 11525 }, { "epoch": 0.9032915360501568, "grad_norm": 3.1452159881591797, "learning_rate": 8.809130122475954e-06, "loss": 1.5522, "step": 11526 }, { "epoch": 0.9033699059561129, "grad_norm": 4.740380764007568, "learning_rate": 8.808930363938866e-06, "loss": 1.3546, "step": 11527 }, { "epoch": 0.903448275862069, "grad_norm": 7.664656162261963, "learning_rate": 8.808730590914563e-06, "loss": 0.7338, "step": 11528 }, { "epoch": 0.9035266457680251, "grad_norm": 2.542790651321411, "learning_rate": 8.808530803403806e-06, "loss": 1.2105, "step": 11529 }, { "epoch": 0.9036050156739812, "grad_norm": 2.413829803466797, "learning_rate": 8.808331001407352e-06, "loss": 1.4825, "step": 11530 }, { "epoch": 0.9036833855799373, "grad_norm": 3.1103272438049316, "learning_rate": 8.808131184925964e-06, "loss": 1.335, "step": 11531 }, { "epoch": 0.9037617554858934, "grad_norm": 2.776777744293213, "learning_rate": 8.807931353960402e-06, "loss": 1.5014, "step": 11532 }, { "epoch": 0.9038401253918495, "grad_norm": 2.9844348430633545, "learning_rate": 8.807731508511424e-06, "loss": 1.5503, "step": 11533 }, { "epoch": 0.9039184952978057, "grad_norm": 2.0684804916381836, "learning_rate": 8.80753164857979e-06, "loss": 1.3524, "step": 11534 }, { "epoch": 0.9039968652037618, "grad_norm": 3.026931047439575, "learning_rate": 8.807331774166263e-06, "loss": 1.5079, "step": 11535 }, { "epoch": 0.9040752351097179, "grad_norm": 2.963019371032715, "learning_rate": 8.8071318852716e-06, "loss": 1.7048, "step": 11536 }, { "epoch": 0.904153605015674, "grad_norm": 3.596863269805908, "learning_rate": 8.806931981896564e-06, "loss": 1.6572, "step": 11537 }, { "epoch": 0.9042319749216301, "grad_norm": 1.7980910539627075, "learning_rate": 8.806732064041914e-06, "loss": 1.3701, "step": 11538 }, { "epoch": 0.9043103448275862, "grad_norm": 4.115890026092529, "learning_rate": 8.80653213170841e-06, "loss": 1.6307, "step": 11539 }, { "epoch": 0.9043887147335423, "grad_norm": 3.2465660572052, "learning_rate": 8.806332184896813e-06, "loss": 1.2176, "step": 11540 }, { "epoch": 0.9044670846394984, "grad_norm": 3.6761794090270996, "learning_rate": 8.806132223607883e-06, "loss": 1.7025, "step": 11541 }, { "epoch": 0.9045454545454545, "grad_norm": 2.6048262119293213, "learning_rate": 8.805932247842383e-06, "loss": 1.297, "step": 11542 }, { "epoch": 0.9046238244514107, "grad_norm": 4.137948036193848, "learning_rate": 8.80573225760107e-06, "loss": 1.0918, "step": 11543 }, { "epoch": 0.9047021943573668, "grad_norm": 9.933605194091797, "learning_rate": 8.805532252884705e-06, "loss": 1.5361, "step": 11544 }, { "epoch": 0.9047805642633229, "grad_norm": 5.498217582702637, "learning_rate": 8.805332233694053e-06, "loss": 0.8901, "step": 11545 }, { "epoch": 0.904858934169279, "grad_norm": 2.950120449066162, "learning_rate": 8.80513220002987e-06, "loss": 1.43, "step": 11546 }, { "epoch": 0.9049373040752351, "grad_norm": 3.7204110622406006, "learning_rate": 8.804932151892918e-06, "loss": 1.6396, "step": 11547 }, { "epoch": 0.9050156739811912, "grad_norm": 1.874651312828064, "learning_rate": 8.804732089283959e-06, "loss": 1.1465, "step": 11548 }, { "epoch": 0.9050940438871473, "grad_norm": 3.289612293243408, "learning_rate": 8.804532012203754e-06, "loss": 1.0675, "step": 11549 }, { "epoch": 0.9051724137931034, "grad_norm": 3.075643301010132, "learning_rate": 8.804331920653063e-06, "loss": 1.4583, "step": 11550 }, { "epoch": 0.9052507836990595, "grad_norm": 2.1153292655944824, "learning_rate": 8.804131814632647e-06, "loss": 1.4031, "step": 11551 }, { "epoch": 0.9053291536050156, "grad_norm": 4.732600688934326, "learning_rate": 8.803931694143267e-06, "loss": 1.409, "step": 11552 }, { "epoch": 0.9054075235109718, "grad_norm": 3.250702381134033, "learning_rate": 8.803731559185686e-06, "loss": 1.8, "step": 11553 }, { "epoch": 0.9054858934169279, "grad_norm": 2.4007036685943604, "learning_rate": 8.803531409760663e-06, "loss": 1.1849, "step": 11554 }, { "epoch": 0.905564263322884, "grad_norm": 2.6635236740112305, "learning_rate": 8.80333124586896e-06, "loss": 1.8505, "step": 11555 }, { "epoch": 0.9056426332288401, "grad_norm": 4.477409839630127, "learning_rate": 8.803131067511338e-06, "loss": 1.3243, "step": 11556 }, { "epoch": 0.9057210031347962, "grad_norm": 3.719958543777466, "learning_rate": 8.802930874688562e-06, "loss": 1.9054, "step": 11557 }, { "epoch": 0.9057993730407523, "grad_norm": 2.4799838066101074, "learning_rate": 8.802730667401386e-06, "loss": 1.0987, "step": 11558 }, { "epoch": 0.9058777429467084, "grad_norm": 2.921482563018799, "learning_rate": 8.802530445650579e-06, "loss": 1.4695, "step": 11559 }, { "epoch": 0.9059561128526645, "grad_norm": 3.7409474849700928, "learning_rate": 8.802330209436898e-06, "loss": 0.741, "step": 11560 }, { "epoch": 0.9060344827586206, "grad_norm": 2.9328737258911133, "learning_rate": 8.802129958761105e-06, "loss": 1.2853, "step": 11561 }, { "epoch": 0.9061128526645768, "grad_norm": 4.03608512878418, "learning_rate": 8.801929693623963e-06, "loss": 1.4237, "step": 11562 }, { "epoch": 0.9061912225705329, "grad_norm": 1.905802607536316, "learning_rate": 8.801729414026234e-06, "loss": 1.0789, "step": 11563 }, { "epoch": 0.9062695924764891, "grad_norm": 5.505167007446289, "learning_rate": 8.80152911996868e-06, "loss": 1.1717, "step": 11564 }, { "epoch": 0.9063479623824452, "grad_norm": 13.948257446289062, "learning_rate": 8.80132881145206e-06, "loss": 1.2467, "step": 11565 }, { "epoch": 0.9064263322884013, "grad_norm": 2.0799076557159424, "learning_rate": 8.801128488477137e-06, "loss": 1.0493, "step": 11566 }, { "epoch": 0.9065047021943574, "grad_norm": 2.9644994735717773, "learning_rate": 8.800928151044674e-06, "loss": 1.3595, "step": 11567 }, { "epoch": 0.9065830721003135, "grad_norm": 3.1940207481384277, "learning_rate": 8.800727799155434e-06, "loss": 1.4532, "step": 11568 }, { "epoch": 0.9066614420062696, "grad_norm": 4.129640102386475, "learning_rate": 8.800527432810177e-06, "loss": 1.3025, "step": 11569 }, { "epoch": 0.9067398119122257, "grad_norm": 2.7019779682159424, "learning_rate": 8.800327052009666e-06, "loss": 0.9616, "step": 11570 }, { "epoch": 0.9068181818181819, "grad_norm": 2.873049259185791, "learning_rate": 8.800126656754661e-06, "loss": 1.6233, "step": 11571 }, { "epoch": 0.906896551724138, "grad_norm": 3.2331418991088867, "learning_rate": 8.799926247045929e-06, "loss": 1.4093, "step": 11572 }, { "epoch": 0.9069749216300941, "grad_norm": 3.328606367111206, "learning_rate": 8.79972582288423e-06, "loss": 1.3486, "step": 11573 }, { "epoch": 0.9070532915360502, "grad_norm": 2.3891024589538574, "learning_rate": 8.799525384270323e-06, "loss": 1.3317, "step": 11574 }, { "epoch": 0.9071316614420063, "grad_norm": 1.9714910984039307, "learning_rate": 8.799324931204972e-06, "loss": 1.6873, "step": 11575 }, { "epoch": 0.9072100313479624, "grad_norm": 3.085380792617798, "learning_rate": 8.799124463688942e-06, "loss": 1.633, "step": 11576 }, { "epoch": 0.9072884012539185, "grad_norm": 1.6306782960891724, "learning_rate": 8.798923981722995e-06, "loss": 1.0215, "step": 11577 }, { "epoch": 0.9073667711598746, "grad_norm": 3.144719362258911, "learning_rate": 8.79872348530789e-06, "loss": 1.1003, "step": 11578 }, { "epoch": 0.9074451410658307, "grad_norm": 4.339900016784668, "learning_rate": 8.798522974444393e-06, "loss": 1.0296, "step": 11579 }, { "epoch": 0.9075235109717869, "grad_norm": 3.4297821521759033, "learning_rate": 8.798322449133266e-06, "loss": 1.2709, "step": 11580 }, { "epoch": 0.907601880877743, "grad_norm": 3.839866876602173, "learning_rate": 8.798121909375272e-06, "loss": 1.6032, "step": 11581 }, { "epoch": 0.9076802507836991, "grad_norm": 2.899887800216675, "learning_rate": 8.797921355171172e-06, "loss": 1.4196, "step": 11582 }, { "epoch": 0.9077586206896552, "grad_norm": 3.91689395904541, "learning_rate": 8.79772078652173e-06, "loss": 1.278, "step": 11583 }, { "epoch": 0.9078369905956113, "grad_norm": 2.7761952877044678, "learning_rate": 8.79752020342771e-06, "loss": 1.691, "step": 11584 }, { "epoch": 0.9079153605015674, "grad_norm": 3.3272545337677, "learning_rate": 8.797319605889872e-06, "loss": 1.4873, "step": 11585 }, { "epoch": 0.9079937304075235, "grad_norm": 3.8239669799804688, "learning_rate": 8.797118993908982e-06, "loss": 1.4128, "step": 11586 }, { "epoch": 0.9080721003134796, "grad_norm": 9.256009101867676, "learning_rate": 8.796918367485801e-06, "loss": 0.8188, "step": 11587 }, { "epoch": 0.9081504702194357, "grad_norm": 2.203056812286377, "learning_rate": 8.796717726621093e-06, "loss": 1.3111, "step": 11588 }, { "epoch": 0.9082288401253918, "grad_norm": 1.818570852279663, "learning_rate": 8.79651707131562e-06, "loss": 1.3961, "step": 11589 }, { "epoch": 0.908307210031348, "grad_norm": 2.6158103942871094, "learning_rate": 8.796316401570146e-06, "loss": 1.4673, "step": 11590 }, { "epoch": 0.9083855799373041, "grad_norm": 3.1914379596710205, "learning_rate": 8.796115717385437e-06, "loss": 1.4032, "step": 11591 }, { "epoch": 0.9084639498432602, "grad_norm": 1.8215352296829224, "learning_rate": 8.795915018762252e-06, "loss": 1.2674, "step": 11592 }, { "epoch": 0.9085423197492163, "grad_norm": 2.3804235458374023, "learning_rate": 8.795714305701354e-06, "loss": 1.3218, "step": 11593 }, { "epoch": 0.9086206896551724, "grad_norm": 2.8287901878356934, "learning_rate": 8.795513578203509e-06, "loss": 1.3225, "step": 11594 }, { "epoch": 0.9086990595611285, "grad_norm": 2.5971519947052, "learning_rate": 8.795312836269483e-06, "loss": 1.229, "step": 11595 }, { "epoch": 0.9087774294670846, "grad_norm": 2.1976802349090576, "learning_rate": 8.795112079900033e-06, "loss": 1.0421, "step": 11596 }, { "epoch": 0.9088557993730407, "grad_norm": 4.38961124420166, "learning_rate": 8.794911309095929e-06, "loss": 1.0583, "step": 11597 }, { "epoch": 0.9089341692789968, "grad_norm": 3.3958489894866943, "learning_rate": 8.794710523857927e-06, "loss": 1.255, "step": 11598 }, { "epoch": 0.909012539184953, "grad_norm": 5.050680637359619, "learning_rate": 8.7945097241868e-06, "loss": 1.1612, "step": 11599 }, { "epoch": 0.9090909090909091, "grad_norm": 3.8062164783477783, "learning_rate": 8.794308910083305e-06, "loss": 1.3332, "step": 11600 }, { "epoch": 0.9091692789968652, "grad_norm": 7.768503189086914, "learning_rate": 8.794108081548207e-06, "loss": 1.7043, "step": 11601 }, { "epoch": 0.9092476489028213, "grad_norm": 2.769761085510254, "learning_rate": 8.79390723858227e-06, "loss": 1.5904, "step": 11602 }, { "epoch": 0.9093260188087774, "grad_norm": 2.7294206619262695, "learning_rate": 8.793706381186263e-06, "loss": 1.3929, "step": 11603 }, { "epoch": 0.9094043887147335, "grad_norm": 2.0631892681121826, "learning_rate": 8.79350550936094e-06, "loss": 1.4204, "step": 11604 }, { "epoch": 0.9094827586206896, "grad_norm": 2.5449721813201904, "learning_rate": 8.793304623107074e-06, "loss": 1.6064, "step": 11605 }, { "epoch": 0.9095611285266457, "grad_norm": 1.9943238496780396, "learning_rate": 8.793103722425424e-06, "loss": 1.377, "step": 11606 }, { "epoch": 0.9096394984326018, "grad_norm": 2.87091326713562, "learning_rate": 8.792902807316757e-06, "loss": 1.5287, "step": 11607 }, { "epoch": 0.909717868338558, "grad_norm": 8.622441291809082, "learning_rate": 8.792701877781835e-06, "loss": 1.06, "step": 11608 }, { "epoch": 0.909796238244514, "grad_norm": 10.181303024291992, "learning_rate": 8.792500933821423e-06, "loss": 1.1779, "step": 11609 }, { "epoch": 0.9098746081504702, "grad_norm": 3.2990050315856934, "learning_rate": 8.792299975436286e-06, "loss": 1.3213, "step": 11610 }, { "epoch": 0.9099529780564264, "grad_norm": 4.795421123504639, "learning_rate": 8.792099002627189e-06, "loss": 1.5701, "step": 11611 }, { "epoch": 0.9100313479623825, "grad_norm": 4.388187408447266, "learning_rate": 8.791898015394892e-06, "loss": 1.1048, "step": 11612 }, { "epoch": 0.9101097178683386, "grad_norm": 1.9517422914505005, "learning_rate": 8.791697013740163e-06, "loss": 1.5541, "step": 11613 }, { "epoch": 0.9101880877742947, "grad_norm": 2.9748125076293945, "learning_rate": 8.791495997663767e-06, "loss": 1.0274, "step": 11614 }, { "epoch": 0.9102664576802508, "grad_norm": 3.3484106063842773, "learning_rate": 8.79129496716647e-06, "loss": 1.2224, "step": 11615 }, { "epoch": 0.9103448275862069, "grad_norm": 3.5245652198791504, "learning_rate": 8.791093922249031e-06, "loss": 1.4281, "step": 11616 }, { "epoch": 0.910423197492163, "grad_norm": 2.8071277141571045, "learning_rate": 8.790892862912219e-06, "loss": 1.4103, "step": 11617 }, { "epoch": 0.9105015673981192, "grad_norm": 6.434015274047852, "learning_rate": 8.790691789156798e-06, "loss": 1.2059, "step": 11618 }, { "epoch": 0.9105799373040753, "grad_norm": 4.329352855682373, "learning_rate": 8.790490700983533e-06, "loss": 1.5077, "step": 11619 }, { "epoch": 0.9106583072100314, "grad_norm": 3.024867534637451, "learning_rate": 8.790289598393186e-06, "loss": 1.1559, "step": 11620 }, { "epoch": 0.9107366771159875, "grad_norm": 4.570672512054443, "learning_rate": 8.790088481386526e-06, "loss": 1.7016, "step": 11621 }, { "epoch": 0.9108150470219436, "grad_norm": 3.473186492919922, "learning_rate": 8.789887349964317e-06, "loss": 1.3103, "step": 11622 }, { "epoch": 0.9108934169278997, "grad_norm": 3.047731876373291, "learning_rate": 8.789686204127322e-06, "loss": 1.0226, "step": 11623 }, { "epoch": 0.9109717868338558, "grad_norm": 2.489516496658325, "learning_rate": 8.789485043876307e-06, "loss": 1.3844, "step": 11624 }, { "epoch": 0.9110501567398119, "grad_norm": 5.586329460144043, "learning_rate": 8.789283869212038e-06, "loss": 1.0791, "step": 11625 }, { "epoch": 0.911128526645768, "grad_norm": 3.2598378658294678, "learning_rate": 8.789082680135277e-06, "loss": 0.8419, "step": 11626 }, { "epoch": 0.9112068965517242, "grad_norm": 6.795238971710205, "learning_rate": 8.788881476646794e-06, "loss": 1.9255, "step": 11627 }, { "epoch": 0.9112852664576803, "grad_norm": 2.4258556365966797, "learning_rate": 8.788680258747352e-06, "loss": 1.0979, "step": 11628 }, { "epoch": 0.9113636363636364, "grad_norm": 6.695666790008545, "learning_rate": 8.788479026437716e-06, "loss": 1.8918, "step": 11629 }, { "epoch": 0.9114420062695925, "grad_norm": 2.357381582260132, "learning_rate": 8.788277779718652e-06, "loss": 1.1859, "step": 11630 }, { "epoch": 0.9115203761755486, "grad_norm": 2.4671471118927, "learning_rate": 8.788076518590925e-06, "loss": 1.1797, "step": 11631 }, { "epoch": 0.9115987460815047, "grad_norm": 5.4965434074401855, "learning_rate": 8.787875243055299e-06, "loss": 1.6503, "step": 11632 }, { "epoch": 0.9116771159874608, "grad_norm": 2.761075973510742, "learning_rate": 8.787673953112543e-06, "loss": 0.8319, "step": 11633 }, { "epoch": 0.9117554858934169, "grad_norm": 5.1522040367126465, "learning_rate": 8.78747264876342e-06, "loss": 1.5527, "step": 11634 }, { "epoch": 0.911833855799373, "grad_norm": 3.113013505935669, "learning_rate": 8.787271330008697e-06, "loss": 1.1449, "step": 11635 }, { "epoch": 0.9119122257053291, "grad_norm": 3.1648104190826416, "learning_rate": 8.787069996849138e-06, "loss": 1.317, "step": 11636 }, { "epoch": 0.9119905956112853, "grad_norm": 3.1675376892089844, "learning_rate": 8.786868649285511e-06, "loss": 1.4669, "step": 11637 }, { "epoch": 0.9120689655172414, "grad_norm": 3.4138243198394775, "learning_rate": 8.78666728731858e-06, "loss": 1.4726, "step": 11638 }, { "epoch": 0.9121473354231975, "grad_norm": 3.2435665130615234, "learning_rate": 8.78646591094911e-06, "loss": 1.4634, "step": 11639 }, { "epoch": 0.9122257053291536, "grad_norm": 3.365551710128784, "learning_rate": 8.786264520177872e-06, "loss": 1.1736, "step": 11640 }, { "epoch": 0.9123040752351097, "grad_norm": 3.33845853805542, "learning_rate": 8.786063115005628e-06, "loss": 1.5077, "step": 11641 }, { "epoch": 0.9123824451410658, "grad_norm": 4.300153732299805, "learning_rate": 8.785861695433142e-06, "loss": 1.321, "step": 11642 }, { "epoch": 0.9124608150470219, "grad_norm": 2.524083375930786, "learning_rate": 8.785660261461183e-06, "loss": 1.7108, "step": 11643 }, { "epoch": 0.912539184952978, "grad_norm": 2.202592611312866, "learning_rate": 8.785458813090519e-06, "loss": 1.3718, "step": 11644 }, { "epoch": 0.9126175548589341, "grad_norm": 4.378688335418701, "learning_rate": 8.785257350321913e-06, "loss": 0.887, "step": 11645 }, { "epoch": 0.9126959247648903, "grad_norm": 3.463557004928589, "learning_rate": 8.785055873156131e-06, "loss": 1.3902, "step": 11646 }, { "epoch": 0.9127742946708464, "grad_norm": 2.537684440612793, "learning_rate": 8.784854381593944e-06, "loss": 1.3565, "step": 11647 }, { "epoch": 0.9128526645768025, "grad_norm": 2.7930448055267334, "learning_rate": 8.784652875636113e-06, "loss": 1.4153, "step": 11648 }, { "epoch": 0.9129310344827586, "grad_norm": 4.34658670425415, "learning_rate": 8.784451355283406e-06, "loss": 1.701, "step": 11649 }, { "epoch": 0.9130094043887147, "grad_norm": 4.432450771331787, "learning_rate": 8.784249820536588e-06, "loss": 1.3341, "step": 11650 }, { "epoch": 0.9130877742946708, "grad_norm": 11.199958801269531, "learning_rate": 8.78404827139643e-06, "loss": 0.5854, "step": 11651 }, { "epoch": 0.9131661442006269, "grad_norm": 2.672928810119629, "learning_rate": 8.783846707863696e-06, "loss": 1.2472, "step": 11652 }, { "epoch": 0.913244514106583, "grad_norm": 2.477809429168701, "learning_rate": 8.783645129939152e-06, "loss": 1.3397, "step": 11653 }, { "epoch": 0.9133228840125391, "grad_norm": 2.554068088531494, "learning_rate": 8.783443537623566e-06, "loss": 0.9129, "step": 11654 }, { "epoch": 0.9134012539184952, "grad_norm": 2.9762659072875977, "learning_rate": 8.783241930917703e-06, "loss": 1.1156, "step": 11655 }, { "epoch": 0.9134796238244514, "grad_norm": 5.196794033050537, "learning_rate": 8.783040309822331e-06, "loss": 1.6989, "step": 11656 }, { "epoch": 0.9135579937304075, "grad_norm": 3.3421669006347656, "learning_rate": 8.782838674338217e-06, "loss": 1.3465, "step": 11657 }, { "epoch": 0.9136363636363637, "grad_norm": 2.6881866455078125, "learning_rate": 8.782637024466127e-06, "loss": 1.5913, "step": 11658 }, { "epoch": 0.9137147335423198, "grad_norm": 2.594813108444214, "learning_rate": 8.78243536020683e-06, "loss": 1.4517, "step": 11659 }, { "epoch": 0.9137931034482759, "grad_norm": 2.550215244293213, "learning_rate": 8.782233681561092e-06, "loss": 1.5713, "step": 11660 }, { "epoch": 0.913871473354232, "grad_norm": 4.2194132804870605, "learning_rate": 8.782031988529677e-06, "loss": 1.6038, "step": 11661 }, { "epoch": 0.9139498432601881, "grad_norm": 2.3360788822174072, "learning_rate": 8.781830281113357e-06, "loss": 1.0254, "step": 11662 }, { "epoch": 0.9140282131661442, "grad_norm": 4.527843475341797, "learning_rate": 8.781628559312896e-06, "loss": 1.5985, "step": 11663 }, { "epoch": 0.9141065830721004, "grad_norm": 2.6060192584991455, "learning_rate": 8.781426823129064e-06, "loss": 1.3248, "step": 11664 }, { "epoch": 0.9141849529780565, "grad_norm": 3.7077505588531494, "learning_rate": 8.781225072562627e-06, "loss": 1.5058, "step": 11665 }, { "epoch": 0.9142633228840126, "grad_norm": 5.947163105010986, "learning_rate": 8.781023307614349e-06, "loss": 1.4914, "step": 11666 }, { "epoch": 0.9143416927899687, "grad_norm": 2.6413676738739014, "learning_rate": 8.780821528285003e-06, "loss": 1.2197, "step": 11667 }, { "epoch": 0.9144200626959248, "grad_norm": 3.052232265472412, "learning_rate": 8.780619734575353e-06, "loss": 1.3486, "step": 11668 }, { "epoch": 0.9144984326018809, "grad_norm": 2.0072555541992188, "learning_rate": 8.780417926486167e-06, "loss": 1.4369, "step": 11669 }, { "epoch": 0.914576802507837, "grad_norm": 3.898625135421753, "learning_rate": 8.780216104018214e-06, "loss": 1.3173, "step": 11670 }, { "epoch": 0.9146551724137931, "grad_norm": 5.294735908508301, "learning_rate": 8.78001426717226e-06, "loss": 2.001, "step": 11671 }, { "epoch": 0.9147335423197492, "grad_norm": 2.3797876834869385, "learning_rate": 8.779812415949072e-06, "loss": 1.2838, "step": 11672 }, { "epoch": 0.9148119122257053, "grad_norm": 3.0697999000549316, "learning_rate": 8.77961055034942e-06, "loss": 1.3421, "step": 11673 }, { "epoch": 0.9148902821316615, "grad_norm": 4.088498592376709, "learning_rate": 8.779408670374071e-06, "loss": 1.2998, "step": 11674 }, { "epoch": 0.9149686520376176, "grad_norm": 2.967820644378662, "learning_rate": 8.779206776023794e-06, "loss": 1.4049, "step": 11675 }, { "epoch": 0.9150470219435737, "grad_norm": 3.3002243041992188, "learning_rate": 8.779004867299353e-06, "loss": 1.3191, "step": 11676 }, { "epoch": 0.9151253918495298, "grad_norm": 2.326992988586426, "learning_rate": 8.77880294420152e-06, "loss": 0.982, "step": 11677 }, { "epoch": 0.9152037617554859, "grad_norm": 1.873255968093872, "learning_rate": 8.778601006731061e-06, "loss": 1.088, "step": 11678 }, { "epoch": 0.915282131661442, "grad_norm": 2.12469482421875, "learning_rate": 8.778399054888744e-06, "loss": 1.4234, "step": 11679 }, { "epoch": 0.9153605015673981, "grad_norm": 3.6195549964904785, "learning_rate": 8.778197088675339e-06, "loss": 1.7261, "step": 11680 }, { "epoch": 0.9154388714733542, "grad_norm": 2.494935989379883, "learning_rate": 8.777995108091612e-06, "loss": 1.8564, "step": 11681 }, { "epoch": 0.9155172413793103, "grad_norm": 6.029295444488525, "learning_rate": 8.777793113138333e-06, "loss": 0.6369, "step": 11682 }, { "epoch": 0.9155956112852665, "grad_norm": 3.356611728668213, "learning_rate": 8.777591103816269e-06, "loss": 1.2266, "step": 11683 }, { "epoch": 0.9156739811912226, "grad_norm": 3.175180196762085, "learning_rate": 8.777389080126188e-06, "loss": 1.3604, "step": 11684 }, { "epoch": 0.9157523510971787, "grad_norm": 4.101245880126953, "learning_rate": 8.77718704206886e-06, "loss": 0.9133, "step": 11685 }, { "epoch": 0.9158307210031348, "grad_norm": 3.4986820220947266, "learning_rate": 8.776984989645053e-06, "loss": 1.4429, "step": 11686 }, { "epoch": 0.9159090909090909, "grad_norm": 9.141404151916504, "learning_rate": 8.776782922855533e-06, "loss": 1.6928, "step": 11687 }, { "epoch": 0.915987460815047, "grad_norm": 3.233701705932617, "learning_rate": 8.776580841701072e-06, "loss": 1.1338, "step": 11688 }, { "epoch": 0.9160658307210031, "grad_norm": 2.843812942504883, "learning_rate": 8.776378746182437e-06, "loss": 1.3242, "step": 11689 }, { "epoch": 0.9161442006269592, "grad_norm": 3.073720693588257, "learning_rate": 8.776176636300398e-06, "loss": 1.3836, "step": 11690 }, { "epoch": 0.9162225705329153, "grad_norm": 1.9949463605880737, "learning_rate": 8.775974512055722e-06, "loss": 0.8122, "step": 11691 }, { "epoch": 0.9163009404388714, "grad_norm": 2.5056731700897217, "learning_rate": 8.775772373449178e-06, "loss": 0.9455, "step": 11692 }, { "epoch": 0.9163793103448276, "grad_norm": 3.14054274559021, "learning_rate": 8.775570220481535e-06, "loss": 1.2589, "step": 11693 }, { "epoch": 0.9164576802507837, "grad_norm": 2.6505343914031982, "learning_rate": 8.775368053153564e-06, "loss": 1.3797, "step": 11694 }, { "epoch": 0.9165360501567398, "grad_norm": 2.8795149326324463, "learning_rate": 8.775165871466029e-06, "loss": 1.1712, "step": 11695 }, { "epoch": 0.9166144200626959, "grad_norm": 3.9959254264831543, "learning_rate": 8.774963675419703e-06, "loss": 1.3535, "step": 11696 }, { "epoch": 0.916692789968652, "grad_norm": 4.433960914611816, "learning_rate": 8.774761465015356e-06, "loss": 1.2994, "step": 11697 }, { "epoch": 0.9167711598746081, "grad_norm": 2.528247594833374, "learning_rate": 8.774559240253755e-06, "loss": 1.3633, "step": 11698 }, { "epoch": 0.9168495297805642, "grad_norm": 2.4420437812805176, "learning_rate": 8.774357001135669e-06, "loss": 1.1869, "step": 11699 }, { "epoch": 0.9169278996865203, "grad_norm": 6.9112548828125, "learning_rate": 8.774154747661867e-06, "loss": 0.9751, "step": 11700 }, { "epoch": 0.9170062695924764, "grad_norm": 2.8302512168884277, "learning_rate": 8.773952479833119e-06, "loss": 1.4475, "step": 11701 }, { "epoch": 0.9170846394984326, "grad_norm": 3.898449182510376, "learning_rate": 8.773750197650192e-06, "loss": 1.3381, "step": 11702 }, { "epoch": 0.9171630094043887, "grad_norm": 2.335024118423462, "learning_rate": 8.773547901113862e-06, "loss": 1.3542, "step": 11703 }, { "epoch": 0.9172413793103448, "grad_norm": 4.002233505249023, "learning_rate": 8.773345590224891e-06, "loss": 1.078, "step": 11704 }, { "epoch": 0.917319749216301, "grad_norm": 2.2939491271972656, "learning_rate": 8.773143264984053e-06, "loss": 1.0832, "step": 11705 }, { "epoch": 0.9173981191222571, "grad_norm": 2.4035520553588867, "learning_rate": 8.772940925392114e-06, "loss": 1.3843, "step": 11706 }, { "epoch": 0.9174764890282132, "grad_norm": 2.1548712253570557, "learning_rate": 8.772738571449847e-06, "loss": 1.2284, "step": 11707 }, { "epoch": 0.9175548589341693, "grad_norm": 2.4187803268432617, "learning_rate": 8.77253620315802e-06, "loss": 1.2367, "step": 11708 }, { "epoch": 0.9176332288401254, "grad_norm": 2.3488376140594482, "learning_rate": 8.772333820517402e-06, "loss": 1.4942, "step": 11709 }, { "epoch": 0.9177115987460815, "grad_norm": 3.5493640899658203, "learning_rate": 8.772131423528766e-06, "loss": 1.4907, "step": 11710 }, { "epoch": 0.9177899686520377, "grad_norm": 2.2085094451904297, "learning_rate": 8.77192901219288e-06, "loss": 1.2045, "step": 11711 }, { "epoch": 0.9178683385579938, "grad_norm": 1.847357988357544, "learning_rate": 8.77172658651051e-06, "loss": 1.3815, "step": 11712 }, { "epoch": 0.9179467084639499, "grad_norm": 3.0026566982269287, "learning_rate": 8.77152414648243e-06, "loss": 1.5407, "step": 11713 }, { "epoch": 0.918025078369906, "grad_norm": 2.5948216915130615, "learning_rate": 8.771321692109411e-06, "loss": 1.5849, "step": 11714 }, { "epoch": 0.9181034482758621, "grad_norm": 3.404053211212158, "learning_rate": 8.771119223392221e-06, "loss": 1.0934, "step": 11715 }, { "epoch": 0.9181818181818182, "grad_norm": 3.043937921524048, "learning_rate": 8.77091674033163e-06, "loss": 1.0778, "step": 11716 }, { "epoch": 0.9182601880877743, "grad_norm": 2.5674703121185303, "learning_rate": 8.77071424292841e-06, "loss": 1.5809, "step": 11717 }, { "epoch": 0.9183385579937304, "grad_norm": 1.91497004032135, "learning_rate": 8.770511731183328e-06, "loss": 1.2835, "step": 11718 }, { "epoch": 0.9184169278996865, "grad_norm": 10.332919120788574, "learning_rate": 8.770309205097157e-06, "loss": 2.0522, "step": 11719 }, { "epoch": 0.9184952978056427, "grad_norm": 2.8805429935455322, "learning_rate": 8.770106664670666e-06, "loss": 1.9892, "step": 11720 }, { "epoch": 0.9185736677115988, "grad_norm": 5.305834770202637, "learning_rate": 8.769904109904625e-06, "loss": 1.8276, "step": 11721 }, { "epoch": 0.9186520376175549, "grad_norm": 3.6586897373199463, "learning_rate": 8.769701540799806e-06, "loss": 0.9105, "step": 11722 }, { "epoch": 0.918730407523511, "grad_norm": 1.8125174045562744, "learning_rate": 8.769498957356979e-06, "loss": 1.1095, "step": 11723 }, { "epoch": 0.9188087774294671, "grad_norm": 5.0562238693237305, "learning_rate": 8.769296359576913e-06, "loss": 1.5798, "step": 11724 }, { "epoch": 0.9188871473354232, "grad_norm": 10.75778579711914, "learning_rate": 8.769093747460382e-06, "loss": 1.8642, "step": 11725 }, { "epoch": 0.9189655172413793, "grad_norm": 2.706490993499756, "learning_rate": 8.768891121008152e-06, "loss": 1.2957, "step": 11726 }, { "epoch": 0.9190438871473354, "grad_norm": 4.857056140899658, "learning_rate": 8.768688480220998e-06, "loss": 1.1923, "step": 11727 }, { "epoch": 0.9191222570532915, "grad_norm": 4.862668514251709, "learning_rate": 8.768485825099684e-06, "loss": 1.4139, "step": 11728 }, { "epoch": 0.9192006269592476, "grad_norm": 4.234622955322266, "learning_rate": 8.76828315564499e-06, "loss": 1.5651, "step": 11729 }, { "epoch": 0.9192789968652038, "grad_norm": 4.960136413574219, "learning_rate": 8.768080471857681e-06, "loss": 0.9487, "step": 11730 }, { "epoch": 0.9193573667711599, "grad_norm": 3.2593929767608643, "learning_rate": 8.767877773738529e-06, "loss": 1.5439, "step": 11731 }, { "epoch": 0.919435736677116, "grad_norm": 5.437571048736572, "learning_rate": 8.767675061288306e-06, "loss": 1.2312, "step": 11732 }, { "epoch": 0.9195141065830721, "grad_norm": 4.091878890991211, "learning_rate": 8.767472334507782e-06, "loss": 1.5403, "step": 11733 }, { "epoch": 0.9195924764890282, "grad_norm": 3.1287364959716797, "learning_rate": 8.767269593397729e-06, "loss": 1.1693, "step": 11734 }, { "epoch": 0.9196708463949843, "grad_norm": 6.521388053894043, "learning_rate": 8.767066837958917e-06, "loss": 1.3134, "step": 11735 }, { "epoch": 0.9197492163009404, "grad_norm": 16.566333770751953, "learning_rate": 8.766864068192118e-06, "loss": 2.6558, "step": 11736 }, { "epoch": 0.9198275862068965, "grad_norm": 2.818169116973877, "learning_rate": 8.766661284098102e-06, "loss": 1.6057, "step": 11737 }, { "epoch": 0.9199059561128526, "grad_norm": 6.843043327331543, "learning_rate": 8.76645848567764e-06, "loss": 1.3526, "step": 11738 }, { "epoch": 0.9199843260188088, "grad_norm": 2.9467110633850098, "learning_rate": 8.766255672931508e-06, "loss": 1.386, "step": 11739 }, { "epoch": 0.9200626959247649, "grad_norm": 7.2705583572387695, "learning_rate": 8.766052845860472e-06, "loss": 1.2236, "step": 11740 }, { "epoch": 0.920141065830721, "grad_norm": 2.840972900390625, "learning_rate": 8.765850004465304e-06, "loss": 1.5653, "step": 11741 }, { "epoch": 0.9202194357366771, "grad_norm": 3.3947744369506836, "learning_rate": 8.765647148746778e-06, "loss": 1.7829, "step": 11742 }, { "epoch": 0.9202978056426332, "grad_norm": 2.602320671081543, "learning_rate": 8.765444278705664e-06, "loss": 1.5012, "step": 11743 }, { "epoch": 0.9203761755485893, "grad_norm": 2.5513744354248047, "learning_rate": 8.765241394342734e-06, "loss": 1.3129, "step": 11744 }, { "epoch": 0.9204545454545454, "grad_norm": 2.6893813610076904, "learning_rate": 8.765038495658762e-06, "loss": 1.2022, "step": 11745 }, { "epoch": 0.9205329153605015, "grad_norm": 2.2076499462127686, "learning_rate": 8.764835582654513e-06, "loss": 1.2738, "step": 11746 }, { "epoch": 0.9206112852664576, "grad_norm": 2.545470714569092, "learning_rate": 8.764632655330767e-06, "loss": 1.5948, "step": 11747 }, { "epoch": 0.9206896551724137, "grad_norm": 2.138037919998169, "learning_rate": 8.764429713688292e-06, "loss": 1.1122, "step": 11748 }, { "epoch": 0.9207680250783699, "grad_norm": 1.7810935974121094, "learning_rate": 8.764226757727857e-06, "loss": 1.1911, "step": 11749 }, { "epoch": 0.920846394984326, "grad_norm": 2.3369929790496826, "learning_rate": 8.764023787450238e-06, "loss": 1.2417, "step": 11750 }, { "epoch": 0.9209247648902821, "grad_norm": 8.264042854309082, "learning_rate": 8.763820802856207e-06, "loss": 1.0036, "step": 11751 }, { "epoch": 0.9210031347962383, "grad_norm": 4.622709274291992, "learning_rate": 8.763617803946534e-06, "loss": 1.3758, "step": 11752 }, { "epoch": 0.9210815047021944, "grad_norm": 2.924044370651245, "learning_rate": 8.763414790721993e-06, "loss": 1.3109, "step": 11753 }, { "epoch": 0.9211598746081505, "grad_norm": 2.3699758052825928, "learning_rate": 8.763211763183354e-06, "loss": 1.7403, "step": 11754 }, { "epoch": 0.9212382445141066, "grad_norm": 2.5812082290649414, "learning_rate": 8.76300872133139e-06, "loss": 1.2778, "step": 11755 }, { "epoch": 0.9213166144200627, "grad_norm": 2.4943296909332275, "learning_rate": 8.762805665166876e-06, "loss": 1.2469, "step": 11756 }, { "epoch": 0.9213949843260189, "grad_norm": 3.6494243144989014, "learning_rate": 8.762602594690579e-06, "loss": 1.5763, "step": 11757 }, { "epoch": 0.921473354231975, "grad_norm": 3.1928324699401855, "learning_rate": 8.762399509903277e-06, "loss": 1.3348, "step": 11758 }, { "epoch": 0.9215517241379311, "grad_norm": 2.615818738937378, "learning_rate": 8.762196410805738e-06, "loss": 1.3642, "step": 11759 }, { "epoch": 0.9216300940438872, "grad_norm": 2.829925060272217, "learning_rate": 8.761993297398737e-06, "loss": 1.4413, "step": 11760 }, { "epoch": 0.9217084639498433, "grad_norm": 3.041187047958374, "learning_rate": 8.761790169683045e-06, "loss": 1.948, "step": 11761 }, { "epoch": 0.9217868338557994, "grad_norm": 2.9712626934051514, "learning_rate": 8.761587027659434e-06, "loss": 1.2075, "step": 11762 }, { "epoch": 0.9218652037617555, "grad_norm": 9.018040657043457, "learning_rate": 8.761383871328682e-06, "loss": 1.1178, "step": 11763 }, { "epoch": 0.9219435736677116, "grad_norm": 3.6891112327575684, "learning_rate": 8.761180700691553e-06, "loss": 1.5641, "step": 11764 }, { "epoch": 0.9220219435736677, "grad_norm": 2.3920528888702393, "learning_rate": 8.760977515748829e-06, "loss": 1.3539, "step": 11765 }, { "epoch": 0.9221003134796238, "grad_norm": 2.7731688022613525, "learning_rate": 8.760774316501275e-06, "loss": 1.2851, "step": 11766 }, { "epoch": 0.92217868338558, "grad_norm": 3.2011070251464844, "learning_rate": 8.760571102949668e-06, "loss": 1.2978, "step": 11767 }, { "epoch": 0.9222570532915361, "grad_norm": 2.742685317993164, "learning_rate": 8.760367875094783e-06, "loss": 1.3984, "step": 11768 }, { "epoch": 0.9223354231974922, "grad_norm": 2.8943400382995605, "learning_rate": 8.760164632937387e-06, "loss": 1.2843, "step": 11769 }, { "epoch": 0.9224137931034483, "grad_norm": 3.4657633304595947, "learning_rate": 8.759961376478256e-06, "loss": 1.6716, "step": 11770 }, { "epoch": 0.9224921630094044, "grad_norm": 2.5278265476226807, "learning_rate": 8.759758105718164e-06, "loss": 1.3753, "step": 11771 }, { "epoch": 0.9225705329153605, "grad_norm": 3.2655904293060303, "learning_rate": 8.759554820657884e-06, "loss": 1.1367, "step": 11772 }, { "epoch": 0.9226489028213166, "grad_norm": 3.226287841796875, "learning_rate": 8.759351521298188e-06, "loss": 1.1724, "step": 11773 }, { "epoch": 0.9227272727272727, "grad_norm": 2.9728312492370605, "learning_rate": 8.759148207639848e-06, "loss": 1.2699, "step": 11774 }, { "epoch": 0.9228056426332288, "grad_norm": 2.5254337787628174, "learning_rate": 8.75894487968364e-06, "loss": 1.5053, "step": 11775 }, { "epoch": 0.922884012539185, "grad_norm": 3.874541997909546, "learning_rate": 8.758741537430337e-06, "loss": 1.2122, "step": 11776 }, { "epoch": 0.9229623824451411, "grad_norm": 2.997185707092285, "learning_rate": 8.758538180880711e-06, "loss": 0.9894, "step": 11777 }, { "epoch": 0.9230407523510972, "grad_norm": 3.7392594814300537, "learning_rate": 8.758334810035538e-06, "loss": 1.4279, "step": 11778 }, { "epoch": 0.9231191222570533, "grad_norm": 5.334169387817383, "learning_rate": 8.758131424895589e-06, "loss": 1.3522, "step": 11779 }, { "epoch": 0.9231974921630094, "grad_norm": 3.3445065021514893, "learning_rate": 8.757928025461637e-06, "loss": 1.369, "step": 11780 }, { "epoch": 0.9232758620689655, "grad_norm": 2.5580379962921143, "learning_rate": 8.757724611734457e-06, "loss": 1.4159, "step": 11781 }, { "epoch": 0.9233542319749216, "grad_norm": 2.5577993392944336, "learning_rate": 8.757521183714824e-06, "loss": 0.8927, "step": 11782 }, { "epoch": 0.9234326018808777, "grad_norm": 2.806781768798828, "learning_rate": 8.75731774140351e-06, "loss": 1.3789, "step": 11783 }, { "epoch": 0.9235109717868338, "grad_norm": 4.11073112487793, "learning_rate": 8.757114284801289e-06, "loss": 1.4407, "step": 11784 }, { "epoch": 0.92358934169279, "grad_norm": 2.374652624130249, "learning_rate": 8.756910813908936e-06, "loss": 1.2998, "step": 11785 }, { "epoch": 0.923667711598746, "grad_norm": 3.447129487991333, "learning_rate": 8.756707328727223e-06, "loss": 1.3087, "step": 11786 }, { "epoch": 0.9237460815047022, "grad_norm": 2.9482007026672363, "learning_rate": 8.756503829256926e-06, "loss": 1.3162, "step": 11787 }, { "epoch": 0.9238244514106583, "grad_norm": 5.802467346191406, "learning_rate": 8.756300315498815e-06, "loss": 1.5322, "step": 11788 }, { "epoch": 0.9239028213166144, "grad_norm": 3.121189594268799, "learning_rate": 8.75609678745367e-06, "loss": 1.3143, "step": 11789 }, { "epoch": 0.9239811912225705, "grad_norm": 5.491695404052734, "learning_rate": 8.755893245122261e-06, "loss": 1.3636, "step": 11790 }, { "epoch": 0.9240595611285266, "grad_norm": 3.297262191772461, "learning_rate": 8.755689688505362e-06, "loss": 1.1747, "step": 11791 }, { "epoch": 0.9241379310344827, "grad_norm": 2.399789810180664, "learning_rate": 8.755486117603751e-06, "loss": 1.1249, "step": 11792 }, { "epoch": 0.9242163009404388, "grad_norm": 3.251030921936035, "learning_rate": 8.755282532418198e-06, "loss": 1.2753, "step": 11793 }, { "epoch": 0.9242946708463949, "grad_norm": 2.226651906967163, "learning_rate": 8.75507893294948e-06, "loss": 1.4885, "step": 11794 }, { "epoch": 0.924373040752351, "grad_norm": 2.996950626373291, "learning_rate": 8.75487531919837e-06, "loss": 1.5152, "step": 11795 }, { "epoch": 0.9244514106583072, "grad_norm": 3.210001230239868, "learning_rate": 8.754671691165644e-06, "loss": 1.2195, "step": 11796 }, { "epoch": 0.9245297805642633, "grad_norm": 2.81720232963562, "learning_rate": 8.754468048852074e-06, "loss": 1.2317, "step": 11797 }, { "epoch": 0.9246081504702194, "grad_norm": 2.7840890884399414, "learning_rate": 8.754264392258435e-06, "loss": 1.5078, "step": 11798 }, { "epoch": 0.9246865203761756, "grad_norm": 3.0080454349517822, "learning_rate": 8.754060721385505e-06, "loss": 1.6832, "step": 11799 }, { "epoch": 0.9247648902821317, "grad_norm": 2.998795509338379, "learning_rate": 8.753857036234055e-06, "loss": 1.5185, "step": 11800 }, { "epoch": 0.9248432601880878, "grad_norm": 4.164860725402832, "learning_rate": 8.75365333680486e-06, "loss": 1.3576, "step": 11801 }, { "epoch": 0.9249216300940439, "grad_norm": 3.478119373321533, "learning_rate": 8.753449623098697e-06, "loss": 1.0408, "step": 11802 }, { "epoch": 0.925, "grad_norm": 3.0072619915008545, "learning_rate": 8.75324589511634e-06, "loss": 1.5866, "step": 11803 }, { "epoch": 0.9250783699059562, "grad_norm": 17.47745132446289, "learning_rate": 8.753042152858561e-06, "loss": 1.6782, "step": 11804 }, { "epoch": 0.9251567398119123, "grad_norm": 2.2998013496398926, "learning_rate": 8.75283839632614e-06, "loss": 1.1764, "step": 11805 }, { "epoch": 0.9252351097178684, "grad_norm": 2.644803524017334, "learning_rate": 8.75263462551985e-06, "loss": 1.2411, "step": 11806 }, { "epoch": 0.9253134796238245, "grad_norm": 4.120710372924805, "learning_rate": 8.752430840440463e-06, "loss": 1.7017, "step": 11807 }, { "epoch": 0.9253918495297806, "grad_norm": 2.9939663410186768, "learning_rate": 8.752227041088756e-06, "loss": 1.3113, "step": 11808 }, { "epoch": 0.9254702194357367, "grad_norm": 2.3323662281036377, "learning_rate": 8.752023227465505e-06, "loss": 1.6129, "step": 11809 }, { "epoch": 0.9255485893416928, "grad_norm": 3.064777135848999, "learning_rate": 8.751819399571486e-06, "loss": 1.1407, "step": 11810 }, { "epoch": 0.9256269592476489, "grad_norm": 3.036231279373169, "learning_rate": 8.751615557407473e-06, "loss": 1.5031, "step": 11811 }, { "epoch": 0.925705329153605, "grad_norm": 3.929548501968384, "learning_rate": 8.751411700974242e-06, "loss": 1.3357, "step": 11812 }, { "epoch": 0.9257836990595611, "grad_norm": 2.8857791423797607, "learning_rate": 8.751207830272565e-06, "loss": 1.087, "step": 11813 }, { "epoch": 0.9258620689655173, "grad_norm": 4.5695343017578125, "learning_rate": 8.751003945303223e-06, "loss": 1.2214, "step": 11814 }, { "epoch": 0.9259404388714734, "grad_norm": 3.029998302459717, "learning_rate": 8.750800046066987e-06, "loss": 1.2616, "step": 11815 }, { "epoch": 0.9260188087774295, "grad_norm": 3.204399824142456, "learning_rate": 8.750596132564635e-06, "loss": 0.8959, "step": 11816 }, { "epoch": 0.9260971786833856, "grad_norm": 2.2875492572784424, "learning_rate": 8.75039220479694e-06, "loss": 1.4306, "step": 11817 }, { "epoch": 0.9261755485893417, "grad_norm": 1.9231288433074951, "learning_rate": 8.750188262764682e-06, "loss": 1.2881, "step": 11818 }, { "epoch": 0.9262539184952978, "grad_norm": 3.020806312561035, "learning_rate": 8.749984306468633e-06, "loss": 1.4037, "step": 11819 }, { "epoch": 0.9263322884012539, "grad_norm": 4.5391387939453125, "learning_rate": 8.74978033590957e-06, "loss": 1.506, "step": 11820 }, { "epoch": 0.92641065830721, "grad_norm": 2.6172120571136475, "learning_rate": 8.749576351088266e-06, "loss": 1.5177, "step": 11821 }, { "epoch": 0.9264890282131661, "grad_norm": 3.085970163345337, "learning_rate": 8.749372352005503e-06, "loss": 0.9417, "step": 11822 }, { "epoch": 0.9265673981191223, "grad_norm": 2.462869644165039, "learning_rate": 8.749168338662053e-06, "loss": 1.4079, "step": 11823 }, { "epoch": 0.9266457680250784, "grad_norm": 5.109785079956055, "learning_rate": 8.748964311058691e-06, "loss": 1.6671, "step": 11824 }, { "epoch": 0.9267241379310345, "grad_norm": 2.2867751121520996, "learning_rate": 8.748760269196195e-06, "loss": 0.7743, "step": 11825 }, { "epoch": 0.9268025078369906, "grad_norm": 2.939763307571411, "learning_rate": 8.74855621307534e-06, "loss": 1.4556, "step": 11826 }, { "epoch": 0.9268808777429467, "grad_norm": 7.830463886260986, "learning_rate": 8.748352142696902e-06, "loss": 1.3637, "step": 11827 }, { "epoch": 0.9269592476489028, "grad_norm": 4.276706695556641, "learning_rate": 8.748148058061659e-06, "loss": 1.2736, "step": 11828 }, { "epoch": 0.9270376175548589, "grad_norm": 2.5782952308654785, "learning_rate": 8.747943959170385e-06, "loss": 1.4431, "step": 11829 }, { "epoch": 0.927115987460815, "grad_norm": 4.030335426330566, "learning_rate": 8.747739846023858e-06, "loss": 1.6516, "step": 11830 }, { "epoch": 0.9271943573667711, "grad_norm": 4.702578067779541, "learning_rate": 8.747535718622852e-06, "loss": 1.2826, "step": 11831 }, { "epoch": 0.9272727272727272, "grad_norm": 5.625157356262207, "learning_rate": 8.747331576968147e-06, "loss": 1.3562, "step": 11832 }, { "epoch": 0.9273510971786834, "grad_norm": 2.528278350830078, "learning_rate": 8.747127421060517e-06, "loss": 1.4322, "step": 11833 }, { "epoch": 0.9274294670846395, "grad_norm": 1.765492558479309, "learning_rate": 8.746923250900738e-06, "loss": 1.209, "step": 11834 }, { "epoch": 0.9275078369905956, "grad_norm": 2.0040769577026367, "learning_rate": 8.746719066489588e-06, "loss": 1.4872, "step": 11835 }, { "epoch": 0.9275862068965517, "grad_norm": 2.6029324531555176, "learning_rate": 8.746514867827845e-06, "loss": 1.4248, "step": 11836 }, { "epoch": 0.9276645768025078, "grad_norm": 2.3138551712036133, "learning_rate": 8.746310654916282e-06, "loss": 1.3214, "step": 11837 }, { "epoch": 0.9277429467084639, "grad_norm": 3.9465084075927734, "learning_rate": 8.746106427755676e-06, "loss": 1.5585, "step": 11838 }, { "epoch": 0.92782131661442, "grad_norm": 2.6979570388793945, "learning_rate": 8.745902186346807e-06, "loss": 1.5348, "step": 11839 }, { "epoch": 0.9278996865203761, "grad_norm": 2.554236650466919, "learning_rate": 8.74569793069045e-06, "loss": 1.2883, "step": 11840 }, { "epoch": 0.9279780564263322, "grad_norm": 6.883528709411621, "learning_rate": 8.745493660787382e-06, "loss": 1.3782, "step": 11841 }, { "epoch": 0.9280564263322884, "grad_norm": 4.845592021942139, "learning_rate": 8.74528937663838e-06, "loss": 1.5445, "step": 11842 }, { "epoch": 0.9281347962382445, "grad_norm": 3.5289783477783203, "learning_rate": 8.74508507824422e-06, "loss": 1.2646, "step": 11843 }, { "epoch": 0.9282131661442006, "grad_norm": 2.576807975769043, "learning_rate": 8.744880765605681e-06, "loss": 1.051, "step": 11844 }, { "epoch": 0.9282915360501567, "grad_norm": 2.5871291160583496, "learning_rate": 8.744676438723538e-06, "loss": 1.1036, "step": 11845 }, { "epoch": 0.9283699059561129, "grad_norm": 3.4239344596862793, "learning_rate": 8.744472097598571e-06, "loss": 1.3449, "step": 11846 }, { "epoch": 0.928448275862069, "grad_norm": 4.418472766876221, "learning_rate": 8.744267742231553e-06, "loss": 2.2032, "step": 11847 }, { "epoch": 0.9285266457680251, "grad_norm": 2.853816032409668, "learning_rate": 8.744063372623267e-06, "loss": 1.1682, "step": 11848 }, { "epoch": 0.9286050156739812, "grad_norm": 2.729231119155884, "learning_rate": 8.743858988774483e-06, "loss": 1.4694, "step": 11849 }, { "epoch": 0.9286833855799373, "grad_norm": 2.169297695159912, "learning_rate": 8.743654590685985e-06, "loss": 1.6087, "step": 11850 }, { "epoch": 0.9287617554858935, "grad_norm": 8.36086368560791, "learning_rate": 8.743450178358549e-06, "loss": 1.8642, "step": 11851 }, { "epoch": 0.9288401253918496, "grad_norm": 2.271606683731079, "learning_rate": 8.743245751792948e-06, "loss": 1.7403, "step": 11852 }, { "epoch": 0.9289184952978057, "grad_norm": 3.7527670860290527, "learning_rate": 8.743041310989964e-06, "loss": 1.2553, "step": 11853 }, { "epoch": 0.9289968652037618, "grad_norm": 2.7390902042388916, "learning_rate": 8.742836855950373e-06, "loss": 1.421, "step": 11854 }, { "epoch": 0.9290752351097179, "grad_norm": 2.8420209884643555, "learning_rate": 8.742632386674954e-06, "loss": 1.2527, "step": 11855 }, { "epoch": 0.929153605015674, "grad_norm": 4.380852699279785, "learning_rate": 8.742427903164482e-06, "loss": 1.1799, "step": 11856 }, { "epoch": 0.9292319749216301, "grad_norm": 3.529024839401245, "learning_rate": 8.74222340541974e-06, "loss": 1.0282, "step": 11857 }, { "epoch": 0.9293103448275862, "grad_norm": 12.089776992797852, "learning_rate": 8.742018893441499e-06, "loss": 1.5515, "step": 11858 }, { "epoch": 0.9293887147335423, "grad_norm": 4.726650238037109, "learning_rate": 8.741814367230543e-06, "loss": 0.9598, "step": 11859 }, { "epoch": 0.9294670846394985, "grad_norm": 3.5132458209991455, "learning_rate": 8.741609826787644e-06, "loss": 1.5541, "step": 11860 }, { "epoch": 0.9295454545454546, "grad_norm": 2.4930896759033203, "learning_rate": 8.741405272113586e-06, "loss": 1.2419, "step": 11861 }, { "epoch": 0.9296238244514107, "grad_norm": 9.289270401000977, "learning_rate": 8.741200703209142e-06, "loss": 0.6072, "step": 11862 }, { "epoch": 0.9297021943573668, "grad_norm": 2.5138943195343018, "learning_rate": 8.740996120075093e-06, "loss": 1.2434, "step": 11863 }, { "epoch": 0.9297805642633229, "grad_norm": 2.996140480041504, "learning_rate": 8.740791522712216e-06, "loss": 1.444, "step": 11864 }, { "epoch": 0.929858934169279, "grad_norm": 2.1757588386535645, "learning_rate": 8.740586911121289e-06, "loss": 1.6906, "step": 11865 }, { "epoch": 0.9299373040752351, "grad_norm": 2.8996622562408447, "learning_rate": 8.740382285303091e-06, "loss": 1.2235, "step": 11866 }, { "epoch": 0.9300156739811912, "grad_norm": 3.310856819152832, "learning_rate": 8.7401776452584e-06, "loss": 1.2511, "step": 11867 }, { "epoch": 0.9300940438871473, "grad_norm": 5.67152214050293, "learning_rate": 8.739972990987995e-06, "loss": 1.608, "step": 11868 }, { "epoch": 0.9301724137931034, "grad_norm": 2.194531202316284, "learning_rate": 8.739768322492653e-06, "loss": 1.4365, "step": 11869 }, { "epoch": 0.9302507836990596, "grad_norm": 7.276558876037598, "learning_rate": 8.739563639773152e-06, "loss": 1.1739, "step": 11870 }, { "epoch": 0.9303291536050157, "grad_norm": 2.7773938179016113, "learning_rate": 8.739358942830274e-06, "loss": 1.1944, "step": 11871 }, { "epoch": 0.9304075235109718, "grad_norm": 3.213402032852173, "learning_rate": 8.739154231664792e-06, "loss": 1.4036, "step": 11872 }, { "epoch": 0.9304858934169279, "grad_norm": 2.645414113998413, "learning_rate": 8.73894950627749e-06, "loss": 1.6825, "step": 11873 }, { "epoch": 0.930564263322884, "grad_norm": 8.654654502868652, "learning_rate": 8.738744766669147e-06, "loss": 1.2065, "step": 11874 }, { "epoch": 0.9306426332288401, "grad_norm": 3.594066619873047, "learning_rate": 8.738540012840536e-06, "loss": 1.4258, "step": 11875 }, { "epoch": 0.9307210031347962, "grad_norm": 3.0694868564605713, "learning_rate": 8.73833524479244e-06, "loss": 1.4704, "step": 11876 }, { "epoch": 0.9307993730407523, "grad_norm": 3.520951986312866, "learning_rate": 8.738130462525636e-06, "loss": 1.2375, "step": 11877 }, { "epoch": 0.9308777429467084, "grad_norm": 2.509748697280884, "learning_rate": 8.737925666040904e-06, "loss": 1.3336, "step": 11878 }, { "epoch": 0.9309561128526646, "grad_norm": 5.842484951019287, "learning_rate": 8.737720855339023e-06, "loss": 1.9627, "step": 11879 }, { "epoch": 0.9310344827586207, "grad_norm": 2.2770907878875732, "learning_rate": 8.737516030420773e-06, "loss": 1.4002, "step": 11880 }, { "epoch": 0.9311128526645768, "grad_norm": 9.922164916992188, "learning_rate": 8.73731119128693e-06, "loss": 0.802, "step": 11881 }, { "epoch": 0.9311912225705329, "grad_norm": 5.261145114898682, "learning_rate": 8.737106337938275e-06, "loss": 0.4172, "step": 11882 }, { "epoch": 0.931269592476489, "grad_norm": 13.182177543640137, "learning_rate": 8.736901470375587e-06, "loss": 0.5092, "step": 11883 }, { "epoch": 0.9313479623824451, "grad_norm": 3.4683187007904053, "learning_rate": 8.736696588599646e-06, "loss": 1.3855, "step": 11884 }, { "epoch": 0.9314263322884012, "grad_norm": 2.7716128826141357, "learning_rate": 8.73649169261123e-06, "loss": 1.4119, "step": 11885 }, { "epoch": 0.9315047021943573, "grad_norm": 3.5625736713409424, "learning_rate": 8.736286782411118e-06, "loss": 1.35, "step": 11886 }, { "epoch": 0.9315830721003134, "grad_norm": 9.158926963806152, "learning_rate": 8.736081858000092e-06, "loss": 1.353, "step": 11887 }, { "epoch": 0.9316614420062695, "grad_norm": 2.772589683532715, "learning_rate": 8.735876919378927e-06, "loss": 0.8479, "step": 11888 }, { "epoch": 0.9317398119122257, "grad_norm": 4.1847028732299805, "learning_rate": 8.735671966548407e-06, "loss": 1.0788, "step": 11889 }, { "epoch": 0.9318181818181818, "grad_norm": 3.4037773609161377, "learning_rate": 8.73546699950931e-06, "loss": 1.2163, "step": 11890 }, { "epoch": 0.9318965517241379, "grad_norm": 2.185466766357422, "learning_rate": 8.735262018262413e-06, "loss": 1.5176, "step": 11891 }, { "epoch": 0.931974921630094, "grad_norm": 3.1305058002471924, "learning_rate": 8.7350570228085e-06, "loss": 2.0561, "step": 11892 }, { "epoch": 0.9320532915360502, "grad_norm": 2.3046984672546387, "learning_rate": 8.734852013148347e-06, "loss": 1.6068, "step": 11893 }, { "epoch": 0.9321316614420063, "grad_norm": 3.4746479988098145, "learning_rate": 8.734646989282737e-06, "loss": 1.4765, "step": 11894 }, { "epoch": 0.9322100313479624, "grad_norm": 2.8051581382751465, "learning_rate": 8.734441951212446e-06, "loss": 1.5301, "step": 11895 }, { "epoch": 0.9322884012539185, "grad_norm": 2.0969326496124268, "learning_rate": 8.734236898938258e-06, "loss": 1.4615, "step": 11896 }, { "epoch": 0.9323667711598747, "grad_norm": 2.2668018341064453, "learning_rate": 8.73403183246095e-06, "loss": 1.2992, "step": 11897 }, { "epoch": 0.9324451410658308, "grad_norm": 2.6857643127441406, "learning_rate": 8.733826751781303e-06, "loss": 1.9153, "step": 11898 }, { "epoch": 0.9325235109717869, "grad_norm": 7.977649211883545, "learning_rate": 8.733621656900095e-06, "loss": 1.6329, "step": 11899 }, { "epoch": 0.932601880877743, "grad_norm": 4.160550117492676, "learning_rate": 8.73341654781811e-06, "loss": 1.5948, "step": 11900 }, { "epoch": 0.9326802507836991, "grad_norm": 16.379901885986328, "learning_rate": 8.733211424536125e-06, "loss": 0.8852, "step": 11901 }, { "epoch": 0.9327586206896552, "grad_norm": 2.1927731037139893, "learning_rate": 8.733006287054923e-06, "loss": 1.3801, "step": 11902 }, { "epoch": 0.9328369905956113, "grad_norm": 6.098167419433594, "learning_rate": 8.732801135375282e-06, "loss": 1.1864, "step": 11903 }, { "epoch": 0.9329153605015674, "grad_norm": 2.778985023498535, "learning_rate": 8.732595969497982e-06, "loss": 1.5285, "step": 11904 }, { "epoch": 0.9329937304075235, "grad_norm": 3.06559157371521, "learning_rate": 8.732390789423803e-06, "loss": 1.8611, "step": 11905 }, { "epoch": 0.9330721003134796, "grad_norm": 3.3043980598449707, "learning_rate": 8.732185595153528e-06, "loss": 1.3235, "step": 11906 }, { "epoch": 0.9331504702194358, "grad_norm": 2.9105844497680664, "learning_rate": 8.731980386687934e-06, "loss": 1.3279, "step": 11907 }, { "epoch": 0.9332288401253919, "grad_norm": 3.075631856918335, "learning_rate": 8.731775164027806e-06, "loss": 1.2944, "step": 11908 }, { "epoch": 0.933307210031348, "grad_norm": 2.916581630706787, "learning_rate": 8.731569927173922e-06, "loss": 1.6404, "step": 11909 }, { "epoch": 0.9333855799373041, "grad_norm": 4.110358238220215, "learning_rate": 8.731364676127061e-06, "loss": 1.6103, "step": 11910 }, { "epoch": 0.9334639498432602, "grad_norm": 5.052664279937744, "learning_rate": 8.731159410888007e-06, "loss": 1.2127, "step": 11911 }, { "epoch": 0.9335423197492163, "grad_norm": 3.476860284805298, "learning_rate": 8.730954131457539e-06, "loss": 1.4422, "step": 11912 }, { "epoch": 0.9336206896551724, "grad_norm": 3.539992570877075, "learning_rate": 8.730748837836435e-06, "loss": 1.3138, "step": 11913 }, { "epoch": 0.9336990595611285, "grad_norm": 12.003140449523926, "learning_rate": 8.730543530025481e-06, "loss": 1.1899, "step": 11914 }, { "epoch": 0.9337774294670846, "grad_norm": 3.3529844284057617, "learning_rate": 8.730338208025455e-06, "loss": 1.6489, "step": 11915 }, { "epoch": 0.9338557993730408, "grad_norm": 6.362948894500732, "learning_rate": 8.730132871837139e-06, "loss": 1.2189, "step": 11916 }, { "epoch": 0.9339341692789969, "grad_norm": 4.608683109283447, "learning_rate": 8.729927521461314e-06, "loss": 2.4111, "step": 11917 }, { "epoch": 0.934012539184953, "grad_norm": 4.226072788238525, "learning_rate": 8.72972215689876e-06, "loss": 0.7916, "step": 11918 }, { "epoch": 0.9340909090909091, "grad_norm": 2.8960964679718018, "learning_rate": 8.729516778150258e-06, "loss": 0.8889, "step": 11919 }, { "epoch": 0.9341692789968652, "grad_norm": 4.183039665222168, "learning_rate": 8.72931138521659e-06, "loss": 1.0873, "step": 11920 }, { "epoch": 0.9342476489028213, "grad_norm": 4.476930141448975, "learning_rate": 8.729105978098537e-06, "loss": 1.323, "step": 11921 }, { "epoch": 0.9343260188087774, "grad_norm": 3.3665287494659424, "learning_rate": 8.72890055679688e-06, "loss": 1.6734, "step": 11922 }, { "epoch": 0.9344043887147335, "grad_norm": 4.610522747039795, "learning_rate": 8.7286951213124e-06, "loss": 1.7964, "step": 11923 }, { "epoch": 0.9344827586206896, "grad_norm": 4.934337615966797, "learning_rate": 8.728489671645881e-06, "loss": 1.4112, "step": 11924 }, { "epoch": 0.9345611285266457, "grad_norm": 2.3053832054138184, "learning_rate": 8.7282842077981e-06, "loss": 1.4254, "step": 11925 }, { "epoch": 0.9346394984326019, "grad_norm": 2.8454909324645996, "learning_rate": 8.728078729769842e-06, "loss": 1.3552, "step": 11926 }, { "epoch": 0.934717868338558, "grad_norm": 6.000471591949463, "learning_rate": 8.727873237561889e-06, "loss": 1.2743, "step": 11927 }, { "epoch": 0.9347962382445141, "grad_norm": 11.088276863098145, "learning_rate": 8.727667731175018e-06, "loss": 0.9084, "step": 11928 }, { "epoch": 0.9348746081504702, "grad_norm": 3.175183057785034, "learning_rate": 8.727462210610016e-06, "loss": 1.7477, "step": 11929 }, { "epoch": 0.9349529780564263, "grad_norm": 3.0489447116851807, "learning_rate": 8.72725667586766e-06, "loss": 1.2913, "step": 11930 }, { "epoch": 0.9350313479623824, "grad_norm": 2.375640392303467, "learning_rate": 8.727051126948735e-06, "loss": 1.1147, "step": 11931 }, { "epoch": 0.9351097178683385, "grad_norm": 2.2518136501312256, "learning_rate": 8.726845563854023e-06, "loss": 1.1759, "step": 11932 }, { "epoch": 0.9351880877742946, "grad_norm": 10.349796295166016, "learning_rate": 8.726639986584302e-06, "loss": 1.4823, "step": 11933 }, { "epoch": 0.9352664576802507, "grad_norm": 3.987065076828003, "learning_rate": 8.72643439514036e-06, "loss": 1.4224, "step": 11934 }, { "epoch": 0.9353448275862069, "grad_norm": 7.73882532119751, "learning_rate": 8.726228789522974e-06, "loss": 1.1201, "step": 11935 }, { "epoch": 0.935423197492163, "grad_norm": 3.3715269565582275, "learning_rate": 8.726023169732928e-06, "loss": 1.1756, "step": 11936 }, { "epoch": 0.9355015673981191, "grad_norm": 2.5609853267669678, "learning_rate": 8.725817535771004e-06, "loss": 1.1035, "step": 11937 }, { "epoch": 0.9355799373040752, "grad_norm": 2.9727442264556885, "learning_rate": 8.725611887637982e-06, "loss": 1.5869, "step": 11938 }, { "epoch": 0.9356583072100313, "grad_norm": 4.687160491943359, "learning_rate": 8.725406225334648e-06, "loss": 1.6874, "step": 11939 }, { "epoch": 0.9357366771159875, "grad_norm": 3.799830198287964, "learning_rate": 8.72520054886178e-06, "loss": 1.4657, "step": 11940 }, { "epoch": 0.9358150470219436, "grad_norm": 5.166547775268555, "learning_rate": 8.724994858220164e-06, "loss": 1.6131, "step": 11941 }, { "epoch": 0.9358934169278997, "grad_norm": 2.110732078552246, "learning_rate": 8.72478915341058e-06, "loss": 0.9017, "step": 11942 }, { "epoch": 0.9359717868338558, "grad_norm": 2.4598097801208496, "learning_rate": 8.724583434433813e-06, "loss": 1.4644, "step": 11943 }, { "epoch": 0.936050156739812, "grad_norm": 5.083011627197266, "learning_rate": 8.724377701290643e-06, "loss": 1.5531, "step": 11944 }, { "epoch": 0.9361285266457681, "grad_norm": 2.270725965499878, "learning_rate": 8.724171953981852e-06, "loss": 1.25, "step": 11945 }, { "epoch": 0.9362068965517242, "grad_norm": 3.1896584033966064, "learning_rate": 8.723966192508224e-06, "loss": 1.6529, "step": 11946 }, { "epoch": 0.9362852664576803, "grad_norm": 2.6635539531707764, "learning_rate": 8.723760416870541e-06, "loss": 1.3595, "step": 11947 }, { "epoch": 0.9363636363636364, "grad_norm": 2.731288194656372, "learning_rate": 8.723554627069588e-06, "loss": 1.445, "step": 11948 }, { "epoch": 0.9364420062695925, "grad_norm": 3.523876428604126, "learning_rate": 8.723348823106143e-06, "loss": 0.9519, "step": 11949 }, { "epoch": 0.9365203761755486, "grad_norm": 3.157898187637329, "learning_rate": 8.723143004980995e-06, "loss": 1.5112, "step": 11950 }, { "epoch": 0.9365987460815047, "grad_norm": 8.45236873626709, "learning_rate": 8.72293717269492e-06, "loss": 1.0856, "step": 11951 }, { "epoch": 0.9366771159874608, "grad_norm": 2.970602035522461, "learning_rate": 8.722731326248706e-06, "loss": 1.4436, "step": 11952 }, { "epoch": 0.936755485893417, "grad_norm": 1.960007905960083, "learning_rate": 8.722525465643135e-06, "loss": 1.4397, "step": 11953 }, { "epoch": 0.9368338557993731, "grad_norm": 8.820005416870117, "learning_rate": 8.722319590878988e-06, "loss": 1.4873, "step": 11954 }, { "epoch": 0.9369122257053292, "grad_norm": 4.639423370361328, "learning_rate": 8.722113701957048e-06, "loss": 0.9471, "step": 11955 }, { "epoch": 0.9369905956112853, "grad_norm": 2.8072948455810547, "learning_rate": 8.7219077988781e-06, "loss": 1.4167, "step": 11956 }, { "epoch": 0.9370689655172414, "grad_norm": 2.9394521713256836, "learning_rate": 8.721701881642928e-06, "loss": 1.5121, "step": 11957 }, { "epoch": 0.9371473354231975, "grad_norm": 3.0631935596466064, "learning_rate": 8.721495950252312e-06, "loss": 1.5524, "step": 11958 }, { "epoch": 0.9372257053291536, "grad_norm": 6.151480197906494, "learning_rate": 8.721290004707037e-06, "loss": 1.5629, "step": 11959 }, { "epoch": 0.9373040752351097, "grad_norm": 2.1669602394104004, "learning_rate": 8.721084045007887e-06, "loss": 1.0971, "step": 11960 }, { "epoch": 0.9373824451410658, "grad_norm": 2.9037928581237793, "learning_rate": 8.720878071155643e-06, "loss": 1.1599, "step": 11961 }, { "epoch": 0.937460815047022, "grad_norm": 4.386253356933594, "learning_rate": 8.720672083151091e-06, "loss": 1.1288, "step": 11962 }, { "epoch": 0.937539184952978, "grad_norm": 2.7460360527038574, "learning_rate": 8.720466080995014e-06, "loss": 1.3064, "step": 11963 }, { "epoch": 0.9376175548589342, "grad_norm": 3.041900634765625, "learning_rate": 8.720260064688195e-06, "loss": 1.3154, "step": 11964 }, { "epoch": 0.9376959247648903, "grad_norm": 10.182820320129395, "learning_rate": 8.720054034231416e-06, "loss": 1.1526, "step": 11965 }, { "epoch": 0.9377742946708464, "grad_norm": 3.4933435916900635, "learning_rate": 8.719847989625465e-06, "loss": 1.5345, "step": 11966 }, { "epoch": 0.9378526645768025, "grad_norm": 7.613632678985596, "learning_rate": 8.71964193087112e-06, "loss": 1.3228, "step": 11967 }, { "epoch": 0.9379310344827586, "grad_norm": 4.2623066902160645, "learning_rate": 8.719435857969168e-06, "loss": 1.6421, "step": 11968 }, { "epoch": 0.9380094043887147, "grad_norm": 2.5159666538238525, "learning_rate": 8.719229770920395e-06, "loss": 1.4077, "step": 11969 }, { "epoch": 0.9380877742946708, "grad_norm": 2.182373285293579, "learning_rate": 8.719023669725579e-06, "loss": 1.49, "step": 11970 }, { "epoch": 0.9381661442006269, "grad_norm": 4.587958812713623, "learning_rate": 8.718817554385509e-06, "loss": 1.4387, "step": 11971 }, { "epoch": 0.938244514106583, "grad_norm": 2.5787343978881836, "learning_rate": 8.718611424900968e-06, "loss": 1.3454, "step": 11972 }, { "epoch": 0.9383228840125392, "grad_norm": 2.2329721450805664, "learning_rate": 8.718405281272736e-06, "loss": 1.5696, "step": 11973 }, { "epoch": 0.9384012539184953, "grad_norm": 2.147944688796997, "learning_rate": 8.718199123501603e-06, "loss": 1.2656, "step": 11974 }, { "epoch": 0.9384796238244514, "grad_norm": 4.421748161315918, "learning_rate": 8.717992951588348e-06, "loss": 1.2535, "step": 11975 }, { "epoch": 0.9385579937304075, "grad_norm": 6.788452625274658, "learning_rate": 8.71778676553376e-06, "loss": 1.6563, "step": 11976 }, { "epoch": 0.9386363636363636, "grad_norm": 3.4603803157806396, "learning_rate": 8.71758056533862e-06, "loss": 1.6154, "step": 11977 }, { "epoch": 0.9387147335423197, "grad_norm": 3.651189088821411, "learning_rate": 8.717374351003712e-06, "loss": 1.5271, "step": 11978 }, { "epoch": 0.9387931034482758, "grad_norm": 2.6463608741760254, "learning_rate": 8.717168122529822e-06, "loss": 1.0845, "step": 11979 }, { "epoch": 0.9388714733542319, "grad_norm": 2.1843972206115723, "learning_rate": 8.716961879917734e-06, "loss": 1.1479, "step": 11980 }, { "epoch": 0.938949843260188, "grad_norm": 4.130764484405518, "learning_rate": 8.716755623168232e-06, "loss": 1.62, "step": 11981 }, { "epoch": 0.9390282131661442, "grad_norm": 2.516108989715576, "learning_rate": 8.7165493522821e-06, "loss": 1.3965, "step": 11982 }, { "epoch": 0.9391065830721003, "grad_norm": 6.972297191619873, "learning_rate": 8.716343067260124e-06, "loss": 1.809, "step": 11983 }, { "epoch": 0.9391849529780564, "grad_norm": 2.4554505348205566, "learning_rate": 8.716136768103088e-06, "loss": 1.5422, "step": 11984 }, { "epoch": 0.9392633228840125, "grad_norm": 2.6208014488220215, "learning_rate": 8.715930454811774e-06, "loss": 1.308, "step": 11985 }, { "epoch": 0.9393416927899687, "grad_norm": 2.4877655506134033, "learning_rate": 8.715724127386971e-06, "loss": 1.4247, "step": 11986 }, { "epoch": 0.9394200626959248, "grad_norm": 1.8368382453918457, "learning_rate": 8.715517785829463e-06, "loss": 1.3973, "step": 11987 }, { "epoch": 0.9394984326018809, "grad_norm": 3.3089957237243652, "learning_rate": 8.715311430140032e-06, "loss": 1.4114, "step": 11988 }, { "epoch": 0.939576802507837, "grad_norm": 2.3371264934539795, "learning_rate": 8.715105060319467e-06, "loss": 1.2881, "step": 11989 }, { "epoch": 0.9396551724137931, "grad_norm": 3.236987829208374, "learning_rate": 8.71489867636855e-06, "loss": 1.4288, "step": 11990 }, { "epoch": 0.9397335423197493, "grad_norm": 2.267052173614502, "learning_rate": 8.714692278288063e-06, "loss": 1.4615, "step": 11991 }, { "epoch": 0.9398119122257054, "grad_norm": 8.100772857666016, "learning_rate": 8.714485866078798e-06, "loss": 2.0152, "step": 11992 }, { "epoch": 0.9398902821316615, "grad_norm": 3.2796413898468018, "learning_rate": 8.714279439741535e-06, "loss": 1.6304, "step": 11993 }, { "epoch": 0.9399686520376176, "grad_norm": 7.9579758644104, "learning_rate": 8.71407299927706e-06, "loss": 1.3584, "step": 11994 }, { "epoch": 0.9400470219435737, "grad_norm": 2.5502750873565674, "learning_rate": 8.713866544686163e-06, "loss": 1.4336, "step": 11995 }, { "epoch": 0.9401253918495298, "grad_norm": 3.208461284637451, "learning_rate": 8.713660075969623e-06, "loss": 1.0788, "step": 11996 }, { "epoch": 0.9402037617554859, "grad_norm": 3.9815144538879395, "learning_rate": 8.713453593128227e-06, "loss": 1.3492, "step": 11997 }, { "epoch": 0.940282131661442, "grad_norm": 2.682772397994995, "learning_rate": 8.71324709616276e-06, "loss": 1.5641, "step": 11998 }, { "epoch": 0.9403605015673981, "grad_norm": 2.0254874229431152, "learning_rate": 8.713040585074009e-06, "loss": 1.4214, "step": 11999 }, { "epoch": 0.9404388714733543, "grad_norm": 3.281292200088501, "learning_rate": 8.71283405986276e-06, "loss": 1.3243, "step": 12000 }, { "epoch": 0.9405172413793104, "grad_norm": 2.6212046146392822, "learning_rate": 8.712627520529796e-06, "loss": 1.2641, "step": 12001 }, { "epoch": 0.9405956112852665, "grad_norm": 2.276409387588501, "learning_rate": 8.712420967075904e-06, "loss": 1.4038, "step": 12002 }, { "epoch": 0.9406739811912226, "grad_norm": 1.9174264669418335, "learning_rate": 8.712214399501871e-06, "loss": 1.07, "step": 12003 }, { "epoch": 0.9407523510971787, "grad_norm": 3.063432216644287, "learning_rate": 8.712007817808479e-06, "loss": 0.5467, "step": 12004 }, { "epoch": 0.9408307210031348, "grad_norm": 2.695356845855713, "learning_rate": 8.711801221996517e-06, "loss": 1.5003, "step": 12005 }, { "epoch": 0.9409090909090909, "grad_norm": 4.247498035430908, "learning_rate": 8.71159461206677e-06, "loss": 1.4283, "step": 12006 }, { "epoch": 0.940987460815047, "grad_norm": 2.9486770629882812, "learning_rate": 8.711387988020023e-06, "loss": 1.3833, "step": 12007 }, { "epoch": 0.9410658307210031, "grad_norm": 2.296950101852417, "learning_rate": 8.71118134985706e-06, "loss": 1.5829, "step": 12008 }, { "epoch": 0.9411442006269592, "grad_norm": 2.7029154300689697, "learning_rate": 8.710974697578674e-06, "loss": 1.4021, "step": 12009 }, { "epoch": 0.9412225705329154, "grad_norm": 3.3216681480407715, "learning_rate": 8.710768031185643e-06, "loss": 1.0691, "step": 12010 }, { "epoch": 0.9413009404388715, "grad_norm": 3.4935262203216553, "learning_rate": 8.710561350678756e-06, "loss": 1.6976, "step": 12011 }, { "epoch": 0.9413793103448276, "grad_norm": 3.275029182434082, "learning_rate": 8.710354656058801e-06, "loss": 1.1457, "step": 12012 }, { "epoch": 0.9414576802507837, "grad_norm": 2.6171069145202637, "learning_rate": 8.710147947326564e-06, "loss": 1.3886, "step": 12013 }, { "epoch": 0.9415360501567398, "grad_norm": 3.022341251373291, "learning_rate": 8.709941224482828e-06, "loss": 1.2624, "step": 12014 }, { "epoch": 0.9416144200626959, "grad_norm": 3.4164390563964844, "learning_rate": 8.70973448752838e-06, "loss": 0.7996, "step": 12015 }, { "epoch": 0.941692789968652, "grad_norm": 2.4687623977661133, "learning_rate": 8.70952773646401e-06, "loss": 1.4848, "step": 12016 }, { "epoch": 0.9417711598746081, "grad_norm": 2.517008066177368, "learning_rate": 8.7093209712905e-06, "loss": 1.3705, "step": 12017 }, { "epoch": 0.9418495297805642, "grad_norm": 2.545062780380249, "learning_rate": 8.70911419200864e-06, "loss": 1.3515, "step": 12018 }, { "epoch": 0.9419278996865204, "grad_norm": 2.4208221435546875, "learning_rate": 8.708907398619214e-06, "loss": 1.6604, "step": 12019 }, { "epoch": 0.9420062695924765, "grad_norm": 3.2785415649414062, "learning_rate": 8.708700591123008e-06, "loss": 1.4179, "step": 12020 }, { "epoch": 0.9420846394984326, "grad_norm": 10.110891342163086, "learning_rate": 8.708493769520812e-06, "loss": 1.8792, "step": 12021 }, { "epoch": 0.9421630094043887, "grad_norm": 3.269347667694092, "learning_rate": 8.708286933813408e-06, "loss": 1.2933, "step": 12022 }, { "epoch": 0.9422413793103448, "grad_norm": 2.426886558532715, "learning_rate": 8.708080084001586e-06, "loss": 1.3477, "step": 12023 }, { "epoch": 0.9423197492163009, "grad_norm": 2.0029070377349854, "learning_rate": 8.707873220086134e-06, "loss": 1.324, "step": 12024 }, { "epoch": 0.942398119122257, "grad_norm": 4.533164978027344, "learning_rate": 8.707666342067835e-06, "loss": 1.4934, "step": 12025 }, { "epoch": 0.9424764890282131, "grad_norm": 3.0832359790802, "learning_rate": 8.707459449947478e-06, "loss": 1.271, "step": 12026 }, { "epoch": 0.9425548589341692, "grad_norm": 9.803574562072754, "learning_rate": 8.707252543725849e-06, "loss": 1.6292, "step": 12027 }, { "epoch": 0.9426332288401253, "grad_norm": 2.140977144241333, "learning_rate": 8.707045623403736e-06, "loss": 1.3437, "step": 12028 }, { "epoch": 0.9427115987460815, "grad_norm": 3.0514307022094727, "learning_rate": 8.706838688981924e-06, "loss": 1.2213, "step": 12029 }, { "epoch": 0.9427899686520376, "grad_norm": 3.0221798419952393, "learning_rate": 8.706631740461204e-06, "loss": 1.6373, "step": 12030 }, { "epoch": 0.9428683385579937, "grad_norm": 2.9634644985198975, "learning_rate": 8.706424777842359e-06, "loss": 1.8061, "step": 12031 }, { "epoch": 0.9429467084639498, "grad_norm": 2.5796990394592285, "learning_rate": 8.706217801126179e-06, "loss": 1.3683, "step": 12032 }, { "epoch": 0.943025078369906, "grad_norm": 3.4710381031036377, "learning_rate": 8.70601081031345e-06, "loss": 1.0265, "step": 12033 }, { "epoch": 0.9431034482758621, "grad_norm": 3.7068471908569336, "learning_rate": 8.70580380540496e-06, "loss": 1.3785, "step": 12034 }, { "epoch": 0.9431818181818182, "grad_norm": 1.956381916999817, "learning_rate": 8.705596786401493e-06, "loss": 1.5088, "step": 12035 }, { "epoch": 0.9432601880877743, "grad_norm": 2.675291061401367, "learning_rate": 8.705389753303842e-06, "loss": 1.5906, "step": 12036 }, { "epoch": 0.9433385579937305, "grad_norm": 2.9411091804504395, "learning_rate": 8.70518270611279e-06, "loss": 0.8587, "step": 12037 }, { "epoch": 0.9434169278996866, "grad_norm": 2.5472588539123535, "learning_rate": 8.704975644829127e-06, "loss": 1.4626, "step": 12038 }, { "epoch": 0.9434952978056427, "grad_norm": 3.946258068084717, "learning_rate": 8.70476856945364e-06, "loss": 1.4967, "step": 12039 }, { "epoch": 0.9435736677115988, "grad_norm": 6.560395240783691, "learning_rate": 8.704561479987115e-06, "loss": 1.855, "step": 12040 }, { "epoch": 0.9436520376175549, "grad_norm": 2.540860414505005, "learning_rate": 8.70435437643034e-06, "loss": 1.9075, "step": 12041 }, { "epoch": 0.943730407523511, "grad_norm": 2.8604156970977783, "learning_rate": 8.704147258784106e-06, "loss": 1.1113, "step": 12042 }, { "epoch": 0.9438087774294671, "grad_norm": 3.480015516281128, "learning_rate": 8.703940127049197e-06, "loss": 1.5738, "step": 12043 }, { "epoch": 0.9438871473354232, "grad_norm": 5.0133957862854, "learning_rate": 8.703732981226402e-06, "loss": 1.0727, "step": 12044 }, { "epoch": 0.9439655172413793, "grad_norm": 2.009530782699585, "learning_rate": 8.703525821316511e-06, "loss": 1.3362, "step": 12045 }, { "epoch": 0.9440438871473354, "grad_norm": 4.362300872802734, "learning_rate": 8.70331864732031e-06, "loss": 1.2224, "step": 12046 }, { "epoch": 0.9441222570532916, "grad_norm": 3.5376176834106445, "learning_rate": 8.703111459238585e-06, "loss": 1.2854, "step": 12047 }, { "epoch": 0.9442006269592477, "grad_norm": 3.057650327682495, "learning_rate": 8.702904257072126e-06, "loss": 1.4645, "step": 12048 }, { "epoch": 0.9442789968652038, "grad_norm": 2.318087339401245, "learning_rate": 8.702697040821722e-06, "loss": 1.2382, "step": 12049 }, { "epoch": 0.9443573667711599, "grad_norm": 3.086311101913452, "learning_rate": 8.702489810488161e-06, "loss": 1.622, "step": 12050 }, { "epoch": 0.944435736677116, "grad_norm": 7.713751792907715, "learning_rate": 8.70228256607223e-06, "loss": 1.3719, "step": 12051 }, { "epoch": 0.9445141065830721, "grad_norm": 6.887241840362549, "learning_rate": 8.702075307574716e-06, "loss": 2.0767, "step": 12052 }, { "epoch": 0.9445924764890282, "grad_norm": 3.829507827758789, "learning_rate": 8.701868034996411e-06, "loss": 1.5932, "step": 12053 }, { "epoch": 0.9446708463949843, "grad_norm": 2.823227643966675, "learning_rate": 8.7016607483381e-06, "loss": 1.3933, "step": 12054 }, { "epoch": 0.9447492163009404, "grad_norm": 2.7948036193847656, "learning_rate": 8.701453447600574e-06, "loss": 1.1349, "step": 12055 }, { "epoch": 0.9448275862068966, "grad_norm": 14.50996208190918, "learning_rate": 8.701246132784622e-06, "loss": 1.8841, "step": 12056 }, { "epoch": 0.9449059561128527, "grad_norm": 4.627837181091309, "learning_rate": 8.701038803891027e-06, "loss": 1.1322, "step": 12057 }, { "epoch": 0.9449843260188088, "grad_norm": 3.098175525665283, "learning_rate": 8.700831460920584e-06, "loss": 1.1837, "step": 12058 }, { "epoch": 0.9450626959247649, "grad_norm": 3.1827807426452637, "learning_rate": 8.700624103874078e-06, "loss": 1.6042, "step": 12059 }, { "epoch": 0.945141065830721, "grad_norm": 2.841165065765381, "learning_rate": 8.700416732752298e-06, "loss": 1.1775, "step": 12060 }, { "epoch": 0.9452194357366771, "grad_norm": 2.777409791946411, "learning_rate": 8.700209347556034e-06, "loss": 1.639, "step": 12061 }, { "epoch": 0.9452978056426332, "grad_norm": 3.2858569622039795, "learning_rate": 8.700001948286074e-06, "loss": 1.5377, "step": 12062 }, { "epoch": 0.9453761755485893, "grad_norm": 3.3120691776275635, "learning_rate": 8.699794534943208e-06, "loss": 1.4231, "step": 12063 }, { "epoch": 0.9454545454545454, "grad_norm": 2.345195770263672, "learning_rate": 8.699587107528224e-06, "loss": 1.271, "step": 12064 }, { "epoch": 0.9455329153605015, "grad_norm": 3.0974652767181396, "learning_rate": 8.69937966604191e-06, "loss": 1.3798, "step": 12065 }, { "epoch": 0.9456112852664577, "grad_norm": 6.817758083343506, "learning_rate": 8.699172210485056e-06, "loss": 1.5047, "step": 12066 }, { "epoch": 0.9456896551724138, "grad_norm": 3.0643532276153564, "learning_rate": 8.69896474085845e-06, "loss": 0.8045, "step": 12067 }, { "epoch": 0.9457680250783699, "grad_norm": 2.23126220703125, "learning_rate": 8.698757257162883e-06, "loss": 1.4671, "step": 12068 }, { "epoch": 0.945846394984326, "grad_norm": 2.3533737659454346, "learning_rate": 8.698549759399144e-06, "loss": 1.2948, "step": 12069 }, { "epoch": 0.9459247648902821, "grad_norm": 4.802262306213379, "learning_rate": 8.698342247568021e-06, "loss": 0.8701, "step": 12070 }, { "epoch": 0.9460031347962382, "grad_norm": 2.699481725692749, "learning_rate": 8.698134721670304e-06, "loss": 1.0372, "step": 12071 }, { "epoch": 0.9460815047021943, "grad_norm": 3.1774632930755615, "learning_rate": 8.69792718170678e-06, "loss": 1.2277, "step": 12072 }, { "epoch": 0.9461598746081504, "grad_norm": 2.189523220062256, "learning_rate": 8.697719627678243e-06, "loss": 1.1025, "step": 12073 }, { "epoch": 0.9462382445141065, "grad_norm": 2.800542116165161, "learning_rate": 8.697512059585478e-06, "loss": 1.1354, "step": 12074 }, { "epoch": 0.9463166144200627, "grad_norm": 2.446057081222534, "learning_rate": 8.697304477429275e-06, "loss": 1.602, "step": 12075 }, { "epoch": 0.9463949843260188, "grad_norm": 2.5143933296203613, "learning_rate": 8.697096881210428e-06, "loss": 1.2913, "step": 12076 }, { "epoch": 0.9464733542319749, "grad_norm": 4.593845367431641, "learning_rate": 8.696889270929722e-06, "loss": 1.4444, "step": 12077 }, { "epoch": 0.946551724137931, "grad_norm": 3.582167863845825, "learning_rate": 8.696681646587947e-06, "loss": 2.0325, "step": 12078 }, { "epoch": 0.9466300940438871, "grad_norm": 2.1884000301361084, "learning_rate": 8.696474008185895e-06, "loss": 1.5205, "step": 12079 }, { "epoch": 0.9467084639498433, "grad_norm": 3.8507237434387207, "learning_rate": 8.696266355724355e-06, "loss": 1.3311, "step": 12080 }, { "epoch": 0.9467868338557994, "grad_norm": 4.009225368499756, "learning_rate": 8.696058689204117e-06, "loss": 0.8845, "step": 12081 }, { "epoch": 0.9468652037617555, "grad_norm": 2.4644081592559814, "learning_rate": 8.695851008625969e-06, "loss": 1.3249, "step": 12082 }, { "epoch": 0.9469435736677116, "grad_norm": 4.277853488922119, "learning_rate": 8.6956433139907e-06, "loss": 1.472, "step": 12083 }, { "epoch": 0.9470219435736678, "grad_norm": 5.166592597961426, "learning_rate": 8.695435605299104e-06, "loss": 1.3193, "step": 12084 }, { "epoch": 0.9471003134796239, "grad_norm": 2.582306146621704, "learning_rate": 8.69522788255197e-06, "loss": 1.2015, "step": 12085 }, { "epoch": 0.94717868338558, "grad_norm": 2.7948741912841797, "learning_rate": 8.695020145750086e-06, "loss": 1.3794, "step": 12086 }, { "epoch": 0.9472570532915361, "grad_norm": 4.085352420806885, "learning_rate": 8.694812394894242e-06, "loss": 1.3483, "step": 12087 }, { "epoch": 0.9473354231974922, "grad_norm": 2.8474485874176025, "learning_rate": 8.694604629985231e-06, "loss": 1.3405, "step": 12088 }, { "epoch": 0.9474137931034483, "grad_norm": 2.28778076171875, "learning_rate": 8.694396851023843e-06, "loss": 1.2003, "step": 12089 }, { "epoch": 0.9474921630094044, "grad_norm": 3.3584439754486084, "learning_rate": 8.694189058010864e-06, "loss": 1.4228, "step": 12090 }, { "epoch": 0.9475705329153605, "grad_norm": 1.922336220741272, "learning_rate": 8.69398125094709e-06, "loss": 1.3023, "step": 12091 }, { "epoch": 0.9476489028213166, "grad_norm": 3.455941915512085, "learning_rate": 8.693773429833307e-06, "loss": 1.3235, "step": 12092 }, { "epoch": 0.9477272727272728, "grad_norm": 2.6778934001922607, "learning_rate": 8.693565594670308e-06, "loss": 1.1766, "step": 12093 }, { "epoch": 0.9478056426332289, "grad_norm": 2.1089119911193848, "learning_rate": 8.693357745458882e-06, "loss": 1.5352, "step": 12094 }, { "epoch": 0.947884012539185, "grad_norm": 3.0270636081695557, "learning_rate": 8.69314988219982e-06, "loss": 1.3887, "step": 12095 }, { "epoch": 0.9479623824451411, "grad_norm": 2.9980762004852295, "learning_rate": 8.692942004893912e-06, "loss": 1.7428, "step": 12096 }, { "epoch": 0.9480407523510972, "grad_norm": 2.599012613296509, "learning_rate": 8.692734113541951e-06, "loss": 0.9516, "step": 12097 }, { "epoch": 0.9481191222570533, "grad_norm": 3.4455606937408447, "learning_rate": 8.692526208144724e-06, "loss": 1.4351, "step": 12098 }, { "epoch": 0.9481974921630094, "grad_norm": 6.923947811126709, "learning_rate": 8.692318288703027e-06, "loss": 0.9833, "step": 12099 }, { "epoch": 0.9482758620689655, "grad_norm": 7.6985182762146, "learning_rate": 8.692110355217646e-06, "loss": 1.0718, "step": 12100 }, { "epoch": 0.9483542319749216, "grad_norm": 3.9225168228149414, "learning_rate": 8.691902407689373e-06, "loss": 2.2886, "step": 12101 }, { "epoch": 0.9484326018808777, "grad_norm": 2.0601837635040283, "learning_rate": 8.691694446119001e-06, "loss": 1.1183, "step": 12102 }, { "epoch": 0.9485109717868339, "grad_norm": 2.6893346309661865, "learning_rate": 8.691486470507318e-06, "loss": 1.4106, "step": 12103 }, { "epoch": 0.94858934169279, "grad_norm": 2.2413930892944336, "learning_rate": 8.69127848085512e-06, "loss": 1.3331, "step": 12104 }, { "epoch": 0.9486677115987461, "grad_norm": 2.151475667953491, "learning_rate": 8.69107047716319e-06, "loss": 1.3039, "step": 12105 }, { "epoch": 0.9487460815047022, "grad_norm": 1.8557024002075195, "learning_rate": 8.690862459432325e-06, "loss": 1.3061, "step": 12106 }, { "epoch": 0.9488244514106583, "grad_norm": 2.9951038360595703, "learning_rate": 8.690654427663317e-06, "loss": 0.9402, "step": 12107 }, { "epoch": 0.9489028213166144, "grad_norm": 2.8658533096313477, "learning_rate": 8.690446381856954e-06, "loss": 1.479, "step": 12108 }, { "epoch": 0.9489811912225705, "grad_norm": 1.7546405792236328, "learning_rate": 8.690238322014028e-06, "loss": 1.499, "step": 12109 }, { "epoch": 0.9490595611285266, "grad_norm": 3.7473251819610596, "learning_rate": 8.69003024813533e-06, "loss": 1.4703, "step": 12110 }, { "epoch": 0.9491379310344827, "grad_norm": 2.7593955993652344, "learning_rate": 8.689822160221655e-06, "loss": 1.4375, "step": 12111 }, { "epoch": 0.9492163009404389, "grad_norm": 4.195639610290527, "learning_rate": 8.68961405827379e-06, "loss": 1.2595, "step": 12112 }, { "epoch": 0.949294670846395, "grad_norm": 2.8074400424957275, "learning_rate": 8.689405942292528e-06, "loss": 1.6645, "step": 12113 }, { "epoch": 0.9493730407523511, "grad_norm": 1.9051889181137085, "learning_rate": 8.68919781227866e-06, "loss": 1.139, "step": 12114 }, { "epoch": 0.9494514106583072, "grad_norm": 4.288721561431885, "learning_rate": 8.688989668232982e-06, "loss": 1.2831, "step": 12115 }, { "epoch": 0.9495297805642633, "grad_norm": 7.998006820678711, "learning_rate": 8.688781510156279e-06, "loss": 1.6885, "step": 12116 }, { "epoch": 0.9496081504702194, "grad_norm": 2.548492193222046, "learning_rate": 8.688573338049346e-06, "loss": 1.5049, "step": 12117 }, { "epoch": 0.9496865203761755, "grad_norm": 3.4847822189331055, "learning_rate": 8.688365151912974e-06, "loss": 0.9896, "step": 12118 }, { "epoch": 0.9497648902821316, "grad_norm": 3.421598434448242, "learning_rate": 8.688156951747957e-06, "loss": 1.7924, "step": 12119 }, { "epoch": 0.9498432601880877, "grad_norm": 7.603511810302734, "learning_rate": 8.687948737555086e-06, "loss": 1.5355, "step": 12120 }, { "epoch": 0.9499216300940438, "grad_norm": 2.9790048599243164, "learning_rate": 8.687740509335149e-06, "loss": 1.6165, "step": 12121 }, { "epoch": 0.95, "grad_norm": 2.5920701026916504, "learning_rate": 8.687532267088945e-06, "loss": 1.3654, "step": 12122 }, { "epoch": 0.9500783699059561, "grad_norm": 2.8317601680755615, "learning_rate": 8.68732401081726e-06, "loss": 1.1497, "step": 12123 }, { "epoch": 0.9501567398119122, "grad_norm": 3.1882121562957764, "learning_rate": 8.68711574052089e-06, "loss": 1.5012, "step": 12124 }, { "epoch": 0.9502351097178683, "grad_norm": 2.0895581245422363, "learning_rate": 8.686907456200625e-06, "loss": 1.4669, "step": 12125 }, { "epoch": 0.9503134796238244, "grad_norm": 2.759326457977295, "learning_rate": 8.686699157857256e-06, "loss": 1.297, "step": 12126 }, { "epoch": 0.9503918495297806, "grad_norm": 2.8080801963806152, "learning_rate": 8.68649084549158e-06, "loss": 1.1938, "step": 12127 }, { "epoch": 0.9504702194357367, "grad_norm": 2.9553749561309814, "learning_rate": 8.686282519104383e-06, "loss": 1.6578, "step": 12128 }, { "epoch": 0.9505485893416928, "grad_norm": 2.465940237045288, "learning_rate": 8.686074178696463e-06, "loss": 1.5086, "step": 12129 }, { "epoch": 0.950626959247649, "grad_norm": 4.345949649810791, "learning_rate": 8.685865824268608e-06, "loss": 1.8096, "step": 12130 }, { "epoch": 0.9507053291536051, "grad_norm": 2.557741641998291, "learning_rate": 8.685657455821615e-06, "loss": 1.5855, "step": 12131 }, { "epoch": 0.9507836990595612, "grad_norm": 2.032822608947754, "learning_rate": 8.685449073356274e-06, "loss": 1.1157, "step": 12132 }, { "epoch": 0.9508620689655173, "grad_norm": 6.242372035980225, "learning_rate": 8.685240676873377e-06, "loss": 1.2364, "step": 12133 }, { "epoch": 0.9509404388714734, "grad_norm": 2.9410147666931152, "learning_rate": 8.685032266373717e-06, "loss": 1.41, "step": 12134 }, { "epoch": 0.9510188087774295, "grad_norm": 6.481821537017822, "learning_rate": 8.684823841858086e-06, "loss": 1.9528, "step": 12135 }, { "epoch": 0.9510971786833856, "grad_norm": 3.5526020526885986, "learning_rate": 8.684615403327278e-06, "loss": 1.3692, "step": 12136 }, { "epoch": 0.9511755485893417, "grad_norm": 2.5717246532440186, "learning_rate": 8.684406950782087e-06, "loss": 1.4645, "step": 12137 }, { "epoch": 0.9512539184952978, "grad_norm": 5.125346660614014, "learning_rate": 8.684198484223305e-06, "loss": 0.9338, "step": 12138 }, { "epoch": 0.951332288401254, "grad_norm": 2.726985454559326, "learning_rate": 8.683990003651723e-06, "loss": 1.446, "step": 12139 }, { "epoch": 0.95141065830721, "grad_norm": 2.580155849456787, "learning_rate": 8.683781509068134e-06, "loss": 1.3657, "step": 12140 }, { "epoch": 0.9514890282131662, "grad_norm": 2.6173324584960938, "learning_rate": 8.683573000473336e-06, "loss": 1.4146, "step": 12141 }, { "epoch": 0.9515673981191223, "grad_norm": 3.2408807277679443, "learning_rate": 8.683364477868115e-06, "loss": 1.8812, "step": 12142 }, { "epoch": 0.9516457680250784, "grad_norm": 4.684264659881592, "learning_rate": 8.683155941253269e-06, "loss": 1.6306, "step": 12143 }, { "epoch": 0.9517241379310345, "grad_norm": 2.017106771469116, "learning_rate": 8.682947390629587e-06, "loss": 1.46, "step": 12144 }, { "epoch": 0.9518025078369906, "grad_norm": 2.335533857345581, "learning_rate": 8.682738825997868e-06, "loss": 1.2134, "step": 12145 }, { "epoch": 0.9518808777429467, "grad_norm": 3.397529125213623, "learning_rate": 8.6825302473589e-06, "loss": 1.5619, "step": 12146 }, { "epoch": 0.9519592476489028, "grad_norm": 3.05100154876709, "learning_rate": 8.68232165471348e-06, "loss": 1.3865, "step": 12147 }, { "epoch": 0.9520376175548589, "grad_norm": 2.704803943634033, "learning_rate": 8.6821130480624e-06, "loss": 1.3188, "step": 12148 }, { "epoch": 0.952115987460815, "grad_norm": 2.35787296295166, "learning_rate": 8.681904427406452e-06, "loss": 1.7987, "step": 12149 }, { "epoch": 0.9521943573667712, "grad_norm": 2.559446334838867, "learning_rate": 8.68169579274643e-06, "loss": 1.3433, "step": 12150 }, { "epoch": 0.9522727272727273, "grad_norm": 5.092039585113525, "learning_rate": 8.68148714408313e-06, "loss": 1.8208, "step": 12151 }, { "epoch": 0.9523510971786834, "grad_norm": 8.456901550292969, "learning_rate": 8.681278481417343e-06, "loss": 1.4008, "step": 12152 }, { "epoch": 0.9524294670846395, "grad_norm": 5.826716899871826, "learning_rate": 8.681069804749863e-06, "loss": 1.314, "step": 12153 }, { "epoch": 0.9525078369905956, "grad_norm": 9.595178604125977, "learning_rate": 8.680861114081484e-06, "loss": 1.2581, "step": 12154 }, { "epoch": 0.9525862068965517, "grad_norm": 2.7091712951660156, "learning_rate": 8.680652409413002e-06, "loss": 1.3553, "step": 12155 }, { "epoch": 0.9526645768025078, "grad_norm": 2.621370315551758, "learning_rate": 8.680443690745206e-06, "loss": 1.0387, "step": 12156 }, { "epoch": 0.9527429467084639, "grad_norm": 2.4411933422088623, "learning_rate": 8.680234958078895e-06, "loss": 1.4447, "step": 12157 }, { "epoch": 0.95282131661442, "grad_norm": 6.767532825469971, "learning_rate": 8.68002621141486e-06, "loss": 1.7608, "step": 12158 }, { "epoch": 0.9528996865203762, "grad_norm": 2.0021467208862305, "learning_rate": 8.679817450753895e-06, "loss": 1.4674, "step": 12159 }, { "epoch": 0.9529780564263323, "grad_norm": 2.2565693855285645, "learning_rate": 8.679608676096793e-06, "loss": 1.1763, "step": 12160 }, { "epoch": 0.9530564263322884, "grad_norm": 2.2717254161834717, "learning_rate": 8.679399887444351e-06, "loss": 1.1798, "step": 12161 }, { "epoch": 0.9531347962382445, "grad_norm": 2.4325759410858154, "learning_rate": 8.679191084797363e-06, "loss": 0.9788, "step": 12162 }, { "epoch": 0.9532131661442006, "grad_norm": 3.0102598667144775, "learning_rate": 8.67898226815662e-06, "loss": 1.3844, "step": 12163 }, { "epoch": 0.9532915360501567, "grad_norm": 2.5585107803344727, "learning_rate": 8.678773437522917e-06, "loss": 1.4991, "step": 12164 }, { "epoch": 0.9533699059561128, "grad_norm": 2.291456937789917, "learning_rate": 8.678564592897052e-06, "loss": 1.129, "step": 12165 }, { "epoch": 0.9534482758620689, "grad_norm": 2.3733768463134766, "learning_rate": 8.678355734279816e-06, "loss": 1.3868, "step": 12166 }, { "epoch": 0.953526645768025, "grad_norm": 1.960233211517334, "learning_rate": 8.678146861672003e-06, "loss": 1.5711, "step": 12167 }, { "epoch": 0.9536050156739811, "grad_norm": 4.855475902557373, "learning_rate": 8.677937975074412e-06, "loss": 1.4671, "step": 12168 }, { "epoch": 0.9536833855799373, "grad_norm": 6.279525279998779, "learning_rate": 8.67772907448783e-06, "loss": 1.8643, "step": 12169 }, { "epoch": 0.9537617554858934, "grad_norm": 2.9183528423309326, "learning_rate": 8.677520159913057e-06, "loss": 1.5497, "step": 12170 }, { "epoch": 0.9538401253918495, "grad_norm": 1.92551589012146, "learning_rate": 8.677311231350887e-06, "loss": 1.1749, "step": 12171 }, { "epoch": 0.9539184952978056, "grad_norm": 3.574672222137451, "learning_rate": 8.677102288802113e-06, "loss": 1.7069, "step": 12172 }, { "epoch": 0.9539968652037617, "grad_norm": 4.380678653717041, "learning_rate": 8.676893332267531e-06, "loss": 1.6194, "step": 12173 }, { "epoch": 0.9540752351097179, "grad_norm": 5.604573726654053, "learning_rate": 8.676684361747933e-06, "loss": 1.4673, "step": 12174 }, { "epoch": 0.954153605015674, "grad_norm": 2.672248601913452, "learning_rate": 8.67647537724412e-06, "loss": 1.6255, "step": 12175 }, { "epoch": 0.9542319749216301, "grad_norm": 2.3821427822113037, "learning_rate": 8.676266378756882e-06, "loss": 1.3373, "step": 12176 }, { "epoch": 0.9543103448275863, "grad_norm": 2.995574474334717, "learning_rate": 8.676057366287013e-06, "loss": 1.5437, "step": 12177 }, { "epoch": 0.9543887147335424, "grad_norm": 2.927851915359497, "learning_rate": 8.675848339835311e-06, "loss": 1.1637, "step": 12178 }, { "epoch": 0.9544670846394985, "grad_norm": 2.1364974975585938, "learning_rate": 8.67563929940257e-06, "loss": 1.4861, "step": 12179 }, { "epoch": 0.9545454545454546, "grad_norm": 2.710462808609009, "learning_rate": 8.675430244989584e-06, "loss": 1.2976, "step": 12180 }, { "epoch": 0.9546238244514107, "grad_norm": 3.7794206142425537, "learning_rate": 8.67522117659715e-06, "loss": 1.7798, "step": 12181 }, { "epoch": 0.9547021943573668, "grad_norm": 2.2033586502075195, "learning_rate": 8.675012094226063e-06, "loss": 1.8131, "step": 12182 }, { "epoch": 0.9547805642633229, "grad_norm": 6.688734531402588, "learning_rate": 8.674802997877117e-06, "loss": 1.6897, "step": 12183 }, { "epoch": 0.954858934169279, "grad_norm": 3.0505456924438477, "learning_rate": 8.67459388755111e-06, "loss": 1.3391, "step": 12184 }, { "epoch": 0.9549373040752351, "grad_norm": 2.3284621238708496, "learning_rate": 8.674384763248833e-06, "loss": 1.5364, "step": 12185 }, { "epoch": 0.9550156739811912, "grad_norm": 3.645297050476074, "learning_rate": 8.674175624971084e-06, "loss": 1.3614, "step": 12186 }, { "epoch": 0.9550940438871474, "grad_norm": 2.604163408279419, "learning_rate": 8.673966472718658e-06, "loss": 1.3839, "step": 12187 }, { "epoch": 0.9551724137931035, "grad_norm": 2.586655378341675, "learning_rate": 8.67375730649235e-06, "loss": 1.4407, "step": 12188 }, { "epoch": 0.9552507836990596, "grad_norm": 2.3121907711029053, "learning_rate": 8.673548126292957e-06, "loss": 1.3938, "step": 12189 }, { "epoch": 0.9553291536050157, "grad_norm": 2.259018898010254, "learning_rate": 8.673338932121274e-06, "loss": 1.4757, "step": 12190 }, { "epoch": 0.9554075235109718, "grad_norm": 2.34858775138855, "learning_rate": 8.673129723978096e-06, "loss": 1.355, "step": 12191 }, { "epoch": 0.9554858934169279, "grad_norm": 6.1812543869018555, "learning_rate": 8.672920501864219e-06, "loss": 0.6508, "step": 12192 }, { "epoch": 0.955564263322884, "grad_norm": 3.216550588607788, "learning_rate": 8.672711265780439e-06, "loss": 1.9562, "step": 12193 }, { "epoch": 0.9556426332288401, "grad_norm": 2.736825942993164, "learning_rate": 8.672502015727551e-06, "loss": 1.5168, "step": 12194 }, { "epoch": 0.9557210031347962, "grad_norm": 6.71078634262085, "learning_rate": 8.672292751706353e-06, "loss": 1.2815, "step": 12195 }, { "epoch": 0.9557993730407524, "grad_norm": 2.5851571559906006, "learning_rate": 8.67208347371764e-06, "loss": 1.4991, "step": 12196 }, { "epoch": 0.9558777429467085, "grad_norm": 2.8747479915618896, "learning_rate": 8.671874181762206e-06, "loss": 1.2554, "step": 12197 }, { "epoch": 0.9559561128526646, "grad_norm": 2.170532703399658, "learning_rate": 8.671664875840851e-06, "loss": 1.3982, "step": 12198 }, { "epoch": 0.9560344827586207, "grad_norm": 3.9356534481048584, "learning_rate": 8.671455555954366e-06, "loss": 1.5574, "step": 12199 }, { "epoch": 0.9561128526645768, "grad_norm": 5.134881019592285, "learning_rate": 8.671246222103552e-06, "loss": 1.0045, "step": 12200 }, { "epoch": 0.9561912225705329, "grad_norm": 2.2511799335479736, "learning_rate": 8.671036874289201e-06, "loss": 1.0621, "step": 12201 }, { "epoch": 0.956269592476489, "grad_norm": 4.72021484375, "learning_rate": 8.670827512512112e-06, "loss": 1.2131, "step": 12202 }, { "epoch": 0.9563479623824451, "grad_norm": 3.31168270111084, "learning_rate": 8.670618136773082e-06, "loss": 1.7529, "step": 12203 }, { "epoch": 0.9564263322884012, "grad_norm": 2.1122853755950928, "learning_rate": 8.670408747072903e-06, "loss": 1.393, "step": 12204 }, { "epoch": 0.9565047021943573, "grad_norm": 13.165884971618652, "learning_rate": 8.670199343412377e-06, "loss": 2.4261, "step": 12205 }, { "epoch": 0.9565830721003135, "grad_norm": 3.0243942737579346, "learning_rate": 8.669989925792297e-06, "loss": 1.2846, "step": 12206 }, { "epoch": 0.9566614420062696, "grad_norm": 6.1249165534973145, "learning_rate": 8.66978049421346e-06, "loss": 1.6165, "step": 12207 }, { "epoch": 0.9567398119122257, "grad_norm": 4.0450758934021, "learning_rate": 8.669571048676664e-06, "loss": 1.4588, "step": 12208 }, { "epoch": 0.9568181818181818, "grad_norm": 3.3158421516418457, "learning_rate": 8.669361589182703e-06, "loss": 1.5772, "step": 12209 }, { "epoch": 0.9568965517241379, "grad_norm": 4.994800567626953, "learning_rate": 8.669152115732375e-06, "loss": 1.3818, "step": 12210 }, { "epoch": 0.956974921630094, "grad_norm": 3.106916904449463, "learning_rate": 8.66894262832648e-06, "loss": 0.8144, "step": 12211 }, { "epoch": 0.9570532915360501, "grad_norm": 2.308349609375, "learning_rate": 8.66873312696581e-06, "loss": 1.3692, "step": 12212 }, { "epoch": 0.9571316614420062, "grad_norm": 2.9639346599578857, "learning_rate": 8.668523611651161e-06, "loss": 1.0109, "step": 12213 }, { "epoch": 0.9572100313479623, "grad_norm": 3.0846245288848877, "learning_rate": 8.668314082383335e-06, "loss": 1.4671, "step": 12214 }, { "epoch": 0.9572884012539185, "grad_norm": 2.5726988315582275, "learning_rate": 8.668104539163127e-06, "loss": 1.8017, "step": 12215 }, { "epoch": 0.9573667711598746, "grad_norm": 3.362922430038452, "learning_rate": 8.667894981991331e-06, "loss": 2.1607, "step": 12216 }, { "epoch": 0.9574451410658307, "grad_norm": 6.797001361846924, "learning_rate": 8.66768541086875e-06, "loss": 1.7018, "step": 12217 }, { "epoch": 0.9575235109717868, "grad_norm": 3.0821428298950195, "learning_rate": 8.667475825796174e-06, "loss": 1.4528, "step": 12218 }, { "epoch": 0.9576018808777429, "grad_norm": 5.595123291015625, "learning_rate": 8.667266226774408e-06, "loss": 1.2869, "step": 12219 }, { "epoch": 0.957680250783699, "grad_norm": 4.65902042388916, "learning_rate": 8.66705661380424e-06, "loss": 1.1014, "step": 12220 }, { "epoch": 0.9577586206896552, "grad_norm": 3.2199316024780273, "learning_rate": 8.666846986886475e-06, "loss": 1.8503, "step": 12221 }, { "epoch": 0.9578369905956113, "grad_norm": 3.291405439376831, "learning_rate": 8.666637346021906e-06, "loss": 1.5332, "step": 12222 }, { "epoch": 0.9579153605015674, "grad_norm": 2.821934700012207, "learning_rate": 8.666427691211333e-06, "loss": 1.3799, "step": 12223 }, { "epoch": 0.9579937304075236, "grad_norm": 3.380357027053833, "learning_rate": 8.666218022455551e-06, "loss": 1.7166, "step": 12224 }, { "epoch": 0.9580721003134797, "grad_norm": 3.323467969894409, "learning_rate": 8.66600833975536e-06, "loss": 1.8198, "step": 12225 }, { "epoch": 0.9581504702194358, "grad_norm": 2.8128702640533447, "learning_rate": 8.665798643111556e-06, "loss": 1.6082, "step": 12226 }, { "epoch": 0.9582288401253919, "grad_norm": 4.645358562469482, "learning_rate": 8.665588932524936e-06, "loss": 1.6305, "step": 12227 }, { "epoch": 0.958307210031348, "grad_norm": 2.264899492263794, "learning_rate": 8.665379207996299e-06, "loss": 1.3962, "step": 12228 }, { "epoch": 0.9583855799373041, "grad_norm": 2.5921545028686523, "learning_rate": 8.665169469526443e-06, "loss": 1.1937, "step": 12229 }, { "epoch": 0.9584639498432602, "grad_norm": 2.253896713256836, "learning_rate": 8.664959717116164e-06, "loss": 1.1812, "step": 12230 }, { "epoch": 0.9585423197492163, "grad_norm": 3.220560312271118, "learning_rate": 8.66474995076626e-06, "loss": 1.6469, "step": 12231 }, { "epoch": 0.9586206896551724, "grad_norm": 5.1500563621521, "learning_rate": 8.664540170477531e-06, "loss": 1.1104, "step": 12232 }, { "epoch": 0.9586990595611286, "grad_norm": 2.3148040771484375, "learning_rate": 8.664330376250773e-06, "loss": 1.3766, "step": 12233 }, { "epoch": 0.9587774294670847, "grad_norm": 3.0001182556152344, "learning_rate": 8.664120568086783e-06, "loss": 1.516, "step": 12234 }, { "epoch": 0.9588557993730408, "grad_norm": 3.934734344482422, "learning_rate": 8.663910745986361e-06, "loss": 1.7601, "step": 12235 }, { "epoch": 0.9589341692789969, "grad_norm": 2.5039827823638916, "learning_rate": 8.663700909950305e-06, "loss": 1.3767, "step": 12236 }, { "epoch": 0.959012539184953, "grad_norm": 3.394637107849121, "learning_rate": 8.663491059979412e-06, "loss": 1.3791, "step": 12237 }, { "epoch": 0.9590909090909091, "grad_norm": 2.825472116470337, "learning_rate": 8.663281196074482e-06, "loss": 0.9922, "step": 12238 }, { "epoch": 0.9591692789968652, "grad_norm": 2.528433084487915, "learning_rate": 8.663071318236312e-06, "loss": 1.5263, "step": 12239 }, { "epoch": 0.9592476489028213, "grad_norm": 5.9035468101501465, "learning_rate": 8.662861426465698e-06, "loss": 1.4988, "step": 12240 }, { "epoch": 0.9593260188087774, "grad_norm": 3.0047836303710938, "learning_rate": 8.662651520763444e-06, "loss": 1.7407, "step": 12241 }, { "epoch": 0.9594043887147335, "grad_norm": 4.533620834350586, "learning_rate": 8.66244160113034e-06, "loss": 1.5243, "step": 12242 }, { "epoch": 0.9594827586206897, "grad_norm": 3.2031941413879395, "learning_rate": 8.662231667567194e-06, "loss": 1.3733, "step": 12243 }, { "epoch": 0.9595611285266458, "grad_norm": 4.900393962860107, "learning_rate": 8.662021720074798e-06, "loss": 1.2565, "step": 12244 }, { "epoch": 0.9596394984326019, "grad_norm": 3.4888343811035156, "learning_rate": 8.66181175865395e-06, "loss": 1.3267, "step": 12245 }, { "epoch": 0.959717868338558, "grad_norm": 2.238548994064331, "learning_rate": 8.661601783305454e-06, "loss": 1.6173, "step": 12246 }, { "epoch": 0.9597962382445141, "grad_norm": 3.3832356929779053, "learning_rate": 8.661391794030105e-06, "loss": 1.2491, "step": 12247 }, { "epoch": 0.9598746081504702, "grad_norm": 3.1886820793151855, "learning_rate": 8.661181790828702e-06, "loss": 1.2416, "step": 12248 }, { "epoch": 0.9599529780564263, "grad_norm": 4.356995582580566, "learning_rate": 8.660971773702044e-06, "loss": 1.4404, "step": 12249 }, { "epoch": 0.9600313479623824, "grad_norm": 2.8883941173553467, "learning_rate": 8.660761742650928e-06, "loss": 1.8073, "step": 12250 }, { "epoch": 0.9601097178683385, "grad_norm": 2.5480329990386963, "learning_rate": 8.660551697676157e-06, "loss": 1.8535, "step": 12251 }, { "epoch": 0.9601880877742947, "grad_norm": 2.305708408355713, "learning_rate": 8.660341638778526e-06, "loss": 1.6182, "step": 12252 }, { "epoch": 0.9602664576802508, "grad_norm": 8.088750839233398, "learning_rate": 8.660131565958837e-06, "loss": 0.9544, "step": 12253 }, { "epoch": 0.9603448275862069, "grad_norm": 4.376911640167236, "learning_rate": 8.659921479217886e-06, "loss": 1.6123, "step": 12254 }, { "epoch": 0.960423197492163, "grad_norm": 3.0693249702453613, "learning_rate": 8.659711378556473e-06, "loss": 1.3469, "step": 12255 }, { "epoch": 0.9605015673981191, "grad_norm": 3.5296144485473633, "learning_rate": 8.659501263975399e-06, "loss": 1.9855, "step": 12256 }, { "epoch": 0.9605799373040752, "grad_norm": 3.569321393966675, "learning_rate": 8.65929113547546e-06, "loss": 1.428, "step": 12257 }, { "epoch": 0.9606583072100313, "grad_norm": 2.686945676803589, "learning_rate": 8.65908099305746e-06, "loss": 1.4491, "step": 12258 }, { "epoch": 0.9607366771159874, "grad_norm": 2.9513416290283203, "learning_rate": 8.658870836722194e-06, "loss": 1.3646, "step": 12259 }, { "epoch": 0.9608150470219435, "grad_norm": 5.154874801635742, "learning_rate": 8.658660666470462e-06, "loss": 1.3603, "step": 12260 }, { "epoch": 0.9608934169278996, "grad_norm": 2.6039838790893555, "learning_rate": 8.658450482303064e-06, "loss": 1.4344, "step": 12261 }, { "epoch": 0.9609717868338558, "grad_norm": 5.226714611053467, "learning_rate": 8.6582402842208e-06, "loss": 1.0766, "step": 12262 }, { "epoch": 0.9610501567398119, "grad_norm": 2.1901230812072754, "learning_rate": 8.65803007222447e-06, "loss": 1.5638, "step": 12263 }, { "epoch": 0.961128526645768, "grad_norm": 2.539990186691284, "learning_rate": 8.65781984631487e-06, "loss": 1.2076, "step": 12264 }, { "epoch": 0.9612068965517241, "grad_norm": 3.5960254669189453, "learning_rate": 8.657609606492805e-06, "loss": 1.4074, "step": 12265 }, { "epoch": 0.9612852664576802, "grad_norm": 2.60624361038208, "learning_rate": 8.65739935275907e-06, "loss": 1.0884, "step": 12266 }, { "epoch": 0.9613636363636363, "grad_norm": 2.3518762588500977, "learning_rate": 8.657189085114465e-06, "loss": 1.3728, "step": 12267 }, { "epoch": 0.9614420062695925, "grad_norm": 5.319581031799316, "learning_rate": 8.656978803559794e-06, "loss": 1.2894, "step": 12268 }, { "epoch": 0.9615203761755486, "grad_norm": 2.3867268562316895, "learning_rate": 8.656768508095853e-06, "loss": 1.3563, "step": 12269 }, { "epoch": 0.9615987460815048, "grad_norm": 2.4504926204681396, "learning_rate": 8.656558198723443e-06, "loss": 1.609, "step": 12270 }, { "epoch": 0.9616771159874609, "grad_norm": 2.273531436920166, "learning_rate": 8.656347875443363e-06, "loss": 1.442, "step": 12271 }, { "epoch": 0.961755485893417, "grad_norm": 3.1959969997406006, "learning_rate": 8.656137538256415e-06, "loss": 1.3716, "step": 12272 }, { "epoch": 0.9618338557993731, "grad_norm": 2.9377832412719727, "learning_rate": 8.655927187163397e-06, "loss": 1.4952, "step": 12273 }, { "epoch": 0.9619122257053292, "grad_norm": 2.571951150894165, "learning_rate": 8.655716822165108e-06, "loss": 1.3738, "step": 12274 }, { "epoch": 0.9619905956112853, "grad_norm": 2.8769805431365967, "learning_rate": 8.655506443262353e-06, "loss": 1.3606, "step": 12275 }, { "epoch": 0.9620689655172414, "grad_norm": 8.957640647888184, "learning_rate": 8.655296050455927e-06, "loss": 1.1786, "step": 12276 }, { "epoch": 0.9621473354231975, "grad_norm": 3.5775678157806396, "learning_rate": 8.655085643746632e-06, "loss": 1.633, "step": 12277 }, { "epoch": 0.9622257053291536, "grad_norm": 2.6821069717407227, "learning_rate": 8.65487522313527e-06, "loss": 1.3365, "step": 12278 }, { "epoch": 0.9623040752351097, "grad_norm": 3.908259391784668, "learning_rate": 8.65466478862264e-06, "loss": 1.3957, "step": 12279 }, { "epoch": 0.9623824451410659, "grad_norm": 6.482378005981445, "learning_rate": 8.654454340209542e-06, "loss": 1.2457, "step": 12280 }, { "epoch": 0.962460815047022, "grad_norm": 6.202136516571045, "learning_rate": 8.654243877896775e-06, "loss": 0.8199, "step": 12281 }, { "epoch": 0.9625391849529781, "grad_norm": 2.4150390625, "learning_rate": 8.654033401685144e-06, "loss": 1.2843, "step": 12282 }, { "epoch": 0.9626175548589342, "grad_norm": 2.0025198459625244, "learning_rate": 8.653822911575445e-06, "loss": 1.3486, "step": 12283 }, { "epoch": 0.9626959247648903, "grad_norm": 4.625406742095947, "learning_rate": 8.65361240756848e-06, "loss": 0.9537, "step": 12284 }, { "epoch": 0.9627742946708464, "grad_norm": 2.04288649559021, "learning_rate": 8.653401889665053e-06, "loss": 1.4097, "step": 12285 }, { "epoch": 0.9628526645768025, "grad_norm": 2.6628448963165283, "learning_rate": 8.653191357865958e-06, "loss": 1.5691, "step": 12286 }, { "epoch": 0.9629310344827586, "grad_norm": 2.7894785404205322, "learning_rate": 8.652980812172e-06, "loss": 1.482, "step": 12287 }, { "epoch": 0.9630094043887147, "grad_norm": 2.9525868892669678, "learning_rate": 8.652770252583982e-06, "loss": 1.1841, "step": 12288 }, { "epoch": 0.9630877742946709, "grad_norm": 4.725566864013672, "learning_rate": 8.6525596791027e-06, "loss": 1.3604, "step": 12289 }, { "epoch": 0.963166144200627, "grad_norm": 2.6493730545043945, "learning_rate": 8.652349091728958e-06, "loss": 1.2225, "step": 12290 }, { "epoch": 0.9632445141065831, "grad_norm": 3.100008726119995, "learning_rate": 8.652138490463556e-06, "loss": 1.3835, "step": 12291 }, { "epoch": 0.9633228840125392, "grad_norm": 2.397158622741699, "learning_rate": 8.651927875307293e-06, "loss": 1.2318, "step": 12292 }, { "epoch": 0.9634012539184953, "grad_norm": 5.170130729675293, "learning_rate": 8.651717246260974e-06, "loss": 0.7171, "step": 12293 }, { "epoch": 0.9634796238244514, "grad_norm": 2.429281234741211, "learning_rate": 8.651506603325397e-06, "loss": 1.4242, "step": 12294 }, { "epoch": 0.9635579937304075, "grad_norm": 3.165560483932495, "learning_rate": 8.651295946501366e-06, "loss": 1.4468, "step": 12295 }, { "epoch": 0.9636363636363636, "grad_norm": 4.22146463394165, "learning_rate": 8.651085275789679e-06, "loss": 1.4806, "step": 12296 }, { "epoch": 0.9637147335423197, "grad_norm": 2.6241724491119385, "learning_rate": 8.650874591191141e-06, "loss": 1.5182, "step": 12297 }, { "epoch": 0.9637931034482758, "grad_norm": 2.936563014984131, "learning_rate": 8.65066389270655e-06, "loss": 1.5158, "step": 12298 }, { "epoch": 0.963871473354232, "grad_norm": 2.784956216812134, "learning_rate": 8.650453180336709e-06, "loss": 1.2279, "step": 12299 }, { "epoch": 0.9639498432601881, "grad_norm": 2.157975912094116, "learning_rate": 8.650242454082419e-06, "loss": 1.1082, "step": 12300 }, { "epoch": 0.9640282131661442, "grad_norm": 8.83572769165039, "learning_rate": 8.650031713944481e-06, "loss": 1.3879, "step": 12301 }, { "epoch": 0.9641065830721003, "grad_norm": 2.896263837814331, "learning_rate": 8.649820959923699e-06, "loss": 1.3296, "step": 12302 }, { "epoch": 0.9641849529780564, "grad_norm": 3.5183374881744385, "learning_rate": 8.64961019202087e-06, "loss": 1.2826, "step": 12303 }, { "epoch": 0.9642633228840125, "grad_norm": 4.75965690612793, "learning_rate": 8.649399410236799e-06, "loss": 1.7103, "step": 12304 }, { "epoch": 0.9643416927899686, "grad_norm": 4.338253021240234, "learning_rate": 8.649188614572287e-06, "loss": 1.9742, "step": 12305 }, { "epoch": 0.9644200626959247, "grad_norm": 2.477519989013672, "learning_rate": 8.648977805028138e-06, "loss": 1.3727, "step": 12306 }, { "epoch": 0.9644984326018808, "grad_norm": 2.4527368545532227, "learning_rate": 8.64876698160515e-06, "loss": 1.2432, "step": 12307 }, { "epoch": 0.964576802507837, "grad_norm": 2.495857000350952, "learning_rate": 8.648556144304128e-06, "loss": 1.0517, "step": 12308 }, { "epoch": 0.9646551724137931, "grad_norm": 3.6988818645477295, "learning_rate": 8.64834529312587e-06, "loss": 1.6588, "step": 12309 }, { "epoch": 0.9647335423197492, "grad_norm": 3.1792826652526855, "learning_rate": 8.648134428071182e-06, "loss": 1.032, "step": 12310 }, { "epoch": 0.9648119122257053, "grad_norm": 2.0044305324554443, "learning_rate": 8.647923549140864e-06, "loss": 0.9918, "step": 12311 }, { "epoch": 0.9648902821316614, "grad_norm": 5.267408847808838, "learning_rate": 8.647712656335719e-06, "loss": 2.0488, "step": 12312 }, { "epoch": 0.9649686520376175, "grad_norm": 2.2931764125823975, "learning_rate": 8.647501749656548e-06, "loss": 0.9791, "step": 12313 }, { "epoch": 0.9650470219435736, "grad_norm": 4.424292087554932, "learning_rate": 8.647290829104155e-06, "loss": 1.3068, "step": 12314 }, { "epoch": 0.9651253918495298, "grad_norm": 6.515335559844971, "learning_rate": 8.64707989467934e-06, "loss": 0.8313, "step": 12315 }, { "epoch": 0.965203761755486, "grad_norm": 2.3041186332702637, "learning_rate": 8.646868946382906e-06, "loss": 1.5115, "step": 12316 }, { "epoch": 0.965282131661442, "grad_norm": 3.5574586391448975, "learning_rate": 8.646657984215656e-06, "loss": 1.4292, "step": 12317 }, { "epoch": 0.9653605015673982, "grad_norm": 3.945939540863037, "learning_rate": 8.646447008178392e-06, "loss": 1.9533, "step": 12318 }, { "epoch": 0.9654388714733543, "grad_norm": 2.2436351776123047, "learning_rate": 8.646236018271917e-06, "loss": 1.3697, "step": 12319 }, { "epoch": 0.9655172413793104, "grad_norm": 2.9805638790130615, "learning_rate": 8.646025014497033e-06, "loss": 1.5588, "step": 12320 }, { "epoch": 0.9655956112852665, "grad_norm": 3.393036127090454, "learning_rate": 8.645813996854543e-06, "loss": 1.4262, "step": 12321 }, { "epoch": 0.9656739811912226, "grad_norm": 10.191813468933105, "learning_rate": 8.645602965345248e-06, "loss": 1.9996, "step": 12322 }, { "epoch": 0.9657523510971787, "grad_norm": 3.8849005699157715, "learning_rate": 8.645391919969952e-06, "loss": 1.6412, "step": 12323 }, { "epoch": 0.9658307210031348, "grad_norm": 3.758561849594116, "learning_rate": 8.645180860729458e-06, "loss": 1.4257, "step": 12324 }, { "epoch": 0.9659090909090909, "grad_norm": 4.848199844360352, "learning_rate": 8.644969787624568e-06, "loss": 1.6461, "step": 12325 }, { "epoch": 0.965987460815047, "grad_norm": 5.525611877441406, "learning_rate": 8.644758700656087e-06, "loss": 1.2655, "step": 12326 }, { "epoch": 0.9660658307210032, "grad_norm": 2.953315258026123, "learning_rate": 8.644547599824813e-06, "loss": 1.1035, "step": 12327 }, { "epoch": 0.9661442006269593, "grad_norm": 2.422664165496826, "learning_rate": 8.644336485131552e-06, "loss": 1.0637, "step": 12328 }, { "epoch": 0.9662225705329154, "grad_norm": 2.0179057121276855, "learning_rate": 8.644125356577109e-06, "loss": 1.0689, "step": 12329 }, { "epoch": 0.9663009404388715, "grad_norm": 2.41184663772583, "learning_rate": 8.643914214162283e-06, "loss": 1.3256, "step": 12330 }, { "epoch": 0.9663793103448276, "grad_norm": 4.913356304168701, "learning_rate": 8.64370305788788e-06, "loss": 1.4223, "step": 12331 }, { "epoch": 0.9664576802507837, "grad_norm": 2.625123977661133, "learning_rate": 8.643491887754703e-06, "loss": 1.4661, "step": 12332 }, { "epoch": 0.9665360501567398, "grad_norm": 4.374452590942383, "learning_rate": 8.643280703763553e-06, "loss": 1.4098, "step": 12333 }, { "epoch": 0.9666144200626959, "grad_norm": 2.3474674224853516, "learning_rate": 8.643069505915234e-06, "loss": 1.3739, "step": 12334 }, { "epoch": 0.966692789968652, "grad_norm": 10.573722839355469, "learning_rate": 8.64285829421055e-06, "loss": 1.8089, "step": 12335 }, { "epoch": 0.9667711598746082, "grad_norm": 2.862314462661743, "learning_rate": 8.642647068650305e-06, "loss": 1.249, "step": 12336 }, { "epoch": 0.9668495297805643, "grad_norm": 8.91076374053955, "learning_rate": 8.6424358292353e-06, "loss": 1.7659, "step": 12337 }, { "epoch": 0.9669278996865204, "grad_norm": 1.950711965560913, "learning_rate": 8.642224575966343e-06, "loss": 1.3237, "step": 12338 }, { "epoch": 0.9670062695924765, "grad_norm": 4.281718730926514, "learning_rate": 8.642013308844232e-06, "loss": 1.5333, "step": 12339 }, { "epoch": 0.9670846394984326, "grad_norm": 9.451624870300293, "learning_rate": 8.641802027869774e-06, "loss": 2.0697, "step": 12340 }, { "epoch": 0.9671630094043887, "grad_norm": 3.1271564960479736, "learning_rate": 8.641590733043773e-06, "loss": 1.303, "step": 12341 }, { "epoch": 0.9672413793103448, "grad_norm": 3.7465195655822754, "learning_rate": 8.641379424367028e-06, "loss": 1.3932, "step": 12342 }, { "epoch": 0.9673197492163009, "grad_norm": 2.7170655727386475, "learning_rate": 8.641168101840347e-06, "loss": 1.6866, "step": 12343 }, { "epoch": 0.967398119122257, "grad_norm": 2.908971071243286, "learning_rate": 8.640956765464534e-06, "loss": 1.2693, "step": 12344 }, { "epoch": 0.9674764890282131, "grad_norm": 2.740875005722046, "learning_rate": 8.64074541524039e-06, "loss": 1.5775, "step": 12345 }, { "epoch": 0.9675548589341693, "grad_norm": 2.9779765605926514, "learning_rate": 8.640534051168723e-06, "loss": 1.7832, "step": 12346 }, { "epoch": 0.9676332288401254, "grad_norm": 3.0585169792175293, "learning_rate": 8.640322673250333e-06, "loss": 1.6319, "step": 12347 }, { "epoch": 0.9677115987460815, "grad_norm": 10.039931297302246, "learning_rate": 8.640111281486024e-06, "loss": 0.85, "step": 12348 }, { "epoch": 0.9677899686520376, "grad_norm": 2.2262771129608154, "learning_rate": 8.639899875876602e-06, "loss": 1.4306, "step": 12349 }, { "epoch": 0.9678683385579937, "grad_norm": 2.065891981124878, "learning_rate": 8.63968845642287e-06, "loss": 1.2081, "step": 12350 }, { "epoch": 0.9679467084639498, "grad_norm": 2.5999505519866943, "learning_rate": 8.639477023125634e-06, "loss": 1.4408, "step": 12351 }, { "epoch": 0.9680250783699059, "grad_norm": 3.643174648284912, "learning_rate": 8.639265575985695e-06, "loss": 1.6329, "step": 12352 }, { "epoch": 0.968103448275862, "grad_norm": 3.2937331199645996, "learning_rate": 8.639054115003861e-06, "loss": 1.3476, "step": 12353 }, { "epoch": 0.9681818181818181, "grad_norm": 2.1808300018310547, "learning_rate": 8.638842640180935e-06, "loss": 0.9468, "step": 12354 }, { "epoch": 0.9682601880877743, "grad_norm": 2.676621198654175, "learning_rate": 8.638631151517718e-06, "loss": 1.3847, "step": 12355 }, { "epoch": 0.9683385579937304, "grad_norm": 3.0644824504852295, "learning_rate": 8.638419649015019e-06, "loss": 1.4425, "step": 12356 }, { "epoch": 0.9684169278996865, "grad_norm": 2.764378786087036, "learning_rate": 8.638208132673639e-06, "loss": 1.4388, "step": 12357 }, { "epoch": 0.9684952978056426, "grad_norm": 2.6274092197418213, "learning_rate": 8.637996602494385e-06, "loss": 1.2863, "step": 12358 }, { "epoch": 0.9685736677115987, "grad_norm": 3.023855686187744, "learning_rate": 8.637785058478058e-06, "loss": 2.0115, "step": 12359 }, { "epoch": 0.9686520376175548, "grad_norm": 2.559831380844116, "learning_rate": 8.637573500625469e-06, "loss": 1.4583, "step": 12360 }, { "epoch": 0.9687304075235109, "grad_norm": 2.6337087154388428, "learning_rate": 8.637361928937416e-06, "loss": 1.7486, "step": 12361 }, { "epoch": 0.9688087774294671, "grad_norm": 2.551591157913208, "learning_rate": 8.637150343414708e-06, "loss": 1.259, "step": 12362 }, { "epoch": 0.9688871473354232, "grad_norm": 2.578996419906616, "learning_rate": 8.636938744058148e-06, "loss": 1.6571, "step": 12363 }, { "epoch": 0.9689655172413794, "grad_norm": 2.600879192352295, "learning_rate": 8.63672713086854e-06, "loss": 1.5267, "step": 12364 }, { "epoch": 0.9690438871473355, "grad_norm": 2.4412479400634766, "learning_rate": 8.636515503846692e-06, "loss": 1.2159, "step": 12365 }, { "epoch": 0.9691222570532916, "grad_norm": 1.9567569494247437, "learning_rate": 8.636303862993405e-06, "loss": 1.0407, "step": 12366 }, { "epoch": 0.9692006269592477, "grad_norm": 2.788698434829712, "learning_rate": 8.636092208309486e-06, "loss": 1.1757, "step": 12367 }, { "epoch": 0.9692789968652038, "grad_norm": 3.3269197940826416, "learning_rate": 8.63588053979574e-06, "loss": 1.314, "step": 12368 }, { "epoch": 0.9693573667711599, "grad_norm": 4.945663928985596, "learning_rate": 8.635668857452972e-06, "loss": 0.7473, "step": 12369 }, { "epoch": 0.969435736677116, "grad_norm": 2.3676068782806396, "learning_rate": 8.635457161281988e-06, "loss": 1.3183, "step": 12370 }, { "epoch": 0.9695141065830721, "grad_norm": 2.612140417098999, "learning_rate": 8.635245451283592e-06, "loss": 1.477, "step": 12371 }, { "epoch": 0.9695924764890282, "grad_norm": 7.574486255645752, "learning_rate": 8.635033727458588e-06, "loss": 1.4651, "step": 12372 }, { "epoch": 0.9696708463949844, "grad_norm": 2.316480875015259, "learning_rate": 8.634821989807784e-06, "loss": 1.0181, "step": 12373 }, { "epoch": 0.9697492163009405, "grad_norm": 2.0875728130340576, "learning_rate": 8.634610238331984e-06, "loss": 1.3461, "step": 12374 }, { "epoch": 0.9698275862068966, "grad_norm": 7.285407543182373, "learning_rate": 8.634398473031993e-06, "loss": 1.4373, "step": 12375 }, { "epoch": 0.9699059561128527, "grad_norm": 2.59312105178833, "learning_rate": 8.634186693908618e-06, "loss": 1.6425, "step": 12376 }, { "epoch": 0.9699843260188088, "grad_norm": 2.065006971359253, "learning_rate": 8.633974900962662e-06, "loss": 1.2709, "step": 12377 }, { "epoch": 0.9700626959247649, "grad_norm": 3.3660216331481934, "learning_rate": 8.633763094194933e-06, "loss": 1.3113, "step": 12378 }, { "epoch": 0.970141065830721, "grad_norm": 2.762061595916748, "learning_rate": 8.633551273606235e-06, "loss": 1.4915, "step": 12379 }, { "epoch": 0.9702194357366771, "grad_norm": 3.566056251525879, "learning_rate": 8.633339439197375e-06, "loss": 1.2101, "step": 12380 }, { "epoch": 0.9702978056426332, "grad_norm": 2.8652515411376953, "learning_rate": 8.633127590969159e-06, "loss": 1.2941, "step": 12381 }, { "epoch": 0.9703761755485893, "grad_norm": 3.8452048301696777, "learning_rate": 8.63291572892239e-06, "loss": 1.266, "step": 12382 }, { "epoch": 0.9704545454545455, "grad_norm": 6.238363265991211, "learning_rate": 8.632703853057875e-06, "loss": 1.8827, "step": 12383 }, { "epoch": 0.9705329153605016, "grad_norm": 2.4624876976013184, "learning_rate": 8.632491963376422e-06, "loss": 1.2406, "step": 12384 }, { "epoch": 0.9706112852664577, "grad_norm": 3.5725557804107666, "learning_rate": 8.632280059878835e-06, "loss": 1.2578, "step": 12385 }, { "epoch": 0.9706896551724138, "grad_norm": 2.7671055793762207, "learning_rate": 8.632068142565919e-06, "loss": 1.7344, "step": 12386 }, { "epoch": 0.9707680250783699, "grad_norm": 2.6568992137908936, "learning_rate": 8.631856211438483e-06, "loss": 1.4038, "step": 12387 }, { "epoch": 0.970846394984326, "grad_norm": 1.816171407699585, "learning_rate": 8.63164426649733e-06, "loss": 1.3726, "step": 12388 }, { "epoch": 0.9709247648902821, "grad_norm": 2.655855655670166, "learning_rate": 8.631432307743268e-06, "loss": 1.457, "step": 12389 }, { "epoch": 0.9710031347962382, "grad_norm": 2.530871868133545, "learning_rate": 8.631220335177105e-06, "loss": 1.2239, "step": 12390 }, { "epoch": 0.9710815047021943, "grad_norm": 2.3026225566864014, "learning_rate": 8.631008348799641e-06, "loss": 1.5659, "step": 12391 }, { "epoch": 0.9711598746081505, "grad_norm": 2.404573917388916, "learning_rate": 8.63079634861169e-06, "loss": 1.4874, "step": 12392 }, { "epoch": 0.9712382445141066, "grad_norm": 2.787179708480835, "learning_rate": 8.630584334614052e-06, "loss": 1.2871, "step": 12393 }, { "epoch": 0.9713166144200627, "grad_norm": 2.145784854888916, "learning_rate": 8.630372306807535e-06, "loss": 1.3157, "step": 12394 }, { "epoch": 0.9713949843260188, "grad_norm": 12.663481712341309, "learning_rate": 8.63016026519295e-06, "loss": 0.8467, "step": 12395 }, { "epoch": 0.9714733542319749, "grad_norm": 6.211315155029297, "learning_rate": 8.629948209771097e-06, "loss": 1.06, "step": 12396 }, { "epoch": 0.971551724137931, "grad_norm": 2.1033856868743896, "learning_rate": 8.629736140542786e-06, "loss": 1.3637, "step": 12397 }, { "epoch": 0.9716300940438871, "grad_norm": 2.549452543258667, "learning_rate": 8.629524057508824e-06, "loss": 0.9462, "step": 12398 }, { "epoch": 0.9717084639498432, "grad_norm": 2.6654748916625977, "learning_rate": 8.629311960670017e-06, "loss": 1.4101, "step": 12399 }, { "epoch": 0.9717868338557993, "grad_norm": 3.36511492729187, "learning_rate": 8.629099850027172e-06, "loss": 1.644, "step": 12400 }, { "epoch": 0.9718652037617554, "grad_norm": 2.831817150115967, "learning_rate": 8.628887725581092e-06, "loss": 1.4038, "step": 12401 }, { "epoch": 0.9719435736677116, "grad_norm": 5.204043388366699, "learning_rate": 8.62867558733259e-06, "loss": 1.0174, "step": 12402 }, { "epoch": 0.9720219435736677, "grad_norm": 2.277784585952759, "learning_rate": 8.628463435282469e-06, "loss": 0.9993, "step": 12403 }, { "epoch": 0.9721003134796238, "grad_norm": 1.977579951286316, "learning_rate": 8.628251269431536e-06, "loss": 1.3609, "step": 12404 }, { "epoch": 0.9721786833855799, "grad_norm": 1.9459888935089111, "learning_rate": 8.6280390897806e-06, "loss": 1.4054, "step": 12405 }, { "epoch": 0.972257053291536, "grad_norm": 1.7861148118972778, "learning_rate": 8.627826896330466e-06, "loss": 1.3605, "step": 12406 }, { "epoch": 0.9723354231974921, "grad_norm": 3.996267795562744, "learning_rate": 8.627614689081942e-06, "loss": 1.2114, "step": 12407 }, { "epoch": 0.9724137931034482, "grad_norm": 4.664403915405273, "learning_rate": 8.627402468035834e-06, "loss": 1.801, "step": 12408 }, { "epoch": 0.9724921630094044, "grad_norm": 2.393223762512207, "learning_rate": 8.627190233192953e-06, "loss": 1.1714, "step": 12409 }, { "epoch": 0.9725705329153606, "grad_norm": 2.789288282394409, "learning_rate": 8.6269779845541e-06, "loss": 1.2073, "step": 12410 }, { "epoch": 0.9726489028213167, "grad_norm": 8.185050010681152, "learning_rate": 8.626765722120087e-06, "loss": 1.0581, "step": 12411 }, { "epoch": 0.9727272727272728, "grad_norm": 2.1192593574523926, "learning_rate": 8.62655344589172e-06, "loss": 1.6498, "step": 12412 }, { "epoch": 0.9728056426332289, "grad_norm": 3.44097638130188, "learning_rate": 8.626341155869808e-06, "loss": 1.3487, "step": 12413 }, { "epoch": 0.972884012539185, "grad_norm": 3.8046929836273193, "learning_rate": 8.626128852055155e-06, "loss": 1.4428, "step": 12414 }, { "epoch": 0.9729623824451411, "grad_norm": 2.954116106033325, "learning_rate": 8.62591653444857e-06, "loss": 1.504, "step": 12415 }, { "epoch": 0.9730407523510972, "grad_norm": 9.414138793945312, "learning_rate": 8.625704203050864e-06, "loss": 1.3579, "step": 12416 }, { "epoch": 0.9731191222570533, "grad_norm": 4.423169136047363, "learning_rate": 8.625491857862837e-06, "loss": 1.3073, "step": 12417 }, { "epoch": 0.9731974921630094, "grad_norm": 3.65144419670105, "learning_rate": 8.625279498885304e-06, "loss": 1.5458, "step": 12418 }, { "epoch": 0.9732758620689655, "grad_norm": 2.303457736968994, "learning_rate": 8.62506712611907e-06, "loss": 1.3406, "step": 12419 }, { "epoch": 0.9733542319749217, "grad_norm": 4.41561222076416, "learning_rate": 8.624854739564943e-06, "loss": 1.5932, "step": 12420 }, { "epoch": 0.9734326018808778, "grad_norm": 3.8576250076293945, "learning_rate": 8.624642339223728e-06, "loss": 1.3872, "step": 12421 }, { "epoch": 0.9735109717868339, "grad_norm": 3.3087635040283203, "learning_rate": 8.624429925096236e-06, "loss": 1.8258, "step": 12422 }, { "epoch": 0.97358934169279, "grad_norm": 2.28403639793396, "learning_rate": 8.624217497183276e-06, "loss": 1.4322, "step": 12423 }, { "epoch": 0.9736677115987461, "grad_norm": 2.779780626296997, "learning_rate": 8.624005055485653e-06, "loss": 1.4953, "step": 12424 }, { "epoch": 0.9737460815047022, "grad_norm": 25.86878204345703, "learning_rate": 8.623792600004176e-06, "loss": 1.4452, "step": 12425 }, { "epoch": 0.9738244514106583, "grad_norm": 2.2630960941314697, "learning_rate": 8.623580130739652e-06, "loss": 1.5216, "step": 12426 }, { "epoch": 0.9739028213166144, "grad_norm": 2.0365257263183594, "learning_rate": 8.623367647692894e-06, "loss": 1.3902, "step": 12427 }, { "epoch": 0.9739811912225705, "grad_norm": 3.0907905101776123, "learning_rate": 8.623155150864702e-06, "loss": 1.4549, "step": 12428 }, { "epoch": 0.9740595611285267, "grad_norm": 8.345251083374023, "learning_rate": 8.622942640255891e-06, "loss": 1.7287, "step": 12429 }, { "epoch": 0.9741379310344828, "grad_norm": 3.522524356842041, "learning_rate": 8.622730115867268e-06, "loss": 1.287, "step": 12430 }, { "epoch": 0.9742163009404389, "grad_norm": 3.613264560699463, "learning_rate": 8.622517577699639e-06, "loss": 1.4567, "step": 12431 }, { "epoch": 0.974294670846395, "grad_norm": 3.2059431076049805, "learning_rate": 8.622305025753814e-06, "loss": 1.4478, "step": 12432 }, { "epoch": 0.9743730407523511, "grad_norm": 2.8424370288848877, "learning_rate": 8.622092460030602e-06, "loss": 1.4759, "step": 12433 }, { "epoch": 0.9744514106583072, "grad_norm": 2.1535089015960693, "learning_rate": 8.62187988053081e-06, "loss": 1.5685, "step": 12434 }, { "epoch": 0.9745297805642633, "grad_norm": 2.6140315532684326, "learning_rate": 8.621667287255248e-06, "loss": 1.8999, "step": 12435 }, { "epoch": 0.9746081504702194, "grad_norm": 3.1206984519958496, "learning_rate": 8.621454680204722e-06, "loss": 1.4372, "step": 12436 }, { "epoch": 0.9746865203761755, "grad_norm": 3.8424408435821533, "learning_rate": 8.621242059380043e-06, "loss": 1.5441, "step": 12437 }, { "epoch": 0.9747648902821316, "grad_norm": 2.1242311000823975, "learning_rate": 8.62102942478202e-06, "loss": 1.5255, "step": 12438 }, { "epoch": 0.9748432601880878, "grad_norm": 3.074866533279419, "learning_rate": 8.620816776411463e-06, "loss": 1.1062, "step": 12439 }, { "epoch": 0.9749216300940439, "grad_norm": 2.8763604164123535, "learning_rate": 8.620604114269176e-06, "loss": 0.8943, "step": 12440 }, { "epoch": 0.975, "grad_norm": 2.642481803894043, "learning_rate": 8.620391438355973e-06, "loss": 1.3676, "step": 12441 }, { "epoch": 0.9750783699059561, "grad_norm": 2.5294721126556396, "learning_rate": 8.620178748672658e-06, "loss": 1.2827, "step": 12442 }, { "epoch": 0.9751567398119122, "grad_norm": 3.0039095878601074, "learning_rate": 8.619966045220042e-06, "loss": 1.3261, "step": 12443 }, { "epoch": 0.9752351097178683, "grad_norm": 4.5633463859558105, "learning_rate": 8.619753327998936e-06, "loss": 1.3119, "step": 12444 }, { "epoch": 0.9753134796238244, "grad_norm": 2.2949042320251465, "learning_rate": 8.619540597010148e-06, "loss": 1.4098, "step": 12445 }, { "epoch": 0.9753918495297805, "grad_norm": 2.5940632820129395, "learning_rate": 8.619327852254485e-06, "loss": 1.6331, "step": 12446 }, { "epoch": 0.9754702194357366, "grad_norm": 6.468184471130371, "learning_rate": 8.619115093732759e-06, "loss": 1.8453, "step": 12447 }, { "epoch": 0.9755485893416928, "grad_norm": 2.321577310562134, "learning_rate": 8.618902321445778e-06, "loss": 1.3673, "step": 12448 }, { "epoch": 0.9756269592476489, "grad_norm": 2.742340087890625, "learning_rate": 8.618689535394351e-06, "loss": 1.2542, "step": 12449 }, { "epoch": 0.975705329153605, "grad_norm": 2.2345261573791504, "learning_rate": 8.618476735579288e-06, "loss": 1.5195, "step": 12450 }, { "epoch": 0.9757836990595611, "grad_norm": 3.26511287689209, "learning_rate": 8.618263922001398e-06, "loss": 1.1358, "step": 12451 }, { "epoch": 0.9758620689655172, "grad_norm": 5.778514862060547, "learning_rate": 8.61805109466149e-06, "loss": 0.8901, "step": 12452 }, { "epoch": 0.9759404388714733, "grad_norm": 3.253742218017578, "learning_rate": 8.617838253560376e-06, "loss": 1.6146, "step": 12453 }, { "epoch": 0.9760188087774294, "grad_norm": 2.4079067707061768, "learning_rate": 8.617625398698861e-06, "loss": 1.433, "step": 12454 }, { "epoch": 0.9760971786833855, "grad_norm": 2.126429796218872, "learning_rate": 8.61741253007776e-06, "loss": 1.2154, "step": 12455 }, { "epoch": 0.9761755485893417, "grad_norm": 3.8150157928466797, "learning_rate": 8.617199647697877e-06, "loss": 1.6995, "step": 12456 }, { "epoch": 0.9762539184952979, "grad_norm": 2.2729594707489014, "learning_rate": 8.616986751560026e-06, "loss": 1.0015, "step": 12457 }, { "epoch": 0.976332288401254, "grad_norm": 3.0142948627471924, "learning_rate": 8.616773841665015e-06, "loss": 1.0502, "step": 12458 }, { "epoch": 0.9764106583072101, "grad_norm": 2.644789934158325, "learning_rate": 8.616560918013653e-06, "loss": 1.3305, "step": 12459 }, { "epoch": 0.9764890282131662, "grad_norm": 4.469286918640137, "learning_rate": 8.616347980606749e-06, "loss": 1.5138, "step": 12460 }, { "epoch": 0.9765673981191223, "grad_norm": 3.0812389850616455, "learning_rate": 8.616135029445118e-06, "loss": 1.3951, "step": 12461 }, { "epoch": 0.9766457680250784, "grad_norm": 2.3041749000549316, "learning_rate": 8.615922064529567e-06, "loss": 1.211, "step": 12462 }, { "epoch": 0.9767241379310345, "grad_norm": 3.039885997772217, "learning_rate": 8.615709085860903e-06, "loss": 0.6436, "step": 12463 }, { "epoch": 0.9768025078369906, "grad_norm": 2.6965436935424805, "learning_rate": 8.615496093439941e-06, "loss": 1.5333, "step": 12464 }, { "epoch": 0.9768808777429467, "grad_norm": 3.1837339401245117, "learning_rate": 8.61528308726749e-06, "loss": 1.26, "step": 12465 }, { "epoch": 0.9769592476489029, "grad_norm": 4.491218566894531, "learning_rate": 8.615070067344355e-06, "loss": 1.4354, "step": 12466 }, { "epoch": 0.977037617554859, "grad_norm": 4.367273330688477, "learning_rate": 8.614857033671354e-06, "loss": 1.4678, "step": 12467 }, { "epoch": 0.9771159874608151, "grad_norm": 1.9691076278686523, "learning_rate": 8.614643986249292e-06, "loss": 1.1558, "step": 12468 }, { "epoch": 0.9771943573667712, "grad_norm": 3.532594919204712, "learning_rate": 8.61443092507898e-06, "loss": 1.5663, "step": 12469 }, { "epoch": 0.9772727272727273, "grad_norm": 9.885066986083984, "learning_rate": 8.61421785016123e-06, "loss": 0.9743, "step": 12470 }, { "epoch": 0.9773510971786834, "grad_norm": 3.307202100753784, "learning_rate": 8.614004761496853e-06, "loss": 1.2227, "step": 12471 }, { "epoch": 0.9774294670846395, "grad_norm": 3.269443988800049, "learning_rate": 8.613791659086657e-06, "loss": 1.7588, "step": 12472 }, { "epoch": 0.9775078369905956, "grad_norm": 3.1175665855407715, "learning_rate": 8.613578542931453e-06, "loss": 1.5233, "step": 12473 }, { "epoch": 0.9775862068965517, "grad_norm": 3.6000075340270996, "learning_rate": 8.613365413032054e-06, "loss": 1.5466, "step": 12474 }, { "epoch": 0.9776645768025078, "grad_norm": 4.719791889190674, "learning_rate": 8.613152269389269e-06, "loss": 0.8814, "step": 12475 }, { "epoch": 0.977742946708464, "grad_norm": 2.5234196186065674, "learning_rate": 8.612939112003907e-06, "loss": 1.6643, "step": 12476 }, { "epoch": 0.9778213166144201, "grad_norm": 6.05275821685791, "learning_rate": 8.612725940876781e-06, "loss": 1.2199, "step": 12477 }, { "epoch": 0.9778996865203762, "grad_norm": 4.614652633666992, "learning_rate": 8.6125127560087e-06, "loss": 1.1348, "step": 12478 }, { "epoch": 0.9779780564263323, "grad_norm": 5.601801872253418, "learning_rate": 8.612299557400478e-06, "loss": 1.8309, "step": 12479 }, { "epoch": 0.9780564263322884, "grad_norm": 2.5094566345214844, "learning_rate": 8.612086345052923e-06, "loss": 0.6842, "step": 12480 }, { "epoch": 0.9781347962382445, "grad_norm": 3.667151927947998, "learning_rate": 8.611873118966847e-06, "loss": 1.4485, "step": 12481 }, { "epoch": 0.9782131661442006, "grad_norm": 4.004659175872803, "learning_rate": 8.61165987914306e-06, "loss": 1.3748, "step": 12482 }, { "epoch": 0.9782915360501567, "grad_norm": 3.062854766845703, "learning_rate": 8.611446625582374e-06, "loss": 1.4287, "step": 12483 }, { "epoch": 0.9783699059561128, "grad_norm": 3.3520519733428955, "learning_rate": 8.6112333582856e-06, "loss": 1.183, "step": 12484 }, { "epoch": 0.978448275862069, "grad_norm": 2.5938637256622314, "learning_rate": 8.611020077253551e-06, "loss": 1.5473, "step": 12485 }, { "epoch": 0.9785266457680251, "grad_norm": 3.011321544647217, "learning_rate": 8.610806782487035e-06, "loss": 1.3749, "step": 12486 }, { "epoch": 0.9786050156739812, "grad_norm": 7.112464427947998, "learning_rate": 8.610593473986863e-06, "loss": 1.4682, "step": 12487 }, { "epoch": 0.9786833855799373, "grad_norm": 2.6326797008514404, "learning_rate": 8.61038015175385e-06, "loss": 1.4356, "step": 12488 }, { "epoch": 0.9787617554858934, "grad_norm": 3.8883347511291504, "learning_rate": 8.610166815788805e-06, "loss": 1.3642, "step": 12489 }, { "epoch": 0.9788401253918495, "grad_norm": 1.890497088432312, "learning_rate": 8.60995346609254e-06, "loss": 1.349, "step": 12490 }, { "epoch": 0.9789184952978056, "grad_norm": 4.400054454803467, "learning_rate": 8.609740102665866e-06, "loss": 1.3153, "step": 12491 }, { "epoch": 0.9789968652037617, "grad_norm": 2.2295892238616943, "learning_rate": 8.609526725509593e-06, "loss": 1.1936, "step": 12492 }, { "epoch": 0.9790752351097178, "grad_norm": 2.9539735317230225, "learning_rate": 8.609313334624538e-06, "loss": 1.2847, "step": 12493 }, { "epoch": 0.979153605015674, "grad_norm": 2.5852408409118652, "learning_rate": 8.609099930011506e-06, "loss": 1.0908, "step": 12494 }, { "epoch": 0.97923197492163, "grad_norm": 3.916191577911377, "learning_rate": 8.608886511671312e-06, "loss": 1.3487, "step": 12495 }, { "epoch": 0.9793103448275862, "grad_norm": 2.3754312992095947, "learning_rate": 8.608673079604769e-06, "loss": 1.2721, "step": 12496 }, { "epoch": 0.9793887147335423, "grad_norm": 2.749619483947754, "learning_rate": 8.608459633812686e-06, "loss": 1.2827, "step": 12497 }, { "epoch": 0.9794670846394984, "grad_norm": 7.085475444793701, "learning_rate": 8.608246174295876e-06, "loss": 1.369, "step": 12498 }, { "epoch": 0.9795454545454545, "grad_norm": 2.165722370147705, "learning_rate": 8.60803270105515e-06, "loss": 1.136, "step": 12499 }, { "epoch": 0.9796238244514106, "grad_norm": 1.6186435222625732, "learning_rate": 8.607819214091322e-06, "loss": 1.238, "step": 12500 }, { "epoch": 0.9797021943573667, "grad_norm": 4.417695999145508, "learning_rate": 8.607605713405203e-06, "loss": 1.4524, "step": 12501 }, { "epoch": 0.9797805642633229, "grad_norm": 3.731288194656372, "learning_rate": 8.607392198997604e-06, "loss": 1.8332, "step": 12502 }, { "epoch": 0.979858934169279, "grad_norm": 2.0662992000579834, "learning_rate": 8.607178670869336e-06, "loss": 1.2728, "step": 12503 }, { "epoch": 0.9799373040752352, "grad_norm": 6.844398498535156, "learning_rate": 8.606965129021216e-06, "loss": 1.6637, "step": 12504 }, { "epoch": 0.9800156739811913, "grad_norm": 2.3645763397216797, "learning_rate": 8.606751573454053e-06, "loss": 1.3998, "step": 12505 }, { "epoch": 0.9800940438871474, "grad_norm": 2.1438939571380615, "learning_rate": 8.60653800416866e-06, "loss": 1.2519, "step": 12506 }, { "epoch": 0.9801724137931035, "grad_norm": 3.1099753379821777, "learning_rate": 8.606324421165847e-06, "loss": 1.2277, "step": 12507 }, { "epoch": 0.9802507836990596, "grad_norm": 2.2914910316467285, "learning_rate": 8.60611082444643e-06, "loss": 1.43, "step": 12508 }, { "epoch": 0.9803291536050157, "grad_norm": 2.8901524543762207, "learning_rate": 8.60589721401122e-06, "loss": 1.3026, "step": 12509 }, { "epoch": 0.9804075235109718, "grad_norm": 3.359100580215454, "learning_rate": 8.605683589861027e-06, "loss": 1.3701, "step": 12510 }, { "epoch": 0.9804858934169279, "grad_norm": 6.626933574676514, "learning_rate": 8.605469951996666e-06, "loss": 1.1055, "step": 12511 }, { "epoch": 0.980564263322884, "grad_norm": 3.6689178943634033, "learning_rate": 8.60525630041895e-06, "loss": 1.6324, "step": 12512 }, { "epoch": 0.9806426332288402, "grad_norm": 4.663429260253906, "learning_rate": 8.60504263512869e-06, "loss": 1.2459, "step": 12513 }, { "epoch": 0.9807210031347963, "grad_norm": 3.4841718673706055, "learning_rate": 8.604828956126699e-06, "loss": 1.4563, "step": 12514 }, { "epoch": 0.9807993730407524, "grad_norm": 3.434943199157715, "learning_rate": 8.604615263413792e-06, "loss": 1.2088, "step": 12515 }, { "epoch": 0.9808777429467085, "grad_norm": 3.462817430496216, "learning_rate": 8.60440155699078e-06, "loss": 1.7031, "step": 12516 }, { "epoch": 0.9809561128526646, "grad_norm": 3.439774751663208, "learning_rate": 8.604187836858475e-06, "loss": 1.3148, "step": 12517 }, { "epoch": 0.9810344827586207, "grad_norm": 3.83504581451416, "learning_rate": 8.603974103017688e-06, "loss": 1.8152, "step": 12518 }, { "epoch": 0.9811128526645768, "grad_norm": 2.375653028488159, "learning_rate": 8.60376035546924e-06, "loss": 1.3677, "step": 12519 }, { "epoch": 0.9811912225705329, "grad_norm": 1.701369285583496, "learning_rate": 8.603546594213935e-06, "loss": 1.2161, "step": 12520 }, { "epoch": 0.981269592476489, "grad_norm": 3.0813446044921875, "learning_rate": 8.603332819252591e-06, "loss": 1.1072, "step": 12521 }, { "epoch": 0.9813479623824451, "grad_norm": 3.845568895339966, "learning_rate": 8.60311903058602e-06, "loss": 1.428, "step": 12522 }, { "epoch": 0.9814263322884013, "grad_norm": 2.220874309539795, "learning_rate": 8.602905228215034e-06, "loss": 1.315, "step": 12523 }, { "epoch": 0.9815047021943574, "grad_norm": 4.239662170410156, "learning_rate": 8.602691412140447e-06, "loss": 1.6633, "step": 12524 }, { "epoch": 0.9815830721003135, "grad_norm": 2.7777321338653564, "learning_rate": 8.602477582363074e-06, "loss": 1.6017, "step": 12525 }, { "epoch": 0.9816614420062696, "grad_norm": 2.626893997192383, "learning_rate": 8.602263738883724e-06, "loss": 1.396, "step": 12526 }, { "epoch": 0.9817398119122257, "grad_norm": 3.8171045780181885, "learning_rate": 8.602049881703214e-06, "loss": 1.6027, "step": 12527 }, { "epoch": 0.9818181818181818, "grad_norm": 3.449986457824707, "learning_rate": 8.601836010822357e-06, "loss": 1.5794, "step": 12528 }, { "epoch": 0.9818965517241379, "grad_norm": 3.5116348266601562, "learning_rate": 8.601622126241966e-06, "loss": 1.0874, "step": 12529 }, { "epoch": 0.981974921630094, "grad_norm": 2.0935027599334717, "learning_rate": 8.601408227962853e-06, "loss": 1.0699, "step": 12530 }, { "epoch": 0.9820532915360501, "grad_norm": 2.516420364379883, "learning_rate": 8.601194315985834e-06, "loss": 1.2382, "step": 12531 }, { "epoch": 0.9821316614420063, "grad_norm": 3.0605218410491943, "learning_rate": 8.600980390311721e-06, "loss": 1.2043, "step": 12532 }, { "epoch": 0.9822100313479624, "grad_norm": 5.44143009185791, "learning_rate": 8.60076645094133e-06, "loss": 1.3508, "step": 12533 }, { "epoch": 0.9822884012539185, "grad_norm": 4.258394718170166, "learning_rate": 8.600552497875469e-06, "loss": 1.8144, "step": 12534 }, { "epoch": 0.9823667711598746, "grad_norm": 3.4266395568847656, "learning_rate": 8.600338531114959e-06, "loss": 1.5018, "step": 12535 }, { "epoch": 0.9824451410658307, "grad_norm": 2.2127273082733154, "learning_rate": 8.60012455066061e-06, "loss": 1.2721, "step": 12536 }, { "epoch": 0.9825235109717868, "grad_norm": 4.284083843231201, "learning_rate": 8.599910556513236e-06, "loss": 1.2565, "step": 12537 }, { "epoch": 0.9826018808777429, "grad_norm": 3.6269407272338867, "learning_rate": 8.599696548673651e-06, "loss": 2.0267, "step": 12538 }, { "epoch": 0.982680250783699, "grad_norm": 2.6731557846069336, "learning_rate": 8.599482527142668e-06, "loss": 1.7297, "step": 12539 }, { "epoch": 0.9827586206896551, "grad_norm": 2.242112636566162, "learning_rate": 8.599268491921104e-06, "loss": 1.2443, "step": 12540 }, { "epoch": 0.9828369905956112, "grad_norm": 2.287033796310425, "learning_rate": 8.599054443009771e-06, "loss": 0.9272, "step": 12541 }, { "epoch": 0.9829153605015674, "grad_norm": 4.187699317932129, "learning_rate": 8.598840380409484e-06, "loss": 1.0288, "step": 12542 }, { "epoch": 0.9829937304075235, "grad_norm": 2.914496421813965, "learning_rate": 8.598626304121056e-06, "loss": 0.9925, "step": 12543 }, { "epoch": 0.9830721003134796, "grad_norm": 3.6324052810668945, "learning_rate": 8.598412214145303e-06, "loss": 1.0151, "step": 12544 }, { "epoch": 0.9831504702194357, "grad_norm": 3.11525559425354, "learning_rate": 8.598198110483035e-06, "loss": 1.993, "step": 12545 }, { "epoch": 0.9832288401253918, "grad_norm": 2.5226709842681885, "learning_rate": 8.597983993135073e-06, "loss": 1.4975, "step": 12546 }, { "epoch": 0.9833072100313479, "grad_norm": 3.111124277114868, "learning_rate": 8.597769862102227e-06, "loss": 1.2117, "step": 12547 }, { "epoch": 0.983385579937304, "grad_norm": 2.1200339794158936, "learning_rate": 8.597555717385313e-06, "loss": 0.8017, "step": 12548 }, { "epoch": 0.9834639498432602, "grad_norm": 4.367138385772705, "learning_rate": 8.597341558985143e-06, "loss": 1.3815, "step": 12549 }, { "epoch": 0.9835423197492164, "grad_norm": 2.319340229034424, "learning_rate": 8.597127386902536e-06, "loss": 1.6507, "step": 12550 }, { "epoch": 0.9836206896551725, "grad_norm": 9.832923889160156, "learning_rate": 8.5969132011383e-06, "loss": 0.9778, "step": 12551 }, { "epoch": 0.9836990595611286, "grad_norm": 2.439357280731201, "learning_rate": 8.596699001693257e-06, "loss": 1.6069, "step": 12552 }, { "epoch": 0.9837774294670847, "grad_norm": 6.1882805824279785, "learning_rate": 8.596484788568217e-06, "loss": 1.4691, "step": 12553 }, { "epoch": 0.9838557993730408, "grad_norm": 2.8462307453155518, "learning_rate": 8.596270561763996e-06, "loss": 1.4692, "step": 12554 }, { "epoch": 0.9839341692789969, "grad_norm": 4.559489727020264, "learning_rate": 8.59605632128141e-06, "loss": 1.7556, "step": 12555 }, { "epoch": 0.984012539184953, "grad_norm": 3.0732178688049316, "learning_rate": 8.595842067121272e-06, "loss": 0.9221, "step": 12556 }, { "epoch": 0.9840909090909091, "grad_norm": 3.3513762950897217, "learning_rate": 8.595627799284399e-06, "loss": 1.259, "step": 12557 }, { "epoch": 0.9841692789968652, "grad_norm": 4.362037181854248, "learning_rate": 8.595413517771603e-06, "loss": 1.3233, "step": 12558 }, { "epoch": 0.9842476489028213, "grad_norm": 1.9229296445846558, "learning_rate": 8.595199222583702e-06, "loss": 1.3489, "step": 12559 }, { "epoch": 0.9843260188087775, "grad_norm": 4.124935150146484, "learning_rate": 8.594984913721509e-06, "loss": 1.1775, "step": 12560 }, { "epoch": 0.9844043887147336, "grad_norm": 2.692808151245117, "learning_rate": 8.59477059118584e-06, "loss": 1.7033, "step": 12561 }, { "epoch": 0.9844827586206897, "grad_norm": 6.818306922912598, "learning_rate": 8.59455625497751e-06, "loss": 1.669, "step": 12562 }, { "epoch": 0.9845611285266458, "grad_norm": 4.507667541503906, "learning_rate": 8.594341905097334e-06, "loss": 0.9465, "step": 12563 }, { "epoch": 0.9846394984326019, "grad_norm": 1.9573535919189453, "learning_rate": 8.594127541546129e-06, "loss": 1.3722, "step": 12564 }, { "epoch": 0.984717868338558, "grad_norm": 3.0790297985076904, "learning_rate": 8.593913164324708e-06, "loss": 1.5697, "step": 12565 }, { "epoch": 0.9847962382445141, "grad_norm": 1.9409465789794922, "learning_rate": 8.593698773433887e-06, "loss": 1.2366, "step": 12566 }, { "epoch": 0.9848746081504702, "grad_norm": 2.4190380573272705, "learning_rate": 8.593484368874481e-06, "loss": 1.469, "step": 12567 }, { "epoch": 0.9849529780564263, "grad_norm": 3.6635587215423584, "learning_rate": 8.593269950647308e-06, "loss": 1.812, "step": 12568 }, { "epoch": 0.9850313479623825, "grad_norm": 3.0031185150146484, "learning_rate": 8.59305551875318e-06, "loss": 1.1659, "step": 12569 }, { "epoch": 0.9851097178683386, "grad_norm": 3.1662235260009766, "learning_rate": 8.592841073192916e-06, "loss": 0.7156, "step": 12570 }, { "epoch": 0.9851880877742947, "grad_norm": 4.280407905578613, "learning_rate": 8.592626613967328e-06, "loss": 1.4066, "step": 12571 }, { "epoch": 0.9852664576802508, "grad_norm": 2.738905906677246, "learning_rate": 8.592412141077236e-06, "loss": 1.0487, "step": 12572 }, { "epoch": 0.9853448275862069, "grad_norm": 3.423675298690796, "learning_rate": 8.59219765452345e-06, "loss": 1.3554, "step": 12573 }, { "epoch": 0.985423197492163, "grad_norm": 4.873727798461914, "learning_rate": 8.591983154306793e-06, "loss": 1.3626, "step": 12574 }, { "epoch": 0.9855015673981191, "grad_norm": 2.2147769927978516, "learning_rate": 8.591768640428075e-06, "loss": 1.2775, "step": 12575 }, { "epoch": 0.9855799373040752, "grad_norm": 2.5620696544647217, "learning_rate": 8.591554112888114e-06, "loss": 1.5951, "step": 12576 }, { "epoch": 0.9856583072100313, "grad_norm": 9.314894676208496, "learning_rate": 8.591339571687727e-06, "loss": 1.5163, "step": 12577 }, { "epoch": 0.9857366771159874, "grad_norm": 1.651898741722107, "learning_rate": 8.591125016827729e-06, "loss": 1.1898, "step": 12578 }, { "epoch": 0.9858150470219436, "grad_norm": 6.900002479553223, "learning_rate": 8.590910448308933e-06, "loss": 1.0372, "step": 12579 }, { "epoch": 0.9858934169278997, "grad_norm": 2.166783571243286, "learning_rate": 8.590695866132162e-06, "loss": 1.5105, "step": 12580 }, { "epoch": 0.9859717868338558, "grad_norm": 2.0949697494506836, "learning_rate": 8.590481270298225e-06, "loss": 1.0864, "step": 12581 }, { "epoch": 0.9860501567398119, "grad_norm": 2.7729456424713135, "learning_rate": 8.590266660807943e-06, "loss": 1.3577, "step": 12582 }, { "epoch": 0.986128526645768, "grad_norm": 7.7831220626831055, "learning_rate": 8.59005203766213e-06, "loss": 1.1805, "step": 12583 }, { "epoch": 0.9862068965517241, "grad_norm": 3.0350756645202637, "learning_rate": 8.589837400861603e-06, "loss": 1.3519, "step": 12584 }, { "epoch": 0.9862852664576802, "grad_norm": 2.3742547035217285, "learning_rate": 8.58962275040718e-06, "loss": 1.6533, "step": 12585 }, { "epoch": 0.9863636363636363, "grad_norm": 2.9256062507629395, "learning_rate": 8.589408086299674e-06, "loss": 0.7768, "step": 12586 }, { "epoch": 0.9864420062695924, "grad_norm": 4.421408176422119, "learning_rate": 8.589193408539904e-06, "loss": 2.0003, "step": 12587 }, { "epoch": 0.9865203761755486, "grad_norm": 4.623698711395264, "learning_rate": 8.588978717128685e-06, "loss": 1.9247, "step": 12588 }, { "epoch": 0.9865987460815047, "grad_norm": 2.6931722164154053, "learning_rate": 8.588764012066836e-06, "loss": 1.1094, "step": 12589 }, { "epoch": 0.9866771159874608, "grad_norm": 4.2743377685546875, "learning_rate": 8.58854929335517e-06, "loss": 1.6807, "step": 12590 }, { "epoch": 0.9867554858934169, "grad_norm": 1.83441162109375, "learning_rate": 8.588334560994509e-06, "loss": 1.2696, "step": 12591 }, { "epoch": 0.986833855799373, "grad_norm": 3.042069911956787, "learning_rate": 8.588119814985663e-06, "loss": 1.6535, "step": 12592 }, { "epoch": 0.9869122257053291, "grad_norm": 3.246325731277466, "learning_rate": 8.587905055329452e-06, "loss": 1.1041, "step": 12593 }, { "epoch": 0.9869905956112852, "grad_norm": 2.8347012996673584, "learning_rate": 8.587690282026696e-06, "loss": 1.5597, "step": 12594 }, { "epoch": 0.9870689655172413, "grad_norm": 2.4426324367523193, "learning_rate": 8.587475495078208e-06, "loss": 1.4051, "step": 12595 }, { "epoch": 0.9871473354231975, "grad_norm": 4.217627048492432, "learning_rate": 8.587260694484804e-06, "loss": 1.6488, "step": 12596 }, { "epoch": 0.9872257053291537, "grad_norm": 3.6830978393554688, "learning_rate": 8.587045880247304e-06, "loss": 1.6744, "step": 12597 }, { "epoch": 0.9873040752351098, "grad_norm": 2.0316851139068604, "learning_rate": 8.586831052366525e-06, "loss": 1.1794, "step": 12598 }, { "epoch": 0.9873824451410659, "grad_norm": 3.4998340606689453, "learning_rate": 8.58661621084328e-06, "loss": 1.5756, "step": 12599 }, { "epoch": 0.987460815047022, "grad_norm": 3.7045929431915283, "learning_rate": 8.586401355678393e-06, "loss": 1.2162, "step": 12600 }, { "epoch": 0.9875391849529781, "grad_norm": 3.9413888454437256, "learning_rate": 8.586186486872676e-06, "loss": 0.389, "step": 12601 }, { "epoch": 0.9876175548589342, "grad_norm": 2.652290105819702, "learning_rate": 8.585971604426946e-06, "loss": 1.5785, "step": 12602 }, { "epoch": 0.9876959247648903, "grad_norm": 6.065511703491211, "learning_rate": 8.585756708342024e-06, "loss": 1.5529, "step": 12603 }, { "epoch": 0.9877742946708464, "grad_norm": 2.6202175617218018, "learning_rate": 8.585541798618724e-06, "loss": 1.4923, "step": 12604 }, { "epoch": 0.9878526645768025, "grad_norm": 3.1505584716796875, "learning_rate": 8.585326875257864e-06, "loss": 1.3837, "step": 12605 }, { "epoch": 0.9879310344827587, "grad_norm": 4.320906639099121, "learning_rate": 8.585111938260263e-06, "loss": 1.4905, "step": 12606 }, { "epoch": 0.9880094043887148, "grad_norm": 3.5669944286346436, "learning_rate": 8.584896987626737e-06, "loss": 1.1727, "step": 12607 }, { "epoch": 0.9880877742946709, "grad_norm": 2.791304588317871, "learning_rate": 8.584682023358106e-06, "loss": 1.474, "step": 12608 }, { "epoch": 0.988166144200627, "grad_norm": 2.749948740005493, "learning_rate": 8.584467045455183e-06, "loss": 1.3536, "step": 12609 }, { "epoch": 0.9882445141065831, "grad_norm": 4.206143856048584, "learning_rate": 8.58425205391879e-06, "loss": 1.9826, "step": 12610 }, { "epoch": 0.9883228840125392, "grad_norm": 2.0007097721099854, "learning_rate": 8.584037048749744e-06, "loss": 1.3619, "step": 12611 }, { "epoch": 0.9884012539184953, "grad_norm": 2.9153053760528564, "learning_rate": 8.58382202994886e-06, "loss": 0.7979, "step": 12612 }, { "epoch": 0.9884796238244514, "grad_norm": 3.468355894088745, "learning_rate": 8.58360699751696e-06, "loss": 1.4966, "step": 12613 }, { "epoch": 0.9885579937304075, "grad_norm": 2.430809736251831, "learning_rate": 8.583391951454858e-06, "loss": 1.6394, "step": 12614 }, { "epoch": 0.9886363636363636, "grad_norm": 6.559545993804932, "learning_rate": 8.583176891763374e-06, "loss": 2.0189, "step": 12615 }, { "epoch": 0.9887147335423198, "grad_norm": 2.8410043716430664, "learning_rate": 8.582961818443324e-06, "loss": 1.3028, "step": 12616 }, { "epoch": 0.9887931034482759, "grad_norm": 3.2128732204437256, "learning_rate": 8.58274673149553e-06, "loss": 1.1855, "step": 12617 }, { "epoch": 0.988871473354232, "grad_norm": 2.521470308303833, "learning_rate": 8.582531630920808e-06, "loss": 1.4918, "step": 12618 }, { "epoch": 0.9889498432601881, "grad_norm": 4.169057846069336, "learning_rate": 8.582316516719976e-06, "loss": 0.8688, "step": 12619 }, { "epoch": 0.9890282131661442, "grad_norm": 1.713484287261963, "learning_rate": 8.58210138889385e-06, "loss": 1.4164, "step": 12620 }, { "epoch": 0.9891065830721003, "grad_norm": 3.727468252182007, "learning_rate": 8.58188624744325e-06, "loss": 1.4797, "step": 12621 }, { "epoch": 0.9891849529780564, "grad_norm": 4.369228363037109, "learning_rate": 8.581671092368996e-06, "loss": 0.8659, "step": 12622 }, { "epoch": 0.9892633228840125, "grad_norm": 2.752593994140625, "learning_rate": 8.581455923671903e-06, "loss": 1.2207, "step": 12623 }, { "epoch": 0.9893416927899686, "grad_norm": 2.571495294570923, "learning_rate": 8.581240741352791e-06, "loss": 1.0707, "step": 12624 }, { "epoch": 0.9894200626959248, "grad_norm": 5.04171895980835, "learning_rate": 8.581025545412482e-06, "loss": 1.6408, "step": 12625 }, { "epoch": 0.9894984326018809, "grad_norm": 11.418739318847656, "learning_rate": 8.580810335851787e-06, "loss": 1.2604, "step": 12626 }, { "epoch": 0.989576802507837, "grad_norm": 2.7076268196105957, "learning_rate": 8.580595112671533e-06, "loss": 0.9091, "step": 12627 }, { "epoch": 0.9896551724137931, "grad_norm": 2.762874126434326, "learning_rate": 8.580379875872531e-06, "loss": 1.3735, "step": 12628 }, { "epoch": 0.9897335423197492, "grad_norm": 2.984055280685425, "learning_rate": 8.580164625455604e-06, "loss": 1.3946, "step": 12629 }, { "epoch": 0.9898119122257053, "grad_norm": 5.6592888832092285, "learning_rate": 8.579949361421569e-06, "loss": 1.3185, "step": 12630 }, { "epoch": 0.9898902821316614, "grad_norm": 2.5255913734436035, "learning_rate": 8.579734083771247e-06, "loss": 1.5198, "step": 12631 }, { "epoch": 0.9899686520376175, "grad_norm": 2.9197709560394287, "learning_rate": 8.579518792505454e-06, "loss": 1.7763, "step": 12632 }, { "epoch": 0.9900470219435736, "grad_norm": 3.0198252201080322, "learning_rate": 8.57930348762501e-06, "loss": 1.4351, "step": 12633 }, { "epoch": 0.9901253918495297, "grad_norm": 2.8512330055236816, "learning_rate": 8.579088169130736e-06, "loss": 0.4985, "step": 12634 }, { "epoch": 0.9902037617554859, "grad_norm": 7.724583625793457, "learning_rate": 8.578872837023446e-06, "loss": 0.9981, "step": 12635 }, { "epoch": 0.990282131661442, "grad_norm": 3.387269973754883, "learning_rate": 8.578657491303963e-06, "loss": 1.4409, "step": 12636 }, { "epoch": 0.9903605015673981, "grad_norm": 7.918820381164551, "learning_rate": 8.578442131973106e-06, "loss": 0.8932, "step": 12637 }, { "epoch": 0.9904388714733542, "grad_norm": 4.131274223327637, "learning_rate": 8.578226759031692e-06, "loss": 1.2912, "step": 12638 }, { "epoch": 0.9905172413793103, "grad_norm": 2.527047872543335, "learning_rate": 8.578011372480541e-06, "loss": 1.3517, "step": 12639 }, { "epoch": 0.9905956112852664, "grad_norm": 2.366787910461426, "learning_rate": 8.577795972320475e-06, "loss": 1.3254, "step": 12640 }, { "epoch": 0.9906739811912225, "grad_norm": 3.052622079849243, "learning_rate": 8.577580558552307e-06, "loss": 1.7211, "step": 12641 }, { "epoch": 0.9907523510971786, "grad_norm": 4.249399185180664, "learning_rate": 8.577365131176863e-06, "loss": 0.5943, "step": 12642 }, { "epoch": 0.9908307210031349, "grad_norm": 2.4851458072662354, "learning_rate": 8.57714969019496e-06, "loss": 1.0916, "step": 12643 }, { "epoch": 0.990909090909091, "grad_norm": 3.7061827182769775, "learning_rate": 8.576934235607416e-06, "loss": 1.4631, "step": 12644 }, { "epoch": 0.9909874608150471, "grad_norm": 2.271324872970581, "learning_rate": 8.576718767415048e-06, "loss": 1.2543, "step": 12645 }, { "epoch": 0.9910658307210032, "grad_norm": 2.4377598762512207, "learning_rate": 8.576503285618683e-06, "loss": 1.6509, "step": 12646 }, { "epoch": 0.9911442006269593, "grad_norm": 5.642116546630859, "learning_rate": 8.576287790219136e-06, "loss": 0.7931, "step": 12647 }, { "epoch": 0.9912225705329154, "grad_norm": 4.112650394439697, "learning_rate": 8.576072281217226e-06, "loss": 1.6285, "step": 12648 }, { "epoch": 0.9913009404388715, "grad_norm": 2.851050853729248, "learning_rate": 8.575856758613775e-06, "loss": 1.4642, "step": 12649 }, { "epoch": 0.9913793103448276, "grad_norm": 3.052213430404663, "learning_rate": 8.5756412224096e-06, "loss": 1.7526, "step": 12650 }, { "epoch": 0.9914576802507837, "grad_norm": 3.189229726791382, "learning_rate": 8.575425672605524e-06, "loss": 1.1934, "step": 12651 }, { "epoch": 0.9915360501567398, "grad_norm": 2.9267466068267822, "learning_rate": 8.575210109202364e-06, "loss": 1.4846, "step": 12652 }, { "epoch": 0.991614420062696, "grad_norm": 2.9053242206573486, "learning_rate": 8.574994532200942e-06, "loss": 1.4336, "step": 12653 }, { "epoch": 0.9916927899686521, "grad_norm": 5.395382404327393, "learning_rate": 8.574778941602076e-06, "loss": 1.5499, "step": 12654 }, { "epoch": 0.9917711598746082, "grad_norm": 3.1158814430236816, "learning_rate": 8.574563337406587e-06, "loss": 1.4325, "step": 12655 }, { "epoch": 0.9918495297805643, "grad_norm": 2.7646079063415527, "learning_rate": 8.574347719615298e-06, "loss": 0.9043, "step": 12656 }, { "epoch": 0.9919278996865204, "grad_norm": 2.750913381576538, "learning_rate": 8.574132088229024e-06, "loss": 1.1489, "step": 12657 }, { "epoch": 0.9920062695924765, "grad_norm": 4.934013843536377, "learning_rate": 8.573916443248587e-06, "loss": 1.3616, "step": 12658 }, { "epoch": 0.9920846394984326, "grad_norm": 2.8818883895874023, "learning_rate": 8.573700784674806e-06, "loss": 1.2392, "step": 12659 }, { "epoch": 0.9921630094043887, "grad_norm": 1.91861093044281, "learning_rate": 8.573485112508506e-06, "loss": 1.3139, "step": 12660 }, { "epoch": 0.9922413793103448, "grad_norm": 2.548786163330078, "learning_rate": 8.573269426750501e-06, "loss": 1.7968, "step": 12661 }, { "epoch": 0.992319749216301, "grad_norm": 3.302321195602417, "learning_rate": 8.573053727401616e-06, "loss": 1.2159, "step": 12662 }, { "epoch": 0.9923981191222571, "grad_norm": 1.6008557081222534, "learning_rate": 8.572838014462672e-06, "loss": 1.3893, "step": 12663 }, { "epoch": 0.9924764890282132, "grad_norm": 2.498239278793335, "learning_rate": 8.572622287934485e-06, "loss": 1.8609, "step": 12664 }, { "epoch": 0.9925548589341693, "grad_norm": 2.429408550262451, "learning_rate": 8.572406547817877e-06, "loss": 1.2554, "step": 12665 }, { "epoch": 0.9926332288401254, "grad_norm": 3.479555606842041, "learning_rate": 8.572190794113671e-06, "loss": 1.4196, "step": 12666 }, { "epoch": 0.9927115987460815, "grad_norm": 2.492579698562622, "learning_rate": 8.571975026822686e-06, "loss": 1.1715, "step": 12667 }, { "epoch": 0.9927899686520376, "grad_norm": 2.4703264236450195, "learning_rate": 8.571759245945741e-06, "loss": 1.5496, "step": 12668 }, { "epoch": 0.9928683385579937, "grad_norm": 5.783741474151611, "learning_rate": 8.571543451483659e-06, "loss": 2.0805, "step": 12669 }, { "epoch": 0.9929467084639498, "grad_norm": 6.853323936462402, "learning_rate": 8.571327643437261e-06, "loss": 0.5793, "step": 12670 }, { "epoch": 0.993025078369906, "grad_norm": 2.0889322757720947, "learning_rate": 8.571111821807368e-06, "loss": 1.2679, "step": 12671 }, { "epoch": 0.993103448275862, "grad_norm": 4.171865463256836, "learning_rate": 8.570895986594798e-06, "loss": 1.2972, "step": 12672 }, { "epoch": 0.9931818181818182, "grad_norm": 2.555022716522217, "learning_rate": 8.570680137800373e-06, "loss": 1.3959, "step": 12673 }, { "epoch": 0.9932601880877743, "grad_norm": 2.051722764968872, "learning_rate": 8.570464275424916e-06, "loss": 1.2508, "step": 12674 }, { "epoch": 0.9933385579937304, "grad_norm": 2.654219388961792, "learning_rate": 8.570248399469248e-06, "loss": 0.9495, "step": 12675 }, { "epoch": 0.9934169278996865, "grad_norm": 2.491583824157715, "learning_rate": 8.570032509934189e-06, "loss": 1.2503, "step": 12676 }, { "epoch": 0.9934952978056426, "grad_norm": 3.064004898071289, "learning_rate": 8.569816606820557e-06, "loss": 1.4833, "step": 12677 }, { "epoch": 0.9935736677115987, "grad_norm": 1.8651384115219116, "learning_rate": 8.569600690129179e-06, "loss": 1.2391, "step": 12678 }, { "epoch": 0.9936520376175548, "grad_norm": 2.8181958198547363, "learning_rate": 8.569384759860873e-06, "loss": 1.624, "step": 12679 }, { "epoch": 0.9937304075235109, "grad_norm": 2.145174026489258, "learning_rate": 8.569168816016459e-06, "loss": 1.2909, "step": 12680 }, { "epoch": 0.993808777429467, "grad_norm": 2.678962230682373, "learning_rate": 8.568952858596762e-06, "loss": 1.3818, "step": 12681 }, { "epoch": 0.9938871473354232, "grad_norm": 3.625947952270508, "learning_rate": 8.568736887602602e-06, "loss": 1.2626, "step": 12682 }, { "epoch": 0.9939655172413793, "grad_norm": 3.483318328857422, "learning_rate": 8.568520903034799e-06, "loss": 1.2361, "step": 12683 }, { "epoch": 0.9940438871473354, "grad_norm": 2.7552976608276367, "learning_rate": 8.568304904894174e-06, "loss": 1.4288, "step": 12684 }, { "epoch": 0.9941222570532915, "grad_norm": 2.8537676334381104, "learning_rate": 8.568088893181551e-06, "loss": 0.9082, "step": 12685 }, { "epoch": 0.9942006269592476, "grad_norm": 2.8663175106048584, "learning_rate": 8.56787286789775e-06, "loss": 1.3245, "step": 12686 }, { "epoch": 0.9942789968652037, "grad_norm": 2.9593615531921387, "learning_rate": 8.567656829043593e-06, "loss": 1.1844, "step": 12687 }, { "epoch": 0.9943573667711598, "grad_norm": 2.1073505878448486, "learning_rate": 8.567440776619903e-06, "loss": 1.3389, "step": 12688 }, { "epoch": 0.9944357366771159, "grad_norm": 2.230405569076538, "learning_rate": 8.5672247106275e-06, "loss": 1.5115, "step": 12689 }, { "epoch": 0.9945141065830722, "grad_norm": 2.7415194511413574, "learning_rate": 8.567008631067208e-06, "loss": 1.522, "step": 12690 }, { "epoch": 0.9945924764890283, "grad_norm": 2.6337504386901855, "learning_rate": 8.566792537939846e-06, "loss": 1.7057, "step": 12691 }, { "epoch": 0.9946708463949844, "grad_norm": 3.6402885913848877, "learning_rate": 8.566576431246237e-06, "loss": 1.4219, "step": 12692 }, { "epoch": 0.9947492163009405, "grad_norm": 2.8046512603759766, "learning_rate": 8.566360310987203e-06, "loss": 1.5902, "step": 12693 }, { "epoch": 0.9948275862068966, "grad_norm": 2.649371862411499, "learning_rate": 8.566144177163566e-06, "loss": 1.3495, "step": 12694 }, { "epoch": 0.9949059561128527, "grad_norm": 2.116948127746582, "learning_rate": 8.56592802977615e-06, "loss": 1.4028, "step": 12695 }, { "epoch": 0.9949843260188088, "grad_norm": 2.733903408050537, "learning_rate": 8.565711868825774e-06, "loss": 1.6482, "step": 12696 }, { "epoch": 0.9950626959247649, "grad_norm": 3.9674904346466064, "learning_rate": 8.565495694313262e-06, "loss": 1.7054, "step": 12697 }, { "epoch": 0.995141065830721, "grad_norm": 3.368262767791748, "learning_rate": 8.565279506239434e-06, "loss": 1.5247, "step": 12698 }, { "epoch": 0.9952194357366771, "grad_norm": 2.273878335952759, "learning_rate": 8.565063304605116e-06, "loss": 1.3231, "step": 12699 }, { "epoch": 0.9952978056426333, "grad_norm": 2.8312692642211914, "learning_rate": 8.564847089411128e-06, "loss": 1.5522, "step": 12700 }, { "epoch": 0.9953761755485894, "grad_norm": 11.83459758758545, "learning_rate": 8.564630860658292e-06, "loss": 1.2378, "step": 12701 }, { "epoch": 0.9954545454545455, "grad_norm": 2.3715286254882812, "learning_rate": 8.56441461834743e-06, "loss": 1.1726, "step": 12702 }, { "epoch": 0.9955329153605016, "grad_norm": 2.643608331680298, "learning_rate": 8.564198362479367e-06, "loss": 1.0731, "step": 12703 }, { "epoch": 0.9956112852664577, "grad_norm": 3.917184829711914, "learning_rate": 8.563982093054925e-06, "loss": 1.3632, "step": 12704 }, { "epoch": 0.9956896551724138, "grad_norm": 6.488761901855469, "learning_rate": 8.563765810074925e-06, "loss": 1.0289, "step": 12705 }, { "epoch": 0.9957680250783699, "grad_norm": 4.363621234893799, "learning_rate": 8.56354951354019e-06, "loss": 1.5049, "step": 12706 }, { "epoch": 0.995846394984326, "grad_norm": 2.9750144481658936, "learning_rate": 8.563333203451544e-06, "loss": 1.5036, "step": 12707 }, { "epoch": 0.9959247648902821, "grad_norm": 3.63209867477417, "learning_rate": 8.563116879809807e-06, "loss": 1.2099, "step": 12708 }, { "epoch": 0.9960031347962383, "grad_norm": 2.6894168853759766, "learning_rate": 8.562900542615805e-06, "loss": 1.3801, "step": 12709 }, { "epoch": 0.9960815047021944, "grad_norm": 3.200413703918457, "learning_rate": 8.562684191870359e-06, "loss": 1.5341, "step": 12710 }, { "epoch": 0.9961598746081505, "grad_norm": 2.9593021869659424, "learning_rate": 8.562467827574293e-06, "loss": 1.0497, "step": 12711 }, { "epoch": 0.9962382445141066, "grad_norm": 7.413888454437256, "learning_rate": 8.562251449728427e-06, "loss": 0.655, "step": 12712 }, { "epoch": 0.9963166144200627, "grad_norm": 3.5589520931243896, "learning_rate": 8.562035058333588e-06, "loss": 1.6613, "step": 12713 }, { "epoch": 0.9963949843260188, "grad_norm": 1.9139680862426758, "learning_rate": 8.561818653390596e-06, "loss": 1.4077, "step": 12714 }, { "epoch": 0.9964733542319749, "grad_norm": 3.0380592346191406, "learning_rate": 8.561602234900276e-06, "loss": 1.8867, "step": 12715 }, { "epoch": 0.996551724137931, "grad_norm": 3.548987865447998, "learning_rate": 8.561385802863452e-06, "loss": 1.8773, "step": 12716 }, { "epoch": 0.9966300940438871, "grad_norm": 2.428945541381836, "learning_rate": 8.561169357280942e-06, "loss": 0.9659, "step": 12717 }, { "epoch": 0.9967084639498432, "grad_norm": 3.7518558502197266, "learning_rate": 8.560952898153576e-06, "loss": 1.2886, "step": 12718 }, { "epoch": 0.9967868338557994, "grad_norm": 1.967769980430603, "learning_rate": 8.560736425482174e-06, "loss": 1.2437, "step": 12719 }, { "epoch": 0.9968652037617555, "grad_norm": 2.690253973007202, "learning_rate": 8.560519939267558e-06, "loss": 1.2499, "step": 12720 }, { "epoch": 0.9969435736677116, "grad_norm": 4.528698444366455, "learning_rate": 8.560303439510553e-06, "loss": 1.4324, "step": 12721 }, { "epoch": 0.9970219435736677, "grad_norm": 3.8281431198120117, "learning_rate": 8.560086926211984e-06, "loss": 1.0237, "step": 12722 }, { "epoch": 0.9971003134796238, "grad_norm": 4.100081920623779, "learning_rate": 8.559870399372672e-06, "loss": 1.6211, "step": 12723 }, { "epoch": 0.9971786833855799, "grad_norm": 2.583376169204712, "learning_rate": 8.55965385899344e-06, "loss": 1.3165, "step": 12724 }, { "epoch": 0.997257053291536, "grad_norm": 12.791356086730957, "learning_rate": 8.559437305075116e-06, "loss": 2.562, "step": 12725 }, { "epoch": 0.9973354231974921, "grad_norm": 4.9267401695251465, "learning_rate": 8.559220737618519e-06, "loss": 1.1916, "step": 12726 }, { "epoch": 0.9974137931034482, "grad_norm": 2.527432441711426, "learning_rate": 8.559004156624474e-06, "loss": 1.4489, "step": 12727 }, { "epoch": 0.9974921630094044, "grad_norm": 2.797638177871704, "learning_rate": 8.558787562093806e-06, "loss": 1.546, "step": 12728 }, { "epoch": 0.9975705329153605, "grad_norm": 10.76860237121582, "learning_rate": 8.558570954027338e-06, "loss": 1.2063, "step": 12729 }, { "epoch": 0.9976489028213166, "grad_norm": 3.0121099948883057, "learning_rate": 8.558354332425893e-06, "loss": 1.281, "step": 12730 }, { "epoch": 0.9977272727272727, "grad_norm": 4.773163795471191, "learning_rate": 8.558137697290297e-06, "loss": 1.1605, "step": 12731 }, { "epoch": 0.9978056426332288, "grad_norm": 3.111365556716919, "learning_rate": 8.557921048621374e-06, "loss": 0.9911, "step": 12732 }, { "epoch": 0.9978840125391849, "grad_norm": 2.3876943588256836, "learning_rate": 8.557704386419944e-06, "loss": 1.3793, "step": 12733 }, { "epoch": 0.997962382445141, "grad_norm": 3.1423094272613525, "learning_rate": 8.557487710686836e-06, "loss": 1.2046, "step": 12734 }, { "epoch": 0.9980407523510971, "grad_norm": 2.281649589538574, "learning_rate": 8.55727102142287e-06, "loss": 1.2415, "step": 12735 }, { "epoch": 0.9981191222570532, "grad_norm": 5.679161548614502, "learning_rate": 8.557054318628874e-06, "loss": 1.0543, "step": 12736 }, { "epoch": 0.9981974921630095, "grad_norm": 3.155712366104126, "learning_rate": 8.55683760230567e-06, "loss": 1.7376, "step": 12737 }, { "epoch": 0.9982758620689656, "grad_norm": 2.8027641773223877, "learning_rate": 8.556620872454082e-06, "loss": 1.4164, "step": 12738 }, { "epoch": 0.9983542319749217, "grad_norm": 4.257358551025391, "learning_rate": 8.556404129074935e-06, "loss": 1.4277, "step": 12739 }, { "epoch": 0.9984326018808778, "grad_norm": 2.7715446949005127, "learning_rate": 8.556187372169054e-06, "loss": 1.2105, "step": 12740 }, { "epoch": 0.9985109717868339, "grad_norm": 1.9646192789077759, "learning_rate": 8.555970601737261e-06, "loss": 1.1925, "step": 12741 }, { "epoch": 0.99858934169279, "grad_norm": 3.417363166809082, "learning_rate": 8.555753817780383e-06, "loss": 1.1459, "step": 12742 }, { "epoch": 0.9986677115987461, "grad_norm": 3.3400254249572754, "learning_rate": 8.555537020299245e-06, "loss": 1.5036, "step": 12743 }, { "epoch": 0.9987460815047022, "grad_norm": 2.099525213241577, "learning_rate": 8.555320209294671e-06, "loss": 1.1966, "step": 12744 }, { "epoch": 0.9988244514106583, "grad_norm": 3.5284950733184814, "learning_rate": 8.555103384767485e-06, "loss": 1.7351, "step": 12745 }, { "epoch": 0.9989028213166145, "grad_norm": 2.1027164459228516, "learning_rate": 8.55488654671851e-06, "loss": 1.317, "step": 12746 }, { "epoch": 0.9989811912225706, "grad_norm": 2.901682138442993, "learning_rate": 8.554669695148572e-06, "loss": 0.8695, "step": 12747 }, { "epoch": 0.9990595611285267, "grad_norm": 3.3259010314941406, "learning_rate": 8.554452830058498e-06, "loss": 1.2054, "step": 12748 }, { "epoch": 0.9991379310344828, "grad_norm": 4.632752418518066, "learning_rate": 8.55423595144911e-06, "loss": 1.8884, "step": 12749 }, { "epoch": 0.9992163009404389, "grad_norm": 4.423877716064453, "learning_rate": 8.554019059321236e-06, "loss": 0.6102, "step": 12750 }, { "epoch": 0.999294670846395, "grad_norm": 3.4516451358795166, "learning_rate": 8.553802153675698e-06, "loss": 1.2986, "step": 12751 }, { "epoch": 0.9993730407523511, "grad_norm": 2.6741952896118164, "learning_rate": 8.55358523451332e-06, "loss": 1.5614, "step": 12752 }, { "epoch": 0.9994514106583072, "grad_norm": 4.653217792510986, "learning_rate": 8.553368301834932e-06, "loss": 1.7107, "step": 12753 }, { "epoch": 0.9995297805642633, "grad_norm": 2.3043267726898193, "learning_rate": 8.553151355641354e-06, "loss": 1.0954, "step": 12754 }, { "epoch": 0.9996081504702194, "grad_norm": 3.252101421356201, "learning_rate": 8.552934395933414e-06, "loss": 1.9188, "step": 12755 }, { "epoch": 0.9996865203761756, "grad_norm": 4.028286457061768, "learning_rate": 8.552717422711936e-06, "loss": 1.5121, "step": 12756 }, { "epoch": 0.9997648902821317, "grad_norm": 2.9001107215881348, "learning_rate": 8.552500435977746e-06, "loss": 1.5463, "step": 12757 }, { "epoch": 0.9998432601880878, "grad_norm": 2.9361319541931152, "learning_rate": 8.55228343573167e-06, "loss": 1.5772, "step": 12758 }, { "epoch": 0.9999216300940439, "grad_norm": 4.5222907066345215, "learning_rate": 8.552066421974532e-06, "loss": 1.1448, "step": 12759 }, { "epoch": 1.0, "grad_norm": 3.2433459758758545, "learning_rate": 8.551849394707158e-06, "loss": 1.2149, "step": 12760 }, { "epoch": 1.0, "eval_loss": 1.3089427947998047, "eval_runtime": 106.8592, "eval_samples_per_second": 25.145, "eval_steps_per_second": 6.289, "step": 12760 }, { "epoch": 1.000078369905956, "grad_norm": 2.99858021736145, "learning_rate": 8.551632353930371e-06, "loss": 1.5115, "step": 12761 }, { "epoch": 1.0001567398119122, "grad_norm": 3.314622640609741, "learning_rate": 8.551415299645e-06, "loss": 1.2785, "step": 12762 }, { "epoch": 1.0002351097178683, "grad_norm": 4.573703765869141, "learning_rate": 8.551198231851869e-06, "loss": 1.1831, "step": 12763 }, { "epoch": 1.0003134796238244, "grad_norm": 2.799522638320923, "learning_rate": 8.550981150551805e-06, "loss": 1.303, "step": 12764 }, { "epoch": 1.0003918495297806, "grad_norm": 9.815847396850586, "learning_rate": 8.550764055745633e-06, "loss": 1.4801, "step": 12765 }, { "epoch": 1.0004702194357367, "grad_norm": 2.6504507064819336, "learning_rate": 8.550546947434178e-06, "loss": 1.6108, "step": 12766 }, { "epoch": 1.0005485893416928, "grad_norm": 3.4595682621002197, "learning_rate": 8.550329825618265e-06, "loss": 1.2973, "step": 12767 }, { "epoch": 1.0006269592476489, "grad_norm": 4.353537559509277, "learning_rate": 8.550112690298722e-06, "loss": 1.2068, "step": 12768 }, { "epoch": 1.000705329153605, "grad_norm": 7.015102863311768, "learning_rate": 8.549895541476371e-06, "loss": 0.975, "step": 12769 }, { "epoch": 1.000783699059561, "grad_norm": 3.7346174716949463, "learning_rate": 8.549678379152044e-06, "loss": 1.8998, "step": 12770 }, { "epoch": 1.0008620689655172, "grad_norm": 2.0979840755462646, "learning_rate": 8.549461203326563e-06, "loss": 1.3947, "step": 12771 }, { "epoch": 1.0009404388714733, "grad_norm": 6.0119428634643555, "learning_rate": 8.549244014000755e-06, "loss": 1.0599, "step": 12772 }, { "epoch": 1.0010188087774294, "grad_norm": 11.669394493103027, "learning_rate": 8.549026811175446e-06, "loss": 1.5387, "step": 12773 }, { "epoch": 1.0010971786833855, "grad_norm": 2.2430543899536133, "learning_rate": 8.548809594851459e-06, "loss": 1.4625, "step": 12774 }, { "epoch": 1.0011755485893417, "grad_norm": 1.6574326753616333, "learning_rate": 8.548592365029626e-06, "loss": 0.1508, "step": 12775 }, { "epoch": 1.0012539184952978, "grad_norm": 2.6858153343200684, "learning_rate": 8.54837512171077e-06, "loss": 1.3313, "step": 12776 }, { "epoch": 1.0013322884012539, "grad_norm": 2.688775062561035, "learning_rate": 8.548157864895716e-06, "loss": 1.5377, "step": 12777 }, { "epoch": 1.00141065830721, "grad_norm": 3.0439889430999756, "learning_rate": 8.547940594585293e-06, "loss": 1.2556, "step": 12778 }, { "epoch": 1.001489028213166, "grad_norm": 3.6552436351776123, "learning_rate": 8.547723310780328e-06, "loss": 1.4562, "step": 12779 }, { "epoch": 1.0015673981191222, "grad_norm": 6.194640636444092, "learning_rate": 8.547506013481643e-06, "loss": 0.7845, "step": 12780 }, { "epoch": 1.0016457680250783, "grad_norm": 3.467989683151245, "learning_rate": 8.547288702690067e-06, "loss": 1.4805, "step": 12781 }, { "epoch": 1.0017241379310344, "grad_norm": 2.3245558738708496, "learning_rate": 8.547071378406428e-06, "loss": 0.4411, "step": 12782 }, { "epoch": 1.0018025078369905, "grad_norm": 5.135527610778809, "learning_rate": 8.54685404063155e-06, "loss": 1.2873, "step": 12783 }, { "epoch": 1.0018808777429467, "grad_norm": 2.9189953804016113, "learning_rate": 8.546636689366264e-06, "loss": 1.4873, "step": 12784 }, { "epoch": 1.0019592476489028, "grad_norm": 1.8477758169174194, "learning_rate": 8.546419324611393e-06, "loss": 1.2974, "step": 12785 }, { "epoch": 1.0020376175548589, "grad_norm": 3.5214293003082275, "learning_rate": 8.546201946367763e-06, "loss": 1.0253, "step": 12786 }, { "epoch": 1.002115987460815, "grad_norm": 3.767371416091919, "learning_rate": 8.545984554636204e-06, "loss": 1.3561, "step": 12787 }, { "epoch": 1.002194357366771, "grad_norm": 2.0975840091705322, "learning_rate": 8.545767149417539e-06, "loss": 1.2013, "step": 12788 }, { "epoch": 1.0022727272727272, "grad_norm": 2.7078804969787598, "learning_rate": 8.545549730712598e-06, "loss": 1.3612, "step": 12789 }, { "epoch": 1.0023510971786833, "grad_norm": 2.1050806045532227, "learning_rate": 8.545332298522207e-06, "loss": 1.4917, "step": 12790 }, { "epoch": 1.0024294670846394, "grad_norm": 6.804091930389404, "learning_rate": 8.545114852847192e-06, "loss": 1.8107, "step": 12791 }, { "epoch": 1.0025078369905955, "grad_norm": 2.952054500579834, "learning_rate": 8.544897393688383e-06, "loss": 1.261, "step": 12792 }, { "epoch": 1.0025862068965516, "grad_norm": 2.559403419494629, "learning_rate": 8.544679921046604e-06, "loss": 1.1749, "step": 12793 }, { "epoch": 1.0026645768025078, "grad_norm": 2.217090129852295, "learning_rate": 8.544462434922682e-06, "loss": 1.3185, "step": 12794 }, { "epoch": 1.0027429467084639, "grad_norm": 13.23897647857666, "learning_rate": 8.544244935317445e-06, "loss": 0.996, "step": 12795 }, { "epoch": 1.00282131661442, "grad_norm": 4.351098537445068, "learning_rate": 8.544027422231723e-06, "loss": 1.6853, "step": 12796 }, { "epoch": 1.002899686520376, "grad_norm": 2.2052724361419678, "learning_rate": 8.54380989566634e-06, "loss": 1.4979, "step": 12797 }, { "epoch": 1.0029780564263322, "grad_norm": 4.228521347045898, "learning_rate": 8.543592355622126e-06, "loss": 1.5586, "step": 12798 }, { "epoch": 1.0030564263322883, "grad_norm": 2.543215036392212, "learning_rate": 8.543374802099903e-06, "loss": 1.4351, "step": 12799 }, { "epoch": 1.0031347962382444, "grad_norm": 4.579706192016602, "learning_rate": 8.543157235100505e-06, "loss": 1.2558, "step": 12800 }, { "epoch": 1.0032131661442005, "grad_norm": 5.155387878417969, "learning_rate": 8.542939654624757e-06, "loss": 1.6799, "step": 12801 }, { "epoch": 1.0032915360501566, "grad_norm": 1.9706945419311523, "learning_rate": 8.542722060673487e-06, "loss": 1.367, "step": 12802 }, { "epoch": 1.0033699059561128, "grad_norm": 5.290753364562988, "learning_rate": 8.542504453247519e-06, "loss": 1.3877, "step": 12803 }, { "epoch": 1.0034482758620689, "grad_norm": 3.4581477642059326, "learning_rate": 8.542286832347686e-06, "loss": 1.3603, "step": 12804 }, { "epoch": 1.003526645768025, "grad_norm": 2.6592488288879395, "learning_rate": 8.542069197974815e-06, "loss": 1.409, "step": 12805 }, { "epoch": 1.003605015673981, "grad_norm": 3.4132988452911377, "learning_rate": 8.541851550129727e-06, "loss": 1.4397, "step": 12806 }, { "epoch": 1.0036833855799374, "grad_norm": 7.107194900512695, "learning_rate": 8.541633888813259e-06, "loss": 0.778, "step": 12807 }, { "epoch": 1.0037617554858935, "grad_norm": 2.965402603149414, "learning_rate": 8.541416214026233e-06, "loss": 1.3201, "step": 12808 }, { "epoch": 1.0038401253918496, "grad_norm": 2.6012797355651855, "learning_rate": 8.541198525769478e-06, "loss": 1.5153, "step": 12809 }, { "epoch": 1.0039184952978057, "grad_norm": 3.0031161308288574, "learning_rate": 8.540980824043825e-06, "loss": 1.0372, "step": 12810 }, { "epoch": 1.0039968652037619, "grad_norm": 4.1155548095703125, "learning_rate": 8.540763108850098e-06, "loss": 1.5302, "step": 12811 }, { "epoch": 1.004075235109718, "grad_norm": 9.726351737976074, "learning_rate": 8.540545380189128e-06, "loss": 0.8747, "step": 12812 }, { "epoch": 1.004153605015674, "grad_norm": 6.9155402183532715, "learning_rate": 8.540327638061741e-06, "loss": 0.8802, "step": 12813 }, { "epoch": 1.0042319749216302, "grad_norm": 2.2770867347717285, "learning_rate": 8.540109882468765e-06, "loss": 1.2044, "step": 12814 }, { "epoch": 1.0043103448275863, "grad_norm": 2.316169500350952, "learning_rate": 8.539892113411031e-06, "loss": 1.5824, "step": 12815 }, { "epoch": 1.0043887147335424, "grad_norm": 3.5299408435821533, "learning_rate": 8.539674330889365e-06, "loss": 1.4589, "step": 12816 }, { "epoch": 1.0044670846394985, "grad_norm": 2.913917303085327, "learning_rate": 8.539456534904596e-06, "loss": 1.3268, "step": 12817 }, { "epoch": 1.0045454545454546, "grad_norm": 4.396604537963867, "learning_rate": 8.539238725457553e-06, "loss": 1.4021, "step": 12818 }, { "epoch": 1.0046238244514107, "grad_norm": 2.204977035522461, "learning_rate": 8.539020902549063e-06, "loss": 1.3065, "step": 12819 }, { "epoch": 1.0047021943573669, "grad_norm": 2.617170572280884, "learning_rate": 8.538803066179955e-06, "loss": 1.639, "step": 12820 }, { "epoch": 1.004780564263323, "grad_norm": 2.3309590816497803, "learning_rate": 8.538585216351058e-06, "loss": 1.1476, "step": 12821 }, { "epoch": 1.004858934169279, "grad_norm": 2.0124855041503906, "learning_rate": 8.538367353063201e-06, "loss": 1.2116, "step": 12822 }, { "epoch": 1.0049373040752352, "grad_norm": 2.8232321739196777, "learning_rate": 8.538149476317212e-06, "loss": 1.3529, "step": 12823 }, { "epoch": 1.0050156739811913, "grad_norm": 5.575414180755615, "learning_rate": 8.53793158611392e-06, "loss": 1.4345, "step": 12824 }, { "epoch": 1.0050940438871474, "grad_norm": 4.842805862426758, "learning_rate": 8.537713682454151e-06, "loss": 1.811, "step": 12825 }, { "epoch": 1.0051724137931035, "grad_norm": 1.976209044456482, "learning_rate": 8.53749576533874e-06, "loss": 1.3347, "step": 12826 }, { "epoch": 1.0052507836990596, "grad_norm": 3.3163161277770996, "learning_rate": 8.537277834768509e-06, "loss": 1.1168, "step": 12827 }, { "epoch": 1.0053291536050157, "grad_norm": 2.9606168270111084, "learning_rate": 8.53705989074429e-06, "loss": 1.2406, "step": 12828 }, { "epoch": 1.0054075235109718, "grad_norm": 2.1805367469787598, "learning_rate": 8.536841933266913e-06, "loss": 1.3103, "step": 12829 }, { "epoch": 1.005485893416928, "grad_norm": 2.4706358909606934, "learning_rate": 8.536623962337206e-06, "loss": 0.7198, "step": 12830 }, { "epoch": 1.005564263322884, "grad_norm": 2.2892165184020996, "learning_rate": 8.536405977955998e-06, "loss": 1.2043, "step": 12831 }, { "epoch": 1.0056426332288402, "grad_norm": 3.449859619140625, "learning_rate": 8.536187980124118e-06, "loss": 1.3794, "step": 12832 }, { "epoch": 1.0057210031347963, "grad_norm": 2.4129457473754883, "learning_rate": 8.535969968842395e-06, "loss": 0.8294, "step": 12833 }, { "epoch": 1.0057993730407524, "grad_norm": 2.511030673980713, "learning_rate": 8.535751944111659e-06, "loss": 1.5445, "step": 12834 }, { "epoch": 1.0058777429467085, "grad_norm": 3.338425636291504, "learning_rate": 8.535533905932739e-06, "loss": 1.2426, "step": 12835 }, { "epoch": 1.0059561128526646, "grad_norm": 3.0571181774139404, "learning_rate": 8.535315854306464e-06, "loss": 1.4976, "step": 12836 }, { "epoch": 1.0060344827586207, "grad_norm": 3.9204773902893066, "learning_rate": 8.535097789233661e-06, "loss": 1.3172, "step": 12837 }, { "epoch": 1.0061128526645768, "grad_norm": 5.8214850425720215, "learning_rate": 8.534879710715164e-06, "loss": 0.9329, "step": 12838 }, { "epoch": 1.006191222570533, "grad_norm": 3.308328628540039, "learning_rate": 8.5346616187518e-06, "loss": 1.6994, "step": 12839 }, { "epoch": 1.006269592476489, "grad_norm": 2.139037609100342, "learning_rate": 8.534443513344397e-06, "loss": 1.352, "step": 12840 }, { "epoch": 1.0063479623824452, "grad_norm": 2.924690008163452, "learning_rate": 8.534225394493788e-06, "loss": 1.3504, "step": 12841 }, { "epoch": 1.0064263322884013, "grad_norm": 2.202942371368408, "learning_rate": 8.534007262200801e-06, "loss": 1.0343, "step": 12842 }, { "epoch": 1.0065047021943574, "grad_norm": 4.306562900543213, "learning_rate": 8.533789116466264e-06, "loss": 1.658, "step": 12843 }, { "epoch": 1.0065830721003135, "grad_norm": 5.617998123168945, "learning_rate": 8.53357095729101e-06, "loss": 1.4604, "step": 12844 }, { "epoch": 1.0066614420062696, "grad_norm": 3.1121342182159424, "learning_rate": 8.533352784675866e-06, "loss": 1.3599, "step": 12845 }, { "epoch": 1.0067398119122257, "grad_norm": 2.4361095428466797, "learning_rate": 8.533134598621665e-06, "loss": 1.2016, "step": 12846 }, { "epoch": 1.0068181818181818, "grad_norm": 2.8082661628723145, "learning_rate": 8.532916399129231e-06, "loss": 1.7291, "step": 12847 }, { "epoch": 1.006896551724138, "grad_norm": 3.312918186187744, "learning_rate": 8.532698186199399e-06, "loss": 1.3006, "step": 12848 }, { "epoch": 1.006974921630094, "grad_norm": 2.032118558883667, "learning_rate": 8.532479959832998e-06, "loss": 1.0623, "step": 12849 }, { "epoch": 1.0070532915360502, "grad_norm": 2.5598349571228027, "learning_rate": 8.53226172003086e-06, "loss": 1.5775, "step": 12850 }, { "epoch": 1.0071316614420063, "grad_norm": 2.976355791091919, "learning_rate": 8.53204346679381e-06, "loss": 1.4067, "step": 12851 }, { "epoch": 1.0072100313479624, "grad_norm": 1.735172986984253, "learning_rate": 8.531825200122681e-06, "loss": 1.5351, "step": 12852 }, { "epoch": 1.0072884012539185, "grad_norm": 2.7078325748443604, "learning_rate": 8.531606920018303e-06, "loss": 1.8686, "step": 12853 }, { "epoch": 1.0073667711598746, "grad_norm": 2.6962215900421143, "learning_rate": 8.531388626481505e-06, "loss": 1.1207, "step": 12854 }, { "epoch": 1.0074451410658307, "grad_norm": 10.511505126953125, "learning_rate": 8.531170319513122e-06, "loss": 1.2079, "step": 12855 }, { "epoch": 1.0075235109717868, "grad_norm": 4.086256504058838, "learning_rate": 8.530951999113977e-06, "loss": 0.9225, "step": 12856 }, { "epoch": 1.007601880877743, "grad_norm": 4.32431697845459, "learning_rate": 8.530733665284906e-06, "loss": 1.4202, "step": 12857 }, { "epoch": 1.007680250783699, "grad_norm": 4.952696323394775, "learning_rate": 8.530515318026738e-06, "loss": 2.0334, "step": 12858 }, { "epoch": 1.0077586206896552, "grad_norm": 7.388339996337891, "learning_rate": 8.530296957340302e-06, "loss": 0.8243, "step": 12859 }, { "epoch": 1.0078369905956113, "grad_norm": 2.612337589263916, "learning_rate": 8.530078583226431e-06, "loss": 1.0649, "step": 12860 }, { "epoch": 1.0079153605015674, "grad_norm": 6.996596813201904, "learning_rate": 8.529860195685954e-06, "loss": 1.6216, "step": 12861 }, { "epoch": 1.0079937304075235, "grad_norm": 2.505988359451294, "learning_rate": 8.5296417947197e-06, "loss": 1.1487, "step": 12862 }, { "epoch": 1.0080721003134796, "grad_norm": 2.9358954429626465, "learning_rate": 8.529423380328506e-06, "loss": 1.3863, "step": 12863 }, { "epoch": 1.0081504702194357, "grad_norm": 3.9804162979125977, "learning_rate": 8.529204952513195e-06, "loss": 1.2286, "step": 12864 }, { "epoch": 1.0082288401253918, "grad_norm": 6.6866655349731445, "learning_rate": 8.5289865112746e-06, "loss": 0.8319, "step": 12865 }, { "epoch": 1.008307210031348, "grad_norm": 2.0157833099365234, "learning_rate": 8.528768056613554e-06, "loss": 1.521, "step": 12866 }, { "epoch": 1.008385579937304, "grad_norm": 3.4458117485046387, "learning_rate": 8.528549588530888e-06, "loss": 1.4774, "step": 12867 }, { "epoch": 1.0084639498432602, "grad_norm": 2.6353588104248047, "learning_rate": 8.528331107027432e-06, "loss": 1.3337, "step": 12868 }, { "epoch": 1.0085423197492163, "grad_norm": 3.645805835723877, "learning_rate": 8.528112612104014e-06, "loss": 1.4625, "step": 12869 }, { "epoch": 1.0086206896551724, "grad_norm": 1.8649407625198364, "learning_rate": 8.52789410376147e-06, "loss": 1.2885, "step": 12870 }, { "epoch": 1.0086990595611285, "grad_norm": 2.0593507289886475, "learning_rate": 8.527675582000627e-06, "loss": 1.5861, "step": 12871 }, { "epoch": 1.0087774294670846, "grad_norm": 2.2636356353759766, "learning_rate": 8.527457046822321e-06, "loss": 1.4529, "step": 12872 }, { "epoch": 1.0088557993730407, "grad_norm": 2.795264959335327, "learning_rate": 8.527238498227377e-06, "loss": 1.4746, "step": 12873 }, { "epoch": 1.0089341692789968, "grad_norm": 5.769553184509277, "learning_rate": 8.527019936216631e-06, "loss": 1.2186, "step": 12874 }, { "epoch": 1.009012539184953, "grad_norm": 1.9561785459518433, "learning_rate": 8.526801360790913e-06, "loss": 1.3167, "step": 12875 }, { "epoch": 1.009090909090909, "grad_norm": 8.033516883850098, "learning_rate": 8.526582771951055e-06, "loss": 1.7325, "step": 12876 }, { "epoch": 1.0091692789968651, "grad_norm": 2.8227407932281494, "learning_rate": 8.526364169697888e-06, "loss": 1.4301, "step": 12877 }, { "epoch": 1.0092476489028213, "grad_norm": 4.764492034912109, "learning_rate": 8.52614555403224e-06, "loss": 1.8303, "step": 12878 }, { "epoch": 1.0093260188087774, "grad_norm": 3.124629259109497, "learning_rate": 8.525926924954946e-06, "loss": 1.4828, "step": 12879 }, { "epoch": 1.0094043887147335, "grad_norm": 2.9421603679656982, "learning_rate": 8.525708282466839e-06, "loss": 1.0439, "step": 12880 }, { "epoch": 1.0094827586206896, "grad_norm": 1.4732837677001953, "learning_rate": 8.525489626568746e-06, "loss": 1.3274, "step": 12881 }, { "epoch": 1.0095611285266457, "grad_norm": 6.790871620178223, "learning_rate": 8.525270957261503e-06, "loss": 1.5996, "step": 12882 }, { "epoch": 1.0096394984326018, "grad_norm": 3.5675196647644043, "learning_rate": 8.52505227454594e-06, "loss": 1.2338, "step": 12883 }, { "epoch": 1.009717868338558, "grad_norm": 2.482029438018799, "learning_rate": 8.524833578422888e-06, "loss": 1.6257, "step": 12884 }, { "epoch": 1.009796238244514, "grad_norm": 2.272437572479248, "learning_rate": 8.52461486889318e-06, "loss": 1.4101, "step": 12885 }, { "epoch": 1.0098746081504701, "grad_norm": 2.265866756439209, "learning_rate": 8.524396145957648e-06, "loss": 1.4091, "step": 12886 }, { "epoch": 1.0099529780564263, "grad_norm": 4.261518955230713, "learning_rate": 8.524177409617121e-06, "loss": 1.1303, "step": 12887 }, { "epoch": 1.0100313479623824, "grad_norm": 2.030233383178711, "learning_rate": 8.523958659872436e-06, "loss": 1.5998, "step": 12888 }, { "epoch": 1.0101097178683385, "grad_norm": 2.2005186080932617, "learning_rate": 8.52373989672442e-06, "loss": 1.4577, "step": 12889 }, { "epoch": 1.0101880877742946, "grad_norm": 3.490309476852417, "learning_rate": 8.523521120173907e-06, "loss": 1.3361, "step": 12890 }, { "epoch": 1.0102664576802507, "grad_norm": 2.432539224624634, "learning_rate": 8.523302330221733e-06, "loss": 1.1719, "step": 12891 }, { "epoch": 1.0103448275862068, "grad_norm": 3.552152633666992, "learning_rate": 8.523083526868725e-06, "loss": 1.2395, "step": 12892 }, { "epoch": 1.010423197492163, "grad_norm": 2.0287044048309326, "learning_rate": 8.522864710115715e-06, "loss": 1.3842, "step": 12893 }, { "epoch": 1.010501567398119, "grad_norm": 2.4932355880737305, "learning_rate": 8.522645879963539e-06, "loss": 1.1073, "step": 12894 }, { "epoch": 1.0105799373040751, "grad_norm": 5.093968391418457, "learning_rate": 8.522427036413027e-06, "loss": 1.6112, "step": 12895 }, { "epoch": 1.0106583072100312, "grad_norm": 2.3422951698303223, "learning_rate": 8.522208179465012e-06, "loss": 1.0876, "step": 12896 }, { "epoch": 1.0107366771159874, "grad_norm": 3.471529483795166, "learning_rate": 8.521989309120324e-06, "loss": 1.3962, "step": 12897 }, { "epoch": 1.0108150470219435, "grad_norm": 2.4791555404663086, "learning_rate": 8.5217704253798e-06, "loss": 1.183, "step": 12898 }, { "epoch": 1.0108934169278996, "grad_norm": 4.066202640533447, "learning_rate": 8.52155152824427e-06, "loss": 1.432, "step": 12899 }, { "epoch": 1.0109717868338557, "grad_norm": 3.766331911087036, "learning_rate": 8.521332617714567e-06, "loss": 0.8251, "step": 12900 }, { "epoch": 1.0110501567398118, "grad_norm": 2.5642669200897217, "learning_rate": 8.521113693791524e-06, "loss": 1.6446, "step": 12901 }, { "epoch": 1.0111285266457681, "grad_norm": 6.712491512298584, "learning_rate": 8.52089475647597e-06, "loss": 1.5144, "step": 12902 }, { "epoch": 1.0112068965517242, "grad_norm": 3.2079315185546875, "learning_rate": 8.520675805768744e-06, "loss": 1.0852, "step": 12903 }, { "epoch": 1.0112852664576804, "grad_norm": 3.4200732707977295, "learning_rate": 8.520456841670675e-06, "loss": 1.4883, "step": 12904 }, { "epoch": 1.0113636363636365, "grad_norm": 3.288625717163086, "learning_rate": 8.520237864182594e-06, "loss": 1.2677, "step": 12905 }, { "epoch": 1.0114420062695926, "grad_norm": 3.490400552749634, "learning_rate": 8.520018873305338e-06, "loss": 1.3239, "step": 12906 }, { "epoch": 1.0115203761755487, "grad_norm": 2.5963215827941895, "learning_rate": 8.519799869039739e-06, "loss": 1.3101, "step": 12907 }, { "epoch": 1.0115987460815048, "grad_norm": 5.0322136878967285, "learning_rate": 8.519580851386628e-06, "loss": 1.3433, "step": 12908 }, { "epoch": 1.011677115987461, "grad_norm": 4.004941940307617, "learning_rate": 8.51936182034684e-06, "loss": 1.381, "step": 12909 }, { "epoch": 1.011755485893417, "grad_norm": 2.8393211364746094, "learning_rate": 8.519142775921207e-06, "loss": 1.3522, "step": 12910 }, { "epoch": 1.0118338557993731, "grad_norm": 3.8037238121032715, "learning_rate": 8.518923718110562e-06, "loss": 1.9056, "step": 12911 }, { "epoch": 1.0119122257053292, "grad_norm": 2.4688878059387207, "learning_rate": 8.518704646915738e-06, "loss": 1.4469, "step": 12912 }, { "epoch": 1.0119905956112853, "grad_norm": 12.220754623413086, "learning_rate": 8.518485562337571e-06, "loss": 1.5166, "step": 12913 }, { "epoch": 1.0120689655172415, "grad_norm": 1.8697328567504883, "learning_rate": 8.51826646437689e-06, "loss": 1.5419, "step": 12914 }, { "epoch": 1.0121473354231976, "grad_norm": 3.648228645324707, "learning_rate": 8.518047353034532e-06, "loss": 1.3145, "step": 12915 }, { "epoch": 1.0122257053291537, "grad_norm": 2.757167100906372, "learning_rate": 8.517828228311327e-06, "loss": 1.4173, "step": 12916 }, { "epoch": 1.0123040752351098, "grad_norm": 3.336040735244751, "learning_rate": 8.517609090208112e-06, "loss": 1.3814, "step": 12917 }, { "epoch": 1.012382445141066, "grad_norm": 4.521623134613037, "learning_rate": 8.51738993872572e-06, "loss": 1.3471, "step": 12918 }, { "epoch": 1.012460815047022, "grad_norm": 2.2742888927459717, "learning_rate": 8.517170773864981e-06, "loss": 1.2705, "step": 12919 }, { "epoch": 1.0125391849529781, "grad_norm": 3.243018627166748, "learning_rate": 8.51695159562673e-06, "loss": 0.8904, "step": 12920 }, { "epoch": 1.0126175548589342, "grad_norm": 2.448795795440674, "learning_rate": 8.516732404011805e-06, "loss": 0.892, "step": 12921 }, { "epoch": 1.0126959247648903, "grad_norm": 4.278282165527344, "learning_rate": 8.516513199021034e-06, "loss": 1.8343, "step": 12922 }, { "epoch": 1.0127742946708465, "grad_norm": 3.5081946849823, "learning_rate": 8.516293980655252e-06, "loss": 1.1916, "step": 12923 }, { "epoch": 1.0128526645768026, "grad_norm": 2.897698402404785, "learning_rate": 8.516074748915297e-06, "loss": 1.2097, "step": 12924 }, { "epoch": 1.0129310344827587, "grad_norm": 2.7614948749542236, "learning_rate": 8.515855503801996e-06, "loss": 1.6391, "step": 12925 }, { "epoch": 1.0130094043887148, "grad_norm": 3.4945011138916016, "learning_rate": 8.515636245316188e-06, "loss": 1.9843, "step": 12926 }, { "epoch": 1.013087774294671, "grad_norm": 2.655365228652954, "learning_rate": 8.515416973458705e-06, "loss": 1.0303, "step": 12927 }, { "epoch": 1.013166144200627, "grad_norm": 3.6749415397644043, "learning_rate": 8.515197688230383e-06, "loss": 1.2641, "step": 12928 }, { "epoch": 1.0132445141065831, "grad_norm": 2.3358232975006104, "learning_rate": 8.514978389632054e-06, "loss": 1.509, "step": 12929 }, { "epoch": 1.0133228840125392, "grad_norm": 1.954274296760559, "learning_rate": 8.514759077664552e-06, "loss": 1.2727, "step": 12930 }, { "epoch": 1.0134012539184953, "grad_norm": 3.000075101852417, "learning_rate": 8.514539752328712e-06, "loss": 1.2674, "step": 12931 }, { "epoch": 1.0134796238244514, "grad_norm": 2.911184787750244, "learning_rate": 8.514320413625367e-06, "loss": 1.3397, "step": 12932 }, { "epoch": 1.0135579937304076, "grad_norm": 2.7244129180908203, "learning_rate": 8.514101061555356e-06, "loss": 1.1828, "step": 12933 }, { "epoch": 1.0136363636363637, "grad_norm": 5.271502494812012, "learning_rate": 8.513881696119506e-06, "loss": 1.6168, "step": 12934 }, { "epoch": 1.0137147335423198, "grad_norm": 3.3172526359558105, "learning_rate": 8.513662317318656e-06, "loss": 1.2454, "step": 12935 }, { "epoch": 1.013793103448276, "grad_norm": 2.047152519226074, "learning_rate": 8.513442925153638e-06, "loss": 1.1235, "step": 12936 }, { "epoch": 1.013871473354232, "grad_norm": 2.5017807483673096, "learning_rate": 8.513223519625289e-06, "loss": 1.2494, "step": 12937 }, { "epoch": 1.013949843260188, "grad_norm": 4.38839054107666, "learning_rate": 8.51300410073444e-06, "loss": 1.2671, "step": 12938 }, { "epoch": 1.0140282131661442, "grad_norm": 4.835701942443848, "learning_rate": 8.512784668481931e-06, "loss": 1.5079, "step": 12939 }, { "epoch": 1.0141065830721003, "grad_norm": 1.7312184572219849, "learning_rate": 8.512565222868592e-06, "loss": 1.4755, "step": 12940 }, { "epoch": 1.0141849529780564, "grad_norm": 1.955987811088562, "learning_rate": 8.51234576389526e-06, "loss": 1.2251, "step": 12941 }, { "epoch": 1.0142633228840126, "grad_norm": 3.4617667198181152, "learning_rate": 8.512126291562766e-06, "loss": 1.5185, "step": 12942 }, { "epoch": 1.0143416927899687, "grad_norm": 1.948965311050415, "learning_rate": 8.51190680587195e-06, "loss": 1.2567, "step": 12943 }, { "epoch": 1.0144200626959248, "grad_norm": 3.2297110557556152, "learning_rate": 8.511687306823644e-06, "loss": 1.2973, "step": 12944 }, { "epoch": 1.0144984326018809, "grad_norm": 4.730377197265625, "learning_rate": 8.511467794418684e-06, "loss": 1.7566, "step": 12945 }, { "epoch": 1.014576802507837, "grad_norm": 3.0542056560516357, "learning_rate": 8.5112482686579e-06, "loss": 1.5511, "step": 12946 }, { "epoch": 1.014655172413793, "grad_norm": 4.731690406799316, "learning_rate": 8.511028729542137e-06, "loss": 1.4077, "step": 12947 }, { "epoch": 1.0147335423197492, "grad_norm": 3.8159148693084717, "learning_rate": 8.510809177072222e-06, "loss": 1.5752, "step": 12948 }, { "epoch": 1.0148119122257053, "grad_norm": 2.2237401008605957, "learning_rate": 8.51058961124899e-06, "loss": 1.3234, "step": 12949 }, { "epoch": 1.0148902821316614, "grad_norm": 4.128610610961914, "learning_rate": 8.510370032073279e-06, "loss": 1.6518, "step": 12950 }, { "epoch": 1.0149686520376175, "grad_norm": 2.4394707679748535, "learning_rate": 8.510150439545925e-06, "loss": 1.125, "step": 12951 }, { "epoch": 1.0150470219435737, "grad_norm": 3.7617645263671875, "learning_rate": 8.509930833667762e-06, "loss": 1.7807, "step": 12952 }, { "epoch": 1.0151253918495298, "grad_norm": 2.763890027999878, "learning_rate": 8.509711214439622e-06, "loss": 1.6215, "step": 12953 }, { "epoch": 1.0152037617554859, "grad_norm": 4.738397121429443, "learning_rate": 8.509491581862345e-06, "loss": 1.9122, "step": 12954 }, { "epoch": 1.015282131661442, "grad_norm": 2.9979043006896973, "learning_rate": 8.509271935936765e-06, "loss": 1.3508, "step": 12955 }, { "epoch": 1.015360501567398, "grad_norm": 4.3720879554748535, "learning_rate": 8.509052276663717e-06, "loss": 1.558, "step": 12956 }, { "epoch": 1.0154388714733542, "grad_norm": 2.683040142059326, "learning_rate": 8.508832604044035e-06, "loss": 1.3213, "step": 12957 }, { "epoch": 1.0155172413793103, "grad_norm": 3.196303606033325, "learning_rate": 8.508612918078557e-06, "loss": 1.3101, "step": 12958 }, { "epoch": 1.0155956112852664, "grad_norm": 5.735143661499023, "learning_rate": 8.508393218768118e-06, "loss": 1.3225, "step": 12959 }, { "epoch": 1.0156739811912225, "grad_norm": 2.6210296154022217, "learning_rate": 8.508173506113554e-06, "loss": 1.2375, "step": 12960 }, { "epoch": 1.0157523510971787, "grad_norm": 8.32910442352295, "learning_rate": 8.507953780115698e-06, "loss": 0.8476, "step": 12961 }, { "epoch": 1.0158307210031348, "grad_norm": 2.2491095066070557, "learning_rate": 8.507734040775388e-06, "loss": 1.4057, "step": 12962 }, { "epoch": 1.0159090909090909, "grad_norm": 2.8520591259002686, "learning_rate": 8.50751428809346e-06, "loss": 1.153, "step": 12963 }, { "epoch": 1.015987460815047, "grad_norm": 1.7296258211135864, "learning_rate": 8.507294522070749e-06, "loss": 1.2976, "step": 12964 }, { "epoch": 1.016065830721003, "grad_norm": 3.922119140625, "learning_rate": 8.507074742708091e-06, "loss": 1.5618, "step": 12965 }, { "epoch": 1.0161442006269592, "grad_norm": 2.8487086296081543, "learning_rate": 8.506854950006323e-06, "loss": 1.3932, "step": 12966 }, { "epoch": 1.0162225705329153, "grad_norm": 2.7390153408050537, "learning_rate": 8.506635143966278e-06, "loss": 1.0973, "step": 12967 }, { "epoch": 1.0163009404388714, "grad_norm": 4.046895503997803, "learning_rate": 8.506415324588795e-06, "loss": 1.6806, "step": 12968 }, { "epoch": 1.0163793103448275, "grad_norm": 3.023407459259033, "learning_rate": 8.506195491874709e-06, "loss": 1.652, "step": 12969 }, { "epoch": 1.0164576802507836, "grad_norm": 3.1787869930267334, "learning_rate": 8.505975645824858e-06, "loss": 1.4751, "step": 12970 }, { "epoch": 1.0165360501567398, "grad_norm": 2.3077516555786133, "learning_rate": 8.505755786440074e-06, "loss": 1.4929, "step": 12971 }, { "epoch": 1.0166144200626959, "grad_norm": 2.7630302906036377, "learning_rate": 8.505535913721197e-06, "loss": 1.2972, "step": 12972 }, { "epoch": 1.016692789968652, "grad_norm": 5.220200061798096, "learning_rate": 8.505316027669062e-06, "loss": 1.1663, "step": 12973 }, { "epoch": 1.016771159874608, "grad_norm": 4.17115592956543, "learning_rate": 8.505096128284505e-06, "loss": 1.3185, "step": 12974 }, { "epoch": 1.0168495297805642, "grad_norm": 1.7900241613388062, "learning_rate": 8.504876215568362e-06, "loss": 1.1191, "step": 12975 }, { "epoch": 1.0169278996865203, "grad_norm": 7.975823402404785, "learning_rate": 8.504656289521472e-06, "loss": 0.9765, "step": 12976 }, { "epoch": 1.0170062695924764, "grad_norm": 12.108115196228027, "learning_rate": 8.504436350144667e-06, "loss": 2.319, "step": 12977 }, { "epoch": 1.0170846394984325, "grad_norm": 2.266309976577759, "learning_rate": 8.504216397438787e-06, "loss": 1.5106, "step": 12978 }, { "epoch": 1.0171630094043886, "grad_norm": 1.664834976196289, "learning_rate": 8.503996431404667e-06, "loss": 1.4994, "step": 12979 }, { "epoch": 1.0172413793103448, "grad_norm": 6.94889497756958, "learning_rate": 8.503776452043147e-06, "loss": 1.4526, "step": 12980 }, { "epoch": 1.0173197492163009, "grad_norm": 3.942714214324951, "learning_rate": 8.503556459355058e-06, "loss": 1.4859, "step": 12981 }, { "epoch": 1.017398119122257, "grad_norm": 2.591731548309326, "learning_rate": 8.503336453341243e-06, "loss": 1.0729, "step": 12982 }, { "epoch": 1.017476489028213, "grad_norm": 4.264618396759033, "learning_rate": 8.503116434002532e-06, "loss": 1.0789, "step": 12983 }, { "epoch": 1.0175548589341692, "grad_norm": 2.1445226669311523, "learning_rate": 8.502896401339767e-06, "loss": 1.2963, "step": 12984 }, { "epoch": 1.0176332288401253, "grad_norm": 3.0818843841552734, "learning_rate": 8.502676355353783e-06, "loss": 1.4118, "step": 12985 }, { "epoch": 1.0177115987460814, "grad_norm": 2.8416662216186523, "learning_rate": 8.502456296045418e-06, "loss": 1.3638, "step": 12986 }, { "epoch": 1.0177899686520375, "grad_norm": 5.127749919891357, "learning_rate": 8.502236223415507e-06, "loss": 1.5305, "step": 12987 }, { "epoch": 1.0178683385579936, "grad_norm": 10.447790145874023, "learning_rate": 8.50201613746489e-06, "loss": 0.9341, "step": 12988 }, { "epoch": 1.0179467084639497, "grad_norm": 2.7079622745513916, "learning_rate": 8.501796038194402e-06, "loss": 1.0742, "step": 12989 }, { "epoch": 1.0180250783699059, "grad_norm": 3.79653000831604, "learning_rate": 8.50157592560488e-06, "loss": 1.3942, "step": 12990 }, { "epoch": 1.018103448275862, "grad_norm": 6.098392963409424, "learning_rate": 8.501355799697161e-06, "loss": 0.971, "step": 12991 }, { "epoch": 1.018181818181818, "grad_norm": 2.871561050415039, "learning_rate": 8.501135660472084e-06, "loss": 1.191, "step": 12992 }, { "epoch": 1.0182601880877742, "grad_norm": 3.2159714698791504, "learning_rate": 8.500915507930486e-06, "loss": 1.5081, "step": 12993 }, { "epoch": 1.0183385579937303, "grad_norm": 2.4827065467834473, "learning_rate": 8.500695342073201e-06, "loss": 1.299, "step": 12994 }, { "epoch": 1.0184169278996866, "grad_norm": 2.8331122398376465, "learning_rate": 8.500475162901072e-06, "loss": 1.3848, "step": 12995 }, { "epoch": 1.0184952978056427, "grad_norm": 1.7854934930801392, "learning_rate": 8.500254970414932e-06, "loss": 1.3968, "step": 12996 }, { "epoch": 1.0185736677115989, "grad_norm": 2.0690271854400635, "learning_rate": 8.50003476461562e-06, "loss": 1.2617, "step": 12997 }, { "epoch": 1.018652037617555, "grad_norm": 2.8080050945281982, "learning_rate": 8.499814545503973e-06, "loss": 1.2607, "step": 12998 }, { "epoch": 1.018730407523511, "grad_norm": 2.6034958362579346, "learning_rate": 8.499594313080827e-06, "loss": 1.628, "step": 12999 }, { "epoch": 1.0188087774294672, "grad_norm": 2.3515515327453613, "learning_rate": 8.499374067347026e-06, "loss": 1.3804, "step": 13000 }, { "epoch": 1.0188871473354233, "grad_norm": 2.385056972503662, "learning_rate": 8.499153808303402e-06, "loss": 1.3553, "step": 13001 }, { "epoch": 1.0189655172413794, "grad_norm": 2.795081377029419, "learning_rate": 8.498933535950793e-06, "loss": 1.471, "step": 13002 }, { "epoch": 1.0190438871473355, "grad_norm": 4.152731895446777, "learning_rate": 8.498713250290038e-06, "loss": 1.7305, "step": 13003 }, { "epoch": 1.0191222570532916, "grad_norm": 2.6200976371765137, "learning_rate": 8.498492951321978e-06, "loss": 1.5095, "step": 13004 }, { "epoch": 1.0192006269592477, "grad_norm": 2.773066520690918, "learning_rate": 8.498272639047444e-06, "loss": 1.642, "step": 13005 }, { "epoch": 1.0192789968652038, "grad_norm": 5.329742431640625, "learning_rate": 8.49805231346728e-06, "loss": 1.7302, "step": 13006 }, { "epoch": 1.01935736677116, "grad_norm": 7.470069885253906, "learning_rate": 8.49783197458232e-06, "loss": 0.4526, "step": 13007 }, { "epoch": 1.019435736677116, "grad_norm": 5.917247295379639, "learning_rate": 8.497611622393404e-06, "loss": 1.6526, "step": 13008 }, { "epoch": 1.0195141065830722, "grad_norm": 2.7504947185516357, "learning_rate": 8.49739125690137e-06, "loss": 1.1832, "step": 13009 }, { "epoch": 1.0195924764890283, "grad_norm": 3.3879284858703613, "learning_rate": 8.497170878107058e-06, "loss": 0.7446, "step": 13010 }, { "epoch": 1.0196708463949844, "grad_norm": 2.443167209625244, "learning_rate": 8.496950486011301e-06, "loss": 1.2392, "step": 13011 }, { "epoch": 1.0197492163009405, "grad_norm": 2.1996774673461914, "learning_rate": 8.496730080614943e-06, "loss": 1.1194, "step": 13012 }, { "epoch": 1.0198275862068966, "grad_norm": 4.86698055267334, "learning_rate": 8.49650966191882e-06, "loss": 1.3927, "step": 13013 }, { "epoch": 1.0199059561128527, "grad_norm": 2.4106576442718506, "learning_rate": 8.496289229923768e-06, "loss": 1.4618, "step": 13014 }, { "epoch": 1.0199843260188088, "grad_norm": 2.2325756549835205, "learning_rate": 8.49606878463063e-06, "loss": 1.3579, "step": 13015 }, { "epoch": 1.020062695924765, "grad_norm": 3.753737211227417, "learning_rate": 8.495848326040241e-06, "loss": 1.3472, "step": 13016 }, { "epoch": 1.020141065830721, "grad_norm": 2.489016532897949, "learning_rate": 8.49562785415344e-06, "loss": 1.4536, "step": 13017 }, { "epoch": 1.0202194357366772, "grad_norm": 2.696202278137207, "learning_rate": 8.495407368971066e-06, "loss": 1.4005, "step": 13018 }, { "epoch": 1.0202978056426333, "grad_norm": 20.63490104675293, "learning_rate": 8.495186870493959e-06, "loss": 2.0934, "step": 13019 }, { "epoch": 1.0203761755485894, "grad_norm": 4.046456813812256, "learning_rate": 8.494966358722954e-06, "loss": 1.4658, "step": 13020 }, { "epoch": 1.0204545454545455, "grad_norm": 2.9957613945007324, "learning_rate": 8.494745833658895e-06, "loss": 1.4841, "step": 13021 }, { "epoch": 1.0205329153605016, "grad_norm": 3.6484251022338867, "learning_rate": 8.494525295302616e-06, "loss": 1.2267, "step": 13022 }, { "epoch": 1.0206112852664577, "grad_norm": 2.1014437675476074, "learning_rate": 8.494304743654958e-06, "loss": 1.5226, "step": 13023 }, { "epoch": 1.0206896551724138, "grad_norm": 2.9325101375579834, "learning_rate": 8.494084178716759e-06, "loss": 1.1734, "step": 13024 }, { "epoch": 1.02076802507837, "grad_norm": 2.6455790996551514, "learning_rate": 8.49386360048886e-06, "loss": 1.3036, "step": 13025 }, { "epoch": 1.020846394984326, "grad_norm": 2.7148735523223877, "learning_rate": 8.493643008972097e-06, "loss": 1.4455, "step": 13026 }, { "epoch": 1.0209247648902822, "grad_norm": 5.413984775543213, "learning_rate": 8.493422404167311e-06, "loss": 1.2475, "step": 13027 }, { "epoch": 1.0210031347962383, "grad_norm": 3.255486488342285, "learning_rate": 8.493201786075342e-06, "loss": 1.7561, "step": 13028 }, { "epoch": 1.0210815047021944, "grad_norm": 2.7645351886749268, "learning_rate": 8.492981154697027e-06, "loss": 0.8825, "step": 13029 }, { "epoch": 1.0211598746081505, "grad_norm": 4.171568393707275, "learning_rate": 8.492760510033203e-06, "loss": 1.715, "step": 13030 }, { "epoch": 1.0212382445141066, "grad_norm": 4.081132411956787, "learning_rate": 8.492539852084715e-06, "loss": 1.4023, "step": 13031 }, { "epoch": 1.0213166144200627, "grad_norm": 2.121393918991089, "learning_rate": 8.492319180852397e-06, "loss": 1.7339, "step": 13032 }, { "epoch": 1.0213949843260188, "grad_norm": 5.168642044067383, "learning_rate": 8.492098496337093e-06, "loss": 1.965, "step": 13033 }, { "epoch": 1.021473354231975, "grad_norm": 3.2387876510620117, "learning_rate": 8.491877798539638e-06, "loss": 0.8944, "step": 13034 }, { "epoch": 1.021551724137931, "grad_norm": 2.138096570968628, "learning_rate": 8.491657087460874e-06, "loss": 1.1191, "step": 13035 }, { "epoch": 1.0216300940438872, "grad_norm": 3.106729507446289, "learning_rate": 8.491436363101642e-06, "loss": 1.3255, "step": 13036 }, { "epoch": 1.0217084639498433, "grad_norm": 2.2660555839538574, "learning_rate": 8.491215625462776e-06, "loss": 1.1721, "step": 13037 }, { "epoch": 1.0217868338557994, "grad_norm": 2.784590721130371, "learning_rate": 8.490994874545122e-06, "loss": 1.2458, "step": 13038 }, { "epoch": 1.0218652037617555, "grad_norm": 2.095233678817749, "learning_rate": 8.490774110349513e-06, "loss": 1.4153, "step": 13039 }, { "epoch": 1.0219435736677116, "grad_norm": 2.2864832878112793, "learning_rate": 8.490553332876797e-06, "loss": 1.3381, "step": 13040 }, { "epoch": 1.0220219435736677, "grad_norm": 2.541813611984253, "learning_rate": 8.490332542127805e-06, "loss": 1.3251, "step": 13041 }, { "epoch": 1.0221003134796238, "grad_norm": 2.0936973094940186, "learning_rate": 8.490111738103382e-06, "loss": 1.3597, "step": 13042 }, { "epoch": 1.02217868338558, "grad_norm": 6.976686954498291, "learning_rate": 8.489890920804366e-06, "loss": 1.1473, "step": 13043 }, { "epoch": 1.022257053291536, "grad_norm": 2.9297780990600586, "learning_rate": 8.489670090231598e-06, "loss": 1.2673, "step": 13044 }, { "epoch": 1.0223354231974922, "grad_norm": 2.1408169269561768, "learning_rate": 8.489449246385919e-06, "loss": 1.2425, "step": 13045 }, { "epoch": 1.0224137931034483, "grad_norm": 4.184651851654053, "learning_rate": 8.489228389268165e-06, "loss": 1.1299, "step": 13046 }, { "epoch": 1.0224921630094044, "grad_norm": 2.0302257537841797, "learning_rate": 8.48900751887918e-06, "loss": 1.3314, "step": 13047 }, { "epoch": 1.0225705329153605, "grad_norm": 2.853036880493164, "learning_rate": 8.488786635219802e-06, "loss": 1.0988, "step": 13048 }, { "epoch": 1.0226489028213166, "grad_norm": 2.3712759017944336, "learning_rate": 8.48856573829087e-06, "loss": 1.0043, "step": 13049 }, { "epoch": 1.0227272727272727, "grad_norm": 9.5169095993042, "learning_rate": 8.488344828093228e-06, "loss": 1.4617, "step": 13050 }, { "epoch": 1.0228056426332288, "grad_norm": 4.526905059814453, "learning_rate": 8.488123904627714e-06, "loss": 1.291, "step": 13051 }, { "epoch": 1.022884012539185, "grad_norm": 2.5291194915771484, "learning_rate": 8.487902967895165e-06, "loss": 1.4335, "step": 13052 }, { "epoch": 1.022962382445141, "grad_norm": 2.600583076477051, "learning_rate": 8.487682017896428e-06, "loss": 1.2791, "step": 13053 }, { "epoch": 1.0230407523510971, "grad_norm": 3.432241439819336, "learning_rate": 8.48746105463234e-06, "loss": 1.6406, "step": 13054 }, { "epoch": 1.0231191222570533, "grad_norm": 2.6079537868499756, "learning_rate": 8.48724007810374e-06, "loss": 1.2229, "step": 13055 }, { "epoch": 1.0231974921630094, "grad_norm": 3.5413196086883545, "learning_rate": 8.48701908831147e-06, "loss": 1.7147, "step": 13056 }, { "epoch": 1.0232758620689655, "grad_norm": 2.3840951919555664, "learning_rate": 8.486798085256371e-06, "loss": 1.5752, "step": 13057 }, { "epoch": 1.0233542319749216, "grad_norm": 2.2058348655700684, "learning_rate": 8.486577068939281e-06, "loss": 0.9414, "step": 13058 }, { "epoch": 1.0234326018808777, "grad_norm": 2.1174795627593994, "learning_rate": 8.486356039361045e-06, "loss": 1.3128, "step": 13059 }, { "epoch": 1.0235109717868338, "grad_norm": 3.5817759037017822, "learning_rate": 8.486134996522502e-06, "loss": 1.8057, "step": 13060 }, { "epoch": 1.02358934169279, "grad_norm": 3.827462673187256, "learning_rate": 8.485913940424491e-06, "loss": 1.4308, "step": 13061 }, { "epoch": 1.023667711598746, "grad_norm": 3.4693844318389893, "learning_rate": 8.485692871067853e-06, "loss": 1.3089, "step": 13062 }, { "epoch": 1.0237460815047021, "grad_norm": 2.2551345825195312, "learning_rate": 8.48547178845343e-06, "loss": 1.2179, "step": 13063 }, { "epoch": 1.0238244514106583, "grad_norm": 5.171680450439453, "learning_rate": 8.485250692582063e-06, "loss": 0.8423, "step": 13064 }, { "epoch": 1.0239028213166144, "grad_norm": 2.3684444427490234, "learning_rate": 8.485029583454592e-06, "loss": 1.5833, "step": 13065 }, { "epoch": 1.0239811912225705, "grad_norm": 2.1641244888305664, "learning_rate": 8.48480846107186e-06, "loss": 1.1086, "step": 13066 }, { "epoch": 1.0240595611285266, "grad_norm": 2.2884490489959717, "learning_rate": 8.484587325434706e-06, "loss": 1.2965, "step": 13067 }, { "epoch": 1.0241379310344827, "grad_norm": 2.837189197540283, "learning_rate": 8.48436617654397e-06, "loss": 1.5975, "step": 13068 }, { "epoch": 1.0242163009404388, "grad_norm": 4.3813347816467285, "learning_rate": 8.484145014400496e-06, "loss": 1.4129, "step": 13069 }, { "epoch": 1.024294670846395, "grad_norm": 3.337296485900879, "learning_rate": 8.483923839005125e-06, "loss": 1.9218, "step": 13070 }, { "epoch": 1.024373040752351, "grad_norm": 2.8439881801605225, "learning_rate": 8.483702650358694e-06, "loss": 1.2012, "step": 13071 }, { "epoch": 1.0244514106583071, "grad_norm": 3.1341516971588135, "learning_rate": 8.48348144846205e-06, "loss": 1.2937, "step": 13072 }, { "epoch": 1.0245297805642632, "grad_norm": 4.815248012542725, "learning_rate": 8.483260233316031e-06, "loss": 1.2206, "step": 13073 }, { "epoch": 1.0246081504702194, "grad_norm": 2.4116718769073486, "learning_rate": 8.483039004921478e-06, "loss": 1.5582, "step": 13074 }, { "epoch": 1.0246865203761755, "grad_norm": 3.505514621734619, "learning_rate": 8.482817763279237e-06, "loss": 1.6075, "step": 13075 }, { "epoch": 1.0247648902821316, "grad_norm": 4.29448938369751, "learning_rate": 8.482596508390144e-06, "loss": 1.8958, "step": 13076 }, { "epoch": 1.0248432601880877, "grad_norm": 3.247809886932373, "learning_rate": 8.482375240255043e-06, "loss": 1.2459, "step": 13077 }, { "epoch": 1.0249216300940438, "grad_norm": 3.737745523452759, "learning_rate": 8.482153958874774e-06, "loss": 1.6606, "step": 13078 }, { "epoch": 1.025, "grad_norm": 2.156136989593506, "learning_rate": 8.481932664250181e-06, "loss": 0.7047, "step": 13079 }, { "epoch": 1.025078369905956, "grad_norm": 5.8963189125061035, "learning_rate": 8.481711356382105e-06, "loss": 1.5026, "step": 13080 }, { "epoch": 1.0251567398119121, "grad_norm": 4.111942291259766, "learning_rate": 8.481490035271386e-06, "loss": 1.2015, "step": 13081 }, { "epoch": 1.0252351097178682, "grad_norm": 3.691237211227417, "learning_rate": 8.481268700918868e-06, "loss": 1.5618, "step": 13082 }, { "epoch": 1.0253134796238244, "grad_norm": 2.645742654800415, "learning_rate": 8.481047353325392e-06, "loss": 1.234, "step": 13083 }, { "epoch": 1.0253918495297805, "grad_norm": 3.2275826930999756, "learning_rate": 8.4808259924918e-06, "loss": 1.4709, "step": 13084 }, { "epoch": 1.0254702194357366, "grad_norm": 2.2791261672973633, "learning_rate": 8.480604618418933e-06, "loss": 1.4572, "step": 13085 }, { "epoch": 1.0255485893416927, "grad_norm": 4.688914775848389, "learning_rate": 8.480383231107633e-06, "loss": 0.7694, "step": 13086 }, { "epoch": 1.0256269592476488, "grad_norm": 5.291653633117676, "learning_rate": 8.480161830558745e-06, "loss": 1.9678, "step": 13087 }, { "epoch": 1.0257053291536051, "grad_norm": 2.919802665710449, "learning_rate": 8.47994041677311e-06, "loss": 1.2081, "step": 13088 }, { "epoch": 1.025783699059561, "grad_norm": 3.0749616622924805, "learning_rate": 8.479718989751567e-06, "loss": 1.8443, "step": 13089 }, { "epoch": 1.0258620689655173, "grad_norm": 6.8709187507629395, "learning_rate": 8.47949754949496e-06, "loss": 1.4949, "step": 13090 }, { "epoch": 1.0259404388714735, "grad_norm": 4.565186023712158, "learning_rate": 8.47927609600413e-06, "loss": 1.0887, "step": 13091 }, { "epoch": 1.0260188087774296, "grad_norm": 3.452176570892334, "learning_rate": 8.479054629279924e-06, "loss": 1.3123, "step": 13092 }, { "epoch": 1.0260971786833857, "grad_norm": 2.6845853328704834, "learning_rate": 8.478833149323179e-06, "loss": 1.5602, "step": 13093 }, { "epoch": 1.0261755485893418, "grad_norm": 6.006680965423584, "learning_rate": 8.47861165613474e-06, "loss": 0.6513, "step": 13094 }, { "epoch": 1.026253918495298, "grad_norm": 2.788881301879883, "learning_rate": 8.47839014971545e-06, "loss": 1.2809, "step": 13095 }, { "epoch": 1.026332288401254, "grad_norm": 3.8842415809631348, "learning_rate": 8.478168630066151e-06, "loss": 1.5568, "step": 13096 }, { "epoch": 1.0264106583072101, "grad_norm": 3.162940502166748, "learning_rate": 8.477947097187682e-06, "loss": 1.5572, "step": 13097 }, { "epoch": 1.0264890282131662, "grad_norm": 2.101842164993286, "learning_rate": 8.477725551080891e-06, "loss": 1.423, "step": 13098 }, { "epoch": 1.0265673981191223, "grad_norm": 3.274608850479126, "learning_rate": 8.477503991746617e-06, "loss": 1.3208, "step": 13099 }, { "epoch": 1.0266457680250785, "grad_norm": 2.1892077922821045, "learning_rate": 8.477282419185704e-06, "loss": 1.4986, "step": 13100 }, { "epoch": 1.0267241379310346, "grad_norm": 2.3729608058929443, "learning_rate": 8.477060833398996e-06, "loss": 1.1044, "step": 13101 }, { "epoch": 1.0268025078369907, "grad_norm": 8.106757164001465, "learning_rate": 8.476839234387335e-06, "loss": 1.3208, "step": 13102 }, { "epoch": 1.0268808777429468, "grad_norm": 2.7928097248077393, "learning_rate": 8.476617622151562e-06, "loss": 1.4945, "step": 13103 }, { "epoch": 1.026959247648903, "grad_norm": 3.4301602840423584, "learning_rate": 8.47639599669252e-06, "loss": 1.3577, "step": 13104 }, { "epoch": 1.027037617554859, "grad_norm": 2.2194504737854004, "learning_rate": 8.476174358011056e-06, "loss": 1.2709, "step": 13105 }, { "epoch": 1.0271159874608151, "grad_norm": 2.363543748855591, "learning_rate": 8.475952706108009e-06, "loss": 1.3417, "step": 13106 }, { "epoch": 1.0271943573667712, "grad_norm": 11.756484985351562, "learning_rate": 8.475731040984223e-06, "loss": 1.4123, "step": 13107 }, { "epoch": 1.0272727272727273, "grad_norm": 4.523585796356201, "learning_rate": 8.475509362640541e-06, "loss": 1.2737, "step": 13108 }, { "epoch": 1.0273510971786834, "grad_norm": 3.9594364166259766, "learning_rate": 8.475287671077808e-06, "loss": 1.0759, "step": 13109 }, { "epoch": 1.0274294670846396, "grad_norm": 3.5558013916015625, "learning_rate": 8.475065966296866e-06, "loss": 1.2108, "step": 13110 }, { "epoch": 1.0275078369905957, "grad_norm": 3.72145676612854, "learning_rate": 8.474844248298556e-06, "loss": 1.0053, "step": 13111 }, { "epoch": 1.0275862068965518, "grad_norm": 3.0308189392089844, "learning_rate": 8.474622517083725e-06, "loss": 1.6632, "step": 13112 }, { "epoch": 1.027664576802508, "grad_norm": 3.4171481132507324, "learning_rate": 8.474400772653214e-06, "loss": 1.342, "step": 13113 }, { "epoch": 1.027742946708464, "grad_norm": 2.8787741661071777, "learning_rate": 8.474179015007865e-06, "loss": 1.9346, "step": 13114 }, { "epoch": 1.02782131661442, "grad_norm": 2.8263165950775146, "learning_rate": 8.473957244148526e-06, "loss": 1.5492, "step": 13115 }, { "epoch": 1.0278996865203762, "grad_norm": 3.4322521686553955, "learning_rate": 8.473735460076037e-06, "loss": 1.2735, "step": 13116 }, { "epoch": 1.0279780564263323, "grad_norm": 5.105262279510498, "learning_rate": 8.473513662791243e-06, "loss": 1.5037, "step": 13117 }, { "epoch": 1.0280564263322884, "grad_norm": 2.769575595855713, "learning_rate": 8.473291852294986e-06, "loss": 1.3992, "step": 13118 }, { "epoch": 1.0281347962382446, "grad_norm": 3.7095370292663574, "learning_rate": 8.473070028588112e-06, "loss": 1.477, "step": 13119 }, { "epoch": 1.0282131661442007, "grad_norm": 2.4768993854522705, "learning_rate": 8.472848191671465e-06, "loss": 1.2353, "step": 13120 }, { "epoch": 1.0282915360501568, "grad_norm": 2.3304038047790527, "learning_rate": 8.472626341545884e-06, "loss": 1.3654, "step": 13121 }, { "epoch": 1.0283699059561129, "grad_norm": 4.589145660400391, "learning_rate": 8.472404478212218e-06, "loss": 1.3891, "step": 13122 }, { "epoch": 1.028448275862069, "grad_norm": 2.2886884212493896, "learning_rate": 8.472182601671308e-06, "loss": 1.2674, "step": 13123 }, { "epoch": 1.028526645768025, "grad_norm": 2.3959789276123047, "learning_rate": 8.471960711924e-06, "loss": 1.195, "step": 13124 }, { "epoch": 1.0286050156739812, "grad_norm": 4.829338550567627, "learning_rate": 8.471738808971135e-06, "loss": 1.7074, "step": 13125 }, { "epoch": 1.0286833855799373, "grad_norm": 2.3840079307556152, "learning_rate": 8.47151689281356e-06, "loss": 0.8469, "step": 13126 }, { "epoch": 1.0287617554858934, "grad_norm": 2.1945254802703857, "learning_rate": 8.471294963452117e-06, "loss": 1.3084, "step": 13127 }, { "epoch": 1.0288401253918495, "grad_norm": 8.430405616760254, "learning_rate": 8.471073020887652e-06, "loss": 1.3476, "step": 13128 }, { "epoch": 1.0289184952978057, "grad_norm": 3.6130707263946533, "learning_rate": 8.47085106512101e-06, "loss": 0.925, "step": 13129 }, { "epoch": 1.0289968652037618, "grad_norm": 2.7210636138916016, "learning_rate": 8.47062909615303e-06, "loss": 1.5194, "step": 13130 }, { "epoch": 1.0290752351097179, "grad_norm": 2.237138509750366, "learning_rate": 8.47040711398456e-06, "loss": 1.4868, "step": 13131 }, { "epoch": 1.029153605015674, "grad_norm": 3.3593201637268066, "learning_rate": 8.470185118616444e-06, "loss": 1.6236, "step": 13132 }, { "epoch": 1.02923197492163, "grad_norm": 6.062631607055664, "learning_rate": 8.469963110049526e-06, "loss": 1.7474, "step": 13133 }, { "epoch": 1.0293103448275862, "grad_norm": 3.909909963607788, "learning_rate": 8.469741088284652e-06, "loss": 1.2428, "step": 13134 }, { "epoch": 1.0293887147335423, "grad_norm": 3.204425573348999, "learning_rate": 8.469519053322663e-06, "loss": 1.3138, "step": 13135 }, { "epoch": 1.0294670846394984, "grad_norm": 3.7764997482299805, "learning_rate": 8.469297005164408e-06, "loss": 1.2546, "step": 13136 }, { "epoch": 1.0295454545454545, "grad_norm": 2.1726338863372803, "learning_rate": 8.469074943810726e-06, "loss": 1.1652, "step": 13137 }, { "epoch": 1.0296238244514107, "grad_norm": 13.022903442382812, "learning_rate": 8.468852869262468e-06, "loss": 0.8635, "step": 13138 }, { "epoch": 1.0297021943573668, "grad_norm": 2.9414541721343994, "learning_rate": 8.468630781520474e-06, "loss": 1.3374, "step": 13139 }, { "epoch": 1.0297805642633229, "grad_norm": 3.7586677074432373, "learning_rate": 8.46840868058559e-06, "loss": 1.7141, "step": 13140 }, { "epoch": 1.029858934169279, "grad_norm": 2.9326248168945312, "learning_rate": 8.46818656645866e-06, "loss": 1.421, "step": 13141 }, { "epoch": 1.029937304075235, "grad_norm": 4.322661399841309, "learning_rate": 8.467964439140532e-06, "loss": 1.1028, "step": 13142 }, { "epoch": 1.0300156739811912, "grad_norm": 6.00874137878418, "learning_rate": 8.467742298632047e-06, "loss": 1.4425, "step": 13143 }, { "epoch": 1.0300940438871473, "grad_norm": 5.194835186004639, "learning_rate": 8.467520144934052e-06, "loss": 1.7982, "step": 13144 }, { "epoch": 1.0301724137931034, "grad_norm": 2.1071550846099854, "learning_rate": 8.46729797804739e-06, "loss": 1.1908, "step": 13145 }, { "epoch": 1.0302507836990595, "grad_norm": 2.4420297145843506, "learning_rate": 8.46707579797291e-06, "loss": 1.2127, "step": 13146 }, { "epoch": 1.0303291536050156, "grad_norm": 3.9570059776306152, "learning_rate": 8.466853604711451e-06, "loss": 1.3085, "step": 13147 }, { "epoch": 1.0304075235109718, "grad_norm": 10.57911491394043, "learning_rate": 8.466631398263864e-06, "loss": 1.0359, "step": 13148 }, { "epoch": 1.0304858934169279, "grad_norm": 6.1592020988464355, "learning_rate": 8.466409178630992e-06, "loss": 1.9666, "step": 13149 }, { "epoch": 1.030564263322884, "grad_norm": 3.5894744396209717, "learning_rate": 8.46618694581368e-06, "loss": 1.5394, "step": 13150 }, { "epoch": 1.03064263322884, "grad_norm": 1.9359883069992065, "learning_rate": 8.465964699812771e-06, "loss": 1.5825, "step": 13151 }, { "epoch": 1.0307210031347962, "grad_norm": 7.265305995941162, "learning_rate": 8.465742440629115e-06, "loss": 1.7747, "step": 13152 }, { "epoch": 1.0307993730407523, "grad_norm": 3.9555702209472656, "learning_rate": 8.465520168263553e-06, "loss": 1.4262, "step": 13153 }, { "epoch": 1.0308777429467084, "grad_norm": 2.654946804046631, "learning_rate": 8.465297882716933e-06, "loss": 1.1782, "step": 13154 }, { "epoch": 1.0309561128526645, "grad_norm": 4.234556198120117, "learning_rate": 8.465075583990099e-06, "loss": 1.4257, "step": 13155 }, { "epoch": 1.0310344827586206, "grad_norm": 1.909905195236206, "learning_rate": 8.464853272083899e-06, "loss": 1.5428, "step": 13156 }, { "epoch": 1.0311128526645768, "grad_norm": 2.3282668590545654, "learning_rate": 8.464630946999177e-06, "loss": 1.4487, "step": 13157 }, { "epoch": 1.0311912225705329, "grad_norm": 12.941886901855469, "learning_rate": 8.464408608736777e-06, "loss": 0.4592, "step": 13158 }, { "epoch": 1.031269592476489, "grad_norm": 4.186945915222168, "learning_rate": 8.464186257297545e-06, "loss": 1.4924, "step": 13159 }, { "epoch": 1.031347962382445, "grad_norm": 2.86547589302063, "learning_rate": 8.46396389268233e-06, "loss": 1.1407, "step": 13160 }, { "epoch": 1.0314263322884012, "grad_norm": 4.27647590637207, "learning_rate": 8.463741514891974e-06, "loss": 1.5062, "step": 13161 }, { "epoch": 1.0315047021943573, "grad_norm": 2.8126320838928223, "learning_rate": 8.463519123927325e-06, "loss": 1.5453, "step": 13162 }, { "epoch": 1.0315830721003134, "grad_norm": 2.749579429626465, "learning_rate": 8.463296719789228e-06, "loss": 1.3644, "step": 13163 }, { "epoch": 1.0316614420062695, "grad_norm": 3.639347553253174, "learning_rate": 8.46307430247853e-06, "loss": 1.3831, "step": 13164 }, { "epoch": 1.0317398119122256, "grad_norm": 2.528528928756714, "learning_rate": 8.462851871996076e-06, "loss": 1.2672, "step": 13165 }, { "epoch": 1.0318181818181817, "grad_norm": 3.5815536975860596, "learning_rate": 8.46262942834271e-06, "loss": 1.5423, "step": 13166 }, { "epoch": 1.0318965517241379, "grad_norm": 3.5467028617858887, "learning_rate": 8.462406971519284e-06, "loss": 1.1532, "step": 13167 }, { "epoch": 1.031974921630094, "grad_norm": 7.765077114105225, "learning_rate": 8.462184501526637e-06, "loss": 2.0515, "step": 13168 }, { "epoch": 1.03205329153605, "grad_norm": 4.992148399353027, "learning_rate": 8.46196201836562e-06, "loss": 1.314, "step": 13169 }, { "epoch": 1.0321316614420062, "grad_norm": 3.1792378425598145, "learning_rate": 8.461739522037079e-06, "loss": 1.535, "step": 13170 }, { "epoch": 1.0322100313479623, "grad_norm": 2.296168565750122, "learning_rate": 8.461517012541856e-06, "loss": 1.4447, "step": 13171 }, { "epoch": 1.0322884012539184, "grad_norm": 2.389047861099243, "learning_rate": 8.4612944898808e-06, "loss": 1.1882, "step": 13172 }, { "epoch": 1.0323667711598745, "grad_norm": 2.399946928024292, "learning_rate": 8.461071954054759e-06, "loss": 1.2905, "step": 13173 }, { "epoch": 1.0324451410658306, "grad_norm": 2.5001697540283203, "learning_rate": 8.46084940506458e-06, "loss": 1.6784, "step": 13174 }, { "epoch": 1.0325235109717867, "grad_norm": 5.325777530670166, "learning_rate": 8.460626842911103e-06, "loss": 1.5386, "step": 13175 }, { "epoch": 1.0326018808777429, "grad_norm": 5.579620361328125, "learning_rate": 8.460404267595183e-06, "loss": 1.6148, "step": 13176 }, { "epoch": 1.032680250783699, "grad_norm": 3.430945873260498, "learning_rate": 8.46018167911766e-06, "loss": 1.4138, "step": 13177 }, { "epoch": 1.032758620689655, "grad_norm": 3.7428553104400635, "learning_rate": 8.459959077479384e-06, "loss": 1.5551, "step": 13178 }, { "epoch": 1.0328369905956112, "grad_norm": 9.840651512145996, "learning_rate": 8.4597364626812e-06, "loss": 1.3464, "step": 13179 }, { "epoch": 1.0329153605015673, "grad_norm": 3.2776522636413574, "learning_rate": 8.459513834723957e-06, "loss": 1.2663, "step": 13180 }, { "epoch": 1.0329937304075236, "grad_norm": 3.8034260272979736, "learning_rate": 8.4592911936085e-06, "loss": 1.5785, "step": 13181 }, { "epoch": 1.0330721003134795, "grad_norm": 2.9848594665527344, "learning_rate": 8.459068539335677e-06, "loss": 1.5342, "step": 13182 }, { "epoch": 1.0331504702194358, "grad_norm": 2.064032793045044, "learning_rate": 8.458845871906331e-06, "loss": 1.3367, "step": 13183 }, { "epoch": 1.033228840125392, "grad_norm": 4.0402727127075195, "learning_rate": 8.458623191321316e-06, "loss": 1.3485, "step": 13184 }, { "epoch": 1.033307210031348, "grad_norm": 2.2831480503082275, "learning_rate": 8.458400497581472e-06, "loss": 1.0603, "step": 13185 }, { "epoch": 1.0333855799373042, "grad_norm": 3.8635735511779785, "learning_rate": 8.45817779068765e-06, "loss": 1.2672, "step": 13186 }, { "epoch": 1.0334639498432603, "grad_norm": 2.5054914951324463, "learning_rate": 8.457955070640696e-06, "loss": 1.2702, "step": 13187 }, { "epoch": 1.0335423197492164, "grad_norm": 4.262599468231201, "learning_rate": 8.457732337441454e-06, "loss": 0.9448, "step": 13188 }, { "epoch": 1.0336206896551725, "grad_norm": 11.337747573852539, "learning_rate": 8.457509591090778e-06, "loss": 0.638, "step": 13189 }, { "epoch": 1.0336990595611286, "grad_norm": 2.674621343612671, "learning_rate": 8.45728683158951e-06, "loss": 1.3127, "step": 13190 }, { "epoch": 1.0337774294670847, "grad_norm": 2.2364003658294678, "learning_rate": 8.457064058938498e-06, "loss": 1.4989, "step": 13191 }, { "epoch": 1.0338557993730408, "grad_norm": 3.1221816539764404, "learning_rate": 8.45684127313859e-06, "loss": 1.3297, "step": 13192 }, { "epoch": 1.033934169278997, "grad_norm": 3.3472397327423096, "learning_rate": 8.456618474190634e-06, "loss": 1.563, "step": 13193 }, { "epoch": 1.034012539184953, "grad_norm": 2.9278457164764404, "learning_rate": 8.456395662095477e-06, "loss": 1.5473, "step": 13194 }, { "epoch": 1.0340909090909092, "grad_norm": 2.416627883911133, "learning_rate": 8.456172836853967e-06, "loss": 1.4988, "step": 13195 }, { "epoch": 1.0341692789968653, "grad_norm": 3.5531985759735107, "learning_rate": 8.455949998466948e-06, "loss": 1.8706, "step": 13196 }, { "epoch": 1.0342476489028214, "grad_norm": 3.1544177532196045, "learning_rate": 8.455727146935273e-06, "loss": 1.6736, "step": 13197 }, { "epoch": 1.0343260188087775, "grad_norm": 5.643148422241211, "learning_rate": 8.455504282259786e-06, "loss": 1.6802, "step": 13198 }, { "epoch": 1.0344043887147336, "grad_norm": 3.961421251296997, "learning_rate": 8.455281404441336e-06, "loss": 1.7554, "step": 13199 }, { "epoch": 1.0344827586206897, "grad_norm": 3.5160958766937256, "learning_rate": 8.45505851348077e-06, "loss": 1.4084, "step": 13200 }, { "epoch": 1.0345611285266458, "grad_norm": 3.1599583625793457, "learning_rate": 8.454835609378934e-06, "loss": 1.2662, "step": 13201 }, { "epoch": 1.034639498432602, "grad_norm": 3.103299617767334, "learning_rate": 8.45461269213668e-06, "loss": 0.8947, "step": 13202 }, { "epoch": 1.034717868338558, "grad_norm": 3.340557336807251, "learning_rate": 8.454389761754852e-06, "loss": 1.3331, "step": 13203 }, { "epoch": 1.0347962382445142, "grad_norm": 2.699083089828491, "learning_rate": 8.454166818234303e-06, "loss": 1.4827, "step": 13204 }, { "epoch": 1.0348746081504703, "grad_norm": 2.478095531463623, "learning_rate": 8.453943861575876e-06, "loss": 1.3741, "step": 13205 }, { "epoch": 1.0349529780564264, "grad_norm": 3.3559160232543945, "learning_rate": 8.453720891780419e-06, "loss": 1.3981, "step": 13206 }, { "epoch": 1.0350313479623825, "grad_norm": 3.3561713695526123, "learning_rate": 8.453497908848783e-06, "loss": 1.2073, "step": 13207 }, { "epoch": 1.0351097178683386, "grad_norm": 4.555943965911865, "learning_rate": 8.453274912781814e-06, "loss": 1.4967, "step": 13208 }, { "epoch": 1.0351880877742947, "grad_norm": 1.9574586153030396, "learning_rate": 8.453051903580363e-06, "loss": 1.4988, "step": 13209 }, { "epoch": 1.0352664576802508, "grad_norm": 3.6520562171936035, "learning_rate": 8.452828881245273e-06, "loss": 1.0642, "step": 13210 }, { "epoch": 1.035344827586207, "grad_norm": 5.009136199951172, "learning_rate": 8.452605845777397e-06, "loss": 1.1402, "step": 13211 }, { "epoch": 1.035423197492163, "grad_norm": 3.3548333644866943, "learning_rate": 8.452382797177584e-06, "loss": 1.5169, "step": 13212 }, { "epoch": 1.0355015673981192, "grad_norm": 4.008763313293457, "learning_rate": 8.452159735446676e-06, "loss": 1.4957, "step": 13213 }, { "epoch": 1.0355799373040753, "grad_norm": 2.608321189880371, "learning_rate": 8.451936660585528e-06, "loss": 1.309, "step": 13214 }, { "epoch": 1.0356583072100314, "grad_norm": 4.014878273010254, "learning_rate": 8.451713572594987e-06, "loss": 1.2378, "step": 13215 }, { "epoch": 1.0357366771159875, "grad_norm": 2.8112986087799072, "learning_rate": 8.451490471475899e-06, "loss": 1.1093, "step": 13216 }, { "epoch": 1.0358150470219436, "grad_norm": 2.4779937267303467, "learning_rate": 8.451267357229113e-06, "loss": 1.8605, "step": 13217 }, { "epoch": 1.0358934169278997, "grad_norm": 5.135185718536377, "learning_rate": 8.451044229855478e-06, "loss": 1.5701, "step": 13218 }, { "epoch": 1.0359717868338558, "grad_norm": 3.040553092956543, "learning_rate": 8.450821089355846e-06, "loss": 1.7386, "step": 13219 }, { "epoch": 1.036050156739812, "grad_norm": 2.1628267765045166, "learning_rate": 8.450597935731061e-06, "loss": 1.4489, "step": 13220 }, { "epoch": 1.036128526645768, "grad_norm": 2.313045024871826, "learning_rate": 8.450374768981977e-06, "loss": 1.4906, "step": 13221 }, { "epoch": 1.0362068965517242, "grad_norm": 1.8747512102127075, "learning_rate": 8.450151589109436e-06, "loss": 1.517, "step": 13222 }, { "epoch": 1.0362852664576803, "grad_norm": 4.433048248291016, "learning_rate": 8.449928396114292e-06, "loss": 1.6812, "step": 13223 }, { "epoch": 1.0363636363636364, "grad_norm": 2.737337589263916, "learning_rate": 8.449705189997393e-06, "loss": 1.1411, "step": 13224 }, { "epoch": 1.0364420062695925, "grad_norm": 1.9657974243164062, "learning_rate": 8.449481970759586e-06, "loss": 1.2408, "step": 13225 }, { "epoch": 1.0365203761755486, "grad_norm": 2.6438565254211426, "learning_rate": 8.449258738401723e-06, "loss": 1.0734, "step": 13226 }, { "epoch": 1.0365987460815047, "grad_norm": 3.2771270275115967, "learning_rate": 8.44903549292465e-06, "loss": 1.6398, "step": 13227 }, { "epoch": 1.0366771159874608, "grad_norm": 2.7978098392486572, "learning_rate": 8.448812234329218e-06, "loss": 1.4573, "step": 13228 }, { "epoch": 1.036755485893417, "grad_norm": 2.3639254570007324, "learning_rate": 8.448588962616274e-06, "loss": 1.6378, "step": 13229 }, { "epoch": 1.036833855799373, "grad_norm": 3.24955153465271, "learning_rate": 8.44836567778667e-06, "loss": 1.1794, "step": 13230 }, { "epoch": 1.0369122257053291, "grad_norm": 5.335944652557373, "learning_rate": 8.448142379841255e-06, "loss": 1.1103, "step": 13231 }, { "epoch": 1.0369905956112853, "grad_norm": 4.333547115325928, "learning_rate": 8.447919068780877e-06, "loss": 1.2998, "step": 13232 }, { "epoch": 1.0370689655172414, "grad_norm": 5.362766742706299, "learning_rate": 8.447695744606386e-06, "loss": 1.7139, "step": 13233 }, { "epoch": 1.0371473354231975, "grad_norm": 3.130265235900879, "learning_rate": 8.44747240731863e-06, "loss": 1.3231, "step": 13234 }, { "epoch": 1.0372257053291536, "grad_norm": 3.3641245365142822, "learning_rate": 8.447249056918461e-06, "loss": 1.3235, "step": 13235 }, { "epoch": 1.0373040752351097, "grad_norm": 2.525418281555176, "learning_rate": 8.447025693406728e-06, "loss": 1.3892, "step": 13236 }, { "epoch": 1.0373824451410658, "grad_norm": 5.302795886993408, "learning_rate": 8.44680231678428e-06, "loss": 1.119, "step": 13237 }, { "epoch": 1.037460815047022, "grad_norm": 6.5733842849731445, "learning_rate": 8.446578927051965e-06, "loss": 1.5809, "step": 13238 }, { "epoch": 1.037539184952978, "grad_norm": 4.18161678314209, "learning_rate": 8.446355524210634e-06, "loss": 1.459, "step": 13239 }, { "epoch": 1.0376175548589341, "grad_norm": 2.3628084659576416, "learning_rate": 8.446132108261136e-06, "loss": 1.2996, "step": 13240 }, { "epoch": 1.0376959247648903, "grad_norm": 8.845600128173828, "learning_rate": 8.445908679204323e-06, "loss": 0.5647, "step": 13241 }, { "epoch": 1.0377742946708464, "grad_norm": 5.722842693328857, "learning_rate": 8.445685237041042e-06, "loss": 1.187, "step": 13242 }, { "epoch": 1.0378526645768025, "grad_norm": 10.700858116149902, "learning_rate": 8.445461781772146e-06, "loss": 2.3025, "step": 13243 }, { "epoch": 1.0379310344827586, "grad_norm": 2.7140214443206787, "learning_rate": 8.445238313398481e-06, "loss": 1.0956, "step": 13244 }, { "epoch": 1.0380094043887147, "grad_norm": 2.2772319316864014, "learning_rate": 8.445014831920902e-06, "loss": 1.3025, "step": 13245 }, { "epoch": 1.0380877742946708, "grad_norm": 3.709502935409546, "learning_rate": 8.444791337340253e-06, "loss": 1.251, "step": 13246 }, { "epoch": 1.038166144200627, "grad_norm": 10.65761947631836, "learning_rate": 8.444567829657388e-06, "loss": 1.1121, "step": 13247 }, { "epoch": 1.038244514106583, "grad_norm": 2.687899112701416, "learning_rate": 8.444344308873156e-06, "loss": 1.5717, "step": 13248 }, { "epoch": 1.0383228840125391, "grad_norm": 5.185573577880859, "learning_rate": 8.444120774988408e-06, "loss": 1.3251, "step": 13249 }, { "epoch": 1.0384012539184952, "grad_norm": 3.451512575149536, "learning_rate": 8.443897228003994e-06, "loss": 1.3529, "step": 13250 }, { "epoch": 1.0384796238244514, "grad_norm": 4.030796527862549, "learning_rate": 8.443673667920763e-06, "loss": 1.2992, "step": 13251 }, { "epoch": 1.0385579937304075, "grad_norm": 3.1655972003936768, "learning_rate": 8.443450094739564e-06, "loss": 1.8622, "step": 13252 }, { "epoch": 1.0386363636363636, "grad_norm": 3.1432383060455322, "learning_rate": 8.443226508461253e-06, "loss": 1.2135, "step": 13253 }, { "epoch": 1.0387147335423197, "grad_norm": 2.4697000980377197, "learning_rate": 8.443002909086677e-06, "loss": 1.4032, "step": 13254 }, { "epoch": 1.0387931034482758, "grad_norm": 5.697281837463379, "learning_rate": 8.442779296616685e-06, "loss": 1.0591, "step": 13255 }, { "epoch": 1.038871473354232, "grad_norm": 6.686534404754639, "learning_rate": 8.442555671052128e-06, "loss": 1.8806, "step": 13256 }, { "epoch": 1.038949843260188, "grad_norm": 3.133199691772461, "learning_rate": 8.442332032393858e-06, "loss": 1.5405, "step": 13257 }, { "epoch": 1.0390282131661441, "grad_norm": 7.928924083709717, "learning_rate": 8.442108380642727e-06, "loss": 1.8691, "step": 13258 }, { "epoch": 1.0391065830721002, "grad_norm": 3.113248825073242, "learning_rate": 8.441884715799582e-06, "loss": 1.4648, "step": 13259 }, { "epoch": 1.0391849529780564, "grad_norm": 2.09372878074646, "learning_rate": 8.441661037865277e-06, "loss": 1.2633, "step": 13260 }, { "epoch": 1.0392633228840125, "grad_norm": 4.852972507476807, "learning_rate": 8.441437346840657e-06, "loss": 1.0643, "step": 13261 }, { "epoch": 1.0393416927899686, "grad_norm": 2.1525440216064453, "learning_rate": 8.441213642726581e-06, "loss": 1.3502, "step": 13262 }, { "epoch": 1.0394200626959247, "grad_norm": 3.322349786758423, "learning_rate": 8.440989925523895e-06, "loss": 1.2875, "step": 13263 }, { "epoch": 1.0394984326018808, "grad_norm": 7.6988749504089355, "learning_rate": 8.44076619523345e-06, "loss": 1.3152, "step": 13264 }, { "epoch": 1.039576802507837, "grad_norm": 3.1268951892852783, "learning_rate": 8.440542451856097e-06, "loss": 1.2622, "step": 13265 }, { "epoch": 1.039655172413793, "grad_norm": 5.283077239990234, "learning_rate": 8.44031869539269e-06, "loss": 1.6328, "step": 13266 }, { "epoch": 1.0397335423197491, "grad_norm": 5.0676984786987305, "learning_rate": 8.440094925844075e-06, "loss": 1.5233, "step": 13267 }, { "epoch": 1.0398119122257052, "grad_norm": 5.443979740142822, "learning_rate": 8.439871143211107e-06, "loss": 1.4813, "step": 13268 }, { "epoch": 1.0398902821316613, "grad_norm": 4.304274082183838, "learning_rate": 8.439647347494638e-06, "loss": 1.5461, "step": 13269 }, { "epoch": 1.0399686520376175, "grad_norm": 2.96303129196167, "learning_rate": 8.439423538695515e-06, "loss": 1.6472, "step": 13270 }, { "epoch": 1.0400470219435736, "grad_norm": 3.0853981971740723, "learning_rate": 8.439199716814592e-06, "loss": 1.4756, "step": 13271 }, { "epoch": 1.0401253918495297, "grad_norm": 2.6887333393096924, "learning_rate": 8.43897588185272e-06, "loss": 1.3141, "step": 13272 }, { "epoch": 1.0402037617554858, "grad_norm": 7.209558486938477, "learning_rate": 8.43875203381075e-06, "loss": 1.5091, "step": 13273 }, { "epoch": 1.040282131661442, "grad_norm": 2.230184316635132, "learning_rate": 8.438528172689532e-06, "loss": 1.3055, "step": 13274 }, { "epoch": 1.040360501567398, "grad_norm": 2.861884832382202, "learning_rate": 8.43830429848992e-06, "loss": 1.4714, "step": 13275 }, { "epoch": 1.0404388714733543, "grad_norm": 3.203622341156006, "learning_rate": 8.438080411212766e-06, "loss": 1.5131, "step": 13276 }, { "epoch": 1.0405172413793102, "grad_norm": 3.3056740760803223, "learning_rate": 8.437856510858919e-06, "loss": 1.1777, "step": 13277 }, { "epoch": 1.0405956112852666, "grad_norm": 2.931443691253662, "learning_rate": 8.43763259742923e-06, "loss": 1.3776, "step": 13278 }, { "epoch": 1.0406739811912227, "grad_norm": 4.327452182769775, "learning_rate": 8.437408670924554e-06, "loss": 1.9165, "step": 13279 }, { "epoch": 1.0407523510971788, "grad_norm": 4.106869697570801, "learning_rate": 8.43718473134574e-06, "loss": 0.8596, "step": 13280 }, { "epoch": 1.040830721003135, "grad_norm": 3.091890335083008, "learning_rate": 8.436960778693643e-06, "loss": 1.3936, "step": 13281 }, { "epoch": 1.040909090909091, "grad_norm": 3.926582098007202, "learning_rate": 8.436736812969111e-06, "loss": 1.3253, "step": 13282 }, { "epoch": 1.0409874608150471, "grad_norm": 4.171814441680908, "learning_rate": 8.436512834172996e-06, "loss": 1.2452, "step": 13283 }, { "epoch": 1.0410658307210032, "grad_norm": 3.4893124103546143, "learning_rate": 8.436288842306152e-06, "loss": 1.6195, "step": 13284 }, { "epoch": 1.0411442006269593, "grad_norm": 3.067878007888794, "learning_rate": 8.436064837369433e-06, "loss": 1.5018, "step": 13285 }, { "epoch": 1.0412225705329154, "grad_norm": 6.4507551193237305, "learning_rate": 8.435840819363685e-06, "loss": 1.0664, "step": 13286 }, { "epoch": 1.0413009404388716, "grad_norm": 3.6697187423706055, "learning_rate": 8.435616788289764e-06, "loss": 1.644, "step": 13287 }, { "epoch": 1.0413793103448277, "grad_norm": 4.577412128448486, "learning_rate": 8.435392744148523e-06, "loss": 1.6844, "step": 13288 }, { "epoch": 1.0414576802507838, "grad_norm": 2.672208786010742, "learning_rate": 8.43516868694081e-06, "loss": 1.2343, "step": 13289 }, { "epoch": 1.04153605015674, "grad_norm": 3.0243020057678223, "learning_rate": 8.434944616667482e-06, "loss": 1.6228, "step": 13290 }, { "epoch": 1.041614420062696, "grad_norm": 4.743002414703369, "learning_rate": 8.434720533329388e-06, "loss": 1.7587, "step": 13291 }, { "epoch": 1.041692789968652, "grad_norm": 5.678651809692383, "learning_rate": 8.434496436927381e-06, "loss": 1.5701, "step": 13292 }, { "epoch": 1.0417711598746082, "grad_norm": 3.8078560829162598, "learning_rate": 8.434272327462314e-06, "loss": 1.3627, "step": 13293 }, { "epoch": 1.0418495297805643, "grad_norm": 3.0045223236083984, "learning_rate": 8.43404820493504e-06, "loss": 1.8873, "step": 13294 }, { "epoch": 1.0419278996865204, "grad_norm": 2.66194486618042, "learning_rate": 8.433824069346407e-06, "loss": 0.8529, "step": 13295 }, { "epoch": 1.0420062695924766, "grad_norm": 4.727532863616943, "learning_rate": 8.433599920697274e-06, "loss": 1.4524, "step": 13296 }, { "epoch": 1.0420846394984327, "grad_norm": 3.3198928833007812, "learning_rate": 8.433375758988492e-06, "loss": 1.3216, "step": 13297 }, { "epoch": 1.0421630094043888, "grad_norm": 3.8084065914154053, "learning_rate": 8.43315158422091e-06, "loss": 1.503, "step": 13298 }, { "epoch": 1.0422413793103449, "grad_norm": 2.492170572280884, "learning_rate": 8.43292739639538e-06, "loss": 1.5593, "step": 13299 }, { "epoch": 1.042319749216301, "grad_norm": 5.947071075439453, "learning_rate": 8.432703195512761e-06, "loss": 1.5007, "step": 13300 }, { "epoch": 1.042398119122257, "grad_norm": 17.579574584960938, "learning_rate": 8.432478981573901e-06, "loss": 1.2167, "step": 13301 }, { "epoch": 1.0424764890282132, "grad_norm": 3.285614013671875, "learning_rate": 8.432254754579655e-06, "loss": 1.3798, "step": 13302 }, { "epoch": 1.0425548589341693, "grad_norm": 2.100429058074951, "learning_rate": 8.432030514530874e-06, "loss": 1.5148, "step": 13303 }, { "epoch": 1.0426332288401254, "grad_norm": 11.371580123901367, "learning_rate": 8.431806261428412e-06, "loss": 1.716, "step": 13304 }, { "epoch": 1.0427115987460815, "grad_norm": 3.2815346717834473, "learning_rate": 8.431581995273122e-06, "loss": 1.7161, "step": 13305 }, { "epoch": 1.0427899686520377, "grad_norm": 3.073960304260254, "learning_rate": 8.431357716065856e-06, "loss": 1.3689, "step": 13306 }, { "epoch": 1.0428683385579938, "grad_norm": 2.85760235786438, "learning_rate": 8.431133423807467e-06, "loss": 1.4359, "step": 13307 }, { "epoch": 1.0429467084639499, "grad_norm": 2.9275922775268555, "learning_rate": 8.430909118498809e-06, "loss": 1.5071, "step": 13308 }, { "epoch": 1.043025078369906, "grad_norm": 3.6717216968536377, "learning_rate": 8.430684800140736e-06, "loss": 1.4681, "step": 13309 }, { "epoch": 1.043103448275862, "grad_norm": 3.7777087688446045, "learning_rate": 8.4304604687341e-06, "loss": 1.2019, "step": 13310 }, { "epoch": 1.0431818181818182, "grad_norm": 3.033188581466675, "learning_rate": 8.430236124279755e-06, "loss": 1.1913, "step": 13311 }, { "epoch": 1.0432601880877743, "grad_norm": 3.4297380447387695, "learning_rate": 8.430011766778553e-06, "loss": 1.4684, "step": 13312 }, { "epoch": 1.0433385579937304, "grad_norm": 1.9062230587005615, "learning_rate": 8.429787396231346e-06, "loss": 1.1698, "step": 13313 }, { "epoch": 1.0434169278996865, "grad_norm": 2.9145822525024414, "learning_rate": 8.429563012638992e-06, "loss": 1.6424, "step": 13314 }, { "epoch": 1.0434952978056427, "grad_norm": 2.8779938220977783, "learning_rate": 8.42933861600234e-06, "loss": 1.5427, "step": 13315 }, { "epoch": 1.0435736677115988, "grad_norm": 5.645954132080078, "learning_rate": 8.429114206322249e-06, "loss": 1.3885, "step": 13316 }, { "epoch": 1.0436520376175549, "grad_norm": 2.2259016036987305, "learning_rate": 8.428889783599566e-06, "loss": 1.3231, "step": 13317 }, { "epoch": 1.043730407523511, "grad_norm": 4.509953022003174, "learning_rate": 8.428665347835146e-06, "loss": 1.2376, "step": 13318 }, { "epoch": 1.043808777429467, "grad_norm": 2.142228841781616, "learning_rate": 8.428440899029847e-06, "loss": 1.5509, "step": 13319 }, { "epoch": 1.0438871473354232, "grad_norm": 3.2121856212615967, "learning_rate": 8.428216437184518e-06, "loss": 1.1895, "step": 13320 }, { "epoch": 1.0439655172413793, "grad_norm": 4.194362163543701, "learning_rate": 8.427991962300015e-06, "loss": 1.7089, "step": 13321 }, { "epoch": 1.0440438871473354, "grad_norm": 2.330104351043701, "learning_rate": 8.427767474377191e-06, "loss": 1.2035, "step": 13322 }, { "epoch": 1.0441222570532915, "grad_norm": 2.4247846603393555, "learning_rate": 8.4275429734169e-06, "loss": 1.2468, "step": 13323 }, { "epoch": 1.0442006269592476, "grad_norm": 3.064593553543091, "learning_rate": 8.427318459419997e-06, "loss": 1.2705, "step": 13324 }, { "epoch": 1.0442789968652038, "grad_norm": 6.672586441040039, "learning_rate": 8.427093932387336e-06, "loss": 0.8397, "step": 13325 }, { "epoch": 1.0443573667711599, "grad_norm": 4.896103382110596, "learning_rate": 8.426869392319768e-06, "loss": 1.3893, "step": 13326 }, { "epoch": 1.044435736677116, "grad_norm": 2.8779520988464355, "learning_rate": 8.42664483921815e-06, "loss": 1.2211, "step": 13327 }, { "epoch": 1.044514106583072, "grad_norm": 4.663296699523926, "learning_rate": 8.426420273083333e-06, "loss": 1.4241, "step": 13328 }, { "epoch": 1.0445924764890282, "grad_norm": 4.321670055389404, "learning_rate": 8.426195693916176e-06, "loss": 1.4997, "step": 13329 }, { "epoch": 1.0446708463949843, "grad_norm": 1.8805159330368042, "learning_rate": 8.425971101717528e-06, "loss": 0.9396, "step": 13330 }, { "epoch": 1.0447492163009404, "grad_norm": 3.0254180431365967, "learning_rate": 8.425746496488247e-06, "loss": 1.8277, "step": 13331 }, { "epoch": 1.0448275862068965, "grad_norm": 3.708425521850586, "learning_rate": 8.425521878229186e-06, "loss": 1.4563, "step": 13332 }, { "epoch": 1.0449059561128526, "grad_norm": 3.5718626976013184, "learning_rate": 8.4252972469412e-06, "loss": 1.5288, "step": 13333 }, { "epoch": 1.0449843260188088, "grad_norm": 2.2890732288360596, "learning_rate": 8.42507260262514e-06, "loss": 1.6243, "step": 13334 }, { "epoch": 1.0450626959247649, "grad_norm": 2.776388168334961, "learning_rate": 8.424847945281864e-06, "loss": 1.2674, "step": 13335 }, { "epoch": 1.045141065830721, "grad_norm": 2.478729009628296, "learning_rate": 8.424623274912226e-06, "loss": 1.3528, "step": 13336 }, { "epoch": 1.045219435736677, "grad_norm": 2.3901689052581787, "learning_rate": 8.42439859151708e-06, "loss": 1.5653, "step": 13337 }, { "epoch": 1.0452978056426332, "grad_norm": 7.615940570831299, "learning_rate": 8.42417389509728e-06, "loss": 1.1903, "step": 13338 }, { "epoch": 1.0453761755485893, "grad_norm": 3.2366232872009277, "learning_rate": 8.423949185653684e-06, "loss": 1.3249, "step": 13339 }, { "epoch": 1.0454545454545454, "grad_norm": 2.6348965167999268, "learning_rate": 8.423724463187142e-06, "loss": 1.0161, "step": 13340 }, { "epoch": 1.0455329153605015, "grad_norm": 6.556733131408691, "learning_rate": 8.42349972769851e-06, "loss": 0.9725, "step": 13341 }, { "epoch": 1.0456112852664576, "grad_norm": 3.86804461479187, "learning_rate": 8.423274979188646e-06, "loss": 1.5953, "step": 13342 }, { "epoch": 1.0456896551724137, "grad_norm": 4.612515449523926, "learning_rate": 8.423050217658398e-06, "loss": 1.413, "step": 13343 }, { "epoch": 1.0457680250783699, "grad_norm": 3.1843044757843018, "learning_rate": 8.422825443108627e-06, "loss": 1.3815, "step": 13344 }, { "epoch": 1.045846394984326, "grad_norm": 1.8394217491149902, "learning_rate": 8.422600655540187e-06, "loss": 1.392, "step": 13345 }, { "epoch": 1.045924764890282, "grad_norm": 3.0398144721984863, "learning_rate": 8.422375854953933e-06, "loss": 1.6396, "step": 13346 }, { "epoch": 1.0460031347962382, "grad_norm": 4.492502689361572, "learning_rate": 8.422151041350717e-06, "loss": 1.2839, "step": 13347 }, { "epoch": 1.0460815047021943, "grad_norm": 3.9566450119018555, "learning_rate": 8.421926214731398e-06, "loss": 0.9171, "step": 13348 }, { "epoch": 1.0461598746081504, "grad_norm": 2.1514883041381836, "learning_rate": 8.421701375096827e-06, "loss": 1.3625, "step": 13349 }, { "epoch": 1.0462382445141065, "grad_norm": 2.4513134956359863, "learning_rate": 8.421476522447862e-06, "loss": 1.3988, "step": 13350 }, { "epoch": 1.0463166144200626, "grad_norm": 2.1062984466552734, "learning_rate": 8.421251656785359e-06, "loss": 0.9742, "step": 13351 }, { "epoch": 1.0463949843260187, "grad_norm": 4.915331840515137, "learning_rate": 8.421026778110172e-06, "loss": 1.2804, "step": 13352 }, { "epoch": 1.0464733542319749, "grad_norm": 3.0626580715179443, "learning_rate": 8.420801886423156e-06, "loss": 1.1588, "step": 13353 }, { "epoch": 1.046551724137931, "grad_norm": 1.6253844499588013, "learning_rate": 8.420576981725167e-06, "loss": 1.1713, "step": 13354 }, { "epoch": 1.046630094043887, "grad_norm": 3.01161527633667, "learning_rate": 8.420352064017058e-06, "loss": 1.4014, "step": 13355 }, { "epoch": 1.0467084639498432, "grad_norm": 2.9931368827819824, "learning_rate": 8.420127133299687e-06, "loss": 1.1549, "step": 13356 }, { "epoch": 1.0467868338557993, "grad_norm": 2.1722075939178467, "learning_rate": 8.419902189573911e-06, "loss": 1.5194, "step": 13357 }, { "epoch": 1.0468652037617554, "grad_norm": 2.887359619140625, "learning_rate": 8.419677232840582e-06, "loss": 1.2355, "step": 13358 }, { "epoch": 1.0469435736677115, "grad_norm": 3.4223337173461914, "learning_rate": 8.419452263100558e-06, "loss": 1.268, "step": 13359 }, { "epoch": 1.0470219435736676, "grad_norm": 2.798557758331299, "learning_rate": 8.419227280354693e-06, "loss": 1.4371, "step": 13360 }, { "epoch": 1.0471003134796237, "grad_norm": 3.1729698181152344, "learning_rate": 8.419002284603845e-06, "loss": 1.4797, "step": 13361 }, { "epoch": 1.0471786833855798, "grad_norm": 2.1977341175079346, "learning_rate": 8.418777275848868e-06, "loss": 1.4303, "step": 13362 }, { "epoch": 1.047257053291536, "grad_norm": 3.2377076148986816, "learning_rate": 8.418552254090619e-06, "loss": 1.6796, "step": 13363 }, { "epoch": 1.047335423197492, "grad_norm": 3.312351703643799, "learning_rate": 8.418327219329952e-06, "loss": 1.7969, "step": 13364 }, { "epoch": 1.0474137931034482, "grad_norm": 2.9646286964416504, "learning_rate": 8.418102171567725e-06, "loss": 1.5108, "step": 13365 }, { "epoch": 1.0474921630094043, "grad_norm": 3.74006724357605, "learning_rate": 8.417877110804791e-06, "loss": 1.3845, "step": 13366 }, { "epoch": 1.0475705329153604, "grad_norm": 2.536135196685791, "learning_rate": 8.41765203704201e-06, "loss": 1.3084, "step": 13367 }, { "epoch": 1.0476489028213165, "grad_norm": 3.7267682552337646, "learning_rate": 8.417426950280236e-06, "loss": 1.256, "step": 13368 }, { "epoch": 1.0477272727272728, "grad_norm": 5.8868327140808105, "learning_rate": 8.417201850520323e-06, "loss": 2.2348, "step": 13369 }, { "epoch": 1.0478056426332287, "grad_norm": 2.794905662536621, "learning_rate": 8.416976737763132e-06, "loss": 1.5338, "step": 13370 }, { "epoch": 1.047884012539185, "grad_norm": 2.689424753189087, "learning_rate": 8.416751612009515e-06, "loss": 1.5036, "step": 13371 }, { "epoch": 1.0479623824451412, "grad_norm": 2.156572103500366, "learning_rate": 8.416526473260331e-06, "loss": 1.3782, "step": 13372 }, { "epoch": 1.0480407523510973, "grad_norm": 2.079923391342163, "learning_rate": 8.416301321516433e-06, "loss": 1.1621, "step": 13373 }, { "epoch": 1.0481191222570534, "grad_norm": 2.20526123046875, "learning_rate": 8.416076156778682e-06, "loss": 1.2077, "step": 13374 }, { "epoch": 1.0481974921630095, "grad_norm": 3.3976354598999023, "learning_rate": 8.41585097904793e-06, "loss": 1.3137, "step": 13375 }, { "epoch": 1.0482758620689656, "grad_norm": 9.615546226501465, "learning_rate": 8.415625788325037e-06, "loss": 1.0943, "step": 13376 }, { "epoch": 1.0483542319749217, "grad_norm": 2.4371278285980225, "learning_rate": 8.415400584610855e-06, "loss": 1.3273, "step": 13377 }, { "epoch": 1.0484326018808778, "grad_norm": 4.990485668182373, "learning_rate": 8.415175367906247e-06, "loss": 2.0764, "step": 13378 }, { "epoch": 1.048510971786834, "grad_norm": 12.04956340789795, "learning_rate": 8.414950138212065e-06, "loss": 1.8316, "step": 13379 }, { "epoch": 1.04858934169279, "grad_norm": 2.957730770111084, "learning_rate": 8.414724895529166e-06, "loss": 1.581, "step": 13380 }, { "epoch": 1.0486677115987462, "grad_norm": 3.277837038040161, "learning_rate": 8.414499639858408e-06, "loss": 1.2144, "step": 13381 }, { "epoch": 1.0487460815047023, "grad_norm": 2.3401920795440674, "learning_rate": 8.414274371200646e-06, "loss": 1.5416, "step": 13382 }, { "epoch": 1.0488244514106584, "grad_norm": 1.872633695602417, "learning_rate": 8.41404908955674e-06, "loss": 1.3225, "step": 13383 }, { "epoch": 1.0489028213166145, "grad_norm": 2.187685966491699, "learning_rate": 8.413823794927543e-06, "loss": 0.9831, "step": 13384 }, { "epoch": 1.0489811912225706, "grad_norm": 2.1502525806427, "learning_rate": 8.413598487313914e-06, "loss": 1.3486, "step": 13385 }, { "epoch": 1.0490595611285267, "grad_norm": 3.1901514530181885, "learning_rate": 8.41337316671671e-06, "loss": 1.2538, "step": 13386 }, { "epoch": 1.0491379310344828, "grad_norm": 1.921825885772705, "learning_rate": 8.413147833136788e-06, "loss": 1.4295, "step": 13387 }, { "epoch": 1.049216300940439, "grad_norm": 4.16799259185791, "learning_rate": 8.412922486575004e-06, "loss": 2.1857, "step": 13388 }, { "epoch": 1.049294670846395, "grad_norm": 2.177361011505127, "learning_rate": 8.412697127032216e-06, "loss": 1.6381, "step": 13389 }, { "epoch": 1.0493730407523512, "grad_norm": 13.397650718688965, "learning_rate": 8.412471754509282e-06, "loss": 0.9377, "step": 13390 }, { "epoch": 1.0494514106583073, "grad_norm": 4.0266900062561035, "learning_rate": 8.412246369007057e-06, "loss": 1.4071, "step": 13391 }, { "epoch": 1.0495297805642634, "grad_norm": 3.921616554260254, "learning_rate": 8.4120209705264e-06, "loss": 1.35, "step": 13392 }, { "epoch": 1.0496081504702195, "grad_norm": 2.672206401824951, "learning_rate": 8.411795559068169e-06, "loss": 1.2296, "step": 13393 }, { "epoch": 1.0496865203761756, "grad_norm": 11.15247917175293, "learning_rate": 8.411570134633218e-06, "loss": 0.667, "step": 13394 }, { "epoch": 1.0497648902821317, "grad_norm": 2.3061509132385254, "learning_rate": 8.411344697222408e-06, "loss": 1.2088, "step": 13395 }, { "epoch": 1.0498432601880878, "grad_norm": 2.7439968585968018, "learning_rate": 8.411119246836594e-06, "loss": 1.4672, "step": 13396 }, { "epoch": 1.049921630094044, "grad_norm": 2.3352231979370117, "learning_rate": 8.410893783476635e-06, "loss": 1.3574, "step": 13397 }, { "epoch": 1.05, "grad_norm": 2.594351053237915, "learning_rate": 8.410668307143388e-06, "loss": 1.1539, "step": 13398 }, { "epoch": 1.0500783699059562, "grad_norm": 1.9822602272033691, "learning_rate": 8.41044281783771e-06, "loss": 0.9465, "step": 13399 }, { "epoch": 1.0501567398119123, "grad_norm": 2.352076530456543, "learning_rate": 8.410217315560459e-06, "loss": 1.5141, "step": 13400 }, { "epoch": 1.0502351097178684, "grad_norm": 2.1924400329589844, "learning_rate": 8.409991800312493e-06, "loss": 0.9989, "step": 13401 }, { "epoch": 1.0503134796238245, "grad_norm": 3.65924334526062, "learning_rate": 8.409766272094671e-06, "loss": 1.0921, "step": 13402 }, { "epoch": 1.0503918495297806, "grad_norm": 8.2733154296875, "learning_rate": 8.409540730907848e-06, "loss": 1.8283, "step": 13403 }, { "epoch": 1.0504702194357367, "grad_norm": 3.8907361030578613, "learning_rate": 8.409315176752883e-06, "loss": 1.6622, "step": 13404 }, { "epoch": 1.0505485893416928, "grad_norm": 2.6821975708007812, "learning_rate": 8.409089609630636e-06, "loss": 1.0907, "step": 13405 }, { "epoch": 1.050626959247649, "grad_norm": 5.805487155914307, "learning_rate": 8.408864029541963e-06, "loss": 1.2139, "step": 13406 }, { "epoch": 1.050705329153605, "grad_norm": 4.645816802978516, "learning_rate": 8.408638436487721e-06, "loss": 1.3395, "step": 13407 }, { "epoch": 1.0507836990595611, "grad_norm": 2.694829225540161, "learning_rate": 8.40841283046877e-06, "loss": 1.3567, "step": 13408 }, { "epoch": 1.0508620689655173, "grad_norm": 2.043790817260742, "learning_rate": 8.408187211485966e-06, "loss": 1.517, "step": 13409 }, { "epoch": 1.0509404388714734, "grad_norm": 2.610466957092285, "learning_rate": 8.40796157954017e-06, "loss": 0.9558, "step": 13410 }, { "epoch": 1.0510188087774295, "grad_norm": 7.7400288581848145, "learning_rate": 8.407735934632237e-06, "loss": 1.0685, "step": 13411 }, { "epoch": 1.0510971786833856, "grad_norm": 3.4227352142333984, "learning_rate": 8.407510276763028e-06, "loss": 1.1995, "step": 13412 }, { "epoch": 1.0511755485893417, "grad_norm": 3.6625664234161377, "learning_rate": 8.4072846059334e-06, "loss": 1.3947, "step": 13413 }, { "epoch": 1.0512539184952978, "grad_norm": 2.9473352432250977, "learning_rate": 8.407058922144212e-06, "loss": 1.4732, "step": 13414 }, { "epoch": 1.051332288401254, "grad_norm": 3.5120649337768555, "learning_rate": 8.40683322539632e-06, "loss": 1.6133, "step": 13415 }, { "epoch": 1.05141065830721, "grad_norm": 3.0872230529785156, "learning_rate": 8.406607515690586e-06, "loss": 1.2842, "step": 13416 }, { "epoch": 1.0514890282131661, "grad_norm": 3.824678659439087, "learning_rate": 8.406381793027866e-06, "loss": 1.3267, "step": 13417 }, { "epoch": 1.0515673981191223, "grad_norm": 7.421202659606934, "learning_rate": 8.406156057409019e-06, "loss": 0.6706, "step": 13418 }, { "epoch": 1.0516457680250784, "grad_norm": 2.577648401260376, "learning_rate": 8.405930308834905e-06, "loss": 0.8603, "step": 13419 }, { "epoch": 1.0517241379310345, "grad_norm": 3.438746929168701, "learning_rate": 8.405704547306379e-06, "loss": 1.7566, "step": 13420 }, { "epoch": 1.0518025078369906, "grad_norm": 4.2059431076049805, "learning_rate": 8.405478772824305e-06, "loss": 1.2461, "step": 13421 }, { "epoch": 1.0518808777429467, "grad_norm": 4.211129665374756, "learning_rate": 8.405252985389538e-06, "loss": 1.4541, "step": 13422 }, { "epoch": 1.0519592476489028, "grad_norm": 2.5367062091827393, "learning_rate": 8.405027185002936e-06, "loss": 1.5189, "step": 13423 }, { "epoch": 1.052037617554859, "grad_norm": 4.212551116943359, "learning_rate": 8.404801371665363e-06, "loss": 1.5182, "step": 13424 }, { "epoch": 1.052115987460815, "grad_norm": 3.21335506439209, "learning_rate": 8.404575545377671e-06, "loss": 1.1228, "step": 13425 }, { "epoch": 1.0521943573667711, "grad_norm": 4.182920455932617, "learning_rate": 8.404349706140725e-06, "loss": 1.4876, "step": 13426 }, { "epoch": 1.0522727272727272, "grad_norm": 3.3986117839813232, "learning_rate": 8.40412385395538e-06, "loss": 1.5091, "step": 13427 }, { "epoch": 1.0523510971786834, "grad_norm": 4.811435699462891, "learning_rate": 8.403897988822496e-06, "loss": 1.3618, "step": 13428 }, { "epoch": 1.0524294670846395, "grad_norm": 7.842459201812744, "learning_rate": 8.403672110742933e-06, "loss": 1.3258, "step": 13429 }, { "epoch": 1.0525078369905956, "grad_norm": 3.712512969970703, "learning_rate": 8.403446219717549e-06, "loss": 1.7142, "step": 13430 }, { "epoch": 1.0525862068965517, "grad_norm": 3.2398064136505127, "learning_rate": 8.403220315747204e-06, "loss": 1.1418, "step": 13431 }, { "epoch": 1.0526645768025078, "grad_norm": 4.885566234588623, "learning_rate": 8.402994398832757e-06, "loss": 1.6694, "step": 13432 }, { "epoch": 1.052742946708464, "grad_norm": 4.8894147872924805, "learning_rate": 8.402768468975066e-06, "loss": 1.0315, "step": 13433 }, { "epoch": 1.05282131661442, "grad_norm": 3.999018907546997, "learning_rate": 8.402542526174994e-06, "loss": 1.3405, "step": 13434 }, { "epoch": 1.0528996865203761, "grad_norm": 3.3791325092315674, "learning_rate": 8.402316570433398e-06, "loss": 1.3667, "step": 13435 }, { "epoch": 1.0529780564263322, "grad_norm": 3.0184829235076904, "learning_rate": 8.402090601751133e-06, "loss": 1.205, "step": 13436 }, { "epoch": 1.0530564263322884, "grad_norm": 2.3377511501312256, "learning_rate": 8.401864620129068e-06, "loss": 1.2311, "step": 13437 }, { "epoch": 1.0531347962382445, "grad_norm": 2.2384722232818604, "learning_rate": 8.401638625568053e-06, "loss": 1.1005, "step": 13438 }, { "epoch": 1.0532131661442006, "grad_norm": 2.3095340728759766, "learning_rate": 8.401412618068956e-06, "loss": 1.4186, "step": 13439 }, { "epoch": 1.0532915360501567, "grad_norm": 3.641944408416748, "learning_rate": 8.401186597632629e-06, "loss": 1.6351, "step": 13440 }, { "epoch": 1.0533699059561128, "grad_norm": 3.1661367416381836, "learning_rate": 8.400960564259936e-06, "loss": 1.3157, "step": 13441 }, { "epoch": 1.053448275862069, "grad_norm": 3.309349536895752, "learning_rate": 8.400734517951737e-06, "loss": 1.2737, "step": 13442 }, { "epoch": 1.053526645768025, "grad_norm": 3.4827818870544434, "learning_rate": 8.400508458708889e-06, "loss": 1.7327, "step": 13443 }, { "epoch": 1.0536050156739811, "grad_norm": 2.249472141265869, "learning_rate": 8.400282386532254e-06, "loss": 1.7481, "step": 13444 }, { "epoch": 1.0536833855799372, "grad_norm": 2.6388189792633057, "learning_rate": 8.400056301422692e-06, "loss": 1.2346, "step": 13445 }, { "epoch": 1.0537617554858933, "grad_norm": 3.0733983516693115, "learning_rate": 8.399830203381063e-06, "loss": 1.4066, "step": 13446 }, { "epoch": 1.0538401253918495, "grad_norm": 3.895796060562134, "learning_rate": 8.399604092408224e-06, "loss": 1.4619, "step": 13447 }, { "epoch": 1.0539184952978056, "grad_norm": 4.173047065734863, "learning_rate": 8.399377968505038e-06, "loss": 1.4266, "step": 13448 }, { "epoch": 1.0539968652037617, "grad_norm": 3.8717925548553467, "learning_rate": 8.399151831672366e-06, "loss": 1.3874, "step": 13449 }, { "epoch": 1.0540752351097178, "grad_norm": 6.65088415145874, "learning_rate": 8.398925681911064e-06, "loss": 1.4168, "step": 13450 }, { "epoch": 1.054153605015674, "grad_norm": 2.2791635990142822, "learning_rate": 8.398699519221995e-06, "loss": 1.6044, "step": 13451 }, { "epoch": 1.05423197492163, "grad_norm": 3.1590194702148438, "learning_rate": 8.398473343606019e-06, "loss": 1.5237, "step": 13452 }, { "epoch": 1.0543103448275861, "grad_norm": 2.448552131652832, "learning_rate": 8.398247155063996e-06, "loss": 1.2276, "step": 13453 }, { "epoch": 1.0543887147335422, "grad_norm": 2.305138111114502, "learning_rate": 8.398020953596786e-06, "loss": 1.2865, "step": 13454 }, { "epoch": 1.0544670846394983, "grad_norm": 2.0295779705047607, "learning_rate": 8.397794739205251e-06, "loss": 1.4133, "step": 13455 }, { "epoch": 1.0545454545454545, "grad_norm": 3.193354845046997, "learning_rate": 8.397568511890248e-06, "loss": 1.2829, "step": 13456 }, { "epoch": 1.0546238244514106, "grad_norm": 3.387437105178833, "learning_rate": 8.39734227165264e-06, "loss": 1.5172, "step": 13457 }, { "epoch": 1.0547021943573667, "grad_norm": 4.6500983238220215, "learning_rate": 8.397116018493287e-06, "loss": 1.0544, "step": 13458 }, { "epoch": 1.0547805642633228, "grad_norm": 3.2670207023620605, "learning_rate": 8.39688975241305e-06, "loss": 1.0861, "step": 13459 }, { "epoch": 1.054858934169279, "grad_norm": 7.716726779937744, "learning_rate": 8.39666347341279e-06, "loss": 1.3142, "step": 13460 }, { "epoch": 1.054937304075235, "grad_norm": 2.620182752609253, "learning_rate": 8.396437181493366e-06, "loss": 1.2964, "step": 13461 }, { "epoch": 1.0550156739811911, "grad_norm": 2.520934820175171, "learning_rate": 8.39621087665564e-06, "loss": 1.2632, "step": 13462 }, { "epoch": 1.0550940438871472, "grad_norm": 2.660379648208618, "learning_rate": 8.39598455890047e-06, "loss": 1.5821, "step": 13463 }, { "epoch": 1.0551724137931036, "grad_norm": 4.568061351776123, "learning_rate": 8.395758228228722e-06, "loss": 1.771, "step": 13464 }, { "epoch": 1.0552507836990594, "grad_norm": 2.689476490020752, "learning_rate": 8.395531884641252e-06, "loss": 1.5383, "step": 13465 }, { "epoch": 1.0553291536050158, "grad_norm": 4.854791641235352, "learning_rate": 8.395305528138923e-06, "loss": 1.2402, "step": 13466 }, { "epoch": 1.055407523510972, "grad_norm": 2.538553237915039, "learning_rate": 8.395079158722595e-06, "loss": 1.4434, "step": 13467 }, { "epoch": 1.055485893416928, "grad_norm": 2.08620023727417, "learning_rate": 8.39485277639313e-06, "loss": 1.3772, "step": 13468 }, { "epoch": 1.055564263322884, "grad_norm": 3.4400386810302734, "learning_rate": 8.394626381151391e-06, "loss": 0.9426, "step": 13469 }, { "epoch": 1.0556426332288402, "grad_norm": 2.7434237003326416, "learning_rate": 8.394399972998235e-06, "loss": 1.4914, "step": 13470 }, { "epoch": 1.0557210031347963, "grad_norm": 4.067943096160889, "learning_rate": 8.394173551934525e-06, "loss": 1.794, "step": 13471 }, { "epoch": 1.0557993730407524, "grad_norm": 3.0852158069610596, "learning_rate": 8.393947117961121e-06, "loss": 1.5198, "step": 13472 }, { "epoch": 1.0558777429467086, "grad_norm": 2.4272236824035645, "learning_rate": 8.393720671078888e-06, "loss": 1.3904, "step": 13473 }, { "epoch": 1.0559561128526647, "grad_norm": 2.246058225631714, "learning_rate": 8.393494211288683e-06, "loss": 0.9353, "step": 13474 }, { "epoch": 1.0560344827586208, "grad_norm": 4.953530311584473, "learning_rate": 8.393267738591369e-06, "loss": 1.8911, "step": 13475 }, { "epoch": 1.0561128526645769, "grad_norm": 2.569854497909546, "learning_rate": 8.393041252987807e-06, "loss": 1.3326, "step": 13476 }, { "epoch": 1.056191222570533, "grad_norm": 3.976273775100708, "learning_rate": 8.392814754478862e-06, "loss": 1.4155, "step": 13477 }, { "epoch": 1.056269592476489, "grad_norm": 2.6036148071289062, "learning_rate": 8.39258824306539e-06, "loss": 1.4104, "step": 13478 }, { "epoch": 1.0563479623824452, "grad_norm": 1.781184434890747, "learning_rate": 8.392361718748253e-06, "loss": 1.3611, "step": 13479 }, { "epoch": 1.0564263322884013, "grad_norm": 12.69182300567627, "learning_rate": 8.392135181528318e-06, "loss": 1.9152, "step": 13480 }, { "epoch": 1.0565047021943574, "grad_norm": 3.5327417850494385, "learning_rate": 8.391908631406442e-06, "loss": 1.1803, "step": 13481 }, { "epoch": 1.0565830721003135, "grad_norm": 7.277730464935303, "learning_rate": 8.391682068383486e-06, "loss": 1.5571, "step": 13482 }, { "epoch": 1.0566614420062697, "grad_norm": 5.042743682861328, "learning_rate": 8.391455492460315e-06, "loss": 1.5293, "step": 13483 }, { "epoch": 1.0567398119122258, "grad_norm": 2.435750722885132, "learning_rate": 8.39122890363779e-06, "loss": 1.3743, "step": 13484 }, { "epoch": 1.0568181818181819, "grad_norm": 6.330252647399902, "learning_rate": 8.39100230191677e-06, "loss": 1.1788, "step": 13485 }, { "epoch": 1.056896551724138, "grad_norm": 3.5035083293914795, "learning_rate": 8.39077568729812e-06, "loss": 1.4447, "step": 13486 }, { "epoch": 1.056974921630094, "grad_norm": 2.8107635974884033, "learning_rate": 8.3905490597827e-06, "loss": 1.0262, "step": 13487 }, { "epoch": 1.0570532915360502, "grad_norm": 2.205852508544922, "learning_rate": 8.390322419371376e-06, "loss": 1.5369, "step": 13488 }, { "epoch": 1.0571316614420063, "grad_norm": 5.931280612945557, "learning_rate": 8.390095766065003e-06, "loss": 1.9255, "step": 13489 }, { "epoch": 1.0572100313479624, "grad_norm": 10.996716499328613, "learning_rate": 8.38986909986445e-06, "loss": 1.2814, "step": 13490 }, { "epoch": 1.0572884012539185, "grad_norm": 2.3127670288085938, "learning_rate": 8.389642420770574e-06, "loss": 1.3636, "step": 13491 }, { "epoch": 1.0573667711598747, "grad_norm": 2.422323703765869, "learning_rate": 8.38941572878424e-06, "loss": 1.0669, "step": 13492 }, { "epoch": 1.0574451410658308, "grad_norm": 5.034599304199219, "learning_rate": 8.38918902390631e-06, "loss": 1.273, "step": 13493 }, { "epoch": 1.0575235109717869, "grad_norm": 3.226956367492676, "learning_rate": 8.388962306137644e-06, "loss": 1.3753, "step": 13494 }, { "epoch": 1.057601880877743, "grad_norm": 4.482888221740723, "learning_rate": 8.388735575479108e-06, "loss": 1.4582, "step": 13495 }, { "epoch": 1.057680250783699, "grad_norm": 4.766706466674805, "learning_rate": 8.38850883193156e-06, "loss": 1.6264, "step": 13496 }, { "epoch": 1.0577586206896552, "grad_norm": 1.941304087638855, "learning_rate": 8.388282075495868e-06, "loss": 1.2997, "step": 13497 }, { "epoch": 1.0578369905956113, "grad_norm": 5.101015567779541, "learning_rate": 8.38805530617289e-06, "loss": 1.1579, "step": 13498 }, { "epoch": 1.0579153605015674, "grad_norm": 3.109020948410034, "learning_rate": 8.387828523963491e-06, "loss": 1.3957, "step": 13499 }, { "epoch": 1.0579937304075235, "grad_norm": 4.854037284851074, "learning_rate": 8.387601728868531e-06, "loss": 1.2124, "step": 13500 }, { "epoch": 1.0580721003134796, "grad_norm": 2.721198081970215, "learning_rate": 8.387374920888873e-06, "loss": 1.3609, "step": 13501 }, { "epoch": 1.0581504702194358, "grad_norm": 4.097742557525635, "learning_rate": 8.387148100025381e-06, "loss": 1.6336, "step": 13502 }, { "epoch": 1.0582288401253919, "grad_norm": 2.9311540126800537, "learning_rate": 8.38692126627892e-06, "loss": 1.0227, "step": 13503 }, { "epoch": 1.058307210031348, "grad_norm": 3.343120813369751, "learning_rate": 8.386694419650347e-06, "loss": 1.2703, "step": 13504 }, { "epoch": 1.058385579937304, "grad_norm": 5.536196708679199, "learning_rate": 8.38646756014053e-06, "loss": 1.4343, "step": 13505 }, { "epoch": 1.0584639498432602, "grad_norm": 2.793384552001953, "learning_rate": 8.386240687750328e-06, "loss": 1.2925, "step": 13506 }, { "epoch": 1.0585423197492163, "grad_norm": 3.339240312576294, "learning_rate": 8.386013802480608e-06, "loss": 1.0436, "step": 13507 }, { "epoch": 1.0586206896551724, "grad_norm": 3.08461856842041, "learning_rate": 8.38578690433223e-06, "loss": 1.1508, "step": 13508 }, { "epoch": 1.0586990595611285, "grad_norm": 2.228928565979004, "learning_rate": 8.385559993306057e-06, "loss": 0.9829, "step": 13509 }, { "epoch": 1.0587774294670846, "grad_norm": 3.5890488624572754, "learning_rate": 8.385333069402952e-06, "loss": 1.1462, "step": 13510 }, { "epoch": 1.0588557993730408, "grad_norm": 13.916231155395508, "learning_rate": 8.385106132623779e-06, "loss": 1.6761, "step": 13511 }, { "epoch": 1.0589341692789969, "grad_norm": 2.197772264480591, "learning_rate": 8.384879182969404e-06, "loss": 1.2978, "step": 13512 }, { "epoch": 1.059012539184953, "grad_norm": 2.4995734691619873, "learning_rate": 8.384652220440684e-06, "loss": 1.7207, "step": 13513 }, { "epoch": 1.059090909090909, "grad_norm": 6.359238147735596, "learning_rate": 8.384425245038485e-06, "loss": 1.5743, "step": 13514 }, { "epoch": 1.0591692789968652, "grad_norm": 2.204554319381714, "learning_rate": 8.384198256763672e-06, "loss": 1.5546, "step": 13515 }, { "epoch": 1.0592476489028213, "grad_norm": 1.8959176540374756, "learning_rate": 8.383971255617107e-06, "loss": 1.3302, "step": 13516 }, { "epoch": 1.0593260188087774, "grad_norm": 2.639526844024658, "learning_rate": 8.383744241599654e-06, "loss": 1.0987, "step": 13517 }, { "epoch": 1.0594043887147335, "grad_norm": 3.3892710208892822, "learning_rate": 8.383517214712176e-06, "loss": 1.4823, "step": 13518 }, { "epoch": 1.0594827586206896, "grad_norm": 2.6879055500030518, "learning_rate": 8.383290174955535e-06, "loss": 1.4495, "step": 13519 }, { "epoch": 1.0595611285266457, "grad_norm": 2.74100661277771, "learning_rate": 8.383063122330597e-06, "loss": 1.9642, "step": 13520 }, { "epoch": 1.0596394984326019, "grad_norm": 4.17286491394043, "learning_rate": 8.382836056838224e-06, "loss": 1.1899, "step": 13521 }, { "epoch": 1.059717868338558, "grad_norm": 6.244968891143799, "learning_rate": 8.382608978479281e-06, "loss": 1.6593, "step": 13522 }, { "epoch": 1.059796238244514, "grad_norm": 3.438218116760254, "learning_rate": 8.38238188725463e-06, "loss": 1.4036, "step": 13523 }, { "epoch": 1.0598746081504702, "grad_norm": 2.2523152828216553, "learning_rate": 8.382154783165135e-06, "loss": 1.478, "step": 13524 }, { "epoch": 1.0599529780564263, "grad_norm": 6.431358814239502, "learning_rate": 8.381927666211663e-06, "loss": 1.263, "step": 13525 }, { "epoch": 1.0600313479623824, "grad_norm": 2.5975935459136963, "learning_rate": 8.381700536395072e-06, "loss": 1.4885, "step": 13526 }, { "epoch": 1.0601097178683385, "grad_norm": 8.174845695495605, "learning_rate": 8.38147339371623e-06, "loss": 1.173, "step": 13527 }, { "epoch": 1.0601880877742946, "grad_norm": 2.711179256439209, "learning_rate": 8.381246238176001e-06, "loss": 1.2566, "step": 13528 }, { "epoch": 1.0602664576802507, "grad_norm": 3.3808977603912354, "learning_rate": 8.381019069775249e-06, "loss": 1.4756, "step": 13529 }, { "epoch": 1.0603448275862069, "grad_norm": 3.97060227394104, "learning_rate": 8.380791888514834e-06, "loss": 1.4435, "step": 13530 }, { "epoch": 1.060423197492163, "grad_norm": 3.1595470905303955, "learning_rate": 8.380564694395625e-06, "loss": 1.5959, "step": 13531 }, { "epoch": 1.060501567398119, "grad_norm": 2.8705759048461914, "learning_rate": 8.380337487418484e-06, "loss": 1.068, "step": 13532 }, { "epoch": 1.0605799373040752, "grad_norm": 2.9972846508026123, "learning_rate": 8.380110267584276e-06, "loss": 1.2188, "step": 13533 }, { "epoch": 1.0606583072100313, "grad_norm": 2.1748692989349365, "learning_rate": 8.379883034893865e-06, "loss": 0.8172, "step": 13534 }, { "epoch": 1.0607366771159874, "grad_norm": 2.2330446243286133, "learning_rate": 8.379655789348114e-06, "loss": 1.3514, "step": 13535 }, { "epoch": 1.0608150470219435, "grad_norm": 2.1216182708740234, "learning_rate": 8.379428530947888e-06, "loss": 1.2422, "step": 13536 }, { "epoch": 1.0608934169278996, "grad_norm": 3.6388938426971436, "learning_rate": 8.379201259694053e-06, "loss": 1.6604, "step": 13537 }, { "epoch": 1.0609717868338557, "grad_norm": 2.439112663269043, "learning_rate": 8.378973975587471e-06, "loss": 1.4167, "step": 13538 }, { "epoch": 1.0610501567398118, "grad_norm": 2.9501953125, "learning_rate": 8.378746678629007e-06, "loss": 1.5837, "step": 13539 }, { "epoch": 1.061128526645768, "grad_norm": 2.4232850074768066, "learning_rate": 8.378519368819528e-06, "loss": 1.436, "step": 13540 }, { "epoch": 1.061206896551724, "grad_norm": 2.6122689247131348, "learning_rate": 8.378292046159894e-06, "loss": 1.3687, "step": 13541 }, { "epoch": 1.0612852664576802, "grad_norm": 2.725621223449707, "learning_rate": 8.378064710650975e-06, "loss": 1.7137, "step": 13542 }, { "epoch": 1.0613636363636363, "grad_norm": 2.9898698329925537, "learning_rate": 8.377837362293631e-06, "loss": 1.1836, "step": 13543 }, { "epoch": 1.0614420062695924, "grad_norm": 2.576145887374878, "learning_rate": 8.37761000108873e-06, "loss": 1.2669, "step": 13544 }, { "epoch": 1.0615203761755485, "grad_norm": 2.613152503967285, "learning_rate": 8.377382627037133e-06, "loss": 1.1167, "step": 13545 }, { "epoch": 1.0615987460815046, "grad_norm": 2.3962600231170654, "learning_rate": 8.37715524013971e-06, "loss": 1.4905, "step": 13546 }, { "epoch": 1.0616771159874607, "grad_norm": 2.5007903575897217, "learning_rate": 8.376927840397322e-06, "loss": 1.3001, "step": 13547 }, { "epoch": 1.0617554858934168, "grad_norm": 3.265901803970337, "learning_rate": 8.376700427810835e-06, "loss": 1.0646, "step": 13548 }, { "epoch": 1.061833855799373, "grad_norm": 5.116260051727295, "learning_rate": 8.376473002381114e-06, "loss": 1.5558, "step": 13549 }, { "epoch": 1.061912225705329, "grad_norm": 2.7369537353515625, "learning_rate": 8.376245564109024e-06, "loss": 1.2545, "step": 13550 }, { "epoch": 1.0619905956112852, "grad_norm": 3.4161107540130615, "learning_rate": 8.37601811299543e-06, "loss": 1.1784, "step": 13551 }, { "epoch": 1.0620689655172413, "grad_norm": 2.7192888259887695, "learning_rate": 8.375790649041198e-06, "loss": 1.3159, "step": 13552 }, { "epoch": 1.0621473354231974, "grad_norm": 2.7240283489227295, "learning_rate": 8.375563172247191e-06, "loss": 1.2638, "step": 13553 }, { "epoch": 1.0622257053291535, "grad_norm": 3.059149742126465, "learning_rate": 8.375335682614276e-06, "loss": 1.0917, "step": 13554 }, { "epoch": 1.0623040752351096, "grad_norm": 3.136648654937744, "learning_rate": 8.375108180143317e-06, "loss": 1.5157, "step": 13555 }, { "epoch": 1.0623824451410657, "grad_norm": 2.116395950317383, "learning_rate": 8.374880664835181e-06, "loss": 1.4978, "step": 13556 }, { "epoch": 1.062460815047022, "grad_norm": 2.970625638961792, "learning_rate": 8.37465313669073e-06, "loss": 1.6463, "step": 13557 }, { "epoch": 1.062539184952978, "grad_norm": 6.251101493835449, "learning_rate": 8.374425595710836e-06, "loss": 0.6219, "step": 13558 }, { "epoch": 1.0626175548589343, "grad_norm": 7.587581157684326, "learning_rate": 8.374198041896357e-06, "loss": 1.3133, "step": 13559 }, { "epoch": 1.0626959247648902, "grad_norm": 4.210686683654785, "learning_rate": 8.373970475248163e-06, "loss": 1.1783, "step": 13560 }, { "epoch": 1.0627742946708465, "grad_norm": 21.345666885375977, "learning_rate": 8.373742895767119e-06, "loss": 1.7496, "step": 13561 }, { "epoch": 1.0628526645768026, "grad_norm": 7.357020854949951, "learning_rate": 8.37351530345409e-06, "loss": 0.5446, "step": 13562 }, { "epoch": 1.0629310344827587, "grad_norm": 5.122198104858398, "learning_rate": 8.37328769830994e-06, "loss": 1.2927, "step": 13563 }, { "epoch": 1.0630094043887148, "grad_norm": 2.7821686267852783, "learning_rate": 8.373060080335537e-06, "loss": 1.4203, "step": 13564 }, { "epoch": 1.063087774294671, "grad_norm": 2.955880880355835, "learning_rate": 8.372832449531745e-06, "loss": 0.9732, "step": 13565 }, { "epoch": 1.063166144200627, "grad_norm": 3.5324015617370605, "learning_rate": 8.372604805899431e-06, "loss": 1.2783, "step": 13566 }, { "epoch": 1.0632445141065832, "grad_norm": 6.783792972564697, "learning_rate": 8.372377149439462e-06, "loss": 1.5071, "step": 13567 }, { "epoch": 1.0633228840125393, "grad_norm": 2.3116455078125, "learning_rate": 8.372149480152701e-06, "loss": 1.4006, "step": 13568 }, { "epoch": 1.0634012539184954, "grad_norm": 7.104086875915527, "learning_rate": 8.371921798040017e-06, "loss": 1.2292, "step": 13569 }, { "epoch": 1.0634796238244515, "grad_norm": 3.5461161136627197, "learning_rate": 8.371694103102272e-06, "loss": 1.456, "step": 13570 }, { "epoch": 1.0635579937304076, "grad_norm": 1.9514427185058594, "learning_rate": 8.371466395340336e-06, "loss": 1.3545, "step": 13571 }, { "epoch": 1.0636363636363637, "grad_norm": 3.3442962169647217, "learning_rate": 8.371238674755073e-06, "loss": 1.3544, "step": 13572 }, { "epoch": 1.0637147335423198, "grad_norm": 2.6557130813598633, "learning_rate": 8.371010941347351e-06, "loss": 1.8079, "step": 13573 }, { "epoch": 1.063793103448276, "grad_norm": 2.99833345413208, "learning_rate": 8.370783195118034e-06, "loss": 1.3687, "step": 13574 }, { "epoch": 1.063871473354232, "grad_norm": 4.292763710021973, "learning_rate": 8.370555436067988e-06, "loss": 1.3709, "step": 13575 }, { "epoch": 1.0639498432601882, "grad_norm": 3.2954986095428467, "learning_rate": 8.370327664198083e-06, "loss": 1.4525, "step": 13576 }, { "epoch": 1.0640282131661443, "grad_norm": 3.3279712200164795, "learning_rate": 8.370099879509182e-06, "loss": 1.4876, "step": 13577 }, { "epoch": 1.0641065830721004, "grad_norm": 2.7554335594177246, "learning_rate": 8.36987208200215e-06, "loss": 1.1586, "step": 13578 }, { "epoch": 1.0641849529780565, "grad_norm": 10.183682441711426, "learning_rate": 8.369644271677856e-06, "loss": 1.2667, "step": 13579 }, { "epoch": 1.0642633228840126, "grad_norm": 2.276954174041748, "learning_rate": 8.369416448537167e-06, "loss": 1.2165, "step": 13580 }, { "epoch": 1.0643416927899687, "grad_norm": 4.57388162612915, "learning_rate": 8.369188612580948e-06, "loss": 1.2935, "step": 13581 }, { "epoch": 1.0644200626959248, "grad_norm": 2.8942699432373047, "learning_rate": 8.368960763810066e-06, "loss": 1.0427, "step": 13582 }, { "epoch": 1.064498432601881, "grad_norm": 2.5967540740966797, "learning_rate": 8.368732902225386e-06, "loss": 1.6906, "step": 13583 }, { "epoch": 1.064576802507837, "grad_norm": 4.586245059967041, "learning_rate": 8.368505027827778e-06, "loss": 1.2918, "step": 13584 }, { "epoch": 1.0646551724137931, "grad_norm": 3.309760570526123, "learning_rate": 8.36827714061811e-06, "loss": 1.356, "step": 13585 }, { "epoch": 1.0647335423197493, "grad_norm": 12.658743858337402, "learning_rate": 8.36804924059724e-06, "loss": 1.4693, "step": 13586 }, { "epoch": 1.0648119122257054, "grad_norm": 3.0100014209747314, "learning_rate": 8.367821327766044e-06, "loss": 1.4392, "step": 13587 }, { "epoch": 1.0648902821316615, "grad_norm": 6.652431011199951, "learning_rate": 8.367593402125385e-06, "loss": 1.4546, "step": 13588 }, { "epoch": 1.0649686520376176, "grad_norm": 4.449180603027344, "learning_rate": 8.367365463676132e-06, "loss": 0.7417, "step": 13589 }, { "epoch": 1.0650470219435737, "grad_norm": 2.1123855113983154, "learning_rate": 8.367137512419147e-06, "loss": 1.2342, "step": 13590 }, { "epoch": 1.0651253918495298, "grad_norm": 3.7264204025268555, "learning_rate": 8.366909548355302e-06, "loss": 1.0858, "step": 13591 }, { "epoch": 1.065203761755486, "grad_norm": 2.579638957977295, "learning_rate": 8.366681571485463e-06, "loss": 1.4537, "step": 13592 }, { "epoch": 1.065282131661442, "grad_norm": 6.741729259490967, "learning_rate": 8.366453581810496e-06, "loss": 1.9278, "step": 13593 }, { "epoch": 1.0653605015673981, "grad_norm": 5.450695037841797, "learning_rate": 8.366225579331269e-06, "loss": 1.7443, "step": 13594 }, { "epoch": 1.0654388714733543, "grad_norm": 2.149749279022217, "learning_rate": 8.36599756404865e-06, "loss": 1.5711, "step": 13595 }, { "epoch": 1.0655172413793104, "grad_norm": 4.364523887634277, "learning_rate": 8.365769535963503e-06, "loss": 1.4977, "step": 13596 }, { "epoch": 1.0655956112852665, "grad_norm": 1.8416500091552734, "learning_rate": 8.365541495076697e-06, "loss": 0.9978, "step": 13597 }, { "epoch": 1.0656739811912226, "grad_norm": 4.143197536468506, "learning_rate": 8.365313441389103e-06, "loss": 1.303, "step": 13598 }, { "epoch": 1.0657523510971787, "grad_norm": 3.212796211242676, "learning_rate": 8.365085374901581e-06, "loss": 1.1326, "step": 13599 }, { "epoch": 1.0658307210031348, "grad_norm": 2.9330332279205322, "learning_rate": 8.364857295615006e-06, "loss": 1.3338, "step": 13600 }, { "epoch": 1.065909090909091, "grad_norm": 2.955617904663086, "learning_rate": 8.36462920353024e-06, "loss": 1.3381, "step": 13601 }, { "epoch": 1.065987460815047, "grad_norm": 3.3764994144439697, "learning_rate": 8.364401098648154e-06, "loss": 1.3397, "step": 13602 }, { "epoch": 1.0660658307210031, "grad_norm": 2.056213140487671, "learning_rate": 8.364172980969616e-06, "loss": 1.3536, "step": 13603 }, { "epoch": 1.0661442006269592, "grad_norm": 2.8142874240875244, "learning_rate": 8.36394485049549e-06, "loss": 1.5487, "step": 13604 }, { "epoch": 1.0662225705329154, "grad_norm": 2.7529118061065674, "learning_rate": 8.363716707226644e-06, "loss": 1.2985, "step": 13605 }, { "epoch": 1.0663009404388715, "grad_norm": 5.918830394744873, "learning_rate": 8.36348855116395e-06, "loss": 1.3453, "step": 13606 }, { "epoch": 1.0663793103448276, "grad_norm": 3.189504861831665, "learning_rate": 8.363260382308272e-06, "loss": 1.0926, "step": 13607 }, { "epoch": 1.0664576802507837, "grad_norm": 1.8518251180648804, "learning_rate": 8.36303220066048e-06, "loss": 1.3314, "step": 13608 }, { "epoch": 1.0665360501567398, "grad_norm": 2.861219882965088, "learning_rate": 8.362804006221438e-06, "loss": 1.2739, "step": 13609 }, { "epoch": 1.066614420062696, "grad_norm": 3.955817461013794, "learning_rate": 8.362575798992019e-06, "loss": 1.5421, "step": 13610 }, { "epoch": 1.066692789968652, "grad_norm": 3.428011655807495, "learning_rate": 8.36234757897309e-06, "loss": 0.8156, "step": 13611 }, { "epoch": 1.0667711598746081, "grad_norm": 2.797098159790039, "learning_rate": 8.362119346165514e-06, "loss": 1.589, "step": 13612 }, { "epoch": 1.0668495297805642, "grad_norm": 1.8044553995132446, "learning_rate": 8.361891100570166e-06, "loss": 1.1987, "step": 13613 }, { "epoch": 1.0669278996865204, "grad_norm": 3.179899215698242, "learning_rate": 8.36166284218791e-06, "loss": 1.88, "step": 13614 }, { "epoch": 1.0670062695924765, "grad_norm": 2.3665096759796143, "learning_rate": 8.361434571019617e-06, "loss": 1.1, "step": 13615 }, { "epoch": 1.0670846394984326, "grad_norm": 3.523472309112549, "learning_rate": 8.361206287066151e-06, "loss": 1.5985, "step": 13616 }, { "epoch": 1.0671630094043887, "grad_norm": 2.3544399738311768, "learning_rate": 8.360977990328385e-06, "loss": 1.0761, "step": 13617 }, { "epoch": 1.0672413793103448, "grad_norm": 4.088952541351318, "learning_rate": 8.360749680807183e-06, "loss": 1.0392, "step": 13618 }, { "epoch": 1.067319749216301, "grad_norm": 3.0537426471710205, "learning_rate": 8.360521358503415e-06, "loss": 1.2694, "step": 13619 }, { "epoch": 1.067398119122257, "grad_norm": 1.940498948097229, "learning_rate": 8.360293023417951e-06, "loss": 1.2645, "step": 13620 }, { "epoch": 1.0674764890282131, "grad_norm": 2.544529438018799, "learning_rate": 8.360064675551659e-06, "loss": 1.2053, "step": 13621 }, { "epoch": 1.0675548589341692, "grad_norm": 2.5339462757110596, "learning_rate": 8.359836314905406e-06, "loss": 1.0345, "step": 13622 }, { "epoch": 1.0676332288401253, "grad_norm": 4.259823799133301, "learning_rate": 8.359607941480061e-06, "loss": 1.3555, "step": 13623 }, { "epoch": 1.0677115987460815, "grad_norm": 2.986360788345337, "learning_rate": 8.359379555276495e-06, "loss": 1.4152, "step": 13624 }, { "epoch": 1.0677899686520376, "grad_norm": 2.5660173892974854, "learning_rate": 8.35915115629557e-06, "loss": 1.389, "step": 13625 }, { "epoch": 1.0678683385579937, "grad_norm": 3.7810657024383545, "learning_rate": 8.358922744538165e-06, "loss": 1.1947, "step": 13626 }, { "epoch": 1.0679467084639498, "grad_norm": 2.233781337738037, "learning_rate": 8.35869432000514e-06, "loss": 1.0216, "step": 13627 }, { "epoch": 1.068025078369906, "grad_norm": 2.366393804550171, "learning_rate": 8.358465882697367e-06, "loss": 1.4165, "step": 13628 }, { "epoch": 1.068103448275862, "grad_norm": 2.9092752933502197, "learning_rate": 8.358237432615715e-06, "loss": 1.3867, "step": 13629 }, { "epoch": 1.0681818181818181, "grad_norm": 2.7984349727630615, "learning_rate": 8.358008969761054e-06, "loss": 1.7302, "step": 13630 }, { "epoch": 1.0682601880877742, "grad_norm": 3.2883622646331787, "learning_rate": 8.357780494134251e-06, "loss": 1.142, "step": 13631 }, { "epoch": 1.0683385579937303, "grad_norm": 8.190718650817871, "learning_rate": 8.357552005736173e-06, "loss": 1.6099, "step": 13632 }, { "epoch": 1.0684169278996865, "grad_norm": 6.567223072052002, "learning_rate": 8.357323504567696e-06, "loss": 0.9983, "step": 13633 }, { "epoch": 1.0684952978056426, "grad_norm": 4.173842430114746, "learning_rate": 8.357094990629682e-06, "loss": 1.0264, "step": 13634 }, { "epoch": 1.0685736677115987, "grad_norm": 4.450650215148926, "learning_rate": 8.356866463923005e-06, "loss": 1.3854, "step": 13635 }, { "epoch": 1.0686520376175548, "grad_norm": 5.055824279785156, "learning_rate": 8.35663792444853e-06, "loss": 0.987, "step": 13636 }, { "epoch": 1.068730407523511, "grad_norm": 4.384795188903809, "learning_rate": 8.35640937220713e-06, "loss": 1.3534, "step": 13637 }, { "epoch": 1.068808777429467, "grad_norm": 2.990509510040283, "learning_rate": 8.356180807199671e-06, "loss": 1.1458, "step": 13638 }, { "epoch": 1.0688871473354231, "grad_norm": 2.7878096103668213, "learning_rate": 8.355952229427027e-06, "loss": 1.5129, "step": 13639 }, { "epoch": 1.0689655172413792, "grad_norm": 3.6362955570220947, "learning_rate": 8.35572363889006e-06, "loss": 1.777, "step": 13640 }, { "epoch": 1.0690438871473353, "grad_norm": 8.206583023071289, "learning_rate": 8.355495035589647e-06, "loss": 0.7073, "step": 13641 }, { "epoch": 1.0691222570532914, "grad_norm": 2.4999399185180664, "learning_rate": 8.355266419526655e-06, "loss": 1.3961, "step": 13642 }, { "epoch": 1.0692006269592476, "grad_norm": 4.005369663238525, "learning_rate": 8.355037790701951e-06, "loss": 1.2046, "step": 13643 }, { "epoch": 1.0692789968652037, "grad_norm": 3.601876974105835, "learning_rate": 8.354809149116408e-06, "loss": 1.415, "step": 13644 }, { "epoch": 1.0693573667711598, "grad_norm": 2.520338773727417, "learning_rate": 8.354580494770894e-06, "loss": 1.3241, "step": 13645 }, { "epoch": 1.069435736677116, "grad_norm": 11.554008483886719, "learning_rate": 8.354351827666278e-06, "loss": 1.6022, "step": 13646 }, { "epoch": 1.069514106583072, "grad_norm": 5.410002708435059, "learning_rate": 8.35412314780343e-06, "loss": 1.4233, "step": 13647 }, { "epoch": 1.069592476489028, "grad_norm": 3.8595592975616455, "learning_rate": 8.353894455183222e-06, "loss": 1.8184, "step": 13648 }, { "epoch": 1.0696708463949842, "grad_norm": 8.32068157196045, "learning_rate": 8.353665749806521e-06, "loss": 1.4055, "step": 13649 }, { "epoch": 1.0697492163009406, "grad_norm": 6.60568380355835, "learning_rate": 8.353437031674199e-06, "loss": 1.4048, "step": 13650 }, { "epoch": 1.0698275862068964, "grad_norm": 3.2727315425872803, "learning_rate": 8.353208300787124e-06, "loss": 1.6588, "step": 13651 }, { "epoch": 1.0699059561128528, "grad_norm": 2.566262722015381, "learning_rate": 8.352979557146169e-06, "loss": 1.409, "step": 13652 }, { "epoch": 1.0699843260188087, "grad_norm": 3.218101739883423, "learning_rate": 8.3527508007522e-06, "loss": 1.7011, "step": 13653 }, { "epoch": 1.070062695924765, "grad_norm": 10.403045654296875, "learning_rate": 8.352522031606088e-06, "loss": 1.9211, "step": 13654 }, { "epoch": 1.070141065830721, "grad_norm": 2.1903350353240967, "learning_rate": 8.352293249708705e-06, "loss": 1.6854, "step": 13655 }, { "epoch": 1.0702194357366772, "grad_norm": 3.1212470531463623, "learning_rate": 8.352064455060922e-06, "loss": 1.452, "step": 13656 }, { "epoch": 1.0702978056426333, "grad_norm": 4.680207252502441, "learning_rate": 8.351835647663605e-06, "loss": 1.2105, "step": 13657 }, { "epoch": 1.0703761755485894, "grad_norm": 2.2651145458221436, "learning_rate": 8.35160682751763e-06, "loss": 1.3287, "step": 13658 }, { "epoch": 1.0704545454545455, "grad_norm": 3.8634328842163086, "learning_rate": 8.35137799462386e-06, "loss": 1.3355, "step": 13659 }, { "epoch": 1.0705329153605017, "grad_norm": 2.2362072467803955, "learning_rate": 8.351149148983173e-06, "loss": 1.6046, "step": 13660 }, { "epoch": 1.0706112852664578, "grad_norm": 1.9721869230270386, "learning_rate": 8.350920290596435e-06, "loss": 1.4762, "step": 13661 }, { "epoch": 1.0706896551724139, "grad_norm": 3.1311357021331787, "learning_rate": 8.350691419464517e-06, "loss": 1.3978, "step": 13662 }, { "epoch": 1.07076802507837, "grad_norm": 2.7234442234039307, "learning_rate": 8.350462535588289e-06, "loss": 1.3156, "step": 13663 }, { "epoch": 1.070846394984326, "grad_norm": 2.2054173946380615, "learning_rate": 8.350233638968622e-06, "loss": 1.4735, "step": 13664 }, { "epoch": 1.0709247648902822, "grad_norm": 2.749271869659424, "learning_rate": 8.350004729606389e-06, "loss": 0.9065, "step": 13665 }, { "epoch": 1.0710031347962383, "grad_norm": 4.311010837554932, "learning_rate": 8.349775807502458e-06, "loss": 1.2097, "step": 13666 }, { "epoch": 1.0710815047021944, "grad_norm": 3.1267507076263428, "learning_rate": 8.3495468726577e-06, "loss": 1.7182, "step": 13667 }, { "epoch": 1.0711598746081505, "grad_norm": 2.2427632808685303, "learning_rate": 8.349317925072986e-06, "loss": 1.2464, "step": 13668 }, { "epoch": 1.0712382445141067, "grad_norm": 3.3870480060577393, "learning_rate": 8.349088964749188e-06, "loss": 1.4493, "step": 13669 }, { "epoch": 1.0713166144200628, "grad_norm": 3.3571853637695312, "learning_rate": 8.348859991687174e-06, "loss": 1.239, "step": 13670 }, { "epoch": 1.0713949843260189, "grad_norm": 4.0068230628967285, "learning_rate": 8.348631005887819e-06, "loss": 1.8401, "step": 13671 }, { "epoch": 1.071473354231975, "grad_norm": 4.462873458862305, "learning_rate": 8.34840200735199e-06, "loss": 1.5525, "step": 13672 }, { "epoch": 1.071551724137931, "grad_norm": 3.326732635498047, "learning_rate": 8.34817299608056e-06, "loss": 1.5439, "step": 13673 }, { "epoch": 1.0716300940438872, "grad_norm": 4.24921178817749, "learning_rate": 8.3479439720744e-06, "loss": 1.5921, "step": 13674 }, { "epoch": 1.0717084639498433, "grad_norm": 2.671997547149658, "learning_rate": 8.347714935334382e-06, "loss": 1.7249, "step": 13675 }, { "epoch": 1.0717868338557994, "grad_norm": 2.6442739963531494, "learning_rate": 8.347485885861373e-06, "loss": 1.2987, "step": 13676 }, { "epoch": 1.0718652037617555, "grad_norm": 2.121241331100464, "learning_rate": 8.34725682365625e-06, "loss": 1.6509, "step": 13677 }, { "epoch": 1.0719435736677116, "grad_norm": 2.5248653888702393, "learning_rate": 8.347027748719878e-06, "loss": 1.3131, "step": 13678 }, { "epoch": 1.0720219435736678, "grad_norm": 3.01493239402771, "learning_rate": 8.346798661053132e-06, "loss": 1.3144, "step": 13679 }, { "epoch": 1.0721003134796239, "grad_norm": 2.1490085124969482, "learning_rate": 8.346569560656885e-06, "loss": 1.1684, "step": 13680 }, { "epoch": 1.07217868338558, "grad_norm": 2.3106436729431152, "learning_rate": 8.346340447532005e-06, "loss": 1.5394, "step": 13681 }, { "epoch": 1.072257053291536, "grad_norm": 3.8903415203094482, "learning_rate": 8.346111321679367e-06, "loss": 1.351, "step": 13682 }, { "epoch": 1.0723354231974922, "grad_norm": 2.7653894424438477, "learning_rate": 8.345882183099838e-06, "loss": 1.3925, "step": 13683 }, { "epoch": 1.0724137931034483, "grad_norm": 2.7212629318237305, "learning_rate": 8.345653031794292e-06, "loss": 1.6154, "step": 13684 }, { "epoch": 1.0724921630094044, "grad_norm": 6.132538318634033, "learning_rate": 8.3454238677636e-06, "loss": 1.6732, "step": 13685 }, { "epoch": 1.0725705329153605, "grad_norm": 4.217459201812744, "learning_rate": 8.345194691008634e-06, "loss": 1.6729, "step": 13686 }, { "epoch": 1.0726489028213166, "grad_norm": 3.490267753601074, "learning_rate": 8.344965501530266e-06, "loss": 1.3943, "step": 13687 }, { "epoch": 1.0727272727272728, "grad_norm": 3.608258008956909, "learning_rate": 8.344736299329368e-06, "loss": 1.0065, "step": 13688 }, { "epoch": 1.0728056426332289, "grad_norm": 1.715820074081421, "learning_rate": 8.34450708440681e-06, "loss": 1.6335, "step": 13689 }, { "epoch": 1.072884012539185, "grad_norm": 2.284259080886841, "learning_rate": 8.344277856763465e-06, "loss": 1.5623, "step": 13690 }, { "epoch": 1.072962382445141, "grad_norm": 2.3966615200042725, "learning_rate": 8.344048616400207e-06, "loss": 1.3839, "step": 13691 }, { "epoch": 1.0730407523510972, "grad_norm": 7.239594459533691, "learning_rate": 8.343819363317901e-06, "loss": 1.2057, "step": 13692 }, { "epoch": 1.0731191222570533, "grad_norm": 2.197791576385498, "learning_rate": 8.343590097517428e-06, "loss": 1.4854, "step": 13693 }, { "epoch": 1.0731974921630094, "grad_norm": 3.4979910850524902, "learning_rate": 8.343360818999653e-06, "loss": 1.0898, "step": 13694 }, { "epoch": 1.0732758620689655, "grad_norm": 5.003958702087402, "learning_rate": 8.343131527765453e-06, "loss": 1.3717, "step": 13695 }, { "epoch": 1.0733542319749216, "grad_norm": 2.8413825035095215, "learning_rate": 8.342902223815696e-06, "loss": 1.4796, "step": 13696 }, { "epoch": 1.0734326018808777, "grad_norm": 2.5945932865142822, "learning_rate": 8.342672907151257e-06, "loss": 1.621, "step": 13697 }, { "epoch": 1.0735109717868339, "grad_norm": 4.023177623748779, "learning_rate": 8.342443577773006e-06, "loss": 1.4556, "step": 13698 }, { "epoch": 1.07358934169279, "grad_norm": 7.170156002044678, "learning_rate": 8.342214235681818e-06, "loss": 1.4984, "step": 13699 }, { "epoch": 1.073667711598746, "grad_norm": 2.169957160949707, "learning_rate": 8.341984880878561e-06, "loss": 1.3875, "step": 13700 }, { "epoch": 1.0737460815047022, "grad_norm": 2.3748772144317627, "learning_rate": 8.341755513364113e-06, "loss": 1.3687, "step": 13701 }, { "epoch": 1.0738244514106583, "grad_norm": 3.088357448577881, "learning_rate": 8.341526133139343e-06, "loss": 1.3495, "step": 13702 }, { "epoch": 1.0739028213166144, "grad_norm": 2.803605079650879, "learning_rate": 8.341296740205123e-06, "loss": 1.0803, "step": 13703 }, { "epoch": 1.0739811912225705, "grad_norm": 2.7861552238464355, "learning_rate": 8.341067334562327e-06, "loss": 1.2505, "step": 13704 }, { "epoch": 1.0740595611285266, "grad_norm": 2.1292152404785156, "learning_rate": 8.340837916211827e-06, "loss": 1.3928, "step": 13705 }, { "epoch": 1.0741379310344827, "grad_norm": 3.441039562225342, "learning_rate": 8.340608485154493e-06, "loss": 1.4309, "step": 13706 }, { "epoch": 1.0742163009404389, "grad_norm": 2.005946159362793, "learning_rate": 8.340379041391202e-06, "loss": 1.3968, "step": 13707 }, { "epoch": 1.074294670846395, "grad_norm": 4.07476282119751, "learning_rate": 8.340149584922826e-06, "loss": 1.1337, "step": 13708 }, { "epoch": 1.074373040752351, "grad_norm": 4.158647060394287, "learning_rate": 8.339920115750235e-06, "loss": 1.9068, "step": 13709 }, { "epoch": 1.0744514106583072, "grad_norm": 3.008772373199463, "learning_rate": 8.339690633874304e-06, "loss": 1.0091, "step": 13710 }, { "epoch": 1.0745297805642633, "grad_norm": 3.600313663482666, "learning_rate": 8.339461139295905e-06, "loss": 1.6838, "step": 13711 }, { "epoch": 1.0746081504702194, "grad_norm": 2.2034835815429688, "learning_rate": 8.339231632015913e-06, "loss": 1.4198, "step": 13712 }, { "epoch": 1.0746865203761755, "grad_norm": 3.4800894260406494, "learning_rate": 8.339002112035198e-06, "loss": 1.4305, "step": 13713 }, { "epoch": 1.0747648902821316, "grad_norm": 2.851292610168457, "learning_rate": 8.338772579354633e-06, "loss": 1.1733, "step": 13714 }, { "epoch": 1.0748432601880877, "grad_norm": 2.656980037689209, "learning_rate": 8.338543033975092e-06, "loss": 1.3074, "step": 13715 }, { "epoch": 1.0749216300940438, "grad_norm": 2.5283164978027344, "learning_rate": 8.338313475897448e-06, "loss": 1.2388, "step": 13716 }, { "epoch": 1.075, "grad_norm": 2.8164165019989014, "learning_rate": 8.338083905122576e-06, "loss": 1.3608, "step": 13717 }, { "epoch": 1.075078369905956, "grad_norm": 2.543213367462158, "learning_rate": 8.337854321651344e-06, "loss": 1.2216, "step": 13718 }, { "epoch": 1.0751567398119122, "grad_norm": 3.9613404273986816, "learning_rate": 8.337624725484633e-06, "loss": 1.2653, "step": 13719 }, { "epoch": 1.0752351097178683, "grad_norm": 2.648202657699585, "learning_rate": 8.337395116623308e-06, "loss": 1.5161, "step": 13720 }, { "epoch": 1.0753134796238244, "grad_norm": 2.001361131668091, "learning_rate": 8.337165495068247e-06, "loss": 1.3106, "step": 13721 }, { "epoch": 1.0753918495297805, "grad_norm": 3.239062786102295, "learning_rate": 8.336935860820323e-06, "loss": 1.3416, "step": 13722 }, { "epoch": 1.0754702194357366, "grad_norm": 3.014568567276001, "learning_rate": 8.336706213880411e-06, "loss": 1.5622, "step": 13723 }, { "epoch": 1.0755485893416927, "grad_norm": 2.580918788909912, "learning_rate": 8.33647655424938e-06, "loss": 1.5603, "step": 13724 }, { "epoch": 1.0756269592476488, "grad_norm": 4.962240219116211, "learning_rate": 8.336246881928105e-06, "loss": 1.3857, "step": 13725 }, { "epoch": 1.075705329153605, "grad_norm": 13.85176944732666, "learning_rate": 8.336017196917462e-06, "loss": 1.522, "step": 13726 }, { "epoch": 1.075783699059561, "grad_norm": 5.419173717498779, "learning_rate": 8.335787499218324e-06, "loss": 0.9511, "step": 13727 }, { "epoch": 1.0758620689655172, "grad_norm": 3.1340818405151367, "learning_rate": 8.335557788831563e-06, "loss": 1.2414, "step": 13728 }, { "epoch": 1.0759404388714733, "grad_norm": 4.67362117767334, "learning_rate": 8.335328065758052e-06, "loss": 1.415, "step": 13729 }, { "epoch": 1.0760188087774294, "grad_norm": 2.994696617126465, "learning_rate": 8.335098329998668e-06, "loss": 1.216, "step": 13730 }, { "epoch": 1.0760971786833855, "grad_norm": 4.717992782592773, "learning_rate": 8.334868581554281e-06, "loss": 1.3442, "step": 13731 }, { "epoch": 1.0761755485893416, "grad_norm": 4.678876876831055, "learning_rate": 8.33463882042577e-06, "loss": 1.5446, "step": 13732 }, { "epoch": 1.0762539184952977, "grad_norm": 3.389692783355713, "learning_rate": 8.334409046614003e-06, "loss": 1.1838, "step": 13733 }, { "epoch": 1.0763322884012538, "grad_norm": 3.3841640949249268, "learning_rate": 8.334179260119857e-06, "loss": 1.2733, "step": 13734 }, { "epoch": 1.07641065830721, "grad_norm": 2.043687343597412, "learning_rate": 8.333949460944206e-06, "loss": 1.0251, "step": 13735 }, { "epoch": 1.076489028213166, "grad_norm": 4.490939140319824, "learning_rate": 8.333719649087924e-06, "loss": 1.2555, "step": 13736 }, { "epoch": 1.0765673981191222, "grad_norm": 11.066393852233887, "learning_rate": 8.333489824551884e-06, "loss": 1.8585, "step": 13737 }, { "epoch": 1.0766457680250783, "grad_norm": 3.490720272064209, "learning_rate": 8.333259987336962e-06, "loss": 1.4826, "step": 13738 }, { "epoch": 1.0767241379310344, "grad_norm": 2.5779364109039307, "learning_rate": 8.33303013744403e-06, "loss": 1.1921, "step": 13739 }, { "epoch": 1.0768025078369905, "grad_norm": 5.797022342681885, "learning_rate": 8.332800274873964e-06, "loss": 1.5407, "step": 13740 }, { "epoch": 1.0768808777429466, "grad_norm": 2.0296645164489746, "learning_rate": 8.332570399627637e-06, "loss": 0.9374, "step": 13741 }, { "epoch": 1.0769592476489027, "grad_norm": 3.4888744354248047, "learning_rate": 8.332340511705925e-06, "loss": 1.5578, "step": 13742 }, { "epoch": 1.077037617554859, "grad_norm": 2.174391031265259, "learning_rate": 8.3321106111097e-06, "loss": 1.0583, "step": 13743 }, { "epoch": 1.077115987460815, "grad_norm": 6.243290424346924, "learning_rate": 8.331880697839838e-06, "loss": 1.2587, "step": 13744 }, { "epoch": 1.0771943573667713, "grad_norm": 2.272467851638794, "learning_rate": 8.331650771897215e-06, "loss": 1.3764, "step": 13745 }, { "epoch": 1.0772727272727272, "grad_norm": 2.690626382827759, "learning_rate": 8.3314208332827e-06, "loss": 1.0336, "step": 13746 }, { "epoch": 1.0773510971786835, "grad_norm": 4.17507791519165, "learning_rate": 8.331190881997175e-06, "loss": 1.1954, "step": 13747 }, { "epoch": 1.0774294670846394, "grad_norm": 7.383242130279541, "learning_rate": 8.330960918041509e-06, "loss": 1.1053, "step": 13748 }, { "epoch": 1.0775078369905957, "grad_norm": 3.58441162109375, "learning_rate": 8.330730941416578e-06, "loss": 0.8369, "step": 13749 }, { "epoch": 1.0775862068965518, "grad_norm": 2.4374940395355225, "learning_rate": 8.330500952123259e-06, "loss": 1.4995, "step": 13750 }, { "epoch": 1.077664576802508, "grad_norm": 4.684671401977539, "learning_rate": 8.330270950162424e-06, "loss": 1.7438, "step": 13751 }, { "epoch": 1.077742946708464, "grad_norm": 5.028148651123047, "learning_rate": 8.330040935534949e-06, "loss": 1.5859, "step": 13752 }, { "epoch": 1.0778213166144202, "grad_norm": 1.9548759460449219, "learning_rate": 8.329810908241708e-06, "loss": 1.6478, "step": 13753 }, { "epoch": 1.0778996865203763, "grad_norm": 2.1298956871032715, "learning_rate": 8.329580868283578e-06, "loss": 1.1754, "step": 13754 }, { "epoch": 1.0779780564263324, "grad_norm": 2.174729824066162, "learning_rate": 8.329350815661432e-06, "loss": 1.5078, "step": 13755 }, { "epoch": 1.0780564263322885, "grad_norm": 8.2648286819458, "learning_rate": 8.329120750376145e-06, "loss": 1.8092, "step": 13756 }, { "epoch": 1.0781347962382446, "grad_norm": 3.4642322063446045, "learning_rate": 8.328890672428593e-06, "loss": 1.4262, "step": 13757 }, { "epoch": 1.0782131661442007, "grad_norm": 3.607828378677368, "learning_rate": 8.328660581819651e-06, "loss": 1.1884, "step": 13758 }, { "epoch": 1.0782915360501568, "grad_norm": 3.9604666233062744, "learning_rate": 8.328430478550193e-06, "loss": 1.7672, "step": 13759 }, { "epoch": 1.078369905956113, "grad_norm": 7.307593822479248, "learning_rate": 8.328200362621097e-06, "loss": 1.8173, "step": 13760 }, { "epoch": 1.078448275862069, "grad_norm": 4.005928993225098, "learning_rate": 8.327970234033235e-06, "loss": 1.1778, "step": 13761 }, { "epoch": 1.0785266457680251, "grad_norm": 4.8446245193481445, "learning_rate": 8.327740092787485e-06, "loss": 1.313, "step": 13762 }, { "epoch": 1.0786050156739813, "grad_norm": 2.1951229572296143, "learning_rate": 8.327509938884718e-06, "loss": 1.4923, "step": 13763 }, { "epoch": 1.0786833855799374, "grad_norm": 4.917087078094482, "learning_rate": 8.327279772325814e-06, "loss": 0.7233, "step": 13764 }, { "epoch": 1.0787617554858935, "grad_norm": 5.0763444900512695, "learning_rate": 8.327049593111646e-06, "loss": 1.4104, "step": 13765 }, { "epoch": 1.0788401253918496, "grad_norm": 6.496805191040039, "learning_rate": 8.326819401243092e-06, "loss": 1.8028, "step": 13766 }, { "epoch": 1.0789184952978057, "grad_norm": 3.163224697113037, "learning_rate": 8.326589196721023e-06, "loss": 1.7304, "step": 13767 }, { "epoch": 1.0789968652037618, "grad_norm": 2.397050142288208, "learning_rate": 8.32635897954632e-06, "loss": 1.2417, "step": 13768 }, { "epoch": 1.079075235109718, "grad_norm": 2.867673397064209, "learning_rate": 8.326128749719854e-06, "loss": 1.5001, "step": 13769 }, { "epoch": 1.079153605015674, "grad_norm": 2.212068557739258, "learning_rate": 8.325898507242503e-06, "loss": 1.5554, "step": 13770 }, { "epoch": 1.0792319749216301, "grad_norm": 1.9481438398361206, "learning_rate": 8.325668252115145e-06, "loss": 1.5149, "step": 13771 }, { "epoch": 1.0793103448275863, "grad_norm": 2.3806369304656982, "learning_rate": 8.32543798433865e-06, "loss": 1.4684, "step": 13772 }, { "epoch": 1.0793887147335424, "grad_norm": 3.7826731204986572, "learning_rate": 8.325207703913899e-06, "loss": 1.6516, "step": 13773 }, { "epoch": 1.0794670846394985, "grad_norm": 8.61374568939209, "learning_rate": 8.324977410841763e-06, "loss": 1.717, "step": 13774 }, { "epoch": 1.0795454545454546, "grad_norm": 2.5129034519195557, "learning_rate": 8.324747105123123e-06, "loss": 1.533, "step": 13775 }, { "epoch": 1.0796238244514107, "grad_norm": 3.8253746032714844, "learning_rate": 8.324516786758852e-06, "loss": 1.6191, "step": 13776 }, { "epoch": 1.0797021943573668, "grad_norm": 3.3609330654144287, "learning_rate": 8.324286455749827e-06, "loss": 1.2918, "step": 13777 }, { "epoch": 1.079780564263323, "grad_norm": 4.614315032958984, "learning_rate": 8.324056112096923e-06, "loss": 1.2471, "step": 13778 }, { "epoch": 1.079858934169279, "grad_norm": 5.871201038360596, "learning_rate": 8.323825755801019e-06, "loss": 1.588, "step": 13779 }, { "epoch": 1.0799373040752351, "grad_norm": 2.7490298748016357, "learning_rate": 8.323595386862985e-06, "loss": 1.3064, "step": 13780 }, { "epoch": 1.0800156739811912, "grad_norm": 2.8209636211395264, "learning_rate": 8.323365005283705e-06, "loss": 1.3232, "step": 13781 }, { "epoch": 1.0800940438871474, "grad_norm": 2.821545124053955, "learning_rate": 8.323134611064052e-06, "loss": 1.3786, "step": 13782 }, { "epoch": 1.0801724137931035, "grad_norm": 2.385573148727417, "learning_rate": 8.3229042042049e-06, "loss": 1.4396, "step": 13783 }, { "epoch": 1.0802507836990596, "grad_norm": 2.515150547027588, "learning_rate": 8.322673784707127e-06, "loss": 1.2578, "step": 13784 }, { "epoch": 1.0803291536050157, "grad_norm": 2.296360492706299, "learning_rate": 8.322443352571609e-06, "loss": 1.1048, "step": 13785 }, { "epoch": 1.0804075235109718, "grad_norm": 2.4212100505828857, "learning_rate": 8.322212907799226e-06, "loss": 1.2066, "step": 13786 }, { "epoch": 1.080485893416928, "grad_norm": 2.3063743114471436, "learning_rate": 8.321982450390849e-06, "loss": 1.4163, "step": 13787 }, { "epoch": 1.080564263322884, "grad_norm": 3.8374643325805664, "learning_rate": 8.321751980347358e-06, "loss": 0.9221, "step": 13788 }, { "epoch": 1.0806426332288401, "grad_norm": 2.1712708473205566, "learning_rate": 8.321521497669626e-06, "loss": 1.2077, "step": 13789 }, { "epoch": 1.0807210031347962, "grad_norm": 4.9949774742126465, "learning_rate": 8.321291002358536e-06, "loss": 1.3758, "step": 13790 }, { "epoch": 1.0807993730407524, "grad_norm": 3.0397746562957764, "learning_rate": 8.321060494414959e-06, "loss": 1.6301, "step": 13791 }, { "epoch": 1.0808777429467085, "grad_norm": 3.5580408573150635, "learning_rate": 8.320829973839774e-06, "loss": 1.2599, "step": 13792 }, { "epoch": 1.0809561128526646, "grad_norm": 2.6317670345306396, "learning_rate": 8.320599440633856e-06, "loss": 1.5886, "step": 13793 }, { "epoch": 1.0810344827586207, "grad_norm": 3.5535812377929688, "learning_rate": 8.320368894798085e-06, "loss": 1.3892, "step": 13794 }, { "epoch": 1.0811128526645768, "grad_norm": 2.8587779998779297, "learning_rate": 8.320138336333337e-06, "loss": 0.826, "step": 13795 }, { "epoch": 1.081191222570533, "grad_norm": 4.574744701385498, "learning_rate": 8.319907765240487e-06, "loss": 1.3481, "step": 13796 }, { "epoch": 1.081269592476489, "grad_norm": 2.005664348602295, "learning_rate": 8.319677181520414e-06, "loss": 1.1941, "step": 13797 }, { "epoch": 1.0813479623824451, "grad_norm": 1.7204872369766235, "learning_rate": 8.319446585173992e-06, "loss": 1.2466, "step": 13798 }, { "epoch": 1.0814263322884012, "grad_norm": 6.2686262130737305, "learning_rate": 8.319215976202101e-06, "loss": 1.1197, "step": 13799 }, { "epoch": 1.0815047021943573, "grad_norm": 1.9453558921813965, "learning_rate": 8.318985354605619e-06, "loss": 1.5656, "step": 13800 }, { "epoch": 1.0815830721003135, "grad_norm": 2.247616767883301, "learning_rate": 8.31875472038542e-06, "loss": 1.1516, "step": 13801 }, { "epoch": 1.0816614420062696, "grad_norm": 2.9597630500793457, "learning_rate": 8.318524073542383e-06, "loss": 1.3165, "step": 13802 }, { "epoch": 1.0817398119122257, "grad_norm": 7.586538791656494, "learning_rate": 8.318293414077385e-06, "loss": 1.5447, "step": 13803 }, { "epoch": 1.0818181818181818, "grad_norm": 2.509415864944458, "learning_rate": 8.318062741991302e-06, "loss": 1.3357, "step": 13804 }, { "epoch": 1.081896551724138, "grad_norm": 2.1204283237457275, "learning_rate": 8.317832057285014e-06, "loss": 1.5993, "step": 13805 }, { "epoch": 1.081974921630094, "grad_norm": 2.789931058883667, "learning_rate": 8.317601359959395e-06, "loss": 1.4763, "step": 13806 }, { "epoch": 1.0820532915360501, "grad_norm": 3.7360265254974365, "learning_rate": 8.317370650015327e-06, "loss": 1.2209, "step": 13807 }, { "epoch": 1.0821316614420062, "grad_norm": 4.108473777770996, "learning_rate": 8.317139927453683e-06, "loss": 0.9279, "step": 13808 }, { "epoch": 1.0822100313479623, "grad_norm": 2.792928457260132, "learning_rate": 8.316909192275344e-06, "loss": 1.7901, "step": 13809 }, { "epoch": 1.0822884012539185, "grad_norm": 5.659300804138184, "learning_rate": 8.316678444481186e-06, "loss": 1.5266, "step": 13810 }, { "epoch": 1.0823667711598746, "grad_norm": 3.786763906478882, "learning_rate": 8.316447684072086e-06, "loss": 1.4568, "step": 13811 }, { "epoch": 1.0824451410658307, "grad_norm": 2.1529815196990967, "learning_rate": 8.316216911048925e-06, "loss": 1.4393, "step": 13812 }, { "epoch": 1.0825235109717868, "grad_norm": 7.749760627746582, "learning_rate": 8.315986125412574e-06, "loss": 1.7862, "step": 13813 }, { "epoch": 1.082601880877743, "grad_norm": 2.9708850383758545, "learning_rate": 8.315755327163917e-06, "loss": 1.4591, "step": 13814 }, { "epoch": 1.082680250783699, "grad_norm": 2.651921033859253, "learning_rate": 8.31552451630383e-06, "loss": 1.3465, "step": 13815 }, { "epoch": 1.0827586206896551, "grad_norm": 2.380377769470215, "learning_rate": 8.31529369283319e-06, "loss": 1.2351, "step": 13816 }, { "epoch": 1.0828369905956112, "grad_norm": 7.675155162811279, "learning_rate": 8.315062856752876e-06, "loss": 1.5994, "step": 13817 }, { "epoch": 1.0829153605015673, "grad_norm": 2.614765167236328, "learning_rate": 8.314832008063767e-06, "loss": 1.3325, "step": 13818 }, { "epoch": 1.0829937304075234, "grad_norm": 2.141784906387329, "learning_rate": 8.314601146766738e-06, "loss": 1.3689, "step": 13819 }, { "epoch": 1.0830721003134796, "grad_norm": 5.873722553253174, "learning_rate": 8.31437027286267e-06, "loss": 1.6526, "step": 13820 }, { "epoch": 1.0831504702194357, "grad_norm": 3.373234987258911, "learning_rate": 8.314139386352438e-06, "loss": 1.7514, "step": 13821 }, { "epoch": 1.0832288401253918, "grad_norm": 5.904409885406494, "learning_rate": 8.313908487236923e-06, "loss": 1.8081, "step": 13822 }, { "epoch": 1.083307210031348, "grad_norm": 6.320091724395752, "learning_rate": 8.313677575517e-06, "loss": 1.414, "step": 13823 }, { "epoch": 1.083385579937304, "grad_norm": 2.1006784439086914, "learning_rate": 8.313446651193553e-06, "loss": 1.385, "step": 13824 }, { "epoch": 1.08346394984326, "grad_norm": 10.599513053894043, "learning_rate": 8.313215714267454e-06, "loss": 1.8238, "step": 13825 }, { "epoch": 1.0835423197492162, "grad_norm": 3.7304513454437256, "learning_rate": 8.312984764739588e-06, "loss": 1.6759, "step": 13826 }, { "epoch": 1.0836206896551723, "grad_norm": 2.4858479499816895, "learning_rate": 8.312753802610827e-06, "loss": 1.5679, "step": 13827 }, { "epoch": 1.0836990595611284, "grad_norm": 2.6060893535614014, "learning_rate": 8.312522827882053e-06, "loss": 1.2529, "step": 13828 }, { "epoch": 1.0837774294670846, "grad_norm": 3.516587495803833, "learning_rate": 8.312291840554143e-06, "loss": 0.9586, "step": 13829 }, { "epoch": 1.0838557993730407, "grad_norm": 2.438506841659546, "learning_rate": 8.312060840627976e-06, "loss": 1.4953, "step": 13830 }, { "epoch": 1.0839341692789968, "grad_norm": 1.8276567459106445, "learning_rate": 8.311829828104429e-06, "loss": 1.458, "step": 13831 }, { "epoch": 1.0840125391849529, "grad_norm": 3.7359209060668945, "learning_rate": 8.311598802984386e-06, "loss": 1.6666, "step": 13832 }, { "epoch": 1.084090909090909, "grad_norm": 3.678480863571167, "learning_rate": 8.31136776526872e-06, "loss": 1.208, "step": 13833 }, { "epoch": 1.084169278996865, "grad_norm": 4.684277057647705, "learning_rate": 8.311136714958313e-06, "loss": 1.462, "step": 13834 }, { "epoch": 1.0842476489028212, "grad_norm": 5.028182029724121, "learning_rate": 8.31090565205404e-06, "loss": 1.53, "step": 13835 }, { "epoch": 1.0843260188087773, "grad_norm": 3.039677858352661, "learning_rate": 8.310674576556786e-06, "loss": 1.1496, "step": 13836 }, { "epoch": 1.0844043887147334, "grad_norm": 3.9787540435791016, "learning_rate": 8.310443488467423e-06, "loss": 1.1006, "step": 13837 }, { "epoch": 1.0844827586206898, "grad_norm": 2.924405574798584, "learning_rate": 8.310212387786835e-06, "loss": 1.2849, "step": 13838 }, { "epoch": 1.0845611285266457, "grad_norm": 3.2031030654907227, "learning_rate": 8.309981274515902e-06, "loss": 0.9463, "step": 13839 }, { "epoch": 1.084639498432602, "grad_norm": 2.7552027702331543, "learning_rate": 8.309750148655496e-06, "loss": 1.398, "step": 13840 }, { "epoch": 1.0847178683385579, "grad_norm": 5.974921226501465, "learning_rate": 8.309519010206502e-06, "loss": 1.7358, "step": 13841 }, { "epoch": 1.0847962382445142, "grad_norm": 3.827005386352539, "learning_rate": 8.309287859169799e-06, "loss": 1.1389, "step": 13842 }, { "epoch": 1.0848746081504703, "grad_norm": 6.714731216430664, "learning_rate": 8.309056695546262e-06, "loss": 1.6969, "step": 13843 }, { "epoch": 1.0849529780564264, "grad_norm": 3.485173463821411, "learning_rate": 8.308825519336775e-06, "loss": 1.3952, "step": 13844 }, { "epoch": 1.0850313479623825, "grad_norm": 5.9825029373168945, "learning_rate": 8.308594330542215e-06, "loss": 1.6021, "step": 13845 }, { "epoch": 1.0851097178683387, "grad_norm": 3.054218292236328, "learning_rate": 8.308363129163462e-06, "loss": 1.0696, "step": 13846 }, { "epoch": 1.0851880877742948, "grad_norm": 2.7515435218811035, "learning_rate": 8.308131915201392e-06, "loss": 1.3346, "step": 13847 }, { "epoch": 1.0852664576802509, "grad_norm": 1.9013571739196777, "learning_rate": 8.30790068865689e-06, "loss": 1.3105, "step": 13848 }, { "epoch": 1.085344827586207, "grad_norm": 3.2655527591705322, "learning_rate": 8.307669449530833e-06, "loss": 0.8568, "step": 13849 }, { "epoch": 1.085423197492163, "grad_norm": 7.747859477996826, "learning_rate": 8.307438197824101e-06, "loss": 1.0457, "step": 13850 }, { "epoch": 1.0855015673981192, "grad_norm": 2.5740647315979004, "learning_rate": 8.30720693353757e-06, "loss": 1.5844, "step": 13851 }, { "epoch": 1.0855799373040753, "grad_norm": 6.230587005615234, "learning_rate": 8.306975656672125e-06, "loss": 1.8695, "step": 13852 }, { "epoch": 1.0856583072100314, "grad_norm": 2.356943368911743, "learning_rate": 8.30674436722864e-06, "loss": 1.4784, "step": 13853 }, { "epoch": 1.0857366771159875, "grad_norm": 3.1391208171844482, "learning_rate": 8.306513065208001e-06, "loss": 1.1191, "step": 13854 }, { "epoch": 1.0858150470219436, "grad_norm": 3.585399866104126, "learning_rate": 8.306281750611086e-06, "loss": 1.6847, "step": 13855 }, { "epoch": 1.0858934169278998, "grad_norm": 4.02368688583374, "learning_rate": 8.30605042343877e-06, "loss": 1.4784, "step": 13856 }, { "epoch": 1.0859717868338559, "grad_norm": 2.366091012954712, "learning_rate": 8.305819083691937e-06, "loss": 1.341, "step": 13857 }, { "epoch": 1.086050156739812, "grad_norm": 2.955892562866211, "learning_rate": 8.305587731371467e-06, "loss": 1.1197, "step": 13858 }, { "epoch": 1.086128526645768, "grad_norm": 2.0807883739471436, "learning_rate": 8.305356366478239e-06, "loss": 1.3936, "step": 13859 }, { "epoch": 1.0862068965517242, "grad_norm": 3.068176507949829, "learning_rate": 8.305124989013132e-06, "loss": 1.3458, "step": 13860 }, { "epoch": 1.0862852664576803, "grad_norm": 7.306464195251465, "learning_rate": 8.304893598977028e-06, "loss": 1.0287, "step": 13861 }, { "epoch": 1.0863636363636364, "grad_norm": 2.4756922721862793, "learning_rate": 8.304662196370807e-06, "loss": 1.4088, "step": 13862 }, { "epoch": 1.0864420062695925, "grad_norm": 2.176548719406128, "learning_rate": 8.304430781195348e-06, "loss": 1.2728, "step": 13863 }, { "epoch": 1.0865203761755486, "grad_norm": 3.464186429977417, "learning_rate": 8.304199353451533e-06, "loss": 1.3794, "step": 13864 }, { "epoch": 1.0865987460815048, "grad_norm": 3.8742942810058594, "learning_rate": 8.303967913140239e-06, "loss": 1.6021, "step": 13865 }, { "epoch": 1.0866771159874609, "grad_norm": 3.78938364982605, "learning_rate": 8.303736460262347e-06, "loss": 0.751, "step": 13866 }, { "epoch": 1.086755485893417, "grad_norm": 3.6429243087768555, "learning_rate": 8.303504994818738e-06, "loss": 1.4988, "step": 13867 }, { "epoch": 1.086833855799373, "grad_norm": 4.068004131317139, "learning_rate": 8.303273516810297e-06, "loss": 1.007, "step": 13868 }, { "epoch": 1.0869122257053292, "grad_norm": 2.280017852783203, "learning_rate": 8.303042026237896e-06, "loss": 1.0683, "step": 13869 }, { "epoch": 1.0869905956112853, "grad_norm": 2.6607611179351807, "learning_rate": 8.302810523102422e-06, "loss": 1.3591, "step": 13870 }, { "epoch": 1.0870689655172414, "grad_norm": 2.8992578983306885, "learning_rate": 8.302579007404752e-06, "loss": 1.1151, "step": 13871 }, { "epoch": 1.0871473354231975, "grad_norm": 2.358743906021118, "learning_rate": 8.30234747914577e-06, "loss": 1.5886, "step": 13872 }, { "epoch": 1.0872257053291536, "grad_norm": 2.777188301086426, "learning_rate": 8.302115938326352e-06, "loss": 1.3015, "step": 13873 }, { "epoch": 1.0873040752351097, "grad_norm": 3.319122552871704, "learning_rate": 8.301884384947381e-06, "loss": 1.2885, "step": 13874 }, { "epoch": 1.0873824451410659, "grad_norm": 2.74302339553833, "learning_rate": 8.301652819009737e-06, "loss": 1.4712, "step": 13875 }, { "epoch": 1.087460815047022, "grad_norm": 2.6415436267852783, "learning_rate": 8.301421240514305e-06, "loss": 1.447, "step": 13876 }, { "epoch": 1.087539184952978, "grad_norm": 3.34256649017334, "learning_rate": 8.30118964946196e-06, "loss": 1.0164, "step": 13877 }, { "epoch": 1.0876175548589342, "grad_norm": 2.283601999282837, "learning_rate": 8.300958045853584e-06, "loss": 1.5796, "step": 13878 }, { "epoch": 1.0876959247648903, "grad_norm": 4.170226573944092, "learning_rate": 8.30072642969006e-06, "loss": 1.4082, "step": 13879 }, { "epoch": 1.0877742946708464, "grad_norm": 2.361675977706909, "learning_rate": 8.30049480097227e-06, "loss": 1.5753, "step": 13880 }, { "epoch": 1.0878526645768025, "grad_norm": 9.79837417602539, "learning_rate": 8.30026315970109e-06, "loss": 0.5295, "step": 13881 }, { "epoch": 1.0879310344827586, "grad_norm": 4.320400714874268, "learning_rate": 8.300031505877405e-06, "loss": 1.4535, "step": 13882 }, { "epoch": 1.0880094043887147, "grad_norm": 3.2047038078308105, "learning_rate": 8.299799839502098e-06, "loss": 1.0966, "step": 13883 }, { "epoch": 1.0880877742946709, "grad_norm": 2.9988739490509033, "learning_rate": 8.299568160576044e-06, "loss": 1.5745, "step": 13884 }, { "epoch": 1.088166144200627, "grad_norm": 3.30682110786438, "learning_rate": 8.29933646910013e-06, "loss": 1.3561, "step": 13885 }, { "epoch": 1.088244514106583, "grad_norm": 20.356969833374023, "learning_rate": 8.299104765075233e-06, "loss": 1.4505, "step": 13886 }, { "epoch": 1.0883228840125392, "grad_norm": 2.398475408554077, "learning_rate": 8.298873048502235e-06, "loss": 0.9377, "step": 13887 }, { "epoch": 1.0884012539184953, "grad_norm": 2.409853219985962, "learning_rate": 8.29864131938202e-06, "loss": 1.3599, "step": 13888 }, { "epoch": 1.0884796238244514, "grad_norm": 2.120710849761963, "learning_rate": 8.298409577715467e-06, "loss": 1.1802, "step": 13889 }, { "epoch": 1.0885579937304075, "grad_norm": 3.964672803878784, "learning_rate": 8.29817782350346e-06, "loss": 1.5319, "step": 13890 }, { "epoch": 1.0886363636363636, "grad_norm": 2.5040550231933594, "learning_rate": 8.297946056746876e-06, "loss": 1.0574, "step": 13891 }, { "epoch": 1.0887147335423197, "grad_norm": 5.336669445037842, "learning_rate": 8.297714277446601e-06, "loss": 1.3936, "step": 13892 }, { "epoch": 1.0887931034482758, "grad_norm": 2.466745376586914, "learning_rate": 8.297482485603515e-06, "loss": 1.2873, "step": 13893 }, { "epoch": 1.088871473354232, "grad_norm": 4.062961101531982, "learning_rate": 8.297250681218497e-06, "loss": 1.3065, "step": 13894 }, { "epoch": 1.088949843260188, "grad_norm": 3.3148083686828613, "learning_rate": 8.297018864292432e-06, "loss": 1.3363, "step": 13895 }, { "epoch": 1.0890282131661442, "grad_norm": 2.4623773097991943, "learning_rate": 8.296787034826201e-06, "loss": 1.3658, "step": 13896 }, { "epoch": 1.0891065830721003, "grad_norm": 7.164298057556152, "learning_rate": 8.296555192820684e-06, "loss": 0.7795, "step": 13897 }, { "epoch": 1.0891849529780564, "grad_norm": 3.5048909187316895, "learning_rate": 8.296323338276767e-06, "loss": 1.4993, "step": 13898 }, { "epoch": 1.0892633228840125, "grad_norm": 3.413569450378418, "learning_rate": 8.296091471195327e-06, "loss": 1.5052, "step": 13899 }, { "epoch": 1.0893416927899686, "grad_norm": 3.2761850357055664, "learning_rate": 8.295859591577249e-06, "loss": 1.2618, "step": 13900 }, { "epoch": 1.0894200626959247, "grad_norm": 1.853305459022522, "learning_rate": 8.295627699423414e-06, "loss": 1.2696, "step": 13901 }, { "epoch": 1.0894984326018808, "grad_norm": 4.513680934906006, "learning_rate": 8.295395794734704e-06, "loss": 1.4628, "step": 13902 }, { "epoch": 1.089576802507837, "grad_norm": 2.518249750137329, "learning_rate": 8.295163877512e-06, "loss": 1.2511, "step": 13903 }, { "epoch": 1.089655172413793, "grad_norm": 2.900618314743042, "learning_rate": 8.294931947756187e-06, "loss": 1.5017, "step": 13904 }, { "epoch": 1.0897335423197492, "grad_norm": 2.473813056945801, "learning_rate": 8.294700005468144e-06, "loss": 1.5649, "step": 13905 }, { "epoch": 1.0898119122257053, "grad_norm": 4.57975959777832, "learning_rate": 8.294468050648754e-06, "loss": 1.1719, "step": 13906 }, { "epoch": 1.0898902821316614, "grad_norm": 2.7888824939727783, "learning_rate": 8.294236083298898e-06, "loss": 1.523, "step": 13907 }, { "epoch": 1.0899686520376175, "grad_norm": 3.6145262718200684, "learning_rate": 8.294004103419463e-06, "loss": 1.5593, "step": 13908 }, { "epoch": 1.0900470219435736, "grad_norm": 6.543888568878174, "learning_rate": 8.293772111011328e-06, "loss": 1.7842, "step": 13909 }, { "epoch": 1.0901253918495297, "grad_norm": 16.38064193725586, "learning_rate": 8.293540106075375e-06, "loss": 1.5121, "step": 13910 }, { "epoch": 1.0902037617554858, "grad_norm": 2.013969898223877, "learning_rate": 8.293308088612488e-06, "loss": 1.3685, "step": 13911 }, { "epoch": 1.090282131661442, "grad_norm": 2.420718193054199, "learning_rate": 8.293076058623548e-06, "loss": 1.0857, "step": 13912 }, { "epoch": 1.090360501567398, "grad_norm": 3.6927738189697266, "learning_rate": 8.292844016109438e-06, "loss": 1.7519, "step": 13913 }, { "epoch": 1.0904388714733542, "grad_norm": 1.9982070922851562, "learning_rate": 8.29261196107104e-06, "loss": 1.398, "step": 13914 }, { "epoch": 1.0905172413793103, "grad_norm": 2.6812527179718018, "learning_rate": 8.292379893509238e-06, "loss": 1.6537, "step": 13915 }, { "epoch": 1.0905956112852664, "grad_norm": 3.4227828979492188, "learning_rate": 8.292147813424913e-06, "loss": 1.1526, "step": 13916 }, { "epoch": 1.0906739811912225, "grad_norm": 3.909158229827881, "learning_rate": 8.29191572081895e-06, "loss": 1.7329, "step": 13917 }, { "epoch": 1.0907523510971786, "grad_norm": 2.888089656829834, "learning_rate": 8.29168361569223e-06, "loss": 1.5565, "step": 13918 }, { "epoch": 1.0908307210031347, "grad_norm": 2.1884377002716064, "learning_rate": 8.291451498045637e-06, "loss": 1.2049, "step": 13919 }, { "epoch": 1.0909090909090908, "grad_norm": 3.808403968811035, "learning_rate": 8.291219367880053e-06, "loss": 1.445, "step": 13920 }, { "epoch": 1.090987460815047, "grad_norm": 2.064229726791382, "learning_rate": 8.29098722519636e-06, "loss": 1.3975, "step": 13921 }, { "epoch": 1.091065830721003, "grad_norm": 2.7599241733551025, "learning_rate": 8.290755069995442e-06, "loss": 1.4773, "step": 13922 }, { "epoch": 1.0911442006269592, "grad_norm": 2.9701807498931885, "learning_rate": 8.290522902278182e-06, "loss": 1.1659, "step": 13923 }, { "epoch": 1.0912225705329153, "grad_norm": 2.1446175575256348, "learning_rate": 8.290290722045464e-06, "loss": 1.509, "step": 13924 }, { "epoch": 1.0913009404388714, "grad_norm": 3.371589422225952, "learning_rate": 8.290058529298169e-06, "loss": 1.6292, "step": 13925 }, { "epoch": 1.0913793103448275, "grad_norm": 2.7806732654571533, "learning_rate": 8.289826324037181e-06, "loss": 1.2512, "step": 13926 }, { "epoch": 1.0914576802507836, "grad_norm": 3.9553208351135254, "learning_rate": 8.289594106263383e-06, "loss": 1.4991, "step": 13927 }, { "epoch": 1.0915360501567397, "grad_norm": 2.921186923980713, "learning_rate": 8.289361875977661e-06, "loss": 0.7237, "step": 13928 }, { "epoch": 1.0916144200626958, "grad_norm": 2.0521342754364014, "learning_rate": 8.289129633180895e-06, "loss": 1.4645, "step": 13929 }, { "epoch": 1.091692789968652, "grad_norm": 1.954664945602417, "learning_rate": 8.288897377873967e-06, "loss": 1.2721, "step": 13930 }, { "epoch": 1.0917711598746083, "grad_norm": 3.176253318786621, "learning_rate": 8.288665110057766e-06, "loss": 1.4848, "step": 13931 }, { "epoch": 1.0918495297805642, "grad_norm": 3.8245670795440674, "learning_rate": 8.288432829733171e-06, "loss": 1.7679, "step": 13932 }, { "epoch": 1.0919278996865205, "grad_norm": 3.564714193344116, "learning_rate": 8.288200536901066e-06, "loss": 1.3181, "step": 13933 }, { "epoch": 1.0920062695924764, "grad_norm": 3.3668713569641113, "learning_rate": 8.287968231562335e-06, "loss": 1.3698, "step": 13934 }, { "epoch": 1.0920846394984327, "grad_norm": 2.1547751426696777, "learning_rate": 8.28773591371786e-06, "loss": 1.167, "step": 13935 }, { "epoch": 1.0921630094043888, "grad_norm": 4.33212947845459, "learning_rate": 8.28750358336853e-06, "loss": 1.5081, "step": 13936 }, { "epoch": 1.092241379310345, "grad_norm": 7.684242248535156, "learning_rate": 8.287271240515222e-06, "loss": 1.7321, "step": 13937 }, { "epoch": 1.092319749216301, "grad_norm": 1.9129589796066284, "learning_rate": 8.287038885158825e-06, "loss": 1.2226, "step": 13938 }, { "epoch": 1.0923981191222571, "grad_norm": 2.5459344387054443, "learning_rate": 8.286806517300219e-06, "loss": 1.7027, "step": 13939 }, { "epoch": 1.0924764890282133, "grad_norm": 4.744163990020752, "learning_rate": 8.28657413694029e-06, "loss": 1.1805, "step": 13940 }, { "epoch": 1.0925548589341694, "grad_norm": 2.778625726699829, "learning_rate": 8.28634174407992e-06, "loss": 1.3924, "step": 13941 }, { "epoch": 1.0926332288401255, "grad_norm": 3.436886787414551, "learning_rate": 8.286109338719995e-06, "loss": 1.2185, "step": 13942 }, { "epoch": 1.0927115987460816, "grad_norm": 3.446159601211548, "learning_rate": 8.285876920861397e-06, "loss": 1.4153, "step": 13943 }, { "epoch": 1.0927899686520377, "grad_norm": 2.084934949874878, "learning_rate": 8.28564449050501e-06, "loss": 1.247, "step": 13944 }, { "epoch": 1.0928683385579938, "grad_norm": 2.1619863510131836, "learning_rate": 8.28541204765172e-06, "loss": 1.2347, "step": 13945 }, { "epoch": 1.09294670846395, "grad_norm": 2.2624382972717285, "learning_rate": 8.28517959230241e-06, "loss": 1.4731, "step": 13946 }, { "epoch": 1.093025078369906, "grad_norm": 2.004613161087036, "learning_rate": 8.284947124457966e-06, "loss": 1.3241, "step": 13947 }, { "epoch": 1.0931034482758621, "grad_norm": 3.9429500102996826, "learning_rate": 8.284714644119268e-06, "loss": 1.2464, "step": 13948 }, { "epoch": 1.0931818181818183, "grad_norm": 3.4653782844543457, "learning_rate": 8.284482151287204e-06, "loss": 1.4565, "step": 13949 }, { "epoch": 1.0932601880877744, "grad_norm": 3.9173386096954346, "learning_rate": 8.284249645962657e-06, "loss": 1.5421, "step": 13950 }, { "epoch": 1.0933385579937305, "grad_norm": 4.494433879852295, "learning_rate": 8.28401712814651e-06, "loss": 1.7846, "step": 13951 }, { "epoch": 1.0934169278996866, "grad_norm": 2.3314478397369385, "learning_rate": 8.28378459783965e-06, "loss": 0.9832, "step": 13952 }, { "epoch": 1.0934952978056427, "grad_norm": 3.882885694503784, "learning_rate": 8.283552055042959e-06, "loss": 1.3368, "step": 13953 }, { "epoch": 1.0935736677115988, "grad_norm": 6.753300666809082, "learning_rate": 8.283319499757322e-06, "loss": 1.3962, "step": 13954 }, { "epoch": 1.093652037617555, "grad_norm": 5.805604934692383, "learning_rate": 8.283086931983627e-06, "loss": 1.3125, "step": 13955 }, { "epoch": 1.093730407523511, "grad_norm": 2.9959614276885986, "learning_rate": 8.282854351722753e-06, "loss": 1.2885, "step": 13956 }, { "epoch": 1.0938087774294671, "grad_norm": 2.2704648971557617, "learning_rate": 8.282621758975587e-06, "loss": 1.5551, "step": 13957 }, { "epoch": 1.0938871473354232, "grad_norm": 4.429991245269775, "learning_rate": 8.282389153743016e-06, "loss": 1.0746, "step": 13958 }, { "epoch": 1.0939655172413794, "grad_norm": 4.065300941467285, "learning_rate": 8.28215653602592e-06, "loss": 1.9359, "step": 13959 }, { "epoch": 1.0940438871473355, "grad_norm": 3.2350080013275146, "learning_rate": 8.281923905825188e-06, "loss": 1.4926, "step": 13960 }, { "epoch": 1.0941222570532916, "grad_norm": 5.138515949249268, "learning_rate": 8.281691263141703e-06, "loss": 1.2541, "step": 13961 }, { "epoch": 1.0942006269592477, "grad_norm": 3.1658644676208496, "learning_rate": 8.28145860797635e-06, "loss": 1.5421, "step": 13962 }, { "epoch": 1.0942789968652038, "grad_norm": 3.215993881225586, "learning_rate": 8.281225940330013e-06, "loss": 1.4613, "step": 13963 }, { "epoch": 1.09435736677116, "grad_norm": 6.911121368408203, "learning_rate": 8.280993260203578e-06, "loss": 1.6498, "step": 13964 }, { "epoch": 1.094435736677116, "grad_norm": 2.469409465789795, "learning_rate": 8.280760567597933e-06, "loss": 1.6228, "step": 13965 }, { "epoch": 1.0945141065830721, "grad_norm": 3.6832263469696045, "learning_rate": 8.280527862513957e-06, "loss": 1.1717, "step": 13966 }, { "epoch": 1.0945924764890282, "grad_norm": 1.8502968549728394, "learning_rate": 8.280295144952537e-06, "loss": 1.5594, "step": 13967 }, { "epoch": 1.0946708463949844, "grad_norm": 3.218003511428833, "learning_rate": 8.280062414914561e-06, "loss": 1.2668, "step": 13968 }, { "epoch": 1.0947492163009405, "grad_norm": 4.466447353363037, "learning_rate": 8.279829672400913e-06, "loss": 2.6958, "step": 13969 }, { "epoch": 1.0948275862068966, "grad_norm": 1.9177279472351074, "learning_rate": 8.279596917412475e-06, "loss": 1.5561, "step": 13970 }, { "epoch": 1.0949059561128527, "grad_norm": 2.4558494091033936, "learning_rate": 8.279364149950137e-06, "loss": 1.411, "step": 13971 }, { "epoch": 1.0949843260188088, "grad_norm": 2.2581682205200195, "learning_rate": 8.279131370014781e-06, "loss": 0.832, "step": 13972 }, { "epoch": 1.095062695924765, "grad_norm": 3.0358283519744873, "learning_rate": 8.278898577607294e-06, "loss": 1.2723, "step": 13973 }, { "epoch": 1.095141065830721, "grad_norm": 8.951953887939453, "learning_rate": 8.278665772728561e-06, "loss": 1.506, "step": 13974 }, { "epoch": 1.0952194357366771, "grad_norm": 4.509147644042969, "learning_rate": 8.278432955379467e-06, "loss": 2.1659, "step": 13975 }, { "epoch": 1.0952978056426332, "grad_norm": 2.7103073596954346, "learning_rate": 8.278200125560898e-06, "loss": 1.3641, "step": 13976 }, { "epoch": 1.0953761755485893, "grad_norm": 3.2687649726867676, "learning_rate": 8.277967283273741e-06, "loss": 0.909, "step": 13977 }, { "epoch": 1.0954545454545455, "grad_norm": 1.9514223337173462, "learning_rate": 8.277734428518877e-06, "loss": 1.0753, "step": 13978 }, { "epoch": 1.0955329153605016, "grad_norm": 3.805856466293335, "learning_rate": 8.277501561297198e-06, "loss": 1.8251, "step": 13979 }, { "epoch": 1.0956112852664577, "grad_norm": 6.8302154541015625, "learning_rate": 8.277268681609585e-06, "loss": 1.6424, "step": 13980 }, { "epoch": 1.0956896551724138, "grad_norm": 5.0558342933654785, "learning_rate": 8.277035789456924e-06, "loss": 1.5664, "step": 13981 }, { "epoch": 1.09576802507837, "grad_norm": 6.580596446990967, "learning_rate": 8.276802884840104e-06, "loss": 1.669, "step": 13982 }, { "epoch": 1.095846394984326, "grad_norm": 1.9972628355026245, "learning_rate": 8.27656996776001e-06, "loss": 1.4295, "step": 13983 }, { "epoch": 1.0959247648902821, "grad_norm": 2.482717990875244, "learning_rate": 8.276337038217524e-06, "loss": 1.3916, "step": 13984 }, { "epoch": 1.0960031347962382, "grad_norm": 3.0913264751434326, "learning_rate": 8.276104096213535e-06, "loss": 0.3638, "step": 13985 }, { "epoch": 1.0960815047021943, "grad_norm": 3.304381847381592, "learning_rate": 8.27587114174893e-06, "loss": 0.9831, "step": 13986 }, { "epoch": 1.0961598746081505, "grad_norm": 2.9476723670959473, "learning_rate": 8.275638174824593e-06, "loss": 1.1306, "step": 13987 }, { "epoch": 1.0962382445141066, "grad_norm": 2.8529038429260254, "learning_rate": 8.275405195441413e-06, "loss": 0.9517, "step": 13988 }, { "epoch": 1.0963166144200627, "grad_norm": 2.261133909225464, "learning_rate": 8.27517220360027e-06, "loss": 1.3568, "step": 13989 }, { "epoch": 1.0963949843260188, "grad_norm": 7.215985298156738, "learning_rate": 8.274939199302058e-06, "loss": 1.3234, "step": 13990 }, { "epoch": 1.096473354231975, "grad_norm": 3.2323670387268066, "learning_rate": 8.274706182547657e-06, "loss": 1.2666, "step": 13991 }, { "epoch": 1.096551724137931, "grad_norm": 2.94953989982605, "learning_rate": 8.274473153337956e-06, "loss": 1.7821, "step": 13992 }, { "epoch": 1.0966300940438871, "grad_norm": 3.604644775390625, "learning_rate": 8.274240111673842e-06, "loss": 1.6682, "step": 13993 }, { "epoch": 1.0967084639498432, "grad_norm": 3.599047899246216, "learning_rate": 8.2740070575562e-06, "loss": 1.0445, "step": 13994 }, { "epoch": 1.0967868338557993, "grad_norm": 2.7921719551086426, "learning_rate": 8.273773990985917e-06, "loss": 1.3605, "step": 13995 }, { "epoch": 1.0968652037617554, "grad_norm": 2.9708149433135986, "learning_rate": 8.273540911963878e-06, "loss": 1.3055, "step": 13996 }, { "epoch": 1.0969435736677116, "grad_norm": 3.047966957092285, "learning_rate": 8.273307820490972e-06, "loss": 1.5526, "step": 13997 }, { "epoch": 1.0970219435736677, "grad_norm": 2.746605157852173, "learning_rate": 8.273074716568084e-06, "loss": 1.5006, "step": 13998 }, { "epoch": 1.0971003134796238, "grad_norm": 5.487582206726074, "learning_rate": 8.272841600196102e-06, "loss": 1.1166, "step": 13999 }, { "epoch": 1.09717868338558, "grad_norm": 6.506438255310059, "learning_rate": 8.27260847137591e-06, "loss": 1.3217, "step": 14000 }, { "epoch": 1.097257053291536, "grad_norm": 5.077670097351074, "learning_rate": 8.272375330108398e-06, "loss": 1.3501, "step": 14001 }, { "epoch": 1.097335423197492, "grad_norm": 3.2514283657073975, "learning_rate": 8.272142176394448e-06, "loss": 1.2319, "step": 14002 }, { "epoch": 1.0974137931034482, "grad_norm": 3.353706121444702, "learning_rate": 8.271909010234952e-06, "loss": 1.1533, "step": 14003 }, { "epoch": 1.0974921630094043, "grad_norm": 4.43610954284668, "learning_rate": 8.271675831630794e-06, "loss": 1.8491, "step": 14004 }, { "epoch": 1.0975705329153604, "grad_norm": 3.965965747833252, "learning_rate": 8.271442640582863e-06, "loss": 1.3027, "step": 14005 }, { "epoch": 1.0976489028213166, "grad_norm": 4.435725212097168, "learning_rate": 8.271209437092043e-06, "loss": 1.5055, "step": 14006 }, { "epoch": 1.0977272727272727, "grad_norm": 1.915554404258728, "learning_rate": 8.270976221159224e-06, "loss": 1.2425, "step": 14007 }, { "epoch": 1.0978056426332288, "grad_norm": 2.315411329269409, "learning_rate": 8.27074299278529e-06, "loss": 1.3088, "step": 14008 }, { "epoch": 1.0978840125391849, "grad_norm": 2.063412666320801, "learning_rate": 8.27050975197113e-06, "loss": 1.1463, "step": 14009 }, { "epoch": 1.097962382445141, "grad_norm": 6.588497638702393, "learning_rate": 8.270276498717631e-06, "loss": 1.6972, "step": 14010 }, { "epoch": 1.098040752351097, "grad_norm": 2.307133197784424, "learning_rate": 8.27004323302568e-06, "loss": 1.3251, "step": 14011 }, { "epoch": 1.0981191222570532, "grad_norm": 2.5942485332489014, "learning_rate": 8.269809954896164e-06, "loss": 1.6539, "step": 14012 }, { "epoch": 1.0981974921630093, "grad_norm": 3.561509132385254, "learning_rate": 8.26957666432997e-06, "loss": 1.5832, "step": 14013 }, { "epoch": 1.0982758620689654, "grad_norm": 3.7840306758880615, "learning_rate": 8.269343361327985e-06, "loss": 0.8403, "step": 14014 }, { "epoch": 1.0983542319749215, "grad_norm": 2.5558533668518066, "learning_rate": 8.269110045891099e-06, "loss": 1.3745, "step": 14015 }, { "epoch": 1.0984326018808777, "grad_norm": 4.511143684387207, "learning_rate": 8.268876718020196e-06, "loss": 1.705, "step": 14016 }, { "epoch": 1.0985109717868338, "grad_norm": 2.1687679290771484, "learning_rate": 8.268643377716166e-06, "loss": 1.1622, "step": 14017 }, { "epoch": 1.0985893416927899, "grad_norm": 2.2948155403137207, "learning_rate": 8.268410024979895e-06, "loss": 1.4387, "step": 14018 }, { "epoch": 1.098667711598746, "grad_norm": 2.5536060333251953, "learning_rate": 8.268176659812272e-06, "loss": 1.3561, "step": 14019 }, { "epoch": 1.098746081504702, "grad_norm": 30.9576416015625, "learning_rate": 8.267943282214182e-06, "loss": 1.9262, "step": 14020 }, { "epoch": 1.0988244514106582, "grad_norm": 2.4219794273376465, "learning_rate": 8.267709892186517e-06, "loss": 1.4941, "step": 14021 }, { "epoch": 1.0989028213166143, "grad_norm": 4.394311904907227, "learning_rate": 8.267476489730158e-06, "loss": 1.916, "step": 14022 }, { "epoch": 1.0989811912225704, "grad_norm": 2.50395131111145, "learning_rate": 8.267243074846e-06, "loss": 1.4776, "step": 14023 }, { "epoch": 1.0990595611285268, "grad_norm": 3.5609006881713867, "learning_rate": 8.267009647534925e-06, "loss": 1.6944, "step": 14024 }, { "epoch": 1.0991379310344827, "grad_norm": 2.1216037273406982, "learning_rate": 8.266776207797825e-06, "loss": 1.1648, "step": 14025 }, { "epoch": 1.099216300940439, "grad_norm": 4.657314300537109, "learning_rate": 8.266542755635586e-06, "loss": 1.5165, "step": 14026 }, { "epoch": 1.0992946708463949, "grad_norm": 3.5698931217193604, "learning_rate": 8.266309291049096e-06, "loss": 1.5704, "step": 14027 }, { "epoch": 1.0993730407523512, "grad_norm": 2.669574022293091, "learning_rate": 8.266075814039242e-06, "loss": 1.7219, "step": 14028 }, { "epoch": 1.099451410658307, "grad_norm": 2.106327533721924, "learning_rate": 8.265842324606915e-06, "loss": 1.1284, "step": 14029 }, { "epoch": 1.0995297805642634, "grad_norm": 2.6058664321899414, "learning_rate": 8.265608822753001e-06, "loss": 1.1008, "step": 14030 }, { "epoch": 1.0996081504702195, "grad_norm": 3.4486613273620605, "learning_rate": 8.265375308478387e-06, "loss": 1.0748, "step": 14031 }, { "epoch": 1.0996865203761756, "grad_norm": 6.046733856201172, "learning_rate": 8.265141781783963e-06, "loss": 1.8792, "step": 14032 }, { "epoch": 1.0997648902821318, "grad_norm": 7.924396514892578, "learning_rate": 8.264908242670617e-06, "loss": 1.1342, "step": 14033 }, { "epoch": 1.0998432601880879, "grad_norm": 2.462237596511841, "learning_rate": 8.264674691139238e-06, "loss": 1.4916, "step": 14034 }, { "epoch": 1.099921630094044, "grad_norm": 1.7677526473999023, "learning_rate": 8.264441127190714e-06, "loss": 1.0473, "step": 14035 }, { "epoch": 1.1, "grad_norm": 2.1170032024383545, "learning_rate": 8.264207550825931e-06, "loss": 1.3824, "step": 14036 }, { "epoch": 1.1000783699059562, "grad_norm": 2.562742233276367, "learning_rate": 8.263973962045778e-06, "loss": 1.2795, "step": 14037 }, { "epoch": 1.1001567398119123, "grad_norm": 3.6825575828552246, "learning_rate": 8.263740360851146e-06, "loss": 1.6592, "step": 14038 }, { "epoch": 1.1002351097178684, "grad_norm": 3.9502105712890625, "learning_rate": 8.263506747242921e-06, "loss": 1.4932, "step": 14039 }, { "epoch": 1.1003134796238245, "grad_norm": 3.101958751678467, "learning_rate": 8.263273121221994e-06, "loss": 1.7576, "step": 14040 }, { "epoch": 1.1003918495297806, "grad_norm": 6.753018856048584, "learning_rate": 8.263039482789253e-06, "loss": 1.4957, "step": 14041 }, { "epoch": 1.1004702194357368, "grad_norm": 2.25748610496521, "learning_rate": 8.262805831945585e-06, "loss": 1.4176, "step": 14042 }, { "epoch": 1.1005485893416929, "grad_norm": 4.611484527587891, "learning_rate": 8.262572168691876e-06, "loss": 1.2497, "step": 14043 }, { "epoch": 1.100626959247649, "grad_norm": 3.1281604766845703, "learning_rate": 8.262338493029023e-06, "loss": 1.3457, "step": 14044 }, { "epoch": 1.100705329153605, "grad_norm": 8.718018531799316, "learning_rate": 8.262104804957906e-06, "loss": 0.7977, "step": 14045 }, { "epoch": 1.1007836990595612, "grad_norm": 2.297868013381958, "learning_rate": 8.26187110447942e-06, "loss": 1.3196, "step": 14046 }, { "epoch": 1.1008620689655173, "grad_norm": 2.2801475524902344, "learning_rate": 8.261637391594453e-06, "loss": 1.6304, "step": 14047 }, { "epoch": 1.1009404388714734, "grad_norm": 2.021066665649414, "learning_rate": 8.26140366630389e-06, "loss": 1.186, "step": 14048 }, { "epoch": 1.1010188087774295, "grad_norm": 7.222036361694336, "learning_rate": 8.261169928608625e-06, "loss": 0.75, "step": 14049 }, { "epoch": 1.1010971786833856, "grad_norm": 12.65796947479248, "learning_rate": 8.260936178509543e-06, "loss": 1.5011, "step": 14050 }, { "epoch": 1.1011755485893417, "grad_norm": 2.0070526599884033, "learning_rate": 8.260702416007536e-06, "loss": 1.0634, "step": 14051 }, { "epoch": 1.1012539184952979, "grad_norm": 3.423922061920166, "learning_rate": 8.26046864110349e-06, "loss": 1.4205, "step": 14052 }, { "epoch": 1.101332288401254, "grad_norm": 2.5386064052581787, "learning_rate": 8.260234853798298e-06, "loss": 1.415, "step": 14053 }, { "epoch": 1.10141065830721, "grad_norm": 2.7009661197662354, "learning_rate": 8.260001054092845e-06, "loss": 1.6566, "step": 14054 }, { "epoch": 1.1014890282131662, "grad_norm": 2.9339754581451416, "learning_rate": 8.259767241988024e-06, "loss": 1.3001, "step": 14055 }, { "epoch": 1.1015673981191223, "grad_norm": 2.2610995769500732, "learning_rate": 8.259533417484723e-06, "loss": 1.4339, "step": 14056 }, { "epoch": 1.1016457680250784, "grad_norm": 2.320427179336548, "learning_rate": 8.25929958058383e-06, "loss": 1.3815, "step": 14057 }, { "epoch": 1.1017241379310345, "grad_norm": 2.55991530418396, "learning_rate": 8.259065731286235e-06, "loss": 1.1558, "step": 14058 }, { "epoch": 1.1018025078369906, "grad_norm": 7.565680503845215, "learning_rate": 8.25883186959283e-06, "loss": 1.959, "step": 14059 }, { "epoch": 1.1018808777429467, "grad_norm": 3.1899688243865967, "learning_rate": 8.258597995504502e-06, "loss": 1.3818, "step": 14060 }, { "epoch": 1.1019592476489029, "grad_norm": 2.062808036804199, "learning_rate": 8.25836410902214e-06, "loss": 1.4388, "step": 14061 }, { "epoch": 1.102037617554859, "grad_norm": 3.102534055709839, "learning_rate": 8.258130210146635e-06, "loss": 1.4354, "step": 14062 }, { "epoch": 1.102115987460815, "grad_norm": 3.4054512977600098, "learning_rate": 8.257896298878876e-06, "loss": 1.3873, "step": 14063 }, { "epoch": 1.1021943573667712, "grad_norm": 2.120077133178711, "learning_rate": 8.257662375219753e-06, "loss": 1.2145, "step": 14064 }, { "epoch": 1.1022727272727273, "grad_norm": 2.6184213161468506, "learning_rate": 8.257428439170154e-06, "loss": 1.6079, "step": 14065 }, { "epoch": 1.1023510971786834, "grad_norm": 2.6097676753997803, "learning_rate": 8.257194490730974e-06, "loss": 1.3743, "step": 14066 }, { "epoch": 1.1024294670846395, "grad_norm": 4.824437618255615, "learning_rate": 8.256960529903095e-06, "loss": 1.4599, "step": 14067 }, { "epoch": 1.1025078369905956, "grad_norm": 3.147639036178589, "learning_rate": 8.256726556687413e-06, "loss": 1.246, "step": 14068 }, { "epoch": 1.1025862068965517, "grad_norm": 4.767794132232666, "learning_rate": 8.256492571084815e-06, "loss": 1.4099, "step": 14069 }, { "epoch": 1.1026645768025078, "grad_norm": 2.0636775493621826, "learning_rate": 8.256258573096195e-06, "loss": 1.2727, "step": 14070 }, { "epoch": 1.102742946708464, "grad_norm": 2.7906084060668945, "learning_rate": 8.256024562722436e-06, "loss": 1.2424, "step": 14071 }, { "epoch": 1.10282131661442, "grad_norm": 14.001814842224121, "learning_rate": 8.255790539964435e-06, "loss": 0.5081, "step": 14072 }, { "epoch": 1.1028996865203762, "grad_norm": 3.6784188747406006, "learning_rate": 8.255556504823077e-06, "loss": 1.0876, "step": 14073 }, { "epoch": 1.1029780564263323, "grad_norm": 2.3437743186950684, "learning_rate": 8.255322457299253e-06, "loss": 1.2995, "step": 14074 }, { "epoch": 1.1030564263322884, "grad_norm": 3.257174253463745, "learning_rate": 8.255088397393858e-06, "loss": 1.3743, "step": 14075 }, { "epoch": 1.1031347962382445, "grad_norm": 2.4452950954437256, "learning_rate": 8.254854325107776e-06, "loss": 1.0936, "step": 14076 }, { "epoch": 1.1032131661442006, "grad_norm": 2.611030101776123, "learning_rate": 8.254620240441902e-06, "loss": 1.1727, "step": 14077 }, { "epoch": 1.1032915360501567, "grad_norm": 4.024354934692383, "learning_rate": 8.254386143397122e-06, "loss": 0.9075, "step": 14078 }, { "epoch": 1.1033699059561128, "grad_norm": 3.2321019172668457, "learning_rate": 8.254152033974329e-06, "loss": 0.9119, "step": 14079 }, { "epoch": 1.103448275862069, "grad_norm": 3.2037010192871094, "learning_rate": 8.253917912174415e-06, "loss": 1.334, "step": 14080 }, { "epoch": 1.103526645768025, "grad_norm": 3.652240514755249, "learning_rate": 8.253683777998267e-06, "loss": 1.3632, "step": 14081 }, { "epoch": 1.1036050156739812, "grad_norm": 2.8920347690582275, "learning_rate": 8.253449631446777e-06, "loss": 1.6963, "step": 14082 }, { "epoch": 1.1036833855799373, "grad_norm": 4.613712787628174, "learning_rate": 8.253215472520837e-06, "loss": 1.5584, "step": 14083 }, { "epoch": 1.1037617554858934, "grad_norm": 3.084916830062866, "learning_rate": 8.252981301221335e-06, "loss": 1.4169, "step": 14084 }, { "epoch": 1.1038401253918495, "grad_norm": 6.3317646980285645, "learning_rate": 8.252747117549164e-06, "loss": 0.8958, "step": 14085 }, { "epoch": 1.1039184952978056, "grad_norm": 2.1537859439849854, "learning_rate": 8.252512921505213e-06, "loss": 1.4422, "step": 14086 }, { "epoch": 1.1039968652037617, "grad_norm": 3.2326831817626953, "learning_rate": 8.252278713090374e-06, "loss": 1.459, "step": 14087 }, { "epoch": 1.1040752351097178, "grad_norm": 2.008673667907715, "learning_rate": 8.252044492305537e-06, "loss": 1.3485, "step": 14088 }, { "epoch": 1.104153605015674, "grad_norm": 3.6301515102386475, "learning_rate": 8.251810259151594e-06, "loss": 1.4205, "step": 14089 }, { "epoch": 1.10423197492163, "grad_norm": 2.1595356464385986, "learning_rate": 8.251576013629433e-06, "loss": 1.5947, "step": 14090 }, { "epoch": 1.1043103448275862, "grad_norm": 4.170566082000732, "learning_rate": 8.251341755739949e-06, "loss": 1.2941, "step": 14091 }, { "epoch": 1.1043887147335423, "grad_norm": 2.1684958934783936, "learning_rate": 8.251107485484029e-06, "loss": 1.4414, "step": 14092 }, { "epoch": 1.1044670846394984, "grad_norm": 17.252756118774414, "learning_rate": 8.250873202862567e-06, "loss": 1.7741, "step": 14093 }, { "epoch": 1.1045454545454545, "grad_norm": 7.175241470336914, "learning_rate": 8.250638907876453e-06, "loss": 1.0057, "step": 14094 }, { "epoch": 1.1046238244514106, "grad_norm": 3.066014528274536, "learning_rate": 8.250404600526579e-06, "loss": 0.9348, "step": 14095 }, { "epoch": 1.1047021943573667, "grad_norm": 2.9936413764953613, "learning_rate": 8.250170280813833e-06, "loss": 1.2405, "step": 14096 }, { "epoch": 1.1047805642633228, "grad_norm": 4.499913215637207, "learning_rate": 8.249935948739112e-06, "loss": 1.3047, "step": 14097 }, { "epoch": 1.104858934169279, "grad_norm": 2.858703136444092, "learning_rate": 8.249701604303302e-06, "loss": 1.4591, "step": 14098 }, { "epoch": 1.104937304075235, "grad_norm": 2.926743745803833, "learning_rate": 8.249467247507295e-06, "loss": 1.5719, "step": 14099 }, { "epoch": 1.1050156739811912, "grad_norm": 2.265169620513916, "learning_rate": 8.249232878351985e-06, "loss": 1.2706, "step": 14100 }, { "epoch": 1.1050940438871473, "grad_norm": 2.710895538330078, "learning_rate": 8.248998496838263e-06, "loss": 1.2781, "step": 14101 }, { "epoch": 1.1051724137931034, "grad_norm": 4.63694953918457, "learning_rate": 8.24876410296702e-06, "loss": 1.475, "step": 14102 }, { "epoch": 1.1052507836990595, "grad_norm": 2.0425729751586914, "learning_rate": 8.248529696739145e-06, "loss": 1.2927, "step": 14103 }, { "epoch": 1.1053291536050156, "grad_norm": 2.4764790534973145, "learning_rate": 8.248295278155532e-06, "loss": 1.3595, "step": 14104 }, { "epoch": 1.1054075235109717, "grad_norm": 2.5368149280548096, "learning_rate": 8.248060847217072e-06, "loss": 1.6144, "step": 14105 }, { "epoch": 1.1054858934169278, "grad_norm": 1.7505460977554321, "learning_rate": 8.247826403924656e-06, "loss": 0.8555, "step": 14106 }, { "epoch": 1.105564263322884, "grad_norm": 2.6091175079345703, "learning_rate": 8.247591948279179e-06, "loss": 1.7078, "step": 14107 }, { "epoch": 1.10564263322884, "grad_norm": 2.7218732833862305, "learning_rate": 8.24735748028153e-06, "loss": 1.5004, "step": 14108 }, { "epoch": 1.1057210031347962, "grad_norm": 3.395294427871704, "learning_rate": 8.2471229999326e-06, "loss": 1.6709, "step": 14109 }, { "epoch": 1.1057993730407523, "grad_norm": 7.4938225746154785, "learning_rate": 8.246888507233281e-06, "loss": 1.4809, "step": 14110 }, { "epoch": 1.1058777429467084, "grad_norm": 3.2692105770111084, "learning_rate": 8.246654002184468e-06, "loss": 1.5844, "step": 14111 }, { "epoch": 1.1059561128526645, "grad_norm": 2.2902255058288574, "learning_rate": 8.246419484787049e-06, "loss": 1.1399, "step": 14112 }, { "epoch": 1.1060344827586206, "grad_norm": 2.26511549949646, "learning_rate": 8.246184955041918e-06, "loss": 1.5172, "step": 14113 }, { "epoch": 1.1061128526645767, "grad_norm": 2.1459615230560303, "learning_rate": 8.245950412949968e-06, "loss": 1.2185, "step": 14114 }, { "epoch": 1.1061912225705328, "grad_norm": 4.00058126449585, "learning_rate": 8.24571585851209e-06, "loss": 1.5763, "step": 14115 }, { "epoch": 1.106269592476489, "grad_norm": 2.1321258544921875, "learning_rate": 8.245481291729174e-06, "loss": 1.3942, "step": 14116 }, { "epoch": 1.106347962382445, "grad_norm": 3.8609752655029297, "learning_rate": 8.245246712602116e-06, "loss": 1.5932, "step": 14117 }, { "epoch": 1.1064263322884011, "grad_norm": 5.204695224761963, "learning_rate": 8.245012121131806e-06, "loss": 1.593, "step": 14118 }, { "epoch": 1.1065047021943575, "grad_norm": 8.423480987548828, "learning_rate": 8.244777517319136e-06, "loss": 1.5553, "step": 14119 }, { "epoch": 1.1065830721003134, "grad_norm": 3.4136812686920166, "learning_rate": 8.244542901165e-06, "loss": 1.4338, "step": 14120 }, { "epoch": 1.1066614420062697, "grad_norm": 2.8758952617645264, "learning_rate": 8.244308272670289e-06, "loss": 0.9888, "step": 14121 }, { "epoch": 1.1067398119122256, "grad_norm": 2.7845003604888916, "learning_rate": 8.244073631835895e-06, "loss": 1.3918, "step": 14122 }, { "epoch": 1.106818181818182, "grad_norm": 3.9210939407348633, "learning_rate": 8.24383897866271e-06, "loss": 1.4944, "step": 14123 }, { "epoch": 1.106896551724138, "grad_norm": 3.941228151321411, "learning_rate": 8.243604313151629e-06, "loss": 1.7786, "step": 14124 }, { "epoch": 1.1069749216300941, "grad_norm": 4.522463798522949, "learning_rate": 8.243369635303543e-06, "loss": 1.2591, "step": 14125 }, { "epoch": 1.1070532915360503, "grad_norm": 2.762702703475952, "learning_rate": 8.243134945119346e-06, "loss": 1.6127, "step": 14126 }, { "epoch": 1.1071316614420064, "grad_norm": 2.7128281593322754, "learning_rate": 8.24290024259993e-06, "loss": 1.3916, "step": 14127 }, { "epoch": 1.1072100313479625, "grad_norm": 5.6979522705078125, "learning_rate": 8.242665527746186e-06, "loss": 1.3456, "step": 14128 }, { "epoch": 1.1072884012539186, "grad_norm": 2.4285852909088135, "learning_rate": 8.242430800559005e-06, "loss": 1.4063, "step": 14129 }, { "epoch": 1.1073667711598747, "grad_norm": 1.974514365196228, "learning_rate": 8.242196061039288e-06, "loss": 1.679, "step": 14130 }, { "epoch": 1.1074451410658308, "grad_norm": 2.9243226051330566, "learning_rate": 8.241961309187919e-06, "loss": 1.178, "step": 14131 }, { "epoch": 1.107523510971787, "grad_norm": 2.531373977661133, "learning_rate": 8.241726545005793e-06, "loss": 0.9215, "step": 14132 }, { "epoch": 1.107601880877743, "grad_norm": 2.3564558029174805, "learning_rate": 8.241491768493807e-06, "loss": 1.406, "step": 14133 }, { "epoch": 1.1076802507836991, "grad_norm": 5.297183513641357, "learning_rate": 8.241256979652852e-06, "loss": 0.9156, "step": 14134 }, { "epoch": 1.1077586206896552, "grad_norm": 3.9888007640838623, "learning_rate": 8.241022178483818e-06, "loss": 1.4728, "step": 14135 }, { "epoch": 1.1078369905956114, "grad_norm": 2.8026137351989746, "learning_rate": 8.240787364987602e-06, "loss": 1.9286, "step": 14136 }, { "epoch": 1.1079153605015675, "grad_norm": 1.9831230640411377, "learning_rate": 8.240552539165095e-06, "loss": 1.2416, "step": 14137 }, { "epoch": 1.1079937304075236, "grad_norm": 2.6557106971740723, "learning_rate": 8.240317701017189e-06, "loss": 1.5058, "step": 14138 }, { "epoch": 1.1080721003134797, "grad_norm": 1.7933598756790161, "learning_rate": 8.24008285054478e-06, "loss": 1.2612, "step": 14139 }, { "epoch": 1.1081504702194358, "grad_norm": 2.1599929332733154, "learning_rate": 8.23984798774876e-06, "loss": 1.2446, "step": 14140 }, { "epoch": 1.108228840125392, "grad_norm": 2.979238271713257, "learning_rate": 8.239613112630022e-06, "loss": 0.9539, "step": 14141 }, { "epoch": 1.108307210031348, "grad_norm": 2.613091468811035, "learning_rate": 8.239378225189461e-06, "loss": 1.1902, "step": 14142 }, { "epoch": 1.1083855799373041, "grad_norm": 3.3105132579803467, "learning_rate": 8.239143325427968e-06, "loss": 1.0151, "step": 14143 }, { "epoch": 1.1084639498432602, "grad_norm": 2.9769418239593506, "learning_rate": 8.238908413346436e-06, "loss": 1.4226, "step": 14144 }, { "epoch": 1.1085423197492164, "grad_norm": 3.869412899017334, "learning_rate": 8.238673488945763e-06, "loss": 0.8868, "step": 14145 }, { "epoch": 1.1086206896551725, "grad_norm": 1.7583937644958496, "learning_rate": 8.238438552226838e-06, "loss": 1.0223, "step": 14146 }, { "epoch": 1.1086990595611286, "grad_norm": 8.419597625732422, "learning_rate": 8.238203603190556e-06, "loss": 1.8509, "step": 14147 }, { "epoch": 1.1087774294670847, "grad_norm": 4.143278121948242, "learning_rate": 8.237968641837812e-06, "loss": 1.1485, "step": 14148 }, { "epoch": 1.1088557993730408, "grad_norm": 3.212097644805908, "learning_rate": 8.237733668169497e-06, "loss": 1.5891, "step": 14149 }, { "epoch": 1.108934169278997, "grad_norm": 3.4886221885681152, "learning_rate": 8.237498682186507e-06, "loss": 1.4918, "step": 14150 }, { "epoch": 1.109012539184953, "grad_norm": 4.133720397949219, "learning_rate": 8.237263683889733e-06, "loss": 1.4103, "step": 14151 }, { "epoch": 1.1090909090909091, "grad_norm": 3.904512882232666, "learning_rate": 8.237028673280073e-06, "loss": 1.226, "step": 14152 }, { "epoch": 1.1091692789968652, "grad_norm": 2.8364834785461426, "learning_rate": 8.236793650358417e-06, "loss": 1.1869, "step": 14153 }, { "epoch": 1.1092476489028213, "grad_norm": 3.9687952995300293, "learning_rate": 8.236558615125662e-06, "loss": 1.5233, "step": 14154 }, { "epoch": 1.1093260188087775, "grad_norm": 2.329087018966675, "learning_rate": 8.236323567582701e-06, "loss": 1.3527, "step": 14155 }, { "epoch": 1.1094043887147336, "grad_norm": 2.492323875427246, "learning_rate": 8.236088507730425e-06, "loss": 1.1933, "step": 14156 }, { "epoch": 1.1094827586206897, "grad_norm": 2.826920986175537, "learning_rate": 8.235853435569732e-06, "loss": 1.0767, "step": 14157 }, { "epoch": 1.1095611285266458, "grad_norm": 4.933052062988281, "learning_rate": 8.235618351101513e-06, "loss": 1.6355, "step": 14158 }, { "epoch": 1.109639498432602, "grad_norm": 2.3744993209838867, "learning_rate": 8.235383254326665e-06, "loss": 1.439, "step": 14159 }, { "epoch": 1.109717868338558, "grad_norm": 2.412797451019287, "learning_rate": 8.23514814524608e-06, "loss": 1.3241, "step": 14160 }, { "epoch": 1.1097962382445141, "grad_norm": 2.197761297225952, "learning_rate": 8.234913023860656e-06, "loss": 1.2472, "step": 14161 }, { "epoch": 1.1098746081504702, "grad_norm": 2.098740577697754, "learning_rate": 8.234677890171281e-06, "loss": 1.1912, "step": 14162 }, { "epoch": 1.1099529780564263, "grad_norm": 2.736706256866455, "learning_rate": 8.234442744178855e-06, "loss": 1.7034, "step": 14163 }, { "epoch": 1.1100313479623825, "grad_norm": 2.180826425552368, "learning_rate": 8.234207585884269e-06, "loss": 0.7186, "step": 14164 }, { "epoch": 1.1101097178683386, "grad_norm": 3.3510992527008057, "learning_rate": 8.233972415288419e-06, "loss": 1.2394, "step": 14165 }, { "epoch": 1.1101880877742947, "grad_norm": 2.3499412536621094, "learning_rate": 8.233737232392198e-06, "loss": 1.2438, "step": 14166 }, { "epoch": 1.1102664576802508, "grad_norm": 2.9500625133514404, "learning_rate": 8.233502037196502e-06, "loss": 1.6086, "step": 14167 }, { "epoch": 1.110344827586207, "grad_norm": 3.2842235565185547, "learning_rate": 8.233266829702226e-06, "loss": 1.416, "step": 14168 }, { "epoch": 1.110423197492163, "grad_norm": 2.8592560291290283, "learning_rate": 8.233031609910262e-06, "loss": 1.3996, "step": 14169 }, { "epoch": 1.1105015673981191, "grad_norm": 2.1482949256896973, "learning_rate": 8.232796377821509e-06, "loss": 1.296, "step": 14170 }, { "epoch": 1.1105799373040752, "grad_norm": 2.0116593837738037, "learning_rate": 8.232561133436857e-06, "loss": 1.6509, "step": 14171 }, { "epoch": 1.1106583072100313, "grad_norm": 3.0383875370025635, "learning_rate": 8.232325876757203e-06, "loss": 1.3198, "step": 14172 }, { "epoch": 1.1107366771159874, "grad_norm": 2.5347704887390137, "learning_rate": 8.232090607783441e-06, "loss": 1.5451, "step": 14173 }, { "epoch": 1.1108150470219436, "grad_norm": 2.406022787094116, "learning_rate": 8.231855326516468e-06, "loss": 1.1309, "step": 14174 }, { "epoch": 1.1108934169278997, "grad_norm": 1.7610265016555786, "learning_rate": 8.231620032957177e-06, "loss": 1.5389, "step": 14175 }, { "epoch": 1.1109717868338558, "grad_norm": 4.176907062530518, "learning_rate": 8.231384727106463e-06, "loss": 1.2288, "step": 14176 }, { "epoch": 1.111050156739812, "grad_norm": 3.5399720668792725, "learning_rate": 8.23114940896522e-06, "loss": 1.2199, "step": 14177 }, { "epoch": 1.111128526645768, "grad_norm": 2.458935022354126, "learning_rate": 8.230914078534346e-06, "loss": 1.5842, "step": 14178 }, { "epoch": 1.111206896551724, "grad_norm": 3.3284833431243896, "learning_rate": 8.230678735814733e-06, "loss": 1.5001, "step": 14179 }, { "epoch": 1.1112852664576802, "grad_norm": 2.1255548000335693, "learning_rate": 8.230443380807278e-06, "loss": 0.8183, "step": 14180 }, { "epoch": 1.1113636363636363, "grad_norm": 2.628843307495117, "learning_rate": 8.230208013512877e-06, "loss": 1.3962, "step": 14181 }, { "epoch": 1.1114420062695924, "grad_norm": 4.704686164855957, "learning_rate": 8.229972633932422e-06, "loss": 1.6975, "step": 14182 }, { "epoch": 1.1115203761755486, "grad_norm": 2.4508893489837646, "learning_rate": 8.22973724206681e-06, "loss": 1.2078, "step": 14183 }, { "epoch": 1.1115987460815047, "grad_norm": 4.661377429962158, "learning_rate": 8.229501837916937e-06, "loss": 2.1109, "step": 14184 }, { "epoch": 1.1116771159874608, "grad_norm": 1.616087794303894, "learning_rate": 8.2292664214837e-06, "loss": 1.197, "step": 14185 }, { "epoch": 1.1117554858934169, "grad_norm": 3.9783027172088623, "learning_rate": 8.229030992767989e-06, "loss": 1.2824, "step": 14186 }, { "epoch": 1.111833855799373, "grad_norm": 2.681137800216675, "learning_rate": 8.228795551770704e-06, "loss": 1.2953, "step": 14187 }, { "epoch": 1.111912225705329, "grad_norm": 6.516366004943848, "learning_rate": 8.228560098492739e-06, "loss": 1.3329, "step": 14188 }, { "epoch": 1.1119905956112852, "grad_norm": 2.067373275756836, "learning_rate": 8.22832463293499e-06, "loss": 1.6232, "step": 14189 }, { "epoch": 1.1120689655172413, "grad_norm": 2.9086101055145264, "learning_rate": 8.228089155098351e-06, "loss": 1.4443, "step": 14190 }, { "epoch": 1.1121473354231974, "grad_norm": 2.6995625495910645, "learning_rate": 8.22785366498372e-06, "loss": 1.462, "step": 14191 }, { "epoch": 1.1122257053291535, "grad_norm": 6.330233097076416, "learning_rate": 8.22761816259199e-06, "loss": 0.7724, "step": 14192 }, { "epoch": 1.1123040752351097, "grad_norm": 2.543444871902466, "learning_rate": 8.227382647924061e-06, "loss": 1.6743, "step": 14193 }, { "epoch": 1.1123824451410658, "grad_norm": 2.7741029262542725, "learning_rate": 8.227147120980825e-06, "loss": 1.4463, "step": 14194 }, { "epoch": 1.1124608150470219, "grad_norm": 4.360218048095703, "learning_rate": 8.22691158176318e-06, "loss": 1.8908, "step": 14195 }, { "epoch": 1.112539184952978, "grad_norm": 6.09552526473999, "learning_rate": 8.226676030272019e-06, "loss": 1.5348, "step": 14196 }, { "epoch": 1.112617554858934, "grad_norm": 3.9686357975006104, "learning_rate": 8.22644046650824e-06, "loss": 1.4261, "step": 14197 }, { "epoch": 1.1126959247648902, "grad_norm": 2.7110846042633057, "learning_rate": 8.226204890472737e-06, "loss": 1.4897, "step": 14198 }, { "epoch": 1.1127742946708463, "grad_norm": 2.8117494583129883, "learning_rate": 8.225969302166411e-06, "loss": 1.5571, "step": 14199 }, { "epoch": 1.1128526645768024, "grad_norm": 3.261319398880005, "learning_rate": 8.225733701590153e-06, "loss": 1.3023, "step": 14200 }, { "epoch": 1.1129310344827585, "grad_norm": 6.774687767028809, "learning_rate": 8.22549808874486e-06, "loss": 1.1406, "step": 14201 }, { "epoch": 1.1130094043887147, "grad_norm": 3.3901891708374023, "learning_rate": 8.225262463631432e-06, "loss": 1.1225, "step": 14202 }, { "epoch": 1.1130877742946708, "grad_norm": 2.6691341400146484, "learning_rate": 8.225026826250761e-06, "loss": 1.6223, "step": 14203 }, { "epoch": 1.1131661442006269, "grad_norm": 2.951613187789917, "learning_rate": 8.224791176603744e-06, "loss": 1.3039, "step": 14204 }, { "epoch": 1.113244514106583, "grad_norm": 3.886079788208008, "learning_rate": 8.224555514691277e-06, "loss": 1.7186, "step": 14205 }, { "epoch": 1.113322884012539, "grad_norm": 3.2500321865081787, "learning_rate": 8.224319840514257e-06, "loss": 1.3583, "step": 14206 }, { "epoch": 1.1134012539184952, "grad_norm": 2.419110059738159, "learning_rate": 8.22408415407358e-06, "loss": 1.4656, "step": 14207 }, { "epoch": 1.1134796238244513, "grad_norm": 8.415013313293457, "learning_rate": 8.223848455370144e-06, "loss": 1.7438, "step": 14208 }, { "epoch": 1.1135579937304074, "grad_norm": 2.4534995555877686, "learning_rate": 8.223612744404844e-06, "loss": 1.0346, "step": 14209 }, { "epoch": 1.1136363636363635, "grad_norm": 2.5378947257995605, "learning_rate": 8.223377021178579e-06, "loss": 1.0325, "step": 14210 }, { "epoch": 1.1137147335423196, "grad_norm": 3.148878812789917, "learning_rate": 8.22314128569224e-06, "loss": 1.1778, "step": 14211 }, { "epoch": 1.113793103448276, "grad_norm": 3.85241436958313, "learning_rate": 8.22290553794673e-06, "loss": 1.2396, "step": 14212 }, { "epoch": 1.1138714733542319, "grad_norm": 7.992580890655518, "learning_rate": 8.22266977794294e-06, "loss": 1.0158, "step": 14213 }, { "epoch": 1.1139498432601882, "grad_norm": 4.553802490234375, "learning_rate": 8.22243400568177e-06, "loss": 1.1302, "step": 14214 }, { "epoch": 1.114028213166144, "grad_norm": 3.095365285873413, "learning_rate": 8.222198221164118e-06, "loss": 1.2643, "step": 14215 }, { "epoch": 1.1141065830721004, "grad_norm": 7.200237274169922, "learning_rate": 8.221962424390877e-06, "loss": 1.1587, "step": 14216 }, { "epoch": 1.1141849529780563, "grad_norm": 2.9464364051818848, "learning_rate": 8.221726615362949e-06, "loss": 1.2071, "step": 14217 }, { "epoch": 1.1142633228840126, "grad_norm": 2.4373483657836914, "learning_rate": 8.221490794081223e-06, "loss": 1.2097, "step": 14218 }, { "epoch": 1.1143416927899688, "grad_norm": 6.286491870880127, "learning_rate": 8.221254960546605e-06, "loss": 1.6538, "step": 14219 }, { "epoch": 1.1144200626959249, "grad_norm": 2.4508888721466064, "learning_rate": 8.221019114759984e-06, "loss": 1.1672, "step": 14220 }, { "epoch": 1.114498432601881, "grad_norm": 5.287596225738525, "learning_rate": 8.220783256722264e-06, "loss": 1.574, "step": 14221 }, { "epoch": 1.114576802507837, "grad_norm": 5.754870891571045, "learning_rate": 8.220547386434337e-06, "loss": 1.4945, "step": 14222 }, { "epoch": 1.1146551724137932, "grad_norm": 2.841865062713623, "learning_rate": 8.220311503897103e-06, "loss": 1.5076, "step": 14223 }, { "epoch": 1.1147335423197493, "grad_norm": 4.058159828186035, "learning_rate": 8.220075609111456e-06, "loss": 0.8292, "step": 14224 }, { "epoch": 1.1148119122257054, "grad_norm": 2.7256577014923096, "learning_rate": 8.219839702078299e-06, "loss": 1.0887, "step": 14225 }, { "epoch": 1.1148902821316615, "grad_norm": 3.8592278957366943, "learning_rate": 8.219603782798522e-06, "loss": 1.3277, "step": 14226 }, { "epoch": 1.1149686520376176, "grad_norm": 5.458192825317383, "learning_rate": 8.219367851273027e-06, "loss": 1.3686, "step": 14227 }, { "epoch": 1.1150470219435737, "grad_norm": 3.9153637886047363, "learning_rate": 8.219131907502711e-06, "loss": 1.2634, "step": 14228 }, { "epoch": 1.1151253918495299, "grad_norm": 3.46410870552063, "learning_rate": 8.21889595148847e-06, "loss": 1.2799, "step": 14229 }, { "epoch": 1.115203761755486, "grad_norm": 5.146234512329102, "learning_rate": 8.218659983231203e-06, "loss": 1.9823, "step": 14230 }, { "epoch": 1.115282131661442, "grad_norm": 4.894218921661377, "learning_rate": 8.218424002731805e-06, "loss": 1.5563, "step": 14231 }, { "epoch": 1.1153605015673982, "grad_norm": 9.547916412353516, "learning_rate": 8.218188009991176e-06, "loss": 0.8763, "step": 14232 }, { "epoch": 1.1154388714733543, "grad_norm": 3.4705421924591064, "learning_rate": 8.217952005010212e-06, "loss": 0.8165, "step": 14233 }, { "epoch": 1.1155172413793104, "grad_norm": 2.499176025390625, "learning_rate": 8.217715987789813e-06, "loss": 1.2758, "step": 14234 }, { "epoch": 1.1155956112852665, "grad_norm": 3.12536358833313, "learning_rate": 8.217479958330874e-06, "loss": 1.1855, "step": 14235 }, { "epoch": 1.1156739811912226, "grad_norm": 5.561826705932617, "learning_rate": 8.217243916634294e-06, "loss": 0.9563, "step": 14236 }, { "epoch": 1.1157523510971787, "grad_norm": 3.321497917175293, "learning_rate": 8.217007862700972e-06, "loss": 1.5798, "step": 14237 }, { "epoch": 1.1158307210031349, "grad_norm": 3.696774959564209, "learning_rate": 8.216771796531802e-06, "loss": 1.1506, "step": 14238 }, { "epoch": 1.115909090909091, "grad_norm": 3.9472692012786865, "learning_rate": 8.216535718127684e-06, "loss": 1.1669, "step": 14239 }, { "epoch": 1.115987460815047, "grad_norm": 1.9870994091033936, "learning_rate": 8.216299627489519e-06, "loss": 1.55, "step": 14240 }, { "epoch": 1.1160658307210032, "grad_norm": 1.6931103467941284, "learning_rate": 8.216063524618202e-06, "loss": 1.3359, "step": 14241 }, { "epoch": 1.1161442006269593, "grad_norm": 2.4225621223449707, "learning_rate": 8.21582740951463e-06, "loss": 1.2917, "step": 14242 }, { "epoch": 1.1162225705329154, "grad_norm": 5.608778953552246, "learning_rate": 8.215591282179702e-06, "loss": 1.3077, "step": 14243 }, { "epoch": 1.1163009404388715, "grad_norm": 4.372252464294434, "learning_rate": 8.215355142614317e-06, "loss": 1.0841, "step": 14244 }, { "epoch": 1.1163793103448276, "grad_norm": 2.2346138954162598, "learning_rate": 8.215118990819373e-06, "loss": 1.3396, "step": 14245 }, { "epoch": 1.1164576802507837, "grad_norm": 4.331778526306152, "learning_rate": 8.214882826795767e-06, "loss": 2.229, "step": 14246 }, { "epoch": 1.1165360501567398, "grad_norm": 4.871946811676025, "learning_rate": 8.214646650544399e-06, "loss": 1.5734, "step": 14247 }, { "epoch": 1.116614420062696, "grad_norm": 5.349807262420654, "learning_rate": 8.214410462066164e-06, "loss": 0.6638, "step": 14248 }, { "epoch": 1.116692789968652, "grad_norm": 3.916354179382324, "learning_rate": 8.214174261361964e-06, "loss": 1.0793, "step": 14249 }, { "epoch": 1.1167711598746082, "grad_norm": 3.137946605682373, "learning_rate": 8.213938048432697e-06, "loss": 1.3737, "step": 14250 }, { "epoch": 1.1168495297805643, "grad_norm": 2.419599771499634, "learning_rate": 8.21370182327926e-06, "loss": 1.1848, "step": 14251 }, { "epoch": 1.1169278996865204, "grad_norm": 2.39117169380188, "learning_rate": 8.213465585902552e-06, "loss": 1.4524, "step": 14252 }, { "epoch": 1.1170062695924765, "grad_norm": 2.2386038303375244, "learning_rate": 8.213229336303472e-06, "loss": 1.4099, "step": 14253 }, { "epoch": 1.1170846394984326, "grad_norm": 4.475034713745117, "learning_rate": 8.212993074482916e-06, "loss": 1.7339, "step": 14254 }, { "epoch": 1.1171630094043887, "grad_norm": 3.0505363941192627, "learning_rate": 8.212756800441786e-06, "loss": 0.583, "step": 14255 }, { "epoch": 1.1172413793103448, "grad_norm": 3.332542896270752, "learning_rate": 8.212520514180979e-06, "loss": 1.2575, "step": 14256 }, { "epoch": 1.117319749216301, "grad_norm": 3.9840586185455322, "learning_rate": 8.212284215701394e-06, "loss": 1.2218, "step": 14257 }, { "epoch": 1.117398119122257, "grad_norm": 10.78153133392334, "learning_rate": 8.21204790500393e-06, "loss": 1.5338, "step": 14258 }, { "epoch": 1.1174764890282132, "grad_norm": 3.5700700283050537, "learning_rate": 8.211811582089485e-06, "loss": 1.332, "step": 14259 }, { "epoch": 1.1175548589341693, "grad_norm": 6.459782600402832, "learning_rate": 8.211575246958959e-06, "loss": 0.8556, "step": 14260 }, { "epoch": 1.1176332288401254, "grad_norm": 5.531834602355957, "learning_rate": 8.211338899613249e-06, "loss": 1.6209, "step": 14261 }, { "epoch": 1.1177115987460815, "grad_norm": 7.88934850692749, "learning_rate": 8.211102540053258e-06, "loss": 1.7275, "step": 14262 }, { "epoch": 1.1177899686520376, "grad_norm": 2.252084732055664, "learning_rate": 8.21086616827988e-06, "loss": 1.2006, "step": 14263 }, { "epoch": 1.1178683385579937, "grad_norm": 2.631425619125366, "learning_rate": 8.210629784294016e-06, "loss": 1.339, "step": 14264 }, { "epoch": 1.1179467084639498, "grad_norm": 1.9578696489334106, "learning_rate": 8.210393388096567e-06, "loss": 1.1285, "step": 14265 }, { "epoch": 1.118025078369906, "grad_norm": 2.177988290786743, "learning_rate": 8.21015697968843e-06, "loss": 1.0175, "step": 14266 }, { "epoch": 1.118103448275862, "grad_norm": 5.118309497833252, "learning_rate": 8.209920559070504e-06, "loss": 1.0243, "step": 14267 }, { "epoch": 1.1181818181818182, "grad_norm": 1.9529542922973633, "learning_rate": 8.20968412624369e-06, "loss": 1.4101, "step": 14268 }, { "epoch": 1.1182601880877743, "grad_norm": 4.408604145050049, "learning_rate": 8.209447681208884e-06, "loss": 1.847, "step": 14269 }, { "epoch": 1.1183385579937304, "grad_norm": 2.8293585777282715, "learning_rate": 8.209211223966991e-06, "loss": 0.8192, "step": 14270 }, { "epoch": 1.1184169278996865, "grad_norm": 3.120392322540283, "learning_rate": 8.208974754518904e-06, "loss": 1.3959, "step": 14271 }, { "epoch": 1.1184952978056426, "grad_norm": 5.022556781768799, "learning_rate": 8.208738272865527e-06, "loss": 1.4014, "step": 14272 }, { "epoch": 1.1185736677115987, "grad_norm": 2.5390055179595947, "learning_rate": 8.208501779007757e-06, "loss": 1.228, "step": 14273 }, { "epoch": 1.1186520376175548, "grad_norm": 4.664909362792969, "learning_rate": 8.208265272946494e-06, "loss": 0.9682, "step": 14274 }, { "epoch": 1.118730407523511, "grad_norm": 5.289271831512451, "learning_rate": 8.208028754682637e-06, "loss": 1.3121, "step": 14275 }, { "epoch": 1.118808777429467, "grad_norm": 1.992178201675415, "learning_rate": 8.207792224217088e-06, "loss": 1.4274, "step": 14276 }, { "epoch": 1.1188871473354232, "grad_norm": 2.6547160148620605, "learning_rate": 8.207555681550746e-06, "loss": 1.3829, "step": 14277 }, { "epoch": 1.1189655172413793, "grad_norm": 7.512078762054443, "learning_rate": 8.207319126684506e-06, "loss": 2.0589, "step": 14278 }, { "epoch": 1.1190438871473354, "grad_norm": 2.6944122314453125, "learning_rate": 8.207082559619273e-06, "loss": 1.4223, "step": 14279 }, { "epoch": 1.1191222570532915, "grad_norm": 3.428757429122925, "learning_rate": 8.206845980355946e-06, "loss": 1.4425, "step": 14280 }, { "epoch": 1.1192006269592476, "grad_norm": 5.031243801116943, "learning_rate": 8.206609388895423e-06, "loss": 1.559, "step": 14281 }, { "epoch": 1.1192789968652037, "grad_norm": 3.2936015129089355, "learning_rate": 8.206372785238607e-06, "loss": 1.317, "step": 14282 }, { "epoch": 1.1193573667711598, "grad_norm": 4.250743389129639, "learning_rate": 8.206136169386394e-06, "loss": 1.184, "step": 14283 }, { "epoch": 1.119435736677116, "grad_norm": 3.336487293243408, "learning_rate": 8.205899541339686e-06, "loss": 1.5108, "step": 14284 }, { "epoch": 1.119514106583072, "grad_norm": 2.368389844894409, "learning_rate": 8.205662901099383e-06, "loss": 1.4788, "step": 14285 }, { "epoch": 1.1195924764890282, "grad_norm": 3.413466453552246, "learning_rate": 8.205426248666387e-06, "loss": 1.3149, "step": 14286 }, { "epoch": 1.1196708463949843, "grad_norm": 2.9019529819488525, "learning_rate": 8.205189584041591e-06, "loss": 1.3305, "step": 14287 }, { "epoch": 1.1197492163009404, "grad_norm": 2.892688512802124, "learning_rate": 8.204952907225904e-06, "loss": 1.9998, "step": 14288 }, { "epoch": 1.1198275862068965, "grad_norm": 4.812099456787109, "learning_rate": 8.20471621822022e-06, "loss": 1.2419, "step": 14289 }, { "epoch": 1.1199059561128526, "grad_norm": 3.994727849960327, "learning_rate": 8.204479517025445e-06, "loss": 1.2885, "step": 14290 }, { "epoch": 1.1199843260188087, "grad_norm": 2.8144001960754395, "learning_rate": 8.204242803642473e-06, "loss": 1.2369, "step": 14291 }, { "epoch": 1.1200626959247648, "grad_norm": 2.9057345390319824, "learning_rate": 8.204006078072209e-06, "loss": 1.4423, "step": 14292 }, { "epoch": 1.120141065830721, "grad_norm": 3.057180643081665, "learning_rate": 8.203769340315549e-06, "loss": 1.3848, "step": 14293 }, { "epoch": 1.120219435736677, "grad_norm": 2.9084157943725586, "learning_rate": 8.203532590373397e-06, "loss": 1.6658, "step": 14294 }, { "epoch": 1.1202978056426331, "grad_norm": 13.551237106323242, "learning_rate": 8.203295828246654e-06, "loss": 0.9231, "step": 14295 }, { "epoch": 1.1203761755485893, "grad_norm": 4.9276580810546875, "learning_rate": 8.203059053936218e-06, "loss": 1.2351, "step": 14296 }, { "epoch": 1.1204545454545454, "grad_norm": 5.684816837310791, "learning_rate": 8.20282226744299e-06, "loss": 1.6953, "step": 14297 }, { "epoch": 1.1205329153605015, "grad_norm": 2.558058500289917, "learning_rate": 8.202585468767871e-06, "loss": 1.5926, "step": 14298 }, { "epoch": 1.1206112852664576, "grad_norm": 22.704273223876953, "learning_rate": 8.202348657911762e-06, "loss": 2.2605, "step": 14299 }, { "epoch": 1.1206896551724137, "grad_norm": 4.166938781738281, "learning_rate": 8.202111834875564e-06, "loss": 1.6386, "step": 14300 }, { "epoch": 1.1207680250783698, "grad_norm": 2.8719449043273926, "learning_rate": 8.201874999660177e-06, "loss": 1.6105, "step": 14301 }, { "epoch": 1.120846394984326, "grad_norm": 3.6487019062042236, "learning_rate": 8.201638152266502e-06, "loss": 1.0603, "step": 14302 }, { "epoch": 1.120924764890282, "grad_norm": 10.012845993041992, "learning_rate": 8.201401292695442e-06, "loss": 1.6662, "step": 14303 }, { "epoch": 1.1210031347962381, "grad_norm": 6.038884162902832, "learning_rate": 8.201164420947894e-06, "loss": 1.7803, "step": 14304 }, { "epoch": 1.1210815047021943, "grad_norm": 2.746009588241577, "learning_rate": 8.20092753702476e-06, "loss": 1.5422, "step": 14305 }, { "epoch": 1.1211598746081504, "grad_norm": 3.430473566055298, "learning_rate": 8.200690640926943e-06, "loss": 1.0819, "step": 14306 }, { "epoch": 1.1212382445141067, "grad_norm": 3.4840471744537354, "learning_rate": 8.200453732655343e-06, "loss": 1.5519, "step": 14307 }, { "epoch": 1.1213166144200626, "grad_norm": 9.867730140686035, "learning_rate": 8.200216812210859e-06, "loss": 1.223, "step": 14308 }, { "epoch": 1.121394984326019, "grad_norm": 2.598759412765503, "learning_rate": 8.199979879594395e-06, "loss": 1.3432, "step": 14309 }, { "epoch": 1.1214733542319748, "grad_norm": 2.1357274055480957, "learning_rate": 8.199742934806852e-06, "loss": 1.4851, "step": 14310 }, { "epoch": 1.1215517241379311, "grad_norm": 3.989337921142578, "learning_rate": 8.19950597784913e-06, "loss": 1.2843, "step": 14311 }, { "epoch": 1.1216300940438872, "grad_norm": 2.618358612060547, "learning_rate": 8.199269008722128e-06, "loss": 1.1425, "step": 14312 }, { "epoch": 1.1217084639498434, "grad_norm": 2.15740966796875, "learning_rate": 8.199032027426754e-06, "loss": 1.0804, "step": 14313 }, { "epoch": 1.1217868338557995, "grad_norm": 2.596632480621338, "learning_rate": 8.198795033963903e-06, "loss": 1.3048, "step": 14314 }, { "epoch": 1.1218652037617556, "grad_norm": 2.887983560562134, "learning_rate": 8.198558028334478e-06, "loss": 1.3074, "step": 14315 }, { "epoch": 1.1219435736677117, "grad_norm": 4.801840305328369, "learning_rate": 8.198321010539383e-06, "loss": 1.8691, "step": 14316 }, { "epoch": 1.1220219435736678, "grad_norm": 2.841390609741211, "learning_rate": 8.198083980579518e-06, "loss": 1.4923, "step": 14317 }, { "epoch": 1.122100313479624, "grad_norm": 2.679126501083374, "learning_rate": 8.197846938455784e-06, "loss": 1.3679, "step": 14318 }, { "epoch": 1.12217868338558, "grad_norm": 3.6107940673828125, "learning_rate": 8.197609884169081e-06, "loss": 1.4833, "step": 14319 }, { "epoch": 1.1222570532915361, "grad_norm": 3.5774025917053223, "learning_rate": 8.197372817720314e-06, "loss": 1.0047, "step": 14320 }, { "epoch": 1.1223354231974922, "grad_norm": 3.1919546127319336, "learning_rate": 8.197135739110383e-06, "loss": 1.5078, "step": 14321 }, { "epoch": 1.1224137931034484, "grad_norm": 4.4439544677734375, "learning_rate": 8.196898648340187e-06, "loss": 1.2136, "step": 14322 }, { "epoch": 1.1224921630094045, "grad_norm": 3.014988899230957, "learning_rate": 8.196661545410634e-06, "loss": 1.7325, "step": 14323 }, { "epoch": 1.1225705329153606, "grad_norm": 1.9138717651367188, "learning_rate": 8.196424430322621e-06, "loss": 1.3276, "step": 14324 }, { "epoch": 1.1226489028213167, "grad_norm": 3.8274149894714355, "learning_rate": 8.196187303077051e-06, "loss": 1.6687, "step": 14325 }, { "epoch": 1.1227272727272728, "grad_norm": 1.9931267499923706, "learning_rate": 8.195950163674827e-06, "loss": 1.5653, "step": 14326 }, { "epoch": 1.122805642633229, "grad_norm": 9.21087646484375, "learning_rate": 8.19571301211685e-06, "loss": 1.2722, "step": 14327 }, { "epoch": 1.122884012539185, "grad_norm": 1.8341010808944702, "learning_rate": 8.195475848404022e-06, "loss": 1.2226, "step": 14328 }, { "epoch": 1.1229623824451411, "grad_norm": 2.437544584274292, "learning_rate": 8.195238672537245e-06, "loss": 1.4067, "step": 14329 }, { "epoch": 1.1230407523510972, "grad_norm": 3.669116735458374, "learning_rate": 8.195001484517421e-06, "loss": 1.3675, "step": 14330 }, { "epoch": 1.1231191222570533, "grad_norm": 1.9484236240386963, "learning_rate": 8.194764284345453e-06, "loss": 1.2426, "step": 14331 }, { "epoch": 1.1231974921630095, "grad_norm": 2.5552725791931152, "learning_rate": 8.194527072022242e-06, "loss": 1.3012, "step": 14332 }, { "epoch": 1.1232758620689656, "grad_norm": 2.7329623699188232, "learning_rate": 8.19428984754869e-06, "loss": 1.3962, "step": 14333 }, { "epoch": 1.1233542319749217, "grad_norm": 2.849116325378418, "learning_rate": 8.194052610925702e-06, "loss": 1.1793, "step": 14334 }, { "epoch": 1.1234326018808778, "grad_norm": 4.368166923522949, "learning_rate": 8.193815362154178e-06, "loss": 1.1565, "step": 14335 }, { "epoch": 1.123510971786834, "grad_norm": 2.8375790119171143, "learning_rate": 8.19357810123502e-06, "loss": 1.3655, "step": 14336 }, { "epoch": 1.12358934169279, "grad_norm": 2.905630350112915, "learning_rate": 8.193340828169133e-06, "loss": 1.0661, "step": 14337 }, { "epoch": 1.1236677115987461, "grad_norm": 4.692920684814453, "learning_rate": 8.193103542957416e-06, "loss": 1.6265, "step": 14338 }, { "epoch": 1.1237460815047022, "grad_norm": 3.14652681350708, "learning_rate": 8.192866245600775e-06, "loss": 1.2912, "step": 14339 }, { "epoch": 1.1238244514106583, "grad_norm": 2.330662727355957, "learning_rate": 8.192628936100108e-06, "loss": 1.2551, "step": 14340 }, { "epoch": 1.1239028213166145, "grad_norm": 2.1623332500457764, "learning_rate": 8.192391614456321e-06, "loss": 1.173, "step": 14341 }, { "epoch": 1.1239811912225706, "grad_norm": 2.1189262866973877, "learning_rate": 8.192154280670318e-06, "loss": 1.0252, "step": 14342 }, { "epoch": 1.1240595611285267, "grad_norm": 2.190574884414673, "learning_rate": 8.191916934742998e-06, "loss": 1.5131, "step": 14343 }, { "epoch": 1.1241379310344828, "grad_norm": 2.259718418121338, "learning_rate": 8.191679576675266e-06, "loss": 1.2045, "step": 14344 }, { "epoch": 1.124216300940439, "grad_norm": 4.975721836090088, "learning_rate": 8.191442206468024e-06, "loss": 1.6784, "step": 14345 }, { "epoch": 1.124294670846395, "grad_norm": 3.0873613357543945, "learning_rate": 8.191204824122176e-06, "loss": 1.7856, "step": 14346 }, { "epoch": 1.1243730407523511, "grad_norm": 2.2562685012817383, "learning_rate": 8.190967429638623e-06, "loss": 1.1359, "step": 14347 }, { "epoch": 1.1244514106583072, "grad_norm": 4.067415237426758, "learning_rate": 8.19073002301827e-06, "loss": 1.2978, "step": 14348 }, { "epoch": 1.1245297805642633, "grad_norm": 4.27618932723999, "learning_rate": 8.190492604262017e-06, "loss": 1.2803, "step": 14349 }, { "epoch": 1.1246081504702194, "grad_norm": 3.241579055786133, "learning_rate": 8.190255173370768e-06, "loss": 1.0667, "step": 14350 }, { "epoch": 1.1246865203761756, "grad_norm": 2.502016305923462, "learning_rate": 8.19001773034543e-06, "loss": 1.3951, "step": 14351 }, { "epoch": 1.1247648902821317, "grad_norm": 4.053908824920654, "learning_rate": 8.189780275186903e-06, "loss": 1.0547, "step": 14352 }, { "epoch": 1.1248432601880878, "grad_norm": 2.8792004585266113, "learning_rate": 8.189542807896088e-06, "loss": 1.5584, "step": 14353 }, { "epoch": 1.124921630094044, "grad_norm": 2.828606367111206, "learning_rate": 8.189305328473892e-06, "loss": 1.3289, "step": 14354 }, { "epoch": 1.125, "grad_norm": 2.5768656730651855, "learning_rate": 8.189067836921214e-06, "loss": 1.4102, "step": 14355 }, { "epoch": 1.125078369905956, "grad_norm": 3.0907256603240967, "learning_rate": 8.188830333238964e-06, "loss": 1.4091, "step": 14356 }, { "epoch": 1.1251567398119122, "grad_norm": 2.738959312438965, "learning_rate": 8.188592817428039e-06, "loss": 1.7002, "step": 14357 }, { "epoch": 1.1252351097178683, "grad_norm": 2.4812076091766357, "learning_rate": 8.188355289489345e-06, "loss": 1.3635, "step": 14358 }, { "epoch": 1.1253134796238244, "grad_norm": 4.680724143981934, "learning_rate": 8.188117749423784e-06, "loss": 1.5507, "step": 14359 }, { "epoch": 1.1253918495297806, "grad_norm": 2.978893518447876, "learning_rate": 8.187880197232262e-06, "loss": 0.9761, "step": 14360 }, { "epoch": 1.1254702194357367, "grad_norm": 4.073607444763184, "learning_rate": 8.187642632915682e-06, "loss": 0.7551, "step": 14361 }, { "epoch": 1.1255485893416928, "grad_norm": 2.200965642929077, "learning_rate": 8.187405056474945e-06, "loss": 1.1256, "step": 14362 }, { "epoch": 1.1256269592476489, "grad_norm": 3.007537364959717, "learning_rate": 8.187167467910958e-06, "loss": 1.354, "step": 14363 }, { "epoch": 1.125705329153605, "grad_norm": 2.2975573539733887, "learning_rate": 8.186929867224622e-06, "loss": 1.3101, "step": 14364 }, { "epoch": 1.125783699059561, "grad_norm": 3.2451231479644775, "learning_rate": 8.186692254416841e-06, "loss": 1.2199, "step": 14365 }, { "epoch": 1.1258620689655172, "grad_norm": 8.297745704650879, "learning_rate": 8.186454629488521e-06, "loss": 1.6442, "step": 14366 }, { "epoch": 1.1259404388714733, "grad_norm": 3.054986000061035, "learning_rate": 8.186216992440562e-06, "loss": 1.4558, "step": 14367 }, { "epoch": 1.1260188087774294, "grad_norm": 4.180680751800537, "learning_rate": 8.185979343273871e-06, "loss": 1.2815, "step": 14368 }, { "epoch": 1.1260971786833855, "grad_norm": 7.709150314331055, "learning_rate": 8.185741681989352e-06, "loss": 1.2087, "step": 14369 }, { "epoch": 1.1261755485893417, "grad_norm": 10.692361831665039, "learning_rate": 8.185504008587909e-06, "loss": 1.7759, "step": 14370 }, { "epoch": 1.1262539184952978, "grad_norm": 4.620105743408203, "learning_rate": 8.185266323070443e-06, "loss": 0.9795, "step": 14371 }, { "epoch": 1.1263322884012539, "grad_norm": 19.721904754638672, "learning_rate": 8.18502862543786e-06, "loss": 1.9231, "step": 14372 }, { "epoch": 1.12641065830721, "grad_norm": 3.327420234680176, "learning_rate": 8.184790915691064e-06, "loss": 1.7054, "step": 14373 }, { "epoch": 1.126489028213166, "grad_norm": 3.6406452655792236, "learning_rate": 8.18455319383096e-06, "loss": 1.1415, "step": 14374 }, { "epoch": 1.1265673981191222, "grad_norm": 2.81811261177063, "learning_rate": 8.18431545985845e-06, "loss": 1.262, "step": 14375 }, { "epoch": 1.1266457680250783, "grad_norm": 4.256862640380859, "learning_rate": 8.184077713774441e-06, "loss": 1.1871, "step": 14376 }, { "epoch": 1.1267241379310344, "grad_norm": 2.446539878845215, "learning_rate": 8.183839955579834e-06, "loss": 1.4346, "step": 14377 }, { "epoch": 1.1268025078369905, "grad_norm": 2.1752946376800537, "learning_rate": 8.183602185275536e-06, "loss": 1.4108, "step": 14378 }, { "epoch": 1.1268808777429467, "grad_norm": 3.1017110347747803, "learning_rate": 8.183364402862451e-06, "loss": 1.1271, "step": 14379 }, { "epoch": 1.1269592476489028, "grad_norm": 2.9779531955718994, "learning_rate": 8.183126608341483e-06, "loss": 1.7551, "step": 14380 }, { "epoch": 1.1270376175548589, "grad_norm": 2.892845392227173, "learning_rate": 8.182888801713535e-06, "loss": 1.4332, "step": 14381 }, { "epoch": 1.127115987460815, "grad_norm": 2.8994908332824707, "learning_rate": 8.182650982979515e-06, "loss": 1.1558, "step": 14382 }, { "epoch": 1.127194357366771, "grad_norm": 6.791423797607422, "learning_rate": 8.182413152140325e-06, "loss": 1.2586, "step": 14383 }, { "epoch": 1.1272727272727272, "grad_norm": 6.280320167541504, "learning_rate": 8.18217530919687e-06, "loss": 0.6597, "step": 14384 }, { "epoch": 1.1273510971786833, "grad_norm": 5.588624954223633, "learning_rate": 8.181937454150054e-06, "loss": 1.6881, "step": 14385 }, { "epoch": 1.1274294670846394, "grad_norm": 5.512336730957031, "learning_rate": 8.18169958700078e-06, "loss": 1.2548, "step": 14386 }, { "epoch": 1.1275078369905955, "grad_norm": 3.008204460144043, "learning_rate": 8.181461707749958e-06, "loss": 1.351, "step": 14387 }, { "epoch": 1.1275862068965516, "grad_norm": 4.118428707122803, "learning_rate": 8.181223816398488e-06, "loss": 1.1949, "step": 14388 }, { "epoch": 1.1276645768025078, "grad_norm": 3.5538108348846436, "learning_rate": 8.180985912947277e-06, "loss": 1.6039, "step": 14389 }, { "epoch": 1.1277429467084639, "grad_norm": 3.6429219245910645, "learning_rate": 8.180747997397231e-06, "loss": 1.3019, "step": 14390 }, { "epoch": 1.12782131661442, "grad_norm": 2.775305986404419, "learning_rate": 8.180510069749251e-06, "loss": 0.822, "step": 14391 }, { "epoch": 1.127899686520376, "grad_norm": 2.5425338745117188, "learning_rate": 8.180272130004245e-06, "loss": 1.3639, "step": 14392 }, { "epoch": 1.1279780564263322, "grad_norm": 3.8500993251800537, "learning_rate": 8.180034178163119e-06, "loss": 1.1151, "step": 14393 }, { "epoch": 1.1280564263322883, "grad_norm": 2.209296226501465, "learning_rate": 8.179796214226774e-06, "loss": 1.273, "step": 14394 }, { "epoch": 1.1281347962382444, "grad_norm": 2.7476985454559326, "learning_rate": 8.17955823819612e-06, "loss": 0.9609, "step": 14395 }, { "epoch": 1.1282131661442005, "grad_norm": 4.558408737182617, "learning_rate": 8.179320250072057e-06, "loss": 1.1747, "step": 14396 }, { "epoch": 1.1282915360501566, "grad_norm": 6.3112640380859375, "learning_rate": 8.179082249855495e-06, "loss": 1.6841, "step": 14397 }, { "epoch": 1.128369905956113, "grad_norm": 2.921268939971924, "learning_rate": 8.178844237547336e-06, "loss": 1.5141, "step": 14398 }, { "epoch": 1.1284482758620689, "grad_norm": 5.546384334564209, "learning_rate": 8.178606213148485e-06, "loss": 1.4314, "step": 14399 }, { "epoch": 1.1285266457680252, "grad_norm": 3.7080137729644775, "learning_rate": 8.178368176659851e-06, "loss": 1.5872, "step": 14400 }, { "epoch": 1.128605015673981, "grad_norm": 6.58235502243042, "learning_rate": 8.178130128082335e-06, "loss": 0.7005, "step": 14401 }, { "epoch": 1.1286833855799374, "grad_norm": 6.183610916137695, "learning_rate": 8.177892067416846e-06, "loss": 2.205, "step": 14402 }, { "epoch": 1.1287617554858933, "grad_norm": 5.223085403442383, "learning_rate": 8.177653994664288e-06, "loss": 1.3222, "step": 14403 }, { "epoch": 1.1288401253918496, "grad_norm": 3.0325217247009277, "learning_rate": 8.177415909825565e-06, "loss": 1.1404, "step": 14404 }, { "epoch": 1.1289184952978055, "grad_norm": 2.1981945037841797, "learning_rate": 8.177177812901583e-06, "loss": 1.1517, "step": 14405 }, { "epoch": 1.1289968652037619, "grad_norm": 2.3455464839935303, "learning_rate": 8.17693970389325e-06, "loss": 1.3814, "step": 14406 }, { "epoch": 1.1290752351097177, "grad_norm": 2.779078245162964, "learning_rate": 8.17670158280147e-06, "loss": 1.248, "step": 14407 }, { "epoch": 1.129153605015674, "grad_norm": 6.062966823577881, "learning_rate": 8.17646344962715e-06, "loss": 1.4461, "step": 14408 }, { "epoch": 1.1292319749216302, "grad_norm": 9.653276443481445, "learning_rate": 8.176225304371193e-06, "loss": 0.918, "step": 14409 }, { "epoch": 1.1293103448275863, "grad_norm": 2.8400075435638428, "learning_rate": 8.175987147034505e-06, "loss": 1.4954, "step": 14410 }, { "epoch": 1.1293887147335424, "grad_norm": 2.65626859664917, "learning_rate": 8.175748977617997e-06, "loss": 1.2446, "step": 14411 }, { "epoch": 1.1294670846394985, "grad_norm": 2.661034345626831, "learning_rate": 8.175510796122568e-06, "loss": 1.6582, "step": 14412 }, { "epoch": 1.1295454545454546, "grad_norm": 5.792517185211182, "learning_rate": 8.17527260254913e-06, "loss": 1.3816, "step": 14413 }, { "epoch": 1.1296238244514107, "grad_norm": 3.3764662742614746, "learning_rate": 8.175034396898582e-06, "loss": 1.4525, "step": 14414 }, { "epoch": 1.1297021943573669, "grad_norm": 2.898921489715576, "learning_rate": 8.174796179171835e-06, "loss": 1.4235, "step": 14415 }, { "epoch": 1.129780564263323, "grad_norm": 3.105806589126587, "learning_rate": 8.174557949369795e-06, "loss": 1.2394, "step": 14416 }, { "epoch": 1.129858934169279, "grad_norm": 2.8706045150756836, "learning_rate": 8.174319707493367e-06, "loss": 1.4052, "step": 14417 }, { "epoch": 1.1299373040752352, "grad_norm": 2.5149898529052734, "learning_rate": 8.174081453543458e-06, "loss": 1.0916, "step": 14418 }, { "epoch": 1.1300156739811913, "grad_norm": 2.662362813949585, "learning_rate": 8.173843187520974e-06, "loss": 1.768, "step": 14419 }, { "epoch": 1.1300940438871474, "grad_norm": 2.2823987007141113, "learning_rate": 8.173604909426817e-06, "loss": 1.4102, "step": 14420 }, { "epoch": 1.1301724137931035, "grad_norm": 3.5239949226379395, "learning_rate": 8.1733666192619e-06, "loss": 1.6813, "step": 14421 }, { "epoch": 1.1302507836990596, "grad_norm": 2.8684136867523193, "learning_rate": 8.173128317027125e-06, "loss": 1.1567, "step": 14422 }, { "epoch": 1.1303291536050157, "grad_norm": 3.273493766784668, "learning_rate": 8.172890002723401e-06, "loss": 1.3538, "step": 14423 }, { "epoch": 1.1304075235109718, "grad_norm": 1.8991707563400269, "learning_rate": 8.172651676351632e-06, "loss": 1.252, "step": 14424 }, { "epoch": 1.130485893416928, "grad_norm": 3.466177225112915, "learning_rate": 8.172413337912727e-06, "loss": 1.2846, "step": 14425 }, { "epoch": 1.130564263322884, "grad_norm": 2.6952195167541504, "learning_rate": 8.172174987407591e-06, "loss": 1.628, "step": 14426 }, { "epoch": 1.1306426332288402, "grad_norm": 2.8391804695129395, "learning_rate": 8.171936624837129e-06, "loss": 1.0635, "step": 14427 }, { "epoch": 1.1307210031347963, "grad_norm": 3.0756688117980957, "learning_rate": 8.171698250202251e-06, "loss": 1.573, "step": 14428 }, { "epoch": 1.1307993730407524, "grad_norm": 3.7630302906036377, "learning_rate": 8.171459863503861e-06, "loss": 1.3624, "step": 14429 }, { "epoch": 1.1308777429467085, "grad_norm": 4.301482677459717, "learning_rate": 8.171221464742866e-06, "loss": 1.2892, "step": 14430 }, { "epoch": 1.1309561128526646, "grad_norm": 2.3753821849823, "learning_rate": 8.170983053920175e-06, "loss": 1.4309, "step": 14431 }, { "epoch": 1.1310344827586207, "grad_norm": 1.775187373161316, "learning_rate": 8.170744631036693e-06, "loss": 0.9429, "step": 14432 }, { "epoch": 1.1311128526645768, "grad_norm": 3.1145026683807373, "learning_rate": 8.170506196093325e-06, "loss": 1.3208, "step": 14433 }, { "epoch": 1.131191222570533, "grad_norm": 5.553412437438965, "learning_rate": 8.170267749090982e-06, "loss": 1.522, "step": 14434 }, { "epoch": 1.131269592476489, "grad_norm": 2.2904539108276367, "learning_rate": 8.170029290030568e-06, "loss": 1.4188, "step": 14435 }, { "epoch": 1.1313479623824452, "grad_norm": 2.2653205394744873, "learning_rate": 8.16979081891299e-06, "loss": 1.3125, "step": 14436 }, { "epoch": 1.1314263322884013, "grad_norm": 2.549997091293335, "learning_rate": 8.169552335739157e-06, "loss": 1.5479, "step": 14437 }, { "epoch": 1.1315047021943574, "grad_norm": 2.2737081050872803, "learning_rate": 8.169313840509976e-06, "loss": 1.3765, "step": 14438 }, { "epoch": 1.1315830721003135, "grad_norm": 2.681260824203491, "learning_rate": 8.16907533322635e-06, "loss": 1.4453, "step": 14439 }, { "epoch": 1.1316614420062696, "grad_norm": 1.993285059928894, "learning_rate": 8.168836813889192e-06, "loss": 1.5542, "step": 14440 }, { "epoch": 1.1317398119122257, "grad_norm": 2.563770055770874, "learning_rate": 8.168598282499404e-06, "loss": 1.3877, "step": 14441 }, { "epoch": 1.1318181818181818, "grad_norm": 2.3313052654266357, "learning_rate": 8.168359739057899e-06, "loss": 1.4162, "step": 14442 }, { "epoch": 1.131896551724138, "grad_norm": 1.936279058456421, "learning_rate": 8.168121183565578e-06, "loss": 1.7203, "step": 14443 }, { "epoch": 1.131974921630094, "grad_norm": 4.083842754364014, "learning_rate": 8.167882616023352e-06, "loss": 1.1586, "step": 14444 }, { "epoch": 1.1320532915360502, "grad_norm": 2.967728614807129, "learning_rate": 8.167644036432127e-06, "loss": 0.7551, "step": 14445 }, { "epoch": 1.1321316614420063, "grad_norm": 2.9647412300109863, "learning_rate": 8.167405444792812e-06, "loss": 1.3922, "step": 14446 }, { "epoch": 1.1322100313479624, "grad_norm": 4.1399641036987305, "learning_rate": 8.167166841106314e-06, "loss": 1.4717, "step": 14447 }, { "epoch": 1.1322884012539185, "grad_norm": 2.8379249572753906, "learning_rate": 8.166928225373539e-06, "loss": 1.2505, "step": 14448 }, { "epoch": 1.1323667711598746, "grad_norm": 2.78656005859375, "learning_rate": 8.166689597595396e-06, "loss": 1.531, "step": 14449 }, { "epoch": 1.1324451410658307, "grad_norm": 3.3373637199401855, "learning_rate": 8.166450957772792e-06, "loss": 1.627, "step": 14450 }, { "epoch": 1.1325235109717868, "grad_norm": 3.2587223052978516, "learning_rate": 8.166212305906636e-06, "loss": 1.4296, "step": 14451 }, { "epoch": 1.132601880877743, "grad_norm": 9.174209594726562, "learning_rate": 8.165973641997833e-06, "loss": 0.9778, "step": 14452 }, { "epoch": 1.132680250783699, "grad_norm": 2.465850591659546, "learning_rate": 8.165734966047295e-06, "loss": 1.3891, "step": 14453 }, { "epoch": 1.1327586206896552, "grad_norm": 2.2778921127319336, "learning_rate": 8.165496278055926e-06, "loss": 1.187, "step": 14454 }, { "epoch": 1.1328369905956113, "grad_norm": 3.8276846408843994, "learning_rate": 8.165257578024635e-06, "loss": 1.3346, "step": 14455 }, { "epoch": 1.1329153605015674, "grad_norm": 6.266128063201904, "learning_rate": 8.165018865954328e-06, "loss": 1.718, "step": 14456 }, { "epoch": 1.1329937304075235, "grad_norm": 2.615581750869751, "learning_rate": 8.164780141845917e-06, "loss": 1.4895, "step": 14457 }, { "epoch": 1.1330721003134796, "grad_norm": 2.2792327404022217, "learning_rate": 8.164541405700308e-06, "loss": 0.7884, "step": 14458 }, { "epoch": 1.1331504702194357, "grad_norm": 3.491126775741577, "learning_rate": 8.164302657518408e-06, "loss": 1.2352, "step": 14459 }, { "epoch": 1.1332288401253918, "grad_norm": 3.0551066398620605, "learning_rate": 8.164063897301126e-06, "loss": 1.4052, "step": 14460 }, { "epoch": 1.133307210031348, "grad_norm": 2.4545884132385254, "learning_rate": 8.163825125049371e-06, "loss": 1.7656, "step": 14461 }, { "epoch": 1.133385579937304, "grad_norm": 5.886127948760986, "learning_rate": 8.16358634076405e-06, "loss": 1.861, "step": 14462 }, { "epoch": 1.1334639498432602, "grad_norm": 4.349639415740967, "learning_rate": 8.16334754444607e-06, "loss": 1.6108, "step": 14463 }, { "epoch": 1.1335423197492163, "grad_norm": 2.3346269130706787, "learning_rate": 8.163108736096344e-06, "loss": 1.417, "step": 14464 }, { "epoch": 1.1336206896551724, "grad_norm": 3.7875006198883057, "learning_rate": 8.162869915715774e-06, "loss": 1.8458, "step": 14465 }, { "epoch": 1.1336990595611285, "grad_norm": 3.0243537425994873, "learning_rate": 8.162631083305272e-06, "loss": 1.2404, "step": 14466 }, { "epoch": 1.1337774294670846, "grad_norm": 6.552737236022949, "learning_rate": 8.162392238865746e-06, "loss": 1.6129, "step": 14467 }, { "epoch": 1.1338557993730407, "grad_norm": 3.281454563140869, "learning_rate": 8.162153382398106e-06, "loss": 1.4477, "step": 14468 }, { "epoch": 1.1339341692789968, "grad_norm": 3.3452541828155518, "learning_rate": 8.161914513903257e-06, "loss": 1.3537, "step": 14469 }, { "epoch": 1.134012539184953, "grad_norm": 2.2015771865844727, "learning_rate": 8.161675633382109e-06, "loss": 1.6242, "step": 14470 }, { "epoch": 1.134090909090909, "grad_norm": 2.8266170024871826, "learning_rate": 8.161436740835571e-06, "loss": 1.0944, "step": 14471 }, { "epoch": 1.1341692789968651, "grad_norm": 6.051517963409424, "learning_rate": 8.161197836264552e-06, "loss": 1.2108, "step": 14472 }, { "epoch": 1.1342476489028213, "grad_norm": 7.5078325271606445, "learning_rate": 8.160958919669958e-06, "loss": 1.3026, "step": 14473 }, { "epoch": 1.1343260188087774, "grad_norm": 3.2972970008850098, "learning_rate": 8.1607199910527e-06, "loss": 1.4451, "step": 14474 }, { "epoch": 1.1344043887147335, "grad_norm": 3.4258198738098145, "learning_rate": 8.16048105041369e-06, "loss": 1.052, "step": 14475 }, { "epoch": 1.1344827586206896, "grad_norm": 3.094250202178955, "learning_rate": 8.160242097753831e-06, "loss": 1.2835, "step": 14476 }, { "epoch": 1.1345611285266457, "grad_norm": 2.270259380340576, "learning_rate": 8.160003133074036e-06, "loss": 1.269, "step": 14477 }, { "epoch": 1.1346394984326018, "grad_norm": 7.887714862823486, "learning_rate": 8.15976415637521e-06, "loss": 1.5673, "step": 14478 }, { "epoch": 1.134717868338558, "grad_norm": 3.174956798553467, "learning_rate": 8.159525167658266e-06, "loss": 1.2479, "step": 14479 }, { "epoch": 1.134796238244514, "grad_norm": 2.128804922103882, "learning_rate": 8.15928616692411e-06, "loss": 1.3438, "step": 14480 }, { "epoch": 1.1348746081504701, "grad_norm": 2.5862326622009277, "learning_rate": 8.159047154173653e-06, "loss": 0.7017, "step": 14481 }, { "epoch": 1.1349529780564263, "grad_norm": 2.9415316581726074, "learning_rate": 8.158808129407802e-06, "loss": 1.1905, "step": 14482 }, { "epoch": 1.1350313479623824, "grad_norm": 2.974571704864502, "learning_rate": 8.158569092627468e-06, "loss": 1.1603, "step": 14483 }, { "epoch": 1.1351097178683385, "grad_norm": 2.5542588233947754, "learning_rate": 8.15833004383356e-06, "loss": 1.1404, "step": 14484 }, { "epoch": 1.1351880877742946, "grad_norm": 4.516720771789551, "learning_rate": 8.158090983026985e-06, "loss": 1.4699, "step": 14485 }, { "epoch": 1.1352664576802507, "grad_norm": 3.149374485015869, "learning_rate": 8.157851910208657e-06, "loss": 1.7175, "step": 14486 }, { "epoch": 1.1353448275862068, "grad_norm": 2.93558406829834, "learning_rate": 8.157612825379479e-06, "loss": 1.3287, "step": 14487 }, { "epoch": 1.135423197492163, "grad_norm": 2.314418315887451, "learning_rate": 8.157373728540365e-06, "loss": 1.2215, "step": 14488 }, { "epoch": 1.135501567398119, "grad_norm": 2.6792092323303223, "learning_rate": 8.157134619692225e-06, "loss": 1.5789, "step": 14489 }, { "epoch": 1.1355799373040751, "grad_norm": 2.3901498317718506, "learning_rate": 8.156895498835964e-06, "loss": 1.2786, "step": 14490 }, { "epoch": 1.1356583072100315, "grad_norm": 2.813595771789551, "learning_rate": 8.156656365972495e-06, "loss": 1.629, "step": 14491 }, { "epoch": 1.1357366771159874, "grad_norm": 2.2381725311279297, "learning_rate": 8.156417221102727e-06, "loss": 1.1972, "step": 14492 }, { "epoch": 1.1358150470219437, "grad_norm": 2.8941400051116943, "learning_rate": 8.156178064227567e-06, "loss": 1.7523, "step": 14493 }, { "epoch": 1.1358934169278996, "grad_norm": 2.599116086959839, "learning_rate": 8.155938895347928e-06, "loss": 1.3456, "step": 14494 }, { "epoch": 1.135971786833856, "grad_norm": 3.3953592777252197, "learning_rate": 8.15569971446472e-06, "loss": 0.666, "step": 14495 }, { "epoch": 1.1360501567398118, "grad_norm": 3.3030049800872803, "learning_rate": 8.155460521578849e-06, "loss": 1.7075, "step": 14496 }, { "epoch": 1.1361285266457681, "grad_norm": 6.348661422729492, "learning_rate": 8.155221316691228e-06, "loss": 1.4625, "step": 14497 }, { "epoch": 1.136206896551724, "grad_norm": 12.63577651977539, "learning_rate": 8.154982099802765e-06, "loss": 0.7207, "step": 14498 }, { "epoch": 1.1362852664576804, "grad_norm": 2.494093418121338, "learning_rate": 8.15474287091437e-06, "loss": 1.5078, "step": 14499 }, { "epoch": 1.1363636363636362, "grad_norm": 3.515749216079712, "learning_rate": 8.154503630026955e-06, "loss": 1.8683, "step": 14500 }, { "epoch": 1.1364420062695926, "grad_norm": 3.6246256828308105, "learning_rate": 8.154264377141427e-06, "loss": 1.7091, "step": 14501 }, { "epoch": 1.1365203761755487, "grad_norm": 4.240995407104492, "learning_rate": 8.154025112258698e-06, "loss": 1.2521, "step": 14502 }, { "epoch": 1.1365987460815048, "grad_norm": 2.154552459716797, "learning_rate": 8.15378583537968e-06, "loss": 0.9482, "step": 14503 }, { "epoch": 1.136677115987461, "grad_norm": 1.8265933990478516, "learning_rate": 8.153546546505278e-06, "loss": 1.4121, "step": 14504 }, { "epoch": 1.136755485893417, "grad_norm": 3.0883777141571045, "learning_rate": 8.153307245636405e-06, "loss": 1.2664, "step": 14505 }, { "epoch": 1.1368338557993731, "grad_norm": 4.038535118103027, "learning_rate": 8.15306793277397e-06, "loss": 1.8777, "step": 14506 }, { "epoch": 1.1369122257053292, "grad_norm": 2.9863505363464355, "learning_rate": 8.152828607918886e-06, "loss": 1.1287, "step": 14507 }, { "epoch": 1.1369905956112853, "grad_norm": 3.369288444519043, "learning_rate": 8.15258927107206e-06, "loss": 0.5112, "step": 14508 }, { "epoch": 1.1370689655172415, "grad_norm": 2.5074682235717773, "learning_rate": 8.152349922234403e-06, "loss": 1.346, "step": 14509 }, { "epoch": 1.1371473354231976, "grad_norm": 6.801239013671875, "learning_rate": 8.152110561406829e-06, "loss": 0.9496, "step": 14510 }, { "epoch": 1.1372257053291537, "grad_norm": 2.7504642009735107, "learning_rate": 8.151871188590244e-06, "loss": 1.2525, "step": 14511 }, { "epoch": 1.1373040752351098, "grad_norm": 8.593118667602539, "learning_rate": 8.151631803785559e-06, "loss": 1.928, "step": 14512 }, { "epoch": 1.137382445141066, "grad_norm": 3.604461431503296, "learning_rate": 8.151392406993686e-06, "loss": 1.017, "step": 14513 }, { "epoch": 1.137460815047022, "grad_norm": 2.4672086238861084, "learning_rate": 8.151152998215534e-06, "loss": 1.3475, "step": 14514 }, { "epoch": 1.1375391849529781, "grad_norm": 3.7369813919067383, "learning_rate": 8.150913577452017e-06, "loss": 1.4038, "step": 14515 }, { "epoch": 1.1376175548589342, "grad_norm": 2.097076892852783, "learning_rate": 8.150674144704041e-06, "loss": 1.3877, "step": 14516 }, { "epoch": 1.1376959247648903, "grad_norm": 3.57706356048584, "learning_rate": 8.150434699972519e-06, "loss": 1.3603, "step": 14517 }, { "epoch": 1.1377742946708465, "grad_norm": 3.090561866760254, "learning_rate": 8.150195243258364e-06, "loss": 1.5695, "step": 14518 }, { "epoch": 1.1378526645768026, "grad_norm": 2.851130247116089, "learning_rate": 8.149955774562482e-06, "loss": 1.2813, "step": 14519 }, { "epoch": 1.1379310344827587, "grad_norm": 2.7598700523376465, "learning_rate": 8.149716293885785e-06, "loss": 0.7894, "step": 14520 }, { "epoch": 1.1380094043887148, "grad_norm": 4.027021408081055, "learning_rate": 8.149476801229186e-06, "loss": 1.7513, "step": 14521 }, { "epoch": 1.138087774294671, "grad_norm": 4.2717061042785645, "learning_rate": 8.149237296593597e-06, "loss": 1.0695, "step": 14522 }, { "epoch": 1.138166144200627, "grad_norm": 2.7848081588745117, "learning_rate": 8.148997779979926e-06, "loss": 1.4044, "step": 14523 }, { "epoch": 1.1382445141065831, "grad_norm": 3.0869054794311523, "learning_rate": 8.148758251389083e-06, "loss": 1.536, "step": 14524 }, { "epoch": 1.1383228840125392, "grad_norm": 4.582344055175781, "learning_rate": 8.148518710821981e-06, "loss": 1.3201, "step": 14525 }, { "epoch": 1.1384012539184953, "grad_norm": 3.1454391479492188, "learning_rate": 8.148279158279533e-06, "loss": 1.2297, "step": 14526 }, { "epoch": 1.1384796238244514, "grad_norm": 4.03372859954834, "learning_rate": 8.148039593762648e-06, "loss": 1.2941, "step": 14527 }, { "epoch": 1.1385579937304076, "grad_norm": 2.208106756210327, "learning_rate": 8.147800017272235e-06, "loss": 1.361, "step": 14528 }, { "epoch": 1.1386363636363637, "grad_norm": 2.8499948978424072, "learning_rate": 8.14756042880921e-06, "loss": 1.3818, "step": 14529 }, { "epoch": 1.1387147335423198, "grad_norm": 2.8964884281158447, "learning_rate": 8.14732082837448e-06, "loss": 1.7567, "step": 14530 }, { "epoch": 1.138793103448276, "grad_norm": 1.8987668752670288, "learning_rate": 8.14708121596896e-06, "loss": 1.2066, "step": 14531 }, { "epoch": 1.138871473354232, "grad_norm": 2.1467700004577637, "learning_rate": 8.146841591593559e-06, "loss": 1.2651, "step": 14532 }, { "epoch": 1.138949843260188, "grad_norm": 2.619281768798828, "learning_rate": 8.146601955249187e-06, "loss": 1.0375, "step": 14533 }, { "epoch": 1.1390282131661442, "grad_norm": 2.2279155254364014, "learning_rate": 8.146362306936759e-06, "loss": 1.2628, "step": 14534 }, { "epoch": 1.1391065830721003, "grad_norm": 3.90828800201416, "learning_rate": 8.146122646657187e-06, "loss": 1.6197, "step": 14535 }, { "epoch": 1.1391849529780564, "grad_norm": 3.7392799854278564, "learning_rate": 8.145882974411376e-06, "loss": 1.3085, "step": 14536 }, { "epoch": 1.1392633228840126, "grad_norm": 3.570406198501587, "learning_rate": 8.145643290200245e-06, "loss": 1.4173, "step": 14537 }, { "epoch": 1.1393416927899687, "grad_norm": 2.2034659385681152, "learning_rate": 8.1454035940247e-06, "loss": 1.4685, "step": 14538 }, { "epoch": 1.1394200626959248, "grad_norm": 3.0720527172088623, "learning_rate": 8.145163885885659e-06, "loss": 1.5462, "step": 14539 }, { "epoch": 1.1394984326018809, "grad_norm": 3.004033327102661, "learning_rate": 8.144924165784027e-06, "loss": 1.5195, "step": 14540 }, { "epoch": 1.139576802507837, "grad_norm": 3.3442180156707764, "learning_rate": 8.14468443372072e-06, "loss": 1.3554, "step": 14541 }, { "epoch": 1.139655172413793, "grad_norm": 2.3407135009765625, "learning_rate": 8.144444689696648e-06, "loss": 1.2382, "step": 14542 }, { "epoch": 1.1397335423197492, "grad_norm": 4.197659969329834, "learning_rate": 8.144204933712726e-06, "loss": 1.4872, "step": 14543 }, { "epoch": 1.1398119122257053, "grad_norm": 2.613037347793579, "learning_rate": 8.14396516576986e-06, "loss": 1.459, "step": 14544 }, { "epoch": 1.1398902821316614, "grad_norm": 7.630178451538086, "learning_rate": 8.143725385868967e-06, "loss": 1.7056, "step": 14545 }, { "epoch": 1.1399686520376175, "grad_norm": 7.864829063415527, "learning_rate": 8.143485594010956e-06, "loss": 1.3104, "step": 14546 }, { "epoch": 1.1400470219435737, "grad_norm": 3.1335813999176025, "learning_rate": 8.143245790196743e-06, "loss": 1.0698, "step": 14547 }, { "epoch": 1.1401253918495298, "grad_norm": 2.528867483139038, "learning_rate": 8.143005974427236e-06, "loss": 1.2954, "step": 14548 }, { "epoch": 1.1402037617554859, "grad_norm": 2.150367259979248, "learning_rate": 8.142766146703348e-06, "loss": 1.4075, "step": 14549 }, { "epoch": 1.140282131661442, "grad_norm": 3.006726026535034, "learning_rate": 8.142526307025992e-06, "loss": 1.4354, "step": 14550 }, { "epoch": 1.140360501567398, "grad_norm": 2.62161922454834, "learning_rate": 8.142286455396082e-06, "loss": 1.3796, "step": 14551 }, { "epoch": 1.1404388714733542, "grad_norm": 3.1148180961608887, "learning_rate": 8.142046591814526e-06, "loss": 1.2894, "step": 14552 }, { "epoch": 1.1405172413793103, "grad_norm": 2.0468177795410156, "learning_rate": 8.14180671628224e-06, "loss": 1.3155, "step": 14553 }, { "epoch": 1.1405956112852664, "grad_norm": 2.222947359085083, "learning_rate": 8.141566828800133e-06, "loss": 1.5672, "step": 14554 }, { "epoch": 1.1406739811912225, "grad_norm": 3.2379724979400635, "learning_rate": 8.141326929369123e-06, "loss": 1.3432, "step": 14555 }, { "epoch": 1.1407523510971787, "grad_norm": 4.403846263885498, "learning_rate": 8.141087017990114e-06, "loss": 1.5533, "step": 14556 }, { "epoch": 1.1408307210031348, "grad_norm": 2.687443733215332, "learning_rate": 8.140847094664026e-06, "loss": 1.6094, "step": 14557 }, { "epoch": 1.1409090909090909, "grad_norm": 15.006888389587402, "learning_rate": 8.140607159391769e-06, "loss": 1.5295, "step": 14558 }, { "epoch": 1.140987460815047, "grad_norm": 4.06438684463501, "learning_rate": 8.140367212174256e-06, "loss": 1.0138, "step": 14559 }, { "epoch": 1.141065830721003, "grad_norm": 2.641422748565674, "learning_rate": 8.140127253012398e-06, "loss": 1.2647, "step": 14560 }, { "epoch": 1.1411442006269592, "grad_norm": 2.191617250442505, "learning_rate": 8.13988728190711e-06, "loss": 1.3686, "step": 14561 }, { "epoch": 1.1412225705329153, "grad_norm": 2.2284047603607178, "learning_rate": 8.139647298859302e-06, "loss": 1.3435, "step": 14562 }, { "epoch": 1.1413009404388714, "grad_norm": 4.772433280944824, "learning_rate": 8.13940730386989e-06, "loss": 0.694, "step": 14563 }, { "epoch": 1.1413793103448275, "grad_norm": 2.621561050415039, "learning_rate": 8.139167296939783e-06, "loss": 1.0541, "step": 14564 }, { "epoch": 1.1414576802507836, "grad_norm": 2.8666818141937256, "learning_rate": 8.138927278069896e-06, "loss": 1.6604, "step": 14565 }, { "epoch": 1.1415360501567398, "grad_norm": 4.201103210449219, "learning_rate": 8.138687247261145e-06, "loss": 1.4038, "step": 14566 }, { "epoch": 1.1416144200626959, "grad_norm": 2.025667428970337, "learning_rate": 8.138447204514437e-06, "loss": 1.1573, "step": 14567 }, { "epoch": 1.141692789968652, "grad_norm": 2.7593512535095215, "learning_rate": 8.138207149830691e-06, "loss": 1.3345, "step": 14568 }, { "epoch": 1.141771159874608, "grad_norm": 2.225407600402832, "learning_rate": 8.137967083210813e-06, "loss": 1.4085, "step": 14569 }, { "epoch": 1.1418495297805642, "grad_norm": 4.629512786865234, "learning_rate": 8.137727004655722e-06, "loss": 0.8332, "step": 14570 }, { "epoch": 1.1419278996865203, "grad_norm": 3.2628276348114014, "learning_rate": 8.137486914166329e-06, "loss": 1.6714, "step": 14571 }, { "epoch": 1.1420062695924764, "grad_norm": 3.095057964324951, "learning_rate": 8.137246811743547e-06, "loss": 1.4676, "step": 14572 }, { "epoch": 1.1420846394984325, "grad_norm": 3.1094067096710205, "learning_rate": 8.137006697388289e-06, "loss": 1.3312, "step": 14573 }, { "epoch": 1.1421630094043886, "grad_norm": 6.678469657897949, "learning_rate": 8.136766571101468e-06, "loss": 1.2847, "step": 14574 }, { "epoch": 1.1422413793103448, "grad_norm": 2.220777750015259, "learning_rate": 8.136526432884e-06, "loss": 0.8295, "step": 14575 }, { "epoch": 1.1423197492163009, "grad_norm": 3.9308507442474365, "learning_rate": 8.136286282736796e-06, "loss": 1.282, "step": 14576 }, { "epoch": 1.142398119122257, "grad_norm": 3.131197690963745, "learning_rate": 8.13604612066077e-06, "loss": 1.4304, "step": 14577 }, { "epoch": 1.142476489028213, "grad_norm": 3.381681203842163, "learning_rate": 8.135805946656835e-06, "loss": 1.156, "step": 14578 }, { "epoch": 1.1425548589341692, "grad_norm": 3.299656391143799, "learning_rate": 8.135565760725906e-06, "loss": 1.3106, "step": 14579 }, { "epoch": 1.1426332288401253, "grad_norm": 5.087945938110352, "learning_rate": 8.135325562868894e-06, "loss": 1.7133, "step": 14580 }, { "epoch": 1.1427115987460814, "grad_norm": 3.02573561668396, "learning_rate": 8.135085353086714e-06, "loss": 1.0137, "step": 14581 }, { "epoch": 1.1427899686520375, "grad_norm": 2.2893450260162354, "learning_rate": 8.13484513138028e-06, "loss": 1.3287, "step": 14582 }, { "epoch": 1.1428683385579936, "grad_norm": 4.988935947418213, "learning_rate": 8.134604897750505e-06, "loss": 1.2002, "step": 14583 }, { "epoch": 1.14294670846395, "grad_norm": 2.6653385162353516, "learning_rate": 8.134364652198303e-06, "loss": 1.4462, "step": 14584 }, { "epoch": 1.1430250783699059, "grad_norm": 2.7715179920196533, "learning_rate": 8.134124394724587e-06, "loss": 1.326, "step": 14585 }, { "epoch": 1.1431034482758622, "grad_norm": 2.4868009090423584, "learning_rate": 8.133884125330274e-06, "loss": 1.5584, "step": 14586 }, { "epoch": 1.143181818181818, "grad_norm": 2.4426891803741455, "learning_rate": 8.133643844016274e-06, "loss": 1.3292, "step": 14587 }, { "epoch": 1.1432601880877744, "grad_norm": 4.414348125457764, "learning_rate": 8.133403550783502e-06, "loss": 1.5287, "step": 14588 }, { "epoch": 1.1433385579937303, "grad_norm": 2.8967652320861816, "learning_rate": 8.133163245632872e-06, "loss": 1.6111, "step": 14589 }, { "epoch": 1.1434169278996866, "grad_norm": 2.3774585723876953, "learning_rate": 8.1329229285653e-06, "loss": 1.3977, "step": 14590 }, { "epoch": 1.1434952978056425, "grad_norm": 4.372030258178711, "learning_rate": 8.132682599581696e-06, "loss": 1.1107, "step": 14591 }, { "epoch": 1.1435736677115989, "grad_norm": 2.1096396446228027, "learning_rate": 8.132442258682978e-06, "loss": 1.2018, "step": 14592 }, { "epoch": 1.1436520376175547, "grad_norm": 3.1721854209899902, "learning_rate": 8.132201905870059e-06, "loss": 0.9615, "step": 14593 }, { "epoch": 1.143730407523511, "grad_norm": 3.108119487762451, "learning_rate": 8.131961541143852e-06, "loss": 1.3512, "step": 14594 }, { "epoch": 1.143808777429467, "grad_norm": 2.6390984058380127, "learning_rate": 8.131721164505272e-06, "loss": 1.4371, "step": 14595 }, { "epoch": 1.1438871473354233, "grad_norm": 2.0042901039123535, "learning_rate": 8.131480775955233e-06, "loss": 1.1827, "step": 14596 }, { "epoch": 1.1439655172413794, "grad_norm": 3.008841037750244, "learning_rate": 8.13124037549465e-06, "loss": 1.5719, "step": 14597 }, { "epoch": 1.1440438871473355, "grad_norm": 6.66417121887207, "learning_rate": 8.130999963124438e-06, "loss": 0.8097, "step": 14598 }, { "epoch": 1.1441222570532916, "grad_norm": 7.525975704193115, "learning_rate": 8.130759538845508e-06, "loss": 1.2665, "step": 14599 }, { "epoch": 1.1442006269592477, "grad_norm": 4.246124744415283, "learning_rate": 8.130519102658779e-06, "loss": 1.4586, "step": 14600 }, { "epoch": 1.1442789968652038, "grad_norm": 1.9431556463241577, "learning_rate": 8.13027865456516e-06, "loss": 1.2276, "step": 14601 }, { "epoch": 1.14435736677116, "grad_norm": 2.2436141967773438, "learning_rate": 8.130038194565573e-06, "loss": 1.3982, "step": 14602 }, { "epoch": 1.144435736677116, "grad_norm": 2.552060127258301, "learning_rate": 8.129797722660924e-06, "loss": 1.4774, "step": 14603 }, { "epoch": 1.1445141065830722, "grad_norm": 2.0578274726867676, "learning_rate": 8.129557238852136e-06, "loss": 1.501, "step": 14604 }, { "epoch": 1.1445924764890283, "grad_norm": 3.926002264022827, "learning_rate": 8.129316743140117e-06, "loss": 1.5634, "step": 14605 }, { "epoch": 1.1446708463949844, "grad_norm": 2.638176679611206, "learning_rate": 8.129076235525786e-06, "loss": 1.3283, "step": 14606 }, { "epoch": 1.1447492163009405, "grad_norm": 2.2725682258605957, "learning_rate": 8.128835716010056e-06, "loss": 1.2484, "step": 14607 }, { "epoch": 1.1448275862068966, "grad_norm": 2.728236436843872, "learning_rate": 8.128595184593841e-06, "loss": 1.0326, "step": 14608 }, { "epoch": 1.1449059561128527, "grad_norm": 4.863288879394531, "learning_rate": 8.128354641278057e-06, "loss": 1.526, "step": 14609 }, { "epoch": 1.1449843260188088, "grad_norm": 3.0702152252197266, "learning_rate": 8.12811408606362e-06, "loss": 1.8842, "step": 14610 }, { "epoch": 1.145062695924765, "grad_norm": 10.254940032958984, "learning_rate": 8.127873518951444e-06, "loss": 1.3653, "step": 14611 }, { "epoch": 1.145141065830721, "grad_norm": 5.015505313873291, "learning_rate": 8.127632939942442e-06, "loss": 1.7563, "step": 14612 }, { "epoch": 1.1452194357366772, "grad_norm": 3.054995536804199, "learning_rate": 8.127392349037532e-06, "loss": 1.5084, "step": 14613 }, { "epoch": 1.1452978056426333, "grad_norm": 2.8470029830932617, "learning_rate": 8.127151746237627e-06, "loss": 1.1948, "step": 14614 }, { "epoch": 1.1453761755485894, "grad_norm": 2.6229851245880127, "learning_rate": 8.126911131543643e-06, "loss": 1.5784, "step": 14615 }, { "epoch": 1.1454545454545455, "grad_norm": 5.95338249206543, "learning_rate": 8.126670504956495e-06, "loss": 1.8471, "step": 14616 }, { "epoch": 1.1455329153605016, "grad_norm": 1.6250382661819458, "learning_rate": 8.126429866477098e-06, "loss": 1.2213, "step": 14617 }, { "epoch": 1.1456112852664577, "grad_norm": 3.046936273574829, "learning_rate": 8.126189216106369e-06, "loss": 1.0711, "step": 14618 }, { "epoch": 1.1456896551724138, "grad_norm": 2.7867519855499268, "learning_rate": 8.125948553845223e-06, "loss": 1.4558, "step": 14619 }, { "epoch": 1.14576802507837, "grad_norm": 3.226229667663574, "learning_rate": 8.125707879694572e-06, "loss": 1.2407, "step": 14620 }, { "epoch": 1.145846394984326, "grad_norm": 3.9082562923431396, "learning_rate": 8.125467193655334e-06, "loss": 1.5638, "step": 14621 }, { "epoch": 1.1459247648902822, "grad_norm": 3.657508611679077, "learning_rate": 8.125226495728424e-06, "loss": 1.5197, "step": 14622 }, { "epoch": 1.1460031347962383, "grad_norm": 2.5418343544006348, "learning_rate": 8.124985785914758e-06, "loss": 1.3702, "step": 14623 }, { "epoch": 1.1460815047021944, "grad_norm": 3.021454334259033, "learning_rate": 8.124745064215252e-06, "loss": 0.904, "step": 14624 }, { "epoch": 1.1461598746081505, "grad_norm": 2.531491994857788, "learning_rate": 8.12450433063082e-06, "loss": 1.8117, "step": 14625 }, { "epoch": 1.1462382445141066, "grad_norm": 2.4483048915863037, "learning_rate": 8.124263585162378e-06, "loss": 1.3596, "step": 14626 }, { "epoch": 1.1463166144200627, "grad_norm": 2.1281542778015137, "learning_rate": 8.124022827810843e-06, "loss": 1.272, "step": 14627 }, { "epoch": 1.1463949843260188, "grad_norm": 2.3013386726379395, "learning_rate": 8.12378205857713e-06, "loss": 1.5893, "step": 14628 }, { "epoch": 1.146473354231975, "grad_norm": 3.2757115364074707, "learning_rate": 8.123541277462152e-06, "loss": 1.4866, "step": 14629 }, { "epoch": 1.146551724137931, "grad_norm": 2.282592296600342, "learning_rate": 8.123300484466829e-06, "loss": 1.6568, "step": 14630 }, { "epoch": 1.1466300940438872, "grad_norm": 3.63836407661438, "learning_rate": 8.123059679592075e-06, "loss": 1.2925, "step": 14631 }, { "epoch": 1.1467084639498433, "grad_norm": 2.449841260910034, "learning_rate": 8.122818862838808e-06, "loss": 1.4222, "step": 14632 }, { "epoch": 1.1467868338557994, "grad_norm": 2.1177008152008057, "learning_rate": 8.12257803420794e-06, "loss": 1.0311, "step": 14633 }, { "epoch": 1.1468652037617555, "grad_norm": 2.0694921016693115, "learning_rate": 8.122337193700389e-06, "loss": 1.571, "step": 14634 }, { "epoch": 1.1469435736677116, "grad_norm": 1.5548583269119263, "learning_rate": 8.122096341317071e-06, "loss": 1.2734, "step": 14635 }, { "epoch": 1.1470219435736677, "grad_norm": 2.33559250831604, "learning_rate": 8.121855477058903e-06, "loss": 1.4441, "step": 14636 }, { "epoch": 1.1471003134796238, "grad_norm": 2.506057024002075, "learning_rate": 8.1216146009268e-06, "loss": 1.5323, "step": 14637 }, { "epoch": 1.14717868338558, "grad_norm": 2.727540969848633, "learning_rate": 8.121373712921678e-06, "loss": 1.4271, "step": 14638 }, { "epoch": 1.147257053291536, "grad_norm": 2.729508399963379, "learning_rate": 8.121132813044454e-06, "loss": 1.711, "step": 14639 }, { "epoch": 1.1473354231974922, "grad_norm": 3.0905113220214844, "learning_rate": 8.120891901296044e-06, "loss": 1.2213, "step": 14640 }, { "epoch": 1.1474137931034483, "grad_norm": 4.987992763519287, "learning_rate": 8.120650977677364e-06, "loss": 1.5026, "step": 14641 }, { "epoch": 1.1474921630094044, "grad_norm": 2.915990114212036, "learning_rate": 8.12041004218933e-06, "loss": 1.1489, "step": 14642 }, { "epoch": 1.1475705329153605, "grad_norm": 2.697054624557495, "learning_rate": 8.12016909483286e-06, "loss": 1.2471, "step": 14643 }, { "epoch": 1.1476489028213166, "grad_norm": 2.113414764404297, "learning_rate": 8.119928135608869e-06, "loss": 1.5089, "step": 14644 }, { "epoch": 1.1477272727272727, "grad_norm": 1.9683778285980225, "learning_rate": 8.119687164518273e-06, "loss": 1.1685, "step": 14645 }, { "epoch": 1.1478056426332288, "grad_norm": 1.8871257305145264, "learning_rate": 8.11944618156199e-06, "loss": 1.6425, "step": 14646 }, { "epoch": 1.147884012539185, "grad_norm": 2.4328298568725586, "learning_rate": 8.119205186740936e-06, "loss": 1.5536, "step": 14647 }, { "epoch": 1.147962382445141, "grad_norm": 6.114497661590576, "learning_rate": 8.118964180056028e-06, "loss": 1.3255, "step": 14648 }, { "epoch": 1.1480407523510971, "grad_norm": 1.9696401357650757, "learning_rate": 8.118723161508182e-06, "loss": 1.2818, "step": 14649 }, { "epoch": 1.1481191222570533, "grad_norm": 2.7120797634124756, "learning_rate": 8.118482131098316e-06, "loss": 1.1661, "step": 14650 }, { "epoch": 1.1481974921630094, "grad_norm": 2.1641154289245605, "learning_rate": 8.118241088827344e-06, "loss": 1.1593, "step": 14651 }, { "epoch": 1.1482758620689655, "grad_norm": 3.6535167694091797, "learning_rate": 8.118000034696185e-06, "loss": 1.7153, "step": 14652 }, { "epoch": 1.1483542319749216, "grad_norm": 3.4149394035339355, "learning_rate": 8.117758968705754e-06, "loss": 1.4575, "step": 14653 }, { "epoch": 1.1484326018808777, "grad_norm": 4.276657581329346, "learning_rate": 8.117517890856971e-06, "loss": 1.9245, "step": 14654 }, { "epoch": 1.1485109717868338, "grad_norm": 3.8995864391326904, "learning_rate": 8.117276801150752e-06, "loss": 1.5327, "step": 14655 }, { "epoch": 1.14858934169279, "grad_norm": 3.020411252975464, "learning_rate": 8.117035699588012e-06, "loss": 1.6671, "step": 14656 }, { "epoch": 1.148667711598746, "grad_norm": 6.4532318115234375, "learning_rate": 8.11679458616967e-06, "loss": 1.5945, "step": 14657 }, { "epoch": 1.1487460815047021, "grad_norm": 2.325985908508301, "learning_rate": 8.116553460896642e-06, "loss": 1.4138, "step": 14658 }, { "epoch": 1.1488244514106583, "grad_norm": 3.4046928882598877, "learning_rate": 8.116312323769843e-06, "loss": 1.4461, "step": 14659 }, { "epoch": 1.1489028213166144, "grad_norm": 4.160439968109131, "learning_rate": 8.116071174790196e-06, "loss": 1.4976, "step": 14660 }, { "epoch": 1.1489811912225705, "grad_norm": 4.713739395141602, "learning_rate": 8.115830013958612e-06, "loss": 0.9875, "step": 14661 }, { "epoch": 1.1490595611285266, "grad_norm": 2.7445740699768066, "learning_rate": 8.115588841276015e-06, "loss": 1.1024, "step": 14662 }, { "epoch": 1.1491379310344827, "grad_norm": 2.452953577041626, "learning_rate": 8.115347656743315e-06, "loss": 1.029, "step": 14663 }, { "epoch": 1.1492163009404388, "grad_norm": 4.434083461761475, "learning_rate": 8.115106460361432e-06, "loss": 0.8793, "step": 14664 }, { "epoch": 1.149294670846395, "grad_norm": 2.3354790210723877, "learning_rate": 8.114865252131285e-06, "loss": 1.3244, "step": 14665 }, { "epoch": 1.149373040752351, "grad_norm": 2.244013547897339, "learning_rate": 8.114624032053791e-06, "loss": 1.3904, "step": 14666 }, { "epoch": 1.1494514106583071, "grad_norm": 2.2117154598236084, "learning_rate": 8.114382800129867e-06, "loss": 1.3753, "step": 14667 }, { "epoch": 1.1495297805642632, "grad_norm": 3.529160261154175, "learning_rate": 8.114141556360432e-06, "loss": 1.2243, "step": 14668 }, { "epoch": 1.1496081504702194, "grad_norm": 3.086405038833618, "learning_rate": 8.1139003007464e-06, "loss": 1.4024, "step": 14669 }, { "epoch": 1.1496865203761755, "grad_norm": 3.717878818511963, "learning_rate": 8.11365903328869e-06, "loss": 1.0968, "step": 14670 }, { "epoch": 1.1497648902821316, "grad_norm": 8.508761405944824, "learning_rate": 8.113417753988222e-06, "loss": 1.6318, "step": 14671 }, { "epoch": 1.1498432601880877, "grad_norm": 3.0609304904937744, "learning_rate": 8.11317646284591e-06, "loss": 1.1656, "step": 14672 }, { "epoch": 1.1499216300940438, "grad_norm": 2.7427098751068115, "learning_rate": 8.112935159862677e-06, "loss": 1.2778, "step": 14673 }, { "epoch": 1.15, "grad_norm": 2.4630980491638184, "learning_rate": 8.112693845039435e-06, "loss": 1.3587, "step": 14674 }, { "epoch": 1.150078369905956, "grad_norm": 2.8759872913360596, "learning_rate": 8.112452518377106e-06, "loss": 1.2912, "step": 14675 }, { "epoch": 1.1501567398119121, "grad_norm": 5.529915809631348, "learning_rate": 8.112211179876603e-06, "loss": 1.4708, "step": 14676 }, { "epoch": 1.1502351097178682, "grad_norm": 5.799744129180908, "learning_rate": 8.11196982953885e-06, "loss": 1.0959, "step": 14677 }, { "epoch": 1.1503134796238244, "grad_norm": 3.1314773559570312, "learning_rate": 8.111728467364763e-06, "loss": 1.2831, "step": 14678 }, { "epoch": 1.1503918495297807, "grad_norm": 6.0644073486328125, "learning_rate": 8.111487093355256e-06, "loss": 1.4377, "step": 14679 }, { "epoch": 1.1504702194357366, "grad_norm": 1.7955741882324219, "learning_rate": 8.111245707511253e-06, "loss": 1.23, "step": 14680 }, { "epoch": 1.150548589341693, "grad_norm": 2.3629038333892822, "learning_rate": 8.11100430983367e-06, "loss": 1.4315, "step": 14681 }, { "epoch": 1.1506269592476488, "grad_norm": 2.285606622695923, "learning_rate": 8.110762900323423e-06, "loss": 1.4641, "step": 14682 }, { "epoch": 1.1507053291536051, "grad_norm": 2.6710779666900635, "learning_rate": 8.11052147898143e-06, "loss": 1.1412, "step": 14683 }, { "epoch": 1.150783699059561, "grad_norm": 2.8551580905914307, "learning_rate": 8.110280045808614e-06, "loss": 1.414, "step": 14684 }, { "epoch": 1.1508620689655173, "grad_norm": 4.2842817306518555, "learning_rate": 8.110038600805888e-06, "loss": 1.8257, "step": 14685 }, { "epoch": 1.1509404388714732, "grad_norm": 3.9015777111053467, "learning_rate": 8.109797143974174e-06, "loss": 1.5986, "step": 14686 }, { "epoch": 1.1510188087774296, "grad_norm": 2.2168641090393066, "learning_rate": 8.10955567531439e-06, "loss": 1.566, "step": 14687 }, { "epoch": 1.1510971786833855, "grad_norm": 3.2129125595092773, "learning_rate": 8.109314194827451e-06, "loss": 1.5528, "step": 14688 }, { "epoch": 1.1511755485893418, "grad_norm": 3.2095396518707275, "learning_rate": 8.109072702514278e-06, "loss": 1.5376, "step": 14689 }, { "epoch": 1.151253918495298, "grad_norm": 3.535811424255371, "learning_rate": 8.108831198375791e-06, "loss": 1.2837, "step": 14690 }, { "epoch": 1.151332288401254, "grad_norm": 2.799708127975464, "learning_rate": 8.108589682412906e-06, "loss": 1.5326, "step": 14691 }, { "epoch": 1.1514106583072101, "grad_norm": 2.608199119567871, "learning_rate": 8.108348154626542e-06, "loss": 1.575, "step": 14692 }, { "epoch": 1.1514890282131662, "grad_norm": 4.695515155792236, "learning_rate": 8.108106615017619e-06, "loss": 1.9402, "step": 14693 }, { "epoch": 1.1515673981191223, "grad_norm": 2.4454329013824463, "learning_rate": 8.107865063587055e-06, "loss": 1.4041, "step": 14694 }, { "epoch": 1.1516457680250785, "grad_norm": 2.8207805156707764, "learning_rate": 8.107623500335768e-06, "loss": 1.2307, "step": 14695 }, { "epoch": 1.1517241379310346, "grad_norm": 7.213973045349121, "learning_rate": 8.107381925264678e-06, "loss": 1.9991, "step": 14696 }, { "epoch": 1.1518025078369907, "grad_norm": 2.1343188285827637, "learning_rate": 8.107140338374702e-06, "loss": 1.5873, "step": 14697 }, { "epoch": 1.1518808777429468, "grad_norm": 2.540862560272217, "learning_rate": 8.106898739666759e-06, "loss": 1.2135, "step": 14698 }, { "epoch": 1.151959247648903, "grad_norm": 3.9867067337036133, "learning_rate": 8.106657129141773e-06, "loss": 1.4311, "step": 14699 }, { "epoch": 1.152037617554859, "grad_norm": 4.1805500984191895, "learning_rate": 8.106415506800656e-06, "loss": 1.812, "step": 14700 }, { "epoch": 1.1521159874608151, "grad_norm": 2.2034685611724854, "learning_rate": 8.106173872644332e-06, "loss": 1.3893, "step": 14701 }, { "epoch": 1.1521943573667712, "grad_norm": 7.126672744750977, "learning_rate": 8.105932226673717e-06, "loss": 0.7411, "step": 14702 }, { "epoch": 1.1522727272727273, "grad_norm": 2.7980754375457764, "learning_rate": 8.10569056888973e-06, "loss": 1.5294, "step": 14703 }, { "epoch": 1.1523510971786834, "grad_norm": 2.2732203006744385, "learning_rate": 8.105448899293291e-06, "loss": 1.3688, "step": 14704 }, { "epoch": 1.1524294670846396, "grad_norm": 3.2520806789398193, "learning_rate": 8.105207217885322e-06, "loss": 0.8517, "step": 14705 }, { "epoch": 1.1525078369905957, "grad_norm": 3.059238910675049, "learning_rate": 8.104965524666738e-06, "loss": 1.2887, "step": 14706 }, { "epoch": 1.1525862068965518, "grad_norm": 3.634643793106079, "learning_rate": 8.10472381963846e-06, "loss": 1.6952, "step": 14707 }, { "epoch": 1.152664576802508, "grad_norm": 2.1598799228668213, "learning_rate": 8.104482102801408e-06, "loss": 1.3357, "step": 14708 }, { "epoch": 1.152742946708464, "grad_norm": 2.370497703552246, "learning_rate": 8.1042403741565e-06, "loss": 1.6872, "step": 14709 }, { "epoch": 1.15282131661442, "grad_norm": 3.447197675704956, "learning_rate": 8.103998633704657e-06, "loss": 1.4372, "step": 14710 }, { "epoch": 1.1528996865203762, "grad_norm": 3.9249093532562256, "learning_rate": 8.103756881446796e-06, "loss": 1.2603, "step": 14711 }, { "epoch": 1.1529780564263323, "grad_norm": 3.0194313526153564, "learning_rate": 8.103515117383839e-06, "loss": 1.0351, "step": 14712 }, { "epoch": 1.1530564263322884, "grad_norm": 5.265835285186768, "learning_rate": 8.103273341516703e-06, "loss": 1.501, "step": 14713 }, { "epoch": 1.1531347962382446, "grad_norm": 4.670560836791992, "learning_rate": 8.103031553846311e-06, "loss": 1.4206, "step": 14714 }, { "epoch": 1.1532131661442007, "grad_norm": 3.7488269805908203, "learning_rate": 8.102789754373582e-06, "loss": 1.326, "step": 14715 }, { "epoch": 1.1532915360501568, "grad_norm": 3.023404598236084, "learning_rate": 8.10254794309943e-06, "loss": 1.7373, "step": 14716 }, { "epoch": 1.1533699059561129, "grad_norm": 3.7627506256103516, "learning_rate": 8.102306120024784e-06, "loss": 1.2638, "step": 14717 }, { "epoch": 1.153448275862069, "grad_norm": 3.888974905014038, "learning_rate": 8.102064285150557e-06, "loss": 1.1269, "step": 14718 }, { "epoch": 1.153526645768025, "grad_norm": 3.2515997886657715, "learning_rate": 8.10182243847767e-06, "loss": 1.7605, "step": 14719 }, { "epoch": 1.1536050156739812, "grad_norm": 5.794195175170898, "learning_rate": 8.101580580007045e-06, "loss": 1.2664, "step": 14720 }, { "epoch": 1.1536833855799373, "grad_norm": 2.6408839225769043, "learning_rate": 8.101338709739598e-06, "loss": 1.3108, "step": 14721 }, { "epoch": 1.1537617554858934, "grad_norm": 3.552304744720459, "learning_rate": 8.101096827676254e-06, "loss": 1.4806, "step": 14722 }, { "epoch": 1.1538401253918495, "grad_norm": 2.5591163635253906, "learning_rate": 8.10085493381793e-06, "loss": 1.5964, "step": 14723 }, { "epoch": 1.1539184952978057, "grad_norm": 3.0696511268615723, "learning_rate": 8.100613028165547e-06, "loss": 1.8138, "step": 14724 }, { "epoch": 1.1539968652037618, "grad_norm": 3.313572883605957, "learning_rate": 8.100371110720024e-06, "loss": 1.5369, "step": 14725 }, { "epoch": 1.1540752351097179, "grad_norm": 4.648499965667725, "learning_rate": 8.10012918148228e-06, "loss": 0.71, "step": 14726 }, { "epoch": 1.154153605015674, "grad_norm": 3.2289767265319824, "learning_rate": 8.09988724045324e-06, "loss": 1.5735, "step": 14727 }, { "epoch": 1.15423197492163, "grad_norm": 3.2733492851257324, "learning_rate": 8.09964528763382e-06, "loss": 1.1467, "step": 14728 }, { "epoch": 1.1543103448275862, "grad_norm": 4.295113563537598, "learning_rate": 8.099403323024941e-06, "loss": 0.9954, "step": 14729 }, { "epoch": 1.1543887147335423, "grad_norm": 2.4378745555877686, "learning_rate": 8.099161346627524e-06, "loss": 1.4347, "step": 14730 }, { "epoch": 1.1544670846394984, "grad_norm": 2.49357008934021, "learning_rate": 8.098919358442488e-06, "loss": 1.6252, "step": 14731 }, { "epoch": 1.1545454545454545, "grad_norm": 5.664924621582031, "learning_rate": 8.098677358470757e-06, "loss": 1.4084, "step": 14732 }, { "epoch": 1.1546238244514107, "grad_norm": 3.8740928173065186, "learning_rate": 8.098435346713246e-06, "loss": 1.3279, "step": 14733 }, { "epoch": 1.1547021943573668, "grad_norm": 2.13297700881958, "learning_rate": 8.09819332317088e-06, "loss": 1.3295, "step": 14734 }, { "epoch": 1.1547805642633229, "grad_norm": 2.6616716384887695, "learning_rate": 8.097951287844578e-06, "loss": 1.2701, "step": 14735 }, { "epoch": 1.154858934169279, "grad_norm": 2.330190658569336, "learning_rate": 8.097709240735261e-06, "loss": 1.4924, "step": 14736 }, { "epoch": 1.154937304075235, "grad_norm": 3.194432258605957, "learning_rate": 8.097467181843848e-06, "loss": 1.413, "step": 14737 }, { "epoch": 1.1550156739811912, "grad_norm": 2.916607141494751, "learning_rate": 8.097225111171261e-06, "loss": 1.3612, "step": 14738 }, { "epoch": 1.1550940438871473, "grad_norm": 2.579975128173828, "learning_rate": 8.096983028718419e-06, "loss": 0.968, "step": 14739 }, { "epoch": 1.1551724137931034, "grad_norm": 3.4610748291015625, "learning_rate": 8.096740934486247e-06, "loss": 1.4939, "step": 14740 }, { "epoch": 1.1552507836990595, "grad_norm": 3.1314399242401123, "learning_rate": 8.096498828475662e-06, "loss": 1.4515, "step": 14741 }, { "epoch": 1.1553291536050156, "grad_norm": 1.978857398033142, "learning_rate": 8.096256710687587e-06, "loss": 0.9194, "step": 14742 }, { "epoch": 1.1554075235109718, "grad_norm": 3.702739953994751, "learning_rate": 8.09601458112294e-06, "loss": 1.2278, "step": 14743 }, { "epoch": 1.1554858934169279, "grad_norm": 1.9596439599990845, "learning_rate": 8.095772439782645e-06, "loss": 1.5017, "step": 14744 }, { "epoch": 1.155564263322884, "grad_norm": 3.058242082595825, "learning_rate": 8.09553028666762e-06, "loss": 1.0996, "step": 14745 }, { "epoch": 1.15564263322884, "grad_norm": 1.8808656930923462, "learning_rate": 8.095288121778788e-06, "loss": 1.3003, "step": 14746 }, { "epoch": 1.1557210031347962, "grad_norm": 1.9257786273956299, "learning_rate": 8.095045945117073e-06, "loss": 1.2915, "step": 14747 }, { "epoch": 1.1557993730407523, "grad_norm": 4.0720672607421875, "learning_rate": 8.09480375668339e-06, "loss": 1.5481, "step": 14748 }, { "epoch": 1.1558777429467084, "grad_norm": 5.050224781036377, "learning_rate": 8.094561556478665e-06, "loss": 1.8434, "step": 14749 }, { "epoch": 1.1559561128526645, "grad_norm": 7.847919464111328, "learning_rate": 8.094319344503814e-06, "loss": 1.4994, "step": 14750 }, { "epoch": 1.1560344827586206, "grad_norm": 2.8616793155670166, "learning_rate": 8.094077120759764e-06, "loss": 0.8529, "step": 14751 }, { "epoch": 1.1561128526645768, "grad_norm": 2.9396426677703857, "learning_rate": 8.093834885247433e-06, "loss": 1.0543, "step": 14752 }, { "epoch": 1.1561912225705329, "grad_norm": 2.287003993988037, "learning_rate": 8.093592637967745e-06, "loss": 1.1459, "step": 14753 }, { "epoch": 1.156269592476489, "grad_norm": 3.2095611095428467, "learning_rate": 8.093350378921616e-06, "loss": 1.2205, "step": 14754 }, { "epoch": 1.156347962382445, "grad_norm": 2.192329168319702, "learning_rate": 8.093108108109975e-06, "loss": 1.2239, "step": 14755 }, { "epoch": 1.1564263322884012, "grad_norm": 2.64493727684021, "learning_rate": 8.092865825533738e-06, "loss": 1.5981, "step": 14756 }, { "epoch": 1.1565047021943573, "grad_norm": 3.532341241836548, "learning_rate": 8.092623531193827e-06, "loss": 1.4205, "step": 14757 }, { "epoch": 1.1565830721003134, "grad_norm": 2.4764461517333984, "learning_rate": 8.092381225091166e-06, "loss": 1.3919, "step": 14758 }, { "epoch": 1.1566614420062695, "grad_norm": 2.803009033203125, "learning_rate": 8.092138907226675e-06, "loss": 1.3241, "step": 14759 }, { "epoch": 1.1567398119122256, "grad_norm": 5.731475353240967, "learning_rate": 8.091896577601274e-06, "loss": 1.7373, "step": 14760 }, { "epoch": 1.1568181818181817, "grad_norm": 2.389833927154541, "learning_rate": 8.091654236215888e-06, "loss": 1.4271, "step": 14761 }, { "epoch": 1.1568965517241379, "grad_norm": 3.4200966358184814, "learning_rate": 8.091411883071436e-06, "loss": 1.6388, "step": 14762 }, { "epoch": 1.156974921630094, "grad_norm": 2.8568413257598877, "learning_rate": 8.091169518168843e-06, "loss": 1.7094, "step": 14763 }, { "epoch": 1.15705329153605, "grad_norm": 2.2703640460968018, "learning_rate": 8.090927141509029e-06, "loss": 1.2531, "step": 14764 }, { "epoch": 1.1571316614420062, "grad_norm": 3.3094918727874756, "learning_rate": 8.090684753092914e-06, "loss": 1.559, "step": 14765 }, { "epoch": 1.1572100313479623, "grad_norm": 6.770715713500977, "learning_rate": 8.090442352921422e-06, "loss": 1.3284, "step": 14766 }, { "epoch": 1.1572884012539184, "grad_norm": 4.213158130645752, "learning_rate": 8.090199940995476e-06, "loss": 1.2536, "step": 14767 }, { "epoch": 1.1573667711598745, "grad_norm": 5.298280239105225, "learning_rate": 8.089957517315995e-06, "loss": 1.4363, "step": 14768 }, { "epoch": 1.1574451410658306, "grad_norm": 4.067042350769043, "learning_rate": 8.089715081883904e-06, "loss": 1.6354, "step": 14769 }, { "epoch": 1.1575235109717867, "grad_norm": 3.921457290649414, "learning_rate": 8.089472634700123e-06, "loss": 1.1265, "step": 14770 }, { "epoch": 1.1576018808777429, "grad_norm": 2.5629498958587646, "learning_rate": 8.089230175765575e-06, "loss": 1.3956, "step": 14771 }, { "epoch": 1.1576802507836992, "grad_norm": 5.099235534667969, "learning_rate": 8.088987705081182e-06, "loss": 1.1331, "step": 14772 }, { "epoch": 1.157758620689655, "grad_norm": 5.087965488433838, "learning_rate": 8.088745222647868e-06, "loss": 1.5835, "step": 14773 }, { "epoch": 1.1578369905956114, "grad_norm": 3.22092604637146, "learning_rate": 8.088502728466552e-06, "loss": 1.4463, "step": 14774 }, { "epoch": 1.1579153605015673, "grad_norm": 2.813983678817749, "learning_rate": 8.088260222538158e-06, "loss": 1.3341, "step": 14775 }, { "epoch": 1.1579937304075236, "grad_norm": 2.6257517337799072, "learning_rate": 8.08801770486361e-06, "loss": 1.085, "step": 14776 }, { "epoch": 1.1580721003134795, "grad_norm": 2.3348894119262695, "learning_rate": 8.087775175443826e-06, "loss": 1.3716, "step": 14777 }, { "epoch": 1.1581504702194358, "grad_norm": 2.9232404232025146, "learning_rate": 8.087532634279732e-06, "loss": 1.3545, "step": 14778 }, { "epoch": 1.1582288401253917, "grad_norm": 2.5974931716918945, "learning_rate": 8.08729008137225e-06, "loss": 1.3402, "step": 14779 }, { "epoch": 1.158307210031348, "grad_norm": 4.9727091789245605, "learning_rate": 8.087047516722305e-06, "loss": 1.1258, "step": 14780 }, { "epoch": 1.158385579937304, "grad_norm": 2.1807656288146973, "learning_rate": 8.086804940330812e-06, "loss": 1.3431, "step": 14781 }, { "epoch": 1.1584639498432603, "grad_norm": 3.032986879348755, "learning_rate": 8.0865623521987e-06, "loss": 1.3996, "step": 14782 }, { "epoch": 1.1585423197492162, "grad_norm": 4.454694747924805, "learning_rate": 8.086319752326893e-06, "loss": 1.3108, "step": 14783 }, { "epoch": 1.1586206896551725, "grad_norm": 4.641650199890137, "learning_rate": 8.086077140716309e-06, "loss": 1.277, "step": 14784 }, { "epoch": 1.1586990595611286, "grad_norm": 4.152586936950684, "learning_rate": 8.085834517367873e-06, "loss": 1.5812, "step": 14785 }, { "epoch": 1.1587774294670847, "grad_norm": 3.3340275287628174, "learning_rate": 8.085591882282506e-06, "loss": 1.0027, "step": 14786 }, { "epoch": 1.1588557993730408, "grad_norm": 2.4598336219787598, "learning_rate": 8.085349235461135e-06, "loss": 1.4617, "step": 14787 }, { "epoch": 1.158934169278997, "grad_norm": 3.1872363090515137, "learning_rate": 8.085106576904679e-06, "loss": 1.8564, "step": 14788 }, { "epoch": 1.159012539184953, "grad_norm": 2.622143507003784, "learning_rate": 8.084863906614063e-06, "loss": 1.0313, "step": 14789 }, { "epoch": 1.1590909090909092, "grad_norm": 3.545461654663086, "learning_rate": 8.084621224590206e-06, "loss": 0.6742, "step": 14790 }, { "epoch": 1.1591692789968653, "grad_norm": 2.9179694652557373, "learning_rate": 8.084378530834037e-06, "loss": 1.4281, "step": 14791 }, { "epoch": 1.1592476489028214, "grad_norm": 2.7804360389709473, "learning_rate": 8.084135825346475e-06, "loss": 1.4016, "step": 14792 }, { "epoch": 1.1593260188087775, "grad_norm": 3.256584882736206, "learning_rate": 8.083893108128445e-06, "loss": 1.339, "step": 14793 }, { "epoch": 1.1594043887147336, "grad_norm": 3.954228639602661, "learning_rate": 8.08365037918087e-06, "loss": 0.9837, "step": 14794 }, { "epoch": 1.1594827586206897, "grad_norm": 3.5396063327789307, "learning_rate": 8.083407638504673e-06, "loss": 1.1894, "step": 14795 }, { "epoch": 1.1595611285266458, "grad_norm": 8.993210792541504, "learning_rate": 8.083164886100776e-06, "loss": 1.122, "step": 14796 }, { "epoch": 1.159639498432602, "grad_norm": 2.69427752494812, "learning_rate": 8.082922121970105e-06, "loss": 1.177, "step": 14797 }, { "epoch": 1.159717868338558, "grad_norm": 2.278916835784912, "learning_rate": 8.08267934611358e-06, "loss": 0.9051, "step": 14798 }, { "epoch": 1.1597962382445142, "grad_norm": 10.117316246032715, "learning_rate": 8.082436558532127e-06, "loss": 1.003, "step": 14799 }, { "epoch": 1.1598746081504703, "grad_norm": 5.146306037902832, "learning_rate": 8.082193759226669e-06, "loss": 0.8807, "step": 14800 }, { "epoch": 1.1599529780564264, "grad_norm": 2.985123872756958, "learning_rate": 8.081950948198127e-06, "loss": 0.9765, "step": 14801 }, { "epoch": 1.1600313479623825, "grad_norm": 9.823501586914062, "learning_rate": 8.081708125447429e-06, "loss": 1.5964, "step": 14802 }, { "epoch": 1.1601097178683386, "grad_norm": 2.379068374633789, "learning_rate": 8.081465290975495e-06, "loss": 1.1878, "step": 14803 }, { "epoch": 1.1601880877742947, "grad_norm": 3.8498144149780273, "learning_rate": 8.08122244478325e-06, "loss": 1.5003, "step": 14804 }, { "epoch": 1.1602664576802508, "grad_norm": 5.251445770263672, "learning_rate": 8.080979586871618e-06, "loss": 0.8085, "step": 14805 }, { "epoch": 1.160344827586207, "grad_norm": 2.6771023273468018, "learning_rate": 8.080736717241522e-06, "loss": 1.4312, "step": 14806 }, { "epoch": 1.160423197492163, "grad_norm": 2.5124738216400146, "learning_rate": 8.080493835893885e-06, "loss": 1.1685, "step": 14807 }, { "epoch": 1.1605015673981192, "grad_norm": 2.756788969039917, "learning_rate": 8.080250942829632e-06, "loss": 1.6882, "step": 14808 }, { "epoch": 1.1605799373040753, "grad_norm": 3.4132981300354004, "learning_rate": 8.080008038049686e-06, "loss": 1.0979, "step": 14809 }, { "epoch": 1.1606583072100314, "grad_norm": 2.6285715103149414, "learning_rate": 8.079765121554972e-06, "loss": 0.9518, "step": 14810 }, { "epoch": 1.1607366771159875, "grad_norm": 2.9688944816589355, "learning_rate": 8.079522193346413e-06, "loss": 1.5976, "step": 14811 }, { "epoch": 1.1608150470219436, "grad_norm": 8.088519096374512, "learning_rate": 8.079279253424934e-06, "loss": 1.5722, "step": 14812 }, { "epoch": 1.1608934169278997, "grad_norm": 4.355169773101807, "learning_rate": 8.079036301791458e-06, "loss": 1.7952, "step": 14813 }, { "epoch": 1.1609717868338558, "grad_norm": 8.103381156921387, "learning_rate": 8.07879333844691e-06, "loss": 0.5617, "step": 14814 }, { "epoch": 1.161050156739812, "grad_norm": 2.8469409942626953, "learning_rate": 8.078550363392214e-06, "loss": 1.5173, "step": 14815 }, { "epoch": 1.161128526645768, "grad_norm": 2.303741216659546, "learning_rate": 8.078307376628292e-06, "loss": 1.2854, "step": 14816 }, { "epoch": 1.1612068965517242, "grad_norm": 2.092367172241211, "learning_rate": 8.07806437815607e-06, "loss": 1.2864, "step": 14817 }, { "epoch": 1.1612852664576803, "grad_norm": 2.46795916557312, "learning_rate": 8.077821367976474e-06, "loss": 1.4474, "step": 14818 }, { "epoch": 1.1613636363636364, "grad_norm": 3.2699058055877686, "learning_rate": 8.077578346090424e-06, "loss": 1.2623, "step": 14819 }, { "epoch": 1.1614420062695925, "grad_norm": 3.823992967605591, "learning_rate": 8.077335312498849e-06, "loss": 1.3965, "step": 14820 }, { "epoch": 1.1615203761755486, "grad_norm": 2.906388521194458, "learning_rate": 8.077092267202668e-06, "loss": 1.4992, "step": 14821 }, { "epoch": 1.1615987460815047, "grad_norm": 3.936159372329712, "learning_rate": 8.076849210202811e-06, "loss": 1.6851, "step": 14822 }, { "epoch": 1.1616771159874608, "grad_norm": 2.3078739643096924, "learning_rate": 8.0766061415002e-06, "loss": 1.2268, "step": 14823 }, { "epoch": 1.161755485893417, "grad_norm": 2.4660863876342773, "learning_rate": 8.07636306109576e-06, "loss": 1.0895, "step": 14824 }, { "epoch": 1.161833855799373, "grad_norm": 2.7155303955078125, "learning_rate": 8.076119968990412e-06, "loss": 1.2799, "step": 14825 }, { "epoch": 1.1619122257053291, "grad_norm": 7.404566287994385, "learning_rate": 8.075876865185086e-06, "loss": 2.2618, "step": 14826 }, { "epoch": 1.1619905956112853, "grad_norm": 4.427624702453613, "learning_rate": 8.075633749680703e-06, "loss": 2.0062, "step": 14827 }, { "epoch": 1.1620689655172414, "grad_norm": 4.039854526519775, "learning_rate": 8.075390622478189e-06, "loss": 1.6524, "step": 14828 }, { "epoch": 1.1621473354231975, "grad_norm": 2.1999704837799072, "learning_rate": 8.07514748357847e-06, "loss": 1.2042, "step": 14829 }, { "epoch": 1.1622257053291536, "grad_norm": 3.817786455154419, "learning_rate": 8.074904332982469e-06, "loss": 1.4406, "step": 14830 }, { "epoch": 1.1623040752351097, "grad_norm": 2.5643787384033203, "learning_rate": 8.074661170691112e-06, "loss": 1.2059, "step": 14831 }, { "epoch": 1.1623824451410658, "grad_norm": 2.616328477859497, "learning_rate": 8.07441799670532e-06, "loss": 1.5009, "step": 14832 }, { "epoch": 1.162460815047022, "grad_norm": 6.8690972328186035, "learning_rate": 8.074174811026026e-06, "loss": 0.8163, "step": 14833 }, { "epoch": 1.162539184952978, "grad_norm": 2.5377085208892822, "learning_rate": 8.073931613654148e-06, "loss": 0.9723, "step": 14834 }, { "epoch": 1.1626175548589341, "grad_norm": 4.662415504455566, "learning_rate": 8.07368840459061e-06, "loss": 1.1289, "step": 14835 }, { "epoch": 1.1626959247648903, "grad_norm": 3.0907559394836426, "learning_rate": 8.073445183836344e-06, "loss": 1.5081, "step": 14836 }, { "epoch": 1.1627742946708464, "grad_norm": 3.688506841659546, "learning_rate": 8.073201951392269e-06, "loss": 2.0101, "step": 14837 }, { "epoch": 1.1628526645768025, "grad_norm": 3.522467851638794, "learning_rate": 8.072958707259313e-06, "loss": 1.2437, "step": 14838 }, { "epoch": 1.1629310344827586, "grad_norm": 3.8496885299682617, "learning_rate": 8.0727154514384e-06, "loss": 1.5206, "step": 14839 }, { "epoch": 1.1630094043887147, "grad_norm": 2.4373512268066406, "learning_rate": 8.072472183930456e-06, "loss": 1.2436, "step": 14840 }, { "epoch": 1.1630877742946708, "grad_norm": 3.7432196140289307, "learning_rate": 8.072228904736405e-06, "loss": 1.2089, "step": 14841 }, { "epoch": 1.163166144200627, "grad_norm": 3.8437135219573975, "learning_rate": 8.071985613857175e-06, "loss": 1.3881, "step": 14842 }, { "epoch": 1.163244514106583, "grad_norm": 3.5747742652893066, "learning_rate": 8.07174231129369e-06, "loss": 1.2659, "step": 14843 }, { "epoch": 1.1633228840125391, "grad_norm": 1.8904246091842651, "learning_rate": 8.071498997046873e-06, "loss": 1.1388, "step": 14844 }, { "epoch": 1.1634012539184952, "grad_norm": 2.404846668243408, "learning_rate": 8.071255671117652e-06, "loss": 1.3511, "step": 14845 }, { "epoch": 1.1634796238244514, "grad_norm": 2.92710542678833, "learning_rate": 8.071012333506952e-06, "loss": 1.3009, "step": 14846 }, { "epoch": 1.1635579937304075, "grad_norm": 3.4646501541137695, "learning_rate": 8.070768984215698e-06, "loss": 1.2588, "step": 14847 }, { "epoch": 1.1636363636363636, "grad_norm": 2.6094582080841064, "learning_rate": 8.070525623244818e-06, "loss": 1.5194, "step": 14848 }, { "epoch": 1.1637147335423197, "grad_norm": 3.3620822429656982, "learning_rate": 8.070282250595233e-06, "loss": 1.3418, "step": 14849 }, { "epoch": 1.1637931034482758, "grad_norm": 3.7140543460845947, "learning_rate": 8.070038866267872e-06, "loss": 1.1741, "step": 14850 }, { "epoch": 1.163871473354232, "grad_norm": 2.518253803253174, "learning_rate": 8.069795470263661e-06, "loss": 1.0006, "step": 14851 }, { "epoch": 1.163949843260188, "grad_norm": 3.2080202102661133, "learning_rate": 8.069552062583526e-06, "loss": 1.2129, "step": 14852 }, { "epoch": 1.1640282131661441, "grad_norm": 3.516631603240967, "learning_rate": 8.069308643228388e-06, "loss": 1.3254, "step": 14853 }, { "epoch": 1.1641065830721002, "grad_norm": 3.134516716003418, "learning_rate": 8.069065212199178e-06, "loss": 1.3892, "step": 14854 }, { "epoch": 1.1641849529780564, "grad_norm": 4.001633167266846, "learning_rate": 8.068821769496822e-06, "loss": 0.5522, "step": 14855 }, { "epoch": 1.1642633228840125, "grad_norm": 2.378207206726074, "learning_rate": 8.068578315122242e-06, "loss": 1.313, "step": 14856 }, { "epoch": 1.1643416927899686, "grad_norm": 3.349130392074585, "learning_rate": 8.068334849076367e-06, "loss": 1.539, "step": 14857 }, { "epoch": 1.1644200626959247, "grad_norm": 2.0881104469299316, "learning_rate": 8.068091371360122e-06, "loss": 1.4942, "step": 14858 }, { "epoch": 1.1644984326018808, "grad_norm": 2.0151188373565674, "learning_rate": 8.067847881974435e-06, "loss": 1.1801, "step": 14859 }, { "epoch": 1.164576802507837, "grad_norm": 3.0497689247131348, "learning_rate": 8.067604380920228e-06, "loss": 1.3171, "step": 14860 }, { "epoch": 1.164655172413793, "grad_norm": 3.7087349891662598, "learning_rate": 8.067360868198431e-06, "loss": 1.5833, "step": 14861 }, { "epoch": 1.1647335423197491, "grad_norm": 2.873058557510376, "learning_rate": 8.06711734380997e-06, "loss": 1.5775, "step": 14862 }, { "epoch": 1.1648119122257052, "grad_norm": 2.3057780265808105, "learning_rate": 8.066873807755767e-06, "loss": 1.2271, "step": 14863 }, { "epoch": 1.1648902821316613, "grad_norm": 3.374234676361084, "learning_rate": 8.066630260036754e-06, "loss": 1.0498, "step": 14864 }, { "epoch": 1.1649686520376175, "grad_norm": 2.9013209342956543, "learning_rate": 8.066386700653853e-06, "loss": 1.3059, "step": 14865 }, { "epoch": 1.1650470219435736, "grad_norm": 3.1941561698913574, "learning_rate": 8.06614312960799e-06, "loss": 1.8059, "step": 14866 }, { "epoch": 1.16512539184953, "grad_norm": 2.5561037063598633, "learning_rate": 8.065899546900098e-06, "loss": 1.5507, "step": 14867 }, { "epoch": 1.1652037617554858, "grad_norm": 7.18826961517334, "learning_rate": 8.065655952531096e-06, "loss": 1.1287, "step": 14868 }, { "epoch": 1.1652821316614421, "grad_norm": 3.891122817993164, "learning_rate": 8.065412346501912e-06, "loss": 1.5433, "step": 14869 }, { "epoch": 1.165360501567398, "grad_norm": 2.970238208770752, "learning_rate": 8.065168728813476e-06, "loss": 1.4034, "step": 14870 }, { "epoch": 1.1654388714733543, "grad_norm": 1.8373123407363892, "learning_rate": 8.064925099466714e-06, "loss": 1.4103, "step": 14871 }, { "epoch": 1.1655172413793102, "grad_norm": 2.7864878177642822, "learning_rate": 8.064681458462548e-06, "loss": 1.1815, "step": 14872 }, { "epoch": 1.1655956112852666, "grad_norm": 10.968676567077637, "learning_rate": 8.06443780580191e-06, "loss": 0.4547, "step": 14873 }, { "epoch": 1.1656739811912225, "grad_norm": 4.089930534362793, "learning_rate": 8.064194141485724e-06, "loss": 1.9005, "step": 14874 }, { "epoch": 1.1657523510971788, "grad_norm": 4.259558200836182, "learning_rate": 8.063950465514918e-06, "loss": 1.4288, "step": 14875 }, { "epoch": 1.1658307210031347, "grad_norm": 3.675748348236084, "learning_rate": 8.063706777890417e-06, "loss": 1.3174, "step": 14876 }, { "epoch": 1.165909090909091, "grad_norm": 2.7849464416503906, "learning_rate": 8.06346307861315e-06, "loss": 1.5876, "step": 14877 }, { "epoch": 1.1659874608150471, "grad_norm": 1.918384075164795, "learning_rate": 8.063219367684042e-06, "loss": 1.1679, "step": 14878 }, { "epoch": 1.1660658307210032, "grad_norm": 4.876883029937744, "learning_rate": 8.06297564510402e-06, "loss": 1.3111, "step": 14879 }, { "epoch": 1.1661442006269593, "grad_norm": 1.6338787078857422, "learning_rate": 8.062731910874015e-06, "loss": 1.0535, "step": 14880 }, { "epoch": 1.1662225705329154, "grad_norm": 4.551833152770996, "learning_rate": 8.062488164994949e-06, "loss": 1.6817, "step": 14881 }, { "epoch": 1.1663009404388716, "grad_norm": 4.846878528594971, "learning_rate": 8.062244407467751e-06, "loss": 1.3756, "step": 14882 }, { "epoch": 1.1663793103448277, "grad_norm": 2.4855191707611084, "learning_rate": 8.062000638293348e-06, "loss": 1.2102, "step": 14883 }, { "epoch": 1.1664576802507838, "grad_norm": 2.7312018871307373, "learning_rate": 8.061756857472668e-06, "loss": 1.3541, "step": 14884 }, { "epoch": 1.16653605015674, "grad_norm": 2.7256569862365723, "learning_rate": 8.061513065006636e-06, "loss": 1.287, "step": 14885 }, { "epoch": 1.166614420062696, "grad_norm": 3.1462597846984863, "learning_rate": 8.061269260896182e-06, "loss": 0.8353, "step": 14886 }, { "epoch": 1.166692789968652, "grad_norm": 2.5715668201446533, "learning_rate": 8.061025445142232e-06, "loss": 1.4688, "step": 14887 }, { "epoch": 1.1667711598746082, "grad_norm": 4.116706371307373, "learning_rate": 8.060781617745715e-06, "loss": 1.4155, "step": 14888 }, { "epoch": 1.1668495297805643, "grad_norm": 2.7771058082580566, "learning_rate": 8.060537778707554e-06, "loss": 1.4718, "step": 14889 }, { "epoch": 1.1669278996865204, "grad_norm": 4.862083435058594, "learning_rate": 8.060293928028681e-06, "loss": 1.0942, "step": 14890 }, { "epoch": 1.1670062695924766, "grad_norm": 4.6323981285095215, "learning_rate": 8.060050065710022e-06, "loss": 1.4865, "step": 14891 }, { "epoch": 1.1670846394984327, "grad_norm": 3.7561168670654297, "learning_rate": 8.059806191752503e-06, "loss": 1.8683, "step": 14892 }, { "epoch": 1.1671630094043888, "grad_norm": 2.576817750930786, "learning_rate": 8.059562306157054e-06, "loss": 1.0851, "step": 14893 }, { "epoch": 1.1672413793103449, "grad_norm": 3.8378043174743652, "learning_rate": 8.0593184089246e-06, "loss": 1.6418, "step": 14894 }, { "epoch": 1.167319749216301, "grad_norm": 2.8987112045288086, "learning_rate": 8.059074500056073e-06, "loss": 1.7436, "step": 14895 }, { "epoch": 1.167398119122257, "grad_norm": 2.7058558464050293, "learning_rate": 8.058830579552394e-06, "loss": 1.3233, "step": 14896 }, { "epoch": 1.1674764890282132, "grad_norm": 2.4870057106018066, "learning_rate": 8.058586647414496e-06, "loss": 1.2073, "step": 14897 }, { "epoch": 1.1675548589341693, "grad_norm": 3.7491226196289062, "learning_rate": 8.058342703643307e-06, "loss": 1.0545, "step": 14898 }, { "epoch": 1.1676332288401254, "grad_norm": 3.078876256942749, "learning_rate": 8.058098748239754e-06, "loss": 1.8971, "step": 14899 }, { "epoch": 1.1677115987460815, "grad_norm": 3.4938011169433594, "learning_rate": 8.05785478120476e-06, "loss": 1.1114, "step": 14900 }, { "epoch": 1.1677899686520377, "grad_norm": 5.60572624206543, "learning_rate": 8.05761080253926e-06, "loss": 1.0228, "step": 14901 }, { "epoch": 1.1678683385579938, "grad_norm": 2.3716933727264404, "learning_rate": 8.057366812244178e-06, "loss": 1.4055, "step": 14902 }, { "epoch": 1.1679467084639499, "grad_norm": 9.652381896972656, "learning_rate": 8.057122810320443e-06, "loss": 1.569, "step": 14903 }, { "epoch": 1.168025078369906, "grad_norm": 1.9162217378616333, "learning_rate": 8.056878796768984e-06, "loss": 1.3102, "step": 14904 }, { "epoch": 1.168103448275862, "grad_norm": 2.449509620666504, "learning_rate": 8.056634771590728e-06, "loss": 1.2287, "step": 14905 }, { "epoch": 1.1681818181818182, "grad_norm": 3.023365020751953, "learning_rate": 8.056390734786603e-06, "loss": 0.9291, "step": 14906 }, { "epoch": 1.1682601880877743, "grad_norm": 2.0098507404327393, "learning_rate": 8.056146686357536e-06, "loss": 0.8824, "step": 14907 }, { "epoch": 1.1683385579937304, "grad_norm": 2.121971845626831, "learning_rate": 8.05590262630446e-06, "loss": 1.4772, "step": 14908 }, { "epoch": 1.1684169278996865, "grad_norm": 3.2367918491363525, "learning_rate": 8.055658554628297e-06, "loss": 1.5835, "step": 14909 }, { "epoch": 1.1684952978056427, "grad_norm": 3.0528182983398438, "learning_rate": 8.05541447132998e-06, "loss": 1.2549, "step": 14910 }, { "epoch": 1.1685736677115988, "grad_norm": 6.325485706329346, "learning_rate": 8.055170376410435e-06, "loss": 1.5643, "step": 14911 }, { "epoch": 1.1686520376175549, "grad_norm": 3.137375593185425, "learning_rate": 8.054926269870592e-06, "loss": 1.6508, "step": 14912 }, { "epoch": 1.168730407523511, "grad_norm": 3.0632050037384033, "learning_rate": 8.054682151711377e-06, "loss": 1.3695, "step": 14913 }, { "epoch": 1.168808777429467, "grad_norm": 4.68892765045166, "learning_rate": 8.054438021933721e-06, "loss": 1.3059, "step": 14914 }, { "epoch": 1.1688871473354232, "grad_norm": 4.065738201141357, "learning_rate": 8.054193880538553e-06, "loss": 0.9843, "step": 14915 }, { "epoch": 1.1689655172413793, "grad_norm": 3.8788259029388428, "learning_rate": 8.053949727526798e-06, "loss": 1.4094, "step": 14916 }, { "epoch": 1.1690438871473354, "grad_norm": 8.130131721496582, "learning_rate": 8.053705562899388e-06, "loss": 1.6559, "step": 14917 }, { "epoch": 1.1691222570532915, "grad_norm": 2.6204404830932617, "learning_rate": 8.053461386657249e-06, "loss": 1.5559, "step": 14918 }, { "epoch": 1.1692006269592476, "grad_norm": 3.136613368988037, "learning_rate": 8.053217198801314e-06, "loss": 1.4461, "step": 14919 }, { "epoch": 1.1692789968652038, "grad_norm": 2.8353655338287354, "learning_rate": 8.052972999332506e-06, "loss": 1.6056, "step": 14920 }, { "epoch": 1.1693573667711599, "grad_norm": 2.7554540634155273, "learning_rate": 8.052728788251757e-06, "loss": 1.4114, "step": 14921 }, { "epoch": 1.169435736677116, "grad_norm": 2.568319797515869, "learning_rate": 8.052484565559997e-06, "loss": 1.1645, "step": 14922 }, { "epoch": 1.169514106583072, "grad_norm": 3.6944496631622314, "learning_rate": 8.052240331258155e-06, "loss": 1.0937, "step": 14923 }, { "epoch": 1.1695924764890282, "grad_norm": 4.5300188064575195, "learning_rate": 8.051996085347156e-06, "loss": 1.314, "step": 14924 }, { "epoch": 1.1696708463949843, "grad_norm": 2.7449285984039307, "learning_rate": 8.051751827827934e-06, "loss": 1.0676, "step": 14925 }, { "epoch": 1.1697492163009404, "grad_norm": 3.6332638263702393, "learning_rate": 8.051507558701412e-06, "loss": 1.3472, "step": 14926 }, { "epoch": 1.1698275862068965, "grad_norm": 2.130596876144409, "learning_rate": 8.051263277968526e-06, "loss": 1.1957, "step": 14927 }, { "epoch": 1.1699059561128526, "grad_norm": 2.405036687850952, "learning_rate": 8.0510189856302e-06, "loss": 1.6142, "step": 14928 }, { "epoch": 1.1699843260188088, "grad_norm": 5.200263500213623, "learning_rate": 8.050774681687366e-06, "loss": 1.4705, "step": 14929 }, { "epoch": 1.1700626959247649, "grad_norm": 2.6203572750091553, "learning_rate": 8.050530366140951e-06, "loss": 1.5921, "step": 14930 }, { "epoch": 1.170141065830721, "grad_norm": 2.4078352451324463, "learning_rate": 8.050286038991886e-06, "loss": 1.4627, "step": 14931 }, { "epoch": 1.170219435736677, "grad_norm": 4.369567394256592, "learning_rate": 8.050041700241099e-06, "loss": 1.2099, "step": 14932 }, { "epoch": 1.1702978056426332, "grad_norm": 2.9388680458068848, "learning_rate": 8.04979734988952e-06, "loss": 1.0754, "step": 14933 }, { "epoch": 1.1703761755485893, "grad_norm": 3.351367235183716, "learning_rate": 8.049552987938078e-06, "loss": 1.4847, "step": 14934 }, { "epoch": 1.1704545454545454, "grad_norm": 2.6274497509002686, "learning_rate": 8.049308614387703e-06, "loss": 1.3244, "step": 14935 }, { "epoch": 1.1705329153605015, "grad_norm": 1.8926799297332764, "learning_rate": 8.049064229239325e-06, "loss": 0.9229, "step": 14936 }, { "epoch": 1.1706112852664576, "grad_norm": 3.036285877227783, "learning_rate": 8.048819832493873e-06, "loss": 1.6229, "step": 14937 }, { "epoch": 1.1706896551724137, "grad_norm": 1.9632047414779663, "learning_rate": 8.048575424152276e-06, "loss": 1.5005, "step": 14938 }, { "epoch": 1.1707680250783699, "grad_norm": 3.686123847961426, "learning_rate": 8.048331004215463e-06, "loss": 1.5017, "step": 14939 }, { "epoch": 1.170846394984326, "grad_norm": 3.2280311584472656, "learning_rate": 8.048086572684366e-06, "loss": 1.1979, "step": 14940 }, { "epoch": 1.170924764890282, "grad_norm": 2.7206568717956543, "learning_rate": 8.047842129559911e-06, "loss": 1.4277, "step": 14941 }, { "epoch": 1.1710031347962382, "grad_norm": 2.870516061782837, "learning_rate": 8.047597674843032e-06, "loss": 1.0218, "step": 14942 }, { "epoch": 1.1710815047021943, "grad_norm": 3.2513632774353027, "learning_rate": 8.047353208534656e-06, "loss": 1.1186, "step": 14943 }, { "epoch": 1.1711598746081504, "grad_norm": 2.1265432834625244, "learning_rate": 8.047108730635716e-06, "loss": 1.6359, "step": 14944 }, { "epoch": 1.1712382445141065, "grad_norm": 3.0379374027252197, "learning_rate": 8.046864241147137e-06, "loss": 1.6244, "step": 14945 }, { "epoch": 1.1713166144200626, "grad_norm": 3.1623475551605225, "learning_rate": 8.04661974006985e-06, "loss": 1.4306, "step": 14946 }, { "epoch": 1.1713949843260187, "grad_norm": 3.209291934967041, "learning_rate": 8.046375227404788e-06, "loss": 1.2551, "step": 14947 }, { "epoch": 1.1714733542319749, "grad_norm": 2.9499616622924805, "learning_rate": 8.04613070315288e-06, "loss": 1.6731, "step": 14948 }, { "epoch": 1.171551724137931, "grad_norm": 3.3528285026550293, "learning_rate": 8.045886167315055e-06, "loss": 1.3628, "step": 14949 }, { "epoch": 1.171630094043887, "grad_norm": 2.633425235748291, "learning_rate": 8.045641619892243e-06, "loss": 1.2915, "step": 14950 }, { "epoch": 1.1717084639498432, "grad_norm": 2.5837461948394775, "learning_rate": 8.045397060885376e-06, "loss": 1.1561, "step": 14951 }, { "epoch": 1.1717868338557993, "grad_norm": 2.854851484298706, "learning_rate": 8.045152490295381e-06, "loss": 1.6359, "step": 14952 }, { "epoch": 1.1718652037617554, "grad_norm": 2.3237154483795166, "learning_rate": 8.044907908123192e-06, "loss": 1.3117, "step": 14953 }, { "epoch": 1.1719435736677115, "grad_norm": 2.576101779937744, "learning_rate": 8.044663314369736e-06, "loss": 1.176, "step": 14954 }, { "epoch": 1.1720219435736676, "grad_norm": 2.6629250049591064, "learning_rate": 8.044418709035943e-06, "loss": 1.5412, "step": 14955 }, { "epoch": 1.1721003134796237, "grad_norm": 3.1083486080169678, "learning_rate": 8.044174092122747e-06, "loss": 1.298, "step": 14956 }, { "epoch": 1.1721786833855798, "grad_norm": 2.007880926132202, "learning_rate": 8.043929463631077e-06, "loss": 1.6857, "step": 14957 }, { "epoch": 1.172257053291536, "grad_norm": 2.086148977279663, "learning_rate": 8.043684823561861e-06, "loss": 1.4584, "step": 14958 }, { "epoch": 1.172335423197492, "grad_norm": 2.018810987472534, "learning_rate": 8.043440171916032e-06, "loss": 1.6118, "step": 14959 }, { "epoch": 1.1724137931034484, "grad_norm": 1.9768667221069336, "learning_rate": 8.043195508694521e-06, "loss": 1.3958, "step": 14960 }, { "epoch": 1.1724921630094043, "grad_norm": 2.9556496143341064, "learning_rate": 8.042950833898255e-06, "loss": 1.2098, "step": 14961 }, { "epoch": 1.1725705329153606, "grad_norm": 7.9660325050354, "learning_rate": 8.042706147528169e-06, "loss": 1.2795, "step": 14962 }, { "epoch": 1.1726489028213165, "grad_norm": 2.3172144889831543, "learning_rate": 8.042461449585191e-06, "loss": 1.4359, "step": 14963 }, { "epoch": 1.1727272727272728, "grad_norm": 2.823979139328003, "learning_rate": 8.042216740070253e-06, "loss": 1.1716, "step": 14964 }, { "epoch": 1.1728056426332287, "grad_norm": 2.9558310508728027, "learning_rate": 8.041972018984284e-06, "loss": 1.565, "step": 14965 }, { "epoch": 1.172884012539185, "grad_norm": 2.0517826080322266, "learning_rate": 8.041727286328216e-06, "loss": 1.2439, "step": 14966 }, { "epoch": 1.172962382445141, "grad_norm": 2.3066816329956055, "learning_rate": 8.041482542102978e-06, "loss": 1.2593, "step": 14967 }, { "epoch": 1.1730407523510973, "grad_norm": 6.5152506828308105, "learning_rate": 8.041237786309506e-06, "loss": 1.3559, "step": 14968 }, { "epoch": 1.1731191222570532, "grad_norm": 3.539954900741577, "learning_rate": 8.040993018948726e-06, "loss": 1.1364, "step": 14969 }, { "epoch": 1.1731974921630095, "grad_norm": 2.3213579654693604, "learning_rate": 8.04074824002157e-06, "loss": 1.4525, "step": 14970 }, { "epoch": 1.1732758620689656, "grad_norm": 3.1180946826934814, "learning_rate": 8.040503449528971e-06, "loss": 1.5601, "step": 14971 }, { "epoch": 1.1733542319749217, "grad_norm": 2.4538142681121826, "learning_rate": 8.040258647471856e-06, "loss": 1.1185, "step": 14972 }, { "epoch": 1.1734326018808778, "grad_norm": 4.144439697265625, "learning_rate": 8.040013833851161e-06, "loss": 1.4596, "step": 14973 }, { "epoch": 1.173510971786834, "grad_norm": 2.565816879272461, "learning_rate": 8.039769008667814e-06, "loss": 1.4403, "step": 14974 }, { "epoch": 1.17358934169279, "grad_norm": 3.1232001781463623, "learning_rate": 8.039524171922747e-06, "loss": 1.6232, "step": 14975 }, { "epoch": 1.1736677115987462, "grad_norm": 4.69441032409668, "learning_rate": 8.039279323616891e-06, "loss": 1.1639, "step": 14976 }, { "epoch": 1.1737460815047023, "grad_norm": 5.7707839012146, "learning_rate": 8.039034463751179e-06, "loss": 1.6302, "step": 14977 }, { "epoch": 1.1738244514106584, "grad_norm": 2.7914352416992188, "learning_rate": 8.038789592326538e-06, "loss": 1.154, "step": 14978 }, { "epoch": 1.1739028213166145, "grad_norm": 4.716634273529053, "learning_rate": 8.038544709343906e-06, "loss": 1.0362, "step": 14979 }, { "epoch": 1.1739811912225706, "grad_norm": 8.75217342376709, "learning_rate": 8.038299814804209e-06, "loss": 1.8821, "step": 14980 }, { "epoch": 1.1740595611285267, "grad_norm": 2.326237678527832, "learning_rate": 8.038054908708378e-06, "loss": 1.4158, "step": 14981 }, { "epoch": 1.1741379310344828, "grad_norm": 2.195044994354248, "learning_rate": 8.03780999105735e-06, "loss": 1.3964, "step": 14982 }, { "epoch": 1.174216300940439, "grad_norm": 10.468575477600098, "learning_rate": 8.03756506185205e-06, "loss": 2.1293, "step": 14983 }, { "epoch": 1.174294670846395, "grad_norm": 3.0205862522125244, "learning_rate": 8.037320121093416e-06, "loss": 1.2905, "step": 14984 }, { "epoch": 1.1743730407523512, "grad_norm": 3.0619893074035645, "learning_rate": 8.037075168782373e-06, "loss": 1.3768, "step": 14985 }, { "epoch": 1.1744514106583073, "grad_norm": 2.0846669673919678, "learning_rate": 8.036830204919858e-06, "loss": 1.3438, "step": 14986 }, { "epoch": 1.1745297805642634, "grad_norm": 3.1088271141052246, "learning_rate": 8.036585229506798e-06, "loss": 1.6294, "step": 14987 }, { "epoch": 1.1746081504702195, "grad_norm": 3.431196689605713, "learning_rate": 8.036340242544131e-06, "loss": 1.2537, "step": 14988 }, { "epoch": 1.1746865203761756, "grad_norm": 2.4499661922454834, "learning_rate": 8.036095244032785e-06, "loss": 1.4421, "step": 14989 }, { "epoch": 1.1747648902821317, "grad_norm": 2.6004366874694824, "learning_rate": 8.03585023397369e-06, "loss": 1.3167, "step": 14990 }, { "epoch": 1.1748432601880878, "grad_norm": 5.459076404571533, "learning_rate": 8.03560521236778e-06, "loss": 1.8224, "step": 14991 }, { "epoch": 1.174921630094044, "grad_norm": 3.405475378036499, "learning_rate": 8.03536017921599e-06, "loss": 1.49, "step": 14992 }, { "epoch": 1.175, "grad_norm": 5.044186115264893, "learning_rate": 8.035115134519247e-06, "loss": 1.9628, "step": 14993 }, { "epoch": 1.1750783699059562, "grad_norm": 2.9462597370147705, "learning_rate": 8.034870078278485e-06, "loss": 1.6061, "step": 14994 }, { "epoch": 1.1751567398119123, "grad_norm": 10.584321975708008, "learning_rate": 8.034625010494634e-06, "loss": 1.025, "step": 14995 }, { "epoch": 1.1752351097178684, "grad_norm": 4.662937641143799, "learning_rate": 8.034379931168631e-06, "loss": 1.7123, "step": 14996 }, { "epoch": 1.1753134796238245, "grad_norm": 2.360166549682617, "learning_rate": 8.034134840301405e-06, "loss": 1.5376, "step": 14997 }, { "epoch": 1.1753918495297806, "grad_norm": 2.1266608238220215, "learning_rate": 8.033889737893888e-06, "loss": 0.8503, "step": 14998 }, { "epoch": 1.1754702194357367, "grad_norm": 3.254627227783203, "learning_rate": 8.033644623947013e-06, "loss": 1.1418, "step": 14999 }, { "epoch": 1.1755485893416928, "grad_norm": 2.343245029449463, "learning_rate": 8.03339949846171e-06, "loss": 1.188, "step": 15000 }, { "epoch": 1.175626959247649, "grad_norm": 2.911182403564453, "learning_rate": 8.033154361438917e-06, "loss": 1.4997, "step": 15001 }, { "epoch": 1.175705329153605, "grad_norm": 3.171144962310791, "learning_rate": 8.03290921287956e-06, "loss": 1.4196, "step": 15002 }, { "epoch": 1.1757836990595611, "grad_norm": 3.5551226139068604, "learning_rate": 8.032664052784574e-06, "loss": 1.0185, "step": 15003 }, { "epoch": 1.1758620689655173, "grad_norm": 3.107947826385498, "learning_rate": 8.032418881154892e-06, "loss": 1.5351, "step": 15004 }, { "epoch": 1.1759404388714734, "grad_norm": 6.80659294128418, "learning_rate": 8.032173697991448e-06, "loss": 1.3023, "step": 15005 }, { "epoch": 1.1760188087774295, "grad_norm": 2.401397228240967, "learning_rate": 8.031928503295171e-06, "loss": 1.5732, "step": 15006 }, { "epoch": 1.1760971786833856, "grad_norm": 2.7692348957061768, "learning_rate": 8.031683297066996e-06, "loss": 1.0349, "step": 15007 }, { "epoch": 1.1761755485893417, "grad_norm": 4.24854040145874, "learning_rate": 8.031438079307854e-06, "loss": 1.3429, "step": 15008 }, { "epoch": 1.1762539184952978, "grad_norm": 3.4375319480895996, "learning_rate": 8.03119285001868e-06, "loss": 0.9933, "step": 15009 }, { "epoch": 1.176332288401254, "grad_norm": 2.4894490242004395, "learning_rate": 8.030947609200404e-06, "loss": 1.1767, "step": 15010 }, { "epoch": 1.17641065830721, "grad_norm": 3.078305959701538, "learning_rate": 8.03070235685396e-06, "loss": 1.5046, "step": 15011 }, { "epoch": 1.1764890282131661, "grad_norm": 1.8626270294189453, "learning_rate": 8.030457092980282e-06, "loss": 1.3222, "step": 15012 }, { "epoch": 1.1765673981191223, "grad_norm": 3.9094059467315674, "learning_rate": 8.0302118175803e-06, "loss": 1.3857, "step": 15013 }, { "epoch": 1.1766457680250784, "grad_norm": 3.607039213180542, "learning_rate": 8.029966530654952e-06, "loss": 1.2989, "step": 15014 }, { "epoch": 1.1767241379310345, "grad_norm": 4.9026899337768555, "learning_rate": 8.029721232205163e-06, "loss": 1.2492, "step": 15015 }, { "epoch": 1.1768025078369906, "grad_norm": 2.1070995330810547, "learning_rate": 8.029475922231875e-06, "loss": 1.4002, "step": 15016 }, { "epoch": 1.1768808777429467, "grad_norm": 2.8254823684692383, "learning_rate": 8.029230600736013e-06, "loss": 1.9521, "step": 15017 }, { "epoch": 1.1769592476489028, "grad_norm": 2.321647882461548, "learning_rate": 8.028985267718517e-06, "loss": 1.6752, "step": 15018 }, { "epoch": 1.177037617554859, "grad_norm": 12.68572998046875, "learning_rate": 8.028739923180313e-06, "loss": 1.54, "step": 15019 }, { "epoch": 1.177115987460815, "grad_norm": 1.9309797286987305, "learning_rate": 8.028494567122341e-06, "loss": 1.3185, "step": 15020 }, { "epoch": 1.1771943573667711, "grad_norm": 8.20701789855957, "learning_rate": 8.02824919954553e-06, "loss": 2.135, "step": 15021 }, { "epoch": 1.1772727272727272, "grad_norm": 2.687776803970337, "learning_rate": 8.028003820450815e-06, "loss": 1.5491, "step": 15022 }, { "epoch": 1.1773510971786834, "grad_norm": 3.6316826343536377, "learning_rate": 8.027758429839127e-06, "loss": 1.4127, "step": 15023 }, { "epoch": 1.1774294670846395, "grad_norm": 2.8790128231048584, "learning_rate": 8.027513027711404e-06, "loss": 1.2395, "step": 15024 }, { "epoch": 1.1775078369905956, "grad_norm": 2.5631840229034424, "learning_rate": 8.027267614068576e-06, "loss": 1.0982, "step": 15025 }, { "epoch": 1.1775862068965517, "grad_norm": 2.8300869464874268, "learning_rate": 8.027022188911574e-06, "loss": 0.9514, "step": 15026 }, { "epoch": 1.1776645768025078, "grad_norm": 3.6987850666046143, "learning_rate": 8.02677675224134e-06, "loss": 1.2538, "step": 15027 }, { "epoch": 1.177742946708464, "grad_norm": 2.13724684715271, "learning_rate": 8.026531304058795e-06, "loss": 1.3064, "step": 15028 }, { "epoch": 1.17782131661442, "grad_norm": 3.37479829788208, "learning_rate": 8.026285844364885e-06, "loss": 1.4092, "step": 15029 }, { "epoch": 1.1778996865203761, "grad_norm": 6.895569801330566, "learning_rate": 8.026040373160534e-06, "loss": 1.7004, "step": 15030 }, { "epoch": 1.1779780564263322, "grad_norm": 4.271060943603516, "learning_rate": 8.025794890446681e-06, "loss": 1.377, "step": 15031 }, { "epoch": 1.1780564263322884, "grad_norm": 2.9919497966766357, "learning_rate": 8.02554939622426e-06, "loss": 1.3703, "step": 15032 }, { "epoch": 1.1781347962382445, "grad_norm": 6.365891456604004, "learning_rate": 8.025303890494203e-06, "loss": 0.8596, "step": 15033 }, { "epoch": 1.1782131661442006, "grad_norm": 2.190504789352417, "learning_rate": 8.025058373257442e-06, "loss": 1.2308, "step": 15034 }, { "epoch": 1.1782915360501567, "grad_norm": 2.8890159130096436, "learning_rate": 8.024812844514912e-06, "loss": 1.1141, "step": 15035 }, { "epoch": 1.1783699059561128, "grad_norm": 2.8136308193206787, "learning_rate": 8.024567304267549e-06, "loss": 1.6046, "step": 15036 }, { "epoch": 1.178448275862069, "grad_norm": 2.178666114807129, "learning_rate": 8.024321752516286e-06, "loss": 1.464, "step": 15037 }, { "epoch": 1.178526645768025, "grad_norm": 2.3055267333984375, "learning_rate": 8.024076189262057e-06, "loss": 1.5299, "step": 15038 }, { "epoch": 1.1786050156739811, "grad_norm": 2.8201189041137695, "learning_rate": 8.023830614505794e-06, "loss": 1.49, "step": 15039 }, { "epoch": 1.1786833855799372, "grad_norm": 6.659368515014648, "learning_rate": 8.023585028248435e-06, "loss": 0.7708, "step": 15040 }, { "epoch": 1.1787617554858933, "grad_norm": 4.055975914001465, "learning_rate": 8.023339430490909e-06, "loss": 1.5315, "step": 15041 }, { "epoch": 1.1788401253918495, "grad_norm": 2.893855571746826, "learning_rate": 8.023093821234154e-06, "loss": 1.643, "step": 15042 }, { "epoch": 1.1789184952978056, "grad_norm": 1.9346531629562378, "learning_rate": 8.0228482004791e-06, "loss": 1.459, "step": 15043 }, { "epoch": 1.1789968652037617, "grad_norm": 2.673008680343628, "learning_rate": 8.022602568226688e-06, "loss": 1.3197, "step": 15044 }, { "epoch": 1.1790752351097178, "grad_norm": 4.235307216644287, "learning_rate": 8.022356924477845e-06, "loss": 1.5979, "step": 15045 }, { "epoch": 1.179153605015674, "grad_norm": 2.456880569458008, "learning_rate": 8.02211126923351e-06, "loss": 1.5505, "step": 15046 }, { "epoch": 1.17923197492163, "grad_norm": 2.138749122619629, "learning_rate": 8.021865602494616e-06, "loss": 1.2251, "step": 15047 }, { "epoch": 1.1793103448275861, "grad_norm": 2.592137098312378, "learning_rate": 8.021619924262098e-06, "loss": 1.0584, "step": 15048 }, { "epoch": 1.1793887147335422, "grad_norm": 2.37608003616333, "learning_rate": 8.021374234536889e-06, "loss": 1.4882, "step": 15049 }, { "epoch": 1.1794670846394983, "grad_norm": 9.003028869628906, "learning_rate": 8.021128533319925e-06, "loss": 0.8049, "step": 15050 }, { "epoch": 1.1795454545454545, "grad_norm": 3.455228567123413, "learning_rate": 8.020882820612139e-06, "loss": 1.5091, "step": 15051 }, { "epoch": 1.1796238244514106, "grad_norm": 4.0923051834106445, "learning_rate": 8.020637096414467e-06, "loss": 1.2074, "step": 15052 }, { "epoch": 1.179702194357367, "grad_norm": 2.4730069637298584, "learning_rate": 8.020391360727843e-06, "loss": 0.7981, "step": 15053 }, { "epoch": 1.1797805642633228, "grad_norm": 2.1286563873291016, "learning_rate": 8.020145613553202e-06, "loss": 1.2581, "step": 15054 }, { "epoch": 1.1798589341692791, "grad_norm": 2.211533546447754, "learning_rate": 8.019899854891478e-06, "loss": 0.9581, "step": 15055 }, { "epoch": 1.179937304075235, "grad_norm": 3.3596558570861816, "learning_rate": 8.019654084743606e-06, "loss": 1.5027, "step": 15056 }, { "epoch": 1.1800156739811913, "grad_norm": 5.482618808746338, "learning_rate": 8.01940830311052e-06, "loss": 1.2483, "step": 15057 }, { "epoch": 1.1800940438871472, "grad_norm": 4.349824905395508, "learning_rate": 8.019162509993158e-06, "loss": 1.2936, "step": 15058 }, { "epoch": 1.1801724137931036, "grad_norm": 4.034682273864746, "learning_rate": 8.01891670539245e-06, "loss": 1.2607, "step": 15059 }, { "epoch": 1.1802507836990594, "grad_norm": 2.529552459716797, "learning_rate": 8.018670889309335e-06, "loss": 1.1899, "step": 15060 }, { "epoch": 1.1803291536050158, "grad_norm": 4.003371238708496, "learning_rate": 8.018425061744746e-06, "loss": 0.4444, "step": 15061 }, { "epoch": 1.1804075235109717, "grad_norm": 2.8038363456726074, "learning_rate": 8.018179222699617e-06, "loss": 1.3674, "step": 15062 }, { "epoch": 1.180485893416928, "grad_norm": 2.323221445083618, "learning_rate": 8.017933372174886e-06, "loss": 1.5071, "step": 15063 }, { "epoch": 1.1805642633228839, "grad_norm": 3.1203086376190186, "learning_rate": 8.017687510171485e-06, "loss": 1.2572, "step": 15064 }, { "epoch": 1.1806426332288402, "grad_norm": 2.0154666900634766, "learning_rate": 8.017441636690353e-06, "loss": 1.1607, "step": 15065 }, { "epoch": 1.1807210031347963, "grad_norm": 13.70081901550293, "learning_rate": 8.017195751732421e-06, "loss": 0.7483, "step": 15066 }, { "epoch": 1.1807993730407524, "grad_norm": 2.492504835128784, "learning_rate": 8.016949855298629e-06, "loss": 1.191, "step": 15067 }, { "epoch": 1.1808777429467086, "grad_norm": 6.474167346954346, "learning_rate": 8.016703947389906e-06, "loss": 1.5016, "step": 15068 }, { "epoch": 1.1809561128526647, "grad_norm": 2.810659885406494, "learning_rate": 8.016458028007192e-06, "loss": 1.3625, "step": 15069 }, { "epoch": 1.1810344827586208, "grad_norm": 2.5214905738830566, "learning_rate": 8.01621209715142e-06, "loss": 0.9542, "step": 15070 }, { "epoch": 1.1811128526645769, "grad_norm": 2.202355146408081, "learning_rate": 8.015966154823528e-06, "loss": 1.2214, "step": 15071 }, { "epoch": 1.181191222570533, "grad_norm": 7.27257776260376, "learning_rate": 8.015720201024449e-06, "loss": 0.8506, "step": 15072 }, { "epoch": 1.181269592476489, "grad_norm": 2.43666934967041, "learning_rate": 8.015474235755121e-06, "loss": 1.36, "step": 15073 }, { "epoch": 1.1813479623824452, "grad_norm": 4.1314239501953125, "learning_rate": 8.015228259016476e-06, "loss": 2.0749, "step": 15074 }, { "epoch": 1.1814263322884013, "grad_norm": 2.552732229232788, "learning_rate": 8.01498227080945e-06, "loss": 1.5791, "step": 15075 }, { "epoch": 1.1815047021943574, "grad_norm": 2.8269236087799072, "learning_rate": 8.014736271134984e-06, "loss": 1.1691, "step": 15076 }, { "epoch": 1.1815830721003135, "grad_norm": 4.178977012634277, "learning_rate": 8.014490259994005e-06, "loss": 1.2596, "step": 15077 }, { "epoch": 1.1816614420062697, "grad_norm": 3.57885479927063, "learning_rate": 8.014244237387457e-06, "loss": 1.4301, "step": 15078 }, { "epoch": 1.1817398119122258, "grad_norm": 3.92459774017334, "learning_rate": 8.01399820331627e-06, "loss": 0.8582, "step": 15079 }, { "epoch": 1.1818181818181819, "grad_norm": 3.0298914909362793, "learning_rate": 8.013752157781386e-06, "loss": 1.3992, "step": 15080 }, { "epoch": 1.181896551724138, "grad_norm": 4.084907531738281, "learning_rate": 8.013506100783733e-06, "loss": 1.2265, "step": 15081 }, { "epoch": 1.181974921630094, "grad_norm": 3.1662111282348633, "learning_rate": 8.01326003232425e-06, "loss": 1.2802, "step": 15082 }, { "epoch": 1.1820532915360502, "grad_norm": 2.502436399459839, "learning_rate": 8.013013952403876e-06, "loss": 1.7052, "step": 15083 }, { "epoch": 1.1821316614420063, "grad_norm": 3.042760133743286, "learning_rate": 8.012767861023543e-06, "loss": 1.3709, "step": 15084 }, { "epoch": 1.1822100313479624, "grad_norm": 5.1977386474609375, "learning_rate": 8.01252175818419e-06, "loss": 0.9863, "step": 15085 }, { "epoch": 1.1822884012539185, "grad_norm": 3.085094928741455, "learning_rate": 8.01227564388675e-06, "loss": 1.7125, "step": 15086 }, { "epoch": 1.1823667711598747, "grad_norm": 2.352832317352295, "learning_rate": 8.012029518132161e-06, "loss": 1.4294, "step": 15087 }, { "epoch": 1.1824451410658308, "grad_norm": 2.47587513923645, "learning_rate": 8.011783380921359e-06, "loss": 1.0191, "step": 15088 }, { "epoch": 1.1825235109717869, "grad_norm": 2.1908681392669678, "learning_rate": 8.01153723225528e-06, "loss": 1.6155, "step": 15089 }, { "epoch": 1.182601880877743, "grad_norm": 2.6593408584594727, "learning_rate": 8.011291072134858e-06, "loss": 1.5402, "step": 15090 }, { "epoch": 1.182680250783699, "grad_norm": 5.893653392791748, "learning_rate": 8.011044900561036e-06, "loss": 1.4564, "step": 15091 }, { "epoch": 1.1827586206896552, "grad_norm": 3.415478467941284, "learning_rate": 8.010798717534743e-06, "loss": 1.2852, "step": 15092 }, { "epoch": 1.1828369905956113, "grad_norm": 2.3612053394317627, "learning_rate": 8.010552523056919e-06, "loss": 1.4287, "step": 15093 }, { "epoch": 1.1829153605015674, "grad_norm": 2.500272274017334, "learning_rate": 8.010306317128498e-06, "loss": 1.3113, "step": 15094 }, { "epoch": 1.1829937304075235, "grad_norm": 2.2666256427764893, "learning_rate": 8.010060099750419e-06, "loss": 1.1235, "step": 15095 }, { "epoch": 1.1830721003134796, "grad_norm": 3.5331549644470215, "learning_rate": 8.009813870923618e-06, "loss": 1.2509, "step": 15096 }, { "epoch": 1.1831504702194358, "grad_norm": 3.131065845489502, "learning_rate": 8.009567630649028e-06, "loss": 0.813, "step": 15097 }, { "epoch": 1.1832288401253919, "grad_norm": 3.509141206741333, "learning_rate": 8.009321378927593e-06, "loss": 1.4985, "step": 15098 }, { "epoch": 1.183307210031348, "grad_norm": 7.771664142608643, "learning_rate": 8.009075115760243e-06, "loss": 1.4401, "step": 15099 }, { "epoch": 1.183385579937304, "grad_norm": 2.262347459793091, "learning_rate": 8.008828841147915e-06, "loss": 1.021, "step": 15100 }, { "epoch": 1.1834639498432602, "grad_norm": 3.0410103797912598, "learning_rate": 8.00858255509155e-06, "loss": 1.3972, "step": 15101 }, { "epoch": 1.1835423197492163, "grad_norm": 3.103609085083008, "learning_rate": 8.008336257592082e-06, "loss": 1.0737, "step": 15102 }, { "epoch": 1.1836206896551724, "grad_norm": 3.0538556575775146, "learning_rate": 8.008089948650448e-06, "loss": 1.2331, "step": 15103 }, { "epoch": 1.1836990595611285, "grad_norm": 4.015758991241455, "learning_rate": 8.007843628267584e-06, "loss": 1.5008, "step": 15104 }, { "epoch": 1.1837774294670846, "grad_norm": 3.616393804550171, "learning_rate": 8.00759729644443e-06, "loss": 1.5231, "step": 15105 }, { "epoch": 1.1838557993730408, "grad_norm": 2.7081098556518555, "learning_rate": 8.007350953181919e-06, "loss": 1.6078, "step": 15106 }, { "epoch": 1.1839341692789969, "grad_norm": 2.3259313106536865, "learning_rate": 8.007104598480988e-06, "loss": 1.1912, "step": 15107 }, { "epoch": 1.184012539184953, "grad_norm": 2.6653454303741455, "learning_rate": 8.006858232342579e-06, "loss": 1.4414, "step": 15108 }, { "epoch": 1.184090909090909, "grad_norm": 6.627895355224609, "learning_rate": 8.006611854767622e-06, "loss": 0.8685, "step": 15109 }, { "epoch": 1.1841692789968652, "grad_norm": 3.9046425819396973, "learning_rate": 8.00636546575706e-06, "loss": 1.0425, "step": 15110 }, { "epoch": 1.1842476489028213, "grad_norm": 2.6575121879577637, "learning_rate": 8.006119065311829e-06, "loss": 1.3609, "step": 15111 }, { "epoch": 1.1843260188087774, "grad_norm": 6.348826885223389, "learning_rate": 8.005872653432863e-06, "loss": 1.7282, "step": 15112 }, { "epoch": 1.1844043887147335, "grad_norm": 2.988104820251465, "learning_rate": 8.005626230121103e-06, "loss": 1.0582, "step": 15113 }, { "epoch": 1.1844827586206896, "grad_norm": 2.62283992767334, "learning_rate": 8.005379795377484e-06, "loss": 1.249, "step": 15114 }, { "epoch": 1.1845611285266457, "grad_norm": 2.4612724781036377, "learning_rate": 8.005133349202944e-06, "loss": 1.0, "step": 15115 }, { "epoch": 1.1846394984326019, "grad_norm": 7.237183570861816, "learning_rate": 8.004886891598419e-06, "loss": 1.7178, "step": 15116 }, { "epoch": 1.184717868338558, "grad_norm": 3.7974045276641846, "learning_rate": 8.004640422564848e-06, "loss": 1.5237, "step": 15117 }, { "epoch": 1.184796238244514, "grad_norm": 2.4553494453430176, "learning_rate": 8.00439394210317e-06, "loss": 1.4426, "step": 15118 }, { "epoch": 1.1848746081504702, "grad_norm": 2.651886463165283, "learning_rate": 8.004147450214319e-06, "loss": 1.1303, "step": 15119 }, { "epoch": 1.1849529780564263, "grad_norm": 2.7295899391174316, "learning_rate": 8.003900946899234e-06, "loss": 1.7278, "step": 15120 }, { "epoch": 1.1850313479623824, "grad_norm": 2.429314374923706, "learning_rate": 8.003654432158854e-06, "loss": 1.1318, "step": 15121 }, { "epoch": 1.1851097178683385, "grad_norm": 3.4210455417633057, "learning_rate": 8.003407905994114e-06, "loss": 1.156, "step": 15122 }, { "epoch": 1.1851880877742946, "grad_norm": 2.18048095703125, "learning_rate": 8.003161368405954e-06, "loss": 1.3095, "step": 15123 }, { "epoch": 1.1852664576802507, "grad_norm": 6.980059623718262, "learning_rate": 8.002914819395312e-06, "loss": 1.3051, "step": 15124 }, { "epoch": 1.1853448275862069, "grad_norm": 2.9327051639556885, "learning_rate": 8.00266825896312e-06, "loss": 1.2155, "step": 15125 }, { "epoch": 1.185423197492163, "grad_norm": 2.853090763092041, "learning_rate": 8.002421687110324e-06, "loss": 1.2621, "step": 15126 }, { "epoch": 1.185501567398119, "grad_norm": 2.6232738494873047, "learning_rate": 8.002175103837856e-06, "loss": 1.3581, "step": 15127 }, { "epoch": 1.1855799373040752, "grad_norm": 2.3566668033599854, "learning_rate": 8.001928509146657e-06, "loss": 1.6478, "step": 15128 }, { "epoch": 1.1856583072100313, "grad_norm": 5.57275390625, "learning_rate": 8.001681903037664e-06, "loss": 1.7035, "step": 15129 }, { "epoch": 1.1857366771159874, "grad_norm": 4.182372093200684, "learning_rate": 8.001435285511815e-06, "loss": 1.632, "step": 15130 }, { "epoch": 1.1858150470219435, "grad_norm": 3.472350597381592, "learning_rate": 8.001188656570046e-06, "loss": 1.3806, "step": 15131 }, { "epoch": 1.1858934169278996, "grad_norm": 4.210264682769775, "learning_rate": 8.0009420162133e-06, "loss": 1.2401, "step": 15132 }, { "epoch": 1.1859717868338557, "grad_norm": 2.771054983139038, "learning_rate": 8.00069536444251e-06, "loss": 1.9195, "step": 15133 }, { "epoch": 1.1860501567398118, "grad_norm": 2.235746145248413, "learning_rate": 8.000448701258614e-06, "loss": 1.3136, "step": 15134 }, { "epoch": 1.186128526645768, "grad_norm": 3.855473518371582, "learning_rate": 8.000202026662556e-06, "loss": 1.4026, "step": 15135 }, { "epoch": 1.186206896551724, "grad_norm": 2.328632116317749, "learning_rate": 7.999955340655267e-06, "loss": 0.9661, "step": 15136 }, { "epoch": 1.1862852664576802, "grad_norm": 2.2699246406555176, "learning_rate": 7.999708643237692e-06, "loss": 1.2363, "step": 15137 }, { "epoch": 1.1863636363636363, "grad_norm": 2.519256591796875, "learning_rate": 7.999461934410766e-06, "loss": 1.176, "step": 15138 }, { "epoch": 1.1864420062695924, "grad_norm": 2.008873462677002, "learning_rate": 7.999215214175426e-06, "loss": 1.421, "step": 15139 }, { "epoch": 1.1865203761755485, "grad_norm": 4.079747676849365, "learning_rate": 7.998968482532612e-06, "loss": 1.5152, "step": 15140 }, { "epoch": 1.1865987460815046, "grad_norm": 2.7778232097625732, "learning_rate": 7.998721739483262e-06, "loss": 1.2912, "step": 15141 }, { "epoch": 1.1866771159874607, "grad_norm": 2.683175563812256, "learning_rate": 7.998474985028315e-06, "loss": 1.1606, "step": 15142 }, { "epoch": 1.1867554858934168, "grad_norm": 2.4535329341888428, "learning_rate": 7.998228219168709e-06, "loss": 1.6356, "step": 15143 }, { "epoch": 1.186833855799373, "grad_norm": 4.111512184143066, "learning_rate": 7.997981441905383e-06, "loss": 2.0062, "step": 15144 }, { "epoch": 1.186912225705329, "grad_norm": 2.969707727432251, "learning_rate": 7.997734653239276e-06, "loss": 1.3365, "step": 15145 }, { "epoch": 1.1869905956112852, "grad_norm": 2.4006693363189697, "learning_rate": 7.997487853171324e-06, "loss": 1.1776, "step": 15146 }, { "epoch": 1.1870689655172413, "grad_norm": 2.5148563385009766, "learning_rate": 7.99724104170247e-06, "loss": 1.2685, "step": 15147 }, { "epoch": 1.1871473354231976, "grad_norm": 2.4865124225616455, "learning_rate": 7.99699421883365e-06, "loss": 1.3675, "step": 15148 }, { "epoch": 1.1872257053291535, "grad_norm": 3.0612916946411133, "learning_rate": 7.996747384565804e-06, "loss": 1.5098, "step": 15149 }, { "epoch": 1.1873040752351098, "grad_norm": 2.119309186935425, "learning_rate": 7.99650053889987e-06, "loss": 1.2008, "step": 15150 }, { "epoch": 1.1873824451410657, "grad_norm": 3.007605791091919, "learning_rate": 7.996253681836786e-06, "loss": 1.2927, "step": 15151 }, { "epoch": 1.187460815047022, "grad_norm": 3.5446314811706543, "learning_rate": 7.996006813377493e-06, "loss": 1.4311, "step": 15152 }, { "epoch": 1.187539184952978, "grad_norm": 3.5176331996917725, "learning_rate": 7.99575993352293e-06, "loss": 1.2099, "step": 15153 }, { "epoch": 1.1876175548589343, "grad_norm": 3.630462646484375, "learning_rate": 7.995513042274031e-06, "loss": 1.3721, "step": 15154 }, { "epoch": 1.1876959247648902, "grad_norm": 8.276093482971191, "learning_rate": 7.995266139631741e-06, "loss": 1.2839, "step": 15155 }, { "epoch": 1.1877742946708465, "grad_norm": 2.3519725799560547, "learning_rate": 7.995019225596998e-06, "loss": 1.2547, "step": 15156 }, { "epoch": 1.1878526645768024, "grad_norm": 2.654627561569214, "learning_rate": 7.994772300170738e-06, "loss": 1.4972, "step": 15157 }, { "epoch": 1.1879310344827587, "grad_norm": 2.3996872901916504, "learning_rate": 7.994525363353907e-06, "loss": 1.2664, "step": 15158 }, { "epoch": 1.1880094043887148, "grad_norm": 3.5915777683258057, "learning_rate": 7.994278415147435e-06, "loss": 1.4874, "step": 15159 }, { "epoch": 1.188087774294671, "grad_norm": 3.3278443813323975, "learning_rate": 7.994031455552267e-06, "loss": 1.4417, "step": 15160 }, { "epoch": 1.188166144200627, "grad_norm": 2.5762743949890137, "learning_rate": 7.993784484569342e-06, "loss": 1.3515, "step": 15161 }, { "epoch": 1.1882445141065832, "grad_norm": 2.388167142868042, "learning_rate": 7.9935375021996e-06, "loss": 1.6312, "step": 15162 }, { "epoch": 1.1883228840125393, "grad_norm": 2.7028896808624268, "learning_rate": 7.993290508443976e-06, "loss": 1.2648, "step": 15163 }, { "epoch": 1.1884012539184954, "grad_norm": 2.7690324783325195, "learning_rate": 7.993043503303413e-06, "loss": 1.3503, "step": 15164 }, { "epoch": 1.1884796238244515, "grad_norm": 4.77494478225708, "learning_rate": 7.99279648677885e-06, "loss": 1.3056, "step": 15165 }, { "epoch": 1.1885579937304076, "grad_norm": 5.8099470138549805, "learning_rate": 7.992549458871228e-06, "loss": 1.7665, "step": 15166 }, { "epoch": 1.1886363636363637, "grad_norm": 3.1651740074157715, "learning_rate": 7.992302419581484e-06, "loss": 1.576, "step": 15167 }, { "epoch": 1.1887147335423198, "grad_norm": 5.3523478507995605, "learning_rate": 7.992055368910558e-06, "loss": 1.1479, "step": 15168 }, { "epoch": 1.188793103448276, "grad_norm": 3.1125495433807373, "learning_rate": 7.99180830685939e-06, "loss": 1.0019, "step": 15169 }, { "epoch": 1.188871473354232, "grad_norm": 3.400393486022949, "learning_rate": 7.991561233428919e-06, "loss": 1.3997, "step": 15170 }, { "epoch": 1.1889498432601882, "grad_norm": 2.6404850482940674, "learning_rate": 7.991314148620088e-06, "loss": 1.6894, "step": 15171 }, { "epoch": 1.1890282131661443, "grad_norm": 3.162799119949341, "learning_rate": 7.991067052433833e-06, "loss": 1.5507, "step": 15172 }, { "epoch": 1.1891065830721004, "grad_norm": 3.4956297874450684, "learning_rate": 7.990819944871096e-06, "loss": 1.4, "step": 15173 }, { "epoch": 1.1891849529780565, "grad_norm": 2.528989315032959, "learning_rate": 7.990572825932816e-06, "loss": 1.2326, "step": 15174 }, { "epoch": 1.1892633228840126, "grad_norm": 5.722925662994385, "learning_rate": 7.990325695619933e-06, "loss": 1.3101, "step": 15175 }, { "epoch": 1.1893416927899687, "grad_norm": 2.8862648010253906, "learning_rate": 7.990078553933386e-06, "loss": 1.3797, "step": 15176 }, { "epoch": 1.1894200626959248, "grad_norm": 3.635026454925537, "learning_rate": 7.989831400874117e-06, "loss": 0.9557, "step": 15177 }, { "epoch": 1.189498432601881, "grad_norm": 2.1757447719573975, "learning_rate": 7.989584236443064e-06, "loss": 1.6483, "step": 15178 }, { "epoch": 1.189576802507837, "grad_norm": 1.8923362493515015, "learning_rate": 7.989337060641168e-06, "loss": 1.0519, "step": 15179 }, { "epoch": 1.1896551724137931, "grad_norm": 3.8485076427459717, "learning_rate": 7.989089873469372e-06, "loss": 1.4552, "step": 15180 }, { "epoch": 1.1897335423197493, "grad_norm": 2.779792547225952, "learning_rate": 7.98884267492861e-06, "loss": 1.6613, "step": 15181 }, { "epoch": 1.1898119122257054, "grad_norm": 3.7622742652893066, "learning_rate": 7.988595465019829e-06, "loss": 1.3487, "step": 15182 }, { "epoch": 1.1898902821316615, "grad_norm": 2.05206561088562, "learning_rate": 7.988348243743963e-06, "loss": 1.3612, "step": 15183 }, { "epoch": 1.1899686520376176, "grad_norm": 3.2811717987060547, "learning_rate": 7.988101011101956e-06, "loss": 1.582, "step": 15184 }, { "epoch": 1.1900470219435737, "grad_norm": 2.548516035079956, "learning_rate": 7.987853767094748e-06, "loss": 1.3828, "step": 15185 }, { "epoch": 1.1901253918495298, "grad_norm": 2.705620527267456, "learning_rate": 7.987606511723279e-06, "loss": 1.5089, "step": 15186 }, { "epoch": 1.190203761755486, "grad_norm": 2.62900447845459, "learning_rate": 7.987359244988488e-06, "loss": 1.0622, "step": 15187 }, { "epoch": 1.190282131661442, "grad_norm": 4.871829032897949, "learning_rate": 7.987111966891318e-06, "loss": 1.1418, "step": 15188 }, { "epoch": 1.1903605015673981, "grad_norm": 2.8916285037994385, "learning_rate": 7.986864677432709e-06, "loss": 1.448, "step": 15189 }, { "epoch": 1.1904388714733543, "grad_norm": 2.025484085083008, "learning_rate": 7.986617376613599e-06, "loss": 1.3992, "step": 15190 }, { "epoch": 1.1905172413793104, "grad_norm": 4.1500749588012695, "learning_rate": 7.986370064434931e-06, "loss": 1.2224, "step": 15191 }, { "epoch": 1.1905956112852665, "grad_norm": 3.9166042804718018, "learning_rate": 7.986122740897647e-06, "loss": 1.4366, "step": 15192 }, { "epoch": 1.1906739811912226, "grad_norm": 3.6866934299468994, "learning_rate": 7.985875406002686e-06, "loss": 1.7271, "step": 15193 }, { "epoch": 1.1907523510971787, "grad_norm": 3.3035295009613037, "learning_rate": 7.985628059750987e-06, "loss": 1.3595, "step": 15194 }, { "epoch": 1.1908307210031348, "grad_norm": 2.761276960372925, "learning_rate": 7.985380702143493e-06, "loss": 1.5532, "step": 15195 }, { "epoch": 1.190909090909091, "grad_norm": 2.982178211212158, "learning_rate": 7.985133333181143e-06, "loss": 1.4916, "step": 15196 }, { "epoch": 1.190987460815047, "grad_norm": 3.144644021987915, "learning_rate": 7.98488595286488e-06, "loss": 1.6851, "step": 15197 }, { "epoch": 1.1910658307210031, "grad_norm": 2.4948477745056152, "learning_rate": 7.984638561195644e-06, "loss": 1.5308, "step": 15198 }, { "epoch": 1.1911442006269592, "grad_norm": 2.2427661418914795, "learning_rate": 7.984391158174375e-06, "loss": 1.3512, "step": 15199 }, { "epoch": 1.1912225705329154, "grad_norm": 3.4530749320983887, "learning_rate": 7.984143743802016e-06, "loss": 1.4495, "step": 15200 }, { "epoch": 1.1913009404388715, "grad_norm": 2.75836181640625, "learning_rate": 7.983896318079507e-06, "loss": 1.5883, "step": 15201 }, { "epoch": 1.1913793103448276, "grad_norm": 2.258399248123169, "learning_rate": 7.983648881007788e-06, "loss": 1.3955, "step": 15202 }, { "epoch": 1.1914576802507837, "grad_norm": 15.017009735107422, "learning_rate": 7.983401432587803e-06, "loss": 2.0435, "step": 15203 }, { "epoch": 1.1915360501567398, "grad_norm": 2.7954981327056885, "learning_rate": 7.983153972820489e-06, "loss": 1.2793, "step": 15204 }, { "epoch": 1.191614420062696, "grad_norm": 2.523639678955078, "learning_rate": 7.98290650170679e-06, "loss": 1.4199, "step": 15205 }, { "epoch": 1.191692789968652, "grad_norm": 2.6658966541290283, "learning_rate": 7.982659019247648e-06, "loss": 1.3125, "step": 15206 }, { "epoch": 1.1917711598746081, "grad_norm": 3.1667370796203613, "learning_rate": 7.982411525444002e-06, "loss": 1.0521, "step": 15207 }, { "epoch": 1.1918495297805642, "grad_norm": 3.0163464546203613, "learning_rate": 7.982164020296796e-06, "loss": 1.5553, "step": 15208 }, { "epoch": 1.1919278996865204, "grad_norm": 2.7326271533966064, "learning_rate": 7.981916503806968e-06, "loss": 1.3735, "step": 15209 }, { "epoch": 1.1920062695924765, "grad_norm": 2.1635189056396484, "learning_rate": 7.981668975975463e-06, "loss": 1.3902, "step": 15210 }, { "epoch": 1.1920846394984326, "grad_norm": 3.1642637252807617, "learning_rate": 7.981421436803219e-06, "loss": 1.4128, "step": 15211 }, { "epoch": 1.1921630094043887, "grad_norm": 3.174403429031372, "learning_rate": 7.98117388629118e-06, "loss": 1.4267, "step": 15212 }, { "epoch": 1.1922413793103448, "grad_norm": 3.146480083465576, "learning_rate": 7.980926324440286e-06, "loss": 1.4566, "step": 15213 }, { "epoch": 1.192319749216301, "grad_norm": 3.303295850753784, "learning_rate": 7.980678751251481e-06, "loss": 1.4415, "step": 15214 }, { "epoch": 1.192398119122257, "grad_norm": 3.3572864532470703, "learning_rate": 7.980431166725704e-06, "loss": 1.5, "step": 15215 }, { "epoch": 1.1924764890282131, "grad_norm": 2.5376715660095215, "learning_rate": 7.980183570863898e-06, "loss": 1.4161, "step": 15216 }, { "epoch": 1.1925548589341692, "grad_norm": 3.296294689178467, "learning_rate": 7.979935963667004e-06, "loss": 1.5358, "step": 15217 }, { "epoch": 1.1926332288401253, "grad_norm": 2.084974765777588, "learning_rate": 7.979688345135964e-06, "loss": 1.4835, "step": 15218 }, { "epoch": 1.1927115987460815, "grad_norm": 3.3516805171966553, "learning_rate": 7.97944071527172e-06, "loss": 1.6876, "step": 15219 }, { "epoch": 1.1927899686520376, "grad_norm": 2.646353244781494, "learning_rate": 7.979193074075216e-06, "loss": 1.4197, "step": 15220 }, { "epoch": 1.1928683385579937, "grad_norm": 3.3288681507110596, "learning_rate": 7.978945421547389e-06, "loss": 1.6672, "step": 15221 }, { "epoch": 1.1929467084639498, "grad_norm": 4.24831485748291, "learning_rate": 7.978697757689185e-06, "loss": 1.502, "step": 15222 }, { "epoch": 1.193025078369906, "grad_norm": 3.5034544467926025, "learning_rate": 7.978450082501544e-06, "loss": 1.817, "step": 15223 }, { "epoch": 1.193103448275862, "grad_norm": 1.74493408203125, "learning_rate": 7.978202395985409e-06, "loss": 1.3645, "step": 15224 }, { "epoch": 1.1931818181818181, "grad_norm": 2.7405333518981934, "learning_rate": 7.977954698141722e-06, "loss": 1.5285, "step": 15225 }, { "epoch": 1.1932601880877742, "grad_norm": 4.526121616363525, "learning_rate": 7.977706988971425e-06, "loss": 0.9764, "step": 15226 }, { "epoch": 1.1933385579937303, "grad_norm": 4.3058390617370605, "learning_rate": 7.977459268475459e-06, "loss": 1.7204, "step": 15227 }, { "epoch": 1.1934169278996865, "grad_norm": 2.4690098762512207, "learning_rate": 7.977211536654768e-06, "loss": 1.2431, "step": 15228 }, { "epoch": 1.1934952978056426, "grad_norm": 4.777683258056641, "learning_rate": 7.976963793510294e-06, "loss": 1.1056, "step": 15229 }, { "epoch": 1.1935736677115987, "grad_norm": 2.668998956680298, "learning_rate": 7.976716039042977e-06, "loss": 1.2689, "step": 15230 }, { "epoch": 1.1936520376175548, "grad_norm": 3.9201841354370117, "learning_rate": 7.976468273253764e-06, "loss": 1.0547, "step": 15231 }, { "epoch": 1.193730407523511, "grad_norm": 2.252608060836792, "learning_rate": 7.976220496143592e-06, "loss": 1.4461, "step": 15232 }, { "epoch": 1.193808777429467, "grad_norm": 3.2576112747192383, "learning_rate": 7.975972707713408e-06, "loss": 1.7084, "step": 15233 }, { "epoch": 1.1938871473354231, "grad_norm": 2.4531009197235107, "learning_rate": 7.97572490796415e-06, "loss": 0.8818, "step": 15234 }, { "epoch": 1.1939655172413792, "grad_norm": 2.7706451416015625, "learning_rate": 7.975477096896765e-06, "loss": 1.5489, "step": 15235 }, { "epoch": 1.1940438871473353, "grad_norm": 2.8754959106445312, "learning_rate": 7.975229274512193e-06, "loss": 1.5366, "step": 15236 }, { "epoch": 1.1941222570532914, "grad_norm": 2.3270838260650635, "learning_rate": 7.974981440811376e-06, "loss": 1.3403, "step": 15237 }, { "epoch": 1.1942006269592476, "grad_norm": 2.477560520172119, "learning_rate": 7.974733595795259e-06, "loss": 1.5338, "step": 15238 }, { "epoch": 1.1942789968652037, "grad_norm": 3.1220715045928955, "learning_rate": 7.974485739464781e-06, "loss": 1.6272, "step": 15239 }, { "epoch": 1.1943573667711598, "grad_norm": 2.1949028968811035, "learning_rate": 7.97423787182089e-06, "loss": 1.1698, "step": 15240 }, { "epoch": 1.1944357366771161, "grad_norm": 4.241308689117432, "learning_rate": 7.973989992864525e-06, "loss": 1.2745, "step": 15241 }, { "epoch": 1.194514106583072, "grad_norm": 6.443285942077637, "learning_rate": 7.973742102596628e-06, "loss": 1.7413, "step": 15242 }, { "epoch": 1.1945924764890283, "grad_norm": 3.3207616806030273, "learning_rate": 7.973494201018143e-06, "loss": 1.3016, "step": 15243 }, { "epoch": 1.1946708463949842, "grad_norm": 3.700485944747925, "learning_rate": 7.973246288130017e-06, "loss": 1.303, "step": 15244 }, { "epoch": 1.1947492163009406, "grad_norm": 3.1804518699645996, "learning_rate": 7.972998363933187e-06, "loss": 1.4512, "step": 15245 }, { "epoch": 1.1948275862068964, "grad_norm": 2.856534719467163, "learning_rate": 7.972750428428596e-06, "loss": 1.2201, "step": 15246 }, { "epoch": 1.1949059561128528, "grad_norm": 5.964324474334717, "learning_rate": 7.972502481617193e-06, "loss": 1.1031, "step": 15247 }, { "epoch": 1.1949843260188087, "grad_norm": 3.817836284637451, "learning_rate": 7.972254523499917e-06, "loss": 1.2455, "step": 15248 }, { "epoch": 1.195062695924765, "grad_norm": 3.806013584136963, "learning_rate": 7.972006554077711e-06, "loss": 2.0516, "step": 15249 }, { "epoch": 1.1951410658307209, "grad_norm": 2.853929281234741, "learning_rate": 7.971758573351517e-06, "loss": 0.84, "step": 15250 }, { "epoch": 1.1952194357366772, "grad_norm": 2.730647325515747, "learning_rate": 7.971510581322282e-06, "loss": 0.8354, "step": 15251 }, { "epoch": 1.195297805642633, "grad_norm": 3.4196321964263916, "learning_rate": 7.971262577990943e-06, "loss": 0.9315, "step": 15252 }, { "epoch": 1.1953761755485894, "grad_norm": 2.0433290004730225, "learning_rate": 7.971014563358452e-06, "loss": 1.2526, "step": 15253 }, { "epoch": 1.1954545454545455, "grad_norm": 2.48915433883667, "learning_rate": 7.970766537425744e-06, "loss": 1.4198, "step": 15254 }, { "epoch": 1.1955329153605017, "grad_norm": 3.629119396209717, "learning_rate": 7.970518500193769e-06, "loss": 1.2898, "step": 15255 }, { "epoch": 1.1956112852664578, "grad_norm": 2.4218780994415283, "learning_rate": 7.970270451663465e-06, "loss": 1.2292, "step": 15256 }, { "epoch": 1.1956896551724139, "grad_norm": 2.7879815101623535, "learning_rate": 7.97002239183578e-06, "loss": 1.7094, "step": 15257 }, { "epoch": 1.19576802507837, "grad_norm": 2.654757499694824, "learning_rate": 7.969774320711655e-06, "loss": 1.0732, "step": 15258 }, { "epoch": 1.195846394984326, "grad_norm": 3.1235415935516357, "learning_rate": 7.969526238292032e-06, "loss": 2.0216, "step": 15259 }, { "epoch": 1.1959247648902822, "grad_norm": 3.2980599403381348, "learning_rate": 7.969278144577856e-06, "loss": 1.3123, "step": 15260 }, { "epoch": 1.1960031347962383, "grad_norm": 8.203692436218262, "learning_rate": 7.969030039570074e-06, "loss": 1.7315, "step": 15261 }, { "epoch": 1.1960815047021944, "grad_norm": 4.071310043334961, "learning_rate": 7.968781923269625e-06, "loss": 1.7705, "step": 15262 }, { "epoch": 1.1961598746081505, "grad_norm": 6.603845596313477, "learning_rate": 7.968533795677455e-06, "loss": 1.1799, "step": 15263 }, { "epoch": 1.1962382445141067, "grad_norm": 2.2193551063537598, "learning_rate": 7.968285656794508e-06, "loss": 1.1407, "step": 15264 }, { "epoch": 1.1963166144200628, "grad_norm": 3.176166296005249, "learning_rate": 7.968037506621724e-06, "loss": 0.9703, "step": 15265 }, { "epoch": 1.1963949843260189, "grad_norm": 2.5254580974578857, "learning_rate": 7.967789345160053e-06, "loss": 0.9857, "step": 15266 }, { "epoch": 1.196473354231975, "grad_norm": 3.168081760406494, "learning_rate": 7.967541172410434e-06, "loss": 1.0965, "step": 15267 }, { "epoch": 1.196551724137931, "grad_norm": 2.2328009605407715, "learning_rate": 7.967292988373812e-06, "loss": 1.6237, "step": 15268 }, { "epoch": 1.1966300940438872, "grad_norm": 3.8758444786071777, "learning_rate": 7.967044793051133e-06, "loss": 1.9224, "step": 15269 }, { "epoch": 1.1967084639498433, "grad_norm": 28.701059341430664, "learning_rate": 7.96679658644334e-06, "loss": 1.3685, "step": 15270 }, { "epoch": 1.1967868338557994, "grad_norm": 3.3435986042022705, "learning_rate": 7.966548368551374e-06, "loss": 1.7999, "step": 15271 }, { "epoch": 1.1968652037617555, "grad_norm": 2.5948240756988525, "learning_rate": 7.966300139376183e-06, "loss": 1.7209, "step": 15272 }, { "epoch": 1.1969435736677116, "grad_norm": 3.803863286972046, "learning_rate": 7.96605189891871e-06, "loss": 1.8089, "step": 15273 }, { "epoch": 1.1970219435736678, "grad_norm": 2.3402554988861084, "learning_rate": 7.9658036471799e-06, "loss": 1.5476, "step": 15274 }, { "epoch": 1.1971003134796239, "grad_norm": 4.07432222366333, "learning_rate": 7.965555384160696e-06, "loss": 1.2276, "step": 15275 }, { "epoch": 1.19717868338558, "grad_norm": 1.9597361087799072, "learning_rate": 7.965307109862042e-06, "loss": 1.2596, "step": 15276 }, { "epoch": 1.197257053291536, "grad_norm": 1.975192904472351, "learning_rate": 7.965058824284884e-06, "loss": 1.2781, "step": 15277 }, { "epoch": 1.1973354231974922, "grad_norm": 3.7401504516601562, "learning_rate": 7.964810527430162e-06, "loss": 1.5314, "step": 15278 }, { "epoch": 1.1974137931034483, "grad_norm": 6.057153224945068, "learning_rate": 7.964562219298827e-06, "loss": 1.9238, "step": 15279 }, { "epoch": 1.1974921630094044, "grad_norm": 3.1097488403320312, "learning_rate": 7.964313899891818e-06, "loss": 1.6265, "step": 15280 }, { "epoch": 1.1975705329153605, "grad_norm": 3.5045666694641113, "learning_rate": 7.964065569210082e-06, "loss": 1.5741, "step": 15281 }, { "epoch": 1.1976489028213166, "grad_norm": 2.291585683822632, "learning_rate": 7.963817227254564e-06, "loss": 1.3253, "step": 15282 }, { "epoch": 1.1977272727272728, "grad_norm": 2.6548726558685303, "learning_rate": 7.963568874026207e-06, "loss": 1.2105, "step": 15283 }, { "epoch": 1.1978056426332289, "grad_norm": 5.561882972717285, "learning_rate": 7.963320509525956e-06, "loss": 1.3044, "step": 15284 }, { "epoch": 1.197884012539185, "grad_norm": 3.113757610321045, "learning_rate": 7.963072133754756e-06, "loss": 1.295, "step": 15285 }, { "epoch": 1.197962382445141, "grad_norm": 3.578737497329712, "learning_rate": 7.96282374671355e-06, "loss": 1.2509, "step": 15286 }, { "epoch": 1.1980407523510972, "grad_norm": 2.1191859245300293, "learning_rate": 7.962575348403287e-06, "loss": 1.2514, "step": 15287 }, { "epoch": 1.1981191222570533, "grad_norm": 3.1937508583068848, "learning_rate": 7.962326938824908e-06, "loss": 0.8828, "step": 15288 }, { "epoch": 1.1981974921630094, "grad_norm": 7.308216094970703, "learning_rate": 7.962078517979358e-06, "loss": 0.9614, "step": 15289 }, { "epoch": 1.1982758620689655, "grad_norm": 7.092773914337158, "learning_rate": 7.961830085867583e-06, "loss": 1.5914, "step": 15290 }, { "epoch": 1.1983542319749216, "grad_norm": 1.960655689239502, "learning_rate": 7.961581642490528e-06, "loss": 1.1896, "step": 15291 }, { "epoch": 1.1984326018808777, "grad_norm": 5.711411952972412, "learning_rate": 7.961333187849138e-06, "loss": 1.0118, "step": 15292 }, { "epoch": 1.1985109717868339, "grad_norm": 2.753448724746704, "learning_rate": 7.961084721944357e-06, "loss": 1.3772, "step": 15293 }, { "epoch": 1.19858934169279, "grad_norm": 5.288435935974121, "learning_rate": 7.96083624477713e-06, "loss": 1.4445, "step": 15294 }, { "epoch": 1.198667711598746, "grad_norm": 3.672154664993286, "learning_rate": 7.960587756348403e-06, "loss": 1.1038, "step": 15295 }, { "epoch": 1.1987460815047022, "grad_norm": 2.3855319023132324, "learning_rate": 7.960339256659122e-06, "loss": 1.56, "step": 15296 }, { "epoch": 1.1988244514106583, "grad_norm": 5.229935169219971, "learning_rate": 7.960090745710229e-06, "loss": 1.2895, "step": 15297 }, { "epoch": 1.1989028213166144, "grad_norm": 4.904116630554199, "learning_rate": 7.959842223502673e-06, "loss": 1.1601, "step": 15298 }, { "epoch": 1.1989811912225705, "grad_norm": 2.6191325187683105, "learning_rate": 7.959593690037398e-06, "loss": 1.5021, "step": 15299 }, { "epoch": 1.1990595611285266, "grad_norm": 6.468853950500488, "learning_rate": 7.959345145315348e-06, "loss": 1.3559, "step": 15300 }, { "epoch": 1.1991379310344827, "grad_norm": 2.2442522048950195, "learning_rate": 7.959096589337466e-06, "loss": 1.4036, "step": 15301 }, { "epoch": 1.1992163009404389, "grad_norm": 2.499638080596924, "learning_rate": 7.958848022104704e-06, "loss": 1.5214, "step": 15302 }, { "epoch": 1.199294670846395, "grad_norm": 1.997179388999939, "learning_rate": 7.958599443618003e-06, "loss": 1.3467, "step": 15303 }, { "epoch": 1.199373040752351, "grad_norm": 2.328406572341919, "learning_rate": 7.95835085387831e-06, "loss": 1.4667, "step": 15304 }, { "epoch": 1.1994514106583072, "grad_norm": 4.95890474319458, "learning_rate": 7.95810225288657e-06, "loss": 0.7648, "step": 15305 }, { "epoch": 1.1995297805642633, "grad_norm": 2.513950824737549, "learning_rate": 7.957853640643727e-06, "loss": 1.4636, "step": 15306 }, { "epoch": 1.1996081504702194, "grad_norm": 3.395246982574463, "learning_rate": 7.957605017150728e-06, "loss": 1.4802, "step": 15307 }, { "epoch": 1.1996865203761755, "grad_norm": 7.397528171539307, "learning_rate": 7.95735638240852e-06, "loss": 2.3292, "step": 15308 }, { "epoch": 1.1997648902821316, "grad_norm": 2.6849329471588135, "learning_rate": 7.957107736418047e-06, "loss": 1.229, "step": 15309 }, { "epoch": 1.1998432601880877, "grad_norm": 4.580414295196533, "learning_rate": 7.956859079180255e-06, "loss": 1.312, "step": 15310 }, { "epoch": 1.1999216300940438, "grad_norm": 3.1960039138793945, "learning_rate": 7.95661041069609e-06, "loss": 1.711, "step": 15311 }, { "epoch": 1.2, "grad_norm": 4.7277045249938965, "learning_rate": 7.956361730966495e-06, "loss": 1.7172, "step": 15312 }, { "epoch": 1.200078369905956, "grad_norm": 7.378077030181885, "learning_rate": 7.95611303999242e-06, "loss": 0.745, "step": 15313 }, { "epoch": 1.2001567398119122, "grad_norm": 3.8792002201080322, "learning_rate": 7.95586433777481e-06, "loss": 1.2293, "step": 15314 }, { "epoch": 1.2002351097178683, "grad_norm": 2.851825475692749, "learning_rate": 7.955615624314612e-06, "loss": 1.1625, "step": 15315 }, { "epoch": 1.2003134796238244, "grad_norm": 3.717266082763672, "learning_rate": 7.955366899612769e-06, "loss": 1.4613, "step": 15316 }, { "epoch": 1.2003918495297805, "grad_norm": 3.269986391067505, "learning_rate": 7.955118163670229e-06, "loss": 1.4193, "step": 15317 }, { "epoch": 1.2004702194357366, "grad_norm": 3.961113214492798, "learning_rate": 7.954869416487934e-06, "loss": 1.1207, "step": 15318 }, { "epoch": 1.2005485893416927, "grad_norm": 2.0879342555999756, "learning_rate": 7.954620658066835e-06, "loss": 1.2014, "step": 15319 }, { "epoch": 1.2006269592476488, "grad_norm": 2.0671138763427734, "learning_rate": 7.954371888407878e-06, "loss": 1.1987, "step": 15320 }, { "epoch": 1.200705329153605, "grad_norm": 2.445824146270752, "learning_rate": 7.954123107512007e-06, "loss": 1.4777, "step": 15321 }, { "epoch": 1.200783699059561, "grad_norm": 12.71091365814209, "learning_rate": 7.953874315380169e-06, "loss": 1.3789, "step": 15322 }, { "epoch": 1.2008620689655172, "grad_norm": 2.781553268432617, "learning_rate": 7.95362551201331e-06, "loss": 1.3747, "step": 15323 }, { "epoch": 1.2009404388714733, "grad_norm": 3.391655206680298, "learning_rate": 7.953376697412377e-06, "loss": 1.3383, "step": 15324 }, { "epoch": 1.2010188087774294, "grad_norm": 3.945558547973633, "learning_rate": 7.953127871578316e-06, "loss": 1.637, "step": 15325 }, { "epoch": 1.2010971786833855, "grad_norm": 2.407902956008911, "learning_rate": 7.952879034512073e-06, "loss": 1.5907, "step": 15326 }, { "epoch": 1.2011755485893416, "grad_norm": 3.309763193130493, "learning_rate": 7.952630186214596e-06, "loss": 1.7636, "step": 15327 }, { "epoch": 1.2012539184952977, "grad_norm": 2.3634111881256104, "learning_rate": 7.952381326686828e-06, "loss": 1.3833, "step": 15328 }, { "epoch": 1.2013322884012538, "grad_norm": 3.759882926940918, "learning_rate": 7.95213245592972e-06, "loss": 1.239, "step": 15329 }, { "epoch": 1.20141065830721, "grad_norm": 5.498821258544922, "learning_rate": 7.951883573944217e-06, "loss": 1.0069, "step": 15330 }, { "epoch": 1.201489028213166, "grad_norm": 2.366680860519409, "learning_rate": 7.951634680731262e-06, "loss": 1.0449, "step": 15331 }, { "epoch": 1.2015673981191222, "grad_norm": 5.5763654708862305, "learning_rate": 7.951385776291807e-06, "loss": 1.1708, "step": 15332 }, { "epoch": 1.2016457680250783, "grad_norm": 4.194911479949951, "learning_rate": 7.951136860626797e-06, "loss": 1.4754, "step": 15333 }, { "epoch": 1.2017241379310344, "grad_norm": 1.8415223360061646, "learning_rate": 7.950887933737178e-06, "loss": 1.4187, "step": 15334 }, { "epoch": 1.2018025078369905, "grad_norm": 2.163470983505249, "learning_rate": 7.950638995623896e-06, "loss": 1.5505, "step": 15335 }, { "epoch": 1.2018808777429468, "grad_norm": 3.0123391151428223, "learning_rate": 7.950390046287898e-06, "loss": 1.0393, "step": 15336 }, { "epoch": 1.2019592476489027, "grad_norm": 2.806976079940796, "learning_rate": 7.950141085730133e-06, "loss": 1.3287, "step": 15337 }, { "epoch": 1.202037617554859, "grad_norm": 2.3161637783050537, "learning_rate": 7.949892113951545e-06, "loss": 1.3833, "step": 15338 }, { "epoch": 1.202115987460815, "grad_norm": 9.046208381652832, "learning_rate": 7.949643130953084e-06, "loss": 1.0944, "step": 15339 }, { "epoch": 1.2021943573667713, "grad_norm": 2.0400166511535645, "learning_rate": 7.949394136735696e-06, "loss": 1.2283, "step": 15340 }, { "epoch": 1.2022727272727272, "grad_norm": 2.6883857250213623, "learning_rate": 7.949145131300328e-06, "loss": 1.6176, "step": 15341 }, { "epoch": 1.2023510971786835, "grad_norm": 2.37977933883667, "learning_rate": 7.948896114647925e-06, "loss": 1.6352, "step": 15342 }, { "epoch": 1.2024294670846394, "grad_norm": 3.0919768810272217, "learning_rate": 7.948647086779438e-06, "loss": 1.485, "step": 15343 }, { "epoch": 1.2025078369905957, "grad_norm": 2.309906005859375, "learning_rate": 7.948398047695808e-06, "loss": 1.1678, "step": 15344 }, { "epoch": 1.2025862068965516, "grad_norm": 3.276237726211548, "learning_rate": 7.948148997397991e-06, "loss": 1.2759, "step": 15345 }, { "epoch": 1.202664576802508, "grad_norm": 3.770193338394165, "learning_rate": 7.947899935886927e-06, "loss": 1.4929, "step": 15346 }, { "epoch": 1.202742946708464, "grad_norm": 4.117105960845947, "learning_rate": 7.947650863163565e-06, "loss": 1.6557, "step": 15347 }, { "epoch": 1.2028213166144202, "grad_norm": 2.494969367980957, "learning_rate": 7.947401779228853e-06, "loss": 1.4522, "step": 15348 }, { "epoch": 1.2028996865203763, "grad_norm": 2.263446807861328, "learning_rate": 7.947152684083742e-06, "loss": 1.2403, "step": 15349 }, { "epoch": 1.2029780564263324, "grad_norm": 4.214143753051758, "learning_rate": 7.946903577729172e-06, "loss": 2.0492, "step": 15350 }, { "epoch": 1.2030564263322885, "grad_norm": 5.836842060089111, "learning_rate": 7.946654460166097e-06, "loss": 1.2066, "step": 15351 }, { "epoch": 1.2031347962382446, "grad_norm": 3.093445062637329, "learning_rate": 7.946405331395462e-06, "loss": 1.4389, "step": 15352 }, { "epoch": 1.2032131661442007, "grad_norm": 3.750591993331909, "learning_rate": 7.946156191418211e-06, "loss": 1.2477, "step": 15353 }, { "epoch": 1.2032915360501568, "grad_norm": 1.9632985591888428, "learning_rate": 7.945907040235299e-06, "loss": 1.5723, "step": 15354 }, { "epoch": 1.203369905956113, "grad_norm": 5.877727031707764, "learning_rate": 7.945657877847669e-06, "loss": 1.1782, "step": 15355 }, { "epoch": 1.203448275862069, "grad_norm": 3.29814076423645, "learning_rate": 7.945408704256268e-06, "loss": 1.5836, "step": 15356 }, { "epoch": 1.2035266457680251, "grad_norm": 2.7950685024261475, "learning_rate": 7.945159519462045e-06, "loss": 1.337, "step": 15357 }, { "epoch": 1.2036050156739813, "grad_norm": 6.227348327636719, "learning_rate": 7.944910323465948e-06, "loss": 0.88, "step": 15358 }, { "epoch": 1.2036833855799374, "grad_norm": 2.5928525924682617, "learning_rate": 7.944661116268927e-06, "loss": 1.5654, "step": 15359 }, { "epoch": 1.2037617554858935, "grad_norm": 2.0439698696136475, "learning_rate": 7.944411897871927e-06, "loss": 1.438, "step": 15360 }, { "epoch": 1.2038401253918496, "grad_norm": 2.9355146884918213, "learning_rate": 7.944162668275894e-06, "loss": 1.2501, "step": 15361 }, { "epoch": 1.2039184952978057, "grad_norm": 2.506960868835449, "learning_rate": 7.94391342748178e-06, "loss": 1.653, "step": 15362 }, { "epoch": 1.2039968652037618, "grad_norm": 3.0368640422821045, "learning_rate": 7.943664175490532e-06, "loss": 1.4191, "step": 15363 }, { "epoch": 1.204075235109718, "grad_norm": 2.4211645126342773, "learning_rate": 7.943414912303096e-06, "loss": 1.2966, "step": 15364 }, { "epoch": 1.204153605015674, "grad_norm": 3.2398014068603516, "learning_rate": 7.943165637920423e-06, "loss": 1.3927, "step": 15365 }, { "epoch": 1.2042319749216301, "grad_norm": 3.8388216495513916, "learning_rate": 7.94291635234346e-06, "loss": 1.2165, "step": 15366 }, { "epoch": 1.2043103448275863, "grad_norm": 3.4029014110565186, "learning_rate": 7.942667055573151e-06, "loss": 1.1934, "step": 15367 }, { "epoch": 1.2043887147335424, "grad_norm": 2.6596872806549072, "learning_rate": 7.942417747610451e-06, "loss": 1.2567, "step": 15368 }, { "epoch": 1.2044670846394985, "grad_norm": 5.176405906677246, "learning_rate": 7.942168428456306e-06, "loss": 1.1303, "step": 15369 }, { "epoch": 1.2045454545454546, "grad_norm": 3.9503629207611084, "learning_rate": 7.941919098111662e-06, "loss": 1.2838, "step": 15370 }, { "epoch": 1.2046238244514107, "grad_norm": 3.3438358306884766, "learning_rate": 7.94166975657747e-06, "loss": 1.5472, "step": 15371 }, { "epoch": 1.2047021943573668, "grad_norm": 4.2717390060424805, "learning_rate": 7.941420403854677e-06, "loss": 1.491, "step": 15372 }, { "epoch": 1.204780564263323, "grad_norm": 9.244856834411621, "learning_rate": 7.941171039944229e-06, "loss": 1.2833, "step": 15373 }, { "epoch": 1.204858934169279, "grad_norm": 3.0586321353912354, "learning_rate": 7.940921664847082e-06, "loss": 1.2793, "step": 15374 }, { "epoch": 1.2049373040752351, "grad_norm": 3.396214485168457, "learning_rate": 7.940672278564174e-06, "loss": 1.2757, "step": 15375 }, { "epoch": 1.2050156739811912, "grad_norm": 2.7593483924865723, "learning_rate": 7.940422881096462e-06, "loss": 1.3062, "step": 15376 }, { "epoch": 1.2050940438871474, "grad_norm": 4.706377983093262, "learning_rate": 7.940173472444891e-06, "loss": 1.4379, "step": 15377 }, { "epoch": 1.2051724137931035, "grad_norm": 2.382356882095337, "learning_rate": 7.939924052610411e-06, "loss": 1.7307, "step": 15378 }, { "epoch": 1.2052507836990596, "grad_norm": 2.8054044246673584, "learning_rate": 7.93967462159397e-06, "loss": 1.4359, "step": 15379 }, { "epoch": 1.2053291536050157, "grad_norm": 4.94146203994751, "learning_rate": 7.939425179396515e-06, "loss": 1.8106, "step": 15380 }, { "epoch": 1.2054075235109718, "grad_norm": 2.201107978820801, "learning_rate": 7.939175726018999e-06, "loss": 1.4356, "step": 15381 }, { "epoch": 1.205485893416928, "grad_norm": 3.033989906311035, "learning_rate": 7.938926261462366e-06, "loss": 0.982, "step": 15382 }, { "epoch": 1.205564263322884, "grad_norm": 2.9667797088623047, "learning_rate": 7.938676785727569e-06, "loss": 1.2564, "step": 15383 }, { "epoch": 1.2056426332288401, "grad_norm": 2.2076568603515625, "learning_rate": 7.938427298815554e-06, "loss": 1.2134, "step": 15384 }, { "epoch": 1.2057210031347962, "grad_norm": 2.4761955738067627, "learning_rate": 7.938177800727269e-06, "loss": 1.2717, "step": 15385 }, { "epoch": 1.2057993730407524, "grad_norm": 3.358887195587158, "learning_rate": 7.937928291463667e-06, "loss": 1.7319, "step": 15386 }, { "epoch": 1.2058777429467085, "grad_norm": 2.714589834213257, "learning_rate": 7.937678771025694e-06, "loss": 1.4239, "step": 15387 }, { "epoch": 1.2059561128526646, "grad_norm": 2.35978627204895, "learning_rate": 7.9374292394143e-06, "loss": 1.2978, "step": 15388 }, { "epoch": 1.2060344827586207, "grad_norm": 2.5609116554260254, "learning_rate": 7.937179696630435e-06, "loss": 0.946, "step": 15389 }, { "epoch": 1.2061128526645768, "grad_norm": 2.5515360832214355, "learning_rate": 7.936930142675045e-06, "loss": 1.428, "step": 15390 }, { "epoch": 1.206191222570533, "grad_norm": 2.50203013420105, "learning_rate": 7.936680577549083e-06, "loss": 1.4472, "step": 15391 }, { "epoch": 1.206269592476489, "grad_norm": 22.583721160888672, "learning_rate": 7.936431001253496e-06, "loss": 0.8278, "step": 15392 }, { "epoch": 1.2063479623824451, "grad_norm": 4.740391731262207, "learning_rate": 7.936181413789232e-06, "loss": 1.8394, "step": 15393 }, { "epoch": 1.2064263322884012, "grad_norm": 3.4434962272644043, "learning_rate": 7.935931815157244e-06, "loss": 1.1467, "step": 15394 }, { "epoch": 1.2065047021943573, "grad_norm": 2.7780561447143555, "learning_rate": 7.935682205358478e-06, "loss": 1.3843, "step": 15395 }, { "epoch": 1.2065830721003135, "grad_norm": 2.218073844909668, "learning_rate": 7.935432584393885e-06, "loss": 1.2974, "step": 15396 }, { "epoch": 1.2066614420062696, "grad_norm": 3.845414161682129, "learning_rate": 7.935182952264415e-06, "loss": 1.6798, "step": 15397 }, { "epoch": 1.2067398119122257, "grad_norm": 3.5231878757476807, "learning_rate": 7.934933308971017e-06, "loss": 1.4737, "step": 15398 }, { "epoch": 1.2068181818181818, "grad_norm": 3.207334518432617, "learning_rate": 7.93468365451464e-06, "loss": 1.4302, "step": 15399 }, { "epoch": 1.206896551724138, "grad_norm": 2.628352642059326, "learning_rate": 7.934433988896233e-06, "loss": 1.1646, "step": 15400 }, { "epoch": 1.206974921630094, "grad_norm": 3.490971326828003, "learning_rate": 7.934184312116746e-06, "loss": 1.3153, "step": 15401 }, { "epoch": 1.2070532915360501, "grad_norm": 4.392449378967285, "learning_rate": 7.93393462417713e-06, "loss": 1.1255, "step": 15402 }, { "epoch": 1.2071316614420062, "grad_norm": 2.6360697746276855, "learning_rate": 7.933684925078334e-06, "loss": 1.5421, "step": 15403 }, { "epoch": 1.2072100313479623, "grad_norm": 4.875308036804199, "learning_rate": 7.933435214821306e-06, "loss": 1.4291, "step": 15404 }, { "epoch": 1.2072884012539185, "grad_norm": 8.79981517791748, "learning_rate": 7.933185493406998e-06, "loss": 1.637, "step": 15405 }, { "epoch": 1.2073667711598746, "grad_norm": 9.701094627380371, "learning_rate": 7.932935760836358e-06, "loss": 0.8148, "step": 15406 }, { "epoch": 1.2074451410658307, "grad_norm": 2.893756866455078, "learning_rate": 7.932686017110338e-06, "loss": 1.5151, "step": 15407 }, { "epoch": 1.2075235109717868, "grad_norm": 3.5540571212768555, "learning_rate": 7.932436262229885e-06, "loss": 1.3203, "step": 15408 }, { "epoch": 1.207601880877743, "grad_norm": 2.871612310409546, "learning_rate": 7.932186496195953e-06, "loss": 1.1859, "step": 15409 }, { "epoch": 1.207680250783699, "grad_norm": 2.4565560817718506, "learning_rate": 7.931936719009487e-06, "loss": 1.3905, "step": 15410 }, { "epoch": 1.2077586206896551, "grad_norm": 1.9955518245697021, "learning_rate": 7.931686930671443e-06, "loss": 1.2598, "step": 15411 }, { "epoch": 1.2078369905956112, "grad_norm": 2.173961639404297, "learning_rate": 7.931437131182765e-06, "loss": 1.1536, "step": 15412 }, { "epoch": 1.2079153605015673, "grad_norm": 3.8790643215179443, "learning_rate": 7.931187320544407e-06, "loss": 1.3352, "step": 15413 }, { "epoch": 1.2079937304075234, "grad_norm": 4.27134895324707, "learning_rate": 7.930937498757318e-06, "loss": 1.3274, "step": 15414 }, { "epoch": 1.2080721003134796, "grad_norm": 2.3018407821655273, "learning_rate": 7.930687665822447e-06, "loss": 0.8003, "step": 15415 }, { "epoch": 1.2081504702194357, "grad_norm": 2.1337249279022217, "learning_rate": 7.930437821740746e-06, "loss": 1.3664, "step": 15416 }, { "epoch": 1.2082288401253918, "grad_norm": 2.542238473892212, "learning_rate": 7.930187966513165e-06, "loss": 1.4041, "step": 15417 }, { "epoch": 1.208307210031348, "grad_norm": 3.92218279838562, "learning_rate": 7.929938100140656e-06, "loss": 1.0636, "step": 15418 }, { "epoch": 1.208385579937304, "grad_norm": 2.904531955718994, "learning_rate": 7.929688222624164e-06, "loss": 1.831, "step": 15419 }, { "epoch": 1.20846394984326, "grad_norm": 3.3027069568634033, "learning_rate": 7.929438333964643e-06, "loss": 1.2416, "step": 15420 }, { "epoch": 1.2085423197492162, "grad_norm": 2.7407660484313965, "learning_rate": 7.929188434163047e-06, "loss": 1.3838, "step": 15421 }, { "epoch": 1.2086206896551723, "grad_norm": 3.470618486404419, "learning_rate": 7.928938523220319e-06, "loss": 1.0307, "step": 15422 }, { "epoch": 1.2086990595611284, "grad_norm": 2.8349854946136475, "learning_rate": 7.928688601137414e-06, "loss": 1.2604, "step": 15423 }, { "epoch": 1.2087774294670846, "grad_norm": 6.174540042877197, "learning_rate": 7.928438667915282e-06, "loss": 1.3826, "step": 15424 }, { "epoch": 1.2088557993730407, "grad_norm": 4.1893086433410645, "learning_rate": 7.928188723554875e-06, "loss": 1.5486, "step": 15425 }, { "epoch": 1.2089341692789968, "grad_norm": 6.488694190979004, "learning_rate": 7.92793876805714e-06, "loss": 1.027, "step": 15426 }, { "epoch": 1.2090125391849529, "grad_norm": 2.903738498687744, "learning_rate": 7.927688801423031e-06, "loss": 1.4053, "step": 15427 }, { "epoch": 1.209090909090909, "grad_norm": 2.658902168273926, "learning_rate": 7.927438823653497e-06, "loss": 1.3647, "step": 15428 }, { "epoch": 1.2091692789968653, "grad_norm": 2.7235913276672363, "learning_rate": 7.927188834749489e-06, "loss": 1.6209, "step": 15429 }, { "epoch": 1.2092476489028212, "grad_norm": 2.440300226211548, "learning_rate": 7.92693883471196e-06, "loss": 1.1636, "step": 15430 }, { "epoch": 1.2093260188087775, "grad_norm": 2.743499517440796, "learning_rate": 7.926688823541858e-06, "loss": 1.259, "step": 15431 }, { "epoch": 1.2094043887147334, "grad_norm": 3.6780169010162354, "learning_rate": 7.926438801240133e-06, "loss": 1.4995, "step": 15432 }, { "epoch": 1.2094827586206898, "grad_norm": 4.8716511726379395, "learning_rate": 7.926188767807739e-06, "loss": 2.0702, "step": 15433 }, { "epoch": 1.2095611285266457, "grad_norm": 4.584115505218506, "learning_rate": 7.925938723245628e-06, "loss": 1.6887, "step": 15434 }, { "epoch": 1.209639498432602, "grad_norm": 3.336196184158325, "learning_rate": 7.925688667554748e-06, "loss": 1.4724, "step": 15435 }, { "epoch": 1.2097178683385579, "grad_norm": 5.037459850311279, "learning_rate": 7.92543860073605e-06, "loss": 1.7953, "step": 15436 }, { "epoch": 1.2097962382445142, "grad_norm": 3.1526541709899902, "learning_rate": 7.925188522790485e-06, "loss": 1.1923, "step": 15437 }, { "epoch": 1.20987460815047, "grad_norm": 2.9466793537139893, "learning_rate": 7.924938433719009e-06, "loss": 1.6342, "step": 15438 }, { "epoch": 1.2099529780564264, "grad_norm": 4.973341464996338, "learning_rate": 7.924688333522566e-06, "loss": 1.5318, "step": 15439 }, { "epoch": 1.2100313479623825, "grad_norm": 3.9150285720825195, "learning_rate": 7.924438222202113e-06, "loss": 1.6615, "step": 15440 }, { "epoch": 1.2101097178683387, "grad_norm": 2.8439621925354004, "learning_rate": 7.924188099758599e-06, "loss": 1.3124, "step": 15441 }, { "epoch": 1.2101880877742948, "grad_norm": 1.9669126272201538, "learning_rate": 7.923937966192973e-06, "loss": 1.2592, "step": 15442 }, { "epoch": 1.2102664576802509, "grad_norm": 2.431568145751953, "learning_rate": 7.923687821506193e-06, "loss": 1.1962, "step": 15443 }, { "epoch": 1.210344827586207, "grad_norm": 2.3794450759887695, "learning_rate": 7.923437665699203e-06, "loss": 1.4114, "step": 15444 }, { "epoch": 1.210423197492163, "grad_norm": 2.938631057739258, "learning_rate": 7.923187498772959e-06, "loss": 1.3784, "step": 15445 }, { "epoch": 1.2105015673981192, "grad_norm": 3.4949450492858887, "learning_rate": 7.92293732072841e-06, "loss": 1.5334, "step": 15446 }, { "epoch": 1.2105799373040753, "grad_norm": 2.558519124984741, "learning_rate": 7.92268713156651e-06, "loss": 1.8109, "step": 15447 }, { "epoch": 1.2106583072100314, "grad_norm": 2.5500986576080322, "learning_rate": 7.922436931288208e-06, "loss": 1.469, "step": 15448 }, { "epoch": 1.2107366771159875, "grad_norm": 4.126236915588379, "learning_rate": 7.922186719894457e-06, "loss": 1.4664, "step": 15449 }, { "epoch": 1.2108150470219436, "grad_norm": 7.252036094665527, "learning_rate": 7.921936497386212e-06, "loss": 1.7472, "step": 15450 }, { "epoch": 1.2108934169278998, "grad_norm": 4.155810356140137, "learning_rate": 7.921686263764417e-06, "loss": 0.9266, "step": 15451 }, { "epoch": 1.2109717868338559, "grad_norm": 2.497344493865967, "learning_rate": 7.92143601903003e-06, "loss": 1.1645, "step": 15452 }, { "epoch": 1.211050156739812, "grad_norm": 2.241835355758667, "learning_rate": 7.921185763184002e-06, "loss": 1.3903, "step": 15453 }, { "epoch": 1.211128526645768, "grad_norm": 2.9763853549957275, "learning_rate": 7.920935496227283e-06, "loss": 1.1702, "step": 15454 }, { "epoch": 1.2112068965517242, "grad_norm": 4.343430519104004, "learning_rate": 7.920685218160824e-06, "loss": 1.4774, "step": 15455 }, { "epoch": 1.2112852664576803, "grad_norm": 4.528934001922607, "learning_rate": 7.920434928985581e-06, "loss": 1.4295, "step": 15456 }, { "epoch": 1.2113636363636364, "grad_norm": 3.445505380630493, "learning_rate": 7.920184628702503e-06, "loss": 1.075, "step": 15457 }, { "epoch": 1.2114420062695925, "grad_norm": 5.207956314086914, "learning_rate": 7.919934317312542e-06, "loss": 1.2341, "step": 15458 }, { "epoch": 1.2115203761755486, "grad_norm": 4.459601402282715, "learning_rate": 7.919683994816649e-06, "loss": 1.5863, "step": 15459 }, { "epoch": 1.2115987460815048, "grad_norm": 13.562170028686523, "learning_rate": 7.91943366121578e-06, "loss": 1.7522, "step": 15460 }, { "epoch": 1.2116771159874609, "grad_norm": 2.629063129425049, "learning_rate": 7.919183316510883e-06, "loss": 1.4993, "step": 15461 }, { "epoch": 1.211755485893417, "grad_norm": 2.8731892108917236, "learning_rate": 7.918932960702914e-06, "loss": 1.2945, "step": 15462 }, { "epoch": 1.211833855799373, "grad_norm": 2.993530511856079, "learning_rate": 7.918682593792822e-06, "loss": 1.4898, "step": 15463 }, { "epoch": 1.2119122257053292, "grad_norm": 1.9524210691452026, "learning_rate": 7.918432215781561e-06, "loss": 1.1158, "step": 15464 }, { "epoch": 1.2119905956112853, "grad_norm": 2.258833885192871, "learning_rate": 7.918181826670083e-06, "loss": 1.0068, "step": 15465 }, { "epoch": 1.2120689655172414, "grad_norm": 3.1174421310424805, "learning_rate": 7.917931426459343e-06, "loss": 1.0955, "step": 15466 }, { "epoch": 1.2121473354231975, "grad_norm": 5.08599328994751, "learning_rate": 7.917681015150286e-06, "loss": 1.644, "step": 15467 }, { "epoch": 1.2122257053291536, "grad_norm": 3.189955949783325, "learning_rate": 7.917430592743872e-06, "loss": 0.9999, "step": 15468 }, { "epoch": 1.2123040752351097, "grad_norm": 2.6701412200927734, "learning_rate": 7.917180159241047e-06, "loss": 1.2376, "step": 15469 }, { "epoch": 1.2123824451410659, "grad_norm": 5.3832902908325195, "learning_rate": 7.91692971464277e-06, "loss": 0.748, "step": 15470 }, { "epoch": 1.212460815047022, "grad_norm": 3.2549612522125244, "learning_rate": 7.91667925894999e-06, "loss": 1.4919, "step": 15471 }, { "epoch": 1.212539184952978, "grad_norm": 3.6926729679107666, "learning_rate": 7.91642879216366e-06, "loss": 1.3394, "step": 15472 }, { "epoch": 1.2126175548589342, "grad_norm": 5.003183364868164, "learning_rate": 7.916178314284735e-06, "loss": 1.0896, "step": 15473 }, { "epoch": 1.2126959247648903, "grad_norm": 2.760989189147949, "learning_rate": 7.915927825314162e-06, "loss": 0.9266, "step": 15474 }, { "epoch": 1.2127742946708464, "grad_norm": 4.530246257781982, "learning_rate": 7.9156773252529e-06, "loss": 1.6203, "step": 15475 }, { "epoch": 1.2128526645768025, "grad_norm": 2.5583128929138184, "learning_rate": 7.915426814101897e-06, "loss": 1.5983, "step": 15476 }, { "epoch": 1.2129310344827586, "grad_norm": 2.925293207168579, "learning_rate": 7.915176291862108e-06, "loss": 1.5212, "step": 15477 }, { "epoch": 1.2130094043887147, "grad_norm": 8.89316463470459, "learning_rate": 7.914925758534487e-06, "loss": 1.897, "step": 15478 }, { "epoch": 1.2130877742946709, "grad_norm": 2.764181613922119, "learning_rate": 7.914675214119984e-06, "loss": 1.1451, "step": 15479 }, { "epoch": 1.213166144200627, "grad_norm": 5.625870227813721, "learning_rate": 7.914424658619556e-06, "loss": 1.221, "step": 15480 }, { "epoch": 1.213244514106583, "grad_norm": 6.753697872161865, "learning_rate": 7.914174092034153e-06, "loss": 1.0604, "step": 15481 }, { "epoch": 1.2133228840125392, "grad_norm": 4.245622634887695, "learning_rate": 7.913923514364725e-06, "loss": 1.5849, "step": 15482 }, { "epoch": 1.2134012539184953, "grad_norm": 2.353416681289673, "learning_rate": 7.913672925612233e-06, "loss": 1.3897, "step": 15483 }, { "epoch": 1.2134796238244514, "grad_norm": 2.3858938217163086, "learning_rate": 7.913422325777624e-06, "loss": 1.5082, "step": 15484 }, { "epoch": 1.2135579937304075, "grad_norm": 2.7588863372802734, "learning_rate": 7.913171714861853e-06, "loss": 1.749, "step": 15485 }, { "epoch": 1.2136363636363636, "grad_norm": 2.85893177986145, "learning_rate": 7.912921092865873e-06, "loss": 1.2238, "step": 15486 }, { "epoch": 1.2137147335423197, "grad_norm": 3.0253264904022217, "learning_rate": 7.912670459790637e-06, "loss": 1.5438, "step": 15487 }, { "epoch": 1.2137931034482758, "grad_norm": 3.3075027465820312, "learning_rate": 7.9124198156371e-06, "loss": 1.2711, "step": 15488 }, { "epoch": 1.213871473354232, "grad_norm": 2.5384562015533447, "learning_rate": 7.912169160406213e-06, "loss": 1.1216, "step": 15489 }, { "epoch": 1.213949843260188, "grad_norm": 3.1625287532806396, "learning_rate": 7.911918494098928e-06, "loss": 1.3972, "step": 15490 }, { "epoch": 1.2140282131661442, "grad_norm": 2.517672300338745, "learning_rate": 7.911667816716204e-06, "loss": 1.3929, "step": 15491 }, { "epoch": 1.2141065830721003, "grad_norm": 9.159749984741211, "learning_rate": 7.911417128258991e-06, "loss": 1.6873, "step": 15492 }, { "epoch": 1.2141849529780564, "grad_norm": 8.266932487487793, "learning_rate": 7.91116642872824e-06, "loss": 1.8069, "step": 15493 }, { "epoch": 1.2142633228840125, "grad_norm": 2.015662431716919, "learning_rate": 7.91091571812491e-06, "loss": 1.1952, "step": 15494 }, { "epoch": 1.2143416927899686, "grad_norm": 2.3398964405059814, "learning_rate": 7.91066499644995e-06, "loss": 1.1155, "step": 15495 }, { "epoch": 1.2144200626959247, "grad_norm": 8.304723739624023, "learning_rate": 7.910414263704317e-06, "loss": 0.976, "step": 15496 }, { "epoch": 1.2144984326018808, "grad_norm": 2.6477248668670654, "learning_rate": 7.910163519888963e-06, "loss": 1.123, "step": 15497 }, { "epoch": 1.214576802507837, "grad_norm": 2.6465914249420166, "learning_rate": 7.90991276500484e-06, "loss": 1.5272, "step": 15498 }, { "epoch": 1.214655172413793, "grad_norm": 6.758338928222656, "learning_rate": 7.909661999052905e-06, "loss": 1.6966, "step": 15499 }, { "epoch": 1.2147335423197492, "grad_norm": 2.1937031745910645, "learning_rate": 7.90941122203411e-06, "loss": 1.3288, "step": 15500 }, { "epoch": 1.2148119122257053, "grad_norm": 2.253223419189453, "learning_rate": 7.909160433949408e-06, "loss": 1.6074, "step": 15501 }, { "epoch": 1.2148902821316614, "grad_norm": 7.372450828552246, "learning_rate": 7.908909634799755e-06, "loss": 1.1842, "step": 15502 }, { "epoch": 1.2149686520376175, "grad_norm": 3.915599822998047, "learning_rate": 7.908658824586103e-06, "loss": 1.4881, "step": 15503 }, { "epoch": 1.2150470219435736, "grad_norm": 2.226069688796997, "learning_rate": 7.908408003309407e-06, "loss": 1.1884, "step": 15504 }, { "epoch": 1.2151253918495297, "grad_norm": 3.312378168106079, "learning_rate": 7.908157170970621e-06, "loss": 1.3225, "step": 15505 }, { "epoch": 1.2152037617554858, "grad_norm": 12.647987365722656, "learning_rate": 7.907906327570698e-06, "loss": 1.6764, "step": 15506 }, { "epoch": 1.215282131661442, "grad_norm": 3.766268253326416, "learning_rate": 7.907655473110594e-06, "loss": 1.2114, "step": 15507 }, { "epoch": 1.215360501567398, "grad_norm": 4.035310745239258, "learning_rate": 7.907404607591262e-06, "loss": 1.3357, "step": 15508 }, { "epoch": 1.2154388714733542, "grad_norm": 2.954458475112915, "learning_rate": 7.907153731013658e-06, "loss": 1.3468, "step": 15509 }, { "epoch": 1.2155172413793103, "grad_norm": 3.0141091346740723, "learning_rate": 7.906902843378731e-06, "loss": 1.5749, "step": 15510 }, { "epoch": 1.2155956112852664, "grad_norm": 2.1540160179138184, "learning_rate": 7.90665194468744e-06, "loss": 1.2452, "step": 15511 }, { "epoch": 1.2156739811912225, "grad_norm": 2.3774449825286865, "learning_rate": 7.906401034940738e-06, "loss": 1.3035, "step": 15512 }, { "epoch": 1.2157523510971786, "grad_norm": 2.6476385593414307, "learning_rate": 7.906150114139579e-06, "loss": 1.0039, "step": 15513 }, { "epoch": 1.2158307210031347, "grad_norm": 4.672003269195557, "learning_rate": 7.90589918228492e-06, "loss": 1.6347, "step": 15514 }, { "epoch": 1.2159090909090908, "grad_norm": 5.6341400146484375, "learning_rate": 7.90564823937771e-06, "loss": 1.0876, "step": 15515 }, { "epoch": 1.215987460815047, "grad_norm": 3.8748092651367188, "learning_rate": 7.905397285418908e-06, "loss": 1.451, "step": 15516 }, { "epoch": 1.216065830721003, "grad_norm": 8.03612995147705, "learning_rate": 7.905146320409466e-06, "loss": 0.743, "step": 15517 }, { "epoch": 1.2161442006269592, "grad_norm": 2.7849929332733154, "learning_rate": 7.90489534435034e-06, "loss": 1.1564, "step": 15518 }, { "epoch": 1.2162225705329153, "grad_norm": 2.879382848739624, "learning_rate": 7.904644357242484e-06, "loss": 1.8583, "step": 15519 }, { "epoch": 1.2163009404388714, "grad_norm": 2.895155906677246, "learning_rate": 7.904393359086854e-06, "loss": 1.3902, "step": 15520 }, { "epoch": 1.2163793103448275, "grad_norm": 3.1898281574249268, "learning_rate": 7.904142349884402e-06, "loss": 1.2271, "step": 15521 }, { "epoch": 1.2164576802507838, "grad_norm": 3.8431215286254883, "learning_rate": 7.903891329636085e-06, "loss": 1.5585, "step": 15522 }, { "epoch": 1.2165360501567397, "grad_norm": 3.560433864593506, "learning_rate": 7.903640298342856e-06, "loss": 1.462, "step": 15523 }, { "epoch": 1.216614420062696, "grad_norm": 2.260829210281372, "learning_rate": 7.903389256005672e-06, "loss": 1.3678, "step": 15524 }, { "epoch": 1.216692789968652, "grad_norm": 2.9855289459228516, "learning_rate": 7.903138202625486e-06, "loss": 1.048, "step": 15525 }, { "epoch": 1.2167711598746083, "grad_norm": 4.949279308319092, "learning_rate": 7.902887138203254e-06, "loss": 1.4298, "step": 15526 }, { "epoch": 1.2168495297805642, "grad_norm": 2.7985236644744873, "learning_rate": 7.902636062739928e-06, "loss": 1.4687, "step": 15527 }, { "epoch": 1.2169278996865205, "grad_norm": 2.7133948802948, "learning_rate": 7.902384976236468e-06, "loss": 1.5978, "step": 15528 }, { "epoch": 1.2170062695924764, "grad_norm": 2.640739917755127, "learning_rate": 7.902133878693826e-06, "loss": 1.2012, "step": 15529 }, { "epoch": 1.2170846394984327, "grad_norm": 2.541888952255249, "learning_rate": 7.901882770112956e-06, "loss": 1.4386, "step": 15530 }, { "epoch": 1.2171630094043886, "grad_norm": 2.5115807056427, "learning_rate": 7.901631650494817e-06, "loss": 1.2962, "step": 15531 }, { "epoch": 1.217241379310345, "grad_norm": 3.59431791305542, "learning_rate": 7.901380519840358e-06, "loss": 1.3711, "step": 15532 }, { "epoch": 1.2173197492163008, "grad_norm": 5.484794616699219, "learning_rate": 7.90112937815054e-06, "loss": 1.2718, "step": 15533 }, { "epoch": 1.2173981191222571, "grad_norm": 3.652967691421509, "learning_rate": 7.900878225426317e-06, "loss": 1.3829, "step": 15534 }, { "epoch": 1.2174764890282133, "grad_norm": 4.267850875854492, "learning_rate": 7.90062706166864e-06, "loss": 1.3249, "step": 15535 }, { "epoch": 1.2175548589341694, "grad_norm": 3.382530689239502, "learning_rate": 7.900375886878468e-06, "loss": 1.6426, "step": 15536 }, { "epoch": 1.2176332288401255, "grad_norm": 3.3802506923675537, "learning_rate": 7.900124701056758e-06, "loss": 1.673, "step": 15537 }, { "epoch": 1.2177115987460816, "grad_norm": 3.736461639404297, "learning_rate": 7.899873504204463e-06, "loss": 1.4873, "step": 15538 }, { "epoch": 1.2177899686520377, "grad_norm": 3.0761897563934326, "learning_rate": 7.899622296322538e-06, "loss": 1.2143, "step": 15539 }, { "epoch": 1.2178683385579938, "grad_norm": 3.5722124576568604, "learning_rate": 7.899371077411939e-06, "loss": 1.3031, "step": 15540 }, { "epoch": 1.21794670846395, "grad_norm": 3.1877875328063965, "learning_rate": 7.899119847473622e-06, "loss": 1.2625, "step": 15541 }, { "epoch": 1.218025078369906, "grad_norm": 3.3783555030822754, "learning_rate": 7.898868606508542e-06, "loss": 1.298, "step": 15542 }, { "epoch": 1.2181034482758621, "grad_norm": 2.7102580070495605, "learning_rate": 7.898617354517653e-06, "loss": 1.4883, "step": 15543 }, { "epoch": 1.2181818181818183, "grad_norm": 3.277573585510254, "learning_rate": 7.898366091501914e-06, "loss": 1.5985, "step": 15544 }, { "epoch": 1.2182601880877744, "grad_norm": 3.342994451522827, "learning_rate": 7.89811481746228e-06, "loss": 1.2191, "step": 15545 }, { "epoch": 1.2183385579937305, "grad_norm": 8.281974792480469, "learning_rate": 7.897863532399706e-06, "loss": 1.4018, "step": 15546 }, { "epoch": 1.2184169278996866, "grad_norm": 2.3374593257904053, "learning_rate": 7.897612236315145e-06, "loss": 1.2074, "step": 15547 }, { "epoch": 1.2184952978056427, "grad_norm": 10.883734703063965, "learning_rate": 7.897360929209557e-06, "loss": 1.3295, "step": 15548 }, { "epoch": 1.2185736677115988, "grad_norm": 2.113224983215332, "learning_rate": 7.897109611083896e-06, "loss": 1.4287, "step": 15549 }, { "epoch": 1.218652037617555, "grad_norm": 3.3582537174224854, "learning_rate": 7.896858281939118e-06, "loss": 1.1638, "step": 15550 }, { "epoch": 1.218730407523511, "grad_norm": 3.557008981704712, "learning_rate": 7.896606941776179e-06, "loss": 0.944, "step": 15551 }, { "epoch": 1.2188087774294671, "grad_norm": 2.4801621437072754, "learning_rate": 7.896355590596036e-06, "loss": 1.2402, "step": 15552 }, { "epoch": 1.2188871473354232, "grad_norm": 2.8638784885406494, "learning_rate": 7.896104228399642e-06, "loss": 1.45, "step": 15553 }, { "epoch": 1.2189655172413794, "grad_norm": 6.196885585784912, "learning_rate": 7.895852855187957e-06, "loss": 1.5406, "step": 15554 }, { "epoch": 1.2190438871473355, "grad_norm": 4.899753093719482, "learning_rate": 7.895601470961935e-06, "loss": 1.9642, "step": 15555 }, { "epoch": 1.2191222570532916, "grad_norm": 3.1557140350341797, "learning_rate": 7.89535007572253e-06, "loss": 1.1187, "step": 15556 }, { "epoch": 1.2192006269592477, "grad_norm": 2.652515172958374, "learning_rate": 7.895098669470704e-06, "loss": 1.5168, "step": 15557 }, { "epoch": 1.2192789968652038, "grad_norm": 4.770417213439941, "learning_rate": 7.894847252207407e-06, "loss": 1.6299, "step": 15558 }, { "epoch": 1.21935736677116, "grad_norm": 5.808052062988281, "learning_rate": 7.894595823933598e-06, "loss": 0.9979, "step": 15559 }, { "epoch": 1.219435736677116, "grad_norm": 4.087418556213379, "learning_rate": 7.894344384650235e-06, "loss": 1.3168, "step": 15560 }, { "epoch": 1.2195141065830721, "grad_norm": 3.720743179321289, "learning_rate": 7.89409293435827e-06, "loss": 1.3931, "step": 15561 }, { "epoch": 1.2195924764890282, "grad_norm": 4.788809776306152, "learning_rate": 7.893841473058664e-06, "loss": 0.8743, "step": 15562 }, { "epoch": 1.2196708463949844, "grad_norm": 2.517836332321167, "learning_rate": 7.893590000752371e-06, "loss": 1.6185, "step": 15563 }, { "epoch": 1.2197492163009405, "grad_norm": 7.868228435516357, "learning_rate": 7.893338517440347e-06, "loss": 1.6887, "step": 15564 }, { "epoch": 1.2198275862068966, "grad_norm": 4.3607683181762695, "learning_rate": 7.89308702312355e-06, "loss": 1.1501, "step": 15565 }, { "epoch": 1.2199059561128527, "grad_norm": 9.362008094787598, "learning_rate": 7.892835517802937e-06, "loss": 1.7904, "step": 15566 }, { "epoch": 1.2199843260188088, "grad_norm": 3.4023170471191406, "learning_rate": 7.892584001479462e-06, "loss": 1.2576, "step": 15567 }, { "epoch": 1.220062695924765, "grad_norm": 2.121365785598755, "learning_rate": 7.892332474154085e-06, "loss": 1.4264, "step": 15568 }, { "epoch": 1.220141065830721, "grad_norm": 3.2957184314727783, "learning_rate": 7.892080935827759e-06, "loss": 1.4507, "step": 15569 }, { "epoch": 1.2202194357366771, "grad_norm": 7.348779201507568, "learning_rate": 7.891829386501444e-06, "loss": 0.9292, "step": 15570 }, { "epoch": 1.2202978056426332, "grad_norm": 3.5357553958892822, "learning_rate": 7.891577826176094e-06, "loss": 1.0558, "step": 15571 }, { "epoch": 1.2203761755485893, "grad_norm": 3.401956796646118, "learning_rate": 7.891326254852667e-06, "loss": 1.4238, "step": 15572 }, { "epoch": 1.2204545454545455, "grad_norm": 2.7598979473114014, "learning_rate": 7.891074672532122e-06, "loss": 1.059, "step": 15573 }, { "epoch": 1.2205329153605016, "grad_norm": 2.3279175758361816, "learning_rate": 7.890823079215413e-06, "loss": 1.3352, "step": 15574 }, { "epoch": 1.2206112852664577, "grad_norm": 4.509481430053711, "learning_rate": 7.890571474903497e-06, "loss": 1.254, "step": 15575 }, { "epoch": 1.2206896551724138, "grad_norm": 3.215458393096924, "learning_rate": 7.890319859597333e-06, "loss": 0.7475, "step": 15576 }, { "epoch": 1.22076802507837, "grad_norm": 5.075228691101074, "learning_rate": 7.890068233297876e-06, "loss": 1.6174, "step": 15577 }, { "epoch": 1.220846394984326, "grad_norm": 3.1927056312561035, "learning_rate": 7.889816596006084e-06, "loss": 1.4197, "step": 15578 }, { "epoch": 1.2209247648902821, "grad_norm": 2.680569887161255, "learning_rate": 7.889564947722914e-06, "loss": 1.503, "step": 15579 }, { "epoch": 1.2210031347962382, "grad_norm": 4.321735858917236, "learning_rate": 7.889313288449323e-06, "loss": 1.0004, "step": 15580 }, { "epoch": 1.2210815047021943, "grad_norm": 2.453024387359619, "learning_rate": 7.889061618186268e-06, "loss": 1.4889, "step": 15581 }, { "epoch": 1.2211598746081505, "grad_norm": 5.664204120635986, "learning_rate": 7.888809936934708e-06, "loss": 1.4853, "step": 15582 }, { "epoch": 1.2212382445141066, "grad_norm": 2.4831578731536865, "learning_rate": 7.888558244695597e-06, "loss": 1.4755, "step": 15583 }, { "epoch": 1.2213166144200627, "grad_norm": 2.036774158477783, "learning_rate": 7.888306541469894e-06, "loss": 1.3595, "step": 15584 }, { "epoch": 1.2213949843260188, "grad_norm": 2.883185386657715, "learning_rate": 7.888054827258557e-06, "loss": 1.3446, "step": 15585 }, { "epoch": 1.221473354231975, "grad_norm": 3.623507022857666, "learning_rate": 7.887803102062543e-06, "loss": 1.4649, "step": 15586 }, { "epoch": 1.221551724137931, "grad_norm": 3.8529295921325684, "learning_rate": 7.88755136588281e-06, "loss": 1.4911, "step": 15587 }, { "epoch": 1.2216300940438871, "grad_norm": 1.8676611185073853, "learning_rate": 7.887299618720314e-06, "loss": 1.2719, "step": 15588 }, { "epoch": 1.2217084639498432, "grad_norm": 3.1896204948425293, "learning_rate": 7.887047860576014e-06, "loss": 1.577, "step": 15589 }, { "epoch": 1.2217868338557993, "grad_norm": 3.1089882850646973, "learning_rate": 7.886796091450865e-06, "loss": 1.5647, "step": 15590 }, { "epoch": 1.2218652037617554, "grad_norm": 4.429203510284424, "learning_rate": 7.886544311345828e-06, "loss": 2.1958, "step": 15591 }, { "epoch": 1.2219435736677116, "grad_norm": 3.1996817588806152, "learning_rate": 7.88629252026186e-06, "loss": 1.2631, "step": 15592 }, { "epoch": 1.2220219435736677, "grad_norm": 2.4125850200653076, "learning_rate": 7.886040718199917e-06, "loss": 1.3679, "step": 15593 }, { "epoch": 1.2221003134796238, "grad_norm": 3.0813169479370117, "learning_rate": 7.885788905160956e-06, "loss": 1.2743, "step": 15594 }, { "epoch": 1.22217868338558, "grad_norm": 2.8131062984466553, "learning_rate": 7.885537081145938e-06, "loss": 1.4327, "step": 15595 }, { "epoch": 1.222257053291536, "grad_norm": 4.675086975097656, "learning_rate": 7.885285246155816e-06, "loss": 1.3483, "step": 15596 }, { "epoch": 1.222335423197492, "grad_norm": 2.551488161087036, "learning_rate": 7.885033400191555e-06, "loss": 1.5705, "step": 15597 }, { "epoch": 1.2224137931034482, "grad_norm": 3.5123445987701416, "learning_rate": 7.884781543254106e-06, "loss": 1.3227, "step": 15598 }, { "epoch": 1.2224921630094043, "grad_norm": 2.525324583053589, "learning_rate": 7.88452967534443e-06, "loss": 1.3573, "step": 15599 }, { "epoch": 1.2225705329153604, "grad_norm": 3.14738392829895, "learning_rate": 7.884277796463487e-06, "loss": 1.6456, "step": 15600 }, { "epoch": 1.2226489028213166, "grad_norm": 5.633378982543945, "learning_rate": 7.88402590661223e-06, "loss": 1.8633, "step": 15601 }, { "epoch": 1.2227272727272727, "grad_norm": 3.3717708587646484, "learning_rate": 7.88377400579162e-06, "loss": 1.3832, "step": 15602 }, { "epoch": 1.2228056426332288, "grad_norm": 2.9403111934661865, "learning_rate": 7.883522094002616e-06, "loss": 0.8994, "step": 15603 }, { "epoch": 1.2228840125391849, "grad_norm": 3.3510894775390625, "learning_rate": 7.883270171246176e-06, "loss": 1.236, "step": 15604 }, { "epoch": 1.222962382445141, "grad_norm": 3.8495748043060303, "learning_rate": 7.883018237523254e-06, "loss": 1.1797, "step": 15605 }, { "epoch": 1.223040752351097, "grad_norm": 6.722247123718262, "learning_rate": 7.882766292834815e-06, "loss": 1.3307, "step": 15606 }, { "epoch": 1.2231191222570532, "grad_norm": 2.4157211780548096, "learning_rate": 7.88251433718181e-06, "loss": 1.3636, "step": 15607 }, { "epoch": 1.2231974921630093, "grad_norm": 2.863865375518799, "learning_rate": 7.882262370565205e-06, "loss": 1.5946, "step": 15608 }, { "epoch": 1.2232758620689654, "grad_norm": 2.0988268852233887, "learning_rate": 7.882010392985951e-06, "loss": 1.0786, "step": 15609 }, { "epoch": 1.2233542319749215, "grad_norm": 12.499938011169434, "learning_rate": 7.881758404445012e-06, "loss": 1.3788, "step": 15610 }, { "epoch": 1.2234326018808777, "grad_norm": 2.5629706382751465, "learning_rate": 7.881506404943344e-06, "loss": 1.6049, "step": 15611 }, { "epoch": 1.2235109717868338, "grad_norm": 1.9909775257110596, "learning_rate": 7.881254394481905e-06, "loss": 0.9585, "step": 15612 }, { "epoch": 1.2235893416927899, "grad_norm": 2.6553537845611572, "learning_rate": 7.881002373061653e-06, "loss": 1.4964, "step": 15613 }, { "epoch": 1.223667711598746, "grad_norm": 3.0121772289276123, "learning_rate": 7.880750340683549e-06, "loss": 1.3223, "step": 15614 }, { "epoch": 1.223746081504702, "grad_norm": 4.049540042877197, "learning_rate": 7.880498297348551e-06, "loss": 1.9257, "step": 15615 }, { "epoch": 1.2238244514106582, "grad_norm": 3.977036714553833, "learning_rate": 7.880246243057616e-06, "loss": 1.4345, "step": 15616 }, { "epoch": 1.2239028213166145, "grad_norm": 3.3478941917419434, "learning_rate": 7.879994177811705e-06, "loss": 1.4442, "step": 15617 }, { "epoch": 1.2239811912225704, "grad_norm": 1.9716368913650513, "learning_rate": 7.879742101611773e-06, "loss": 1.1672, "step": 15618 }, { "epoch": 1.2240595611285268, "grad_norm": 2.38055157661438, "learning_rate": 7.879490014458784e-06, "loss": 1.0875, "step": 15619 }, { "epoch": 1.2241379310344827, "grad_norm": 6.267146110534668, "learning_rate": 7.879237916353692e-06, "loss": 1.5857, "step": 15620 }, { "epoch": 1.224216300940439, "grad_norm": 3.473456859588623, "learning_rate": 7.878985807297459e-06, "loss": 1.3913, "step": 15621 }, { "epoch": 1.2242946708463949, "grad_norm": 2.0081264972686768, "learning_rate": 7.878733687291041e-06, "loss": 1.3908, "step": 15622 }, { "epoch": 1.2243730407523512, "grad_norm": 3.157236337661743, "learning_rate": 7.8784815563354e-06, "loss": 1.5041, "step": 15623 }, { "epoch": 1.224451410658307, "grad_norm": 10.983642578125, "learning_rate": 7.878229414431493e-06, "loss": 0.747, "step": 15624 }, { "epoch": 1.2245297805642634, "grad_norm": 3.0500969886779785, "learning_rate": 7.877977261580281e-06, "loss": 1.3935, "step": 15625 }, { "epoch": 1.2246081504702193, "grad_norm": 4.054320335388184, "learning_rate": 7.877725097782721e-06, "loss": 1.3876, "step": 15626 }, { "epoch": 1.2246865203761756, "grad_norm": 4.642301559448242, "learning_rate": 7.877472923039772e-06, "loss": 0.9261, "step": 15627 }, { "epoch": 1.2247648902821318, "grad_norm": 5.157763957977295, "learning_rate": 7.877220737352395e-06, "loss": 1.5476, "step": 15628 }, { "epoch": 1.2248432601880879, "grad_norm": 2.7208847999572754, "learning_rate": 7.876968540721548e-06, "loss": 0.9444, "step": 15629 }, { "epoch": 1.224921630094044, "grad_norm": 3.6369144916534424, "learning_rate": 7.87671633314819e-06, "loss": 1.7689, "step": 15630 }, { "epoch": 1.225, "grad_norm": 2.381193161010742, "learning_rate": 7.87646411463328e-06, "loss": 1.61, "step": 15631 }, { "epoch": 1.2250783699059562, "grad_norm": 2.3252475261688232, "learning_rate": 7.87621188517778e-06, "loss": 1.2323, "step": 15632 }, { "epoch": 1.2251567398119123, "grad_norm": 2.3289694786071777, "learning_rate": 7.875959644782645e-06, "loss": 1.2173, "step": 15633 }, { "epoch": 1.2252351097178684, "grad_norm": 4.833980560302734, "learning_rate": 7.875707393448838e-06, "loss": 0.9746, "step": 15634 }, { "epoch": 1.2253134796238245, "grad_norm": 4.221931457519531, "learning_rate": 7.875455131177317e-06, "loss": 1.544, "step": 15635 }, { "epoch": 1.2253918495297806, "grad_norm": 2.5153868198394775, "learning_rate": 7.875202857969041e-06, "loss": 1.4518, "step": 15636 }, { "epoch": 1.2254702194357368, "grad_norm": 2.2223658561706543, "learning_rate": 7.87495057382497e-06, "loss": 1.2905, "step": 15637 }, { "epoch": 1.2255485893416929, "grad_norm": 2.7438735961914062, "learning_rate": 7.874698278746065e-06, "loss": 1.3435, "step": 15638 }, { "epoch": 1.225626959247649, "grad_norm": 3.338853120803833, "learning_rate": 7.874445972733284e-06, "loss": 1.1741, "step": 15639 }, { "epoch": 1.225705329153605, "grad_norm": 2.4277431964874268, "learning_rate": 7.874193655787586e-06, "loss": 1.769, "step": 15640 }, { "epoch": 1.2257836990595612, "grad_norm": 2.8653764724731445, "learning_rate": 7.873941327909933e-06, "loss": 1.5194, "step": 15641 }, { "epoch": 1.2258620689655173, "grad_norm": 2.4362363815307617, "learning_rate": 7.873688989101282e-06, "loss": 1.3281, "step": 15642 }, { "epoch": 1.2259404388714734, "grad_norm": 2.5423073768615723, "learning_rate": 7.873436639362595e-06, "loss": 1.3026, "step": 15643 }, { "epoch": 1.2260188087774295, "grad_norm": 4.19612455368042, "learning_rate": 7.87318427869483e-06, "loss": 1.6325, "step": 15644 }, { "epoch": 1.2260971786833856, "grad_norm": 2.2940897941589355, "learning_rate": 7.872931907098949e-06, "loss": 1.2744, "step": 15645 }, { "epoch": 1.2261755485893417, "grad_norm": 2.5787765979766846, "learning_rate": 7.87267952457591e-06, "loss": 1.0398, "step": 15646 }, { "epoch": 1.2262539184952979, "grad_norm": 2.729789972305298, "learning_rate": 7.872427131126673e-06, "loss": 1.4245, "step": 15647 }, { "epoch": 1.226332288401254, "grad_norm": 4.166090488433838, "learning_rate": 7.872174726752198e-06, "loss": 2.001, "step": 15648 }, { "epoch": 1.22641065830721, "grad_norm": 4.028538703918457, "learning_rate": 7.871922311453449e-06, "loss": 1.3063, "step": 15649 }, { "epoch": 1.2264890282131662, "grad_norm": 7.101220607757568, "learning_rate": 7.871669885231378e-06, "loss": 1.6468, "step": 15650 }, { "epoch": 1.2265673981191223, "grad_norm": 4.7458271980285645, "learning_rate": 7.871417448086953e-06, "loss": 2.5863, "step": 15651 }, { "epoch": 1.2266457680250784, "grad_norm": 3.540407657623291, "learning_rate": 7.87116500002113e-06, "loss": 1.2897, "step": 15652 }, { "epoch": 1.2267241379310345, "grad_norm": 2.5558319091796875, "learning_rate": 7.870912541034869e-06, "loss": 1.6038, "step": 15653 }, { "epoch": 1.2268025078369906, "grad_norm": 6.8429388999938965, "learning_rate": 7.870660071129134e-06, "loss": 1.3649, "step": 15654 }, { "epoch": 1.2268808777429467, "grad_norm": 3.781719446182251, "learning_rate": 7.87040759030488e-06, "loss": 1.0918, "step": 15655 }, { "epoch": 1.2269592476489029, "grad_norm": 8.36611557006836, "learning_rate": 7.870155098563071e-06, "loss": 1.0855, "step": 15656 }, { "epoch": 1.227037617554859, "grad_norm": 3.010113477706909, "learning_rate": 7.869902595904664e-06, "loss": 1.1763, "step": 15657 }, { "epoch": 1.227115987460815, "grad_norm": 2.7271597385406494, "learning_rate": 7.869650082330625e-06, "loss": 1.5246, "step": 15658 }, { "epoch": 1.2271943573667712, "grad_norm": 3.1083059310913086, "learning_rate": 7.86939755784191e-06, "loss": 1.2894, "step": 15659 }, { "epoch": 1.2272727272727273, "grad_norm": 5.369534969329834, "learning_rate": 7.869145022439479e-06, "loss": 1.0058, "step": 15660 }, { "epoch": 1.2273510971786834, "grad_norm": 4.985074043273926, "learning_rate": 7.868892476124294e-06, "loss": 0.885, "step": 15661 }, { "epoch": 1.2274294670846395, "grad_norm": 2.213956356048584, "learning_rate": 7.868639918897318e-06, "loss": 1.4259, "step": 15662 }, { "epoch": 1.2275078369905956, "grad_norm": 1.9711483716964722, "learning_rate": 7.868387350759508e-06, "loss": 1.3616, "step": 15663 }, { "epoch": 1.2275862068965517, "grad_norm": 3.2628817558288574, "learning_rate": 7.868134771711825e-06, "loss": 1.2496, "step": 15664 }, { "epoch": 1.2276645768025078, "grad_norm": 2.4740676879882812, "learning_rate": 7.86788218175523e-06, "loss": 1.6339, "step": 15665 }, { "epoch": 1.227742946708464, "grad_norm": 2.655592203140259, "learning_rate": 7.867629580890687e-06, "loss": 1.2317, "step": 15666 }, { "epoch": 1.22782131661442, "grad_norm": 3.0646352767944336, "learning_rate": 7.867376969119151e-06, "loss": 1.3418, "step": 15667 }, { "epoch": 1.2278996865203762, "grad_norm": 5.572106838226318, "learning_rate": 7.867124346441588e-06, "loss": 1.3943, "step": 15668 }, { "epoch": 1.2279780564263323, "grad_norm": 2.323817491531372, "learning_rate": 7.866871712858957e-06, "loss": 1.5163, "step": 15669 }, { "epoch": 1.2280564263322884, "grad_norm": 2.5853652954101562, "learning_rate": 7.866619068372217e-06, "loss": 1.2084, "step": 15670 }, { "epoch": 1.2281347962382445, "grad_norm": 5.157432556152344, "learning_rate": 7.86636641298233e-06, "loss": 1.1425, "step": 15671 }, { "epoch": 1.2282131661442006, "grad_norm": 2.897639036178589, "learning_rate": 7.866113746690258e-06, "loss": 0.7938, "step": 15672 }, { "epoch": 1.2282915360501567, "grad_norm": 3.9820544719696045, "learning_rate": 7.86586106949696e-06, "loss": 1.4138, "step": 15673 }, { "epoch": 1.2283699059561128, "grad_norm": 3.755021572113037, "learning_rate": 7.865608381403401e-06, "loss": 1.4812, "step": 15674 }, { "epoch": 1.228448275862069, "grad_norm": 4.490444660186768, "learning_rate": 7.86535568241054e-06, "loss": 1.6389, "step": 15675 }, { "epoch": 1.228526645768025, "grad_norm": 3.1342060565948486, "learning_rate": 7.865102972519336e-06, "loss": 1.4623, "step": 15676 }, { "epoch": 1.2286050156739812, "grad_norm": 4.6687822341918945, "learning_rate": 7.864850251730751e-06, "loss": 1.2495, "step": 15677 }, { "epoch": 1.2286833855799373, "grad_norm": 5.457071781158447, "learning_rate": 7.864597520045749e-06, "loss": 1.5632, "step": 15678 }, { "epoch": 1.2287617554858934, "grad_norm": 2.6566717624664307, "learning_rate": 7.86434477746529e-06, "loss": 1.6863, "step": 15679 }, { "epoch": 1.2288401253918495, "grad_norm": 2.7542567253112793, "learning_rate": 7.864092023990332e-06, "loss": 1.2323, "step": 15680 }, { "epoch": 1.2289184952978056, "grad_norm": 2.1832332611083984, "learning_rate": 7.86383925962184e-06, "loss": 1.3885, "step": 15681 }, { "epoch": 1.2289968652037617, "grad_norm": 2.0070080757141113, "learning_rate": 7.863586484360777e-06, "loss": 1.1736, "step": 15682 }, { "epoch": 1.2290752351097178, "grad_norm": 2.4857919216156006, "learning_rate": 7.863333698208098e-06, "loss": 1.5269, "step": 15683 }, { "epoch": 1.229153605015674, "grad_norm": 2.989241361618042, "learning_rate": 7.863080901164771e-06, "loss": 1.722, "step": 15684 }, { "epoch": 1.22923197492163, "grad_norm": 2.60636568069458, "learning_rate": 7.862828093231753e-06, "loss": 1.46, "step": 15685 }, { "epoch": 1.2293103448275862, "grad_norm": 2.4286351203918457, "learning_rate": 7.86257527441001e-06, "loss": 1.5819, "step": 15686 }, { "epoch": 1.2293887147335423, "grad_norm": 2.482999563217163, "learning_rate": 7.862322444700498e-06, "loss": 1.3724, "step": 15687 }, { "epoch": 1.2294670846394984, "grad_norm": 3.1910147666931152, "learning_rate": 7.862069604104183e-06, "loss": 1.4179, "step": 15688 }, { "epoch": 1.2295454545454545, "grad_norm": 2.55145001411438, "learning_rate": 7.861816752622025e-06, "loss": 0.9555, "step": 15689 }, { "epoch": 1.2296238244514106, "grad_norm": 2.375462293624878, "learning_rate": 7.861563890254987e-06, "loss": 1.0986, "step": 15690 }, { "epoch": 1.2297021943573667, "grad_norm": 7.465587139129639, "learning_rate": 7.861311017004028e-06, "loss": 1.6177, "step": 15691 }, { "epoch": 1.2297805642633228, "grad_norm": 2.8769290447235107, "learning_rate": 7.861058132870111e-06, "loss": 1.5756, "step": 15692 }, { "epoch": 1.229858934169279, "grad_norm": 2.977564573287964, "learning_rate": 7.860805237854202e-06, "loss": 1.5139, "step": 15693 }, { "epoch": 1.229937304075235, "grad_norm": 2.4708011150360107, "learning_rate": 7.860552331957256e-06, "loss": 1.6165, "step": 15694 }, { "epoch": 1.2300156739811912, "grad_norm": 2.8031344413757324, "learning_rate": 7.860299415180238e-06, "loss": 1.3049, "step": 15695 }, { "epoch": 1.2300940438871473, "grad_norm": 2.9344794750213623, "learning_rate": 7.860046487524113e-06, "loss": 1.4743, "step": 15696 }, { "epoch": 1.2301724137931034, "grad_norm": 2.4991796016693115, "learning_rate": 7.859793548989837e-06, "loss": 1.2356, "step": 15697 }, { "epoch": 1.2302507836990595, "grad_norm": 3.170841693878174, "learning_rate": 7.859540599578376e-06, "loss": 1.2946, "step": 15698 }, { "epoch": 1.2303291536050156, "grad_norm": 2.2259745597839355, "learning_rate": 7.859287639290694e-06, "loss": 1.3905, "step": 15699 }, { "epoch": 1.2304075235109717, "grad_norm": 2.4399406909942627, "learning_rate": 7.859034668127749e-06, "loss": 1.2504, "step": 15700 }, { "epoch": 1.2304858934169278, "grad_norm": 3.824962854385376, "learning_rate": 7.858781686090504e-06, "loss": 1.4134, "step": 15701 }, { "epoch": 1.230564263322884, "grad_norm": 2.2937710285186768, "learning_rate": 7.858528693179922e-06, "loss": 1.0725, "step": 15702 }, { "epoch": 1.23064263322884, "grad_norm": 2.4605557918548584, "learning_rate": 7.858275689396965e-06, "loss": 1.5722, "step": 15703 }, { "epoch": 1.2307210031347962, "grad_norm": 2.330394744873047, "learning_rate": 7.858022674742595e-06, "loss": 1.4232, "step": 15704 }, { "epoch": 1.2307993730407523, "grad_norm": 4.354261875152588, "learning_rate": 7.857769649217775e-06, "loss": 1.2702, "step": 15705 }, { "epoch": 1.2308777429467084, "grad_norm": 2.144026517868042, "learning_rate": 7.857516612823468e-06, "loss": 1.4822, "step": 15706 }, { "epoch": 1.2309561128526645, "grad_norm": 2.471595287322998, "learning_rate": 7.857263565560635e-06, "loss": 1.2992, "step": 15707 }, { "epoch": 1.2310344827586206, "grad_norm": 3.0163328647613525, "learning_rate": 7.857010507430239e-06, "loss": 1.2734, "step": 15708 }, { "epoch": 1.2311128526645767, "grad_norm": 2.8294568061828613, "learning_rate": 7.856757438433242e-06, "loss": 1.302, "step": 15709 }, { "epoch": 1.231191222570533, "grad_norm": 2.772805690765381, "learning_rate": 7.856504358570608e-06, "loss": 1.3902, "step": 15710 }, { "epoch": 1.231269592476489, "grad_norm": 3.008974075317383, "learning_rate": 7.856251267843298e-06, "loss": 1.1604, "step": 15711 }, { "epoch": 1.2313479623824453, "grad_norm": 3.653461217880249, "learning_rate": 7.855998166252275e-06, "loss": 1.5661, "step": 15712 }, { "epoch": 1.2314263322884011, "grad_norm": 3.0634498596191406, "learning_rate": 7.855745053798503e-06, "loss": 1.6295, "step": 15713 }, { "epoch": 1.2315047021943575, "grad_norm": 2.677502155303955, "learning_rate": 7.855491930482942e-06, "loss": 1.5843, "step": 15714 }, { "epoch": 1.2315830721003134, "grad_norm": 2.8748748302459717, "learning_rate": 7.855238796306558e-06, "loss": 1.1339, "step": 15715 }, { "epoch": 1.2316614420062697, "grad_norm": 4.6500163078308105, "learning_rate": 7.85498565127031e-06, "loss": 1.3086, "step": 15716 }, { "epoch": 1.2317398119122256, "grad_norm": 3.306960105895996, "learning_rate": 7.854732495375163e-06, "loss": 1.5744, "step": 15717 }, { "epoch": 1.231818181818182, "grad_norm": 4.016134738922119, "learning_rate": 7.854479328622082e-06, "loss": 1.4058, "step": 15718 }, { "epoch": 1.2318965517241378, "grad_norm": 3.4114956855773926, "learning_rate": 7.854226151012027e-06, "loss": 1.0305, "step": 15719 }, { "epoch": 1.2319749216300941, "grad_norm": 3.1480422019958496, "learning_rate": 7.853972962545962e-06, "loss": 1.499, "step": 15720 }, { "epoch": 1.23205329153605, "grad_norm": 18.977279663085938, "learning_rate": 7.853719763224849e-06, "loss": 1.3733, "step": 15721 }, { "epoch": 1.2321316614420064, "grad_norm": 7.0251665115356445, "learning_rate": 7.853466553049652e-06, "loss": 1.7829, "step": 15722 }, { "epoch": 1.2322100313479625, "grad_norm": 2.905148983001709, "learning_rate": 7.853213332021335e-06, "loss": 1.4164, "step": 15723 }, { "epoch": 1.2322884012539186, "grad_norm": 6.255733013153076, "learning_rate": 7.852960100140858e-06, "loss": 1.0463, "step": 15724 }, { "epoch": 1.2323667711598747, "grad_norm": 2.0743675231933594, "learning_rate": 7.852706857409187e-06, "loss": 1.3215, "step": 15725 }, { "epoch": 1.2324451410658308, "grad_norm": 2.67737078666687, "learning_rate": 7.852453603827285e-06, "loss": 1.4378, "step": 15726 }, { "epoch": 1.232523510971787, "grad_norm": 3.383357048034668, "learning_rate": 7.852200339396115e-06, "loss": 1.339, "step": 15727 }, { "epoch": 1.232601880877743, "grad_norm": 3.6419572830200195, "learning_rate": 7.851947064116639e-06, "loss": 1.4116, "step": 15728 }, { "epoch": 1.2326802507836991, "grad_norm": 4.946371555328369, "learning_rate": 7.85169377798982e-06, "loss": 1.3063, "step": 15729 }, { "epoch": 1.2327586206896552, "grad_norm": 3.6314613819122314, "learning_rate": 7.851440481016623e-06, "loss": 1.2825, "step": 15730 }, { "epoch": 1.2328369905956114, "grad_norm": 2.0280070304870605, "learning_rate": 7.851187173198014e-06, "loss": 1.2376, "step": 15731 }, { "epoch": 1.2329153605015675, "grad_norm": 2.945047616958618, "learning_rate": 7.85093385453495e-06, "loss": 1.464, "step": 15732 }, { "epoch": 1.2329937304075236, "grad_norm": 2.200756311416626, "learning_rate": 7.8506805250284e-06, "loss": 0.9098, "step": 15733 }, { "epoch": 1.2330721003134797, "grad_norm": 2.2592647075653076, "learning_rate": 7.850427184679323e-06, "loss": 1.3709, "step": 15734 }, { "epoch": 1.2331504702194358, "grad_norm": 2.6237566471099854, "learning_rate": 7.850173833488688e-06, "loss": 1.4742, "step": 15735 }, { "epoch": 1.233228840125392, "grad_norm": 3.800201177597046, "learning_rate": 7.849920471457453e-06, "loss": 1.4778, "step": 15736 }, { "epoch": 1.233307210031348, "grad_norm": 3.942075490951538, "learning_rate": 7.849667098586587e-06, "loss": 1.5295, "step": 15737 }, { "epoch": 1.2333855799373041, "grad_norm": 7.772106647491455, "learning_rate": 7.84941371487705e-06, "loss": 1.9251, "step": 15738 }, { "epoch": 1.2334639498432602, "grad_norm": 3.581692695617676, "learning_rate": 7.849160320329807e-06, "loss": 2.0942, "step": 15739 }, { "epoch": 1.2335423197492164, "grad_norm": 2.6041247844696045, "learning_rate": 7.848906914945821e-06, "loss": 1.3594, "step": 15740 }, { "epoch": 1.2336206896551725, "grad_norm": 6.416703701019287, "learning_rate": 7.848653498726057e-06, "loss": 0.7912, "step": 15741 }, { "epoch": 1.2336990595611286, "grad_norm": 4.508624076843262, "learning_rate": 7.848400071671479e-06, "loss": 1.3673, "step": 15742 }, { "epoch": 1.2337774294670847, "grad_norm": 7.079256057739258, "learning_rate": 7.848146633783048e-06, "loss": 0.8936, "step": 15743 }, { "epoch": 1.2338557993730408, "grad_norm": 3.788968086242676, "learning_rate": 7.847893185061731e-06, "loss": 1.054, "step": 15744 }, { "epoch": 1.233934169278997, "grad_norm": 3.59912371635437, "learning_rate": 7.847639725508491e-06, "loss": 1.403, "step": 15745 }, { "epoch": 1.234012539184953, "grad_norm": 4.717572212219238, "learning_rate": 7.847386255124291e-06, "loss": 1.5787, "step": 15746 }, { "epoch": 1.2340909090909091, "grad_norm": 2.6874873638153076, "learning_rate": 7.847132773910098e-06, "loss": 1.1034, "step": 15747 }, { "epoch": 1.2341692789968652, "grad_norm": 2.2398288249969482, "learning_rate": 7.846879281866875e-06, "loss": 1.3505, "step": 15748 }, { "epoch": 1.2342476489028213, "grad_norm": 2.7964839935302734, "learning_rate": 7.846625778995582e-06, "loss": 1.2563, "step": 15749 }, { "epoch": 1.2343260188087775, "grad_norm": 10.667512893676758, "learning_rate": 7.84637226529719e-06, "loss": 1.5019, "step": 15750 }, { "epoch": 1.2344043887147336, "grad_norm": 3.0319745540618896, "learning_rate": 7.84611874077266e-06, "loss": 1.3933, "step": 15751 }, { "epoch": 1.2344827586206897, "grad_norm": 2.682670831680298, "learning_rate": 7.845865205422954e-06, "loss": 1.0881, "step": 15752 }, { "epoch": 1.2345611285266458, "grad_norm": 3.8512511253356934, "learning_rate": 7.84561165924904e-06, "loss": 1.058, "step": 15753 }, { "epoch": 1.234639498432602, "grad_norm": 3.010651111602783, "learning_rate": 7.84535810225188e-06, "loss": 1.4442, "step": 15754 }, { "epoch": 1.234717868338558, "grad_norm": 3.871445894241333, "learning_rate": 7.84510453443244e-06, "loss": 1.2711, "step": 15755 }, { "epoch": 1.2347962382445141, "grad_norm": 2.7246530055999756, "learning_rate": 7.844850955791683e-06, "loss": 1.2338, "step": 15756 }, { "epoch": 1.2348746081504702, "grad_norm": 5.049811363220215, "learning_rate": 7.844597366330575e-06, "loss": 2.0519, "step": 15757 }, { "epoch": 1.2349529780564263, "grad_norm": 3.394136428833008, "learning_rate": 7.844343766050078e-06, "loss": 1.0546, "step": 15758 }, { "epoch": 1.2350313479623825, "grad_norm": 3.512547254562378, "learning_rate": 7.84409015495116e-06, "loss": 1.6381, "step": 15759 }, { "epoch": 1.2351097178683386, "grad_norm": 3.438558578491211, "learning_rate": 7.843836533034784e-06, "loss": 1.5235, "step": 15760 }, { "epoch": 1.2351880877742947, "grad_norm": 4.099548816680908, "learning_rate": 7.843582900301914e-06, "loss": 1.1365, "step": 15761 }, { "epoch": 1.2352664576802508, "grad_norm": 2.5327742099761963, "learning_rate": 7.843329256753514e-06, "loss": 0.9083, "step": 15762 }, { "epoch": 1.235344827586207, "grad_norm": 3.902205228805542, "learning_rate": 7.843075602390551e-06, "loss": 0.9528, "step": 15763 }, { "epoch": 1.235423197492163, "grad_norm": 2.5631296634674072, "learning_rate": 7.842821937213988e-06, "loss": 1.3376, "step": 15764 }, { "epoch": 1.2355015673981191, "grad_norm": 3.5999839305877686, "learning_rate": 7.84256826122479e-06, "loss": 1.474, "step": 15765 }, { "epoch": 1.2355799373040752, "grad_norm": 2.7378697395324707, "learning_rate": 7.842314574423926e-06, "loss": 1.2756, "step": 15766 }, { "epoch": 1.2356583072100313, "grad_norm": 2.7139968872070312, "learning_rate": 7.842060876812354e-06, "loss": 1.429, "step": 15767 }, { "epoch": 1.2357366771159874, "grad_norm": 3.653942584991455, "learning_rate": 7.841807168391042e-06, "loss": 1.2384, "step": 15768 }, { "epoch": 1.2358150470219436, "grad_norm": 3.978834390640259, "learning_rate": 7.841553449160956e-06, "loss": 1.0296, "step": 15769 }, { "epoch": 1.2358934169278997, "grad_norm": 2.984682083129883, "learning_rate": 7.84129971912306e-06, "loss": 1.0599, "step": 15770 }, { "epoch": 1.2359717868338558, "grad_norm": 2.4276785850524902, "learning_rate": 7.84104597827832e-06, "loss": 1.2099, "step": 15771 }, { "epoch": 1.236050156739812, "grad_norm": 2.796221971511841, "learning_rate": 7.8407922266277e-06, "loss": 1.1557, "step": 15772 }, { "epoch": 1.236128526645768, "grad_norm": 4.279824733734131, "learning_rate": 7.840538464172166e-06, "loss": 1.676, "step": 15773 }, { "epoch": 1.236206896551724, "grad_norm": 5.6032867431640625, "learning_rate": 7.840284690912683e-06, "loss": 1.1261, "step": 15774 }, { "epoch": 1.2362852664576802, "grad_norm": 4.371493339538574, "learning_rate": 7.840030906850214e-06, "loss": 1.349, "step": 15775 }, { "epoch": 1.2363636363636363, "grad_norm": 2.9105043411254883, "learning_rate": 7.839777111985727e-06, "loss": 1.5308, "step": 15776 }, { "epoch": 1.2364420062695924, "grad_norm": 2.2866921424865723, "learning_rate": 7.839523306320187e-06, "loss": 1.3669, "step": 15777 }, { "epoch": 1.2365203761755486, "grad_norm": 4.839547634124756, "learning_rate": 7.839269489854559e-06, "loss": 0.9477, "step": 15778 }, { "epoch": 1.2365987460815047, "grad_norm": 3.1913790702819824, "learning_rate": 7.839015662589807e-06, "loss": 1.1946, "step": 15779 }, { "epoch": 1.2366771159874608, "grad_norm": 9.218789100646973, "learning_rate": 7.838761824526898e-06, "loss": 1.4537, "step": 15780 }, { "epoch": 1.2367554858934169, "grad_norm": 2.1970162391662598, "learning_rate": 7.838507975666798e-06, "loss": 1.3023, "step": 15781 }, { "epoch": 1.236833855799373, "grad_norm": 5.954183101654053, "learning_rate": 7.838254116010471e-06, "loss": 1.801, "step": 15782 }, { "epoch": 1.236912225705329, "grad_norm": 5.732160568237305, "learning_rate": 7.838000245558883e-06, "loss": 1.1158, "step": 15783 }, { "epoch": 1.2369905956112852, "grad_norm": 2.2476718425750732, "learning_rate": 7.837746364313e-06, "loss": 1.1727, "step": 15784 }, { "epoch": 1.2370689655172413, "grad_norm": 4.0937886238098145, "learning_rate": 7.837492472273786e-06, "loss": 1.4792, "step": 15785 }, { "epoch": 1.2371473354231974, "grad_norm": 2.34673810005188, "learning_rate": 7.837238569442211e-06, "loss": 1.2594, "step": 15786 }, { "epoch": 1.2372257053291535, "grad_norm": 3.2521910667419434, "learning_rate": 7.836984655819235e-06, "loss": 1.1001, "step": 15787 }, { "epoch": 1.2373040752351097, "grad_norm": 4.763226509094238, "learning_rate": 7.836730731405828e-06, "loss": 2.1799, "step": 15788 }, { "epoch": 1.2373824451410658, "grad_norm": 9.08338451385498, "learning_rate": 7.836476796202954e-06, "loss": 1.1525, "step": 15789 }, { "epoch": 1.2374608150470219, "grad_norm": 9.815740585327148, "learning_rate": 7.836222850211579e-06, "loss": 1.2125, "step": 15790 }, { "epoch": 1.237539184952978, "grad_norm": 3.3451828956604004, "learning_rate": 7.83596889343267e-06, "loss": 1.0637, "step": 15791 }, { "epoch": 1.237617554858934, "grad_norm": 2.5139458179473877, "learning_rate": 7.83571492586719e-06, "loss": 1.3421, "step": 15792 }, { "epoch": 1.2376959247648902, "grad_norm": 2.825547933578491, "learning_rate": 7.835460947516109e-06, "loss": 1.4545, "step": 15793 }, { "epoch": 1.2377742946708463, "grad_norm": 2.6167027950286865, "learning_rate": 7.83520695838039e-06, "loss": 1.3493, "step": 15794 }, { "epoch": 1.2378526645768024, "grad_norm": 2.246387243270874, "learning_rate": 7.834952958461e-06, "loss": 1.2706, "step": 15795 }, { "epoch": 1.2379310344827585, "grad_norm": 2.9955854415893555, "learning_rate": 7.834698947758904e-06, "loss": 1.4278, "step": 15796 }, { "epoch": 1.2380094043887147, "grad_norm": 5.204796314239502, "learning_rate": 7.834444926275071e-06, "loss": 1.2199, "step": 15797 }, { "epoch": 1.2380877742946708, "grad_norm": 2.8064417839050293, "learning_rate": 7.834190894010465e-06, "loss": 1.0006, "step": 15798 }, { "epoch": 1.2381661442006269, "grad_norm": 2.183457374572754, "learning_rate": 7.833936850966053e-06, "loss": 1.4015, "step": 15799 }, { "epoch": 1.238244514106583, "grad_norm": 3.5835514068603516, "learning_rate": 7.8336827971428e-06, "loss": 1.8128, "step": 15800 }, { "epoch": 1.238322884012539, "grad_norm": 4.8094987869262695, "learning_rate": 7.833428732541672e-06, "loss": 1.5019, "step": 15801 }, { "epoch": 1.2384012539184952, "grad_norm": 3.1691479682922363, "learning_rate": 7.83317465716364e-06, "loss": 1.4924, "step": 15802 }, { "epoch": 1.2384796238244513, "grad_norm": 7.607959747314453, "learning_rate": 7.832920571009664e-06, "loss": 1.317, "step": 15803 }, { "epoch": 1.2385579937304074, "grad_norm": 2.425865888595581, "learning_rate": 7.832666474080713e-06, "loss": 1.393, "step": 15804 }, { "epoch": 1.2386363636363638, "grad_norm": 2.661759376525879, "learning_rate": 7.832412366377755e-06, "loss": 1.1265, "step": 15805 }, { "epoch": 1.2387147335423196, "grad_norm": 3.087855815887451, "learning_rate": 7.832158247901754e-06, "loss": 1.5084, "step": 15806 }, { "epoch": 1.238793103448276, "grad_norm": 4.2854461669921875, "learning_rate": 7.831904118653678e-06, "loss": 0.9469, "step": 15807 }, { "epoch": 1.2388714733542319, "grad_norm": 2.4849791526794434, "learning_rate": 7.831649978634495e-06, "loss": 1.341, "step": 15808 }, { "epoch": 1.2389498432601882, "grad_norm": 2.7846765518188477, "learning_rate": 7.831395827845168e-06, "loss": 1.1762, "step": 15809 }, { "epoch": 1.239028213166144, "grad_norm": 2.563441514968872, "learning_rate": 7.831141666286666e-06, "loss": 1.3733, "step": 15810 }, { "epoch": 1.2391065830721004, "grad_norm": 3.160618782043457, "learning_rate": 7.830887493959956e-06, "loss": 1.5556, "step": 15811 }, { "epoch": 1.2391849529780563, "grad_norm": 3.2776358127593994, "learning_rate": 7.830633310866002e-06, "loss": 1.1271, "step": 15812 }, { "epoch": 1.2392633228840126, "grad_norm": 2.541457414627075, "learning_rate": 7.830379117005773e-06, "loss": 1.4297, "step": 15813 }, { "epoch": 1.2393416927899685, "grad_norm": 2.8442859649658203, "learning_rate": 7.830124912380237e-06, "loss": 1.2314, "step": 15814 }, { "epoch": 1.2394200626959249, "grad_norm": 4.301679611206055, "learning_rate": 7.82987069699036e-06, "loss": 1.1566, "step": 15815 }, { "epoch": 1.239498432601881, "grad_norm": 4.380280494689941, "learning_rate": 7.829616470837107e-06, "loss": 1.3396, "step": 15816 }, { "epoch": 1.239576802507837, "grad_norm": 3.4293336868286133, "learning_rate": 7.829362233921448e-06, "loss": 1.3763, "step": 15817 }, { "epoch": 1.2396551724137932, "grad_norm": 2.283017873764038, "learning_rate": 7.829107986244347e-06, "loss": 1.3385, "step": 15818 }, { "epoch": 1.2397335423197493, "grad_norm": 3.372995138168335, "learning_rate": 7.828853727806772e-06, "loss": 1.4021, "step": 15819 }, { "epoch": 1.2398119122257054, "grad_norm": 2.8190665245056152, "learning_rate": 7.828599458609691e-06, "loss": 1.4235, "step": 15820 }, { "epoch": 1.2398902821316615, "grad_norm": 3.166823625564575, "learning_rate": 7.82834517865407e-06, "loss": 1.3096, "step": 15821 }, { "epoch": 1.2399686520376176, "grad_norm": 2.0593488216400146, "learning_rate": 7.828090887940878e-06, "loss": 0.9017, "step": 15822 }, { "epoch": 1.2400470219435737, "grad_norm": 3.7648251056671143, "learning_rate": 7.827836586471081e-06, "loss": 1.5506, "step": 15823 }, { "epoch": 1.2401253918495299, "grad_norm": 3.3467776775360107, "learning_rate": 7.827582274245645e-06, "loss": 1.415, "step": 15824 }, { "epoch": 1.240203761755486, "grad_norm": 2.1291394233703613, "learning_rate": 7.827327951265538e-06, "loss": 1.4372, "step": 15825 }, { "epoch": 1.240282131661442, "grad_norm": 6.751909255981445, "learning_rate": 7.827073617531728e-06, "loss": 0.9985, "step": 15826 }, { "epoch": 1.2403605015673982, "grad_norm": 9.571041107177734, "learning_rate": 7.826819273045182e-06, "loss": 0.4909, "step": 15827 }, { "epoch": 1.2404388714733543, "grad_norm": 4.918368816375732, "learning_rate": 7.826564917806867e-06, "loss": 1.5337, "step": 15828 }, { "epoch": 1.2405172413793104, "grad_norm": 3.6417455673217773, "learning_rate": 7.826310551817752e-06, "loss": 1.4195, "step": 15829 }, { "epoch": 1.2405956112852665, "grad_norm": 2.0339343547821045, "learning_rate": 7.826056175078802e-06, "loss": 1.2314, "step": 15830 }, { "epoch": 1.2406739811912226, "grad_norm": 3.70736026763916, "learning_rate": 7.825801787590987e-06, "loss": 1.201, "step": 15831 }, { "epoch": 1.2407523510971787, "grad_norm": 3.952981948852539, "learning_rate": 7.825547389355274e-06, "loss": 1.5705, "step": 15832 }, { "epoch": 1.2408307210031349, "grad_norm": 2.931838035583496, "learning_rate": 7.825292980372627e-06, "loss": 1.3466, "step": 15833 }, { "epoch": 1.240909090909091, "grad_norm": 2.9439194202423096, "learning_rate": 7.825038560644021e-06, "loss": 1.4191, "step": 15834 }, { "epoch": 1.240987460815047, "grad_norm": 2.077453374862671, "learning_rate": 7.824784130170417e-06, "loss": 1.1146, "step": 15835 }, { "epoch": 1.2410658307210032, "grad_norm": 4.751809120178223, "learning_rate": 7.824529688952785e-06, "loss": 1.5533, "step": 15836 }, { "epoch": 1.2411442006269593, "grad_norm": 6.539947032928467, "learning_rate": 7.824275236992094e-06, "loss": 0.9676, "step": 15837 }, { "epoch": 1.2412225705329154, "grad_norm": 2.946887493133545, "learning_rate": 7.824020774289308e-06, "loss": 1.2552, "step": 15838 }, { "epoch": 1.2413009404388715, "grad_norm": 2.607780933380127, "learning_rate": 7.8237663008454e-06, "loss": 1.5475, "step": 15839 }, { "epoch": 1.2413793103448276, "grad_norm": 3.061596155166626, "learning_rate": 7.823511816661333e-06, "loss": 1.4082, "step": 15840 }, { "epoch": 1.2414576802507837, "grad_norm": 3.605194568634033, "learning_rate": 7.82325732173808e-06, "loss": 0.8773, "step": 15841 }, { "epoch": 1.2415360501567398, "grad_norm": 4.328125953674316, "learning_rate": 7.823002816076605e-06, "loss": 1.3245, "step": 15842 }, { "epoch": 1.241614420062696, "grad_norm": 2.398606300354004, "learning_rate": 7.822748299677878e-06, "loss": 1.3947, "step": 15843 }, { "epoch": 1.241692789968652, "grad_norm": 2.2190868854522705, "learning_rate": 7.822493772542865e-06, "loss": 1.1989, "step": 15844 }, { "epoch": 1.2417711598746082, "grad_norm": 1.99909245967865, "learning_rate": 7.822239234672537e-06, "loss": 1.429, "step": 15845 }, { "epoch": 1.2418495297805643, "grad_norm": 6.285212516784668, "learning_rate": 7.821984686067858e-06, "loss": 1.0951, "step": 15846 }, { "epoch": 1.2419278996865204, "grad_norm": 3.196688175201416, "learning_rate": 7.8217301267298e-06, "loss": 1.4193, "step": 15847 }, { "epoch": 1.2420062695924765, "grad_norm": 2.8574202060699463, "learning_rate": 7.821475556659332e-06, "loss": 1.7091, "step": 15848 }, { "epoch": 1.2420846394984326, "grad_norm": 3.2928626537323, "learning_rate": 7.82122097585742e-06, "loss": 1.4057, "step": 15849 }, { "epoch": 1.2421630094043887, "grad_norm": 1.925126314163208, "learning_rate": 7.82096638432503e-06, "loss": 1.1736, "step": 15850 }, { "epoch": 1.2422413793103448, "grad_norm": 4.173044204711914, "learning_rate": 7.820711782063133e-06, "loss": 2.2843, "step": 15851 }, { "epoch": 1.242319749216301, "grad_norm": 2.3180861473083496, "learning_rate": 7.820457169072699e-06, "loss": 1.2368, "step": 15852 }, { "epoch": 1.242398119122257, "grad_norm": 2.0915353298187256, "learning_rate": 7.820202545354693e-06, "loss": 1.2818, "step": 15853 }, { "epoch": 1.2424764890282132, "grad_norm": 2.7498934268951416, "learning_rate": 7.819947910910087e-06, "loss": 1.6609, "step": 15854 }, { "epoch": 1.2425548589341693, "grad_norm": 4.9641499519348145, "learning_rate": 7.819693265739845e-06, "loss": 1.2888, "step": 15855 }, { "epoch": 1.2426332288401254, "grad_norm": 3.6507973670959473, "learning_rate": 7.819438609844939e-06, "loss": 1.5228, "step": 15856 }, { "epoch": 1.2427115987460815, "grad_norm": 4.98912239074707, "learning_rate": 7.819183943226337e-06, "loss": 1.6035, "step": 15857 }, { "epoch": 1.2427899686520376, "grad_norm": 3.410982847213745, "learning_rate": 7.818929265885008e-06, "loss": 1.2792, "step": 15858 }, { "epoch": 1.2428683385579937, "grad_norm": 3.7405412197113037, "learning_rate": 7.818674577821918e-06, "loss": 1.6144, "step": 15859 }, { "epoch": 1.2429467084639498, "grad_norm": 3.0097596645355225, "learning_rate": 7.81841987903804e-06, "loss": 0.8535, "step": 15860 }, { "epoch": 1.243025078369906, "grad_norm": 7.0258355140686035, "learning_rate": 7.818165169534339e-06, "loss": 1.3852, "step": 15861 }, { "epoch": 1.243103448275862, "grad_norm": 7.826570510864258, "learning_rate": 7.817910449311785e-06, "loss": 1.8368, "step": 15862 }, { "epoch": 1.2431818181818182, "grad_norm": 4.018733978271484, "learning_rate": 7.817655718371348e-06, "loss": 1.7102, "step": 15863 }, { "epoch": 1.2432601880877743, "grad_norm": 4.803266525268555, "learning_rate": 7.817400976713995e-06, "loss": 1.3374, "step": 15864 }, { "epoch": 1.2433385579937304, "grad_norm": 2.3066375255584717, "learning_rate": 7.817146224340697e-06, "loss": 1.376, "step": 15865 }, { "epoch": 1.2434169278996865, "grad_norm": 4.28411865234375, "learning_rate": 7.81689146125242e-06, "loss": 1.3211, "step": 15866 }, { "epoch": 1.2434952978056426, "grad_norm": 2.4582273960113525, "learning_rate": 7.816636687450137e-06, "loss": 1.2354, "step": 15867 }, { "epoch": 1.2435736677115987, "grad_norm": 4.743219375610352, "learning_rate": 7.816381902934813e-06, "loss": 1.5472, "step": 15868 }, { "epoch": 1.2436520376175548, "grad_norm": 4.469650745391846, "learning_rate": 7.816127107707419e-06, "loss": 1.7036, "step": 15869 }, { "epoch": 1.243730407523511, "grad_norm": 2.803212881088257, "learning_rate": 7.815872301768924e-06, "loss": 1.3816, "step": 15870 }, { "epoch": 1.243808777429467, "grad_norm": 2.559993267059326, "learning_rate": 7.815617485120299e-06, "loss": 1.2772, "step": 15871 }, { "epoch": 1.2438871473354232, "grad_norm": 4.460194110870361, "learning_rate": 7.81536265776251e-06, "loss": 1.2851, "step": 15872 }, { "epoch": 1.2439655172413793, "grad_norm": 2.8966312408447266, "learning_rate": 7.815107819696526e-06, "loss": 1.4165, "step": 15873 }, { "epoch": 1.2440438871473354, "grad_norm": 3.121995687484741, "learning_rate": 7.814852970923317e-06, "loss": 1.2662, "step": 15874 }, { "epoch": 1.2441222570532915, "grad_norm": 2.335545301437378, "learning_rate": 7.814598111443856e-06, "loss": 1.2191, "step": 15875 }, { "epoch": 1.2442006269592476, "grad_norm": 2.6923282146453857, "learning_rate": 7.814343241259109e-06, "loss": 1.6157, "step": 15876 }, { "epoch": 1.2442789968652037, "grad_norm": 3.353889226913452, "learning_rate": 7.814088360370046e-06, "loss": 1.4131, "step": 15877 }, { "epoch": 1.2443573667711598, "grad_norm": 2.6391091346740723, "learning_rate": 7.813833468777635e-06, "loss": 1.2967, "step": 15878 }, { "epoch": 1.244435736677116, "grad_norm": 2.673123359680176, "learning_rate": 7.813578566482847e-06, "loss": 1.3806, "step": 15879 }, { "epoch": 1.244514106583072, "grad_norm": 3.449328899383545, "learning_rate": 7.813323653486654e-06, "loss": 1.1451, "step": 15880 }, { "epoch": 1.2445924764890282, "grad_norm": 4.993951320648193, "learning_rate": 7.813068729790018e-06, "loss": 1.5528, "step": 15881 }, { "epoch": 1.2446708463949843, "grad_norm": 2.6681981086730957, "learning_rate": 7.812813795393917e-06, "loss": 1.6176, "step": 15882 }, { "epoch": 1.2447492163009404, "grad_norm": 2.8115732669830322, "learning_rate": 7.812558850299315e-06, "loss": 1.3884, "step": 15883 }, { "epoch": 1.2448275862068965, "grad_norm": 2.7102885246276855, "learning_rate": 7.812303894507185e-06, "loss": 1.5093, "step": 15884 }, { "epoch": 1.2449059561128526, "grad_norm": 2.9168708324432373, "learning_rate": 7.812048928018494e-06, "loss": 1.3676, "step": 15885 }, { "epoch": 1.2449843260188087, "grad_norm": 2.3789830207824707, "learning_rate": 7.811793950834215e-06, "loss": 1.1172, "step": 15886 }, { "epoch": 1.2450626959247648, "grad_norm": 2.630687713623047, "learning_rate": 7.811538962955317e-06, "loss": 1.2128, "step": 15887 }, { "epoch": 1.245141065830721, "grad_norm": 2.969374895095825, "learning_rate": 7.811283964382766e-06, "loss": 1.6457, "step": 15888 }, { "epoch": 1.245219435736677, "grad_norm": 3.9890012741088867, "learning_rate": 7.811028955117536e-06, "loss": 1.606, "step": 15889 }, { "epoch": 1.2452978056426331, "grad_norm": 2.620835065841675, "learning_rate": 7.810773935160596e-06, "loss": 1.0178, "step": 15890 }, { "epoch": 1.2453761755485893, "grad_norm": 2.6218247413635254, "learning_rate": 7.810518904512917e-06, "loss": 1.401, "step": 15891 }, { "epoch": 1.2454545454545454, "grad_norm": 3.573686361312866, "learning_rate": 7.810263863175465e-06, "loss": 1.237, "step": 15892 }, { "epoch": 1.2455329153605015, "grad_norm": 4.529457092285156, "learning_rate": 7.810008811149215e-06, "loss": 1.1144, "step": 15893 }, { "epoch": 1.2456112852664576, "grad_norm": 3.11087965965271, "learning_rate": 7.809753748435133e-06, "loss": 1.2901, "step": 15894 }, { "epoch": 1.2456896551724137, "grad_norm": 2.875356912612915, "learning_rate": 7.809498675034191e-06, "loss": 1.3889, "step": 15895 }, { "epoch": 1.2457680250783698, "grad_norm": 2.6543445587158203, "learning_rate": 7.80924359094736e-06, "loss": 1.0597, "step": 15896 }, { "epoch": 1.245846394984326, "grad_norm": 3.079808473587036, "learning_rate": 7.808988496175608e-06, "loss": 1.3022, "step": 15897 }, { "epoch": 1.2459247648902823, "grad_norm": 4.037564754486084, "learning_rate": 7.808733390719907e-06, "loss": 1.0608, "step": 15898 }, { "epoch": 1.2460031347962381, "grad_norm": 2.701441764831543, "learning_rate": 7.808478274581229e-06, "loss": 1.5873, "step": 15899 }, { "epoch": 1.2460815047021945, "grad_norm": 2.5032644271850586, "learning_rate": 7.80822314776054e-06, "loss": 1.2347, "step": 15900 }, { "epoch": 1.2461598746081504, "grad_norm": 9.726851463317871, "learning_rate": 7.807968010258814e-06, "loss": 1.2897, "step": 15901 }, { "epoch": 1.2462382445141067, "grad_norm": 2.8845038414001465, "learning_rate": 7.807712862077017e-06, "loss": 1.2567, "step": 15902 }, { "epoch": 1.2463166144200626, "grad_norm": 5.253558158874512, "learning_rate": 7.807457703216126e-06, "loss": 1.4507, "step": 15903 }, { "epoch": 1.246394984326019, "grad_norm": 4.052727699279785, "learning_rate": 7.807202533677107e-06, "loss": 1.0745, "step": 15904 }, { "epoch": 1.2464733542319748, "grad_norm": 2.170985221862793, "learning_rate": 7.80694735346093e-06, "loss": 1.3555, "step": 15905 }, { "epoch": 1.2465517241379311, "grad_norm": 5.790942192077637, "learning_rate": 7.806692162568568e-06, "loss": 0.978, "step": 15906 }, { "epoch": 1.246630094043887, "grad_norm": 3.8137121200561523, "learning_rate": 7.806436961000989e-06, "loss": 1.0011, "step": 15907 }, { "epoch": 1.2467084639498434, "grad_norm": 2.0627875328063965, "learning_rate": 7.806181748759168e-06, "loss": 1.2191, "step": 15908 }, { "epoch": 1.2467868338557995, "grad_norm": 3.1817452907562256, "learning_rate": 7.805926525844071e-06, "loss": 1.3675, "step": 15909 }, { "epoch": 1.2468652037617556, "grad_norm": 2.7874410152435303, "learning_rate": 7.805671292256671e-06, "loss": 1.3519, "step": 15910 }, { "epoch": 1.2469435736677117, "grad_norm": 2.1249005794525146, "learning_rate": 7.805416047997939e-06, "loss": 1.2646, "step": 15911 }, { "epoch": 1.2470219435736678, "grad_norm": 3.111213207244873, "learning_rate": 7.805160793068846e-06, "loss": 1.1241, "step": 15912 }, { "epoch": 1.247100313479624, "grad_norm": 23.721899032592773, "learning_rate": 7.804905527470361e-06, "loss": 2.2892, "step": 15913 }, { "epoch": 1.24717868338558, "grad_norm": 3.153485059738159, "learning_rate": 7.804650251203457e-06, "loss": 1.6356, "step": 15914 }, { "epoch": 1.2472570532915361, "grad_norm": 2.708004951477051, "learning_rate": 7.804394964269102e-06, "loss": 1.183, "step": 15915 }, { "epoch": 1.2473354231974922, "grad_norm": 4.213184833526611, "learning_rate": 7.804139666668271e-06, "loss": 1.524, "step": 15916 }, { "epoch": 1.2474137931034484, "grad_norm": 16.6431827545166, "learning_rate": 7.803884358401932e-06, "loss": 2.518, "step": 15917 }, { "epoch": 1.2474921630094045, "grad_norm": 3.0737340450286865, "learning_rate": 7.803629039471056e-06, "loss": 1.2175, "step": 15918 }, { "epoch": 1.2475705329153606, "grad_norm": 3.251558542251587, "learning_rate": 7.803373709876619e-06, "loss": 1.4617, "step": 15919 }, { "epoch": 1.2476489028213167, "grad_norm": 7.295462131500244, "learning_rate": 7.803118369619585e-06, "loss": 1.4096, "step": 15920 }, { "epoch": 1.2477272727272728, "grad_norm": 4.181959629058838, "learning_rate": 7.80286301870093e-06, "loss": 1.5246, "step": 15921 }, { "epoch": 1.247805642633229, "grad_norm": 2.704882860183716, "learning_rate": 7.802607657121621e-06, "loss": 1.6804, "step": 15922 }, { "epoch": 1.247884012539185, "grad_norm": 2.5870707035064697, "learning_rate": 7.802352284882636e-06, "loss": 1.1959, "step": 15923 }, { "epoch": 1.2479623824451411, "grad_norm": 3.005282402038574, "learning_rate": 7.802096901984938e-06, "loss": 1.1665, "step": 15924 }, { "epoch": 1.2480407523510972, "grad_norm": 3.2931737899780273, "learning_rate": 7.801841508429508e-06, "loss": 1.2941, "step": 15925 }, { "epoch": 1.2481191222570533, "grad_norm": 2.87756609916687, "learning_rate": 7.801586104217308e-06, "loss": 0.6071, "step": 15926 }, { "epoch": 1.2481974921630095, "grad_norm": 3.4988856315612793, "learning_rate": 7.801330689349314e-06, "loss": 1.8063, "step": 15927 }, { "epoch": 1.2482758620689656, "grad_norm": 8.179986953735352, "learning_rate": 7.801075263826498e-06, "loss": 1.0864, "step": 15928 }, { "epoch": 1.2483542319749217, "grad_norm": 2.475790023803711, "learning_rate": 7.80081982764983e-06, "loss": 1.3762, "step": 15929 }, { "epoch": 1.2484326018808778, "grad_norm": 3.788313627243042, "learning_rate": 7.800564380820281e-06, "loss": 1.3449, "step": 15930 }, { "epoch": 1.248510971786834, "grad_norm": 3.104844331741333, "learning_rate": 7.800308923338824e-06, "loss": 1.6889, "step": 15931 }, { "epoch": 1.24858934169279, "grad_norm": 2.1642098426818848, "learning_rate": 7.80005345520643e-06, "loss": 1.2097, "step": 15932 }, { "epoch": 1.2486677115987461, "grad_norm": 12.031868934631348, "learning_rate": 7.799797976424074e-06, "loss": 2.5855, "step": 15933 }, { "epoch": 1.2487460815047022, "grad_norm": 4.608710765838623, "learning_rate": 7.799542486992722e-06, "loss": 1.0469, "step": 15934 }, { "epoch": 1.2488244514106583, "grad_norm": 2.5421364307403564, "learning_rate": 7.799286986913345e-06, "loss": 0.9328, "step": 15935 }, { "epoch": 1.2489028213166145, "grad_norm": 3.679426670074463, "learning_rate": 7.799031476186923e-06, "loss": 1.5347, "step": 15936 }, { "epoch": 1.2489811912225706, "grad_norm": 5.780064105987549, "learning_rate": 7.798775954814421e-06, "loss": 1.8031, "step": 15937 }, { "epoch": 1.2490595611285267, "grad_norm": 4.168468952178955, "learning_rate": 7.798520422796814e-06, "loss": 1.5445, "step": 15938 }, { "epoch": 1.2491379310344828, "grad_norm": 2.6512227058410645, "learning_rate": 7.798264880135072e-06, "loss": 1.5675, "step": 15939 }, { "epoch": 1.249216300940439, "grad_norm": 5.6301703453063965, "learning_rate": 7.798009326830167e-06, "loss": 1.1872, "step": 15940 }, { "epoch": 1.249294670846395, "grad_norm": 3.627654552459717, "learning_rate": 7.797753762883072e-06, "loss": 1.7393, "step": 15941 }, { "epoch": 1.2493730407523511, "grad_norm": 3.5730183124542236, "learning_rate": 7.797498188294759e-06, "loss": 1.4769, "step": 15942 }, { "epoch": 1.2494514106583072, "grad_norm": 3.596623659133911, "learning_rate": 7.7972426030662e-06, "loss": 1.6743, "step": 15943 }, { "epoch": 1.2495297805642633, "grad_norm": 3.8103926181793213, "learning_rate": 7.796987007198367e-06, "loss": 1.3396, "step": 15944 }, { "epoch": 1.2496081504702194, "grad_norm": 4.321627140045166, "learning_rate": 7.79673140069223e-06, "loss": 1.478, "step": 15945 }, { "epoch": 1.2496865203761756, "grad_norm": 2.5808558464050293, "learning_rate": 7.796475783548765e-06, "loss": 1.3588, "step": 15946 }, { "epoch": 1.2497648902821317, "grad_norm": 2.6774303913116455, "learning_rate": 7.796220155768944e-06, "loss": 1.558, "step": 15947 }, { "epoch": 1.2498432601880878, "grad_norm": 1.888143539428711, "learning_rate": 7.795964517353734e-06, "loss": 1.3229, "step": 15948 }, { "epoch": 1.249921630094044, "grad_norm": 3.2043583393096924, "learning_rate": 7.795708868304115e-06, "loss": 1.24, "step": 15949 }, { "epoch": 1.25, "grad_norm": 2.763932466506958, "learning_rate": 7.795453208621054e-06, "loss": 1.6017, "step": 15950 }, { "epoch": 1.25, "eval_loss": 1.2916802167892456, "eval_runtime": 108.5406, "eval_samples_per_second": 24.756, "eval_steps_per_second": 6.191, "step": 15950 }, { "epoch": 1.250078369905956, "grad_norm": 3.2386701107025146, "learning_rate": 7.795197538305523e-06, "loss": 1.2794, "step": 15951 }, { "epoch": 1.2501567398119122, "grad_norm": 2.8717639446258545, "learning_rate": 7.794941857358498e-06, "loss": 1.4892, "step": 15952 }, { "epoch": 1.2502351097178683, "grad_norm": 2.9362406730651855, "learning_rate": 7.79468616578095e-06, "loss": 1.3355, "step": 15953 }, { "epoch": 1.2503134796238244, "grad_norm": 6.063574314117432, "learning_rate": 7.794430463573849e-06, "loss": 1.8308, "step": 15954 }, { "epoch": 1.2503918495297806, "grad_norm": 3.869192600250244, "learning_rate": 7.794174750738172e-06, "loss": 0.948, "step": 15955 }, { "epoch": 1.2504702194357367, "grad_norm": 2.3569393157958984, "learning_rate": 7.79391902727489e-06, "loss": 1.5377, "step": 15956 }, { "epoch": 1.2505485893416928, "grad_norm": 3.547574281692505, "learning_rate": 7.793663293184974e-06, "loss": 0.9552, "step": 15957 }, { "epoch": 1.2506269592476489, "grad_norm": 9.222808837890625, "learning_rate": 7.793407548469396e-06, "loss": 2.0683, "step": 15958 }, { "epoch": 1.250705329153605, "grad_norm": 3.3834893703460693, "learning_rate": 7.793151793129134e-06, "loss": 1.4162, "step": 15959 }, { "epoch": 1.250783699059561, "grad_norm": 3.5683844089508057, "learning_rate": 7.792896027165155e-06, "loss": 0.5779, "step": 15960 }, { "epoch": 1.2508620689655172, "grad_norm": 3.5113792419433594, "learning_rate": 7.792640250578434e-06, "loss": 1.8985, "step": 15961 }, { "epoch": 1.2509404388714733, "grad_norm": 2.1437532901763916, "learning_rate": 7.792384463369944e-06, "loss": 1.2299, "step": 15962 }, { "epoch": 1.2510188087774294, "grad_norm": 3.7115137577056885, "learning_rate": 7.792128665540659e-06, "loss": 1.8599, "step": 15963 }, { "epoch": 1.2510971786833855, "grad_norm": 3.13385009765625, "learning_rate": 7.791872857091549e-06, "loss": 1.3845, "step": 15964 }, { "epoch": 1.2511755485893417, "grad_norm": 2.8956992626190186, "learning_rate": 7.79161703802359e-06, "loss": 1.6812, "step": 15965 }, { "epoch": 1.2512539184952978, "grad_norm": 4.480795860290527, "learning_rate": 7.791361208337753e-06, "loss": 1.4899, "step": 15966 }, { "epoch": 1.2513322884012539, "grad_norm": 3.016638994216919, "learning_rate": 7.791105368035012e-06, "loss": 1.2818, "step": 15967 }, { "epoch": 1.25141065830721, "grad_norm": 7.392183303833008, "learning_rate": 7.790849517116339e-06, "loss": 0.8218, "step": 15968 }, { "epoch": 1.251489028213166, "grad_norm": 2.1988019943237305, "learning_rate": 7.79059365558271e-06, "loss": 1.3472, "step": 15969 }, { "epoch": 1.2515673981191222, "grad_norm": 3.5634331703186035, "learning_rate": 7.790337783435093e-06, "loss": 1.8394, "step": 15970 }, { "epoch": 1.2516457680250783, "grad_norm": 2.2201690673828125, "learning_rate": 7.790081900674466e-06, "loss": 1.5185, "step": 15971 }, { "epoch": 1.2517241379310344, "grad_norm": 3.8660483360290527, "learning_rate": 7.789826007301802e-06, "loss": 1.2758, "step": 15972 }, { "epoch": 1.2518025078369905, "grad_norm": 2.521399974822998, "learning_rate": 7.789570103318069e-06, "loss": 1.1286, "step": 15973 }, { "epoch": 1.2518808777429467, "grad_norm": 3.5623960494995117, "learning_rate": 7.789314188724247e-06, "loss": 1.0987, "step": 15974 }, { "epoch": 1.2519592476489028, "grad_norm": 3.402193784713745, "learning_rate": 7.789058263521308e-06, "loss": 1.1821, "step": 15975 }, { "epoch": 1.2520376175548589, "grad_norm": 2.601938486099243, "learning_rate": 7.78880232771022e-06, "loss": 1.049, "step": 15976 }, { "epoch": 1.252115987460815, "grad_norm": 4.0117363929748535, "learning_rate": 7.788546381291963e-06, "loss": 1.6058, "step": 15977 }, { "epoch": 1.252194357366771, "grad_norm": 3.476515531539917, "learning_rate": 7.788290424267506e-06, "loss": 1.4144, "step": 15978 }, { "epoch": 1.2522727272727272, "grad_norm": 2.68646240234375, "learning_rate": 7.788034456637825e-06, "loss": 1.2091, "step": 15979 }, { "epoch": 1.2523510971786833, "grad_norm": 2.31000018119812, "learning_rate": 7.787778478403894e-06, "loss": 1.1927, "step": 15980 }, { "epoch": 1.2524294670846394, "grad_norm": 3.290473222732544, "learning_rate": 7.787522489566686e-06, "loss": 1.3971, "step": 15981 }, { "epoch": 1.2525078369905955, "grad_norm": 3.442716121673584, "learning_rate": 7.787266490127173e-06, "loss": 0.7405, "step": 15982 }, { "epoch": 1.2525862068965516, "grad_norm": 1.9925695657730103, "learning_rate": 7.787010480086329e-06, "loss": 1.7158, "step": 15983 }, { "epoch": 1.2526645768025078, "grad_norm": 2.140167236328125, "learning_rate": 7.78675445944513e-06, "loss": 1.289, "step": 15984 }, { "epoch": 1.2527429467084639, "grad_norm": 1.9530795812606812, "learning_rate": 7.786498428204547e-06, "loss": 1.5364, "step": 15985 }, { "epoch": 1.25282131661442, "grad_norm": 2.4892544746398926, "learning_rate": 7.786242386365556e-06, "loss": 1.4582, "step": 15986 }, { "epoch": 1.2528996865203763, "grad_norm": 3.321009635925293, "learning_rate": 7.785986333929132e-06, "loss": 1.6872, "step": 15987 }, { "epoch": 1.2529780564263322, "grad_norm": 3.4687907695770264, "learning_rate": 7.785730270896245e-06, "loss": 1.0328, "step": 15988 }, { "epoch": 1.2530564263322885, "grad_norm": 5.0674357414245605, "learning_rate": 7.78547419726787e-06, "loss": 1.2834, "step": 15989 }, { "epoch": 1.2531347962382444, "grad_norm": 2.0780770778656006, "learning_rate": 7.785218113044984e-06, "loss": 1.0488, "step": 15990 }, { "epoch": 1.2532131661442008, "grad_norm": 3.0219593048095703, "learning_rate": 7.784962018228558e-06, "loss": 1.5068, "step": 15991 }, { "epoch": 1.2532915360501566, "grad_norm": 2.183161497116089, "learning_rate": 7.784705912819568e-06, "loss": 1.1899, "step": 15992 }, { "epoch": 1.253369905956113, "grad_norm": 3.755746603012085, "learning_rate": 7.784449796818984e-06, "loss": 1.5382, "step": 15993 }, { "epoch": 1.2534482758620689, "grad_norm": 2.515740156173706, "learning_rate": 7.784193670227786e-06, "loss": 1.2801, "step": 15994 }, { "epoch": 1.2535266457680252, "grad_norm": 4.188596248626709, "learning_rate": 7.783937533046944e-06, "loss": 1.7863, "step": 15995 }, { "epoch": 1.253605015673981, "grad_norm": 5.348414421081543, "learning_rate": 7.783681385277433e-06, "loss": 1.1586, "step": 15996 }, { "epoch": 1.2536833855799374, "grad_norm": 10.929865837097168, "learning_rate": 7.783425226920229e-06, "loss": 1.2613, "step": 15997 }, { "epoch": 1.2537617554858933, "grad_norm": 3.494363307952881, "learning_rate": 7.783169057976306e-06, "loss": 1.6184, "step": 15998 }, { "epoch": 1.2538401253918496, "grad_norm": 4.7393012046813965, "learning_rate": 7.782912878446635e-06, "loss": 1.2016, "step": 15999 }, { "epoch": 1.2539184952978055, "grad_norm": 2.20767879486084, "learning_rate": 7.782656688332194e-06, "loss": 1.2461, "step": 16000 }, { "epoch": 1.2539968652037619, "grad_norm": 2.5729570388793945, "learning_rate": 7.782400487633956e-06, "loss": 1.2937, "step": 16001 }, { "epoch": 1.2540752351097177, "grad_norm": 2.062063217163086, "learning_rate": 7.782144276352896e-06, "loss": 1.2136, "step": 16002 }, { "epoch": 1.254153605015674, "grad_norm": 3.270663261413574, "learning_rate": 7.781888054489988e-06, "loss": 1.0569, "step": 16003 }, { "epoch": 1.25423197492163, "grad_norm": 3.5654423236846924, "learning_rate": 7.781631822046207e-06, "loss": 1.6897, "step": 16004 }, { "epoch": 1.2543103448275863, "grad_norm": 13.042304992675781, "learning_rate": 7.781375579022527e-06, "loss": 2.1856, "step": 16005 }, { "epoch": 1.2543887147335422, "grad_norm": 4.981933116912842, "learning_rate": 7.781119325419924e-06, "loss": 1.9278, "step": 16006 }, { "epoch": 1.2544670846394985, "grad_norm": 2.5273754596710205, "learning_rate": 7.78086306123937e-06, "loss": 1.2984, "step": 16007 }, { "epoch": 1.2545454545454544, "grad_norm": 4.050480365753174, "learning_rate": 7.780606786481841e-06, "loss": 1.9415, "step": 16008 }, { "epoch": 1.2546238244514107, "grad_norm": 4.554530143737793, "learning_rate": 7.780350501148313e-06, "loss": 1.3296, "step": 16009 }, { "epoch": 1.2547021943573669, "grad_norm": 12.021025657653809, "learning_rate": 7.78009420523976e-06, "loss": 2.1838, "step": 16010 }, { "epoch": 1.254780564263323, "grad_norm": 5.29920768737793, "learning_rate": 7.779837898757157e-06, "loss": 1.6395, "step": 16011 }, { "epoch": 1.254858934169279, "grad_norm": 2.6395974159240723, "learning_rate": 7.779581581701476e-06, "loss": 1.5868, "step": 16012 }, { "epoch": 1.2549373040752352, "grad_norm": 5.493203163146973, "learning_rate": 7.779325254073697e-06, "loss": 1.0809, "step": 16013 }, { "epoch": 1.2550156739811913, "grad_norm": 3.8113176822662354, "learning_rate": 7.77906891587479e-06, "loss": 1.1215, "step": 16014 }, { "epoch": 1.2550940438871474, "grad_norm": 3.4091978073120117, "learning_rate": 7.778812567105735e-06, "loss": 1.2462, "step": 16015 }, { "epoch": 1.2551724137931035, "grad_norm": 3.0803282260894775, "learning_rate": 7.778556207767502e-06, "loss": 1.4706, "step": 16016 }, { "epoch": 1.2552507836990596, "grad_norm": 3.3596224784851074, "learning_rate": 7.77829983786107e-06, "loss": 1.2883, "step": 16017 }, { "epoch": 1.2553291536050157, "grad_norm": 3.936126232147217, "learning_rate": 7.77804345738741e-06, "loss": 1.4168, "step": 16018 }, { "epoch": 1.2554075235109718, "grad_norm": 13.793538093566895, "learning_rate": 7.777787066347503e-06, "loss": 2.5824, "step": 16019 }, { "epoch": 1.255485893416928, "grad_norm": 5.128481864929199, "learning_rate": 7.77753066474232e-06, "loss": 1.3057, "step": 16020 }, { "epoch": 1.255564263322884, "grad_norm": 7.158870697021484, "learning_rate": 7.777274252572833e-06, "loss": 1.9566, "step": 16021 }, { "epoch": 1.2556426332288402, "grad_norm": 4.772810935974121, "learning_rate": 7.777017829840025e-06, "loss": 1.9747, "step": 16022 }, { "epoch": 1.2557210031347963, "grad_norm": 2.500332832336426, "learning_rate": 7.776761396544866e-06, "loss": 1.444, "step": 16023 }, { "epoch": 1.2557993730407524, "grad_norm": 3.300774574279785, "learning_rate": 7.776504952688334e-06, "loss": 1.7371, "step": 16024 }, { "epoch": 1.2558777429467085, "grad_norm": 1.7233704328536987, "learning_rate": 7.776248498271402e-06, "loss": 1.4189, "step": 16025 }, { "epoch": 1.2559561128526646, "grad_norm": 3.718132495880127, "learning_rate": 7.775992033295049e-06, "loss": 1.4563, "step": 16026 }, { "epoch": 1.2560344827586207, "grad_norm": 2.3930792808532715, "learning_rate": 7.775735557760245e-06, "loss": 1.3913, "step": 16027 }, { "epoch": 1.2561128526645768, "grad_norm": 2.7341649532318115, "learning_rate": 7.77547907166797e-06, "loss": 1.6792, "step": 16028 }, { "epoch": 1.256191222570533, "grad_norm": 3.807126760482788, "learning_rate": 7.775222575019199e-06, "loss": 0.7856, "step": 16029 }, { "epoch": 1.256269592476489, "grad_norm": 6.554472923278809, "learning_rate": 7.774966067814906e-06, "loss": 1.046, "step": 16030 }, { "epoch": 1.2563479623824452, "grad_norm": 3.0110881328582764, "learning_rate": 7.774709550056069e-06, "loss": 1.3927, "step": 16031 }, { "epoch": 1.2564263322884013, "grad_norm": 3.557230234146118, "learning_rate": 7.77445302174366e-06, "loss": 0.9134, "step": 16032 }, { "epoch": 1.2565047021943574, "grad_norm": 3.035546064376831, "learning_rate": 7.774196482878658e-06, "loss": 1.3933, "step": 16033 }, { "epoch": 1.2565830721003135, "grad_norm": 3.4131572246551514, "learning_rate": 7.773939933462036e-06, "loss": 1.6678, "step": 16034 }, { "epoch": 1.2566614420062696, "grad_norm": 9.176066398620605, "learning_rate": 7.773683373494773e-06, "loss": 1.0126, "step": 16035 }, { "epoch": 1.2567398119122257, "grad_norm": 2.3212544918060303, "learning_rate": 7.77342680297784e-06, "loss": 1.3675, "step": 16036 }, { "epoch": 1.2568181818181818, "grad_norm": 5.356493949890137, "learning_rate": 7.77317022191222e-06, "loss": 1.6473, "step": 16037 }, { "epoch": 1.256896551724138, "grad_norm": 2.4633119106292725, "learning_rate": 7.772913630298883e-06, "loss": 1.6166, "step": 16038 }, { "epoch": 1.256974921630094, "grad_norm": 3.1818103790283203, "learning_rate": 7.772657028138809e-06, "loss": 1.301, "step": 16039 }, { "epoch": 1.2570532915360502, "grad_norm": 2.075490713119507, "learning_rate": 7.772400415432967e-06, "loss": 1.5335, "step": 16040 }, { "epoch": 1.2571316614420063, "grad_norm": 2.015442132949829, "learning_rate": 7.772143792182343e-06, "loss": 1.4652, "step": 16041 }, { "epoch": 1.2572100313479624, "grad_norm": 3.85931658744812, "learning_rate": 7.771887158387904e-06, "loss": 1.6957, "step": 16042 }, { "epoch": 1.2572884012539185, "grad_norm": 2.113445281982422, "learning_rate": 7.771630514050632e-06, "loss": 1.375, "step": 16043 }, { "epoch": 1.2573667711598746, "grad_norm": 4.303491115570068, "learning_rate": 7.771373859171501e-06, "loss": 1.3321, "step": 16044 }, { "epoch": 1.2574451410658307, "grad_norm": 3.36628794670105, "learning_rate": 7.771117193751488e-06, "loss": 1.6407, "step": 16045 }, { "epoch": 1.2575235109717868, "grad_norm": 2.8863472938537598, "learning_rate": 7.770860517791567e-06, "loss": 1.227, "step": 16046 }, { "epoch": 1.257601880877743, "grad_norm": 2.6050515174865723, "learning_rate": 7.770603831292715e-06, "loss": 1.3515, "step": 16047 }, { "epoch": 1.257680250783699, "grad_norm": 2.581249237060547, "learning_rate": 7.77034713425591e-06, "loss": 1.3869, "step": 16048 }, { "epoch": 1.2577586206896552, "grad_norm": 3.864982843399048, "learning_rate": 7.770090426682127e-06, "loss": 1.8381, "step": 16049 }, { "epoch": 1.2578369905956113, "grad_norm": 3.178222179412842, "learning_rate": 7.769833708572346e-06, "loss": 1.5955, "step": 16050 }, { "epoch": 1.2579153605015674, "grad_norm": 4.651072978973389, "learning_rate": 7.769576979927536e-06, "loss": 1.3212, "step": 16051 }, { "epoch": 1.2579937304075235, "grad_norm": 4.8650994300842285, "learning_rate": 7.76932024074868e-06, "loss": 1.3295, "step": 16052 }, { "epoch": 1.2580721003134796, "grad_norm": 2.0822255611419678, "learning_rate": 7.769063491036752e-06, "loss": 1.3422, "step": 16053 }, { "epoch": 1.2581504702194357, "grad_norm": 1.9157103300094604, "learning_rate": 7.768806730792728e-06, "loss": 1.3424, "step": 16054 }, { "epoch": 1.2582288401253918, "grad_norm": 2.68229079246521, "learning_rate": 7.768549960017585e-06, "loss": 1.1454, "step": 16055 }, { "epoch": 1.258307210031348, "grad_norm": 4.652844429016113, "learning_rate": 7.768293178712302e-06, "loss": 1.8408, "step": 16056 }, { "epoch": 1.258385579937304, "grad_norm": 3.1989169120788574, "learning_rate": 7.768036386877853e-06, "loss": 1.1028, "step": 16057 }, { "epoch": 1.2584639498432602, "grad_norm": 2.3614439964294434, "learning_rate": 7.767779584515214e-06, "loss": 1.1145, "step": 16058 }, { "epoch": 1.2585423197492163, "grad_norm": 3.928994655609131, "learning_rate": 7.767522771625365e-06, "loss": 1.2962, "step": 16059 }, { "epoch": 1.2586206896551724, "grad_norm": 3.3479342460632324, "learning_rate": 7.767265948209278e-06, "loss": 1.9446, "step": 16060 }, { "epoch": 1.2586990595611285, "grad_norm": 2.340801477432251, "learning_rate": 7.767009114267936e-06, "loss": 1.1595, "step": 16061 }, { "epoch": 1.2587774294670846, "grad_norm": 3.944277048110962, "learning_rate": 7.766752269802308e-06, "loss": 1.3627, "step": 16062 }, { "epoch": 1.2588557993730407, "grad_norm": 3.5396347045898438, "learning_rate": 7.76649541481338e-06, "loss": 1.5982, "step": 16063 }, { "epoch": 1.2589341692789968, "grad_norm": 2.6736481189727783, "learning_rate": 7.766238549302123e-06, "loss": 1.6669, "step": 16064 }, { "epoch": 1.259012539184953, "grad_norm": 2.189680337905884, "learning_rate": 7.765981673269516e-06, "loss": 1.2846, "step": 16065 }, { "epoch": 1.259090909090909, "grad_norm": 3.1862716674804688, "learning_rate": 7.765724786716532e-06, "loss": 1.0985, "step": 16066 }, { "epoch": 1.2591692789968651, "grad_norm": 2.408569574356079, "learning_rate": 7.765467889644156e-06, "loss": 1.3517, "step": 16067 }, { "epoch": 1.2592476489028213, "grad_norm": 4.1115336418151855, "learning_rate": 7.765210982053359e-06, "loss": 1.6493, "step": 16068 }, { "epoch": 1.2593260188087774, "grad_norm": 2.8836474418640137, "learning_rate": 7.76495406394512e-06, "loss": 1.3689, "step": 16069 }, { "epoch": 1.2594043887147335, "grad_norm": 2.5354506969451904, "learning_rate": 7.764697135320414e-06, "loss": 1.6493, "step": 16070 }, { "epoch": 1.2594827586206896, "grad_norm": 3.157860517501831, "learning_rate": 7.764440196180223e-06, "loss": 0.9197, "step": 16071 }, { "epoch": 1.2595611285266457, "grad_norm": 2.116563320159912, "learning_rate": 7.764183246525518e-06, "loss": 1.2597, "step": 16072 }, { "epoch": 1.2596394984326018, "grad_norm": 4.08419132232666, "learning_rate": 7.763926286357282e-06, "loss": 1.3471, "step": 16073 }, { "epoch": 1.259717868338558, "grad_norm": 2.64372181892395, "learning_rate": 7.76366931567649e-06, "loss": 1.1378, "step": 16074 }, { "epoch": 1.259796238244514, "grad_norm": 2.2062747478485107, "learning_rate": 7.76341233448412e-06, "loss": 1.4563, "step": 16075 }, { "epoch": 1.2598746081504701, "grad_norm": 3.100874662399292, "learning_rate": 7.763155342781148e-06, "loss": 1.7173, "step": 16076 }, { "epoch": 1.2599529780564263, "grad_norm": 2.8160409927368164, "learning_rate": 7.76289834056855e-06, "loss": 1.3834, "step": 16077 }, { "epoch": 1.2600313479623824, "grad_norm": 10.43942642211914, "learning_rate": 7.762641327847309e-06, "loss": 0.8754, "step": 16078 }, { "epoch": 1.2601097178683385, "grad_norm": 3.937692642211914, "learning_rate": 7.7623843046184e-06, "loss": 1.2926, "step": 16079 }, { "epoch": 1.2601880877742948, "grad_norm": 3.493014097213745, "learning_rate": 7.762127270882797e-06, "loss": 1.4135, "step": 16080 }, { "epoch": 1.2602664576802507, "grad_norm": 2.779508352279663, "learning_rate": 7.761870226641482e-06, "loss": 1.4285, "step": 16081 }, { "epoch": 1.260344827586207, "grad_norm": 2.4599926471710205, "learning_rate": 7.76161317189543e-06, "loss": 1.3347, "step": 16082 }, { "epoch": 1.260423197492163, "grad_norm": 1.8063081502914429, "learning_rate": 7.761356106645622e-06, "loss": 0.9388, "step": 16083 }, { "epoch": 1.2605015673981192, "grad_norm": 5.022633075714111, "learning_rate": 7.761099030893033e-06, "loss": 1.2161, "step": 16084 }, { "epoch": 1.2605799373040751, "grad_norm": 3.2408053874969482, "learning_rate": 7.760841944638642e-06, "loss": 1.1353, "step": 16085 }, { "epoch": 1.2606583072100315, "grad_norm": 2.591672658920288, "learning_rate": 7.760584847883425e-06, "loss": 1.5333, "step": 16086 }, { "epoch": 1.2607366771159874, "grad_norm": 2.5269181728363037, "learning_rate": 7.760327740628362e-06, "loss": 1.6806, "step": 16087 }, { "epoch": 1.2608150470219437, "grad_norm": 4.352431774139404, "learning_rate": 7.76007062287443e-06, "loss": 1.4901, "step": 16088 }, { "epoch": 1.2608934169278996, "grad_norm": 3.4484469890594482, "learning_rate": 7.759813494622608e-06, "loss": 1.3165, "step": 16089 }, { "epoch": 1.260971786833856, "grad_norm": 2.079380750656128, "learning_rate": 7.75955635587387e-06, "loss": 1.1885, "step": 16090 }, { "epoch": 1.2610501567398118, "grad_norm": 6.178481578826904, "learning_rate": 7.7592992066292e-06, "loss": 0.8994, "step": 16091 }, { "epoch": 1.2611285266457681, "grad_norm": 2.6435158252716064, "learning_rate": 7.759042046889572e-06, "loss": 1.1376, "step": 16092 }, { "epoch": 1.261206896551724, "grad_norm": 2.277383327484131, "learning_rate": 7.758784876655965e-06, "loss": 1.4645, "step": 16093 }, { "epoch": 1.2612852664576804, "grad_norm": 2.210406541824341, "learning_rate": 7.758527695929357e-06, "loss": 1.0906, "step": 16094 }, { "epoch": 1.2613636363636362, "grad_norm": 4.341009140014648, "learning_rate": 7.758270504710727e-06, "loss": 1.8407, "step": 16095 }, { "epoch": 1.2614420062695926, "grad_norm": 6.011364459991455, "learning_rate": 7.758013303001053e-06, "loss": 0.6014, "step": 16096 }, { "epoch": 1.2615203761755485, "grad_norm": 3.0409154891967773, "learning_rate": 7.757756090801312e-06, "loss": 1.266, "step": 16097 }, { "epoch": 1.2615987460815048, "grad_norm": 5.460184574127197, "learning_rate": 7.757498868112485e-06, "loss": 1.4562, "step": 16098 }, { "epoch": 1.2616771159874607, "grad_norm": 3.0497138500213623, "learning_rate": 7.757241634935548e-06, "loss": 1.4536, "step": 16099 }, { "epoch": 1.261755485893417, "grad_norm": 4.1214118003845215, "learning_rate": 7.756984391271481e-06, "loss": 1.9758, "step": 16100 }, { "epoch": 1.261833855799373, "grad_norm": 3.5463340282440186, "learning_rate": 7.756727137121258e-06, "loss": 1.3735, "step": 16101 }, { "epoch": 1.2619122257053292, "grad_norm": 3.20943021774292, "learning_rate": 7.756469872485863e-06, "loss": 1.307, "step": 16102 }, { "epoch": 1.2619905956112853, "grad_norm": 2.893571376800537, "learning_rate": 7.756212597366273e-06, "loss": 1.2186, "step": 16103 }, { "epoch": 1.2620689655172415, "grad_norm": 2.458792209625244, "learning_rate": 7.755955311763465e-06, "loss": 1.3546, "step": 16104 }, { "epoch": 1.2621473354231976, "grad_norm": 3.968595266342163, "learning_rate": 7.75569801567842e-06, "loss": 1.1971, "step": 16105 }, { "epoch": 1.2622257053291537, "grad_norm": 1.701171636581421, "learning_rate": 7.755440709112114e-06, "loss": 1.3066, "step": 16106 }, { "epoch": 1.2623040752351098, "grad_norm": 10.2024507522583, "learning_rate": 7.755183392065525e-06, "loss": 1.0356, "step": 16107 }, { "epoch": 1.262382445141066, "grad_norm": 3.1230087280273438, "learning_rate": 7.754926064539637e-06, "loss": 1.4117, "step": 16108 }, { "epoch": 1.262460815047022, "grad_norm": 1.8736428022384644, "learning_rate": 7.754668726535424e-06, "loss": 1.4426, "step": 16109 }, { "epoch": 1.2625391849529781, "grad_norm": 2.926642417907715, "learning_rate": 7.754411378053866e-06, "loss": 1.2123, "step": 16110 }, { "epoch": 1.2626175548589342, "grad_norm": 3.8856146335601807, "learning_rate": 7.754154019095941e-06, "loss": 1.5751, "step": 16111 }, { "epoch": 1.2626959247648903, "grad_norm": 2.8041186332702637, "learning_rate": 7.75389664966263e-06, "loss": 1.6117, "step": 16112 }, { "epoch": 1.2627742946708465, "grad_norm": 3.2230043411254883, "learning_rate": 7.753639269754911e-06, "loss": 1.4926, "step": 16113 }, { "epoch": 1.2628526645768026, "grad_norm": 1.9574425220489502, "learning_rate": 7.75338187937376e-06, "loss": 1.1502, "step": 16114 }, { "epoch": 1.2629310344827587, "grad_norm": 2.3404955863952637, "learning_rate": 7.753124478520162e-06, "loss": 1.3224, "step": 16115 }, { "epoch": 1.2630094043887148, "grad_norm": 2.5446135997772217, "learning_rate": 7.75286706719509e-06, "loss": 1.2953, "step": 16116 }, { "epoch": 1.263087774294671, "grad_norm": 3.407573699951172, "learning_rate": 7.752609645399526e-06, "loss": 1.366, "step": 16117 }, { "epoch": 1.263166144200627, "grad_norm": 2.9400579929351807, "learning_rate": 7.75235221313445e-06, "loss": 1.3061, "step": 16118 }, { "epoch": 1.2632445141065831, "grad_norm": 2.958383798599243, "learning_rate": 7.752094770400838e-06, "loss": 0.9425, "step": 16119 }, { "epoch": 1.2633228840125392, "grad_norm": 2.651344060897827, "learning_rate": 7.751837317199673e-06, "loss": 1.5321, "step": 16120 }, { "epoch": 1.2634012539184953, "grad_norm": 3.295156478881836, "learning_rate": 7.751579853531931e-06, "loss": 1.8395, "step": 16121 }, { "epoch": 1.2634796238244514, "grad_norm": 1.7887859344482422, "learning_rate": 7.751322379398593e-06, "loss": 1.3331, "step": 16122 }, { "epoch": 1.2635579937304076, "grad_norm": 1.9108691215515137, "learning_rate": 7.751064894800638e-06, "loss": 1.2763, "step": 16123 }, { "epoch": 1.2636363636363637, "grad_norm": 2.059065103530884, "learning_rate": 7.750807399739044e-06, "loss": 1.1163, "step": 16124 }, { "epoch": 1.2637147335423198, "grad_norm": 4.347363471984863, "learning_rate": 7.750549894214792e-06, "loss": 1.3725, "step": 16125 }, { "epoch": 1.263793103448276, "grad_norm": 3.0936849117279053, "learning_rate": 7.750292378228862e-06, "loss": 1.4801, "step": 16126 }, { "epoch": 1.263871473354232, "grad_norm": 2.1380772590637207, "learning_rate": 7.750034851782232e-06, "loss": 1.2509, "step": 16127 }, { "epoch": 1.263949843260188, "grad_norm": 3.0707013607025146, "learning_rate": 7.749777314875882e-06, "loss": 1.2945, "step": 16128 }, { "epoch": 1.2640282131661442, "grad_norm": 4.487125873565674, "learning_rate": 7.749519767510792e-06, "loss": 1.241, "step": 16129 }, { "epoch": 1.2641065830721003, "grad_norm": 2.5846989154815674, "learning_rate": 7.74926220968794e-06, "loss": 1.5236, "step": 16130 }, { "epoch": 1.2641849529780564, "grad_norm": 2.4044792652130127, "learning_rate": 7.749004641408307e-06, "loss": 1.3853, "step": 16131 }, { "epoch": 1.2642633228840126, "grad_norm": 3.1562716960906982, "learning_rate": 7.748747062672872e-06, "loss": 1.3204, "step": 16132 }, { "epoch": 1.2643416927899687, "grad_norm": 2.3276219367980957, "learning_rate": 7.748489473482615e-06, "loss": 1.504, "step": 16133 }, { "epoch": 1.2644200626959248, "grad_norm": 2.2204771041870117, "learning_rate": 7.748231873838516e-06, "loss": 1.3669, "step": 16134 }, { "epoch": 1.2644984326018809, "grad_norm": 3.856301784515381, "learning_rate": 7.747974263741555e-06, "loss": 1.3579, "step": 16135 }, { "epoch": 1.264576802507837, "grad_norm": 2.5249838829040527, "learning_rate": 7.747716643192711e-06, "loss": 1.5832, "step": 16136 }, { "epoch": 1.264655172413793, "grad_norm": 3.6511282920837402, "learning_rate": 7.747459012192964e-06, "loss": 1.5058, "step": 16137 }, { "epoch": 1.2647335423197492, "grad_norm": 2.7886338233947754, "learning_rate": 7.747201370743294e-06, "loss": 1.551, "step": 16138 }, { "epoch": 1.2648119122257053, "grad_norm": 3.072448968887329, "learning_rate": 7.746943718844681e-06, "loss": 1.1009, "step": 16139 }, { "epoch": 1.2648902821316614, "grad_norm": 3.259770393371582, "learning_rate": 7.746686056498107e-06, "loss": 1.5455, "step": 16140 }, { "epoch": 1.2649686520376175, "grad_norm": 2.8724265098571777, "learning_rate": 7.746428383704549e-06, "loss": 1.65, "step": 16141 }, { "epoch": 1.2650470219435737, "grad_norm": 5.7899956703186035, "learning_rate": 7.746170700464985e-06, "loss": 0.9701, "step": 16142 }, { "epoch": 1.2651253918495298, "grad_norm": 1.8930584192276, "learning_rate": 7.745913006780403e-06, "loss": 1.1864, "step": 16143 }, { "epoch": 1.2652037617554859, "grad_norm": 4.004149436950684, "learning_rate": 7.745655302651775e-06, "loss": 1.492, "step": 16144 }, { "epoch": 1.265282131661442, "grad_norm": 8.146449089050293, "learning_rate": 7.745397588080085e-06, "loss": 0.7386, "step": 16145 }, { "epoch": 1.265360501567398, "grad_norm": 3.5370121002197266, "learning_rate": 7.745139863066313e-06, "loss": 1.28, "step": 16146 }, { "epoch": 1.2654388714733542, "grad_norm": 3.236870288848877, "learning_rate": 7.74488212761144e-06, "loss": 1.1442, "step": 16147 }, { "epoch": 1.2655172413793103, "grad_norm": 4.800334453582764, "learning_rate": 7.744624381716444e-06, "loss": 1.4819, "step": 16148 }, { "epoch": 1.2655956112852664, "grad_norm": 2.096682071685791, "learning_rate": 7.744366625382306e-06, "loss": 1.3, "step": 16149 }, { "epoch": 1.2656739811912225, "grad_norm": 2.8899548053741455, "learning_rate": 7.744108858610008e-06, "loss": 1.2289, "step": 16150 }, { "epoch": 1.2657523510971787, "grad_norm": 4.0198140144348145, "learning_rate": 7.74385108140053e-06, "loss": 1.139, "step": 16151 }, { "epoch": 1.2658307210031348, "grad_norm": 3.048322916030884, "learning_rate": 7.74359329375485e-06, "loss": 1.0877, "step": 16152 }, { "epoch": 1.2659090909090909, "grad_norm": 2.6372156143188477, "learning_rate": 7.743335495673953e-06, "loss": 1.4515, "step": 16153 }, { "epoch": 1.265987460815047, "grad_norm": 7.151068210601807, "learning_rate": 7.743077687158815e-06, "loss": 1.3173, "step": 16154 }, { "epoch": 1.266065830721003, "grad_norm": 9.017237663269043, "learning_rate": 7.742819868210418e-06, "loss": 1.4064, "step": 16155 }, { "epoch": 1.2661442006269592, "grad_norm": 2.8451013565063477, "learning_rate": 7.742562038829744e-06, "loss": 1.282, "step": 16156 }, { "epoch": 1.2662225705329153, "grad_norm": 4.4839019775390625, "learning_rate": 7.742304199017772e-06, "loss": 1.7044, "step": 16157 }, { "epoch": 1.2663009404388714, "grad_norm": 2.288912534713745, "learning_rate": 7.742046348775485e-06, "loss": 1.0817, "step": 16158 }, { "epoch": 1.2663793103448275, "grad_norm": 2.1888861656188965, "learning_rate": 7.741788488103859e-06, "loss": 1.3966, "step": 16159 }, { "epoch": 1.2664576802507836, "grad_norm": 2.348966360092163, "learning_rate": 7.74153061700388e-06, "loss": 1.5313, "step": 16160 }, { "epoch": 1.2665360501567398, "grad_norm": 11.473735809326172, "learning_rate": 7.741272735476526e-06, "loss": 0.9272, "step": 16161 }, { "epoch": 1.2666144200626959, "grad_norm": 3.2950992584228516, "learning_rate": 7.74101484352278e-06, "loss": 0.9985, "step": 16162 }, { "epoch": 1.266692789968652, "grad_norm": 2.809173583984375, "learning_rate": 7.740756941143619e-06, "loss": 1.287, "step": 16163 }, { "epoch": 1.266771159874608, "grad_norm": 5.158835411071777, "learning_rate": 7.740499028340029e-06, "loss": 1.059, "step": 16164 }, { "epoch": 1.2668495297805642, "grad_norm": 3.5011234283447266, "learning_rate": 7.740241105112987e-06, "loss": 1.0757, "step": 16165 }, { "epoch": 1.2669278996865203, "grad_norm": 8.038414001464844, "learning_rate": 7.739983171463474e-06, "loss": 0.7234, "step": 16166 }, { "epoch": 1.2670062695924764, "grad_norm": 2.344451427459717, "learning_rate": 7.739725227392474e-06, "loss": 1.3397, "step": 16167 }, { "epoch": 1.2670846394984325, "grad_norm": 2.9187171459198, "learning_rate": 7.739467272900967e-06, "loss": 1.2871, "step": 16168 }, { "epoch": 1.2671630094043886, "grad_norm": 1.9927983283996582, "learning_rate": 7.739209307989933e-06, "loss": 1.2971, "step": 16169 }, { "epoch": 1.2672413793103448, "grad_norm": 3.086310386657715, "learning_rate": 7.738951332660351e-06, "loss": 1.1692, "step": 16170 }, { "epoch": 1.2673197492163009, "grad_norm": 3.1333038806915283, "learning_rate": 7.73869334691321e-06, "loss": 1.4801, "step": 16171 }, { "epoch": 1.267398119122257, "grad_norm": 2.0917227268218994, "learning_rate": 7.738435350749482e-06, "loss": 1.1908, "step": 16172 }, { "epoch": 1.2674764890282133, "grad_norm": 4.702539443969727, "learning_rate": 7.738177344170154e-06, "loss": 1.712, "step": 16173 }, { "epoch": 1.2675548589341692, "grad_norm": 2.9790706634521484, "learning_rate": 7.737919327176204e-06, "loss": 1.3208, "step": 16174 }, { "epoch": 1.2676332288401255, "grad_norm": 2.316063165664673, "learning_rate": 7.737661299768618e-06, "loss": 1.3316, "step": 16175 }, { "epoch": 1.2677115987460814, "grad_norm": 2.348137378692627, "learning_rate": 7.737403261948372e-06, "loss": 1.5074, "step": 16176 }, { "epoch": 1.2677899686520377, "grad_norm": 3.222494602203369, "learning_rate": 7.737145213716452e-06, "loss": 1.3864, "step": 16177 }, { "epoch": 1.2678683385579936, "grad_norm": 6.661417484283447, "learning_rate": 7.736887155073838e-06, "loss": 1.2925, "step": 16178 }, { "epoch": 1.26794670846395, "grad_norm": 3.8606629371643066, "learning_rate": 7.736629086021509e-06, "loss": 1.3085, "step": 16179 }, { "epoch": 1.2680250783699059, "grad_norm": 3.840789556503296, "learning_rate": 7.73637100656045e-06, "loss": 1.4015, "step": 16180 }, { "epoch": 1.2681034482758622, "grad_norm": 2.0661089420318604, "learning_rate": 7.73611291669164e-06, "loss": 1.5318, "step": 16181 }, { "epoch": 1.268181818181818, "grad_norm": 3.4444994926452637, "learning_rate": 7.735854816416065e-06, "loss": 1.4748, "step": 16182 }, { "epoch": 1.2682601880877744, "grad_norm": 2.8329296112060547, "learning_rate": 7.7355967057347e-06, "loss": 1.1032, "step": 16183 }, { "epoch": 1.2683385579937303, "grad_norm": 2.1210155487060547, "learning_rate": 7.735338584648531e-06, "loss": 1.4716, "step": 16184 }, { "epoch": 1.2684169278996866, "grad_norm": 3.2844409942626953, "learning_rate": 7.735080453158539e-06, "loss": 1.7072, "step": 16185 }, { "epoch": 1.2684952978056425, "grad_norm": 3.2577991485595703, "learning_rate": 7.734822311265707e-06, "loss": 1.6975, "step": 16186 }, { "epoch": 1.2685736677115989, "grad_norm": 6.634128093719482, "learning_rate": 7.734564158971013e-06, "loss": 1.3546, "step": 16187 }, { "epoch": 1.2686520376175547, "grad_norm": 3.3353421688079834, "learning_rate": 7.734305996275445e-06, "loss": 1.3329, "step": 16188 }, { "epoch": 1.268730407523511, "grad_norm": 2.6881065368652344, "learning_rate": 7.734047823179979e-06, "loss": 1.2062, "step": 16189 }, { "epoch": 1.268808777429467, "grad_norm": 3.367417812347412, "learning_rate": 7.733789639685601e-06, "loss": 1.2236, "step": 16190 }, { "epoch": 1.2688871473354233, "grad_norm": 2.778080940246582, "learning_rate": 7.733531445793289e-06, "loss": 1.4129, "step": 16191 }, { "epoch": 1.2689655172413792, "grad_norm": 2.5861217975616455, "learning_rate": 7.733273241504029e-06, "loss": 0.6474, "step": 16192 }, { "epoch": 1.2690438871473355, "grad_norm": 3.9372401237487793, "learning_rate": 7.733015026818801e-06, "loss": 1.7275, "step": 16193 }, { "epoch": 1.2691222570532914, "grad_norm": 3.919895648956299, "learning_rate": 7.732756801738588e-06, "loss": 1.8234, "step": 16194 }, { "epoch": 1.2692006269592477, "grad_norm": 4.299433708190918, "learning_rate": 7.73249856626437e-06, "loss": 1.0698, "step": 16195 }, { "epoch": 1.2692789968652036, "grad_norm": 14.677425384521484, "learning_rate": 7.732240320397134e-06, "loss": 1.6643, "step": 16196 }, { "epoch": 1.26935736677116, "grad_norm": 4.00283670425415, "learning_rate": 7.731982064137857e-06, "loss": 1.3705, "step": 16197 }, { "epoch": 1.269435736677116, "grad_norm": 3.9888453483581543, "learning_rate": 7.731723797487524e-06, "loss": 1.0608, "step": 16198 }, { "epoch": 1.2695141065830722, "grad_norm": 16.10050392150879, "learning_rate": 7.731465520447119e-06, "loss": 1.8835, "step": 16199 }, { "epoch": 1.2695924764890283, "grad_norm": 3.1855316162109375, "learning_rate": 7.73120723301762e-06, "loss": 1.1127, "step": 16200 }, { "epoch": 1.2696708463949844, "grad_norm": 11.386340141296387, "learning_rate": 7.730948935200011e-06, "loss": 0.9549, "step": 16201 }, { "epoch": 1.2697492163009405, "grad_norm": 3.7722156047821045, "learning_rate": 7.730690626995275e-06, "loss": 0.5426, "step": 16202 }, { "epoch": 1.2698275862068966, "grad_norm": 6.505066871643066, "learning_rate": 7.730432308404395e-06, "loss": 1.5549, "step": 16203 }, { "epoch": 1.2699059561128527, "grad_norm": 4.729238033294678, "learning_rate": 7.730173979428353e-06, "loss": 1.1469, "step": 16204 }, { "epoch": 1.2699843260188088, "grad_norm": 2.570857286453247, "learning_rate": 7.729915640068132e-06, "loss": 1.2383, "step": 16205 }, { "epoch": 1.270062695924765, "grad_norm": 3.2463557720184326, "learning_rate": 7.729657290324712e-06, "loss": 0.9018, "step": 16206 }, { "epoch": 1.270141065830721, "grad_norm": 6.633151054382324, "learning_rate": 7.729398930199081e-06, "loss": 1.7038, "step": 16207 }, { "epoch": 1.2702194357366772, "grad_norm": 9.436614036560059, "learning_rate": 7.729140559692217e-06, "loss": 1.1972, "step": 16208 }, { "epoch": 1.2702978056426333, "grad_norm": 2.2887909412384033, "learning_rate": 7.728882178805103e-06, "loss": 1.2492, "step": 16209 }, { "epoch": 1.2703761755485894, "grad_norm": 2.4443161487579346, "learning_rate": 7.728623787538722e-06, "loss": 1.4857, "step": 16210 }, { "epoch": 1.2704545454545455, "grad_norm": 3.6049399375915527, "learning_rate": 7.728365385894062e-06, "loss": 1.0611, "step": 16211 }, { "epoch": 1.2705329153605016, "grad_norm": 2.8488762378692627, "learning_rate": 7.728106973872097e-06, "loss": 1.6339, "step": 16212 }, { "epoch": 1.2706112852664577, "grad_norm": 2.1190593242645264, "learning_rate": 7.727848551473816e-06, "loss": 1.2693, "step": 16213 }, { "epoch": 1.2706896551724138, "grad_norm": 5.342695713043213, "learning_rate": 7.727590118700202e-06, "loss": 1.5465, "step": 16214 }, { "epoch": 1.27076802507837, "grad_norm": 3.044463872909546, "learning_rate": 7.727331675552233e-06, "loss": 1.1029, "step": 16215 }, { "epoch": 1.270846394984326, "grad_norm": 5.111049652099609, "learning_rate": 7.727073222030897e-06, "loss": 1.1171, "step": 16216 }, { "epoch": 1.2709247648902822, "grad_norm": 2.0867600440979004, "learning_rate": 7.726814758137173e-06, "loss": 1.5137, "step": 16217 }, { "epoch": 1.2710031347962383, "grad_norm": 6.165126323699951, "learning_rate": 7.726556283872048e-06, "loss": 1.7583, "step": 16218 }, { "epoch": 1.2710815047021944, "grad_norm": 3.2562906742095947, "learning_rate": 7.726297799236504e-06, "loss": 1.3544, "step": 16219 }, { "epoch": 1.2711598746081505, "grad_norm": 3.169727325439453, "learning_rate": 7.726039304231523e-06, "loss": 1.5264, "step": 16220 }, { "epoch": 1.2712382445141066, "grad_norm": 2.4873311519622803, "learning_rate": 7.725780798858087e-06, "loss": 1.4153, "step": 16221 }, { "epoch": 1.2713166144200627, "grad_norm": 5.152953147888184, "learning_rate": 7.725522283117182e-06, "loss": 1.3339, "step": 16222 }, { "epoch": 1.2713949843260188, "grad_norm": 2.6099131107330322, "learning_rate": 7.725263757009792e-06, "loss": 1.3496, "step": 16223 }, { "epoch": 1.271473354231975, "grad_norm": 3.9835994243621826, "learning_rate": 7.725005220536897e-06, "loss": 1.6341, "step": 16224 }, { "epoch": 1.271551724137931, "grad_norm": 6.8603386878967285, "learning_rate": 7.72474667369948e-06, "loss": 1.6393, "step": 16225 }, { "epoch": 1.2716300940438872, "grad_norm": 9.865520477294922, "learning_rate": 7.724488116498527e-06, "loss": 1.9816, "step": 16226 }, { "epoch": 1.2717084639498433, "grad_norm": 2.4798479080200195, "learning_rate": 7.724229548935023e-06, "loss": 1.0673, "step": 16227 }, { "epoch": 1.2717868338557994, "grad_norm": 1.972059726715088, "learning_rate": 7.723970971009948e-06, "loss": 1.4562, "step": 16228 }, { "epoch": 1.2718652037617555, "grad_norm": 2.3386998176574707, "learning_rate": 7.723712382724285e-06, "loss": 1.1055, "step": 16229 }, { "epoch": 1.2719435736677116, "grad_norm": 2.373263120651245, "learning_rate": 7.72345378407902e-06, "loss": 1.3447, "step": 16230 }, { "epoch": 1.2720219435736677, "grad_norm": 3.1172070503234863, "learning_rate": 7.723195175075136e-06, "loss": 1.5415, "step": 16231 }, { "epoch": 1.2721003134796238, "grad_norm": 3.1669013500213623, "learning_rate": 7.722936555713616e-06, "loss": 1.3422, "step": 16232 }, { "epoch": 1.27217868338558, "grad_norm": 3.830130100250244, "learning_rate": 7.722677925995444e-06, "loss": 1.077, "step": 16233 }, { "epoch": 1.272257053291536, "grad_norm": 3.066835641860962, "learning_rate": 7.722419285921604e-06, "loss": 1.7034, "step": 16234 }, { "epoch": 1.2723354231974922, "grad_norm": 2.59659743309021, "learning_rate": 7.722160635493078e-06, "loss": 1.4602, "step": 16235 }, { "epoch": 1.2724137931034483, "grad_norm": 2.9458444118499756, "learning_rate": 7.721901974710852e-06, "loss": 1.2195, "step": 16236 }, { "epoch": 1.2724921630094044, "grad_norm": 7.3633809089660645, "learning_rate": 7.72164330357591e-06, "loss": 0.9543, "step": 16237 }, { "epoch": 1.2725705329153605, "grad_norm": 2.5287094116210938, "learning_rate": 7.721384622089233e-06, "loss": 0.4836, "step": 16238 }, { "epoch": 1.2726489028213166, "grad_norm": 2.575784206390381, "learning_rate": 7.721125930251806e-06, "loss": 1.3181, "step": 16239 }, { "epoch": 1.2727272727272727, "grad_norm": 2.951369524002075, "learning_rate": 7.720867228064616e-06, "loss": 1.4694, "step": 16240 }, { "epoch": 1.2728056426332288, "grad_norm": 2.408015012741089, "learning_rate": 7.720608515528645e-06, "loss": 1.297, "step": 16241 }, { "epoch": 1.272884012539185, "grad_norm": 2.6440916061401367, "learning_rate": 7.720349792644876e-06, "loss": 1.6378, "step": 16242 }, { "epoch": 1.272962382445141, "grad_norm": 3.2641618251800537, "learning_rate": 7.720091059414292e-06, "loss": 1.371, "step": 16243 }, { "epoch": 1.2730407523510971, "grad_norm": 4.942193031311035, "learning_rate": 7.71983231583788e-06, "loss": 1.1409, "step": 16244 }, { "epoch": 1.2731191222570533, "grad_norm": 3.9012510776519775, "learning_rate": 7.719573561916622e-06, "loss": 1.8425, "step": 16245 }, { "epoch": 1.2731974921630094, "grad_norm": 9.08509635925293, "learning_rate": 7.719314797651505e-06, "loss": 1.6796, "step": 16246 }, { "epoch": 1.2732758620689655, "grad_norm": 2.667137622833252, "learning_rate": 7.71905602304351e-06, "loss": 1.4741, "step": 16247 }, { "epoch": 1.2733542319749216, "grad_norm": 6.511796951293945, "learning_rate": 7.718797238093621e-06, "loss": 1.4097, "step": 16248 }, { "epoch": 1.2734326018808777, "grad_norm": 2.3913509845733643, "learning_rate": 7.718538442802826e-06, "loss": 1.0262, "step": 16249 }, { "epoch": 1.2735109717868338, "grad_norm": 3.967681407928467, "learning_rate": 7.718279637172106e-06, "loss": 1.1349, "step": 16250 }, { "epoch": 1.27358934169279, "grad_norm": 2.5503482818603516, "learning_rate": 7.718020821202445e-06, "loss": 1.1815, "step": 16251 }, { "epoch": 1.273667711598746, "grad_norm": 3.4662156105041504, "learning_rate": 7.717761994894833e-06, "loss": 1.4616, "step": 16252 }, { "epoch": 1.2737460815047021, "grad_norm": 4.633604526519775, "learning_rate": 7.717503158250247e-06, "loss": 0.9696, "step": 16253 }, { "epoch": 1.2738244514106583, "grad_norm": 2.1388773918151855, "learning_rate": 7.717244311269677e-06, "loss": 1.2038, "step": 16254 }, { "epoch": 1.2739028213166144, "grad_norm": 3.2998619079589844, "learning_rate": 7.716985453954103e-06, "loss": 1.8918, "step": 16255 }, { "epoch": 1.2739811912225705, "grad_norm": 3.4315383434295654, "learning_rate": 7.716726586304512e-06, "loss": 1.8689, "step": 16256 }, { "epoch": 1.2740595611285266, "grad_norm": 2.657219409942627, "learning_rate": 7.71646770832189e-06, "loss": 1.5143, "step": 16257 }, { "epoch": 1.2741379310344827, "grad_norm": 2.852074146270752, "learning_rate": 7.716208820007219e-06, "loss": 1.4509, "step": 16258 }, { "epoch": 1.2742163009404388, "grad_norm": 3.4400017261505127, "learning_rate": 7.715949921361485e-06, "loss": 1.291, "step": 16259 }, { "epoch": 1.274294670846395, "grad_norm": 3.1917035579681396, "learning_rate": 7.715691012385672e-06, "loss": 1.4922, "step": 16260 }, { "epoch": 1.274373040752351, "grad_norm": 5.686941146850586, "learning_rate": 7.715432093080766e-06, "loss": 1.3211, "step": 16261 }, { "epoch": 1.2744514106583071, "grad_norm": 20.792919158935547, "learning_rate": 7.715173163447751e-06, "loss": 1.4697, "step": 16262 }, { "epoch": 1.2745297805642632, "grad_norm": 4.977180480957031, "learning_rate": 7.714914223487612e-06, "loss": 1.015, "step": 16263 }, { "epoch": 1.2746081504702194, "grad_norm": 3.2359273433685303, "learning_rate": 7.714655273201332e-06, "loss": 1.4017, "step": 16264 }, { "epoch": 1.2746865203761755, "grad_norm": 1.9242311716079712, "learning_rate": 7.7143963125899e-06, "loss": 1.2836, "step": 16265 }, { "epoch": 1.2747648902821316, "grad_norm": 2.7153255939483643, "learning_rate": 7.714137341654298e-06, "loss": 1.3839, "step": 16266 }, { "epoch": 1.2748432601880877, "grad_norm": 2.052138566970825, "learning_rate": 7.713878360395513e-06, "loss": 1.2994, "step": 16267 }, { "epoch": 1.274921630094044, "grad_norm": 2.3395588397979736, "learning_rate": 7.713619368814527e-06, "loss": 1.2069, "step": 16268 }, { "epoch": 1.275, "grad_norm": 2.5882558822631836, "learning_rate": 7.713360366912327e-06, "loss": 1.4489, "step": 16269 }, { "epoch": 1.2750783699059562, "grad_norm": 3.4121603965759277, "learning_rate": 7.713101354689897e-06, "loss": 1.2129, "step": 16270 }, { "epoch": 1.2751567398119121, "grad_norm": 6.221714496612549, "learning_rate": 7.712842332148222e-06, "loss": 1.3638, "step": 16271 }, { "epoch": 1.2752351097178685, "grad_norm": 2.6835830211639404, "learning_rate": 7.71258329928829e-06, "loss": 1.3672, "step": 16272 }, { "epoch": 1.2753134796238244, "grad_norm": 3.179028034210205, "learning_rate": 7.712324256111086e-06, "loss": 1.294, "step": 16273 }, { "epoch": 1.2753918495297807, "grad_norm": 4.634096622467041, "learning_rate": 7.712065202617592e-06, "loss": 1.3341, "step": 16274 }, { "epoch": 1.2754702194357366, "grad_norm": 2.734025478363037, "learning_rate": 7.711806138808794e-06, "loss": 1.491, "step": 16275 }, { "epoch": 1.275548589341693, "grad_norm": 3.0263712406158447, "learning_rate": 7.71154706468568e-06, "loss": 1.3697, "step": 16276 }, { "epoch": 1.2756269592476488, "grad_norm": 3.8894312381744385, "learning_rate": 7.711287980249232e-06, "loss": 1.3462, "step": 16277 }, { "epoch": 1.2757053291536051, "grad_norm": 2.708524465560913, "learning_rate": 7.711028885500438e-06, "loss": 1.2169, "step": 16278 }, { "epoch": 1.275783699059561, "grad_norm": 6.090775966644287, "learning_rate": 7.710769780440282e-06, "loss": 1.4579, "step": 16279 }, { "epoch": 1.2758620689655173, "grad_norm": 3.6771481037139893, "learning_rate": 7.71051066506975e-06, "loss": 1.4707, "step": 16280 }, { "epoch": 1.2759404388714732, "grad_norm": 2.3331518173217773, "learning_rate": 7.710251539389831e-06, "loss": 1.2228, "step": 16281 }, { "epoch": 1.2760188087774296, "grad_norm": 3.127203941345215, "learning_rate": 7.709992403401503e-06, "loss": 1.4898, "step": 16282 }, { "epoch": 1.2760971786833855, "grad_norm": 2.8431034088134766, "learning_rate": 7.709733257105759e-06, "loss": 1.3384, "step": 16283 }, { "epoch": 1.2761755485893418, "grad_norm": 3.610248565673828, "learning_rate": 7.709474100503578e-06, "loss": 1.6232, "step": 16284 }, { "epoch": 1.2762539184952977, "grad_norm": 10.835589408874512, "learning_rate": 7.709214933595953e-06, "loss": 1.2721, "step": 16285 }, { "epoch": 1.276332288401254, "grad_norm": 2.1188883781433105, "learning_rate": 7.708955756383863e-06, "loss": 1.2141, "step": 16286 }, { "epoch": 1.27641065830721, "grad_norm": 4.72061824798584, "learning_rate": 7.708696568868298e-06, "loss": 1.517, "step": 16287 }, { "epoch": 1.2764890282131662, "grad_norm": 5.721791744232178, "learning_rate": 7.708437371050242e-06, "loss": 0.679, "step": 16288 }, { "epoch": 1.2765673981191221, "grad_norm": 1.9971857070922852, "learning_rate": 7.708178162930682e-06, "loss": 1.2018, "step": 16289 }, { "epoch": 1.2766457680250785, "grad_norm": 3.358370542526245, "learning_rate": 7.707918944510602e-06, "loss": 0.7384, "step": 16290 }, { "epoch": 1.2767241379310346, "grad_norm": 3.0518383979797363, "learning_rate": 7.707659715790991e-06, "loss": 0.9828, "step": 16291 }, { "epoch": 1.2768025078369907, "grad_norm": 3.812589168548584, "learning_rate": 7.707400476772832e-06, "loss": 1.2737, "step": 16292 }, { "epoch": 1.2768808777429468, "grad_norm": 2.1986019611358643, "learning_rate": 7.707141227457113e-06, "loss": 1.392, "step": 16293 }, { "epoch": 1.276959247648903, "grad_norm": 2.330517292022705, "learning_rate": 7.706881967844818e-06, "loss": 1.1049, "step": 16294 }, { "epoch": 1.277037617554859, "grad_norm": 5.645579814910889, "learning_rate": 7.706622697936936e-06, "loss": 1.3519, "step": 16295 }, { "epoch": 1.2771159874608151, "grad_norm": 3.232555866241455, "learning_rate": 7.70636341773445e-06, "loss": 1.1996, "step": 16296 }, { "epoch": 1.2771943573667712, "grad_norm": 2.7475101947784424, "learning_rate": 7.706104127238348e-06, "loss": 1.3339, "step": 16297 }, { "epoch": 1.2772727272727273, "grad_norm": 3.8491227626800537, "learning_rate": 7.705844826449617e-06, "loss": 1.568, "step": 16298 }, { "epoch": 1.2773510971786834, "grad_norm": 6.718956470489502, "learning_rate": 7.70558551536924e-06, "loss": 1.4987, "step": 16299 }, { "epoch": 1.2774294670846396, "grad_norm": 2.1045596599578857, "learning_rate": 7.705326193998207e-06, "loss": 1.5269, "step": 16300 }, { "epoch": 1.2775078369905957, "grad_norm": 1.8888804912567139, "learning_rate": 7.705066862337503e-06, "loss": 1.327, "step": 16301 }, { "epoch": 1.2775862068965518, "grad_norm": 2.3487625122070312, "learning_rate": 7.704807520388114e-06, "loss": 1.1827, "step": 16302 }, { "epoch": 1.277664576802508, "grad_norm": 3.7742862701416016, "learning_rate": 7.704548168151024e-06, "loss": 1.1101, "step": 16303 }, { "epoch": 1.277742946708464, "grad_norm": 2.7412962913513184, "learning_rate": 7.704288805627225e-06, "loss": 1.0616, "step": 16304 }, { "epoch": 1.27782131661442, "grad_norm": 2.5906472206115723, "learning_rate": 7.704029432817697e-06, "loss": 1.5731, "step": 16305 }, { "epoch": 1.2778996865203762, "grad_norm": 3.1920278072357178, "learning_rate": 7.703770049723433e-06, "loss": 1.2752, "step": 16306 }, { "epoch": 1.2779780564263323, "grad_norm": 2.8962485790252686, "learning_rate": 7.703510656345415e-06, "loss": 1.7962, "step": 16307 }, { "epoch": 1.2780564263322884, "grad_norm": 2.5468270778656006, "learning_rate": 7.703251252684633e-06, "loss": 1.3565, "step": 16308 }, { "epoch": 1.2781347962382446, "grad_norm": 7.281578063964844, "learning_rate": 7.70299183874207e-06, "loss": 1.6986, "step": 16309 }, { "epoch": 1.2782131661442007, "grad_norm": 10.44620132446289, "learning_rate": 7.702732414518715e-06, "loss": 1.6612, "step": 16310 }, { "epoch": 1.2782915360501568, "grad_norm": 5.467554092407227, "learning_rate": 7.702472980015554e-06, "loss": 1.7357, "step": 16311 }, { "epoch": 1.2783699059561129, "grad_norm": 7.5151519775390625, "learning_rate": 7.702213535233571e-06, "loss": 1.2145, "step": 16312 }, { "epoch": 1.278448275862069, "grad_norm": 2.5079524517059326, "learning_rate": 7.70195408017376e-06, "loss": 1.7265, "step": 16313 }, { "epoch": 1.278526645768025, "grad_norm": 3.4018867015838623, "learning_rate": 7.701694614837101e-06, "loss": 1.8023, "step": 16314 }, { "epoch": 1.2786050156739812, "grad_norm": 5.910706043243408, "learning_rate": 7.701435139224584e-06, "loss": 1.2305, "step": 16315 }, { "epoch": 1.2786833855799373, "grad_norm": 2.663273572921753, "learning_rate": 7.701175653337195e-06, "loss": 1.3762, "step": 16316 }, { "epoch": 1.2787617554858934, "grad_norm": 4.853041648864746, "learning_rate": 7.700916157175923e-06, "loss": 1.4302, "step": 16317 }, { "epoch": 1.2788401253918495, "grad_norm": 2.9391913414001465, "learning_rate": 7.700656650741751e-06, "loss": 1.2836, "step": 16318 }, { "epoch": 1.2789184952978057, "grad_norm": 5.739162921905518, "learning_rate": 7.70039713403567e-06, "loss": 1.4422, "step": 16319 }, { "epoch": 1.2789968652037618, "grad_norm": 3.075531244277954, "learning_rate": 7.700137607058665e-06, "loss": 1.2556, "step": 16320 }, { "epoch": 1.2790752351097179, "grad_norm": 3.2185754776000977, "learning_rate": 7.69987806981172e-06, "loss": 1.2956, "step": 16321 }, { "epoch": 1.279153605015674, "grad_norm": 2.20625376701355, "learning_rate": 7.69961852229583e-06, "loss": 1.3726, "step": 16322 }, { "epoch": 1.27923197492163, "grad_norm": 4.896811008453369, "learning_rate": 7.699358964511977e-06, "loss": 1.4978, "step": 16323 }, { "epoch": 1.2793103448275862, "grad_norm": 2.294447660446167, "learning_rate": 7.699099396461146e-06, "loss": 1.1676, "step": 16324 }, { "epoch": 1.2793887147335423, "grad_norm": 2.1644656658172607, "learning_rate": 7.69883981814433e-06, "loss": 1.384, "step": 16325 }, { "epoch": 1.2794670846394984, "grad_norm": 11.657171249389648, "learning_rate": 7.698580229562513e-06, "loss": 1.4665, "step": 16326 }, { "epoch": 1.2795454545454545, "grad_norm": 4.171713829040527, "learning_rate": 7.698320630716684e-06, "loss": 1.6415, "step": 16327 }, { "epoch": 1.2796238244514107, "grad_norm": 2.9722580909729004, "learning_rate": 7.698061021607827e-06, "loss": 1.4252, "step": 16328 }, { "epoch": 1.2797021943573668, "grad_norm": 2.04369854927063, "learning_rate": 7.697801402236933e-06, "loss": 1.4379, "step": 16329 }, { "epoch": 1.2797805642633229, "grad_norm": 3.1687045097351074, "learning_rate": 7.697541772604988e-06, "loss": 1.2269, "step": 16330 }, { "epoch": 1.279858934169279, "grad_norm": 3.7889339923858643, "learning_rate": 7.69728213271298e-06, "loss": 1.6725, "step": 16331 }, { "epoch": 1.279937304075235, "grad_norm": 2.437189817428589, "learning_rate": 7.697022482561895e-06, "loss": 1.2622, "step": 16332 }, { "epoch": 1.2800156739811912, "grad_norm": 4.172094821929932, "learning_rate": 7.696762822152723e-06, "loss": 1.3978, "step": 16333 }, { "epoch": 1.2800940438871473, "grad_norm": 2.54617977142334, "learning_rate": 7.69650315148645e-06, "loss": 0.5284, "step": 16334 }, { "epoch": 1.2801724137931034, "grad_norm": 2.785905599594116, "learning_rate": 7.696243470564062e-06, "loss": 1.242, "step": 16335 }, { "epoch": 1.2802507836990595, "grad_norm": 3.9116904735565186, "learning_rate": 7.695983779386552e-06, "loss": 1.2553, "step": 16336 }, { "epoch": 1.2803291536050156, "grad_norm": 3.6671359539031982, "learning_rate": 7.695724077954902e-06, "loss": 1.4051, "step": 16337 }, { "epoch": 1.2804075235109718, "grad_norm": 2.3764472007751465, "learning_rate": 7.695464366270105e-06, "loss": 1.5274, "step": 16338 }, { "epoch": 1.2804858934169279, "grad_norm": 3.4693286418914795, "learning_rate": 7.695204644333145e-06, "loss": 0.8012, "step": 16339 }, { "epoch": 1.280564263322884, "grad_norm": 2.5304243564605713, "learning_rate": 7.694944912145009e-06, "loss": 1.2881, "step": 16340 }, { "epoch": 1.28064263322884, "grad_norm": 3.2745676040649414, "learning_rate": 7.694685169706688e-06, "loss": 1.0988, "step": 16341 }, { "epoch": 1.2807210031347962, "grad_norm": 4.386086463928223, "learning_rate": 7.694425417019167e-06, "loss": 1.5763, "step": 16342 }, { "epoch": 1.2807993730407523, "grad_norm": 2.6479806900024414, "learning_rate": 7.694165654083439e-06, "loss": 1.3673, "step": 16343 }, { "epoch": 1.2808777429467084, "grad_norm": 2.204665184020996, "learning_rate": 7.693905880900487e-06, "loss": 1.5081, "step": 16344 }, { "epoch": 1.2809561128526645, "grad_norm": 4.458737850189209, "learning_rate": 7.6936460974713e-06, "loss": 1.0181, "step": 16345 }, { "epoch": 1.2810344827586206, "grad_norm": 4.021216869354248, "learning_rate": 7.693386303796867e-06, "loss": 1.4137, "step": 16346 }, { "epoch": 1.2811128526645768, "grad_norm": 2.499983310699463, "learning_rate": 7.693126499878176e-06, "loss": 1.5754, "step": 16347 }, { "epoch": 1.2811912225705329, "grad_norm": 3.083090305328369, "learning_rate": 7.692866685716216e-06, "loss": 1.4981, "step": 16348 }, { "epoch": 1.281269592476489, "grad_norm": 4.282631874084473, "learning_rate": 7.692606861311972e-06, "loss": 1.4152, "step": 16349 }, { "epoch": 1.281347962382445, "grad_norm": 2.6819474697113037, "learning_rate": 7.692347026666438e-06, "loss": 1.3112, "step": 16350 }, { "epoch": 1.2814263322884012, "grad_norm": 10.670987129211426, "learning_rate": 7.692087181780597e-06, "loss": 1.6901, "step": 16351 }, { "epoch": 1.2815047021943573, "grad_norm": 3.9691336154937744, "learning_rate": 7.69182732665544e-06, "loss": 0.9345, "step": 16352 }, { "epoch": 1.2815830721003134, "grad_norm": 2.8475630283355713, "learning_rate": 7.691567461291953e-06, "loss": 1.406, "step": 16353 }, { "epoch": 1.2816614420062695, "grad_norm": 4.233014106750488, "learning_rate": 7.691307585691126e-06, "loss": 1.9314, "step": 16354 }, { "epoch": 1.2817398119122256, "grad_norm": 2.636573553085327, "learning_rate": 7.691047699853947e-06, "loss": 1.3994, "step": 16355 }, { "epoch": 1.2818181818181817, "grad_norm": 8.481795310974121, "learning_rate": 7.690787803781406e-06, "loss": 0.6628, "step": 16356 }, { "epoch": 1.2818965517241379, "grad_norm": 6.918747425079346, "learning_rate": 7.690527897474489e-06, "loss": 1.2362, "step": 16357 }, { "epoch": 1.281974921630094, "grad_norm": 2.7078170776367188, "learning_rate": 7.690267980934187e-06, "loss": 1.5412, "step": 16358 }, { "epoch": 1.28205329153605, "grad_norm": 2.2230801582336426, "learning_rate": 7.690008054161487e-06, "loss": 1.3861, "step": 16359 }, { "epoch": 1.2821316614420062, "grad_norm": 4.827703952789307, "learning_rate": 7.689748117157379e-06, "loss": 1.2367, "step": 16360 }, { "epoch": 1.2822100313479625, "grad_norm": 2.1416823863983154, "learning_rate": 7.689488169922848e-06, "loss": 1.3406, "step": 16361 }, { "epoch": 1.2822884012539184, "grad_norm": 2.8653759956359863, "learning_rate": 7.689228212458888e-06, "loss": 1.3878, "step": 16362 }, { "epoch": 1.2823667711598747, "grad_norm": 3.687127113342285, "learning_rate": 7.688968244766483e-06, "loss": 1.7389, "step": 16363 }, { "epoch": 1.2824451410658306, "grad_norm": 3.085132122039795, "learning_rate": 7.688708266846625e-06, "loss": 1.0307, "step": 16364 }, { "epoch": 1.282523510971787, "grad_norm": 4.709840297698975, "learning_rate": 7.688448278700303e-06, "loss": 1.3022, "step": 16365 }, { "epoch": 1.2826018808777429, "grad_norm": 2.644890069961548, "learning_rate": 7.688188280328504e-06, "loss": 1.3287, "step": 16366 }, { "epoch": 1.2826802507836992, "grad_norm": 2.505798578262329, "learning_rate": 7.687928271732216e-06, "loss": 1.2179, "step": 16367 }, { "epoch": 1.282758620689655, "grad_norm": 3.6416587829589844, "learning_rate": 7.68766825291243e-06, "loss": 1.4263, "step": 16368 }, { "epoch": 1.2828369905956114, "grad_norm": 2.3070895671844482, "learning_rate": 7.687408223870136e-06, "loss": 1.4339, "step": 16369 }, { "epoch": 1.2829153605015673, "grad_norm": 4.092777729034424, "learning_rate": 7.68714818460632e-06, "loss": 1.2793, "step": 16370 }, { "epoch": 1.2829937304075236, "grad_norm": 4.722558498382568, "learning_rate": 7.686888135121973e-06, "loss": 2.0604, "step": 16371 }, { "epoch": 1.2830721003134795, "grad_norm": 4.727590560913086, "learning_rate": 7.686628075418084e-06, "loss": 1.4095, "step": 16372 }, { "epoch": 1.2831504702194358, "grad_norm": 1.8565889596939087, "learning_rate": 7.686368005495642e-06, "loss": 1.2917, "step": 16373 }, { "epoch": 1.2832288401253917, "grad_norm": 2.433610677719116, "learning_rate": 7.686107925355634e-06, "loss": 1.3817, "step": 16374 }, { "epoch": 1.283307210031348, "grad_norm": 3.6884448528289795, "learning_rate": 7.685847834999052e-06, "loss": 1.3023, "step": 16375 }, { "epoch": 1.283385579937304, "grad_norm": 6.7480854988098145, "learning_rate": 7.685587734426885e-06, "loss": 1.4004, "step": 16376 }, { "epoch": 1.2834639498432603, "grad_norm": 2.623189687728882, "learning_rate": 7.685327623640121e-06, "loss": 1.3727, "step": 16377 }, { "epoch": 1.2835423197492162, "grad_norm": 2.280224561691284, "learning_rate": 7.68506750263975e-06, "loss": 1.2648, "step": 16378 }, { "epoch": 1.2836206896551725, "grad_norm": 3.3874075412750244, "learning_rate": 7.684807371426763e-06, "loss": 1.505, "step": 16379 }, { "epoch": 1.2836990595611284, "grad_norm": 2.4014852046966553, "learning_rate": 7.684547230002147e-06, "loss": 1.3525, "step": 16380 }, { "epoch": 1.2837774294670847, "grad_norm": 8.948784828186035, "learning_rate": 7.684287078366892e-06, "loss": 0.6755, "step": 16381 }, { "epoch": 1.2838557993730406, "grad_norm": 11.540122985839844, "learning_rate": 7.684026916521988e-06, "loss": 0.6107, "step": 16382 }, { "epoch": 1.283934169278997, "grad_norm": 3.3967065811157227, "learning_rate": 7.683766744468422e-06, "loss": 0.9809, "step": 16383 }, { "epoch": 1.284012539184953, "grad_norm": 4.995034217834473, "learning_rate": 7.683506562207189e-06, "loss": 1.0854, "step": 16384 }, { "epoch": 1.2840909090909092, "grad_norm": 3.5947985649108887, "learning_rate": 7.683246369739272e-06, "loss": 1.3223, "step": 16385 }, { "epoch": 1.2841692789968653, "grad_norm": 3.3294708728790283, "learning_rate": 7.682986167065667e-06, "loss": 1.0113, "step": 16386 }, { "epoch": 1.2842476489028214, "grad_norm": 2.820073127746582, "learning_rate": 7.682725954187357e-06, "loss": 1.4408, "step": 16387 }, { "epoch": 1.2843260188087775, "grad_norm": 2.760593891143799, "learning_rate": 7.68246573110534e-06, "loss": 1.3699, "step": 16388 }, { "epoch": 1.2844043887147336, "grad_norm": 4.031617164611816, "learning_rate": 7.682205497820597e-06, "loss": 1.0176, "step": 16389 }, { "epoch": 1.2844827586206897, "grad_norm": 2.426588535308838, "learning_rate": 7.681945254334126e-06, "loss": 1.3547, "step": 16390 }, { "epoch": 1.2845611285266458, "grad_norm": 3.4215073585510254, "learning_rate": 7.68168500064691e-06, "loss": 1.4578, "step": 16391 }, { "epoch": 1.284639498432602, "grad_norm": 3.439107656478882, "learning_rate": 7.681424736759942e-06, "loss": 1.1253, "step": 16392 }, { "epoch": 1.284717868338558, "grad_norm": 3.0730552673339844, "learning_rate": 7.681164462674214e-06, "loss": 1.1919, "step": 16393 }, { "epoch": 1.2847962382445142, "grad_norm": 2.700451374053955, "learning_rate": 7.68090417839071e-06, "loss": 1.313, "step": 16394 }, { "epoch": 1.2848746081504703, "grad_norm": 2.4841668605804443, "learning_rate": 7.680643883910426e-06, "loss": 1.3188, "step": 16395 }, { "epoch": 1.2849529780564264, "grad_norm": 2.7987754344940186, "learning_rate": 7.680383579234347e-06, "loss": 1.2477, "step": 16396 }, { "epoch": 1.2850313479623825, "grad_norm": 5.992220401763916, "learning_rate": 7.680123264363468e-06, "loss": 0.8122, "step": 16397 }, { "epoch": 1.2851097178683386, "grad_norm": 2.408818244934082, "learning_rate": 7.679862939298776e-06, "loss": 1.3995, "step": 16398 }, { "epoch": 1.2851880877742947, "grad_norm": 2.9105286598205566, "learning_rate": 7.67960260404126e-06, "loss": 1.3222, "step": 16399 }, { "epoch": 1.2852664576802508, "grad_norm": 3.0785024166107178, "learning_rate": 7.679342258591916e-06, "loss": 1.055, "step": 16400 }, { "epoch": 1.285344827586207, "grad_norm": 3.01450777053833, "learning_rate": 7.679081902951727e-06, "loss": 0.9177, "step": 16401 }, { "epoch": 1.285423197492163, "grad_norm": 3.253758430480957, "learning_rate": 7.678821537121687e-06, "loss": 1.4401, "step": 16402 }, { "epoch": 1.2855015673981192, "grad_norm": 2.352163076400757, "learning_rate": 7.678561161102786e-06, "loss": 1.3697, "step": 16403 }, { "epoch": 1.2855799373040753, "grad_norm": 6.707089424133301, "learning_rate": 7.678300774896015e-06, "loss": 1.9707, "step": 16404 }, { "epoch": 1.2856583072100314, "grad_norm": 2.8988373279571533, "learning_rate": 7.678040378502362e-06, "loss": 1.3202, "step": 16405 }, { "epoch": 1.2857366771159875, "grad_norm": 2.5769104957580566, "learning_rate": 7.677779971922821e-06, "loss": 1.2258, "step": 16406 }, { "epoch": 1.2858150470219436, "grad_norm": 2.6400630474090576, "learning_rate": 7.67751955515838e-06, "loss": 1.2164, "step": 16407 }, { "epoch": 1.2858934169278997, "grad_norm": 5.064312934875488, "learning_rate": 7.677259128210029e-06, "loss": 0.5221, "step": 16408 }, { "epoch": 1.2859717868338558, "grad_norm": 2.08535099029541, "learning_rate": 7.676998691078758e-06, "loss": 1.648, "step": 16409 }, { "epoch": 1.286050156739812, "grad_norm": 2.6185855865478516, "learning_rate": 7.67673824376556e-06, "loss": 1.4878, "step": 16410 }, { "epoch": 1.286128526645768, "grad_norm": 2.6308577060699463, "learning_rate": 7.676477786271426e-06, "loss": 1.5302, "step": 16411 }, { "epoch": 1.2862068965517242, "grad_norm": 3.63163685798645, "learning_rate": 7.676217318597343e-06, "loss": 1.5021, "step": 16412 }, { "epoch": 1.2862852664576803, "grad_norm": 5.147244453430176, "learning_rate": 7.675956840744306e-06, "loss": 1.6106, "step": 16413 }, { "epoch": 1.2863636363636364, "grad_norm": 3.622012138366699, "learning_rate": 7.675696352713304e-06, "loss": 1.3584, "step": 16414 }, { "epoch": 1.2864420062695925, "grad_norm": 2.924107074737549, "learning_rate": 7.675435854505325e-06, "loss": 1.6079, "step": 16415 }, { "epoch": 1.2865203761755486, "grad_norm": 3.4959876537323, "learning_rate": 7.675175346121365e-06, "loss": 1.4288, "step": 16416 }, { "epoch": 1.2865987460815047, "grad_norm": 3.409684419631958, "learning_rate": 7.674914827562408e-06, "loss": 1.2618, "step": 16417 }, { "epoch": 1.2866771159874608, "grad_norm": 3.742481231689453, "learning_rate": 7.674654298829452e-06, "loss": 1.3072, "step": 16418 }, { "epoch": 1.286755485893417, "grad_norm": 3.165694236755371, "learning_rate": 7.674393759923484e-06, "loss": 1.4283, "step": 16419 }, { "epoch": 1.286833855799373, "grad_norm": 3.36917781829834, "learning_rate": 7.674133210845496e-06, "loss": 1.2756, "step": 16420 }, { "epoch": 1.2869122257053291, "grad_norm": 2.062262535095215, "learning_rate": 7.673872651596481e-06, "loss": 1.2446, "step": 16421 }, { "epoch": 1.2869905956112853, "grad_norm": 3.8652312755584717, "learning_rate": 7.673612082177424e-06, "loss": 0.8756, "step": 16422 }, { "epoch": 1.2870689655172414, "grad_norm": 3.6540470123291016, "learning_rate": 7.673351502589321e-06, "loss": 1.2183, "step": 16423 }, { "epoch": 1.2871473354231975, "grad_norm": 3.377664804458618, "learning_rate": 7.673090912833162e-06, "loss": 1.5865, "step": 16424 }, { "epoch": 1.2872257053291536, "grad_norm": 6.3883056640625, "learning_rate": 7.67283031290994e-06, "loss": 1.236, "step": 16425 }, { "epoch": 1.2873040752351097, "grad_norm": 8.208599090576172, "learning_rate": 7.672569702820643e-06, "loss": 1.378, "step": 16426 }, { "epoch": 1.2873824451410658, "grad_norm": 3.670849561691284, "learning_rate": 7.672309082566263e-06, "loss": 1.497, "step": 16427 }, { "epoch": 1.287460815047022, "grad_norm": 2.963365077972412, "learning_rate": 7.672048452147792e-06, "loss": 1.2982, "step": 16428 }, { "epoch": 1.287539184952978, "grad_norm": 2.050729274749756, "learning_rate": 7.671787811566222e-06, "loss": 1.383, "step": 16429 }, { "epoch": 1.2876175548589341, "grad_norm": 3.7474310398101807, "learning_rate": 7.671527160822542e-06, "loss": 1.3627, "step": 16430 }, { "epoch": 1.2876959247648903, "grad_norm": 3.743760585784912, "learning_rate": 7.671266499917747e-06, "loss": 1.4481, "step": 16431 }, { "epoch": 1.2877742946708464, "grad_norm": 2.676891803741455, "learning_rate": 7.671005828852825e-06, "loss": 1.1906, "step": 16432 }, { "epoch": 1.2878526645768025, "grad_norm": 2.9238815307617188, "learning_rate": 7.670745147628768e-06, "loss": 1.3509, "step": 16433 }, { "epoch": 1.2879310344827586, "grad_norm": 2.2425568103790283, "learning_rate": 7.67048445624657e-06, "loss": 1.192, "step": 16434 }, { "epoch": 1.2880094043887147, "grad_norm": 3.455517292022705, "learning_rate": 7.67022375470722e-06, "loss": 1.0194, "step": 16435 }, { "epoch": 1.2880877742946708, "grad_norm": 3.3625168800354004, "learning_rate": 7.66996304301171e-06, "loss": 1.0865, "step": 16436 }, { "epoch": 1.288166144200627, "grad_norm": 2.318411111831665, "learning_rate": 7.669702321161032e-06, "loss": 1.605, "step": 16437 }, { "epoch": 1.288244514106583, "grad_norm": 2.5769004821777344, "learning_rate": 7.669441589156179e-06, "loss": 1.4885, "step": 16438 }, { "epoch": 1.2883228840125391, "grad_norm": 3.984901189804077, "learning_rate": 7.669180846998139e-06, "loss": 1.2046, "step": 16439 }, { "epoch": 1.2884012539184952, "grad_norm": 2.702789545059204, "learning_rate": 7.668920094687909e-06, "loss": 1.2342, "step": 16440 }, { "epoch": 1.2884796238244514, "grad_norm": 6.580308437347412, "learning_rate": 7.668659332226474e-06, "loss": 0.862, "step": 16441 }, { "epoch": 1.2885579937304075, "grad_norm": 4.4283366203308105, "learning_rate": 7.668398559614833e-06, "loss": 1.3837, "step": 16442 }, { "epoch": 1.2886363636363636, "grad_norm": 2.794806480407715, "learning_rate": 7.668137776853972e-06, "loss": 1.6732, "step": 16443 }, { "epoch": 1.2887147335423197, "grad_norm": 2.143392562866211, "learning_rate": 7.667876983944886e-06, "loss": 1.222, "step": 16444 }, { "epoch": 1.2887931034482758, "grad_norm": 2.600459337234497, "learning_rate": 7.667616180888568e-06, "loss": 1.088, "step": 16445 }, { "epoch": 1.288871473354232, "grad_norm": 2.9046661853790283, "learning_rate": 7.667355367686007e-06, "loss": 1.2891, "step": 16446 }, { "epoch": 1.288949843260188, "grad_norm": 10.546079635620117, "learning_rate": 7.667094544338197e-06, "loss": 1.1903, "step": 16447 }, { "epoch": 1.2890282131661441, "grad_norm": 4.187384128570557, "learning_rate": 7.666833710846127e-06, "loss": 0.9892, "step": 16448 }, { "epoch": 1.2891065830721002, "grad_norm": 3.3450090885162354, "learning_rate": 7.666572867210794e-06, "loss": 1.6072, "step": 16449 }, { "epoch": 1.2891849529780564, "grad_norm": 4.232439041137695, "learning_rate": 7.666312013433183e-06, "loss": 1.2235, "step": 16450 }, { "epoch": 1.2892633228840125, "grad_norm": 2.911961793899536, "learning_rate": 7.666051149514295e-06, "loss": 1.2672, "step": 16451 }, { "epoch": 1.2893416927899686, "grad_norm": 6.200765132904053, "learning_rate": 7.665790275455118e-06, "loss": 1.3404, "step": 16452 }, { "epoch": 1.2894200626959247, "grad_norm": 2.741269826889038, "learning_rate": 7.665529391256642e-06, "loss": 1.1107, "step": 16453 }, { "epoch": 1.2894984326018808, "grad_norm": 3.7696168422698975, "learning_rate": 7.665268496919863e-06, "loss": 1.3654, "step": 16454 }, { "epoch": 1.289576802507837, "grad_norm": 6.7991042137146, "learning_rate": 7.66500759244577e-06, "loss": 1.7221, "step": 16455 }, { "epoch": 1.2896551724137932, "grad_norm": 3.9788622856140137, "learning_rate": 7.664746677835357e-06, "loss": 1.5219, "step": 16456 }, { "epoch": 1.2897335423197491, "grad_norm": 2.7093276977539062, "learning_rate": 7.664485753089617e-06, "loss": 1.8065, "step": 16457 }, { "epoch": 1.2898119122257055, "grad_norm": 2.8805012702941895, "learning_rate": 7.66422481820954e-06, "loss": 1.0641, "step": 16458 }, { "epoch": 1.2898902821316613, "grad_norm": 6.100564479827881, "learning_rate": 7.663963873196123e-06, "loss": 1.6109, "step": 16459 }, { "epoch": 1.2899686520376177, "grad_norm": 2.2814016342163086, "learning_rate": 7.663702918050353e-06, "loss": 1.337, "step": 16460 }, { "epoch": 1.2900470219435736, "grad_norm": 2.2505335807800293, "learning_rate": 7.66344195277323e-06, "loss": 1.4431, "step": 16461 }, { "epoch": 1.29012539184953, "grad_norm": 3.456069231033325, "learning_rate": 7.663180977365736e-06, "loss": 0.9851, "step": 16462 }, { "epoch": 1.2902037617554858, "grad_norm": 2.902083396911621, "learning_rate": 7.662919991828872e-06, "loss": 1.0397, "step": 16463 }, { "epoch": 1.2902821316614421, "grad_norm": 3.412647008895874, "learning_rate": 7.662658996163628e-06, "loss": 1.0908, "step": 16464 }, { "epoch": 1.290360501567398, "grad_norm": 3.02001690864563, "learning_rate": 7.662397990370998e-06, "loss": 1.5004, "step": 16465 }, { "epoch": 1.2904388714733543, "grad_norm": 3.423245906829834, "learning_rate": 7.662136974451972e-06, "loss": 1.4912, "step": 16466 }, { "epoch": 1.2905172413793102, "grad_norm": 6.076349258422852, "learning_rate": 7.661875948407545e-06, "loss": 1.7943, "step": 16467 }, { "epoch": 1.2905956112852666, "grad_norm": 7.473971366882324, "learning_rate": 7.66161491223871e-06, "loss": 2.6169, "step": 16468 }, { "epoch": 1.2906739811912225, "grad_norm": 3.0381784439086914, "learning_rate": 7.661353865946457e-06, "loss": 1.1394, "step": 16469 }, { "epoch": 1.2907523510971788, "grad_norm": 4.310506343841553, "learning_rate": 7.661092809531783e-06, "loss": 1.6292, "step": 16470 }, { "epoch": 1.2908307210031347, "grad_norm": 2.2677440643310547, "learning_rate": 7.660831742995676e-06, "loss": 1.4496, "step": 16471 }, { "epoch": 1.290909090909091, "grad_norm": 5.314913749694824, "learning_rate": 7.660570666339133e-06, "loss": 2.1825, "step": 16472 }, { "epoch": 1.290987460815047, "grad_norm": 2.5621306896209717, "learning_rate": 7.660309579563146e-06, "loss": 1.39, "step": 16473 }, { "epoch": 1.2910658307210032, "grad_norm": 2.811131715774536, "learning_rate": 7.660048482668709e-06, "loss": 0.5514, "step": 16474 }, { "epoch": 1.2911442006269591, "grad_norm": 2.5573511123657227, "learning_rate": 7.659787375656812e-06, "loss": 1.4712, "step": 16475 }, { "epoch": 1.2912225705329154, "grad_norm": 2.4663474559783936, "learning_rate": 7.65952625852845e-06, "loss": 1.3627, "step": 16476 }, { "epoch": 1.2913009404388713, "grad_norm": 4.688687801361084, "learning_rate": 7.659265131284617e-06, "loss": 1.3229, "step": 16477 }, { "epoch": 1.2913793103448277, "grad_norm": 2.5380337238311768, "learning_rate": 7.659003993926304e-06, "loss": 1.1899, "step": 16478 }, { "epoch": 1.2914576802507838, "grad_norm": 2.8612399101257324, "learning_rate": 7.658742846454508e-06, "loss": 1.2956, "step": 16479 }, { "epoch": 1.29153605015674, "grad_norm": 2.3973147869110107, "learning_rate": 7.658481688870218e-06, "loss": 1.294, "step": 16480 }, { "epoch": 1.291614420062696, "grad_norm": 3.518118381500244, "learning_rate": 7.658220521174431e-06, "loss": 1.6336, "step": 16481 }, { "epoch": 1.291692789968652, "grad_norm": 4.7505364418029785, "learning_rate": 7.657959343368137e-06, "loss": 0.4872, "step": 16482 }, { "epoch": 1.2917711598746082, "grad_norm": 5.08241605758667, "learning_rate": 7.65769815545233e-06, "loss": 2.0198, "step": 16483 }, { "epoch": 1.2918495297805643, "grad_norm": 7.5826921463012695, "learning_rate": 7.657436957428006e-06, "loss": 1.2517, "step": 16484 }, { "epoch": 1.2919278996865204, "grad_norm": 2.105182647705078, "learning_rate": 7.657175749296155e-06, "loss": 1.4302, "step": 16485 }, { "epoch": 1.2920062695924766, "grad_norm": 3.849926710128784, "learning_rate": 7.656914531057773e-06, "loss": 1.3889, "step": 16486 }, { "epoch": 1.2920846394984327, "grad_norm": 2.585273027420044, "learning_rate": 7.656653302713855e-06, "loss": 1.4842, "step": 16487 }, { "epoch": 1.2921630094043888, "grad_norm": 2.6815695762634277, "learning_rate": 7.656392064265389e-06, "loss": 1.3205, "step": 16488 }, { "epoch": 1.2922413793103449, "grad_norm": 2.474555015563965, "learning_rate": 7.656130815713374e-06, "loss": 1.4596, "step": 16489 }, { "epoch": 1.292319749216301, "grad_norm": 2.3236684799194336, "learning_rate": 7.655869557058801e-06, "loss": 1.4544, "step": 16490 }, { "epoch": 1.292398119122257, "grad_norm": 3.4937617778778076, "learning_rate": 7.655608288302665e-06, "loss": 1.2506, "step": 16491 }, { "epoch": 1.2924764890282132, "grad_norm": 2.713679552078247, "learning_rate": 7.655347009445959e-06, "loss": 1.4002, "step": 16492 }, { "epoch": 1.2925548589341693, "grad_norm": 3.2440595626831055, "learning_rate": 7.655085720489675e-06, "loss": 0.6942, "step": 16493 }, { "epoch": 1.2926332288401254, "grad_norm": 3.452975273132324, "learning_rate": 7.654824421434811e-06, "loss": 0.9129, "step": 16494 }, { "epoch": 1.2927115987460815, "grad_norm": 2.4999022483825684, "learning_rate": 7.654563112282355e-06, "loss": 1.3462, "step": 16495 }, { "epoch": 1.2927899686520377, "grad_norm": 3.6075682640075684, "learning_rate": 7.65430179303331e-06, "loss": 0.7657, "step": 16496 }, { "epoch": 1.2928683385579938, "grad_norm": 3.619126796722412, "learning_rate": 7.65404046368866e-06, "loss": 1.7982, "step": 16497 }, { "epoch": 1.2929467084639499, "grad_norm": 2.6489052772521973, "learning_rate": 7.653779124249405e-06, "loss": 1.2725, "step": 16498 }, { "epoch": 1.293025078369906, "grad_norm": 6.118031978607178, "learning_rate": 7.653517774716537e-06, "loss": 1.4387, "step": 16499 }, { "epoch": 1.293103448275862, "grad_norm": 2.7031939029693604, "learning_rate": 7.65325641509105e-06, "loss": 1.5501, "step": 16500 }, { "epoch": 1.2931818181818182, "grad_norm": 3.0142412185668945, "learning_rate": 7.65299504537394e-06, "loss": 1.3832, "step": 16501 }, { "epoch": 1.2932601880877743, "grad_norm": 3.4151549339294434, "learning_rate": 7.652733665566197e-06, "loss": 0.8886, "step": 16502 }, { "epoch": 1.2933385579937304, "grad_norm": 4.301278591156006, "learning_rate": 7.65247227566882e-06, "loss": 1.1488, "step": 16503 }, { "epoch": 1.2934169278996865, "grad_norm": 2.6543235778808594, "learning_rate": 7.652210875682797e-06, "loss": 1.1224, "step": 16504 }, { "epoch": 1.2934952978056427, "grad_norm": 4.022940158843994, "learning_rate": 7.65194946560913e-06, "loss": 1.4025, "step": 16505 }, { "epoch": 1.2935736677115988, "grad_norm": 9.981492042541504, "learning_rate": 7.651688045448807e-06, "loss": 1.3734, "step": 16506 }, { "epoch": 1.2936520376175549, "grad_norm": 3.104168176651001, "learning_rate": 7.651426615202825e-06, "loss": 1.5776, "step": 16507 }, { "epoch": 1.293730407523511, "grad_norm": 1.9496510028839111, "learning_rate": 7.65116517487218e-06, "loss": 1.4112, "step": 16508 }, { "epoch": 1.293808777429467, "grad_norm": 3.115755081176758, "learning_rate": 7.650903724457864e-06, "loss": 1.6544, "step": 16509 }, { "epoch": 1.2938871473354232, "grad_norm": 3.860330104827881, "learning_rate": 7.65064226396087e-06, "loss": 1.1634, "step": 16510 }, { "epoch": 1.2939655172413793, "grad_norm": 2.6268811225891113, "learning_rate": 7.650380793382195e-06, "loss": 1.1923, "step": 16511 }, { "epoch": 1.2940438871473354, "grad_norm": 2.656034231185913, "learning_rate": 7.650119312722833e-06, "loss": 1.2075, "step": 16512 }, { "epoch": 1.2941222570532915, "grad_norm": 2.454899549484253, "learning_rate": 7.649857821983778e-06, "loss": 1.1846, "step": 16513 }, { "epoch": 1.2942006269592476, "grad_norm": 3.5434584617614746, "learning_rate": 7.649596321166024e-06, "loss": 1.5127, "step": 16514 }, { "epoch": 1.2942789968652038, "grad_norm": 2.910693407058716, "learning_rate": 7.64933481027057e-06, "loss": 1.6858, "step": 16515 }, { "epoch": 1.2943573667711599, "grad_norm": 3.527571201324463, "learning_rate": 7.649073289298403e-06, "loss": 1.4472, "step": 16516 }, { "epoch": 1.294435736677116, "grad_norm": 2.9222166538238525, "learning_rate": 7.648811758250525e-06, "loss": 1.6193, "step": 16517 }, { "epoch": 1.294514106583072, "grad_norm": 6.483218193054199, "learning_rate": 7.648550217127926e-06, "loss": 1.827, "step": 16518 }, { "epoch": 1.2945924764890282, "grad_norm": 3.8528060913085938, "learning_rate": 7.648288665931602e-06, "loss": 1.543, "step": 16519 }, { "epoch": 1.2946708463949843, "grad_norm": 3.683649778366089, "learning_rate": 7.64802710466255e-06, "loss": 1.1427, "step": 16520 }, { "epoch": 1.2947492163009404, "grad_norm": 2.9044077396392822, "learning_rate": 7.647765533321762e-06, "loss": 1.1397, "step": 16521 }, { "epoch": 1.2948275862068965, "grad_norm": 5.1130523681640625, "learning_rate": 7.647503951910233e-06, "loss": 1.0458, "step": 16522 }, { "epoch": 1.2949059561128526, "grad_norm": 2.528052806854248, "learning_rate": 7.647242360428958e-06, "loss": 1.5915, "step": 16523 }, { "epoch": 1.2949843260188088, "grad_norm": 2.7349393367767334, "learning_rate": 7.646980758878937e-06, "loss": 1.3457, "step": 16524 }, { "epoch": 1.2950626959247649, "grad_norm": 3.9031448364257812, "learning_rate": 7.646719147261156e-06, "loss": 1.5231, "step": 16525 }, { "epoch": 1.295141065830721, "grad_norm": 3.2620174884796143, "learning_rate": 7.646457525576618e-06, "loss": 1.3844, "step": 16526 }, { "epoch": 1.295219435736677, "grad_norm": 4.06638240814209, "learning_rate": 7.646195893826315e-06, "loss": 1.2036, "step": 16527 }, { "epoch": 1.2952978056426332, "grad_norm": 2.4297664165496826, "learning_rate": 7.64593425201124e-06, "loss": 1.5121, "step": 16528 }, { "epoch": 1.2953761755485893, "grad_norm": 7.047866344451904, "learning_rate": 7.64567260013239e-06, "loss": 1.6518, "step": 16529 }, { "epoch": 1.2954545454545454, "grad_norm": 2.4337644577026367, "learning_rate": 7.645410938190762e-06, "loss": 1.4874, "step": 16530 }, { "epoch": 1.2955329153605015, "grad_norm": 4.251836776733398, "learning_rate": 7.645149266187349e-06, "loss": 1.5813, "step": 16531 }, { "epoch": 1.2956112852664576, "grad_norm": 2.8135693073272705, "learning_rate": 7.644887584123146e-06, "loss": 1.2793, "step": 16532 }, { "epoch": 1.2956896551724137, "grad_norm": 2.2535698413848877, "learning_rate": 7.644625891999151e-06, "loss": 1.3944, "step": 16533 }, { "epoch": 1.2957680250783699, "grad_norm": 2.9191195964813232, "learning_rate": 7.644364189816356e-06, "loss": 1.5042, "step": 16534 }, { "epoch": 1.295846394984326, "grad_norm": 3.8147919178009033, "learning_rate": 7.644102477575759e-06, "loss": 1.5033, "step": 16535 }, { "epoch": 1.295924764890282, "grad_norm": 4.866990566253662, "learning_rate": 7.643840755278353e-06, "loss": 1.8862, "step": 16536 }, { "epoch": 1.2960031347962382, "grad_norm": 3.567401170730591, "learning_rate": 7.643579022925133e-06, "loss": 1.4676, "step": 16537 }, { "epoch": 1.2960815047021943, "grad_norm": 2.1621744632720947, "learning_rate": 7.6433172805171e-06, "loss": 1.0198, "step": 16538 }, { "epoch": 1.2961598746081504, "grad_norm": 5.643536567687988, "learning_rate": 7.643055528055244e-06, "loss": 1.3944, "step": 16539 }, { "epoch": 1.2962382445141065, "grad_norm": 2.7675445079803467, "learning_rate": 7.642793765540561e-06, "loss": 1.3705, "step": 16540 }, { "epoch": 1.2963166144200626, "grad_norm": 3.2757389545440674, "learning_rate": 7.642531992974052e-06, "loss": 1.3286, "step": 16541 }, { "epoch": 1.2963949843260187, "grad_norm": 3.034994125366211, "learning_rate": 7.642270210356703e-06, "loss": 1.3297, "step": 16542 }, { "epoch": 1.2964733542319749, "grad_norm": 3.185872793197632, "learning_rate": 7.64200841768952e-06, "loss": 1.5547, "step": 16543 }, { "epoch": 1.296551724137931, "grad_norm": 5.041598320007324, "learning_rate": 7.641746614973492e-06, "loss": 1.2364, "step": 16544 }, { "epoch": 1.296630094043887, "grad_norm": 6.46451997756958, "learning_rate": 7.641484802209616e-06, "loss": 1.3078, "step": 16545 }, { "epoch": 1.2967084639498432, "grad_norm": 3.121375560760498, "learning_rate": 7.64122297939889e-06, "loss": 1.4833, "step": 16546 }, { "epoch": 1.2967868338557993, "grad_norm": 2.098330497741699, "learning_rate": 7.640961146542307e-06, "loss": 1.3514, "step": 16547 }, { "epoch": 1.2968652037617554, "grad_norm": 2.8841583728790283, "learning_rate": 7.640699303640866e-06, "loss": 1.4813, "step": 16548 }, { "epoch": 1.2969435736677117, "grad_norm": 6.661746025085449, "learning_rate": 7.64043745069556e-06, "loss": 1.5745, "step": 16549 }, { "epoch": 1.2970219435736676, "grad_norm": 2.524019479751587, "learning_rate": 7.640175587707388e-06, "loss": 1.344, "step": 16550 }, { "epoch": 1.297100313479624, "grad_norm": 1.8520630598068237, "learning_rate": 7.639913714677342e-06, "loss": 1.2232, "step": 16551 }, { "epoch": 1.2971786833855798, "grad_norm": 2.4454426765441895, "learning_rate": 7.63965183160642e-06, "loss": 1.3379, "step": 16552 }, { "epoch": 1.2972570532915362, "grad_norm": 1.748419165611267, "learning_rate": 7.63938993849562e-06, "loss": 0.9415, "step": 16553 }, { "epoch": 1.297335423197492, "grad_norm": 2.101567268371582, "learning_rate": 7.639128035345935e-06, "loss": 1.2371, "step": 16554 }, { "epoch": 1.2974137931034484, "grad_norm": 4.228922367095947, "learning_rate": 7.638866122158364e-06, "loss": 1.8771, "step": 16555 }, { "epoch": 1.2974921630094043, "grad_norm": 3.1466076374053955, "learning_rate": 7.6386041989339e-06, "loss": 1.1657, "step": 16556 }, { "epoch": 1.2975705329153606, "grad_norm": 3.433441638946533, "learning_rate": 7.638342265673542e-06, "loss": 1.4314, "step": 16557 }, { "epoch": 1.2976489028213165, "grad_norm": 2.990483045578003, "learning_rate": 7.638080322378284e-06, "loss": 1.0861, "step": 16558 }, { "epoch": 1.2977272727272728, "grad_norm": 2.0709195137023926, "learning_rate": 7.637818369049124e-06, "loss": 1.4167, "step": 16559 }, { "epoch": 1.2978056426332287, "grad_norm": 2.6124866008758545, "learning_rate": 7.63755640568706e-06, "loss": 1.4427, "step": 16560 }, { "epoch": 1.297884012539185, "grad_norm": 3.372972011566162, "learning_rate": 7.637294432293083e-06, "loss": 1.1345, "step": 16561 }, { "epoch": 1.297962382445141, "grad_norm": 2.0566205978393555, "learning_rate": 7.637032448868194e-06, "loss": 1.5787, "step": 16562 }, { "epoch": 1.2980407523510973, "grad_norm": 2.794464111328125, "learning_rate": 7.636770455413387e-06, "loss": 0.7876, "step": 16563 }, { "epoch": 1.2981191222570532, "grad_norm": 3.196183919906616, "learning_rate": 7.636508451929661e-06, "loss": 1.3012, "step": 16564 }, { "epoch": 1.2981974921630095, "grad_norm": 2.6284561157226562, "learning_rate": 7.63624643841801e-06, "loss": 1.468, "step": 16565 }, { "epoch": 1.2982758620689654, "grad_norm": 2.4230194091796875, "learning_rate": 7.635984414879432e-06, "loss": 1.5813, "step": 16566 }, { "epoch": 1.2983542319749217, "grad_norm": 2.9863994121551514, "learning_rate": 7.635722381314923e-06, "loss": 1.1714, "step": 16567 }, { "epoch": 1.2984326018808776, "grad_norm": 2.892989158630371, "learning_rate": 7.635460337725479e-06, "loss": 1.6991, "step": 16568 }, { "epoch": 1.298510971786834, "grad_norm": 3.9785728454589844, "learning_rate": 7.6351982841121e-06, "loss": 1.341, "step": 16569 }, { "epoch": 1.2985893416927898, "grad_norm": 2.694089651107788, "learning_rate": 7.634936220475777e-06, "loss": 1.4624, "step": 16570 }, { "epoch": 1.2986677115987462, "grad_norm": 3.061006546020508, "learning_rate": 7.634674146817514e-06, "loss": 1.7081, "step": 16571 }, { "epoch": 1.2987460815047023, "grad_norm": 2.7055957317352295, "learning_rate": 7.6344120631383e-06, "loss": 1.0925, "step": 16572 }, { "epoch": 1.2988244514106584, "grad_norm": 2.5854532718658447, "learning_rate": 7.634149969439138e-06, "loss": 1.3007, "step": 16573 }, { "epoch": 1.2989028213166145, "grad_norm": 2.0764076709747314, "learning_rate": 7.633887865721021e-06, "loss": 1.4743, "step": 16574 }, { "epoch": 1.2989811912225706, "grad_norm": 2.8567845821380615, "learning_rate": 7.633625751984949e-06, "loss": 1.6555, "step": 16575 }, { "epoch": 1.2990595611285267, "grad_norm": 2.1940596103668213, "learning_rate": 7.633363628231917e-06, "loss": 1.5716, "step": 16576 }, { "epoch": 1.2991379310344828, "grad_norm": 3.4744863510131836, "learning_rate": 7.63310149446292e-06, "loss": 1.0717, "step": 16577 }, { "epoch": 1.299216300940439, "grad_norm": 2.7471165657043457, "learning_rate": 7.63283935067896e-06, "loss": 1.3957, "step": 16578 }, { "epoch": 1.299294670846395, "grad_norm": 2.0466959476470947, "learning_rate": 7.63257719688103e-06, "loss": 1.4555, "step": 16579 }, { "epoch": 1.2993730407523512, "grad_norm": 2.1082916259765625, "learning_rate": 7.63231503307013e-06, "loss": 1.3826, "step": 16580 }, { "epoch": 1.2994514106583073, "grad_norm": 5.2828049659729, "learning_rate": 7.632052859247254e-06, "loss": 1.3245, "step": 16581 }, { "epoch": 1.2995297805642634, "grad_norm": 2.68257212638855, "learning_rate": 7.631790675413403e-06, "loss": 0.7962, "step": 16582 }, { "epoch": 1.2996081504702195, "grad_norm": 2.451580286026001, "learning_rate": 7.63152848156957e-06, "loss": 1.2961, "step": 16583 }, { "epoch": 1.2996865203761756, "grad_norm": 2.652113437652588, "learning_rate": 7.631266277716755e-06, "loss": 1.5542, "step": 16584 }, { "epoch": 1.2997648902821317, "grad_norm": 4.910404205322266, "learning_rate": 7.631004063855953e-06, "loss": 1.3205, "step": 16585 }, { "epoch": 1.2998432601880878, "grad_norm": 3.067051410675049, "learning_rate": 7.630741839988164e-06, "loss": 1.4255, "step": 16586 }, { "epoch": 1.299921630094044, "grad_norm": 4.382308483123779, "learning_rate": 7.630479606114387e-06, "loss": 1.4425, "step": 16587 }, { "epoch": 1.3, "grad_norm": 4.894184589385986, "learning_rate": 7.630217362235615e-06, "loss": 1.4613, "step": 16588 }, { "epoch": 1.3000783699059562, "grad_norm": 2.510474681854248, "learning_rate": 7.629955108352846e-06, "loss": 1.3139, "step": 16589 }, { "epoch": 1.3001567398119123, "grad_norm": 3.4961817264556885, "learning_rate": 7.629692844467078e-06, "loss": 1.6811, "step": 16590 }, { "epoch": 1.3002351097178684, "grad_norm": 1.7908865213394165, "learning_rate": 7.62943057057931e-06, "loss": 1.2914, "step": 16591 }, { "epoch": 1.3003134796238245, "grad_norm": 3.190248727798462, "learning_rate": 7.62916828669054e-06, "loss": 1.3913, "step": 16592 }, { "epoch": 1.3003918495297806, "grad_norm": 2.9592998027801514, "learning_rate": 7.628905992801763e-06, "loss": 1.2388, "step": 16593 }, { "epoch": 1.3004702194357367, "grad_norm": 2.666811227798462, "learning_rate": 7.628643688913977e-06, "loss": 1.4253, "step": 16594 }, { "epoch": 1.3005485893416928, "grad_norm": 3.7618651390075684, "learning_rate": 7.628381375028183e-06, "loss": 1.5931, "step": 16595 }, { "epoch": 1.300626959247649, "grad_norm": 2.901804208755493, "learning_rate": 7.628119051145374e-06, "loss": 1.2351, "step": 16596 }, { "epoch": 1.300705329153605, "grad_norm": 2.449510097503662, "learning_rate": 7.627856717266552e-06, "loss": 1.0969, "step": 16597 }, { "epoch": 1.3007836990595611, "grad_norm": 2.8723626136779785, "learning_rate": 7.62759437339271e-06, "loss": 1.1952, "step": 16598 }, { "epoch": 1.3008620689655173, "grad_norm": 7.07858419418335, "learning_rate": 7.627332019524851e-06, "loss": 2.1411, "step": 16599 }, { "epoch": 1.3009404388714734, "grad_norm": 6.139581203460693, "learning_rate": 7.62706965566397e-06, "loss": 1.321, "step": 16600 }, { "epoch": 1.3010188087774295, "grad_norm": 2.71591854095459, "learning_rate": 7.626807281811066e-06, "loss": 1.425, "step": 16601 }, { "epoch": 1.3010971786833856, "grad_norm": 3.6571125984191895, "learning_rate": 7.626544897967134e-06, "loss": 2.0439, "step": 16602 }, { "epoch": 1.3011755485893417, "grad_norm": 3.5219035148620605, "learning_rate": 7.626282504133176e-06, "loss": 1.2992, "step": 16603 }, { "epoch": 1.3012539184952978, "grad_norm": 3.717193603515625, "learning_rate": 7.626020100310189e-06, "loss": 1.3391, "step": 16604 }, { "epoch": 1.301332288401254, "grad_norm": 5.32027530670166, "learning_rate": 7.625757686499169e-06, "loss": 1.2427, "step": 16605 }, { "epoch": 1.30141065830721, "grad_norm": 2.2123069763183594, "learning_rate": 7.625495262701117e-06, "loss": 1.5994, "step": 16606 }, { "epoch": 1.3014890282131661, "grad_norm": 2.3199074268341064, "learning_rate": 7.625232828917029e-06, "loss": 1.0527, "step": 16607 }, { "epoch": 1.3015673981191223, "grad_norm": 2.729884147644043, "learning_rate": 7.6249703851479016e-06, "loss": 0.8533, "step": 16608 }, { "epoch": 1.3016457680250784, "grad_norm": 8.50602912902832, "learning_rate": 7.624707931394737e-06, "loss": 1.4475, "step": 16609 }, { "epoch": 1.3017241379310345, "grad_norm": 3.363072633743286, "learning_rate": 7.624445467658531e-06, "loss": 1.6663, "step": 16610 }, { "epoch": 1.3018025078369906, "grad_norm": 3.7438409328460693, "learning_rate": 7.624182993940282e-06, "loss": 1.0569, "step": 16611 }, { "epoch": 1.3018808777429467, "grad_norm": 2.9438424110412598, "learning_rate": 7.623920510240989e-06, "loss": 1.0208, "step": 16612 }, { "epoch": 1.3019592476489028, "grad_norm": 2.1281652450561523, "learning_rate": 7.623658016561651e-06, "loss": 1.1185, "step": 16613 }, { "epoch": 1.302037617554859, "grad_norm": 3.6587228775024414, "learning_rate": 7.6233955129032645e-06, "loss": 1.621, "step": 16614 }, { "epoch": 1.302115987460815, "grad_norm": 3.533362865447998, "learning_rate": 7.6231329992668305e-06, "loss": 1.1807, "step": 16615 }, { "epoch": 1.3021943573667711, "grad_norm": 2.6672370433807373, "learning_rate": 7.622870475653345e-06, "loss": 1.2096, "step": 16616 }, { "epoch": 1.3022727272727272, "grad_norm": 3.4780194759368896, "learning_rate": 7.622607942063808e-06, "loss": 1.3735, "step": 16617 }, { "epoch": 1.3023510971786834, "grad_norm": 4.071853160858154, "learning_rate": 7.622345398499215e-06, "loss": 1.3766, "step": 16618 }, { "epoch": 1.3024294670846395, "grad_norm": 3.2704577445983887, "learning_rate": 7.6220828449605695e-06, "loss": 1.2491, "step": 16619 }, { "epoch": 1.3025078369905956, "grad_norm": 2.4362852573394775, "learning_rate": 7.621820281448867e-06, "loss": 1.2474, "step": 16620 }, { "epoch": 1.3025862068965517, "grad_norm": 2.8033645153045654, "learning_rate": 7.621557707965106e-06, "loss": 1.2421, "step": 16621 }, { "epoch": 1.3026645768025078, "grad_norm": 2.6374990940093994, "learning_rate": 7.621295124510287e-06, "loss": 1.5255, "step": 16622 }, { "epoch": 1.302742946708464, "grad_norm": 5.017779350280762, "learning_rate": 7.621032531085408e-06, "loss": 1.0869, "step": 16623 }, { "epoch": 1.30282131661442, "grad_norm": 3.8311729431152344, "learning_rate": 7.620769927691467e-06, "loss": 1.2612, "step": 16624 }, { "epoch": 1.3028996865203761, "grad_norm": 3.489100217819214, "learning_rate": 7.620507314329464e-06, "loss": 1.0997, "step": 16625 }, { "epoch": 1.3029780564263322, "grad_norm": 8.093880653381348, "learning_rate": 7.620244691000395e-06, "loss": 0.9649, "step": 16626 }, { "epoch": 1.3030564263322884, "grad_norm": 3.186898946762085, "learning_rate": 7.619982057705263e-06, "loss": 1.4329, "step": 16627 }, { "epoch": 1.3031347962382445, "grad_norm": 3.8867812156677246, "learning_rate": 7.619719414445064e-06, "loss": 1.3374, "step": 16628 }, { "epoch": 1.3032131661442006, "grad_norm": 2.18774151802063, "learning_rate": 7.6194567612207984e-06, "loss": 1.2849, "step": 16629 }, { "epoch": 1.3032915360501567, "grad_norm": 2.683929443359375, "learning_rate": 7.619194098033466e-06, "loss": 1.5903, "step": 16630 }, { "epoch": 1.3033699059561128, "grad_norm": 3.981588840484619, "learning_rate": 7.6189314248840615e-06, "loss": 1.4008, "step": 16631 }, { "epoch": 1.303448275862069, "grad_norm": 2.3757615089416504, "learning_rate": 7.61866874177359e-06, "loss": 1.7988, "step": 16632 }, { "epoch": 1.303526645768025, "grad_norm": 2.5813255310058594, "learning_rate": 7.618406048703045e-06, "loss": 1.4566, "step": 16633 }, { "epoch": 1.3036050156739811, "grad_norm": 3.7237064838409424, "learning_rate": 7.61814334567343e-06, "loss": 1.0842, "step": 16634 }, { "epoch": 1.3036833855799372, "grad_norm": 4.764947414398193, "learning_rate": 7.617880632685742e-06, "loss": 1.5278, "step": 16635 }, { "epoch": 1.3037617554858933, "grad_norm": 2.405763626098633, "learning_rate": 7.617617909740982e-06, "loss": 1.1239, "step": 16636 }, { "epoch": 1.3038401253918495, "grad_norm": 3.8771884441375732, "learning_rate": 7.617355176840147e-06, "loss": 1.5801, "step": 16637 }, { "epoch": 1.3039184952978056, "grad_norm": 3.5175552368164062, "learning_rate": 7.617092433984236e-06, "loss": 1.1276, "step": 16638 }, { "epoch": 1.3039968652037617, "grad_norm": 3.6776041984558105, "learning_rate": 7.61682968117425e-06, "loss": 1.2595, "step": 16639 }, { "epoch": 1.3040752351097178, "grad_norm": 5.063980579376221, "learning_rate": 7.616566918411189e-06, "loss": 1.5211, "step": 16640 }, { "epoch": 1.304153605015674, "grad_norm": 4.74043083190918, "learning_rate": 7.6163041456960505e-06, "loss": 1.502, "step": 16641 }, { "epoch": 1.3042319749216302, "grad_norm": 12.14653205871582, "learning_rate": 7.616041363029836e-06, "loss": 1.4116, "step": 16642 }, { "epoch": 1.3043103448275861, "grad_norm": 7.895711421966553, "learning_rate": 7.6157785704135435e-06, "loss": 0.835, "step": 16643 }, { "epoch": 1.3043887147335425, "grad_norm": 5.664938449859619, "learning_rate": 7.6155157678481725e-06, "loss": 1.2993, "step": 16644 }, { "epoch": 1.3044670846394983, "grad_norm": 2.1691319942474365, "learning_rate": 7.615252955334723e-06, "loss": 1.118, "step": 16645 }, { "epoch": 1.3045454545454547, "grad_norm": 2.8769948482513428, "learning_rate": 7.614990132874194e-06, "loss": 1.135, "step": 16646 }, { "epoch": 1.3046238244514106, "grad_norm": 3.00679349899292, "learning_rate": 7.6147273004675855e-06, "loss": 1.6268, "step": 16647 }, { "epoch": 1.304702194357367, "grad_norm": 2.891235589981079, "learning_rate": 7.614464458115899e-06, "loss": 1.266, "step": 16648 }, { "epoch": 1.3047805642633228, "grad_norm": 3.3361332416534424, "learning_rate": 7.6142016058201325e-06, "loss": 1.2336, "step": 16649 }, { "epoch": 1.3048589341692791, "grad_norm": 2.007606267929077, "learning_rate": 7.613938743581284e-06, "loss": 0.9887, "step": 16650 }, { "epoch": 1.304937304075235, "grad_norm": 2.5959177017211914, "learning_rate": 7.613675871400357e-06, "loss": 1.4451, "step": 16651 }, { "epoch": 1.3050156739811913, "grad_norm": 2.4399940967559814, "learning_rate": 7.613412989278348e-06, "loss": 1.0522, "step": 16652 }, { "epoch": 1.3050940438871472, "grad_norm": 2.1626181602478027, "learning_rate": 7.61315009721626e-06, "loss": 1.1881, "step": 16653 }, { "epoch": 1.3051724137931036, "grad_norm": 2.947742462158203, "learning_rate": 7.612887195215089e-06, "loss": 1.3508, "step": 16654 }, { "epoch": 1.3052507836990594, "grad_norm": 10.411250114440918, "learning_rate": 7.612624283275838e-06, "loss": 0.9985, "step": 16655 }, { "epoch": 1.3053291536050158, "grad_norm": 3.333345890045166, "learning_rate": 7.612361361399507e-06, "loss": 1.4158, "step": 16656 }, { "epoch": 1.3054075235109717, "grad_norm": 3.2426888942718506, "learning_rate": 7.612098429587093e-06, "loss": 1.4447, "step": 16657 }, { "epoch": 1.305485893416928, "grad_norm": 4.641050338745117, "learning_rate": 7.611835487839601e-06, "loss": 1.5355, "step": 16658 }, { "epoch": 1.3055642633228839, "grad_norm": 2.376335859298706, "learning_rate": 7.611572536158027e-06, "loss": 1.1908, "step": 16659 }, { "epoch": 1.3056426332288402, "grad_norm": 4.292912006378174, "learning_rate": 7.611309574543373e-06, "loss": 1.7083, "step": 16660 }, { "epoch": 1.305721003134796, "grad_norm": 1.9800652265548706, "learning_rate": 7.6110466029966375e-06, "loss": 0.9748, "step": 16661 }, { "epoch": 1.3057993730407524, "grad_norm": 3.447007894515991, "learning_rate": 7.610783621518824e-06, "loss": 1.1442, "step": 16662 }, { "epoch": 1.3058777429467083, "grad_norm": 9.829310417175293, "learning_rate": 7.610520630110929e-06, "loss": 1.176, "step": 16663 }, { "epoch": 1.3059561128526647, "grad_norm": 10.255220413208008, "learning_rate": 7.610257628773954e-06, "loss": 2.0431, "step": 16664 }, { "epoch": 1.3060344827586206, "grad_norm": 7.544529914855957, "learning_rate": 7.6099946175089e-06, "loss": 1.6601, "step": 16665 }, { "epoch": 1.3061128526645769, "grad_norm": 2.8222389221191406, "learning_rate": 7.6097315963167675e-06, "loss": 1.2017, "step": 16666 }, { "epoch": 1.306191222570533, "grad_norm": 3.1106209754943848, "learning_rate": 7.609468565198556e-06, "loss": 1.8094, "step": 16667 }, { "epoch": 1.306269592476489, "grad_norm": 3.71807861328125, "learning_rate": 7.609205524155268e-06, "loss": 1.1884, "step": 16668 }, { "epoch": 1.3063479623824452, "grad_norm": 6.574819087982178, "learning_rate": 7.608942473187899e-06, "loss": 0.5667, "step": 16669 }, { "epoch": 1.3064263322884013, "grad_norm": 3.988954544067383, "learning_rate": 7.608679412297458e-06, "loss": 1.8978, "step": 16670 }, { "epoch": 1.3065047021943574, "grad_norm": 2.7457644939422607, "learning_rate": 7.608416341484935e-06, "loss": 1.6003, "step": 16671 }, { "epoch": 1.3065830721003135, "grad_norm": 3.258857488632202, "learning_rate": 7.608153260751338e-06, "loss": 1.2745, "step": 16672 }, { "epoch": 1.3066614420062697, "grad_norm": 6.195892333984375, "learning_rate": 7.607890170097667e-06, "loss": 1.4037, "step": 16673 }, { "epoch": 1.3067398119122258, "grad_norm": 3.109868049621582, "learning_rate": 7.6076270695249195e-06, "loss": 1.6157, "step": 16674 }, { "epoch": 1.3068181818181819, "grad_norm": 8.448070526123047, "learning_rate": 7.607363959034099e-06, "loss": 1.2871, "step": 16675 }, { "epoch": 1.306896551724138, "grad_norm": 2.731595754623413, "learning_rate": 7.607100838626205e-06, "loss": 1.6629, "step": 16676 }, { "epoch": 1.306974921630094, "grad_norm": 2.673252582550049, "learning_rate": 7.60683770830224e-06, "loss": 1.3144, "step": 16677 }, { "epoch": 1.3070532915360502, "grad_norm": 5.001166820526123, "learning_rate": 7.6065745680632e-06, "loss": 1.5497, "step": 16678 }, { "epoch": 1.3071316614420063, "grad_norm": 3.746015787124634, "learning_rate": 7.606311417910092e-06, "loss": 1.1148, "step": 16679 }, { "epoch": 1.3072100313479624, "grad_norm": 3.101151466369629, "learning_rate": 7.606048257843912e-06, "loss": 1.2466, "step": 16680 }, { "epoch": 1.3072884012539185, "grad_norm": 2.726555585861206, "learning_rate": 7.605785087865665e-06, "loss": 1.005, "step": 16681 }, { "epoch": 1.3073667711598747, "grad_norm": 4.891838550567627, "learning_rate": 7.605521907976348e-06, "loss": 1.8206, "step": 16682 }, { "epoch": 1.3074451410658308, "grad_norm": 2.3664982318878174, "learning_rate": 7.605258718176964e-06, "loss": 1.3136, "step": 16683 }, { "epoch": 1.3075235109717869, "grad_norm": 4.082950115203857, "learning_rate": 7.604995518468515e-06, "loss": 1.6789, "step": 16684 }, { "epoch": 1.307601880877743, "grad_norm": 3.049487352371216, "learning_rate": 7.604732308852001e-06, "loss": 1.0334, "step": 16685 }, { "epoch": 1.307680250783699, "grad_norm": 2.9798636436462402, "learning_rate": 7.604469089328423e-06, "loss": 1.6605, "step": 16686 }, { "epoch": 1.3077586206896552, "grad_norm": 4.445632457733154, "learning_rate": 7.604205859898783e-06, "loss": 1.1179, "step": 16687 }, { "epoch": 1.3078369905956113, "grad_norm": 2.3795645236968994, "learning_rate": 7.603942620564081e-06, "loss": 1.2344, "step": 16688 }, { "epoch": 1.3079153605015674, "grad_norm": 2.2829761505126953, "learning_rate": 7.603679371325319e-06, "loss": 1.3612, "step": 16689 }, { "epoch": 1.3079937304075235, "grad_norm": 3.040189743041992, "learning_rate": 7.603416112183497e-06, "loss": 1.5921, "step": 16690 }, { "epoch": 1.3080721003134796, "grad_norm": 3.6430516242980957, "learning_rate": 7.603152843139619e-06, "loss": 1.6347, "step": 16691 }, { "epoch": 1.3081504702194358, "grad_norm": 2.8445661067962646, "learning_rate": 7.602889564194682e-06, "loss": 1.4908, "step": 16692 }, { "epoch": 1.3082288401253919, "grad_norm": 2.328861713409424, "learning_rate": 7.602626275349691e-06, "loss": 1.3606, "step": 16693 }, { "epoch": 1.308307210031348, "grad_norm": 3.108671188354492, "learning_rate": 7.602362976605649e-06, "loss": 1.3867, "step": 16694 }, { "epoch": 1.308385579937304, "grad_norm": 4.064202785491943, "learning_rate": 7.602099667963552e-06, "loss": 1.04, "step": 16695 }, { "epoch": 1.3084639498432602, "grad_norm": 3.481959342956543, "learning_rate": 7.601836349424405e-06, "loss": 1.3323, "step": 16696 }, { "epoch": 1.3085423197492163, "grad_norm": 2.0853941440582275, "learning_rate": 7.601573020989209e-06, "loss": 1.2793, "step": 16697 }, { "epoch": 1.3086206896551724, "grad_norm": 2.6654257774353027, "learning_rate": 7.601309682658966e-06, "loss": 0.9952, "step": 16698 }, { "epoch": 1.3086990595611285, "grad_norm": 2.2354676723480225, "learning_rate": 7.601046334434677e-06, "loss": 1.0817, "step": 16699 }, { "epoch": 1.3087774294670846, "grad_norm": 3.407175302505493, "learning_rate": 7.6007829763173445e-06, "loss": 1.3659, "step": 16700 }, { "epoch": 1.3088557993730408, "grad_norm": 2.832439661026001, "learning_rate": 7.600519608307968e-06, "loss": 1.3634, "step": 16701 }, { "epoch": 1.3089341692789969, "grad_norm": 11.492843627929688, "learning_rate": 7.60025623040755e-06, "loss": 1.6841, "step": 16702 }, { "epoch": 1.309012539184953, "grad_norm": 5.495255470275879, "learning_rate": 7.599992842617094e-06, "loss": 1.616, "step": 16703 }, { "epoch": 1.309090909090909, "grad_norm": 4.899061679840088, "learning_rate": 7.5997294449376e-06, "loss": 1.8153, "step": 16704 }, { "epoch": 1.3091692789968652, "grad_norm": 2.966017007827759, "learning_rate": 7.599466037370073e-06, "loss": 1.2405, "step": 16705 }, { "epoch": 1.3092476489028213, "grad_norm": 4.6728057861328125, "learning_rate": 7.599202619915509e-06, "loss": 1.7238, "step": 16706 }, { "epoch": 1.3093260188087774, "grad_norm": 3.596491813659668, "learning_rate": 7.598939192574915e-06, "loss": 1.4844, "step": 16707 }, { "epoch": 1.3094043887147335, "grad_norm": 5.484676837921143, "learning_rate": 7.598675755349292e-06, "loss": 1.402, "step": 16708 }, { "epoch": 1.3094827586206896, "grad_norm": 2.224348783493042, "learning_rate": 7.59841230823964e-06, "loss": 1.2699, "step": 16709 }, { "epoch": 1.3095611285266457, "grad_norm": 3.6734559535980225, "learning_rate": 7.598148851246962e-06, "loss": 1.5875, "step": 16710 }, { "epoch": 1.3096394984326019, "grad_norm": 4.4467926025390625, "learning_rate": 7.597885384372261e-06, "loss": 1.7942, "step": 16711 }, { "epoch": 1.309717868338558, "grad_norm": 2.347022294998169, "learning_rate": 7.597621907616537e-06, "loss": 1.4673, "step": 16712 }, { "epoch": 1.309796238244514, "grad_norm": 3.2257540225982666, "learning_rate": 7.5973584209807945e-06, "loss": 1.1875, "step": 16713 }, { "epoch": 1.3098746081504702, "grad_norm": 4.209676265716553, "learning_rate": 7.597094924466035e-06, "loss": 1.4857, "step": 16714 }, { "epoch": 1.3099529780564263, "grad_norm": 4.3672404289245605, "learning_rate": 7.596831418073259e-06, "loss": 1.4603, "step": 16715 }, { "epoch": 1.3100313479623824, "grad_norm": 2.53912091255188, "learning_rate": 7.596567901803473e-06, "loss": 1.1998, "step": 16716 }, { "epoch": 1.3101097178683385, "grad_norm": 2.7422354221343994, "learning_rate": 7.5963043756576735e-06, "loss": 1.4895, "step": 16717 }, { "epoch": 1.3101880877742946, "grad_norm": 2.444643259048462, "learning_rate": 7.5960408396368665e-06, "loss": 1.7471, "step": 16718 }, { "epoch": 1.3102664576802507, "grad_norm": 2.951838254928589, "learning_rate": 7.595777293742054e-06, "loss": 1.2738, "step": 16719 }, { "epoch": 1.3103448275862069, "grad_norm": 3.0045838356018066, "learning_rate": 7.595513737974238e-06, "loss": 1.1116, "step": 16720 }, { "epoch": 1.310423197492163, "grad_norm": 2.8480069637298584, "learning_rate": 7.5952501723344205e-06, "loss": 1.3736, "step": 16721 }, { "epoch": 1.310501567398119, "grad_norm": 7.285274505615234, "learning_rate": 7.594986596823603e-06, "loss": 1.8682, "step": 16722 }, { "epoch": 1.3105799373040752, "grad_norm": 2.520263671875, "learning_rate": 7.594723011442792e-06, "loss": 1.4454, "step": 16723 }, { "epoch": 1.3106583072100313, "grad_norm": 3.830489158630371, "learning_rate": 7.594459416192986e-06, "loss": 1.4901, "step": 16724 }, { "epoch": 1.3107366771159874, "grad_norm": 4.827056407928467, "learning_rate": 7.594195811075189e-06, "loss": 1.0115, "step": 16725 }, { "epoch": 1.3108150470219435, "grad_norm": 2.1074793338775635, "learning_rate": 7.5939321960904036e-06, "loss": 1.2394, "step": 16726 }, { "epoch": 1.3108934169278996, "grad_norm": 3.530158519744873, "learning_rate": 7.593668571239633e-06, "loss": 1.6373, "step": 16727 }, { "epoch": 1.3109717868338557, "grad_norm": 4.714639186859131, "learning_rate": 7.593404936523878e-06, "loss": 1.4971, "step": 16728 }, { "epoch": 1.3110501567398118, "grad_norm": 2.0284390449523926, "learning_rate": 7.593141291944144e-06, "loss": 1.3041, "step": 16729 }, { "epoch": 1.311128526645768, "grad_norm": 3.463137149810791, "learning_rate": 7.592877637501432e-06, "loss": 0.8641, "step": 16730 }, { "epoch": 1.311206896551724, "grad_norm": 2.295969009399414, "learning_rate": 7.592613973196746e-06, "loss": 1.3428, "step": 16731 }, { "epoch": 1.3112852664576802, "grad_norm": 3.9811856746673584, "learning_rate": 7.592350299031086e-06, "loss": 1.7429, "step": 16732 }, { "epoch": 1.3113636363636363, "grad_norm": 8.250802993774414, "learning_rate": 7.592086615005461e-06, "loss": 1.6215, "step": 16733 }, { "epoch": 1.3114420062695924, "grad_norm": 4.7793121337890625, "learning_rate": 7.591822921120865e-06, "loss": 1.7531, "step": 16734 }, { "epoch": 1.3115203761755485, "grad_norm": 1.8334988355636597, "learning_rate": 7.5915592173783095e-06, "loss": 1.1688, "step": 16735 }, { "epoch": 1.3115987460815046, "grad_norm": 8.28044605255127, "learning_rate": 7.591295503778792e-06, "loss": 1.2696, "step": 16736 }, { "epoch": 1.311677115987461, "grad_norm": 2.5400350093841553, "learning_rate": 7.591031780323317e-06, "loss": 1.2739, "step": 16737 }, { "epoch": 1.3117554858934168, "grad_norm": 2.4977071285247803, "learning_rate": 7.5907680470128885e-06, "loss": 1.42, "step": 16738 }, { "epoch": 1.3118338557993732, "grad_norm": 6.145055294036865, "learning_rate": 7.590504303848509e-06, "loss": 1.1534, "step": 16739 }, { "epoch": 1.311912225705329, "grad_norm": 2.590034008026123, "learning_rate": 7.590240550831183e-06, "loss": 0.9259, "step": 16740 }, { "epoch": 1.3119905956112854, "grad_norm": 2.175999402999878, "learning_rate": 7.589976787961911e-06, "loss": 1.658, "step": 16741 }, { "epoch": 1.3120689655172413, "grad_norm": 3.960782527923584, "learning_rate": 7.5897130152417e-06, "loss": 1.6757, "step": 16742 }, { "epoch": 1.3121473354231976, "grad_norm": 3.6194000244140625, "learning_rate": 7.5894492326715485e-06, "loss": 1.1838, "step": 16743 }, { "epoch": 1.3122257053291535, "grad_norm": 2.356693983078003, "learning_rate": 7.589185440252462e-06, "loss": 1.4057, "step": 16744 }, { "epoch": 1.3123040752351098, "grad_norm": 2.322251319885254, "learning_rate": 7.588921637985445e-06, "loss": 1.3788, "step": 16745 }, { "epoch": 1.3123824451410657, "grad_norm": 2.8408925533294678, "learning_rate": 7.588657825871498e-06, "loss": 1.4889, "step": 16746 }, { "epoch": 1.312460815047022, "grad_norm": 2.7843315601348877, "learning_rate": 7.588394003911628e-06, "loss": 1.1264, "step": 16747 }, { "epoch": 1.312539184952978, "grad_norm": 3.422258138656616, "learning_rate": 7.588130172106835e-06, "loss": 1.5176, "step": 16748 }, { "epoch": 1.3126175548589343, "grad_norm": 2.2011172771453857, "learning_rate": 7.5878663304581265e-06, "loss": 1.0308, "step": 16749 }, { "epoch": 1.3126959247648902, "grad_norm": 2.672222852706909, "learning_rate": 7.587602478966503e-06, "loss": 1.3516, "step": 16750 }, { "epoch": 1.3127742946708465, "grad_norm": 2.2001163959503174, "learning_rate": 7.587338617632967e-06, "loss": 1.7115, "step": 16751 }, { "epoch": 1.3128526645768024, "grad_norm": 3.482722520828247, "learning_rate": 7.587074746458527e-06, "loss": 1.2422, "step": 16752 }, { "epoch": 1.3129310344827587, "grad_norm": 4.018467426300049, "learning_rate": 7.5868108654441805e-06, "loss": 1.8847, "step": 16753 }, { "epoch": 1.3130094043887146, "grad_norm": 2.626631498336792, "learning_rate": 7.5865469745909356e-06, "loss": 1.4563, "step": 16754 }, { "epoch": 1.313087774294671, "grad_norm": 3.5660853385925293, "learning_rate": 7.586283073899794e-06, "loss": 1.54, "step": 16755 }, { "epoch": 1.3131661442006268, "grad_norm": 1.7910740375518799, "learning_rate": 7.58601916337176e-06, "loss": 1.2353, "step": 16756 }, { "epoch": 1.3132445141065832, "grad_norm": 2.7609684467315674, "learning_rate": 7.585755243007838e-06, "loss": 1.1845, "step": 16757 }, { "epoch": 1.313322884012539, "grad_norm": 8.19917106628418, "learning_rate": 7.58549131280903e-06, "loss": 1.9416, "step": 16758 }, { "epoch": 1.3134012539184954, "grad_norm": 6.116833209991455, "learning_rate": 7.585227372776343e-06, "loss": 0.9144, "step": 16759 }, { "epoch": 1.3134796238244515, "grad_norm": 2.220292091369629, "learning_rate": 7.584963422910777e-06, "loss": 0.7314, "step": 16760 }, { "epoch": 1.3135579937304076, "grad_norm": 2.2684080600738525, "learning_rate": 7.5846994632133395e-06, "loss": 0.7282, "step": 16761 }, { "epoch": 1.3136363636363637, "grad_norm": 5.965960502624512, "learning_rate": 7.584435493685032e-06, "loss": 1.5849, "step": 16762 }, { "epoch": 1.3137147335423198, "grad_norm": 10.918498992919922, "learning_rate": 7.584171514326859e-06, "loss": 1.5497, "step": 16763 }, { "epoch": 1.313793103448276, "grad_norm": 3.001494884490967, "learning_rate": 7.583907525139825e-06, "loss": 1.5238, "step": 16764 }, { "epoch": 1.313871473354232, "grad_norm": 2.7167444229125977, "learning_rate": 7.583643526124935e-06, "loss": 1.4031, "step": 16765 }, { "epoch": 1.3139498432601882, "grad_norm": 3.4075851440429688, "learning_rate": 7.5833795172831905e-06, "loss": 1.14, "step": 16766 }, { "epoch": 1.3140282131661443, "grad_norm": 2.534247636795044, "learning_rate": 7.5831154986155985e-06, "loss": 1.6691, "step": 16767 }, { "epoch": 1.3141065830721004, "grad_norm": 1.9070812463760376, "learning_rate": 7.5828514701231605e-06, "loss": 1.3385, "step": 16768 }, { "epoch": 1.3141849529780565, "grad_norm": 2.5587024688720703, "learning_rate": 7.582587431806883e-06, "loss": 1.3515, "step": 16769 }, { "epoch": 1.3142633228840126, "grad_norm": 4.73985481262207, "learning_rate": 7.58232338366777e-06, "loss": 1.2292, "step": 16770 }, { "epoch": 1.3143416927899687, "grad_norm": 2.6020684242248535, "learning_rate": 7.582059325706826e-06, "loss": 1.6544, "step": 16771 }, { "epoch": 1.3144200626959248, "grad_norm": 2.8640928268432617, "learning_rate": 7.581795257925051e-06, "loss": 1.4335, "step": 16772 }, { "epoch": 1.314498432601881, "grad_norm": 3.4018914699554443, "learning_rate": 7.581531180323456e-06, "loss": 1.4064, "step": 16773 }, { "epoch": 1.314576802507837, "grad_norm": 2.8211100101470947, "learning_rate": 7.581267092903041e-06, "loss": 1.5374, "step": 16774 }, { "epoch": 1.3146551724137931, "grad_norm": 2.8670804500579834, "learning_rate": 7.581002995664812e-06, "loss": 1.3852, "step": 16775 }, { "epoch": 1.3147335423197493, "grad_norm": 4.359316349029541, "learning_rate": 7.580738888609774e-06, "loss": 1.3347, "step": 16776 }, { "epoch": 1.3148119122257054, "grad_norm": 4.10687255859375, "learning_rate": 7.58047477173893e-06, "loss": 1.4399, "step": 16777 }, { "epoch": 1.3148902821316615, "grad_norm": 2.5850062370300293, "learning_rate": 7.580210645053286e-06, "loss": 1.4166, "step": 16778 }, { "epoch": 1.3149686520376176, "grad_norm": 3.23549222946167, "learning_rate": 7.579946508553845e-06, "loss": 0.733, "step": 16779 }, { "epoch": 1.3150470219435737, "grad_norm": 2.5225133895874023, "learning_rate": 7.579682362241613e-06, "loss": 1.2655, "step": 16780 }, { "epoch": 1.3151253918495298, "grad_norm": 3.1229004859924316, "learning_rate": 7.579418206117594e-06, "loss": 1.5919, "step": 16781 }, { "epoch": 1.315203761755486, "grad_norm": 3.6362340450286865, "learning_rate": 7.579154040182794e-06, "loss": 1.1789, "step": 16782 }, { "epoch": 1.315282131661442, "grad_norm": 2.484971046447754, "learning_rate": 7.578889864438217e-06, "loss": 1.323, "step": 16783 }, { "epoch": 1.3153605015673981, "grad_norm": 2.9768805503845215, "learning_rate": 7.578625678884868e-06, "loss": 1.4509, "step": 16784 }, { "epoch": 1.3154388714733543, "grad_norm": 6.065636157989502, "learning_rate": 7.57836148352375e-06, "loss": 1.4421, "step": 16785 }, { "epoch": 1.3155172413793104, "grad_norm": 3.0137133598327637, "learning_rate": 7.578097278355869e-06, "loss": 1.6261, "step": 16786 }, { "epoch": 1.3155956112852665, "grad_norm": 2.221318483352661, "learning_rate": 7.577833063382232e-06, "loss": 1.2845, "step": 16787 }, { "epoch": 1.3156739811912226, "grad_norm": 3.5588443279266357, "learning_rate": 7.5775688386038395e-06, "loss": 1.018, "step": 16788 }, { "epoch": 1.3157523510971787, "grad_norm": 3.4793334007263184, "learning_rate": 7.5773046040217014e-06, "loss": 1.2043, "step": 16789 }, { "epoch": 1.3158307210031348, "grad_norm": 2.525496006011963, "learning_rate": 7.577040359636819e-06, "loss": 1.264, "step": 16790 }, { "epoch": 1.315909090909091, "grad_norm": 3.0298171043395996, "learning_rate": 7.576776105450199e-06, "loss": 1.248, "step": 16791 }, { "epoch": 1.315987460815047, "grad_norm": 7.609665870666504, "learning_rate": 7.576511841462846e-06, "loss": 0.8517, "step": 16792 }, { "epoch": 1.3160658307210031, "grad_norm": 9.24437427520752, "learning_rate": 7.576247567675765e-06, "loss": 1.3557, "step": 16793 }, { "epoch": 1.3161442006269592, "grad_norm": 3.41306734085083, "learning_rate": 7.575983284089962e-06, "loss": 1.8337, "step": 16794 }, { "epoch": 1.3162225705329154, "grad_norm": 6.529296398162842, "learning_rate": 7.575718990706442e-06, "loss": 1.4084, "step": 16795 }, { "epoch": 1.3163009404388715, "grad_norm": 2.7564198970794678, "learning_rate": 7.57545468752621e-06, "loss": 1.5636, "step": 16796 }, { "epoch": 1.3163793103448276, "grad_norm": 3.929619550704956, "learning_rate": 7.575190374550272e-06, "loss": 1.0915, "step": 16797 }, { "epoch": 1.3164576802507837, "grad_norm": 2.168613910675049, "learning_rate": 7.574926051779632e-06, "loss": 1.2385, "step": 16798 }, { "epoch": 1.3165360501567398, "grad_norm": 2.3160195350646973, "learning_rate": 7.574661719215295e-06, "loss": 1.4088, "step": 16799 }, { "epoch": 1.316614420062696, "grad_norm": 3.3987393379211426, "learning_rate": 7.574397376858267e-06, "loss": 1.1012, "step": 16800 }, { "epoch": 1.316692789968652, "grad_norm": 3.43111252784729, "learning_rate": 7.574133024709555e-06, "loss": 1.3618, "step": 16801 }, { "epoch": 1.3167711598746081, "grad_norm": 3.197906494140625, "learning_rate": 7.573868662770162e-06, "loss": 1.4482, "step": 16802 }, { "epoch": 1.3168495297805642, "grad_norm": 4.903282642364502, "learning_rate": 7.573604291041095e-06, "loss": 1.3709, "step": 16803 }, { "epoch": 1.3169278996865204, "grad_norm": 3.0020785331726074, "learning_rate": 7.57333990952336e-06, "loss": 1.2635, "step": 16804 }, { "epoch": 1.3170062695924765, "grad_norm": 2.302581548690796, "learning_rate": 7.57307551821796e-06, "loss": 1.3886, "step": 16805 }, { "epoch": 1.3170846394984326, "grad_norm": 2.566876173019409, "learning_rate": 7.572811117125905e-06, "loss": 1.3367, "step": 16806 }, { "epoch": 1.3171630094043887, "grad_norm": 2.9398720264434814, "learning_rate": 7.572546706248194e-06, "loss": 1.3288, "step": 16807 }, { "epoch": 1.3172413793103448, "grad_norm": 8.452606201171875, "learning_rate": 7.572282285585839e-06, "loss": 1.8947, "step": 16808 }, { "epoch": 1.317319749216301, "grad_norm": 2.866466760635376, "learning_rate": 7.572017855139844e-06, "loss": 1.6836, "step": 16809 }, { "epoch": 1.317398119122257, "grad_norm": 2.4983291625976562, "learning_rate": 7.571753414911213e-06, "loss": 1.501, "step": 16810 }, { "epoch": 1.3174764890282131, "grad_norm": 16.21087646484375, "learning_rate": 7.571488964900954e-06, "loss": 1.9462, "step": 16811 }, { "epoch": 1.3175548589341692, "grad_norm": 3.6784369945526123, "learning_rate": 7.571224505110071e-06, "loss": 1.2365, "step": 16812 }, { "epoch": 1.3176332288401253, "grad_norm": 3.9017670154571533, "learning_rate": 7.570960035539572e-06, "loss": 1.3883, "step": 16813 }, { "epoch": 1.3177115987460815, "grad_norm": 3.493070602416992, "learning_rate": 7.570695556190459e-06, "loss": 1.3522, "step": 16814 }, { "epoch": 1.3177899686520376, "grad_norm": 3.09775447845459, "learning_rate": 7.570431067063742e-06, "loss": 1.7542, "step": 16815 }, { "epoch": 1.3178683385579937, "grad_norm": 1.9419951438903809, "learning_rate": 7.570166568160425e-06, "loss": 1.347, "step": 16816 }, { "epoch": 1.3179467084639498, "grad_norm": 2.193772077560425, "learning_rate": 7.5699020594815155e-06, "loss": 1.1767, "step": 16817 }, { "epoch": 1.318025078369906, "grad_norm": 5.6811933517456055, "learning_rate": 7.569637541028019e-06, "loss": 1.1553, "step": 16818 }, { "epoch": 1.318103448275862, "grad_norm": 2.8941116333007812, "learning_rate": 7.569373012800939e-06, "loss": 1.8509, "step": 16819 }, { "epoch": 1.3181818181818181, "grad_norm": 3.223015069961548, "learning_rate": 7.569108474801286e-06, "loss": 1.1218, "step": 16820 }, { "epoch": 1.3182601880877742, "grad_norm": 3.2152559757232666, "learning_rate": 7.568843927030062e-06, "loss": 1.4608, "step": 16821 }, { "epoch": 1.3183385579937303, "grad_norm": 13.757827758789062, "learning_rate": 7.568579369488276e-06, "loss": 2.4137, "step": 16822 }, { "epoch": 1.3184169278996865, "grad_norm": 4.194719314575195, "learning_rate": 7.568314802176933e-06, "loss": 1.3044, "step": 16823 }, { "epoch": 1.3184952978056426, "grad_norm": 3.2298924922943115, "learning_rate": 7.568050225097039e-06, "loss": 1.7211, "step": 16824 }, { "epoch": 1.3185736677115987, "grad_norm": 2.1160621643066406, "learning_rate": 7.567785638249603e-06, "loss": 1.2643, "step": 16825 }, { "epoch": 1.3186520376175548, "grad_norm": 5.479483127593994, "learning_rate": 7.567521041635628e-06, "loss": 1.2068, "step": 16826 }, { "epoch": 1.318730407523511, "grad_norm": 2.849029302597046, "learning_rate": 7.567256435256123e-06, "loss": 1.2165, "step": 16827 }, { "epoch": 1.318808777429467, "grad_norm": 1.8270434141159058, "learning_rate": 7.566991819112092e-06, "loss": 1.439, "step": 16828 }, { "epoch": 1.3188871473354231, "grad_norm": 4.80741548538208, "learning_rate": 7.566727193204541e-06, "loss": 1.8546, "step": 16829 }, { "epoch": 1.3189655172413794, "grad_norm": 3.046992778778076, "learning_rate": 7.566462557534481e-06, "loss": 1.3191, "step": 16830 }, { "epoch": 1.3190438871473353, "grad_norm": 4.832958698272705, "learning_rate": 7.566197912102915e-06, "loss": 1.3635, "step": 16831 }, { "epoch": 1.3191222570532917, "grad_norm": 2.6743323802948, "learning_rate": 7.56593325691085e-06, "loss": 1.3858, "step": 16832 }, { "epoch": 1.3192006269592476, "grad_norm": 3.7479052543640137, "learning_rate": 7.565668591959292e-06, "loss": 1.1815, "step": 16833 }, { "epoch": 1.319278996865204, "grad_norm": 5.805828094482422, "learning_rate": 7.56540391724925e-06, "loss": 1.4122, "step": 16834 }, { "epoch": 1.3193573667711598, "grad_norm": 4.556481838226318, "learning_rate": 7.5651392327817265e-06, "loss": 1.3685, "step": 16835 }, { "epoch": 1.3194357366771161, "grad_norm": 2.1397767066955566, "learning_rate": 7.564874538557734e-06, "loss": 1.5238, "step": 16836 }, { "epoch": 1.319514106583072, "grad_norm": 3.2559115886688232, "learning_rate": 7.564609834578274e-06, "loss": 1.274, "step": 16837 }, { "epoch": 1.3195924764890283, "grad_norm": 3.0233662128448486, "learning_rate": 7.564345120844356e-06, "loss": 1.6041, "step": 16838 }, { "epoch": 1.3196708463949842, "grad_norm": 5.206564903259277, "learning_rate": 7.564080397356988e-06, "loss": 1.3866, "step": 16839 }, { "epoch": 1.3197492163009406, "grad_norm": 2.4495623111724854, "learning_rate": 7.563815664117173e-06, "loss": 1.3352, "step": 16840 }, { "epoch": 1.3198275862068964, "grad_norm": 3.1101176738739014, "learning_rate": 7.563550921125922e-06, "loss": 1.5428, "step": 16841 }, { "epoch": 1.3199059561128528, "grad_norm": 2.7568020820617676, "learning_rate": 7.563286168384236e-06, "loss": 1.6658, "step": 16842 }, { "epoch": 1.3199843260188087, "grad_norm": 3.299774169921875, "learning_rate": 7.563021405893131e-06, "loss": 1.3771, "step": 16843 }, { "epoch": 1.320062695924765, "grad_norm": 4.308231353759766, "learning_rate": 7.5627566336536066e-06, "loss": 1.2227, "step": 16844 }, { "epoch": 1.3201410658307209, "grad_norm": 2.9255573749542236, "learning_rate": 7.562491851666673e-06, "loss": 1.1921, "step": 16845 }, { "epoch": 1.3202194357366772, "grad_norm": 1.9636237621307373, "learning_rate": 7.562227059933336e-06, "loss": 1.1934, "step": 16846 }, { "epoch": 1.320297805642633, "grad_norm": 9.894444465637207, "learning_rate": 7.5619622584546024e-06, "loss": 2.9704, "step": 16847 }, { "epoch": 1.3203761755485894, "grad_norm": 2.7839877605438232, "learning_rate": 7.561697447231482e-06, "loss": 1.3903, "step": 16848 }, { "epoch": 1.3204545454545453, "grad_norm": 2.490312337875366, "learning_rate": 7.561432626264979e-06, "loss": 1.2332, "step": 16849 }, { "epoch": 1.3205329153605017, "grad_norm": 3.758683204650879, "learning_rate": 7.5611677955561016e-06, "loss": 1.1701, "step": 16850 }, { "epoch": 1.3206112852664575, "grad_norm": 3.734606981277466, "learning_rate": 7.56090295510586e-06, "loss": 1.266, "step": 16851 }, { "epoch": 1.3206896551724139, "grad_norm": 4.86650276184082, "learning_rate": 7.560638104915256e-06, "loss": 1.3783, "step": 16852 }, { "epoch": 1.3207680250783698, "grad_norm": 4.534850597381592, "learning_rate": 7.560373244985302e-06, "loss": 1.5081, "step": 16853 }, { "epoch": 1.320846394984326, "grad_norm": 2.250288963317871, "learning_rate": 7.5601083753170015e-06, "loss": 1.4152, "step": 16854 }, { "epoch": 1.3209247648902822, "grad_norm": 4.913544654846191, "learning_rate": 7.559843495911365e-06, "loss": 1.2857, "step": 16855 }, { "epoch": 1.3210031347962383, "grad_norm": 2.3730878829956055, "learning_rate": 7.559578606769397e-06, "loss": 1.1541, "step": 16856 }, { "epoch": 1.3210815047021944, "grad_norm": 3.628162384033203, "learning_rate": 7.559313707892108e-06, "loss": 1.5742, "step": 16857 }, { "epoch": 1.3211598746081505, "grad_norm": 3.3429934978485107, "learning_rate": 7.559048799280504e-06, "loss": 1.2838, "step": 16858 }, { "epoch": 1.3212382445141067, "grad_norm": 4.409390926361084, "learning_rate": 7.558783880935594e-06, "loss": 1.395, "step": 16859 }, { "epoch": 1.3213166144200628, "grad_norm": 2.329225778579712, "learning_rate": 7.558518952858384e-06, "loss": 1.4194, "step": 16860 }, { "epoch": 1.3213949843260189, "grad_norm": 2.8310253620147705, "learning_rate": 7.55825401504988e-06, "loss": 1.2669, "step": 16861 }, { "epoch": 1.321473354231975, "grad_norm": 3.430233955383301, "learning_rate": 7.5579890675110935e-06, "loss": 1.7642, "step": 16862 }, { "epoch": 1.321551724137931, "grad_norm": 3.1734139919281006, "learning_rate": 7.55772411024303e-06, "loss": 1.4328, "step": 16863 }, { "epoch": 1.3216300940438872, "grad_norm": 2.343717575073242, "learning_rate": 7.557459143246697e-06, "loss": 1.291, "step": 16864 }, { "epoch": 1.3217084639498433, "grad_norm": 2.7306737899780273, "learning_rate": 7.557194166523104e-06, "loss": 1.5873, "step": 16865 }, { "epoch": 1.3217868338557994, "grad_norm": 3.092400312423706, "learning_rate": 7.556929180073258e-06, "loss": 1.21, "step": 16866 }, { "epoch": 1.3218652037617555, "grad_norm": 7.380795478820801, "learning_rate": 7.556664183898166e-06, "loss": 0.5521, "step": 16867 }, { "epoch": 1.3219435736677116, "grad_norm": 7.459310531616211, "learning_rate": 7.5563991779988366e-06, "loss": 0.7328, "step": 16868 }, { "epoch": 1.3220219435736678, "grad_norm": 6.007792949676514, "learning_rate": 7.556134162376279e-06, "loss": 1.7258, "step": 16869 }, { "epoch": 1.3221003134796239, "grad_norm": 2.003159523010254, "learning_rate": 7.555869137031497e-06, "loss": 0.9078, "step": 16870 }, { "epoch": 1.32217868338558, "grad_norm": 2.582329273223877, "learning_rate": 7.555604101965505e-06, "loss": 1.2081, "step": 16871 }, { "epoch": 1.322257053291536, "grad_norm": 4.087313652038574, "learning_rate": 7.555339057179306e-06, "loss": 1.2108, "step": 16872 }, { "epoch": 1.3223354231974922, "grad_norm": 3.4381775856018066, "learning_rate": 7.55507400267391e-06, "loss": 1.6338, "step": 16873 }, { "epoch": 1.3224137931034483, "grad_norm": 4.1799635887146, "learning_rate": 7.554808938450324e-06, "loss": 1.0092, "step": 16874 }, { "epoch": 1.3224921630094044, "grad_norm": 2.891342878341675, "learning_rate": 7.554543864509558e-06, "loss": 1.4332, "step": 16875 }, { "epoch": 1.3225705329153605, "grad_norm": 3.0525364875793457, "learning_rate": 7.5542787808526195e-06, "loss": 1.0715, "step": 16876 }, { "epoch": 1.3226489028213166, "grad_norm": 3.0503315925598145, "learning_rate": 7.554013687480517e-06, "loss": 1.5169, "step": 16877 }, { "epoch": 1.3227272727272728, "grad_norm": 3.2999179363250732, "learning_rate": 7.553748584394256e-06, "loss": 1.3857, "step": 16878 }, { "epoch": 1.3228056426332289, "grad_norm": 2.5413355827331543, "learning_rate": 7.553483471594849e-06, "loss": 1.4014, "step": 16879 }, { "epoch": 1.322884012539185, "grad_norm": 2.876256227493286, "learning_rate": 7.553218349083302e-06, "loss": 1.5975, "step": 16880 }, { "epoch": 1.322962382445141, "grad_norm": 2.0778679847717285, "learning_rate": 7.552953216860624e-06, "loss": 1.1913, "step": 16881 }, { "epoch": 1.3230407523510972, "grad_norm": 4.468914985656738, "learning_rate": 7.552688074927823e-06, "loss": 1.6225, "step": 16882 }, { "epoch": 1.3231191222570533, "grad_norm": 6.028054714202881, "learning_rate": 7.552422923285907e-06, "loss": 0.7424, "step": 16883 }, { "epoch": 1.3231974921630094, "grad_norm": 10.537078857421875, "learning_rate": 7.552157761935886e-06, "loss": 1.6548, "step": 16884 }, { "epoch": 1.3232758620689655, "grad_norm": 6.33555269241333, "learning_rate": 7.551892590878766e-06, "loss": 1.3242, "step": 16885 }, { "epoch": 1.3233542319749216, "grad_norm": 2.8597054481506348, "learning_rate": 7.551627410115561e-06, "loss": 1.2641, "step": 16886 }, { "epoch": 1.3234326018808777, "grad_norm": 3.631248950958252, "learning_rate": 7.5513622196472734e-06, "loss": 1.0057, "step": 16887 }, { "epoch": 1.3235109717868339, "grad_norm": 13.614810943603516, "learning_rate": 7.551097019474915e-06, "loss": 1.2231, "step": 16888 }, { "epoch": 1.32358934169279, "grad_norm": 3.3818774223327637, "learning_rate": 7.550831809599493e-06, "loss": 1.6149, "step": 16889 }, { "epoch": 1.323667711598746, "grad_norm": 3.3566575050354004, "learning_rate": 7.550566590022018e-06, "loss": 1.3654, "step": 16890 }, { "epoch": 1.3237460815047022, "grad_norm": 3.8202524185180664, "learning_rate": 7.550301360743497e-06, "loss": 1.6364, "step": 16891 }, { "epoch": 1.3238244514106583, "grad_norm": 1.97482430934906, "learning_rate": 7.550036121764941e-06, "loss": 1.2869, "step": 16892 }, { "epoch": 1.3239028213166144, "grad_norm": 2.789095878601074, "learning_rate": 7.549770873087355e-06, "loss": 1.1442, "step": 16893 }, { "epoch": 1.3239811912225705, "grad_norm": 5.32555627822876, "learning_rate": 7.549505614711752e-06, "loss": 1.813, "step": 16894 }, { "epoch": 1.3240595611285266, "grad_norm": 2.702115058898926, "learning_rate": 7.54924034663914e-06, "loss": 1.4893, "step": 16895 }, { "epoch": 1.3241379310344827, "grad_norm": 2.5372345447540283, "learning_rate": 7.548975068870524e-06, "loss": 1.4637, "step": 16896 }, { "epoch": 1.3242163009404389, "grad_norm": 2.178497791290283, "learning_rate": 7.548709781406919e-06, "loss": 1.4313, "step": 16897 }, { "epoch": 1.324294670846395, "grad_norm": 2.888273000717163, "learning_rate": 7.54844448424933e-06, "loss": 1.4702, "step": 16898 }, { "epoch": 1.324373040752351, "grad_norm": 4.329638481140137, "learning_rate": 7.548179177398766e-06, "loss": 1.1493, "step": 16899 }, { "epoch": 1.3244514106583072, "grad_norm": 2.315783977508545, "learning_rate": 7.547913860856239e-06, "loss": 1.4888, "step": 16900 }, { "epoch": 1.3245297805642633, "grad_norm": 4.307356357574463, "learning_rate": 7.547648534622756e-06, "loss": 2.156, "step": 16901 }, { "epoch": 1.3246081504702194, "grad_norm": 3.1872916221618652, "learning_rate": 7.547383198699325e-06, "loss": 1.1689, "step": 16902 }, { "epoch": 1.3246865203761755, "grad_norm": 3.9599409103393555, "learning_rate": 7.5471178530869585e-06, "loss": 1.214, "step": 16903 }, { "epoch": 1.3247648902821316, "grad_norm": 2.4720020294189453, "learning_rate": 7.546852497786663e-06, "loss": 1.113, "step": 16904 }, { "epoch": 1.3248432601880877, "grad_norm": 4.278650283813477, "learning_rate": 7.54658713279945e-06, "loss": 1.6068, "step": 16905 }, { "epoch": 1.3249216300940438, "grad_norm": 2.1490705013275146, "learning_rate": 7.546321758126326e-06, "loss": 1.5425, "step": 16906 }, { "epoch": 1.325, "grad_norm": 5.033570766448975, "learning_rate": 7.5460563737683034e-06, "loss": 1.6555, "step": 16907 }, { "epoch": 1.325078369905956, "grad_norm": 3.0964930057525635, "learning_rate": 7.5457909797263894e-06, "loss": 1.4286, "step": 16908 }, { "epoch": 1.3251567398119122, "grad_norm": 2.9057905673980713, "learning_rate": 7.545525576001593e-06, "loss": 1.4979, "step": 16909 }, { "epoch": 1.3252351097178683, "grad_norm": 3.3959879875183105, "learning_rate": 7.545260162594928e-06, "loss": 1.2548, "step": 16910 }, { "epoch": 1.3253134796238244, "grad_norm": 2.7987117767333984, "learning_rate": 7.544994739507397e-06, "loss": 1.3341, "step": 16911 }, { "epoch": 1.3253918495297805, "grad_norm": 4.68636417388916, "learning_rate": 7.544729306740015e-06, "loss": 1.7256, "step": 16912 }, { "epoch": 1.3254702194357366, "grad_norm": 3.98091197013855, "learning_rate": 7.54446386429379e-06, "loss": 1.3829, "step": 16913 }, { "epoch": 1.3255485893416927, "grad_norm": 4.202126979827881, "learning_rate": 7.54419841216973e-06, "loss": 1.3445, "step": 16914 }, { "epoch": 1.3256269592476488, "grad_norm": 2.281276226043701, "learning_rate": 7.543932950368847e-06, "loss": 1.3552, "step": 16915 }, { "epoch": 1.325705329153605, "grad_norm": 2.172656297683716, "learning_rate": 7.54366747889215e-06, "loss": 1.2287, "step": 16916 }, { "epoch": 1.325783699059561, "grad_norm": 6.744328022003174, "learning_rate": 7.543401997740647e-06, "loss": 1.2055, "step": 16917 }, { "epoch": 1.3258620689655172, "grad_norm": 4.156723976135254, "learning_rate": 7.54313650691535e-06, "loss": 1.3336, "step": 16918 }, { "epoch": 1.3259404388714733, "grad_norm": 3.1559033393859863, "learning_rate": 7.542871006417269e-06, "loss": 1.4098, "step": 16919 }, { "epoch": 1.3260188087774294, "grad_norm": 2.7136356830596924, "learning_rate": 7.542605496247412e-06, "loss": 1.145, "step": 16920 }, { "epoch": 1.3260971786833855, "grad_norm": 2.264867067337036, "learning_rate": 7.542339976406789e-06, "loss": 1.3225, "step": 16921 }, { "epoch": 1.3261755485893416, "grad_norm": 3.471665143966675, "learning_rate": 7.54207444689641e-06, "loss": 1.3903, "step": 16922 }, { "epoch": 1.3262539184952977, "grad_norm": 2.214233636856079, "learning_rate": 7.541808907717288e-06, "loss": 1.5179, "step": 16923 }, { "epoch": 1.3263322884012538, "grad_norm": 2.0221939086914062, "learning_rate": 7.541543358870429e-06, "loss": 0.579, "step": 16924 }, { "epoch": 1.3264106583072102, "grad_norm": 7.363254070281982, "learning_rate": 7.541277800356846e-06, "loss": 0.8034, "step": 16925 }, { "epoch": 1.326489028213166, "grad_norm": 4.836443901062012, "learning_rate": 7.541012232177545e-06, "loss": 1.4329, "step": 16926 }, { "epoch": 1.3265673981191224, "grad_norm": 2.762517213821411, "learning_rate": 7.540746654333539e-06, "loss": 1.4042, "step": 16927 }, { "epoch": 1.3266457680250783, "grad_norm": 2.2911860942840576, "learning_rate": 7.540481066825839e-06, "loss": 1.2264, "step": 16928 }, { "epoch": 1.3267241379310346, "grad_norm": 2.900259494781494, "learning_rate": 7.540215469655453e-06, "loss": 1.5348, "step": 16929 }, { "epoch": 1.3268025078369905, "grad_norm": 3.220958709716797, "learning_rate": 7.5399498628233925e-06, "loss": 1.4488, "step": 16930 }, { "epoch": 1.3268808777429468, "grad_norm": 2.4591310024261475, "learning_rate": 7.539684246330667e-06, "loss": 1.2478, "step": 16931 }, { "epoch": 1.3269592476489027, "grad_norm": 2.6962943077087402, "learning_rate": 7.539418620178287e-06, "loss": 1.3614, "step": 16932 }, { "epoch": 1.327037617554859, "grad_norm": 2.871842622756958, "learning_rate": 7.539152984367264e-06, "loss": 1.4692, "step": 16933 }, { "epoch": 1.327115987460815, "grad_norm": 6.303317546844482, "learning_rate": 7.538887338898606e-06, "loss": 0.9622, "step": 16934 }, { "epoch": 1.3271943573667713, "grad_norm": 2.3130922317504883, "learning_rate": 7.538621683773326e-06, "loss": 1.1508, "step": 16935 }, { "epoch": 1.3272727272727272, "grad_norm": 3.4439878463745117, "learning_rate": 7.538356018992432e-06, "loss": 1.1604, "step": 16936 }, { "epoch": 1.3273510971786835, "grad_norm": 2.6710104942321777, "learning_rate": 7.538090344556936e-06, "loss": 1.3734, "step": 16937 }, { "epoch": 1.3274294670846394, "grad_norm": 2.2200608253479004, "learning_rate": 7.537824660467847e-06, "loss": 1.4261, "step": 16938 }, { "epoch": 1.3275078369905957, "grad_norm": 2.705263614654541, "learning_rate": 7.5375589667261775e-06, "loss": 1.3023, "step": 16939 }, { "epoch": 1.3275862068965516, "grad_norm": 3.211916923522949, "learning_rate": 7.537293263332937e-06, "loss": 1.1357, "step": 16940 }, { "epoch": 1.327664576802508, "grad_norm": 3.4153640270233154, "learning_rate": 7.537027550289136e-06, "loss": 1.7746, "step": 16941 }, { "epoch": 1.3277429467084638, "grad_norm": 2.5145761966705322, "learning_rate": 7.5367618275957845e-06, "loss": 1.3278, "step": 16942 }, { "epoch": 1.3278213166144202, "grad_norm": 1.924803614616394, "learning_rate": 7.5364960952538936e-06, "loss": 1.2455, "step": 16943 }, { "epoch": 1.327899686520376, "grad_norm": 3.13753080368042, "learning_rate": 7.536230353264476e-06, "loss": 1.5834, "step": 16944 }, { "epoch": 1.3279780564263324, "grad_norm": 3.299433708190918, "learning_rate": 7.53596460162854e-06, "loss": 1.3635, "step": 16945 }, { "epoch": 1.3280564263322883, "grad_norm": 6.211390018463135, "learning_rate": 7.535698840347098e-06, "loss": 1.687, "step": 16946 }, { "epoch": 1.3281347962382446, "grad_norm": 7.0333781242370605, "learning_rate": 7.535433069421159e-06, "loss": 1.3733, "step": 16947 }, { "epoch": 1.3282131661442007, "grad_norm": 2.1602864265441895, "learning_rate": 7.535167288851735e-06, "loss": 1.6068, "step": 16948 }, { "epoch": 1.3282915360501568, "grad_norm": 2.8650450706481934, "learning_rate": 7.5349014986398375e-06, "loss": 1.0927, "step": 16949 }, { "epoch": 1.328369905956113, "grad_norm": 3.733562707901001, "learning_rate": 7.534635698786476e-06, "loss": 2.2945, "step": 16950 }, { "epoch": 1.328448275862069, "grad_norm": 4.656535625457764, "learning_rate": 7.534369889292663e-06, "loss": 1.6447, "step": 16951 }, { "epoch": 1.3285266457680251, "grad_norm": 3.110546350479126, "learning_rate": 7.534104070159409e-06, "loss": 1.2712, "step": 16952 }, { "epoch": 1.3286050156739813, "grad_norm": 2.993845224380493, "learning_rate": 7.533838241387723e-06, "loss": 1.2915, "step": 16953 }, { "epoch": 1.3286833855799374, "grad_norm": 3.475229263305664, "learning_rate": 7.533572402978619e-06, "loss": 1.0565, "step": 16954 }, { "epoch": 1.3287617554858935, "grad_norm": 3.02449893951416, "learning_rate": 7.533306554933106e-06, "loss": 0.9086, "step": 16955 }, { "epoch": 1.3288401253918496, "grad_norm": 2.4190990924835205, "learning_rate": 7.533040697252197e-06, "loss": 1.3095, "step": 16956 }, { "epoch": 1.3289184952978057, "grad_norm": 5.423582077026367, "learning_rate": 7.532774829936902e-06, "loss": 1.3324, "step": 16957 }, { "epoch": 1.3289968652037618, "grad_norm": 3.839035987854004, "learning_rate": 7.532508952988231e-06, "loss": 1.3392, "step": 16958 }, { "epoch": 1.329075235109718, "grad_norm": 2.8018946647644043, "learning_rate": 7.532243066407199e-06, "loss": 1.3059, "step": 16959 }, { "epoch": 1.329153605015674, "grad_norm": 6.630230903625488, "learning_rate": 7.531977170194813e-06, "loss": 1.424, "step": 16960 }, { "epoch": 1.3292319749216301, "grad_norm": 8.318155288696289, "learning_rate": 7.531711264352088e-06, "loss": 1.1103, "step": 16961 }, { "epoch": 1.3293103448275863, "grad_norm": 14.02871322631836, "learning_rate": 7.5314453488800335e-06, "loss": 0.8824, "step": 16962 }, { "epoch": 1.3293887147335424, "grad_norm": 3.2935824394226074, "learning_rate": 7.531179423779662e-06, "loss": 1.4222, "step": 16963 }, { "epoch": 1.3294670846394985, "grad_norm": 7.400113105773926, "learning_rate": 7.530913489051982e-06, "loss": 1.2722, "step": 16964 }, { "epoch": 1.3295454545454546, "grad_norm": 2.198911666870117, "learning_rate": 7.530647544698008e-06, "loss": 1.4485, "step": 16965 }, { "epoch": 1.3296238244514107, "grad_norm": 3.9480109214782715, "learning_rate": 7.530381590718749e-06, "loss": 1.1414, "step": 16966 }, { "epoch": 1.3297021943573668, "grad_norm": 3.7182061672210693, "learning_rate": 7.5301156271152195e-06, "loss": 1.225, "step": 16967 }, { "epoch": 1.329780564263323, "grad_norm": 3.073716878890991, "learning_rate": 7.5298496538884324e-06, "loss": 1.4405, "step": 16968 }, { "epoch": 1.329858934169279, "grad_norm": 8.785473823547363, "learning_rate": 7.529583671039392e-06, "loss": 1.7798, "step": 16969 }, { "epoch": 1.3299373040752351, "grad_norm": 4.156060695648193, "learning_rate": 7.529317678569117e-06, "loss": 1.3743, "step": 16970 }, { "epoch": 1.3300156739811912, "grad_norm": 2.2632546424865723, "learning_rate": 7.529051676478616e-06, "loss": 1.2942, "step": 16971 }, { "epoch": 1.3300940438871474, "grad_norm": 3.483726739883423, "learning_rate": 7.528785664768903e-06, "loss": 1.4783, "step": 16972 }, { "epoch": 1.3301724137931035, "grad_norm": 2.8238978385925293, "learning_rate": 7.528519643440986e-06, "loss": 1.2465, "step": 16973 }, { "epoch": 1.3302507836990596, "grad_norm": 3.3342957496643066, "learning_rate": 7.52825361249588e-06, "loss": 1.4544, "step": 16974 }, { "epoch": 1.3303291536050157, "grad_norm": 2.5163674354553223, "learning_rate": 7.527987571934596e-06, "loss": 1.2113, "step": 16975 }, { "epoch": 1.3304075235109718, "grad_norm": 2.981245994567871, "learning_rate": 7.527721521758145e-06, "loss": 1.3782, "step": 16976 }, { "epoch": 1.330485893416928, "grad_norm": 3.5372166633605957, "learning_rate": 7.52745546196754e-06, "loss": 1.1843, "step": 16977 }, { "epoch": 1.330564263322884, "grad_norm": 3.8591606616973877, "learning_rate": 7.527189392563793e-06, "loss": 1.6606, "step": 16978 }, { "epoch": 1.3306426332288401, "grad_norm": 2.343461751937866, "learning_rate": 7.526923313547916e-06, "loss": 1.5196, "step": 16979 }, { "epoch": 1.3307210031347962, "grad_norm": 2.3952083587646484, "learning_rate": 7.526657224920919e-06, "loss": 1.3864, "step": 16980 }, { "epoch": 1.3307993730407524, "grad_norm": 2.427366256713867, "learning_rate": 7.526391126683817e-06, "loss": 1.3352, "step": 16981 }, { "epoch": 1.3308777429467085, "grad_norm": 4.485159397125244, "learning_rate": 7.5261250188376204e-06, "loss": 1.8963, "step": 16982 }, { "epoch": 1.3309561128526646, "grad_norm": 2.491534471511841, "learning_rate": 7.525858901383343e-06, "loss": 1.0338, "step": 16983 }, { "epoch": 1.3310344827586207, "grad_norm": 2.3092498779296875, "learning_rate": 7.525592774321993e-06, "loss": 1.4727, "step": 16984 }, { "epoch": 1.3311128526645768, "grad_norm": 3.1369714736938477, "learning_rate": 7.525326637654587e-06, "loss": 1.3977, "step": 16985 }, { "epoch": 1.331191222570533, "grad_norm": 2.0229644775390625, "learning_rate": 7.525060491382135e-06, "loss": 1.0704, "step": 16986 }, { "epoch": 1.331269592476489, "grad_norm": 2.628551483154297, "learning_rate": 7.524794335505652e-06, "loss": 1.1274, "step": 16987 }, { "epoch": 1.3313479623824451, "grad_norm": 2.4206185340881348, "learning_rate": 7.524528170026144e-06, "loss": 1.0995, "step": 16988 }, { "epoch": 1.3314263322884012, "grad_norm": 2.426746129989624, "learning_rate": 7.524261994944631e-06, "loss": 0.9695, "step": 16989 }, { "epoch": 1.3315047021943573, "grad_norm": 2.3971402645111084, "learning_rate": 7.52399581026212e-06, "loss": 1.2728, "step": 16990 }, { "epoch": 1.3315830721003135, "grad_norm": 2.7844815254211426, "learning_rate": 7.523729615979627e-06, "loss": 1.4128, "step": 16991 }, { "epoch": 1.3316614420062696, "grad_norm": 3.8402791023254395, "learning_rate": 7.523463412098162e-06, "loss": 1.5484, "step": 16992 }, { "epoch": 1.3317398119122257, "grad_norm": 3.8092029094696045, "learning_rate": 7.523197198618738e-06, "loss": 1.5062, "step": 16993 }, { "epoch": 1.3318181818181818, "grad_norm": 2.481853723526001, "learning_rate": 7.5229309755423686e-06, "loss": 1.1483, "step": 16994 }, { "epoch": 1.331896551724138, "grad_norm": 2.33146595954895, "learning_rate": 7.522664742870065e-06, "loss": 1.6646, "step": 16995 }, { "epoch": 1.331974921630094, "grad_norm": 3.4015676975250244, "learning_rate": 7.522398500602841e-06, "loss": 1.3094, "step": 16996 }, { "epoch": 1.3320532915360501, "grad_norm": 3.2716479301452637, "learning_rate": 7.522132248741707e-06, "loss": 1.3296, "step": 16997 }, { "epoch": 1.3321316614420062, "grad_norm": 3.8385748863220215, "learning_rate": 7.52186598728768e-06, "loss": 1.6128, "step": 16998 }, { "epoch": 1.3322100313479623, "grad_norm": 1.964892029762268, "learning_rate": 7.521599716241767e-06, "loss": 1.0852, "step": 16999 }, { "epoch": 1.3322884012539185, "grad_norm": 23.912927627563477, "learning_rate": 7.521333435604987e-06, "loss": 1.4109, "step": 17000 }, { "epoch": 1.3323667711598746, "grad_norm": 2.2972192764282227, "learning_rate": 7.521067145378348e-06, "loss": 1.1946, "step": 17001 }, { "epoch": 1.3324451410658307, "grad_norm": 4.666977882385254, "learning_rate": 7.520800845562865e-06, "loss": 1.6833, "step": 17002 }, { "epoch": 1.3325235109717868, "grad_norm": 6.426468849182129, "learning_rate": 7.52053453615955e-06, "loss": 1.4075, "step": 17003 }, { "epoch": 1.332601880877743, "grad_norm": 6.764322757720947, "learning_rate": 7.520268217169415e-06, "loss": 1.5378, "step": 17004 }, { "epoch": 1.332680250783699, "grad_norm": 3.5366485118865967, "learning_rate": 7.520001888593475e-06, "loss": 1.3917, "step": 17005 }, { "epoch": 1.3327586206896551, "grad_norm": 3.881932020187378, "learning_rate": 7.5197355504327455e-06, "loss": 1.2326, "step": 17006 }, { "epoch": 1.3328369905956112, "grad_norm": 3.325584888458252, "learning_rate": 7.519469202688232e-06, "loss": 1.3097, "step": 17007 }, { "epoch": 1.3329153605015673, "grad_norm": 3.9125783443450928, "learning_rate": 7.519202845360954e-06, "loss": 1.3516, "step": 17008 }, { "epoch": 1.3329937304075234, "grad_norm": 2.7845757007598877, "learning_rate": 7.518936478451921e-06, "loss": 1.6354, "step": 17009 }, { "epoch": 1.3330721003134796, "grad_norm": 3.4445383548736572, "learning_rate": 7.518670101962148e-06, "loss": 1.3755, "step": 17010 }, { "epoch": 1.3331504702194357, "grad_norm": 3.1983957290649414, "learning_rate": 7.518403715892647e-06, "loss": 1.8861, "step": 17011 }, { "epoch": 1.3332288401253918, "grad_norm": 3.9568862915039062, "learning_rate": 7.518137320244432e-06, "loss": 1.3859, "step": 17012 }, { "epoch": 1.333307210031348, "grad_norm": 5.502223014831543, "learning_rate": 7.517870915018517e-06, "loss": 1.5128, "step": 17013 }, { "epoch": 1.333385579937304, "grad_norm": 2.211682081222534, "learning_rate": 7.5176045002159126e-06, "loss": 1.1907, "step": 17014 }, { "epoch": 1.33346394984326, "grad_norm": 5.5565619468688965, "learning_rate": 7.517338075837636e-06, "loss": 0.7968, "step": 17015 }, { "epoch": 1.3335423197492162, "grad_norm": 2.9776992797851562, "learning_rate": 7.517071641884697e-06, "loss": 1.6587, "step": 17016 }, { "epoch": 1.3336206896551723, "grad_norm": 4.446997165679932, "learning_rate": 7.5168051983581105e-06, "loss": 1.3906, "step": 17017 }, { "epoch": 1.3336990595611287, "grad_norm": 4.082415580749512, "learning_rate": 7.51653874525889e-06, "loss": 2.0491, "step": 17018 }, { "epoch": 1.3337774294670846, "grad_norm": 4.368404388427734, "learning_rate": 7.516272282588047e-06, "loss": 1.2234, "step": 17019 }, { "epoch": 1.3338557993730409, "grad_norm": 3.375187635421753, "learning_rate": 7.5160058103465985e-06, "loss": 1.6132, "step": 17020 }, { "epoch": 1.3339341692789968, "grad_norm": 4.55340051651001, "learning_rate": 7.515739328535556e-06, "loss": 0.8048, "step": 17021 }, { "epoch": 1.334012539184953, "grad_norm": 4.836673259735107, "learning_rate": 7.515472837155934e-06, "loss": 0.8512, "step": 17022 }, { "epoch": 1.334090909090909, "grad_norm": 7.1749725341796875, "learning_rate": 7.5152063362087426e-06, "loss": 1.1641, "step": 17023 }, { "epoch": 1.3341692789968653, "grad_norm": 2.9579029083251953, "learning_rate": 7.514939825695001e-06, "loss": 1.3943, "step": 17024 }, { "epoch": 1.3342476489028212, "grad_norm": 5.500136852264404, "learning_rate": 7.514673305615719e-06, "loss": 1.6605, "step": 17025 }, { "epoch": 1.3343260188087775, "grad_norm": 3.7383852005004883, "learning_rate": 7.514406775971912e-06, "loss": 1.1864, "step": 17026 }, { "epoch": 1.3344043887147334, "grad_norm": 2.4551947116851807, "learning_rate": 7.5141402367645925e-06, "loss": 1.3885, "step": 17027 }, { "epoch": 1.3344827586206898, "grad_norm": 3.066952705383301, "learning_rate": 7.513873687994775e-06, "loss": 1.7718, "step": 17028 }, { "epoch": 1.3345611285266457, "grad_norm": 3.4781391620635986, "learning_rate": 7.513607129663472e-06, "loss": 1.6052, "step": 17029 }, { "epoch": 1.334639498432602, "grad_norm": 4.750423431396484, "learning_rate": 7.5133405617717e-06, "loss": 1.4349, "step": 17030 }, { "epoch": 1.3347178683385579, "grad_norm": 3.334169387817383, "learning_rate": 7.513073984320471e-06, "loss": 1.2211, "step": 17031 }, { "epoch": 1.3347962382445142, "grad_norm": 4.385118007659912, "learning_rate": 7.512807397310799e-06, "loss": 1.0334, "step": 17032 }, { "epoch": 1.33487460815047, "grad_norm": 3.354602336883545, "learning_rate": 7.5125408007436994e-06, "loss": 1.6428, "step": 17033 }, { "epoch": 1.3349529780564264, "grad_norm": 2.601346254348755, "learning_rate": 7.512274194620185e-06, "loss": 1.301, "step": 17034 }, { "epoch": 1.3350313479623823, "grad_norm": 3.4256296157836914, "learning_rate": 7.512007578941269e-06, "loss": 0.8163, "step": 17035 }, { "epoch": 1.3351097178683387, "grad_norm": 2.029930591583252, "learning_rate": 7.5117409537079665e-06, "loss": 1.4274, "step": 17036 }, { "epoch": 1.3351880877742945, "grad_norm": 3.1806371212005615, "learning_rate": 7.5114743189212925e-06, "loss": 1.1963, "step": 17037 }, { "epoch": 1.3352664576802509, "grad_norm": 14.697831153869629, "learning_rate": 7.51120767458226e-06, "loss": 1.7375, "step": 17038 }, { "epoch": 1.3353448275862068, "grad_norm": 3.1441681385040283, "learning_rate": 7.5109410206918824e-06, "loss": 1.2463, "step": 17039 }, { "epoch": 1.335423197492163, "grad_norm": 3.783600091934204, "learning_rate": 7.510674357251176e-06, "loss": 1.2242, "step": 17040 }, { "epoch": 1.3355015673981192, "grad_norm": 7.021510124206543, "learning_rate": 7.510407684261153e-06, "loss": 1.4503, "step": 17041 }, { "epoch": 1.3355799373040753, "grad_norm": 4.295192718505859, "learning_rate": 7.5101410017228286e-06, "loss": 0.7993, "step": 17042 }, { "epoch": 1.3356583072100314, "grad_norm": 3.5830118656158447, "learning_rate": 7.5098743096372176e-06, "loss": 1.3231, "step": 17043 }, { "epoch": 1.3357366771159875, "grad_norm": 2.30759859085083, "learning_rate": 7.509607608005332e-06, "loss": 1.3953, "step": 17044 }, { "epoch": 1.3358150470219436, "grad_norm": 2.341116189956665, "learning_rate": 7.50934089682819e-06, "loss": 1.2763, "step": 17045 }, { "epoch": 1.3358934169278998, "grad_norm": 6.087860107421875, "learning_rate": 7.509074176106804e-06, "loss": 1.7307, "step": 17046 }, { "epoch": 1.3359717868338559, "grad_norm": 2.914125442504883, "learning_rate": 7.508807445842188e-06, "loss": 1.6277, "step": 17047 }, { "epoch": 1.336050156739812, "grad_norm": 2.6806530952453613, "learning_rate": 7.508540706035359e-06, "loss": 1.462, "step": 17048 }, { "epoch": 1.336128526645768, "grad_norm": 2.2360329627990723, "learning_rate": 7.508273956687326e-06, "loss": 1.5137, "step": 17049 }, { "epoch": 1.3362068965517242, "grad_norm": 3.39589262008667, "learning_rate": 7.508007197799111e-06, "loss": 1.3662, "step": 17050 }, { "epoch": 1.3362852664576803, "grad_norm": 3.111471652984619, "learning_rate": 7.507740429371721e-06, "loss": 1.5287, "step": 17051 }, { "epoch": 1.3363636363636364, "grad_norm": 7.944911003112793, "learning_rate": 7.5074736514061785e-06, "loss": 0.9633, "step": 17052 }, { "epoch": 1.3364420062695925, "grad_norm": 5.699385643005371, "learning_rate": 7.507206863903493e-06, "loss": 1.0175, "step": 17053 }, { "epoch": 1.3365203761755486, "grad_norm": 2.3496949672698975, "learning_rate": 7.506940066864679e-06, "loss": 1.27, "step": 17054 }, { "epoch": 1.3365987460815048, "grad_norm": 2.227872133255005, "learning_rate": 7.506673260290753e-06, "loss": 1.2926, "step": 17055 }, { "epoch": 1.3366771159874609, "grad_norm": 6.909761428833008, "learning_rate": 7.50640644418273e-06, "loss": 1.5645, "step": 17056 }, { "epoch": 1.336755485893417, "grad_norm": 6.046655654907227, "learning_rate": 7.506139618541624e-06, "loss": 1.5101, "step": 17057 }, { "epoch": 1.336833855799373, "grad_norm": 2.220210075378418, "learning_rate": 7.505872783368449e-06, "loss": 1.6135, "step": 17058 }, { "epoch": 1.3369122257053292, "grad_norm": 2.8337173461914062, "learning_rate": 7.505605938664222e-06, "loss": 1.3869, "step": 17059 }, { "epoch": 1.3369905956112853, "grad_norm": 1.7434866428375244, "learning_rate": 7.505339084429958e-06, "loss": 1.341, "step": 17060 }, { "epoch": 1.3370689655172414, "grad_norm": 2.409644603729248, "learning_rate": 7.505072220666669e-06, "loss": 1.3375, "step": 17061 }, { "epoch": 1.3371473354231975, "grad_norm": 2.087085247039795, "learning_rate": 7.504805347375374e-06, "loss": 1.5482, "step": 17062 }, { "epoch": 1.3372257053291536, "grad_norm": 4.072905540466309, "learning_rate": 7.504538464557085e-06, "loss": 1.7115, "step": 17063 }, { "epoch": 1.3373040752351097, "grad_norm": 2.604952573776245, "learning_rate": 7.504271572212818e-06, "loss": 1.2187, "step": 17064 }, { "epoch": 1.3373824451410659, "grad_norm": 3.125906229019165, "learning_rate": 7.504004670343588e-06, "loss": 1.2387, "step": 17065 }, { "epoch": 1.337460815047022, "grad_norm": 2.426384210586548, "learning_rate": 7.50373775895041e-06, "loss": 1.3424, "step": 17066 }, { "epoch": 1.337539184952978, "grad_norm": 2.871649742126465, "learning_rate": 7.5034708380343e-06, "loss": 1.7611, "step": 17067 }, { "epoch": 1.3376175548589342, "grad_norm": 3.416311740875244, "learning_rate": 7.5032039075962725e-06, "loss": 1.1837, "step": 17068 }, { "epoch": 1.3376959247648903, "grad_norm": 2.6578280925750732, "learning_rate": 7.5029369676373454e-06, "loss": 1.1094, "step": 17069 }, { "epoch": 1.3377742946708464, "grad_norm": 2.9782094955444336, "learning_rate": 7.502670018158529e-06, "loss": 1.5254, "step": 17070 }, { "epoch": 1.3378526645768025, "grad_norm": 3.5685808658599854, "learning_rate": 7.502403059160843e-06, "loss": 1.2648, "step": 17071 }, { "epoch": 1.3379310344827586, "grad_norm": 2.9519777297973633, "learning_rate": 7.502136090645298e-06, "loss": 0.8799, "step": 17072 }, { "epoch": 1.3380094043887147, "grad_norm": 2.1476192474365234, "learning_rate": 7.501869112612916e-06, "loss": 1.2368, "step": 17073 }, { "epoch": 1.3380877742946709, "grad_norm": 2.487821578979492, "learning_rate": 7.501602125064708e-06, "loss": 1.208, "step": 17074 }, { "epoch": 1.338166144200627, "grad_norm": 8.035805702209473, "learning_rate": 7.50133512800169e-06, "loss": 1.7607, "step": 17075 }, { "epoch": 1.338244514106583, "grad_norm": 4.11090612411499, "learning_rate": 7.501068121424879e-06, "loss": 1.3898, "step": 17076 }, { "epoch": 1.3383228840125392, "grad_norm": 3.546140193939209, "learning_rate": 7.500801105335287e-06, "loss": 1.2319, "step": 17077 }, { "epoch": 1.3384012539184953, "grad_norm": 2.595763921737671, "learning_rate": 7.500534079733935e-06, "loss": 1.2779, "step": 17078 }, { "epoch": 1.3384796238244514, "grad_norm": 2.5358424186706543, "learning_rate": 7.500267044621832e-06, "loss": 1.3545, "step": 17079 }, { "epoch": 1.3385579937304075, "grad_norm": 3.743821859359741, "learning_rate": 7.500000000000001e-06, "loss": 1.4605, "step": 17080 }, { "epoch": 1.3386363636363636, "grad_norm": 4.090908527374268, "learning_rate": 7.499732945869454e-06, "loss": 1.0478, "step": 17081 }, { "epoch": 1.3387147335423197, "grad_norm": 3.4821856021881104, "learning_rate": 7.499465882231204e-06, "loss": 1.3741, "step": 17082 }, { "epoch": 1.3387931034482758, "grad_norm": 3.1042604446411133, "learning_rate": 7.499198809086271e-06, "loss": 1.6275, "step": 17083 }, { "epoch": 1.338871473354232, "grad_norm": 4.669977188110352, "learning_rate": 7.49893172643567e-06, "loss": 1.5497, "step": 17084 }, { "epoch": 1.338949843260188, "grad_norm": 3.1913492679595947, "learning_rate": 7.498664634280415e-06, "loss": 1.4852, "step": 17085 }, { "epoch": 1.3390282131661442, "grad_norm": 3.707003355026245, "learning_rate": 7.498397532621524e-06, "loss": 1.3027, "step": 17086 }, { "epoch": 1.3391065830721003, "grad_norm": 3.6093034744262695, "learning_rate": 7.498130421460013e-06, "loss": 0.9153, "step": 17087 }, { "epoch": 1.3391849529780564, "grad_norm": 2.042262554168701, "learning_rate": 7.497863300796896e-06, "loss": 0.808, "step": 17088 }, { "epoch": 1.3392633228840125, "grad_norm": 2.256377935409546, "learning_rate": 7.497596170633189e-06, "loss": 1.5057, "step": 17089 }, { "epoch": 1.3393416927899686, "grad_norm": 3.092057466506958, "learning_rate": 7.497329030969911e-06, "loss": 1.567, "step": 17090 }, { "epoch": 1.3394200626959247, "grad_norm": 2.2358059883117676, "learning_rate": 7.497061881808074e-06, "loss": 1.4291, "step": 17091 }, { "epoch": 1.3394984326018808, "grad_norm": 2.2677628993988037, "learning_rate": 7.496794723148698e-06, "loss": 1.21, "step": 17092 }, { "epoch": 1.339576802507837, "grad_norm": 5.594866752624512, "learning_rate": 7.496527554992796e-06, "loss": 1.5227, "step": 17093 }, { "epoch": 1.339655172413793, "grad_norm": 3.089848756790161, "learning_rate": 7.496260377341387e-06, "loss": 1.5621, "step": 17094 }, { "epoch": 1.3397335423197492, "grad_norm": 2.832930326461792, "learning_rate": 7.495993190195485e-06, "loss": 1.0479, "step": 17095 }, { "epoch": 1.3398119122257053, "grad_norm": 2.295011043548584, "learning_rate": 7.495725993556106e-06, "loss": 1.3081, "step": 17096 }, { "epoch": 1.3398902821316614, "grad_norm": 3.030608654022217, "learning_rate": 7.495458787424268e-06, "loss": 1.528, "step": 17097 }, { "epoch": 1.3399686520376175, "grad_norm": 3.543172597885132, "learning_rate": 7.495191571800985e-06, "loss": 1.5158, "step": 17098 }, { "epoch": 1.3400470219435736, "grad_norm": 3.392310857772827, "learning_rate": 7.4949243466872775e-06, "loss": 1.3708, "step": 17099 }, { "epoch": 1.3401253918495297, "grad_norm": 2.4462900161743164, "learning_rate": 7.4946571120841585e-06, "loss": 1.5207, "step": 17100 }, { "epoch": 1.3402037617554858, "grad_norm": 2.4128613471984863, "learning_rate": 7.494389867992644e-06, "loss": 1.0144, "step": 17101 }, { "epoch": 1.340282131661442, "grad_norm": 2.144078493118286, "learning_rate": 7.494122614413753e-06, "loss": 1.0202, "step": 17102 }, { "epoch": 1.340360501567398, "grad_norm": 4.237800121307373, "learning_rate": 7.493855351348499e-06, "loss": 1.7047, "step": 17103 }, { "epoch": 1.3404388714733542, "grad_norm": 5.402835369110107, "learning_rate": 7.493588078797902e-06, "loss": 1.1912, "step": 17104 }, { "epoch": 1.3405172413793103, "grad_norm": 3.316185474395752, "learning_rate": 7.4933207967629755e-06, "loss": 0.9711, "step": 17105 }, { "epoch": 1.3405956112852664, "grad_norm": 3.9705004692077637, "learning_rate": 7.493053505244737e-06, "loss": 1.4127, "step": 17106 }, { "epoch": 1.3406739811912225, "grad_norm": 2.610386371612549, "learning_rate": 7.492786204244205e-06, "loss": 1.3603, "step": 17107 }, { "epoch": 1.3407523510971786, "grad_norm": 2.8345718383789062, "learning_rate": 7.492518893762393e-06, "loss": 1.2746, "step": 17108 }, { "epoch": 1.3408307210031347, "grad_norm": 2.4866654872894287, "learning_rate": 7.49225157380032e-06, "loss": 1.3028, "step": 17109 }, { "epoch": 1.3409090909090908, "grad_norm": 4.320090293884277, "learning_rate": 7.491984244359003e-06, "loss": 1.6705, "step": 17110 }, { "epoch": 1.340987460815047, "grad_norm": 7.674319267272949, "learning_rate": 7.491716905439456e-06, "loss": 1.1995, "step": 17111 }, { "epoch": 1.341065830721003, "grad_norm": 9.2221097946167, "learning_rate": 7.491449557042699e-06, "loss": 1.2584, "step": 17112 }, { "epoch": 1.3411442006269594, "grad_norm": 3.1859593391418457, "learning_rate": 7.491182199169748e-06, "loss": 1.5705, "step": 17113 }, { "epoch": 1.3412225705329153, "grad_norm": 2.3822779655456543, "learning_rate": 7.490914831821618e-06, "loss": 1.1545, "step": 17114 }, { "epoch": 1.3413009404388716, "grad_norm": 2.384934186935425, "learning_rate": 7.490647454999329e-06, "loss": 1.2723, "step": 17115 }, { "epoch": 1.3413793103448275, "grad_norm": 2.5899887084960938, "learning_rate": 7.490380068703897e-06, "loss": 1.5863, "step": 17116 }, { "epoch": 1.3414576802507838, "grad_norm": 2.2105629444122314, "learning_rate": 7.490112672936336e-06, "loss": 1.0926, "step": 17117 }, { "epoch": 1.3415360501567397, "grad_norm": 3.7262420654296875, "learning_rate": 7.489845267697666e-06, "loss": 1.5977, "step": 17118 }, { "epoch": 1.341614420062696, "grad_norm": 4.2769880294799805, "learning_rate": 7.489577852988905e-06, "loss": 1.4053, "step": 17119 }, { "epoch": 1.341692789968652, "grad_norm": 3.2577221393585205, "learning_rate": 7.489310428811068e-06, "loss": 1.1206, "step": 17120 }, { "epoch": 1.3417711598746083, "grad_norm": 2.500965118408203, "learning_rate": 7.4890429951651714e-06, "loss": 1.5025, "step": 17121 }, { "epoch": 1.3418495297805642, "grad_norm": 1.8431041240692139, "learning_rate": 7.4887755520522344e-06, "loss": 1.3433, "step": 17122 }, { "epoch": 1.3419278996865205, "grad_norm": 2.0973896980285645, "learning_rate": 7.488508099473275e-06, "loss": 1.2147, "step": 17123 }, { "epoch": 1.3420062695924764, "grad_norm": 2.491734743118286, "learning_rate": 7.488240637429306e-06, "loss": 1.4314, "step": 17124 }, { "epoch": 1.3420846394984327, "grad_norm": 6.172510147094727, "learning_rate": 7.487973165921349e-06, "loss": 1.1807, "step": 17125 }, { "epoch": 1.3421630094043886, "grad_norm": 2.524611473083496, "learning_rate": 7.48770568495042e-06, "loss": 1.3665, "step": 17126 }, { "epoch": 1.342241379310345, "grad_norm": 3.862323522567749, "learning_rate": 7.487438194517538e-06, "loss": 1.2644, "step": 17127 }, { "epoch": 1.3423197492163008, "grad_norm": 2.1939308643341064, "learning_rate": 7.487170694623717e-06, "loss": 1.3712, "step": 17128 }, { "epoch": 1.3423981191222571, "grad_norm": 3.3919553756713867, "learning_rate": 7.486903185269976e-06, "loss": 1.3141, "step": 17129 }, { "epoch": 1.342476489028213, "grad_norm": 4.248289108276367, "learning_rate": 7.4866356664573334e-06, "loss": 1.4415, "step": 17130 }, { "epoch": 1.3425548589341694, "grad_norm": 2.144108772277832, "learning_rate": 7.486368138186806e-06, "loss": 1.1831, "step": 17131 }, { "epoch": 1.3426332288401253, "grad_norm": 3.758820056915283, "learning_rate": 7.486100600459412e-06, "loss": 1.1717, "step": 17132 }, { "epoch": 1.3427115987460816, "grad_norm": 7.559966087341309, "learning_rate": 7.485833053276166e-06, "loss": 2.0082, "step": 17133 }, { "epoch": 1.3427899686520375, "grad_norm": 3.9328601360321045, "learning_rate": 7.4855654966380896e-06, "loss": 1.1076, "step": 17134 }, { "epoch": 1.3428683385579938, "grad_norm": 2.4713313579559326, "learning_rate": 7.485297930546198e-06, "loss": 1.3433, "step": 17135 }, { "epoch": 1.34294670846395, "grad_norm": 2.780609607696533, "learning_rate": 7.48503035500151e-06, "loss": 1.4521, "step": 17136 }, { "epoch": 1.343025078369906, "grad_norm": 9.620870590209961, "learning_rate": 7.4847627700050425e-06, "loss": 1.2238, "step": 17137 }, { "epoch": 1.3431034482758621, "grad_norm": 7.633688926696777, "learning_rate": 7.484495175557814e-06, "loss": 1.9806, "step": 17138 }, { "epoch": 1.3431818181818183, "grad_norm": 2.2082343101501465, "learning_rate": 7.484227571660841e-06, "loss": 0.9674, "step": 17139 }, { "epoch": 1.3432601880877744, "grad_norm": 3.015021800994873, "learning_rate": 7.483959958315143e-06, "loss": 1.7425, "step": 17140 }, { "epoch": 1.3433385579937305, "grad_norm": 3.660999059677124, "learning_rate": 7.483692335521737e-06, "loss": 1.4035, "step": 17141 }, { "epoch": 1.3434169278996866, "grad_norm": 2.546508550643921, "learning_rate": 7.483424703281643e-06, "loss": 1.3542, "step": 17142 }, { "epoch": 1.3434952978056427, "grad_norm": 6.089187145233154, "learning_rate": 7.483157061595874e-06, "loss": 1.4119, "step": 17143 }, { "epoch": 1.3435736677115988, "grad_norm": 1.6123480796813965, "learning_rate": 7.482889410465453e-06, "loss": 1.3254, "step": 17144 }, { "epoch": 1.343652037617555, "grad_norm": 3.56099796295166, "learning_rate": 7.482621749891395e-06, "loss": 1.3467, "step": 17145 }, { "epoch": 1.343730407523511, "grad_norm": 3.4026167392730713, "learning_rate": 7.48235407987472e-06, "loss": 1.5935, "step": 17146 }, { "epoch": 1.3438087774294671, "grad_norm": 10.142470359802246, "learning_rate": 7.482086400416444e-06, "loss": 1.2503, "step": 17147 }, { "epoch": 1.3438871473354232, "grad_norm": 4.854825496673584, "learning_rate": 7.481818711517587e-06, "loss": 2.4886, "step": 17148 }, { "epoch": 1.3439655172413794, "grad_norm": 5.680945873260498, "learning_rate": 7.481551013179167e-06, "loss": 1.8813, "step": 17149 }, { "epoch": 1.3440438871473355, "grad_norm": 4.006632328033447, "learning_rate": 7.4812833054022e-06, "loss": 0.9476, "step": 17150 }, { "epoch": 1.3441222570532916, "grad_norm": 4.22556734085083, "learning_rate": 7.481015588187707e-06, "loss": 1.0161, "step": 17151 }, { "epoch": 1.3442006269592477, "grad_norm": 3.2120249271392822, "learning_rate": 7.480747861536702e-06, "loss": 1.3621, "step": 17152 }, { "epoch": 1.3442789968652038, "grad_norm": 3.839782953262329, "learning_rate": 7.48048012545021e-06, "loss": 1.2541, "step": 17153 }, { "epoch": 1.34435736677116, "grad_norm": 3.0298218727111816, "learning_rate": 7.480212379929244e-06, "loss": 1.2548, "step": 17154 }, { "epoch": 1.344435736677116, "grad_norm": 2.569185972213745, "learning_rate": 7.479944624974824e-06, "loss": 1.4307, "step": 17155 }, { "epoch": 1.3445141065830721, "grad_norm": 6.075372219085693, "learning_rate": 7.47967686058797e-06, "loss": 1.6855, "step": 17156 }, { "epoch": 1.3445924764890282, "grad_norm": 3.6827845573425293, "learning_rate": 7.479409086769698e-06, "loss": 1.3932, "step": 17157 }, { "epoch": 1.3446708463949844, "grad_norm": 2.6624529361724854, "learning_rate": 7.479141303521026e-06, "loss": 1.2858, "step": 17158 }, { "epoch": 1.3447492163009405, "grad_norm": 3.5087711811065674, "learning_rate": 7.478873510842974e-06, "loss": 1.4599, "step": 17159 }, { "epoch": 1.3448275862068966, "grad_norm": 4.579308032989502, "learning_rate": 7.478605708736563e-06, "loss": 1.4562, "step": 17160 }, { "epoch": 1.3449059561128527, "grad_norm": 3.041123390197754, "learning_rate": 7.478337897202807e-06, "loss": 1.3957, "step": 17161 }, { "epoch": 1.3449843260188088, "grad_norm": 3.355168104171753, "learning_rate": 7.478070076242727e-06, "loss": 1.1892, "step": 17162 }, { "epoch": 1.345062695924765, "grad_norm": 3.3822243213653564, "learning_rate": 7.477802245857341e-06, "loss": 1.4589, "step": 17163 }, { "epoch": 1.345141065830721, "grad_norm": 2.8364627361297607, "learning_rate": 7.477534406047667e-06, "loss": 1.444, "step": 17164 }, { "epoch": 1.3452194357366771, "grad_norm": 5.25086784362793, "learning_rate": 7.477266556814726e-06, "loss": 1.5704, "step": 17165 }, { "epoch": 1.3452978056426332, "grad_norm": 3.104715347290039, "learning_rate": 7.476998698159535e-06, "loss": 1.2698, "step": 17166 }, { "epoch": 1.3453761755485893, "grad_norm": 3.207225799560547, "learning_rate": 7.4767308300831136e-06, "loss": 1.347, "step": 17167 }, { "epoch": 1.3454545454545455, "grad_norm": 2.511467933654785, "learning_rate": 7.4764629525864795e-06, "loss": 1.4362, "step": 17168 }, { "epoch": 1.3455329153605016, "grad_norm": 2.8604609966278076, "learning_rate": 7.476195065670653e-06, "loss": 1.2319, "step": 17169 }, { "epoch": 1.3456112852664577, "grad_norm": 3.4606494903564453, "learning_rate": 7.475927169336653e-06, "loss": 1.4703, "step": 17170 }, { "epoch": 1.3456896551724138, "grad_norm": 4.499128341674805, "learning_rate": 7.475659263585495e-06, "loss": 1.3558, "step": 17171 }, { "epoch": 1.34576802507837, "grad_norm": 3.8515608310699463, "learning_rate": 7.475391348418204e-06, "loss": 1.6226, "step": 17172 }, { "epoch": 1.345846394984326, "grad_norm": 3.3988287448883057, "learning_rate": 7.475123423835795e-06, "loss": 1.4177, "step": 17173 }, { "epoch": 1.3459247648902821, "grad_norm": 3.2034411430358887, "learning_rate": 7.474855489839287e-06, "loss": 1.3044, "step": 17174 }, { "epoch": 1.3460031347962382, "grad_norm": 2.733793258666992, "learning_rate": 7.474587546429699e-06, "loss": 1.0801, "step": 17175 }, { "epoch": 1.3460815047021943, "grad_norm": 3.0954387187957764, "learning_rate": 7.474319593608053e-06, "loss": 1.526, "step": 17176 }, { "epoch": 1.3461598746081505, "grad_norm": 7.659568786621094, "learning_rate": 7.474051631375365e-06, "loss": 0.7877, "step": 17177 }, { "epoch": 1.3462382445141066, "grad_norm": 4.8828535079956055, "learning_rate": 7.473783659732655e-06, "loss": 1.504, "step": 17178 }, { "epoch": 1.3463166144200627, "grad_norm": 2.701805591583252, "learning_rate": 7.473515678680943e-06, "loss": 0.9404, "step": 17179 }, { "epoch": 1.3463949843260188, "grad_norm": 4.517419815063477, "learning_rate": 7.473247688221248e-06, "loss": 1.0682, "step": 17180 }, { "epoch": 1.346473354231975, "grad_norm": 2.461588144302368, "learning_rate": 7.47297968835459e-06, "loss": 1.2726, "step": 17181 }, { "epoch": 1.346551724137931, "grad_norm": 4.144700050354004, "learning_rate": 7.472711679081986e-06, "loss": 1.5216, "step": 17182 }, { "epoch": 1.3466300940438871, "grad_norm": 1.984686255455017, "learning_rate": 7.4724436604044584e-06, "loss": 1.4272, "step": 17183 }, { "epoch": 1.3467084639498432, "grad_norm": 2.2603940963745117, "learning_rate": 7.4721756323230236e-06, "loss": 1.217, "step": 17184 }, { "epoch": 1.3467868338557993, "grad_norm": 3.7404913902282715, "learning_rate": 7.471907594838703e-06, "loss": 1.1549, "step": 17185 }, { "epoch": 1.3468652037617554, "grad_norm": 3.3039627075195312, "learning_rate": 7.471639547952515e-06, "loss": 1.5992, "step": 17186 }, { "epoch": 1.3469435736677116, "grad_norm": 3.737276554107666, "learning_rate": 7.47137149166548e-06, "loss": 1.4651, "step": 17187 }, { "epoch": 1.3470219435736677, "grad_norm": 2.6727328300476074, "learning_rate": 7.471103425978618e-06, "loss": 0.8968, "step": 17188 }, { "epoch": 1.3471003134796238, "grad_norm": 2.4536330699920654, "learning_rate": 7.4708353508929474e-06, "loss": 1.6316, "step": 17189 }, { "epoch": 1.34717868338558, "grad_norm": 2.1638145446777344, "learning_rate": 7.470567266409488e-06, "loss": 1.2714, "step": 17190 }, { "epoch": 1.347257053291536, "grad_norm": 3.2159647941589355, "learning_rate": 7.4702991725292595e-06, "loss": 1.1507, "step": 17191 }, { "epoch": 1.347335423197492, "grad_norm": 2.636002779006958, "learning_rate": 7.470031069253282e-06, "loss": 1.2476, "step": 17192 }, { "epoch": 1.3474137931034482, "grad_norm": 4.127438068389893, "learning_rate": 7.469762956582575e-06, "loss": 1.4805, "step": 17193 }, { "epoch": 1.3474921630094043, "grad_norm": 6.720308303833008, "learning_rate": 7.469494834518159e-06, "loss": 0.9358, "step": 17194 }, { "epoch": 1.3475705329153604, "grad_norm": 2.499800205230713, "learning_rate": 7.469226703061051e-06, "loss": 1.5428, "step": 17195 }, { "epoch": 1.3476489028213166, "grad_norm": 2.800976276397705, "learning_rate": 7.468958562212275e-06, "loss": 1.2083, "step": 17196 }, { "epoch": 1.3477272727272727, "grad_norm": 2.4430174827575684, "learning_rate": 7.468690411972847e-06, "loss": 1.3727, "step": 17197 }, { "epoch": 1.3478056426332288, "grad_norm": 7.262555122375488, "learning_rate": 7.468422252343791e-06, "loss": 1.3011, "step": 17198 }, { "epoch": 1.3478840125391849, "grad_norm": 8.485854148864746, "learning_rate": 7.468154083326122e-06, "loss": 1.969, "step": 17199 }, { "epoch": 1.347962382445141, "grad_norm": 3.28240704536438, "learning_rate": 7.467885904920864e-06, "loss": 1.461, "step": 17200 }, { "epoch": 1.348040752351097, "grad_norm": 4.213860511779785, "learning_rate": 7.467617717129036e-06, "loss": 1.2794, "step": 17201 }, { "epoch": 1.3481191222570532, "grad_norm": 4.976203918457031, "learning_rate": 7.467349519951657e-06, "loss": 1.9697, "step": 17202 }, { "epoch": 1.3481974921630093, "grad_norm": 5.17507791519165, "learning_rate": 7.467081313389748e-06, "loss": 1.359, "step": 17203 }, { "epoch": 1.3482758620689654, "grad_norm": 2.868234634399414, "learning_rate": 7.466813097444328e-06, "loss": 1.44, "step": 17204 }, { "epoch": 1.3483542319749215, "grad_norm": 4.278392791748047, "learning_rate": 7.46654487211642e-06, "loss": 1.4138, "step": 17205 }, { "epoch": 1.3484326018808779, "grad_norm": 1.924471139907837, "learning_rate": 7.46627663740704e-06, "loss": 1.3773, "step": 17206 }, { "epoch": 1.3485109717868338, "grad_norm": 2.5951526165008545, "learning_rate": 7.466008393317213e-06, "loss": 1.3813, "step": 17207 }, { "epoch": 1.34858934169279, "grad_norm": 3.107790231704712, "learning_rate": 7.465740139847953e-06, "loss": 1.2211, "step": 17208 }, { "epoch": 1.348667711598746, "grad_norm": 3.5852978229522705, "learning_rate": 7.465471877000287e-06, "loss": 1.5914, "step": 17209 }, { "epoch": 1.3487460815047023, "grad_norm": 9.854165077209473, "learning_rate": 7.465203604775231e-06, "loss": 1.294, "step": 17210 }, { "epoch": 1.3488244514106582, "grad_norm": 2.124387502670288, "learning_rate": 7.4649353231738076e-06, "loss": 1.2996, "step": 17211 }, { "epoch": 1.3489028213166145, "grad_norm": 7.086082935333252, "learning_rate": 7.464667032197035e-06, "loss": 1.5108, "step": 17212 }, { "epoch": 1.3489811912225704, "grad_norm": 2.5832347869873047, "learning_rate": 7.464398731845936e-06, "loss": 1.4637, "step": 17213 }, { "epoch": 1.3490595611285268, "grad_norm": 2.1604509353637695, "learning_rate": 7.46413042212153e-06, "loss": 1.2356, "step": 17214 }, { "epoch": 1.3491379310344827, "grad_norm": 3.2116482257843018, "learning_rate": 7.463862103024837e-06, "loss": 1.5603, "step": 17215 }, { "epoch": 1.349216300940439, "grad_norm": 4.622010231018066, "learning_rate": 7.463593774556879e-06, "loss": 1.274, "step": 17216 }, { "epoch": 1.3492946708463949, "grad_norm": 6.015075206756592, "learning_rate": 7.463325436718675e-06, "loss": 1.3768, "step": 17217 }, { "epoch": 1.3493730407523512, "grad_norm": 5.0285139083862305, "learning_rate": 7.463057089511246e-06, "loss": 1.8648, "step": 17218 }, { "epoch": 1.349451410658307, "grad_norm": 1.796379566192627, "learning_rate": 7.462788732935613e-06, "loss": 1.3786, "step": 17219 }, { "epoch": 1.3495297805642634, "grad_norm": 4.505695343017578, "learning_rate": 7.462520366992796e-06, "loss": 1.6146, "step": 17220 }, { "epoch": 1.3496081504702193, "grad_norm": 2.7603819370269775, "learning_rate": 7.462251991683818e-06, "loss": 1.3756, "step": 17221 }, { "epoch": 1.3496865203761756, "grad_norm": 2.651211977005005, "learning_rate": 7.461983607009697e-06, "loss": 1.3425, "step": 17222 }, { "epoch": 1.3497648902821315, "grad_norm": 8.343387603759766, "learning_rate": 7.461715212971454e-06, "loss": 1.1325, "step": 17223 }, { "epoch": 1.3498432601880879, "grad_norm": 8.382566452026367, "learning_rate": 7.461446809570114e-06, "loss": 1.4295, "step": 17224 }, { "epoch": 1.3499216300940438, "grad_norm": 4.575356960296631, "learning_rate": 7.461178396806692e-06, "loss": 1.7183, "step": 17225 }, { "epoch": 1.35, "grad_norm": 4.218506813049316, "learning_rate": 7.460909974682212e-06, "loss": 1.4165, "step": 17226 }, { "epoch": 1.350078369905956, "grad_norm": 2.0949153900146484, "learning_rate": 7.460641543197694e-06, "loss": 1.5094, "step": 17227 }, { "epoch": 1.3501567398119123, "grad_norm": 2.359565258026123, "learning_rate": 7.46037310235416e-06, "loss": 0.8724, "step": 17228 }, { "epoch": 1.3502351097178684, "grad_norm": 2.8994858264923096, "learning_rate": 7.460104652152631e-06, "loss": 1.1803, "step": 17229 }, { "epoch": 1.3503134796238245, "grad_norm": 3.8454647064208984, "learning_rate": 7.459836192594127e-06, "loss": 1.371, "step": 17230 }, { "epoch": 1.3503918495297806, "grad_norm": 2.316986560821533, "learning_rate": 7.459567723679671e-06, "loss": 1.2351, "step": 17231 }, { "epoch": 1.3504702194357368, "grad_norm": 2.6635212898254395, "learning_rate": 7.4592992454102795e-06, "loss": 1.2972, "step": 17232 }, { "epoch": 1.3505485893416929, "grad_norm": 2.2383995056152344, "learning_rate": 7.459030757786979e-06, "loss": 1.3279, "step": 17233 }, { "epoch": 1.350626959247649, "grad_norm": 3.5304698944091797, "learning_rate": 7.458762260810788e-06, "loss": 1.5857, "step": 17234 }, { "epoch": 1.350705329153605, "grad_norm": 2.328751564025879, "learning_rate": 7.458493754482728e-06, "loss": 1.1718, "step": 17235 }, { "epoch": 1.3507836990595612, "grad_norm": 3.934812545776367, "learning_rate": 7.458225238803821e-06, "loss": 1.447, "step": 17236 }, { "epoch": 1.3508620689655173, "grad_norm": 3.0281453132629395, "learning_rate": 7.457956713775087e-06, "loss": 1.1814, "step": 17237 }, { "epoch": 1.3509404388714734, "grad_norm": 2.0538015365600586, "learning_rate": 7.45768817939755e-06, "loss": 1.3215, "step": 17238 }, { "epoch": 1.3510188087774295, "grad_norm": 4.254906177520752, "learning_rate": 7.4574196356722275e-06, "loss": 1.2894, "step": 17239 }, { "epoch": 1.3510971786833856, "grad_norm": 2.3431997299194336, "learning_rate": 7.457151082600145e-06, "loss": 1.3481, "step": 17240 }, { "epoch": 1.3511755485893417, "grad_norm": 3.243665933609009, "learning_rate": 7.45688252018232e-06, "loss": 1.4183, "step": 17241 }, { "epoch": 1.3512539184952979, "grad_norm": 2.7301647663116455, "learning_rate": 7.456613948419775e-06, "loss": 1.4722, "step": 17242 }, { "epoch": 1.351332288401254, "grad_norm": 5.5882744789123535, "learning_rate": 7.456345367313535e-06, "loss": 1.9203, "step": 17243 }, { "epoch": 1.35141065830721, "grad_norm": 3.2269067764282227, "learning_rate": 7.456076776864618e-06, "loss": 1.2933, "step": 17244 }, { "epoch": 1.3514890282131662, "grad_norm": 2.879117250442505, "learning_rate": 7.455808177074046e-06, "loss": 1.2825, "step": 17245 }, { "epoch": 1.3515673981191223, "grad_norm": 2.1112730503082275, "learning_rate": 7.455539567942841e-06, "loss": 1.4709, "step": 17246 }, { "epoch": 1.3516457680250784, "grad_norm": 2.715599536895752, "learning_rate": 7.455270949472025e-06, "loss": 1.1316, "step": 17247 }, { "epoch": 1.3517241379310345, "grad_norm": 6.013223648071289, "learning_rate": 7.455002321662619e-06, "loss": 1.3618, "step": 17248 }, { "epoch": 1.3518025078369906, "grad_norm": 3.248236894607544, "learning_rate": 7.454733684515646e-06, "loss": 1.23, "step": 17249 }, { "epoch": 1.3518808777429467, "grad_norm": 3.3584327697753906, "learning_rate": 7.454465038032127e-06, "loss": 1.6423, "step": 17250 }, { "epoch": 1.3519592476489029, "grad_norm": 3.612607955932617, "learning_rate": 7.454196382213082e-06, "loss": 1.2449, "step": 17251 }, { "epoch": 1.352037617554859, "grad_norm": 3.0644354820251465, "learning_rate": 7.453927717059537e-06, "loss": 1.3062, "step": 17252 }, { "epoch": 1.352115987460815, "grad_norm": 2.487067222595215, "learning_rate": 7.4536590425725085e-06, "loss": 1.2828, "step": 17253 }, { "epoch": 1.3521943573667712, "grad_norm": 2.2425808906555176, "learning_rate": 7.453390358753023e-06, "loss": 1.4259, "step": 17254 }, { "epoch": 1.3522727272727273, "grad_norm": 3.983323097229004, "learning_rate": 7.453121665602101e-06, "loss": 1.1567, "step": 17255 }, { "epoch": 1.3523510971786834, "grad_norm": 5.173330783843994, "learning_rate": 7.452852963120763e-06, "loss": 1.4633, "step": 17256 }, { "epoch": 1.3524294670846395, "grad_norm": 2.561298370361328, "learning_rate": 7.4525842513100346e-06, "loss": 1.3151, "step": 17257 }, { "epoch": 1.3525078369905956, "grad_norm": 2.2784817218780518, "learning_rate": 7.452315530170933e-06, "loss": 1.3272, "step": 17258 }, { "epoch": 1.3525862068965517, "grad_norm": 3.5064573287963867, "learning_rate": 7.452046799704483e-06, "loss": 1.2994, "step": 17259 }, { "epoch": 1.3526645768025078, "grad_norm": 5.552616596221924, "learning_rate": 7.451778059911706e-06, "loss": 1.8952, "step": 17260 }, { "epoch": 1.352742946708464, "grad_norm": 2.9420902729034424, "learning_rate": 7.4515093107936265e-06, "loss": 1.4151, "step": 17261 }, { "epoch": 1.35282131661442, "grad_norm": 7.789029121398926, "learning_rate": 7.451240552351263e-06, "loss": 1.176, "step": 17262 }, { "epoch": 1.3528996865203762, "grad_norm": 6.0069708824157715, "learning_rate": 7.450971784585641e-06, "loss": 1.2114, "step": 17263 }, { "epoch": 1.3529780564263323, "grad_norm": 3.4591829776763916, "learning_rate": 7.450703007497779e-06, "loss": 1.3798, "step": 17264 }, { "epoch": 1.3530564263322884, "grad_norm": 3.629916191101074, "learning_rate": 7.450434221088704e-06, "loss": 1.0799, "step": 17265 }, { "epoch": 1.3531347962382445, "grad_norm": 5.96800422668457, "learning_rate": 7.450165425359435e-06, "loss": 1.5393, "step": 17266 }, { "epoch": 1.3532131661442006, "grad_norm": 3.0667176246643066, "learning_rate": 7.4498966203109945e-06, "loss": 1.4839, "step": 17267 }, { "epoch": 1.3532915360501567, "grad_norm": 11.981147766113281, "learning_rate": 7.449627805944405e-06, "loss": 1.1125, "step": 17268 }, { "epoch": 1.3533699059561128, "grad_norm": 3.633566379547119, "learning_rate": 7.44935898226069e-06, "loss": 1.5469, "step": 17269 }, { "epoch": 1.353448275862069, "grad_norm": 2.8679542541503906, "learning_rate": 7.449090149260873e-06, "loss": 1.3456, "step": 17270 }, { "epoch": 1.353526645768025, "grad_norm": 2.1441915035247803, "learning_rate": 7.448821306945975e-06, "loss": 1.3463, "step": 17271 }, { "epoch": 1.3536050156739812, "grad_norm": 2.5689492225646973, "learning_rate": 7.4485524553170175e-06, "loss": 1.5711, "step": 17272 }, { "epoch": 1.3536833855799373, "grad_norm": 3.5262773036956787, "learning_rate": 7.448283594375023e-06, "loss": 1.2132, "step": 17273 }, { "epoch": 1.3537617554858934, "grad_norm": 3.105652093887329, "learning_rate": 7.4480147241210174e-06, "loss": 1.4484, "step": 17274 }, { "epoch": 1.3538401253918495, "grad_norm": 2.9114086627960205, "learning_rate": 7.44774584455602e-06, "loss": 1.0854, "step": 17275 }, { "epoch": 1.3539184952978056, "grad_norm": 3.0205368995666504, "learning_rate": 7.447476955681054e-06, "loss": 1.6896, "step": 17276 }, { "epoch": 1.3539968652037617, "grad_norm": 5.898192405700684, "learning_rate": 7.447208057497144e-06, "loss": 1.3362, "step": 17277 }, { "epoch": 1.3540752351097178, "grad_norm": 2.672818183898926, "learning_rate": 7.446939150005312e-06, "loss": 1.0991, "step": 17278 }, { "epoch": 1.354153605015674, "grad_norm": 2.2514126300811768, "learning_rate": 7.446670233206579e-06, "loss": 1.1899, "step": 17279 }, { "epoch": 1.35423197492163, "grad_norm": 2.660973072052002, "learning_rate": 7.44640130710197e-06, "loss": 1.3892, "step": 17280 }, { "epoch": 1.3543103448275862, "grad_norm": 2.3168270587921143, "learning_rate": 7.446132371692505e-06, "loss": 1.3107, "step": 17281 }, { "epoch": 1.3543887147335423, "grad_norm": 2.387380599975586, "learning_rate": 7.445863426979212e-06, "loss": 1.3627, "step": 17282 }, { "epoch": 1.3544670846394984, "grad_norm": 2.8907408714294434, "learning_rate": 7.445594472963109e-06, "loss": 1.6389, "step": 17283 }, { "epoch": 1.3545454545454545, "grad_norm": 3.485483407974243, "learning_rate": 7.44532550964522e-06, "loss": 1.5684, "step": 17284 }, { "epoch": 1.3546238244514106, "grad_norm": 2.5559308528900146, "learning_rate": 7.445056537026572e-06, "loss": 1.2455, "step": 17285 }, { "epoch": 1.3547021943573667, "grad_norm": 2.4141430854797363, "learning_rate": 7.444787555108181e-06, "loss": 1.2871, "step": 17286 }, { "epoch": 1.3547805642633228, "grad_norm": 3.3006296157836914, "learning_rate": 7.444518563891076e-06, "loss": 1.3053, "step": 17287 }, { "epoch": 1.354858934169279, "grad_norm": 2.3792591094970703, "learning_rate": 7.444249563376277e-06, "loss": 1.2607, "step": 17288 }, { "epoch": 1.354937304075235, "grad_norm": 5.825366973876953, "learning_rate": 7.4439805535648095e-06, "loss": 1.5041, "step": 17289 }, { "epoch": 1.3550156739811912, "grad_norm": 2.811645030975342, "learning_rate": 7.4437115344576935e-06, "loss": 1.4596, "step": 17290 }, { "epoch": 1.3550940438871473, "grad_norm": 5.062646865844727, "learning_rate": 7.4434425060559546e-06, "loss": 0.9715, "step": 17291 }, { "epoch": 1.3551724137931034, "grad_norm": 2.472560405731201, "learning_rate": 7.443173468360616e-06, "loss": 1.4679, "step": 17292 }, { "epoch": 1.3552507836990595, "grad_norm": 4.002954006195068, "learning_rate": 7.442904421372699e-06, "loss": 1.5103, "step": 17293 }, { "epoch": 1.3553291536050156, "grad_norm": 3.0846521854400635, "learning_rate": 7.4426353650932295e-06, "loss": 1.3506, "step": 17294 }, { "epoch": 1.3554075235109717, "grad_norm": 2.8206496238708496, "learning_rate": 7.442366299523229e-06, "loss": 1.2507, "step": 17295 }, { "epoch": 1.3554858934169278, "grad_norm": 4.8888373374938965, "learning_rate": 7.442097224663723e-06, "loss": 1.3404, "step": 17296 }, { "epoch": 1.355564263322884, "grad_norm": 1.99112868309021, "learning_rate": 7.4418281405157334e-06, "loss": 1.3857, "step": 17297 }, { "epoch": 1.35564263322884, "grad_norm": 5.870800018310547, "learning_rate": 7.441559047080281e-06, "loss": 1.1612, "step": 17298 }, { "epoch": 1.3557210031347964, "grad_norm": 2.9440648555755615, "learning_rate": 7.441289944358395e-06, "loss": 0.8557, "step": 17299 }, { "epoch": 1.3557993730407523, "grad_norm": 2.636884927749634, "learning_rate": 7.441020832351096e-06, "loss": 1.4255, "step": 17300 }, { "epoch": 1.3558777429467086, "grad_norm": 5.1447930335998535, "learning_rate": 7.4407517110594065e-06, "loss": 1.0862, "step": 17301 }, { "epoch": 1.3559561128526645, "grad_norm": 3.430198907852173, "learning_rate": 7.4404825804843505e-06, "loss": 1.3638, "step": 17302 }, { "epoch": 1.3560344827586208, "grad_norm": 2.354865074157715, "learning_rate": 7.440213440626952e-06, "loss": 1.2061, "step": 17303 }, { "epoch": 1.3561128526645767, "grad_norm": 2.3185136318206787, "learning_rate": 7.439944291488237e-06, "loss": 1.501, "step": 17304 }, { "epoch": 1.356191222570533, "grad_norm": 2.1187057495117188, "learning_rate": 7.439675133069225e-06, "loss": 1.003, "step": 17305 }, { "epoch": 1.356269592476489, "grad_norm": 2.470611095428467, "learning_rate": 7.4394059653709435e-06, "loss": 1.5579, "step": 17306 }, { "epoch": 1.3563479623824453, "grad_norm": 12.42461109161377, "learning_rate": 7.439136788394413e-06, "loss": 1.1015, "step": 17307 }, { "epoch": 1.3564263322884011, "grad_norm": 4.004929065704346, "learning_rate": 7.438867602140662e-06, "loss": 0.8797, "step": 17308 }, { "epoch": 1.3565047021943575, "grad_norm": 4.537670612335205, "learning_rate": 7.438598406610709e-06, "loss": 1.3421, "step": 17309 }, { "epoch": 1.3565830721003134, "grad_norm": 3.7789549827575684, "learning_rate": 7.438329201805581e-06, "loss": 1.2613, "step": 17310 }, { "epoch": 1.3566614420062697, "grad_norm": 4.322999954223633, "learning_rate": 7.4380599877263e-06, "loss": 1.7247, "step": 17311 }, { "epoch": 1.3567398119122256, "grad_norm": 2.06919002532959, "learning_rate": 7.4377907643738915e-06, "loss": 1.12, "step": 17312 }, { "epoch": 1.356818181818182, "grad_norm": 3.7072181701660156, "learning_rate": 7.437521531749381e-06, "loss": 1.0909, "step": 17313 }, { "epoch": 1.3568965517241378, "grad_norm": 3.3844869136810303, "learning_rate": 7.437252289853787e-06, "loss": 1.4359, "step": 17314 }, { "epoch": 1.3569749216300941, "grad_norm": 2.7664108276367188, "learning_rate": 7.43698303868814e-06, "loss": 1.353, "step": 17315 }, { "epoch": 1.35705329153605, "grad_norm": 2.3832077980041504, "learning_rate": 7.43671377825346e-06, "loss": 1.5365, "step": 17316 }, { "epoch": 1.3571316614420064, "grad_norm": 2.430656671524048, "learning_rate": 7.436444508550773e-06, "loss": 1.5701, "step": 17317 }, { "epoch": 1.3572100313479623, "grad_norm": 2.848609209060669, "learning_rate": 7.436175229581102e-06, "loss": 1.2284, "step": 17318 }, { "epoch": 1.3572884012539186, "grad_norm": 1.9872311353683472, "learning_rate": 7.435905941345471e-06, "loss": 1.5903, "step": 17319 }, { "epoch": 1.3573667711598745, "grad_norm": 1.76265549659729, "learning_rate": 7.4356366438449065e-06, "loss": 1.2746, "step": 17320 }, { "epoch": 1.3574451410658308, "grad_norm": 3.8311057090759277, "learning_rate": 7.4353673370804304e-06, "loss": 1.836, "step": 17321 }, { "epoch": 1.3575235109717867, "grad_norm": 3.2782113552093506, "learning_rate": 7.435098021053069e-06, "loss": 1.4272, "step": 17322 }, { "epoch": 1.357601880877743, "grad_norm": 3.693742513656616, "learning_rate": 7.434828695763844e-06, "loss": 1.307, "step": 17323 }, { "epoch": 1.3576802507836991, "grad_norm": 2.874896764755249, "learning_rate": 7.434559361213781e-06, "loss": 1.3965, "step": 17324 }, { "epoch": 1.3577586206896552, "grad_norm": 2.9887897968292236, "learning_rate": 7.434290017403907e-06, "loss": 0.7799, "step": 17325 }, { "epoch": 1.3578369905956114, "grad_norm": 2.837996244430542, "learning_rate": 7.434020664335241e-06, "loss": 1.1552, "step": 17326 }, { "epoch": 1.3579153605015675, "grad_norm": 2.2867541313171387, "learning_rate": 7.433751302008812e-06, "loss": 1.3847, "step": 17327 }, { "epoch": 1.3579937304075236, "grad_norm": 2.8395707607269287, "learning_rate": 7.433481930425644e-06, "loss": 1.3046, "step": 17328 }, { "epoch": 1.3580721003134797, "grad_norm": 2.6815667152404785, "learning_rate": 7.43321254958676e-06, "loss": 1.4139, "step": 17329 }, { "epoch": 1.3581504702194358, "grad_norm": 3.4267072677612305, "learning_rate": 7.432943159493184e-06, "loss": 1.5071, "step": 17330 }, { "epoch": 1.358228840125392, "grad_norm": 2.1840977668762207, "learning_rate": 7.4326737601459435e-06, "loss": 1.4512, "step": 17331 }, { "epoch": 1.358307210031348, "grad_norm": 2.2491283416748047, "learning_rate": 7.432404351546062e-06, "loss": 1.3889, "step": 17332 }, { "epoch": 1.3583855799373041, "grad_norm": 1.9852616786956787, "learning_rate": 7.432134933694561e-06, "loss": 1.4911, "step": 17333 }, { "epoch": 1.3584639498432602, "grad_norm": 3.086803913116455, "learning_rate": 7.431865506592471e-06, "loss": 1.4603, "step": 17334 }, { "epoch": 1.3585423197492164, "grad_norm": 3.7169830799102783, "learning_rate": 7.431596070240812e-06, "loss": 1.7517, "step": 17335 }, { "epoch": 1.3586206896551725, "grad_norm": 4.080387592315674, "learning_rate": 7.431326624640611e-06, "loss": 1.6171, "step": 17336 }, { "epoch": 1.3586990595611286, "grad_norm": 2.817579984664917, "learning_rate": 7.431057169792892e-06, "loss": 1.5219, "step": 17337 }, { "epoch": 1.3587774294670847, "grad_norm": 3.1886632442474365, "learning_rate": 7.430787705698681e-06, "loss": 1.2377, "step": 17338 }, { "epoch": 1.3588557993730408, "grad_norm": 2.3694448471069336, "learning_rate": 7.430518232359001e-06, "loss": 1.2023, "step": 17339 }, { "epoch": 1.358934169278997, "grad_norm": 2.243123769760132, "learning_rate": 7.430248749774878e-06, "loss": 1.1605, "step": 17340 }, { "epoch": 1.359012539184953, "grad_norm": 3.056011438369751, "learning_rate": 7.429979257947338e-06, "loss": 1.4436, "step": 17341 }, { "epoch": 1.3590909090909091, "grad_norm": 3.3406870365142822, "learning_rate": 7.429709756877404e-06, "loss": 0.8861, "step": 17342 }, { "epoch": 1.3591692789968652, "grad_norm": 2.438993453979492, "learning_rate": 7.429440246566103e-06, "loss": 1.5461, "step": 17343 }, { "epoch": 1.3592476489028213, "grad_norm": 2.809682846069336, "learning_rate": 7.429170727014459e-06, "loss": 0.9678, "step": 17344 }, { "epoch": 1.3593260188087775, "grad_norm": 2.8017923831939697, "learning_rate": 7.428901198223497e-06, "loss": 1.186, "step": 17345 }, { "epoch": 1.3594043887147336, "grad_norm": 9.162637710571289, "learning_rate": 7.428631660194242e-06, "loss": 1.5388, "step": 17346 }, { "epoch": 1.3594827586206897, "grad_norm": 3.1002795696258545, "learning_rate": 7.428362112927721e-06, "loss": 1.2334, "step": 17347 }, { "epoch": 1.3595611285266458, "grad_norm": 1.8654773235321045, "learning_rate": 7.428092556424955e-06, "loss": 1.2538, "step": 17348 }, { "epoch": 1.359639498432602, "grad_norm": 2.123992681503296, "learning_rate": 7.427822990686975e-06, "loss": 1.3125, "step": 17349 }, { "epoch": 1.359717868338558, "grad_norm": 12.196428298950195, "learning_rate": 7.427553415714801e-06, "loss": 2.5576, "step": 17350 }, { "epoch": 1.3597962382445141, "grad_norm": 3.789994239807129, "learning_rate": 7.427283831509463e-06, "loss": 1.3407, "step": 17351 }, { "epoch": 1.3598746081504702, "grad_norm": 2.987849712371826, "learning_rate": 7.427014238071982e-06, "loss": 1.6993, "step": 17352 }, { "epoch": 1.3599529780564263, "grad_norm": 2.5477380752563477, "learning_rate": 7.4267446354033865e-06, "loss": 1.2548, "step": 17353 }, { "epoch": 1.3600313479623825, "grad_norm": 8.628501892089844, "learning_rate": 7.4264750235047e-06, "loss": 1.7213, "step": 17354 }, { "epoch": 1.3601097178683386, "grad_norm": 3.304802417755127, "learning_rate": 7.42620540237695e-06, "loss": 0.9342, "step": 17355 }, { "epoch": 1.3601880877742947, "grad_norm": 2.9943621158599854, "learning_rate": 7.4259357720211615e-06, "loss": 1.5973, "step": 17356 }, { "epoch": 1.3602664576802508, "grad_norm": 2.3523690700531006, "learning_rate": 7.425666132438358e-06, "loss": 1.3604, "step": 17357 }, { "epoch": 1.360344827586207, "grad_norm": 2.683478593826294, "learning_rate": 7.425396483629567e-06, "loss": 1.3586, "step": 17358 }, { "epoch": 1.360423197492163, "grad_norm": 2.0308244228363037, "learning_rate": 7.425126825595812e-06, "loss": 1.3737, "step": 17359 }, { "epoch": 1.3605015673981191, "grad_norm": 7.402695178985596, "learning_rate": 7.4248571583381225e-06, "loss": 1.3543, "step": 17360 }, { "epoch": 1.3605799373040752, "grad_norm": 4.230079174041748, "learning_rate": 7.424587481857519e-06, "loss": 1.7001, "step": 17361 }, { "epoch": 1.3606583072100313, "grad_norm": 3.6104841232299805, "learning_rate": 7.424317796155032e-06, "loss": 1.6228, "step": 17362 }, { "epoch": 1.3607366771159874, "grad_norm": 3.2547264099121094, "learning_rate": 7.424048101231687e-06, "loss": 1.4734, "step": 17363 }, { "epoch": 1.3608150470219436, "grad_norm": 3.319819927215576, "learning_rate": 7.423778397088506e-06, "loss": 1.0818, "step": 17364 }, { "epoch": 1.3608934169278997, "grad_norm": 4.867660045623779, "learning_rate": 7.423508683726517e-06, "loss": 1.2645, "step": 17365 }, { "epoch": 1.3609717868338558, "grad_norm": 2.6430585384368896, "learning_rate": 7.423238961146745e-06, "loss": 1.3659, "step": 17366 }, { "epoch": 1.361050156739812, "grad_norm": 3.3887929916381836, "learning_rate": 7.422969229350218e-06, "loss": 1.3549, "step": 17367 }, { "epoch": 1.361128526645768, "grad_norm": 13.210249900817871, "learning_rate": 7.4226994883379585e-06, "loss": 1.1611, "step": 17368 }, { "epoch": 1.361206896551724, "grad_norm": 2.7866709232330322, "learning_rate": 7.422429738110998e-06, "loss": 1.1039, "step": 17369 }, { "epoch": 1.3612852664576802, "grad_norm": 3.2585198879241943, "learning_rate": 7.4221599786703556e-06, "loss": 1.4551, "step": 17370 }, { "epoch": 1.3613636363636363, "grad_norm": 2.8291006088256836, "learning_rate": 7.421890210017063e-06, "loss": 1.1982, "step": 17371 }, { "epoch": 1.3614420062695924, "grad_norm": 2.753951072692871, "learning_rate": 7.421620432152144e-06, "loss": 1.0927, "step": 17372 }, { "epoch": 1.3615203761755486, "grad_norm": 3.5669522285461426, "learning_rate": 7.421350645076624e-06, "loss": 1.2259, "step": 17373 }, { "epoch": 1.3615987460815047, "grad_norm": 5.343728065490723, "learning_rate": 7.4210808487915285e-06, "loss": 1.5485, "step": 17374 }, { "epoch": 1.3616771159874608, "grad_norm": 2.795386552810669, "learning_rate": 7.420811043297887e-06, "loss": 1.1594, "step": 17375 }, { "epoch": 1.3617554858934169, "grad_norm": 2.3089921474456787, "learning_rate": 7.420541228596722e-06, "loss": 1.4623, "step": 17376 }, { "epoch": 1.361833855799373, "grad_norm": 5.041741371154785, "learning_rate": 7.420271404689064e-06, "loss": 1.3058, "step": 17377 }, { "epoch": 1.361912225705329, "grad_norm": 2.991452217102051, "learning_rate": 7.420001571575934e-06, "loss": 1.572, "step": 17378 }, { "epoch": 1.3619905956112852, "grad_norm": 3.2186622619628906, "learning_rate": 7.4197317292583625e-06, "loss": 0.9532, "step": 17379 }, { "epoch": 1.3620689655172413, "grad_norm": 8.525214195251465, "learning_rate": 7.419461877737373e-06, "loss": 1.4385, "step": 17380 }, { "epoch": 1.3621473354231974, "grad_norm": 3.2568421363830566, "learning_rate": 7.419192017013995e-06, "loss": 1.5156, "step": 17381 }, { "epoch": 1.3622257053291535, "grad_norm": 2.960029363632202, "learning_rate": 7.418922147089251e-06, "loss": 1.3002, "step": 17382 }, { "epoch": 1.3623040752351097, "grad_norm": 2.323565721511841, "learning_rate": 7.418652267964173e-06, "loss": 1.278, "step": 17383 }, { "epoch": 1.3623824451410658, "grad_norm": 3.4437789916992188, "learning_rate": 7.4183823796397815e-06, "loss": 1.179, "step": 17384 }, { "epoch": 1.3624608150470219, "grad_norm": 2.5713000297546387, "learning_rate": 7.418112482117106e-06, "loss": 0.9643, "step": 17385 }, { "epoch": 1.362539184952978, "grad_norm": 5.481871604919434, "learning_rate": 7.417842575397174e-06, "loss": 1.3617, "step": 17386 }, { "epoch": 1.362617554858934, "grad_norm": 4.360400676727295, "learning_rate": 7.417572659481008e-06, "loss": 1.5883, "step": 17387 }, { "epoch": 1.3626959247648902, "grad_norm": 4.1967010498046875, "learning_rate": 7.41730273436964e-06, "loss": 1.5103, "step": 17388 }, { "epoch": 1.3627742946708463, "grad_norm": 2.3237764835357666, "learning_rate": 7.4170328000640915e-06, "loss": 1.2821, "step": 17389 }, { "epoch": 1.3628526645768024, "grad_norm": 2.5496675968170166, "learning_rate": 7.4167628565653935e-06, "loss": 1.0992, "step": 17390 }, { "epoch": 1.3629310344827585, "grad_norm": 3.417773485183716, "learning_rate": 7.41649290387457e-06, "loss": 1.4336, "step": 17391 }, { "epoch": 1.3630094043887147, "grad_norm": 2.7503557205200195, "learning_rate": 7.41622294199265e-06, "loss": 1.2459, "step": 17392 }, { "epoch": 1.3630877742946708, "grad_norm": 3.9985198974609375, "learning_rate": 7.415952970920658e-06, "loss": 1.161, "step": 17393 }, { "epoch": 1.363166144200627, "grad_norm": 2.3359525203704834, "learning_rate": 7.415682990659621e-06, "loss": 1.2392, "step": 17394 }, { "epoch": 1.363244514106583, "grad_norm": 2.150886297225952, "learning_rate": 7.4154130012105675e-06, "loss": 1.2002, "step": 17395 }, { "epoch": 1.3633228840125393, "grad_norm": 2.5035881996154785, "learning_rate": 7.415143002574524e-06, "loss": 1.531, "step": 17396 }, { "epoch": 1.3634012539184952, "grad_norm": 3.3884530067443848, "learning_rate": 7.414872994752518e-06, "loss": 0.6039, "step": 17397 }, { "epoch": 1.3634796238244515, "grad_norm": 2.1826016902923584, "learning_rate": 7.414602977745573e-06, "loss": 1.4883, "step": 17398 }, { "epoch": 1.3635579937304074, "grad_norm": 3.2078864574432373, "learning_rate": 7.41433295155472e-06, "loss": 1.3078, "step": 17399 }, { "epoch": 1.3636363636363638, "grad_norm": 2.2461700439453125, "learning_rate": 7.414062916180984e-06, "loss": 1.3826, "step": 17400 }, { "epoch": 1.3637147335423196, "grad_norm": 4.066821098327637, "learning_rate": 7.4137928716253925e-06, "loss": 0.9151, "step": 17401 }, { "epoch": 1.363793103448276, "grad_norm": 6.810795783996582, "learning_rate": 7.413522817888973e-06, "loss": 1.8591, "step": 17402 }, { "epoch": 1.3638714733542319, "grad_norm": 3.576735019683838, "learning_rate": 7.413252754972753e-06, "loss": 1.2839, "step": 17403 }, { "epoch": 1.3639498432601882, "grad_norm": 2.309746026992798, "learning_rate": 7.412982682877759e-06, "loss": 1.0437, "step": 17404 }, { "epoch": 1.364028213166144, "grad_norm": 2.112651824951172, "learning_rate": 7.4127126016050175e-06, "loss": 0.9917, "step": 17405 }, { "epoch": 1.3641065830721004, "grad_norm": 5.850866317749023, "learning_rate": 7.412442511155555e-06, "loss": 1.5933, "step": 17406 }, { "epoch": 1.3641849529780563, "grad_norm": 3.908311128616333, "learning_rate": 7.412172411530403e-06, "loss": 1.6065, "step": 17407 }, { "epoch": 1.3642633228840126, "grad_norm": 2.846334457397461, "learning_rate": 7.411902302730585e-06, "loss": 1.572, "step": 17408 }, { "epoch": 1.3643416927899685, "grad_norm": 2.544917345046997, "learning_rate": 7.411632184757129e-06, "loss": 1.4954, "step": 17409 }, { "epoch": 1.3644200626959249, "grad_norm": 2.851111888885498, "learning_rate": 7.411362057611065e-06, "loss": 1.3155, "step": 17410 }, { "epoch": 1.3644984326018808, "grad_norm": 3.2506282329559326, "learning_rate": 7.411091921293416e-06, "loss": 1.2358, "step": 17411 }, { "epoch": 1.364576802507837, "grad_norm": 3.174712657928467, "learning_rate": 7.4108217758052126e-06, "loss": 1.65, "step": 17412 }, { "epoch": 1.364655172413793, "grad_norm": 2.176945209503174, "learning_rate": 7.410551621147481e-06, "loss": 1.3927, "step": 17413 }, { "epoch": 1.3647335423197493, "grad_norm": 14.586994171142578, "learning_rate": 7.41028145732125e-06, "loss": 0.8772, "step": 17414 }, { "epoch": 1.3648119122257052, "grad_norm": 2.67810320854187, "learning_rate": 7.410011284327545e-06, "loss": 1.1096, "step": 17415 }, { "epoch": 1.3648902821316615, "grad_norm": 3.270209550857544, "learning_rate": 7.409741102167396e-06, "loss": 1.3544, "step": 17416 }, { "epoch": 1.3649686520376176, "grad_norm": 3.1654956340789795, "learning_rate": 7.409470910841829e-06, "loss": 1.5098, "step": 17417 }, { "epoch": 1.3650470219435737, "grad_norm": 3.444714307785034, "learning_rate": 7.409200710351874e-06, "loss": 1.3264, "step": 17418 }, { "epoch": 1.3651253918495299, "grad_norm": 2.5393714904785156, "learning_rate": 7.4089305006985544e-06, "loss": 1.368, "step": 17419 }, { "epoch": 1.365203761755486, "grad_norm": 2.2013754844665527, "learning_rate": 7.408660281882902e-06, "loss": 1.8273, "step": 17420 }, { "epoch": 1.365282131661442, "grad_norm": 3.283548593521118, "learning_rate": 7.408390053905943e-06, "loss": 1.3241, "step": 17421 }, { "epoch": 1.3653605015673982, "grad_norm": 4.386909008026123, "learning_rate": 7.408119816768704e-06, "loss": 1.2031, "step": 17422 }, { "epoch": 1.3654388714733543, "grad_norm": 3.541766405105591, "learning_rate": 7.407849570472215e-06, "loss": 1.5775, "step": 17423 }, { "epoch": 1.3655172413793104, "grad_norm": 2.6033668518066406, "learning_rate": 7.407579315017503e-06, "loss": 1.59, "step": 17424 }, { "epoch": 1.3655956112852665, "grad_norm": 4.3386549949646, "learning_rate": 7.4073090504055964e-06, "loss": 1.9092, "step": 17425 }, { "epoch": 1.3656739811912226, "grad_norm": 2.0728986263275146, "learning_rate": 7.4070387766375215e-06, "loss": 1.5121, "step": 17426 }, { "epoch": 1.3657523510971787, "grad_norm": 2.482409954071045, "learning_rate": 7.40676849371431e-06, "loss": 1.4662, "step": 17427 }, { "epoch": 1.3658307210031349, "grad_norm": 2.2112934589385986, "learning_rate": 7.406498201636984e-06, "loss": 1.4383, "step": 17428 }, { "epoch": 1.365909090909091, "grad_norm": 3.1053555011749268, "learning_rate": 7.406227900406576e-06, "loss": 1.432, "step": 17429 }, { "epoch": 1.365987460815047, "grad_norm": 2.908205270767212, "learning_rate": 7.405957590024113e-06, "loss": 1.7497, "step": 17430 }, { "epoch": 1.3660658307210032, "grad_norm": 2.9670538902282715, "learning_rate": 7.405687270490623e-06, "loss": 1.3606, "step": 17431 }, { "epoch": 1.3661442006269593, "grad_norm": 2.948859691619873, "learning_rate": 7.405416941807135e-06, "loss": 1.7861, "step": 17432 }, { "epoch": 1.3662225705329154, "grad_norm": 3.326666831970215, "learning_rate": 7.405146603974677e-06, "loss": 1.5781, "step": 17433 }, { "epoch": 1.3663009404388715, "grad_norm": 3.49280047416687, "learning_rate": 7.404876256994275e-06, "loss": 1.5191, "step": 17434 }, { "epoch": 1.3663793103448276, "grad_norm": 3.537517547607422, "learning_rate": 7.40460590086696e-06, "loss": 0.7915, "step": 17435 }, { "epoch": 1.3664576802507837, "grad_norm": 2.1851327419281006, "learning_rate": 7.4043355355937585e-06, "loss": 1.3362, "step": 17436 }, { "epoch": 1.3665360501567398, "grad_norm": 2.9139416217803955, "learning_rate": 7.404065161175701e-06, "loss": 1.4134, "step": 17437 }, { "epoch": 1.366614420062696, "grad_norm": 2.753351926803589, "learning_rate": 7.403794777613813e-06, "loss": 1.8256, "step": 17438 }, { "epoch": 1.366692789968652, "grad_norm": 3.049565076828003, "learning_rate": 7.4035243849091246e-06, "loss": 1.2149, "step": 17439 }, { "epoch": 1.3667711598746082, "grad_norm": 2.9294517040252686, "learning_rate": 7.403253983062665e-06, "loss": 1.1911, "step": 17440 }, { "epoch": 1.3668495297805643, "grad_norm": 2.586395502090454, "learning_rate": 7.40298357207546e-06, "loss": 1.417, "step": 17441 }, { "epoch": 1.3669278996865204, "grad_norm": 3.263075351715088, "learning_rate": 7.402713151948542e-06, "loss": 0.8184, "step": 17442 }, { "epoch": 1.3670062695924765, "grad_norm": 2.099687337875366, "learning_rate": 7.402442722682934e-06, "loss": 1.359, "step": 17443 }, { "epoch": 1.3670846394984326, "grad_norm": 2.304063320159912, "learning_rate": 7.402172284279671e-06, "loss": 1.3984, "step": 17444 }, { "epoch": 1.3671630094043887, "grad_norm": 3.149207353591919, "learning_rate": 7.401901836739776e-06, "loss": 1.3482, "step": 17445 }, { "epoch": 1.3672413793103448, "grad_norm": 2.476231575012207, "learning_rate": 7.401631380064282e-06, "loss": 1.3125, "step": 17446 }, { "epoch": 1.367319749216301, "grad_norm": 3.533691883087158, "learning_rate": 7.401360914254214e-06, "loss": 1.1678, "step": 17447 }, { "epoch": 1.367398119122257, "grad_norm": 3.320646047592163, "learning_rate": 7.401090439310604e-06, "loss": 1.4336, "step": 17448 }, { "epoch": 1.3674764890282132, "grad_norm": 3.0801479816436768, "learning_rate": 7.400819955234478e-06, "loss": 1.2702, "step": 17449 }, { "epoch": 1.3675548589341693, "grad_norm": 2.546611785888672, "learning_rate": 7.400549462026867e-06, "loss": 1.394, "step": 17450 }, { "epoch": 1.3676332288401254, "grad_norm": 2.5610506534576416, "learning_rate": 7.400278959688799e-06, "loss": 1.4127, "step": 17451 }, { "epoch": 1.3677115987460815, "grad_norm": 2.7964694499969482, "learning_rate": 7.400008448221303e-06, "loss": 1.6512, "step": 17452 }, { "epoch": 1.3677899686520376, "grad_norm": 2.2496228218078613, "learning_rate": 7.399737927625408e-06, "loss": 1.0364, "step": 17453 }, { "epoch": 1.3678683385579937, "grad_norm": 4.008937358856201, "learning_rate": 7.39946739790214e-06, "loss": 1.9212, "step": 17454 }, { "epoch": 1.3679467084639498, "grad_norm": 5.920269012451172, "learning_rate": 7.399196859052532e-06, "loss": 1.4092, "step": 17455 }, { "epoch": 1.368025078369906, "grad_norm": 2.6202361583709717, "learning_rate": 7.398926311077612e-06, "loss": 1.3078, "step": 17456 }, { "epoch": 1.368103448275862, "grad_norm": 2.442067861557007, "learning_rate": 7.398655753978408e-06, "loss": 1.4896, "step": 17457 }, { "epoch": 1.3681818181818182, "grad_norm": 6.225067615509033, "learning_rate": 7.398385187755949e-06, "loss": 2.0983, "step": 17458 }, { "epoch": 1.3682601880877743, "grad_norm": 16.442909240722656, "learning_rate": 7.3981146124112654e-06, "loss": 2.5899, "step": 17459 }, { "epoch": 1.3683385579937304, "grad_norm": 2.9053773880004883, "learning_rate": 7.397844027945385e-06, "loss": 1.2328, "step": 17460 }, { "epoch": 1.3684169278996865, "grad_norm": 3.5427911281585693, "learning_rate": 7.397573434359338e-06, "loss": 1.4828, "step": 17461 }, { "epoch": 1.3684952978056426, "grad_norm": 3.941983222961426, "learning_rate": 7.397302831654152e-06, "loss": 1.6942, "step": 17462 }, { "epoch": 1.3685736677115987, "grad_norm": 7.050436973571777, "learning_rate": 7.397032219830859e-06, "loss": 1.2295, "step": 17463 }, { "epoch": 1.3686520376175548, "grad_norm": 7.5570573806762695, "learning_rate": 7.396761598890485e-06, "loss": 1.3501, "step": 17464 }, { "epoch": 1.368730407523511, "grad_norm": 3.0527734756469727, "learning_rate": 7.396490968834063e-06, "loss": 1.822, "step": 17465 }, { "epoch": 1.368808777429467, "grad_norm": 3.14510178565979, "learning_rate": 7.396220329662618e-06, "loss": 1.3402, "step": 17466 }, { "epoch": 1.3688871473354232, "grad_norm": 3.5824856758117676, "learning_rate": 7.395949681377183e-06, "loss": 1.5009, "step": 17467 }, { "epoch": 1.3689655172413793, "grad_norm": 6.0366716384887695, "learning_rate": 7.3956790239787855e-06, "loss": 1.9062, "step": 17468 }, { "epoch": 1.3690438871473354, "grad_norm": 2.1653993129730225, "learning_rate": 7.395408357468455e-06, "loss": 1.3298, "step": 17469 }, { "epoch": 1.3691222570532915, "grad_norm": 3.2838351726531982, "learning_rate": 7.395137681847223e-06, "loss": 1.1732, "step": 17470 }, { "epoch": 1.3692006269592476, "grad_norm": 2.2691292762756348, "learning_rate": 7.394866997116117e-06, "loss": 1.31, "step": 17471 }, { "epoch": 1.3692789968652037, "grad_norm": 3.548936605453491, "learning_rate": 7.394596303276167e-06, "loss": 1.1491, "step": 17472 }, { "epoch": 1.3693573667711598, "grad_norm": 2.077491283416748, "learning_rate": 7.394325600328402e-06, "loss": 0.9626, "step": 17473 }, { "epoch": 1.369435736677116, "grad_norm": 2.208711862564087, "learning_rate": 7.394054888273852e-06, "loss": 1.4831, "step": 17474 }, { "epoch": 1.369514106583072, "grad_norm": 2.583951711654663, "learning_rate": 7.3937841671135466e-06, "loss": 1.1555, "step": 17475 }, { "epoch": 1.3695924764890282, "grad_norm": 6.210395812988281, "learning_rate": 7.3935134368485174e-06, "loss": 1.217, "step": 17476 }, { "epoch": 1.3696708463949843, "grad_norm": 3.340000629425049, "learning_rate": 7.393242697479791e-06, "loss": 1.5896, "step": 17477 }, { "epoch": 1.3697492163009404, "grad_norm": 4.911624908447266, "learning_rate": 7.392971949008399e-06, "loss": 1.0667, "step": 17478 }, { "epoch": 1.3698275862068965, "grad_norm": 2.5275039672851562, "learning_rate": 7.392701191435371e-06, "loss": 1.6084, "step": 17479 }, { "epoch": 1.3699059561128526, "grad_norm": 3.479464292526245, "learning_rate": 7.3924304247617376e-06, "loss": 1.3796, "step": 17480 }, { "epoch": 1.3699843260188087, "grad_norm": 2.97762131690979, "learning_rate": 7.392159648988526e-06, "loss": 1.3298, "step": 17481 }, { "epoch": 1.3700626959247648, "grad_norm": 11.270793914794922, "learning_rate": 7.39188886411677e-06, "loss": 0.6862, "step": 17482 }, { "epoch": 1.370141065830721, "grad_norm": 5.901028156280518, "learning_rate": 7.391618070147495e-06, "loss": 1.0479, "step": 17483 }, { "epoch": 1.370219435736677, "grad_norm": 2.121192216873169, "learning_rate": 7.391347267081734e-06, "loss": 1.1881, "step": 17484 }, { "epoch": 1.3702978056426331, "grad_norm": 2.520258903503418, "learning_rate": 7.391076454920516e-06, "loss": 1.4614, "step": 17485 }, { "epoch": 1.3703761755485893, "grad_norm": 3.4733145236968994, "learning_rate": 7.390805633664873e-06, "loss": 1.2841, "step": 17486 }, { "epoch": 1.3704545454545456, "grad_norm": 3.637885570526123, "learning_rate": 7.390534803315832e-06, "loss": 1.6943, "step": 17487 }, { "epoch": 1.3705329153605015, "grad_norm": 2.878906726837158, "learning_rate": 7.3902639638744235e-06, "loss": 0.9791, "step": 17488 }, { "epoch": 1.3706112852664578, "grad_norm": 3.072946786880493, "learning_rate": 7.3899931153416805e-06, "loss": 1.2876, "step": 17489 }, { "epoch": 1.3706896551724137, "grad_norm": 2.1177921295166016, "learning_rate": 7.389722257718629e-06, "loss": 1.5287, "step": 17490 }, { "epoch": 1.37076802507837, "grad_norm": 2.6848304271698, "learning_rate": 7.389451391006303e-06, "loss": 1.449, "step": 17491 }, { "epoch": 1.370846394984326, "grad_norm": 21.8765869140625, "learning_rate": 7.389180515205732e-06, "loss": 2.0048, "step": 17492 }, { "epoch": 1.3709247648902823, "grad_norm": 5.3467864990234375, "learning_rate": 7.388909630317944e-06, "loss": 1.4293, "step": 17493 }, { "epoch": 1.3710031347962381, "grad_norm": 2.480348587036133, "learning_rate": 7.388638736343973e-06, "loss": 1.6993, "step": 17494 }, { "epoch": 1.3710815047021945, "grad_norm": 3.0474352836608887, "learning_rate": 7.3883678332848444e-06, "loss": 1.8373, "step": 17495 }, { "epoch": 1.3711598746081504, "grad_norm": 2.8135459423065186, "learning_rate": 7.388096921141593e-06, "loss": 1.5519, "step": 17496 }, { "epoch": 1.3712382445141067, "grad_norm": 2.472655773162842, "learning_rate": 7.387825999915246e-06, "loss": 1.3305, "step": 17497 }, { "epoch": 1.3713166144200626, "grad_norm": 2.922520637512207, "learning_rate": 7.387555069606836e-06, "loss": 0.7901, "step": 17498 }, { "epoch": 1.371394984326019, "grad_norm": 3.0729148387908936, "learning_rate": 7.387284130217394e-06, "loss": 1.3637, "step": 17499 }, { "epoch": 1.3714733542319748, "grad_norm": 3.257453680038452, "learning_rate": 7.387013181747949e-06, "loss": 0.3214, "step": 17500 }, { "epoch": 1.3715517241379311, "grad_norm": 2.556894540786743, "learning_rate": 7.38674222419953e-06, "loss": 1.5704, "step": 17501 }, { "epoch": 1.371630094043887, "grad_norm": 2.2507309913635254, "learning_rate": 7.386471257573171e-06, "loss": 1.523, "step": 17502 }, { "epoch": 1.3717084639498434, "grad_norm": 4.1697998046875, "learning_rate": 7.3862002818699005e-06, "loss": 1.673, "step": 17503 }, { "epoch": 1.3717868338557992, "grad_norm": 3.3894660472869873, "learning_rate": 7.38592929709075e-06, "loss": 1.4302, "step": 17504 }, { "epoch": 1.3718652037617556, "grad_norm": 2.3406455516815186, "learning_rate": 7.385658303236751e-06, "loss": 1.237, "step": 17505 }, { "epoch": 1.3719435736677115, "grad_norm": 2.92730975151062, "learning_rate": 7.385387300308933e-06, "loss": 1.4433, "step": 17506 }, { "epoch": 1.3720219435736678, "grad_norm": 3.2440154552459717, "learning_rate": 7.385116288308325e-06, "loss": 1.5663, "step": 17507 }, { "epoch": 1.3721003134796237, "grad_norm": 2.2905497550964355, "learning_rate": 7.384845267235961e-06, "loss": 1.4239, "step": 17508 }, { "epoch": 1.37217868338558, "grad_norm": 4.186412811279297, "learning_rate": 7.384574237092872e-06, "loss": 1.2979, "step": 17509 }, { "epoch": 1.3722570532915361, "grad_norm": 4.328677177429199, "learning_rate": 7.384303197880085e-06, "loss": 1.6542, "step": 17510 }, { "epoch": 1.3723354231974922, "grad_norm": 2.556178092956543, "learning_rate": 7.384032149598634e-06, "loss": 1.177, "step": 17511 }, { "epoch": 1.3724137931034484, "grad_norm": 4.191436290740967, "learning_rate": 7.38376109224955e-06, "loss": 1.2864, "step": 17512 }, { "epoch": 1.3724921630094045, "grad_norm": 2.8075249195098877, "learning_rate": 7.383490025833863e-06, "loss": 1.2722, "step": 17513 }, { "epoch": 1.3725705329153606, "grad_norm": 4.38962459564209, "learning_rate": 7.383218950352603e-06, "loss": 1.7963, "step": 17514 }, { "epoch": 1.3726489028213167, "grad_norm": 3.0127573013305664, "learning_rate": 7.3829478658068035e-06, "loss": 1.4015, "step": 17515 }, { "epoch": 1.3727272727272728, "grad_norm": 4.832974433898926, "learning_rate": 7.382676772197492e-06, "loss": 1.5288, "step": 17516 }, { "epoch": 1.372805642633229, "grad_norm": 2.157933473587036, "learning_rate": 7.382405669525705e-06, "loss": 1.0933, "step": 17517 }, { "epoch": 1.372884012539185, "grad_norm": 2.8896312713623047, "learning_rate": 7.3821345577924695e-06, "loss": 1.214, "step": 17518 }, { "epoch": 1.3729623824451411, "grad_norm": 2.300891399383545, "learning_rate": 7.381863436998817e-06, "loss": 1.3956, "step": 17519 }, { "epoch": 1.3730407523510972, "grad_norm": 2.3106606006622314, "learning_rate": 7.381592307145779e-06, "loss": 1.4967, "step": 17520 }, { "epoch": 1.3731191222570533, "grad_norm": 11.335358619689941, "learning_rate": 7.381321168234389e-06, "loss": 2.5751, "step": 17521 }, { "epoch": 1.3731974921630095, "grad_norm": 3.817732334136963, "learning_rate": 7.381050020265675e-06, "loss": 1.3544, "step": 17522 }, { "epoch": 1.3732758620689656, "grad_norm": 2.67960786819458, "learning_rate": 7.38077886324067e-06, "loss": 1.8005, "step": 17523 }, { "epoch": 1.3733542319749217, "grad_norm": 3.8027760982513428, "learning_rate": 7.380507697160405e-06, "loss": 1.0426, "step": 17524 }, { "epoch": 1.3734326018808778, "grad_norm": 3.837005376815796, "learning_rate": 7.380236522025911e-06, "loss": 0.9461, "step": 17525 }, { "epoch": 1.373510971786834, "grad_norm": 4.209864616394043, "learning_rate": 7.379965337838221e-06, "loss": 1.7558, "step": 17526 }, { "epoch": 1.37358934169279, "grad_norm": 2.9220707416534424, "learning_rate": 7.379694144598364e-06, "loss": 1.4459, "step": 17527 }, { "epoch": 1.3736677115987461, "grad_norm": 2.6545023918151855, "learning_rate": 7.3794229423073735e-06, "loss": 1.3671, "step": 17528 }, { "epoch": 1.3737460815047022, "grad_norm": 5.057577610015869, "learning_rate": 7.3791517309662795e-06, "loss": 1.6189, "step": 17529 }, { "epoch": 1.3738244514106583, "grad_norm": 12.327705383300781, "learning_rate": 7.378880510576115e-06, "loss": 1.4695, "step": 17530 }, { "epoch": 1.3739028213166145, "grad_norm": 4.612191677093506, "learning_rate": 7.378609281137911e-06, "loss": 1.2993, "step": 17531 }, { "epoch": 1.3739811912225706, "grad_norm": 2.5929107666015625, "learning_rate": 7.378338042652699e-06, "loss": 1.6109, "step": 17532 }, { "epoch": 1.3740595611285267, "grad_norm": 2.6397323608398438, "learning_rate": 7.37806679512151e-06, "loss": 1.0939, "step": 17533 }, { "epoch": 1.3741379310344828, "grad_norm": 3.8130693435668945, "learning_rate": 7.377795538545378e-06, "loss": 1.1433, "step": 17534 }, { "epoch": 1.374216300940439, "grad_norm": 2.6459405422210693, "learning_rate": 7.37752427292533e-06, "loss": 1.1469, "step": 17535 }, { "epoch": 1.374294670846395, "grad_norm": 2.739133596420288, "learning_rate": 7.3772529982624045e-06, "loss": 1.375, "step": 17536 }, { "epoch": 1.3743730407523511, "grad_norm": 5.962523937225342, "learning_rate": 7.376981714557627e-06, "loss": 0.8069, "step": 17537 }, { "epoch": 1.3744514106583072, "grad_norm": 6.5211405754089355, "learning_rate": 7.376710421812033e-06, "loss": 1.4831, "step": 17538 }, { "epoch": 1.3745297805642633, "grad_norm": 5.341365337371826, "learning_rate": 7.376439120026654e-06, "loss": 1.0794, "step": 17539 }, { "epoch": 1.3746081504702194, "grad_norm": 2.309622049331665, "learning_rate": 7.37616780920252e-06, "loss": 1.1298, "step": 17540 }, { "epoch": 1.3746865203761756, "grad_norm": 3.79226016998291, "learning_rate": 7.375896489340666e-06, "loss": 1.3415, "step": 17541 }, { "epoch": 1.3747648902821317, "grad_norm": 2.364818811416626, "learning_rate": 7.375625160442119e-06, "loss": 1.65, "step": 17542 }, { "epoch": 1.3748432601880878, "grad_norm": 3.777318239212036, "learning_rate": 7.375353822507917e-06, "loss": 1.2761, "step": 17543 }, { "epoch": 1.374921630094044, "grad_norm": 2.7477633953094482, "learning_rate": 7.375082475539086e-06, "loss": 1.0828, "step": 17544 }, { "epoch": 1.375, "grad_norm": 3.6748132705688477, "learning_rate": 7.374811119536664e-06, "loss": 1.4101, "step": 17545 }, { "epoch": 1.375078369905956, "grad_norm": 2.9440977573394775, "learning_rate": 7.374539754501679e-06, "loss": 1.122, "step": 17546 }, { "epoch": 1.3751567398119122, "grad_norm": 2.5145041942596436, "learning_rate": 7.374268380435166e-06, "loss": 1.2066, "step": 17547 }, { "epoch": 1.3752351097178683, "grad_norm": 2.736724615097046, "learning_rate": 7.373996997338153e-06, "loss": 1.3228, "step": 17548 }, { "epoch": 1.3753134796238244, "grad_norm": 3.1106069087982178, "learning_rate": 7.373725605211677e-06, "loss": 1.4317, "step": 17549 }, { "epoch": 1.3753918495297806, "grad_norm": 10.892958641052246, "learning_rate": 7.3734542040567666e-06, "loss": 1.135, "step": 17550 }, { "epoch": 1.3754702194357367, "grad_norm": 4.574949264526367, "learning_rate": 7.373182793874456e-06, "loss": 1.1129, "step": 17551 }, { "epoch": 1.3755485893416928, "grad_norm": 8.599154472351074, "learning_rate": 7.372911374665777e-06, "loss": 0.9178, "step": 17552 }, { "epoch": 1.3756269592476489, "grad_norm": 2.3735947608947754, "learning_rate": 7.3726399464317625e-06, "loss": 1.8489, "step": 17553 }, { "epoch": 1.375705329153605, "grad_norm": 4.423234462738037, "learning_rate": 7.3723685091734435e-06, "loss": 1.777, "step": 17554 }, { "epoch": 1.375783699059561, "grad_norm": 2.5899195671081543, "learning_rate": 7.3720970628918535e-06, "loss": 1.4226, "step": 17555 }, { "epoch": 1.3758620689655172, "grad_norm": 2.5644357204437256, "learning_rate": 7.371825607588024e-06, "loss": 1.4617, "step": 17556 }, { "epoch": 1.3759404388714733, "grad_norm": 5.762386322021484, "learning_rate": 7.3715541432629885e-06, "loss": 1.173, "step": 17557 }, { "epoch": 1.3760188087774294, "grad_norm": 5.733558654785156, "learning_rate": 7.3712826699177794e-06, "loss": 1.8425, "step": 17558 }, { "epoch": 1.3760971786833855, "grad_norm": 5.298183917999268, "learning_rate": 7.3710111875534296e-06, "loss": 1.2583, "step": 17559 }, { "epoch": 1.3761755485893417, "grad_norm": 3.509277105331421, "learning_rate": 7.370739696170971e-06, "loss": 1.23, "step": 17560 }, { "epoch": 1.3762539184952978, "grad_norm": 2.8028604984283447, "learning_rate": 7.370468195771434e-06, "loss": 1.2903, "step": 17561 }, { "epoch": 1.3763322884012539, "grad_norm": 2.7220001220703125, "learning_rate": 7.370196686355857e-06, "loss": 1.6179, "step": 17562 }, { "epoch": 1.37641065830721, "grad_norm": 3.3309125900268555, "learning_rate": 7.369925167925266e-06, "loss": 1.2519, "step": 17563 }, { "epoch": 1.376489028213166, "grad_norm": 2.6371066570281982, "learning_rate": 7.369653640480699e-06, "loss": 1.5321, "step": 17564 }, { "epoch": 1.3765673981191222, "grad_norm": 6.7793965339660645, "learning_rate": 7.369382104023186e-06, "loss": 1.5489, "step": 17565 }, { "epoch": 1.3766457680250783, "grad_norm": 2.424034595489502, "learning_rate": 7.36911055855376e-06, "loss": 1.2045, "step": 17566 }, { "epoch": 1.3767241379310344, "grad_norm": 3.2103054523468018, "learning_rate": 7.368839004073456e-06, "loss": 1.4149, "step": 17567 }, { "epoch": 1.3768025078369905, "grad_norm": 4.511981010437012, "learning_rate": 7.368567440583303e-06, "loss": 1.1574, "step": 17568 }, { "epoch": 1.3768808777429467, "grad_norm": 2.413179874420166, "learning_rate": 7.368295868084338e-06, "loss": 1.3077, "step": 17569 }, { "epoch": 1.3769592476489028, "grad_norm": 4.649902820587158, "learning_rate": 7.368024286577591e-06, "loss": 1.1956, "step": 17570 }, { "epoch": 1.3770376175548589, "grad_norm": 3.035205602645874, "learning_rate": 7.367752696064096e-06, "loss": 1.3392, "step": 17571 }, { "epoch": 1.377115987460815, "grad_norm": 11.121600151062012, "learning_rate": 7.367481096544886e-06, "loss": 1.2293, "step": 17572 }, { "epoch": 1.377194357366771, "grad_norm": 3.183112382888794, "learning_rate": 7.367209488020993e-06, "loss": 1.4285, "step": 17573 }, { "epoch": 1.3772727272727272, "grad_norm": 2.885925531387329, "learning_rate": 7.366937870493451e-06, "loss": 1.0179, "step": 17574 }, { "epoch": 1.3773510971786833, "grad_norm": 2.7486138343811035, "learning_rate": 7.366666243963294e-06, "loss": 1.5837, "step": 17575 }, { "epoch": 1.3774294670846394, "grad_norm": 2.6403372287750244, "learning_rate": 7.366394608431555e-06, "loss": 1.4008, "step": 17576 }, { "epoch": 1.3775078369905955, "grad_norm": 3.874567747116089, "learning_rate": 7.3661229638992644e-06, "loss": 1.6101, "step": 17577 }, { "epoch": 1.3775862068965516, "grad_norm": 11.371635437011719, "learning_rate": 7.36585131036746e-06, "loss": 1.6241, "step": 17578 }, { "epoch": 1.3776645768025078, "grad_norm": 4.961742877960205, "learning_rate": 7.365579647837169e-06, "loss": 1.4351, "step": 17579 }, { "epoch": 1.3777429467084639, "grad_norm": 3.0132358074188232, "learning_rate": 7.36530797630943e-06, "loss": 1.4217, "step": 17580 }, { "epoch": 1.37782131661442, "grad_norm": 4.007899761199951, "learning_rate": 7.365036295785275e-06, "loss": 1.6998, "step": 17581 }, { "epoch": 1.3778996865203763, "grad_norm": 3.3049447536468506, "learning_rate": 7.364764606265735e-06, "loss": 1.3762, "step": 17582 }, { "epoch": 1.3779780564263322, "grad_norm": 2.758876085281372, "learning_rate": 7.364492907751845e-06, "loss": 1.4184, "step": 17583 }, { "epoch": 1.3780564263322885, "grad_norm": 2.1351442337036133, "learning_rate": 7.3642212002446405e-06, "loss": 1.2034, "step": 17584 }, { "epoch": 1.3781347962382444, "grad_norm": 3.0180654525756836, "learning_rate": 7.363949483745152e-06, "loss": 1.4896, "step": 17585 }, { "epoch": 1.3782131661442008, "grad_norm": 2.6564977169036865, "learning_rate": 7.363677758254413e-06, "loss": 0.9012, "step": 17586 }, { "epoch": 1.3782915360501566, "grad_norm": 3.1275103092193604, "learning_rate": 7.363406023773458e-06, "loss": 1.4125, "step": 17587 }, { "epoch": 1.378369905956113, "grad_norm": 2.6517744064331055, "learning_rate": 7.363134280303322e-06, "loss": 1.3921, "step": 17588 }, { "epoch": 1.3784482758620689, "grad_norm": 2.3503713607788086, "learning_rate": 7.362862527845035e-06, "loss": 1.2595, "step": 17589 }, { "epoch": 1.3785266457680252, "grad_norm": 2.8767004013061523, "learning_rate": 7.362590766399635e-06, "loss": 1.4208, "step": 17590 }, { "epoch": 1.378605015673981, "grad_norm": 2.3238751888275146, "learning_rate": 7.3623189959681495e-06, "loss": 1.3043, "step": 17591 }, { "epoch": 1.3786833855799374, "grad_norm": 3.3346149921417236, "learning_rate": 7.3620472165516185e-06, "loss": 1.3796, "step": 17592 }, { "epoch": 1.3787617554858933, "grad_norm": 3.8247592449188232, "learning_rate": 7.361775428151073e-06, "loss": 1.5818, "step": 17593 }, { "epoch": 1.3788401253918496, "grad_norm": 5.503970146179199, "learning_rate": 7.361503630767547e-06, "loss": 2.5806, "step": 17594 }, { "epoch": 1.3789184952978055, "grad_norm": 2.4780495166778564, "learning_rate": 7.361231824402074e-06, "loss": 0.6923, "step": 17595 }, { "epoch": 1.3789968652037619, "grad_norm": 2.148242712020874, "learning_rate": 7.360960009055687e-06, "loss": 1.5583, "step": 17596 }, { "epoch": 1.3790752351097177, "grad_norm": 3.2407665252685547, "learning_rate": 7.3606881847294224e-06, "loss": 1.2519, "step": 17597 }, { "epoch": 1.379153605015674, "grad_norm": 3.535630464553833, "learning_rate": 7.36041635142431e-06, "loss": 1.5718, "step": 17598 }, { "epoch": 1.37923197492163, "grad_norm": 2.880763292312622, "learning_rate": 7.360144509141388e-06, "loss": 1.2353, "step": 17599 }, { "epoch": 1.3793103448275863, "grad_norm": 8.31786823272705, "learning_rate": 7.359872657881688e-06, "loss": 1.4914, "step": 17600 }, { "epoch": 1.3793887147335422, "grad_norm": 4.100274085998535, "learning_rate": 7.359600797646245e-06, "loss": 1.4255, "step": 17601 }, { "epoch": 1.3794670846394985, "grad_norm": 10.244105339050293, "learning_rate": 7.359328928436092e-06, "loss": 1.7719, "step": 17602 }, { "epoch": 1.3795454545454544, "grad_norm": 2.7810170650482178, "learning_rate": 7.3590570502522644e-06, "loss": 1.36, "step": 17603 }, { "epoch": 1.3796238244514107, "grad_norm": 2.633274793624878, "learning_rate": 7.358785163095794e-06, "loss": 1.3138, "step": 17604 }, { "epoch": 1.3797021943573669, "grad_norm": 3.16011118888855, "learning_rate": 7.358513266967717e-06, "loss": 1.5762, "step": 17605 }, { "epoch": 1.379780564263323, "grad_norm": 2.512176036834717, "learning_rate": 7.358241361869068e-06, "loss": 1.325, "step": 17606 }, { "epoch": 1.379858934169279, "grad_norm": 4.486955165863037, "learning_rate": 7.35796944780088e-06, "loss": 1.2881, "step": 17607 }, { "epoch": 1.3799373040752352, "grad_norm": 5.348795413970947, "learning_rate": 7.357697524764187e-06, "loss": 1.3555, "step": 17608 }, { "epoch": 1.3800156739811913, "grad_norm": 2.9322359561920166, "learning_rate": 7.357425592760022e-06, "loss": 1.0549, "step": 17609 }, { "epoch": 1.3800940438871474, "grad_norm": 3.268585681915283, "learning_rate": 7.357153651789423e-06, "loss": 1.6947, "step": 17610 }, { "epoch": 1.3801724137931035, "grad_norm": 2.694117784500122, "learning_rate": 7.3568817018534215e-06, "loss": 1.1998, "step": 17611 }, { "epoch": 1.3802507836990596, "grad_norm": 3.7223708629608154, "learning_rate": 7.356609742953052e-06, "loss": 1.3824, "step": 17612 }, { "epoch": 1.3803291536050157, "grad_norm": 2.667914867401123, "learning_rate": 7.35633777508935e-06, "loss": 1.3844, "step": 17613 }, { "epoch": 1.3804075235109718, "grad_norm": 2.0805885791778564, "learning_rate": 7.35606579826335e-06, "loss": 1.5045, "step": 17614 }, { "epoch": 1.380485893416928, "grad_norm": 1.8320324420928955, "learning_rate": 7.355793812476084e-06, "loss": 1.5355, "step": 17615 }, { "epoch": 1.380564263322884, "grad_norm": 3.9715564250946045, "learning_rate": 7.355521817728591e-06, "loss": 1.323, "step": 17616 }, { "epoch": 1.3806426332288402, "grad_norm": 3.0741939544677734, "learning_rate": 7.3552498140219e-06, "loss": 1.6856, "step": 17617 }, { "epoch": 1.3807210031347963, "grad_norm": 3.9142894744873047, "learning_rate": 7.354977801357051e-06, "loss": 1.7139, "step": 17618 }, { "epoch": 1.3807993730407524, "grad_norm": 2.987393379211426, "learning_rate": 7.354705779735074e-06, "loss": 1.4533, "step": 17619 }, { "epoch": 1.3808777429467085, "grad_norm": 2.3376293182373047, "learning_rate": 7.3544337491570075e-06, "loss": 1.379, "step": 17620 }, { "epoch": 1.3809561128526646, "grad_norm": 4.734549045562744, "learning_rate": 7.354161709623883e-06, "loss": 1.425, "step": 17621 }, { "epoch": 1.3810344827586207, "grad_norm": 2.9328479766845703, "learning_rate": 7.353889661136737e-06, "loss": 1.2481, "step": 17622 }, { "epoch": 1.3811128526645768, "grad_norm": 2.142475128173828, "learning_rate": 7.353617603696605e-06, "loss": 1.4866, "step": 17623 }, { "epoch": 1.381191222570533, "grad_norm": 4.638413906097412, "learning_rate": 7.353345537304517e-06, "loss": 0.8536, "step": 17624 }, { "epoch": 1.381269592476489, "grad_norm": 2.5752289295196533, "learning_rate": 7.3530734619615155e-06, "loss": 1.289, "step": 17625 }, { "epoch": 1.3813479623824452, "grad_norm": 2.145441770553589, "learning_rate": 7.352801377668628e-06, "loss": 1.1828, "step": 17626 }, { "epoch": 1.3814263322884013, "grad_norm": 2.424623489379883, "learning_rate": 7.352529284426895e-06, "loss": 1.2481, "step": 17627 }, { "epoch": 1.3815047021943574, "grad_norm": 4.642831802368164, "learning_rate": 7.352257182237347e-06, "loss": 1.3575, "step": 17628 }, { "epoch": 1.3815830721003135, "grad_norm": 1.9387726783752441, "learning_rate": 7.351985071101021e-06, "loss": 1.2038, "step": 17629 }, { "epoch": 1.3816614420062696, "grad_norm": 2.9676389694213867, "learning_rate": 7.351712951018952e-06, "loss": 1.0516, "step": 17630 }, { "epoch": 1.3817398119122257, "grad_norm": 2.808492660522461, "learning_rate": 7.351440821992175e-06, "loss": 1.2903, "step": 17631 }, { "epoch": 1.3818181818181818, "grad_norm": 2.0726964473724365, "learning_rate": 7.351168684021726e-06, "loss": 1.549, "step": 17632 }, { "epoch": 1.381896551724138, "grad_norm": 4.506444931030273, "learning_rate": 7.350896537108639e-06, "loss": 1.337, "step": 17633 }, { "epoch": 1.381974921630094, "grad_norm": 3.5682899951934814, "learning_rate": 7.350624381253948e-06, "loss": 0.9798, "step": 17634 }, { "epoch": 1.3820532915360502, "grad_norm": 2.1394965648651123, "learning_rate": 7.350352216458689e-06, "loss": 1.2836, "step": 17635 }, { "epoch": 1.3821316614420063, "grad_norm": 2.919863224029541, "learning_rate": 7.350080042723898e-06, "loss": 1.1931, "step": 17636 }, { "epoch": 1.3822100313479624, "grad_norm": 3.6584885120391846, "learning_rate": 7.34980786005061e-06, "loss": 1.1459, "step": 17637 }, { "epoch": 1.3822884012539185, "grad_norm": 3.352916955947876, "learning_rate": 7.349535668439858e-06, "loss": 1.7258, "step": 17638 }, { "epoch": 1.3823667711598746, "grad_norm": 4.822422981262207, "learning_rate": 7.34926346789268e-06, "loss": 1.1015, "step": 17639 }, { "epoch": 1.3824451410658307, "grad_norm": 2.8586151599884033, "learning_rate": 7.348991258410111e-06, "loss": 1.3747, "step": 17640 }, { "epoch": 1.3825235109717868, "grad_norm": 2.2317352294921875, "learning_rate": 7.3487190399931865e-06, "loss": 1.5337, "step": 17641 }, { "epoch": 1.382601880877743, "grad_norm": 1.9079008102416992, "learning_rate": 7.34844681264294e-06, "loss": 1.2143, "step": 17642 }, { "epoch": 1.382680250783699, "grad_norm": 2.453853130340576, "learning_rate": 7.348174576360407e-06, "loss": 1.4589, "step": 17643 }, { "epoch": 1.3827586206896552, "grad_norm": 2.8170595169067383, "learning_rate": 7.347902331146627e-06, "loss": 1.7225, "step": 17644 }, { "epoch": 1.3828369905956113, "grad_norm": 3.006697416305542, "learning_rate": 7.347630077002629e-06, "loss": 1.3014, "step": 17645 }, { "epoch": 1.3829153605015674, "grad_norm": 3.0405659675598145, "learning_rate": 7.347357813929455e-06, "loss": 1.3155, "step": 17646 }, { "epoch": 1.3829937304075235, "grad_norm": 2.383425235748291, "learning_rate": 7.347085541928136e-06, "loss": 1.5193, "step": 17647 }, { "epoch": 1.3830721003134796, "grad_norm": 3.491567850112915, "learning_rate": 7.346813260999709e-06, "loss": 1.5833, "step": 17648 }, { "epoch": 1.3831504702194357, "grad_norm": 3.715484142303467, "learning_rate": 7.3465409711452115e-06, "loss": 1.1564, "step": 17649 }, { "epoch": 1.3832288401253918, "grad_norm": 5.274644374847412, "learning_rate": 7.346268672365675e-06, "loss": 0.9035, "step": 17650 }, { "epoch": 1.383307210031348, "grad_norm": 3.205585479736328, "learning_rate": 7.3459963646621405e-06, "loss": 0.9935, "step": 17651 }, { "epoch": 1.383385579937304, "grad_norm": 2.444882392883301, "learning_rate": 7.345724048035637e-06, "loss": 1.2709, "step": 17652 }, { "epoch": 1.3834639498432602, "grad_norm": 2.635118246078491, "learning_rate": 7.345451722487207e-06, "loss": 1.2212, "step": 17653 }, { "epoch": 1.3835423197492163, "grad_norm": 2.7121100425720215, "learning_rate": 7.345179388017882e-06, "loss": 1.6457, "step": 17654 }, { "epoch": 1.3836206896551724, "grad_norm": 2.6905157566070557, "learning_rate": 7.3449070446286995e-06, "loss": 1.2268, "step": 17655 }, { "epoch": 1.3836990595611285, "grad_norm": 2.419236660003662, "learning_rate": 7.344634692320696e-06, "loss": 1.2986, "step": 17656 }, { "epoch": 1.3837774294670846, "grad_norm": 3.245723009109497, "learning_rate": 7.344362331094905e-06, "loss": 1.2583, "step": 17657 }, { "epoch": 1.3838557993730407, "grad_norm": 3.9892795085906982, "learning_rate": 7.344089960952364e-06, "loss": 1.317, "step": 17658 }, { "epoch": 1.3839341692789968, "grad_norm": 5.8867716789245605, "learning_rate": 7.343817581894109e-06, "loss": 1.4149, "step": 17659 }, { "epoch": 1.384012539184953, "grad_norm": 3.8944602012634277, "learning_rate": 7.3435451939211754e-06, "loss": 1.486, "step": 17660 }, { "epoch": 1.384090909090909, "grad_norm": 3.7271039485931396, "learning_rate": 7.343272797034602e-06, "loss": 1.3801, "step": 17661 }, { "epoch": 1.3841692789968651, "grad_norm": 3.775578498840332, "learning_rate": 7.343000391235419e-06, "loss": 0.9491, "step": 17662 }, { "epoch": 1.3842476489028213, "grad_norm": 6.108189582824707, "learning_rate": 7.342727976524668e-06, "loss": 1.4534, "step": 17663 }, { "epoch": 1.3843260188087774, "grad_norm": 2.9047820568084717, "learning_rate": 7.342455552903383e-06, "loss": 1.6126, "step": 17664 }, { "epoch": 1.3844043887147335, "grad_norm": 2.695545196533203, "learning_rate": 7.342183120372599e-06, "loss": 1.2244, "step": 17665 }, { "epoch": 1.3844827586206896, "grad_norm": 3.829498767852783, "learning_rate": 7.341910678933355e-06, "loss": 1.4386, "step": 17666 }, { "epoch": 1.3845611285266457, "grad_norm": 2.796596050262451, "learning_rate": 7.341638228586685e-06, "loss": 1.4241, "step": 17667 }, { "epoch": 1.3846394984326018, "grad_norm": 2.8489327430725098, "learning_rate": 7.341365769333626e-06, "loss": 1.2963, "step": 17668 }, { "epoch": 1.384717868338558, "grad_norm": 3.417576313018799, "learning_rate": 7.341093301175213e-06, "loss": 1.4375, "step": 17669 }, { "epoch": 1.384796238244514, "grad_norm": 3.4246742725372314, "learning_rate": 7.340820824112485e-06, "loss": 1.4501, "step": 17670 }, { "epoch": 1.3848746081504701, "grad_norm": 5.111609935760498, "learning_rate": 7.340548338146476e-06, "loss": 1.2293, "step": 17671 }, { "epoch": 1.3849529780564263, "grad_norm": 9.836384773254395, "learning_rate": 7.340275843278225e-06, "loss": 1.0456, "step": 17672 }, { "epoch": 1.3850313479623824, "grad_norm": 3.2157578468322754, "learning_rate": 7.340003339508766e-06, "loss": 1.36, "step": 17673 }, { "epoch": 1.3851097178683385, "grad_norm": 2.187345027923584, "learning_rate": 7.339730826839135e-06, "loss": 1.125, "step": 17674 }, { "epoch": 1.3851880877742948, "grad_norm": 1.933506727218628, "learning_rate": 7.339458305270371e-06, "loss": 1.5082, "step": 17675 }, { "epoch": 1.3852664576802507, "grad_norm": 2.9228403568267822, "learning_rate": 7.3391857748035076e-06, "loss": 1.5019, "step": 17676 }, { "epoch": 1.385344827586207, "grad_norm": 3.115513563156128, "learning_rate": 7.3389132354395855e-06, "loss": 1.2612, "step": 17677 }, { "epoch": 1.385423197492163, "grad_norm": 2.8869707584381104, "learning_rate": 7.3386406871796364e-06, "loss": 1.6448, "step": 17678 }, { "epoch": 1.3855015673981192, "grad_norm": 13.966543197631836, "learning_rate": 7.338368130024701e-06, "loss": 1.6051, "step": 17679 }, { "epoch": 1.3855799373040751, "grad_norm": 2.8520610332489014, "learning_rate": 7.338095563975813e-06, "loss": 1.4903, "step": 17680 }, { "epoch": 1.3856583072100315, "grad_norm": 10.674673080444336, "learning_rate": 7.3378229890340115e-06, "loss": 0.805, "step": 17681 }, { "epoch": 1.3857366771159874, "grad_norm": 4.268157958984375, "learning_rate": 7.337550405200331e-06, "loss": 1.2918, "step": 17682 }, { "epoch": 1.3858150470219437, "grad_norm": 3.4610917568206787, "learning_rate": 7.33727781247581e-06, "loss": 1.5072, "step": 17683 }, { "epoch": 1.3858934169278996, "grad_norm": 2.042236804962158, "learning_rate": 7.337005210861485e-06, "loss": 1.0333, "step": 17684 }, { "epoch": 1.385971786833856, "grad_norm": 2.4168429374694824, "learning_rate": 7.336732600358392e-06, "loss": 1.1854, "step": 17685 }, { "epoch": 1.3860501567398118, "grad_norm": 3.1817679405212402, "learning_rate": 7.336459980967568e-06, "loss": 1.3963, "step": 17686 }, { "epoch": 1.3861285266457681, "grad_norm": 4.867456912994385, "learning_rate": 7.3361873526900515e-06, "loss": 1.4828, "step": 17687 }, { "epoch": 1.386206896551724, "grad_norm": 2.876469373703003, "learning_rate": 7.335914715526877e-06, "loss": 1.1719, "step": 17688 }, { "epoch": 1.3862852664576804, "grad_norm": 2.577033281326294, "learning_rate": 7.335642069479085e-06, "loss": 1.2925, "step": 17689 }, { "epoch": 1.3863636363636362, "grad_norm": 1.8634138107299805, "learning_rate": 7.335369414547708e-06, "loss": 1.4119, "step": 17690 }, { "epoch": 1.3864420062695926, "grad_norm": 3.983595609664917, "learning_rate": 7.335096750733787e-06, "loss": 1.4633, "step": 17691 }, { "epoch": 1.3865203761755485, "grad_norm": 3.716261863708496, "learning_rate": 7.334824078038356e-06, "loss": 1.6786, "step": 17692 }, { "epoch": 1.3865987460815048, "grad_norm": 2.2420437335968018, "learning_rate": 7.334551396462454e-06, "loss": 1.3012, "step": 17693 }, { "epoch": 1.3866771159874607, "grad_norm": 4.4700093269348145, "learning_rate": 7.334278706007117e-06, "loss": 0.9473, "step": 17694 }, { "epoch": 1.386755485893417, "grad_norm": 2.1444668769836426, "learning_rate": 7.334006006673383e-06, "loss": 1.2177, "step": 17695 }, { "epoch": 1.386833855799373, "grad_norm": 2.532327651977539, "learning_rate": 7.33373329846229e-06, "loss": 1.0496, "step": 17696 }, { "epoch": 1.3869122257053292, "grad_norm": 3.938831329345703, "learning_rate": 7.333460581374872e-06, "loss": 1.066, "step": 17697 }, { "epoch": 1.3869905956112853, "grad_norm": 2.620739221572876, "learning_rate": 7.3331878554121705e-06, "loss": 1.3317, "step": 17698 }, { "epoch": 1.3870689655172415, "grad_norm": 2.369821786880493, "learning_rate": 7.3329151205752196e-06, "loss": 1.2505, "step": 17699 }, { "epoch": 1.3871473354231976, "grad_norm": 4.567276954650879, "learning_rate": 7.332642376865059e-06, "loss": 1.651, "step": 17700 }, { "epoch": 1.3872257053291537, "grad_norm": 6.028604984283447, "learning_rate": 7.332369624282724e-06, "loss": 1.0913, "step": 17701 }, { "epoch": 1.3873040752351098, "grad_norm": 2.182521343231201, "learning_rate": 7.332096862829253e-06, "loss": 1.1985, "step": 17702 }, { "epoch": 1.387382445141066, "grad_norm": 4.089737415313721, "learning_rate": 7.331824092505684e-06, "loss": 1.2952, "step": 17703 }, { "epoch": 1.387460815047022, "grad_norm": 2.251483917236328, "learning_rate": 7.3315513133130525e-06, "loss": 1.6486, "step": 17704 }, { "epoch": 1.3875391849529781, "grad_norm": 2.6523804664611816, "learning_rate": 7.331278525252398e-06, "loss": 1.6158, "step": 17705 }, { "epoch": 1.3876175548589342, "grad_norm": 3.3272624015808105, "learning_rate": 7.331005728324759e-06, "loss": 1.1383, "step": 17706 }, { "epoch": 1.3876959247648903, "grad_norm": 2.517360210418701, "learning_rate": 7.33073292253117e-06, "loss": 1.4017, "step": 17707 }, { "epoch": 1.3877742946708465, "grad_norm": 2.893174171447754, "learning_rate": 7.33046010787267e-06, "loss": 1.6076, "step": 17708 }, { "epoch": 1.3878526645768026, "grad_norm": 1.9681063890457153, "learning_rate": 7.330187284350298e-06, "loss": 1.3269, "step": 17709 }, { "epoch": 1.3879310344827587, "grad_norm": 2.78840708732605, "learning_rate": 7.329914451965089e-06, "loss": 2.0243, "step": 17710 }, { "epoch": 1.3880094043887148, "grad_norm": 4.089331150054932, "learning_rate": 7.329641610718082e-06, "loss": 1.3044, "step": 17711 }, { "epoch": 1.388087774294671, "grad_norm": 2.235595226287842, "learning_rate": 7.329368760610316e-06, "loss": 1.2071, "step": 17712 }, { "epoch": 1.388166144200627, "grad_norm": 3.7820420265197754, "learning_rate": 7.329095901642826e-06, "loss": 1.6158, "step": 17713 }, { "epoch": 1.3882445141065831, "grad_norm": 3.664832353591919, "learning_rate": 7.328823033816653e-06, "loss": 1.7664, "step": 17714 }, { "epoch": 1.3883228840125392, "grad_norm": 7.841136932373047, "learning_rate": 7.328550157132834e-06, "loss": 0.769, "step": 17715 }, { "epoch": 1.3884012539184953, "grad_norm": 3.3543381690979004, "learning_rate": 7.328277271592403e-06, "loss": 1.6851, "step": 17716 }, { "epoch": 1.3884796238244514, "grad_norm": 1.9936037063598633, "learning_rate": 7.328004377196405e-06, "loss": 1.307, "step": 17717 }, { "epoch": 1.3885579937304076, "grad_norm": 2.58302903175354, "learning_rate": 7.327731473945873e-06, "loss": 1.2303, "step": 17718 }, { "epoch": 1.3886363636363637, "grad_norm": 3.4354724884033203, "learning_rate": 7.3274585618418445e-06, "loss": 1.0598, "step": 17719 }, { "epoch": 1.3887147335423198, "grad_norm": 2.446979284286499, "learning_rate": 7.327185640885361e-06, "loss": 0.9124, "step": 17720 }, { "epoch": 1.388793103448276, "grad_norm": 3.383327007293701, "learning_rate": 7.326912711077457e-06, "loss": 0.949, "step": 17721 }, { "epoch": 1.388871473354232, "grad_norm": 5.2875590324401855, "learning_rate": 7.326639772419173e-06, "loss": 1.5197, "step": 17722 }, { "epoch": 1.388949843260188, "grad_norm": 3.551206350326538, "learning_rate": 7.326366824911546e-06, "loss": 1.9346, "step": 17723 }, { "epoch": 1.3890282131661442, "grad_norm": 4.041664123535156, "learning_rate": 7.326093868555616e-06, "loss": 1.4008, "step": 17724 }, { "epoch": 1.3891065830721003, "grad_norm": 3.1327924728393555, "learning_rate": 7.325820903352418e-06, "loss": 1.1634, "step": 17725 }, { "epoch": 1.3891849529780564, "grad_norm": 1.62923264503479, "learning_rate": 7.325547929302992e-06, "loss": 1.3186, "step": 17726 }, { "epoch": 1.3892633228840126, "grad_norm": 2.7218525409698486, "learning_rate": 7.325274946408377e-06, "loss": 1.1727, "step": 17727 }, { "epoch": 1.3893416927899687, "grad_norm": 3.3707222938537598, "learning_rate": 7.32500195466961e-06, "loss": 1.3529, "step": 17728 }, { "epoch": 1.3894200626959248, "grad_norm": 5.235511779785156, "learning_rate": 7.32472895408773e-06, "loss": 1.7773, "step": 17729 }, { "epoch": 1.3894984326018809, "grad_norm": 2.2925047874450684, "learning_rate": 7.324455944663774e-06, "loss": 1.3319, "step": 17730 }, { "epoch": 1.389576802507837, "grad_norm": 2.3211352825164795, "learning_rate": 7.324182926398784e-06, "loss": 1.1829, "step": 17731 }, { "epoch": 1.389655172413793, "grad_norm": 2.7426252365112305, "learning_rate": 7.323909899293793e-06, "loss": 1.6912, "step": 17732 }, { "epoch": 1.3897335423197492, "grad_norm": 2.9956371784210205, "learning_rate": 7.323636863349844e-06, "loss": 1.8852, "step": 17733 }, { "epoch": 1.3898119122257053, "grad_norm": 2.8379530906677246, "learning_rate": 7.323363818567972e-06, "loss": 1.2913, "step": 17734 }, { "epoch": 1.3898902821316614, "grad_norm": 4.014809608459473, "learning_rate": 7.323090764949221e-06, "loss": 1.0475, "step": 17735 }, { "epoch": 1.3899686520376175, "grad_norm": 4.563238143920898, "learning_rate": 7.322817702494624e-06, "loss": 0.9397, "step": 17736 }, { "epoch": 1.3900470219435737, "grad_norm": 3.8903582096099854, "learning_rate": 7.32254463120522e-06, "loss": 1.0574, "step": 17737 }, { "epoch": 1.3901253918495298, "grad_norm": 2.96919584274292, "learning_rate": 7.32227155108205e-06, "loss": 1.0825, "step": 17738 }, { "epoch": 1.3902037617554859, "grad_norm": 2.600586414337158, "learning_rate": 7.321998462126153e-06, "loss": 1.467, "step": 17739 }, { "epoch": 1.390282131661442, "grad_norm": 11.263191223144531, "learning_rate": 7.321725364338566e-06, "loss": 2.146, "step": 17740 }, { "epoch": 1.390360501567398, "grad_norm": 2.881044387817383, "learning_rate": 7.321452257720327e-06, "loss": 1.4066, "step": 17741 }, { "epoch": 1.3904388714733542, "grad_norm": 2.595273733139038, "learning_rate": 7.321179142272477e-06, "loss": 1.5897, "step": 17742 }, { "epoch": 1.3905172413793103, "grad_norm": 4.616004943847656, "learning_rate": 7.3209060179960555e-06, "loss": 1.1393, "step": 17743 }, { "epoch": 1.3905956112852664, "grad_norm": 2.9084506034851074, "learning_rate": 7.320632884892097e-06, "loss": 1.4377, "step": 17744 }, { "epoch": 1.3906739811912225, "grad_norm": 4.119938850402832, "learning_rate": 7.320359742961644e-06, "loss": 1.1155, "step": 17745 }, { "epoch": 1.3907523510971787, "grad_norm": 4.684494972229004, "learning_rate": 7.320086592205735e-06, "loss": 1.8159, "step": 17746 }, { "epoch": 1.3908307210031348, "grad_norm": 4.186325550079346, "learning_rate": 7.319813432625408e-06, "loss": 1.4421, "step": 17747 }, { "epoch": 1.3909090909090909, "grad_norm": 2.3539986610412598, "learning_rate": 7.3195402642217e-06, "loss": 1.3693, "step": 17748 }, { "epoch": 1.390987460815047, "grad_norm": 3.1950953006744385, "learning_rate": 7.319267086995655e-06, "loss": 1.5835, "step": 17749 }, { "epoch": 1.391065830721003, "grad_norm": 2.423030376434326, "learning_rate": 7.318993900948309e-06, "loss": 1.3148, "step": 17750 }, { "epoch": 1.3911442006269592, "grad_norm": 4.072287082672119, "learning_rate": 7.3187207060807e-06, "loss": 1.22, "step": 17751 }, { "epoch": 1.3912225705329153, "grad_norm": 3.234590768814087, "learning_rate": 7.318447502393869e-06, "loss": 1.5607, "step": 17752 }, { "epoch": 1.3913009404388714, "grad_norm": 2.4934284687042236, "learning_rate": 7.318174289888854e-06, "loss": 1.3885, "step": 17753 }, { "epoch": 1.3913793103448275, "grad_norm": 3.542701244354248, "learning_rate": 7.317901068566696e-06, "loss": 1.3988, "step": 17754 }, { "epoch": 1.3914576802507836, "grad_norm": 4.888550281524658, "learning_rate": 7.317627838428431e-06, "loss": 1.7466, "step": 17755 }, { "epoch": 1.3915360501567398, "grad_norm": 1.993988275527954, "learning_rate": 7.317354599475101e-06, "loss": 1.25, "step": 17756 }, { "epoch": 1.3916144200626959, "grad_norm": 4.120991230010986, "learning_rate": 7.317081351707746e-06, "loss": 1.1544, "step": 17757 }, { "epoch": 1.391692789968652, "grad_norm": 4.165445327758789, "learning_rate": 7.316808095127402e-06, "loss": 1.3662, "step": 17758 }, { "epoch": 1.391771159874608, "grad_norm": 2.686490535736084, "learning_rate": 7.316534829735111e-06, "loss": 1.159, "step": 17759 }, { "epoch": 1.3918495297805642, "grad_norm": 6.473156452178955, "learning_rate": 7.31626155553191e-06, "loss": 1.6104, "step": 17760 }, { "epoch": 1.3919278996865203, "grad_norm": 2.252558946609497, "learning_rate": 7.3159882725188404e-06, "loss": 1.2819, "step": 17761 }, { "epoch": 1.3920062695924764, "grad_norm": 1.7938036918640137, "learning_rate": 7.315714980696941e-06, "loss": 1.1963, "step": 17762 }, { "epoch": 1.3920846394984325, "grad_norm": 2.7258005142211914, "learning_rate": 7.315441680067251e-06, "loss": 1.373, "step": 17763 }, { "epoch": 1.3921630094043886, "grad_norm": 2.812133550643921, "learning_rate": 7.315168370630809e-06, "loss": 0.7132, "step": 17764 }, { "epoch": 1.3922413793103448, "grad_norm": 5.059963703155518, "learning_rate": 7.314895052388657e-06, "loss": 1.1551, "step": 17765 }, { "epoch": 1.3923197492163009, "grad_norm": 3.4444713592529297, "learning_rate": 7.314621725341834e-06, "loss": 1.3948, "step": 17766 }, { "epoch": 1.392398119122257, "grad_norm": 5.546612739562988, "learning_rate": 7.314348389491378e-06, "loss": 1.3578, "step": 17767 }, { "epoch": 1.3924764890282133, "grad_norm": 2.092420816421509, "learning_rate": 7.314075044838329e-06, "loss": 1.2537, "step": 17768 }, { "epoch": 1.3925548589341692, "grad_norm": 1.8936480283737183, "learning_rate": 7.3138016913837284e-06, "loss": 1.2213, "step": 17769 }, { "epoch": 1.3926332288401255, "grad_norm": 6.895576000213623, "learning_rate": 7.313528329128613e-06, "loss": 1.0217, "step": 17770 }, { "epoch": 1.3927115987460814, "grad_norm": 5.107694149017334, "learning_rate": 7.313254958074025e-06, "loss": 1.1438, "step": 17771 }, { "epoch": 1.3927899686520377, "grad_norm": 3.1533796787261963, "learning_rate": 7.312981578221004e-06, "loss": 1.4502, "step": 17772 }, { "epoch": 1.3928683385579936, "grad_norm": 5.720098495483398, "learning_rate": 7.312708189570588e-06, "loss": 1.4865, "step": 17773 }, { "epoch": 1.39294670846395, "grad_norm": 1.947256326675415, "learning_rate": 7.312434792123818e-06, "loss": 1.0888, "step": 17774 }, { "epoch": 1.3930250783699059, "grad_norm": 3.589719533920288, "learning_rate": 7.312161385881734e-06, "loss": 1.6076, "step": 17775 }, { "epoch": 1.3931034482758622, "grad_norm": 2.8018338680267334, "learning_rate": 7.311887970845376e-06, "loss": 1.0391, "step": 17776 }, { "epoch": 1.393181818181818, "grad_norm": 3.1292099952697754, "learning_rate": 7.311614547015783e-06, "loss": 1.5889, "step": 17777 }, { "epoch": 1.3932601880877744, "grad_norm": 7.250228404998779, "learning_rate": 7.311341114393997e-06, "loss": 1.4158, "step": 17778 }, { "epoch": 1.3933385579937303, "grad_norm": 2.8288559913635254, "learning_rate": 7.311067672981054e-06, "loss": 1.4733, "step": 17779 }, { "epoch": 1.3934169278996866, "grad_norm": 1.9646308422088623, "learning_rate": 7.310794222778e-06, "loss": 1.1812, "step": 17780 }, { "epoch": 1.3934952978056425, "grad_norm": 2.715128183364868, "learning_rate": 7.310520763785868e-06, "loss": 1.1071, "step": 17781 }, { "epoch": 1.3935736677115989, "grad_norm": 2.20413875579834, "learning_rate": 7.3102472960057034e-06, "loss": 1.3087, "step": 17782 }, { "epoch": 1.3936520376175547, "grad_norm": 7.770510673522949, "learning_rate": 7.309973819438545e-06, "loss": 1.3551, "step": 17783 }, { "epoch": 1.393730407523511, "grad_norm": 23.862770080566406, "learning_rate": 7.309700334085434e-06, "loss": 1.6398, "step": 17784 }, { "epoch": 1.393808777429467, "grad_norm": 3.0079824924468994, "learning_rate": 7.309426839947407e-06, "loss": 1.6579, "step": 17785 }, { "epoch": 1.3938871473354233, "grad_norm": 3.1392982006073, "learning_rate": 7.309153337025507e-06, "loss": 1.3101, "step": 17786 }, { "epoch": 1.3939655172413792, "grad_norm": 4.6201252937316895, "learning_rate": 7.308879825320775e-06, "loss": 1.4389, "step": 17787 }, { "epoch": 1.3940438871473355, "grad_norm": 4.887069225311279, "learning_rate": 7.3086063048342485e-06, "loss": 1.9742, "step": 17788 }, { "epoch": 1.3941222570532914, "grad_norm": 3.575085163116455, "learning_rate": 7.308332775566971e-06, "loss": 1.374, "step": 17789 }, { "epoch": 1.3942006269592477, "grad_norm": 3.1906373500823975, "learning_rate": 7.30805923751998e-06, "loss": 0.5418, "step": 17790 }, { "epoch": 1.3942789968652036, "grad_norm": 2.0186538696289062, "learning_rate": 7.307785690694316e-06, "loss": 1.3786, "step": 17791 }, { "epoch": 1.39435736677116, "grad_norm": 2.127871513366699, "learning_rate": 7.307512135091023e-06, "loss": 1.5803, "step": 17792 }, { "epoch": 1.394435736677116, "grad_norm": 3.4354286193847656, "learning_rate": 7.307238570711137e-06, "loss": 1.1975, "step": 17793 }, { "epoch": 1.3945141065830722, "grad_norm": 2.359908103942871, "learning_rate": 7.306964997555702e-06, "loss": 1.3745, "step": 17794 }, { "epoch": 1.3945924764890283, "grad_norm": 2.4947869777679443, "learning_rate": 7.306691415625757e-06, "loss": 1.4187, "step": 17795 }, { "epoch": 1.3946708463949844, "grad_norm": 2.671443462371826, "learning_rate": 7.306417824922341e-06, "loss": 1.1285, "step": 17796 }, { "epoch": 1.3947492163009405, "grad_norm": 6.544776439666748, "learning_rate": 7.3061442254465e-06, "loss": 1.7002, "step": 17797 }, { "epoch": 1.3948275862068966, "grad_norm": 3.4372849464416504, "learning_rate": 7.3058706171992664e-06, "loss": 1.1893, "step": 17798 }, { "epoch": 1.3949059561128527, "grad_norm": 9.274222373962402, "learning_rate": 7.305597000181689e-06, "loss": 1.1001, "step": 17799 }, { "epoch": 1.3949843260188088, "grad_norm": 3.34177565574646, "learning_rate": 7.305323374394802e-06, "loss": 1.0193, "step": 17800 }, { "epoch": 1.395062695924765, "grad_norm": 3.294640302658081, "learning_rate": 7.3050497398396515e-06, "loss": 1.3613, "step": 17801 }, { "epoch": 1.395141065830721, "grad_norm": 3.8572726249694824, "learning_rate": 7.304776096517274e-06, "loss": 1.046, "step": 17802 }, { "epoch": 1.3952194357366772, "grad_norm": 2.574312925338745, "learning_rate": 7.304502444428711e-06, "loss": 1.5314, "step": 17803 }, { "epoch": 1.3952978056426333, "grad_norm": 2.5855283737182617, "learning_rate": 7.304228783575008e-06, "loss": 1.5611, "step": 17804 }, { "epoch": 1.3953761755485894, "grad_norm": 3.7235372066497803, "learning_rate": 7.303955113957199e-06, "loss": 1.0059, "step": 17805 }, { "epoch": 1.3954545454545455, "grad_norm": 2.4228320121765137, "learning_rate": 7.303681435576329e-06, "loss": 1.5442, "step": 17806 }, { "epoch": 1.3955329153605016, "grad_norm": 3.2249326705932617, "learning_rate": 7.303407748433438e-06, "loss": 1.249, "step": 17807 }, { "epoch": 1.3956112852664577, "grad_norm": 2.0601537227630615, "learning_rate": 7.303134052529568e-06, "loss": 0.9939, "step": 17808 }, { "epoch": 1.3956896551724138, "grad_norm": 3.769315481185913, "learning_rate": 7.3028603478657586e-06, "loss": 1.4082, "step": 17809 }, { "epoch": 1.39576802507837, "grad_norm": 3.179776191711426, "learning_rate": 7.3025866344430505e-06, "loss": 1.4278, "step": 17810 }, { "epoch": 1.395846394984326, "grad_norm": 3.126394748687744, "learning_rate": 7.302312912262486e-06, "loss": 1.4654, "step": 17811 }, { "epoch": 1.3959247648902822, "grad_norm": 3.03847336769104, "learning_rate": 7.302039181325106e-06, "loss": 1.3072, "step": 17812 }, { "epoch": 1.3960031347962383, "grad_norm": 3.354961395263672, "learning_rate": 7.301765441631951e-06, "loss": 1.0423, "step": 17813 }, { "epoch": 1.3960815047021944, "grad_norm": 2.878986358642578, "learning_rate": 7.3014916931840625e-06, "loss": 1.5789, "step": 17814 }, { "epoch": 1.3961598746081505, "grad_norm": 5.441644191741943, "learning_rate": 7.301217935982483e-06, "loss": 1.3399, "step": 17815 }, { "epoch": 1.3962382445141066, "grad_norm": 2.882906675338745, "learning_rate": 7.300944170028252e-06, "loss": 1.0466, "step": 17816 }, { "epoch": 1.3963166144200627, "grad_norm": 4.798153400421143, "learning_rate": 7.300670395322409e-06, "loss": 1.1942, "step": 17817 }, { "epoch": 1.3963949843260188, "grad_norm": 2.8486814498901367, "learning_rate": 7.300396611866e-06, "loss": 1.4154, "step": 17818 }, { "epoch": 1.396473354231975, "grad_norm": 3.523798704147339, "learning_rate": 7.300122819660063e-06, "loss": 1.8236, "step": 17819 }, { "epoch": 1.396551724137931, "grad_norm": 2.31894588470459, "learning_rate": 7.299849018705639e-06, "loss": 1.3729, "step": 17820 }, { "epoch": 1.3966300940438872, "grad_norm": 3.844879150390625, "learning_rate": 7.2995752090037715e-06, "loss": 1.6052, "step": 17821 }, { "epoch": 1.3967084639498433, "grad_norm": 3.596179246902466, "learning_rate": 7.299301390555501e-06, "loss": 1.5209, "step": 17822 }, { "epoch": 1.3967868338557994, "grad_norm": 3.6962618827819824, "learning_rate": 7.29902756336187e-06, "loss": 1.5335, "step": 17823 }, { "epoch": 1.3968652037617555, "grad_norm": 2.084930896759033, "learning_rate": 7.298753727423917e-06, "loss": 1.1261, "step": 17824 }, { "epoch": 1.3969435736677116, "grad_norm": 1.9337605237960815, "learning_rate": 7.298479882742688e-06, "loss": 0.9309, "step": 17825 }, { "epoch": 1.3970219435736677, "grad_norm": 4.623090744018555, "learning_rate": 7.29820602931922e-06, "loss": 1.3818, "step": 17826 }, { "epoch": 1.3971003134796238, "grad_norm": 4.774962425231934, "learning_rate": 7.2979321671545585e-06, "loss": 1.2365, "step": 17827 }, { "epoch": 1.39717868338558, "grad_norm": 9.247745513916016, "learning_rate": 7.2976582962497425e-06, "loss": 1.6136, "step": 17828 }, { "epoch": 1.397257053291536, "grad_norm": 2.56081485748291, "learning_rate": 7.297384416605814e-06, "loss": 1.3359, "step": 17829 }, { "epoch": 1.3973354231974922, "grad_norm": 2.8566486835479736, "learning_rate": 7.297110528223817e-06, "loss": 1.5932, "step": 17830 }, { "epoch": 1.3974137931034483, "grad_norm": 14.506032943725586, "learning_rate": 7.296836631104789e-06, "loss": 2.0647, "step": 17831 }, { "epoch": 1.3974921630094044, "grad_norm": 2.2911570072174072, "learning_rate": 7.296562725249776e-06, "loss": 1.4264, "step": 17832 }, { "epoch": 1.3975705329153605, "grad_norm": 3.6265106201171875, "learning_rate": 7.296288810659816e-06, "loss": 1.1346, "step": 17833 }, { "epoch": 1.3976489028213166, "grad_norm": 3.2900726795196533, "learning_rate": 7.296014887335955e-06, "loss": 1.6283, "step": 17834 }, { "epoch": 1.3977272727272727, "grad_norm": 4.069955348968506, "learning_rate": 7.29574095527923e-06, "loss": 1.0284, "step": 17835 }, { "epoch": 1.3978056426332288, "grad_norm": 4.186624050140381, "learning_rate": 7.295467014490689e-06, "loss": 1.351, "step": 17836 }, { "epoch": 1.397884012539185, "grad_norm": 4.455310821533203, "learning_rate": 7.295193064971368e-06, "loss": 0.9635, "step": 17837 }, { "epoch": 1.397962382445141, "grad_norm": 3.039381980895996, "learning_rate": 7.294919106722312e-06, "loss": 1.2034, "step": 17838 }, { "epoch": 1.3980407523510971, "grad_norm": 2.4860310554504395, "learning_rate": 7.294645139744563e-06, "loss": 1.5189, "step": 17839 }, { "epoch": 1.3981191222570533, "grad_norm": 3.3298234939575195, "learning_rate": 7.2943711640391626e-06, "loss": 1.4898, "step": 17840 }, { "epoch": 1.3981974921630094, "grad_norm": 1.9804686307907104, "learning_rate": 7.294097179607152e-06, "loss": 1.3636, "step": 17841 }, { "epoch": 1.3982758620689655, "grad_norm": 2.1180508136749268, "learning_rate": 7.293823186449573e-06, "loss": 1.4067, "step": 17842 }, { "epoch": 1.3983542319749216, "grad_norm": 3.68670392036438, "learning_rate": 7.2935491845674725e-06, "loss": 1.9024, "step": 17843 }, { "epoch": 1.3984326018808777, "grad_norm": 2.3817873001098633, "learning_rate": 7.293275173961885e-06, "loss": 1.0848, "step": 17844 }, { "epoch": 1.3985109717868338, "grad_norm": 2.6971542835235596, "learning_rate": 7.2930011546338585e-06, "loss": 1.7643, "step": 17845 }, { "epoch": 1.39858934169279, "grad_norm": 3.19616961479187, "learning_rate": 7.292727126584433e-06, "loss": 1.4948, "step": 17846 }, { "epoch": 1.398667711598746, "grad_norm": 3.4184272289276123, "learning_rate": 7.292453089814651e-06, "loss": 1.1616, "step": 17847 }, { "epoch": 1.3987460815047021, "grad_norm": 2.530794382095337, "learning_rate": 7.2921790443255555e-06, "loss": 1.3919, "step": 17848 }, { "epoch": 1.3988244514106583, "grad_norm": 4.884702682495117, "learning_rate": 7.291904990118188e-06, "loss": 1.4676, "step": 17849 }, { "epoch": 1.3989028213166144, "grad_norm": 3.3849704265594482, "learning_rate": 7.291630927193592e-06, "loss": 0.6461, "step": 17850 }, { "epoch": 1.3989811912225705, "grad_norm": 2.2943060398101807, "learning_rate": 7.291356855552808e-06, "loss": 1.1771, "step": 17851 }, { "epoch": 1.3990595611285266, "grad_norm": 1.936705470085144, "learning_rate": 7.291082775196878e-06, "loss": 0.8432, "step": 17852 }, { "epoch": 1.3991379310344827, "grad_norm": 4.265669345855713, "learning_rate": 7.290808686126849e-06, "loss": 0.7665, "step": 17853 }, { "epoch": 1.3992163009404388, "grad_norm": 2.711259126663208, "learning_rate": 7.290534588343757e-06, "loss": 1.6375, "step": 17854 }, { "epoch": 1.399294670846395, "grad_norm": 3.1702842712402344, "learning_rate": 7.290260481848651e-06, "loss": 1.6491, "step": 17855 }, { "epoch": 1.399373040752351, "grad_norm": 2.2638590335845947, "learning_rate": 7.289986366642569e-06, "loss": 1.3394, "step": 17856 }, { "epoch": 1.3994514106583071, "grad_norm": 2.7943813800811768, "learning_rate": 7.289712242726555e-06, "loss": 1.2847, "step": 17857 }, { "epoch": 1.3995297805642632, "grad_norm": 5.391011714935303, "learning_rate": 7.289438110101653e-06, "loss": 1.646, "step": 17858 }, { "epoch": 1.3996081504702194, "grad_norm": 2.2727556228637695, "learning_rate": 7.289163968768902e-06, "loss": 1.1099, "step": 17859 }, { "epoch": 1.3996865203761755, "grad_norm": 3.2345635890960693, "learning_rate": 7.28888981872935e-06, "loss": 1.3485, "step": 17860 }, { "epoch": 1.3997648902821316, "grad_norm": 3.5826022624969482, "learning_rate": 7.288615659984033e-06, "loss": 1.5736, "step": 17861 }, { "epoch": 1.3998432601880877, "grad_norm": 2.9726035594940186, "learning_rate": 7.288341492534001e-06, "loss": 1.3627, "step": 17862 }, { "epoch": 1.399921630094044, "grad_norm": 2.1886894702911377, "learning_rate": 7.288067316380292e-06, "loss": 1.489, "step": 17863 }, { "epoch": 1.4, "grad_norm": 4.779746055603027, "learning_rate": 7.2877931315239505e-06, "loss": 1.5251, "step": 17864 }, { "epoch": 1.4000783699059562, "grad_norm": 5.860010147094727, "learning_rate": 7.287518937966018e-06, "loss": 1.3349, "step": 17865 }, { "epoch": 1.4001567398119121, "grad_norm": 2.1499898433685303, "learning_rate": 7.287244735707539e-06, "loss": 1.6988, "step": 17866 }, { "epoch": 1.4002351097178685, "grad_norm": 3.098989963531494, "learning_rate": 7.286970524749556e-06, "loss": 1.1741, "step": 17867 }, { "epoch": 1.4003134796238244, "grad_norm": 3.199402332305908, "learning_rate": 7.2866963050931115e-06, "loss": 1.2884, "step": 17868 }, { "epoch": 1.4003918495297807, "grad_norm": 3.1260292530059814, "learning_rate": 7.286422076739248e-06, "loss": 1.5661, "step": 17869 }, { "epoch": 1.4004702194357366, "grad_norm": 2.480175495147705, "learning_rate": 7.286147839689011e-06, "loss": 1.4958, "step": 17870 }, { "epoch": 1.400548589341693, "grad_norm": 2.783205986022949, "learning_rate": 7.2858735939434425e-06, "loss": 1.3173, "step": 17871 }, { "epoch": 1.4006269592476488, "grad_norm": 3.906284809112549, "learning_rate": 7.285599339503583e-06, "loss": 1.2104, "step": 17872 }, { "epoch": 1.4007053291536051, "grad_norm": 2.5433225631713867, "learning_rate": 7.285325076370477e-06, "loss": 1.4206, "step": 17873 }, { "epoch": 1.400783699059561, "grad_norm": 1.8349894285202026, "learning_rate": 7.2850508045451705e-06, "loss": 1.4298, "step": 17874 }, { "epoch": 1.4008620689655173, "grad_norm": 2.8914198875427246, "learning_rate": 7.284776524028703e-06, "loss": 1.3114, "step": 17875 }, { "epoch": 1.4009404388714732, "grad_norm": 2.8869311809539795, "learning_rate": 7.28450223482212e-06, "loss": 1.3165, "step": 17876 }, { "epoch": 1.4010188087774296, "grad_norm": 2.7284140586853027, "learning_rate": 7.284227936926462e-06, "loss": 1.4644, "step": 17877 }, { "epoch": 1.4010971786833855, "grad_norm": 3.2195634841918945, "learning_rate": 7.2839536303427745e-06, "loss": 1.2333, "step": 17878 }, { "epoch": 1.4011755485893418, "grad_norm": 2.1860294342041016, "learning_rate": 7.283679315072103e-06, "loss": 1.0283, "step": 17879 }, { "epoch": 1.4012539184952977, "grad_norm": 2.075106620788574, "learning_rate": 7.283404991115486e-06, "loss": 1.3869, "step": 17880 }, { "epoch": 1.401332288401254, "grad_norm": 3.5219321250915527, "learning_rate": 7.2831306584739715e-06, "loss": 1.4081, "step": 17881 }, { "epoch": 1.40141065830721, "grad_norm": 2.3934404850006104, "learning_rate": 7.282856317148599e-06, "loss": 1.2951, "step": 17882 }, { "epoch": 1.4014890282131662, "grad_norm": 4.879157066345215, "learning_rate": 7.282581967140414e-06, "loss": 1.3391, "step": 17883 }, { "epoch": 1.4015673981191221, "grad_norm": 2.886043071746826, "learning_rate": 7.282307608450458e-06, "loss": 1.2761, "step": 17884 }, { "epoch": 1.4016457680250785, "grad_norm": 3.0113744735717773, "learning_rate": 7.282033241079779e-06, "loss": 1.2635, "step": 17885 }, { "epoch": 1.4017241379310346, "grad_norm": 1.833416223526001, "learning_rate": 7.281758865029417e-06, "loss": 1.3478, "step": 17886 }, { "epoch": 1.4018025078369907, "grad_norm": 3.6874263286590576, "learning_rate": 7.281484480300414e-06, "loss": 1.3761, "step": 17887 }, { "epoch": 1.4018808777429468, "grad_norm": 2.2123448848724365, "learning_rate": 7.281210086893818e-06, "loss": 1.4304, "step": 17888 }, { "epoch": 1.401959247648903, "grad_norm": 2.067140817642212, "learning_rate": 7.280935684810669e-06, "loss": 1.4844, "step": 17889 }, { "epoch": 1.402037617554859, "grad_norm": 5.243889331817627, "learning_rate": 7.280661274052014e-06, "loss": 1.9297, "step": 17890 }, { "epoch": 1.4021159874608151, "grad_norm": 4.160791397094727, "learning_rate": 7.2803868546188945e-06, "loss": 1.2768, "step": 17891 }, { "epoch": 1.4021943573667712, "grad_norm": 2.915567636489868, "learning_rate": 7.280112426512353e-06, "loss": 1.4993, "step": 17892 }, { "epoch": 1.4022727272727273, "grad_norm": 11.596925735473633, "learning_rate": 7.279837989733437e-06, "loss": 0.7147, "step": 17893 }, { "epoch": 1.4023510971786834, "grad_norm": 2.746544361114502, "learning_rate": 7.279563544283188e-06, "loss": 1.2301, "step": 17894 }, { "epoch": 1.4024294670846396, "grad_norm": 3.8169007301330566, "learning_rate": 7.279289090162649e-06, "loss": 1.5685, "step": 17895 }, { "epoch": 1.4025078369905957, "grad_norm": 2.8633766174316406, "learning_rate": 7.2790146273728655e-06, "loss": 1.2818, "step": 17896 }, { "epoch": 1.4025862068965518, "grad_norm": 3.4513068199157715, "learning_rate": 7.2787401559148815e-06, "loss": 1.7505, "step": 17897 }, { "epoch": 1.402664576802508, "grad_norm": 3.8608322143554688, "learning_rate": 7.278465675789741e-06, "loss": 1.1429, "step": 17898 }, { "epoch": 1.402742946708464, "grad_norm": 4.169771194458008, "learning_rate": 7.278191186998487e-06, "loss": 1.0607, "step": 17899 }, { "epoch": 1.40282131661442, "grad_norm": 4.8649773597717285, "learning_rate": 7.277916689542163e-06, "loss": 1.3737, "step": 17900 }, { "epoch": 1.4028996865203762, "grad_norm": 2.467424154281616, "learning_rate": 7.277642183421814e-06, "loss": 1.3661, "step": 17901 }, { "epoch": 1.4029780564263323, "grad_norm": 3.112006664276123, "learning_rate": 7.277367668638484e-06, "loss": 1.1981, "step": 17902 }, { "epoch": 1.4030564263322884, "grad_norm": 2.587913990020752, "learning_rate": 7.277093145193217e-06, "loss": 1.0947, "step": 17903 }, { "epoch": 1.4031347962382446, "grad_norm": 2.3877856731414795, "learning_rate": 7.276818613087058e-06, "loss": 1.389, "step": 17904 }, { "epoch": 1.4032131661442007, "grad_norm": 2.128495216369629, "learning_rate": 7.27654407232105e-06, "loss": 1.3773, "step": 17905 }, { "epoch": 1.4032915360501568, "grad_norm": 3.9810049533843994, "learning_rate": 7.2762695228962365e-06, "loss": 1.1861, "step": 17906 }, { "epoch": 1.4033699059561129, "grad_norm": 2.597562074661255, "learning_rate": 7.275994964813665e-06, "loss": 1.0715, "step": 17907 }, { "epoch": 1.403448275862069, "grad_norm": 3.6494197845458984, "learning_rate": 7.2757203980743764e-06, "loss": 1.5772, "step": 17908 }, { "epoch": 1.403526645768025, "grad_norm": 4.84579610824585, "learning_rate": 7.275445822679416e-06, "loss": 1.3832, "step": 17909 }, { "epoch": 1.4036050156739812, "grad_norm": 7.180422782897949, "learning_rate": 7.27517123862983e-06, "loss": 1.5849, "step": 17910 }, { "epoch": 1.4036833855799373, "grad_norm": 1.9309935569763184, "learning_rate": 7.27489664592666e-06, "loss": 1.4423, "step": 17911 }, { "epoch": 1.4037617554858934, "grad_norm": 3.7988760471343994, "learning_rate": 7.2746220445709516e-06, "loss": 1.0139, "step": 17912 }, { "epoch": 1.4038401253918495, "grad_norm": 2.37065052986145, "learning_rate": 7.27434743456375e-06, "loss": 1.4431, "step": 17913 }, { "epoch": 1.4039184952978057, "grad_norm": 3.55826473236084, "learning_rate": 7.274072815906099e-06, "loss": 1.7869, "step": 17914 }, { "epoch": 1.4039968652037618, "grad_norm": 1.6212084293365479, "learning_rate": 7.2737981885990415e-06, "loss": 1.1761, "step": 17915 }, { "epoch": 1.4040752351097179, "grad_norm": 2.07576847076416, "learning_rate": 7.273523552643625e-06, "loss": 1.4168, "step": 17916 }, { "epoch": 1.404153605015674, "grad_norm": 2.0966830253601074, "learning_rate": 7.273248908040893e-06, "loss": 1.4475, "step": 17917 }, { "epoch": 1.40423197492163, "grad_norm": 2.0543692111968994, "learning_rate": 7.272974254791889e-06, "loss": 1.7949, "step": 17918 }, { "epoch": 1.4043103448275862, "grad_norm": 2.7758235931396484, "learning_rate": 7.2726995928976585e-06, "loss": 1.4641, "step": 17919 }, { "epoch": 1.4043887147335423, "grad_norm": 3.1985080242156982, "learning_rate": 7.272424922359246e-06, "loss": 1.418, "step": 17920 }, { "epoch": 1.4044670846394984, "grad_norm": 3.0657877922058105, "learning_rate": 7.272150243177696e-06, "loss": 1.0806, "step": 17921 }, { "epoch": 1.4045454545454545, "grad_norm": 9.238110542297363, "learning_rate": 7.271875555354053e-06, "loss": 2.1395, "step": 17922 }, { "epoch": 1.4046238244514107, "grad_norm": 7.462965965270996, "learning_rate": 7.271600858889364e-06, "loss": 1.3012, "step": 17923 }, { "epoch": 1.4047021943573668, "grad_norm": 4.667387962341309, "learning_rate": 7.271326153784672e-06, "loss": 1.2393, "step": 17924 }, { "epoch": 1.4047805642633229, "grad_norm": 7.502115249633789, "learning_rate": 7.271051440041021e-06, "loss": 1.5218, "step": 17925 }, { "epoch": 1.404858934169279, "grad_norm": 2.6153080463409424, "learning_rate": 7.270776717659457e-06, "loss": 1.3113, "step": 17926 }, { "epoch": 1.404937304075235, "grad_norm": 6.210639476776123, "learning_rate": 7.270501986641024e-06, "loss": 1.3403, "step": 17927 }, { "epoch": 1.4050156739811912, "grad_norm": 2.487013578414917, "learning_rate": 7.270227246986768e-06, "loss": 1.6191, "step": 17928 }, { "epoch": 1.4050940438871473, "grad_norm": 6.966084957122803, "learning_rate": 7.269952498697734e-06, "loss": 1.4953, "step": 17929 }, { "epoch": 1.4051724137931034, "grad_norm": 8.248346328735352, "learning_rate": 7.269677741774968e-06, "loss": 1.2341, "step": 17930 }, { "epoch": 1.4052507836990595, "grad_norm": 2.5315048694610596, "learning_rate": 7.269402976219512e-06, "loss": 1.255, "step": 17931 }, { "epoch": 1.4053291536050156, "grad_norm": 2.4528608322143555, "learning_rate": 7.269128202032412e-06, "loss": 1.5744, "step": 17932 }, { "epoch": 1.4054075235109718, "grad_norm": 4.145180702209473, "learning_rate": 7.268853419214717e-06, "loss": 1.3889, "step": 17933 }, { "epoch": 1.4054858934169279, "grad_norm": 2.9457263946533203, "learning_rate": 7.268578627767464e-06, "loss": 0.8524, "step": 17934 }, { "epoch": 1.405564263322884, "grad_norm": 2.247342586517334, "learning_rate": 7.268303827691707e-06, "loss": 1.5455, "step": 17935 }, { "epoch": 1.40564263322884, "grad_norm": 2.8402915000915527, "learning_rate": 7.268029018988485e-06, "loss": 1.4943, "step": 17936 }, { "epoch": 1.4057210031347962, "grad_norm": 2.8463099002838135, "learning_rate": 7.267754201658848e-06, "loss": 1.4303, "step": 17937 }, { "epoch": 1.4057993730407523, "grad_norm": 2.4451098442077637, "learning_rate": 7.2674793757038375e-06, "loss": 1.4873, "step": 17938 }, { "epoch": 1.4058777429467084, "grad_norm": 3.2522385120391846, "learning_rate": 7.267204541124499e-06, "loss": 1.5039, "step": 17939 }, { "epoch": 1.4059561128526645, "grad_norm": 2.194361925125122, "learning_rate": 7.2669296979218805e-06, "loss": 1.5541, "step": 17940 }, { "epoch": 1.4060344827586206, "grad_norm": 1.9245973825454712, "learning_rate": 7.266654846097025e-06, "loss": 1.3488, "step": 17941 }, { "epoch": 1.4061128526645768, "grad_norm": 2.993044376373291, "learning_rate": 7.2663799856509805e-06, "loss": 1.3441, "step": 17942 }, { "epoch": 1.4061912225705329, "grad_norm": 3.474870443344116, "learning_rate": 7.266105116584788e-06, "loss": 1.4142, "step": 17943 }, { "epoch": 1.406269592476489, "grad_norm": 3.212970733642578, "learning_rate": 7.265830238899499e-06, "loss": 1.169, "step": 17944 }, { "epoch": 1.406347962382445, "grad_norm": 3.8803465366363525, "learning_rate": 7.265555352596152e-06, "loss": 1.9182, "step": 17945 }, { "epoch": 1.4064263322884012, "grad_norm": 6.788388252258301, "learning_rate": 7.265280457675797e-06, "loss": 1.2371, "step": 17946 }, { "epoch": 1.4065047021943573, "grad_norm": 2.407911777496338, "learning_rate": 7.2650055541394796e-06, "loss": 1.5572, "step": 17947 }, { "epoch": 1.4065830721003134, "grad_norm": 2.5110440254211426, "learning_rate": 7.264730641988243e-06, "loss": 1.4165, "step": 17948 }, { "epoch": 1.4066614420062695, "grad_norm": 2.7909934520721436, "learning_rate": 7.264455721223136e-06, "loss": 1.7316, "step": 17949 }, { "epoch": 1.4067398119122256, "grad_norm": 4.409964561462402, "learning_rate": 7.264180791845201e-06, "loss": 1.5353, "step": 17950 }, { "epoch": 1.4068181818181817, "grad_norm": 6.988717079162598, "learning_rate": 7.263905853855486e-06, "loss": 1.8494, "step": 17951 }, { "epoch": 1.4068965517241379, "grad_norm": 3.4113502502441406, "learning_rate": 7.263630907255037e-06, "loss": 1.3395, "step": 17952 }, { "epoch": 1.406974921630094, "grad_norm": 4.987044334411621, "learning_rate": 7.263355952044897e-06, "loss": 1.5783, "step": 17953 }, { "epoch": 1.40705329153605, "grad_norm": 2.2117795944213867, "learning_rate": 7.263080988226115e-06, "loss": 1.2131, "step": 17954 }, { "epoch": 1.4071316614420062, "grad_norm": 2.303241729736328, "learning_rate": 7.262806015799733e-06, "loss": 1.3981, "step": 17955 }, { "epoch": 1.4072100313479625, "grad_norm": 3.8909599781036377, "learning_rate": 7.262531034766802e-06, "loss": 0.7131, "step": 17956 }, { "epoch": 1.4072884012539184, "grad_norm": 2.4100797176361084, "learning_rate": 7.262256045128363e-06, "loss": 1.3581, "step": 17957 }, { "epoch": 1.4073667711598747, "grad_norm": 2.111172914505005, "learning_rate": 7.261981046885465e-06, "loss": 0.5838, "step": 17958 }, { "epoch": 1.4074451410658306, "grad_norm": 3.4856975078582764, "learning_rate": 7.2617060400391534e-06, "loss": 1.6936, "step": 17959 }, { "epoch": 1.407523510971787, "grad_norm": 2.8245253562927246, "learning_rate": 7.261431024590472e-06, "loss": 1.1671, "step": 17960 }, { "epoch": 1.4076018808777429, "grad_norm": 2.3564271926879883, "learning_rate": 7.26115600054047e-06, "loss": 1.3788, "step": 17961 }, { "epoch": 1.4076802507836992, "grad_norm": 2.84220552444458, "learning_rate": 7.2608809678901904e-06, "loss": 1.1133, "step": 17962 }, { "epoch": 1.407758620689655, "grad_norm": 3.364871025085449, "learning_rate": 7.260605926640682e-06, "loss": 1.3659, "step": 17963 }, { "epoch": 1.4078369905956114, "grad_norm": 3.98884916305542, "learning_rate": 7.260330876792989e-06, "loss": 0.9967, "step": 17964 }, { "epoch": 1.4079153605015673, "grad_norm": 3.1504576206207275, "learning_rate": 7.260055818348159e-06, "loss": 1.7031, "step": 17965 }, { "epoch": 1.4079937304075236, "grad_norm": 3.5202765464782715, "learning_rate": 7.259780751307238e-06, "loss": 1.6555, "step": 17966 }, { "epoch": 1.4080721003134795, "grad_norm": 2.5883471965789795, "learning_rate": 7.259505675671271e-06, "loss": 1.4017, "step": 17967 }, { "epoch": 1.4081504702194358, "grad_norm": 5.251824855804443, "learning_rate": 7.2592305914413054e-06, "loss": 1.4023, "step": 17968 }, { "epoch": 1.4082288401253917, "grad_norm": 3.016810894012451, "learning_rate": 7.258955498618386e-06, "loss": 1.3556, "step": 17969 }, { "epoch": 1.408307210031348, "grad_norm": 6.018458366394043, "learning_rate": 7.258680397203563e-06, "loss": 1.4165, "step": 17970 }, { "epoch": 1.408385579937304, "grad_norm": 2.6215991973876953, "learning_rate": 7.258405287197876e-06, "loss": 1.3404, "step": 17971 }, { "epoch": 1.4084639498432603, "grad_norm": 2.1780035495758057, "learning_rate": 7.258130168602378e-06, "loss": 1.3139, "step": 17972 }, { "epoch": 1.4085423197492162, "grad_norm": 3.4504354000091553, "learning_rate": 7.257855041418112e-06, "loss": 1.4881, "step": 17973 }, { "epoch": 1.4086206896551725, "grad_norm": 2.1735851764678955, "learning_rate": 7.2575799056461235e-06, "loss": 0.911, "step": 17974 }, { "epoch": 1.4086990595611284, "grad_norm": 2.0727643966674805, "learning_rate": 7.257304761287462e-06, "loss": 1.2345, "step": 17975 }, { "epoch": 1.4087774294670847, "grad_norm": 4.282217502593994, "learning_rate": 7.257029608343174e-06, "loss": 1.1906, "step": 17976 }, { "epoch": 1.4088557993730406, "grad_norm": 3.6737875938415527, "learning_rate": 7.256754446814303e-06, "loss": 1.7251, "step": 17977 }, { "epoch": 1.408934169278997, "grad_norm": 6.357816696166992, "learning_rate": 7.256479276701897e-06, "loss": 1.2066, "step": 17978 }, { "epoch": 1.409012539184953, "grad_norm": 2.318035125732422, "learning_rate": 7.256204098007003e-06, "loss": 1.1967, "step": 17979 }, { "epoch": 1.4090909090909092, "grad_norm": 7.31436824798584, "learning_rate": 7.255928910730669e-06, "loss": 0.7418, "step": 17980 }, { "epoch": 1.4091692789968653, "grad_norm": 2.7597358226776123, "learning_rate": 7.255653714873938e-06, "loss": 1.3469, "step": 17981 }, { "epoch": 1.4092476489028214, "grad_norm": 3.569566011428833, "learning_rate": 7.25537851043786e-06, "loss": 1.4909, "step": 17982 }, { "epoch": 1.4093260188087775, "grad_norm": 4.3416218757629395, "learning_rate": 7.255103297423481e-06, "loss": 0.9226, "step": 17983 }, { "epoch": 1.4094043887147336, "grad_norm": 5.253042221069336, "learning_rate": 7.2548280758318455e-06, "loss": 1.6377, "step": 17984 }, { "epoch": 1.4094827586206897, "grad_norm": 4.942294597625732, "learning_rate": 7.254552845664004e-06, "loss": 1.7684, "step": 17985 }, { "epoch": 1.4095611285266458, "grad_norm": 3.747781276702881, "learning_rate": 7.2542776069210005e-06, "loss": 1.1321, "step": 17986 }, { "epoch": 1.409639498432602, "grad_norm": 3.7750775814056396, "learning_rate": 7.254002359603885e-06, "loss": 1.6947, "step": 17987 }, { "epoch": 1.409717868338558, "grad_norm": 2.2393691539764404, "learning_rate": 7.253727103713699e-06, "loss": 1.0733, "step": 17988 }, { "epoch": 1.4097962382445142, "grad_norm": 3.054581880569458, "learning_rate": 7.2534518392514955e-06, "loss": 1.4481, "step": 17989 }, { "epoch": 1.4098746081504703, "grad_norm": 5.0391316413879395, "learning_rate": 7.253176566218315e-06, "loss": 1.3988, "step": 17990 }, { "epoch": 1.4099529780564264, "grad_norm": 6.479074478149414, "learning_rate": 7.252901284615212e-06, "loss": 1.6687, "step": 17991 }, { "epoch": 1.4100313479623825, "grad_norm": 3.1198418140411377, "learning_rate": 7.252625994443228e-06, "loss": 1.3292, "step": 17992 }, { "epoch": 1.4101097178683386, "grad_norm": 9.16590690612793, "learning_rate": 7.2523506957034115e-06, "loss": 0.7333, "step": 17993 }, { "epoch": 1.4101880877742947, "grad_norm": 2.3975603580474854, "learning_rate": 7.252075388396811e-06, "loss": 1.4242, "step": 17994 }, { "epoch": 1.4102664576802508, "grad_norm": 9.792996406555176, "learning_rate": 7.251800072524471e-06, "loss": 1.5788, "step": 17995 }, { "epoch": 1.410344827586207, "grad_norm": 2.2124710083007812, "learning_rate": 7.251524748087442e-06, "loss": 1.4301, "step": 17996 }, { "epoch": 1.410423197492163, "grad_norm": 2.4823880195617676, "learning_rate": 7.251249415086766e-06, "loss": 1.3921, "step": 17997 }, { "epoch": 1.4105015673981192, "grad_norm": 3.9338858127593994, "learning_rate": 7.250974073523497e-06, "loss": 1.2395, "step": 17998 }, { "epoch": 1.4105799373040753, "grad_norm": 2.7091917991638184, "learning_rate": 7.250698723398677e-06, "loss": 1.3235, "step": 17999 }, { "epoch": 1.4106583072100314, "grad_norm": 3.024059534072876, "learning_rate": 7.250423364713356e-06, "loss": 1.1101, "step": 18000 }, { "epoch": 1.4107366771159875, "grad_norm": 3.1550347805023193, "learning_rate": 7.250147997468579e-06, "loss": 1.2734, "step": 18001 }, { "epoch": 1.4108150470219436, "grad_norm": 5.237597942352295, "learning_rate": 7.249872621665397e-06, "loss": 1.4482, "step": 18002 }, { "epoch": 1.4108934169278997, "grad_norm": 2.5884268283843994, "learning_rate": 7.249597237304853e-06, "loss": 1.2667, "step": 18003 }, { "epoch": 1.4109717868338558, "grad_norm": 5.272827625274658, "learning_rate": 7.249321844387997e-06, "loss": 0.8425, "step": 18004 }, { "epoch": 1.411050156739812, "grad_norm": 6.467191696166992, "learning_rate": 7.2490464429158765e-06, "loss": 1.1957, "step": 18005 }, { "epoch": 1.411128526645768, "grad_norm": 2.6990902423858643, "learning_rate": 7.24877103288954e-06, "loss": 1.8893, "step": 18006 }, { "epoch": 1.4112068965517242, "grad_norm": 3.186929702758789, "learning_rate": 7.248495614310031e-06, "loss": 1.6253, "step": 18007 }, { "epoch": 1.4112852664576803, "grad_norm": 7.586770534515381, "learning_rate": 7.248220187178401e-06, "loss": 1.1179, "step": 18008 }, { "epoch": 1.4113636363636364, "grad_norm": 3.7150750160217285, "learning_rate": 7.247944751495695e-06, "loss": 1.5166, "step": 18009 }, { "epoch": 1.4114420062695925, "grad_norm": 2.610682249069214, "learning_rate": 7.247669307262964e-06, "loss": 1.6013, "step": 18010 }, { "epoch": 1.4115203761755486, "grad_norm": 2.4405295848846436, "learning_rate": 7.2473938544812505e-06, "loss": 1.2512, "step": 18011 }, { "epoch": 1.4115987460815047, "grad_norm": 2.792278289794922, "learning_rate": 7.247118393151607e-06, "loss": 1.0938, "step": 18012 }, { "epoch": 1.4116771159874608, "grad_norm": 2.640139102935791, "learning_rate": 7.246842923275079e-06, "loss": 1.2741, "step": 18013 }, { "epoch": 1.411755485893417, "grad_norm": 2.8689029216766357, "learning_rate": 7.246567444852714e-06, "loss": 1.0596, "step": 18014 }, { "epoch": 1.411833855799373, "grad_norm": 4.854043006896973, "learning_rate": 7.246291957885562e-06, "loss": 1.4388, "step": 18015 }, { "epoch": 1.4119122257053291, "grad_norm": 2.5972492694854736, "learning_rate": 7.246016462374667e-06, "loss": 1.3145, "step": 18016 }, { "epoch": 1.4119905956112853, "grad_norm": 3.149155855178833, "learning_rate": 7.245740958321081e-06, "loss": 1.2143, "step": 18017 }, { "epoch": 1.4120689655172414, "grad_norm": 3.3959484100341797, "learning_rate": 7.245465445725849e-06, "loss": 2.1251, "step": 18018 }, { "epoch": 1.4121473354231975, "grad_norm": 2.33990216255188, "learning_rate": 7.245189924590019e-06, "loss": 0.8687, "step": 18019 }, { "epoch": 1.4122257053291536, "grad_norm": 2.6099624633789062, "learning_rate": 7.244914394914641e-06, "loss": 1.2699, "step": 18020 }, { "epoch": 1.4123040752351097, "grad_norm": 4.874489784240723, "learning_rate": 7.244638856700761e-06, "loss": 1.1409, "step": 18021 }, { "epoch": 1.4123824451410658, "grad_norm": 3.740678548812866, "learning_rate": 7.244363309949427e-06, "loss": 1.5191, "step": 18022 }, { "epoch": 1.412460815047022, "grad_norm": 2.765127658843994, "learning_rate": 7.244087754661689e-06, "loss": 1.509, "step": 18023 }, { "epoch": 1.412539184952978, "grad_norm": 2.1360700130462646, "learning_rate": 7.2438121908385935e-06, "loss": 1.4121, "step": 18024 }, { "epoch": 1.4126175548589341, "grad_norm": 6.118231773376465, "learning_rate": 7.243536618481188e-06, "loss": 0.9065, "step": 18025 }, { "epoch": 1.4126959247648903, "grad_norm": 2.6420249938964844, "learning_rate": 7.243261037590523e-06, "loss": 1.4037, "step": 18026 }, { "epoch": 1.4127742946708464, "grad_norm": 2.919883966445923, "learning_rate": 7.242985448167645e-06, "loss": 1.323, "step": 18027 }, { "epoch": 1.4128526645768025, "grad_norm": 2.4658310413360596, "learning_rate": 7.2427098502136015e-06, "loss": 1.6882, "step": 18028 }, { "epoch": 1.4129310344827586, "grad_norm": 4.420916557312012, "learning_rate": 7.242434243729443e-06, "loss": 0.8526, "step": 18029 }, { "epoch": 1.4130094043887147, "grad_norm": 2.4759278297424316, "learning_rate": 7.242158628716215e-06, "loss": 1.4179, "step": 18030 }, { "epoch": 1.4130877742946708, "grad_norm": 2.4193339347839355, "learning_rate": 7.2418830051749675e-06, "loss": 1.538, "step": 18031 }, { "epoch": 1.413166144200627, "grad_norm": 10.441139221191406, "learning_rate": 7.24160737310675e-06, "loss": 1.3177, "step": 18032 }, { "epoch": 1.413244514106583, "grad_norm": 3.345750093460083, "learning_rate": 7.2413317325126065e-06, "loss": 1.4875, "step": 18033 }, { "epoch": 1.4133228840125391, "grad_norm": 4.390337944030762, "learning_rate": 7.241056083393592e-06, "loss": 1.4306, "step": 18034 }, { "epoch": 1.4134012539184952, "grad_norm": 2.2457826137542725, "learning_rate": 7.240780425750747e-06, "loss": 1.3159, "step": 18035 }, { "epoch": 1.4134796238244514, "grad_norm": 5.571081638336182, "learning_rate": 7.240504759585127e-06, "loss": 0.8691, "step": 18036 }, { "epoch": 1.4135579937304075, "grad_norm": 2.567887544631958, "learning_rate": 7.2402290848977785e-06, "loss": 1.3036, "step": 18037 }, { "epoch": 1.4136363636363636, "grad_norm": 3.6718595027923584, "learning_rate": 7.239953401689747e-06, "loss": 1.5028, "step": 18038 }, { "epoch": 1.4137147335423197, "grad_norm": 5.276634693145752, "learning_rate": 7.239677709962085e-06, "loss": 1.5461, "step": 18039 }, { "epoch": 1.4137931034482758, "grad_norm": 3.3832664489746094, "learning_rate": 7.239402009715838e-06, "loss": 1.6988, "step": 18040 }, { "epoch": 1.413871473354232, "grad_norm": 2.6811649799346924, "learning_rate": 7.239126300952057e-06, "loss": 1.0534, "step": 18041 }, { "epoch": 1.413949843260188, "grad_norm": 3.1481516361236572, "learning_rate": 7.238850583671788e-06, "loss": 1.2026, "step": 18042 }, { "epoch": 1.4140282131661441, "grad_norm": 2.6292247772216797, "learning_rate": 7.238574857876084e-06, "loss": 1.3133, "step": 18043 }, { "epoch": 1.4141065830721002, "grad_norm": 6.041173934936523, "learning_rate": 7.238299123565988e-06, "loss": 1.4216, "step": 18044 }, { "epoch": 1.4141849529780564, "grad_norm": 3.6036314964294434, "learning_rate": 7.2380233807425545e-06, "loss": 1.3545, "step": 18045 }, { "epoch": 1.4142633228840125, "grad_norm": 2.6449592113494873, "learning_rate": 7.237747629406828e-06, "loss": 1.1147, "step": 18046 }, { "epoch": 1.4143416927899686, "grad_norm": 1.8615261316299438, "learning_rate": 7.23747186955986e-06, "loss": 0.9815, "step": 18047 }, { "epoch": 1.4144200626959247, "grad_norm": 2.876506805419922, "learning_rate": 7.237196101202697e-06, "loss": 1.4276, "step": 18048 }, { "epoch": 1.4144984326018808, "grad_norm": 2.425313949584961, "learning_rate": 7.2369203243363895e-06, "loss": 1.3796, "step": 18049 }, { "epoch": 1.414576802507837, "grad_norm": 3.7408878803253174, "learning_rate": 7.236644538961986e-06, "loss": 1.2875, "step": 18050 }, { "epoch": 1.4146551724137932, "grad_norm": 4.2184977531433105, "learning_rate": 7.236368745080536e-06, "loss": 1.6542, "step": 18051 }, { "epoch": 1.4147335423197491, "grad_norm": 2.2191083431243896, "learning_rate": 7.236092942693089e-06, "loss": 1.3393, "step": 18052 }, { "epoch": 1.4148119122257055, "grad_norm": 2.3285324573516846, "learning_rate": 7.235817131800691e-06, "loss": 1.2389, "step": 18053 }, { "epoch": 1.4148902821316613, "grad_norm": 2.9138600826263428, "learning_rate": 7.235541312404394e-06, "loss": 1.1758, "step": 18054 }, { "epoch": 1.4149686520376177, "grad_norm": 4.06806755065918, "learning_rate": 7.235265484505245e-06, "loss": 1.1773, "step": 18055 }, { "epoch": 1.4150470219435736, "grad_norm": 2.7483768463134766, "learning_rate": 7.234989648104296e-06, "loss": 1.745, "step": 18056 }, { "epoch": 1.41512539184953, "grad_norm": 2.8702945709228516, "learning_rate": 7.234713803202593e-06, "loss": 1.1063, "step": 18057 }, { "epoch": 1.4152037617554858, "grad_norm": 2.1793148517608643, "learning_rate": 7.2344379498011865e-06, "loss": 1.1717, "step": 18058 }, { "epoch": 1.4152821316614421, "grad_norm": 15.431866645812988, "learning_rate": 7.2341620879011265e-06, "loss": 1.8343, "step": 18059 }, { "epoch": 1.415360501567398, "grad_norm": 4.919205188751221, "learning_rate": 7.233886217503461e-06, "loss": 1.9488, "step": 18060 }, { "epoch": 1.4154388714733543, "grad_norm": 3.5306477546691895, "learning_rate": 7.233610338609238e-06, "loss": 1.343, "step": 18061 }, { "epoch": 1.4155172413793102, "grad_norm": 2.693544626235962, "learning_rate": 7.2333344512195115e-06, "loss": 1.2186, "step": 18062 }, { "epoch": 1.4155956112852666, "grad_norm": 3.9336612224578857, "learning_rate": 7.233058555335326e-06, "loss": 1.2603, "step": 18063 }, { "epoch": 1.4156739811912225, "grad_norm": 1.9892127513885498, "learning_rate": 7.232782650957735e-06, "loss": 1.2207, "step": 18064 }, { "epoch": 1.4157523510971788, "grad_norm": 2.948671817779541, "learning_rate": 7.232506738087783e-06, "loss": 1.1812, "step": 18065 }, { "epoch": 1.4158307210031347, "grad_norm": 3.613020658493042, "learning_rate": 7.232230816726523e-06, "loss": 0.8936, "step": 18066 }, { "epoch": 1.415909090909091, "grad_norm": 2.42295241355896, "learning_rate": 7.231954886875004e-06, "loss": 1.0051, "step": 18067 }, { "epoch": 1.415987460815047, "grad_norm": 5.622279167175293, "learning_rate": 7.2316789485342745e-06, "loss": 1.1985, "step": 18068 }, { "epoch": 1.4160658307210032, "grad_norm": 2.5005738735198975, "learning_rate": 7.231403001705385e-06, "loss": 1.2566, "step": 18069 }, { "epoch": 1.4161442006269591, "grad_norm": 6.102827072143555, "learning_rate": 7.231127046389384e-06, "loss": 0.6093, "step": 18070 }, { "epoch": 1.4162225705329154, "grad_norm": 2.6616320610046387, "learning_rate": 7.230851082587322e-06, "loss": 1.5573, "step": 18071 }, { "epoch": 1.4163009404388713, "grad_norm": 4.5524725914001465, "learning_rate": 7.230575110300249e-06, "loss": 1.5786, "step": 18072 }, { "epoch": 1.4163793103448277, "grad_norm": 2.0436086654663086, "learning_rate": 7.230299129529212e-06, "loss": 1.2838, "step": 18073 }, { "epoch": 1.4164576802507838, "grad_norm": 4.06051778793335, "learning_rate": 7.2300231402752644e-06, "loss": 0.9007, "step": 18074 }, { "epoch": 1.41653605015674, "grad_norm": 2.6334636211395264, "learning_rate": 7.229747142539454e-06, "loss": 1.5535, "step": 18075 }, { "epoch": 1.416614420062696, "grad_norm": 6.651711940765381, "learning_rate": 7.22947113632283e-06, "loss": 1.2213, "step": 18076 }, { "epoch": 1.416692789968652, "grad_norm": 2.4302279949188232, "learning_rate": 7.2291951216264424e-06, "loss": 1.3992, "step": 18077 }, { "epoch": 1.4167711598746082, "grad_norm": 3.990424633026123, "learning_rate": 7.228919098451342e-06, "loss": 1.4385, "step": 18078 }, { "epoch": 1.4168495297805643, "grad_norm": 3.727975845336914, "learning_rate": 7.22864306679858e-06, "loss": 1.2927, "step": 18079 }, { "epoch": 1.4169278996865204, "grad_norm": 3.095914125442505, "learning_rate": 7.228367026669203e-06, "loss": 1.2788, "step": 18080 }, { "epoch": 1.4170062695924766, "grad_norm": 4.053891658782959, "learning_rate": 7.228090978064263e-06, "loss": 1.7519, "step": 18081 }, { "epoch": 1.4170846394984327, "grad_norm": 2.3777849674224854, "learning_rate": 7.227814920984808e-06, "loss": 1.4072, "step": 18082 }, { "epoch": 1.4171630094043888, "grad_norm": 3.1233086585998535, "learning_rate": 7.22753885543189e-06, "loss": 1.4558, "step": 18083 }, { "epoch": 1.4172413793103449, "grad_norm": 2.6955323219299316, "learning_rate": 7.227262781406559e-06, "loss": 1.1989, "step": 18084 }, { "epoch": 1.417319749216301, "grad_norm": 2.3226120471954346, "learning_rate": 7.2269866989098636e-06, "loss": 1.3827, "step": 18085 }, { "epoch": 1.417398119122257, "grad_norm": 4.294455051422119, "learning_rate": 7.226710607942855e-06, "loss": 1.4839, "step": 18086 }, { "epoch": 1.4174764890282132, "grad_norm": 2.2420737743377686, "learning_rate": 7.2264345085065835e-06, "loss": 1.4199, "step": 18087 }, { "epoch": 1.4175548589341693, "grad_norm": 2.7631921768188477, "learning_rate": 7.226158400602099e-06, "loss": 1.3804, "step": 18088 }, { "epoch": 1.4176332288401254, "grad_norm": 3.4584462642669678, "learning_rate": 7.225882284230449e-06, "loss": 1.0033, "step": 18089 }, { "epoch": 1.4177115987460815, "grad_norm": 3.3216664791107178, "learning_rate": 7.225606159392689e-06, "loss": 2.0179, "step": 18090 }, { "epoch": 1.4177899686520377, "grad_norm": 2.1277973651885986, "learning_rate": 7.225330026089863e-06, "loss": 1.384, "step": 18091 }, { "epoch": 1.4178683385579938, "grad_norm": 4.670810222625732, "learning_rate": 7.2250538843230275e-06, "loss": 1.4506, "step": 18092 }, { "epoch": 1.4179467084639499, "grad_norm": 3.754765510559082, "learning_rate": 7.224777734093229e-06, "loss": 1.8437, "step": 18093 }, { "epoch": 1.418025078369906, "grad_norm": 3.4271204471588135, "learning_rate": 7.224501575401518e-06, "loss": 1.5188, "step": 18094 }, { "epoch": 1.418103448275862, "grad_norm": 2.755134344100952, "learning_rate": 7.224225408248947e-06, "loss": 1.219, "step": 18095 }, { "epoch": 1.4181818181818182, "grad_norm": 2.238398551940918, "learning_rate": 7.223949232636563e-06, "loss": 1.1757, "step": 18096 }, { "epoch": 1.4182601880877743, "grad_norm": 2.776031255722046, "learning_rate": 7.22367304856542e-06, "loss": 1.428, "step": 18097 }, { "epoch": 1.4183385579937304, "grad_norm": 2.9208879470825195, "learning_rate": 7.223396856036566e-06, "loss": 1.3511, "step": 18098 }, { "epoch": 1.4184169278996865, "grad_norm": 5.13970947265625, "learning_rate": 7.223120655051054e-06, "loss": 1.4885, "step": 18099 }, { "epoch": 1.4184952978056427, "grad_norm": 3.4539377689361572, "learning_rate": 7.222844445609931e-06, "loss": 1.4153, "step": 18100 }, { "epoch": 1.4185736677115988, "grad_norm": 2.4030303955078125, "learning_rate": 7.2225682277142505e-06, "loss": 1.66, "step": 18101 }, { "epoch": 1.4186520376175549, "grad_norm": 2.040381908416748, "learning_rate": 7.2222920013650614e-06, "loss": 1.4411, "step": 18102 }, { "epoch": 1.418730407523511, "grad_norm": 2.446645498275757, "learning_rate": 7.222015766563415e-06, "loss": 1.4905, "step": 18103 }, { "epoch": 1.418808777429467, "grad_norm": 2.2797698974609375, "learning_rate": 7.2217395233103615e-06, "loss": 1.0755, "step": 18104 }, { "epoch": 1.4188871473354232, "grad_norm": 2.7039291858673096, "learning_rate": 7.221463271606953e-06, "loss": 1.2091, "step": 18105 }, { "epoch": 1.4189655172413793, "grad_norm": 2.987555742263794, "learning_rate": 7.221187011454238e-06, "loss": 1.4401, "step": 18106 }, { "epoch": 1.4190438871473354, "grad_norm": 2.304945230484009, "learning_rate": 7.220910742853271e-06, "loss": 1.3949, "step": 18107 }, { "epoch": 1.4191222570532915, "grad_norm": 2.3131821155548096, "learning_rate": 7.220634465805099e-06, "loss": 1.3825, "step": 18108 }, { "epoch": 1.4192006269592476, "grad_norm": 2.4975497722625732, "learning_rate": 7.220358180310774e-06, "loss": 1.1333, "step": 18109 }, { "epoch": 1.4192789968652038, "grad_norm": 4.199270725250244, "learning_rate": 7.220081886371346e-06, "loss": 1.2347, "step": 18110 }, { "epoch": 1.4193573667711599, "grad_norm": 8.989749908447266, "learning_rate": 7.219805583987867e-06, "loss": 0.9197, "step": 18111 }, { "epoch": 1.419435736677116, "grad_norm": 7.03555154800415, "learning_rate": 7.2195292731613884e-06, "loss": 1.7818, "step": 18112 }, { "epoch": 1.419514106583072, "grad_norm": 2.9706411361694336, "learning_rate": 7.21925295389296e-06, "loss": 1.2221, "step": 18113 }, { "epoch": 1.4195924764890282, "grad_norm": 2.677168369293213, "learning_rate": 7.218976626183635e-06, "loss": 1.3853, "step": 18114 }, { "epoch": 1.4196708463949843, "grad_norm": 2.323982000350952, "learning_rate": 7.218700290034459e-06, "loss": 1.4651, "step": 18115 }, { "epoch": 1.4197492163009404, "grad_norm": 2.292867660522461, "learning_rate": 7.21842394544649e-06, "loss": 1.3482, "step": 18116 }, { "epoch": 1.4198275862068965, "grad_norm": 1.9554364681243896, "learning_rate": 7.218147592420773e-06, "loss": 1.3768, "step": 18117 }, { "epoch": 1.4199059561128526, "grad_norm": 5.245334148406982, "learning_rate": 7.217871230958365e-06, "loss": 1.1951, "step": 18118 }, { "epoch": 1.4199843260188088, "grad_norm": 2.6623125076293945, "learning_rate": 7.217594861060313e-06, "loss": 1.103, "step": 18119 }, { "epoch": 1.4200626959247649, "grad_norm": 3.285984992980957, "learning_rate": 7.2173184827276685e-06, "loss": 1.2233, "step": 18120 }, { "epoch": 1.420141065830721, "grad_norm": 5.122029781341553, "learning_rate": 7.217042095961483e-06, "loss": 1.2405, "step": 18121 }, { "epoch": 1.420219435736677, "grad_norm": 2.397890329360962, "learning_rate": 7.216765700762809e-06, "loss": 1.4522, "step": 18122 }, { "epoch": 1.4202978056426332, "grad_norm": 3.825244188308716, "learning_rate": 7.216489297132697e-06, "loss": 1.4048, "step": 18123 }, { "epoch": 1.4203761755485893, "grad_norm": 3.5702545642852783, "learning_rate": 7.2162128850721965e-06, "loss": 1.1859, "step": 18124 }, { "epoch": 1.4204545454545454, "grad_norm": 4.236247539520264, "learning_rate": 7.215936464582363e-06, "loss": 1.2283, "step": 18125 }, { "epoch": 1.4205329153605015, "grad_norm": 3.073535442352295, "learning_rate": 7.2156600356642434e-06, "loss": 1.5173, "step": 18126 }, { "epoch": 1.4206112852664576, "grad_norm": 3.035749673843384, "learning_rate": 7.215383598318892e-06, "loss": 1.5279, "step": 18127 }, { "epoch": 1.4206896551724137, "grad_norm": 18.609663009643555, "learning_rate": 7.215107152547359e-06, "loss": 3.1889, "step": 18128 }, { "epoch": 1.4207680250783699, "grad_norm": 3.889690637588501, "learning_rate": 7.214830698350696e-06, "loss": 1.1509, "step": 18129 }, { "epoch": 1.420846394984326, "grad_norm": 3.0339131355285645, "learning_rate": 7.214554235729955e-06, "loss": 1.1465, "step": 18130 }, { "epoch": 1.420924764890282, "grad_norm": 1.9211891889572144, "learning_rate": 7.214277764686186e-06, "loss": 1.2508, "step": 18131 }, { "epoch": 1.4210031347962382, "grad_norm": 2.5913946628570557, "learning_rate": 7.214001285220443e-06, "loss": 1.2749, "step": 18132 }, { "epoch": 1.4210815047021943, "grad_norm": 3.8977465629577637, "learning_rate": 7.213724797333777e-06, "loss": 1.2163, "step": 18133 }, { "epoch": 1.4211598746081504, "grad_norm": 3.6303229331970215, "learning_rate": 7.213448301027237e-06, "loss": 1.5352, "step": 18134 }, { "epoch": 1.4212382445141065, "grad_norm": 2.528064489364624, "learning_rate": 7.213171796301878e-06, "loss": 1.4323, "step": 18135 }, { "epoch": 1.4213166144200626, "grad_norm": 6.380125045776367, "learning_rate": 7.212895283158749e-06, "loss": 1.5937, "step": 18136 }, { "epoch": 1.4213949843260187, "grad_norm": 4.376086711883545, "learning_rate": 7.212618761598904e-06, "loss": 1.009, "step": 18137 }, { "epoch": 1.4214733542319749, "grad_norm": 2.3990633487701416, "learning_rate": 7.212342231623394e-06, "loss": 1.5121, "step": 18138 }, { "epoch": 1.421551724137931, "grad_norm": 3.6904549598693848, "learning_rate": 7.2120656932332685e-06, "loss": 1.4815, "step": 18139 }, { "epoch": 1.421630094043887, "grad_norm": 6.22389030456543, "learning_rate": 7.211789146429583e-06, "loss": 1.4819, "step": 18140 }, { "epoch": 1.4217084639498432, "grad_norm": 1.9273127317428589, "learning_rate": 7.211512591213388e-06, "loss": 1.2881, "step": 18141 }, { "epoch": 1.4217868338557993, "grad_norm": 2.702092170715332, "learning_rate": 7.211236027585734e-06, "loss": 1.3349, "step": 18142 }, { "epoch": 1.4218652037617554, "grad_norm": 3.6602580547332764, "learning_rate": 7.210959455547674e-06, "loss": 1.1998, "step": 18143 }, { "epoch": 1.4219435736677117, "grad_norm": 2.587568998336792, "learning_rate": 7.21068287510026e-06, "loss": 1.5001, "step": 18144 }, { "epoch": 1.4220219435736676, "grad_norm": 2.4029955863952637, "learning_rate": 7.210406286244543e-06, "loss": 0.7686, "step": 18145 }, { "epoch": 1.422100313479624, "grad_norm": 3.3532321453094482, "learning_rate": 7.210129688981577e-06, "loss": 1.4875, "step": 18146 }, { "epoch": 1.4221786833855798, "grad_norm": 4.623039245605469, "learning_rate": 7.209853083312412e-06, "loss": 1.4748, "step": 18147 }, { "epoch": 1.4222570532915362, "grad_norm": 3.2142086029052734, "learning_rate": 7.209576469238102e-06, "loss": 1.3958, "step": 18148 }, { "epoch": 1.422335423197492, "grad_norm": 2.9566409587860107, "learning_rate": 7.209299846759697e-06, "loss": 1.4473, "step": 18149 }, { "epoch": 1.4224137931034484, "grad_norm": 3.1402788162231445, "learning_rate": 7.209023215878251e-06, "loss": 1.4307, "step": 18150 }, { "epoch": 1.4224921630094043, "grad_norm": 5.870482921600342, "learning_rate": 7.208746576594816e-06, "loss": 1.2045, "step": 18151 }, { "epoch": 1.4225705329153606, "grad_norm": 3.372614622116089, "learning_rate": 7.208469928910442e-06, "loss": 1.1078, "step": 18152 }, { "epoch": 1.4226489028213165, "grad_norm": 3.0756301879882812, "learning_rate": 7.208193272826184e-06, "loss": 1.442, "step": 18153 }, { "epoch": 1.4227272727272728, "grad_norm": 2.73868989944458, "learning_rate": 7.207916608343092e-06, "loss": 1.2018, "step": 18154 }, { "epoch": 1.4228056426332287, "grad_norm": 1.9128719568252563, "learning_rate": 7.207639935462218e-06, "loss": 1.3829, "step": 18155 }, { "epoch": 1.422884012539185, "grad_norm": 4.570829391479492, "learning_rate": 7.207363254184618e-06, "loss": 1.6573, "step": 18156 }, { "epoch": 1.422962382445141, "grad_norm": 2.545370101928711, "learning_rate": 7.207086564511341e-06, "loss": 1.1337, "step": 18157 }, { "epoch": 1.4230407523510973, "grad_norm": 2.606994390487671, "learning_rate": 7.2068098664434406e-06, "loss": 1.4021, "step": 18158 }, { "epoch": 1.4231191222570532, "grad_norm": 4.260374546051025, "learning_rate": 7.206533159981968e-06, "loss": 1.9167, "step": 18159 }, { "epoch": 1.4231974921630095, "grad_norm": 2.078916072845459, "learning_rate": 7.206256445127977e-06, "loss": 1.1539, "step": 18160 }, { "epoch": 1.4232758620689654, "grad_norm": 5.00300407409668, "learning_rate": 7.205979721882521e-06, "loss": 1.4037, "step": 18161 }, { "epoch": 1.4233542319749217, "grad_norm": 3.120375156402588, "learning_rate": 7.205702990246649e-06, "loss": 1.3659, "step": 18162 }, { "epoch": 1.4234326018808776, "grad_norm": 3.4452121257781982, "learning_rate": 7.205426250221418e-06, "loss": 1.2721, "step": 18163 }, { "epoch": 1.423510971786834, "grad_norm": 3.8510286808013916, "learning_rate": 7.205149501807878e-06, "loss": 1.24, "step": 18164 }, { "epoch": 1.4235893416927898, "grad_norm": 3.2185542583465576, "learning_rate": 7.20487274500708e-06, "loss": 1.4825, "step": 18165 }, { "epoch": 1.4236677115987462, "grad_norm": 5.811490535736084, "learning_rate": 7.204595979820079e-06, "loss": 1.4608, "step": 18166 }, { "epoch": 1.4237460815047023, "grad_norm": 2.605017900466919, "learning_rate": 7.2043192062479284e-06, "loss": 1.2242, "step": 18167 }, { "epoch": 1.4238244514106584, "grad_norm": 2.867433786392212, "learning_rate": 7.204042424291681e-06, "loss": 1.3438, "step": 18168 }, { "epoch": 1.4239028213166145, "grad_norm": 2.088287353515625, "learning_rate": 7.203765633952386e-06, "loss": 1.4183, "step": 18169 }, { "epoch": 1.4239811912225706, "grad_norm": 2.8890862464904785, "learning_rate": 7.203488835231099e-06, "loss": 1.5948, "step": 18170 }, { "epoch": 1.4240595611285267, "grad_norm": 2.2024102210998535, "learning_rate": 7.203212028128871e-06, "loss": 1.5304, "step": 18171 }, { "epoch": 1.4241379310344828, "grad_norm": 2.858185052871704, "learning_rate": 7.202935212646759e-06, "loss": 1.1611, "step": 18172 }, { "epoch": 1.424216300940439, "grad_norm": 17.586164474487305, "learning_rate": 7.20265838878581e-06, "loss": 0.7952, "step": 18173 }, { "epoch": 1.424294670846395, "grad_norm": 8.193583488464355, "learning_rate": 7.202381556547082e-06, "loss": 1.4566, "step": 18174 }, { "epoch": 1.4243730407523512, "grad_norm": 2.4792490005493164, "learning_rate": 7.202104715931625e-06, "loss": 1.4356, "step": 18175 }, { "epoch": 1.4244514106583073, "grad_norm": 2.7895028591156006, "learning_rate": 7.201827866940493e-06, "loss": 1.3637, "step": 18176 }, { "epoch": 1.4245297805642634, "grad_norm": 2.3847694396972656, "learning_rate": 7.2015510095747375e-06, "loss": 1.4617, "step": 18177 }, { "epoch": 1.4246081504702195, "grad_norm": 4.456268310546875, "learning_rate": 7.201274143835414e-06, "loss": 1.7983, "step": 18178 }, { "epoch": 1.4246865203761756, "grad_norm": 3.3367934226989746, "learning_rate": 7.200997269723576e-06, "loss": 1.2567, "step": 18179 }, { "epoch": 1.4247648902821317, "grad_norm": 2.7658143043518066, "learning_rate": 7.200720387240272e-06, "loss": 1.0284, "step": 18180 }, { "epoch": 1.4248432601880878, "grad_norm": 3.433835983276367, "learning_rate": 7.200443496386559e-06, "loss": 1.685, "step": 18181 }, { "epoch": 1.424921630094044, "grad_norm": 5.123569011688232, "learning_rate": 7.200166597163488e-06, "loss": 0.9613, "step": 18182 }, { "epoch": 1.425, "grad_norm": 4.203614234924316, "learning_rate": 7.1998896895721145e-06, "loss": 1.4973, "step": 18183 }, { "epoch": 1.4250783699059562, "grad_norm": 2.571471929550171, "learning_rate": 7.19961277361349e-06, "loss": 1.4381, "step": 18184 }, { "epoch": 1.4251567398119123, "grad_norm": 2.4239354133605957, "learning_rate": 7.199335849288669e-06, "loss": 0.8808, "step": 18185 }, { "epoch": 1.4252351097178684, "grad_norm": 4.155561447143555, "learning_rate": 7.199058916598705e-06, "loss": 1.3481, "step": 18186 }, { "epoch": 1.4253134796238245, "grad_norm": 2.7115049362182617, "learning_rate": 7.198781975544649e-06, "loss": 1.335, "step": 18187 }, { "epoch": 1.4253918495297806, "grad_norm": 2.1526272296905518, "learning_rate": 7.198505026127555e-06, "loss": 1.227, "step": 18188 }, { "epoch": 1.4254702194357367, "grad_norm": 3.8531243801116943, "learning_rate": 7.198228068348478e-06, "loss": 1.6749, "step": 18189 }, { "epoch": 1.4255485893416928, "grad_norm": 3.8315863609313965, "learning_rate": 7.19795110220847e-06, "loss": 1.4433, "step": 18190 }, { "epoch": 1.425626959247649, "grad_norm": 4.7741379737854, "learning_rate": 7.197674127708587e-06, "loss": 1.6464, "step": 18191 }, { "epoch": 1.425705329153605, "grad_norm": 4.604709148406982, "learning_rate": 7.197397144849878e-06, "loss": 1.7121, "step": 18192 }, { "epoch": 1.4257836990595611, "grad_norm": 4.297584533691406, "learning_rate": 7.1971201536334e-06, "loss": 2.1364, "step": 18193 }, { "epoch": 1.4258620689655173, "grad_norm": 3.286268711090088, "learning_rate": 7.196843154060205e-06, "loss": 1.7047, "step": 18194 }, { "epoch": 1.4259404388714734, "grad_norm": 14.23165225982666, "learning_rate": 7.196566146131347e-06, "loss": 0.897, "step": 18195 }, { "epoch": 1.4260188087774295, "grad_norm": 3.4015183448791504, "learning_rate": 7.196289129847882e-06, "loss": 1.4843, "step": 18196 }, { "epoch": 1.4260971786833856, "grad_norm": 3.1049132347106934, "learning_rate": 7.196012105210857e-06, "loss": 1.1116, "step": 18197 }, { "epoch": 1.4261755485893417, "grad_norm": 2.930142879486084, "learning_rate": 7.195735072221334e-06, "loss": 1.4642, "step": 18198 }, { "epoch": 1.4262539184952978, "grad_norm": 3.005887508392334, "learning_rate": 7.195458030880359e-06, "loss": 1.328, "step": 18199 }, { "epoch": 1.426332288401254, "grad_norm": 3.283231496810913, "learning_rate": 7.195180981188991e-06, "loss": 1.4945, "step": 18200 }, { "epoch": 1.42641065830721, "grad_norm": 3.613881826400757, "learning_rate": 7.194903923148283e-06, "loss": 0.7308, "step": 18201 }, { "epoch": 1.4264890282131661, "grad_norm": 3.1052732467651367, "learning_rate": 7.194626856759287e-06, "loss": 1.3359, "step": 18202 }, { "epoch": 1.4265673981191223, "grad_norm": 2.388650894165039, "learning_rate": 7.1943497820230565e-06, "loss": 1.3007, "step": 18203 }, { "epoch": 1.4266457680250784, "grad_norm": 4.457905292510986, "learning_rate": 7.194072698940648e-06, "loss": 1.182, "step": 18204 }, { "epoch": 1.4267241379310345, "grad_norm": 2.1271727085113525, "learning_rate": 7.193795607513113e-06, "loss": 1.1923, "step": 18205 }, { "epoch": 1.4268025078369906, "grad_norm": 3.3097686767578125, "learning_rate": 7.193518507741507e-06, "loss": 1.3832, "step": 18206 }, { "epoch": 1.4268808777429467, "grad_norm": 2.5037693977355957, "learning_rate": 7.193241399626885e-06, "loss": 1.346, "step": 18207 }, { "epoch": 1.4269592476489028, "grad_norm": 2.8758368492126465, "learning_rate": 7.1929642831702965e-06, "loss": 1.4417, "step": 18208 }, { "epoch": 1.427037617554859, "grad_norm": 3.595674991607666, "learning_rate": 7.192687158372799e-06, "loss": 1.391, "step": 18209 }, { "epoch": 1.427115987460815, "grad_norm": 2.8275647163391113, "learning_rate": 7.192410025235446e-06, "loss": 1.6459, "step": 18210 }, { "epoch": 1.4271943573667711, "grad_norm": 2.9634947776794434, "learning_rate": 7.192132883759291e-06, "loss": 0.8737, "step": 18211 }, { "epoch": 1.4272727272727272, "grad_norm": 2.9419634342193604, "learning_rate": 7.191855733945388e-06, "loss": 1.2608, "step": 18212 }, { "epoch": 1.4273510971786834, "grad_norm": 2.8124256134033203, "learning_rate": 7.191578575794792e-06, "loss": 1.0716, "step": 18213 }, { "epoch": 1.4274294670846395, "grad_norm": 2.356177806854248, "learning_rate": 7.191301409308557e-06, "loss": 0.9561, "step": 18214 }, { "epoch": 1.4275078369905956, "grad_norm": 2.4044008255004883, "learning_rate": 7.1910242344877375e-06, "loss": 1.5226, "step": 18215 }, { "epoch": 1.4275862068965517, "grad_norm": 2.642965793609619, "learning_rate": 7.1907470513333855e-06, "loss": 1.5807, "step": 18216 }, { "epoch": 1.4276645768025078, "grad_norm": 3.4493930339813232, "learning_rate": 7.1904698598465595e-06, "loss": 1.5979, "step": 18217 }, { "epoch": 1.427742946708464, "grad_norm": 2.617266893386841, "learning_rate": 7.190192660028308e-06, "loss": 1.4173, "step": 18218 }, { "epoch": 1.42782131661442, "grad_norm": 2.4003686904907227, "learning_rate": 7.189915451879692e-06, "loss": 1.6087, "step": 18219 }, { "epoch": 1.4278996865203761, "grad_norm": 9.422929763793945, "learning_rate": 7.18963823540176e-06, "loss": 0.6905, "step": 18220 }, { "epoch": 1.4279780564263322, "grad_norm": 2.902784585952759, "learning_rate": 7.18936101059557e-06, "loss": 1.1459, "step": 18221 }, { "epoch": 1.4280564263322884, "grad_norm": 2.4211416244506836, "learning_rate": 7.189083777462176e-06, "loss": 1.294, "step": 18222 }, { "epoch": 1.4281347962382445, "grad_norm": 3.5545833110809326, "learning_rate": 7.18880653600263e-06, "loss": 1.5166, "step": 18223 }, { "epoch": 1.4282131661442006, "grad_norm": 5.119802474975586, "learning_rate": 7.188529286217989e-06, "loss": 1.7666, "step": 18224 }, { "epoch": 1.4282915360501567, "grad_norm": 2.714531898498535, "learning_rate": 7.188252028109305e-06, "loss": 1.4686, "step": 18225 }, { "epoch": 1.4283699059561128, "grad_norm": 3.030271291732788, "learning_rate": 7.187974761677636e-06, "loss": 1.2578, "step": 18226 }, { "epoch": 1.428448275862069, "grad_norm": 5.977012634277344, "learning_rate": 7.187697486924034e-06, "loss": 0.5779, "step": 18227 }, { "epoch": 1.428526645768025, "grad_norm": 1.858985185623169, "learning_rate": 7.187420203849555e-06, "loss": 1.3129, "step": 18228 }, { "epoch": 1.4286050156739811, "grad_norm": 2.6896071434020996, "learning_rate": 7.187142912455254e-06, "loss": 1.4026, "step": 18229 }, { "epoch": 1.4286833855799372, "grad_norm": 2.903207540512085, "learning_rate": 7.186865612742185e-06, "loss": 1.2293, "step": 18230 }, { "epoch": 1.4287617554858933, "grad_norm": 4.500511169433594, "learning_rate": 7.186588304711399e-06, "loss": 0.7306, "step": 18231 }, { "epoch": 1.4288401253918495, "grad_norm": 3.0228075981140137, "learning_rate": 7.186310988363957e-06, "loss": 1.4349, "step": 18232 }, { "epoch": 1.4289184952978056, "grad_norm": 2.7432243824005127, "learning_rate": 7.186033663700911e-06, "loss": 1.1322, "step": 18233 }, { "epoch": 1.4289968652037617, "grad_norm": 3.616250514984131, "learning_rate": 7.185756330723316e-06, "loss": 0.8835, "step": 18234 }, { "epoch": 1.4290752351097178, "grad_norm": 2.701563596725464, "learning_rate": 7.185478989432225e-06, "loss": 1.3103, "step": 18235 }, { "epoch": 1.429153605015674, "grad_norm": 3.983916759490967, "learning_rate": 7.185201639828695e-06, "loss": 1.387, "step": 18236 }, { "epoch": 1.4292319749216302, "grad_norm": 4.542185306549072, "learning_rate": 7.184924281913781e-06, "loss": 1.8265, "step": 18237 }, { "epoch": 1.4293103448275861, "grad_norm": 2.6160571575164795, "learning_rate": 7.184646915688536e-06, "loss": 1.6332, "step": 18238 }, { "epoch": 1.4293887147335425, "grad_norm": 2.8926594257354736, "learning_rate": 7.184369541154017e-06, "loss": 1.5455, "step": 18239 }, { "epoch": 1.4294670846394983, "grad_norm": 2.987117052078247, "learning_rate": 7.184092158311279e-06, "loss": 1.3437, "step": 18240 }, { "epoch": 1.4295454545454547, "grad_norm": 2.8216707706451416, "learning_rate": 7.183814767161375e-06, "loss": 1.5899, "step": 18241 }, { "epoch": 1.4296238244514106, "grad_norm": 2.5451886653900146, "learning_rate": 7.183537367705362e-06, "loss": 1.4535, "step": 18242 }, { "epoch": 1.429702194357367, "grad_norm": 3.264815330505371, "learning_rate": 7.183259959944296e-06, "loss": 1.6469, "step": 18243 }, { "epoch": 1.4297805642633228, "grad_norm": 6.006311893463135, "learning_rate": 7.1829825438792265e-06, "loss": 1.4895, "step": 18244 }, { "epoch": 1.4298589341692791, "grad_norm": 2.4470512866973877, "learning_rate": 7.182705119511215e-06, "loss": 1.2956, "step": 18245 }, { "epoch": 1.429937304075235, "grad_norm": 2.0108425617218018, "learning_rate": 7.182427686841314e-06, "loss": 1.4978, "step": 18246 }, { "epoch": 1.4300156739811913, "grad_norm": 2.59230637550354, "learning_rate": 7.182150245870581e-06, "loss": 1.0812, "step": 18247 }, { "epoch": 1.4300940438871472, "grad_norm": 4.054750919342041, "learning_rate": 7.181872796600067e-06, "loss": 1.4235, "step": 18248 }, { "epoch": 1.4301724137931036, "grad_norm": 3.75075364112854, "learning_rate": 7.181595339030829e-06, "loss": 1.0521, "step": 18249 }, { "epoch": 1.4302507836990594, "grad_norm": 2.237117290496826, "learning_rate": 7.1813178731639255e-06, "loss": 1.2837, "step": 18250 }, { "epoch": 1.4303291536050158, "grad_norm": 3.964329719543457, "learning_rate": 7.181040399000407e-06, "loss": 1.4504, "step": 18251 }, { "epoch": 1.4304075235109717, "grad_norm": 2.3795433044433594, "learning_rate": 7.1807629165413325e-06, "loss": 1.156, "step": 18252 }, { "epoch": 1.430485893416928, "grad_norm": 3.137723922729492, "learning_rate": 7.180485425787754e-06, "loss": 1.1451, "step": 18253 }, { "epoch": 1.4305642633228839, "grad_norm": 2.993372678756714, "learning_rate": 7.180207926740731e-06, "loss": 1.2327, "step": 18254 }, { "epoch": 1.4306426332288402, "grad_norm": 3.0438015460968018, "learning_rate": 7.179930419401315e-06, "loss": 1.7588, "step": 18255 }, { "epoch": 1.430721003134796, "grad_norm": 2.9873104095458984, "learning_rate": 7.179652903770565e-06, "loss": 1.2746, "step": 18256 }, { "epoch": 1.4307993730407524, "grad_norm": 2.3213088512420654, "learning_rate": 7.179375379849535e-06, "loss": 1.3294, "step": 18257 }, { "epoch": 1.4308777429467083, "grad_norm": 4.0290398597717285, "learning_rate": 7.179097847639279e-06, "loss": 1.5027, "step": 18258 }, { "epoch": 1.4309561128526647, "grad_norm": 2.85666561126709, "learning_rate": 7.178820307140855e-06, "loss": 1.1421, "step": 18259 }, { "epoch": 1.4310344827586206, "grad_norm": 2.696660280227661, "learning_rate": 7.178542758355317e-06, "loss": 1.5364, "step": 18260 }, { "epoch": 1.4311128526645769, "grad_norm": 3.8839006423950195, "learning_rate": 7.178265201283723e-06, "loss": 1.3493, "step": 18261 }, { "epoch": 1.431191222570533, "grad_norm": 3.2740635871887207, "learning_rate": 7.177987635927126e-06, "loss": 1.4181, "step": 18262 }, { "epoch": 1.431269592476489, "grad_norm": 6.028969764709473, "learning_rate": 7.177710062286583e-06, "loss": 1.0811, "step": 18263 }, { "epoch": 1.4313479623824452, "grad_norm": 3.2071008682250977, "learning_rate": 7.177432480363148e-06, "loss": 1.3087, "step": 18264 }, { "epoch": 1.4314263322884013, "grad_norm": 2.1208457946777344, "learning_rate": 7.1771548901578794e-06, "loss": 1.578, "step": 18265 }, { "epoch": 1.4315047021943574, "grad_norm": 10.01616382598877, "learning_rate": 7.176877291671831e-06, "loss": 0.898, "step": 18266 }, { "epoch": 1.4315830721003135, "grad_norm": 5.135559558868408, "learning_rate": 7.17659968490606e-06, "loss": 0.9687, "step": 18267 }, { "epoch": 1.4316614420062697, "grad_norm": 3.208343029022217, "learning_rate": 7.176322069861622e-06, "loss": 1.4561, "step": 18268 }, { "epoch": 1.4317398119122258, "grad_norm": 4.724156379699707, "learning_rate": 7.176044446539574e-06, "loss": 1.6875, "step": 18269 }, { "epoch": 1.4318181818181819, "grad_norm": 4.0337724685668945, "learning_rate": 7.175766814940967e-06, "loss": 1.3391, "step": 18270 }, { "epoch": 1.431896551724138, "grad_norm": 4.3010029792785645, "learning_rate": 7.175489175066864e-06, "loss": 1.5651, "step": 18271 }, { "epoch": 1.431974921630094, "grad_norm": 3.901639699935913, "learning_rate": 7.175211526918315e-06, "loss": 1.5471, "step": 18272 }, { "epoch": 1.4320532915360502, "grad_norm": 2.440340995788574, "learning_rate": 7.174933870496381e-06, "loss": 1.2532, "step": 18273 }, { "epoch": 1.4321316614420063, "grad_norm": 6.845313549041748, "learning_rate": 7.174656205802115e-06, "loss": 1.1703, "step": 18274 }, { "epoch": 1.4322100313479624, "grad_norm": 1.8010942935943604, "learning_rate": 7.174378532836574e-06, "loss": 0.9116, "step": 18275 }, { "epoch": 1.4322884012539185, "grad_norm": 2.1139047145843506, "learning_rate": 7.1741008516008135e-06, "loss": 1.5593, "step": 18276 }, { "epoch": 1.4323667711598747, "grad_norm": 2.8702285289764404, "learning_rate": 7.173823162095889e-06, "loss": 0.7634, "step": 18277 }, { "epoch": 1.4324451410658308, "grad_norm": 2.362046718597412, "learning_rate": 7.17354546432286e-06, "loss": 1.3358, "step": 18278 }, { "epoch": 1.4325235109717869, "grad_norm": 5.054678440093994, "learning_rate": 7.173267758282777e-06, "loss": 2.155, "step": 18279 }, { "epoch": 1.432601880877743, "grad_norm": 2.1571555137634277, "learning_rate": 7.172990043976703e-06, "loss": 1.4309, "step": 18280 }, { "epoch": 1.432680250783699, "grad_norm": 2.1325340270996094, "learning_rate": 7.172712321405689e-06, "loss": 1.4638, "step": 18281 }, { "epoch": 1.4327586206896552, "grad_norm": 5.5535359382629395, "learning_rate": 7.172434590570795e-06, "loss": 1.4155, "step": 18282 }, { "epoch": 1.4328369905956113, "grad_norm": 3.218503952026367, "learning_rate": 7.172156851473075e-06, "loss": 1.3065, "step": 18283 }, { "epoch": 1.4329153605015674, "grad_norm": 2.4864773750305176, "learning_rate": 7.171879104113585e-06, "loss": 1.3403, "step": 18284 }, { "epoch": 1.4329937304075235, "grad_norm": 7.538715839385986, "learning_rate": 7.1716013484933835e-06, "loss": 1.1948, "step": 18285 }, { "epoch": 1.4330721003134796, "grad_norm": 2.794966697692871, "learning_rate": 7.171323584613526e-06, "loss": 1.3274, "step": 18286 }, { "epoch": 1.4331504702194358, "grad_norm": 3.265685796737671, "learning_rate": 7.171045812475067e-06, "loss": 1.3149, "step": 18287 }, { "epoch": 1.4332288401253919, "grad_norm": 4.749534606933594, "learning_rate": 7.170768032079066e-06, "loss": 1.9247, "step": 18288 }, { "epoch": 1.433307210031348, "grad_norm": 2.400534152984619, "learning_rate": 7.17049024342658e-06, "loss": 1.0176, "step": 18289 }, { "epoch": 1.433385579937304, "grad_norm": 3.6247427463531494, "learning_rate": 7.170212446518663e-06, "loss": 1.5748, "step": 18290 }, { "epoch": 1.4334639498432602, "grad_norm": 1.8072361946105957, "learning_rate": 7.169934641356371e-06, "loss": 0.8856, "step": 18291 }, { "epoch": 1.4335423197492163, "grad_norm": 2.224215269088745, "learning_rate": 7.169656827940763e-06, "loss": 1.2224, "step": 18292 }, { "epoch": 1.4336206896551724, "grad_norm": 4.544886112213135, "learning_rate": 7.169379006272895e-06, "loss": 1.4564, "step": 18293 }, { "epoch": 1.4336990595611285, "grad_norm": 2.2323429584503174, "learning_rate": 7.169101176353824e-06, "loss": 1.4745, "step": 18294 }, { "epoch": 1.4337774294670846, "grad_norm": 3.1865735054016113, "learning_rate": 7.168823338184606e-06, "loss": 1.0035, "step": 18295 }, { "epoch": 1.4338557993730408, "grad_norm": 3.5506269931793213, "learning_rate": 7.168545491766298e-06, "loss": 1.4378, "step": 18296 }, { "epoch": 1.4339341692789969, "grad_norm": 9.562896728515625, "learning_rate": 7.168267637099957e-06, "loss": 0.492, "step": 18297 }, { "epoch": 1.434012539184953, "grad_norm": 2.8691868782043457, "learning_rate": 7.167989774186639e-06, "loss": 1.4483, "step": 18298 }, { "epoch": 1.434090909090909, "grad_norm": 5.287923336029053, "learning_rate": 7.167711903027402e-06, "loss": 1.2773, "step": 18299 }, { "epoch": 1.4341692789968652, "grad_norm": 3.1689460277557373, "learning_rate": 7.167434023623301e-06, "loss": 1.3248, "step": 18300 }, { "epoch": 1.4342476489028213, "grad_norm": 2.6372058391571045, "learning_rate": 7.167156135975396e-06, "loss": 1.7938, "step": 18301 }, { "epoch": 1.4343260188087774, "grad_norm": 2.301081657409668, "learning_rate": 7.166878240084741e-06, "loss": 1.2709, "step": 18302 }, { "epoch": 1.4344043887147335, "grad_norm": 3.221874237060547, "learning_rate": 7.166600335952395e-06, "loss": 1.3262, "step": 18303 }, { "epoch": 1.4344827586206896, "grad_norm": 6.224596977233887, "learning_rate": 7.166322423579415e-06, "loss": 1.5453, "step": 18304 }, { "epoch": 1.4345611285266457, "grad_norm": 2.277796983718872, "learning_rate": 7.166044502966855e-06, "loss": 1.2703, "step": 18305 }, { "epoch": 1.4346394984326019, "grad_norm": 2.564897060394287, "learning_rate": 7.165766574115776e-06, "loss": 1.5639, "step": 18306 }, { "epoch": 1.434717868338558, "grad_norm": 5.028066635131836, "learning_rate": 7.165488637027232e-06, "loss": 1.6079, "step": 18307 }, { "epoch": 1.434796238244514, "grad_norm": 2.183976650238037, "learning_rate": 7.165210691702282e-06, "loss": 1.635, "step": 18308 }, { "epoch": 1.4348746081504702, "grad_norm": 2.249938488006592, "learning_rate": 7.164932738141983e-06, "loss": 1.1295, "step": 18309 }, { "epoch": 1.4349529780564263, "grad_norm": 6.572813034057617, "learning_rate": 7.1646547763473916e-06, "loss": 2.1762, "step": 18310 }, { "epoch": 1.4350313479623824, "grad_norm": 2.634946346282959, "learning_rate": 7.1643768063195655e-06, "loss": 1.4196, "step": 18311 }, { "epoch": 1.4351097178683385, "grad_norm": 2.5471551418304443, "learning_rate": 7.164098828059561e-06, "loss": 1.1601, "step": 18312 }, { "epoch": 1.4351880877742946, "grad_norm": 2.243629217147827, "learning_rate": 7.163820841568437e-06, "loss": 1.3321, "step": 18313 }, { "epoch": 1.4352664576802507, "grad_norm": 3.98876953125, "learning_rate": 7.163542846847249e-06, "loss": 1.6977, "step": 18314 }, { "epoch": 1.4353448275862069, "grad_norm": 2.641580581665039, "learning_rate": 7.1632648438970555e-06, "loss": 1.2928, "step": 18315 }, { "epoch": 1.435423197492163, "grad_norm": 2.0562100410461426, "learning_rate": 7.1629868327189145e-06, "loss": 1.4826, "step": 18316 }, { "epoch": 1.435501567398119, "grad_norm": 3.802537202835083, "learning_rate": 7.162708813313882e-06, "loss": 1.4431, "step": 18317 }, { "epoch": 1.4355799373040752, "grad_norm": 2.962512493133545, "learning_rate": 7.1624307856830164e-06, "loss": 1.869, "step": 18318 }, { "epoch": 1.4356583072100313, "grad_norm": 2.2735371589660645, "learning_rate": 7.162152749827374e-06, "loss": 1.1422, "step": 18319 }, { "epoch": 1.4357366771159874, "grad_norm": 2.6014339923858643, "learning_rate": 7.161874705748014e-06, "loss": 1.233, "step": 18320 }, { "epoch": 1.4358150470219435, "grad_norm": 2.6969587802886963, "learning_rate": 7.161596653445992e-06, "loss": 1.1789, "step": 18321 }, { "epoch": 1.4358934169278996, "grad_norm": 3.663889169692993, "learning_rate": 7.161318592922367e-06, "loss": 1.2903, "step": 18322 }, { "epoch": 1.4359717868338557, "grad_norm": 1.8514740467071533, "learning_rate": 7.161040524178198e-06, "loss": 1.4088, "step": 18323 }, { "epoch": 1.4360501567398118, "grad_norm": 3.2045273780822754, "learning_rate": 7.160762447214537e-06, "loss": 1.7246, "step": 18324 }, { "epoch": 1.436128526645768, "grad_norm": 3.8100292682647705, "learning_rate": 7.160484362032449e-06, "loss": 1.6955, "step": 18325 }, { "epoch": 1.436206896551724, "grad_norm": 3.193997621536255, "learning_rate": 7.160206268632985e-06, "loss": 1.2648, "step": 18326 }, { "epoch": 1.4362852664576802, "grad_norm": 3.1700387001037598, "learning_rate": 7.159928167017209e-06, "loss": 1.1124, "step": 18327 }, { "epoch": 1.4363636363636363, "grad_norm": 2.4987919330596924, "learning_rate": 7.159650057186175e-06, "loss": 1.5651, "step": 18328 }, { "epoch": 1.4364420062695924, "grad_norm": 2.2228176593780518, "learning_rate": 7.15937193914094e-06, "loss": 1.1509, "step": 18329 }, { "epoch": 1.4365203761755485, "grad_norm": 3.1132962703704834, "learning_rate": 7.159093812882565e-06, "loss": 1.0984, "step": 18330 }, { "epoch": 1.4365987460815046, "grad_norm": 2.136406183242798, "learning_rate": 7.1588156784121055e-06, "loss": 1.3664, "step": 18331 }, { "epoch": 1.436677115987461, "grad_norm": 2.525317907333374, "learning_rate": 7.15853753573062e-06, "loss": 1.3537, "step": 18332 }, { "epoch": 1.4367554858934168, "grad_norm": 3.139263868331909, "learning_rate": 7.158259384839165e-06, "loss": 1.2802, "step": 18333 }, { "epoch": 1.4368338557993732, "grad_norm": 2.267878770828247, "learning_rate": 7.1579812257388015e-06, "loss": 1.3053, "step": 18334 }, { "epoch": 1.436912225705329, "grad_norm": 4.993031024932861, "learning_rate": 7.157703058430585e-06, "loss": 1.4769, "step": 18335 }, { "epoch": 1.4369905956112854, "grad_norm": 3.4883553981781006, "learning_rate": 7.157424882915575e-06, "loss": 1.2914, "step": 18336 }, { "epoch": 1.4370689655172413, "grad_norm": 2.3648815155029297, "learning_rate": 7.157146699194828e-06, "loss": 1.4925, "step": 18337 }, { "epoch": 1.4371473354231976, "grad_norm": 2.564811944961548, "learning_rate": 7.156868507269403e-06, "loss": 1.534, "step": 18338 }, { "epoch": 1.4372257053291535, "grad_norm": 2.05000901222229, "learning_rate": 7.156590307140358e-06, "loss": 1.3262, "step": 18339 }, { "epoch": 1.4373040752351098, "grad_norm": 3.924039602279663, "learning_rate": 7.156312098808753e-06, "loss": 1.2264, "step": 18340 }, { "epoch": 1.4373824451410657, "grad_norm": 2.629669189453125, "learning_rate": 7.156033882275643e-06, "loss": 1.277, "step": 18341 }, { "epoch": 1.437460815047022, "grad_norm": 8.549025535583496, "learning_rate": 7.155755657542087e-06, "loss": 0.7372, "step": 18342 }, { "epoch": 1.437539184952978, "grad_norm": 13.383816719055176, "learning_rate": 7.155477424609144e-06, "loss": 1.3008, "step": 18343 }, { "epoch": 1.4376175548589343, "grad_norm": 3.973660707473755, "learning_rate": 7.155199183477873e-06, "loss": 1.6125, "step": 18344 }, { "epoch": 1.4376959247648902, "grad_norm": 2.2812798023223877, "learning_rate": 7.154920934149331e-06, "loss": 1.2046, "step": 18345 }, { "epoch": 1.4377742946708465, "grad_norm": 4.529092788696289, "learning_rate": 7.154642676624576e-06, "loss": 0.9386, "step": 18346 }, { "epoch": 1.4378526645768024, "grad_norm": 2.562333345413208, "learning_rate": 7.1543644109046675e-06, "loss": 1.2971, "step": 18347 }, { "epoch": 1.4379310344827587, "grad_norm": 2.4054982662200928, "learning_rate": 7.154086136990662e-06, "loss": 1.2913, "step": 18348 }, { "epoch": 1.4380094043887146, "grad_norm": 2.8605165481567383, "learning_rate": 7.153807854883621e-06, "loss": 1.3755, "step": 18349 }, { "epoch": 1.438087774294671, "grad_norm": 2.357447862625122, "learning_rate": 7.1535295645845996e-06, "loss": 1.7885, "step": 18350 }, { "epoch": 1.4381661442006268, "grad_norm": 2.970769166946411, "learning_rate": 7.15325126609466e-06, "loss": 1.2915, "step": 18351 }, { "epoch": 1.4382445141065832, "grad_norm": 4.045078754425049, "learning_rate": 7.152972959414855e-06, "loss": 1.4035, "step": 18352 }, { "epoch": 1.438322884012539, "grad_norm": 2.929232597351074, "learning_rate": 7.15269464454625e-06, "loss": 1.229, "step": 18353 }, { "epoch": 1.4384012539184954, "grad_norm": 3.0848708152770996, "learning_rate": 7.152416321489899e-06, "loss": 1.2957, "step": 18354 }, { "epoch": 1.4384796238244515, "grad_norm": 2.526496410369873, "learning_rate": 7.152137990246862e-06, "loss": 1.4766, "step": 18355 }, { "epoch": 1.4385579937304076, "grad_norm": 2.8334553241729736, "learning_rate": 7.151859650818198e-06, "loss": 1.3154, "step": 18356 }, { "epoch": 1.4386363636363637, "grad_norm": 5.761535167694092, "learning_rate": 7.151581303204964e-06, "loss": 1.0808, "step": 18357 }, { "epoch": 1.4387147335423198, "grad_norm": 6.912318706512451, "learning_rate": 7.151302947408221e-06, "loss": 1.3657, "step": 18358 }, { "epoch": 1.438793103448276, "grad_norm": 3.6644136905670166, "learning_rate": 7.151024583429026e-06, "loss": 0.9894, "step": 18359 }, { "epoch": 1.438871473354232, "grad_norm": 3.1119797229766846, "learning_rate": 7.150746211268439e-06, "loss": 1.3763, "step": 18360 }, { "epoch": 1.4389498432601882, "grad_norm": 5.185423374176025, "learning_rate": 7.150467830927516e-06, "loss": 1.1727, "step": 18361 }, { "epoch": 1.4390282131661443, "grad_norm": 2.7893643379211426, "learning_rate": 7.1501894424073206e-06, "loss": 1.6208, "step": 18362 }, { "epoch": 1.4391065830721004, "grad_norm": 3.139305353164673, "learning_rate": 7.149911045708908e-06, "loss": 0.8846, "step": 18363 }, { "epoch": 1.4391849529780565, "grad_norm": 1.9860360622406006, "learning_rate": 7.149632640833337e-06, "loss": 1.5674, "step": 18364 }, { "epoch": 1.4392633228840126, "grad_norm": 2.7711477279663086, "learning_rate": 7.149354227781669e-06, "loss": 1.0423, "step": 18365 }, { "epoch": 1.4393416927899687, "grad_norm": 1.983652114868164, "learning_rate": 7.14907580655496e-06, "loss": 1.0259, "step": 18366 }, { "epoch": 1.4394200626959248, "grad_norm": 13.475505828857422, "learning_rate": 7.148797377154272e-06, "loss": 1.517, "step": 18367 }, { "epoch": 1.439498432601881, "grad_norm": 4.105361461639404, "learning_rate": 7.148518939580662e-06, "loss": 1.8706, "step": 18368 }, { "epoch": 1.439576802507837, "grad_norm": 4.019039154052734, "learning_rate": 7.148240493835189e-06, "loss": 1.6949, "step": 18369 }, { "epoch": 1.4396551724137931, "grad_norm": 3.297497510910034, "learning_rate": 7.147962039918913e-06, "loss": 1.0909, "step": 18370 }, { "epoch": 1.4397335423197493, "grad_norm": 4.957165718078613, "learning_rate": 7.147683577832892e-06, "loss": 1.1246, "step": 18371 }, { "epoch": 1.4398119122257054, "grad_norm": 5.66048526763916, "learning_rate": 7.147405107578187e-06, "loss": 1.6241, "step": 18372 }, { "epoch": 1.4398902821316615, "grad_norm": 2.207463264465332, "learning_rate": 7.1471266291558546e-06, "loss": 1.0747, "step": 18373 }, { "epoch": 1.4399686520376176, "grad_norm": 2.8757801055908203, "learning_rate": 7.146848142566957e-06, "loss": 0.9573, "step": 18374 }, { "epoch": 1.4400470219435737, "grad_norm": 3.2213332653045654, "learning_rate": 7.14656964781255e-06, "loss": 1.2565, "step": 18375 }, { "epoch": 1.4401253918495298, "grad_norm": 3.876638889312744, "learning_rate": 7.146291144893696e-06, "loss": 0.9394, "step": 18376 }, { "epoch": 1.440203761755486, "grad_norm": 4.191919326782227, "learning_rate": 7.146012633811454e-06, "loss": 1.6233, "step": 18377 }, { "epoch": 1.440282131661442, "grad_norm": 3.90712571144104, "learning_rate": 7.1457341145668776e-06, "loss": 1.3415, "step": 18378 }, { "epoch": 1.4403605015673981, "grad_norm": 2.7022275924682617, "learning_rate": 7.145455587161036e-06, "loss": 1.172, "step": 18379 }, { "epoch": 1.4404388714733543, "grad_norm": 3.2781922817230225, "learning_rate": 7.145177051594979e-06, "loss": 1.3688, "step": 18380 }, { "epoch": 1.4405172413793104, "grad_norm": 5.455151081085205, "learning_rate": 7.144898507869773e-06, "loss": 1.0913, "step": 18381 }, { "epoch": 1.4405956112852665, "grad_norm": 3.647374153137207, "learning_rate": 7.144619955986474e-06, "loss": 1.0649, "step": 18382 }, { "epoch": 1.4406739811912226, "grad_norm": 2.3244547843933105, "learning_rate": 7.1443413959461424e-06, "loss": 1.1537, "step": 18383 }, { "epoch": 1.4407523510971787, "grad_norm": 5.593031883239746, "learning_rate": 7.144062827749837e-06, "loss": 1.3993, "step": 18384 }, { "epoch": 1.4408307210031348, "grad_norm": 2.0765838623046875, "learning_rate": 7.143784251398618e-06, "loss": 1.1294, "step": 18385 }, { "epoch": 1.440909090909091, "grad_norm": 2.464982032775879, "learning_rate": 7.1435056668935455e-06, "loss": 1.4196, "step": 18386 }, { "epoch": 1.440987460815047, "grad_norm": 3.078486442565918, "learning_rate": 7.143227074235678e-06, "loss": 1.165, "step": 18387 }, { "epoch": 1.4410658307210031, "grad_norm": 7.650554180145264, "learning_rate": 7.142948473426077e-06, "loss": 1.5326, "step": 18388 }, { "epoch": 1.4411442006269592, "grad_norm": 3.7681901454925537, "learning_rate": 7.142669864465798e-06, "loss": 1.3168, "step": 18389 }, { "epoch": 1.4412225705329154, "grad_norm": 2.213036060333252, "learning_rate": 7.142391247355904e-06, "loss": 1.4611, "step": 18390 }, { "epoch": 1.4413009404388715, "grad_norm": 2.8683385848999023, "learning_rate": 7.142112622097455e-06, "loss": 0.9536, "step": 18391 }, { "epoch": 1.4413793103448276, "grad_norm": 3.160684108734131, "learning_rate": 7.14183398869151e-06, "loss": 1.6147, "step": 18392 }, { "epoch": 1.4414576802507837, "grad_norm": 3.537757396697998, "learning_rate": 7.141555347139129e-06, "loss": 1.373, "step": 18393 }, { "epoch": 1.4415360501567398, "grad_norm": 3.3858914375305176, "learning_rate": 7.14127669744137e-06, "loss": 1.5114, "step": 18394 }, { "epoch": 1.441614420062696, "grad_norm": 2.4531030654907227, "learning_rate": 7.140998039599294e-06, "loss": 1.5144, "step": 18395 }, { "epoch": 1.441692789968652, "grad_norm": 2.817208766937256, "learning_rate": 7.1407193736139625e-06, "loss": 1.6082, "step": 18396 }, { "epoch": 1.4417711598746081, "grad_norm": 2.17797589302063, "learning_rate": 7.140440699486433e-06, "loss": 1.0507, "step": 18397 }, { "epoch": 1.4418495297805642, "grad_norm": 10.692920684814453, "learning_rate": 7.140162017217769e-06, "loss": 0.6884, "step": 18398 }, { "epoch": 1.4419278996865204, "grad_norm": 2.891265869140625, "learning_rate": 7.139883326809024e-06, "loss": 1.199, "step": 18399 }, { "epoch": 1.4420062695924765, "grad_norm": 2.545936107635498, "learning_rate": 7.139604628261265e-06, "loss": 1.2779, "step": 18400 }, { "epoch": 1.4420846394984326, "grad_norm": 2.044477939605713, "learning_rate": 7.139325921575549e-06, "loss": 1.05, "step": 18401 }, { "epoch": 1.4421630094043887, "grad_norm": 2.7014806270599365, "learning_rate": 7.139047206752934e-06, "loss": 1.2253, "step": 18402 }, { "epoch": 1.4422413793103448, "grad_norm": 2.4976325035095215, "learning_rate": 7.1387684837944825e-06, "loss": 1.6876, "step": 18403 }, { "epoch": 1.442319749216301, "grad_norm": 2.3281872272491455, "learning_rate": 7.138489752701256e-06, "loss": 1.1087, "step": 18404 }, { "epoch": 1.442398119122257, "grad_norm": 3.10269832611084, "learning_rate": 7.138211013474312e-06, "loss": 0.7059, "step": 18405 }, { "epoch": 1.4424764890282131, "grad_norm": 1.8146989345550537, "learning_rate": 7.13793226611471e-06, "loss": 1.5017, "step": 18406 }, { "epoch": 1.4425548589341692, "grad_norm": 3.8260984420776367, "learning_rate": 7.137653510623513e-06, "loss": 1.6762, "step": 18407 }, { "epoch": 1.4426332288401253, "grad_norm": 4.135952472686768, "learning_rate": 7.137374747001779e-06, "loss": 1.5903, "step": 18408 }, { "epoch": 1.4427115987460815, "grad_norm": 5.186459541320801, "learning_rate": 7.137095975250572e-06, "loss": 1.2341, "step": 18409 }, { "epoch": 1.4427899686520376, "grad_norm": 6.090764999389648, "learning_rate": 7.136817195370948e-06, "loss": 1.5388, "step": 18410 }, { "epoch": 1.4428683385579937, "grad_norm": 2.5204226970672607, "learning_rate": 7.136538407363968e-06, "loss": 1.4094, "step": 18411 }, { "epoch": 1.4429467084639498, "grad_norm": 4.405819892883301, "learning_rate": 7.1362596112306935e-06, "loss": 2.0405, "step": 18412 }, { "epoch": 1.443025078369906, "grad_norm": 2.509866237640381, "learning_rate": 7.135980806972186e-06, "loss": 1.5697, "step": 18413 }, { "epoch": 1.443103448275862, "grad_norm": 3.689668893814087, "learning_rate": 7.135701994589503e-06, "loss": 1.4635, "step": 18414 }, { "epoch": 1.4431818181818181, "grad_norm": 3.0276780128479004, "learning_rate": 7.135423174083708e-06, "loss": 1.4572, "step": 18415 }, { "epoch": 1.4432601880877742, "grad_norm": 3.383906841278076, "learning_rate": 7.13514434545586e-06, "loss": 1.3351, "step": 18416 }, { "epoch": 1.4433385579937303, "grad_norm": 4.283149242401123, "learning_rate": 7.134865508707019e-06, "loss": 1.2783, "step": 18417 }, { "epoch": 1.4434169278996865, "grad_norm": 7.640678882598877, "learning_rate": 7.1345866638382465e-06, "loss": 2.3101, "step": 18418 }, { "epoch": 1.4434952978056426, "grad_norm": 2.798313617706299, "learning_rate": 7.134307810850602e-06, "loss": 1.3878, "step": 18419 }, { "epoch": 1.4435736677115987, "grad_norm": 2.2803099155426025, "learning_rate": 7.134028949745147e-06, "loss": 0.9874, "step": 18420 }, { "epoch": 1.4436520376175548, "grad_norm": 2.859917163848877, "learning_rate": 7.133750080522943e-06, "loss": 1.1887, "step": 18421 }, { "epoch": 1.443730407523511, "grad_norm": 2.6208276748657227, "learning_rate": 7.13347120318505e-06, "loss": 1.2586, "step": 18422 }, { "epoch": 1.443808777429467, "grad_norm": 2.119110345840454, "learning_rate": 7.133192317732526e-06, "loss": 1.2239, "step": 18423 }, { "epoch": 1.4438871473354231, "grad_norm": 2.65116810798645, "learning_rate": 7.132913424166436e-06, "loss": 0.7114, "step": 18424 }, { "epoch": 1.4439655172413794, "grad_norm": 2.9095265865325928, "learning_rate": 7.132634522487838e-06, "loss": 1.4973, "step": 18425 }, { "epoch": 1.4440438871473353, "grad_norm": 8.6322021484375, "learning_rate": 7.132355612697794e-06, "loss": 1.1888, "step": 18426 }, { "epoch": 1.4441222570532917, "grad_norm": 3.0028610229492188, "learning_rate": 7.132076694797364e-06, "loss": 1.3838, "step": 18427 }, { "epoch": 1.4442006269592476, "grad_norm": 2.759246349334717, "learning_rate": 7.131797768787611e-06, "loss": 1.4754, "step": 18428 }, { "epoch": 1.444278996865204, "grad_norm": 2.413465738296509, "learning_rate": 7.131518834669593e-06, "loss": 1.3127, "step": 18429 }, { "epoch": 1.4443573667711598, "grad_norm": 13.416040420532227, "learning_rate": 7.131239892444371e-06, "loss": 1.007, "step": 18430 }, { "epoch": 1.4444357366771161, "grad_norm": 4.306731700897217, "learning_rate": 7.13096094211301e-06, "loss": 1.1847, "step": 18431 }, { "epoch": 1.444514106583072, "grad_norm": 3.0761587619781494, "learning_rate": 7.130681983676565e-06, "loss": 1.339, "step": 18432 }, { "epoch": 1.4445924764890283, "grad_norm": 2.3501572608947754, "learning_rate": 7.130403017136103e-06, "loss": 1.421, "step": 18433 }, { "epoch": 1.4446708463949842, "grad_norm": 3.4043784141540527, "learning_rate": 7.1301240424926796e-06, "loss": 1.2776, "step": 18434 }, { "epoch": 1.4447492163009406, "grad_norm": 2.978665351867676, "learning_rate": 7.12984505974736e-06, "loss": 1.534, "step": 18435 }, { "epoch": 1.4448275862068964, "grad_norm": 2.2011947631835938, "learning_rate": 7.129566068901204e-06, "loss": 1.7236, "step": 18436 }, { "epoch": 1.4449059561128528, "grad_norm": 3.24362850189209, "learning_rate": 7.129287069955272e-06, "loss": 1.1592, "step": 18437 }, { "epoch": 1.4449843260188087, "grad_norm": 2.0573995113372803, "learning_rate": 7.129008062910625e-06, "loss": 1.2175, "step": 18438 }, { "epoch": 1.445062695924765, "grad_norm": 3.5407180786132812, "learning_rate": 7.128729047768326e-06, "loss": 1.4523, "step": 18439 }, { "epoch": 1.4451410658307209, "grad_norm": 2.101945161819458, "learning_rate": 7.128450024529434e-06, "loss": 1.1898, "step": 18440 }, { "epoch": 1.4452194357366772, "grad_norm": 9.87392520904541, "learning_rate": 7.1281709931950115e-06, "loss": 1.7716, "step": 18441 }, { "epoch": 1.445297805642633, "grad_norm": 2.323056221008301, "learning_rate": 7.127891953766119e-06, "loss": 1.2593, "step": 18442 }, { "epoch": 1.4453761755485894, "grad_norm": 2.834071636199951, "learning_rate": 7.1276129062438195e-06, "loss": 1.2804, "step": 18443 }, { "epoch": 1.4454545454545453, "grad_norm": 3.474637985229492, "learning_rate": 7.1273338506291745e-06, "loss": 1.1428, "step": 18444 }, { "epoch": 1.4455329153605017, "grad_norm": 2.981071949005127, "learning_rate": 7.127054786923243e-06, "loss": 1.3152, "step": 18445 }, { "epoch": 1.4456112852664575, "grad_norm": 2.1236228942871094, "learning_rate": 7.126775715127088e-06, "loss": 1.0295, "step": 18446 }, { "epoch": 1.4456896551724139, "grad_norm": 3.4710917472839355, "learning_rate": 7.1264966352417695e-06, "loss": 1.0329, "step": 18447 }, { "epoch": 1.4457680250783698, "grad_norm": 2.8621582984924316, "learning_rate": 7.12621754726835e-06, "loss": 1.4352, "step": 18448 }, { "epoch": 1.445846394984326, "grad_norm": 5.292993068695068, "learning_rate": 7.125938451207892e-06, "loss": 1.9339, "step": 18449 }, { "epoch": 1.4459247648902822, "grad_norm": 3.883796453475952, "learning_rate": 7.125659347061455e-06, "loss": 1.4742, "step": 18450 }, { "epoch": 1.4460031347962383, "grad_norm": 2.1268579959869385, "learning_rate": 7.125380234830103e-06, "loss": 0.7041, "step": 18451 }, { "epoch": 1.4460815047021944, "grad_norm": 4.527515411376953, "learning_rate": 7.125101114514896e-06, "loss": 1.3376, "step": 18452 }, { "epoch": 1.4461598746081505, "grad_norm": 6.487316608428955, "learning_rate": 7.124821986116895e-06, "loss": 1.8608, "step": 18453 }, { "epoch": 1.4462382445141067, "grad_norm": 1.9489251375198364, "learning_rate": 7.124542849637164e-06, "loss": 1.4778, "step": 18454 }, { "epoch": 1.4463166144200628, "grad_norm": 2.905163288116455, "learning_rate": 7.124263705076761e-06, "loss": 1.3023, "step": 18455 }, { "epoch": 1.4463949843260189, "grad_norm": 2.770113229751587, "learning_rate": 7.123984552436752e-06, "loss": 1.3453, "step": 18456 }, { "epoch": 1.446473354231975, "grad_norm": 2.257136583328247, "learning_rate": 7.1237053917181965e-06, "loss": 1.5946, "step": 18457 }, { "epoch": 1.446551724137931, "grad_norm": 9.747455596923828, "learning_rate": 7.123426222922155e-06, "loss": 1.088, "step": 18458 }, { "epoch": 1.4466300940438872, "grad_norm": 3.0778279304504395, "learning_rate": 7.123147046049693e-06, "loss": 1.6372, "step": 18459 }, { "epoch": 1.4467084639498433, "grad_norm": 4.843617916107178, "learning_rate": 7.122867861101868e-06, "loss": 1.6988, "step": 18460 }, { "epoch": 1.4467868338557994, "grad_norm": 7.029633045196533, "learning_rate": 7.122588668079747e-06, "loss": 0.8643, "step": 18461 }, { "epoch": 1.4468652037617555, "grad_norm": 3.2094075679779053, "learning_rate": 7.122309466984385e-06, "loss": 1.9474, "step": 18462 }, { "epoch": 1.4469435736677116, "grad_norm": 3.406146764755249, "learning_rate": 7.1220302578168495e-06, "loss": 1.609, "step": 18463 }, { "epoch": 1.4470219435736678, "grad_norm": 3.564581871032715, "learning_rate": 7.121751040578202e-06, "loss": 1.1703, "step": 18464 }, { "epoch": 1.4471003134796239, "grad_norm": 2.8782737255096436, "learning_rate": 7.1214718152695024e-06, "loss": 1.3126, "step": 18465 }, { "epoch": 1.44717868338558, "grad_norm": 2.1965911388397217, "learning_rate": 7.121192581891814e-06, "loss": 1.4565, "step": 18466 }, { "epoch": 1.447257053291536, "grad_norm": 2.235240936279297, "learning_rate": 7.120913340446199e-06, "loss": 1.5915, "step": 18467 }, { "epoch": 1.4473354231974922, "grad_norm": 2.6312108039855957, "learning_rate": 7.1206340909337164e-06, "loss": 1.5073, "step": 18468 }, { "epoch": 1.4474137931034483, "grad_norm": 2.0327370166778564, "learning_rate": 7.120354833355433e-06, "loss": 1.3233, "step": 18469 }, { "epoch": 1.4474921630094044, "grad_norm": 2.5759494304656982, "learning_rate": 7.120075567712409e-06, "loss": 1.2456, "step": 18470 }, { "epoch": 1.4475705329153605, "grad_norm": 2.3697876930236816, "learning_rate": 7.119796294005708e-06, "loss": 1.5004, "step": 18471 }, { "epoch": 1.4476489028213166, "grad_norm": 2.512317180633545, "learning_rate": 7.119517012236388e-06, "loss": 1.5632, "step": 18472 }, { "epoch": 1.4477272727272728, "grad_norm": 2.9064626693725586, "learning_rate": 7.119237722405514e-06, "loss": 1.3673, "step": 18473 }, { "epoch": 1.4478056426332289, "grad_norm": 3.3259201049804688, "learning_rate": 7.11895842451415e-06, "loss": 1.0613, "step": 18474 }, { "epoch": 1.447884012539185, "grad_norm": 6.453789234161377, "learning_rate": 7.118679118563356e-06, "loss": 1.8919, "step": 18475 }, { "epoch": 1.447962382445141, "grad_norm": 4.8574700355529785, "learning_rate": 7.118399804554193e-06, "loss": 1.546, "step": 18476 }, { "epoch": 1.4480407523510972, "grad_norm": 4.024721145629883, "learning_rate": 7.118120482487727e-06, "loss": 1.0901, "step": 18477 }, { "epoch": 1.4481191222570533, "grad_norm": 4.45288610458374, "learning_rate": 7.11784115236502e-06, "loss": 1.5734, "step": 18478 }, { "epoch": 1.4481974921630094, "grad_norm": 4.959612846374512, "learning_rate": 7.117561814187129e-06, "loss": 1.7484, "step": 18479 }, { "epoch": 1.4482758620689655, "grad_norm": 3.2227790355682373, "learning_rate": 7.117282467955124e-06, "loss": 1.1345, "step": 18480 }, { "epoch": 1.4483542319749216, "grad_norm": 4.391576290130615, "learning_rate": 7.117003113670063e-06, "loss": 1.3827, "step": 18481 }, { "epoch": 1.4484326018808777, "grad_norm": 2.636927366256714, "learning_rate": 7.116723751333009e-06, "loss": 1.114, "step": 18482 }, { "epoch": 1.4485109717868339, "grad_norm": 2.277916193008423, "learning_rate": 7.116444380945026e-06, "loss": 0.9287, "step": 18483 }, { "epoch": 1.44858934169279, "grad_norm": 3.9878549575805664, "learning_rate": 7.116165002507174e-06, "loss": 1.3115, "step": 18484 }, { "epoch": 1.448667711598746, "grad_norm": 2.0997183322906494, "learning_rate": 7.115885616020519e-06, "loss": 1.1437, "step": 18485 }, { "epoch": 1.4487460815047022, "grad_norm": 2.9160654544830322, "learning_rate": 7.115606221486121e-06, "loss": 1.2582, "step": 18486 }, { "epoch": 1.4488244514106583, "grad_norm": 2.7700300216674805, "learning_rate": 7.115326818905044e-06, "loss": 1.2567, "step": 18487 }, { "epoch": 1.4489028213166144, "grad_norm": 2.3755342960357666, "learning_rate": 7.115047408278349e-06, "loss": 1.0888, "step": 18488 }, { "epoch": 1.4489811912225705, "grad_norm": 1.9025386571884155, "learning_rate": 7.114767989607102e-06, "loss": 1.2803, "step": 18489 }, { "epoch": 1.4490595611285266, "grad_norm": 2.4922118186950684, "learning_rate": 7.114488562892363e-06, "loss": 0.9342, "step": 18490 }, { "epoch": 1.4491379310344827, "grad_norm": 5.159653186798096, "learning_rate": 7.1142091281351954e-06, "loss": 1.1862, "step": 18491 }, { "epoch": 1.4492163009404389, "grad_norm": 6.292453289031982, "learning_rate": 7.113929685336662e-06, "loss": 1.5165, "step": 18492 }, { "epoch": 1.449294670846395, "grad_norm": 4.0362162590026855, "learning_rate": 7.1136502344978265e-06, "loss": 1.6266, "step": 18493 }, { "epoch": 1.449373040752351, "grad_norm": 2.099323272705078, "learning_rate": 7.11337077561975e-06, "loss": 1.4717, "step": 18494 }, { "epoch": 1.4494514106583072, "grad_norm": 4.145229339599609, "learning_rate": 7.113091308703498e-06, "loss": 1.198, "step": 18495 }, { "epoch": 1.4495297805642633, "grad_norm": 3.0135550498962402, "learning_rate": 7.112811833750131e-06, "loss": 1.1695, "step": 18496 }, { "epoch": 1.4496081504702194, "grad_norm": 5.088078022003174, "learning_rate": 7.1125323507607134e-06, "loss": 1.5841, "step": 18497 }, { "epoch": 1.4496865203761755, "grad_norm": 2.367839813232422, "learning_rate": 7.112252859736307e-06, "loss": 1.3092, "step": 18498 }, { "epoch": 1.4497648902821316, "grad_norm": 2.067880392074585, "learning_rate": 7.111973360677978e-06, "loss": 1.4142, "step": 18499 }, { "epoch": 1.4498432601880877, "grad_norm": 2.919325828552246, "learning_rate": 7.111693853586785e-06, "loss": 1.3448, "step": 18500 }, { "epoch": 1.4499216300940438, "grad_norm": 2.2308802604675293, "learning_rate": 7.111414338463793e-06, "loss": 1.3411, "step": 18501 }, { "epoch": 1.45, "grad_norm": 3.5261385440826416, "learning_rate": 7.111134815310067e-06, "loss": 1.4398, "step": 18502 }, { "epoch": 1.450078369905956, "grad_norm": 3.8363633155822754, "learning_rate": 7.110855284126668e-06, "loss": 1.6124, "step": 18503 }, { "epoch": 1.4501567398119122, "grad_norm": 3.2199113368988037, "learning_rate": 7.110575744914659e-06, "loss": 1.6005, "step": 18504 }, { "epoch": 1.4502351097178683, "grad_norm": 2.856494188308716, "learning_rate": 7.110296197675104e-06, "loss": 1.2689, "step": 18505 }, { "epoch": 1.4503134796238244, "grad_norm": 2.5881049633026123, "learning_rate": 7.110016642409067e-06, "loss": 1.4132, "step": 18506 }, { "epoch": 1.4503918495297805, "grad_norm": 2.1630451679229736, "learning_rate": 7.109737079117609e-06, "loss": 1.3293, "step": 18507 }, { "epoch": 1.4504702194357366, "grad_norm": 3.002267837524414, "learning_rate": 7.109457507801796e-06, "loss": 1.4626, "step": 18508 }, { "epoch": 1.4505485893416927, "grad_norm": 2.6755359172821045, "learning_rate": 7.109177928462689e-06, "loss": 1.5634, "step": 18509 }, { "epoch": 1.4506269592476488, "grad_norm": 6.086198329925537, "learning_rate": 7.108898341101353e-06, "loss": 0.7772, "step": 18510 }, { "epoch": 1.450705329153605, "grad_norm": 3.7851250171661377, "learning_rate": 7.108618745718851e-06, "loss": 1.1799, "step": 18511 }, { "epoch": 1.450783699059561, "grad_norm": 3.9478933811187744, "learning_rate": 7.1083391423162475e-06, "loss": 1.3259, "step": 18512 }, { "epoch": 1.4508620689655172, "grad_norm": 2.317789316177368, "learning_rate": 7.108059530894604e-06, "loss": 1.0627, "step": 18513 }, { "epoch": 1.4509404388714733, "grad_norm": 8.631206512451172, "learning_rate": 7.107779911454984e-06, "loss": 1.0051, "step": 18514 }, { "epoch": 1.4510188087774294, "grad_norm": 3.3718369007110596, "learning_rate": 7.107500283998454e-06, "loss": 1.5119, "step": 18515 }, { "epoch": 1.4510971786833855, "grad_norm": 3.991279363632202, "learning_rate": 7.107220648526072e-06, "loss": 1.4369, "step": 18516 }, { "epoch": 1.4511755485893416, "grad_norm": 2.969928503036499, "learning_rate": 7.106941005038908e-06, "loss": 1.5916, "step": 18517 }, { "epoch": 1.4512539184952977, "grad_norm": 2.2959177494049072, "learning_rate": 7.106661353538022e-06, "loss": 1.2683, "step": 18518 }, { "epoch": 1.4513322884012538, "grad_norm": 2.7342429161071777, "learning_rate": 7.106381694024478e-06, "loss": 1.4204, "step": 18519 }, { "epoch": 1.4514106583072102, "grad_norm": 2.3587582111358643, "learning_rate": 7.106102026499339e-06, "loss": 1.2795, "step": 18520 }, { "epoch": 1.451489028213166, "grad_norm": 2.8574607372283936, "learning_rate": 7.10582235096367e-06, "loss": 1.3873, "step": 18521 }, { "epoch": 1.4515673981191224, "grad_norm": 3.80450177192688, "learning_rate": 7.105542667418535e-06, "loss": 1.2258, "step": 18522 }, { "epoch": 1.4516457680250783, "grad_norm": 4.962334632873535, "learning_rate": 7.105262975864997e-06, "loss": 1.0953, "step": 18523 }, { "epoch": 1.4517241379310346, "grad_norm": 4.371170520782471, "learning_rate": 7.104983276304119e-06, "loss": 1.4425, "step": 18524 }, { "epoch": 1.4518025078369905, "grad_norm": 3.915440082550049, "learning_rate": 7.104703568736968e-06, "loss": 1.1972, "step": 18525 }, { "epoch": 1.4518808777429468, "grad_norm": 6.514601707458496, "learning_rate": 7.104423853164604e-06, "loss": 1.3186, "step": 18526 }, { "epoch": 1.4519592476489027, "grad_norm": 6.111024856567383, "learning_rate": 7.104144129588093e-06, "loss": 1.3861, "step": 18527 }, { "epoch": 1.452037617554859, "grad_norm": 2.530576467514038, "learning_rate": 7.103864398008498e-06, "loss": 1.6644, "step": 18528 }, { "epoch": 1.452115987460815, "grad_norm": 1.860317349433899, "learning_rate": 7.103584658426884e-06, "loss": 1.2898, "step": 18529 }, { "epoch": 1.4521943573667713, "grad_norm": 2.2559521198272705, "learning_rate": 7.1033049108443145e-06, "loss": 1.0955, "step": 18530 }, { "epoch": 1.4522727272727272, "grad_norm": 4.751891136169434, "learning_rate": 7.103025155261854e-06, "loss": 1.4758, "step": 18531 }, { "epoch": 1.4523510971786835, "grad_norm": 3.9951391220092773, "learning_rate": 7.102745391680565e-06, "loss": 1.2469, "step": 18532 }, { "epoch": 1.4524294670846394, "grad_norm": 3.3911821842193604, "learning_rate": 7.102465620101512e-06, "loss": 1.7055, "step": 18533 }, { "epoch": 1.4525078369905957, "grad_norm": 3.3387291431427, "learning_rate": 7.102185840525762e-06, "loss": 1.206, "step": 18534 }, { "epoch": 1.4525862068965516, "grad_norm": 3.9869842529296875, "learning_rate": 7.101906052954372e-06, "loss": 1.6042, "step": 18535 }, { "epoch": 1.452664576802508, "grad_norm": 2.267763376235962, "learning_rate": 7.101626257388417e-06, "loss": 1.2037, "step": 18536 }, { "epoch": 1.4527429467084638, "grad_norm": 3.6245980262756348, "learning_rate": 7.101346453828951e-06, "loss": 1.6253, "step": 18537 }, { "epoch": 1.4528213166144202, "grad_norm": 2.907547950744629, "learning_rate": 7.1010666422770446e-06, "loss": 1.8763, "step": 18538 }, { "epoch": 1.452899686520376, "grad_norm": 3.8716859817504883, "learning_rate": 7.100786822733758e-06, "loss": 1.7251, "step": 18539 }, { "epoch": 1.4529780564263324, "grad_norm": 2.5923192501068115, "learning_rate": 7.100506995200159e-06, "loss": 1.0675, "step": 18540 }, { "epoch": 1.4530564263322883, "grad_norm": 3.4924118518829346, "learning_rate": 7.100227159677311e-06, "loss": 1.1026, "step": 18541 }, { "epoch": 1.4531347962382446, "grad_norm": 1.9389958381652832, "learning_rate": 7.099947316166275e-06, "loss": 1.2957, "step": 18542 }, { "epoch": 1.4532131661442007, "grad_norm": 2.8846070766448975, "learning_rate": 7.099667464668121e-06, "loss": 1.3035, "step": 18543 }, { "epoch": 1.4532915360501568, "grad_norm": 2.503993272781372, "learning_rate": 7.099387605183907e-06, "loss": 1.5908, "step": 18544 }, { "epoch": 1.453369905956113, "grad_norm": 2.4439663887023926, "learning_rate": 7.0991077377147035e-06, "loss": 1.2688, "step": 18545 }, { "epoch": 1.453448275862069, "grad_norm": 17.51683235168457, "learning_rate": 7.0988278622615716e-06, "loss": 1.2388, "step": 18546 }, { "epoch": 1.4535266457680251, "grad_norm": 2.8694043159484863, "learning_rate": 7.098547978825576e-06, "loss": 1.3665, "step": 18547 }, { "epoch": 1.4536050156739813, "grad_norm": 2.792006731033325, "learning_rate": 7.098268087407782e-06, "loss": 1.3603, "step": 18548 }, { "epoch": 1.4536833855799374, "grad_norm": 2.626948356628418, "learning_rate": 7.097988188009255e-06, "loss": 1.2958, "step": 18549 }, { "epoch": 1.4537617554858935, "grad_norm": 3.1572365760803223, "learning_rate": 7.097708280631057e-06, "loss": 1.3637, "step": 18550 }, { "epoch": 1.4538401253918496, "grad_norm": 7.548600196838379, "learning_rate": 7.097428365274255e-06, "loss": 1.061, "step": 18551 }, { "epoch": 1.4539184952978057, "grad_norm": 2.2676424980163574, "learning_rate": 7.097148441939912e-06, "loss": 1.313, "step": 18552 }, { "epoch": 1.4539968652037618, "grad_norm": 6.805093288421631, "learning_rate": 7.096868510629096e-06, "loss": 1.2336, "step": 18553 }, { "epoch": 1.454075235109718, "grad_norm": 2.3928756713867188, "learning_rate": 7.096588571342866e-06, "loss": 1.4361, "step": 18554 }, { "epoch": 1.454153605015674, "grad_norm": 3.38810133934021, "learning_rate": 7.096308624082292e-06, "loss": 1.5831, "step": 18555 }, { "epoch": 1.4542319749216301, "grad_norm": 4.532074928283691, "learning_rate": 7.096028668848436e-06, "loss": 1.6344, "step": 18556 }, { "epoch": 1.4543103448275863, "grad_norm": 2.3821845054626465, "learning_rate": 7.095748705642362e-06, "loss": 1.3719, "step": 18557 }, { "epoch": 1.4543887147335424, "grad_norm": 2.8540101051330566, "learning_rate": 7.095468734465138e-06, "loss": 1.3572, "step": 18558 }, { "epoch": 1.4544670846394985, "grad_norm": 2.0022976398468018, "learning_rate": 7.095188755317828e-06, "loss": 1.5324, "step": 18559 }, { "epoch": 1.4545454545454546, "grad_norm": 3.924206256866455, "learning_rate": 7.0949087682014965e-06, "loss": 1.685, "step": 18560 }, { "epoch": 1.4546238244514107, "grad_norm": 2.2762441635131836, "learning_rate": 7.094628773117205e-06, "loss": 1.2439, "step": 18561 }, { "epoch": 1.4547021943573668, "grad_norm": 3.7575113773345947, "learning_rate": 7.094348770066024e-06, "loss": 1.4399, "step": 18562 }, { "epoch": 1.454780564263323, "grad_norm": 2.2037291526794434, "learning_rate": 7.0940687590490155e-06, "loss": 1.5064, "step": 18563 }, { "epoch": 1.454858934169279, "grad_norm": 3.3281025886535645, "learning_rate": 7.093788740067244e-06, "loss": 1.3109, "step": 18564 }, { "epoch": 1.4549373040752351, "grad_norm": 3.527496576309204, "learning_rate": 7.093508713121778e-06, "loss": 1.1656, "step": 18565 }, { "epoch": 1.4550156739811912, "grad_norm": 2.8999762535095215, "learning_rate": 7.093228678213677e-06, "loss": 1.2423, "step": 18566 }, { "epoch": 1.4550940438871474, "grad_norm": 3.770636796951294, "learning_rate": 7.092948635344011e-06, "loss": 1.3644, "step": 18567 }, { "epoch": 1.4551724137931035, "grad_norm": 5.450827121734619, "learning_rate": 7.092668584513844e-06, "loss": 1.4813, "step": 18568 }, { "epoch": 1.4552507836990596, "grad_norm": 1.953432559967041, "learning_rate": 7.092388525724241e-06, "loss": 1.4684, "step": 18569 }, { "epoch": 1.4553291536050157, "grad_norm": 4.801266670227051, "learning_rate": 7.092108458976263e-06, "loss": 1.5324, "step": 18570 }, { "epoch": 1.4554075235109718, "grad_norm": 3.342780828475952, "learning_rate": 7.091828384270984e-06, "loss": 1.1021, "step": 18571 }, { "epoch": 1.455485893416928, "grad_norm": 3.17317795753479, "learning_rate": 7.091548301609461e-06, "loss": 1.4273, "step": 18572 }, { "epoch": 1.455564263322884, "grad_norm": 3.215165615081787, "learning_rate": 7.091268210992763e-06, "loss": 1.6983, "step": 18573 }, { "epoch": 1.4556426332288401, "grad_norm": 3.481069803237915, "learning_rate": 7.090988112421956e-06, "loss": 1.6343, "step": 18574 }, { "epoch": 1.4557210031347962, "grad_norm": 3.1445398330688477, "learning_rate": 7.090708005898103e-06, "loss": 1.1618, "step": 18575 }, { "epoch": 1.4557993730407524, "grad_norm": 2.9853081703186035, "learning_rate": 7.090427891422271e-06, "loss": 1.6965, "step": 18576 }, { "epoch": 1.4558777429467085, "grad_norm": 3.7948403358459473, "learning_rate": 7.090147768995525e-06, "loss": 1.4468, "step": 18577 }, { "epoch": 1.4559561128526646, "grad_norm": 2.4972033500671387, "learning_rate": 7.089867638618931e-06, "loss": 1.485, "step": 18578 }, { "epoch": 1.4560344827586207, "grad_norm": 2.3092849254608154, "learning_rate": 7.0895875002935535e-06, "loss": 1.1268, "step": 18579 }, { "epoch": 1.4561128526645768, "grad_norm": 2.867847204208374, "learning_rate": 7.089307354020459e-06, "loss": 1.3082, "step": 18580 }, { "epoch": 1.456191222570533, "grad_norm": 2.900465965270996, "learning_rate": 7.089027199800711e-06, "loss": 1.455, "step": 18581 }, { "epoch": 1.456269592476489, "grad_norm": 6.09663724899292, "learning_rate": 7.088747037635377e-06, "loss": 1.7313, "step": 18582 }, { "epoch": 1.4563479623824451, "grad_norm": 3.3828184604644775, "learning_rate": 7.088466867525523e-06, "loss": 1.5543, "step": 18583 }, { "epoch": 1.4564263322884012, "grad_norm": 2.771428346633911, "learning_rate": 7.088186689472213e-06, "loss": 1.3667, "step": 18584 }, { "epoch": 1.4565047021943573, "grad_norm": 5.736284255981445, "learning_rate": 7.0879065034765135e-06, "loss": 1.3979, "step": 18585 }, { "epoch": 1.4565830721003135, "grad_norm": 2.8080978393554688, "learning_rate": 7.0876263095394915e-06, "loss": 1.4301, "step": 18586 }, { "epoch": 1.4566614420062696, "grad_norm": 4.567803859710693, "learning_rate": 7.087346107662208e-06, "loss": 1.0936, "step": 18587 }, { "epoch": 1.4567398119122257, "grad_norm": 3.182807207107544, "learning_rate": 7.087065897845735e-06, "loss": 1.062, "step": 18588 }, { "epoch": 1.4568181818181818, "grad_norm": 2.192871570587158, "learning_rate": 7.086785680091134e-06, "loss": 1.326, "step": 18589 }, { "epoch": 1.456896551724138, "grad_norm": 3.031449317932129, "learning_rate": 7.086505454399473e-06, "loss": 1.4436, "step": 18590 }, { "epoch": 1.456974921630094, "grad_norm": 2.339423179626465, "learning_rate": 7.086225220771816e-06, "loss": 1.3071, "step": 18591 }, { "epoch": 1.4570532915360501, "grad_norm": 2.585123062133789, "learning_rate": 7.085944979209231e-06, "loss": 1.2068, "step": 18592 }, { "epoch": 1.4571316614420062, "grad_norm": 3.1212074756622314, "learning_rate": 7.085664729712781e-06, "loss": 1.538, "step": 18593 }, { "epoch": 1.4572100313479623, "grad_norm": 3.7409021854400635, "learning_rate": 7.085384472283535e-06, "loss": 1.4569, "step": 18594 }, { "epoch": 1.4572884012539185, "grad_norm": 3.8260574340820312, "learning_rate": 7.085104206922557e-06, "loss": 1.2952, "step": 18595 }, { "epoch": 1.4573667711598746, "grad_norm": 12.886631965637207, "learning_rate": 7.084823933630913e-06, "loss": 1.7942, "step": 18596 }, { "epoch": 1.4574451410658307, "grad_norm": 3.854499578475952, "learning_rate": 7.084543652409673e-06, "loss": 1.5959, "step": 18597 }, { "epoch": 1.4575235109717868, "grad_norm": 3.2587203979492188, "learning_rate": 7.0842633632598945e-06, "loss": 1.4995, "step": 18598 }, { "epoch": 1.457601880877743, "grad_norm": 1.7964768409729004, "learning_rate": 7.083983066182652e-06, "loss": 1.2902, "step": 18599 }, { "epoch": 1.457680250783699, "grad_norm": 2.7750699520111084, "learning_rate": 7.0837027611790075e-06, "loss": 1.2166, "step": 18600 }, { "epoch": 1.4577586206896551, "grad_norm": 2.252661943435669, "learning_rate": 7.083422448250028e-06, "loss": 1.367, "step": 18601 }, { "epoch": 1.4578369905956112, "grad_norm": 2.391519069671631, "learning_rate": 7.083142127396779e-06, "loss": 1.354, "step": 18602 }, { "epoch": 1.4579153605015673, "grad_norm": 2.237419605255127, "learning_rate": 7.082861798620329e-06, "loss": 1.3528, "step": 18603 }, { "epoch": 1.4579937304075234, "grad_norm": 3.0614686012268066, "learning_rate": 7.0825814619217406e-06, "loss": 1.1736, "step": 18604 }, { "epoch": 1.4580721003134796, "grad_norm": 2.464071750640869, "learning_rate": 7.082301117302083e-06, "loss": 1.4032, "step": 18605 }, { "epoch": 1.4581504702194357, "grad_norm": 7.341878890991211, "learning_rate": 7.082020764762421e-06, "loss": 1.2637, "step": 18606 }, { "epoch": 1.4582288401253918, "grad_norm": 6.3884735107421875, "learning_rate": 7.081740404303823e-06, "loss": 1.0262, "step": 18607 }, { "epoch": 1.458307210031348, "grad_norm": 2.1164867877960205, "learning_rate": 7.081460035927351e-06, "loss": 1.3668, "step": 18608 }, { "epoch": 1.458385579937304, "grad_norm": 3.0371344089508057, "learning_rate": 7.081179659634076e-06, "loss": 0.7508, "step": 18609 }, { "epoch": 1.45846394984326, "grad_norm": 2.6943447589874268, "learning_rate": 7.080899275425063e-06, "loss": 1.0003, "step": 18610 }, { "epoch": 1.4585423197492162, "grad_norm": 3.1515097618103027, "learning_rate": 7.0806188833013765e-06, "loss": 1.2298, "step": 18611 }, { "epoch": 1.4586206896551723, "grad_norm": 4.740994930267334, "learning_rate": 7.080338483264084e-06, "loss": 1.4345, "step": 18612 }, { "epoch": 1.4586990595611287, "grad_norm": 1.9809904098510742, "learning_rate": 7.080058075314254e-06, "loss": 1.4332, "step": 18613 }, { "epoch": 1.4587774294670846, "grad_norm": 3.1943459510803223, "learning_rate": 7.0797776594529525e-06, "loss": 1.2617, "step": 18614 }, { "epoch": 1.4588557993730409, "grad_norm": 7.935259819030762, "learning_rate": 7.079497235681242e-06, "loss": 1.0166, "step": 18615 }, { "epoch": 1.4589341692789968, "grad_norm": 2.3213446140289307, "learning_rate": 7.079216804000194e-06, "loss": 1.4504, "step": 18616 }, { "epoch": 1.459012539184953, "grad_norm": 2.141535520553589, "learning_rate": 7.078936364410872e-06, "loss": 1.1391, "step": 18617 }, { "epoch": 1.459090909090909, "grad_norm": 3.487478256225586, "learning_rate": 7.078655916914345e-06, "loss": 1.1653, "step": 18618 }, { "epoch": 1.4591692789968653, "grad_norm": 13.350846290588379, "learning_rate": 7.078375461511679e-06, "loss": 1.4766, "step": 18619 }, { "epoch": 1.4592476489028212, "grad_norm": 2.4244706630706787, "learning_rate": 7.078094998203939e-06, "loss": 1.8301, "step": 18620 }, { "epoch": 1.4593260188087775, "grad_norm": 3.5183451175689697, "learning_rate": 7.077814526992194e-06, "loss": 1.4722, "step": 18621 }, { "epoch": 1.4594043887147334, "grad_norm": 2.6303353309631348, "learning_rate": 7.077534047877509e-06, "loss": 1.1492, "step": 18622 }, { "epoch": 1.4594827586206898, "grad_norm": 2.3875107765197754, "learning_rate": 7.0772535608609515e-06, "loss": 1.4388, "step": 18623 }, { "epoch": 1.4595611285266457, "grad_norm": 2.1637516021728516, "learning_rate": 7.076973065943589e-06, "loss": 1.3016, "step": 18624 }, { "epoch": 1.459639498432602, "grad_norm": 14.415205955505371, "learning_rate": 7.076692563126488e-06, "loss": 2.5613, "step": 18625 }, { "epoch": 1.4597178683385579, "grad_norm": 3.52378249168396, "learning_rate": 7.076412052410715e-06, "loss": 1.5044, "step": 18626 }, { "epoch": 1.4597962382445142, "grad_norm": 3.565343141555786, "learning_rate": 7.0761315337973365e-06, "loss": 1.2801, "step": 18627 }, { "epoch": 1.45987460815047, "grad_norm": 2.954172372817993, "learning_rate": 7.07585100728742e-06, "loss": 1.5451, "step": 18628 }, { "epoch": 1.4599529780564264, "grad_norm": 2.6803934574127197, "learning_rate": 7.075570472882033e-06, "loss": 1.2208, "step": 18629 }, { "epoch": 1.4600313479623823, "grad_norm": 3.2604081630706787, "learning_rate": 7.075289930582241e-06, "loss": 1.2783, "step": 18630 }, { "epoch": 1.4601097178683387, "grad_norm": 2.1738898754119873, "learning_rate": 7.075009380389113e-06, "loss": 1.3148, "step": 18631 }, { "epoch": 1.4601880877742945, "grad_norm": 3.5116777420043945, "learning_rate": 7.074728822303715e-06, "loss": 1.2863, "step": 18632 }, { "epoch": 1.4602664576802509, "grad_norm": 2.125452995300293, "learning_rate": 7.074448256327114e-06, "loss": 1.5899, "step": 18633 }, { "epoch": 1.4603448275862068, "grad_norm": 2.753005266189575, "learning_rate": 7.074167682460378e-06, "loss": 1.4206, "step": 18634 }, { "epoch": 1.460423197492163, "grad_norm": 5.057666778564453, "learning_rate": 7.0738871007045725e-06, "loss": 1.27, "step": 18635 }, { "epoch": 1.4605015673981192, "grad_norm": 4.028746128082275, "learning_rate": 7.073606511060765e-06, "loss": 1.275, "step": 18636 }, { "epoch": 1.4605799373040753, "grad_norm": 4.1836466789245605, "learning_rate": 7.073325913530026e-06, "loss": 1.4412, "step": 18637 }, { "epoch": 1.4606583072100314, "grad_norm": 2.917609453201294, "learning_rate": 7.073045308113419e-06, "loss": 1.4029, "step": 18638 }, { "epoch": 1.4607366771159875, "grad_norm": 2.79343843460083, "learning_rate": 7.072764694812011e-06, "loss": 1.0514, "step": 18639 }, { "epoch": 1.4608150470219436, "grad_norm": 2.8999979496002197, "learning_rate": 7.072484073626872e-06, "loss": 1.2754, "step": 18640 }, { "epoch": 1.4608934169278998, "grad_norm": 3.330866575241089, "learning_rate": 7.072203444559068e-06, "loss": 1.3792, "step": 18641 }, { "epoch": 1.4609717868338559, "grad_norm": 3.1863889694213867, "learning_rate": 7.071922807609667e-06, "loss": 1.6055, "step": 18642 }, { "epoch": 1.461050156739812, "grad_norm": 2.110473155975342, "learning_rate": 7.071642162779735e-06, "loss": 1.0149, "step": 18643 }, { "epoch": 1.461128526645768, "grad_norm": 3.694087266921997, "learning_rate": 7.07136151007034e-06, "loss": 0.9907, "step": 18644 }, { "epoch": 1.4612068965517242, "grad_norm": 3.4511444568634033, "learning_rate": 7.0710808494825515e-06, "loss": 1.3657, "step": 18645 }, { "epoch": 1.4612852664576803, "grad_norm": 2.0897111892700195, "learning_rate": 7.0708001810174335e-06, "loss": 0.957, "step": 18646 }, { "epoch": 1.4613636363636364, "grad_norm": 9.380558013916016, "learning_rate": 7.070519504676056e-06, "loss": 1.0438, "step": 18647 }, { "epoch": 1.4614420062695925, "grad_norm": 2.2721004486083984, "learning_rate": 7.070238820459485e-06, "loss": 1.1696, "step": 18648 }, { "epoch": 1.4615203761755486, "grad_norm": 2.2975451946258545, "learning_rate": 7.069958128368789e-06, "loss": 1.0004, "step": 18649 }, { "epoch": 1.4615987460815048, "grad_norm": 4.706953048706055, "learning_rate": 7.069677428405035e-06, "loss": 1.3517, "step": 18650 }, { "epoch": 1.4616771159874609, "grad_norm": 2.6038575172424316, "learning_rate": 7.069396720569293e-06, "loss": 1.523, "step": 18651 }, { "epoch": 1.461755485893417, "grad_norm": 4.983109951019287, "learning_rate": 7.0691160048626275e-06, "loss": 1.155, "step": 18652 }, { "epoch": 1.461833855799373, "grad_norm": 3.0346667766571045, "learning_rate": 7.068835281286109e-06, "loss": 1.3416, "step": 18653 }, { "epoch": 1.4619122257053292, "grad_norm": 2.0547454357147217, "learning_rate": 7.068554549840801e-06, "loss": 1.2239, "step": 18654 }, { "epoch": 1.4619905956112853, "grad_norm": 2.516456127166748, "learning_rate": 7.068273810527776e-06, "loss": 1.3026, "step": 18655 }, { "epoch": 1.4620689655172414, "grad_norm": 1.9382879734039307, "learning_rate": 7.067993063348099e-06, "loss": 1.1324, "step": 18656 }, { "epoch": 1.4621473354231975, "grad_norm": 6.656585693359375, "learning_rate": 7.067712308302839e-06, "loss": 1.5839, "step": 18657 }, { "epoch": 1.4622257053291536, "grad_norm": 2.5713953971862793, "learning_rate": 7.067431545393062e-06, "loss": 1.4932, "step": 18658 }, { "epoch": 1.4623040752351097, "grad_norm": 2.3013782501220703, "learning_rate": 7.067150774619839e-06, "loss": 1.3092, "step": 18659 }, { "epoch": 1.4623824451410659, "grad_norm": 4.014723777770996, "learning_rate": 7.066869995984236e-06, "loss": 1.4362, "step": 18660 }, { "epoch": 1.462460815047022, "grad_norm": 3.260978937149048, "learning_rate": 7.066589209487321e-06, "loss": 1.3693, "step": 18661 }, { "epoch": 1.462539184952978, "grad_norm": 4.113279342651367, "learning_rate": 7.066308415130161e-06, "loss": 1.8352, "step": 18662 }, { "epoch": 1.4626175548589342, "grad_norm": 5.0011162757873535, "learning_rate": 7.066027612913827e-06, "loss": 1.6593, "step": 18663 }, { "epoch": 1.4626959247648903, "grad_norm": 3.366511583328247, "learning_rate": 7.065746802839383e-06, "loss": 1.5086, "step": 18664 }, { "epoch": 1.4627742946708464, "grad_norm": 2.4622886180877686, "learning_rate": 7.065465984907901e-06, "loss": 1.0789, "step": 18665 }, { "epoch": 1.4628526645768025, "grad_norm": 6.782904624938965, "learning_rate": 7.065185159120447e-06, "loss": 1.1883, "step": 18666 }, { "epoch": 1.4629310344827586, "grad_norm": 2.6591782569885254, "learning_rate": 7.064904325478089e-06, "loss": 1.5172, "step": 18667 }, { "epoch": 1.4630094043887147, "grad_norm": 4.026864051818848, "learning_rate": 7.0646234839818965e-06, "loss": 1.58, "step": 18668 }, { "epoch": 1.4630877742946709, "grad_norm": 2.418093204498291, "learning_rate": 7.064342634632935e-06, "loss": 1.1913, "step": 18669 }, { "epoch": 1.463166144200627, "grad_norm": 5.363358020782471, "learning_rate": 7.064061777432276e-06, "loss": 1.2008, "step": 18670 }, { "epoch": 1.463244514106583, "grad_norm": 6.434842586517334, "learning_rate": 7.063780912380984e-06, "loss": 1.3535, "step": 18671 }, { "epoch": 1.4633228840125392, "grad_norm": 4.1752495765686035, "learning_rate": 7.063500039480131e-06, "loss": 1.3294, "step": 18672 }, { "epoch": 1.4634012539184953, "grad_norm": 2.4373395442962646, "learning_rate": 7.063219158730784e-06, "loss": 1.3251, "step": 18673 }, { "epoch": 1.4634796238244514, "grad_norm": 5.104575157165527, "learning_rate": 7.062938270134011e-06, "loss": 1.6717, "step": 18674 }, { "epoch": 1.4635579937304075, "grad_norm": 8.714685440063477, "learning_rate": 7.06265737369088e-06, "loss": 1.8953, "step": 18675 }, { "epoch": 1.4636363636363636, "grad_norm": 10.02941608428955, "learning_rate": 7.062376469402459e-06, "loss": 1.4249, "step": 18676 }, { "epoch": 1.4637147335423197, "grad_norm": 6.032079696655273, "learning_rate": 7.062095557269818e-06, "loss": 1.6688, "step": 18677 }, { "epoch": 1.4637931034482758, "grad_norm": 6.197020530700684, "learning_rate": 7.061814637294024e-06, "loss": 1.4982, "step": 18678 }, { "epoch": 1.463871473354232, "grad_norm": 2.5703914165496826, "learning_rate": 7.061533709476147e-06, "loss": 0.988, "step": 18679 }, { "epoch": 1.463949843260188, "grad_norm": 2.2809081077575684, "learning_rate": 7.0612527738172565e-06, "loss": 1.113, "step": 18680 }, { "epoch": 1.4640282131661442, "grad_norm": 3.384381055831909, "learning_rate": 7.060971830318417e-06, "loss": 1.0868, "step": 18681 }, { "epoch": 1.4641065830721003, "grad_norm": 3.5754776000976562, "learning_rate": 7.060690878980699e-06, "loss": 1.1973, "step": 18682 }, { "epoch": 1.4641849529780564, "grad_norm": 2.650820016860962, "learning_rate": 7.060409919805171e-06, "loss": 1.5328, "step": 18683 }, { "epoch": 1.4642633228840125, "grad_norm": 3.763366222381592, "learning_rate": 7.060128952792904e-06, "loss": 1.2268, "step": 18684 }, { "epoch": 1.4643416927899686, "grad_norm": 3.0387701988220215, "learning_rate": 7.059847977944962e-06, "loss": 1.2047, "step": 18685 }, { "epoch": 1.4644200626959247, "grad_norm": 3.574368715286255, "learning_rate": 7.0595669952624175e-06, "loss": 1.339, "step": 18686 }, { "epoch": 1.4644984326018808, "grad_norm": 3.9014062881469727, "learning_rate": 7.059286004746339e-06, "loss": 1.4219, "step": 18687 }, { "epoch": 1.464576802507837, "grad_norm": 3.092184066772461, "learning_rate": 7.059005006397792e-06, "loss": 1.4534, "step": 18688 }, { "epoch": 1.464655172413793, "grad_norm": 2.629739761352539, "learning_rate": 7.058724000217851e-06, "loss": 1.2028, "step": 18689 }, { "epoch": 1.4647335423197492, "grad_norm": 3.5936121940612793, "learning_rate": 7.058442986207577e-06, "loss": 2.0567, "step": 18690 }, { "epoch": 1.4648119122257053, "grad_norm": 3.9036946296691895, "learning_rate": 7.058161964368047e-06, "loss": 1.5355, "step": 18691 }, { "epoch": 1.4648902821316614, "grad_norm": 2.164557933807373, "learning_rate": 7.057880934700324e-06, "loss": 1.3386, "step": 18692 }, { "epoch": 1.4649686520376175, "grad_norm": 2.627394199371338, "learning_rate": 7.057599897205479e-06, "loss": 1.4186, "step": 18693 }, { "epoch": 1.4650470219435736, "grad_norm": 6.372649669647217, "learning_rate": 7.057318851884581e-06, "loss": 0.7632, "step": 18694 }, { "epoch": 1.4651253918495297, "grad_norm": 9.281352043151855, "learning_rate": 7.0570377987387e-06, "loss": 1.457, "step": 18695 }, { "epoch": 1.4652037617554858, "grad_norm": 4.327761173248291, "learning_rate": 7.056756737768904e-06, "loss": 1.7475, "step": 18696 }, { "epoch": 1.465282131661442, "grad_norm": 4.6281232833862305, "learning_rate": 7.056475668976259e-06, "loss": 1.7001, "step": 18697 }, { "epoch": 1.465360501567398, "grad_norm": 2.375215768814087, "learning_rate": 7.056194592361839e-06, "loss": 1.2088, "step": 18698 }, { "epoch": 1.4654388714733542, "grad_norm": 3.1255486011505127, "learning_rate": 7.0559135079267096e-06, "loss": 1.2323, "step": 18699 }, { "epoch": 1.4655172413793103, "grad_norm": 4.224027156829834, "learning_rate": 7.055632415671942e-06, "loss": 1.2208, "step": 18700 }, { "epoch": 1.4655956112852664, "grad_norm": 2.4648826122283936, "learning_rate": 7.055351315598604e-06, "loss": 1.2489, "step": 18701 }, { "epoch": 1.4656739811912225, "grad_norm": 2.3049652576446533, "learning_rate": 7.055070207707766e-06, "loss": 1.3541, "step": 18702 }, { "epoch": 1.4657523510971786, "grad_norm": 3.077383041381836, "learning_rate": 7.054789092000496e-06, "loss": 1.2787, "step": 18703 }, { "epoch": 1.4658307210031347, "grad_norm": 2.1071553230285645, "learning_rate": 7.054507968477864e-06, "loss": 1.1141, "step": 18704 }, { "epoch": 1.4659090909090908, "grad_norm": 7.637364864349365, "learning_rate": 7.054226837140939e-06, "loss": 1.219, "step": 18705 }, { "epoch": 1.465987460815047, "grad_norm": 3.69821834564209, "learning_rate": 7.053945697990789e-06, "loss": 1.2989, "step": 18706 }, { "epoch": 1.466065830721003, "grad_norm": 3.072134494781494, "learning_rate": 7.053664551028486e-06, "loss": 1.4394, "step": 18707 }, { "epoch": 1.4661442006269594, "grad_norm": 3.0603413581848145, "learning_rate": 7.053383396255098e-06, "loss": 1.3931, "step": 18708 }, { "epoch": 1.4662225705329153, "grad_norm": 2.431607246398926, "learning_rate": 7.053102233671694e-06, "loss": 0.8831, "step": 18709 }, { "epoch": 1.4663009404388716, "grad_norm": 2.518925666809082, "learning_rate": 7.052821063279343e-06, "loss": 1.2527, "step": 18710 }, { "epoch": 1.4663793103448275, "grad_norm": 2.9458093643188477, "learning_rate": 7.052539885079116e-06, "loss": 1.5516, "step": 18711 }, { "epoch": 1.4664576802507838, "grad_norm": 2.817054033279419, "learning_rate": 7.052258699072081e-06, "loss": 1.2894, "step": 18712 }, { "epoch": 1.4665360501567397, "grad_norm": 2.5560953617095947, "learning_rate": 7.051977505259308e-06, "loss": 0.7301, "step": 18713 }, { "epoch": 1.466614420062696, "grad_norm": 1.981782078742981, "learning_rate": 7.051696303641866e-06, "loss": 1.4057, "step": 18714 }, { "epoch": 1.466692789968652, "grad_norm": 6.580178260803223, "learning_rate": 7.051415094220826e-06, "loss": 0.9464, "step": 18715 }, { "epoch": 1.4667711598746083, "grad_norm": 2.6951894760131836, "learning_rate": 7.051133876997256e-06, "loss": 1.3396, "step": 18716 }, { "epoch": 1.4668495297805642, "grad_norm": 5.33828592300415, "learning_rate": 7.050852651972226e-06, "loss": 1.7058, "step": 18717 }, { "epoch": 1.4669278996865205, "grad_norm": 1.8343749046325684, "learning_rate": 7.050571419146806e-06, "loss": 1.3754, "step": 18718 }, { "epoch": 1.4670062695924764, "grad_norm": 2.4296934604644775, "learning_rate": 7.050290178522067e-06, "loss": 1.2514, "step": 18719 }, { "epoch": 1.4670846394984327, "grad_norm": 2.671271562576294, "learning_rate": 7.050008930099076e-06, "loss": 1.5688, "step": 18720 }, { "epoch": 1.4671630094043886, "grad_norm": 2.4496564865112305, "learning_rate": 7.049727673878905e-06, "loss": 1.2251, "step": 18721 }, { "epoch": 1.467241379310345, "grad_norm": 4.560339450836182, "learning_rate": 7.0494464098626225e-06, "loss": 1.5272, "step": 18722 }, { "epoch": 1.4673197492163008, "grad_norm": 5.202225685119629, "learning_rate": 7.049165138051297e-06, "loss": 1.7268, "step": 18723 }, { "epoch": 1.4673981191222571, "grad_norm": 2.6475582122802734, "learning_rate": 7.048883858446001e-06, "loss": 1.0815, "step": 18724 }, { "epoch": 1.467476489028213, "grad_norm": 2.4453442096710205, "learning_rate": 7.048602571047802e-06, "loss": 1.3719, "step": 18725 }, { "epoch": 1.4675548589341694, "grad_norm": 2.9983766078948975, "learning_rate": 7.048321275857773e-06, "loss": 1.4904, "step": 18726 }, { "epoch": 1.4676332288401253, "grad_norm": 3.0980217456817627, "learning_rate": 7.048039972876981e-06, "loss": 1.1742, "step": 18727 }, { "epoch": 1.4677115987460816, "grad_norm": 3.393832206726074, "learning_rate": 7.0477586621064965e-06, "loss": 1.1759, "step": 18728 }, { "epoch": 1.4677899686520375, "grad_norm": 3.0244054794311523, "learning_rate": 7.0474773435473895e-06, "loss": 1.1767, "step": 18729 }, { "epoch": 1.4678683385579938, "grad_norm": 3.716050148010254, "learning_rate": 7.047196017200731e-06, "loss": 1.3396, "step": 18730 }, { "epoch": 1.46794670846395, "grad_norm": 3.6513936519622803, "learning_rate": 7.046914683067591e-06, "loss": 1.3692, "step": 18731 }, { "epoch": 1.468025078369906, "grad_norm": 10.17541790008545, "learning_rate": 7.046633341149038e-06, "loss": 0.7981, "step": 18732 }, { "epoch": 1.4681034482758621, "grad_norm": 8.349817276000977, "learning_rate": 7.046351991446143e-06, "loss": 1.3007, "step": 18733 }, { "epoch": 1.4681818181818183, "grad_norm": 2.8829383850097656, "learning_rate": 7.046070633959977e-06, "loss": 1.4336, "step": 18734 }, { "epoch": 1.4682601880877744, "grad_norm": 3.5467212200164795, "learning_rate": 7.0457892686916075e-06, "loss": 1.4757, "step": 18735 }, { "epoch": 1.4683385579937305, "grad_norm": 4.976625919342041, "learning_rate": 7.045507895642109e-06, "loss": 1.3155, "step": 18736 }, { "epoch": 1.4684169278996866, "grad_norm": 3.002105236053467, "learning_rate": 7.045226514812548e-06, "loss": 1.5657, "step": 18737 }, { "epoch": 1.4684952978056427, "grad_norm": 2.292170763015747, "learning_rate": 7.044945126203995e-06, "loss": 1.1446, "step": 18738 }, { "epoch": 1.4685736677115988, "grad_norm": 2.3117973804473877, "learning_rate": 7.044663729817521e-06, "loss": 1.4992, "step": 18739 }, { "epoch": 1.468652037617555, "grad_norm": 2.4293620586395264, "learning_rate": 7.0443823256541966e-06, "loss": 1.3126, "step": 18740 }, { "epoch": 1.468730407523511, "grad_norm": 2.9653544425964355, "learning_rate": 7.044100913715093e-06, "loss": 1.5266, "step": 18741 }, { "epoch": 1.4688087774294671, "grad_norm": 3.5819814205169678, "learning_rate": 7.043819494001278e-06, "loss": 1.5524, "step": 18742 }, { "epoch": 1.4688871473354232, "grad_norm": 2.3183932304382324, "learning_rate": 7.043538066513824e-06, "loss": 1.2624, "step": 18743 }, { "epoch": 1.4689655172413794, "grad_norm": 4.647316932678223, "learning_rate": 7.043256631253801e-06, "loss": 1.0517, "step": 18744 }, { "epoch": 1.4690438871473355, "grad_norm": 7.046375751495361, "learning_rate": 7.04297518822228e-06, "loss": 1.1111, "step": 18745 }, { "epoch": 1.4691222570532916, "grad_norm": 2.0983049869537354, "learning_rate": 7.04269373742033e-06, "loss": 1.5661, "step": 18746 }, { "epoch": 1.4692006269592477, "grad_norm": 3.9114480018615723, "learning_rate": 7.042412278849021e-06, "loss": 1.6822, "step": 18747 }, { "epoch": 1.4692789968652038, "grad_norm": 2.7550487518310547, "learning_rate": 7.042130812509427e-06, "loss": 1.5558, "step": 18748 }, { "epoch": 1.46935736677116, "grad_norm": 7.237314701080322, "learning_rate": 7.041849338402614e-06, "loss": 1.2889, "step": 18749 }, { "epoch": 1.469435736677116, "grad_norm": 6.121330738067627, "learning_rate": 7.041567856529657e-06, "loss": 1.4407, "step": 18750 }, { "epoch": 1.4695141065830721, "grad_norm": 2.7621400356292725, "learning_rate": 7.041286366891622e-06, "loss": 1.6459, "step": 18751 }, { "epoch": 1.4695924764890282, "grad_norm": 3.021165609359741, "learning_rate": 7.041004869489585e-06, "loss": 1.4369, "step": 18752 }, { "epoch": 1.4696708463949844, "grad_norm": 3.038992166519165, "learning_rate": 7.040723364324611e-06, "loss": 1.2752, "step": 18753 }, { "epoch": 1.4697492163009405, "grad_norm": 2.8415656089782715, "learning_rate": 7.040441851397776e-06, "loss": 1.3458, "step": 18754 }, { "epoch": 1.4698275862068966, "grad_norm": 3.706987142562866, "learning_rate": 7.040160330710147e-06, "loss": 1.183, "step": 18755 }, { "epoch": 1.4699059561128527, "grad_norm": 2.8174853324890137, "learning_rate": 7.039878802262796e-06, "loss": 1.425, "step": 18756 }, { "epoch": 1.4699843260188088, "grad_norm": 2.1585514545440674, "learning_rate": 7.039597266056795e-06, "loss": 1.3481, "step": 18757 }, { "epoch": 1.470062695924765, "grad_norm": 3.4158644676208496, "learning_rate": 7.0393157220932115e-06, "loss": 1.9712, "step": 18758 }, { "epoch": 1.470141065830721, "grad_norm": 3.2336373329162598, "learning_rate": 7.039034170373119e-06, "loss": 1.4084, "step": 18759 }, { "epoch": 1.4702194357366771, "grad_norm": 3.887178659439087, "learning_rate": 7.038752610897589e-06, "loss": 1.3813, "step": 18760 }, { "epoch": 1.4702978056426332, "grad_norm": 2.9005885124206543, "learning_rate": 7.0384710436676895e-06, "loss": 1.1348, "step": 18761 }, { "epoch": 1.4703761755485893, "grad_norm": 2.690000295639038, "learning_rate": 7.038189468684495e-06, "loss": 1.335, "step": 18762 }, { "epoch": 1.4704545454545455, "grad_norm": 2.764327049255371, "learning_rate": 7.037907885949074e-06, "loss": 1.6073, "step": 18763 }, { "epoch": 1.4705329153605016, "grad_norm": 2.925410032272339, "learning_rate": 7.037626295462498e-06, "loss": 1.5947, "step": 18764 }, { "epoch": 1.4706112852664577, "grad_norm": 2.9436581134796143, "learning_rate": 7.037344697225839e-06, "loss": 1.5825, "step": 18765 }, { "epoch": 1.4706896551724138, "grad_norm": 2.9056320190429688, "learning_rate": 7.037063091240166e-06, "loss": 1.6933, "step": 18766 }, { "epoch": 1.47076802507837, "grad_norm": 2.4602246284484863, "learning_rate": 7.036781477506551e-06, "loss": 1.1856, "step": 18767 }, { "epoch": 1.470846394984326, "grad_norm": 3.306603193283081, "learning_rate": 7.036499856026066e-06, "loss": 1.3438, "step": 18768 }, { "epoch": 1.4709247648902821, "grad_norm": 5.065360069274902, "learning_rate": 7.036218226799782e-06, "loss": 0.9812, "step": 18769 }, { "epoch": 1.4710031347962382, "grad_norm": 2.79463267326355, "learning_rate": 7.035936589828769e-06, "loss": 1.7511, "step": 18770 }, { "epoch": 1.4710815047021943, "grad_norm": 3.153960943222046, "learning_rate": 7.035654945114101e-06, "loss": 1.2629, "step": 18771 }, { "epoch": 1.4711598746081505, "grad_norm": 3.3053808212280273, "learning_rate": 7.035373292656844e-06, "loss": 1.5461, "step": 18772 }, { "epoch": 1.4712382445141066, "grad_norm": 2.1121771335601807, "learning_rate": 7.035091632458075e-06, "loss": 1.306, "step": 18773 }, { "epoch": 1.4713166144200627, "grad_norm": 1.9597992897033691, "learning_rate": 7.03480996451886e-06, "loss": 1.4183, "step": 18774 }, { "epoch": 1.4713949843260188, "grad_norm": 8.349935531616211, "learning_rate": 7.034528288840276e-06, "loss": 1.6575, "step": 18775 }, { "epoch": 1.471473354231975, "grad_norm": 5.975426197052002, "learning_rate": 7.034246605423389e-06, "loss": 1.2743, "step": 18776 }, { "epoch": 1.471551724137931, "grad_norm": 3.4273605346679688, "learning_rate": 7.033964914269275e-06, "loss": 1.4358, "step": 18777 }, { "epoch": 1.4716300940438871, "grad_norm": 4.201061725616455, "learning_rate": 7.033683215379002e-06, "loss": 1.4812, "step": 18778 }, { "epoch": 1.4717084639498432, "grad_norm": 3.0970771312713623, "learning_rate": 7.033401508753642e-06, "loss": 1.2715, "step": 18779 }, { "epoch": 1.4717868338557993, "grad_norm": 2.1737301349639893, "learning_rate": 7.0331197943942685e-06, "loss": 1.0957, "step": 18780 }, { "epoch": 1.4718652037617554, "grad_norm": 3.3013720512390137, "learning_rate": 7.032838072301951e-06, "loss": 1.1881, "step": 18781 }, { "epoch": 1.4719435736677116, "grad_norm": 3.221196174621582, "learning_rate": 7.032556342477761e-06, "loss": 1.6273, "step": 18782 }, { "epoch": 1.4720219435736677, "grad_norm": 2.8429739475250244, "learning_rate": 7.032274604922771e-06, "loss": 1.6063, "step": 18783 }, { "epoch": 1.4721003134796238, "grad_norm": 2.0764055252075195, "learning_rate": 7.031992859638052e-06, "loss": 1.3847, "step": 18784 }, { "epoch": 1.47217868338558, "grad_norm": 2.9543404579162598, "learning_rate": 7.031711106624677e-06, "loss": 1.5391, "step": 18785 }, { "epoch": 1.472257053291536, "grad_norm": 4.456919193267822, "learning_rate": 7.031429345883715e-06, "loss": 1.0638, "step": 18786 }, { "epoch": 1.472335423197492, "grad_norm": 2.6750402450561523, "learning_rate": 7.03114757741624e-06, "loss": 1.2619, "step": 18787 }, { "epoch": 1.4724137931034482, "grad_norm": 3.441014051437378, "learning_rate": 7.0308658012233234e-06, "loss": 1.9997, "step": 18788 }, { "epoch": 1.4724921630094043, "grad_norm": 7.329529762268066, "learning_rate": 7.030584017306035e-06, "loss": 0.9835, "step": 18789 }, { "epoch": 1.4725705329153604, "grad_norm": 2.442694664001465, "learning_rate": 7.03030222566545e-06, "loss": 1.6817, "step": 18790 }, { "epoch": 1.4726489028213166, "grad_norm": 2.7932324409484863, "learning_rate": 7.030020426302635e-06, "loss": 1.0287, "step": 18791 }, { "epoch": 1.4727272727272727, "grad_norm": 2.907111167907715, "learning_rate": 7.029738619218668e-06, "loss": 1.536, "step": 18792 }, { "epoch": 1.4728056426332288, "grad_norm": 3.935609817504883, "learning_rate": 7.0294568044146165e-06, "loss": 1.3752, "step": 18793 }, { "epoch": 1.4728840125391849, "grad_norm": 5.996408462524414, "learning_rate": 7.029174981891553e-06, "loss": 1.6353, "step": 18794 }, { "epoch": 1.472962382445141, "grad_norm": 2.3360610008239746, "learning_rate": 7.028893151650551e-06, "loss": 1.343, "step": 18795 }, { "epoch": 1.473040752351097, "grad_norm": 3.554523229598999, "learning_rate": 7.028611313692681e-06, "loss": 1.2552, "step": 18796 }, { "epoch": 1.4731191222570532, "grad_norm": 2.5737240314483643, "learning_rate": 7.028329468019018e-06, "loss": 1.7374, "step": 18797 }, { "epoch": 1.4731974921630093, "grad_norm": 2.0028889179229736, "learning_rate": 7.028047614630628e-06, "loss": 1.5008, "step": 18798 }, { "epoch": 1.4732758620689654, "grad_norm": 2.980045795440674, "learning_rate": 7.027765753528589e-06, "loss": 1.3304, "step": 18799 }, { "epoch": 1.4733542319749215, "grad_norm": 3.725830316543579, "learning_rate": 7.02748388471397e-06, "loss": 1.3405, "step": 18800 }, { "epoch": 1.4734326018808779, "grad_norm": 3.642458200454712, "learning_rate": 7.027202008187843e-06, "loss": 1.5202, "step": 18801 }, { "epoch": 1.4735109717868338, "grad_norm": 6.277907848358154, "learning_rate": 7.026920123951282e-06, "loss": 1.1388, "step": 18802 }, { "epoch": 1.47358934169279, "grad_norm": 2.90464186668396, "learning_rate": 7.026638232005357e-06, "loss": 1.2935, "step": 18803 }, { "epoch": 1.473667711598746, "grad_norm": 2.8014791011810303, "learning_rate": 7.026356332351141e-06, "loss": 1.3103, "step": 18804 }, { "epoch": 1.4737460815047023, "grad_norm": 5.2584075927734375, "learning_rate": 7.026074424989706e-06, "loss": 1.2764, "step": 18805 }, { "epoch": 1.4738244514106582, "grad_norm": 3.8235280513763428, "learning_rate": 7.025792509922127e-06, "loss": 1.464, "step": 18806 }, { "epoch": 1.4739028213166145, "grad_norm": 2.307107448577881, "learning_rate": 7.025510587149471e-06, "loss": 1.5449, "step": 18807 }, { "epoch": 1.4739811912225704, "grad_norm": 2.7052228450775146, "learning_rate": 7.025228656672814e-06, "loss": 1.2837, "step": 18808 }, { "epoch": 1.4740595611285268, "grad_norm": 6.8056464195251465, "learning_rate": 7.024946718493229e-06, "loss": 1.7783, "step": 18809 }, { "epoch": 1.4741379310344827, "grad_norm": 3.5707504749298096, "learning_rate": 7.024664772611784e-06, "loss": 1.1428, "step": 18810 }, { "epoch": 1.474216300940439, "grad_norm": 3.186516761779785, "learning_rate": 7.024382819029555e-06, "loss": 1.1327, "step": 18811 }, { "epoch": 1.4742946708463949, "grad_norm": 3.7665855884552, "learning_rate": 7.024100857747614e-06, "loss": 1.5076, "step": 18812 }, { "epoch": 1.4743730407523512, "grad_norm": 2.602057695388794, "learning_rate": 7.0238188887670335e-06, "loss": 1.2011, "step": 18813 }, { "epoch": 1.474451410658307, "grad_norm": 2.2867133617401123, "learning_rate": 7.023536912088884e-06, "loss": 1.0848, "step": 18814 }, { "epoch": 1.4745297805642634, "grad_norm": 2.299309492111206, "learning_rate": 7.02325492771424e-06, "loss": 0.9971, "step": 18815 }, { "epoch": 1.4746081504702193, "grad_norm": 4.577622890472412, "learning_rate": 7.022972935644175e-06, "loss": 0.8888, "step": 18816 }, { "epoch": 1.4746865203761756, "grad_norm": 2.844586133956909, "learning_rate": 7.022690935879758e-06, "loss": 1.2216, "step": 18817 }, { "epoch": 1.4747648902821315, "grad_norm": 2.4987480640411377, "learning_rate": 7.022408928422064e-06, "loss": 1.5527, "step": 18818 }, { "epoch": 1.4748432601880879, "grad_norm": 2.170804977416992, "learning_rate": 7.022126913272166e-06, "loss": 0.8825, "step": 18819 }, { "epoch": 1.4749216300940438, "grad_norm": 4.5149149894714355, "learning_rate": 7.021844890431136e-06, "loss": 1.5818, "step": 18820 }, { "epoch": 1.475, "grad_norm": 9.582289695739746, "learning_rate": 7.021562859900045e-06, "loss": 1.1749, "step": 18821 }, { "epoch": 1.475078369905956, "grad_norm": 3.5379014015197754, "learning_rate": 7.021280821679969e-06, "loss": 1.2796, "step": 18822 }, { "epoch": 1.4751567398119123, "grad_norm": 2.310340166091919, "learning_rate": 7.020998775771979e-06, "loss": 1.3471, "step": 18823 }, { "epoch": 1.4752351097178684, "grad_norm": 2.3264527320861816, "learning_rate": 7.020716722177146e-06, "loss": 1.1278, "step": 18824 }, { "epoch": 1.4753134796238245, "grad_norm": 2.3203022480010986, "learning_rate": 7.020434660896546e-06, "loss": 1.3617, "step": 18825 }, { "epoch": 1.4753918495297806, "grad_norm": 4.567938804626465, "learning_rate": 7.020152591931248e-06, "loss": 1.5382, "step": 18826 }, { "epoch": 1.4754702194357368, "grad_norm": 3.5414106845855713, "learning_rate": 7.01987051528233e-06, "loss": 2.0159, "step": 18827 }, { "epoch": 1.4755485893416929, "grad_norm": 8.1693754196167, "learning_rate": 7.019588430950861e-06, "loss": 1.5168, "step": 18828 }, { "epoch": 1.475626959247649, "grad_norm": 2.5752785205841064, "learning_rate": 7.019306338937914e-06, "loss": 1.2347, "step": 18829 }, { "epoch": 1.475705329153605, "grad_norm": 2.064775228500366, "learning_rate": 7.019024239244564e-06, "loss": 1.4372, "step": 18830 }, { "epoch": 1.4757836990595612, "grad_norm": 4.439849853515625, "learning_rate": 7.018742131871882e-06, "loss": 0.9048, "step": 18831 }, { "epoch": 1.4758620689655173, "grad_norm": 2.795032024383545, "learning_rate": 7.018460016820943e-06, "loss": 1.1851, "step": 18832 }, { "epoch": 1.4759404388714734, "grad_norm": 9.745865821838379, "learning_rate": 7.018177894092817e-06, "loss": 1.2292, "step": 18833 }, { "epoch": 1.4760188087774295, "grad_norm": 3.231041431427002, "learning_rate": 7.017895763688581e-06, "loss": 1.2618, "step": 18834 }, { "epoch": 1.4760971786833856, "grad_norm": 2.6812102794647217, "learning_rate": 7.017613625609306e-06, "loss": 1.3682, "step": 18835 }, { "epoch": 1.4761755485893417, "grad_norm": 2.3542909622192383, "learning_rate": 7.017331479856063e-06, "loss": 2.1021, "step": 18836 }, { "epoch": 1.4762539184952979, "grad_norm": 2.768918991088867, "learning_rate": 7.017049326429928e-06, "loss": 1.8218, "step": 18837 }, { "epoch": 1.476332288401254, "grad_norm": 2.78395414352417, "learning_rate": 7.016767165331974e-06, "loss": 1.1733, "step": 18838 }, { "epoch": 1.47641065830721, "grad_norm": 2.698317527770996, "learning_rate": 7.016484996563273e-06, "loss": 1.7249, "step": 18839 }, { "epoch": 1.4764890282131662, "grad_norm": 3.650233507156372, "learning_rate": 7.016202820124899e-06, "loss": 1.6306, "step": 18840 }, { "epoch": 1.4765673981191223, "grad_norm": 2.000305652618408, "learning_rate": 7.015920636017924e-06, "loss": 1.1776, "step": 18841 }, { "epoch": 1.4766457680250784, "grad_norm": 3.167116641998291, "learning_rate": 7.015638444243426e-06, "loss": 1.361, "step": 18842 }, { "epoch": 1.4767241379310345, "grad_norm": 3.043318271636963, "learning_rate": 7.01535624480247e-06, "loss": 1.8147, "step": 18843 }, { "epoch": 1.4768025078369906, "grad_norm": 4.195157527923584, "learning_rate": 7.015074037696137e-06, "loss": 1.6554, "step": 18844 }, { "epoch": 1.4768808777429467, "grad_norm": 3.07230544090271, "learning_rate": 7.014791822925495e-06, "loss": 1.0488, "step": 18845 }, { "epoch": 1.4769592476489029, "grad_norm": 2.6200852394104004, "learning_rate": 7.014509600491623e-06, "loss": 1.3828, "step": 18846 }, { "epoch": 1.477037617554859, "grad_norm": 3.004706621170044, "learning_rate": 7.014227370395589e-06, "loss": 1.7795, "step": 18847 }, { "epoch": 1.477115987460815, "grad_norm": 2.804715156555176, "learning_rate": 7.0139451326384686e-06, "loss": 1.5157, "step": 18848 }, { "epoch": 1.4771943573667712, "grad_norm": 3.69884991645813, "learning_rate": 7.0136628872213365e-06, "loss": 1.3882, "step": 18849 }, { "epoch": 1.4772727272727273, "grad_norm": 2.215334892272949, "learning_rate": 7.013380634145264e-06, "loss": 1.3079, "step": 18850 }, { "epoch": 1.4773510971786834, "grad_norm": 4.085166931152344, "learning_rate": 7.0130983734113266e-06, "loss": 1.6169, "step": 18851 }, { "epoch": 1.4774294670846395, "grad_norm": 2.717249870300293, "learning_rate": 7.012816105020596e-06, "loss": 1.0842, "step": 18852 }, { "epoch": 1.4775078369905956, "grad_norm": 9.482766151428223, "learning_rate": 7.0125338289741485e-06, "loss": 0.7401, "step": 18853 }, { "epoch": 1.4775862068965517, "grad_norm": 6.273816108703613, "learning_rate": 7.012251545273053e-06, "loss": 1.5072, "step": 18854 }, { "epoch": 1.4776645768025078, "grad_norm": 3.242489814758301, "learning_rate": 7.011969253918389e-06, "loss": 1.5093, "step": 18855 }, { "epoch": 1.477742946708464, "grad_norm": 2.1857900619506836, "learning_rate": 7.011686954911226e-06, "loss": 1.3959, "step": 18856 }, { "epoch": 1.47782131661442, "grad_norm": 4.278509140014648, "learning_rate": 7.01140464825264e-06, "loss": 1.472, "step": 18857 }, { "epoch": 1.4778996865203762, "grad_norm": 2.427645683288574, "learning_rate": 7.011122333943703e-06, "loss": 1.4887, "step": 18858 }, { "epoch": 1.4779780564263323, "grad_norm": 2.1173653602600098, "learning_rate": 7.010840011985489e-06, "loss": 1.5617, "step": 18859 }, { "epoch": 1.4780564263322884, "grad_norm": 2.710629940032959, "learning_rate": 7.010557682379073e-06, "loss": 1.7025, "step": 18860 }, { "epoch": 1.4781347962382445, "grad_norm": 2.4646787643432617, "learning_rate": 7.010275345125529e-06, "loss": 1.3907, "step": 18861 }, { "epoch": 1.4782131661442006, "grad_norm": 7.260997772216797, "learning_rate": 7.009993000225931e-06, "loss": 1.1405, "step": 18862 }, { "epoch": 1.4782915360501567, "grad_norm": 2.6199769973754883, "learning_rate": 7.00971064768135e-06, "loss": 1.4784, "step": 18863 }, { "epoch": 1.4783699059561128, "grad_norm": 2.253598213195801, "learning_rate": 7.009428287492863e-06, "loss": 1.5823, "step": 18864 }, { "epoch": 1.478448275862069, "grad_norm": 5.459826469421387, "learning_rate": 7.009145919661542e-06, "loss": 1.4246, "step": 18865 }, { "epoch": 1.478526645768025, "grad_norm": 2.816023588180542, "learning_rate": 7.008863544188462e-06, "loss": 1.2976, "step": 18866 }, { "epoch": 1.4786050156739812, "grad_norm": 5.000065803527832, "learning_rate": 7.008581161074698e-06, "loss": 1.1648, "step": 18867 }, { "epoch": 1.4786833855799373, "grad_norm": 2.8695034980773926, "learning_rate": 7.008298770321323e-06, "loss": 1.359, "step": 18868 }, { "epoch": 1.4787617554858934, "grad_norm": 3.295102119445801, "learning_rate": 7.008016371929409e-06, "loss": 1.2891, "step": 18869 }, { "epoch": 1.4788401253918495, "grad_norm": 1.7770322561264038, "learning_rate": 7.0077339659000345e-06, "loss": 1.3517, "step": 18870 }, { "epoch": 1.4789184952978056, "grad_norm": 3.710691213607788, "learning_rate": 7.007451552234269e-06, "loss": 1.8577, "step": 18871 }, { "epoch": 1.4789968652037617, "grad_norm": 3.1410164833068848, "learning_rate": 7.0071691309331915e-06, "loss": 1.8559, "step": 18872 }, { "epoch": 1.4790752351097178, "grad_norm": 18.302888870239258, "learning_rate": 7.006886701997871e-06, "loss": 1.5087, "step": 18873 }, { "epoch": 1.479153605015674, "grad_norm": 3.558577060699463, "learning_rate": 7.006604265429386e-06, "loss": 1.5922, "step": 18874 }, { "epoch": 1.47923197492163, "grad_norm": 2.3030543327331543, "learning_rate": 7.006321821228809e-06, "loss": 1.2717, "step": 18875 }, { "epoch": 1.4793103448275862, "grad_norm": 4.251152038574219, "learning_rate": 7.006039369397213e-06, "loss": 1.6154, "step": 18876 }, { "epoch": 1.4793887147335423, "grad_norm": 1.9646327495574951, "learning_rate": 7.005756909935675e-06, "loss": 1.3473, "step": 18877 }, { "epoch": 1.4794670846394984, "grad_norm": 3.5045156478881836, "learning_rate": 7.0054744428452655e-06, "loss": 1.0917, "step": 18878 }, { "epoch": 1.4795454545454545, "grad_norm": 2.7035746574401855, "learning_rate": 7.005191968127065e-06, "loss": 1.1152, "step": 18879 }, { "epoch": 1.4796238244514106, "grad_norm": 9.063811302185059, "learning_rate": 7.004909485782141e-06, "loss": 1.2453, "step": 18880 }, { "epoch": 1.4797021943573667, "grad_norm": 2.4918994903564453, "learning_rate": 7.004626995811573e-06, "loss": 1.1523, "step": 18881 }, { "epoch": 1.4797805642633228, "grad_norm": 4.0114946365356445, "learning_rate": 7.004344498216433e-06, "loss": 1.4012, "step": 18882 }, { "epoch": 1.479858934169279, "grad_norm": 3.7936666011810303, "learning_rate": 7.004061992997796e-06, "loss": 1.5436, "step": 18883 }, { "epoch": 1.479937304075235, "grad_norm": 6.77782678604126, "learning_rate": 7.003779480156736e-06, "loss": 0.6305, "step": 18884 }, { "epoch": 1.4800156739811912, "grad_norm": 3.1781671047210693, "learning_rate": 7.0034969596943295e-06, "loss": 1.2608, "step": 18885 }, { "epoch": 1.4800940438871473, "grad_norm": 2.2737348079681396, "learning_rate": 7.003214431611648e-06, "loss": 1.3577, "step": 18886 }, { "epoch": 1.4801724137931034, "grad_norm": 2.259486436843872, "learning_rate": 7.002931895909768e-06, "loss": 1.5891, "step": 18887 }, { "epoch": 1.4802507836990595, "grad_norm": 3.702347993850708, "learning_rate": 7.0026493525897644e-06, "loss": 1.415, "step": 18888 }, { "epoch": 1.4803291536050156, "grad_norm": 6.8094892501831055, "learning_rate": 7.002366801652711e-06, "loss": 1.3145, "step": 18889 }, { "epoch": 1.4804075235109717, "grad_norm": 2.9190633296966553, "learning_rate": 7.002084243099683e-06, "loss": 1.3217, "step": 18890 }, { "epoch": 1.4804858934169278, "grad_norm": 2.340735673904419, "learning_rate": 7.001801676931754e-06, "loss": 1.2042, "step": 18891 }, { "epoch": 1.480564263322884, "grad_norm": 2.292315721511841, "learning_rate": 7.00151910315e-06, "loss": 1.5779, "step": 18892 }, { "epoch": 1.48064263322884, "grad_norm": 2.2216827869415283, "learning_rate": 7.0012365217554964e-06, "loss": 1.3763, "step": 18893 }, { "epoch": 1.4807210031347964, "grad_norm": 1.8233052492141724, "learning_rate": 7.000953932749317e-06, "loss": 1.5379, "step": 18894 }, { "epoch": 1.4807993730407523, "grad_norm": 4.172451019287109, "learning_rate": 7.000671336132535e-06, "loss": 1.1762, "step": 18895 }, { "epoch": 1.4808777429467086, "grad_norm": 1.65017569065094, "learning_rate": 7.000388731906228e-06, "loss": 1.1471, "step": 18896 }, { "epoch": 1.4809561128526645, "grad_norm": 3.0033559799194336, "learning_rate": 7.0001061200714695e-06, "loss": 1.3824, "step": 18897 }, { "epoch": 1.4810344827586208, "grad_norm": 2.5106711387634277, "learning_rate": 6.999823500629335e-06, "loss": 1.4561, "step": 18898 }, { "epoch": 1.4811128526645767, "grad_norm": 2.0508368015289307, "learning_rate": 6.999540873580898e-06, "loss": 1.451, "step": 18899 }, { "epoch": 1.481191222570533, "grad_norm": 3.0162672996520996, "learning_rate": 6.9992582389272355e-06, "loss": 1.5756, "step": 18900 }, { "epoch": 1.481269592476489, "grad_norm": 4.70667028427124, "learning_rate": 6.998975596669421e-06, "loss": 1.1562, "step": 18901 }, { "epoch": 1.4813479623824453, "grad_norm": 2.6368510723114014, "learning_rate": 6.998692946808531e-06, "loss": 1.3662, "step": 18902 }, { "epoch": 1.4814263322884011, "grad_norm": 2.520890951156616, "learning_rate": 6.998410289345639e-06, "loss": 1.1412, "step": 18903 }, { "epoch": 1.4815047021943575, "grad_norm": 2.5057590007781982, "learning_rate": 6.998127624281821e-06, "loss": 0.9727, "step": 18904 }, { "epoch": 1.4815830721003134, "grad_norm": 8.30229377746582, "learning_rate": 6.997844951618151e-06, "loss": 1.6908, "step": 18905 }, { "epoch": 1.4816614420062697, "grad_norm": 2.409412384033203, "learning_rate": 6.9975622713557045e-06, "loss": 1.2868, "step": 18906 }, { "epoch": 1.4817398119122256, "grad_norm": 3.386953830718994, "learning_rate": 6.99727958349556e-06, "loss": 1.3785, "step": 18907 }, { "epoch": 1.481818181818182, "grad_norm": 3.8096392154693604, "learning_rate": 6.996996888038786e-06, "loss": 1.4667, "step": 18908 }, { "epoch": 1.4818965517241378, "grad_norm": 6.028322219848633, "learning_rate": 6.996714184986465e-06, "loss": 1.1121, "step": 18909 }, { "epoch": 1.4819749216300941, "grad_norm": 4.021376132965088, "learning_rate": 6.996431474339666e-06, "loss": 1.3317, "step": 18910 }, { "epoch": 1.48205329153605, "grad_norm": 3.4268648624420166, "learning_rate": 6.996148756099469e-06, "loss": 1.2285, "step": 18911 }, { "epoch": 1.4821316614420064, "grad_norm": 2.3130836486816406, "learning_rate": 6.995866030266947e-06, "loss": 1.038, "step": 18912 }, { "epoch": 1.4822100313479623, "grad_norm": 5.835732460021973, "learning_rate": 6.9955832968431755e-06, "loss": 1.8636, "step": 18913 }, { "epoch": 1.4822884012539186, "grad_norm": 3.252286434173584, "learning_rate": 6.995300555829229e-06, "loss": 1.3027, "step": 18914 }, { "epoch": 1.4823667711598745, "grad_norm": 7.926952362060547, "learning_rate": 6.995017807226186e-06, "loss": 1.6194, "step": 18915 }, { "epoch": 1.4824451410658308, "grad_norm": 2.6529922485351562, "learning_rate": 6.99473505103512e-06, "loss": 1.2416, "step": 18916 }, { "epoch": 1.4825235109717867, "grad_norm": 2.609936475753784, "learning_rate": 6.994452287257105e-06, "loss": 1.7539, "step": 18917 }, { "epoch": 1.482601880877743, "grad_norm": 5.623257637023926, "learning_rate": 6.994169515893219e-06, "loss": 1.2342, "step": 18918 }, { "epoch": 1.4826802507836991, "grad_norm": 2.8685495853424072, "learning_rate": 6.993886736944536e-06, "loss": 1.505, "step": 18919 }, { "epoch": 1.4827586206896552, "grad_norm": 2.423647403717041, "learning_rate": 6.993603950412132e-06, "loss": 1.0995, "step": 18920 }, { "epoch": 1.4828369905956114, "grad_norm": 2.9115376472473145, "learning_rate": 6.993321156297083e-06, "loss": 1.3665, "step": 18921 }, { "epoch": 1.4829153605015675, "grad_norm": 6.123973846435547, "learning_rate": 6.993038354600463e-06, "loss": 1.0178, "step": 18922 }, { "epoch": 1.4829937304075236, "grad_norm": 3.9283323287963867, "learning_rate": 6.992755545323348e-06, "loss": 1.4601, "step": 18923 }, { "epoch": 1.4830721003134797, "grad_norm": 4.063938140869141, "learning_rate": 6.992472728466818e-06, "loss": 1.6463, "step": 18924 }, { "epoch": 1.4831504702194358, "grad_norm": 2.649643659591675, "learning_rate": 6.992189904031942e-06, "loss": 1.6026, "step": 18925 }, { "epoch": 1.483228840125392, "grad_norm": 3.630591869354248, "learning_rate": 6.991907072019801e-06, "loss": 1.2243, "step": 18926 }, { "epoch": 1.483307210031348, "grad_norm": 3.7511308193206787, "learning_rate": 6.991624232431467e-06, "loss": 1.5736, "step": 18927 }, { "epoch": 1.4833855799373041, "grad_norm": 2.813140630722046, "learning_rate": 6.991341385268018e-06, "loss": 1.3976, "step": 18928 }, { "epoch": 1.4834639498432602, "grad_norm": 2.007941961288452, "learning_rate": 6.991058530530529e-06, "loss": 1.4728, "step": 18929 }, { "epoch": 1.4835423197492164, "grad_norm": 4.260928630828857, "learning_rate": 6.9907756682200756e-06, "loss": 1.6192, "step": 18930 }, { "epoch": 1.4836206896551725, "grad_norm": 2.4260826110839844, "learning_rate": 6.9904927983377355e-06, "loss": 0.9899, "step": 18931 }, { "epoch": 1.4836990595611286, "grad_norm": 3.5892624855041504, "learning_rate": 6.990209920884582e-06, "loss": 1.3391, "step": 18932 }, { "epoch": 1.4837774294670847, "grad_norm": 2.8192689418792725, "learning_rate": 6.989927035861693e-06, "loss": 0.9968, "step": 18933 }, { "epoch": 1.4838557993730408, "grad_norm": 3.8843133449554443, "learning_rate": 6.989644143270142e-06, "loss": 1.8989, "step": 18934 }, { "epoch": 1.483934169278997, "grad_norm": 2.285698413848877, "learning_rate": 6.989361243111008e-06, "loss": 1.3892, "step": 18935 }, { "epoch": 1.484012539184953, "grad_norm": 1.7292144298553467, "learning_rate": 6.9890783353853665e-06, "loss": 1.3361, "step": 18936 }, { "epoch": 1.4840909090909091, "grad_norm": 2.8101515769958496, "learning_rate": 6.988795420094291e-06, "loss": 1.6391, "step": 18937 }, { "epoch": 1.4841692789968652, "grad_norm": 3.1720194816589355, "learning_rate": 6.98851249723886e-06, "loss": 1.2943, "step": 18938 }, { "epoch": 1.4842476489028213, "grad_norm": 3.110236167907715, "learning_rate": 6.988229566820148e-06, "loss": 1.3558, "step": 18939 }, { "epoch": 1.4843260188087775, "grad_norm": 2.881937265396118, "learning_rate": 6.987946628839232e-06, "loss": 1.4251, "step": 18940 }, { "epoch": 1.4844043887147336, "grad_norm": 2.287078857421875, "learning_rate": 6.987663683297189e-06, "loss": 1.2265, "step": 18941 }, { "epoch": 1.4844827586206897, "grad_norm": 2.4236788749694824, "learning_rate": 6.987380730195093e-06, "loss": 1.4192, "step": 18942 }, { "epoch": 1.4845611285266458, "grad_norm": 3.543736457824707, "learning_rate": 6.987097769534024e-06, "loss": 1.5132, "step": 18943 }, { "epoch": 1.484639498432602, "grad_norm": 4.0696492195129395, "learning_rate": 6.986814801315053e-06, "loss": 1.8164, "step": 18944 }, { "epoch": 1.484717868338558, "grad_norm": 2.7036917209625244, "learning_rate": 6.9865318255392615e-06, "loss": 1.2938, "step": 18945 }, { "epoch": 1.4847962382445141, "grad_norm": 2.960935354232788, "learning_rate": 6.986248842207721e-06, "loss": 1.7286, "step": 18946 }, { "epoch": 1.4848746081504702, "grad_norm": 2.798316478729248, "learning_rate": 6.985965851321511e-06, "loss": 1.3697, "step": 18947 }, { "epoch": 1.4849529780564263, "grad_norm": 3.4082279205322266, "learning_rate": 6.985682852881708e-06, "loss": 1.5926, "step": 18948 }, { "epoch": 1.4850313479623825, "grad_norm": 2.3436684608459473, "learning_rate": 6.985399846889386e-06, "loss": 1.4591, "step": 18949 }, { "epoch": 1.4851097178683386, "grad_norm": 2.0760550498962402, "learning_rate": 6.985116833345622e-06, "loss": 1.2574, "step": 18950 }, { "epoch": 1.4851880877742947, "grad_norm": 2.9385640621185303, "learning_rate": 6.984833812251495e-06, "loss": 1.4014, "step": 18951 }, { "epoch": 1.4852664576802508, "grad_norm": 2.8409812450408936, "learning_rate": 6.98455078360808e-06, "loss": 1.3694, "step": 18952 }, { "epoch": 1.485344827586207, "grad_norm": 8.054047584533691, "learning_rate": 6.98426774741645e-06, "loss": 1.1216, "step": 18953 }, { "epoch": 1.485423197492163, "grad_norm": 2.7094876766204834, "learning_rate": 6.983984703677688e-06, "loss": 1.4805, "step": 18954 }, { "epoch": 1.4855015673981191, "grad_norm": 2.6157588958740234, "learning_rate": 6.983701652392864e-06, "loss": 1.5193, "step": 18955 }, { "epoch": 1.4855799373040752, "grad_norm": 2.3692660331726074, "learning_rate": 6.983418593563062e-06, "loss": 1.0289, "step": 18956 }, { "epoch": 1.4856583072100313, "grad_norm": 7.026470184326172, "learning_rate": 6.983135527189351e-06, "loss": 1.5019, "step": 18957 }, { "epoch": 1.4857366771159874, "grad_norm": 3.595508337020874, "learning_rate": 6.982852453272812e-06, "loss": 1.4981, "step": 18958 }, { "epoch": 1.4858150470219436, "grad_norm": 8.236085891723633, "learning_rate": 6.982569371814523e-06, "loss": 0.9695, "step": 18959 }, { "epoch": 1.4858934169278997, "grad_norm": 6.855197906494141, "learning_rate": 6.982286282815555e-06, "loss": 1.2814, "step": 18960 }, { "epoch": 1.4859717868338558, "grad_norm": 2.463562250137329, "learning_rate": 6.98200318627699e-06, "loss": 1.4157, "step": 18961 }, { "epoch": 1.486050156739812, "grad_norm": 3.537888526916504, "learning_rate": 6.981720082199901e-06, "loss": 0.8511, "step": 18962 }, { "epoch": 1.486128526645768, "grad_norm": 2.2532958984375, "learning_rate": 6.981436970585369e-06, "loss": 1.3904, "step": 18963 }, { "epoch": 1.486206896551724, "grad_norm": 2.0218374729156494, "learning_rate": 6.9811538514344675e-06, "loss": 1.2992, "step": 18964 }, { "epoch": 1.4862852664576802, "grad_norm": 1.9730262756347656, "learning_rate": 6.9808707247482745e-06, "loss": 1.1553, "step": 18965 }, { "epoch": 1.4863636363636363, "grad_norm": 3.022021532058716, "learning_rate": 6.9805875905278665e-06, "loss": 1.3098, "step": 18966 }, { "epoch": 1.4864420062695924, "grad_norm": 1.9927444458007812, "learning_rate": 6.980304448774321e-06, "loss": 1.1799, "step": 18967 }, { "epoch": 1.4865203761755486, "grad_norm": 5.486480712890625, "learning_rate": 6.980021299488714e-06, "loss": 1.1457, "step": 18968 }, { "epoch": 1.4865987460815047, "grad_norm": 5.947285175323486, "learning_rate": 6.979738142672123e-06, "loss": 1.2439, "step": 18969 }, { "epoch": 1.4866771159874608, "grad_norm": 2.955876350402832, "learning_rate": 6.979454978325625e-06, "loss": 0.9236, "step": 18970 }, { "epoch": 1.4867554858934169, "grad_norm": 2.3396852016448975, "learning_rate": 6.9791718064502976e-06, "loss": 1.3305, "step": 18971 }, { "epoch": 1.486833855799373, "grad_norm": 3.19663405418396, "learning_rate": 6.978888627047216e-06, "loss": 1.2944, "step": 18972 }, { "epoch": 1.486912225705329, "grad_norm": 2.983253002166748, "learning_rate": 6.97860544011746e-06, "loss": 1.6262, "step": 18973 }, { "epoch": 1.4869905956112852, "grad_norm": 5.007349491119385, "learning_rate": 6.978322245662104e-06, "loss": 1.6991, "step": 18974 }, { "epoch": 1.4870689655172413, "grad_norm": 3.921778440475464, "learning_rate": 6.978039043682228e-06, "loss": 1.7773, "step": 18975 }, { "epoch": 1.4871473354231974, "grad_norm": 8.270561218261719, "learning_rate": 6.977755834178905e-06, "loss": 1.55, "step": 18976 }, { "epoch": 1.4872257053291535, "grad_norm": 2.323427677154541, "learning_rate": 6.977472617153216e-06, "loss": 1.2701, "step": 18977 }, { "epoch": 1.4873040752351097, "grad_norm": 2.734978437423706, "learning_rate": 6.977189392606239e-06, "loss": 1.3493, "step": 18978 }, { "epoch": 1.4873824451410658, "grad_norm": 2.6083033084869385, "learning_rate": 6.9769061605390445e-06, "loss": 1.3131, "step": 18979 }, { "epoch": 1.4874608150470219, "grad_norm": 5.281616687774658, "learning_rate": 6.9766229209527185e-06, "loss": 1.3373, "step": 18980 }, { "epoch": 1.487539184952978, "grad_norm": 3.28590726852417, "learning_rate": 6.9763396738483304e-06, "loss": 1.4274, "step": 18981 }, { "epoch": 1.487617554858934, "grad_norm": 2.8051018714904785, "learning_rate": 6.976056419226965e-06, "loss": 1.8078, "step": 18982 }, { "epoch": 1.4876959247648902, "grad_norm": 2.2452242374420166, "learning_rate": 6.975773157089693e-06, "loss": 1.1003, "step": 18983 }, { "epoch": 1.4877742946708463, "grad_norm": 2.2602696418762207, "learning_rate": 6.9754898874375965e-06, "loss": 1.3534, "step": 18984 }, { "epoch": 1.4878526645768024, "grad_norm": 3.4302868843078613, "learning_rate": 6.97520661027175e-06, "loss": 0.9795, "step": 18985 }, { "epoch": 1.4879310344827585, "grad_norm": 1.8508845567703247, "learning_rate": 6.974923325593233e-06, "loss": 1.1695, "step": 18986 }, { "epoch": 1.4880094043887147, "grad_norm": 5.262948513031006, "learning_rate": 6.9746400334031216e-06, "loss": 1.4426, "step": 18987 }, { "epoch": 1.4880877742946708, "grad_norm": 5.0313873291015625, "learning_rate": 6.974356733702493e-06, "loss": 1.4564, "step": 18988 }, { "epoch": 1.488166144200627, "grad_norm": 2.6972053050994873, "learning_rate": 6.974073426492427e-06, "loss": 1.4285, "step": 18989 }, { "epoch": 1.488244514106583, "grad_norm": 4.654870986938477, "learning_rate": 6.9737901117739995e-06, "loss": 1.1967, "step": 18990 }, { "epoch": 1.4883228840125393, "grad_norm": 4.799703598022461, "learning_rate": 6.973506789548288e-06, "loss": 0.6917, "step": 18991 }, { "epoch": 1.4884012539184952, "grad_norm": 14.606062889099121, "learning_rate": 6.97322345981637e-06, "loss": 0.7116, "step": 18992 }, { "epoch": 1.4884796238244515, "grad_norm": 2.337524652481079, "learning_rate": 6.972940122579323e-06, "loss": 1.1152, "step": 18993 }, { "epoch": 1.4885579937304074, "grad_norm": 2.386012315750122, "learning_rate": 6.972656777838226e-06, "loss": 1.4574, "step": 18994 }, { "epoch": 1.4886363636363638, "grad_norm": 3.312549591064453, "learning_rate": 6.972373425594155e-06, "loss": 1.3479, "step": 18995 }, { "epoch": 1.4887147335423196, "grad_norm": 2.8953397274017334, "learning_rate": 6.9720900658481896e-06, "loss": 1.096, "step": 18996 }, { "epoch": 1.488793103448276, "grad_norm": 2.8915483951568604, "learning_rate": 6.971806698601406e-06, "loss": 1.0594, "step": 18997 }, { "epoch": 1.4888714733542319, "grad_norm": 1.930067777633667, "learning_rate": 6.971523323854882e-06, "loss": 1.5238, "step": 18998 }, { "epoch": 1.4889498432601882, "grad_norm": 9.28554630279541, "learning_rate": 6.971239941609696e-06, "loss": 1.1419, "step": 18999 }, { "epoch": 1.489028213166144, "grad_norm": 1.9072662591934204, "learning_rate": 6.970956551866925e-06, "loss": 1.1709, "step": 19000 }, { "epoch": 1.4891065830721004, "grad_norm": 2.698132038116455, "learning_rate": 6.97067315462765e-06, "loss": 1.3778, "step": 19001 }, { "epoch": 1.4891849529780563, "grad_norm": 8.76003646850586, "learning_rate": 6.970389749892945e-06, "loss": 0.3948, "step": 19002 }, { "epoch": 1.4892633228840126, "grad_norm": 2.6994917392730713, "learning_rate": 6.97010633766389e-06, "loss": 1.092, "step": 19003 }, { "epoch": 1.4893416927899685, "grad_norm": 2.14912748336792, "learning_rate": 6.9698229179415624e-06, "loss": 1.3658, "step": 19004 }, { "epoch": 1.4894200626959249, "grad_norm": 2.9307034015655518, "learning_rate": 6.9695394907270395e-06, "loss": 1.2278, "step": 19005 }, { "epoch": 1.4894984326018808, "grad_norm": 2.958488702774048, "learning_rate": 6.969256056021402e-06, "loss": 1.3948, "step": 19006 }, { "epoch": 1.489576802507837, "grad_norm": 3.2624480724334717, "learning_rate": 6.968972613825723e-06, "loss": 1.8948, "step": 19007 }, { "epoch": 1.489655172413793, "grad_norm": 2.3625009059906006, "learning_rate": 6.968689164141085e-06, "loss": 1.1974, "step": 19008 }, { "epoch": 1.4897335423197493, "grad_norm": 4.080782413482666, "learning_rate": 6.968405706968564e-06, "loss": 1.7284, "step": 19009 }, { "epoch": 1.4898119122257052, "grad_norm": 3.8414902687072754, "learning_rate": 6.96812224230924e-06, "loss": 1.3093, "step": 19010 }, { "epoch": 1.4898902821316615, "grad_norm": 2.371304512023926, "learning_rate": 6.967838770164189e-06, "loss": 1.3924, "step": 19011 }, { "epoch": 1.4899686520376176, "grad_norm": 3.307781934738159, "learning_rate": 6.96755529053449e-06, "loss": 1.1261, "step": 19012 }, { "epoch": 1.4900470219435737, "grad_norm": 2.846411943435669, "learning_rate": 6.967271803421221e-06, "loss": 1.2281, "step": 19013 }, { "epoch": 1.4901253918495299, "grad_norm": 2.6805856227874756, "learning_rate": 6.966988308825461e-06, "loss": 1.598, "step": 19014 }, { "epoch": 1.490203761755486, "grad_norm": 11.333165168762207, "learning_rate": 6.966704806748288e-06, "loss": 1.4818, "step": 19015 }, { "epoch": 1.490282131661442, "grad_norm": 4.758626461029053, "learning_rate": 6.966421297190779e-06, "loss": 0.964, "step": 19016 }, { "epoch": 1.4903605015673982, "grad_norm": 2.2147014141082764, "learning_rate": 6.9661377801540155e-06, "loss": 1.1079, "step": 19017 }, { "epoch": 1.4904388714733543, "grad_norm": 3.4243006706237793, "learning_rate": 6.965854255639071e-06, "loss": 1.3425, "step": 19018 }, { "epoch": 1.4905172413793104, "grad_norm": 3.9977471828460693, "learning_rate": 6.9655707236470295e-06, "loss": 1.0568, "step": 19019 }, { "epoch": 1.4905956112852665, "grad_norm": 5.680472373962402, "learning_rate": 6.965287184178965e-06, "loss": 1.2562, "step": 19020 }, { "epoch": 1.4906739811912226, "grad_norm": 2.549633026123047, "learning_rate": 6.965003637235958e-06, "loss": 1.7093, "step": 19021 }, { "epoch": 1.4907523510971787, "grad_norm": 3.3125529289245605, "learning_rate": 6.964720082819085e-06, "loss": 1.2097, "step": 19022 }, { "epoch": 1.4908307210031349, "grad_norm": 1.9578834772109985, "learning_rate": 6.964436520929427e-06, "loss": 1.0558, "step": 19023 }, { "epoch": 1.490909090909091, "grad_norm": 2.8364133834838867, "learning_rate": 6.964152951568062e-06, "loss": 1.3587, "step": 19024 }, { "epoch": 1.490987460815047, "grad_norm": 3.8583285808563232, "learning_rate": 6.963869374736068e-06, "loss": 1.414, "step": 19025 }, { "epoch": 1.4910658307210032, "grad_norm": 3.3462395668029785, "learning_rate": 6.963585790434522e-06, "loss": 1.2346, "step": 19026 }, { "epoch": 1.4911442006269593, "grad_norm": 2.318002462387085, "learning_rate": 6.963302198664505e-06, "loss": 1.6411, "step": 19027 }, { "epoch": 1.4912225705329154, "grad_norm": 2.4158830642700195, "learning_rate": 6.963018599427095e-06, "loss": 1.4028, "step": 19028 }, { "epoch": 1.4913009404388715, "grad_norm": 2.9211678504943848, "learning_rate": 6.962734992723372e-06, "loss": 1.5529, "step": 19029 }, { "epoch": 1.4913793103448276, "grad_norm": 4.4508819580078125, "learning_rate": 6.962451378554411e-06, "loss": 1.5103, "step": 19030 }, { "epoch": 1.4914576802507837, "grad_norm": 9.582559585571289, "learning_rate": 6.962167756921294e-06, "loss": 1.5607, "step": 19031 }, { "epoch": 1.4915360501567398, "grad_norm": 7.870630264282227, "learning_rate": 6.961884127825101e-06, "loss": 0.9918, "step": 19032 }, { "epoch": 1.491614420062696, "grad_norm": 2.549454689025879, "learning_rate": 6.961600491266903e-06, "loss": 1.3807, "step": 19033 }, { "epoch": 1.491692789968652, "grad_norm": 4.334681034088135, "learning_rate": 6.961316847247789e-06, "loss": 1.3021, "step": 19034 }, { "epoch": 1.4917711598746082, "grad_norm": 3.8289506435394287, "learning_rate": 6.961033195768829e-06, "loss": 1.7114, "step": 19035 }, { "epoch": 1.4918495297805643, "grad_norm": 2.8210813999176025, "learning_rate": 6.960749536831111e-06, "loss": 1.4264, "step": 19036 }, { "epoch": 1.4919278996865204, "grad_norm": 2.5436203479766846, "learning_rate": 6.960465870435705e-06, "loss": 1.2966, "step": 19037 }, { "epoch": 1.4920062695924765, "grad_norm": 1.8601077795028687, "learning_rate": 6.9601821965836956e-06, "loss": 1.2381, "step": 19038 }, { "epoch": 1.4920846394984326, "grad_norm": 3.049830436706543, "learning_rate": 6.959898515276159e-06, "loss": 1.353, "step": 19039 }, { "epoch": 1.4921630094043887, "grad_norm": 3.6119725704193115, "learning_rate": 6.959614826514175e-06, "loss": 1.4196, "step": 19040 }, { "epoch": 1.4922413793103448, "grad_norm": 2.8854379653930664, "learning_rate": 6.959331130298823e-06, "loss": 1.5053, "step": 19041 }, { "epoch": 1.492319749216301, "grad_norm": 4.721161842346191, "learning_rate": 6.959047426631182e-06, "loss": 1.735, "step": 19042 }, { "epoch": 1.492398119122257, "grad_norm": 3.494220018386841, "learning_rate": 6.958763715512332e-06, "loss": 1.9355, "step": 19043 }, { "epoch": 1.4924764890282132, "grad_norm": 3.0223476886749268, "learning_rate": 6.958479996943349e-06, "loss": 1.3811, "step": 19044 }, { "epoch": 1.4925548589341693, "grad_norm": 2.4430503845214844, "learning_rate": 6.958196270925315e-06, "loss": 1.4269, "step": 19045 }, { "epoch": 1.4926332288401254, "grad_norm": 5.604470252990723, "learning_rate": 6.957912537459308e-06, "loss": 1.8364, "step": 19046 }, { "epoch": 1.4927115987460815, "grad_norm": 2.6729025840759277, "learning_rate": 6.957628796546407e-06, "loss": 1.5739, "step": 19047 }, { "epoch": 1.4927899686520376, "grad_norm": 3.450878143310547, "learning_rate": 6.957345048187692e-06, "loss": 1.3827, "step": 19048 }, { "epoch": 1.4928683385579937, "grad_norm": 3.5052287578582764, "learning_rate": 6.957061292384241e-06, "loss": 1.2529, "step": 19049 }, { "epoch": 1.4929467084639498, "grad_norm": 2.928123712539673, "learning_rate": 6.956777529137134e-06, "loss": 1.2848, "step": 19050 }, { "epoch": 1.493025078369906, "grad_norm": 3.2331650257110596, "learning_rate": 6.956493758447451e-06, "loss": 1.5014, "step": 19051 }, { "epoch": 1.493103448275862, "grad_norm": 4.652995586395264, "learning_rate": 6.956209980316271e-06, "loss": 1.5785, "step": 19052 }, { "epoch": 1.4931818181818182, "grad_norm": 4.80622673034668, "learning_rate": 6.955926194744673e-06, "loss": 1.2113, "step": 19053 }, { "epoch": 1.4932601880877743, "grad_norm": 3.3062806129455566, "learning_rate": 6.9556424017337354e-06, "loss": 1.6559, "step": 19054 }, { "epoch": 1.4933385579937304, "grad_norm": 2.9246506690979004, "learning_rate": 6.955358601284541e-06, "loss": 1.2099, "step": 19055 }, { "epoch": 1.4934169278996865, "grad_norm": 3.061375617980957, "learning_rate": 6.955074793398164e-06, "loss": 1.642, "step": 19056 }, { "epoch": 1.4934952978056426, "grad_norm": 6.524726867675781, "learning_rate": 6.9547909780756875e-06, "loss": 1.3407, "step": 19057 }, { "epoch": 1.4935736677115987, "grad_norm": 2.5275721549987793, "learning_rate": 6.954507155318191e-06, "loss": 1.0999, "step": 19058 }, { "epoch": 1.4936520376175548, "grad_norm": 3.5802061557769775, "learning_rate": 6.954223325126752e-06, "loss": 1.3269, "step": 19059 }, { "epoch": 1.493730407523511, "grad_norm": 4.123996257781982, "learning_rate": 6.9539394875024525e-06, "loss": 1.4751, "step": 19060 }, { "epoch": 1.493808777429467, "grad_norm": 1.8591740131378174, "learning_rate": 6.953655642446368e-06, "loss": 0.6499, "step": 19061 }, { "epoch": 1.4938871473354232, "grad_norm": 8.062348365783691, "learning_rate": 6.953371789959585e-06, "loss": 1.1974, "step": 19062 }, { "epoch": 1.4939655172413793, "grad_norm": 4.433302879333496, "learning_rate": 6.953087930043176e-06, "loss": 1.3659, "step": 19063 }, { "epoch": 1.4940438871473354, "grad_norm": 2.842799663543701, "learning_rate": 6.952804062698225e-06, "loss": 1.0847, "step": 19064 }, { "epoch": 1.4941222570532915, "grad_norm": 3.305523157119751, "learning_rate": 6.952520187925811e-06, "loss": 0.7573, "step": 19065 }, { "epoch": 1.4942006269592476, "grad_norm": 4.4743452072143555, "learning_rate": 6.952236305727012e-06, "loss": 1.1864, "step": 19066 }, { "epoch": 1.4942789968652037, "grad_norm": 3.169236660003662, "learning_rate": 6.9519524161029084e-06, "loss": 1.2442, "step": 19067 }, { "epoch": 1.4943573667711598, "grad_norm": 4.821457386016846, "learning_rate": 6.951668519054581e-06, "loss": 0.7692, "step": 19068 }, { "epoch": 1.494435736677116, "grad_norm": 2.4736928939819336, "learning_rate": 6.951384614583109e-06, "loss": 1.269, "step": 19069 }, { "epoch": 1.494514106583072, "grad_norm": 3.7548887729644775, "learning_rate": 6.951100702689572e-06, "loss": 0.7253, "step": 19070 }, { "epoch": 1.4945924764890282, "grad_norm": 2.689218759536743, "learning_rate": 6.950816783375051e-06, "loss": 1.3615, "step": 19071 }, { "epoch": 1.4946708463949843, "grad_norm": 6.4102935791015625, "learning_rate": 6.950532856640624e-06, "loss": 1.4107, "step": 19072 }, { "epoch": 1.4947492163009404, "grad_norm": 2.448988437652588, "learning_rate": 6.950248922487371e-06, "loss": 1.1865, "step": 19073 }, { "epoch": 1.4948275862068965, "grad_norm": 1.9402073621749878, "learning_rate": 6.949964980916375e-06, "loss": 1.3375, "step": 19074 }, { "epoch": 1.4949059561128526, "grad_norm": 3.4181013107299805, "learning_rate": 6.949681031928712e-06, "loss": 1.4489, "step": 19075 }, { "epoch": 1.4949843260188087, "grad_norm": 10.839095115661621, "learning_rate": 6.949397075525465e-06, "loss": 0.9887, "step": 19076 }, { "epoch": 1.4950626959247648, "grad_norm": 2.1827056407928467, "learning_rate": 6.949113111707711e-06, "loss": 1.4993, "step": 19077 }, { "epoch": 1.495141065830721, "grad_norm": 2.846514940261841, "learning_rate": 6.948829140476534e-06, "loss": 1.3392, "step": 19078 }, { "epoch": 1.495219435736677, "grad_norm": 3.3673439025878906, "learning_rate": 6.9485451618330115e-06, "loss": 1.2925, "step": 19079 }, { "epoch": 1.4952978056426331, "grad_norm": 4.645277976989746, "learning_rate": 6.948261175778223e-06, "loss": 1.3396, "step": 19080 }, { "epoch": 1.4953761755485893, "grad_norm": 2.6548261642456055, "learning_rate": 6.94797718231325e-06, "loss": 1.1822, "step": 19081 }, { "epoch": 1.4954545454545456, "grad_norm": 12.617006301879883, "learning_rate": 6.947693181439172e-06, "loss": 2.1139, "step": 19082 }, { "epoch": 1.4955329153605015, "grad_norm": 4.773101806640625, "learning_rate": 6.947409173157071e-06, "loss": 1.3043, "step": 19083 }, { "epoch": 1.4956112852664578, "grad_norm": 5.147011756896973, "learning_rate": 6.947125157468024e-06, "loss": 1.148, "step": 19084 }, { "epoch": 1.4956896551724137, "grad_norm": 3.120218276977539, "learning_rate": 6.946841134373114e-06, "loss": 1.6232, "step": 19085 }, { "epoch": 1.49576802507837, "grad_norm": 5.059209823608398, "learning_rate": 6.946557103873421e-06, "loss": 1.1102, "step": 19086 }, { "epoch": 1.495846394984326, "grad_norm": 3.28352952003479, "learning_rate": 6.946273065970024e-06, "loss": 1.1525, "step": 19087 }, { "epoch": 1.4959247648902823, "grad_norm": 2.4560348987579346, "learning_rate": 6.945989020664003e-06, "loss": 1.2528, "step": 19088 }, { "epoch": 1.4960031347962381, "grad_norm": 2.1388869285583496, "learning_rate": 6.945704967956439e-06, "loss": 1.6313, "step": 19089 }, { "epoch": 1.4960815047021945, "grad_norm": 2.541170358657837, "learning_rate": 6.945420907848415e-06, "loss": 1.5585, "step": 19090 }, { "epoch": 1.4961598746081504, "grad_norm": 2.6453423500061035, "learning_rate": 6.945136840341008e-06, "loss": 1.4949, "step": 19091 }, { "epoch": 1.4962382445141067, "grad_norm": 2.833972692489624, "learning_rate": 6.944852765435299e-06, "loss": 1.7445, "step": 19092 }, { "epoch": 1.4963166144200626, "grad_norm": 3.2874815464019775, "learning_rate": 6.944568683132369e-06, "loss": 1.404, "step": 19093 }, { "epoch": 1.496394984326019, "grad_norm": 2.4792158603668213, "learning_rate": 6.944284593433299e-06, "loss": 1.4514, "step": 19094 }, { "epoch": 1.4964733542319748, "grad_norm": 3.078054428100586, "learning_rate": 6.94400049633917e-06, "loss": 1.4714, "step": 19095 }, { "epoch": 1.4965517241379311, "grad_norm": 2.9962828159332275, "learning_rate": 6.943716391851059e-06, "loss": 1.1425, "step": 19096 }, { "epoch": 1.496630094043887, "grad_norm": 1.7269552946090698, "learning_rate": 6.94343227997005e-06, "loss": 1.4081, "step": 19097 }, { "epoch": 1.4967084639498434, "grad_norm": 2.9476208686828613, "learning_rate": 6.943148160697225e-06, "loss": 1.5189, "step": 19098 }, { "epoch": 1.4967868338557992, "grad_norm": 2.786344051361084, "learning_rate": 6.942864034033661e-06, "loss": 1.056, "step": 19099 }, { "epoch": 1.4968652037617556, "grad_norm": 3.360323667526245, "learning_rate": 6.942579899980441e-06, "loss": 1.3635, "step": 19100 }, { "epoch": 1.4969435736677115, "grad_norm": 2.03412127494812, "learning_rate": 6.942295758538643e-06, "loss": 1.1586, "step": 19101 }, { "epoch": 1.4970219435736678, "grad_norm": 3.1894617080688477, "learning_rate": 6.94201160970935e-06, "loss": 1.1978, "step": 19102 }, { "epoch": 1.4971003134796237, "grad_norm": 2.9877073764801025, "learning_rate": 6.941727453493643e-06, "loss": 1.4868, "step": 19103 }, { "epoch": 1.49717868338558, "grad_norm": 2.6500298976898193, "learning_rate": 6.9414432898926024e-06, "loss": 1.2398, "step": 19104 }, { "epoch": 1.4972570532915361, "grad_norm": 3.2247262001037598, "learning_rate": 6.941159118907308e-06, "loss": 1.123, "step": 19105 }, { "epoch": 1.4973354231974922, "grad_norm": 4.9861836433410645, "learning_rate": 6.940874940538841e-06, "loss": 1.1512, "step": 19106 }, { "epoch": 1.4974137931034484, "grad_norm": 3.094938039779663, "learning_rate": 6.940590754788285e-06, "loss": 1.4638, "step": 19107 }, { "epoch": 1.4974921630094045, "grad_norm": 2.388259172439575, "learning_rate": 6.9403065616567155e-06, "loss": 1.2525, "step": 19108 }, { "epoch": 1.4975705329153606, "grad_norm": 5.024774551391602, "learning_rate": 6.940022361145218e-06, "loss": 1.47, "step": 19109 }, { "epoch": 1.4976489028213167, "grad_norm": 2.3155267238616943, "learning_rate": 6.93973815325487e-06, "loss": 1.2917, "step": 19110 }, { "epoch": 1.4977272727272728, "grad_norm": 3.263920783996582, "learning_rate": 6.939453937986757e-06, "loss": 1.3375, "step": 19111 }, { "epoch": 1.497805642633229, "grad_norm": 3.0973336696624756, "learning_rate": 6.9391697153419564e-06, "loss": 1.398, "step": 19112 }, { "epoch": 1.497884012539185, "grad_norm": 4.260322570800781, "learning_rate": 6.93888548532155e-06, "loss": 1.0978, "step": 19113 }, { "epoch": 1.4979623824451411, "grad_norm": 6.760804653167725, "learning_rate": 6.938601247926621e-06, "loss": 1.9322, "step": 19114 }, { "epoch": 1.4980407523510972, "grad_norm": 2.0128297805786133, "learning_rate": 6.9383170031582445e-06, "loss": 0.9604, "step": 19115 }, { "epoch": 1.4981191222570533, "grad_norm": 3.3530988693237305, "learning_rate": 6.938032751017509e-06, "loss": 1.7106, "step": 19116 }, { "epoch": 1.4981974921630095, "grad_norm": 4.075582504272461, "learning_rate": 6.93774849150549e-06, "loss": 1.5812, "step": 19117 }, { "epoch": 1.4982758620689656, "grad_norm": 2.7619261741638184, "learning_rate": 6.937464224623274e-06, "loss": 1.2632, "step": 19118 }, { "epoch": 1.4983542319749217, "grad_norm": 2.4210925102233887, "learning_rate": 6.937179950371937e-06, "loss": 1.2216, "step": 19119 }, { "epoch": 1.4984326018808778, "grad_norm": 3.220318078994751, "learning_rate": 6.936895668752564e-06, "loss": 2.0391, "step": 19120 }, { "epoch": 1.498510971786834, "grad_norm": 3.2184107303619385, "learning_rate": 6.936611379766234e-06, "loss": 1.168, "step": 19121 }, { "epoch": 1.49858934169279, "grad_norm": 2.635202646255493, "learning_rate": 6.936327083414028e-06, "loss": 1.1247, "step": 19122 }, { "epoch": 1.4986677115987461, "grad_norm": 2.624380588531494, "learning_rate": 6.93604277969703e-06, "loss": 1.3359, "step": 19123 }, { "epoch": 1.4987460815047022, "grad_norm": 2.351191282272339, "learning_rate": 6.935758468616318e-06, "loss": 1.0114, "step": 19124 }, { "epoch": 1.4988244514106583, "grad_norm": 3.9950056076049805, "learning_rate": 6.935474150172976e-06, "loss": 1.5712, "step": 19125 }, { "epoch": 1.4989028213166145, "grad_norm": 1.8581829071044922, "learning_rate": 6.9351898243680846e-06, "loss": 1.4776, "step": 19126 }, { "epoch": 1.4989811912225706, "grad_norm": 2.3562936782836914, "learning_rate": 6.9349054912027255e-06, "loss": 1.3185, "step": 19127 }, { "epoch": 1.4990595611285267, "grad_norm": 4.385297775268555, "learning_rate": 6.934621150677978e-06, "loss": 1.2788, "step": 19128 }, { "epoch": 1.4991379310344828, "grad_norm": 3.1229605674743652, "learning_rate": 6.934336802794928e-06, "loss": 1.2767, "step": 19129 }, { "epoch": 1.499216300940439, "grad_norm": 4.265829086303711, "learning_rate": 6.934052447554653e-06, "loss": 1.9146, "step": 19130 }, { "epoch": 1.499294670846395, "grad_norm": 1.969163417816162, "learning_rate": 6.933768084958234e-06, "loss": 1.2876, "step": 19131 }, { "epoch": 1.4993730407523511, "grad_norm": 2.75614070892334, "learning_rate": 6.933483715006757e-06, "loss": 1.4174, "step": 19132 }, { "epoch": 1.4994514106583072, "grad_norm": 3.1053273677825928, "learning_rate": 6.9331993377013005e-06, "loss": 1.2885, "step": 19133 }, { "epoch": 1.4995297805642633, "grad_norm": 3.167226791381836, "learning_rate": 6.932914953042945e-06, "loss": 1.5371, "step": 19134 }, { "epoch": 1.4996081504702194, "grad_norm": 2.2835638523101807, "learning_rate": 6.932630561032776e-06, "loss": 1.305, "step": 19135 }, { "epoch": 1.4996865203761756, "grad_norm": 2.5391054153442383, "learning_rate": 6.932346161671871e-06, "loss": 1.582, "step": 19136 }, { "epoch": 1.4997648902821317, "grad_norm": 2.702436685562134, "learning_rate": 6.932061754961316e-06, "loss": 1.4412, "step": 19137 }, { "epoch": 1.4998432601880878, "grad_norm": 2.4900870323181152, "learning_rate": 6.931777340902189e-06, "loss": 1.3715, "step": 19138 }, { "epoch": 1.499921630094044, "grad_norm": 2.639378547668457, "learning_rate": 6.9314929194955734e-06, "loss": 1.1829, "step": 19139 }, { "epoch": 1.5, "grad_norm": 2.356058120727539, "learning_rate": 6.931208490742552e-06, "loss": 1.1827, "step": 19140 }, { "epoch": 1.5, "eval_loss": 1.2826650142669678, "eval_runtime": 108.0081, "eval_samples_per_second": 24.878, "eval_steps_per_second": 6.222, "step": 19140 }, { "epoch": 1.500078369905956, "grad_norm": 3.52469539642334, "learning_rate": 6.9309240546442035e-06, "loss": 1.5827, "step": 19141 }, { "epoch": 1.5001567398119122, "grad_norm": 1.9197760820388794, "learning_rate": 6.930639611201614e-06, "loss": 1.294, "step": 19142 }, { "epoch": 1.5002351097178683, "grad_norm": 2.6620068550109863, "learning_rate": 6.930355160415861e-06, "loss": 1.4025, "step": 19143 }, { "epoch": 1.5003134796238244, "grad_norm": 6.710373401641846, "learning_rate": 6.930070702288029e-06, "loss": 1.3736, "step": 19144 }, { "epoch": 1.5003918495297806, "grad_norm": 2.9960222244262695, "learning_rate": 6.929786236819199e-06, "loss": 1.3057, "step": 19145 }, { "epoch": 1.5004702194357367, "grad_norm": 2.842865228652954, "learning_rate": 6.9295017640104544e-06, "loss": 1.1401, "step": 19146 }, { "epoch": 1.5005485893416928, "grad_norm": 2.14298415184021, "learning_rate": 6.929217283862875e-06, "loss": 1.4757, "step": 19147 }, { "epoch": 1.5006269592476489, "grad_norm": 2.6554453372955322, "learning_rate": 6.928932796377544e-06, "loss": 1.6044, "step": 19148 }, { "epoch": 1.500705329153605, "grad_norm": 3.228414535522461, "learning_rate": 6.928648301555545e-06, "loss": 1.8241, "step": 19149 }, { "epoch": 1.500783699059561, "grad_norm": 2.9037013053894043, "learning_rate": 6.9283637993979565e-06, "loss": 1.2102, "step": 19150 }, { "epoch": 1.5008620689655172, "grad_norm": 5.134494304656982, "learning_rate": 6.928079289905863e-06, "loss": 1.4114, "step": 19151 }, { "epoch": 1.5009404388714733, "grad_norm": 2.357055187225342, "learning_rate": 6.927794773080347e-06, "loss": 1.4674, "step": 19152 }, { "epoch": 1.5010188087774294, "grad_norm": 1.8932057619094849, "learning_rate": 6.92751024892249e-06, "loss": 1.0852, "step": 19153 }, { "epoch": 1.5010971786833855, "grad_norm": 2.5918424129486084, "learning_rate": 6.927225717433375e-06, "loss": 1.1841, "step": 19154 }, { "epoch": 1.5011755485893417, "grad_norm": 4.436528205871582, "learning_rate": 6.926941178614082e-06, "loss": 1.3426, "step": 19155 }, { "epoch": 1.5012539184952978, "grad_norm": 3.1396117210388184, "learning_rate": 6.926656632465694e-06, "loss": 1.4107, "step": 19156 }, { "epoch": 1.5013322884012539, "grad_norm": 3.634206771850586, "learning_rate": 6.926372078989294e-06, "loss": 1.4496, "step": 19157 }, { "epoch": 1.50141065830721, "grad_norm": 3.8293027877807617, "learning_rate": 6.926087518185965e-06, "loss": 1.2671, "step": 19158 }, { "epoch": 1.501489028213166, "grad_norm": 2.1095550060272217, "learning_rate": 6.925802950056786e-06, "loss": 1.3179, "step": 19159 }, { "epoch": 1.5015673981191222, "grad_norm": 3.3359642028808594, "learning_rate": 6.9255183746028435e-06, "loss": 1.0878, "step": 19160 }, { "epoch": 1.5016457680250783, "grad_norm": 2.681993007659912, "learning_rate": 6.925233791825219e-06, "loss": 1.3019, "step": 19161 }, { "epoch": 1.5017241379310344, "grad_norm": 2.6129581928253174, "learning_rate": 6.924949201724992e-06, "loss": 1.2627, "step": 19162 }, { "epoch": 1.5018025078369905, "grad_norm": 3.673616409301758, "learning_rate": 6.924664604303249e-06, "loss": 1.7059, "step": 19163 }, { "epoch": 1.5018808777429467, "grad_norm": 3.2029318809509277, "learning_rate": 6.924379999561067e-06, "loss": 1.3193, "step": 19164 }, { "epoch": 1.501959247648903, "grad_norm": 18.419416427612305, "learning_rate": 6.924095387499535e-06, "loss": 1.7511, "step": 19165 }, { "epoch": 1.5020376175548589, "grad_norm": 2.6577394008636475, "learning_rate": 6.923810768119732e-06, "loss": 1.3557, "step": 19166 }, { "epoch": 1.5021159874608152, "grad_norm": 4.967996597290039, "learning_rate": 6.92352614142274e-06, "loss": 1.3637, "step": 19167 }, { "epoch": 1.502194357366771, "grad_norm": 2.2820301055908203, "learning_rate": 6.9232415074096446e-06, "loss": 1.1877, "step": 19168 }, { "epoch": 1.5022727272727274, "grad_norm": 1.8701528310775757, "learning_rate": 6.922956866081524e-06, "loss": 1.2328, "step": 19169 }, { "epoch": 1.5023510971786833, "grad_norm": 2.7231507301330566, "learning_rate": 6.922672217439465e-06, "loss": 1.4653, "step": 19170 }, { "epoch": 1.5024294670846396, "grad_norm": 2.6218743324279785, "learning_rate": 6.922387561484547e-06, "loss": 1.3673, "step": 19171 }, { "epoch": 1.5025078369905955, "grad_norm": 5.745115280151367, "learning_rate": 6.9221028982178555e-06, "loss": 1.6706, "step": 19172 }, { "epoch": 1.5025862068965519, "grad_norm": 2.4252071380615234, "learning_rate": 6.92181822764047e-06, "loss": 1.4929, "step": 19173 }, { "epoch": 1.5026645768025078, "grad_norm": 2.3292455673217773, "learning_rate": 6.921533549753477e-06, "loss": 1.0333, "step": 19174 }, { "epoch": 1.502742946708464, "grad_norm": 3.241281032562256, "learning_rate": 6.921248864557956e-06, "loss": 1.3898, "step": 19175 }, { "epoch": 1.50282131661442, "grad_norm": 3.5204970836639404, "learning_rate": 6.9209641720549914e-06, "loss": 1.2195, "step": 19176 }, { "epoch": 1.5028996865203763, "grad_norm": 2.796245574951172, "learning_rate": 6.920679472245665e-06, "loss": 1.4295, "step": 19177 }, { "epoch": 1.5029780564263322, "grad_norm": 2.5707340240478516, "learning_rate": 6.920394765131061e-06, "loss": 1.0555, "step": 19178 }, { "epoch": 1.5030564263322885, "grad_norm": 3.714994192123413, "learning_rate": 6.920110050712262e-06, "loss": 1.7835, "step": 19179 }, { "epoch": 1.5031347962382444, "grad_norm": 3.7874703407287598, "learning_rate": 6.9198253289903515e-06, "loss": 1.6107, "step": 19180 }, { "epoch": 1.5032131661442008, "grad_norm": 6.1017303466796875, "learning_rate": 6.91954059996641e-06, "loss": 1.243, "step": 19181 }, { "epoch": 1.5032915360501566, "grad_norm": 2.6677820682525635, "learning_rate": 6.919255863641523e-06, "loss": 1.2285, "step": 19182 }, { "epoch": 1.503369905956113, "grad_norm": 2.7069547176361084, "learning_rate": 6.918971120016772e-06, "loss": 1.5509, "step": 19183 }, { "epoch": 1.5034482758620689, "grad_norm": 2.604978322982788, "learning_rate": 6.91868636909324e-06, "loss": 1.5732, "step": 19184 }, { "epoch": 1.5035266457680252, "grad_norm": 3.9231321811676025, "learning_rate": 6.918401610872012e-06, "loss": 1.7804, "step": 19185 }, { "epoch": 1.503605015673981, "grad_norm": 4.492701530456543, "learning_rate": 6.918116845354168e-06, "loss": 1.4942, "step": 19186 }, { "epoch": 1.5036833855799374, "grad_norm": 2.798867702484131, "learning_rate": 6.9178320725407945e-06, "loss": 1.0605, "step": 19187 }, { "epoch": 1.5037617554858933, "grad_norm": 2.785858631134033, "learning_rate": 6.917547292432971e-06, "loss": 1.3193, "step": 19188 }, { "epoch": 1.5038401253918496, "grad_norm": 2.968139886856079, "learning_rate": 6.9172625050317846e-06, "loss": 1.038, "step": 19189 }, { "epoch": 1.5039184952978055, "grad_norm": 2.608151435852051, "learning_rate": 6.916977710338314e-06, "loss": 1.3192, "step": 19190 }, { "epoch": 1.5039968652037619, "grad_norm": 4.264720439910889, "learning_rate": 6.916692908353646e-06, "loss": 1.5364, "step": 19191 }, { "epoch": 1.5040752351097177, "grad_norm": 3.3548216819763184, "learning_rate": 6.916408099078863e-06, "loss": 1.6413, "step": 19192 }, { "epoch": 1.504153605015674, "grad_norm": 3.386774778366089, "learning_rate": 6.916123282515048e-06, "loss": 1.3375, "step": 19193 }, { "epoch": 1.50423197492163, "grad_norm": 2.0628561973571777, "learning_rate": 6.915838458663283e-06, "loss": 1.4836, "step": 19194 }, { "epoch": 1.5043103448275863, "grad_norm": 2.9869589805603027, "learning_rate": 6.915553627524654e-06, "loss": 1.321, "step": 19195 }, { "epoch": 1.5043887147335422, "grad_norm": 3.291513204574585, "learning_rate": 6.915268789100242e-06, "loss": 1.2848, "step": 19196 }, { "epoch": 1.5044670846394985, "grad_norm": 3.123387575149536, "learning_rate": 6.914983943391132e-06, "loss": 1.3114, "step": 19197 }, { "epoch": 1.5045454545454544, "grad_norm": 5.756241321563721, "learning_rate": 6.914699090398406e-06, "loss": 1.5574, "step": 19198 }, { "epoch": 1.5046238244514107, "grad_norm": 1.8658716678619385, "learning_rate": 6.914414230123148e-06, "loss": 1.3743, "step": 19199 }, { "epoch": 1.5047021943573666, "grad_norm": 8.637383460998535, "learning_rate": 6.9141293625664414e-06, "loss": 1.7215, "step": 19200 }, { "epoch": 1.504780564263323, "grad_norm": 2.5914671421051025, "learning_rate": 6.913844487729371e-06, "loss": 1.4738, "step": 19201 }, { "epoch": 1.5048589341692789, "grad_norm": 4.435746192932129, "learning_rate": 6.913559605613018e-06, "loss": 1.7554, "step": 19202 }, { "epoch": 1.5049373040752352, "grad_norm": 2.363626003265381, "learning_rate": 6.913274716218467e-06, "loss": 1.1848, "step": 19203 }, { "epoch": 1.505015673981191, "grad_norm": 2.4083569049835205, "learning_rate": 6.912989819546802e-06, "loss": 1.4605, "step": 19204 }, { "epoch": 1.5050940438871474, "grad_norm": 3.3438286781311035, "learning_rate": 6.9127049155991065e-06, "loss": 1.3127, "step": 19205 }, { "epoch": 1.5051724137931033, "grad_norm": 2.287961006164551, "learning_rate": 6.912420004376464e-06, "loss": 1.3977, "step": 19206 }, { "epoch": 1.5052507836990596, "grad_norm": 2.621729612350464, "learning_rate": 6.912135085879958e-06, "loss": 1.4022, "step": 19207 }, { "epoch": 1.5053291536050155, "grad_norm": 3.338238477706909, "learning_rate": 6.911850160110674e-06, "loss": 1.7544, "step": 19208 }, { "epoch": 1.5054075235109718, "grad_norm": 2.379539966583252, "learning_rate": 6.911565227069691e-06, "loss": 1.6145, "step": 19209 }, { "epoch": 1.5054858934169277, "grad_norm": 3.0088961124420166, "learning_rate": 6.911280286758097e-06, "loss": 1.42, "step": 19210 }, { "epoch": 1.505564263322884, "grad_norm": 2.5136547088623047, "learning_rate": 6.910995339176975e-06, "loss": 1.3481, "step": 19211 }, { "epoch": 1.5056426332288402, "grad_norm": 3.1689953804016113, "learning_rate": 6.910710384327408e-06, "loss": 1.4096, "step": 19212 }, { "epoch": 1.5057210031347963, "grad_norm": 3.050278902053833, "learning_rate": 6.91042542221048e-06, "loss": 1.152, "step": 19213 }, { "epoch": 1.5057993730407524, "grad_norm": 2.73376727104187, "learning_rate": 6.910140452827274e-06, "loss": 1.3245, "step": 19214 }, { "epoch": 1.5058777429467085, "grad_norm": 3.3660733699798584, "learning_rate": 6.909855476178877e-06, "loss": 1.7224, "step": 19215 }, { "epoch": 1.5059561128526646, "grad_norm": 4.371905326843262, "learning_rate": 6.909570492266369e-06, "loss": 1.4267, "step": 19216 }, { "epoch": 1.5060344827586207, "grad_norm": 2.1398088932037354, "learning_rate": 6.9092855010908365e-06, "loss": 0.9585, "step": 19217 }, { "epoch": 1.5061128526645768, "grad_norm": 2.3646700382232666, "learning_rate": 6.909000502653361e-06, "loss": 1.4611, "step": 19218 }, { "epoch": 1.506191222570533, "grad_norm": 3.2532920837402344, "learning_rate": 6.90871549695503e-06, "loss": 1.5545, "step": 19219 }, { "epoch": 1.506269592476489, "grad_norm": 16.307708740234375, "learning_rate": 6.908430483996925e-06, "loss": 0.5606, "step": 19220 }, { "epoch": 1.5063479623824452, "grad_norm": 2.843168020248413, "learning_rate": 6.908145463780131e-06, "loss": 1.055, "step": 19221 }, { "epoch": 1.5064263322884013, "grad_norm": 4.673534870147705, "learning_rate": 6.907860436305731e-06, "loss": 1.0856, "step": 19222 }, { "epoch": 1.5065047021943574, "grad_norm": 4.1111249923706055, "learning_rate": 6.907575401574811e-06, "loss": 1.143, "step": 19223 }, { "epoch": 1.5065830721003135, "grad_norm": 10.651597023010254, "learning_rate": 6.9072903595884535e-06, "loss": 1.8751, "step": 19224 }, { "epoch": 1.5066614420062696, "grad_norm": 4.47423791885376, "learning_rate": 6.907005310347742e-06, "loss": 1.3099, "step": 19225 }, { "epoch": 1.5067398119122257, "grad_norm": 3.948681116104126, "learning_rate": 6.9067202538537634e-06, "loss": 1.4513, "step": 19226 }, { "epoch": 1.5068181818181818, "grad_norm": 2.9368011951446533, "learning_rate": 6.906435190107599e-06, "loss": 1.3603, "step": 19227 }, { "epoch": 1.506896551724138, "grad_norm": 2.4714839458465576, "learning_rate": 6.906150119110335e-06, "loss": 1.7411, "step": 19228 }, { "epoch": 1.506974921630094, "grad_norm": 2.126983404159546, "learning_rate": 6.905865040863054e-06, "loss": 1.1186, "step": 19229 }, { "epoch": 1.5070532915360502, "grad_norm": 8.073463439941406, "learning_rate": 6.905579955366844e-06, "loss": 1.3143, "step": 19230 }, { "epoch": 1.5071316614420063, "grad_norm": 2.6718385219573975, "learning_rate": 6.9052948626227835e-06, "loss": 1.2789, "step": 19231 }, { "epoch": 1.5072100313479624, "grad_norm": 4.579092025756836, "learning_rate": 6.9050097626319615e-06, "loss": 0.9293, "step": 19232 }, { "epoch": 1.5072884012539185, "grad_norm": 3.162386417388916, "learning_rate": 6.9047246553954605e-06, "loss": 1.3881, "step": 19233 }, { "epoch": 1.5073667711598746, "grad_norm": 2.0810279846191406, "learning_rate": 6.904439540914366e-06, "loss": 1.4626, "step": 19234 }, { "epoch": 1.5074451410658307, "grad_norm": 2.666846752166748, "learning_rate": 6.904154419189759e-06, "loss": 1.1421, "step": 19235 }, { "epoch": 1.5075235109717868, "grad_norm": 6.61128044128418, "learning_rate": 6.90386929022273e-06, "loss": 1.3192, "step": 19236 }, { "epoch": 1.507601880877743, "grad_norm": 3.1946895122528076, "learning_rate": 6.903584154014357e-06, "loss": 0.9819, "step": 19237 }, { "epoch": 1.507680250783699, "grad_norm": 6.546694278717041, "learning_rate": 6.903299010565732e-06, "loss": 1.2965, "step": 19238 }, { "epoch": 1.5077586206896552, "grad_norm": 4.580959796905518, "learning_rate": 6.903013859877932e-06, "loss": 1.1917, "step": 19239 }, { "epoch": 1.5078369905956113, "grad_norm": 3.3686134815216064, "learning_rate": 6.902728701952045e-06, "loss": 1.7456, "step": 19240 }, { "epoch": 1.5079153605015674, "grad_norm": 4.720685958862305, "learning_rate": 6.902443536789157e-06, "loss": 1.4753, "step": 19241 }, { "epoch": 1.5079937304075235, "grad_norm": 7.826314926147461, "learning_rate": 6.902158364390347e-06, "loss": 1.5651, "step": 19242 }, { "epoch": 1.5080721003134796, "grad_norm": 4.842555999755859, "learning_rate": 6.901873184756708e-06, "loss": 1.7048, "step": 19243 }, { "epoch": 1.5081504702194357, "grad_norm": 4.772430896759033, "learning_rate": 6.9015879978893165e-06, "loss": 1.3321, "step": 19244 }, { "epoch": 1.5082288401253918, "grad_norm": 2.3606374263763428, "learning_rate": 6.901302803789264e-06, "loss": 1.0436, "step": 19245 }, { "epoch": 1.508307210031348, "grad_norm": 2.373560667037964, "learning_rate": 6.901017602457631e-06, "loss": 1.3413, "step": 19246 }, { "epoch": 1.508385579937304, "grad_norm": 2.354595422744751, "learning_rate": 6.900732393895504e-06, "loss": 1.2121, "step": 19247 }, { "epoch": 1.5084639498432602, "grad_norm": 3.3383805751800537, "learning_rate": 6.900447178103966e-06, "loss": 0.3173, "step": 19248 }, { "epoch": 1.5085423197492163, "grad_norm": 2.125138521194458, "learning_rate": 6.900161955084103e-06, "loss": 1.0868, "step": 19249 }, { "epoch": 1.5086206896551724, "grad_norm": 3.853712320327759, "learning_rate": 6.899876724837001e-06, "loss": 1.5052, "step": 19250 }, { "epoch": 1.5086990595611285, "grad_norm": 3.414640188217163, "learning_rate": 6.899591487363744e-06, "loss": 1.4214, "step": 19251 }, { "epoch": 1.5087774294670846, "grad_norm": 2.8394153118133545, "learning_rate": 6.899306242665417e-06, "loss": 1.3432, "step": 19252 }, { "epoch": 1.5088557993730407, "grad_norm": 1.8507513999938965, "learning_rate": 6.899020990743104e-06, "loss": 1.4499, "step": 19253 }, { "epoch": 1.5089341692789968, "grad_norm": 2.0271565914154053, "learning_rate": 6.89873573159789e-06, "loss": 1.2005, "step": 19254 }, { "epoch": 1.509012539184953, "grad_norm": 2.394771099090576, "learning_rate": 6.89845046523086e-06, "loss": 1.3121, "step": 19255 }, { "epoch": 1.509090909090909, "grad_norm": 2.674110174179077, "learning_rate": 6.8981651916431e-06, "loss": 1.4107, "step": 19256 }, { "epoch": 1.5091692789968651, "grad_norm": 5.7563323974609375, "learning_rate": 6.897879910835696e-06, "loss": 1.4029, "step": 19257 }, { "epoch": 1.5092476489028215, "grad_norm": 3.6552276611328125, "learning_rate": 6.8975946228097305e-06, "loss": 1.0635, "step": 19258 }, { "epoch": 1.5093260188087774, "grad_norm": 2.600823402404785, "learning_rate": 6.8973093275662895e-06, "loss": 1.3009, "step": 19259 }, { "epoch": 1.5094043887147337, "grad_norm": 4.571048736572266, "learning_rate": 6.897024025106458e-06, "loss": 1.3587, "step": 19260 }, { "epoch": 1.5094827586206896, "grad_norm": 5.8471221923828125, "learning_rate": 6.896738715431322e-06, "loss": 1.0595, "step": 19261 }, { "epoch": 1.509561128526646, "grad_norm": 2.8597335815429688, "learning_rate": 6.896453398541968e-06, "loss": 1.1369, "step": 19262 }, { "epoch": 1.5096394984326018, "grad_norm": 3.594160795211792, "learning_rate": 6.8961680744394764e-06, "loss": 1.3334, "step": 19263 }, { "epoch": 1.5097178683385581, "grad_norm": 3.169502019882202, "learning_rate": 6.8958827431249375e-06, "loss": 1.9359, "step": 19264 }, { "epoch": 1.509796238244514, "grad_norm": 2.5486981868743896, "learning_rate": 6.8955974045994325e-06, "loss": 1.0959, "step": 19265 }, { "epoch": 1.5098746081504704, "grad_norm": 2.5718483924865723, "learning_rate": 6.89531205886405e-06, "loss": 1.447, "step": 19266 }, { "epoch": 1.5099529780564263, "grad_norm": 3.2399191856384277, "learning_rate": 6.895026705919874e-06, "loss": 1.2638, "step": 19267 }, { "epoch": 1.5100313479623826, "grad_norm": 5.843527317047119, "learning_rate": 6.894741345767989e-06, "loss": 1.1106, "step": 19268 }, { "epoch": 1.5101097178683385, "grad_norm": 5.212405204772949, "learning_rate": 6.894455978409482e-06, "loss": 1.1805, "step": 19269 }, { "epoch": 1.5101880877742948, "grad_norm": 3.168083667755127, "learning_rate": 6.894170603845436e-06, "loss": 1.0056, "step": 19270 }, { "epoch": 1.5102664576802507, "grad_norm": 2.3228847980499268, "learning_rate": 6.89388522207694e-06, "loss": 1.2399, "step": 19271 }, { "epoch": 1.510344827586207, "grad_norm": 2.972646951675415, "learning_rate": 6.893599833105075e-06, "loss": 1.4734, "step": 19272 }, { "epoch": 1.510423197492163, "grad_norm": 2.49446964263916, "learning_rate": 6.8933144369309315e-06, "loss": 1.3687, "step": 19273 }, { "epoch": 1.5105015673981192, "grad_norm": 4.577274799346924, "learning_rate": 6.893029033555591e-06, "loss": 0.9497, "step": 19274 }, { "epoch": 1.5105799373040751, "grad_norm": 2.182650566101074, "learning_rate": 6.89274362298014e-06, "loss": 1.2463, "step": 19275 }, { "epoch": 1.5106583072100315, "grad_norm": 3.982165575027466, "learning_rate": 6.892458205205664e-06, "loss": 0.9856, "step": 19276 }, { "epoch": 1.5107366771159874, "grad_norm": 2.889289617538452, "learning_rate": 6.892172780233251e-06, "loss": 1.4127, "step": 19277 }, { "epoch": 1.5108150470219437, "grad_norm": 2.6669762134552, "learning_rate": 6.891887348063983e-06, "loss": 1.6794, "step": 19278 }, { "epoch": 1.5108934169278996, "grad_norm": 2.0731780529022217, "learning_rate": 6.891601908698947e-06, "loss": 1.3606, "step": 19279 }, { "epoch": 1.510971786833856, "grad_norm": 3.200760841369629, "learning_rate": 6.89131646213923e-06, "loss": 1.1749, "step": 19280 }, { "epoch": 1.5110501567398118, "grad_norm": 2.394211530685425, "learning_rate": 6.891031008385916e-06, "loss": 1.1915, "step": 19281 }, { "epoch": 1.5111285266457681, "grad_norm": 2.3687148094177246, "learning_rate": 6.890745547440092e-06, "loss": 1.3233, "step": 19282 }, { "epoch": 1.511206896551724, "grad_norm": 2.3870561122894287, "learning_rate": 6.890460079302843e-06, "loss": 0.8897, "step": 19283 }, { "epoch": 1.5112852664576804, "grad_norm": 2.562406301498413, "learning_rate": 6.890174603975255e-06, "loss": 1.2554, "step": 19284 }, { "epoch": 1.5113636363636362, "grad_norm": 3.337315320968628, "learning_rate": 6.889889121458413e-06, "loss": 1.2798, "step": 19285 }, { "epoch": 1.5114420062695926, "grad_norm": 2.8504817485809326, "learning_rate": 6.889603631753403e-06, "loss": 1.2074, "step": 19286 }, { "epoch": 1.5115203761755485, "grad_norm": 3.412578582763672, "learning_rate": 6.8893181348613116e-06, "loss": 0.6332, "step": 19287 }, { "epoch": 1.5115987460815048, "grad_norm": 4.02721643447876, "learning_rate": 6.8890326307832256e-06, "loss": 1.5243, "step": 19288 }, { "epoch": 1.5116771159874607, "grad_norm": 2.98006534576416, "learning_rate": 6.888747119520228e-06, "loss": 0.8258, "step": 19289 }, { "epoch": 1.511755485893417, "grad_norm": 2.2968783378601074, "learning_rate": 6.888461601073408e-06, "loss": 1.3057, "step": 19290 }, { "epoch": 1.511833855799373, "grad_norm": 3.529954195022583, "learning_rate": 6.888176075443849e-06, "loss": 1.3119, "step": 19291 }, { "epoch": 1.5119122257053292, "grad_norm": 9.614359855651855, "learning_rate": 6.88789054263264e-06, "loss": 1.515, "step": 19292 }, { "epoch": 1.5119905956112851, "grad_norm": 4.171742916107178, "learning_rate": 6.8876050026408635e-06, "loss": 1.5403, "step": 19293 }, { "epoch": 1.5120689655172415, "grad_norm": 2.8596184253692627, "learning_rate": 6.887319455469607e-06, "loss": 1.7205, "step": 19294 }, { "epoch": 1.5121473354231973, "grad_norm": 4.645768165588379, "learning_rate": 6.887033901119957e-06, "loss": 1.2785, "step": 19295 }, { "epoch": 1.5122257053291537, "grad_norm": 2.4182016849517822, "learning_rate": 6.886748339592999e-06, "loss": 1.0027, "step": 19296 }, { "epoch": 1.5123040752351096, "grad_norm": 2.506876230239868, "learning_rate": 6.88646277088982e-06, "loss": 1.4757, "step": 19297 }, { "epoch": 1.512382445141066, "grad_norm": 3.1785552501678467, "learning_rate": 6.886177195011504e-06, "loss": 1.4457, "step": 19298 }, { "epoch": 1.5124608150470218, "grad_norm": 2.110520362854004, "learning_rate": 6.885891611959141e-06, "loss": 1.4094, "step": 19299 }, { "epoch": 1.5125391849529781, "grad_norm": 4.033464431762695, "learning_rate": 6.885606021733814e-06, "loss": 1.3395, "step": 19300 }, { "epoch": 1.512617554858934, "grad_norm": 5.974521160125732, "learning_rate": 6.88532042433661e-06, "loss": 2.0217, "step": 19301 }, { "epoch": 1.5126959247648903, "grad_norm": 3.4325010776519775, "learning_rate": 6.885034819768616e-06, "loss": 1.1003, "step": 19302 }, { "epoch": 1.5127742946708462, "grad_norm": 2.417926073074341, "learning_rate": 6.884749208030917e-06, "loss": 1.369, "step": 19303 }, { "epoch": 1.5128526645768026, "grad_norm": 2.36903715133667, "learning_rate": 6.884463589124599e-06, "loss": 1.1755, "step": 19304 }, { "epoch": 1.5129310344827587, "grad_norm": 2.301910877227783, "learning_rate": 6.884177963050751e-06, "loss": 1.2709, "step": 19305 }, { "epoch": 1.5130094043887148, "grad_norm": 6.018330097198486, "learning_rate": 6.883892329810458e-06, "loss": 1.2035, "step": 19306 }, { "epoch": 1.513087774294671, "grad_norm": 3.5474963188171387, "learning_rate": 6.883606689404806e-06, "loss": 1.1436, "step": 19307 }, { "epoch": 1.513166144200627, "grad_norm": 2.375356674194336, "learning_rate": 6.8833210418348825e-06, "loss": 1.6088, "step": 19308 }, { "epoch": 1.5132445141065831, "grad_norm": 3.0110669136047363, "learning_rate": 6.883035387101772e-06, "loss": 1.2665, "step": 19309 }, { "epoch": 1.5133228840125392, "grad_norm": 2.8845713138580322, "learning_rate": 6.882749725206561e-06, "loss": 1.561, "step": 19310 }, { "epoch": 1.5134012539184953, "grad_norm": 3.1268222332000732, "learning_rate": 6.882464056150339e-06, "loss": 1.3473, "step": 19311 }, { "epoch": 1.5134796238244514, "grad_norm": 2.7903037071228027, "learning_rate": 6.8821783799341905e-06, "loss": 1.1819, "step": 19312 }, { "epoch": 1.5135579937304076, "grad_norm": 3.358964443206787, "learning_rate": 6.8818926965592e-06, "loss": 1.5688, "step": 19313 }, { "epoch": 1.5136363636363637, "grad_norm": 2.486565113067627, "learning_rate": 6.8816070060264585e-06, "loss": 1.379, "step": 19314 }, { "epoch": 1.5137147335423198, "grad_norm": 2.970235586166382, "learning_rate": 6.881321308337049e-06, "loss": 1.2731, "step": 19315 }, { "epoch": 1.513793103448276, "grad_norm": 4.013711452484131, "learning_rate": 6.881035603492061e-06, "loss": 1.9492, "step": 19316 }, { "epoch": 1.513871473354232, "grad_norm": 2.1327285766601562, "learning_rate": 6.880749891492578e-06, "loss": 1.4598, "step": 19317 }, { "epoch": 1.513949843260188, "grad_norm": 4.282514572143555, "learning_rate": 6.88046417233969e-06, "loss": 1.2638, "step": 19318 }, { "epoch": 1.5140282131661442, "grad_norm": 3.020573616027832, "learning_rate": 6.880178446034481e-06, "loss": 1.5684, "step": 19319 }, { "epoch": 1.5141065830721003, "grad_norm": 4.205358028411865, "learning_rate": 6.8798927125780405e-06, "loss": 1.7004, "step": 19320 }, { "epoch": 1.5141849529780564, "grad_norm": 3.3420841693878174, "learning_rate": 6.879606971971453e-06, "loss": 1.6023, "step": 19321 }, { "epoch": 1.5142633228840126, "grad_norm": 3.972247362136841, "learning_rate": 6.879321224215805e-06, "loss": 1.1325, "step": 19322 }, { "epoch": 1.5143416927899687, "grad_norm": 3.4996259212493896, "learning_rate": 6.879035469312186e-06, "loss": 1.3234, "step": 19323 }, { "epoch": 1.5144200626959248, "grad_norm": 2.4819416999816895, "learning_rate": 6.87874970726168e-06, "loss": 1.3126, "step": 19324 }, { "epoch": 1.5144984326018809, "grad_norm": 2.0135762691497803, "learning_rate": 6.878463938065376e-06, "loss": 1.5246, "step": 19325 }, { "epoch": 1.514576802507837, "grad_norm": 3.1487271785736084, "learning_rate": 6.878178161724358e-06, "loss": 1.6858, "step": 19326 }, { "epoch": 1.514655172413793, "grad_norm": 2.2437045574188232, "learning_rate": 6.877892378239717e-06, "loss": 1.6195, "step": 19327 }, { "epoch": 1.5147335423197492, "grad_norm": 2.321467638015747, "learning_rate": 6.877606587612537e-06, "loss": 1.1917, "step": 19328 }, { "epoch": 1.5148119122257053, "grad_norm": 2.5811350345611572, "learning_rate": 6.877320789843907e-06, "loss": 1.4408, "step": 19329 }, { "epoch": 1.5148902821316614, "grad_norm": 5.412712097167969, "learning_rate": 6.877034984934912e-06, "loss": 1.4731, "step": 19330 }, { "epoch": 1.5149686520376175, "grad_norm": 2.802821636199951, "learning_rate": 6.87674917288664e-06, "loss": 1.4415, "step": 19331 }, { "epoch": 1.5150470219435737, "grad_norm": 4.246772289276123, "learning_rate": 6.876463353700178e-06, "loss": 1.2941, "step": 19332 }, { "epoch": 1.5151253918495298, "grad_norm": 2.335143804550171, "learning_rate": 6.876177527376613e-06, "loss": 1.5822, "step": 19333 }, { "epoch": 1.5152037617554859, "grad_norm": 3.412741184234619, "learning_rate": 6.8758916939170325e-06, "loss": 1.4198, "step": 19334 }, { "epoch": 1.515282131661442, "grad_norm": 6.078891277313232, "learning_rate": 6.875605853322525e-06, "loss": 1.4472, "step": 19335 }, { "epoch": 1.515360501567398, "grad_norm": 3.260162115097046, "learning_rate": 6.875320005594175e-06, "loss": 1.4336, "step": 19336 }, { "epoch": 1.5154388714733542, "grad_norm": 7.8378496170043945, "learning_rate": 6.875034150733071e-06, "loss": 1.2311, "step": 19337 }, { "epoch": 1.5155172413793103, "grad_norm": 3.292686700820923, "learning_rate": 6.8747482887403e-06, "loss": 1.4169, "step": 19338 }, { "epoch": 1.5155956112852664, "grad_norm": 3.4269816875457764, "learning_rate": 6.874462419616949e-06, "loss": 1.4106, "step": 19339 }, { "epoch": 1.5156739811912225, "grad_norm": 2.0878260135650635, "learning_rate": 6.874176543364108e-06, "loss": 1.1393, "step": 19340 }, { "epoch": 1.5157523510971787, "grad_norm": 3.903775453567505, "learning_rate": 6.873890659982859e-06, "loss": 1.4302, "step": 19341 }, { "epoch": 1.5158307210031348, "grad_norm": 3.2481377124786377, "learning_rate": 6.873604769474296e-06, "loss": 1.429, "step": 19342 }, { "epoch": 1.5159090909090909, "grad_norm": 4.3975510597229, "learning_rate": 6.873318871839498e-06, "loss": 0.6933, "step": 19343 }, { "epoch": 1.515987460815047, "grad_norm": 4.777590751647949, "learning_rate": 6.873032967079562e-06, "loss": 1.2987, "step": 19344 }, { "epoch": 1.516065830721003, "grad_norm": 2.5850143432617188, "learning_rate": 6.872747055195567e-06, "loss": 1.5351, "step": 19345 }, { "epoch": 1.5161442006269592, "grad_norm": 4.15206241607666, "learning_rate": 6.872461136188607e-06, "loss": 1.4992, "step": 19346 }, { "epoch": 1.5162225705329153, "grad_norm": 2.303276300430298, "learning_rate": 6.872175210059764e-06, "loss": 1.5168, "step": 19347 }, { "epoch": 1.5163009404388714, "grad_norm": 2.8768043518066406, "learning_rate": 6.87188927681013e-06, "loss": 1.2756, "step": 19348 }, { "epoch": 1.5163793103448275, "grad_norm": 5.165632724761963, "learning_rate": 6.87160333644079e-06, "loss": 1.0969, "step": 19349 }, { "epoch": 1.5164576802507836, "grad_norm": 3.0623199939727783, "learning_rate": 6.871317388952831e-06, "loss": 1.3098, "step": 19350 }, { "epoch": 1.5165360501567398, "grad_norm": 1.8958911895751953, "learning_rate": 6.871031434347344e-06, "loss": 1.3767, "step": 19351 }, { "epoch": 1.5166144200626959, "grad_norm": 4.477323055267334, "learning_rate": 6.870745472625413e-06, "loss": 1.4923, "step": 19352 }, { "epoch": 1.5166927899686522, "grad_norm": 5.1887335777282715, "learning_rate": 6.8704595037881284e-06, "loss": 1.745, "step": 19353 }, { "epoch": 1.516771159874608, "grad_norm": 1.8672575950622559, "learning_rate": 6.8701735278365756e-06, "loss": 1.1504, "step": 19354 }, { "epoch": 1.5168495297805644, "grad_norm": 2.8947510719299316, "learning_rate": 6.8698875447718425e-06, "loss": 1.6305, "step": 19355 }, { "epoch": 1.5169278996865203, "grad_norm": 2.437103033065796, "learning_rate": 6.869601554595019e-06, "loss": 1.4124, "step": 19356 }, { "epoch": 1.5170062695924766, "grad_norm": 3.9511144161224365, "learning_rate": 6.869315557307191e-06, "loss": 1.109, "step": 19357 }, { "epoch": 1.5170846394984325, "grad_norm": 2.33418345451355, "learning_rate": 6.869029552909448e-06, "loss": 1.4027, "step": 19358 }, { "epoch": 1.5171630094043889, "grad_norm": 3.0210437774658203, "learning_rate": 6.868743541402874e-06, "loss": 1.1536, "step": 19359 }, { "epoch": 1.5172413793103448, "grad_norm": 2.55861234664917, "learning_rate": 6.868457522788561e-06, "loss": 1.0518, "step": 19360 }, { "epoch": 1.517319749216301, "grad_norm": 1.9892265796661377, "learning_rate": 6.868171497067595e-06, "loss": 1.4123, "step": 19361 }, { "epoch": 1.517398119122257, "grad_norm": 3.274127960205078, "learning_rate": 6.8678854642410644e-06, "loss": 1.2816, "step": 19362 }, { "epoch": 1.5174764890282133, "grad_norm": 4.025475978851318, "learning_rate": 6.867599424310058e-06, "loss": 1.9106, "step": 19363 }, { "epoch": 1.5175548589341692, "grad_norm": 2.643108367919922, "learning_rate": 6.867313377275661e-06, "loss": 1.4657, "step": 19364 }, { "epoch": 1.5176332288401255, "grad_norm": 7.228840351104736, "learning_rate": 6.867027323138963e-06, "loss": 1.8619, "step": 19365 }, { "epoch": 1.5177115987460814, "grad_norm": 2.514460802078247, "learning_rate": 6.866741261901052e-06, "loss": 1.5301, "step": 19366 }, { "epoch": 1.5177899686520377, "grad_norm": 6.19735050201416, "learning_rate": 6.866455193563017e-06, "loss": 1.0896, "step": 19367 }, { "epoch": 1.5178683385579936, "grad_norm": 5.767351150512695, "learning_rate": 6.866169118125944e-06, "loss": 1.1935, "step": 19368 }, { "epoch": 1.51794670846395, "grad_norm": 2.6856439113616943, "learning_rate": 6.8658830355909225e-06, "loss": 1.2387, "step": 19369 }, { "epoch": 1.5180250783699059, "grad_norm": 2.2108683586120605, "learning_rate": 6.865596945959041e-06, "loss": 1.2368, "step": 19370 }, { "epoch": 1.5181034482758622, "grad_norm": 2.705396890640259, "learning_rate": 6.865310849231386e-06, "loss": 1.4883, "step": 19371 }, { "epoch": 1.518181818181818, "grad_norm": 2.947387218475342, "learning_rate": 6.865024745409048e-06, "loss": 1.1858, "step": 19372 }, { "epoch": 1.5182601880877744, "grad_norm": 2.0452120304107666, "learning_rate": 6.864738634493112e-06, "loss": 1.2912, "step": 19373 }, { "epoch": 1.5183385579937303, "grad_norm": 4.270471572875977, "learning_rate": 6.86445251648467e-06, "loss": 1.2815, "step": 19374 }, { "epoch": 1.5184169278996866, "grad_norm": 4.326528549194336, "learning_rate": 6.864166391384808e-06, "loss": 0.9215, "step": 19375 }, { "epoch": 1.5184952978056425, "grad_norm": 3.250654935836792, "learning_rate": 6.863880259194613e-06, "loss": 1.2862, "step": 19376 }, { "epoch": 1.5185736677115989, "grad_norm": 2.979522228240967, "learning_rate": 6.863594119915177e-06, "loss": 1.2236, "step": 19377 }, { "epoch": 1.5186520376175547, "grad_norm": 3.070305585861206, "learning_rate": 6.863307973547583e-06, "loss": 1.4062, "step": 19378 }, { "epoch": 1.518730407523511, "grad_norm": 3.7011077404022217, "learning_rate": 6.863021820092926e-06, "loss": 1.2256, "step": 19379 }, { "epoch": 1.518808777429467, "grad_norm": 2.045788288116455, "learning_rate": 6.862735659552288e-06, "loss": 0.9774, "step": 19380 }, { "epoch": 1.5188871473354233, "grad_norm": 3.885990619659424, "learning_rate": 6.862449491926761e-06, "loss": 1.254, "step": 19381 }, { "epoch": 1.5189655172413792, "grad_norm": 3.112450122833252, "learning_rate": 6.862163317217434e-06, "loss": 1.4343, "step": 19382 }, { "epoch": 1.5190438871473355, "grad_norm": 2.713855504989624, "learning_rate": 6.861877135425394e-06, "loss": 1.5209, "step": 19383 }, { "epoch": 1.5191222570532914, "grad_norm": 2.5736217498779297, "learning_rate": 6.861590946551728e-06, "loss": 1.4082, "step": 19384 }, { "epoch": 1.5192006269592477, "grad_norm": 3.5643999576568604, "learning_rate": 6.861304750597529e-06, "loss": 1.3277, "step": 19385 }, { "epoch": 1.5192789968652036, "grad_norm": 2.7451577186584473, "learning_rate": 6.8610185475638805e-06, "loss": 1.3514, "step": 19386 }, { "epoch": 1.51935736677116, "grad_norm": 3.4225070476531982, "learning_rate": 6.860732337451874e-06, "loss": 1.4089, "step": 19387 }, { "epoch": 1.5194357366771158, "grad_norm": 2.4294049739837646, "learning_rate": 6.860446120262597e-06, "loss": 1.5904, "step": 19388 }, { "epoch": 1.5195141065830722, "grad_norm": 5.390928745269775, "learning_rate": 6.860159895997141e-06, "loss": 1.308, "step": 19389 }, { "epoch": 1.519592476489028, "grad_norm": 2.386261463165283, "learning_rate": 6.859873664656588e-06, "loss": 1.6327, "step": 19390 }, { "epoch": 1.5196708463949844, "grad_norm": 2.8506534099578857, "learning_rate": 6.859587426242034e-06, "loss": 1.5362, "step": 19391 }, { "epoch": 1.5197492163009403, "grad_norm": 1.9105148315429688, "learning_rate": 6.8593011807545625e-06, "loss": 1.2413, "step": 19392 }, { "epoch": 1.5198275862068966, "grad_norm": 2.888059139251709, "learning_rate": 6.859014928195266e-06, "loss": 1.1668, "step": 19393 }, { "epoch": 1.5199059561128525, "grad_norm": 3.021732807159424, "learning_rate": 6.858728668565231e-06, "loss": 1.2987, "step": 19394 }, { "epoch": 1.5199843260188088, "grad_norm": 4.610328197479248, "learning_rate": 6.858442401865547e-06, "loss": 1.7226, "step": 19395 }, { "epoch": 1.5200626959247647, "grad_norm": 4.473758697509766, "learning_rate": 6.858156128097303e-06, "loss": 1.2185, "step": 19396 }, { "epoch": 1.520141065830721, "grad_norm": 2.684285879135132, "learning_rate": 6.857869847261587e-06, "loss": 1.2996, "step": 19397 }, { "epoch": 1.520219435736677, "grad_norm": 2.5998642444610596, "learning_rate": 6.857583559359489e-06, "loss": 1.2912, "step": 19398 }, { "epoch": 1.5202978056426333, "grad_norm": 3.7778685092926025, "learning_rate": 6.8572972643920956e-06, "loss": 1.5312, "step": 19399 }, { "epoch": 1.5203761755485894, "grad_norm": 4.3593573570251465, "learning_rate": 6.857010962360498e-06, "loss": 1.5459, "step": 19400 }, { "epoch": 1.5204545454545455, "grad_norm": 2.5779998302459717, "learning_rate": 6.856724653265785e-06, "loss": 1.5528, "step": 19401 }, { "epoch": 1.5205329153605016, "grad_norm": 2.034196615219116, "learning_rate": 6.856438337109046e-06, "loss": 1.5142, "step": 19402 }, { "epoch": 1.5206112852664577, "grad_norm": 2.914944648742676, "learning_rate": 6.8561520138913675e-06, "loss": 2.0765, "step": 19403 }, { "epoch": 1.5206896551724138, "grad_norm": 5.389361381530762, "learning_rate": 6.8558656836138405e-06, "loss": 1.3248, "step": 19404 }, { "epoch": 1.52076802507837, "grad_norm": 3.329514265060425, "learning_rate": 6.855579346277554e-06, "loss": 1.3079, "step": 19405 }, { "epoch": 1.520846394984326, "grad_norm": 2.6317317485809326, "learning_rate": 6.8552930018835966e-06, "loss": 1.3168, "step": 19406 }, { "epoch": 1.5209247648902822, "grad_norm": 2.835925817489624, "learning_rate": 6.855006650433058e-06, "loss": 1.48, "step": 19407 }, { "epoch": 1.5210031347962383, "grad_norm": 3.459076166152954, "learning_rate": 6.8547202919270265e-06, "loss": 1.5295, "step": 19408 }, { "epoch": 1.5210815047021944, "grad_norm": 4.987934589385986, "learning_rate": 6.854433926366591e-06, "loss": 1.558, "step": 19409 }, { "epoch": 1.5211598746081505, "grad_norm": 5.038321018218994, "learning_rate": 6.854147553752842e-06, "loss": 1.1496, "step": 19410 }, { "epoch": 1.5212382445141066, "grad_norm": 1.9969297647476196, "learning_rate": 6.853861174086869e-06, "loss": 1.3457, "step": 19411 }, { "epoch": 1.5213166144200627, "grad_norm": 3.8587746620178223, "learning_rate": 6.853574787369759e-06, "loss": 1.2763, "step": 19412 }, { "epoch": 1.5213949843260188, "grad_norm": 3.4192540645599365, "learning_rate": 6.853288393602603e-06, "loss": 1.1837, "step": 19413 }, { "epoch": 1.521473354231975, "grad_norm": 2.6494505405426025, "learning_rate": 6.85300199278649e-06, "loss": 1.1427, "step": 19414 }, { "epoch": 1.521551724137931, "grad_norm": 3.571239948272705, "learning_rate": 6.852715584922508e-06, "loss": 1.3176, "step": 19415 }, { "epoch": 1.5216300940438872, "grad_norm": 2.7457776069641113, "learning_rate": 6.852429170011749e-06, "loss": 1.3364, "step": 19416 }, { "epoch": 1.5217084639498433, "grad_norm": 3.7080180644989014, "learning_rate": 6.852142748055302e-06, "loss": 1.2304, "step": 19417 }, { "epoch": 1.5217868338557994, "grad_norm": 7.287753582000732, "learning_rate": 6.851856319054251e-06, "loss": 0.7148, "step": 19418 }, { "epoch": 1.5218652037617555, "grad_norm": 4.854672431945801, "learning_rate": 6.851569883009694e-06, "loss": 1.346, "step": 19419 }, { "epoch": 1.5219435736677116, "grad_norm": 2.3020904064178467, "learning_rate": 6.851283439922714e-06, "loss": 1.6145, "step": 19420 }, { "epoch": 1.5220219435736677, "grad_norm": 2.406273365020752, "learning_rate": 6.850996989794404e-06, "loss": 1.0018, "step": 19421 }, { "epoch": 1.5221003134796238, "grad_norm": 4.102041721343994, "learning_rate": 6.850710532625851e-06, "loss": 1.7274, "step": 19422 }, { "epoch": 1.52217868338558, "grad_norm": 2.322871208190918, "learning_rate": 6.8504240684181464e-06, "loss": 1.3656, "step": 19423 }, { "epoch": 1.522257053291536, "grad_norm": 4.206876277923584, "learning_rate": 6.8501375971723794e-06, "loss": 1.0964, "step": 19424 }, { "epoch": 1.5223354231974922, "grad_norm": 1.936326503753662, "learning_rate": 6.849851118889638e-06, "loss": 1.1469, "step": 19425 }, { "epoch": 1.5224137931034483, "grad_norm": 1.9984447956085205, "learning_rate": 6.849564633571014e-06, "loss": 1.4865, "step": 19426 }, { "epoch": 1.5224921630094044, "grad_norm": 2.590442657470703, "learning_rate": 6.849278141217594e-06, "loss": 1.5016, "step": 19427 }, { "epoch": 1.5225705329153605, "grad_norm": 2.547030448913574, "learning_rate": 6.848991641830474e-06, "loss": 1.1396, "step": 19428 }, { "epoch": 1.5226489028213166, "grad_norm": 2.1968302726745605, "learning_rate": 6.848705135410737e-06, "loss": 1.5872, "step": 19429 }, { "epoch": 1.5227272727272727, "grad_norm": 7.748044013977051, "learning_rate": 6.848418621959476e-06, "loss": 1.5469, "step": 19430 }, { "epoch": 1.5228056426332288, "grad_norm": 4.462237358093262, "learning_rate": 6.848132101477779e-06, "loss": 1.3809, "step": 19431 }, { "epoch": 1.522884012539185, "grad_norm": 3.352725028991699, "learning_rate": 6.847845573966737e-06, "loss": 1.2978, "step": 19432 }, { "epoch": 1.522962382445141, "grad_norm": 2.4355645179748535, "learning_rate": 6.84755903942744e-06, "loss": 1.7471, "step": 19433 }, { "epoch": 1.5230407523510971, "grad_norm": 3.3683743476867676, "learning_rate": 6.847272497860977e-06, "loss": 1.1988, "step": 19434 }, { "epoch": 1.5231191222570533, "grad_norm": 2.607652187347412, "learning_rate": 6.846985949268439e-06, "loss": 1.0826, "step": 19435 }, { "epoch": 1.5231974921630094, "grad_norm": 3.931209087371826, "learning_rate": 6.846699393650914e-06, "loss": 1.2255, "step": 19436 }, { "epoch": 1.5232758620689655, "grad_norm": 3.515437364578247, "learning_rate": 6.846412831009493e-06, "loss": 1.3458, "step": 19437 }, { "epoch": 1.5233542319749216, "grad_norm": 4.080391883850098, "learning_rate": 6.846126261345267e-06, "loss": 2.0254, "step": 19438 }, { "epoch": 1.5234326018808777, "grad_norm": 2.627762794494629, "learning_rate": 6.845839684659325e-06, "loss": 1.3106, "step": 19439 }, { "epoch": 1.5235109717868338, "grad_norm": 2.20402455329895, "learning_rate": 6.845553100952755e-06, "loss": 1.2837, "step": 19440 }, { "epoch": 1.52358934169279, "grad_norm": 5.842122554779053, "learning_rate": 6.845266510226651e-06, "loss": 1.5997, "step": 19441 }, { "epoch": 1.523667711598746, "grad_norm": 2.225052833557129, "learning_rate": 6.8449799124821e-06, "loss": 1.468, "step": 19442 }, { "epoch": 1.5237460815047021, "grad_norm": 2.346708297729492, "learning_rate": 6.844693307720193e-06, "loss": 1.7799, "step": 19443 }, { "epoch": 1.5238244514106583, "grad_norm": 2.5982959270477295, "learning_rate": 6.84440669594202e-06, "loss": 1.4615, "step": 19444 }, { "epoch": 1.5239028213166144, "grad_norm": 3.7255566120147705, "learning_rate": 6.844120077148673e-06, "loss": 1.6213, "step": 19445 }, { "epoch": 1.5239811912225707, "grad_norm": 2.7462587356567383, "learning_rate": 6.843833451341238e-06, "loss": 1.3208, "step": 19446 }, { "epoch": 1.5240595611285266, "grad_norm": 2.4305450916290283, "learning_rate": 6.84354681852081e-06, "loss": 1.0465, "step": 19447 }, { "epoch": 1.524137931034483, "grad_norm": 5.413342475891113, "learning_rate": 6.8432601786884755e-06, "loss": 1.5173, "step": 19448 }, { "epoch": 1.5242163009404388, "grad_norm": 3.6337997913360596, "learning_rate": 6.842973531845326e-06, "loss": 1.6351, "step": 19449 }, { "epoch": 1.5242946708463951, "grad_norm": 2.1572961807250977, "learning_rate": 6.842686877992453e-06, "loss": 1.3662, "step": 19450 }, { "epoch": 1.524373040752351, "grad_norm": 2.3204398155212402, "learning_rate": 6.842400217130943e-06, "loss": 1.3782, "step": 19451 }, { "epoch": 1.5244514106583074, "grad_norm": 3.843912124633789, "learning_rate": 6.842113549261892e-06, "loss": 1.4553, "step": 19452 }, { "epoch": 1.5245297805642632, "grad_norm": 3.542534351348877, "learning_rate": 6.8418268743863836e-06, "loss": 1.086, "step": 19453 }, { "epoch": 1.5246081504702196, "grad_norm": 6.957254409790039, "learning_rate": 6.841540192505516e-06, "loss": 0.6987, "step": 19454 }, { "epoch": 1.5246865203761755, "grad_norm": 2.330120801925659, "learning_rate": 6.841253503620372e-06, "loss": 1.3069, "step": 19455 }, { "epoch": 1.5247648902821318, "grad_norm": 7.415915489196777, "learning_rate": 6.840966807732047e-06, "loss": 1.4626, "step": 19456 }, { "epoch": 1.5248432601880877, "grad_norm": 1.9969903230667114, "learning_rate": 6.840680104841629e-06, "loss": 1.4117, "step": 19457 }, { "epoch": 1.524921630094044, "grad_norm": 2.4258475303649902, "learning_rate": 6.84039339495021e-06, "loss": 1.3391, "step": 19458 }, { "epoch": 1.525, "grad_norm": 2.659412145614624, "learning_rate": 6.84010667805888e-06, "loss": 1.4779, "step": 19459 }, { "epoch": 1.5250783699059562, "grad_norm": 2.172266721725464, "learning_rate": 6.8398199541687284e-06, "loss": 1.3586, "step": 19460 }, { "epoch": 1.5251567398119121, "grad_norm": 7.739894866943359, "learning_rate": 6.839533223280848e-06, "loss": 1.4772, "step": 19461 }, { "epoch": 1.5252351097178685, "grad_norm": 2.431816816329956, "learning_rate": 6.8392464853963255e-06, "loss": 1.0043, "step": 19462 }, { "epoch": 1.5253134796238244, "grad_norm": 3.7836077213287354, "learning_rate": 6.838959740516257e-06, "loss": 1.3705, "step": 19463 }, { "epoch": 1.5253918495297807, "grad_norm": 3.920397996902466, "learning_rate": 6.838672988641729e-06, "loss": 1.2936, "step": 19464 }, { "epoch": 1.5254702194357366, "grad_norm": 3.7288084030151367, "learning_rate": 6.838386229773834e-06, "loss": 1.104, "step": 19465 }, { "epoch": 1.525548589341693, "grad_norm": 3.261013984680176, "learning_rate": 6.8380994639136615e-06, "loss": 1.2806, "step": 19466 }, { "epoch": 1.5256269592476488, "grad_norm": 2.321632146835327, "learning_rate": 6.837812691062302e-06, "loss": 1.4832, "step": 19467 }, { "epoch": 1.5257053291536051, "grad_norm": 2.325047492980957, "learning_rate": 6.837525911220847e-06, "loss": 1.0688, "step": 19468 }, { "epoch": 1.525783699059561, "grad_norm": 2.2372260093688965, "learning_rate": 6.837239124390388e-06, "loss": 1.2133, "step": 19469 }, { "epoch": 1.5258620689655173, "grad_norm": 4.514373779296875, "learning_rate": 6.836952330572015e-06, "loss": 1.2752, "step": 19470 }, { "epoch": 1.5259404388714732, "grad_norm": 2.1187942028045654, "learning_rate": 6.83666552976682e-06, "loss": 1.2808, "step": 19471 }, { "epoch": 1.5260188087774296, "grad_norm": 4.460914134979248, "learning_rate": 6.836378721975889e-06, "loss": 1.4574, "step": 19472 }, { "epoch": 1.5260971786833855, "grad_norm": 3.97052264213562, "learning_rate": 6.836091907200321e-06, "loss": 1.0919, "step": 19473 }, { "epoch": 1.5261755485893418, "grad_norm": 4.0926432609558105, "learning_rate": 6.8358050854411995e-06, "loss": 1.5508, "step": 19474 }, { "epoch": 1.5262539184952977, "grad_norm": 9.657179832458496, "learning_rate": 6.835518256699621e-06, "loss": 1.2533, "step": 19475 }, { "epoch": 1.526332288401254, "grad_norm": 2.1737966537475586, "learning_rate": 6.835231420976672e-06, "loss": 1.7368, "step": 19476 }, { "epoch": 1.52641065830721, "grad_norm": 3.339015007019043, "learning_rate": 6.834944578273444e-06, "loss": 1.6542, "step": 19477 }, { "epoch": 1.5264890282131662, "grad_norm": 5.102196216583252, "learning_rate": 6.8346577285910324e-06, "loss": 1.5903, "step": 19478 }, { "epoch": 1.5265673981191221, "grad_norm": 3.598731279373169, "learning_rate": 6.8343708719305225e-06, "loss": 1.4793, "step": 19479 }, { "epoch": 1.5266457680250785, "grad_norm": 3.6440470218658447, "learning_rate": 6.834084008293009e-06, "loss": 1.2352, "step": 19480 }, { "epoch": 1.5267241379310343, "grad_norm": 2.8697707653045654, "learning_rate": 6.833797137679582e-06, "loss": 1.2845, "step": 19481 }, { "epoch": 1.5268025078369907, "grad_norm": 4.910980224609375, "learning_rate": 6.833510260091334e-06, "loss": 1.9022, "step": 19482 }, { "epoch": 1.5268808777429466, "grad_norm": 7.269101619720459, "learning_rate": 6.833223375529352e-06, "loss": 1.4124, "step": 19483 }, { "epoch": 1.526959247648903, "grad_norm": 3.4404115676879883, "learning_rate": 6.832936483994732e-06, "loss": 1.3031, "step": 19484 }, { "epoch": 1.5270376175548588, "grad_norm": 2.817453384399414, "learning_rate": 6.832649585488562e-06, "loss": 1.4867, "step": 19485 }, { "epoch": 1.5271159874608151, "grad_norm": 2.5824596881866455, "learning_rate": 6.832362680011935e-06, "loss": 1.4979, "step": 19486 }, { "epoch": 1.527194357366771, "grad_norm": 2.2693605422973633, "learning_rate": 6.83207576756594e-06, "loss": 1.4499, "step": 19487 }, { "epoch": 1.5272727272727273, "grad_norm": 5.228864669799805, "learning_rate": 6.831788848151671e-06, "loss": 1.6617, "step": 19488 }, { "epoch": 1.5273510971786832, "grad_norm": 11.09290599822998, "learning_rate": 6.831501921770219e-06, "loss": 1.6264, "step": 19489 }, { "epoch": 1.5274294670846396, "grad_norm": 2.973501682281494, "learning_rate": 6.831214988422673e-06, "loss": 1.5152, "step": 19490 }, { "epoch": 1.5275078369905954, "grad_norm": 1.9340921640396118, "learning_rate": 6.8309280481101255e-06, "loss": 1.2609, "step": 19491 }, { "epoch": 1.5275862068965518, "grad_norm": 1.9446266889572144, "learning_rate": 6.830641100833669e-06, "loss": 1.3339, "step": 19492 }, { "epoch": 1.527664576802508, "grad_norm": 3.2104551792144775, "learning_rate": 6.8303541465943945e-06, "loss": 1.3757, "step": 19493 }, { "epoch": 1.527742946708464, "grad_norm": 3.2464375495910645, "learning_rate": 6.830067185393392e-06, "loss": 1.2534, "step": 19494 }, { "epoch": 1.52782131661442, "grad_norm": 3.496441602706909, "learning_rate": 6.829780217231754e-06, "loss": 1.5284, "step": 19495 }, { "epoch": 1.5278996865203762, "grad_norm": 2.1941401958465576, "learning_rate": 6.829493242110572e-06, "loss": 1.3333, "step": 19496 }, { "epoch": 1.5279780564263323, "grad_norm": 2.816765546798706, "learning_rate": 6.829206260030939e-06, "loss": 1.4758, "step": 19497 }, { "epoch": 1.5280564263322884, "grad_norm": 2.7205562591552734, "learning_rate": 6.828919270993941e-06, "loss": 1.4348, "step": 19498 }, { "epoch": 1.5281347962382446, "grad_norm": 2.794994354248047, "learning_rate": 6.828632275000677e-06, "loss": 1.2721, "step": 19499 }, { "epoch": 1.5282131661442007, "grad_norm": 3.517221689224243, "learning_rate": 6.828345272052233e-06, "loss": 1.66, "step": 19500 }, { "epoch": 1.5282915360501568, "grad_norm": 1.7273246049880981, "learning_rate": 6.828058262149705e-06, "loss": 1.2951, "step": 19501 }, { "epoch": 1.5283699059561129, "grad_norm": 5.762658596038818, "learning_rate": 6.827771245294181e-06, "loss": 1.1716, "step": 19502 }, { "epoch": 1.528448275862069, "grad_norm": 2.4552907943725586, "learning_rate": 6.827484221486753e-06, "loss": 1.4771, "step": 19503 }, { "epoch": 1.528526645768025, "grad_norm": 4.18879508972168, "learning_rate": 6.827197190728515e-06, "loss": 1.4928, "step": 19504 }, { "epoch": 1.5286050156739812, "grad_norm": 3.610539674758911, "learning_rate": 6.826910153020557e-06, "loss": 1.3572, "step": 19505 }, { "epoch": 1.5286833855799373, "grad_norm": 2.5351343154907227, "learning_rate": 6.826623108363972e-06, "loss": 1.2485, "step": 19506 }, { "epoch": 1.5287617554858934, "grad_norm": 3.2958011627197266, "learning_rate": 6.826336056759849e-06, "loss": 1.1505, "step": 19507 }, { "epoch": 1.5288401253918495, "grad_norm": 3.562098503112793, "learning_rate": 6.826048998209284e-06, "loss": 1.3753, "step": 19508 }, { "epoch": 1.5289184952978057, "grad_norm": 8.134197235107422, "learning_rate": 6.825761932713365e-06, "loss": 1.2633, "step": 19509 }, { "epoch": 1.5289968652037618, "grad_norm": 3.2889487743377686, "learning_rate": 6.825474860273186e-06, "loss": 1.5897, "step": 19510 }, { "epoch": 1.5290752351097179, "grad_norm": 2.187926769256592, "learning_rate": 6.825187780889837e-06, "loss": 1.5945, "step": 19511 }, { "epoch": 1.529153605015674, "grad_norm": 3.2242965698242188, "learning_rate": 6.824900694564413e-06, "loss": 1.4208, "step": 19512 }, { "epoch": 1.52923197492163, "grad_norm": 3.0049140453338623, "learning_rate": 6.824613601298003e-06, "loss": 1.4417, "step": 19513 }, { "epoch": 1.5293103448275862, "grad_norm": 3.766120195388794, "learning_rate": 6.824326501091699e-06, "loss": 1.7926, "step": 19514 }, { "epoch": 1.5293887147335423, "grad_norm": 2.676332950592041, "learning_rate": 6.824039393946595e-06, "loss": 1.6036, "step": 19515 }, { "epoch": 1.5294670846394984, "grad_norm": 3.7059569358825684, "learning_rate": 6.823752279863782e-06, "loss": 1.6442, "step": 19516 }, { "epoch": 1.5295454545454545, "grad_norm": 2.888326406478882, "learning_rate": 6.823465158844354e-06, "loss": 1.4307, "step": 19517 }, { "epoch": 1.5296238244514107, "grad_norm": 2.910255193710327, "learning_rate": 6.823178030889399e-06, "loss": 1.2654, "step": 19518 }, { "epoch": 1.5297021943573668, "grad_norm": 3.1059226989746094, "learning_rate": 6.822890896000011e-06, "loss": 1.2929, "step": 19519 }, { "epoch": 1.5297805642633229, "grad_norm": 2.299304962158203, "learning_rate": 6.8226037541772835e-06, "loss": 1.0208, "step": 19520 }, { "epoch": 1.529858934169279, "grad_norm": 12.138104438781738, "learning_rate": 6.822316605422307e-06, "loss": 1.5971, "step": 19521 }, { "epoch": 1.529937304075235, "grad_norm": 2.8397719860076904, "learning_rate": 6.822029449736174e-06, "loss": 1.6087, "step": 19522 }, { "epoch": 1.5300156739811912, "grad_norm": 7.832496643066406, "learning_rate": 6.821742287119977e-06, "loss": 1.3237, "step": 19523 }, { "epoch": 1.5300940438871473, "grad_norm": 2.4676196575164795, "learning_rate": 6.821455117574808e-06, "loss": 1.3137, "step": 19524 }, { "epoch": 1.5301724137931034, "grad_norm": 3.074486494064331, "learning_rate": 6.82116794110176e-06, "loss": 1.1873, "step": 19525 }, { "epoch": 1.5302507836990595, "grad_norm": 2.5862860679626465, "learning_rate": 6.820880757701922e-06, "loss": 1.2917, "step": 19526 }, { "epoch": 1.5303291536050156, "grad_norm": 4.548020839691162, "learning_rate": 6.820593567376392e-06, "loss": 1.0773, "step": 19527 }, { "epoch": 1.5304075235109718, "grad_norm": 2.4935262203216553, "learning_rate": 6.820306370126258e-06, "loss": 1.2518, "step": 19528 }, { "epoch": 1.5304858934169279, "grad_norm": 3.183347225189209, "learning_rate": 6.820019165952614e-06, "loss": 1.4996, "step": 19529 }, { "epoch": 1.530564263322884, "grad_norm": 3.320075035095215, "learning_rate": 6.81973195485655e-06, "loss": 1.6269, "step": 19530 }, { "epoch": 1.53064263322884, "grad_norm": 2.1376001834869385, "learning_rate": 6.819444736839162e-06, "loss": 1.3788, "step": 19531 }, { "epoch": 1.5307210031347962, "grad_norm": 3.0248513221740723, "learning_rate": 6.819157511901542e-06, "loss": 0.9158, "step": 19532 }, { "epoch": 1.5307993730407523, "grad_norm": 3.97519850730896, "learning_rate": 6.818870280044778e-06, "loss": 1.1123, "step": 19533 }, { "epoch": 1.5308777429467084, "grad_norm": 3.1988046169281006, "learning_rate": 6.818583041269968e-06, "loss": 1.3298, "step": 19534 }, { "epoch": 1.5309561128526645, "grad_norm": 2.947751998901367, "learning_rate": 6.8182957955782005e-06, "loss": 1.5303, "step": 19535 }, { "epoch": 1.5310344827586206, "grad_norm": 3.518477439880371, "learning_rate": 6.8180085429705715e-06, "loss": 1.4154, "step": 19536 }, { "epoch": 1.5311128526645768, "grad_norm": 4.339463233947754, "learning_rate": 6.817721283448171e-06, "loss": 1.3632, "step": 19537 }, { "epoch": 1.5311912225705329, "grad_norm": 3.2121057510375977, "learning_rate": 6.8174340170120924e-06, "loss": 1.0027, "step": 19538 }, { "epoch": 1.5312695924764892, "grad_norm": 3.5332984924316406, "learning_rate": 6.817146743663428e-06, "loss": 1.7167, "step": 19539 }, { "epoch": 1.531347962382445, "grad_norm": 6.903076648712158, "learning_rate": 6.816859463403271e-06, "loss": 1.464, "step": 19540 }, { "epoch": 1.5314263322884014, "grad_norm": 1.8843591213226318, "learning_rate": 6.816572176232713e-06, "loss": 1.3687, "step": 19541 }, { "epoch": 1.5315047021943573, "grad_norm": 10.255890846252441, "learning_rate": 6.816284882152848e-06, "loss": 1.4082, "step": 19542 }, { "epoch": 1.5315830721003136, "grad_norm": 3.938840627670288, "learning_rate": 6.815997581164768e-06, "loss": 1.4513, "step": 19543 }, { "epoch": 1.5316614420062695, "grad_norm": 11.022997856140137, "learning_rate": 6.8157102732695675e-06, "loss": 1.9668, "step": 19544 }, { "epoch": 1.5317398119122259, "grad_norm": 2.3815181255340576, "learning_rate": 6.815422958468338e-06, "loss": 1.5153, "step": 19545 }, { "epoch": 1.5318181818181817, "grad_norm": 4.07989501953125, "learning_rate": 6.81513563676217e-06, "loss": 1.4556, "step": 19546 }, { "epoch": 1.531896551724138, "grad_norm": 13.39150619506836, "learning_rate": 6.814848308152159e-06, "loss": 0.8875, "step": 19547 }, { "epoch": 1.531974921630094, "grad_norm": 2.9924862384796143, "learning_rate": 6.814560972639397e-06, "loss": 1.3908, "step": 19548 }, { "epoch": 1.5320532915360503, "grad_norm": 3.9195854663848877, "learning_rate": 6.814273630224977e-06, "loss": 1.4502, "step": 19549 }, { "epoch": 1.5321316614420062, "grad_norm": 5.466148376464844, "learning_rate": 6.813986280909992e-06, "loss": 1.4124, "step": 19550 }, { "epoch": 1.5322100313479625, "grad_norm": 2.0493016242980957, "learning_rate": 6.8136989246955364e-06, "loss": 1.2302, "step": 19551 }, { "epoch": 1.5322884012539184, "grad_norm": 7.425384521484375, "learning_rate": 6.813411561582699e-06, "loss": 1.1788, "step": 19552 }, { "epoch": 1.5323667711598747, "grad_norm": 2.93571400642395, "learning_rate": 6.813124191572578e-06, "loss": 1.7214, "step": 19553 }, { "epoch": 1.5324451410658306, "grad_norm": 2.531015634536743, "learning_rate": 6.812836814666262e-06, "loss": 1.2453, "step": 19554 }, { "epoch": 1.532523510971787, "grad_norm": 3.8089826107025146, "learning_rate": 6.812549430864847e-06, "loss": 1.4409, "step": 19555 }, { "epoch": 1.5326018808777429, "grad_norm": 3.831406593322754, "learning_rate": 6.812262040169425e-06, "loss": 1.4617, "step": 19556 }, { "epoch": 1.5326802507836992, "grad_norm": 2.8824105262756348, "learning_rate": 6.811974642581089e-06, "loss": 1.5654, "step": 19557 }, { "epoch": 1.532758620689655, "grad_norm": 5.60394287109375, "learning_rate": 6.811687238100932e-06, "loss": 1.1521, "step": 19558 }, { "epoch": 1.5328369905956114, "grad_norm": 3.3668999671936035, "learning_rate": 6.811399826730046e-06, "loss": 1.5008, "step": 19559 }, { "epoch": 1.5329153605015673, "grad_norm": 6.267885208129883, "learning_rate": 6.811112408469528e-06, "loss": 1.0296, "step": 19560 }, { "epoch": 1.5329937304075236, "grad_norm": 3.1753134727478027, "learning_rate": 6.810824983320465e-06, "loss": 1.797, "step": 19561 }, { "epoch": 1.5330721003134795, "grad_norm": 2.427408218383789, "learning_rate": 6.8105375512839574e-06, "loss": 1.2941, "step": 19562 }, { "epoch": 1.5331504702194358, "grad_norm": 2.5475594997406006, "learning_rate": 6.810250112361093e-06, "loss": 1.5179, "step": 19563 }, { "epoch": 1.5332288401253917, "grad_norm": 2.048433542251587, "learning_rate": 6.8099626665529684e-06, "loss": 0.6378, "step": 19564 }, { "epoch": 1.533307210031348, "grad_norm": 3.484276533126831, "learning_rate": 6.809675213860674e-06, "loss": 1.1894, "step": 19565 }, { "epoch": 1.533385579937304, "grad_norm": 2.3113062381744385, "learning_rate": 6.809387754285306e-06, "loss": 1.4752, "step": 19566 }, { "epoch": 1.5334639498432603, "grad_norm": 4.07543420791626, "learning_rate": 6.809100287827956e-06, "loss": 1.2356, "step": 19567 }, { "epoch": 1.5335423197492162, "grad_norm": 2.7546939849853516, "learning_rate": 6.8088128144897155e-06, "loss": 1.6012, "step": 19568 }, { "epoch": 1.5336206896551725, "grad_norm": 4.445809841156006, "learning_rate": 6.808525334271683e-06, "loss": 1.0959, "step": 19569 }, { "epoch": 1.5336990595611284, "grad_norm": 3.359405517578125, "learning_rate": 6.808237847174948e-06, "loss": 1.4261, "step": 19570 }, { "epoch": 1.5337774294670847, "grad_norm": 2.4480702877044678, "learning_rate": 6.807950353200604e-06, "loss": 1.1486, "step": 19571 }, { "epoch": 1.5338557993730406, "grad_norm": 1.9941966533660889, "learning_rate": 6.807662852349748e-06, "loss": 1.4029, "step": 19572 }, { "epoch": 1.533934169278997, "grad_norm": 11.231508255004883, "learning_rate": 6.807375344623469e-06, "loss": 1.6145, "step": 19573 }, { "epoch": 1.5340125391849528, "grad_norm": 5.202138900756836, "learning_rate": 6.807087830022863e-06, "loss": 1.5014, "step": 19574 }, { "epoch": 1.5340909090909092, "grad_norm": 9.125970840454102, "learning_rate": 6.806800308549022e-06, "loss": 1.113, "step": 19575 }, { "epoch": 1.534169278996865, "grad_norm": 2.558129072189331, "learning_rate": 6.806512780203042e-06, "loss": 0.8645, "step": 19576 }, { "epoch": 1.5342476489028214, "grad_norm": 3.5165445804595947, "learning_rate": 6.806225244986016e-06, "loss": 1.5057, "step": 19577 }, { "epoch": 1.5343260188087773, "grad_norm": 3.924978256225586, "learning_rate": 6.805937702899035e-06, "loss": 2.0008, "step": 19578 }, { "epoch": 1.5344043887147336, "grad_norm": 2.534363269805908, "learning_rate": 6.805650153943196e-06, "loss": 1.6981, "step": 19579 }, { "epoch": 1.5344827586206895, "grad_norm": 4.555238723754883, "learning_rate": 6.805362598119589e-06, "loss": 1.5657, "step": 19580 }, { "epoch": 1.5345611285266458, "grad_norm": 2.6942026615142822, "learning_rate": 6.805075035429312e-06, "loss": 1.3022, "step": 19581 }, { "epoch": 1.5346394984326017, "grad_norm": 5.192937850952148, "learning_rate": 6.8047874658734555e-06, "loss": 1.2206, "step": 19582 }, { "epoch": 1.534717868338558, "grad_norm": 2.6571457386016846, "learning_rate": 6.804499889453117e-06, "loss": 1.8014, "step": 19583 }, { "epoch": 1.534796238244514, "grad_norm": 5.409742832183838, "learning_rate": 6.804212306169386e-06, "loss": 1.4793, "step": 19584 }, { "epoch": 1.5348746081504703, "grad_norm": 2.5010855197906494, "learning_rate": 6.803924716023358e-06, "loss": 1.2632, "step": 19585 }, { "epoch": 1.5349529780564264, "grad_norm": 2.803394317626953, "learning_rate": 6.803637119016127e-06, "loss": 1.6057, "step": 19586 }, { "epoch": 1.5350313479623825, "grad_norm": 2.860895872116089, "learning_rate": 6.803349515148787e-06, "loss": 1.2947, "step": 19587 }, { "epoch": 1.5351097178683386, "grad_norm": 6.664331912994385, "learning_rate": 6.803061904422432e-06, "loss": 1.0826, "step": 19588 }, { "epoch": 1.5351880877742947, "grad_norm": 3.291919231414795, "learning_rate": 6.802774286838154e-06, "loss": 1.1762, "step": 19589 }, { "epoch": 1.5352664576802508, "grad_norm": 2.2551896572113037, "learning_rate": 6.802486662397051e-06, "loss": 1.3161, "step": 19590 }, { "epoch": 1.535344827586207, "grad_norm": 2.210698366165161, "learning_rate": 6.802199031100213e-06, "loss": 1.1707, "step": 19591 }, { "epoch": 1.535423197492163, "grad_norm": 3.6216039657592773, "learning_rate": 6.801911392948736e-06, "loss": 0.9293, "step": 19592 }, { "epoch": 1.5355015673981192, "grad_norm": 4.301240921020508, "learning_rate": 6.801623747943713e-06, "loss": 1.4774, "step": 19593 }, { "epoch": 1.5355799373040753, "grad_norm": 2.3172543048858643, "learning_rate": 6.80133609608624e-06, "loss": 1.3992, "step": 19594 }, { "epoch": 1.5356583072100314, "grad_norm": 12.434146881103516, "learning_rate": 6.801048437377407e-06, "loss": 1.3917, "step": 19595 }, { "epoch": 1.5357366771159875, "grad_norm": 4.378405570983887, "learning_rate": 6.800760771818313e-06, "loss": 1.2977, "step": 19596 }, { "epoch": 1.5358150470219436, "grad_norm": 2.6476802825927734, "learning_rate": 6.80047309941005e-06, "loss": 1.1278, "step": 19597 }, { "epoch": 1.5358934169278997, "grad_norm": 2.513420820236206, "learning_rate": 6.800185420153713e-06, "loss": 1.2801, "step": 19598 }, { "epoch": 1.5359717868338558, "grad_norm": 2.9429430961608887, "learning_rate": 6.799897734050392e-06, "loss": 1.123, "step": 19599 }, { "epoch": 1.536050156739812, "grad_norm": 2.3143818378448486, "learning_rate": 6.799610041101188e-06, "loss": 1.2851, "step": 19600 }, { "epoch": 1.536128526645768, "grad_norm": 4.413442611694336, "learning_rate": 6.799322341307191e-06, "loss": 1.0413, "step": 19601 }, { "epoch": 1.5362068965517242, "grad_norm": 2.961146116256714, "learning_rate": 6.7990346346694945e-06, "loss": 1.4747, "step": 19602 }, { "epoch": 1.5362852664576803, "grad_norm": 2.4029276371002197, "learning_rate": 6.798746921189194e-06, "loss": 1.4114, "step": 19603 }, { "epoch": 1.5363636363636364, "grad_norm": 3.236114740371704, "learning_rate": 6.798459200867387e-06, "loss": 1.2363, "step": 19604 }, { "epoch": 1.5364420062695925, "grad_norm": 1.9919475317001343, "learning_rate": 6.798171473705163e-06, "loss": 1.4854, "step": 19605 }, { "epoch": 1.5365203761755486, "grad_norm": 4.382026195526123, "learning_rate": 6.797883739703617e-06, "loss": 1.4132, "step": 19606 }, { "epoch": 1.5365987460815047, "grad_norm": 3.087437629699707, "learning_rate": 6.7975959988638474e-06, "loss": 1.5709, "step": 19607 }, { "epoch": 1.5366771159874608, "grad_norm": 2.519569158554077, "learning_rate": 6.797308251186944e-06, "loss": 1.5744, "step": 19608 }, { "epoch": 1.536755485893417, "grad_norm": 3.3691599369049072, "learning_rate": 6.797020496674004e-06, "loss": 1.6114, "step": 19609 }, { "epoch": 1.536833855799373, "grad_norm": 2.5931029319763184, "learning_rate": 6.796732735326122e-06, "loss": 1.4163, "step": 19610 }, { "epoch": 1.5369122257053291, "grad_norm": 3.724353075027466, "learning_rate": 6.796444967144389e-06, "loss": 1.7175, "step": 19611 }, { "epoch": 1.5369905956112853, "grad_norm": 3.458712577819824, "learning_rate": 6.796157192129904e-06, "loss": 1.7394, "step": 19612 }, { "epoch": 1.5370689655172414, "grad_norm": 2.8729050159454346, "learning_rate": 6.795869410283759e-06, "loss": 1.1264, "step": 19613 }, { "epoch": 1.5371473354231975, "grad_norm": 2.5065419673919678, "learning_rate": 6.795581621607051e-06, "loss": 0.9602, "step": 19614 }, { "epoch": 1.5372257053291536, "grad_norm": 3.8284292221069336, "learning_rate": 6.795293826100868e-06, "loss": 1.1617, "step": 19615 }, { "epoch": 1.5373040752351097, "grad_norm": 2.871647357940674, "learning_rate": 6.795006023766314e-06, "loss": 0.9912, "step": 19616 }, { "epoch": 1.5373824451410658, "grad_norm": 3.6321310997009277, "learning_rate": 6.794718214604476e-06, "loss": 1.3945, "step": 19617 }, { "epoch": 1.537460815047022, "grad_norm": 2.7250165939331055, "learning_rate": 6.794430398616454e-06, "loss": 1.4901, "step": 19618 }, { "epoch": 1.537539184952978, "grad_norm": 3.452343225479126, "learning_rate": 6.7941425758033394e-06, "loss": 1.4288, "step": 19619 }, { "epoch": 1.5376175548589341, "grad_norm": 2.0635809898376465, "learning_rate": 6.793854746166229e-06, "loss": 1.0964, "step": 19620 }, { "epoch": 1.5376959247648903, "grad_norm": 2.5636982917785645, "learning_rate": 6.793566909706215e-06, "loss": 1.3253, "step": 19621 }, { "epoch": 1.5377742946708464, "grad_norm": 4.017321586608887, "learning_rate": 6.793279066424394e-06, "loss": 1.4137, "step": 19622 }, { "epoch": 1.5378526645768025, "grad_norm": 6.243327617645264, "learning_rate": 6.792991216321862e-06, "loss": 1.754, "step": 19623 }, { "epoch": 1.5379310344827586, "grad_norm": 2.2740249633789062, "learning_rate": 6.792703359399711e-06, "loss": 1.4811, "step": 19624 }, { "epoch": 1.5380094043887147, "grad_norm": 3.1805946826934814, "learning_rate": 6.792415495659037e-06, "loss": 1.8697, "step": 19625 }, { "epoch": 1.5380877742946708, "grad_norm": 3.6690428256988525, "learning_rate": 6.792127625100937e-06, "loss": 1.6147, "step": 19626 }, { "epoch": 1.538166144200627, "grad_norm": 3.8868625164031982, "learning_rate": 6.7918397477265e-06, "loss": 1.2519, "step": 19627 }, { "epoch": 1.538244514106583, "grad_norm": 3.3569529056549072, "learning_rate": 6.791551863536829e-06, "loss": 1.1818, "step": 19628 }, { "epoch": 1.5383228840125391, "grad_norm": 4.138058662414551, "learning_rate": 6.791263972533014e-06, "loss": 1.4086, "step": 19629 }, { "epoch": 1.5384012539184952, "grad_norm": 4.36472225189209, "learning_rate": 6.790976074716151e-06, "loss": 1.8927, "step": 19630 }, { "epoch": 1.5384796238244514, "grad_norm": 3.536313533782959, "learning_rate": 6.790688170087334e-06, "loss": 1.4819, "step": 19631 }, { "epoch": 1.5385579937304075, "grad_norm": 2.8803884983062744, "learning_rate": 6.790400258647659e-06, "loss": 1.317, "step": 19632 }, { "epoch": 1.5386363636363636, "grad_norm": 4.060188293457031, "learning_rate": 6.790112340398223e-06, "loss": 1.8514, "step": 19633 }, { "epoch": 1.53871473354232, "grad_norm": 7.209377765655518, "learning_rate": 6.789824415340116e-06, "loss": 0.9412, "step": 19634 }, { "epoch": 1.5387931034482758, "grad_norm": 1.6463807821273804, "learning_rate": 6.789536483474438e-06, "loss": 1.1129, "step": 19635 }, { "epoch": 1.5388714733542321, "grad_norm": 2.4424421787261963, "learning_rate": 6.789248544802282e-06, "loss": 0.965, "step": 19636 }, { "epoch": 1.538949843260188, "grad_norm": 3.0236425399780273, "learning_rate": 6.788960599324744e-06, "loss": 1.7118, "step": 19637 }, { "epoch": 1.5390282131661444, "grad_norm": 6.0972514152526855, "learning_rate": 6.788672647042918e-06, "loss": 1.2191, "step": 19638 }, { "epoch": 1.5391065830721002, "grad_norm": 2.8461966514587402, "learning_rate": 6.7883846879579e-06, "loss": 1.4369, "step": 19639 }, { "epoch": 1.5391849529780566, "grad_norm": 3.9270524978637695, "learning_rate": 6.788096722070785e-06, "loss": 1.8285, "step": 19640 }, { "epoch": 1.5392633228840125, "grad_norm": 7.2085371017456055, "learning_rate": 6.787808749382669e-06, "loss": 0.9916, "step": 19641 }, { "epoch": 1.5393416927899688, "grad_norm": 4.174142837524414, "learning_rate": 6.787520769894649e-06, "loss": 1.4922, "step": 19642 }, { "epoch": 1.5394200626959247, "grad_norm": 2.2905027866363525, "learning_rate": 6.787232783607815e-06, "loss": 1.4684, "step": 19643 }, { "epoch": 1.539498432601881, "grad_norm": 2.1884820461273193, "learning_rate": 6.786944790523267e-06, "loss": 1.3184, "step": 19644 }, { "epoch": 1.539576802507837, "grad_norm": 2.574381113052368, "learning_rate": 6.7866567906420975e-06, "loss": 1.2592, "step": 19645 }, { "epoch": 1.5396551724137932, "grad_norm": 4.763769626617432, "learning_rate": 6.786368783965405e-06, "loss": 0.9936, "step": 19646 }, { "epoch": 1.5397335423197491, "grad_norm": 2.417576551437378, "learning_rate": 6.7860807704942815e-06, "loss": 1.4727, "step": 19647 }, { "epoch": 1.5398119122257055, "grad_norm": 5.254388332366943, "learning_rate": 6.785792750229825e-06, "loss": 1.8252, "step": 19648 }, { "epoch": 1.5398902821316613, "grad_norm": 5.86484432220459, "learning_rate": 6.785504723173128e-06, "loss": 1.8247, "step": 19649 }, { "epoch": 1.5399686520376177, "grad_norm": 3.1920602321624756, "learning_rate": 6.785216689325292e-06, "loss": 1.3205, "step": 19650 }, { "epoch": 1.5400470219435736, "grad_norm": 5.020763874053955, "learning_rate": 6.784928648687405e-06, "loss": 1.3482, "step": 19651 }, { "epoch": 1.54012539184953, "grad_norm": 5.452658176422119, "learning_rate": 6.784640601260571e-06, "loss": 1.2361, "step": 19652 }, { "epoch": 1.5402037617554858, "grad_norm": 3.133650779724121, "learning_rate": 6.784352547045876e-06, "loss": 1.3988, "step": 19653 }, { "epoch": 1.5402821316614421, "grad_norm": 4.115521430969238, "learning_rate": 6.784064486044423e-06, "loss": 1.5182, "step": 19654 }, { "epoch": 1.540360501567398, "grad_norm": 3.701643228530884, "learning_rate": 6.7837764182573015e-06, "loss": 1.4272, "step": 19655 }, { "epoch": 1.5404388714733543, "grad_norm": 2.4001123905181885, "learning_rate": 6.783488343685616e-06, "loss": 1.5098, "step": 19656 }, { "epoch": 1.5405172413793102, "grad_norm": 8.310111045837402, "learning_rate": 6.783200262330454e-06, "loss": 1.1583, "step": 19657 }, { "epoch": 1.5405956112852666, "grad_norm": 4.023614406585693, "learning_rate": 6.782912174192914e-06, "loss": 1.2728, "step": 19658 }, { "epoch": 1.5406739811912225, "grad_norm": 3.1167402267456055, "learning_rate": 6.782624079274091e-06, "loss": 1.659, "step": 19659 }, { "epoch": 1.5407523510971788, "grad_norm": 2.0087201595306396, "learning_rate": 6.782335977575084e-06, "loss": 1.5133, "step": 19660 }, { "epoch": 1.5408307210031347, "grad_norm": 4.05758810043335, "learning_rate": 6.782047869096986e-06, "loss": 1.2138, "step": 19661 }, { "epoch": 1.540909090909091, "grad_norm": 2.8833422660827637, "learning_rate": 6.781759753840892e-06, "loss": 1.274, "step": 19662 }, { "epoch": 1.540987460815047, "grad_norm": 3.4813756942749023, "learning_rate": 6.7814716318078995e-06, "loss": 1.3323, "step": 19663 }, { "epoch": 1.5410658307210032, "grad_norm": 2.2885079383850098, "learning_rate": 6.781183502999105e-06, "loss": 1.1582, "step": 19664 }, { "epoch": 1.5411442006269591, "grad_norm": 3.0073320865631104, "learning_rate": 6.780895367415602e-06, "loss": 1.246, "step": 19665 }, { "epoch": 1.5412225705329154, "grad_norm": 2.1960902214050293, "learning_rate": 6.780607225058488e-06, "loss": 1.5019, "step": 19666 }, { "epoch": 1.5413009404388713, "grad_norm": 4.9937849044799805, "learning_rate": 6.780319075928858e-06, "loss": 1.5675, "step": 19667 }, { "epoch": 1.5413793103448277, "grad_norm": 3.408682346343994, "learning_rate": 6.7800309200278104e-06, "loss": 1.2079, "step": 19668 }, { "epoch": 1.5414576802507836, "grad_norm": 3.007707118988037, "learning_rate": 6.779742757356438e-06, "loss": 1.4893, "step": 19669 }, { "epoch": 1.54153605015674, "grad_norm": 2.082869529724121, "learning_rate": 6.779454587915839e-06, "loss": 1.1764, "step": 19670 }, { "epoch": 1.5416144200626958, "grad_norm": 2.8824193477630615, "learning_rate": 6.779166411707108e-06, "loss": 1.2767, "step": 19671 }, { "epoch": 1.541692789968652, "grad_norm": 3.405123710632324, "learning_rate": 6.778878228731343e-06, "loss": 1.6376, "step": 19672 }, { "epoch": 1.541771159874608, "grad_norm": 3.0016250610351562, "learning_rate": 6.778590038989639e-06, "loss": 1.1603, "step": 19673 }, { "epoch": 1.5418495297805643, "grad_norm": 3.638305902481079, "learning_rate": 6.7783018424830905e-06, "loss": 1.2455, "step": 19674 }, { "epoch": 1.5419278996865202, "grad_norm": 3.1802279949188232, "learning_rate": 6.778013639212796e-06, "loss": 1.3063, "step": 19675 }, { "epoch": 1.5420062695924766, "grad_norm": 2.6171796321868896, "learning_rate": 6.7777254291798515e-06, "loss": 1.5365, "step": 19676 }, { "epoch": 1.5420846394984324, "grad_norm": 8.082660675048828, "learning_rate": 6.777437212385351e-06, "loss": 2.5266, "step": 19677 }, { "epoch": 1.5421630094043888, "grad_norm": 2.997347831726074, "learning_rate": 6.777148988830393e-06, "loss": 1.2825, "step": 19678 }, { "epoch": 1.5422413793103447, "grad_norm": 4.288064956665039, "learning_rate": 6.776860758516073e-06, "loss": 1.566, "step": 19679 }, { "epoch": 1.542319749216301, "grad_norm": 3.057318925857544, "learning_rate": 6.77657252144349e-06, "loss": 1.0304, "step": 19680 }, { "epoch": 1.542398119122257, "grad_norm": 4.973690986633301, "learning_rate": 6.776284277613732e-06, "loss": 1.1235, "step": 19681 }, { "epoch": 1.5424764890282132, "grad_norm": 1.9343143701553345, "learning_rate": 6.7759960270279064e-06, "loss": 1.3388, "step": 19682 }, { "epoch": 1.5425548589341693, "grad_norm": 2.42138409614563, "learning_rate": 6.775707769687101e-06, "loss": 1.5222, "step": 19683 }, { "epoch": 1.5426332288401254, "grad_norm": 5.01594877243042, "learning_rate": 6.775419505592417e-06, "loss": 1.5015, "step": 19684 }, { "epoch": 1.5427115987460815, "grad_norm": 10.270724296569824, "learning_rate": 6.775131234744948e-06, "loss": 0.6367, "step": 19685 }, { "epoch": 1.5427899686520377, "grad_norm": 2.7596194744110107, "learning_rate": 6.774842957145792e-06, "loss": 1.895, "step": 19686 }, { "epoch": 1.5428683385579938, "grad_norm": 2.0049986839294434, "learning_rate": 6.774554672796045e-06, "loss": 1.3052, "step": 19687 }, { "epoch": 1.5429467084639499, "grad_norm": 2.6037490367889404, "learning_rate": 6.774266381696803e-06, "loss": 1.3442, "step": 19688 }, { "epoch": 1.543025078369906, "grad_norm": 5.913359642028809, "learning_rate": 6.773978083849164e-06, "loss": 1.5185, "step": 19689 }, { "epoch": 1.543103448275862, "grad_norm": 11.072441101074219, "learning_rate": 6.773689779254222e-06, "loss": 0.599, "step": 19690 }, { "epoch": 1.5431818181818182, "grad_norm": 3.1195054054260254, "learning_rate": 6.773401467913077e-06, "loss": 1.4975, "step": 19691 }, { "epoch": 1.5432601880877743, "grad_norm": 2.2523250579833984, "learning_rate": 6.7731131498268225e-06, "loss": 1.1541, "step": 19692 }, { "epoch": 1.5433385579937304, "grad_norm": 2.585740089416504, "learning_rate": 6.772824824996556e-06, "loss": 1.0914, "step": 19693 }, { "epoch": 1.5434169278996865, "grad_norm": 3.373182535171509, "learning_rate": 6.772536493423375e-06, "loss": 0.9976, "step": 19694 }, { "epoch": 1.5434952978056427, "grad_norm": 2.9882116317749023, "learning_rate": 6.772248155108377e-06, "loss": 1.3833, "step": 19695 }, { "epoch": 1.5435736677115988, "grad_norm": 5.2952728271484375, "learning_rate": 6.771959810052655e-06, "loss": 0.4883, "step": 19696 }, { "epoch": 1.5436520376175549, "grad_norm": 4.236733913421631, "learning_rate": 6.771671458257309e-06, "loss": 1.741, "step": 19697 }, { "epoch": 1.543730407523511, "grad_norm": 10.960768699645996, "learning_rate": 6.771383099723435e-06, "loss": 1.9952, "step": 19698 }, { "epoch": 1.543808777429467, "grad_norm": 2.80887508392334, "learning_rate": 6.771094734452128e-06, "loss": 1.3349, "step": 19699 }, { "epoch": 1.5438871473354232, "grad_norm": 4.046226501464844, "learning_rate": 6.770806362444489e-06, "loss": 1.5774, "step": 19700 }, { "epoch": 1.5439655172413793, "grad_norm": 5.98681640625, "learning_rate": 6.77051798370161e-06, "loss": 1.5278, "step": 19701 }, { "epoch": 1.5440438871473354, "grad_norm": 5.250062942504883, "learning_rate": 6.77022959822459e-06, "loss": 1.5681, "step": 19702 }, { "epoch": 1.5441222570532915, "grad_norm": 3.933408737182617, "learning_rate": 6.769941206014527e-06, "loss": 1.3757, "step": 19703 }, { "epoch": 1.5442006269592476, "grad_norm": 2.285183906555176, "learning_rate": 6.7696528070725155e-06, "loss": 1.1013, "step": 19704 }, { "epoch": 1.5442789968652038, "grad_norm": 4.184068202972412, "learning_rate": 6.769364401399655e-06, "loss": 1.4345, "step": 19705 }, { "epoch": 1.5443573667711599, "grad_norm": 3.8703629970550537, "learning_rate": 6.7690759889970415e-06, "loss": 1.4849, "step": 19706 }, { "epoch": 1.544435736677116, "grad_norm": 6.881159782409668, "learning_rate": 6.768787569865769e-06, "loss": 1.1212, "step": 19707 }, { "epoch": 1.544514106583072, "grad_norm": 2.3515305519104004, "learning_rate": 6.7684991440069405e-06, "loss": 1.4725, "step": 19708 }, { "epoch": 1.5445924764890282, "grad_norm": 3.087955951690674, "learning_rate": 6.768210711421646e-06, "loss": 1.1952, "step": 19709 }, { "epoch": 1.5446708463949843, "grad_norm": 5.246851921081543, "learning_rate": 6.76792227211099e-06, "loss": 1.2176, "step": 19710 }, { "epoch": 1.5447492163009404, "grad_norm": 2.5023560523986816, "learning_rate": 6.7676338260760635e-06, "loss": 1.5591, "step": 19711 }, { "epoch": 1.5448275862068965, "grad_norm": 4.027304649353027, "learning_rate": 6.767345373317965e-06, "loss": 1.2926, "step": 19712 }, { "epoch": 1.5449059561128526, "grad_norm": 6.440156936645508, "learning_rate": 6.767056913837794e-06, "loss": 1.3843, "step": 19713 }, { "epoch": 1.5449843260188088, "grad_norm": 3.1252846717834473, "learning_rate": 6.766768447636645e-06, "loss": 1.1961, "step": 19714 }, { "epoch": 1.5450626959247649, "grad_norm": 2.4560739994049072, "learning_rate": 6.766479974715618e-06, "loss": 1.2705, "step": 19715 }, { "epoch": 1.545141065830721, "grad_norm": 3.089015245437622, "learning_rate": 6.766191495075807e-06, "loss": 1.6332, "step": 19716 }, { "epoch": 1.545219435736677, "grad_norm": 5.574956893920898, "learning_rate": 6.765903008718312e-06, "loss": 1.7641, "step": 19717 }, { "epoch": 1.5452978056426332, "grad_norm": 2.1352741718292236, "learning_rate": 6.7656145156442265e-06, "loss": 1.3562, "step": 19718 }, { "epoch": 1.5453761755485893, "grad_norm": 2.881600856781006, "learning_rate": 6.765326015854653e-06, "loss": 1.2035, "step": 19719 }, { "epoch": 1.5454545454545454, "grad_norm": 6.554505825042725, "learning_rate": 6.765037509350685e-06, "loss": 1.5287, "step": 19720 }, { "epoch": 1.5455329153605015, "grad_norm": 2.525810956954956, "learning_rate": 6.764748996133421e-06, "loss": 1.7552, "step": 19721 }, { "epoch": 1.5456112852664576, "grad_norm": 2.9712555408477783, "learning_rate": 6.7644604762039575e-06, "loss": 1.5879, "step": 19722 }, { "epoch": 1.5456896551724137, "grad_norm": 3.5259511470794678, "learning_rate": 6.764171949563394e-06, "loss": 1.5194, "step": 19723 }, { "epoch": 1.5457680250783699, "grad_norm": 6.182732582092285, "learning_rate": 6.763883416212825e-06, "loss": 1.0788, "step": 19724 }, { "epoch": 1.545846394984326, "grad_norm": 3.0918643474578857, "learning_rate": 6.76359487615335e-06, "loss": 1.1729, "step": 19725 }, { "epoch": 1.545924764890282, "grad_norm": 2.2275638580322266, "learning_rate": 6.763306329386068e-06, "loss": 1.1587, "step": 19726 }, { "epoch": 1.5460031347962384, "grad_norm": 2.436105728149414, "learning_rate": 6.763017775912072e-06, "loss": 1.0512, "step": 19727 }, { "epoch": 1.5460815047021943, "grad_norm": 3.6587929725646973, "learning_rate": 6.762729215732462e-06, "loss": 1.6097, "step": 19728 }, { "epoch": 1.5461598746081506, "grad_norm": 6.144601821899414, "learning_rate": 6.7624406488483364e-06, "loss": 1.3039, "step": 19729 }, { "epoch": 1.5462382445141065, "grad_norm": 3.4738645553588867, "learning_rate": 6.7621520752607906e-06, "loss": 1.1306, "step": 19730 }, { "epoch": 1.5463166144200629, "grad_norm": 2.933295726776123, "learning_rate": 6.7618634949709246e-06, "loss": 1.3843, "step": 19731 }, { "epoch": 1.5463949843260187, "grad_norm": 2.1220250129699707, "learning_rate": 6.7615749079798345e-06, "loss": 1.5068, "step": 19732 }, { "epoch": 1.546473354231975, "grad_norm": 2.180162191390991, "learning_rate": 6.761286314288617e-06, "loss": 1.37, "step": 19733 }, { "epoch": 1.546551724137931, "grad_norm": 2.7094318866729736, "learning_rate": 6.760997713898374e-06, "loss": 1.2502, "step": 19734 }, { "epoch": 1.5466300940438873, "grad_norm": 2.3264260292053223, "learning_rate": 6.760709106810197e-06, "loss": 1.451, "step": 19735 }, { "epoch": 1.5467084639498432, "grad_norm": 3.6602392196655273, "learning_rate": 6.760420493025189e-06, "loss": 1.1347, "step": 19736 }, { "epoch": 1.5467868338557995, "grad_norm": 2.3094658851623535, "learning_rate": 6.760131872544444e-06, "loss": 1.4411, "step": 19737 }, { "epoch": 1.5468652037617554, "grad_norm": 7.761471271514893, "learning_rate": 6.759843245369064e-06, "loss": 1.4436, "step": 19738 }, { "epoch": 1.5469435736677117, "grad_norm": 3.0973358154296875, "learning_rate": 6.759554611500142e-06, "loss": 1.5174, "step": 19739 }, { "epoch": 1.5470219435736676, "grad_norm": 1.9829665422439575, "learning_rate": 6.759265970938778e-06, "loss": 1.4469, "step": 19740 }, { "epoch": 1.547100313479624, "grad_norm": 3.261164665222168, "learning_rate": 6.758977323686072e-06, "loss": 1.2538, "step": 19741 }, { "epoch": 1.5471786833855798, "grad_norm": 2.459167957305908, "learning_rate": 6.758688669743118e-06, "loss": 1.4119, "step": 19742 }, { "epoch": 1.5472570532915362, "grad_norm": 2.832700729370117, "learning_rate": 6.758400009111015e-06, "loss": 1.5098, "step": 19743 }, { "epoch": 1.547335423197492, "grad_norm": 3.398859977722168, "learning_rate": 6.758111341790863e-06, "loss": 1.7601, "step": 19744 }, { "epoch": 1.5474137931034484, "grad_norm": 2.248649835586548, "learning_rate": 6.757822667783758e-06, "loss": 0.9826, "step": 19745 }, { "epoch": 1.5474921630094043, "grad_norm": 2.367119073867798, "learning_rate": 6.757533987090798e-06, "loss": 0.9197, "step": 19746 }, { "epoch": 1.5475705329153606, "grad_norm": 3.6281697750091553, "learning_rate": 6.757245299713082e-06, "loss": 1.633, "step": 19747 }, { "epoch": 1.5476489028213165, "grad_norm": 3.639554500579834, "learning_rate": 6.756956605651708e-06, "loss": 1.3578, "step": 19748 }, { "epoch": 1.5477272727272728, "grad_norm": 2.100397825241089, "learning_rate": 6.7566679049077725e-06, "loss": 1.4906, "step": 19749 }, { "epoch": 1.5478056426332287, "grad_norm": 3.508875846862793, "learning_rate": 6.756379197482374e-06, "loss": 1.4877, "step": 19750 }, { "epoch": 1.547884012539185, "grad_norm": 5.7700934410095215, "learning_rate": 6.756090483376612e-06, "loss": 1.9082, "step": 19751 }, { "epoch": 1.547962382445141, "grad_norm": 4.960399627685547, "learning_rate": 6.755801762591583e-06, "loss": 1.3125, "step": 19752 }, { "epoch": 1.5480407523510973, "grad_norm": 3.6724791526794434, "learning_rate": 6.755513035128388e-06, "loss": 1.2816, "step": 19753 }, { "epoch": 1.5481191222570532, "grad_norm": 11.388436317443848, "learning_rate": 6.755224300988121e-06, "loss": 1.0464, "step": 19754 }, { "epoch": 1.5481974921630095, "grad_norm": 4.4003167152404785, "learning_rate": 6.754935560171884e-06, "loss": 1.3734, "step": 19755 }, { "epoch": 1.5482758620689654, "grad_norm": 2.568641424179077, "learning_rate": 6.754646812680772e-06, "loss": 0.9713, "step": 19756 }, { "epoch": 1.5483542319749217, "grad_norm": 3.3354103565216064, "learning_rate": 6.754358058515886e-06, "loss": 1.4538, "step": 19757 }, { "epoch": 1.5484326018808776, "grad_norm": 3.8587911128997803, "learning_rate": 6.754069297678322e-06, "loss": 1.2999, "step": 19758 }, { "epoch": 1.548510971786834, "grad_norm": 3.517578363418579, "learning_rate": 6.7537805301691795e-06, "loss": 0.9105, "step": 19759 }, { "epoch": 1.5485893416927898, "grad_norm": 3.7853035926818848, "learning_rate": 6.753491755989556e-06, "loss": 0.853, "step": 19760 }, { "epoch": 1.5486677115987462, "grad_norm": 2.157139539718628, "learning_rate": 6.753202975140552e-06, "loss": 1.1218, "step": 19761 }, { "epoch": 1.548746081504702, "grad_norm": 4.729036808013916, "learning_rate": 6.752914187623265e-06, "loss": 1.7365, "step": 19762 }, { "epoch": 1.5488244514106584, "grad_norm": 2.6870992183685303, "learning_rate": 6.75262539343879e-06, "loss": 1.3463, "step": 19763 }, { "epoch": 1.5489028213166143, "grad_norm": 2.480473756790161, "learning_rate": 6.752336592588231e-06, "loss": 1.1888, "step": 19764 }, { "epoch": 1.5489811912225706, "grad_norm": 6.045192241668701, "learning_rate": 6.752047785072682e-06, "loss": 1.7248, "step": 19765 }, { "epoch": 1.5490595611285265, "grad_norm": 2.887904167175293, "learning_rate": 6.751758970893243e-06, "loss": 1.2621, "step": 19766 }, { "epoch": 1.5491379310344828, "grad_norm": 2.946415424346924, "learning_rate": 6.751470150051013e-06, "loss": 1.5044, "step": 19767 }, { "epoch": 1.5492163009404387, "grad_norm": 4.419124126434326, "learning_rate": 6.7511813225470914e-06, "loss": 2.02, "step": 19768 }, { "epoch": 1.549294670846395, "grad_norm": 6.7040019035339355, "learning_rate": 6.750892488382574e-06, "loss": 1.1151, "step": 19769 }, { "epoch": 1.549373040752351, "grad_norm": 4.993077278137207, "learning_rate": 6.750603647558561e-06, "loss": 0.9641, "step": 19770 }, { "epoch": 1.5494514106583073, "grad_norm": 3.654115915298462, "learning_rate": 6.750314800076152e-06, "loss": 1.8899, "step": 19771 }, { "epoch": 1.5495297805642632, "grad_norm": 2.5247230529785156, "learning_rate": 6.7500259459364425e-06, "loss": 1.2551, "step": 19772 }, { "epoch": 1.5496081504702195, "grad_norm": 4.090636253356934, "learning_rate": 6.749737085140535e-06, "loss": 1.4835, "step": 19773 }, { "epoch": 1.5496865203761756, "grad_norm": 5.245147228240967, "learning_rate": 6.749448217689527e-06, "loss": 1.2809, "step": 19774 }, { "epoch": 1.5497648902821317, "grad_norm": 1.868345022201538, "learning_rate": 6.749159343584515e-06, "loss": 1.2876, "step": 19775 }, { "epoch": 1.5498432601880878, "grad_norm": 2.7926363945007324, "learning_rate": 6.748870462826599e-06, "loss": 1.2525, "step": 19776 }, { "epoch": 1.549921630094044, "grad_norm": 4.353845596313477, "learning_rate": 6.748581575416879e-06, "loss": 1.2892, "step": 19777 }, { "epoch": 1.55, "grad_norm": 2.058147430419922, "learning_rate": 6.748292681356452e-06, "loss": 1.3146, "step": 19778 }, { "epoch": 1.5500783699059562, "grad_norm": 2.524081230163574, "learning_rate": 6.7480037806464185e-06, "loss": 1.2694, "step": 19779 }, { "epoch": 1.5501567398119123, "grad_norm": 3.4846725463867188, "learning_rate": 6.747714873287876e-06, "loss": 1.5334, "step": 19780 }, { "epoch": 1.5502351097178684, "grad_norm": 4.519648551940918, "learning_rate": 6.7474259592819246e-06, "loss": 0.6576, "step": 19781 }, { "epoch": 1.5503134796238245, "grad_norm": 5.178365707397461, "learning_rate": 6.747137038629661e-06, "loss": 1.3287, "step": 19782 }, { "epoch": 1.5503918495297806, "grad_norm": 2.6504616737365723, "learning_rate": 6.746848111332186e-06, "loss": 1.3697, "step": 19783 }, { "epoch": 1.5504702194357367, "grad_norm": 3.302349328994751, "learning_rate": 6.746559177390598e-06, "loss": 1.3374, "step": 19784 }, { "epoch": 1.5505485893416928, "grad_norm": 4.040644645690918, "learning_rate": 6.746270236805996e-06, "loss": 1.7338, "step": 19785 }, { "epoch": 1.550626959247649, "grad_norm": 4.639869213104248, "learning_rate": 6.7459812895794795e-06, "loss": 1.018, "step": 19786 }, { "epoch": 1.550705329153605, "grad_norm": 2.985692024230957, "learning_rate": 6.745692335712146e-06, "loss": 1.2994, "step": 19787 }, { "epoch": 1.5507836990595611, "grad_norm": 2.910421371459961, "learning_rate": 6.745403375205097e-06, "loss": 0.8662, "step": 19788 }, { "epoch": 1.5508620689655173, "grad_norm": 2.680497884750366, "learning_rate": 6.745114408059428e-06, "loss": 1.1863, "step": 19789 }, { "epoch": 1.5509404388714734, "grad_norm": 4.927672863006592, "learning_rate": 6.744825434276243e-06, "loss": 1.3511, "step": 19790 }, { "epoch": 1.5510188087774295, "grad_norm": 2.881345510482788, "learning_rate": 6.744536453856635e-06, "loss": 1.3223, "step": 19791 }, { "epoch": 1.5510971786833856, "grad_norm": 9.273438453674316, "learning_rate": 6.744247466801709e-06, "loss": 1.5219, "step": 19792 }, { "epoch": 1.5511755485893417, "grad_norm": 3.8435490131378174, "learning_rate": 6.743958473112559e-06, "loss": 1.2889, "step": 19793 }, { "epoch": 1.5512539184952978, "grad_norm": 2.700714111328125, "learning_rate": 6.743669472790288e-06, "loss": 1.06, "step": 19794 }, { "epoch": 1.551332288401254, "grad_norm": 3.321209192276001, "learning_rate": 6.7433804658359955e-06, "loss": 1.183, "step": 19795 }, { "epoch": 1.55141065830721, "grad_norm": 2.210597276687622, "learning_rate": 6.7430914522507774e-06, "loss": 1.2995, "step": 19796 }, { "epoch": 1.5514890282131661, "grad_norm": 3.5962939262390137, "learning_rate": 6.742802432035736e-06, "loss": 1.1715, "step": 19797 }, { "epoch": 1.5515673981191223, "grad_norm": 2.496917724609375, "learning_rate": 6.742513405191967e-06, "loss": 1.6681, "step": 19798 }, { "epoch": 1.5516457680250784, "grad_norm": 10.17677116394043, "learning_rate": 6.7422243717205735e-06, "loss": 1.2299, "step": 19799 }, { "epoch": 1.5517241379310345, "grad_norm": 2.480036497116089, "learning_rate": 6.741935331622654e-06, "loss": 1.1638, "step": 19800 }, { "epoch": 1.5518025078369906, "grad_norm": 2.230311155319214, "learning_rate": 6.741646284899306e-06, "loss": 1.2049, "step": 19801 }, { "epoch": 1.5518808777429467, "grad_norm": 2.185467481613159, "learning_rate": 6.741357231551631e-06, "loss": 1.0953, "step": 19802 }, { "epoch": 1.5519592476489028, "grad_norm": 2.713982343673706, "learning_rate": 6.741068171580727e-06, "loss": 1.4325, "step": 19803 }, { "epoch": 1.552037617554859, "grad_norm": 3.7168538570404053, "learning_rate": 6.740779104987693e-06, "loss": 1.5955, "step": 19804 }, { "epoch": 1.552115987460815, "grad_norm": 3.180109977722168, "learning_rate": 6.740490031773631e-06, "loss": 0.7675, "step": 19805 }, { "epoch": 1.5521943573667711, "grad_norm": 3.483581781387329, "learning_rate": 6.740200951939638e-06, "loss": 1.3709, "step": 19806 }, { "epoch": 1.5522727272727272, "grad_norm": 2.05146861076355, "learning_rate": 6.739911865486816e-06, "loss": 1.3019, "step": 19807 }, { "epoch": 1.5523510971786834, "grad_norm": 2.9018054008483887, "learning_rate": 6.7396227724162615e-06, "loss": 1.4299, "step": 19808 }, { "epoch": 1.5524294670846395, "grad_norm": 6.3949079513549805, "learning_rate": 6.739333672729076e-06, "loss": 1.1163, "step": 19809 }, { "epoch": 1.5525078369905956, "grad_norm": 3.6303627490997314, "learning_rate": 6.7390445664263586e-06, "loss": 1.3155, "step": 19810 }, { "epoch": 1.5525862068965517, "grad_norm": 4.633110523223877, "learning_rate": 6.738755453509207e-06, "loss": 1.6899, "step": 19811 }, { "epoch": 1.5526645768025078, "grad_norm": 8.05700969696045, "learning_rate": 6.738466333978725e-06, "loss": 1.5344, "step": 19812 }, { "epoch": 1.552742946708464, "grad_norm": 5.901303768157959, "learning_rate": 6.738177207836009e-06, "loss": 1.1812, "step": 19813 }, { "epoch": 1.55282131661442, "grad_norm": 2.152113199234009, "learning_rate": 6.737888075082159e-06, "loss": 1.2204, "step": 19814 }, { "epoch": 1.5528996865203761, "grad_norm": 3.2060954570770264, "learning_rate": 6.737598935718277e-06, "loss": 1.1816, "step": 19815 }, { "epoch": 1.5529780564263322, "grad_norm": 2.598205089569092, "learning_rate": 6.737309789745461e-06, "loss": 1.1889, "step": 19816 }, { "epoch": 1.5530564263322884, "grad_norm": 3.2225990295410156, "learning_rate": 6.737020637164809e-06, "loss": 1.649, "step": 19817 }, { "epoch": 1.5531347962382445, "grad_norm": 3.2131223678588867, "learning_rate": 6.736731477977425e-06, "loss": 1.2202, "step": 19818 }, { "epoch": 1.5532131661442006, "grad_norm": 6.027105808258057, "learning_rate": 6.736442312184405e-06, "loss": 1.7934, "step": 19819 }, { "epoch": 1.5532915360501567, "grad_norm": 5.0634284019470215, "learning_rate": 6.736153139786852e-06, "loss": 1.1825, "step": 19820 }, { "epoch": 1.5533699059561128, "grad_norm": 2.2129085063934326, "learning_rate": 6.735863960785863e-06, "loss": 1.5626, "step": 19821 }, { "epoch": 1.5534482758620691, "grad_norm": 3.9916396141052246, "learning_rate": 6.735574775182538e-06, "loss": 1.2905, "step": 19822 }, { "epoch": 1.553526645768025, "grad_norm": 2.870542287826538, "learning_rate": 6.735285582977979e-06, "loss": 1.3665, "step": 19823 }, { "epoch": 1.5536050156739813, "grad_norm": 2.940364122390747, "learning_rate": 6.734996384173287e-06, "loss": 1.1031, "step": 19824 }, { "epoch": 1.5536833855799372, "grad_norm": 6.12410831451416, "learning_rate": 6.734707178769559e-06, "loss": 1.5054, "step": 19825 }, { "epoch": 1.5537617554858936, "grad_norm": 5.636028289794922, "learning_rate": 6.734417966767893e-06, "loss": 0.9225, "step": 19826 }, { "epoch": 1.5538401253918495, "grad_norm": 1.9246745109558105, "learning_rate": 6.734128748169395e-06, "loss": 1.1758, "step": 19827 }, { "epoch": 1.5539184952978058, "grad_norm": 7.913265228271484, "learning_rate": 6.733839522975161e-06, "loss": 1.5191, "step": 19828 }, { "epoch": 1.5539968652037617, "grad_norm": 6.561695098876953, "learning_rate": 6.733550291186292e-06, "loss": 0.9169, "step": 19829 }, { "epoch": 1.554075235109718, "grad_norm": 2.807724714279175, "learning_rate": 6.733261052803887e-06, "loss": 1.6493, "step": 19830 }, { "epoch": 1.554153605015674, "grad_norm": 4.253044605255127, "learning_rate": 6.7329718078290475e-06, "loss": 0.7352, "step": 19831 }, { "epoch": 1.5542319749216302, "grad_norm": 4.303863525390625, "learning_rate": 6.732682556262875e-06, "loss": 1.4537, "step": 19832 }, { "epoch": 1.5543103448275861, "grad_norm": 2.8042285442352295, "learning_rate": 6.732393298106466e-06, "loss": 1.0263, "step": 19833 }, { "epoch": 1.5543887147335425, "grad_norm": 3.3050589561462402, "learning_rate": 6.732104033360924e-06, "loss": 1.7366, "step": 19834 }, { "epoch": 1.5544670846394983, "grad_norm": 3.259629964828491, "learning_rate": 6.731814762027349e-06, "loss": 1.1661, "step": 19835 }, { "epoch": 1.5545454545454547, "grad_norm": 3.1229658126831055, "learning_rate": 6.731525484106837e-06, "loss": 1.1553, "step": 19836 }, { "epoch": 1.5546238244514106, "grad_norm": 2.688854455947876, "learning_rate": 6.731236199600495e-06, "loss": 1.3319, "step": 19837 }, { "epoch": 1.554702194357367, "grad_norm": 2.2826595306396484, "learning_rate": 6.730946908509417e-06, "loss": 1.4914, "step": 19838 }, { "epoch": 1.5547805642633228, "grad_norm": 3.2886605262756348, "learning_rate": 6.730657610834707e-06, "loss": 1.545, "step": 19839 }, { "epoch": 1.5548589341692791, "grad_norm": 3.6713345050811768, "learning_rate": 6.730368306577464e-06, "loss": 1.2519, "step": 19840 }, { "epoch": 1.554937304075235, "grad_norm": 8.673883438110352, "learning_rate": 6.730078995738789e-06, "loss": 1.5738, "step": 19841 }, { "epoch": 1.5550156739811913, "grad_norm": 3.49349045753479, "learning_rate": 6.729789678319783e-06, "loss": 1.8623, "step": 19842 }, { "epoch": 1.5550940438871472, "grad_norm": 3.2205827236175537, "learning_rate": 6.729500354321543e-06, "loss": 1.011, "step": 19843 }, { "epoch": 1.5551724137931036, "grad_norm": 2.1167924404144287, "learning_rate": 6.729211023745176e-06, "loss": 1.5198, "step": 19844 }, { "epoch": 1.5552507836990594, "grad_norm": 4.838353633880615, "learning_rate": 6.728921686591774e-06, "loss": 1.0536, "step": 19845 }, { "epoch": 1.5553291536050158, "grad_norm": 2.627906084060669, "learning_rate": 6.728632342862446e-06, "loss": 1.0164, "step": 19846 }, { "epoch": 1.5554075235109717, "grad_norm": 2.254985809326172, "learning_rate": 6.728342992558286e-06, "loss": 1.3981, "step": 19847 }, { "epoch": 1.555485893416928, "grad_norm": 2.8500423431396484, "learning_rate": 6.7280536356803975e-06, "loss": 1.154, "step": 19848 }, { "epoch": 1.5555642633228839, "grad_norm": 5.758751392364502, "learning_rate": 6.72776427222988e-06, "loss": 1.1362, "step": 19849 }, { "epoch": 1.5556426332288402, "grad_norm": 2.3065669536590576, "learning_rate": 6.727474902207836e-06, "loss": 1.5639, "step": 19850 }, { "epoch": 1.555721003134796, "grad_norm": 2.108666181564331, "learning_rate": 6.727185525615363e-06, "loss": 1.1133, "step": 19851 }, { "epoch": 1.5557993730407524, "grad_norm": 2.031771183013916, "learning_rate": 6.726896142453565e-06, "loss": 1.3117, "step": 19852 }, { "epoch": 1.5558777429467083, "grad_norm": 3.1359570026397705, "learning_rate": 6.7266067527235415e-06, "loss": 1.9126, "step": 19853 }, { "epoch": 1.5559561128526647, "grad_norm": 2.747453212738037, "learning_rate": 6.726317356426391e-06, "loss": 1.5563, "step": 19854 }, { "epoch": 1.5560344827586206, "grad_norm": 3.931555986404419, "learning_rate": 6.7260279535632165e-06, "loss": 1.3034, "step": 19855 }, { "epoch": 1.5561128526645769, "grad_norm": 2.0889999866485596, "learning_rate": 6.725738544135119e-06, "loss": 1.587, "step": 19856 }, { "epoch": 1.5561912225705328, "grad_norm": 4.034072399139404, "learning_rate": 6.725449128143198e-06, "loss": 1.1934, "step": 19857 }, { "epoch": 1.556269592476489, "grad_norm": 3.624724864959717, "learning_rate": 6.7251597055885535e-06, "loss": 1.2587, "step": 19858 }, { "epoch": 1.556347962382445, "grad_norm": 2.8050272464752197, "learning_rate": 6.7248702764722886e-06, "loss": 1.5866, "step": 19859 }, { "epoch": 1.5564263322884013, "grad_norm": 4.545240879058838, "learning_rate": 6.724580840795503e-06, "loss": 1.7914, "step": 19860 }, { "epoch": 1.5565047021943572, "grad_norm": 1.6866003274917603, "learning_rate": 6.7242913985593e-06, "loss": 1.1894, "step": 19861 }, { "epoch": 1.5565830721003135, "grad_norm": 8.009635925292969, "learning_rate": 6.724001949764774e-06, "loss": 1.3301, "step": 19862 }, { "epoch": 1.5566614420062694, "grad_norm": 2.8646128177642822, "learning_rate": 6.723712494413033e-06, "loss": 1.1525, "step": 19863 }, { "epoch": 1.5567398119122258, "grad_norm": 4.456772804260254, "learning_rate": 6.723423032505173e-06, "loss": 2.0939, "step": 19864 }, { "epoch": 1.5568181818181817, "grad_norm": 4.235828876495361, "learning_rate": 6.723133564042299e-06, "loss": 1.8219, "step": 19865 }, { "epoch": 1.556896551724138, "grad_norm": 2.149167776107788, "learning_rate": 6.722844089025509e-06, "loss": 1.3025, "step": 19866 }, { "epoch": 1.5569749216300939, "grad_norm": 3.80285382270813, "learning_rate": 6.722554607455905e-06, "loss": 1.4335, "step": 19867 }, { "epoch": 1.5570532915360502, "grad_norm": 4.192538738250732, "learning_rate": 6.7222651193345875e-06, "loss": 1.5971, "step": 19868 }, { "epoch": 1.5571316614420063, "grad_norm": 2.8850433826446533, "learning_rate": 6.72197562466266e-06, "loss": 1.273, "step": 19869 }, { "epoch": 1.5572100313479624, "grad_norm": 3.399113655090332, "learning_rate": 6.721686123441221e-06, "loss": 1.3359, "step": 19870 }, { "epoch": 1.5572884012539185, "grad_norm": 3.8314948081970215, "learning_rate": 6.721396615671371e-06, "loss": 1.5273, "step": 19871 }, { "epoch": 1.5573667711598747, "grad_norm": 3.06771183013916, "learning_rate": 6.721107101354214e-06, "loss": 0.9847, "step": 19872 }, { "epoch": 1.5574451410658308, "grad_norm": 4.875725746154785, "learning_rate": 6.720817580490847e-06, "loss": 1.428, "step": 19873 }, { "epoch": 1.5575235109717869, "grad_norm": 2.3243608474731445, "learning_rate": 6.720528053082377e-06, "loss": 1.5601, "step": 19874 }, { "epoch": 1.557601880877743, "grad_norm": 8.892292022705078, "learning_rate": 6.7202385191299e-06, "loss": 1.052, "step": 19875 }, { "epoch": 1.557680250783699, "grad_norm": 2.2048861980438232, "learning_rate": 6.719948978634521e-06, "loss": 1.6831, "step": 19876 }, { "epoch": 1.5577586206896552, "grad_norm": 2.8081820011138916, "learning_rate": 6.719659431597337e-06, "loss": 1.2138, "step": 19877 }, { "epoch": 1.5578369905956113, "grad_norm": 2.0883688926696777, "learning_rate": 6.719369878019454e-06, "loss": 1.5323, "step": 19878 }, { "epoch": 1.5579153605015674, "grad_norm": 2.5025975704193115, "learning_rate": 6.719080317901969e-06, "loss": 1.3206, "step": 19879 }, { "epoch": 1.5579937304075235, "grad_norm": 5.093723297119141, "learning_rate": 6.718790751245987e-06, "loss": 1.2873, "step": 19880 }, { "epoch": 1.5580721003134796, "grad_norm": 3.91322922706604, "learning_rate": 6.718501178052609e-06, "loss": 1.3457, "step": 19881 }, { "epoch": 1.5581504702194358, "grad_norm": 2.7829411029815674, "learning_rate": 6.718211598322933e-06, "loss": 1.3788, "step": 19882 }, { "epoch": 1.5582288401253919, "grad_norm": 3.867213010787964, "learning_rate": 6.717922012058064e-06, "loss": 1.3249, "step": 19883 }, { "epoch": 1.558307210031348, "grad_norm": 3.517167329788208, "learning_rate": 6.717632419259101e-06, "loss": 1.5186, "step": 19884 }, { "epoch": 1.558385579937304, "grad_norm": 3.9423534870147705, "learning_rate": 6.717342819927146e-06, "loss": 1.3078, "step": 19885 }, { "epoch": 1.5584639498432602, "grad_norm": 2.4793782234191895, "learning_rate": 6.7170532140633025e-06, "loss": 1.3579, "step": 19886 }, { "epoch": 1.5585423197492163, "grad_norm": 2.4629030227661133, "learning_rate": 6.71676360166867e-06, "loss": 1.2133, "step": 19887 }, { "epoch": 1.5586206896551724, "grad_norm": 2.775559663772583, "learning_rate": 6.7164739827443495e-06, "loss": 1.6194, "step": 19888 }, { "epoch": 1.5586990595611285, "grad_norm": 3.281752586364746, "learning_rate": 6.716184357291446e-06, "loss": 1.5978, "step": 19889 }, { "epoch": 1.5587774294670846, "grad_norm": 3.3344197273254395, "learning_rate": 6.7158947253110565e-06, "loss": 1.0817, "step": 19890 }, { "epoch": 1.5588557993730408, "grad_norm": 3.3943581581115723, "learning_rate": 6.7156050868042855e-06, "loss": 1.3051, "step": 19891 }, { "epoch": 1.5589341692789969, "grad_norm": 2.915447235107422, "learning_rate": 6.7153154417722335e-06, "loss": 1.3689, "step": 19892 }, { "epoch": 1.559012539184953, "grad_norm": 3.8276638984680176, "learning_rate": 6.715025790216003e-06, "loss": 1.4899, "step": 19893 }, { "epoch": 1.559090909090909, "grad_norm": 5.087416172027588, "learning_rate": 6.714736132136696e-06, "loss": 1.7219, "step": 19894 }, { "epoch": 1.5591692789968652, "grad_norm": 7.412559986114502, "learning_rate": 6.714446467535413e-06, "loss": 1.2766, "step": 19895 }, { "epoch": 1.5592476489028213, "grad_norm": 8.184821128845215, "learning_rate": 6.714156796413257e-06, "loss": 1.8014, "step": 19896 }, { "epoch": 1.5593260188087774, "grad_norm": 3.0607802867889404, "learning_rate": 6.7138671187713266e-06, "loss": 1.3623, "step": 19897 }, { "epoch": 1.5594043887147335, "grad_norm": 4.474493503570557, "learning_rate": 6.713577434610728e-06, "loss": 1.6676, "step": 19898 }, { "epoch": 1.5594827586206896, "grad_norm": 2.4231958389282227, "learning_rate": 6.713287743932559e-06, "loss": 0.9691, "step": 19899 }, { "epoch": 1.5595611285266457, "grad_norm": 3.500488758087158, "learning_rate": 6.7129980467379265e-06, "loss": 1.6521, "step": 19900 }, { "epoch": 1.5596394984326019, "grad_norm": 2.0834803581237793, "learning_rate": 6.712708343027927e-06, "loss": 1.2799, "step": 19901 }, { "epoch": 1.559717868338558, "grad_norm": 8.743485450744629, "learning_rate": 6.712418632803666e-06, "loss": 1.1771, "step": 19902 }, { "epoch": 1.559796238244514, "grad_norm": 1.9692946672439575, "learning_rate": 6.712128916066243e-06, "loss": 0.9764, "step": 19903 }, { "epoch": 1.5598746081504702, "grad_norm": 2.625056028366089, "learning_rate": 6.711839192816761e-06, "loss": 1.0849, "step": 19904 }, { "epoch": 1.5599529780564263, "grad_norm": 2.1020145416259766, "learning_rate": 6.711549463056323e-06, "loss": 1.3858, "step": 19905 }, { "epoch": 1.5600313479623824, "grad_norm": 4.624995231628418, "learning_rate": 6.711259726786029e-06, "loss": 1.0991, "step": 19906 }, { "epoch": 1.5601097178683385, "grad_norm": 2.5206079483032227, "learning_rate": 6.710969984006983e-06, "loss": 1.4708, "step": 19907 }, { "epoch": 1.5601880877742946, "grad_norm": 3.4187066555023193, "learning_rate": 6.710680234720285e-06, "loss": 1.3098, "step": 19908 }, { "epoch": 1.5602664576802507, "grad_norm": 2.932166814804077, "learning_rate": 6.710390478927038e-06, "loss": 1.5408, "step": 19909 }, { "epoch": 1.5603448275862069, "grad_norm": 3.5114002227783203, "learning_rate": 6.710100716628345e-06, "loss": 1.6327, "step": 19910 }, { "epoch": 1.560423197492163, "grad_norm": 3.806645154953003, "learning_rate": 6.709810947825306e-06, "loss": 1.773, "step": 19911 }, { "epoch": 1.560501567398119, "grad_norm": 2.3487846851348877, "learning_rate": 6.7095211725190245e-06, "loss": 1.2698, "step": 19912 }, { "epoch": 1.5605799373040752, "grad_norm": 2.4981095790863037, "learning_rate": 6.709231390710603e-06, "loss": 1.5076, "step": 19913 }, { "epoch": 1.5606583072100313, "grad_norm": 2.247864007949829, "learning_rate": 6.708941602401145e-06, "loss": 1.2055, "step": 19914 }, { "epoch": 1.5607366771159876, "grad_norm": 2.264784336090088, "learning_rate": 6.708651807591748e-06, "loss": 1.4289, "step": 19915 }, { "epoch": 1.5608150470219435, "grad_norm": 5.791467666625977, "learning_rate": 6.708362006283518e-06, "loss": 1.5848, "step": 19916 }, { "epoch": 1.5608934169278998, "grad_norm": 3.8821449279785156, "learning_rate": 6.7080721984775575e-06, "loss": 1.3518, "step": 19917 }, { "epoch": 1.5609717868338557, "grad_norm": 3.6380817890167236, "learning_rate": 6.707782384174966e-06, "loss": 1.2711, "step": 19918 }, { "epoch": 1.561050156739812, "grad_norm": 2.922433614730835, "learning_rate": 6.707492563376849e-06, "loss": 1.2699, "step": 19919 }, { "epoch": 1.561128526645768, "grad_norm": 2.612377405166626, "learning_rate": 6.707202736084307e-06, "loss": 1.2544, "step": 19920 }, { "epoch": 1.5612068965517243, "grad_norm": 2.7902143001556396, "learning_rate": 6.7069129022984424e-06, "loss": 1.4914, "step": 19921 }, { "epoch": 1.5612852664576802, "grad_norm": 4.167262554168701, "learning_rate": 6.706623062020358e-06, "loss": 1.8918, "step": 19922 }, { "epoch": 1.5613636363636365, "grad_norm": 3.7211005687713623, "learning_rate": 6.706333215251155e-06, "loss": 1.1407, "step": 19923 }, { "epoch": 1.5614420062695924, "grad_norm": 5.01893949508667, "learning_rate": 6.706043361991938e-06, "loss": 1.3978, "step": 19924 }, { "epoch": 1.5615203761755487, "grad_norm": 3.268428087234497, "learning_rate": 6.705753502243808e-06, "loss": 1.3696, "step": 19925 }, { "epoch": 1.5615987460815046, "grad_norm": 4.066439628601074, "learning_rate": 6.705463636007868e-06, "loss": 1.6332, "step": 19926 }, { "epoch": 1.561677115987461, "grad_norm": 3.0341429710388184, "learning_rate": 6.70517376328522e-06, "loss": 1.7151, "step": 19927 }, { "epoch": 1.5617554858934168, "grad_norm": 3.201557159423828, "learning_rate": 6.704883884076968e-06, "loss": 0.9865, "step": 19928 }, { "epoch": 1.5618338557993732, "grad_norm": 3.070553779602051, "learning_rate": 6.704593998384212e-06, "loss": 2.005, "step": 19929 }, { "epoch": 1.561912225705329, "grad_norm": 3.568315029144287, "learning_rate": 6.704304106208056e-06, "loss": 1.785, "step": 19930 }, { "epoch": 1.5619905956112854, "grad_norm": 3.3467962741851807, "learning_rate": 6.7040142075496025e-06, "loss": 1.7141, "step": 19931 }, { "epoch": 1.5620689655172413, "grad_norm": 10.379989624023438, "learning_rate": 6.703724302409956e-06, "loss": 0.6708, "step": 19932 }, { "epoch": 1.5621473354231976, "grad_norm": 3.1053149700164795, "learning_rate": 6.703434390790215e-06, "loss": 1.3335, "step": 19933 }, { "epoch": 1.5622257053291535, "grad_norm": 2.071150302886963, "learning_rate": 6.703144472691485e-06, "loss": 1.3017, "step": 19934 }, { "epoch": 1.5623040752351098, "grad_norm": 2.7889981269836426, "learning_rate": 6.70285454811487e-06, "loss": 1.6288, "step": 19935 }, { "epoch": 1.5623824451410657, "grad_norm": 2.7977488040924072, "learning_rate": 6.70256461706147e-06, "loss": 0.9373, "step": 19936 }, { "epoch": 1.562460815047022, "grad_norm": 2.703094720840454, "learning_rate": 6.702274679532388e-06, "loss": 1.1907, "step": 19937 }, { "epoch": 1.562539184952978, "grad_norm": 2.067887544631958, "learning_rate": 6.701984735528728e-06, "loss": 1.2029, "step": 19938 }, { "epoch": 1.5626175548589343, "grad_norm": 3.570897102355957, "learning_rate": 6.701694785051591e-06, "loss": 1.7719, "step": 19939 }, { "epoch": 1.5626959247648902, "grad_norm": 3.324688196182251, "learning_rate": 6.701404828102083e-06, "loss": 1.5132, "step": 19940 }, { "epoch": 1.5627742946708465, "grad_norm": 3.4821345806121826, "learning_rate": 6.701114864681304e-06, "loss": 1.3413, "step": 19941 }, { "epoch": 1.5628526645768024, "grad_norm": 3.064541816711426, "learning_rate": 6.700824894790357e-06, "loss": 1.4062, "step": 19942 }, { "epoch": 1.5629310344827587, "grad_norm": 3.0345442295074463, "learning_rate": 6.700534918430348e-06, "loss": 1.379, "step": 19943 }, { "epoch": 1.5630094043887146, "grad_norm": 2.9832358360290527, "learning_rate": 6.700244935602375e-06, "loss": 1.4519, "step": 19944 }, { "epoch": 1.563087774294671, "grad_norm": 2.061530113220215, "learning_rate": 6.699954946307545e-06, "loss": 1.0525, "step": 19945 }, { "epoch": 1.5631661442006268, "grad_norm": 1.8237671852111816, "learning_rate": 6.699664950546959e-06, "loss": 1.5255, "step": 19946 }, { "epoch": 1.5632445141065832, "grad_norm": 2.9525742530822754, "learning_rate": 6.699374948321722e-06, "loss": 0.8671, "step": 19947 }, { "epoch": 1.563322884012539, "grad_norm": 4.555646896362305, "learning_rate": 6.699084939632934e-06, "loss": 1.533, "step": 19948 }, { "epoch": 1.5634012539184954, "grad_norm": 11.109675407409668, "learning_rate": 6.6987949244817e-06, "loss": 1.1258, "step": 19949 }, { "epoch": 1.5634796238244513, "grad_norm": 3.9757986068725586, "learning_rate": 6.698504902869123e-06, "loss": 1.6665, "step": 19950 }, { "epoch": 1.5635579937304076, "grad_norm": 3.2247867584228516, "learning_rate": 6.698214874796305e-06, "loss": 1.077, "step": 19951 }, { "epoch": 1.5636363636363635, "grad_norm": 2.1391396522521973, "learning_rate": 6.697924840264351e-06, "loss": 1.3514, "step": 19952 }, { "epoch": 1.5637147335423198, "grad_norm": 9.513111114501953, "learning_rate": 6.6976347992743605e-06, "loss": 1.3157, "step": 19953 }, { "epoch": 1.5637931034482757, "grad_norm": 2.5340845584869385, "learning_rate": 6.697344751827442e-06, "loss": 1.9098, "step": 19954 }, { "epoch": 1.563871473354232, "grad_norm": 2.466708183288574, "learning_rate": 6.697054697924695e-06, "loss": 1.213, "step": 19955 }, { "epoch": 1.563949843260188, "grad_norm": 3.235349416732788, "learning_rate": 6.696764637567224e-06, "loss": 1.4089, "step": 19956 }, { "epoch": 1.5640282131661443, "grad_norm": 5.291232585906982, "learning_rate": 6.696474570756131e-06, "loss": 1.5626, "step": 19957 }, { "epoch": 1.5641065830721002, "grad_norm": 2.287569761276245, "learning_rate": 6.6961844974925195e-06, "loss": 1.17, "step": 19958 }, { "epoch": 1.5641849529780565, "grad_norm": 2.356963872909546, "learning_rate": 6.695894417777494e-06, "loss": 1.425, "step": 19959 }, { "epoch": 1.5642633228840124, "grad_norm": 2.709702491760254, "learning_rate": 6.695604331612158e-06, "loss": 1.1751, "step": 19960 }, { "epoch": 1.5643416927899687, "grad_norm": 2.6840310096740723, "learning_rate": 6.695314238997614e-06, "loss": 1.2127, "step": 19961 }, { "epoch": 1.5644200626959248, "grad_norm": 3.214933395385742, "learning_rate": 6.695024139934965e-06, "loss": 0.9886, "step": 19962 }, { "epoch": 1.564498432601881, "grad_norm": 3.492217540740967, "learning_rate": 6.694734034425314e-06, "loss": 1.4934, "step": 19963 }, { "epoch": 1.564576802507837, "grad_norm": 5.508564472198486, "learning_rate": 6.694443922469767e-06, "loss": 0.9508, "step": 19964 }, { "epoch": 1.5646551724137931, "grad_norm": 4.593308448791504, "learning_rate": 6.694153804069423e-06, "loss": 1.1163, "step": 19965 }, { "epoch": 1.5647335423197493, "grad_norm": 2.5670671463012695, "learning_rate": 6.69386367922539e-06, "loss": 1.1846, "step": 19966 }, { "epoch": 1.5648119122257054, "grad_norm": 2.3915796279907227, "learning_rate": 6.69357354793877e-06, "loss": 1.2871, "step": 19967 }, { "epoch": 1.5648902821316615, "grad_norm": 3.023113489151001, "learning_rate": 6.693283410210664e-06, "loss": 1.4942, "step": 19968 }, { "epoch": 1.5649686520376176, "grad_norm": 3.485793352127075, "learning_rate": 6.692993266042179e-06, "loss": 1.1526, "step": 19969 }, { "epoch": 1.5650470219435737, "grad_norm": 4.681019306182861, "learning_rate": 6.692703115434418e-06, "loss": 1.592, "step": 19970 }, { "epoch": 1.5651253918495298, "grad_norm": 7.708241939544678, "learning_rate": 6.6924129583884835e-06, "loss": 0.8421, "step": 19971 }, { "epoch": 1.565203761755486, "grad_norm": 2.2421605587005615, "learning_rate": 6.6921227949054775e-06, "loss": 1.1927, "step": 19972 }, { "epoch": 1.565282131661442, "grad_norm": 4.190406799316406, "learning_rate": 6.691832624986508e-06, "loss": 1.5864, "step": 19973 }, { "epoch": 1.5653605015673981, "grad_norm": 2.138190746307373, "learning_rate": 6.691542448632674e-06, "loss": 1.4944, "step": 19974 }, { "epoch": 1.5654388714733543, "grad_norm": 2.549553394317627, "learning_rate": 6.691252265845083e-06, "loss": 1.1897, "step": 19975 }, { "epoch": 1.5655172413793104, "grad_norm": 1.8546634912490845, "learning_rate": 6.690962076624836e-06, "loss": 1.1852, "step": 19976 }, { "epoch": 1.5655956112852665, "grad_norm": 5.995599269866943, "learning_rate": 6.690671880973039e-06, "loss": 1.0033, "step": 19977 }, { "epoch": 1.5656739811912226, "grad_norm": 2.960116147994995, "learning_rate": 6.690381678890794e-06, "loss": 1.2619, "step": 19978 }, { "epoch": 1.5657523510971787, "grad_norm": 5.722209453582764, "learning_rate": 6.690091470379204e-06, "loss": 1.6613, "step": 19979 }, { "epoch": 1.5658307210031348, "grad_norm": 2.2669882774353027, "learning_rate": 6.689801255439376e-06, "loss": 1.1908, "step": 19980 }, { "epoch": 1.565909090909091, "grad_norm": 2.7027108669281006, "learning_rate": 6.68951103407241e-06, "loss": 1.5011, "step": 19981 }, { "epoch": 1.565987460815047, "grad_norm": 3.1652565002441406, "learning_rate": 6.689220806279414e-06, "loss": 0.7736, "step": 19982 }, { "epoch": 1.5660658307210031, "grad_norm": 2.1568517684936523, "learning_rate": 6.6889305720614895e-06, "loss": 1.3531, "step": 19983 }, { "epoch": 1.5661442006269592, "grad_norm": 13.768672943115234, "learning_rate": 6.688640331419739e-06, "loss": 1.2547, "step": 19984 }, { "epoch": 1.5662225705329154, "grad_norm": 2.7776753902435303, "learning_rate": 6.6883500843552685e-06, "loss": 1.3873, "step": 19985 }, { "epoch": 1.5663009404388715, "grad_norm": 4.968955993652344, "learning_rate": 6.688059830869183e-06, "loss": 1.1272, "step": 19986 }, { "epoch": 1.5663793103448276, "grad_norm": 3.9920787811279297, "learning_rate": 6.687769570962582e-06, "loss": 1.9689, "step": 19987 }, { "epoch": 1.5664576802507837, "grad_norm": 2.7082667350769043, "learning_rate": 6.687479304636575e-06, "loss": 1.2711, "step": 19988 }, { "epoch": 1.5665360501567398, "grad_norm": 2.204509973526001, "learning_rate": 6.687189031892262e-06, "loss": 1.4839, "step": 19989 }, { "epoch": 1.566614420062696, "grad_norm": 3.001612901687622, "learning_rate": 6.686898752730751e-06, "loss": 1.1164, "step": 19990 }, { "epoch": 1.566692789968652, "grad_norm": 2.4295027256011963, "learning_rate": 6.686608467153141e-06, "loss": 1.3883, "step": 19991 }, { "epoch": 1.5667711598746081, "grad_norm": 18.971216201782227, "learning_rate": 6.686318175160539e-06, "loss": 1.7599, "step": 19992 }, { "epoch": 1.5668495297805642, "grad_norm": 3.118922233581543, "learning_rate": 6.686027876754049e-06, "loss": 1.4601, "step": 19993 }, { "epoch": 1.5669278996865204, "grad_norm": 1.9370118379592896, "learning_rate": 6.685737571934775e-06, "loss": 1.3878, "step": 19994 }, { "epoch": 1.5670062695924765, "grad_norm": 2.6992833614349365, "learning_rate": 6.685447260703821e-06, "loss": 1.0901, "step": 19995 }, { "epoch": 1.5670846394984326, "grad_norm": 5.085611820220947, "learning_rate": 6.685156943062292e-06, "loss": 1.3726, "step": 19996 }, { "epoch": 1.5671630094043887, "grad_norm": 2.717467784881592, "learning_rate": 6.684866619011292e-06, "loss": 1.4659, "step": 19997 }, { "epoch": 1.5672413793103448, "grad_norm": 4.227867126464844, "learning_rate": 6.684576288551923e-06, "loss": 1.7238, "step": 19998 }, { "epoch": 1.567319749216301, "grad_norm": 3.3225364685058594, "learning_rate": 6.684285951685293e-06, "loss": 1.7906, "step": 19999 }, { "epoch": 1.567398119122257, "grad_norm": 5.364062309265137, "learning_rate": 6.6839956084125025e-06, "loss": 0.9639, "step": 20000 }, { "epoch": 1.5674764890282131, "grad_norm": 2.948234796524048, "learning_rate": 6.683705258734658e-06, "loss": 1.1367, "step": 20001 }, { "epoch": 1.5675548589341692, "grad_norm": 4.093371391296387, "learning_rate": 6.683414902652865e-06, "loss": 1.1544, "step": 20002 }, { "epoch": 1.5676332288401253, "grad_norm": 3.5361766815185547, "learning_rate": 6.683124540168225e-06, "loss": 1.6252, "step": 20003 }, { "epoch": 1.5677115987460815, "grad_norm": 6.155594348907471, "learning_rate": 6.6828341712818445e-06, "loss": 1.1998, "step": 20004 }, { "epoch": 1.5677899686520376, "grad_norm": 2.0982816219329834, "learning_rate": 6.682543795994827e-06, "loss": 0.9103, "step": 20005 }, { "epoch": 1.5678683385579937, "grad_norm": 5.113744735717773, "learning_rate": 6.682253414308277e-06, "loss": 1.27, "step": 20006 }, { "epoch": 1.5679467084639498, "grad_norm": 3.935379981994629, "learning_rate": 6.681963026223298e-06, "loss": 1.2578, "step": 20007 }, { "epoch": 1.5680250783699061, "grad_norm": 2.9079034328460693, "learning_rate": 6.681672631740997e-06, "loss": 1.6493, "step": 20008 }, { "epoch": 1.568103448275862, "grad_norm": 4.14784574508667, "learning_rate": 6.681382230862478e-06, "loss": 1.4939, "step": 20009 }, { "epoch": 1.5681818181818183, "grad_norm": 12.133569717407227, "learning_rate": 6.681091823588842e-06, "loss": 1.0521, "step": 20010 }, { "epoch": 1.5682601880877742, "grad_norm": 3.3743112087249756, "learning_rate": 6.680801409921198e-06, "loss": 0.8842, "step": 20011 }, { "epoch": 1.5683385579937306, "grad_norm": 3.311866283416748, "learning_rate": 6.680510989860648e-06, "loss": 1.4311, "step": 20012 }, { "epoch": 1.5684169278996865, "grad_norm": 2.996621608734131, "learning_rate": 6.680220563408297e-06, "loss": 1.5514, "step": 20013 }, { "epoch": 1.5684952978056428, "grad_norm": 2.6642074584960938, "learning_rate": 6.6799301305652505e-06, "loss": 1.4439, "step": 20014 }, { "epoch": 1.5685736677115987, "grad_norm": 3.407433032989502, "learning_rate": 6.6796396913326125e-06, "loss": 1.4035, "step": 20015 }, { "epoch": 1.568652037617555, "grad_norm": 3.8151872158050537, "learning_rate": 6.6793492457114894e-06, "loss": 1.5816, "step": 20016 }, { "epoch": 1.568730407523511, "grad_norm": 5.740756511688232, "learning_rate": 6.6790587937029815e-06, "loss": 1.6901, "step": 20017 }, { "epoch": 1.5688087774294672, "grad_norm": 3.0651228427886963, "learning_rate": 6.678768335308199e-06, "loss": 1.1263, "step": 20018 }, { "epoch": 1.5688871473354231, "grad_norm": 2.605318546295166, "learning_rate": 6.678477870528243e-06, "loss": 1.3482, "step": 20019 }, { "epoch": 1.5689655172413794, "grad_norm": 2.300872325897217, "learning_rate": 6.678187399364219e-06, "loss": 1.3235, "step": 20020 }, { "epoch": 1.5690438871473353, "grad_norm": 3.0463385581970215, "learning_rate": 6.677896921817232e-06, "loss": 1.8116, "step": 20021 }, { "epoch": 1.5691222570532917, "grad_norm": 7.469266414642334, "learning_rate": 6.6776064378883866e-06, "loss": 1.217, "step": 20022 }, { "epoch": 1.5692006269592476, "grad_norm": 2.187727212905884, "learning_rate": 6.6773159475787886e-06, "loss": 1.6517, "step": 20023 }, { "epoch": 1.569278996865204, "grad_norm": 4.6239800453186035, "learning_rate": 6.677025450889542e-06, "loss": 1.4283, "step": 20024 }, { "epoch": 1.5693573667711598, "grad_norm": 3.3640716075897217, "learning_rate": 6.676734947821752e-06, "loss": 1.255, "step": 20025 }, { "epoch": 1.5694357366771161, "grad_norm": 2.0158345699310303, "learning_rate": 6.676444438376524e-06, "loss": 1.2491, "step": 20026 }, { "epoch": 1.569514106583072, "grad_norm": 12.436601638793945, "learning_rate": 6.6761539225549625e-06, "loss": 0.8558, "step": 20027 }, { "epoch": 1.5695924764890283, "grad_norm": 2.15083646774292, "learning_rate": 6.675863400358171e-06, "loss": 1.1517, "step": 20028 }, { "epoch": 1.5696708463949842, "grad_norm": 2.292633533477783, "learning_rate": 6.675572871787257e-06, "loss": 1.1801, "step": 20029 }, { "epoch": 1.5697492163009406, "grad_norm": 2.7403106689453125, "learning_rate": 6.675282336843325e-06, "loss": 1.2583, "step": 20030 }, { "epoch": 1.5698275862068964, "grad_norm": 5.082547664642334, "learning_rate": 6.674991795527479e-06, "loss": 1.2115, "step": 20031 }, { "epoch": 1.5699059561128528, "grad_norm": 3.3826422691345215, "learning_rate": 6.674701247840823e-06, "loss": 1.5721, "step": 20032 }, { "epoch": 1.5699843260188087, "grad_norm": 2.0651416778564453, "learning_rate": 6.674410693784466e-06, "loss": 1.0433, "step": 20033 }, { "epoch": 1.570062695924765, "grad_norm": 3.50545334815979, "learning_rate": 6.67412013335951e-06, "loss": 1.0837, "step": 20034 }, { "epoch": 1.5701410658307209, "grad_norm": 2.4348697662353516, "learning_rate": 6.673829566567059e-06, "loss": 1.3914, "step": 20035 }, { "epoch": 1.5702194357366772, "grad_norm": 2.828073263168335, "learning_rate": 6.6735389934082225e-06, "loss": 0.8735, "step": 20036 }, { "epoch": 1.570297805642633, "grad_norm": 2.170436382293701, "learning_rate": 6.6732484138841034e-06, "loss": 1.5326, "step": 20037 }, { "epoch": 1.5703761755485894, "grad_norm": 2.665287733078003, "learning_rate": 6.672957827995805e-06, "loss": 1.6258, "step": 20038 }, { "epoch": 1.5704545454545453, "grad_norm": 1.6907376050949097, "learning_rate": 6.6726672357444356e-06, "loss": 1.2657, "step": 20039 }, { "epoch": 1.5705329153605017, "grad_norm": 2.1727588176727295, "learning_rate": 6.672376637131098e-06, "loss": 1.4617, "step": 20040 }, { "epoch": 1.5706112852664575, "grad_norm": 2.9806976318359375, "learning_rate": 6.6720860321568995e-06, "loss": 1.6314, "step": 20041 }, { "epoch": 1.5706896551724139, "grad_norm": 4.182343482971191, "learning_rate": 6.671795420822944e-06, "loss": 1.3028, "step": 20042 }, { "epoch": 1.5707680250783698, "grad_norm": 5.477461338043213, "learning_rate": 6.671504803130337e-06, "loss": 1.474, "step": 20043 }, { "epoch": 1.570846394984326, "grad_norm": 3.527970314025879, "learning_rate": 6.6712141790801865e-06, "loss": 1.5376, "step": 20044 }, { "epoch": 1.570924764890282, "grad_norm": 2.719691514968872, "learning_rate": 6.670923548673594e-06, "loss": 1.1387, "step": 20045 }, { "epoch": 1.5710031347962383, "grad_norm": 4.631570816040039, "learning_rate": 6.670632911911667e-06, "loss": 1.5206, "step": 20046 }, { "epoch": 1.5710815047021942, "grad_norm": 4.359252452850342, "learning_rate": 6.6703422687955095e-06, "loss": 1.1767, "step": 20047 }, { "epoch": 1.5711598746081505, "grad_norm": 3.0401694774627686, "learning_rate": 6.67005161932623e-06, "loss": 1.0155, "step": 20048 }, { "epoch": 1.5712382445141064, "grad_norm": 3.4487905502319336, "learning_rate": 6.66976096350493e-06, "loss": 1.2791, "step": 20049 }, { "epoch": 1.5713166144200628, "grad_norm": 3.5234622955322266, "learning_rate": 6.669470301332718e-06, "loss": 1.4013, "step": 20050 }, { "epoch": 1.5713949843260187, "grad_norm": 5.747122764587402, "learning_rate": 6.669179632810698e-06, "loss": 2.181, "step": 20051 }, { "epoch": 1.571473354231975, "grad_norm": 5.971458435058594, "learning_rate": 6.668888957939975e-06, "loss": 1.5259, "step": 20052 }, { "epoch": 1.5715517241379309, "grad_norm": 2.9203784465789795, "learning_rate": 6.668598276721659e-06, "loss": 1.0416, "step": 20053 }, { "epoch": 1.5716300940438872, "grad_norm": 4.025247573852539, "learning_rate": 6.668307589156848e-06, "loss": 1.3523, "step": 20054 }, { "epoch": 1.5717084639498433, "grad_norm": 2.3364741802215576, "learning_rate": 6.668016895246654e-06, "loss": 1.2925, "step": 20055 }, { "epoch": 1.5717868338557994, "grad_norm": 3.517144203186035, "learning_rate": 6.66772619499218e-06, "loss": 1.3975, "step": 20056 }, { "epoch": 1.5718652037617555, "grad_norm": 5.116374492645264, "learning_rate": 6.667435488394533e-06, "loss": 1.7988, "step": 20057 }, { "epoch": 1.5719435736677116, "grad_norm": 4.19935417175293, "learning_rate": 6.667144775454817e-06, "loss": 1.3549, "step": 20058 }, { "epoch": 1.5720219435736678, "grad_norm": 3.0872204303741455, "learning_rate": 6.666854056174139e-06, "loss": 1.0142, "step": 20059 }, { "epoch": 1.5721003134796239, "grad_norm": 3.3747127056121826, "learning_rate": 6.666563330553603e-06, "loss": 1.8091, "step": 20060 }, { "epoch": 1.57217868338558, "grad_norm": 3.5479657649993896, "learning_rate": 6.666272598594318e-06, "loss": 1.6256, "step": 20061 }, { "epoch": 1.572257053291536, "grad_norm": 2.412977933883667, "learning_rate": 6.665981860297388e-06, "loss": 1.3883, "step": 20062 }, { "epoch": 1.5723354231974922, "grad_norm": 2.3756000995635986, "learning_rate": 6.665691115663917e-06, "loss": 1.2695, "step": 20063 }, { "epoch": 1.5724137931034483, "grad_norm": 4.805812358856201, "learning_rate": 6.665400364695014e-06, "loss": 1.3759, "step": 20064 }, { "epoch": 1.5724921630094044, "grad_norm": 2.3536014556884766, "learning_rate": 6.665109607391782e-06, "loss": 1.5273, "step": 20065 }, { "epoch": 1.5725705329153605, "grad_norm": 4.434027194976807, "learning_rate": 6.66481884375533e-06, "loss": 1.3919, "step": 20066 }, { "epoch": 1.5726489028213166, "grad_norm": 3.0608887672424316, "learning_rate": 6.66452807378676e-06, "loss": 1.4213, "step": 20067 }, { "epoch": 1.5727272727272728, "grad_norm": 3.8821284770965576, "learning_rate": 6.664237297487182e-06, "loss": 1.3928, "step": 20068 }, { "epoch": 1.5728056426332289, "grad_norm": 3.804046869277954, "learning_rate": 6.663946514857699e-06, "loss": 1.3507, "step": 20069 }, { "epoch": 1.572884012539185, "grad_norm": 6.878783702850342, "learning_rate": 6.663655725899418e-06, "loss": 1.4736, "step": 20070 }, { "epoch": 1.572962382445141, "grad_norm": 3.6976990699768066, "learning_rate": 6.663364930613446e-06, "loss": 1.8086, "step": 20071 }, { "epoch": 1.5730407523510972, "grad_norm": 2.960062265396118, "learning_rate": 6.66307412900089e-06, "loss": 0.939, "step": 20072 }, { "epoch": 1.5731191222570533, "grad_norm": 3.715928792953491, "learning_rate": 6.66278332106285e-06, "loss": 1.3967, "step": 20073 }, { "epoch": 1.5731974921630094, "grad_norm": 5.342013835906982, "learning_rate": 6.66249250680044e-06, "loss": 1.1737, "step": 20074 }, { "epoch": 1.5732758620689655, "grad_norm": 4.143653869628906, "learning_rate": 6.66220168621476e-06, "loss": 1.3663, "step": 20075 }, { "epoch": 1.5733542319749216, "grad_norm": 3.3619890213012695, "learning_rate": 6.661910859306919e-06, "loss": 1.4418, "step": 20076 }, { "epoch": 1.5734326018808777, "grad_norm": 3.207871198654175, "learning_rate": 6.661620026078024e-06, "loss": 1.0904, "step": 20077 }, { "epoch": 1.5735109717868339, "grad_norm": 2.461219310760498, "learning_rate": 6.661329186529179e-06, "loss": 1.4222, "step": 20078 }, { "epoch": 1.57358934169279, "grad_norm": 3.7749130725860596, "learning_rate": 6.661038340661492e-06, "loss": 1.7797, "step": 20079 }, { "epoch": 1.573667711598746, "grad_norm": 3.5075736045837402, "learning_rate": 6.660747488476066e-06, "loss": 1.4579, "step": 20080 }, { "epoch": 1.5737460815047022, "grad_norm": 2.467238426208496, "learning_rate": 6.660456629974013e-06, "loss": 1.5521, "step": 20081 }, { "epoch": 1.5738244514106583, "grad_norm": 3.858497142791748, "learning_rate": 6.660165765156432e-06, "loss": 1.7292, "step": 20082 }, { "epoch": 1.5739028213166144, "grad_norm": 2.5244696140289307, "learning_rate": 6.659874894024436e-06, "loss": 1.6458, "step": 20083 }, { "epoch": 1.5739811912225705, "grad_norm": 2.4497740268707275, "learning_rate": 6.659584016579127e-06, "loss": 1.6467, "step": 20084 }, { "epoch": 1.5740595611285266, "grad_norm": 2.448829174041748, "learning_rate": 6.6592931328216135e-06, "loss": 1.2997, "step": 20085 }, { "epoch": 1.5741379310344827, "grad_norm": 3.340491771697998, "learning_rate": 6.659002242753001e-06, "loss": 1.3962, "step": 20086 }, { "epoch": 1.5742163009404389, "grad_norm": 5.507172107696533, "learning_rate": 6.658711346374396e-06, "loss": 1.5259, "step": 20087 }, { "epoch": 1.574294670846395, "grad_norm": 2.746593952178955, "learning_rate": 6.658420443686904e-06, "loss": 1.2801, "step": 20088 }, { "epoch": 1.574373040752351, "grad_norm": 3.7209818363189697, "learning_rate": 6.658129534691634e-06, "loss": 1.5005, "step": 20089 }, { "epoch": 1.5744514106583072, "grad_norm": 5.048358917236328, "learning_rate": 6.6578386193896915e-06, "loss": 1.3776, "step": 20090 }, { "epoch": 1.5745297805642633, "grad_norm": 3.3167364597320557, "learning_rate": 6.657547697782181e-06, "loss": 1.2976, "step": 20091 }, { "epoch": 1.5746081504702194, "grad_norm": 2.1406984329223633, "learning_rate": 6.65725676987021e-06, "loss": 1.3685, "step": 20092 }, { "epoch": 1.5746865203761755, "grad_norm": 4.235101222991943, "learning_rate": 6.6569658356548876e-06, "loss": 1.5596, "step": 20093 }, { "epoch": 1.5747648902821316, "grad_norm": 2.0009982585906982, "learning_rate": 6.656674895137316e-06, "loss": 1.373, "step": 20094 }, { "epoch": 1.5748432601880877, "grad_norm": 2.893345832824707, "learning_rate": 6.656383948318604e-06, "loss": 1.0458, "step": 20095 }, { "epoch": 1.5749216300940438, "grad_norm": 3.097886800765991, "learning_rate": 6.65609299519986e-06, "loss": 1.8163, "step": 20096 }, { "epoch": 1.575, "grad_norm": 2.6587319374084473, "learning_rate": 6.655802035782187e-06, "loss": 1.1576, "step": 20097 }, { "epoch": 1.575078369905956, "grad_norm": 3.209346294403076, "learning_rate": 6.655511070066696e-06, "loss": 1.6552, "step": 20098 }, { "epoch": 1.5751567398119122, "grad_norm": 2.615614414215088, "learning_rate": 6.655220098054489e-06, "loss": 1.1004, "step": 20099 }, { "epoch": 1.5752351097178683, "grad_norm": 2.2742409706115723, "learning_rate": 6.6549291197466755e-06, "loss": 1.2886, "step": 20100 }, { "epoch": 1.5753134796238244, "grad_norm": 2.619105577468872, "learning_rate": 6.654638135144361e-06, "loss": 1.0462, "step": 20101 }, { "epoch": 1.5753918495297805, "grad_norm": 2.332566738128662, "learning_rate": 6.654347144248655e-06, "loss": 1.4648, "step": 20102 }, { "epoch": 1.5754702194357368, "grad_norm": 2.907257318496704, "learning_rate": 6.654056147060661e-06, "loss": 1.6372, "step": 20103 }, { "epoch": 1.5755485893416927, "grad_norm": 2.1638195514678955, "learning_rate": 6.6537651435814864e-06, "loss": 1.2784, "step": 20104 }, { "epoch": 1.575626959247649, "grad_norm": 6.611091613769531, "learning_rate": 6.65347413381224e-06, "loss": 0.9486, "step": 20105 }, { "epoch": 1.575705329153605, "grad_norm": 4.635976791381836, "learning_rate": 6.653183117754025e-06, "loss": 2.0471, "step": 20106 }, { "epoch": 1.5757836990595613, "grad_norm": 2.0637035369873047, "learning_rate": 6.652892095407953e-06, "loss": 1.6002, "step": 20107 }, { "epoch": 1.5758620689655172, "grad_norm": 2.1951050758361816, "learning_rate": 6.6526010667751265e-06, "loss": 1.4717, "step": 20108 }, { "epoch": 1.5759404388714735, "grad_norm": 2.9493846893310547, "learning_rate": 6.652310031856657e-06, "loss": 1.3095, "step": 20109 }, { "epoch": 1.5760188087774294, "grad_norm": 2.8106377124786377, "learning_rate": 6.652018990653646e-06, "loss": 1.5497, "step": 20110 }, { "epoch": 1.5760971786833857, "grad_norm": 2.986820936203003, "learning_rate": 6.651727943167205e-06, "loss": 1.3301, "step": 20111 }, { "epoch": 1.5761755485893416, "grad_norm": 2.577772378921509, "learning_rate": 6.651436889398438e-06, "loss": 1.3628, "step": 20112 }, { "epoch": 1.576253918495298, "grad_norm": 2.2534313201904297, "learning_rate": 6.651145829348454e-06, "loss": 1.4273, "step": 20113 }, { "epoch": 1.5763322884012538, "grad_norm": 3.439460277557373, "learning_rate": 6.65085476301836e-06, "loss": 1.2896, "step": 20114 }, { "epoch": 1.5764106583072102, "grad_norm": 3.8927412033081055, "learning_rate": 6.650563690409261e-06, "loss": 1.8118, "step": 20115 }, { "epoch": 1.576489028213166, "grad_norm": 2.870150566101074, "learning_rate": 6.650272611522266e-06, "loss": 1.2695, "step": 20116 }, { "epoch": 1.5765673981191224, "grad_norm": 8.896941184997559, "learning_rate": 6.649981526358483e-06, "loss": 1.2485, "step": 20117 }, { "epoch": 1.5766457680250783, "grad_norm": 7.395367622375488, "learning_rate": 6.649690434919016e-06, "loss": 1.833, "step": 20118 }, { "epoch": 1.5767241379310346, "grad_norm": 3.4362030029296875, "learning_rate": 6.649399337204975e-06, "loss": 1.3168, "step": 20119 }, { "epoch": 1.5768025078369905, "grad_norm": 2.371674060821533, "learning_rate": 6.649108233217465e-06, "loss": 1.452, "step": 20120 }, { "epoch": 1.5768808777429468, "grad_norm": 3.240497589111328, "learning_rate": 6.648817122957595e-06, "loss": 1.4446, "step": 20121 }, { "epoch": 1.5769592476489027, "grad_norm": 10.05712890625, "learning_rate": 6.648526006426471e-06, "loss": 1.1322, "step": 20122 }, { "epoch": 1.577037617554859, "grad_norm": 3.515024423599243, "learning_rate": 6.648234883625202e-06, "loss": 0.8546, "step": 20123 }, { "epoch": 1.577115987460815, "grad_norm": 8.201712608337402, "learning_rate": 6.6479437545548934e-06, "loss": 1.2658, "step": 20124 }, { "epoch": 1.5771943573667713, "grad_norm": 3.744607925415039, "learning_rate": 6.6476526192166525e-06, "loss": 1.5848, "step": 20125 }, { "epoch": 1.5772727272727272, "grad_norm": 2.7728350162506104, "learning_rate": 6.647361477611589e-06, "loss": 1.5241, "step": 20126 }, { "epoch": 1.5773510971786835, "grad_norm": 2.3884449005126953, "learning_rate": 6.647070329740807e-06, "loss": 1.3202, "step": 20127 }, { "epoch": 1.5774294670846394, "grad_norm": 4.5633721351623535, "learning_rate": 6.646779175605417e-06, "loss": 1.4649, "step": 20128 }, { "epoch": 1.5775078369905957, "grad_norm": 5.010488033294678, "learning_rate": 6.646488015206523e-06, "loss": 1.653, "step": 20129 }, { "epoch": 1.5775862068965516, "grad_norm": 3.044825792312622, "learning_rate": 6.6461968485452355e-06, "loss": 1.6159, "step": 20130 }, { "epoch": 1.577664576802508, "grad_norm": 1.9198657274246216, "learning_rate": 6.64590567562266e-06, "loss": 1.4015, "step": 20131 }, { "epoch": 1.5777429467084638, "grad_norm": 3.586561918258667, "learning_rate": 6.645614496439905e-06, "loss": 1.1793, "step": 20132 }, { "epoch": 1.5778213166144202, "grad_norm": 8.838685035705566, "learning_rate": 6.64532331099808e-06, "loss": 1.935, "step": 20133 }, { "epoch": 1.577899686520376, "grad_norm": 4.1241559982299805, "learning_rate": 6.645032119298286e-06, "loss": 1.1324, "step": 20134 }, { "epoch": 1.5779780564263324, "grad_norm": 2.0474636554718018, "learning_rate": 6.644740921341638e-06, "loss": 1.1917, "step": 20135 }, { "epoch": 1.5780564263322883, "grad_norm": 4.60486364364624, "learning_rate": 6.644449717129238e-06, "loss": 1.9133, "step": 20136 }, { "epoch": 1.5781347962382446, "grad_norm": 4.2680983543396, "learning_rate": 6.6441585066621975e-06, "loss": 1.3411, "step": 20137 }, { "epoch": 1.5782131661442005, "grad_norm": 4.317613124847412, "learning_rate": 6.643867289941622e-06, "loss": 1.5515, "step": 20138 }, { "epoch": 1.5782915360501568, "grad_norm": 2.015261650085449, "learning_rate": 6.64357606696862e-06, "loss": 1.1115, "step": 20139 }, { "epoch": 1.5783699059561127, "grad_norm": 2.92688250541687, "learning_rate": 6.643284837744298e-06, "loss": 1.2114, "step": 20140 }, { "epoch": 1.578448275862069, "grad_norm": 4.669975757598877, "learning_rate": 6.642993602269765e-06, "loss": 1.3493, "step": 20141 }, { "epoch": 1.578526645768025, "grad_norm": 2.542793035507202, "learning_rate": 6.642702360546128e-06, "loss": 1.299, "step": 20142 }, { "epoch": 1.5786050156739813, "grad_norm": 5.723719120025635, "learning_rate": 6.6424111125744965e-06, "loss": 1.3542, "step": 20143 }, { "epoch": 1.5786833855799371, "grad_norm": 2.702897548675537, "learning_rate": 6.642119858355974e-06, "loss": 1.0324, "step": 20144 }, { "epoch": 1.5787617554858935, "grad_norm": 2.7327232360839844, "learning_rate": 6.641828597891674e-06, "loss": 1.7259, "step": 20145 }, { "epoch": 1.5788401253918494, "grad_norm": 3.375586748123169, "learning_rate": 6.641537331182699e-06, "loss": 0.903, "step": 20146 }, { "epoch": 1.5789184952978057, "grad_norm": 3.1520824432373047, "learning_rate": 6.641246058230161e-06, "loss": 1.3597, "step": 20147 }, { "epoch": 1.5789968652037616, "grad_norm": 4.749095916748047, "learning_rate": 6.6409547790351646e-06, "loss": 1.7389, "step": 20148 }, { "epoch": 1.579075235109718, "grad_norm": 3.7851853370666504, "learning_rate": 6.640663493598819e-06, "loss": 1.2887, "step": 20149 }, { "epoch": 1.579153605015674, "grad_norm": 2.925891876220703, "learning_rate": 6.640372201922233e-06, "loss": 1.4009, "step": 20150 }, { "epoch": 1.5792319749216301, "grad_norm": 4.386595249176025, "learning_rate": 6.640080904006513e-06, "loss": 1.4689, "step": 20151 }, { "epoch": 1.5793103448275863, "grad_norm": 4.257596015930176, "learning_rate": 6.639789599852769e-06, "loss": 1.5838, "step": 20152 }, { "epoch": 1.5793887147335424, "grad_norm": 4.557248115539551, "learning_rate": 6.639498289462105e-06, "loss": 1.0982, "step": 20153 }, { "epoch": 1.5794670846394985, "grad_norm": 2.331505537033081, "learning_rate": 6.639206972835635e-06, "loss": 1.6002, "step": 20154 }, { "epoch": 1.5795454545454546, "grad_norm": 2.520062208175659, "learning_rate": 6.63891564997446e-06, "loss": 1.4305, "step": 20155 }, { "epoch": 1.5796238244514107, "grad_norm": 3.12764048576355, "learning_rate": 6.638624320879694e-06, "loss": 1.2082, "step": 20156 }, { "epoch": 1.5797021943573668, "grad_norm": 6.519546985626221, "learning_rate": 6.638332985552443e-06, "loss": 1.5167, "step": 20157 }, { "epoch": 1.579780564263323, "grad_norm": 4.74826192855835, "learning_rate": 6.638041643993813e-06, "loss": 1.6485, "step": 20158 }, { "epoch": 1.579858934169279, "grad_norm": 3.967533826828003, "learning_rate": 6.637750296204917e-06, "loss": 1.0206, "step": 20159 }, { "epoch": 1.5799373040752351, "grad_norm": 2.287372350692749, "learning_rate": 6.637458942186856e-06, "loss": 1.5369, "step": 20160 }, { "epoch": 1.5800156739811912, "grad_norm": 3.1052985191345215, "learning_rate": 6.637167581940746e-06, "loss": 1.4875, "step": 20161 }, { "epoch": 1.5800940438871474, "grad_norm": 5.555791854858398, "learning_rate": 6.636876215467688e-06, "loss": 1.3546, "step": 20162 }, { "epoch": 1.5801724137931035, "grad_norm": 4.829458236694336, "learning_rate": 6.636584842768796e-06, "loss": 1.8031, "step": 20163 }, { "epoch": 1.5802507836990596, "grad_norm": 2.5604066848754883, "learning_rate": 6.6362934638451746e-06, "loss": 1.6199, "step": 20164 }, { "epoch": 1.5803291536050157, "grad_norm": 2.9205992221832275, "learning_rate": 6.636002078697934e-06, "loss": 1.3222, "step": 20165 }, { "epoch": 1.5804075235109718, "grad_norm": 5.094528675079346, "learning_rate": 6.635710687328181e-06, "loss": 1.6038, "step": 20166 }, { "epoch": 1.580485893416928, "grad_norm": 3.2696988582611084, "learning_rate": 6.635419289737026e-06, "loss": 1.8688, "step": 20167 }, { "epoch": 1.580564263322884, "grad_norm": 4.126356601715088, "learning_rate": 6.635127885925575e-06, "loss": 1.0579, "step": 20168 }, { "epoch": 1.5806426332288401, "grad_norm": 1.7602914571762085, "learning_rate": 6.634836475894938e-06, "loss": 1.5204, "step": 20169 }, { "epoch": 1.5807210031347962, "grad_norm": 2.2932536602020264, "learning_rate": 6.6345450596462224e-06, "loss": 1.5842, "step": 20170 }, { "epoch": 1.5807993730407524, "grad_norm": 4.207779407501221, "learning_rate": 6.634253637180536e-06, "loss": 1.389, "step": 20171 }, { "epoch": 1.5808777429467085, "grad_norm": 2.7604386806488037, "learning_rate": 6.633962208498992e-06, "loss": 1.4397, "step": 20172 }, { "epoch": 1.5809561128526646, "grad_norm": 5.226703643798828, "learning_rate": 6.633670773602691e-06, "loss": 1.6026, "step": 20173 }, { "epoch": 1.5810344827586207, "grad_norm": 2.208725690841675, "learning_rate": 6.633379332492746e-06, "loss": 1.2788, "step": 20174 }, { "epoch": 1.5811128526645768, "grad_norm": 3.2000060081481934, "learning_rate": 6.633087885170266e-06, "loss": 1.6632, "step": 20175 }, { "epoch": 1.581191222570533, "grad_norm": 2.2466883659362793, "learning_rate": 6.6327964316363595e-06, "loss": 1.6303, "step": 20176 }, { "epoch": 1.581269592476489, "grad_norm": 9.572805404663086, "learning_rate": 6.632504971892132e-06, "loss": 1.2346, "step": 20177 }, { "epoch": 1.5813479623824451, "grad_norm": 3.311377763748169, "learning_rate": 6.632213505938695e-06, "loss": 1.4692, "step": 20178 }, { "epoch": 1.5814263322884012, "grad_norm": 3.1355063915252686, "learning_rate": 6.631922033777156e-06, "loss": 1.4853, "step": 20179 }, { "epoch": 1.5815047021943573, "grad_norm": 2.2012081146240234, "learning_rate": 6.631630555408625e-06, "loss": 1.4776, "step": 20180 }, { "epoch": 1.5815830721003135, "grad_norm": 2.5489354133605957, "learning_rate": 6.631339070834206e-06, "loss": 1.3084, "step": 20181 }, { "epoch": 1.5816614420062696, "grad_norm": 3.880403995513916, "learning_rate": 6.631047580055015e-06, "loss": 1.4652, "step": 20182 }, { "epoch": 1.5817398119122257, "grad_norm": 2.7997617721557617, "learning_rate": 6.630756083072154e-06, "loss": 1.6368, "step": 20183 }, { "epoch": 1.5818181818181818, "grad_norm": 2.144491195678711, "learning_rate": 6.630464579886737e-06, "loss": 1.2337, "step": 20184 }, { "epoch": 1.581896551724138, "grad_norm": 3.0942718982696533, "learning_rate": 6.630173070499868e-06, "loss": 1.3729, "step": 20185 }, { "epoch": 1.581974921630094, "grad_norm": 3.154601812362671, "learning_rate": 6.629881554912658e-06, "loss": 1.0137, "step": 20186 }, { "epoch": 1.5820532915360501, "grad_norm": 3.261253833770752, "learning_rate": 6.629590033126218e-06, "loss": 1.3632, "step": 20187 }, { "epoch": 1.5821316614420062, "grad_norm": 6.703990459442139, "learning_rate": 6.629298505141651e-06, "loss": 1.0802, "step": 20188 }, { "epoch": 1.5822100313479623, "grad_norm": 1.832072138786316, "learning_rate": 6.629006970960072e-06, "loss": 1.2552, "step": 20189 }, { "epoch": 1.5822884012539185, "grad_norm": 4.445194721221924, "learning_rate": 6.628715430582586e-06, "loss": 1.8999, "step": 20190 }, { "epoch": 1.5823667711598746, "grad_norm": 3.7236063480377197, "learning_rate": 6.628423884010305e-06, "loss": 1.4812, "step": 20191 }, { "epoch": 1.5824451410658307, "grad_norm": 2.6256203651428223, "learning_rate": 6.628132331244335e-06, "loss": 1.1783, "step": 20192 }, { "epoch": 1.5825235109717868, "grad_norm": 2.3543777465820312, "learning_rate": 6.627840772285784e-06, "loss": 1.0824, "step": 20193 }, { "epoch": 1.582601880877743, "grad_norm": 3.8400394916534424, "learning_rate": 6.627549207135764e-06, "loss": 1.3452, "step": 20194 }, { "epoch": 1.582680250783699, "grad_norm": 2.4466753005981445, "learning_rate": 6.627257635795383e-06, "loss": 1.435, "step": 20195 }, { "epoch": 1.5827586206896553, "grad_norm": 4.7909626960754395, "learning_rate": 6.62696605826575e-06, "loss": 1.4906, "step": 20196 }, { "epoch": 1.5828369905956112, "grad_norm": 4.191154956817627, "learning_rate": 6.6266744745479735e-06, "loss": 1.5152, "step": 20197 }, { "epoch": 1.5829153605015676, "grad_norm": 3.181001901626587, "learning_rate": 6.626382884643162e-06, "loss": 1.3772, "step": 20198 }, { "epoch": 1.5829937304075234, "grad_norm": 1.8754216432571411, "learning_rate": 6.626091288552428e-06, "loss": 1.4397, "step": 20199 }, { "epoch": 1.5830721003134798, "grad_norm": 22.33770179748535, "learning_rate": 6.625799686276876e-06, "loss": 2.038, "step": 20200 }, { "epoch": 1.5831504702194357, "grad_norm": 3.5512380599975586, "learning_rate": 6.625508077817617e-06, "loss": 1.6648, "step": 20201 }, { "epoch": 1.583228840125392, "grad_norm": 2.58207106590271, "learning_rate": 6.62521646317576e-06, "loss": 1.2004, "step": 20202 }, { "epoch": 1.583307210031348, "grad_norm": 2.5141444206237793, "learning_rate": 6.624924842352415e-06, "loss": 1.3922, "step": 20203 }, { "epoch": 1.5833855799373042, "grad_norm": 3.0959835052490234, "learning_rate": 6.62463321534869e-06, "loss": 1.6438, "step": 20204 }, { "epoch": 1.58346394984326, "grad_norm": 3.1276462078094482, "learning_rate": 6.624341582165695e-06, "loss": 1.183, "step": 20205 }, { "epoch": 1.5835423197492164, "grad_norm": 2.6496472358703613, "learning_rate": 6.62404994280454e-06, "loss": 1.3576, "step": 20206 }, { "epoch": 1.5836206896551723, "grad_norm": 6.147639751434326, "learning_rate": 6.623758297266331e-06, "loss": 1.5845, "step": 20207 }, { "epoch": 1.5836990595611287, "grad_norm": 13.57062816619873, "learning_rate": 6.623466645552181e-06, "loss": 1.5901, "step": 20208 }, { "epoch": 1.5837774294670846, "grad_norm": 2.850569248199463, "learning_rate": 6.623174987663196e-06, "loss": 1.627, "step": 20209 }, { "epoch": 1.5838557993730409, "grad_norm": 3.4168355464935303, "learning_rate": 6.62288332360049e-06, "loss": 1.3307, "step": 20210 }, { "epoch": 1.5839341692789968, "grad_norm": 3.37555193901062, "learning_rate": 6.622591653365167e-06, "loss": 1.7655, "step": 20211 }, { "epoch": 1.584012539184953, "grad_norm": 1.9785236120224, "learning_rate": 6.62229997695834e-06, "loss": 1.2539, "step": 20212 }, { "epoch": 1.584090909090909, "grad_norm": 3.4313759803771973, "learning_rate": 6.622008294381116e-06, "loss": 0.6604, "step": 20213 }, { "epoch": 1.5841692789968653, "grad_norm": 3.436110258102417, "learning_rate": 6.621716605634606e-06, "loss": 1.4874, "step": 20214 }, { "epoch": 1.5842476489028212, "grad_norm": 2.1674036979675293, "learning_rate": 6.62142491071992e-06, "loss": 1.3398, "step": 20215 }, { "epoch": 1.5843260188087775, "grad_norm": 2.284842014312744, "learning_rate": 6.6211332096381645e-06, "loss": 0.7438, "step": 20216 }, { "epoch": 1.5844043887147334, "grad_norm": 3.675607919692993, "learning_rate": 6.620841502390454e-06, "loss": 1.5761, "step": 20217 }, { "epoch": 1.5844827586206898, "grad_norm": 7.239332675933838, "learning_rate": 6.620549788977892e-06, "loss": 0.9566, "step": 20218 }, { "epoch": 1.5845611285266457, "grad_norm": 2.274132013320923, "learning_rate": 6.620258069401593e-06, "loss": 1.2033, "step": 20219 }, { "epoch": 1.584639498432602, "grad_norm": 2.6275362968444824, "learning_rate": 6.619966343662662e-06, "loss": 1.6225, "step": 20220 }, { "epoch": 1.5847178683385579, "grad_norm": 2.620053768157959, "learning_rate": 6.6196746117622134e-06, "loss": 1.5978, "step": 20221 }, { "epoch": 1.5847962382445142, "grad_norm": 2.2891106605529785, "learning_rate": 6.619382873701353e-06, "loss": 1.3782, "step": 20222 }, { "epoch": 1.58487460815047, "grad_norm": 3.194049119949341, "learning_rate": 6.619091129481194e-06, "loss": 1.3324, "step": 20223 }, { "epoch": 1.5849529780564264, "grad_norm": 2.267995834350586, "learning_rate": 6.618799379102841e-06, "loss": 1.2426, "step": 20224 }, { "epoch": 1.5850313479623823, "grad_norm": 2.1232988834381104, "learning_rate": 6.618507622567408e-06, "loss": 1.2076, "step": 20225 }, { "epoch": 1.5851097178683387, "grad_norm": 2.4820570945739746, "learning_rate": 6.618215859876004e-06, "loss": 1.5109, "step": 20226 }, { "epoch": 1.5851880877742945, "grad_norm": 5.752037048339844, "learning_rate": 6.617924091029739e-06, "loss": 1.4367, "step": 20227 }, { "epoch": 1.5852664576802509, "grad_norm": 2.525472402572632, "learning_rate": 6.61763231602972e-06, "loss": 1.1065, "step": 20228 }, { "epoch": 1.5853448275862068, "grad_norm": 2.642575263977051, "learning_rate": 6.6173405348770594e-06, "loss": 1.3791, "step": 20229 }, { "epoch": 1.585423197492163, "grad_norm": 7.823812007904053, "learning_rate": 6.617048747572865e-06, "loss": 1.2501, "step": 20230 }, { "epoch": 1.585501567398119, "grad_norm": 3.776660442352295, "learning_rate": 6.616756954118249e-06, "loss": 1.6996, "step": 20231 }, { "epoch": 1.5855799373040753, "grad_norm": 4.460891246795654, "learning_rate": 6.616465154514318e-06, "loss": 1.1191, "step": 20232 }, { "epoch": 1.5856583072100312, "grad_norm": 6.871138572692871, "learning_rate": 6.6161733487621846e-06, "loss": 1.6549, "step": 20233 }, { "epoch": 1.5857366771159875, "grad_norm": 3.660749673843384, "learning_rate": 6.61588153686296e-06, "loss": 1.3527, "step": 20234 }, { "epoch": 1.5858150470219434, "grad_norm": 2.296220302581787, "learning_rate": 6.61558971881775e-06, "loss": 1.5357, "step": 20235 }, { "epoch": 1.5858934169278998, "grad_norm": 3.1196484565734863, "learning_rate": 6.615297894627668e-06, "loss": 1.34, "step": 20236 }, { "epoch": 1.5859717868338556, "grad_norm": 4.785458087921143, "learning_rate": 6.61500606429382e-06, "loss": 1.2137, "step": 20237 }, { "epoch": 1.586050156739812, "grad_norm": 3.8081319332122803, "learning_rate": 6.61471422781732e-06, "loss": 1.4204, "step": 20238 }, { "epoch": 1.5861285266457679, "grad_norm": 3.176731586456299, "learning_rate": 6.614422385199277e-06, "loss": 1.5094, "step": 20239 }, { "epoch": 1.5862068965517242, "grad_norm": 4.443747043609619, "learning_rate": 6.614130536440799e-06, "loss": 1.5629, "step": 20240 }, { "epoch": 1.58628526645768, "grad_norm": 3.576172351837158, "learning_rate": 6.613838681542999e-06, "loss": 1.4995, "step": 20241 }, { "epoch": 1.5863636363636364, "grad_norm": 25.239200592041016, "learning_rate": 6.613546820506984e-06, "loss": 1.7659, "step": 20242 }, { "epoch": 1.5864420062695925, "grad_norm": 6.523921012878418, "learning_rate": 6.6132549533338684e-06, "loss": 2.0871, "step": 20243 }, { "epoch": 1.5865203761755486, "grad_norm": 2.9879393577575684, "learning_rate": 6.612963080024757e-06, "loss": 1.3376, "step": 20244 }, { "epoch": 1.5865987460815048, "grad_norm": 2.601501226425171, "learning_rate": 6.612671200580764e-06, "loss": 1.6614, "step": 20245 }, { "epoch": 1.5866771159874609, "grad_norm": 2.2223494052886963, "learning_rate": 6.612379315002997e-06, "loss": 1.4201, "step": 20246 }, { "epoch": 1.586755485893417, "grad_norm": 3.4547481536865234, "learning_rate": 6.612087423292568e-06, "loss": 1.5326, "step": 20247 }, { "epoch": 1.586833855799373, "grad_norm": 3.83905291557312, "learning_rate": 6.6117955254505865e-06, "loss": 1.2219, "step": 20248 }, { "epoch": 1.5869122257053292, "grad_norm": 5.227468013763428, "learning_rate": 6.6115036214781616e-06, "loss": 1.3674, "step": 20249 }, { "epoch": 1.5869905956112853, "grad_norm": 2.940505027770996, "learning_rate": 6.611211711376404e-06, "loss": 1.6113, "step": 20250 }, { "epoch": 1.5870689655172414, "grad_norm": 3.3389151096343994, "learning_rate": 6.610919795146426e-06, "loss": 1.1712, "step": 20251 }, { "epoch": 1.5871473354231975, "grad_norm": 3.4127490520477295, "learning_rate": 6.610627872789337e-06, "loss": 1.3515, "step": 20252 }, { "epoch": 1.5872257053291536, "grad_norm": 3.465424060821533, "learning_rate": 6.610335944306247e-06, "loss": 1.409, "step": 20253 }, { "epoch": 1.5873040752351097, "grad_norm": 2.109576940536499, "learning_rate": 6.6100440096982644e-06, "loss": 1.1083, "step": 20254 }, { "epoch": 1.5873824451410659, "grad_norm": 2.2496283054351807, "learning_rate": 6.609752068966504e-06, "loss": 1.1717, "step": 20255 }, { "epoch": 1.587460815047022, "grad_norm": 5.120632171630859, "learning_rate": 6.609460122112072e-06, "loss": 1.4442, "step": 20256 }, { "epoch": 1.587539184952978, "grad_norm": 3.7629621028900146, "learning_rate": 6.6091681691360795e-06, "loss": 1.2606, "step": 20257 }, { "epoch": 1.5876175548589342, "grad_norm": 9.156678199768066, "learning_rate": 6.608876210039639e-06, "loss": 1.2829, "step": 20258 }, { "epoch": 1.5876959247648903, "grad_norm": 4.505120277404785, "learning_rate": 6.608584244823859e-06, "loss": 1.041, "step": 20259 }, { "epoch": 1.5877742946708464, "grad_norm": 4.161468505859375, "learning_rate": 6.608292273489851e-06, "loss": 1.2608, "step": 20260 }, { "epoch": 1.5878526645768025, "grad_norm": 3.2774319648742676, "learning_rate": 6.608000296038726e-06, "loss": 1.2467, "step": 20261 }, { "epoch": 1.5879310344827586, "grad_norm": 2.5440807342529297, "learning_rate": 6.607708312471593e-06, "loss": 1.3917, "step": 20262 }, { "epoch": 1.5880094043887147, "grad_norm": 3.2670304775238037, "learning_rate": 6.607416322789563e-06, "loss": 1.5323, "step": 20263 }, { "epoch": 1.5880877742946709, "grad_norm": 6.677184581756592, "learning_rate": 6.607124326993748e-06, "loss": 1.3989, "step": 20264 }, { "epoch": 1.588166144200627, "grad_norm": 2.883328437805176, "learning_rate": 6.6068323250852575e-06, "loss": 1.1378, "step": 20265 }, { "epoch": 1.588244514106583, "grad_norm": 3.0402109622955322, "learning_rate": 6.606540317065202e-06, "loss": 1.1542, "step": 20266 }, { "epoch": 1.5883228840125392, "grad_norm": 2.8268449306488037, "learning_rate": 6.606248302934691e-06, "loss": 1.2885, "step": 20267 }, { "epoch": 1.5884012539184953, "grad_norm": 6.24373722076416, "learning_rate": 6.605956282694838e-06, "loss": 1.8532, "step": 20268 }, { "epoch": 1.5884796238244514, "grad_norm": 2.834343194961548, "learning_rate": 6.605664256346751e-06, "loss": 1.3142, "step": 20269 }, { "epoch": 1.5885579937304075, "grad_norm": 7.76966667175293, "learning_rate": 6.605372223891543e-06, "loss": 0.4736, "step": 20270 }, { "epoch": 1.5886363636363636, "grad_norm": 2.1989834308624268, "learning_rate": 6.6050801853303235e-06, "loss": 1.3606, "step": 20271 }, { "epoch": 1.5887147335423197, "grad_norm": 2.3358802795410156, "learning_rate": 6.604788140664203e-06, "loss": 1.4829, "step": 20272 }, { "epoch": 1.5887931034482758, "grad_norm": 3.4185903072357178, "learning_rate": 6.604496089894294e-06, "loss": 1.0451, "step": 20273 }, { "epoch": 1.588871473354232, "grad_norm": 2.7066731452941895, "learning_rate": 6.6042040330217035e-06, "loss": 1.3557, "step": 20274 }, { "epoch": 1.588949843260188, "grad_norm": 2.2820446491241455, "learning_rate": 6.603911970047546e-06, "loss": 1.4963, "step": 20275 }, { "epoch": 1.5890282131661442, "grad_norm": 3.6652793884277344, "learning_rate": 6.60361990097293e-06, "loss": 1.5706, "step": 20276 }, { "epoch": 1.5891065830721003, "grad_norm": 3.4533426761627197, "learning_rate": 6.603327825798969e-06, "loss": 1.4493, "step": 20277 }, { "epoch": 1.5891849529780564, "grad_norm": 3.4680001735687256, "learning_rate": 6.603035744526773e-06, "loss": 1.1601, "step": 20278 }, { "epoch": 1.5892633228840125, "grad_norm": 2.6254546642303467, "learning_rate": 6.602743657157452e-06, "loss": 1.4871, "step": 20279 }, { "epoch": 1.5893416927899686, "grad_norm": 2.358469247817993, "learning_rate": 6.602451563692117e-06, "loss": 1.3917, "step": 20280 }, { "epoch": 1.5894200626959247, "grad_norm": 2.902177095413208, "learning_rate": 6.60215946413188e-06, "loss": 1.3922, "step": 20281 }, { "epoch": 1.5894984326018808, "grad_norm": 2.611945867538452, "learning_rate": 6.601867358477849e-06, "loss": 1.4286, "step": 20282 }, { "epoch": 1.589576802507837, "grad_norm": 15.70847225189209, "learning_rate": 6.60157524673114e-06, "loss": 1.3255, "step": 20283 }, { "epoch": 1.589655172413793, "grad_norm": 2.787834644317627, "learning_rate": 6.601283128892861e-06, "loss": 1.3894, "step": 20284 }, { "epoch": 1.5897335423197492, "grad_norm": 2.5421853065490723, "learning_rate": 6.600991004964123e-06, "loss": 1.2391, "step": 20285 }, { "epoch": 1.5898119122257053, "grad_norm": 2.6220126152038574, "learning_rate": 6.600698874946037e-06, "loss": 1.3391, "step": 20286 }, { "epoch": 1.5898902821316614, "grad_norm": 3.415037155151367, "learning_rate": 6.6004067388397155e-06, "loss": 1.8925, "step": 20287 }, { "epoch": 1.5899686520376175, "grad_norm": 4.26564884185791, "learning_rate": 6.600114596646269e-06, "loss": 1.5183, "step": 20288 }, { "epoch": 1.5900470219435736, "grad_norm": 9.24328899383545, "learning_rate": 6.599822448366806e-06, "loss": 0.8081, "step": 20289 }, { "epoch": 1.5901253918495297, "grad_norm": 10.890756607055664, "learning_rate": 6.599530294002443e-06, "loss": 0.4226, "step": 20290 }, { "epoch": 1.590203761755486, "grad_norm": 2.599773406982422, "learning_rate": 6.599238133554287e-06, "loss": 1.0758, "step": 20291 }, { "epoch": 1.590282131661442, "grad_norm": 2.999995470046997, "learning_rate": 6.598945967023451e-06, "loss": 1.3017, "step": 20292 }, { "epoch": 1.5903605015673983, "grad_norm": 3.7087111473083496, "learning_rate": 6.5986537944110455e-06, "loss": 1.6332, "step": 20293 }, { "epoch": 1.5904388714733542, "grad_norm": 2.2549519538879395, "learning_rate": 6.598361615718183e-06, "loss": 1.5091, "step": 20294 }, { "epoch": 1.5905172413793105, "grad_norm": 2.350966453552246, "learning_rate": 6.598069430945972e-06, "loss": 1.2132, "step": 20295 }, { "epoch": 1.5905956112852664, "grad_norm": 2.6290528774261475, "learning_rate": 6.597777240095526e-06, "loss": 1.4197, "step": 20296 }, { "epoch": 1.5906739811912227, "grad_norm": 2.5804057121276855, "learning_rate": 6.597485043167958e-06, "loss": 1.1275, "step": 20297 }, { "epoch": 1.5907523510971786, "grad_norm": 3.6173911094665527, "learning_rate": 6.597192840164376e-06, "loss": 1.5037, "step": 20298 }, { "epoch": 1.590830721003135, "grad_norm": 2.559811592102051, "learning_rate": 6.596900631085894e-06, "loss": 0.9463, "step": 20299 }, { "epoch": 1.5909090909090908, "grad_norm": 3.635720729827881, "learning_rate": 6.596608415933621e-06, "loss": 1.1549, "step": 20300 }, { "epoch": 1.5909874608150472, "grad_norm": 2.5789361000061035, "learning_rate": 6.596316194708669e-06, "loss": 1.3234, "step": 20301 }, { "epoch": 1.591065830721003, "grad_norm": 4.42014741897583, "learning_rate": 6.596023967412152e-06, "loss": 1.6707, "step": 20302 }, { "epoch": 1.5911442006269594, "grad_norm": 3.100999593734741, "learning_rate": 6.5957317340451786e-06, "loss": 1.1087, "step": 20303 }, { "epoch": 1.5912225705329153, "grad_norm": 6.218932151794434, "learning_rate": 6.595439494608862e-06, "loss": 1.7453, "step": 20304 }, { "epoch": 1.5913009404388716, "grad_norm": 5.283812522888184, "learning_rate": 6.595147249104311e-06, "loss": 1.3117, "step": 20305 }, { "epoch": 1.5913793103448275, "grad_norm": 5.276617527008057, "learning_rate": 6.59485499753264e-06, "loss": 1.2794, "step": 20306 }, { "epoch": 1.5914576802507838, "grad_norm": 7.933109760284424, "learning_rate": 6.594562739894963e-06, "loss": 2.4258, "step": 20307 }, { "epoch": 1.5915360501567397, "grad_norm": 2.7819275856018066, "learning_rate": 6.594270476192384e-06, "loss": 1.3812, "step": 20308 }, { "epoch": 1.591614420062696, "grad_norm": 2.3268041610717773, "learning_rate": 6.593978206426022e-06, "loss": 1.4332, "step": 20309 }, { "epoch": 1.591692789968652, "grad_norm": 3.0476503372192383, "learning_rate": 6.5936859305969825e-06, "loss": 1.1011, "step": 20310 }, { "epoch": 1.5917711598746083, "grad_norm": 3.1463310718536377, "learning_rate": 6.593393648706383e-06, "loss": 1.6033, "step": 20311 }, { "epoch": 1.5918495297805642, "grad_norm": 3.2566444873809814, "learning_rate": 6.593101360755332e-06, "loss": 1.4986, "step": 20312 }, { "epoch": 1.5919278996865205, "grad_norm": 3.207908868789673, "learning_rate": 6.59280906674494e-06, "loss": 1.4322, "step": 20313 }, { "epoch": 1.5920062695924764, "grad_norm": 3.004103183746338, "learning_rate": 6.592516766676324e-06, "loss": 1.0833, "step": 20314 }, { "epoch": 1.5920846394984327, "grad_norm": 3.492830276489258, "learning_rate": 6.592224460550588e-06, "loss": 1.4373, "step": 20315 }, { "epoch": 1.5921630094043886, "grad_norm": 2.571131944656372, "learning_rate": 6.591932148368851e-06, "loss": 1.1831, "step": 20316 }, { "epoch": 1.592241379310345, "grad_norm": 2.6118557453155518, "learning_rate": 6.59163983013222e-06, "loss": 1.3408, "step": 20317 }, { "epoch": 1.5923197492163008, "grad_norm": 3.4442696571350098, "learning_rate": 6.59134750584181e-06, "loss": 1.0289, "step": 20318 }, { "epoch": 1.5923981191222571, "grad_norm": 2.7042698860168457, "learning_rate": 6.5910551754987305e-06, "loss": 1.1565, "step": 20319 }, { "epoch": 1.592476489028213, "grad_norm": 3.294293165206909, "learning_rate": 6.5907628391040945e-06, "loss": 1.8582, "step": 20320 }, { "epoch": 1.5925548589341694, "grad_norm": 2.9088058471679688, "learning_rate": 6.590470496659012e-06, "loss": 1.4023, "step": 20321 }, { "epoch": 1.5926332288401253, "grad_norm": 5.816065311431885, "learning_rate": 6.590178148164599e-06, "loss": 0.6162, "step": 20322 }, { "epoch": 1.5927115987460816, "grad_norm": 3.4472568035125732, "learning_rate": 6.589885793621965e-06, "loss": 1.7044, "step": 20323 }, { "epoch": 1.5927899686520375, "grad_norm": 7.026486873626709, "learning_rate": 6.589593433032221e-06, "loss": 1.3868, "step": 20324 }, { "epoch": 1.5928683385579938, "grad_norm": 5.084288597106934, "learning_rate": 6.589301066396481e-06, "loss": 1.1758, "step": 20325 }, { "epoch": 1.5929467084639497, "grad_norm": 3.4105026721954346, "learning_rate": 6.589008693715855e-06, "loss": 1.4875, "step": 20326 }, { "epoch": 1.593025078369906, "grad_norm": 1.9558639526367188, "learning_rate": 6.588716314991458e-06, "loss": 1.3702, "step": 20327 }, { "epoch": 1.593103448275862, "grad_norm": 3.415365695953369, "learning_rate": 6.588423930224399e-06, "loss": 1.4083, "step": 20328 }, { "epoch": 1.5931818181818183, "grad_norm": 4.023480415344238, "learning_rate": 6.5881315394157905e-06, "loss": 1.5425, "step": 20329 }, { "epoch": 1.5932601880877741, "grad_norm": 2.3071136474609375, "learning_rate": 6.587839142566745e-06, "loss": 1.231, "step": 20330 }, { "epoch": 1.5933385579937305, "grad_norm": 2.493318796157837, "learning_rate": 6.587546739678376e-06, "loss": 1.5141, "step": 20331 }, { "epoch": 1.5934169278996864, "grad_norm": 3.536134719848633, "learning_rate": 6.587254330751796e-06, "loss": 1.0372, "step": 20332 }, { "epoch": 1.5934952978056427, "grad_norm": 2.9697978496551514, "learning_rate": 6.586961915788114e-06, "loss": 1.5526, "step": 20333 }, { "epoch": 1.5935736677115986, "grad_norm": 3.5156052112579346, "learning_rate": 6.586669494788443e-06, "loss": 1.3702, "step": 20334 }, { "epoch": 1.593652037617555, "grad_norm": 2.784640312194824, "learning_rate": 6.586377067753899e-06, "loss": 1.3121, "step": 20335 }, { "epoch": 1.5937304075235108, "grad_norm": 3.427539587020874, "learning_rate": 6.586084634685589e-06, "loss": 1.6075, "step": 20336 }, { "epoch": 1.5938087774294671, "grad_norm": 2.9735653400421143, "learning_rate": 6.585792195584629e-06, "loss": 0.8337, "step": 20337 }, { "epoch": 1.5938871473354232, "grad_norm": 2.4844186305999756, "learning_rate": 6.585499750452129e-06, "loss": 1.2653, "step": 20338 }, { "epoch": 1.5939655172413794, "grad_norm": 3.1118247509002686, "learning_rate": 6.5852072992892045e-06, "loss": 1.2582, "step": 20339 }, { "epoch": 1.5940438871473355, "grad_norm": 2.4943349361419678, "learning_rate": 6.584914842096963e-06, "loss": 1.2731, "step": 20340 }, { "epoch": 1.5941222570532916, "grad_norm": 6.517461776733398, "learning_rate": 6.584622378876522e-06, "loss": 1.6246, "step": 20341 }, { "epoch": 1.5942006269592477, "grad_norm": 3.3613617420196533, "learning_rate": 6.58432990962899e-06, "loss": 1.2136, "step": 20342 }, { "epoch": 1.5942789968652038, "grad_norm": 13.491312980651855, "learning_rate": 6.58403743435548e-06, "loss": 1.4229, "step": 20343 }, { "epoch": 1.59435736677116, "grad_norm": 3.4511308670043945, "learning_rate": 6.583744953057107e-06, "loss": 1.2409, "step": 20344 }, { "epoch": 1.594435736677116, "grad_norm": 5.115262985229492, "learning_rate": 6.583452465734979e-06, "loss": 1.3887, "step": 20345 }, { "epoch": 1.5945141065830721, "grad_norm": 4.3928046226501465, "learning_rate": 6.583159972390214e-06, "loss": 1.3488, "step": 20346 }, { "epoch": 1.5945924764890282, "grad_norm": 2.8200180530548096, "learning_rate": 6.58286747302392e-06, "loss": 1.1746, "step": 20347 }, { "epoch": 1.5946708463949844, "grad_norm": 2.5815582275390625, "learning_rate": 6.582574967637212e-06, "loss": 1.1838, "step": 20348 }, { "epoch": 1.5947492163009405, "grad_norm": 3.298797369003296, "learning_rate": 6.582282456231201e-06, "loss": 1.3313, "step": 20349 }, { "epoch": 1.5948275862068966, "grad_norm": 3.88710880279541, "learning_rate": 6.581989938807001e-06, "loss": 1.3664, "step": 20350 }, { "epoch": 1.5949059561128527, "grad_norm": 2.9226419925689697, "learning_rate": 6.581697415365723e-06, "loss": 1.1038, "step": 20351 }, { "epoch": 1.5949843260188088, "grad_norm": 5.387098789215088, "learning_rate": 6.581404885908481e-06, "loss": 1.2029, "step": 20352 }, { "epoch": 1.595062695924765, "grad_norm": 2.3727850914001465, "learning_rate": 6.581112350436387e-06, "loss": 1.4997, "step": 20353 }, { "epoch": 1.595141065830721, "grad_norm": 2.368767738342285, "learning_rate": 6.5808198089505535e-06, "loss": 1.4164, "step": 20354 }, { "epoch": 1.5952194357366771, "grad_norm": 3.301959276199341, "learning_rate": 6.580527261452094e-06, "loss": 1.6218, "step": 20355 }, { "epoch": 1.5952978056426332, "grad_norm": 2.1103551387786865, "learning_rate": 6.58023470794212e-06, "loss": 1.1698, "step": 20356 }, { "epoch": 1.5953761755485893, "grad_norm": 2.8929383754730225, "learning_rate": 6.579942148421745e-06, "loss": 1.1635, "step": 20357 }, { "epoch": 1.5954545454545455, "grad_norm": 3.834820508956909, "learning_rate": 6.579649582892082e-06, "loss": 1.1386, "step": 20358 }, { "epoch": 1.5955329153605016, "grad_norm": 3.5625505447387695, "learning_rate": 6.579357011354243e-06, "loss": 1.0585, "step": 20359 }, { "epoch": 1.5956112852664577, "grad_norm": 4.350480079650879, "learning_rate": 6.57906443380934e-06, "loss": 1.2936, "step": 20360 }, { "epoch": 1.5956896551724138, "grad_norm": 2.170128345489502, "learning_rate": 6.57877185025849e-06, "loss": 1.2207, "step": 20361 }, { "epoch": 1.59576802507837, "grad_norm": 2.5836362838745117, "learning_rate": 6.578479260702799e-06, "loss": 1.5527, "step": 20362 }, { "epoch": 1.595846394984326, "grad_norm": 2.3516838550567627, "learning_rate": 6.578186665143387e-06, "loss": 1.2241, "step": 20363 }, { "epoch": 1.5959247648902821, "grad_norm": 2.8110954761505127, "learning_rate": 6.5778940635813605e-06, "loss": 1.4074, "step": 20364 }, { "epoch": 1.5960031347962382, "grad_norm": 2.6548595428466797, "learning_rate": 6.577601456017837e-06, "loss": 1.0802, "step": 20365 }, { "epoch": 1.5960815047021943, "grad_norm": 2.5536811351776123, "learning_rate": 6.577308842453928e-06, "loss": 1.1732, "step": 20366 }, { "epoch": 1.5961598746081505, "grad_norm": 2.745652675628662, "learning_rate": 6.577016222890746e-06, "loss": 1.2277, "step": 20367 }, { "epoch": 1.5962382445141066, "grad_norm": 2.86782169342041, "learning_rate": 6.576723597329404e-06, "loss": 1.133, "step": 20368 }, { "epoch": 1.5963166144200627, "grad_norm": 7.1098551750183105, "learning_rate": 6.576430965771016e-06, "loss": 1.74, "step": 20369 }, { "epoch": 1.5963949843260188, "grad_norm": 4.591586589813232, "learning_rate": 6.5761383282166945e-06, "loss": 1.102, "step": 20370 }, { "epoch": 1.596473354231975, "grad_norm": 2.1661012172698975, "learning_rate": 6.575845684667551e-06, "loss": 1.2418, "step": 20371 }, { "epoch": 1.596551724137931, "grad_norm": 3.738727569580078, "learning_rate": 6.575553035124702e-06, "loss": 1.1394, "step": 20372 }, { "epoch": 1.5966300940438871, "grad_norm": 7.9936394691467285, "learning_rate": 6.575260379589257e-06, "loss": 1.2691, "step": 20373 }, { "epoch": 1.5967084639498432, "grad_norm": 3.973041296005249, "learning_rate": 6.57496771806233e-06, "loss": 1.1072, "step": 20374 }, { "epoch": 1.5967868338557993, "grad_norm": 2.020277976989746, "learning_rate": 6.574675050545036e-06, "loss": 1.4432, "step": 20375 }, { "epoch": 1.5968652037617554, "grad_norm": 4.166606426239014, "learning_rate": 6.574382377038487e-06, "loss": 1.486, "step": 20376 }, { "epoch": 1.5969435736677116, "grad_norm": 4.2095561027526855, "learning_rate": 6.574089697543795e-06, "loss": 1.4494, "step": 20377 }, { "epoch": 1.5970219435736677, "grad_norm": 2.2641940116882324, "learning_rate": 6.573797012062075e-06, "loss": 1.1198, "step": 20378 }, { "epoch": 1.5971003134796238, "grad_norm": 3.551724672317505, "learning_rate": 6.5735043205944395e-06, "loss": 1.1464, "step": 20379 }, { "epoch": 1.59717868338558, "grad_norm": 6.40797233581543, "learning_rate": 6.573211623142002e-06, "loss": 1.3534, "step": 20380 }, { "epoch": 1.597257053291536, "grad_norm": 2.1665854454040527, "learning_rate": 6.572918919705875e-06, "loss": 1.249, "step": 20381 }, { "epoch": 1.597335423197492, "grad_norm": 3.2368125915527344, "learning_rate": 6.572626210287174e-06, "loss": 1.5456, "step": 20382 }, { "epoch": 1.5974137931034482, "grad_norm": 3.4716086387634277, "learning_rate": 6.572333494887009e-06, "loss": 1.8512, "step": 20383 }, { "epoch": 1.5974921630094046, "grad_norm": 2.575340509414673, "learning_rate": 6.572040773506495e-06, "loss": 1.0817, "step": 20384 }, { "epoch": 1.5975705329153604, "grad_norm": 3.208371877670288, "learning_rate": 6.571748046146745e-06, "loss": 1.1587, "step": 20385 }, { "epoch": 1.5976489028213168, "grad_norm": 2.8751397132873535, "learning_rate": 6.5714553128088735e-06, "loss": 1.2402, "step": 20386 }, { "epoch": 1.5977272727272727, "grad_norm": 3.129242420196533, "learning_rate": 6.571162573493993e-06, "loss": 1.5058, "step": 20387 }, { "epoch": 1.597805642633229, "grad_norm": 2.7199785709381104, "learning_rate": 6.570869828203217e-06, "loss": 1.0161, "step": 20388 }, { "epoch": 1.5978840125391849, "grad_norm": 2.3134477138519287, "learning_rate": 6.57057707693766e-06, "loss": 1.2674, "step": 20389 }, { "epoch": 1.5979623824451412, "grad_norm": 2.312692403793335, "learning_rate": 6.570284319698432e-06, "loss": 1.4202, "step": 20390 }, { "epoch": 1.598040752351097, "grad_norm": 4.909838676452637, "learning_rate": 6.569991556486653e-06, "loss": 1.2094, "step": 20391 }, { "epoch": 1.5981191222570534, "grad_norm": 3.25127911567688, "learning_rate": 6.569698787303428e-06, "loss": 1.3348, "step": 20392 }, { "epoch": 1.5981974921630093, "grad_norm": 3.449458599090576, "learning_rate": 6.569406012149877e-06, "loss": 1.3019, "step": 20393 }, { "epoch": 1.5982758620689657, "grad_norm": 5.391209602355957, "learning_rate": 6.569113231027112e-06, "loss": 1.2871, "step": 20394 }, { "epoch": 1.5983542319749215, "grad_norm": 6.246533393859863, "learning_rate": 6.568820443936246e-06, "loss": 1.3221, "step": 20395 }, { "epoch": 1.5984326018808779, "grad_norm": 2.8684229850769043, "learning_rate": 6.5685276508783935e-06, "loss": 0.9301, "step": 20396 }, { "epoch": 1.5985109717868338, "grad_norm": 2.9043843746185303, "learning_rate": 6.568234851854666e-06, "loss": 1.3715, "step": 20397 }, { "epoch": 1.59858934169279, "grad_norm": 2.113544225692749, "learning_rate": 6.567942046866181e-06, "loss": 1.2581, "step": 20398 }, { "epoch": 1.598667711598746, "grad_norm": 2.3184072971343994, "learning_rate": 6.567649235914046e-06, "loss": 0.9335, "step": 20399 }, { "epoch": 1.5987460815047023, "grad_norm": 4.871512413024902, "learning_rate": 6.567356418999382e-06, "loss": 0.9484, "step": 20400 }, { "epoch": 1.5988244514106582, "grad_norm": 2.1245498657226562, "learning_rate": 6.567063596123297e-06, "loss": 1.2752, "step": 20401 }, { "epoch": 1.5989028213166145, "grad_norm": 2.852027177810669, "learning_rate": 6.566770767286909e-06, "loss": 1.4962, "step": 20402 }, { "epoch": 1.5989811912225704, "grad_norm": 5.711973190307617, "learning_rate": 6.566477932491329e-06, "loss": 1.7046, "step": 20403 }, { "epoch": 1.5990595611285268, "grad_norm": 2.5000243186950684, "learning_rate": 6.56618509173767e-06, "loss": 1.1046, "step": 20404 }, { "epoch": 1.5991379310344827, "grad_norm": 3.577610492706299, "learning_rate": 6.565892245027049e-06, "loss": 1.5473, "step": 20405 }, { "epoch": 1.599216300940439, "grad_norm": 2.3409924507141113, "learning_rate": 6.5655993923605775e-06, "loss": 1.4508, "step": 20406 }, { "epoch": 1.5992946708463949, "grad_norm": 2.9924590587615967, "learning_rate": 6.56530653373937e-06, "loss": 1.2639, "step": 20407 }, { "epoch": 1.5993730407523512, "grad_norm": 4.295770645141602, "learning_rate": 6.565013669164541e-06, "loss": 1.5571, "step": 20408 }, { "epoch": 1.599451410658307, "grad_norm": 3.192549705505371, "learning_rate": 6.564720798637203e-06, "loss": 1.474, "step": 20409 }, { "epoch": 1.5995297805642634, "grad_norm": 2.253486156463623, "learning_rate": 6.564427922158472e-06, "loss": 1.3717, "step": 20410 }, { "epoch": 1.5996081504702193, "grad_norm": 2.685194730758667, "learning_rate": 6.56413503972946e-06, "loss": 0.7947, "step": 20411 }, { "epoch": 1.5996865203761756, "grad_norm": 3.3875930309295654, "learning_rate": 6.563842151351281e-06, "loss": 1.4206, "step": 20412 }, { "epoch": 1.5997648902821315, "grad_norm": 2.1742212772369385, "learning_rate": 6.563549257025051e-06, "loss": 1.5793, "step": 20413 }, { "epoch": 1.5998432601880879, "grad_norm": 3.2373158931732178, "learning_rate": 6.563256356751882e-06, "loss": 0.9503, "step": 20414 }, { "epoch": 1.5999216300940438, "grad_norm": 3.332965850830078, "learning_rate": 6.562963450532891e-06, "loss": 1.4891, "step": 20415 }, { "epoch": 1.6, "grad_norm": 1.8261489868164062, "learning_rate": 6.562670538369186e-06, "loss": 0.858, "step": 20416 }, { "epoch": 1.600078369905956, "grad_norm": 5.648465633392334, "learning_rate": 6.562377620261887e-06, "loss": 1.5122, "step": 20417 }, { "epoch": 1.6001567398119123, "grad_norm": 3.3411827087402344, "learning_rate": 6.562084696212105e-06, "loss": 1.6472, "step": 20418 }, { "epoch": 1.6002351097178682, "grad_norm": 3.8236114978790283, "learning_rate": 6.5617917662209574e-06, "loss": 1.6377, "step": 20419 }, { "epoch": 1.6003134796238245, "grad_norm": 3.041069984436035, "learning_rate": 6.561498830289554e-06, "loss": 1.322, "step": 20420 }, { "epoch": 1.6003918495297804, "grad_norm": 2.8084957599639893, "learning_rate": 6.561205888419011e-06, "loss": 1.1739, "step": 20421 }, { "epoch": 1.6004702194357368, "grad_norm": 4.442378997802734, "learning_rate": 6.560912940610443e-06, "loss": 1.1016, "step": 20422 }, { "epoch": 1.6005485893416926, "grad_norm": 2.2163915634155273, "learning_rate": 6.560619986864964e-06, "loss": 1.4191, "step": 20423 }, { "epoch": 1.600626959247649, "grad_norm": 2.6365058422088623, "learning_rate": 6.5603270271836905e-06, "loss": 1.1061, "step": 20424 }, { "epoch": 1.6007053291536049, "grad_norm": 3.588143825531006, "learning_rate": 6.56003406156773e-06, "loss": 1.6987, "step": 20425 }, { "epoch": 1.6007836990595612, "grad_norm": 4.172966957092285, "learning_rate": 6.559741090018206e-06, "loss": 1.4347, "step": 20426 }, { "epoch": 1.600862068965517, "grad_norm": 5.08535623550415, "learning_rate": 6.559448112536224e-06, "loss": 1.2711, "step": 20427 }, { "epoch": 1.6009404388714734, "grad_norm": 2.5113863945007324, "learning_rate": 6.559155129122905e-06, "loss": 0.969, "step": 20428 }, { "epoch": 1.6010188087774293, "grad_norm": 3.2333600521087646, "learning_rate": 6.558862139779358e-06, "loss": 1.8409, "step": 20429 }, { "epoch": 1.6010971786833856, "grad_norm": 4.192272663116455, "learning_rate": 6.558569144506702e-06, "loss": 0.9268, "step": 20430 }, { "epoch": 1.6011755485893417, "grad_norm": 2.430800199508667, "learning_rate": 6.558276143306049e-06, "loss": 1.0913, "step": 20431 }, { "epoch": 1.6012539184952979, "grad_norm": 3.2680869102478027, "learning_rate": 6.557983136178515e-06, "loss": 1.4631, "step": 20432 }, { "epoch": 1.601332288401254, "grad_norm": 2.548340320587158, "learning_rate": 6.557690123125212e-06, "loss": 1.4369, "step": 20433 }, { "epoch": 1.60141065830721, "grad_norm": 3.6843371391296387, "learning_rate": 6.5573971041472564e-06, "loss": 1.5282, "step": 20434 }, { "epoch": 1.6014890282131662, "grad_norm": 3.214481830596924, "learning_rate": 6.557104079245762e-06, "loss": 1.4226, "step": 20435 }, { "epoch": 1.6015673981191223, "grad_norm": 4.065465450286865, "learning_rate": 6.556811048421845e-06, "loss": 1.3033, "step": 20436 }, { "epoch": 1.6016457680250784, "grad_norm": 3.9962716102600098, "learning_rate": 6.556518011676617e-06, "loss": 1.0245, "step": 20437 }, { "epoch": 1.6017241379310345, "grad_norm": 5.136277675628662, "learning_rate": 6.556224969011194e-06, "loss": 1.3364, "step": 20438 }, { "epoch": 1.6018025078369906, "grad_norm": 2.292729139328003, "learning_rate": 6.55593192042669e-06, "loss": 1.4896, "step": 20439 }, { "epoch": 1.6018808777429467, "grad_norm": 3.3249623775482178, "learning_rate": 6.555638865924221e-06, "loss": 1.1227, "step": 20440 }, { "epoch": 1.6019592476489029, "grad_norm": 3.330888271331787, "learning_rate": 6.5553458055049e-06, "loss": 1.4032, "step": 20441 }, { "epoch": 1.602037617554859, "grad_norm": 2.486952781677246, "learning_rate": 6.555052739169843e-06, "loss": 0.964, "step": 20442 }, { "epoch": 1.602115987460815, "grad_norm": 2.520038604736328, "learning_rate": 6.554759666920165e-06, "loss": 1.1107, "step": 20443 }, { "epoch": 1.6021943573667712, "grad_norm": 5.918888092041016, "learning_rate": 6.554466588756977e-06, "loss": 1.3856, "step": 20444 }, { "epoch": 1.6022727272727273, "grad_norm": 4.507502555847168, "learning_rate": 6.554173504681399e-06, "loss": 0.7432, "step": 20445 }, { "epoch": 1.6023510971786834, "grad_norm": 2.8349578380584717, "learning_rate": 6.553880414694542e-06, "loss": 1.3276, "step": 20446 }, { "epoch": 1.6024294670846395, "grad_norm": 3.6941330432891846, "learning_rate": 6.553587318797523e-06, "loss": 1.3647, "step": 20447 }, { "epoch": 1.6025078369905956, "grad_norm": 3.249868154525757, "learning_rate": 6.553294216991456e-06, "loss": 1.4087, "step": 20448 }, { "epoch": 1.6025862068965517, "grad_norm": 1.951720118522644, "learning_rate": 6.553001109277454e-06, "loss": 1.4448, "step": 20449 }, { "epoch": 1.6026645768025078, "grad_norm": 5.788397312164307, "learning_rate": 6.5527079956566355e-06, "loss": 1.8449, "step": 20450 }, { "epoch": 1.602742946708464, "grad_norm": 2.416837692260742, "learning_rate": 6.552414876130112e-06, "loss": 1.2238, "step": 20451 }, { "epoch": 1.60282131661442, "grad_norm": 2.33414626121521, "learning_rate": 6.552121750699001e-06, "loss": 1.6901, "step": 20452 }, { "epoch": 1.6028996865203762, "grad_norm": 4.3213276863098145, "learning_rate": 6.551828619364413e-06, "loss": 1.5291, "step": 20453 }, { "epoch": 1.6029780564263323, "grad_norm": 2.082754611968994, "learning_rate": 6.551535482127469e-06, "loss": 1.3001, "step": 20454 }, { "epoch": 1.6030564263322884, "grad_norm": 5.065915107727051, "learning_rate": 6.551242338989281e-06, "loss": 1.349, "step": 20455 }, { "epoch": 1.6031347962382445, "grad_norm": 3.2866501808166504, "learning_rate": 6.550949189950963e-06, "loss": 1.8491, "step": 20456 }, { "epoch": 1.6032131661442006, "grad_norm": 3.4374680519104004, "learning_rate": 6.550656035013629e-06, "loss": 1.0566, "step": 20457 }, { "epoch": 1.6032915360501567, "grad_norm": 1.9443389177322388, "learning_rate": 6.5503628741783985e-06, "loss": 0.994, "step": 20458 }, { "epoch": 1.6033699059561128, "grad_norm": 2.6736438274383545, "learning_rate": 6.550069707446383e-06, "loss": 1.4282, "step": 20459 }, { "epoch": 1.603448275862069, "grad_norm": 3.3510894775390625, "learning_rate": 6.549776534818699e-06, "loss": 1.2173, "step": 20460 }, { "epoch": 1.603526645768025, "grad_norm": 4.02199125289917, "learning_rate": 6.54948335629646e-06, "loss": 1.3418, "step": 20461 }, { "epoch": 1.6036050156739812, "grad_norm": 2.3520596027374268, "learning_rate": 6.549190171880784e-06, "loss": 1.2411, "step": 20462 }, { "epoch": 1.6036833855799373, "grad_norm": 4.67193603515625, "learning_rate": 6.548896981572782e-06, "loss": 0.7487, "step": 20463 }, { "epoch": 1.6037617554858934, "grad_norm": 1.9426263570785522, "learning_rate": 6.548603785373574e-06, "loss": 1.5375, "step": 20464 }, { "epoch": 1.6038401253918495, "grad_norm": 3.319369077682495, "learning_rate": 6.548310583284271e-06, "loss": 1.6087, "step": 20465 }, { "epoch": 1.6039184952978056, "grad_norm": 4.390105247497559, "learning_rate": 6.5480173753059905e-06, "loss": 1.4093, "step": 20466 }, { "epoch": 1.6039968652037617, "grad_norm": 2.9018490314483643, "learning_rate": 6.5477241614398455e-06, "loss": 1.7417, "step": 20467 }, { "epoch": 1.6040752351097178, "grad_norm": 4.792440891265869, "learning_rate": 6.547430941686954e-06, "loss": 0.8418, "step": 20468 }, { "epoch": 1.604153605015674, "grad_norm": 2.210937023162842, "learning_rate": 6.54713771604843e-06, "loss": 1.2679, "step": 20469 }, { "epoch": 1.60423197492163, "grad_norm": 2.216834783554077, "learning_rate": 6.546844484525389e-06, "loss": 1.1454, "step": 20470 }, { "epoch": 1.6043103448275862, "grad_norm": 3.225260019302368, "learning_rate": 6.546551247118946e-06, "loss": 1.4519, "step": 20471 }, { "epoch": 1.6043887147335423, "grad_norm": 3.1199357509613037, "learning_rate": 6.546258003830215e-06, "loss": 1.5737, "step": 20472 }, { "epoch": 1.6044670846394984, "grad_norm": 2.5610415935516357, "learning_rate": 6.545964754660314e-06, "loss": 1.3898, "step": 20473 }, { "epoch": 1.6045454545454545, "grad_norm": 2.9185893535614014, "learning_rate": 6.5456714996103575e-06, "loss": 1.4609, "step": 20474 }, { "epoch": 1.6046238244514106, "grad_norm": 11.161867141723633, "learning_rate": 6.54537823868146e-06, "loss": 2.0432, "step": 20475 }, { "epoch": 1.6047021943573667, "grad_norm": 3.1131906509399414, "learning_rate": 6.545084971874738e-06, "loss": 1.2335, "step": 20476 }, { "epoch": 1.604780564263323, "grad_norm": 2.9333698749542236, "learning_rate": 6.544791699191307e-06, "loss": 1.4585, "step": 20477 }, { "epoch": 1.604858934169279, "grad_norm": 3.859267473220825, "learning_rate": 6.544498420632279e-06, "loss": 1.5904, "step": 20478 }, { "epoch": 1.6049373040752353, "grad_norm": 3.641453742980957, "learning_rate": 6.544205136198773e-06, "loss": 1.6899, "step": 20479 }, { "epoch": 1.6050156739811912, "grad_norm": 7.010354042053223, "learning_rate": 6.543911845891907e-06, "loss": 1.4735, "step": 20480 }, { "epoch": 1.6050940438871475, "grad_norm": 2.298098087310791, "learning_rate": 6.54361854971279e-06, "loss": 1.4375, "step": 20481 }, { "epoch": 1.6051724137931034, "grad_norm": 1.9706159830093384, "learning_rate": 6.543325247662543e-06, "loss": 1.3174, "step": 20482 }, { "epoch": 1.6052507836990597, "grad_norm": 1.826998233795166, "learning_rate": 6.5430319397422785e-06, "loss": 0.8277, "step": 20483 }, { "epoch": 1.6053291536050156, "grad_norm": 2.997347831726074, "learning_rate": 6.5427386259531135e-06, "loss": 1.4752, "step": 20484 }, { "epoch": 1.605407523510972, "grad_norm": 2.942066192626953, "learning_rate": 6.542445306296163e-06, "loss": 1.2308, "step": 20485 }, { "epoch": 1.6054858934169278, "grad_norm": 3.7120137214660645, "learning_rate": 6.542151980772542e-06, "loss": 1.5516, "step": 20486 }, { "epoch": 1.6055642633228842, "grad_norm": 2.2627716064453125, "learning_rate": 6.5418586493833665e-06, "loss": 1.1681, "step": 20487 }, { "epoch": 1.60564263322884, "grad_norm": 3.2208385467529297, "learning_rate": 6.541565312129754e-06, "loss": 0.9623, "step": 20488 }, { "epoch": 1.6057210031347964, "grad_norm": 2.6805758476257324, "learning_rate": 6.541271969012819e-06, "loss": 1.2059, "step": 20489 }, { "epoch": 1.6057993730407523, "grad_norm": 5.185912609100342, "learning_rate": 6.540978620033677e-06, "loss": 1.1683, "step": 20490 }, { "epoch": 1.6058777429467086, "grad_norm": 5.348283290863037, "learning_rate": 6.540685265193443e-06, "loss": 1.2299, "step": 20491 }, { "epoch": 1.6059561128526645, "grad_norm": 2.429433584213257, "learning_rate": 6.5403919044932344e-06, "loss": 1.4302, "step": 20492 }, { "epoch": 1.6060344827586208, "grad_norm": 2.8720216751098633, "learning_rate": 6.540098537934165e-06, "loss": 1.484, "step": 20493 }, { "epoch": 1.6061128526645767, "grad_norm": 11.404552459716797, "learning_rate": 6.539805165517352e-06, "loss": 1.2555, "step": 20494 }, { "epoch": 1.606191222570533, "grad_norm": 4.1568803787231445, "learning_rate": 6.539511787243913e-06, "loss": 0.9639, "step": 20495 }, { "epoch": 1.606269592476489, "grad_norm": 2.6882593631744385, "learning_rate": 6.53921840311496e-06, "loss": 1.2783, "step": 20496 }, { "epoch": 1.6063479623824453, "grad_norm": 2.236147403717041, "learning_rate": 6.538925013131612e-06, "loss": 1.2619, "step": 20497 }, { "epoch": 1.6064263322884011, "grad_norm": 2.9322266578674316, "learning_rate": 6.538631617294981e-06, "loss": 1.4429, "step": 20498 }, { "epoch": 1.6065047021943575, "grad_norm": 8.483556747436523, "learning_rate": 6.53833821560619e-06, "loss": 1.11, "step": 20499 }, { "epoch": 1.6065830721003134, "grad_norm": 3.4706926345825195, "learning_rate": 6.538044808066346e-06, "loss": 1.6315, "step": 20500 }, { "epoch": 1.6066614420062697, "grad_norm": 2.894517183303833, "learning_rate": 6.5377513946765735e-06, "loss": 1.869, "step": 20501 }, { "epoch": 1.6067398119122256, "grad_norm": 2.719844102859497, "learning_rate": 6.537457975437983e-06, "loss": 0.9354, "step": 20502 }, { "epoch": 1.606818181818182, "grad_norm": 2.3995652198791504, "learning_rate": 6.537164550351691e-06, "loss": 1.5442, "step": 20503 }, { "epoch": 1.6068965517241378, "grad_norm": 3.066329002380371, "learning_rate": 6.536871119418816e-06, "loss": 1.3254, "step": 20504 }, { "epoch": 1.6069749216300941, "grad_norm": 3.621267557144165, "learning_rate": 6.536577682640472e-06, "loss": 1.7785, "step": 20505 }, { "epoch": 1.60705329153605, "grad_norm": 2.9306156635284424, "learning_rate": 6.536284240017775e-06, "loss": 1.9619, "step": 20506 }, { "epoch": 1.6071316614420064, "grad_norm": 2.5589499473571777, "learning_rate": 6.535990791551843e-06, "loss": 1.0552, "step": 20507 }, { "epoch": 1.6072100313479623, "grad_norm": 2.7773070335388184, "learning_rate": 6.535697337243791e-06, "loss": 1.5266, "step": 20508 }, { "epoch": 1.6072884012539186, "grad_norm": 2.140899419784546, "learning_rate": 6.535403877094734e-06, "loss": 1.1091, "step": 20509 }, { "epoch": 1.6073667711598745, "grad_norm": 4.999030590057373, "learning_rate": 6.535110411105792e-06, "loss": 1.8506, "step": 20510 }, { "epoch": 1.6074451410658308, "grad_norm": 3.196087121963501, "learning_rate": 6.534816939278076e-06, "loss": 1.1009, "step": 20511 }, { "epoch": 1.6075235109717867, "grad_norm": 2.575993537902832, "learning_rate": 6.534523461612705e-06, "loss": 1.3053, "step": 20512 }, { "epoch": 1.607601880877743, "grad_norm": 3.412975311279297, "learning_rate": 6.534229978110795e-06, "loss": 1.4675, "step": 20513 }, { "epoch": 1.607680250783699, "grad_norm": 2.1025893688201904, "learning_rate": 6.533936488773463e-06, "loss": 1.4134, "step": 20514 }, { "epoch": 1.6077586206896552, "grad_norm": 2.2996652126312256, "learning_rate": 6.533642993601823e-06, "loss": 1.254, "step": 20515 }, { "epoch": 1.6078369905956111, "grad_norm": 2.9462969303131104, "learning_rate": 6.533349492596994e-06, "loss": 1.4755, "step": 20516 }, { "epoch": 1.6079153605015675, "grad_norm": 5.434114456176758, "learning_rate": 6.533055985760089e-06, "loss": 1.536, "step": 20517 }, { "epoch": 1.6079937304075234, "grad_norm": 2.141427993774414, "learning_rate": 6.532762473092229e-06, "loss": 1.4551, "step": 20518 }, { "epoch": 1.6080721003134797, "grad_norm": 25.11931037902832, "learning_rate": 6.532468954594526e-06, "loss": 2.7887, "step": 20519 }, { "epoch": 1.6081504702194356, "grad_norm": 2.0691139698028564, "learning_rate": 6.5321754302681e-06, "loss": 1.3667, "step": 20520 }, { "epoch": 1.608228840125392, "grad_norm": 3.021650791168213, "learning_rate": 6.531881900114066e-06, "loss": 1.2788, "step": 20521 }, { "epoch": 1.6083072100313478, "grad_norm": 2.3677151203155518, "learning_rate": 6.531588364133539e-06, "loss": 1.212, "step": 20522 }, { "epoch": 1.6083855799373041, "grad_norm": 2.53163480758667, "learning_rate": 6.531294822327635e-06, "loss": 1.3281, "step": 20523 }, { "epoch": 1.6084639498432602, "grad_norm": 1.9281631708145142, "learning_rate": 6.531001274697474e-06, "loss": 1.1674, "step": 20524 }, { "epoch": 1.6085423197492164, "grad_norm": 5.522040367126465, "learning_rate": 6.53070772124417e-06, "loss": 1.5779, "step": 20525 }, { "epoch": 1.6086206896551725, "grad_norm": 3.5065274238586426, "learning_rate": 6.53041416196884e-06, "loss": 1.1427, "step": 20526 }, { "epoch": 1.6086990595611286, "grad_norm": 3.0538110733032227, "learning_rate": 6.530120596872601e-06, "loss": 1.367, "step": 20527 }, { "epoch": 1.6087774294670847, "grad_norm": 7.936227798461914, "learning_rate": 6.5298270259565685e-06, "loss": 1.1382, "step": 20528 }, { "epoch": 1.6088557993730408, "grad_norm": 2.1092023849487305, "learning_rate": 6.529533449221861e-06, "loss": 1.3866, "step": 20529 }, { "epoch": 1.608934169278997, "grad_norm": 3.569117307662964, "learning_rate": 6.529239866669592e-06, "loss": 1.7871, "step": 20530 }, { "epoch": 1.609012539184953, "grad_norm": 12.352157592773438, "learning_rate": 6.52894627830088e-06, "loss": 1.938, "step": 20531 }, { "epoch": 1.6090909090909091, "grad_norm": 3.376403570175171, "learning_rate": 6.528652684116843e-06, "loss": 1.2036, "step": 20532 }, { "epoch": 1.6091692789968652, "grad_norm": 4.359912395477295, "learning_rate": 6.528359084118596e-06, "loss": 1.7113, "step": 20533 }, { "epoch": 1.6092476489028213, "grad_norm": 3.395113229751587, "learning_rate": 6.528065478307256e-06, "loss": 1.3766, "step": 20534 }, { "epoch": 1.6093260188087775, "grad_norm": 3.2423508167266846, "learning_rate": 6.527771866683941e-06, "loss": 1.4563, "step": 20535 }, { "epoch": 1.6094043887147336, "grad_norm": 2.082165241241455, "learning_rate": 6.527478249249765e-06, "loss": 1.3923, "step": 20536 }, { "epoch": 1.6094827586206897, "grad_norm": 3.286954164505005, "learning_rate": 6.527184626005847e-06, "loss": 1.2627, "step": 20537 }, { "epoch": 1.6095611285266458, "grad_norm": 3.473792552947998, "learning_rate": 6.526890996953303e-06, "loss": 1.5909, "step": 20538 }, { "epoch": 1.609639498432602, "grad_norm": 2.506467580795288, "learning_rate": 6.526597362093249e-06, "loss": 1.0141, "step": 20539 }, { "epoch": 1.609717868338558, "grad_norm": 2.7335174083709717, "learning_rate": 6.526303721426803e-06, "loss": 1.1496, "step": 20540 }, { "epoch": 1.6097962382445141, "grad_norm": 2.389814853668213, "learning_rate": 6.526010074955083e-06, "loss": 1.3584, "step": 20541 }, { "epoch": 1.6098746081504702, "grad_norm": 2.2328789234161377, "learning_rate": 6.525716422679202e-06, "loss": 1.6908, "step": 20542 }, { "epoch": 1.6099529780564263, "grad_norm": 4.5849385261535645, "learning_rate": 6.525422764600282e-06, "loss": 0.7978, "step": 20543 }, { "epoch": 1.6100313479623825, "grad_norm": 2.061988353729248, "learning_rate": 6.5251291007194364e-06, "loss": 1.6069, "step": 20544 }, { "epoch": 1.6101097178683386, "grad_norm": 4.673431396484375, "learning_rate": 6.524835431037781e-06, "loss": 1.5215, "step": 20545 }, { "epoch": 1.6101880877742947, "grad_norm": 2.8028926849365234, "learning_rate": 6.524541755556439e-06, "loss": 1.3499, "step": 20546 }, { "epoch": 1.6102664576802508, "grad_norm": 5.158871173858643, "learning_rate": 6.524248074276519e-06, "loss": 1.3525, "step": 20547 }, { "epoch": 1.610344827586207, "grad_norm": 1.932655692100525, "learning_rate": 6.5239543871991454e-06, "loss": 1.1781, "step": 20548 }, { "epoch": 1.610423197492163, "grad_norm": 4.494194507598877, "learning_rate": 6.523660694325431e-06, "loss": 1.1546, "step": 20549 }, { "epoch": 1.6105015673981191, "grad_norm": 2.476419687271118, "learning_rate": 6.523366995656493e-06, "loss": 0.9713, "step": 20550 }, { "epoch": 1.6105799373040752, "grad_norm": 3.7368741035461426, "learning_rate": 6.523073291193451e-06, "loss": 1.7051, "step": 20551 }, { "epoch": 1.6106583072100313, "grad_norm": 5.021912097930908, "learning_rate": 6.522779580937419e-06, "loss": 1.6592, "step": 20552 }, { "epoch": 1.6107366771159874, "grad_norm": 16.38756561279297, "learning_rate": 6.522485864889516e-06, "loss": 1.3788, "step": 20553 }, { "epoch": 1.6108150470219436, "grad_norm": 8.976479530334473, "learning_rate": 6.522192143050858e-06, "loss": 1.4361, "step": 20554 }, { "epoch": 1.6108934169278997, "grad_norm": 9.737736701965332, "learning_rate": 6.521898415422565e-06, "loss": 0.9938, "step": 20555 }, { "epoch": 1.6109717868338558, "grad_norm": 2.1310927867889404, "learning_rate": 6.521604682005751e-06, "loss": 1.1674, "step": 20556 }, { "epoch": 1.611050156739812, "grad_norm": 2.6448514461517334, "learning_rate": 6.521310942801535e-06, "loss": 1.4154, "step": 20557 }, { "epoch": 1.611128526645768, "grad_norm": 4.934900760650635, "learning_rate": 6.521017197811033e-06, "loss": 1.4886, "step": 20558 }, { "epoch": 1.611206896551724, "grad_norm": 2.58927845954895, "learning_rate": 6.520723447035362e-06, "loss": 1.2814, "step": 20559 }, { "epoch": 1.6112852664576802, "grad_norm": 3.0138871669769287, "learning_rate": 6.5204296904756405e-06, "loss": 0.798, "step": 20560 }, { "epoch": 1.6113636363636363, "grad_norm": 5.655742645263672, "learning_rate": 6.520135928132987e-06, "loss": 2.1543, "step": 20561 }, { "epoch": 1.6114420062695924, "grad_norm": 4.174717903137207, "learning_rate": 6.519842160008515e-06, "loss": 1.5471, "step": 20562 }, { "epoch": 1.6115203761755486, "grad_norm": 4.017707824707031, "learning_rate": 6.519548386103346e-06, "loss": 1.2599, "step": 20563 }, { "epoch": 1.6115987460815047, "grad_norm": 8.04269790649414, "learning_rate": 6.519254606418595e-06, "loss": 2.1841, "step": 20564 }, { "epoch": 1.6116771159874608, "grad_norm": 2.5638792514801025, "learning_rate": 6.518960820955379e-06, "loss": 0.9864, "step": 20565 }, { "epoch": 1.6117554858934169, "grad_norm": 3.267648458480835, "learning_rate": 6.518667029714816e-06, "loss": 1.4195, "step": 20566 }, { "epoch": 1.611833855799373, "grad_norm": 2.8924567699432373, "learning_rate": 6.518373232698024e-06, "loss": 1.2287, "step": 20567 }, { "epoch": 1.611912225705329, "grad_norm": 4.050848484039307, "learning_rate": 6.51807942990612e-06, "loss": 1.4793, "step": 20568 }, { "epoch": 1.6119905956112852, "grad_norm": 2.477020025253296, "learning_rate": 6.5177856213402226e-06, "loss": 1.638, "step": 20569 }, { "epoch": 1.6120689655172413, "grad_norm": 3.7741317749023438, "learning_rate": 6.517491807001448e-06, "loss": 1.3874, "step": 20570 }, { "epoch": 1.6121473354231974, "grad_norm": 4.905696868896484, "learning_rate": 6.517197986890913e-06, "loss": 1.6759, "step": 20571 }, { "epoch": 1.6122257053291538, "grad_norm": 3.7197794914245605, "learning_rate": 6.516904161009737e-06, "loss": 1.5227, "step": 20572 }, { "epoch": 1.6123040752351097, "grad_norm": 3.5723297595977783, "learning_rate": 6.516610329359035e-06, "loss": 1.6235, "step": 20573 }, { "epoch": 1.612382445141066, "grad_norm": 3.008063554763794, "learning_rate": 6.516316491939929e-06, "loss": 1.2162, "step": 20574 }, { "epoch": 1.6124608150470219, "grad_norm": 2.4874684810638428, "learning_rate": 6.516022648753533e-06, "loss": 1.1587, "step": 20575 }, { "epoch": 1.6125391849529782, "grad_norm": 2.0782899856567383, "learning_rate": 6.515728799800965e-06, "loss": 1.5695, "step": 20576 }, { "epoch": 1.612617554858934, "grad_norm": 4.713322639465332, "learning_rate": 6.515434945083343e-06, "loss": 1.1648, "step": 20577 }, { "epoch": 1.6126959247648904, "grad_norm": 3.7003958225250244, "learning_rate": 6.515141084601785e-06, "loss": 1.2532, "step": 20578 }, { "epoch": 1.6127742946708463, "grad_norm": 1.8384588956832886, "learning_rate": 6.51484721835741e-06, "loss": 1.2347, "step": 20579 }, { "epoch": 1.6128526645768027, "grad_norm": 3.201658010482788, "learning_rate": 6.514553346351331e-06, "loss": 1.2641, "step": 20580 }, { "epoch": 1.6129310344827585, "grad_norm": 7.436569690704346, "learning_rate": 6.514259468584672e-06, "loss": 1.1861, "step": 20581 }, { "epoch": 1.6130094043887149, "grad_norm": 4.059823513031006, "learning_rate": 6.513965585058546e-06, "loss": 1.4775, "step": 20582 }, { "epoch": 1.6130877742946708, "grad_norm": 5.6734619140625, "learning_rate": 6.513671695774073e-06, "loss": 1.4953, "step": 20583 }, { "epoch": 1.613166144200627, "grad_norm": 3.24320125579834, "learning_rate": 6.513377800732371e-06, "loss": 1.3599, "step": 20584 }, { "epoch": 1.613244514106583, "grad_norm": 2.266465425491333, "learning_rate": 6.5130838999345565e-06, "loss": 1.3936, "step": 20585 }, { "epoch": 1.6133228840125393, "grad_norm": 1.7519090175628662, "learning_rate": 6.512789993381749e-06, "loss": 1.191, "step": 20586 }, { "epoch": 1.6134012539184952, "grad_norm": 13.310980796813965, "learning_rate": 6.512496081075064e-06, "loss": 0.6305, "step": 20587 }, { "epoch": 1.6134796238244515, "grad_norm": 9.654537200927734, "learning_rate": 6.5122021630156206e-06, "loss": 1.4246, "step": 20588 }, { "epoch": 1.6135579937304074, "grad_norm": 6.266297340393066, "learning_rate": 6.511908239204538e-06, "loss": 1.7411, "step": 20589 }, { "epoch": 1.6136363636363638, "grad_norm": 2.285745620727539, "learning_rate": 6.511614309642933e-06, "loss": 1.2113, "step": 20590 }, { "epoch": 1.6137147335423196, "grad_norm": 3.574293375015259, "learning_rate": 6.5113203743319245e-06, "loss": 0.9492, "step": 20591 }, { "epoch": 1.613793103448276, "grad_norm": 2.5271308422088623, "learning_rate": 6.511026433272627e-06, "loss": 1.0048, "step": 20592 }, { "epoch": 1.6138714733542319, "grad_norm": 4.938534259796143, "learning_rate": 6.5107324864661626e-06, "loss": 1.1674, "step": 20593 }, { "epoch": 1.6139498432601882, "grad_norm": 2.13401460647583, "learning_rate": 6.510438533913648e-06, "loss": 1.5124, "step": 20594 }, { "epoch": 1.614028213166144, "grad_norm": 2.521199941635132, "learning_rate": 6.5101445756162e-06, "loss": 1.4667, "step": 20595 }, { "epoch": 1.6141065830721004, "grad_norm": 1.9143232107162476, "learning_rate": 6.509850611574939e-06, "loss": 1.2402, "step": 20596 }, { "epoch": 1.6141849529780563, "grad_norm": 2.4650020599365234, "learning_rate": 6.509556641790981e-06, "loss": 1.3245, "step": 20597 }, { "epoch": 1.6142633228840126, "grad_norm": 2.681656837463379, "learning_rate": 6.509262666265445e-06, "loss": 1.6571, "step": 20598 }, { "epoch": 1.6143416927899685, "grad_norm": 2.9728763103485107, "learning_rate": 6.508968684999448e-06, "loss": 1.5073, "step": 20599 }, { "epoch": 1.6144200626959249, "grad_norm": 3.691005229949951, "learning_rate": 6.508674697994111e-06, "loss": 1.3144, "step": 20600 }, { "epoch": 1.6144984326018808, "grad_norm": 2.6635236740112305, "learning_rate": 6.508380705250549e-06, "loss": 1.3144, "step": 20601 }, { "epoch": 1.614576802507837, "grad_norm": 2.5002517700195312, "learning_rate": 6.508086706769883e-06, "loss": 1.4747, "step": 20602 }, { "epoch": 1.614655172413793, "grad_norm": 3.751093626022339, "learning_rate": 6.507792702553228e-06, "loss": 1.5915, "step": 20603 }, { "epoch": 1.6147335423197493, "grad_norm": 3.592813730239868, "learning_rate": 6.507498692601705e-06, "loss": 1.3189, "step": 20604 }, { "epoch": 1.6148119122257052, "grad_norm": 2.5372207164764404, "learning_rate": 6.507204676916432e-06, "loss": 1.5848, "step": 20605 }, { "epoch": 1.6148902821316615, "grad_norm": 3.0603811740875244, "learning_rate": 6.506910655498524e-06, "loss": 1.5419, "step": 20606 }, { "epoch": 1.6149686520376174, "grad_norm": 2.9759674072265625, "learning_rate": 6.506616628349105e-06, "loss": 1.0812, "step": 20607 }, { "epoch": 1.6150470219435737, "grad_norm": 3.28544545173645, "learning_rate": 6.506322595469287e-06, "loss": 1.3169, "step": 20608 }, { "epoch": 1.6151253918495296, "grad_norm": 3.67924165725708, "learning_rate": 6.506028556860196e-06, "loss": 1.305, "step": 20609 }, { "epoch": 1.615203761755486, "grad_norm": 3.5156946182250977, "learning_rate": 6.505734512522942e-06, "loss": 1.3589, "step": 20610 }, { "epoch": 1.6152821316614419, "grad_norm": 2.915250778198242, "learning_rate": 6.505440462458649e-06, "loss": 1.382, "step": 20611 }, { "epoch": 1.6153605015673982, "grad_norm": 2.982074499130249, "learning_rate": 6.505146406668432e-06, "loss": 1.3907, "step": 20612 }, { "epoch": 1.615438871473354, "grad_norm": 4.724081039428711, "learning_rate": 6.5048523451534125e-06, "loss": 1.7046, "step": 20613 }, { "epoch": 1.6155172413793104, "grad_norm": 3.858325958251953, "learning_rate": 6.504558277914708e-06, "loss": 1.3637, "step": 20614 }, { "epoch": 1.6155956112852663, "grad_norm": 3.9188883304595947, "learning_rate": 6.504264204953435e-06, "loss": 1.1589, "step": 20615 }, { "epoch": 1.6156739811912226, "grad_norm": 3.207094669342041, "learning_rate": 6.503970126270715e-06, "loss": 1.5362, "step": 20616 }, { "epoch": 1.6157523510971785, "grad_norm": 3.158665418624878, "learning_rate": 6.503676041867666e-06, "loss": 1.2668, "step": 20617 }, { "epoch": 1.6158307210031349, "grad_norm": 2.2311534881591797, "learning_rate": 6.503381951745403e-06, "loss": 1.537, "step": 20618 }, { "epoch": 1.615909090909091, "grad_norm": 3.647775650024414, "learning_rate": 6.503087855905051e-06, "loss": 1.5712, "step": 20619 }, { "epoch": 1.615987460815047, "grad_norm": 4.0865797996521, "learning_rate": 6.502793754347721e-06, "loss": 1.6802, "step": 20620 }, { "epoch": 1.6160658307210032, "grad_norm": 3.2979187965393066, "learning_rate": 6.502499647074538e-06, "loss": 1.7313, "step": 20621 }, { "epoch": 1.6161442006269593, "grad_norm": 2.8073842525482178, "learning_rate": 6.502205534086616e-06, "loss": 0.9811, "step": 20622 }, { "epoch": 1.6162225705329154, "grad_norm": 3.2612433433532715, "learning_rate": 6.501911415385077e-06, "loss": 0.8947, "step": 20623 }, { "epoch": 1.6163009404388715, "grad_norm": 2.6208627223968506, "learning_rate": 6.50161729097104e-06, "loss": 1.1764, "step": 20624 }, { "epoch": 1.6163793103448276, "grad_norm": 1.9838380813598633, "learning_rate": 6.5013231608456185e-06, "loss": 1.4692, "step": 20625 }, { "epoch": 1.6164576802507837, "grad_norm": 2.6273410320281982, "learning_rate": 6.501029025009937e-06, "loss": 1.443, "step": 20626 }, { "epoch": 1.6165360501567398, "grad_norm": 4.993374824523926, "learning_rate": 6.500734883465112e-06, "loss": 1.4272, "step": 20627 }, { "epoch": 1.616614420062696, "grad_norm": 3.431861639022827, "learning_rate": 6.500440736212262e-06, "loss": 1.2436, "step": 20628 }, { "epoch": 1.616692789968652, "grad_norm": 2.626190423965454, "learning_rate": 6.500146583252507e-06, "loss": 1.3755, "step": 20629 }, { "epoch": 1.6167711598746082, "grad_norm": 6.551988124847412, "learning_rate": 6.499852424586963e-06, "loss": 1.2389, "step": 20630 }, { "epoch": 1.6168495297805643, "grad_norm": 15.742302894592285, "learning_rate": 6.499558260216752e-06, "loss": 1.9397, "step": 20631 }, { "epoch": 1.6169278996865204, "grad_norm": 3.5956320762634277, "learning_rate": 6.499264090142992e-06, "loss": 1.573, "step": 20632 }, { "epoch": 1.6170062695924765, "grad_norm": 8.725645065307617, "learning_rate": 6.4989699143668016e-06, "loss": 1.2134, "step": 20633 }, { "epoch": 1.6170846394984326, "grad_norm": 2.4233627319335938, "learning_rate": 6.498675732889298e-06, "loss": 1.3605, "step": 20634 }, { "epoch": 1.6171630094043887, "grad_norm": 6.220555782318115, "learning_rate": 6.498381545711603e-06, "loss": 1.509, "step": 20635 }, { "epoch": 1.6172413793103448, "grad_norm": 3.0254058837890625, "learning_rate": 6.4980873528348335e-06, "loss": 1.2796, "step": 20636 }, { "epoch": 1.617319749216301, "grad_norm": 2.284093141555786, "learning_rate": 6.4977931542601105e-06, "loss": 1.4385, "step": 20637 }, { "epoch": 1.617398119122257, "grad_norm": 5.848351955413818, "learning_rate": 6.49749894998855e-06, "loss": 0.6774, "step": 20638 }, { "epoch": 1.6174764890282132, "grad_norm": 3.057077646255493, "learning_rate": 6.4972047400212725e-06, "loss": 1.4253, "step": 20639 }, { "epoch": 1.6175548589341693, "grad_norm": 2.4779531955718994, "learning_rate": 6.496910524359399e-06, "loss": 1.2771, "step": 20640 }, { "epoch": 1.6176332288401254, "grad_norm": 4.151228427886963, "learning_rate": 6.4966163030040456e-06, "loss": 1.7097, "step": 20641 }, { "epoch": 1.6177115987460815, "grad_norm": 3.026479959487915, "learning_rate": 6.496322075956332e-06, "loss": 1.3929, "step": 20642 }, { "epoch": 1.6177899686520376, "grad_norm": 4.444183826446533, "learning_rate": 6.496027843217379e-06, "loss": 1.3966, "step": 20643 }, { "epoch": 1.6178683385579937, "grad_norm": 2.8669683933258057, "learning_rate": 6.495733604788302e-06, "loss": 1.3231, "step": 20644 }, { "epoch": 1.6179467084639498, "grad_norm": 3.4493017196655273, "learning_rate": 6.495439360670226e-06, "loss": 1.3882, "step": 20645 }, { "epoch": 1.618025078369906, "grad_norm": 3.943706512451172, "learning_rate": 6.495145110864265e-06, "loss": 1.5498, "step": 20646 }, { "epoch": 1.618103448275862, "grad_norm": 2.48441743850708, "learning_rate": 6.49485085537154e-06, "loss": 1.6677, "step": 20647 }, { "epoch": 1.6181818181818182, "grad_norm": 11.029603958129883, "learning_rate": 6.494556594193171e-06, "loss": 0.8222, "step": 20648 }, { "epoch": 1.6182601880877743, "grad_norm": 2.170717716217041, "learning_rate": 6.494262327330275e-06, "loss": 0.8713, "step": 20649 }, { "epoch": 1.6183385579937304, "grad_norm": 2.632394790649414, "learning_rate": 6.493968054783973e-06, "loss": 1.0939, "step": 20650 }, { "epoch": 1.6184169278996865, "grad_norm": 2.5761516094207764, "learning_rate": 6.493673776555385e-06, "loss": 1.448, "step": 20651 }, { "epoch": 1.6184952978056426, "grad_norm": 8.483644485473633, "learning_rate": 6.493379492645628e-06, "loss": 1.9031, "step": 20652 }, { "epoch": 1.6185736677115987, "grad_norm": 2.2082273960113525, "learning_rate": 6.493085203055822e-06, "loss": 1.3911, "step": 20653 }, { "epoch": 1.6186520376175548, "grad_norm": 2.075787305831909, "learning_rate": 6.49279090778709e-06, "loss": 1.14, "step": 20654 }, { "epoch": 1.618730407523511, "grad_norm": 5.134134769439697, "learning_rate": 6.492496606840544e-06, "loss": 2.2071, "step": 20655 }, { "epoch": 1.618808777429467, "grad_norm": 4.5091352462768555, "learning_rate": 6.4922023002173094e-06, "loss": 1.6629, "step": 20656 }, { "epoch": 1.6188871473354232, "grad_norm": 2.8843441009521484, "learning_rate": 6.491907987918505e-06, "loss": 1.5867, "step": 20657 }, { "epoch": 1.6189655172413793, "grad_norm": 3.0789361000061035, "learning_rate": 6.491613669945247e-06, "loss": 1.2663, "step": 20658 }, { "epoch": 1.6190438871473354, "grad_norm": 8.793185234069824, "learning_rate": 6.491319346298657e-06, "loss": 1.5601, "step": 20659 }, { "epoch": 1.6191222570532915, "grad_norm": 2.7707738876342773, "learning_rate": 6.491025016979854e-06, "loss": 1.286, "step": 20660 }, { "epoch": 1.6192006269592476, "grad_norm": 3.883511543273926, "learning_rate": 6.490730681989958e-06, "loss": 1.6024, "step": 20661 }, { "epoch": 1.6192789968652037, "grad_norm": 3.875438690185547, "learning_rate": 6.490436341330087e-06, "loss": 1.6764, "step": 20662 }, { "epoch": 1.6193573667711598, "grad_norm": 3.6478235721588135, "learning_rate": 6.4901419950013645e-06, "loss": 1.6967, "step": 20663 }, { "epoch": 1.619435736677116, "grad_norm": 3.1078715324401855, "learning_rate": 6.489847643004905e-06, "loss": 1.4605, "step": 20664 }, { "epoch": 1.6195141065830723, "grad_norm": 2.187591791152954, "learning_rate": 6.4895532853418315e-06, "loss": 1.2766, "step": 20665 }, { "epoch": 1.6195924764890282, "grad_norm": 3.9735450744628906, "learning_rate": 6.489258922013262e-06, "loss": 1.047, "step": 20666 }, { "epoch": 1.6196708463949845, "grad_norm": 2.3414981365203857, "learning_rate": 6.488964553020316e-06, "loss": 0.9507, "step": 20667 }, { "epoch": 1.6197492163009404, "grad_norm": 2.6087937355041504, "learning_rate": 6.4886701783641135e-06, "loss": 1.619, "step": 20668 }, { "epoch": 1.6198275862068967, "grad_norm": 4.400938987731934, "learning_rate": 6.4883757980457754e-06, "loss": 1.0474, "step": 20669 }, { "epoch": 1.6199059561128526, "grad_norm": 3.6221203804016113, "learning_rate": 6.488081412066419e-06, "loss": 1.2844, "step": 20670 }, { "epoch": 1.619984326018809, "grad_norm": 2.630802869796753, "learning_rate": 6.487787020427166e-06, "loss": 1.26, "step": 20671 }, { "epoch": 1.6200626959247648, "grad_norm": 4.131011009216309, "learning_rate": 6.487492623129135e-06, "loss": 1.4425, "step": 20672 }, { "epoch": 1.6201410658307211, "grad_norm": 4.063344478607178, "learning_rate": 6.487198220173447e-06, "loss": 1.4116, "step": 20673 }, { "epoch": 1.620219435736677, "grad_norm": 6.147233009338379, "learning_rate": 6.4869038115612195e-06, "loss": 1.237, "step": 20674 }, { "epoch": 1.6202978056426334, "grad_norm": 6.401065349578857, "learning_rate": 6.486609397293575e-06, "loss": 1.9274, "step": 20675 }, { "epoch": 1.6203761755485893, "grad_norm": 2.542468786239624, "learning_rate": 6.48631497737163e-06, "loss": 1.5018, "step": 20676 }, { "epoch": 1.6204545454545456, "grad_norm": 3.433068037033081, "learning_rate": 6.486020551796508e-06, "loss": 1.5509, "step": 20677 }, { "epoch": 1.6205329153605015, "grad_norm": 3.1350162029266357, "learning_rate": 6.485726120569327e-06, "loss": 1.751, "step": 20678 }, { "epoch": 1.6206112852664578, "grad_norm": 5.58050012588501, "learning_rate": 6.485431683691206e-06, "loss": 1.5298, "step": 20679 }, { "epoch": 1.6206896551724137, "grad_norm": 5.513754367828369, "learning_rate": 6.485137241163266e-06, "loss": 1.1854, "step": 20680 }, { "epoch": 1.62076802507837, "grad_norm": 3.959433078765869, "learning_rate": 6.484842792986626e-06, "loss": 1.3324, "step": 20681 }, { "epoch": 1.620846394984326, "grad_norm": 7.069729328155518, "learning_rate": 6.484548339162409e-06, "loss": 1.2223, "step": 20682 }, { "epoch": 1.6209247648902823, "grad_norm": 3.4541141986846924, "learning_rate": 6.484253879691731e-06, "loss": 1.2315, "step": 20683 }, { "epoch": 1.6210031347962381, "grad_norm": 4.492924213409424, "learning_rate": 6.483959414575714e-06, "loss": 1.4384, "step": 20684 }, { "epoch": 1.6210815047021945, "grad_norm": 3.721754789352417, "learning_rate": 6.4836649438154774e-06, "loss": 1.2286, "step": 20685 }, { "epoch": 1.6211598746081504, "grad_norm": 8.747522354125977, "learning_rate": 6.483370467412143e-06, "loss": 1.9103, "step": 20686 }, { "epoch": 1.6212382445141067, "grad_norm": 2.67409348487854, "learning_rate": 6.4830759853668265e-06, "loss": 1.1959, "step": 20687 }, { "epoch": 1.6213166144200626, "grad_norm": 3.5179085731506348, "learning_rate": 6.482781497680652e-06, "loss": 1.2731, "step": 20688 }, { "epoch": 1.621394984326019, "grad_norm": 2.5070650577545166, "learning_rate": 6.482487004354739e-06, "loss": 1.213, "step": 20689 }, { "epoch": 1.6214733542319748, "grad_norm": 4.369427680969238, "learning_rate": 6.482192505390206e-06, "loss": 1.6439, "step": 20690 }, { "epoch": 1.6215517241379311, "grad_norm": 3.0421535968780518, "learning_rate": 6.4818980007881736e-06, "loss": 1.4958, "step": 20691 }, { "epoch": 1.621630094043887, "grad_norm": 3.6704001426696777, "learning_rate": 6.481603490549764e-06, "loss": 1.1101, "step": 20692 }, { "epoch": 1.6217084639498434, "grad_norm": 2.1939857006073, "learning_rate": 6.481308974676094e-06, "loss": 1.6922, "step": 20693 }, { "epoch": 1.6217868338557992, "grad_norm": 2.9280483722686768, "learning_rate": 6.481014453168286e-06, "loss": 1.2199, "step": 20694 }, { "epoch": 1.6218652037617556, "grad_norm": 2.94187593460083, "learning_rate": 6.48071992602746e-06, "loss": 1.322, "step": 20695 }, { "epoch": 1.6219435736677115, "grad_norm": 2.5304207801818848, "learning_rate": 6.480425393254736e-06, "loss": 0.8059, "step": 20696 }, { "epoch": 1.6220219435736678, "grad_norm": 3.1059417724609375, "learning_rate": 6.480130854851233e-06, "loss": 0.9536, "step": 20697 }, { "epoch": 1.6221003134796237, "grad_norm": 3.1021714210510254, "learning_rate": 6.479836310818074e-06, "loss": 1.3581, "step": 20698 }, { "epoch": 1.62217868338558, "grad_norm": 10.733001708984375, "learning_rate": 6.479541761156378e-06, "loss": 1.2825, "step": 20699 }, { "epoch": 1.622257053291536, "grad_norm": 2.3322408199310303, "learning_rate": 6.479247205867263e-06, "loss": 1.4861, "step": 20700 }, { "epoch": 1.6223354231974922, "grad_norm": 2.1597084999084473, "learning_rate": 6.478952644951854e-06, "loss": 1.2847, "step": 20701 }, { "epoch": 1.6224137931034481, "grad_norm": 3.014500617980957, "learning_rate": 6.478658078411268e-06, "loss": 1.0839, "step": 20702 }, { "epoch": 1.6224921630094045, "grad_norm": 2.760063648223877, "learning_rate": 6.4783635062466256e-06, "loss": 1.4383, "step": 20703 }, { "epoch": 1.6225705329153604, "grad_norm": 3.683678388595581, "learning_rate": 6.478068928459047e-06, "loss": 1.4588, "step": 20704 }, { "epoch": 1.6226489028213167, "grad_norm": 4.244441509246826, "learning_rate": 6.477774345049654e-06, "loss": 1.004, "step": 20705 }, { "epoch": 1.6227272727272726, "grad_norm": 2.5201334953308105, "learning_rate": 6.477479756019568e-06, "loss": 1.295, "step": 20706 }, { "epoch": 1.622805642633229, "grad_norm": 3.473531723022461, "learning_rate": 6.4771851613699055e-06, "loss": 1.5328, "step": 20707 }, { "epoch": 1.6228840125391848, "grad_norm": 3.1617274284362793, "learning_rate": 6.476890561101792e-06, "loss": 1.4787, "step": 20708 }, { "epoch": 1.6229623824451411, "grad_norm": 2.2366580963134766, "learning_rate": 6.476595955216341e-06, "loss": 1.348, "step": 20709 }, { "epoch": 1.623040752351097, "grad_norm": 2.278479814529419, "learning_rate": 6.476301343714682e-06, "loss": 1.0004, "step": 20710 }, { "epoch": 1.6231191222570533, "grad_norm": 3.1565041542053223, "learning_rate": 6.4760067265979286e-06, "loss": 1.9018, "step": 20711 }, { "epoch": 1.6231974921630095, "grad_norm": 13.60552978515625, "learning_rate": 6.475712103867204e-06, "loss": 1.3549, "step": 20712 }, { "epoch": 1.6232758620689656, "grad_norm": 2.5901246070861816, "learning_rate": 6.475417475523629e-06, "loss": 1.2451, "step": 20713 }, { "epoch": 1.6233542319749217, "grad_norm": 2.3234643936157227, "learning_rate": 6.475122841568325e-06, "loss": 1.3136, "step": 20714 }, { "epoch": 1.6234326018808778, "grad_norm": 2.654917001724243, "learning_rate": 6.474828202002409e-06, "loss": 1.669, "step": 20715 }, { "epoch": 1.623510971786834, "grad_norm": 2.683490514755249, "learning_rate": 6.4745335568270055e-06, "loss": 1.3257, "step": 20716 }, { "epoch": 1.62358934169279, "grad_norm": 1.7248092889785767, "learning_rate": 6.4742389060432335e-06, "loss": 1.5398, "step": 20717 }, { "epoch": 1.6236677115987461, "grad_norm": 3.258472204208374, "learning_rate": 6.4739442496522135e-06, "loss": 1.4177, "step": 20718 }, { "epoch": 1.6237460815047022, "grad_norm": 5.329360485076904, "learning_rate": 6.473649587655066e-06, "loss": 1.3355, "step": 20719 }, { "epoch": 1.6238244514106583, "grad_norm": 2.4719929695129395, "learning_rate": 6.473354920052913e-06, "loss": 1.0422, "step": 20720 }, { "epoch": 1.6239028213166145, "grad_norm": 2.635725259780884, "learning_rate": 6.473060246846875e-06, "loss": 1.7216, "step": 20721 }, { "epoch": 1.6239811912225706, "grad_norm": 3.292969226837158, "learning_rate": 6.472765568038073e-06, "loss": 1.4042, "step": 20722 }, { "epoch": 1.6240595611285267, "grad_norm": 2.5321834087371826, "learning_rate": 6.472470883627626e-06, "loss": 1.272, "step": 20723 }, { "epoch": 1.6241379310344828, "grad_norm": 2.694453001022339, "learning_rate": 6.472176193616656e-06, "loss": 1.5724, "step": 20724 }, { "epoch": 1.624216300940439, "grad_norm": 1.9266588687896729, "learning_rate": 6.471881498006286e-06, "loss": 1.4647, "step": 20725 }, { "epoch": 1.624294670846395, "grad_norm": 2.5315451622009277, "learning_rate": 6.471586796797632e-06, "loss": 1.3317, "step": 20726 }, { "epoch": 1.6243730407523511, "grad_norm": 2.12933611869812, "learning_rate": 6.4712920899918194e-06, "loss": 1.232, "step": 20727 }, { "epoch": 1.6244514106583072, "grad_norm": 8.866003036499023, "learning_rate": 6.470997377589966e-06, "loss": 1.6508, "step": 20728 }, { "epoch": 1.6245297805642633, "grad_norm": 2.724490165710449, "learning_rate": 6.470702659593197e-06, "loss": 1.1974, "step": 20729 }, { "epoch": 1.6246081504702194, "grad_norm": 6.822302341461182, "learning_rate": 6.470407936002627e-06, "loss": 1.3286, "step": 20730 }, { "epoch": 1.6246865203761756, "grad_norm": 2.3270866870880127, "learning_rate": 6.470113206819383e-06, "loss": 1.1158, "step": 20731 }, { "epoch": 1.6247648902821317, "grad_norm": 2.777775764465332, "learning_rate": 6.469818472044582e-06, "loss": 1.4578, "step": 20732 }, { "epoch": 1.6248432601880878, "grad_norm": 4.2833380699157715, "learning_rate": 6.469523731679347e-06, "loss": 1.6944, "step": 20733 }, { "epoch": 1.624921630094044, "grad_norm": 8.0176362991333, "learning_rate": 6.4692289857247996e-06, "loss": 0.8538, "step": 20734 }, { "epoch": 1.625, "grad_norm": 2.3990285396575928, "learning_rate": 6.468934234182057e-06, "loss": 1.2199, "step": 20735 }, { "epoch": 1.625078369905956, "grad_norm": 1.954927921295166, "learning_rate": 6.468639477052245e-06, "loss": 0.6305, "step": 20736 }, { "epoch": 1.6251567398119122, "grad_norm": 3.515948534011841, "learning_rate": 6.468344714336482e-06, "loss": 1.3385, "step": 20737 }, { "epoch": 1.6252351097178683, "grad_norm": 2.2362003326416016, "learning_rate": 6.468049946035892e-06, "loss": 1.3083, "step": 20738 }, { "epoch": 1.6253134796238244, "grad_norm": 3.6066019535064697, "learning_rate": 6.4677551721515915e-06, "loss": 1.2679, "step": 20739 }, { "epoch": 1.6253918495297806, "grad_norm": 2.277221202850342, "learning_rate": 6.467460392684706e-06, "loss": 1.4092, "step": 20740 }, { "epoch": 1.6254702194357367, "grad_norm": 3.2907040119171143, "learning_rate": 6.467165607636353e-06, "loss": 1.7909, "step": 20741 }, { "epoch": 1.6255485893416928, "grad_norm": 2.1073477268218994, "learning_rate": 6.4668708170076565e-06, "loss": 1.2487, "step": 20742 }, { "epoch": 1.6256269592476489, "grad_norm": 2.5211069583892822, "learning_rate": 6.466576020799737e-06, "loss": 1.2928, "step": 20743 }, { "epoch": 1.625705329153605, "grad_norm": 2.4377429485321045, "learning_rate": 6.466281219013716e-06, "loss": 1.4445, "step": 20744 }, { "epoch": 1.625783699059561, "grad_norm": 5.286473751068115, "learning_rate": 6.465986411650716e-06, "loss": 1.5102, "step": 20745 }, { "epoch": 1.6258620689655172, "grad_norm": 5.985886096954346, "learning_rate": 6.465691598711853e-06, "loss": 1.5748, "step": 20746 }, { "epoch": 1.6259404388714733, "grad_norm": 5.001277446746826, "learning_rate": 6.4653967801982546e-06, "loss": 1.673, "step": 20747 }, { "epoch": 1.6260188087774294, "grad_norm": 2.902909278869629, "learning_rate": 6.465101956111038e-06, "loss": 0.9991, "step": 20748 }, { "epoch": 1.6260971786833855, "grad_norm": 11.105753898620605, "learning_rate": 6.464807126451326e-06, "loss": 1.671, "step": 20749 }, { "epoch": 1.6261755485893417, "grad_norm": 5.237392902374268, "learning_rate": 6.464512291220241e-06, "loss": 1.198, "step": 20750 }, { "epoch": 1.6262539184952978, "grad_norm": 4.175321102142334, "learning_rate": 6.464217450418903e-06, "loss": 1.5527, "step": 20751 }, { "epoch": 1.6263322884012539, "grad_norm": 2.036918878555298, "learning_rate": 6.463922604048434e-06, "loss": 1.3794, "step": 20752 }, { "epoch": 1.62641065830721, "grad_norm": 7.351240634918213, "learning_rate": 6.463627752109955e-06, "loss": 1.2444, "step": 20753 }, { "epoch": 1.626489028213166, "grad_norm": 4.614598751068115, "learning_rate": 6.463332894604588e-06, "loss": 1.533, "step": 20754 }, { "epoch": 1.6265673981191222, "grad_norm": 3.595479965209961, "learning_rate": 6.463038031533455e-06, "loss": 1.4734, "step": 20755 }, { "epoch": 1.6266457680250783, "grad_norm": 3.6125638484954834, "learning_rate": 6.462743162897676e-06, "loss": 1.4261, "step": 20756 }, { "epoch": 1.6267241379310344, "grad_norm": 3.3613479137420654, "learning_rate": 6.462448288698374e-06, "loss": 1.1514, "step": 20757 }, { "epoch": 1.6268025078369905, "grad_norm": 2.4057769775390625, "learning_rate": 6.462153408936669e-06, "loss": 1.4747, "step": 20758 }, { "epoch": 1.6268808777429467, "grad_norm": 3.8152647018432617, "learning_rate": 6.461858523613684e-06, "loss": 1.7447, "step": 20759 }, { "epoch": 1.626959247648903, "grad_norm": 3.2662057876586914, "learning_rate": 6.461563632730541e-06, "loss": 1.3863, "step": 20760 }, { "epoch": 1.6270376175548589, "grad_norm": 3.920545816421509, "learning_rate": 6.461268736288358e-06, "loss": 0.7583, "step": 20761 }, { "epoch": 1.6271159874608152, "grad_norm": 5.4261393547058105, "learning_rate": 6.4609738342882625e-06, "loss": 1.4473, "step": 20762 }, { "epoch": 1.627194357366771, "grad_norm": 9.767740249633789, "learning_rate": 6.4606789267313696e-06, "loss": 1.1551, "step": 20763 }, { "epoch": 1.6272727272727274, "grad_norm": 7.121468544006348, "learning_rate": 6.460384013618808e-06, "loss": 1.6513, "step": 20764 }, { "epoch": 1.6273510971786833, "grad_norm": 3.0034432411193848, "learning_rate": 6.460089094951692e-06, "loss": 1.2128, "step": 20765 }, { "epoch": 1.6274294670846396, "grad_norm": 3.9925389289855957, "learning_rate": 6.45979417073115e-06, "loss": 1.8458, "step": 20766 }, { "epoch": 1.6275078369905955, "grad_norm": 3.608966112136841, "learning_rate": 6.459499240958299e-06, "loss": 1.2968, "step": 20767 }, { "epoch": 1.6275862068965519, "grad_norm": 8.87270736694336, "learning_rate": 6.459204305634263e-06, "loss": 1.8185, "step": 20768 }, { "epoch": 1.6276645768025078, "grad_norm": 2.2831783294677734, "learning_rate": 6.458909364760163e-06, "loss": 0.8405, "step": 20769 }, { "epoch": 1.627742946708464, "grad_norm": 2.495169162750244, "learning_rate": 6.4586144183371215e-06, "loss": 1.2348, "step": 20770 }, { "epoch": 1.62782131661442, "grad_norm": 4.172596454620361, "learning_rate": 6.458319466366259e-06, "loss": 1.3152, "step": 20771 }, { "epoch": 1.6278996865203763, "grad_norm": 2.5650768280029297, "learning_rate": 6.458024508848702e-06, "loss": 1.2608, "step": 20772 }, { "epoch": 1.6279780564263322, "grad_norm": 2.3375282287597656, "learning_rate": 6.4577295457855646e-06, "loss": 1.4038, "step": 20773 }, { "epoch": 1.6280564263322885, "grad_norm": 6.106447219848633, "learning_rate": 6.457434577177975e-06, "loss": 1.1899, "step": 20774 }, { "epoch": 1.6281347962382444, "grad_norm": 3.6429860591888428, "learning_rate": 6.457139603027052e-06, "loss": 1.4396, "step": 20775 }, { "epoch": 1.6282131661442008, "grad_norm": 4.7375569343566895, "learning_rate": 6.456844623333918e-06, "loss": 0.7598, "step": 20776 }, { "epoch": 1.6282915360501566, "grad_norm": 2.2925429344177246, "learning_rate": 6.456549638099695e-06, "loss": 0.9499, "step": 20777 }, { "epoch": 1.628369905956113, "grad_norm": 2.2515857219696045, "learning_rate": 6.456254647325506e-06, "loss": 1.1324, "step": 20778 }, { "epoch": 1.6284482758620689, "grad_norm": 6.873986721038818, "learning_rate": 6.455959651012471e-06, "loss": 1.7617, "step": 20779 }, { "epoch": 1.6285266457680252, "grad_norm": 2.2765939235687256, "learning_rate": 6.4556646491617145e-06, "loss": 1.3101, "step": 20780 }, { "epoch": 1.628605015673981, "grad_norm": 3.3771517276763916, "learning_rate": 6.455369641774359e-06, "loss": 1.4713, "step": 20781 }, { "epoch": 1.6286833855799374, "grad_norm": 3.9696292877197266, "learning_rate": 6.455074628851522e-06, "loss": 1.3908, "step": 20782 }, { "epoch": 1.6287617554858933, "grad_norm": 2.011267900466919, "learning_rate": 6.45477961039433e-06, "loss": 1.4775, "step": 20783 }, { "epoch": 1.6288401253918496, "grad_norm": 6.816311836242676, "learning_rate": 6.454484586403904e-06, "loss": 1.1575, "step": 20784 }, { "epoch": 1.6289184952978055, "grad_norm": 2.9397335052490234, "learning_rate": 6.454189556881365e-06, "loss": 1.4278, "step": 20785 }, { "epoch": 1.6289968652037619, "grad_norm": 2.345200777053833, "learning_rate": 6.453894521827836e-06, "loss": 1.2061, "step": 20786 }, { "epoch": 1.6290752351097177, "grad_norm": 3.6644175052642822, "learning_rate": 6.453599481244439e-06, "loss": 1.6509, "step": 20787 }, { "epoch": 1.629153605015674, "grad_norm": 3.1314239501953125, "learning_rate": 6.453304435132297e-06, "loss": 1.0761, "step": 20788 }, { "epoch": 1.62923197492163, "grad_norm": 3.025465250015259, "learning_rate": 6.453009383492529e-06, "loss": 1.5021, "step": 20789 }, { "epoch": 1.6293103448275863, "grad_norm": 8.4404935836792, "learning_rate": 6.452714326326262e-06, "loss": 1.9965, "step": 20790 }, { "epoch": 1.6293887147335422, "grad_norm": 4.06135368347168, "learning_rate": 6.4524192636346135e-06, "loss": 1.0504, "step": 20791 }, { "epoch": 1.6294670846394985, "grad_norm": 3.6288623809814453, "learning_rate": 6.452124195418711e-06, "loss": 1.7185, "step": 20792 }, { "epoch": 1.6295454545454544, "grad_norm": 3.3369762897491455, "learning_rate": 6.451829121679673e-06, "loss": 1.2155, "step": 20793 }, { "epoch": 1.6296238244514107, "grad_norm": 2.6942076683044434, "learning_rate": 6.451534042418621e-06, "loss": 1.3852, "step": 20794 }, { "epoch": 1.6297021943573666, "grad_norm": 4.67692232131958, "learning_rate": 6.451238957636681e-06, "loss": 1.2453, "step": 20795 }, { "epoch": 1.629780564263323, "grad_norm": 9.233853340148926, "learning_rate": 6.450943867334972e-06, "loss": 1.4769, "step": 20796 }, { "epoch": 1.6298589341692789, "grad_norm": 2.138134479522705, "learning_rate": 6.450648771514618e-06, "loss": 1.283, "step": 20797 }, { "epoch": 1.6299373040752352, "grad_norm": 2.383007287979126, "learning_rate": 6.450353670176742e-06, "loss": 1.2627, "step": 20798 }, { "epoch": 1.630015673981191, "grad_norm": 2.20448899269104, "learning_rate": 6.4500585633224655e-06, "loss": 1.3242, "step": 20799 }, { "epoch": 1.6300940438871474, "grad_norm": 6.962295055389404, "learning_rate": 6.449763450952912e-06, "loss": 1.5808, "step": 20800 }, { "epoch": 1.6301724137931033, "grad_norm": 2.4898927211761475, "learning_rate": 6.449468333069201e-06, "loss": 1.2104, "step": 20801 }, { "epoch": 1.6302507836990596, "grad_norm": 2.095961332321167, "learning_rate": 6.449173209672458e-06, "loss": 1.2371, "step": 20802 }, { "epoch": 1.6303291536050155, "grad_norm": 3.643204927444458, "learning_rate": 6.448878080763804e-06, "loss": 1.2444, "step": 20803 }, { "epoch": 1.6304075235109718, "grad_norm": 2.440054178237915, "learning_rate": 6.448582946344361e-06, "loss": 1.288, "step": 20804 }, { "epoch": 1.6304858934169277, "grad_norm": 11.335838317871094, "learning_rate": 6.448287806415254e-06, "loss": 1.939, "step": 20805 }, { "epoch": 1.630564263322884, "grad_norm": 1.8579288721084595, "learning_rate": 6.447992660977605e-06, "loss": 0.9418, "step": 20806 }, { "epoch": 1.6306426332288402, "grad_norm": 4.425692558288574, "learning_rate": 6.447697510032534e-06, "loss": 0.9095, "step": 20807 }, { "epoch": 1.6307210031347963, "grad_norm": 2.904136896133423, "learning_rate": 6.447402353581164e-06, "loss": 1.2598, "step": 20808 }, { "epoch": 1.6307993730407524, "grad_norm": 2.4804489612579346, "learning_rate": 6.447107191624622e-06, "loss": 1.2691, "step": 20809 }, { "epoch": 1.6308777429467085, "grad_norm": 3.7015273571014404, "learning_rate": 6.446812024164026e-06, "loss": 1.0725, "step": 20810 }, { "epoch": 1.6309561128526646, "grad_norm": 3.538177967071533, "learning_rate": 6.446516851200501e-06, "loss": 0.9261, "step": 20811 }, { "epoch": 1.6310344827586207, "grad_norm": 5.27646541595459, "learning_rate": 6.446221672735167e-06, "loss": 1.9224, "step": 20812 }, { "epoch": 1.6311128526645768, "grad_norm": 5.904325008392334, "learning_rate": 6.44592648876915e-06, "loss": 1.2632, "step": 20813 }, { "epoch": 1.631191222570533, "grad_norm": 3.8986175060272217, "learning_rate": 6.445631299303572e-06, "loss": 1.2631, "step": 20814 }, { "epoch": 1.631269592476489, "grad_norm": 2.794556140899658, "learning_rate": 6.445336104339552e-06, "loss": 1.4094, "step": 20815 }, { "epoch": 1.6313479623824452, "grad_norm": 3.7882072925567627, "learning_rate": 6.445040903878219e-06, "loss": 1.5799, "step": 20816 }, { "epoch": 1.6314263322884013, "grad_norm": 9.510628700256348, "learning_rate": 6.444745697920691e-06, "loss": 1.2329, "step": 20817 }, { "epoch": 1.6315047021943574, "grad_norm": 2.474616289138794, "learning_rate": 6.444450486468094e-06, "loss": 1.276, "step": 20818 }, { "epoch": 1.6315830721003135, "grad_norm": 8.274250030517578, "learning_rate": 6.444155269521548e-06, "loss": 1.1275, "step": 20819 }, { "epoch": 1.6316614420062696, "grad_norm": 6.117449760437012, "learning_rate": 6.443860047082177e-06, "loss": 1.0152, "step": 20820 }, { "epoch": 1.6317398119122257, "grad_norm": 2.1790852546691895, "learning_rate": 6.443564819151105e-06, "loss": 1.4729, "step": 20821 }, { "epoch": 1.6318181818181818, "grad_norm": 2.8268697261810303, "learning_rate": 6.443269585729453e-06, "loss": 1.4214, "step": 20822 }, { "epoch": 1.631896551724138, "grad_norm": 3.810249090194702, "learning_rate": 6.442974346818345e-06, "loss": 1.098, "step": 20823 }, { "epoch": 1.631974921630094, "grad_norm": 2.39312481880188, "learning_rate": 6.442679102418903e-06, "loss": 0.847, "step": 20824 }, { "epoch": 1.6320532915360502, "grad_norm": 3.0718259811401367, "learning_rate": 6.442383852532252e-06, "loss": 1.2957, "step": 20825 }, { "epoch": 1.6321316614420063, "grad_norm": 2.3240394592285156, "learning_rate": 6.442088597159514e-06, "loss": 1.571, "step": 20826 }, { "epoch": 1.6322100313479624, "grad_norm": 9.564406394958496, "learning_rate": 6.441793336301809e-06, "loss": 1.4084, "step": 20827 }, { "epoch": 1.6322884012539185, "grad_norm": 2.028167486190796, "learning_rate": 6.441498069960266e-06, "loss": 1.6339, "step": 20828 }, { "epoch": 1.6323667711598746, "grad_norm": 4.695261478424072, "learning_rate": 6.441202798136003e-06, "loss": 1.8596, "step": 20829 }, { "epoch": 1.6324451410658307, "grad_norm": 3.228250026702881, "learning_rate": 6.4409075208301454e-06, "loss": 1.2433, "step": 20830 }, { "epoch": 1.6325235109717868, "grad_norm": 2.481369733810425, "learning_rate": 6.440612238043815e-06, "loss": 1.4958, "step": 20831 }, { "epoch": 1.632601880877743, "grad_norm": 2.3308908939361572, "learning_rate": 6.440316949778136e-06, "loss": 1.2962, "step": 20832 }, { "epoch": 1.632680250783699, "grad_norm": 3.4539458751678467, "learning_rate": 6.44002165603423e-06, "loss": 1.2831, "step": 20833 }, { "epoch": 1.6327586206896552, "grad_norm": 4.34423828125, "learning_rate": 6.439726356813223e-06, "loss": 1.5242, "step": 20834 }, { "epoch": 1.6328369905956113, "grad_norm": 3.6312358379364014, "learning_rate": 6.439431052116236e-06, "loss": 1.0465, "step": 20835 }, { "epoch": 1.6329153605015674, "grad_norm": 2.9936530590057373, "learning_rate": 6.439135741944391e-06, "loss": 1.2932, "step": 20836 }, { "epoch": 1.6329937304075235, "grad_norm": 3.3111629486083984, "learning_rate": 6.438840426298815e-06, "loss": 1.3549, "step": 20837 }, { "epoch": 1.6330721003134796, "grad_norm": 2.5303258895874023, "learning_rate": 6.438545105180629e-06, "loss": 1.5276, "step": 20838 }, { "epoch": 1.6331504702194357, "grad_norm": 9.857606887817383, "learning_rate": 6.438249778590955e-06, "loss": 1.5524, "step": 20839 }, { "epoch": 1.6332288401253918, "grad_norm": 4.45313024520874, "learning_rate": 6.437954446530919e-06, "loss": 0.9777, "step": 20840 }, { "epoch": 1.633307210031348, "grad_norm": 3.4764251708984375, "learning_rate": 6.4376591090016416e-06, "loss": 1.208, "step": 20841 }, { "epoch": 1.633385579937304, "grad_norm": 3.0680832862854004, "learning_rate": 6.437363766004248e-06, "loss": 1.6162, "step": 20842 }, { "epoch": 1.6334639498432602, "grad_norm": 2.483319044113159, "learning_rate": 6.4370684175398605e-06, "loss": 1.1817, "step": 20843 }, { "epoch": 1.6335423197492163, "grad_norm": 2.924804925918579, "learning_rate": 6.436773063609604e-06, "loss": 1.5184, "step": 20844 }, { "epoch": 1.6336206896551724, "grad_norm": 2.433915138244629, "learning_rate": 6.436477704214599e-06, "loss": 1.4355, "step": 20845 }, { "epoch": 1.6336990595611285, "grad_norm": 2.5893893241882324, "learning_rate": 6.436182339355972e-06, "loss": 1.4343, "step": 20846 }, { "epoch": 1.6337774294670846, "grad_norm": 2.0061287879943848, "learning_rate": 6.435886969034844e-06, "loss": 1.2316, "step": 20847 }, { "epoch": 1.6338557993730407, "grad_norm": 4.279733180999756, "learning_rate": 6.43559159325234e-06, "loss": 1.5659, "step": 20848 }, { "epoch": 1.6339341692789968, "grad_norm": 2.312519073486328, "learning_rate": 6.435296212009583e-06, "loss": 1.3473, "step": 20849 }, { "epoch": 1.634012539184953, "grad_norm": 2.143399238586426, "learning_rate": 6.435000825307696e-06, "loss": 1.223, "step": 20850 }, { "epoch": 1.634090909090909, "grad_norm": 2.0252933502197266, "learning_rate": 6.434705433147803e-06, "loss": 1.0156, "step": 20851 }, { "epoch": 1.6341692789968651, "grad_norm": 2.898984909057617, "learning_rate": 6.434410035531027e-06, "loss": 1.2685, "step": 20852 }, { "epoch": 1.6342476489028215, "grad_norm": 4.283604145050049, "learning_rate": 6.434114632458493e-06, "loss": 1.2231, "step": 20853 }, { "epoch": 1.6343260188087774, "grad_norm": 5.079522132873535, "learning_rate": 6.433819223931324e-06, "loss": 1.2888, "step": 20854 }, { "epoch": 1.6344043887147337, "grad_norm": 1.9802908897399902, "learning_rate": 6.433523809950642e-06, "loss": 1.2754, "step": 20855 }, { "epoch": 1.6344827586206896, "grad_norm": 2.4416089057922363, "learning_rate": 6.433228390517572e-06, "loss": 1.5425, "step": 20856 }, { "epoch": 1.634561128526646, "grad_norm": 5.374349594116211, "learning_rate": 6.432932965633238e-06, "loss": 1.5211, "step": 20857 }, { "epoch": 1.6346394984326018, "grad_norm": 3.161346435546875, "learning_rate": 6.4326375352987624e-06, "loss": 1.1432, "step": 20858 }, { "epoch": 1.6347178683385581, "grad_norm": 4.308370113372803, "learning_rate": 6.432342099515269e-06, "loss": 1.5413, "step": 20859 }, { "epoch": 1.634796238244514, "grad_norm": 3.3321645259857178, "learning_rate": 6.432046658283882e-06, "loss": 1.383, "step": 20860 }, { "epoch": 1.6348746081504704, "grad_norm": 3.5593674182891846, "learning_rate": 6.431751211605727e-06, "loss": 0.9655, "step": 20861 }, { "epoch": 1.6349529780564263, "grad_norm": 2.367950201034546, "learning_rate": 6.431455759481923e-06, "loss": 1.5121, "step": 20862 }, { "epoch": 1.6350313479623826, "grad_norm": 6.948366641998291, "learning_rate": 6.431160301913599e-06, "loss": 1.375, "step": 20863 }, { "epoch": 1.6351097178683385, "grad_norm": 2.173990488052368, "learning_rate": 6.4308648389018735e-06, "loss": 1.261, "step": 20864 }, { "epoch": 1.6351880877742948, "grad_norm": 2.1607394218444824, "learning_rate": 6.430569370447877e-06, "loss": 1.2218, "step": 20865 }, { "epoch": 1.6352664576802507, "grad_norm": 2.57767391204834, "learning_rate": 6.430273896552726e-06, "loss": 1.5048, "step": 20866 }, { "epoch": 1.635344827586207, "grad_norm": 3.83054780960083, "learning_rate": 6.429978417217548e-06, "loss": 1.3628, "step": 20867 }, { "epoch": 1.635423197492163, "grad_norm": 7.056385040283203, "learning_rate": 6.429682932443468e-06, "loss": 2.8479, "step": 20868 }, { "epoch": 1.6355015673981192, "grad_norm": 2.0363919734954834, "learning_rate": 6.429387442231608e-06, "loss": 1.5125, "step": 20869 }, { "epoch": 1.6355799373040751, "grad_norm": 3.1570723056793213, "learning_rate": 6.429091946583093e-06, "loss": 1.1659, "step": 20870 }, { "epoch": 1.6356583072100315, "grad_norm": 2.9559412002563477, "learning_rate": 6.428796445499045e-06, "loss": 1.4107, "step": 20871 }, { "epoch": 1.6357366771159874, "grad_norm": 3.225382089614868, "learning_rate": 6.428500938980591e-06, "loss": 1.3874, "step": 20872 }, { "epoch": 1.6358150470219437, "grad_norm": 6.2771782875061035, "learning_rate": 6.428205427028851e-06, "loss": 1.2405, "step": 20873 }, { "epoch": 1.6358934169278996, "grad_norm": 2.353229522705078, "learning_rate": 6.427909909644951e-06, "loss": 1.6065, "step": 20874 }, { "epoch": 1.635971786833856, "grad_norm": 6.70040225982666, "learning_rate": 6.4276143868300164e-06, "loss": 0.5953, "step": 20875 }, { "epoch": 1.6360501567398118, "grad_norm": 3.521907329559326, "learning_rate": 6.42731885858517e-06, "loss": 0.927, "step": 20876 }, { "epoch": 1.6361285266457681, "grad_norm": 3.2254579067230225, "learning_rate": 6.427023324911535e-06, "loss": 1.6627, "step": 20877 }, { "epoch": 1.636206896551724, "grad_norm": 3.230703115463257, "learning_rate": 6.426727785810237e-06, "loss": 1.4523, "step": 20878 }, { "epoch": 1.6362852664576804, "grad_norm": 3.451662302017212, "learning_rate": 6.426432241282399e-06, "loss": 1.3116, "step": 20879 }, { "epoch": 1.6363636363636362, "grad_norm": 2.7846171855926514, "learning_rate": 6.4261366913291456e-06, "loss": 1.2418, "step": 20880 }, { "epoch": 1.6364420062695926, "grad_norm": 4.535923004150391, "learning_rate": 6.4258411359516e-06, "loss": 1.0569, "step": 20881 }, { "epoch": 1.6365203761755485, "grad_norm": 3.425485849380493, "learning_rate": 6.425545575150889e-06, "loss": 1.1777, "step": 20882 }, { "epoch": 1.6365987460815048, "grad_norm": 2.7447476387023926, "learning_rate": 6.425250008928132e-06, "loss": 1.7335, "step": 20883 }, { "epoch": 1.6366771159874607, "grad_norm": 2.408874750137329, "learning_rate": 6.424954437284458e-06, "loss": 1.4716, "step": 20884 }, { "epoch": 1.636755485893417, "grad_norm": 7.930013179779053, "learning_rate": 6.424658860220989e-06, "loss": 0.7158, "step": 20885 }, { "epoch": 1.636833855799373, "grad_norm": 2.3598735332489014, "learning_rate": 6.4243632777388486e-06, "loss": 1.2368, "step": 20886 }, { "epoch": 1.6369122257053292, "grad_norm": 2.7484774589538574, "learning_rate": 6.424067689839163e-06, "loss": 1.2367, "step": 20887 }, { "epoch": 1.6369905956112851, "grad_norm": 5.946088790893555, "learning_rate": 6.4237720965230545e-06, "loss": 0.8335, "step": 20888 }, { "epoch": 1.6370689655172415, "grad_norm": 3.3918328285217285, "learning_rate": 6.42347649779165e-06, "loss": 1.4544, "step": 20889 }, { "epoch": 1.6371473354231973, "grad_norm": 5.058273792266846, "learning_rate": 6.423180893646068e-06, "loss": 1.6108, "step": 20890 }, { "epoch": 1.6372257053291537, "grad_norm": 10.939276695251465, "learning_rate": 6.422885284087441e-06, "loss": 0.8545, "step": 20891 }, { "epoch": 1.6373040752351096, "grad_norm": 3.6262550354003906, "learning_rate": 6.422589669116886e-06, "loss": 1.12, "step": 20892 }, { "epoch": 1.637382445141066, "grad_norm": 4.815717697143555, "learning_rate": 6.422294048735534e-06, "loss": 1.641, "step": 20893 }, { "epoch": 1.6374608150470218, "grad_norm": 3.5141680240631104, "learning_rate": 6.4219984229445045e-06, "loss": 1.6037, "step": 20894 }, { "epoch": 1.6375391849529781, "grad_norm": 3.184952974319458, "learning_rate": 6.421702791744923e-06, "loss": 1.4815, "step": 20895 }, { "epoch": 1.637617554858934, "grad_norm": 2.7142534255981445, "learning_rate": 6.421407155137913e-06, "loss": 1.2251, "step": 20896 }, { "epoch": 1.6376959247648903, "grad_norm": 4.523089408874512, "learning_rate": 6.421111513124603e-06, "loss": 1.2268, "step": 20897 }, { "epoch": 1.6377742946708462, "grad_norm": 3.508687973022461, "learning_rate": 6.420815865706114e-06, "loss": 1.5234, "step": 20898 }, { "epoch": 1.6378526645768026, "grad_norm": 4.463326930999756, "learning_rate": 6.4205202128835695e-06, "loss": 1.7142, "step": 20899 }, { "epoch": 1.6379310344827587, "grad_norm": 2.7429378032684326, "learning_rate": 6.420224554658098e-06, "loss": 1.5237, "step": 20900 }, { "epoch": 1.6380094043887148, "grad_norm": 4.606085300445557, "learning_rate": 6.41992889103082e-06, "loss": 1.3, "step": 20901 }, { "epoch": 1.638087774294671, "grad_norm": 2.4577887058258057, "learning_rate": 6.419633222002862e-06, "loss": 1.4805, "step": 20902 }, { "epoch": 1.638166144200627, "grad_norm": 15.297701835632324, "learning_rate": 6.419337547575349e-06, "loss": 1.145, "step": 20903 }, { "epoch": 1.6382445141065831, "grad_norm": 2.071031093597412, "learning_rate": 6.419041867749405e-06, "loss": 1.06, "step": 20904 }, { "epoch": 1.6383228840125392, "grad_norm": 2.6760575771331787, "learning_rate": 6.418746182526154e-06, "loss": 1.1564, "step": 20905 }, { "epoch": 1.6384012539184953, "grad_norm": 2.607179880142212, "learning_rate": 6.418450491906722e-06, "loss": 1.0794, "step": 20906 }, { "epoch": 1.6384796238244514, "grad_norm": 2.8634700775146484, "learning_rate": 6.418154795892233e-06, "loss": 1.3017, "step": 20907 }, { "epoch": 1.6385579937304076, "grad_norm": 2.120145320892334, "learning_rate": 6.417859094483812e-06, "loss": 1.1588, "step": 20908 }, { "epoch": 1.6386363636363637, "grad_norm": 5.763986110687256, "learning_rate": 6.41756338768258e-06, "loss": 1.6835, "step": 20909 }, { "epoch": 1.6387147335423198, "grad_norm": 2.358246326446533, "learning_rate": 6.417267675489669e-06, "loss": 1.5055, "step": 20910 }, { "epoch": 1.638793103448276, "grad_norm": 3.03165864944458, "learning_rate": 6.416971957906198e-06, "loss": 1.6266, "step": 20911 }, { "epoch": 1.638871473354232, "grad_norm": 2.5759780406951904, "learning_rate": 6.416676234933293e-06, "loss": 1.0089, "step": 20912 }, { "epoch": 1.638949843260188, "grad_norm": 2.092346668243408, "learning_rate": 6.4163805065720804e-06, "loss": 1.4648, "step": 20913 }, { "epoch": 1.6390282131661442, "grad_norm": 2.7331936359405518, "learning_rate": 6.416084772823683e-06, "loss": 1.4297, "step": 20914 }, { "epoch": 1.6391065830721003, "grad_norm": 2.144188165664673, "learning_rate": 6.415789033689228e-06, "loss": 1.1282, "step": 20915 }, { "epoch": 1.6391849529780564, "grad_norm": 3.931094169616699, "learning_rate": 6.415493289169836e-06, "loss": 1.2978, "step": 20916 }, { "epoch": 1.6392633228840126, "grad_norm": 1.902237057685852, "learning_rate": 6.415197539266638e-06, "loss": 1.3152, "step": 20917 }, { "epoch": 1.6393416927899687, "grad_norm": 6.07900857925415, "learning_rate": 6.414901783980751e-06, "loss": 1.3292, "step": 20918 }, { "epoch": 1.6394200626959248, "grad_norm": 4.459927082061768, "learning_rate": 6.414606023313308e-06, "loss": 1.8155, "step": 20919 }, { "epoch": 1.6394984326018809, "grad_norm": 2.399127244949341, "learning_rate": 6.41431025726543e-06, "loss": 1.5647, "step": 20920 }, { "epoch": 1.639576802507837, "grad_norm": 8.277502059936523, "learning_rate": 6.414014485838241e-06, "loss": 1.669, "step": 20921 }, { "epoch": 1.639655172413793, "grad_norm": 3.0044126510620117, "learning_rate": 6.413718709032867e-06, "loss": 1.284, "step": 20922 }, { "epoch": 1.6397335423197492, "grad_norm": 2.052356243133545, "learning_rate": 6.413422926850434e-06, "loss": 1.4737, "step": 20923 }, { "epoch": 1.6398119122257053, "grad_norm": 3.020526647567749, "learning_rate": 6.413127139292065e-06, "loss": 1.2077, "step": 20924 }, { "epoch": 1.6398902821316614, "grad_norm": 2.678157329559326, "learning_rate": 6.412831346358888e-06, "loss": 1.3836, "step": 20925 }, { "epoch": 1.6399686520376175, "grad_norm": 2.6301708221435547, "learning_rate": 6.412535548052025e-06, "loss": 0.8561, "step": 20926 }, { "epoch": 1.6400470219435737, "grad_norm": 2.5192477703094482, "learning_rate": 6.412239744372603e-06, "loss": 0.9344, "step": 20927 }, { "epoch": 1.6401253918495298, "grad_norm": 5.613051891326904, "learning_rate": 6.411943935321746e-06, "loss": 1.5125, "step": 20928 }, { "epoch": 1.6402037617554859, "grad_norm": 2.5223445892333984, "learning_rate": 6.41164812090058e-06, "loss": 1.7635, "step": 20929 }, { "epoch": 1.640282131661442, "grad_norm": 5.3068037033081055, "learning_rate": 6.4113523011102284e-06, "loss": 1.6559, "step": 20930 }, { "epoch": 1.640360501567398, "grad_norm": 4.519196510314941, "learning_rate": 6.411056475951818e-06, "loss": 1.4585, "step": 20931 }, { "epoch": 1.6404388714733542, "grad_norm": 2.2630045413970947, "learning_rate": 6.410760645426473e-06, "loss": 0.8659, "step": 20932 }, { "epoch": 1.6405172413793103, "grad_norm": 3.685741901397705, "learning_rate": 6.41046480953532e-06, "loss": 1.4541, "step": 20933 }, { "epoch": 1.6405956112852664, "grad_norm": 2.3010218143463135, "learning_rate": 6.410168968279483e-06, "loss": 1.271, "step": 20934 }, { "epoch": 1.6406739811912225, "grad_norm": 3.2125606536865234, "learning_rate": 6.409873121660088e-06, "loss": 1.4667, "step": 20935 }, { "epoch": 1.6407523510971787, "grad_norm": 2.1088435649871826, "learning_rate": 6.40957726967826e-06, "loss": 1.4468, "step": 20936 }, { "epoch": 1.6408307210031348, "grad_norm": 3.705383062362671, "learning_rate": 6.409281412335123e-06, "loss": 1.3869, "step": 20937 }, { "epoch": 1.6409090909090909, "grad_norm": 3.1237051486968994, "learning_rate": 6.408985549631805e-06, "loss": 1.3705, "step": 20938 }, { "epoch": 1.640987460815047, "grad_norm": 4.1409173011779785, "learning_rate": 6.408689681569429e-06, "loss": 1.2837, "step": 20939 }, { "epoch": 1.641065830721003, "grad_norm": 6.653587341308594, "learning_rate": 6.40839380814912e-06, "loss": 1.3994, "step": 20940 }, { "epoch": 1.6411442006269592, "grad_norm": 4.803877830505371, "learning_rate": 6.408097929372004e-06, "loss": 0.8634, "step": 20941 }, { "epoch": 1.6412225705329153, "grad_norm": 4.02607536315918, "learning_rate": 6.4078020452392084e-06, "loss": 1.5469, "step": 20942 }, { "epoch": 1.6413009404388714, "grad_norm": 2.932661771774292, "learning_rate": 6.4075061557518566e-06, "loss": 1.3143, "step": 20943 }, { "epoch": 1.6413793103448275, "grad_norm": 3.8384201526641846, "learning_rate": 6.407210260911074e-06, "loss": 1.3466, "step": 20944 }, { "epoch": 1.6414576802507836, "grad_norm": 3.3582184314727783, "learning_rate": 6.406914360717987e-06, "loss": 1.2052, "step": 20945 }, { "epoch": 1.6415360501567398, "grad_norm": 4.980443000793457, "learning_rate": 6.406618455173718e-06, "loss": 1.2186, "step": 20946 }, { "epoch": 1.6416144200626959, "grad_norm": 2.3013553619384766, "learning_rate": 6.406322544279398e-06, "loss": 1.2278, "step": 20947 }, { "epoch": 1.6416927899686522, "grad_norm": 4.759364604949951, "learning_rate": 6.406026628036149e-06, "loss": 1.1462, "step": 20948 }, { "epoch": 1.641771159874608, "grad_norm": 3.26910400390625, "learning_rate": 6.405730706445094e-06, "loss": 1.1282, "step": 20949 }, { "epoch": 1.6418495297805644, "grad_norm": 2.862990617752075, "learning_rate": 6.405434779507363e-06, "loss": 1.4296, "step": 20950 }, { "epoch": 1.6419278996865203, "grad_norm": 2.5700523853302, "learning_rate": 6.405138847224082e-06, "loss": 1.7036, "step": 20951 }, { "epoch": 1.6420062695924766, "grad_norm": 3.0334525108337402, "learning_rate": 6.404842909596372e-06, "loss": 1.282, "step": 20952 }, { "epoch": 1.6420846394984325, "grad_norm": 3.6501271724700928, "learning_rate": 6.404546966625362e-06, "loss": 1.234, "step": 20953 }, { "epoch": 1.6421630094043889, "grad_norm": 4.005332946777344, "learning_rate": 6.404251018312177e-06, "loss": 1.195, "step": 20954 }, { "epoch": 1.6422413793103448, "grad_norm": 4.460129261016846, "learning_rate": 6.403955064657943e-06, "loss": 2.4797, "step": 20955 }, { "epoch": 1.642319749216301, "grad_norm": 1.8280905485153198, "learning_rate": 6.403659105663783e-06, "loss": 1.3738, "step": 20956 }, { "epoch": 1.642398119122257, "grad_norm": 5.138123989105225, "learning_rate": 6.403363141330826e-06, "loss": 1.1102, "step": 20957 }, { "epoch": 1.6424764890282133, "grad_norm": 2.510070562362671, "learning_rate": 6.403067171660196e-06, "loss": 1.3047, "step": 20958 }, { "epoch": 1.6425548589341692, "grad_norm": 2.038646936416626, "learning_rate": 6.402771196653019e-06, "loss": 1.1683, "step": 20959 }, { "epoch": 1.6426332288401255, "grad_norm": 3.448668956756592, "learning_rate": 6.402475216310422e-06, "loss": 1.0751, "step": 20960 }, { "epoch": 1.6427115987460814, "grad_norm": 2.1532394886016846, "learning_rate": 6.402179230633529e-06, "loss": 1.2309, "step": 20961 }, { "epoch": 1.6427899686520377, "grad_norm": 3.3265256881713867, "learning_rate": 6.4018832396234674e-06, "loss": 1.4224, "step": 20962 }, { "epoch": 1.6428683385579936, "grad_norm": 6.049869060516357, "learning_rate": 6.40158724328136e-06, "loss": 1.0158, "step": 20963 }, { "epoch": 1.64294670846395, "grad_norm": 4.636307716369629, "learning_rate": 6.4012912416083364e-06, "loss": 1.965, "step": 20964 }, { "epoch": 1.6430250783699059, "grad_norm": 1.919538974761963, "learning_rate": 6.400995234605518e-06, "loss": 1.3087, "step": 20965 }, { "epoch": 1.6431034482758622, "grad_norm": 2.3100125789642334, "learning_rate": 6.400699222274036e-06, "loss": 1.2491, "step": 20966 }, { "epoch": 1.643181818181818, "grad_norm": 7.202815532684326, "learning_rate": 6.400403204615014e-06, "loss": 0.8791, "step": 20967 }, { "epoch": 1.6432601880877744, "grad_norm": 2.966984510421753, "learning_rate": 6.4001071816295755e-06, "loss": 1.3969, "step": 20968 }, { "epoch": 1.6433385579937303, "grad_norm": 4.497280597686768, "learning_rate": 6.39981115331885e-06, "loss": 0.6597, "step": 20969 }, { "epoch": 1.6434169278996866, "grad_norm": 4.507579803466797, "learning_rate": 6.39951511968396e-06, "loss": 1.503, "step": 20970 }, { "epoch": 1.6434952978056425, "grad_norm": 2.507899284362793, "learning_rate": 6.399219080726035e-06, "loss": 1.3575, "step": 20971 }, { "epoch": 1.6435736677115989, "grad_norm": 3.844255208969116, "learning_rate": 6.398923036446198e-06, "loss": 1.5375, "step": 20972 }, { "epoch": 1.6436520376175547, "grad_norm": 3.7735395431518555, "learning_rate": 6.398626986845577e-06, "loss": 1.4514, "step": 20973 }, { "epoch": 1.643730407523511, "grad_norm": 4.805870056152344, "learning_rate": 6.398330931925299e-06, "loss": 1.4194, "step": 20974 }, { "epoch": 1.643808777429467, "grad_norm": 2.516664743423462, "learning_rate": 6.398034871686486e-06, "loss": 1.5632, "step": 20975 }, { "epoch": 1.6438871473354233, "grad_norm": 3.4644932746887207, "learning_rate": 6.397738806130267e-06, "loss": 0.7974, "step": 20976 }, { "epoch": 1.6439655172413792, "grad_norm": 3.4318056106567383, "learning_rate": 6.397442735257767e-06, "loss": 1.6584, "step": 20977 }, { "epoch": 1.6440438871473355, "grad_norm": 2.4531686305999756, "learning_rate": 6.397146659070114e-06, "loss": 1.7774, "step": 20978 }, { "epoch": 1.6441222570532914, "grad_norm": 6.239988803863525, "learning_rate": 6.396850577568432e-06, "loss": 1.0493, "step": 20979 }, { "epoch": 1.6442006269592477, "grad_norm": 2.8548688888549805, "learning_rate": 6.396554490753848e-06, "loss": 1.1272, "step": 20980 }, { "epoch": 1.6442789968652036, "grad_norm": 3.479120969772339, "learning_rate": 6.396258398627488e-06, "loss": 1.3615, "step": 20981 }, { "epoch": 1.64435736677116, "grad_norm": 3.9308156967163086, "learning_rate": 6.395962301190478e-06, "loss": 1.8584, "step": 20982 }, { "epoch": 1.6444357366771158, "grad_norm": 4.787549018859863, "learning_rate": 6.3956661984439435e-06, "loss": 1.5119, "step": 20983 }, { "epoch": 1.6445141065830722, "grad_norm": 2.4501965045928955, "learning_rate": 6.395370090389014e-06, "loss": 1.3147, "step": 20984 }, { "epoch": 1.644592476489028, "grad_norm": 3.835063934326172, "learning_rate": 6.395073977026811e-06, "loss": 1.4766, "step": 20985 }, { "epoch": 1.6446708463949844, "grad_norm": 3.203030586242676, "learning_rate": 6.394777858358465e-06, "loss": 1.4318, "step": 20986 }, { "epoch": 1.6447492163009403, "grad_norm": 3.0977790355682373, "learning_rate": 6.394481734385099e-06, "loss": 1.4988, "step": 20987 }, { "epoch": 1.6448275862068966, "grad_norm": 2.621947765350342, "learning_rate": 6.394185605107841e-06, "loss": 1.5519, "step": 20988 }, { "epoch": 1.6449059561128525, "grad_norm": 4.578357696533203, "learning_rate": 6.393889470527818e-06, "loss": 1.1872, "step": 20989 }, { "epoch": 1.6449843260188088, "grad_norm": 4.226833343505859, "learning_rate": 6.393593330646155e-06, "loss": 1.4505, "step": 20990 }, { "epoch": 1.6450626959247647, "grad_norm": 2.605816602706909, "learning_rate": 6.393297185463978e-06, "loss": 1.4802, "step": 20991 }, { "epoch": 1.645141065830721, "grad_norm": 5.310192108154297, "learning_rate": 6.393001034982416e-06, "loss": 0.8296, "step": 20992 }, { "epoch": 1.645219435736677, "grad_norm": 2.1637754440307617, "learning_rate": 6.39270487920259e-06, "loss": 1.1376, "step": 20993 }, { "epoch": 1.6452978056426333, "grad_norm": 4.131079196929932, "learning_rate": 6.392408718125633e-06, "loss": 1.3296, "step": 20994 }, { "epoch": 1.6453761755485894, "grad_norm": 2.0263431072235107, "learning_rate": 6.392112551752668e-06, "loss": 1.2297, "step": 20995 }, { "epoch": 1.6454545454545455, "grad_norm": 2.0111491680145264, "learning_rate": 6.391816380084822e-06, "loss": 0.8712, "step": 20996 }, { "epoch": 1.6455329153605016, "grad_norm": 3.666245460510254, "learning_rate": 6.3915202031232215e-06, "loss": 1.3266, "step": 20997 }, { "epoch": 1.6456112852664577, "grad_norm": 4.769146919250488, "learning_rate": 6.391224020868992e-06, "loss": 1.7106, "step": 20998 }, { "epoch": 1.6456896551724138, "grad_norm": 3.289175510406494, "learning_rate": 6.390927833323261e-06, "loss": 1.0696, "step": 20999 }, { "epoch": 1.64576802507837, "grad_norm": 2.6774110794067383, "learning_rate": 6.390631640487155e-06, "loss": 1.189, "step": 21000 }, { "epoch": 1.645846394984326, "grad_norm": 3.4130218029022217, "learning_rate": 6.390335442361802e-06, "loss": 1.2598, "step": 21001 }, { "epoch": 1.6459247648902822, "grad_norm": 3.272852659225464, "learning_rate": 6.390039238948325e-06, "loss": 1.1872, "step": 21002 }, { "epoch": 1.6460031347962383, "grad_norm": 2.2474024295806885, "learning_rate": 6.3897430302478524e-06, "loss": 1.4479, "step": 21003 }, { "epoch": 1.6460815047021944, "grad_norm": 2.6820690631866455, "learning_rate": 6.389446816261513e-06, "loss": 1.2009, "step": 21004 }, { "epoch": 1.6461598746081505, "grad_norm": 4.354439735412598, "learning_rate": 6.38915059699043e-06, "loss": 1.4708, "step": 21005 }, { "epoch": 1.6462382445141066, "grad_norm": 2.442283868789673, "learning_rate": 6.388854372435733e-06, "loss": 1.4864, "step": 21006 }, { "epoch": 1.6463166144200627, "grad_norm": 2.9577207565307617, "learning_rate": 6.388558142598547e-06, "loss": 1.3953, "step": 21007 }, { "epoch": 1.6463949843260188, "grad_norm": 2.835421323776245, "learning_rate": 6.388261907479998e-06, "loss": 1.4947, "step": 21008 }, { "epoch": 1.646473354231975, "grad_norm": 4.524563789367676, "learning_rate": 6.387965667081215e-06, "loss": 1.4649, "step": 21009 }, { "epoch": 1.646551724137931, "grad_norm": 7.888511657714844, "learning_rate": 6.387669421403324e-06, "loss": 1.3504, "step": 21010 }, { "epoch": 1.6466300940438872, "grad_norm": 6.055876731872559, "learning_rate": 6.387373170447449e-06, "loss": 1.263, "step": 21011 }, { "epoch": 1.6467084639498433, "grad_norm": 2.6630358695983887, "learning_rate": 6.3870769142147204e-06, "loss": 1.3408, "step": 21012 }, { "epoch": 1.6467868338557994, "grad_norm": 5.457272529602051, "learning_rate": 6.386780652706264e-06, "loss": 1.5451, "step": 21013 }, { "epoch": 1.6468652037617555, "grad_norm": 2.369293689727783, "learning_rate": 6.386484385923206e-06, "loss": 1.3909, "step": 21014 }, { "epoch": 1.6469435736677116, "grad_norm": 2.495300054550171, "learning_rate": 6.3861881138666734e-06, "loss": 1.5835, "step": 21015 }, { "epoch": 1.6470219435736677, "grad_norm": 2.8132011890411377, "learning_rate": 6.3858918365377945e-06, "loss": 1.4311, "step": 21016 }, { "epoch": 1.6471003134796238, "grad_norm": 2.2559683322906494, "learning_rate": 6.385595553937692e-06, "loss": 1.0581, "step": 21017 }, { "epoch": 1.64717868338558, "grad_norm": 2.2586991786956787, "learning_rate": 6.3852992660674985e-06, "loss": 1.1959, "step": 21018 }, { "epoch": 1.647257053291536, "grad_norm": 4.393288612365723, "learning_rate": 6.3850029729283355e-06, "loss": 0.9204, "step": 21019 }, { "epoch": 1.6473354231974922, "grad_norm": 2.186633348464966, "learning_rate": 6.384706674521335e-06, "loss": 1.3316, "step": 21020 }, { "epoch": 1.6474137931034483, "grad_norm": 4.007196426391602, "learning_rate": 6.384410370847621e-06, "loss": 1.3813, "step": 21021 }, { "epoch": 1.6474921630094044, "grad_norm": 2.3266303539276123, "learning_rate": 6.3841140619083205e-06, "loss": 1.3489, "step": 21022 }, { "epoch": 1.6475705329153605, "grad_norm": 12.676495552062988, "learning_rate": 6.383817747704562e-06, "loss": 1.4072, "step": 21023 }, { "epoch": 1.6476489028213166, "grad_norm": 3.5085394382476807, "learning_rate": 6.383521428237469e-06, "loss": 1.3468, "step": 21024 }, { "epoch": 1.6477272727272727, "grad_norm": 4.752745628356934, "learning_rate": 6.383225103508175e-06, "loss": 1.425, "step": 21025 }, { "epoch": 1.6478056426332288, "grad_norm": 6.697200775146484, "learning_rate": 6.382928773517799e-06, "loss": 1.6757, "step": 21026 }, { "epoch": 1.647884012539185, "grad_norm": 3.1540300846099854, "learning_rate": 6.382632438267475e-06, "loss": 1.1142, "step": 21027 }, { "epoch": 1.647962382445141, "grad_norm": 2.7072737216949463, "learning_rate": 6.382336097758327e-06, "loss": 1.2577, "step": 21028 }, { "epoch": 1.6480407523510971, "grad_norm": 4.991259574890137, "learning_rate": 6.3820397519914825e-06, "loss": 0.9734, "step": 21029 }, { "epoch": 1.6481191222570533, "grad_norm": 3.515424966812134, "learning_rate": 6.381743400968068e-06, "loss": 1.1942, "step": 21030 }, { "epoch": 1.6481974921630094, "grad_norm": 3.78818678855896, "learning_rate": 6.381447044689212e-06, "loss": 1.7362, "step": 21031 }, { "epoch": 1.6482758620689655, "grad_norm": 2.34808611869812, "learning_rate": 6.38115068315604e-06, "loss": 1.3075, "step": 21032 }, { "epoch": 1.6483542319749216, "grad_norm": 3.0612587928771973, "learning_rate": 6.38085431636968e-06, "loss": 1.3543, "step": 21033 }, { "epoch": 1.6484326018808777, "grad_norm": 2.365074634552002, "learning_rate": 6.3805579443312605e-06, "loss": 1.3418, "step": 21034 }, { "epoch": 1.6485109717868338, "grad_norm": 3.525501012802124, "learning_rate": 6.380261567041908e-06, "loss": 1.6865, "step": 21035 }, { "epoch": 1.64858934169279, "grad_norm": 2.638252019882202, "learning_rate": 6.379965184502749e-06, "loss": 0.8501, "step": 21036 }, { "epoch": 1.648667711598746, "grad_norm": 5.760338306427002, "learning_rate": 6.379668796714912e-06, "loss": 1.0194, "step": 21037 }, { "epoch": 1.6487460815047021, "grad_norm": 2.4933972358703613, "learning_rate": 6.379372403679522e-06, "loss": 1.3024, "step": 21038 }, { "epoch": 1.6488244514106583, "grad_norm": 6.624270915985107, "learning_rate": 6.379076005397709e-06, "loss": 1.3011, "step": 21039 }, { "epoch": 1.6489028213166144, "grad_norm": 3.435570478439331, "learning_rate": 6.378779601870598e-06, "loss": 1.2625, "step": 21040 }, { "epoch": 1.6489811912225707, "grad_norm": 3.967144012451172, "learning_rate": 6.378483193099318e-06, "loss": 0.8861, "step": 21041 }, { "epoch": 1.6490595611285266, "grad_norm": 2.3395261764526367, "learning_rate": 6.378186779084996e-06, "loss": 1.3157, "step": 21042 }, { "epoch": 1.649137931034483, "grad_norm": 2.6241087913513184, "learning_rate": 6.37789035982876e-06, "loss": 1.2581, "step": 21043 }, { "epoch": 1.6492163009404388, "grad_norm": 3.5248892307281494, "learning_rate": 6.377593935331737e-06, "loss": 1.4386, "step": 21044 }, { "epoch": 1.6492946708463951, "grad_norm": 2.5682685375213623, "learning_rate": 6.377297505595052e-06, "loss": 1.3268, "step": 21045 }, { "epoch": 1.649373040752351, "grad_norm": 5.817646503448486, "learning_rate": 6.377001070619838e-06, "loss": 1.1682, "step": 21046 }, { "epoch": 1.6494514106583074, "grad_norm": 2.3565638065338135, "learning_rate": 6.3767046304072165e-06, "loss": 1.4949, "step": 21047 }, { "epoch": 1.6495297805642632, "grad_norm": 2.435333251953125, "learning_rate": 6.37640818495832e-06, "loss": 1.5771, "step": 21048 }, { "epoch": 1.6496081504702196, "grad_norm": 4.1372528076171875, "learning_rate": 6.376111734274273e-06, "loss": 0.9639, "step": 21049 }, { "epoch": 1.6496865203761755, "grad_norm": 8.35297679901123, "learning_rate": 6.375815278356203e-06, "loss": 1.8251, "step": 21050 }, { "epoch": 1.6497648902821318, "grad_norm": 1.7587624788284302, "learning_rate": 6.37551881720524e-06, "loss": 1.408, "step": 21051 }, { "epoch": 1.6498432601880877, "grad_norm": 2.1692733764648438, "learning_rate": 6.375222350822507e-06, "loss": 1.1873, "step": 21052 }, { "epoch": 1.649921630094044, "grad_norm": 3.794163942337036, "learning_rate": 6.374925879209138e-06, "loss": 1.7367, "step": 21053 }, { "epoch": 1.65, "grad_norm": 4.2226762771606445, "learning_rate": 6.3746294023662524e-06, "loss": 1.3569, "step": 21054 }, { "epoch": 1.6500783699059562, "grad_norm": 2.0778045654296875, "learning_rate": 6.374332920294987e-06, "loss": 1.4261, "step": 21055 }, { "epoch": 1.6501567398119121, "grad_norm": 2.591062307357788, "learning_rate": 6.3740364329964634e-06, "loss": 1.0514, "step": 21056 }, { "epoch": 1.6502351097178685, "grad_norm": 3.5199732780456543, "learning_rate": 6.373739940471812e-06, "loss": 1.277, "step": 21057 }, { "epoch": 1.6503134796238244, "grad_norm": 8.288809776306152, "learning_rate": 6.373443442722158e-06, "loss": 1.2216, "step": 21058 }, { "epoch": 1.6503918495297807, "grad_norm": 7.643355369567871, "learning_rate": 6.373146939748632e-06, "loss": 1.588, "step": 21059 }, { "epoch": 1.6504702194357366, "grad_norm": 3.2071409225463867, "learning_rate": 6.372850431552358e-06, "loss": 1.4138, "step": 21060 }, { "epoch": 1.650548589341693, "grad_norm": 3.327024459838867, "learning_rate": 6.372553918134468e-06, "loss": 1.5776, "step": 21061 }, { "epoch": 1.6506269592476488, "grad_norm": 2.395437717437744, "learning_rate": 6.372257399496087e-06, "loss": 1.4297, "step": 21062 }, { "epoch": 1.6507053291536051, "grad_norm": 2.8470382690429688, "learning_rate": 6.371960875638346e-06, "loss": 1.3127, "step": 21063 }, { "epoch": 1.650783699059561, "grad_norm": 5.50128698348999, "learning_rate": 6.371664346562368e-06, "loss": 1.3918, "step": 21064 }, { "epoch": 1.6508620689655173, "grad_norm": 3.092773675918579, "learning_rate": 6.371367812269284e-06, "loss": 1.4564, "step": 21065 }, { "epoch": 1.6509404388714732, "grad_norm": 3.0222432613372803, "learning_rate": 6.371071272760221e-06, "loss": 1.4584, "step": 21066 }, { "epoch": 1.6510188087774296, "grad_norm": 2.167551279067993, "learning_rate": 6.370774728036307e-06, "loss": 1.2764, "step": 21067 }, { "epoch": 1.6510971786833855, "grad_norm": 2.854295253753662, "learning_rate": 6.3704781780986705e-06, "loss": 1.1988, "step": 21068 }, { "epoch": 1.6511755485893418, "grad_norm": 3.06048321723938, "learning_rate": 6.370181622948439e-06, "loss": 0.8921, "step": 21069 }, { "epoch": 1.6512539184952977, "grad_norm": 3.8062517642974854, "learning_rate": 6.369885062586741e-06, "loss": 1.5196, "step": 21070 }, { "epoch": 1.651332288401254, "grad_norm": 3.260383129119873, "learning_rate": 6.369588497014701e-06, "loss": 0.8752, "step": 21071 }, { "epoch": 1.65141065830721, "grad_norm": 2.719414472579956, "learning_rate": 6.369291926233454e-06, "loss": 0.8566, "step": 21072 }, { "epoch": 1.6514890282131662, "grad_norm": 2.607250690460205, "learning_rate": 6.368995350244121e-06, "loss": 1.2628, "step": 21073 }, { "epoch": 1.6515673981191221, "grad_norm": 6.922784328460693, "learning_rate": 6.368698769047834e-06, "loss": 1.0086, "step": 21074 }, { "epoch": 1.6516457680250785, "grad_norm": 6.259767532348633, "learning_rate": 6.368402182645719e-06, "loss": 1.0874, "step": 21075 }, { "epoch": 1.6517241379310343, "grad_norm": 2.7682878971099854, "learning_rate": 6.368105591038906e-06, "loss": 1.4953, "step": 21076 }, { "epoch": 1.6518025078369907, "grad_norm": 6.712653160095215, "learning_rate": 6.367808994228521e-06, "loss": 1.492, "step": 21077 }, { "epoch": 1.6518808777429466, "grad_norm": 3.7649240493774414, "learning_rate": 6.367512392215693e-06, "loss": 1.0211, "step": 21078 }, { "epoch": 1.651959247648903, "grad_norm": 2.6237261295318604, "learning_rate": 6.3672157850015515e-06, "loss": 1.5529, "step": 21079 }, { "epoch": 1.6520376175548588, "grad_norm": 2.551036834716797, "learning_rate": 6.366919172587221e-06, "loss": 1.5978, "step": 21080 }, { "epoch": 1.6521159874608151, "grad_norm": 3.2811808586120605, "learning_rate": 6.366622554973836e-06, "loss": 1.1793, "step": 21081 }, { "epoch": 1.652194357366771, "grad_norm": 2.0163722038269043, "learning_rate": 6.3663259321625185e-06, "loss": 1.1097, "step": 21082 }, { "epoch": 1.6522727272727273, "grad_norm": 2.2313430309295654, "learning_rate": 6.366029304154399e-06, "loss": 1.255, "step": 21083 }, { "epoch": 1.6523510971786832, "grad_norm": 2.3160126209259033, "learning_rate": 6.365732670950605e-06, "loss": 1.3392, "step": 21084 }, { "epoch": 1.6524294670846396, "grad_norm": 1.8349791765213013, "learning_rate": 6.365436032552266e-06, "loss": 1.3842, "step": 21085 }, { "epoch": 1.6525078369905954, "grad_norm": 2.789858818054199, "learning_rate": 6.3651393889605095e-06, "loss": 1.0686, "step": 21086 }, { "epoch": 1.6525862068965518, "grad_norm": 2.661885976791382, "learning_rate": 6.364842740176464e-06, "loss": 1.178, "step": 21087 }, { "epoch": 1.652664576802508, "grad_norm": 3.2613492012023926, "learning_rate": 6.364546086201258e-06, "loss": 1.489, "step": 21088 }, { "epoch": 1.652742946708464, "grad_norm": 2.366556406021118, "learning_rate": 6.364249427036019e-06, "loss": 1.4953, "step": 21089 }, { "epoch": 1.65282131661442, "grad_norm": 3.2027745246887207, "learning_rate": 6.363952762681877e-06, "loss": 1.2996, "step": 21090 }, { "epoch": 1.6528996865203762, "grad_norm": 2.961869478225708, "learning_rate": 6.363656093139958e-06, "loss": 0.9433, "step": 21091 }, { "epoch": 1.6529780564263323, "grad_norm": 3.9625110626220703, "learning_rate": 6.363359418411392e-06, "loss": 1.4345, "step": 21092 }, { "epoch": 1.6530564263322884, "grad_norm": 2.2463505268096924, "learning_rate": 6.363062738497308e-06, "loss": 1.4688, "step": 21093 }, { "epoch": 1.6531347962382446, "grad_norm": 2.9470527172088623, "learning_rate": 6.362766053398832e-06, "loss": 1.34, "step": 21094 }, { "epoch": 1.6532131661442007, "grad_norm": 2.7466838359832764, "learning_rate": 6.362469363117095e-06, "loss": 0.7095, "step": 21095 }, { "epoch": 1.6532915360501568, "grad_norm": 3.214240312576294, "learning_rate": 6.362172667653225e-06, "loss": 1.298, "step": 21096 }, { "epoch": 1.6533699059561129, "grad_norm": 5.709709644317627, "learning_rate": 6.361875967008348e-06, "loss": 1.1027, "step": 21097 }, { "epoch": 1.653448275862069, "grad_norm": 2.7675893306732178, "learning_rate": 6.3615792611835955e-06, "loss": 1.2616, "step": 21098 }, { "epoch": 1.653526645768025, "grad_norm": 4.847761631011963, "learning_rate": 6.3612825501800944e-06, "loss": 1.4141, "step": 21099 }, { "epoch": 1.6536050156739812, "grad_norm": 2.2793405055999756, "learning_rate": 6.360985833998974e-06, "loss": 1.5743, "step": 21100 }, { "epoch": 1.6536833855799373, "grad_norm": 2.0130300521850586, "learning_rate": 6.360689112641362e-06, "loss": 1.5357, "step": 21101 }, { "epoch": 1.6537617554858934, "grad_norm": 2.3454556465148926, "learning_rate": 6.360392386108388e-06, "loss": 1.5607, "step": 21102 }, { "epoch": 1.6538401253918495, "grad_norm": 3.8477044105529785, "learning_rate": 6.360095654401179e-06, "loss": 1.7679, "step": 21103 }, { "epoch": 1.6539184952978057, "grad_norm": 5.231746673583984, "learning_rate": 6.359798917520867e-06, "loss": 1.2304, "step": 21104 }, { "epoch": 1.6539968652037618, "grad_norm": 1.9238535165786743, "learning_rate": 6.359502175468577e-06, "loss": 1.3259, "step": 21105 }, { "epoch": 1.6540752351097179, "grad_norm": 5.731003761291504, "learning_rate": 6.359205428245439e-06, "loss": 0.816, "step": 21106 }, { "epoch": 1.654153605015674, "grad_norm": 2.5548832416534424, "learning_rate": 6.358908675852582e-06, "loss": 1.7642, "step": 21107 }, { "epoch": 1.65423197492163, "grad_norm": 3.3845458030700684, "learning_rate": 6.3586119182911335e-06, "loss": 1.2727, "step": 21108 }, { "epoch": 1.6543103448275862, "grad_norm": 3.4776947498321533, "learning_rate": 6.358315155562224e-06, "loss": 1.0676, "step": 21109 }, { "epoch": 1.6543887147335423, "grad_norm": 3.1872010231018066, "learning_rate": 6.358018387666981e-06, "loss": 1.3536, "step": 21110 }, { "epoch": 1.6544670846394984, "grad_norm": 2.9500772953033447, "learning_rate": 6.357721614606533e-06, "loss": 1.5426, "step": 21111 }, { "epoch": 1.6545454545454545, "grad_norm": 2.754807949066162, "learning_rate": 6.35742483638201e-06, "loss": 1.5236, "step": 21112 }, { "epoch": 1.6546238244514107, "grad_norm": 3.5521047115325928, "learning_rate": 6.35712805299454e-06, "loss": 1.3842, "step": 21113 }, { "epoch": 1.6547021943573668, "grad_norm": 3.192646026611328, "learning_rate": 6.356831264445251e-06, "loss": 1.1434, "step": 21114 }, { "epoch": 1.6547805642633229, "grad_norm": 2.7623345851898193, "learning_rate": 6.356534470735274e-06, "loss": 1.3426, "step": 21115 }, { "epoch": 1.654858934169279, "grad_norm": 4.305774688720703, "learning_rate": 6.356237671865737e-06, "loss": 1.7649, "step": 21116 }, { "epoch": 1.654937304075235, "grad_norm": 3.0738179683685303, "learning_rate": 6.355940867837769e-06, "loss": 1.4427, "step": 21117 }, { "epoch": 1.6550156739811912, "grad_norm": 2.274454116821289, "learning_rate": 6.355644058652496e-06, "loss": 1.3622, "step": 21118 }, { "epoch": 1.6550940438871473, "grad_norm": 2.300029993057251, "learning_rate": 6.355347244311052e-06, "loss": 1.3185, "step": 21119 }, { "epoch": 1.6551724137931034, "grad_norm": 3.923229932785034, "learning_rate": 6.3550504248145595e-06, "loss": 1.2621, "step": 21120 }, { "epoch": 1.6552507836990595, "grad_norm": 3.1874771118164062, "learning_rate": 6.354753600164155e-06, "loss": 1.1886, "step": 21121 }, { "epoch": 1.6553291536050156, "grad_norm": 7.097949981689453, "learning_rate": 6.354456770360962e-06, "loss": 0.9238, "step": 21122 }, { "epoch": 1.6554075235109718, "grad_norm": 3.322714328765869, "learning_rate": 6.354159935406111e-06, "loss": 1.4936, "step": 21123 }, { "epoch": 1.6554858934169279, "grad_norm": 3.1633520126342773, "learning_rate": 6.353863095300733e-06, "loss": 1.2784, "step": 21124 }, { "epoch": 1.655564263322884, "grad_norm": 2.0727367401123047, "learning_rate": 6.353566250045953e-06, "loss": 1.5118, "step": 21125 }, { "epoch": 1.65564263322884, "grad_norm": 5.640665054321289, "learning_rate": 6.3532693996429025e-06, "loss": 1.3706, "step": 21126 }, { "epoch": 1.6557210031347962, "grad_norm": 4.741607666015625, "learning_rate": 6.352972544092711e-06, "loss": 1.646, "step": 21127 }, { "epoch": 1.6557993730407523, "grad_norm": 2.2556779384613037, "learning_rate": 6.3526756833965065e-06, "loss": 1.1225, "step": 21128 }, { "epoch": 1.6558777429467084, "grad_norm": 2.8157660961151123, "learning_rate": 6.352378817555418e-06, "loss": 1.6783, "step": 21129 }, { "epoch": 1.6559561128526645, "grad_norm": 3.657871723175049, "learning_rate": 6.352081946570577e-06, "loss": 1.1853, "step": 21130 }, { "epoch": 1.6560344827586206, "grad_norm": 2.6624820232391357, "learning_rate": 6.3517850704431085e-06, "loss": 1.2992, "step": 21131 }, { "epoch": 1.6561128526645768, "grad_norm": 2.260465145111084, "learning_rate": 6.351488189174144e-06, "loss": 0.9705, "step": 21132 }, { "epoch": 1.6561912225705329, "grad_norm": 3.8773574829101562, "learning_rate": 6.3511913027648134e-06, "loss": 1.061, "step": 21133 }, { "epoch": 1.6562695924764892, "grad_norm": 5.536136150360107, "learning_rate": 6.350894411216245e-06, "loss": 1.2249, "step": 21134 }, { "epoch": 1.656347962382445, "grad_norm": 3.1806533336639404, "learning_rate": 6.350597514529568e-06, "loss": 1.4515, "step": 21135 }, { "epoch": 1.6564263322884014, "grad_norm": 5.25092887878418, "learning_rate": 6.350300612705913e-06, "loss": 1.5926, "step": 21136 }, { "epoch": 1.6565047021943573, "grad_norm": 2.393298387527466, "learning_rate": 6.350003705746407e-06, "loss": 1.3285, "step": 21137 }, { "epoch": 1.6565830721003136, "grad_norm": 3.255038022994995, "learning_rate": 6.34970679365218e-06, "loss": 1.6356, "step": 21138 }, { "epoch": 1.6566614420062695, "grad_norm": 2.825274705886841, "learning_rate": 6.349409876424361e-06, "loss": 1.2023, "step": 21139 }, { "epoch": 1.6567398119122259, "grad_norm": 2.8036293983459473, "learning_rate": 6.3491129540640815e-06, "loss": 1.3103, "step": 21140 }, { "epoch": 1.6568181818181817, "grad_norm": 3.541033983230591, "learning_rate": 6.3488160265724685e-06, "loss": 1.6219, "step": 21141 }, { "epoch": 1.656896551724138, "grad_norm": 4.949773788452148, "learning_rate": 6.348519093950651e-06, "loss": 1.4649, "step": 21142 }, { "epoch": 1.656974921630094, "grad_norm": 3.574305772781372, "learning_rate": 6.348222156199762e-06, "loss": 1.5044, "step": 21143 }, { "epoch": 1.6570532915360503, "grad_norm": 2.9254438877105713, "learning_rate": 6.347925213320927e-06, "loss": 1.4134, "step": 21144 }, { "epoch": 1.6571316614420062, "grad_norm": 5.118002414703369, "learning_rate": 6.347628265315278e-06, "loss": 1.1599, "step": 21145 }, { "epoch": 1.6572100313479625, "grad_norm": 3.514101505279541, "learning_rate": 6.347331312183941e-06, "loss": 1.318, "step": 21146 }, { "epoch": 1.6572884012539184, "grad_norm": 2.895254135131836, "learning_rate": 6.34703435392805e-06, "loss": 1.2883, "step": 21147 }, { "epoch": 1.6573667711598747, "grad_norm": 3.1757652759552, "learning_rate": 6.346737390548729e-06, "loss": 1.5015, "step": 21148 }, { "epoch": 1.6574451410658306, "grad_norm": 6.0175862312316895, "learning_rate": 6.346440422047114e-06, "loss": 1.7298, "step": 21149 }, { "epoch": 1.657523510971787, "grad_norm": 2.6367530822753906, "learning_rate": 6.34614344842433e-06, "loss": 1.5769, "step": 21150 }, { "epoch": 1.6576018808777429, "grad_norm": 3.466700792312622, "learning_rate": 6.345846469681508e-06, "loss": 1.4186, "step": 21151 }, { "epoch": 1.6576802507836992, "grad_norm": 5.524541854858398, "learning_rate": 6.3455494858197774e-06, "loss": 1.1876, "step": 21152 }, { "epoch": 1.657758620689655, "grad_norm": 2.7790768146514893, "learning_rate": 6.3452524968402665e-06, "loss": 1.323, "step": 21153 }, { "epoch": 1.6578369905956114, "grad_norm": 4.178127765655518, "learning_rate": 6.344955502744108e-06, "loss": 1.4946, "step": 21154 }, { "epoch": 1.6579153605015673, "grad_norm": 3.5403082370758057, "learning_rate": 6.3446585035324275e-06, "loss": 1.0473, "step": 21155 }, { "epoch": 1.6579937304075236, "grad_norm": 3.194875717163086, "learning_rate": 6.344361499206358e-06, "loss": 1.1642, "step": 21156 }, { "epoch": 1.6580721003134795, "grad_norm": 2.323539972305298, "learning_rate": 6.3440644897670275e-06, "loss": 1.5006, "step": 21157 }, { "epoch": 1.6581504702194358, "grad_norm": 2.1722934246063232, "learning_rate": 6.3437674752155665e-06, "loss": 1.3666, "step": 21158 }, { "epoch": 1.6582288401253917, "grad_norm": 3.8727240562438965, "learning_rate": 6.3434704555531045e-06, "loss": 1.5498, "step": 21159 }, { "epoch": 1.658307210031348, "grad_norm": 4.178417682647705, "learning_rate": 6.343173430780769e-06, "loss": 1.0154, "step": 21160 }, { "epoch": 1.658385579937304, "grad_norm": 2.1140947341918945, "learning_rate": 6.342876400899694e-06, "loss": 1.4365, "step": 21161 }, { "epoch": 1.6584639498432603, "grad_norm": 2.914828300476074, "learning_rate": 6.342579365911006e-06, "loss": 1.4156, "step": 21162 }, { "epoch": 1.6585423197492162, "grad_norm": 2.0545740127563477, "learning_rate": 6.342282325815836e-06, "loss": 1.353, "step": 21163 }, { "epoch": 1.6586206896551725, "grad_norm": 2.684751272201538, "learning_rate": 6.341985280615312e-06, "loss": 1.5331, "step": 21164 }, { "epoch": 1.6586990595611284, "grad_norm": 2.2777233123779297, "learning_rate": 6.341688230310568e-06, "loss": 1.4649, "step": 21165 }, { "epoch": 1.6587774294670847, "grad_norm": 3.282972574234009, "learning_rate": 6.341391174902728e-06, "loss": 1.0192, "step": 21166 }, { "epoch": 1.6588557993730406, "grad_norm": 2.7821619510650635, "learning_rate": 6.3410941143929265e-06, "loss": 1.5065, "step": 21167 }, { "epoch": 1.658934169278997, "grad_norm": 2.436014413833618, "learning_rate": 6.3407970487822915e-06, "loss": 1.2094, "step": 21168 }, { "epoch": 1.6590125391849528, "grad_norm": 2.707636594772339, "learning_rate": 6.340499978071953e-06, "loss": 0.9878, "step": 21169 }, { "epoch": 1.6590909090909092, "grad_norm": 2.5842068195343018, "learning_rate": 6.3402029022630415e-06, "loss": 1.6914, "step": 21170 }, { "epoch": 1.659169278996865, "grad_norm": 3.2265126705169678, "learning_rate": 6.339905821356688e-06, "loss": 1.4098, "step": 21171 }, { "epoch": 1.6592476489028214, "grad_norm": 2.969433307647705, "learning_rate": 6.339608735354018e-06, "loss": 1.0844, "step": 21172 }, { "epoch": 1.6593260188087773, "grad_norm": 2.995741128921509, "learning_rate": 6.339311644256167e-06, "loss": 1.455, "step": 21173 }, { "epoch": 1.6594043887147336, "grad_norm": 1.929498553276062, "learning_rate": 6.33901454806426e-06, "loss": 1.3843, "step": 21174 }, { "epoch": 1.6594827586206895, "grad_norm": 2.0787079334259033, "learning_rate": 6.338717446779432e-06, "loss": 1.6123, "step": 21175 }, { "epoch": 1.6595611285266458, "grad_norm": 7.358410358428955, "learning_rate": 6.338420340402809e-06, "loss": 0.46, "step": 21176 }, { "epoch": 1.6596394984326017, "grad_norm": 5.684992790222168, "learning_rate": 6.338123228935522e-06, "loss": 0.9019, "step": 21177 }, { "epoch": 1.659717868338558, "grad_norm": 3.3791754245758057, "learning_rate": 6.337826112378705e-06, "loss": 1.1229, "step": 21178 }, { "epoch": 1.659796238244514, "grad_norm": 4.730266094207764, "learning_rate": 6.337528990733479e-06, "loss": 1.2582, "step": 21179 }, { "epoch": 1.6598746081504703, "grad_norm": 5.1602911949157715, "learning_rate": 6.337231864000985e-06, "loss": 1.3032, "step": 21180 }, { "epoch": 1.6599529780564264, "grad_norm": 3.285871744155884, "learning_rate": 6.336934732182343e-06, "loss": 1.0031, "step": 21181 }, { "epoch": 1.6600313479623825, "grad_norm": 2.7370455265045166, "learning_rate": 6.336637595278692e-06, "loss": 1.4403, "step": 21182 }, { "epoch": 1.6601097178683386, "grad_norm": 5.904160022735596, "learning_rate": 6.3363404532911565e-06, "loss": 1.6865, "step": 21183 }, { "epoch": 1.6601880877742947, "grad_norm": 12.260773658752441, "learning_rate": 6.336043306220868e-06, "loss": 1.5023, "step": 21184 }, { "epoch": 1.6602664576802508, "grad_norm": 2.853947877883911, "learning_rate": 6.335746154068957e-06, "loss": 1.6562, "step": 21185 }, { "epoch": 1.660344827586207, "grad_norm": 2.8961284160614014, "learning_rate": 6.335448996836554e-06, "loss": 1.599, "step": 21186 }, { "epoch": 1.660423197492163, "grad_norm": 4.202459335327148, "learning_rate": 6.335151834524789e-06, "loss": 1.3182, "step": 21187 }, { "epoch": 1.6605015673981192, "grad_norm": 5.2471466064453125, "learning_rate": 6.334854667134792e-06, "loss": 1.7116, "step": 21188 }, { "epoch": 1.6605799373040753, "grad_norm": 4.315226078033447, "learning_rate": 6.334557494667694e-06, "loss": 1.6283, "step": 21189 }, { "epoch": 1.6606583072100314, "grad_norm": 3.550447702407837, "learning_rate": 6.334260317124623e-06, "loss": 1.4004, "step": 21190 }, { "epoch": 1.6607366771159875, "grad_norm": 2.8510565757751465, "learning_rate": 6.333963134506714e-06, "loss": 1.1813, "step": 21191 }, { "epoch": 1.6608150470219436, "grad_norm": 3.3842411041259766, "learning_rate": 6.333665946815093e-06, "loss": 1.3036, "step": 21192 }, { "epoch": 1.6608934169278997, "grad_norm": 3.6057040691375732, "learning_rate": 6.333368754050892e-06, "loss": 1.4838, "step": 21193 }, { "epoch": 1.6609717868338558, "grad_norm": 2.840223789215088, "learning_rate": 6.33307155621524e-06, "loss": 1.5146, "step": 21194 }, { "epoch": 1.661050156739812, "grad_norm": 2.8903894424438477, "learning_rate": 6.332774353309269e-06, "loss": 1.7761, "step": 21195 }, { "epoch": 1.661128526645768, "grad_norm": 2.657315254211426, "learning_rate": 6.332477145334109e-06, "loss": 1.2034, "step": 21196 }, { "epoch": 1.6612068965517242, "grad_norm": 3.125922203063965, "learning_rate": 6.33217993229089e-06, "loss": 1.3729, "step": 21197 }, { "epoch": 1.6612852664576803, "grad_norm": 3.2724523544311523, "learning_rate": 6.331882714180743e-06, "loss": 1.6031, "step": 21198 }, { "epoch": 1.6613636363636364, "grad_norm": 2.8091678619384766, "learning_rate": 6.331585491004801e-06, "loss": 0.7216, "step": 21199 }, { "epoch": 1.6614420062695925, "grad_norm": 2.0902957916259766, "learning_rate": 6.331288262764188e-06, "loss": 1.2569, "step": 21200 }, { "epoch": 1.6615203761755486, "grad_norm": 17.76266098022461, "learning_rate": 6.330991029460041e-06, "loss": 1.8385, "step": 21201 }, { "epoch": 1.6615987460815047, "grad_norm": 2.4115374088287354, "learning_rate": 6.330693791093484e-06, "loss": 1.3035, "step": 21202 }, { "epoch": 1.6616771159874608, "grad_norm": 2.7214269638061523, "learning_rate": 6.330396547665655e-06, "loss": 1.5185, "step": 21203 }, { "epoch": 1.661755485893417, "grad_norm": 3.278857469558716, "learning_rate": 6.3300992991776786e-06, "loss": 1.4111, "step": 21204 }, { "epoch": 1.661833855799373, "grad_norm": 2.1494081020355225, "learning_rate": 6.3298020456306885e-06, "loss": 1.4739, "step": 21205 }, { "epoch": 1.6619122257053291, "grad_norm": 5.703860282897949, "learning_rate": 6.3295047870258165e-06, "loss": 1.3138, "step": 21206 }, { "epoch": 1.6619905956112853, "grad_norm": 3.030820608139038, "learning_rate": 6.329207523364187e-06, "loss": 1.6836, "step": 21207 }, { "epoch": 1.6620689655172414, "grad_norm": 2.8413426876068115, "learning_rate": 6.328910254646938e-06, "loss": 1.4683, "step": 21208 }, { "epoch": 1.6621473354231975, "grad_norm": 3.1211025714874268, "learning_rate": 6.328612980875195e-06, "loss": 1.3965, "step": 21209 }, { "epoch": 1.6622257053291536, "grad_norm": 9.774949073791504, "learning_rate": 6.328315702050093e-06, "loss": 1.5802, "step": 21210 }, { "epoch": 1.6623040752351097, "grad_norm": 2.691899299621582, "learning_rate": 6.328018418172759e-06, "loss": 1.3795, "step": 21211 }, { "epoch": 1.6623824451410658, "grad_norm": 2.477155923843384, "learning_rate": 6.327721129244324e-06, "loss": 1.6017, "step": 21212 }, { "epoch": 1.662460815047022, "grad_norm": 3.5628345012664795, "learning_rate": 6.327423835265921e-06, "loss": 1.708, "step": 21213 }, { "epoch": 1.662539184952978, "grad_norm": 3.8043060302734375, "learning_rate": 6.327126536238679e-06, "loss": 1.7224, "step": 21214 }, { "epoch": 1.6626175548589341, "grad_norm": 2.284806728363037, "learning_rate": 6.326829232163729e-06, "loss": 1.3352, "step": 21215 }, { "epoch": 1.6626959247648903, "grad_norm": 2.0862200260162354, "learning_rate": 6.326531923042203e-06, "loss": 1.0903, "step": 21216 }, { "epoch": 1.6627742946708464, "grad_norm": 2.111302137374878, "learning_rate": 6.3262346088752305e-06, "loss": 1.501, "step": 21217 }, { "epoch": 1.6628526645768025, "grad_norm": 5.971668243408203, "learning_rate": 6.325937289663943e-06, "loss": 1.5023, "step": 21218 }, { "epoch": 1.6629310344827586, "grad_norm": 3.694734811782837, "learning_rate": 6.325639965409471e-06, "loss": 1.3579, "step": 21219 }, { "epoch": 1.6630094043887147, "grad_norm": 12.268906593322754, "learning_rate": 6.325342636112945e-06, "loss": 1.2221, "step": 21220 }, { "epoch": 1.6630877742946708, "grad_norm": 2.099513292312622, "learning_rate": 6.325045301775498e-06, "loss": 1.3344, "step": 21221 }, { "epoch": 1.663166144200627, "grad_norm": 5.44715690612793, "learning_rate": 6.324747962398258e-06, "loss": 1.6732, "step": 21222 }, { "epoch": 1.663244514106583, "grad_norm": 4.93310546875, "learning_rate": 6.324450617982356e-06, "loss": 1.2476, "step": 21223 }, { "epoch": 1.6633228840125391, "grad_norm": 7.7172393798828125, "learning_rate": 6.324153268528925e-06, "loss": 1.34, "step": 21224 }, { "epoch": 1.6634012539184952, "grad_norm": 2.2517454624176025, "learning_rate": 6.323855914039098e-06, "loss": 1.5754, "step": 21225 }, { "epoch": 1.6634796238244514, "grad_norm": 3.6946024894714355, "learning_rate": 6.323558554513999e-06, "loss": 1.4716, "step": 21226 }, { "epoch": 1.6635579937304075, "grad_norm": 4.672304153442383, "learning_rate": 6.323261189954766e-06, "loss": 1.5329, "step": 21227 }, { "epoch": 1.6636363636363636, "grad_norm": 11.55490779876709, "learning_rate": 6.322963820362524e-06, "loss": 1.1877, "step": 21228 }, { "epoch": 1.66371473354232, "grad_norm": 2.3089020252227783, "learning_rate": 6.32266644573841e-06, "loss": 1.023, "step": 21229 }, { "epoch": 1.6637931034482758, "grad_norm": 2.4046263694763184, "learning_rate": 6.3223690660835505e-06, "loss": 1.3428, "step": 21230 }, { "epoch": 1.6638714733542321, "grad_norm": 2.502232551574707, "learning_rate": 6.322071681399079e-06, "loss": 1.2436, "step": 21231 }, { "epoch": 1.663949843260188, "grad_norm": 4.3749542236328125, "learning_rate": 6.321774291686125e-06, "loss": 1.0337, "step": 21232 }, { "epoch": 1.6640282131661444, "grad_norm": 4.655486583709717, "learning_rate": 6.321476896945822e-06, "loss": 1.5756, "step": 21233 }, { "epoch": 1.6641065830721002, "grad_norm": 3.105135202407837, "learning_rate": 6.321179497179299e-06, "loss": 0.8237, "step": 21234 }, { "epoch": 1.6641849529780566, "grad_norm": 3.3868305683135986, "learning_rate": 6.320882092387687e-06, "loss": 1.3622, "step": 21235 }, { "epoch": 1.6642633228840125, "grad_norm": 3.47039794921875, "learning_rate": 6.320584682572119e-06, "loss": 1.3768, "step": 21236 }, { "epoch": 1.6643416927899688, "grad_norm": 2.475285768508911, "learning_rate": 6.3202872677337244e-06, "loss": 0.9433, "step": 21237 }, { "epoch": 1.6644200626959247, "grad_norm": 2.716416358947754, "learning_rate": 6.3199898478736355e-06, "loss": 1.6826, "step": 21238 }, { "epoch": 1.664498432601881, "grad_norm": 2.3559153079986572, "learning_rate": 6.319692422992983e-06, "loss": 1.2728, "step": 21239 }, { "epoch": 1.664576802507837, "grad_norm": 3.7465996742248535, "learning_rate": 6.319394993092899e-06, "loss": 1.2876, "step": 21240 }, { "epoch": 1.6646551724137932, "grad_norm": 2.478262186050415, "learning_rate": 6.319097558174514e-06, "loss": 1.646, "step": 21241 }, { "epoch": 1.6647335423197491, "grad_norm": 6.8479533195495605, "learning_rate": 6.318800118238959e-06, "loss": 1.3855, "step": 21242 }, { "epoch": 1.6648119122257055, "grad_norm": 1.777091383934021, "learning_rate": 6.318502673287366e-06, "loss": 1.0142, "step": 21243 }, { "epoch": 1.6648902821316613, "grad_norm": 3.048487663269043, "learning_rate": 6.318205223320865e-06, "loss": 1.253, "step": 21244 }, { "epoch": 1.6649686520376177, "grad_norm": 7.0086870193481445, "learning_rate": 6.31790776834059e-06, "loss": 1.2468, "step": 21245 }, { "epoch": 1.6650470219435736, "grad_norm": 3.949019193649292, "learning_rate": 6.317610308347671e-06, "loss": 1.3479, "step": 21246 }, { "epoch": 1.66512539184953, "grad_norm": 2.3804028034210205, "learning_rate": 6.317312843343237e-06, "loss": 1.2301, "step": 21247 }, { "epoch": 1.6652037617554858, "grad_norm": 3.2613251209259033, "learning_rate": 6.317015373328423e-06, "loss": 1.2834, "step": 21248 }, { "epoch": 1.6652821316614421, "grad_norm": 2.4457781314849854, "learning_rate": 6.316717898304359e-06, "loss": 1.4016, "step": 21249 }, { "epoch": 1.665360501567398, "grad_norm": 2.205212116241455, "learning_rate": 6.316420418272176e-06, "loss": 1.2837, "step": 21250 }, { "epoch": 1.6654388714733543, "grad_norm": 4.435801029205322, "learning_rate": 6.3161229332330065e-06, "loss": 1.7184, "step": 21251 }, { "epoch": 1.6655172413793102, "grad_norm": 2.851666212081909, "learning_rate": 6.315825443187981e-06, "loss": 0.9777, "step": 21252 }, { "epoch": 1.6655956112852666, "grad_norm": 3.824082136154175, "learning_rate": 6.315527948138232e-06, "loss": 1.3586, "step": 21253 }, { "epoch": 1.6656739811912225, "grad_norm": 2.2288670539855957, "learning_rate": 6.315230448084888e-06, "loss": 1.5168, "step": 21254 }, { "epoch": 1.6657523510971788, "grad_norm": 1.847184181213379, "learning_rate": 6.314932943029086e-06, "loss": 1.1366, "step": 21255 }, { "epoch": 1.6658307210031347, "grad_norm": 5.351232528686523, "learning_rate": 6.314635432971952e-06, "loss": 1.1525, "step": 21256 }, { "epoch": 1.665909090909091, "grad_norm": 3.037163019180298, "learning_rate": 6.314337917914623e-06, "loss": 1.5802, "step": 21257 }, { "epoch": 1.665987460815047, "grad_norm": 4.252137660980225, "learning_rate": 6.3140403978582256e-06, "loss": 0.6407, "step": 21258 }, { "epoch": 1.6660658307210032, "grad_norm": 5.30271053314209, "learning_rate": 6.313742872803893e-06, "loss": 0.8579, "step": 21259 }, { "epoch": 1.6661442006269591, "grad_norm": 2.540039539337158, "learning_rate": 6.313445342752759e-06, "loss": 1.6762, "step": 21260 }, { "epoch": 1.6662225705329154, "grad_norm": 3.857863187789917, "learning_rate": 6.31314780770595e-06, "loss": 1.7036, "step": 21261 }, { "epoch": 1.6663009404388713, "grad_norm": 2.7255265712738037, "learning_rate": 6.312850267664605e-06, "loss": 1.4808, "step": 21262 }, { "epoch": 1.6663793103448277, "grad_norm": 2.3530941009521484, "learning_rate": 6.31255272262985e-06, "loss": 1.491, "step": 21263 }, { "epoch": 1.6664576802507836, "grad_norm": 2.198230028152466, "learning_rate": 6.31225517260282e-06, "loss": 1.2633, "step": 21264 }, { "epoch": 1.66653605015674, "grad_norm": 2.282320976257324, "learning_rate": 6.311957617584646e-06, "loss": 1.267, "step": 21265 }, { "epoch": 1.6666144200626958, "grad_norm": 2.2149600982666016, "learning_rate": 6.311660057576456e-06, "loss": 1.3462, "step": 21266 }, { "epoch": 1.666692789968652, "grad_norm": 2.7917709350585938, "learning_rate": 6.311362492579387e-06, "loss": 1.1535, "step": 21267 }, { "epoch": 1.666771159874608, "grad_norm": 2.5618200302124023, "learning_rate": 6.311064922594567e-06, "loss": 1.5547, "step": 21268 }, { "epoch": 1.6668495297805643, "grad_norm": 2.2862675189971924, "learning_rate": 6.310767347623132e-06, "loss": 1.6052, "step": 21269 }, { "epoch": 1.6669278996865202, "grad_norm": 2.613182544708252, "learning_rate": 6.310469767666209e-06, "loss": 1.6148, "step": 21270 }, { "epoch": 1.6670062695924766, "grad_norm": 2.085179567337036, "learning_rate": 6.3101721827249326e-06, "loss": 1.4407, "step": 21271 }, { "epoch": 1.6670846394984324, "grad_norm": 2.2819371223449707, "learning_rate": 6.309874592800436e-06, "loss": 1.2848, "step": 21272 }, { "epoch": 1.6671630094043888, "grad_norm": 3.1484925746917725, "learning_rate": 6.309576997893847e-06, "loss": 1.5117, "step": 21273 }, { "epoch": 1.6672413793103447, "grad_norm": 3.0058038234710693, "learning_rate": 6.309279398006302e-06, "loss": 1.2759, "step": 21274 }, { "epoch": 1.667319749216301, "grad_norm": 4.3619489669799805, "learning_rate": 6.308981793138928e-06, "loss": 2.0457, "step": 21275 }, { "epoch": 1.667398119122257, "grad_norm": 9.716022491455078, "learning_rate": 6.308684183292861e-06, "loss": 1.3148, "step": 21276 }, { "epoch": 1.6674764890282132, "grad_norm": 2.3047101497650146, "learning_rate": 6.308386568469232e-06, "loss": 1.3809, "step": 21277 }, { "epoch": 1.6675548589341693, "grad_norm": 3.7866177558898926, "learning_rate": 6.308088948669171e-06, "loss": 1.3235, "step": 21278 }, { "epoch": 1.6676332288401254, "grad_norm": 2.3750312328338623, "learning_rate": 6.307791323893813e-06, "loss": 1.188, "step": 21279 }, { "epoch": 1.6677115987460815, "grad_norm": 2.056457996368408, "learning_rate": 6.3074936941442865e-06, "loss": 1.1816, "step": 21280 }, { "epoch": 1.6677899686520377, "grad_norm": 2.9340455532073975, "learning_rate": 6.307196059421728e-06, "loss": 1.4392, "step": 21281 }, { "epoch": 1.6678683385579938, "grad_norm": 7.6968231201171875, "learning_rate": 6.306898419727265e-06, "loss": 1.5567, "step": 21282 }, { "epoch": 1.6679467084639499, "grad_norm": 5.498025417327881, "learning_rate": 6.306600775062034e-06, "loss": 1.7711, "step": 21283 }, { "epoch": 1.668025078369906, "grad_norm": 3.005600929260254, "learning_rate": 6.306303125427163e-06, "loss": 0.7698, "step": 21284 }, { "epoch": 1.668103448275862, "grad_norm": 3.906874656677246, "learning_rate": 6.306005470823786e-06, "loss": 1.3318, "step": 21285 }, { "epoch": 1.6681818181818182, "grad_norm": 3.5609467029571533, "learning_rate": 6.305707811253036e-06, "loss": 1.4511, "step": 21286 }, { "epoch": 1.6682601880877743, "grad_norm": 3.0727035999298096, "learning_rate": 6.3054101467160435e-06, "loss": 1.1059, "step": 21287 }, { "epoch": 1.6683385579937304, "grad_norm": 2.7207870483398438, "learning_rate": 6.305112477213941e-06, "loss": 1.4812, "step": 21288 }, { "epoch": 1.6684169278996865, "grad_norm": 7.325997352600098, "learning_rate": 6.304814802747861e-06, "loss": 1.9218, "step": 21289 }, { "epoch": 1.6684952978056427, "grad_norm": 2.7784156799316406, "learning_rate": 6.3045171233189375e-06, "loss": 1.1369, "step": 21290 }, { "epoch": 1.6685736677115988, "grad_norm": 3.2595860958099365, "learning_rate": 6.304219438928298e-06, "loss": 1.0783, "step": 21291 }, { "epoch": 1.6686520376175549, "grad_norm": 3.366915464401245, "learning_rate": 6.30392174957708e-06, "loss": 1.213, "step": 21292 }, { "epoch": 1.668730407523511, "grad_norm": 2.3461146354675293, "learning_rate": 6.303624055266412e-06, "loss": 1.4031, "step": 21293 }, { "epoch": 1.668808777429467, "grad_norm": 3.7470126152038574, "learning_rate": 6.303326355997428e-06, "loss": 1.1395, "step": 21294 }, { "epoch": 1.6688871473354232, "grad_norm": 2.4177613258361816, "learning_rate": 6.30302865177126e-06, "loss": 1.5658, "step": 21295 }, { "epoch": 1.6689655172413793, "grad_norm": 6.068456649780273, "learning_rate": 6.302730942589041e-06, "loss": 1.2094, "step": 21296 }, { "epoch": 1.6690438871473354, "grad_norm": 3.2843987941741943, "learning_rate": 6.3024332284519e-06, "loss": 1.4027, "step": 21297 }, { "epoch": 1.6691222570532915, "grad_norm": 4.961142539978027, "learning_rate": 6.302135509360975e-06, "loss": 1.3704, "step": 21298 }, { "epoch": 1.6692006269592476, "grad_norm": 3.3914246559143066, "learning_rate": 6.301837785317395e-06, "loss": 0.9862, "step": 21299 }, { "epoch": 1.6692789968652038, "grad_norm": 3.6189465522766113, "learning_rate": 6.301540056322293e-06, "loss": 1.1317, "step": 21300 }, { "epoch": 1.6693573667711599, "grad_norm": 3.074791431427002, "learning_rate": 6.301242322376798e-06, "loss": 1.1904, "step": 21301 }, { "epoch": 1.669435736677116, "grad_norm": 3.590898275375366, "learning_rate": 6.3009445834820496e-06, "loss": 1.3337, "step": 21302 }, { "epoch": 1.669514106583072, "grad_norm": 3.9051787853240967, "learning_rate": 6.300646839639174e-06, "loss": 1.309, "step": 21303 }, { "epoch": 1.6695924764890282, "grad_norm": 3.3124170303344727, "learning_rate": 6.300349090849306e-06, "loss": 1.2019, "step": 21304 }, { "epoch": 1.6696708463949843, "grad_norm": 1.99443519115448, "learning_rate": 6.300051337113578e-06, "loss": 1.024, "step": 21305 }, { "epoch": 1.6697492163009404, "grad_norm": 3.1661643981933594, "learning_rate": 6.299753578433123e-06, "loss": 1.805, "step": 21306 }, { "epoch": 1.6698275862068965, "grad_norm": 2.724029541015625, "learning_rate": 6.299455814809074e-06, "loss": 1.4138, "step": 21307 }, { "epoch": 1.6699059561128526, "grad_norm": 2.686342477798462, "learning_rate": 6.299158046242559e-06, "loss": 1.1921, "step": 21308 }, { "epoch": 1.6699843260188088, "grad_norm": 2.847172498703003, "learning_rate": 6.298860272734717e-06, "loss": 1.1876, "step": 21309 }, { "epoch": 1.6700626959247649, "grad_norm": 5.689830780029297, "learning_rate": 6.2985624942866764e-06, "loss": 1.9251, "step": 21310 }, { "epoch": 1.670141065830721, "grad_norm": 2.850341796875, "learning_rate": 6.298264710899572e-06, "loss": 1.31, "step": 21311 }, { "epoch": 1.670219435736677, "grad_norm": 3.8682100772857666, "learning_rate": 6.2979669225745346e-06, "loss": 1.3733, "step": 21312 }, { "epoch": 1.6702978056426332, "grad_norm": 2.341205596923828, "learning_rate": 6.297669129312698e-06, "loss": 1.0117, "step": 21313 }, { "epoch": 1.6703761755485893, "grad_norm": 2.0591773986816406, "learning_rate": 6.297371331115195e-06, "loss": 1.1841, "step": 21314 }, { "epoch": 1.6704545454545454, "grad_norm": 7.401251316070557, "learning_rate": 6.2970735279831575e-06, "loss": 1.2612, "step": 21315 }, { "epoch": 1.6705329153605015, "grad_norm": 2.5617568492889404, "learning_rate": 6.29677571991772e-06, "loss": 1.595, "step": 21316 }, { "epoch": 1.6706112852664576, "grad_norm": 3.3853201866149902, "learning_rate": 6.296477906920011e-06, "loss": 1.2543, "step": 21317 }, { "epoch": 1.6706896551724137, "grad_norm": 7.236016273498535, "learning_rate": 6.296180088991167e-06, "loss": 1.3987, "step": 21318 }, { "epoch": 1.6707680250783699, "grad_norm": 3.391824960708618, "learning_rate": 6.29588226613232e-06, "loss": 1.6113, "step": 21319 }, { "epoch": 1.670846394984326, "grad_norm": 2.930229663848877, "learning_rate": 6.295584438344603e-06, "loss": 1.3074, "step": 21320 }, { "epoch": 1.670924764890282, "grad_norm": 2.4117228984832764, "learning_rate": 6.295286605629147e-06, "loss": 1.2283, "step": 21321 }, { "epoch": 1.6710031347962384, "grad_norm": 2.4849274158477783, "learning_rate": 6.294988767987087e-06, "loss": 1.1348, "step": 21322 }, { "epoch": 1.6710815047021943, "grad_norm": 2.4564692974090576, "learning_rate": 6.294690925419554e-06, "loss": 1.449, "step": 21323 }, { "epoch": 1.6711598746081506, "grad_norm": 2.73246431350708, "learning_rate": 6.294393077927683e-06, "loss": 1.5924, "step": 21324 }, { "epoch": 1.6712382445141065, "grad_norm": 3.2371273040771484, "learning_rate": 6.294095225512604e-06, "loss": 0.8779, "step": 21325 }, { "epoch": 1.6713166144200629, "grad_norm": 2.613861322402954, "learning_rate": 6.293797368175454e-06, "loss": 1.322, "step": 21326 }, { "epoch": 1.6713949843260187, "grad_norm": 3.566063642501831, "learning_rate": 6.29349950591736e-06, "loss": 1.6488, "step": 21327 }, { "epoch": 1.671473354231975, "grad_norm": 11.828252792358398, "learning_rate": 6.293201638739461e-06, "loss": 1.3688, "step": 21328 }, { "epoch": 1.671551724137931, "grad_norm": 40.3821907043457, "learning_rate": 6.292903766642885e-06, "loss": 1.205, "step": 21329 }, { "epoch": 1.6716300940438873, "grad_norm": 7.329074382781982, "learning_rate": 6.292605889628769e-06, "loss": 1.9156, "step": 21330 }, { "epoch": 1.6717084639498432, "grad_norm": 3.0751919746398926, "learning_rate": 6.292308007698244e-06, "loss": 1.4861, "step": 21331 }, { "epoch": 1.6717868338557995, "grad_norm": 2.3425827026367188, "learning_rate": 6.292010120852442e-06, "loss": 1.329, "step": 21332 }, { "epoch": 1.6718652037617554, "grad_norm": 3.8638715744018555, "learning_rate": 6.291712229092498e-06, "loss": 1.542, "step": 21333 }, { "epoch": 1.6719435736677117, "grad_norm": 2.627776622772217, "learning_rate": 6.291414332419543e-06, "loss": 1.3981, "step": 21334 }, { "epoch": 1.6720219435736676, "grad_norm": 2.4537367820739746, "learning_rate": 6.291116430834713e-06, "loss": 1.5079, "step": 21335 }, { "epoch": 1.672100313479624, "grad_norm": 2.2311489582061768, "learning_rate": 6.2908185243391374e-06, "loss": 1.291, "step": 21336 }, { "epoch": 1.6721786833855798, "grad_norm": 3.744889974594116, "learning_rate": 6.290520612933952e-06, "loss": 1.7872, "step": 21337 }, { "epoch": 1.6722570532915362, "grad_norm": 2.6757636070251465, "learning_rate": 6.2902226966202895e-06, "loss": 1.3705, "step": 21338 }, { "epoch": 1.672335423197492, "grad_norm": 6.186439037322998, "learning_rate": 6.289924775399281e-06, "loss": 1.3333, "step": 21339 }, { "epoch": 1.6724137931034484, "grad_norm": 2.9875123500823975, "learning_rate": 6.289626849272062e-06, "loss": 1.6067, "step": 21340 }, { "epoch": 1.6724921630094043, "grad_norm": 2.9316489696502686, "learning_rate": 6.2893289182397655e-06, "loss": 1.0119, "step": 21341 }, { "epoch": 1.6725705329153606, "grad_norm": 2.825627326965332, "learning_rate": 6.2890309823035235e-06, "loss": 1.3287, "step": 21342 }, { "epoch": 1.6726489028213165, "grad_norm": 2.5771477222442627, "learning_rate": 6.28873304146447e-06, "loss": 1.2373, "step": 21343 }, { "epoch": 1.6727272727272728, "grad_norm": 4.853254795074463, "learning_rate": 6.2884350957237386e-06, "loss": 1.2502, "step": 21344 }, { "epoch": 1.6728056426332287, "grad_norm": 2.422214984893799, "learning_rate": 6.28813714508246e-06, "loss": 1.4574, "step": 21345 }, { "epoch": 1.672884012539185, "grad_norm": 2.9405274391174316, "learning_rate": 6.2878391895417715e-06, "loss": 1.6196, "step": 21346 }, { "epoch": 1.672962382445141, "grad_norm": 3.484498977661133, "learning_rate": 6.287541229102803e-06, "loss": 1.5141, "step": 21347 }, { "epoch": 1.6730407523510973, "grad_norm": 3.923663854598999, "learning_rate": 6.287243263766689e-06, "loss": 1.8371, "step": 21348 }, { "epoch": 1.6731191222570532, "grad_norm": 2.319100856781006, "learning_rate": 6.286945293534563e-06, "loss": 1.114, "step": 21349 }, { "epoch": 1.6731974921630095, "grad_norm": 2.835237979888916, "learning_rate": 6.2866473184075575e-06, "loss": 1.4815, "step": 21350 }, { "epoch": 1.6732758620689654, "grad_norm": 4.009203910827637, "learning_rate": 6.286349338386807e-06, "loss": 1.5788, "step": 21351 }, { "epoch": 1.6733542319749217, "grad_norm": 2.3721466064453125, "learning_rate": 6.286051353473445e-06, "loss": 1.4207, "step": 21352 }, { "epoch": 1.6734326018808776, "grad_norm": 2.4306724071502686, "learning_rate": 6.285753363668603e-06, "loss": 1.2484, "step": 21353 }, { "epoch": 1.673510971786834, "grad_norm": 3.6779189109802246, "learning_rate": 6.285455368973416e-06, "loss": 1.3905, "step": 21354 }, { "epoch": 1.6735893416927898, "grad_norm": 3.2170722484588623, "learning_rate": 6.285157369389017e-06, "loss": 1.54, "step": 21355 }, { "epoch": 1.6736677115987462, "grad_norm": 3.806004285812378, "learning_rate": 6.28485936491654e-06, "loss": 1.1939, "step": 21356 }, { "epoch": 1.673746081504702, "grad_norm": 3.1413040161132812, "learning_rate": 6.284561355557117e-06, "loss": 1.3467, "step": 21357 }, { "epoch": 1.6738244514106584, "grad_norm": 2.674138307571411, "learning_rate": 6.2842633413118835e-06, "loss": 1.6631, "step": 21358 }, { "epoch": 1.6739028213166143, "grad_norm": 2.705606460571289, "learning_rate": 6.283965322181971e-06, "loss": 1.5863, "step": 21359 }, { "epoch": 1.6739811912225706, "grad_norm": 2.6990365982055664, "learning_rate": 6.283667298168514e-06, "loss": 1.2284, "step": 21360 }, { "epoch": 1.6740595611285265, "grad_norm": 6.218212604522705, "learning_rate": 6.283369269272648e-06, "loss": 1.3349, "step": 21361 }, { "epoch": 1.6741379310344828, "grad_norm": 4.103358745574951, "learning_rate": 6.2830712354955015e-06, "loss": 1.5415, "step": 21362 }, { "epoch": 1.6742163009404387, "grad_norm": 2.8838772773742676, "learning_rate": 6.282773196838212e-06, "loss": 1.4364, "step": 21363 }, { "epoch": 1.674294670846395, "grad_norm": 2.057807207107544, "learning_rate": 6.2824751533019126e-06, "loss": 1.6014, "step": 21364 }, { "epoch": 1.674373040752351, "grad_norm": 2.497910976409912, "learning_rate": 6.282177104887737e-06, "loss": 1.368, "step": 21365 }, { "epoch": 1.6744514106583073, "grad_norm": 1.919748067855835, "learning_rate": 6.281879051596816e-06, "loss": 1.5467, "step": 21366 }, { "epoch": 1.6745297805642632, "grad_norm": 2.367280960083008, "learning_rate": 6.281580993430288e-06, "loss": 0.9374, "step": 21367 }, { "epoch": 1.6746081504702195, "grad_norm": 2.2940940856933594, "learning_rate": 6.281282930389283e-06, "loss": 1.3814, "step": 21368 }, { "epoch": 1.6746865203761756, "grad_norm": 2.5082621574401855, "learning_rate": 6.280984862474935e-06, "loss": 1.5617, "step": 21369 }, { "epoch": 1.6747648902821317, "grad_norm": 1.8745125532150269, "learning_rate": 6.2806867896883795e-06, "loss": 1.3096, "step": 21370 }, { "epoch": 1.6748432601880878, "grad_norm": 2.6188628673553467, "learning_rate": 6.28038871203075e-06, "loss": 1.3091, "step": 21371 }, { "epoch": 1.674921630094044, "grad_norm": 3.141638994216919, "learning_rate": 6.280090629503179e-06, "loss": 1.6152, "step": 21372 }, { "epoch": 1.675, "grad_norm": 3.0891995429992676, "learning_rate": 6.2797925421068005e-06, "loss": 1.6278, "step": 21373 }, { "epoch": 1.6750783699059562, "grad_norm": 2.5512750148773193, "learning_rate": 6.279494449842748e-06, "loss": 1.2788, "step": 21374 }, { "epoch": 1.6751567398119123, "grad_norm": 3.299220085144043, "learning_rate": 6.279196352712156e-06, "loss": 1.5325, "step": 21375 }, { "epoch": 1.6752351097178684, "grad_norm": 7.412543296813965, "learning_rate": 6.278898250716159e-06, "loss": 1.1689, "step": 21376 }, { "epoch": 1.6753134796238245, "grad_norm": 3.376033306121826, "learning_rate": 6.278600143855888e-06, "loss": 1.3393, "step": 21377 }, { "epoch": 1.6753918495297806, "grad_norm": 2.4312758445739746, "learning_rate": 6.2783020321324805e-06, "loss": 1.4048, "step": 21378 }, { "epoch": 1.6754702194357367, "grad_norm": 2.791646718978882, "learning_rate": 6.278003915547069e-06, "loss": 1.3884, "step": 21379 }, { "epoch": 1.6755485893416928, "grad_norm": 4.903043746948242, "learning_rate": 6.2777057941007864e-06, "loss": 1.3121, "step": 21380 }, { "epoch": 1.675626959247649, "grad_norm": 3.651729106903076, "learning_rate": 6.277407667794766e-06, "loss": 1.2284, "step": 21381 }, { "epoch": 1.675705329153605, "grad_norm": 1.9291177988052368, "learning_rate": 6.277109536630146e-06, "loss": 1.2281, "step": 21382 }, { "epoch": 1.6757836990595611, "grad_norm": 2.587827205657959, "learning_rate": 6.276811400608054e-06, "loss": 1.2213, "step": 21383 }, { "epoch": 1.6758620689655173, "grad_norm": 4.62948751449585, "learning_rate": 6.276513259729628e-06, "loss": 0.9958, "step": 21384 }, { "epoch": 1.6759404388714734, "grad_norm": 3.4095938205718994, "learning_rate": 6.2762151139960026e-06, "loss": 1.1672, "step": 21385 }, { "epoch": 1.6760188087774295, "grad_norm": 2.649141788482666, "learning_rate": 6.27591696340831e-06, "loss": 1.4432, "step": 21386 }, { "epoch": 1.6760971786833856, "grad_norm": 18.88389778137207, "learning_rate": 6.275618807967684e-06, "loss": 2.3325, "step": 21387 }, { "epoch": 1.6761755485893417, "grad_norm": 2.422227144241333, "learning_rate": 6.27532064767526e-06, "loss": 1.7411, "step": 21388 }, { "epoch": 1.6762539184952978, "grad_norm": 2.9257805347442627, "learning_rate": 6.275022482532171e-06, "loss": 1.3505, "step": 21389 }, { "epoch": 1.676332288401254, "grad_norm": 2.8138678073883057, "learning_rate": 6.274724312539549e-06, "loss": 1.1652, "step": 21390 }, { "epoch": 1.67641065830721, "grad_norm": 5.5972490310668945, "learning_rate": 6.274426137698535e-06, "loss": 1.1628, "step": 21391 }, { "epoch": 1.6764890282131661, "grad_norm": 2.4341509342193604, "learning_rate": 6.274127958010256e-06, "loss": 1.355, "step": 21392 }, { "epoch": 1.6765673981191223, "grad_norm": 3.205087423324585, "learning_rate": 6.273829773475848e-06, "loss": 1.296, "step": 21393 }, { "epoch": 1.6766457680250784, "grad_norm": 2.6555521488189697, "learning_rate": 6.273531584096447e-06, "loss": 1.4481, "step": 21394 }, { "epoch": 1.6767241379310345, "grad_norm": 2.1816515922546387, "learning_rate": 6.273233389873185e-06, "loss": 1.187, "step": 21395 }, { "epoch": 1.6768025078369906, "grad_norm": 3.079099655151367, "learning_rate": 6.272935190807199e-06, "loss": 1.7108, "step": 21396 }, { "epoch": 1.6768808777429467, "grad_norm": 4.006982803344727, "learning_rate": 6.272636986899619e-06, "loss": 1.6019, "step": 21397 }, { "epoch": 1.6769592476489028, "grad_norm": 4.5217413902282715, "learning_rate": 6.272338778151583e-06, "loss": 1.3247, "step": 21398 }, { "epoch": 1.677037617554859, "grad_norm": 1.91411292552948, "learning_rate": 6.2720405645642235e-06, "loss": 1.3018, "step": 21399 }, { "epoch": 1.677115987460815, "grad_norm": 3.905763626098633, "learning_rate": 6.271742346138676e-06, "loss": 1.7356, "step": 21400 }, { "epoch": 1.6771943573667711, "grad_norm": 8.804561614990234, "learning_rate": 6.271444122876072e-06, "loss": 0.821, "step": 21401 }, { "epoch": 1.6772727272727272, "grad_norm": 3.6358323097229004, "learning_rate": 6.27114589477755e-06, "loss": 1.3125, "step": 21402 }, { "epoch": 1.6773510971786834, "grad_norm": 2.155418872833252, "learning_rate": 6.27084766184424e-06, "loss": 1.345, "step": 21403 }, { "epoch": 1.6774294670846395, "grad_norm": 3.5215625762939453, "learning_rate": 6.270549424077278e-06, "loss": 1.6821, "step": 21404 }, { "epoch": 1.6775078369905956, "grad_norm": 2.639072895050049, "learning_rate": 6.2702511814778e-06, "loss": 1.3396, "step": 21405 }, { "epoch": 1.6775862068965517, "grad_norm": 5.031935214996338, "learning_rate": 6.269952934046938e-06, "loss": 1.6696, "step": 21406 }, { "epoch": 1.6776645768025078, "grad_norm": 2.3789567947387695, "learning_rate": 6.269654681785827e-06, "loss": 1.1149, "step": 21407 }, { "epoch": 1.677742946708464, "grad_norm": 2.7333579063415527, "learning_rate": 6.269356424695603e-06, "loss": 1.2389, "step": 21408 }, { "epoch": 1.67782131661442, "grad_norm": 2.820560932159424, "learning_rate": 6.269058162777397e-06, "loss": 1.5093, "step": 21409 }, { "epoch": 1.6778996865203761, "grad_norm": 3.681971311569214, "learning_rate": 6.268759896032348e-06, "loss": 1.4992, "step": 21410 }, { "epoch": 1.6779780564263322, "grad_norm": 3.3996543884277344, "learning_rate": 6.268461624461586e-06, "loss": 1.3944, "step": 21411 }, { "epoch": 1.6780564263322884, "grad_norm": 2.7224442958831787, "learning_rate": 6.26816334806625e-06, "loss": 1.5184, "step": 21412 }, { "epoch": 1.6781347962382445, "grad_norm": 2.31077241897583, "learning_rate": 6.267865066847471e-06, "loss": 1.356, "step": 21413 }, { "epoch": 1.6782131661442006, "grad_norm": 2.8807919025421143, "learning_rate": 6.267566780806383e-06, "loss": 1.3916, "step": 21414 }, { "epoch": 1.6782915360501567, "grad_norm": 2.3767037391662598, "learning_rate": 6.267268489944125e-06, "loss": 1.0889, "step": 21415 }, { "epoch": 1.6783699059561128, "grad_norm": 3.821063995361328, "learning_rate": 6.266970194261824e-06, "loss": 1.0517, "step": 21416 }, { "epoch": 1.6784482758620691, "grad_norm": 3.315952777862549, "learning_rate": 6.266671893760623e-06, "loss": 1.1751, "step": 21417 }, { "epoch": 1.678526645768025, "grad_norm": 3.2711870670318604, "learning_rate": 6.266373588441651e-06, "loss": 1.3125, "step": 21418 }, { "epoch": 1.6786050156739813, "grad_norm": 10.744762420654297, "learning_rate": 6.266075278306045e-06, "loss": 1.4973, "step": 21419 }, { "epoch": 1.6786833855799372, "grad_norm": 4.136627197265625, "learning_rate": 6.26577696335494e-06, "loss": 1.5024, "step": 21420 }, { "epoch": 1.6787617554858936, "grad_norm": 3.065891742706299, "learning_rate": 6.265478643589467e-06, "loss": 1.7669, "step": 21421 }, { "epoch": 1.6788401253918495, "grad_norm": 1.9769535064697266, "learning_rate": 6.265180319010765e-06, "loss": 1.2811, "step": 21422 }, { "epoch": 1.6789184952978058, "grad_norm": 1.944138765335083, "learning_rate": 6.264881989619966e-06, "loss": 0.8971, "step": 21423 }, { "epoch": 1.6789968652037617, "grad_norm": 2.7221744060516357, "learning_rate": 6.264583655418206e-06, "loss": 1.0905, "step": 21424 }, { "epoch": 1.679075235109718, "grad_norm": 3.204524040222168, "learning_rate": 6.26428531640662e-06, "loss": 1.3321, "step": 21425 }, { "epoch": 1.679153605015674, "grad_norm": 3.1666312217712402, "learning_rate": 6.263986972586341e-06, "loss": 1.6053, "step": 21426 }, { "epoch": 1.6792319749216302, "grad_norm": 2.4784703254699707, "learning_rate": 6.263688623958506e-06, "loss": 1.5147, "step": 21427 }, { "epoch": 1.6793103448275861, "grad_norm": 3.6947555541992188, "learning_rate": 6.263390270524248e-06, "loss": 1.5652, "step": 21428 }, { "epoch": 1.6793887147335425, "grad_norm": 9.821039199829102, "learning_rate": 6.263091912284702e-06, "loss": 1.7853, "step": 21429 }, { "epoch": 1.6794670846394983, "grad_norm": 2.1424612998962402, "learning_rate": 6.262793549241003e-06, "loss": 1.4163, "step": 21430 }, { "epoch": 1.6795454545454547, "grad_norm": 2.2886264324188232, "learning_rate": 6.262495181394286e-06, "loss": 1.1171, "step": 21431 }, { "epoch": 1.6796238244514106, "grad_norm": 2.955004930496216, "learning_rate": 6.2621968087456865e-06, "loss": 1.363, "step": 21432 }, { "epoch": 1.679702194357367, "grad_norm": 4.600681304931641, "learning_rate": 6.261898431296337e-06, "loss": 1.1871, "step": 21433 }, { "epoch": 1.6797805642633228, "grad_norm": 2.1482701301574707, "learning_rate": 6.261600049047376e-06, "loss": 1.4875, "step": 21434 }, { "epoch": 1.6798589341692791, "grad_norm": 2.1676154136657715, "learning_rate": 6.261301661999935e-06, "loss": 1.281, "step": 21435 }, { "epoch": 1.679937304075235, "grad_norm": 12.553572654724121, "learning_rate": 6.261003270155152e-06, "loss": 0.764, "step": 21436 }, { "epoch": 1.6800156739811913, "grad_norm": 3.4923129081726074, "learning_rate": 6.260704873514158e-06, "loss": 1.1126, "step": 21437 }, { "epoch": 1.6800940438871472, "grad_norm": 2.7994608879089355, "learning_rate": 6.260406472078091e-06, "loss": 1.2689, "step": 21438 }, { "epoch": 1.6801724137931036, "grad_norm": 2.4289801120758057, "learning_rate": 6.260108065848086e-06, "loss": 1.3788, "step": 21439 }, { "epoch": 1.6802507836990594, "grad_norm": 3.1962954998016357, "learning_rate": 6.259809654825276e-06, "loss": 1.4318, "step": 21440 }, { "epoch": 1.6803291536050158, "grad_norm": 3.3640661239624023, "learning_rate": 6.2595112390107986e-06, "loss": 1.6576, "step": 21441 }, { "epoch": 1.6804075235109717, "grad_norm": 2.4906251430511475, "learning_rate": 6.259212818405786e-06, "loss": 1.3727, "step": 21442 }, { "epoch": 1.680485893416928, "grad_norm": 3.024596691131592, "learning_rate": 6.258914393011376e-06, "loss": 1.0553, "step": 21443 }, { "epoch": 1.6805642633228839, "grad_norm": 2.333953619003296, "learning_rate": 6.2586159628287e-06, "loss": 1.5093, "step": 21444 }, { "epoch": 1.6806426332288402, "grad_norm": 2.6707139015197754, "learning_rate": 6.258317527858898e-06, "loss": 1.1201, "step": 21445 }, { "epoch": 1.680721003134796, "grad_norm": 2.6091437339782715, "learning_rate": 6.258019088103101e-06, "loss": 1.4158, "step": 21446 }, { "epoch": 1.6807993730407524, "grad_norm": 2.953768253326416, "learning_rate": 6.257720643562446e-06, "loss": 1.3007, "step": 21447 }, { "epoch": 1.6808777429467083, "grad_norm": 2.666269302368164, "learning_rate": 6.257422194238067e-06, "loss": 1.2801, "step": 21448 }, { "epoch": 1.6809561128526647, "grad_norm": 9.25232982635498, "learning_rate": 6.2571237401311e-06, "loss": 1.236, "step": 21449 }, { "epoch": 1.6810344827586206, "grad_norm": 7.225214004516602, "learning_rate": 6.25682528124268e-06, "loss": 1.4, "step": 21450 }, { "epoch": 1.6811128526645769, "grad_norm": 2.624174118041992, "learning_rate": 6.256526817573942e-06, "loss": 0.8306, "step": 21451 }, { "epoch": 1.6811912225705328, "grad_norm": 2.727658987045288, "learning_rate": 6.256228349126022e-06, "loss": 1.518, "step": 21452 }, { "epoch": 1.681269592476489, "grad_norm": 3.195758819580078, "learning_rate": 6.255929875900054e-06, "loss": 1.0417, "step": 21453 }, { "epoch": 1.681347962382445, "grad_norm": 3.469228506088257, "learning_rate": 6.255631397897175e-06, "loss": 1.4556, "step": 21454 }, { "epoch": 1.6814263322884013, "grad_norm": 2.854686737060547, "learning_rate": 6.2553329151185195e-06, "loss": 1.4082, "step": 21455 }, { "epoch": 1.6815047021943572, "grad_norm": 2.6724119186401367, "learning_rate": 6.255034427565221e-06, "loss": 1.1714, "step": 21456 }, { "epoch": 1.6815830721003135, "grad_norm": 2.7878787517547607, "learning_rate": 6.254735935238416e-06, "loss": 1.8397, "step": 21457 }, { "epoch": 1.6816614420062694, "grad_norm": 3.6395835876464844, "learning_rate": 6.254437438139241e-06, "loss": 1.3677, "step": 21458 }, { "epoch": 1.6817398119122258, "grad_norm": 6.032602310180664, "learning_rate": 6.2541389362688285e-06, "loss": 1.0693, "step": 21459 }, { "epoch": 1.6818181818181817, "grad_norm": 3.0787010192871094, "learning_rate": 6.253840429628317e-06, "loss": 1.2293, "step": 21460 }, { "epoch": 1.681896551724138, "grad_norm": 3.95192813873291, "learning_rate": 6.25354191821884e-06, "loss": 1.1756, "step": 21461 }, { "epoch": 1.6819749216300939, "grad_norm": 2.9059648513793945, "learning_rate": 6.2532434020415345e-06, "loss": 0.9493, "step": 21462 }, { "epoch": 1.6820532915360502, "grad_norm": 2.5337939262390137, "learning_rate": 6.252944881097534e-06, "loss": 1.5973, "step": 21463 }, { "epoch": 1.6821316614420063, "grad_norm": 13.719645500183105, "learning_rate": 6.252646355387975e-06, "loss": 1.3205, "step": 21464 }, { "epoch": 1.6822100313479624, "grad_norm": 10.645143508911133, "learning_rate": 6.252347824913991e-06, "loss": 1.5139, "step": 21465 }, { "epoch": 1.6822884012539185, "grad_norm": 4.647288799285889, "learning_rate": 6.252049289676722e-06, "loss": 0.7619, "step": 21466 }, { "epoch": 1.6823667711598747, "grad_norm": 6.533790111541748, "learning_rate": 6.2517507496772985e-06, "loss": 1.3518, "step": 21467 }, { "epoch": 1.6824451410658308, "grad_norm": 2.6391379833221436, "learning_rate": 6.251452204916859e-06, "loss": 0.761, "step": 21468 }, { "epoch": 1.6825235109717869, "grad_norm": 2.9037749767303467, "learning_rate": 6.2511536553965364e-06, "loss": 1.1299, "step": 21469 }, { "epoch": 1.682601880877743, "grad_norm": 2.963369131088257, "learning_rate": 6.25085510111747e-06, "loss": 0.9629, "step": 21470 }, { "epoch": 1.682680250783699, "grad_norm": 3.995436668395996, "learning_rate": 6.2505565420807934e-06, "loss": 1.4024, "step": 21471 }, { "epoch": 1.6827586206896552, "grad_norm": 2.2721855640411377, "learning_rate": 6.2502579782876405e-06, "loss": 1.3671, "step": 21472 }, { "epoch": 1.6828369905956113, "grad_norm": 3.2533276081085205, "learning_rate": 6.2499594097391505e-06, "loss": 1.517, "step": 21473 }, { "epoch": 1.6829153605015674, "grad_norm": 2.4325673580169678, "learning_rate": 6.249660836436455e-06, "loss": 1.4651, "step": 21474 }, { "epoch": 1.6829937304075235, "grad_norm": 3.398801803588867, "learning_rate": 6.249362258380692e-06, "loss": 1.591, "step": 21475 }, { "epoch": 1.6830721003134796, "grad_norm": 5.075812816619873, "learning_rate": 6.249063675572997e-06, "loss": 0.8206, "step": 21476 }, { "epoch": 1.6831504702194358, "grad_norm": 3.6748433113098145, "learning_rate": 6.248765088014504e-06, "loss": 1.4234, "step": 21477 }, { "epoch": 1.6832288401253919, "grad_norm": 2.3701565265655518, "learning_rate": 6.2484664957063515e-06, "loss": 1.7033, "step": 21478 }, { "epoch": 1.683307210031348, "grad_norm": 3.1683874130249023, "learning_rate": 6.248167898649673e-06, "loss": 1.3819, "step": 21479 }, { "epoch": 1.683385579937304, "grad_norm": 4.258184909820557, "learning_rate": 6.247869296845605e-06, "loss": 1.3194, "step": 21480 }, { "epoch": 1.6834639498432602, "grad_norm": 3.3855557441711426, "learning_rate": 6.2475706902952845e-06, "loss": 1.1262, "step": 21481 }, { "epoch": 1.6835423197492163, "grad_norm": 10.920014381408691, "learning_rate": 6.2472720789998445e-06, "loss": 1.3621, "step": 21482 }, { "epoch": 1.6836206896551724, "grad_norm": 3.8285586833953857, "learning_rate": 6.246973462960423e-06, "loss": 1.5476, "step": 21483 }, { "epoch": 1.6836990595611285, "grad_norm": 2.2840158939361572, "learning_rate": 6.246674842178155e-06, "loss": 1.4114, "step": 21484 }, { "epoch": 1.6837774294670846, "grad_norm": 2.1603829860687256, "learning_rate": 6.2463762166541756e-06, "loss": 1.2914, "step": 21485 }, { "epoch": 1.6838557993730408, "grad_norm": 4.037728786468506, "learning_rate": 6.246077586389621e-06, "loss": 1.387, "step": 21486 }, { "epoch": 1.6839341692789969, "grad_norm": 3.7790353298187256, "learning_rate": 6.245778951385627e-06, "loss": 1.1392, "step": 21487 }, { "epoch": 1.684012539184953, "grad_norm": 3.9073288440704346, "learning_rate": 6.245480311643332e-06, "loss": 1.3862, "step": 21488 }, { "epoch": 1.684090909090909, "grad_norm": 2.6009607315063477, "learning_rate": 6.245181667163867e-06, "loss": 1.3696, "step": 21489 }, { "epoch": 1.6841692789968652, "grad_norm": 2.5560929775238037, "learning_rate": 6.244883017948371e-06, "loss": 1.4193, "step": 21490 }, { "epoch": 1.6842476489028213, "grad_norm": 5.023924350738525, "learning_rate": 6.24458436399798e-06, "loss": 1.3169, "step": 21491 }, { "epoch": 1.6843260188087774, "grad_norm": 2.4345781803131104, "learning_rate": 6.24428570531383e-06, "loss": 1.1393, "step": 21492 }, { "epoch": 1.6844043887147335, "grad_norm": 2.721467971801758, "learning_rate": 6.243987041897055e-06, "loss": 1.0325, "step": 21493 }, { "epoch": 1.6844827586206896, "grad_norm": 6.259023189544678, "learning_rate": 6.243688373748793e-06, "loss": 1.2035, "step": 21494 }, { "epoch": 1.6845611285266457, "grad_norm": 2.6701831817626953, "learning_rate": 6.243389700870179e-06, "loss": 1.4, "step": 21495 }, { "epoch": 1.6846394984326019, "grad_norm": 4.689807891845703, "learning_rate": 6.243091023262349e-06, "loss": 1.5709, "step": 21496 }, { "epoch": 1.684717868338558, "grad_norm": 2.6093060970306396, "learning_rate": 6.242792340926439e-06, "loss": 1.2868, "step": 21497 }, { "epoch": 1.684796238244514, "grad_norm": 4.273171424865723, "learning_rate": 6.242493653863585e-06, "loss": 1.3899, "step": 21498 }, { "epoch": 1.6848746081504702, "grad_norm": 4.264240741729736, "learning_rate": 6.2421949620749254e-06, "loss": 1.5221, "step": 21499 }, { "epoch": 1.6849529780564263, "grad_norm": 3.0122082233428955, "learning_rate": 6.241896265561591e-06, "loss": 1.5406, "step": 21500 }, { "epoch": 1.6850313479623824, "grad_norm": 4.979034900665283, "learning_rate": 6.241597564324724e-06, "loss": 1.2701, "step": 21501 }, { "epoch": 1.6851097178683385, "grad_norm": 2.7306859493255615, "learning_rate": 6.241298858365456e-06, "loss": 1.1523, "step": 21502 }, { "epoch": 1.6851880877742946, "grad_norm": 2.2982561588287354, "learning_rate": 6.241000147684926e-06, "loss": 1.2911, "step": 21503 }, { "epoch": 1.6852664576802507, "grad_norm": 8.603107452392578, "learning_rate": 6.240701432284267e-06, "loss": 1.4213, "step": 21504 }, { "epoch": 1.6853448275862069, "grad_norm": 3.8477108478546143, "learning_rate": 6.240402712164618e-06, "loss": 1.1624, "step": 21505 }, { "epoch": 1.685423197492163, "grad_norm": 3.7356197834014893, "learning_rate": 6.240103987327116e-06, "loss": 1.3379, "step": 21506 }, { "epoch": 1.685501567398119, "grad_norm": 2.8900647163391113, "learning_rate": 6.239805257772892e-06, "loss": 1.2504, "step": 21507 }, { "epoch": 1.6855799373040752, "grad_norm": 6.1658525466918945, "learning_rate": 6.239506523503087e-06, "loss": 1.6621, "step": 21508 }, { "epoch": 1.6856583072100313, "grad_norm": 2.5535621643066406, "learning_rate": 6.239207784518837e-06, "loss": 1.2176, "step": 21509 }, { "epoch": 1.6857366771159876, "grad_norm": 3.775421380996704, "learning_rate": 6.238909040821275e-06, "loss": 1.3726, "step": 21510 }, { "epoch": 1.6858150470219435, "grad_norm": 12.905235290527344, "learning_rate": 6.238610292411542e-06, "loss": 1.3188, "step": 21511 }, { "epoch": 1.6858934169278998, "grad_norm": 3.051213502883911, "learning_rate": 6.23831153929077e-06, "loss": 1.2722, "step": 21512 }, { "epoch": 1.6859717868338557, "grad_norm": 4.1377177238464355, "learning_rate": 6.238012781460096e-06, "loss": 1.2718, "step": 21513 }, { "epoch": 1.686050156739812, "grad_norm": 2.242041826248169, "learning_rate": 6.23771401892066e-06, "loss": 1.3976, "step": 21514 }, { "epoch": 1.686128526645768, "grad_norm": 2.469794988632202, "learning_rate": 6.237415251673593e-06, "loss": 1.17, "step": 21515 }, { "epoch": 1.6862068965517243, "grad_norm": 2.4423298835754395, "learning_rate": 6.237116479720036e-06, "loss": 1.4383, "step": 21516 }, { "epoch": 1.6862852664576802, "grad_norm": 2.306248903274536, "learning_rate": 6.236817703061121e-06, "loss": 1.5828, "step": 21517 }, { "epoch": 1.6863636363636365, "grad_norm": 4.2611260414123535, "learning_rate": 6.23651892169799e-06, "loss": 1.3663, "step": 21518 }, { "epoch": 1.6864420062695924, "grad_norm": 2.4326705932617188, "learning_rate": 6.236220135631772e-06, "loss": 1.2498, "step": 21519 }, { "epoch": 1.6865203761755487, "grad_norm": 6.179162979125977, "learning_rate": 6.2359213448636104e-06, "loss": 1.2141, "step": 21520 }, { "epoch": 1.6865987460815046, "grad_norm": 2.131901741027832, "learning_rate": 6.235622549394637e-06, "loss": 1.0674, "step": 21521 }, { "epoch": 1.686677115987461, "grad_norm": 3.6896135807037354, "learning_rate": 6.2353237492259925e-06, "loss": 1.1809, "step": 21522 }, { "epoch": 1.6867554858934168, "grad_norm": 2.7357687950134277, "learning_rate": 6.235024944358809e-06, "loss": 1.1141, "step": 21523 }, { "epoch": 1.6868338557993732, "grad_norm": 3.708568572998047, "learning_rate": 6.2347261347942255e-06, "loss": 1.0524, "step": 21524 }, { "epoch": 1.686912225705329, "grad_norm": 2.8871984481811523, "learning_rate": 6.234427320533379e-06, "loss": 1.4212, "step": 21525 }, { "epoch": 1.6869905956112854, "grad_norm": 5.476912975311279, "learning_rate": 6.234128501577403e-06, "loss": 2.0008, "step": 21526 }, { "epoch": 1.6870689655172413, "grad_norm": 2.3431334495544434, "learning_rate": 6.233829677927439e-06, "loss": 1.1907, "step": 21527 }, { "epoch": 1.6871473354231976, "grad_norm": 2.7662668228149414, "learning_rate": 6.233530849584618e-06, "loss": 1.4471, "step": 21528 }, { "epoch": 1.6872257053291535, "grad_norm": 2.3826382160186768, "learning_rate": 6.23323201655008e-06, "loss": 1.4687, "step": 21529 }, { "epoch": 1.6873040752351098, "grad_norm": 5.461125373840332, "learning_rate": 6.23293317882496e-06, "loss": 1.4108, "step": 21530 }, { "epoch": 1.6873824451410657, "grad_norm": 2.158806085586548, "learning_rate": 6.232634336410397e-06, "loss": 1.1526, "step": 21531 }, { "epoch": 1.687460815047022, "grad_norm": 7.430636882781982, "learning_rate": 6.232335489307525e-06, "loss": 1.6398, "step": 21532 }, { "epoch": 1.687539184952978, "grad_norm": 3.002398729324341, "learning_rate": 6.2320366375174825e-06, "loss": 1.3636, "step": 21533 }, { "epoch": 1.6876175548589343, "grad_norm": 2.942572593688965, "learning_rate": 6.231737781041405e-06, "loss": 1.2525, "step": 21534 }, { "epoch": 1.6876959247648902, "grad_norm": 2.9545419216156006, "learning_rate": 6.23143891988043e-06, "loss": 1.1982, "step": 21535 }, { "epoch": 1.6877742946708465, "grad_norm": 3.8745455741882324, "learning_rate": 6.231140054035693e-06, "loss": 1.8291, "step": 21536 }, { "epoch": 1.6878526645768024, "grad_norm": 4.94205379486084, "learning_rate": 6.230841183508334e-06, "loss": 0.8803, "step": 21537 }, { "epoch": 1.6879310344827587, "grad_norm": 2.5130743980407715, "learning_rate": 6.230542308299484e-06, "loss": 1.4821, "step": 21538 }, { "epoch": 1.6880094043887146, "grad_norm": 4.300607204437256, "learning_rate": 6.2302434284102855e-06, "loss": 1.0983, "step": 21539 }, { "epoch": 1.688087774294671, "grad_norm": 2.7642362117767334, "learning_rate": 6.229944543841872e-06, "loss": 1.5423, "step": 21540 }, { "epoch": 1.6881661442006268, "grad_norm": 1.947503924369812, "learning_rate": 6.2296456545953806e-06, "loss": 1.2069, "step": 21541 }, { "epoch": 1.6882445141065832, "grad_norm": 2.5041942596435547, "learning_rate": 6.2293467606719496e-06, "loss": 1.2982, "step": 21542 }, { "epoch": 1.688322884012539, "grad_norm": 10.38148021697998, "learning_rate": 6.2290478620727146e-06, "loss": 1.2905, "step": 21543 }, { "epoch": 1.6884012539184954, "grad_norm": 3.1165757179260254, "learning_rate": 6.228748958798814e-06, "loss": 1.1781, "step": 21544 }, { "epoch": 1.6884796238244513, "grad_norm": 3.132523536682129, "learning_rate": 6.228450050851381e-06, "loss": 1.7849, "step": 21545 }, { "epoch": 1.6885579937304076, "grad_norm": 4.348504543304443, "learning_rate": 6.228151138231557e-06, "loss": 1.1473, "step": 21546 }, { "epoch": 1.6886363636363635, "grad_norm": 2.006765604019165, "learning_rate": 6.227852220940477e-06, "loss": 1.4241, "step": 21547 }, { "epoch": 1.6887147335423198, "grad_norm": 3.2559938430786133, "learning_rate": 6.227553298979276e-06, "loss": 1.3033, "step": 21548 }, { "epoch": 1.6887931034482757, "grad_norm": 2.096280336380005, "learning_rate": 6.227254372349094e-06, "loss": 0.9406, "step": 21549 }, { "epoch": 1.688871473354232, "grad_norm": 3.2737107276916504, "learning_rate": 6.226955441051067e-06, "loss": 1.3848, "step": 21550 }, { "epoch": 1.688949843260188, "grad_norm": 3.105346441268921, "learning_rate": 6.226656505086331e-06, "loss": 1.7295, "step": 21551 }, { "epoch": 1.6890282131661443, "grad_norm": 3.1789400577545166, "learning_rate": 6.226357564456025e-06, "loss": 1.2183, "step": 21552 }, { "epoch": 1.6891065830721002, "grad_norm": 4.255615234375, "learning_rate": 6.226058619161284e-06, "loss": 1.8991, "step": 21553 }, { "epoch": 1.6891849529780565, "grad_norm": 2.3613624572753906, "learning_rate": 6.225759669203244e-06, "loss": 1.1146, "step": 21554 }, { "epoch": 1.6892633228840124, "grad_norm": 3.9626047611236572, "learning_rate": 6.225460714583047e-06, "loss": 1.6334, "step": 21555 }, { "epoch": 1.6893416927899687, "grad_norm": 2.510953426361084, "learning_rate": 6.225161755301825e-06, "loss": 1.0338, "step": 21556 }, { "epoch": 1.6894200626959248, "grad_norm": 4.6713738441467285, "learning_rate": 6.224862791360716e-06, "loss": 1.2631, "step": 21557 }, { "epoch": 1.689498432601881, "grad_norm": 9.916573524475098, "learning_rate": 6.22456382276086e-06, "loss": 1.1467, "step": 21558 }, { "epoch": 1.689576802507837, "grad_norm": 2.587183952331543, "learning_rate": 6.2242648495033905e-06, "loss": 1.0347, "step": 21559 }, { "epoch": 1.6896551724137931, "grad_norm": 2.4380834102630615, "learning_rate": 6.223965871589448e-06, "loss": 1.2831, "step": 21560 }, { "epoch": 1.6897335423197493, "grad_norm": 3.0228271484375, "learning_rate": 6.223666889020167e-06, "loss": 1.4, "step": 21561 }, { "epoch": 1.6898119122257054, "grad_norm": 3.6638259887695312, "learning_rate": 6.223367901796685e-06, "loss": 1.3904, "step": 21562 }, { "epoch": 1.6898902821316615, "grad_norm": 2.748483180999756, "learning_rate": 6.2230689099201424e-06, "loss": 1.1367, "step": 21563 }, { "epoch": 1.6899686520376176, "grad_norm": 2.214125394821167, "learning_rate": 6.222769913391671e-06, "loss": 1.0639, "step": 21564 }, { "epoch": 1.6900470219435737, "grad_norm": 3.3912909030914307, "learning_rate": 6.222470912212413e-06, "loss": 1.2001, "step": 21565 }, { "epoch": 1.6901253918495298, "grad_norm": 5.08004093170166, "learning_rate": 6.222171906383502e-06, "loss": 1.9583, "step": 21566 }, { "epoch": 1.690203761755486, "grad_norm": 1.95085871219635, "learning_rate": 6.221872895906078e-06, "loss": 1.3009, "step": 21567 }, { "epoch": 1.690282131661442, "grad_norm": 2.307823657989502, "learning_rate": 6.221573880781276e-06, "loss": 1.4092, "step": 21568 }, { "epoch": 1.6903605015673981, "grad_norm": 2.2392947673797607, "learning_rate": 6.221274861010236e-06, "loss": 1.3776, "step": 21569 }, { "epoch": 1.6904388714733543, "grad_norm": 2.4643659591674805, "learning_rate": 6.220975836594093e-06, "loss": 1.5127, "step": 21570 }, { "epoch": 1.6905172413793104, "grad_norm": 2.7855443954467773, "learning_rate": 6.220676807533984e-06, "loss": 1.1269, "step": 21571 }, { "epoch": 1.6905956112852665, "grad_norm": 4.0339226722717285, "learning_rate": 6.220377773831049e-06, "loss": 1.3083, "step": 21572 }, { "epoch": 1.6906739811912226, "grad_norm": 2.936497688293457, "learning_rate": 6.220078735486421e-06, "loss": 1.1573, "step": 21573 }, { "epoch": 1.6907523510971787, "grad_norm": 4.011308670043945, "learning_rate": 6.219779692501244e-06, "loss": 1.1421, "step": 21574 }, { "epoch": 1.6908307210031348, "grad_norm": 3.7067103385925293, "learning_rate": 6.2194806448766495e-06, "loss": 1.6709, "step": 21575 }, { "epoch": 1.690909090909091, "grad_norm": 2.744379758834839, "learning_rate": 6.219181592613777e-06, "loss": 1.29, "step": 21576 }, { "epoch": 1.690987460815047, "grad_norm": 2.450906991958618, "learning_rate": 6.218882535713764e-06, "loss": 1.4729, "step": 21577 }, { "epoch": 1.6910658307210031, "grad_norm": 4.514129161834717, "learning_rate": 6.218583474177748e-06, "loss": 1.3445, "step": 21578 }, { "epoch": 1.6911442006269592, "grad_norm": 1.8488852977752686, "learning_rate": 6.218284408006866e-06, "loss": 1.3662, "step": 21579 }, { "epoch": 1.6912225705329154, "grad_norm": 2.3945162296295166, "learning_rate": 6.2179853372022555e-06, "loss": 1.0434, "step": 21580 }, { "epoch": 1.6913009404388715, "grad_norm": 3.918022394180298, "learning_rate": 6.217686261765057e-06, "loss": 1.3296, "step": 21581 }, { "epoch": 1.6913793103448276, "grad_norm": 4.393553733825684, "learning_rate": 6.217387181696404e-06, "loss": 1.8903, "step": 21582 }, { "epoch": 1.6914576802507837, "grad_norm": 3.5731635093688965, "learning_rate": 6.217088096997435e-06, "loss": 1.2724, "step": 21583 }, { "epoch": 1.6915360501567398, "grad_norm": 3.148513078689575, "learning_rate": 6.216789007669287e-06, "loss": 1.4733, "step": 21584 }, { "epoch": 1.691614420062696, "grad_norm": 4.5926008224487305, "learning_rate": 6.2164899137131e-06, "loss": 1.244, "step": 21585 }, { "epoch": 1.691692789968652, "grad_norm": 3.422738552093506, "learning_rate": 6.216190815130009e-06, "loss": 1.2822, "step": 21586 }, { "epoch": 1.6917711598746081, "grad_norm": 3.8009097576141357, "learning_rate": 6.215891711921154e-06, "loss": 1.3565, "step": 21587 }, { "epoch": 1.6918495297805642, "grad_norm": 4.460323810577393, "learning_rate": 6.2155926040876706e-06, "loss": 1.3927, "step": 21588 }, { "epoch": 1.6919278996865204, "grad_norm": 3.0024776458740234, "learning_rate": 6.215293491630698e-06, "loss": 1.142, "step": 21589 }, { "epoch": 1.6920062695924765, "grad_norm": 3.1758499145507812, "learning_rate": 6.214994374551372e-06, "loss": 1.3309, "step": 21590 }, { "epoch": 1.6920846394984326, "grad_norm": 2.9302821159362793, "learning_rate": 6.214695252850833e-06, "loss": 1.3305, "step": 21591 }, { "epoch": 1.6921630094043887, "grad_norm": 1.8520253896713257, "learning_rate": 6.214396126530215e-06, "loss": 1.4144, "step": 21592 }, { "epoch": 1.6922413793103448, "grad_norm": 4.963454246520996, "learning_rate": 6.214096995590661e-06, "loss": 1.3483, "step": 21593 }, { "epoch": 1.692319749216301, "grad_norm": 2.3207244873046875, "learning_rate": 6.213797860033304e-06, "loss": 1.2785, "step": 21594 }, { "epoch": 1.692398119122257, "grad_norm": 3.729750871658325, "learning_rate": 6.213498719859284e-06, "loss": 1.2418, "step": 21595 }, { "epoch": 1.6924764890282131, "grad_norm": 3.666327714920044, "learning_rate": 6.213199575069737e-06, "loss": 1.2814, "step": 21596 }, { "epoch": 1.6925548589341692, "grad_norm": 3.5677692890167236, "learning_rate": 6.212900425665802e-06, "loss": 1.4944, "step": 21597 }, { "epoch": 1.6926332288401253, "grad_norm": 5.361937046051025, "learning_rate": 6.2126012716486185e-06, "loss": 1.7923, "step": 21598 }, { "epoch": 1.6927115987460815, "grad_norm": 3.0387730598449707, "learning_rate": 6.212302113019321e-06, "loss": 0.9592, "step": 21599 }, { "epoch": 1.6927899686520376, "grad_norm": 2.0364480018615723, "learning_rate": 6.2120029497790505e-06, "loss": 1.5551, "step": 21600 }, { "epoch": 1.6928683385579937, "grad_norm": 3.2261927127838135, "learning_rate": 6.21170378192894e-06, "loss": 1.4124, "step": 21601 }, { "epoch": 1.6929467084639498, "grad_norm": 4.229267597198486, "learning_rate": 6.211404609470134e-06, "loss": 1.8926, "step": 21602 }, { "epoch": 1.6930250783699061, "grad_norm": 2.5888381004333496, "learning_rate": 6.211105432403765e-06, "loss": 1.2109, "step": 21603 }, { "epoch": 1.693103448275862, "grad_norm": 3.160808563232422, "learning_rate": 6.210806250730975e-06, "loss": 1.5622, "step": 21604 }, { "epoch": 1.6931818181818183, "grad_norm": 2.6797866821289062, "learning_rate": 6.210507064452899e-06, "loss": 1.2402, "step": 21605 }, { "epoch": 1.6932601880877742, "grad_norm": 6.382140636444092, "learning_rate": 6.210207873570675e-06, "loss": 1.4408, "step": 21606 }, { "epoch": 1.6933385579937306, "grad_norm": 3.14335298538208, "learning_rate": 6.209908678085442e-06, "loss": 1.4833, "step": 21607 }, { "epoch": 1.6934169278996865, "grad_norm": 2.5374691486358643, "learning_rate": 6.209609477998339e-06, "loss": 0.4947, "step": 21608 }, { "epoch": 1.6934952978056428, "grad_norm": 3.613018035888672, "learning_rate": 6.209310273310502e-06, "loss": 1.6276, "step": 21609 }, { "epoch": 1.6935736677115987, "grad_norm": 5.759538650512695, "learning_rate": 6.209011064023072e-06, "loss": 0.7186, "step": 21610 }, { "epoch": 1.693652037617555, "grad_norm": 16.84559440612793, "learning_rate": 6.208711850137182e-06, "loss": 2.0493, "step": 21611 }, { "epoch": 1.693730407523511, "grad_norm": 2.7117154598236084, "learning_rate": 6.208412631653974e-06, "loss": 1.2972, "step": 21612 }, { "epoch": 1.6938087774294672, "grad_norm": 4.149866580963135, "learning_rate": 6.208113408574585e-06, "loss": 1.0127, "step": 21613 }, { "epoch": 1.6938871473354231, "grad_norm": 3.5330865383148193, "learning_rate": 6.207814180900152e-06, "loss": 1.3507, "step": 21614 }, { "epoch": 1.6939655172413794, "grad_norm": 3.353769540786743, "learning_rate": 6.207514948631816e-06, "loss": 1.2523, "step": 21615 }, { "epoch": 1.6940438871473353, "grad_norm": 4.135389804840088, "learning_rate": 6.207215711770713e-06, "loss": 1.4529, "step": 21616 }, { "epoch": 1.6941222570532917, "grad_norm": 5.013324737548828, "learning_rate": 6.206916470317982e-06, "loss": 1.1717, "step": 21617 }, { "epoch": 1.6942006269592476, "grad_norm": 2.742811679840088, "learning_rate": 6.206617224274758e-06, "loss": 1.1187, "step": 21618 }, { "epoch": 1.694278996865204, "grad_norm": 3.8790624141693115, "learning_rate": 6.206317973642185e-06, "loss": 1.5147, "step": 21619 }, { "epoch": 1.6943573667711598, "grad_norm": 2.2003440856933594, "learning_rate": 6.206018718421396e-06, "loss": 1.4042, "step": 21620 }, { "epoch": 1.6944357366771161, "grad_norm": 2.2776217460632324, "learning_rate": 6.205719458613533e-06, "loss": 1.0792, "step": 21621 }, { "epoch": 1.694514106583072, "grad_norm": 2.778477430343628, "learning_rate": 6.2054201942197315e-06, "loss": 1.1928, "step": 21622 }, { "epoch": 1.6945924764890283, "grad_norm": 6.792442321777344, "learning_rate": 6.2051209252411305e-06, "loss": 1.0446, "step": 21623 }, { "epoch": 1.6946708463949842, "grad_norm": 2.7171099185943604, "learning_rate": 6.204821651678869e-06, "loss": 1.6231, "step": 21624 }, { "epoch": 1.6947492163009406, "grad_norm": 5.028123378753662, "learning_rate": 6.204522373534083e-06, "loss": 1.1264, "step": 21625 }, { "epoch": 1.6948275862068964, "grad_norm": 3.220527410507202, "learning_rate": 6.204223090807916e-06, "loss": 1.3091, "step": 21626 }, { "epoch": 1.6949059561128528, "grad_norm": 2.7199759483337402, "learning_rate": 6.2039238035014995e-06, "loss": 1.466, "step": 21627 }, { "epoch": 1.6949843260188087, "grad_norm": 3.401181936264038, "learning_rate": 6.2036245116159775e-06, "loss": 0.9379, "step": 21628 }, { "epoch": 1.695062695924765, "grad_norm": 2.2371318340301514, "learning_rate": 6.203325215152485e-06, "loss": 1.0456, "step": 21629 }, { "epoch": 1.6951410658307209, "grad_norm": 2.7952897548675537, "learning_rate": 6.203025914112162e-06, "loss": 1.2399, "step": 21630 }, { "epoch": 1.6952194357366772, "grad_norm": 2.3336145877838135, "learning_rate": 6.202726608496146e-06, "loss": 1.0386, "step": 21631 }, { "epoch": 1.695297805642633, "grad_norm": 2.7667245864868164, "learning_rate": 6.2024272983055754e-06, "loss": 1.0428, "step": 21632 }, { "epoch": 1.6953761755485894, "grad_norm": 2.9533276557922363, "learning_rate": 6.2021279835415885e-06, "loss": 1.0347, "step": 21633 }, { "epoch": 1.6954545454545453, "grad_norm": 2.118053913116455, "learning_rate": 6.201828664205325e-06, "loss": 1.2097, "step": 21634 }, { "epoch": 1.6955329153605017, "grad_norm": 2.5487656593322754, "learning_rate": 6.201529340297923e-06, "loss": 1.1829, "step": 21635 }, { "epoch": 1.6956112852664575, "grad_norm": 5.082744598388672, "learning_rate": 6.2012300118205205e-06, "loss": 0.7335, "step": 21636 }, { "epoch": 1.6956896551724139, "grad_norm": 3.4285194873809814, "learning_rate": 6.200930678774255e-06, "loss": 1.4975, "step": 21637 }, { "epoch": 1.6957680250783698, "grad_norm": 3.337137460708618, "learning_rate": 6.200631341160267e-06, "loss": 1.0405, "step": 21638 }, { "epoch": 1.695846394984326, "grad_norm": 2.783007860183716, "learning_rate": 6.2003319989796935e-06, "loss": 1.338, "step": 21639 }, { "epoch": 1.695924764890282, "grad_norm": 3.6211681365966797, "learning_rate": 6.200032652233674e-06, "loss": 1.0651, "step": 21640 }, { "epoch": 1.6960031347962383, "grad_norm": 3.0054898262023926, "learning_rate": 6.199733300923347e-06, "loss": 1.4956, "step": 21641 }, { "epoch": 1.6960815047021942, "grad_norm": 3.171402931213379, "learning_rate": 6.19943394504985e-06, "loss": 0.7425, "step": 21642 }, { "epoch": 1.6961598746081505, "grad_norm": 2.602766275405884, "learning_rate": 6.199134584614324e-06, "loss": 1.4634, "step": 21643 }, { "epoch": 1.6962382445141064, "grad_norm": 2.298846960067749, "learning_rate": 6.198835219617903e-06, "loss": 1.6883, "step": 21644 }, { "epoch": 1.6963166144200628, "grad_norm": 3.2908718585968018, "learning_rate": 6.198535850061731e-06, "loss": 1.0513, "step": 21645 }, { "epoch": 1.6963949843260187, "grad_norm": 5.93988037109375, "learning_rate": 6.198236475946942e-06, "loss": 1.9292, "step": 21646 }, { "epoch": 1.696473354231975, "grad_norm": 8.892009735107422, "learning_rate": 6.197937097274679e-06, "loss": 1.7277, "step": 21647 }, { "epoch": 1.6965517241379309, "grad_norm": 3.572920083999634, "learning_rate": 6.197637714046078e-06, "loss": 1.2242, "step": 21648 }, { "epoch": 1.6966300940438872, "grad_norm": 6.93674373626709, "learning_rate": 6.197338326262278e-06, "loss": 1.7066, "step": 21649 }, { "epoch": 1.6967084639498433, "grad_norm": 3.828350305557251, "learning_rate": 6.197038933924418e-06, "loss": 1.5675, "step": 21650 }, { "epoch": 1.6967868338557994, "grad_norm": 2.9525156021118164, "learning_rate": 6.196739537033636e-06, "loss": 1.473, "step": 21651 }, { "epoch": 1.6968652037617555, "grad_norm": 7.251319408416748, "learning_rate": 6.196440135591073e-06, "loss": 1.2164, "step": 21652 }, { "epoch": 1.6969435736677116, "grad_norm": 6.673832416534424, "learning_rate": 6.196140729597865e-06, "loss": 1.6712, "step": 21653 }, { "epoch": 1.6970219435736678, "grad_norm": 3.7084977626800537, "learning_rate": 6.195841319055153e-06, "loss": 1.3294, "step": 21654 }, { "epoch": 1.6971003134796239, "grad_norm": 3.7483510971069336, "learning_rate": 6.195541903964074e-06, "loss": 1.2197, "step": 21655 }, { "epoch": 1.69717868338558, "grad_norm": 2.441276788711548, "learning_rate": 6.195242484325769e-06, "loss": 1.4291, "step": 21656 }, { "epoch": 1.697257053291536, "grad_norm": 7.078649520874023, "learning_rate": 6.194943060141374e-06, "loss": 1.2378, "step": 21657 }, { "epoch": 1.6973354231974922, "grad_norm": 4.392245769500732, "learning_rate": 6.19464363141203e-06, "loss": 1.499, "step": 21658 }, { "epoch": 1.6974137931034483, "grad_norm": 8.9793062210083, "learning_rate": 6.194344198138875e-06, "loss": 1.5979, "step": 21659 }, { "epoch": 1.6974921630094044, "grad_norm": 3.9443295001983643, "learning_rate": 6.194044760323049e-06, "loss": 1.2341, "step": 21660 }, { "epoch": 1.6975705329153605, "grad_norm": 7.912234783172607, "learning_rate": 6.193745317965689e-06, "loss": 1.8925, "step": 21661 }, { "epoch": 1.6976489028213166, "grad_norm": 2.1077821254730225, "learning_rate": 6.193445871067935e-06, "loss": 1.3112, "step": 21662 }, { "epoch": 1.6977272727272728, "grad_norm": 2.9412055015563965, "learning_rate": 6.193146419630927e-06, "loss": 1.2485, "step": 21663 }, { "epoch": 1.6978056426332289, "grad_norm": 2.8261983394622803, "learning_rate": 6.192846963655803e-06, "loss": 1.4164, "step": 21664 }, { "epoch": 1.697884012539185, "grad_norm": 3.1657323837280273, "learning_rate": 6.1925475031437e-06, "loss": 1.3961, "step": 21665 }, { "epoch": 1.697962382445141, "grad_norm": 3.3122825622558594, "learning_rate": 6.19224803809576e-06, "loss": 1.5542, "step": 21666 }, { "epoch": 1.6980407523510972, "grad_norm": 3.8763768672943115, "learning_rate": 6.191948568513121e-06, "loss": 1.7602, "step": 21667 }, { "epoch": 1.6981191222570533, "grad_norm": 2.9271535873413086, "learning_rate": 6.191649094396921e-06, "loss": 1.3055, "step": 21668 }, { "epoch": 1.6981974921630094, "grad_norm": 2.5920143127441406, "learning_rate": 6.1913496157483e-06, "loss": 0.9965, "step": 21669 }, { "epoch": 1.6982758620689655, "grad_norm": 1.9885956048965454, "learning_rate": 6.191050132568397e-06, "loss": 1.2291, "step": 21670 }, { "epoch": 1.6983542319749216, "grad_norm": 3.9861464500427246, "learning_rate": 6.190750644858353e-06, "loss": 1.1736, "step": 21671 }, { "epoch": 1.6984326018808777, "grad_norm": 2.83624267578125, "learning_rate": 6.190451152619302e-06, "loss": 1.0489, "step": 21672 }, { "epoch": 1.6985109717868339, "grad_norm": 2.7174980640411377, "learning_rate": 6.190151655852389e-06, "loss": 1.2995, "step": 21673 }, { "epoch": 1.69858934169279, "grad_norm": 2.427565336227417, "learning_rate": 6.189852154558748e-06, "loss": 1.3166, "step": 21674 }, { "epoch": 1.698667711598746, "grad_norm": 3.3564209938049316, "learning_rate": 6.189552648739523e-06, "loss": 1.1956, "step": 21675 }, { "epoch": 1.6987460815047022, "grad_norm": 3.684307813644409, "learning_rate": 6.18925313839585e-06, "loss": 1.125, "step": 21676 }, { "epoch": 1.6988244514106583, "grad_norm": 2.4510231018066406, "learning_rate": 6.1889536235288674e-06, "loss": 1.4361, "step": 21677 }, { "epoch": 1.6989028213166144, "grad_norm": 3.733790636062622, "learning_rate": 6.188654104139717e-06, "loss": 1.6562, "step": 21678 }, { "epoch": 1.6989811912225705, "grad_norm": 4.740199565887451, "learning_rate": 6.1883545802295365e-06, "loss": 1.3627, "step": 21679 }, { "epoch": 1.6990595611285266, "grad_norm": 8.719825744628906, "learning_rate": 6.188055051799468e-06, "loss": 1.4954, "step": 21680 }, { "epoch": 1.6991379310344827, "grad_norm": 3.9269909858703613, "learning_rate": 6.187755518850645e-06, "loss": 1.699, "step": 21681 }, { "epoch": 1.6992163009404389, "grad_norm": 2.8713793754577637, "learning_rate": 6.1874559813842115e-06, "loss": 1.4042, "step": 21682 }, { "epoch": 1.699294670846395, "grad_norm": 2.6989455223083496, "learning_rate": 6.1871564394013055e-06, "loss": 1.4454, "step": 21683 }, { "epoch": 1.699373040752351, "grad_norm": 3.8946382999420166, "learning_rate": 6.186856892903065e-06, "loss": 1.3756, "step": 21684 }, { "epoch": 1.6994514106583072, "grad_norm": 3.3147547245025635, "learning_rate": 6.186557341890631e-06, "loss": 1.2914, "step": 21685 }, { "epoch": 1.6995297805642633, "grad_norm": 4.882882118225098, "learning_rate": 6.186257786365143e-06, "loss": 1.3264, "step": 21686 }, { "epoch": 1.6996081504702194, "grad_norm": 3.1634156703948975, "learning_rate": 6.185958226327739e-06, "loss": 1.7844, "step": 21687 }, { "epoch": 1.6996865203761755, "grad_norm": 4.184216499328613, "learning_rate": 6.185658661779559e-06, "loss": 1.2857, "step": 21688 }, { "epoch": 1.6997648902821316, "grad_norm": 3.1621899604797363, "learning_rate": 6.185359092721744e-06, "loss": 0.8846, "step": 21689 }, { "epoch": 1.6998432601880877, "grad_norm": 5.167660236358643, "learning_rate": 6.185059519155431e-06, "loss": 1.9012, "step": 21690 }, { "epoch": 1.6999216300940438, "grad_norm": 2.7190535068511963, "learning_rate": 6.184759941081758e-06, "loss": 1.1646, "step": 21691 }, { "epoch": 1.7, "grad_norm": 3.8851914405822754, "learning_rate": 6.184460358501871e-06, "loss": 1.3277, "step": 21692 }, { "epoch": 1.700078369905956, "grad_norm": 3.407895088195801, "learning_rate": 6.184160771416901e-06, "loss": 1.2208, "step": 21693 }, { "epoch": 1.7001567398119122, "grad_norm": 2.378528594970703, "learning_rate": 6.1838611798279934e-06, "loss": 1.137, "step": 21694 }, { "epoch": 1.7002351097178683, "grad_norm": 3.4416306018829346, "learning_rate": 6.183561583736286e-06, "loss": 1.3203, "step": 21695 }, { "epoch": 1.7003134796238244, "grad_norm": 6.371809482574463, "learning_rate": 6.183261983142919e-06, "loss": 1.2928, "step": 21696 }, { "epoch": 1.7003918495297805, "grad_norm": 2.6525917053222656, "learning_rate": 6.18296237804903e-06, "loss": 1.5524, "step": 21697 }, { "epoch": 1.7004702194357368, "grad_norm": 2.4331037998199463, "learning_rate": 6.18266276845576e-06, "loss": 1.3182, "step": 21698 }, { "epoch": 1.7005485893416927, "grad_norm": 3.103764772415161, "learning_rate": 6.182363154364249e-06, "loss": 1.5153, "step": 21699 }, { "epoch": 1.700626959247649, "grad_norm": 2.6490612030029297, "learning_rate": 6.182063535775634e-06, "loss": 1.4869, "step": 21700 }, { "epoch": 1.700705329153605, "grad_norm": 3.9068734645843506, "learning_rate": 6.181763912691059e-06, "loss": 1.5387, "step": 21701 }, { "epoch": 1.7007836990595613, "grad_norm": 6.2497639656066895, "learning_rate": 6.18146428511166e-06, "loss": 1.2188, "step": 21702 }, { "epoch": 1.7008620689655172, "grad_norm": 3.9358699321746826, "learning_rate": 6.181164653038577e-06, "loss": 0.995, "step": 21703 }, { "epoch": 1.7009404388714735, "grad_norm": 2.8048083782196045, "learning_rate": 6.180865016472951e-06, "loss": 0.897, "step": 21704 }, { "epoch": 1.7010188087774294, "grad_norm": 3.440415143966675, "learning_rate": 6.180565375415921e-06, "loss": 1.3072, "step": 21705 }, { "epoch": 1.7010971786833857, "grad_norm": 2.493767261505127, "learning_rate": 6.180265729868627e-06, "loss": 1.2438, "step": 21706 }, { "epoch": 1.7011755485893416, "grad_norm": 2.064147710800171, "learning_rate": 6.179966079832207e-06, "loss": 1.4178, "step": 21707 }, { "epoch": 1.701253918495298, "grad_norm": 4.7786784172058105, "learning_rate": 6.179666425307804e-06, "loss": 1.4047, "step": 21708 }, { "epoch": 1.7013322884012538, "grad_norm": 3.2687361240386963, "learning_rate": 6.179366766296554e-06, "loss": 1.3299, "step": 21709 }, { "epoch": 1.7014106583072102, "grad_norm": 4.001611232757568, "learning_rate": 6.179067102799601e-06, "loss": 1.551, "step": 21710 }, { "epoch": 1.701489028213166, "grad_norm": 2.873399019241333, "learning_rate": 6.178767434818081e-06, "loss": 1.0396, "step": 21711 }, { "epoch": 1.7015673981191224, "grad_norm": 4.397907257080078, "learning_rate": 6.178467762353134e-06, "loss": 1.2842, "step": 21712 }, { "epoch": 1.7016457680250783, "grad_norm": 4.625427722930908, "learning_rate": 6.178168085405901e-06, "loss": 1.2852, "step": 21713 }, { "epoch": 1.7017241379310346, "grad_norm": 5.90447998046875, "learning_rate": 6.177868403977523e-06, "loss": 1.0911, "step": 21714 }, { "epoch": 1.7018025078369905, "grad_norm": 13.266386032104492, "learning_rate": 6.1775687180691376e-06, "loss": 1.3904, "step": 21715 }, { "epoch": 1.7018808777429468, "grad_norm": 11.020991325378418, "learning_rate": 6.177269027681887e-06, "loss": 1.9019, "step": 21716 }, { "epoch": 1.7019592476489027, "grad_norm": 6.220468997955322, "learning_rate": 6.176969332816909e-06, "loss": 1.0396, "step": 21717 }, { "epoch": 1.702037617554859, "grad_norm": 4.002222061157227, "learning_rate": 6.1766696334753455e-06, "loss": 1.3689, "step": 21718 }, { "epoch": 1.702115987460815, "grad_norm": 4.510912895202637, "learning_rate": 6.176369929658331e-06, "loss": 1.0074, "step": 21719 }, { "epoch": 1.7021943573667713, "grad_norm": 4.645577907562256, "learning_rate": 6.176070221367015e-06, "loss": 1.132, "step": 21720 }, { "epoch": 1.7022727272727272, "grad_norm": 3.502850294113159, "learning_rate": 6.175770508602528e-06, "loss": 1.4374, "step": 21721 }, { "epoch": 1.7023510971786835, "grad_norm": 4.4486985206604, "learning_rate": 6.175470791366015e-06, "loss": 1.2794, "step": 21722 }, { "epoch": 1.7024294670846394, "grad_norm": 2.7741079330444336, "learning_rate": 6.1751710696586144e-06, "loss": 1.5304, "step": 21723 }, { "epoch": 1.7025078369905957, "grad_norm": 2.3716416358947754, "learning_rate": 6.174871343481468e-06, "loss": 1.1039, "step": 21724 }, { "epoch": 1.7025862068965516, "grad_norm": 12.514822959899902, "learning_rate": 6.174571612835713e-06, "loss": 2.0342, "step": 21725 }, { "epoch": 1.702664576802508, "grad_norm": 2.9160163402557373, "learning_rate": 6.174271877722491e-06, "loss": 1.479, "step": 21726 }, { "epoch": 1.7027429467084638, "grad_norm": 2.3310585021972656, "learning_rate": 6.1739721381429445e-06, "loss": 1.3422, "step": 21727 }, { "epoch": 1.7028213166144202, "grad_norm": 7.3835601806640625, "learning_rate": 6.173672394098207e-06, "loss": 1.7007, "step": 21728 }, { "epoch": 1.702899686520376, "grad_norm": 2.403449058532715, "learning_rate": 6.173372645589426e-06, "loss": 1.2559, "step": 21729 }, { "epoch": 1.7029780564263324, "grad_norm": 2.7806990146636963, "learning_rate": 6.173072892617737e-06, "loss": 0.8498, "step": 21730 }, { "epoch": 1.7030564263322883, "grad_norm": 3.4407148361206055, "learning_rate": 6.17277313518428e-06, "loss": 1.5036, "step": 21731 }, { "epoch": 1.7031347962382446, "grad_norm": 3.419402599334717, "learning_rate": 6.1724733732901965e-06, "loss": 1.4827, "step": 21732 }, { "epoch": 1.7032131661442005, "grad_norm": 6.885343074798584, "learning_rate": 6.172173606936628e-06, "loss": 0.7726, "step": 21733 }, { "epoch": 1.7032915360501568, "grad_norm": 2.894437789916992, "learning_rate": 6.171873836124712e-06, "loss": 1.4172, "step": 21734 }, { "epoch": 1.7033699059561127, "grad_norm": 2.5033159255981445, "learning_rate": 6.1715740608555895e-06, "loss": 1.361, "step": 21735 }, { "epoch": 1.703448275862069, "grad_norm": 3.857839822769165, "learning_rate": 6.171274281130404e-06, "loss": 0.9324, "step": 21736 }, { "epoch": 1.703526645768025, "grad_norm": 3.3138673305511475, "learning_rate": 6.170974496950289e-06, "loss": 1.0455, "step": 21737 }, { "epoch": 1.7036050156739813, "grad_norm": 3.316572666168213, "learning_rate": 6.170674708316391e-06, "loss": 1.3236, "step": 21738 }, { "epoch": 1.7036833855799371, "grad_norm": 6.217922687530518, "learning_rate": 6.170374915229847e-06, "loss": 1.5362, "step": 21739 }, { "epoch": 1.7037617554858935, "grad_norm": 2.3813254833221436, "learning_rate": 6.170075117691796e-06, "loss": 1.0605, "step": 21740 }, { "epoch": 1.7038401253918494, "grad_norm": 2.349721908569336, "learning_rate": 6.169775315703382e-06, "loss": 1.2739, "step": 21741 }, { "epoch": 1.7039184952978057, "grad_norm": 5.92578125, "learning_rate": 6.169475509265745e-06, "loss": 1.5304, "step": 21742 }, { "epoch": 1.7039968652037616, "grad_norm": 2.5088372230529785, "learning_rate": 6.169175698380021e-06, "loss": 1.1794, "step": 21743 }, { "epoch": 1.704075235109718, "grad_norm": 3.9263176918029785, "learning_rate": 6.168875883047356e-06, "loss": 1.6233, "step": 21744 }, { "epoch": 1.704153605015674, "grad_norm": 4.706567287445068, "learning_rate": 6.168576063268884e-06, "loss": 1.3313, "step": 21745 }, { "epoch": 1.7042319749216301, "grad_norm": 5.3971052169799805, "learning_rate": 6.168276239045751e-06, "loss": 1.387, "step": 21746 }, { "epoch": 1.7043103448275863, "grad_norm": 2.5306620597839355, "learning_rate": 6.167976410379095e-06, "loss": 1.4198, "step": 21747 }, { "epoch": 1.7043887147335424, "grad_norm": 2.3257670402526855, "learning_rate": 6.167676577270057e-06, "loss": 1.314, "step": 21748 }, { "epoch": 1.7044670846394985, "grad_norm": 3.596709728240967, "learning_rate": 6.167376739719777e-06, "loss": 2.3559, "step": 21749 }, { "epoch": 1.7045454545454546, "grad_norm": 3.7726590633392334, "learning_rate": 6.167076897729396e-06, "loss": 1.2128, "step": 21750 }, { "epoch": 1.7046238244514107, "grad_norm": 4.955302715301514, "learning_rate": 6.166777051300053e-06, "loss": 1.6884, "step": 21751 }, { "epoch": 1.7047021943573668, "grad_norm": 2.7668042182922363, "learning_rate": 6.1664772004328885e-06, "loss": 1.3755, "step": 21752 }, { "epoch": 1.704780564263323, "grad_norm": 2.205207109451294, "learning_rate": 6.166177345129046e-06, "loss": 1.3884, "step": 21753 }, { "epoch": 1.704858934169279, "grad_norm": 2.9624977111816406, "learning_rate": 6.1658774853896615e-06, "loss": 1.7264, "step": 21754 }, { "epoch": 1.7049373040752351, "grad_norm": 4.077892780303955, "learning_rate": 6.1655776212158815e-06, "loss": 0.8924, "step": 21755 }, { "epoch": 1.7050156739811912, "grad_norm": 2.212522506713867, "learning_rate": 6.1652777526088395e-06, "loss": 1.2027, "step": 21756 }, { "epoch": 1.7050940438871474, "grad_norm": 7.046557903289795, "learning_rate": 6.16497787956968e-06, "loss": 1.2065, "step": 21757 }, { "epoch": 1.7051724137931035, "grad_norm": 3.110642433166504, "learning_rate": 6.164678002099544e-06, "loss": 1.6102, "step": 21758 }, { "epoch": 1.7052507836990596, "grad_norm": 3.6462247371673584, "learning_rate": 6.164378120199571e-06, "loss": 1.1104, "step": 21759 }, { "epoch": 1.7053291536050157, "grad_norm": 2.2502567768096924, "learning_rate": 6.164078233870902e-06, "loss": 1.5568, "step": 21760 }, { "epoch": 1.7054075235109718, "grad_norm": 3.103151798248291, "learning_rate": 6.163778343114676e-06, "loss": 1.2883, "step": 21761 }, { "epoch": 1.705485893416928, "grad_norm": 2.993515968322754, "learning_rate": 6.163478447932035e-06, "loss": 1.3276, "step": 21762 }, { "epoch": 1.705564263322884, "grad_norm": 3.0455358028411865, "learning_rate": 6.16317854832412e-06, "loss": 1.28, "step": 21763 }, { "epoch": 1.7056426332288401, "grad_norm": 4.094277381896973, "learning_rate": 6.162878644292073e-06, "loss": 1.3117, "step": 21764 }, { "epoch": 1.7057210031347962, "grad_norm": 2.1389546394348145, "learning_rate": 6.162578735837032e-06, "loss": 1.3302, "step": 21765 }, { "epoch": 1.7057993730407524, "grad_norm": 2.068110466003418, "learning_rate": 6.162278822960137e-06, "loss": 1.472, "step": 21766 }, { "epoch": 1.7058777429467085, "grad_norm": 4.696996688842773, "learning_rate": 6.161978905662531e-06, "loss": 1.3432, "step": 21767 }, { "epoch": 1.7059561128526646, "grad_norm": 2.9773106575012207, "learning_rate": 6.161678983945354e-06, "loss": 1.3306, "step": 21768 }, { "epoch": 1.7060344827586207, "grad_norm": 8.188494682312012, "learning_rate": 6.161379057809748e-06, "loss": 1.6862, "step": 21769 }, { "epoch": 1.7061128526645768, "grad_norm": 6.389729976654053, "learning_rate": 6.161079127256851e-06, "loss": 1.4175, "step": 21770 }, { "epoch": 1.706191222570533, "grad_norm": 6.652254104614258, "learning_rate": 6.160779192287807e-06, "loss": 1.5344, "step": 21771 }, { "epoch": 1.706269592476489, "grad_norm": 2.659350633621216, "learning_rate": 6.160479252903755e-06, "loss": 1.1566, "step": 21772 }, { "epoch": 1.7063479623824451, "grad_norm": 2.1703171730041504, "learning_rate": 6.1601793091058335e-06, "loss": 1.4259, "step": 21773 }, { "epoch": 1.7064263322884012, "grad_norm": 3.2768499851226807, "learning_rate": 6.159879360895189e-06, "loss": 1.4219, "step": 21774 }, { "epoch": 1.7065047021943573, "grad_norm": 2.399564266204834, "learning_rate": 6.1595794082729565e-06, "loss": 1.5233, "step": 21775 }, { "epoch": 1.7065830721003135, "grad_norm": 10.751065254211426, "learning_rate": 6.159279451240282e-06, "loss": 1.7555, "step": 21776 }, { "epoch": 1.7066614420062696, "grad_norm": 2.4739627838134766, "learning_rate": 6.1589794897983026e-06, "loss": 1.3931, "step": 21777 }, { "epoch": 1.7067398119122257, "grad_norm": 2.5526528358459473, "learning_rate": 6.1586795239481614e-06, "loss": 1.1631, "step": 21778 }, { "epoch": 1.7068181818181818, "grad_norm": 3.59808349609375, "learning_rate": 6.1583795536909985e-06, "loss": 1.5158, "step": 21779 }, { "epoch": 1.706896551724138, "grad_norm": 6.968013286590576, "learning_rate": 6.158079579027952e-06, "loss": 0.9756, "step": 21780 }, { "epoch": 1.706974921630094, "grad_norm": 4.337058067321777, "learning_rate": 6.1577795999601685e-06, "loss": 1.4922, "step": 21781 }, { "epoch": 1.7070532915360501, "grad_norm": 2.965151786804199, "learning_rate": 6.157479616488784e-06, "loss": 1.5849, "step": 21782 }, { "epoch": 1.7071316614420062, "grad_norm": 3.339592933654785, "learning_rate": 6.1571796286149445e-06, "loss": 1.3675, "step": 21783 }, { "epoch": 1.7072100313479623, "grad_norm": 2.998967170715332, "learning_rate": 6.156879636339785e-06, "loss": 1.4968, "step": 21784 }, { "epoch": 1.7072884012539185, "grad_norm": 3.042229652404785, "learning_rate": 6.156579639664451e-06, "loss": 1.286, "step": 21785 }, { "epoch": 1.7073667711598746, "grad_norm": 2.6944844722747803, "learning_rate": 6.156279638590083e-06, "loss": 1.4167, "step": 21786 }, { "epoch": 1.7074451410658307, "grad_norm": 2.938894271850586, "learning_rate": 6.155979633117819e-06, "loss": 1.26, "step": 21787 }, { "epoch": 1.7075235109717868, "grad_norm": 1.7903809547424316, "learning_rate": 6.155679623248804e-06, "loss": 1.1809, "step": 21788 }, { "epoch": 1.707601880877743, "grad_norm": 2.2870025634765625, "learning_rate": 6.155379608984175e-06, "loss": 1.1397, "step": 21789 }, { "epoch": 1.707680250783699, "grad_norm": 3.0083422660827637, "learning_rate": 6.155079590325079e-06, "loss": 1.1446, "step": 21790 }, { "epoch": 1.7077586206896553, "grad_norm": 1.9042651653289795, "learning_rate": 6.154779567272651e-06, "loss": 1.3473, "step": 21791 }, { "epoch": 1.7078369905956112, "grad_norm": 3.911632776260376, "learning_rate": 6.154479539828035e-06, "loss": 1.1976, "step": 21792 }, { "epoch": 1.7079153605015676, "grad_norm": 3.542768716812134, "learning_rate": 6.1541795079923726e-06, "loss": 1.5687, "step": 21793 }, { "epoch": 1.7079937304075234, "grad_norm": 3.0291426181793213, "learning_rate": 6.153879471766804e-06, "loss": 1.3434, "step": 21794 }, { "epoch": 1.7080721003134798, "grad_norm": 2.0965983867645264, "learning_rate": 6.153579431152469e-06, "loss": 1.043, "step": 21795 }, { "epoch": 1.7081504702194357, "grad_norm": 3.3404531478881836, "learning_rate": 6.153279386150511e-06, "loss": 1.48, "step": 21796 }, { "epoch": 1.708228840125392, "grad_norm": 2.0826175212860107, "learning_rate": 6.152979336762072e-06, "loss": 0.9969, "step": 21797 }, { "epoch": 1.708307210031348, "grad_norm": 2.1209733486175537, "learning_rate": 6.1526792829882905e-06, "loss": 1.364, "step": 21798 }, { "epoch": 1.7083855799373042, "grad_norm": 3.14274001121521, "learning_rate": 6.15237922483031e-06, "loss": 1.2626, "step": 21799 }, { "epoch": 1.70846394984326, "grad_norm": 4.360422134399414, "learning_rate": 6.152079162289271e-06, "loss": 1.1084, "step": 21800 }, { "epoch": 1.7085423197492164, "grad_norm": 4.442715644836426, "learning_rate": 6.151779095366312e-06, "loss": 0.8202, "step": 21801 }, { "epoch": 1.7086206896551723, "grad_norm": 2.604771852493286, "learning_rate": 6.15147902406258e-06, "loss": 1.2326, "step": 21802 }, { "epoch": 1.7086990595611287, "grad_norm": 1.9246002435684204, "learning_rate": 6.1511789483792115e-06, "loss": 1.3054, "step": 21803 }, { "epoch": 1.7087774294670846, "grad_norm": 2.798654317855835, "learning_rate": 6.15087886831735e-06, "loss": 1.6048, "step": 21804 }, { "epoch": 1.7088557993730409, "grad_norm": 2.5849685668945312, "learning_rate": 6.1505787838781375e-06, "loss": 1.292, "step": 21805 }, { "epoch": 1.7089341692789968, "grad_norm": 4.122807025909424, "learning_rate": 6.150278695062712e-06, "loss": 1.6424, "step": 21806 }, { "epoch": 1.709012539184953, "grad_norm": 3.0177526473999023, "learning_rate": 6.14997860187222e-06, "loss": 1.1086, "step": 21807 }, { "epoch": 1.709090909090909, "grad_norm": 3.3950657844543457, "learning_rate": 6.149678504307798e-06, "loss": 1.2749, "step": 21808 }, { "epoch": 1.7091692789968653, "grad_norm": 4.834907054901123, "learning_rate": 6.149378402370591e-06, "loss": 1.2512, "step": 21809 }, { "epoch": 1.7092476489028212, "grad_norm": 4.280876636505127, "learning_rate": 6.149078296061737e-06, "loss": 1.8955, "step": 21810 }, { "epoch": 1.7093260188087775, "grad_norm": 3.141486167907715, "learning_rate": 6.148778185382381e-06, "loss": 1.0072, "step": 21811 }, { "epoch": 1.7094043887147334, "grad_norm": 2.758042335510254, "learning_rate": 6.1484780703336615e-06, "loss": 1.54, "step": 21812 }, { "epoch": 1.7094827586206898, "grad_norm": 2.708095073699951, "learning_rate": 6.1481779509167214e-06, "loss": 1.2773, "step": 21813 }, { "epoch": 1.7095611285266457, "grad_norm": 3.6617350578308105, "learning_rate": 6.147877827132703e-06, "loss": 1.5664, "step": 21814 }, { "epoch": 1.709639498432602, "grad_norm": 2.5389742851257324, "learning_rate": 6.147577698982746e-06, "loss": 1.4033, "step": 21815 }, { "epoch": 1.7097178683385579, "grad_norm": 6.1248297691345215, "learning_rate": 6.147277566467994e-06, "loss": 0.9964, "step": 21816 }, { "epoch": 1.7097962382445142, "grad_norm": 2.668531656265259, "learning_rate": 6.1469774295895865e-06, "loss": 1.2916, "step": 21817 }, { "epoch": 1.70987460815047, "grad_norm": 2.3425936698913574, "learning_rate": 6.146677288348668e-06, "loss": 1.2739, "step": 21818 }, { "epoch": 1.7099529780564264, "grad_norm": 5.366852760314941, "learning_rate": 6.146377142746376e-06, "loss": 0.7661, "step": 21819 }, { "epoch": 1.7100313479623823, "grad_norm": 2.2132411003112793, "learning_rate": 6.1460769927838535e-06, "loss": 1.3295, "step": 21820 }, { "epoch": 1.7101097178683387, "grad_norm": 3.354360342025757, "learning_rate": 6.145776838462245e-06, "loss": 1.3025, "step": 21821 }, { "epoch": 1.7101880877742945, "grad_norm": 2.713728189468384, "learning_rate": 6.145476679782688e-06, "loss": 1.1976, "step": 21822 }, { "epoch": 1.7102664576802509, "grad_norm": 3.1305642127990723, "learning_rate": 6.1451765167463275e-06, "loss": 0.9839, "step": 21823 }, { "epoch": 1.7103448275862068, "grad_norm": 2.9546871185302734, "learning_rate": 6.144876349354303e-06, "loss": 1.4518, "step": 21824 }, { "epoch": 1.710423197492163, "grad_norm": 7.561145305633545, "learning_rate": 6.144576177607756e-06, "loss": 1.3749, "step": 21825 }, { "epoch": 1.710501567398119, "grad_norm": 3.1222715377807617, "learning_rate": 6.144276001507832e-06, "loss": 1.3472, "step": 21826 }, { "epoch": 1.7105799373040753, "grad_norm": 4.6714701652526855, "learning_rate": 6.143975821055666e-06, "loss": 1.8667, "step": 21827 }, { "epoch": 1.7106583072100312, "grad_norm": 4.137197017669678, "learning_rate": 6.143675636252407e-06, "loss": 1.4819, "step": 21828 }, { "epoch": 1.7107366771159875, "grad_norm": 3.9228837490081787, "learning_rate": 6.14337544709919e-06, "loss": 1.1947, "step": 21829 }, { "epoch": 1.7108150470219434, "grad_norm": 3.9198544025421143, "learning_rate": 6.143075253597163e-06, "loss": 1.2576, "step": 21830 }, { "epoch": 1.7108934169278998, "grad_norm": 2.122628927230835, "learning_rate": 6.142775055747464e-06, "loss": 1.2612, "step": 21831 }, { "epoch": 1.7109717868338556, "grad_norm": 2.5639755725860596, "learning_rate": 6.142474853551235e-06, "loss": 1.3662, "step": 21832 }, { "epoch": 1.711050156739812, "grad_norm": 5.493491172790527, "learning_rate": 6.14217464700962e-06, "loss": 1.8301, "step": 21833 }, { "epoch": 1.7111285266457679, "grad_norm": 2.540111780166626, "learning_rate": 6.141874436123757e-06, "loss": 1.0568, "step": 21834 }, { "epoch": 1.7112068965517242, "grad_norm": 2.388798475265503, "learning_rate": 6.141574220894793e-06, "loss": 1.0879, "step": 21835 }, { "epoch": 1.71128526645768, "grad_norm": 2.049013137817383, "learning_rate": 6.141274001323864e-06, "loss": 1.1282, "step": 21836 }, { "epoch": 1.7113636363636364, "grad_norm": 5.578799724578857, "learning_rate": 6.140973777412117e-06, "loss": 1.4235, "step": 21837 }, { "epoch": 1.7114420062695925, "grad_norm": 2.5486435890197754, "learning_rate": 6.140673549160691e-06, "loss": 1.5054, "step": 21838 }, { "epoch": 1.7115203761755486, "grad_norm": 2.913266181945801, "learning_rate": 6.140373316570728e-06, "loss": 1.4172, "step": 21839 }, { "epoch": 1.7115987460815048, "grad_norm": 2.2613911628723145, "learning_rate": 6.140073079643372e-06, "loss": 1.2561, "step": 21840 }, { "epoch": 1.7116771159874609, "grad_norm": 2.2498984336853027, "learning_rate": 6.139772838379761e-06, "loss": 1.2099, "step": 21841 }, { "epoch": 1.711755485893417, "grad_norm": 2.6463682651519775, "learning_rate": 6.139472592781043e-06, "loss": 1.1612, "step": 21842 }, { "epoch": 1.711833855799373, "grad_norm": 6.820378303527832, "learning_rate": 6.139172342848354e-06, "loss": 1.4858, "step": 21843 }, { "epoch": 1.7119122257053292, "grad_norm": 4.361903190612793, "learning_rate": 6.13887208858284e-06, "loss": 1.699, "step": 21844 }, { "epoch": 1.7119905956112853, "grad_norm": 2.9566617012023926, "learning_rate": 6.1385718299856424e-06, "loss": 1.2198, "step": 21845 }, { "epoch": 1.7120689655172414, "grad_norm": 3.3388454914093018, "learning_rate": 6.138271567057901e-06, "loss": 1.4005, "step": 21846 }, { "epoch": 1.7121473354231975, "grad_norm": 2.73410964012146, "learning_rate": 6.137971299800759e-06, "loss": 1.4127, "step": 21847 }, { "epoch": 1.7122257053291536, "grad_norm": 4.169677257537842, "learning_rate": 6.137671028215358e-06, "loss": 1.2643, "step": 21848 }, { "epoch": 1.7123040752351097, "grad_norm": 3.2626757621765137, "learning_rate": 6.137370752302843e-06, "loss": 1.4961, "step": 21849 }, { "epoch": 1.7123824451410659, "grad_norm": 3.5377273559570312, "learning_rate": 6.137070472064351e-06, "loss": 1.3368, "step": 21850 }, { "epoch": 1.712460815047022, "grad_norm": 5.844203948974609, "learning_rate": 6.136770187501029e-06, "loss": 1.7367, "step": 21851 }, { "epoch": 1.712539184952978, "grad_norm": 3.090388774871826, "learning_rate": 6.136469898614016e-06, "loss": 1.5227, "step": 21852 }, { "epoch": 1.7126175548589342, "grad_norm": 2.345679759979248, "learning_rate": 6.136169605404456e-06, "loss": 1.4446, "step": 21853 }, { "epoch": 1.7126959247648903, "grad_norm": 2.4682233333587646, "learning_rate": 6.135869307873491e-06, "loss": 1.5842, "step": 21854 }, { "epoch": 1.7127742946708464, "grad_norm": 9.954429626464844, "learning_rate": 6.135569006022261e-06, "loss": 1.0813, "step": 21855 }, { "epoch": 1.7128526645768025, "grad_norm": 8.99574089050293, "learning_rate": 6.135268699851911e-06, "loss": 0.5796, "step": 21856 }, { "epoch": 1.7129310344827586, "grad_norm": 2.106492757797241, "learning_rate": 6.134968389363581e-06, "loss": 1.4982, "step": 21857 }, { "epoch": 1.7130094043887147, "grad_norm": 3.1398651599884033, "learning_rate": 6.134668074558416e-06, "loss": 1.3276, "step": 21858 }, { "epoch": 1.7130877742946709, "grad_norm": 4.753034591674805, "learning_rate": 6.1343677554375545e-06, "loss": 1.3842, "step": 21859 }, { "epoch": 1.713166144200627, "grad_norm": 4.527919769287109, "learning_rate": 6.134067432002141e-06, "loss": 1.0635, "step": 21860 }, { "epoch": 1.713244514106583, "grad_norm": 2.160217761993408, "learning_rate": 6.1337671042533185e-06, "loss": 1.1934, "step": 21861 }, { "epoch": 1.7133228840125392, "grad_norm": 2.67034912109375, "learning_rate": 6.133466772192227e-06, "loss": 1.3372, "step": 21862 }, { "epoch": 1.7134012539184953, "grad_norm": 6.928555011749268, "learning_rate": 6.133166435820011e-06, "loss": 1.3361, "step": 21863 }, { "epoch": 1.7134796238244514, "grad_norm": 2.7147674560546875, "learning_rate": 6.13286609513781e-06, "loss": 1.1857, "step": 21864 }, { "epoch": 1.7135579937304075, "grad_norm": 2.1730844974517822, "learning_rate": 6.132565750146771e-06, "loss": 1.2682, "step": 21865 }, { "epoch": 1.7136363636363636, "grad_norm": 4.172915458679199, "learning_rate": 6.132265400848033e-06, "loss": 1.6001, "step": 21866 }, { "epoch": 1.7137147335423197, "grad_norm": 3.3841941356658936, "learning_rate": 6.131965047242738e-06, "loss": 1.5841, "step": 21867 }, { "epoch": 1.7137931034482758, "grad_norm": 3.0340802669525146, "learning_rate": 6.131664689332031e-06, "loss": 1.2169, "step": 21868 }, { "epoch": 1.713871473354232, "grad_norm": 2.496514320373535, "learning_rate": 6.131364327117051e-06, "loss": 1.4631, "step": 21869 }, { "epoch": 1.713949843260188, "grad_norm": 1.6341465711593628, "learning_rate": 6.131063960598942e-06, "loss": 0.9921, "step": 21870 }, { "epoch": 1.7140282131661442, "grad_norm": 3.7855544090270996, "learning_rate": 6.130763589778847e-06, "loss": 1.8022, "step": 21871 }, { "epoch": 1.7141065830721003, "grad_norm": 3.273456335067749, "learning_rate": 6.130463214657909e-06, "loss": 1.3001, "step": 21872 }, { "epoch": 1.7141849529780564, "grad_norm": 3.450920343399048, "learning_rate": 6.13016283523727e-06, "loss": 1.5211, "step": 21873 }, { "epoch": 1.7142633228840125, "grad_norm": 2.866962194442749, "learning_rate": 6.129862451518071e-06, "loss": 1.5391, "step": 21874 }, { "epoch": 1.7143416927899686, "grad_norm": 2.7414774894714355, "learning_rate": 6.129562063501456e-06, "loss": 1.4032, "step": 21875 }, { "epoch": 1.7144200626959247, "grad_norm": 4.100192070007324, "learning_rate": 6.129261671188567e-06, "loss": 1.4524, "step": 21876 }, { "epoch": 1.7144984326018808, "grad_norm": 2.6152801513671875, "learning_rate": 6.128961274580546e-06, "loss": 1.3832, "step": 21877 }, { "epoch": 1.714576802507837, "grad_norm": 2.3071072101593018, "learning_rate": 6.128660873678537e-06, "loss": 1.2413, "step": 21878 }, { "epoch": 1.714655172413793, "grad_norm": 2.0844807624816895, "learning_rate": 6.128360468483682e-06, "loss": 1.2277, "step": 21879 }, { "epoch": 1.7147335423197492, "grad_norm": 2.7450594902038574, "learning_rate": 6.1280600589971225e-06, "loss": 1.2008, "step": 21880 }, { "epoch": 1.7148119122257053, "grad_norm": 5.50192403793335, "learning_rate": 6.127759645220002e-06, "loss": 1.7179, "step": 21881 }, { "epoch": 1.7148902821316614, "grad_norm": 2.2883267402648926, "learning_rate": 6.1274592271534635e-06, "loss": 1.2008, "step": 21882 }, { "epoch": 1.7149686520376175, "grad_norm": 2.7351462841033936, "learning_rate": 6.127158804798648e-06, "loss": 1.5625, "step": 21883 }, { "epoch": 1.7150470219435736, "grad_norm": 2.733729362487793, "learning_rate": 6.126858378156701e-06, "loss": 0.9688, "step": 21884 }, { "epoch": 1.7151253918495297, "grad_norm": 5.086834907531738, "learning_rate": 6.126557947228763e-06, "loss": 1.68, "step": 21885 }, { "epoch": 1.715203761755486, "grad_norm": 3.9001786708831787, "learning_rate": 6.126257512015977e-06, "loss": 1.3185, "step": 21886 }, { "epoch": 1.715282131661442, "grad_norm": 2.635955572128296, "learning_rate": 6.125957072519486e-06, "loss": 1.1633, "step": 21887 }, { "epoch": 1.7153605015673983, "grad_norm": 3.1468088626861572, "learning_rate": 6.125656628740433e-06, "loss": 1.2888, "step": 21888 }, { "epoch": 1.7154388714733542, "grad_norm": 3.2937300205230713, "learning_rate": 6.12535618067996e-06, "loss": 1.148, "step": 21889 }, { "epoch": 1.7155172413793105, "grad_norm": 3.527827262878418, "learning_rate": 6.125055728339209e-06, "loss": 0.4923, "step": 21890 }, { "epoch": 1.7155956112852664, "grad_norm": 2.223773717880249, "learning_rate": 6.124755271719326e-06, "loss": 1.2893, "step": 21891 }, { "epoch": 1.7156739811912227, "grad_norm": 3.915323495864868, "learning_rate": 6.124454810821451e-06, "loss": 1.4384, "step": 21892 }, { "epoch": 1.7157523510971786, "grad_norm": 3.2230420112609863, "learning_rate": 6.124154345646727e-06, "loss": 1.4929, "step": 21893 }, { "epoch": 1.715830721003135, "grad_norm": 3.168186902999878, "learning_rate": 6.123853876196297e-06, "loss": 1.4438, "step": 21894 }, { "epoch": 1.7159090909090908, "grad_norm": 4.697486400604248, "learning_rate": 6.123553402471304e-06, "loss": 1.4198, "step": 21895 }, { "epoch": 1.7159874608150472, "grad_norm": 2.081204414367676, "learning_rate": 6.123252924472891e-06, "loss": 0.9246, "step": 21896 }, { "epoch": 1.716065830721003, "grad_norm": 3.1719112396240234, "learning_rate": 6.122952442202202e-06, "loss": 1.3924, "step": 21897 }, { "epoch": 1.7161442006269594, "grad_norm": 3.475531816482544, "learning_rate": 6.122651955660377e-06, "loss": 1.3523, "step": 21898 }, { "epoch": 1.7162225705329153, "grad_norm": 2.740743398666382, "learning_rate": 6.122351464848563e-06, "loss": 1.3475, "step": 21899 }, { "epoch": 1.7163009404388716, "grad_norm": 2.022176504135132, "learning_rate": 6.122050969767898e-06, "loss": 1.4216, "step": 21900 }, { "epoch": 1.7163793103448275, "grad_norm": 3.0893867015838623, "learning_rate": 6.121750470419528e-06, "loss": 1.3725, "step": 21901 }, { "epoch": 1.7164576802507838, "grad_norm": 3.49269700050354, "learning_rate": 6.121449966804597e-06, "loss": 1.5347, "step": 21902 }, { "epoch": 1.7165360501567397, "grad_norm": 3.8329875469207764, "learning_rate": 6.1211494589242444e-06, "loss": 1.3091, "step": 21903 }, { "epoch": 1.716614420062696, "grad_norm": 2.924341917037964, "learning_rate": 6.120848946779616e-06, "loss": 1.0487, "step": 21904 }, { "epoch": 1.716692789968652, "grad_norm": 4.390443801879883, "learning_rate": 6.120548430371853e-06, "loss": 1.2921, "step": 21905 }, { "epoch": 1.7167711598746083, "grad_norm": 3.021239995956421, "learning_rate": 6.120247909702101e-06, "loss": 1.6301, "step": 21906 }, { "epoch": 1.7168495297805642, "grad_norm": 2.9376895427703857, "learning_rate": 6.1199473847715e-06, "loss": 1.4157, "step": 21907 }, { "epoch": 1.7169278996865205, "grad_norm": 9.676614761352539, "learning_rate": 6.119646855581196e-06, "loss": 1.0188, "step": 21908 }, { "epoch": 1.7170062695924764, "grad_norm": 3.008791446685791, "learning_rate": 6.119346322132328e-06, "loss": 1.4368, "step": 21909 }, { "epoch": 1.7170846394984327, "grad_norm": 5.591065883636475, "learning_rate": 6.1190457844260434e-06, "loss": 1.5183, "step": 21910 }, { "epoch": 1.7171630094043886, "grad_norm": 5.005906105041504, "learning_rate": 6.118745242463482e-06, "loss": 1.9748, "step": 21911 }, { "epoch": 1.717241379310345, "grad_norm": 4.024512767791748, "learning_rate": 6.1184446962457896e-06, "loss": 1.4159, "step": 21912 }, { "epoch": 1.7173197492163008, "grad_norm": 7.287861347198486, "learning_rate": 6.118144145774108e-06, "loss": 1.6385, "step": 21913 }, { "epoch": 1.7173981191222571, "grad_norm": 1.8990956544876099, "learning_rate": 6.117843591049579e-06, "loss": 1.5814, "step": 21914 }, { "epoch": 1.717476489028213, "grad_norm": 2.2577099800109863, "learning_rate": 6.1175430320733485e-06, "loss": 1.2623, "step": 21915 }, { "epoch": 1.7175548589341694, "grad_norm": 3.4515838623046875, "learning_rate": 6.1172424688465586e-06, "loss": 0.7691, "step": 21916 }, { "epoch": 1.7176332288401253, "grad_norm": 2.952864646911621, "learning_rate": 6.116941901370351e-06, "loss": 1.3681, "step": 21917 }, { "epoch": 1.7177115987460816, "grad_norm": 2.898678779602051, "learning_rate": 6.116641329645871e-06, "loss": 1.3664, "step": 21918 }, { "epoch": 1.7177899686520375, "grad_norm": 4.57528829574585, "learning_rate": 6.11634075367426e-06, "loss": 1.5039, "step": 21919 }, { "epoch": 1.7178683385579938, "grad_norm": 3.9994471073150635, "learning_rate": 6.116040173456663e-06, "loss": 1.4708, "step": 21920 }, { "epoch": 1.7179467084639497, "grad_norm": 2.980125665664673, "learning_rate": 6.115739588994222e-06, "loss": 1.5477, "step": 21921 }, { "epoch": 1.718025078369906, "grad_norm": 3.0386805534362793, "learning_rate": 6.11543900028808e-06, "loss": 1.2819, "step": 21922 }, { "epoch": 1.718103448275862, "grad_norm": 3.0346176624298096, "learning_rate": 6.115138407339383e-06, "loss": 1.2826, "step": 21923 }, { "epoch": 1.7181818181818183, "grad_norm": 2.9160821437835693, "learning_rate": 6.11483781014927e-06, "loss": 1.4385, "step": 21924 }, { "epoch": 1.7182601880877741, "grad_norm": 3.5291829109191895, "learning_rate": 6.114537208718888e-06, "loss": 1.1352, "step": 21925 }, { "epoch": 1.7183385579937305, "grad_norm": 2.967397451400757, "learning_rate": 6.114236603049378e-06, "loss": 1.4202, "step": 21926 }, { "epoch": 1.7184169278996864, "grad_norm": 2.7250351905822754, "learning_rate": 6.113935993141886e-06, "loss": 1.1317, "step": 21927 }, { "epoch": 1.7184952978056427, "grad_norm": 5.130458354949951, "learning_rate": 6.113635378997551e-06, "loss": 1.514, "step": 21928 }, { "epoch": 1.7185736677115986, "grad_norm": 5.946170806884766, "learning_rate": 6.113334760617522e-06, "loss": 1.4296, "step": 21929 }, { "epoch": 1.718652037617555, "grad_norm": 2.484161615371704, "learning_rate": 6.113034138002938e-06, "loss": 1.4906, "step": 21930 }, { "epoch": 1.7187304075235108, "grad_norm": 2.360349416732788, "learning_rate": 6.1127335111549436e-06, "loss": 1.5931, "step": 21931 }, { "epoch": 1.7188087774294671, "grad_norm": 1.8727526664733887, "learning_rate": 6.112432880074682e-06, "loss": 0.6292, "step": 21932 }, { "epoch": 1.7188871473354232, "grad_norm": 3.540405035018921, "learning_rate": 6.112132244763298e-06, "loss": 1.8126, "step": 21933 }, { "epoch": 1.7189655172413794, "grad_norm": 2.9025042057037354, "learning_rate": 6.111831605221934e-06, "loss": 1.5224, "step": 21934 }, { "epoch": 1.7190438871473355, "grad_norm": 1.9830845594406128, "learning_rate": 6.111530961451733e-06, "loss": 1.2661, "step": 21935 }, { "epoch": 1.7191222570532916, "grad_norm": 2.171963930130005, "learning_rate": 6.11123031345384e-06, "loss": 1.4269, "step": 21936 }, { "epoch": 1.7192006269592477, "grad_norm": 2.0992472171783447, "learning_rate": 6.110929661229397e-06, "loss": 1.2789, "step": 21937 }, { "epoch": 1.7192789968652038, "grad_norm": 3.1058740615844727, "learning_rate": 6.110629004779549e-06, "loss": 1.1993, "step": 21938 }, { "epoch": 1.71935736677116, "grad_norm": 3.839346408843994, "learning_rate": 6.110328344105439e-06, "loss": 1.0715, "step": 21939 }, { "epoch": 1.719435736677116, "grad_norm": 4.644327163696289, "learning_rate": 6.110027679208208e-06, "loss": 1.4639, "step": 21940 }, { "epoch": 1.7195141065830721, "grad_norm": 2.2672739028930664, "learning_rate": 6.109727010089004e-06, "loss": 1.3871, "step": 21941 }, { "epoch": 1.7195924764890282, "grad_norm": 2.3522391319274902, "learning_rate": 6.109426336748967e-06, "loss": 1.2328, "step": 21942 }, { "epoch": 1.7196708463949844, "grad_norm": 2.887636661529541, "learning_rate": 6.109125659189244e-06, "loss": 1.6332, "step": 21943 }, { "epoch": 1.7197492163009405, "grad_norm": 2.6076881885528564, "learning_rate": 6.108824977410975e-06, "loss": 0.8513, "step": 21944 }, { "epoch": 1.7198275862068966, "grad_norm": 2.538820505142212, "learning_rate": 6.108524291415306e-06, "loss": 1.5229, "step": 21945 }, { "epoch": 1.7199059561128527, "grad_norm": 2.9857516288757324, "learning_rate": 6.10822360120338e-06, "loss": 1.227, "step": 21946 }, { "epoch": 1.7199843260188088, "grad_norm": 2.9954421520233154, "learning_rate": 6.10792290677634e-06, "loss": 1.247, "step": 21947 }, { "epoch": 1.720062695924765, "grad_norm": 2.7401323318481445, "learning_rate": 6.107622208135331e-06, "loss": 1.286, "step": 21948 }, { "epoch": 1.720141065830721, "grad_norm": 2.967841148376465, "learning_rate": 6.107321505281495e-06, "loss": 1.6604, "step": 21949 }, { "epoch": 1.7202194357366771, "grad_norm": 3.375746250152588, "learning_rate": 6.107020798215978e-06, "loss": 1.1593, "step": 21950 }, { "epoch": 1.7202978056426332, "grad_norm": 5.544539451599121, "learning_rate": 6.106720086939921e-06, "loss": 1.1187, "step": 21951 }, { "epoch": 1.7203761755485893, "grad_norm": 3.1003758907318115, "learning_rate": 6.10641937145447e-06, "loss": 1.4001, "step": 21952 }, { "epoch": 1.7204545454545455, "grad_norm": 4.017236232757568, "learning_rate": 6.1061186517607684e-06, "loss": 1.4684, "step": 21953 }, { "epoch": 1.7205329153605016, "grad_norm": 8.29569149017334, "learning_rate": 6.105817927859959e-06, "loss": 0.9769, "step": 21954 }, { "epoch": 1.7206112852664577, "grad_norm": 2.6914942264556885, "learning_rate": 6.105517199753188e-06, "loss": 1.6009, "step": 21955 }, { "epoch": 1.7206896551724138, "grad_norm": 6.01509952545166, "learning_rate": 6.105216467441595e-06, "loss": 1.7735, "step": 21956 }, { "epoch": 1.72076802507837, "grad_norm": 4.330306053161621, "learning_rate": 6.104915730926329e-06, "loss": 1.2028, "step": 21957 }, { "epoch": 1.720846394984326, "grad_norm": 3.700418710708618, "learning_rate": 6.104614990208528e-06, "loss": 1.3267, "step": 21958 }, { "epoch": 1.7209247648902821, "grad_norm": 3.663283586502075, "learning_rate": 6.1043142452893414e-06, "loss": 1.3098, "step": 21959 }, { "epoch": 1.7210031347962382, "grad_norm": 3.2283589839935303, "learning_rate": 6.104013496169909e-06, "loss": 1.5058, "step": 21960 }, { "epoch": 1.7210815047021943, "grad_norm": 2.9509456157684326, "learning_rate": 6.103712742851377e-06, "loss": 1.1686, "step": 21961 }, { "epoch": 1.7211598746081505, "grad_norm": 5.011630535125732, "learning_rate": 6.1034119853348885e-06, "loss": 1.2694, "step": 21962 }, { "epoch": 1.7212382445141066, "grad_norm": 5.376773834228516, "learning_rate": 6.103111223621587e-06, "loss": 1.264, "step": 21963 }, { "epoch": 1.7213166144200627, "grad_norm": 3.785268783569336, "learning_rate": 6.102810457712618e-06, "loss": 1.4429, "step": 21964 }, { "epoch": 1.7213949843260188, "grad_norm": 1.894707441329956, "learning_rate": 6.102509687609123e-06, "loss": 1.1092, "step": 21965 }, { "epoch": 1.721473354231975, "grad_norm": 2.494478702545166, "learning_rate": 6.10220891331225e-06, "loss": 1.6367, "step": 21966 }, { "epoch": 1.721551724137931, "grad_norm": 4.023717880249023, "learning_rate": 6.101908134823138e-06, "loss": 1.5908, "step": 21967 }, { "epoch": 1.7216300940438871, "grad_norm": 4.800234794616699, "learning_rate": 6.101607352142934e-06, "loss": 1.2537, "step": 21968 }, { "epoch": 1.7217084639498432, "grad_norm": 3.575270414352417, "learning_rate": 6.101306565272782e-06, "loss": 1.4715, "step": 21969 }, { "epoch": 1.7217868338557993, "grad_norm": 3.1324455738067627, "learning_rate": 6.1010057742138255e-06, "loss": 1.5356, "step": 21970 }, { "epoch": 1.7218652037617554, "grad_norm": 3.9622995853424072, "learning_rate": 6.100704978967208e-06, "loss": 1.2936, "step": 21971 }, { "epoch": 1.7219435736677116, "grad_norm": 4.312772750854492, "learning_rate": 6.100404179534074e-06, "loss": 1.4459, "step": 21972 }, { "epoch": 1.7220219435736677, "grad_norm": 3.7885518074035645, "learning_rate": 6.100103375915569e-06, "loss": 1.3138, "step": 21973 }, { "epoch": 1.7221003134796238, "grad_norm": 6.82962703704834, "learning_rate": 6.0998025681128346e-06, "loss": 1.5541, "step": 21974 }, { "epoch": 1.72217868338558, "grad_norm": 2.705777645111084, "learning_rate": 6.099501756127017e-06, "loss": 1.3374, "step": 21975 }, { "epoch": 1.722257053291536, "grad_norm": 5.4300360679626465, "learning_rate": 6.099200939959258e-06, "loss": 1.4201, "step": 21976 }, { "epoch": 1.722335423197492, "grad_norm": 4.015382766723633, "learning_rate": 6.098900119610704e-06, "loss": 1.6451, "step": 21977 }, { "epoch": 1.7224137931034482, "grad_norm": 4.057199478149414, "learning_rate": 6.098599295082498e-06, "loss": 1.5295, "step": 21978 }, { "epoch": 1.7224921630094046, "grad_norm": 3.357220411300659, "learning_rate": 6.098298466375785e-06, "loss": 1.5355, "step": 21979 }, { "epoch": 1.7225705329153604, "grad_norm": 2.638278007507324, "learning_rate": 6.097997633491709e-06, "loss": 1.4722, "step": 21980 }, { "epoch": 1.7226489028213168, "grad_norm": 3.5302789211273193, "learning_rate": 6.097696796431415e-06, "loss": 1.5541, "step": 21981 }, { "epoch": 1.7227272727272727, "grad_norm": 2.3325698375701904, "learning_rate": 6.097395955196044e-06, "loss": 1.705, "step": 21982 }, { "epoch": 1.722805642633229, "grad_norm": 9.709606170654297, "learning_rate": 6.097095109786744e-06, "loss": 0.4324, "step": 21983 }, { "epoch": 1.7228840125391849, "grad_norm": 2.5481557846069336, "learning_rate": 6.096794260204656e-06, "loss": 1.5047, "step": 21984 }, { "epoch": 1.7229623824451412, "grad_norm": 2.6593592166900635, "learning_rate": 6.096493406450927e-06, "loss": 1.2761, "step": 21985 }, { "epoch": 1.723040752351097, "grad_norm": 3.2120540142059326, "learning_rate": 6.0961925485267e-06, "loss": 0.945, "step": 21986 }, { "epoch": 1.7231191222570534, "grad_norm": 3.6908464431762695, "learning_rate": 6.095891686433121e-06, "loss": 1.5391, "step": 21987 }, { "epoch": 1.7231974921630093, "grad_norm": 7.2144670486450195, "learning_rate": 6.095590820171332e-06, "loss": 1.5313, "step": 21988 }, { "epoch": 1.7232758620689657, "grad_norm": 2.57594895362854, "learning_rate": 6.0952899497424775e-06, "loss": 1.1376, "step": 21989 }, { "epoch": 1.7233542319749215, "grad_norm": 3.772080421447754, "learning_rate": 6.0949890751477034e-06, "loss": 1.6086, "step": 21990 }, { "epoch": 1.7234326018808779, "grad_norm": 2.898768663406372, "learning_rate": 6.094688196388152e-06, "loss": 1.2368, "step": 21991 }, { "epoch": 1.7235109717868338, "grad_norm": 2.543445348739624, "learning_rate": 6.094387313464971e-06, "loss": 1.0221, "step": 21992 }, { "epoch": 1.72358934169279, "grad_norm": 2.902468204498291, "learning_rate": 6.094086426379302e-06, "loss": 1.4353, "step": 21993 }, { "epoch": 1.723667711598746, "grad_norm": 2.511579751968384, "learning_rate": 6.09378553513229e-06, "loss": 1.8668, "step": 21994 }, { "epoch": 1.7237460815047023, "grad_norm": 3.077694892883301, "learning_rate": 6.093484639725079e-06, "loss": 1.4517, "step": 21995 }, { "epoch": 1.7238244514106582, "grad_norm": 7.112820625305176, "learning_rate": 6.093183740158815e-06, "loss": 0.92, "step": 21996 }, { "epoch": 1.7239028213166145, "grad_norm": 4.624963283538818, "learning_rate": 6.092882836434641e-06, "loss": 1.0861, "step": 21997 }, { "epoch": 1.7239811912225704, "grad_norm": 3.2037954330444336, "learning_rate": 6.0925819285537025e-06, "loss": 1.2797, "step": 21998 }, { "epoch": 1.7240595611285268, "grad_norm": 4.31157112121582, "learning_rate": 6.092281016517143e-06, "loss": 1.4561, "step": 21999 }, { "epoch": 1.7241379310344827, "grad_norm": 1.9938297271728516, "learning_rate": 6.091980100326109e-06, "loss": 1.0209, "step": 22000 }, { "epoch": 1.724216300940439, "grad_norm": 3.4020283222198486, "learning_rate": 6.091679179981742e-06, "loss": 1.4771, "step": 22001 }, { "epoch": 1.7242946708463949, "grad_norm": 2.336670160293579, "learning_rate": 6.091378255485188e-06, "loss": 1.4484, "step": 22002 }, { "epoch": 1.7243730407523512, "grad_norm": 4.596964359283447, "learning_rate": 6.091077326837594e-06, "loss": 1.0154, "step": 22003 }, { "epoch": 1.724451410658307, "grad_norm": 3.3099358081817627, "learning_rate": 6.0907763940400984e-06, "loss": 1.5692, "step": 22004 }, { "epoch": 1.7245297805642634, "grad_norm": 4.439035892486572, "learning_rate": 6.090475457093853e-06, "loss": 1.5573, "step": 22005 }, { "epoch": 1.7246081504702193, "grad_norm": 3.7591030597686768, "learning_rate": 6.090174515999997e-06, "loss": 1.1572, "step": 22006 }, { "epoch": 1.7246865203761756, "grad_norm": 4.230563640594482, "learning_rate": 6.089873570759678e-06, "loss": 1.1507, "step": 22007 }, { "epoch": 1.7247648902821315, "grad_norm": 3.030644178390503, "learning_rate": 6.089572621374039e-06, "loss": 1.6315, "step": 22008 }, { "epoch": 1.7248432601880879, "grad_norm": 3.3403377532958984, "learning_rate": 6.089271667844228e-06, "loss": 1.163, "step": 22009 }, { "epoch": 1.7249216300940438, "grad_norm": 3.8772568702697754, "learning_rate": 6.088970710171384e-06, "loss": 1.6128, "step": 22010 }, { "epoch": 1.725, "grad_norm": 3.5867204666137695, "learning_rate": 6.088669748356656e-06, "loss": 1.4622, "step": 22011 }, { "epoch": 1.725078369905956, "grad_norm": 2.8973841667175293, "learning_rate": 6.088368782401186e-06, "loss": 1.0306, "step": 22012 }, { "epoch": 1.7251567398119123, "grad_norm": 3.8052186965942383, "learning_rate": 6.0880678123061235e-06, "loss": 1.3595, "step": 22013 }, { "epoch": 1.7252351097178682, "grad_norm": 4.078466892242432, "learning_rate": 6.087766838072607e-06, "loss": 1.4872, "step": 22014 }, { "epoch": 1.7253134796238245, "grad_norm": 3.669384002685547, "learning_rate": 6.087465859701784e-06, "loss": 1.4387, "step": 22015 }, { "epoch": 1.7253918495297804, "grad_norm": 2.562380075454712, "learning_rate": 6.087164877194802e-06, "loss": 1.347, "step": 22016 }, { "epoch": 1.7254702194357368, "grad_norm": 3.4504966735839844, "learning_rate": 6.086863890552801e-06, "loss": 1.6398, "step": 22017 }, { "epoch": 1.7255485893416926, "grad_norm": 2.8820443153381348, "learning_rate": 6.08656289977693e-06, "loss": 1.4458, "step": 22018 }, { "epoch": 1.725626959247649, "grad_norm": 2.000474691390991, "learning_rate": 6.086261904868329e-06, "loss": 1.3307, "step": 22019 }, { "epoch": 1.7257053291536049, "grad_norm": 4.463290691375732, "learning_rate": 6.085960905828147e-06, "loss": 1.4488, "step": 22020 }, { "epoch": 1.7257836990595612, "grad_norm": 5.520115375518799, "learning_rate": 6.0856599026575284e-06, "loss": 1.4081, "step": 22021 }, { "epoch": 1.725862068965517, "grad_norm": 4.027019500732422, "learning_rate": 6.085358895357616e-06, "loss": 1.7265, "step": 22022 }, { "epoch": 1.7259404388714734, "grad_norm": 3.642429828643799, "learning_rate": 6.085057883929557e-06, "loss": 1.2012, "step": 22023 }, { "epoch": 1.7260188087774293, "grad_norm": 3.975694179534912, "learning_rate": 6.084756868374494e-06, "loss": 1.6674, "step": 22024 }, { "epoch": 1.7260971786833856, "grad_norm": 5.771420478820801, "learning_rate": 6.084455848693573e-06, "loss": 1.4661, "step": 22025 }, { "epoch": 1.7261755485893417, "grad_norm": 5.528365612030029, "learning_rate": 6.08415482488794e-06, "loss": 1.6446, "step": 22026 }, { "epoch": 1.7262539184952979, "grad_norm": 3.1632165908813477, "learning_rate": 6.083853796958737e-06, "loss": 1.2065, "step": 22027 }, { "epoch": 1.726332288401254, "grad_norm": 2.6305549144744873, "learning_rate": 6.083552764907114e-06, "loss": 1.3934, "step": 22028 }, { "epoch": 1.72641065830721, "grad_norm": 5.3131818771362305, "learning_rate": 6.083251728734211e-06, "loss": 2.2722, "step": 22029 }, { "epoch": 1.7264890282131662, "grad_norm": 4.448073387145996, "learning_rate": 6.082950688441174e-06, "loss": 1.4446, "step": 22030 }, { "epoch": 1.7265673981191223, "grad_norm": 2.3116164207458496, "learning_rate": 6.082649644029151e-06, "loss": 1.4682, "step": 22031 }, { "epoch": 1.7266457680250784, "grad_norm": 1.813544750213623, "learning_rate": 6.082348595499282e-06, "loss": 1.4591, "step": 22032 }, { "epoch": 1.7267241379310345, "grad_norm": 3.5021846294403076, "learning_rate": 6.082047542852716e-06, "loss": 1.3413, "step": 22033 }, { "epoch": 1.7268025078369906, "grad_norm": 3.165745496749878, "learning_rate": 6.081746486090597e-06, "loss": 1.6638, "step": 22034 }, { "epoch": 1.7268808777429467, "grad_norm": 2.168778896331787, "learning_rate": 6.081445425214072e-06, "loss": 1.3653, "step": 22035 }, { "epoch": 1.7269592476489029, "grad_norm": 2.8351261615753174, "learning_rate": 6.08114436022428e-06, "loss": 1.0071, "step": 22036 }, { "epoch": 1.727037617554859, "grad_norm": 2.838042736053467, "learning_rate": 6.080843291122375e-06, "loss": 1.7406, "step": 22037 }, { "epoch": 1.727115987460815, "grad_norm": 3.9369728565216064, "learning_rate": 6.080542217909493e-06, "loss": 1.7444, "step": 22038 }, { "epoch": 1.7271943573667712, "grad_norm": 3.4110124111175537, "learning_rate": 6.080241140586786e-06, "loss": 1.8469, "step": 22039 }, { "epoch": 1.7272727272727273, "grad_norm": 2.4878957271575928, "learning_rate": 6.0799400591553956e-06, "loss": 1.2468, "step": 22040 }, { "epoch": 1.7273510971786834, "grad_norm": 3.21274733543396, "learning_rate": 6.079638973616469e-06, "loss": 1.4455, "step": 22041 }, { "epoch": 1.7274294670846395, "grad_norm": 2.0966994762420654, "learning_rate": 6.079337883971149e-06, "loss": 1.1275, "step": 22042 }, { "epoch": 1.7275078369905956, "grad_norm": 6.645318984985352, "learning_rate": 6.079036790220581e-06, "loss": 0.8407, "step": 22043 }, { "epoch": 1.7275862068965517, "grad_norm": 3.2409095764160156, "learning_rate": 6.078735692365913e-06, "loss": 1.4584, "step": 22044 }, { "epoch": 1.7276645768025078, "grad_norm": 5.570581912994385, "learning_rate": 6.078434590408287e-06, "loss": 1.2917, "step": 22045 }, { "epoch": 1.727742946708464, "grad_norm": 2.288142204284668, "learning_rate": 6.0781334843488524e-06, "loss": 1.1983, "step": 22046 }, { "epoch": 1.72782131661442, "grad_norm": 4.216975212097168, "learning_rate": 6.07783237418875e-06, "loss": 1.2817, "step": 22047 }, { "epoch": 1.7278996865203762, "grad_norm": 3.2725887298583984, "learning_rate": 6.077531259929128e-06, "loss": 0.996, "step": 22048 }, { "epoch": 1.7279780564263323, "grad_norm": 2.8449926376342773, "learning_rate": 6.077230141571129e-06, "loss": 1.1701, "step": 22049 }, { "epoch": 1.7280564263322884, "grad_norm": 3.802929639816284, "learning_rate": 6.0769290191159e-06, "loss": 1.1465, "step": 22050 }, { "epoch": 1.7281347962382445, "grad_norm": 11.775915145874023, "learning_rate": 6.076627892564587e-06, "loss": 1.1547, "step": 22051 }, { "epoch": 1.7282131661442006, "grad_norm": 2.909067392349243, "learning_rate": 6.076326761918334e-06, "loss": 1.2177, "step": 22052 }, { "epoch": 1.7282915360501567, "grad_norm": 2.0426599979400635, "learning_rate": 6.076025627178287e-06, "loss": 1.2487, "step": 22053 }, { "epoch": 1.7283699059561128, "grad_norm": 2.085231065750122, "learning_rate": 6.075724488345592e-06, "loss": 1.148, "step": 22054 }, { "epoch": 1.728448275862069, "grad_norm": 2.428865671157837, "learning_rate": 6.075423345421393e-06, "loss": 1.1783, "step": 22055 }, { "epoch": 1.728526645768025, "grad_norm": 10.728094100952148, "learning_rate": 6.075122198406836e-06, "loss": 2.4206, "step": 22056 }, { "epoch": 1.7286050156739812, "grad_norm": 2.7878308296203613, "learning_rate": 6.0748210473030655e-06, "loss": 1.1931, "step": 22057 }, { "epoch": 1.7286833855799373, "grad_norm": 4.705859661102295, "learning_rate": 6.074519892111228e-06, "loss": 0.9711, "step": 22058 }, { "epoch": 1.7287617554858934, "grad_norm": 6.748517036437988, "learning_rate": 6.074218732832469e-06, "loss": 0.7744, "step": 22059 }, { "epoch": 1.7288401253918495, "grad_norm": 2.7964820861816406, "learning_rate": 6.073917569467934e-06, "loss": 1.5058, "step": 22060 }, { "epoch": 1.7289184952978056, "grad_norm": 2.754115104675293, "learning_rate": 6.073616402018768e-06, "loss": 1.6144, "step": 22061 }, { "epoch": 1.7289968652037617, "grad_norm": 2.2187278270721436, "learning_rate": 6.073315230486117e-06, "loss": 1.5394, "step": 22062 }, { "epoch": 1.7290752351097178, "grad_norm": 3.630462408065796, "learning_rate": 6.073014054871126e-06, "loss": 1.4619, "step": 22063 }, { "epoch": 1.729153605015674, "grad_norm": 2.8414011001586914, "learning_rate": 6.07271287517494e-06, "loss": 1.085, "step": 22064 }, { "epoch": 1.72923197492163, "grad_norm": 5.8126983642578125, "learning_rate": 6.072411691398707e-06, "loss": 1.4731, "step": 22065 }, { "epoch": 1.7293103448275862, "grad_norm": 6.910077095031738, "learning_rate": 6.072110503543567e-06, "loss": 1.2883, "step": 22066 }, { "epoch": 1.7293887147335423, "grad_norm": 3.385397434234619, "learning_rate": 6.071809311610672e-06, "loss": 1.6852, "step": 22067 }, { "epoch": 1.7294670846394984, "grad_norm": 3.118823528289795, "learning_rate": 6.071508115601166e-06, "loss": 1.8685, "step": 22068 }, { "epoch": 1.7295454545454545, "grad_norm": 3.2931344509124756, "learning_rate": 6.071206915516191e-06, "loss": 1.1629, "step": 22069 }, { "epoch": 1.7296238244514106, "grad_norm": 1.8802305459976196, "learning_rate": 6.070905711356898e-06, "loss": 1.2004, "step": 22070 }, { "epoch": 1.7297021943573667, "grad_norm": 3.512840986251831, "learning_rate": 6.0706045031244265e-06, "loss": 1.3636, "step": 22071 }, { "epoch": 1.729780564263323, "grad_norm": 2.2271437644958496, "learning_rate": 6.070303290819928e-06, "loss": 1.5025, "step": 22072 }, { "epoch": 1.729858934169279, "grad_norm": 3.2918200492858887, "learning_rate": 6.070002074444543e-06, "loss": 1.5953, "step": 22073 }, { "epoch": 1.7299373040752353, "grad_norm": 1.9517030715942383, "learning_rate": 6.069700853999422e-06, "loss": 1.2583, "step": 22074 }, { "epoch": 1.7300156739811912, "grad_norm": 3.5082948207855225, "learning_rate": 6.0693996294857074e-06, "loss": 0.9824, "step": 22075 }, { "epoch": 1.7300940438871475, "grad_norm": 2.7559401988983154, "learning_rate": 6.069098400904545e-06, "loss": 1.489, "step": 22076 }, { "epoch": 1.7301724137931034, "grad_norm": 1.7779558897018433, "learning_rate": 6.068797168257083e-06, "loss": 1.1475, "step": 22077 }, { "epoch": 1.7302507836990597, "grad_norm": 3.0032145977020264, "learning_rate": 6.0684959315444645e-06, "loss": 1.177, "step": 22078 }, { "epoch": 1.7303291536050156, "grad_norm": 4.001339912414551, "learning_rate": 6.068194690767837e-06, "loss": 1.0017, "step": 22079 }, { "epoch": 1.730407523510972, "grad_norm": 2.5783097743988037, "learning_rate": 6.0678934459283445e-06, "loss": 1.1746, "step": 22080 }, { "epoch": 1.7304858934169278, "grad_norm": 2.7567126750946045, "learning_rate": 6.067592197027134e-06, "loss": 1.3459, "step": 22081 }, { "epoch": 1.7305642633228842, "grad_norm": 3.2671308517456055, "learning_rate": 6.067290944065354e-06, "loss": 1.2358, "step": 22082 }, { "epoch": 1.73064263322884, "grad_norm": 2.3299169540405273, "learning_rate": 6.066989687044145e-06, "loss": 0.7912, "step": 22083 }, { "epoch": 1.7307210031347964, "grad_norm": 6.093560218811035, "learning_rate": 6.066688425964654e-06, "loss": 0.8761, "step": 22084 }, { "epoch": 1.7307993730407523, "grad_norm": 2.626537561416626, "learning_rate": 6.06638716082803e-06, "loss": 1.1209, "step": 22085 }, { "epoch": 1.7308777429467086, "grad_norm": 2.760287046432495, "learning_rate": 6.066085891635416e-06, "loss": 1.3116, "step": 22086 }, { "epoch": 1.7309561128526645, "grad_norm": 2.311403274536133, "learning_rate": 6.06578461838796e-06, "loss": 1.3335, "step": 22087 }, { "epoch": 1.7310344827586208, "grad_norm": 3.0068345069885254, "learning_rate": 6.065483341086805e-06, "loss": 1.0927, "step": 22088 }, { "epoch": 1.7311128526645767, "grad_norm": 2.1610195636749268, "learning_rate": 6.065182059733101e-06, "loss": 1.2908, "step": 22089 }, { "epoch": 1.731191222570533, "grad_norm": 2.456005811691284, "learning_rate": 6.064880774327989e-06, "loss": 1.3849, "step": 22090 }, { "epoch": 1.731269592476489, "grad_norm": 3.800266742706299, "learning_rate": 6.06457948487262e-06, "loss": 1.6877, "step": 22091 }, { "epoch": 1.7313479623824453, "grad_norm": 3.5842015743255615, "learning_rate": 6.064278191368135e-06, "loss": 1.0361, "step": 22092 }, { "epoch": 1.7314263322884011, "grad_norm": 3.46284556388855, "learning_rate": 6.063976893815684e-06, "loss": 1.4577, "step": 22093 }, { "epoch": 1.7315047021943575, "grad_norm": 2.92044997215271, "learning_rate": 6.0636755922164115e-06, "loss": 1.264, "step": 22094 }, { "epoch": 1.7315830721003134, "grad_norm": 3.837916135787964, "learning_rate": 6.063374286571463e-06, "loss": 1.7357, "step": 22095 }, { "epoch": 1.7316614420062697, "grad_norm": 4.460016250610352, "learning_rate": 6.063072976881984e-06, "loss": 1.4867, "step": 22096 }, { "epoch": 1.7317398119122256, "grad_norm": 3.6255502700805664, "learning_rate": 6.062771663149122e-06, "loss": 1.6866, "step": 22097 }, { "epoch": 1.731818181818182, "grad_norm": 4.036574363708496, "learning_rate": 6.062470345374024e-06, "loss": 1.1319, "step": 22098 }, { "epoch": 1.7318965517241378, "grad_norm": 2.277956247329712, "learning_rate": 6.062169023557832e-06, "loss": 1.1078, "step": 22099 }, { "epoch": 1.7319749216300941, "grad_norm": 4.455769062042236, "learning_rate": 6.061867697701697e-06, "loss": 1.1921, "step": 22100 }, { "epoch": 1.73205329153605, "grad_norm": 3.6761093139648438, "learning_rate": 6.061566367806762e-06, "loss": 1.4559, "step": 22101 }, { "epoch": 1.7321316614420064, "grad_norm": 7.3331990242004395, "learning_rate": 6.0612650338741715e-06, "loss": 1.4216, "step": 22102 }, { "epoch": 1.7322100313479623, "grad_norm": 3.398773431777954, "learning_rate": 6.060963695905077e-06, "loss": 1.3012, "step": 22103 }, { "epoch": 1.7322884012539186, "grad_norm": 2.566925048828125, "learning_rate": 6.060662353900619e-06, "loss": 0.7985, "step": 22104 }, { "epoch": 1.7323667711598745, "grad_norm": 3.112854242324829, "learning_rate": 6.060361007861948e-06, "loss": 1.4038, "step": 22105 }, { "epoch": 1.7324451410658308, "grad_norm": 2.2575714588165283, "learning_rate": 6.0600596577902075e-06, "loss": 1.6611, "step": 22106 }, { "epoch": 1.7325235109717867, "grad_norm": 1.7449285984039307, "learning_rate": 6.059758303686545e-06, "loss": 0.9798, "step": 22107 }, { "epoch": 1.732601880877743, "grad_norm": 4.042872428894043, "learning_rate": 6.0594569455521056e-06, "loss": 1.4797, "step": 22108 }, { "epoch": 1.732680250783699, "grad_norm": 4.655035018920898, "learning_rate": 6.059155583388037e-06, "loss": 1.4385, "step": 22109 }, { "epoch": 1.7327586206896552, "grad_norm": 2.5328967571258545, "learning_rate": 6.058854217195484e-06, "loss": 1.4769, "step": 22110 }, { "epoch": 1.7328369905956111, "grad_norm": 9.18782901763916, "learning_rate": 6.058552846975594e-06, "loss": 0.534, "step": 22111 }, { "epoch": 1.7329153605015675, "grad_norm": 5.101609706878662, "learning_rate": 6.058251472729512e-06, "loss": 1.3437, "step": 22112 }, { "epoch": 1.7329937304075234, "grad_norm": 2.3744657039642334, "learning_rate": 6.0579500944583845e-06, "loss": 1.1966, "step": 22113 }, { "epoch": 1.7330721003134797, "grad_norm": 3.22015380859375, "learning_rate": 6.0576487121633595e-06, "loss": 1.4683, "step": 22114 }, { "epoch": 1.7331504702194356, "grad_norm": 2.392787456512451, "learning_rate": 6.057347325845582e-06, "loss": 0.819, "step": 22115 }, { "epoch": 1.733228840125392, "grad_norm": 5.397763729095459, "learning_rate": 6.057045935506197e-06, "loss": 1.674, "step": 22116 }, { "epoch": 1.7333072100313478, "grad_norm": 2.988675117492676, "learning_rate": 6.056744541146354e-06, "loss": 1.0188, "step": 22117 }, { "epoch": 1.7333855799373041, "grad_norm": 3.2420706748962402, "learning_rate": 6.056443142767195e-06, "loss": 1.6573, "step": 22118 }, { "epoch": 1.7334639498432602, "grad_norm": 7.068515300750732, "learning_rate": 6.056141740369872e-06, "loss": 1.558, "step": 22119 }, { "epoch": 1.7335423197492164, "grad_norm": 3.210388660430908, "learning_rate": 6.055840333955526e-06, "loss": 1.5641, "step": 22120 }, { "epoch": 1.7336206896551725, "grad_norm": 3.4627926349639893, "learning_rate": 6.055538923525307e-06, "loss": 1.2818, "step": 22121 }, { "epoch": 1.7336990595611286, "grad_norm": 2.9010841846466064, "learning_rate": 6.055237509080359e-06, "loss": 1.532, "step": 22122 }, { "epoch": 1.7337774294670847, "grad_norm": 2.0438039302825928, "learning_rate": 6.05493609062183e-06, "loss": 1.4117, "step": 22123 }, { "epoch": 1.7338557993730408, "grad_norm": 2.4420437812805176, "learning_rate": 6.054634668150866e-06, "loss": 1.5906, "step": 22124 }, { "epoch": 1.733934169278997, "grad_norm": 2.789433240890503, "learning_rate": 6.054333241668614e-06, "loss": 0.9176, "step": 22125 }, { "epoch": 1.734012539184953, "grad_norm": 2.750281810760498, "learning_rate": 6.0540318111762196e-06, "loss": 1.608, "step": 22126 }, { "epoch": 1.7340909090909091, "grad_norm": 2.74453067779541, "learning_rate": 6.053730376674829e-06, "loss": 1.5093, "step": 22127 }, { "epoch": 1.7341692789968652, "grad_norm": 2.2318851947784424, "learning_rate": 6.053428938165591e-06, "loss": 1.2287, "step": 22128 }, { "epoch": 1.7342476489028213, "grad_norm": 4.689579010009766, "learning_rate": 6.053127495649649e-06, "loss": 1.0402, "step": 22129 }, { "epoch": 1.7343260188087775, "grad_norm": 2.9114389419555664, "learning_rate": 6.05282604912815e-06, "loss": 1.5132, "step": 22130 }, { "epoch": 1.7344043887147336, "grad_norm": 2.682263135910034, "learning_rate": 6.052524598602243e-06, "loss": 1.4116, "step": 22131 }, { "epoch": 1.7344827586206897, "grad_norm": 4.392446994781494, "learning_rate": 6.052223144073071e-06, "loss": 1.5143, "step": 22132 }, { "epoch": 1.7345611285266458, "grad_norm": 2.568732261657715, "learning_rate": 6.051921685541785e-06, "loss": 1.061, "step": 22133 }, { "epoch": 1.734639498432602, "grad_norm": 4.464893817901611, "learning_rate": 6.051620223009527e-06, "loss": 1.3967, "step": 22134 }, { "epoch": 1.734717868338558, "grad_norm": 6.5840935707092285, "learning_rate": 6.051318756477448e-06, "loss": 1.3844, "step": 22135 }, { "epoch": 1.7347962382445141, "grad_norm": 6.935147285461426, "learning_rate": 6.051017285946692e-06, "loss": 1.9836, "step": 22136 }, { "epoch": 1.7348746081504702, "grad_norm": 3.6474764347076416, "learning_rate": 6.050715811418405e-06, "loss": 0.8265, "step": 22137 }, { "epoch": 1.7349529780564263, "grad_norm": 2.4461123943328857, "learning_rate": 6.050414332893736e-06, "loss": 1.6346, "step": 22138 }, { "epoch": 1.7350313479623825, "grad_norm": 2.425081491470337, "learning_rate": 6.050112850373829e-06, "loss": 1.1035, "step": 22139 }, { "epoch": 1.7351097178683386, "grad_norm": 6.034768104553223, "learning_rate": 6.049811363859833e-06, "loss": 1.3475, "step": 22140 }, { "epoch": 1.7351880877742947, "grad_norm": 3.9154293537139893, "learning_rate": 6.049509873352894e-06, "loss": 1.5443, "step": 22141 }, { "epoch": 1.7352664576802508, "grad_norm": 4.0477495193481445, "learning_rate": 6.0492083788541575e-06, "loss": 1.5045, "step": 22142 }, { "epoch": 1.735344827586207, "grad_norm": 7.605324745178223, "learning_rate": 6.048906880364773e-06, "loss": 0.5089, "step": 22143 }, { "epoch": 1.735423197492163, "grad_norm": 1.9012455940246582, "learning_rate": 6.048605377885883e-06, "loss": 0.933, "step": 22144 }, { "epoch": 1.7355015673981191, "grad_norm": 2.9729833602905273, "learning_rate": 6.04830387141864e-06, "loss": 1.4325, "step": 22145 }, { "epoch": 1.7355799373040752, "grad_norm": 4.59864616394043, "learning_rate": 6.048002360964184e-06, "loss": 1.9313, "step": 22146 }, { "epoch": 1.7356583072100313, "grad_norm": 4.6579766273498535, "learning_rate": 6.0477008465236676e-06, "loss": 1.2578, "step": 22147 }, { "epoch": 1.7357366771159874, "grad_norm": 3.658470392227173, "learning_rate": 6.047399328098235e-06, "loss": 1.6287, "step": 22148 }, { "epoch": 1.7358150470219436, "grad_norm": 1.96504807472229, "learning_rate": 6.047097805689033e-06, "loss": 1.1721, "step": 22149 }, { "epoch": 1.7358934169278997, "grad_norm": 2.9016244411468506, "learning_rate": 6.046796279297208e-06, "loss": 1.5767, "step": 22150 }, { "epoch": 1.7359717868338558, "grad_norm": 2.3028736114501953, "learning_rate": 6.046494748923909e-06, "loss": 1.2164, "step": 22151 }, { "epoch": 1.736050156739812, "grad_norm": 2.413510322570801, "learning_rate": 6.046193214570281e-06, "loss": 1.2247, "step": 22152 }, { "epoch": 1.736128526645768, "grad_norm": 3.151238203048706, "learning_rate": 6.045891676237471e-06, "loss": 1.5661, "step": 22153 }, { "epoch": 1.736206896551724, "grad_norm": 4.473540306091309, "learning_rate": 6.045590133926627e-06, "loss": 1.9073, "step": 22154 }, { "epoch": 1.7362852664576802, "grad_norm": 2.487889051437378, "learning_rate": 6.045288587638895e-06, "loss": 1.026, "step": 22155 }, { "epoch": 1.7363636363636363, "grad_norm": 5.902001857757568, "learning_rate": 6.044987037375422e-06, "loss": 2.2778, "step": 22156 }, { "epoch": 1.7364420062695924, "grad_norm": 2.2563343048095703, "learning_rate": 6.044685483137354e-06, "loss": 1.1205, "step": 22157 }, { "epoch": 1.7365203761755486, "grad_norm": 3.011624336242676, "learning_rate": 6.04438392492584e-06, "loss": 1.178, "step": 22158 }, { "epoch": 1.7365987460815047, "grad_norm": 4.1367716789245605, "learning_rate": 6.044082362742025e-06, "loss": 1.4969, "step": 22159 }, { "epoch": 1.7366771159874608, "grad_norm": 23.56886863708496, "learning_rate": 6.043780796587057e-06, "loss": 1.3049, "step": 22160 }, { "epoch": 1.7367554858934169, "grad_norm": 8.150520324707031, "learning_rate": 6.043479226462085e-06, "loss": 1.3723, "step": 22161 }, { "epoch": 1.736833855799373, "grad_norm": 3.3581786155700684, "learning_rate": 6.043177652368252e-06, "loss": 1.2689, "step": 22162 }, { "epoch": 1.736912225705329, "grad_norm": 2.0586273670196533, "learning_rate": 6.042876074306707e-06, "loss": 1.4622, "step": 22163 }, { "epoch": 1.7369905956112852, "grad_norm": 2.0563056468963623, "learning_rate": 6.042574492278599e-06, "loss": 1.0847, "step": 22164 }, { "epoch": 1.7370689655172413, "grad_norm": 4.661405086517334, "learning_rate": 6.042272906285072e-06, "loss": 1.8241, "step": 22165 }, { "epoch": 1.7371473354231974, "grad_norm": 2.9055755138397217, "learning_rate": 6.0419713163272745e-06, "loss": 1.2489, "step": 22166 }, { "epoch": 1.7372257053291538, "grad_norm": 2.937228202819824, "learning_rate": 6.041669722406352e-06, "loss": 1.3447, "step": 22167 }, { "epoch": 1.7373040752351097, "grad_norm": 1.9973191022872925, "learning_rate": 6.041368124523455e-06, "loss": 1.1593, "step": 22168 }, { "epoch": 1.737382445141066, "grad_norm": 2.8147988319396973, "learning_rate": 6.041066522679727e-06, "loss": 1.7, "step": 22169 }, { "epoch": 1.7374608150470219, "grad_norm": 3.3162143230438232, "learning_rate": 6.0407649168763175e-06, "loss": 1.5522, "step": 22170 }, { "epoch": 1.7375391849529782, "grad_norm": 3.3989250659942627, "learning_rate": 6.040463307114372e-06, "loss": 1.121, "step": 22171 }, { "epoch": 1.737617554858934, "grad_norm": 4.238213539123535, "learning_rate": 6.040161693395038e-06, "loss": 1.7253, "step": 22172 }, { "epoch": 1.7376959247648904, "grad_norm": 3.4615893363952637, "learning_rate": 6.039860075719466e-06, "loss": 1.2152, "step": 22173 }, { "epoch": 1.7377742946708463, "grad_norm": 5.969131946563721, "learning_rate": 6.039558454088796e-06, "loss": 1.4207, "step": 22174 }, { "epoch": 1.7378526645768027, "grad_norm": 2.537229537963867, "learning_rate": 6.039256828504184e-06, "loss": 1.4216, "step": 22175 }, { "epoch": 1.7379310344827585, "grad_norm": 2.22495174407959, "learning_rate": 6.038955198966771e-06, "loss": 1.2047, "step": 22176 }, { "epoch": 1.7380094043887149, "grad_norm": 2.174360513687134, "learning_rate": 6.038653565477706e-06, "loss": 1.636, "step": 22177 }, { "epoch": 1.7380877742946708, "grad_norm": 2.086987018585205, "learning_rate": 6.038351928038136e-06, "loss": 1.2304, "step": 22178 }, { "epoch": 1.738166144200627, "grad_norm": 3.3680102825164795, "learning_rate": 6.0380502866492085e-06, "loss": 1.272, "step": 22179 }, { "epoch": 1.738244514106583, "grad_norm": 2.572178363800049, "learning_rate": 6.037748641312071e-06, "loss": 1.1793, "step": 22180 }, { "epoch": 1.7383228840125393, "grad_norm": 2.4959287643432617, "learning_rate": 6.037446992027871e-06, "loss": 1.5884, "step": 22181 }, { "epoch": 1.7384012539184952, "grad_norm": 3.628795862197876, "learning_rate": 6.037145338797756e-06, "loss": 1.3967, "step": 22182 }, { "epoch": 1.7384796238244515, "grad_norm": 2.6431853771209717, "learning_rate": 6.036843681622872e-06, "loss": 1.44, "step": 22183 }, { "epoch": 1.7385579937304074, "grad_norm": 8.834220886230469, "learning_rate": 6.0365420205043676e-06, "loss": 1.416, "step": 22184 }, { "epoch": 1.7386363636363638, "grad_norm": 3.795788288116455, "learning_rate": 6.03624035544339e-06, "loss": 1.2059, "step": 22185 }, { "epoch": 1.7387147335423196, "grad_norm": 1.7001056671142578, "learning_rate": 6.035938686441085e-06, "loss": 1.3186, "step": 22186 }, { "epoch": 1.738793103448276, "grad_norm": 2.7573812007904053, "learning_rate": 6.035637013498603e-06, "loss": 1.2225, "step": 22187 }, { "epoch": 1.7388714733542319, "grad_norm": 5.486629962921143, "learning_rate": 6.035335336617089e-06, "loss": 1.7258, "step": 22188 }, { "epoch": 1.7389498432601882, "grad_norm": 3.2713255882263184, "learning_rate": 6.03503365579769e-06, "loss": 1.3569, "step": 22189 }, { "epoch": 1.739028213166144, "grad_norm": 3.0891480445861816, "learning_rate": 6.034731971041557e-06, "loss": 1.4761, "step": 22190 }, { "epoch": 1.7391065830721004, "grad_norm": 4.980393409729004, "learning_rate": 6.034430282349833e-06, "loss": 0.9761, "step": 22191 }, { "epoch": 1.7391849529780563, "grad_norm": 4.096583843231201, "learning_rate": 6.034128589723669e-06, "loss": 1.6201, "step": 22192 }, { "epoch": 1.7392633228840126, "grad_norm": 4.655981063842773, "learning_rate": 6.03382689316421e-06, "loss": 1.3138, "step": 22193 }, { "epoch": 1.7393416927899685, "grad_norm": 3.615353584289551, "learning_rate": 6.033525192672606e-06, "loss": 1.264, "step": 22194 }, { "epoch": 1.7394200626959249, "grad_norm": 2.580279588699341, "learning_rate": 6.033223488250001e-06, "loss": 1.2394, "step": 22195 }, { "epoch": 1.7394984326018808, "grad_norm": 2.28005313873291, "learning_rate": 6.0329217798975455e-06, "loss": 1.3476, "step": 22196 }, { "epoch": 1.739576802507837, "grad_norm": 3.4115359783172607, "learning_rate": 6.032620067616387e-06, "loss": 1.0393, "step": 22197 }, { "epoch": 1.739655172413793, "grad_norm": 5.4822611808776855, "learning_rate": 6.03231835140767e-06, "loss": 1.1782, "step": 22198 }, { "epoch": 1.7397335423197493, "grad_norm": 4.644200325012207, "learning_rate": 6.032016631272546e-06, "loss": 1.9891, "step": 22199 }, { "epoch": 1.7398119122257052, "grad_norm": 5.45955228805542, "learning_rate": 6.031714907212159e-06, "loss": 1.7936, "step": 22200 }, { "epoch": 1.7398902821316615, "grad_norm": 5.2244873046875, "learning_rate": 6.031413179227661e-06, "loss": 0.9567, "step": 22201 }, { "epoch": 1.7399686520376174, "grad_norm": 1.9848004579544067, "learning_rate": 6.031111447320196e-06, "loss": 1.3515, "step": 22202 }, { "epoch": 1.7400470219435737, "grad_norm": 4.2280354499816895, "learning_rate": 6.030809711490912e-06, "loss": 1.3922, "step": 22203 }, { "epoch": 1.7401253918495296, "grad_norm": 4.172384262084961, "learning_rate": 6.030507971740958e-06, "loss": 1.5039, "step": 22204 }, { "epoch": 1.740203761755486, "grad_norm": 2.7678961753845215, "learning_rate": 6.03020622807148e-06, "loss": 1.5186, "step": 22205 }, { "epoch": 1.7402821316614419, "grad_norm": 5.413852691650391, "learning_rate": 6.029904480483627e-06, "loss": 2.2388, "step": 22206 }, { "epoch": 1.7403605015673982, "grad_norm": 3.1235506534576416, "learning_rate": 6.029602728978546e-06, "loss": 1.3933, "step": 22207 }, { "epoch": 1.740438871473354, "grad_norm": 4.4658050537109375, "learning_rate": 6.029300973557386e-06, "loss": 1.4865, "step": 22208 }, { "epoch": 1.7405172413793104, "grad_norm": 2.5356862545013428, "learning_rate": 6.028999214221294e-06, "loss": 1.3746, "step": 22209 }, { "epoch": 1.7405956112852663, "grad_norm": 5.579040050506592, "learning_rate": 6.028697450971417e-06, "loss": 1.0113, "step": 22210 }, { "epoch": 1.7406739811912226, "grad_norm": 3.7862720489501953, "learning_rate": 6.028395683808903e-06, "loss": 1.4557, "step": 22211 }, { "epoch": 1.7407523510971785, "grad_norm": 3.06774640083313, "learning_rate": 6.0280939127349e-06, "loss": 1.444, "step": 22212 }, { "epoch": 1.7408307210031349, "grad_norm": 3.6159257888793945, "learning_rate": 6.027792137750557e-06, "loss": 1.3791, "step": 22213 }, { "epoch": 1.740909090909091, "grad_norm": 2.296285390853882, "learning_rate": 6.02749035885702e-06, "loss": 1.0165, "step": 22214 }, { "epoch": 1.740987460815047, "grad_norm": 16.67243766784668, "learning_rate": 6.027188576055436e-06, "loss": 1.1705, "step": 22215 }, { "epoch": 1.7410658307210032, "grad_norm": 3.072267532348633, "learning_rate": 6.026886789346957e-06, "loss": 1.3488, "step": 22216 }, { "epoch": 1.7411442006269593, "grad_norm": 3.6864755153656006, "learning_rate": 6.026584998732725e-06, "loss": 1.1576, "step": 22217 }, { "epoch": 1.7412225705329154, "grad_norm": 2.3945746421813965, "learning_rate": 6.026283204213894e-06, "loss": 1.2323, "step": 22218 }, { "epoch": 1.7413009404388715, "grad_norm": 1.8503062725067139, "learning_rate": 6.025981405791605e-06, "loss": 1.3216, "step": 22219 }, { "epoch": 1.7413793103448276, "grad_norm": 2.5262668132781982, "learning_rate": 6.025679603467014e-06, "loss": 0.8225, "step": 22220 }, { "epoch": 1.7414576802507837, "grad_norm": 9.464040756225586, "learning_rate": 6.025377797241261e-06, "loss": 1.4989, "step": 22221 }, { "epoch": 1.7415360501567398, "grad_norm": 2.5385513305664062, "learning_rate": 6.025075987115501e-06, "loss": 1.4007, "step": 22222 }, { "epoch": 1.741614420062696, "grad_norm": 4.14299201965332, "learning_rate": 6.024774173090876e-06, "loss": 1.8706, "step": 22223 }, { "epoch": 1.741692789968652, "grad_norm": 2.4916324615478516, "learning_rate": 6.024472355168537e-06, "loss": 1.5469, "step": 22224 }, { "epoch": 1.7417711598746082, "grad_norm": 3.999039649963379, "learning_rate": 6.024170533349631e-06, "loss": 1.3348, "step": 22225 }, { "epoch": 1.7418495297805643, "grad_norm": 3.830770969390869, "learning_rate": 6.023868707635306e-06, "loss": 1.7872, "step": 22226 }, { "epoch": 1.7419278996865204, "grad_norm": 2.5680651664733887, "learning_rate": 6.023566878026712e-06, "loss": 1.3105, "step": 22227 }, { "epoch": 1.7420062695924765, "grad_norm": 2.0592594146728516, "learning_rate": 6.023265044524993e-06, "loss": 1.4099, "step": 22228 }, { "epoch": 1.7420846394984326, "grad_norm": 3.346726655960083, "learning_rate": 6.022963207131303e-06, "loss": 1.1617, "step": 22229 }, { "epoch": 1.7421630094043887, "grad_norm": 2.573143243789673, "learning_rate": 6.022661365846783e-06, "loss": 1.3563, "step": 22230 }, { "epoch": 1.7422413793103448, "grad_norm": 3.297447681427002, "learning_rate": 6.022359520672586e-06, "loss": 1.1162, "step": 22231 }, { "epoch": 1.742319749216301, "grad_norm": 3.3379335403442383, "learning_rate": 6.0220576716098576e-06, "loss": 1.433, "step": 22232 }, { "epoch": 1.742398119122257, "grad_norm": 3.8800859451293945, "learning_rate": 6.021755818659747e-06, "loss": 1.5704, "step": 22233 }, { "epoch": 1.7424764890282132, "grad_norm": 4.327786922454834, "learning_rate": 6.0214539618234025e-06, "loss": 1.0387, "step": 22234 }, { "epoch": 1.7425548589341693, "grad_norm": 2.9914464950561523, "learning_rate": 6.021152101101972e-06, "loss": 1.4654, "step": 22235 }, { "epoch": 1.7426332288401254, "grad_norm": 5.38045072555542, "learning_rate": 6.020850236496602e-06, "loss": 0.8804, "step": 22236 }, { "epoch": 1.7427115987460815, "grad_norm": 3.9595069885253906, "learning_rate": 6.020548368008444e-06, "loss": 1.2662, "step": 22237 }, { "epoch": 1.7427899686520376, "grad_norm": 3.608133316040039, "learning_rate": 6.020246495638643e-06, "loss": 1.7517, "step": 22238 }, { "epoch": 1.7428683385579937, "grad_norm": 3.2345595359802246, "learning_rate": 6.0199446193883484e-06, "loss": 1.3945, "step": 22239 }, { "epoch": 1.7429467084639498, "grad_norm": 3.213015079498291, "learning_rate": 6.0196427392587085e-06, "loss": 1.0868, "step": 22240 }, { "epoch": 1.743025078369906, "grad_norm": 7.9553351402282715, "learning_rate": 6.0193408552508714e-06, "loss": 1.4665, "step": 22241 }, { "epoch": 1.743103448275862, "grad_norm": 2.3823742866516113, "learning_rate": 6.019038967365984e-06, "loss": 1.3171, "step": 22242 }, { "epoch": 1.7431818181818182, "grad_norm": 4.087495803833008, "learning_rate": 6.018737075605198e-06, "loss": 1.3517, "step": 22243 }, { "epoch": 1.7432601880877743, "grad_norm": 3.0068130493164062, "learning_rate": 6.018435179969659e-06, "loss": 1.3039, "step": 22244 }, { "epoch": 1.7433385579937304, "grad_norm": 1.919818639755249, "learning_rate": 6.0181332804605145e-06, "loss": 1.4891, "step": 22245 }, { "epoch": 1.7434169278996865, "grad_norm": 2.2652597427368164, "learning_rate": 6.0178313770789155e-06, "loss": 1.5689, "step": 22246 }, { "epoch": 1.7434952978056426, "grad_norm": 3.2556567192077637, "learning_rate": 6.017529469826007e-06, "loss": 1.1758, "step": 22247 }, { "epoch": 1.7435736677115987, "grad_norm": 3.1667721271514893, "learning_rate": 6.017227558702941e-06, "loss": 1.4227, "step": 22248 }, { "epoch": 1.7436520376175548, "grad_norm": 2.6236517429351807, "learning_rate": 6.016925643710861e-06, "loss": 1.3131, "step": 22249 }, { "epoch": 1.743730407523511, "grad_norm": 2.5206992626190186, "learning_rate": 6.016623724850921e-06, "loss": 1.4609, "step": 22250 }, { "epoch": 1.743808777429467, "grad_norm": 4.597697734832764, "learning_rate": 6.016321802124265e-06, "loss": 0.9437, "step": 22251 }, { "epoch": 1.7438871473354232, "grad_norm": 3.1688385009765625, "learning_rate": 6.0160198755320424e-06, "loss": 1.7204, "step": 22252 }, { "epoch": 1.7439655172413793, "grad_norm": 4.032832145690918, "learning_rate": 6.015717945075405e-06, "loss": 1.3995, "step": 22253 }, { "epoch": 1.7440438871473354, "grad_norm": 2.656284809112549, "learning_rate": 6.0154160107554936e-06, "loss": 1.6509, "step": 22254 }, { "epoch": 1.7441222570532915, "grad_norm": 2.1598563194274902, "learning_rate": 6.015114072573465e-06, "loss": 0.9887, "step": 22255 }, { "epoch": 1.7442006269592476, "grad_norm": 3.190885066986084, "learning_rate": 6.014812130530463e-06, "loss": 1.1759, "step": 22256 }, { "epoch": 1.7442789968652037, "grad_norm": 2.851447105407715, "learning_rate": 6.014510184627636e-06, "loss": 1.3915, "step": 22257 }, { "epoch": 1.7443573667711598, "grad_norm": 3.0494260787963867, "learning_rate": 6.0142082348661335e-06, "loss": 1.6713, "step": 22258 }, { "epoch": 1.744435736677116, "grad_norm": 3.420762777328491, "learning_rate": 6.013906281247104e-06, "loss": 1.2761, "step": 22259 }, { "epoch": 1.7445141065830723, "grad_norm": 2.9123454093933105, "learning_rate": 6.013604323771695e-06, "loss": 1.2567, "step": 22260 }, { "epoch": 1.7445924764890282, "grad_norm": 3.148327350616455, "learning_rate": 6.013302362441058e-06, "loss": 1.4007, "step": 22261 }, { "epoch": 1.7446708463949845, "grad_norm": 2.726766347885132, "learning_rate": 6.013000397256338e-06, "loss": 0.9679, "step": 22262 }, { "epoch": 1.7447492163009404, "grad_norm": 3.166313409805298, "learning_rate": 6.012698428218684e-06, "loss": 1.2401, "step": 22263 }, { "epoch": 1.7448275862068967, "grad_norm": 2.94425892829895, "learning_rate": 6.012396455329246e-06, "loss": 1.3495, "step": 22264 }, { "epoch": 1.7449059561128526, "grad_norm": 2.887382984161377, "learning_rate": 6.012094478589173e-06, "loss": 1.1881, "step": 22265 }, { "epoch": 1.744984326018809, "grad_norm": 3.0649795532226562, "learning_rate": 6.0117924979996115e-06, "loss": 1.2015, "step": 22266 }, { "epoch": 1.7450626959247648, "grad_norm": 2.2034265995025635, "learning_rate": 6.0114905135617105e-06, "loss": 1.1902, "step": 22267 }, { "epoch": 1.7451410658307211, "grad_norm": 2.1864821910858154, "learning_rate": 6.01118852527662e-06, "loss": 1.3992, "step": 22268 }, { "epoch": 1.745219435736677, "grad_norm": 7.179172992706299, "learning_rate": 6.010886533145487e-06, "loss": 1.5108, "step": 22269 }, { "epoch": 1.7452978056426334, "grad_norm": 4.09971809387207, "learning_rate": 6.0105845371694615e-06, "loss": 1.0974, "step": 22270 }, { "epoch": 1.7453761755485893, "grad_norm": 2.404404640197754, "learning_rate": 6.010282537349691e-06, "loss": 1.4318, "step": 22271 }, { "epoch": 1.7454545454545456, "grad_norm": 4.136593341827393, "learning_rate": 6.009980533687325e-06, "loss": 1.2028, "step": 22272 }, { "epoch": 1.7455329153605015, "grad_norm": 2.520453691482544, "learning_rate": 6.009678526183511e-06, "loss": 1.2723, "step": 22273 }, { "epoch": 1.7456112852664578, "grad_norm": 2.8550713062286377, "learning_rate": 6.0093765148394e-06, "loss": 1.1938, "step": 22274 }, { "epoch": 1.7456896551724137, "grad_norm": 2.7573914527893066, "learning_rate": 6.009074499656138e-06, "loss": 1.1057, "step": 22275 }, { "epoch": 1.74576802507837, "grad_norm": 3.197819948196411, "learning_rate": 6.008772480634876e-06, "loss": 1.0137, "step": 22276 }, { "epoch": 1.745846394984326, "grad_norm": 3.409107208251953, "learning_rate": 6.008470457776761e-06, "loss": 1.2882, "step": 22277 }, { "epoch": 1.7459247648902823, "grad_norm": 2.224205493927002, "learning_rate": 6.008168431082941e-06, "loss": 1.2367, "step": 22278 }, { "epoch": 1.7460031347962381, "grad_norm": 3.789458990097046, "learning_rate": 6.007866400554568e-06, "loss": 1.6185, "step": 22279 }, { "epoch": 1.7460815047021945, "grad_norm": 3.9231228828430176, "learning_rate": 6.007564366192788e-06, "loss": 1.2247, "step": 22280 }, { "epoch": 1.7461598746081504, "grad_norm": 2.8784303665161133, "learning_rate": 6.007262327998752e-06, "loss": 1.2106, "step": 22281 }, { "epoch": 1.7462382445141067, "grad_norm": 2.856325387954712, "learning_rate": 6.006960285973604e-06, "loss": 1.4091, "step": 22282 }, { "epoch": 1.7463166144200626, "grad_norm": 14.446459770202637, "learning_rate": 6.0066582401185005e-06, "loss": 1.9964, "step": 22283 }, { "epoch": 1.746394984326019, "grad_norm": 3.116471529006958, "learning_rate": 6.006356190434583e-06, "loss": 1.2027, "step": 22284 }, { "epoch": 1.7464733542319748, "grad_norm": 2.2887158393859863, "learning_rate": 6.0060541369230055e-06, "loss": 1.3586, "step": 22285 }, { "epoch": 1.7465517241379311, "grad_norm": 1.9885936975479126, "learning_rate": 6.005752079584913e-06, "loss": 1.1766, "step": 22286 }, { "epoch": 1.746630094043887, "grad_norm": 3.3249683380126953, "learning_rate": 6.005450018421458e-06, "loss": 1.4973, "step": 22287 }, { "epoch": 1.7467084639498434, "grad_norm": 5.079562187194824, "learning_rate": 6.005147953433786e-06, "loss": 1.4632, "step": 22288 }, { "epoch": 1.7467868338557992, "grad_norm": 2.2597506046295166, "learning_rate": 6.004845884623048e-06, "loss": 1.1849, "step": 22289 }, { "epoch": 1.7468652037617556, "grad_norm": 2.23176908493042, "learning_rate": 6.004543811990393e-06, "loss": 1.3827, "step": 22290 }, { "epoch": 1.7469435736677115, "grad_norm": 5.221247673034668, "learning_rate": 6.00424173553697e-06, "loss": 1.2023, "step": 22291 }, { "epoch": 1.7470219435736678, "grad_norm": 3.7595441341400146, "learning_rate": 6.003939655263925e-06, "loss": 0.9642, "step": 22292 }, { "epoch": 1.7471003134796237, "grad_norm": 2.2141175270080566, "learning_rate": 6.003637571172411e-06, "loss": 1.5208, "step": 22293 }, { "epoch": 1.74717868338558, "grad_norm": 1.8847358226776123, "learning_rate": 6.0033354832635745e-06, "loss": 1.0196, "step": 22294 }, { "epoch": 1.747257053291536, "grad_norm": 5.096343040466309, "learning_rate": 6.003033391538565e-06, "loss": 1.691, "step": 22295 }, { "epoch": 1.7473354231974922, "grad_norm": 2.0681538581848145, "learning_rate": 6.002731295998531e-06, "loss": 0.7796, "step": 22296 }, { "epoch": 1.7474137931034481, "grad_norm": 3.666400194168091, "learning_rate": 6.002429196644623e-06, "loss": 1.5026, "step": 22297 }, { "epoch": 1.7474921630094045, "grad_norm": 5.131645679473877, "learning_rate": 6.0021270934779904e-06, "loss": 0.9216, "step": 22298 }, { "epoch": 1.7475705329153604, "grad_norm": 5.124171257019043, "learning_rate": 6.001824986499779e-06, "loss": 1.3496, "step": 22299 }, { "epoch": 1.7476489028213167, "grad_norm": 3.5070717334747314, "learning_rate": 6.001522875711142e-06, "loss": 1.2129, "step": 22300 }, { "epoch": 1.7477272727272726, "grad_norm": 3.5063462257385254, "learning_rate": 6.001220761113223e-06, "loss": 1.1099, "step": 22301 }, { "epoch": 1.747805642633229, "grad_norm": 2.437185287475586, "learning_rate": 6.0009186427071786e-06, "loss": 1.528, "step": 22302 }, { "epoch": 1.7478840125391848, "grad_norm": 3.0655782222747803, "learning_rate": 6.000616520494151e-06, "loss": 1.4454, "step": 22303 }, { "epoch": 1.7479623824451411, "grad_norm": 2.7344448566436768, "learning_rate": 6.0003143944752925e-06, "loss": 1.2779, "step": 22304 }, { "epoch": 1.748040752351097, "grad_norm": 3.293900489807129, "learning_rate": 6.0000122646517535e-06, "loss": 1.263, "step": 22305 }, { "epoch": 1.7481191222570533, "grad_norm": 2.106981039047241, "learning_rate": 5.99971013102468e-06, "loss": 1.5591, "step": 22306 }, { "epoch": 1.7481974921630095, "grad_norm": 2.6366660594940186, "learning_rate": 5.999407993595223e-06, "loss": 1.2895, "step": 22307 }, { "epoch": 1.7482758620689656, "grad_norm": 10.648036003112793, "learning_rate": 5.99910585236453e-06, "loss": 1.6376, "step": 22308 }, { "epoch": 1.7483542319749217, "grad_norm": 4.3511881828308105, "learning_rate": 5.998803707333755e-06, "loss": 1.2006, "step": 22309 }, { "epoch": 1.7484326018808778, "grad_norm": 2.802827835083008, "learning_rate": 5.998501558504041e-06, "loss": 1.1042, "step": 22310 }, { "epoch": 1.748510971786834, "grad_norm": 2.611687183380127, "learning_rate": 5.99819940587654e-06, "loss": 1.2145, "step": 22311 }, { "epoch": 1.74858934169279, "grad_norm": 2.183842420578003, "learning_rate": 5.997897249452402e-06, "loss": 1.2996, "step": 22312 }, { "epoch": 1.7486677115987461, "grad_norm": 5.160345077514648, "learning_rate": 5.997595089232775e-06, "loss": 1.9036, "step": 22313 }, { "epoch": 1.7487460815047022, "grad_norm": 1.9223742485046387, "learning_rate": 5.997292925218809e-06, "loss": 1.4325, "step": 22314 }, { "epoch": 1.7488244514106583, "grad_norm": 2.536547899246216, "learning_rate": 5.996990757411652e-06, "loss": 1.6339, "step": 22315 }, { "epoch": 1.7489028213166145, "grad_norm": 2.3549931049346924, "learning_rate": 5.996688585812455e-06, "loss": 1.5081, "step": 22316 }, { "epoch": 1.7489811912225706, "grad_norm": 4.911789417266846, "learning_rate": 5.996386410422366e-06, "loss": 1.2797, "step": 22317 }, { "epoch": 1.7490595611285267, "grad_norm": 2.3916025161743164, "learning_rate": 5.996084231242535e-06, "loss": 1.4649, "step": 22318 }, { "epoch": 1.7491379310344828, "grad_norm": 2.3176090717315674, "learning_rate": 5.995782048274112e-06, "loss": 1.4787, "step": 22319 }, { "epoch": 1.749216300940439, "grad_norm": 6.454819202423096, "learning_rate": 5.995479861518245e-06, "loss": 0.9409, "step": 22320 }, { "epoch": 1.749294670846395, "grad_norm": 2.97676420211792, "learning_rate": 5.995177670976085e-06, "loss": 1.4134, "step": 22321 }, { "epoch": 1.7493730407523511, "grad_norm": 2.733978271484375, "learning_rate": 5.9948754766487785e-06, "loss": 1.6876, "step": 22322 }, { "epoch": 1.7494514106583072, "grad_norm": 2.905089855194092, "learning_rate": 5.994573278537478e-06, "loss": 1.4838, "step": 22323 }, { "epoch": 1.7495297805642633, "grad_norm": 3.320859432220459, "learning_rate": 5.99427107664333e-06, "loss": 1.3743, "step": 22324 }, { "epoch": 1.7496081504702194, "grad_norm": 2.521437883377075, "learning_rate": 5.9939688709674875e-06, "loss": 0.888, "step": 22325 }, { "epoch": 1.7496865203761756, "grad_norm": 2.6897194385528564, "learning_rate": 5.993666661511097e-06, "loss": 1.4449, "step": 22326 }, { "epoch": 1.7497648902821317, "grad_norm": 6.650660514831543, "learning_rate": 5.9933644482753084e-06, "loss": 1.2229, "step": 22327 }, { "epoch": 1.7498432601880878, "grad_norm": 2.312988758087158, "learning_rate": 5.9930622312612726e-06, "loss": 1.1848, "step": 22328 }, { "epoch": 1.749921630094044, "grad_norm": 4.799710273742676, "learning_rate": 5.992760010470136e-06, "loss": 0.7221, "step": 22329 }, { "epoch": 1.75, "grad_norm": 3.53605580329895, "learning_rate": 5.992457785903054e-06, "loss": 0.9565, "step": 22330 }, { "epoch": 1.75, "eval_loss": 1.2739139795303345, "eval_runtime": 108.8922, "eval_samples_per_second": 24.676, "eval_steps_per_second": 6.171, "step": 22330 }, { "epoch": 1.750078369905956, "grad_norm": 2.743624687194824, "learning_rate": 5.99215555756117e-06, "loss": 1.4175, "step": 22331 }, { "epoch": 1.7501567398119122, "grad_norm": 2.816754102706909, "learning_rate": 5.991853325445636e-06, "loss": 1.4732, "step": 22332 }, { "epoch": 1.7502351097178683, "grad_norm": 3.793320417404175, "learning_rate": 5.991551089557601e-06, "loss": 0.9863, "step": 22333 }, { "epoch": 1.7503134796238244, "grad_norm": 2.005314588546753, "learning_rate": 5.991248849898216e-06, "loss": 1.2437, "step": 22334 }, { "epoch": 1.7503918495297806, "grad_norm": 2.172039031982422, "learning_rate": 5.99094660646863e-06, "loss": 0.9212, "step": 22335 }, { "epoch": 1.7504702194357367, "grad_norm": 2.262347936630249, "learning_rate": 5.9906443592699904e-06, "loss": 1.4587, "step": 22336 }, { "epoch": 1.7505485893416928, "grad_norm": 3.858375310897827, "learning_rate": 5.9903421083034495e-06, "loss": 1.5731, "step": 22337 }, { "epoch": 1.7506269592476489, "grad_norm": 1.9379764795303345, "learning_rate": 5.9900398535701555e-06, "loss": 1.2514, "step": 22338 }, { "epoch": 1.750705329153605, "grad_norm": 4.911422252655029, "learning_rate": 5.989737595071259e-06, "loss": 1.2228, "step": 22339 }, { "epoch": 1.750783699059561, "grad_norm": 2.638943672180176, "learning_rate": 5.989435332807908e-06, "loss": 1.1173, "step": 22340 }, { "epoch": 1.7508620689655172, "grad_norm": 2.386218786239624, "learning_rate": 5.989133066781254e-06, "loss": 1.2558, "step": 22341 }, { "epoch": 1.7509404388714733, "grad_norm": 2.3574378490448, "learning_rate": 5.988830796992445e-06, "loss": 1.266, "step": 22342 }, { "epoch": 1.7510188087774294, "grad_norm": 5.324767112731934, "learning_rate": 5.988528523442631e-06, "loss": 1.3564, "step": 22343 }, { "epoch": 1.7510971786833855, "grad_norm": 2.8547463417053223, "learning_rate": 5.988226246132964e-06, "loss": 1.6281, "step": 22344 }, { "epoch": 1.7511755485893417, "grad_norm": 9.576192855834961, "learning_rate": 5.9879239650645925e-06, "loss": 1.3711, "step": 22345 }, { "epoch": 1.7512539184952978, "grad_norm": 14.065950393676758, "learning_rate": 5.987621680238663e-06, "loss": 1.281, "step": 22346 }, { "epoch": 1.7513322884012539, "grad_norm": 3.610276699066162, "learning_rate": 5.987319391656331e-06, "loss": 1.2646, "step": 22347 }, { "epoch": 1.75141065830721, "grad_norm": 4.244760036468506, "learning_rate": 5.987017099318742e-06, "loss": 1.7399, "step": 22348 }, { "epoch": 1.751489028213166, "grad_norm": 3.0172693729400635, "learning_rate": 5.986714803227046e-06, "loss": 1.4489, "step": 22349 }, { "epoch": 1.7515673981191222, "grad_norm": 2.7110447883605957, "learning_rate": 5.986412503382394e-06, "loss": 1.3276, "step": 22350 }, { "epoch": 1.7516457680250783, "grad_norm": 3.3073489665985107, "learning_rate": 5.986110199785935e-06, "loss": 1.3549, "step": 22351 }, { "epoch": 1.7517241379310344, "grad_norm": 4.040341377258301, "learning_rate": 5.985807892438822e-06, "loss": 1.7343, "step": 22352 }, { "epoch": 1.7518025078369905, "grad_norm": 2.627589464187622, "learning_rate": 5.985505581342198e-06, "loss": 1.3625, "step": 22353 }, { "epoch": 1.7518808777429467, "grad_norm": 2.2785916328430176, "learning_rate": 5.985203266497221e-06, "loss": 1.4925, "step": 22354 }, { "epoch": 1.751959247648903, "grad_norm": 3.5606186389923096, "learning_rate": 5.984900947905032e-06, "loss": 1.4953, "step": 22355 }, { "epoch": 1.7520376175548589, "grad_norm": 3.00862979888916, "learning_rate": 5.984598625566791e-06, "loss": 1.0548, "step": 22356 }, { "epoch": 1.7521159874608152, "grad_norm": 3.4524784088134766, "learning_rate": 5.9842962994836396e-06, "loss": 1.3522, "step": 22357 }, { "epoch": 1.752194357366771, "grad_norm": 2.2709124088287354, "learning_rate": 5.98399396965673e-06, "loss": 1.4085, "step": 22358 }, { "epoch": 1.7522727272727274, "grad_norm": 3.454003095626831, "learning_rate": 5.9836916360872145e-06, "loss": 0.4772, "step": 22359 }, { "epoch": 1.7523510971786833, "grad_norm": 3.5654592514038086, "learning_rate": 5.983389298776241e-06, "loss": 1.1945, "step": 22360 }, { "epoch": 1.7524294670846396, "grad_norm": 1.8278594017028809, "learning_rate": 5.98308695772496e-06, "loss": 1.1532, "step": 22361 }, { "epoch": 1.7525078369905955, "grad_norm": 3.4892070293426514, "learning_rate": 5.982784612934519e-06, "loss": 1.4508, "step": 22362 }, { "epoch": 1.7525862068965519, "grad_norm": 2.9111478328704834, "learning_rate": 5.9824822644060735e-06, "loss": 1.3137, "step": 22363 }, { "epoch": 1.7526645768025078, "grad_norm": 2.4178895950317383, "learning_rate": 5.982179912140767e-06, "loss": 1.2854, "step": 22364 }, { "epoch": 1.752742946708464, "grad_norm": 2.4373576641082764, "learning_rate": 5.981877556139754e-06, "loss": 1.3702, "step": 22365 }, { "epoch": 1.75282131661442, "grad_norm": 5.975520133972168, "learning_rate": 5.981575196404184e-06, "loss": 1.3496, "step": 22366 }, { "epoch": 1.7528996865203763, "grad_norm": 2.582645893096924, "learning_rate": 5.981272832935204e-06, "loss": 1.3426, "step": 22367 }, { "epoch": 1.7529780564263322, "grad_norm": 4.453884124755859, "learning_rate": 5.9809704657339675e-06, "loss": 1.0103, "step": 22368 }, { "epoch": 1.7530564263322885, "grad_norm": 5.134122848510742, "learning_rate": 5.9806680948016225e-06, "loss": 1.0244, "step": 22369 }, { "epoch": 1.7531347962382444, "grad_norm": 3.303650140762329, "learning_rate": 5.9803657201393205e-06, "loss": 1.4507, "step": 22370 }, { "epoch": 1.7532131661442008, "grad_norm": 1.9838175773620605, "learning_rate": 5.98006334174821e-06, "loss": 1.1201, "step": 22371 }, { "epoch": 1.7532915360501566, "grad_norm": 2.120877504348755, "learning_rate": 5.979760959629442e-06, "loss": 1.1416, "step": 22372 }, { "epoch": 1.753369905956113, "grad_norm": 6.163236141204834, "learning_rate": 5.979458573784168e-06, "loss": 1.6676, "step": 22373 }, { "epoch": 1.7534482758620689, "grad_norm": 2.9326655864715576, "learning_rate": 5.979156184213535e-06, "loss": 1.288, "step": 22374 }, { "epoch": 1.7535266457680252, "grad_norm": 4.033809185028076, "learning_rate": 5.978853790918696e-06, "loss": 1.3263, "step": 22375 }, { "epoch": 1.753605015673981, "grad_norm": 12.406140327453613, "learning_rate": 5.9785513939008e-06, "loss": 1.0517, "step": 22376 }, { "epoch": 1.7536833855799374, "grad_norm": 3.640744209289551, "learning_rate": 5.9782489931609955e-06, "loss": 1.1473, "step": 22377 }, { "epoch": 1.7537617554858933, "grad_norm": 2.669454574584961, "learning_rate": 5.977946588700436e-06, "loss": 1.2663, "step": 22378 }, { "epoch": 1.7538401253918496, "grad_norm": 3.422764778137207, "learning_rate": 5.977644180520269e-06, "loss": 1.0112, "step": 22379 }, { "epoch": 1.7539184952978055, "grad_norm": 2.0432474613189697, "learning_rate": 5.977341768621647e-06, "loss": 1.2613, "step": 22380 }, { "epoch": 1.7539968652037619, "grad_norm": 2.5937540531158447, "learning_rate": 5.977039353005717e-06, "loss": 1.1885, "step": 22381 }, { "epoch": 1.7540752351097177, "grad_norm": 3.546963691711426, "learning_rate": 5.9767369336736325e-06, "loss": 1.7431, "step": 22382 }, { "epoch": 1.754153605015674, "grad_norm": 2.46256947517395, "learning_rate": 5.976434510626541e-06, "loss": 1.2438, "step": 22383 }, { "epoch": 1.75423197492163, "grad_norm": 5.012724876403809, "learning_rate": 5.9761320838655955e-06, "loss": 1.2371, "step": 22384 }, { "epoch": 1.7543103448275863, "grad_norm": 2.7470009326934814, "learning_rate": 5.9758296533919445e-06, "loss": 1.0261, "step": 22385 }, { "epoch": 1.7543887147335422, "grad_norm": 2.70272159576416, "learning_rate": 5.975527219206739e-06, "loss": 0.9312, "step": 22386 }, { "epoch": 1.7544670846394985, "grad_norm": 3.1469693183898926, "learning_rate": 5.975224781311129e-06, "loss": 1.4713, "step": 22387 }, { "epoch": 1.7545454545454544, "grad_norm": 3.1463241577148438, "learning_rate": 5.974922339706264e-06, "loss": 1.4698, "step": 22388 }, { "epoch": 1.7546238244514107, "grad_norm": 3.912306308746338, "learning_rate": 5.974619894393297e-06, "loss": 1.3335, "step": 22389 }, { "epoch": 1.7547021943573666, "grad_norm": 4.02325439453125, "learning_rate": 5.974317445373374e-06, "loss": 1.3559, "step": 22390 }, { "epoch": 1.754780564263323, "grad_norm": 2.919306755065918, "learning_rate": 5.97401499264765e-06, "loss": 1.3163, "step": 22391 }, { "epoch": 1.7548589341692789, "grad_norm": 2.29267954826355, "learning_rate": 5.973712536217273e-06, "loss": 1.2875, "step": 22392 }, { "epoch": 1.7549373040752352, "grad_norm": 4.6434783935546875, "learning_rate": 5.973410076083392e-06, "loss": 1.3369, "step": 22393 }, { "epoch": 1.755015673981191, "grad_norm": 3.3581857681274414, "learning_rate": 5.973107612247161e-06, "loss": 1.5244, "step": 22394 }, { "epoch": 1.7550940438871474, "grad_norm": 11.58947467803955, "learning_rate": 5.972805144709727e-06, "loss": 0.739, "step": 22395 }, { "epoch": 1.7551724137931033, "grad_norm": 3.7022511959075928, "learning_rate": 5.972502673472243e-06, "loss": 1.6168, "step": 22396 }, { "epoch": 1.7552507836990596, "grad_norm": 2.588245153427124, "learning_rate": 5.972200198535857e-06, "loss": 1.101, "step": 22397 }, { "epoch": 1.7553291536050155, "grad_norm": 2.7288010120391846, "learning_rate": 5.971897719901721e-06, "loss": 1.319, "step": 22398 }, { "epoch": 1.7554075235109718, "grad_norm": 3.3626511096954346, "learning_rate": 5.9715952375709876e-06, "loss": 1.3241, "step": 22399 }, { "epoch": 1.7554858934169277, "grad_norm": 2.548035144805908, "learning_rate": 5.971292751544802e-06, "loss": 1.1766, "step": 22400 }, { "epoch": 1.755564263322884, "grad_norm": 4.245662689208984, "learning_rate": 5.97099026182432e-06, "loss": 1.4928, "step": 22401 }, { "epoch": 1.7556426332288402, "grad_norm": 2.26853346824646, "learning_rate": 5.970687768410688e-06, "loss": 1.1954, "step": 22402 }, { "epoch": 1.7557210031347963, "grad_norm": 2.7338972091674805, "learning_rate": 5.970385271305059e-06, "loss": 1.2927, "step": 22403 }, { "epoch": 1.7557993730407524, "grad_norm": 3.655884265899658, "learning_rate": 5.970082770508583e-06, "loss": 1.4157, "step": 22404 }, { "epoch": 1.7558777429467085, "grad_norm": 2.5548486709594727, "learning_rate": 5.96978026602241e-06, "loss": 1.3365, "step": 22405 }, { "epoch": 1.7559561128526646, "grad_norm": 3.2649919986724854, "learning_rate": 5.969477757847692e-06, "loss": 1.8348, "step": 22406 }, { "epoch": 1.7560344827586207, "grad_norm": 4.046197891235352, "learning_rate": 5.9691752459855765e-06, "loss": 1.1632, "step": 22407 }, { "epoch": 1.7561128526645768, "grad_norm": 2.629105806350708, "learning_rate": 5.968872730437219e-06, "loss": 1.4883, "step": 22408 }, { "epoch": 1.756191222570533, "grad_norm": 4.340452671051025, "learning_rate": 5.968570211203764e-06, "loss": 1.1416, "step": 22409 }, { "epoch": 1.756269592476489, "grad_norm": 2.19219708442688, "learning_rate": 5.968267688286369e-06, "loss": 1.0907, "step": 22410 }, { "epoch": 1.7563479623824452, "grad_norm": 5.774455547332764, "learning_rate": 5.967965161686179e-06, "loss": 1.2812, "step": 22411 }, { "epoch": 1.7564263322884013, "grad_norm": 2.6619105339050293, "learning_rate": 5.9676626314043466e-06, "loss": 1.4223, "step": 22412 }, { "epoch": 1.7565047021943574, "grad_norm": 2.149005651473999, "learning_rate": 5.967360097442022e-06, "loss": 1.282, "step": 22413 }, { "epoch": 1.7565830721003135, "grad_norm": 3.964000701904297, "learning_rate": 5.967057559800358e-06, "loss": 1.4921, "step": 22414 }, { "epoch": 1.7566614420062696, "grad_norm": 3.452216148376465, "learning_rate": 5.966755018480503e-06, "loss": 1.5928, "step": 22415 }, { "epoch": 1.7567398119122257, "grad_norm": 3.5408949851989746, "learning_rate": 5.966452473483609e-06, "loss": 1.3953, "step": 22416 }, { "epoch": 1.7568181818181818, "grad_norm": 4.071410179138184, "learning_rate": 5.966149924810825e-06, "loss": 1.4094, "step": 22417 }, { "epoch": 1.756896551724138, "grad_norm": 2.7729713916778564, "learning_rate": 5.9658473724633025e-06, "loss": 1.494, "step": 22418 }, { "epoch": 1.756974921630094, "grad_norm": 3.621317148208618, "learning_rate": 5.965544816442196e-06, "loss": 1.6036, "step": 22419 }, { "epoch": 1.7570532915360502, "grad_norm": 2.2342865467071533, "learning_rate": 5.96524225674865e-06, "loss": 1.0106, "step": 22420 }, { "epoch": 1.7571316614420063, "grad_norm": 4.059857368469238, "learning_rate": 5.964939693383819e-06, "loss": 1.406, "step": 22421 }, { "epoch": 1.7572100313479624, "grad_norm": 3.4759902954101562, "learning_rate": 5.964637126348854e-06, "loss": 1.7444, "step": 22422 }, { "epoch": 1.7572884012539185, "grad_norm": 3.129694938659668, "learning_rate": 5.964334555644903e-06, "loss": 1.0766, "step": 22423 }, { "epoch": 1.7573667711598746, "grad_norm": 3.8745570182800293, "learning_rate": 5.96403198127312e-06, "loss": 1.2381, "step": 22424 }, { "epoch": 1.7574451410658307, "grad_norm": 3.621776580810547, "learning_rate": 5.963729403234653e-06, "loss": 1.3835, "step": 22425 }, { "epoch": 1.7575235109717868, "grad_norm": 2.7713375091552734, "learning_rate": 5.963426821530655e-06, "loss": 1.16, "step": 22426 }, { "epoch": 1.757601880877743, "grad_norm": 3.9424962997436523, "learning_rate": 5.963124236162278e-06, "loss": 1.0738, "step": 22427 }, { "epoch": 1.757680250783699, "grad_norm": 1.7260226011276245, "learning_rate": 5.9628216471306686e-06, "loss": 0.886, "step": 22428 }, { "epoch": 1.7577586206896552, "grad_norm": 2.2339839935302734, "learning_rate": 5.9625190544369826e-06, "loss": 1.371, "step": 22429 }, { "epoch": 1.7578369905956113, "grad_norm": 2.7842204570770264, "learning_rate": 5.962216458082366e-06, "loss": 1.2175, "step": 22430 }, { "epoch": 1.7579153605015674, "grad_norm": 3.5913853645324707, "learning_rate": 5.961913858067973e-06, "loss": 1.0583, "step": 22431 }, { "epoch": 1.7579937304075235, "grad_norm": 3.886150598526001, "learning_rate": 5.9616112543949556e-06, "loss": 1.0298, "step": 22432 }, { "epoch": 1.7580721003134796, "grad_norm": 3.5526530742645264, "learning_rate": 5.961308647064461e-06, "loss": 1.3738, "step": 22433 }, { "epoch": 1.7581504702194357, "grad_norm": 2.581732988357544, "learning_rate": 5.961006036077644e-06, "loss": 1.3503, "step": 22434 }, { "epoch": 1.7582288401253918, "grad_norm": 5.9792094230651855, "learning_rate": 5.960703421435651e-06, "loss": 1.4035, "step": 22435 }, { "epoch": 1.758307210031348, "grad_norm": 3.195692777633667, "learning_rate": 5.960400803139638e-06, "loss": 1.282, "step": 22436 }, { "epoch": 1.758385579937304, "grad_norm": 2.883653402328491, "learning_rate": 5.960098181190751e-06, "loss": 1.4265, "step": 22437 }, { "epoch": 1.7584639498432602, "grad_norm": 2.594597578048706, "learning_rate": 5.959795555590146e-06, "loss": 1.2197, "step": 22438 }, { "epoch": 1.7585423197492163, "grad_norm": 2.9918086528778076, "learning_rate": 5.95949292633897e-06, "loss": 1.0067, "step": 22439 }, { "epoch": 1.7586206896551724, "grad_norm": 8.040680885314941, "learning_rate": 5.9591902934383776e-06, "loss": 0.7556, "step": 22440 }, { "epoch": 1.7586990595611285, "grad_norm": 4.549802303314209, "learning_rate": 5.958887656889516e-06, "loss": 1.4133, "step": 22441 }, { "epoch": 1.7587774294670846, "grad_norm": 2.791416883468628, "learning_rate": 5.958585016693539e-06, "loss": 1.3658, "step": 22442 }, { "epoch": 1.7588557993730407, "grad_norm": 11.280641555786133, "learning_rate": 5.958282372851597e-06, "loss": 0.745, "step": 22443 }, { "epoch": 1.7589341692789968, "grad_norm": 3.5820364952087402, "learning_rate": 5.957979725364842e-06, "loss": 1.7654, "step": 22444 }, { "epoch": 1.759012539184953, "grad_norm": 5.085467338562012, "learning_rate": 5.9576770742344225e-06, "loss": 1.5991, "step": 22445 }, { "epoch": 1.759090909090909, "grad_norm": 3.691535234451294, "learning_rate": 5.957374419461494e-06, "loss": 1.6806, "step": 22446 }, { "epoch": 1.7591692789968651, "grad_norm": 6.19473123550415, "learning_rate": 5.957071761047203e-06, "loss": 1.4092, "step": 22447 }, { "epoch": 1.7592476489028215, "grad_norm": 3.7110440731048584, "learning_rate": 5.956769098992702e-06, "loss": 1.2699, "step": 22448 }, { "epoch": 1.7593260188087774, "grad_norm": 2.4326000213623047, "learning_rate": 5.956466433299144e-06, "loss": 1.5284, "step": 22449 }, { "epoch": 1.7594043887147337, "grad_norm": 3.1004810333251953, "learning_rate": 5.956163763967678e-06, "loss": 1.1759, "step": 22450 }, { "epoch": 1.7594827586206896, "grad_norm": 2.5509417057037354, "learning_rate": 5.955861090999457e-06, "loss": 1.3197, "step": 22451 }, { "epoch": 1.759561128526646, "grad_norm": 3.5305798053741455, "learning_rate": 5.95555841439563e-06, "loss": 0.7976, "step": 22452 }, { "epoch": 1.7596394984326018, "grad_norm": 4.770324230194092, "learning_rate": 5.955255734157353e-06, "loss": 1.4006, "step": 22453 }, { "epoch": 1.7597178683385581, "grad_norm": 2.5437684059143066, "learning_rate": 5.954953050285769e-06, "loss": 1.1152, "step": 22454 }, { "epoch": 1.759796238244514, "grad_norm": 2.3270838260650635, "learning_rate": 5.954650362782038e-06, "loss": 1.4506, "step": 22455 }, { "epoch": 1.7598746081504704, "grad_norm": 3.314138650894165, "learning_rate": 5.954347671647305e-06, "loss": 1.5823, "step": 22456 }, { "epoch": 1.7599529780564263, "grad_norm": 3.8518407344818115, "learning_rate": 5.954044976882725e-06, "loss": 1.1654, "step": 22457 }, { "epoch": 1.7600313479623826, "grad_norm": 3.4777939319610596, "learning_rate": 5.9537422784894475e-06, "loss": 1.2657, "step": 22458 }, { "epoch": 1.7601097178683385, "grad_norm": 2.5836803913116455, "learning_rate": 5.9534395764686235e-06, "loss": 1.2938, "step": 22459 }, { "epoch": 1.7601880877742948, "grad_norm": 2.3524765968322754, "learning_rate": 5.953136870821407e-06, "loss": 1.1528, "step": 22460 }, { "epoch": 1.7602664576802507, "grad_norm": 3.415479898452759, "learning_rate": 5.952834161548945e-06, "loss": 1.1025, "step": 22461 }, { "epoch": 1.760344827586207, "grad_norm": 2.398353099822998, "learning_rate": 5.952531448652394e-06, "loss": 1.5167, "step": 22462 }, { "epoch": 1.760423197492163, "grad_norm": 2.423645257949829, "learning_rate": 5.9522287321328995e-06, "loss": 1.4765, "step": 22463 }, { "epoch": 1.7605015673981192, "grad_norm": 2.869586944580078, "learning_rate": 5.951926011991619e-06, "loss": 1.3194, "step": 22464 }, { "epoch": 1.7605799373040751, "grad_norm": 2.0839569568634033, "learning_rate": 5.951623288229699e-06, "loss": 1.1463, "step": 22465 }, { "epoch": 1.7606583072100315, "grad_norm": 2.078456401824951, "learning_rate": 5.951320560848293e-06, "loss": 1.2481, "step": 22466 }, { "epoch": 1.7607366771159874, "grad_norm": 2.5745444297790527, "learning_rate": 5.951017829848553e-06, "loss": 1.4904, "step": 22467 }, { "epoch": 1.7608150470219437, "grad_norm": 2.0572049617767334, "learning_rate": 5.950715095231629e-06, "loss": 1.1544, "step": 22468 }, { "epoch": 1.7608934169278996, "grad_norm": 2.518157720565796, "learning_rate": 5.950412356998672e-06, "loss": 1.4487, "step": 22469 }, { "epoch": 1.760971786833856, "grad_norm": 2.5658814907073975, "learning_rate": 5.950109615150836e-06, "loss": 1.3165, "step": 22470 }, { "epoch": 1.7610501567398118, "grad_norm": 2.3179285526275635, "learning_rate": 5.949806869689271e-06, "loss": 1.3518, "step": 22471 }, { "epoch": 1.7611285266457681, "grad_norm": 2.976501703262329, "learning_rate": 5.949504120615128e-06, "loss": 1.2786, "step": 22472 }, { "epoch": 1.761206896551724, "grad_norm": 5.655116081237793, "learning_rate": 5.94920136792956e-06, "loss": 1.3892, "step": 22473 }, { "epoch": 1.7612852664576804, "grad_norm": 7.52768087387085, "learning_rate": 5.948898611633718e-06, "loss": 0.9884, "step": 22474 }, { "epoch": 1.7613636363636362, "grad_norm": 3.930570363998413, "learning_rate": 5.948595851728751e-06, "loss": 1.3684, "step": 22475 }, { "epoch": 1.7614420062695926, "grad_norm": 7.938914775848389, "learning_rate": 5.9482930882158145e-06, "loss": 1.5964, "step": 22476 }, { "epoch": 1.7615203761755485, "grad_norm": 3.817577838897705, "learning_rate": 5.9479903210960565e-06, "loss": 1.4088, "step": 22477 }, { "epoch": 1.7615987460815048, "grad_norm": 2.447552442550659, "learning_rate": 5.947687550370631e-06, "loss": 1.4024, "step": 22478 }, { "epoch": 1.7616771159874607, "grad_norm": 3.7189903259277344, "learning_rate": 5.947384776040689e-06, "loss": 1.1945, "step": 22479 }, { "epoch": 1.761755485893417, "grad_norm": 3.0564801692962646, "learning_rate": 5.947081998107381e-06, "loss": 1.4059, "step": 22480 }, { "epoch": 1.761833855799373, "grad_norm": 3.508668899536133, "learning_rate": 5.946779216571862e-06, "loss": 1.2801, "step": 22481 }, { "epoch": 1.7619122257053292, "grad_norm": 2.691479444503784, "learning_rate": 5.946476431435278e-06, "loss": 1.3335, "step": 22482 }, { "epoch": 1.7619905956112851, "grad_norm": 2.425283670425415, "learning_rate": 5.946173642698786e-06, "loss": 1.2553, "step": 22483 }, { "epoch": 1.7620689655172415, "grad_norm": 2.660205364227295, "learning_rate": 5.945870850363533e-06, "loss": 1.3344, "step": 22484 }, { "epoch": 1.7621473354231973, "grad_norm": 2.52297043800354, "learning_rate": 5.945568054430676e-06, "loss": 1.2503, "step": 22485 }, { "epoch": 1.7622257053291537, "grad_norm": 2.254760265350342, "learning_rate": 5.9452652549013625e-06, "loss": 1.3083, "step": 22486 }, { "epoch": 1.7623040752351096, "grad_norm": 2.5808873176574707, "learning_rate": 5.944962451776745e-06, "loss": 1.4198, "step": 22487 }, { "epoch": 1.762382445141066, "grad_norm": 3.6528079509735107, "learning_rate": 5.944659645057978e-06, "loss": 1.4564, "step": 22488 }, { "epoch": 1.7624608150470218, "grad_norm": 2.6639177799224854, "learning_rate": 5.944356834746207e-06, "loss": 1.2578, "step": 22489 }, { "epoch": 1.7625391849529781, "grad_norm": 3.0283865928649902, "learning_rate": 5.944054020842591e-06, "loss": 1.3471, "step": 22490 }, { "epoch": 1.762617554858934, "grad_norm": 2.7819838523864746, "learning_rate": 5.943751203348276e-06, "loss": 1.4979, "step": 22491 }, { "epoch": 1.7626959247648903, "grad_norm": 3.40705943107605, "learning_rate": 5.943448382264418e-06, "loss": 1.1419, "step": 22492 }, { "epoch": 1.7627742946708462, "grad_norm": 2.693307399749756, "learning_rate": 5.943145557592167e-06, "loss": 1.3483, "step": 22493 }, { "epoch": 1.7628526645768026, "grad_norm": 5.768039226531982, "learning_rate": 5.942842729332674e-06, "loss": 1.2817, "step": 22494 }, { "epoch": 1.7629310344827587, "grad_norm": 4.59761905670166, "learning_rate": 5.942539897487092e-06, "loss": 1.6448, "step": 22495 }, { "epoch": 1.7630094043887148, "grad_norm": 3.8176393508911133, "learning_rate": 5.942237062056572e-06, "loss": 1.329, "step": 22496 }, { "epoch": 1.763087774294671, "grad_norm": 2.4046640396118164, "learning_rate": 5.941934223042265e-06, "loss": 1.3748, "step": 22497 }, { "epoch": 1.763166144200627, "grad_norm": 3.6469080448150635, "learning_rate": 5.941631380445325e-06, "loss": 1.8454, "step": 22498 }, { "epoch": 1.7632445141065831, "grad_norm": 10.125755310058594, "learning_rate": 5.941328534266903e-06, "loss": 0.9125, "step": 22499 }, { "epoch": 1.7633228840125392, "grad_norm": 2.932742118835449, "learning_rate": 5.941025684508152e-06, "loss": 1.0391, "step": 22500 }, { "epoch": 1.7634012539184953, "grad_norm": 6.294980525970459, "learning_rate": 5.940722831170221e-06, "loss": 0.8221, "step": 22501 }, { "epoch": 1.7634796238244514, "grad_norm": 2.190140962600708, "learning_rate": 5.940419974254265e-06, "loss": 1.4156, "step": 22502 }, { "epoch": 1.7635579937304076, "grad_norm": 2.2095401287078857, "learning_rate": 5.9401171137614325e-06, "loss": 1.0435, "step": 22503 }, { "epoch": 1.7636363636363637, "grad_norm": 4.918929100036621, "learning_rate": 5.939814249692878e-06, "loss": 0.8926, "step": 22504 }, { "epoch": 1.7637147335423198, "grad_norm": 5.427492618560791, "learning_rate": 5.939511382049753e-06, "loss": 1.6316, "step": 22505 }, { "epoch": 1.763793103448276, "grad_norm": 3.2426278591156006, "learning_rate": 5.93920851083321e-06, "loss": 1.5118, "step": 22506 }, { "epoch": 1.763871473354232, "grad_norm": 2.534911632537842, "learning_rate": 5.9389056360443995e-06, "loss": 1.3384, "step": 22507 }, { "epoch": 1.763949843260188, "grad_norm": 3.701683521270752, "learning_rate": 5.938602757684473e-06, "loss": 1.723, "step": 22508 }, { "epoch": 1.7640282131661442, "grad_norm": 3.71230149269104, "learning_rate": 5.938299875754586e-06, "loss": 1.7917, "step": 22509 }, { "epoch": 1.7641065830721003, "grad_norm": 7.010741233825684, "learning_rate": 5.937996990255886e-06, "loss": 0.6343, "step": 22510 }, { "epoch": 1.7641849529780564, "grad_norm": 5.12568998336792, "learning_rate": 5.937694101189529e-06, "loss": 1.752, "step": 22511 }, { "epoch": 1.7642633228840126, "grad_norm": 2.8036980628967285, "learning_rate": 5.937391208556665e-06, "loss": 1.6198, "step": 22512 }, { "epoch": 1.7643416927899687, "grad_norm": 5.426863670349121, "learning_rate": 5.937088312358447e-06, "loss": 0.8473, "step": 22513 }, { "epoch": 1.7644200626959248, "grad_norm": 2.4717252254486084, "learning_rate": 5.9367854125960246e-06, "loss": 1.381, "step": 22514 }, { "epoch": 1.7644984326018809, "grad_norm": 2.4459657669067383, "learning_rate": 5.936482509270552e-06, "loss": 1.2972, "step": 22515 }, { "epoch": 1.764576802507837, "grad_norm": 3.259971857070923, "learning_rate": 5.936179602383183e-06, "loss": 1.3495, "step": 22516 }, { "epoch": 1.764655172413793, "grad_norm": 2.261570692062378, "learning_rate": 5.9358766919350655e-06, "loss": 1.0532, "step": 22517 }, { "epoch": 1.7647335423197492, "grad_norm": 2.8386781215667725, "learning_rate": 5.935573777927355e-06, "loss": 1.2761, "step": 22518 }, { "epoch": 1.7648119122257053, "grad_norm": 3.4084560871124268, "learning_rate": 5.9352708603612e-06, "loss": 1.3964, "step": 22519 }, { "epoch": 1.7648902821316614, "grad_norm": 5.396134853363037, "learning_rate": 5.934967939237758e-06, "loss": 1.3321, "step": 22520 }, { "epoch": 1.7649686520376175, "grad_norm": 17.884944915771484, "learning_rate": 5.9346650145581765e-06, "loss": 0.5154, "step": 22521 }, { "epoch": 1.7650470219435737, "grad_norm": 2.210803270339966, "learning_rate": 5.934362086323609e-06, "loss": 1.2857, "step": 22522 }, { "epoch": 1.7651253918495298, "grad_norm": 2.781428575515747, "learning_rate": 5.93405915453521e-06, "loss": 1.2914, "step": 22523 }, { "epoch": 1.7652037617554859, "grad_norm": 1.7561933994293213, "learning_rate": 5.933756219194128e-06, "loss": 1.1242, "step": 22524 }, { "epoch": 1.765282131661442, "grad_norm": 1.6806529760360718, "learning_rate": 5.9334532803015174e-06, "loss": 1.2053, "step": 22525 }, { "epoch": 1.765360501567398, "grad_norm": 2.524167537689209, "learning_rate": 5.93315033785853e-06, "loss": 1.3635, "step": 22526 }, { "epoch": 1.7654388714733542, "grad_norm": 2.0290772914886475, "learning_rate": 5.932847391866318e-06, "loss": 1.2509, "step": 22527 }, { "epoch": 1.7655172413793103, "grad_norm": 2.9389190673828125, "learning_rate": 5.9325444423260344e-06, "loss": 1.826, "step": 22528 }, { "epoch": 1.7655956112852664, "grad_norm": 2.9950368404388428, "learning_rate": 5.932241489238829e-06, "loss": 1.3992, "step": 22529 }, { "epoch": 1.7656739811912225, "grad_norm": 3.795581102371216, "learning_rate": 5.931938532605858e-06, "loss": 1.3702, "step": 22530 }, { "epoch": 1.7657523510971787, "grad_norm": 4.024925231933594, "learning_rate": 5.93163557242827e-06, "loss": 1.487, "step": 22531 }, { "epoch": 1.7658307210031348, "grad_norm": 3.1311371326446533, "learning_rate": 5.931332608707219e-06, "loss": 1.7207, "step": 22532 }, { "epoch": 1.7659090909090909, "grad_norm": 3.4642174243927, "learning_rate": 5.931029641443857e-06, "loss": 1.3333, "step": 22533 }, { "epoch": 1.765987460815047, "grad_norm": 3.4177660942077637, "learning_rate": 5.930726670639337e-06, "loss": 1.5, "step": 22534 }, { "epoch": 1.766065830721003, "grad_norm": 2.8368022441864014, "learning_rate": 5.930423696294811e-06, "loss": 1.4246, "step": 22535 }, { "epoch": 1.7661442006269592, "grad_norm": 9.016653060913086, "learning_rate": 5.9301207184114295e-06, "loss": 1.3747, "step": 22536 }, { "epoch": 1.7662225705329153, "grad_norm": 4.94528341293335, "learning_rate": 5.9298177369903496e-06, "loss": 1.5151, "step": 22537 }, { "epoch": 1.7663009404388714, "grad_norm": 4.309382915496826, "learning_rate": 5.9295147520327165e-06, "loss": 1.2056, "step": 22538 }, { "epoch": 1.7663793103448275, "grad_norm": 1.987345814704895, "learning_rate": 5.929211763539691e-06, "loss": 1.2576, "step": 22539 }, { "epoch": 1.7664576802507836, "grad_norm": 1.781982660293579, "learning_rate": 5.928908771512418e-06, "loss": 1.2752, "step": 22540 }, { "epoch": 1.7665360501567398, "grad_norm": 2.6577858924865723, "learning_rate": 5.9286057759520555e-06, "loss": 1.3632, "step": 22541 }, { "epoch": 1.7666144200626959, "grad_norm": 5.077421188354492, "learning_rate": 5.928302776859752e-06, "loss": 1.1411, "step": 22542 }, { "epoch": 1.7666927899686522, "grad_norm": 3.7484567165374756, "learning_rate": 5.927999774236661e-06, "loss": 1.5127, "step": 22543 }, { "epoch": 1.766771159874608, "grad_norm": 1.7431602478027344, "learning_rate": 5.927696768083938e-06, "loss": 1.1461, "step": 22544 }, { "epoch": 1.7668495297805644, "grad_norm": 3.4984169006347656, "learning_rate": 5.927393758402731e-06, "loss": 1.5713, "step": 22545 }, { "epoch": 1.7669278996865203, "grad_norm": 3.374142646789551, "learning_rate": 5.927090745194195e-06, "loss": 1.055, "step": 22546 }, { "epoch": 1.7670062695924766, "grad_norm": 2.351356267929077, "learning_rate": 5.9267877284594825e-06, "loss": 1.3408, "step": 22547 }, { "epoch": 1.7670846394984325, "grad_norm": 4.7565484046936035, "learning_rate": 5.926484708199745e-06, "loss": 1.5229, "step": 22548 }, { "epoch": 1.7671630094043889, "grad_norm": 2.013641834259033, "learning_rate": 5.926181684416135e-06, "loss": 1.5944, "step": 22549 }, { "epoch": 1.7672413793103448, "grad_norm": 1.9795441627502441, "learning_rate": 5.925878657109806e-06, "loss": 1.3165, "step": 22550 }, { "epoch": 1.767319749216301, "grad_norm": 3.2066099643707275, "learning_rate": 5.9255756262819095e-06, "loss": 1.6673, "step": 22551 }, { "epoch": 1.767398119122257, "grad_norm": 4.787670612335205, "learning_rate": 5.9252725919336e-06, "loss": 1.2681, "step": 22552 }, { "epoch": 1.7674764890282133, "grad_norm": 3.48875093460083, "learning_rate": 5.924969554066028e-06, "loss": 1.3436, "step": 22553 }, { "epoch": 1.7675548589341692, "grad_norm": 5.640324592590332, "learning_rate": 5.924666512680349e-06, "loss": 1.6358, "step": 22554 }, { "epoch": 1.7676332288401255, "grad_norm": 2.5536117553710938, "learning_rate": 5.924363467777709e-06, "loss": 1.1747, "step": 22555 }, { "epoch": 1.7677115987460814, "grad_norm": 2.4156458377838135, "learning_rate": 5.92406041935927e-06, "loss": 1.4736, "step": 22556 }, { "epoch": 1.7677899686520377, "grad_norm": 2.5907366275787354, "learning_rate": 5.923757367426176e-06, "loss": 1.3125, "step": 22557 }, { "epoch": 1.7678683385579936, "grad_norm": 2.914381742477417, "learning_rate": 5.923454311979586e-06, "loss": 1.4259, "step": 22558 }, { "epoch": 1.76794670846395, "grad_norm": 2.048905611038208, "learning_rate": 5.923151253020649e-06, "loss": 1.2141, "step": 22559 }, { "epoch": 1.7680250783699059, "grad_norm": 3.607928514480591, "learning_rate": 5.922848190550519e-06, "loss": 1.529, "step": 22560 }, { "epoch": 1.7681034482758622, "grad_norm": 3.01676607131958, "learning_rate": 5.922545124570349e-06, "loss": 1.273, "step": 22561 }, { "epoch": 1.768181818181818, "grad_norm": 5.073705673217773, "learning_rate": 5.92224205508129e-06, "loss": 1.3623, "step": 22562 }, { "epoch": 1.7682601880877744, "grad_norm": 2.4631283283233643, "learning_rate": 5.921938982084498e-06, "loss": 1.5959, "step": 22563 }, { "epoch": 1.7683385579937303, "grad_norm": 2.3260042667388916, "learning_rate": 5.92163590558112e-06, "loss": 1.3182, "step": 22564 }, { "epoch": 1.7684169278996866, "grad_norm": 2.6595280170440674, "learning_rate": 5.921332825572317e-06, "loss": 1.5011, "step": 22565 }, { "epoch": 1.7684952978056425, "grad_norm": 4.102800369262695, "learning_rate": 5.921029742059234e-06, "loss": 1.2456, "step": 22566 }, { "epoch": 1.7685736677115989, "grad_norm": 2.720595598220825, "learning_rate": 5.920726655043027e-06, "loss": 1.3575, "step": 22567 }, { "epoch": 1.7686520376175547, "grad_norm": 2.6573801040649414, "learning_rate": 5.92042356452485e-06, "loss": 1.5322, "step": 22568 }, { "epoch": 1.768730407523511, "grad_norm": 8.92387866973877, "learning_rate": 5.920120470505853e-06, "loss": 0.7113, "step": 22569 }, { "epoch": 1.768808777429467, "grad_norm": 2.7534284591674805, "learning_rate": 5.919817372987192e-06, "loss": 1.2253, "step": 22570 }, { "epoch": 1.7688871473354233, "grad_norm": 2.916684150695801, "learning_rate": 5.9195142719700175e-06, "loss": 1.657, "step": 22571 }, { "epoch": 1.7689655172413792, "grad_norm": 10.292221069335938, "learning_rate": 5.919211167455484e-06, "loss": 1.5211, "step": 22572 }, { "epoch": 1.7690438871473355, "grad_norm": 2.7564754486083984, "learning_rate": 5.918908059444741e-06, "loss": 0.9511, "step": 22573 }, { "epoch": 1.7691222570532914, "grad_norm": 2.507657527923584, "learning_rate": 5.918604947938946e-06, "loss": 1.4048, "step": 22574 }, { "epoch": 1.7692006269592477, "grad_norm": 2.878086566925049, "learning_rate": 5.918301832939248e-06, "loss": 1.5649, "step": 22575 }, { "epoch": 1.7692789968652036, "grad_norm": 2.423842668533325, "learning_rate": 5.917998714446802e-06, "loss": 0.8611, "step": 22576 }, { "epoch": 1.76935736677116, "grad_norm": 3.099806308746338, "learning_rate": 5.917695592462761e-06, "loss": 1.4851, "step": 22577 }, { "epoch": 1.7694357366771158, "grad_norm": 10.30378532409668, "learning_rate": 5.917392466988276e-06, "loss": 1.2893, "step": 22578 }, { "epoch": 1.7695141065830722, "grad_norm": 3.8850927352905273, "learning_rate": 5.9170893380245035e-06, "loss": 1.3928, "step": 22579 }, { "epoch": 1.769592476489028, "grad_norm": 3.1612610816955566, "learning_rate": 5.916786205572592e-06, "loss": 1.652, "step": 22580 }, { "epoch": 1.7696708463949844, "grad_norm": 2.8106119632720947, "learning_rate": 5.916483069633697e-06, "loss": 1.1683, "step": 22581 }, { "epoch": 1.7697492163009403, "grad_norm": 2.123464822769165, "learning_rate": 5.916179930208973e-06, "loss": 1.5105, "step": 22582 }, { "epoch": 1.7698275862068966, "grad_norm": 2.783834934234619, "learning_rate": 5.915876787299568e-06, "loss": 1.1343, "step": 22583 }, { "epoch": 1.7699059561128525, "grad_norm": 2.8082592487335205, "learning_rate": 5.915573640906642e-06, "loss": 1.4281, "step": 22584 }, { "epoch": 1.7699843260188088, "grad_norm": 5.312647342681885, "learning_rate": 5.915270491031341e-06, "loss": 1.2359, "step": 22585 }, { "epoch": 1.7700626959247647, "grad_norm": 4.9071149826049805, "learning_rate": 5.914967337674823e-06, "loss": 0.7474, "step": 22586 }, { "epoch": 1.770141065830721, "grad_norm": 8.41566276550293, "learning_rate": 5.914664180838238e-06, "loss": 0.8806, "step": 22587 }, { "epoch": 1.770219435736677, "grad_norm": 2.7476966381073, "learning_rate": 5.914361020522741e-06, "loss": 1.386, "step": 22588 }, { "epoch": 1.7702978056426333, "grad_norm": 3.4431140422821045, "learning_rate": 5.9140578567294845e-06, "loss": 1.5821, "step": 22589 }, { "epoch": 1.7703761755485894, "grad_norm": 5.315615177154541, "learning_rate": 5.91375468945962e-06, "loss": 0.8245, "step": 22590 }, { "epoch": 1.7704545454545455, "grad_norm": 6.418537616729736, "learning_rate": 5.913451518714304e-06, "loss": 0.8648, "step": 22591 }, { "epoch": 1.7705329153605016, "grad_norm": 2.744978427886963, "learning_rate": 5.913148344494684e-06, "loss": 1.0697, "step": 22592 }, { "epoch": 1.7706112852664577, "grad_norm": 6.751672744750977, "learning_rate": 5.912845166801922e-06, "loss": 1.2051, "step": 22593 }, { "epoch": 1.7706896551724138, "grad_norm": 3.2525668144226074, "learning_rate": 5.912541985637163e-06, "loss": 1.4697, "step": 22594 }, { "epoch": 1.77076802507837, "grad_norm": 7.191387176513672, "learning_rate": 5.912238801001562e-06, "loss": 1.0818, "step": 22595 }, { "epoch": 1.770846394984326, "grad_norm": 2.473994016647339, "learning_rate": 5.911935612896274e-06, "loss": 0.9916, "step": 22596 }, { "epoch": 1.7709247648902822, "grad_norm": 2.2116246223449707, "learning_rate": 5.911632421322452e-06, "loss": 1.1662, "step": 22597 }, { "epoch": 1.7710031347962383, "grad_norm": 2.724364757537842, "learning_rate": 5.9113292262812475e-06, "loss": 1.0708, "step": 22598 }, { "epoch": 1.7710815047021944, "grad_norm": 2.2832236289978027, "learning_rate": 5.911026027773815e-06, "loss": 1.2106, "step": 22599 }, { "epoch": 1.7711598746081505, "grad_norm": 3.4872756004333496, "learning_rate": 5.9107228258013085e-06, "loss": 1.3759, "step": 22600 }, { "epoch": 1.7712382445141066, "grad_norm": 2.6819753646850586, "learning_rate": 5.91041962036488e-06, "loss": 1.5227, "step": 22601 }, { "epoch": 1.7713166144200627, "grad_norm": 3.689300060272217, "learning_rate": 5.910116411465682e-06, "loss": 1.5015, "step": 22602 }, { "epoch": 1.7713949843260188, "grad_norm": 2.051405906677246, "learning_rate": 5.909813199104868e-06, "loss": 1.4127, "step": 22603 }, { "epoch": 1.771473354231975, "grad_norm": 2.278912305831909, "learning_rate": 5.909509983283594e-06, "loss": 1.2705, "step": 22604 }, { "epoch": 1.771551724137931, "grad_norm": 5.698517799377441, "learning_rate": 5.909206764003008e-06, "loss": 1.5395, "step": 22605 }, { "epoch": 1.7716300940438872, "grad_norm": 2.799813985824585, "learning_rate": 5.9089035412642695e-06, "loss": 1.3834, "step": 22606 }, { "epoch": 1.7717084639498433, "grad_norm": 17.798734664916992, "learning_rate": 5.908600315068527e-06, "loss": 1.3908, "step": 22607 }, { "epoch": 1.7717868338557994, "grad_norm": 3.2734780311584473, "learning_rate": 5.908297085416938e-06, "loss": 0.9777, "step": 22608 }, { "epoch": 1.7718652037617555, "grad_norm": 3.7395527362823486, "learning_rate": 5.9079938523106495e-06, "loss": 1.2297, "step": 22609 }, { "epoch": 1.7719435736677116, "grad_norm": 3.288443088531494, "learning_rate": 5.907690615750822e-06, "loss": 0.3753, "step": 22610 }, { "epoch": 1.7720219435736677, "grad_norm": 5.370203971862793, "learning_rate": 5.9073873757386034e-06, "loss": 1.54, "step": 22611 }, { "epoch": 1.7721003134796238, "grad_norm": 7.333498001098633, "learning_rate": 5.907084132275151e-06, "loss": 0.7231, "step": 22612 }, { "epoch": 1.77217868338558, "grad_norm": 2.932626247406006, "learning_rate": 5.906780885361616e-06, "loss": 1.2577, "step": 22613 }, { "epoch": 1.772257053291536, "grad_norm": 3.531118154525757, "learning_rate": 5.9064776349991525e-06, "loss": 1.0027, "step": 22614 }, { "epoch": 1.7723354231974922, "grad_norm": 5.437429904937744, "learning_rate": 5.906174381188914e-06, "loss": 1.1825, "step": 22615 }, { "epoch": 1.7724137931034483, "grad_norm": 2.5723743438720703, "learning_rate": 5.905871123932052e-06, "loss": 1.8006, "step": 22616 }, { "epoch": 1.7724921630094044, "grad_norm": 11.6184663772583, "learning_rate": 5.905567863229723e-06, "loss": 1.5966, "step": 22617 }, { "epoch": 1.7725705329153605, "grad_norm": 3.837576150894165, "learning_rate": 5.905264599083078e-06, "loss": 1.3577, "step": 22618 }, { "epoch": 1.7726489028213166, "grad_norm": 3.0493037700653076, "learning_rate": 5.904961331493273e-06, "loss": 1.6816, "step": 22619 }, { "epoch": 1.7727272727272727, "grad_norm": 3.0844578742980957, "learning_rate": 5.904658060461457e-06, "loss": 1.6044, "step": 22620 }, { "epoch": 1.7728056426332288, "grad_norm": 6.495530128479004, "learning_rate": 5.904354785988789e-06, "loss": 1.4483, "step": 22621 }, { "epoch": 1.772884012539185, "grad_norm": 4.445104122161865, "learning_rate": 5.904051508076419e-06, "loss": 1.6863, "step": 22622 }, { "epoch": 1.772962382445141, "grad_norm": 4.412736415863037, "learning_rate": 5.903748226725502e-06, "loss": 1.2936, "step": 22623 }, { "epoch": 1.7730407523510971, "grad_norm": 3.189044952392578, "learning_rate": 5.90344494193719e-06, "loss": 1.4485, "step": 22624 }, { "epoch": 1.7731191222570533, "grad_norm": 2.5898044109344482, "learning_rate": 5.903141653712638e-06, "loss": 1.7181, "step": 22625 }, { "epoch": 1.7731974921630094, "grad_norm": 2.9479691982269287, "learning_rate": 5.9028383620529985e-06, "loss": 1.3777, "step": 22626 }, { "epoch": 1.7732758620689655, "grad_norm": 2.9263556003570557, "learning_rate": 5.902535066959426e-06, "loss": 1.437, "step": 22627 }, { "epoch": 1.7733542319749216, "grad_norm": 2.5700573921203613, "learning_rate": 5.902231768433075e-06, "loss": 1.1403, "step": 22628 }, { "epoch": 1.7734326018808777, "grad_norm": 4.167846202850342, "learning_rate": 5.9019284664750964e-06, "loss": 1.121, "step": 22629 }, { "epoch": 1.7735109717868338, "grad_norm": 4.321710109710693, "learning_rate": 5.901625161086645e-06, "loss": 1.6864, "step": 22630 }, { "epoch": 1.77358934169279, "grad_norm": 4.839382171630859, "learning_rate": 5.901321852268875e-06, "loss": 1.3409, "step": 22631 }, { "epoch": 1.773667711598746, "grad_norm": 2.3521883487701416, "learning_rate": 5.901018540022939e-06, "loss": 1.2663, "step": 22632 }, { "epoch": 1.7737460815047021, "grad_norm": 4.248964309692383, "learning_rate": 5.900715224349992e-06, "loss": 1.2341, "step": 22633 }, { "epoch": 1.7738244514106583, "grad_norm": 8.953577995300293, "learning_rate": 5.9004119052511865e-06, "loss": 1.3517, "step": 22634 }, { "epoch": 1.7739028213166144, "grad_norm": 2.643122434616089, "learning_rate": 5.900108582727677e-06, "loss": 1.3737, "step": 22635 }, { "epoch": 1.7739811912225707, "grad_norm": 3.8213820457458496, "learning_rate": 5.899805256780617e-06, "loss": 1.037, "step": 22636 }, { "epoch": 1.7740595611285266, "grad_norm": 2.4979491233825684, "learning_rate": 5.899501927411159e-06, "loss": 0.7282, "step": 22637 }, { "epoch": 1.774137931034483, "grad_norm": 3.89609956741333, "learning_rate": 5.899198594620459e-06, "loss": 1.4604, "step": 22638 }, { "epoch": 1.7742163009404388, "grad_norm": 4.225036144256592, "learning_rate": 5.898895258409667e-06, "loss": 1.5888, "step": 22639 }, { "epoch": 1.7742946708463951, "grad_norm": 3.2604622840881348, "learning_rate": 5.898591918779942e-06, "loss": 1.2205, "step": 22640 }, { "epoch": 1.774373040752351, "grad_norm": 2.5219483375549316, "learning_rate": 5.898288575732434e-06, "loss": 1.4098, "step": 22641 }, { "epoch": 1.7744514106583074, "grad_norm": 6.271130084991455, "learning_rate": 5.8979852292682975e-06, "loss": 1.7408, "step": 22642 }, { "epoch": 1.7745297805642632, "grad_norm": 2.676171064376831, "learning_rate": 5.897681879388686e-06, "loss": 1.4726, "step": 22643 }, { "epoch": 1.7746081504702196, "grad_norm": 3.112762928009033, "learning_rate": 5.897378526094753e-06, "loss": 1.39, "step": 22644 }, { "epoch": 1.7746865203761755, "grad_norm": 3.0495805740356445, "learning_rate": 5.897075169387655e-06, "loss": 1.4069, "step": 22645 }, { "epoch": 1.7747648902821318, "grad_norm": 3.0238771438598633, "learning_rate": 5.896771809268542e-06, "loss": 1.2943, "step": 22646 }, { "epoch": 1.7748432601880877, "grad_norm": 3.608285427093506, "learning_rate": 5.896468445738571e-06, "loss": 1.2099, "step": 22647 }, { "epoch": 1.774921630094044, "grad_norm": 2.3560307025909424, "learning_rate": 5.896165078798894e-06, "loss": 1.054, "step": 22648 }, { "epoch": 1.775, "grad_norm": 6.324848651885986, "learning_rate": 5.895861708450665e-06, "loss": 0.9357, "step": 22649 }, { "epoch": 1.7750783699059562, "grad_norm": 3.369474411010742, "learning_rate": 5.895558334695039e-06, "loss": 1.1666, "step": 22650 }, { "epoch": 1.7751567398119121, "grad_norm": 2.7938289642333984, "learning_rate": 5.8952549575331675e-06, "loss": 1.2003, "step": 22651 }, { "epoch": 1.7752351097178685, "grad_norm": 2.2017698287963867, "learning_rate": 5.894951576966207e-06, "loss": 1.5192, "step": 22652 }, { "epoch": 1.7753134796238244, "grad_norm": 2.51877498626709, "learning_rate": 5.89464819299531e-06, "loss": 1.3282, "step": 22653 }, { "epoch": 1.7753918495297807, "grad_norm": 4.131408214569092, "learning_rate": 5.894344805621631e-06, "loss": 1.7548, "step": 22654 }, { "epoch": 1.7754702194357366, "grad_norm": 3.2961175441741943, "learning_rate": 5.894041414846325e-06, "loss": 1.3517, "step": 22655 }, { "epoch": 1.775548589341693, "grad_norm": 3.673858642578125, "learning_rate": 5.893738020670543e-06, "loss": 1.424, "step": 22656 }, { "epoch": 1.7756269592476488, "grad_norm": 3.3003363609313965, "learning_rate": 5.893434623095442e-06, "loss": 1.1638, "step": 22657 }, { "epoch": 1.7757053291536051, "grad_norm": 3.2548155784606934, "learning_rate": 5.893131222122174e-06, "loss": 1.3017, "step": 22658 }, { "epoch": 1.775783699059561, "grad_norm": 3.2924633026123047, "learning_rate": 5.892827817751893e-06, "loss": 1.0693, "step": 22659 }, { "epoch": 1.7758620689655173, "grad_norm": 3.39411997795105, "learning_rate": 5.892524409985754e-06, "loss": 1.3937, "step": 22660 }, { "epoch": 1.7759404388714732, "grad_norm": 2.740694522857666, "learning_rate": 5.892220998824909e-06, "loss": 1.4797, "step": 22661 }, { "epoch": 1.7760188087774296, "grad_norm": 1.9169130325317383, "learning_rate": 5.891917584270517e-06, "loss": 1.433, "step": 22662 }, { "epoch": 1.7760971786833855, "grad_norm": 10.84631061553955, "learning_rate": 5.8916141663237245e-06, "loss": 1.0438, "step": 22663 }, { "epoch": 1.7761755485893418, "grad_norm": 8.610417366027832, "learning_rate": 5.891310744985693e-06, "loss": 1.1814, "step": 22664 }, { "epoch": 1.7762539184952977, "grad_norm": 2.0407283306121826, "learning_rate": 5.891007320257571e-06, "loss": 1.5254, "step": 22665 }, { "epoch": 1.776332288401254, "grad_norm": 5.322789669036865, "learning_rate": 5.890703892140517e-06, "loss": 1.7254, "step": 22666 }, { "epoch": 1.77641065830721, "grad_norm": 2.940145492553711, "learning_rate": 5.890400460635682e-06, "loss": 1.2135, "step": 22667 }, { "epoch": 1.7764890282131662, "grad_norm": 8.983156204223633, "learning_rate": 5.89009702574422e-06, "loss": 0.9562, "step": 22668 }, { "epoch": 1.7765673981191221, "grad_norm": 2.98423171043396, "learning_rate": 5.889793587467287e-06, "loss": 1.4371, "step": 22669 }, { "epoch": 1.7766457680250785, "grad_norm": 3.6259167194366455, "learning_rate": 5.889490145806036e-06, "loss": 1.4446, "step": 22670 }, { "epoch": 1.7767241379310343, "grad_norm": 5.55975866317749, "learning_rate": 5.889186700761623e-06, "loss": 1.2673, "step": 22671 }, { "epoch": 1.7768025078369907, "grad_norm": 1.9716655015945435, "learning_rate": 5.888883252335198e-06, "loss": 1.1684, "step": 22672 }, { "epoch": 1.7768808777429466, "grad_norm": 3.6440091133117676, "learning_rate": 5.8885798005279195e-06, "loss": 1.6978, "step": 22673 }, { "epoch": 1.776959247648903, "grad_norm": 6.408882141113281, "learning_rate": 5.888276345340938e-06, "loss": 1.7421, "step": 22674 }, { "epoch": 1.7770376175548588, "grad_norm": 3.598283529281616, "learning_rate": 5.887972886775411e-06, "loss": 1.133, "step": 22675 }, { "epoch": 1.7771159874608151, "grad_norm": 7.046808242797852, "learning_rate": 5.887669424832491e-06, "loss": 1.3603, "step": 22676 }, { "epoch": 1.777194357366771, "grad_norm": 3.4893174171447754, "learning_rate": 5.887365959513331e-06, "loss": 1.3749, "step": 22677 }, { "epoch": 1.7772727272727273, "grad_norm": 3.074643611907959, "learning_rate": 5.8870624908190885e-06, "loss": 0.8828, "step": 22678 }, { "epoch": 1.7773510971786832, "grad_norm": 2.628075122833252, "learning_rate": 5.8867590187509145e-06, "loss": 1.5191, "step": 22679 }, { "epoch": 1.7774294670846396, "grad_norm": 2.938704252243042, "learning_rate": 5.8864555433099654e-06, "loss": 1.5433, "step": 22680 }, { "epoch": 1.7775078369905954, "grad_norm": 3.6042470932006836, "learning_rate": 5.886152064497395e-06, "loss": 1.5898, "step": 22681 }, { "epoch": 1.7775862068965518, "grad_norm": 2.688977003097534, "learning_rate": 5.8858485823143575e-06, "loss": 1.3878, "step": 22682 }, { "epoch": 1.777664576802508, "grad_norm": 5.896588325500488, "learning_rate": 5.885545096762007e-06, "loss": 2.1379, "step": 22683 }, { "epoch": 1.777742946708464, "grad_norm": 2.2949883937835693, "learning_rate": 5.8852416078414974e-06, "loss": 1.3181, "step": 22684 }, { "epoch": 1.77782131661442, "grad_norm": 2.7121951580047607, "learning_rate": 5.884938115553983e-06, "loss": 1.1056, "step": 22685 }, { "epoch": 1.7778996865203762, "grad_norm": 3.727884531021118, "learning_rate": 5.884634619900619e-06, "loss": 1.899, "step": 22686 }, { "epoch": 1.7779780564263323, "grad_norm": 2.560946226119995, "learning_rate": 5.884331120882559e-06, "loss": 1.536, "step": 22687 }, { "epoch": 1.7780564263322884, "grad_norm": 3.1721861362457275, "learning_rate": 5.884027618500958e-06, "loss": 1.6318, "step": 22688 }, { "epoch": 1.7781347962382446, "grad_norm": 2.1547183990478516, "learning_rate": 5.88372411275697e-06, "loss": 1.2443, "step": 22689 }, { "epoch": 1.7782131661442007, "grad_norm": 3.805420398712158, "learning_rate": 5.883420603651749e-06, "loss": 1.125, "step": 22690 }, { "epoch": 1.7782915360501568, "grad_norm": 3.5737946033477783, "learning_rate": 5.88311709118645e-06, "loss": 1.4658, "step": 22691 }, { "epoch": 1.7783699059561129, "grad_norm": 5.764222145080566, "learning_rate": 5.882813575362227e-06, "loss": 1.6523, "step": 22692 }, { "epoch": 1.778448275862069, "grad_norm": 3.057241201400757, "learning_rate": 5.882510056180234e-06, "loss": 1.389, "step": 22693 }, { "epoch": 1.778526645768025, "grad_norm": 2.883796215057373, "learning_rate": 5.8822065336416275e-06, "loss": 1.5126, "step": 22694 }, { "epoch": 1.7786050156739812, "grad_norm": 2.818845510482788, "learning_rate": 5.88190300774756e-06, "loss": 1.2588, "step": 22695 }, { "epoch": 1.7786833855799373, "grad_norm": 3.003399610519409, "learning_rate": 5.881599478499186e-06, "loss": 1.4324, "step": 22696 }, { "epoch": 1.7787617554858934, "grad_norm": 2.309199810028076, "learning_rate": 5.881295945897662e-06, "loss": 1.3507, "step": 22697 }, { "epoch": 1.7788401253918495, "grad_norm": 4.602280139923096, "learning_rate": 5.880992409944138e-06, "loss": 1.5273, "step": 22698 }, { "epoch": 1.7789184952978057, "grad_norm": 3.140953540802002, "learning_rate": 5.880688870639773e-06, "loss": 1.572, "step": 22699 }, { "epoch": 1.7789968652037618, "grad_norm": 3.301799774169922, "learning_rate": 5.88038532798572e-06, "loss": 0.897, "step": 22700 }, { "epoch": 1.7790752351097179, "grad_norm": 2.2185492515563965, "learning_rate": 5.880081781983134e-06, "loss": 1.6268, "step": 22701 }, { "epoch": 1.779153605015674, "grad_norm": 4.264637470245361, "learning_rate": 5.879778232633168e-06, "loss": 1.6742, "step": 22702 }, { "epoch": 1.77923197492163, "grad_norm": 8.546209335327148, "learning_rate": 5.879474679936977e-06, "loss": 0.8301, "step": 22703 }, { "epoch": 1.7793103448275862, "grad_norm": 3.4067792892456055, "learning_rate": 5.879171123895716e-06, "loss": 1.3949, "step": 22704 }, { "epoch": 1.7793887147335423, "grad_norm": 6.009164333343506, "learning_rate": 5.878867564510541e-06, "loss": 1.659, "step": 22705 }, { "epoch": 1.7794670846394984, "grad_norm": 2.370187282562256, "learning_rate": 5.878564001782604e-06, "loss": 1.5716, "step": 22706 }, { "epoch": 1.7795454545454545, "grad_norm": 9.946573257446289, "learning_rate": 5.878260435713061e-06, "loss": 2.1844, "step": 22707 }, { "epoch": 1.7796238244514107, "grad_norm": 3.1726579666137695, "learning_rate": 5.877956866303066e-06, "loss": 1.1091, "step": 22708 }, { "epoch": 1.7797021943573668, "grad_norm": 2.297828197479248, "learning_rate": 5.877653293553775e-06, "loss": 1.0343, "step": 22709 }, { "epoch": 1.7797805642633229, "grad_norm": 4.81138277053833, "learning_rate": 5.877349717466341e-06, "loss": 1.3788, "step": 22710 }, { "epoch": 1.779858934169279, "grad_norm": 3.636591911315918, "learning_rate": 5.877046138041921e-06, "loss": 0.9168, "step": 22711 }, { "epoch": 1.779937304075235, "grad_norm": 2.331000566482544, "learning_rate": 5.876742555281667e-06, "loss": 1.5551, "step": 22712 }, { "epoch": 1.7800156739811912, "grad_norm": 3.7989296913146973, "learning_rate": 5.876438969186733e-06, "loss": 1.1437, "step": 22713 }, { "epoch": 1.7800940438871473, "grad_norm": 3.2756099700927734, "learning_rate": 5.876135379758277e-06, "loss": 1.1966, "step": 22714 }, { "epoch": 1.7801724137931034, "grad_norm": 2.772813081741333, "learning_rate": 5.875831786997451e-06, "loss": 1.5409, "step": 22715 }, { "epoch": 1.7802507836990595, "grad_norm": 3.1447219848632812, "learning_rate": 5.875528190905412e-06, "loss": 1.35, "step": 22716 }, { "epoch": 1.7803291536050156, "grad_norm": 2.380516767501831, "learning_rate": 5.875224591483313e-06, "loss": 1.6932, "step": 22717 }, { "epoch": 1.7804075235109718, "grad_norm": 3.177812337875366, "learning_rate": 5.874920988732309e-06, "loss": 1.7998, "step": 22718 }, { "epoch": 1.7804858934169279, "grad_norm": 2.847055673599243, "learning_rate": 5.874617382653555e-06, "loss": 1.4165, "step": 22719 }, { "epoch": 1.780564263322884, "grad_norm": 13.286463737487793, "learning_rate": 5.874313773248206e-06, "loss": 1.3969, "step": 22720 }, { "epoch": 1.78064263322884, "grad_norm": 2.77776837348938, "learning_rate": 5.874010160517416e-06, "loss": 1.4569, "step": 22721 }, { "epoch": 1.7807210031347962, "grad_norm": 2.569204330444336, "learning_rate": 5.87370654446234e-06, "loss": 1.0939, "step": 22722 }, { "epoch": 1.7807993730407523, "grad_norm": 2.396151304244995, "learning_rate": 5.873402925084134e-06, "loss": 1.3886, "step": 22723 }, { "epoch": 1.7808777429467084, "grad_norm": 2.834460735321045, "learning_rate": 5.8730993023839515e-06, "loss": 1.9087, "step": 22724 }, { "epoch": 1.7809561128526645, "grad_norm": 2.755127191543579, "learning_rate": 5.872795676362948e-06, "loss": 1.2801, "step": 22725 }, { "epoch": 1.7810344827586206, "grad_norm": 2.4553027153015137, "learning_rate": 5.872492047022276e-06, "loss": 1.4353, "step": 22726 }, { "epoch": 1.7811128526645768, "grad_norm": 3.3439109325408936, "learning_rate": 5.872188414363096e-06, "loss": 1.3579, "step": 22727 }, { "epoch": 1.7811912225705329, "grad_norm": 2.7804958820343018, "learning_rate": 5.8718847783865555e-06, "loss": 1.1473, "step": 22728 }, { "epoch": 1.7812695924764892, "grad_norm": 3.8473262786865234, "learning_rate": 5.871581139093816e-06, "loss": 1.4783, "step": 22729 }, { "epoch": 1.781347962382445, "grad_norm": 2.8155460357666016, "learning_rate": 5.871277496486027e-06, "loss": 1.3098, "step": 22730 }, { "epoch": 1.7814263322884014, "grad_norm": 2.1886332035064697, "learning_rate": 5.870973850564348e-06, "loss": 1.1025, "step": 22731 }, { "epoch": 1.7815047021943573, "grad_norm": 1.8592402935028076, "learning_rate": 5.870670201329931e-06, "loss": 1.4544, "step": 22732 }, { "epoch": 1.7815830721003136, "grad_norm": 3.9364736080169678, "learning_rate": 5.870366548783932e-06, "loss": 1.3878, "step": 22733 }, { "epoch": 1.7816614420062695, "grad_norm": 4.386201858520508, "learning_rate": 5.870062892927506e-06, "loss": 1.5169, "step": 22734 }, { "epoch": 1.7817398119122259, "grad_norm": 2.286181688308716, "learning_rate": 5.869759233761807e-06, "loss": 1.3891, "step": 22735 }, { "epoch": 1.7818181818181817, "grad_norm": 13.282700538635254, "learning_rate": 5.8694555712879905e-06, "loss": 2.0333, "step": 22736 }, { "epoch": 1.781896551724138, "grad_norm": 3.4139795303344727, "learning_rate": 5.8691519055072135e-06, "loss": 0.9506, "step": 22737 }, { "epoch": 1.781974921630094, "grad_norm": 7.657077789306641, "learning_rate": 5.868848236420625e-06, "loss": 1.386, "step": 22738 }, { "epoch": 1.7820532915360503, "grad_norm": 2.638814926147461, "learning_rate": 5.868544564029388e-06, "loss": 1.4254, "step": 22739 }, { "epoch": 1.7821316614420062, "grad_norm": 3.312014102935791, "learning_rate": 5.8682408883346535e-06, "loss": 1.5041, "step": 22740 }, { "epoch": 1.7822100313479625, "grad_norm": 3.1378531455993652, "learning_rate": 5.867937209337574e-06, "loss": 1.3579, "step": 22741 }, { "epoch": 1.7822884012539184, "grad_norm": 2.375108480453491, "learning_rate": 5.867633527039309e-06, "loss": 1.2495, "step": 22742 }, { "epoch": 1.7823667711598747, "grad_norm": 4.569724082946777, "learning_rate": 5.867329841441011e-06, "loss": 1.123, "step": 22743 }, { "epoch": 1.7824451410658306, "grad_norm": 2.9656753540039062, "learning_rate": 5.8670261525438366e-06, "loss": 1.4629, "step": 22744 }, { "epoch": 1.782523510971787, "grad_norm": 3.708371639251709, "learning_rate": 5.8667224603489384e-06, "loss": 1.0584, "step": 22745 }, { "epoch": 1.7826018808777429, "grad_norm": 2.6511759757995605, "learning_rate": 5.866418764857475e-06, "loss": 1.5288, "step": 22746 }, { "epoch": 1.7826802507836992, "grad_norm": 3.9057953357696533, "learning_rate": 5.8661150660705975e-06, "loss": 1.4813, "step": 22747 }, { "epoch": 1.782758620689655, "grad_norm": 2.911726951599121, "learning_rate": 5.865811363989465e-06, "loss": 1.5954, "step": 22748 }, { "epoch": 1.7828369905956114, "grad_norm": 9.552093505859375, "learning_rate": 5.86550765861523e-06, "loss": 1.2445, "step": 22749 }, { "epoch": 1.7829153605015673, "grad_norm": 2.09431791305542, "learning_rate": 5.86520394994905e-06, "loss": 1.2391, "step": 22750 }, { "epoch": 1.7829937304075236, "grad_norm": 3.845529317855835, "learning_rate": 5.864900237992076e-06, "loss": 1.6378, "step": 22751 }, { "epoch": 1.7830721003134795, "grad_norm": 2.5864853858947754, "learning_rate": 5.864596522745467e-06, "loss": 1.1021, "step": 22752 }, { "epoch": 1.7831504702194358, "grad_norm": 2.985586643218994, "learning_rate": 5.864292804210379e-06, "loss": 1.2682, "step": 22753 }, { "epoch": 1.7832288401253917, "grad_norm": 2.5398738384246826, "learning_rate": 5.863989082387961e-06, "loss": 1.3784, "step": 22754 }, { "epoch": 1.783307210031348, "grad_norm": 2.7913477420806885, "learning_rate": 5.863685357279375e-06, "loss": 1.2809, "step": 22755 }, { "epoch": 1.783385579937304, "grad_norm": 2.566384792327881, "learning_rate": 5.863381628885772e-06, "loss": 1.0276, "step": 22756 }, { "epoch": 1.7834639498432603, "grad_norm": 2.641305923461914, "learning_rate": 5.86307789720831e-06, "loss": 1.0925, "step": 22757 }, { "epoch": 1.7835423197492162, "grad_norm": 2.049847364425659, "learning_rate": 5.862774162248143e-06, "loss": 1.4819, "step": 22758 }, { "epoch": 1.7836206896551725, "grad_norm": 3.8020107746124268, "learning_rate": 5.862470424006425e-06, "loss": 1.4342, "step": 22759 }, { "epoch": 1.7836990595611284, "grad_norm": 3.2229533195495605, "learning_rate": 5.8621666824843136e-06, "loss": 1.3487, "step": 22760 }, { "epoch": 1.7837774294670847, "grad_norm": 2.8812310695648193, "learning_rate": 5.861862937682963e-06, "loss": 1.1637, "step": 22761 }, { "epoch": 1.7838557993730406, "grad_norm": 3.7660129070281982, "learning_rate": 5.861559189603527e-06, "loss": 1.3871, "step": 22762 }, { "epoch": 1.783934169278997, "grad_norm": 2.4385526180267334, "learning_rate": 5.861255438247165e-06, "loss": 1.4082, "step": 22763 }, { "epoch": 1.7840125391849528, "grad_norm": 3.1920063495635986, "learning_rate": 5.860951683615027e-06, "loss": 1.3108, "step": 22764 }, { "epoch": 1.7840909090909092, "grad_norm": 12.35398006439209, "learning_rate": 5.860647925708272e-06, "loss": 2.1575, "step": 22765 }, { "epoch": 1.784169278996865, "grad_norm": 3.1670215129852295, "learning_rate": 5.860344164528055e-06, "loss": 1.5536, "step": 22766 }, { "epoch": 1.7842476489028214, "grad_norm": 1.7713781595230103, "learning_rate": 5.860040400075531e-06, "loss": 1.2788, "step": 22767 }, { "epoch": 1.7843260188087773, "grad_norm": 2.5588018894195557, "learning_rate": 5.859736632351853e-06, "loss": 1.6193, "step": 22768 }, { "epoch": 1.7844043887147336, "grad_norm": 2.6345224380493164, "learning_rate": 5.85943286135818e-06, "loss": 1.4738, "step": 22769 }, { "epoch": 1.7844827586206895, "grad_norm": 3.9452807903289795, "learning_rate": 5.859129087095665e-06, "loss": 1.5719, "step": 22770 }, { "epoch": 1.7845611285266458, "grad_norm": 4.827881336212158, "learning_rate": 5.8588253095654656e-06, "loss": 1.5599, "step": 22771 }, { "epoch": 1.7846394984326017, "grad_norm": 4.544920444488525, "learning_rate": 5.858521528768736e-06, "loss": 1.2368, "step": 22772 }, { "epoch": 1.784717868338558, "grad_norm": 6.182962417602539, "learning_rate": 5.858217744706629e-06, "loss": 1.3425, "step": 22773 }, { "epoch": 1.784796238244514, "grad_norm": 2.4071202278137207, "learning_rate": 5.857913957380304e-06, "loss": 1.7091, "step": 22774 }, { "epoch": 1.7848746081504703, "grad_norm": 6.085063934326172, "learning_rate": 5.857610166790915e-06, "loss": 1.8805, "step": 22775 }, { "epoch": 1.7849529780564264, "grad_norm": 2.210055112838745, "learning_rate": 5.8573063729396175e-06, "loss": 1.3323, "step": 22776 }, { "epoch": 1.7850313479623825, "grad_norm": 3.5664048194885254, "learning_rate": 5.857002575827567e-06, "loss": 1.1408, "step": 22777 }, { "epoch": 1.7851097178683386, "grad_norm": 2.6681125164031982, "learning_rate": 5.856698775455918e-06, "loss": 1.0536, "step": 22778 }, { "epoch": 1.7851880877742947, "grad_norm": 2.723288059234619, "learning_rate": 5.8563949718258275e-06, "loss": 1.3258, "step": 22779 }, { "epoch": 1.7852664576802508, "grad_norm": 2.753429651260376, "learning_rate": 5.856091164938451e-06, "loss": 1.1006, "step": 22780 }, { "epoch": 1.785344827586207, "grad_norm": 4.067387104034424, "learning_rate": 5.855787354794944e-06, "loss": 1.2141, "step": 22781 }, { "epoch": 1.785423197492163, "grad_norm": 2.7431366443634033, "learning_rate": 5.8554835413964584e-06, "loss": 1.3484, "step": 22782 }, { "epoch": 1.7855015673981192, "grad_norm": 3.9469947814941406, "learning_rate": 5.855179724744156e-06, "loss": 1.1261, "step": 22783 }, { "epoch": 1.7855799373040753, "grad_norm": 1.9810059070587158, "learning_rate": 5.854875904839188e-06, "loss": 1.3342, "step": 22784 }, { "epoch": 1.7856583072100314, "grad_norm": 2.6799962520599365, "learning_rate": 5.85457208168271e-06, "loss": 1.3975, "step": 22785 }, { "epoch": 1.7857366771159875, "grad_norm": 5.1710357666015625, "learning_rate": 5.85426825527588e-06, "loss": 1.482, "step": 22786 }, { "epoch": 1.7858150470219436, "grad_norm": 3.244117498397827, "learning_rate": 5.853964425619852e-06, "loss": 1.4021, "step": 22787 }, { "epoch": 1.7858934169278997, "grad_norm": 6.544559478759766, "learning_rate": 5.8536605927157815e-06, "loss": 1.4139, "step": 22788 }, { "epoch": 1.7859717868338558, "grad_norm": 2.9447102546691895, "learning_rate": 5.853356756564825e-06, "loss": 1.3255, "step": 22789 }, { "epoch": 1.786050156739812, "grad_norm": 2.86651349067688, "learning_rate": 5.853052917168139e-06, "loss": 1.48, "step": 22790 }, { "epoch": 1.786128526645768, "grad_norm": 4.158804893493652, "learning_rate": 5.852749074526878e-06, "loss": 1.2932, "step": 22791 }, { "epoch": 1.7862068965517242, "grad_norm": 5.486031532287598, "learning_rate": 5.852445228642194e-06, "loss": 1.52, "step": 22792 }, { "epoch": 1.7862852664576803, "grad_norm": 6.811756134033203, "learning_rate": 5.852141379515249e-06, "loss": 1.8292, "step": 22793 }, { "epoch": 1.7863636363636364, "grad_norm": 2.8992366790771484, "learning_rate": 5.851837527147195e-06, "loss": 1.6841, "step": 22794 }, { "epoch": 1.7864420062695925, "grad_norm": 3.0526340007781982, "learning_rate": 5.85153367153919e-06, "loss": 1.5897, "step": 22795 }, { "epoch": 1.7865203761755486, "grad_norm": 3.020641565322876, "learning_rate": 5.851229812692387e-06, "loss": 1.453, "step": 22796 }, { "epoch": 1.7865987460815047, "grad_norm": 1.988101840019226, "learning_rate": 5.850925950607943e-06, "loss": 1.3627, "step": 22797 }, { "epoch": 1.7866771159874608, "grad_norm": 2.0890913009643555, "learning_rate": 5.8506220852870145e-06, "loss": 1.3735, "step": 22798 }, { "epoch": 1.786755485893417, "grad_norm": 2.8943891525268555, "learning_rate": 5.8503182167307545e-06, "loss": 1.132, "step": 22799 }, { "epoch": 1.786833855799373, "grad_norm": 2.67128849029541, "learning_rate": 5.850014344940324e-06, "loss": 1.1169, "step": 22800 }, { "epoch": 1.7869122257053291, "grad_norm": 9.056537628173828, "learning_rate": 5.8497104699168715e-06, "loss": 1.5445, "step": 22801 }, { "epoch": 1.7869905956112853, "grad_norm": 9.139581680297852, "learning_rate": 5.849406591661561e-06, "loss": 1.1104, "step": 22802 }, { "epoch": 1.7870689655172414, "grad_norm": 2.8952276706695557, "learning_rate": 5.849102710175541e-06, "loss": 1.4796, "step": 22803 }, { "epoch": 1.7871473354231975, "grad_norm": 2.0112133026123047, "learning_rate": 5.848798825459973e-06, "loss": 1.4725, "step": 22804 }, { "epoch": 1.7872257053291536, "grad_norm": 3.1171975135803223, "learning_rate": 5.848494937516008e-06, "loss": 1.3629, "step": 22805 }, { "epoch": 1.7873040752351097, "grad_norm": 3.1166789531707764, "learning_rate": 5.848191046344804e-06, "loss": 1.4836, "step": 22806 }, { "epoch": 1.7873824451410658, "grad_norm": 4.7972187995910645, "learning_rate": 5.847887151947518e-06, "loss": 1.3606, "step": 22807 }, { "epoch": 1.787460815047022, "grad_norm": 6.614680767059326, "learning_rate": 5.847583254325304e-06, "loss": 0.898, "step": 22808 }, { "epoch": 1.787539184952978, "grad_norm": 1.9320191144943237, "learning_rate": 5.8472793534793206e-06, "loss": 1.368, "step": 22809 }, { "epoch": 1.7876175548589341, "grad_norm": 2.91925311088562, "learning_rate": 5.8469754494107215e-06, "loss": 1.778, "step": 22810 }, { "epoch": 1.7876959247648903, "grad_norm": 4.214756965637207, "learning_rate": 5.846671542120661e-06, "loss": 1.1294, "step": 22811 }, { "epoch": 1.7877742946708464, "grad_norm": 3.542607069015503, "learning_rate": 5.846367631610297e-06, "loss": 1.1666, "step": 22812 }, { "epoch": 1.7878526645768025, "grad_norm": 2.0520105361938477, "learning_rate": 5.846063717880786e-06, "loss": 1.4008, "step": 22813 }, { "epoch": 1.7879310344827586, "grad_norm": 4.487125396728516, "learning_rate": 5.845759800933283e-06, "loss": 1.578, "step": 22814 }, { "epoch": 1.7880094043887147, "grad_norm": 3.241201639175415, "learning_rate": 5.8454558807689435e-06, "loss": 1.1436, "step": 22815 }, { "epoch": 1.7880877742946708, "grad_norm": 2.8495137691497803, "learning_rate": 5.845151957388925e-06, "loss": 1.864, "step": 22816 }, { "epoch": 1.788166144200627, "grad_norm": 3.777362585067749, "learning_rate": 5.844848030794381e-06, "loss": 1.2056, "step": 22817 }, { "epoch": 1.788244514106583, "grad_norm": 4.039756774902344, "learning_rate": 5.84454410098647e-06, "loss": 1.2984, "step": 22818 }, { "epoch": 1.7883228840125391, "grad_norm": 3.4331045150756836, "learning_rate": 5.844240167966348e-06, "loss": 0.779, "step": 22819 }, { "epoch": 1.7884012539184952, "grad_norm": 2.800899028778076, "learning_rate": 5.843936231735169e-06, "loss": 1.5852, "step": 22820 }, { "epoch": 1.7884796238244514, "grad_norm": 3.2148420810699463, "learning_rate": 5.843632292294092e-06, "loss": 1.1385, "step": 22821 }, { "epoch": 1.7885579937304075, "grad_norm": 4.426966667175293, "learning_rate": 5.843328349644268e-06, "loss": 1.1079, "step": 22822 }, { "epoch": 1.7886363636363636, "grad_norm": 5.202485084533691, "learning_rate": 5.8430244037868575e-06, "loss": 1.4478, "step": 22823 }, { "epoch": 1.78871473354232, "grad_norm": 5.638864994049072, "learning_rate": 5.842720454723015e-06, "loss": 1.1393, "step": 22824 }, { "epoch": 1.7887931034482758, "grad_norm": 3.7978007793426514, "learning_rate": 5.842416502453897e-06, "loss": 1.6282, "step": 22825 }, { "epoch": 1.7888714733542321, "grad_norm": 4.2636590003967285, "learning_rate": 5.842112546980661e-06, "loss": 1.692, "step": 22826 }, { "epoch": 1.788949843260188, "grad_norm": 2.3406310081481934, "learning_rate": 5.841808588304458e-06, "loss": 1.3948, "step": 22827 }, { "epoch": 1.7890282131661444, "grad_norm": 4.541144371032715, "learning_rate": 5.841504626426449e-06, "loss": 1.4779, "step": 22828 }, { "epoch": 1.7891065830721002, "grad_norm": 1.923925518989563, "learning_rate": 5.841200661347789e-06, "loss": 1.0576, "step": 22829 }, { "epoch": 1.7891849529780566, "grad_norm": 2.6496660709381104, "learning_rate": 5.840896693069635e-06, "loss": 1.4094, "step": 22830 }, { "epoch": 1.7892633228840125, "grad_norm": 2.72906494140625, "learning_rate": 5.840592721593139e-06, "loss": 1.3183, "step": 22831 }, { "epoch": 1.7893416927899688, "grad_norm": 2.3884809017181396, "learning_rate": 5.840288746919461e-06, "loss": 1.2908, "step": 22832 }, { "epoch": 1.7894200626959247, "grad_norm": 4.03427791595459, "learning_rate": 5.839984769049758e-06, "loss": 1.2271, "step": 22833 }, { "epoch": 1.789498432601881, "grad_norm": 3.565778970718384, "learning_rate": 5.839680787985182e-06, "loss": 1.1302, "step": 22834 }, { "epoch": 1.789576802507837, "grad_norm": 4.786384582519531, "learning_rate": 5.8393768037268925e-06, "loss": 1.4255, "step": 22835 }, { "epoch": 1.7896551724137932, "grad_norm": 14.08421516418457, "learning_rate": 5.839072816276046e-06, "loss": 0.7533, "step": 22836 }, { "epoch": 1.7897335423197491, "grad_norm": 2.1849582195281982, "learning_rate": 5.838768825633796e-06, "loss": 1.1707, "step": 22837 }, { "epoch": 1.7898119122257055, "grad_norm": 3.8637099266052246, "learning_rate": 5.8384648318013e-06, "loss": 1.2158, "step": 22838 }, { "epoch": 1.7898902821316613, "grad_norm": 3.47829270362854, "learning_rate": 5.838160834779716e-06, "loss": 1.2202, "step": 22839 }, { "epoch": 1.7899686520376177, "grad_norm": 14.89626693725586, "learning_rate": 5.837856834570197e-06, "loss": 1.3364, "step": 22840 }, { "epoch": 1.7900470219435736, "grad_norm": 2.561129570007324, "learning_rate": 5.837552831173901e-06, "loss": 1.4909, "step": 22841 }, { "epoch": 1.79012539184953, "grad_norm": 2.2091493606567383, "learning_rate": 5.837248824591985e-06, "loss": 1.1767, "step": 22842 }, { "epoch": 1.7902037617554858, "grad_norm": 2.9066851139068604, "learning_rate": 5.836944814825605e-06, "loss": 1.3686, "step": 22843 }, { "epoch": 1.7902821316614421, "grad_norm": 5.421438694000244, "learning_rate": 5.836640801875917e-06, "loss": 1.3752, "step": 22844 }, { "epoch": 1.790360501567398, "grad_norm": 2.760983467102051, "learning_rate": 5.836336785744076e-06, "loss": 1.1428, "step": 22845 }, { "epoch": 1.7904388714733543, "grad_norm": 7.844782829284668, "learning_rate": 5.836032766431239e-06, "loss": 1.9201, "step": 22846 }, { "epoch": 1.7905172413793102, "grad_norm": 2.2071101665496826, "learning_rate": 5.835728743938565e-06, "loss": 1.2084, "step": 22847 }, { "epoch": 1.7905956112852666, "grad_norm": 3.8217875957489014, "learning_rate": 5.835424718267205e-06, "loss": 1.474, "step": 22848 }, { "epoch": 1.7906739811912225, "grad_norm": 2.881380319595337, "learning_rate": 5.835120689418321e-06, "loss": 1.558, "step": 22849 }, { "epoch": 1.7907523510971788, "grad_norm": 2.765040636062622, "learning_rate": 5.834816657393067e-06, "loss": 1.5538, "step": 22850 }, { "epoch": 1.7908307210031347, "grad_norm": 4.318172454833984, "learning_rate": 5.8345126221925976e-06, "loss": 1.4465, "step": 22851 }, { "epoch": 1.790909090909091, "grad_norm": 5.276190757751465, "learning_rate": 5.8342085838180715e-06, "loss": 1.7469, "step": 22852 }, { "epoch": 1.790987460815047, "grad_norm": 3.7399802207946777, "learning_rate": 5.833904542270643e-06, "loss": 1.0932, "step": 22853 }, { "epoch": 1.7910658307210032, "grad_norm": 2.608577251434326, "learning_rate": 5.833600497551473e-06, "loss": 1.3207, "step": 22854 }, { "epoch": 1.7911442006269591, "grad_norm": 2.3971364498138428, "learning_rate": 5.833296449661711e-06, "loss": 1.2114, "step": 22855 }, { "epoch": 1.7912225705329154, "grad_norm": 2.366140127182007, "learning_rate": 5.832992398602521e-06, "loss": 0.986, "step": 22856 }, { "epoch": 1.7913009404388713, "grad_norm": 3.190887212753296, "learning_rate": 5.8326883443750545e-06, "loss": 1.4663, "step": 22857 }, { "epoch": 1.7913793103448277, "grad_norm": 3.7652807235717773, "learning_rate": 5.832384286980468e-06, "loss": 1.8267, "step": 22858 }, { "epoch": 1.7914576802507836, "grad_norm": 2.179182529449463, "learning_rate": 5.8320802264199205e-06, "loss": 1.2202, "step": 22859 }, { "epoch": 1.79153605015674, "grad_norm": 2.0893120765686035, "learning_rate": 5.8317761626945665e-06, "loss": 1.2509, "step": 22860 }, { "epoch": 1.7916144200626958, "grad_norm": 4.180143356323242, "learning_rate": 5.831472095805564e-06, "loss": 1.6046, "step": 22861 }, { "epoch": 1.791692789968652, "grad_norm": 5.873291969299316, "learning_rate": 5.831168025754068e-06, "loss": 1.3236, "step": 22862 }, { "epoch": 1.791771159874608, "grad_norm": 2.9024956226348877, "learning_rate": 5.8308639525412355e-06, "loss": 1.476, "step": 22863 }, { "epoch": 1.7918495297805643, "grad_norm": 2.8617641925811768, "learning_rate": 5.830559876168226e-06, "loss": 1.0006, "step": 22864 }, { "epoch": 1.7919278996865202, "grad_norm": 2.555516004562378, "learning_rate": 5.83025579663619e-06, "loss": 1.2078, "step": 22865 }, { "epoch": 1.7920062695924766, "grad_norm": 3.084355115890503, "learning_rate": 5.829951713946289e-06, "loss": 1.7601, "step": 22866 }, { "epoch": 1.7920846394984324, "grad_norm": 3.0475473403930664, "learning_rate": 5.829647628099678e-06, "loss": 1.7643, "step": 22867 }, { "epoch": 1.7921630094043888, "grad_norm": 3.102189779281616, "learning_rate": 5.829343539097512e-06, "loss": 1.6184, "step": 22868 }, { "epoch": 1.7922413793103447, "grad_norm": 2.8201069831848145, "learning_rate": 5.829039446940951e-06, "loss": 1.3761, "step": 22869 }, { "epoch": 1.792319749216301, "grad_norm": 2.5750062465667725, "learning_rate": 5.828735351631149e-06, "loss": 1.2018, "step": 22870 }, { "epoch": 1.792398119122257, "grad_norm": 2.6188366413116455, "learning_rate": 5.828431253169262e-06, "loss": 1.2903, "step": 22871 }, { "epoch": 1.7924764890282132, "grad_norm": 3.5549612045288086, "learning_rate": 5.82812715155645e-06, "loss": 1.8122, "step": 22872 }, { "epoch": 1.7925548589341693, "grad_norm": 4.820079326629639, "learning_rate": 5.8278230467938675e-06, "loss": 1.15, "step": 22873 }, { "epoch": 1.7926332288401254, "grad_norm": 4.548864364624023, "learning_rate": 5.8275189388826706e-06, "loss": 1.828, "step": 22874 }, { "epoch": 1.7927115987460815, "grad_norm": 7.7469072341918945, "learning_rate": 5.827214827824017e-06, "loss": 1.3094, "step": 22875 }, { "epoch": 1.7927899686520377, "grad_norm": 3.6620516777038574, "learning_rate": 5.826910713619064e-06, "loss": 1.2535, "step": 22876 }, { "epoch": 1.7928683385579938, "grad_norm": 2.463122606277466, "learning_rate": 5.826606596268965e-06, "loss": 1.0069, "step": 22877 }, { "epoch": 1.7929467084639499, "grad_norm": 2.6550614833831787, "learning_rate": 5.826302475774881e-06, "loss": 1.5464, "step": 22878 }, { "epoch": 1.793025078369906, "grad_norm": 4.315481185913086, "learning_rate": 5.825998352137967e-06, "loss": 1.2898, "step": 22879 }, { "epoch": 1.793103448275862, "grad_norm": 3.0387980937957764, "learning_rate": 5.82569422535938e-06, "loss": 1.391, "step": 22880 }, { "epoch": 1.7931818181818182, "grad_norm": 2.702788829803467, "learning_rate": 5.825390095440273e-06, "loss": 1.4077, "step": 22881 }, { "epoch": 1.7932601880877743, "grad_norm": 2.9042561054229736, "learning_rate": 5.825085962381809e-06, "loss": 1.1146, "step": 22882 }, { "epoch": 1.7933385579937304, "grad_norm": 2.453132390975952, "learning_rate": 5.82478182618514e-06, "loss": 1.4695, "step": 22883 }, { "epoch": 1.7934169278996865, "grad_norm": 3.3149662017822266, "learning_rate": 5.824477686851426e-06, "loss": 1.5829, "step": 22884 }, { "epoch": 1.7934952978056427, "grad_norm": 1.9726938009262085, "learning_rate": 5.824173544381823e-06, "loss": 1.1669, "step": 22885 }, { "epoch": 1.7935736677115988, "grad_norm": 2.895750045776367, "learning_rate": 5.823869398777485e-06, "loss": 1.4153, "step": 22886 }, { "epoch": 1.7936520376175549, "grad_norm": 2.9776899814605713, "learning_rate": 5.823565250039572e-06, "loss": 1.1777, "step": 22887 }, { "epoch": 1.793730407523511, "grad_norm": 2.7701892852783203, "learning_rate": 5.82326109816924e-06, "loss": 1.1894, "step": 22888 }, { "epoch": 1.793808777429467, "grad_norm": 3.5105316638946533, "learning_rate": 5.822956943167646e-06, "loss": 1.3848, "step": 22889 }, { "epoch": 1.7938871473354232, "grad_norm": 2.7988839149475098, "learning_rate": 5.822652785035945e-06, "loss": 1.8167, "step": 22890 }, { "epoch": 1.7939655172413793, "grad_norm": 3.877408027648926, "learning_rate": 5.822348623775296e-06, "loss": 1.2814, "step": 22891 }, { "epoch": 1.7940438871473354, "grad_norm": 3.424070119857788, "learning_rate": 5.822044459386856e-06, "loss": 0.7346, "step": 22892 }, { "epoch": 1.7941222570532915, "grad_norm": 2.428574323654175, "learning_rate": 5.821740291871781e-06, "loss": 1.5232, "step": 22893 }, { "epoch": 1.7942006269592476, "grad_norm": 7.15269660949707, "learning_rate": 5.821436121231227e-06, "loss": 1.571, "step": 22894 }, { "epoch": 1.7942789968652038, "grad_norm": 2.922403335571289, "learning_rate": 5.821131947466353e-06, "loss": 0.9715, "step": 22895 }, { "epoch": 1.7943573667711599, "grad_norm": 6.277759552001953, "learning_rate": 5.820827770578314e-06, "loss": 1.0037, "step": 22896 }, { "epoch": 1.794435736677116, "grad_norm": 2.481635332107544, "learning_rate": 5.8205235905682674e-06, "loss": 1.2262, "step": 22897 }, { "epoch": 1.794514106583072, "grad_norm": 2.763247489929199, "learning_rate": 5.820219407437372e-06, "loss": 1.6882, "step": 22898 }, { "epoch": 1.7945924764890282, "grad_norm": 6.168356418609619, "learning_rate": 5.8199152211867835e-06, "loss": 1.3543, "step": 22899 }, { "epoch": 1.7946708463949843, "grad_norm": 1.962951898574829, "learning_rate": 5.819611031817657e-06, "loss": 1.3894, "step": 22900 }, { "epoch": 1.7947492163009404, "grad_norm": 3.880080461502075, "learning_rate": 5.819306839331153e-06, "loss": 1.3309, "step": 22901 }, { "epoch": 1.7948275862068965, "grad_norm": 3.0954954624176025, "learning_rate": 5.819002643728423e-06, "loss": 1.1729, "step": 22902 }, { "epoch": 1.7949059561128526, "grad_norm": 2.659188747406006, "learning_rate": 5.818698445010631e-06, "loss": 1.1932, "step": 22903 }, { "epoch": 1.7949843260188088, "grad_norm": 3.2194762229919434, "learning_rate": 5.818394243178929e-06, "loss": 1.3361, "step": 22904 }, { "epoch": 1.7950626959247649, "grad_norm": 14.806975364685059, "learning_rate": 5.818090038234476e-06, "loss": 1.2159, "step": 22905 }, { "epoch": 1.795141065830721, "grad_norm": 3.0196545124053955, "learning_rate": 5.817785830178431e-06, "loss": 1.4224, "step": 22906 }, { "epoch": 1.795219435736677, "grad_norm": 2.5804367065429688, "learning_rate": 5.817481619011946e-06, "loss": 1.7914, "step": 22907 }, { "epoch": 1.7952978056426332, "grad_norm": 3.5204408168792725, "learning_rate": 5.8171774047361815e-06, "loss": 1.0537, "step": 22908 }, { "epoch": 1.7953761755485893, "grad_norm": 3.4526634216308594, "learning_rate": 5.8168731873522924e-06, "loss": 1.7526, "step": 22909 }, { "epoch": 1.7954545454545454, "grad_norm": 1.8586070537567139, "learning_rate": 5.81656896686144e-06, "loss": 1.2174, "step": 22910 }, { "epoch": 1.7955329153605015, "grad_norm": 2.6543617248535156, "learning_rate": 5.816264743264777e-06, "loss": 1.1916, "step": 22911 }, { "epoch": 1.7956112852664576, "grad_norm": 2.664809226989746, "learning_rate": 5.815960516563463e-06, "loss": 1.3398, "step": 22912 }, { "epoch": 1.7956896551724137, "grad_norm": 5.767043590545654, "learning_rate": 5.815656286758653e-06, "loss": 0.8258, "step": 22913 }, { "epoch": 1.7957680250783699, "grad_norm": 2.427968978881836, "learning_rate": 5.815352053851506e-06, "loss": 0.9594, "step": 22914 }, { "epoch": 1.795846394984326, "grad_norm": 3.470196008682251, "learning_rate": 5.815047817843179e-06, "loss": 1.7152, "step": 22915 }, { "epoch": 1.795924764890282, "grad_norm": 2.3865883350372314, "learning_rate": 5.81474357873483e-06, "loss": 1.276, "step": 22916 }, { "epoch": 1.7960031347962384, "grad_norm": 13.401872634887695, "learning_rate": 5.814439336527612e-06, "loss": 0.4245, "step": 22917 }, { "epoch": 1.7960815047021943, "grad_norm": 2.6380882263183594, "learning_rate": 5.814135091222688e-06, "loss": 1.2565, "step": 22918 }, { "epoch": 1.7961598746081506, "grad_norm": 2.66218900680542, "learning_rate": 5.8138308428212095e-06, "loss": 1.4942, "step": 22919 }, { "epoch": 1.7962382445141065, "grad_norm": 3.9183545112609863, "learning_rate": 5.813526591324339e-06, "loss": 1.5915, "step": 22920 }, { "epoch": 1.7963166144200629, "grad_norm": 3.0696861743927, "learning_rate": 5.81322233673323e-06, "loss": 1.4307, "step": 22921 }, { "epoch": 1.7963949843260187, "grad_norm": 2.101773738861084, "learning_rate": 5.8129180790490406e-06, "loss": 1.1583, "step": 22922 }, { "epoch": 1.796473354231975, "grad_norm": 2.7017812728881836, "learning_rate": 5.812613818272928e-06, "loss": 1.414, "step": 22923 }, { "epoch": 1.796551724137931, "grad_norm": 3.507812738418579, "learning_rate": 5.8123095544060515e-06, "loss": 1.0122, "step": 22924 }, { "epoch": 1.7966300940438873, "grad_norm": 2.758570671081543, "learning_rate": 5.812005287449566e-06, "loss": 1.4562, "step": 22925 }, { "epoch": 1.7967084639498432, "grad_norm": 3.044161081314087, "learning_rate": 5.811701017404628e-06, "loss": 1.2035, "step": 22926 }, { "epoch": 1.7967868338557995, "grad_norm": 4.88434362411499, "learning_rate": 5.8113967442724e-06, "loss": 1.3463, "step": 22927 }, { "epoch": 1.7968652037617554, "grad_norm": 3.558777093887329, "learning_rate": 5.81109246805403e-06, "loss": 1.2202, "step": 22928 }, { "epoch": 1.7969435736677117, "grad_norm": 2.890312433242798, "learning_rate": 5.810788188750686e-06, "loss": 1.4053, "step": 22929 }, { "epoch": 1.7970219435736676, "grad_norm": 2.6582233905792236, "learning_rate": 5.8104839063635164e-06, "loss": 0.9613, "step": 22930 }, { "epoch": 1.797100313479624, "grad_norm": 4.504414081573486, "learning_rate": 5.810179620893686e-06, "loss": 1.2742, "step": 22931 }, { "epoch": 1.7971786833855798, "grad_norm": 2.8102049827575684, "learning_rate": 5.809875332342345e-06, "loss": 1.0558, "step": 22932 }, { "epoch": 1.7972570532915362, "grad_norm": 3.6845579147338867, "learning_rate": 5.809571040710655e-06, "loss": 1.404, "step": 22933 }, { "epoch": 1.797335423197492, "grad_norm": 2.630342483520508, "learning_rate": 5.809266745999775e-06, "loss": 1.2082, "step": 22934 }, { "epoch": 1.7974137931034484, "grad_norm": 2.9968135356903076, "learning_rate": 5.808962448210858e-06, "loss": 1.5088, "step": 22935 }, { "epoch": 1.7974921630094043, "grad_norm": 3.9446122646331787, "learning_rate": 5.808658147345063e-06, "loss": 1.0416, "step": 22936 }, { "epoch": 1.7975705329153606, "grad_norm": 2.311232566833496, "learning_rate": 5.8083538434035465e-06, "loss": 1.6201, "step": 22937 }, { "epoch": 1.7976489028213165, "grad_norm": 2.086106300354004, "learning_rate": 5.808049536387471e-06, "loss": 1.0017, "step": 22938 }, { "epoch": 1.7977272727272728, "grad_norm": 2.8542022705078125, "learning_rate": 5.8077452262979886e-06, "loss": 1.3264, "step": 22939 }, { "epoch": 1.7978056426332287, "grad_norm": 1.76553475856781, "learning_rate": 5.807440913136257e-06, "loss": 1.3083, "step": 22940 }, { "epoch": 1.797884012539185, "grad_norm": 3.5561118125915527, "learning_rate": 5.807136596903435e-06, "loss": 1.6011, "step": 22941 }, { "epoch": 1.797962382445141, "grad_norm": 4.724587440490723, "learning_rate": 5.80683227760068e-06, "loss": 1.4929, "step": 22942 }, { "epoch": 1.7980407523510973, "grad_norm": 7.887779712677002, "learning_rate": 5.806527955229151e-06, "loss": 1.4726, "step": 22943 }, { "epoch": 1.7981191222570532, "grad_norm": 13.729771614074707, "learning_rate": 5.806223629790002e-06, "loss": 1.082, "step": 22944 }, { "epoch": 1.7981974921630095, "grad_norm": 11.111054420471191, "learning_rate": 5.805919301284394e-06, "loss": 1.0549, "step": 22945 }, { "epoch": 1.7982758620689654, "grad_norm": 6.687718391418457, "learning_rate": 5.805614969713483e-06, "loss": 1.8465, "step": 22946 }, { "epoch": 1.7983542319749217, "grad_norm": 2.5958828926086426, "learning_rate": 5.805310635078425e-06, "loss": 1.4068, "step": 22947 }, { "epoch": 1.7984326018808776, "grad_norm": 2.5281407833099365, "learning_rate": 5.80500629738038e-06, "loss": 1.2925, "step": 22948 }, { "epoch": 1.798510971786834, "grad_norm": 3.8945512771606445, "learning_rate": 5.8047019566205046e-06, "loss": 1.2389, "step": 22949 }, { "epoch": 1.7985893416927898, "grad_norm": 3.4737319946289062, "learning_rate": 5.804397612799956e-06, "loss": 1.6928, "step": 22950 }, { "epoch": 1.7986677115987462, "grad_norm": 2.393016815185547, "learning_rate": 5.804093265919891e-06, "loss": 0.9175, "step": 22951 }, { "epoch": 1.798746081504702, "grad_norm": 3.000471591949463, "learning_rate": 5.80378891598147e-06, "loss": 1.8952, "step": 22952 }, { "epoch": 1.7988244514106584, "grad_norm": 2.370319366455078, "learning_rate": 5.803484562985848e-06, "loss": 1.5424, "step": 22953 }, { "epoch": 1.7989028213166143, "grad_norm": 2.8146982192993164, "learning_rate": 5.803180206934183e-06, "loss": 1.18, "step": 22954 }, { "epoch": 1.7989811912225706, "grad_norm": 2.3982045650482178, "learning_rate": 5.8028758478276335e-06, "loss": 1.409, "step": 22955 }, { "epoch": 1.7990595611285265, "grad_norm": 2.759737730026245, "learning_rate": 5.802571485667355e-06, "loss": 1.4698, "step": 22956 }, { "epoch": 1.7991379310344828, "grad_norm": 3.4891157150268555, "learning_rate": 5.80226712045451e-06, "loss": 0.8072, "step": 22957 }, { "epoch": 1.7992163009404387, "grad_norm": 2.442293405532837, "learning_rate": 5.801962752190251e-06, "loss": 1.3656, "step": 22958 }, { "epoch": 1.799294670846395, "grad_norm": 4.070591926574707, "learning_rate": 5.801658380875737e-06, "loss": 0.9701, "step": 22959 }, { "epoch": 1.799373040752351, "grad_norm": 2.8248603343963623, "learning_rate": 5.801354006512127e-06, "loss": 1.6472, "step": 22960 }, { "epoch": 1.7994514106583073, "grad_norm": 2.3230321407318115, "learning_rate": 5.801049629100579e-06, "loss": 1.2694, "step": 22961 }, { "epoch": 1.7995297805642632, "grad_norm": 2.548804998397827, "learning_rate": 5.80074524864225e-06, "loss": 1.4265, "step": 22962 }, { "epoch": 1.7996081504702195, "grad_norm": 3.2610442638397217, "learning_rate": 5.800440865138295e-06, "loss": 1.6984, "step": 22963 }, { "epoch": 1.7996865203761756, "grad_norm": 1.688718318939209, "learning_rate": 5.8001364785898764e-06, "loss": 1.2949, "step": 22964 }, { "epoch": 1.7997648902821317, "grad_norm": 2.1599748134613037, "learning_rate": 5.799832088998147e-06, "loss": 0.9151, "step": 22965 }, { "epoch": 1.7998432601880878, "grad_norm": 5.461043834686279, "learning_rate": 5.79952769636427e-06, "loss": 1.1779, "step": 22966 }, { "epoch": 1.799921630094044, "grad_norm": 6.834898471832275, "learning_rate": 5.799223300689398e-06, "loss": 0.7385, "step": 22967 }, { "epoch": 1.8, "grad_norm": 1.887016773223877, "learning_rate": 5.798918901974693e-06, "loss": 1.1129, "step": 22968 }, { "epoch": 1.8000783699059562, "grad_norm": 3.972571611404419, "learning_rate": 5.798614500221309e-06, "loss": 1.4782, "step": 22969 }, { "epoch": 1.8001567398119123, "grad_norm": 2.7341301441192627, "learning_rate": 5.798310095430406e-06, "loss": 1.3346, "step": 22970 }, { "epoch": 1.8002351097178684, "grad_norm": 3.1438167095184326, "learning_rate": 5.798005687603142e-06, "loss": 1.4487, "step": 22971 }, { "epoch": 1.8003134796238245, "grad_norm": 2.923401117324829, "learning_rate": 5.797701276740674e-06, "loss": 1.3601, "step": 22972 }, { "epoch": 1.8003918495297806, "grad_norm": 2.6156527996063232, "learning_rate": 5.79739686284416e-06, "loss": 1.377, "step": 22973 }, { "epoch": 1.8004702194357367, "grad_norm": 4.405646800994873, "learning_rate": 5.7970924459147595e-06, "loss": 1.2401, "step": 22974 }, { "epoch": 1.8005485893416928, "grad_norm": 1.8687688112258911, "learning_rate": 5.7967880259536255e-06, "loss": 1.3744, "step": 22975 }, { "epoch": 1.800626959247649, "grad_norm": 2.1183197498321533, "learning_rate": 5.796483602961922e-06, "loss": 1.3077, "step": 22976 }, { "epoch": 1.800705329153605, "grad_norm": 2.359574556350708, "learning_rate": 5.796179176940804e-06, "loss": 1.1993, "step": 22977 }, { "epoch": 1.8007836990595611, "grad_norm": 3.810854434967041, "learning_rate": 5.795874747891428e-06, "loss": 1.3318, "step": 22978 }, { "epoch": 1.8008620689655173, "grad_norm": 3.069708824157715, "learning_rate": 5.795570315814953e-06, "loss": 1.1284, "step": 22979 }, { "epoch": 1.8009404388714734, "grad_norm": 2.205850839614868, "learning_rate": 5.795265880712537e-06, "loss": 1.2186, "step": 22980 }, { "epoch": 1.8010188087774295, "grad_norm": 2.5724945068359375, "learning_rate": 5.7949614425853405e-06, "loss": 1.44, "step": 22981 }, { "epoch": 1.8010971786833856, "grad_norm": 2.0400912761688232, "learning_rate": 5.794657001434516e-06, "loss": 1.3836, "step": 22982 }, { "epoch": 1.8011755485893417, "grad_norm": 5.220730781555176, "learning_rate": 5.794352557261227e-06, "loss": 1.4985, "step": 22983 }, { "epoch": 1.8012539184952978, "grad_norm": 3.252885341644287, "learning_rate": 5.794048110066626e-06, "loss": 1.2546, "step": 22984 }, { "epoch": 1.801332288401254, "grad_norm": 4.030269145965576, "learning_rate": 5.793743659851876e-06, "loss": 1.6529, "step": 22985 }, { "epoch": 1.80141065830721, "grad_norm": 5.137441158294678, "learning_rate": 5.793439206618132e-06, "loss": 1.4076, "step": 22986 }, { "epoch": 1.8014890282131661, "grad_norm": 2.69240403175354, "learning_rate": 5.793134750366553e-06, "loss": 0.9185, "step": 22987 }, { "epoch": 1.8015673981191223, "grad_norm": 4.586521625518799, "learning_rate": 5.792830291098297e-06, "loss": 1.3208, "step": 22988 }, { "epoch": 1.8016457680250784, "grad_norm": 2.0663981437683105, "learning_rate": 5.792525828814521e-06, "loss": 1.4794, "step": 22989 }, { "epoch": 1.8017241379310345, "grad_norm": 1.892392635345459, "learning_rate": 5.792221363516386e-06, "loss": 1.2037, "step": 22990 }, { "epoch": 1.8018025078369906, "grad_norm": 2.558967351913452, "learning_rate": 5.791916895205045e-06, "loss": 1.439, "step": 22991 }, { "epoch": 1.8018808777429467, "grad_norm": 3.095452308654785, "learning_rate": 5.791612423881661e-06, "loss": 1.6154, "step": 22992 }, { "epoch": 1.8019592476489028, "grad_norm": 3.0979461669921875, "learning_rate": 5.791307949547389e-06, "loss": 1.2481, "step": 22993 }, { "epoch": 1.802037617554859, "grad_norm": 5.685220718383789, "learning_rate": 5.791003472203388e-06, "loss": 1.4322, "step": 22994 }, { "epoch": 1.802115987460815, "grad_norm": 4.752599239349365, "learning_rate": 5.790698991850816e-06, "loss": 1.2232, "step": 22995 }, { "epoch": 1.8021943573667711, "grad_norm": 2.959845781326294, "learning_rate": 5.790394508490831e-06, "loss": 0.9391, "step": 22996 }, { "epoch": 1.8022727272727272, "grad_norm": 5.915397644042969, "learning_rate": 5.790090022124591e-06, "loss": 1.0731, "step": 22997 }, { "epoch": 1.8023510971786834, "grad_norm": 3.2157068252563477, "learning_rate": 5.789785532753256e-06, "loss": 1.0691, "step": 22998 }, { "epoch": 1.8024294670846395, "grad_norm": 2.157076358795166, "learning_rate": 5.789481040377981e-06, "loss": 1.1142, "step": 22999 }, { "epoch": 1.8025078369905956, "grad_norm": 4.269254684448242, "learning_rate": 5.7891765449999275e-06, "loss": 1.3803, "step": 23000 }, { "epoch": 1.8025862068965517, "grad_norm": 2.0776052474975586, "learning_rate": 5.7888720466202485e-06, "loss": 1.6645, "step": 23001 }, { "epoch": 1.8026645768025078, "grad_norm": 2.736388921737671, "learning_rate": 5.788567545240108e-06, "loss": 1.2126, "step": 23002 }, { "epoch": 1.802742946708464, "grad_norm": 2.6203248500823975, "learning_rate": 5.78826304086066e-06, "loss": 1.0384, "step": 23003 }, { "epoch": 1.80282131661442, "grad_norm": 3.2858572006225586, "learning_rate": 5.787958533483066e-06, "loss": 1.4513, "step": 23004 }, { "epoch": 1.8028996865203761, "grad_norm": 4.011447906494141, "learning_rate": 5.787654023108481e-06, "loss": 1.5307, "step": 23005 }, { "epoch": 1.8029780564263322, "grad_norm": 3.6999382972717285, "learning_rate": 5.787349509738065e-06, "loss": 1.1593, "step": 23006 }, { "epoch": 1.8030564263322884, "grad_norm": 3.5892913341522217, "learning_rate": 5.787044993372978e-06, "loss": 1.0715, "step": 23007 }, { "epoch": 1.8031347962382445, "grad_norm": 3.1684489250183105, "learning_rate": 5.7867404740143725e-06, "loss": 1.3876, "step": 23008 }, { "epoch": 1.8032131661442006, "grad_norm": 2.3512680530548096, "learning_rate": 5.786435951663413e-06, "loss": 1.5056, "step": 23009 }, { "epoch": 1.8032915360501567, "grad_norm": 5.387323379516602, "learning_rate": 5.786131426321253e-06, "loss": 1.5237, "step": 23010 }, { "epoch": 1.8033699059561128, "grad_norm": 3.3795113563537598, "learning_rate": 5.785826897989055e-06, "loss": 1.5456, "step": 23011 }, { "epoch": 1.8034482758620691, "grad_norm": 4.727719783782959, "learning_rate": 5.7855223666679735e-06, "loss": 0.7037, "step": 23012 }, { "epoch": 1.803526645768025, "grad_norm": 3.9167165756225586, "learning_rate": 5.785217832359169e-06, "loss": 1.1426, "step": 23013 }, { "epoch": 1.8036050156739813, "grad_norm": 2.6113100051879883, "learning_rate": 5.7849132950638e-06, "loss": 1.4683, "step": 23014 }, { "epoch": 1.8036833855799372, "grad_norm": 4.92494010925293, "learning_rate": 5.784608754783022e-06, "loss": 1.0647, "step": 23015 }, { "epoch": 1.8037617554858936, "grad_norm": 3.50998854637146, "learning_rate": 5.784304211517997e-06, "loss": 1.6099, "step": 23016 }, { "epoch": 1.8038401253918495, "grad_norm": 5.873226642608643, "learning_rate": 5.783999665269881e-06, "loss": 1.0091, "step": 23017 }, { "epoch": 1.8039184952978058, "grad_norm": 2.655646562576294, "learning_rate": 5.783695116039833e-06, "loss": 1.3852, "step": 23018 }, { "epoch": 1.8039968652037617, "grad_norm": 3.343968391418457, "learning_rate": 5.7833905638290125e-06, "loss": 1.4627, "step": 23019 }, { "epoch": 1.804075235109718, "grad_norm": 2.9278838634490967, "learning_rate": 5.7830860086385746e-06, "loss": 1.2304, "step": 23020 }, { "epoch": 1.804153605015674, "grad_norm": 3.8741114139556885, "learning_rate": 5.782781450469681e-06, "loss": 1.0713, "step": 23021 }, { "epoch": 1.8042319749216302, "grad_norm": 2.4556355476379395, "learning_rate": 5.782476889323488e-06, "loss": 1.1236, "step": 23022 }, { "epoch": 1.8043103448275861, "grad_norm": 3.8791990280151367, "learning_rate": 5.782172325201155e-06, "loss": 1.1187, "step": 23023 }, { "epoch": 1.8043887147335425, "grad_norm": 2.2409958839416504, "learning_rate": 5.781867758103841e-06, "loss": 1.1642, "step": 23024 }, { "epoch": 1.8044670846394983, "grad_norm": 2.4478607177734375, "learning_rate": 5.781563188032703e-06, "loss": 1.3494, "step": 23025 }, { "epoch": 1.8045454545454547, "grad_norm": 6.135236740112305, "learning_rate": 5.781258614988899e-06, "loss": 1.7336, "step": 23026 }, { "epoch": 1.8046238244514106, "grad_norm": 3.2045907974243164, "learning_rate": 5.7809540389735896e-06, "loss": 1.6065, "step": 23027 }, { "epoch": 1.804702194357367, "grad_norm": 6.015456676483154, "learning_rate": 5.780649459987934e-06, "loss": 2.0581, "step": 23028 }, { "epoch": 1.8047805642633228, "grad_norm": 2.0801520347595215, "learning_rate": 5.780344878033086e-06, "loss": 1.3623, "step": 23029 }, { "epoch": 1.8048589341692791, "grad_norm": 4.70746374130249, "learning_rate": 5.780040293110208e-06, "loss": 1.5434, "step": 23030 }, { "epoch": 1.804937304075235, "grad_norm": 6.117354393005371, "learning_rate": 5.779735705220457e-06, "loss": 2.3308, "step": 23031 }, { "epoch": 1.8050156739811913, "grad_norm": 3.6430110931396484, "learning_rate": 5.7794311143649925e-06, "loss": 1.4275, "step": 23032 }, { "epoch": 1.8050940438871472, "grad_norm": 4.1613664627075195, "learning_rate": 5.779126520544973e-06, "loss": 1.7197, "step": 23033 }, { "epoch": 1.8051724137931036, "grad_norm": 2.049506664276123, "learning_rate": 5.778821923761555e-06, "loss": 1.2026, "step": 23034 }, { "epoch": 1.8052507836990594, "grad_norm": 14.711586952209473, "learning_rate": 5.7785173240159e-06, "loss": 0.6161, "step": 23035 }, { "epoch": 1.8053291536050158, "grad_norm": 3.047266960144043, "learning_rate": 5.778212721309162e-06, "loss": 1.2592, "step": 23036 }, { "epoch": 1.8054075235109717, "grad_norm": 2.4399750232696533, "learning_rate": 5.777908115642505e-06, "loss": 0.7945, "step": 23037 }, { "epoch": 1.805485893416928, "grad_norm": 1.959885597229004, "learning_rate": 5.777603507017084e-06, "loss": 0.9613, "step": 23038 }, { "epoch": 1.8055642633228839, "grad_norm": 3.1504805088043213, "learning_rate": 5.77729889543406e-06, "loss": 0.8911, "step": 23039 }, { "epoch": 1.8056426332288402, "grad_norm": 8.79568099975586, "learning_rate": 5.77699428089459e-06, "loss": 1.7404, "step": 23040 }, { "epoch": 1.805721003134796, "grad_norm": 4.036530494689941, "learning_rate": 5.776689663399833e-06, "loss": 1.2245, "step": 23041 }, { "epoch": 1.8057993730407524, "grad_norm": 2.316725254058838, "learning_rate": 5.776385042950946e-06, "loss": 1.3579, "step": 23042 }, { "epoch": 1.8058777429467083, "grad_norm": 3.7672548294067383, "learning_rate": 5.776080419549089e-06, "loss": 1.4534, "step": 23043 }, { "epoch": 1.8059561128526647, "grad_norm": 5.114150524139404, "learning_rate": 5.775775793195422e-06, "loss": 1.9191, "step": 23044 }, { "epoch": 1.8060344827586206, "grad_norm": 2.688422679901123, "learning_rate": 5.775471163891102e-06, "loss": 1.4366, "step": 23045 }, { "epoch": 1.8061128526645769, "grad_norm": 3.1779158115386963, "learning_rate": 5.775166531637288e-06, "loss": 1.3808, "step": 23046 }, { "epoch": 1.8061912225705328, "grad_norm": 1.9974488019943237, "learning_rate": 5.774861896435139e-06, "loss": 1.1495, "step": 23047 }, { "epoch": 1.806269592476489, "grad_norm": 3.8094961643218994, "learning_rate": 5.774557258285812e-06, "loss": 1.5315, "step": 23048 }, { "epoch": 1.806347962382445, "grad_norm": 3.3597447872161865, "learning_rate": 5.7742526171904675e-06, "loss": 1.402, "step": 23049 }, { "epoch": 1.8064263322884013, "grad_norm": 2.446976900100708, "learning_rate": 5.773947973150265e-06, "loss": 1.4759, "step": 23050 }, { "epoch": 1.8065047021943572, "grad_norm": 2.3262412548065186, "learning_rate": 5.7736433261663606e-06, "loss": 1.3818, "step": 23051 }, { "epoch": 1.8065830721003135, "grad_norm": 2.2641477584838867, "learning_rate": 5.773338676239915e-06, "loss": 1.6438, "step": 23052 }, { "epoch": 1.8066614420062694, "grad_norm": 3.683049201965332, "learning_rate": 5.773034023372086e-06, "loss": 1.2647, "step": 23053 }, { "epoch": 1.8067398119122258, "grad_norm": 2.6775944232940674, "learning_rate": 5.772729367564034e-06, "loss": 1.2425, "step": 23054 }, { "epoch": 1.8068181818181817, "grad_norm": 2.76275634765625, "learning_rate": 5.7724247088169136e-06, "loss": 1.2337, "step": 23055 }, { "epoch": 1.806896551724138, "grad_norm": 4.100339889526367, "learning_rate": 5.772120047131889e-06, "loss": 1.3747, "step": 23056 }, { "epoch": 1.8069749216300939, "grad_norm": 2.899509906768799, "learning_rate": 5.771815382510113e-06, "loss": 1.6591, "step": 23057 }, { "epoch": 1.8070532915360502, "grad_norm": 6.287993431091309, "learning_rate": 5.771510714952751e-06, "loss": 1.5463, "step": 23058 }, { "epoch": 1.8071316614420063, "grad_norm": 8.724695205688477, "learning_rate": 5.771206044460957e-06, "loss": 0.8897, "step": 23059 }, { "epoch": 1.8072100313479624, "grad_norm": 3.15787935256958, "learning_rate": 5.770901371035893e-06, "loss": 1.1029, "step": 23060 }, { "epoch": 1.8072884012539185, "grad_norm": 2.2406418323516846, "learning_rate": 5.770596694678715e-06, "loss": 1.3511, "step": 23061 }, { "epoch": 1.8073667711598747, "grad_norm": 2.57330584526062, "learning_rate": 5.770292015390582e-06, "loss": 1.1563, "step": 23062 }, { "epoch": 1.8074451410658308, "grad_norm": 4.511711597442627, "learning_rate": 5.769987333172655e-06, "loss": 1.7756, "step": 23063 }, { "epoch": 1.8075235109717869, "grad_norm": 2.354806661605835, "learning_rate": 5.76968264802609e-06, "loss": 1.3029, "step": 23064 }, { "epoch": 1.807601880877743, "grad_norm": 3.7202892303466797, "learning_rate": 5.76937795995205e-06, "loss": 1.5573, "step": 23065 }, { "epoch": 1.807680250783699, "grad_norm": 4.888895034790039, "learning_rate": 5.76907326895169e-06, "loss": 1.5586, "step": 23066 }, { "epoch": 1.8077586206896552, "grad_norm": 4.051336765289307, "learning_rate": 5.768768575026171e-06, "loss": 1.3459, "step": 23067 }, { "epoch": 1.8078369905956113, "grad_norm": 3.482631206512451, "learning_rate": 5.768463878176651e-06, "loss": 1.2004, "step": 23068 }, { "epoch": 1.8079153605015674, "grad_norm": 3.3100945949554443, "learning_rate": 5.768159178404289e-06, "loss": 1.1568, "step": 23069 }, { "epoch": 1.8079937304075235, "grad_norm": 2.634657144546509, "learning_rate": 5.767854475710244e-06, "loss": 1.421, "step": 23070 }, { "epoch": 1.8080721003134796, "grad_norm": 3.3376598358154297, "learning_rate": 5.767549770095675e-06, "loss": 1.1249, "step": 23071 }, { "epoch": 1.8081504702194358, "grad_norm": 3.1759965419769287, "learning_rate": 5.767245061561741e-06, "loss": 1.1489, "step": 23072 }, { "epoch": 1.8082288401253919, "grad_norm": 3.2446513175964355, "learning_rate": 5.766940350109601e-06, "loss": 0.812, "step": 23073 }, { "epoch": 1.808307210031348, "grad_norm": 2.2345781326293945, "learning_rate": 5.766635635740414e-06, "loss": 1.5059, "step": 23074 }, { "epoch": 1.808385579937304, "grad_norm": 6.404743671417236, "learning_rate": 5.766330918455338e-06, "loss": 0.4878, "step": 23075 }, { "epoch": 1.8084639498432602, "grad_norm": 2.850402593612671, "learning_rate": 5.766026198255534e-06, "loss": 1.3175, "step": 23076 }, { "epoch": 1.8085423197492163, "grad_norm": 2.112523078918457, "learning_rate": 5.765721475142157e-06, "loss": 1.4921, "step": 23077 }, { "epoch": 1.8086206896551724, "grad_norm": 2.5490729808807373, "learning_rate": 5.765416749116372e-06, "loss": 1.2507, "step": 23078 }, { "epoch": 1.8086990595611285, "grad_norm": 2.9112226963043213, "learning_rate": 5.765112020179333e-06, "loss": 1.6851, "step": 23079 }, { "epoch": 1.8087774294670846, "grad_norm": 5.06282377243042, "learning_rate": 5.764807288332202e-06, "loss": 1.519, "step": 23080 }, { "epoch": 1.8088557993730408, "grad_norm": 4.121644973754883, "learning_rate": 5.764502553576136e-06, "loss": 1.6094, "step": 23081 }, { "epoch": 1.8089341692789969, "grad_norm": 2.411846399307251, "learning_rate": 5.764197815912296e-06, "loss": 1.1489, "step": 23082 }, { "epoch": 1.809012539184953, "grad_norm": 6.826808452606201, "learning_rate": 5.7638930753418375e-06, "loss": 1.8152, "step": 23083 }, { "epoch": 1.809090909090909, "grad_norm": 3.5277466773986816, "learning_rate": 5.7635883318659255e-06, "loss": 1.2713, "step": 23084 }, { "epoch": 1.8091692789968652, "grad_norm": 2.4583075046539307, "learning_rate": 5.763283585485713e-06, "loss": 1.5569, "step": 23085 }, { "epoch": 1.8092476489028213, "grad_norm": 3.5126426219940186, "learning_rate": 5.762978836202364e-06, "loss": 1.2558, "step": 23086 }, { "epoch": 1.8093260188087774, "grad_norm": 2.8735134601593018, "learning_rate": 5.762674084017033e-06, "loss": 1.3784, "step": 23087 }, { "epoch": 1.8094043887147335, "grad_norm": 12.14175033569336, "learning_rate": 5.762369328930884e-06, "loss": 1.4808, "step": 23088 }, { "epoch": 1.8094827586206896, "grad_norm": 3.8676681518554688, "learning_rate": 5.762064570945073e-06, "loss": 0.886, "step": 23089 }, { "epoch": 1.8095611285266457, "grad_norm": 8.342325210571289, "learning_rate": 5.761759810060759e-06, "loss": 1.3153, "step": 23090 }, { "epoch": 1.8096394984326019, "grad_norm": 2.745457649230957, "learning_rate": 5.761455046279103e-06, "loss": 1.296, "step": 23091 }, { "epoch": 1.809717868338558, "grad_norm": 2.5329818725585938, "learning_rate": 5.761150279601261e-06, "loss": 1.3265, "step": 23092 }, { "epoch": 1.809796238244514, "grad_norm": 4.771634101867676, "learning_rate": 5.760845510028397e-06, "loss": 1.3085, "step": 23093 }, { "epoch": 1.8098746081504702, "grad_norm": 1.4942008256912231, "learning_rate": 5.760540737561665e-06, "loss": 1.3954, "step": 23094 }, { "epoch": 1.8099529780564263, "grad_norm": 2.7719919681549072, "learning_rate": 5.7602359622022276e-06, "loss": 1.379, "step": 23095 }, { "epoch": 1.8100313479623824, "grad_norm": 3.457249879837036, "learning_rate": 5.759931183951244e-06, "loss": 1.3679, "step": 23096 }, { "epoch": 1.8101097178683385, "grad_norm": 2.456312894821167, "learning_rate": 5.759626402809872e-06, "loss": 1.2527, "step": 23097 }, { "epoch": 1.8101880877742946, "grad_norm": 3.2403945922851562, "learning_rate": 5.75932161877927e-06, "loss": 1.5749, "step": 23098 }, { "epoch": 1.8102664576802507, "grad_norm": 2.882887840270996, "learning_rate": 5.7590168318606e-06, "loss": 1.5336, "step": 23099 }, { "epoch": 1.8103448275862069, "grad_norm": 2.2649986743927, "learning_rate": 5.758712042055019e-06, "loss": 1.6386, "step": 23100 }, { "epoch": 1.810423197492163, "grad_norm": 2.175046682357788, "learning_rate": 5.758407249363689e-06, "loss": 0.939, "step": 23101 }, { "epoch": 1.810501567398119, "grad_norm": 3.8818140029907227, "learning_rate": 5.758102453787766e-06, "loss": 1.4633, "step": 23102 }, { "epoch": 1.8105799373040752, "grad_norm": 3.2161011695861816, "learning_rate": 5.75779765532841e-06, "loss": 1.562, "step": 23103 }, { "epoch": 1.8106583072100313, "grad_norm": 2.0284578800201416, "learning_rate": 5.757492853986781e-06, "loss": 1.1518, "step": 23104 }, { "epoch": 1.8107366771159876, "grad_norm": 2.3440847396850586, "learning_rate": 5.75718804976404e-06, "loss": 1.0553, "step": 23105 }, { "epoch": 1.8108150470219435, "grad_norm": 3.0321619510650635, "learning_rate": 5.756883242661342e-06, "loss": 1.309, "step": 23106 }, { "epoch": 1.8108934169278998, "grad_norm": 2.237293004989624, "learning_rate": 5.756578432679851e-06, "loss": 1.2971, "step": 23107 }, { "epoch": 1.8109717868338557, "grad_norm": 4.093634605407715, "learning_rate": 5.7562736198207245e-06, "loss": 1.5747, "step": 23108 }, { "epoch": 1.811050156739812, "grad_norm": 3.6202433109283447, "learning_rate": 5.7559688040851196e-06, "loss": 1.4636, "step": 23109 }, { "epoch": 1.811128526645768, "grad_norm": 2.4399967193603516, "learning_rate": 5.7556639854741995e-06, "loss": 0.8496, "step": 23110 }, { "epoch": 1.8112068965517243, "grad_norm": 2.4713101387023926, "learning_rate": 5.7553591639891195e-06, "loss": 1.3399, "step": 23111 }, { "epoch": 1.8112852664576802, "grad_norm": 3.73022723197937, "learning_rate": 5.755054339631044e-06, "loss": 1.1327, "step": 23112 }, { "epoch": 1.8113636363636365, "grad_norm": 2.3157126903533936, "learning_rate": 5.754749512401128e-06, "loss": 1.2424, "step": 23113 }, { "epoch": 1.8114420062695924, "grad_norm": 2.819607973098755, "learning_rate": 5.754444682300533e-06, "loss": 1.7245, "step": 23114 }, { "epoch": 1.8115203761755487, "grad_norm": 2.3246045112609863, "learning_rate": 5.754139849330417e-06, "loss": 1.5298, "step": 23115 }, { "epoch": 1.8115987460815046, "grad_norm": 3.346737861633301, "learning_rate": 5.75383501349194e-06, "loss": 1.6548, "step": 23116 }, { "epoch": 1.811677115987461, "grad_norm": 3.701261520385742, "learning_rate": 5.753530174786265e-06, "loss": 1.5322, "step": 23117 }, { "epoch": 1.8117554858934168, "grad_norm": 2.7707741260528564, "learning_rate": 5.753225333214545e-06, "loss": 1.7422, "step": 23118 }, { "epoch": 1.8118338557993732, "grad_norm": 2.4035160541534424, "learning_rate": 5.752920488777944e-06, "loss": 1.2531, "step": 23119 }, { "epoch": 1.811912225705329, "grad_norm": 2.261916399002075, "learning_rate": 5.75261564147762e-06, "loss": 1.1926, "step": 23120 }, { "epoch": 1.8119905956112854, "grad_norm": 2.8455939292907715, "learning_rate": 5.7523107913147336e-06, "loss": 1.4623, "step": 23121 }, { "epoch": 1.8120689655172413, "grad_norm": 2.0853962898254395, "learning_rate": 5.752005938290441e-06, "loss": 1.2561, "step": 23122 }, { "epoch": 1.8121473354231976, "grad_norm": 2.5827252864837646, "learning_rate": 5.7517010824059065e-06, "loss": 1.29, "step": 23123 }, { "epoch": 1.8122257053291535, "grad_norm": 2.0602829456329346, "learning_rate": 5.751396223662286e-06, "loss": 1.479, "step": 23124 }, { "epoch": 1.8123040752351098, "grad_norm": 3.255777359008789, "learning_rate": 5.75109136206074e-06, "loss": 1.693, "step": 23125 }, { "epoch": 1.8123824451410657, "grad_norm": 3.3647255897521973, "learning_rate": 5.750786497602429e-06, "loss": 1.358, "step": 23126 }, { "epoch": 1.812460815047022, "grad_norm": 6.434975624084473, "learning_rate": 5.750481630288511e-06, "loss": 1.3676, "step": 23127 }, { "epoch": 1.812539184952978, "grad_norm": 2.4079277515411377, "learning_rate": 5.7501767601201465e-06, "loss": 1.1531, "step": 23128 }, { "epoch": 1.8126175548589343, "grad_norm": 3.3105626106262207, "learning_rate": 5.749871887098497e-06, "loss": 1.3744, "step": 23129 }, { "epoch": 1.8126959247648902, "grad_norm": 8.230779647827148, "learning_rate": 5.749567011224717e-06, "loss": 2.277, "step": 23130 }, { "epoch": 1.8127742946708465, "grad_norm": 3.2777717113494873, "learning_rate": 5.749262132499971e-06, "loss": 1.7241, "step": 23131 }, { "epoch": 1.8128526645768024, "grad_norm": 2.5523786544799805, "learning_rate": 5.748957250925416e-06, "loss": 1.5977, "step": 23132 }, { "epoch": 1.8129310344827587, "grad_norm": 2.080242872238159, "learning_rate": 5.748652366502212e-06, "loss": 0.7126, "step": 23133 }, { "epoch": 1.8130094043887146, "grad_norm": 5.238704681396484, "learning_rate": 5.748347479231518e-06, "loss": 0.6757, "step": 23134 }, { "epoch": 1.813087774294671, "grad_norm": 2.7693443298339844, "learning_rate": 5.748042589114496e-06, "loss": 1.4778, "step": 23135 }, { "epoch": 1.8131661442006268, "grad_norm": 4.029564380645752, "learning_rate": 5.747737696152305e-06, "loss": 1.206, "step": 23136 }, { "epoch": 1.8132445141065832, "grad_norm": 3.947936534881592, "learning_rate": 5.7474328003461025e-06, "loss": 1.8506, "step": 23137 }, { "epoch": 1.813322884012539, "grad_norm": 2.6315619945526123, "learning_rate": 5.74712790169705e-06, "loss": 1.1964, "step": 23138 }, { "epoch": 1.8134012539184954, "grad_norm": 3.5086305141448975, "learning_rate": 5.746823000206306e-06, "loss": 1.3985, "step": 23139 }, { "epoch": 1.8134796238244513, "grad_norm": 3.5553770065307617, "learning_rate": 5.746518095875033e-06, "loss": 1.7648, "step": 23140 }, { "epoch": 1.8135579937304076, "grad_norm": 6.633967876434326, "learning_rate": 5.746213188704387e-06, "loss": 1.2605, "step": 23141 }, { "epoch": 1.8136363636363635, "grad_norm": 3.0621562004089355, "learning_rate": 5.745908278695529e-06, "loss": 1.5458, "step": 23142 }, { "epoch": 1.8137147335423198, "grad_norm": 9.595025062561035, "learning_rate": 5.7456033658496205e-06, "loss": 1.2015, "step": 23143 }, { "epoch": 1.8137931034482757, "grad_norm": 1.868491768836975, "learning_rate": 5.7452984501678175e-06, "loss": 1.4153, "step": 23144 }, { "epoch": 1.813871473354232, "grad_norm": 3.9171576499938965, "learning_rate": 5.744993531651284e-06, "loss": 1.1303, "step": 23145 }, { "epoch": 1.813949843260188, "grad_norm": 5.304259777069092, "learning_rate": 5.7446886103011755e-06, "loss": 1.4378, "step": 23146 }, { "epoch": 1.8140282131661443, "grad_norm": 4.572198390960693, "learning_rate": 5.744383686118657e-06, "loss": 1.4195, "step": 23147 }, { "epoch": 1.8141065830721002, "grad_norm": 2.218571662902832, "learning_rate": 5.744078759104884e-06, "loss": 1.1267, "step": 23148 }, { "epoch": 1.8141849529780565, "grad_norm": 8.40436840057373, "learning_rate": 5.743773829261016e-06, "loss": 1.4732, "step": 23149 }, { "epoch": 1.8142633228840124, "grad_norm": 4.723679542541504, "learning_rate": 5.743468896588217e-06, "loss": 1.3001, "step": 23150 }, { "epoch": 1.8143416927899687, "grad_norm": 2.7854442596435547, "learning_rate": 5.743163961087642e-06, "loss": 1.6923, "step": 23151 }, { "epoch": 1.8144200626959248, "grad_norm": 3.3426008224487305, "learning_rate": 5.742859022760454e-06, "loss": 0.8445, "step": 23152 }, { "epoch": 1.814498432601881, "grad_norm": 2.2944695949554443, "learning_rate": 5.742554081607812e-06, "loss": 1.5898, "step": 23153 }, { "epoch": 1.814576802507837, "grad_norm": 6.084152698516846, "learning_rate": 5.742249137630876e-06, "loss": 1.3534, "step": 23154 }, { "epoch": 1.8146551724137931, "grad_norm": 3.4798526763916016, "learning_rate": 5.741944190830806e-06, "loss": 1.3916, "step": 23155 }, { "epoch": 1.8147335423197493, "grad_norm": 8.010255813598633, "learning_rate": 5.741639241208759e-06, "loss": 1.2436, "step": 23156 }, { "epoch": 1.8148119122257054, "grad_norm": 3.6554112434387207, "learning_rate": 5.741334288765899e-06, "loss": 1.4077, "step": 23157 }, { "epoch": 1.8148902821316615, "grad_norm": 2.1425774097442627, "learning_rate": 5.741029333503384e-06, "loss": 1.6403, "step": 23158 }, { "epoch": 1.8149686520376176, "grad_norm": 4.707508087158203, "learning_rate": 5.740724375422374e-06, "loss": 1.5019, "step": 23159 }, { "epoch": 1.8150470219435737, "grad_norm": 4.183837890625, "learning_rate": 5.7404194145240276e-06, "loss": 1.134, "step": 23160 }, { "epoch": 1.8151253918495298, "grad_norm": 4.333556652069092, "learning_rate": 5.740114450809508e-06, "loss": 1.1844, "step": 23161 }, { "epoch": 1.815203761755486, "grad_norm": 2.6220409870147705, "learning_rate": 5.739809484279974e-06, "loss": 1.4473, "step": 23162 }, { "epoch": 1.815282131661442, "grad_norm": 2.6929454803466797, "learning_rate": 5.739504514936581e-06, "loss": 1.4285, "step": 23163 }, { "epoch": 1.8153605015673981, "grad_norm": 6.5436882972717285, "learning_rate": 5.739199542780496e-06, "loss": 0.5054, "step": 23164 }, { "epoch": 1.8154388714733543, "grad_norm": 2.425408124923706, "learning_rate": 5.738894567812873e-06, "loss": 1.3956, "step": 23165 }, { "epoch": 1.8155172413793104, "grad_norm": 5.259180068969727, "learning_rate": 5.738589590034877e-06, "loss": 1.1289, "step": 23166 }, { "epoch": 1.8155956112852665, "grad_norm": 6.323754787445068, "learning_rate": 5.738284609447664e-06, "loss": 1.5806, "step": 23167 }, { "epoch": 1.8156739811912226, "grad_norm": 5.329339981079102, "learning_rate": 5.737979626052397e-06, "loss": 1.2456, "step": 23168 }, { "epoch": 1.8157523510971787, "grad_norm": 2.4111170768737793, "learning_rate": 5.737674639850232e-06, "loss": 1.4281, "step": 23169 }, { "epoch": 1.8158307210031348, "grad_norm": 2.694512367248535, "learning_rate": 5.737369650842334e-06, "loss": 1.4504, "step": 23170 }, { "epoch": 1.815909090909091, "grad_norm": 3.9142608642578125, "learning_rate": 5.737064659029861e-06, "loss": 1.1206, "step": 23171 }, { "epoch": 1.815987460815047, "grad_norm": 4.006898880004883, "learning_rate": 5.73675966441397e-06, "loss": 2.2777, "step": 23172 }, { "epoch": 1.8160658307210031, "grad_norm": 2.2794883251190186, "learning_rate": 5.736454666995825e-06, "loss": 1.1285, "step": 23173 }, { "epoch": 1.8161442006269592, "grad_norm": 1.7663604021072388, "learning_rate": 5.736149666776585e-06, "loss": 1.4933, "step": 23174 }, { "epoch": 1.8162225705329154, "grad_norm": 5.01282262802124, "learning_rate": 5.735844663757409e-06, "loss": 1.3742, "step": 23175 }, { "epoch": 1.8163009404388715, "grad_norm": 4.461760520935059, "learning_rate": 5.735539657939459e-06, "loss": 1.6465, "step": 23176 }, { "epoch": 1.8163793103448276, "grad_norm": 11.98300838470459, "learning_rate": 5.735234649323893e-06, "loss": 0.3773, "step": 23177 }, { "epoch": 1.8164576802507837, "grad_norm": 7.821341514587402, "learning_rate": 5.734929637911872e-06, "loss": 1.7029, "step": 23178 }, { "epoch": 1.8165360501567398, "grad_norm": 3.382047176361084, "learning_rate": 5.734624623704556e-06, "loss": 1.3806, "step": 23179 }, { "epoch": 1.816614420062696, "grad_norm": 3.0590786933898926, "learning_rate": 5.734319606703107e-06, "loss": 1.4273, "step": 23180 }, { "epoch": 1.816692789968652, "grad_norm": 4.498565673828125, "learning_rate": 5.7340145869086815e-06, "loss": 1.3493, "step": 23181 }, { "epoch": 1.8167711598746081, "grad_norm": 2.5532889366149902, "learning_rate": 5.733709564322443e-06, "loss": 1.2439, "step": 23182 }, { "epoch": 1.8168495297805642, "grad_norm": 3.5008857250213623, "learning_rate": 5.73340453894555e-06, "loss": 1.7976, "step": 23183 }, { "epoch": 1.8169278996865204, "grad_norm": 1.7451698780059814, "learning_rate": 5.7330995107791605e-06, "loss": 1.3663, "step": 23184 }, { "epoch": 1.8170062695924765, "grad_norm": 3.100677251815796, "learning_rate": 5.73279447982444e-06, "loss": 1.4658, "step": 23185 }, { "epoch": 1.8170846394984326, "grad_norm": 4.416818141937256, "learning_rate": 5.732489446082544e-06, "loss": 1.4096, "step": 23186 }, { "epoch": 1.8171630094043887, "grad_norm": 4.496610164642334, "learning_rate": 5.732184409554634e-06, "loss": 1.5474, "step": 23187 }, { "epoch": 1.8172413793103448, "grad_norm": 3.2606942653656006, "learning_rate": 5.7318793702418716e-06, "loss": 1.699, "step": 23188 }, { "epoch": 1.817319749216301, "grad_norm": 2.770120143890381, "learning_rate": 5.7315743281454164e-06, "loss": 1.7841, "step": 23189 }, { "epoch": 1.817398119122257, "grad_norm": 2.9614596366882324, "learning_rate": 5.731269283266428e-06, "loss": 1.3026, "step": 23190 }, { "epoch": 1.8174764890282131, "grad_norm": 4.216982364654541, "learning_rate": 5.730964235606065e-06, "loss": 1.8361, "step": 23191 }, { "epoch": 1.8175548589341692, "grad_norm": 2.715778112411499, "learning_rate": 5.730659185165492e-06, "loss": 1.2487, "step": 23192 }, { "epoch": 1.8176332288401253, "grad_norm": 4.312142848968506, "learning_rate": 5.730354131945865e-06, "loss": 1.3222, "step": 23193 }, { "epoch": 1.8177115987460815, "grad_norm": 2.8641791343688965, "learning_rate": 5.730049075948347e-06, "loss": 1.6951, "step": 23194 }, { "epoch": 1.8177899686520376, "grad_norm": 4.4488420486450195, "learning_rate": 5.729744017174097e-06, "loss": 1.3065, "step": 23195 }, { "epoch": 1.8178683385579937, "grad_norm": 4.407907009124756, "learning_rate": 5.729438955624276e-06, "loss": 1.2911, "step": 23196 }, { "epoch": 1.8179467084639498, "grad_norm": 2.504059076309204, "learning_rate": 5.729133891300043e-06, "loss": 1.3315, "step": 23197 }, { "epoch": 1.8180250783699061, "grad_norm": 2.1327757835388184, "learning_rate": 5.728828824202559e-06, "loss": 1.583, "step": 23198 }, { "epoch": 1.818103448275862, "grad_norm": 2.1604886054992676, "learning_rate": 5.728523754332986e-06, "loss": 1.3428, "step": 23199 }, { "epoch": 1.8181818181818183, "grad_norm": 2.835024356842041, "learning_rate": 5.728218681692482e-06, "loss": 1.7779, "step": 23200 }, { "epoch": 1.8182601880877742, "grad_norm": 3.444262742996216, "learning_rate": 5.727913606282207e-06, "loss": 1.4872, "step": 23201 }, { "epoch": 1.8183385579937306, "grad_norm": 3.8338708877563477, "learning_rate": 5.727608528103324e-06, "loss": 0.967, "step": 23202 }, { "epoch": 1.8184169278996865, "grad_norm": 3.7065372467041016, "learning_rate": 5.727303447156991e-06, "loss": 1.6332, "step": 23203 }, { "epoch": 1.8184952978056428, "grad_norm": 9.073384284973145, "learning_rate": 5.72699836344437e-06, "loss": 0.9321, "step": 23204 }, { "epoch": 1.8185736677115987, "grad_norm": 3.730870485305786, "learning_rate": 5.726693276966621e-06, "loss": 1.3759, "step": 23205 }, { "epoch": 1.818652037617555, "grad_norm": 4.169299602508545, "learning_rate": 5.726388187724903e-06, "loss": 1.7487, "step": 23206 }, { "epoch": 1.818730407523511, "grad_norm": 3.0530130863189697, "learning_rate": 5.726083095720377e-06, "loss": 0.8792, "step": 23207 }, { "epoch": 1.8188087774294672, "grad_norm": 2.914379119873047, "learning_rate": 5.725778000954206e-06, "loss": 1.0746, "step": 23208 }, { "epoch": 1.8188871473354231, "grad_norm": 3.4800374507904053, "learning_rate": 5.725472903427547e-06, "loss": 1.3469, "step": 23209 }, { "epoch": 1.8189655172413794, "grad_norm": 2.2386462688446045, "learning_rate": 5.72516780314156e-06, "loss": 1.2986, "step": 23210 }, { "epoch": 1.8190438871473353, "grad_norm": 1.9001580476760864, "learning_rate": 5.72486270009741e-06, "loss": 1.4293, "step": 23211 }, { "epoch": 1.8191222570532917, "grad_norm": 15.148691177368164, "learning_rate": 5.724557594296253e-06, "loss": 1.0691, "step": 23212 }, { "epoch": 1.8192006269592476, "grad_norm": 3.3542585372924805, "learning_rate": 5.724252485739252e-06, "loss": 1.295, "step": 23213 }, { "epoch": 1.819278996865204, "grad_norm": 4.32048225402832, "learning_rate": 5.723947374427565e-06, "loss": 1.5231, "step": 23214 }, { "epoch": 1.8193573667711598, "grad_norm": 5.4797515869140625, "learning_rate": 5.723642260362356e-06, "loss": 0.9941, "step": 23215 }, { "epoch": 1.8194357366771161, "grad_norm": 2.739680290222168, "learning_rate": 5.7233371435447824e-06, "loss": 1.8431, "step": 23216 }, { "epoch": 1.819514106583072, "grad_norm": 2.2228775024414062, "learning_rate": 5.7230320239760055e-06, "loss": 1.4867, "step": 23217 }, { "epoch": 1.8195924764890283, "grad_norm": 2.8700850009918213, "learning_rate": 5.722726901657187e-06, "loss": 1.1215, "step": 23218 }, { "epoch": 1.8196708463949842, "grad_norm": 1.7860599756240845, "learning_rate": 5.7224217765894865e-06, "loss": 1.3658, "step": 23219 }, { "epoch": 1.8197492163009406, "grad_norm": 2.2469635009765625, "learning_rate": 5.7221166487740645e-06, "loss": 1.2719, "step": 23220 }, { "epoch": 1.8198275862068964, "grad_norm": 3.982011556625366, "learning_rate": 5.7218115182120814e-06, "loss": 1.1401, "step": 23221 }, { "epoch": 1.8199059561128528, "grad_norm": 2.4573869705200195, "learning_rate": 5.721506384904698e-06, "loss": 1.2318, "step": 23222 }, { "epoch": 1.8199843260188087, "grad_norm": 3.7326035499572754, "learning_rate": 5.7212012488530755e-06, "loss": 1.6438, "step": 23223 }, { "epoch": 1.820062695924765, "grad_norm": 2.7462003231048584, "learning_rate": 5.720896110058373e-06, "loss": 1.0822, "step": 23224 }, { "epoch": 1.8201410658307209, "grad_norm": 2.458725690841675, "learning_rate": 5.720590968521752e-06, "loss": 1.0526, "step": 23225 }, { "epoch": 1.8202194357366772, "grad_norm": 5.798128128051758, "learning_rate": 5.720285824244374e-06, "loss": 1.0719, "step": 23226 }, { "epoch": 1.820297805642633, "grad_norm": 4.751236915588379, "learning_rate": 5.719980677227399e-06, "loss": 1.552, "step": 23227 }, { "epoch": 1.8203761755485894, "grad_norm": 5.36222505569458, "learning_rate": 5.719675527471985e-06, "loss": 1.3064, "step": 23228 }, { "epoch": 1.8204545454545453, "grad_norm": 3.5754334926605225, "learning_rate": 5.719370374979297e-06, "loss": 1.168, "step": 23229 }, { "epoch": 1.8205329153605017, "grad_norm": 5.696514129638672, "learning_rate": 5.719065219750493e-06, "loss": 1.7377, "step": 23230 }, { "epoch": 1.8206112852664575, "grad_norm": 2.539548635482788, "learning_rate": 5.7187600617867344e-06, "loss": 1.0719, "step": 23231 }, { "epoch": 1.8206896551724139, "grad_norm": 7.8299970626831055, "learning_rate": 5.718454901089181e-06, "loss": 1.6521, "step": 23232 }, { "epoch": 1.8207680250783698, "grad_norm": 2.8891689777374268, "learning_rate": 5.718149737658996e-06, "loss": 1.0382, "step": 23233 }, { "epoch": 1.820846394984326, "grad_norm": 3.6189358234405518, "learning_rate": 5.717844571497336e-06, "loss": 1.5762, "step": 23234 }, { "epoch": 1.820924764890282, "grad_norm": 3.326874256134033, "learning_rate": 5.7175394026053654e-06, "loss": 1.2646, "step": 23235 }, { "epoch": 1.8210031347962383, "grad_norm": 2.5770089626312256, "learning_rate": 5.717234230984242e-06, "loss": 1.3876, "step": 23236 }, { "epoch": 1.8210815047021942, "grad_norm": 2.4899368286132812, "learning_rate": 5.71692905663513e-06, "loss": 1.2169, "step": 23237 }, { "epoch": 1.8211598746081505, "grad_norm": 2.8408257961273193, "learning_rate": 5.716623879559187e-06, "loss": 1.3273, "step": 23238 }, { "epoch": 1.8212382445141064, "grad_norm": 5.538912773132324, "learning_rate": 5.716318699757574e-06, "loss": 0.9637, "step": 23239 }, { "epoch": 1.8213166144200628, "grad_norm": 3.1086151599884033, "learning_rate": 5.716013517231453e-06, "loss": 1.6986, "step": 23240 }, { "epoch": 1.8213949843260187, "grad_norm": 7.398880481719971, "learning_rate": 5.7157083319819865e-06, "loss": 1.7795, "step": 23241 }, { "epoch": 1.821473354231975, "grad_norm": 2.5823476314544678, "learning_rate": 5.715403144010332e-06, "loss": 1.4645, "step": 23242 }, { "epoch": 1.8215517241379309, "grad_norm": 2.5269243717193604, "learning_rate": 5.7150979533176494e-06, "loss": 1.5499, "step": 23243 }, { "epoch": 1.8216300940438872, "grad_norm": 4.6742987632751465, "learning_rate": 5.714792759905104e-06, "loss": 1.3399, "step": 23244 }, { "epoch": 1.8217084639498433, "grad_norm": 6.643151760101318, "learning_rate": 5.714487563773853e-06, "loss": 1.7876, "step": 23245 }, { "epoch": 1.8217868338557994, "grad_norm": 3.1259446144104004, "learning_rate": 5.7141823649250585e-06, "loss": 1.6817, "step": 23246 }, { "epoch": 1.8218652037617555, "grad_norm": 2.2515854835510254, "learning_rate": 5.71387716335988e-06, "loss": 1.6221, "step": 23247 }, { "epoch": 1.8219435736677116, "grad_norm": 2.161139965057373, "learning_rate": 5.713571959079482e-06, "loss": 1.4084, "step": 23248 }, { "epoch": 1.8220219435736678, "grad_norm": 2.5049831867218018, "learning_rate": 5.713266752085021e-06, "loss": 1.489, "step": 23249 }, { "epoch": 1.8221003134796239, "grad_norm": 2.6601805686950684, "learning_rate": 5.71296154237766e-06, "loss": 1.5252, "step": 23250 }, { "epoch": 1.82217868338558, "grad_norm": 2.654510259628296, "learning_rate": 5.712656329958559e-06, "loss": 1.8155, "step": 23251 }, { "epoch": 1.822257053291536, "grad_norm": 2.591304302215576, "learning_rate": 5.712351114828879e-06, "loss": 0.9193, "step": 23252 }, { "epoch": 1.8223354231974922, "grad_norm": 2.4569220542907715, "learning_rate": 5.712045896989782e-06, "loss": 1.36, "step": 23253 }, { "epoch": 1.8224137931034483, "grad_norm": 3.080165386199951, "learning_rate": 5.711740676442428e-06, "loss": 1.5898, "step": 23254 }, { "epoch": 1.8224921630094044, "grad_norm": 2.0316455364227295, "learning_rate": 5.711435453187979e-06, "loss": 1.3421, "step": 23255 }, { "epoch": 1.8225705329153605, "grad_norm": 4.875380516052246, "learning_rate": 5.711130227227593e-06, "loss": 1.6757, "step": 23256 }, { "epoch": 1.8226489028213166, "grad_norm": 2.82859206199646, "learning_rate": 5.710824998562434e-06, "loss": 1.7952, "step": 23257 }, { "epoch": 1.8227272727272728, "grad_norm": 4.789337635040283, "learning_rate": 5.710519767193662e-06, "loss": 1.499, "step": 23258 }, { "epoch": 1.8228056426332289, "grad_norm": 4.355620384216309, "learning_rate": 5.710214533122438e-06, "loss": 1.0349, "step": 23259 }, { "epoch": 1.822884012539185, "grad_norm": 2.269012689590454, "learning_rate": 5.709909296349921e-06, "loss": 1.5553, "step": 23260 }, { "epoch": 1.822962382445141, "grad_norm": 6.4393463134765625, "learning_rate": 5.709604056877276e-06, "loss": 1.1235, "step": 23261 }, { "epoch": 1.8230407523510972, "grad_norm": 2.6324615478515625, "learning_rate": 5.70929881470566e-06, "loss": 1.7852, "step": 23262 }, { "epoch": 1.8231191222570533, "grad_norm": 3.0943350791931152, "learning_rate": 5.708993569836237e-06, "loss": 1.2648, "step": 23263 }, { "epoch": 1.8231974921630094, "grad_norm": 2.715167760848999, "learning_rate": 5.708688322270165e-06, "loss": 1.2262, "step": 23264 }, { "epoch": 1.8232758620689655, "grad_norm": 3.0366384983062744, "learning_rate": 5.708383072008607e-06, "loss": 1.9005, "step": 23265 }, { "epoch": 1.8233542319749216, "grad_norm": 2.937828302383423, "learning_rate": 5.708077819052724e-06, "loss": 0.9315, "step": 23266 }, { "epoch": 1.8234326018808777, "grad_norm": 2.2523443698883057, "learning_rate": 5.707772563403678e-06, "loss": 1.3749, "step": 23267 }, { "epoch": 1.8235109717868339, "grad_norm": 2.931497097015381, "learning_rate": 5.7074673050626265e-06, "loss": 1.3718, "step": 23268 }, { "epoch": 1.82358934169279, "grad_norm": 2.705724000930786, "learning_rate": 5.707162044030734e-06, "loss": 1.5249, "step": 23269 }, { "epoch": 1.823667711598746, "grad_norm": 5.340031623840332, "learning_rate": 5.70685678030916e-06, "loss": 1.4291, "step": 23270 }, { "epoch": 1.8237460815047022, "grad_norm": 2.978714942932129, "learning_rate": 5.706551513899066e-06, "loss": 0.9408, "step": 23271 }, { "epoch": 1.8238244514106583, "grad_norm": 2.7415502071380615, "learning_rate": 5.7062462448016135e-06, "loss": 1.8549, "step": 23272 }, { "epoch": 1.8239028213166144, "grad_norm": 1.9118082523345947, "learning_rate": 5.705940973017961e-06, "loss": 1.0215, "step": 23273 }, { "epoch": 1.8239811912225705, "grad_norm": 2.3934881687164307, "learning_rate": 5.7056356985492735e-06, "loss": 1.6416, "step": 23274 }, { "epoch": 1.8240595611285266, "grad_norm": 2.0865697860717773, "learning_rate": 5.70533042139671e-06, "loss": 1.4701, "step": 23275 }, { "epoch": 1.8241379310344827, "grad_norm": 7.641992092132568, "learning_rate": 5.705025141561431e-06, "loss": 1.6916, "step": 23276 }, { "epoch": 1.8242163009404389, "grad_norm": 2.576564073562622, "learning_rate": 5.704719859044599e-06, "loss": 1.2414, "step": 23277 }, { "epoch": 1.824294670846395, "grad_norm": 3.5903258323669434, "learning_rate": 5.7044145738473745e-06, "loss": 1.7818, "step": 23278 }, { "epoch": 1.824373040752351, "grad_norm": 21.33578109741211, "learning_rate": 5.704109285970919e-06, "loss": 1.7085, "step": 23279 }, { "epoch": 1.8244514106583072, "grad_norm": 7.909531593322754, "learning_rate": 5.703803995416393e-06, "loss": 1.2999, "step": 23280 }, { "epoch": 1.8245297805642633, "grad_norm": 3.9853923320770264, "learning_rate": 5.703498702184958e-06, "loss": 1.1634, "step": 23281 }, { "epoch": 1.8246081504702194, "grad_norm": 3.9642326831817627, "learning_rate": 5.703193406277776e-06, "loss": 1.2936, "step": 23282 }, { "epoch": 1.8246865203761755, "grad_norm": 4.160394191741943, "learning_rate": 5.702888107696009e-06, "loss": 1.5529, "step": 23283 }, { "epoch": 1.8247648902821316, "grad_norm": 3.7091548442840576, "learning_rate": 5.702582806440814e-06, "loss": 1.3927, "step": 23284 }, { "epoch": 1.8248432601880877, "grad_norm": 3.056042432785034, "learning_rate": 5.702277502513356e-06, "loss": 1.0515, "step": 23285 }, { "epoch": 1.8249216300940438, "grad_norm": 2.594917058944702, "learning_rate": 5.7019721959147946e-06, "loss": 1.3228, "step": 23286 }, { "epoch": 1.825, "grad_norm": 2.4619436264038086, "learning_rate": 5.701666886646291e-06, "loss": 1.2257, "step": 23287 }, { "epoch": 1.825078369905956, "grad_norm": 4.04658842086792, "learning_rate": 5.701361574709008e-06, "loss": 1.6488, "step": 23288 }, { "epoch": 1.8251567398119122, "grad_norm": 3.4698541164398193, "learning_rate": 5.701056260104105e-06, "loss": 1.1214, "step": 23289 }, { "epoch": 1.8252351097178683, "grad_norm": 5.569337844848633, "learning_rate": 5.700750942832744e-06, "loss": 0.9636, "step": 23290 }, { "epoch": 1.8253134796238244, "grad_norm": 2.5024807453155518, "learning_rate": 5.700445622896089e-06, "loss": 1.4767, "step": 23291 }, { "epoch": 1.8253918495297805, "grad_norm": 5.03836727142334, "learning_rate": 5.700140300295295e-06, "loss": 1.2843, "step": 23292 }, { "epoch": 1.8254702194357368, "grad_norm": 3.565420389175415, "learning_rate": 5.6998349750315295e-06, "loss": 1.1253, "step": 23293 }, { "epoch": 1.8255485893416927, "grad_norm": 2.3749306201934814, "learning_rate": 5.699529647105949e-06, "loss": 1.2396, "step": 23294 }, { "epoch": 1.825626959247649, "grad_norm": 4.940044403076172, "learning_rate": 5.699224316519719e-06, "loss": 1.7939, "step": 23295 }, { "epoch": 1.825705329153605, "grad_norm": 3.112776517868042, "learning_rate": 5.698918983273998e-06, "loss": 1.4429, "step": 23296 }, { "epoch": 1.8257836990595613, "grad_norm": 2.638981342315674, "learning_rate": 5.6986136473699484e-06, "loss": 1.5516, "step": 23297 }, { "epoch": 1.8258620689655172, "grad_norm": 4.44307804107666, "learning_rate": 5.698308308808732e-06, "loss": 1.2879, "step": 23298 }, { "epoch": 1.8259404388714735, "grad_norm": 4.149982929229736, "learning_rate": 5.6980029675915074e-06, "loss": 1.6823, "step": 23299 }, { "epoch": 1.8260188087774294, "grad_norm": 5.773571491241455, "learning_rate": 5.69769762371944e-06, "loss": 1.3267, "step": 23300 }, { "epoch": 1.8260971786833857, "grad_norm": 12.41440200805664, "learning_rate": 5.697392277193687e-06, "loss": 1.2502, "step": 23301 }, { "epoch": 1.8261755485893416, "grad_norm": 4.374749660491943, "learning_rate": 5.697086928015415e-06, "loss": 1.3514, "step": 23302 }, { "epoch": 1.826253918495298, "grad_norm": 2.9615886211395264, "learning_rate": 5.69678157618578e-06, "loss": 1.4224, "step": 23303 }, { "epoch": 1.8263322884012538, "grad_norm": 4.017961502075195, "learning_rate": 5.6964762217059465e-06, "loss": 1.4566, "step": 23304 }, { "epoch": 1.8264106583072102, "grad_norm": 4.722839832305908, "learning_rate": 5.696170864577075e-06, "loss": 1.1218, "step": 23305 }, { "epoch": 1.826489028213166, "grad_norm": 4.711735248565674, "learning_rate": 5.695865504800328e-06, "loss": 1.5434, "step": 23306 }, { "epoch": 1.8265673981191224, "grad_norm": 4.403921604156494, "learning_rate": 5.695560142376865e-06, "loss": 1.3455, "step": 23307 }, { "epoch": 1.8266457680250783, "grad_norm": 2.084416389465332, "learning_rate": 5.695254777307849e-06, "loss": 1.2993, "step": 23308 }, { "epoch": 1.8267241379310346, "grad_norm": 2.701631784439087, "learning_rate": 5.69494940959444e-06, "loss": 1.3416, "step": 23309 }, { "epoch": 1.8268025078369905, "grad_norm": 2.6418867111206055, "learning_rate": 5.694644039237802e-06, "loss": 1.4412, "step": 23310 }, { "epoch": 1.8268808777429468, "grad_norm": 4.587369441986084, "learning_rate": 5.694338666239095e-06, "loss": 2.1111, "step": 23311 }, { "epoch": 1.8269592476489027, "grad_norm": 2.981238603591919, "learning_rate": 5.6940332905994786e-06, "loss": 1.3372, "step": 23312 }, { "epoch": 1.827037617554859, "grad_norm": 2.235466241836548, "learning_rate": 5.693727912320117e-06, "loss": 1.2199, "step": 23313 }, { "epoch": 1.827115987460815, "grad_norm": 2.113583564758301, "learning_rate": 5.6934225314021705e-06, "loss": 1.4474, "step": 23314 }, { "epoch": 1.8271943573667713, "grad_norm": 3.1231915950775146, "learning_rate": 5.6931171478468e-06, "loss": 1.5559, "step": 23315 }, { "epoch": 1.8272727272727272, "grad_norm": 2.0061097145080566, "learning_rate": 5.6928117616551695e-06, "loss": 1.5941, "step": 23316 }, { "epoch": 1.8273510971786835, "grad_norm": 10.757543563842773, "learning_rate": 5.692506372828439e-06, "loss": 2.4443, "step": 23317 }, { "epoch": 1.8274294670846394, "grad_norm": 5.626668930053711, "learning_rate": 5.692200981367768e-06, "loss": 1.7303, "step": 23318 }, { "epoch": 1.8275078369905957, "grad_norm": 3.6448299884796143, "learning_rate": 5.691895587274323e-06, "loss": 1.385, "step": 23319 }, { "epoch": 1.8275862068965516, "grad_norm": 4.089784622192383, "learning_rate": 5.6915901905492586e-06, "loss": 1.2763, "step": 23320 }, { "epoch": 1.827664576802508, "grad_norm": 3.800755262374878, "learning_rate": 5.691284791193744e-06, "loss": 1.5587, "step": 23321 }, { "epoch": 1.8277429467084638, "grad_norm": 2.229515314102173, "learning_rate": 5.690979389208936e-06, "loss": 1.5042, "step": 23322 }, { "epoch": 1.8278213166144202, "grad_norm": 2.3962202072143555, "learning_rate": 5.690673984595995e-06, "loss": 1.4618, "step": 23323 }, { "epoch": 1.827899686520376, "grad_norm": 3.672121524810791, "learning_rate": 5.690368577356088e-06, "loss": 1.3447, "step": 23324 }, { "epoch": 1.8279780564263324, "grad_norm": 2.117629289627075, "learning_rate": 5.690063167490373e-06, "loss": 1.2101, "step": 23325 }, { "epoch": 1.8280564263322883, "grad_norm": 2.831361770629883, "learning_rate": 5.689757755000011e-06, "loss": 1.1183, "step": 23326 }, { "epoch": 1.8281347962382446, "grad_norm": 4.265957832336426, "learning_rate": 5.6894523398861645e-06, "loss": 1.4297, "step": 23327 }, { "epoch": 1.8282131661442005, "grad_norm": 3.033250570297241, "learning_rate": 5.689146922149998e-06, "loss": 1.1782, "step": 23328 }, { "epoch": 1.8282915360501568, "grad_norm": 1.8296473026275635, "learning_rate": 5.688841501792668e-06, "loss": 1.3446, "step": 23329 }, { "epoch": 1.8283699059561127, "grad_norm": 4.089858055114746, "learning_rate": 5.68853607881534e-06, "loss": 1.5541, "step": 23330 }, { "epoch": 1.828448275862069, "grad_norm": 3.9820127487182617, "learning_rate": 5.688230653219173e-06, "loss": 1.2224, "step": 23331 }, { "epoch": 1.828526645768025, "grad_norm": 3.31646466255188, "learning_rate": 5.687925225005332e-06, "loss": 1.5115, "step": 23332 }, { "epoch": 1.8286050156739813, "grad_norm": 2.3153555393218994, "learning_rate": 5.687619794174976e-06, "loss": 1.506, "step": 23333 }, { "epoch": 1.8286833855799371, "grad_norm": 3.065321922302246, "learning_rate": 5.687314360729267e-06, "loss": 1.2689, "step": 23334 }, { "epoch": 1.8287617554858935, "grad_norm": 5.678764820098877, "learning_rate": 5.6870089246693675e-06, "loss": 1.3456, "step": 23335 }, { "epoch": 1.8288401253918494, "grad_norm": 4.384570598602295, "learning_rate": 5.6867034859964385e-06, "loss": 0.7233, "step": 23336 }, { "epoch": 1.8289184952978057, "grad_norm": 2.2817142009735107, "learning_rate": 5.686398044711642e-06, "loss": 1.1106, "step": 23337 }, { "epoch": 1.8289968652037616, "grad_norm": 3.5690619945526123, "learning_rate": 5.686092600816142e-06, "loss": 1.158, "step": 23338 }, { "epoch": 1.829075235109718, "grad_norm": 2.9351046085357666, "learning_rate": 5.685787154311096e-06, "loss": 1.2786, "step": 23339 }, { "epoch": 1.829153605015674, "grad_norm": 3.962301254272461, "learning_rate": 5.685481705197669e-06, "loss": 1.0721, "step": 23340 }, { "epoch": 1.8292319749216301, "grad_norm": 1.973017692565918, "learning_rate": 5.6851762534770205e-06, "loss": 1.1357, "step": 23341 }, { "epoch": 1.8293103448275863, "grad_norm": 2.108818531036377, "learning_rate": 5.6848707991503145e-06, "loss": 1.3299, "step": 23342 }, { "epoch": 1.8293887147335424, "grad_norm": 2.5209829807281494, "learning_rate": 5.684565342218711e-06, "loss": 1.1781, "step": 23343 }, { "epoch": 1.8294670846394985, "grad_norm": 2.7893950939178467, "learning_rate": 5.6842598826833725e-06, "loss": 1.5171, "step": 23344 }, { "epoch": 1.8295454545454546, "grad_norm": 3.2338473796844482, "learning_rate": 5.683954420545463e-06, "loss": 1.6059, "step": 23345 }, { "epoch": 1.8296238244514107, "grad_norm": 7.679914474487305, "learning_rate": 5.683648955806139e-06, "loss": 2.0764, "step": 23346 }, { "epoch": 1.8297021943573668, "grad_norm": 2.932521104812622, "learning_rate": 5.683343488466568e-06, "loss": 1.5277, "step": 23347 }, { "epoch": 1.829780564263323, "grad_norm": 2.406951904296875, "learning_rate": 5.6830380185279075e-06, "loss": 1.2638, "step": 23348 }, { "epoch": 1.829858934169279, "grad_norm": 2.9540393352508545, "learning_rate": 5.682732545991323e-06, "loss": 1.5503, "step": 23349 }, { "epoch": 1.8299373040752351, "grad_norm": 2.409513473510742, "learning_rate": 5.682427070857973e-06, "loss": 1.2263, "step": 23350 }, { "epoch": 1.8300156739811912, "grad_norm": 4.37819242477417, "learning_rate": 5.682121593129022e-06, "loss": 1.3072, "step": 23351 }, { "epoch": 1.8300940438871474, "grad_norm": 3.038583993911743, "learning_rate": 5.68181611280563e-06, "loss": 1.4728, "step": 23352 }, { "epoch": 1.8301724137931035, "grad_norm": 2.573535680770874, "learning_rate": 5.68151062988896e-06, "loss": 1.3485, "step": 23353 }, { "epoch": 1.8302507836990596, "grad_norm": 2.8413422107696533, "learning_rate": 5.6812051443801754e-06, "loss": 1.4693, "step": 23354 }, { "epoch": 1.8303291536050157, "grad_norm": 2.1576130390167236, "learning_rate": 5.680899656280433e-06, "loss": 1.3379, "step": 23355 }, { "epoch": 1.8304075235109718, "grad_norm": 3.1370913982391357, "learning_rate": 5.680594165590901e-06, "loss": 1.2448, "step": 23356 }, { "epoch": 1.830485893416928, "grad_norm": 2.8480260372161865, "learning_rate": 5.680288672312737e-06, "loss": 1.1717, "step": 23357 }, { "epoch": 1.830564263322884, "grad_norm": 5.56207799911499, "learning_rate": 5.6799831764471044e-06, "loss": 1.3508, "step": 23358 }, { "epoch": 1.8306426332288401, "grad_norm": 3.320967674255371, "learning_rate": 5.679677677995165e-06, "loss": 1.3066, "step": 23359 }, { "epoch": 1.8307210031347962, "grad_norm": 3.4536309242248535, "learning_rate": 5.67937217695808e-06, "loss": 1.1741, "step": 23360 }, { "epoch": 1.8307993730407524, "grad_norm": 4.0621747970581055, "learning_rate": 5.679066673337013e-06, "loss": 1.5543, "step": 23361 }, { "epoch": 1.8308777429467085, "grad_norm": 3.8284711837768555, "learning_rate": 5.678761167133126e-06, "loss": 1.3976, "step": 23362 }, { "epoch": 1.8309561128526646, "grad_norm": 2.8636045455932617, "learning_rate": 5.678455658347579e-06, "loss": 1.3338, "step": 23363 }, { "epoch": 1.8310344827586207, "grad_norm": 2.5690367221832275, "learning_rate": 5.678150146981536e-06, "loss": 1.4153, "step": 23364 }, { "epoch": 1.8311128526645768, "grad_norm": 2.177156925201416, "learning_rate": 5.677844633036157e-06, "loss": 1.6215, "step": 23365 }, { "epoch": 1.831191222570533, "grad_norm": 2.371443271636963, "learning_rate": 5.677539116512607e-06, "loss": 1.4952, "step": 23366 }, { "epoch": 1.831269592476489, "grad_norm": 2.72481107711792, "learning_rate": 5.677233597412045e-06, "loss": 1.4078, "step": 23367 }, { "epoch": 1.8313479623824451, "grad_norm": 2.138186454772949, "learning_rate": 5.676928075735634e-06, "loss": 1.3815, "step": 23368 }, { "epoch": 1.8314263322884012, "grad_norm": 3.937443971633911, "learning_rate": 5.676622551484536e-06, "loss": 1.8345, "step": 23369 }, { "epoch": 1.8315047021943573, "grad_norm": 2.4857664108276367, "learning_rate": 5.676317024659914e-06, "loss": 1.3853, "step": 23370 }, { "epoch": 1.8315830721003135, "grad_norm": 5.6934943199157715, "learning_rate": 5.676011495262931e-06, "loss": 1.6594, "step": 23371 }, { "epoch": 1.8316614420062696, "grad_norm": 2.216541051864624, "learning_rate": 5.675705963294744e-06, "loss": 1.4054, "step": 23372 }, { "epoch": 1.8317398119122257, "grad_norm": 2.9482262134552, "learning_rate": 5.675400428756521e-06, "loss": 1.3172, "step": 23373 }, { "epoch": 1.8318181818181818, "grad_norm": 2.3612735271453857, "learning_rate": 5.6750948916494205e-06, "loss": 1.3767, "step": 23374 }, { "epoch": 1.831896551724138, "grad_norm": 3.5328176021575928, "learning_rate": 5.674789351974607e-06, "loss": 1.1068, "step": 23375 }, { "epoch": 1.831974921630094, "grad_norm": 2.5262985229492188, "learning_rate": 5.674483809733241e-06, "loss": 1.337, "step": 23376 }, { "epoch": 1.8320532915360501, "grad_norm": 10.517257690429688, "learning_rate": 5.674178264926485e-06, "loss": 1.4568, "step": 23377 }, { "epoch": 1.8321316614420062, "grad_norm": 4.470142364501953, "learning_rate": 5.673872717555501e-06, "loss": 1.3732, "step": 23378 }, { "epoch": 1.8322100313479623, "grad_norm": 2.5736491680145264, "learning_rate": 5.673567167621451e-06, "loss": 1.9241, "step": 23379 }, { "epoch": 1.8322884012539185, "grad_norm": 3.316554069519043, "learning_rate": 5.673261615125498e-06, "loss": 1.2475, "step": 23380 }, { "epoch": 1.8323667711598746, "grad_norm": 2.453186511993408, "learning_rate": 5.672956060068803e-06, "loss": 1.2062, "step": 23381 }, { "epoch": 1.8324451410658307, "grad_norm": 4.982770919799805, "learning_rate": 5.67265050245253e-06, "loss": 1.6789, "step": 23382 }, { "epoch": 1.8325235109717868, "grad_norm": 9.497761726379395, "learning_rate": 5.672344942277838e-06, "loss": 1.4286, "step": 23383 }, { "epoch": 1.832601880877743, "grad_norm": 2.281402587890625, "learning_rate": 5.672039379545894e-06, "loss": 1.3425, "step": 23384 }, { "epoch": 1.832680250783699, "grad_norm": 5.323732376098633, "learning_rate": 5.671733814257855e-06, "loss": 1.0842, "step": 23385 }, { "epoch": 1.8327586206896553, "grad_norm": 2.1574413776397705, "learning_rate": 5.671428246414886e-06, "loss": 1.3449, "step": 23386 }, { "epoch": 1.8328369905956112, "grad_norm": 2.94162654876709, "learning_rate": 5.671122676018149e-06, "loss": 1.2608, "step": 23387 }, { "epoch": 1.8329153605015676, "grad_norm": 2.452671766281128, "learning_rate": 5.670817103068806e-06, "loss": 1.5095, "step": 23388 }, { "epoch": 1.8329937304075234, "grad_norm": 2.2609949111938477, "learning_rate": 5.6705115275680205e-06, "loss": 1.0896, "step": 23389 }, { "epoch": 1.8330721003134798, "grad_norm": 2.737985134124756, "learning_rate": 5.670205949516952e-06, "loss": 1.1873, "step": 23390 }, { "epoch": 1.8331504702194357, "grad_norm": 5.900731086730957, "learning_rate": 5.669900368916764e-06, "loss": 1.5428, "step": 23391 }, { "epoch": 1.833228840125392, "grad_norm": 2.3265793323516846, "learning_rate": 5.669594785768622e-06, "loss": 1.255, "step": 23392 }, { "epoch": 1.833307210031348, "grad_norm": 2.083190441131592, "learning_rate": 5.669289200073682e-06, "loss": 1.2983, "step": 23393 }, { "epoch": 1.8333855799373042, "grad_norm": 3.001246929168701, "learning_rate": 5.668983611833112e-06, "loss": 1.2927, "step": 23394 }, { "epoch": 1.83346394984326, "grad_norm": 2.4902503490448, "learning_rate": 5.668678021048071e-06, "loss": 1.493, "step": 23395 }, { "epoch": 1.8335423197492164, "grad_norm": 2.8069307804107666, "learning_rate": 5.668372427719722e-06, "loss": 1.1595, "step": 23396 }, { "epoch": 1.8336206896551723, "grad_norm": 4.01255989074707, "learning_rate": 5.668066831849227e-06, "loss": 1.0371, "step": 23397 }, { "epoch": 1.8336990595611287, "grad_norm": 2.6545071601867676, "learning_rate": 5.66776123343775e-06, "loss": 1.1851, "step": 23398 }, { "epoch": 1.8337774294670846, "grad_norm": 2.5318100452423096, "learning_rate": 5.667455632486453e-06, "loss": 1.2956, "step": 23399 }, { "epoch": 1.8338557993730409, "grad_norm": 2.956723928451538, "learning_rate": 5.667150028996495e-06, "loss": 1.2317, "step": 23400 }, { "epoch": 1.8339341692789968, "grad_norm": 7.303894519805908, "learning_rate": 5.666844422969043e-06, "loss": 1.0368, "step": 23401 }, { "epoch": 1.834012539184953, "grad_norm": 2.6939644813537598, "learning_rate": 5.6665388144052556e-06, "loss": 1.4385, "step": 23402 }, { "epoch": 1.834090909090909, "grad_norm": 2.983790159225464, "learning_rate": 5.666233203306299e-06, "loss": 1.3452, "step": 23403 }, { "epoch": 1.8341692789968653, "grad_norm": 4.03061580657959, "learning_rate": 5.665927589673332e-06, "loss": 1.5406, "step": 23404 }, { "epoch": 1.8342476489028212, "grad_norm": 2.443721055984497, "learning_rate": 5.665621973507519e-06, "loss": 1.5384, "step": 23405 }, { "epoch": 1.8343260188087775, "grad_norm": 3.4817678928375244, "learning_rate": 5.665316354810022e-06, "loss": 1.1958, "step": 23406 }, { "epoch": 1.8344043887147334, "grad_norm": 5.980639457702637, "learning_rate": 5.665010733582004e-06, "loss": 1.106, "step": 23407 }, { "epoch": 1.8344827586206898, "grad_norm": 7.4803314208984375, "learning_rate": 5.664705109824626e-06, "loss": 1.2611, "step": 23408 }, { "epoch": 1.8345611285266457, "grad_norm": 4.64027214050293, "learning_rate": 5.66439948353905e-06, "loss": 1.0716, "step": 23409 }, { "epoch": 1.834639498432602, "grad_norm": 5.119842052459717, "learning_rate": 5.664093854726442e-06, "loss": 1.3232, "step": 23410 }, { "epoch": 1.8347178683385579, "grad_norm": 2.19448184967041, "learning_rate": 5.6637882233879615e-06, "loss": 1.3028, "step": 23411 }, { "epoch": 1.8347962382445142, "grad_norm": 2.9684431552886963, "learning_rate": 5.66348258952477e-06, "loss": 1.4534, "step": 23412 }, { "epoch": 1.83487460815047, "grad_norm": 2.4805877208709717, "learning_rate": 5.663176953138033e-06, "loss": 1.322, "step": 23413 }, { "epoch": 1.8349529780564264, "grad_norm": 2.119049310684204, "learning_rate": 5.66287131422891e-06, "loss": 1.4011, "step": 23414 }, { "epoch": 1.8350313479623823, "grad_norm": 5.723996162414551, "learning_rate": 5.662565672798566e-06, "loss": 1.7229, "step": 23415 }, { "epoch": 1.8351097178683387, "grad_norm": 3.3804526329040527, "learning_rate": 5.662260028848163e-06, "loss": 1.4127, "step": 23416 }, { "epoch": 1.8351880877742945, "grad_norm": 5.252063274383545, "learning_rate": 5.661954382378863e-06, "loss": 0.8742, "step": 23417 }, { "epoch": 1.8352664576802509, "grad_norm": 5.608338356018066, "learning_rate": 5.661648733391828e-06, "loss": 1.3465, "step": 23418 }, { "epoch": 1.8353448275862068, "grad_norm": 4.668772220611572, "learning_rate": 5.66134308188822e-06, "loss": 1.4618, "step": 23419 }, { "epoch": 1.835423197492163, "grad_norm": 1.745435118675232, "learning_rate": 5.661037427869205e-06, "loss": 1.3328, "step": 23420 }, { "epoch": 1.835501567398119, "grad_norm": 2.021486282348633, "learning_rate": 5.66073177133594e-06, "loss": 1.1302, "step": 23421 }, { "epoch": 1.8355799373040753, "grad_norm": 2.891117811203003, "learning_rate": 5.6604261122895945e-06, "loss": 1.066, "step": 23422 }, { "epoch": 1.8356583072100312, "grad_norm": 4.873149394989014, "learning_rate": 5.6601204507313245e-06, "loss": 1.6851, "step": 23423 }, { "epoch": 1.8357366771159875, "grad_norm": 2.2215418815612793, "learning_rate": 5.659814786662296e-06, "loss": 1.4932, "step": 23424 }, { "epoch": 1.8358150470219434, "grad_norm": 2.441906690597534, "learning_rate": 5.6595091200836714e-06, "loss": 1.3549, "step": 23425 }, { "epoch": 1.8358934169278998, "grad_norm": 2.7863190174102783, "learning_rate": 5.659203450996612e-06, "loss": 0.9025, "step": 23426 }, { "epoch": 1.8359717868338556, "grad_norm": 2.5714731216430664, "learning_rate": 5.658897779402283e-06, "loss": 1.4728, "step": 23427 }, { "epoch": 1.836050156739812, "grad_norm": 2.719367504119873, "learning_rate": 5.6585921053018435e-06, "loss": 1.862, "step": 23428 }, { "epoch": 1.8361285266457679, "grad_norm": 4.4669342041015625, "learning_rate": 5.658286428696459e-06, "loss": 1.2365, "step": 23429 }, { "epoch": 1.8362068965517242, "grad_norm": 2.549945592880249, "learning_rate": 5.6579807495872905e-06, "loss": 1.4165, "step": 23430 }, { "epoch": 1.83628526645768, "grad_norm": 4.650808811187744, "learning_rate": 5.657675067975501e-06, "loss": 1.4147, "step": 23431 }, { "epoch": 1.8363636363636364, "grad_norm": 4.21261739730835, "learning_rate": 5.657369383862253e-06, "loss": 1.5926, "step": 23432 }, { "epoch": 1.8364420062695925, "grad_norm": 3.315377950668335, "learning_rate": 5.65706369724871e-06, "loss": 1.329, "step": 23433 }, { "epoch": 1.8365203761755486, "grad_norm": 6.178709506988525, "learning_rate": 5.656758008136035e-06, "loss": 1.3886, "step": 23434 }, { "epoch": 1.8365987460815048, "grad_norm": 2.0055384635925293, "learning_rate": 5.656452316525389e-06, "loss": 1.1075, "step": 23435 }, { "epoch": 1.8366771159874609, "grad_norm": 2.966150999069214, "learning_rate": 5.656146622417936e-06, "loss": 1.7037, "step": 23436 }, { "epoch": 1.836755485893417, "grad_norm": 2.973194122314453, "learning_rate": 5.6558409258148375e-06, "loss": 1.5109, "step": 23437 }, { "epoch": 1.836833855799373, "grad_norm": 3.1903738975524902, "learning_rate": 5.655535226717258e-06, "loss": 0.9303, "step": 23438 }, { "epoch": 1.8369122257053292, "grad_norm": 3.5859534740448, "learning_rate": 5.655229525126359e-06, "loss": 0.4991, "step": 23439 }, { "epoch": 1.8369905956112853, "grad_norm": 2.363893508911133, "learning_rate": 5.6549238210433035e-06, "loss": 1.2378, "step": 23440 }, { "epoch": 1.8370689655172414, "grad_norm": 5.000827312469482, "learning_rate": 5.654618114469252e-06, "loss": 1.3862, "step": 23441 }, { "epoch": 1.8371473354231975, "grad_norm": 3.362212896347046, "learning_rate": 5.654312405405372e-06, "loss": 1.1999, "step": 23442 }, { "epoch": 1.8372257053291536, "grad_norm": 2.964369297027588, "learning_rate": 5.6540066938528224e-06, "loss": 1.3464, "step": 23443 }, { "epoch": 1.8373040752351097, "grad_norm": 11.946467399597168, "learning_rate": 5.653700979812769e-06, "loss": 0.8948, "step": 23444 }, { "epoch": 1.8373824451410659, "grad_norm": 2.2438015937805176, "learning_rate": 5.653395263286371e-06, "loss": 1.1157, "step": 23445 }, { "epoch": 1.837460815047022, "grad_norm": 4.637772083282471, "learning_rate": 5.653089544274795e-06, "loss": 1.3267, "step": 23446 }, { "epoch": 1.837539184952978, "grad_norm": 3.831064462661743, "learning_rate": 5.652783822779199e-06, "loss": 1.3074, "step": 23447 }, { "epoch": 1.8376175548589342, "grad_norm": 9.042391777038574, "learning_rate": 5.65247809880075e-06, "loss": 1.6597, "step": 23448 }, { "epoch": 1.8376959247648903, "grad_norm": 19.200212478637695, "learning_rate": 5.652172372340609e-06, "loss": 1.2395, "step": 23449 }, { "epoch": 1.8377742946708464, "grad_norm": 2.62988018989563, "learning_rate": 5.6518666433999404e-06, "loss": 1.4513, "step": 23450 }, { "epoch": 1.8378526645768025, "grad_norm": 4.524738311767578, "learning_rate": 5.651560911979905e-06, "loss": 1.263, "step": 23451 }, { "epoch": 1.8379310344827586, "grad_norm": 6.719449996948242, "learning_rate": 5.651255178081667e-06, "loss": 0.6549, "step": 23452 }, { "epoch": 1.8380094043887147, "grad_norm": 5.353320121765137, "learning_rate": 5.6509494417063895e-06, "loss": 1.6733, "step": 23453 }, { "epoch": 1.8380877742946709, "grad_norm": 3.9914660453796387, "learning_rate": 5.650643702855233e-06, "loss": 1.445, "step": 23454 }, { "epoch": 1.838166144200627, "grad_norm": 3.1092307567596436, "learning_rate": 5.650337961529364e-06, "loss": 1.3698, "step": 23455 }, { "epoch": 1.838244514106583, "grad_norm": 4.594555854797363, "learning_rate": 5.65003221772994e-06, "loss": 1.7703, "step": 23456 }, { "epoch": 1.8383228840125392, "grad_norm": 2.3812944889068604, "learning_rate": 5.649726471458132e-06, "loss": 1.2965, "step": 23457 }, { "epoch": 1.8384012539184953, "grad_norm": 3.2319223880767822, "learning_rate": 5.649420722715096e-06, "loss": 1.3948, "step": 23458 }, { "epoch": 1.8384796238244514, "grad_norm": 2.8595781326293945, "learning_rate": 5.649114971501996e-06, "loss": 1.3657, "step": 23459 }, { "epoch": 1.8385579937304075, "grad_norm": 2.6321308612823486, "learning_rate": 5.648809217819997e-06, "loss": 1.4539, "step": 23460 }, { "epoch": 1.8386363636363636, "grad_norm": 3.0072431564331055, "learning_rate": 5.6485034616702605e-06, "loss": 1.31, "step": 23461 }, { "epoch": 1.8387147335423197, "grad_norm": 5.878625392913818, "learning_rate": 5.64819770305395e-06, "loss": 1.5446, "step": 23462 }, { "epoch": 1.8387931034482758, "grad_norm": 5.124680995941162, "learning_rate": 5.647891941972229e-06, "loss": 1.1136, "step": 23463 }, { "epoch": 1.838871473354232, "grad_norm": 3.7015175819396973, "learning_rate": 5.647586178426261e-06, "loss": 0.9007, "step": 23464 }, { "epoch": 1.838949843260188, "grad_norm": 2.4543495178222656, "learning_rate": 5.647280412417207e-06, "loss": 1.1591, "step": 23465 }, { "epoch": 1.8390282131661442, "grad_norm": 3.208756446838379, "learning_rate": 5.646974643946229e-06, "loss": 1.5517, "step": 23466 }, { "epoch": 1.8391065830721003, "grad_norm": 3.1550421714782715, "learning_rate": 5.646668873014493e-06, "loss": 1.3196, "step": 23467 }, { "epoch": 1.8391849529780564, "grad_norm": 3.4661245346069336, "learning_rate": 5.646363099623161e-06, "loss": 1.2834, "step": 23468 }, { "epoch": 1.8392633228840125, "grad_norm": 3.571376085281372, "learning_rate": 5.646057323773395e-06, "loss": 0.7987, "step": 23469 }, { "epoch": 1.8393416927899686, "grad_norm": 15.364500045776367, "learning_rate": 5.6457515454663595e-06, "loss": 1.6303, "step": 23470 }, { "epoch": 1.8394200626959247, "grad_norm": 4.681215763092041, "learning_rate": 5.645445764703217e-06, "loss": 1.0558, "step": 23471 }, { "epoch": 1.8394984326018808, "grad_norm": 7.902752876281738, "learning_rate": 5.64513998148513e-06, "loss": 1.251, "step": 23472 }, { "epoch": 1.839576802507837, "grad_norm": 2.505215644836426, "learning_rate": 5.64483419581326e-06, "loss": 1.3739, "step": 23473 }, { "epoch": 1.839655172413793, "grad_norm": 3.8827757835388184, "learning_rate": 5.644528407688775e-06, "loss": 1.6889, "step": 23474 }, { "epoch": 1.8397335423197492, "grad_norm": 2.6072490215301514, "learning_rate": 5.6442226171128336e-06, "loss": 1.4835, "step": 23475 }, { "epoch": 1.8398119122257053, "grad_norm": 4.890881061553955, "learning_rate": 5.6439168240866e-06, "loss": 1.7412, "step": 23476 }, { "epoch": 1.8398902821316614, "grad_norm": 2.9335732460021973, "learning_rate": 5.643611028611239e-06, "loss": 1.5123, "step": 23477 }, { "epoch": 1.8399686520376175, "grad_norm": 4.32396125793457, "learning_rate": 5.64330523068791e-06, "loss": 0.8637, "step": 23478 }, { "epoch": 1.8400470219435736, "grad_norm": 2.131800889968872, "learning_rate": 5.6429994303177795e-06, "loss": 1.1444, "step": 23479 }, { "epoch": 1.8401253918495297, "grad_norm": 3.7884955406188965, "learning_rate": 5.642693627502009e-06, "loss": 1.4493, "step": 23480 }, { "epoch": 1.840203761755486, "grad_norm": 4.03344202041626, "learning_rate": 5.642387822241764e-06, "loss": 1.8422, "step": 23481 }, { "epoch": 1.840282131661442, "grad_norm": 1.9733397960662842, "learning_rate": 5.642082014538203e-06, "loss": 1.6288, "step": 23482 }, { "epoch": 1.8403605015673983, "grad_norm": 2.5185635089874268, "learning_rate": 5.641776204392494e-06, "loss": 0.9074, "step": 23483 }, { "epoch": 1.8404388714733542, "grad_norm": 5.576341152191162, "learning_rate": 5.641470391805797e-06, "loss": 1.5739, "step": 23484 }, { "epoch": 1.8405172413793105, "grad_norm": 3.179859161376953, "learning_rate": 5.641164576779276e-06, "loss": 1.5396, "step": 23485 }, { "epoch": 1.8405956112852664, "grad_norm": 2.4074246883392334, "learning_rate": 5.640858759314094e-06, "loss": 1.3012, "step": 23486 }, { "epoch": 1.8406739811912227, "grad_norm": 1.8838196992874146, "learning_rate": 5.640552939411414e-06, "loss": 1.2935, "step": 23487 }, { "epoch": 1.8407523510971786, "grad_norm": 3.1552467346191406, "learning_rate": 5.6402471170724005e-06, "loss": 1.1298, "step": 23488 }, { "epoch": 1.840830721003135, "grad_norm": 2.2961859703063965, "learning_rate": 5.639941292298215e-06, "loss": 1.4253, "step": 23489 }, { "epoch": 1.8409090909090908, "grad_norm": 2.633282423019409, "learning_rate": 5.639635465090022e-06, "loss": 1.047, "step": 23490 }, { "epoch": 1.8409874608150472, "grad_norm": 2.3986096382141113, "learning_rate": 5.639329635448985e-06, "loss": 1.1546, "step": 23491 }, { "epoch": 1.841065830721003, "grad_norm": 3.0044806003570557, "learning_rate": 5.639023803376267e-06, "loss": 1.1995, "step": 23492 }, { "epoch": 1.8411442006269594, "grad_norm": 2.6017158031463623, "learning_rate": 5.638717968873029e-06, "loss": 1.2939, "step": 23493 }, { "epoch": 1.8412225705329153, "grad_norm": 4.029433250427246, "learning_rate": 5.638412131940436e-06, "loss": 1.5286, "step": 23494 }, { "epoch": 1.8413009404388716, "grad_norm": 3.6453616619110107, "learning_rate": 5.6381062925796516e-06, "loss": 1.4296, "step": 23495 }, { "epoch": 1.8413793103448275, "grad_norm": 2.486064910888672, "learning_rate": 5.637800450791838e-06, "loss": 1.2832, "step": 23496 }, { "epoch": 1.8414576802507838, "grad_norm": 3.182093858718872, "learning_rate": 5.63749460657816e-06, "loss": 1.3848, "step": 23497 }, { "epoch": 1.8415360501567397, "grad_norm": 3.9570472240448, "learning_rate": 5.63718875993978e-06, "loss": 1.1149, "step": 23498 }, { "epoch": 1.841614420062696, "grad_norm": 2.086688995361328, "learning_rate": 5.636882910877861e-06, "loss": 1.4109, "step": 23499 }, { "epoch": 1.841692789968652, "grad_norm": 3.020994186401367, "learning_rate": 5.6365770593935665e-06, "loss": 1.1099, "step": 23500 }, { "epoch": 1.8417711598746083, "grad_norm": 3.196815013885498, "learning_rate": 5.636271205488059e-06, "loss": 1.2757, "step": 23501 }, { "epoch": 1.8418495297805642, "grad_norm": 2.651021957397461, "learning_rate": 5.6359653491625044e-06, "loss": 1.1432, "step": 23502 }, { "epoch": 1.8419278996865205, "grad_norm": 1.8695313930511475, "learning_rate": 5.635659490418062e-06, "loss": 1.0721, "step": 23503 }, { "epoch": 1.8420062695924764, "grad_norm": 2.8434665203094482, "learning_rate": 5.635353629255899e-06, "loss": 1.0852, "step": 23504 }, { "epoch": 1.8420846394984327, "grad_norm": 3.061880350112915, "learning_rate": 5.635047765677177e-06, "loss": 1.2387, "step": 23505 }, { "epoch": 1.8421630094043886, "grad_norm": 4.276272296905518, "learning_rate": 5.634741899683059e-06, "loss": 1.8485, "step": 23506 }, { "epoch": 1.842241379310345, "grad_norm": 3.2132279872894287, "learning_rate": 5.63443603127471e-06, "loss": 1.2805, "step": 23507 }, { "epoch": 1.8423197492163008, "grad_norm": 4.1770195960998535, "learning_rate": 5.634130160453291e-06, "loss": 1.0229, "step": 23508 }, { "epoch": 1.8423981191222571, "grad_norm": 6.441213607788086, "learning_rate": 5.6338242872199675e-06, "loss": 0.9166, "step": 23509 }, { "epoch": 1.842476489028213, "grad_norm": 3.665247917175293, "learning_rate": 5.6335184115759e-06, "loss": 1.5894, "step": 23510 }, { "epoch": 1.8425548589341694, "grad_norm": 3.0329477787017822, "learning_rate": 5.633212533522256e-06, "loss": 1.4803, "step": 23511 }, { "epoch": 1.8426332288401253, "grad_norm": 4.239943027496338, "learning_rate": 5.632906653060196e-06, "loss": 1.3158, "step": 23512 }, { "epoch": 1.8427115987460816, "grad_norm": 3.2697668075561523, "learning_rate": 5.632600770190883e-06, "loss": 1.5305, "step": 23513 }, { "epoch": 1.8427899686520375, "grad_norm": 2.9482545852661133, "learning_rate": 5.632294884915483e-06, "loss": 1.3531, "step": 23514 }, { "epoch": 1.8428683385579938, "grad_norm": 3.1390669345855713, "learning_rate": 5.631988997235159e-06, "loss": 1.3763, "step": 23515 }, { "epoch": 1.8429467084639497, "grad_norm": 1.8735401630401611, "learning_rate": 5.631683107151071e-06, "loss": 1.3081, "step": 23516 }, { "epoch": 1.843025078369906, "grad_norm": 2.2775566577911377, "learning_rate": 5.631377214664386e-06, "loss": 1.3809, "step": 23517 }, { "epoch": 1.843103448275862, "grad_norm": 3.149439573287964, "learning_rate": 5.631071319776267e-06, "loss": 1.3173, "step": 23518 }, { "epoch": 1.8431818181818183, "grad_norm": 2.868260383605957, "learning_rate": 5.630765422487877e-06, "loss": 1.3494, "step": 23519 }, { "epoch": 1.8432601880877741, "grad_norm": 4.775551795959473, "learning_rate": 5.630459522800379e-06, "loss": 1.0798, "step": 23520 }, { "epoch": 1.8433385579937305, "grad_norm": 4.22043514251709, "learning_rate": 5.630153620714935e-06, "loss": 1.4187, "step": 23521 }, { "epoch": 1.8434169278996864, "grad_norm": 2.95076847076416, "learning_rate": 5.629847716232713e-06, "loss": 1.4322, "step": 23522 }, { "epoch": 1.8434952978056427, "grad_norm": 4.26750373840332, "learning_rate": 5.6295418093548725e-06, "loss": 1.426, "step": 23523 }, { "epoch": 1.8435736677115986, "grad_norm": 3.7296462059020996, "learning_rate": 5.629235900082578e-06, "loss": 1.5316, "step": 23524 }, { "epoch": 1.843652037617555, "grad_norm": 3.477825403213501, "learning_rate": 5.628929988416994e-06, "loss": 1.2893, "step": 23525 }, { "epoch": 1.8437304075235108, "grad_norm": 6.78740119934082, "learning_rate": 5.628624074359283e-06, "loss": 0.4943, "step": 23526 }, { "epoch": 1.8438087774294671, "grad_norm": 2.4519855976104736, "learning_rate": 5.628318157910608e-06, "loss": 1.5591, "step": 23527 }, { "epoch": 1.8438871473354232, "grad_norm": 3.083444356918335, "learning_rate": 5.628012239072136e-06, "loss": 0.6037, "step": 23528 }, { "epoch": 1.8439655172413794, "grad_norm": 11.670271873474121, "learning_rate": 5.627706317845025e-06, "loss": 1.7688, "step": 23529 }, { "epoch": 1.8440438871473355, "grad_norm": 3.0065343379974365, "learning_rate": 5.627400394230443e-06, "loss": 1.3202, "step": 23530 }, { "epoch": 1.8441222570532916, "grad_norm": 2.2458369731903076, "learning_rate": 5.627094468229552e-06, "loss": 1.254, "step": 23531 }, { "epoch": 1.8442006269592477, "grad_norm": 4.499751091003418, "learning_rate": 5.626788539843516e-06, "loss": 1.5552, "step": 23532 }, { "epoch": 1.8442789968652038, "grad_norm": 6.138880252838135, "learning_rate": 5.626482609073498e-06, "loss": 0.8802, "step": 23533 }, { "epoch": 1.84435736677116, "grad_norm": 5.8865180015563965, "learning_rate": 5.626176675920663e-06, "loss": 1.0282, "step": 23534 }, { "epoch": 1.844435736677116, "grad_norm": 2.3561742305755615, "learning_rate": 5.625870740386173e-06, "loss": 1.4867, "step": 23535 }, { "epoch": 1.8445141065830721, "grad_norm": 2.721632957458496, "learning_rate": 5.625564802471191e-06, "loss": 1.5732, "step": 23536 }, { "epoch": 1.8445924764890282, "grad_norm": 2.498826026916504, "learning_rate": 5.6252588621768836e-06, "loss": 1.3233, "step": 23537 }, { "epoch": 1.8446708463949844, "grad_norm": 2.6852948665618896, "learning_rate": 5.62495291950441e-06, "loss": 1.294, "step": 23538 }, { "epoch": 1.8447492163009405, "grad_norm": 3.532158613204956, "learning_rate": 5.6246469744549395e-06, "loss": 1.6116, "step": 23539 }, { "epoch": 1.8448275862068966, "grad_norm": 3.2472927570343018, "learning_rate": 5.624341027029631e-06, "loss": 1.2871, "step": 23540 }, { "epoch": 1.8449059561128527, "grad_norm": 4.577721118927002, "learning_rate": 5.62403507722965e-06, "loss": 1.7484, "step": 23541 }, { "epoch": 1.8449843260188088, "grad_norm": 2.585674285888672, "learning_rate": 5.623729125056161e-06, "loss": 1.1601, "step": 23542 }, { "epoch": 1.845062695924765, "grad_norm": 3.3357980251312256, "learning_rate": 5.623423170510326e-06, "loss": 1.3297, "step": 23543 }, { "epoch": 1.845141065830721, "grad_norm": 3.0329980850219727, "learning_rate": 5.62311721359331e-06, "loss": 1.5677, "step": 23544 }, { "epoch": 1.8452194357366771, "grad_norm": 3.4800047874450684, "learning_rate": 5.622811254306275e-06, "loss": 1.1266, "step": 23545 }, { "epoch": 1.8452978056426332, "grad_norm": 5.4431986808776855, "learning_rate": 5.6225052926503876e-06, "loss": 1.8459, "step": 23546 }, { "epoch": 1.8453761755485893, "grad_norm": 3.141434669494629, "learning_rate": 5.62219932862681e-06, "loss": 1.1051, "step": 23547 }, { "epoch": 1.8454545454545455, "grad_norm": 2.353559732437134, "learning_rate": 5.621893362236705e-06, "loss": 1.24, "step": 23548 }, { "epoch": 1.8455329153605016, "grad_norm": 7.541935443878174, "learning_rate": 5.621587393481236e-06, "loss": 1.6497, "step": 23549 }, { "epoch": 1.8456112852664577, "grad_norm": 2.559555768966675, "learning_rate": 5.621281422361568e-06, "loss": 1.48, "step": 23550 }, { "epoch": 1.8456896551724138, "grad_norm": 1.9749683141708374, "learning_rate": 5.620975448878866e-06, "loss": 1.3902, "step": 23551 }, { "epoch": 1.84576802507837, "grad_norm": 6.735489845275879, "learning_rate": 5.62066947303429e-06, "loss": 1.1715, "step": 23552 }, { "epoch": 1.845846394984326, "grad_norm": 2.9463136196136475, "learning_rate": 5.6203634948290084e-06, "loss": 1.7086, "step": 23553 }, { "epoch": 1.8459247648902821, "grad_norm": 3.4243717193603516, "learning_rate": 5.620057514264183e-06, "loss": 1.166, "step": 23554 }, { "epoch": 1.8460031347962382, "grad_norm": 3.064772605895996, "learning_rate": 5.619751531340974e-06, "loss": 1.2565, "step": 23555 }, { "epoch": 1.8460815047021943, "grad_norm": 4.884822845458984, "learning_rate": 5.6194455460605525e-06, "loss": 1.206, "step": 23556 }, { "epoch": 1.8461598746081505, "grad_norm": 4.366864204406738, "learning_rate": 5.619139558424076e-06, "loss": 1.2984, "step": 23557 }, { "epoch": 1.8462382445141066, "grad_norm": 3.322434186935425, "learning_rate": 5.618833568432712e-06, "loss": 1.5735, "step": 23558 }, { "epoch": 1.8463166144200627, "grad_norm": 2.355452537536621, "learning_rate": 5.618527576087623e-06, "loss": 1.0933, "step": 23559 }, { "epoch": 1.8463949843260188, "grad_norm": 2.619122266769409, "learning_rate": 5.618221581389971e-06, "loss": 1.4535, "step": 23560 }, { "epoch": 1.846473354231975, "grad_norm": 3.1509785652160645, "learning_rate": 5.617915584340921e-06, "loss": 0.9428, "step": 23561 }, { "epoch": 1.846551724137931, "grad_norm": 2.548771619796753, "learning_rate": 5.61760958494164e-06, "loss": 1.3355, "step": 23562 }, { "epoch": 1.8466300940438871, "grad_norm": 2.9181790351867676, "learning_rate": 5.61730358319329e-06, "loss": 1.4638, "step": 23563 }, { "epoch": 1.8467084639498432, "grad_norm": 2.5415008068084717, "learning_rate": 5.616997579097031e-06, "loss": 1.2348, "step": 23564 }, { "epoch": 1.8467868338557993, "grad_norm": 2.1314444541931152, "learning_rate": 5.6166915726540325e-06, "loss": 1.1532, "step": 23565 }, { "epoch": 1.8468652037617554, "grad_norm": 4.843535900115967, "learning_rate": 5.6163855638654555e-06, "loss": 1.6816, "step": 23566 }, { "epoch": 1.8469435736677116, "grad_norm": 2.73348331451416, "learning_rate": 5.616079552732464e-06, "loss": 0.7997, "step": 23567 }, { "epoch": 1.8470219435736677, "grad_norm": 2.5101478099823, "learning_rate": 5.615773539256223e-06, "loss": 1.4383, "step": 23568 }, { "epoch": 1.8471003134796238, "grad_norm": 9.320666313171387, "learning_rate": 5.615467523437894e-06, "loss": 1.4316, "step": 23569 }, { "epoch": 1.84717868338558, "grad_norm": 2.003713607788086, "learning_rate": 5.615161505278645e-06, "loss": 1.4815, "step": 23570 }, { "epoch": 1.847257053291536, "grad_norm": 6.995102405548096, "learning_rate": 5.6148554847796355e-06, "loss": 1.1419, "step": 23571 }, { "epoch": 1.847335423197492, "grad_norm": 3.439985752105713, "learning_rate": 5.614549461942033e-06, "loss": 1.0222, "step": 23572 }, { "epoch": 1.8474137931034482, "grad_norm": 2.987222909927368, "learning_rate": 5.614243436767001e-06, "loss": 1.2424, "step": 23573 }, { "epoch": 1.8474921630094046, "grad_norm": 2.992262363433838, "learning_rate": 5.6139374092557e-06, "loss": 1.4667, "step": 23574 }, { "epoch": 1.8475705329153604, "grad_norm": 3.590700626373291, "learning_rate": 5.613631379409298e-06, "loss": 1.1564, "step": 23575 }, { "epoch": 1.8476489028213168, "grad_norm": 4.635181903839111, "learning_rate": 5.613325347228956e-06, "loss": 1.0983, "step": 23576 }, { "epoch": 1.8477272727272727, "grad_norm": 2.5428719520568848, "learning_rate": 5.61301931271584e-06, "loss": 1.5764, "step": 23577 }, { "epoch": 1.847805642633229, "grad_norm": 2.69248366355896, "learning_rate": 5.6127132758711146e-06, "loss": 1.0974, "step": 23578 }, { "epoch": 1.8478840125391849, "grad_norm": 3.5800583362579346, "learning_rate": 5.612407236695941e-06, "loss": 1.1207, "step": 23579 }, { "epoch": 1.8479623824451412, "grad_norm": 2.9166078567504883, "learning_rate": 5.612101195191486e-06, "loss": 1.1964, "step": 23580 }, { "epoch": 1.848040752351097, "grad_norm": 4.285132884979248, "learning_rate": 5.611795151358912e-06, "loss": 1.5741, "step": 23581 }, { "epoch": 1.8481191222570534, "grad_norm": 4.361025810241699, "learning_rate": 5.611489105199385e-06, "loss": 1.576, "step": 23582 }, { "epoch": 1.8481974921630093, "grad_norm": 5.845180988311768, "learning_rate": 5.611183056714063e-06, "loss": 1.2473, "step": 23583 }, { "epoch": 1.8482758620689657, "grad_norm": 6.619462013244629, "learning_rate": 5.61087700590412e-06, "loss": 1.4151, "step": 23584 }, { "epoch": 1.8483542319749215, "grad_norm": 2.0808868408203125, "learning_rate": 5.610570952770712e-06, "loss": 1.1865, "step": 23585 }, { "epoch": 1.8484326018808779, "grad_norm": 4.65831184387207, "learning_rate": 5.610264897315005e-06, "loss": 0.6931, "step": 23586 }, { "epoch": 1.8485109717868338, "grad_norm": 1.985766887664795, "learning_rate": 5.609958839538166e-06, "loss": 1.1177, "step": 23587 }, { "epoch": 1.84858934169279, "grad_norm": 2.295764923095703, "learning_rate": 5.609652779441355e-06, "loss": 1.354, "step": 23588 }, { "epoch": 1.848667711598746, "grad_norm": 2.241725444793701, "learning_rate": 5.609346717025738e-06, "loss": 1.4156, "step": 23589 }, { "epoch": 1.8487460815047023, "grad_norm": 2.36442494392395, "learning_rate": 5.609040652292479e-06, "loss": 1.2587, "step": 23590 }, { "epoch": 1.8488244514106582, "grad_norm": 3.5928592681884766, "learning_rate": 5.608734585242744e-06, "loss": 1.3338, "step": 23591 }, { "epoch": 1.8489028213166145, "grad_norm": 11.42246150970459, "learning_rate": 5.608428515877693e-06, "loss": 2.0075, "step": 23592 }, { "epoch": 1.8489811912225704, "grad_norm": 2.266230583190918, "learning_rate": 5.608122444198495e-06, "loss": 1.3932, "step": 23593 }, { "epoch": 1.8490595611285268, "grad_norm": 6.084579944610596, "learning_rate": 5.607816370206309e-06, "loss": 1.7193, "step": 23594 }, { "epoch": 1.8491379310344827, "grad_norm": 2.1109046936035156, "learning_rate": 5.607510293902303e-06, "loss": 1.3244, "step": 23595 }, { "epoch": 1.849216300940439, "grad_norm": 4.088217735290527, "learning_rate": 5.60720421528764e-06, "loss": 1.2381, "step": 23596 }, { "epoch": 1.8492946708463949, "grad_norm": 3.306211233139038, "learning_rate": 5.606898134363483e-06, "loss": 1.5634, "step": 23597 }, { "epoch": 1.8493730407523512, "grad_norm": 3.8505377769470215, "learning_rate": 5.606592051130998e-06, "loss": 2.1186, "step": 23598 }, { "epoch": 1.849451410658307, "grad_norm": 3.2995803356170654, "learning_rate": 5.606285965591348e-06, "loss": 1.3191, "step": 23599 }, { "epoch": 1.8495297805642634, "grad_norm": 2.6533799171447754, "learning_rate": 5.605979877745699e-06, "loss": 1.542, "step": 23600 }, { "epoch": 1.8496081504702193, "grad_norm": 4.592413902282715, "learning_rate": 5.605673787595213e-06, "loss": 1.023, "step": 23601 }, { "epoch": 1.8496865203761756, "grad_norm": 2.5159823894500732, "learning_rate": 5.605367695141054e-06, "loss": 1.3166, "step": 23602 }, { "epoch": 1.8497648902821315, "grad_norm": 3.2043142318725586, "learning_rate": 5.605061600384388e-06, "loss": 1.2245, "step": 23603 }, { "epoch": 1.8498432601880879, "grad_norm": 3.32772159576416, "learning_rate": 5.604755503326379e-06, "loss": 1.0549, "step": 23604 }, { "epoch": 1.8499216300940438, "grad_norm": 2.8059706687927246, "learning_rate": 5.60444940396819e-06, "loss": 0.7711, "step": 23605 }, { "epoch": 1.85, "grad_norm": 3.772517681121826, "learning_rate": 5.604143302310986e-06, "loss": 1.2168, "step": 23606 }, { "epoch": 1.850078369905956, "grad_norm": 2.7391653060913086, "learning_rate": 5.603837198355932e-06, "loss": 1.1497, "step": 23607 }, { "epoch": 1.8501567398119123, "grad_norm": 3.3862292766571045, "learning_rate": 5.6035310921041905e-06, "loss": 1.2468, "step": 23608 }, { "epoch": 1.8502351097178682, "grad_norm": 2.8257462978363037, "learning_rate": 5.6032249835569255e-06, "loss": 1.7724, "step": 23609 }, { "epoch": 1.8503134796238245, "grad_norm": 4.139300346374512, "learning_rate": 5.602918872715305e-06, "loss": 1.3532, "step": 23610 }, { "epoch": 1.8503918495297804, "grad_norm": 2.528698682785034, "learning_rate": 5.602612759580489e-06, "loss": 1.2154, "step": 23611 }, { "epoch": 1.8504702194357368, "grad_norm": 3.4218850135803223, "learning_rate": 5.602306644153645e-06, "loss": 1.1109, "step": 23612 }, { "epoch": 1.8505485893416926, "grad_norm": 3.160829544067383, "learning_rate": 5.602000526435934e-06, "loss": 1.1323, "step": 23613 }, { "epoch": 1.850626959247649, "grad_norm": 3.773005723953247, "learning_rate": 5.601694406428524e-06, "loss": 1.1759, "step": 23614 }, { "epoch": 1.8507053291536049, "grad_norm": 2.699235200881958, "learning_rate": 5.601388284132577e-06, "loss": 1.4928, "step": 23615 }, { "epoch": 1.8507836990595612, "grad_norm": 6.824491024017334, "learning_rate": 5.601082159549258e-06, "loss": 1.4852, "step": 23616 }, { "epoch": 1.850862068965517, "grad_norm": 2.6439998149871826, "learning_rate": 5.60077603267973e-06, "loss": 1.5287, "step": 23617 }, { "epoch": 1.8509404388714734, "grad_norm": 4.971560001373291, "learning_rate": 5.600469903525159e-06, "loss": 1.2684, "step": 23618 }, { "epoch": 1.8510188087774293, "grad_norm": 4.6622114181518555, "learning_rate": 5.60016377208671e-06, "loss": 1.5531, "step": 23619 }, { "epoch": 1.8510971786833856, "grad_norm": 7.250678539276123, "learning_rate": 5.599857638365547e-06, "loss": 1.3148, "step": 23620 }, { "epoch": 1.8511755485893417, "grad_norm": 3.254436492919922, "learning_rate": 5.599551502362832e-06, "loss": 0.6346, "step": 23621 }, { "epoch": 1.8512539184952979, "grad_norm": 2.4218883514404297, "learning_rate": 5.5992453640797316e-06, "loss": 1.3211, "step": 23622 }, { "epoch": 1.851332288401254, "grad_norm": 2.5967369079589844, "learning_rate": 5.598939223517409e-06, "loss": 1.3105, "step": 23623 }, { "epoch": 1.85141065830721, "grad_norm": 9.218771934509277, "learning_rate": 5.59863308067703e-06, "loss": 1.7591, "step": 23624 }, { "epoch": 1.8514890282131662, "grad_norm": 3.2732362747192383, "learning_rate": 5.598326935559757e-06, "loss": 1.2569, "step": 23625 }, { "epoch": 1.8515673981191223, "grad_norm": 11.557043075561523, "learning_rate": 5.598020788166757e-06, "loss": 2.2785, "step": 23626 }, { "epoch": 1.8516457680250784, "grad_norm": 4.688180446624756, "learning_rate": 5.5977146384991935e-06, "loss": 1.5395, "step": 23627 }, { "epoch": 1.8517241379310345, "grad_norm": 3.6696813106536865, "learning_rate": 5.597408486558228e-06, "loss": 1.0644, "step": 23628 }, { "epoch": 1.8518025078369906, "grad_norm": 3.6139800548553467, "learning_rate": 5.59710233234503e-06, "loss": 1.5476, "step": 23629 }, { "epoch": 1.8518808777429467, "grad_norm": 2.374372959136963, "learning_rate": 5.596796175860759e-06, "loss": 1.1342, "step": 23630 }, { "epoch": 1.8519592476489029, "grad_norm": 3.5442557334899902, "learning_rate": 5.596490017106585e-06, "loss": 1.5357, "step": 23631 }, { "epoch": 1.852037617554859, "grad_norm": 2.9481217861175537, "learning_rate": 5.596183856083668e-06, "loss": 1.3019, "step": 23632 }, { "epoch": 1.852115987460815, "grad_norm": 2.9680392742156982, "learning_rate": 5.5958776927931736e-06, "loss": 1.11, "step": 23633 }, { "epoch": 1.8521943573667712, "grad_norm": 3.7343459129333496, "learning_rate": 5.5955715272362654e-06, "loss": 1.4165, "step": 23634 }, { "epoch": 1.8522727272727273, "grad_norm": 2.66162371635437, "learning_rate": 5.59526535941411e-06, "loss": 1.2851, "step": 23635 }, { "epoch": 1.8523510971786834, "grad_norm": 2.422522783279419, "learning_rate": 5.594959189327872e-06, "loss": 1.4594, "step": 23636 }, { "epoch": 1.8524294670846395, "grad_norm": 4.464406967163086, "learning_rate": 5.5946530169787125e-06, "loss": 1.4078, "step": 23637 }, { "epoch": 1.8525078369905956, "grad_norm": 2.0514743328094482, "learning_rate": 5.5943468423678014e-06, "loss": 1.1345, "step": 23638 }, { "epoch": 1.8525862068965517, "grad_norm": 3.0464670658111572, "learning_rate": 5.5940406654962975e-06, "loss": 1.174, "step": 23639 }, { "epoch": 1.8526645768025078, "grad_norm": 4.642536640167236, "learning_rate": 5.59373448636537e-06, "loss": 1.4819, "step": 23640 }, { "epoch": 1.852742946708464, "grad_norm": 2.6813104152679443, "learning_rate": 5.593428304976181e-06, "loss": 1.2511, "step": 23641 }, { "epoch": 1.85282131661442, "grad_norm": 4.14697790145874, "learning_rate": 5.593122121329894e-06, "loss": 1.5446, "step": 23642 }, { "epoch": 1.8528996865203762, "grad_norm": 5.37375020980835, "learning_rate": 5.592815935427675e-06, "loss": 1.0397, "step": 23643 }, { "epoch": 1.8529780564263323, "grad_norm": 2.754668951034546, "learning_rate": 5.5925097472706915e-06, "loss": 1.0012, "step": 23644 }, { "epoch": 1.8530564263322884, "grad_norm": 4.4431352615356445, "learning_rate": 5.592203556860103e-06, "loss": 1.4232, "step": 23645 }, { "epoch": 1.8531347962382445, "grad_norm": 7.680891036987305, "learning_rate": 5.5918973641970766e-06, "loss": 1.0711, "step": 23646 }, { "epoch": 1.8532131661442006, "grad_norm": 4.053092002868652, "learning_rate": 5.591591169282777e-06, "loss": 1.5534, "step": 23647 }, { "epoch": 1.8532915360501567, "grad_norm": 4.063625812530518, "learning_rate": 5.591284972118369e-06, "loss": 1.5097, "step": 23648 }, { "epoch": 1.8533699059561128, "grad_norm": 3.056241035461426, "learning_rate": 5.590978772705016e-06, "loss": 1.0804, "step": 23649 }, { "epoch": 1.853448275862069, "grad_norm": 3.3925259113311768, "learning_rate": 5.590672571043883e-06, "loss": 1.1968, "step": 23650 }, { "epoch": 1.853526645768025, "grad_norm": 2.9226861000061035, "learning_rate": 5.5903663671361355e-06, "loss": 1.4895, "step": 23651 }, { "epoch": 1.8536050156739812, "grad_norm": 2.983043670654297, "learning_rate": 5.590060160982937e-06, "loss": 1.4281, "step": 23652 }, { "epoch": 1.8536833855799373, "grad_norm": 4.556900978088379, "learning_rate": 5.589753952585452e-06, "loss": 1.1516, "step": 23653 }, { "epoch": 1.8537617554858934, "grad_norm": 3.25227427482605, "learning_rate": 5.589447741944848e-06, "loss": 0.8064, "step": 23654 }, { "epoch": 1.8538401253918495, "grad_norm": 3.668348550796509, "learning_rate": 5.589141529062287e-06, "loss": 1.1444, "step": 23655 }, { "epoch": 1.8539184952978056, "grad_norm": 3.390336036682129, "learning_rate": 5.588835313938933e-06, "loss": 1.4119, "step": 23656 }, { "epoch": 1.8539968652037617, "grad_norm": 2.212186574935913, "learning_rate": 5.588529096575954e-06, "loss": 1.1325, "step": 23657 }, { "epoch": 1.8540752351097178, "grad_norm": 4.6008758544921875, "learning_rate": 5.588222876974509e-06, "loss": 1.4964, "step": 23658 }, { "epoch": 1.854153605015674, "grad_norm": 4.260002136230469, "learning_rate": 5.58791665513577e-06, "loss": 1.7486, "step": 23659 }, { "epoch": 1.85423197492163, "grad_norm": 3.958948850631714, "learning_rate": 5.587610431060897e-06, "loss": 1.4627, "step": 23660 }, { "epoch": 1.8543103448275862, "grad_norm": 6.805309295654297, "learning_rate": 5.587304204751055e-06, "loss": 1.1365, "step": 23661 }, { "epoch": 1.8543887147335423, "grad_norm": 2.8228700160980225, "learning_rate": 5.586997976207411e-06, "loss": 1.6872, "step": 23662 }, { "epoch": 1.8544670846394984, "grad_norm": 1.8640938997268677, "learning_rate": 5.586691745431126e-06, "loss": 1.2857, "step": 23663 }, { "epoch": 1.8545454545454545, "grad_norm": 3.3892509937286377, "learning_rate": 5.586385512423369e-06, "loss": 1.3957, "step": 23664 }, { "epoch": 1.8546238244514106, "grad_norm": 2.8062756061553955, "learning_rate": 5.5860792771853e-06, "loss": 1.7589, "step": 23665 }, { "epoch": 1.8547021943573667, "grad_norm": 2.8394100666046143, "learning_rate": 5.585773039718088e-06, "loss": 1.4285, "step": 23666 }, { "epoch": 1.854780564263323, "grad_norm": 4.482953071594238, "learning_rate": 5.585466800022897e-06, "loss": 1.6666, "step": 23667 }, { "epoch": 1.854858934169279, "grad_norm": 5.352048397064209, "learning_rate": 5.585160558100891e-06, "loss": 1.9644, "step": 23668 }, { "epoch": 1.8549373040752353, "grad_norm": 3.3283934593200684, "learning_rate": 5.584854313953233e-06, "loss": 1.3219, "step": 23669 }, { "epoch": 1.8550156739811912, "grad_norm": 2.4024605751037598, "learning_rate": 5.5845480675810905e-06, "loss": 1.2862, "step": 23670 }, { "epoch": 1.8550940438871475, "grad_norm": 2.549563407897949, "learning_rate": 5.5842418189856276e-06, "loss": 1.4047, "step": 23671 }, { "epoch": 1.8551724137931034, "grad_norm": 3.5729761123657227, "learning_rate": 5.5839355681680084e-06, "loss": 1.4798, "step": 23672 }, { "epoch": 1.8552507836990597, "grad_norm": 4.069349765777588, "learning_rate": 5.5836293151293995e-06, "loss": 1.4383, "step": 23673 }, { "epoch": 1.8553291536050156, "grad_norm": 2.6196882724761963, "learning_rate": 5.5833230598709645e-06, "loss": 1.3801, "step": 23674 }, { "epoch": 1.855407523510972, "grad_norm": 3.111680746078491, "learning_rate": 5.583016802393867e-06, "loss": 1.2886, "step": 23675 }, { "epoch": 1.8554858934169278, "grad_norm": 35.335182189941406, "learning_rate": 5.582710542699273e-06, "loss": 3.0075, "step": 23676 }, { "epoch": 1.8555642633228842, "grad_norm": 3.3591291904449463, "learning_rate": 5.582404280788348e-06, "loss": 1.3362, "step": 23677 }, { "epoch": 1.85564263322884, "grad_norm": 3.225691080093384, "learning_rate": 5.582098016662256e-06, "loss": 0.9008, "step": 23678 }, { "epoch": 1.8557210031347964, "grad_norm": 2.9072933197021484, "learning_rate": 5.581791750322162e-06, "loss": 0.8401, "step": 23679 }, { "epoch": 1.8557993730407523, "grad_norm": 2.9686009883880615, "learning_rate": 5.581485481769231e-06, "loss": 1.2691, "step": 23680 }, { "epoch": 1.8558777429467086, "grad_norm": 2.464902639389038, "learning_rate": 5.581179211004629e-06, "loss": 1.5442, "step": 23681 }, { "epoch": 1.8559561128526645, "grad_norm": 5.361387252807617, "learning_rate": 5.5808729380295175e-06, "loss": 1.1171, "step": 23682 }, { "epoch": 1.8560344827586208, "grad_norm": 5.101444721221924, "learning_rate": 5.580566662845067e-06, "loss": 1.466, "step": 23683 }, { "epoch": 1.8561128526645767, "grad_norm": 2.234873056411743, "learning_rate": 5.580260385452436e-06, "loss": 1.3645, "step": 23684 }, { "epoch": 1.856191222570533, "grad_norm": 4.749307155609131, "learning_rate": 5.579954105852794e-06, "loss": 1.2353, "step": 23685 }, { "epoch": 1.856269592476489, "grad_norm": 5.679443836212158, "learning_rate": 5.579647824047305e-06, "loss": 1.4597, "step": 23686 }, { "epoch": 1.8563479623824453, "grad_norm": 3.7717440128326416, "learning_rate": 5.579341540037132e-06, "loss": 1.288, "step": 23687 }, { "epoch": 1.8564263322884011, "grad_norm": 5.681687831878662, "learning_rate": 5.5790352538234415e-06, "loss": 1.5237, "step": 23688 }, { "epoch": 1.8565047021943575, "grad_norm": 7.063530921936035, "learning_rate": 5.5787289654073995e-06, "loss": 1.1933, "step": 23689 }, { "epoch": 1.8565830721003134, "grad_norm": 5.291553497314453, "learning_rate": 5.57842267479017e-06, "loss": 1.0239, "step": 23690 }, { "epoch": 1.8566614420062697, "grad_norm": 3.0446739196777344, "learning_rate": 5.578116381972916e-06, "loss": 1.3579, "step": 23691 }, { "epoch": 1.8567398119122256, "grad_norm": 3.686183214187622, "learning_rate": 5.577810086956806e-06, "loss": 1.31, "step": 23692 }, { "epoch": 1.856818181818182, "grad_norm": 3.049363374710083, "learning_rate": 5.577503789743001e-06, "loss": 1.3082, "step": 23693 }, { "epoch": 1.8568965517241378, "grad_norm": 3.549724578857422, "learning_rate": 5.577197490332669e-06, "loss": 1.75, "step": 23694 }, { "epoch": 1.8569749216300941, "grad_norm": 2.9305787086486816, "learning_rate": 5.576891188726975e-06, "loss": 1.7897, "step": 23695 }, { "epoch": 1.85705329153605, "grad_norm": 5.014636516571045, "learning_rate": 5.576584884927083e-06, "loss": 1.3381, "step": 23696 }, { "epoch": 1.8571316614420064, "grad_norm": 3.653883218765259, "learning_rate": 5.576278578934158e-06, "loss": 1.0171, "step": 23697 }, { "epoch": 1.8572100313479623, "grad_norm": 2.3145511150360107, "learning_rate": 5.575972270749366e-06, "loss": 1.1555, "step": 23698 }, { "epoch": 1.8572884012539186, "grad_norm": 2.5315558910369873, "learning_rate": 5.575665960373869e-06, "loss": 1.6258, "step": 23699 }, { "epoch": 1.8573667711598745, "grad_norm": 4.332711219787598, "learning_rate": 5.575359647808837e-06, "loss": 1.2367, "step": 23700 }, { "epoch": 1.8574451410658308, "grad_norm": 10.015076637268066, "learning_rate": 5.575053333055433e-06, "loss": 1.345, "step": 23701 }, { "epoch": 1.8575235109717867, "grad_norm": 3.074211835861206, "learning_rate": 5.574747016114818e-06, "loss": 1.1115, "step": 23702 }, { "epoch": 1.857601880877743, "grad_norm": 2.46136474609375, "learning_rate": 5.574440696988164e-06, "loss": 1.2677, "step": 23703 }, { "epoch": 1.857680250783699, "grad_norm": 2.6595096588134766, "learning_rate": 5.574134375676631e-06, "loss": 1.3631, "step": 23704 }, { "epoch": 1.8577586206896552, "grad_norm": 2.7910611629486084, "learning_rate": 5.573828052181386e-06, "loss": 1.4169, "step": 23705 }, { "epoch": 1.8578369905956111, "grad_norm": 3.4912281036376953, "learning_rate": 5.573521726503594e-06, "loss": 1.9792, "step": 23706 }, { "epoch": 1.8579153605015675, "grad_norm": 3.1265792846679688, "learning_rate": 5.573215398644418e-06, "loss": 1.549, "step": 23707 }, { "epoch": 1.8579937304075234, "grad_norm": 2.4099764823913574, "learning_rate": 5.572909068605027e-06, "loss": 1.5634, "step": 23708 }, { "epoch": 1.8580721003134797, "grad_norm": 2.096141815185547, "learning_rate": 5.572602736386585e-06, "loss": 1.1957, "step": 23709 }, { "epoch": 1.8581504702194356, "grad_norm": 2.76423978805542, "learning_rate": 5.5722964019902535e-06, "loss": 1.4137, "step": 23710 }, { "epoch": 1.858228840125392, "grad_norm": 3.172985076904297, "learning_rate": 5.5719900654172036e-06, "loss": 1.6798, "step": 23711 }, { "epoch": 1.8583072100313478, "grad_norm": 5.105000972747803, "learning_rate": 5.571683726668595e-06, "loss": 1.7616, "step": 23712 }, { "epoch": 1.8583855799373041, "grad_norm": 2.6344354152679443, "learning_rate": 5.571377385745597e-06, "loss": 1.4598, "step": 23713 }, { "epoch": 1.8584639498432602, "grad_norm": 3.312527894973755, "learning_rate": 5.5710710426493716e-06, "loss": 1.2554, "step": 23714 }, { "epoch": 1.8585423197492164, "grad_norm": 2.6574177742004395, "learning_rate": 5.5707646973810855e-06, "loss": 1.3215, "step": 23715 }, { "epoch": 1.8586206896551725, "grad_norm": 2.586575508117676, "learning_rate": 5.570458349941904e-06, "loss": 1.5646, "step": 23716 }, { "epoch": 1.8586990595611286, "grad_norm": 3.9796829223632812, "learning_rate": 5.5701520003329915e-06, "loss": 1.5506, "step": 23717 }, { "epoch": 1.8587774294670847, "grad_norm": 2.3236167430877686, "learning_rate": 5.569845648555514e-06, "loss": 1.1686, "step": 23718 }, { "epoch": 1.8588557993730408, "grad_norm": 4.182004451751709, "learning_rate": 5.569539294610636e-06, "loss": 0.9805, "step": 23719 }, { "epoch": 1.858934169278997, "grad_norm": 1.884503722190857, "learning_rate": 5.569232938499524e-06, "loss": 1.3624, "step": 23720 }, { "epoch": 1.859012539184953, "grad_norm": 3.2789053916931152, "learning_rate": 5.56892658022334e-06, "loss": 1.1708, "step": 23721 }, { "epoch": 1.8590909090909091, "grad_norm": 3.6883163452148438, "learning_rate": 5.568620219783253e-06, "loss": 1.8628, "step": 23722 }, { "epoch": 1.8591692789968652, "grad_norm": 2.507456064224243, "learning_rate": 5.568313857180428e-06, "loss": 1.2367, "step": 23723 }, { "epoch": 1.8592476489028213, "grad_norm": 2.887052059173584, "learning_rate": 5.5680074924160265e-06, "loss": 1.1357, "step": 23724 }, { "epoch": 1.8593260188087775, "grad_norm": 3.023247718811035, "learning_rate": 5.567701125491217e-06, "loss": 1.3926, "step": 23725 }, { "epoch": 1.8594043887147336, "grad_norm": 4.827486991882324, "learning_rate": 5.567394756407165e-06, "loss": 1.4133, "step": 23726 }, { "epoch": 1.8594827586206897, "grad_norm": 4.401556491851807, "learning_rate": 5.5670883851650335e-06, "loss": 1.5859, "step": 23727 }, { "epoch": 1.8595611285266458, "grad_norm": 3.591085433959961, "learning_rate": 5.566782011765991e-06, "loss": 1.3973, "step": 23728 }, { "epoch": 1.859639498432602, "grad_norm": 2.272071361541748, "learning_rate": 5.566475636211199e-06, "loss": 1.3776, "step": 23729 }, { "epoch": 1.859717868338558, "grad_norm": 2.054161787033081, "learning_rate": 5.566169258501826e-06, "loss": 1.2405, "step": 23730 }, { "epoch": 1.8597962382445141, "grad_norm": 6.656524181365967, "learning_rate": 5.5658628786390344e-06, "loss": 1.8358, "step": 23731 }, { "epoch": 1.8598746081504702, "grad_norm": 3.3086118698120117, "learning_rate": 5.565556496623992e-06, "loss": 1.6793, "step": 23732 }, { "epoch": 1.8599529780564263, "grad_norm": 1.9664512872695923, "learning_rate": 5.565250112457863e-06, "loss": 1.1735, "step": 23733 }, { "epoch": 1.8600313479623825, "grad_norm": 7.09398078918457, "learning_rate": 5.564943726141812e-06, "loss": 1.2422, "step": 23734 }, { "epoch": 1.8601097178683386, "grad_norm": 5.677492141723633, "learning_rate": 5.564637337677007e-06, "loss": 1.0981, "step": 23735 }, { "epoch": 1.8601880877742947, "grad_norm": 2.8665900230407715, "learning_rate": 5.564330947064611e-06, "loss": 1.6811, "step": 23736 }, { "epoch": 1.8602664576802508, "grad_norm": 2.3977720737457275, "learning_rate": 5.56402455430579e-06, "loss": 1.349, "step": 23737 }, { "epoch": 1.860344827586207, "grad_norm": 4.217711448669434, "learning_rate": 5.563718159401707e-06, "loss": 1.2515, "step": 23738 }, { "epoch": 1.860423197492163, "grad_norm": 3.0221519470214844, "learning_rate": 5.563411762353533e-06, "loss": 1.522, "step": 23739 }, { "epoch": 1.8605015673981191, "grad_norm": 2.977646589279175, "learning_rate": 5.56310536316243e-06, "loss": 1.169, "step": 23740 }, { "epoch": 1.8605799373040752, "grad_norm": 2.6737923622131348, "learning_rate": 5.562798961829562e-06, "loss": 1.3098, "step": 23741 }, { "epoch": 1.8606583072100313, "grad_norm": 3.134049654006958, "learning_rate": 5.562492558356095e-06, "loss": 1.1798, "step": 23742 }, { "epoch": 1.8607366771159874, "grad_norm": 2.577604055404663, "learning_rate": 5.5621861527431965e-06, "loss": 1.449, "step": 23743 }, { "epoch": 1.8608150470219436, "grad_norm": 4.314911365509033, "learning_rate": 5.561879744992032e-06, "loss": 1.587, "step": 23744 }, { "epoch": 1.8608934169278997, "grad_norm": 3.0727663040161133, "learning_rate": 5.561573335103761e-06, "loss": 1.5211, "step": 23745 }, { "epoch": 1.8609717868338558, "grad_norm": 4.351261138916016, "learning_rate": 5.5612669230795576e-06, "loss": 1.1726, "step": 23746 }, { "epoch": 1.861050156739812, "grad_norm": 2.362974166870117, "learning_rate": 5.560960508920581e-06, "loss": 1.2273, "step": 23747 }, { "epoch": 1.861128526645768, "grad_norm": 5.955233573913574, "learning_rate": 5.560654092628e-06, "loss": 1.2881, "step": 23748 }, { "epoch": 1.861206896551724, "grad_norm": 2.1152570247650146, "learning_rate": 5.5603476742029795e-06, "loss": 1.3024, "step": 23749 }, { "epoch": 1.8612852664576802, "grad_norm": 3.0885202884674072, "learning_rate": 5.560041253646683e-06, "loss": 1.4197, "step": 23750 }, { "epoch": 1.8613636363636363, "grad_norm": 5.0047688484191895, "learning_rate": 5.559734830960277e-06, "loss": 1.2905, "step": 23751 }, { "epoch": 1.8614420062695924, "grad_norm": 10.88546085357666, "learning_rate": 5.559428406144928e-06, "loss": 1.7108, "step": 23752 }, { "epoch": 1.8615203761755486, "grad_norm": 3.5974183082580566, "learning_rate": 5.559121979201799e-06, "loss": 1.5227, "step": 23753 }, { "epoch": 1.8615987460815047, "grad_norm": 7.205098628997803, "learning_rate": 5.558815550132058e-06, "loss": 1.5307, "step": 23754 }, { "epoch": 1.8616771159874608, "grad_norm": 3.3306145668029785, "learning_rate": 5.5585091189368714e-06, "loss": 1.2527, "step": 23755 }, { "epoch": 1.8617554858934169, "grad_norm": 2.409137487411499, "learning_rate": 5.558202685617402e-06, "loss": 1.2406, "step": 23756 }, { "epoch": 1.861833855799373, "grad_norm": 8.44356632232666, "learning_rate": 5.557896250174816e-06, "loss": 2.0034, "step": 23757 }, { "epoch": 1.861912225705329, "grad_norm": 2.283271074295044, "learning_rate": 5.557589812610279e-06, "loss": 1.413, "step": 23758 }, { "epoch": 1.8619905956112852, "grad_norm": 2.3975489139556885, "learning_rate": 5.557283372924956e-06, "loss": 1.0964, "step": 23759 }, { "epoch": 1.8620689655172413, "grad_norm": 2.8809165954589844, "learning_rate": 5.556976931120015e-06, "loss": 1.5466, "step": 23760 }, { "epoch": 1.8621473354231974, "grad_norm": 3.3475499153137207, "learning_rate": 5.556670487196619e-06, "loss": 1.2882, "step": 23761 }, { "epoch": 1.8622257053291538, "grad_norm": 4.544661045074463, "learning_rate": 5.556364041155935e-06, "loss": 1.4306, "step": 23762 }, { "epoch": 1.8623040752351097, "grad_norm": 2.0260918140411377, "learning_rate": 5.556057592999128e-06, "loss": 1.1087, "step": 23763 }, { "epoch": 1.862382445141066, "grad_norm": 4.004471302032471, "learning_rate": 5.555751142727361e-06, "loss": 1.3392, "step": 23764 }, { "epoch": 1.8624608150470219, "grad_norm": 3.4388601779937744, "learning_rate": 5.555444690341806e-06, "loss": 1.7397, "step": 23765 }, { "epoch": 1.8625391849529782, "grad_norm": 1.792779564857483, "learning_rate": 5.555138235843621e-06, "loss": 1.1775, "step": 23766 }, { "epoch": 1.862617554858934, "grad_norm": 4.507354736328125, "learning_rate": 5.5548317792339775e-06, "loss": 1.1607, "step": 23767 }, { "epoch": 1.8626959247648904, "grad_norm": 2.5497241020202637, "learning_rate": 5.554525320514039e-06, "loss": 1.3331, "step": 23768 }, { "epoch": 1.8627742946708463, "grad_norm": 3.111996650695801, "learning_rate": 5.55421885968497e-06, "loss": 1.565, "step": 23769 }, { "epoch": 1.8628526645768027, "grad_norm": 4.620918273925781, "learning_rate": 5.553912396747938e-06, "loss": 1.5694, "step": 23770 }, { "epoch": 1.8629310344827585, "grad_norm": 4.51743221282959, "learning_rate": 5.553605931704107e-06, "loss": 1.2483, "step": 23771 }, { "epoch": 1.8630094043887149, "grad_norm": 4.537753105163574, "learning_rate": 5.553299464554644e-06, "loss": 0.9313, "step": 23772 }, { "epoch": 1.8630877742946708, "grad_norm": 2.126206874847412, "learning_rate": 5.552992995300713e-06, "loss": 1.054, "step": 23773 }, { "epoch": 1.863166144200627, "grad_norm": 2.0246524810791016, "learning_rate": 5.552686523943483e-06, "loss": 1.1006, "step": 23774 }, { "epoch": 1.863244514106583, "grad_norm": 3.1650781631469727, "learning_rate": 5.5523800504841156e-06, "loss": 1.2021, "step": 23775 }, { "epoch": 1.8633228840125393, "grad_norm": 3.3406600952148438, "learning_rate": 5.552073574923777e-06, "loss": 1.5497, "step": 23776 }, { "epoch": 1.8634012539184952, "grad_norm": 2.733527421951294, "learning_rate": 5.551767097263636e-06, "loss": 1.406, "step": 23777 }, { "epoch": 1.8634796238244515, "grad_norm": 2.6378684043884277, "learning_rate": 5.551460617504856e-06, "loss": 1.1407, "step": 23778 }, { "epoch": 1.8635579937304074, "grad_norm": 2.234753370285034, "learning_rate": 5.551154135648602e-06, "loss": 1.3528, "step": 23779 }, { "epoch": 1.8636363636363638, "grad_norm": 2.521501302719116, "learning_rate": 5.550847651696042e-06, "loss": 1.5675, "step": 23780 }, { "epoch": 1.8637147335423196, "grad_norm": 2.656970739364624, "learning_rate": 5.55054116564834e-06, "loss": 1.2311, "step": 23781 }, { "epoch": 1.863793103448276, "grad_norm": 6.484748840332031, "learning_rate": 5.550234677506663e-06, "loss": 1.5417, "step": 23782 }, { "epoch": 1.8638714733542319, "grad_norm": 3.7991838455200195, "learning_rate": 5.5499281872721744e-06, "loss": 1.4966, "step": 23783 }, { "epoch": 1.8639498432601882, "grad_norm": 6.924911022186279, "learning_rate": 5.549621694946044e-06, "loss": 1.3129, "step": 23784 }, { "epoch": 1.864028213166144, "grad_norm": 17.44729232788086, "learning_rate": 5.5493152005294325e-06, "loss": 0.95, "step": 23785 }, { "epoch": 1.8641065830721004, "grad_norm": 2.548997402191162, "learning_rate": 5.5490087040235085e-06, "loss": 1.4331, "step": 23786 }, { "epoch": 1.8641849529780563, "grad_norm": 2.4831063747406006, "learning_rate": 5.548702205429437e-06, "loss": 1.2863, "step": 23787 }, { "epoch": 1.8642633228840126, "grad_norm": 4.436575412750244, "learning_rate": 5.548395704748385e-06, "loss": 1.326, "step": 23788 }, { "epoch": 1.8643416927899685, "grad_norm": 3.001549243927002, "learning_rate": 5.548089201981517e-06, "loss": 1.4473, "step": 23789 }, { "epoch": 1.8644200626959249, "grad_norm": 2.128237247467041, "learning_rate": 5.54778269713e-06, "loss": 1.1334, "step": 23790 }, { "epoch": 1.8644984326018808, "grad_norm": 4.66090202331543, "learning_rate": 5.547476190194999e-06, "loss": 1.5995, "step": 23791 }, { "epoch": 1.864576802507837, "grad_norm": 4.536486625671387, "learning_rate": 5.5471696811776775e-06, "loss": 1.6014, "step": 23792 }, { "epoch": 1.864655172413793, "grad_norm": 5.325764179229736, "learning_rate": 5.546863170079206e-06, "loss": 1.3806, "step": 23793 }, { "epoch": 1.8647335423197493, "grad_norm": 3.8044445514678955, "learning_rate": 5.5465566569007466e-06, "loss": 1.6652, "step": 23794 }, { "epoch": 1.8648119122257052, "grad_norm": 4.5234599113464355, "learning_rate": 5.546250141643468e-06, "loss": 1.5425, "step": 23795 }, { "epoch": 1.8648902821316615, "grad_norm": 2.3483808040618896, "learning_rate": 5.545943624308533e-06, "loss": 1.1619, "step": 23796 }, { "epoch": 1.8649686520376174, "grad_norm": 2.821368932723999, "learning_rate": 5.545637104897109e-06, "loss": 1.2744, "step": 23797 }, { "epoch": 1.8650470219435737, "grad_norm": 5.057024955749512, "learning_rate": 5.545330583410362e-06, "loss": 1.4, "step": 23798 }, { "epoch": 1.8651253918495296, "grad_norm": 2.9775617122650146, "learning_rate": 5.5450240598494576e-06, "loss": 1.5166, "step": 23799 }, { "epoch": 1.865203761755486, "grad_norm": 4.00328254699707, "learning_rate": 5.544717534215562e-06, "loss": 1.3697, "step": 23800 }, { "epoch": 1.8652821316614419, "grad_norm": 5.1412458419799805, "learning_rate": 5.54441100650984e-06, "loss": 1.2255, "step": 23801 }, { "epoch": 1.8653605015673982, "grad_norm": 3.211091995239258, "learning_rate": 5.544104476733458e-06, "loss": 1.4007, "step": 23802 }, { "epoch": 1.865438871473354, "grad_norm": 2.610048770904541, "learning_rate": 5.543797944887583e-06, "loss": 1.3608, "step": 23803 }, { "epoch": 1.8655172413793104, "grad_norm": 2.6345510482788086, "learning_rate": 5.543491410973379e-06, "loss": 1.5389, "step": 23804 }, { "epoch": 1.8655956112852663, "grad_norm": 2.6559319496154785, "learning_rate": 5.543184874992013e-06, "loss": 1.2571, "step": 23805 }, { "epoch": 1.8656739811912226, "grad_norm": 3.3236546516418457, "learning_rate": 5.542878336944652e-06, "loss": 1.6346, "step": 23806 }, { "epoch": 1.8657523510971785, "grad_norm": 3.3132357597351074, "learning_rate": 5.542571796832459e-06, "loss": 1.3516, "step": 23807 }, { "epoch": 1.8658307210031349, "grad_norm": 2.911489963531494, "learning_rate": 5.542265254656603e-06, "loss": 0.8406, "step": 23808 }, { "epoch": 1.865909090909091, "grad_norm": 2.6225802898406982, "learning_rate": 5.541958710418246e-06, "loss": 0.605, "step": 23809 }, { "epoch": 1.865987460815047, "grad_norm": 2.8477516174316406, "learning_rate": 5.5416521641185596e-06, "loss": 1.1459, "step": 23810 }, { "epoch": 1.8660658307210032, "grad_norm": 2.329636812210083, "learning_rate": 5.541345615758704e-06, "loss": 1.5176, "step": 23811 }, { "epoch": 1.8661442006269593, "grad_norm": 4.2676520347595215, "learning_rate": 5.54103906533985e-06, "loss": 1.2247, "step": 23812 }, { "epoch": 1.8662225705329154, "grad_norm": 2.21771240234375, "learning_rate": 5.540732512863159e-06, "loss": 1.3707, "step": 23813 }, { "epoch": 1.8663009404388715, "grad_norm": 5.902869701385498, "learning_rate": 5.540425958329801e-06, "loss": 2.0348, "step": 23814 }, { "epoch": 1.8663793103448276, "grad_norm": 3.846371650695801, "learning_rate": 5.540119401740939e-06, "loss": 1.3588, "step": 23815 }, { "epoch": 1.8664576802507837, "grad_norm": 2.6051158905029297, "learning_rate": 5.539812843097741e-06, "loss": 1.568, "step": 23816 }, { "epoch": 1.8665360501567398, "grad_norm": 2.260491132736206, "learning_rate": 5.539506282401372e-06, "loss": 1.3847, "step": 23817 }, { "epoch": 1.866614420062696, "grad_norm": 3.0222723484039307, "learning_rate": 5.539199719652998e-06, "loss": 1.2771, "step": 23818 }, { "epoch": 1.866692789968652, "grad_norm": 3.0350427627563477, "learning_rate": 5.538893154853786e-06, "loss": 1.2943, "step": 23819 }, { "epoch": 1.8667711598746082, "grad_norm": 2.496654510498047, "learning_rate": 5.538586588004899e-06, "loss": 0.9108, "step": 23820 }, { "epoch": 1.8668495297805643, "grad_norm": 2.3583343029022217, "learning_rate": 5.538280019107508e-06, "loss": 0.9903, "step": 23821 }, { "epoch": 1.8669278996865204, "grad_norm": 7.509052753448486, "learning_rate": 5.537973448162776e-06, "loss": 0.7933, "step": 23822 }, { "epoch": 1.8670062695924765, "grad_norm": 2.5833728313446045, "learning_rate": 5.537666875171868e-06, "loss": 1.2994, "step": 23823 }, { "epoch": 1.8670846394984326, "grad_norm": 3.846885919570923, "learning_rate": 5.537360300135952e-06, "loss": 1.3864, "step": 23824 }, { "epoch": 1.8671630094043887, "grad_norm": 3.4085285663604736, "learning_rate": 5.537053723056193e-06, "loss": 1.2822, "step": 23825 }, { "epoch": 1.8672413793103448, "grad_norm": 2.9420816898345947, "learning_rate": 5.5367471439337576e-06, "loss": 1.061, "step": 23826 }, { "epoch": 1.867319749216301, "grad_norm": 2.731684446334839, "learning_rate": 5.536440562769811e-06, "loss": 1.0637, "step": 23827 }, { "epoch": 1.867398119122257, "grad_norm": 1.9910353422164917, "learning_rate": 5.536133979565522e-06, "loss": 0.8844, "step": 23828 }, { "epoch": 1.8674764890282132, "grad_norm": 3.5977299213409424, "learning_rate": 5.535827394322054e-06, "loss": 1.4204, "step": 23829 }, { "epoch": 1.8675548589341693, "grad_norm": 3.3164546489715576, "learning_rate": 5.535520807040574e-06, "loss": 1.2719, "step": 23830 }, { "epoch": 1.8676332288401254, "grad_norm": 2.386953830718994, "learning_rate": 5.5352142177222465e-06, "loss": 1.458, "step": 23831 }, { "epoch": 1.8677115987460815, "grad_norm": 4.102747917175293, "learning_rate": 5.53490762636824e-06, "loss": 0.9308, "step": 23832 }, { "epoch": 1.8677899686520376, "grad_norm": 5.807549476623535, "learning_rate": 5.53460103297972e-06, "loss": 1.1497, "step": 23833 }, { "epoch": 1.8678683385579937, "grad_norm": 4.94497537612915, "learning_rate": 5.534294437557851e-06, "loss": 1.4795, "step": 23834 }, { "epoch": 1.8679467084639498, "grad_norm": 2.7715625762939453, "learning_rate": 5.533987840103802e-06, "loss": 1.4916, "step": 23835 }, { "epoch": 1.868025078369906, "grad_norm": 3.62669038772583, "learning_rate": 5.533681240618735e-06, "loss": 1.5783, "step": 23836 }, { "epoch": 1.868103448275862, "grad_norm": 3.1898534297943115, "learning_rate": 5.533374639103821e-06, "loss": 1.5739, "step": 23837 }, { "epoch": 1.8681818181818182, "grad_norm": 3.3164050579071045, "learning_rate": 5.533068035560225e-06, "loss": 1.2333, "step": 23838 }, { "epoch": 1.8682601880877743, "grad_norm": 4.931968688964844, "learning_rate": 5.532761429989109e-06, "loss": 1.2046, "step": 23839 }, { "epoch": 1.8683385579937304, "grad_norm": 13.408923149108887, "learning_rate": 5.532454822391644e-06, "loss": 1.0377, "step": 23840 }, { "epoch": 1.8684169278996865, "grad_norm": 2.778925657272339, "learning_rate": 5.532148212768994e-06, "loss": 0.8349, "step": 23841 }, { "epoch": 1.8684952978056426, "grad_norm": 4.844809532165527, "learning_rate": 5.531841601122325e-06, "loss": 1.4479, "step": 23842 }, { "epoch": 1.8685736677115987, "grad_norm": 2.901451826095581, "learning_rate": 5.531534987452804e-06, "loss": 1.3485, "step": 23843 }, { "epoch": 1.8686520376175548, "grad_norm": 5.335193157196045, "learning_rate": 5.531228371761596e-06, "loss": 1.9314, "step": 23844 }, { "epoch": 1.868730407523511, "grad_norm": 2.588693141937256, "learning_rate": 5.5309217540498705e-06, "loss": 1.1094, "step": 23845 }, { "epoch": 1.868808777429467, "grad_norm": 5.417827606201172, "learning_rate": 5.5306151343187886e-06, "loss": 1.5361, "step": 23846 }, { "epoch": 1.8688871473354232, "grad_norm": 3.1015548706054688, "learning_rate": 5.530308512569522e-06, "loss": 1.4198, "step": 23847 }, { "epoch": 1.8689655172413793, "grad_norm": 2.9989054203033447, "learning_rate": 5.530001888803232e-06, "loss": 1.4454, "step": 23848 }, { "epoch": 1.8690438871473354, "grad_norm": 2.3399417400360107, "learning_rate": 5.529695263021089e-06, "loss": 1.4755, "step": 23849 }, { "epoch": 1.8691222570532915, "grad_norm": 2.5025241374969482, "learning_rate": 5.529388635224256e-06, "loss": 1.1921, "step": 23850 }, { "epoch": 1.8692006269592476, "grad_norm": 9.67910099029541, "learning_rate": 5.529082005413899e-06, "loss": 1.7445, "step": 23851 }, { "epoch": 1.8692789968652037, "grad_norm": 3.4438328742980957, "learning_rate": 5.528775373591189e-06, "loss": 1.5074, "step": 23852 }, { "epoch": 1.8693573667711598, "grad_norm": 2.7635998725891113, "learning_rate": 5.528468739757286e-06, "loss": 1.2792, "step": 23853 }, { "epoch": 1.869435736677116, "grad_norm": 3.2286674976348877, "learning_rate": 5.528162103913362e-06, "loss": 1.2722, "step": 23854 }, { "epoch": 1.8695141065830723, "grad_norm": 6.068853855133057, "learning_rate": 5.527855466060579e-06, "loss": 1.2459, "step": 23855 }, { "epoch": 1.8695924764890282, "grad_norm": 3.9087870121002197, "learning_rate": 5.527548826200106e-06, "loss": 0.7206, "step": 23856 }, { "epoch": 1.8696708463949845, "grad_norm": 3.7011094093322754, "learning_rate": 5.527242184333108e-06, "loss": 1.7897, "step": 23857 }, { "epoch": 1.8697492163009404, "grad_norm": 3.102351665496826, "learning_rate": 5.5269355404607505e-06, "loss": 1.1036, "step": 23858 }, { "epoch": 1.8698275862068967, "grad_norm": 7.031285762786865, "learning_rate": 5.526628894584202e-06, "loss": 1.1968, "step": 23859 }, { "epoch": 1.8699059561128526, "grad_norm": 3.9523348808288574, "learning_rate": 5.526322246704628e-06, "loss": 1.7775, "step": 23860 }, { "epoch": 1.869984326018809, "grad_norm": 3.1271400451660156, "learning_rate": 5.526015596823192e-06, "loss": 1.8377, "step": 23861 }, { "epoch": 1.8700626959247648, "grad_norm": 5.125463485717773, "learning_rate": 5.525708944941065e-06, "loss": 1.2426, "step": 23862 }, { "epoch": 1.8701410658307211, "grad_norm": 3.7678699493408203, "learning_rate": 5.5254022910594105e-06, "loss": 1.5994, "step": 23863 }, { "epoch": 1.870219435736677, "grad_norm": 15.986435890197754, "learning_rate": 5.5250956351793975e-06, "loss": 2.3385, "step": 23864 }, { "epoch": 1.8702978056426334, "grad_norm": 2.5662379264831543, "learning_rate": 5.524788977302187e-06, "loss": 0.9412, "step": 23865 }, { "epoch": 1.8703761755485893, "grad_norm": 2.4649484157562256, "learning_rate": 5.524482317428952e-06, "loss": 1.3561, "step": 23866 }, { "epoch": 1.8704545454545456, "grad_norm": 3.814659357070923, "learning_rate": 5.5241756555608526e-06, "loss": 1.2037, "step": 23867 }, { "epoch": 1.8705329153605015, "grad_norm": 2.999732494354248, "learning_rate": 5.523868991699061e-06, "loss": 1.3504, "step": 23868 }, { "epoch": 1.8706112852664578, "grad_norm": 2.58419132232666, "learning_rate": 5.523562325844739e-06, "loss": 1.3232, "step": 23869 }, { "epoch": 1.8706896551724137, "grad_norm": 3.0037293434143066, "learning_rate": 5.5232556579990556e-06, "loss": 1.1203, "step": 23870 }, { "epoch": 1.87076802507837, "grad_norm": 5.279471397399902, "learning_rate": 5.522948988163177e-06, "loss": 1.5021, "step": 23871 }, { "epoch": 1.870846394984326, "grad_norm": 1.819580078125, "learning_rate": 5.522642316338268e-06, "loss": 1.1703, "step": 23872 }, { "epoch": 1.8709247648902823, "grad_norm": 2.7685415744781494, "learning_rate": 5.522335642525497e-06, "loss": 1.2737, "step": 23873 }, { "epoch": 1.8710031347962381, "grad_norm": 5.2402729988098145, "learning_rate": 5.522028966726028e-06, "loss": 1.28, "step": 23874 }, { "epoch": 1.8710815047021945, "grad_norm": 2.846696376800537, "learning_rate": 5.52172228894103e-06, "loss": 1.1957, "step": 23875 }, { "epoch": 1.8711598746081504, "grad_norm": 3.914860725402832, "learning_rate": 5.5214156091716685e-06, "loss": 1.2361, "step": 23876 }, { "epoch": 1.8712382445141067, "grad_norm": 5.569454193115234, "learning_rate": 5.52110892741911e-06, "loss": 1.1582, "step": 23877 }, { "epoch": 1.8713166144200626, "grad_norm": 3.2190918922424316, "learning_rate": 5.52080224368452e-06, "loss": 1.6858, "step": 23878 }, { "epoch": 1.871394984326019, "grad_norm": 3.075047016143799, "learning_rate": 5.5204955579690665e-06, "loss": 1.4749, "step": 23879 }, { "epoch": 1.8714733542319748, "grad_norm": 4.066319465637207, "learning_rate": 5.520188870273915e-06, "loss": 1.1719, "step": 23880 }, { "epoch": 1.8715517241379311, "grad_norm": 2.114302396774292, "learning_rate": 5.5198821806002326e-06, "loss": 1.4427, "step": 23881 }, { "epoch": 1.871630094043887, "grad_norm": 2.4299890995025635, "learning_rate": 5.519575488949184e-06, "loss": 1.4325, "step": 23882 }, { "epoch": 1.8717084639498434, "grad_norm": 2.84829044342041, "learning_rate": 5.51926879532194e-06, "loss": 1.2349, "step": 23883 }, { "epoch": 1.8717868338557992, "grad_norm": 2.324265480041504, "learning_rate": 5.518962099719661e-06, "loss": 1.317, "step": 23884 }, { "epoch": 1.8718652037617556, "grad_norm": 2.4115066528320312, "learning_rate": 5.518655402143518e-06, "loss": 1.5708, "step": 23885 }, { "epoch": 1.8719435736677115, "grad_norm": 2.3175175189971924, "learning_rate": 5.518348702594677e-06, "loss": 1.6492, "step": 23886 }, { "epoch": 1.8720219435736678, "grad_norm": 3.7399628162384033, "learning_rate": 5.518042001074302e-06, "loss": 1.2528, "step": 23887 }, { "epoch": 1.8721003134796237, "grad_norm": 2.2420403957366943, "learning_rate": 5.517735297583563e-06, "loss": 1.784, "step": 23888 }, { "epoch": 1.87217868338558, "grad_norm": 2.2309279441833496, "learning_rate": 5.517428592123624e-06, "loss": 1.5723, "step": 23889 }, { "epoch": 1.872257053291536, "grad_norm": 3.6354167461395264, "learning_rate": 5.517121884695652e-06, "loss": 1.5377, "step": 23890 }, { "epoch": 1.8723354231974922, "grad_norm": 3.4192848205566406, "learning_rate": 5.516815175300814e-06, "loss": 1.7541, "step": 23891 }, { "epoch": 1.8724137931034481, "grad_norm": 2.481086015701294, "learning_rate": 5.516508463940279e-06, "loss": 1.4369, "step": 23892 }, { "epoch": 1.8724921630094045, "grad_norm": 2.346738815307617, "learning_rate": 5.516201750615208e-06, "loss": 1.2397, "step": 23893 }, { "epoch": 1.8725705329153604, "grad_norm": 2.8932089805603027, "learning_rate": 5.5158950353267725e-06, "loss": 1.2839, "step": 23894 }, { "epoch": 1.8726489028213167, "grad_norm": 2.604374647140503, "learning_rate": 5.515588318076137e-06, "loss": 1.1489, "step": 23895 }, { "epoch": 1.8727272727272726, "grad_norm": 4.210163116455078, "learning_rate": 5.515281598864468e-06, "loss": 1.295, "step": 23896 }, { "epoch": 1.872805642633229, "grad_norm": 10.088747024536133, "learning_rate": 5.514974877692932e-06, "loss": 1.5699, "step": 23897 }, { "epoch": 1.8728840125391848, "grad_norm": 5.411129951477051, "learning_rate": 5.514668154562697e-06, "loss": 1.7464, "step": 23898 }, { "epoch": 1.8729623824451411, "grad_norm": 4.200960159301758, "learning_rate": 5.514361429474928e-06, "loss": 1.7153, "step": 23899 }, { "epoch": 1.873040752351097, "grad_norm": 4.636017799377441, "learning_rate": 5.514054702430792e-06, "loss": 1.8448, "step": 23900 }, { "epoch": 1.8731191222570533, "grad_norm": 2.2474308013916016, "learning_rate": 5.513747973431458e-06, "loss": 1.2573, "step": 23901 }, { "epoch": 1.8731974921630095, "grad_norm": 2.186352252960205, "learning_rate": 5.513441242478087e-06, "loss": 1.3006, "step": 23902 }, { "epoch": 1.8732758620689656, "grad_norm": 8.1397066116333, "learning_rate": 5.513134509571853e-06, "loss": 1.8409, "step": 23903 }, { "epoch": 1.8733542319749217, "grad_norm": 8.152997016906738, "learning_rate": 5.512827774713917e-06, "loss": 1.055, "step": 23904 }, { "epoch": 1.8734326018808778, "grad_norm": 2.8725483417510986, "learning_rate": 5.512521037905447e-06, "loss": 1.5949, "step": 23905 }, { "epoch": 1.873510971786834, "grad_norm": 2.845428943634033, "learning_rate": 5.512214299147611e-06, "loss": 1.4787, "step": 23906 }, { "epoch": 1.87358934169279, "grad_norm": 3.800340414047241, "learning_rate": 5.511907558441575e-06, "loss": 1.455, "step": 23907 }, { "epoch": 1.8736677115987461, "grad_norm": 5.190035820007324, "learning_rate": 5.511600815788506e-06, "loss": 1.5451, "step": 23908 }, { "epoch": 1.8737460815047022, "grad_norm": 2.6897621154785156, "learning_rate": 5.511294071189569e-06, "loss": 1.2215, "step": 23909 }, { "epoch": 1.8738244514106583, "grad_norm": 6.009715557098389, "learning_rate": 5.5109873246459325e-06, "loss": 1.2184, "step": 23910 }, { "epoch": 1.8739028213166145, "grad_norm": 2.50630784034729, "learning_rate": 5.510680576158765e-06, "loss": 1.2023, "step": 23911 }, { "epoch": 1.8739811912225706, "grad_norm": 2.7547895908355713, "learning_rate": 5.510373825729227e-06, "loss": 1.444, "step": 23912 }, { "epoch": 1.8740595611285267, "grad_norm": 3.022538423538208, "learning_rate": 5.510067073358492e-06, "loss": 1.1623, "step": 23913 }, { "epoch": 1.8741379310344828, "grad_norm": 3.2963039875030518, "learning_rate": 5.509760319047723e-06, "loss": 1.7305, "step": 23914 }, { "epoch": 1.874216300940439, "grad_norm": 4.348820209503174, "learning_rate": 5.509453562798086e-06, "loss": 1.6868, "step": 23915 }, { "epoch": 1.874294670846395, "grad_norm": 4.628509044647217, "learning_rate": 5.509146804610751e-06, "loss": 1.4013, "step": 23916 }, { "epoch": 1.8743730407523511, "grad_norm": 2.5526387691497803, "learning_rate": 5.508840044486883e-06, "loss": 0.8882, "step": 23917 }, { "epoch": 1.8744514106583072, "grad_norm": 3.6925160884857178, "learning_rate": 5.508533282427649e-06, "loss": 1.5312, "step": 23918 }, { "epoch": 1.8745297805642633, "grad_norm": 1.9821940660476685, "learning_rate": 5.508226518434214e-06, "loss": 1.3003, "step": 23919 }, { "epoch": 1.8746081504702194, "grad_norm": 2.446563959121704, "learning_rate": 5.507919752507749e-06, "loss": 1.3124, "step": 23920 }, { "epoch": 1.8746865203761756, "grad_norm": 5.053555488586426, "learning_rate": 5.507612984649414e-06, "loss": 1.3696, "step": 23921 }, { "epoch": 1.8747648902821317, "grad_norm": 2.5384366512298584, "learning_rate": 5.507306214860384e-06, "loss": 1.1608, "step": 23922 }, { "epoch": 1.8748432601880878, "grad_norm": 6.336667060852051, "learning_rate": 5.50699944314182e-06, "loss": 1.5072, "step": 23923 }, { "epoch": 1.874921630094044, "grad_norm": 4.464630126953125, "learning_rate": 5.5066926694948906e-06, "loss": 1.5969, "step": 23924 }, { "epoch": 1.875, "grad_norm": 3.337557077407837, "learning_rate": 5.506385893920764e-06, "loss": 1.7371, "step": 23925 }, { "epoch": 1.875078369905956, "grad_norm": 2.7892582416534424, "learning_rate": 5.506079116420603e-06, "loss": 1.2072, "step": 23926 }, { "epoch": 1.8751567398119122, "grad_norm": 6.058197498321533, "learning_rate": 5.50577233699558e-06, "loss": 1.2365, "step": 23927 }, { "epoch": 1.8752351097178683, "grad_norm": 4.559134006500244, "learning_rate": 5.505465555646855e-06, "loss": 0.9565, "step": 23928 }, { "epoch": 1.8753134796238244, "grad_norm": 2.4899754524230957, "learning_rate": 5.505158772375602e-06, "loss": 0.9453, "step": 23929 }, { "epoch": 1.8753918495297806, "grad_norm": 2.1365809440612793, "learning_rate": 5.504851987182983e-06, "loss": 1.4025, "step": 23930 }, { "epoch": 1.8754702194357367, "grad_norm": 4.4081878662109375, "learning_rate": 5.504545200070167e-06, "loss": 1.191, "step": 23931 }, { "epoch": 1.8755485893416928, "grad_norm": 7.6323394775390625, "learning_rate": 5.504238411038319e-06, "loss": 0.601, "step": 23932 }, { "epoch": 1.8756269592476489, "grad_norm": 3.355337381362915, "learning_rate": 5.503931620088608e-06, "loss": 1.3299, "step": 23933 }, { "epoch": 1.875705329153605, "grad_norm": 2.7455649375915527, "learning_rate": 5.5036248272222e-06, "loss": 1.0892, "step": 23934 }, { "epoch": 1.875783699059561, "grad_norm": 2.71046781539917, "learning_rate": 5.503318032440261e-06, "loss": 1.3631, "step": 23935 }, { "epoch": 1.8758620689655172, "grad_norm": 2.950294017791748, "learning_rate": 5.50301123574396e-06, "loss": 1.3312, "step": 23936 }, { "epoch": 1.8759404388714733, "grad_norm": 5.126803398132324, "learning_rate": 5.502704437134463e-06, "loss": 1.4388, "step": 23937 }, { "epoch": 1.8760188087774294, "grad_norm": 3.1418979167938232, "learning_rate": 5.502397636612934e-06, "loss": 1.4073, "step": 23938 }, { "epoch": 1.8760971786833855, "grad_norm": 2.110370635986328, "learning_rate": 5.502090834180545e-06, "loss": 1.6843, "step": 23939 }, { "epoch": 1.8761755485893417, "grad_norm": 2.3532774448394775, "learning_rate": 5.501784029838459e-06, "loss": 1.2474, "step": 23940 }, { "epoch": 1.8762539184952978, "grad_norm": 4.700955390930176, "learning_rate": 5.501477223587844e-06, "loss": 1.5264, "step": 23941 }, { "epoch": 1.8763322884012539, "grad_norm": 3.9231605529785156, "learning_rate": 5.501170415429868e-06, "loss": 1.5782, "step": 23942 }, { "epoch": 1.87641065830721, "grad_norm": 4.490715980529785, "learning_rate": 5.500863605365697e-06, "loss": 1.8031, "step": 23943 }, { "epoch": 1.876489028213166, "grad_norm": 3.631983518600464, "learning_rate": 5.5005567933964984e-06, "loss": 1.2501, "step": 23944 }, { "epoch": 1.8765673981191222, "grad_norm": 3.9417991638183594, "learning_rate": 5.500249979523439e-06, "loss": 1.0466, "step": 23945 }, { "epoch": 1.8766457680250783, "grad_norm": 2.4921417236328125, "learning_rate": 5.499943163747685e-06, "loss": 1.3321, "step": 23946 }, { "epoch": 1.8767241379310344, "grad_norm": 2.368018865585327, "learning_rate": 5.499636346070404e-06, "loss": 1.3328, "step": 23947 }, { "epoch": 1.8768025078369905, "grad_norm": 2.557539701461792, "learning_rate": 5.499329526492764e-06, "loss": 1.5844, "step": 23948 }, { "epoch": 1.8768808777429467, "grad_norm": 3.2483248710632324, "learning_rate": 5.499022705015928e-06, "loss": 1.3045, "step": 23949 }, { "epoch": 1.876959247648903, "grad_norm": 3.4888100624084473, "learning_rate": 5.498715881641069e-06, "loss": 1.6267, "step": 23950 }, { "epoch": 1.8770376175548589, "grad_norm": 2.924635887145996, "learning_rate": 5.498409056369351e-06, "loss": 1.1833, "step": 23951 }, { "epoch": 1.8771159874608152, "grad_norm": 14.433538436889648, "learning_rate": 5.498102229201939e-06, "loss": 1.4421, "step": 23952 }, { "epoch": 1.877194357366771, "grad_norm": 2.753033399581909, "learning_rate": 5.497795400140004e-06, "loss": 1.3742, "step": 23953 }, { "epoch": 1.8772727272727274, "grad_norm": 2.40665602684021, "learning_rate": 5.497488569184709e-06, "loss": 1.651, "step": 23954 }, { "epoch": 1.8773510971786833, "grad_norm": 3.226360559463501, "learning_rate": 5.4971817363372245e-06, "loss": 1.1044, "step": 23955 }, { "epoch": 1.8774294670846396, "grad_norm": 2.5016770362854004, "learning_rate": 5.496874901598714e-06, "loss": 1.4187, "step": 23956 }, { "epoch": 1.8775078369905955, "grad_norm": 2.5707693099975586, "learning_rate": 5.496568064970349e-06, "loss": 1.3158, "step": 23957 }, { "epoch": 1.8775862068965519, "grad_norm": 4.706180095672607, "learning_rate": 5.496261226453293e-06, "loss": 1.4035, "step": 23958 }, { "epoch": 1.8776645768025078, "grad_norm": 11.389781951904297, "learning_rate": 5.4959543860487145e-06, "loss": 1.5242, "step": 23959 }, { "epoch": 1.877742946708464, "grad_norm": 4.696955680847168, "learning_rate": 5.495647543757781e-06, "loss": 1.2801, "step": 23960 }, { "epoch": 1.87782131661442, "grad_norm": 4.280242443084717, "learning_rate": 5.495340699581657e-06, "loss": 1.2466, "step": 23961 }, { "epoch": 1.8778996865203763, "grad_norm": 3.170701265335083, "learning_rate": 5.495033853521512e-06, "loss": 1.1372, "step": 23962 }, { "epoch": 1.8779780564263322, "grad_norm": 3.650477170944214, "learning_rate": 5.494727005578513e-06, "loss": 1.614, "step": 23963 }, { "epoch": 1.8780564263322885, "grad_norm": 3.626981735229492, "learning_rate": 5.494420155753826e-06, "loss": 1.316, "step": 23964 }, { "epoch": 1.8781347962382444, "grad_norm": 2.236525774002075, "learning_rate": 5.49411330404862e-06, "loss": 1.1124, "step": 23965 }, { "epoch": 1.8782131661442008, "grad_norm": 3.6993608474731445, "learning_rate": 5.49380645046406e-06, "loss": 1.0836, "step": 23966 }, { "epoch": 1.8782915360501566, "grad_norm": 2.695164203643799, "learning_rate": 5.493499595001314e-06, "loss": 1.2765, "step": 23967 }, { "epoch": 1.878369905956113, "grad_norm": 2.9632771015167236, "learning_rate": 5.4931927376615476e-06, "loss": 1.0253, "step": 23968 }, { "epoch": 1.8784482758620689, "grad_norm": 6.951590538024902, "learning_rate": 5.492885878445931e-06, "loss": 1.5736, "step": 23969 }, { "epoch": 1.8785266457680252, "grad_norm": 6.022696495056152, "learning_rate": 5.4925790173556295e-06, "loss": 1.4742, "step": 23970 }, { "epoch": 1.878605015673981, "grad_norm": 4.365288257598877, "learning_rate": 5.492272154391809e-06, "loss": 1.217, "step": 23971 }, { "epoch": 1.8786833855799374, "grad_norm": 7.437793731689453, "learning_rate": 5.491965289555641e-06, "loss": 1.0041, "step": 23972 }, { "epoch": 1.8787617554858933, "grad_norm": 2.8765692710876465, "learning_rate": 5.491658422848286e-06, "loss": 1.0332, "step": 23973 }, { "epoch": 1.8788401253918496, "grad_norm": 3.2689361572265625, "learning_rate": 5.491351554270918e-06, "loss": 1.2724, "step": 23974 }, { "epoch": 1.8789184952978055, "grad_norm": 3.2022764682769775, "learning_rate": 5.491044683824699e-06, "loss": 1.2969, "step": 23975 }, { "epoch": 1.8789968652037619, "grad_norm": 5.7583842277526855, "learning_rate": 5.4907378115108005e-06, "loss": 1.5573, "step": 23976 }, { "epoch": 1.8790752351097177, "grad_norm": 2.003249406814575, "learning_rate": 5.490430937330385e-06, "loss": 1.4516, "step": 23977 }, { "epoch": 1.879153605015674, "grad_norm": 4.605665683746338, "learning_rate": 5.490124061284623e-06, "loss": 1.2625, "step": 23978 }, { "epoch": 1.87923197492163, "grad_norm": 2.126633644104004, "learning_rate": 5.4898171833746805e-06, "loss": 1.4814, "step": 23979 }, { "epoch": 1.8793103448275863, "grad_norm": 2.552981376647949, "learning_rate": 5.489510303601726e-06, "loss": 1.5789, "step": 23980 }, { "epoch": 1.8793887147335422, "grad_norm": 2.2145543098449707, "learning_rate": 5.4892034219669266e-06, "loss": 1.1065, "step": 23981 }, { "epoch": 1.8794670846394985, "grad_norm": 2.642096519470215, "learning_rate": 5.488896538471446e-06, "loss": 1.3554, "step": 23982 }, { "epoch": 1.8795454545454544, "grad_norm": 2.499974250793457, "learning_rate": 5.488589653116456e-06, "loss": 0.6757, "step": 23983 }, { "epoch": 1.8796238244514107, "grad_norm": 3.0082430839538574, "learning_rate": 5.4882827659031214e-06, "loss": 1.4782, "step": 23984 }, { "epoch": 1.8797021943573666, "grad_norm": 3.3447248935699463, "learning_rate": 5.48797587683261e-06, "loss": 1.6915, "step": 23985 }, { "epoch": 1.879780564263323, "grad_norm": 4.8521904945373535, "learning_rate": 5.4876689859060885e-06, "loss": 1.5206, "step": 23986 }, { "epoch": 1.8798589341692789, "grad_norm": 3.547194480895996, "learning_rate": 5.487362093124724e-06, "loss": 1.019, "step": 23987 }, { "epoch": 1.8799373040752352, "grad_norm": 7.705770492553711, "learning_rate": 5.487055198489686e-06, "loss": 0.9931, "step": 23988 }, { "epoch": 1.880015673981191, "grad_norm": 2.6815576553344727, "learning_rate": 5.486748302002139e-06, "loss": 1.2469, "step": 23989 }, { "epoch": 1.8800940438871474, "grad_norm": 2.45896315574646, "learning_rate": 5.486441403663252e-06, "loss": 0.5395, "step": 23990 }, { "epoch": 1.8801724137931033, "grad_norm": 3.3786659240722656, "learning_rate": 5.486134503474192e-06, "loss": 1.3569, "step": 23991 }, { "epoch": 1.8802507836990596, "grad_norm": 3.598345994949341, "learning_rate": 5.485827601436126e-06, "loss": 1.1903, "step": 23992 }, { "epoch": 1.8803291536050155, "grad_norm": 2.7671751976013184, "learning_rate": 5.485520697550222e-06, "loss": 1.5046, "step": 23993 }, { "epoch": 1.8804075235109718, "grad_norm": 3.2527143955230713, "learning_rate": 5.485213791817645e-06, "loss": 1.3558, "step": 23994 }, { "epoch": 1.8804858934169277, "grad_norm": 6.1242804527282715, "learning_rate": 5.484906884239564e-06, "loss": 1.3179, "step": 23995 }, { "epoch": 1.880564263322884, "grad_norm": 5.016081809997559, "learning_rate": 5.484599974817147e-06, "loss": 1.3758, "step": 23996 }, { "epoch": 1.8806426332288402, "grad_norm": 3.8859221935272217, "learning_rate": 5.484293063551561e-06, "loss": 1.2014, "step": 23997 }, { "epoch": 1.8807210031347963, "grad_norm": 2.1731181144714355, "learning_rate": 5.483986150443973e-06, "loss": 1.0438, "step": 23998 }, { "epoch": 1.8807993730407524, "grad_norm": 2.6648643016815186, "learning_rate": 5.483679235495549e-06, "loss": 1.5282, "step": 23999 }, { "epoch": 1.8808777429467085, "grad_norm": 2.4536404609680176, "learning_rate": 5.48337231870746e-06, "loss": 1.0807, "step": 24000 }, { "epoch": 1.8809561128526646, "grad_norm": 2.089266061782837, "learning_rate": 5.483065400080868e-06, "loss": 0.9271, "step": 24001 }, { "epoch": 1.8810344827586207, "grad_norm": 5.145242214202881, "learning_rate": 5.482758479616947e-06, "loss": 1.0114, "step": 24002 }, { "epoch": 1.8811128526645768, "grad_norm": 2.6471874713897705, "learning_rate": 5.482451557316857e-06, "loss": 1.2688, "step": 24003 }, { "epoch": 1.881191222570533, "grad_norm": 2.2953133583068848, "learning_rate": 5.482144633181772e-06, "loss": 1.4306, "step": 24004 }, { "epoch": 1.881269592476489, "grad_norm": 2.953270673751831, "learning_rate": 5.481837707212854e-06, "loss": 1.5752, "step": 24005 }, { "epoch": 1.8813479623824452, "grad_norm": 2.536466121673584, "learning_rate": 5.481530779411273e-06, "loss": 1.2361, "step": 24006 }, { "epoch": 1.8814263322884013, "grad_norm": 4.284120082855225, "learning_rate": 5.481223849778198e-06, "loss": 1.2481, "step": 24007 }, { "epoch": 1.8815047021943574, "grad_norm": 2.3966891765594482, "learning_rate": 5.480916918314794e-06, "loss": 1.2831, "step": 24008 }, { "epoch": 1.8815830721003135, "grad_norm": 2.8213589191436768, "learning_rate": 5.480609985022229e-06, "loss": 1.4609, "step": 24009 }, { "epoch": 1.8816614420062696, "grad_norm": 3.664590358734131, "learning_rate": 5.480303049901669e-06, "loss": 1.5749, "step": 24010 }, { "epoch": 1.8817398119122257, "grad_norm": 2.2521095275878906, "learning_rate": 5.4799961129542855e-06, "loss": 1.4547, "step": 24011 }, { "epoch": 1.8818181818181818, "grad_norm": 2.0550625324249268, "learning_rate": 5.479689174181242e-06, "loss": 1.2863, "step": 24012 }, { "epoch": 1.881896551724138, "grad_norm": 2.917419910430908, "learning_rate": 5.479382233583708e-06, "loss": 1.7467, "step": 24013 }, { "epoch": 1.881974921630094, "grad_norm": 3.229452133178711, "learning_rate": 5.479075291162849e-06, "loss": 1.0994, "step": 24014 }, { "epoch": 1.8820532915360502, "grad_norm": 3.3205347061157227, "learning_rate": 5.478768346919834e-06, "loss": 1.1322, "step": 24015 }, { "epoch": 1.8821316614420063, "grad_norm": 2.6618363857269287, "learning_rate": 5.47846140085583e-06, "loss": 1.5655, "step": 24016 }, { "epoch": 1.8822100313479624, "grad_norm": 2.7886033058166504, "learning_rate": 5.478154452972006e-06, "loss": 1.3737, "step": 24017 }, { "epoch": 1.8822884012539185, "grad_norm": 3.546732187271118, "learning_rate": 5.477847503269525e-06, "loss": 1.6657, "step": 24018 }, { "epoch": 1.8823667711598746, "grad_norm": 2.3445467948913574, "learning_rate": 5.477540551749561e-06, "loss": 1.4111, "step": 24019 }, { "epoch": 1.8824451410658307, "grad_norm": 6.826140880584717, "learning_rate": 5.477233598413277e-06, "loss": 1.1743, "step": 24020 }, { "epoch": 1.8825235109717868, "grad_norm": 3.2364954948425293, "learning_rate": 5.476926643261841e-06, "loss": 1.2719, "step": 24021 }, { "epoch": 1.882601880877743, "grad_norm": 2.0754804611206055, "learning_rate": 5.476619686296421e-06, "loss": 1.1925, "step": 24022 }, { "epoch": 1.882680250783699, "grad_norm": 1.8801411390304565, "learning_rate": 5.476312727518184e-06, "loss": 1.2601, "step": 24023 }, { "epoch": 1.8827586206896552, "grad_norm": 2.732562780380249, "learning_rate": 5.476005766928299e-06, "loss": 1.1128, "step": 24024 }, { "epoch": 1.8828369905956113, "grad_norm": 7.306100845336914, "learning_rate": 5.475698804527931e-06, "loss": 1.1259, "step": 24025 }, { "epoch": 1.8829153605015674, "grad_norm": 8.231517791748047, "learning_rate": 5.475391840318252e-06, "loss": 1.6917, "step": 24026 }, { "epoch": 1.8829937304075235, "grad_norm": 2.42213773727417, "learning_rate": 5.475084874300423e-06, "loss": 1.1746, "step": 24027 }, { "epoch": 1.8830721003134796, "grad_norm": 2.5998547077178955, "learning_rate": 5.474777906475619e-06, "loss": 1.4034, "step": 24028 }, { "epoch": 1.8831504702194357, "grad_norm": 2.5905370712280273, "learning_rate": 5.474470936845e-06, "loss": 1.072, "step": 24029 }, { "epoch": 1.8832288401253918, "grad_norm": 5.359162330627441, "learning_rate": 5.47416396540974e-06, "loss": 1.1201, "step": 24030 }, { "epoch": 1.883307210031348, "grad_norm": 2.726696491241455, "learning_rate": 5.4738569921710025e-06, "loss": 1.35, "step": 24031 }, { "epoch": 1.883385579937304, "grad_norm": 3.9509294033050537, "learning_rate": 5.4735500171299576e-06, "loss": 0.9126, "step": 24032 }, { "epoch": 1.8834639498432602, "grad_norm": 8.398825645446777, "learning_rate": 5.47324304028777e-06, "loss": 1.9156, "step": 24033 }, { "epoch": 1.8835423197492163, "grad_norm": 5.939862251281738, "learning_rate": 5.472936061645609e-06, "loss": 1.3743, "step": 24034 }, { "epoch": 1.8836206896551724, "grad_norm": 7.270759105682373, "learning_rate": 5.472629081204643e-06, "loss": 1.653, "step": 24035 }, { "epoch": 1.8836990595611285, "grad_norm": 2.854599952697754, "learning_rate": 5.472322098966039e-06, "loss": 1.5697, "step": 24036 }, { "epoch": 1.8837774294670846, "grad_norm": 3.8041532039642334, "learning_rate": 5.4720151149309644e-06, "loss": 1.763, "step": 24037 }, { "epoch": 1.8838557993730407, "grad_norm": 3.917574882507324, "learning_rate": 5.471708129100586e-06, "loss": 1.54, "step": 24038 }, { "epoch": 1.8839341692789968, "grad_norm": 3.3924648761749268, "learning_rate": 5.471401141476073e-06, "loss": 1.84, "step": 24039 }, { "epoch": 1.884012539184953, "grad_norm": 2.655320405960083, "learning_rate": 5.471094152058592e-06, "loss": 1.0033, "step": 24040 }, { "epoch": 1.884090909090909, "grad_norm": 3.6154539585113525, "learning_rate": 5.470787160849309e-06, "loss": 1.5721, "step": 24041 }, { "epoch": 1.8841692789968651, "grad_norm": 3.3723318576812744, "learning_rate": 5.470480167849395e-06, "loss": 0.9785, "step": 24042 }, { "epoch": 1.8842476489028215, "grad_norm": 4.0445051193237305, "learning_rate": 5.470173173060016e-06, "loss": 1.8011, "step": 24043 }, { "epoch": 1.8843260188087774, "grad_norm": 2.860734224319458, "learning_rate": 5.469866176482339e-06, "loss": 1.4961, "step": 24044 }, { "epoch": 1.8844043887147337, "grad_norm": 2.6715991497039795, "learning_rate": 5.4695591781175335e-06, "loss": 1.0485, "step": 24045 }, { "epoch": 1.8844827586206896, "grad_norm": 3.071669340133667, "learning_rate": 5.469252177966765e-06, "loss": 1.6018, "step": 24046 }, { "epoch": 1.884561128526646, "grad_norm": 4.532118797302246, "learning_rate": 5.468945176031204e-06, "loss": 1.8096, "step": 24047 }, { "epoch": 1.8846394984326018, "grad_norm": 2.0882105827331543, "learning_rate": 5.468638172312014e-06, "loss": 1.4438, "step": 24048 }, { "epoch": 1.8847178683385581, "grad_norm": 3.6340062618255615, "learning_rate": 5.468331166810368e-06, "loss": 1.6804, "step": 24049 }, { "epoch": 1.884796238244514, "grad_norm": 2.4405877590179443, "learning_rate": 5.468024159527429e-06, "loss": 1.3143, "step": 24050 }, { "epoch": 1.8848746081504704, "grad_norm": 2.3482367992401123, "learning_rate": 5.467717150464367e-06, "loss": 1.4223, "step": 24051 }, { "epoch": 1.8849529780564263, "grad_norm": 3.171556234359741, "learning_rate": 5.467410139622348e-06, "loss": 1.4911, "step": 24052 }, { "epoch": 1.8850313479623826, "grad_norm": 3.6464014053344727, "learning_rate": 5.467103127002542e-06, "loss": 1.5773, "step": 24053 }, { "epoch": 1.8851097178683385, "grad_norm": 2.6126813888549805, "learning_rate": 5.466796112606117e-06, "loss": 1.3017, "step": 24054 }, { "epoch": 1.8851880877742948, "grad_norm": 4.23748779296875, "learning_rate": 5.466489096434235e-06, "loss": 0.6653, "step": 24055 }, { "epoch": 1.8852664576802507, "grad_norm": 3.3083078861236572, "learning_rate": 5.466182078488073e-06, "loss": 0.8948, "step": 24056 }, { "epoch": 1.885344827586207, "grad_norm": 2.9535763263702393, "learning_rate": 5.46587505876879e-06, "loss": 1.1259, "step": 24057 }, { "epoch": 1.885423197492163, "grad_norm": 2.0851211547851562, "learning_rate": 5.465568037277561e-06, "loss": 1.2235, "step": 24058 }, { "epoch": 1.8855015673981192, "grad_norm": 3.2645211219787598, "learning_rate": 5.465261014015548e-06, "loss": 1.4373, "step": 24059 }, { "epoch": 1.8855799373040751, "grad_norm": 3.7704336643218994, "learning_rate": 5.464953988983921e-06, "loss": 1.7821, "step": 24060 }, { "epoch": 1.8856583072100315, "grad_norm": 2.245814800262451, "learning_rate": 5.464646962183849e-06, "loss": 1.5709, "step": 24061 }, { "epoch": 1.8857366771159874, "grad_norm": 3.17830753326416, "learning_rate": 5.464339933616498e-06, "loss": 0.6434, "step": 24062 }, { "epoch": 1.8858150470219437, "grad_norm": 2.3998496532440186, "learning_rate": 5.464032903283036e-06, "loss": 1.0433, "step": 24063 }, { "epoch": 1.8858934169278996, "grad_norm": 3.8445966243743896, "learning_rate": 5.463725871184632e-06, "loss": 0.8733, "step": 24064 }, { "epoch": 1.885971786833856, "grad_norm": 1.967462182044983, "learning_rate": 5.463418837322454e-06, "loss": 1.2763, "step": 24065 }, { "epoch": 1.8860501567398118, "grad_norm": 3.484734296798706, "learning_rate": 5.463111801697667e-06, "loss": 1.1347, "step": 24066 }, { "epoch": 1.8861285266457681, "grad_norm": 6.091213226318359, "learning_rate": 5.46280476431144e-06, "loss": 1.4369, "step": 24067 }, { "epoch": 1.886206896551724, "grad_norm": 2.973724603652954, "learning_rate": 5.462497725164943e-06, "loss": 1.2671, "step": 24068 }, { "epoch": 1.8862852664576804, "grad_norm": 2.9862804412841797, "learning_rate": 5.462190684259341e-06, "loss": 1.3455, "step": 24069 }, { "epoch": 1.8863636363636362, "grad_norm": 3.087761640548706, "learning_rate": 5.461883641595804e-06, "loss": 1.4434, "step": 24070 }, { "epoch": 1.8864420062695926, "grad_norm": 2.7040019035339355, "learning_rate": 5.4615765971754985e-06, "loss": 1.5032, "step": 24071 }, { "epoch": 1.8865203761755485, "grad_norm": 2.7525508403778076, "learning_rate": 5.4612695509995925e-06, "loss": 1.4124, "step": 24072 }, { "epoch": 1.8865987460815048, "grad_norm": 2.9938340187072754, "learning_rate": 5.460962503069255e-06, "loss": 1.2131, "step": 24073 }, { "epoch": 1.8866771159874607, "grad_norm": 2.6672418117523193, "learning_rate": 5.46065545338565e-06, "loss": 1.4256, "step": 24074 }, { "epoch": 1.886755485893417, "grad_norm": 2.055039167404175, "learning_rate": 5.460348401949952e-06, "loss": 1.3811, "step": 24075 }, { "epoch": 1.886833855799373, "grad_norm": 9.045796394348145, "learning_rate": 5.460041348763323e-06, "loss": 1.048, "step": 24076 }, { "epoch": 1.8869122257053292, "grad_norm": 2.620220184326172, "learning_rate": 5.459734293826934e-06, "loss": 1.1451, "step": 24077 }, { "epoch": 1.8869905956112851, "grad_norm": 3.2425293922424316, "learning_rate": 5.459427237141951e-06, "loss": 1.6429, "step": 24078 }, { "epoch": 1.8870689655172415, "grad_norm": 10.068544387817383, "learning_rate": 5.4591201787095435e-06, "loss": 1.9812, "step": 24079 }, { "epoch": 1.8871473354231973, "grad_norm": 8.802413940429688, "learning_rate": 5.458813118530879e-06, "loss": 1.4127, "step": 24080 }, { "epoch": 1.8872257053291537, "grad_norm": 2.9982504844665527, "learning_rate": 5.458506056607124e-06, "loss": 1.4462, "step": 24081 }, { "epoch": 1.8873040752351096, "grad_norm": 2.217083692550659, "learning_rate": 5.458198992939449e-06, "loss": 1.4563, "step": 24082 }, { "epoch": 1.887382445141066, "grad_norm": 2.687072277069092, "learning_rate": 5.457891927529018e-06, "loss": 1.2964, "step": 24083 }, { "epoch": 1.8874608150470218, "grad_norm": 2.3630599975585938, "learning_rate": 5.457584860377003e-06, "loss": 1.3226, "step": 24084 }, { "epoch": 1.8875391849529781, "grad_norm": 3.4070985317230225, "learning_rate": 5.45727779148457e-06, "loss": 1.0464, "step": 24085 }, { "epoch": 1.887617554858934, "grad_norm": 2.647223711013794, "learning_rate": 5.456970720852887e-06, "loss": 1.5281, "step": 24086 }, { "epoch": 1.8876959247648903, "grad_norm": 2.7617762088775635, "learning_rate": 5.456663648483121e-06, "loss": 1.2611, "step": 24087 }, { "epoch": 1.8877742946708462, "grad_norm": 5.129970073699951, "learning_rate": 5.456356574376442e-06, "loss": 1.0621, "step": 24088 }, { "epoch": 1.8878526645768026, "grad_norm": 3.267948627471924, "learning_rate": 5.4560494985340176e-06, "loss": 1.2983, "step": 24089 }, { "epoch": 1.8879310344827587, "grad_norm": 6.572132587432861, "learning_rate": 5.4557424209570145e-06, "loss": 0.8102, "step": 24090 }, { "epoch": 1.8880094043887148, "grad_norm": 3.5730111598968506, "learning_rate": 5.455435341646601e-06, "loss": 1.2413, "step": 24091 }, { "epoch": 1.888087774294671, "grad_norm": 6.203868389129639, "learning_rate": 5.4551282606039465e-06, "loss": 1.405, "step": 24092 }, { "epoch": 1.888166144200627, "grad_norm": 10.571846008300781, "learning_rate": 5.4548211778302176e-06, "loss": 0.7773, "step": 24093 }, { "epoch": 1.8882445141065831, "grad_norm": 3.4468071460723877, "learning_rate": 5.454514093326581e-06, "loss": 1.377, "step": 24094 }, { "epoch": 1.8883228840125392, "grad_norm": 4.251000881195068, "learning_rate": 5.454207007094208e-06, "loss": 1.3375, "step": 24095 }, { "epoch": 1.8884012539184953, "grad_norm": 3.1538469791412354, "learning_rate": 5.453899919134263e-06, "loss": 1.6596, "step": 24096 }, { "epoch": 1.8884796238244514, "grad_norm": 2.2116732597351074, "learning_rate": 5.453592829447918e-06, "loss": 1.2648, "step": 24097 }, { "epoch": 1.8885579937304076, "grad_norm": 2.0527946949005127, "learning_rate": 5.4532857380363365e-06, "loss": 1.345, "step": 24098 }, { "epoch": 1.8886363636363637, "grad_norm": 5.854623317718506, "learning_rate": 5.45297864490069e-06, "loss": 1.5085, "step": 24099 }, { "epoch": 1.8887147335423198, "grad_norm": 5.126428127288818, "learning_rate": 5.4526715500421465e-06, "loss": 1.4808, "step": 24100 }, { "epoch": 1.888793103448276, "grad_norm": 5.479525089263916, "learning_rate": 5.452364453461874e-06, "loss": 1.761, "step": 24101 }, { "epoch": 1.888871473354232, "grad_norm": 2.4140639305114746, "learning_rate": 5.452057355161036e-06, "loss": 1.2078, "step": 24102 }, { "epoch": 1.888949843260188, "grad_norm": 3.8967647552490234, "learning_rate": 5.451750255140806e-06, "loss": 0.7759, "step": 24103 }, { "epoch": 1.8890282131661442, "grad_norm": 2.2919583320617676, "learning_rate": 5.4514431534023484e-06, "loss": 1.2964, "step": 24104 }, { "epoch": 1.8891065830721003, "grad_norm": 3.3376312255859375, "learning_rate": 5.451136049946835e-06, "loss": 1.6098, "step": 24105 }, { "epoch": 1.8891849529780564, "grad_norm": 2.4444146156311035, "learning_rate": 5.45082894477543e-06, "loss": 1.4131, "step": 24106 }, { "epoch": 1.8892633228840126, "grad_norm": 3.8881406784057617, "learning_rate": 5.450521837889305e-06, "loss": 1.8892, "step": 24107 }, { "epoch": 1.8893416927899687, "grad_norm": 3.2523603439331055, "learning_rate": 5.450214729289627e-06, "loss": 1.7777, "step": 24108 }, { "epoch": 1.8894200626959248, "grad_norm": 4.6697869300842285, "learning_rate": 5.44990761897756e-06, "loss": 1.2024, "step": 24109 }, { "epoch": 1.8894984326018809, "grad_norm": 10.477845191955566, "learning_rate": 5.449600506954279e-06, "loss": 1.7578, "step": 24110 }, { "epoch": 1.889576802507837, "grad_norm": 2.5598368644714355, "learning_rate": 5.449293393220946e-06, "loss": 1.3163, "step": 24111 }, { "epoch": 1.889655172413793, "grad_norm": 1.936277985572815, "learning_rate": 5.448986277778732e-06, "loss": 1.5528, "step": 24112 }, { "epoch": 1.8897335423197492, "grad_norm": 2.8191897869110107, "learning_rate": 5.448679160628807e-06, "loss": 1.7363, "step": 24113 }, { "epoch": 1.8898119122257053, "grad_norm": 6.5673828125, "learning_rate": 5.4483720417723356e-06, "loss": 1.1235, "step": 24114 }, { "epoch": 1.8898902821316614, "grad_norm": 2.488762617111206, "learning_rate": 5.4480649212104865e-06, "loss": 1.3407, "step": 24115 }, { "epoch": 1.8899686520376175, "grad_norm": 2.894080877304077, "learning_rate": 5.447757798944428e-06, "loss": 1.2502, "step": 24116 }, { "epoch": 1.8900470219435737, "grad_norm": 5.082192897796631, "learning_rate": 5.447450674975331e-06, "loss": 1.1246, "step": 24117 }, { "epoch": 1.8901253918495298, "grad_norm": 4.0545477867126465, "learning_rate": 5.44714354930436e-06, "loss": 1.7099, "step": 24118 }, { "epoch": 1.8902037617554859, "grad_norm": 6.814266204833984, "learning_rate": 5.446836421932685e-06, "loss": 1.0325, "step": 24119 }, { "epoch": 1.890282131661442, "grad_norm": 2.264012575149536, "learning_rate": 5.446529292861475e-06, "loss": 1.3588, "step": 24120 }, { "epoch": 1.890360501567398, "grad_norm": 4.5064239501953125, "learning_rate": 5.446222162091896e-06, "loss": 1.5592, "step": 24121 }, { "epoch": 1.8904388714733542, "grad_norm": 3.430046558380127, "learning_rate": 5.445915029625116e-06, "loss": 1.1499, "step": 24122 }, { "epoch": 1.8905172413793103, "grad_norm": 2.5508742332458496, "learning_rate": 5.445607895462306e-06, "loss": 0.5604, "step": 24123 }, { "epoch": 1.8905956112852664, "grad_norm": 2.489030122756958, "learning_rate": 5.445300759604632e-06, "loss": 0.9467, "step": 24124 }, { "epoch": 1.8906739811912225, "grad_norm": 2.724482297897339, "learning_rate": 5.444993622053261e-06, "loss": 1.1139, "step": 24125 }, { "epoch": 1.8907523510971787, "grad_norm": 2.5492653846740723, "learning_rate": 5.444686482809364e-06, "loss": 1.323, "step": 24126 }, { "epoch": 1.8908307210031348, "grad_norm": 2.601891040802002, "learning_rate": 5.444379341874109e-06, "loss": 0.886, "step": 24127 }, { "epoch": 1.8909090909090909, "grad_norm": 1.6771939992904663, "learning_rate": 5.4440721992486615e-06, "loss": 1.3081, "step": 24128 }, { "epoch": 1.890987460815047, "grad_norm": 3.670304775238037, "learning_rate": 5.443765054934193e-06, "loss": 1.1026, "step": 24129 }, { "epoch": 1.891065830721003, "grad_norm": 2.8580856323242188, "learning_rate": 5.443457908931868e-06, "loss": 0.7883, "step": 24130 }, { "epoch": 1.8911442006269592, "grad_norm": 2.7826170921325684, "learning_rate": 5.443150761242859e-06, "loss": 0.5998, "step": 24131 }, { "epoch": 1.8912225705329153, "grad_norm": 2.386355400085449, "learning_rate": 5.442843611868331e-06, "loss": 1.3782, "step": 24132 }, { "epoch": 1.8913009404388714, "grad_norm": 3.0137088298797607, "learning_rate": 5.442536460809454e-06, "loss": 0.7796, "step": 24133 }, { "epoch": 1.8913793103448275, "grad_norm": 2.1949424743652344, "learning_rate": 5.442229308067397e-06, "loss": 1.251, "step": 24134 }, { "epoch": 1.8914576802507836, "grad_norm": 2.732649087905884, "learning_rate": 5.441922153643323e-06, "loss": 1.5094, "step": 24135 }, { "epoch": 1.8915360501567398, "grad_norm": 2.928525447845459, "learning_rate": 5.441614997538408e-06, "loss": 1.7454, "step": 24136 }, { "epoch": 1.8916144200626959, "grad_norm": 3.633172035217285, "learning_rate": 5.441307839753814e-06, "loss": 1.331, "step": 24137 }, { "epoch": 1.8916927899686522, "grad_norm": 3.424309492111206, "learning_rate": 5.4410006802907136e-06, "loss": 1.5023, "step": 24138 }, { "epoch": 1.891771159874608, "grad_norm": 3.0774691104888916, "learning_rate": 5.440693519150271e-06, "loss": 1.0205, "step": 24139 }, { "epoch": 1.8918495297805644, "grad_norm": 2.578242540359497, "learning_rate": 5.440386356333658e-06, "loss": 1.2006, "step": 24140 }, { "epoch": 1.8919278996865203, "grad_norm": 2.658015489578247, "learning_rate": 5.440079191842041e-06, "loss": 1.2152, "step": 24141 }, { "epoch": 1.8920062695924766, "grad_norm": 4.7470831871032715, "learning_rate": 5.439772025676588e-06, "loss": 1.6627, "step": 24142 }, { "epoch": 1.8920846394984325, "grad_norm": 4.391317367553711, "learning_rate": 5.439464857838469e-06, "loss": 1.4518, "step": 24143 }, { "epoch": 1.8921630094043889, "grad_norm": 2.398822069168091, "learning_rate": 5.439157688328851e-06, "loss": 1.3021, "step": 24144 }, { "epoch": 1.8922413793103448, "grad_norm": 2.537830114364624, "learning_rate": 5.438850517148902e-06, "loss": 1.0835, "step": 24145 }, { "epoch": 1.892319749216301, "grad_norm": 4.159834861755371, "learning_rate": 5.438543344299792e-06, "loss": 1.4532, "step": 24146 }, { "epoch": 1.892398119122257, "grad_norm": 4.459961891174316, "learning_rate": 5.438236169782689e-06, "loss": 1.5112, "step": 24147 }, { "epoch": 1.8924764890282133, "grad_norm": 2.555860996246338, "learning_rate": 5.437928993598761e-06, "loss": 1.2601, "step": 24148 }, { "epoch": 1.8925548589341692, "grad_norm": 2.9067564010620117, "learning_rate": 5.4376218157491755e-06, "loss": 1.227, "step": 24149 }, { "epoch": 1.8926332288401255, "grad_norm": 3.314582347869873, "learning_rate": 5.4373146362351005e-06, "loss": 1.5056, "step": 24150 }, { "epoch": 1.8927115987460814, "grad_norm": 2.749809980392456, "learning_rate": 5.437007455057706e-06, "loss": 1.6963, "step": 24151 }, { "epoch": 1.8927899686520377, "grad_norm": 5.383923053741455, "learning_rate": 5.436700272218159e-06, "loss": 1.1116, "step": 24152 }, { "epoch": 1.8928683385579936, "grad_norm": 3.3250460624694824, "learning_rate": 5.4363930877176285e-06, "loss": 1.37, "step": 24153 }, { "epoch": 1.89294670846395, "grad_norm": 4.79852819442749, "learning_rate": 5.436085901557283e-06, "loss": 0.8406, "step": 24154 }, { "epoch": 1.8930250783699059, "grad_norm": 2.745112180709839, "learning_rate": 5.435778713738292e-06, "loss": 1.232, "step": 24155 }, { "epoch": 1.8931034482758622, "grad_norm": 3.258066177368164, "learning_rate": 5.435471524261821e-06, "loss": 1.2982, "step": 24156 }, { "epoch": 1.893181818181818, "grad_norm": 3.125359535217285, "learning_rate": 5.435164333129041e-06, "loss": 1.5108, "step": 24157 }, { "epoch": 1.8932601880877744, "grad_norm": 2.6274712085723877, "learning_rate": 5.434857140341117e-06, "loss": 1.2136, "step": 24158 }, { "epoch": 1.8933385579937303, "grad_norm": 8.73509407043457, "learning_rate": 5.434549945899223e-06, "loss": 1.096, "step": 24159 }, { "epoch": 1.8934169278996866, "grad_norm": 4.5323944091796875, "learning_rate": 5.434242749804523e-06, "loss": 1.3412, "step": 24160 }, { "epoch": 1.8934952978056425, "grad_norm": 2.675879716873169, "learning_rate": 5.433935552058187e-06, "loss": 1.2999, "step": 24161 }, { "epoch": 1.8935736677115989, "grad_norm": 7.928797245025635, "learning_rate": 5.433628352661383e-06, "loss": 1.3629, "step": 24162 }, { "epoch": 1.8936520376175547, "grad_norm": 3.1696858406066895, "learning_rate": 5.433321151615278e-06, "loss": 1.3907, "step": 24163 }, { "epoch": 1.893730407523511, "grad_norm": 2.5192596912384033, "learning_rate": 5.433013948921045e-06, "loss": 1.4806, "step": 24164 }, { "epoch": 1.893808777429467, "grad_norm": 7.5835113525390625, "learning_rate": 5.432706744579845e-06, "loss": 1.6478, "step": 24165 }, { "epoch": 1.8938871473354233, "grad_norm": 3.3132574558258057, "learning_rate": 5.432399538592854e-06, "loss": 1.351, "step": 24166 }, { "epoch": 1.8939655172413792, "grad_norm": 2.9764230251312256, "learning_rate": 5.432092330961236e-06, "loss": 1.3236, "step": 24167 }, { "epoch": 1.8940438871473355, "grad_norm": 7.166687488555908, "learning_rate": 5.431785121686162e-06, "loss": 1.4671, "step": 24168 }, { "epoch": 1.8941222570532914, "grad_norm": 3.10502290725708, "learning_rate": 5.431477910768798e-06, "loss": 1.327, "step": 24169 }, { "epoch": 1.8942006269592477, "grad_norm": 2.276918411254883, "learning_rate": 5.431170698210313e-06, "loss": 1.1651, "step": 24170 }, { "epoch": 1.8942789968652036, "grad_norm": 2.5388989448547363, "learning_rate": 5.430863484011877e-06, "loss": 1.6987, "step": 24171 }, { "epoch": 1.89435736677116, "grad_norm": 3.327497720718384, "learning_rate": 5.430556268174657e-06, "loss": 1.6451, "step": 24172 }, { "epoch": 1.8944357366771158, "grad_norm": 2.747687578201294, "learning_rate": 5.430249050699823e-06, "loss": 1.2223, "step": 24173 }, { "epoch": 1.8945141065830722, "grad_norm": 3.088106155395508, "learning_rate": 5.429941831588543e-06, "loss": 1.5053, "step": 24174 }, { "epoch": 1.894592476489028, "grad_norm": 1.994483470916748, "learning_rate": 5.429634610841984e-06, "loss": 1.6736, "step": 24175 }, { "epoch": 1.8946708463949844, "grad_norm": 3.003161668777466, "learning_rate": 5.429327388461316e-06, "loss": 1.5303, "step": 24176 }, { "epoch": 1.8947492163009403, "grad_norm": 2.9758827686309814, "learning_rate": 5.4290201644477085e-06, "loss": 1.3391, "step": 24177 }, { "epoch": 1.8948275862068966, "grad_norm": 5.71444845199585, "learning_rate": 5.428712938802326e-06, "loss": 1.2677, "step": 24178 }, { "epoch": 1.8949059561128525, "grad_norm": 4.387885570526123, "learning_rate": 5.428405711526341e-06, "loss": 1.37, "step": 24179 }, { "epoch": 1.8949843260188088, "grad_norm": 2.706505060195923, "learning_rate": 5.428098482620921e-06, "loss": 1.3775, "step": 24180 }, { "epoch": 1.8950626959247647, "grad_norm": 4.048715114593506, "learning_rate": 5.427791252087234e-06, "loss": 1.1019, "step": 24181 }, { "epoch": 1.895141065830721, "grad_norm": 3.2602131366729736, "learning_rate": 5.427484019926449e-06, "loss": 0.8361, "step": 24182 }, { "epoch": 1.895219435736677, "grad_norm": 2.895725727081299, "learning_rate": 5.427176786139735e-06, "loss": 1.2957, "step": 24183 }, { "epoch": 1.8952978056426333, "grad_norm": 2.032384157180786, "learning_rate": 5.426869550728257e-06, "loss": 1.5588, "step": 24184 }, { "epoch": 1.8953761755485894, "grad_norm": 2.725212812423706, "learning_rate": 5.4265623136931896e-06, "loss": 1.0802, "step": 24185 }, { "epoch": 1.8954545454545455, "grad_norm": 3.20621657371521, "learning_rate": 5.426255075035698e-06, "loss": 1.301, "step": 24186 }, { "epoch": 1.8955329153605016, "grad_norm": 4.335665225982666, "learning_rate": 5.42594783475695e-06, "loss": 1.4284, "step": 24187 }, { "epoch": 1.8956112852664577, "grad_norm": 2.314060688018799, "learning_rate": 5.425640592858115e-06, "loss": 1.4174, "step": 24188 }, { "epoch": 1.8956896551724138, "grad_norm": 4.697861194610596, "learning_rate": 5.4253333493403625e-06, "loss": 1.1788, "step": 24189 }, { "epoch": 1.89576802507837, "grad_norm": 2.528610944747925, "learning_rate": 5.42502610420486e-06, "loss": 1.3654, "step": 24190 }, { "epoch": 1.895846394984326, "grad_norm": 3.2965452671051025, "learning_rate": 5.424718857452776e-06, "loss": 1.1905, "step": 24191 }, { "epoch": 1.8959247648902822, "grad_norm": 2.6580164432525635, "learning_rate": 5.424411609085282e-06, "loss": 1.7026, "step": 24192 }, { "epoch": 1.8960031347962383, "grad_norm": 22.852861404418945, "learning_rate": 5.424104359103541e-06, "loss": 1.2276, "step": 24193 }, { "epoch": 1.8960815047021944, "grad_norm": 3.2792656421661377, "learning_rate": 5.423797107508727e-06, "loss": 1.3139, "step": 24194 }, { "epoch": 1.8961598746081505, "grad_norm": 2.5615434646606445, "learning_rate": 5.423489854302005e-06, "loss": 1.4728, "step": 24195 }, { "epoch": 1.8962382445141066, "grad_norm": 3.6331284046173096, "learning_rate": 5.423182599484547e-06, "loss": 1.5858, "step": 24196 }, { "epoch": 1.8963166144200627, "grad_norm": 2.831334352493286, "learning_rate": 5.422875343057519e-06, "loss": 1.2489, "step": 24197 }, { "epoch": 1.8963949843260188, "grad_norm": 14.386825561523438, "learning_rate": 5.4225680850220895e-06, "loss": 1.6533, "step": 24198 }, { "epoch": 1.896473354231975, "grad_norm": 4.9859538078308105, "learning_rate": 5.422260825379429e-06, "loss": 1.2492, "step": 24199 }, { "epoch": 1.896551724137931, "grad_norm": 2.6976680755615234, "learning_rate": 5.421953564130704e-06, "loss": 1.4849, "step": 24200 }, { "epoch": 1.8966300940438872, "grad_norm": 2.012253522872925, "learning_rate": 5.421646301277085e-06, "loss": 1.568, "step": 24201 }, { "epoch": 1.8967084639498433, "grad_norm": 2.7589187622070312, "learning_rate": 5.421339036819741e-06, "loss": 1.3138, "step": 24202 }, { "epoch": 1.8967868338557994, "grad_norm": 2.5043604373931885, "learning_rate": 5.421031770759838e-06, "loss": 1.3959, "step": 24203 }, { "epoch": 1.8968652037617555, "grad_norm": 4.365967750549316, "learning_rate": 5.420724503098549e-06, "loss": 1.7504, "step": 24204 }, { "epoch": 1.8969435736677116, "grad_norm": 2.378653049468994, "learning_rate": 5.420417233837038e-06, "loss": 1.0267, "step": 24205 }, { "epoch": 1.8970219435736677, "grad_norm": 4.878999710083008, "learning_rate": 5.420109962976476e-06, "loss": 1.2888, "step": 24206 }, { "epoch": 1.8971003134796238, "grad_norm": 2.8363749980926514, "learning_rate": 5.419802690518032e-06, "loss": 1.2868, "step": 24207 }, { "epoch": 1.89717868338558, "grad_norm": 6.613383769989014, "learning_rate": 5.419495416462873e-06, "loss": 1.2086, "step": 24208 }, { "epoch": 1.897257053291536, "grad_norm": 3.552774667739868, "learning_rate": 5.419188140812172e-06, "loss": 1.1199, "step": 24209 }, { "epoch": 1.8973354231974922, "grad_norm": 4.249411106109619, "learning_rate": 5.418880863567091e-06, "loss": 1.2797, "step": 24210 }, { "epoch": 1.8974137931034483, "grad_norm": 3.4443013668060303, "learning_rate": 5.418573584728805e-06, "loss": 1.3274, "step": 24211 }, { "epoch": 1.8974921630094044, "grad_norm": 3.072186231613159, "learning_rate": 5.4182663042984775e-06, "loss": 1.2498, "step": 24212 }, { "epoch": 1.8975705329153605, "grad_norm": 2.0209972858428955, "learning_rate": 5.417959022277282e-06, "loss": 1.4856, "step": 24213 }, { "epoch": 1.8976489028213166, "grad_norm": 7.308231830596924, "learning_rate": 5.417651738666384e-06, "loss": 1.0485, "step": 24214 }, { "epoch": 1.8977272727272727, "grad_norm": 3.6359312534332275, "learning_rate": 5.417344453466953e-06, "loss": 1.2017, "step": 24215 }, { "epoch": 1.8978056426332288, "grad_norm": 3.8314714431762695, "learning_rate": 5.417037166680159e-06, "loss": 1.1984, "step": 24216 }, { "epoch": 1.897884012539185, "grad_norm": 4.962128639221191, "learning_rate": 5.416729878307168e-06, "loss": 1.4542, "step": 24217 }, { "epoch": 1.897962382445141, "grad_norm": 2.9394655227661133, "learning_rate": 5.416422588349154e-06, "loss": 1.1565, "step": 24218 }, { "epoch": 1.8980407523510971, "grad_norm": 2.1105997562408447, "learning_rate": 5.416115296807279e-06, "loss": 1.4597, "step": 24219 }, { "epoch": 1.8981191222570533, "grad_norm": 3.40242862701416, "learning_rate": 5.415808003682717e-06, "loss": 1.0299, "step": 24220 }, { "epoch": 1.8981974921630094, "grad_norm": 2.367147207260132, "learning_rate": 5.4155007089766354e-06, "loss": 1.2469, "step": 24221 }, { "epoch": 1.8982758620689655, "grad_norm": 5.18325662612915, "learning_rate": 5.415193412690202e-06, "loss": 1.455, "step": 24222 }, { "epoch": 1.8983542319749216, "grad_norm": 2.613783359527588, "learning_rate": 5.414886114824585e-06, "loss": 1.2353, "step": 24223 }, { "epoch": 1.8984326018808777, "grad_norm": 5.758330345153809, "learning_rate": 5.414578815380955e-06, "loss": 1.2728, "step": 24224 }, { "epoch": 1.8985109717868338, "grad_norm": 2.791612148284912, "learning_rate": 5.414271514360479e-06, "loss": 1.4466, "step": 24225 }, { "epoch": 1.89858934169279, "grad_norm": 2.4540274143218994, "learning_rate": 5.413964211764329e-06, "loss": 1.3944, "step": 24226 }, { "epoch": 1.898667711598746, "grad_norm": 3.035753011703491, "learning_rate": 5.413656907593671e-06, "loss": 1.6544, "step": 24227 }, { "epoch": 1.8987460815047021, "grad_norm": 1.8207998275756836, "learning_rate": 5.413349601849674e-06, "loss": 1.2273, "step": 24228 }, { "epoch": 1.8988244514106583, "grad_norm": 2.5494818687438965, "learning_rate": 5.413042294533507e-06, "loss": 0.9061, "step": 24229 }, { "epoch": 1.8989028213166144, "grad_norm": 7.106064319610596, "learning_rate": 5.412734985646342e-06, "loss": 1.2917, "step": 24230 }, { "epoch": 1.8989811912225707, "grad_norm": 3.259378433227539, "learning_rate": 5.412427675189342e-06, "loss": 1.1016, "step": 24231 }, { "epoch": 1.8990595611285266, "grad_norm": 4.840518951416016, "learning_rate": 5.412120363163681e-06, "loss": 1.0514, "step": 24232 }, { "epoch": 1.899137931034483, "grad_norm": 2.9714555740356445, "learning_rate": 5.411813049570525e-06, "loss": 1.0512, "step": 24233 }, { "epoch": 1.8992163009404388, "grad_norm": 2.2618587017059326, "learning_rate": 5.4115057344110435e-06, "loss": 1.3996, "step": 24234 }, { "epoch": 1.8992946708463951, "grad_norm": 2.595341920852661, "learning_rate": 5.411198417686406e-06, "loss": 1.3143, "step": 24235 }, { "epoch": 1.899373040752351, "grad_norm": 3.748995304107666, "learning_rate": 5.41089109939778e-06, "loss": 1.2461, "step": 24236 }, { "epoch": 1.8994514106583074, "grad_norm": 3.053713083267212, "learning_rate": 5.410583779546337e-06, "loss": 1.9434, "step": 24237 }, { "epoch": 1.8995297805642632, "grad_norm": 3.4122793674468994, "learning_rate": 5.410276458133242e-06, "loss": 1.4259, "step": 24238 }, { "epoch": 1.8996081504702196, "grad_norm": 2.9973747730255127, "learning_rate": 5.4099691351596685e-06, "loss": 1.2822, "step": 24239 }, { "epoch": 1.8996865203761755, "grad_norm": 4.513333797454834, "learning_rate": 5.409661810626781e-06, "loss": 1.7007, "step": 24240 }, { "epoch": 1.8997648902821318, "grad_norm": 3.8227450847625732, "learning_rate": 5.409354484535752e-06, "loss": 1.1207, "step": 24241 }, { "epoch": 1.8998432601880877, "grad_norm": 4.132935047149658, "learning_rate": 5.409047156887747e-06, "loss": 1.2256, "step": 24242 }, { "epoch": 1.899921630094044, "grad_norm": 4.228967189788818, "learning_rate": 5.408739827683938e-06, "loss": 1.9108, "step": 24243 }, { "epoch": 1.9, "grad_norm": 4.434548377990723, "learning_rate": 5.408432496925492e-06, "loss": 1.3056, "step": 24244 }, { "epoch": 1.9000783699059562, "grad_norm": 2.5685677528381348, "learning_rate": 5.408125164613578e-06, "loss": 1.3711, "step": 24245 }, { "epoch": 1.9001567398119121, "grad_norm": 4.917937755584717, "learning_rate": 5.407817830749368e-06, "loss": 1.4425, "step": 24246 }, { "epoch": 1.9002351097178685, "grad_norm": 2.1664023399353027, "learning_rate": 5.407510495334025e-06, "loss": 1.3057, "step": 24247 }, { "epoch": 1.9003134796238244, "grad_norm": 5.696625232696533, "learning_rate": 5.407203158368725e-06, "loss": 1.6333, "step": 24248 }, { "epoch": 1.9003918495297807, "grad_norm": 2.680241823196411, "learning_rate": 5.406895819854631e-06, "loss": 1.4462, "step": 24249 }, { "epoch": 1.9004702194357366, "grad_norm": 3.1347477436065674, "learning_rate": 5.406588479792915e-06, "loss": 1.7493, "step": 24250 }, { "epoch": 1.900548589341693, "grad_norm": 2.6369190216064453, "learning_rate": 5.406281138184745e-06, "loss": 1.3911, "step": 24251 }, { "epoch": 1.9006269592476488, "grad_norm": 3.0532779693603516, "learning_rate": 5.40597379503129e-06, "loss": 1.3926, "step": 24252 }, { "epoch": 1.9007053291536051, "grad_norm": 3.2772819995880127, "learning_rate": 5.405666450333719e-06, "loss": 1.8012, "step": 24253 }, { "epoch": 1.900783699059561, "grad_norm": 5.645914554595947, "learning_rate": 5.405359104093202e-06, "loss": 1.4098, "step": 24254 }, { "epoch": 1.9008620689655173, "grad_norm": 3.6344738006591797, "learning_rate": 5.405051756310906e-06, "loss": 1.8378, "step": 24255 }, { "epoch": 1.9009404388714732, "grad_norm": 3.147400379180908, "learning_rate": 5.4047444069880026e-06, "loss": 1.0243, "step": 24256 }, { "epoch": 1.9010188087774296, "grad_norm": 3.5741326808929443, "learning_rate": 5.404437056125657e-06, "loss": 1.6966, "step": 24257 }, { "epoch": 1.9010971786833855, "grad_norm": 3.380504846572876, "learning_rate": 5.404129703725044e-06, "loss": 1.7851, "step": 24258 }, { "epoch": 1.9011755485893418, "grad_norm": 2.8194591999053955, "learning_rate": 5.403822349787328e-06, "loss": 1.4932, "step": 24259 }, { "epoch": 1.9012539184952977, "grad_norm": 2.886939525604248, "learning_rate": 5.403514994313679e-06, "loss": 1.3684, "step": 24260 }, { "epoch": 1.901332288401254, "grad_norm": 2.390378952026367, "learning_rate": 5.4032076373052654e-06, "loss": 1.2994, "step": 24261 }, { "epoch": 1.90141065830721, "grad_norm": 3.149198532104492, "learning_rate": 5.402900278763258e-06, "loss": 1.5629, "step": 24262 }, { "epoch": 1.9014890282131662, "grad_norm": 2.951448440551758, "learning_rate": 5.402592918688825e-06, "loss": 1.4904, "step": 24263 }, { "epoch": 1.9015673981191221, "grad_norm": 5.857303619384766, "learning_rate": 5.402285557083134e-06, "loss": 1.4312, "step": 24264 }, { "epoch": 1.9016457680250785, "grad_norm": 2.4974617958068848, "learning_rate": 5.401978193947357e-06, "loss": 1.1245, "step": 24265 }, { "epoch": 1.9017241379310343, "grad_norm": 2.983858585357666, "learning_rate": 5.401670829282659e-06, "loss": 1.2889, "step": 24266 }, { "epoch": 1.9018025078369907, "grad_norm": 2.729550838470459, "learning_rate": 5.401363463090215e-06, "loss": 1.4054, "step": 24267 }, { "epoch": 1.9018808777429466, "grad_norm": 1.966504693031311, "learning_rate": 5.401056095371188e-06, "loss": 1.5811, "step": 24268 }, { "epoch": 1.901959247648903, "grad_norm": 2.8085546493530273, "learning_rate": 5.40074872612675e-06, "loss": 1.1995, "step": 24269 }, { "epoch": 1.9020376175548588, "grad_norm": 5.302279949188232, "learning_rate": 5.400441355358069e-06, "loss": 0.6907, "step": 24270 }, { "epoch": 1.9021159874608151, "grad_norm": 4.005594253540039, "learning_rate": 5.400133983066315e-06, "loss": 1.0671, "step": 24271 }, { "epoch": 1.902194357366771, "grad_norm": 3.046865224838257, "learning_rate": 5.399826609252659e-06, "loss": 1.6365, "step": 24272 }, { "epoch": 1.9022727272727273, "grad_norm": 2.0559988021850586, "learning_rate": 5.399519233918265e-06, "loss": 1.5199, "step": 24273 }, { "epoch": 1.9023510971786832, "grad_norm": 3.51278018951416, "learning_rate": 5.3992118570643085e-06, "loss": 1.5211, "step": 24274 }, { "epoch": 1.9024294670846396, "grad_norm": 2.6277897357940674, "learning_rate": 5.3989044786919535e-06, "loss": 1.4188, "step": 24275 }, { "epoch": 1.9025078369905954, "grad_norm": 3.5917534828186035, "learning_rate": 5.3985970988023695e-06, "loss": 1.3299, "step": 24276 }, { "epoch": 1.9025862068965518, "grad_norm": 3.1919193267822266, "learning_rate": 5.398289717396728e-06, "loss": 1.7745, "step": 24277 }, { "epoch": 1.902664576802508, "grad_norm": 2.7763078212738037, "learning_rate": 5.397982334476196e-06, "loss": 1.2948, "step": 24278 }, { "epoch": 1.902742946708464, "grad_norm": 3.7436609268188477, "learning_rate": 5.397674950041944e-06, "loss": 1.0066, "step": 24279 }, { "epoch": 1.90282131661442, "grad_norm": 5.218647003173828, "learning_rate": 5.397367564095142e-06, "loss": 1.7464, "step": 24280 }, { "epoch": 1.9028996865203762, "grad_norm": 2.6238303184509277, "learning_rate": 5.397060176636957e-06, "loss": 1.3691, "step": 24281 }, { "epoch": 1.9029780564263323, "grad_norm": 1.8130699396133423, "learning_rate": 5.39675278766856e-06, "loss": 1.28, "step": 24282 }, { "epoch": 1.9030564263322884, "grad_norm": 2.5915706157684326, "learning_rate": 5.396445397191117e-06, "loss": 1.4369, "step": 24283 }, { "epoch": 1.9031347962382446, "grad_norm": 2.6369502544403076, "learning_rate": 5.396138005205801e-06, "loss": 1.4547, "step": 24284 }, { "epoch": 1.9032131661442007, "grad_norm": 3.0136454105377197, "learning_rate": 5.395830611713778e-06, "loss": 1.0097, "step": 24285 }, { "epoch": 1.9032915360501568, "grad_norm": 2.1709659099578857, "learning_rate": 5.395523216716221e-06, "loss": 1.2315, "step": 24286 }, { "epoch": 1.9033699059561129, "grad_norm": 3.6644601821899414, "learning_rate": 5.395215820214295e-06, "loss": 1.68, "step": 24287 }, { "epoch": 1.903448275862069, "grad_norm": 3.5145621299743652, "learning_rate": 5.394908422209171e-06, "loss": 1.5943, "step": 24288 }, { "epoch": 1.903526645768025, "grad_norm": 6.9187188148498535, "learning_rate": 5.39460102270202e-06, "loss": 1.1034, "step": 24289 }, { "epoch": 1.9036050156739812, "grad_norm": 4.289222717285156, "learning_rate": 5.394293621694007e-06, "loss": 2.0735, "step": 24290 }, { "epoch": 1.9036833855799373, "grad_norm": 3.026393175125122, "learning_rate": 5.393986219186305e-06, "loss": 1.4, "step": 24291 }, { "epoch": 1.9037617554858934, "grad_norm": 3.603207588195801, "learning_rate": 5.393678815180081e-06, "loss": 1.8521, "step": 24292 }, { "epoch": 1.9038401253918495, "grad_norm": 3.331864595413208, "learning_rate": 5.393371409676506e-06, "loss": 1.2005, "step": 24293 }, { "epoch": 1.9039184952978057, "grad_norm": 3.0495171546936035, "learning_rate": 5.393064002676748e-06, "loss": 1.3531, "step": 24294 }, { "epoch": 1.9039968652037618, "grad_norm": 5.234751224517822, "learning_rate": 5.392756594181976e-06, "loss": 1.5085, "step": 24295 }, { "epoch": 1.9040752351097179, "grad_norm": 3.9583020210266113, "learning_rate": 5.39244918419336e-06, "loss": 1.6676, "step": 24296 }, { "epoch": 1.904153605015674, "grad_norm": 3.68349552154541, "learning_rate": 5.392141772712068e-06, "loss": 1.7268, "step": 24297 }, { "epoch": 1.90423197492163, "grad_norm": 4.040513515472412, "learning_rate": 5.3918343597392705e-06, "loss": 1.2992, "step": 24298 }, { "epoch": 1.9043103448275862, "grad_norm": 2.368783712387085, "learning_rate": 5.391526945276138e-06, "loss": 1.3477, "step": 24299 }, { "epoch": 1.9043887147335423, "grad_norm": 3.7896032333374023, "learning_rate": 5.391219529323835e-06, "loss": 1.149, "step": 24300 }, { "epoch": 1.9044670846394984, "grad_norm": 2.4592607021331787, "learning_rate": 5.390912111883536e-06, "loss": 1.9134, "step": 24301 }, { "epoch": 1.9045454545454545, "grad_norm": 2.307274580001831, "learning_rate": 5.39060469295641e-06, "loss": 1.1811, "step": 24302 }, { "epoch": 1.9046238244514107, "grad_norm": 3.3704490661621094, "learning_rate": 5.390297272543621e-06, "loss": 1.4619, "step": 24303 }, { "epoch": 1.9047021943573668, "grad_norm": 2.8367679119110107, "learning_rate": 5.389989850646342e-06, "loss": 1.0325, "step": 24304 }, { "epoch": 1.9047805642633229, "grad_norm": 3.4743478298187256, "learning_rate": 5.389682427265743e-06, "loss": 1.165, "step": 24305 }, { "epoch": 1.904858934169279, "grad_norm": 3.1445086002349854, "learning_rate": 5.389375002402992e-06, "loss": 2.0007, "step": 24306 }, { "epoch": 1.904937304075235, "grad_norm": 4.303769588470459, "learning_rate": 5.389067576059258e-06, "loss": 1.2103, "step": 24307 }, { "epoch": 1.9050156739811912, "grad_norm": 3.591580629348755, "learning_rate": 5.388760148235711e-06, "loss": 1.1772, "step": 24308 }, { "epoch": 1.9050940438871473, "grad_norm": 2.1742215156555176, "learning_rate": 5.38845271893352e-06, "loss": 0.9622, "step": 24309 }, { "epoch": 1.9051724137931034, "grad_norm": 3.850329875946045, "learning_rate": 5.388145288153855e-06, "loss": 1.613, "step": 24310 }, { "epoch": 1.9052507836990595, "grad_norm": 3.5239131450653076, "learning_rate": 5.387837855897884e-06, "loss": 1.4551, "step": 24311 }, { "epoch": 1.9053291536050156, "grad_norm": 2.820180892944336, "learning_rate": 5.3875304221667775e-06, "loss": 1.4092, "step": 24312 }, { "epoch": 1.9054075235109718, "grad_norm": 2.695873498916626, "learning_rate": 5.3872229869617034e-06, "loss": 1.2258, "step": 24313 }, { "epoch": 1.9054858934169279, "grad_norm": 2.388824701309204, "learning_rate": 5.386915550283833e-06, "loss": 1.2512, "step": 24314 }, { "epoch": 1.905564263322884, "grad_norm": 2.0146563053131104, "learning_rate": 5.386608112134335e-06, "loss": 1.4829, "step": 24315 }, { "epoch": 1.90564263322884, "grad_norm": 3.9361989498138428, "learning_rate": 5.386300672514377e-06, "loss": 1.4558, "step": 24316 }, { "epoch": 1.9057210031347962, "grad_norm": 2.4944772720336914, "learning_rate": 5.38599323142513e-06, "loss": 1.4149, "step": 24317 }, { "epoch": 1.9057993730407523, "grad_norm": 2.327937126159668, "learning_rate": 5.385685788867763e-06, "loss": 1.4227, "step": 24318 }, { "epoch": 1.9058777429467084, "grad_norm": 2.887342691421509, "learning_rate": 5.385378344843447e-06, "loss": 1.5434, "step": 24319 }, { "epoch": 1.9059561128526645, "grad_norm": 4.54204797744751, "learning_rate": 5.3850708993533465e-06, "loss": 1.7263, "step": 24320 }, { "epoch": 1.9060344827586206, "grad_norm": 3.6967413425445557, "learning_rate": 5.384763452398637e-06, "loss": 1.2049, "step": 24321 }, { "epoch": 1.9061128526645768, "grad_norm": 4.084228515625, "learning_rate": 5.3844560039804834e-06, "loss": 1.4539, "step": 24322 }, { "epoch": 1.9061912225705329, "grad_norm": 5.328677177429199, "learning_rate": 5.384148554100057e-06, "loss": 0.9075, "step": 24323 }, { "epoch": 1.9062695924764892, "grad_norm": 2.3644888401031494, "learning_rate": 5.383841102758527e-06, "loss": 1.5954, "step": 24324 }, { "epoch": 1.906347962382445, "grad_norm": 2.365718364715576, "learning_rate": 5.383533649957062e-06, "loss": 1.4999, "step": 24325 }, { "epoch": 1.9064263322884014, "grad_norm": 4.301462650299072, "learning_rate": 5.383226195696833e-06, "loss": 1.7082, "step": 24326 }, { "epoch": 1.9065047021943573, "grad_norm": 2.2976651191711426, "learning_rate": 5.3829187399790075e-06, "loss": 1.0965, "step": 24327 }, { "epoch": 1.9065830721003136, "grad_norm": 3.1218924522399902, "learning_rate": 5.382611282804756e-06, "loss": 1.3039, "step": 24328 }, { "epoch": 1.9066614420062695, "grad_norm": 10.789459228515625, "learning_rate": 5.382303824175249e-06, "loss": 1.357, "step": 24329 }, { "epoch": 1.9067398119122259, "grad_norm": 5.7106242179870605, "learning_rate": 5.381996364091653e-06, "loss": 0.9579, "step": 24330 }, { "epoch": 1.9068181818181817, "grad_norm": 2.7382731437683105, "learning_rate": 5.381688902555139e-06, "loss": 1.2703, "step": 24331 }, { "epoch": 1.906896551724138, "grad_norm": 2.082728624343872, "learning_rate": 5.381381439566878e-06, "loss": 1.2093, "step": 24332 }, { "epoch": 1.906974921630094, "grad_norm": 5.865113258361816, "learning_rate": 5.381073975128036e-06, "loss": 1.697, "step": 24333 }, { "epoch": 1.9070532915360503, "grad_norm": 2.6050126552581787, "learning_rate": 5.3807665092397855e-06, "loss": 1.3342, "step": 24334 }, { "epoch": 1.9071316614420062, "grad_norm": 4.431262969970703, "learning_rate": 5.380459041903293e-06, "loss": 1.6077, "step": 24335 }, { "epoch": 1.9072100313479625, "grad_norm": 5.028085231781006, "learning_rate": 5.3801515731197325e-06, "loss": 1.5292, "step": 24336 }, { "epoch": 1.9072884012539184, "grad_norm": 1.8852323293685913, "learning_rate": 5.379844102890268e-06, "loss": 1.2462, "step": 24337 }, { "epoch": 1.9073667711598747, "grad_norm": 2.909528970718384, "learning_rate": 5.379536631216074e-06, "loss": 1.8118, "step": 24338 }, { "epoch": 1.9074451410658306, "grad_norm": 8.559548377990723, "learning_rate": 5.379229158098316e-06, "loss": 2.4502, "step": 24339 }, { "epoch": 1.907523510971787, "grad_norm": 3.6340317726135254, "learning_rate": 5.378921683538166e-06, "loss": 1.8409, "step": 24340 }, { "epoch": 1.9076018808777429, "grad_norm": 5.8750901222229, "learning_rate": 5.378614207536792e-06, "loss": 1.1908, "step": 24341 }, { "epoch": 1.9076802507836992, "grad_norm": 3.269658327102661, "learning_rate": 5.378306730095364e-06, "loss": 1.4428, "step": 24342 }, { "epoch": 1.907758620689655, "grad_norm": 3.1770641803741455, "learning_rate": 5.377999251215052e-06, "loss": 1.676, "step": 24343 }, { "epoch": 1.9078369905956114, "grad_norm": 5.832804203033447, "learning_rate": 5.377691770897025e-06, "loss": 1.3912, "step": 24344 }, { "epoch": 1.9079153605015673, "grad_norm": 2.052218437194824, "learning_rate": 5.377384289142454e-06, "loss": 1.1723, "step": 24345 }, { "epoch": 1.9079937304075236, "grad_norm": 16.382524490356445, "learning_rate": 5.377076805952502e-06, "loss": 1.7141, "step": 24346 }, { "epoch": 1.9080721003134795, "grad_norm": 4.325671195983887, "learning_rate": 5.376769321328349e-06, "loss": 1.5903, "step": 24347 }, { "epoch": 1.9081504702194358, "grad_norm": 2.808431386947632, "learning_rate": 5.376461835271156e-06, "loss": 1.8792, "step": 24348 }, { "epoch": 1.9082288401253917, "grad_norm": 2.310772180557251, "learning_rate": 5.376154347782097e-06, "loss": 1.4508, "step": 24349 }, { "epoch": 1.908307210031348, "grad_norm": 3.731424570083618, "learning_rate": 5.375846858862338e-06, "loss": 1.6717, "step": 24350 }, { "epoch": 1.908385579937304, "grad_norm": 6.551516532897949, "learning_rate": 5.375539368513052e-06, "loss": 1.1651, "step": 24351 }, { "epoch": 1.9084639498432603, "grad_norm": 2.038712978363037, "learning_rate": 5.375231876735407e-06, "loss": 1.1085, "step": 24352 }, { "epoch": 1.9085423197492162, "grad_norm": 4.6737060546875, "learning_rate": 5.374924383530573e-06, "loss": 0.5255, "step": 24353 }, { "epoch": 1.9086206896551725, "grad_norm": 8.70528793334961, "learning_rate": 5.374616888899719e-06, "loss": 0.8259, "step": 24354 }, { "epoch": 1.9086990595611284, "grad_norm": 2.8304169178009033, "learning_rate": 5.374309392844014e-06, "loss": 1.5667, "step": 24355 }, { "epoch": 1.9087774294670847, "grad_norm": 2.425954818725586, "learning_rate": 5.3740018953646285e-06, "loss": 1.5903, "step": 24356 }, { "epoch": 1.9088557993730406, "grad_norm": 2.3399131298065186, "learning_rate": 5.373694396462734e-06, "loss": 1.1913, "step": 24357 }, { "epoch": 1.908934169278997, "grad_norm": 2.771052598953247, "learning_rate": 5.373386896139496e-06, "loss": 1.4086, "step": 24358 }, { "epoch": 1.9090125391849528, "grad_norm": 2.2372899055480957, "learning_rate": 5.373079394396087e-06, "loss": 1.461, "step": 24359 }, { "epoch": 1.9090909090909092, "grad_norm": 2.9125993251800537, "learning_rate": 5.372771891233675e-06, "loss": 1.0376, "step": 24360 }, { "epoch": 1.909169278996865, "grad_norm": 2.3077573776245117, "learning_rate": 5.37246438665343e-06, "loss": 1.2893, "step": 24361 }, { "epoch": 1.9092476489028214, "grad_norm": 2.350140333175659, "learning_rate": 5.372156880656522e-06, "loss": 1.3107, "step": 24362 }, { "epoch": 1.9093260188087773, "grad_norm": 7.172300338745117, "learning_rate": 5.37184937324412e-06, "loss": 0.8502, "step": 24363 }, { "epoch": 1.9094043887147336, "grad_norm": 2.814326524734497, "learning_rate": 5.371541864417395e-06, "loss": 1.1822, "step": 24364 }, { "epoch": 1.9094827586206895, "grad_norm": 2.9985780715942383, "learning_rate": 5.371234354177514e-06, "loss": 1.4202, "step": 24365 }, { "epoch": 1.9095611285266458, "grad_norm": 4.611415863037109, "learning_rate": 5.37092684252565e-06, "loss": 1.1534, "step": 24366 }, { "epoch": 1.9096394984326017, "grad_norm": 4.333442687988281, "learning_rate": 5.37061932946297e-06, "loss": 1.115, "step": 24367 }, { "epoch": 1.909717868338558, "grad_norm": 5.089881896972656, "learning_rate": 5.370311814990644e-06, "loss": 1.2511, "step": 24368 }, { "epoch": 1.909796238244514, "grad_norm": 2.3439555168151855, "learning_rate": 5.370004299109842e-06, "loss": 1.1543, "step": 24369 }, { "epoch": 1.9098746081504703, "grad_norm": 4.286592960357666, "learning_rate": 5.369696781821735e-06, "loss": 1.4114, "step": 24370 }, { "epoch": 1.9099529780564264, "grad_norm": 3.2577295303344727, "learning_rate": 5.3693892631274915e-06, "loss": 1.3309, "step": 24371 }, { "epoch": 1.9100313479623825, "grad_norm": 2.6186537742614746, "learning_rate": 5.3690817430282785e-06, "loss": 1.2368, "step": 24372 }, { "epoch": 1.9101097178683386, "grad_norm": 2.4407472610473633, "learning_rate": 5.36877422152527e-06, "loss": 0.8399, "step": 24373 }, { "epoch": 1.9101880877742947, "grad_norm": 2.8152568340301514, "learning_rate": 5.368466698619632e-06, "loss": 1.6709, "step": 24374 }, { "epoch": 1.9102664576802508, "grad_norm": 2.4478819370269775, "learning_rate": 5.368159174312538e-06, "loss": 1.3303, "step": 24375 }, { "epoch": 1.910344827586207, "grad_norm": 3.298264741897583, "learning_rate": 5.367851648605156e-06, "loss": 1.5247, "step": 24376 }, { "epoch": 1.910423197492163, "grad_norm": 3.23098087310791, "learning_rate": 5.3675441214986535e-06, "loss": 1.089, "step": 24377 }, { "epoch": 1.9105015673981192, "grad_norm": 2.6336634159088135, "learning_rate": 5.367236592994203e-06, "loss": 1.4515, "step": 24378 }, { "epoch": 1.9105799373040753, "grad_norm": 12.740741729736328, "learning_rate": 5.366929063092972e-06, "loss": 1.8872, "step": 24379 }, { "epoch": 1.9106583072100314, "grad_norm": 2.545457363128662, "learning_rate": 5.366621531796132e-06, "loss": 1.6369, "step": 24380 }, { "epoch": 1.9107366771159875, "grad_norm": 2.359672784805298, "learning_rate": 5.366313999104852e-06, "loss": 1.5316, "step": 24381 }, { "epoch": 1.9108150470219436, "grad_norm": 2.951737403869629, "learning_rate": 5.366006465020302e-06, "loss": 1.2015, "step": 24382 }, { "epoch": 1.9108934169278997, "grad_norm": 2.6815199851989746, "learning_rate": 5.365698929543652e-06, "loss": 1.1987, "step": 24383 }, { "epoch": 1.9109717868338558, "grad_norm": 1.9985103607177734, "learning_rate": 5.365391392676069e-06, "loss": 0.8204, "step": 24384 }, { "epoch": 1.911050156739812, "grad_norm": 3.0626935958862305, "learning_rate": 5.365083854418727e-06, "loss": 1.3603, "step": 24385 }, { "epoch": 1.911128526645768, "grad_norm": 2.121211051940918, "learning_rate": 5.364776314772792e-06, "loss": 1.2647, "step": 24386 }, { "epoch": 1.9112068965517242, "grad_norm": 3.8814187049865723, "learning_rate": 5.364468773739436e-06, "loss": 0.9086, "step": 24387 }, { "epoch": 1.9112852664576803, "grad_norm": 4.640308380126953, "learning_rate": 5.364161231319828e-06, "loss": 1.6179, "step": 24388 }, { "epoch": 1.9113636363636364, "grad_norm": 2.4815292358398438, "learning_rate": 5.363853687515137e-06, "loss": 1.3006, "step": 24389 }, { "epoch": 1.9114420062695925, "grad_norm": 2.9044442176818848, "learning_rate": 5.363546142326536e-06, "loss": 1.2996, "step": 24390 }, { "epoch": 1.9115203761755486, "grad_norm": 2.5141680240631104, "learning_rate": 5.363238595755188e-06, "loss": 1.3157, "step": 24391 }, { "epoch": 1.9115987460815047, "grad_norm": 2.811309576034546, "learning_rate": 5.36293104780227e-06, "loss": 1.3499, "step": 24392 }, { "epoch": 1.9116771159874608, "grad_norm": 4.863748073577881, "learning_rate": 5.3626234984689475e-06, "loss": 1.2429, "step": 24393 }, { "epoch": 1.911755485893417, "grad_norm": 1.8797900676727295, "learning_rate": 5.362315947756393e-06, "loss": 1.2806, "step": 24394 }, { "epoch": 1.911833855799373, "grad_norm": 2.447329044342041, "learning_rate": 5.3620083956657735e-06, "loss": 1.3015, "step": 24395 }, { "epoch": 1.9119122257053291, "grad_norm": 2.915060520172119, "learning_rate": 5.36170084219826e-06, "loss": 0.857, "step": 24396 }, { "epoch": 1.9119905956112853, "grad_norm": 2.1800055503845215, "learning_rate": 5.361393287355021e-06, "loss": 1.0559, "step": 24397 }, { "epoch": 1.9120689655172414, "grad_norm": 3.220637798309326, "learning_rate": 5.36108573113723e-06, "loss": 1.4626, "step": 24398 }, { "epoch": 1.9121473354231975, "grad_norm": 4.869452476501465, "learning_rate": 5.360778173546054e-06, "loss": 1.5392, "step": 24399 }, { "epoch": 1.9122257053291536, "grad_norm": 2.4452431201934814, "learning_rate": 5.360470614582661e-06, "loss": 1.2822, "step": 24400 }, { "epoch": 1.9123040752351097, "grad_norm": 2.249560832977295, "learning_rate": 5.3601630542482255e-06, "loss": 0.9879, "step": 24401 }, { "epoch": 1.9123824451410658, "grad_norm": 8.665628433227539, "learning_rate": 5.359855492543912e-06, "loss": 1.1404, "step": 24402 }, { "epoch": 1.912460815047022, "grad_norm": 2.9541702270507812, "learning_rate": 5.359547929470895e-06, "loss": 1.323, "step": 24403 }, { "epoch": 1.912539184952978, "grad_norm": 3.0492584705352783, "learning_rate": 5.3592403650303425e-06, "loss": 1.4593, "step": 24404 }, { "epoch": 1.9126175548589341, "grad_norm": 3.281481981277466, "learning_rate": 5.358932799223423e-06, "loss": 1.1671, "step": 24405 }, { "epoch": 1.9126959247648903, "grad_norm": 2.779737949371338, "learning_rate": 5.358625232051309e-06, "loss": 1.4329, "step": 24406 }, { "epoch": 1.9127742946708464, "grad_norm": 3.336304187774658, "learning_rate": 5.3583176635151665e-06, "loss": 1.4896, "step": 24407 }, { "epoch": 1.9128526645768025, "grad_norm": 3.621854305267334, "learning_rate": 5.358010093616169e-06, "loss": 1.5934, "step": 24408 }, { "epoch": 1.9129310344827586, "grad_norm": 3.8474931716918945, "learning_rate": 5.357702522355485e-06, "loss": 1.0975, "step": 24409 }, { "epoch": 1.9130094043887147, "grad_norm": 3.5669002532958984, "learning_rate": 5.357394949734285e-06, "loss": 1.4243, "step": 24410 }, { "epoch": 1.9130877742946708, "grad_norm": 2.944617509841919, "learning_rate": 5.357087375753739e-06, "loss": 1.1007, "step": 24411 }, { "epoch": 1.913166144200627, "grad_norm": 5.531489372253418, "learning_rate": 5.356779800415013e-06, "loss": 1.6128, "step": 24412 }, { "epoch": 1.913244514106583, "grad_norm": 2.7294323444366455, "learning_rate": 5.356472223719283e-06, "loss": 1.1631, "step": 24413 }, { "epoch": 1.9133228840125391, "grad_norm": 5.290228366851807, "learning_rate": 5.356164645667714e-06, "loss": 1.3533, "step": 24414 }, { "epoch": 1.9134012539184952, "grad_norm": 2.7211415767669678, "learning_rate": 5.355857066261477e-06, "loss": 1.2684, "step": 24415 }, { "epoch": 1.9134796238244514, "grad_norm": 2.886873245239258, "learning_rate": 5.355549485501743e-06, "loss": 1.3187, "step": 24416 }, { "epoch": 1.9135579937304075, "grad_norm": 4.5072126388549805, "learning_rate": 5.355241903389682e-06, "loss": 1.4111, "step": 24417 }, { "epoch": 1.9136363636363636, "grad_norm": 2.645878791809082, "learning_rate": 5.354934319926465e-06, "loss": 1.4865, "step": 24418 }, { "epoch": 1.91371473354232, "grad_norm": 2.5283381938934326, "learning_rate": 5.354626735113256e-06, "loss": 1.3612, "step": 24419 }, { "epoch": 1.9137931034482758, "grad_norm": 9.561456680297852, "learning_rate": 5.3543191489512325e-06, "loss": 1.7404, "step": 24420 }, { "epoch": 1.9138714733542321, "grad_norm": 5.02354621887207, "learning_rate": 5.354011561441559e-06, "loss": 1.6002, "step": 24421 }, { "epoch": 1.913949843260188, "grad_norm": 4.0104875564575195, "learning_rate": 5.353703972585409e-06, "loss": 1.1956, "step": 24422 }, { "epoch": 1.9140282131661444, "grad_norm": 3.2158737182617188, "learning_rate": 5.35339638238395e-06, "loss": 1.5251, "step": 24423 }, { "epoch": 1.9141065830721002, "grad_norm": 2.2306511402130127, "learning_rate": 5.353088790838352e-06, "loss": 1.3777, "step": 24424 }, { "epoch": 1.9141849529780566, "grad_norm": 3.1637399196624756, "learning_rate": 5.352781197949786e-06, "loss": 0.939, "step": 24425 }, { "epoch": 1.9142633228840125, "grad_norm": 3.1288044452667236, "learning_rate": 5.3524736037194216e-06, "loss": 1.2507, "step": 24426 }, { "epoch": 1.9143416927899688, "grad_norm": 3.5905611515045166, "learning_rate": 5.3521660081484295e-06, "loss": 1.2966, "step": 24427 }, { "epoch": 1.9144200626959247, "grad_norm": 3.03173565864563, "learning_rate": 5.351858411237977e-06, "loss": 1.5214, "step": 24428 }, { "epoch": 1.914498432601881, "grad_norm": 3.293771266937256, "learning_rate": 5.351550812989237e-06, "loss": 1.3813, "step": 24429 }, { "epoch": 1.914576802507837, "grad_norm": 2.4801814556121826, "learning_rate": 5.351243213403378e-06, "loss": 1.5864, "step": 24430 }, { "epoch": 1.9146551724137932, "grad_norm": 4.205289363861084, "learning_rate": 5.35093561248157e-06, "loss": 1.1993, "step": 24431 }, { "epoch": 1.9147335423197491, "grad_norm": 3.5253539085388184, "learning_rate": 5.350628010224983e-06, "loss": 1.4729, "step": 24432 }, { "epoch": 1.9148119122257055, "grad_norm": 3.1216094493865967, "learning_rate": 5.350320406634788e-06, "loss": 1.5138, "step": 24433 }, { "epoch": 1.9148902821316613, "grad_norm": 5.618022918701172, "learning_rate": 5.350012801712154e-06, "loss": 1.3, "step": 24434 }, { "epoch": 1.9149686520376177, "grad_norm": 6.1088337898254395, "learning_rate": 5.34970519545825e-06, "loss": 1.0431, "step": 24435 }, { "epoch": 1.9150470219435736, "grad_norm": 3.43920636177063, "learning_rate": 5.349397587874248e-06, "loss": 1.2956, "step": 24436 }, { "epoch": 1.91512539184953, "grad_norm": 5.564638614654541, "learning_rate": 5.349089978961318e-06, "loss": 1.6457, "step": 24437 }, { "epoch": 1.9152037617554858, "grad_norm": 2.706382989883423, "learning_rate": 5.348782368720627e-06, "loss": 1.2742, "step": 24438 }, { "epoch": 1.9152821316614421, "grad_norm": 1.7692561149597168, "learning_rate": 5.348474757153348e-06, "loss": 1.233, "step": 24439 }, { "epoch": 1.915360501567398, "grad_norm": 9.221280097961426, "learning_rate": 5.34816714426065e-06, "loss": 1.3524, "step": 24440 }, { "epoch": 1.9154388714733543, "grad_norm": 6.537792205810547, "learning_rate": 5.347859530043703e-06, "loss": 0.5535, "step": 24441 }, { "epoch": 1.9155172413793102, "grad_norm": 2.191192865371704, "learning_rate": 5.347551914503677e-06, "loss": 1.3811, "step": 24442 }, { "epoch": 1.9155956112852666, "grad_norm": 3.358104705810547, "learning_rate": 5.347244297641741e-06, "loss": 1.3924, "step": 24443 }, { "epoch": 1.9156739811912225, "grad_norm": 3.438775062561035, "learning_rate": 5.346936679459069e-06, "loss": 1.3239, "step": 24444 }, { "epoch": 1.9157523510971788, "grad_norm": 2.860163927078247, "learning_rate": 5.346629059956825e-06, "loss": 1.4285, "step": 24445 }, { "epoch": 1.9158307210031347, "grad_norm": 5.451385498046875, "learning_rate": 5.346321439136184e-06, "loss": 1.2567, "step": 24446 }, { "epoch": 1.915909090909091, "grad_norm": 3.306892156600952, "learning_rate": 5.3460138169983125e-06, "loss": 1.3885, "step": 24447 }, { "epoch": 1.915987460815047, "grad_norm": 3.8379194736480713, "learning_rate": 5.345706193544383e-06, "loss": 1.1054, "step": 24448 }, { "epoch": 1.9160658307210032, "grad_norm": 2.378230094909668, "learning_rate": 5.345398568775565e-06, "loss": 1.1493, "step": 24449 }, { "epoch": 1.9161442006269591, "grad_norm": 2.285261869430542, "learning_rate": 5.345090942693027e-06, "loss": 1.2506, "step": 24450 }, { "epoch": 1.9162225705329154, "grad_norm": 2.305964469909668, "learning_rate": 5.344783315297942e-06, "loss": 1.2295, "step": 24451 }, { "epoch": 1.9163009404388713, "grad_norm": 2.4312334060668945, "learning_rate": 5.344475686591478e-06, "loss": 1.123, "step": 24452 }, { "epoch": 1.9163793103448277, "grad_norm": 5.2614569664001465, "learning_rate": 5.344168056574805e-06, "loss": 1.8635, "step": 24453 }, { "epoch": 1.9164576802507836, "grad_norm": 22.45663070678711, "learning_rate": 5.3438604252490934e-06, "loss": 1.4225, "step": 24454 }, { "epoch": 1.91653605015674, "grad_norm": 1.8356324434280396, "learning_rate": 5.343552792615514e-06, "loss": 1.4127, "step": 24455 }, { "epoch": 1.9166144200626958, "grad_norm": 8.63986587524414, "learning_rate": 5.343245158675234e-06, "loss": 1.1186, "step": 24456 }, { "epoch": 1.916692789968652, "grad_norm": 3.411478281021118, "learning_rate": 5.342937523429429e-06, "loss": 1.0514, "step": 24457 }, { "epoch": 1.916771159874608, "grad_norm": 2.695963144302368, "learning_rate": 5.342629886879265e-06, "loss": 1.0892, "step": 24458 }, { "epoch": 1.9168495297805643, "grad_norm": 4.759791851043701, "learning_rate": 5.342322249025911e-06, "loss": 1.2213, "step": 24459 }, { "epoch": 1.9169278996865202, "grad_norm": 3.284658670425415, "learning_rate": 5.3420146098705404e-06, "loss": 1.5325, "step": 24460 }, { "epoch": 1.9170062695924766, "grad_norm": 2.6083314418792725, "learning_rate": 5.341706969414322e-06, "loss": 1.3597, "step": 24461 }, { "epoch": 1.9170846394984324, "grad_norm": 3.127041816711426, "learning_rate": 5.341399327658425e-06, "loss": 1.3651, "step": 24462 }, { "epoch": 1.9171630094043888, "grad_norm": 3.7744598388671875, "learning_rate": 5.34109168460402e-06, "loss": 1.2289, "step": 24463 }, { "epoch": 1.9172413793103447, "grad_norm": 4.212949275970459, "learning_rate": 5.340784040252278e-06, "loss": 1.2291, "step": 24464 }, { "epoch": 1.917319749216301, "grad_norm": 2.4181597232818604, "learning_rate": 5.34047639460437e-06, "loss": 1.4502, "step": 24465 }, { "epoch": 1.917398119122257, "grad_norm": 4.365541934967041, "learning_rate": 5.340168747661463e-06, "loss": 2.1747, "step": 24466 }, { "epoch": 1.9174764890282132, "grad_norm": 2.308530569076538, "learning_rate": 5.339861099424729e-06, "loss": 1.3686, "step": 24467 }, { "epoch": 1.9175548589341693, "grad_norm": 3.8302955627441406, "learning_rate": 5.339553449895337e-06, "loss": 1.6907, "step": 24468 }, { "epoch": 1.9176332288401254, "grad_norm": 3.1451199054718018, "learning_rate": 5.33924579907446e-06, "loss": 1.2569, "step": 24469 }, { "epoch": 1.9177115987460815, "grad_norm": 3.144824504852295, "learning_rate": 5.338938146963266e-06, "loss": 1.545, "step": 24470 }, { "epoch": 1.9177899686520377, "grad_norm": 2.3964385986328125, "learning_rate": 5.338630493562925e-06, "loss": 1.3238, "step": 24471 }, { "epoch": 1.9178683385579938, "grad_norm": 8.923807144165039, "learning_rate": 5.3383228388746075e-06, "loss": 1.5248, "step": 24472 }, { "epoch": 1.9179467084639499, "grad_norm": 3.4923524856567383, "learning_rate": 5.338015182899483e-06, "loss": 1.3957, "step": 24473 }, { "epoch": 1.918025078369906, "grad_norm": 4.840236663818359, "learning_rate": 5.337707525638724e-06, "loss": 1.4199, "step": 24474 }, { "epoch": 1.918103448275862, "grad_norm": 3.0421745777130127, "learning_rate": 5.337399867093497e-06, "loss": 0.6928, "step": 24475 }, { "epoch": 1.9181818181818182, "grad_norm": 2.2662336826324463, "learning_rate": 5.337092207264976e-06, "loss": 1.4579, "step": 24476 }, { "epoch": 1.9182601880877743, "grad_norm": 3.4489598274230957, "learning_rate": 5.336784546154328e-06, "loss": 0.7116, "step": 24477 }, { "epoch": 1.9183385579937304, "grad_norm": 3.9932703971862793, "learning_rate": 5.336476883762726e-06, "loss": 1.6163, "step": 24478 }, { "epoch": 1.9184169278996865, "grad_norm": 2.7473127841949463, "learning_rate": 5.336169220091338e-06, "loss": 1.3399, "step": 24479 }, { "epoch": 1.9184952978056427, "grad_norm": 2.8958823680877686, "learning_rate": 5.335861555141334e-06, "loss": 1.3153, "step": 24480 }, { "epoch": 1.9185736677115988, "grad_norm": 3.0772335529327393, "learning_rate": 5.335553888913886e-06, "loss": 1.4837, "step": 24481 }, { "epoch": 1.9186520376175549, "grad_norm": 2.021346092224121, "learning_rate": 5.335246221410163e-06, "loss": 1.2348, "step": 24482 }, { "epoch": 1.918730407523511, "grad_norm": 2.2077090740203857, "learning_rate": 5.3349385526313365e-06, "loss": 1.4432, "step": 24483 }, { "epoch": 1.918808777429467, "grad_norm": 4.672721862792969, "learning_rate": 5.334630882578575e-06, "loss": 1.3803, "step": 24484 }, { "epoch": 1.9188871473354232, "grad_norm": 2.8544466495513916, "learning_rate": 5.33432321125305e-06, "loss": 1.5676, "step": 24485 }, { "epoch": 1.9189655172413793, "grad_norm": 8.071662902832031, "learning_rate": 5.334015538655931e-06, "loss": 1.3582, "step": 24486 }, { "epoch": 1.9190438871473354, "grad_norm": 1.9163496494293213, "learning_rate": 5.333707864788388e-06, "loss": 1.6035, "step": 24487 }, { "epoch": 1.9191222570532915, "grad_norm": 2.969949960708618, "learning_rate": 5.333400189651592e-06, "loss": 1.5093, "step": 24488 }, { "epoch": 1.9192006269592476, "grad_norm": 3.432894706726074, "learning_rate": 5.333092513246713e-06, "loss": 1.2501, "step": 24489 }, { "epoch": 1.9192789968652038, "grad_norm": 4.998518943786621, "learning_rate": 5.33278483557492e-06, "loss": 1.2257, "step": 24490 }, { "epoch": 1.9193573667711599, "grad_norm": 3.6653428077697754, "learning_rate": 5.332477156637387e-06, "loss": 1.3664, "step": 24491 }, { "epoch": 1.919435736677116, "grad_norm": 3.2451400756835938, "learning_rate": 5.332169476435279e-06, "loss": 1.4375, "step": 24492 }, { "epoch": 1.919514106583072, "grad_norm": 2.5052411556243896, "learning_rate": 5.331861794969771e-06, "loss": 1.6803, "step": 24493 }, { "epoch": 1.9195924764890282, "grad_norm": 4.44008207321167, "learning_rate": 5.33155411224203e-06, "loss": 1.5737, "step": 24494 }, { "epoch": 1.9196708463949843, "grad_norm": 2.0853054523468018, "learning_rate": 5.331246428253228e-06, "loss": 1.3269, "step": 24495 }, { "epoch": 1.9197492163009404, "grad_norm": 3.9087483882904053, "learning_rate": 5.330938743004535e-06, "loss": 1.1837, "step": 24496 }, { "epoch": 1.9198275862068965, "grad_norm": 2.7167322635650635, "learning_rate": 5.33063105649712e-06, "loss": 1.3416, "step": 24497 }, { "epoch": 1.9199059561128526, "grad_norm": 2.73909330368042, "learning_rate": 5.330323368732155e-06, "loss": 1.4389, "step": 24498 }, { "epoch": 1.9199843260188088, "grad_norm": 4.834291934967041, "learning_rate": 5.330015679710809e-06, "loss": 0.9021, "step": 24499 }, { "epoch": 1.9200626959247649, "grad_norm": 3.8794894218444824, "learning_rate": 5.329707989434254e-06, "loss": 1.3508, "step": 24500 }, { "epoch": 1.920141065830721, "grad_norm": 2.5631654262542725, "learning_rate": 5.329400297903658e-06, "loss": 1.3302, "step": 24501 }, { "epoch": 1.920219435736677, "grad_norm": 2.654005765914917, "learning_rate": 5.329092605120193e-06, "loss": 0.953, "step": 24502 }, { "epoch": 1.9202978056426332, "grad_norm": 3.4151456356048584, "learning_rate": 5.328784911085029e-06, "loss": 1.2907, "step": 24503 }, { "epoch": 1.9203761755485893, "grad_norm": 2.2772698402404785, "learning_rate": 5.328477215799335e-06, "loss": 1.5024, "step": 24504 }, { "epoch": 1.9204545454545454, "grad_norm": 2.0590922832489014, "learning_rate": 5.328169519264284e-06, "loss": 1.4682, "step": 24505 }, { "epoch": 1.9205329153605015, "grad_norm": 3.4653356075286865, "learning_rate": 5.327861821481043e-06, "loss": 1.1646, "step": 24506 }, { "epoch": 1.9206112852664576, "grad_norm": 3.8493902683258057, "learning_rate": 5.327554122450784e-06, "loss": 1.301, "step": 24507 }, { "epoch": 1.9206896551724137, "grad_norm": 2.586369514465332, "learning_rate": 5.327246422174679e-06, "loss": 1.3057, "step": 24508 }, { "epoch": 1.9207680250783699, "grad_norm": 3.040827512741089, "learning_rate": 5.3269387206538955e-06, "loss": 1.2241, "step": 24509 }, { "epoch": 1.920846394984326, "grad_norm": 3.9653210639953613, "learning_rate": 5.326631017889605e-06, "loss": 1.335, "step": 24510 }, { "epoch": 1.920924764890282, "grad_norm": 3.426414728164673, "learning_rate": 5.32632331388298e-06, "loss": 1.0915, "step": 24511 }, { "epoch": 1.9210031347962384, "grad_norm": 2.6441798210144043, "learning_rate": 5.326015608635187e-06, "loss": 1.2613, "step": 24512 }, { "epoch": 1.9210815047021943, "grad_norm": 3.2534475326538086, "learning_rate": 5.325707902147397e-06, "loss": 1.2233, "step": 24513 }, { "epoch": 1.9211598746081506, "grad_norm": 3.776146411895752, "learning_rate": 5.325400194420783e-06, "loss": 1.4102, "step": 24514 }, { "epoch": 1.9212382445141065, "grad_norm": 3.3214547634124756, "learning_rate": 5.325092485456513e-06, "loss": 1.6101, "step": 24515 }, { "epoch": 1.9213166144200629, "grad_norm": 2.348557233810425, "learning_rate": 5.324784775255758e-06, "loss": 1.1445, "step": 24516 }, { "epoch": 1.9213949843260187, "grad_norm": 1.8994477987289429, "learning_rate": 5.324477063819689e-06, "loss": 1.1614, "step": 24517 }, { "epoch": 1.921473354231975, "grad_norm": 2.8818466663360596, "learning_rate": 5.324169351149476e-06, "loss": 1.4602, "step": 24518 }, { "epoch": 1.921551724137931, "grad_norm": 3.334523916244507, "learning_rate": 5.3238616372462905e-06, "loss": 0.9719, "step": 24519 }, { "epoch": 1.9216300940438873, "grad_norm": 7.712810039520264, "learning_rate": 5.323553922111299e-06, "loss": 1.1909, "step": 24520 }, { "epoch": 1.9217084639498432, "grad_norm": 8.752166748046875, "learning_rate": 5.323246205745677e-06, "loss": 1.6864, "step": 24521 }, { "epoch": 1.9217868338557995, "grad_norm": 11.745826721191406, "learning_rate": 5.32293848815059e-06, "loss": 1.2288, "step": 24522 }, { "epoch": 1.9218652037617554, "grad_norm": 2.248370885848999, "learning_rate": 5.322630769327213e-06, "loss": 1.5128, "step": 24523 }, { "epoch": 1.9219435736677117, "grad_norm": 3.581937313079834, "learning_rate": 5.322323049276713e-06, "loss": 1.587, "step": 24524 }, { "epoch": 1.9220219435736676, "grad_norm": 3.0082554817199707, "learning_rate": 5.322015328000263e-06, "loss": 1.366, "step": 24525 }, { "epoch": 1.922100313479624, "grad_norm": 8.664509773254395, "learning_rate": 5.321707605499033e-06, "loss": 0.8915, "step": 24526 }, { "epoch": 1.9221786833855798, "grad_norm": 7.989385604858398, "learning_rate": 5.321399881774188e-06, "loss": 2.1739, "step": 24527 }, { "epoch": 1.9222570532915362, "grad_norm": 4.407105922698975, "learning_rate": 5.321092156826907e-06, "loss": 2.2109, "step": 24528 }, { "epoch": 1.922335423197492, "grad_norm": 2.2129461765289307, "learning_rate": 5.320784430658354e-06, "loss": 1.1854, "step": 24529 }, { "epoch": 1.9224137931034484, "grad_norm": 2.168984889984131, "learning_rate": 5.320476703269704e-06, "loss": 1.4169, "step": 24530 }, { "epoch": 1.9224921630094043, "grad_norm": 3.1107304096221924, "learning_rate": 5.320168974662125e-06, "loss": 1.6042, "step": 24531 }, { "epoch": 1.9225705329153606, "grad_norm": 4.316341400146484, "learning_rate": 5.319861244836786e-06, "loss": 1.5682, "step": 24532 }, { "epoch": 1.9226489028213165, "grad_norm": 5.346182346343994, "learning_rate": 5.319553513794859e-06, "loss": 1.1882, "step": 24533 }, { "epoch": 1.9227272727272728, "grad_norm": 4.6767706871032715, "learning_rate": 5.319245781537517e-06, "loss": 1.3431, "step": 24534 }, { "epoch": 1.9228056426332287, "grad_norm": 2.460157871246338, "learning_rate": 5.318938048065926e-06, "loss": 1.1385, "step": 24535 }, { "epoch": 1.922884012539185, "grad_norm": 3.805800437927246, "learning_rate": 5.3186303133812575e-06, "loss": 1.6712, "step": 24536 }, { "epoch": 1.922962382445141, "grad_norm": 2.309429883956909, "learning_rate": 5.3183225774846846e-06, "loss": 1.256, "step": 24537 }, { "epoch": 1.9230407523510973, "grad_norm": 4.712227821350098, "learning_rate": 5.318014840377377e-06, "loss": 1.2987, "step": 24538 }, { "epoch": 1.9231191222570532, "grad_norm": 2.046823263168335, "learning_rate": 5.317707102060503e-06, "loss": 0.934, "step": 24539 }, { "epoch": 1.9231974921630095, "grad_norm": 2.045700788497925, "learning_rate": 5.317399362535235e-06, "loss": 1.1994, "step": 24540 }, { "epoch": 1.9232758620689654, "grad_norm": 9.504439353942871, "learning_rate": 5.317091621802743e-06, "loss": 0.8161, "step": 24541 }, { "epoch": 1.9233542319749217, "grad_norm": 3.9293012619018555, "learning_rate": 5.316783879864196e-06, "loss": 1.5594, "step": 24542 }, { "epoch": 1.9234326018808776, "grad_norm": 5.131062030792236, "learning_rate": 5.316476136720766e-06, "loss": 0.867, "step": 24543 }, { "epoch": 1.923510971786834, "grad_norm": 4.40160608291626, "learning_rate": 5.316168392373624e-06, "loss": 1.2223, "step": 24544 }, { "epoch": 1.9235893416927898, "grad_norm": 2.721299886703491, "learning_rate": 5.315860646823942e-06, "loss": 1.666, "step": 24545 }, { "epoch": 1.9236677115987462, "grad_norm": 3.8394997119903564, "learning_rate": 5.315552900072885e-06, "loss": 1.3212, "step": 24546 }, { "epoch": 1.923746081504702, "grad_norm": 3.1209280490875244, "learning_rate": 5.315245152121628e-06, "loss": 1.4433, "step": 24547 }, { "epoch": 1.9238244514106584, "grad_norm": 3.7395734786987305, "learning_rate": 5.314937402971339e-06, "loss": 1.5736, "step": 24548 }, { "epoch": 1.9239028213166143, "grad_norm": 3.815302610397339, "learning_rate": 5.314629652623192e-06, "loss": 1.4695, "step": 24549 }, { "epoch": 1.9239811912225706, "grad_norm": 2.4180142879486084, "learning_rate": 5.314321901078355e-06, "loss": 1.2257, "step": 24550 }, { "epoch": 1.9240595611285265, "grad_norm": 3.5173280239105225, "learning_rate": 5.314014148337998e-06, "loss": 1.3689, "step": 24551 }, { "epoch": 1.9241379310344828, "grad_norm": 2.9095325469970703, "learning_rate": 5.313706394403293e-06, "loss": 1.5195, "step": 24552 }, { "epoch": 1.9242163009404387, "grad_norm": 3.2936275005340576, "learning_rate": 5.31339863927541e-06, "loss": 1.3304, "step": 24553 }, { "epoch": 1.924294670846395, "grad_norm": 6.365375995635986, "learning_rate": 5.313090882955521e-06, "loss": 0.2559, "step": 24554 }, { "epoch": 1.924373040752351, "grad_norm": 3.3794240951538086, "learning_rate": 5.312783125444791e-06, "loss": 1.4807, "step": 24555 }, { "epoch": 1.9244514106583073, "grad_norm": 3.73923921585083, "learning_rate": 5.3124753667444e-06, "loss": 1.3933, "step": 24556 }, { "epoch": 1.9245297805642632, "grad_norm": 2.836660861968994, "learning_rate": 5.3121676068555085e-06, "loss": 1.293, "step": 24557 }, { "epoch": 1.9246081504702195, "grad_norm": 3.641728639602661, "learning_rate": 5.311859845779295e-06, "loss": 0.9112, "step": 24558 }, { "epoch": 1.9246865203761756, "grad_norm": 3.6649744510650635, "learning_rate": 5.311552083516926e-06, "loss": 1.2241, "step": 24559 }, { "epoch": 1.9247648902821317, "grad_norm": 2.121642827987671, "learning_rate": 5.311244320069572e-06, "loss": 0.9949, "step": 24560 }, { "epoch": 1.9248432601880878, "grad_norm": 4.127458095550537, "learning_rate": 5.310936555438405e-06, "loss": 2.1536, "step": 24561 }, { "epoch": 1.924921630094044, "grad_norm": 2.782313346862793, "learning_rate": 5.310628789624594e-06, "loss": 0.9936, "step": 24562 }, { "epoch": 1.925, "grad_norm": 3.87740159034729, "learning_rate": 5.3103210226293125e-06, "loss": 1.5251, "step": 24563 }, { "epoch": 1.9250783699059562, "grad_norm": 3.6925559043884277, "learning_rate": 5.310013254453728e-06, "loss": 1.5328, "step": 24564 }, { "epoch": 1.9251567398119123, "grad_norm": 1.969544529914856, "learning_rate": 5.309705485099013e-06, "loss": 0.9535, "step": 24565 }, { "epoch": 1.9252351097178684, "grad_norm": 2.4011752605438232, "learning_rate": 5.3093977145663376e-06, "loss": 1.4472, "step": 24566 }, { "epoch": 1.9253134796238245, "grad_norm": 2.92787504196167, "learning_rate": 5.309089942856871e-06, "loss": 1.0455, "step": 24567 }, { "epoch": 1.9253918495297806, "grad_norm": 2.445129871368408, "learning_rate": 5.3087821699717856e-06, "loss": 1.7219, "step": 24568 }, { "epoch": 1.9254702194357367, "grad_norm": 1.9942950010299683, "learning_rate": 5.308474395912252e-06, "loss": 1.1591, "step": 24569 }, { "epoch": 1.9255485893416928, "grad_norm": 3.054525136947632, "learning_rate": 5.308166620679439e-06, "loss": 1.0205, "step": 24570 }, { "epoch": 1.925626959247649, "grad_norm": 2.666867971420288, "learning_rate": 5.307858844274519e-06, "loss": 1.4039, "step": 24571 }, { "epoch": 1.925705329153605, "grad_norm": 2.759824275970459, "learning_rate": 5.307551066698663e-06, "loss": 1.2374, "step": 24572 }, { "epoch": 1.9257836990595611, "grad_norm": 11.522663116455078, "learning_rate": 5.3072432879530414e-06, "loss": 1.3833, "step": 24573 }, { "epoch": 1.9258620689655173, "grad_norm": 13.477392196655273, "learning_rate": 5.306935508038821e-06, "loss": 2.706, "step": 24574 }, { "epoch": 1.9259404388714734, "grad_norm": 1.789648413658142, "learning_rate": 5.306627726957179e-06, "loss": 1.3416, "step": 24575 }, { "epoch": 1.9260188087774295, "grad_norm": 9.198328971862793, "learning_rate": 5.30631994470928e-06, "loss": 1.5574, "step": 24576 }, { "epoch": 1.9260971786833856, "grad_norm": 3.3160345554351807, "learning_rate": 5.306012161296299e-06, "loss": 1.4283, "step": 24577 }, { "epoch": 1.9261755485893417, "grad_norm": 2.0905280113220215, "learning_rate": 5.305704376719404e-06, "loss": 1.2051, "step": 24578 }, { "epoch": 1.9262539184952978, "grad_norm": 2.5266361236572266, "learning_rate": 5.305396590979767e-06, "loss": 1.7189, "step": 24579 }, { "epoch": 1.926332288401254, "grad_norm": 2.1360061168670654, "learning_rate": 5.305088804078559e-06, "loss": 1.5257, "step": 24580 }, { "epoch": 1.92641065830721, "grad_norm": 3.1163363456726074, "learning_rate": 5.304781016016947e-06, "loss": 1.1349, "step": 24581 }, { "epoch": 1.9264890282131661, "grad_norm": 2.4596803188323975, "learning_rate": 5.3044732267961075e-06, "loss": 1.3162, "step": 24582 }, { "epoch": 1.9265673981191223, "grad_norm": 6.284365177154541, "learning_rate": 5.304165436417206e-06, "loss": 1.6517, "step": 24583 }, { "epoch": 1.9266457680250784, "grad_norm": 3.684450149536133, "learning_rate": 5.303857644881417e-06, "loss": 1.2117, "step": 24584 }, { "epoch": 1.9267241379310345, "grad_norm": 9.863750457763672, "learning_rate": 5.30354985218991e-06, "loss": 1.4164, "step": 24585 }, { "epoch": 1.9268025078369906, "grad_norm": 2.461594343185425, "learning_rate": 5.303242058343854e-06, "loss": 1.1011, "step": 24586 }, { "epoch": 1.9268808777429467, "grad_norm": 6.476289749145508, "learning_rate": 5.302934263344421e-06, "loss": 0.9632, "step": 24587 }, { "epoch": 1.9269592476489028, "grad_norm": 6.949549674987793, "learning_rate": 5.3026264671927816e-06, "loss": 1.0746, "step": 24588 }, { "epoch": 1.927037617554859, "grad_norm": 2.017784357070923, "learning_rate": 5.302318669890108e-06, "loss": 1.2992, "step": 24589 }, { "epoch": 1.927115987460815, "grad_norm": 3.14432692527771, "learning_rate": 5.3020108714375674e-06, "loss": 1.7266, "step": 24590 }, { "epoch": 1.9271943573667711, "grad_norm": 5.022832870483398, "learning_rate": 5.3017030718363335e-06, "loss": 1.4665, "step": 24591 }, { "epoch": 1.9272727272727272, "grad_norm": 2.7909345626831055, "learning_rate": 5.301395271087577e-06, "loss": 1.272, "step": 24592 }, { "epoch": 1.9273510971786834, "grad_norm": 3.0058305263519287, "learning_rate": 5.301087469192466e-06, "loss": 1.086, "step": 24593 }, { "epoch": 1.9274294670846395, "grad_norm": 4.107060432434082, "learning_rate": 5.300779666152175e-06, "loss": 1.273, "step": 24594 }, { "epoch": 1.9275078369905956, "grad_norm": 2.005129337310791, "learning_rate": 5.3004718619678705e-06, "loss": 1.1562, "step": 24595 }, { "epoch": 1.9275862068965517, "grad_norm": 3.288473606109619, "learning_rate": 5.300164056640727e-06, "loss": 1.6085, "step": 24596 }, { "epoch": 1.9276645768025078, "grad_norm": 4.250787734985352, "learning_rate": 5.299856250171913e-06, "loss": 1.3718, "step": 24597 }, { "epoch": 1.927742946708464, "grad_norm": 4.491955280303955, "learning_rate": 5.2995484425626e-06, "loss": 2.1576, "step": 24598 }, { "epoch": 1.92782131661442, "grad_norm": 2.240438222885132, "learning_rate": 5.2992406338139575e-06, "loss": 1.1218, "step": 24599 }, { "epoch": 1.9278996865203761, "grad_norm": 2.4741008281707764, "learning_rate": 5.298932823927158e-06, "loss": 1.2221, "step": 24600 }, { "epoch": 1.9279780564263322, "grad_norm": 5.004998683929443, "learning_rate": 5.298625012903373e-06, "loss": 0.8501, "step": 24601 }, { "epoch": 1.9280564263322884, "grad_norm": 2.161280870437622, "learning_rate": 5.298317200743769e-06, "loss": 1.2295, "step": 24602 }, { "epoch": 1.9281347962382445, "grad_norm": 2.3122448921203613, "learning_rate": 5.298009387449523e-06, "loss": 1.4411, "step": 24603 }, { "epoch": 1.9282131661442006, "grad_norm": 2.9825894832611084, "learning_rate": 5.2977015730218e-06, "loss": 1.3419, "step": 24604 }, { "epoch": 1.9282915360501567, "grad_norm": 3.080493688583374, "learning_rate": 5.297393757461774e-06, "loss": 1.6215, "step": 24605 }, { "epoch": 1.9283699059561128, "grad_norm": 7.3090338706970215, "learning_rate": 5.297085940770614e-06, "loss": 1.4967, "step": 24606 }, { "epoch": 1.9284482758620691, "grad_norm": 2.4867005348205566, "learning_rate": 5.296778122949492e-06, "loss": 1.2532, "step": 24607 }, { "epoch": 1.928526645768025, "grad_norm": 3.110360622406006, "learning_rate": 5.296470303999581e-06, "loss": 1.1167, "step": 24608 }, { "epoch": 1.9286050156739813, "grad_norm": 7.07588529586792, "learning_rate": 5.296162483922045e-06, "loss": 0.5229, "step": 24609 }, { "epoch": 1.9286833855799372, "grad_norm": 3.2578542232513428, "learning_rate": 5.295854662718062e-06, "loss": 1.243, "step": 24610 }, { "epoch": 1.9287617554858936, "grad_norm": 2.7841343879699707, "learning_rate": 5.295546840388797e-06, "loss": 1.3322, "step": 24611 }, { "epoch": 1.9288401253918495, "grad_norm": 2.7738311290740967, "learning_rate": 5.2952390169354266e-06, "loss": 1.3079, "step": 24612 }, { "epoch": 1.9289184952978058, "grad_norm": 2.6312928199768066, "learning_rate": 5.294931192359117e-06, "loss": 1.3537, "step": 24613 }, { "epoch": 1.9289968652037617, "grad_norm": 6.499255657196045, "learning_rate": 5.294623366661041e-06, "loss": 1.893, "step": 24614 }, { "epoch": 1.929075235109718, "grad_norm": 5.036432266235352, "learning_rate": 5.294315539842369e-06, "loss": 1.2564, "step": 24615 }, { "epoch": 1.929153605015674, "grad_norm": 3.793062925338745, "learning_rate": 5.294007711904272e-06, "loss": 1.3053, "step": 24616 }, { "epoch": 1.9292319749216302, "grad_norm": 7.562624454498291, "learning_rate": 5.293699882847921e-06, "loss": 0.9959, "step": 24617 }, { "epoch": 1.9293103448275861, "grad_norm": 3.898768186569214, "learning_rate": 5.293392052674485e-06, "loss": 1.6146, "step": 24618 }, { "epoch": 1.9293887147335425, "grad_norm": 2.291872978210449, "learning_rate": 5.293084221385139e-06, "loss": 0.9456, "step": 24619 }, { "epoch": 1.9294670846394983, "grad_norm": 5.086706638336182, "learning_rate": 5.29277638898105e-06, "loss": 1.6382, "step": 24620 }, { "epoch": 1.9295454545454547, "grad_norm": 2.8449630737304688, "learning_rate": 5.292468555463388e-06, "loss": 1.6296, "step": 24621 }, { "epoch": 1.9296238244514106, "grad_norm": 2.4780237674713135, "learning_rate": 5.292160720833328e-06, "loss": 1.496, "step": 24622 }, { "epoch": 1.929702194357367, "grad_norm": 14.184612274169922, "learning_rate": 5.291852885092039e-06, "loss": 1.5542, "step": 24623 }, { "epoch": 1.9297805642633228, "grad_norm": 2.7720348834991455, "learning_rate": 5.29154504824069e-06, "loss": 1.1894, "step": 24624 }, { "epoch": 1.9298589341692791, "grad_norm": 3.9765424728393555, "learning_rate": 5.291237210280455e-06, "loss": 0.5645, "step": 24625 }, { "epoch": 1.929937304075235, "grad_norm": 4.18750524520874, "learning_rate": 5.290929371212501e-06, "loss": 1.5515, "step": 24626 }, { "epoch": 1.9300156739811913, "grad_norm": 3.874499797821045, "learning_rate": 5.290621531038004e-06, "loss": 1.3389, "step": 24627 }, { "epoch": 1.9300940438871472, "grad_norm": 3.284468412399292, "learning_rate": 5.290313689758128e-06, "loss": 1.3607, "step": 24628 }, { "epoch": 1.9301724137931036, "grad_norm": 2.2170398235321045, "learning_rate": 5.290005847374052e-06, "loss": 1.3325, "step": 24629 }, { "epoch": 1.9302507836990594, "grad_norm": 3.2965548038482666, "learning_rate": 5.289698003886939e-06, "loss": 1.5938, "step": 24630 }, { "epoch": 1.9303291536050158, "grad_norm": 1.928925633430481, "learning_rate": 5.289390159297967e-06, "loss": 1.117, "step": 24631 }, { "epoch": 1.9304075235109717, "grad_norm": 3.1162471771240234, "learning_rate": 5.289082313608302e-06, "loss": 1.2916, "step": 24632 }, { "epoch": 1.930485893416928, "grad_norm": 2.0938735008239746, "learning_rate": 5.288774466819117e-06, "loss": 1.3258, "step": 24633 }, { "epoch": 1.9305642633228839, "grad_norm": 3.6560230255126953, "learning_rate": 5.288466618931581e-06, "loss": 1.2022, "step": 24634 }, { "epoch": 1.9306426332288402, "grad_norm": 2.550771474838257, "learning_rate": 5.288158769946866e-06, "loss": 1.1768, "step": 24635 }, { "epoch": 1.930721003134796, "grad_norm": 2.3842074871063232, "learning_rate": 5.2878509198661456e-06, "loss": 1.4198, "step": 24636 }, { "epoch": 1.9307993730407524, "grad_norm": 3.9125983715057373, "learning_rate": 5.287543068690585e-06, "loss": 1.334, "step": 24637 }, { "epoch": 1.9308777429467083, "grad_norm": 2.381021499633789, "learning_rate": 5.287235216421361e-06, "loss": 1.2834, "step": 24638 }, { "epoch": 1.9309561128526647, "grad_norm": 2.3051552772521973, "learning_rate": 5.28692736305964e-06, "loss": 1.0443, "step": 24639 }, { "epoch": 1.9310344827586206, "grad_norm": 2.6694023609161377, "learning_rate": 5.286619508606595e-06, "loss": 1.2536, "step": 24640 }, { "epoch": 1.9311128526645769, "grad_norm": 2.647110939025879, "learning_rate": 5.2863116530633965e-06, "loss": 1.3023, "step": 24641 }, { "epoch": 1.9311912225705328, "grad_norm": 5.164106369018555, "learning_rate": 5.286003796431215e-06, "loss": 0.9424, "step": 24642 }, { "epoch": 1.931269592476489, "grad_norm": 5.2831950187683105, "learning_rate": 5.285695938711223e-06, "loss": 1.6504, "step": 24643 }, { "epoch": 1.931347962382445, "grad_norm": 1.9454128742218018, "learning_rate": 5.2853880799045896e-06, "loss": 1.1582, "step": 24644 }, { "epoch": 1.9314263322884013, "grad_norm": 2.792381525039673, "learning_rate": 5.285080220012487e-06, "loss": 1.6591, "step": 24645 }, { "epoch": 1.9315047021943572, "grad_norm": 2.8726582527160645, "learning_rate": 5.284772359036087e-06, "loss": 1.0995, "step": 24646 }, { "epoch": 1.9315830721003135, "grad_norm": 2.421776056289673, "learning_rate": 5.284464496976557e-06, "loss": 1.2689, "step": 24647 }, { "epoch": 1.9316614420062694, "grad_norm": 4.845242023468018, "learning_rate": 5.284156633835071e-06, "loss": 1.5886, "step": 24648 }, { "epoch": 1.9317398119122258, "grad_norm": 2.68851637840271, "learning_rate": 5.283848769612798e-06, "loss": 1.216, "step": 24649 }, { "epoch": 1.9318181818181817, "grad_norm": 2.7509491443634033, "learning_rate": 5.283540904310913e-06, "loss": 1.546, "step": 24650 }, { "epoch": 1.931896551724138, "grad_norm": 2.419097661972046, "learning_rate": 5.283233037930582e-06, "loss": 1.5262, "step": 24651 }, { "epoch": 1.9319749216300939, "grad_norm": 3.6908724308013916, "learning_rate": 5.282925170472978e-06, "loss": 1.9767, "step": 24652 }, { "epoch": 1.9320532915360502, "grad_norm": 3.4497008323669434, "learning_rate": 5.282617301939272e-06, "loss": 1.1094, "step": 24653 }, { "epoch": 1.9321316614420063, "grad_norm": 3.5227365493774414, "learning_rate": 5.282309432330636e-06, "loss": 1.5463, "step": 24654 }, { "epoch": 1.9322100313479624, "grad_norm": 2.3849380016326904, "learning_rate": 5.282001561648239e-06, "loss": 1.4603, "step": 24655 }, { "epoch": 1.9322884012539185, "grad_norm": 2.2344231605529785, "learning_rate": 5.281693689893252e-06, "loss": 1.3249, "step": 24656 }, { "epoch": 1.9323667711598747, "grad_norm": 7.7468647956848145, "learning_rate": 5.281385817066848e-06, "loss": 0.6323, "step": 24657 }, { "epoch": 1.9324451410658308, "grad_norm": 5.129304885864258, "learning_rate": 5.281077943170196e-06, "loss": 1.6881, "step": 24658 }, { "epoch": 1.9325235109717869, "grad_norm": 2.919297218322754, "learning_rate": 5.28077006820447e-06, "loss": 1.3589, "step": 24659 }, { "epoch": 1.932601880877743, "grad_norm": 3.10257625579834, "learning_rate": 5.280462192170838e-06, "loss": 1.0453, "step": 24660 }, { "epoch": 1.932680250783699, "grad_norm": 2.3557465076446533, "learning_rate": 5.28015431507047e-06, "loss": 1.3638, "step": 24661 }, { "epoch": 1.9327586206896552, "grad_norm": 2.8278157711029053, "learning_rate": 5.279846436904541e-06, "loss": 1.1641, "step": 24662 }, { "epoch": 1.9328369905956113, "grad_norm": 15.040692329406738, "learning_rate": 5.279538557674219e-06, "loss": 2.1339, "step": 24663 }, { "epoch": 1.9329153605015674, "grad_norm": 3.0055506229400635, "learning_rate": 5.2792306773806765e-06, "loss": 1.4046, "step": 24664 }, { "epoch": 1.9329937304075235, "grad_norm": 9.413166046142578, "learning_rate": 5.278922796025084e-06, "loss": 0.8635, "step": 24665 }, { "epoch": 1.9330721003134796, "grad_norm": 2.776700496673584, "learning_rate": 5.278614913608612e-06, "loss": 1.5072, "step": 24666 }, { "epoch": 1.9331504702194358, "grad_norm": 3.941835880279541, "learning_rate": 5.278307030132431e-06, "loss": 1.3677, "step": 24667 }, { "epoch": 1.9332288401253919, "grad_norm": 8.852404594421387, "learning_rate": 5.277999145597715e-06, "loss": 1.5868, "step": 24668 }, { "epoch": 1.933307210031348, "grad_norm": 3.6580185890197754, "learning_rate": 5.2776912600056314e-06, "loss": 1.5296, "step": 24669 }, { "epoch": 1.933385579937304, "grad_norm": 2.8626770973205566, "learning_rate": 5.277383373357353e-06, "loss": 1.1915, "step": 24670 }, { "epoch": 1.9334639498432602, "grad_norm": 2.1085798740386963, "learning_rate": 5.2770754856540515e-06, "loss": 1.0881, "step": 24671 }, { "epoch": 1.9335423197492163, "grad_norm": 2.6856536865234375, "learning_rate": 5.276767596896897e-06, "loss": 1.4283, "step": 24672 }, { "epoch": 1.9336206896551724, "grad_norm": 2.2414615154266357, "learning_rate": 5.2764597070870604e-06, "loss": 1.4197, "step": 24673 }, { "epoch": 1.9336990595611285, "grad_norm": 3.299252510070801, "learning_rate": 5.276151816225714e-06, "loss": 1.3264, "step": 24674 }, { "epoch": 1.9337774294670846, "grad_norm": 4.520290851593018, "learning_rate": 5.275843924314026e-06, "loss": 1.456, "step": 24675 }, { "epoch": 1.9338557993730408, "grad_norm": 2.1385865211486816, "learning_rate": 5.275536031353171e-06, "loss": 1.5925, "step": 24676 }, { "epoch": 1.9339341692789969, "grad_norm": 3.748685836791992, "learning_rate": 5.275228137344317e-06, "loss": 1.5343, "step": 24677 }, { "epoch": 1.934012539184953, "grad_norm": 2.6854803562164307, "learning_rate": 5.274920242288639e-06, "loss": 1.3061, "step": 24678 }, { "epoch": 1.934090909090909, "grad_norm": 3.97202205657959, "learning_rate": 5.2746123461873035e-06, "loss": 1.174, "step": 24679 }, { "epoch": 1.9341692789968652, "grad_norm": 2.769465684890747, "learning_rate": 5.274304449041484e-06, "loss": 1.4089, "step": 24680 }, { "epoch": 1.9342476489028213, "grad_norm": 2.800751209259033, "learning_rate": 5.2739965508523525e-06, "loss": 1.8438, "step": 24681 }, { "epoch": 1.9343260188087774, "grad_norm": 4.492627143859863, "learning_rate": 5.273688651621077e-06, "loss": 1.3121, "step": 24682 }, { "epoch": 1.9344043887147335, "grad_norm": 2.632565498352051, "learning_rate": 5.273380751348832e-06, "loss": 1.3564, "step": 24683 }, { "epoch": 1.9344827586206896, "grad_norm": 13.371661186218262, "learning_rate": 5.273072850036786e-06, "loss": 1.3727, "step": 24684 }, { "epoch": 1.9345611285266457, "grad_norm": 2.0820658206939697, "learning_rate": 5.272764947686112e-06, "loss": 1.2409, "step": 24685 }, { "epoch": 1.9346394984326019, "grad_norm": 3.22506046295166, "learning_rate": 5.272457044297981e-06, "loss": 1.2805, "step": 24686 }, { "epoch": 1.934717868338558, "grad_norm": 3.6907951831817627, "learning_rate": 5.2721491398735615e-06, "loss": 1.3995, "step": 24687 }, { "epoch": 1.934796238244514, "grad_norm": 4.004587650299072, "learning_rate": 5.271841234414027e-06, "loss": 1.9289, "step": 24688 }, { "epoch": 1.9348746081504702, "grad_norm": 2.817803382873535, "learning_rate": 5.271533327920548e-06, "loss": 0.7786, "step": 24689 }, { "epoch": 1.9349529780564263, "grad_norm": 4.136071681976318, "learning_rate": 5.2712254203942954e-06, "loss": 1.5685, "step": 24690 }, { "epoch": 1.9350313479623824, "grad_norm": 3.8663885593414307, "learning_rate": 5.270917511836441e-06, "loss": 1.2837, "step": 24691 }, { "epoch": 1.9351097178683385, "grad_norm": 3.3569564819335938, "learning_rate": 5.270609602248157e-06, "loss": 1.6888, "step": 24692 }, { "epoch": 1.9351880877742946, "grad_norm": 3.7218024730682373, "learning_rate": 5.2703016916306114e-06, "loss": 1.4065, "step": 24693 }, { "epoch": 1.9352664576802507, "grad_norm": 3.1884920597076416, "learning_rate": 5.269993779984977e-06, "loss": 1.7084, "step": 24694 }, { "epoch": 1.9353448275862069, "grad_norm": 5.502549171447754, "learning_rate": 5.269685867312425e-06, "loss": 1.1467, "step": 24695 }, { "epoch": 1.935423197492163, "grad_norm": 5.026683330535889, "learning_rate": 5.269377953614127e-06, "loss": 0.9482, "step": 24696 }, { "epoch": 1.935501567398119, "grad_norm": 2.741166353225708, "learning_rate": 5.269070038891253e-06, "loss": 0.9356, "step": 24697 }, { "epoch": 1.9355799373040752, "grad_norm": 1.9914860725402832, "learning_rate": 5.268762123144975e-06, "loss": 1.1636, "step": 24698 }, { "epoch": 1.9356583072100313, "grad_norm": 8.005188941955566, "learning_rate": 5.268454206376464e-06, "loss": 1.0233, "step": 24699 }, { "epoch": 1.9357366771159876, "grad_norm": 2.77959942817688, "learning_rate": 5.268146288586893e-06, "loss": 1.1053, "step": 24700 }, { "epoch": 1.9358150470219435, "grad_norm": 3.0160317420959473, "learning_rate": 5.2678383697774275e-06, "loss": 1.406, "step": 24701 }, { "epoch": 1.9358934169278998, "grad_norm": 2.7697248458862305, "learning_rate": 5.267530449949245e-06, "loss": 1.25, "step": 24702 }, { "epoch": 1.9359717868338557, "grad_norm": 3.1440188884735107, "learning_rate": 5.2672225291035125e-06, "loss": 1.7731, "step": 24703 }, { "epoch": 1.936050156739812, "grad_norm": 1.9275130033493042, "learning_rate": 5.266914607241405e-06, "loss": 1.1618, "step": 24704 }, { "epoch": 1.936128526645768, "grad_norm": 2.625988483428955, "learning_rate": 5.266606684364089e-06, "loss": 0.9258, "step": 24705 }, { "epoch": 1.9362068965517243, "grad_norm": 2.931429147720337, "learning_rate": 5.266298760472739e-06, "loss": 1.5105, "step": 24706 }, { "epoch": 1.9362852664576802, "grad_norm": 3.511211395263672, "learning_rate": 5.265990835568526e-06, "loss": 1.4813, "step": 24707 }, { "epoch": 1.9363636363636365, "grad_norm": 3.177882671356201, "learning_rate": 5.265682909652619e-06, "loss": 1.4721, "step": 24708 }, { "epoch": 1.9364420062695924, "grad_norm": 1.8700132369995117, "learning_rate": 5.265374982726193e-06, "loss": 1.3227, "step": 24709 }, { "epoch": 1.9365203761755487, "grad_norm": 2.748396396636963, "learning_rate": 5.265067054790415e-06, "loss": 1.447, "step": 24710 }, { "epoch": 1.9365987460815046, "grad_norm": 2.9021999835968018, "learning_rate": 5.264759125846459e-06, "loss": 0.8152, "step": 24711 }, { "epoch": 1.936677115987461, "grad_norm": 2.9375357627868652, "learning_rate": 5.264451195895494e-06, "loss": 1.3691, "step": 24712 }, { "epoch": 1.9367554858934168, "grad_norm": 5.255838871002197, "learning_rate": 5.264143264938693e-06, "loss": 2.0265, "step": 24713 }, { "epoch": 1.9368338557993732, "grad_norm": 2.8794729709625244, "learning_rate": 5.263835332977227e-06, "loss": 1.2243, "step": 24714 }, { "epoch": 1.936912225705329, "grad_norm": 4.542822360992432, "learning_rate": 5.263527400012268e-06, "loss": 1.7635, "step": 24715 }, { "epoch": 1.9369905956112854, "grad_norm": 2.016211986541748, "learning_rate": 5.263219466044984e-06, "loss": 1.1516, "step": 24716 }, { "epoch": 1.9370689655172413, "grad_norm": 5.635812282562256, "learning_rate": 5.26291153107655e-06, "loss": 1.5975, "step": 24717 }, { "epoch": 1.9371473354231976, "grad_norm": 3.7040228843688965, "learning_rate": 5.262603595108135e-06, "loss": 1.1998, "step": 24718 }, { "epoch": 1.9372257053291535, "grad_norm": 3.190655469894409, "learning_rate": 5.26229565814091e-06, "loss": 1.2915, "step": 24719 }, { "epoch": 1.9373040752351098, "grad_norm": 4.040642738342285, "learning_rate": 5.261987720176049e-06, "loss": 1.5938, "step": 24720 }, { "epoch": 1.9373824451410657, "grad_norm": 3.6986043453216553, "learning_rate": 5.2616797812147205e-06, "loss": 1.7979, "step": 24721 }, { "epoch": 1.937460815047022, "grad_norm": 3.333353042602539, "learning_rate": 5.261371841258096e-06, "loss": 1.2407, "step": 24722 }, { "epoch": 1.937539184952978, "grad_norm": 2.2225358486175537, "learning_rate": 5.261063900307347e-06, "loss": 1.0388, "step": 24723 }, { "epoch": 1.9376175548589343, "grad_norm": 2.6835646629333496, "learning_rate": 5.260755958363646e-06, "loss": 1.1383, "step": 24724 }, { "epoch": 1.9376959247648902, "grad_norm": 2.9006247520446777, "learning_rate": 5.260448015428162e-06, "loss": 1.3586, "step": 24725 }, { "epoch": 1.9377742946708465, "grad_norm": 7.836759090423584, "learning_rate": 5.260140071502068e-06, "loss": 1.2937, "step": 24726 }, { "epoch": 1.9378526645768024, "grad_norm": 3.4957621097564697, "learning_rate": 5.259832126586535e-06, "loss": 1.3172, "step": 24727 }, { "epoch": 1.9379310344827587, "grad_norm": 2.347444534301758, "learning_rate": 5.259524180682734e-06, "loss": 1.6709, "step": 24728 }, { "epoch": 1.9380094043887146, "grad_norm": 2.3981094360351562, "learning_rate": 5.259216233791836e-06, "loss": 1.6193, "step": 24729 }, { "epoch": 1.938087774294671, "grad_norm": 3.417088270187378, "learning_rate": 5.258908285915014e-06, "loss": 1.47, "step": 24730 }, { "epoch": 1.9381661442006268, "grad_norm": 5.298465251922607, "learning_rate": 5.258600337053436e-06, "loss": 1.5694, "step": 24731 }, { "epoch": 1.9382445141065832, "grad_norm": 2.4781124591827393, "learning_rate": 5.258292387208276e-06, "loss": 1.3234, "step": 24732 }, { "epoch": 1.938322884012539, "grad_norm": 2.5781409740448, "learning_rate": 5.257984436380704e-06, "loss": 1.3379, "step": 24733 }, { "epoch": 1.9384012539184954, "grad_norm": 3.0386178493499756, "learning_rate": 5.2576764845718916e-06, "loss": 1.5197, "step": 24734 }, { "epoch": 1.9384796238244513, "grad_norm": 5.661633491516113, "learning_rate": 5.257368531783013e-06, "loss": 1.503, "step": 24735 }, { "epoch": 1.9385579937304076, "grad_norm": 3.711230516433716, "learning_rate": 5.257060578015233e-06, "loss": 1.8287, "step": 24736 }, { "epoch": 1.9386363636363635, "grad_norm": 2.7936527729034424, "learning_rate": 5.256752623269728e-06, "loss": 1.3515, "step": 24737 }, { "epoch": 1.9387147335423198, "grad_norm": 2.123933792114258, "learning_rate": 5.2564446675476675e-06, "loss": 1.3776, "step": 24738 }, { "epoch": 1.9387931034482757, "grad_norm": 3.108711004257202, "learning_rate": 5.2561367108502245e-06, "loss": 1.3738, "step": 24739 }, { "epoch": 1.938871473354232, "grad_norm": 3.224459409713745, "learning_rate": 5.255828753178568e-06, "loss": 0.9656, "step": 24740 }, { "epoch": 1.938949843260188, "grad_norm": 2.3278141021728516, "learning_rate": 5.25552079453387e-06, "loss": 1.2035, "step": 24741 }, { "epoch": 1.9390282131661443, "grad_norm": 4.756519317626953, "learning_rate": 5.255212834917303e-06, "loss": 1.3547, "step": 24742 }, { "epoch": 1.9391065830721002, "grad_norm": 3.561386823654175, "learning_rate": 5.254904874330037e-06, "loss": 1.1394, "step": 24743 }, { "epoch": 1.9391849529780565, "grad_norm": 3.58752703666687, "learning_rate": 5.254596912773243e-06, "loss": 1.0242, "step": 24744 }, { "epoch": 1.9392633228840124, "grad_norm": 1.979660987854004, "learning_rate": 5.254288950248094e-06, "loss": 1.2289, "step": 24745 }, { "epoch": 1.9393416927899687, "grad_norm": 3.2224440574645996, "learning_rate": 5.25398098675576e-06, "loss": 1.5142, "step": 24746 }, { "epoch": 1.9394200626959248, "grad_norm": 4.971282005310059, "learning_rate": 5.253673022297414e-06, "loss": 1.4053, "step": 24747 }, { "epoch": 1.939498432601881, "grad_norm": 2.7340526580810547, "learning_rate": 5.253365056874226e-06, "loss": 1.4189, "step": 24748 }, { "epoch": 1.939576802507837, "grad_norm": 3.282297134399414, "learning_rate": 5.2530570904873655e-06, "loss": 1.5031, "step": 24749 }, { "epoch": 1.9396551724137931, "grad_norm": 3.3917722702026367, "learning_rate": 5.252749123138008e-06, "loss": 1.3965, "step": 24750 }, { "epoch": 1.9397335423197493, "grad_norm": 3.907764196395874, "learning_rate": 5.2524411548273215e-06, "loss": 1.1952, "step": 24751 }, { "epoch": 1.9398119122257054, "grad_norm": 2.1776881217956543, "learning_rate": 5.252133185556478e-06, "loss": 1.1586, "step": 24752 }, { "epoch": 1.9398902821316615, "grad_norm": 2.8699147701263428, "learning_rate": 5.2518252153266506e-06, "loss": 1.5676, "step": 24753 }, { "epoch": 1.9399686520376176, "grad_norm": 6.79036283493042, "learning_rate": 5.251517244139008e-06, "loss": 1.7418, "step": 24754 }, { "epoch": 1.9400470219435737, "grad_norm": 3.073145866394043, "learning_rate": 5.251209271994724e-06, "loss": 1.1629, "step": 24755 }, { "epoch": 1.9401253918495298, "grad_norm": 2.0742735862731934, "learning_rate": 5.250901298894969e-06, "loss": 1.4817, "step": 24756 }, { "epoch": 1.940203761755486, "grad_norm": 2.802090644836426, "learning_rate": 5.250593324840913e-06, "loss": 1.3415, "step": 24757 }, { "epoch": 1.940282131661442, "grad_norm": 3.932370901107788, "learning_rate": 5.250285349833732e-06, "loss": 1.3675, "step": 24758 }, { "epoch": 1.9403605015673981, "grad_norm": 2.3710453510284424, "learning_rate": 5.249977373874591e-06, "loss": 1.1087, "step": 24759 }, { "epoch": 1.9404388714733543, "grad_norm": 3.859257459640503, "learning_rate": 5.249669396964665e-06, "loss": 1.3088, "step": 24760 }, { "epoch": 1.9405172413793104, "grad_norm": 2.401404619216919, "learning_rate": 5.249361419105125e-06, "loss": 1.4492, "step": 24761 }, { "epoch": 1.9405956112852665, "grad_norm": 8.753183364868164, "learning_rate": 5.2490534402971425e-06, "loss": 1.813, "step": 24762 }, { "epoch": 1.9406739811912226, "grad_norm": 2.7111666202545166, "learning_rate": 5.24874546054189e-06, "loss": 1.6287, "step": 24763 }, { "epoch": 1.9407523510971787, "grad_norm": 3.7337112426757812, "learning_rate": 5.248437479840535e-06, "loss": 0.999, "step": 24764 }, { "epoch": 1.9408307210031348, "grad_norm": 2.844309091567993, "learning_rate": 5.248129498194254e-06, "loss": 1.5193, "step": 24765 }, { "epoch": 1.940909090909091, "grad_norm": 2.1248505115509033, "learning_rate": 5.247821515604213e-06, "loss": 1.4463, "step": 24766 }, { "epoch": 1.940987460815047, "grad_norm": 2.736370325088501, "learning_rate": 5.247513532071588e-06, "loss": 1.7316, "step": 24767 }, { "epoch": 1.9410658307210031, "grad_norm": 8.367180824279785, "learning_rate": 5.247205547597549e-06, "loss": 1.1348, "step": 24768 }, { "epoch": 1.9411442006269592, "grad_norm": 2.9411168098449707, "learning_rate": 5.2468975621832665e-06, "loss": 1.0247, "step": 24769 }, { "epoch": 1.9412225705329154, "grad_norm": 2.920644760131836, "learning_rate": 5.246589575829913e-06, "loss": 1.2645, "step": 24770 }, { "epoch": 1.9413009404388715, "grad_norm": 2.590744972229004, "learning_rate": 5.246281588538658e-06, "loss": 1.3433, "step": 24771 }, { "epoch": 1.9413793103448276, "grad_norm": 2.269879102706909, "learning_rate": 5.245973600310674e-06, "loss": 1.486, "step": 24772 }, { "epoch": 1.9414576802507837, "grad_norm": 3.6763405799865723, "learning_rate": 5.245665611147135e-06, "loss": 1.6877, "step": 24773 }, { "epoch": 1.9415360501567398, "grad_norm": 3.106323003768921, "learning_rate": 5.2453576210492085e-06, "loss": 1.4092, "step": 24774 }, { "epoch": 1.941614420062696, "grad_norm": 2.9773032665252686, "learning_rate": 5.2450496300180685e-06, "loss": 1.2108, "step": 24775 }, { "epoch": 1.941692789968652, "grad_norm": 1.969048261642456, "learning_rate": 5.244741638054886e-06, "loss": 1.2369, "step": 24776 }, { "epoch": 1.9417711598746081, "grad_norm": 3.01840877532959, "learning_rate": 5.244433645160831e-06, "loss": 1.4592, "step": 24777 }, { "epoch": 1.9418495297805642, "grad_norm": 4.615586757659912, "learning_rate": 5.244125651337076e-06, "loss": 1.2078, "step": 24778 }, { "epoch": 1.9419278996865204, "grad_norm": 1.9975749254226685, "learning_rate": 5.2438176565847934e-06, "loss": 1.3762, "step": 24779 }, { "epoch": 1.9420062695924765, "grad_norm": 2.3321421146392822, "learning_rate": 5.243509660905153e-06, "loss": 1.4321, "step": 24780 }, { "epoch": 1.9420846394984326, "grad_norm": 4.156912326812744, "learning_rate": 5.2432016642993255e-06, "loss": 1.8019, "step": 24781 }, { "epoch": 1.9421630094043887, "grad_norm": 3.118842363357544, "learning_rate": 5.2428936667684875e-06, "loss": 1.3177, "step": 24782 }, { "epoch": 1.9422413793103448, "grad_norm": 2.419811725616455, "learning_rate": 5.2425856683138025e-06, "loss": 1.2978, "step": 24783 }, { "epoch": 1.942319749216301, "grad_norm": 3.8410394191741943, "learning_rate": 5.242277668936449e-06, "loss": 1.595, "step": 24784 }, { "epoch": 1.942398119122257, "grad_norm": 8.197518348693848, "learning_rate": 5.2419696686375935e-06, "loss": 1.1091, "step": 24785 }, { "epoch": 1.9424764890282131, "grad_norm": 4.123672962188721, "learning_rate": 5.2416616674184115e-06, "loss": 1.7972, "step": 24786 }, { "epoch": 1.9425548589341692, "grad_norm": 6.77444314956665, "learning_rate": 5.2413536652800726e-06, "loss": 1.2496, "step": 24787 }, { "epoch": 1.9426332288401253, "grad_norm": 6.61969518661499, "learning_rate": 5.241045662223748e-06, "loss": 0.8049, "step": 24788 }, { "epoch": 1.9427115987460815, "grad_norm": 3.574662446975708, "learning_rate": 5.24073765825061e-06, "loss": 1.2345, "step": 24789 }, { "epoch": 1.9427899686520376, "grad_norm": 2.711979627609253, "learning_rate": 5.2404296533618285e-06, "loss": 1.5326, "step": 24790 }, { "epoch": 1.9428683385579937, "grad_norm": 3.1372549533843994, "learning_rate": 5.240121647558577e-06, "loss": 1.2387, "step": 24791 }, { "epoch": 1.9429467084639498, "grad_norm": 2.270298957824707, "learning_rate": 5.239813640842024e-06, "loss": 0.813, "step": 24792 }, { "epoch": 1.9430250783699061, "grad_norm": 2.7198331356048584, "learning_rate": 5.239505633213346e-06, "loss": 1.5025, "step": 24793 }, { "epoch": 1.943103448275862, "grad_norm": 2.7598495483398438, "learning_rate": 5.239197624673711e-06, "loss": 1.3841, "step": 24794 }, { "epoch": 1.9431818181818183, "grad_norm": 1.9344456195831299, "learning_rate": 5.238889615224289e-06, "loss": 1.3739, "step": 24795 }, { "epoch": 1.9432601880877742, "grad_norm": 3.0409841537475586, "learning_rate": 5.238581604866255e-06, "loss": 1.2383, "step": 24796 }, { "epoch": 1.9433385579937306, "grad_norm": 2.131493091583252, "learning_rate": 5.23827359360078e-06, "loss": 1.4077, "step": 24797 }, { "epoch": 1.9434169278996865, "grad_norm": 2.4993133544921875, "learning_rate": 5.237965581429033e-06, "loss": 1.346, "step": 24798 }, { "epoch": 1.9434952978056428, "grad_norm": 4.203141212463379, "learning_rate": 5.237657568352188e-06, "loss": 1.2058, "step": 24799 }, { "epoch": 1.9435736677115987, "grad_norm": 6.381051540374756, "learning_rate": 5.237349554371416e-06, "loss": 2.0336, "step": 24800 }, { "epoch": 1.943652037617555, "grad_norm": 5.019135475158691, "learning_rate": 5.2370415394878885e-06, "loss": 0.9951, "step": 24801 }, { "epoch": 1.943730407523511, "grad_norm": 1.8598135709762573, "learning_rate": 5.236733523702775e-06, "loss": 1.2172, "step": 24802 }, { "epoch": 1.9438087774294672, "grad_norm": 4.099278926849365, "learning_rate": 5.236425507017251e-06, "loss": 1.902, "step": 24803 }, { "epoch": 1.9438871473354231, "grad_norm": 2.9439291954040527, "learning_rate": 5.236117489432484e-06, "loss": 1.3278, "step": 24804 }, { "epoch": 1.9439655172413794, "grad_norm": 2.5384786128997803, "learning_rate": 5.235809470949647e-06, "loss": 1.2372, "step": 24805 }, { "epoch": 1.9440438871473353, "grad_norm": 3.436345100402832, "learning_rate": 5.235501451569913e-06, "loss": 1.3448, "step": 24806 }, { "epoch": 1.9441222570532917, "grad_norm": 2.0793299674987793, "learning_rate": 5.235193431294453e-06, "loss": 1.4607, "step": 24807 }, { "epoch": 1.9442006269592476, "grad_norm": 2.0487864017486572, "learning_rate": 5.2348854101244375e-06, "loss": 1.2054, "step": 24808 }, { "epoch": 1.944278996865204, "grad_norm": 3.560590982437134, "learning_rate": 5.234577388061038e-06, "loss": 1.3014, "step": 24809 }, { "epoch": 1.9443573667711598, "grad_norm": 3.1060564517974854, "learning_rate": 5.234269365105427e-06, "loss": 1.1595, "step": 24810 }, { "epoch": 1.9444357366771161, "grad_norm": 5.982565879821777, "learning_rate": 5.233961341258775e-06, "loss": 1.951, "step": 24811 }, { "epoch": 1.944514106583072, "grad_norm": 4.113454818725586, "learning_rate": 5.233653316522256e-06, "loss": 1.7795, "step": 24812 }, { "epoch": 1.9445924764890283, "grad_norm": 6.1847825050354, "learning_rate": 5.233345290897038e-06, "loss": 1.0144, "step": 24813 }, { "epoch": 1.9446708463949842, "grad_norm": 4.023298263549805, "learning_rate": 5.233037264384296e-06, "loss": 1.9014, "step": 24814 }, { "epoch": 1.9447492163009406, "grad_norm": 3.072547674179077, "learning_rate": 5.232729236985198e-06, "loss": 1.1124, "step": 24815 }, { "epoch": 1.9448275862068964, "grad_norm": 1.9816555976867676, "learning_rate": 5.232421208700919e-06, "loss": 1.2239, "step": 24816 }, { "epoch": 1.9449059561128528, "grad_norm": 3.693676233291626, "learning_rate": 5.232113179532629e-06, "loss": 0.9483, "step": 24817 }, { "epoch": 1.9449843260188087, "grad_norm": 2.4897871017456055, "learning_rate": 5.231805149481498e-06, "loss": 1.6064, "step": 24818 }, { "epoch": 1.945062695924765, "grad_norm": 2.764552593231201, "learning_rate": 5.231497118548702e-06, "loss": 1.3412, "step": 24819 }, { "epoch": 1.9451410658307209, "grad_norm": 4.004270553588867, "learning_rate": 5.231189086735406e-06, "loss": 0.8753, "step": 24820 }, { "epoch": 1.9452194357366772, "grad_norm": 3.8671653270721436, "learning_rate": 5.230881054042789e-06, "loss": 1.3457, "step": 24821 }, { "epoch": 1.945297805642633, "grad_norm": 4.543491363525391, "learning_rate": 5.230573020472019e-06, "loss": 1.1467, "step": 24822 }, { "epoch": 1.9453761755485894, "grad_norm": 7.564594745635986, "learning_rate": 5.230264986024266e-06, "loss": 1.5392, "step": 24823 }, { "epoch": 1.9454545454545453, "grad_norm": 3.148879289627075, "learning_rate": 5.229956950700703e-06, "loss": 1.2095, "step": 24824 }, { "epoch": 1.9455329153605017, "grad_norm": 2.871558666229248, "learning_rate": 5.229648914502503e-06, "loss": 1.38, "step": 24825 }, { "epoch": 1.9456112852664575, "grad_norm": 3.6216421127319336, "learning_rate": 5.2293408774308365e-06, "loss": 1.2287, "step": 24826 }, { "epoch": 1.9456896551724139, "grad_norm": 3.762768030166626, "learning_rate": 5.229032839486874e-06, "loss": 1.2455, "step": 24827 }, { "epoch": 1.9457680250783698, "grad_norm": 3.711205244064331, "learning_rate": 5.228724800671789e-06, "loss": 1.0664, "step": 24828 }, { "epoch": 1.945846394984326, "grad_norm": 5.020691394805908, "learning_rate": 5.228416760986753e-06, "loss": 1.481, "step": 24829 }, { "epoch": 1.945924764890282, "grad_norm": 3.7252590656280518, "learning_rate": 5.228108720432936e-06, "loss": 1.3561, "step": 24830 }, { "epoch": 1.9460031347962383, "grad_norm": 3.9338669776916504, "learning_rate": 5.227800679011511e-06, "loss": 1.5091, "step": 24831 }, { "epoch": 1.9460815047021942, "grad_norm": 5.4159016609191895, "learning_rate": 5.22749263672365e-06, "loss": 1.1334, "step": 24832 }, { "epoch": 1.9461598746081505, "grad_norm": 2.5548288822174072, "learning_rate": 5.2271845935705225e-06, "loss": 1.1576, "step": 24833 }, { "epoch": 1.9462382445141064, "grad_norm": 2.9221999645233154, "learning_rate": 5.226876549553302e-06, "loss": 1.7484, "step": 24834 }, { "epoch": 1.9463166144200628, "grad_norm": 2.889540195465088, "learning_rate": 5.226568504673159e-06, "loss": 1.0417, "step": 24835 }, { "epoch": 1.9463949843260187, "grad_norm": 3.4056074619293213, "learning_rate": 5.226260458931267e-06, "loss": 1.6094, "step": 24836 }, { "epoch": 1.946473354231975, "grad_norm": 2.6660454273223877, "learning_rate": 5.2259524123287955e-06, "loss": 1.3967, "step": 24837 }, { "epoch": 1.9465517241379309, "grad_norm": 6.370896816253662, "learning_rate": 5.225644364866919e-06, "loss": 1.2734, "step": 24838 }, { "epoch": 1.9466300940438872, "grad_norm": 4.752179145812988, "learning_rate": 5.2253363165468044e-06, "loss": 1.2035, "step": 24839 }, { "epoch": 1.9467084639498433, "grad_norm": 2.5682449340820312, "learning_rate": 5.2250282673696275e-06, "loss": 1.6195, "step": 24840 }, { "epoch": 1.9467868338557994, "grad_norm": 13.027873992919922, "learning_rate": 5.2247202173365585e-06, "loss": 0.6827, "step": 24841 }, { "epoch": 1.9468652037617555, "grad_norm": 2.7609524726867676, "learning_rate": 5.224412166448769e-06, "loss": 1.3611, "step": 24842 }, { "epoch": 1.9469435736677116, "grad_norm": 2.816101551055908, "learning_rate": 5.224104114707431e-06, "loss": 1.2509, "step": 24843 }, { "epoch": 1.9470219435736678, "grad_norm": 2.3940629959106445, "learning_rate": 5.223796062113716e-06, "loss": 1.3492, "step": 24844 }, { "epoch": 1.9471003134796239, "grad_norm": 3.1995961666107178, "learning_rate": 5.2234880086687975e-06, "loss": 1.5243, "step": 24845 }, { "epoch": 1.94717868338558, "grad_norm": 2.974865436553955, "learning_rate": 5.223179954373842e-06, "loss": 1.2043, "step": 24846 }, { "epoch": 1.947257053291536, "grad_norm": 2.1231682300567627, "learning_rate": 5.222871899230027e-06, "loss": 1.2805, "step": 24847 }, { "epoch": 1.9473354231974922, "grad_norm": 2.650775909423828, "learning_rate": 5.222563843238522e-06, "loss": 1.4784, "step": 24848 }, { "epoch": 1.9474137931034483, "grad_norm": 2.424107789993286, "learning_rate": 5.222255786400498e-06, "loss": 1.259, "step": 24849 }, { "epoch": 1.9474921630094044, "grad_norm": 10.691422462463379, "learning_rate": 5.221947728717126e-06, "loss": 1.7337, "step": 24850 }, { "epoch": 1.9475705329153605, "grad_norm": 3.7797155380249023, "learning_rate": 5.22163967018958e-06, "loss": 1.9949, "step": 24851 }, { "epoch": 1.9476489028213166, "grad_norm": 5.18836784362793, "learning_rate": 5.2213316108190304e-06, "loss": 0.6859, "step": 24852 }, { "epoch": 1.9477272727272728, "grad_norm": 3.5890722274780273, "learning_rate": 5.221023550606649e-06, "loss": 1.3264, "step": 24853 }, { "epoch": 1.9478056426332289, "grad_norm": 4.038149356842041, "learning_rate": 5.220715489553606e-06, "loss": 1.0972, "step": 24854 }, { "epoch": 1.947884012539185, "grad_norm": 2.7739715576171875, "learning_rate": 5.220407427661076e-06, "loss": 1.3974, "step": 24855 }, { "epoch": 1.947962382445141, "grad_norm": 4.568730354309082, "learning_rate": 5.22009936493023e-06, "loss": 1.6204, "step": 24856 }, { "epoch": 1.9480407523510972, "grad_norm": 5.186561584472656, "learning_rate": 5.219791301362239e-06, "loss": 1.784, "step": 24857 }, { "epoch": 1.9481191222570533, "grad_norm": 3.5121734142303467, "learning_rate": 5.2194832369582724e-06, "loss": 1.5931, "step": 24858 }, { "epoch": 1.9481974921630094, "grad_norm": 2.827667474746704, "learning_rate": 5.2191751717195085e-06, "loss": 1.2912, "step": 24859 }, { "epoch": 1.9482758620689655, "grad_norm": 4.612046718597412, "learning_rate": 5.218867105647112e-06, "loss": 1.4385, "step": 24860 }, { "epoch": 1.9483542319749216, "grad_norm": 2.328209400177002, "learning_rate": 5.218559038742258e-06, "loss": 1.6768, "step": 24861 }, { "epoch": 1.9484326018808777, "grad_norm": 5.402577877044678, "learning_rate": 5.218250971006117e-06, "loss": 1.6477, "step": 24862 }, { "epoch": 1.9485109717868339, "grad_norm": 5.033542633056641, "learning_rate": 5.217942902439862e-06, "loss": 1.6537, "step": 24863 }, { "epoch": 1.94858934169279, "grad_norm": 2.3970065116882324, "learning_rate": 5.217634833044664e-06, "loss": 1.5166, "step": 24864 }, { "epoch": 1.948667711598746, "grad_norm": 3.543348789215088, "learning_rate": 5.217326762821695e-06, "loss": 1.2475, "step": 24865 }, { "epoch": 1.9487460815047022, "grad_norm": 2.865307331085205, "learning_rate": 5.217018691772128e-06, "loss": 1.3338, "step": 24866 }, { "epoch": 1.9488244514106583, "grad_norm": 2.9439327716827393, "learning_rate": 5.21671061989713e-06, "loss": 0.8183, "step": 24867 }, { "epoch": 1.9489028213166144, "grad_norm": 3.0660903453826904, "learning_rate": 5.216402547197879e-06, "loss": 0.9064, "step": 24868 }, { "epoch": 1.9489811912225705, "grad_norm": 2.592533588409424, "learning_rate": 5.216094473675542e-06, "loss": 1.3659, "step": 24869 }, { "epoch": 1.9490595611285266, "grad_norm": 2.6615986824035645, "learning_rate": 5.215786399331294e-06, "loss": 1.0814, "step": 24870 }, { "epoch": 1.9491379310344827, "grad_norm": 4.520389556884766, "learning_rate": 5.215478324166306e-06, "loss": 1.2634, "step": 24871 }, { "epoch": 1.9492163009404389, "grad_norm": 5.079651355743408, "learning_rate": 5.215170248181747e-06, "loss": 1.0196, "step": 24872 }, { "epoch": 1.949294670846395, "grad_norm": 3.3174710273742676, "learning_rate": 5.214862171378793e-06, "loss": 1.3951, "step": 24873 }, { "epoch": 1.949373040752351, "grad_norm": 3.94309663772583, "learning_rate": 5.2145540937586114e-06, "loss": 1.5063, "step": 24874 }, { "epoch": 1.9494514106583072, "grad_norm": 5.264861583709717, "learning_rate": 5.214246015322379e-06, "loss": 0.9259, "step": 24875 }, { "epoch": 1.9495297805642633, "grad_norm": 3.0873732566833496, "learning_rate": 5.213937936071263e-06, "loss": 1.6103, "step": 24876 }, { "epoch": 1.9496081504702194, "grad_norm": 2.3334081172943115, "learning_rate": 5.213629856006437e-06, "loss": 1.4997, "step": 24877 }, { "epoch": 1.9496865203761755, "grad_norm": 2.613431692123413, "learning_rate": 5.213321775129073e-06, "loss": 1.2016, "step": 24878 }, { "epoch": 1.9497648902821316, "grad_norm": 3.2655038833618164, "learning_rate": 5.213013693440342e-06, "loss": 1.701, "step": 24879 }, { "epoch": 1.9498432601880877, "grad_norm": 3.0910582542419434, "learning_rate": 5.212705610941417e-06, "loss": 1.0934, "step": 24880 }, { "epoch": 1.9499216300940438, "grad_norm": 2.2130162715911865, "learning_rate": 5.212397527633469e-06, "loss": 1.066, "step": 24881 }, { "epoch": 1.95, "grad_norm": 5.72467041015625, "learning_rate": 5.21208944351767e-06, "loss": 1.8152, "step": 24882 }, { "epoch": 1.950078369905956, "grad_norm": 4.630728721618652, "learning_rate": 5.2117813585951916e-06, "loss": 1.7153, "step": 24883 }, { "epoch": 1.9501567398119122, "grad_norm": 11.905736923217773, "learning_rate": 5.211473272867205e-06, "loss": 0.9715, "step": 24884 }, { "epoch": 1.9502351097178683, "grad_norm": 2.2904469966888428, "learning_rate": 5.2111651863348845e-06, "loss": 1.457, "step": 24885 }, { "epoch": 1.9503134796238244, "grad_norm": 3.4774813652038574, "learning_rate": 5.210857098999398e-06, "loss": 1.4526, "step": 24886 }, { "epoch": 1.9503918495297805, "grad_norm": 9.620189666748047, "learning_rate": 5.2105490108619204e-06, "loss": 1.6508, "step": 24887 }, { "epoch": 1.9504702194357368, "grad_norm": 6.477749347686768, "learning_rate": 5.210240921923623e-06, "loss": 1.032, "step": 24888 }, { "epoch": 1.9505485893416927, "grad_norm": 3.221571207046509, "learning_rate": 5.209932832185676e-06, "loss": 1.2994, "step": 24889 }, { "epoch": 1.950626959247649, "grad_norm": 4.302340507507324, "learning_rate": 5.209624741649254e-06, "loss": 1.1275, "step": 24890 }, { "epoch": 1.950705329153605, "grad_norm": 2.9818694591522217, "learning_rate": 5.209316650315524e-06, "loss": 1.2829, "step": 24891 }, { "epoch": 1.9507836990595613, "grad_norm": 4.589247703552246, "learning_rate": 5.209008558185664e-06, "loss": 1.0927, "step": 24892 }, { "epoch": 1.9508620689655172, "grad_norm": 2.0745036602020264, "learning_rate": 5.2087004652608405e-06, "loss": 1.2764, "step": 24893 }, { "epoch": 1.9509404388714735, "grad_norm": 3.8174617290496826, "learning_rate": 5.20839237154223e-06, "loss": 1.1898, "step": 24894 }, { "epoch": 1.9510188087774294, "grad_norm": 6.724817752838135, "learning_rate": 5.208084277031001e-06, "loss": 1.2172, "step": 24895 }, { "epoch": 1.9510971786833857, "grad_norm": 4.9523820877075195, "learning_rate": 5.207776181728327e-06, "loss": 1.4684, "step": 24896 }, { "epoch": 1.9511755485893416, "grad_norm": 4.232523441314697, "learning_rate": 5.207468085635378e-06, "loss": 1.7609, "step": 24897 }, { "epoch": 1.951253918495298, "grad_norm": 2.4656941890716553, "learning_rate": 5.2071599887533265e-06, "loss": 1.446, "step": 24898 }, { "epoch": 1.9513322884012538, "grad_norm": 3.7776858806610107, "learning_rate": 5.206851891083346e-06, "loss": 1.2727, "step": 24899 }, { "epoch": 1.9514106583072102, "grad_norm": 2.236992120742798, "learning_rate": 5.206543792626606e-06, "loss": 1.5868, "step": 24900 }, { "epoch": 1.951489028213166, "grad_norm": 3.109058380126953, "learning_rate": 5.206235693384282e-06, "loss": 1.4475, "step": 24901 }, { "epoch": 1.9515673981191224, "grad_norm": 4.618882656097412, "learning_rate": 5.205927593357541e-06, "loss": 1.3862, "step": 24902 }, { "epoch": 1.9516457680250783, "grad_norm": 3.8114802837371826, "learning_rate": 5.205619492547557e-06, "loss": 1.3012, "step": 24903 }, { "epoch": 1.9517241379310346, "grad_norm": 2.592994213104248, "learning_rate": 5.205311390955503e-06, "loss": 1.3283, "step": 24904 }, { "epoch": 1.9518025078369905, "grad_norm": 2.000706672668457, "learning_rate": 5.2050032885825505e-06, "loss": 1.3331, "step": 24905 }, { "epoch": 1.9518808777429468, "grad_norm": 4.477371692657471, "learning_rate": 5.20469518542987e-06, "loss": 1.7341, "step": 24906 }, { "epoch": 1.9519592476489027, "grad_norm": 3.0450708866119385, "learning_rate": 5.2043870814986345e-06, "loss": 1.453, "step": 24907 }, { "epoch": 1.952037617554859, "grad_norm": 2.4627838134765625, "learning_rate": 5.204078976790015e-06, "loss": 1.1355, "step": 24908 }, { "epoch": 1.952115987460815, "grad_norm": 3.218581438064575, "learning_rate": 5.203770871305183e-06, "loss": 1.7551, "step": 24909 }, { "epoch": 1.9521943573667713, "grad_norm": 4.6206231117248535, "learning_rate": 5.203462765045313e-06, "loss": 1.6906, "step": 24910 }, { "epoch": 1.9522727272727272, "grad_norm": 2.874800682067871, "learning_rate": 5.203154658011576e-06, "loss": 1.423, "step": 24911 }, { "epoch": 1.9523510971786835, "grad_norm": 3.7303287982940674, "learning_rate": 5.20284655020514e-06, "loss": 1.5046, "step": 24912 }, { "epoch": 1.9524294670846394, "grad_norm": 3.8235628604888916, "learning_rate": 5.202538441627182e-06, "loss": 1.7522, "step": 24913 }, { "epoch": 1.9525078369905957, "grad_norm": 6.451874732971191, "learning_rate": 5.2022303322788715e-06, "loss": 1.2109, "step": 24914 }, { "epoch": 1.9525862068965516, "grad_norm": 2.811818838119507, "learning_rate": 5.20192222216138e-06, "loss": 1.6075, "step": 24915 }, { "epoch": 1.952664576802508, "grad_norm": 6.852560043334961, "learning_rate": 5.20161411127588e-06, "loss": 1.4706, "step": 24916 }, { "epoch": 1.9527429467084638, "grad_norm": 2.988910436630249, "learning_rate": 5.201305999623545e-06, "loss": 1.2486, "step": 24917 }, { "epoch": 1.9528213166144202, "grad_norm": 5.334326267242432, "learning_rate": 5.200997887205545e-06, "loss": 2.0319, "step": 24918 }, { "epoch": 1.952899686520376, "grad_norm": 4.520298480987549, "learning_rate": 5.20068977402305e-06, "loss": 1.0409, "step": 24919 }, { "epoch": 1.9529780564263324, "grad_norm": 3.5247814655303955, "learning_rate": 5.200381660077237e-06, "loss": 1.2663, "step": 24920 }, { "epoch": 1.9530564263322883, "grad_norm": 4.326263904571533, "learning_rate": 5.200073545369273e-06, "loss": 1.2519, "step": 24921 }, { "epoch": 1.9531347962382446, "grad_norm": 2.9090030193328857, "learning_rate": 5.199765429900334e-06, "loss": 1.327, "step": 24922 }, { "epoch": 1.9532131661442005, "grad_norm": 3.6451642513275146, "learning_rate": 5.199457313671588e-06, "loss": 1.2414, "step": 24923 }, { "epoch": 1.9532915360501568, "grad_norm": 3.6832475662231445, "learning_rate": 5.19914919668421e-06, "loss": 1.7176, "step": 24924 }, { "epoch": 1.9533699059561127, "grad_norm": 3.51456618309021, "learning_rate": 5.198841078939371e-06, "loss": 1.5351, "step": 24925 }, { "epoch": 1.953448275862069, "grad_norm": 4.559899806976318, "learning_rate": 5.198532960438242e-06, "loss": 1.3098, "step": 24926 }, { "epoch": 1.953526645768025, "grad_norm": 2.048217535018921, "learning_rate": 5.198224841181995e-06, "loss": 1.3636, "step": 24927 }, { "epoch": 1.9536050156739813, "grad_norm": 2.763734817504883, "learning_rate": 5.197916721171803e-06, "loss": 1.4137, "step": 24928 }, { "epoch": 1.9536833855799371, "grad_norm": 2.8873002529144287, "learning_rate": 5.197608600408839e-06, "loss": 1.3307, "step": 24929 }, { "epoch": 1.9537617554858935, "grad_norm": 7.417638301849365, "learning_rate": 5.1973004788942716e-06, "loss": 0.8725, "step": 24930 }, { "epoch": 1.9538401253918494, "grad_norm": 2.939638137817383, "learning_rate": 5.1969923566292745e-06, "loss": 1.5389, "step": 24931 }, { "epoch": 1.9539184952978057, "grad_norm": 2.281031847000122, "learning_rate": 5.19668423361502e-06, "loss": 1.6103, "step": 24932 }, { "epoch": 1.9539968652037616, "grad_norm": 3.9471471309661865, "learning_rate": 5.19637610985268e-06, "loss": 1.6436, "step": 24933 }, { "epoch": 1.954075235109718, "grad_norm": 3.341170072555542, "learning_rate": 5.196067985343426e-06, "loss": 1.1926, "step": 24934 }, { "epoch": 1.954153605015674, "grad_norm": 7.4633259773254395, "learning_rate": 5.195759860088429e-06, "loss": 1.0054, "step": 24935 }, { "epoch": 1.9542319749216301, "grad_norm": 2.0009076595306396, "learning_rate": 5.195451734088863e-06, "loss": 1.27, "step": 24936 }, { "epoch": 1.9543103448275863, "grad_norm": 6.489184379577637, "learning_rate": 5.1951436073459e-06, "loss": 1.3056, "step": 24937 }, { "epoch": 1.9543887147335424, "grad_norm": 3.2393672466278076, "learning_rate": 5.194835479860709e-06, "loss": 1.0331, "step": 24938 }, { "epoch": 1.9544670846394985, "grad_norm": 7.627380847930908, "learning_rate": 5.194527351634466e-06, "loss": 0.7912, "step": 24939 }, { "epoch": 1.9545454545454546, "grad_norm": 4.73383092880249, "learning_rate": 5.1942192226683385e-06, "loss": 1.6158, "step": 24940 }, { "epoch": 1.9546238244514107, "grad_norm": 2.46279239654541, "learning_rate": 5.193911092963504e-06, "loss": 1.4756, "step": 24941 }, { "epoch": 1.9547021943573668, "grad_norm": 4.945756912231445, "learning_rate": 5.1936029625211295e-06, "loss": 0.8839, "step": 24942 }, { "epoch": 1.954780564263323, "grad_norm": 3.8082423210144043, "learning_rate": 5.193294831342388e-06, "loss": 1.569, "step": 24943 }, { "epoch": 1.954858934169279, "grad_norm": 3.941859483718872, "learning_rate": 5.192986699428454e-06, "loss": 1.4836, "step": 24944 }, { "epoch": 1.9549373040752351, "grad_norm": 5.084774971008301, "learning_rate": 5.192678566780496e-06, "loss": 1.4574, "step": 24945 }, { "epoch": 1.9550156739811912, "grad_norm": 2.9672656059265137, "learning_rate": 5.19237043339969e-06, "loss": 1.1198, "step": 24946 }, { "epoch": 1.9550940438871474, "grad_norm": 2.630574941635132, "learning_rate": 5.192062299287201e-06, "loss": 1.4698, "step": 24947 }, { "epoch": 1.9551724137931035, "grad_norm": 11.083925247192383, "learning_rate": 5.1917541644442105e-06, "loss": 0.9815, "step": 24948 }, { "epoch": 1.9552507836990596, "grad_norm": 5.327718257904053, "learning_rate": 5.191446028871884e-06, "loss": 0.723, "step": 24949 }, { "epoch": 1.9553291536050157, "grad_norm": 3.0290982723236084, "learning_rate": 5.191137892571394e-06, "loss": 1.276, "step": 24950 }, { "epoch": 1.9554075235109718, "grad_norm": 4.8899922370910645, "learning_rate": 5.1908297555439145e-06, "loss": 1.3837, "step": 24951 }, { "epoch": 1.955485893416928, "grad_norm": 5.994209289550781, "learning_rate": 5.190521617790617e-06, "loss": 1.164, "step": 24952 }, { "epoch": 1.955564263322884, "grad_norm": 2.587139368057251, "learning_rate": 5.190213479312672e-06, "loss": 1.3301, "step": 24953 }, { "epoch": 1.9556426332288401, "grad_norm": 3.8327534198760986, "learning_rate": 5.189905340111254e-06, "loss": 1.578, "step": 24954 }, { "epoch": 1.9557210031347962, "grad_norm": 2.1618130207061768, "learning_rate": 5.1895972001875325e-06, "loss": 1.6001, "step": 24955 }, { "epoch": 1.9557993730407524, "grad_norm": 3.247004270553589, "learning_rate": 5.189289059542683e-06, "loss": 1.7188, "step": 24956 }, { "epoch": 1.9558777429467085, "grad_norm": 3.7575249671936035, "learning_rate": 5.188980918177874e-06, "loss": 1.2024, "step": 24957 }, { "epoch": 1.9559561128526646, "grad_norm": 3.587135076522827, "learning_rate": 5.188672776094277e-06, "loss": 1.1984, "step": 24958 }, { "epoch": 1.9560344827586207, "grad_norm": 3.803316831588745, "learning_rate": 5.188364633293067e-06, "loss": 1.3944, "step": 24959 }, { "epoch": 1.9561128526645768, "grad_norm": 3.279357671737671, "learning_rate": 5.188056489775414e-06, "loss": 1.3552, "step": 24960 }, { "epoch": 1.956191222570533, "grad_norm": 2.0107476711273193, "learning_rate": 5.187748345542491e-06, "loss": 0.5872, "step": 24961 }, { "epoch": 1.956269592476489, "grad_norm": 2.343083381652832, "learning_rate": 5.187440200595471e-06, "loss": 1.24, "step": 24962 }, { "epoch": 1.9563479623824451, "grad_norm": 2.3003768920898438, "learning_rate": 5.187132054935522e-06, "loss": 1.1826, "step": 24963 }, { "epoch": 1.9564263322884012, "grad_norm": 2.324948310852051, "learning_rate": 5.186823908563821e-06, "loss": 1.2945, "step": 24964 }, { "epoch": 1.9565047021943573, "grad_norm": 4.014414310455322, "learning_rate": 5.1865157614815375e-06, "loss": 1.4375, "step": 24965 }, { "epoch": 1.9565830721003135, "grad_norm": 6.10013484954834, "learning_rate": 5.186207613689843e-06, "loss": 0.7151, "step": 24966 }, { "epoch": 1.9566614420062696, "grad_norm": 2.3342373371124268, "learning_rate": 5.185899465189912e-06, "loss": 1.152, "step": 24967 }, { "epoch": 1.9567398119122257, "grad_norm": 3.6343209743499756, "learning_rate": 5.185591315982912e-06, "loss": 1.3296, "step": 24968 }, { "epoch": 1.9568181818181818, "grad_norm": 4.883218765258789, "learning_rate": 5.185283166070021e-06, "loss": 1.5078, "step": 24969 }, { "epoch": 1.956896551724138, "grad_norm": 4.149533271789551, "learning_rate": 5.184975015452407e-06, "loss": 1.0622, "step": 24970 }, { "epoch": 1.956974921630094, "grad_norm": 3.557079792022705, "learning_rate": 5.1846668641312424e-06, "loss": 1.2172, "step": 24971 }, { "epoch": 1.9570532915360501, "grad_norm": 4.53836727142334, "learning_rate": 5.184358712107701e-06, "loss": 1.5647, "step": 24972 }, { "epoch": 1.9571316614420062, "grad_norm": 6.236188888549805, "learning_rate": 5.184050559382953e-06, "loss": 1.4071, "step": 24973 }, { "epoch": 1.9572100313479623, "grad_norm": 2.1664116382598877, "learning_rate": 5.183742405958172e-06, "loss": 1.3286, "step": 24974 }, { "epoch": 1.9572884012539185, "grad_norm": 2.1595964431762695, "learning_rate": 5.183434251834527e-06, "loss": 1.4005, "step": 24975 }, { "epoch": 1.9573667711598746, "grad_norm": 2.6905195713043213, "learning_rate": 5.183126097013195e-06, "loss": 1.6143, "step": 24976 }, { "epoch": 1.9574451410658307, "grad_norm": 3.0546505451202393, "learning_rate": 5.1828179414953435e-06, "loss": 1.3097, "step": 24977 }, { "epoch": 1.9575235109717868, "grad_norm": 2.1805286407470703, "learning_rate": 5.182509785282147e-06, "loss": 1.3171, "step": 24978 }, { "epoch": 1.957601880877743, "grad_norm": 3.2853074073791504, "learning_rate": 5.182201628374778e-06, "loss": 1.1488, "step": 24979 }, { "epoch": 1.957680250783699, "grad_norm": 2.23874568939209, "learning_rate": 5.181893470774406e-06, "loss": 0.958, "step": 24980 }, { "epoch": 1.9577586206896553, "grad_norm": 3.6261065006256104, "learning_rate": 5.1815853124822054e-06, "loss": 1.6289, "step": 24981 }, { "epoch": 1.9578369905956112, "grad_norm": 1.8455888032913208, "learning_rate": 5.181277153499347e-06, "loss": 1.2536, "step": 24982 }, { "epoch": 1.9579153605015676, "grad_norm": 2.07919979095459, "learning_rate": 5.1809689938270045e-06, "loss": 1.2926, "step": 24983 }, { "epoch": 1.9579937304075234, "grad_norm": 3.523939371109009, "learning_rate": 5.18066083346635e-06, "loss": 1.5557, "step": 24984 }, { "epoch": 1.9580721003134798, "grad_norm": 7.604977607727051, "learning_rate": 5.180352672418553e-06, "loss": 1.138, "step": 24985 }, { "epoch": 1.9581504702194357, "grad_norm": 2.291914939880371, "learning_rate": 5.180044510684787e-06, "loss": 1.4361, "step": 24986 }, { "epoch": 1.958228840125392, "grad_norm": 9.488127708435059, "learning_rate": 5.179736348266223e-06, "loss": 0.859, "step": 24987 }, { "epoch": 1.958307210031348, "grad_norm": 3.434972047805786, "learning_rate": 5.179428185164036e-06, "loss": 1.4108, "step": 24988 }, { "epoch": 1.9583855799373042, "grad_norm": 7.237236499786377, "learning_rate": 5.179120021379396e-06, "loss": 1.089, "step": 24989 }, { "epoch": 1.95846394984326, "grad_norm": 2.669700860977173, "learning_rate": 5.178811856913475e-06, "loss": 1.3446, "step": 24990 }, { "epoch": 1.9585423197492164, "grad_norm": 4.019262790679932, "learning_rate": 5.178503691767447e-06, "loss": 0.6205, "step": 24991 }, { "epoch": 1.9586206896551723, "grad_norm": 6.572936534881592, "learning_rate": 5.17819552594248e-06, "loss": 1.4581, "step": 24992 }, { "epoch": 1.9586990595611287, "grad_norm": 4.644083499908447, "learning_rate": 5.17788735943975e-06, "loss": 1.3599, "step": 24993 }, { "epoch": 1.9587774294670846, "grad_norm": 3.031719446182251, "learning_rate": 5.177579192260428e-06, "loss": 1.5858, "step": 24994 }, { "epoch": 1.9588557993730409, "grad_norm": 4.044564247131348, "learning_rate": 5.177271024405687e-06, "loss": 0.861, "step": 24995 }, { "epoch": 1.9589341692789968, "grad_norm": 5.8018693923950195, "learning_rate": 5.176962855876696e-06, "loss": 2.0394, "step": 24996 }, { "epoch": 1.959012539184953, "grad_norm": 9.610889434814453, "learning_rate": 5.1766546866746305e-06, "loss": 0.7676, "step": 24997 }, { "epoch": 1.959090909090909, "grad_norm": 3.1508431434631348, "learning_rate": 5.176346516800661e-06, "loss": 0.8145, "step": 24998 }, { "epoch": 1.9591692789968653, "grad_norm": 2.618518114089966, "learning_rate": 5.176038346255959e-06, "loss": 1.3812, "step": 24999 }, { "epoch": 1.9592476489028212, "grad_norm": 4.135262489318848, "learning_rate": 5.1757301750416996e-06, "loss": 1.4561, "step": 25000 }, { "epoch": 1.9593260188087775, "grad_norm": 3.0944690704345703, "learning_rate": 5.175422003159051e-06, "loss": 1.5403, "step": 25001 }, { "epoch": 1.9594043887147334, "grad_norm": 1.9440189599990845, "learning_rate": 5.17511383060919e-06, "loss": 1.5474, "step": 25002 }, { "epoch": 1.9594827586206898, "grad_norm": 1.861055612564087, "learning_rate": 5.174805657393283e-06, "loss": 1.1255, "step": 25003 }, { "epoch": 1.9595611285266457, "grad_norm": 3.9784035682678223, "learning_rate": 5.174497483512506e-06, "loss": 1.1284, "step": 25004 }, { "epoch": 1.959639498432602, "grad_norm": 2.255598545074463, "learning_rate": 5.174189308968029e-06, "loss": 1.2945, "step": 25005 }, { "epoch": 1.9597178683385579, "grad_norm": 4.892000198364258, "learning_rate": 5.1738811337610275e-06, "loss": 1.3094, "step": 25006 }, { "epoch": 1.9597962382445142, "grad_norm": 3.63023042678833, "learning_rate": 5.173572957892669e-06, "loss": 1.6758, "step": 25007 }, { "epoch": 1.95987460815047, "grad_norm": 5.620214462280273, "learning_rate": 5.17326478136413e-06, "loss": 1.4016, "step": 25008 }, { "epoch": 1.9599529780564264, "grad_norm": 6.081930160522461, "learning_rate": 5.172956604176581e-06, "loss": 0.8464, "step": 25009 }, { "epoch": 1.9600313479623823, "grad_norm": 6.732357025146484, "learning_rate": 5.172648426331193e-06, "loss": 0.8921, "step": 25010 }, { "epoch": 1.9601097178683387, "grad_norm": 2.363828420639038, "learning_rate": 5.17234024782914e-06, "loss": 1.2198, "step": 25011 }, { "epoch": 1.9601880877742945, "grad_norm": 2.447741985321045, "learning_rate": 5.1720320686715934e-06, "loss": 1.4852, "step": 25012 }, { "epoch": 1.9602664576802509, "grad_norm": 3.2724359035491943, "learning_rate": 5.171723888859723e-06, "loss": 1.4986, "step": 25013 }, { "epoch": 1.9603448275862068, "grad_norm": 3.1807072162628174, "learning_rate": 5.171415708394705e-06, "loss": 1.3594, "step": 25014 }, { "epoch": 1.960423197492163, "grad_norm": 2.8610308170318604, "learning_rate": 5.171107527277709e-06, "loss": 1.4846, "step": 25015 }, { "epoch": 1.960501567398119, "grad_norm": 3.6072099208831787, "learning_rate": 5.170799345509908e-06, "loss": 1.1389, "step": 25016 }, { "epoch": 1.9605799373040753, "grad_norm": 1.8868610858917236, "learning_rate": 5.170491163092474e-06, "loss": 1.0086, "step": 25017 }, { "epoch": 1.9606583072100312, "grad_norm": 2.6382813453674316, "learning_rate": 5.170182980026579e-06, "loss": 1.5719, "step": 25018 }, { "epoch": 1.9607366771159875, "grad_norm": 3.8956782817840576, "learning_rate": 5.169874796313397e-06, "loss": 1.4399, "step": 25019 }, { "epoch": 1.9608150470219434, "grad_norm": 4.627285003662109, "learning_rate": 5.169566611954095e-06, "loss": 1.8738, "step": 25020 }, { "epoch": 1.9608934169278998, "grad_norm": 3.246636152267456, "learning_rate": 5.169258426949852e-06, "loss": 1.4645, "step": 25021 }, { "epoch": 1.9609717868338556, "grad_norm": 2.3187286853790283, "learning_rate": 5.1689502413018345e-06, "loss": 1.1917, "step": 25022 }, { "epoch": 1.961050156739812, "grad_norm": 4.0850605964660645, "learning_rate": 5.1686420550112195e-06, "loss": 1.2723, "step": 25023 }, { "epoch": 1.9611285266457679, "grad_norm": 2.579010486602783, "learning_rate": 5.168333868079175e-06, "loss": 1.7008, "step": 25024 }, { "epoch": 1.9612068965517242, "grad_norm": 3.396542549133301, "learning_rate": 5.168025680506876e-06, "loss": 1.6261, "step": 25025 }, { "epoch": 1.96128526645768, "grad_norm": 2.5816402435302734, "learning_rate": 5.167717492295493e-06, "loss": 1.399, "step": 25026 }, { "epoch": 1.9613636363636364, "grad_norm": 3.6101787090301514, "learning_rate": 5.167409303446198e-06, "loss": 1.4121, "step": 25027 }, { "epoch": 1.9614420062695925, "grad_norm": 2.133117198944092, "learning_rate": 5.167101113960167e-06, "loss": 1.2824, "step": 25028 }, { "epoch": 1.9615203761755486, "grad_norm": 2.699610710144043, "learning_rate": 5.166792923838565e-06, "loss": 1.4974, "step": 25029 }, { "epoch": 1.9615987460815048, "grad_norm": 3.6510159969329834, "learning_rate": 5.166484733082572e-06, "loss": 1.6279, "step": 25030 }, { "epoch": 1.9616771159874609, "grad_norm": 3.106250762939453, "learning_rate": 5.166176541693354e-06, "loss": 1.4511, "step": 25031 }, { "epoch": 1.961755485893417, "grad_norm": 5.162753105163574, "learning_rate": 5.165868349672087e-06, "loss": 1.7107, "step": 25032 }, { "epoch": 1.961833855799373, "grad_norm": 2.7088234424591064, "learning_rate": 5.165560157019942e-06, "loss": 1.4598, "step": 25033 }, { "epoch": 1.9619122257053292, "grad_norm": 3.701296806335449, "learning_rate": 5.165251963738091e-06, "loss": 1.6432, "step": 25034 }, { "epoch": 1.9619905956112853, "grad_norm": 5.4829254150390625, "learning_rate": 5.164943769827706e-06, "loss": 0.5243, "step": 25035 }, { "epoch": 1.9620689655172414, "grad_norm": 2.196960210800171, "learning_rate": 5.16463557528996e-06, "loss": 1.38, "step": 25036 }, { "epoch": 1.9621473354231975, "grad_norm": 3.3871119022369385, "learning_rate": 5.164327380126024e-06, "loss": 1.6347, "step": 25037 }, { "epoch": 1.9622257053291536, "grad_norm": 3.920588254928589, "learning_rate": 5.1640191843370735e-06, "loss": 1.5187, "step": 25038 }, { "epoch": 1.9623040752351097, "grad_norm": 2.852564573287964, "learning_rate": 5.163710987924275e-06, "loss": 1.2814, "step": 25039 }, { "epoch": 1.9623824451410659, "grad_norm": 3.57082200050354, "learning_rate": 5.1634027908888075e-06, "loss": 1.1341, "step": 25040 }, { "epoch": 1.962460815047022, "grad_norm": 3.9384021759033203, "learning_rate": 5.163094593231838e-06, "loss": 1.7187, "step": 25041 }, { "epoch": 1.962539184952978, "grad_norm": 2.8692028522491455, "learning_rate": 5.1627863949545396e-06, "loss": 1.4304, "step": 25042 }, { "epoch": 1.9626175548589342, "grad_norm": 2.4303927421569824, "learning_rate": 5.162478196058086e-06, "loss": 1.456, "step": 25043 }, { "epoch": 1.9626959247648903, "grad_norm": 7.108695983886719, "learning_rate": 5.162169996543649e-06, "loss": 1.8582, "step": 25044 }, { "epoch": 1.9627742946708464, "grad_norm": 2.851398468017578, "learning_rate": 5.1618617964124e-06, "loss": 1.1567, "step": 25045 }, { "epoch": 1.9628526645768025, "grad_norm": 3.27311110496521, "learning_rate": 5.161553595665511e-06, "loss": 0.9726, "step": 25046 }, { "epoch": 1.9629310344827586, "grad_norm": 9.125344276428223, "learning_rate": 5.1612453943041585e-06, "loss": 1.6156, "step": 25047 }, { "epoch": 1.9630094043887147, "grad_norm": 2.337637424468994, "learning_rate": 5.160937192329508e-06, "loss": 1.1989, "step": 25048 }, { "epoch": 1.9630877742946709, "grad_norm": 5.335158824920654, "learning_rate": 5.160628989742737e-06, "loss": 1.6357, "step": 25049 }, { "epoch": 1.963166144200627, "grad_norm": 16.907421112060547, "learning_rate": 5.160320786545016e-06, "loss": 2.3508, "step": 25050 }, { "epoch": 1.963244514106583, "grad_norm": 6.073760986328125, "learning_rate": 5.160012582737516e-06, "loss": 1.5902, "step": 25051 }, { "epoch": 1.9633228840125392, "grad_norm": 2.935199022293091, "learning_rate": 5.15970437832141e-06, "loss": 1.4695, "step": 25052 }, { "epoch": 1.9634012539184953, "grad_norm": 4.413129806518555, "learning_rate": 5.1593961732978715e-06, "loss": 1.4042, "step": 25053 }, { "epoch": 1.9634796238244514, "grad_norm": 2.402117967605591, "learning_rate": 5.159087967668073e-06, "loss": 1.2213, "step": 25054 }, { "epoch": 1.9635579937304075, "grad_norm": 3.2678816318511963, "learning_rate": 5.158779761433182e-06, "loss": 1.0925, "step": 25055 }, { "epoch": 1.9636363636363636, "grad_norm": 2.482408285140991, "learning_rate": 5.158471554594377e-06, "loss": 1.4377, "step": 25056 }, { "epoch": 1.9637147335423197, "grad_norm": 4.752016544342041, "learning_rate": 5.158163347152828e-06, "loss": 1.2333, "step": 25057 }, { "epoch": 1.9637931034482758, "grad_norm": 9.825006484985352, "learning_rate": 5.157855139109705e-06, "loss": 1.7253, "step": 25058 }, { "epoch": 1.963871473354232, "grad_norm": 3.499372720718384, "learning_rate": 5.157546930466183e-06, "loss": 1.7578, "step": 25059 }, { "epoch": 1.963949843260188, "grad_norm": 4.217711925506592, "learning_rate": 5.157238721223433e-06, "loss": 1.4337, "step": 25060 }, { "epoch": 1.9640282131661442, "grad_norm": 2.54646635055542, "learning_rate": 5.1569305113826284e-06, "loss": 1.8229, "step": 25061 }, { "epoch": 1.9641065830721003, "grad_norm": 7.095046043395996, "learning_rate": 5.15662230094494e-06, "loss": 1.3222, "step": 25062 }, { "epoch": 1.9641849529780564, "grad_norm": 4.349380016326904, "learning_rate": 5.156314089911542e-06, "loss": 1.4723, "step": 25063 }, { "epoch": 1.9642633228840125, "grad_norm": 2.777318000793457, "learning_rate": 5.156005878283604e-06, "loss": 0.9524, "step": 25064 }, { "epoch": 1.9643416927899686, "grad_norm": 4.339447021484375, "learning_rate": 5.1556976660623e-06, "loss": 0.6653, "step": 25065 }, { "epoch": 1.9644200626959247, "grad_norm": 2.005506992340088, "learning_rate": 5.1553894532488045e-06, "loss": 1.5402, "step": 25066 }, { "epoch": 1.9644984326018808, "grad_norm": 3.004972219467163, "learning_rate": 5.155081239844284e-06, "loss": 0.8058, "step": 25067 }, { "epoch": 1.964576802507837, "grad_norm": 2.283912181854248, "learning_rate": 5.154773025849916e-06, "loss": 1.1316, "step": 25068 }, { "epoch": 1.964655172413793, "grad_norm": 2.916773557662964, "learning_rate": 5.15446481126687e-06, "loss": 1.4248, "step": 25069 }, { "epoch": 1.9647335423197492, "grad_norm": 2.6592631340026855, "learning_rate": 5.154156596096321e-06, "loss": 1.7386, "step": 25070 }, { "epoch": 1.9648119122257053, "grad_norm": 6.584362506866455, "learning_rate": 5.153848380339437e-06, "loss": 1.7285, "step": 25071 }, { "epoch": 1.9648902821316614, "grad_norm": 2.794527292251587, "learning_rate": 5.153540163997394e-06, "loss": 1.3989, "step": 25072 }, { "epoch": 1.9649686520376175, "grad_norm": 4.834535598754883, "learning_rate": 5.153231947071364e-06, "loss": 0.95, "step": 25073 }, { "epoch": 1.9650470219435736, "grad_norm": 8.630623817443848, "learning_rate": 5.152923729562516e-06, "loss": 2.1049, "step": 25074 }, { "epoch": 1.9651253918495297, "grad_norm": 3.1083486080169678, "learning_rate": 5.1526155114720265e-06, "loss": 1.3679, "step": 25075 }, { "epoch": 1.965203761755486, "grad_norm": 3.8609132766723633, "learning_rate": 5.152307292801066e-06, "loss": 1.328, "step": 25076 }, { "epoch": 1.965282131661442, "grad_norm": 3.348554849624634, "learning_rate": 5.151999073550806e-06, "loss": 1.243, "step": 25077 }, { "epoch": 1.9653605015673983, "grad_norm": 11.879929542541504, "learning_rate": 5.15169085372242e-06, "loss": 1.4434, "step": 25078 }, { "epoch": 1.9654388714733542, "grad_norm": 2.991119861602783, "learning_rate": 5.1513826333170795e-06, "loss": 1.6277, "step": 25079 }, { "epoch": 1.9655172413793105, "grad_norm": 3.003170967102051, "learning_rate": 5.151074412335958e-06, "loss": 1.1906, "step": 25080 }, { "epoch": 1.9655956112852664, "grad_norm": 2.666341781616211, "learning_rate": 5.150766190780225e-06, "loss": 1.4796, "step": 25081 }, { "epoch": 1.9656739811912227, "grad_norm": 3.2863821983337402, "learning_rate": 5.150457968651058e-06, "loss": 1.46, "step": 25082 }, { "epoch": 1.9657523510971786, "grad_norm": 3.5857973098754883, "learning_rate": 5.150149745949624e-06, "loss": 1.0979, "step": 25083 }, { "epoch": 1.965830721003135, "grad_norm": 2.467393398284912, "learning_rate": 5.1498415226770985e-06, "loss": 1.25, "step": 25084 }, { "epoch": 1.9659090909090908, "grad_norm": 3.142982244491577, "learning_rate": 5.149533298834653e-06, "loss": 1.5303, "step": 25085 }, { "epoch": 1.9659874608150472, "grad_norm": 2.028716564178467, "learning_rate": 5.149225074423459e-06, "loss": 1.2171, "step": 25086 }, { "epoch": 1.966065830721003, "grad_norm": 2.862107753753662, "learning_rate": 5.1489168494446885e-06, "loss": 1.6332, "step": 25087 }, { "epoch": 1.9661442006269594, "grad_norm": 2.646038293838501, "learning_rate": 5.148608623899516e-06, "loss": 1.542, "step": 25088 }, { "epoch": 1.9662225705329153, "grad_norm": 5.270538806915283, "learning_rate": 5.148300397789112e-06, "loss": 2.4683, "step": 25089 }, { "epoch": 1.9663009404388716, "grad_norm": 2.9752838611602783, "learning_rate": 5.1479921711146495e-06, "loss": 0.8675, "step": 25090 }, { "epoch": 1.9663793103448275, "grad_norm": 4.455269813537598, "learning_rate": 5.1476839438773006e-06, "loss": 1.5135, "step": 25091 }, { "epoch": 1.9664576802507838, "grad_norm": 2.7646124362945557, "learning_rate": 5.14737571607824e-06, "loss": 1.4405, "step": 25092 }, { "epoch": 1.9665360501567397, "grad_norm": 2.766510248184204, "learning_rate": 5.147067487718635e-06, "loss": 0.9791, "step": 25093 }, { "epoch": 1.966614420062696, "grad_norm": 2.4642279148101807, "learning_rate": 5.146759258799663e-06, "loss": 1.2074, "step": 25094 }, { "epoch": 1.966692789968652, "grad_norm": 2.2092442512512207, "learning_rate": 5.1464510293224924e-06, "loss": 1.3232, "step": 25095 }, { "epoch": 1.9667711598746083, "grad_norm": 2.7882304191589355, "learning_rate": 5.146142799288298e-06, "loss": 1.3439, "step": 25096 }, { "epoch": 1.9668495297805642, "grad_norm": 4.8378496170043945, "learning_rate": 5.145834568698252e-06, "loss": 1.2654, "step": 25097 }, { "epoch": 1.9669278996865205, "grad_norm": 2.6363682746887207, "learning_rate": 5.145526337553526e-06, "loss": 1.5696, "step": 25098 }, { "epoch": 1.9670062695924764, "grad_norm": 7.743223190307617, "learning_rate": 5.1452181058552934e-06, "loss": 0.9278, "step": 25099 }, { "epoch": 1.9670846394984327, "grad_norm": 2.4304187297821045, "learning_rate": 5.144909873604723e-06, "loss": 1.0919, "step": 25100 }, { "epoch": 1.9671630094043886, "grad_norm": 3.128587007522583, "learning_rate": 5.144601640802993e-06, "loss": 1.227, "step": 25101 }, { "epoch": 1.967241379310345, "grad_norm": 3.345771551132202, "learning_rate": 5.14429340745127e-06, "loss": 1.2565, "step": 25102 }, { "epoch": 1.9673197492163008, "grad_norm": 16.649015426635742, "learning_rate": 5.14398517355073e-06, "loss": 0.7832, "step": 25103 }, { "epoch": 1.9673981191222571, "grad_norm": 9.414231300354004, "learning_rate": 5.143676939102545e-06, "loss": 1.1889, "step": 25104 }, { "epoch": 1.967476489028213, "grad_norm": 2.2297096252441406, "learning_rate": 5.143368704107886e-06, "loss": 1.3045, "step": 25105 }, { "epoch": 1.9675548589341694, "grad_norm": 8.648735046386719, "learning_rate": 5.143060468567926e-06, "loss": 0.9897, "step": 25106 }, { "epoch": 1.9676332288401253, "grad_norm": 2.121422052383423, "learning_rate": 5.142752232483838e-06, "loss": 1.344, "step": 25107 }, { "epoch": 1.9677115987460816, "grad_norm": 2.814234733581543, "learning_rate": 5.142443995856794e-06, "loss": 1.3025, "step": 25108 }, { "epoch": 1.9677899686520375, "grad_norm": 3.2953007221221924, "learning_rate": 5.142135758687965e-06, "loss": 1.0135, "step": 25109 }, { "epoch": 1.9678683385579938, "grad_norm": 2.2420413494110107, "learning_rate": 5.141827520978526e-06, "loss": 1.3375, "step": 25110 }, { "epoch": 1.9679467084639497, "grad_norm": 3.2986361980438232, "learning_rate": 5.141519282729648e-06, "loss": 1.0542, "step": 25111 }, { "epoch": 1.968025078369906, "grad_norm": 3.15215802192688, "learning_rate": 5.141211043942502e-06, "loss": 1.024, "step": 25112 }, { "epoch": 1.968103448275862, "grad_norm": 2.974055528640747, "learning_rate": 5.140902804618263e-06, "loss": 1.3154, "step": 25113 }, { "epoch": 1.9681818181818183, "grad_norm": 4.521073818206787, "learning_rate": 5.140594564758101e-06, "loss": 1.2587, "step": 25114 }, { "epoch": 1.9682601880877741, "grad_norm": 3.517241954803467, "learning_rate": 5.1402863243631905e-06, "loss": 1.1098, "step": 25115 }, { "epoch": 1.9683385579937305, "grad_norm": 20.154953002929688, "learning_rate": 5.1399780834347025e-06, "loss": 1.2788, "step": 25116 }, { "epoch": 1.9684169278996864, "grad_norm": 3.283747911453247, "learning_rate": 5.139669841973809e-06, "loss": 1.389, "step": 25117 }, { "epoch": 1.9684952978056427, "grad_norm": 2.6257660388946533, "learning_rate": 5.139361599981684e-06, "loss": 0.8561, "step": 25118 }, { "epoch": 1.9685736677115986, "grad_norm": 2.484187364578247, "learning_rate": 5.139053357459498e-06, "loss": 1.1809, "step": 25119 }, { "epoch": 1.968652037617555, "grad_norm": 9.634330749511719, "learning_rate": 5.138745114408427e-06, "loss": 1.7581, "step": 25120 }, { "epoch": 1.9687304075235108, "grad_norm": 2.5379397869110107, "learning_rate": 5.138436870829637e-06, "loss": 1.2835, "step": 25121 }, { "epoch": 1.9688087774294671, "grad_norm": 3.5248618125915527, "learning_rate": 5.138128626724308e-06, "loss": 1.549, "step": 25122 }, { "epoch": 1.9688871473354232, "grad_norm": 7.2365217208862305, "learning_rate": 5.137820382093605e-06, "loss": 1.7636, "step": 25123 }, { "epoch": 1.9689655172413794, "grad_norm": 3.7281196117401123, "learning_rate": 5.1375121369387085e-06, "loss": 1.8202, "step": 25124 }, { "epoch": 1.9690438871473355, "grad_norm": 4.976504802703857, "learning_rate": 5.137203891260783e-06, "loss": 1.2682, "step": 25125 }, { "epoch": 1.9691222570532916, "grad_norm": 2.9693567752838135, "learning_rate": 5.136895645061005e-06, "loss": 1.0695, "step": 25126 }, { "epoch": 1.9692006269592477, "grad_norm": 4.223176956176758, "learning_rate": 5.136587398340548e-06, "loss": 1.6731, "step": 25127 }, { "epoch": 1.9692789968652038, "grad_norm": 6.438902378082275, "learning_rate": 5.136279151100581e-06, "loss": 1.7547, "step": 25128 }, { "epoch": 1.96935736677116, "grad_norm": 2.3839216232299805, "learning_rate": 5.135970903342279e-06, "loss": 1.4009, "step": 25129 }, { "epoch": 1.969435736677116, "grad_norm": 2.620375871658325, "learning_rate": 5.135662655066812e-06, "loss": 1.3434, "step": 25130 }, { "epoch": 1.9695141065830721, "grad_norm": 9.15445613861084, "learning_rate": 5.135354406275356e-06, "loss": 1.7104, "step": 25131 }, { "epoch": 1.9695924764890282, "grad_norm": 3.868373155593872, "learning_rate": 5.13504615696908e-06, "loss": 1.0032, "step": 25132 }, { "epoch": 1.9696708463949844, "grad_norm": 2.97055983543396, "learning_rate": 5.134737907149158e-06, "loss": 1.111, "step": 25133 }, { "epoch": 1.9697492163009405, "grad_norm": 2.909594774246216, "learning_rate": 5.134429656816762e-06, "loss": 1.275, "step": 25134 }, { "epoch": 1.9698275862068966, "grad_norm": 4.765994071960449, "learning_rate": 5.134121405973065e-06, "loss": 1.3879, "step": 25135 }, { "epoch": 1.9699059561128527, "grad_norm": 3.0396454334259033, "learning_rate": 5.1338131546192395e-06, "loss": 1.0427, "step": 25136 }, { "epoch": 1.9699843260188088, "grad_norm": 3.813526153564453, "learning_rate": 5.133504902756457e-06, "loss": 1.7425, "step": 25137 }, { "epoch": 1.970062695924765, "grad_norm": 2.2705726623535156, "learning_rate": 5.133196650385891e-06, "loss": 1.6533, "step": 25138 }, { "epoch": 1.970141065830721, "grad_norm": 3.0353589057922363, "learning_rate": 5.132888397508713e-06, "loss": 1.5222, "step": 25139 }, { "epoch": 1.9702194357366771, "grad_norm": 3.184201240539551, "learning_rate": 5.132580144126095e-06, "loss": 1.1261, "step": 25140 }, { "epoch": 1.9702978056426332, "grad_norm": 2.664479970932007, "learning_rate": 5.132271890239211e-06, "loss": 1.4458, "step": 25141 }, { "epoch": 1.9703761755485893, "grad_norm": 3.9609739780426025, "learning_rate": 5.131963635849232e-06, "loss": 0.7285, "step": 25142 }, { "epoch": 1.9704545454545455, "grad_norm": 3.1098809242248535, "learning_rate": 5.131655380957331e-06, "loss": 1.458, "step": 25143 }, { "epoch": 1.9705329153605016, "grad_norm": 2.7274765968322754, "learning_rate": 5.131347125564682e-06, "loss": 1.3928, "step": 25144 }, { "epoch": 1.9706112852664577, "grad_norm": 10.00522518157959, "learning_rate": 5.131038869672454e-06, "loss": 0.9523, "step": 25145 }, { "epoch": 1.9706896551724138, "grad_norm": 2.9554038047790527, "learning_rate": 5.130730613281825e-06, "loss": 1.5748, "step": 25146 }, { "epoch": 1.97076802507837, "grad_norm": 2.6128835678100586, "learning_rate": 5.1304223563939595e-06, "loss": 1.5389, "step": 25147 }, { "epoch": 1.970846394984326, "grad_norm": 3.45145583152771, "learning_rate": 5.130114099010037e-06, "loss": 1.212, "step": 25148 }, { "epoch": 1.9709247648902821, "grad_norm": 4.471240520477295, "learning_rate": 5.129805841131226e-06, "loss": 1.7507, "step": 25149 }, { "epoch": 1.9710031347962382, "grad_norm": 5.484139442443848, "learning_rate": 5.1294975827587015e-06, "loss": 1.5715, "step": 25150 }, { "epoch": 1.9710815047021943, "grad_norm": 3.61783766746521, "learning_rate": 5.1291893238936344e-06, "loss": 1.4894, "step": 25151 }, { "epoch": 1.9711598746081505, "grad_norm": 3.656717538833618, "learning_rate": 5.128881064537197e-06, "loss": 1.0561, "step": 25152 }, { "epoch": 1.9712382445141066, "grad_norm": 3.098846435546875, "learning_rate": 5.128572804690563e-06, "loss": 1.7422, "step": 25153 }, { "epoch": 1.9713166144200627, "grad_norm": 2.603938102722168, "learning_rate": 5.1282645443549015e-06, "loss": 1.4694, "step": 25154 }, { "epoch": 1.9713949843260188, "grad_norm": 4.315499305725098, "learning_rate": 5.127956283531391e-06, "loss": 1.064, "step": 25155 }, { "epoch": 1.971473354231975, "grad_norm": 2.7572126388549805, "learning_rate": 5.127648022221197e-06, "loss": 1.2629, "step": 25156 }, { "epoch": 1.971551724137931, "grad_norm": 14.025517463684082, "learning_rate": 5.1273397604255e-06, "loss": 1.7155, "step": 25157 }, { "epoch": 1.9716300940438871, "grad_norm": 3.235407829284668, "learning_rate": 5.127031498145465e-06, "loss": 1.4102, "step": 25158 }, { "epoch": 1.9717084639498432, "grad_norm": 5.1367926597595215, "learning_rate": 5.126723235382269e-06, "loss": 1.6458, "step": 25159 }, { "epoch": 1.9717868338557993, "grad_norm": 3.445035934448242, "learning_rate": 5.126414972137081e-06, "loss": 1.4295, "step": 25160 }, { "epoch": 1.9718652037617554, "grad_norm": 5.88496732711792, "learning_rate": 5.126106708411076e-06, "loss": 1.3115, "step": 25161 }, { "epoch": 1.9719435736677116, "grad_norm": 4.70039701461792, "learning_rate": 5.125798444205426e-06, "loss": 1.4825, "step": 25162 }, { "epoch": 1.9720219435736677, "grad_norm": 2.781745672225952, "learning_rate": 5.125490179521303e-06, "loss": 1.5615, "step": 25163 }, { "epoch": 1.9721003134796238, "grad_norm": 3.5271852016448975, "learning_rate": 5.1251819143598805e-06, "loss": 1.1405, "step": 25164 }, { "epoch": 1.97217868338558, "grad_norm": 2.059292793273926, "learning_rate": 5.124873648722329e-06, "loss": 1.3239, "step": 25165 }, { "epoch": 1.972257053291536, "grad_norm": 2.5398967266082764, "learning_rate": 5.124565382609825e-06, "loss": 1.9179, "step": 25166 }, { "epoch": 1.972335423197492, "grad_norm": 1.9843075275421143, "learning_rate": 5.124257116023537e-06, "loss": 1.3957, "step": 25167 }, { "epoch": 1.9724137931034482, "grad_norm": 2.854907512664795, "learning_rate": 5.123948848964637e-06, "loss": 1.2428, "step": 25168 }, { "epoch": 1.9724921630094046, "grad_norm": 3.2795352935791016, "learning_rate": 5.1236405814343e-06, "loss": 1.3359, "step": 25169 }, { "epoch": 1.9725705329153604, "grad_norm": 2.1562438011169434, "learning_rate": 5.123332313433699e-06, "loss": 1.3447, "step": 25170 }, { "epoch": 1.9726489028213168, "grad_norm": 3.383823871612549, "learning_rate": 5.123024044964004e-06, "loss": 1.2561, "step": 25171 }, { "epoch": 1.9727272727272727, "grad_norm": 4.22222375869751, "learning_rate": 5.122715776026389e-06, "loss": 1.591, "step": 25172 }, { "epoch": 1.972805642633229, "grad_norm": 2.622642755508423, "learning_rate": 5.122407506622027e-06, "loss": 1.4894, "step": 25173 }, { "epoch": 1.9728840125391849, "grad_norm": 3.4027369022369385, "learning_rate": 5.12209923675209e-06, "loss": 1.4326, "step": 25174 }, { "epoch": 1.9729623824451412, "grad_norm": 2.940481185913086, "learning_rate": 5.121790966417748e-06, "loss": 1.2415, "step": 25175 }, { "epoch": 1.973040752351097, "grad_norm": 2.7217540740966797, "learning_rate": 5.121482695620178e-06, "loss": 1.313, "step": 25176 }, { "epoch": 1.9731191222570534, "grad_norm": 5.222758769989014, "learning_rate": 5.121174424360548e-06, "loss": 1.4254, "step": 25177 }, { "epoch": 1.9731974921630093, "grad_norm": 2.7840018272399902, "learning_rate": 5.120866152640035e-06, "loss": 1.498, "step": 25178 }, { "epoch": 1.9732758620689657, "grad_norm": 2.3041887283325195, "learning_rate": 5.120557880459809e-06, "loss": 1.3483, "step": 25179 }, { "epoch": 1.9733542319749215, "grad_norm": 2.7433981895446777, "learning_rate": 5.1202496078210415e-06, "loss": 1.2851, "step": 25180 }, { "epoch": 1.9734326018808779, "grad_norm": 2.2641072273254395, "learning_rate": 5.119941334724907e-06, "loss": 1.5025, "step": 25181 }, { "epoch": 1.9735109717868338, "grad_norm": 4.54233980178833, "learning_rate": 5.119633061172577e-06, "loss": 1.4383, "step": 25182 }, { "epoch": 1.97358934169279, "grad_norm": 2.681922674179077, "learning_rate": 5.119324787165225e-06, "loss": 1.6493, "step": 25183 }, { "epoch": 1.973667711598746, "grad_norm": 10.45671272277832, "learning_rate": 5.11901651270402e-06, "loss": 1.0685, "step": 25184 }, { "epoch": 1.9737460815047023, "grad_norm": 3.888598680496216, "learning_rate": 5.118708237790141e-06, "loss": 1.3012, "step": 25185 }, { "epoch": 1.9738244514106582, "grad_norm": 2.9839038848876953, "learning_rate": 5.118399962424755e-06, "loss": 0.8327, "step": 25186 }, { "epoch": 1.9739028213166145, "grad_norm": 6.807955741882324, "learning_rate": 5.118091686609037e-06, "loss": 1.6135, "step": 25187 }, { "epoch": 1.9739811912225704, "grad_norm": 15.845337867736816, "learning_rate": 5.117783410344158e-06, "loss": 1.7824, "step": 25188 }, { "epoch": 1.9740595611285268, "grad_norm": 8.5415678024292, "learning_rate": 5.117475133631292e-06, "loss": 1.6291, "step": 25189 }, { "epoch": 1.9741379310344827, "grad_norm": 2.705723285675049, "learning_rate": 5.11716685647161e-06, "loss": 1.4981, "step": 25190 }, { "epoch": 1.974216300940439, "grad_norm": 3.8161520957946777, "learning_rate": 5.116858578866287e-06, "loss": 1.0557, "step": 25191 }, { "epoch": 1.9742946708463949, "grad_norm": 4.4670820236206055, "learning_rate": 5.116550300816493e-06, "loss": 0.8804, "step": 25192 }, { "epoch": 1.9743730407523512, "grad_norm": 2.856837749481201, "learning_rate": 5.1162420223234035e-06, "loss": 1.2901, "step": 25193 }, { "epoch": 1.974451410658307, "grad_norm": 3.1225838661193848, "learning_rate": 5.1159337433881864e-06, "loss": 1.5682, "step": 25194 }, { "epoch": 1.9745297805642634, "grad_norm": 3.537140130996704, "learning_rate": 5.115625464012018e-06, "loss": 1.6907, "step": 25195 }, { "epoch": 1.9746081504702193, "grad_norm": 2.591722249984741, "learning_rate": 5.115317184196069e-06, "loss": 1.5497, "step": 25196 }, { "epoch": 1.9746865203761756, "grad_norm": 4.797138690948486, "learning_rate": 5.115008903941514e-06, "loss": 1.7662, "step": 25197 }, { "epoch": 1.9747648902821315, "grad_norm": 2.8709423542022705, "learning_rate": 5.114700623249523e-06, "loss": 1.3645, "step": 25198 }, { "epoch": 1.9748432601880879, "grad_norm": 2.9958713054656982, "learning_rate": 5.114392342121271e-06, "loss": 1.5749, "step": 25199 }, { "epoch": 1.9749216300940438, "grad_norm": 1.858715534210205, "learning_rate": 5.114084060557929e-06, "loss": 1.0335, "step": 25200 }, { "epoch": 1.975, "grad_norm": 2.335430383682251, "learning_rate": 5.113775778560668e-06, "loss": 1.2888, "step": 25201 }, { "epoch": 1.975078369905956, "grad_norm": 14.694744110107422, "learning_rate": 5.113467496130664e-06, "loss": 1.136, "step": 25202 }, { "epoch": 1.9751567398119123, "grad_norm": 2.8178603649139404, "learning_rate": 5.113159213269087e-06, "loss": 1.64, "step": 25203 }, { "epoch": 1.9752351097178682, "grad_norm": 2.178549289703369, "learning_rate": 5.112850929977111e-06, "loss": 1.6326, "step": 25204 }, { "epoch": 1.9753134796238245, "grad_norm": 3.2875494956970215, "learning_rate": 5.1125426462559095e-06, "loss": 1.0664, "step": 25205 }, { "epoch": 1.9753918495297804, "grad_norm": 2.7699036598205566, "learning_rate": 5.1122343621066515e-06, "loss": 1.5231, "step": 25206 }, { "epoch": 1.9754702194357368, "grad_norm": 3.3187224864959717, "learning_rate": 5.111926077530513e-06, "loss": 1.6229, "step": 25207 }, { "epoch": 1.9755485893416926, "grad_norm": 3.1239826679229736, "learning_rate": 5.111617792528663e-06, "loss": 1.3277, "step": 25208 }, { "epoch": 1.975626959247649, "grad_norm": 6.783665180206299, "learning_rate": 5.111309507102279e-06, "loss": 1.0795, "step": 25209 }, { "epoch": 1.9757053291536049, "grad_norm": 4.39021635055542, "learning_rate": 5.111001221252528e-06, "loss": 0.9132, "step": 25210 }, { "epoch": 1.9757836990595612, "grad_norm": 3.8806042671203613, "learning_rate": 5.110692934980589e-06, "loss": 1.6079, "step": 25211 }, { "epoch": 1.975862068965517, "grad_norm": 2.7743263244628906, "learning_rate": 5.110384648287628e-06, "loss": 1.2755, "step": 25212 }, { "epoch": 1.9759404388714734, "grad_norm": 3.8911869525909424, "learning_rate": 5.110076361174821e-06, "loss": 1.6865, "step": 25213 }, { "epoch": 1.9760188087774293, "grad_norm": 16.48988914489746, "learning_rate": 5.1097680736433405e-06, "loss": 0.6074, "step": 25214 }, { "epoch": 1.9760971786833856, "grad_norm": 3.1400210857391357, "learning_rate": 5.109459785694359e-06, "loss": 1.4736, "step": 25215 }, { "epoch": 1.9761755485893417, "grad_norm": 4.053014755249023, "learning_rate": 5.109151497329048e-06, "loss": 1.4613, "step": 25216 }, { "epoch": 1.9762539184952979, "grad_norm": 2.8650176525115967, "learning_rate": 5.108843208548581e-06, "loss": 1.6131, "step": 25217 }, { "epoch": 1.976332288401254, "grad_norm": 3.935511827468872, "learning_rate": 5.10853491935413e-06, "loss": 1.4328, "step": 25218 }, { "epoch": 1.97641065830721, "grad_norm": 4.789463520050049, "learning_rate": 5.108226629746868e-06, "loss": 1.2962, "step": 25219 }, { "epoch": 1.9764890282131662, "grad_norm": 5.24176025390625, "learning_rate": 5.107918339727968e-06, "loss": 1.7363, "step": 25220 }, { "epoch": 1.9765673981191223, "grad_norm": 3.9032554626464844, "learning_rate": 5.107610049298602e-06, "loss": 1.722, "step": 25221 }, { "epoch": 1.9766457680250784, "grad_norm": 2.9942731857299805, "learning_rate": 5.107301758459942e-06, "loss": 1.4548, "step": 25222 }, { "epoch": 1.9767241379310345, "grad_norm": 8.782692909240723, "learning_rate": 5.106993467213162e-06, "loss": 1.8258, "step": 25223 }, { "epoch": 1.9768025078369906, "grad_norm": 3.7011184692382812, "learning_rate": 5.106685175559434e-06, "loss": 1.2414, "step": 25224 }, { "epoch": 1.9768808777429467, "grad_norm": 2.9414620399475098, "learning_rate": 5.106376883499929e-06, "loss": 1.5779, "step": 25225 }, { "epoch": 1.9769592476489029, "grad_norm": 3.2008578777313232, "learning_rate": 5.106068591035822e-06, "loss": 1.5094, "step": 25226 }, { "epoch": 1.977037617554859, "grad_norm": 4.0642409324646, "learning_rate": 5.105760298168285e-06, "loss": 1.0979, "step": 25227 }, { "epoch": 1.977115987460815, "grad_norm": 3.3311586380004883, "learning_rate": 5.10545200489849e-06, "loss": 1.2195, "step": 25228 }, { "epoch": 1.9771943573667712, "grad_norm": 3.222663640975952, "learning_rate": 5.105143711227608e-06, "loss": 1.4439, "step": 25229 }, { "epoch": 1.9772727272727273, "grad_norm": 3.597053289413452, "learning_rate": 5.104835417156816e-06, "loss": 1.0578, "step": 25230 }, { "epoch": 1.9773510971786834, "grad_norm": 8.031994819641113, "learning_rate": 5.104527122687282e-06, "loss": 1.1193, "step": 25231 }, { "epoch": 1.9774294670846395, "grad_norm": 3.661991596221924, "learning_rate": 5.104218827820182e-06, "loss": 1.7701, "step": 25232 }, { "epoch": 1.9775078369905956, "grad_norm": 2.7182722091674805, "learning_rate": 5.103910532556685e-06, "loss": 1.1663, "step": 25233 }, { "epoch": 1.9775862068965517, "grad_norm": 3.466252326965332, "learning_rate": 5.103602236897967e-06, "loss": 0.9656, "step": 25234 }, { "epoch": 1.9776645768025078, "grad_norm": 2.321167469024658, "learning_rate": 5.1032939408451995e-06, "loss": 1.4863, "step": 25235 }, { "epoch": 1.977742946708464, "grad_norm": 2.5056724548339844, "learning_rate": 5.102985644399554e-06, "loss": 0.9467, "step": 25236 }, { "epoch": 1.97782131661442, "grad_norm": 2.7801551818847656, "learning_rate": 5.102677347562206e-06, "loss": 1.2334, "step": 25237 }, { "epoch": 1.9778996865203762, "grad_norm": 7.664423942565918, "learning_rate": 5.102369050334324e-06, "loss": 1.2863, "step": 25238 }, { "epoch": 1.9779780564263323, "grad_norm": 3.52361798286438, "learning_rate": 5.102060752717084e-06, "loss": 1.3562, "step": 25239 }, { "epoch": 1.9780564263322884, "grad_norm": 4.155674934387207, "learning_rate": 5.101752454711657e-06, "loss": 1.4072, "step": 25240 }, { "epoch": 1.9781347962382445, "grad_norm": 2.7721636295318604, "learning_rate": 5.101444156319215e-06, "loss": 1.4568, "step": 25241 }, { "epoch": 1.9782131661442006, "grad_norm": 2.8951992988586426, "learning_rate": 5.101135857540932e-06, "loss": 1.5242, "step": 25242 }, { "epoch": 1.9782915360501567, "grad_norm": 2.258280038833618, "learning_rate": 5.1008275583779806e-06, "loss": 1.3784, "step": 25243 }, { "epoch": 1.9783699059561128, "grad_norm": 2.0127274990081787, "learning_rate": 5.100519258831532e-06, "loss": 1.2394, "step": 25244 }, { "epoch": 1.978448275862069, "grad_norm": 3.751973867416382, "learning_rate": 5.10021095890276e-06, "loss": 0.9576, "step": 25245 }, { "epoch": 1.978526645768025, "grad_norm": 2.2261767387390137, "learning_rate": 5.099902658592838e-06, "loss": 1.1077, "step": 25246 }, { "epoch": 1.9786050156739812, "grad_norm": 2.4127089977264404, "learning_rate": 5.099594357902938e-06, "loss": 1.2141, "step": 25247 }, { "epoch": 1.9786833855799373, "grad_norm": 2.9471395015716553, "learning_rate": 5.099286056834231e-06, "loss": 1.4301, "step": 25248 }, { "epoch": 1.9787617554858934, "grad_norm": 3.2239911556243896, "learning_rate": 5.098977755387892e-06, "loss": 1.6971, "step": 25249 }, { "epoch": 1.9788401253918495, "grad_norm": 2.7687156200408936, "learning_rate": 5.0986694535650904e-06, "loss": 1.4188, "step": 25250 }, { "epoch": 1.9789184952978056, "grad_norm": 2.6224775314331055, "learning_rate": 5.098361151367003e-06, "loss": 1.4659, "step": 25251 }, { "epoch": 1.9789968652037617, "grad_norm": 3.2719972133636475, "learning_rate": 5.0980528487947985e-06, "loss": 1.2169, "step": 25252 }, { "epoch": 1.9790752351097178, "grad_norm": 3.623044729232788, "learning_rate": 5.0977445458496524e-06, "loss": 1.479, "step": 25253 }, { "epoch": 1.979153605015674, "grad_norm": 2.6647961139678955, "learning_rate": 5.097436242532738e-06, "loss": 1.2661, "step": 25254 }, { "epoch": 1.97923197492163, "grad_norm": 3.177154302597046, "learning_rate": 5.097127938845222e-06, "loss": 1.397, "step": 25255 }, { "epoch": 1.9793103448275862, "grad_norm": 3.4253203868865967, "learning_rate": 5.0968196347882835e-06, "loss": 1.2953, "step": 25256 }, { "epoch": 1.9793887147335423, "grad_norm": 2.752135992050171, "learning_rate": 5.096511330363092e-06, "loss": 1.1186, "step": 25257 }, { "epoch": 1.9794670846394984, "grad_norm": 2.7929821014404297, "learning_rate": 5.096203025570823e-06, "loss": 1.2789, "step": 25258 }, { "epoch": 1.9795454545454545, "grad_norm": 3.3974802494049072, "learning_rate": 5.095894720412645e-06, "loss": 1.6092, "step": 25259 }, { "epoch": 1.9796238244514106, "grad_norm": 3.0378353595733643, "learning_rate": 5.095586414889734e-06, "loss": 1.4485, "step": 25260 }, { "epoch": 1.9797021943573667, "grad_norm": 3.8895020484924316, "learning_rate": 5.095278109003261e-06, "loss": 1.7462, "step": 25261 }, { "epoch": 1.979780564263323, "grad_norm": 2.6059188842773438, "learning_rate": 5.0949698027544e-06, "loss": 1.4935, "step": 25262 }, { "epoch": 1.979858934169279, "grad_norm": 4.239710807800293, "learning_rate": 5.09466149614432e-06, "loss": 0.8715, "step": 25263 }, { "epoch": 1.9799373040752353, "grad_norm": 3.897763252258301, "learning_rate": 5.0943531891741985e-06, "loss": 1.3514, "step": 25264 }, { "epoch": 1.9800156739811912, "grad_norm": 9.058754920959473, "learning_rate": 5.094044881845206e-06, "loss": 1.5854, "step": 25265 }, { "epoch": 1.9800940438871475, "grad_norm": 3.8335931301116943, "learning_rate": 5.093736574158513e-06, "loss": 1.4895, "step": 25266 }, { "epoch": 1.9801724137931034, "grad_norm": 2.323503255844116, "learning_rate": 5.0934282661152965e-06, "loss": 1.2063, "step": 25267 }, { "epoch": 1.9802507836990597, "grad_norm": 3.993501901626587, "learning_rate": 5.093119957716725e-06, "loss": 1.7368, "step": 25268 }, { "epoch": 1.9803291536050156, "grad_norm": 3.562122344970703, "learning_rate": 5.092811648963975e-06, "loss": 1.2082, "step": 25269 }, { "epoch": 1.980407523510972, "grad_norm": 2.562349557876587, "learning_rate": 5.092503339858216e-06, "loss": 1.5254, "step": 25270 }, { "epoch": 1.9804858934169278, "grad_norm": 3.3208279609680176, "learning_rate": 5.092195030400622e-06, "loss": 1.2606, "step": 25271 }, { "epoch": 1.9805642633228842, "grad_norm": 2.1017892360687256, "learning_rate": 5.091886720592365e-06, "loss": 1.2717, "step": 25272 }, { "epoch": 1.98064263322884, "grad_norm": 2.8940157890319824, "learning_rate": 5.0915784104346184e-06, "loss": 1.5802, "step": 25273 }, { "epoch": 1.9807210031347964, "grad_norm": 2.152625560760498, "learning_rate": 5.091270099928555e-06, "loss": 1.2871, "step": 25274 }, { "epoch": 1.9807993730407523, "grad_norm": 4.0936737060546875, "learning_rate": 5.090961789075347e-06, "loss": 1.672, "step": 25275 }, { "epoch": 1.9808777429467086, "grad_norm": 3.2989401817321777, "learning_rate": 5.0906534778761666e-06, "loss": 1.4371, "step": 25276 }, { "epoch": 1.9809561128526645, "grad_norm": 3.5625081062316895, "learning_rate": 5.090345166332188e-06, "loss": 1.3926, "step": 25277 }, { "epoch": 1.9810344827586208, "grad_norm": 2.2246358394622803, "learning_rate": 5.090036854444581e-06, "loss": 1.1586, "step": 25278 }, { "epoch": 1.9811128526645767, "grad_norm": 4.546529769897461, "learning_rate": 5.089728542214521e-06, "loss": 1.7277, "step": 25279 }, { "epoch": 1.981191222570533, "grad_norm": 3.5436654090881348, "learning_rate": 5.08942022964318e-06, "loss": 0.7038, "step": 25280 }, { "epoch": 1.981269592476489, "grad_norm": 3.2003233432769775, "learning_rate": 5.0891119167317295e-06, "loss": 0.6931, "step": 25281 }, { "epoch": 1.9813479623824453, "grad_norm": 3.270021915435791, "learning_rate": 5.088803603481345e-06, "loss": 1.6886, "step": 25282 }, { "epoch": 1.9814263322884011, "grad_norm": 1.9660401344299316, "learning_rate": 5.0884952898931935e-06, "loss": 1.22, "step": 25283 }, { "epoch": 1.9815047021943575, "grad_norm": 7.3859028816223145, "learning_rate": 5.088186975968456e-06, "loss": 1.4954, "step": 25284 }, { "epoch": 1.9815830721003134, "grad_norm": 2.9956576824188232, "learning_rate": 5.087878661708296e-06, "loss": 1.6136, "step": 25285 }, { "epoch": 1.9816614420062697, "grad_norm": 2.9430108070373535, "learning_rate": 5.087570347113894e-06, "loss": 1.1294, "step": 25286 }, { "epoch": 1.9817398119122256, "grad_norm": 3.0487160682678223, "learning_rate": 5.087262032186418e-06, "loss": 0.7423, "step": 25287 }, { "epoch": 1.981818181818182, "grad_norm": 4.145255088806152, "learning_rate": 5.086953716927042e-06, "loss": 0.8281, "step": 25288 }, { "epoch": 1.9818965517241378, "grad_norm": 2.5642335414886475, "learning_rate": 5.086645401336939e-06, "loss": 1.2358, "step": 25289 }, { "epoch": 1.9819749216300941, "grad_norm": 4.436520576477051, "learning_rate": 5.086337085417282e-06, "loss": 1.8303, "step": 25290 }, { "epoch": 1.98205329153605, "grad_norm": 3.1956145763397217, "learning_rate": 5.086028769169242e-06, "loss": 0.7798, "step": 25291 }, { "epoch": 1.9821316614420064, "grad_norm": 5.323418617248535, "learning_rate": 5.085720452593994e-06, "loss": 1.5562, "step": 25292 }, { "epoch": 1.9822100313479623, "grad_norm": 2.807419776916504, "learning_rate": 5.085412135692709e-06, "loss": 1.6383, "step": 25293 }, { "epoch": 1.9822884012539186, "grad_norm": 2.978271961212158, "learning_rate": 5.08510381846656e-06, "loss": 1.0179, "step": 25294 }, { "epoch": 1.9823667711598745, "grad_norm": 4.8989338874816895, "learning_rate": 5.084795500916719e-06, "loss": 0.8948, "step": 25295 }, { "epoch": 1.9824451410658308, "grad_norm": 3.5305604934692383, "learning_rate": 5.08448718304436e-06, "loss": 1.2065, "step": 25296 }, { "epoch": 1.9825235109717867, "grad_norm": 2.2976057529449463, "learning_rate": 5.084178864850655e-06, "loss": 1.3348, "step": 25297 }, { "epoch": 1.982601880877743, "grad_norm": 3.02618145942688, "learning_rate": 5.083870546336778e-06, "loss": 1.4584, "step": 25298 }, { "epoch": 1.982680250783699, "grad_norm": 6.809010982513428, "learning_rate": 5.0835622275038984e-06, "loss": 1.0834, "step": 25299 }, { "epoch": 1.9827586206896552, "grad_norm": 2.3190672397613525, "learning_rate": 5.083253908353193e-06, "loss": 1.2877, "step": 25300 }, { "epoch": 1.9828369905956111, "grad_norm": 2.763692855834961, "learning_rate": 5.082945588885833e-06, "loss": 1.1614, "step": 25301 }, { "epoch": 1.9829153605015675, "grad_norm": 6.180742263793945, "learning_rate": 5.082637269102987e-06, "loss": 0.8604, "step": 25302 }, { "epoch": 1.9829937304075234, "grad_norm": 2.960907220840454, "learning_rate": 5.082328949005835e-06, "loss": 1.2325, "step": 25303 }, { "epoch": 1.9830721003134797, "grad_norm": 3.5803637504577637, "learning_rate": 5.0820206285955434e-06, "loss": 1.4792, "step": 25304 }, { "epoch": 1.9831504702194356, "grad_norm": 5.870364665985107, "learning_rate": 5.081712307873289e-06, "loss": 1.0609, "step": 25305 }, { "epoch": 1.983228840125392, "grad_norm": 3.021376848220825, "learning_rate": 5.081403986840242e-06, "loss": 1.3992, "step": 25306 }, { "epoch": 1.9833072100313478, "grad_norm": 2.8819572925567627, "learning_rate": 5.081095665497577e-06, "loss": 1.3896, "step": 25307 }, { "epoch": 1.9833855799373041, "grad_norm": 3.421692371368408, "learning_rate": 5.080787343846466e-06, "loss": 1.4077, "step": 25308 }, { "epoch": 1.9834639498432602, "grad_norm": 2.881199359893799, "learning_rate": 5.080479021888079e-06, "loss": 1.6086, "step": 25309 }, { "epoch": 1.9835423197492164, "grad_norm": 3.7791101932525635, "learning_rate": 5.080170699623595e-06, "loss": 1.2684, "step": 25310 }, { "epoch": 1.9836206896551725, "grad_norm": 2.7148218154907227, "learning_rate": 5.079862377054179e-06, "loss": 1.3148, "step": 25311 }, { "epoch": 1.9836990595611286, "grad_norm": 3.6894514560699463, "learning_rate": 5.079554054181011e-06, "loss": 1.5475, "step": 25312 }, { "epoch": 1.9837774294670847, "grad_norm": 6.0557637214660645, "learning_rate": 5.079245731005258e-06, "loss": 1.4635, "step": 25313 }, { "epoch": 1.9838557993730408, "grad_norm": 3.9789578914642334, "learning_rate": 5.078937407528095e-06, "loss": 1.4985, "step": 25314 }, { "epoch": 1.983934169278997, "grad_norm": 2.5292136669158936, "learning_rate": 5.0786290837506945e-06, "loss": 1.1856, "step": 25315 }, { "epoch": 1.984012539184953, "grad_norm": 5.116399765014648, "learning_rate": 5.07832075967423e-06, "loss": 1.7433, "step": 25316 }, { "epoch": 1.9840909090909091, "grad_norm": 2.4271326065063477, "learning_rate": 5.078012435299874e-06, "loss": 1.3651, "step": 25317 }, { "epoch": 1.9841692789968652, "grad_norm": 2.4839670658111572, "learning_rate": 5.077704110628799e-06, "loss": 1.2417, "step": 25318 }, { "epoch": 1.9842476489028213, "grad_norm": 2.9100451469421387, "learning_rate": 5.077395785662177e-06, "loss": 1.46, "step": 25319 }, { "epoch": 1.9843260188087775, "grad_norm": 7.475956439971924, "learning_rate": 5.077087460401179e-06, "loss": 0.8334, "step": 25320 }, { "epoch": 1.9844043887147336, "grad_norm": 2.753793716430664, "learning_rate": 5.076779134846983e-06, "loss": 1.3203, "step": 25321 }, { "epoch": 1.9844827586206897, "grad_norm": 2.258598566055298, "learning_rate": 5.076470809000757e-06, "loss": 1.2456, "step": 25322 }, { "epoch": 1.9845611285266458, "grad_norm": 12.280494689941406, "learning_rate": 5.076162482863676e-06, "loss": 0.9902, "step": 25323 }, { "epoch": 1.984639498432602, "grad_norm": 3.3570737838745117, "learning_rate": 5.0758541564369116e-06, "loss": 1.0121, "step": 25324 }, { "epoch": 1.984717868338558, "grad_norm": 2.786172389984131, "learning_rate": 5.075545829721637e-06, "loss": 1.1888, "step": 25325 }, { "epoch": 1.9847962382445141, "grad_norm": 7.895569801330566, "learning_rate": 5.075237502719025e-06, "loss": 1.0238, "step": 25326 }, { "epoch": 1.9848746081504702, "grad_norm": 11.637457847595215, "learning_rate": 5.074929175430248e-06, "loss": 1.1037, "step": 25327 }, { "epoch": 1.9849529780564263, "grad_norm": 2.5810790061950684, "learning_rate": 5.074620847856479e-06, "loss": 1.1229, "step": 25328 }, { "epoch": 1.9850313479623825, "grad_norm": 4.223343849182129, "learning_rate": 5.074312519998892e-06, "loss": 1.4235, "step": 25329 }, { "epoch": 1.9851097178683386, "grad_norm": 3.5620243549346924, "learning_rate": 5.074004191858656e-06, "loss": 1.5877, "step": 25330 }, { "epoch": 1.9851880877742947, "grad_norm": 2.589972734451294, "learning_rate": 5.073695863436947e-06, "loss": 1.2831, "step": 25331 }, { "epoch": 1.9852664576802508, "grad_norm": 3.293840169906616, "learning_rate": 5.073387534734936e-06, "loss": 1.0469, "step": 25332 }, { "epoch": 1.985344827586207, "grad_norm": 9.202879905700684, "learning_rate": 5.073079205753798e-06, "loss": 1.0057, "step": 25333 }, { "epoch": 1.985423197492163, "grad_norm": 4.062912940979004, "learning_rate": 5.072770876494703e-06, "loss": 1.1981, "step": 25334 }, { "epoch": 1.9855015673981191, "grad_norm": 6.246265888214111, "learning_rate": 5.0724625469588265e-06, "loss": 1.6225, "step": 25335 }, { "epoch": 1.9855799373040752, "grad_norm": 2.7266573905944824, "learning_rate": 5.07215421714734e-06, "loss": 1.2875, "step": 25336 }, { "epoch": 1.9856583072100313, "grad_norm": 6.895386695861816, "learning_rate": 5.071845887061413e-06, "loss": 2.273, "step": 25337 }, { "epoch": 1.9857366771159874, "grad_norm": 3.082855463027954, "learning_rate": 5.071537556702224e-06, "loss": 1.3437, "step": 25338 }, { "epoch": 1.9858150470219436, "grad_norm": 3.2905075550079346, "learning_rate": 5.0712292260709405e-06, "loss": 1.2652, "step": 25339 }, { "epoch": 1.9858934169278997, "grad_norm": 2.1798558235168457, "learning_rate": 5.070920895168741e-06, "loss": 1.2093, "step": 25340 }, { "epoch": 1.9859717868338558, "grad_norm": 3.602478504180908, "learning_rate": 5.070612563996792e-06, "loss": 1.1363, "step": 25341 }, { "epoch": 1.986050156739812, "grad_norm": 10.1185941696167, "learning_rate": 5.070304232556269e-06, "loss": 1.1728, "step": 25342 }, { "epoch": 1.986128526645768, "grad_norm": 2.0370192527770996, "learning_rate": 5.0699959008483455e-06, "loss": 1.2289, "step": 25343 }, { "epoch": 1.986206896551724, "grad_norm": 3.8722753524780273, "learning_rate": 5.0696875688741945e-06, "loss": 1.675, "step": 25344 }, { "epoch": 1.9862852664576802, "grad_norm": 4.286161422729492, "learning_rate": 5.069379236634986e-06, "loss": 0.9209, "step": 25345 }, { "epoch": 1.9863636363636363, "grad_norm": 4.175210952758789, "learning_rate": 5.0690709041318955e-06, "loss": 1.9237, "step": 25346 }, { "epoch": 1.9864420062695924, "grad_norm": 4.701127529144287, "learning_rate": 5.068762571366096e-06, "loss": 0.7395, "step": 25347 }, { "epoch": 1.9865203761755486, "grad_norm": 2.1410000324249268, "learning_rate": 5.068454238338757e-06, "loss": 1.1971, "step": 25348 }, { "epoch": 1.9865987460815047, "grad_norm": 4.444962501525879, "learning_rate": 5.068145905051054e-06, "loss": 1.0849, "step": 25349 }, { "epoch": 1.9866771159874608, "grad_norm": 4.369148254394531, "learning_rate": 5.067837571504158e-06, "loss": 0.7109, "step": 25350 }, { "epoch": 1.9867554858934169, "grad_norm": 3.4579923152923584, "learning_rate": 5.067529237699245e-06, "loss": 1.574, "step": 25351 }, { "epoch": 1.986833855799373, "grad_norm": 2.4871652126312256, "learning_rate": 5.067220903637483e-06, "loss": 1.0634, "step": 25352 }, { "epoch": 1.986912225705329, "grad_norm": 5.019951820373535, "learning_rate": 5.066912569320048e-06, "loss": 2.1531, "step": 25353 }, { "epoch": 1.9869905956112852, "grad_norm": 5.0356950759887695, "learning_rate": 5.066604234748113e-06, "loss": 1.3477, "step": 25354 }, { "epoch": 1.9870689655172413, "grad_norm": 3.4615533351898193, "learning_rate": 5.06629589992285e-06, "loss": 1.7782, "step": 25355 }, { "epoch": 1.9871473354231974, "grad_norm": 3.030595064163208, "learning_rate": 5.065987564845428e-06, "loss": 1.4515, "step": 25356 }, { "epoch": 1.9872257053291538, "grad_norm": 2.8420841693878174, "learning_rate": 5.065679229517028e-06, "loss": 1.3348, "step": 25357 }, { "epoch": 1.9873040752351097, "grad_norm": 3.236558437347412, "learning_rate": 5.065370893938813e-06, "loss": 1.4159, "step": 25358 }, { "epoch": 1.987382445141066, "grad_norm": 3.2517313957214355, "learning_rate": 5.065062558111965e-06, "loss": 1.3907, "step": 25359 }, { "epoch": 1.9874608150470219, "grad_norm": 4.0719499588012695, "learning_rate": 5.06475422203765e-06, "loss": 1.2192, "step": 25360 }, { "epoch": 1.9875391849529782, "grad_norm": 4.007348537445068, "learning_rate": 5.064445885717044e-06, "loss": 1.5797, "step": 25361 }, { "epoch": 1.987617554858934, "grad_norm": 4.054492950439453, "learning_rate": 5.064137549151319e-06, "loss": 1.7322, "step": 25362 }, { "epoch": 1.9876959247648904, "grad_norm": 1.851348638534546, "learning_rate": 5.0638292123416476e-06, "loss": 1.0365, "step": 25363 }, { "epoch": 1.9877742946708463, "grad_norm": 3.121809482574463, "learning_rate": 5.063520875289203e-06, "loss": 1.6176, "step": 25364 }, { "epoch": 1.9878526645768027, "grad_norm": 2.0455362796783447, "learning_rate": 5.0632125379951554e-06, "loss": 1.0311, "step": 25365 }, { "epoch": 1.9879310344827585, "grad_norm": 3.6435298919677734, "learning_rate": 5.062904200460683e-06, "loss": 1.4163, "step": 25366 }, { "epoch": 1.9880094043887149, "grad_norm": 2.6342413425445557, "learning_rate": 5.062595862686954e-06, "loss": 1.5217, "step": 25367 }, { "epoch": 1.9880877742946708, "grad_norm": 4.487494468688965, "learning_rate": 5.062287524675143e-06, "loss": 1.2215, "step": 25368 }, { "epoch": 1.988166144200627, "grad_norm": 2.6753013134002686, "learning_rate": 5.0619791864264215e-06, "loss": 1.048, "step": 25369 }, { "epoch": 1.988244514106583, "grad_norm": 3.4080941677093506, "learning_rate": 5.061670847941963e-06, "loss": 1.4789, "step": 25370 }, { "epoch": 1.9883228840125393, "grad_norm": 3.8934621810913086, "learning_rate": 5.0613625092229406e-06, "loss": 1.8302, "step": 25371 }, { "epoch": 1.9884012539184952, "grad_norm": 2.361726760864258, "learning_rate": 5.061054170270526e-06, "loss": 1.0453, "step": 25372 }, { "epoch": 1.9884796238244515, "grad_norm": 2.7584969997406006, "learning_rate": 5.060745831085893e-06, "loss": 0.9164, "step": 25373 }, { "epoch": 1.9885579937304074, "grad_norm": 2.9561479091644287, "learning_rate": 5.0604374916702136e-06, "loss": 1.106, "step": 25374 }, { "epoch": 1.9886363636363638, "grad_norm": 3.725184202194214, "learning_rate": 5.060129152024663e-06, "loss": 1.2463, "step": 25375 }, { "epoch": 1.9887147335423196, "grad_norm": 3.5095932483673096, "learning_rate": 5.059820812150411e-06, "loss": 1.7849, "step": 25376 }, { "epoch": 1.988793103448276, "grad_norm": 2.9186887741088867, "learning_rate": 5.059512472048631e-06, "loss": 1.249, "step": 25377 }, { "epoch": 1.9888714733542319, "grad_norm": 2.62406587600708, "learning_rate": 5.059204131720495e-06, "loss": 1.0134, "step": 25378 }, { "epoch": 1.9889498432601882, "grad_norm": 2.6090033054351807, "learning_rate": 5.0588957911671785e-06, "loss": 1.4958, "step": 25379 }, { "epoch": 1.989028213166144, "grad_norm": 7.265153884887695, "learning_rate": 5.058587450389852e-06, "loss": 1.0772, "step": 25380 }, { "epoch": 1.9891065830721004, "grad_norm": 17.207927703857422, "learning_rate": 5.05827910938969e-06, "loss": 1.6371, "step": 25381 }, { "epoch": 1.9891849529780563, "grad_norm": 4.424388408660889, "learning_rate": 5.057970768167863e-06, "loss": 1.603, "step": 25382 }, { "epoch": 1.9892633228840126, "grad_norm": 3.4438977241516113, "learning_rate": 5.057662426725546e-06, "loss": 1.4462, "step": 25383 }, { "epoch": 1.9893416927899685, "grad_norm": 2.3619425296783447, "learning_rate": 5.057354085063909e-06, "loss": 1.0115, "step": 25384 }, { "epoch": 1.9894200626959249, "grad_norm": 3.402846097946167, "learning_rate": 5.057045743184129e-06, "loss": 1.3155, "step": 25385 }, { "epoch": 1.9894984326018808, "grad_norm": 8.128303527832031, "learning_rate": 5.056737401087374e-06, "loss": 1.1662, "step": 25386 }, { "epoch": 1.989576802507837, "grad_norm": 3.7530927658081055, "learning_rate": 5.056429058774821e-06, "loss": 1.0628, "step": 25387 }, { "epoch": 1.989655172413793, "grad_norm": 4.35321044921875, "learning_rate": 5.0561207162476385e-06, "loss": 1.9983, "step": 25388 }, { "epoch": 1.9897335423197493, "grad_norm": 4.2756757736206055, "learning_rate": 5.055812373507004e-06, "loss": 1.9239, "step": 25389 }, { "epoch": 1.9898119122257052, "grad_norm": 3.1537392139434814, "learning_rate": 5.055504030554088e-06, "loss": 1.395, "step": 25390 }, { "epoch": 1.9898902821316615, "grad_norm": 3.939520835876465, "learning_rate": 5.05519568739006e-06, "loss": 1.2652, "step": 25391 }, { "epoch": 1.9899686520376174, "grad_norm": 3.274806499481201, "learning_rate": 5.0548873440161e-06, "loss": 1.6457, "step": 25392 }, { "epoch": 1.9900470219435737, "grad_norm": 1.999964952468872, "learning_rate": 5.0545790004333725e-06, "loss": 1.2154, "step": 25393 }, { "epoch": 1.9901253918495296, "grad_norm": 2.098827600479126, "learning_rate": 5.0542706566430576e-06, "loss": 1.1234, "step": 25394 }, { "epoch": 1.990203761755486, "grad_norm": 2.091487169265747, "learning_rate": 5.053962312646324e-06, "loss": 1.5574, "step": 25395 }, { "epoch": 1.9902821316614419, "grad_norm": 7.240144729614258, "learning_rate": 5.053653968444346e-06, "loss": 1.0966, "step": 25396 }, { "epoch": 1.9903605015673982, "grad_norm": 4.171501636505127, "learning_rate": 5.053345624038295e-06, "loss": 1.5944, "step": 25397 }, { "epoch": 1.990438871473354, "grad_norm": 2.0796456336975098, "learning_rate": 5.053037279429346e-06, "loss": 1.1484, "step": 25398 }, { "epoch": 1.9905172413793104, "grad_norm": 9.092097282409668, "learning_rate": 5.0527289346186695e-06, "loss": 2.1229, "step": 25399 }, { "epoch": 1.9905956112852663, "grad_norm": 1.993830919265747, "learning_rate": 5.052420589607439e-06, "loss": 1.1483, "step": 25400 }, { "epoch": 1.9906739811912226, "grad_norm": 2.1799230575561523, "learning_rate": 5.0521122443968276e-06, "loss": 1.127, "step": 25401 }, { "epoch": 1.9907523510971785, "grad_norm": 6.0292649269104, "learning_rate": 5.051803898988009e-06, "loss": 1.4287, "step": 25402 }, { "epoch": 1.9908307210031349, "grad_norm": 2.809769630432129, "learning_rate": 5.051495553382154e-06, "loss": 1.5407, "step": 25403 }, { "epoch": 1.990909090909091, "grad_norm": 4.852634906768799, "learning_rate": 5.051187207580436e-06, "loss": 0.8613, "step": 25404 }, { "epoch": 1.990987460815047, "grad_norm": 2.7231907844543457, "learning_rate": 5.050878861584029e-06, "loss": 1.209, "step": 25405 }, { "epoch": 1.9910658307210032, "grad_norm": 2.252549886703491, "learning_rate": 5.050570515394104e-06, "loss": 1.3444, "step": 25406 }, { "epoch": 1.9911442006269593, "grad_norm": 4.404833793640137, "learning_rate": 5.050262169011836e-06, "loss": 1.547, "step": 25407 }, { "epoch": 1.9912225705329154, "grad_norm": 2.0077528953552246, "learning_rate": 5.049953822438395e-06, "loss": 1.1918, "step": 25408 }, { "epoch": 1.9913009404388715, "grad_norm": 5.6183600425720215, "learning_rate": 5.049645475674958e-06, "loss": 1.5023, "step": 25409 }, { "epoch": 1.9913793103448276, "grad_norm": 6.91878604888916, "learning_rate": 5.049337128722691e-06, "loss": 1.0233, "step": 25410 }, { "epoch": 1.9914576802507837, "grad_norm": 6.418910980224609, "learning_rate": 5.049028781582774e-06, "loss": 1.2702, "step": 25411 }, { "epoch": 1.9915360501567398, "grad_norm": 2.272463321685791, "learning_rate": 5.0487204342563755e-06, "loss": 1.7824, "step": 25412 }, { "epoch": 1.991614420062696, "grad_norm": 2.979193925857544, "learning_rate": 5.0484120867446705e-06, "loss": 1.7466, "step": 25413 }, { "epoch": 1.991692789968652, "grad_norm": 3.3366665840148926, "learning_rate": 5.048103739048831e-06, "loss": 1.1822, "step": 25414 }, { "epoch": 1.9917711598746082, "grad_norm": 2.8539466857910156, "learning_rate": 5.047795391170028e-06, "loss": 1.3612, "step": 25415 }, { "epoch": 1.9918495297805643, "grad_norm": 1.9989886283874512, "learning_rate": 5.047487043109437e-06, "loss": 1.6269, "step": 25416 }, { "epoch": 1.9919278996865204, "grad_norm": 3.1269190311431885, "learning_rate": 5.047178694868229e-06, "loss": 1.448, "step": 25417 }, { "epoch": 1.9920062695924765, "grad_norm": 2.2943854331970215, "learning_rate": 5.046870346447579e-06, "loss": 1.2132, "step": 25418 }, { "epoch": 1.9920846394984326, "grad_norm": 2.388702630996704, "learning_rate": 5.046561997848656e-06, "loss": 0.9172, "step": 25419 }, { "epoch": 1.9921630094043887, "grad_norm": 4.131557941436768, "learning_rate": 5.046253649072637e-06, "loss": 1.2208, "step": 25420 }, { "epoch": 1.9922413793103448, "grad_norm": 2.4727540016174316, "learning_rate": 5.04594530012069e-06, "loss": 1.2889, "step": 25421 }, { "epoch": 1.992319749216301, "grad_norm": 3.1829540729522705, "learning_rate": 5.045636950993994e-06, "loss": 1.3407, "step": 25422 }, { "epoch": 1.992398119122257, "grad_norm": 2.2293269634246826, "learning_rate": 5.045328601693717e-06, "loss": 1.0465, "step": 25423 }, { "epoch": 1.9924764890282132, "grad_norm": 2.700850009918213, "learning_rate": 5.045020252221033e-06, "loss": 1.3953, "step": 25424 }, { "epoch": 1.9925548589341693, "grad_norm": 3.4797980785369873, "learning_rate": 5.044711902577116e-06, "loss": 1.2277, "step": 25425 }, { "epoch": 1.9926332288401254, "grad_norm": 4.195002555847168, "learning_rate": 5.044403552763137e-06, "loss": 1.2123, "step": 25426 }, { "epoch": 1.9927115987460815, "grad_norm": 3.5348358154296875, "learning_rate": 5.0440952027802705e-06, "loss": 1.8912, "step": 25427 }, { "epoch": 1.9927899686520376, "grad_norm": 5.6387763023376465, "learning_rate": 5.043786852629689e-06, "loss": 1.0438, "step": 25428 }, { "epoch": 1.9928683385579937, "grad_norm": 4.261425495147705, "learning_rate": 5.043478502312563e-06, "loss": 1.5329, "step": 25429 }, { "epoch": 1.9929467084639498, "grad_norm": 2.3509113788604736, "learning_rate": 5.043170151830069e-06, "loss": 1.4243, "step": 25430 }, { "epoch": 1.993025078369906, "grad_norm": 2.307008743286133, "learning_rate": 5.042861801183376e-06, "loss": 0.9193, "step": 25431 }, { "epoch": 1.993103448275862, "grad_norm": 3.3362081050872803, "learning_rate": 5.04255345037366e-06, "loss": 1.5798, "step": 25432 }, { "epoch": 1.9931818181818182, "grad_norm": 4.169219017028809, "learning_rate": 5.0422450994020925e-06, "loss": 1.4605, "step": 25433 }, { "epoch": 1.9932601880877743, "grad_norm": 2.7779717445373535, "learning_rate": 5.041936748269846e-06, "loss": 1.217, "step": 25434 }, { "epoch": 1.9933385579937304, "grad_norm": 6.483877658843994, "learning_rate": 5.041628396978093e-06, "loss": 1.0797, "step": 25435 }, { "epoch": 1.9934169278996865, "grad_norm": 3.159148931503296, "learning_rate": 5.041320045528007e-06, "loss": 1.4624, "step": 25436 }, { "epoch": 1.9934952978056426, "grad_norm": 2.0285611152648926, "learning_rate": 5.041011693920763e-06, "loss": 1.2152, "step": 25437 }, { "epoch": 1.9935736677115987, "grad_norm": 11.21717357635498, "learning_rate": 5.040703342157529e-06, "loss": 1.7613, "step": 25438 }, { "epoch": 1.9936520376175548, "grad_norm": 3.1243958473205566, "learning_rate": 5.040394990239482e-06, "loss": 1.5145, "step": 25439 }, { "epoch": 1.993730407523511, "grad_norm": 3.3147478103637695, "learning_rate": 5.0400866381677916e-06, "loss": 1.4481, "step": 25440 }, { "epoch": 1.993808777429467, "grad_norm": 3.630007266998291, "learning_rate": 5.039778285943635e-06, "loss": 1.5943, "step": 25441 }, { "epoch": 1.9938871473354232, "grad_norm": 4.334168434143066, "learning_rate": 5.039469933568179e-06, "loss": 1.1746, "step": 25442 }, { "epoch": 1.9939655172413793, "grad_norm": 7.186563491821289, "learning_rate": 5.039161581042602e-06, "loss": 1.1624, "step": 25443 }, { "epoch": 1.9940438871473354, "grad_norm": 2.2201550006866455, "learning_rate": 5.038853228368073e-06, "loss": 0.9375, "step": 25444 }, { "epoch": 1.9941222570532915, "grad_norm": 5.789028167724609, "learning_rate": 5.038544875545767e-06, "loss": 1.345, "step": 25445 }, { "epoch": 1.9942006269592476, "grad_norm": 2.78533673286438, "learning_rate": 5.038236522576857e-06, "loss": 1.423, "step": 25446 }, { "epoch": 1.9942789968652037, "grad_norm": 4.732938766479492, "learning_rate": 5.037928169462513e-06, "loss": 0.8861, "step": 25447 }, { "epoch": 1.9943573667711598, "grad_norm": 3.2228403091430664, "learning_rate": 5.037619816203911e-06, "loss": 1.4037, "step": 25448 }, { "epoch": 1.994435736677116, "grad_norm": 2.7317304611206055, "learning_rate": 5.037311462802223e-06, "loss": 1.2831, "step": 25449 }, { "epoch": 1.9945141065830723, "grad_norm": 6.1033453941345215, "learning_rate": 5.037003109258619e-06, "loss": 1.4419, "step": 25450 }, { "epoch": 1.9945924764890282, "grad_norm": 2.803584337234497, "learning_rate": 5.036694755574277e-06, "loss": 1.4604, "step": 25451 }, { "epoch": 1.9946708463949845, "grad_norm": 3.941903829574585, "learning_rate": 5.036386401750365e-06, "loss": 1.4102, "step": 25452 }, { "epoch": 1.9947492163009404, "grad_norm": 2.492781400680542, "learning_rate": 5.036078047788058e-06, "loss": 0.9655, "step": 25453 }, { "epoch": 1.9948275862068967, "grad_norm": 5.0550689697265625, "learning_rate": 5.0357696936885296e-06, "loss": 1.6625, "step": 25454 }, { "epoch": 1.9949059561128526, "grad_norm": 3.0103347301483154, "learning_rate": 5.035461339452951e-06, "loss": 1.3666, "step": 25455 }, { "epoch": 1.994984326018809, "grad_norm": 2.1748130321502686, "learning_rate": 5.0351529850824965e-06, "loss": 1.4032, "step": 25456 }, { "epoch": 1.9950626959247648, "grad_norm": 2.04476261138916, "learning_rate": 5.034844630578336e-06, "loss": 1.0562, "step": 25457 }, { "epoch": 1.9951410658307211, "grad_norm": 2.2661335468292236, "learning_rate": 5.034536275941647e-06, "loss": 1.408, "step": 25458 }, { "epoch": 1.995219435736677, "grad_norm": 2.006193161010742, "learning_rate": 5.034227921173598e-06, "loss": 1.2736, "step": 25459 }, { "epoch": 1.9952978056426334, "grad_norm": 2.786972999572754, "learning_rate": 5.033919566275365e-06, "loss": 1.4407, "step": 25460 }, { "epoch": 1.9953761755485893, "grad_norm": 2.810952663421631, "learning_rate": 5.0336112112481185e-06, "loss": 1.3805, "step": 25461 }, { "epoch": 1.9954545454545456, "grad_norm": 3.3297202587127686, "learning_rate": 5.033302856093032e-06, "loss": 1.2937, "step": 25462 }, { "epoch": 1.9955329153605015, "grad_norm": 2.5425853729248047, "learning_rate": 5.032994500811279e-06, "loss": 1.3876, "step": 25463 }, { "epoch": 1.9956112852664578, "grad_norm": 8.722970962524414, "learning_rate": 5.032686145404031e-06, "loss": 1.9582, "step": 25464 }, { "epoch": 1.9956896551724137, "grad_norm": 3.769634485244751, "learning_rate": 5.0323777898724635e-06, "loss": 1.285, "step": 25465 }, { "epoch": 1.99576802507837, "grad_norm": 6.408098220825195, "learning_rate": 5.032069434217745e-06, "loss": 1.6504, "step": 25466 }, { "epoch": 1.995846394984326, "grad_norm": 2.9135053157806396, "learning_rate": 5.031761078441053e-06, "loss": 1.4504, "step": 25467 }, { "epoch": 1.9959247648902823, "grad_norm": 2.8303112983703613, "learning_rate": 5.031452722543557e-06, "loss": 1.5918, "step": 25468 }, { "epoch": 1.9960031347962381, "grad_norm": 2.526762008666992, "learning_rate": 5.031144366526431e-06, "loss": 1.4744, "step": 25469 }, { "epoch": 1.9960815047021945, "grad_norm": 2.790921926498413, "learning_rate": 5.030836010390849e-06, "loss": 1.0629, "step": 25470 }, { "epoch": 1.9961598746081504, "grad_norm": 2.6958272457122803, "learning_rate": 5.030527654137982e-06, "loss": 1.202, "step": 25471 }, { "epoch": 1.9962382445141067, "grad_norm": 5.116800785064697, "learning_rate": 5.030219297769003e-06, "loss": 1.484, "step": 25472 }, { "epoch": 1.9963166144200626, "grad_norm": 4.1630144119262695, "learning_rate": 5.0299109412850856e-06, "loss": 1.6151, "step": 25473 }, { "epoch": 1.996394984326019, "grad_norm": 2.22060489654541, "learning_rate": 5.029602584687403e-06, "loss": 1.3698, "step": 25474 }, { "epoch": 1.9964733542319748, "grad_norm": 4.347837448120117, "learning_rate": 5.029294227977126e-06, "loss": 0.7657, "step": 25475 }, { "epoch": 1.9965517241379311, "grad_norm": 5.846790790557861, "learning_rate": 5.02898587115543e-06, "loss": 1.3245, "step": 25476 }, { "epoch": 1.996630094043887, "grad_norm": 2.8188862800598145, "learning_rate": 5.0286775142234864e-06, "loss": 1.2656, "step": 25477 }, { "epoch": 1.9967084639498434, "grad_norm": 6.81173038482666, "learning_rate": 5.028369157182468e-06, "loss": 1.7061, "step": 25478 }, { "epoch": 1.9967868338557992, "grad_norm": 2.427643299102783, "learning_rate": 5.028060800033549e-06, "loss": 1.5358, "step": 25479 }, { "epoch": 1.9968652037617556, "grad_norm": 1.9376550912857056, "learning_rate": 5.0277524427778986e-06, "loss": 1.5687, "step": 25480 }, { "epoch": 1.9969435736677115, "grad_norm": 2.730198860168457, "learning_rate": 5.027444085416694e-06, "loss": 1.3121, "step": 25481 }, { "epoch": 1.9970219435736678, "grad_norm": 2.1526482105255127, "learning_rate": 5.0271357279511056e-06, "loss": 1.3441, "step": 25482 }, { "epoch": 1.9971003134796237, "grad_norm": 3.1823112964630127, "learning_rate": 5.026827370382308e-06, "loss": 1.5201, "step": 25483 }, { "epoch": 1.99717868338558, "grad_norm": 5.832528591156006, "learning_rate": 5.026519012711473e-06, "loss": 2.1304, "step": 25484 }, { "epoch": 1.997257053291536, "grad_norm": 1.9222663640975952, "learning_rate": 5.0262106549397705e-06, "loss": 1.38, "step": 25485 }, { "epoch": 1.9973354231974922, "grad_norm": 4.114529609680176, "learning_rate": 5.025902297068379e-06, "loss": 1.1149, "step": 25486 }, { "epoch": 1.9974137931034481, "grad_norm": 3.3245208263397217, "learning_rate": 5.025593939098467e-06, "loss": 1.3902, "step": 25487 }, { "epoch": 1.9974921630094045, "grad_norm": 2.4774794578552246, "learning_rate": 5.02528558103121e-06, "loss": 1.3416, "step": 25488 }, { "epoch": 1.9975705329153604, "grad_norm": 1.9589430093765259, "learning_rate": 5.0249772228677775e-06, "loss": 1.3913, "step": 25489 }, { "epoch": 1.9976489028213167, "grad_norm": 4.1237945556640625, "learning_rate": 5.024668864609345e-06, "loss": 1.5012, "step": 25490 }, { "epoch": 1.9977272727272726, "grad_norm": 4.218694686889648, "learning_rate": 5.0243605062570874e-06, "loss": 1.662, "step": 25491 }, { "epoch": 1.997805642633229, "grad_norm": 1.9737187623977661, "learning_rate": 5.024052147812172e-06, "loss": 1.43, "step": 25492 }, { "epoch": 1.9978840125391848, "grad_norm": 2.9416096210479736, "learning_rate": 5.023743789275777e-06, "loss": 1.6082, "step": 25493 }, { "epoch": 1.9979623824451411, "grad_norm": 1.7482362985610962, "learning_rate": 5.02343543064907e-06, "loss": 1.326, "step": 25494 }, { "epoch": 1.998040752351097, "grad_norm": 3.2121987342834473, "learning_rate": 5.0231270719332294e-06, "loss": 1.2637, "step": 25495 }, { "epoch": 1.9981191222570533, "grad_norm": 3.212524175643921, "learning_rate": 5.022818713129424e-06, "loss": 1.0175, "step": 25496 }, { "epoch": 1.9981974921630095, "grad_norm": 3.4164774417877197, "learning_rate": 5.0225103542388276e-06, "loss": 1.2912, "step": 25497 }, { "epoch": 1.9982758620689656, "grad_norm": 2.304319381713867, "learning_rate": 5.022201995262615e-06, "loss": 1.239, "step": 25498 }, { "epoch": 1.9983542319749217, "grad_norm": 4.593746185302734, "learning_rate": 5.021893636201956e-06, "loss": 1.3184, "step": 25499 }, { "epoch": 1.9984326018808778, "grad_norm": 2.645371913909912, "learning_rate": 5.0215852770580245e-06, "loss": 1.527, "step": 25500 }, { "epoch": 1.998510971786834, "grad_norm": 4.5364155769348145, "learning_rate": 5.021276917831994e-06, "loss": 1.9751, "step": 25501 }, { "epoch": 1.99858934169279, "grad_norm": 3.327880620956421, "learning_rate": 5.020968558525038e-06, "loss": 1.4274, "step": 25502 }, { "epoch": 1.9986677115987461, "grad_norm": 2.173755168914795, "learning_rate": 5.0206601991383275e-06, "loss": 1.326, "step": 25503 }, { "epoch": 1.9987460815047022, "grad_norm": 3.6745312213897705, "learning_rate": 5.020351839673037e-06, "loss": 0.9671, "step": 25504 }, { "epoch": 1.9988244514106583, "grad_norm": 2.2376930713653564, "learning_rate": 5.020043480130338e-06, "loss": 1.1948, "step": 25505 }, { "epoch": 1.9989028213166145, "grad_norm": 3.827350378036499, "learning_rate": 5.019735120511403e-06, "loss": 1.3716, "step": 25506 }, { "epoch": 1.9989811912225706, "grad_norm": 3.150724172592163, "learning_rate": 5.0194267608174065e-06, "loss": 1.8795, "step": 25507 }, { "epoch": 1.9990595611285267, "grad_norm": 2.0197315216064453, "learning_rate": 5.019118401049521e-06, "loss": 1.1832, "step": 25508 }, { "epoch": 1.9991379310344828, "grad_norm": 3.988403797149658, "learning_rate": 5.018810041208918e-06, "loss": 0.8818, "step": 25509 }, { "epoch": 1.999216300940439, "grad_norm": 3.499713659286499, "learning_rate": 5.018501681296772e-06, "loss": 1.329, "step": 25510 }, { "epoch": 1.999294670846395, "grad_norm": 4.424577713012695, "learning_rate": 5.018193321314254e-06, "loss": 0.9889, "step": 25511 }, { "epoch": 1.9993730407523511, "grad_norm": 3.335860013961792, "learning_rate": 5.01788496126254e-06, "loss": 1.4679, "step": 25512 }, { "epoch": 1.9994514106583072, "grad_norm": 3.2369909286499023, "learning_rate": 5.0175766011428e-06, "loss": 1.4554, "step": 25513 }, { "epoch": 1.9995297805642633, "grad_norm": 3.0701794624328613, "learning_rate": 5.017268240956208e-06, "loss": 1.1328, "step": 25514 }, { "epoch": 1.9996081504702194, "grad_norm": 2.099522352218628, "learning_rate": 5.016959880703936e-06, "loss": 1.2768, "step": 25515 }, { "epoch": 1.9996865203761756, "grad_norm": 2.3480801582336426, "learning_rate": 5.016651520387156e-06, "loss": 1.2239, "step": 25516 }, { "epoch": 1.9997648902821317, "grad_norm": 3.9941866397857666, "learning_rate": 5.0163431600070435e-06, "loss": 1.2549, "step": 25517 }, { "epoch": 1.9998432601880878, "grad_norm": 2.161329984664917, "learning_rate": 5.01603479956477e-06, "loss": 1.3868, "step": 25518 }, { "epoch": 1.999921630094044, "grad_norm": 4.938161849975586, "learning_rate": 5.015726439061509e-06, "loss": 1.5171, "step": 25519 }, { "epoch": 2.0, "grad_norm": 3.8114335536956787, "learning_rate": 5.015418078498431e-06, "loss": 1.2363, "step": 25520 }, { "epoch": 2.0, "eval_loss": 1.2674051523208618, "eval_runtime": 111.0587, "eval_samples_per_second": 24.194, "eval_steps_per_second": 6.051, "step": 25520 }, { "epoch": 2.0000783699059563, "grad_norm": 2.315239191055298, "learning_rate": 5.015109717876711e-06, "loss": 1.3855, "step": 25521 }, { "epoch": 2.000156739811912, "grad_norm": 2.015854835510254, "learning_rate": 5.0148013571975226e-06, "loss": 1.3362, "step": 25522 }, { "epoch": 2.0002351097178686, "grad_norm": 3.889047861099243, "learning_rate": 5.0144929964620374e-06, "loss": 1.6018, "step": 25523 }, { "epoch": 2.0003134796238244, "grad_norm": 2.595942735671997, "learning_rate": 5.0141846356714266e-06, "loss": 1.3352, "step": 25524 }, { "epoch": 2.0003918495297808, "grad_norm": 3.345827341079712, "learning_rate": 5.013876274826866e-06, "loss": 1.5675, "step": 25525 }, { "epoch": 2.0004702194357367, "grad_norm": 4.112117767333984, "learning_rate": 5.013567913929527e-06, "loss": 1.5872, "step": 25526 }, { "epoch": 2.000548589341693, "grad_norm": 2.390092372894287, "learning_rate": 5.013259552980582e-06, "loss": 1.3626, "step": 25527 }, { "epoch": 2.000626959247649, "grad_norm": 5.436458587646484, "learning_rate": 5.0129511919812044e-06, "loss": 1.3058, "step": 25528 }, { "epoch": 2.000705329153605, "grad_norm": 2.8329451084136963, "learning_rate": 5.012642830932568e-06, "loss": 1.3835, "step": 25529 }, { "epoch": 2.000783699059561, "grad_norm": 2.5470776557922363, "learning_rate": 5.012334469835845e-06, "loss": 0.896, "step": 25530 }, { "epoch": 2.0008620689655174, "grad_norm": 3.187007427215576, "learning_rate": 5.012026108692207e-06, "loss": 0.9468, "step": 25531 }, { "epoch": 2.0009404388714733, "grad_norm": 3.6487252712249756, "learning_rate": 5.011717747502829e-06, "loss": 1.0316, "step": 25532 }, { "epoch": 2.0010188087774297, "grad_norm": 2.8970038890838623, "learning_rate": 5.011409386268882e-06, "loss": 1.3201, "step": 25533 }, { "epoch": 2.0010971786833855, "grad_norm": 3.332141637802124, "learning_rate": 5.011101024991539e-06, "loss": 1.4744, "step": 25534 }, { "epoch": 2.001175548589342, "grad_norm": 2.3994650840759277, "learning_rate": 5.010792663671974e-06, "loss": 1.3684, "step": 25535 }, { "epoch": 2.0012539184952978, "grad_norm": 2.3212108612060547, "learning_rate": 5.010484302311359e-06, "loss": 1.0271, "step": 25536 }, { "epoch": 2.001332288401254, "grad_norm": 2.9089643955230713, "learning_rate": 5.010175940910867e-06, "loss": 1.5864, "step": 25537 }, { "epoch": 2.00141065830721, "grad_norm": 2.9586033821105957, "learning_rate": 5.009867579471672e-06, "loss": 1.6208, "step": 25538 }, { "epoch": 2.0014890282131663, "grad_norm": 1.9722174406051636, "learning_rate": 5.0095592179949444e-06, "loss": 1.1885, "step": 25539 }, { "epoch": 2.001567398119122, "grad_norm": 2.0602266788482666, "learning_rate": 5.00925085648186e-06, "loss": 1.1729, "step": 25540 }, { "epoch": 2.0016457680250785, "grad_norm": 2.8621666431427, "learning_rate": 5.0089424949335885e-06, "loss": 1.4007, "step": 25541 }, { "epoch": 2.0017241379310344, "grad_norm": 2.536125659942627, "learning_rate": 5.008634133351306e-06, "loss": 1.0122, "step": 25542 }, { "epoch": 2.0018025078369908, "grad_norm": 2.1441593170166016, "learning_rate": 5.008325771736182e-06, "loss": 1.3234, "step": 25543 }, { "epoch": 2.0018808777429467, "grad_norm": 2.541980504989624, "learning_rate": 5.008017410089392e-06, "loss": 1.6042, "step": 25544 }, { "epoch": 2.001959247648903, "grad_norm": 2.592603921890259, "learning_rate": 5.007709048412108e-06, "loss": 1.2757, "step": 25545 }, { "epoch": 2.002037617554859, "grad_norm": 2.5148637294769287, "learning_rate": 5.007400686705503e-06, "loss": 1.3722, "step": 25546 }, { "epoch": 2.002115987460815, "grad_norm": 3.3106393814086914, "learning_rate": 5.007092324970749e-06, "loss": 1.4577, "step": 25547 }, { "epoch": 2.002194357366771, "grad_norm": 3.7683777809143066, "learning_rate": 5.0067839632090185e-06, "loss": 1.2759, "step": 25548 }, { "epoch": 2.0022727272727274, "grad_norm": 3.6127238273620605, "learning_rate": 5.006475601421486e-06, "loss": 1.3664, "step": 25549 }, { "epoch": 2.0023510971786833, "grad_norm": 2.5696020126342773, "learning_rate": 5.006167239609324e-06, "loss": 1.3447, "step": 25550 }, { "epoch": 2.0024294670846396, "grad_norm": 6.982426643371582, "learning_rate": 5.005858877773705e-06, "loss": 1.3706, "step": 25551 }, { "epoch": 2.0025078369905955, "grad_norm": 2.6533749103546143, "learning_rate": 5.005550515915802e-06, "loss": 1.018, "step": 25552 }, { "epoch": 2.002586206896552, "grad_norm": 4.772246360778809, "learning_rate": 5.005242154036787e-06, "loss": 1.1715, "step": 25553 }, { "epoch": 2.0026645768025078, "grad_norm": 2.935889959335327, "learning_rate": 5.004933792137834e-06, "loss": 1.0754, "step": 25554 }, { "epoch": 2.002742946708464, "grad_norm": 2.2461745738983154, "learning_rate": 5.0046254302201155e-06, "loss": 1.3888, "step": 25555 }, { "epoch": 2.00282131661442, "grad_norm": 3.187443256378174, "learning_rate": 5.004317068284804e-06, "loss": 1.5014, "step": 25556 }, { "epoch": 2.0028996865203763, "grad_norm": 3.5542972087860107, "learning_rate": 5.0040087063330736e-06, "loss": 1.5693, "step": 25557 }, { "epoch": 2.002978056426332, "grad_norm": 4.178723335266113, "learning_rate": 5.003700344366095e-06, "loss": 1.291, "step": 25558 }, { "epoch": 2.0030564263322885, "grad_norm": 2.903625965118408, "learning_rate": 5.003391982385042e-06, "loss": 1.234, "step": 25559 }, { "epoch": 2.0031347962382444, "grad_norm": 3.265380620956421, "learning_rate": 5.0030836203910885e-06, "loss": 1.5787, "step": 25560 }, { "epoch": 2.0032131661442008, "grad_norm": 4.322426795959473, "learning_rate": 5.002775258385405e-06, "loss": 1.4691, "step": 25561 }, { "epoch": 2.0032915360501566, "grad_norm": 5.357983112335205, "learning_rate": 5.002466896369167e-06, "loss": 0.9691, "step": 25562 }, { "epoch": 2.003369905956113, "grad_norm": 5.911996841430664, "learning_rate": 5.002158534343546e-06, "loss": 1.8739, "step": 25563 }, { "epoch": 2.003448275862069, "grad_norm": 6.426892280578613, "learning_rate": 5.001850172309717e-06, "loss": 1.699, "step": 25564 }, { "epoch": 2.003526645768025, "grad_norm": 2.778627395629883, "learning_rate": 5.0015418102688476e-06, "loss": 1.5506, "step": 25565 }, { "epoch": 2.003605015673981, "grad_norm": 3.545051336288452, "learning_rate": 5.001233448222116e-06, "loss": 0.9138, "step": 25566 }, { "epoch": 2.0036833855799374, "grad_norm": 5.696430683135986, "learning_rate": 5.00092508617069e-06, "loss": 1.0638, "step": 25567 }, { "epoch": 2.0037617554858933, "grad_norm": 3.5822911262512207, "learning_rate": 5.00061672411575e-06, "loss": 1.2036, "step": 25568 }, { "epoch": 2.0038401253918496, "grad_norm": 3.2177016735076904, "learning_rate": 5.000308362058461e-06, "loss": 1.1465, "step": 25569 }, { "epoch": 2.0039184952978055, "grad_norm": 2.869910955429077, "learning_rate": 5e-06, "loss": 0.9027, "step": 25570 }, { "epoch": 2.003996865203762, "grad_norm": 2.4946706295013428, "learning_rate": 4.99969163794154e-06, "loss": 1.4692, "step": 25571 }, { "epoch": 2.0040752351097177, "grad_norm": 3.0132129192352295, "learning_rate": 4.999383275884253e-06, "loss": 1.2496, "step": 25572 }, { "epoch": 2.004153605015674, "grad_norm": 3.7911176681518555, "learning_rate": 4.9990749138293106e-06, "loss": 1.4215, "step": 25573 }, { "epoch": 2.00423197492163, "grad_norm": 3.7012970447540283, "learning_rate": 4.998766551777885e-06, "loss": 1.6932, "step": 25574 }, { "epoch": 2.0043103448275863, "grad_norm": 4.946603775024414, "learning_rate": 4.998458189731153e-06, "loss": 1.1565, "step": 25575 }, { "epoch": 2.004388714733542, "grad_norm": 2.9864211082458496, "learning_rate": 4.998149827690286e-06, "loss": 0.8414, "step": 25576 }, { "epoch": 2.0044670846394985, "grad_norm": 3.395439386367798, "learning_rate": 4.997841465656455e-06, "loss": 1.4179, "step": 25577 }, { "epoch": 2.0045454545454544, "grad_norm": 3.9996790885925293, "learning_rate": 4.997533103630834e-06, "loss": 1.0269, "step": 25578 }, { "epoch": 2.0046238244514107, "grad_norm": 2.858924150466919, "learning_rate": 4.9972247416145945e-06, "loss": 1.3004, "step": 25579 }, { "epoch": 2.0047021943573666, "grad_norm": 2.6360268592834473, "learning_rate": 4.996916379608914e-06, "loss": 1.4509, "step": 25580 }, { "epoch": 2.004780564263323, "grad_norm": 2.0290141105651855, "learning_rate": 4.99660801761496e-06, "loss": 1.6759, "step": 25581 }, { "epoch": 2.004858934169279, "grad_norm": 4.488654613494873, "learning_rate": 4.996299655633907e-06, "loss": 1.6053, "step": 25582 }, { "epoch": 2.004937304075235, "grad_norm": 2.9444386959075928, "learning_rate": 4.995991293666928e-06, "loss": 1.3614, "step": 25583 }, { "epoch": 2.005015673981191, "grad_norm": 8.215251922607422, "learning_rate": 4.995682931715198e-06, "loss": 1.2738, "step": 25584 }, { "epoch": 2.0050940438871474, "grad_norm": 9.03470516204834, "learning_rate": 4.995374569779886e-06, "loss": 1.4069, "step": 25585 }, { "epoch": 2.0051724137931033, "grad_norm": 3.214050531387329, "learning_rate": 4.995066207862166e-06, "loss": 1.3771, "step": 25586 }, { "epoch": 2.0052507836990596, "grad_norm": 17.484481811523438, "learning_rate": 4.994757845963215e-06, "loss": 0.8713, "step": 25587 }, { "epoch": 2.0053291536050155, "grad_norm": 3.286900520324707, "learning_rate": 4.9944494840842e-06, "loss": 1.0785, "step": 25588 }, { "epoch": 2.005407523510972, "grad_norm": 8.024312019348145, "learning_rate": 4.994141122226296e-06, "loss": 1.6806, "step": 25589 }, { "epoch": 2.0054858934169277, "grad_norm": 6.0303955078125, "learning_rate": 4.993832760390677e-06, "loss": 1.85, "step": 25590 }, { "epoch": 2.005564263322884, "grad_norm": 3.22078800201416, "learning_rate": 4.993524398578515e-06, "loss": 0.92, "step": 25591 }, { "epoch": 2.00564263322884, "grad_norm": 4.132789134979248, "learning_rate": 4.993216036790983e-06, "loss": 0.902, "step": 25592 }, { "epoch": 2.0057210031347963, "grad_norm": 3.4684221744537354, "learning_rate": 4.992907675029252e-06, "loss": 1.2384, "step": 25593 }, { "epoch": 2.005799373040752, "grad_norm": 3.283273220062256, "learning_rate": 4.992599313294498e-06, "loss": 1.0186, "step": 25594 }, { "epoch": 2.0058777429467085, "grad_norm": 2.3454842567443848, "learning_rate": 4.992290951587894e-06, "loss": 1.5241, "step": 25595 }, { "epoch": 2.0059561128526644, "grad_norm": 2.096947193145752, "learning_rate": 4.99198258991061e-06, "loss": 1.5032, "step": 25596 }, { "epoch": 2.0060344827586207, "grad_norm": 3.204218626022339, "learning_rate": 4.9916742282638195e-06, "loss": 1.2463, "step": 25597 }, { "epoch": 2.0061128526645766, "grad_norm": 4.5654296875, "learning_rate": 4.991365866648697e-06, "loss": 1.5038, "step": 25598 }, { "epoch": 2.006191222570533, "grad_norm": 2.472823143005371, "learning_rate": 4.991057505066413e-06, "loss": 1.0543, "step": 25599 }, { "epoch": 2.006269592476489, "grad_norm": 3.514237642288208, "learning_rate": 4.990749143518141e-06, "loss": 0.7561, "step": 25600 }, { "epoch": 2.006347962382445, "grad_norm": 3.982825517654419, "learning_rate": 4.990440782005056e-06, "loss": 1.18, "step": 25601 }, { "epoch": 2.006426332288401, "grad_norm": 1.8407471179962158, "learning_rate": 4.990132420528329e-06, "loss": 1.2268, "step": 25602 }, { "epoch": 2.0065047021943574, "grad_norm": 3.5323374271392822, "learning_rate": 4.989824059089134e-06, "loss": 1.4603, "step": 25603 }, { "epoch": 2.0065830721003133, "grad_norm": 3.64400315284729, "learning_rate": 4.989515697688642e-06, "loss": 1.5609, "step": 25604 }, { "epoch": 2.0066614420062696, "grad_norm": 2.3141868114471436, "learning_rate": 4.989207336328026e-06, "loss": 1.218, "step": 25605 }, { "epoch": 2.0067398119122255, "grad_norm": 3.581122398376465, "learning_rate": 4.9888989750084635e-06, "loss": 1.374, "step": 25606 }, { "epoch": 2.006818181818182, "grad_norm": 2.9018218517303467, "learning_rate": 4.98859061373112e-06, "loss": 1.4438, "step": 25607 }, { "epoch": 2.0068965517241377, "grad_norm": 2.06520938873291, "learning_rate": 4.9882822524971725e-06, "loss": 0.234, "step": 25608 }, { "epoch": 2.006974921630094, "grad_norm": 2.9135749340057373, "learning_rate": 4.987973891307794e-06, "loss": 1.2764, "step": 25609 }, { "epoch": 2.00705329153605, "grad_norm": 4.277474880218506, "learning_rate": 4.987665530164157e-06, "loss": 1.426, "step": 25610 }, { "epoch": 2.0071316614420063, "grad_norm": 5.145147323608398, "learning_rate": 4.987357169067434e-06, "loss": 1.3805, "step": 25611 }, { "epoch": 2.007210031347962, "grad_norm": 2.6468164920806885, "learning_rate": 4.9870488080187955e-06, "loss": 1.3478, "step": 25612 }, { "epoch": 2.0072884012539185, "grad_norm": 2.1024904251098633, "learning_rate": 4.98674044701942e-06, "loss": 0.9915, "step": 25613 }, { "epoch": 2.007366771159875, "grad_norm": 2.8288276195526123, "learning_rate": 4.986432086070475e-06, "loss": 1.801, "step": 25614 }, { "epoch": 2.0074451410658307, "grad_norm": 3.220698118209839, "learning_rate": 4.986123725173136e-06, "loss": 1.4235, "step": 25615 }, { "epoch": 2.007523510971787, "grad_norm": 4.192529201507568, "learning_rate": 4.985815364328575e-06, "loss": 1.0726, "step": 25616 }, { "epoch": 2.007601880877743, "grad_norm": 2.844883918762207, "learning_rate": 4.985507003537966e-06, "loss": 1.3483, "step": 25617 }, { "epoch": 2.0076802507836993, "grad_norm": 3.5794413089752197, "learning_rate": 4.985198642802479e-06, "loss": 1.4853, "step": 25618 }, { "epoch": 2.007758620689655, "grad_norm": 2.487239360809326, "learning_rate": 4.9848902821232895e-06, "loss": 1.4786, "step": 25619 }, { "epoch": 2.0078369905956115, "grad_norm": 3.2707271575927734, "learning_rate": 4.98458192150157e-06, "loss": 0.9499, "step": 25620 }, { "epoch": 2.0079153605015674, "grad_norm": 3.420586347579956, "learning_rate": 4.984273560938494e-06, "loss": 1.32, "step": 25621 }, { "epoch": 2.0079937304075237, "grad_norm": 2.8167459964752197, "learning_rate": 4.983965200435232e-06, "loss": 1.2875, "step": 25622 }, { "epoch": 2.0080721003134796, "grad_norm": 3.370725154876709, "learning_rate": 4.983656839992958e-06, "loss": 1.3177, "step": 25623 }, { "epoch": 2.008150470219436, "grad_norm": 4.695559978485107, "learning_rate": 4.983348479612844e-06, "loss": 1.4026, "step": 25624 }, { "epoch": 2.008228840125392, "grad_norm": 6.431939125061035, "learning_rate": 4.983040119296067e-06, "loss": 1.3216, "step": 25625 }, { "epoch": 2.008307210031348, "grad_norm": 4.866130828857422, "learning_rate": 4.9827317590437944e-06, "loss": 1.8139, "step": 25626 }, { "epoch": 2.008385579937304, "grad_norm": 3.5336503982543945, "learning_rate": 4.982423398857201e-06, "loss": 1.4271, "step": 25627 }, { "epoch": 2.0084639498432604, "grad_norm": 2.040964126586914, "learning_rate": 4.982115038737462e-06, "loss": 1.4544, "step": 25628 }, { "epoch": 2.0085423197492163, "grad_norm": 2.3639614582061768, "learning_rate": 4.981806678685747e-06, "loss": 1.5039, "step": 25629 }, { "epoch": 2.0086206896551726, "grad_norm": 3.029175043106079, "learning_rate": 4.9814983187032285e-06, "loss": 1.1573, "step": 25630 }, { "epoch": 2.0086990595611285, "grad_norm": 3.0298380851745605, "learning_rate": 4.981189958791082e-06, "loss": 1.2898, "step": 25631 }, { "epoch": 2.008777429467085, "grad_norm": 3.3510940074920654, "learning_rate": 4.9808815989504815e-06, "loss": 1.472, "step": 25632 }, { "epoch": 2.0088557993730407, "grad_norm": 4.172348499298096, "learning_rate": 4.980573239182595e-06, "loss": 1.5065, "step": 25633 }, { "epoch": 2.008934169278997, "grad_norm": 3.409435272216797, "learning_rate": 4.980264879488598e-06, "loss": 1.1985, "step": 25634 }, { "epoch": 2.009012539184953, "grad_norm": 4.424959659576416, "learning_rate": 4.979956519869663e-06, "loss": 1.6116, "step": 25635 }, { "epoch": 2.0090909090909093, "grad_norm": 4.520177364349365, "learning_rate": 4.979648160326966e-06, "loss": 1.5115, "step": 25636 }, { "epoch": 2.009169278996865, "grad_norm": 2.185114622116089, "learning_rate": 4.979339800861674e-06, "loss": 1.3159, "step": 25637 }, { "epoch": 2.0092476489028215, "grad_norm": 3.6227285861968994, "learning_rate": 4.979031441474964e-06, "loss": 1.5197, "step": 25638 }, { "epoch": 2.0093260188087774, "grad_norm": 3.244185209274292, "learning_rate": 4.978723082168006e-06, "loss": 1.6028, "step": 25639 }, { "epoch": 2.0094043887147337, "grad_norm": 5.677401065826416, "learning_rate": 4.978414722941977e-06, "loss": 0.7578, "step": 25640 }, { "epoch": 2.0094827586206896, "grad_norm": 3.0568299293518066, "learning_rate": 4.978106363798047e-06, "loss": 1.2211, "step": 25641 }, { "epoch": 2.009561128526646, "grad_norm": 2.7529027462005615, "learning_rate": 4.977798004737386e-06, "loss": 1.7658, "step": 25642 }, { "epoch": 2.009639498432602, "grad_norm": 2.173123836517334, "learning_rate": 4.977489645761175e-06, "loss": 1.5434, "step": 25643 }, { "epoch": 2.009717868338558, "grad_norm": 3.471407651901245, "learning_rate": 4.9771812868705776e-06, "loss": 1.1828, "step": 25644 }, { "epoch": 2.009796238244514, "grad_norm": 2.4435842037200928, "learning_rate": 4.976872928066772e-06, "loss": 0.6697, "step": 25645 }, { "epoch": 2.0098746081504704, "grad_norm": 3.67875337600708, "learning_rate": 4.9765645693509304e-06, "loss": 1.2897, "step": 25646 }, { "epoch": 2.0099529780564263, "grad_norm": 1.9550485610961914, "learning_rate": 4.976256210724225e-06, "loss": 1.0876, "step": 25647 }, { "epoch": 2.0100313479623826, "grad_norm": 3.806670904159546, "learning_rate": 4.975947852187829e-06, "loss": 0.8183, "step": 25648 }, { "epoch": 2.0101097178683385, "grad_norm": 9.234003067016602, "learning_rate": 4.975639493742913e-06, "loss": 1.7553, "step": 25649 }, { "epoch": 2.010188087774295, "grad_norm": 3.2875418663024902, "learning_rate": 4.9753311353906546e-06, "loss": 1.5283, "step": 25650 }, { "epoch": 2.0102664576802507, "grad_norm": 3.0824897289276123, "learning_rate": 4.975022777132223e-06, "loss": 1.4895, "step": 25651 }, { "epoch": 2.010344827586207, "grad_norm": 9.123323440551758, "learning_rate": 4.974714418968793e-06, "loss": 1.3382, "step": 25652 }, { "epoch": 2.010423197492163, "grad_norm": 2.2773098945617676, "learning_rate": 4.974406060901534e-06, "loss": 1.5469, "step": 25653 }, { "epoch": 2.0105015673981192, "grad_norm": 2.2232398986816406, "learning_rate": 4.974097702931624e-06, "loss": 1.4695, "step": 25654 }, { "epoch": 2.010579937304075, "grad_norm": 3.145432233810425, "learning_rate": 4.973789345060231e-06, "loss": 1.4289, "step": 25655 }, { "epoch": 2.0106583072100315, "grad_norm": 2.4800844192504883, "learning_rate": 4.9734809872885296e-06, "loss": 1.5014, "step": 25656 }, { "epoch": 2.0107366771159874, "grad_norm": 5.01071310043335, "learning_rate": 4.973172629617693e-06, "loss": 2.0126, "step": 25657 }, { "epoch": 2.0108150470219437, "grad_norm": 2.9497361183166504, "learning_rate": 4.972864272048895e-06, "loss": 1.5519, "step": 25658 }, { "epoch": 2.0108934169278996, "grad_norm": 5.2864298820495605, "learning_rate": 4.972555914583307e-06, "loss": 1.4812, "step": 25659 }, { "epoch": 2.010971786833856, "grad_norm": 2.1714649200439453, "learning_rate": 4.972247557222102e-06, "loss": 1.1931, "step": 25660 }, { "epoch": 2.011050156739812, "grad_norm": 2.1301145553588867, "learning_rate": 4.971939199966452e-06, "loss": 1.3241, "step": 25661 }, { "epoch": 2.011128526645768, "grad_norm": 2.543806314468384, "learning_rate": 4.971630842817534e-06, "loss": 1.101, "step": 25662 }, { "epoch": 2.011206896551724, "grad_norm": 2.7797904014587402, "learning_rate": 4.971322485776514e-06, "loss": 1.382, "step": 25663 }, { "epoch": 2.0112852664576804, "grad_norm": 4.117465972900391, "learning_rate": 4.9710141288445706e-06, "loss": 1.0909, "step": 25664 }, { "epoch": 2.0113636363636362, "grad_norm": 4.763472080230713, "learning_rate": 4.970705772022875e-06, "loss": 1.1149, "step": 25665 }, { "epoch": 2.0114420062695926, "grad_norm": 2.684723377227783, "learning_rate": 4.9703974153125996e-06, "loss": 1.4384, "step": 25666 }, { "epoch": 2.0115203761755485, "grad_norm": 2.1380081176757812, "learning_rate": 4.970089058714916e-06, "loss": 1.332, "step": 25667 }, { "epoch": 2.011598746081505, "grad_norm": 2.6629228591918945, "learning_rate": 4.969780702230998e-06, "loss": 1.6259, "step": 25668 }, { "epoch": 2.0116771159874607, "grad_norm": 2.5693681240081787, "learning_rate": 4.969472345862021e-06, "loss": 0.7306, "step": 25669 }, { "epoch": 2.011755485893417, "grad_norm": 1.991280436515808, "learning_rate": 4.9691639896091526e-06, "loss": 1.1328, "step": 25670 }, { "epoch": 2.011833855799373, "grad_norm": 9.437658309936523, "learning_rate": 4.96885563347357e-06, "loss": 1.5087, "step": 25671 }, { "epoch": 2.0119122257053292, "grad_norm": 2.800494432449341, "learning_rate": 4.968547277456444e-06, "loss": 1.7749, "step": 25672 }, { "epoch": 2.011990595611285, "grad_norm": 3.15429425239563, "learning_rate": 4.96823892155895e-06, "loss": 1.2684, "step": 25673 }, { "epoch": 2.0120689655172415, "grad_norm": 2.7127370834350586, "learning_rate": 4.967930565782256e-06, "loss": 1.5226, "step": 25674 }, { "epoch": 2.0121473354231973, "grad_norm": 4.959653854370117, "learning_rate": 4.967622210127538e-06, "loss": 1.2923, "step": 25675 }, { "epoch": 2.0122257053291537, "grad_norm": 4.030667781829834, "learning_rate": 4.967313854595969e-06, "loss": 0.9918, "step": 25676 }, { "epoch": 2.0123040752351096, "grad_norm": 7.866405010223389, "learning_rate": 4.967005499188724e-06, "loss": 1.3379, "step": 25677 }, { "epoch": 2.012382445141066, "grad_norm": 3.8096415996551514, "learning_rate": 4.96669714390697e-06, "loss": 1.308, "step": 25678 }, { "epoch": 2.012460815047022, "grad_norm": 2.700995445251465, "learning_rate": 4.966388788751883e-06, "loss": 1.2747, "step": 25679 }, { "epoch": 2.012539184952978, "grad_norm": 2.591416358947754, "learning_rate": 4.966080433724636e-06, "loss": 1.4131, "step": 25680 }, { "epoch": 2.012617554858934, "grad_norm": 7.054470539093018, "learning_rate": 4.965772078826404e-06, "loss": 1.4299, "step": 25681 }, { "epoch": 2.0126959247648903, "grad_norm": 5.394391059875488, "learning_rate": 4.965463724058355e-06, "loss": 1.4413, "step": 25682 }, { "epoch": 2.0127742946708462, "grad_norm": 5.384701251983643, "learning_rate": 4.9651553694216646e-06, "loss": 1.0966, "step": 25683 }, { "epoch": 2.0128526645768026, "grad_norm": 2.9259684085845947, "learning_rate": 4.964847014917506e-06, "loss": 1.5581, "step": 25684 }, { "epoch": 2.0129310344827585, "grad_norm": 10.654889106750488, "learning_rate": 4.964538660547051e-06, "loss": 1.1892, "step": 25685 }, { "epoch": 2.013009404388715, "grad_norm": 1.9824544191360474, "learning_rate": 4.964230306311472e-06, "loss": 1.0753, "step": 25686 }, { "epoch": 2.0130877742946707, "grad_norm": 12.57641315460205, "learning_rate": 4.963921952211943e-06, "loss": 1.2799, "step": 25687 }, { "epoch": 2.013166144200627, "grad_norm": 2.477924346923828, "learning_rate": 4.963613598249638e-06, "loss": 1.456, "step": 25688 }, { "epoch": 2.013244514106583, "grad_norm": 4.731114864349365, "learning_rate": 4.963305244425725e-06, "loss": 0.8676, "step": 25689 }, { "epoch": 2.0133228840125392, "grad_norm": 3.3900630474090576, "learning_rate": 4.962996890741382e-06, "loss": 1.5392, "step": 25690 }, { "epoch": 2.013401253918495, "grad_norm": 3.9368884563446045, "learning_rate": 4.962688537197779e-06, "loss": 1.3313, "step": 25691 }, { "epoch": 2.0134796238244514, "grad_norm": 13.627907752990723, "learning_rate": 4.9623801837960915e-06, "loss": 1.5365, "step": 25692 }, { "epoch": 2.0135579937304073, "grad_norm": 2.93629789352417, "learning_rate": 4.962071830537489e-06, "loss": 1.2102, "step": 25693 }, { "epoch": 2.0136363636363637, "grad_norm": 2.737715005874634, "learning_rate": 4.961763477423145e-06, "loss": 1.6303, "step": 25694 }, { "epoch": 2.0137147335423196, "grad_norm": 3.8501439094543457, "learning_rate": 4.961455124454233e-06, "loss": 1.1342, "step": 25695 }, { "epoch": 2.013793103448276, "grad_norm": 3.0535717010498047, "learning_rate": 4.961146771631927e-06, "loss": 1.4425, "step": 25696 }, { "epoch": 2.013871473354232, "grad_norm": 9.836753845214844, "learning_rate": 4.960838418957399e-06, "loss": 1.1327, "step": 25697 }, { "epoch": 2.013949843260188, "grad_norm": 5.8080973625183105, "learning_rate": 4.960530066431821e-06, "loss": 1.4866, "step": 25698 }, { "epoch": 2.014028213166144, "grad_norm": 3.0121312141418457, "learning_rate": 4.960221714056368e-06, "loss": 1.3086, "step": 25699 }, { "epoch": 2.0141065830721003, "grad_norm": 3.147479772567749, "learning_rate": 4.959913361832209e-06, "loss": 1.528, "step": 25700 }, { "epoch": 2.014184952978056, "grad_norm": 3.3415908813476562, "learning_rate": 4.9596050097605185e-06, "loss": 0.8295, "step": 25701 }, { "epoch": 2.0142633228840126, "grad_norm": 4.3946099281311035, "learning_rate": 4.959296657842471e-06, "loss": 1.2735, "step": 25702 }, { "epoch": 2.0143416927899684, "grad_norm": 9.152653694152832, "learning_rate": 4.958988306079239e-06, "loss": 1.262, "step": 25703 }, { "epoch": 2.0144200626959248, "grad_norm": 8.279033660888672, "learning_rate": 4.958679954471993e-06, "loss": 1.1392, "step": 25704 }, { "epoch": 2.0144984326018807, "grad_norm": 5.39618444442749, "learning_rate": 4.958371603021908e-06, "loss": 1.3506, "step": 25705 }, { "epoch": 2.014576802507837, "grad_norm": 2.9428980350494385, "learning_rate": 4.958063251730155e-06, "loss": 1.1056, "step": 25706 }, { "epoch": 2.014655172413793, "grad_norm": 3.9835896492004395, "learning_rate": 4.95775490059791e-06, "loss": 0.7471, "step": 25707 }, { "epoch": 2.014733542319749, "grad_norm": 3.6887075901031494, "learning_rate": 4.957446549626341e-06, "loss": 1.0536, "step": 25708 }, { "epoch": 2.0148119122257055, "grad_norm": 3.818511962890625, "learning_rate": 4.957138198816625e-06, "loss": 1.368, "step": 25709 }, { "epoch": 2.0148902821316614, "grad_norm": 4.570028781890869, "learning_rate": 4.956829848169934e-06, "loss": 1.6014, "step": 25710 }, { "epoch": 2.0149686520376178, "grad_norm": 5.984767436981201, "learning_rate": 4.956521497687439e-06, "loss": 1.7743, "step": 25711 }, { "epoch": 2.0150470219435737, "grad_norm": 5.749345779418945, "learning_rate": 4.956213147370313e-06, "loss": 1.0527, "step": 25712 }, { "epoch": 2.01512539184953, "grad_norm": 2.3602182865142822, "learning_rate": 4.95590479721973e-06, "loss": 1.192, "step": 25713 }, { "epoch": 2.015203761755486, "grad_norm": 2.904987335205078, "learning_rate": 4.955596447236864e-06, "loss": 1.5709, "step": 25714 }, { "epoch": 2.015282131661442, "grad_norm": 2.173133134841919, "learning_rate": 4.955288097422885e-06, "loss": 1.5904, "step": 25715 }, { "epoch": 2.015360501567398, "grad_norm": 4.925267219543457, "learning_rate": 4.954979747778968e-06, "loss": 1.3297, "step": 25716 }, { "epoch": 2.0154388714733544, "grad_norm": 2.252784252166748, "learning_rate": 4.9546713983062846e-06, "loss": 1.2908, "step": 25717 }, { "epoch": 2.0155172413793103, "grad_norm": 3.126627206802368, "learning_rate": 4.9543630490060085e-06, "loss": 1.2097, "step": 25718 }, { "epoch": 2.0155956112852667, "grad_norm": 2.305056095123291, "learning_rate": 4.954054699879311e-06, "loss": 1.0977, "step": 25719 }, { "epoch": 2.0156739811912225, "grad_norm": 2.955531597137451, "learning_rate": 4.953746350927365e-06, "loss": 1.3009, "step": 25720 }, { "epoch": 2.015752351097179, "grad_norm": 5.695986747741699, "learning_rate": 4.953438002151345e-06, "loss": 1.6576, "step": 25721 }, { "epoch": 2.0158307210031348, "grad_norm": 2.2085933685302734, "learning_rate": 4.953129653552424e-06, "loss": 1.2846, "step": 25722 }, { "epoch": 2.015909090909091, "grad_norm": 2.444547414779663, "learning_rate": 4.952821305131773e-06, "loss": 1.1989, "step": 25723 }, { "epoch": 2.015987460815047, "grad_norm": 3.5122764110565186, "learning_rate": 4.952512956890565e-06, "loss": 1.1766, "step": 25724 }, { "epoch": 2.0160658307210033, "grad_norm": 3.2038521766662598, "learning_rate": 4.952204608829975e-06, "loss": 1.2798, "step": 25725 }, { "epoch": 2.016144200626959, "grad_norm": 2.980574131011963, "learning_rate": 4.951896260951171e-06, "loss": 1.5331, "step": 25726 }, { "epoch": 2.0162225705329155, "grad_norm": 4.574360370635986, "learning_rate": 4.951587913255331e-06, "loss": 1.7782, "step": 25727 }, { "epoch": 2.0163009404388714, "grad_norm": 2.6232759952545166, "learning_rate": 4.9512795657436244e-06, "loss": 1.162, "step": 25728 }, { "epoch": 2.0163793103448278, "grad_norm": 2.60569167137146, "learning_rate": 4.950971218417227e-06, "loss": 1.6182, "step": 25729 }, { "epoch": 2.0164576802507836, "grad_norm": 3.9532582759857178, "learning_rate": 4.95066287127731e-06, "loss": 1.5676, "step": 25730 }, { "epoch": 2.01653605015674, "grad_norm": 3.5340378284454346, "learning_rate": 4.950354524325044e-06, "loss": 1.2884, "step": 25731 }, { "epoch": 2.016614420062696, "grad_norm": 2.014606475830078, "learning_rate": 4.950046177561605e-06, "loss": 0.8829, "step": 25732 }, { "epoch": 2.016692789968652, "grad_norm": 4.190402030944824, "learning_rate": 4.949737830988166e-06, "loss": 1.4421, "step": 25733 }, { "epoch": 2.016771159874608, "grad_norm": 2.9577813148498535, "learning_rate": 4.949429484605897e-06, "loss": 1.346, "step": 25734 }, { "epoch": 2.0168495297805644, "grad_norm": 3.9530069828033447, "learning_rate": 4.949121138415973e-06, "loss": 1.3777, "step": 25735 }, { "epoch": 2.0169278996865203, "grad_norm": 2.3965494632720947, "learning_rate": 4.948812792419565e-06, "loss": 0.4983, "step": 25736 }, { "epoch": 2.0170062695924766, "grad_norm": 7.995419979095459, "learning_rate": 4.948504446617849e-06, "loss": 1.2388, "step": 25737 }, { "epoch": 2.0170846394984325, "grad_norm": 5.060823440551758, "learning_rate": 4.948196101011993e-06, "loss": 1.2474, "step": 25738 }, { "epoch": 2.017163009404389, "grad_norm": 3.7766120433807373, "learning_rate": 4.947887755603173e-06, "loss": 1.0513, "step": 25739 }, { "epoch": 2.0172413793103448, "grad_norm": 5.214733123779297, "learning_rate": 4.947579410392563e-06, "loss": 1.1038, "step": 25740 }, { "epoch": 2.017319749216301, "grad_norm": 3.202855348587036, "learning_rate": 4.947271065381333e-06, "loss": 1.3355, "step": 25741 }, { "epoch": 2.017398119122257, "grad_norm": 2.129455327987671, "learning_rate": 4.946962720570656e-06, "loss": 1.3093, "step": 25742 }, { "epoch": 2.0174764890282133, "grad_norm": 2.5972139835357666, "learning_rate": 4.946654375961705e-06, "loss": 1.2758, "step": 25743 }, { "epoch": 2.017554858934169, "grad_norm": 2.5013670921325684, "learning_rate": 4.946346031555656e-06, "loss": 1.1387, "step": 25744 }, { "epoch": 2.0176332288401255, "grad_norm": 3.020272731781006, "learning_rate": 4.946037687353677e-06, "loss": 1.3087, "step": 25745 }, { "epoch": 2.0177115987460814, "grad_norm": 3.597565174102783, "learning_rate": 4.945729343356943e-06, "loss": 1.5885, "step": 25746 }, { "epoch": 2.0177899686520377, "grad_norm": 3.3255808353424072, "learning_rate": 4.9454209995666275e-06, "loss": 1.5274, "step": 25747 }, { "epoch": 2.0178683385579936, "grad_norm": 1.9303773641586304, "learning_rate": 4.945112655983903e-06, "loss": 1.1983, "step": 25748 }, { "epoch": 2.01794670846395, "grad_norm": 4.5087666511535645, "learning_rate": 4.944804312609941e-06, "loss": 1.4151, "step": 25749 }, { "epoch": 2.018025078369906, "grad_norm": 2.479285478591919, "learning_rate": 4.944495969445914e-06, "loss": 1.1401, "step": 25750 }, { "epoch": 2.018103448275862, "grad_norm": 5.293533802032471, "learning_rate": 4.944187626492997e-06, "loss": 1.6372, "step": 25751 }, { "epoch": 2.018181818181818, "grad_norm": 2.805838108062744, "learning_rate": 4.943879283752362e-06, "loss": 1.6188, "step": 25752 }, { "epoch": 2.0182601880877744, "grad_norm": 1.979970932006836, "learning_rate": 4.943570941225181e-06, "loss": 0.999, "step": 25753 }, { "epoch": 2.0183385579937303, "grad_norm": 5.781512260437012, "learning_rate": 4.943262598912626e-06, "loss": 0.9412, "step": 25754 }, { "epoch": 2.0184169278996866, "grad_norm": 2.020012855529785, "learning_rate": 4.942954256815874e-06, "loss": 0.9311, "step": 25755 }, { "epoch": 2.0184952978056425, "grad_norm": 2.815228223800659, "learning_rate": 4.942645914936092e-06, "loss": 1.7157, "step": 25756 }, { "epoch": 2.018573667711599, "grad_norm": 7.176788806915283, "learning_rate": 4.942337573274455e-06, "loss": 1.0974, "step": 25757 }, { "epoch": 2.0186520376175547, "grad_norm": 3.47465181350708, "learning_rate": 4.942029231832137e-06, "loss": 1.4744, "step": 25758 }, { "epoch": 2.018730407523511, "grad_norm": 2.5558102130889893, "learning_rate": 4.941720890610312e-06, "loss": 1.6516, "step": 25759 }, { "epoch": 2.018808777429467, "grad_norm": 3.0742132663726807, "learning_rate": 4.941412549610149e-06, "loss": 1.4008, "step": 25760 }, { "epoch": 2.0188871473354233, "grad_norm": 2.63698148727417, "learning_rate": 4.941104208832822e-06, "loss": 0.8056, "step": 25761 }, { "epoch": 2.018965517241379, "grad_norm": 4.554261684417725, "learning_rate": 4.940795868279504e-06, "loss": 1.2985, "step": 25762 }, { "epoch": 2.0190438871473355, "grad_norm": 6.733950138092041, "learning_rate": 4.940487527951372e-06, "loss": 1.2236, "step": 25763 }, { "epoch": 2.0191222570532914, "grad_norm": 2.5209670066833496, "learning_rate": 4.940179187849591e-06, "loss": 1.1537, "step": 25764 }, { "epoch": 2.0192006269592477, "grad_norm": 3.0547306537628174, "learning_rate": 4.939870847975338e-06, "loss": 1.6545, "step": 25765 }, { "epoch": 2.0192789968652036, "grad_norm": 2.4858837127685547, "learning_rate": 4.939562508329787e-06, "loss": 1.2907, "step": 25766 }, { "epoch": 2.01935736677116, "grad_norm": 7.582021713256836, "learning_rate": 4.939254168914109e-06, "loss": 1.4314, "step": 25767 }, { "epoch": 2.019435736677116, "grad_norm": 3.654564142227173, "learning_rate": 4.938945829729475e-06, "loss": 1.2499, "step": 25768 }, { "epoch": 2.019514106583072, "grad_norm": 4.19820499420166, "learning_rate": 4.93863749077706e-06, "loss": 1.4036, "step": 25769 }, { "epoch": 2.019592476489028, "grad_norm": 2.956181049346924, "learning_rate": 4.93832915205804e-06, "loss": 1.3956, "step": 25770 }, { "epoch": 2.0196708463949844, "grad_norm": 2.2298927307128906, "learning_rate": 4.93802081357358e-06, "loss": 1.5914, "step": 25771 }, { "epoch": 2.0197492163009403, "grad_norm": 21.784656524658203, "learning_rate": 4.937712475324859e-06, "loss": 0.6721, "step": 25772 }, { "epoch": 2.0198275862068966, "grad_norm": 3.3699889183044434, "learning_rate": 4.937404137313047e-06, "loss": 1.6488, "step": 25773 }, { "epoch": 2.0199059561128525, "grad_norm": 2.4121644496917725, "learning_rate": 4.93709579953932e-06, "loss": 0.8089, "step": 25774 }, { "epoch": 2.019984326018809, "grad_norm": 5.4322662353515625, "learning_rate": 4.936787462004845e-06, "loss": 1.0005, "step": 25775 }, { "epoch": 2.0200626959247647, "grad_norm": 2.0881640911102295, "learning_rate": 4.936479124710799e-06, "loss": 1.5182, "step": 25776 }, { "epoch": 2.020141065830721, "grad_norm": 2.2758431434631348, "learning_rate": 4.936170787658353e-06, "loss": 1.5281, "step": 25777 }, { "epoch": 2.020219435736677, "grad_norm": 2.0434927940368652, "learning_rate": 4.9358624508486825e-06, "loss": 1.4417, "step": 25778 }, { "epoch": 2.0202978056426333, "grad_norm": 6.454832077026367, "learning_rate": 4.935554114282957e-06, "loss": 1.1062, "step": 25779 }, { "epoch": 2.020376175548589, "grad_norm": 2.148582935333252, "learning_rate": 4.9352457779623515e-06, "loss": 1.4697, "step": 25780 }, { "epoch": 2.0204545454545455, "grad_norm": 2.843282461166382, "learning_rate": 4.934937441888038e-06, "loss": 1.3083, "step": 25781 }, { "epoch": 2.0205329153605014, "grad_norm": 2.5138845443725586, "learning_rate": 4.934629106061188e-06, "loss": 1.5614, "step": 25782 }, { "epoch": 2.0206112852664577, "grad_norm": 7.14554500579834, "learning_rate": 4.934320770482974e-06, "loss": 1.4525, "step": 25783 }, { "epoch": 2.0206896551724136, "grad_norm": 2.4015161991119385, "learning_rate": 4.934012435154572e-06, "loss": 1.326, "step": 25784 }, { "epoch": 2.02076802507837, "grad_norm": 4.5723347663879395, "learning_rate": 4.933704100077153e-06, "loss": 1.0495, "step": 25785 }, { "epoch": 2.020846394984326, "grad_norm": 10.560503959655762, "learning_rate": 4.933395765251889e-06, "loss": 1.7441, "step": 25786 }, { "epoch": 2.020924764890282, "grad_norm": 5.4739603996276855, "learning_rate": 4.933087430679953e-06, "loss": 1.4221, "step": 25787 }, { "epoch": 2.021003134796238, "grad_norm": 2.7113046646118164, "learning_rate": 4.932779096362517e-06, "loss": 1.3343, "step": 25788 }, { "epoch": 2.0210815047021944, "grad_norm": 2.3677761554718018, "learning_rate": 4.932470762300759e-06, "loss": 1.0765, "step": 25789 }, { "epoch": 2.0211598746081503, "grad_norm": 7.815822124481201, "learning_rate": 4.9321624284958426e-06, "loss": 0.7822, "step": 25790 }, { "epoch": 2.0212382445141066, "grad_norm": 3.6868228912353516, "learning_rate": 4.931854094948948e-06, "loss": 1.4165, "step": 25791 }, { "epoch": 2.0213166144200625, "grad_norm": 3.072108507156372, "learning_rate": 4.931545761661244e-06, "loss": 1.3519, "step": 25792 }, { "epoch": 2.021394984326019, "grad_norm": 2.037367343902588, "learning_rate": 4.931237428633907e-06, "loss": 1.2995, "step": 25793 }, { "epoch": 2.0214733542319747, "grad_norm": 3.3842360973358154, "learning_rate": 4.930929095868106e-06, "loss": 1.1268, "step": 25794 }, { "epoch": 2.021551724137931, "grad_norm": 5.502960681915283, "learning_rate": 4.930620763365014e-06, "loss": 1.5747, "step": 25795 }, { "epoch": 2.021630094043887, "grad_norm": 2.7855477333068848, "learning_rate": 4.930312431125809e-06, "loss": 1.2504, "step": 25796 }, { "epoch": 2.0217084639498433, "grad_norm": 3.331782341003418, "learning_rate": 4.930004099151655e-06, "loss": 1.3911, "step": 25797 }, { "epoch": 2.021786833855799, "grad_norm": 3.7622971534729004, "learning_rate": 4.929695767443733e-06, "loss": 1.0942, "step": 25798 }, { "epoch": 2.0218652037617555, "grad_norm": 2.5107173919677734, "learning_rate": 4.92938743600321e-06, "loss": 1.42, "step": 25799 }, { "epoch": 2.0219435736677114, "grad_norm": 4.493052005767822, "learning_rate": 4.9290791048312624e-06, "loss": 0.8732, "step": 25800 }, { "epoch": 2.0220219435736677, "grad_norm": 4.035653591156006, "learning_rate": 4.92877077392906e-06, "loss": 1.0452, "step": 25801 }, { "epoch": 2.0221003134796236, "grad_norm": 3.9587602615356445, "learning_rate": 4.9284624432977775e-06, "loss": 1.4357, "step": 25802 }, { "epoch": 2.02217868338558, "grad_norm": 2.277223825454712, "learning_rate": 4.928154112938587e-06, "loss": 1.2723, "step": 25803 }, { "epoch": 2.0222570532915363, "grad_norm": 2.1610817909240723, "learning_rate": 4.927845782852663e-06, "loss": 0.9233, "step": 25804 }, { "epoch": 2.022335423197492, "grad_norm": 5.301385402679443, "learning_rate": 4.927537453041174e-06, "loss": 1.6486, "step": 25805 }, { "epoch": 2.0224137931034485, "grad_norm": 3.645707607269287, "learning_rate": 4.927229123505297e-06, "loss": 1.7695, "step": 25806 }, { "epoch": 2.0224921630094044, "grad_norm": 1.9306957721710205, "learning_rate": 4.926920794246202e-06, "loss": 1.4273, "step": 25807 }, { "epoch": 2.0225705329153607, "grad_norm": 3.0487282276153564, "learning_rate": 4.926612465265066e-06, "loss": 1.2812, "step": 25808 }, { "epoch": 2.0226489028213166, "grad_norm": 2.645581007003784, "learning_rate": 4.926304136563054e-06, "loss": 1.4818, "step": 25809 }, { "epoch": 2.022727272727273, "grad_norm": 2.5113844871520996, "learning_rate": 4.925995808141345e-06, "loss": 1.1281, "step": 25810 }, { "epoch": 2.022805642633229, "grad_norm": 2.3454933166503906, "learning_rate": 4.925687480001111e-06, "loss": 1.1599, "step": 25811 }, { "epoch": 2.022884012539185, "grad_norm": 1.7970157861709595, "learning_rate": 4.925379152143522e-06, "loss": 1.3272, "step": 25812 }, { "epoch": 2.022962382445141, "grad_norm": 3.969801187515259, "learning_rate": 4.925070824569753e-06, "loss": 1.6119, "step": 25813 }, { "epoch": 2.0230407523510974, "grad_norm": 2.5600240230560303, "learning_rate": 4.924762497280976e-06, "loss": 1.4767, "step": 25814 }, { "epoch": 2.0231191222570533, "grad_norm": 2.6520726680755615, "learning_rate": 4.924454170278364e-06, "loss": 1.4917, "step": 25815 }, { "epoch": 2.0231974921630096, "grad_norm": 2.8891119956970215, "learning_rate": 4.924145843563089e-06, "loss": 1.2228, "step": 25816 }, { "epoch": 2.0232758620689655, "grad_norm": 3.635493755340576, "learning_rate": 4.9238375171363255e-06, "loss": 1.3542, "step": 25817 }, { "epoch": 2.023354231974922, "grad_norm": 3.854626417160034, "learning_rate": 4.923529190999244e-06, "loss": 0.6407, "step": 25818 }, { "epoch": 2.0234326018808777, "grad_norm": 5.213845729827881, "learning_rate": 4.923220865153019e-06, "loss": 2.3302, "step": 25819 }, { "epoch": 2.023510971786834, "grad_norm": 2.9032604694366455, "learning_rate": 4.9229125395988216e-06, "loss": 1.3463, "step": 25820 }, { "epoch": 2.02358934169279, "grad_norm": 5.988729000091553, "learning_rate": 4.922604214337825e-06, "loss": 1.2171, "step": 25821 }, { "epoch": 2.0236677115987463, "grad_norm": 2.991530656814575, "learning_rate": 4.922295889371204e-06, "loss": 1.4115, "step": 25822 }, { "epoch": 2.023746081504702, "grad_norm": 6.069118976593018, "learning_rate": 4.921987564700128e-06, "loss": 1.0852, "step": 25823 }, { "epoch": 2.0238244514106585, "grad_norm": 2.4573867321014404, "learning_rate": 4.921679240325771e-06, "loss": 1.5797, "step": 25824 }, { "epoch": 2.0239028213166144, "grad_norm": 3.2997326850891113, "learning_rate": 4.921370916249306e-06, "loss": 1.1451, "step": 25825 }, { "epoch": 2.0239811912225707, "grad_norm": 5.452095031738281, "learning_rate": 4.9210625924719075e-06, "loss": 1.3365, "step": 25826 }, { "epoch": 2.0240595611285266, "grad_norm": 4.3657002449035645, "learning_rate": 4.920754268994744e-06, "loss": 1.5659, "step": 25827 }, { "epoch": 2.024137931034483, "grad_norm": 1.7864246368408203, "learning_rate": 4.9204459458189915e-06, "loss": 1.1713, "step": 25828 }, { "epoch": 2.024216300940439, "grad_norm": 2.937541961669922, "learning_rate": 4.920137622945821e-06, "loss": 1.3555, "step": 25829 }, { "epoch": 2.024294670846395, "grad_norm": 2.822070598602295, "learning_rate": 4.919829300376408e-06, "loss": 1.1179, "step": 25830 }, { "epoch": 2.024373040752351, "grad_norm": 7.828766345977783, "learning_rate": 4.919520978111922e-06, "loss": 1.0427, "step": 25831 }, { "epoch": 2.0244514106583074, "grad_norm": 2.9413561820983887, "learning_rate": 4.919212656153535e-06, "loss": 1.4381, "step": 25832 }, { "epoch": 2.0245297805642632, "grad_norm": 3.9669718742370605, "learning_rate": 4.918904334502424e-06, "loss": 0.9677, "step": 25833 }, { "epoch": 2.0246081504702196, "grad_norm": 3.42682147026062, "learning_rate": 4.9185960131597585e-06, "loss": 1.0681, "step": 25834 }, { "epoch": 2.0246865203761755, "grad_norm": 1.954198956489563, "learning_rate": 4.918287692126712e-06, "loss": 1.4775, "step": 25835 }, { "epoch": 2.024764890282132, "grad_norm": 2.9767825603485107, "learning_rate": 4.917979371404457e-06, "loss": 1.6281, "step": 25836 }, { "epoch": 2.0248432601880877, "grad_norm": 2.1641671657562256, "learning_rate": 4.917671050994168e-06, "loss": 1.5873, "step": 25837 }, { "epoch": 2.024921630094044, "grad_norm": 2.8329434394836426, "learning_rate": 4.917362730897014e-06, "loss": 1.4942, "step": 25838 }, { "epoch": 2.025, "grad_norm": 2.2738020420074463, "learning_rate": 4.91705441111417e-06, "loss": 1.2178, "step": 25839 }, { "epoch": 2.0250783699059562, "grad_norm": 2.6649484634399414, "learning_rate": 4.916746091646808e-06, "loss": 1.4041, "step": 25840 }, { "epoch": 2.025156739811912, "grad_norm": 2.729355812072754, "learning_rate": 4.916437772496102e-06, "loss": 1.3537, "step": 25841 }, { "epoch": 2.0252351097178685, "grad_norm": 3.438185930252075, "learning_rate": 4.9161294536632245e-06, "loss": 1.4732, "step": 25842 }, { "epoch": 2.0253134796238244, "grad_norm": 2.134579658508301, "learning_rate": 4.915821135149346e-06, "loss": 0.8476, "step": 25843 }, { "epoch": 2.0253918495297807, "grad_norm": 2.1944596767425537, "learning_rate": 4.91551281695564e-06, "loss": 1.6891, "step": 25844 }, { "epoch": 2.0254702194357366, "grad_norm": 7.839756488800049, "learning_rate": 4.915204499083283e-06, "loss": 1.3195, "step": 25845 }, { "epoch": 2.025548589341693, "grad_norm": 2.5657365322113037, "learning_rate": 4.914896181533442e-06, "loss": 1.6089, "step": 25846 }, { "epoch": 2.025626959247649, "grad_norm": 2.496408462524414, "learning_rate": 4.914587864307292e-06, "loss": 1.1309, "step": 25847 }, { "epoch": 2.025705329153605, "grad_norm": 2.3645853996276855, "learning_rate": 4.914279547406007e-06, "loss": 1.2285, "step": 25848 }, { "epoch": 2.025783699059561, "grad_norm": 2.012002468109131, "learning_rate": 4.913971230830759e-06, "loss": 1.6167, "step": 25849 }, { "epoch": 2.0258620689655173, "grad_norm": 2.9434282779693604, "learning_rate": 4.91366291458272e-06, "loss": 1.199, "step": 25850 }, { "epoch": 2.0259404388714732, "grad_norm": 2.3252217769622803, "learning_rate": 4.913354598663061e-06, "loss": 1.2436, "step": 25851 }, { "epoch": 2.0260188087774296, "grad_norm": 5.974715232849121, "learning_rate": 4.91304628307296e-06, "loss": 0.9359, "step": 25852 }, { "epoch": 2.0260971786833855, "grad_norm": 3.653514862060547, "learning_rate": 4.9127379678135825e-06, "loss": 0.7771, "step": 25853 }, { "epoch": 2.026175548589342, "grad_norm": 1.9900226593017578, "learning_rate": 4.912429652886107e-06, "loss": 1.2134, "step": 25854 }, { "epoch": 2.0262539184952977, "grad_norm": 3.3099920749664307, "learning_rate": 4.912121338291704e-06, "loss": 1.5506, "step": 25855 }, { "epoch": 2.026332288401254, "grad_norm": 2.9179186820983887, "learning_rate": 4.911813024031547e-06, "loss": 1.3854, "step": 25856 }, { "epoch": 2.02641065830721, "grad_norm": 2.7121620178222656, "learning_rate": 4.911504710106807e-06, "loss": 1.189, "step": 25857 }, { "epoch": 2.0264890282131662, "grad_norm": 2.9318439960479736, "learning_rate": 4.9111963965186575e-06, "loss": 1.4464, "step": 25858 }, { "epoch": 2.026567398119122, "grad_norm": 2.937187433242798, "learning_rate": 4.9108880832682705e-06, "loss": 0.9645, "step": 25859 }, { "epoch": 2.0266457680250785, "grad_norm": 3.2161753177642822, "learning_rate": 4.910579770356822e-06, "loss": 1.2062, "step": 25860 }, { "epoch": 2.0267241379310343, "grad_norm": 6.292026996612549, "learning_rate": 4.91027145778548e-06, "loss": 1.1153, "step": 25861 }, { "epoch": 2.0268025078369907, "grad_norm": 3.6495630741119385, "learning_rate": 4.909963145555421e-06, "loss": 1.377, "step": 25862 }, { "epoch": 2.0268808777429466, "grad_norm": 15.028717994689941, "learning_rate": 4.909654833667815e-06, "loss": 3.0297, "step": 25863 }, { "epoch": 2.026959247648903, "grad_norm": 3.3188085556030273, "learning_rate": 4.909346522123836e-06, "loss": 1.8136, "step": 25864 }, { "epoch": 2.027037617554859, "grad_norm": 3.499382734298706, "learning_rate": 4.9090382109246545e-06, "loss": 1.0752, "step": 25865 }, { "epoch": 2.027115987460815, "grad_norm": 4.228094100952148, "learning_rate": 4.908729900071445e-06, "loss": 1.4952, "step": 25866 }, { "epoch": 2.027194357366771, "grad_norm": 5.082976818084717, "learning_rate": 4.908421589565383e-06, "loss": 1.46, "step": 25867 }, { "epoch": 2.0272727272727273, "grad_norm": 2.8558151721954346, "learning_rate": 4.908113279407636e-06, "loss": 1.513, "step": 25868 }, { "epoch": 2.0273510971786832, "grad_norm": 2.6863718032836914, "learning_rate": 4.9078049695993795e-06, "loss": 1.3611, "step": 25869 }, { "epoch": 2.0274294670846396, "grad_norm": 2.6258950233459473, "learning_rate": 4.907496660141784e-06, "loss": 1.1002, "step": 25870 }, { "epoch": 2.0275078369905954, "grad_norm": 3.3924968242645264, "learning_rate": 4.907188351036028e-06, "loss": 1.6654, "step": 25871 }, { "epoch": 2.027586206896552, "grad_norm": 7.192987442016602, "learning_rate": 4.9068800422832756e-06, "loss": 0.9365, "step": 25872 }, { "epoch": 2.0276645768025077, "grad_norm": 3.375898838043213, "learning_rate": 4.906571733884704e-06, "loss": 1.164, "step": 25873 }, { "epoch": 2.027742946708464, "grad_norm": 4.140169143676758, "learning_rate": 4.906263425841487e-06, "loss": 1.0639, "step": 25874 }, { "epoch": 2.02782131661442, "grad_norm": 4.42486572265625, "learning_rate": 4.905955118154796e-06, "loss": 1.5881, "step": 25875 }, { "epoch": 2.027899686520376, "grad_norm": 4.680102825164795, "learning_rate": 4.905646810825803e-06, "loss": 1.3062, "step": 25876 }, { "epoch": 2.027978056426332, "grad_norm": 3.253614664077759, "learning_rate": 4.90533850385568e-06, "loss": 1.4167, "step": 25877 }, { "epoch": 2.0280564263322884, "grad_norm": 3.810643196105957, "learning_rate": 4.905030197245604e-06, "loss": 1.4632, "step": 25878 }, { "epoch": 2.0281347962382443, "grad_norm": 4.764348030090332, "learning_rate": 4.90472189099674e-06, "loss": 1.3819, "step": 25879 }, { "epoch": 2.0282131661442007, "grad_norm": 3.291396141052246, "learning_rate": 4.904413585110267e-06, "loss": 1.1838, "step": 25880 }, { "epoch": 2.0282915360501566, "grad_norm": 6.645533561706543, "learning_rate": 4.904105279587356e-06, "loss": 1.3231, "step": 25881 }, { "epoch": 2.028369905956113, "grad_norm": 7.267966270446777, "learning_rate": 4.9037969744291795e-06, "loss": 1.0555, "step": 25882 }, { "epoch": 2.0284482758620688, "grad_norm": 2.763491630554199, "learning_rate": 4.9034886696369095e-06, "loss": 1.2511, "step": 25883 }, { "epoch": 2.028526645768025, "grad_norm": 6.764620780944824, "learning_rate": 4.903180365211717e-06, "loss": 1.7102, "step": 25884 }, { "epoch": 2.028605015673981, "grad_norm": 6.038356781005859, "learning_rate": 4.902872061154779e-06, "loss": 1.7073, "step": 25885 }, { "epoch": 2.0286833855799373, "grad_norm": 6.701714038848877, "learning_rate": 4.902563757467266e-06, "loss": 1.3163, "step": 25886 }, { "epoch": 2.028761755485893, "grad_norm": 3.4887149333953857, "learning_rate": 4.902255454150349e-06, "loss": 1.3024, "step": 25887 }, { "epoch": 2.0288401253918495, "grad_norm": 2.9257988929748535, "learning_rate": 4.901947151205203e-06, "loss": 1.3336, "step": 25888 }, { "epoch": 2.0289184952978054, "grad_norm": 2.480790853500366, "learning_rate": 4.901638848632998e-06, "loss": 1.17, "step": 25889 }, { "epoch": 2.0289968652037618, "grad_norm": 1.9712533950805664, "learning_rate": 4.901330546434911e-06, "loss": 1.2347, "step": 25890 }, { "epoch": 2.0290752351097177, "grad_norm": 2.665187120437622, "learning_rate": 4.90102224461211e-06, "loss": 1.4742, "step": 25891 }, { "epoch": 2.029153605015674, "grad_norm": 2.6426243782043457, "learning_rate": 4.90071394316577e-06, "loss": 1.5372, "step": 25892 }, { "epoch": 2.02923197492163, "grad_norm": 6.6936798095703125, "learning_rate": 4.9004056420970645e-06, "loss": 1.1854, "step": 25893 }, { "epoch": 2.029310344827586, "grad_norm": 3.616377592086792, "learning_rate": 4.900097341407163e-06, "loss": 1.4379, "step": 25894 }, { "epoch": 2.0293887147335425, "grad_norm": 2.9358205795288086, "learning_rate": 4.899789041097241e-06, "loss": 1.3983, "step": 25895 }, { "epoch": 2.0294670846394984, "grad_norm": 1.909588098526001, "learning_rate": 4.899480741168468e-06, "loss": 1.6276, "step": 25896 }, { "epoch": 2.0295454545454548, "grad_norm": 3.3639907836914062, "learning_rate": 4.899172441622022e-06, "loss": 1.4657, "step": 25897 }, { "epoch": 2.0296238244514107, "grad_norm": 3.5081722736358643, "learning_rate": 4.898864142459069e-06, "loss": 1.326, "step": 25898 }, { "epoch": 2.029702194357367, "grad_norm": 2.9178225994110107, "learning_rate": 4.898555843680786e-06, "loss": 1.1868, "step": 25899 }, { "epoch": 2.029780564263323, "grad_norm": 4.366328716278076, "learning_rate": 4.898247545288345e-06, "loss": 1.5284, "step": 25900 }, { "epoch": 2.029858934169279, "grad_norm": 3.3291642665863037, "learning_rate": 4.897939247282918e-06, "loss": 2.0231, "step": 25901 }, { "epoch": 2.029937304075235, "grad_norm": 5.532717227935791, "learning_rate": 4.897630949665678e-06, "loss": 1.2071, "step": 25902 }, { "epoch": 2.0300156739811914, "grad_norm": 3.7516512870788574, "learning_rate": 4.897322652437795e-06, "loss": 1.5281, "step": 25903 }, { "epoch": 2.0300940438871473, "grad_norm": 4.453003406524658, "learning_rate": 4.897014355600446e-06, "loss": 0.8063, "step": 25904 }, { "epoch": 2.0301724137931036, "grad_norm": 4.475782871246338, "learning_rate": 4.896706059154802e-06, "loss": 1.5287, "step": 25905 }, { "epoch": 2.0302507836990595, "grad_norm": 3.480133295059204, "learning_rate": 4.896397763102034e-06, "loss": 1.0403, "step": 25906 }, { "epoch": 2.030329153605016, "grad_norm": 4.323653221130371, "learning_rate": 4.8960894674433155e-06, "loss": 1.5942, "step": 25907 }, { "epoch": 2.0304075235109718, "grad_norm": 2.800168514251709, "learning_rate": 4.895781172179821e-06, "loss": 1.3623, "step": 25908 }, { "epoch": 2.030485893416928, "grad_norm": 5.4100751876831055, "learning_rate": 4.89547287731272e-06, "loss": 1.3192, "step": 25909 }, { "epoch": 2.030564263322884, "grad_norm": 2.9866697788238525, "learning_rate": 4.8951645828431855e-06, "loss": 0.9631, "step": 25910 }, { "epoch": 2.0306426332288403, "grad_norm": 9.79737377166748, "learning_rate": 4.894856288772392e-06, "loss": 1.32, "step": 25911 }, { "epoch": 2.030721003134796, "grad_norm": 3.1237030029296875, "learning_rate": 4.894547995101513e-06, "loss": 1.3279, "step": 25912 }, { "epoch": 2.0307993730407525, "grad_norm": 2.181504726409912, "learning_rate": 4.8942397018317165e-06, "loss": 1.3757, "step": 25913 }, { "epoch": 2.0308777429467084, "grad_norm": 2.49826717376709, "learning_rate": 4.893931408964179e-06, "loss": 1.5607, "step": 25914 }, { "epoch": 2.0309561128526648, "grad_norm": 3.3921456336975098, "learning_rate": 4.89362311650007e-06, "loss": 1.0234, "step": 25915 }, { "epoch": 2.0310344827586206, "grad_norm": 3.9294745922088623, "learning_rate": 4.893314824440569e-06, "loss": 1.5181, "step": 25916 }, { "epoch": 2.031112852664577, "grad_norm": 3.8023643493652344, "learning_rate": 4.8930065327868385e-06, "loss": 1.6729, "step": 25917 }, { "epoch": 2.031191222570533, "grad_norm": 3.1271378993988037, "learning_rate": 4.8926982415400585e-06, "loss": 1.3511, "step": 25918 }, { "epoch": 2.031269592476489, "grad_norm": 2.843451738357544, "learning_rate": 4.8923899507014e-06, "loss": 1.2429, "step": 25919 }, { "epoch": 2.031347962382445, "grad_norm": 5.015377998352051, "learning_rate": 4.892081660272033e-06, "loss": 0.8773, "step": 25920 }, { "epoch": 2.0314263322884014, "grad_norm": 2.9257519245147705, "learning_rate": 4.891773370253133e-06, "loss": 1.8044, "step": 25921 }, { "epoch": 2.0315047021943573, "grad_norm": 2.5069785118103027, "learning_rate": 4.89146508064587e-06, "loss": 1.3099, "step": 25922 }, { "epoch": 2.0315830721003136, "grad_norm": 2.410628318786621, "learning_rate": 4.891156791451421e-06, "loss": 1.1474, "step": 25923 }, { "epoch": 2.0316614420062695, "grad_norm": 5.8994855880737305, "learning_rate": 4.890848502670953e-06, "loss": 1.3096, "step": 25924 }, { "epoch": 2.031739811912226, "grad_norm": 2.4342005252838135, "learning_rate": 4.890540214305642e-06, "loss": 1.4705, "step": 25925 }, { "epoch": 2.0318181818181817, "grad_norm": 5.182249069213867, "learning_rate": 4.89023192635666e-06, "loss": 1.6013, "step": 25926 }, { "epoch": 2.031896551724138, "grad_norm": 3.4877610206604004, "learning_rate": 4.889923638825181e-06, "loss": 1.5126, "step": 25927 }, { "epoch": 2.031974921630094, "grad_norm": 3.7833073139190674, "learning_rate": 4.889615351712373e-06, "loss": 1.251, "step": 25928 }, { "epoch": 2.0320532915360503, "grad_norm": 4.529951095581055, "learning_rate": 4.889307065019413e-06, "loss": 1.5108, "step": 25929 }, { "epoch": 2.032131661442006, "grad_norm": 2.682692289352417, "learning_rate": 4.8889987787474716e-06, "loss": 1.491, "step": 25930 }, { "epoch": 2.0322100313479625, "grad_norm": 4.669649600982666, "learning_rate": 4.888690492897724e-06, "loss": 1.3903, "step": 25931 }, { "epoch": 2.0322884012539184, "grad_norm": 2.287371873855591, "learning_rate": 4.888382207471337e-06, "loss": 1.2268, "step": 25932 }, { "epoch": 2.0323667711598747, "grad_norm": 3.5699942111968994, "learning_rate": 4.88807392246949e-06, "loss": 1.5726, "step": 25933 }, { "epoch": 2.0324451410658306, "grad_norm": 8.662333488464355, "learning_rate": 4.887765637893351e-06, "loss": 0.6149, "step": 25934 }, { "epoch": 2.032523510971787, "grad_norm": 2.708064556121826, "learning_rate": 4.887457353744093e-06, "loss": 1.7315, "step": 25935 }, { "epoch": 2.032601880877743, "grad_norm": 2.1311473846435547, "learning_rate": 4.88714907002289e-06, "loss": 1.0186, "step": 25936 }, { "epoch": 2.032680250783699, "grad_norm": 5.497957229614258, "learning_rate": 4.886840786730914e-06, "loss": 1.4954, "step": 25937 }, { "epoch": 2.032758620689655, "grad_norm": 4.069248676300049, "learning_rate": 4.886532503869338e-06, "loss": 1.7726, "step": 25938 }, { "epoch": 2.0328369905956114, "grad_norm": 2.80269718170166, "learning_rate": 4.886224221439334e-06, "loss": 1.4137, "step": 25939 }, { "epoch": 2.0329153605015673, "grad_norm": 6.032202243804932, "learning_rate": 4.885915939442073e-06, "loss": 1.342, "step": 25940 }, { "epoch": 2.0329937304075236, "grad_norm": 3.688852310180664, "learning_rate": 4.88560765787873e-06, "loss": 1.7074, "step": 25941 }, { "epoch": 2.0330721003134795, "grad_norm": 3.7203049659729004, "learning_rate": 4.885299376750479e-06, "loss": 1.4015, "step": 25942 }, { "epoch": 2.033150470219436, "grad_norm": 2.281670570373535, "learning_rate": 4.884991096058488e-06, "loss": 1.231, "step": 25943 }, { "epoch": 2.0332288401253917, "grad_norm": 3.179629325866699, "learning_rate": 4.8846828158039326e-06, "loss": 1.1821, "step": 25944 }, { "epoch": 2.033307210031348, "grad_norm": 4.141936779022217, "learning_rate": 4.8843745359879824e-06, "loss": 1.4147, "step": 25945 }, { "epoch": 2.033385579937304, "grad_norm": 4.4787116050720215, "learning_rate": 4.884066256611816e-06, "loss": 1.5683, "step": 25946 }, { "epoch": 2.0334639498432603, "grad_norm": 2.3553481101989746, "learning_rate": 4.883757977676599e-06, "loss": 1.0776, "step": 25947 }, { "epoch": 2.033542319749216, "grad_norm": 1.8829452991485596, "learning_rate": 4.883449699183508e-06, "loss": 1.1986, "step": 25948 }, { "epoch": 2.0336206896551725, "grad_norm": 3.6577699184417725, "learning_rate": 4.8831414211337144e-06, "loss": 1.1723, "step": 25949 }, { "epoch": 2.0336990595611284, "grad_norm": 2.89701771736145, "learning_rate": 4.882833143528391e-06, "loss": 1.7377, "step": 25950 }, { "epoch": 2.0337774294670847, "grad_norm": 2.248581886291504, "learning_rate": 4.88252486636871e-06, "loss": 1.178, "step": 25951 }, { "epoch": 2.0338557993730406, "grad_norm": 2.660724401473999, "learning_rate": 4.882216589655842e-06, "loss": 0.9412, "step": 25952 }, { "epoch": 2.033934169278997, "grad_norm": 3.9923694133758545, "learning_rate": 4.881908313390966e-06, "loss": 1.5631, "step": 25953 }, { "epoch": 2.034012539184953, "grad_norm": 2.8207204341888428, "learning_rate": 4.881600037575246e-06, "loss": 1.3917, "step": 25954 }, { "epoch": 2.034090909090909, "grad_norm": 2.894360303878784, "learning_rate": 4.881291762209861e-06, "loss": 1.5051, "step": 25955 }, { "epoch": 2.034169278996865, "grad_norm": 11.824026107788086, "learning_rate": 4.88098348729598e-06, "loss": 0.9606, "step": 25956 }, { "epoch": 2.0342476489028214, "grad_norm": 9.512093544006348, "learning_rate": 4.880675212834778e-06, "loss": 1.458, "step": 25957 }, { "epoch": 2.0343260188087773, "grad_norm": 4.920641899108887, "learning_rate": 4.880366938827425e-06, "loss": 1.4082, "step": 25958 }, { "epoch": 2.0344043887147336, "grad_norm": 12.125406265258789, "learning_rate": 4.880058665275094e-06, "loss": 1.4552, "step": 25959 }, { "epoch": 2.0344827586206895, "grad_norm": 3.99906063079834, "learning_rate": 4.879750392178959e-06, "loss": 1.6052, "step": 25960 }, { "epoch": 2.034561128526646, "grad_norm": 5.838871002197266, "learning_rate": 4.879442119540193e-06, "loss": 1.2211, "step": 25961 }, { "epoch": 2.0346394984326017, "grad_norm": 6.593120098114014, "learning_rate": 4.879133847359966e-06, "loss": 1.0303, "step": 25962 }, { "epoch": 2.034717868338558, "grad_norm": 2.452972888946533, "learning_rate": 4.878825575639452e-06, "loss": 1.2247, "step": 25963 }, { "epoch": 2.034796238244514, "grad_norm": 4.008107662200928, "learning_rate": 4.878517304379825e-06, "loss": 1.1782, "step": 25964 }, { "epoch": 2.0348746081504703, "grad_norm": 2.34651255607605, "learning_rate": 4.878209033582253e-06, "loss": 1.3511, "step": 25965 }, { "epoch": 2.034952978056426, "grad_norm": 6.246417045593262, "learning_rate": 4.877900763247911e-06, "loss": 1.0188, "step": 25966 }, { "epoch": 2.0350313479623825, "grad_norm": 4.561275005340576, "learning_rate": 4.8775924933779735e-06, "loss": 1.5659, "step": 25967 }, { "epoch": 2.0351097178683384, "grad_norm": 4.247203350067139, "learning_rate": 4.877284223973613e-06, "loss": 1.1726, "step": 25968 }, { "epoch": 2.0351880877742947, "grad_norm": 4.51232385635376, "learning_rate": 4.876975955035997e-06, "loss": 0.7239, "step": 25969 }, { "epoch": 2.0352664576802506, "grad_norm": 2.503394603729248, "learning_rate": 4.876667686566303e-06, "loss": 1.4447, "step": 25970 }, { "epoch": 2.035344827586207, "grad_norm": 12.157805442810059, "learning_rate": 4.8763594185657e-06, "loss": 1.3709, "step": 25971 }, { "epoch": 2.035423197492163, "grad_norm": 2.8118622303009033, "learning_rate": 4.8760511510353655e-06, "loss": 1.4542, "step": 25972 }, { "epoch": 2.035501567398119, "grad_norm": 3.8797404766082764, "learning_rate": 4.875742883976465e-06, "loss": 1.5829, "step": 25973 }, { "epoch": 2.035579937304075, "grad_norm": 2.3209547996520996, "learning_rate": 4.875434617390177e-06, "loss": 1.4358, "step": 25974 }, { "epoch": 2.0356583072100314, "grad_norm": 2.0277156829833984, "learning_rate": 4.875126351277672e-06, "loss": 1.2232, "step": 25975 }, { "epoch": 2.0357366771159873, "grad_norm": 2.6586408615112305, "learning_rate": 4.874818085640122e-06, "loss": 1.2902, "step": 25976 }, { "epoch": 2.0358150470219436, "grad_norm": 6.960887908935547, "learning_rate": 4.874509820478698e-06, "loss": 1.8576, "step": 25977 }, { "epoch": 2.0358934169278995, "grad_norm": 2.6998701095581055, "learning_rate": 4.874201555794575e-06, "loss": 0.7017, "step": 25978 }, { "epoch": 2.035971786833856, "grad_norm": 3.157723903656006, "learning_rate": 4.873893291588927e-06, "loss": 1.3523, "step": 25979 }, { "epoch": 2.0360501567398117, "grad_norm": 6.842494010925293, "learning_rate": 4.873585027862921e-06, "loss": 1.2241, "step": 25980 }, { "epoch": 2.036128526645768, "grad_norm": 3.8076863288879395, "learning_rate": 4.873276764617734e-06, "loss": 0.9743, "step": 25981 }, { "epoch": 2.036206896551724, "grad_norm": 3.3893792629241943, "learning_rate": 4.8729685018545366e-06, "loss": 1.3398, "step": 25982 }, { "epoch": 2.0362852664576803, "grad_norm": 3.0493323802948, "learning_rate": 4.872660239574503e-06, "loss": 1.4262, "step": 25983 }, { "epoch": 2.036363636363636, "grad_norm": 3.907041072845459, "learning_rate": 4.8723519777788035e-06, "loss": 1.0328, "step": 25984 }, { "epoch": 2.0364420062695925, "grad_norm": 2.670689105987549, "learning_rate": 4.87204371646861e-06, "loss": 1.3919, "step": 25985 }, { "epoch": 2.0365203761755484, "grad_norm": 6.8541741371154785, "learning_rate": 4.871735455645098e-06, "loss": 1.1089, "step": 25986 }, { "epoch": 2.0365987460815047, "grad_norm": 2.6567726135253906, "learning_rate": 4.871427195309439e-06, "loss": 1.548, "step": 25987 }, { "epoch": 2.0366771159874606, "grad_norm": 4.037927627563477, "learning_rate": 4.871118935462805e-06, "loss": 1.4445, "step": 25988 }, { "epoch": 2.036755485893417, "grad_norm": 3.5558922290802, "learning_rate": 4.870810676106367e-06, "loss": 1.4778, "step": 25989 }, { "epoch": 2.0368338557993733, "grad_norm": 9.89084529876709, "learning_rate": 4.870502417241301e-06, "loss": 1.9104, "step": 25990 }, { "epoch": 2.036912225705329, "grad_norm": 3.9675471782684326, "learning_rate": 4.870194158868776e-06, "loss": 1.5922, "step": 25991 }, { "epoch": 2.0369905956112855, "grad_norm": 2.668494701385498, "learning_rate": 4.869885900989964e-06, "loss": 0.9482, "step": 25992 }, { "epoch": 2.0370689655172414, "grad_norm": 6.591489791870117, "learning_rate": 4.869577643606041e-06, "loss": 0.8371, "step": 25993 }, { "epoch": 2.0371473354231977, "grad_norm": 3.2488176822662354, "learning_rate": 4.869269386718179e-06, "loss": 1.6404, "step": 25994 }, { "epoch": 2.0372257053291536, "grad_norm": 4.148835182189941, "learning_rate": 4.8689611303275465e-06, "loss": 1.4955, "step": 25995 }, { "epoch": 2.03730407523511, "grad_norm": 11.837851524353027, "learning_rate": 4.86865287443532e-06, "loss": 1.705, "step": 25996 }, { "epoch": 2.037382445141066, "grad_norm": 4.4927287101745605, "learning_rate": 4.868344619042668e-06, "loss": 1.4699, "step": 25997 }, { "epoch": 2.037460815047022, "grad_norm": 8.708577156066895, "learning_rate": 4.86803636415077e-06, "loss": 1.1602, "step": 25998 }, { "epoch": 2.037539184952978, "grad_norm": 2.28161358833313, "learning_rate": 4.867728109760791e-06, "loss": 1.3617, "step": 25999 }, { "epoch": 2.0376175548589344, "grad_norm": 5.2855544090271, "learning_rate": 4.867419855873907e-06, "loss": 2.0161, "step": 26000 }, { "epoch": 2.0376959247648903, "grad_norm": 3.031254768371582, "learning_rate": 4.867111602491289e-06, "loss": 1.4106, "step": 26001 }, { "epoch": 2.0377742946708466, "grad_norm": 2.791415214538574, "learning_rate": 4.8668033496141115e-06, "loss": 1.1836, "step": 26002 }, { "epoch": 2.0378526645768025, "grad_norm": 2.705890655517578, "learning_rate": 4.866495097243545e-06, "loss": 1.2621, "step": 26003 }, { "epoch": 2.037931034482759, "grad_norm": 4.960422039031982, "learning_rate": 4.866186845380761e-06, "loss": 1.1397, "step": 26004 }, { "epoch": 2.0380094043887147, "grad_norm": 2.861456871032715, "learning_rate": 4.865878594026938e-06, "loss": 1.5632, "step": 26005 }, { "epoch": 2.038087774294671, "grad_norm": 4.077563285827637, "learning_rate": 4.865570343183239e-06, "loss": 1.3956, "step": 26006 }, { "epoch": 2.038166144200627, "grad_norm": 2.752326726913452, "learning_rate": 4.865262092850843e-06, "loss": 1.1498, "step": 26007 }, { "epoch": 2.0382445141065832, "grad_norm": 3.841122627258301, "learning_rate": 4.864953843030921e-06, "loss": 1.4811, "step": 26008 }, { "epoch": 2.038322884012539, "grad_norm": 5.349506855010986, "learning_rate": 4.8646455937246465e-06, "loss": 1.41, "step": 26009 }, { "epoch": 2.0384012539184955, "grad_norm": 5.386584281921387, "learning_rate": 4.86433734493319e-06, "loss": 1.5623, "step": 26010 }, { "epoch": 2.0384796238244514, "grad_norm": 3.888171434402466, "learning_rate": 4.864029096657723e-06, "loss": 1.4167, "step": 26011 }, { "epoch": 2.0385579937304077, "grad_norm": 3.386561870574951, "learning_rate": 4.86372084889942e-06, "loss": 1.3664, "step": 26012 }, { "epoch": 2.0386363636363636, "grad_norm": 4.793684959411621, "learning_rate": 4.863412601659454e-06, "loss": 0.6157, "step": 26013 }, { "epoch": 2.03871473354232, "grad_norm": 5.797109127044678, "learning_rate": 4.863104354938996e-06, "loss": 1.5041, "step": 26014 }, { "epoch": 2.038793103448276, "grad_norm": 6.833901882171631, "learning_rate": 4.8627961087392185e-06, "loss": 1.5167, "step": 26015 }, { "epoch": 2.038871473354232, "grad_norm": 3.6164536476135254, "learning_rate": 4.862487863061292e-06, "loss": 1.1486, "step": 26016 }, { "epoch": 2.038949843260188, "grad_norm": 3.4009015560150146, "learning_rate": 4.8621796179063954e-06, "loss": 0.7488, "step": 26017 }, { "epoch": 2.0390282131661444, "grad_norm": 5.08592414855957, "learning_rate": 4.861871373275693e-06, "loss": 1.9798, "step": 26018 }, { "epoch": 2.0391065830721002, "grad_norm": 11.567686080932617, "learning_rate": 4.861563129170363e-06, "loss": 1.4254, "step": 26019 }, { "epoch": 2.0391849529780566, "grad_norm": 3.106234312057495, "learning_rate": 4.8612548855915755e-06, "loss": 1.9129, "step": 26020 }, { "epoch": 2.0392633228840125, "grad_norm": 3.728956937789917, "learning_rate": 4.8609466425405035e-06, "loss": 1.5047, "step": 26021 }, { "epoch": 2.039341692789969, "grad_norm": 1.8243200778961182, "learning_rate": 4.860638400018318e-06, "loss": 1.162, "step": 26022 }, { "epoch": 2.0394200626959247, "grad_norm": 3.2639172077178955, "learning_rate": 4.860330158026191e-06, "loss": 1.3906, "step": 26023 }, { "epoch": 2.039498432601881, "grad_norm": 3.1591055393218994, "learning_rate": 4.860021916565299e-06, "loss": 1.205, "step": 26024 }, { "epoch": 2.039576802507837, "grad_norm": 2.98655366897583, "learning_rate": 4.859713675636811e-06, "loss": 1.4642, "step": 26025 }, { "epoch": 2.0396551724137932, "grad_norm": 8.402037620544434, "learning_rate": 4.8594054352419e-06, "loss": 1.0803, "step": 26026 }, { "epoch": 2.039733542319749, "grad_norm": 3.1395456790924072, "learning_rate": 4.859097195381738e-06, "loss": 1.5019, "step": 26027 }, { "epoch": 2.0398119122257055, "grad_norm": 4.173777103424072, "learning_rate": 4.8587889560575e-06, "loss": 1.5776, "step": 26028 }, { "epoch": 2.0398902821316613, "grad_norm": 2.5138278007507324, "learning_rate": 4.858480717270354e-06, "loss": 1.2265, "step": 26029 }, { "epoch": 2.0399686520376177, "grad_norm": 4.8519439697265625, "learning_rate": 4.858172479021475e-06, "loss": 1.7677, "step": 26030 }, { "epoch": 2.0400470219435736, "grad_norm": 3.318868637084961, "learning_rate": 4.857864241312036e-06, "loss": 1.2354, "step": 26031 }, { "epoch": 2.04012539184953, "grad_norm": 2.2021095752716064, "learning_rate": 4.857556004143208e-06, "loss": 1.4719, "step": 26032 }, { "epoch": 2.040203761755486, "grad_norm": 5.011488437652588, "learning_rate": 4.857247767516163e-06, "loss": 1.5109, "step": 26033 }, { "epoch": 2.040282131661442, "grad_norm": 2.736834764480591, "learning_rate": 4.856939531432075e-06, "loss": 1.2738, "step": 26034 }, { "epoch": 2.040360501567398, "grad_norm": 9.000960350036621, "learning_rate": 4.856631295892116e-06, "loss": 1.0522, "step": 26035 }, { "epoch": 2.0404388714733543, "grad_norm": 2.415696382522583, "learning_rate": 4.856323060897456e-06, "loss": 1.2644, "step": 26036 }, { "epoch": 2.0405172413793102, "grad_norm": 2.4966750144958496, "learning_rate": 4.8560148264492705e-06, "loss": 1.4649, "step": 26037 }, { "epoch": 2.0405956112852666, "grad_norm": 2.47025990486145, "learning_rate": 4.855706592548731e-06, "loss": 1.4667, "step": 26038 }, { "epoch": 2.0406739811912225, "grad_norm": 6.5671868324279785, "learning_rate": 4.85539835919701e-06, "loss": 0.5066, "step": 26039 }, { "epoch": 2.040752351097179, "grad_norm": 3.6360151767730713, "learning_rate": 4.855090126395278e-06, "loss": 1.3626, "step": 26040 }, { "epoch": 2.0408307210031347, "grad_norm": 9.476937294006348, "learning_rate": 4.854781894144708e-06, "loss": 1.2424, "step": 26041 }, { "epoch": 2.040909090909091, "grad_norm": 4.563803672790527, "learning_rate": 4.854473662446475e-06, "loss": 1.746, "step": 26042 }, { "epoch": 2.040987460815047, "grad_norm": 2.982407569885254, "learning_rate": 4.854165431301749e-06, "loss": 1.6674, "step": 26043 }, { "epoch": 2.0410658307210032, "grad_norm": 3.5390255451202393, "learning_rate": 4.8538572007117025e-06, "loss": 0.8835, "step": 26044 }, { "epoch": 2.041144200626959, "grad_norm": 2.34238862991333, "learning_rate": 4.853548970677508e-06, "loss": 1.4742, "step": 26045 }, { "epoch": 2.0412225705329154, "grad_norm": 3.188497543334961, "learning_rate": 4.853240741200339e-06, "loss": 1.7472, "step": 26046 }, { "epoch": 2.0413009404388713, "grad_norm": 2.5165107250213623, "learning_rate": 4.8529325122813665e-06, "loss": 1.3317, "step": 26047 }, { "epoch": 2.0413793103448277, "grad_norm": 2.995307207107544, "learning_rate": 4.852624283921762e-06, "loss": 1.1565, "step": 26048 }, { "epoch": 2.0414576802507836, "grad_norm": 2.4566280841827393, "learning_rate": 4.8523160561226986e-06, "loss": 1.4874, "step": 26049 }, { "epoch": 2.04153605015674, "grad_norm": 2.318803310394287, "learning_rate": 4.852007828885351e-06, "loss": 1.0624, "step": 26050 }, { "epoch": 2.041614420062696, "grad_norm": 2.4944565296173096, "learning_rate": 4.851699602210889e-06, "loss": 1.2425, "step": 26051 }, { "epoch": 2.041692789968652, "grad_norm": 2.135343074798584, "learning_rate": 4.851391376100485e-06, "loss": 1.3509, "step": 26052 }, { "epoch": 2.041771159874608, "grad_norm": 4.985097408294678, "learning_rate": 4.8510831505553115e-06, "loss": 1.8173, "step": 26053 }, { "epoch": 2.0418495297805643, "grad_norm": 3.1042065620422363, "learning_rate": 4.850774925576544e-06, "loss": 1.4385, "step": 26054 }, { "epoch": 2.04192789968652, "grad_norm": 2.9945931434631348, "learning_rate": 4.850466701165349e-06, "loss": 1.3688, "step": 26055 }, { "epoch": 2.0420062695924766, "grad_norm": 2.2661585807800293, "learning_rate": 4.850158477322902e-06, "loss": 1.0008, "step": 26056 }, { "epoch": 2.0420846394984324, "grad_norm": 6.329539775848389, "learning_rate": 4.8498502540503765e-06, "loss": 1.6699, "step": 26057 }, { "epoch": 2.0421630094043888, "grad_norm": 2.3180596828460693, "learning_rate": 4.849542031348944e-06, "loss": 1.2592, "step": 26058 }, { "epoch": 2.0422413793103447, "grad_norm": 3.8321216106414795, "learning_rate": 4.849233809219776e-06, "loss": 1.214, "step": 26059 }, { "epoch": 2.042319749216301, "grad_norm": 3.3118748664855957, "learning_rate": 4.8489255876640425e-06, "loss": 1.5259, "step": 26060 }, { "epoch": 2.042398119122257, "grad_norm": 2.834172248840332, "learning_rate": 4.848617366682923e-06, "loss": 1.0005, "step": 26061 }, { "epoch": 2.042476489028213, "grad_norm": 3.580235481262207, "learning_rate": 4.848309146277582e-06, "loss": 1.202, "step": 26062 }, { "epoch": 2.042554858934169, "grad_norm": 3.8747406005859375, "learning_rate": 4.848000926449195e-06, "loss": 1.4836, "step": 26063 }, { "epoch": 2.0426332288401254, "grad_norm": 3.127202033996582, "learning_rate": 4.847692707198935e-06, "loss": 1.1207, "step": 26064 }, { "epoch": 2.0427115987460813, "grad_norm": 4.556969165802002, "learning_rate": 4.847384488527975e-06, "loss": 1.0584, "step": 26065 }, { "epoch": 2.0427899686520377, "grad_norm": 3.4820053577423096, "learning_rate": 4.847076270437485e-06, "loss": 1.6683, "step": 26066 }, { "epoch": 2.0428683385579935, "grad_norm": 2.4831020832061768, "learning_rate": 4.8467680529286374e-06, "loss": 1.2045, "step": 26067 }, { "epoch": 2.04294670846395, "grad_norm": 2.9080846309661865, "learning_rate": 4.846459836002607e-06, "loss": 1.2725, "step": 26068 }, { "epoch": 2.0430250783699058, "grad_norm": 4.177947044372559, "learning_rate": 4.846151619660564e-06, "loss": 1.3104, "step": 26069 }, { "epoch": 2.043103448275862, "grad_norm": 2.5451135635375977, "learning_rate": 4.845843403903681e-06, "loss": 1.5279, "step": 26070 }, { "epoch": 2.043181818181818, "grad_norm": 4.1766180992126465, "learning_rate": 4.8455351887331305e-06, "loss": 1.0623, "step": 26071 }, { "epoch": 2.0432601880877743, "grad_norm": 3.5561635494232178, "learning_rate": 4.845226974150084e-06, "loss": 1.291, "step": 26072 }, { "epoch": 2.04333855799373, "grad_norm": 1.9312820434570312, "learning_rate": 4.844918760155717e-06, "loss": 1.479, "step": 26073 }, { "epoch": 2.0434169278996865, "grad_norm": 2.6087231636047363, "learning_rate": 4.844610546751197e-06, "loss": 1.4492, "step": 26074 }, { "epoch": 2.0434952978056424, "grad_norm": 4.136590003967285, "learning_rate": 4.8443023339377e-06, "loss": 1.7044, "step": 26075 }, { "epoch": 2.0435736677115988, "grad_norm": 2.5021493434906006, "learning_rate": 4.8439941217163975e-06, "loss": 1.3735, "step": 26076 }, { "epoch": 2.0436520376175547, "grad_norm": 3.3205044269561768, "learning_rate": 4.84368591008846e-06, "loss": 1.6065, "step": 26077 }, { "epoch": 2.043730407523511, "grad_norm": 3.295062780380249, "learning_rate": 4.843377699055061e-06, "loss": 1.1952, "step": 26078 }, { "epoch": 2.043808777429467, "grad_norm": 2.843040704727173, "learning_rate": 4.8430694886173715e-06, "loss": 1.3063, "step": 26079 }, { "epoch": 2.043887147335423, "grad_norm": 2.1147475242614746, "learning_rate": 4.842761278776569e-06, "loss": 0.916, "step": 26080 }, { "epoch": 2.043965517241379, "grad_norm": 3.379232168197632, "learning_rate": 4.842453069533818e-06, "loss": 1.3817, "step": 26081 }, { "epoch": 2.0440438871473354, "grad_norm": 2.2483205795288086, "learning_rate": 4.842144860890296e-06, "loss": 1.0039, "step": 26082 }, { "epoch": 2.0441222570532913, "grad_norm": 6.079575538635254, "learning_rate": 4.841836652847174e-06, "loss": 1.2121, "step": 26083 }, { "epoch": 2.0442006269592476, "grad_norm": 3.630739212036133, "learning_rate": 4.841528445405625e-06, "loss": 1.1942, "step": 26084 }, { "epoch": 2.044278996865204, "grad_norm": 3.601738691329956, "learning_rate": 4.8412202385668195e-06, "loss": 1.4638, "step": 26085 }, { "epoch": 2.04435736677116, "grad_norm": 14.780632972717285, "learning_rate": 4.8409120323319295e-06, "loss": 0.831, "step": 26086 }, { "epoch": 2.044435736677116, "grad_norm": 3.4966816902160645, "learning_rate": 4.840603826702131e-06, "loss": 1.8465, "step": 26087 }, { "epoch": 2.044514106583072, "grad_norm": 2.8744304180145264, "learning_rate": 4.840295621678591e-06, "loss": 1.4635, "step": 26088 }, { "epoch": 2.0445924764890284, "grad_norm": 4.936729431152344, "learning_rate": 4.839987417262485e-06, "loss": 1.4198, "step": 26089 }, { "epoch": 2.0446708463949843, "grad_norm": 3.54899001121521, "learning_rate": 4.839679213454986e-06, "loss": 1.633, "step": 26090 }, { "epoch": 2.0447492163009406, "grad_norm": 2.1551003456115723, "learning_rate": 4.839371010257265e-06, "loss": 1.1444, "step": 26091 }, { "epoch": 2.0448275862068965, "grad_norm": 5.170647144317627, "learning_rate": 4.839062807670493e-06, "loss": 1.3544, "step": 26092 }, { "epoch": 2.044905956112853, "grad_norm": 6.918540000915527, "learning_rate": 4.838754605695843e-06, "loss": 1.5034, "step": 26093 }, { "epoch": 2.0449843260188088, "grad_norm": 2.867208242416382, "learning_rate": 4.838446404334488e-06, "loss": 1.1412, "step": 26094 }, { "epoch": 2.045062695924765, "grad_norm": 3.009847402572632, "learning_rate": 4.838138203587602e-06, "loss": 1.6071, "step": 26095 }, { "epoch": 2.045141065830721, "grad_norm": 3.3910164833068848, "learning_rate": 4.837830003456353e-06, "loss": 1.2347, "step": 26096 }, { "epoch": 2.0452194357366773, "grad_norm": 3.1960461139678955, "learning_rate": 4.837521803941916e-06, "loss": 1.7616, "step": 26097 }, { "epoch": 2.045297805642633, "grad_norm": 3.9357962608337402, "learning_rate": 4.837213605045461e-06, "loss": 1.7743, "step": 26098 }, { "epoch": 2.0453761755485895, "grad_norm": 3.7807579040527344, "learning_rate": 4.8369054067681654e-06, "loss": 0.9019, "step": 26099 }, { "epoch": 2.0454545454545454, "grad_norm": 2.951105833053589, "learning_rate": 4.836597209111194e-06, "loss": 1.8419, "step": 26100 }, { "epoch": 2.0455329153605017, "grad_norm": 2.932037830352783, "learning_rate": 4.836289012075725e-06, "loss": 1.4069, "step": 26101 }, { "epoch": 2.0456112852664576, "grad_norm": 2.534985065460205, "learning_rate": 4.835980815662929e-06, "loss": 1.0061, "step": 26102 }, { "epoch": 2.045689655172414, "grad_norm": 2.530266523361206, "learning_rate": 4.835672619873977e-06, "loss": 0.7482, "step": 26103 }, { "epoch": 2.04576802507837, "grad_norm": 3.0778555870056152, "learning_rate": 4.8353644247100415e-06, "loss": 1.634, "step": 26104 }, { "epoch": 2.045846394984326, "grad_norm": 2.655400037765503, "learning_rate": 4.8350562301722945e-06, "loss": 1.2167, "step": 26105 }, { "epoch": 2.045924764890282, "grad_norm": 2.854539155960083, "learning_rate": 4.834748036261912e-06, "loss": 1.4373, "step": 26106 }, { "epoch": 2.0460031347962384, "grad_norm": 2.178696393966675, "learning_rate": 4.83443984298006e-06, "loss": 1.3582, "step": 26107 }, { "epoch": 2.0460815047021943, "grad_norm": 3.2040696144104004, "learning_rate": 4.834131650327915e-06, "loss": 1.6092, "step": 26108 }, { "epoch": 2.0461598746081506, "grad_norm": 5.005683898925781, "learning_rate": 4.8338234583066475e-06, "loss": 1.0093, "step": 26109 }, { "epoch": 2.0462382445141065, "grad_norm": 6.433727264404297, "learning_rate": 4.833515266917431e-06, "loss": 1.1748, "step": 26110 }, { "epoch": 2.046316614420063, "grad_norm": 5.459755897521973, "learning_rate": 4.8332070761614364e-06, "loss": 1.1108, "step": 26111 }, { "epoch": 2.0463949843260187, "grad_norm": 2.1991069316864014, "learning_rate": 4.8328988860398354e-06, "loss": 1.312, "step": 26112 }, { "epoch": 2.046473354231975, "grad_norm": 2.625906467437744, "learning_rate": 4.832590696553801e-06, "loss": 1.3345, "step": 26113 }, { "epoch": 2.046551724137931, "grad_norm": 3.401704788208008, "learning_rate": 4.832282507704509e-06, "loss": 1.5537, "step": 26114 }, { "epoch": 2.0466300940438873, "grad_norm": 4.174200534820557, "learning_rate": 4.831974319493126e-06, "loss": 1.3868, "step": 26115 }, { "epoch": 2.046708463949843, "grad_norm": 5.54486083984375, "learning_rate": 4.8316661319208255e-06, "loss": 1.249, "step": 26116 }, { "epoch": 2.0467868338557995, "grad_norm": 2.5138301849365234, "learning_rate": 4.831357944988783e-06, "loss": 1.4961, "step": 26117 }, { "epoch": 2.0468652037617554, "grad_norm": 3.4142308235168457, "learning_rate": 4.831049758698166e-06, "loss": 1.5934, "step": 26118 }, { "epoch": 2.0469435736677117, "grad_norm": 2.7239744663238525, "learning_rate": 4.830741573050149e-06, "loss": 1.1756, "step": 26119 }, { "epoch": 2.0470219435736676, "grad_norm": 2.8802239894866943, "learning_rate": 4.830433388045905e-06, "loss": 1.3569, "step": 26120 }, { "epoch": 2.047100313479624, "grad_norm": 2.4893722534179688, "learning_rate": 4.8301252036866055e-06, "loss": 1.4364, "step": 26121 }, { "epoch": 2.04717868338558, "grad_norm": 3.8207032680511475, "learning_rate": 4.829817019973422e-06, "loss": 1.2725, "step": 26122 }, { "epoch": 2.047257053291536, "grad_norm": 6.825625896453857, "learning_rate": 4.8295088369075276e-06, "loss": 1.2353, "step": 26123 }, { "epoch": 2.047335423197492, "grad_norm": 2.5988571643829346, "learning_rate": 4.829200654490092e-06, "loss": 1.5665, "step": 26124 }, { "epoch": 2.0474137931034484, "grad_norm": 3.5361244678497314, "learning_rate": 4.828892472722292e-06, "loss": 1.5386, "step": 26125 }, { "epoch": 2.0474921630094043, "grad_norm": 3.4146664142608643, "learning_rate": 4.828584291605296e-06, "loss": 1.2532, "step": 26126 }, { "epoch": 2.0475705329153606, "grad_norm": 2.772442579269409, "learning_rate": 4.828276111140277e-06, "loss": 1.2077, "step": 26127 }, { "epoch": 2.0476489028213165, "grad_norm": 5.149704933166504, "learning_rate": 4.827967931328409e-06, "loss": 1.7152, "step": 26128 }, { "epoch": 2.047727272727273, "grad_norm": 2.7833359241485596, "learning_rate": 4.827659752170863e-06, "loss": 1.3642, "step": 26129 }, { "epoch": 2.0478056426332287, "grad_norm": 2.9693069458007812, "learning_rate": 4.827351573668808e-06, "loss": 1.3684, "step": 26130 }, { "epoch": 2.047884012539185, "grad_norm": 7.270946502685547, "learning_rate": 4.827043395823419e-06, "loss": 1.9947, "step": 26131 }, { "epoch": 2.047962382445141, "grad_norm": 3.194671392440796, "learning_rate": 4.8267352186358705e-06, "loss": 1.2291, "step": 26132 }, { "epoch": 2.0480407523510973, "grad_norm": 6.724985599517822, "learning_rate": 4.8264270421073315e-06, "loss": 1.0452, "step": 26133 }, { "epoch": 2.048119122257053, "grad_norm": 4.072687149047852, "learning_rate": 4.826118866238974e-06, "loss": 1.2233, "step": 26134 }, { "epoch": 2.0481974921630095, "grad_norm": 7.129438877105713, "learning_rate": 4.825810691031972e-06, "loss": 1.619, "step": 26135 }, { "epoch": 2.0482758620689654, "grad_norm": 4.0171051025390625, "learning_rate": 4.825502516487497e-06, "loss": 1.6765, "step": 26136 }, { "epoch": 2.0483542319749217, "grad_norm": 2.4450740814208984, "learning_rate": 4.825194342606718e-06, "loss": 1.2813, "step": 26137 }, { "epoch": 2.0484326018808776, "grad_norm": 2.328470230102539, "learning_rate": 4.824886169390812e-06, "loss": 1.4547, "step": 26138 }, { "epoch": 2.048510971786834, "grad_norm": 3.3856046199798584, "learning_rate": 4.824577996840949e-06, "loss": 1.6559, "step": 26139 }, { "epoch": 2.04858934169279, "grad_norm": 2.633435010910034, "learning_rate": 4.824269824958303e-06, "loss": 1.1008, "step": 26140 }, { "epoch": 2.048667711598746, "grad_norm": 2.039754867553711, "learning_rate": 4.823961653744042e-06, "loss": 1.1154, "step": 26141 }, { "epoch": 2.048746081504702, "grad_norm": 2.738523483276367, "learning_rate": 4.823653483199339e-06, "loss": 1.2996, "step": 26142 }, { "epoch": 2.0488244514106584, "grad_norm": 4.365992546081543, "learning_rate": 4.823345313325372e-06, "loss": 1.0461, "step": 26143 }, { "epoch": 2.0489028213166143, "grad_norm": 2.771257162094116, "learning_rate": 4.823037144123305e-06, "loss": 1.2713, "step": 26144 }, { "epoch": 2.0489811912225706, "grad_norm": 9.533929824829102, "learning_rate": 4.822728975594315e-06, "loss": 1.2693, "step": 26145 }, { "epoch": 2.0490595611285265, "grad_norm": 3.7621350288391113, "learning_rate": 4.822420807739573e-06, "loss": 1.7773, "step": 26146 }, { "epoch": 2.049137931034483, "grad_norm": 2.732611894607544, "learning_rate": 4.822112640560252e-06, "loss": 0.8491, "step": 26147 }, { "epoch": 2.0492163009404387, "grad_norm": 2.206747055053711, "learning_rate": 4.821804474057522e-06, "loss": 1.0737, "step": 26148 }, { "epoch": 2.049294670846395, "grad_norm": 6.207011699676514, "learning_rate": 4.821496308232555e-06, "loss": 1.7286, "step": 26149 }, { "epoch": 2.049373040752351, "grad_norm": 2.309868097305298, "learning_rate": 4.821188143086526e-06, "loss": 1.3569, "step": 26150 }, { "epoch": 2.0494514106583073, "grad_norm": 2.9138588905334473, "learning_rate": 4.820879978620606e-06, "loss": 1.5334, "step": 26151 }, { "epoch": 2.049529780564263, "grad_norm": 2.2165966033935547, "learning_rate": 4.820571814835965e-06, "loss": 1.5434, "step": 26152 }, { "epoch": 2.0496081504702195, "grad_norm": 2.3718528747558594, "learning_rate": 4.820263651733778e-06, "loss": 1.2793, "step": 26153 }, { "epoch": 2.0496865203761754, "grad_norm": 3.022671937942505, "learning_rate": 4.819955489315213e-06, "loss": 1.1172, "step": 26154 }, { "epoch": 2.0497648902821317, "grad_norm": 2.7631592750549316, "learning_rate": 4.81964732758145e-06, "loss": 1.3268, "step": 26155 }, { "epoch": 2.0498432601880876, "grad_norm": 2.6758689880371094, "learning_rate": 4.819339166533652e-06, "loss": 1.2999, "step": 26156 }, { "epoch": 2.049921630094044, "grad_norm": 2.7855069637298584, "learning_rate": 4.8190310061729954e-06, "loss": 1.2227, "step": 26157 }, { "epoch": 2.05, "grad_norm": 3.627370595932007, "learning_rate": 4.8187228465006534e-06, "loss": 1.2167, "step": 26158 }, { "epoch": 2.050078369905956, "grad_norm": 3.380408763885498, "learning_rate": 4.818414687517795e-06, "loss": 1.551, "step": 26159 }, { "epoch": 2.050156739811912, "grad_norm": 2.655308961868286, "learning_rate": 4.818106529225595e-06, "loss": 1.1099, "step": 26160 }, { "epoch": 2.0502351097178684, "grad_norm": 3.949636220932007, "learning_rate": 4.817798371625222e-06, "loss": 1.5467, "step": 26161 }, { "epoch": 2.0503134796238243, "grad_norm": 2.5418288707733154, "learning_rate": 4.817490214717855e-06, "loss": 1.5224, "step": 26162 }, { "epoch": 2.0503918495297806, "grad_norm": 2.8800272941589355, "learning_rate": 4.817182058504657e-06, "loss": 1.4429, "step": 26163 }, { "epoch": 2.0504702194357365, "grad_norm": 3.8796792030334473, "learning_rate": 4.816873902986807e-06, "loss": 1.4908, "step": 26164 }, { "epoch": 2.050548589341693, "grad_norm": 2.5035715103149414, "learning_rate": 4.816565748165473e-06, "loss": 1.3546, "step": 26165 }, { "epoch": 2.0506269592476487, "grad_norm": 14.095311164855957, "learning_rate": 4.816257594041831e-06, "loss": 0.7628, "step": 26166 }, { "epoch": 2.050705329153605, "grad_norm": 29.449247360229492, "learning_rate": 4.81594944061705e-06, "loss": 1.1976, "step": 26167 }, { "epoch": 2.050783699059561, "grad_norm": 3.6188361644744873, "learning_rate": 4.8156412878923e-06, "loss": 1.1039, "step": 26168 }, { "epoch": 2.0508620689655173, "grad_norm": 3.4348504543304443, "learning_rate": 4.8153331358687575e-06, "loss": 0.8065, "step": 26169 }, { "epoch": 2.050940438871473, "grad_norm": 4.869271278381348, "learning_rate": 4.815024984547595e-06, "loss": 0.4816, "step": 26170 }, { "epoch": 2.0510188087774295, "grad_norm": 2.7582015991210938, "learning_rate": 4.8147168339299806e-06, "loss": 1.404, "step": 26171 }, { "epoch": 2.0510971786833854, "grad_norm": 2.4999213218688965, "learning_rate": 4.814408684017088e-06, "loss": 1.3303, "step": 26172 }, { "epoch": 2.0511755485893417, "grad_norm": 4.039997100830078, "learning_rate": 4.814100534810091e-06, "loss": 1.3282, "step": 26173 }, { "epoch": 2.0512539184952976, "grad_norm": 2.030830144882202, "learning_rate": 4.8137923863101586e-06, "loss": 1.3579, "step": 26174 }, { "epoch": 2.051332288401254, "grad_norm": 3.989286422729492, "learning_rate": 4.813484238518463e-06, "loss": 1.2055, "step": 26175 }, { "epoch": 2.0514106583072103, "grad_norm": 3.3297994136810303, "learning_rate": 4.813176091436179e-06, "loss": 1.4249, "step": 26176 }, { "epoch": 2.051489028213166, "grad_norm": 3.24527907371521, "learning_rate": 4.812867945064479e-06, "loss": 1.2078, "step": 26177 }, { "epoch": 2.051567398119122, "grad_norm": 3.0387041568756104, "learning_rate": 4.812559799404532e-06, "loss": 1.3406, "step": 26178 }, { "epoch": 2.0516457680250784, "grad_norm": 2.9709551334381104, "learning_rate": 4.81225165445751e-06, "loss": 1.1889, "step": 26179 }, { "epoch": 2.0517241379310347, "grad_norm": 2.887956380844116, "learning_rate": 4.811943510224586e-06, "loss": 1.3404, "step": 26180 }, { "epoch": 2.0518025078369906, "grad_norm": 2.6895103454589844, "learning_rate": 4.811635366706936e-06, "loss": 1.323, "step": 26181 }, { "epoch": 2.051880877742947, "grad_norm": 4.146422863006592, "learning_rate": 4.811327223905724e-06, "loss": 1.6522, "step": 26182 }, { "epoch": 2.051959247648903, "grad_norm": 3.4566102027893066, "learning_rate": 4.811019081822128e-06, "loss": 1.4339, "step": 26183 }, { "epoch": 2.052037617554859, "grad_norm": 2.461979389190674, "learning_rate": 4.81071094045732e-06, "loss": 1.0007, "step": 26184 }, { "epoch": 2.052115987460815, "grad_norm": 4.895248889923096, "learning_rate": 4.810402799812468e-06, "loss": 1.3117, "step": 26185 }, { "epoch": 2.0521943573667714, "grad_norm": 4.558495998382568, "learning_rate": 4.810094659888747e-06, "loss": 1.4026, "step": 26186 }, { "epoch": 2.0522727272727272, "grad_norm": 1.6192514896392822, "learning_rate": 4.809786520687327e-06, "loss": 1.1849, "step": 26187 }, { "epoch": 2.0523510971786836, "grad_norm": 2.275960683822632, "learning_rate": 4.809478382209386e-06, "loss": 1.1765, "step": 26188 }, { "epoch": 2.0524294670846395, "grad_norm": 2.6908483505249023, "learning_rate": 4.809170244456086e-06, "loss": 1.2781, "step": 26189 }, { "epoch": 2.052507836990596, "grad_norm": 3.8679776191711426, "learning_rate": 4.808862107428607e-06, "loss": 1.7225, "step": 26190 }, { "epoch": 2.0525862068965517, "grad_norm": 3.23056697845459, "learning_rate": 4.808553971128118e-06, "loss": 1.5869, "step": 26191 }, { "epoch": 2.052664576802508, "grad_norm": 3.9361677169799805, "learning_rate": 4.808245835555792e-06, "loss": 1.8402, "step": 26192 }, { "epoch": 2.052742946708464, "grad_norm": 3.8107528686523438, "learning_rate": 4.8079377007127995e-06, "loss": 1.1361, "step": 26193 }, { "epoch": 2.0528213166144202, "grad_norm": 6.332618236541748, "learning_rate": 4.807629566600313e-06, "loss": 0.9739, "step": 26194 }, { "epoch": 2.052899686520376, "grad_norm": 7.788076877593994, "learning_rate": 4.807321433219505e-06, "loss": 1.7012, "step": 26195 }, { "epoch": 2.0529780564263325, "grad_norm": 4.326021671295166, "learning_rate": 4.8070133005715485e-06, "loss": 1.7317, "step": 26196 }, { "epoch": 2.0530564263322884, "grad_norm": 3.3776347637176514, "learning_rate": 4.806705168657614e-06, "loss": 1.1094, "step": 26197 }, { "epoch": 2.0531347962382447, "grad_norm": 3.4884235858917236, "learning_rate": 4.806397037478873e-06, "loss": 1.4094, "step": 26198 }, { "epoch": 2.0532131661442006, "grad_norm": 2.5873472690582275, "learning_rate": 4.806088907036499e-06, "loss": 1.5712, "step": 26199 }, { "epoch": 2.053291536050157, "grad_norm": 2.7882139682769775, "learning_rate": 4.805780777331662e-06, "loss": 1.3099, "step": 26200 }, { "epoch": 2.053369905956113, "grad_norm": 5.390898704528809, "learning_rate": 4.805472648365535e-06, "loss": 1.3222, "step": 26201 }, { "epoch": 2.053448275862069, "grad_norm": 2.5817205905914307, "learning_rate": 4.805164520139291e-06, "loss": 1.2899, "step": 26202 }, { "epoch": 2.053526645768025, "grad_norm": 4.2361650466918945, "learning_rate": 4.804856392654102e-06, "loss": 1.5613, "step": 26203 }, { "epoch": 2.0536050156739813, "grad_norm": 2.964961051940918, "learning_rate": 4.804548265911138e-06, "loss": 1.4715, "step": 26204 }, { "epoch": 2.0536833855799372, "grad_norm": 2.5704076290130615, "learning_rate": 4.8042401399115715e-06, "loss": 1.48, "step": 26205 }, { "epoch": 2.0537617554858936, "grad_norm": 4.097198963165283, "learning_rate": 4.803932014656574e-06, "loss": 1.3553, "step": 26206 }, { "epoch": 2.0538401253918495, "grad_norm": 2.9255106449127197, "learning_rate": 4.803623890147323e-06, "loss": 1.2793, "step": 26207 }, { "epoch": 2.053918495297806, "grad_norm": 2.851977586746216, "learning_rate": 4.803315766384981e-06, "loss": 1.3658, "step": 26208 }, { "epoch": 2.0539968652037617, "grad_norm": 1.6820917129516602, "learning_rate": 4.803007643370727e-06, "loss": 1.551, "step": 26209 }, { "epoch": 2.054075235109718, "grad_norm": 3.697935104370117, "learning_rate": 4.80269952110573e-06, "loss": 1.3815, "step": 26210 }, { "epoch": 2.054153605015674, "grad_norm": 2.328249216079712, "learning_rate": 4.802391399591164e-06, "loss": 1.337, "step": 26211 }, { "epoch": 2.0542319749216302, "grad_norm": 3.8474161624908447, "learning_rate": 4.802083278828198e-06, "loss": 1.7097, "step": 26212 }, { "epoch": 2.054310344827586, "grad_norm": 4.6399712562561035, "learning_rate": 4.801775158818005e-06, "loss": 1.0441, "step": 26213 }, { "epoch": 2.0543887147335425, "grad_norm": 4.23220157623291, "learning_rate": 4.801467039561762e-06, "loss": 1.2141, "step": 26214 }, { "epoch": 2.0544670846394983, "grad_norm": 2.4074320793151855, "learning_rate": 4.801158921060632e-06, "loss": 1.3414, "step": 26215 }, { "epoch": 2.0545454545454547, "grad_norm": 2.2344534397125244, "learning_rate": 4.800850803315791e-06, "loss": 1.3052, "step": 26216 }, { "epoch": 2.0546238244514106, "grad_norm": 2.498335361480713, "learning_rate": 4.800542686328413e-06, "loss": 1.4503, "step": 26217 }, { "epoch": 2.054702194357367, "grad_norm": 2.920865535736084, "learning_rate": 4.800234570099669e-06, "loss": 1.0106, "step": 26218 }, { "epoch": 2.054780564263323, "grad_norm": 2.6341497898101807, "learning_rate": 4.799926454630729e-06, "loss": 1.2109, "step": 26219 }, { "epoch": 2.054858934169279, "grad_norm": 3.863981246948242, "learning_rate": 4.799618339922764e-06, "loss": 1.5472, "step": 26220 }, { "epoch": 2.054937304075235, "grad_norm": 2.63017201423645, "learning_rate": 4.79931022597695e-06, "loss": 1.7508, "step": 26221 }, { "epoch": 2.0550156739811913, "grad_norm": 3.306631088256836, "learning_rate": 4.799002112794458e-06, "loss": 1.1582, "step": 26222 }, { "epoch": 2.0550940438871472, "grad_norm": 4.540471076965332, "learning_rate": 4.798694000376457e-06, "loss": 1.5963, "step": 26223 }, { "epoch": 2.0551724137931036, "grad_norm": 2.9549319744110107, "learning_rate": 4.798385888724121e-06, "loss": 1.5728, "step": 26224 }, { "epoch": 2.0552507836990594, "grad_norm": 3.2433125972747803, "learning_rate": 4.79807777783862e-06, "loss": 1.3471, "step": 26225 }, { "epoch": 2.055329153605016, "grad_norm": 2.5913383960723877, "learning_rate": 4.79776966772113e-06, "loss": 1.1139, "step": 26226 }, { "epoch": 2.0554075235109717, "grad_norm": 4.080230236053467, "learning_rate": 4.79746155837282e-06, "loss": 1.4792, "step": 26227 }, { "epoch": 2.055485893416928, "grad_norm": 2.665822982788086, "learning_rate": 4.79715344979486e-06, "loss": 1.041, "step": 26228 }, { "epoch": 2.055564263322884, "grad_norm": 3.029179334640503, "learning_rate": 4.796845341988428e-06, "loss": 1.4192, "step": 26229 }, { "epoch": 2.05564263322884, "grad_norm": 2.730947494506836, "learning_rate": 4.796537234954689e-06, "loss": 1.542, "step": 26230 }, { "epoch": 2.055721003134796, "grad_norm": 10.988382339477539, "learning_rate": 4.796229128694818e-06, "loss": 0.7673, "step": 26231 }, { "epoch": 2.0557993730407524, "grad_norm": 5.246099472045898, "learning_rate": 4.795921023209986e-06, "loss": 1.5163, "step": 26232 }, { "epoch": 2.0558777429467083, "grad_norm": 2.898667097091675, "learning_rate": 4.795612918501368e-06, "loss": 1.2409, "step": 26233 }, { "epoch": 2.0559561128526647, "grad_norm": 3.378441333770752, "learning_rate": 4.795304814570132e-06, "loss": 1.2213, "step": 26234 }, { "epoch": 2.0560344827586206, "grad_norm": 3.4410696029663086, "learning_rate": 4.794996711417451e-06, "loss": 1.3881, "step": 26235 }, { "epoch": 2.056112852664577, "grad_norm": 2.3496949672698975, "learning_rate": 4.794688609044498e-06, "loss": 1.1234, "step": 26236 }, { "epoch": 2.0561912225705328, "grad_norm": 2.8407154083251953, "learning_rate": 4.794380507452444e-06, "loss": 1.3229, "step": 26237 }, { "epoch": 2.056269592476489, "grad_norm": 3.860130548477173, "learning_rate": 4.7940724066424606e-06, "loss": 0.864, "step": 26238 }, { "epoch": 2.056347962382445, "grad_norm": 6.558437347412109, "learning_rate": 4.79376430661572e-06, "loss": 0.9264, "step": 26239 }, { "epoch": 2.0564263322884013, "grad_norm": 2.330085515975952, "learning_rate": 4.793456207373395e-06, "loss": 1.274, "step": 26240 }, { "epoch": 2.056504702194357, "grad_norm": 2.9387588500976562, "learning_rate": 4.793148108916655e-06, "loss": 1.2397, "step": 26241 }, { "epoch": 2.0565830721003135, "grad_norm": 3.534618616104126, "learning_rate": 4.792840011246674e-06, "loss": 1.1614, "step": 26242 }, { "epoch": 2.0566614420062694, "grad_norm": 2.2592508792877197, "learning_rate": 4.792531914364623e-06, "loss": 1.1985, "step": 26243 }, { "epoch": 2.0567398119122258, "grad_norm": 3.7159531116485596, "learning_rate": 4.792223818271676e-06, "loss": 1.4868, "step": 26244 }, { "epoch": 2.0568181818181817, "grad_norm": 2.1781020164489746, "learning_rate": 4.791915722969e-06, "loss": 1.2547, "step": 26245 }, { "epoch": 2.056896551724138, "grad_norm": 3.5700578689575195, "learning_rate": 4.791607628457771e-06, "loss": 1.3957, "step": 26246 }, { "epoch": 2.056974921630094, "grad_norm": 3.5964367389678955, "learning_rate": 4.791299534739159e-06, "loss": 1.4117, "step": 26247 }, { "epoch": 2.05705329153605, "grad_norm": 3.8527638912200928, "learning_rate": 4.790991441814338e-06, "loss": 1.6422, "step": 26248 }, { "epoch": 2.057131661442006, "grad_norm": 2.2699975967407227, "learning_rate": 4.7906833496844765e-06, "loss": 1.7159, "step": 26249 }, { "epoch": 2.0572100313479624, "grad_norm": 2.462695598602295, "learning_rate": 4.790375258350748e-06, "loss": 1.4707, "step": 26250 }, { "epoch": 2.0572884012539183, "grad_norm": 3.0864791870117188, "learning_rate": 4.790067167814324e-06, "loss": 1.3887, "step": 26251 }, { "epoch": 2.0573667711598747, "grad_norm": 8.163745880126953, "learning_rate": 4.789759078076379e-06, "loss": 1.4666, "step": 26252 }, { "epoch": 2.0574451410658305, "grad_norm": 4.064269065856934, "learning_rate": 4.789450989138081e-06, "loss": 1.1608, "step": 26253 }, { "epoch": 2.057523510971787, "grad_norm": 2.8937790393829346, "learning_rate": 4.789142901000603e-06, "loss": 1.5068, "step": 26254 }, { "epoch": 2.0576018808777428, "grad_norm": 3.552107334136963, "learning_rate": 4.788834813665119e-06, "loss": 1.5914, "step": 26255 }, { "epoch": 2.057680250783699, "grad_norm": 6.537871837615967, "learning_rate": 4.788526727132797e-06, "loss": 1.7016, "step": 26256 }, { "epoch": 2.057758620689655, "grad_norm": 3.3553502559661865, "learning_rate": 4.788218641404809e-06, "loss": 1.4522, "step": 26257 }, { "epoch": 2.0578369905956113, "grad_norm": 2.522336959838867, "learning_rate": 4.787910556482331e-06, "loss": 1.3873, "step": 26258 }, { "epoch": 2.057915360501567, "grad_norm": 3.092078685760498, "learning_rate": 4.7876024723665325e-06, "loss": 1.003, "step": 26259 }, { "epoch": 2.0579937304075235, "grad_norm": 2.639967679977417, "learning_rate": 4.787294389058584e-06, "loss": 1.4589, "step": 26260 }, { "epoch": 2.0580721003134794, "grad_norm": 3.1292126178741455, "learning_rate": 4.786986306559659e-06, "loss": 1.3633, "step": 26261 }, { "epoch": 2.0581504702194358, "grad_norm": 3.787273645401001, "learning_rate": 4.786678224870928e-06, "loss": 1.5684, "step": 26262 }, { "epoch": 2.0582288401253916, "grad_norm": 2.884794235229492, "learning_rate": 4.786370143993566e-06, "loss": 1.1246, "step": 26263 }, { "epoch": 2.058307210031348, "grad_norm": 3.3979413509368896, "learning_rate": 4.7860620639287385e-06, "loss": 1.3264, "step": 26264 }, { "epoch": 2.058385579937304, "grad_norm": 4.433408260345459, "learning_rate": 4.785753984677623e-06, "loss": 1.7257, "step": 26265 }, { "epoch": 2.05846394984326, "grad_norm": 2.984771966934204, "learning_rate": 4.7854459062413885e-06, "loss": 1.3818, "step": 26266 }, { "epoch": 2.058542319749216, "grad_norm": 3.692291259765625, "learning_rate": 4.7851378286212095e-06, "loss": 1.2023, "step": 26267 }, { "epoch": 2.0586206896551724, "grad_norm": 11.950276374816895, "learning_rate": 4.784829751818254e-06, "loss": 1.5944, "step": 26268 }, { "epoch": 2.0586990595611283, "grad_norm": 3.295398473739624, "learning_rate": 4.784521675833695e-06, "loss": 1.4277, "step": 26269 }, { "epoch": 2.0587774294670846, "grad_norm": 4.176739692687988, "learning_rate": 4.784213600668708e-06, "loss": 0.8535, "step": 26270 }, { "epoch": 2.058855799373041, "grad_norm": 3.9225776195526123, "learning_rate": 4.7839055263244585e-06, "loss": 1.6926, "step": 26271 }, { "epoch": 2.058934169278997, "grad_norm": 2.665003538131714, "learning_rate": 4.783597452802122e-06, "loss": 1.12, "step": 26272 }, { "epoch": 2.059012539184953, "grad_norm": 2.6376028060913086, "learning_rate": 4.78328938010287e-06, "loss": 1.0843, "step": 26273 }, { "epoch": 2.059090909090909, "grad_norm": 5.190847396850586, "learning_rate": 4.782981308227876e-06, "loss": 1.5981, "step": 26274 }, { "epoch": 2.0591692789968654, "grad_norm": 3.3381500244140625, "learning_rate": 4.7826732371783066e-06, "loss": 0.9106, "step": 26275 }, { "epoch": 2.0592476489028213, "grad_norm": 4.36641788482666, "learning_rate": 4.7823651669553365e-06, "loss": 1.5271, "step": 26276 }, { "epoch": 2.0593260188087776, "grad_norm": 2.41085147857666, "learning_rate": 4.782057097560138e-06, "loss": 1.4353, "step": 26277 }, { "epoch": 2.0594043887147335, "grad_norm": 4.150155067443848, "learning_rate": 4.781749028993885e-06, "loss": 1.5346, "step": 26278 }, { "epoch": 2.05948275862069, "grad_norm": 2.800487518310547, "learning_rate": 4.781440961257744e-06, "loss": 1.4482, "step": 26279 }, { "epoch": 2.0595611285266457, "grad_norm": 2.669243574142456, "learning_rate": 4.781132894352889e-06, "loss": 1.5241, "step": 26280 }, { "epoch": 2.059639498432602, "grad_norm": 4.250916004180908, "learning_rate": 4.780824828280492e-06, "loss": 1.2109, "step": 26281 }, { "epoch": 2.059717868338558, "grad_norm": 5.143697738647461, "learning_rate": 4.780516763041728e-06, "loss": 1.4466, "step": 26282 }, { "epoch": 2.0597962382445143, "grad_norm": 4.080282688140869, "learning_rate": 4.780208698637762e-06, "loss": 0.8157, "step": 26283 }, { "epoch": 2.05987460815047, "grad_norm": 3.680971384048462, "learning_rate": 4.7799006350697706e-06, "loss": 1.7253, "step": 26284 }, { "epoch": 2.0599529780564265, "grad_norm": 3.3868894577026367, "learning_rate": 4.779592572338925e-06, "loss": 0.9839, "step": 26285 }, { "epoch": 2.0600313479623824, "grad_norm": 3.201604127883911, "learning_rate": 4.779284510446395e-06, "loss": 1.6225, "step": 26286 }, { "epoch": 2.0601097178683387, "grad_norm": 5.414649486541748, "learning_rate": 4.7789764493933535e-06, "loss": 1.1724, "step": 26287 }, { "epoch": 2.0601880877742946, "grad_norm": 3.0370378494262695, "learning_rate": 4.77866838918097e-06, "loss": 1.349, "step": 26288 }, { "epoch": 2.060266457680251, "grad_norm": 4.596681118011475, "learning_rate": 4.778360329810423e-06, "loss": 1.5618, "step": 26289 }, { "epoch": 2.060344827586207, "grad_norm": 2.0463407039642334, "learning_rate": 4.778052271282875e-06, "loss": 1.2993, "step": 26290 }, { "epoch": 2.060423197492163, "grad_norm": 7.091310977935791, "learning_rate": 4.777744213599504e-06, "loss": 1.5781, "step": 26291 }, { "epoch": 2.060501567398119, "grad_norm": 2.163299083709717, "learning_rate": 4.777436156761479e-06, "loss": 0.3204, "step": 26292 }, { "epoch": 2.0605799373040754, "grad_norm": 3.6229727268218994, "learning_rate": 4.777128100769974e-06, "loss": 1.22, "step": 26293 }, { "epoch": 2.0606583072100313, "grad_norm": 2.634031295776367, "learning_rate": 4.7768200456261584e-06, "loss": 1.1505, "step": 26294 }, { "epoch": 2.0607366771159876, "grad_norm": 5.195488929748535, "learning_rate": 4.776511991331204e-06, "loss": 1.1333, "step": 26295 }, { "epoch": 2.0608150470219435, "grad_norm": 3.366203546524048, "learning_rate": 4.7762039378862865e-06, "loss": 1.5996, "step": 26296 }, { "epoch": 2.0608934169279, "grad_norm": 4.822387218475342, "learning_rate": 4.77589588529257e-06, "loss": 1.5082, "step": 26297 }, { "epoch": 2.0609717868338557, "grad_norm": 3.2099313735961914, "learning_rate": 4.775587833551232e-06, "loss": 1.4072, "step": 26298 }, { "epoch": 2.061050156739812, "grad_norm": 4.397820949554443, "learning_rate": 4.775279782663443e-06, "loss": 1.4571, "step": 26299 }, { "epoch": 2.061128526645768, "grad_norm": 3.9017865657806396, "learning_rate": 4.774971732630375e-06, "loss": 1.3447, "step": 26300 }, { "epoch": 2.0612068965517243, "grad_norm": 3.2893362045288086, "learning_rate": 4.774663683453198e-06, "loss": 1.6705, "step": 26301 }, { "epoch": 2.06128526645768, "grad_norm": 9.0230073928833, "learning_rate": 4.774355635133083e-06, "loss": 0.4314, "step": 26302 }, { "epoch": 2.0613636363636365, "grad_norm": 2.890662670135498, "learning_rate": 4.774047587671205e-06, "loss": 1.4561, "step": 26303 }, { "epoch": 2.0614420062695924, "grad_norm": 2.7314372062683105, "learning_rate": 4.773739541068735e-06, "loss": 1.7446, "step": 26304 }, { "epoch": 2.0615203761755487, "grad_norm": 3.9629709720611572, "learning_rate": 4.773431495326842e-06, "loss": 1.1986, "step": 26305 }, { "epoch": 2.0615987460815046, "grad_norm": 3.827180862426758, "learning_rate": 4.773123450446699e-06, "loss": 1.0338, "step": 26306 }, { "epoch": 2.061677115987461, "grad_norm": 3.0521957874298096, "learning_rate": 4.7728154064294775e-06, "loss": 1.5016, "step": 26307 }, { "epoch": 2.061755485893417, "grad_norm": 3.448566198348999, "learning_rate": 4.772507363276353e-06, "loss": 1.1976, "step": 26308 }, { "epoch": 2.061833855799373, "grad_norm": 3.149676561355591, "learning_rate": 4.77219932098849e-06, "loss": 1.227, "step": 26309 }, { "epoch": 2.061912225705329, "grad_norm": 2.3154184818267822, "learning_rate": 4.771891279567065e-06, "loss": 1.1353, "step": 26310 }, { "epoch": 2.0619905956112854, "grad_norm": 2.567201614379883, "learning_rate": 4.7715832390132485e-06, "loss": 1.338, "step": 26311 }, { "epoch": 2.0620689655172413, "grad_norm": 4.2172627449035645, "learning_rate": 4.7712751993282124e-06, "loss": 1.3155, "step": 26312 }, { "epoch": 2.0621473354231976, "grad_norm": 3.016921281814575, "learning_rate": 4.770967160513127e-06, "loss": 1.0549, "step": 26313 }, { "epoch": 2.0622257053291535, "grad_norm": 2.7495031356811523, "learning_rate": 4.770659122569164e-06, "loss": 1.1553, "step": 26314 }, { "epoch": 2.06230407523511, "grad_norm": 3.4313406944274902, "learning_rate": 4.770351085497499e-06, "loss": 1.1537, "step": 26315 }, { "epoch": 2.0623824451410657, "grad_norm": 4.8649001121521, "learning_rate": 4.770043049299298e-06, "loss": 1.6442, "step": 26316 }, { "epoch": 2.062460815047022, "grad_norm": 2.3608882427215576, "learning_rate": 4.769735013975736e-06, "loss": 1.3313, "step": 26317 }, { "epoch": 2.062539184952978, "grad_norm": 5.256227493286133, "learning_rate": 4.769426979527983e-06, "loss": 1.3081, "step": 26318 }, { "epoch": 2.0626175548589343, "grad_norm": 3.367497682571411, "learning_rate": 4.769118945957213e-06, "loss": 0.938, "step": 26319 }, { "epoch": 2.06269592476489, "grad_norm": 7.859873294830322, "learning_rate": 4.7688109132645945e-06, "loss": 1.2429, "step": 26320 }, { "epoch": 2.0627742946708465, "grad_norm": 3.2119925022125244, "learning_rate": 4.7685028814513e-06, "loss": 1.3355, "step": 26321 }, { "epoch": 2.0628526645768024, "grad_norm": 2.86690616607666, "learning_rate": 4.768194850518502e-06, "loss": 1.6077, "step": 26322 }, { "epoch": 2.0629310344827587, "grad_norm": 2.851393699645996, "learning_rate": 4.767886820467373e-06, "loss": 1.5596, "step": 26323 }, { "epoch": 2.0630094043887146, "grad_norm": 6.461373329162598, "learning_rate": 4.767578791299083e-06, "loss": 1.7112, "step": 26324 }, { "epoch": 2.063087774294671, "grad_norm": 2.6679461002349854, "learning_rate": 4.767270763014804e-06, "loss": 1.4177, "step": 26325 }, { "epoch": 2.063166144200627, "grad_norm": 3.1250245571136475, "learning_rate": 4.766962735615707e-06, "loss": 1.3555, "step": 26326 }, { "epoch": 2.063244514106583, "grad_norm": 2.5569887161254883, "learning_rate": 4.766654709102964e-06, "loss": 1.4457, "step": 26327 }, { "epoch": 2.063322884012539, "grad_norm": 4.214664459228516, "learning_rate": 4.766346683477746e-06, "loss": 0.8501, "step": 26328 }, { "epoch": 2.0634012539184954, "grad_norm": 5.674430847167969, "learning_rate": 4.766038658741226e-06, "loss": 1.6529, "step": 26329 }, { "epoch": 2.0634796238244513, "grad_norm": 4.260318756103516, "learning_rate": 4.765730634894575e-06, "loss": 1.5295, "step": 26330 }, { "epoch": 2.0635579937304076, "grad_norm": 3.126838207244873, "learning_rate": 4.765422611938963e-06, "loss": 1.6374, "step": 26331 }, { "epoch": 2.0636363636363635, "grad_norm": 3.3309528827667236, "learning_rate": 4.765114589875565e-06, "loss": 2.0086, "step": 26332 }, { "epoch": 2.06371473354232, "grad_norm": 2.538351535797119, "learning_rate": 4.764806568705548e-06, "loss": 1.2811, "step": 26333 }, { "epoch": 2.0637931034482757, "grad_norm": 6.835656642913818, "learning_rate": 4.764498548430088e-06, "loss": 1.7153, "step": 26334 }, { "epoch": 2.063871473354232, "grad_norm": 2.5297818183898926, "learning_rate": 4.7641905290503535e-06, "loss": 1.007, "step": 26335 }, { "epoch": 2.063949843260188, "grad_norm": 2.802978515625, "learning_rate": 4.763882510567518e-06, "loss": 1.509, "step": 26336 }, { "epoch": 2.0640282131661443, "grad_norm": 2.3830578327178955, "learning_rate": 4.763574492982751e-06, "loss": 1.3154, "step": 26337 }, { "epoch": 2.0641065830721, "grad_norm": 4.298862457275391, "learning_rate": 4.763266476297227e-06, "loss": 1.4597, "step": 26338 }, { "epoch": 2.0641849529780565, "grad_norm": 2.681795835494995, "learning_rate": 4.762958460512113e-06, "loss": 1.1672, "step": 26339 }, { "epoch": 2.0642633228840124, "grad_norm": 3.0546298027038574, "learning_rate": 4.762650445628585e-06, "loss": 1.5006, "step": 26340 }, { "epoch": 2.0643416927899687, "grad_norm": 3.5010173320770264, "learning_rate": 4.762342431647814e-06, "loss": 1.2378, "step": 26341 }, { "epoch": 2.0644200626959246, "grad_norm": 2.843118190765381, "learning_rate": 4.762034418570968e-06, "loss": 1.3999, "step": 26342 }, { "epoch": 2.064498432601881, "grad_norm": 3.5347886085510254, "learning_rate": 4.761726406399222e-06, "loss": 0.7667, "step": 26343 }, { "epoch": 2.064576802507837, "grad_norm": 2.947831392288208, "learning_rate": 4.761418395133746e-06, "loss": 1.1897, "step": 26344 }, { "epoch": 2.064655172413793, "grad_norm": 2.7906758785247803, "learning_rate": 4.761110384775713e-06, "loss": 1.0258, "step": 26345 }, { "epoch": 2.064733542319749, "grad_norm": 6.699916839599609, "learning_rate": 4.760802375326291e-06, "loss": 2.2289, "step": 26346 }, { "epoch": 2.0648119122257054, "grad_norm": 2.0069077014923096, "learning_rate": 4.760494366786655e-06, "loss": 1.3981, "step": 26347 }, { "epoch": 2.0648902821316613, "grad_norm": 4.362747669219971, "learning_rate": 4.760186359157976e-06, "loss": 1.3653, "step": 26348 }, { "epoch": 2.0649686520376176, "grad_norm": 3.8844215869903564, "learning_rate": 4.7598783524414255e-06, "loss": 0.5862, "step": 26349 }, { "epoch": 2.0650470219435735, "grad_norm": 4.355609893798828, "learning_rate": 4.759570346638174e-06, "loss": 1.096, "step": 26350 }, { "epoch": 2.06512539184953, "grad_norm": 2.271587610244751, "learning_rate": 4.759262341749391e-06, "loss": 1.5118, "step": 26351 }, { "epoch": 2.0652037617554857, "grad_norm": 4.437263011932373, "learning_rate": 4.758954337776255e-06, "loss": 1.1895, "step": 26352 }, { "epoch": 2.065282131661442, "grad_norm": 3.0212292671203613, "learning_rate": 4.758646334719928e-06, "loss": 1.5215, "step": 26353 }, { "epoch": 2.065360501567398, "grad_norm": 2.923135995864868, "learning_rate": 4.758338332581589e-06, "loss": 1.3106, "step": 26354 }, { "epoch": 2.0654388714733543, "grad_norm": 6.0959038734436035, "learning_rate": 4.7580303313624065e-06, "loss": 1.111, "step": 26355 }, { "epoch": 2.06551724137931, "grad_norm": 2.5750176906585693, "learning_rate": 4.757722331063554e-06, "loss": 1.2472, "step": 26356 }, { "epoch": 2.0655956112852665, "grad_norm": 2.513085126876831, "learning_rate": 4.757414331686199e-06, "loss": 1.0103, "step": 26357 }, { "epoch": 2.0656739811912224, "grad_norm": 3.19670033454895, "learning_rate": 4.757106333231515e-06, "loss": 1.8752, "step": 26358 }, { "epoch": 2.0657523510971787, "grad_norm": 2.288520097732544, "learning_rate": 4.7567983357006745e-06, "loss": 1.2987, "step": 26359 }, { "epoch": 2.0658307210031346, "grad_norm": 5.250013828277588, "learning_rate": 4.756490339094849e-06, "loss": 1.3996, "step": 26360 }, { "epoch": 2.065909090909091, "grad_norm": 2.528099775314331, "learning_rate": 4.756182343415209e-06, "loss": 1.4451, "step": 26361 }, { "epoch": 2.0659874608150472, "grad_norm": 5.318690776824951, "learning_rate": 4.7558743486629254e-06, "loss": 1.2712, "step": 26362 }, { "epoch": 2.066065830721003, "grad_norm": 2.7273659706115723, "learning_rate": 4.7555663548391695e-06, "loss": 1.5703, "step": 26363 }, { "epoch": 2.066144200626959, "grad_norm": 2.446761131286621, "learning_rate": 4.755258361945117e-06, "loss": 0.9087, "step": 26364 }, { "epoch": 2.0662225705329154, "grad_norm": 5.1093058586120605, "learning_rate": 4.754950369981932e-06, "loss": 1.1876, "step": 26365 }, { "epoch": 2.0663009404388717, "grad_norm": 2.444775342941284, "learning_rate": 4.7546423789507915e-06, "loss": 1.4149, "step": 26366 }, { "epoch": 2.0663793103448276, "grad_norm": 2.81412672996521, "learning_rate": 4.754334388852867e-06, "loss": 1.2585, "step": 26367 }, { "epoch": 2.066457680250784, "grad_norm": 2.458250045776367, "learning_rate": 4.754026399689327e-06, "loss": 1.4207, "step": 26368 }, { "epoch": 2.06653605015674, "grad_norm": 9.825451850891113, "learning_rate": 4.753718411461344e-06, "loss": 0.9771, "step": 26369 }, { "epoch": 2.066614420062696, "grad_norm": 3.0031898021698, "learning_rate": 4.753410424170088e-06, "loss": 1.6691, "step": 26370 }, { "epoch": 2.066692789968652, "grad_norm": 4.733494758605957, "learning_rate": 4.753102437816736e-06, "loss": 0.9298, "step": 26371 }, { "epoch": 2.0667711598746084, "grad_norm": 2.3904778957366943, "learning_rate": 4.752794452402453e-06, "loss": 1.1662, "step": 26372 }, { "epoch": 2.0668495297805642, "grad_norm": 3.51582407951355, "learning_rate": 4.7524864679284124e-06, "loss": 1.6402, "step": 26373 }, { "epoch": 2.0669278996865206, "grad_norm": 2.9057323932647705, "learning_rate": 4.7521784843957875e-06, "loss": 1.4076, "step": 26374 }, { "epoch": 2.0670062695924765, "grad_norm": 2.1504788398742676, "learning_rate": 4.751870501805748e-06, "loss": 1.3728, "step": 26375 }, { "epoch": 2.067084639498433, "grad_norm": 3.171480894088745, "learning_rate": 4.751562520159466e-06, "loss": 1.479, "step": 26376 }, { "epoch": 2.0671630094043887, "grad_norm": 3.660811424255371, "learning_rate": 4.751254539458111e-06, "loss": 1.1689, "step": 26377 }, { "epoch": 2.067241379310345, "grad_norm": 3.758730411529541, "learning_rate": 4.750946559702857e-06, "loss": 0.9088, "step": 26378 }, { "epoch": 2.067319749216301, "grad_norm": 2.37713885307312, "learning_rate": 4.7506385808948756e-06, "loss": 1.3592, "step": 26379 }, { "epoch": 2.0673981191222572, "grad_norm": 4.251706123352051, "learning_rate": 4.750330603035336e-06, "loss": 1.2266, "step": 26380 }, { "epoch": 2.067476489028213, "grad_norm": 2.0535671710968018, "learning_rate": 4.75002262612541e-06, "loss": 1.2231, "step": 26381 }, { "epoch": 2.0675548589341695, "grad_norm": 2.4593887329101562, "learning_rate": 4.749714650166271e-06, "loss": 1.2276, "step": 26382 }, { "epoch": 2.0676332288401253, "grad_norm": 14.391020774841309, "learning_rate": 4.749406675159088e-06, "loss": 1.1034, "step": 26383 }, { "epoch": 2.0677115987460817, "grad_norm": 2.7343032360076904, "learning_rate": 4.7490987011050315e-06, "loss": 1.247, "step": 26384 }, { "epoch": 2.0677899686520376, "grad_norm": 11.399139404296875, "learning_rate": 4.748790728005277e-06, "loss": 1.5189, "step": 26385 }, { "epoch": 2.067868338557994, "grad_norm": 2.79927659034729, "learning_rate": 4.7484827558609935e-06, "loss": 1.3566, "step": 26386 }, { "epoch": 2.06794670846395, "grad_norm": 3.6236729621887207, "learning_rate": 4.748174784673352e-06, "loss": 1.482, "step": 26387 }, { "epoch": 2.068025078369906, "grad_norm": 2.099759101867676, "learning_rate": 4.747866814443523e-06, "loss": 1.3485, "step": 26388 }, { "epoch": 2.068103448275862, "grad_norm": 7.089600086212158, "learning_rate": 4.747558845172679e-06, "loss": 1.3807, "step": 26389 }, { "epoch": 2.0681818181818183, "grad_norm": 3.264815330505371, "learning_rate": 4.747250876861995e-06, "loss": 1.3413, "step": 26390 }, { "epoch": 2.0682601880877742, "grad_norm": 2.0364906787872314, "learning_rate": 4.746942909512635e-06, "loss": 1.3581, "step": 26391 }, { "epoch": 2.0683385579937306, "grad_norm": 7.158135414123535, "learning_rate": 4.746634943125776e-06, "loss": 1.1532, "step": 26392 }, { "epoch": 2.0684169278996865, "grad_norm": 3.7334723472595215, "learning_rate": 4.746326977702588e-06, "loss": 0.9957, "step": 26393 }, { "epoch": 2.068495297805643, "grad_norm": 3.2155230045318604, "learning_rate": 4.746019013244241e-06, "loss": 1.1331, "step": 26394 }, { "epoch": 2.0685736677115987, "grad_norm": 2.9613118171691895, "learning_rate": 4.745711049751908e-06, "loss": 1.5483, "step": 26395 }, { "epoch": 2.068652037617555, "grad_norm": 4.540999889373779, "learning_rate": 4.745403087226757e-06, "loss": 1.1754, "step": 26396 }, { "epoch": 2.068730407523511, "grad_norm": 4.571605682373047, "learning_rate": 4.745095125669966e-06, "loss": 1.4722, "step": 26397 }, { "epoch": 2.0688087774294672, "grad_norm": 3.4418210983276367, "learning_rate": 4.744787165082699e-06, "loss": 1.1719, "step": 26398 }, { "epoch": 2.068887147335423, "grad_norm": 3.275294065475464, "learning_rate": 4.7444792054661315e-06, "loss": 1.6136, "step": 26399 }, { "epoch": 2.0689655172413794, "grad_norm": 3.3696539402008057, "learning_rate": 4.744171246821433e-06, "loss": 1.5376, "step": 26400 }, { "epoch": 2.0690438871473353, "grad_norm": 2.344475030899048, "learning_rate": 4.743863289149778e-06, "loss": 1.5181, "step": 26401 }, { "epoch": 2.0691222570532917, "grad_norm": 2.4720139503479004, "learning_rate": 4.743555332452334e-06, "loss": 1.3607, "step": 26402 }, { "epoch": 2.0692006269592476, "grad_norm": 3.318528652191162, "learning_rate": 4.7432473767302725e-06, "loss": 1.4764, "step": 26403 }, { "epoch": 2.069278996865204, "grad_norm": 4.871815204620361, "learning_rate": 4.742939421984768e-06, "loss": 1.3238, "step": 26404 }, { "epoch": 2.06935736677116, "grad_norm": 2.526146650314331, "learning_rate": 4.74263146821699e-06, "loss": 1.2377, "step": 26405 }, { "epoch": 2.069435736677116, "grad_norm": 3.2365806102752686, "learning_rate": 4.742323515428109e-06, "loss": 1.3132, "step": 26406 }, { "epoch": 2.069514106583072, "grad_norm": 3.6910653114318848, "learning_rate": 4.742015563619297e-06, "loss": 0.919, "step": 26407 }, { "epoch": 2.0695924764890283, "grad_norm": 2.4293200969696045, "learning_rate": 4.741707612791726e-06, "loss": 1.4331, "step": 26408 }, { "epoch": 2.069670846394984, "grad_norm": 2.4295198917388916, "learning_rate": 4.741399662946566e-06, "loss": 1.4207, "step": 26409 }, { "epoch": 2.0697492163009406, "grad_norm": 3.0635786056518555, "learning_rate": 4.7410917140849875e-06, "loss": 1.059, "step": 26410 }, { "epoch": 2.0698275862068964, "grad_norm": 2.4062914848327637, "learning_rate": 4.7407837662081645e-06, "loss": 1.0897, "step": 26411 }, { "epoch": 2.0699059561128528, "grad_norm": 5.007913112640381, "learning_rate": 4.7404758193172674e-06, "loss": 1.5745, "step": 26412 }, { "epoch": 2.0699843260188087, "grad_norm": 3.327242851257324, "learning_rate": 4.740167873413466e-06, "loss": 1.3962, "step": 26413 }, { "epoch": 2.070062695924765, "grad_norm": 4.563041687011719, "learning_rate": 4.739859928497934e-06, "loss": 1.5658, "step": 26414 }, { "epoch": 2.070141065830721, "grad_norm": 1.842632532119751, "learning_rate": 4.7395519845718385e-06, "loss": 1.1712, "step": 26415 }, { "epoch": 2.070219435736677, "grad_norm": 2.8570899963378906, "learning_rate": 4.7392440416363575e-06, "loss": 1.2583, "step": 26416 }, { "epoch": 2.070297805642633, "grad_norm": 6.955935478210449, "learning_rate": 4.738936099692654e-06, "loss": 0.6599, "step": 26417 }, { "epoch": 2.0703761755485894, "grad_norm": 2.7476441860198975, "learning_rate": 4.738628158741906e-06, "loss": 1.3904, "step": 26418 }, { "epoch": 2.0704545454545453, "grad_norm": 4.544615268707275, "learning_rate": 4.738320218785281e-06, "loss": 1.4891, "step": 26419 }, { "epoch": 2.0705329153605017, "grad_norm": 1.8758928775787354, "learning_rate": 4.738012279823953e-06, "loss": 1.3628, "step": 26420 }, { "epoch": 2.0706112852664575, "grad_norm": 4.685173988342285, "learning_rate": 4.73770434185909e-06, "loss": 1.4507, "step": 26421 }, { "epoch": 2.070689655172414, "grad_norm": 4.745297431945801, "learning_rate": 4.737396404891865e-06, "loss": 1.0023, "step": 26422 }, { "epoch": 2.0707680250783698, "grad_norm": 2.8425333499908447, "learning_rate": 4.7370884689234514e-06, "loss": 1.2874, "step": 26423 }, { "epoch": 2.070846394984326, "grad_norm": 2.754465103149414, "learning_rate": 4.736780533955017e-06, "loss": 1.3275, "step": 26424 }, { "epoch": 2.070924764890282, "grad_norm": 4.8485541343688965, "learning_rate": 4.736472599987733e-06, "loss": 1.6012, "step": 26425 }, { "epoch": 2.0710031347962383, "grad_norm": 3.3458619117736816, "learning_rate": 4.736164667022774e-06, "loss": 1.2579, "step": 26426 }, { "epoch": 2.071081504702194, "grad_norm": 4.057445049285889, "learning_rate": 4.735856735061308e-06, "loss": 1.4905, "step": 26427 }, { "epoch": 2.0711598746081505, "grad_norm": 4.451837062835693, "learning_rate": 4.735548804104508e-06, "loss": 1.2734, "step": 26428 }, { "epoch": 2.0712382445141064, "grad_norm": 4.118666172027588, "learning_rate": 4.735240874153542e-06, "loss": 0.8836, "step": 26429 }, { "epoch": 2.0713166144200628, "grad_norm": 3.3055601119995117, "learning_rate": 4.734932945209586e-06, "loss": 0.8009, "step": 26430 }, { "epoch": 2.0713949843260187, "grad_norm": 4.310542583465576, "learning_rate": 4.734625017273809e-06, "loss": 1.5118, "step": 26431 }, { "epoch": 2.071473354231975, "grad_norm": 5.502954483032227, "learning_rate": 4.7343170903473814e-06, "loss": 1.5346, "step": 26432 }, { "epoch": 2.071551724137931, "grad_norm": 10.194315910339355, "learning_rate": 4.734009164431476e-06, "loss": 1.8769, "step": 26433 }, { "epoch": 2.071630094043887, "grad_norm": 3.5599496364593506, "learning_rate": 4.733701239527261e-06, "loss": 1.1085, "step": 26434 }, { "epoch": 2.071708463949843, "grad_norm": 4.469828128814697, "learning_rate": 4.7333933156359115e-06, "loss": 1.3857, "step": 26435 }, { "epoch": 2.0717868338557994, "grad_norm": 5.458199977874756, "learning_rate": 4.733085392758597e-06, "loss": 1.4822, "step": 26436 }, { "epoch": 2.0718652037617553, "grad_norm": 2.8744561672210693, "learning_rate": 4.732777470896488e-06, "loss": 1.1646, "step": 26437 }, { "epoch": 2.0719435736677116, "grad_norm": 4.269110679626465, "learning_rate": 4.7324695500507575e-06, "loss": 1.1337, "step": 26438 }, { "epoch": 2.0720219435736675, "grad_norm": 3.271811008453369, "learning_rate": 4.732161630222574e-06, "loss": 1.3007, "step": 26439 }, { "epoch": 2.072100313479624, "grad_norm": 5.216278553009033, "learning_rate": 4.731853711413109e-06, "loss": 1.0113, "step": 26440 }, { "epoch": 2.0721786833855798, "grad_norm": 4.52495002746582, "learning_rate": 4.731545793623536e-06, "loss": 1.4172, "step": 26441 }, { "epoch": 2.072257053291536, "grad_norm": 2.2988083362579346, "learning_rate": 4.731237876855026e-06, "loss": 1.4014, "step": 26442 }, { "epoch": 2.072335423197492, "grad_norm": 7.842491626739502, "learning_rate": 4.7309299611087485e-06, "loss": 1.2289, "step": 26443 }, { "epoch": 2.0724137931034483, "grad_norm": 2.0236153602600098, "learning_rate": 4.730622046385874e-06, "loss": 1.2991, "step": 26444 }, { "epoch": 2.072492163009404, "grad_norm": 3.0437445640563965, "learning_rate": 4.730314132687576e-06, "loss": 1.5191, "step": 26445 }, { "epoch": 2.0725705329153605, "grad_norm": 2.31720232963562, "learning_rate": 4.730006220015025e-06, "loss": 1.0295, "step": 26446 }, { "epoch": 2.0726489028213164, "grad_norm": 5.362094879150391, "learning_rate": 4.72969830836939e-06, "loss": 1.0376, "step": 26447 }, { "epoch": 2.0727272727272728, "grad_norm": 3.393799304962158, "learning_rate": 4.729390397751845e-06, "loss": 1.4562, "step": 26448 }, { "epoch": 2.0728056426332286, "grad_norm": 3.1585984230041504, "learning_rate": 4.72908248816356e-06, "loss": 1.3689, "step": 26449 }, { "epoch": 2.072884012539185, "grad_norm": 6.215701580047607, "learning_rate": 4.728774579605706e-06, "loss": 1.1565, "step": 26450 }, { "epoch": 2.072962382445141, "grad_norm": 4.865140914916992, "learning_rate": 4.728466672079453e-06, "loss": 1.0986, "step": 26451 }, { "epoch": 2.073040752351097, "grad_norm": 3.9565343856811523, "learning_rate": 4.728158765585973e-06, "loss": 1.2415, "step": 26452 }, { "epoch": 2.073119122257053, "grad_norm": 2.5418338775634766, "learning_rate": 4.727850860126441e-06, "loss": 1.4076, "step": 26453 }, { "epoch": 2.0731974921630094, "grad_norm": 2.7619540691375732, "learning_rate": 4.727542955702021e-06, "loss": 1.0721, "step": 26454 }, { "epoch": 2.0732758620689653, "grad_norm": 2.454948902130127, "learning_rate": 4.727235052313889e-06, "loss": 1.3263, "step": 26455 }, { "epoch": 2.0733542319749216, "grad_norm": 6.355726718902588, "learning_rate": 4.726927149963214e-06, "loss": 1.4164, "step": 26456 }, { "epoch": 2.073432601880878, "grad_norm": 2.9041271209716797, "learning_rate": 4.726619248651169e-06, "loss": 1.3548, "step": 26457 }, { "epoch": 2.073510971786834, "grad_norm": 13.083271980285645, "learning_rate": 4.7263113483789245e-06, "loss": 1.0529, "step": 26458 }, { "epoch": 2.0735893416927897, "grad_norm": 2.9330341815948486, "learning_rate": 4.726003449147648e-06, "loss": 1.2322, "step": 26459 }, { "epoch": 2.073667711598746, "grad_norm": 3.127166748046875, "learning_rate": 4.725695550958516e-06, "loss": 1.3446, "step": 26460 }, { "epoch": 2.0737460815047024, "grad_norm": 5.115874290466309, "learning_rate": 4.725387653812697e-06, "loss": 0.948, "step": 26461 }, { "epoch": 2.0738244514106583, "grad_norm": 6.481277942657471, "learning_rate": 4.725079757711363e-06, "loss": 1.2615, "step": 26462 }, { "epoch": 2.0739028213166146, "grad_norm": 4.319669246673584, "learning_rate": 4.724771862655683e-06, "loss": 1.2131, "step": 26463 }, { "epoch": 2.0739811912225705, "grad_norm": 2.834604024887085, "learning_rate": 4.724463968646832e-06, "loss": 1.3063, "step": 26464 }, { "epoch": 2.074059561128527, "grad_norm": 4.710607528686523, "learning_rate": 4.724156075685975e-06, "loss": 1.6327, "step": 26465 }, { "epoch": 2.0741379310344827, "grad_norm": 2.387876033782959, "learning_rate": 4.723848183774288e-06, "loss": 1.1241, "step": 26466 }, { "epoch": 2.074216300940439, "grad_norm": 4.6392130851745605, "learning_rate": 4.72354029291294e-06, "loss": 0.8045, "step": 26467 }, { "epoch": 2.074294670846395, "grad_norm": 2.514620065689087, "learning_rate": 4.723232403103105e-06, "loss": 1.2903, "step": 26468 }, { "epoch": 2.0743730407523513, "grad_norm": 3.007484197616577, "learning_rate": 4.72292451434595e-06, "loss": 1.4297, "step": 26469 }, { "epoch": 2.074451410658307, "grad_norm": 1.9990668296813965, "learning_rate": 4.722616626642648e-06, "loss": 1.3507, "step": 26470 }, { "epoch": 2.0745297805642635, "grad_norm": 2.5359954833984375, "learning_rate": 4.722308739994369e-06, "loss": 1.4723, "step": 26471 }, { "epoch": 2.0746081504702194, "grad_norm": 5.7367939949035645, "learning_rate": 4.722000854402288e-06, "loss": 1.0975, "step": 26472 }, { "epoch": 2.0746865203761757, "grad_norm": 2.568979501724243, "learning_rate": 4.7216929698675695e-06, "loss": 1.4552, "step": 26473 }, { "epoch": 2.0747648902821316, "grad_norm": 2.024395704269409, "learning_rate": 4.72138508639139e-06, "loss": 1.2164, "step": 26474 }, { "epoch": 2.074843260188088, "grad_norm": 3.069225549697876, "learning_rate": 4.721077203974917e-06, "loss": 1.4827, "step": 26475 }, { "epoch": 2.074921630094044, "grad_norm": 2.0915439128875732, "learning_rate": 4.720769322619325e-06, "loss": 1.3377, "step": 26476 }, { "epoch": 2.075, "grad_norm": 4.908476829528809, "learning_rate": 4.720461442325782e-06, "loss": 0.7025, "step": 26477 }, { "epoch": 2.075078369905956, "grad_norm": 5.333481311798096, "learning_rate": 4.720153563095459e-06, "loss": 0.8016, "step": 26478 }, { "epoch": 2.0751567398119124, "grad_norm": 2.3593103885650635, "learning_rate": 4.719845684929532e-06, "loss": 1.0483, "step": 26479 }, { "epoch": 2.0752351097178683, "grad_norm": 2.9243507385253906, "learning_rate": 4.719537807829164e-06, "loss": 1.5006, "step": 26480 }, { "epoch": 2.0753134796238246, "grad_norm": 2.7951385974884033, "learning_rate": 4.7192299317955316e-06, "loss": 1.5482, "step": 26481 }, { "epoch": 2.0753918495297805, "grad_norm": 3.8167169094085693, "learning_rate": 4.718922056829804e-06, "loss": 1.1662, "step": 26482 }, { "epoch": 2.075470219435737, "grad_norm": 2.6163697242736816, "learning_rate": 4.718614182933153e-06, "loss": 1.2675, "step": 26483 }, { "epoch": 2.0755485893416927, "grad_norm": 5.0727152824401855, "learning_rate": 4.7183063101067495e-06, "loss": 1.9506, "step": 26484 }, { "epoch": 2.075626959247649, "grad_norm": 2.8042209148406982, "learning_rate": 4.717998438351762e-06, "loss": 1.122, "step": 26485 }, { "epoch": 2.075705329153605, "grad_norm": 3.4018967151641846, "learning_rate": 4.717690567669365e-06, "loss": 0.9958, "step": 26486 }, { "epoch": 2.0757836990595613, "grad_norm": 5.823202610015869, "learning_rate": 4.717382698060729e-06, "loss": 1.1794, "step": 26487 }, { "epoch": 2.075862068965517, "grad_norm": 2.7632243633270264, "learning_rate": 4.717074829527023e-06, "loss": 1.3993, "step": 26488 }, { "epoch": 2.0759404388714735, "grad_norm": 2.5241799354553223, "learning_rate": 4.716766962069419e-06, "loss": 1.447, "step": 26489 }, { "epoch": 2.0760188087774294, "grad_norm": 4.168001651763916, "learning_rate": 4.716459095689088e-06, "loss": 1.2685, "step": 26490 }, { "epoch": 2.0760971786833857, "grad_norm": 9.611148834228516, "learning_rate": 4.7161512303872035e-06, "loss": 0.342, "step": 26491 }, { "epoch": 2.0761755485893416, "grad_norm": 2.0277392864227295, "learning_rate": 4.71584336616493e-06, "loss": 1.3635, "step": 26492 }, { "epoch": 2.076253918495298, "grad_norm": 3.704841375350952, "learning_rate": 4.715535503023444e-06, "loss": 1.501, "step": 26493 }, { "epoch": 2.076332288401254, "grad_norm": 3.601649284362793, "learning_rate": 4.7152276409639155e-06, "loss": 1.2973, "step": 26494 }, { "epoch": 2.07641065830721, "grad_norm": 3.2126524448394775, "learning_rate": 4.7149197799875144e-06, "loss": 1.67, "step": 26495 }, { "epoch": 2.076489028213166, "grad_norm": 2.955368995666504, "learning_rate": 4.714611920095411e-06, "loss": 1.3041, "step": 26496 }, { "epoch": 2.0765673981191224, "grad_norm": 4.919708251953125, "learning_rate": 4.714304061288777e-06, "loss": 1.3274, "step": 26497 }, { "epoch": 2.0766457680250783, "grad_norm": 3.210658311843872, "learning_rate": 4.713996203568787e-06, "loss": 1.3782, "step": 26498 }, { "epoch": 2.0767241379310346, "grad_norm": 3.7608542442321777, "learning_rate": 4.713688346936605e-06, "loss": 1.6638, "step": 26499 }, { "epoch": 2.0768025078369905, "grad_norm": 2.660048723220825, "learning_rate": 4.713380491393407e-06, "loss": 0.9963, "step": 26500 }, { "epoch": 2.076880877742947, "grad_norm": 2.6695759296417236, "learning_rate": 4.713072636940362e-06, "loss": 1.4204, "step": 26501 }, { "epoch": 2.0769592476489027, "grad_norm": 3.1727004051208496, "learning_rate": 4.712764783578642e-06, "loss": 1.2583, "step": 26502 }, { "epoch": 2.077037617554859, "grad_norm": 5.799688816070557, "learning_rate": 4.712456931309417e-06, "loss": 2.382, "step": 26503 }, { "epoch": 2.077115987460815, "grad_norm": 4.322524070739746, "learning_rate": 4.712149080133856e-06, "loss": 1.5041, "step": 26504 }, { "epoch": 2.0771943573667713, "grad_norm": 3.1200766563415527, "learning_rate": 4.7118412300531354e-06, "loss": 1.5319, "step": 26505 }, { "epoch": 2.077272727272727, "grad_norm": 8.779958724975586, "learning_rate": 4.71153338106842e-06, "loss": 1.6795, "step": 26506 }, { "epoch": 2.0773510971786835, "grad_norm": 2.620913028717041, "learning_rate": 4.711225533180885e-06, "loss": 1.4388, "step": 26507 }, { "epoch": 2.0774294670846394, "grad_norm": 2.4185376167297363, "learning_rate": 4.710917686391699e-06, "loss": 1.4502, "step": 26508 }, { "epoch": 2.0775078369905957, "grad_norm": 3.706756591796875, "learning_rate": 4.710609840702035e-06, "loss": 1.157, "step": 26509 }, { "epoch": 2.0775862068965516, "grad_norm": 4.203411102294922, "learning_rate": 4.710301996113062e-06, "loss": 1.6918, "step": 26510 }, { "epoch": 2.077664576802508, "grad_norm": 3.9943902492523193, "learning_rate": 4.709994152625949e-06, "loss": 1.5481, "step": 26511 }, { "epoch": 2.077742946708464, "grad_norm": 12.028943061828613, "learning_rate": 4.709686310241871e-06, "loss": 1.5077, "step": 26512 }, { "epoch": 2.07782131661442, "grad_norm": 3.8473989963531494, "learning_rate": 4.709378468961999e-06, "loss": 1.3693, "step": 26513 }, { "epoch": 2.077899686520376, "grad_norm": 2.634767532348633, "learning_rate": 4.7090706287875e-06, "loss": 1.351, "step": 26514 }, { "epoch": 2.0779780564263324, "grad_norm": 2.3933420181274414, "learning_rate": 4.708762789719547e-06, "loss": 1.5177, "step": 26515 }, { "epoch": 2.0780564263322883, "grad_norm": 16.98501205444336, "learning_rate": 4.7084549517593105e-06, "loss": 1.465, "step": 26516 }, { "epoch": 2.0781347962382446, "grad_norm": 2.7480759620666504, "learning_rate": 4.708147114907964e-06, "loss": 1.0685, "step": 26517 }, { "epoch": 2.0782131661442005, "grad_norm": 3.9889979362487793, "learning_rate": 4.707839279166673e-06, "loss": 1.3518, "step": 26518 }, { "epoch": 2.078291536050157, "grad_norm": 3.1135966777801514, "learning_rate": 4.707531444536612e-06, "loss": 1.0969, "step": 26519 }, { "epoch": 2.0783699059561127, "grad_norm": 2.438098669052124, "learning_rate": 4.707223611018953e-06, "loss": 1.2583, "step": 26520 }, { "epoch": 2.078448275862069, "grad_norm": 2.800342082977295, "learning_rate": 4.706915778614863e-06, "loss": 1.8284, "step": 26521 }, { "epoch": 2.078526645768025, "grad_norm": 2.496931791305542, "learning_rate": 4.706607947325516e-06, "loss": 1.5933, "step": 26522 }, { "epoch": 2.0786050156739813, "grad_norm": 4.365067958831787, "learning_rate": 4.706300117152079e-06, "loss": 1.8018, "step": 26523 }, { "epoch": 2.078683385579937, "grad_norm": 2.6463778018951416, "learning_rate": 4.70599228809573e-06, "loss": 1.0768, "step": 26524 }, { "epoch": 2.0787617554858935, "grad_norm": 5.0622944831848145, "learning_rate": 4.7056844601576325e-06, "loss": 1.3164, "step": 26525 }, { "epoch": 2.0788401253918494, "grad_norm": 3.9518344402313232, "learning_rate": 4.70537663333896e-06, "loss": 1.424, "step": 26526 }, { "epoch": 2.0789184952978057, "grad_norm": 3.84202241897583, "learning_rate": 4.705068807640884e-06, "loss": 1.6117, "step": 26527 }, { "epoch": 2.0789968652037616, "grad_norm": 2.9744069576263428, "learning_rate": 4.704760983064576e-06, "loss": 1.6752, "step": 26528 }, { "epoch": 2.079075235109718, "grad_norm": 3.1029086112976074, "learning_rate": 4.704453159611204e-06, "loss": 1.3397, "step": 26529 }, { "epoch": 2.079153605015674, "grad_norm": 4.195206165313721, "learning_rate": 4.704145337281939e-06, "loss": 1.1075, "step": 26530 }, { "epoch": 2.07923197492163, "grad_norm": 3.1232314109802246, "learning_rate": 4.703837516077956e-06, "loss": 1.9771, "step": 26531 }, { "epoch": 2.079310344827586, "grad_norm": 2.3579070568084717, "learning_rate": 4.703529696000423e-06, "loss": 0.9355, "step": 26532 }, { "epoch": 2.0793887147335424, "grad_norm": 2.744497537612915, "learning_rate": 4.703221877050509e-06, "loss": 1.3074, "step": 26533 }, { "epoch": 2.0794670846394983, "grad_norm": 4.325725555419922, "learning_rate": 4.702914059229387e-06, "loss": 1.4327, "step": 26534 }, { "epoch": 2.0795454545454546, "grad_norm": 2.0286312103271484, "learning_rate": 4.702606242538229e-06, "loss": 1.1839, "step": 26535 }, { "epoch": 2.0796238244514105, "grad_norm": 4.468648433685303, "learning_rate": 4.702298426978201e-06, "loss": 1.6338, "step": 26536 }, { "epoch": 2.079702194357367, "grad_norm": 3.3153133392333984, "learning_rate": 4.7019906125504785e-06, "loss": 1.937, "step": 26537 }, { "epoch": 2.0797805642633227, "grad_norm": 3.2375590801239014, "learning_rate": 4.701682799256231e-06, "loss": 1.4296, "step": 26538 }, { "epoch": 2.079858934169279, "grad_norm": 4.425300598144531, "learning_rate": 4.70137498709663e-06, "loss": 1.5256, "step": 26539 }, { "epoch": 2.079937304075235, "grad_norm": 2.2986271381378174, "learning_rate": 4.701067176072843e-06, "loss": 1.5077, "step": 26540 }, { "epoch": 2.0800156739811912, "grad_norm": 5.363092422485352, "learning_rate": 4.700759366186044e-06, "loss": 1.5121, "step": 26541 }, { "epoch": 2.080094043887147, "grad_norm": 4.312945365905762, "learning_rate": 4.700451557437401e-06, "loss": 1.391, "step": 26542 }, { "epoch": 2.0801724137931035, "grad_norm": 3.307126760482788, "learning_rate": 4.7001437498280885e-06, "loss": 1.2401, "step": 26543 }, { "epoch": 2.0802507836990594, "grad_norm": 2.2968223094940186, "learning_rate": 4.699835943359274e-06, "loss": 1.4707, "step": 26544 }, { "epoch": 2.0803291536050157, "grad_norm": 2.369504451751709, "learning_rate": 4.69952813803213e-06, "loss": 1.2969, "step": 26545 }, { "epoch": 2.0804075235109716, "grad_norm": 2.8567700386047363, "learning_rate": 4.699220333847826e-06, "loss": 1.5121, "step": 26546 }, { "epoch": 2.080485893416928, "grad_norm": 4.961216926574707, "learning_rate": 4.698912530807535e-06, "loss": 1.2436, "step": 26547 }, { "epoch": 2.080564263322884, "grad_norm": 2.492846965789795, "learning_rate": 4.698604728912424e-06, "loss": 1.1781, "step": 26548 }, { "epoch": 2.08064263322884, "grad_norm": 2.608536958694458, "learning_rate": 4.698296928163667e-06, "loss": 1.5366, "step": 26549 }, { "epoch": 2.080721003134796, "grad_norm": 3.3732991218566895, "learning_rate": 4.697989128562434e-06, "loss": 1.5713, "step": 26550 }, { "epoch": 2.0807993730407524, "grad_norm": 6.117767333984375, "learning_rate": 4.697681330109895e-06, "loss": 0.866, "step": 26551 }, { "epoch": 2.0808777429467087, "grad_norm": 6.661434650421143, "learning_rate": 4.697373532807219e-06, "loss": 1.4363, "step": 26552 }, { "epoch": 2.0809561128526646, "grad_norm": 3.5183465480804443, "learning_rate": 4.6970657366555795e-06, "loss": 1.3531, "step": 26553 }, { "epoch": 2.0810344827586205, "grad_norm": 2.3452882766723633, "learning_rate": 4.6967579416561484e-06, "loss": 1.3464, "step": 26554 }, { "epoch": 2.081112852664577, "grad_norm": 2.78706431388855, "learning_rate": 4.6964501478100925e-06, "loss": 1.1217, "step": 26555 }, { "epoch": 2.081191222570533, "grad_norm": 2.9574618339538574, "learning_rate": 4.696142355118584e-06, "loss": 1.1652, "step": 26556 }, { "epoch": 2.081269592476489, "grad_norm": 7.503132343292236, "learning_rate": 4.695834563582794e-06, "loss": 1.5601, "step": 26557 }, { "epoch": 2.0813479623824453, "grad_norm": 4.2842698097229, "learning_rate": 4.695526773203895e-06, "loss": 0.7224, "step": 26558 }, { "epoch": 2.0814263322884012, "grad_norm": 6.270111083984375, "learning_rate": 4.695218983983055e-06, "loss": 1.5254, "step": 26559 }, { "epoch": 2.0815047021943576, "grad_norm": 2.027536630630493, "learning_rate": 4.694911195921443e-06, "loss": 1.4575, "step": 26560 }, { "epoch": 2.0815830721003135, "grad_norm": 4.456570625305176, "learning_rate": 4.694603409020236e-06, "loss": 1.5814, "step": 26561 }, { "epoch": 2.08166144200627, "grad_norm": 1.8829505443572998, "learning_rate": 4.694295623280598e-06, "loss": 1.1057, "step": 26562 }, { "epoch": 2.0817398119122257, "grad_norm": 3.334892511367798, "learning_rate": 4.693987838703703e-06, "loss": 1.68, "step": 26563 }, { "epoch": 2.081818181818182, "grad_norm": 4.35968017578125, "learning_rate": 4.693680055290721e-06, "loss": 1.4024, "step": 26564 }, { "epoch": 2.081896551724138, "grad_norm": 7.995857238769531, "learning_rate": 4.693372273042824e-06, "loss": 2.1157, "step": 26565 }, { "epoch": 2.0819749216300942, "grad_norm": 2.585089683532715, "learning_rate": 4.69306449196118e-06, "loss": 1.2994, "step": 26566 }, { "epoch": 2.08205329153605, "grad_norm": 2.6714158058166504, "learning_rate": 4.69275671204696e-06, "loss": 1.2446, "step": 26567 }, { "epoch": 2.0821316614420065, "grad_norm": 5.653437614440918, "learning_rate": 4.692448933301338e-06, "loss": 1.3301, "step": 26568 }, { "epoch": 2.0822100313479623, "grad_norm": 3.0734198093414307, "learning_rate": 4.692141155725482e-06, "loss": 1.31, "step": 26569 }, { "epoch": 2.0822884012539187, "grad_norm": 3.795736312866211, "learning_rate": 4.691833379320562e-06, "loss": 1.2186, "step": 26570 }, { "epoch": 2.0823667711598746, "grad_norm": 6.428853988647461, "learning_rate": 4.691525604087749e-06, "loss": 1.309, "step": 26571 }, { "epoch": 2.082445141065831, "grad_norm": 2.2789628505706787, "learning_rate": 4.691217830028214e-06, "loss": 1.3348, "step": 26572 }, { "epoch": 2.082523510971787, "grad_norm": 2.6061153411865234, "learning_rate": 4.690910057143131e-06, "loss": 1.3032, "step": 26573 }, { "epoch": 2.082601880877743, "grad_norm": 3.684990406036377, "learning_rate": 4.690602285433664e-06, "loss": 1.7294, "step": 26574 }, { "epoch": 2.082680250783699, "grad_norm": 3.174417495727539, "learning_rate": 4.690294514900988e-06, "loss": 1.4894, "step": 26575 }, { "epoch": 2.0827586206896553, "grad_norm": 7.859870910644531, "learning_rate": 4.689986745546273e-06, "loss": 1.2757, "step": 26576 }, { "epoch": 2.0828369905956112, "grad_norm": 4.116950035095215, "learning_rate": 4.689678977370689e-06, "loss": 1.4115, "step": 26577 }, { "epoch": 2.0829153605015676, "grad_norm": 3.5824406147003174, "learning_rate": 4.689371210375407e-06, "loss": 1.3467, "step": 26578 }, { "epoch": 2.0829937304075234, "grad_norm": 8.260217666625977, "learning_rate": 4.689063444561596e-06, "loss": 1.1893, "step": 26579 }, { "epoch": 2.08307210031348, "grad_norm": 2.3489153385162354, "learning_rate": 4.688755679930431e-06, "loss": 1.1531, "step": 26580 }, { "epoch": 2.0831504702194357, "grad_norm": 3.018960952758789, "learning_rate": 4.688447916483076e-06, "loss": 1.4623, "step": 26581 }, { "epoch": 2.083228840125392, "grad_norm": 3.693089723587036, "learning_rate": 4.688140154220707e-06, "loss": 1.1327, "step": 26582 }, { "epoch": 2.083307210031348, "grad_norm": 6.064723491668701, "learning_rate": 4.687832393144492e-06, "loss": 1.4434, "step": 26583 }, { "epoch": 2.083385579937304, "grad_norm": 6.115511417388916, "learning_rate": 4.6875246332556035e-06, "loss": 2.1862, "step": 26584 }, { "epoch": 2.08346394984326, "grad_norm": 6.372826099395752, "learning_rate": 4.6872168745552096e-06, "loss": 1.527, "step": 26585 }, { "epoch": 2.0835423197492164, "grad_norm": 2.8251137733459473, "learning_rate": 4.686909117044481e-06, "loss": 1.4353, "step": 26586 }, { "epoch": 2.0836206896551723, "grad_norm": 3.1587142944335938, "learning_rate": 4.68660136072459e-06, "loss": 1.1688, "step": 26587 }, { "epoch": 2.0836990595611287, "grad_norm": 3.920586109161377, "learning_rate": 4.686293605596708e-06, "loss": 1.1354, "step": 26588 }, { "epoch": 2.0837774294670846, "grad_norm": 2.094686269760132, "learning_rate": 4.685985851662003e-06, "loss": 1.1542, "step": 26589 }, { "epoch": 2.083855799373041, "grad_norm": 4.140957355499268, "learning_rate": 4.685678098921646e-06, "loss": 1.3997, "step": 26590 }, { "epoch": 2.0839341692789968, "grad_norm": 5.003094673156738, "learning_rate": 4.68537034737681e-06, "loss": 0.9773, "step": 26591 }, { "epoch": 2.084012539184953, "grad_norm": 2.3597309589385986, "learning_rate": 4.685062597028662e-06, "loss": 1.3437, "step": 26592 }, { "epoch": 2.084090909090909, "grad_norm": 2.2549285888671875, "learning_rate": 4.684754847878373e-06, "loss": 1.1912, "step": 26593 }, { "epoch": 2.0841692789968653, "grad_norm": 3.6465959548950195, "learning_rate": 4.684447099927116e-06, "loss": 1.1725, "step": 26594 }, { "epoch": 2.084247648902821, "grad_norm": 3.2003087997436523, "learning_rate": 4.684139353176061e-06, "loss": 1.4917, "step": 26595 }, { "epoch": 2.0843260188087775, "grad_norm": 4.808638095855713, "learning_rate": 4.683831607626377e-06, "loss": 1.3673, "step": 26596 }, { "epoch": 2.0844043887147334, "grad_norm": 4.058971405029297, "learning_rate": 4.683523863279235e-06, "loss": 1.1594, "step": 26597 }, { "epoch": 2.0844827586206898, "grad_norm": 4.093345642089844, "learning_rate": 4.683216120135804e-06, "loss": 1.4018, "step": 26598 }, { "epoch": 2.0845611285266457, "grad_norm": 11.166772842407227, "learning_rate": 4.68290837819726e-06, "loss": 1.1592, "step": 26599 }, { "epoch": 2.084639498432602, "grad_norm": 2.144019842147827, "learning_rate": 4.6826006374647665e-06, "loss": 0.6331, "step": 26600 }, { "epoch": 2.084717868338558, "grad_norm": 3.7644500732421875, "learning_rate": 4.682292897939498e-06, "loss": 1.4001, "step": 26601 }, { "epoch": 2.084796238244514, "grad_norm": 2.771109104156494, "learning_rate": 4.681985159622624e-06, "loss": 1.6179, "step": 26602 }, { "epoch": 2.08487460815047, "grad_norm": 3.9212536811828613, "learning_rate": 4.681677422515316e-06, "loss": 1.5922, "step": 26603 }, { "epoch": 2.0849529780564264, "grad_norm": 4.416985988616943, "learning_rate": 4.681369686618743e-06, "loss": 1.3444, "step": 26604 }, { "epoch": 2.0850313479623823, "grad_norm": 2.161104679107666, "learning_rate": 4.681061951934075e-06, "loss": 1.486, "step": 26605 }, { "epoch": 2.0851097178683387, "grad_norm": 5.357553958892822, "learning_rate": 4.680754218462487e-06, "loss": 1.897, "step": 26606 }, { "epoch": 2.0851880877742945, "grad_norm": 2.937307357788086, "learning_rate": 4.680446486205142e-06, "loss": 1.7504, "step": 26607 }, { "epoch": 2.085266457680251, "grad_norm": 2.9167683124542236, "learning_rate": 4.680138755163215e-06, "loss": 1.6488, "step": 26608 }, { "epoch": 2.0853448275862068, "grad_norm": 3.4658334255218506, "learning_rate": 4.679831025337878e-06, "loss": 1.3275, "step": 26609 }, { "epoch": 2.085423197492163, "grad_norm": 3.417158365249634, "learning_rate": 4.679523296730299e-06, "loss": 1.3368, "step": 26610 }, { "epoch": 2.085501567398119, "grad_norm": 2.5711283683776855, "learning_rate": 4.679215569341647e-06, "loss": 1.3502, "step": 26611 }, { "epoch": 2.0855799373040753, "grad_norm": 2.7937660217285156, "learning_rate": 4.678907843173094e-06, "loss": 1.0721, "step": 26612 }, { "epoch": 2.085658307210031, "grad_norm": 8.10145092010498, "learning_rate": 4.6786001182258116e-06, "loss": 1.7629, "step": 26613 }, { "epoch": 2.0857366771159875, "grad_norm": 2.3489835262298584, "learning_rate": 4.678292394500971e-06, "loss": 1.4278, "step": 26614 }, { "epoch": 2.0858150470219434, "grad_norm": 2.614239454269409, "learning_rate": 4.677984671999739e-06, "loss": 1.6657, "step": 26615 }, { "epoch": 2.0858934169278998, "grad_norm": 3.4363479614257812, "learning_rate": 4.677676950723288e-06, "loss": 1.0374, "step": 26616 }, { "epoch": 2.0859717868338556, "grad_norm": 2.0885705947875977, "learning_rate": 4.677369230672789e-06, "loss": 1.3471, "step": 26617 }, { "epoch": 2.086050156739812, "grad_norm": 3.9016940593719482, "learning_rate": 4.677061511849412e-06, "loss": 1.6923, "step": 26618 }, { "epoch": 2.086128526645768, "grad_norm": 6.712887287139893, "learning_rate": 4.676753794254325e-06, "loss": 1.2496, "step": 26619 }, { "epoch": 2.086206896551724, "grad_norm": 2.5430824756622314, "learning_rate": 4.676446077888702e-06, "loss": 1.2532, "step": 26620 }, { "epoch": 2.08628526645768, "grad_norm": 2.167006015777588, "learning_rate": 4.676138362753713e-06, "loss": 1.4569, "step": 26621 }, { "epoch": 2.0863636363636364, "grad_norm": 3.7686264514923096, "learning_rate": 4.675830648850526e-06, "loss": 1.5709, "step": 26622 }, { "epoch": 2.0864420062695923, "grad_norm": 4.032506465911865, "learning_rate": 4.675522936180312e-06, "loss": 1.8098, "step": 26623 }, { "epoch": 2.0865203761755486, "grad_norm": 2.9723799228668213, "learning_rate": 4.675215224744242e-06, "loss": 1.3405, "step": 26624 }, { "epoch": 2.0865987460815045, "grad_norm": 2.655623435974121, "learning_rate": 4.67490751454349e-06, "loss": 1.1605, "step": 26625 }, { "epoch": 2.086677115987461, "grad_norm": 2.0025720596313477, "learning_rate": 4.674599805579218e-06, "loss": 1.3078, "step": 26626 }, { "epoch": 2.0867554858934168, "grad_norm": 3.3509981632232666, "learning_rate": 4.674292097852604e-06, "loss": 0.9017, "step": 26627 }, { "epoch": 2.086833855799373, "grad_norm": 2.7293756008148193, "learning_rate": 4.673984391364816e-06, "loss": 1.5138, "step": 26628 }, { "epoch": 2.086912225705329, "grad_norm": 3.798827648162842, "learning_rate": 4.673676686117023e-06, "loss": 1.4755, "step": 26629 }, { "epoch": 2.0869905956112853, "grad_norm": 3.4648847579956055, "learning_rate": 4.673368982110396e-06, "loss": 1.2979, "step": 26630 }, { "epoch": 2.087068965517241, "grad_norm": 3.2447197437286377, "learning_rate": 4.673061279346105e-06, "loss": 1.5595, "step": 26631 }, { "epoch": 2.0871473354231975, "grad_norm": 2.382148027420044, "learning_rate": 4.672753577825323e-06, "loss": 1.3147, "step": 26632 }, { "epoch": 2.0872257053291534, "grad_norm": 3.5683655738830566, "learning_rate": 4.672445877549217e-06, "loss": 1.0866, "step": 26633 }, { "epoch": 2.0873040752351097, "grad_norm": 3.658799171447754, "learning_rate": 4.672138178518958e-06, "loss": 1.3188, "step": 26634 }, { "epoch": 2.0873824451410656, "grad_norm": 6.300516605377197, "learning_rate": 4.671830480735719e-06, "loss": 1.6865, "step": 26635 }, { "epoch": 2.087460815047022, "grad_norm": 3.65777587890625, "learning_rate": 4.671522784200667e-06, "loss": 1.4268, "step": 26636 }, { "epoch": 2.087539184952978, "grad_norm": 2.631251811981201, "learning_rate": 4.6712150889149726e-06, "loss": 1.0974, "step": 26637 }, { "epoch": 2.087617554858934, "grad_norm": 2.696730136871338, "learning_rate": 4.670907394879808e-06, "loss": 1.2613, "step": 26638 }, { "epoch": 2.08769592476489, "grad_norm": 3.981137990951538, "learning_rate": 4.670599702096343e-06, "loss": 1.3654, "step": 26639 }, { "epoch": 2.0877742946708464, "grad_norm": 4.041280746459961, "learning_rate": 4.670292010565748e-06, "loss": 1.5855, "step": 26640 }, { "epoch": 2.0878526645768023, "grad_norm": 2.2079896926879883, "learning_rate": 4.669984320289192e-06, "loss": 1.4091, "step": 26641 }, { "epoch": 2.0879310344827586, "grad_norm": 2.581400156021118, "learning_rate": 4.669676631267846e-06, "loss": 1.0416, "step": 26642 }, { "epoch": 2.088009404388715, "grad_norm": 4.9841084480285645, "learning_rate": 4.66936894350288e-06, "loss": 1.1562, "step": 26643 }, { "epoch": 2.088087774294671, "grad_norm": 3.1238107681274414, "learning_rate": 4.669061256995467e-06, "loss": 1.5354, "step": 26644 }, { "epoch": 2.0881661442006267, "grad_norm": 2.700821876525879, "learning_rate": 4.668753571746773e-06, "loss": 1.293, "step": 26645 }, { "epoch": 2.088244514106583, "grad_norm": 3.1350533962249756, "learning_rate": 4.668445887757971e-06, "loss": 1.1887, "step": 26646 }, { "epoch": 2.0883228840125394, "grad_norm": 2.4846272468566895, "learning_rate": 4.668138205030231e-06, "loss": 1.5344, "step": 26647 }, { "epoch": 2.0884012539184953, "grad_norm": 2.330512762069702, "learning_rate": 4.667830523564722e-06, "loss": 1.464, "step": 26648 }, { "epoch": 2.0884796238244516, "grad_norm": 2.6165454387664795, "learning_rate": 4.667522843362614e-06, "loss": 1.2775, "step": 26649 }, { "epoch": 2.0885579937304075, "grad_norm": 11.89302921295166, "learning_rate": 4.66721516442508e-06, "loss": 1.4969, "step": 26650 }, { "epoch": 2.088636363636364, "grad_norm": 2.6443474292755127, "learning_rate": 4.666907486753289e-06, "loss": 1.2119, "step": 26651 }, { "epoch": 2.0887147335423197, "grad_norm": 2.939614772796631, "learning_rate": 4.6665998103484095e-06, "loss": 1.6262, "step": 26652 }, { "epoch": 2.088793103448276, "grad_norm": 3.1315813064575195, "learning_rate": 4.666292135211614e-06, "loss": 1.4231, "step": 26653 }, { "epoch": 2.088871473354232, "grad_norm": 2.324753761291504, "learning_rate": 4.66598446134407e-06, "loss": 1.2261, "step": 26654 }, { "epoch": 2.0889498432601883, "grad_norm": 3.4581868648529053, "learning_rate": 4.665676788746953e-06, "loss": 1.1384, "step": 26655 }, { "epoch": 2.089028213166144, "grad_norm": 4.489522457122803, "learning_rate": 4.665369117421426e-06, "loss": 1.1299, "step": 26656 }, { "epoch": 2.0891065830721005, "grad_norm": 4.377861022949219, "learning_rate": 4.665061447368664e-06, "loss": 1.4465, "step": 26657 }, { "epoch": 2.0891849529780564, "grad_norm": 2.99521541595459, "learning_rate": 4.664753778589839e-06, "loss": 1.4481, "step": 26658 }, { "epoch": 2.0892633228840127, "grad_norm": 5.2290472984313965, "learning_rate": 4.664446111086116e-06, "loss": 1.695, "step": 26659 }, { "epoch": 2.0893416927899686, "grad_norm": 2.7627549171447754, "learning_rate": 4.664138444858668e-06, "loss": 1.3723, "step": 26660 }, { "epoch": 2.089420062695925, "grad_norm": 4.133547782897949, "learning_rate": 4.663830779908663e-06, "loss": 1.9245, "step": 26661 }, { "epoch": 2.089498432601881, "grad_norm": 4.882558822631836, "learning_rate": 4.6635231162372775e-06, "loss": 1.1087, "step": 26662 }, { "epoch": 2.089576802507837, "grad_norm": 2.7153961658477783, "learning_rate": 4.663215453845673e-06, "loss": 1.3631, "step": 26663 }, { "epoch": 2.089655172413793, "grad_norm": 5.036291122436523, "learning_rate": 4.662907792735026e-06, "loss": 1.4764, "step": 26664 }, { "epoch": 2.0897335423197494, "grad_norm": 2.9403269290924072, "learning_rate": 4.662600132906503e-06, "loss": 1.3579, "step": 26665 }, { "epoch": 2.0898119122257053, "grad_norm": 2.6334707736968994, "learning_rate": 4.662292474361279e-06, "loss": 1.1618, "step": 26666 }, { "epoch": 2.0898902821316616, "grad_norm": 7.544633388519287, "learning_rate": 4.661984817100518e-06, "loss": 1.3724, "step": 26667 }, { "epoch": 2.0899686520376175, "grad_norm": 5.046412944793701, "learning_rate": 4.661677161125393e-06, "loss": 1.2469, "step": 26668 }, { "epoch": 2.090047021943574, "grad_norm": 4.986956596374512, "learning_rate": 4.661369506437076e-06, "loss": 1.7489, "step": 26669 }, { "epoch": 2.0901253918495297, "grad_norm": 14.175981521606445, "learning_rate": 4.6610618530367355e-06, "loss": 0.7473, "step": 26670 }, { "epoch": 2.090203761755486, "grad_norm": 3.903850555419922, "learning_rate": 4.66075420092554e-06, "loss": 1.3595, "step": 26671 }, { "epoch": 2.090282131661442, "grad_norm": 3.130023956298828, "learning_rate": 4.660446550104663e-06, "loss": 1.4451, "step": 26672 }, { "epoch": 2.0903605015673983, "grad_norm": 2.403102159500122, "learning_rate": 4.660138900575273e-06, "loss": 1.3669, "step": 26673 }, { "epoch": 2.090438871473354, "grad_norm": 5.895804405212402, "learning_rate": 4.659831252338539e-06, "loss": 0.741, "step": 26674 }, { "epoch": 2.0905172413793105, "grad_norm": 8.667303085327148, "learning_rate": 4.659523605395631e-06, "loss": 1.6424, "step": 26675 }, { "epoch": 2.0905956112852664, "grad_norm": 2.405611276626587, "learning_rate": 4.659215959747722e-06, "loss": 1.1974, "step": 26676 }, { "epoch": 2.0906739811912227, "grad_norm": 4.0407586097717285, "learning_rate": 4.6589083153959805e-06, "loss": 1.4189, "step": 26677 }, { "epoch": 2.0907523510971786, "grad_norm": 2.4046542644500732, "learning_rate": 4.658600672341577e-06, "loss": 1.307, "step": 26678 }, { "epoch": 2.090830721003135, "grad_norm": 3.390475034713745, "learning_rate": 4.65829303058568e-06, "loss": 1.5878, "step": 26679 }, { "epoch": 2.090909090909091, "grad_norm": 3.231501817703247, "learning_rate": 4.65798539012946e-06, "loss": 1.1047, "step": 26680 }, { "epoch": 2.090987460815047, "grad_norm": 4.787170886993408, "learning_rate": 4.6576777509740916e-06, "loss": 1.0145, "step": 26681 }, { "epoch": 2.091065830721003, "grad_norm": 3.4755215644836426, "learning_rate": 4.657370113120737e-06, "loss": 1.334, "step": 26682 }, { "epoch": 2.0911442006269594, "grad_norm": 2.7798397541046143, "learning_rate": 4.657062476570572e-06, "loss": 1.1723, "step": 26683 }, { "epoch": 2.0912225705329153, "grad_norm": 6.669501304626465, "learning_rate": 4.656754841324765e-06, "loss": 1.6339, "step": 26684 }, { "epoch": 2.0913009404388716, "grad_norm": 3.9176886081695557, "learning_rate": 4.656447207384488e-06, "loss": 1.492, "step": 26685 }, { "epoch": 2.0913793103448275, "grad_norm": 2.551868200302124, "learning_rate": 4.656139574750908e-06, "loss": 1.1732, "step": 26686 }, { "epoch": 2.091457680250784, "grad_norm": 2.334972858428955, "learning_rate": 4.655831943425196e-06, "loss": 1.6561, "step": 26687 }, { "epoch": 2.0915360501567397, "grad_norm": 4.187391757965088, "learning_rate": 4.655524313408525e-06, "loss": 1.5606, "step": 26688 }, { "epoch": 2.091614420062696, "grad_norm": 7.961763858795166, "learning_rate": 4.655216684702059e-06, "loss": 1.5278, "step": 26689 }, { "epoch": 2.091692789968652, "grad_norm": 2.711392402648926, "learning_rate": 4.6549090573069735e-06, "loss": 1.1408, "step": 26690 }, { "epoch": 2.0917711598746083, "grad_norm": 2.242285966873169, "learning_rate": 4.654601431224437e-06, "loss": 1.2665, "step": 26691 }, { "epoch": 2.091849529780564, "grad_norm": 2.624830484390259, "learning_rate": 4.6542938064556184e-06, "loss": 1.0755, "step": 26692 }, { "epoch": 2.0919278996865205, "grad_norm": 3.4282243251800537, "learning_rate": 4.653986183001689e-06, "loss": 0.8084, "step": 26693 }, { "epoch": 2.0920062695924764, "grad_norm": 2.5474448204040527, "learning_rate": 4.653678560863817e-06, "loss": 1.192, "step": 26694 }, { "epoch": 2.0920846394984327, "grad_norm": 2.4101881980895996, "learning_rate": 4.653370940043175e-06, "loss": 1.2503, "step": 26695 }, { "epoch": 2.0921630094043886, "grad_norm": 2.8764734268188477, "learning_rate": 4.653063320540933e-06, "loss": 1.3893, "step": 26696 }, { "epoch": 2.092241379310345, "grad_norm": 2.9944961071014404, "learning_rate": 4.6527557023582596e-06, "loss": 1.4161, "step": 26697 }, { "epoch": 2.092319749216301, "grad_norm": 4.003603458404541, "learning_rate": 4.652448085496324e-06, "loss": 1.6278, "step": 26698 }, { "epoch": 2.092398119122257, "grad_norm": 2.583291530609131, "learning_rate": 4.652140469956297e-06, "loss": 1.5717, "step": 26699 }, { "epoch": 2.092476489028213, "grad_norm": 3.428215742111206, "learning_rate": 4.6518328557393526e-06, "loss": 0.9899, "step": 26700 }, { "epoch": 2.0925548589341694, "grad_norm": 2.726114511489868, "learning_rate": 4.651525242846653e-06, "loss": 1.4332, "step": 26701 }, { "epoch": 2.0926332288401253, "grad_norm": 2.7460203170776367, "learning_rate": 4.651217631279374e-06, "loss": 1.0517, "step": 26702 }, { "epoch": 2.0927115987460816, "grad_norm": 3.472468614578247, "learning_rate": 4.6509100210386845e-06, "loss": 1.3259, "step": 26703 }, { "epoch": 2.0927899686520375, "grad_norm": 3.6424367427825928, "learning_rate": 4.650602412125753e-06, "loss": 1.4795, "step": 26704 }, { "epoch": 2.092868338557994, "grad_norm": 5.033779621124268, "learning_rate": 4.650294804541751e-06, "loss": 1.2392, "step": 26705 }, { "epoch": 2.0929467084639497, "grad_norm": 5.36460542678833, "learning_rate": 4.649987198287846e-06, "loss": 2.2399, "step": 26706 }, { "epoch": 2.093025078369906, "grad_norm": 3.2240712642669678, "learning_rate": 4.649679593365214e-06, "loss": 0.8847, "step": 26707 }, { "epoch": 2.093103448275862, "grad_norm": 4.893913745880127, "learning_rate": 4.649371989775018e-06, "loss": 1.3528, "step": 26708 }, { "epoch": 2.0931818181818183, "grad_norm": 4.804601192474365, "learning_rate": 4.649064387518431e-06, "loss": 1.2236, "step": 26709 }, { "epoch": 2.093260188087774, "grad_norm": 4.556411266326904, "learning_rate": 4.648756786596623e-06, "loss": 0.9377, "step": 26710 }, { "epoch": 2.0933385579937305, "grad_norm": 10.123272895812988, "learning_rate": 4.648449187010765e-06, "loss": 1.327, "step": 26711 }, { "epoch": 2.0934169278996864, "grad_norm": 3.797667980194092, "learning_rate": 4.648141588762025e-06, "loss": 1.374, "step": 26712 }, { "epoch": 2.0934952978056427, "grad_norm": 3.438591480255127, "learning_rate": 4.647833991851572e-06, "loss": 1.2536, "step": 26713 }, { "epoch": 2.0935736677115986, "grad_norm": 3.22381329536438, "learning_rate": 4.647526396280582e-06, "loss": 1.4101, "step": 26714 }, { "epoch": 2.093652037617555, "grad_norm": 2.589540958404541, "learning_rate": 4.647218802050216e-06, "loss": 1.0237, "step": 26715 }, { "epoch": 2.093730407523511, "grad_norm": 4.714776992797852, "learning_rate": 4.64691120916165e-06, "loss": 1.5399, "step": 26716 }, { "epoch": 2.093808777429467, "grad_norm": 2.9849929809570312, "learning_rate": 4.6466036176160526e-06, "loss": 1.2725, "step": 26717 }, { "epoch": 2.093887147335423, "grad_norm": 3.1350317001342773, "learning_rate": 4.646296027414594e-06, "loss": 1.5297, "step": 26718 }, { "epoch": 2.0939655172413794, "grad_norm": 4.108931541442871, "learning_rate": 4.645988438558443e-06, "loss": 1.1265, "step": 26719 }, { "epoch": 2.0940438871473352, "grad_norm": 3.0376839637756348, "learning_rate": 4.645680851048769e-06, "loss": 1.3432, "step": 26720 }, { "epoch": 2.0941222570532916, "grad_norm": 2.982577323913574, "learning_rate": 4.6453732648867436e-06, "loss": 0.9419, "step": 26721 }, { "epoch": 2.0942006269592475, "grad_norm": 6.742863655090332, "learning_rate": 4.645065680073538e-06, "loss": 1.2164, "step": 26722 }, { "epoch": 2.094278996865204, "grad_norm": 2.4241065979003906, "learning_rate": 4.644758096610319e-06, "loss": 1.475, "step": 26723 }, { "epoch": 2.0943573667711597, "grad_norm": 2.8545103073120117, "learning_rate": 4.644450514498258e-06, "loss": 1.7637, "step": 26724 }, { "epoch": 2.094435736677116, "grad_norm": 2.4834747314453125, "learning_rate": 4.644142933738523e-06, "loss": 1.4161, "step": 26725 }, { "epoch": 2.094514106583072, "grad_norm": 3.106826066970825, "learning_rate": 4.643835354332289e-06, "loss": 1.4287, "step": 26726 }, { "epoch": 2.0945924764890282, "grad_norm": 2.4570348262786865, "learning_rate": 4.643527776280719e-06, "loss": 1.2628, "step": 26727 }, { "epoch": 2.094670846394984, "grad_norm": 2.3505141735076904, "learning_rate": 4.643220199584987e-06, "loss": 1.183, "step": 26728 }, { "epoch": 2.0947492163009405, "grad_norm": 4.617321491241455, "learning_rate": 4.642912624246264e-06, "loss": 1.3023, "step": 26729 }, { "epoch": 2.0948275862068964, "grad_norm": 3.034623861312866, "learning_rate": 4.6426050502657165e-06, "loss": 1.5276, "step": 26730 }, { "epoch": 2.0949059561128527, "grad_norm": 2.576003074645996, "learning_rate": 4.6422974776445155e-06, "loss": 1.4369, "step": 26731 }, { "epoch": 2.0949843260188086, "grad_norm": 2.6062159538269043, "learning_rate": 4.641989906383831e-06, "loss": 1.3113, "step": 26732 }, { "epoch": 2.095062695924765, "grad_norm": 3.202662229537964, "learning_rate": 4.641682336484834e-06, "loss": 1.3989, "step": 26733 }, { "epoch": 2.095141065830721, "grad_norm": 2.652834415435791, "learning_rate": 4.641374767948693e-06, "loss": 1.5827, "step": 26734 }, { "epoch": 2.095219435736677, "grad_norm": 8.298819541931152, "learning_rate": 4.6410672007765775e-06, "loss": 1.4878, "step": 26735 }, { "epoch": 2.095297805642633, "grad_norm": 6.495259761810303, "learning_rate": 4.640759634969659e-06, "loss": 1.7593, "step": 26736 }, { "epoch": 2.0953761755485893, "grad_norm": 5.415072917938232, "learning_rate": 4.640452070529107e-06, "loss": 1.7531, "step": 26737 }, { "epoch": 2.0954545454545457, "grad_norm": 3.7720863819122314, "learning_rate": 4.640144507456089e-06, "loss": 1.4447, "step": 26738 }, { "epoch": 2.0955329153605016, "grad_norm": 2.4383389949798584, "learning_rate": 4.639836945751776e-06, "loss": 1.0396, "step": 26739 }, { "epoch": 2.0956112852664575, "grad_norm": 7.6364665031433105, "learning_rate": 4.6395293854173395e-06, "loss": 1.0853, "step": 26740 }, { "epoch": 2.095689655172414, "grad_norm": 4.518570899963379, "learning_rate": 4.639221826453949e-06, "loss": 1.2931, "step": 26741 }, { "epoch": 2.09576802507837, "grad_norm": 2.2565619945526123, "learning_rate": 4.638914268862772e-06, "loss": 1.3594, "step": 26742 }, { "epoch": 2.095846394984326, "grad_norm": 4.013350009918213, "learning_rate": 4.63860671264498e-06, "loss": 1.454, "step": 26743 }, { "epoch": 2.0959247648902823, "grad_norm": 3.236138343811035, "learning_rate": 4.638299157801743e-06, "loss": 1.2849, "step": 26744 }, { "epoch": 2.0960031347962382, "grad_norm": 3.9770214557647705, "learning_rate": 4.637991604334229e-06, "loss": 1.3443, "step": 26745 }, { "epoch": 2.0960815047021946, "grad_norm": 3.159379005432129, "learning_rate": 4.637684052243609e-06, "loss": 1.3684, "step": 26746 }, { "epoch": 2.0961598746081505, "grad_norm": 3.697136878967285, "learning_rate": 4.637376501531053e-06, "loss": 1.4025, "step": 26747 }, { "epoch": 2.096238244514107, "grad_norm": 3.2134103775024414, "learning_rate": 4.637068952197732e-06, "loss": 1.2669, "step": 26748 }, { "epoch": 2.0963166144200627, "grad_norm": 7.537684917449951, "learning_rate": 4.636761404244813e-06, "loss": 1.2842, "step": 26749 }, { "epoch": 2.096394984326019, "grad_norm": 2.0941927433013916, "learning_rate": 4.636453857673466e-06, "loss": 1.3233, "step": 26750 }, { "epoch": 2.096473354231975, "grad_norm": 4.9021477699279785, "learning_rate": 4.6361463124848635e-06, "loss": 1.166, "step": 26751 }, { "epoch": 2.0965517241379312, "grad_norm": 2.837970495223999, "learning_rate": 4.635838768680174e-06, "loss": 1.379, "step": 26752 }, { "epoch": 2.096630094043887, "grad_norm": 2.391339063644409, "learning_rate": 4.6355312262605654e-06, "loss": 1.0029, "step": 26753 }, { "epoch": 2.0967084639498434, "grad_norm": 4.299483299255371, "learning_rate": 4.635223685227209e-06, "loss": 1.2733, "step": 26754 }, { "epoch": 2.0967868338557993, "grad_norm": 2.7453224658966064, "learning_rate": 4.6349161455812746e-06, "loss": 1.3679, "step": 26755 }, { "epoch": 2.0968652037617557, "grad_norm": 3.4939112663269043, "learning_rate": 4.6346086073239326e-06, "loss": 1.1433, "step": 26756 }, { "epoch": 2.0969435736677116, "grad_norm": 2.340092658996582, "learning_rate": 4.63430107045635e-06, "loss": 1.2098, "step": 26757 }, { "epoch": 2.097021943573668, "grad_norm": 5.139881610870361, "learning_rate": 4.633993534979699e-06, "loss": 1.1699, "step": 26758 }, { "epoch": 2.097100313479624, "grad_norm": 3.2581372261047363, "learning_rate": 4.63368600089515e-06, "loss": 1.3303, "step": 26759 }, { "epoch": 2.09717868338558, "grad_norm": 3.725118398666382, "learning_rate": 4.63337846820387e-06, "loss": 1.2387, "step": 26760 }, { "epoch": 2.097257053291536, "grad_norm": 2.2367167472839355, "learning_rate": 4.633070936907029e-06, "loss": 1.1503, "step": 26761 }, { "epoch": 2.0973354231974923, "grad_norm": 4.073482990264893, "learning_rate": 4.632763407005798e-06, "loss": 1.4452, "step": 26762 }, { "epoch": 2.097413793103448, "grad_norm": 15.364054679870605, "learning_rate": 4.632455878501349e-06, "loss": 1.4232, "step": 26763 }, { "epoch": 2.0974921630094046, "grad_norm": 2.69301176071167, "learning_rate": 4.632148351394846e-06, "loss": 1.2853, "step": 26764 }, { "epoch": 2.0975705329153604, "grad_norm": 10.052230834960938, "learning_rate": 4.6318408256874626e-06, "loss": 0.9365, "step": 26765 }, { "epoch": 2.0976489028213168, "grad_norm": 3.4782891273498535, "learning_rate": 4.631533301380368e-06, "loss": 1.4691, "step": 26766 }, { "epoch": 2.0977272727272727, "grad_norm": 4.468391418457031, "learning_rate": 4.631225778474732e-06, "loss": 1.3661, "step": 26767 }, { "epoch": 2.097805642633229, "grad_norm": 3.524912118911743, "learning_rate": 4.630918256971723e-06, "loss": 1.1137, "step": 26768 }, { "epoch": 2.097884012539185, "grad_norm": 6.490472316741943, "learning_rate": 4.63061073687251e-06, "loss": 1.3436, "step": 26769 }, { "epoch": 2.097962382445141, "grad_norm": 10.396383285522461, "learning_rate": 4.630303218178268e-06, "loss": 1.9158, "step": 26770 }, { "epoch": 2.098040752351097, "grad_norm": 3.558912515640259, "learning_rate": 4.6299957008901586e-06, "loss": 1.2122, "step": 26771 }, { "epoch": 2.0981191222570534, "grad_norm": 3.0816149711608887, "learning_rate": 4.629688185009357e-06, "loss": 0.8695, "step": 26772 }, { "epoch": 2.0981974921630093, "grad_norm": 3.320889949798584, "learning_rate": 4.629380670537031e-06, "loss": 1.3364, "step": 26773 }, { "epoch": 2.0982758620689657, "grad_norm": 4.100924491882324, "learning_rate": 4.629073157474352e-06, "loss": 1.1201, "step": 26774 }, { "epoch": 2.0983542319749215, "grad_norm": 2.9729838371276855, "learning_rate": 4.628765645822487e-06, "loss": 1.5104, "step": 26775 }, { "epoch": 2.098432601880878, "grad_norm": 3.4790539741516113, "learning_rate": 4.628458135582606e-06, "loss": 1.4644, "step": 26776 }, { "epoch": 2.0985109717868338, "grad_norm": 2.625476598739624, "learning_rate": 4.628150626755881e-06, "loss": 1.4516, "step": 26777 }, { "epoch": 2.09858934169279, "grad_norm": 2.7947347164154053, "learning_rate": 4.62784311934348e-06, "loss": 1.5027, "step": 26778 }, { "epoch": 2.098667711598746, "grad_norm": 5.6053972244262695, "learning_rate": 4.627535613346572e-06, "loss": 1.3448, "step": 26779 }, { "epoch": 2.0987460815047023, "grad_norm": 3.2803571224212646, "learning_rate": 4.627228108766327e-06, "loss": 1.0522, "step": 26780 }, { "epoch": 2.098824451410658, "grad_norm": 2.4922330379486084, "learning_rate": 4.626920605603913e-06, "loss": 1.2551, "step": 26781 }, { "epoch": 2.0989028213166145, "grad_norm": 2.7778208255767822, "learning_rate": 4.626613103860506e-06, "loss": 1.2349, "step": 26782 }, { "epoch": 2.0989811912225704, "grad_norm": 2.904468059539795, "learning_rate": 4.626305603537267e-06, "loss": 1.3853, "step": 26783 }, { "epoch": 2.0990595611285268, "grad_norm": 3.248783826828003, "learning_rate": 4.6259981046353714e-06, "loss": 0.9637, "step": 26784 }, { "epoch": 2.0991379310344827, "grad_norm": 3.698866605758667, "learning_rate": 4.625690607155987e-06, "loss": 1.429, "step": 26785 }, { "epoch": 2.099216300940439, "grad_norm": 2.4617221355438232, "learning_rate": 4.625383111100283e-06, "loss": 1.5207, "step": 26786 }, { "epoch": 2.099294670846395, "grad_norm": 6.348447799682617, "learning_rate": 4.625075616469429e-06, "loss": 1.1939, "step": 26787 }, { "epoch": 2.099373040752351, "grad_norm": 3.8403494358062744, "learning_rate": 4.624768123264593e-06, "loss": 1.2344, "step": 26788 }, { "epoch": 2.099451410658307, "grad_norm": 5.6009039878845215, "learning_rate": 4.62446063148695e-06, "loss": 0.709, "step": 26789 }, { "epoch": 2.0995297805642634, "grad_norm": 2.1038331985473633, "learning_rate": 4.624153141137663e-06, "loss": 1.3272, "step": 26790 }, { "epoch": 2.0996081504702193, "grad_norm": 5.792874813079834, "learning_rate": 4.623845652217906e-06, "loss": 1.4813, "step": 26791 }, { "epoch": 2.0996865203761756, "grad_norm": 4.0162434577941895, "learning_rate": 4.623538164728845e-06, "loss": 1.5474, "step": 26792 }, { "epoch": 2.0997648902821315, "grad_norm": 2.7795841693878174, "learning_rate": 4.6232306786716545e-06, "loss": 0.4303, "step": 26793 }, { "epoch": 2.099843260188088, "grad_norm": 3.0515048503875732, "learning_rate": 4.6229231940474985e-06, "loss": 1.1886, "step": 26794 }, { "epoch": 2.0999216300940438, "grad_norm": 3.1706721782684326, "learning_rate": 4.622615710857549e-06, "loss": 1.5929, "step": 26795 }, { "epoch": 2.1, "grad_norm": 3.6617932319641113, "learning_rate": 4.622308229102976e-06, "loss": 1.2345, "step": 26796 }, { "epoch": 2.100078369905956, "grad_norm": 2.875981092453003, "learning_rate": 4.622000748784949e-06, "loss": 1.2592, "step": 26797 }, { "epoch": 2.1001567398119123, "grad_norm": 8.66759204864502, "learning_rate": 4.6216932699046365e-06, "loss": 0.4813, "step": 26798 }, { "epoch": 2.100235109717868, "grad_norm": 3.4182331562042236, "learning_rate": 4.621385792463209e-06, "loss": 1.4969, "step": 26799 }, { "epoch": 2.1003134796238245, "grad_norm": 3.3265204429626465, "learning_rate": 4.6210783164618365e-06, "loss": 1.6897, "step": 26800 }, { "epoch": 2.1003918495297804, "grad_norm": 2.396043539047241, "learning_rate": 4.6207708419016855e-06, "loss": 1.1945, "step": 26801 }, { "epoch": 2.1004702194357368, "grad_norm": 6.51700496673584, "learning_rate": 4.6204633687839275e-06, "loss": 1.8311, "step": 26802 }, { "epoch": 2.1005485893416926, "grad_norm": 3.5754494667053223, "learning_rate": 4.620155897109732e-06, "loss": 1.0908, "step": 26803 }, { "epoch": 2.100626959247649, "grad_norm": 3.3128163814544678, "learning_rate": 4.619848426880269e-06, "loss": 1.3763, "step": 26804 }, { "epoch": 2.100705329153605, "grad_norm": 4.11736536026001, "learning_rate": 4.6195409580967076e-06, "loss": 1.2619, "step": 26805 }, { "epoch": 2.100783699059561, "grad_norm": 3.052823305130005, "learning_rate": 4.619233490760216e-06, "loss": 1.53, "step": 26806 }, { "epoch": 2.100862068965517, "grad_norm": 16.892606735229492, "learning_rate": 4.618926024871965e-06, "loss": 1.5707, "step": 26807 }, { "epoch": 2.1009404388714734, "grad_norm": 4.756246566772461, "learning_rate": 4.618618560433126e-06, "loss": 1.3007, "step": 26808 }, { "epoch": 2.1010188087774293, "grad_norm": 3.42024564743042, "learning_rate": 4.618311097444862e-06, "loss": 1.8604, "step": 26809 }, { "epoch": 2.1010971786833856, "grad_norm": 4.268091678619385, "learning_rate": 4.618003635908348e-06, "loss": 1.5182, "step": 26810 }, { "epoch": 2.1011755485893415, "grad_norm": 3.540043592453003, "learning_rate": 4.617696175824753e-06, "loss": 1.0558, "step": 26811 }, { "epoch": 2.101253918495298, "grad_norm": 4.943821907043457, "learning_rate": 4.617388717195246e-06, "loss": 1.1624, "step": 26812 }, { "epoch": 2.1013322884012537, "grad_norm": 6.266630172729492, "learning_rate": 4.617081260020994e-06, "loss": 1.8701, "step": 26813 }, { "epoch": 2.10141065830721, "grad_norm": 7.012069225311279, "learning_rate": 4.616773804303168e-06, "loss": 1.3148, "step": 26814 }, { "epoch": 2.101489028213166, "grad_norm": 3.8922648429870605, "learning_rate": 4.61646635004294e-06, "loss": 1.2236, "step": 26815 }, { "epoch": 2.1015673981191223, "grad_norm": 2.3904683589935303, "learning_rate": 4.616158897241474e-06, "loss": 0.8631, "step": 26816 }, { "epoch": 2.101645768025078, "grad_norm": 3.129335403442383, "learning_rate": 4.615851445899945e-06, "loss": 1.5704, "step": 26817 }, { "epoch": 2.1017241379310345, "grad_norm": 2.8529553413391113, "learning_rate": 4.615543996019517e-06, "loss": 1.1289, "step": 26818 }, { "epoch": 2.1018025078369904, "grad_norm": 2.6980745792388916, "learning_rate": 4.6152365476013654e-06, "loss": 1.3754, "step": 26819 }, { "epoch": 2.1018808777429467, "grad_norm": 3.31630277633667, "learning_rate": 4.614929100646654e-06, "loss": 1.4138, "step": 26820 }, { "epoch": 2.1019592476489026, "grad_norm": 3.0610129833221436, "learning_rate": 4.614621655156555e-06, "loss": 1.4042, "step": 26821 }, { "epoch": 2.102037617554859, "grad_norm": 3.7298836708068848, "learning_rate": 4.614314211132237e-06, "loss": 1.061, "step": 26822 }, { "epoch": 2.102115987460815, "grad_norm": 3.841843843460083, "learning_rate": 4.614006768574871e-06, "loss": 1.038, "step": 26823 }, { "epoch": 2.102194357366771, "grad_norm": 3.4332544803619385, "learning_rate": 4.6136993274856245e-06, "loss": 0.9934, "step": 26824 }, { "epoch": 2.102272727272727, "grad_norm": 3.6085638999938965, "learning_rate": 4.613391887865667e-06, "loss": 1.2822, "step": 26825 }, { "epoch": 2.1023510971786834, "grad_norm": 4.1522111892700195, "learning_rate": 4.6130844497161695e-06, "loss": 1.1557, "step": 26826 }, { "epoch": 2.1024294670846393, "grad_norm": 5.877662658691406, "learning_rate": 4.612777013038298e-06, "loss": 1.1474, "step": 26827 }, { "epoch": 2.1025078369905956, "grad_norm": 2.784076690673828, "learning_rate": 4.612469577833223e-06, "loss": 1.4261, "step": 26828 }, { "epoch": 2.1025862068965515, "grad_norm": 9.085208892822266, "learning_rate": 4.612162144102117e-06, "loss": 2.064, "step": 26829 }, { "epoch": 2.102664576802508, "grad_norm": 7.339325904846191, "learning_rate": 4.611854711846147e-06, "loss": 1.2723, "step": 26830 }, { "epoch": 2.1027429467084637, "grad_norm": 3.937363386154175, "learning_rate": 4.611547281066482e-06, "loss": 1.2299, "step": 26831 }, { "epoch": 2.10282131661442, "grad_norm": 2.688750982284546, "learning_rate": 4.61123985176429e-06, "loss": 1.2203, "step": 26832 }, { "epoch": 2.1028996865203764, "grad_norm": 3.264411449432373, "learning_rate": 4.6109324239407425e-06, "loss": 1.3537, "step": 26833 }, { "epoch": 2.1029780564263323, "grad_norm": 3.033419132232666, "learning_rate": 4.6106249975970095e-06, "loss": 1.2045, "step": 26834 }, { "epoch": 2.103056426332288, "grad_norm": 2.557854413986206, "learning_rate": 4.6103175727342585e-06, "loss": 1.3626, "step": 26835 }, { "epoch": 2.1031347962382445, "grad_norm": 2.459573745727539, "learning_rate": 4.6100101493536585e-06, "loss": 1.421, "step": 26836 }, { "epoch": 2.103213166144201, "grad_norm": 2.5491230487823486, "learning_rate": 4.609702727456381e-06, "loss": 1.2459, "step": 26837 }, { "epoch": 2.1032915360501567, "grad_norm": 3.1991047859191895, "learning_rate": 4.609395307043594e-06, "loss": 1.2929, "step": 26838 }, { "epoch": 2.103369905956113, "grad_norm": 7.469272136688232, "learning_rate": 4.609087888116465e-06, "loss": 1.7996, "step": 26839 }, { "epoch": 2.103448275862069, "grad_norm": 3.0089352130889893, "learning_rate": 4.608780470676165e-06, "loss": 1.1828, "step": 26840 }, { "epoch": 2.1035266457680253, "grad_norm": 2.562001943588257, "learning_rate": 4.608473054723864e-06, "loss": 1.3281, "step": 26841 }, { "epoch": 2.103605015673981, "grad_norm": 2.45577073097229, "learning_rate": 4.60816564026073e-06, "loss": 1.2528, "step": 26842 }, { "epoch": 2.1036833855799375, "grad_norm": 4.917741775512695, "learning_rate": 4.607858227287933e-06, "loss": 1.3893, "step": 26843 }, { "epoch": 2.1037617554858934, "grad_norm": 3.662118673324585, "learning_rate": 4.607550815806642e-06, "loss": 1.6143, "step": 26844 }, { "epoch": 2.1038401253918497, "grad_norm": 3.305542230606079, "learning_rate": 4.607243405818027e-06, "loss": 1.126, "step": 26845 }, { "epoch": 2.1039184952978056, "grad_norm": 3.768843650817871, "learning_rate": 4.606935997323253e-06, "loss": 1.4599, "step": 26846 }, { "epoch": 2.103996865203762, "grad_norm": 4.644232749938965, "learning_rate": 4.606628590323494e-06, "loss": 1.5143, "step": 26847 }, { "epoch": 2.104075235109718, "grad_norm": 2.1163063049316406, "learning_rate": 4.6063211848199195e-06, "loss": 1.1326, "step": 26848 }, { "epoch": 2.104153605015674, "grad_norm": 4.713520050048828, "learning_rate": 4.606013780813697e-06, "loss": 0.8907, "step": 26849 }, { "epoch": 2.10423197492163, "grad_norm": 2.467594861984253, "learning_rate": 4.605706378305995e-06, "loss": 1.3976, "step": 26850 }, { "epoch": 2.1043103448275864, "grad_norm": 3.684947967529297, "learning_rate": 4.605398977297981e-06, "loss": 1.2023, "step": 26851 }, { "epoch": 2.1043887147335423, "grad_norm": 2.652472734451294, "learning_rate": 4.605091577790829e-06, "loss": 1.4598, "step": 26852 }, { "epoch": 2.1044670846394986, "grad_norm": 3.3258092403411865, "learning_rate": 4.604784179785706e-06, "loss": 1.391, "step": 26853 }, { "epoch": 2.1045454545454545, "grad_norm": 3.454087257385254, "learning_rate": 4.604476783283781e-06, "loss": 1.5721, "step": 26854 }, { "epoch": 2.104623824451411, "grad_norm": 4.608514785766602, "learning_rate": 4.604169388286223e-06, "loss": 1.9245, "step": 26855 }, { "epoch": 2.1047021943573667, "grad_norm": 3.2517483234405518, "learning_rate": 4.603861994794202e-06, "loss": 1.2802, "step": 26856 }, { "epoch": 2.104780564263323, "grad_norm": 3.9360463619232178, "learning_rate": 4.603554602808885e-06, "loss": 0.8537, "step": 26857 }, { "epoch": 2.104858934169279, "grad_norm": 3.154536724090576, "learning_rate": 4.603247212331442e-06, "loss": 1.2551, "step": 26858 }, { "epoch": 2.1049373040752353, "grad_norm": 2.6563785076141357, "learning_rate": 4.602939823363044e-06, "loss": 0.9872, "step": 26859 }, { "epoch": 2.105015673981191, "grad_norm": 4.465291500091553, "learning_rate": 4.6026324359048605e-06, "loss": 1.0581, "step": 26860 }, { "epoch": 2.1050940438871475, "grad_norm": 3.0559158325195312, "learning_rate": 4.602325049958057e-06, "loss": 0.6373, "step": 26861 }, { "epoch": 2.1051724137931034, "grad_norm": 2.1740097999572754, "learning_rate": 4.602017665523805e-06, "loss": 1.5321, "step": 26862 }, { "epoch": 2.1052507836990597, "grad_norm": 11.812274932861328, "learning_rate": 4.601710282603273e-06, "loss": 1.2393, "step": 26863 }, { "epoch": 2.1053291536050156, "grad_norm": 2.8439414501190186, "learning_rate": 4.601402901197633e-06, "loss": 1.152, "step": 26864 }, { "epoch": 2.105407523510972, "grad_norm": 2.3108906745910645, "learning_rate": 4.601095521308049e-06, "loss": 1.5217, "step": 26865 }, { "epoch": 2.105485893416928, "grad_norm": 3.107296943664551, "learning_rate": 4.600788142935693e-06, "loss": 1.1813, "step": 26866 }, { "epoch": 2.105564263322884, "grad_norm": 5.057979106903076, "learning_rate": 4.600480766081735e-06, "loss": 1.3022, "step": 26867 }, { "epoch": 2.10564263322884, "grad_norm": 2.758615493774414, "learning_rate": 4.600173390747344e-06, "loss": 1.1294, "step": 26868 }, { "epoch": 2.1057210031347964, "grad_norm": 3.0806522369384766, "learning_rate": 4.5998660169336854e-06, "loss": 1.0134, "step": 26869 }, { "epoch": 2.1057993730407523, "grad_norm": 3.3540351390838623, "learning_rate": 4.599558644641931e-06, "loss": 1.4298, "step": 26870 }, { "epoch": 2.1058777429467086, "grad_norm": 2.8522329330444336, "learning_rate": 4.599251273873253e-06, "loss": 1.2378, "step": 26871 }, { "epoch": 2.1059561128526645, "grad_norm": 5.606349468231201, "learning_rate": 4.598943904628814e-06, "loss": 1.7053, "step": 26872 }, { "epoch": 2.106034482758621, "grad_norm": 4.459517002105713, "learning_rate": 4.598636536909788e-06, "loss": 1.4659, "step": 26873 }, { "epoch": 2.1061128526645767, "grad_norm": 3.4714698791503906, "learning_rate": 4.5983291707173414e-06, "loss": 1.6867, "step": 26874 }, { "epoch": 2.106191222570533, "grad_norm": 2.4500842094421387, "learning_rate": 4.598021806052646e-06, "loss": 1.3474, "step": 26875 }, { "epoch": 2.106269592476489, "grad_norm": 5.10873556137085, "learning_rate": 4.597714442916868e-06, "loss": 1.1564, "step": 26876 }, { "epoch": 2.1063479623824453, "grad_norm": 4.99930477142334, "learning_rate": 4.597407081311176e-06, "loss": 2.0625, "step": 26877 }, { "epoch": 2.106426332288401, "grad_norm": 2.6055147647857666, "learning_rate": 4.597099721236743e-06, "loss": 1.2578, "step": 26878 }, { "epoch": 2.1065047021943575, "grad_norm": 2.015690326690674, "learning_rate": 4.596792362694736e-06, "loss": 0.881, "step": 26879 }, { "epoch": 2.1065830721003134, "grad_norm": 5.508071422576904, "learning_rate": 4.596485005686323e-06, "loss": 2.2292, "step": 26880 }, { "epoch": 2.1066614420062697, "grad_norm": 4.242338180541992, "learning_rate": 4.596177650212674e-06, "loss": 1.1943, "step": 26881 }, { "epoch": 2.1067398119122256, "grad_norm": 3.16886568069458, "learning_rate": 4.595870296274958e-06, "loss": 0.5653, "step": 26882 }, { "epoch": 2.106818181818182, "grad_norm": 3.695317268371582, "learning_rate": 4.5955629438743434e-06, "loss": 1.3476, "step": 26883 }, { "epoch": 2.106896551724138, "grad_norm": 4.425429821014404, "learning_rate": 4.595255593011998e-06, "loss": 1.2753, "step": 26884 }, { "epoch": 2.106974921630094, "grad_norm": 6.708230495452881, "learning_rate": 4.594948243689094e-06, "loss": 0.8185, "step": 26885 }, { "epoch": 2.10705329153605, "grad_norm": 3.3756184577941895, "learning_rate": 4.5946408959068e-06, "loss": 1.2769, "step": 26886 }, { "epoch": 2.1071316614420064, "grad_norm": 2.8411364555358887, "learning_rate": 4.594333549666283e-06, "loss": 1.4536, "step": 26887 }, { "epoch": 2.1072100313479623, "grad_norm": 2.7375762462615967, "learning_rate": 4.594026204968712e-06, "loss": 1.2025, "step": 26888 }, { "epoch": 2.1072884012539186, "grad_norm": 5.581597328186035, "learning_rate": 4.593718861815256e-06, "loss": 1.6528, "step": 26889 }, { "epoch": 2.1073667711598745, "grad_norm": 3.7893216609954834, "learning_rate": 4.593411520207089e-06, "loss": 1.3443, "step": 26890 }, { "epoch": 2.107445141065831, "grad_norm": 3.395965099334717, "learning_rate": 4.593104180145371e-06, "loss": 1.1682, "step": 26891 }, { "epoch": 2.1075235109717867, "grad_norm": 3.1166818141937256, "learning_rate": 4.592796841631277e-06, "loss": 1.0462, "step": 26892 }, { "epoch": 2.107601880877743, "grad_norm": 3.992342233657837, "learning_rate": 4.592489504665975e-06, "loss": 1.6392, "step": 26893 }, { "epoch": 2.107680250783699, "grad_norm": 2.982234477996826, "learning_rate": 4.592182169250634e-06, "loss": 1.3265, "step": 26894 }, { "epoch": 2.1077586206896552, "grad_norm": 4.056735038757324, "learning_rate": 4.591874835386423e-06, "loss": 1.3881, "step": 26895 }, { "epoch": 2.107836990595611, "grad_norm": 7.20294713973999, "learning_rate": 4.591567503074509e-06, "loss": 1.3531, "step": 26896 }, { "epoch": 2.1079153605015675, "grad_norm": 4.490516185760498, "learning_rate": 4.591260172316065e-06, "loss": 1.4978, "step": 26897 }, { "epoch": 2.1079937304075234, "grad_norm": 2.683659076690674, "learning_rate": 4.590952843112255e-06, "loss": 1.3935, "step": 26898 }, { "epoch": 2.1080721003134797, "grad_norm": 2.7210476398468018, "learning_rate": 4.59064551546425e-06, "loss": 1.3224, "step": 26899 }, { "epoch": 2.1081504702194356, "grad_norm": 3.080655813217163, "learning_rate": 4.59033818937322e-06, "loss": 1.2223, "step": 26900 }, { "epoch": 2.108228840125392, "grad_norm": 3.7664926052093506, "learning_rate": 4.590030864840335e-06, "loss": 1.3057, "step": 26901 }, { "epoch": 2.108307210031348, "grad_norm": 3.173428773880005, "learning_rate": 4.5897235418667595e-06, "loss": 1.685, "step": 26902 }, { "epoch": 2.108385579937304, "grad_norm": 3.8709259033203125, "learning_rate": 4.5894162204536645e-06, "loss": 1.3186, "step": 26903 }, { "epoch": 2.10846394984326, "grad_norm": 3.5151448249816895, "learning_rate": 4.5891089006022204e-06, "loss": 1.4576, "step": 26904 }, { "epoch": 2.1085423197492164, "grad_norm": 9.702369689941406, "learning_rate": 4.588801582313596e-06, "loss": 1.9201, "step": 26905 }, { "epoch": 2.1086206896551722, "grad_norm": 2.9171693325042725, "learning_rate": 4.588494265588958e-06, "loss": 1.3053, "step": 26906 }, { "epoch": 2.1086990595611286, "grad_norm": 3.206617593765259, "learning_rate": 4.588186950429477e-06, "loss": 1.4087, "step": 26907 }, { "epoch": 2.1087774294670845, "grad_norm": 3.5457701683044434, "learning_rate": 4.58787963683632e-06, "loss": 1.2705, "step": 26908 }, { "epoch": 2.108855799373041, "grad_norm": 6.077681064605713, "learning_rate": 4.5875723248106596e-06, "loss": 1.1843, "step": 26909 }, { "epoch": 2.1089341692789967, "grad_norm": 5.519111156463623, "learning_rate": 4.58726501435366e-06, "loss": 1.2236, "step": 26910 }, { "epoch": 2.109012539184953, "grad_norm": 2.417607307434082, "learning_rate": 4.586957705466493e-06, "loss": 1.2014, "step": 26911 }, { "epoch": 2.109090909090909, "grad_norm": 2.692901134490967, "learning_rate": 4.5866503981503275e-06, "loss": 1.3661, "step": 26912 }, { "epoch": 2.1091692789968652, "grad_norm": 4.196122169494629, "learning_rate": 4.5863430924063315e-06, "loss": 1.3327, "step": 26913 }, { "epoch": 2.109247648902821, "grad_norm": 4.2699408531188965, "learning_rate": 4.586035788235673e-06, "loss": 2.132, "step": 26914 }, { "epoch": 2.1093260188087775, "grad_norm": 4.091533184051514, "learning_rate": 4.585728485639521e-06, "loss": 1.7444, "step": 26915 }, { "epoch": 2.1094043887147333, "grad_norm": 6.521609306335449, "learning_rate": 4.585421184619048e-06, "loss": 1.1395, "step": 26916 }, { "epoch": 2.1094827586206897, "grad_norm": 2.863787889480591, "learning_rate": 4.585113885175417e-06, "loss": 1.5941, "step": 26917 }, { "epoch": 2.1095611285266456, "grad_norm": 1.819787621498108, "learning_rate": 4.584806587309801e-06, "loss": 1.4404, "step": 26918 }, { "epoch": 2.109639498432602, "grad_norm": 6.471523761749268, "learning_rate": 4.584499291023366e-06, "loss": 1.3973, "step": 26919 }, { "epoch": 2.109717868338558, "grad_norm": 2.577890157699585, "learning_rate": 4.584191996317285e-06, "loss": 1.4105, "step": 26920 }, { "epoch": 2.109796238244514, "grad_norm": 2.567450761795044, "learning_rate": 4.5838847031927215e-06, "loss": 1.4684, "step": 26921 }, { "epoch": 2.10987460815047, "grad_norm": 2.701932668685913, "learning_rate": 4.583577411650847e-06, "loss": 0.8987, "step": 26922 }, { "epoch": 2.1099529780564263, "grad_norm": 9.252248764038086, "learning_rate": 4.583270121692831e-06, "loss": 0.9373, "step": 26923 }, { "epoch": 2.1100313479623822, "grad_norm": 3.355879068374634, "learning_rate": 4.582962833319843e-06, "loss": 1.1527, "step": 26924 }, { "epoch": 2.1101097178683386, "grad_norm": 3.567694664001465, "learning_rate": 4.582655546533048e-06, "loss": 1.6097, "step": 26925 }, { "epoch": 2.1101880877742945, "grad_norm": 3.3639988899230957, "learning_rate": 4.582348261333617e-06, "loss": 1.3615, "step": 26926 }, { "epoch": 2.110266457680251, "grad_norm": 4.45464563369751, "learning_rate": 4.582040977722721e-06, "loss": 0.6851, "step": 26927 }, { "epoch": 2.110344827586207, "grad_norm": 2.8975329399108887, "learning_rate": 4.581733695701525e-06, "loss": 0.8961, "step": 26928 }, { "epoch": 2.110423197492163, "grad_norm": 3.813506603240967, "learning_rate": 4.581426415271197e-06, "loss": 1.3452, "step": 26929 }, { "epoch": 2.110501567398119, "grad_norm": 4.219841480255127, "learning_rate": 4.58111913643291e-06, "loss": 1.4989, "step": 26930 }, { "epoch": 2.1105799373040752, "grad_norm": 2.072751522064209, "learning_rate": 4.5808118591878314e-06, "loss": 1.3722, "step": 26931 }, { "epoch": 2.1106583072100316, "grad_norm": 3.3813235759735107, "learning_rate": 4.580504583537128e-06, "loss": 1.2462, "step": 26932 }, { "epoch": 2.1107366771159874, "grad_norm": 2.855646848678589, "learning_rate": 4.5801973094819695e-06, "loss": 1.4184, "step": 26933 }, { "epoch": 2.110815047021944, "grad_norm": 3.2792508602142334, "learning_rate": 4.579890037023525e-06, "loss": 1.1546, "step": 26934 }, { "epoch": 2.1108934169278997, "grad_norm": 3.6817519664764404, "learning_rate": 4.579582766162965e-06, "loss": 1.3493, "step": 26935 }, { "epoch": 2.110971786833856, "grad_norm": 4.663439750671387, "learning_rate": 4.579275496901453e-06, "loss": 1.2887, "step": 26936 }, { "epoch": 2.111050156739812, "grad_norm": 13.894116401672363, "learning_rate": 4.578968229240162e-06, "loss": 1.5175, "step": 26937 }, { "epoch": 2.111128526645768, "grad_norm": 2.9642038345336914, "learning_rate": 4.578660963180261e-06, "loss": 1.6095, "step": 26938 }, { "epoch": 2.111206896551724, "grad_norm": 3.3770813941955566, "learning_rate": 4.578353698722916e-06, "loss": 1.1766, "step": 26939 }, { "epoch": 2.1112852664576804, "grad_norm": 3.529745101928711, "learning_rate": 4.578046435869297e-06, "loss": 1.2567, "step": 26940 }, { "epoch": 2.1113636363636363, "grad_norm": 4.447443008422852, "learning_rate": 4.577739174620572e-06, "loss": 1.8951, "step": 26941 }, { "epoch": 2.1114420062695927, "grad_norm": 3.5675244331359863, "learning_rate": 4.577431914977912e-06, "loss": 2.0376, "step": 26942 }, { "epoch": 2.1115203761755486, "grad_norm": 2.7506234645843506, "learning_rate": 4.577124656942483e-06, "loss": 1.3988, "step": 26943 }, { "epoch": 2.111598746081505, "grad_norm": 3.1426751613616943, "learning_rate": 4.576817400515455e-06, "loss": 1.5702, "step": 26944 }, { "epoch": 2.1116771159874608, "grad_norm": 3.18738055229187, "learning_rate": 4.576510145697996e-06, "loss": 1.5885, "step": 26945 }, { "epoch": 2.111755485893417, "grad_norm": 3.943208694458008, "learning_rate": 4.576202892491275e-06, "loss": 1.0533, "step": 26946 }, { "epoch": 2.111833855799373, "grad_norm": 5.895715713500977, "learning_rate": 4.57589564089646e-06, "loss": 0.7234, "step": 26947 }, { "epoch": 2.1119122257053293, "grad_norm": 4.762284755706787, "learning_rate": 4.57558839091472e-06, "loss": 1.1247, "step": 26948 }, { "epoch": 2.111990595611285, "grad_norm": 2.5065855979919434, "learning_rate": 4.575281142547224e-06, "loss": 1.5551, "step": 26949 }, { "epoch": 2.1120689655172415, "grad_norm": 2.1517882347106934, "learning_rate": 4.574973895795142e-06, "loss": 1.4459, "step": 26950 }, { "epoch": 2.1121473354231974, "grad_norm": 10.671875, "learning_rate": 4.574666650659639e-06, "loss": 1.0374, "step": 26951 }, { "epoch": 2.1122257053291538, "grad_norm": 3.231485605239868, "learning_rate": 4.574359407141886e-06, "loss": 1.3923, "step": 26952 }, { "epoch": 2.1123040752351097, "grad_norm": 3.2282299995422363, "learning_rate": 4.574052165243053e-06, "loss": 1.204, "step": 26953 }, { "epoch": 2.112382445141066, "grad_norm": 2.359009265899658, "learning_rate": 4.573744924964304e-06, "loss": 0.7908, "step": 26954 }, { "epoch": 2.112460815047022, "grad_norm": 2.9842090606689453, "learning_rate": 4.573437686306811e-06, "loss": 1.7264, "step": 26955 }, { "epoch": 2.112539184952978, "grad_norm": 2.9178688526153564, "learning_rate": 4.573130449271743e-06, "loss": 1.2533, "step": 26956 }, { "epoch": 2.112617554858934, "grad_norm": 2.808285713195801, "learning_rate": 4.572823213860268e-06, "loss": 1.5182, "step": 26957 }, { "epoch": 2.1126959247648904, "grad_norm": 3.9919540882110596, "learning_rate": 4.572515980073553e-06, "loss": 1.4964, "step": 26958 }, { "epoch": 2.1127742946708463, "grad_norm": 3.490138292312622, "learning_rate": 4.572208747912767e-06, "loss": 1.2383, "step": 26959 }, { "epoch": 2.1128526645768027, "grad_norm": 7.840473175048828, "learning_rate": 4.571901517379079e-06, "loss": 1.3052, "step": 26960 }, { "epoch": 2.1129310344827585, "grad_norm": 6.082953929901123, "learning_rate": 4.57159428847366e-06, "loss": 1.2821, "step": 26961 }, { "epoch": 2.113009404388715, "grad_norm": 4.294466972351074, "learning_rate": 4.571287061197675e-06, "loss": 1.458, "step": 26962 }, { "epoch": 2.1130877742946708, "grad_norm": 2.528986930847168, "learning_rate": 4.570979835552294e-06, "loss": 1.3284, "step": 26963 }, { "epoch": 2.113166144200627, "grad_norm": 13.552184104919434, "learning_rate": 4.570672611538683e-06, "loss": 0.8245, "step": 26964 }, { "epoch": 2.113244514106583, "grad_norm": 2.5055599212646484, "learning_rate": 4.570365389158018e-06, "loss": 1.1215, "step": 26965 }, { "epoch": 2.1133228840125393, "grad_norm": 2.496969223022461, "learning_rate": 4.5700581684114585e-06, "loss": 1.382, "step": 26966 }, { "epoch": 2.113401253918495, "grad_norm": 2.217283010482788, "learning_rate": 4.569750949300177e-06, "loss": 1.3463, "step": 26967 }, { "epoch": 2.1134796238244515, "grad_norm": 3.5864651203155518, "learning_rate": 4.569443731825344e-06, "loss": 1.4595, "step": 26968 }, { "epoch": 2.1135579937304074, "grad_norm": 2.2150697708129883, "learning_rate": 4.5691365159881245e-06, "loss": 1.2665, "step": 26969 }, { "epoch": 2.1136363636363638, "grad_norm": 3.601788282394409, "learning_rate": 4.5688293017896885e-06, "loss": 1.3643, "step": 26970 }, { "epoch": 2.1137147335423196, "grad_norm": 8.914862632751465, "learning_rate": 4.568522089231203e-06, "loss": 1.1612, "step": 26971 }, { "epoch": 2.113793103448276, "grad_norm": 9.868558883666992, "learning_rate": 4.5682148783138415e-06, "loss": 1.3597, "step": 26972 }, { "epoch": 2.113871473354232, "grad_norm": 2.612860679626465, "learning_rate": 4.567907669038765e-06, "loss": 1.239, "step": 26973 }, { "epoch": 2.113949843260188, "grad_norm": 3.6043906211853027, "learning_rate": 4.5676004614071475e-06, "loss": 1.3766, "step": 26974 }, { "epoch": 2.114028213166144, "grad_norm": 3.294562339782715, "learning_rate": 4.5672932554201555e-06, "loss": 1.3723, "step": 26975 }, { "epoch": 2.1141065830721004, "grad_norm": 3.037632465362549, "learning_rate": 4.566986051078959e-06, "loss": 1.6427, "step": 26976 }, { "epoch": 2.1141849529780563, "grad_norm": 2.3129944801330566, "learning_rate": 4.566678848384723e-06, "loss": 1.4284, "step": 26977 }, { "epoch": 2.1142633228840126, "grad_norm": 18.67852020263672, "learning_rate": 4.5663716473386186e-06, "loss": 1.505, "step": 26978 }, { "epoch": 2.1143416927899685, "grad_norm": 3.3070027828216553, "learning_rate": 4.566064447941816e-06, "loss": 1.3204, "step": 26979 }, { "epoch": 2.114420062695925, "grad_norm": 2.227423667907715, "learning_rate": 4.565757250195478e-06, "loss": 1.306, "step": 26980 }, { "epoch": 2.1144984326018808, "grad_norm": 3.8895492553710938, "learning_rate": 4.565450054100778e-06, "loss": 1.155, "step": 26981 }, { "epoch": 2.114576802507837, "grad_norm": 2.944514274597168, "learning_rate": 4.565142859658883e-06, "loss": 1.173, "step": 26982 }, { "epoch": 2.114655172413793, "grad_norm": 2.9499030113220215, "learning_rate": 4.564835666870961e-06, "loss": 1.6021, "step": 26983 }, { "epoch": 2.1147335423197493, "grad_norm": 2.825984477996826, "learning_rate": 4.564528475738181e-06, "loss": 1.3606, "step": 26984 }, { "epoch": 2.114811912225705, "grad_norm": 3.245281219482422, "learning_rate": 4.564221286261709e-06, "loss": 1.3555, "step": 26985 }, { "epoch": 2.1148902821316615, "grad_norm": 4.067508220672607, "learning_rate": 4.563914098442717e-06, "loss": 1.5316, "step": 26986 }, { "epoch": 2.1149686520376174, "grad_norm": 3.721196413040161, "learning_rate": 4.563606912282373e-06, "loss": 1.2052, "step": 26987 }, { "epoch": 2.1150470219435737, "grad_norm": 3.7360451221466064, "learning_rate": 4.563299727781842e-06, "loss": 1.1183, "step": 26988 }, { "epoch": 2.1151253918495296, "grad_norm": 4.859527587890625, "learning_rate": 4.562992544942296e-06, "loss": 1.6249, "step": 26989 }, { "epoch": 2.115203761755486, "grad_norm": 2.8090224266052246, "learning_rate": 4.5626853637649e-06, "loss": 1.5165, "step": 26990 }, { "epoch": 2.115282131661442, "grad_norm": 4.032922267913818, "learning_rate": 4.562378184250828e-06, "loss": 1.2463, "step": 26991 }, { "epoch": 2.115360501567398, "grad_norm": 7.669566631317139, "learning_rate": 4.56207100640124e-06, "loss": 1.3106, "step": 26992 }, { "epoch": 2.115438871473354, "grad_norm": 2.4077651500701904, "learning_rate": 4.561763830217311e-06, "loss": 1.2966, "step": 26993 }, { "epoch": 2.1155172413793104, "grad_norm": 6.192490577697754, "learning_rate": 4.5614566557002085e-06, "loss": 1.5375, "step": 26994 }, { "epoch": 2.1155956112852663, "grad_norm": 4.375070095062256, "learning_rate": 4.561149482851099e-06, "loss": 0.9629, "step": 26995 }, { "epoch": 2.1156739811912226, "grad_norm": 2.477642059326172, "learning_rate": 4.560842311671151e-06, "loss": 1.1785, "step": 26996 }, { "epoch": 2.1157523510971785, "grad_norm": 2.857815742492676, "learning_rate": 4.560535142161532e-06, "loss": 1.2166, "step": 26997 }, { "epoch": 2.115830721003135, "grad_norm": 2.4312686920166016, "learning_rate": 4.560227974323415e-06, "loss": 1.1735, "step": 26998 }, { "epoch": 2.1159090909090907, "grad_norm": 2.5737404823303223, "learning_rate": 4.559920808157961e-06, "loss": 1.3938, "step": 26999 }, { "epoch": 2.115987460815047, "grad_norm": 3.636612892150879, "learning_rate": 4.559613643666344e-06, "loss": 1.19, "step": 27000 }, { "epoch": 2.116065830721003, "grad_norm": 4.320742130279541, "learning_rate": 4.55930648084973e-06, "loss": 1.4014, "step": 27001 }, { "epoch": 2.1161442006269593, "grad_norm": 2.6278460025787354, "learning_rate": 4.55899931970929e-06, "loss": 1.4267, "step": 27002 }, { "epoch": 2.116222570532915, "grad_norm": 2.045348882675171, "learning_rate": 4.558692160246188e-06, "loss": 0.814, "step": 27003 }, { "epoch": 2.1163009404388715, "grad_norm": 3.1159827709198, "learning_rate": 4.558385002461593e-06, "loss": 1.2582, "step": 27004 }, { "epoch": 2.1163793103448274, "grad_norm": 2.5501601696014404, "learning_rate": 4.558077846356677e-06, "loss": 0.6495, "step": 27005 }, { "epoch": 2.1164576802507837, "grad_norm": 2.423449754714966, "learning_rate": 4.557770691932606e-06, "loss": 1.4732, "step": 27006 }, { "epoch": 2.1165360501567396, "grad_norm": 2.9711251258850098, "learning_rate": 4.557463539190547e-06, "loss": 1.2699, "step": 27007 }, { "epoch": 2.116614420062696, "grad_norm": 2.6742544174194336, "learning_rate": 4.5571563881316695e-06, "loss": 1.3401, "step": 27008 }, { "epoch": 2.116692789968652, "grad_norm": 2.9348931312561035, "learning_rate": 4.5568492387571435e-06, "loss": 1.4278, "step": 27009 }, { "epoch": 2.116771159874608, "grad_norm": 4.2941389083862305, "learning_rate": 4.5565420910681334e-06, "loss": 1.5501, "step": 27010 }, { "epoch": 2.116849529780564, "grad_norm": 5.434820175170898, "learning_rate": 4.556234945065808e-06, "loss": 1.4991, "step": 27011 }, { "epoch": 2.1169278996865204, "grad_norm": 2.078115224838257, "learning_rate": 4.555927800751339e-06, "loss": 1.2736, "step": 27012 }, { "epoch": 2.1170062695924763, "grad_norm": 3.6483073234558105, "learning_rate": 4.555620658125893e-06, "loss": 1.0264, "step": 27013 }, { "epoch": 2.1170846394984326, "grad_norm": 4.252403736114502, "learning_rate": 4.5553135171906375e-06, "loss": 1.9728, "step": 27014 }, { "epoch": 2.1171630094043885, "grad_norm": 7.310468673706055, "learning_rate": 4.55500637794674e-06, "loss": 0.9342, "step": 27015 }, { "epoch": 2.117241379310345, "grad_norm": 7.077987194061279, "learning_rate": 4.554699240395369e-06, "loss": 1.7394, "step": 27016 }, { "epoch": 2.1173197492163007, "grad_norm": 3.2531514167785645, "learning_rate": 4.5543921045376975e-06, "loss": 1.2911, "step": 27017 }, { "epoch": 2.117398119122257, "grad_norm": 4.27161169052124, "learning_rate": 4.5540849703748846e-06, "loss": 1.4042, "step": 27018 }, { "epoch": 2.1174764890282134, "grad_norm": 4.0325093269348145, "learning_rate": 4.553777837908105e-06, "loss": 1.4871, "step": 27019 }, { "epoch": 2.1175548589341693, "grad_norm": 2.1674630641937256, "learning_rate": 4.5534707071385266e-06, "loss": 1.2856, "step": 27020 }, { "epoch": 2.117633228840125, "grad_norm": 2.639763355255127, "learning_rate": 4.553163578067316e-06, "loss": 1.2191, "step": 27021 }, { "epoch": 2.1177115987460815, "grad_norm": 2.6228435039520264, "learning_rate": 4.552856450695641e-06, "loss": 1.2204, "step": 27022 }, { "epoch": 2.117789968652038, "grad_norm": 3.5564424991607666, "learning_rate": 4.552549325024669e-06, "loss": 1.4585, "step": 27023 }, { "epoch": 2.1178683385579937, "grad_norm": 3.9589812755584717, "learning_rate": 4.552242201055573e-06, "loss": 1.193, "step": 27024 }, { "epoch": 2.11794670846395, "grad_norm": 6.3431901931762695, "learning_rate": 4.551935078789515e-06, "loss": 1.1584, "step": 27025 }, { "epoch": 2.118025078369906, "grad_norm": 2.312467336654663, "learning_rate": 4.551627958227666e-06, "loss": 1.3454, "step": 27026 }, { "epoch": 2.1181034482758623, "grad_norm": 5.059167385101318, "learning_rate": 4.551320839371195e-06, "loss": 1.7084, "step": 27027 }, { "epoch": 2.118181818181818, "grad_norm": 4.195366382598877, "learning_rate": 4.551013722221269e-06, "loss": 1.4535, "step": 27028 }, { "epoch": 2.1182601880877745, "grad_norm": 6.078300952911377, "learning_rate": 4.550706606779056e-06, "loss": 0.9343, "step": 27029 }, { "epoch": 2.1183385579937304, "grad_norm": 4.835808753967285, "learning_rate": 4.550399493045723e-06, "loss": 1.5201, "step": 27030 }, { "epoch": 2.1184169278996867, "grad_norm": 3.890596628189087, "learning_rate": 4.5500923810224405e-06, "loss": 1.3959, "step": 27031 }, { "epoch": 2.1184952978056426, "grad_norm": 4.024050712585449, "learning_rate": 4.549785270710376e-06, "loss": 1.0776, "step": 27032 }, { "epoch": 2.118573667711599, "grad_norm": 10.489351272583008, "learning_rate": 4.549478162110697e-06, "loss": 1.6738, "step": 27033 }, { "epoch": 2.118652037617555, "grad_norm": 2.7579848766326904, "learning_rate": 4.549171055224571e-06, "loss": 1.5967, "step": 27034 }, { "epoch": 2.118730407523511, "grad_norm": 3.7507526874542236, "learning_rate": 4.548863950053167e-06, "loss": 1.5092, "step": 27035 }, { "epoch": 2.118808777429467, "grad_norm": 3.990427255630493, "learning_rate": 4.548556846597653e-06, "loss": 1.0742, "step": 27036 }, { "epoch": 2.1188871473354234, "grad_norm": 4.434090614318848, "learning_rate": 4.5482497448591954e-06, "loss": 1.2654, "step": 27037 }, { "epoch": 2.1189655172413793, "grad_norm": 2.429030656814575, "learning_rate": 4.547942644838965e-06, "loss": 1.1356, "step": 27038 }, { "epoch": 2.1190438871473356, "grad_norm": 10.952280044555664, "learning_rate": 4.54763554653813e-06, "loss": 0.8923, "step": 27039 }, { "epoch": 2.1191222570532915, "grad_norm": 8.103649139404297, "learning_rate": 4.547328449957855e-06, "loss": 1.4311, "step": 27040 }, { "epoch": 2.119200626959248, "grad_norm": 17.922388076782227, "learning_rate": 4.547021355099311e-06, "loss": 1.1921, "step": 27041 }, { "epoch": 2.1192789968652037, "grad_norm": 2.9043476581573486, "learning_rate": 4.546714261963663e-06, "loss": 0.7126, "step": 27042 }, { "epoch": 2.11935736677116, "grad_norm": 7.3438944816589355, "learning_rate": 4.546407170552084e-06, "loss": 1.4838, "step": 27043 }, { "epoch": 2.119435736677116, "grad_norm": 4.221888065338135, "learning_rate": 4.546100080865738e-06, "loss": 1.4924, "step": 27044 }, { "epoch": 2.1195141065830723, "grad_norm": 3.6665470600128174, "learning_rate": 4.545792992905794e-06, "loss": 1.3935, "step": 27045 }, { "epoch": 2.119592476489028, "grad_norm": 7.81145715713501, "learning_rate": 4.54548590667342e-06, "loss": 1.1414, "step": 27046 }, { "epoch": 2.1196708463949845, "grad_norm": 6.220129489898682, "learning_rate": 4.545178822169786e-06, "loss": 1.4711, "step": 27047 }, { "epoch": 2.1197492163009404, "grad_norm": 5.563567161560059, "learning_rate": 4.544871739396055e-06, "loss": 1.8692, "step": 27048 }, { "epoch": 2.1198275862068967, "grad_norm": 3.178863525390625, "learning_rate": 4.5445646583534e-06, "loss": 1.0165, "step": 27049 }, { "epoch": 2.1199059561128526, "grad_norm": 3.2976536750793457, "learning_rate": 4.544257579042988e-06, "loss": 1.444, "step": 27050 }, { "epoch": 2.119984326018809, "grad_norm": 2.696153402328491, "learning_rate": 4.543950501465985e-06, "loss": 1.4003, "step": 27051 }, { "epoch": 2.120062695924765, "grad_norm": 5.31800651550293, "learning_rate": 4.54364342562356e-06, "loss": 1.4819, "step": 27052 }, { "epoch": 2.120141065830721, "grad_norm": 3.270967483520508, "learning_rate": 4.54333635151688e-06, "loss": 1.2211, "step": 27053 }, { "epoch": 2.120219435736677, "grad_norm": 2.3325746059417725, "learning_rate": 4.5430292791471165e-06, "loss": 1.3944, "step": 27054 }, { "epoch": 2.1202978056426334, "grad_norm": 8.392824172973633, "learning_rate": 4.542722208515432e-06, "loss": 1.305, "step": 27055 }, { "epoch": 2.1203761755485893, "grad_norm": 3.365147352218628, "learning_rate": 4.5424151396229985e-06, "loss": 1.3056, "step": 27056 }, { "epoch": 2.1204545454545456, "grad_norm": 3.8888754844665527, "learning_rate": 4.542108072470983e-06, "loss": 1.3151, "step": 27057 }, { "epoch": 2.1205329153605015, "grad_norm": 3.1338846683502197, "learning_rate": 4.541801007060555e-06, "loss": 1.3849, "step": 27058 }, { "epoch": 2.120611285266458, "grad_norm": 3.3134355545043945, "learning_rate": 4.541493943392878e-06, "loss": 1.5486, "step": 27059 }, { "epoch": 2.1206896551724137, "grad_norm": 4.762340068817139, "learning_rate": 4.541186881469122e-06, "loss": 1.261, "step": 27060 }, { "epoch": 2.12076802507837, "grad_norm": 2.776372194290161, "learning_rate": 4.540879821290457e-06, "loss": 1.3841, "step": 27061 }, { "epoch": 2.120846394984326, "grad_norm": 5.175329208374023, "learning_rate": 4.54057276285805e-06, "loss": 0.9792, "step": 27062 }, { "epoch": 2.1209247648902823, "grad_norm": 3.686796188354492, "learning_rate": 4.540265706173067e-06, "loss": 1.3302, "step": 27063 }, { "epoch": 2.121003134796238, "grad_norm": 4.935762882232666, "learning_rate": 4.539958651236678e-06, "loss": 1.2403, "step": 27064 }, { "epoch": 2.1210815047021945, "grad_norm": 3.2653579711914062, "learning_rate": 4.53965159805005e-06, "loss": 1.539, "step": 27065 }, { "epoch": 2.1211598746081504, "grad_norm": 3.038104295730591, "learning_rate": 4.539344546614351e-06, "loss": 1.1605, "step": 27066 }, { "epoch": 2.1212382445141067, "grad_norm": 2.8278915882110596, "learning_rate": 4.539037496930747e-06, "loss": 1.3258, "step": 27067 }, { "epoch": 2.1213166144200626, "grad_norm": 2.9210524559020996, "learning_rate": 4.538730449000408e-06, "loss": 1.1971, "step": 27068 }, { "epoch": 2.121394984326019, "grad_norm": 2.6922500133514404, "learning_rate": 4.538423402824503e-06, "loss": 1.3275, "step": 27069 }, { "epoch": 2.121473354231975, "grad_norm": 3.7120344638824463, "learning_rate": 4.538116358404197e-06, "loss": 1.1572, "step": 27070 }, { "epoch": 2.121551724137931, "grad_norm": 2.501119613647461, "learning_rate": 4.53780931574066e-06, "loss": 1.1596, "step": 27071 }, { "epoch": 2.121630094043887, "grad_norm": 2.624913454055786, "learning_rate": 4.537502274835057e-06, "loss": 1.4601, "step": 27072 }, { "epoch": 2.1217084639498434, "grad_norm": 3.006443738937378, "learning_rate": 4.537195235688562e-06, "loss": 1.5059, "step": 27073 }, { "epoch": 2.1217868338557992, "grad_norm": 2.8553943634033203, "learning_rate": 4.536888198302335e-06, "loss": 1.5695, "step": 27074 }, { "epoch": 2.1218652037617556, "grad_norm": 2.9265055656433105, "learning_rate": 4.536581162677548e-06, "loss": 1.4832, "step": 27075 }, { "epoch": 2.1219435736677115, "grad_norm": 2.806490182876587, "learning_rate": 4.5362741288153686e-06, "loss": 0.9321, "step": 27076 }, { "epoch": 2.122021943573668, "grad_norm": 2.745540142059326, "learning_rate": 4.535967096716965e-06, "loss": 1.4537, "step": 27077 }, { "epoch": 2.1221003134796237, "grad_norm": 2.4063010215759277, "learning_rate": 4.535660066383504e-06, "loss": 1.362, "step": 27078 }, { "epoch": 2.12217868338558, "grad_norm": 2.0794801712036133, "learning_rate": 4.535353037816151e-06, "loss": 1.1817, "step": 27079 }, { "epoch": 2.122257053291536, "grad_norm": 2.9218907356262207, "learning_rate": 4.535046011016081e-06, "loss": 1.1858, "step": 27080 }, { "epoch": 2.1223354231974922, "grad_norm": 2.7515017986297607, "learning_rate": 4.534738985984454e-06, "loss": 1.2318, "step": 27081 }, { "epoch": 2.122413793103448, "grad_norm": 2.7426488399505615, "learning_rate": 4.534431962722441e-06, "loss": 1.2261, "step": 27082 }, { "epoch": 2.1224921630094045, "grad_norm": 2.237903594970703, "learning_rate": 4.53412494123121e-06, "loss": 1.377, "step": 27083 }, { "epoch": 2.1225705329153604, "grad_norm": 2.599752187728882, "learning_rate": 4.5338179215119296e-06, "loss": 1.5471, "step": 27084 }, { "epoch": 2.1226489028213167, "grad_norm": 4.097829818725586, "learning_rate": 4.533510903565766e-06, "loss": 1.6669, "step": 27085 }, { "epoch": 2.1227272727272726, "grad_norm": 4.585257053375244, "learning_rate": 4.533203887393885e-06, "loss": 1.2443, "step": 27086 }, { "epoch": 2.122805642633229, "grad_norm": 3.28432035446167, "learning_rate": 4.532896872997458e-06, "loss": 1.2507, "step": 27087 }, { "epoch": 2.122884012539185, "grad_norm": 2.9174611568450928, "learning_rate": 4.532589860377653e-06, "loss": 1.4186, "step": 27088 }, { "epoch": 2.122962382445141, "grad_norm": 3.256467580795288, "learning_rate": 4.5322828495356346e-06, "loss": 1.0394, "step": 27089 }, { "epoch": 2.123040752351097, "grad_norm": 3.1207594871520996, "learning_rate": 4.531975840472572e-06, "loss": 1.4025, "step": 27090 }, { "epoch": 2.1231191222570533, "grad_norm": 2.636077404022217, "learning_rate": 4.531668833189635e-06, "loss": 1.301, "step": 27091 }, { "epoch": 2.1231974921630092, "grad_norm": 7.63735818862915, "learning_rate": 4.5313618276879865e-06, "loss": 1.8208, "step": 27092 }, { "epoch": 2.1232758620689656, "grad_norm": 2.4174540042877197, "learning_rate": 4.531054823968797e-06, "loss": 1.5669, "step": 27093 }, { "epoch": 2.1233542319749215, "grad_norm": 3.08126163482666, "learning_rate": 4.530747822033235e-06, "loss": 1.3508, "step": 27094 }, { "epoch": 2.123432601880878, "grad_norm": 2.3328208923339844, "learning_rate": 4.530440821882468e-06, "loss": 1.164, "step": 27095 }, { "epoch": 2.1235109717868337, "grad_norm": 3.132391929626465, "learning_rate": 4.530133823517662e-06, "loss": 0.8482, "step": 27096 }, { "epoch": 2.12358934169279, "grad_norm": 2.953411817550659, "learning_rate": 4.529826826939986e-06, "loss": 1.1189, "step": 27097 }, { "epoch": 2.123667711598746, "grad_norm": 3.944861888885498, "learning_rate": 4.529519832150605e-06, "loss": 1.1203, "step": 27098 }, { "epoch": 2.1237460815047022, "grad_norm": 2.7338194847106934, "learning_rate": 4.529212839150693e-06, "loss": 1.097, "step": 27099 }, { "epoch": 2.123824451410658, "grad_norm": 4.940197944641113, "learning_rate": 4.528905847941411e-06, "loss": 1.5988, "step": 27100 }, { "epoch": 2.1239028213166145, "grad_norm": 3.1571478843688965, "learning_rate": 4.52859885852393e-06, "loss": 1.4834, "step": 27101 }, { "epoch": 2.1239811912225703, "grad_norm": 7.338258266448975, "learning_rate": 4.528291870899415e-06, "loss": 1.1755, "step": 27102 }, { "epoch": 2.1240595611285267, "grad_norm": 2.866909980773926, "learning_rate": 4.527984885069038e-06, "loss": 1.4032, "step": 27103 }, { "epoch": 2.1241379310344826, "grad_norm": 11.78347396850586, "learning_rate": 4.527677901033963e-06, "loss": 1.3984, "step": 27104 }, { "epoch": 2.124216300940439, "grad_norm": 7.424300670623779, "learning_rate": 4.527370918795358e-06, "loss": 1.5993, "step": 27105 }, { "epoch": 2.124294670846395, "grad_norm": 3.210461378097534, "learning_rate": 4.527063938354393e-06, "loss": 1.461, "step": 27106 }, { "epoch": 2.124373040752351, "grad_norm": 2.8925297260284424, "learning_rate": 4.526756959712232e-06, "loss": 1.6189, "step": 27107 }, { "epoch": 2.124451410658307, "grad_norm": 4.063520908355713, "learning_rate": 4.526449982870045e-06, "loss": 1.3654, "step": 27108 }, { "epoch": 2.1245297805642633, "grad_norm": 3.3801181316375732, "learning_rate": 4.526143007828998e-06, "loss": 1.4256, "step": 27109 }, { "epoch": 2.1246081504702192, "grad_norm": 2.846543788909912, "learning_rate": 4.5258360345902625e-06, "loss": 1.3904, "step": 27110 }, { "epoch": 2.1246865203761756, "grad_norm": 2.8652234077453613, "learning_rate": 4.525529063155001e-06, "loss": 1.2726, "step": 27111 }, { "epoch": 2.1247648902821314, "grad_norm": 4.9502339363098145, "learning_rate": 4.525222093524383e-06, "loss": 1.7122, "step": 27112 }, { "epoch": 2.124843260188088, "grad_norm": 2.765251636505127, "learning_rate": 4.524915125699577e-06, "loss": 1.3825, "step": 27113 }, { "epoch": 2.124921630094044, "grad_norm": 3.694490671157837, "learning_rate": 4.52460815968175e-06, "loss": 1.3094, "step": 27114 }, { "epoch": 2.125, "grad_norm": 6.1616530418396, "learning_rate": 4.52430119547207e-06, "loss": 1.3231, "step": 27115 }, { "epoch": 2.125078369905956, "grad_norm": 4.20853853225708, "learning_rate": 4.523994233071703e-06, "loss": 1.1721, "step": 27116 }, { "epoch": 2.125156739811912, "grad_norm": 3.348724842071533, "learning_rate": 4.523687272481816e-06, "loss": 1.1968, "step": 27117 }, { "epoch": 2.1252351097178686, "grad_norm": 11.874808311462402, "learning_rate": 4.523380313703582e-06, "loss": 0.9517, "step": 27118 }, { "epoch": 2.1253134796238244, "grad_norm": 2.498319625854492, "learning_rate": 4.523073356738161e-06, "loss": 1.5986, "step": 27119 }, { "epoch": 2.1253918495297803, "grad_norm": 4.035677433013916, "learning_rate": 4.522766401586724e-06, "loss": 1.3857, "step": 27120 }, { "epoch": 2.1254702194357367, "grad_norm": 3.2616262435913086, "learning_rate": 4.522459448250441e-06, "loss": 1.5391, "step": 27121 }, { "epoch": 2.125548589341693, "grad_norm": 1.8418680429458618, "learning_rate": 4.522152496730475e-06, "loss": 1.5475, "step": 27122 }, { "epoch": 2.125626959247649, "grad_norm": 6.498587608337402, "learning_rate": 4.521845547027997e-06, "loss": 1.6691, "step": 27123 }, { "epoch": 2.125705329153605, "grad_norm": 3.7661502361297607, "learning_rate": 4.52153859914417e-06, "loss": 0.9822, "step": 27124 }, { "epoch": 2.125783699059561, "grad_norm": 2.2968804836273193, "learning_rate": 4.521231653080169e-06, "loss": 1.1508, "step": 27125 }, { "epoch": 2.1258620689655174, "grad_norm": 3.113579034805298, "learning_rate": 4.520924708837152e-06, "loss": 1.5267, "step": 27126 }, { "epoch": 2.1259404388714733, "grad_norm": 3.7947282791137695, "learning_rate": 4.520617766416294e-06, "loss": 0.9999, "step": 27127 }, { "epoch": 2.1260188087774297, "grad_norm": 4.061516761779785, "learning_rate": 4.5203108258187585e-06, "loss": 1.4603, "step": 27128 }, { "epoch": 2.1260971786833855, "grad_norm": 2.670846462249756, "learning_rate": 4.520003887045716e-06, "loss": 1.3716, "step": 27129 }, { "epoch": 2.126175548589342, "grad_norm": 2.827028512954712, "learning_rate": 4.5196969500983315e-06, "loss": 1.3198, "step": 27130 }, { "epoch": 2.1262539184952978, "grad_norm": 2.4144318103790283, "learning_rate": 4.519390014977772e-06, "loss": 1.0331, "step": 27131 }, { "epoch": 2.126332288401254, "grad_norm": 3.5179598331451416, "learning_rate": 4.519083081685207e-06, "loss": 1.5767, "step": 27132 }, { "epoch": 2.12641065830721, "grad_norm": 4.469582557678223, "learning_rate": 4.518776150221804e-06, "loss": 1.2021, "step": 27133 }, { "epoch": 2.1264890282131663, "grad_norm": 2.6133439540863037, "learning_rate": 4.5184692205887275e-06, "loss": 1.4324, "step": 27134 }, { "epoch": 2.126567398119122, "grad_norm": 2.4477498531341553, "learning_rate": 4.518162292787147e-06, "loss": 1.0525, "step": 27135 }, { "epoch": 2.1266457680250785, "grad_norm": 2.8309624195098877, "learning_rate": 4.517855366818231e-06, "loss": 1.672, "step": 27136 }, { "epoch": 2.1267241379310344, "grad_norm": 8.988116264343262, "learning_rate": 4.517548442683145e-06, "loss": 0.9323, "step": 27137 }, { "epoch": 2.1268025078369908, "grad_norm": 4.580802917480469, "learning_rate": 4.517241520383055e-06, "loss": 1.4059, "step": 27138 }, { "epoch": 2.1268808777429467, "grad_norm": 2.7231225967407227, "learning_rate": 4.516934599919131e-06, "loss": 1.7618, "step": 27139 }, { "epoch": 2.126959247648903, "grad_norm": 3.71643328666687, "learning_rate": 4.516627681292542e-06, "loss": 1.7971, "step": 27140 }, { "epoch": 2.127037617554859, "grad_norm": 3.2350401878356934, "learning_rate": 4.516320764504452e-06, "loss": 1.3472, "step": 27141 }, { "epoch": 2.127115987460815, "grad_norm": 11.548203468322754, "learning_rate": 4.516013849556028e-06, "loss": 1.8024, "step": 27142 }, { "epoch": 2.127194357366771, "grad_norm": 3.1648402214050293, "learning_rate": 4.515706936448439e-06, "loss": 1.3074, "step": 27143 }, { "epoch": 2.1272727272727274, "grad_norm": 7.673740386962891, "learning_rate": 4.515400025182854e-06, "loss": 1.3021, "step": 27144 }, { "epoch": 2.1273510971786833, "grad_norm": 4.114084720611572, "learning_rate": 4.515093115760436e-06, "loss": 1.2599, "step": 27145 }, { "epoch": 2.1274294670846396, "grad_norm": 2.3759162425994873, "learning_rate": 4.514786208182356e-06, "loss": 1.4145, "step": 27146 }, { "epoch": 2.1275078369905955, "grad_norm": 3.0263426303863525, "learning_rate": 4.514479302449781e-06, "loss": 1.3709, "step": 27147 }, { "epoch": 2.127586206896552, "grad_norm": 4.229574680328369, "learning_rate": 4.5141723985638766e-06, "loss": 1.1643, "step": 27148 }, { "epoch": 2.1276645768025078, "grad_norm": 3.626755952835083, "learning_rate": 4.5138654965258096e-06, "loss": 1.6321, "step": 27149 }, { "epoch": 2.127742946708464, "grad_norm": 4.526039123535156, "learning_rate": 4.5135585963367486e-06, "loss": 1.8866, "step": 27150 }, { "epoch": 2.12782131661442, "grad_norm": 2.665837287902832, "learning_rate": 4.5132516979978625e-06, "loss": 1.1691, "step": 27151 }, { "epoch": 2.1278996865203763, "grad_norm": 3.5749199390411377, "learning_rate": 4.512944801510316e-06, "loss": 1.4418, "step": 27152 }, { "epoch": 2.127978056426332, "grad_norm": 4.126018047332764, "learning_rate": 4.512637906875277e-06, "loss": 1.7369, "step": 27153 }, { "epoch": 2.1280564263322885, "grad_norm": 8.55878734588623, "learning_rate": 4.512331014093914e-06, "loss": 0.8909, "step": 27154 }, { "epoch": 2.1281347962382444, "grad_norm": 2.188868522644043, "learning_rate": 4.512024123167394e-06, "loss": 1.3224, "step": 27155 }, { "epoch": 2.1282131661442008, "grad_norm": 9.632761001586914, "learning_rate": 4.511717234096881e-06, "loss": 1.2982, "step": 27156 }, { "epoch": 2.1282915360501566, "grad_norm": 3.5494682788848877, "learning_rate": 4.511410346883546e-06, "loss": 1.5674, "step": 27157 }, { "epoch": 2.128369905956113, "grad_norm": 2.603083372116089, "learning_rate": 4.511103461528555e-06, "loss": 1.3101, "step": 27158 }, { "epoch": 2.128448275862069, "grad_norm": 1.9804935455322266, "learning_rate": 4.510796578033077e-06, "loss": 1.4343, "step": 27159 }, { "epoch": 2.128526645768025, "grad_norm": 6.391092300415039, "learning_rate": 4.510489696398276e-06, "loss": 1.1973, "step": 27160 }, { "epoch": 2.128605015673981, "grad_norm": 2.33520770072937, "learning_rate": 4.5101828166253195e-06, "loss": 1.1764, "step": 27161 }, { "epoch": 2.1286833855799374, "grad_norm": 12.662055969238281, "learning_rate": 4.50987593871538e-06, "loss": 1.3537, "step": 27162 }, { "epoch": 2.1287617554858933, "grad_norm": 3.788010358810425, "learning_rate": 4.509569062669617e-06, "loss": 0.9551, "step": 27163 }, { "epoch": 2.1288401253918496, "grad_norm": 3.4393632411956787, "learning_rate": 4.509262188489202e-06, "loss": 1.5273, "step": 27164 }, { "epoch": 2.1289184952978055, "grad_norm": 5.030027866363525, "learning_rate": 4.5089553161753016e-06, "loss": 1.2508, "step": 27165 }, { "epoch": 2.128996865203762, "grad_norm": 3.2179388999938965, "learning_rate": 4.508648445729085e-06, "loss": 1.6026, "step": 27166 }, { "epoch": 2.1290752351097177, "grad_norm": 3.664806842803955, "learning_rate": 4.5083415771517155e-06, "loss": 1.3629, "step": 27167 }, { "epoch": 2.129153605015674, "grad_norm": 2.670478343963623, "learning_rate": 4.508034710444361e-06, "loss": 1.1783, "step": 27168 }, { "epoch": 2.12923197492163, "grad_norm": 3.247614622116089, "learning_rate": 4.507727845608191e-06, "loss": 1.5558, "step": 27169 }, { "epoch": 2.1293103448275863, "grad_norm": 2.8569769859313965, "learning_rate": 4.507420982644373e-06, "loss": 1.5161, "step": 27170 }, { "epoch": 2.129388714733542, "grad_norm": 2.347430467605591, "learning_rate": 4.507114121554071e-06, "loss": 1.2973, "step": 27171 }, { "epoch": 2.1294670846394985, "grad_norm": 3.094515085220337, "learning_rate": 4.506807262338453e-06, "loss": 1.6553, "step": 27172 }, { "epoch": 2.1295454545454544, "grad_norm": 7.635763168334961, "learning_rate": 4.506500404998687e-06, "loss": 1.6517, "step": 27173 }, { "epoch": 2.1296238244514107, "grad_norm": 4.047317028045654, "learning_rate": 4.506193549535943e-06, "loss": 1.5936, "step": 27174 }, { "epoch": 2.1297021943573666, "grad_norm": 2.8248026371002197, "learning_rate": 4.5058866959513815e-06, "loss": 1.5345, "step": 27175 }, { "epoch": 2.129780564263323, "grad_norm": 5.0929436683654785, "learning_rate": 4.505579844246175e-06, "loss": 1.8703, "step": 27176 }, { "epoch": 2.129858934169279, "grad_norm": 2.179672956466675, "learning_rate": 4.505272994421489e-06, "loss": 1.118, "step": 27177 }, { "epoch": 2.129937304075235, "grad_norm": 5.463028430938721, "learning_rate": 4.50496614647849e-06, "loss": 1.3473, "step": 27178 }, { "epoch": 2.130015673981191, "grad_norm": 3.763241767883301, "learning_rate": 4.504659300418345e-06, "loss": 1.6344, "step": 27179 }, { "epoch": 2.1300940438871474, "grad_norm": 4.411290645599365, "learning_rate": 4.50435245624222e-06, "loss": 1.1161, "step": 27180 }, { "epoch": 2.1301724137931033, "grad_norm": 2.4789912700653076, "learning_rate": 4.504045613951288e-06, "loss": 1.6243, "step": 27181 }, { "epoch": 2.1302507836990596, "grad_norm": 3.877152442932129, "learning_rate": 4.503738773546708e-06, "loss": 1.4448, "step": 27182 }, { "epoch": 2.1303291536050155, "grad_norm": 2.97662353515625, "learning_rate": 4.503431935029652e-06, "loss": 1.4042, "step": 27183 }, { "epoch": 2.130407523510972, "grad_norm": 2.2692320346832275, "learning_rate": 4.503125098401286e-06, "loss": 1.189, "step": 27184 }, { "epoch": 2.1304858934169277, "grad_norm": 4.553552150726318, "learning_rate": 4.502818263662778e-06, "loss": 1.4498, "step": 27185 }, { "epoch": 2.130564263322884, "grad_norm": 3.5005080699920654, "learning_rate": 4.502511430815293e-06, "loss": 1.5137, "step": 27186 }, { "epoch": 2.13064263322884, "grad_norm": 2.1837575435638428, "learning_rate": 4.502204599859997e-06, "loss": 1.1141, "step": 27187 }, { "epoch": 2.1307210031347963, "grad_norm": 12.980423927307129, "learning_rate": 4.501897770798061e-06, "loss": 1.7485, "step": 27188 }, { "epoch": 2.130799373040752, "grad_norm": 3.394014358520508, "learning_rate": 4.501590943630651e-06, "loss": 1.0551, "step": 27189 }, { "epoch": 2.1308777429467085, "grad_norm": 2.3444392681121826, "learning_rate": 4.501284118358932e-06, "loss": 1.3602, "step": 27190 }, { "epoch": 2.1309561128526644, "grad_norm": 2.385687828063965, "learning_rate": 4.500977294984072e-06, "loss": 1.0207, "step": 27191 }, { "epoch": 2.1310344827586207, "grad_norm": 3.099818229675293, "learning_rate": 4.500670473507239e-06, "loss": 1.2114, "step": 27192 }, { "epoch": 2.1311128526645766, "grad_norm": 2.086948871612549, "learning_rate": 4.500363653929598e-06, "loss": 1.3166, "step": 27193 }, { "epoch": 2.131191222570533, "grad_norm": 3.6395511627197266, "learning_rate": 4.500056836252316e-06, "loss": 1.6147, "step": 27194 }, { "epoch": 2.131269592476489, "grad_norm": 3.9188098907470703, "learning_rate": 4.499750020476562e-06, "loss": 1.4324, "step": 27195 }, { "epoch": 2.131347962382445, "grad_norm": 9.672780990600586, "learning_rate": 4.499443206603503e-06, "loss": 0.8829, "step": 27196 }, { "epoch": 2.131426332288401, "grad_norm": 2.937774896621704, "learning_rate": 4.499136394634304e-06, "loss": 1.5834, "step": 27197 }, { "epoch": 2.1315047021943574, "grad_norm": 8.919018745422363, "learning_rate": 4.4988295845701326e-06, "loss": 1.2698, "step": 27198 }, { "epoch": 2.1315830721003133, "grad_norm": 2.291177749633789, "learning_rate": 4.498522776412155e-06, "loss": 1.569, "step": 27199 }, { "epoch": 2.1316614420062696, "grad_norm": 2.864064931869507, "learning_rate": 4.498215970161543e-06, "loss": 1.2896, "step": 27200 }, { "epoch": 2.1317398119122255, "grad_norm": 4.135727882385254, "learning_rate": 4.497909165819456e-06, "loss": 1.7295, "step": 27201 }, { "epoch": 2.131818181818182, "grad_norm": 2.6266913414001465, "learning_rate": 4.497602363387066e-06, "loss": 1.3374, "step": 27202 }, { "epoch": 2.1318965517241377, "grad_norm": 2.894693613052368, "learning_rate": 4.49729556286554e-06, "loss": 1.0118, "step": 27203 }, { "epoch": 2.131974921630094, "grad_norm": 3.2012312412261963, "learning_rate": 4.496988764256042e-06, "loss": 1.367, "step": 27204 }, { "epoch": 2.1320532915360504, "grad_norm": 2.6205520629882812, "learning_rate": 4.49668196755974e-06, "loss": 1.3311, "step": 27205 }, { "epoch": 2.1321316614420063, "grad_norm": 2.797369956970215, "learning_rate": 4.496375172777801e-06, "loss": 1.1676, "step": 27206 }, { "epoch": 2.132210031347962, "grad_norm": 3.25935697555542, "learning_rate": 4.4960683799113945e-06, "loss": 1.1418, "step": 27207 }, { "epoch": 2.1322884012539185, "grad_norm": 3.0848379135131836, "learning_rate": 4.495761588961682e-06, "loss": 1.3906, "step": 27208 }, { "epoch": 2.132366771159875, "grad_norm": 1.8400206565856934, "learning_rate": 4.495454799929835e-06, "loss": 1.2136, "step": 27209 }, { "epoch": 2.1324451410658307, "grad_norm": 4.557611465454102, "learning_rate": 4.495148012817018e-06, "loss": 1.0401, "step": 27210 }, { "epoch": 2.1325235109717866, "grad_norm": 3.7146427631378174, "learning_rate": 4.4948412276244e-06, "loss": 1.2288, "step": 27211 }, { "epoch": 2.132601880877743, "grad_norm": 2.1511056423187256, "learning_rate": 4.494534444353146e-06, "loss": 1.4906, "step": 27212 }, { "epoch": 2.1326802507836993, "grad_norm": 3.5088605880737305, "learning_rate": 4.494227663004422e-06, "loss": 1.4049, "step": 27213 }, { "epoch": 2.132758620689655, "grad_norm": 2.815732002258301, "learning_rate": 4.493920883579397e-06, "loss": 1.3799, "step": 27214 }, { "epoch": 2.1328369905956115, "grad_norm": 3.244919776916504, "learning_rate": 4.493614106079238e-06, "loss": 1.5598, "step": 27215 }, { "epoch": 2.1329153605015674, "grad_norm": 16.86461639404297, "learning_rate": 4.49330733050511e-06, "loss": 1.6541, "step": 27216 }, { "epoch": 2.1329937304075237, "grad_norm": 2.759739875793457, "learning_rate": 4.4930005568581815e-06, "loss": 1.3292, "step": 27217 }, { "epoch": 2.1330721003134796, "grad_norm": 2.795428991317749, "learning_rate": 4.492693785139618e-06, "loss": 1.2136, "step": 27218 }, { "epoch": 2.133150470219436, "grad_norm": 4.3114190101623535, "learning_rate": 4.4923870153505865e-06, "loss": 1.0426, "step": 27219 }, { "epoch": 2.133228840125392, "grad_norm": 2.4442553520202637, "learning_rate": 4.492080247492253e-06, "loss": 1.1973, "step": 27220 }, { "epoch": 2.133307210031348, "grad_norm": 3.558168649673462, "learning_rate": 4.4917734815657864e-06, "loss": 0.8685, "step": 27221 }, { "epoch": 2.133385579937304, "grad_norm": 4.283606052398682, "learning_rate": 4.491466717572353e-06, "loss": 1.4656, "step": 27222 }, { "epoch": 2.1334639498432604, "grad_norm": 4.333035945892334, "learning_rate": 4.491159955513119e-06, "loss": 1.4683, "step": 27223 }, { "epoch": 2.1335423197492163, "grad_norm": 1.8532401323318481, "learning_rate": 4.490853195389251e-06, "loss": 1.0502, "step": 27224 }, { "epoch": 2.1336206896551726, "grad_norm": 3.035320997238159, "learning_rate": 4.490546437201914e-06, "loss": 1.4962, "step": 27225 }, { "epoch": 2.1336990595611285, "grad_norm": 2.864170789718628, "learning_rate": 4.49023968095228e-06, "loss": 1.3137, "step": 27226 }, { "epoch": 2.133777429467085, "grad_norm": 2.207846164703369, "learning_rate": 4.48993292664151e-06, "loss": 1.5632, "step": 27227 }, { "epoch": 2.1338557993730407, "grad_norm": 4.3312296867370605, "learning_rate": 4.489626174270774e-06, "loss": 1.2906, "step": 27228 }, { "epoch": 2.133934169278997, "grad_norm": 2.497291088104248, "learning_rate": 4.489319423841238e-06, "loss": 0.9774, "step": 27229 }, { "epoch": 2.134012539184953, "grad_norm": 2.352566719055176, "learning_rate": 4.489012675354068e-06, "loss": 1.1985, "step": 27230 }, { "epoch": 2.1340909090909093, "grad_norm": 2.061772584915161, "learning_rate": 4.488705928810432e-06, "loss": 1.1923, "step": 27231 }, { "epoch": 2.134169278996865, "grad_norm": 3.2014923095703125, "learning_rate": 4.4883991842114945e-06, "loss": 1.3353, "step": 27232 }, { "epoch": 2.1342476489028215, "grad_norm": 6.963027477264404, "learning_rate": 4.488092441558428e-06, "loss": 1.2796, "step": 27233 }, { "epoch": 2.1343260188087774, "grad_norm": 3.684424877166748, "learning_rate": 4.48778570085239e-06, "loss": 1.4233, "step": 27234 }, { "epoch": 2.1344043887147337, "grad_norm": 3.53521728515625, "learning_rate": 4.487478962094554e-06, "loss": 1.6387, "step": 27235 }, { "epoch": 2.1344827586206896, "grad_norm": 3.343961715698242, "learning_rate": 4.487172225286084e-06, "loss": 1.0226, "step": 27236 }, { "epoch": 2.134561128526646, "grad_norm": 3.326676845550537, "learning_rate": 4.48686549042815e-06, "loss": 1.3093, "step": 27237 }, { "epoch": 2.134639498432602, "grad_norm": 2.225698947906494, "learning_rate": 4.4865587575219135e-06, "loss": 1.1491, "step": 27238 }, { "epoch": 2.134717868338558, "grad_norm": 4.160295486450195, "learning_rate": 4.486252026568544e-06, "loss": 1.3006, "step": 27239 }, { "epoch": 2.134796238244514, "grad_norm": 4.730323791503906, "learning_rate": 4.485945297569208e-06, "loss": 1.2535, "step": 27240 }, { "epoch": 2.1348746081504704, "grad_norm": 3.7169368267059326, "learning_rate": 4.485638570525074e-06, "loss": 1.0862, "step": 27241 }, { "epoch": 2.1349529780564263, "grad_norm": 6.3683600425720215, "learning_rate": 4.485331845437305e-06, "loss": 1.3399, "step": 27242 }, { "epoch": 2.1350313479623826, "grad_norm": 3.3189239501953125, "learning_rate": 4.4850251223070695e-06, "loss": 1.4295, "step": 27243 }, { "epoch": 2.1351097178683385, "grad_norm": 3.9308393001556396, "learning_rate": 4.484718401135535e-06, "loss": 1.5677, "step": 27244 }, { "epoch": 2.135188087774295, "grad_norm": 7.523262023925781, "learning_rate": 4.484411681923865e-06, "loss": 1.1622, "step": 27245 }, { "epoch": 2.1352664576802507, "grad_norm": 2.6822896003723145, "learning_rate": 4.484104964673228e-06, "loss": 1.244, "step": 27246 }, { "epoch": 2.135344827586207, "grad_norm": 7.967254638671875, "learning_rate": 4.483798249384793e-06, "loss": 0.9824, "step": 27247 }, { "epoch": 2.135423197492163, "grad_norm": 4.399447917938232, "learning_rate": 4.483491536059724e-06, "loss": 1.1135, "step": 27248 }, { "epoch": 2.1355015673981192, "grad_norm": 2.9261868000030518, "learning_rate": 4.4831848246991864e-06, "loss": 1.2002, "step": 27249 }, { "epoch": 2.135579937304075, "grad_norm": 4.72969388961792, "learning_rate": 4.482878115304349e-06, "loss": 1.2689, "step": 27250 }, { "epoch": 2.1356583072100315, "grad_norm": 2.5479543209075928, "learning_rate": 4.482571407876377e-06, "loss": 1.1515, "step": 27251 }, { "epoch": 2.1357366771159874, "grad_norm": 3.0351712703704834, "learning_rate": 4.482264702416439e-06, "loss": 1.1617, "step": 27252 }, { "epoch": 2.1358150470219437, "grad_norm": 3.3254690170288086, "learning_rate": 4.4819579989256985e-06, "loss": 0.9739, "step": 27253 }, { "epoch": 2.1358934169278996, "grad_norm": 3.128786325454712, "learning_rate": 4.481651297405325e-06, "loss": 1.0876, "step": 27254 }, { "epoch": 2.135971786833856, "grad_norm": 3.2810301780700684, "learning_rate": 4.481344597856483e-06, "loss": 1.4703, "step": 27255 }, { "epoch": 2.136050156739812, "grad_norm": 1.8072826862335205, "learning_rate": 4.4810379002803414e-06, "loss": 1.4384, "step": 27256 }, { "epoch": 2.136128526645768, "grad_norm": 2.0412604808807373, "learning_rate": 4.480731204678062e-06, "loss": 1.4051, "step": 27257 }, { "epoch": 2.136206896551724, "grad_norm": 3.138692617416382, "learning_rate": 4.480424511050816e-06, "loss": 1.3246, "step": 27258 }, { "epoch": 2.1362852664576804, "grad_norm": 1.850390911102295, "learning_rate": 4.48011781939977e-06, "loss": 1.1177, "step": 27259 }, { "epoch": 2.1363636363636362, "grad_norm": 2.4307966232299805, "learning_rate": 4.479811129726087e-06, "loss": 0.9433, "step": 27260 }, { "epoch": 2.1364420062695926, "grad_norm": 2.4655048847198486, "learning_rate": 4.479504442030935e-06, "loss": 1.2887, "step": 27261 }, { "epoch": 2.1365203761755485, "grad_norm": 2.4558286666870117, "learning_rate": 4.47919775631548e-06, "loss": 1.3398, "step": 27262 }, { "epoch": 2.136598746081505, "grad_norm": 2.3945956230163574, "learning_rate": 4.478891072580893e-06, "loss": 1.3567, "step": 27263 }, { "epoch": 2.1366771159874607, "grad_norm": 5.563399791717529, "learning_rate": 4.478584390828332e-06, "loss": 1.0933, "step": 27264 }, { "epoch": 2.136755485893417, "grad_norm": 2.8925859928131104, "learning_rate": 4.4782777110589705e-06, "loss": 1.1817, "step": 27265 }, { "epoch": 2.136833855799373, "grad_norm": 2.676783800125122, "learning_rate": 4.477971033273972e-06, "loss": 1.151, "step": 27266 }, { "epoch": 2.1369122257053292, "grad_norm": 3.784430980682373, "learning_rate": 4.477664357474506e-06, "loss": 1.4188, "step": 27267 }, { "epoch": 2.136990595611285, "grad_norm": 5.495927333831787, "learning_rate": 4.477357683661734e-06, "loss": 1.3822, "step": 27268 }, { "epoch": 2.1370689655172415, "grad_norm": 2.281233787536621, "learning_rate": 4.477051011836824e-06, "loss": 1.3166, "step": 27269 }, { "epoch": 2.1371473354231973, "grad_norm": 3.4979920387268066, "learning_rate": 4.476744342000944e-06, "loss": 1.3348, "step": 27270 }, { "epoch": 2.1372257053291537, "grad_norm": 3.3070502281188965, "learning_rate": 4.476437674155262e-06, "loss": 1.1447, "step": 27271 }, { "epoch": 2.1373040752351096, "grad_norm": 6.59684419631958, "learning_rate": 4.476131008300941e-06, "loss": 0.8695, "step": 27272 }, { "epoch": 2.137382445141066, "grad_norm": 3.444953680038452, "learning_rate": 4.475824344439147e-06, "loss": 1.2622, "step": 27273 }, { "epoch": 2.137460815047022, "grad_norm": 3.9142374992370605, "learning_rate": 4.475517682571051e-06, "loss": 1.2715, "step": 27274 }, { "epoch": 2.137539184952978, "grad_norm": 2.9968671798706055, "learning_rate": 4.475211022697814e-06, "loss": 1.4371, "step": 27275 }, { "epoch": 2.137617554858934, "grad_norm": 2.162158489227295, "learning_rate": 4.474904364820604e-06, "loss": 1.1039, "step": 27276 }, { "epoch": 2.1376959247648903, "grad_norm": 2.386305570602417, "learning_rate": 4.4745977089405895e-06, "loss": 1.2061, "step": 27277 }, { "epoch": 2.1377742946708462, "grad_norm": 3.9794769287109375, "learning_rate": 4.4742910550589365e-06, "loss": 1.3705, "step": 27278 }, { "epoch": 2.1378526645768026, "grad_norm": 3.705523729324341, "learning_rate": 4.473984403176809e-06, "loss": 1.6158, "step": 27279 }, { "epoch": 2.1379310344827585, "grad_norm": 3.0649938583374023, "learning_rate": 4.473677753295375e-06, "loss": 1.4122, "step": 27280 }, { "epoch": 2.138009404388715, "grad_norm": 3.107977867126465, "learning_rate": 4.473371105415798e-06, "loss": 1.329, "step": 27281 }, { "epoch": 2.1380877742946707, "grad_norm": 2.2058498859405518, "learning_rate": 4.473064459539252e-06, "loss": 1.6383, "step": 27282 }, { "epoch": 2.138166144200627, "grad_norm": 3.1405248641967773, "learning_rate": 4.472757815666894e-06, "loss": 1.735, "step": 27283 }, { "epoch": 2.138244514106583, "grad_norm": 4.373773574829102, "learning_rate": 4.472451173799895e-06, "loss": 0.883, "step": 27284 }, { "epoch": 2.1383228840125392, "grad_norm": 2.9307799339294434, "learning_rate": 4.472144533939422e-06, "loss": 1.5114, "step": 27285 }, { "epoch": 2.138401253918495, "grad_norm": 3.646442174911499, "learning_rate": 4.47183789608664e-06, "loss": 1.5767, "step": 27286 }, { "epoch": 2.1384796238244514, "grad_norm": 3.3885748386383057, "learning_rate": 4.4715312602427145e-06, "loss": 1.3977, "step": 27287 }, { "epoch": 2.1385579937304073, "grad_norm": 3.5329854488372803, "learning_rate": 4.471224626408812e-06, "loss": 1.4281, "step": 27288 }, { "epoch": 2.1386363636363637, "grad_norm": 9.871672630310059, "learning_rate": 4.4709179945861024e-06, "loss": 1.5353, "step": 27289 }, { "epoch": 2.1387147335423196, "grad_norm": 5.060041427612305, "learning_rate": 4.470611364775746e-06, "loss": 1.2571, "step": 27290 }, { "epoch": 2.138793103448276, "grad_norm": 3.328616142272949, "learning_rate": 4.470304736978913e-06, "loss": 1.1416, "step": 27291 }, { "epoch": 2.138871473354232, "grad_norm": 2.3809406757354736, "learning_rate": 4.469998111196769e-06, "loss": 1.4331, "step": 27292 }, { "epoch": 2.138949843260188, "grad_norm": 2.4895920753479004, "learning_rate": 4.469691487430481e-06, "loss": 1.5252, "step": 27293 }, { "epoch": 2.139028213166144, "grad_norm": 2.9612698554992676, "learning_rate": 4.469384865681212e-06, "loss": 1.2249, "step": 27294 }, { "epoch": 2.1391065830721003, "grad_norm": 3.507800579071045, "learning_rate": 4.469078245950131e-06, "loss": 1.5526, "step": 27295 }, { "epoch": 2.139184952978056, "grad_norm": 2.8562886714935303, "learning_rate": 4.468771628238404e-06, "loss": 1.5401, "step": 27296 }, { "epoch": 2.1392633228840126, "grad_norm": 3.824660062789917, "learning_rate": 4.468465012547197e-06, "loss": 1.205, "step": 27297 }, { "epoch": 2.1393416927899684, "grad_norm": 4.095112323760986, "learning_rate": 4.468158398877677e-06, "loss": 1.6403, "step": 27298 }, { "epoch": 2.1394200626959248, "grad_norm": 12.821606636047363, "learning_rate": 4.467851787231008e-06, "loss": 1.4609, "step": 27299 }, { "epoch": 2.139498432601881, "grad_norm": 6.24271297454834, "learning_rate": 4.467545177608358e-06, "loss": 1.397, "step": 27300 }, { "epoch": 2.139576802507837, "grad_norm": 3.901729106903076, "learning_rate": 4.467238570010893e-06, "loss": 1.2194, "step": 27301 }, { "epoch": 2.139655172413793, "grad_norm": 2.6860032081604004, "learning_rate": 4.466931964439777e-06, "loss": 1.5823, "step": 27302 }, { "epoch": 2.139733542319749, "grad_norm": 3.988825798034668, "learning_rate": 4.466625360896179e-06, "loss": 1.3754, "step": 27303 }, { "epoch": 2.1398119122257055, "grad_norm": 3.6965034008026123, "learning_rate": 4.4663187593812655e-06, "loss": 1.2436, "step": 27304 }, { "epoch": 2.1398902821316614, "grad_norm": 3.4705169200897217, "learning_rate": 4.4660121598962e-06, "loss": 1.3927, "step": 27305 }, { "epoch": 2.1399686520376173, "grad_norm": 3.936081886291504, "learning_rate": 4.4657055624421494e-06, "loss": 1.4096, "step": 27306 }, { "epoch": 2.1400470219435737, "grad_norm": 2.6344833374023438, "learning_rate": 4.465398967020281e-06, "loss": 1.0854, "step": 27307 }, { "epoch": 2.14012539184953, "grad_norm": 3.1269311904907227, "learning_rate": 4.465092373631762e-06, "loss": 1.174, "step": 27308 }, { "epoch": 2.140203761755486, "grad_norm": 2.57441782951355, "learning_rate": 4.464785782277754e-06, "loss": 1.1883, "step": 27309 }, { "epoch": 2.140282131661442, "grad_norm": 3.2918150424957275, "learning_rate": 4.4644791929594275e-06, "loss": 1.3015, "step": 27310 }, { "epoch": 2.140360501567398, "grad_norm": 3.0082428455352783, "learning_rate": 4.464172605677947e-06, "loss": 1.5213, "step": 27311 }, { "epoch": 2.1404388714733544, "grad_norm": 4.678092002868652, "learning_rate": 4.463866020434479e-06, "loss": 1.1959, "step": 27312 }, { "epoch": 2.1405172413793103, "grad_norm": 2.884673833847046, "learning_rate": 4.46355943723019e-06, "loss": 1.2044, "step": 27313 }, { "epoch": 2.1405956112852667, "grad_norm": 2.750067949295044, "learning_rate": 4.463252856066243e-06, "loss": 1.3235, "step": 27314 }, { "epoch": 2.1406739811912225, "grad_norm": 2.6519124507904053, "learning_rate": 4.46294627694381e-06, "loss": 1.1919, "step": 27315 }, { "epoch": 2.140752351097179, "grad_norm": 3.2813050746917725, "learning_rate": 4.462639699864049e-06, "loss": 1.3745, "step": 27316 }, { "epoch": 2.1408307210031348, "grad_norm": 3.505557060241699, "learning_rate": 4.462333124828134e-06, "loss": 1.2918, "step": 27317 }, { "epoch": 2.140909090909091, "grad_norm": 9.205095291137695, "learning_rate": 4.462026551837226e-06, "loss": 1.382, "step": 27318 }, { "epoch": 2.140987460815047, "grad_norm": 3.0677855014801025, "learning_rate": 4.461719980892494e-06, "loss": 1.4115, "step": 27319 }, { "epoch": 2.1410658307210033, "grad_norm": 3.525237798690796, "learning_rate": 4.461413411995102e-06, "loss": 1.7119, "step": 27320 }, { "epoch": 2.141144200626959, "grad_norm": 4.094598770141602, "learning_rate": 4.461106845146215e-06, "loss": 1.4761, "step": 27321 }, { "epoch": 2.1412225705329155, "grad_norm": 3.420332670211792, "learning_rate": 4.460800280347002e-06, "loss": 1.561, "step": 27322 }, { "epoch": 2.1413009404388714, "grad_norm": 2.665757417678833, "learning_rate": 4.460493717598629e-06, "loss": 1.4358, "step": 27323 }, { "epoch": 2.1413793103448278, "grad_norm": 2.302537679672241, "learning_rate": 4.460187156902261e-06, "loss": 1.2404, "step": 27324 }, { "epoch": 2.1414576802507836, "grad_norm": 3.975531816482544, "learning_rate": 4.459880598259062e-06, "loss": 1.5861, "step": 27325 }, { "epoch": 2.14153605015674, "grad_norm": 10.367713928222656, "learning_rate": 4.4595740416702e-06, "loss": 0.6582, "step": 27326 }, { "epoch": 2.141614420062696, "grad_norm": 3.9633970260620117, "learning_rate": 4.459267487136843e-06, "loss": 1.8414, "step": 27327 }, { "epoch": 2.141692789968652, "grad_norm": 6.770721912384033, "learning_rate": 4.458960934660152e-06, "loss": 1.5872, "step": 27328 }, { "epoch": 2.141771159874608, "grad_norm": 3.9725146293640137, "learning_rate": 4.458654384241297e-06, "loss": 1.2709, "step": 27329 }, { "epoch": 2.1418495297805644, "grad_norm": 2.7778589725494385, "learning_rate": 4.458347835881443e-06, "loss": 1.0611, "step": 27330 }, { "epoch": 2.1419278996865203, "grad_norm": 2.369391679763794, "learning_rate": 4.458041289581755e-06, "loss": 1.3634, "step": 27331 }, { "epoch": 2.1420062695924766, "grad_norm": 2.847226619720459, "learning_rate": 4.4577347453434e-06, "loss": 1.4872, "step": 27332 }, { "epoch": 2.1420846394984325, "grad_norm": 2.1977710723876953, "learning_rate": 4.457428203167541e-06, "loss": 1.2193, "step": 27333 }, { "epoch": 2.142163009404389, "grad_norm": 3.4237189292907715, "learning_rate": 4.457121663055351e-06, "loss": 1.1463, "step": 27334 }, { "epoch": 2.1422413793103448, "grad_norm": 3.000054359436035, "learning_rate": 4.4568151250079875e-06, "loss": 1.4875, "step": 27335 }, { "epoch": 2.142319749216301, "grad_norm": 7.625284671783447, "learning_rate": 4.4565085890266216e-06, "loss": 1.1208, "step": 27336 }, { "epoch": 2.142398119122257, "grad_norm": 1.9360660314559937, "learning_rate": 4.456202055112418e-06, "loss": 1.36, "step": 27337 }, { "epoch": 2.1424764890282133, "grad_norm": 3.1220967769622803, "learning_rate": 4.455895523266543e-06, "loss": 0.6653, "step": 27338 }, { "epoch": 2.142554858934169, "grad_norm": 2.4441065788269043, "learning_rate": 4.455588993490162e-06, "loss": 1.354, "step": 27339 }, { "epoch": 2.1426332288401255, "grad_norm": 7.140838146209717, "learning_rate": 4.455282465784439e-06, "loss": 1.2833, "step": 27340 }, { "epoch": 2.1427115987460814, "grad_norm": 7.4626617431640625, "learning_rate": 4.454975940150542e-06, "loss": 2.2617, "step": 27341 }, { "epoch": 2.1427899686520377, "grad_norm": 2.625810384750366, "learning_rate": 4.454669416589639e-06, "loss": 1.5342, "step": 27342 }, { "epoch": 2.1428683385579936, "grad_norm": 3.585782766342163, "learning_rate": 4.454362895102892e-06, "loss": 1.3411, "step": 27343 }, { "epoch": 2.14294670846395, "grad_norm": 2.4312994480133057, "learning_rate": 4.454056375691469e-06, "loss": 1.2307, "step": 27344 }, { "epoch": 2.143025078369906, "grad_norm": 4.665245532989502, "learning_rate": 4.453749858356534e-06, "loss": 1.4665, "step": 27345 }, { "epoch": 2.143103448275862, "grad_norm": 2.8641035556793213, "learning_rate": 4.453443343099254e-06, "loss": 1.3736, "step": 27346 }, { "epoch": 2.143181818181818, "grad_norm": 4.438804626464844, "learning_rate": 4.453136829920795e-06, "loss": 1.5473, "step": 27347 }, { "epoch": 2.1432601880877744, "grad_norm": 6.8818254470825195, "learning_rate": 4.4528303188223225e-06, "loss": 0.9103, "step": 27348 }, { "epoch": 2.1433385579937303, "grad_norm": 3.411616563796997, "learning_rate": 4.452523809805004e-06, "loss": 1.503, "step": 27349 }, { "epoch": 2.1434169278996866, "grad_norm": 6.02122163772583, "learning_rate": 4.452217302870002e-06, "loss": 0.917, "step": 27350 }, { "epoch": 2.1434952978056425, "grad_norm": 4.118086814880371, "learning_rate": 4.451910798018484e-06, "loss": 1.2845, "step": 27351 }, { "epoch": 2.143573667711599, "grad_norm": 3.9488630294799805, "learning_rate": 4.451604295251615e-06, "loss": 1.604, "step": 27352 }, { "epoch": 2.1436520376175547, "grad_norm": 3.370584726333618, "learning_rate": 4.451297794570563e-06, "loss": 1.2196, "step": 27353 }, { "epoch": 2.143730407523511, "grad_norm": 2.9197938442230225, "learning_rate": 4.450991295976493e-06, "loss": 1.1945, "step": 27354 }, { "epoch": 2.143808777429467, "grad_norm": 6.571789264678955, "learning_rate": 4.450684799470569e-06, "loss": 1.3072, "step": 27355 }, { "epoch": 2.1438871473354233, "grad_norm": 2.622140645980835, "learning_rate": 4.4503783050539596e-06, "loss": 1.3757, "step": 27356 }, { "epoch": 2.143965517241379, "grad_norm": 3.146001100540161, "learning_rate": 4.450071812727827e-06, "loss": 1.4607, "step": 27357 }, { "epoch": 2.1440438871473355, "grad_norm": 3.592804431915283, "learning_rate": 4.449765322493338e-06, "loss": 1.5007, "step": 27358 }, { "epoch": 2.1441222570532914, "grad_norm": 7.773771286010742, "learning_rate": 4.44945883435166e-06, "loss": 0.8561, "step": 27359 }, { "epoch": 2.1442006269592477, "grad_norm": 2.7961905002593994, "learning_rate": 4.4491523483039595e-06, "loss": 1.8926, "step": 27360 }, { "epoch": 2.1442789968652036, "grad_norm": 4.744790077209473, "learning_rate": 4.448845864351398e-06, "loss": 2.0573, "step": 27361 }, { "epoch": 2.14435736677116, "grad_norm": 2.7225759029388428, "learning_rate": 4.448539382495146e-06, "loss": 1.4346, "step": 27362 }, { "epoch": 2.144435736677116, "grad_norm": 4.380146026611328, "learning_rate": 4.448232902736365e-06, "loss": 1.4107, "step": 27363 }, { "epoch": 2.144514106583072, "grad_norm": 4.747612476348877, "learning_rate": 4.447926425076225e-06, "loss": 0.8344, "step": 27364 }, { "epoch": 2.144592476489028, "grad_norm": 2.876695394515991, "learning_rate": 4.447619949515886e-06, "loss": 1.2775, "step": 27365 }, { "epoch": 2.1446708463949844, "grad_norm": 3.709425210952759, "learning_rate": 4.447313476056519e-06, "loss": 1.0477, "step": 27366 }, { "epoch": 2.1447492163009403, "grad_norm": 3.5968518257141113, "learning_rate": 4.447007004699288e-06, "loss": 1.424, "step": 27367 }, { "epoch": 2.1448275862068966, "grad_norm": 16.612234115600586, "learning_rate": 4.446700535445358e-06, "loss": 1.255, "step": 27368 }, { "epoch": 2.1449059561128525, "grad_norm": 2.150001049041748, "learning_rate": 4.446394068295894e-06, "loss": 1.3009, "step": 27369 }, { "epoch": 2.144984326018809, "grad_norm": 2.9732666015625, "learning_rate": 4.446087603252063e-06, "loss": 1.2735, "step": 27370 }, { "epoch": 2.1450626959247647, "grad_norm": 3.3597307205200195, "learning_rate": 4.445781140315032e-06, "loss": 1.2151, "step": 27371 }, { "epoch": 2.145141065830721, "grad_norm": 3.3814868927001953, "learning_rate": 4.4454746794859625e-06, "loss": 1.4424, "step": 27372 }, { "epoch": 2.145219435736677, "grad_norm": 2.1935973167419434, "learning_rate": 4.445168220766023e-06, "loss": 1.2557, "step": 27373 }, { "epoch": 2.1452978056426333, "grad_norm": 2.1787333488464355, "learning_rate": 4.4448617641563794e-06, "loss": 1.1148, "step": 27374 }, { "epoch": 2.145376175548589, "grad_norm": 5.3720173835754395, "learning_rate": 4.4445553096581964e-06, "loss": 1.2337, "step": 27375 }, { "epoch": 2.1454545454545455, "grad_norm": 3.182462692260742, "learning_rate": 4.44424885727264e-06, "loss": 1.4109, "step": 27376 }, { "epoch": 2.1455329153605014, "grad_norm": 3.940390110015869, "learning_rate": 4.443942407000873e-06, "loss": 2.0131, "step": 27377 }, { "epoch": 2.1456112852664577, "grad_norm": 2.7619476318359375, "learning_rate": 4.443635958844066e-06, "loss": 0.8653, "step": 27378 }, { "epoch": 2.1456896551724136, "grad_norm": 2.6818578243255615, "learning_rate": 4.443329512803383e-06, "loss": 1.4578, "step": 27379 }, { "epoch": 2.14576802507837, "grad_norm": 3.281301975250244, "learning_rate": 4.443023068879987e-06, "loss": 1.4938, "step": 27380 }, { "epoch": 2.145846394984326, "grad_norm": 2.3012804985046387, "learning_rate": 4.442716627075045e-06, "loss": 1.3203, "step": 27381 }, { "epoch": 2.145924764890282, "grad_norm": 3.3002853393554688, "learning_rate": 4.442410187389721e-06, "loss": 1.2611, "step": 27382 }, { "epoch": 2.146003134796238, "grad_norm": 2.6237761974334717, "learning_rate": 4.4421037498251864e-06, "loss": 1.524, "step": 27383 }, { "epoch": 2.1460815047021944, "grad_norm": 3.363554000854492, "learning_rate": 4.4417973143825995e-06, "loss": 1.1404, "step": 27384 }, { "epoch": 2.1461598746081503, "grad_norm": 3.413728952407837, "learning_rate": 4.441490881063129e-06, "loss": 1.4117, "step": 27385 }, { "epoch": 2.1462382445141066, "grad_norm": 2.9554712772369385, "learning_rate": 4.4411844498679425e-06, "loss": 1.5062, "step": 27386 }, { "epoch": 2.1463166144200625, "grad_norm": 2.288133382797241, "learning_rate": 4.440878020798202e-06, "loss": 1.5678, "step": 27387 }, { "epoch": 2.146394984326019, "grad_norm": 10.073701858520508, "learning_rate": 4.440571593855074e-06, "loss": 1.7373, "step": 27388 }, { "epoch": 2.1464733542319747, "grad_norm": 5.741009712219238, "learning_rate": 4.440265169039724e-06, "loss": 1.0951, "step": 27389 }, { "epoch": 2.146551724137931, "grad_norm": 3.2664785385131836, "learning_rate": 4.43995874635332e-06, "loss": 1.1102, "step": 27390 }, { "epoch": 2.1466300940438874, "grad_norm": 4.85213565826416, "learning_rate": 4.439652325797022e-06, "loss": 1.5105, "step": 27391 }, { "epoch": 2.1467084639498433, "grad_norm": 3.725503444671631, "learning_rate": 4.4393459073720005e-06, "loss": 1.7314, "step": 27392 }, { "epoch": 2.146786833855799, "grad_norm": 2.2927374839782715, "learning_rate": 4.439039491079419e-06, "loss": 1.0854, "step": 27393 }, { "epoch": 2.1468652037617555, "grad_norm": 1.9531331062316895, "learning_rate": 4.438733076920444e-06, "loss": 1.289, "step": 27394 }, { "epoch": 2.146943573667712, "grad_norm": 13.792600631713867, "learning_rate": 4.4384266648962395e-06, "loss": 1.1827, "step": 27395 }, { "epoch": 2.1470219435736677, "grad_norm": 2.8248586654663086, "learning_rate": 4.438120255007971e-06, "loss": 0.8432, "step": 27396 }, { "epoch": 2.1471003134796236, "grad_norm": 5.39491081237793, "learning_rate": 4.4378138472568035e-06, "loss": 1.2896, "step": 27397 }, { "epoch": 2.14717868338558, "grad_norm": 3.4065732955932617, "learning_rate": 4.4375074416439065e-06, "loss": 1.2758, "step": 27398 }, { "epoch": 2.1472570532915363, "grad_norm": 4.800049304962158, "learning_rate": 4.43720103817044e-06, "loss": 1.0807, "step": 27399 }, { "epoch": 2.147335423197492, "grad_norm": 5.156898498535156, "learning_rate": 4.436894636837572e-06, "loss": 1.4866, "step": 27400 }, { "epoch": 2.147413793103448, "grad_norm": 3.705702304840088, "learning_rate": 4.436588237646469e-06, "loss": 1.4527, "step": 27401 }, { "epoch": 2.1474921630094044, "grad_norm": 4.0024542808532715, "learning_rate": 4.436281840598294e-06, "loss": 0.8429, "step": 27402 }, { "epoch": 2.1475705329153607, "grad_norm": 2.625115156173706, "learning_rate": 4.435975445694211e-06, "loss": 1.305, "step": 27403 }, { "epoch": 2.1476489028213166, "grad_norm": 3.0965495109558105, "learning_rate": 4.43566905293539e-06, "loss": 1.0322, "step": 27404 }, { "epoch": 2.147727272727273, "grad_norm": 3.4782094955444336, "learning_rate": 4.435362662322994e-06, "loss": 0.9512, "step": 27405 }, { "epoch": 2.147805642633229, "grad_norm": 2.2660794258117676, "learning_rate": 4.4350562738581886e-06, "loss": 1.2615, "step": 27406 }, { "epoch": 2.147884012539185, "grad_norm": 4.465564250946045, "learning_rate": 4.434749887542139e-06, "loss": 1.6621, "step": 27407 }, { "epoch": 2.147962382445141, "grad_norm": 2.6812744140625, "learning_rate": 4.434443503376008e-06, "loss": 1.2773, "step": 27408 }, { "epoch": 2.1480407523510974, "grad_norm": 2.6739132404327393, "learning_rate": 4.434137121360967e-06, "loss": 1.7511, "step": 27409 }, { "epoch": 2.1481191222570533, "grad_norm": 14.975531578063965, "learning_rate": 4.433830741498176e-06, "loss": 1.279, "step": 27410 }, { "epoch": 2.1481974921630096, "grad_norm": 2.9522314071655273, "learning_rate": 4.433524363788802e-06, "loss": 1.3629, "step": 27411 }, { "epoch": 2.1482758620689655, "grad_norm": 4.152809143066406, "learning_rate": 4.433217988234011e-06, "loss": 1.346, "step": 27412 }, { "epoch": 2.148354231974922, "grad_norm": 2.6757874488830566, "learning_rate": 4.432911614834967e-06, "loss": 0.8367, "step": 27413 }, { "epoch": 2.1484326018808777, "grad_norm": 2.9992754459381104, "learning_rate": 4.432605243592837e-06, "loss": 1.1914, "step": 27414 }, { "epoch": 2.148510971786834, "grad_norm": 4.875482082366943, "learning_rate": 4.432298874508783e-06, "loss": 1.6693, "step": 27415 }, { "epoch": 2.14858934169279, "grad_norm": 2.9786837100982666, "learning_rate": 4.431992507583976e-06, "loss": 1.3024, "step": 27416 }, { "epoch": 2.1486677115987463, "grad_norm": 3.037156105041504, "learning_rate": 4.431686142819575e-06, "loss": 1.5907, "step": 27417 }, { "epoch": 2.148746081504702, "grad_norm": 2.2816200256347656, "learning_rate": 4.431379780216748e-06, "loss": 1.2832, "step": 27418 }, { "epoch": 2.1488244514106585, "grad_norm": 3.7126243114471436, "learning_rate": 4.431073419776661e-06, "loss": 1.2939, "step": 27419 }, { "epoch": 2.1489028213166144, "grad_norm": 2.8283417224884033, "learning_rate": 4.430767061500479e-06, "loss": 1.3255, "step": 27420 }, { "epoch": 2.1489811912225707, "grad_norm": 3.4253146648406982, "learning_rate": 4.430460705389367e-06, "loss": 1.3904, "step": 27421 }, { "epoch": 2.1490595611285266, "grad_norm": 2.812884569168091, "learning_rate": 4.430154351444487e-06, "loss": 1.167, "step": 27422 }, { "epoch": 2.149137931034483, "grad_norm": 2.967170238494873, "learning_rate": 4.429847999667009e-06, "loss": 1.5503, "step": 27423 }, { "epoch": 2.149216300940439, "grad_norm": 2.64157772064209, "learning_rate": 4.429541650058098e-06, "loss": 1.3815, "step": 27424 }, { "epoch": 2.149294670846395, "grad_norm": 3.138918399810791, "learning_rate": 4.429235302618916e-06, "loss": 1.6282, "step": 27425 }, { "epoch": 2.149373040752351, "grad_norm": 3.3972420692443848, "learning_rate": 4.42892895735063e-06, "loss": 1.3047, "step": 27426 }, { "epoch": 2.1494514106583074, "grad_norm": 3.5476906299591064, "learning_rate": 4.428622614254406e-06, "loss": 1.4873, "step": 27427 }, { "epoch": 2.1495297805642632, "grad_norm": 2.7607271671295166, "learning_rate": 4.428316273331406e-06, "loss": 1.4329, "step": 27428 }, { "epoch": 2.1496081504702196, "grad_norm": 3.629007339477539, "learning_rate": 4.428009934582797e-06, "loss": 1.3014, "step": 27429 }, { "epoch": 2.1496865203761755, "grad_norm": 3.6365914344787598, "learning_rate": 4.427703598009746e-06, "loss": 1.421, "step": 27430 }, { "epoch": 2.149764890282132, "grad_norm": 3.275883674621582, "learning_rate": 4.427397263613417e-06, "loss": 1.3206, "step": 27431 }, { "epoch": 2.1498432601880877, "grad_norm": 3.542630195617676, "learning_rate": 4.427090931394974e-06, "loss": 1.2269, "step": 27432 }, { "epoch": 2.149921630094044, "grad_norm": 2.510424852371216, "learning_rate": 4.426784601355583e-06, "loss": 1.2718, "step": 27433 }, { "epoch": 2.15, "grad_norm": 2.5062196254730225, "learning_rate": 4.426478273496407e-06, "loss": 1.381, "step": 27434 }, { "epoch": 2.1500783699059562, "grad_norm": 2.8774144649505615, "learning_rate": 4.426171947818617e-06, "loss": 0.7882, "step": 27435 }, { "epoch": 2.150156739811912, "grad_norm": 4.263574123382568, "learning_rate": 4.42586562432337e-06, "loss": 1.0966, "step": 27436 }, { "epoch": 2.1502351097178685, "grad_norm": 15.431478500366211, "learning_rate": 4.425559303011838e-06, "loss": 1.05, "step": 27437 }, { "epoch": 2.1503134796238244, "grad_norm": 3.7994003295898438, "learning_rate": 4.4252529838851825e-06, "loss": 0.848, "step": 27438 }, { "epoch": 2.1503918495297807, "grad_norm": 4.420429229736328, "learning_rate": 4.42494666694457e-06, "loss": 1.7139, "step": 27439 }, { "epoch": 2.1504702194357366, "grad_norm": 2.6586363315582275, "learning_rate": 4.424640352191165e-06, "loss": 1.5026, "step": 27440 }, { "epoch": 2.150548589341693, "grad_norm": 2.0108261108398438, "learning_rate": 4.424334039626131e-06, "loss": 1.0457, "step": 27441 }, { "epoch": 2.150626959247649, "grad_norm": 6.448887348175049, "learning_rate": 4.424027729250636e-06, "loss": 1.1978, "step": 27442 }, { "epoch": 2.150705329153605, "grad_norm": 3.8175172805786133, "learning_rate": 4.4237214210658435e-06, "loss": 0.7545, "step": 27443 }, { "epoch": 2.150783699059561, "grad_norm": 4.039362907409668, "learning_rate": 4.423415115072919e-06, "loss": 1.3987, "step": 27444 }, { "epoch": 2.1508620689655173, "grad_norm": 5.091869831085205, "learning_rate": 4.423108811273027e-06, "loss": 0.966, "step": 27445 }, { "epoch": 2.1509404388714732, "grad_norm": 3.5032737255096436, "learning_rate": 4.422802509667332e-06, "loss": 1.3897, "step": 27446 }, { "epoch": 2.1510188087774296, "grad_norm": 3.6345908641815186, "learning_rate": 4.422496210257002e-06, "loss": 1.5303, "step": 27447 }, { "epoch": 2.1510971786833855, "grad_norm": 11.26568603515625, "learning_rate": 4.422189913043196e-06, "loss": 1.0643, "step": 27448 }, { "epoch": 2.151175548589342, "grad_norm": 5.858158588409424, "learning_rate": 4.421883618027085e-06, "loss": 1.3542, "step": 27449 }, { "epoch": 2.1512539184952977, "grad_norm": 2.357320547103882, "learning_rate": 4.421577325209833e-06, "loss": 1.2111, "step": 27450 }, { "epoch": 2.151332288401254, "grad_norm": 2.4416821002960205, "learning_rate": 4.421271034592602e-06, "loss": 1.2898, "step": 27451 }, { "epoch": 2.15141065830721, "grad_norm": 3.1350114345550537, "learning_rate": 4.420964746176559e-06, "loss": 1.4093, "step": 27452 }, { "epoch": 2.1514890282131662, "grad_norm": 3.2335174083709717, "learning_rate": 4.420658459962868e-06, "loss": 1.3274, "step": 27453 }, { "epoch": 2.151567398119122, "grad_norm": 2.8254687786102295, "learning_rate": 4.420352175952697e-06, "loss": 1.6174, "step": 27454 }, { "epoch": 2.1516457680250785, "grad_norm": 2.6086742877960205, "learning_rate": 4.4200458941472066e-06, "loss": 1.158, "step": 27455 }, { "epoch": 2.1517241379310343, "grad_norm": 5.050987243652344, "learning_rate": 4.419739614547565e-06, "loss": 1.178, "step": 27456 }, { "epoch": 2.1518025078369907, "grad_norm": 8.220598220825195, "learning_rate": 4.4194333371549356e-06, "loss": 1.043, "step": 27457 }, { "epoch": 2.1518808777429466, "grad_norm": 3.8707921504974365, "learning_rate": 4.419127061970483e-06, "loss": 1.4689, "step": 27458 }, { "epoch": 2.151959247648903, "grad_norm": 3.114476203918457, "learning_rate": 4.4188207889953725e-06, "loss": 1.4225, "step": 27459 }, { "epoch": 2.152037617554859, "grad_norm": 3.178917169570923, "learning_rate": 4.418514518230769e-06, "loss": 1.5832, "step": 27460 }, { "epoch": 2.152115987460815, "grad_norm": 3.1683332920074463, "learning_rate": 4.418208249677839e-06, "loss": 1.8852, "step": 27461 }, { "epoch": 2.152194357366771, "grad_norm": 2.3088972568511963, "learning_rate": 4.417901983337745e-06, "loss": 1.2052, "step": 27462 }, { "epoch": 2.1522727272727273, "grad_norm": 3.7371714115142822, "learning_rate": 4.417595719211653e-06, "loss": 1.5171, "step": 27463 }, { "epoch": 2.1523510971786832, "grad_norm": 13.7506103515625, "learning_rate": 4.417289457300728e-06, "loss": 2.0278, "step": 27464 }, { "epoch": 2.1524294670846396, "grad_norm": 2.156330108642578, "learning_rate": 4.416983197606135e-06, "loss": 1.4145, "step": 27465 }, { "epoch": 2.1525078369905954, "grad_norm": 3.0853400230407715, "learning_rate": 4.416676940129037e-06, "loss": 1.405, "step": 27466 }, { "epoch": 2.152586206896552, "grad_norm": 3.066131114959717, "learning_rate": 4.416370684870601e-06, "loss": 1.1635, "step": 27467 }, { "epoch": 2.1526645768025077, "grad_norm": 2.9448235034942627, "learning_rate": 4.416064431831992e-06, "loss": 1.3545, "step": 27468 }, { "epoch": 2.152742946708464, "grad_norm": 4.164117336273193, "learning_rate": 4.415758181014373e-06, "loss": 1.5036, "step": 27469 }, { "epoch": 2.15282131661442, "grad_norm": 2.52962327003479, "learning_rate": 4.41545193241891e-06, "loss": 1.5546, "step": 27470 }, { "epoch": 2.152899686520376, "grad_norm": 3.880915641784668, "learning_rate": 4.415145686046767e-06, "loss": 1.6733, "step": 27471 }, { "epoch": 2.152978056426332, "grad_norm": 3.366081714630127, "learning_rate": 4.414839441899112e-06, "loss": 1.5734, "step": 27472 }, { "epoch": 2.1530564263322884, "grad_norm": 4.651971817016602, "learning_rate": 4.414533199977104e-06, "loss": 1.1784, "step": 27473 }, { "epoch": 2.1531347962382443, "grad_norm": 3.1923277378082275, "learning_rate": 4.4142269602819125e-06, "loss": 1.1587, "step": 27474 }, { "epoch": 2.1532131661442007, "grad_norm": 2.9846465587615967, "learning_rate": 4.413920722814701e-06, "loss": 1.4811, "step": 27475 }, { "epoch": 2.1532915360501566, "grad_norm": 2.9675228595733643, "learning_rate": 4.413614487576634e-06, "loss": 1.2953, "step": 27476 }, { "epoch": 2.153369905956113, "grad_norm": 3.110837459564209, "learning_rate": 4.413308254568876e-06, "loss": 1.846, "step": 27477 }, { "epoch": 2.1534482758620688, "grad_norm": 4.108426570892334, "learning_rate": 4.413002023792591e-06, "loss": 1.4027, "step": 27478 }, { "epoch": 2.153526645768025, "grad_norm": 2.262763023376465, "learning_rate": 4.412695795248945e-06, "loss": 1.3303, "step": 27479 }, { "epoch": 2.153605015673981, "grad_norm": 2.4505462646484375, "learning_rate": 4.412389568939104e-06, "loss": 1.18, "step": 27480 }, { "epoch": 2.1536833855799373, "grad_norm": 3.863738536834717, "learning_rate": 4.4120833448642305e-06, "loss": 1.5732, "step": 27481 }, { "epoch": 2.153761755485893, "grad_norm": 6.748054027557373, "learning_rate": 4.411777123025491e-06, "loss": 1.2338, "step": 27482 }, { "epoch": 2.1538401253918495, "grad_norm": 4.8715291023254395, "learning_rate": 4.4114709034240485e-06, "loss": 1.237, "step": 27483 }, { "epoch": 2.1539184952978054, "grad_norm": 5.564545154571533, "learning_rate": 4.411164686061069e-06, "loss": 1.651, "step": 27484 }, { "epoch": 2.1539968652037618, "grad_norm": 2.9281768798828125, "learning_rate": 4.410858470937714e-06, "loss": 1.1855, "step": 27485 }, { "epoch": 2.154075235109718, "grad_norm": 6.445323944091797, "learning_rate": 4.410552258055152e-06, "loss": 1.4161, "step": 27486 }, { "epoch": 2.154153605015674, "grad_norm": 2.612997055053711, "learning_rate": 4.4102460474145485e-06, "loss": 1.4666, "step": 27487 }, { "epoch": 2.15423197492163, "grad_norm": 2.6991758346557617, "learning_rate": 4.4099398390170644e-06, "loss": 1.1304, "step": 27488 }, { "epoch": 2.154310344827586, "grad_norm": 2.7817394733428955, "learning_rate": 4.409633632863866e-06, "loss": 1.436, "step": 27489 }, { "epoch": 2.1543887147335425, "grad_norm": 2.759547472000122, "learning_rate": 4.4093274289561175e-06, "loss": 1.6133, "step": 27490 }, { "epoch": 2.1544670846394984, "grad_norm": 2.551811695098877, "learning_rate": 4.4090212272949865e-06, "loss": 1.1836, "step": 27491 }, { "epoch": 2.1545454545454543, "grad_norm": 4.361412525177002, "learning_rate": 4.408715027881633e-06, "loss": 1.2062, "step": 27492 }, { "epoch": 2.1546238244514107, "grad_norm": 3.0621654987335205, "learning_rate": 4.4084088307172235e-06, "loss": 0.9334, "step": 27493 }, { "epoch": 2.154702194357367, "grad_norm": 5.953489303588867, "learning_rate": 4.408102635802924e-06, "loss": 1.4255, "step": 27494 }, { "epoch": 2.154780564263323, "grad_norm": 3.977755546569824, "learning_rate": 4.407796443139899e-06, "loss": 1.9774, "step": 27495 }, { "epoch": 2.1548589341692788, "grad_norm": 2.209070920944214, "learning_rate": 4.407490252729311e-06, "loss": 1.3577, "step": 27496 }, { "epoch": 2.154937304075235, "grad_norm": 3.2536354064941406, "learning_rate": 4.407184064572324e-06, "loss": 1.4303, "step": 27497 }, { "epoch": 2.1550156739811914, "grad_norm": 5.038819313049316, "learning_rate": 4.406877878670109e-06, "loss": 1.2884, "step": 27498 }, { "epoch": 2.1550940438871473, "grad_norm": 2.6951420307159424, "learning_rate": 4.406571695023821e-06, "loss": 1.4495, "step": 27499 }, { "epoch": 2.1551724137931036, "grad_norm": 5.361559867858887, "learning_rate": 4.406265513634632e-06, "loss": 1.6738, "step": 27500 }, { "epoch": 2.1552507836990595, "grad_norm": 4.756617069244385, "learning_rate": 4.405959334503703e-06, "loss": 0.9064, "step": 27501 }, { "epoch": 2.155329153605016, "grad_norm": 2.1442532539367676, "learning_rate": 4.405653157632201e-06, "loss": 1.4952, "step": 27502 }, { "epoch": 2.1554075235109718, "grad_norm": 3.467837333679199, "learning_rate": 4.405346983021288e-06, "loss": 1.7821, "step": 27503 }, { "epoch": 2.155485893416928, "grad_norm": 2.528165817260742, "learning_rate": 4.405040810672129e-06, "loss": 1.2531, "step": 27504 }, { "epoch": 2.155564263322884, "grad_norm": 10.55933666229248, "learning_rate": 4.40473464058589e-06, "loss": 1.9022, "step": 27505 }, { "epoch": 2.1556426332288403, "grad_norm": 2.0046095848083496, "learning_rate": 4.404428472763735e-06, "loss": 1.4148, "step": 27506 }, { "epoch": 2.155721003134796, "grad_norm": 3.1291675567626953, "learning_rate": 4.404122307206828e-06, "loss": 1.1368, "step": 27507 }, { "epoch": 2.1557993730407525, "grad_norm": 2.6889758110046387, "learning_rate": 4.403816143916334e-06, "loss": 1.5545, "step": 27508 }, { "epoch": 2.1558777429467084, "grad_norm": 14.859112739562988, "learning_rate": 4.403509982893418e-06, "loss": 0.2886, "step": 27509 }, { "epoch": 2.1559561128526648, "grad_norm": 2.8031532764434814, "learning_rate": 4.403203824139242e-06, "loss": 1.0805, "step": 27510 }, { "epoch": 2.1560344827586206, "grad_norm": 3.4419710636138916, "learning_rate": 4.4028976676549715e-06, "loss": 1.4078, "step": 27511 }, { "epoch": 2.156112852664577, "grad_norm": 3.9600484371185303, "learning_rate": 4.402591513441773e-06, "loss": 1.3634, "step": 27512 }, { "epoch": 2.156191222570533, "grad_norm": 4.028415679931641, "learning_rate": 4.402285361500809e-06, "loss": 1.459, "step": 27513 }, { "epoch": 2.156269592476489, "grad_norm": 2.496816396713257, "learning_rate": 4.4019792118332446e-06, "loss": 1.4195, "step": 27514 }, { "epoch": 2.156347962382445, "grad_norm": 3.6273579597473145, "learning_rate": 4.401673064440244e-06, "loss": 1.1076, "step": 27515 }, { "epoch": 2.1564263322884014, "grad_norm": 1.7668774127960205, "learning_rate": 4.401366919322971e-06, "loss": 1.2268, "step": 27516 }, { "epoch": 2.1565047021943573, "grad_norm": 6.239473342895508, "learning_rate": 4.401060776482594e-06, "loss": 1.5087, "step": 27517 }, { "epoch": 2.1565830721003136, "grad_norm": 4.232467174530029, "learning_rate": 4.40075463592027e-06, "loss": 1.0483, "step": 27518 }, { "epoch": 2.1566614420062695, "grad_norm": 3.6437926292419434, "learning_rate": 4.400448497637169e-06, "loss": 1.5613, "step": 27519 }, { "epoch": 2.156739811912226, "grad_norm": 3.8087844848632812, "learning_rate": 4.400142361634455e-06, "loss": 2.0541, "step": 27520 }, { "epoch": 2.1568181818181817, "grad_norm": 4.0891032218933105, "learning_rate": 4.399836227913291e-06, "loss": 1.6042, "step": 27521 }, { "epoch": 2.156896551724138, "grad_norm": 3.789308547973633, "learning_rate": 4.3995300964748415e-06, "loss": 1.7035, "step": 27522 }, { "epoch": 2.156974921630094, "grad_norm": 2.432968854904175, "learning_rate": 4.3992239673202695e-06, "loss": 1.1216, "step": 27523 }, { "epoch": 2.1570532915360503, "grad_norm": 2.4780008792877197, "learning_rate": 4.398917840450745e-06, "loss": 1.4162, "step": 27524 }, { "epoch": 2.157131661442006, "grad_norm": 3.189100503921509, "learning_rate": 4.398611715867424e-06, "loss": 1.3397, "step": 27525 }, { "epoch": 2.1572100313479625, "grad_norm": 4.899746894836426, "learning_rate": 4.398305593571477e-06, "loss": 1.3601, "step": 27526 }, { "epoch": 2.1572884012539184, "grad_norm": 2.9904377460479736, "learning_rate": 4.3979994735640665e-06, "loss": 1.3139, "step": 27527 }, { "epoch": 2.1573667711598747, "grad_norm": 3.530540704727173, "learning_rate": 4.397693355846358e-06, "loss": 1.6796, "step": 27528 }, { "epoch": 2.1574451410658306, "grad_norm": 3.340993642807007, "learning_rate": 4.3973872404195125e-06, "loss": 0.8835, "step": 27529 }, { "epoch": 2.157523510971787, "grad_norm": 10.353195190429688, "learning_rate": 4.397081127284697e-06, "loss": 0.5584, "step": 27530 }, { "epoch": 2.157601880877743, "grad_norm": 2.8286330699920654, "learning_rate": 4.3967750164430745e-06, "loss": 1.0619, "step": 27531 }, { "epoch": 2.157680250783699, "grad_norm": 3.0896923542022705, "learning_rate": 4.396468907895812e-06, "loss": 1.7036, "step": 27532 }, { "epoch": 2.157758620689655, "grad_norm": 6.227737903594971, "learning_rate": 4.396162801644071e-06, "loss": 1.4189, "step": 27533 }, { "epoch": 2.1578369905956114, "grad_norm": 3.362884759902954, "learning_rate": 4.395856697689015e-06, "loss": 1.4501, "step": 27534 }, { "epoch": 2.1579153605015673, "grad_norm": 2.086864471435547, "learning_rate": 4.395550596031811e-06, "loss": 1.1795, "step": 27535 }, { "epoch": 2.1579937304075236, "grad_norm": 4.16055965423584, "learning_rate": 4.395244496673624e-06, "loss": 1.7938, "step": 27536 }, { "epoch": 2.1580721003134795, "grad_norm": 5.9538421630859375, "learning_rate": 4.3949383996156135e-06, "loss": 1.5904, "step": 27537 }, { "epoch": 2.158150470219436, "grad_norm": 2.797393321990967, "learning_rate": 4.394632304858947e-06, "loss": 1.3368, "step": 27538 }, { "epoch": 2.1582288401253917, "grad_norm": 2.1094589233398438, "learning_rate": 4.394326212404789e-06, "loss": 1.3408, "step": 27539 }, { "epoch": 2.158307210031348, "grad_norm": 2.8679885864257812, "learning_rate": 4.3940201222543036e-06, "loss": 1.0507, "step": 27540 }, { "epoch": 2.158385579937304, "grad_norm": 4.090983867645264, "learning_rate": 4.3937140344086534e-06, "loss": 1.1184, "step": 27541 }, { "epoch": 2.1584639498432603, "grad_norm": 3.6285815238952637, "learning_rate": 4.393407948869002e-06, "loss": 1.692, "step": 27542 }, { "epoch": 2.158542319749216, "grad_norm": 3.667634963989258, "learning_rate": 4.39310186563652e-06, "loss": 1.6237, "step": 27543 }, { "epoch": 2.1586206896551725, "grad_norm": 3.0360419750213623, "learning_rate": 4.392795784712361e-06, "loss": 1.2408, "step": 27544 }, { "epoch": 2.1586990595611284, "grad_norm": 3.394998788833618, "learning_rate": 4.392489706097698e-06, "loss": 1.3224, "step": 27545 }, { "epoch": 2.1587774294670847, "grad_norm": 4.882524013519287, "learning_rate": 4.392183629793692e-06, "loss": 1.1744, "step": 27546 }, { "epoch": 2.1588557993730406, "grad_norm": 3.9907028675079346, "learning_rate": 4.391877555801508e-06, "loss": 1.2683, "step": 27547 }, { "epoch": 2.158934169278997, "grad_norm": 4.193512439727783, "learning_rate": 4.391571484122308e-06, "loss": 1.7247, "step": 27548 }, { "epoch": 2.159012539184953, "grad_norm": 3.1228880882263184, "learning_rate": 4.3912654147572575e-06, "loss": 1.1513, "step": 27549 }, { "epoch": 2.159090909090909, "grad_norm": 7.33929443359375, "learning_rate": 4.390959347707521e-06, "loss": 1.0957, "step": 27550 }, { "epoch": 2.159169278996865, "grad_norm": 13.725715637207031, "learning_rate": 4.390653282974264e-06, "loss": 1.1771, "step": 27551 }, { "epoch": 2.1592476489028214, "grad_norm": 1.8397340774536133, "learning_rate": 4.390347220558647e-06, "loss": 1.201, "step": 27552 }, { "epoch": 2.1593260188087773, "grad_norm": 2.8790957927703857, "learning_rate": 4.390041160461837e-06, "loss": 1.2643, "step": 27553 }, { "epoch": 2.1594043887147336, "grad_norm": 6.591362476348877, "learning_rate": 4.389735102684997e-06, "loss": 1.4647, "step": 27554 }, { "epoch": 2.1594827586206895, "grad_norm": 3.807349443435669, "learning_rate": 4.38942904722929e-06, "loss": 1.3834, "step": 27555 }, { "epoch": 2.159561128526646, "grad_norm": 4.326879501342773, "learning_rate": 4.389122994095882e-06, "loss": 1.529, "step": 27556 }, { "epoch": 2.1596394984326017, "grad_norm": 3.926778554916382, "learning_rate": 4.388816943285936e-06, "loss": 1.6079, "step": 27557 }, { "epoch": 2.159717868338558, "grad_norm": 4.073122024536133, "learning_rate": 4.388510894800618e-06, "loss": 1.2515, "step": 27558 }, { "epoch": 2.159796238244514, "grad_norm": 3.3912954330444336, "learning_rate": 4.388204848641089e-06, "loss": 1.3499, "step": 27559 }, { "epoch": 2.1598746081504703, "grad_norm": 3.96590256690979, "learning_rate": 4.387898804808516e-06, "loss": 1.3599, "step": 27560 }, { "epoch": 2.159952978056426, "grad_norm": 2.5237607955932617, "learning_rate": 4.387592763304059e-06, "loss": 1.1807, "step": 27561 }, { "epoch": 2.1600313479623825, "grad_norm": 3.6611416339874268, "learning_rate": 4.387286724128887e-06, "loss": 1.6968, "step": 27562 }, { "epoch": 2.1601097178683384, "grad_norm": 2.532801389694214, "learning_rate": 4.3869806872841605e-06, "loss": 1.5435, "step": 27563 }, { "epoch": 2.1601880877742947, "grad_norm": 3.102576971054077, "learning_rate": 4.386674652771045e-06, "loss": 1.2061, "step": 27564 }, { "epoch": 2.1602664576802506, "grad_norm": 2.528721809387207, "learning_rate": 4.386368620590705e-06, "loss": 1.2612, "step": 27565 }, { "epoch": 2.160344827586207, "grad_norm": 3.5675854682922363, "learning_rate": 4.3860625907443025e-06, "loss": 1.7007, "step": 27566 }, { "epoch": 2.160423197492163, "grad_norm": 2.730894088745117, "learning_rate": 4.385756563233001e-06, "loss": 1.4347, "step": 27567 }, { "epoch": 2.160501567398119, "grad_norm": 3.3814377784729004, "learning_rate": 4.385450538057967e-06, "loss": 1.2519, "step": 27568 }, { "epoch": 2.160579937304075, "grad_norm": 3.7400083541870117, "learning_rate": 4.385144515220365e-06, "loss": 1.3533, "step": 27569 }, { "epoch": 2.1606583072100314, "grad_norm": 2.411682367324829, "learning_rate": 4.384838494721357e-06, "loss": 1.2673, "step": 27570 }, { "epoch": 2.1607366771159873, "grad_norm": 2.6883389949798584, "learning_rate": 4.384532476562107e-06, "loss": 1.2173, "step": 27571 }, { "epoch": 2.1608150470219436, "grad_norm": 4.428774833679199, "learning_rate": 4.384226460743778e-06, "loss": 1.1173, "step": 27572 }, { "epoch": 2.1608934169278995, "grad_norm": 4.367411136627197, "learning_rate": 4.3839204472675386e-06, "loss": 1.1564, "step": 27573 }, { "epoch": 2.160971786833856, "grad_norm": 3.189002275466919, "learning_rate": 4.383614436134545e-06, "loss": 1.5131, "step": 27574 }, { "epoch": 2.1610501567398117, "grad_norm": 4.363739967346191, "learning_rate": 4.383308427345968e-06, "loss": 1.5882, "step": 27575 }, { "epoch": 2.161128526645768, "grad_norm": 3.374479293823242, "learning_rate": 4.383002420902969e-06, "loss": 1.646, "step": 27576 }, { "epoch": 2.161206896551724, "grad_norm": 5.147483825683594, "learning_rate": 4.3826964168067134e-06, "loss": 1.3085, "step": 27577 }, { "epoch": 2.1612852664576803, "grad_norm": 6.3990254402160645, "learning_rate": 4.382390415058362e-06, "loss": 1.4971, "step": 27578 }, { "epoch": 2.161363636363636, "grad_norm": 3.2318074703216553, "learning_rate": 4.382084415659078e-06, "loss": 1.8408, "step": 27579 }, { "epoch": 2.1614420062695925, "grad_norm": 3.814997911453247, "learning_rate": 4.381778418610032e-06, "loss": 1.3632, "step": 27580 }, { "epoch": 2.161520376175549, "grad_norm": 2.139472246170044, "learning_rate": 4.38147242391238e-06, "loss": 1.3715, "step": 27581 }, { "epoch": 2.1615987460815047, "grad_norm": 3.789217233657837, "learning_rate": 4.38116643156729e-06, "loss": 1.6302, "step": 27582 }, { "epoch": 2.1616771159874606, "grad_norm": 4.7105255126953125, "learning_rate": 4.380860441575925e-06, "loss": 1.8041, "step": 27583 }, { "epoch": 2.161755485893417, "grad_norm": 12.848626136779785, "learning_rate": 4.380554453939449e-06, "loss": 1.4154, "step": 27584 }, { "epoch": 2.1618338557993733, "grad_norm": 2.185610294342041, "learning_rate": 4.3802484686590264e-06, "loss": 1.4003, "step": 27585 }, { "epoch": 2.161912225705329, "grad_norm": 2.734227180480957, "learning_rate": 4.379942485735818e-06, "loss": 1.4799, "step": 27586 }, { "epoch": 2.161990595611285, "grad_norm": 3.3824923038482666, "learning_rate": 4.379636505170992e-06, "loss": 1.4215, "step": 27587 }, { "epoch": 2.1620689655172414, "grad_norm": 8.352127075195312, "learning_rate": 4.3793305269657104e-06, "loss": 0.917, "step": 27588 }, { "epoch": 2.1621473354231977, "grad_norm": 4.5059614181518555, "learning_rate": 4.379024551121136e-06, "loss": 1.287, "step": 27589 }, { "epoch": 2.1622257053291536, "grad_norm": 2.4602935314178467, "learning_rate": 4.3787185776384325e-06, "loss": 1.195, "step": 27590 }, { "epoch": 2.16230407523511, "grad_norm": 2.637622117996216, "learning_rate": 4.378412606518764e-06, "loss": 1.3266, "step": 27591 }, { "epoch": 2.162382445141066, "grad_norm": 2.27791690826416, "learning_rate": 4.378106637763298e-06, "loss": 1.4681, "step": 27592 }, { "epoch": 2.162460815047022, "grad_norm": 4.093493461608887, "learning_rate": 4.377800671373192e-06, "loss": 1.646, "step": 27593 }, { "epoch": 2.162539184952978, "grad_norm": 2.3845155239105225, "learning_rate": 4.377494707349613e-06, "loss": 1.2957, "step": 27594 }, { "epoch": 2.1626175548589344, "grad_norm": 4.121951580047607, "learning_rate": 4.377188745693725e-06, "loss": 0.9144, "step": 27595 }, { "epoch": 2.1626959247648903, "grad_norm": 3.154675006866455, "learning_rate": 4.3768827864066914e-06, "loss": 1.3665, "step": 27596 }, { "epoch": 2.1627742946708466, "grad_norm": 2.6085920333862305, "learning_rate": 4.376576829489675e-06, "loss": 0.8879, "step": 27597 }, { "epoch": 2.1628526645768025, "grad_norm": 2.3987298011779785, "learning_rate": 4.3762708749438394e-06, "loss": 1.5035, "step": 27598 }, { "epoch": 2.162931034482759, "grad_norm": 6.637810707092285, "learning_rate": 4.375964922770352e-06, "loss": 1.3075, "step": 27599 }, { "epoch": 2.1630094043887147, "grad_norm": 3.1120359897613525, "learning_rate": 4.3756589729703695e-06, "loss": 1.155, "step": 27600 }, { "epoch": 2.163087774294671, "grad_norm": 4.173479080200195, "learning_rate": 4.375353025545061e-06, "loss": 1.8642, "step": 27601 }, { "epoch": 2.163166144200627, "grad_norm": 3.221776247024536, "learning_rate": 4.37504708049559e-06, "loss": 1.351, "step": 27602 }, { "epoch": 2.1632445141065832, "grad_norm": 3.191642999649048, "learning_rate": 4.374741137823119e-06, "loss": 1.5674, "step": 27603 }, { "epoch": 2.163322884012539, "grad_norm": 3.063002824783325, "learning_rate": 4.37443519752881e-06, "loss": 1.381, "step": 27604 }, { "epoch": 2.1634012539184955, "grad_norm": 6.184494495391846, "learning_rate": 4.3741292596138275e-06, "loss": 1.5855, "step": 27605 }, { "epoch": 2.1634796238244514, "grad_norm": 7.035276412963867, "learning_rate": 4.373823324079338e-06, "loss": 1.2607, "step": 27606 }, { "epoch": 2.1635579937304077, "grad_norm": 3.4302377700805664, "learning_rate": 4.373517390926503e-06, "loss": 1.7244, "step": 27607 }, { "epoch": 2.1636363636363636, "grad_norm": 3.3201143741607666, "learning_rate": 4.3732114601564855e-06, "loss": 1.3552, "step": 27608 }, { "epoch": 2.16371473354232, "grad_norm": 3.1179800033569336, "learning_rate": 4.372905531770449e-06, "loss": 1.3995, "step": 27609 }, { "epoch": 2.163793103448276, "grad_norm": 2.929522752761841, "learning_rate": 4.372599605769559e-06, "loss": 1.28, "step": 27610 }, { "epoch": 2.163871473354232, "grad_norm": 2.7302019596099854, "learning_rate": 4.372293682154977e-06, "loss": 1.2746, "step": 27611 }, { "epoch": 2.163949843260188, "grad_norm": 3.4311420917510986, "learning_rate": 4.371987760927866e-06, "loss": 1.0943, "step": 27612 }, { "epoch": 2.1640282131661444, "grad_norm": 2.396306037902832, "learning_rate": 4.3716818420893925e-06, "loss": 1.2804, "step": 27613 }, { "epoch": 2.1641065830721002, "grad_norm": 5.234736919403076, "learning_rate": 4.371375925640719e-06, "loss": 1.1948, "step": 27614 }, { "epoch": 2.1641849529780566, "grad_norm": 3.5306406021118164, "learning_rate": 4.371070011583008e-06, "loss": 1.2798, "step": 27615 }, { "epoch": 2.1642633228840125, "grad_norm": 3.4358510971069336, "learning_rate": 4.370764099917424e-06, "loss": 1.6705, "step": 27616 }, { "epoch": 2.164341692789969, "grad_norm": 2.4117074012756348, "learning_rate": 4.370458190645128e-06, "loss": 1.3708, "step": 27617 }, { "epoch": 2.1644200626959247, "grad_norm": 4.346085071563721, "learning_rate": 4.37015228376729e-06, "loss": 1.1799, "step": 27618 }, { "epoch": 2.164498432601881, "grad_norm": 3.974149703979492, "learning_rate": 4.369846379285065e-06, "loss": 1.3517, "step": 27619 }, { "epoch": 2.164576802507837, "grad_norm": 3.1620028018951416, "learning_rate": 4.369540477199623e-06, "loss": 1.3498, "step": 27620 }, { "epoch": 2.1646551724137932, "grad_norm": 2.0666663646698, "learning_rate": 4.3692345775121254e-06, "loss": 1.0717, "step": 27621 }, { "epoch": 2.164733542319749, "grad_norm": 27.665067672729492, "learning_rate": 4.368928680223735e-06, "loss": 1.2499, "step": 27622 }, { "epoch": 2.1648119122257055, "grad_norm": 13.208542823791504, "learning_rate": 4.368622785335615e-06, "loss": 1.2984, "step": 27623 }, { "epoch": 2.1648902821316613, "grad_norm": 3.7873728275299072, "learning_rate": 4.368316892848929e-06, "loss": 1.3824, "step": 27624 }, { "epoch": 2.1649686520376177, "grad_norm": 2.5249147415161133, "learning_rate": 4.368011002764844e-06, "loss": 1.5107, "step": 27625 }, { "epoch": 2.1650470219435736, "grad_norm": 2.3760452270507812, "learning_rate": 4.3677051150845175e-06, "loss": 1.5428, "step": 27626 }, { "epoch": 2.16512539184953, "grad_norm": 2.983670234680176, "learning_rate": 4.367399229809118e-06, "loss": 1.2536, "step": 27627 }, { "epoch": 2.165203761755486, "grad_norm": 2.7978949546813965, "learning_rate": 4.367093346939805e-06, "loss": 1.3551, "step": 27628 }, { "epoch": 2.165282131661442, "grad_norm": 2.6883506774902344, "learning_rate": 4.366787466477746e-06, "loss": 1.3133, "step": 27629 }, { "epoch": 2.165360501567398, "grad_norm": 3.5592188835144043, "learning_rate": 4.366481588424101e-06, "loss": 1.5725, "step": 27630 }, { "epoch": 2.1654388714733543, "grad_norm": 2.8444745540618896, "learning_rate": 4.366175712780034e-06, "loss": 1.3683, "step": 27631 }, { "epoch": 2.1655172413793102, "grad_norm": 3.1276280879974365, "learning_rate": 4.365869839546709e-06, "loss": 1.7504, "step": 27632 }, { "epoch": 2.1655956112852666, "grad_norm": 4.916653633117676, "learning_rate": 4.365563968725292e-06, "loss": 0.8104, "step": 27633 }, { "epoch": 2.1656739811912225, "grad_norm": 4.282370090484619, "learning_rate": 4.365258100316942e-06, "loss": 1.4062, "step": 27634 }, { "epoch": 2.165752351097179, "grad_norm": 12.7926025390625, "learning_rate": 4.3649522343228245e-06, "loss": 1.5331, "step": 27635 }, { "epoch": 2.1658307210031347, "grad_norm": 2.124067544937134, "learning_rate": 4.364646370744103e-06, "loss": 1.3613, "step": 27636 }, { "epoch": 2.165909090909091, "grad_norm": 2.6997623443603516, "learning_rate": 4.36434050958194e-06, "loss": 1.4403, "step": 27637 }, { "epoch": 2.165987460815047, "grad_norm": 3.0826315879821777, "learning_rate": 4.364034650837497e-06, "loss": 0.9676, "step": 27638 }, { "epoch": 2.1660658307210032, "grad_norm": 2.3369104862213135, "learning_rate": 4.363728794511941e-06, "loss": 1.2104, "step": 27639 }, { "epoch": 2.166144200626959, "grad_norm": 7.900994777679443, "learning_rate": 4.363422940606435e-06, "loss": 1.6362, "step": 27640 }, { "epoch": 2.1662225705329154, "grad_norm": 3.1076912879943848, "learning_rate": 4.363117089122141e-06, "loss": 1.7077, "step": 27641 }, { "epoch": 2.1663009404388713, "grad_norm": 6.607662200927734, "learning_rate": 4.362811240060222e-06, "loss": 1.5203, "step": 27642 }, { "epoch": 2.1663793103448277, "grad_norm": 5.045599460601807, "learning_rate": 4.36250539342184e-06, "loss": 1.0033, "step": 27643 }, { "epoch": 2.1664576802507836, "grad_norm": 6.101127624511719, "learning_rate": 4.362199549208164e-06, "loss": 1.4251, "step": 27644 }, { "epoch": 2.16653605015674, "grad_norm": 10.260812759399414, "learning_rate": 4.361893707420349e-06, "loss": 1.8248, "step": 27645 }, { "epoch": 2.166614420062696, "grad_norm": 2.919785976409912, "learning_rate": 4.361587868059565e-06, "loss": 1.0591, "step": 27646 }, { "epoch": 2.166692789968652, "grad_norm": 2.118713617324829, "learning_rate": 4.361282031126972e-06, "loss": 0.9078, "step": 27647 }, { "epoch": 2.166771159874608, "grad_norm": 7.269959926605225, "learning_rate": 4.360976196623736e-06, "loss": 1.2721, "step": 27648 }, { "epoch": 2.1668495297805643, "grad_norm": 2.6894500255584717, "learning_rate": 4.360670364551017e-06, "loss": 1.2296, "step": 27649 }, { "epoch": 2.16692789968652, "grad_norm": 4.556245803833008, "learning_rate": 4.360364534909978e-06, "loss": 0.9007, "step": 27650 }, { "epoch": 2.1670062695924766, "grad_norm": 2.5478193759918213, "learning_rate": 4.360058707701786e-06, "loss": 1.2799, "step": 27651 }, { "epoch": 2.1670846394984324, "grad_norm": 3.780050754547119, "learning_rate": 4.359752882927601e-06, "loss": 1.5054, "step": 27652 }, { "epoch": 2.1671630094043888, "grad_norm": 2.287177562713623, "learning_rate": 4.359447060588587e-06, "loss": 1.1464, "step": 27653 }, { "epoch": 2.1672413793103447, "grad_norm": 5.954824924468994, "learning_rate": 4.359141240685908e-06, "loss": 1.1588, "step": 27654 }, { "epoch": 2.167319749216301, "grad_norm": 8.866533279418945, "learning_rate": 4.3588354232207266e-06, "loss": 1.2059, "step": 27655 }, { "epoch": 2.167398119122257, "grad_norm": 2.0699782371520996, "learning_rate": 4.358529608194205e-06, "loss": 1.1379, "step": 27656 }, { "epoch": 2.167476489028213, "grad_norm": 2.607382297515869, "learning_rate": 4.358223795607507e-06, "loss": 1.2977, "step": 27657 }, { "epoch": 2.167554858934169, "grad_norm": 3.0418460369110107, "learning_rate": 4.357917985461798e-06, "loss": 1.4308, "step": 27658 }, { "epoch": 2.1676332288401254, "grad_norm": 2.540493965148926, "learning_rate": 4.3576121777582385e-06, "loss": 1.3025, "step": 27659 }, { "epoch": 2.1677115987460813, "grad_norm": 8.982221603393555, "learning_rate": 4.357306372497992e-06, "loss": 0.6511, "step": 27660 }, { "epoch": 2.1677899686520377, "grad_norm": 2.792203187942505, "learning_rate": 4.357000569682221e-06, "loss": 1.3261, "step": 27661 }, { "epoch": 2.1678683385579935, "grad_norm": 4.137444019317627, "learning_rate": 4.356694769312089e-06, "loss": 1.6361, "step": 27662 }, { "epoch": 2.16794670846395, "grad_norm": 3.7183384895324707, "learning_rate": 4.356388971388763e-06, "loss": 1.985, "step": 27663 }, { "epoch": 2.1680250783699058, "grad_norm": 3.012207508087158, "learning_rate": 4.356083175913401e-06, "loss": 1.2418, "step": 27664 }, { "epoch": 2.168103448275862, "grad_norm": 2.385422706604004, "learning_rate": 4.355777382887167e-06, "loss": 1.2089, "step": 27665 }, { "epoch": 2.168181818181818, "grad_norm": 3.618751287460327, "learning_rate": 4.3554715923112265e-06, "loss": 1.2167, "step": 27666 }, { "epoch": 2.1682601880877743, "grad_norm": 4.218583106994629, "learning_rate": 4.3551658041867406e-06, "loss": 1.4831, "step": 27667 }, { "epoch": 2.16833855799373, "grad_norm": 2.181687831878662, "learning_rate": 4.3548600185148716e-06, "loss": 1.5967, "step": 27668 }, { "epoch": 2.1684169278996865, "grad_norm": 3.3273351192474365, "learning_rate": 4.354554235296784e-06, "loss": 0.9054, "step": 27669 }, { "epoch": 2.1684952978056424, "grad_norm": 3.0204622745513916, "learning_rate": 4.354248454533642e-06, "loss": 1.6263, "step": 27670 }, { "epoch": 2.1685736677115988, "grad_norm": 2.3835883140563965, "learning_rate": 4.353942676226606e-06, "loss": 1.4263, "step": 27671 }, { "epoch": 2.1686520376175547, "grad_norm": 3.7040700912475586, "learning_rate": 4.3536369003768404e-06, "loss": 1.4837, "step": 27672 }, { "epoch": 2.168730407523511, "grad_norm": 3.561389207839966, "learning_rate": 4.3533311269855075e-06, "loss": 1.8448, "step": 27673 }, { "epoch": 2.168808777429467, "grad_norm": 4.118878364562988, "learning_rate": 4.353025356053774e-06, "loss": 1.64, "step": 27674 }, { "epoch": 2.168887147335423, "grad_norm": 2.7167513370513916, "learning_rate": 4.352719587582796e-06, "loss": 1.5658, "step": 27675 }, { "epoch": 2.1689655172413795, "grad_norm": 2.365746259689331, "learning_rate": 4.35241382157374e-06, "loss": 1.4003, "step": 27676 }, { "epoch": 2.1690438871473354, "grad_norm": 2.893995761871338, "learning_rate": 4.352108058027772e-06, "loss": 1.1963, "step": 27677 }, { "epoch": 2.1691222570532913, "grad_norm": 3.5337021350860596, "learning_rate": 4.351802296946051e-06, "loss": 0.9491, "step": 27678 }, { "epoch": 2.1692006269592476, "grad_norm": 3.7009782791137695, "learning_rate": 4.351496538329741e-06, "loss": 1.4943, "step": 27679 }, { "epoch": 2.169278996865204, "grad_norm": 3.329692840576172, "learning_rate": 4.3511907821800036e-06, "loss": 1.1021, "step": 27680 }, { "epoch": 2.16935736677116, "grad_norm": 3.030017375946045, "learning_rate": 4.350885028498007e-06, "loss": 1.3151, "step": 27681 }, { "epoch": 2.1694357366771158, "grad_norm": 7.211816787719727, "learning_rate": 4.350579277284907e-06, "loss": 1.2983, "step": 27682 }, { "epoch": 2.169514106583072, "grad_norm": 4.1363701820373535, "learning_rate": 4.35027352854187e-06, "loss": 1.4033, "step": 27683 }, { "epoch": 2.1695924764890284, "grad_norm": 3.513789176940918, "learning_rate": 4.34996778227006e-06, "loss": 1.4801, "step": 27684 }, { "epoch": 2.1696708463949843, "grad_norm": 2.6348373889923096, "learning_rate": 4.349662038470639e-06, "loss": 1.5357, "step": 27685 }, { "epoch": 2.1697492163009406, "grad_norm": 2.4778409004211426, "learning_rate": 4.3493562971447685e-06, "loss": 1.1692, "step": 27686 }, { "epoch": 2.1698275862068965, "grad_norm": 3.9239070415496826, "learning_rate": 4.349050558293611e-06, "loss": 1.6585, "step": 27687 }, { "epoch": 2.169905956112853, "grad_norm": 3.345153331756592, "learning_rate": 4.348744821918333e-06, "loss": 1.2532, "step": 27688 }, { "epoch": 2.1699843260188088, "grad_norm": 8.055427551269531, "learning_rate": 4.348439088020096e-06, "loss": 0.9978, "step": 27689 }, { "epoch": 2.170062695924765, "grad_norm": 3.5353968143463135, "learning_rate": 4.348133356600061e-06, "loss": 1.1446, "step": 27690 }, { "epoch": 2.170141065830721, "grad_norm": 4.817378520965576, "learning_rate": 4.347827627659392e-06, "loss": 1.3079, "step": 27691 }, { "epoch": 2.1702194357366773, "grad_norm": 2.577181577682495, "learning_rate": 4.347521901199252e-06, "loss": 1.3167, "step": 27692 }, { "epoch": 2.170297805642633, "grad_norm": 2.791668176651001, "learning_rate": 4.347216177220803e-06, "loss": 0.9699, "step": 27693 }, { "epoch": 2.1703761755485895, "grad_norm": 4.332771301269531, "learning_rate": 4.346910455725207e-06, "loss": 1.7755, "step": 27694 }, { "epoch": 2.1704545454545454, "grad_norm": 18.824434280395508, "learning_rate": 4.34660473671363e-06, "loss": 1.7782, "step": 27695 }, { "epoch": 2.1705329153605017, "grad_norm": 2.588918924331665, "learning_rate": 4.346299020187234e-06, "loss": 1.1063, "step": 27696 }, { "epoch": 2.1706112852664576, "grad_norm": 2.9301228523254395, "learning_rate": 4.345993306147178e-06, "loss": 1.2321, "step": 27697 }, { "epoch": 2.170689655172414, "grad_norm": 4.02938985824585, "learning_rate": 4.345687594594629e-06, "loss": 1.47, "step": 27698 }, { "epoch": 2.17076802507837, "grad_norm": 2.5970723628997803, "learning_rate": 4.345381885530748e-06, "loss": 1.1255, "step": 27699 }, { "epoch": 2.170846394984326, "grad_norm": 6.615966796875, "learning_rate": 4.3450761789567e-06, "loss": 1.2295, "step": 27700 }, { "epoch": 2.170924764890282, "grad_norm": 2.1478068828582764, "learning_rate": 4.344770474873642e-06, "loss": 1.3032, "step": 27701 }, { "epoch": 2.1710031347962384, "grad_norm": 2.5601754188537598, "learning_rate": 4.3444647732827435e-06, "loss": 1.2021, "step": 27702 }, { "epoch": 2.1710815047021943, "grad_norm": 5.424239158630371, "learning_rate": 4.344159074185163e-06, "loss": 1.2093, "step": 27703 }, { "epoch": 2.1711598746081506, "grad_norm": 3.117034435272217, "learning_rate": 4.343853377582066e-06, "loss": 1.1958, "step": 27704 }, { "epoch": 2.1712382445141065, "grad_norm": 3.8268380165100098, "learning_rate": 4.343547683474612e-06, "loss": 1.3642, "step": 27705 }, { "epoch": 2.171316614420063, "grad_norm": 3.0004889965057373, "learning_rate": 4.343241991863966e-06, "loss": 1.7201, "step": 27706 }, { "epoch": 2.1713949843260187, "grad_norm": 1.9251680374145508, "learning_rate": 4.342936302751292e-06, "loss": 0.7629, "step": 27707 }, { "epoch": 2.171473354231975, "grad_norm": 3.400557518005371, "learning_rate": 4.342630616137748e-06, "loss": 1.4432, "step": 27708 }, { "epoch": 2.171551724137931, "grad_norm": 3.3234238624572754, "learning_rate": 4.3423249320245006e-06, "loss": 1.4357, "step": 27709 }, { "epoch": 2.1716300940438873, "grad_norm": 2.5257456302642822, "learning_rate": 4.342019250412711e-06, "loss": 1.2235, "step": 27710 }, { "epoch": 2.171708463949843, "grad_norm": 4.388071060180664, "learning_rate": 4.341713571303543e-06, "loss": 1.0971, "step": 27711 }, { "epoch": 2.1717868338557995, "grad_norm": 4.224931716918945, "learning_rate": 4.341407894698159e-06, "loss": 1.253, "step": 27712 }, { "epoch": 2.1718652037617554, "grad_norm": 3.3273940086364746, "learning_rate": 4.341102220597718e-06, "loss": 1.1543, "step": 27713 }, { "epoch": 2.1719435736677117, "grad_norm": 4.910303115844727, "learning_rate": 4.340796549003388e-06, "loss": 0.9318, "step": 27714 }, { "epoch": 2.1720219435736676, "grad_norm": 2.311967611312866, "learning_rate": 4.34049087991633e-06, "loss": 1.2097, "step": 27715 }, { "epoch": 2.172100313479624, "grad_norm": 2.5709822177886963, "learning_rate": 4.340185213337705e-06, "loss": 1.6948, "step": 27716 }, { "epoch": 2.17217868338558, "grad_norm": 2.2589542865753174, "learning_rate": 4.339879549268677e-06, "loss": 1.3409, "step": 27717 }, { "epoch": 2.172257053291536, "grad_norm": 2.618781566619873, "learning_rate": 4.339573887710406e-06, "loss": 1.3274, "step": 27718 }, { "epoch": 2.172335423197492, "grad_norm": 4.231698036193848, "learning_rate": 4.339268228664061e-06, "loss": 0.8287, "step": 27719 }, { "epoch": 2.1724137931034484, "grad_norm": 3.0797336101531982, "learning_rate": 4.338962572130797e-06, "loss": 1.4337, "step": 27720 }, { "epoch": 2.1724921630094043, "grad_norm": 4.739078044891357, "learning_rate": 4.33865691811178e-06, "loss": 1.6763, "step": 27721 }, { "epoch": 2.1725705329153606, "grad_norm": 6.744109153747559, "learning_rate": 4.338351266608174e-06, "loss": 1.7235, "step": 27722 }, { "epoch": 2.1726489028213165, "grad_norm": 2.336494207382202, "learning_rate": 4.338045617621139e-06, "loss": 1.315, "step": 27723 }, { "epoch": 2.172727272727273, "grad_norm": 3.2462780475616455, "learning_rate": 4.3377399711518385e-06, "loss": 1.4278, "step": 27724 }, { "epoch": 2.1728056426332287, "grad_norm": 2.17740535736084, "learning_rate": 4.337434327201434e-06, "loss": 1.0504, "step": 27725 }, { "epoch": 2.172884012539185, "grad_norm": 4.132634162902832, "learning_rate": 4.3371286857710915e-06, "loss": 1.5601, "step": 27726 }, { "epoch": 2.172962382445141, "grad_norm": 14.774516105651855, "learning_rate": 4.336823046861968e-06, "loss": 0.925, "step": 27727 }, { "epoch": 2.1730407523510973, "grad_norm": 2.948326349258423, "learning_rate": 4.3365174104752305e-06, "loss": 1.2602, "step": 27728 }, { "epoch": 2.173119122257053, "grad_norm": 3.5107979774475098, "learning_rate": 4.33621177661204e-06, "loss": 1.3378, "step": 27729 }, { "epoch": 2.1731974921630095, "grad_norm": 2.16804575920105, "learning_rate": 4.33590614527356e-06, "loss": 1.3362, "step": 27730 }, { "epoch": 2.1732758620689654, "grad_norm": 4.059310436248779, "learning_rate": 4.3356005164609505e-06, "loss": 1.6371, "step": 27731 }, { "epoch": 2.1733542319749217, "grad_norm": 3.7249858379364014, "learning_rate": 4.335294890175375e-06, "loss": 1.3389, "step": 27732 }, { "epoch": 2.1734326018808776, "grad_norm": 3.8931798934936523, "learning_rate": 4.334989266417999e-06, "loss": 1.2099, "step": 27733 }, { "epoch": 2.173510971786834, "grad_norm": 2.8821845054626465, "learning_rate": 4.3346836451899785e-06, "loss": 1.2888, "step": 27734 }, { "epoch": 2.17358934169279, "grad_norm": 5.069382667541504, "learning_rate": 4.334378026492482e-06, "loss": 0.9341, "step": 27735 }, { "epoch": 2.173667711598746, "grad_norm": 23.98382568359375, "learning_rate": 4.334072410326669e-06, "loss": 0.7561, "step": 27736 }, { "epoch": 2.173746081504702, "grad_norm": 3.0418033599853516, "learning_rate": 4.333766796693703e-06, "loss": 1.1283, "step": 27737 }, { "epoch": 2.1738244514106584, "grad_norm": 2.777557134628296, "learning_rate": 4.333461185594745e-06, "loss": 1.0623, "step": 27738 }, { "epoch": 2.1739028213166143, "grad_norm": 6.061855792999268, "learning_rate": 4.333155577030959e-06, "loss": 1.1953, "step": 27739 }, { "epoch": 2.1739811912225706, "grad_norm": 3.43369460105896, "learning_rate": 4.332849971003505e-06, "loss": 1.1227, "step": 27740 }, { "epoch": 2.1740595611285265, "grad_norm": 2.832080125808716, "learning_rate": 4.33254436751355e-06, "loss": 1.117, "step": 27741 }, { "epoch": 2.174137931034483, "grad_norm": 3.4747095108032227, "learning_rate": 4.332238766562252e-06, "loss": 1.1839, "step": 27742 }, { "epoch": 2.1742163009404387, "grad_norm": 3.6881496906280518, "learning_rate": 4.331933168150774e-06, "loss": 1.4356, "step": 27743 }, { "epoch": 2.174294670846395, "grad_norm": 3.5329132080078125, "learning_rate": 4.331627572280279e-06, "loss": 0.9073, "step": 27744 }, { "epoch": 2.174373040752351, "grad_norm": 3.5470168590545654, "learning_rate": 4.331321978951932e-06, "loss": 1.227, "step": 27745 }, { "epoch": 2.1744514106583073, "grad_norm": 3.361105442047119, "learning_rate": 4.3310163881668895e-06, "loss": 1.2995, "step": 27746 }, { "epoch": 2.174529780564263, "grad_norm": 7.033664226531982, "learning_rate": 4.330710799926318e-06, "loss": 1.185, "step": 27747 }, { "epoch": 2.1746081504702195, "grad_norm": 11.277018547058105, "learning_rate": 4.330405214231381e-06, "loss": 1.2675, "step": 27748 }, { "epoch": 2.1746865203761754, "grad_norm": 6.794802188873291, "learning_rate": 4.3300996310832365e-06, "loss": 1.4704, "step": 27749 }, { "epoch": 2.1747648902821317, "grad_norm": 4.832025051116943, "learning_rate": 4.329794050483049e-06, "loss": 1.4485, "step": 27750 }, { "epoch": 2.1748432601880876, "grad_norm": 2.6407370567321777, "learning_rate": 4.32948847243198e-06, "loss": 0.9289, "step": 27751 }, { "epoch": 2.174921630094044, "grad_norm": 7.377638816833496, "learning_rate": 4.329182896931195e-06, "loss": 1.7074, "step": 27752 }, { "epoch": 2.175, "grad_norm": 2.541358470916748, "learning_rate": 4.328877323981852e-06, "loss": 1.4068, "step": 27753 }, { "epoch": 2.175078369905956, "grad_norm": 2.7001888751983643, "learning_rate": 4.328571753585115e-06, "loss": 1.1678, "step": 27754 }, { "epoch": 2.175156739811912, "grad_norm": 3.72640323638916, "learning_rate": 4.328266185742146e-06, "loss": 1.1875, "step": 27755 }, { "epoch": 2.1752351097178684, "grad_norm": 3.2210052013397217, "learning_rate": 4.327960620454109e-06, "loss": 1.5452, "step": 27756 }, { "epoch": 2.1753134796238243, "grad_norm": 3.4441914558410645, "learning_rate": 4.327655057722163e-06, "loss": 1.1605, "step": 27757 }, { "epoch": 2.1753918495297806, "grad_norm": 3.9069106578826904, "learning_rate": 4.327349497547471e-06, "loss": 1.4895, "step": 27758 }, { "epoch": 2.1754702194357365, "grad_norm": 4.60902738571167, "learning_rate": 4.327043939931197e-06, "loss": 1.8022, "step": 27759 }, { "epoch": 2.175548589341693, "grad_norm": 2.737403631210327, "learning_rate": 4.326738384874504e-06, "loss": 1.2342, "step": 27760 }, { "epoch": 2.1756269592476487, "grad_norm": 1.9441184997558594, "learning_rate": 4.32643283237855e-06, "loss": 1.1806, "step": 27761 }, { "epoch": 2.175705329153605, "grad_norm": 10.139679908752441, "learning_rate": 4.3261272824445005e-06, "loss": 1.1644, "step": 27762 }, { "epoch": 2.175783699059561, "grad_norm": 3.7375600337982178, "learning_rate": 4.325821735073517e-06, "loss": 1.5761, "step": 27763 }, { "epoch": 2.1758620689655173, "grad_norm": 3.3910975456237793, "learning_rate": 4.3255161902667595e-06, "loss": 1.4728, "step": 27764 }, { "epoch": 2.175940438871473, "grad_norm": 3.4001801013946533, "learning_rate": 4.325210648025393e-06, "loss": 1.567, "step": 27765 }, { "epoch": 2.1760188087774295, "grad_norm": 3.947530508041382, "learning_rate": 4.3249051083505795e-06, "loss": 1.5411, "step": 27766 }, { "epoch": 2.176097178683386, "grad_norm": 4.6239776611328125, "learning_rate": 4.324599571243481e-06, "loss": 1.4574, "step": 27767 }, { "epoch": 2.1761755485893417, "grad_norm": 2.720879554748535, "learning_rate": 4.324294036705257e-06, "loss": 1.2236, "step": 27768 }, { "epoch": 2.1762539184952976, "grad_norm": 3.0198864936828613, "learning_rate": 4.3239885047370705e-06, "loss": 1.1045, "step": 27769 }, { "epoch": 2.176332288401254, "grad_norm": 3.3166987895965576, "learning_rate": 4.323682975340086e-06, "loss": 1.5032, "step": 27770 }, { "epoch": 2.1764106583072103, "grad_norm": 1.9425691366195679, "learning_rate": 4.323377448515465e-06, "loss": 1.422, "step": 27771 }, { "epoch": 2.176489028213166, "grad_norm": 7.63191556930542, "learning_rate": 4.323071924264368e-06, "loss": 1.2686, "step": 27772 }, { "epoch": 2.176567398119122, "grad_norm": 4.837769985198975, "learning_rate": 4.322766402587957e-06, "loss": 1.1645, "step": 27773 }, { "epoch": 2.1766457680250784, "grad_norm": 3.5377063751220703, "learning_rate": 4.322460883487396e-06, "loss": 1.4343, "step": 27774 }, { "epoch": 2.1767241379310347, "grad_norm": 6.808014869689941, "learning_rate": 4.322155366963845e-06, "loss": 1.8573, "step": 27775 }, { "epoch": 2.1768025078369906, "grad_norm": 6.085888385772705, "learning_rate": 4.321849853018465e-06, "loss": 1.7878, "step": 27776 }, { "epoch": 2.1768808777429465, "grad_norm": 3.3484840393066406, "learning_rate": 4.321544341652422e-06, "loss": 1.8333, "step": 27777 }, { "epoch": 2.176959247648903, "grad_norm": 4.320826053619385, "learning_rate": 4.321238832866876e-06, "loss": 1.3733, "step": 27778 }, { "epoch": 2.177037617554859, "grad_norm": 3.0748093128204346, "learning_rate": 4.3209333266629875e-06, "loss": 1.0178, "step": 27779 }, { "epoch": 2.177115987460815, "grad_norm": 3.4460337162017822, "learning_rate": 4.320627823041921e-06, "loss": 1.6031, "step": 27780 }, { "epoch": 2.1771943573667714, "grad_norm": 2.798994541168213, "learning_rate": 4.320322322004836e-06, "loss": 1.1351, "step": 27781 }, { "epoch": 2.1772727272727272, "grad_norm": 2.5803170204162598, "learning_rate": 4.320016823552898e-06, "loss": 1.2798, "step": 27782 }, { "epoch": 2.1773510971786836, "grad_norm": 2.43473482131958, "learning_rate": 4.319711327687264e-06, "loss": 1.2655, "step": 27783 }, { "epoch": 2.1774294670846395, "grad_norm": 3.52388072013855, "learning_rate": 4.3194058344091e-06, "loss": 1.3452, "step": 27784 }, { "epoch": 2.177507836990596, "grad_norm": 2.742385149002075, "learning_rate": 4.319100343719567e-06, "loss": 1.4964, "step": 27785 }, { "epoch": 2.1775862068965517, "grad_norm": 3.365485429763794, "learning_rate": 4.318794855619827e-06, "loss": 1.5502, "step": 27786 }, { "epoch": 2.177664576802508, "grad_norm": 4.700692176818848, "learning_rate": 4.3184893701110405e-06, "loss": 1.8462, "step": 27787 }, { "epoch": 2.177742946708464, "grad_norm": 10.81640625, "learning_rate": 4.31818388719437e-06, "loss": 1.6353, "step": 27788 }, { "epoch": 2.1778213166144202, "grad_norm": 3.222276449203491, "learning_rate": 4.317878406870981e-06, "loss": 1.3624, "step": 27789 }, { "epoch": 2.177899686520376, "grad_norm": 4.791630268096924, "learning_rate": 4.3175729291420274e-06, "loss": 1.0146, "step": 27790 }, { "epoch": 2.1779780564263325, "grad_norm": 4.5644965171813965, "learning_rate": 4.317267454008679e-06, "loss": 1.6822, "step": 27791 }, { "epoch": 2.1780564263322884, "grad_norm": 4.034019470214844, "learning_rate": 4.316961981472093e-06, "loss": 0.336, "step": 27792 }, { "epoch": 2.1781347962382447, "grad_norm": 2.5939502716064453, "learning_rate": 4.316656511533434e-06, "loss": 1.2304, "step": 27793 }, { "epoch": 2.1782131661442006, "grad_norm": 4.0191755294799805, "learning_rate": 4.316351044193862e-06, "loss": 1.2765, "step": 27794 }, { "epoch": 2.178291536050157, "grad_norm": 4.104156494140625, "learning_rate": 4.316045579454538e-06, "loss": 1.8718, "step": 27795 }, { "epoch": 2.178369905956113, "grad_norm": 3.028002977371216, "learning_rate": 4.3157401173166275e-06, "loss": 1.4344, "step": 27796 }, { "epoch": 2.178448275862069, "grad_norm": 5.114626407623291, "learning_rate": 4.31543465778129e-06, "loss": 1.3497, "step": 27797 }, { "epoch": 2.178526645768025, "grad_norm": 2.6934268474578857, "learning_rate": 4.315129200849687e-06, "loss": 1.604, "step": 27798 }, { "epoch": 2.1786050156739813, "grad_norm": 7.659353733062744, "learning_rate": 4.314823746522981e-06, "loss": 1.3861, "step": 27799 }, { "epoch": 2.1786833855799372, "grad_norm": 5.515778541564941, "learning_rate": 4.314518294802332e-06, "loss": 1.8592, "step": 27800 }, { "epoch": 2.1787617554858936, "grad_norm": 2.8581254482269287, "learning_rate": 4.314212845688906e-06, "loss": 1.2235, "step": 27801 }, { "epoch": 2.1788401253918495, "grad_norm": 3.0409865379333496, "learning_rate": 4.31390739918386e-06, "loss": 1.0228, "step": 27802 }, { "epoch": 2.178918495297806, "grad_norm": 2.9445009231567383, "learning_rate": 4.313601955288358e-06, "loss": 1.1211, "step": 27803 }, { "epoch": 2.1789968652037617, "grad_norm": 9.707901954650879, "learning_rate": 4.313296514003563e-06, "loss": 0.8062, "step": 27804 }, { "epoch": 2.179075235109718, "grad_norm": 5.001545429229736, "learning_rate": 4.312991075330635e-06, "loss": 1.3221, "step": 27805 }, { "epoch": 2.179153605015674, "grad_norm": 8.416911125183105, "learning_rate": 4.312685639270735e-06, "loss": 1.4731, "step": 27806 }, { "epoch": 2.1792319749216302, "grad_norm": 7.834564685821533, "learning_rate": 4.312380205825025e-06, "loss": 1.4171, "step": 27807 }, { "epoch": 2.179310344827586, "grad_norm": 3.065419912338257, "learning_rate": 4.312074774994672e-06, "loss": 0.8836, "step": 27808 }, { "epoch": 2.1793887147335425, "grad_norm": 3.4153037071228027, "learning_rate": 4.3117693467808275e-06, "loss": 1.3602, "step": 27809 }, { "epoch": 2.1794670846394983, "grad_norm": 2.134521722793579, "learning_rate": 4.311463921184662e-06, "loss": 1.3306, "step": 27810 }, { "epoch": 2.1795454545454547, "grad_norm": 3.038006544113159, "learning_rate": 4.311158498207333e-06, "loss": 1.2508, "step": 27811 }, { "epoch": 2.1796238244514106, "grad_norm": 2.92948055267334, "learning_rate": 4.310853077850005e-06, "loss": 0.9584, "step": 27812 }, { "epoch": 2.179702194357367, "grad_norm": 2.8131701946258545, "learning_rate": 4.310547660113836e-06, "loss": 1.6161, "step": 27813 }, { "epoch": 2.179780564263323, "grad_norm": 2.864089012145996, "learning_rate": 4.3102422449999895e-06, "loss": 1.1221, "step": 27814 }, { "epoch": 2.179858934169279, "grad_norm": 2.770883560180664, "learning_rate": 4.309936832509628e-06, "loss": 0.9652, "step": 27815 }, { "epoch": 2.179937304075235, "grad_norm": 2.819169521331787, "learning_rate": 4.309631422643914e-06, "loss": 0.9563, "step": 27816 }, { "epoch": 2.1800156739811913, "grad_norm": 5.224313735961914, "learning_rate": 4.3093260154040055e-06, "loss": 1.5426, "step": 27817 }, { "epoch": 2.1800940438871472, "grad_norm": 2.794314384460449, "learning_rate": 4.309020610791066e-06, "loss": 1.1537, "step": 27818 }, { "epoch": 2.1801724137931036, "grad_norm": 3.189424514770508, "learning_rate": 4.3087152088062585e-06, "loss": 1.2863, "step": 27819 }, { "epoch": 2.1802507836990594, "grad_norm": 4.400903224945068, "learning_rate": 4.308409809450742e-06, "loss": 1.49, "step": 27820 }, { "epoch": 2.180329153605016, "grad_norm": 3.4808404445648193, "learning_rate": 4.308104412725679e-06, "loss": 1.1045, "step": 27821 }, { "epoch": 2.1804075235109717, "grad_norm": 4.138237476348877, "learning_rate": 4.3077990186322314e-06, "loss": 1.5375, "step": 27822 }, { "epoch": 2.180485893416928, "grad_norm": 2.751742124557495, "learning_rate": 4.307493627171563e-06, "loss": 1.7041, "step": 27823 }, { "epoch": 2.180564263322884, "grad_norm": 2.595216989517212, "learning_rate": 4.307188238344831e-06, "loss": 1.5009, "step": 27824 }, { "epoch": 2.18064263322884, "grad_norm": 3.3700807094573975, "learning_rate": 4.306882852153201e-06, "loss": 1.7961, "step": 27825 }, { "epoch": 2.180721003134796, "grad_norm": 4.288511753082275, "learning_rate": 4.3065774685978295e-06, "loss": 1.3802, "step": 27826 }, { "epoch": 2.1807993730407524, "grad_norm": 8.658796310424805, "learning_rate": 4.306272087679885e-06, "loss": 0.458, "step": 27827 }, { "epoch": 2.1808777429467083, "grad_norm": 5.658804893493652, "learning_rate": 4.305966709400522e-06, "loss": 0.9426, "step": 27828 }, { "epoch": 2.1809561128526647, "grad_norm": 11.046478271484375, "learning_rate": 4.305661333760906e-06, "loss": 1.5294, "step": 27829 }, { "epoch": 2.1810344827586206, "grad_norm": 4.953073978424072, "learning_rate": 4.3053559607622e-06, "loss": 1.5478, "step": 27830 }, { "epoch": 2.181112852664577, "grad_norm": 2.8554306030273438, "learning_rate": 4.3050505904055604e-06, "loss": 1.2435, "step": 27831 }, { "epoch": 2.1811912225705328, "grad_norm": 4.798684597015381, "learning_rate": 4.3047452226921525e-06, "loss": 1.5325, "step": 27832 }, { "epoch": 2.181269592476489, "grad_norm": 2.4084460735321045, "learning_rate": 4.304439857623135e-06, "loss": 1.4644, "step": 27833 }, { "epoch": 2.181347962382445, "grad_norm": 2.6276392936706543, "learning_rate": 4.304134495199675e-06, "loss": 1.4218, "step": 27834 }, { "epoch": 2.1814263322884013, "grad_norm": 2.5686724185943604, "learning_rate": 4.303829135422926e-06, "loss": 1.4651, "step": 27835 }, { "epoch": 2.181504702194357, "grad_norm": 3.6876749992370605, "learning_rate": 4.303523778294054e-06, "loss": 1.2901, "step": 27836 }, { "epoch": 2.1815830721003135, "grad_norm": 3.752704620361328, "learning_rate": 4.303218423814221e-06, "loss": 1.666, "step": 27837 }, { "epoch": 2.1816614420062694, "grad_norm": 3.970695734024048, "learning_rate": 4.302913071984588e-06, "loss": 1.3007, "step": 27838 }, { "epoch": 2.1817398119122258, "grad_norm": 3.853147506713867, "learning_rate": 4.302607722806314e-06, "loss": 1.3171, "step": 27839 }, { "epoch": 2.1818181818181817, "grad_norm": 2.4466638565063477, "learning_rate": 4.302302376280561e-06, "loss": 1.3676, "step": 27840 }, { "epoch": 2.181896551724138, "grad_norm": 4.544016361236572, "learning_rate": 4.301997032408493e-06, "loss": 1.4279, "step": 27841 }, { "epoch": 2.181974921630094, "grad_norm": 3.356902599334717, "learning_rate": 4.3016916911912705e-06, "loss": 1.4557, "step": 27842 }, { "epoch": 2.18205329153605, "grad_norm": 5.813718795776367, "learning_rate": 4.301386352630053e-06, "loss": 1.4249, "step": 27843 }, { "epoch": 2.182131661442006, "grad_norm": 2.6143627166748047, "learning_rate": 4.301081016726003e-06, "loss": 1.2812, "step": 27844 }, { "epoch": 2.1822100313479624, "grad_norm": 3.971649169921875, "learning_rate": 4.300775683480283e-06, "loss": 0.7942, "step": 27845 }, { "epoch": 2.1822884012539183, "grad_norm": 3.1118741035461426, "learning_rate": 4.300470352894052e-06, "loss": 1.4118, "step": 27846 }, { "epoch": 2.1823667711598747, "grad_norm": 2.4839589595794678, "learning_rate": 4.300165024968472e-06, "loss": 1.4232, "step": 27847 }, { "epoch": 2.1824451410658305, "grad_norm": 2.9943816661834717, "learning_rate": 4.299859699704706e-06, "loss": 1.3226, "step": 27848 }, { "epoch": 2.182523510971787, "grad_norm": 8.240086555480957, "learning_rate": 4.299554377103914e-06, "loss": 1.0027, "step": 27849 }, { "epoch": 2.1826018808777428, "grad_norm": 5.510960102081299, "learning_rate": 4.299249057167257e-06, "loss": 1.8026, "step": 27850 }, { "epoch": 2.182680250783699, "grad_norm": 2.569819450378418, "learning_rate": 4.298943739895896e-06, "loss": 1.2452, "step": 27851 }, { "epoch": 2.182758620689655, "grad_norm": 4.403101921081543, "learning_rate": 4.298638425290992e-06, "loss": 0.6105, "step": 27852 }, { "epoch": 2.1828369905956113, "grad_norm": 2.5719311237335205, "learning_rate": 4.29833311335371e-06, "loss": 1.3412, "step": 27853 }, { "epoch": 2.182915360501567, "grad_norm": 4.468227386474609, "learning_rate": 4.298027804085207e-06, "loss": 1.9002, "step": 27854 }, { "epoch": 2.1829937304075235, "grad_norm": 3.6123507022857666, "learning_rate": 4.297722497486646e-06, "loss": 1.435, "step": 27855 }, { "epoch": 2.1830721003134794, "grad_norm": 2.4597630500793457, "learning_rate": 4.297417193559187e-06, "loss": 1.2405, "step": 27856 }, { "epoch": 2.1831504702194358, "grad_norm": 2.4753518104553223, "learning_rate": 4.2971118923039945e-06, "loss": 1.3842, "step": 27857 }, { "epoch": 2.1832288401253916, "grad_norm": 2.7889842987060547, "learning_rate": 4.296806593722225e-06, "loss": 1.2079, "step": 27858 }, { "epoch": 2.183307210031348, "grad_norm": 3.214586019515991, "learning_rate": 4.296501297815042e-06, "loss": 1.3098, "step": 27859 }, { "epoch": 2.183385579937304, "grad_norm": 4.042396545410156, "learning_rate": 4.296196004583608e-06, "loss": 0.9628, "step": 27860 }, { "epoch": 2.18346394984326, "grad_norm": 1.9937433004379272, "learning_rate": 4.295890714029083e-06, "loss": 1.1516, "step": 27861 }, { "epoch": 2.1835423197492165, "grad_norm": 4.403209209442139, "learning_rate": 4.295585426152627e-06, "loss": 0.8969, "step": 27862 }, { "epoch": 2.1836206896551724, "grad_norm": 2.8711917400360107, "learning_rate": 4.295280140955402e-06, "loss": 1.5126, "step": 27863 }, { "epoch": 2.1836990595611283, "grad_norm": 5.554233074188232, "learning_rate": 4.294974858438571e-06, "loss": 1.4081, "step": 27864 }, { "epoch": 2.1837774294670846, "grad_norm": 2.994962692260742, "learning_rate": 4.294669578603292e-06, "loss": 1.5231, "step": 27865 }, { "epoch": 2.183855799373041, "grad_norm": 2.666728973388672, "learning_rate": 4.294364301450727e-06, "loss": 1.3057, "step": 27866 }, { "epoch": 2.183934169278997, "grad_norm": 2.7719526290893555, "learning_rate": 4.29405902698204e-06, "loss": 1.2672, "step": 27867 }, { "epoch": 2.1840125391849528, "grad_norm": 2.983128786087036, "learning_rate": 4.293753755198389e-06, "loss": 1.4476, "step": 27868 }, { "epoch": 2.184090909090909, "grad_norm": 3.037216901779175, "learning_rate": 4.293448486100936e-06, "loss": 1.2797, "step": 27869 }, { "epoch": 2.1841692789968654, "grad_norm": 3.2423651218414307, "learning_rate": 4.293143219690841e-06, "loss": 1.2937, "step": 27870 }, { "epoch": 2.1842476489028213, "grad_norm": 4.784511089324951, "learning_rate": 4.292837955969267e-06, "loss": 1.3269, "step": 27871 }, { "epoch": 2.1843260188087776, "grad_norm": 4.73023796081543, "learning_rate": 4.292532694937374e-06, "loss": 1.3395, "step": 27872 }, { "epoch": 2.1844043887147335, "grad_norm": 3.203448534011841, "learning_rate": 4.292227436596324e-06, "loss": 1.3782, "step": 27873 }, { "epoch": 2.18448275862069, "grad_norm": 6.74299430847168, "learning_rate": 4.291922180947277e-06, "loss": 2.1848, "step": 27874 }, { "epoch": 2.1845611285266457, "grad_norm": 3.3918747901916504, "learning_rate": 4.291616927991394e-06, "loss": 1.3101, "step": 27875 }, { "epoch": 2.184639498432602, "grad_norm": 3.090867757797241, "learning_rate": 4.291311677729837e-06, "loss": 1.57, "step": 27876 }, { "epoch": 2.184717868338558, "grad_norm": 3.206040143966675, "learning_rate": 4.291006430163764e-06, "loss": 1.4648, "step": 27877 }, { "epoch": 2.1847962382445143, "grad_norm": 3.75742244720459, "learning_rate": 4.29070118529434e-06, "loss": 1.1, "step": 27878 }, { "epoch": 2.18487460815047, "grad_norm": 3.9183788299560547, "learning_rate": 4.290395943122726e-06, "loss": 1.4641, "step": 27879 }, { "epoch": 2.1849529780564265, "grad_norm": 6.19127893447876, "learning_rate": 4.29009070365008e-06, "loss": 1.6151, "step": 27880 }, { "epoch": 2.1850313479623824, "grad_norm": 2.4582815170288086, "learning_rate": 4.289785466877564e-06, "loss": 1.2351, "step": 27881 }, { "epoch": 2.1851097178683387, "grad_norm": 2.040043830871582, "learning_rate": 4.289480232806338e-06, "loss": 0.9797, "step": 27882 }, { "epoch": 2.1851880877742946, "grad_norm": 3.195204496383667, "learning_rate": 4.289175001437568e-06, "loss": 0.6285, "step": 27883 }, { "epoch": 2.185266457680251, "grad_norm": 2.1440298557281494, "learning_rate": 4.288869772772407e-06, "loss": 1.1547, "step": 27884 }, { "epoch": 2.185344827586207, "grad_norm": 2.917813777923584, "learning_rate": 4.288564546812022e-06, "loss": 1.0077, "step": 27885 }, { "epoch": 2.185423197492163, "grad_norm": 12.743941307067871, "learning_rate": 4.288259323557574e-06, "loss": 1.1221, "step": 27886 }, { "epoch": 2.185501567398119, "grad_norm": 7.682565212249756, "learning_rate": 4.28795410301022e-06, "loss": 1.0877, "step": 27887 }, { "epoch": 2.1855799373040754, "grad_norm": 4.982417106628418, "learning_rate": 4.287648885171122e-06, "loss": 1.09, "step": 27888 }, { "epoch": 2.1856583072100313, "grad_norm": 3.2854483127593994, "learning_rate": 4.287343670041442e-06, "loss": 1.5207, "step": 27889 }, { "epoch": 2.1857366771159876, "grad_norm": 2.581068277359009, "learning_rate": 4.287038457622344e-06, "loss": 1.546, "step": 27890 }, { "epoch": 2.1858150470219435, "grad_norm": 13.64935302734375, "learning_rate": 4.286733247914981e-06, "loss": 1.3684, "step": 27891 }, { "epoch": 2.1858934169279, "grad_norm": 4.29879903793335, "learning_rate": 4.28642804092052e-06, "loss": 1.4368, "step": 27892 }, { "epoch": 2.1859717868338557, "grad_norm": 3.1722640991210938, "learning_rate": 4.28612283664012e-06, "loss": 1.0702, "step": 27893 }, { "epoch": 2.186050156739812, "grad_norm": 3.541189432144165, "learning_rate": 4.285817635074944e-06, "loss": 1.545, "step": 27894 }, { "epoch": 2.186128526645768, "grad_norm": 3.2449777126312256, "learning_rate": 4.285512436226149e-06, "loss": 1.2872, "step": 27895 }, { "epoch": 2.1862068965517243, "grad_norm": 6.837806701660156, "learning_rate": 4.2852072400948966e-06, "loss": 1.3714, "step": 27896 }, { "epoch": 2.18628526645768, "grad_norm": 2.9536733627319336, "learning_rate": 4.28490204668235e-06, "loss": 1.1433, "step": 27897 }, { "epoch": 2.1863636363636365, "grad_norm": 2.9775822162628174, "learning_rate": 4.28459685598967e-06, "loss": 1.4631, "step": 27898 }, { "epoch": 2.1864420062695924, "grad_norm": 3.4171526432037354, "learning_rate": 4.284291668018014e-06, "loss": 1.4615, "step": 27899 }, { "epoch": 2.1865203761755487, "grad_norm": 5.582916259765625, "learning_rate": 4.283986482768547e-06, "loss": 1.5348, "step": 27900 }, { "epoch": 2.1865987460815046, "grad_norm": 2.6485676765441895, "learning_rate": 4.283681300242427e-06, "loss": 1.4379, "step": 27901 }, { "epoch": 2.186677115987461, "grad_norm": 2.9332423210144043, "learning_rate": 4.2833761204408156e-06, "loss": 1.3237, "step": 27902 }, { "epoch": 2.186755485893417, "grad_norm": 2.4757559299468994, "learning_rate": 4.283070943364871e-06, "loss": 1.1117, "step": 27903 }, { "epoch": 2.186833855799373, "grad_norm": 6.490952491760254, "learning_rate": 4.282765769015758e-06, "loss": 0.4065, "step": 27904 }, { "epoch": 2.186912225705329, "grad_norm": 4.587780475616455, "learning_rate": 4.282460597394636e-06, "loss": 1.2907, "step": 27905 }, { "epoch": 2.1869905956112854, "grad_norm": 3.1219711303710938, "learning_rate": 4.282155428502666e-06, "loss": 1.5779, "step": 27906 }, { "epoch": 2.1870689655172413, "grad_norm": 2.8135268688201904, "learning_rate": 4.281850262341006e-06, "loss": 1.4237, "step": 27907 }, { "epoch": 2.1871473354231976, "grad_norm": 3.774035692214966, "learning_rate": 4.281545098910819e-06, "loss": 1.0791, "step": 27908 }, { "epoch": 2.1872257053291535, "grad_norm": 2.5922293663024902, "learning_rate": 4.281239938213268e-06, "loss": 1.5595, "step": 27909 }, { "epoch": 2.18730407523511, "grad_norm": 2.9901387691497803, "learning_rate": 4.280934780249508e-06, "loss": 1.6839, "step": 27910 }, { "epoch": 2.1873824451410657, "grad_norm": 6.052331924438477, "learning_rate": 4.280629625020704e-06, "loss": 1.4458, "step": 27911 }, { "epoch": 2.187460815047022, "grad_norm": 3.279449224472046, "learning_rate": 4.280324472528015e-06, "loss": 1.2175, "step": 27912 }, { "epoch": 2.187539184952978, "grad_norm": 2.5962629318237305, "learning_rate": 4.280019322772604e-06, "loss": 1.2854, "step": 27913 }, { "epoch": 2.1876175548589343, "grad_norm": 4.437948226928711, "learning_rate": 4.279714175755627e-06, "loss": 1.0523, "step": 27914 }, { "epoch": 2.18769592476489, "grad_norm": 8.59419059753418, "learning_rate": 4.279409031478248e-06, "loss": 1.5494, "step": 27915 }, { "epoch": 2.1877742946708465, "grad_norm": 2.7343008518218994, "learning_rate": 4.279103889941629e-06, "loss": 1.0297, "step": 27916 }, { "epoch": 2.1878526645768024, "grad_norm": 3.9342918395996094, "learning_rate": 4.278798751146926e-06, "loss": 1.4837, "step": 27917 }, { "epoch": 2.1879310344827587, "grad_norm": 10.007975578308105, "learning_rate": 4.278493615095303e-06, "loss": 0.7589, "step": 27918 }, { "epoch": 2.1880094043887146, "grad_norm": 9.228096008300781, "learning_rate": 4.27818848178792e-06, "loss": 1.0267, "step": 27919 }, { "epoch": 2.188087774294671, "grad_norm": 2.621788263320923, "learning_rate": 4.277883351225938e-06, "loss": 1.5337, "step": 27920 }, { "epoch": 2.188166144200627, "grad_norm": 2.2635691165924072, "learning_rate": 4.277578223410516e-06, "loss": 1.2686, "step": 27921 }, { "epoch": 2.188244514106583, "grad_norm": 2.184943199157715, "learning_rate": 4.277273098342814e-06, "loss": 1.4761, "step": 27922 }, { "epoch": 2.188322884012539, "grad_norm": 2.7046027183532715, "learning_rate": 4.2769679760239945e-06, "loss": 1.1538, "step": 27923 }, { "epoch": 2.1884012539184954, "grad_norm": 6.189651966094971, "learning_rate": 4.276662856455219e-06, "loss": 0.9564, "step": 27924 }, { "epoch": 2.1884796238244513, "grad_norm": 3.647174596786499, "learning_rate": 4.276357739637645e-06, "loss": 1.4023, "step": 27925 }, { "epoch": 2.1885579937304076, "grad_norm": 3.7941174507141113, "learning_rate": 4.276052625572436e-06, "loss": 1.098, "step": 27926 }, { "epoch": 2.1886363636363635, "grad_norm": 2.5229530334472656, "learning_rate": 4.275747514260749e-06, "loss": 2.014, "step": 27927 }, { "epoch": 2.18871473354232, "grad_norm": 3.6294708251953125, "learning_rate": 4.275442405703749e-06, "loss": 1.5688, "step": 27928 }, { "epoch": 2.1887931034482757, "grad_norm": 2.563229560852051, "learning_rate": 4.2751372999025905e-06, "loss": 1.3597, "step": 27929 }, { "epoch": 2.188871473354232, "grad_norm": 2.127934694290161, "learning_rate": 4.2748321968584396e-06, "loss": 1.0124, "step": 27930 }, { "epoch": 2.188949843260188, "grad_norm": 2.5676980018615723, "learning_rate": 4.274527096572456e-06, "loss": 1.3933, "step": 27931 }, { "epoch": 2.1890282131661443, "grad_norm": 4.260594367980957, "learning_rate": 4.274221999045796e-06, "loss": 0.8762, "step": 27932 }, { "epoch": 2.1891065830721, "grad_norm": 2.87827467918396, "learning_rate": 4.2739169042796236e-06, "loss": 1.3571, "step": 27933 }, { "epoch": 2.1891849529780565, "grad_norm": 3.038727283477783, "learning_rate": 4.273611812275098e-06, "loss": 1.288, "step": 27934 }, { "epoch": 2.1892633228840124, "grad_norm": 4.830597877502441, "learning_rate": 4.273306723033382e-06, "loss": 1.8907, "step": 27935 }, { "epoch": 2.1893416927899687, "grad_norm": 3.6329336166381836, "learning_rate": 4.273001636555632e-06, "loss": 1.3263, "step": 27936 }, { "epoch": 2.1894200626959246, "grad_norm": 3.6196115016937256, "learning_rate": 4.2726965528430095e-06, "loss": 0.9808, "step": 27937 }, { "epoch": 2.189498432601881, "grad_norm": 3.682021141052246, "learning_rate": 4.272391471896678e-06, "loss": 0.9363, "step": 27938 }, { "epoch": 2.189576802507837, "grad_norm": 2.4705424308776855, "learning_rate": 4.272086393717795e-06, "loss": 1.2887, "step": 27939 }, { "epoch": 2.189655172413793, "grad_norm": 2.6366469860076904, "learning_rate": 4.271781318307521e-06, "loss": 1.3381, "step": 27940 }, { "epoch": 2.189733542319749, "grad_norm": 2.628812551498413, "learning_rate": 4.271476245667015e-06, "loss": 1.2371, "step": 27941 }, { "epoch": 2.1898119122257054, "grad_norm": 7.009641170501709, "learning_rate": 4.271171175797444e-06, "loss": 1.8244, "step": 27942 }, { "epoch": 2.1898902821316613, "grad_norm": 6.023574352264404, "learning_rate": 4.27086610869996e-06, "loss": 1.0296, "step": 27943 }, { "epoch": 2.1899686520376176, "grad_norm": 2.7750790119171143, "learning_rate": 4.270561044375726e-06, "loss": 1.1985, "step": 27944 }, { "epoch": 2.1900470219435735, "grad_norm": 2.8623270988464355, "learning_rate": 4.270255982825905e-06, "loss": 1.0485, "step": 27945 }, { "epoch": 2.19012539184953, "grad_norm": 3.5048112869262695, "learning_rate": 4.2699509240516555e-06, "loss": 1.3033, "step": 27946 }, { "epoch": 2.1902037617554857, "grad_norm": 3.8062469959259033, "learning_rate": 4.269645868054137e-06, "loss": 1.2888, "step": 27947 }, { "epoch": 2.190282131661442, "grad_norm": 2.4805190563201904, "learning_rate": 4.2693408148345095e-06, "loss": 1.0622, "step": 27948 }, { "epoch": 2.190360501567398, "grad_norm": 5.163987159729004, "learning_rate": 4.269035764393935e-06, "loss": 1.6262, "step": 27949 }, { "epoch": 2.1904388714733543, "grad_norm": 3.315399169921875, "learning_rate": 4.268730716733575e-06, "loss": 1.5783, "step": 27950 }, { "epoch": 2.19051724137931, "grad_norm": 4.082901954650879, "learning_rate": 4.268425671854586e-06, "loss": 1.1512, "step": 27951 }, { "epoch": 2.1905956112852665, "grad_norm": 3.7538349628448486, "learning_rate": 4.268120629758129e-06, "loss": 1.4482, "step": 27952 }, { "epoch": 2.1906739811912224, "grad_norm": 6.634105682373047, "learning_rate": 4.267815590445366e-06, "loss": 1.4884, "step": 27953 }, { "epoch": 2.1907523510971787, "grad_norm": 2.388148546218872, "learning_rate": 4.2675105539174585e-06, "loss": 1.4615, "step": 27954 }, { "epoch": 2.1908307210031346, "grad_norm": 4.565727233886719, "learning_rate": 4.267205520175562e-06, "loss": 1.4807, "step": 27955 }, { "epoch": 2.190909090909091, "grad_norm": 4.075380802154541, "learning_rate": 4.26690048922084e-06, "loss": 1.0315, "step": 27956 }, { "epoch": 2.1909874608150472, "grad_norm": 2.5079636573791504, "learning_rate": 4.266595461054453e-06, "loss": 1.5776, "step": 27957 }, { "epoch": 2.191065830721003, "grad_norm": 2.8654251098632812, "learning_rate": 4.266290435677559e-06, "loss": 1.2337, "step": 27958 }, { "epoch": 2.191144200626959, "grad_norm": 2.723198652267456, "learning_rate": 4.26598541309132e-06, "loss": 1.1765, "step": 27959 }, { "epoch": 2.1912225705329154, "grad_norm": 6.764906406402588, "learning_rate": 4.2656803932968935e-06, "loss": 1.402, "step": 27960 }, { "epoch": 2.1913009404388717, "grad_norm": 2.325113296508789, "learning_rate": 4.2653753762954446e-06, "loss": 1.4174, "step": 27961 }, { "epoch": 2.1913793103448276, "grad_norm": 4.81480598449707, "learning_rate": 4.265070362088129e-06, "loss": 1.5359, "step": 27962 }, { "epoch": 2.1914576802507835, "grad_norm": 3.301687002182007, "learning_rate": 4.264765350676109e-06, "loss": 1.55, "step": 27963 }, { "epoch": 2.19153605015674, "grad_norm": 2.5610058307647705, "learning_rate": 4.2644603420605424e-06, "loss": 1.525, "step": 27964 }, { "epoch": 2.191614420062696, "grad_norm": 3.0110089778900146, "learning_rate": 4.264155336242592e-06, "loss": 1.2364, "step": 27965 }, { "epoch": 2.191692789968652, "grad_norm": 3.1816012859344482, "learning_rate": 4.263850333223416e-06, "loss": 1.0843, "step": 27966 }, { "epoch": 2.1917711598746084, "grad_norm": 3.5868284702301025, "learning_rate": 4.263545333004176e-06, "loss": 1.3522, "step": 27967 }, { "epoch": 2.1918495297805642, "grad_norm": 6.343742847442627, "learning_rate": 4.263240335586031e-06, "loss": 1.2102, "step": 27968 }, { "epoch": 2.1919278996865206, "grad_norm": 4.833770751953125, "learning_rate": 4.262935340970142e-06, "loss": 1.0375, "step": 27969 }, { "epoch": 2.1920062695924765, "grad_norm": 4.152242660522461, "learning_rate": 4.262630349157668e-06, "loss": 0.7707, "step": 27970 }, { "epoch": 2.192084639498433, "grad_norm": 4.79637336730957, "learning_rate": 4.2623253601497685e-06, "loss": 1.4177, "step": 27971 }, { "epoch": 2.1921630094043887, "grad_norm": 4.830849647521973, "learning_rate": 4.262020373947606e-06, "loss": 1.4008, "step": 27972 }, { "epoch": 2.192241379310345, "grad_norm": 2.9388506412506104, "learning_rate": 4.261715390552337e-06, "loss": 1.4455, "step": 27973 }, { "epoch": 2.192319749216301, "grad_norm": 3.3458964824676514, "learning_rate": 4.261410409965124e-06, "loss": 1.1945, "step": 27974 }, { "epoch": 2.1923981191222572, "grad_norm": 4.667283058166504, "learning_rate": 4.261105432187127e-06, "loss": 1.4461, "step": 27975 }, { "epoch": 2.192476489028213, "grad_norm": 3.2755701541900635, "learning_rate": 4.260800457219506e-06, "loss": 0.8965, "step": 27976 }, { "epoch": 2.1925548589341695, "grad_norm": 5.870696544647217, "learning_rate": 4.2604954850634205e-06, "loss": 1.6223, "step": 27977 }, { "epoch": 2.1926332288401253, "grad_norm": 3.80072283744812, "learning_rate": 4.260190515720029e-06, "loss": 1.4775, "step": 27978 }, { "epoch": 2.1927115987460817, "grad_norm": 2.3198468685150146, "learning_rate": 4.259885549190492e-06, "loss": 1.2949, "step": 27979 }, { "epoch": 2.1927899686520376, "grad_norm": 3.3036820888519287, "learning_rate": 4.259580585475973e-06, "loss": 1.1572, "step": 27980 }, { "epoch": 2.192868338557994, "grad_norm": 5.774562358856201, "learning_rate": 4.259275624577628e-06, "loss": 1.3609, "step": 27981 }, { "epoch": 2.19294670846395, "grad_norm": 2.900017023086548, "learning_rate": 4.2589706664966174e-06, "loss": 1.4727, "step": 27982 }, { "epoch": 2.193025078369906, "grad_norm": 2.4879519939422607, "learning_rate": 4.2586657112341015e-06, "loss": 1.2418, "step": 27983 }, { "epoch": 2.193103448275862, "grad_norm": 3.319324493408203, "learning_rate": 4.258360758791243e-06, "loss": 1.3971, "step": 27984 }, { "epoch": 2.1931818181818183, "grad_norm": 6.9321064949035645, "learning_rate": 4.258055809169196e-06, "loss": 1.4864, "step": 27985 }, { "epoch": 2.1932601880877742, "grad_norm": 2.77175235748291, "learning_rate": 4.257750862369125e-06, "loss": 1.3606, "step": 27986 }, { "epoch": 2.1933385579937306, "grad_norm": 11.753722190856934, "learning_rate": 4.257445918392189e-06, "loss": 1.0446, "step": 27987 }, { "epoch": 2.1934169278996865, "grad_norm": 4.413163661956787, "learning_rate": 4.257140977239547e-06, "loss": 1.2861, "step": 27988 }, { "epoch": 2.193495297805643, "grad_norm": 3.5761873722076416, "learning_rate": 4.256836038912359e-06, "loss": 1.305, "step": 27989 }, { "epoch": 2.1935736677115987, "grad_norm": 2.6686441898345947, "learning_rate": 4.256531103411783e-06, "loss": 1.3139, "step": 27990 }, { "epoch": 2.193652037617555, "grad_norm": 4.847924709320068, "learning_rate": 4.256226170738985e-06, "loss": 1.2195, "step": 27991 }, { "epoch": 2.193730407523511, "grad_norm": 4.541983127593994, "learning_rate": 4.2559212408951175e-06, "loss": 1.238, "step": 27992 }, { "epoch": 2.1938087774294672, "grad_norm": 2.651750326156616, "learning_rate": 4.255616313881344e-06, "loss": 1.2533, "step": 27993 }, { "epoch": 2.193887147335423, "grad_norm": 3.622338056564331, "learning_rate": 4.2553113896988245e-06, "loss": 1.0138, "step": 27994 }, { "epoch": 2.1939655172413794, "grad_norm": 3.774808645248413, "learning_rate": 4.255006468348718e-06, "loss": 1.5935, "step": 27995 }, { "epoch": 2.1940438871473353, "grad_norm": 3.6260526180267334, "learning_rate": 4.254701549832183e-06, "loss": 1.1321, "step": 27996 }, { "epoch": 2.1941222570532917, "grad_norm": 4.787395477294922, "learning_rate": 4.25439663415038e-06, "loss": 1.2882, "step": 27997 }, { "epoch": 2.1942006269592476, "grad_norm": 7.602570056915283, "learning_rate": 4.254091721304473e-06, "loss": 1.2873, "step": 27998 }, { "epoch": 2.194278996865204, "grad_norm": 3.2952663898468018, "learning_rate": 4.253786811295615e-06, "loss": 1.2582, "step": 27999 }, { "epoch": 2.19435736677116, "grad_norm": 3.0357308387756348, "learning_rate": 4.253481904124968e-06, "loss": 1.3372, "step": 28000 }, { "epoch": 2.194435736677116, "grad_norm": 8.631478309631348, "learning_rate": 4.253176999793694e-06, "loss": 0.9362, "step": 28001 }, { "epoch": 2.194514106583072, "grad_norm": 2.8128137588500977, "learning_rate": 4.252872098302951e-06, "loss": 1.1478, "step": 28002 }, { "epoch": 2.1945924764890283, "grad_norm": 8.913751602172852, "learning_rate": 4.252567199653899e-06, "loss": 1.1656, "step": 28003 }, { "epoch": 2.194670846394984, "grad_norm": 5.577325820922852, "learning_rate": 4.252262303847695e-06, "loss": 1.926, "step": 28004 }, { "epoch": 2.1947492163009406, "grad_norm": 2.8688788414001465, "learning_rate": 4.251957410885503e-06, "loss": 1.1909, "step": 28005 }, { "epoch": 2.1948275862068964, "grad_norm": 3.2892208099365234, "learning_rate": 4.2516525207684825e-06, "loss": 1.3589, "step": 28006 }, { "epoch": 2.1949059561128528, "grad_norm": 2.7605462074279785, "learning_rate": 4.251347633497789e-06, "loss": 1.1778, "step": 28007 }, { "epoch": 2.1949843260188087, "grad_norm": 2.5145413875579834, "learning_rate": 4.251042749074585e-06, "loss": 1.2934, "step": 28008 }, { "epoch": 2.195062695924765, "grad_norm": 2.324068307876587, "learning_rate": 4.250737867500029e-06, "loss": 1.1514, "step": 28009 }, { "epoch": 2.195141065830721, "grad_norm": 5.170756816864014, "learning_rate": 4.250432988775285e-06, "loss": 1.1531, "step": 28010 }, { "epoch": 2.195219435736677, "grad_norm": 4.568497180938721, "learning_rate": 4.2501281129015045e-06, "loss": 1.4244, "step": 28011 }, { "epoch": 2.195297805642633, "grad_norm": 2.19578218460083, "learning_rate": 4.2498232398798534e-06, "loss": 1.2276, "step": 28012 }, { "epoch": 2.1953761755485894, "grad_norm": 2.9065840244293213, "learning_rate": 4.2495183697114905e-06, "loss": 1.7043, "step": 28013 }, { "epoch": 2.1954545454545453, "grad_norm": 3.431878089904785, "learning_rate": 4.249213502397573e-06, "loss": 1.4571, "step": 28014 }, { "epoch": 2.1955329153605017, "grad_norm": 2.2318906784057617, "learning_rate": 4.248908637939261e-06, "loss": 1.3489, "step": 28015 }, { "epoch": 2.1956112852664575, "grad_norm": 2.4645938873291016, "learning_rate": 4.248603776337714e-06, "loss": 1.4536, "step": 28016 }, { "epoch": 2.195689655172414, "grad_norm": 3.0279412269592285, "learning_rate": 4.248298917594096e-06, "loss": 1.086, "step": 28017 }, { "epoch": 2.1957680250783698, "grad_norm": 1.9976780414581299, "learning_rate": 4.24799406170956e-06, "loss": 1.2342, "step": 28018 }, { "epoch": 2.195846394984326, "grad_norm": 2.470125198364258, "learning_rate": 4.247689208685268e-06, "loss": 1.3267, "step": 28019 }, { "epoch": 2.195924764890282, "grad_norm": 4.953673362731934, "learning_rate": 4.247384358522381e-06, "loss": 0.948, "step": 28020 }, { "epoch": 2.1960031347962383, "grad_norm": 3.6463816165924072, "learning_rate": 4.2470795112220584e-06, "loss": 1.2196, "step": 28021 }, { "epoch": 2.196081504702194, "grad_norm": 3.329054117202759, "learning_rate": 4.246774666785456e-06, "loss": 1.5252, "step": 28022 }, { "epoch": 2.1961598746081505, "grad_norm": 5.162888526916504, "learning_rate": 4.246469825213736e-06, "loss": 1.3034, "step": 28023 }, { "epoch": 2.1962382445141064, "grad_norm": 2.4083943367004395, "learning_rate": 4.24616498650806e-06, "loss": 1.3602, "step": 28024 }, { "epoch": 2.1963166144200628, "grad_norm": 2.880600929260254, "learning_rate": 4.2458601506695846e-06, "loss": 1.3759, "step": 28025 }, { "epoch": 2.1963949843260187, "grad_norm": 3.930265426635742, "learning_rate": 4.245555317699469e-06, "loss": 1.6996, "step": 28026 }, { "epoch": 2.196473354231975, "grad_norm": 6.108808994293213, "learning_rate": 4.245250487598874e-06, "loss": 1.651, "step": 28027 }, { "epoch": 2.196551724137931, "grad_norm": 3.4466843605041504, "learning_rate": 4.244945660368959e-06, "loss": 1.649, "step": 28028 }, { "epoch": 2.196630094043887, "grad_norm": 3.5831167697906494, "learning_rate": 4.244640836010882e-06, "loss": 1.3435, "step": 28029 }, { "epoch": 2.196708463949843, "grad_norm": 4.260570526123047, "learning_rate": 4.244336014525802e-06, "loss": 0.8757, "step": 28030 }, { "epoch": 2.1967868338557994, "grad_norm": 2.528569459915161, "learning_rate": 4.244031195914881e-06, "loss": 1.1946, "step": 28031 }, { "epoch": 2.1968652037617553, "grad_norm": 3.2055411338806152, "learning_rate": 4.243726380179278e-06, "loss": 1.2327, "step": 28032 }, { "epoch": 2.1969435736677116, "grad_norm": 3.0152132511138916, "learning_rate": 4.24342156732015e-06, "loss": 1.8509, "step": 28033 }, { "epoch": 2.1970219435736675, "grad_norm": 6.313596725463867, "learning_rate": 4.243116757338659e-06, "loss": 1.7877, "step": 28034 }, { "epoch": 2.197100313479624, "grad_norm": 6.202723026275635, "learning_rate": 4.242811950235961e-06, "loss": 1.9009, "step": 28035 }, { "epoch": 2.1971786833855798, "grad_norm": 5.011476039886475, "learning_rate": 4.242507146013221e-06, "loss": 0.9176, "step": 28036 }, { "epoch": 2.197257053291536, "grad_norm": 2.516693592071533, "learning_rate": 4.242202344671592e-06, "loss": 1.2164, "step": 28037 }, { "epoch": 2.197335423197492, "grad_norm": 2.270278215408325, "learning_rate": 4.241897546212236e-06, "loss": 1.31, "step": 28038 }, { "epoch": 2.1974137931034483, "grad_norm": 8.38839340209961, "learning_rate": 4.241592750636314e-06, "loss": 1.2602, "step": 28039 }, { "epoch": 2.197492163009404, "grad_norm": 5.359395980834961, "learning_rate": 4.241287957944982e-06, "loss": 1.5282, "step": 28040 }, { "epoch": 2.1975705329153605, "grad_norm": 2.569976329803467, "learning_rate": 4.240983168139402e-06, "loss": 1.3558, "step": 28041 }, { "epoch": 2.1976489028213164, "grad_norm": 8.698308944702148, "learning_rate": 4.24067838122073e-06, "loss": 1.2506, "step": 28042 }, { "epoch": 2.1977272727272728, "grad_norm": 4.516960620880127, "learning_rate": 4.240373597190132e-06, "loss": 1.3686, "step": 28043 }, { "epoch": 2.1978056426332286, "grad_norm": 3.5816214084625244, "learning_rate": 4.2400688160487576e-06, "loss": 1.1115, "step": 28044 }, { "epoch": 2.197884012539185, "grad_norm": 4.897731304168701, "learning_rate": 4.239764037797773e-06, "loss": 1.5384, "step": 28045 }, { "epoch": 2.197962382445141, "grad_norm": 2.8389999866485596, "learning_rate": 4.239459262438337e-06, "loss": 1.3649, "step": 28046 }, { "epoch": 2.198040752351097, "grad_norm": 3.0826127529144287, "learning_rate": 4.2391544899716065e-06, "loss": 0.9094, "step": 28047 }, { "epoch": 2.1981191222570535, "grad_norm": 4.009069919586182, "learning_rate": 4.2388497203987415e-06, "loss": 1.3339, "step": 28048 }, { "epoch": 2.1981974921630094, "grad_norm": 2.257214307785034, "learning_rate": 4.238544953720899e-06, "loss": 1.4058, "step": 28049 }, { "epoch": 2.1982758620689653, "grad_norm": 2.794821262359619, "learning_rate": 4.238240189939242e-06, "loss": 1.3847, "step": 28050 }, { "epoch": 2.1983542319749216, "grad_norm": 2.4217734336853027, "learning_rate": 4.2379354290549294e-06, "loss": 1.4352, "step": 28051 }, { "epoch": 2.198432601880878, "grad_norm": 2.6250882148742676, "learning_rate": 4.237630671069118e-06, "loss": 1.3502, "step": 28052 }, { "epoch": 2.198510971786834, "grad_norm": 3.652966022491455, "learning_rate": 4.2373259159829675e-06, "loss": 1.3472, "step": 28053 }, { "epoch": 2.1985893416927897, "grad_norm": 3.147498846054077, "learning_rate": 4.237021163797639e-06, "loss": 1.5631, "step": 28054 }, { "epoch": 2.198667711598746, "grad_norm": 3.9958689212799072, "learning_rate": 4.236716414514289e-06, "loss": 1.6416, "step": 28055 }, { "epoch": 2.1987460815047024, "grad_norm": 2.235548257827759, "learning_rate": 4.236411668134076e-06, "loss": 1.4303, "step": 28056 }, { "epoch": 2.1988244514106583, "grad_norm": 4.858445644378662, "learning_rate": 4.2361069246581625e-06, "loss": 1.014, "step": 28057 }, { "epoch": 2.198902821316614, "grad_norm": 4.089315414428711, "learning_rate": 4.2358021840877065e-06, "loss": 1.8549, "step": 28058 }, { "epoch": 2.1989811912225705, "grad_norm": 3.7926926612854004, "learning_rate": 4.235497446423866e-06, "loss": 1.4864, "step": 28059 }, { "epoch": 2.199059561128527, "grad_norm": 6.554837703704834, "learning_rate": 4.235192711667801e-06, "loss": 1.6488, "step": 28060 }, { "epoch": 2.1991379310344827, "grad_norm": 5.530969619750977, "learning_rate": 4.234887979820668e-06, "loss": 1.2508, "step": 28061 }, { "epoch": 2.199216300940439, "grad_norm": 5.356905937194824, "learning_rate": 4.234583250883631e-06, "loss": 1.4697, "step": 28062 }, { "epoch": 2.199294670846395, "grad_norm": 2.9625120162963867, "learning_rate": 4.2342785248578435e-06, "loss": 1.2227, "step": 28063 }, { "epoch": 2.1993730407523513, "grad_norm": 3.2311489582061768, "learning_rate": 4.233973801744469e-06, "loss": 1.4487, "step": 28064 }, { "epoch": 2.199451410658307, "grad_norm": 2.7655935287475586, "learning_rate": 4.233669081544663e-06, "loss": 1.4347, "step": 28065 }, { "epoch": 2.1995297805642635, "grad_norm": 6.465382099151611, "learning_rate": 4.233364364259589e-06, "loss": 1.0004, "step": 28066 }, { "epoch": 2.1996081504702194, "grad_norm": 2.8821091651916504, "learning_rate": 4.2330596498904e-06, "loss": 1.4535, "step": 28067 }, { "epoch": 2.1996865203761757, "grad_norm": 3.3609213829040527, "learning_rate": 4.23275493843826e-06, "loss": 1.1341, "step": 28068 }, { "epoch": 2.1997648902821316, "grad_norm": 4.098123550415039, "learning_rate": 4.232450229904327e-06, "loss": 1.6492, "step": 28069 }, { "epoch": 2.199843260188088, "grad_norm": 6.121530532836914, "learning_rate": 4.232145524289758e-06, "loss": 1.4855, "step": 28070 }, { "epoch": 2.199921630094044, "grad_norm": 1.9048733711242676, "learning_rate": 4.231840821595713e-06, "loss": 1.0067, "step": 28071 }, { "epoch": 2.2, "grad_norm": 3.980502128601074, "learning_rate": 4.23153612182335e-06, "loss": 1.1293, "step": 28072 }, { "epoch": 2.200078369905956, "grad_norm": 2.4955503940582275, "learning_rate": 4.231231424973831e-06, "loss": 1.1838, "step": 28073 }, { "epoch": 2.2001567398119124, "grad_norm": 4.819706439971924, "learning_rate": 4.230926731048311e-06, "loss": 1.4409, "step": 28074 }, { "epoch": 2.2002351097178683, "grad_norm": 3.159688711166382, "learning_rate": 4.230622040047951e-06, "loss": 1.5223, "step": 28075 }, { "epoch": 2.2003134796238246, "grad_norm": 2.3207144737243652, "learning_rate": 4.230317351973909e-06, "loss": 1.1905, "step": 28076 }, { "epoch": 2.2003918495297805, "grad_norm": 3.020643949508667, "learning_rate": 4.230012666827346e-06, "loss": 1.1609, "step": 28077 }, { "epoch": 2.200470219435737, "grad_norm": 2.542156219482422, "learning_rate": 4.2297079846094195e-06, "loss": 0.8264, "step": 28078 }, { "epoch": 2.2005485893416927, "grad_norm": 3.2907235622406006, "learning_rate": 4.2294033053212865e-06, "loss": 1.0731, "step": 28079 }, { "epoch": 2.200626959247649, "grad_norm": 2.443711519241333, "learning_rate": 4.229098628964108e-06, "loss": 1.071, "step": 28080 }, { "epoch": 2.200705329153605, "grad_norm": 6.8371968269348145, "learning_rate": 4.2287939555390436e-06, "loss": 1.1762, "step": 28081 }, { "epoch": 2.2007836990595613, "grad_norm": 4.569899082183838, "learning_rate": 4.2284892850472494e-06, "loss": 1.4126, "step": 28082 }, { "epoch": 2.200862068965517, "grad_norm": 2.518758535385132, "learning_rate": 4.228184617489887e-06, "loss": 1.3512, "step": 28083 }, { "epoch": 2.2009404388714735, "grad_norm": 1.9240853786468506, "learning_rate": 4.227879952868114e-06, "loss": 1.323, "step": 28084 }, { "epoch": 2.2010188087774294, "grad_norm": 2.6294941902160645, "learning_rate": 4.227575291183088e-06, "loss": 1.2158, "step": 28085 }, { "epoch": 2.2010971786833857, "grad_norm": 3.3003883361816406, "learning_rate": 4.227270632435968e-06, "loss": 1.3054, "step": 28086 }, { "epoch": 2.2011755485893416, "grad_norm": 2.8113927841186523, "learning_rate": 4.226965976627915e-06, "loss": 1.2163, "step": 28087 }, { "epoch": 2.201253918495298, "grad_norm": 2.9501490592956543, "learning_rate": 4.226661323760087e-06, "loss": 1.2984, "step": 28088 }, { "epoch": 2.201332288401254, "grad_norm": 5.043436050415039, "learning_rate": 4.226356673833641e-06, "loss": 1.5898, "step": 28089 }, { "epoch": 2.20141065830721, "grad_norm": 4.511906623840332, "learning_rate": 4.226052026849737e-06, "loss": 1.1806, "step": 28090 }, { "epoch": 2.201489028213166, "grad_norm": 7.0210652351379395, "learning_rate": 4.225747382809532e-06, "loss": 1.5301, "step": 28091 }, { "epoch": 2.2015673981191224, "grad_norm": 6.250636100769043, "learning_rate": 4.22544274171419e-06, "loss": 1.0428, "step": 28092 }, { "epoch": 2.2016457680250783, "grad_norm": 2.2650742530822754, "learning_rate": 4.225138103564863e-06, "loss": 1.1732, "step": 28093 }, { "epoch": 2.2017241379310346, "grad_norm": 4.547648906707764, "learning_rate": 4.224833468362713e-06, "loss": 1.0949, "step": 28094 }, { "epoch": 2.2018025078369905, "grad_norm": 5.116152763366699, "learning_rate": 4.2245288361089e-06, "loss": 1.367, "step": 28095 }, { "epoch": 2.201880877742947, "grad_norm": 5.077336311340332, "learning_rate": 4.22422420680458e-06, "loss": 1.5364, "step": 28096 }, { "epoch": 2.2019592476489027, "grad_norm": 2.959786891937256, "learning_rate": 4.223919580450912e-06, "loss": 1.4275, "step": 28097 }, { "epoch": 2.202037617554859, "grad_norm": 4.260414123535156, "learning_rate": 4.223614957049055e-06, "loss": 1.7057, "step": 28098 }, { "epoch": 2.202115987460815, "grad_norm": 5.422021865844727, "learning_rate": 4.22331033660017e-06, "loss": 1.4924, "step": 28099 }, { "epoch": 2.2021943573667713, "grad_norm": 11.00387191772461, "learning_rate": 4.223005719105412e-06, "loss": 1.1657, "step": 28100 }, { "epoch": 2.202272727272727, "grad_norm": 2.866544246673584, "learning_rate": 4.222701104565941e-06, "loss": 1.4884, "step": 28101 }, { "epoch": 2.2023510971786835, "grad_norm": 2.2611887454986572, "learning_rate": 4.222396492982916e-06, "loss": 1.5305, "step": 28102 }, { "epoch": 2.2024294670846394, "grad_norm": 2.3361759185791016, "learning_rate": 4.222091884357497e-06, "loss": 1.4813, "step": 28103 }, { "epoch": 2.2025078369905957, "grad_norm": 14.025162696838379, "learning_rate": 4.221787278690839e-06, "loss": 1.3056, "step": 28104 }, { "epoch": 2.2025862068965516, "grad_norm": 6.837095737457275, "learning_rate": 4.221482675984102e-06, "loss": 0.607, "step": 28105 }, { "epoch": 2.202664576802508, "grad_norm": 2.647801399230957, "learning_rate": 4.221178076238445e-06, "loss": 1.3319, "step": 28106 }, { "epoch": 2.202742946708464, "grad_norm": 4.123846530914307, "learning_rate": 4.220873479455028e-06, "loss": 1.4823, "step": 28107 }, { "epoch": 2.20282131661442, "grad_norm": 2.0380172729492188, "learning_rate": 4.220568885635008e-06, "loss": 1.2596, "step": 28108 }, { "epoch": 2.202899686520376, "grad_norm": 3.4072484970092773, "learning_rate": 4.220264294779543e-06, "loss": 1.3914, "step": 28109 }, { "epoch": 2.2029780564263324, "grad_norm": 2.9940991401672363, "learning_rate": 4.219959706889794e-06, "loss": 1.2999, "step": 28110 }, { "epoch": 2.2030564263322883, "grad_norm": 3.1070668697357178, "learning_rate": 4.219655121966916e-06, "loss": 1.0882, "step": 28111 }, { "epoch": 2.2031347962382446, "grad_norm": 6.75324010848999, "learning_rate": 4.219350540012068e-06, "loss": 1.5166, "step": 28112 }, { "epoch": 2.2032131661442005, "grad_norm": 3.801602840423584, "learning_rate": 4.21904596102641e-06, "loss": 0.9805, "step": 28113 }, { "epoch": 2.203291536050157, "grad_norm": 7.238879680633545, "learning_rate": 4.218741385011102e-06, "loss": 1.3877, "step": 28114 }, { "epoch": 2.2033699059561127, "grad_norm": 3.581808090209961, "learning_rate": 4.218436811967299e-06, "loss": 1.6285, "step": 28115 }, { "epoch": 2.203448275862069, "grad_norm": 4.834822177886963, "learning_rate": 4.2181322418961616e-06, "loss": 1.5918, "step": 28116 }, { "epoch": 2.203526645768025, "grad_norm": 2.0396597385406494, "learning_rate": 4.217827674798845e-06, "loss": 1.207, "step": 28117 }, { "epoch": 2.2036050156739813, "grad_norm": 2.7557389736175537, "learning_rate": 4.217523110676514e-06, "loss": 1.2036, "step": 28118 }, { "epoch": 2.203683385579937, "grad_norm": 4.24862003326416, "learning_rate": 4.217218549530321e-06, "loss": 1.6228, "step": 28119 }, { "epoch": 2.2037617554858935, "grad_norm": 2.586703300476074, "learning_rate": 4.216913991361426e-06, "loss": 1.5177, "step": 28120 }, { "epoch": 2.2038401253918494, "grad_norm": 3.6555166244506836, "learning_rate": 4.21660943617099e-06, "loss": 1.3589, "step": 28121 }, { "epoch": 2.2039184952978057, "grad_norm": 3.0229077339172363, "learning_rate": 4.216304883960169e-06, "loss": 1.3201, "step": 28122 }, { "epoch": 2.2039968652037616, "grad_norm": 3.318336009979248, "learning_rate": 4.216000334730121e-06, "loss": 1.5004, "step": 28123 }, { "epoch": 2.204075235109718, "grad_norm": 2.5479748249053955, "learning_rate": 4.215695788482004e-06, "loss": 1.4045, "step": 28124 }, { "epoch": 2.204153605015674, "grad_norm": 2.336120367050171, "learning_rate": 4.21539124521698e-06, "loss": 1.2516, "step": 28125 }, { "epoch": 2.20423197492163, "grad_norm": 3.3662352561950684, "learning_rate": 4.215086704936202e-06, "loss": 1.7281, "step": 28126 }, { "epoch": 2.204310344827586, "grad_norm": 9.392257690429688, "learning_rate": 4.214782167640832e-06, "loss": 1.6478, "step": 28127 }, { "epoch": 2.2043887147335424, "grad_norm": 2.2690961360931396, "learning_rate": 4.214477633332027e-06, "loss": 1.1554, "step": 28128 }, { "epoch": 2.2044670846394983, "grad_norm": 2.997561454772949, "learning_rate": 4.214173102010947e-06, "loss": 1.3577, "step": 28129 }, { "epoch": 2.2045454545454546, "grad_norm": 2.2162952423095703, "learning_rate": 4.2138685736787485e-06, "loss": 0.9047, "step": 28130 }, { "epoch": 2.2046238244514105, "grad_norm": 3.7273406982421875, "learning_rate": 4.213564048336588e-06, "loss": 1.4238, "step": 28131 }, { "epoch": 2.204702194357367, "grad_norm": 7.413957118988037, "learning_rate": 4.213259525985628e-06, "loss": 1.1589, "step": 28132 }, { "epoch": 2.2047805642633227, "grad_norm": 2.3061001300811768, "learning_rate": 4.212955006627024e-06, "loss": 1.5854, "step": 28133 }, { "epoch": 2.204858934169279, "grad_norm": 8.441032409667969, "learning_rate": 4.212650490261936e-06, "loss": 1.7643, "step": 28134 }, { "epoch": 2.204937304075235, "grad_norm": 2.607048988342285, "learning_rate": 4.21234597689152e-06, "loss": 1.5634, "step": 28135 }, { "epoch": 2.2050156739811912, "grad_norm": 4.182351112365723, "learning_rate": 4.212041466516934e-06, "loss": 1.3843, "step": 28136 }, { "epoch": 2.205094043887147, "grad_norm": 2.495851755142212, "learning_rate": 4.211736959139342e-06, "loss": 1.0728, "step": 28137 }, { "epoch": 2.2051724137931035, "grad_norm": 8.037505149841309, "learning_rate": 4.211432454759893e-06, "loss": 1.1961, "step": 28138 }, { "epoch": 2.2052507836990594, "grad_norm": 2.9262537956237793, "learning_rate": 4.211127953379751e-06, "loss": 1.1011, "step": 28139 }, { "epoch": 2.2053291536050157, "grad_norm": 7.3200201988220215, "learning_rate": 4.210823455000076e-06, "loss": 1.5021, "step": 28140 }, { "epoch": 2.2054075235109716, "grad_norm": 2.813117504119873, "learning_rate": 4.210518959622021e-06, "loss": 1.3367, "step": 28141 }, { "epoch": 2.205485893416928, "grad_norm": 4.620471954345703, "learning_rate": 4.210214467246746e-06, "loss": 1.5022, "step": 28142 }, { "epoch": 2.2055642633228842, "grad_norm": 3.0374107360839844, "learning_rate": 4.209909977875409e-06, "loss": 0.7419, "step": 28143 }, { "epoch": 2.20564263322884, "grad_norm": 3.9134817123413086, "learning_rate": 4.209605491509171e-06, "loss": 1.2636, "step": 28144 }, { "epoch": 2.205721003134796, "grad_norm": 2.811325788497925, "learning_rate": 4.209301008149185e-06, "loss": 1.3953, "step": 28145 }, { "epoch": 2.2057993730407524, "grad_norm": 4.18775749206543, "learning_rate": 4.208996527796614e-06, "loss": 1.9644, "step": 28146 }, { "epoch": 2.2058777429467087, "grad_norm": 3.1776506900787354, "learning_rate": 4.208692050452613e-06, "loss": 1.1394, "step": 28147 }, { "epoch": 2.2059561128526646, "grad_norm": 3.331545114517212, "learning_rate": 4.2083875761183416e-06, "loss": 1.7083, "step": 28148 }, { "epoch": 2.2060344827586205, "grad_norm": 3.3808107376098633, "learning_rate": 4.208083104794957e-06, "loss": 1.2851, "step": 28149 }, { "epoch": 2.206112852664577, "grad_norm": 3.1610910892486572, "learning_rate": 4.207778636483616e-06, "loss": 1.1653, "step": 28150 }, { "epoch": 2.206191222570533, "grad_norm": 5.177190780639648, "learning_rate": 4.207474171185481e-06, "loss": 1.4107, "step": 28151 }, { "epoch": 2.206269592476489, "grad_norm": 8.355463981628418, "learning_rate": 4.207169708901705e-06, "loss": 1.7869, "step": 28152 }, { "epoch": 2.2063479623824453, "grad_norm": 2.2854185104370117, "learning_rate": 4.2068652496334485e-06, "loss": 0.845, "step": 28153 }, { "epoch": 2.2064263322884012, "grad_norm": 3.154304027557373, "learning_rate": 4.206560793381869e-06, "loss": 1.2782, "step": 28154 }, { "epoch": 2.2065047021943576, "grad_norm": 4.533659934997559, "learning_rate": 4.206256340148126e-06, "loss": 1.9582, "step": 28155 }, { "epoch": 2.2065830721003135, "grad_norm": 4.5474138259887695, "learning_rate": 4.205951889933376e-06, "loss": 1.7785, "step": 28156 }, { "epoch": 2.20666144200627, "grad_norm": 3.9839720726013184, "learning_rate": 4.205647442738775e-06, "loss": 1.4098, "step": 28157 }, { "epoch": 2.2067398119122257, "grad_norm": 3.6615118980407715, "learning_rate": 4.205342998565485e-06, "loss": 1.3459, "step": 28158 }, { "epoch": 2.206818181818182, "grad_norm": 4.067523002624512, "learning_rate": 4.205038557414662e-06, "loss": 1.6781, "step": 28159 }, { "epoch": 2.206896551724138, "grad_norm": 3.1580848693847656, "learning_rate": 4.204734119287464e-06, "loss": 1.3279, "step": 28160 }, { "epoch": 2.2069749216300942, "grad_norm": 2.3493783473968506, "learning_rate": 4.204429684185048e-06, "loss": 1.3391, "step": 28161 }, { "epoch": 2.20705329153605, "grad_norm": 3.2181825637817383, "learning_rate": 4.2041252521085725e-06, "loss": 1.2232, "step": 28162 }, { "epoch": 2.2071316614420065, "grad_norm": 2.4717044830322266, "learning_rate": 4.2038208230591984e-06, "loss": 1.3588, "step": 28163 }, { "epoch": 2.2072100313479623, "grad_norm": 4.719052314758301, "learning_rate": 4.203516397038079e-06, "loss": 1.4816, "step": 28164 }, { "epoch": 2.2072884012539187, "grad_norm": 6.693448543548584, "learning_rate": 4.2032119740463745e-06, "loss": 2.4762, "step": 28165 }, { "epoch": 2.2073667711598746, "grad_norm": 10.181559562683105, "learning_rate": 4.202907554085243e-06, "loss": 0.8922, "step": 28166 }, { "epoch": 2.207445141065831, "grad_norm": 3.809706926345825, "learning_rate": 4.202603137155841e-06, "loss": 1.5158, "step": 28167 }, { "epoch": 2.207523510971787, "grad_norm": 2.730041265487671, "learning_rate": 4.202298723259327e-06, "loss": 1.0266, "step": 28168 }, { "epoch": 2.207601880877743, "grad_norm": 2.635120153427124, "learning_rate": 4.201994312396858e-06, "loss": 1.3395, "step": 28169 }, { "epoch": 2.207680250783699, "grad_norm": 2.2515673637390137, "learning_rate": 4.2016899045695955e-06, "loss": 1.2282, "step": 28170 }, { "epoch": 2.2077586206896553, "grad_norm": 2.300902843475342, "learning_rate": 4.201385499778693e-06, "loss": 1.5555, "step": 28171 }, { "epoch": 2.2078369905956112, "grad_norm": 6.201737880706787, "learning_rate": 4.2010810980253095e-06, "loss": 1.1919, "step": 28172 }, { "epoch": 2.2079153605015676, "grad_norm": 2.4694135189056396, "learning_rate": 4.200776699310603e-06, "loss": 1.1828, "step": 28173 }, { "epoch": 2.2079937304075234, "grad_norm": 2.090031862258911, "learning_rate": 4.200472303635733e-06, "loss": 1.123, "step": 28174 }, { "epoch": 2.20807210031348, "grad_norm": 4.396871566772461, "learning_rate": 4.2001679110018535e-06, "loss": 1.3818, "step": 28175 }, { "epoch": 2.2081504702194357, "grad_norm": 4.5556640625, "learning_rate": 4.199863521410125e-06, "loss": 0.8231, "step": 28176 }, { "epoch": 2.208228840125392, "grad_norm": 4.6104865074157715, "learning_rate": 4.199559134861705e-06, "loss": 1.2308, "step": 28177 }, { "epoch": 2.208307210031348, "grad_norm": 3.2067363262176514, "learning_rate": 4.1992547513577525e-06, "loss": 1.2259, "step": 28178 }, { "epoch": 2.208385579937304, "grad_norm": 3.8711419105529785, "learning_rate": 4.198950370899422e-06, "loss": 1.2219, "step": 28179 }, { "epoch": 2.20846394984326, "grad_norm": 2.3501837253570557, "learning_rate": 4.198645993487872e-06, "loss": 1.3355, "step": 28180 }, { "epoch": 2.2085423197492164, "grad_norm": 2.9615437984466553, "learning_rate": 4.198341619124264e-06, "loss": 1.2232, "step": 28181 }, { "epoch": 2.2086206896551723, "grad_norm": 3.683661937713623, "learning_rate": 4.19803724780975e-06, "loss": 1.6794, "step": 28182 }, { "epoch": 2.2086990595611287, "grad_norm": 3.385610580444336, "learning_rate": 4.197732879545492e-06, "loss": 1.4154, "step": 28183 }, { "epoch": 2.2087774294670846, "grad_norm": 3.1725475788116455, "learning_rate": 4.197428514332645e-06, "loss": 1.5225, "step": 28184 }, { "epoch": 2.208855799373041, "grad_norm": 2.735949993133545, "learning_rate": 4.197124152172369e-06, "loss": 1.3493, "step": 28185 }, { "epoch": 2.2089341692789968, "grad_norm": 2.25594425201416, "learning_rate": 4.196819793065819e-06, "loss": 1.2217, "step": 28186 }, { "epoch": 2.209012539184953, "grad_norm": 5.295373916625977, "learning_rate": 4.1965154370141535e-06, "loss": 1.4497, "step": 28187 }, { "epoch": 2.209090909090909, "grad_norm": 5.545726299285889, "learning_rate": 4.196211084018531e-06, "loss": 1.7779, "step": 28188 }, { "epoch": 2.2091692789968653, "grad_norm": 4.016866683959961, "learning_rate": 4.19590673408011e-06, "loss": 1.6166, "step": 28189 }, { "epoch": 2.209247648902821, "grad_norm": 2.6933364868164062, "learning_rate": 4.195602387200046e-06, "loss": 1.246, "step": 28190 }, { "epoch": 2.2093260188087775, "grad_norm": 2.559177875518799, "learning_rate": 4.195298043379497e-06, "loss": 1.2439, "step": 28191 }, { "epoch": 2.2094043887147334, "grad_norm": 3.047788619995117, "learning_rate": 4.1949937026196206e-06, "loss": 1.3734, "step": 28192 }, { "epoch": 2.2094827586206898, "grad_norm": 3.4236466884613037, "learning_rate": 4.194689364921577e-06, "loss": 1.5894, "step": 28193 }, { "epoch": 2.2095611285266457, "grad_norm": 2.2462220191955566, "learning_rate": 4.194385030286518e-06, "loss": 1.2301, "step": 28194 }, { "epoch": 2.209639498432602, "grad_norm": 3.09287691116333, "learning_rate": 4.194080698715607e-06, "loss": 1.2227, "step": 28195 }, { "epoch": 2.209717868338558, "grad_norm": 7.007223606109619, "learning_rate": 4.193776370209999e-06, "loss": 1.9795, "step": 28196 }, { "epoch": 2.209796238244514, "grad_norm": 5.0991740226745605, "learning_rate": 4.193472044770851e-06, "loss": 1.2915, "step": 28197 }, { "epoch": 2.20987460815047, "grad_norm": 3.615968942642212, "learning_rate": 4.1931677223993205e-06, "loss": 1.2076, "step": 28198 }, { "epoch": 2.2099529780564264, "grad_norm": 2.9524691104888916, "learning_rate": 4.192863403096565e-06, "loss": 1.0565, "step": 28199 }, { "epoch": 2.2100313479623823, "grad_norm": 4.73462438583374, "learning_rate": 4.192559086863745e-06, "loss": 1.6749, "step": 28200 }, { "epoch": 2.2101097178683387, "grad_norm": 2.313875675201416, "learning_rate": 4.192254773702014e-06, "loss": 1.2776, "step": 28201 }, { "epoch": 2.2101880877742945, "grad_norm": 3.5979011058807373, "learning_rate": 4.1919504636125304e-06, "loss": 1.3611, "step": 28202 }, { "epoch": 2.210266457680251, "grad_norm": 3.60762357711792, "learning_rate": 4.191646156596453e-06, "loss": 1.3802, "step": 28203 }, { "epoch": 2.2103448275862068, "grad_norm": 2.0482840538024902, "learning_rate": 4.191341852654938e-06, "loss": 0.9169, "step": 28204 }, { "epoch": 2.210423197492163, "grad_norm": 4.499917984008789, "learning_rate": 4.191037551789145e-06, "loss": 0.8013, "step": 28205 }, { "epoch": 2.210501567398119, "grad_norm": 2.6567368507385254, "learning_rate": 4.190733254000226e-06, "loss": 1.2334, "step": 28206 }, { "epoch": 2.2105799373040753, "grad_norm": 3.403481960296631, "learning_rate": 4.190428959289347e-06, "loss": 1.6785, "step": 28207 }, { "epoch": 2.210658307210031, "grad_norm": 2.7834386825561523, "learning_rate": 4.190124667657656e-06, "loss": 1.6892, "step": 28208 }, { "epoch": 2.2107366771159875, "grad_norm": 4.0622944831848145, "learning_rate": 4.189820379106317e-06, "loss": 1.1783, "step": 28209 }, { "epoch": 2.2108150470219434, "grad_norm": 3.117107391357422, "learning_rate": 4.1895160936364835e-06, "loss": 1.1376, "step": 28210 }, { "epoch": 2.2108934169278998, "grad_norm": 3.459653615951538, "learning_rate": 4.189211811249317e-06, "loss": 1.2434, "step": 28211 }, { "epoch": 2.2109717868338556, "grad_norm": 4.408684730529785, "learning_rate": 4.188907531945971e-06, "loss": 1.3251, "step": 28212 }, { "epoch": 2.211050156739812, "grad_norm": 3.3882644176483154, "learning_rate": 4.188603255727603e-06, "loss": 1.3553, "step": 28213 }, { "epoch": 2.211128526645768, "grad_norm": 3.7265350818634033, "learning_rate": 4.188298982595372e-06, "loss": 1.2933, "step": 28214 }, { "epoch": 2.211206896551724, "grad_norm": 4.472093105316162, "learning_rate": 4.187994712550436e-06, "loss": 1.3302, "step": 28215 }, { "epoch": 2.21128526645768, "grad_norm": 2.7826130390167236, "learning_rate": 4.18769044559395e-06, "loss": 1.255, "step": 28216 }, { "epoch": 2.2113636363636364, "grad_norm": 4.4879631996154785, "learning_rate": 4.187386181727072e-06, "loss": 1.3095, "step": 28217 }, { "epoch": 2.2114420062695923, "grad_norm": 2.594954013824463, "learning_rate": 4.187081920950959e-06, "loss": 1.0178, "step": 28218 }, { "epoch": 2.2115203761755486, "grad_norm": 5.100770950317383, "learning_rate": 4.1867776632667725e-06, "loss": 1.0435, "step": 28219 }, { "epoch": 2.2115987460815045, "grad_norm": 3.792423963546753, "learning_rate": 4.186473408675663e-06, "loss": 1.358, "step": 28220 }, { "epoch": 2.211677115987461, "grad_norm": 4.1067399978637695, "learning_rate": 4.186169157178791e-06, "loss": 1.4583, "step": 28221 }, { "epoch": 2.2117554858934168, "grad_norm": 2.5831987857818604, "learning_rate": 4.185864908777315e-06, "loss": 1.4978, "step": 28222 }, { "epoch": 2.211833855799373, "grad_norm": 2.831510066986084, "learning_rate": 4.185560663472389e-06, "loss": 1.2271, "step": 28223 }, { "epoch": 2.211912225705329, "grad_norm": 3.6974196434020996, "learning_rate": 4.185256421265173e-06, "loss": 1.091, "step": 28224 }, { "epoch": 2.2119905956112853, "grad_norm": 4.6847052574157715, "learning_rate": 4.184952182156821e-06, "loss": 1.5631, "step": 28225 }, { "epoch": 2.212068965517241, "grad_norm": 5.171989440917969, "learning_rate": 4.184647946148496e-06, "loss": 0.9916, "step": 28226 }, { "epoch": 2.2121473354231975, "grad_norm": 2.7007124423980713, "learning_rate": 4.184343713241348e-06, "loss": 1.6176, "step": 28227 }, { "epoch": 2.2122257053291534, "grad_norm": 7.455834865570068, "learning_rate": 4.184039483436539e-06, "loss": 0.9993, "step": 28228 }, { "epoch": 2.2123040752351097, "grad_norm": 2.890197992324829, "learning_rate": 4.183735256735224e-06, "loss": 1.2779, "step": 28229 }, { "epoch": 2.2123824451410656, "grad_norm": 2.710401773452759, "learning_rate": 4.183431033138563e-06, "loss": 1.3212, "step": 28230 }, { "epoch": 2.212460815047022, "grad_norm": 5.916323661804199, "learning_rate": 4.183126812647708e-06, "loss": 1.2989, "step": 28231 }, { "epoch": 2.212539184952978, "grad_norm": 4.615233421325684, "learning_rate": 4.18282259526382e-06, "loss": 1.4046, "step": 28232 }, { "epoch": 2.212617554858934, "grad_norm": 4.919708728790283, "learning_rate": 4.182518380988055e-06, "loss": 1.4465, "step": 28233 }, { "epoch": 2.21269592476489, "grad_norm": 3.850904941558838, "learning_rate": 4.1822141698215716e-06, "loss": 1.3979, "step": 28234 }, { "epoch": 2.2127742946708464, "grad_norm": 3.033149003982544, "learning_rate": 4.181909961765524e-06, "loss": 1.1139, "step": 28235 }, { "epoch": 2.2128526645768023, "grad_norm": 2.878857135772705, "learning_rate": 4.1816057568210714e-06, "loss": 0.9151, "step": 28236 }, { "epoch": 2.2129310344827586, "grad_norm": 2.6576991081237793, "learning_rate": 4.181301554989371e-06, "loss": 1.3933, "step": 28237 }, { "epoch": 2.213009404388715, "grad_norm": 2.9014101028442383, "learning_rate": 4.180997356271579e-06, "loss": 1.2298, "step": 28238 }, { "epoch": 2.213087774294671, "grad_norm": 4.7496538162231445, "learning_rate": 4.18069316066885e-06, "loss": 1.5158, "step": 28239 }, { "epoch": 2.2131661442006267, "grad_norm": 3.35214900970459, "learning_rate": 4.180388968182344e-06, "loss": 1.2854, "step": 28240 }, { "epoch": 2.213244514106583, "grad_norm": 3.698505401611328, "learning_rate": 4.180084778813219e-06, "loss": 1.5399, "step": 28241 }, { "epoch": 2.2133228840125394, "grad_norm": 3.1850802898406982, "learning_rate": 4.17978059256263e-06, "loss": 1.0181, "step": 28242 }, { "epoch": 2.2134012539184953, "grad_norm": 4.004579544067383, "learning_rate": 4.179476409431733e-06, "loss": 1.2288, "step": 28243 }, { "epoch": 2.213479623824451, "grad_norm": 13.784845352172852, "learning_rate": 4.1791722294216866e-06, "loss": 0.9224, "step": 28244 }, { "epoch": 2.2135579937304075, "grad_norm": 6.476318359375, "learning_rate": 4.17886805253365e-06, "loss": 1.196, "step": 28245 }, { "epoch": 2.213636363636364, "grad_norm": 4.436473846435547, "learning_rate": 4.178563878768774e-06, "loss": 1.3729, "step": 28246 }, { "epoch": 2.2137147335423197, "grad_norm": 3.664876937866211, "learning_rate": 4.17825970812822e-06, "loss": 1.6539, "step": 28247 }, { "epoch": 2.213793103448276, "grad_norm": 5.293007850646973, "learning_rate": 4.177955540613145e-06, "loss": 1.4368, "step": 28248 }, { "epoch": 2.213871473354232, "grad_norm": 2.2772631645202637, "learning_rate": 4.177651376224705e-06, "loss": 1.2635, "step": 28249 }, { "epoch": 2.2139498432601883, "grad_norm": 2.1742970943450928, "learning_rate": 4.177347214964057e-06, "loss": 1.2797, "step": 28250 }, { "epoch": 2.214028213166144, "grad_norm": 4.012304306030273, "learning_rate": 4.177043056832355e-06, "loss": 1.1027, "step": 28251 }, { "epoch": 2.2141065830721005, "grad_norm": 2.7584850788116455, "learning_rate": 4.176738901830763e-06, "loss": 1.1828, "step": 28252 }, { "epoch": 2.2141849529780564, "grad_norm": 3.7468864917755127, "learning_rate": 4.176434749960429e-06, "loss": 1.1413, "step": 28253 }, { "epoch": 2.2142633228840127, "grad_norm": 2.578230857849121, "learning_rate": 4.1761306012225165e-06, "loss": 1.3376, "step": 28254 }, { "epoch": 2.2143416927899686, "grad_norm": 2.5611796379089355, "learning_rate": 4.17582645561818e-06, "loss": 1.246, "step": 28255 }, { "epoch": 2.214420062695925, "grad_norm": 3.4961254596710205, "learning_rate": 4.1755223131485755e-06, "loss": 1.7921, "step": 28256 }, { "epoch": 2.214498432601881, "grad_norm": 7.019449710845947, "learning_rate": 4.175218173814861e-06, "loss": 0.779, "step": 28257 }, { "epoch": 2.214576802507837, "grad_norm": 5.472051620483398, "learning_rate": 4.174914037618192e-06, "loss": 1.5812, "step": 28258 }, { "epoch": 2.214655172413793, "grad_norm": 2.7416486740112305, "learning_rate": 4.174609904559727e-06, "loss": 1.3705, "step": 28259 }, { "epoch": 2.2147335423197494, "grad_norm": 3.5154476165771484, "learning_rate": 4.174305774640624e-06, "loss": 1.3814, "step": 28260 }, { "epoch": 2.2148119122257053, "grad_norm": 4.7680344581604, "learning_rate": 4.174001647862035e-06, "loss": 1.0053, "step": 28261 }, { "epoch": 2.2148902821316616, "grad_norm": 2.458298921585083, "learning_rate": 4.17369752422512e-06, "loss": 1.4575, "step": 28262 }, { "epoch": 2.2149686520376175, "grad_norm": 3.093526601791382, "learning_rate": 4.1733934037310365e-06, "loss": 0.9095, "step": 28263 }, { "epoch": 2.215047021943574, "grad_norm": 4.377826690673828, "learning_rate": 4.173089286380938e-06, "loss": 1.2605, "step": 28264 }, { "epoch": 2.2151253918495297, "grad_norm": 3.061246871948242, "learning_rate": 4.172785172175984e-06, "loss": 1.6427, "step": 28265 }, { "epoch": 2.215203761755486, "grad_norm": 2.866483211517334, "learning_rate": 4.17248106111733e-06, "loss": 1.4799, "step": 28266 }, { "epoch": 2.215282131661442, "grad_norm": 3.197628974914551, "learning_rate": 4.172176953206135e-06, "loss": 1.235, "step": 28267 }, { "epoch": 2.2153605015673983, "grad_norm": 3.936253070831299, "learning_rate": 4.171872848443552e-06, "loss": 1.3455, "step": 28268 }, { "epoch": 2.215438871473354, "grad_norm": 2.5465824604034424, "learning_rate": 4.1715687468307385e-06, "loss": 1.3472, "step": 28269 }, { "epoch": 2.2155172413793105, "grad_norm": 5.813747406005859, "learning_rate": 4.171264648368852e-06, "loss": 1.7929, "step": 28270 }, { "epoch": 2.2155956112852664, "grad_norm": 2.7875702381134033, "learning_rate": 4.170960553059052e-06, "loss": 1.1842, "step": 28271 }, { "epoch": 2.2156739811912227, "grad_norm": 3.0820858478546143, "learning_rate": 4.170656460902489e-06, "loss": 1.0266, "step": 28272 }, { "epoch": 2.2157523510971786, "grad_norm": 2.4572625160217285, "learning_rate": 4.170352371900324e-06, "loss": 1.1905, "step": 28273 }, { "epoch": 2.215830721003135, "grad_norm": 5.607822418212891, "learning_rate": 4.170048286053713e-06, "loss": 1.1368, "step": 28274 }, { "epoch": 2.215909090909091, "grad_norm": 2.931591272354126, "learning_rate": 4.169744203363812e-06, "loss": 1.2996, "step": 28275 }, { "epoch": 2.215987460815047, "grad_norm": 3.3377685546875, "learning_rate": 4.169440123831776e-06, "loss": 1.3551, "step": 28276 }, { "epoch": 2.216065830721003, "grad_norm": 6.874673843383789, "learning_rate": 4.1691360474587645e-06, "loss": 1.1369, "step": 28277 }, { "epoch": 2.2161442006269594, "grad_norm": 2.2804009914398193, "learning_rate": 4.1688319742459334e-06, "loss": 1.3757, "step": 28278 }, { "epoch": 2.2162225705329153, "grad_norm": 3.780313014984131, "learning_rate": 4.168527904194438e-06, "loss": 1.3957, "step": 28279 }, { "epoch": 2.2163009404388716, "grad_norm": 3.0958683490753174, "learning_rate": 4.168223837305434e-06, "loss": 0.9408, "step": 28280 }, { "epoch": 2.2163793103448275, "grad_norm": 6.935520648956299, "learning_rate": 4.1679197735800795e-06, "loss": 1.8755, "step": 28281 }, { "epoch": 2.216457680250784, "grad_norm": 3.0694258213043213, "learning_rate": 4.167615713019534e-06, "loss": 1.3752, "step": 28282 }, { "epoch": 2.2165360501567397, "grad_norm": 2.83005428314209, "learning_rate": 4.167311655624947e-06, "loss": 1.3954, "step": 28283 }, { "epoch": 2.216614420062696, "grad_norm": 2.368495464324951, "learning_rate": 4.16700760139748e-06, "loss": 1.3221, "step": 28284 }, { "epoch": 2.216692789968652, "grad_norm": 2.542065382003784, "learning_rate": 4.166703550338289e-06, "loss": 1.4761, "step": 28285 }, { "epoch": 2.2167711598746083, "grad_norm": 2.405414342880249, "learning_rate": 4.16639950244853e-06, "loss": 1.0887, "step": 28286 }, { "epoch": 2.216849529780564, "grad_norm": 3.9443161487579346, "learning_rate": 4.166095457729358e-06, "loss": 1.3636, "step": 28287 }, { "epoch": 2.2169278996865205, "grad_norm": 4.819931507110596, "learning_rate": 4.165791416181929e-06, "loss": 1.6006, "step": 28288 }, { "epoch": 2.2170062695924764, "grad_norm": 3.496593713760376, "learning_rate": 4.165487377807403e-06, "loss": 1.2672, "step": 28289 }, { "epoch": 2.2170846394984327, "grad_norm": 2.5818448066711426, "learning_rate": 4.165183342606936e-06, "loss": 1.1287, "step": 28290 }, { "epoch": 2.2171630094043886, "grad_norm": 6.755015850067139, "learning_rate": 4.164879310581681e-06, "loss": 1.0355, "step": 28291 }, { "epoch": 2.217241379310345, "grad_norm": 4.393182754516602, "learning_rate": 4.164575281732796e-06, "loss": 1.2831, "step": 28292 }, { "epoch": 2.217319749216301, "grad_norm": 2.458909511566162, "learning_rate": 4.164271256061438e-06, "loss": 1.1954, "step": 28293 }, { "epoch": 2.217398119122257, "grad_norm": 2.34551739692688, "learning_rate": 4.163967233568762e-06, "loss": 1.3279, "step": 28294 }, { "epoch": 2.217476489028213, "grad_norm": 2.44370174407959, "learning_rate": 4.1636632142559245e-06, "loss": 1.2857, "step": 28295 }, { "epoch": 2.2175548589341694, "grad_norm": 16.937192916870117, "learning_rate": 4.163359198124084e-06, "loss": 0.9526, "step": 28296 }, { "epoch": 2.2176332288401253, "grad_norm": 3.155992031097412, "learning_rate": 4.163055185174396e-06, "loss": 1.3758, "step": 28297 }, { "epoch": 2.2177115987460816, "grad_norm": 3.504347562789917, "learning_rate": 4.1627511754080156e-06, "loss": 1.0982, "step": 28298 }, { "epoch": 2.2177899686520375, "grad_norm": 4.1371870040893555, "learning_rate": 4.1624471688260995e-06, "loss": 1.2155, "step": 28299 }, { "epoch": 2.217868338557994, "grad_norm": 6.082976818084717, "learning_rate": 4.1621431654298024e-06, "loss": 1.369, "step": 28300 }, { "epoch": 2.2179467084639497, "grad_norm": 2.470397472381592, "learning_rate": 4.161839165220286e-06, "loss": 1.1729, "step": 28301 }, { "epoch": 2.218025078369906, "grad_norm": 4.990030288696289, "learning_rate": 4.161535168198701e-06, "loss": 1.0046, "step": 28302 }, { "epoch": 2.218103448275862, "grad_norm": 5.045039653778076, "learning_rate": 4.161231174366205e-06, "loss": 1.258, "step": 28303 }, { "epoch": 2.2181818181818183, "grad_norm": 4.453485488891602, "learning_rate": 4.160927183723956e-06, "loss": 0.9034, "step": 28304 }, { "epoch": 2.218260188087774, "grad_norm": 3.1895787715911865, "learning_rate": 4.160623196273108e-06, "loss": 1.4215, "step": 28305 }, { "epoch": 2.2183385579937305, "grad_norm": 3.04380202293396, "learning_rate": 4.160319212014819e-06, "loss": 1.246, "step": 28306 }, { "epoch": 2.2184169278996864, "grad_norm": 4.608914375305176, "learning_rate": 4.160015230950243e-06, "loss": 1.9904, "step": 28307 }, { "epoch": 2.2184952978056427, "grad_norm": 3.149289846420288, "learning_rate": 4.15971125308054e-06, "loss": 1.4215, "step": 28308 }, { "epoch": 2.2185736677115986, "grad_norm": 2.0384974479675293, "learning_rate": 4.159407278406862e-06, "loss": 1.3304, "step": 28309 }, { "epoch": 2.218652037617555, "grad_norm": 2.3756256103515625, "learning_rate": 4.159103306930367e-06, "loss": 1.3593, "step": 28310 }, { "epoch": 2.218730407523511, "grad_norm": 3.1991963386535645, "learning_rate": 4.158799338652212e-06, "loss": 1.5796, "step": 28311 }, { "epoch": 2.218808777429467, "grad_norm": 2.7988436222076416, "learning_rate": 4.158495373573552e-06, "loss": 1.2192, "step": 28312 }, { "epoch": 2.218887147335423, "grad_norm": 2.495967388153076, "learning_rate": 4.158191411695543e-06, "loss": 1.3377, "step": 28313 }, { "epoch": 2.2189655172413794, "grad_norm": 14.312207221984863, "learning_rate": 4.157887453019341e-06, "loss": 1.4551, "step": 28314 }, { "epoch": 2.2190438871473352, "grad_norm": 2.331333875656128, "learning_rate": 4.157583497546103e-06, "loss": 1.5119, "step": 28315 }, { "epoch": 2.2191222570532916, "grad_norm": 3.0399482250213623, "learning_rate": 4.157279545276986e-06, "loss": 1.4945, "step": 28316 }, { "epoch": 2.2192006269592475, "grad_norm": 17.427473068237305, "learning_rate": 4.156975596213143e-06, "loss": 2.1389, "step": 28317 }, { "epoch": 2.219278996865204, "grad_norm": 2.7850749492645264, "learning_rate": 4.156671650355733e-06, "loss": 1.086, "step": 28318 }, { "epoch": 2.2193573667711597, "grad_norm": 3.239985704421997, "learning_rate": 4.156367707705912e-06, "loss": 1.1241, "step": 28319 }, { "epoch": 2.219435736677116, "grad_norm": 2.6839747428894043, "learning_rate": 4.156063768264832e-06, "loss": 1.6786, "step": 28320 }, { "epoch": 2.219514106583072, "grad_norm": 15.338980674743652, "learning_rate": 4.155759832033653e-06, "loss": 0.8654, "step": 28321 }, { "epoch": 2.2195924764890282, "grad_norm": 2.7958765029907227, "learning_rate": 4.155455899013529e-06, "loss": 1.2584, "step": 28322 }, { "epoch": 2.219670846394984, "grad_norm": 2.9010567665100098, "learning_rate": 4.15515196920562e-06, "loss": 1.5801, "step": 28323 }, { "epoch": 2.2197492163009405, "grad_norm": 2.556100845336914, "learning_rate": 4.154848042611077e-06, "loss": 1.239, "step": 28324 }, { "epoch": 2.2198275862068964, "grad_norm": 2.0385890007019043, "learning_rate": 4.154544119231058e-06, "loss": 1.6432, "step": 28325 }, { "epoch": 2.2199059561128527, "grad_norm": 2.564107656478882, "learning_rate": 4.154240199066718e-06, "loss": 1.0674, "step": 28326 }, { "epoch": 2.2199843260188086, "grad_norm": 4.430568695068359, "learning_rate": 4.153936282119217e-06, "loss": 1.2612, "step": 28327 }, { "epoch": 2.220062695924765, "grad_norm": 3.3106002807617188, "learning_rate": 4.153632368389705e-06, "loss": 1.3613, "step": 28328 }, { "epoch": 2.2201410658307212, "grad_norm": 2.4314210414886475, "learning_rate": 4.153328457879341e-06, "loss": 1.1431, "step": 28329 }, { "epoch": 2.220219435736677, "grad_norm": 3.3268301486968994, "learning_rate": 4.153024550589281e-06, "loss": 0.9348, "step": 28330 }, { "epoch": 2.220297805642633, "grad_norm": 7.183349132537842, "learning_rate": 4.152720646520681e-06, "loss": 1.2259, "step": 28331 }, { "epoch": 2.2203761755485893, "grad_norm": 3.4549412727355957, "learning_rate": 4.152416745674697e-06, "loss": 1.2898, "step": 28332 }, { "epoch": 2.2204545454545457, "grad_norm": 2.220961570739746, "learning_rate": 4.152112848052482e-06, "loss": 1.2466, "step": 28333 }, { "epoch": 2.2205329153605016, "grad_norm": 3.360827922821045, "learning_rate": 4.151808953655198e-06, "loss": 1.4463, "step": 28334 }, { "epoch": 2.2206112852664575, "grad_norm": 2.8829352855682373, "learning_rate": 4.151505062483994e-06, "loss": 1.4952, "step": 28335 }, { "epoch": 2.220689655172414, "grad_norm": 2.7607903480529785, "learning_rate": 4.15120117454003e-06, "loss": 1.5826, "step": 28336 }, { "epoch": 2.22076802507837, "grad_norm": 3.1433684825897217, "learning_rate": 4.15089728982446e-06, "loss": 1.6418, "step": 28337 }, { "epoch": 2.220846394984326, "grad_norm": 9.188019752502441, "learning_rate": 4.150593408338442e-06, "loss": 1.0478, "step": 28338 }, { "epoch": 2.220924764890282, "grad_norm": 3.438114643096924, "learning_rate": 4.150289530083129e-06, "loss": 1.1549, "step": 28339 }, { "epoch": 2.2210031347962382, "grad_norm": 4.905551433563232, "learning_rate": 4.149985655059678e-06, "loss": 1.5918, "step": 28340 }, { "epoch": 2.2210815047021946, "grad_norm": 3.532872200012207, "learning_rate": 4.1496817832692455e-06, "loss": 0.9967, "step": 28341 }, { "epoch": 2.2211598746081505, "grad_norm": 3.0287091732025146, "learning_rate": 4.149377914712987e-06, "loss": 1.5092, "step": 28342 }, { "epoch": 2.221238244514107, "grad_norm": 4.7617621421813965, "learning_rate": 4.1490740493920585e-06, "loss": 1.543, "step": 28343 }, { "epoch": 2.2213166144200627, "grad_norm": 2.100233316421509, "learning_rate": 4.148770187307615e-06, "loss": 1.2875, "step": 28344 }, { "epoch": 2.221394984326019, "grad_norm": 4.120450496673584, "learning_rate": 4.1484663284608105e-06, "loss": 1.3185, "step": 28345 }, { "epoch": 2.221473354231975, "grad_norm": 8.188702583312988, "learning_rate": 4.148162472852806e-06, "loss": 0.7911, "step": 28346 }, { "epoch": 2.2215517241379312, "grad_norm": 2.4325737953186035, "learning_rate": 4.147858620484752e-06, "loss": 1.1818, "step": 28347 }, { "epoch": 2.221630094043887, "grad_norm": 6.1429243087768555, "learning_rate": 4.147554771357806e-06, "loss": 0.7214, "step": 28348 }, { "epoch": 2.2217084639498434, "grad_norm": 4.1536383628845215, "learning_rate": 4.147250925473125e-06, "loss": 1.4398, "step": 28349 }, { "epoch": 2.2217868338557993, "grad_norm": 2.983962059020996, "learning_rate": 4.146947082831862e-06, "loss": 1.3219, "step": 28350 }, { "epoch": 2.2218652037617557, "grad_norm": 2.726162910461426, "learning_rate": 4.1466432434351755e-06, "loss": 1.1153, "step": 28351 }, { "epoch": 2.2219435736677116, "grad_norm": 3.4123899936676025, "learning_rate": 4.146339407284218e-06, "loss": 1.322, "step": 28352 }, { "epoch": 2.222021943573668, "grad_norm": 8.847332000732422, "learning_rate": 4.14603557438015e-06, "loss": 2.0537, "step": 28353 }, { "epoch": 2.222100313479624, "grad_norm": 3.43481707572937, "learning_rate": 4.145731744724121e-06, "loss": 1.1674, "step": 28354 }, { "epoch": 2.22217868338558, "grad_norm": 8.032588005065918, "learning_rate": 4.145427918317291e-06, "loss": 0.9211, "step": 28355 }, { "epoch": 2.222257053291536, "grad_norm": 5.0486063957214355, "learning_rate": 4.145124095160814e-06, "loss": 1.1341, "step": 28356 }, { "epoch": 2.2223354231974923, "grad_norm": 3.365651845932007, "learning_rate": 4.144820275255847e-06, "loss": 1.5014, "step": 28357 }, { "epoch": 2.222413793103448, "grad_norm": 1.8703322410583496, "learning_rate": 4.144516458603542e-06, "loss": 1.3118, "step": 28358 }, { "epoch": 2.2224921630094046, "grad_norm": 13.338081359863281, "learning_rate": 4.144212645205058e-06, "loss": 1.9214, "step": 28359 }, { "epoch": 2.2225705329153604, "grad_norm": 2.5192694664001465, "learning_rate": 4.143908835061551e-06, "loss": 1.3474, "step": 28360 }, { "epoch": 2.2226489028213168, "grad_norm": 2.209461212158203, "learning_rate": 4.143605028174173e-06, "loss": 1.3165, "step": 28361 }, { "epoch": 2.2227272727272727, "grad_norm": 2.8471479415893555, "learning_rate": 4.143301224544083e-06, "loss": 1.2955, "step": 28362 }, { "epoch": 2.222805642633229, "grad_norm": 5.2896223068237305, "learning_rate": 4.142997424172435e-06, "loss": 1.4807, "step": 28363 }, { "epoch": 2.222884012539185, "grad_norm": 2.136108636856079, "learning_rate": 4.142693627060385e-06, "loss": 1.3513, "step": 28364 }, { "epoch": 2.222962382445141, "grad_norm": 3.2139058113098145, "learning_rate": 4.142389833209087e-06, "loss": 0.9858, "step": 28365 }, { "epoch": 2.223040752351097, "grad_norm": 3.237057685852051, "learning_rate": 4.142086042619697e-06, "loss": 1.4792, "step": 28366 }, { "epoch": 2.2231191222570534, "grad_norm": 2.2001726627349854, "learning_rate": 4.141782255293372e-06, "loss": 1.3853, "step": 28367 }, { "epoch": 2.2231974921630093, "grad_norm": 2.9978232383728027, "learning_rate": 4.1414784712312675e-06, "loss": 1.5018, "step": 28368 }, { "epoch": 2.2232758620689657, "grad_norm": 3.8596689701080322, "learning_rate": 4.141174690434536e-06, "loss": 1.577, "step": 28369 }, { "epoch": 2.2233542319749215, "grad_norm": 6.446117877960205, "learning_rate": 4.140870912904336e-06, "loss": 0.4502, "step": 28370 }, { "epoch": 2.223432601880878, "grad_norm": 2.773656129837036, "learning_rate": 4.14056713864182e-06, "loss": 1.5377, "step": 28371 }, { "epoch": 2.2235109717868338, "grad_norm": 7.508632183074951, "learning_rate": 4.140263367648148e-06, "loss": 1.9628, "step": 28372 }, { "epoch": 2.22358934169279, "grad_norm": 2.6387364864349365, "learning_rate": 4.139959599924471e-06, "loss": 1.4757, "step": 28373 }, { "epoch": 2.223667711598746, "grad_norm": 4.6472039222717285, "learning_rate": 4.139655835471946e-06, "loss": 1.3151, "step": 28374 }, { "epoch": 2.2237460815047023, "grad_norm": 3.2590935230255127, "learning_rate": 4.13935207429173e-06, "loss": 1.4221, "step": 28375 }, { "epoch": 2.223824451410658, "grad_norm": 3.281826972961426, "learning_rate": 4.139048316384975e-06, "loss": 1.82, "step": 28376 }, { "epoch": 2.2239028213166145, "grad_norm": 3.3012964725494385, "learning_rate": 4.138744561752836e-06, "loss": 1.3933, "step": 28377 }, { "epoch": 2.2239811912225704, "grad_norm": 5.5462188720703125, "learning_rate": 4.138440810396473e-06, "loss": 1.0948, "step": 28378 }, { "epoch": 2.2240595611285268, "grad_norm": 5.591483116149902, "learning_rate": 4.1381370623170396e-06, "loss": 1.1173, "step": 28379 }, { "epoch": 2.2241379310344827, "grad_norm": 4.36816930770874, "learning_rate": 4.137833317515688e-06, "loss": 1.7295, "step": 28380 }, { "epoch": 2.224216300940439, "grad_norm": 3.7409873008728027, "learning_rate": 4.1375295759935764e-06, "loss": 1.4246, "step": 28381 }, { "epoch": 2.224294670846395, "grad_norm": 3.1209723949432373, "learning_rate": 4.137225837751859e-06, "loss": 1.2634, "step": 28382 }, { "epoch": 2.224373040752351, "grad_norm": 3.853804111480713, "learning_rate": 4.136922102791692e-06, "loss": 1.3859, "step": 28383 }, { "epoch": 2.224451410658307, "grad_norm": 2.2657928466796875, "learning_rate": 4.1366183711142285e-06, "loss": 1.3346, "step": 28384 }, { "epoch": 2.2245297805642634, "grad_norm": 8.713693618774414, "learning_rate": 4.136314642720626e-06, "loss": 0.8489, "step": 28385 }, { "epoch": 2.2246081504702193, "grad_norm": 3.0270862579345703, "learning_rate": 4.136010917612039e-06, "loss": 1.436, "step": 28386 }, { "epoch": 2.2246865203761756, "grad_norm": 2.637014865875244, "learning_rate": 4.1357071957896245e-06, "loss": 1.686, "step": 28387 }, { "epoch": 2.2247648902821315, "grad_norm": 2.9310154914855957, "learning_rate": 4.135403477254534e-06, "loss": 1.2555, "step": 28388 }, { "epoch": 2.224843260188088, "grad_norm": 3.989539384841919, "learning_rate": 4.135099762007923e-06, "loss": 1.4802, "step": 28389 }, { "epoch": 2.2249216300940438, "grad_norm": 3.480648994445801, "learning_rate": 4.134796050050953e-06, "loss": 1.7375, "step": 28390 }, { "epoch": 2.225, "grad_norm": 3.1894001960754395, "learning_rate": 4.134492341384771e-06, "loss": 1.2983, "step": 28391 }, { "epoch": 2.225078369905956, "grad_norm": 2.374169111251831, "learning_rate": 4.134188636010536e-06, "loss": 1.5555, "step": 28392 }, { "epoch": 2.2251567398119123, "grad_norm": 3.3485898971557617, "learning_rate": 4.133884933929403e-06, "loss": 1.3207, "step": 28393 }, { "epoch": 2.225235109717868, "grad_norm": 1.8710993528366089, "learning_rate": 4.133581235142527e-06, "loss": 1.3637, "step": 28394 }, { "epoch": 2.2253134796238245, "grad_norm": 4.469804763793945, "learning_rate": 4.133277539651062e-06, "loss": 1.4938, "step": 28395 }, { "epoch": 2.2253918495297804, "grad_norm": 2.233708381652832, "learning_rate": 4.132973847456165e-06, "loss": 1.3394, "step": 28396 }, { "epoch": 2.2254702194357368, "grad_norm": 6.2418084144592285, "learning_rate": 4.13267015855899e-06, "loss": 1.7229, "step": 28397 }, { "epoch": 2.2255485893416926, "grad_norm": 11.374634742736816, "learning_rate": 4.132366472960693e-06, "loss": 1.2014, "step": 28398 }, { "epoch": 2.225626959247649, "grad_norm": 3.3096346855163574, "learning_rate": 4.132062790662427e-06, "loss": 0.8367, "step": 28399 }, { "epoch": 2.225705329153605, "grad_norm": 5.838473320007324, "learning_rate": 4.131759111665349e-06, "loss": 1.3633, "step": 28400 }, { "epoch": 2.225783699059561, "grad_norm": 4.289334297180176, "learning_rate": 4.131455435970611e-06, "loss": 1.3438, "step": 28401 }, { "epoch": 2.225862068965517, "grad_norm": 6.55430793762207, "learning_rate": 4.131151763579376e-06, "loss": 1.2009, "step": 28402 }, { "epoch": 2.2259404388714734, "grad_norm": 4.708864212036133, "learning_rate": 4.130848094492789e-06, "loss": 1.4165, "step": 28403 }, { "epoch": 2.2260188087774293, "grad_norm": 4.391069412231445, "learning_rate": 4.13054442871201e-06, "loss": 1.6533, "step": 28404 }, { "epoch": 2.2260971786833856, "grad_norm": 2.999922752380371, "learning_rate": 4.130240766238195e-06, "loss": 1.2733, "step": 28405 }, { "epoch": 2.2261755485893415, "grad_norm": 6.172568321228027, "learning_rate": 4.129937107072496e-06, "loss": 1.6267, "step": 28406 }, { "epoch": 2.226253918495298, "grad_norm": 5.806349754333496, "learning_rate": 4.129633451216069e-06, "loss": 1.0807, "step": 28407 }, { "epoch": 2.2263322884012537, "grad_norm": 7.895473003387451, "learning_rate": 4.129329798670069e-06, "loss": 0.9483, "step": 28408 }, { "epoch": 2.22641065830721, "grad_norm": 4.102843284606934, "learning_rate": 4.129026149435654e-06, "loss": 1.4171, "step": 28409 }, { "epoch": 2.226489028213166, "grad_norm": 2.1929657459259033, "learning_rate": 4.128722503513974e-06, "loss": 1.4343, "step": 28410 }, { "epoch": 2.2265673981191223, "grad_norm": 3.2439687252044678, "learning_rate": 4.128418860906186e-06, "loss": 1.1697, "step": 28411 }, { "epoch": 2.226645768025078, "grad_norm": 2.8853466510772705, "learning_rate": 4.128115221613445e-06, "loss": 1.4288, "step": 28412 }, { "epoch": 2.2267241379310345, "grad_norm": 3.876425266265869, "learning_rate": 4.1278115856369075e-06, "loss": 1.2067, "step": 28413 }, { "epoch": 2.2268025078369904, "grad_norm": 2.292660713195801, "learning_rate": 4.127507952977725e-06, "loss": 1.321, "step": 28414 }, { "epoch": 2.2268808777429467, "grad_norm": 5.05194616317749, "learning_rate": 4.127204323637053e-06, "loss": 1.2727, "step": 28415 }, { "epoch": 2.2269592476489026, "grad_norm": 3.3503658771514893, "learning_rate": 4.126900697616052e-06, "loss": 1.4376, "step": 28416 }, { "epoch": 2.227037617554859, "grad_norm": 3.6562185287475586, "learning_rate": 4.126597074915868e-06, "loss": 2.002, "step": 28417 }, { "epoch": 2.227115987460815, "grad_norm": 2.897620677947998, "learning_rate": 4.126293455537661e-06, "loss": 1.172, "step": 28418 }, { "epoch": 2.227194357366771, "grad_norm": 4.6345438957214355, "learning_rate": 4.125989839482585e-06, "loss": 1.341, "step": 28419 }, { "epoch": 2.227272727272727, "grad_norm": 2.757194995880127, "learning_rate": 4.125686226751797e-06, "loss": 1.5871, "step": 28420 }, { "epoch": 2.2273510971786834, "grad_norm": 4.1241021156311035, "learning_rate": 4.125382617346447e-06, "loss": 1.4708, "step": 28421 }, { "epoch": 2.2274294670846393, "grad_norm": 4.295033931732178, "learning_rate": 4.125079011267692e-06, "loss": 0.9444, "step": 28422 }, { "epoch": 2.2275078369905956, "grad_norm": 3.121814489364624, "learning_rate": 4.124775408516688e-06, "loss": 1.2274, "step": 28423 }, { "epoch": 2.227586206896552, "grad_norm": 8.854084968566895, "learning_rate": 4.12447180909459e-06, "loss": 1.0516, "step": 28424 }, { "epoch": 2.227664576802508, "grad_norm": 4.001461029052734, "learning_rate": 4.12416821300255e-06, "loss": 0.8428, "step": 28425 }, { "epoch": 2.2277429467084637, "grad_norm": 3.994541883468628, "learning_rate": 4.123864620241724e-06, "loss": 1.7601, "step": 28426 }, { "epoch": 2.22782131661442, "grad_norm": 3.2051029205322266, "learning_rate": 4.123561030813267e-06, "loss": 1.4105, "step": 28427 }, { "epoch": 2.2278996865203764, "grad_norm": 3.0866987705230713, "learning_rate": 4.123257444718336e-06, "loss": 1.4167, "step": 28428 }, { "epoch": 2.2279780564263323, "grad_norm": 2.467801094055176, "learning_rate": 4.12295386195808e-06, "loss": 1.4051, "step": 28429 }, { "epoch": 2.228056426332288, "grad_norm": 2.486165761947632, "learning_rate": 4.122650282533659e-06, "loss": 1.0316, "step": 28430 }, { "epoch": 2.2281347962382445, "grad_norm": 3.047736406326294, "learning_rate": 4.122346706446226e-06, "loss": 1.5358, "step": 28431 }, { "epoch": 2.228213166144201, "grad_norm": 1.9072465896606445, "learning_rate": 4.122043133696935e-06, "loss": 1.1699, "step": 28432 }, { "epoch": 2.2282915360501567, "grad_norm": 4.584039211273193, "learning_rate": 4.121739564286941e-06, "loss": 1.2472, "step": 28433 }, { "epoch": 2.2283699059561126, "grad_norm": 5.214262008666992, "learning_rate": 4.1214359982173965e-06, "loss": 1.3102, "step": 28434 }, { "epoch": 2.228448275862069, "grad_norm": 3.003649950027466, "learning_rate": 4.1211324354894625e-06, "loss": 1.3297, "step": 28435 }, { "epoch": 2.2285266457680253, "grad_norm": 10.381407737731934, "learning_rate": 4.120828876104285e-06, "loss": 1.9016, "step": 28436 }, { "epoch": 2.228605015673981, "grad_norm": 3.275123119354248, "learning_rate": 4.1205253200630245e-06, "loss": 1.4081, "step": 28437 }, { "epoch": 2.2286833855799375, "grad_norm": 6.017420768737793, "learning_rate": 4.1202217673668345e-06, "loss": 1.4386, "step": 28438 }, { "epoch": 2.2287617554858934, "grad_norm": 5.09971809387207, "learning_rate": 4.119918218016869e-06, "loss": 1.6238, "step": 28439 }, { "epoch": 2.2288401253918497, "grad_norm": 3.3881351947784424, "learning_rate": 4.119614672014282e-06, "loss": 1.3688, "step": 28440 }, { "epoch": 2.2289184952978056, "grad_norm": 2.8385932445526123, "learning_rate": 4.1193111293602276e-06, "loss": 1.3338, "step": 28441 }, { "epoch": 2.228996865203762, "grad_norm": 2.7413082122802734, "learning_rate": 4.119007590055862e-06, "loss": 1.3336, "step": 28442 }, { "epoch": 2.229075235109718, "grad_norm": 4.12923526763916, "learning_rate": 4.118704054102341e-06, "loss": 0.9806, "step": 28443 }, { "epoch": 2.229153605015674, "grad_norm": 2.6660590171813965, "learning_rate": 4.118400521500816e-06, "loss": 1.3483, "step": 28444 }, { "epoch": 2.22923197492163, "grad_norm": 2.462853193283081, "learning_rate": 4.118096992252442e-06, "loss": 1.0745, "step": 28445 }, { "epoch": 2.2293103448275864, "grad_norm": 3.0325095653533936, "learning_rate": 4.117793466358375e-06, "loss": 1.255, "step": 28446 }, { "epoch": 2.2293887147335423, "grad_norm": 2.4524996280670166, "learning_rate": 4.117489943819767e-06, "loss": 1.4022, "step": 28447 }, { "epoch": 2.2294670846394986, "grad_norm": 3.3953983783721924, "learning_rate": 4.117186424637774e-06, "loss": 1.0699, "step": 28448 }, { "epoch": 2.2295454545454545, "grad_norm": 13.973822593688965, "learning_rate": 4.116882908813551e-06, "loss": 1.8671, "step": 28449 }, { "epoch": 2.229623824451411, "grad_norm": 3.2108311653137207, "learning_rate": 4.116579396348253e-06, "loss": 1.2773, "step": 28450 }, { "epoch": 2.2297021943573667, "grad_norm": 8.86376953125, "learning_rate": 4.116275887243032e-06, "loss": 1.8889, "step": 28451 }, { "epoch": 2.229780564263323, "grad_norm": 3.2381813526153564, "learning_rate": 4.115972381499044e-06, "loss": 1.1882, "step": 28452 }, { "epoch": 2.229858934169279, "grad_norm": 2.9214484691619873, "learning_rate": 4.115668879117442e-06, "loss": 1.4421, "step": 28453 }, { "epoch": 2.2299373040752353, "grad_norm": 2.0037002563476562, "learning_rate": 4.115365380099384e-06, "loss": 1.2936, "step": 28454 }, { "epoch": 2.230015673981191, "grad_norm": 4.442870616912842, "learning_rate": 4.115061884446018e-06, "loss": 1.4839, "step": 28455 }, { "epoch": 2.2300940438871475, "grad_norm": 3.1470401287078857, "learning_rate": 4.114758392158504e-06, "loss": 1.2518, "step": 28456 }, { "epoch": 2.2301724137931034, "grad_norm": 3.6152422428131104, "learning_rate": 4.114454903237994e-06, "loss": 1.2813, "step": 28457 }, { "epoch": 2.2302507836990597, "grad_norm": 2.265780210494995, "learning_rate": 4.114151417685644e-06, "loss": 1.4443, "step": 28458 }, { "epoch": 2.2303291536050156, "grad_norm": 3.3898863792419434, "learning_rate": 4.113847935502606e-06, "loss": 1.2647, "step": 28459 }, { "epoch": 2.230407523510972, "grad_norm": 3.8961989879608154, "learning_rate": 4.1135444566900345e-06, "loss": 1.5838, "step": 28460 }, { "epoch": 2.230485893416928, "grad_norm": 5.081026077270508, "learning_rate": 4.113240981249086e-06, "loss": 1.2229, "step": 28461 }, { "epoch": 2.230564263322884, "grad_norm": 2.629035711288452, "learning_rate": 4.112937509180913e-06, "loss": 1.2987, "step": 28462 }, { "epoch": 2.23064263322884, "grad_norm": 2.284595489501953, "learning_rate": 4.11263404048667e-06, "loss": 1.4516, "step": 28463 }, { "epoch": 2.2307210031347964, "grad_norm": 7.150533676147461, "learning_rate": 4.112330575167511e-06, "loss": 0.845, "step": 28464 }, { "epoch": 2.2307993730407523, "grad_norm": 4.302741527557373, "learning_rate": 4.112027113224592e-06, "loss": 1.6339, "step": 28465 }, { "epoch": 2.2308777429467086, "grad_norm": 2.2344064712524414, "learning_rate": 4.111723654659064e-06, "loss": 1.2591, "step": 28466 }, { "epoch": 2.2309561128526645, "grad_norm": 2.589350461959839, "learning_rate": 4.111420199472082e-06, "loss": 1.1785, "step": 28467 }, { "epoch": 2.231034482758621, "grad_norm": 3.0888681411743164, "learning_rate": 4.1111167476648025e-06, "loss": 1.3414, "step": 28468 }, { "epoch": 2.2311128526645767, "grad_norm": 6.663400650024414, "learning_rate": 4.110813299238379e-06, "loss": 0.8737, "step": 28469 }, { "epoch": 2.231191222570533, "grad_norm": 6.146576404571533, "learning_rate": 4.110509854193965e-06, "loss": 1.5973, "step": 28470 }, { "epoch": 2.231269592476489, "grad_norm": 3.1259636878967285, "learning_rate": 4.110206412532714e-06, "loss": 1.1751, "step": 28471 }, { "epoch": 2.2313479623824453, "grad_norm": 4.120420932769775, "learning_rate": 4.109902974255781e-06, "loss": 0.6863, "step": 28472 }, { "epoch": 2.231426332288401, "grad_norm": 4.617086410522461, "learning_rate": 4.10959953936432e-06, "loss": 1.0081, "step": 28473 }, { "epoch": 2.2315047021943575, "grad_norm": 4.420201778411865, "learning_rate": 4.109296107859484e-06, "loss": 1.0372, "step": 28474 }, { "epoch": 2.2315830721003134, "grad_norm": 3.2448105812072754, "learning_rate": 4.108992679742429e-06, "loss": 1.5096, "step": 28475 }, { "epoch": 2.2316614420062697, "grad_norm": 6.125192642211914, "learning_rate": 4.108689255014309e-06, "loss": 1.3465, "step": 28476 }, { "epoch": 2.2317398119122256, "grad_norm": 3.2013614177703857, "learning_rate": 4.108385833676276e-06, "loss": 1.6426, "step": 28477 }, { "epoch": 2.231818181818182, "grad_norm": 3.137986660003662, "learning_rate": 4.108082415729485e-06, "loss": 1.3365, "step": 28478 }, { "epoch": 2.231896551724138, "grad_norm": 4.024570941925049, "learning_rate": 4.1077790011750904e-06, "loss": 1.8106, "step": 28479 }, { "epoch": 2.231974921630094, "grad_norm": 2.692124128341675, "learning_rate": 4.107475590014249e-06, "loss": 1.3372, "step": 28480 }, { "epoch": 2.23205329153605, "grad_norm": 4.660836219787598, "learning_rate": 4.107172182248109e-06, "loss": 1.3434, "step": 28481 }, { "epoch": 2.2321316614420064, "grad_norm": 3.4762513637542725, "learning_rate": 4.106868777877828e-06, "loss": 1.2385, "step": 28482 }, { "epoch": 2.2322100313479623, "grad_norm": 2.6152830123901367, "learning_rate": 4.10656537690456e-06, "loss": 1.4543, "step": 28483 }, { "epoch": 2.2322884012539186, "grad_norm": 3.7318625450134277, "learning_rate": 4.106261979329459e-06, "loss": 1.4089, "step": 28484 }, { "epoch": 2.2323667711598745, "grad_norm": 4.804696083068848, "learning_rate": 4.1059585851536765e-06, "loss": 1.7753, "step": 28485 }, { "epoch": 2.232445141065831, "grad_norm": 5.5032267570495605, "learning_rate": 4.105655194378369e-06, "loss": 1.4843, "step": 28486 }, { "epoch": 2.2325235109717867, "grad_norm": 4.536172389984131, "learning_rate": 4.105351807004691e-06, "loss": 1.2417, "step": 28487 }, { "epoch": 2.232601880877743, "grad_norm": 3.903139114379883, "learning_rate": 4.105048423033795e-06, "loss": 1.5369, "step": 28488 }, { "epoch": 2.232680250783699, "grad_norm": 3.9105186462402344, "learning_rate": 4.104745042466834e-06, "loss": 1.0704, "step": 28489 }, { "epoch": 2.2327586206896552, "grad_norm": 3.1877968311309814, "learning_rate": 4.104441665304962e-06, "loss": 1.2614, "step": 28490 }, { "epoch": 2.232836990595611, "grad_norm": 4.088478088378906, "learning_rate": 4.104138291549338e-06, "loss": 1.1404, "step": 28491 }, { "epoch": 2.2329153605015675, "grad_norm": 4.7719011306762695, "learning_rate": 4.103834921201107e-06, "loss": 1.3389, "step": 28492 }, { "epoch": 2.2329937304075234, "grad_norm": 3.571652889251709, "learning_rate": 4.1035315542614305e-06, "loss": 1.3657, "step": 28493 }, { "epoch": 2.2330721003134797, "grad_norm": 4.757447242736816, "learning_rate": 4.103228190731459e-06, "loss": 1.9096, "step": 28494 }, { "epoch": 2.2331504702194356, "grad_norm": 3.2123913764953613, "learning_rate": 4.102924830612348e-06, "loss": 0.9168, "step": 28495 }, { "epoch": 2.233228840125392, "grad_norm": 2.922506809234619, "learning_rate": 4.1026214739052484e-06, "loss": 1.4256, "step": 28496 }, { "epoch": 2.233307210031348, "grad_norm": 3.1347572803497314, "learning_rate": 4.102318120611315e-06, "loss": 1.5239, "step": 28497 }, { "epoch": 2.233385579937304, "grad_norm": 3.1464576721191406, "learning_rate": 4.102014770731703e-06, "loss": 1.4708, "step": 28498 }, { "epoch": 2.23346394984326, "grad_norm": 3.310091972351074, "learning_rate": 4.101711424267568e-06, "loss": 1.1631, "step": 28499 }, { "epoch": 2.2335423197492164, "grad_norm": 8.952268600463867, "learning_rate": 4.101408081220059e-06, "loss": 0.6796, "step": 28500 }, { "epoch": 2.2336206896551722, "grad_norm": 4.3538618087768555, "learning_rate": 4.101104741590333e-06, "loss": 1.4735, "step": 28501 }, { "epoch": 2.2336990595611286, "grad_norm": 2.1809003353118896, "learning_rate": 4.100801405379543e-06, "loss": 1.3973, "step": 28502 }, { "epoch": 2.2337774294670845, "grad_norm": 2.015662908554077, "learning_rate": 4.100498072588843e-06, "loss": 1.1773, "step": 28503 }, { "epoch": 2.233855799373041, "grad_norm": 3.961660146713257, "learning_rate": 4.100194743219384e-06, "loss": 1.3058, "step": 28504 }, { "epoch": 2.2339341692789967, "grad_norm": 2.6922073364257812, "learning_rate": 4.099891417272324e-06, "loss": 1.2506, "step": 28505 }, { "epoch": 2.234012539184953, "grad_norm": 2.3607702255249023, "learning_rate": 4.099588094748815e-06, "loss": 1.33, "step": 28506 }, { "epoch": 2.234090909090909, "grad_norm": 2.3430349826812744, "learning_rate": 4.099284775650009e-06, "loss": 1.624, "step": 28507 }, { "epoch": 2.2341692789968652, "grad_norm": 4.323636531829834, "learning_rate": 4.0989814599770625e-06, "loss": 1.1567, "step": 28508 }, { "epoch": 2.234247648902821, "grad_norm": 3.4559807777404785, "learning_rate": 4.098678147731126e-06, "loss": 1.314, "step": 28509 }, { "epoch": 2.2343260188087775, "grad_norm": 4.331997394561768, "learning_rate": 4.098374838913357e-06, "loss": 0.9489, "step": 28510 }, { "epoch": 2.2344043887147333, "grad_norm": 2.481356620788574, "learning_rate": 4.098071533524905e-06, "loss": 1.0205, "step": 28511 }, { "epoch": 2.2344827586206897, "grad_norm": 5.1713666915893555, "learning_rate": 4.097768231566926e-06, "loss": 1.2564, "step": 28512 }, { "epoch": 2.2345611285266456, "grad_norm": 9.018072128295898, "learning_rate": 4.0974649330405745e-06, "loss": 1.0658, "step": 28513 }, { "epoch": 2.234639498432602, "grad_norm": 4.097751140594482, "learning_rate": 4.097161637947002e-06, "loss": 1.1056, "step": 28514 }, { "epoch": 2.234717868338558, "grad_norm": 2.0444793701171875, "learning_rate": 4.0968583462873635e-06, "loss": 1.342, "step": 28515 }, { "epoch": 2.234796238244514, "grad_norm": 2.5957138538360596, "learning_rate": 4.09655505806281e-06, "loss": 1.4315, "step": 28516 }, { "epoch": 2.23487460815047, "grad_norm": 3.0237014293670654, "learning_rate": 4.096251773274501e-06, "loss": 1.2599, "step": 28517 }, { "epoch": 2.2349529780564263, "grad_norm": 3.468852996826172, "learning_rate": 4.095948491923582e-06, "loss": 1.3471, "step": 28518 }, { "epoch": 2.2350313479623827, "grad_norm": 3.022622585296631, "learning_rate": 4.0956452140112115e-06, "loss": 1.03, "step": 28519 }, { "epoch": 2.2351097178683386, "grad_norm": 6.289475440979004, "learning_rate": 4.095341939538543e-06, "loss": 1.2927, "step": 28520 }, { "epoch": 2.2351880877742945, "grad_norm": 3.842658281326294, "learning_rate": 4.09503866850673e-06, "loss": 1.4214, "step": 28521 }, { "epoch": 2.235266457680251, "grad_norm": 3.3207168579101562, "learning_rate": 4.094735400916924e-06, "loss": 1.5297, "step": 28522 }, { "epoch": 2.235344827586207, "grad_norm": 18.324472427368164, "learning_rate": 4.094432136770278e-06, "loss": 1.8867, "step": 28523 }, { "epoch": 2.235423197492163, "grad_norm": 2.757797956466675, "learning_rate": 4.094128876067949e-06, "loss": 1.2308, "step": 28524 }, { "epoch": 2.235501567398119, "grad_norm": 2.660645008087158, "learning_rate": 4.093825618811088e-06, "loss": 1.2461, "step": 28525 }, { "epoch": 2.2355799373040752, "grad_norm": 2.1816139221191406, "learning_rate": 4.093522365000849e-06, "loss": 1.2876, "step": 28526 }, { "epoch": 2.2356583072100316, "grad_norm": 3.1371729373931885, "learning_rate": 4.093219114638385e-06, "loss": 1.5515, "step": 28527 }, { "epoch": 2.2357366771159874, "grad_norm": 3.1029350757598877, "learning_rate": 4.092915867724851e-06, "loss": 1.4266, "step": 28528 }, { "epoch": 2.235815047021944, "grad_norm": 3.7026774883270264, "learning_rate": 4.092612624261398e-06, "loss": 1.2511, "step": 28529 }, { "epoch": 2.2358934169278997, "grad_norm": 2.6449837684631348, "learning_rate": 4.092309384249179e-06, "loss": 1.2851, "step": 28530 }, { "epoch": 2.235971786833856, "grad_norm": 2.5084760189056396, "learning_rate": 4.0920061476893505e-06, "loss": 1.372, "step": 28531 }, { "epoch": 2.236050156739812, "grad_norm": 2.161482334136963, "learning_rate": 4.091702914583065e-06, "loss": 1.4764, "step": 28532 }, { "epoch": 2.236128526645768, "grad_norm": 10.389763832092285, "learning_rate": 4.091399684931474e-06, "loss": 1.2817, "step": 28533 }, { "epoch": 2.236206896551724, "grad_norm": 3.6067817211151123, "learning_rate": 4.091096458735733e-06, "loss": 1.1834, "step": 28534 }, { "epoch": 2.2362852664576804, "grad_norm": 7.053145885467529, "learning_rate": 4.090793235996992e-06, "loss": 1.3131, "step": 28535 }, { "epoch": 2.2363636363636363, "grad_norm": 2.3776280879974365, "learning_rate": 4.09049001671641e-06, "loss": 1.4589, "step": 28536 }, { "epoch": 2.2364420062695927, "grad_norm": 4.276851177215576, "learning_rate": 4.090186800895133e-06, "loss": 1.5149, "step": 28537 }, { "epoch": 2.2365203761755486, "grad_norm": 3.485753297805786, "learning_rate": 4.08988358853432e-06, "loss": 1.0183, "step": 28538 }, { "epoch": 2.236598746081505, "grad_norm": 2.810002565383911, "learning_rate": 4.089580379635122e-06, "loss": 1.088, "step": 28539 }, { "epoch": 2.2366771159874608, "grad_norm": 2.939737319946289, "learning_rate": 4.089277174198694e-06, "loss": 1.3613, "step": 28540 }, { "epoch": 2.236755485893417, "grad_norm": 2.679839849472046, "learning_rate": 4.088973972226186e-06, "loss": 1.1118, "step": 28541 }, { "epoch": 2.236833855799373, "grad_norm": 2.2671585083007812, "learning_rate": 4.088670773718753e-06, "loss": 1.3676, "step": 28542 }, { "epoch": 2.2369122257053293, "grad_norm": 2.9446582794189453, "learning_rate": 4.088367578677551e-06, "loss": 1.1501, "step": 28543 }, { "epoch": 2.236990595611285, "grad_norm": 3.7841603755950928, "learning_rate": 4.088064387103727e-06, "loss": 1.3398, "step": 28544 }, { "epoch": 2.2370689655172415, "grad_norm": 2.052943706512451, "learning_rate": 4.087761198998439e-06, "loss": 1.3524, "step": 28545 }, { "epoch": 2.2371473354231974, "grad_norm": 3.8799386024475098, "learning_rate": 4.08745801436284e-06, "loss": 1.1373, "step": 28546 }, { "epoch": 2.2372257053291538, "grad_norm": 3.3841075897216797, "learning_rate": 4.087154833198082e-06, "loss": 1.3811, "step": 28547 }, { "epoch": 2.2373040752351097, "grad_norm": 2.791464328765869, "learning_rate": 4.086851655505316e-06, "loss": 1.2225, "step": 28548 }, { "epoch": 2.237382445141066, "grad_norm": 2.5941760540008545, "learning_rate": 4.086548481285698e-06, "loss": 1.2772, "step": 28549 }, { "epoch": 2.237460815047022, "grad_norm": 2.376437187194824, "learning_rate": 4.086245310540381e-06, "loss": 0.8926, "step": 28550 }, { "epoch": 2.237539184952978, "grad_norm": 5.794025421142578, "learning_rate": 4.085942143270518e-06, "loss": 1.1869, "step": 28551 }, { "epoch": 2.237617554858934, "grad_norm": 5.555271148681641, "learning_rate": 4.085638979477261e-06, "loss": 1.4457, "step": 28552 }, { "epoch": 2.2376959247648904, "grad_norm": 2.895210027694702, "learning_rate": 4.085335819161763e-06, "loss": 1.1409, "step": 28553 }, { "epoch": 2.2377742946708463, "grad_norm": 2.7006595134735107, "learning_rate": 4.085032662325178e-06, "loss": 1.4466, "step": 28554 }, { "epoch": 2.2378526645768027, "grad_norm": 2.480976104736328, "learning_rate": 4.084729508968661e-06, "loss": 1.118, "step": 28555 }, { "epoch": 2.2379310344827585, "grad_norm": 2.8233325481414795, "learning_rate": 4.08442635909336e-06, "loss": 1.2587, "step": 28556 }, { "epoch": 2.238009404388715, "grad_norm": 3.684316635131836, "learning_rate": 4.084123212700431e-06, "loss": 1.0298, "step": 28557 }, { "epoch": 2.2380877742946708, "grad_norm": 4.010429382324219, "learning_rate": 4.083820069791029e-06, "loss": 1.6255, "step": 28558 }, { "epoch": 2.238166144200627, "grad_norm": 3.0801303386688232, "learning_rate": 4.083516930366304e-06, "loss": 1.4686, "step": 28559 }, { "epoch": 2.238244514106583, "grad_norm": 4.521563529968262, "learning_rate": 4.0832137944274094e-06, "loss": 1.4543, "step": 28560 }, { "epoch": 2.2383228840125393, "grad_norm": 2.695296049118042, "learning_rate": 4.082910661975497e-06, "loss": 1.4071, "step": 28561 }, { "epoch": 2.238401253918495, "grad_norm": 2.9843316078186035, "learning_rate": 4.082607533011725e-06, "loss": 1.2882, "step": 28562 }, { "epoch": 2.2384796238244515, "grad_norm": 3.3238115310668945, "learning_rate": 4.08230440753724e-06, "loss": 1.5543, "step": 28563 }, { "epoch": 2.2385579937304074, "grad_norm": 3.4427597522735596, "learning_rate": 4.0820012855531984e-06, "loss": 1.1953, "step": 28564 }, { "epoch": 2.2386363636363638, "grad_norm": 2.734941244125366, "learning_rate": 4.081698167060753e-06, "loss": 1.1614, "step": 28565 }, { "epoch": 2.2387147335423196, "grad_norm": 2.431032180786133, "learning_rate": 4.0813950520610565e-06, "loss": 1.5731, "step": 28566 }, { "epoch": 2.238793103448276, "grad_norm": 3.4200870990753174, "learning_rate": 4.08109194055526e-06, "loss": 1.0698, "step": 28567 }, { "epoch": 2.238871473354232, "grad_norm": 2.5390098094940186, "learning_rate": 4.080788832544518e-06, "loss": 1.2481, "step": 28568 }, { "epoch": 2.238949843260188, "grad_norm": 3.555421829223633, "learning_rate": 4.080485728029983e-06, "loss": 1.247, "step": 28569 }, { "epoch": 2.239028213166144, "grad_norm": 4.137604713439941, "learning_rate": 4.080182627012809e-06, "loss": 1.4136, "step": 28570 }, { "epoch": 2.2391065830721004, "grad_norm": 3.42897629737854, "learning_rate": 4.0798795294941474e-06, "loss": 1.0672, "step": 28571 }, { "epoch": 2.2391849529780563, "grad_norm": 3.9591264724731445, "learning_rate": 4.079576435475152e-06, "loss": 1.2848, "step": 28572 }, { "epoch": 2.2392633228840126, "grad_norm": 4.196384429931641, "learning_rate": 4.079273344956975e-06, "loss": 1.2283, "step": 28573 }, { "epoch": 2.2393416927899685, "grad_norm": 2.3688247203826904, "learning_rate": 4.0789702579407676e-06, "loss": 0.9862, "step": 28574 }, { "epoch": 2.239420062695925, "grad_norm": 3.4106602668762207, "learning_rate": 4.078667174427685e-06, "loss": 1.2777, "step": 28575 }, { "epoch": 2.2394984326018808, "grad_norm": 8.470844268798828, "learning_rate": 4.07836409441888e-06, "loss": 1.2792, "step": 28576 }, { "epoch": 2.239576802507837, "grad_norm": 2.537907361984253, "learning_rate": 4.078061017915505e-06, "loss": 1.5448, "step": 28577 }, { "epoch": 2.239655172413793, "grad_norm": 2.8913562297821045, "learning_rate": 4.077757944918711e-06, "loss": 1.3642, "step": 28578 }, { "epoch": 2.2397335423197493, "grad_norm": 3.3078019618988037, "learning_rate": 4.077454875429653e-06, "loss": 1.3727, "step": 28579 }, { "epoch": 2.239811912225705, "grad_norm": 4.6106486320495605, "learning_rate": 4.077151809449481e-06, "loss": 0.8388, "step": 28580 }, { "epoch": 2.2398902821316615, "grad_norm": 2.890005111694336, "learning_rate": 4.076848746979352e-06, "loss": 1.3079, "step": 28581 }, { "epoch": 2.2399686520376174, "grad_norm": 5.028018474578857, "learning_rate": 4.076545688020415e-06, "loss": 1.2261, "step": 28582 }, { "epoch": 2.2400470219435737, "grad_norm": 3.3086838722229004, "learning_rate": 4.076242632573824e-06, "loss": 1.5063, "step": 28583 }, { "epoch": 2.2401253918495296, "grad_norm": 3.0685930252075195, "learning_rate": 4.075939580640733e-06, "loss": 1.2136, "step": 28584 }, { "epoch": 2.240203761755486, "grad_norm": 2.6862807273864746, "learning_rate": 4.0756365322222915e-06, "loss": 1.3249, "step": 28585 }, { "epoch": 2.240282131661442, "grad_norm": 2.840268611907959, "learning_rate": 4.075333487319653e-06, "loss": 1.2007, "step": 28586 }, { "epoch": 2.240360501567398, "grad_norm": 1.996624231338501, "learning_rate": 4.0750304459339726e-06, "loss": 1.3906, "step": 28587 }, { "epoch": 2.240438871473354, "grad_norm": 2.755434513092041, "learning_rate": 4.074727408066401e-06, "loss": 1.1204, "step": 28588 }, { "epoch": 2.2405172413793104, "grad_norm": 3.7584004402160645, "learning_rate": 4.074424373718091e-06, "loss": 1.0128, "step": 28589 }, { "epoch": 2.2405956112852663, "grad_norm": 2.449032783508301, "learning_rate": 4.074121342890195e-06, "loss": 1.2963, "step": 28590 }, { "epoch": 2.2406739811912226, "grad_norm": 2.4810478687286377, "learning_rate": 4.073818315583865e-06, "loss": 1.5906, "step": 28591 }, { "epoch": 2.2407523510971785, "grad_norm": 4.719036102294922, "learning_rate": 4.073515291800258e-06, "loss": 1.4195, "step": 28592 }, { "epoch": 2.240830721003135, "grad_norm": 4.076385498046875, "learning_rate": 4.073212271540519e-06, "loss": 1.5099, "step": 28593 }, { "epoch": 2.2409090909090907, "grad_norm": 3.006013870239258, "learning_rate": 4.072909254805806e-06, "loss": 1.7444, "step": 28594 }, { "epoch": 2.240987460815047, "grad_norm": 2.927058696746826, "learning_rate": 4.07260624159727e-06, "loss": 1.3103, "step": 28595 }, { "epoch": 2.241065830721003, "grad_norm": 5.264503002166748, "learning_rate": 4.072303231916064e-06, "loss": 1.2563, "step": 28596 }, { "epoch": 2.2411442006269593, "grad_norm": 4.794107913970947, "learning_rate": 4.07200022576334e-06, "loss": 1.2519, "step": 28597 }, { "epoch": 2.241222570532915, "grad_norm": 2.7648091316223145, "learning_rate": 4.0716972231402484e-06, "loss": 1.5553, "step": 28598 }, { "epoch": 2.2413009404388715, "grad_norm": 8.045703887939453, "learning_rate": 4.071394224047948e-06, "loss": 1.0747, "step": 28599 }, { "epoch": 2.2413793103448274, "grad_norm": 3.5612740516662598, "learning_rate": 4.0710912284875825e-06, "loss": 1.5584, "step": 28600 }, { "epoch": 2.2414576802507837, "grad_norm": 3.766524076461792, "learning_rate": 4.070788236460311e-06, "loss": 1.2984, "step": 28601 }, { "epoch": 2.2415360501567396, "grad_norm": 4.38845682144165, "learning_rate": 4.070485247967283e-06, "loss": 0.9731, "step": 28602 }, { "epoch": 2.241614420062696, "grad_norm": 2.783069610595703, "learning_rate": 4.070182263009653e-06, "loss": 1.3919, "step": 28603 }, { "epoch": 2.241692789968652, "grad_norm": 2.466815948486328, "learning_rate": 4.069879281588571e-06, "loss": 1.1934, "step": 28604 }, { "epoch": 2.241771159874608, "grad_norm": 2.559950590133667, "learning_rate": 4.06957630370519e-06, "loss": 1.2435, "step": 28605 }, { "epoch": 2.241849529780564, "grad_norm": 3.795851230621338, "learning_rate": 4.0692733293606635e-06, "loss": 1.4221, "step": 28606 }, { "epoch": 2.2419278996865204, "grad_norm": 2.6812400817871094, "learning_rate": 4.068970358556144e-06, "loss": 1.4564, "step": 28607 }, { "epoch": 2.2420062695924763, "grad_norm": 7.114945411682129, "learning_rate": 4.068667391292782e-06, "loss": 0.9624, "step": 28608 }, { "epoch": 2.2420846394984326, "grad_norm": 2.531815767288208, "learning_rate": 4.068364427571732e-06, "loss": 1.1788, "step": 28609 }, { "epoch": 2.2421630094043885, "grad_norm": 4.770381450653076, "learning_rate": 4.068061467394142e-06, "loss": 1.7665, "step": 28610 }, { "epoch": 2.242241379310345, "grad_norm": 3.836873769760132, "learning_rate": 4.0677585107611725e-06, "loss": 1.3018, "step": 28611 }, { "epoch": 2.2423197492163007, "grad_norm": 3.1225192546844482, "learning_rate": 4.067455557673967e-06, "loss": 1.1856, "step": 28612 }, { "epoch": 2.242398119122257, "grad_norm": 2.4355788230895996, "learning_rate": 4.067152608133683e-06, "loss": 1.39, "step": 28613 }, { "epoch": 2.2424764890282134, "grad_norm": 2.059581995010376, "learning_rate": 4.066849662141472e-06, "loss": 1.0001, "step": 28614 }, { "epoch": 2.2425548589341693, "grad_norm": 5.4406890869140625, "learning_rate": 4.066546719698484e-06, "loss": 1.6574, "step": 28615 }, { "epoch": 2.242633228840125, "grad_norm": 2.9637644290924072, "learning_rate": 4.066243780805873e-06, "loss": 1.1752, "step": 28616 }, { "epoch": 2.2427115987460815, "grad_norm": 2.766920804977417, "learning_rate": 4.065940845464791e-06, "loss": 1.1945, "step": 28617 }, { "epoch": 2.242789968652038, "grad_norm": 2.501011371612549, "learning_rate": 4.0656379136763925e-06, "loss": 1.1107, "step": 28618 }, { "epoch": 2.2428683385579937, "grad_norm": 5.443558216094971, "learning_rate": 4.065334985441825e-06, "loss": 0.7734, "step": 28619 }, { "epoch": 2.2429467084639496, "grad_norm": 2.9227685928344727, "learning_rate": 4.065032060762244e-06, "loss": 1.5557, "step": 28620 }, { "epoch": 2.243025078369906, "grad_norm": 2.233168363571167, "learning_rate": 4.0647291396388e-06, "loss": 1.2831, "step": 28621 }, { "epoch": 2.2431034482758623, "grad_norm": 2.3040928840637207, "learning_rate": 4.064426222072648e-06, "loss": 1.3004, "step": 28622 }, { "epoch": 2.243181818181818, "grad_norm": 5.15696382522583, "learning_rate": 4.064123308064937e-06, "loss": 1.7878, "step": 28623 }, { "epoch": 2.2432601880877745, "grad_norm": 2.9198338985443115, "learning_rate": 4.063820397616819e-06, "loss": 1.4083, "step": 28624 }, { "epoch": 2.2433385579937304, "grad_norm": 3.865652322769165, "learning_rate": 4.06351749072945e-06, "loss": 1.3021, "step": 28625 }, { "epoch": 2.2434169278996867, "grad_norm": 3.792675495147705, "learning_rate": 4.063214587403977e-06, "loss": 1.181, "step": 28626 }, { "epoch": 2.2434952978056426, "grad_norm": 11.878755569458008, "learning_rate": 4.062911687641556e-06, "loss": 1.7071, "step": 28627 }, { "epoch": 2.243573667711599, "grad_norm": 2.577437400817871, "learning_rate": 4.062608791443336e-06, "loss": 1.4699, "step": 28628 }, { "epoch": 2.243652037617555, "grad_norm": 7.337409496307373, "learning_rate": 4.0623058988104725e-06, "loss": 1.9095, "step": 28629 }, { "epoch": 2.243730407523511, "grad_norm": 2.160291910171509, "learning_rate": 4.062003009744115e-06, "loss": 1.7067, "step": 28630 }, { "epoch": 2.243808777429467, "grad_norm": 3.586545467376709, "learning_rate": 4.0617001242454155e-06, "loss": 1.3193, "step": 28631 }, { "epoch": 2.2438871473354234, "grad_norm": 3.313290596008301, "learning_rate": 4.061397242315527e-06, "loss": 1.2295, "step": 28632 }, { "epoch": 2.2439655172413793, "grad_norm": 2.1521472930908203, "learning_rate": 4.061094363955603e-06, "loss": 1.1982, "step": 28633 }, { "epoch": 2.2440438871473356, "grad_norm": 4.015810966491699, "learning_rate": 4.060791489166792e-06, "loss": 1.4285, "step": 28634 }, { "epoch": 2.2441222570532915, "grad_norm": 5.779170036315918, "learning_rate": 4.060488617950248e-06, "loss": 1.3781, "step": 28635 }, { "epoch": 2.244200626959248, "grad_norm": 2.5750532150268555, "learning_rate": 4.060185750307122e-06, "loss": 1.2854, "step": 28636 }, { "epoch": 2.2442789968652037, "grad_norm": 4.359306812286377, "learning_rate": 4.05988288623857e-06, "loss": 0.911, "step": 28637 }, { "epoch": 2.24435736677116, "grad_norm": 3.211150884628296, "learning_rate": 4.059580025745737e-06, "loss": 1.5013, "step": 28638 }, { "epoch": 2.244435736677116, "grad_norm": 6.7478556632995605, "learning_rate": 4.05927716882978e-06, "loss": 1.3082, "step": 28639 }, { "epoch": 2.2445141065830723, "grad_norm": 3.085806369781494, "learning_rate": 4.0589743154918506e-06, "loss": 1.4409, "step": 28640 }, { "epoch": 2.244592476489028, "grad_norm": 2.6182644367218018, "learning_rate": 4.0586714657330985e-06, "loss": 1.21, "step": 28641 }, { "epoch": 2.2446708463949845, "grad_norm": 2.0906620025634766, "learning_rate": 4.058368619554676e-06, "loss": 1.3245, "step": 28642 }, { "epoch": 2.2447492163009404, "grad_norm": 25.466516494750977, "learning_rate": 4.058065776957735e-06, "loss": 0.6594, "step": 28643 }, { "epoch": 2.2448275862068967, "grad_norm": 2.9327402114868164, "learning_rate": 4.057762937943431e-06, "loss": 1.3644, "step": 28644 }, { "epoch": 2.2449059561128526, "grad_norm": 6.129795074462891, "learning_rate": 4.05746010251291e-06, "loss": 1.2923, "step": 28645 }, { "epoch": 2.244984326018809, "grad_norm": 3.1004116535186768, "learning_rate": 4.057157270667327e-06, "loss": 1.0691, "step": 28646 }, { "epoch": 2.245062695924765, "grad_norm": 10.161396026611328, "learning_rate": 4.056854442407834e-06, "loss": 1.6869, "step": 28647 }, { "epoch": 2.245141065830721, "grad_norm": 2.019838809967041, "learning_rate": 4.056551617735583e-06, "loss": 1.4095, "step": 28648 }, { "epoch": 2.245219435736677, "grad_norm": 3.838120698928833, "learning_rate": 4.056248796651725e-06, "loss": 1.1245, "step": 28649 }, { "epoch": 2.2452978056426334, "grad_norm": 3.3805885314941406, "learning_rate": 4.0559459791574105e-06, "loss": 1.549, "step": 28650 }, { "epoch": 2.2453761755485893, "grad_norm": 3.192434310913086, "learning_rate": 4.055643165253793e-06, "loss": 1.4662, "step": 28651 }, { "epoch": 2.2454545454545456, "grad_norm": 2.8836965560913086, "learning_rate": 4.055340354942025e-06, "loss": 1.261, "step": 28652 }, { "epoch": 2.2455329153605015, "grad_norm": 2.7805538177490234, "learning_rate": 4.055037548223256e-06, "loss": 1.3759, "step": 28653 }, { "epoch": 2.245611285266458, "grad_norm": 4.781154632568359, "learning_rate": 4.054734745098639e-06, "loss": 1.5683, "step": 28654 }, { "epoch": 2.2456896551724137, "grad_norm": 2.957277536392212, "learning_rate": 4.054431945569326e-06, "loss": 1.6671, "step": 28655 }, { "epoch": 2.24576802507837, "grad_norm": 6.413886070251465, "learning_rate": 4.054129149636468e-06, "loss": 1.4586, "step": 28656 }, { "epoch": 2.245846394984326, "grad_norm": 5.155442237854004, "learning_rate": 4.053826357301216e-06, "loss": 1.6109, "step": 28657 }, { "epoch": 2.2459247648902823, "grad_norm": 2.7904603481292725, "learning_rate": 4.053523568564723e-06, "loss": 0.9106, "step": 28658 }, { "epoch": 2.246003134796238, "grad_norm": 3.2071447372436523, "learning_rate": 4.053220783428141e-06, "loss": 1.2498, "step": 28659 }, { "epoch": 2.2460815047021945, "grad_norm": 2.3673746585845947, "learning_rate": 4.0529180018926204e-06, "loss": 1.1662, "step": 28660 }, { "epoch": 2.2461598746081504, "grad_norm": 2.5768864154815674, "learning_rate": 4.052615223959313e-06, "loss": 1.5134, "step": 28661 }, { "epoch": 2.2462382445141067, "grad_norm": 3.7046782970428467, "learning_rate": 4.05231244962937e-06, "loss": 1.3346, "step": 28662 }, { "epoch": 2.2463166144200626, "grad_norm": 3.54240345954895, "learning_rate": 4.052009678903946e-06, "loss": 1.5245, "step": 28663 }, { "epoch": 2.246394984326019, "grad_norm": 2.9367125034332275, "learning_rate": 4.051706911784188e-06, "loss": 1.5596, "step": 28664 }, { "epoch": 2.246473354231975, "grad_norm": 4.078030586242676, "learning_rate": 4.0514041482712505e-06, "loss": 1.4952, "step": 28665 }, { "epoch": 2.246551724137931, "grad_norm": 6.577546119689941, "learning_rate": 4.051101388366284e-06, "loss": 0.9893, "step": 28666 }, { "epoch": 2.246630094043887, "grad_norm": 3.867610216140747, "learning_rate": 4.050798632070442e-06, "loss": 1.6787, "step": 28667 }, { "epoch": 2.2467084639498434, "grad_norm": 2.8596127033233643, "learning_rate": 4.050495879384873e-06, "loss": 1.4227, "step": 28668 }, { "epoch": 2.2467868338557992, "grad_norm": 8.178245544433594, "learning_rate": 4.050193130310729e-06, "loss": 1.7248, "step": 28669 }, { "epoch": 2.2468652037617556, "grad_norm": 4.041194915771484, "learning_rate": 4.049890384849165e-06, "loss": 1.0947, "step": 28670 }, { "epoch": 2.2469435736677115, "grad_norm": 2.5211474895477295, "learning_rate": 4.0495876430013285e-06, "loss": 1.6145, "step": 28671 }, { "epoch": 2.247021943573668, "grad_norm": 7.33946418762207, "learning_rate": 4.049284904768372e-06, "loss": 0.9604, "step": 28672 }, { "epoch": 2.2471003134796237, "grad_norm": 3.0895731449127197, "learning_rate": 4.048982170151449e-06, "loss": 1.512, "step": 28673 }, { "epoch": 2.24717868338558, "grad_norm": 5.996201992034912, "learning_rate": 4.048679439151709e-06, "loss": 1.4589, "step": 28674 }, { "epoch": 2.247257053291536, "grad_norm": 3.5954651832580566, "learning_rate": 4.048376711770302e-06, "loss": 1.1036, "step": 28675 }, { "epoch": 2.2473354231974922, "grad_norm": 3.3378541469573975, "learning_rate": 4.048073988008383e-06, "loss": 1.4452, "step": 28676 }, { "epoch": 2.247413793103448, "grad_norm": 2.582394599914551, "learning_rate": 4.0477712678671005e-06, "loss": 1.5134, "step": 28677 }, { "epoch": 2.2474921630094045, "grad_norm": 3.427795648574829, "learning_rate": 4.0474685513476085e-06, "loss": 1.1557, "step": 28678 }, { "epoch": 2.2475705329153604, "grad_norm": 3.1756482124328613, "learning_rate": 4.047165838451056e-06, "loss": 1.7307, "step": 28679 }, { "epoch": 2.2476489028213167, "grad_norm": 3.7911977767944336, "learning_rate": 4.046863129178596e-06, "loss": 1.4151, "step": 28680 }, { "epoch": 2.2477272727272726, "grad_norm": 2.9997899532318115, "learning_rate": 4.046560423531378e-06, "loss": 1.2894, "step": 28681 }, { "epoch": 2.247805642633229, "grad_norm": 2.94340443611145, "learning_rate": 4.046257721510554e-06, "loss": 1.4032, "step": 28682 }, { "epoch": 2.247884012539185, "grad_norm": 2.179941415786743, "learning_rate": 4.045955023117276e-06, "loss": 1.4698, "step": 28683 }, { "epoch": 2.247962382445141, "grad_norm": 4.163334846496582, "learning_rate": 4.045652328352696e-06, "loss": 1.3045, "step": 28684 }, { "epoch": 2.248040752351097, "grad_norm": 3.0320725440979004, "learning_rate": 4.045349637217965e-06, "loss": 1.2023, "step": 28685 }, { "epoch": 2.2481191222570533, "grad_norm": 2.6656315326690674, "learning_rate": 4.045046949714232e-06, "loss": 1.1976, "step": 28686 }, { "epoch": 2.2481974921630092, "grad_norm": 3.9392900466918945, "learning_rate": 4.044744265842649e-06, "loss": 1.2449, "step": 28687 }, { "epoch": 2.2482758620689656, "grad_norm": 8.20882511138916, "learning_rate": 4.04444158560437e-06, "loss": 1.4203, "step": 28688 }, { "epoch": 2.2483542319749215, "grad_norm": 4.177700519561768, "learning_rate": 4.044138909000545e-06, "loss": 1.2105, "step": 28689 }, { "epoch": 2.248432601880878, "grad_norm": 4.027488708496094, "learning_rate": 4.0438362360323235e-06, "loss": 0.9563, "step": 28690 }, { "epoch": 2.2485109717868337, "grad_norm": 1.8461107015609741, "learning_rate": 4.043533566700858e-06, "loss": 1.3928, "step": 28691 }, { "epoch": 2.24858934169279, "grad_norm": 3.674738645553589, "learning_rate": 4.043230901007298e-06, "loss": 0.9461, "step": 28692 }, { "epoch": 2.248667711598746, "grad_norm": 3.979116916656494, "learning_rate": 4.0429282389528e-06, "loss": 1.3264, "step": 28693 }, { "epoch": 2.2487460815047022, "grad_norm": 6.462730407714844, "learning_rate": 4.042625580538508e-06, "loss": 1.062, "step": 28694 }, { "epoch": 2.248824451410658, "grad_norm": 4.301449775695801, "learning_rate": 4.0423229257655775e-06, "loss": 0.7555, "step": 28695 }, { "epoch": 2.2489028213166145, "grad_norm": 4.647109508514404, "learning_rate": 4.04202027463516e-06, "loss": 1.3059, "step": 28696 }, { "epoch": 2.2489811912225703, "grad_norm": 5.1880717277526855, "learning_rate": 4.041717627148405e-06, "loss": 2.1429, "step": 28697 }, { "epoch": 2.2490595611285267, "grad_norm": 3.9327540397644043, "learning_rate": 4.041414983306462e-06, "loss": 1.1874, "step": 28698 }, { "epoch": 2.2491379310344826, "grad_norm": 6.38027811050415, "learning_rate": 4.041112343110484e-06, "loss": 1.1619, "step": 28699 }, { "epoch": 2.249216300940439, "grad_norm": 12.611599922180176, "learning_rate": 4.040809706561626e-06, "loss": 1.1509, "step": 28700 }, { "epoch": 2.249294670846395, "grad_norm": 4.2714056968688965, "learning_rate": 4.040507073661031e-06, "loss": 1.1736, "step": 28701 }, { "epoch": 2.249373040752351, "grad_norm": 2.5820558071136475, "learning_rate": 4.040204444409855e-06, "loss": 1.5616, "step": 28702 }, { "epoch": 2.249451410658307, "grad_norm": 3.526989459991455, "learning_rate": 4.03990181880925e-06, "loss": 1.4677, "step": 28703 }, { "epoch": 2.2495297805642633, "grad_norm": 5.192986011505127, "learning_rate": 4.0395991968603655e-06, "loss": 1.2518, "step": 28704 }, { "epoch": 2.2496081504702197, "grad_norm": 7.235508441925049, "learning_rate": 4.0392965785643515e-06, "loss": 1.1784, "step": 28705 }, { "epoch": 2.2496865203761756, "grad_norm": 2.4960379600524902, "learning_rate": 4.038993963922358e-06, "loss": 1.2071, "step": 28706 }, { "epoch": 2.2497648902821314, "grad_norm": 3.667057514190674, "learning_rate": 4.038691352935539e-06, "loss": 1.1119, "step": 28707 }, { "epoch": 2.249843260188088, "grad_norm": 3.78135085105896, "learning_rate": 4.038388745605046e-06, "loss": 1.4791, "step": 28708 }, { "epoch": 2.249921630094044, "grad_norm": 2.7700421810150146, "learning_rate": 4.0380861419320276e-06, "loss": 1.7383, "step": 28709 }, { "epoch": 2.25, "grad_norm": 2.5831191539764404, "learning_rate": 4.037783541917635e-06, "loss": 1.3477, "step": 28710 }, { "epoch": 2.25, "eval_loss": 1.2596125602722168, "eval_runtime": 110.3059, "eval_samples_per_second": 24.36, "eval_steps_per_second": 6.092, "step": 28710 }, { "epoch": 2.250078369905956, "grad_norm": 2.5367817878723145, "learning_rate": 4.037480945563021e-06, "loss": 1.2674, "step": 28711 }, { "epoch": 2.250156739811912, "grad_norm": 8.49642276763916, "learning_rate": 4.037178352869333e-06, "loss": 1.1483, "step": 28712 }, { "epoch": 2.2502351097178686, "grad_norm": 3.3866965770721436, "learning_rate": 4.036875763837723e-06, "loss": 1.4401, "step": 28713 }, { "epoch": 2.2503134796238244, "grad_norm": 3.9221246242523193, "learning_rate": 4.0365731784693455e-06, "loss": 1.4537, "step": 28714 }, { "epoch": 2.2503918495297803, "grad_norm": 3.219825029373169, "learning_rate": 4.036270596765348e-06, "loss": 1.3903, "step": 28715 }, { "epoch": 2.2504702194357367, "grad_norm": 5.4177069664001465, "learning_rate": 4.035968018726882e-06, "loss": 1.3717, "step": 28716 }, { "epoch": 2.250548589341693, "grad_norm": 2.706704616546631, "learning_rate": 4.035665444355099e-06, "loss": 1.5654, "step": 28717 }, { "epoch": 2.250626959247649, "grad_norm": 3.630117654800415, "learning_rate": 4.035362873651147e-06, "loss": 1.694, "step": 28718 }, { "epoch": 2.2507053291536048, "grad_norm": 6.460152626037598, "learning_rate": 4.035060306616184e-06, "loss": 1.4238, "step": 28719 }, { "epoch": 2.250783699059561, "grad_norm": 2.2951133251190186, "learning_rate": 4.0347577432513515e-06, "loss": 1.3527, "step": 28720 }, { "epoch": 2.2508620689655174, "grad_norm": 3.6314210891723633, "learning_rate": 4.034455183557806e-06, "loss": 0.8698, "step": 28721 }, { "epoch": 2.2509404388714733, "grad_norm": 2.783017635345459, "learning_rate": 4.0341526275366975e-06, "loss": 1.3791, "step": 28722 }, { "epoch": 2.2510188087774297, "grad_norm": 3.543189287185669, "learning_rate": 4.033850075189177e-06, "loss": 1.4256, "step": 28723 }, { "epoch": 2.2510971786833855, "grad_norm": 2.2581162452697754, "learning_rate": 4.033547526516394e-06, "loss": 1.2923, "step": 28724 }, { "epoch": 2.251175548589342, "grad_norm": 5.003109931945801, "learning_rate": 4.033244981519498e-06, "loss": 1.5559, "step": 28725 }, { "epoch": 2.2512539184952978, "grad_norm": 5.495401382446289, "learning_rate": 4.0329424401996446e-06, "loss": 1.2688, "step": 28726 }, { "epoch": 2.251332288401254, "grad_norm": 3.006357192993164, "learning_rate": 4.032639902557979e-06, "loss": 1.4533, "step": 28727 }, { "epoch": 2.25141065830721, "grad_norm": 5.219773769378662, "learning_rate": 4.032337368595655e-06, "loss": 1.2986, "step": 28728 }, { "epoch": 2.2514890282131663, "grad_norm": 4.835130214691162, "learning_rate": 4.032034838313823e-06, "loss": 1.6365, "step": 28729 }, { "epoch": 2.251567398119122, "grad_norm": 5.485454559326172, "learning_rate": 4.0317323117136344e-06, "loss": 1.4581, "step": 28730 }, { "epoch": 2.2516457680250785, "grad_norm": 2.9406509399414062, "learning_rate": 4.031429788796237e-06, "loss": 1.5475, "step": 28731 }, { "epoch": 2.2517241379310344, "grad_norm": 4.887260437011719, "learning_rate": 4.031127269562782e-06, "loss": 1.325, "step": 28732 }, { "epoch": 2.2518025078369908, "grad_norm": 7.602297782897949, "learning_rate": 4.0308247540144235e-06, "loss": 1.2492, "step": 28733 }, { "epoch": 2.2518808777429467, "grad_norm": 2.6626482009887695, "learning_rate": 4.030522242152311e-06, "loss": 1.2756, "step": 28734 }, { "epoch": 2.251959247648903, "grad_norm": 3.8955202102661133, "learning_rate": 4.0302197339775915e-06, "loss": 1.6402, "step": 28735 }, { "epoch": 2.252037617554859, "grad_norm": 3.2233657836914062, "learning_rate": 4.029917229491419e-06, "loss": 1.4036, "step": 28736 }, { "epoch": 2.252115987460815, "grad_norm": 4.235177040100098, "learning_rate": 4.0296147286949435e-06, "loss": 1.8469, "step": 28737 }, { "epoch": 2.252194357366771, "grad_norm": 4.350862979888916, "learning_rate": 4.029312231589314e-06, "loss": 1.4498, "step": 28738 }, { "epoch": 2.2522727272727274, "grad_norm": 2.4220693111419678, "learning_rate": 4.029009738175682e-06, "loss": 1.1738, "step": 28739 }, { "epoch": 2.2523510971786833, "grad_norm": 2.445645809173584, "learning_rate": 4.028707248455198e-06, "loss": 1.1254, "step": 28740 }, { "epoch": 2.2524294670846396, "grad_norm": 3.211308240890503, "learning_rate": 4.028404762429015e-06, "loss": 1.4786, "step": 28741 }, { "epoch": 2.2525078369905955, "grad_norm": 9.368863105773926, "learning_rate": 4.02810228009828e-06, "loss": 0.8969, "step": 28742 }, { "epoch": 2.252586206896552, "grad_norm": 2.1201107501983643, "learning_rate": 4.027799801464144e-06, "loss": 0.6171, "step": 28743 }, { "epoch": 2.2526645768025078, "grad_norm": 4.51328182220459, "learning_rate": 4.027497326527758e-06, "loss": 1.5335, "step": 28744 }, { "epoch": 2.252742946708464, "grad_norm": 5.834029197692871, "learning_rate": 4.0271948552902755e-06, "loss": 1.2061, "step": 28745 }, { "epoch": 2.25282131661442, "grad_norm": 5.253078937530518, "learning_rate": 4.026892387752841e-06, "loss": 1.6772, "step": 28746 }, { "epoch": 2.2528996865203763, "grad_norm": 2.1262755393981934, "learning_rate": 4.026589923916609e-06, "loss": 1.2615, "step": 28747 }, { "epoch": 2.252978056426332, "grad_norm": 2.6117026805877686, "learning_rate": 4.026287463782729e-06, "loss": 1.4438, "step": 28748 }, { "epoch": 2.2530564263322885, "grad_norm": 2.473818302154541, "learning_rate": 4.025985007352352e-06, "loss": 1.4296, "step": 28749 }, { "epoch": 2.2531347962382444, "grad_norm": 4.1901326179504395, "learning_rate": 4.025682554626627e-06, "loss": 1.3411, "step": 28750 }, { "epoch": 2.2532131661442008, "grad_norm": 3.1190054416656494, "learning_rate": 4.025380105606704e-06, "loss": 1.3598, "step": 28751 }, { "epoch": 2.2532915360501566, "grad_norm": 2.3490867614746094, "learning_rate": 4.025077660293738e-06, "loss": 1.2197, "step": 28752 }, { "epoch": 2.253369905956113, "grad_norm": 2.8744499683380127, "learning_rate": 4.024775218688873e-06, "loss": 1.0528, "step": 28753 }, { "epoch": 2.253448275862069, "grad_norm": 3.2985551357269287, "learning_rate": 4.024472780793262e-06, "loss": 1.3357, "step": 28754 }, { "epoch": 2.253526645768025, "grad_norm": 2.567272186279297, "learning_rate": 4.024170346608056e-06, "loss": 1.4162, "step": 28755 }, { "epoch": 2.253605015673981, "grad_norm": 3.7109570503234863, "learning_rate": 4.023867916134406e-06, "loss": 1.2153, "step": 28756 }, { "epoch": 2.2536833855799374, "grad_norm": 2.275672674179077, "learning_rate": 4.023565489373461e-06, "loss": 1.5645, "step": 28757 }, { "epoch": 2.2537617554858933, "grad_norm": 7.535020351409912, "learning_rate": 4.023263066326368e-06, "loss": 1.6173, "step": 28758 }, { "epoch": 2.2538401253918496, "grad_norm": 7.886903285980225, "learning_rate": 4.022960646994284e-06, "loss": 1.2884, "step": 28759 }, { "epoch": 2.2539184952978055, "grad_norm": 2.9840853214263916, "learning_rate": 4.0226582313783555e-06, "loss": 1.3261, "step": 28760 }, { "epoch": 2.253996865203762, "grad_norm": 3.1509056091308594, "learning_rate": 4.022355819479732e-06, "loss": 1.1351, "step": 28761 }, { "epoch": 2.2540752351097177, "grad_norm": 3.1086554527282715, "learning_rate": 4.0220534112995656e-06, "loss": 1.3143, "step": 28762 }, { "epoch": 2.254153605015674, "grad_norm": 2.2781054973602295, "learning_rate": 4.0217510068390045e-06, "loss": 1.4051, "step": 28763 }, { "epoch": 2.25423197492163, "grad_norm": 4.100556373596191, "learning_rate": 4.021448606099203e-06, "loss": 1.4498, "step": 28764 }, { "epoch": 2.2543103448275863, "grad_norm": 7.023443698883057, "learning_rate": 4.0211462090813056e-06, "loss": 1.4022, "step": 28765 }, { "epoch": 2.254388714733542, "grad_norm": 2.2929022312164307, "learning_rate": 4.020843815786466e-06, "loss": 1.1998, "step": 28766 }, { "epoch": 2.2544670846394985, "grad_norm": 2.8938610553741455, "learning_rate": 4.020541426215834e-06, "loss": 1.2785, "step": 28767 }, { "epoch": 2.2545454545454544, "grad_norm": 2.631868839263916, "learning_rate": 4.020239040370559e-06, "loss": 1.3765, "step": 28768 }, { "epoch": 2.2546238244514107, "grad_norm": 3.6504926681518555, "learning_rate": 4.019936658251792e-06, "loss": 1.3059, "step": 28769 }, { "epoch": 2.2547021943573666, "grad_norm": 3.712137222290039, "learning_rate": 4.01963427986068e-06, "loss": 1.6423, "step": 28770 }, { "epoch": 2.254780564263323, "grad_norm": 3.107917070388794, "learning_rate": 4.019331905198379e-06, "loss": 1.2731, "step": 28771 }, { "epoch": 2.254858934169279, "grad_norm": 5.791876792907715, "learning_rate": 4.019029534266034e-06, "loss": 1.0256, "step": 28772 }, { "epoch": 2.254937304075235, "grad_norm": 2.321286678314209, "learning_rate": 4.018727167064797e-06, "loss": 1.4377, "step": 28773 }, { "epoch": 2.255015673981191, "grad_norm": 3.161477565765381, "learning_rate": 4.018424803595818e-06, "loss": 1.0875, "step": 28774 }, { "epoch": 2.2550940438871474, "grad_norm": 1.7801926136016846, "learning_rate": 4.018122443860248e-06, "loss": 1.3731, "step": 28775 }, { "epoch": 2.2551724137931033, "grad_norm": 3.4842512607574463, "learning_rate": 4.017820087859235e-06, "loss": 1.31, "step": 28776 }, { "epoch": 2.2552507836990596, "grad_norm": 2.6520352363586426, "learning_rate": 4.017517735593929e-06, "loss": 1.3797, "step": 28777 }, { "epoch": 2.2553291536050155, "grad_norm": 3.229034662246704, "learning_rate": 4.017215387065481e-06, "loss": 0.8424, "step": 28778 }, { "epoch": 2.255407523510972, "grad_norm": 3.379302740097046, "learning_rate": 4.016913042275043e-06, "loss": 1.9835, "step": 28779 }, { "epoch": 2.2554858934169277, "grad_norm": 4.467007160186768, "learning_rate": 4.016610701223761e-06, "loss": 1.3369, "step": 28780 }, { "epoch": 2.255564263322884, "grad_norm": 2.371229410171509, "learning_rate": 4.016308363912786e-06, "loss": 1.313, "step": 28781 }, { "epoch": 2.25564263322884, "grad_norm": 6.076260566711426, "learning_rate": 4.016006030343271e-06, "loss": 1.9089, "step": 28782 }, { "epoch": 2.2557210031347963, "grad_norm": 2.838590145111084, "learning_rate": 4.015703700516362e-06, "loss": 1.2578, "step": 28783 }, { "epoch": 2.255799373040752, "grad_norm": 2.886784791946411, "learning_rate": 4.015401374433211e-06, "loss": 0.85, "step": 28784 }, { "epoch": 2.2558777429467085, "grad_norm": 6.462858200073242, "learning_rate": 4.015099052094967e-06, "loss": 1.303, "step": 28785 }, { "epoch": 2.2559561128526644, "grad_norm": 3.8194069862365723, "learning_rate": 4.014796733502782e-06, "loss": 1.343, "step": 28786 }, { "epoch": 2.2560344827586207, "grad_norm": 3.930643081665039, "learning_rate": 4.0144944186578025e-06, "loss": 1.0701, "step": 28787 }, { "epoch": 2.2561128526645766, "grad_norm": 7.702260971069336, "learning_rate": 4.01419210756118e-06, "loss": 1.0351, "step": 28788 }, { "epoch": 2.256191222570533, "grad_norm": 3.1842093467712402, "learning_rate": 4.013889800214065e-06, "loss": 1.308, "step": 28789 }, { "epoch": 2.256269592476489, "grad_norm": 2.870166778564453, "learning_rate": 4.013587496617608e-06, "loss": 1.2897, "step": 28790 }, { "epoch": 2.256347962382445, "grad_norm": 4.81305456161499, "learning_rate": 4.013285196772956e-06, "loss": 1.6911, "step": 28791 }, { "epoch": 2.256426332288401, "grad_norm": 2.9363510608673096, "learning_rate": 4.01298290068126e-06, "loss": 1.2841, "step": 28792 }, { "epoch": 2.2565047021943574, "grad_norm": 6.303844928741455, "learning_rate": 4.0126806083436714e-06, "loss": 1.5117, "step": 28793 }, { "epoch": 2.2565830721003133, "grad_norm": 2.7105228900909424, "learning_rate": 4.0123783197613375e-06, "loss": 1.0789, "step": 28794 }, { "epoch": 2.2566614420062696, "grad_norm": 2.9189722537994385, "learning_rate": 4.012076034935409e-06, "loss": 1.6105, "step": 28795 }, { "epoch": 2.256739811912226, "grad_norm": 3.6449222564697266, "learning_rate": 4.011773753867036e-06, "loss": 1.596, "step": 28796 }, { "epoch": 2.256818181818182, "grad_norm": 3.291018009185791, "learning_rate": 4.01147147655737e-06, "loss": 1.4754, "step": 28797 }, { "epoch": 2.2568965517241377, "grad_norm": 4.528032302856445, "learning_rate": 4.011169203007556e-06, "loss": 1.1621, "step": 28798 }, { "epoch": 2.256974921630094, "grad_norm": 3.554363489151001, "learning_rate": 4.010866933218748e-06, "loss": 1.8598, "step": 28799 }, { "epoch": 2.2570532915360504, "grad_norm": 3.0153615474700928, "learning_rate": 4.010564667192092e-06, "loss": 1.2391, "step": 28800 }, { "epoch": 2.2571316614420063, "grad_norm": 20.980457305908203, "learning_rate": 4.0102624049287445e-06, "loss": 1.2434, "step": 28801 }, { "epoch": 2.257210031347962, "grad_norm": 2.6200850009918213, "learning_rate": 4.009960146429846e-06, "loss": 1.4814, "step": 28802 }, { "epoch": 2.2572884012539185, "grad_norm": 3.4449641704559326, "learning_rate": 4.009657891696551e-06, "loss": 1.2632, "step": 28803 }, { "epoch": 2.257366771159875, "grad_norm": 3.7965087890625, "learning_rate": 4.00935564073001e-06, "loss": 1.3205, "step": 28804 }, { "epoch": 2.2574451410658307, "grad_norm": 3.1649587154388428, "learning_rate": 4.009053393531373e-06, "loss": 1.3052, "step": 28805 }, { "epoch": 2.2575235109717866, "grad_norm": 1.8807722330093384, "learning_rate": 4.008751150101786e-06, "loss": 1.0841, "step": 28806 }, { "epoch": 2.257601880877743, "grad_norm": 4.31267786026001, "learning_rate": 4.008448910442399e-06, "loss": 1.0769, "step": 28807 }, { "epoch": 2.2576802507836993, "grad_norm": 9.045514106750488, "learning_rate": 4.0081466745543665e-06, "loss": 0.8495, "step": 28808 }, { "epoch": 2.257758620689655, "grad_norm": 2.4604735374450684, "learning_rate": 4.007844442438832e-06, "loss": 1.2938, "step": 28809 }, { "epoch": 2.257836990595611, "grad_norm": 2.259913444519043, "learning_rate": 4.007542214096947e-06, "loss": 1.2045, "step": 28810 }, { "epoch": 2.2579153605015674, "grad_norm": 3.981829881668091, "learning_rate": 4.007239989529864e-06, "loss": 1.3232, "step": 28811 }, { "epoch": 2.2579937304075237, "grad_norm": 2.8934428691864014, "learning_rate": 4.006937768738729e-06, "loss": 1.4376, "step": 28812 }, { "epoch": 2.2580721003134796, "grad_norm": 3.6040797233581543, "learning_rate": 4.006635551724693e-06, "loss": 1.1963, "step": 28813 }, { "epoch": 2.2581504702194355, "grad_norm": 3.1498725414276123, "learning_rate": 4.0063333384889046e-06, "loss": 1.2785, "step": 28814 }, { "epoch": 2.258228840125392, "grad_norm": 2.1942434310913086, "learning_rate": 4.006031129032513e-06, "loss": 1.1123, "step": 28815 }, { "epoch": 2.258307210031348, "grad_norm": 3.249368667602539, "learning_rate": 4.005728923356671e-06, "loss": 1.5071, "step": 28816 }, { "epoch": 2.258385579937304, "grad_norm": 2.27616548538208, "learning_rate": 4.0054267214625245e-06, "loss": 1.5771, "step": 28817 }, { "epoch": 2.2584639498432604, "grad_norm": 2.738374948501587, "learning_rate": 4.005124523351222e-06, "loss": 1.2767, "step": 28818 }, { "epoch": 2.2585423197492163, "grad_norm": 8.716212272644043, "learning_rate": 4.004822329023916e-06, "loss": 0.5395, "step": 28819 }, { "epoch": 2.2586206896551726, "grad_norm": 2.97652006149292, "learning_rate": 4.004520138481756e-06, "loss": 1.2697, "step": 28820 }, { "epoch": 2.2586990595611285, "grad_norm": 3.067892074584961, "learning_rate": 4.004217951725889e-06, "loss": 1.4547, "step": 28821 }, { "epoch": 2.258777429467085, "grad_norm": 3.2945339679718018, "learning_rate": 4.0039157687574655e-06, "loss": 1.3451, "step": 28822 }, { "epoch": 2.2588557993730407, "grad_norm": 8.587969779968262, "learning_rate": 4.003613589577636e-06, "loss": 0.8375, "step": 28823 }, { "epoch": 2.258934169278997, "grad_norm": 2.5460104942321777, "learning_rate": 4.003311414187547e-06, "loss": 1.5619, "step": 28824 }, { "epoch": 2.259012539184953, "grad_norm": 3.817701578140259, "learning_rate": 4.003009242588349e-06, "loss": 1.5485, "step": 28825 }, { "epoch": 2.2590909090909093, "grad_norm": 3.44486403465271, "learning_rate": 4.002707074781192e-06, "loss": 1.355, "step": 28826 }, { "epoch": 2.259169278996865, "grad_norm": 4.69640588760376, "learning_rate": 4.002404910767228e-06, "loss": 1.046, "step": 28827 }, { "epoch": 2.2592476489028215, "grad_norm": 3.0138072967529297, "learning_rate": 4.002102750547599e-06, "loss": 1.767, "step": 28828 }, { "epoch": 2.2593260188087774, "grad_norm": 2.369939088821411, "learning_rate": 4.001800594123461e-06, "loss": 1.2648, "step": 28829 }, { "epoch": 2.2594043887147337, "grad_norm": 3.9660634994506836, "learning_rate": 4.001498441495961e-06, "loss": 1.4236, "step": 28830 }, { "epoch": 2.2594827586206896, "grad_norm": 4.263728141784668, "learning_rate": 4.001196292666248e-06, "loss": 0.6289, "step": 28831 }, { "epoch": 2.259561128526646, "grad_norm": 6.419029235839844, "learning_rate": 4.0008941476354705e-06, "loss": 1.3948, "step": 28832 }, { "epoch": 2.259639498432602, "grad_norm": 5.153295993804932, "learning_rate": 4.000592006404778e-06, "loss": 1.1869, "step": 28833 }, { "epoch": 2.259717868338558, "grad_norm": 5.215392589569092, "learning_rate": 4.000289868975323e-06, "loss": 0.963, "step": 28834 }, { "epoch": 2.259796238244514, "grad_norm": 4.762544631958008, "learning_rate": 3.999987735348248e-06, "loss": 1.484, "step": 28835 }, { "epoch": 2.2598746081504704, "grad_norm": 2.816850185394287, "learning_rate": 3.999685605524708e-06, "loss": 1.1314, "step": 28836 }, { "epoch": 2.2599529780564263, "grad_norm": 4.494633197784424, "learning_rate": 3.99938347950585e-06, "loss": 1.6245, "step": 28837 }, { "epoch": 2.2600313479623826, "grad_norm": 6.581861972808838, "learning_rate": 3.999081357292825e-06, "loss": 1.7728, "step": 28838 }, { "epoch": 2.2601097178683385, "grad_norm": 2.605651617050171, "learning_rate": 3.9987792388867774e-06, "loss": 1.421, "step": 28839 }, { "epoch": 2.260188087774295, "grad_norm": 3.1726441383361816, "learning_rate": 3.99847712428886e-06, "loss": 0.8208, "step": 28840 }, { "epoch": 2.2602664576802507, "grad_norm": 4.902192115783691, "learning_rate": 3.998175013500222e-06, "loss": 1.6151, "step": 28841 }, { "epoch": 2.260344827586207, "grad_norm": 3.882423162460327, "learning_rate": 3.997872906522012e-06, "loss": 1.4879, "step": 28842 }, { "epoch": 2.260423197492163, "grad_norm": 4.001837253570557, "learning_rate": 3.9975708033553775e-06, "loss": 1.6271, "step": 28843 }, { "epoch": 2.2605015673981192, "grad_norm": 2.0772767066955566, "learning_rate": 3.99726870400147e-06, "loss": 1.3501, "step": 28844 }, { "epoch": 2.260579937304075, "grad_norm": 4.991461277008057, "learning_rate": 3.996966608461436e-06, "loss": 1.4586, "step": 28845 }, { "epoch": 2.2606583072100315, "grad_norm": 5.32226037979126, "learning_rate": 3.996664516736428e-06, "loss": 1.5454, "step": 28846 }, { "epoch": 2.2607366771159874, "grad_norm": 2.9759697914123535, "learning_rate": 3.99636242882759e-06, "loss": 1.4215, "step": 28847 }, { "epoch": 2.2608150470219437, "grad_norm": 2.9431393146514893, "learning_rate": 3.996060344736076e-06, "loss": 0.8978, "step": 28848 }, { "epoch": 2.2608934169278996, "grad_norm": 5.355340957641602, "learning_rate": 3.995758264463032e-06, "loss": 1.2631, "step": 28849 }, { "epoch": 2.260971786833856, "grad_norm": 3.217616319656372, "learning_rate": 3.995456188009609e-06, "loss": 1.3987, "step": 28850 }, { "epoch": 2.261050156739812, "grad_norm": 3.501927614212036, "learning_rate": 3.995154115376953e-06, "loss": 1.3657, "step": 28851 }, { "epoch": 2.261128526645768, "grad_norm": 4.320569038391113, "learning_rate": 3.994852046566214e-06, "loss": 1.8871, "step": 28852 }, { "epoch": 2.261206896551724, "grad_norm": 1.9280834197998047, "learning_rate": 3.994549981578545e-06, "loss": 1.344, "step": 28853 }, { "epoch": 2.2612852664576804, "grad_norm": 2.477280855178833, "learning_rate": 3.994247920415088e-06, "loss": 1.2674, "step": 28854 }, { "epoch": 2.2613636363636362, "grad_norm": 4.931196689605713, "learning_rate": 3.993945863076996e-06, "loss": 1.7408, "step": 28855 }, { "epoch": 2.2614420062695926, "grad_norm": 2.604525089263916, "learning_rate": 3.9936438095654175e-06, "loss": 1.447, "step": 28856 }, { "epoch": 2.2615203761755485, "grad_norm": 6.2624006271362305, "learning_rate": 3.993341759881502e-06, "loss": 1.0375, "step": 28857 }, { "epoch": 2.261598746081505, "grad_norm": 2.5532147884368896, "learning_rate": 3.9930397140263965e-06, "loss": 1.6752, "step": 28858 }, { "epoch": 2.2616771159874607, "grad_norm": 13.733954429626465, "learning_rate": 3.99273767200125e-06, "loss": 0.9986, "step": 28859 }, { "epoch": 2.261755485893417, "grad_norm": 2.5170488357543945, "learning_rate": 3.992435633807212e-06, "loss": 1.1814, "step": 28860 }, { "epoch": 2.261833855799373, "grad_norm": 3.0563125610351562, "learning_rate": 3.992133599445433e-06, "loss": 1.3115, "step": 28861 }, { "epoch": 2.2619122257053292, "grad_norm": 4.727047920227051, "learning_rate": 3.99183156891706e-06, "loss": 1.615, "step": 28862 }, { "epoch": 2.261990595611285, "grad_norm": 2.4370906352996826, "learning_rate": 3.991529542223241e-06, "loss": 1.4551, "step": 28863 }, { "epoch": 2.2620689655172415, "grad_norm": 6.718320846557617, "learning_rate": 3.991227519365127e-06, "loss": 1.7031, "step": 28864 }, { "epoch": 2.2621473354231973, "grad_norm": 2.9790196418762207, "learning_rate": 3.990925500343864e-06, "loss": 1.3259, "step": 28865 }, { "epoch": 2.2622257053291537, "grad_norm": 5.598478317260742, "learning_rate": 3.990623485160602e-06, "loss": 1.4788, "step": 28866 }, { "epoch": 2.2623040752351096, "grad_norm": 4.1618571281433105, "learning_rate": 3.990321473816489e-06, "loss": 1.6969, "step": 28867 }, { "epoch": 2.262382445141066, "grad_norm": 3.432586431503296, "learning_rate": 3.990019466312677e-06, "loss": 1.1836, "step": 28868 }, { "epoch": 2.262460815047022, "grad_norm": 2.9594454765319824, "learning_rate": 3.9897174626503105e-06, "loss": 1.6884, "step": 28869 }, { "epoch": 2.262539184952978, "grad_norm": 5.454934120178223, "learning_rate": 3.98941546283054e-06, "loss": 1.7713, "step": 28870 }, { "epoch": 2.262617554858934, "grad_norm": 2.507236957550049, "learning_rate": 3.989113466854513e-06, "loss": 1.3532, "step": 28871 }, { "epoch": 2.2626959247648903, "grad_norm": 10.809791564941406, "learning_rate": 3.988811474723381e-06, "loss": 0.8604, "step": 28872 }, { "epoch": 2.2627742946708462, "grad_norm": 2.450655937194824, "learning_rate": 3.98850948643829e-06, "loss": 1.2607, "step": 28873 }, { "epoch": 2.2628526645768026, "grad_norm": 5.0616888999938965, "learning_rate": 3.988207502000389e-06, "loss": 1.2092, "step": 28874 }, { "epoch": 2.2629310344827585, "grad_norm": 2.7233965396881104, "learning_rate": 3.987905521410828e-06, "loss": 1.763, "step": 28875 }, { "epoch": 2.263009404388715, "grad_norm": 6.351364612579346, "learning_rate": 3.987603544670755e-06, "loss": 1.5782, "step": 28876 }, { "epoch": 2.2630877742946707, "grad_norm": 3.123491048812866, "learning_rate": 3.9873015717813165e-06, "loss": 1.694, "step": 28877 }, { "epoch": 2.263166144200627, "grad_norm": 2.5673294067382812, "learning_rate": 3.986999602743663e-06, "loss": 1.2438, "step": 28878 }, { "epoch": 2.263244514106583, "grad_norm": 2.764464855194092, "learning_rate": 3.986697637558944e-06, "loss": 1.2011, "step": 28879 }, { "epoch": 2.2633228840125392, "grad_norm": 2.640124797821045, "learning_rate": 3.9863956762283055e-06, "loss": 1.2214, "step": 28880 }, { "epoch": 2.263401253918495, "grad_norm": 2.40429949760437, "learning_rate": 3.986093718752897e-06, "loss": 1.3542, "step": 28881 }, { "epoch": 2.2634796238244514, "grad_norm": 2.9961907863616943, "learning_rate": 3.985791765133868e-06, "loss": 1.3093, "step": 28882 }, { "epoch": 2.2635579937304073, "grad_norm": 4.611904621124268, "learning_rate": 3.985489815372366e-06, "loss": 1.52, "step": 28883 }, { "epoch": 2.2636363636363637, "grad_norm": 4.018280029296875, "learning_rate": 3.985187869469539e-06, "loss": 1.3147, "step": 28884 }, { "epoch": 2.2637147335423196, "grad_norm": 4.0047607421875, "learning_rate": 3.984885927426536e-06, "loss": 1.7006, "step": 28885 }, { "epoch": 2.263793103448276, "grad_norm": 6.444235324859619, "learning_rate": 3.984583989244506e-06, "loss": 1.6397, "step": 28886 }, { "epoch": 2.2638714733542322, "grad_norm": 2.525918960571289, "learning_rate": 3.984282054924599e-06, "loss": 1.296, "step": 28887 }, { "epoch": 2.263949843260188, "grad_norm": 2.9037680625915527, "learning_rate": 3.983980124467958e-06, "loss": 1.5138, "step": 28888 }, { "epoch": 2.264028213166144, "grad_norm": 2.7712464332580566, "learning_rate": 3.9836781978757364e-06, "loss": 1.3595, "step": 28889 }, { "epoch": 2.2641065830721003, "grad_norm": 7.6027913093566895, "learning_rate": 3.983376275149082e-06, "loss": 1.4971, "step": 28890 }, { "epoch": 2.2641849529780567, "grad_norm": 3.123133897781372, "learning_rate": 3.98307435628914e-06, "loss": 1.5401, "step": 28891 }, { "epoch": 2.2642633228840126, "grad_norm": 2.554593801498413, "learning_rate": 3.982772441297062e-06, "loss": 1.3949, "step": 28892 }, { "epoch": 2.2643416927899684, "grad_norm": 9.040421485900879, "learning_rate": 3.982470530173994e-06, "loss": 1.5569, "step": 28893 }, { "epoch": 2.2644200626959248, "grad_norm": 4.153013706207275, "learning_rate": 3.982168622921088e-06, "loss": 1.4443, "step": 28894 }, { "epoch": 2.264498432601881, "grad_norm": 4.300342559814453, "learning_rate": 3.981866719539487e-06, "loss": 1.4058, "step": 28895 }, { "epoch": 2.264576802507837, "grad_norm": 6.429489612579346, "learning_rate": 3.981564820030342e-06, "loss": 1.1217, "step": 28896 }, { "epoch": 2.264655172413793, "grad_norm": 10.181745529174805, "learning_rate": 3.981262924394802e-06, "loss": 1.2253, "step": 28897 }, { "epoch": 2.264733542319749, "grad_norm": 3.965566396713257, "learning_rate": 3.980961032634017e-06, "loss": 1.3139, "step": 28898 }, { "epoch": 2.2648119122257055, "grad_norm": 4.023244380950928, "learning_rate": 3.98065914474913e-06, "loss": 1.2679, "step": 28899 }, { "epoch": 2.2648902821316614, "grad_norm": 7.446450233459473, "learning_rate": 3.980357260741293e-06, "loss": 1.4196, "step": 28900 }, { "epoch": 2.2649686520376173, "grad_norm": 2.5694072246551514, "learning_rate": 3.9800553806116515e-06, "loss": 1.3671, "step": 28901 }, { "epoch": 2.2650470219435737, "grad_norm": 4.266473293304443, "learning_rate": 3.97975350436136e-06, "loss": 1.6581, "step": 28902 }, { "epoch": 2.26512539184953, "grad_norm": 4.230676174163818, "learning_rate": 3.979451631991558e-06, "loss": 1.3289, "step": 28903 }, { "epoch": 2.265203761755486, "grad_norm": 3.473012685775757, "learning_rate": 3.979149763503398e-06, "loss": 1.6271, "step": 28904 }, { "epoch": 2.2652821316614418, "grad_norm": 4.167920112609863, "learning_rate": 3.978847898898031e-06, "loss": 1.9758, "step": 28905 }, { "epoch": 2.265360501567398, "grad_norm": 6.971639633178711, "learning_rate": 3.978546038176599e-06, "loss": 1.0004, "step": 28906 }, { "epoch": 2.2654388714733544, "grad_norm": 2.3534069061279297, "learning_rate": 3.978244181340254e-06, "loss": 1.2394, "step": 28907 }, { "epoch": 2.2655172413793103, "grad_norm": 4.298050403594971, "learning_rate": 3.977942328390143e-06, "loss": 1.3701, "step": 28908 }, { "epoch": 2.265595611285266, "grad_norm": 4.091745376586914, "learning_rate": 3.977640479327417e-06, "loss": 1.6012, "step": 28909 }, { "epoch": 2.2656739811912225, "grad_norm": 6.052587032318115, "learning_rate": 3.977338634153218e-06, "loss": 1.0571, "step": 28910 }, { "epoch": 2.265752351097179, "grad_norm": 3.8588576316833496, "learning_rate": 3.977036792868699e-06, "loss": 1.0326, "step": 28911 }, { "epoch": 2.2658307210031348, "grad_norm": 2.537364959716797, "learning_rate": 3.976734955475007e-06, "loss": 1.4568, "step": 28912 }, { "epoch": 2.265909090909091, "grad_norm": 4.100359916687012, "learning_rate": 3.97643312197329e-06, "loss": 1.7799, "step": 28913 }, { "epoch": 2.265987460815047, "grad_norm": 2.9923207759857178, "learning_rate": 3.976131292364695e-06, "loss": 1.2351, "step": 28914 }, { "epoch": 2.2660658307210033, "grad_norm": 2.86601185798645, "learning_rate": 3.97582946665037e-06, "loss": 1.35, "step": 28915 }, { "epoch": 2.266144200626959, "grad_norm": 2.9799818992614746, "learning_rate": 3.975527644831464e-06, "loss": 1.1573, "step": 28916 }, { "epoch": 2.2662225705329155, "grad_norm": 7.231321334838867, "learning_rate": 3.975225826909126e-06, "loss": 1.6645, "step": 28917 }, { "epoch": 2.2663009404388714, "grad_norm": 2.8438773155212402, "learning_rate": 3.9749240128845016e-06, "loss": 1.3828, "step": 28918 }, { "epoch": 2.2663793103448278, "grad_norm": 3.4081060886383057, "learning_rate": 3.97462220275874e-06, "loss": 1.2027, "step": 28919 }, { "epoch": 2.2664576802507836, "grad_norm": 2.9737772941589355, "learning_rate": 3.974320396532989e-06, "loss": 1.3555, "step": 28920 }, { "epoch": 2.26653605015674, "grad_norm": 3.759338617324829, "learning_rate": 3.974018594208396e-06, "loss": 1.4894, "step": 28921 }, { "epoch": 2.266614420062696, "grad_norm": 3.1370911598205566, "learning_rate": 3.973716795786108e-06, "loss": 1.4275, "step": 28922 }, { "epoch": 2.266692789968652, "grad_norm": 4.647267818450928, "learning_rate": 3.973415001267275e-06, "loss": 1.6136, "step": 28923 }, { "epoch": 2.266771159874608, "grad_norm": 2.781555652618408, "learning_rate": 3.973113210653046e-06, "loss": 0.5764, "step": 28924 }, { "epoch": 2.2668495297805644, "grad_norm": 2.687422752380371, "learning_rate": 3.9728114239445645e-06, "loss": 1.5262, "step": 28925 }, { "epoch": 2.2669278996865203, "grad_norm": 6.907716274261475, "learning_rate": 3.972509641142982e-06, "loss": 1.1225, "step": 28926 }, { "epoch": 2.2670062695924766, "grad_norm": 5.048106670379639, "learning_rate": 3.972207862249443e-06, "loss": 1.0297, "step": 28927 }, { "epoch": 2.2670846394984325, "grad_norm": 2.4717862606048584, "learning_rate": 3.971906087265102e-06, "loss": 1.121, "step": 28928 }, { "epoch": 2.267163009404389, "grad_norm": 2.5523841381073, "learning_rate": 3.971604316191098e-06, "loss": 1.2844, "step": 28929 }, { "epoch": 2.2672413793103448, "grad_norm": 3.3256940841674805, "learning_rate": 3.971302549028584e-06, "loss": 1.6474, "step": 28930 }, { "epoch": 2.267319749216301, "grad_norm": 2.8839972019195557, "learning_rate": 3.9710007857787066e-06, "loss": 1.3613, "step": 28931 }, { "epoch": 2.267398119122257, "grad_norm": 2.6405367851257324, "learning_rate": 3.970699026442616e-06, "loss": 1.185, "step": 28932 }, { "epoch": 2.2674764890282133, "grad_norm": 5.079228401184082, "learning_rate": 3.970397271021455e-06, "loss": 1.3442, "step": 28933 }, { "epoch": 2.267554858934169, "grad_norm": 3.427126407623291, "learning_rate": 3.970095519516373e-06, "loss": 1.1081, "step": 28934 }, { "epoch": 2.2676332288401255, "grad_norm": 7.969059944152832, "learning_rate": 3.969793771928523e-06, "loss": 2.005, "step": 28935 }, { "epoch": 2.2677115987460814, "grad_norm": 3.8635776042938232, "learning_rate": 3.969492028259044e-06, "loss": 1.0358, "step": 28936 }, { "epoch": 2.2677899686520377, "grad_norm": 2.997351884841919, "learning_rate": 3.969190288509089e-06, "loss": 1.6401, "step": 28937 }, { "epoch": 2.2678683385579936, "grad_norm": 4.820173740386963, "learning_rate": 3.968888552679806e-06, "loss": 0.9761, "step": 28938 }, { "epoch": 2.26794670846395, "grad_norm": 2.7509214878082275, "learning_rate": 3.968586820772341e-06, "loss": 1.2094, "step": 28939 }, { "epoch": 2.268025078369906, "grad_norm": 5.746023654937744, "learning_rate": 3.968285092787842e-06, "loss": 1.0104, "step": 28940 }, { "epoch": 2.268103448275862, "grad_norm": 3.768035650253296, "learning_rate": 3.9679833687274545e-06, "loss": 1.5175, "step": 28941 }, { "epoch": 2.268181818181818, "grad_norm": 5.275641918182373, "learning_rate": 3.967681648592331e-06, "loss": 1.1016, "step": 28942 }, { "epoch": 2.2682601880877744, "grad_norm": 3.4453866481781006, "learning_rate": 3.967379932383615e-06, "loss": 1.3731, "step": 28943 }, { "epoch": 2.2683385579937303, "grad_norm": 3.3433828353881836, "learning_rate": 3.967078220102456e-06, "loss": 1.5832, "step": 28944 }, { "epoch": 2.2684169278996866, "grad_norm": 2.1802735328674316, "learning_rate": 3.96677651175e-06, "loss": 1.3765, "step": 28945 }, { "epoch": 2.2684952978056425, "grad_norm": 4.771096229553223, "learning_rate": 3.9664748073273975e-06, "loss": 1.1598, "step": 28946 }, { "epoch": 2.268573667711599, "grad_norm": 2.6709144115448, "learning_rate": 3.966173106835792e-06, "loss": 1.478, "step": 28947 }, { "epoch": 2.2686520376175547, "grad_norm": 6.657306671142578, "learning_rate": 3.965871410276332e-06, "loss": 0.7472, "step": 28948 }, { "epoch": 2.268730407523511, "grad_norm": 2.758265733718872, "learning_rate": 3.965569717650167e-06, "loss": 1.3636, "step": 28949 }, { "epoch": 2.268808777429467, "grad_norm": 9.07697582244873, "learning_rate": 3.965268028958445e-06, "loss": 1.3016, "step": 28950 }, { "epoch": 2.2688871473354233, "grad_norm": 3.6133086681365967, "learning_rate": 3.964966344202311e-06, "loss": 0.9514, "step": 28951 }, { "epoch": 2.268965517241379, "grad_norm": 3.1563847064971924, "learning_rate": 3.964664663382913e-06, "loss": 1.0927, "step": 28952 }, { "epoch": 2.2690438871473355, "grad_norm": 22.11501693725586, "learning_rate": 3.964362986501398e-06, "loss": 0.3386, "step": 28953 }, { "epoch": 2.2691222570532914, "grad_norm": 4.556239604949951, "learning_rate": 3.964061313558917e-06, "loss": 0.832, "step": 28954 }, { "epoch": 2.2692006269592477, "grad_norm": 2.5413103103637695, "learning_rate": 3.963759644556611e-06, "loss": 1.4806, "step": 28955 }, { "epoch": 2.2692789968652036, "grad_norm": 3.7118539810180664, "learning_rate": 3.963457979495634e-06, "loss": 1.4278, "step": 28956 }, { "epoch": 2.26935736677116, "grad_norm": 4.853450775146484, "learning_rate": 3.963156318377129e-06, "loss": 1.6254, "step": 28957 }, { "epoch": 2.269435736677116, "grad_norm": 3.504934072494507, "learning_rate": 3.9628546612022464e-06, "loss": 1.094, "step": 28958 }, { "epoch": 2.269514106583072, "grad_norm": 3.519099712371826, "learning_rate": 3.962553007972131e-06, "loss": 1.3527, "step": 28959 }, { "epoch": 2.269592476489028, "grad_norm": 2.5574734210968018, "learning_rate": 3.96225135868793e-06, "loss": 1.2493, "step": 28960 }, { "epoch": 2.2696708463949844, "grad_norm": 8.874998092651367, "learning_rate": 3.961949713350795e-06, "loss": 1.7384, "step": 28961 }, { "epoch": 2.2697492163009403, "grad_norm": 3.990321397781372, "learning_rate": 3.961648071961866e-06, "loss": 1.2644, "step": 28962 }, { "epoch": 2.2698275862068966, "grad_norm": 5.840339660644531, "learning_rate": 3.961346434522296e-06, "loss": 1.5113, "step": 28963 }, { "epoch": 2.2699059561128525, "grad_norm": 2.8363776206970215, "learning_rate": 3.961044801033231e-06, "loss": 1.0946, "step": 28964 }, { "epoch": 2.269984326018809, "grad_norm": 2.6154816150665283, "learning_rate": 3.960743171495819e-06, "loss": 0.9029, "step": 28965 }, { "epoch": 2.2700626959247647, "grad_norm": 4.192566394805908, "learning_rate": 3.960441545911205e-06, "loss": 1.3035, "step": 28966 }, { "epoch": 2.270141065830721, "grad_norm": 6.894608497619629, "learning_rate": 3.9601399242805364e-06, "loss": 1.4412, "step": 28967 }, { "epoch": 2.270219435736677, "grad_norm": 4.125693321228027, "learning_rate": 3.9598383066049615e-06, "loss": 1.5226, "step": 28968 }, { "epoch": 2.2702978056426333, "grad_norm": 2.9601614475250244, "learning_rate": 3.95953669288563e-06, "loss": 1.235, "step": 28969 }, { "epoch": 2.270376175548589, "grad_norm": 2.603323221206665, "learning_rate": 3.959235083123685e-06, "loss": 1.2562, "step": 28970 }, { "epoch": 2.2704545454545455, "grad_norm": 2.6945199966430664, "learning_rate": 3.958933477320275e-06, "loss": 1.2295, "step": 28971 }, { "epoch": 2.2705329153605014, "grad_norm": 2.7040843963623047, "learning_rate": 3.958631875476546e-06, "loss": 1.1441, "step": 28972 }, { "epoch": 2.2706112852664577, "grad_norm": 3.218773603439331, "learning_rate": 3.958330277593649e-06, "loss": 1.531, "step": 28973 }, { "epoch": 2.2706896551724136, "grad_norm": 3.998441219329834, "learning_rate": 3.958028683672726e-06, "loss": 1.5347, "step": 28974 }, { "epoch": 2.27076802507837, "grad_norm": 2.6760213375091553, "learning_rate": 3.957727093714929e-06, "loss": 1.1599, "step": 28975 }, { "epoch": 2.270846394984326, "grad_norm": 2.518068552017212, "learning_rate": 3.957425507721403e-06, "loss": 1.4006, "step": 28976 }, { "epoch": 2.270924764890282, "grad_norm": 5.125228404998779, "learning_rate": 3.957123925693294e-06, "loss": 1.011, "step": 28977 }, { "epoch": 2.271003134796238, "grad_norm": 2.394035816192627, "learning_rate": 3.956822347631749e-06, "loss": 1.1145, "step": 28978 }, { "epoch": 2.2710815047021944, "grad_norm": 2.9894354343414307, "learning_rate": 3.956520773537916e-06, "loss": 1.4754, "step": 28979 }, { "epoch": 2.2711598746081503, "grad_norm": 2.0865955352783203, "learning_rate": 3.956219203412944e-06, "loss": 1.4487, "step": 28980 }, { "epoch": 2.2712382445141066, "grad_norm": 2.9807486534118652, "learning_rate": 3.955917637257976e-06, "loss": 1.2007, "step": 28981 }, { "epoch": 2.271316614420063, "grad_norm": 3.0788605213165283, "learning_rate": 3.955616075074162e-06, "loss": 1.1445, "step": 28982 }, { "epoch": 2.271394984326019, "grad_norm": 2.752904176712036, "learning_rate": 3.955314516862648e-06, "loss": 1.4759, "step": 28983 }, { "epoch": 2.2714733542319747, "grad_norm": 2.233062982559204, "learning_rate": 3.955012962624581e-06, "loss": 1.1606, "step": 28984 }, { "epoch": 2.271551724137931, "grad_norm": 3.1304972171783447, "learning_rate": 3.954711412361107e-06, "loss": 1.0881, "step": 28985 }, { "epoch": 2.2716300940438874, "grad_norm": 2.0296499729156494, "learning_rate": 3.954409866073374e-06, "loss": 1.3774, "step": 28986 }, { "epoch": 2.2717084639498433, "grad_norm": 6.545528411865234, "learning_rate": 3.95410832376253e-06, "loss": 1.8997, "step": 28987 }, { "epoch": 2.271786833855799, "grad_norm": 4.421164035797119, "learning_rate": 3.9538067854297215e-06, "loss": 1.6198, "step": 28988 }, { "epoch": 2.2718652037617555, "grad_norm": 2.326572895050049, "learning_rate": 3.953505251076093e-06, "loss": 1.4633, "step": 28989 }, { "epoch": 2.271943573667712, "grad_norm": 4.318005561828613, "learning_rate": 3.953203720702793e-06, "loss": 0.9037, "step": 28990 }, { "epoch": 2.2720219435736677, "grad_norm": 6.405950546264648, "learning_rate": 3.9529021943109695e-06, "loss": 1.6853, "step": 28991 }, { "epoch": 2.2721003134796236, "grad_norm": 6.148138999938965, "learning_rate": 3.952600671901766e-06, "loss": 1.5969, "step": 28992 }, { "epoch": 2.27217868338558, "grad_norm": 3.879051685333252, "learning_rate": 3.952299153476333e-06, "loss": 1.1244, "step": 28993 }, { "epoch": 2.2722570532915363, "grad_norm": 2.5171871185302734, "learning_rate": 3.951997639035816e-06, "loss": 1.3911, "step": 28994 }, { "epoch": 2.272335423197492, "grad_norm": 3.7845358848571777, "learning_rate": 3.951696128581363e-06, "loss": 1.4412, "step": 28995 }, { "epoch": 2.272413793103448, "grad_norm": 3.324380874633789, "learning_rate": 3.9513946221141175e-06, "loss": 0.8562, "step": 28996 }, { "epoch": 2.2724921630094044, "grad_norm": 2.5499486923217773, "learning_rate": 3.9510931196352284e-06, "loss": 1.6642, "step": 28997 }, { "epoch": 2.2725705329153607, "grad_norm": 2.536267042160034, "learning_rate": 3.950791621145842e-06, "loss": 1.3853, "step": 28998 }, { "epoch": 2.2726489028213166, "grad_norm": 4.42697811126709, "learning_rate": 3.9504901266471074e-06, "loss": 1.1814, "step": 28999 }, { "epoch": 2.2727272727272725, "grad_norm": 7.209356307983398, "learning_rate": 3.950188636140168e-06, "loss": 1.3017, "step": 29000 }, { "epoch": 2.272805642633229, "grad_norm": 3.1771833896636963, "learning_rate": 3.949887149626172e-06, "loss": 1.0202, "step": 29001 }, { "epoch": 2.272884012539185, "grad_norm": 3.2230658531188965, "learning_rate": 3.949585667106266e-06, "loss": 1.3763, "step": 29002 }, { "epoch": 2.272962382445141, "grad_norm": 4.898065567016602, "learning_rate": 3.949284188581597e-06, "loss": 1.5275, "step": 29003 }, { "epoch": 2.2730407523510974, "grad_norm": 2.2838239669799805, "learning_rate": 3.9489827140533084e-06, "loss": 1.4366, "step": 29004 }, { "epoch": 2.2731191222570533, "grad_norm": 3.2370500564575195, "learning_rate": 3.948681243522552e-06, "loss": 1.2725, "step": 29005 }, { "epoch": 2.2731974921630096, "grad_norm": 2.620518922805786, "learning_rate": 3.948379776990473e-06, "loss": 1.2883, "step": 29006 }, { "epoch": 2.2732758620689655, "grad_norm": 1.8818880319595337, "learning_rate": 3.948078314458217e-06, "loss": 1.414, "step": 29007 }, { "epoch": 2.273354231974922, "grad_norm": 4.753962993621826, "learning_rate": 3.9477768559269295e-06, "loss": 1.2569, "step": 29008 }, { "epoch": 2.2734326018808777, "grad_norm": 11.10775089263916, "learning_rate": 3.947475401397758e-06, "loss": 1.6799, "step": 29009 }, { "epoch": 2.273510971786834, "grad_norm": 2.931304454803467, "learning_rate": 3.947173950871852e-06, "loss": 0.9094, "step": 29010 }, { "epoch": 2.27358934169279, "grad_norm": 3.7009990215301514, "learning_rate": 3.946872504350353e-06, "loss": 1.6796, "step": 29011 }, { "epoch": 2.2736677115987463, "grad_norm": 2.9160332679748535, "learning_rate": 3.946571061834411e-06, "loss": 0.9698, "step": 29012 }, { "epoch": 2.273746081504702, "grad_norm": 2.5624210834503174, "learning_rate": 3.946269623325172e-06, "loss": 1.5043, "step": 29013 }, { "epoch": 2.2738244514106585, "grad_norm": 5.41151237487793, "learning_rate": 3.945968188823782e-06, "loss": 1.3834, "step": 29014 }, { "epoch": 2.2739028213166144, "grad_norm": 15.528502464294434, "learning_rate": 3.945666758331387e-06, "loss": 0.7503, "step": 29015 }, { "epoch": 2.2739811912225707, "grad_norm": 7.76888370513916, "learning_rate": 3.945365331849134e-06, "loss": 1.4111, "step": 29016 }, { "epoch": 2.2740595611285266, "grad_norm": 2.0901918411254883, "learning_rate": 3.945063909378171e-06, "loss": 1.0025, "step": 29017 }, { "epoch": 2.274137931034483, "grad_norm": 2.7819619178771973, "learning_rate": 3.944762490919642e-06, "loss": 1.0554, "step": 29018 }, { "epoch": 2.274216300940439, "grad_norm": 2.5388736724853516, "learning_rate": 3.944461076474694e-06, "loss": 1.2973, "step": 29019 }, { "epoch": 2.274294670846395, "grad_norm": 4.643214225769043, "learning_rate": 3.944159666044475e-06, "loss": 1.1072, "step": 29020 }, { "epoch": 2.274373040752351, "grad_norm": 3.8265910148620605, "learning_rate": 3.943858259630131e-06, "loss": 1.5535, "step": 29021 }, { "epoch": 2.2744514106583074, "grad_norm": 3.019519805908203, "learning_rate": 3.943556857232806e-06, "loss": 0.8811, "step": 29022 }, { "epoch": 2.2745297805642632, "grad_norm": 4.397955417633057, "learning_rate": 3.9432554588536474e-06, "loss": 1.7026, "step": 29023 }, { "epoch": 2.2746081504702196, "grad_norm": 4.473117828369141, "learning_rate": 3.942954064493803e-06, "loss": 1.0268, "step": 29024 }, { "epoch": 2.2746865203761755, "grad_norm": 2.3864829540252686, "learning_rate": 3.9426526741544205e-06, "loss": 1.3793, "step": 29025 }, { "epoch": 2.274764890282132, "grad_norm": 2.9830474853515625, "learning_rate": 3.942351287836642e-06, "loss": 0.7008, "step": 29026 }, { "epoch": 2.2748432601880877, "grad_norm": 3.3599798679351807, "learning_rate": 3.942049905541616e-06, "loss": 1.3601, "step": 29027 }, { "epoch": 2.274921630094044, "grad_norm": 3.4536330699920654, "learning_rate": 3.941748527270488e-06, "loss": 1.2369, "step": 29028 }, { "epoch": 2.275, "grad_norm": 2.3768208026885986, "learning_rate": 3.9414471530244084e-06, "loss": 1.211, "step": 29029 }, { "epoch": 2.2750783699059562, "grad_norm": 4.104726791381836, "learning_rate": 3.941145782804517e-06, "loss": 1.4194, "step": 29030 }, { "epoch": 2.275156739811912, "grad_norm": 2.309478998184204, "learning_rate": 3.940844416611964e-06, "loss": 1.1318, "step": 29031 }, { "epoch": 2.2752351097178685, "grad_norm": 4.745169162750244, "learning_rate": 3.940543054447896e-06, "loss": 0.7742, "step": 29032 }, { "epoch": 2.2753134796238244, "grad_norm": 2.615546226501465, "learning_rate": 3.940241696313457e-06, "loss": 1.3248, "step": 29033 }, { "epoch": 2.2753918495297807, "grad_norm": 4.140808582305908, "learning_rate": 3.939940342209794e-06, "loss": 1.1123, "step": 29034 }, { "epoch": 2.2754702194357366, "grad_norm": 5.29416561126709, "learning_rate": 3.939638992138053e-06, "loss": 1.4294, "step": 29035 }, { "epoch": 2.275548589341693, "grad_norm": 2.3770155906677246, "learning_rate": 3.939337646099383e-06, "loss": 1.4118, "step": 29036 }, { "epoch": 2.275626959247649, "grad_norm": 4.779952049255371, "learning_rate": 3.939036304094926e-06, "loss": 1.4571, "step": 29037 }, { "epoch": 2.275705329153605, "grad_norm": 3.618790864944458, "learning_rate": 3.938734966125829e-06, "loss": 1.3486, "step": 29038 }, { "epoch": 2.275783699059561, "grad_norm": 3.761784315109253, "learning_rate": 3.938433632193241e-06, "loss": 1.4886, "step": 29039 }, { "epoch": 2.2758620689655173, "grad_norm": 3.387991189956665, "learning_rate": 3.938132302298306e-06, "loss": 1.8415, "step": 29040 }, { "epoch": 2.2759404388714732, "grad_norm": 3.0643606185913086, "learning_rate": 3.937830976442169e-06, "loss": 1.7229, "step": 29041 }, { "epoch": 2.2760188087774296, "grad_norm": 2.3028643131256104, "learning_rate": 3.937529654625978e-06, "loss": 1.2113, "step": 29042 }, { "epoch": 2.2760971786833855, "grad_norm": 3.0004727840423584, "learning_rate": 3.937228336850878e-06, "loss": 1.0868, "step": 29043 }, { "epoch": 2.276175548589342, "grad_norm": 3.771624803543091, "learning_rate": 3.936927023118017e-06, "loss": 1.3753, "step": 29044 }, { "epoch": 2.2762539184952977, "grad_norm": 2.233288526535034, "learning_rate": 3.936625713428539e-06, "loss": 1.3125, "step": 29045 }, { "epoch": 2.276332288401254, "grad_norm": 1.9975041151046753, "learning_rate": 3.93632440778359e-06, "loss": 1.1991, "step": 29046 }, { "epoch": 2.27641065830721, "grad_norm": 3.520556926727295, "learning_rate": 3.936023106184318e-06, "loss": 1.2106, "step": 29047 }, { "epoch": 2.2764890282131662, "grad_norm": 3.451897621154785, "learning_rate": 3.935721808631867e-06, "loss": 1.0193, "step": 29048 }, { "epoch": 2.276567398119122, "grad_norm": 3.4841620922088623, "learning_rate": 3.9354205151273815e-06, "loss": 1.3292, "step": 29049 }, { "epoch": 2.2766457680250785, "grad_norm": 3.3302292823791504, "learning_rate": 3.935119225672011e-06, "loss": 2.0603, "step": 29050 }, { "epoch": 2.2767241379310343, "grad_norm": 2.684605836868286, "learning_rate": 3.934817940266901e-06, "loss": 1.1872, "step": 29051 }, { "epoch": 2.2768025078369907, "grad_norm": 14.536033630371094, "learning_rate": 3.934516658913196e-06, "loss": 0.958, "step": 29052 }, { "epoch": 2.2768808777429466, "grad_norm": 3.349125862121582, "learning_rate": 3.9342153816120425e-06, "loss": 1.3547, "step": 29053 }, { "epoch": 2.276959247648903, "grad_norm": 2.1590070724487305, "learning_rate": 3.933914108364584e-06, "loss": 1.4778, "step": 29054 }, { "epoch": 2.277037617554859, "grad_norm": 4.163966655731201, "learning_rate": 3.933612839171973e-06, "loss": 1.4885, "step": 29055 }, { "epoch": 2.277115987460815, "grad_norm": 2.8295278549194336, "learning_rate": 3.933311574035347e-06, "loss": 1.4563, "step": 29056 }, { "epoch": 2.277194357366771, "grad_norm": 5.215262413024902, "learning_rate": 3.933010312955858e-06, "loss": 0.9943, "step": 29057 }, { "epoch": 2.2772727272727273, "grad_norm": 3.097111463546753, "learning_rate": 3.932709055934649e-06, "loss": 1.0098, "step": 29058 }, { "epoch": 2.2773510971786832, "grad_norm": 3.361771583557129, "learning_rate": 3.9324078029728664e-06, "loss": 1.5458, "step": 29059 }, { "epoch": 2.2774294670846396, "grad_norm": 4.528814315795898, "learning_rate": 3.932106554071657e-06, "loss": 1.5907, "step": 29060 }, { "epoch": 2.2775078369905954, "grad_norm": 3.0425305366516113, "learning_rate": 3.931805309232164e-06, "loss": 1.4958, "step": 29061 }, { "epoch": 2.277586206896552, "grad_norm": 5.297169208526611, "learning_rate": 3.931504068455538e-06, "loss": 1.4364, "step": 29062 }, { "epoch": 2.2776645768025077, "grad_norm": 3.3782050609588623, "learning_rate": 3.9312028317429186e-06, "loss": 1.2187, "step": 29063 }, { "epoch": 2.277742946708464, "grad_norm": 4.428186893463135, "learning_rate": 3.930901599095456e-06, "loss": 1.6141, "step": 29064 }, { "epoch": 2.27782131661442, "grad_norm": 5.061298370361328, "learning_rate": 3.930600370514294e-06, "loss": 1.4905, "step": 29065 }, { "epoch": 2.277899686520376, "grad_norm": 3.080202341079712, "learning_rate": 3.930299146000581e-06, "loss": 1.2527, "step": 29066 }, { "epoch": 2.277978056426332, "grad_norm": 2.956961154937744, "learning_rate": 3.9299979255554584e-06, "loss": 1.3088, "step": 29067 }, { "epoch": 2.2780564263322884, "grad_norm": 2.045549154281616, "learning_rate": 3.929696709180074e-06, "loss": 1.2021, "step": 29068 }, { "epoch": 2.2781347962382443, "grad_norm": 5.170082092285156, "learning_rate": 3.9293954968755735e-06, "loss": 1.0384, "step": 29069 }, { "epoch": 2.2782131661442007, "grad_norm": 5.070509433746338, "learning_rate": 3.929094288643105e-06, "loss": 1.1408, "step": 29070 }, { "epoch": 2.2782915360501566, "grad_norm": 4.465334415435791, "learning_rate": 3.9287930844838094e-06, "loss": 1.3874, "step": 29071 }, { "epoch": 2.278369905956113, "grad_norm": 5.553513050079346, "learning_rate": 3.928491884398836e-06, "loss": 0.8397, "step": 29072 }, { "epoch": 2.2784482758620688, "grad_norm": 3.391728401184082, "learning_rate": 3.928190688389329e-06, "loss": 1.2129, "step": 29073 }, { "epoch": 2.278526645768025, "grad_norm": 8.61430835723877, "learning_rate": 3.927889496456434e-06, "loss": 0.4901, "step": 29074 }, { "epoch": 2.278605015673981, "grad_norm": 3.2220406532287598, "learning_rate": 3.927588308601296e-06, "loss": 1.6751, "step": 29075 }, { "epoch": 2.2786833855799373, "grad_norm": 2.086524248123169, "learning_rate": 3.927287124825061e-06, "loss": 1.4983, "step": 29076 }, { "epoch": 2.2787617554858937, "grad_norm": 5.514288425445557, "learning_rate": 3.926985945128876e-06, "loss": 1.5839, "step": 29077 }, { "epoch": 2.2788401253918495, "grad_norm": 2.9963676929473877, "learning_rate": 3.926684769513885e-06, "loss": 1.4002, "step": 29078 }, { "epoch": 2.2789184952978054, "grad_norm": 3.0699825286865234, "learning_rate": 3.926383597981233e-06, "loss": 1.1262, "step": 29079 }, { "epoch": 2.2789968652037618, "grad_norm": 3.272557497024536, "learning_rate": 3.926082430532067e-06, "loss": 1.3453, "step": 29080 }, { "epoch": 2.279075235109718, "grad_norm": 7.0729079246521, "learning_rate": 3.925781267167532e-06, "loss": 1.3263, "step": 29081 }, { "epoch": 2.279153605015674, "grad_norm": 3.777268171310425, "learning_rate": 3.925480107888773e-06, "loss": 1.2985, "step": 29082 }, { "epoch": 2.27923197492163, "grad_norm": 16.095521926879883, "learning_rate": 3.925178952696936e-06, "loss": 1.0484, "step": 29083 }, { "epoch": 2.279310344827586, "grad_norm": 4.376072406768799, "learning_rate": 3.924877801593166e-06, "loss": 1.3933, "step": 29084 }, { "epoch": 2.2793887147335425, "grad_norm": 3.932823419570923, "learning_rate": 3.92457665457861e-06, "loss": 1.3734, "step": 29085 }, { "epoch": 2.2794670846394984, "grad_norm": 3.5130550861358643, "learning_rate": 3.92427551165441e-06, "loss": 1.4177, "step": 29086 }, { "epoch": 2.2795454545454543, "grad_norm": 4.8446044921875, "learning_rate": 3.923974372821713e-06, "loss": 1.6253, "step": 29087 }, { "epoch": 2.2796238244514107, "grad_norm": 4.12554931640625, "learning_rate": 3.9236732380816675e-06, "loss": 1.6436, "step": 29088 }, { "epoch": 2.279702194357367, "grad_norm": 3.052607774734497, "learning_rate": 3.923372107435414e-06, "loss": 1.246, "step": 29089 }, { "epoch": 2.279780564263323, "grad_norm": 7.882229328155518, "learning_rate": 3.923070980884101e-06, "loss": 1.3754, "step": 29090 }, { "epoch": 2.2798589341692788, "grad_norm": 5.126818656921387, "learning_rate": 3.922769858428872e-06, "loss": 1.6968, "step": 29091 }, { "epoch": 2.279937304075235, "grad_norm": 3.855952739715576, "learning_rate": 3.922468740070875e-06, "loss": 1.2543, "step": 29092 }, { "epoch": 2.2800156739811914, "grad_norm": 3.022209882736206, "learning_rate": 3.922167625811251e-06, "loss": 0.8466, "step": 29093 }, { "epoch": 2.2800940438871473, "grad_norm": 2.4669103622436523, "learning_rate": 3.921866515651148e-06, "loss": 1.3277, "step": 29094 }, { "epoch": 2.280172413793103, "grad_norm": 11.260560989379883, "learning_rate": 3.921565409591712e-06, "loss": 0.9035, "step": 29095 }, { "epoch": 2.2802507836990595, "grad_norm": 3.3405215740203857, "learning_rate": 3.921264307634088e-06, "loss": 1.4183, "step": 29096 }, { "epoch": 2.280329153605016, "grad_norm": 3.6787357330322266, "learning_rate": 3.92096320977942e-06, "loss": 1.2841, "step": 29097 }, { "epoch": 2.2804075235109718, "grad_norm": 4.557681083679199, "learning_rate": 3.920662116028852e-06, "loss": 1.6542, "step": 29098 }, { "epoch": 2.280485893416928, "grad_norm": 3.3871448040008545, "learning_rate": 3.9203610263835345e-06, "loss": 1.1779, "step": 29099 }, { "epoch": 2.280564263322884, "grad_norm": 3.4531924724578857, "learning_rate": 3.920059940844606e-06, "loss": 1.1985, "step": 29100 }, { "epoch": 2.2806426332288403, "grad_norm": 2.245670795440674, "learning_rate": 3.919758859413215e-06, "loss": 1.153, "step": 29101 }, { "epoch": 2.280721003134796, "grad_norm": 3.68038010597229, "learning_rate": 3.919457782090508e-06, "loss": 1.2252, "step": 29102 }, { "epoch": 2.2807993730407525, "grad_norm": 2.8409714698791504, "learning_rate": 3.919156708877628e-06, "loss": 1.1991, "step": 29103 }, { "epoch": 2.2808777429467084, "grad_norm": 2.7352945804595947, "learning_rate": 3.918855639775721e-06, "loss": 0.8133, "step": 29104 }, { "epoch": 2.2809561128526648, "grad_norm": 4.881478309631348, "learning_rate": 3.91855457478593e-06, "loss": 1.3342, "step": 29105 }, { "epoch": 2.2810344827586206, "grad_norm": 2.4313430786132812, "learning_rate": 3.918253513909402e-06, "loss": 1.3294, "step": 29106 }, { "epoch": 2.281112852664577, "grad_norm": 3.1073334217071533, "learning_rate": 3.917952457147285e-06, "loss": 1.4064, "step": 29107 }, { "epoch": 2.281191222570533, "grad_norm": 6.771124362945557, "learning_rate": 3.917651404500719e-06, "loss": 0.8606, "step": 29108 }, { "epoch": 2.281269592476489, "grad_norm": 3.7805089950561523, "learning_rate": 3.917350355970852e-06, "loss": 1.7716, "step": 29109 }, { "epoch": 2.281347962382445, "grad_norm": 5.150003910064697, "learning_rate": 3.917049311558826e-06, "loss": 1.3983, "step": 29110 }, { "epoch": 2.2814263322884014, "grad_norm": 3.218848943710327, "learning_rate": 3.916748271265792e-06, "loss": 1.6957, "step": 29111 }, { "epoch": 2.2815047021943573, "grad_norm": 2.739719867706299, "learning_rate": 3.916447235092887e-06, "loss": 1.3663, "step": 29112 }, { "epoch": 2.2815830721003136, "grad_norm": 2.9477620124816895, "learning_rate": 3.916146203041263e-06, "loss": 0.9649, "step": 29113 }, { "epoch": 2.2816614420062695, "grad_norm": 3.861572742462158, "learning_rate": 3.9158451751120615e-06, "loss": 1.622, "step": 29114 }, { "epoch": 2.281739811912226, "grad_norm": 2.9140021800994873, "learning_rate": 3.915544151306428e-06, "loss": 1.3201, "step": 29115 }, { "epoch": 2.2818181818181817, "grad_norm": 2.2633888721466064, "learning_rate": 3.915243131625508e-06, "loss": 1.2272, "step": 29116 }, { "epoch": 2.281896551724138, "grad_norm": 4.588357448577881, "learning_rate": 3.914942116070444e-06, "loss": 1.4868, "step": 29117 }, { "epoch": 2.281974921630094, "grad_norm": 9.364465713500977, "learning_rate": 3.914641104642386e-06, "loss": 1.7478, "step": 29118 }, { "epoch": 2.2820532915360503, "grad_norm": 3.028578996658325, "learning_rate": 3.914340097342472e-06, "loss": 1.3855, "step": 29119 }, { "epoch": 2.282131661442006, "grad_norm": 4.2530927658081055, "learning_rate": 3.9140390941718535e-06, "loss": 0.9269, "step": 29120 }, { "epoch": 2.2822100313479625, "grad_norm": 3.6080753803253174, "learning_rate": 3.913738095131671e-06, "loss": 1.0378, "step": 29121 }, { "epoch": 2.2822884012539184, "grad_norm": 4.637319087982178, "learning_rate": 3.913437100223073e-06, "loss": 1.3316, "step": 29122 }, { "epoch": 2.2823667711598747, "grad_norm": 2.5759835243225098, "learning_rate": 3.913136109447201e-06, "loss": 1.2824, "step": 29123 }, { "epoch": 2.2824451410658306, "grad_norm": 2.8667819499969482, "learning_rate": 3.912835122805199e-06, "loss": 1.376, "step": 29124 }, { "epoch": 2.282523510971787, "grad_norm": 1.9311292171478271, "learning_rate": 3.912534140298216e-06, "loss": 1.337, "step": 29125 }, { "epoch": 2.282601880877743, "grad_norm": 2.8859477043151855, "learning_rate": 3.912233161927394e-06, "loss": 1.3703, "step": 29126 }, { "epoch": 2.282680250783699, "grad_norm": 3.1811516284942627, "learning_rate": 3.911932187693879e-06, "loss": 1.3676, "step": 29127 }, { "epoch": 2.282758620689655, "grad_norm": 2.8398492336273193, "learning_rate": 3.911631217598814e-06, "loss": 1.1378, "step": 29128 }, { "epoch": 2.2828369905956114, "grad_norm": 3.1284286975860596, "learning_rate": 3.911330251643346e-06, "loss": 1.8761, "step": 29129 }, { "epoch": 2.2829153605015673, "grad_norm": 5.285719871520996, "learning_rate": 3.911029289828618e-06, "loss": 1.0811, "step": 29130 }, { "epoch": 2.2829937304075236, "grad_norm": 4.481654644012451, "learning_rate": 3.910728332155774e-06, "loss": 1.6475, "step": 29131 }, { "epoch": 2.2830721003134795, "grad_norm": 2.936052083969116, "learning_rate": 3.910427378625961e-06, "loss": 0.7573, "step": 29132 }, { "epoch": 2.283150470219436, "grad_norm": 5.314664840698242, "learning_rate": 3.910126429240324e-06, "loss": 1.4552, "step": 29133 }, { "epoch": 2.2832288401253917, "grad_norm": 5.080419063568115, "learning_rate": 3.909825484000004e-06, "loss": 1.0701, "step": 29134 }, { "epoch": 2.283307210031348, "grad_norm": 3.0230069160461426, "learning_rate": 3.909524542906149e-06, "loss": 1.244, "step": 29135 }, { "epoch": 2.283385579937304, "grad_norm": 4.043539047241211, "learning_rate": 3.909223605959901e-06, "loss": 1.243, "step": 29136 }, { "epoch": 2.2834639498432603, "grad_norm": 2.746131181716919, "learning_rate": 3.90892267316241e-06, "loss": 1.4609, "step": 29137 }, { "epoch": 2.283542319749216, "grad_norm": 2.489272117614746, "learning_rate": 3.9086217445148125e-06, "loss": 1.5378, "step": 29138 }, { "epoch": 2.2836206896551725, "grad_norm": 2.8512625694274902, "learning_rate": 3.9083208200182595e-06, "loss": 1.3628, "step": 29139 }, { "epoch": 2.2836990595611284, "grad_norm": 2.356679916381836, "learning_rate": 3.908019899673893e-06, "loss": 1.2839, "step": 29140 }, { "epoch": 2.2837774294670847, "grad_norm": 3.8103954792022705, "learning_rate": 3.9077189834828586e-06, "loss": 1.231, "step": 29141 }, { "epoch": 2.2838557993730406, "grad_norm": 4.153494834899902, "learning_rate": 3.907418071446299e-06, "loss": 1.2373, "step": 29142 }, { "epoch": 2.283934169278997, "grad_norm": 3.2141175270080566, "learning_rate": 3.90711716356536e-06, "loss": 1.4179, "step": 29143 }, { "epoch": 2.284012539184953, "grad_norm": 4.447955131530762, "learning_rate": 3.906816259841188e-06, "loss": 0.8598, "step": 29144 }, { "epoch": 2.284090909090909, "grad_norm": 3.883098840713501, "learning_rate": 3.906515360274922e-06, "loss": 1.1895, "step": 29145 }, { "epoch": 2.284169278996865, "grad_norm": 4.345818996429443, "learning_rate": 3.9062144648677115e-06, "loss": 1.1046, "step": 29146 }, { "epoch": 2.2842476489028214, "grad_norm": 2.876688003540039, "learning_rate": 3.905913573620699e-06, "loss": 1.088, "step": 29147 }, { "epoch": 2.2843260188087773, "grad_norm": 3.692842483520508, "learning_rate": 3.905612686535031e-06, "loss": 1.3317, "step": 29148 }, { "epoch": 2.2844043887147336, "grad_norm": 4.151209831237793, "learning_rate": 3.9053118036118494e-06, "loss": 1.3681, "step": 29149 }, { "epoch": 2.2844827586206895, "grad_norm": 2.5099923610687256, "learning_rate": 3.905010924852297e-06, "loss": 1.7037, "step": 29150 }, { "epoch": 2.284561128526646, "grad_norm": 4.258016109466553, "learning_rate": 3.904710050257523e-06, "loss": 1.0674, "step": 29151 }, { "epoch": 2.2846394984326017, "grad_norm": 3.3488199710845947, "learning_rate": 3.90440917982867e-06, "loss": 1.1376, "step": 29152 }, { "epoch": 2.284717868338558, "grad_norm": 2.65376615524292, "learning_rate": 3.9041083135668815e-06, "loss": 1.3036, "step": 29153 }, { "epoch": 2.284796238244514, "grad_norm": 2.836374282836914, "learning_rate": 3.903807451473301e-06, "loss": 1.1062, "step": 29154 }, { "epoch": 2.2848746081504703, "grad_norm": 6.5228424072265625, "learning_rate": 3.903506593549075e-06, "loss": 0.8903, "step": 29155 }, { "epoch": 2.284952978056426, "grad_norm": 2.185504198074341, "learning_rate": 3.903205739795346e-06, "loss": 1.123, "step": 29156 }, { "epoch": 2.2850313479623825, "grad_norm": 4.0615620613098145, "learning_rate": 3.902904890213258e-06, "loss": 1.3728, "step": 29157 }, { "epoch": 2.2851097178683384, "grad_norm": 3.498548984527588, "learning_rate": 3.902604044803957e-06, "loss": 1.2549, "step": 29158 }, { "epoch": 2.2851880877742947, "grad_norm": 3.7173256874084473, "learning_rate": 3.902303203568588e-06, "loss": 1.4574, "step": 29159 }, { "epoch": 2.2852664576802506, "grad_norm": 2.8197860717773438, "learning_rate": 3.902002366508293e-06, "loss": 1.4016, "step": 29160 }, { "epoch": 2.285344827586207, "grad_norm": 5.710289478302002, "learning_rate": 3.9017015336242165e-06, "loss": 1.317, "step": 29161 }, { "epoch": 2.285423197492163, "grad_norm": 2.649691581726074, "learning_rate": 3.9014007049175015e-06, "loss": 1.5322, "step": 29162 }, { "epoch": 2.285501567398119, "grad_norm": 2.5193002223968506, "learning_rate": 3.901099880389298e-06, "loss": 1.2474, "step": 29163 }, { "epoch": 2.285579937304075, "grad_norm": 3.0385043621063232, "learning_rate": 3.9007990600407435e-06, "loss": 1.6704, "step": 29164 }, { "epoch": 2.2856583072100314, "grad_norm": 2.7443466186523438, "learning_rate": 3.900498243872986e-06, "loss": 1.4898, "step": 29165 }, { "epoch": 2.2857366771159873, "grad_norm": 2.251201868057251, "learning_rate": 3.900197431887167e-06, "loss": 1.1923, "step": 29166 }, { "epoch": 2.2858150470219436, "grad_norm": 2.8390839099884033, "learning_rate": 3.899896624084434e-06, "loss": 1.3444, "step": 29167 }, { "epoch": 2.2858934169279, "grad_norm": 3.42976713180542, "learning_rate": 3.899595820465928e-06, "loss": 1.2554, "step": 29168 }, { "epoch": 2.285971786833856, "grad_norm": 3.476573944091797, "learning_rate": 3.899295021032793e-06, "loss": 1.3201, "step": 29169 }, { "epoch": 2.2860501567398117, "grad_norm": 3.717906951904297, "learning_rate": 3.898994225786178e-06, "loss": 1.398, "step": 29170 }, { "epoch": 2.286128526645768, "grad_norm": 2.664396286010742, "learning_rate": 3.8986934347272194e-06, "loss": 1.3104, "step": 29171 }, { "epoch": 2.2862068965517244, "grad_norm": 3.7385826110839844, "learning_rate": 3.898392647857067e-06, "loss": 1.4869, "step": 29172 }, { "epoch": 2.2862852664576803, "grad_norm": 5.736825466156006, "learning_rate": 3.8980918651768635e-06, "loss": 0.8525, "step": 29173 }, { "epoch": 2.286363636363636, "grad_norm": 6.374131202697754, "learning_rate": 3.897791086687753e-06, "loss": 1.1566, "step": 29174 }, { "epoch": 2.2864420062695925, "grad_norm": 4.212503910064697, "learning_rate": 3.8974903123908786e-06, "loss": 1.4296, "step": 29175 }, { "epoch": 2.286520376175549, "grad_norm": 3.6021568775177, "learning_rate": 3.897189542287383e-06, "loss": 1.417, "step": 29176 }, { "epoch": 2.2865987460815047, "grad_norm": 5.10023307800293, "learning_rate": 3.8968887763784135e-06, "loss": 1.2524, "step": 29177 }, { "epoch": 2.2866771159874606, "grad_norm": 3.0657007694244385, "learning_rate": 3.896588014665114e-06, "loss": 1.6281, "step": 29178 }, { "epoch": 2.286755485893417, "grad_norm": 3.6567189693450928, "learning_rate": 3.896287257148625e-06, "loss": 1.419, "step": 29179 }, { "epoch": 2.2868338557993733, "grad_norm": 2.4969489574432373, "learning_rate": 3.895986503830093e-06, "loss": 1.3805, "step": 29180 }, { "epoch": 2.286912225705329, "grad_norm": 4.082997798919678, "learning_rate": 3.895685754710659e-06, "loss": 1.7544, "step": 29181 }, { "epoch": 2.286990595611285, "grad_norm": 3.6395368576049805, "learning_rate": 3.895385009791473e-06, "loss": 1.2943, "step": 29182 }, { "epoch": 2.2870689655172414, "grad_norm": 2.5061514377593994, "learning_rate": 3.895084269073673e-06, "loss": 1.5972, "step": 29183 }, { "epoch": 2.2871473354231977, "grad_norm": 3.6246535778045654, "learning_rate": 3.894783532558405e-06, "loss": 1.7279, "step": 29184 }, { "epoch": 2.2872257053291536, "grad_norm": 3.206181287765503, "learning_rate": 3.894482800246814e-06, "loss": 1.6315, "step": 29185 }, { "epoch": 2.2873040752351095, "grad_norm": 11.14071273803711, "learning_rate": 3.894182072140042e-06, "loss": 1.8096, "step": 29186 }, { "epoch": 2.287382445141066, "grad_norm": 2.768341302871704, "learning_rate": 3.893881348239232e-06, "loss": 1.0511, "step": 29187 }, { "epoch": 2.287460815047022, "grad_norm": 4.182125568389893, "learning_rate": 3.89358062854553e-06, "loss": 1.3775, "step": 29188 }, { "epoch": 2.287539184952978, "grad_norm": 10.449433326721191, "learning_rate": 3.89327991306008e-06, "loss": 2.0646, "step": 29189 }, { "epoch": 2.287617554858934, "grad_norm": 6.1285881996154785, "learning_rate": 3.892979201784024e-06, "loss": 1.3169, "step": 29190 }, { "epoch": 2.2876959247648903, "grad_norm": 2.905759811401367, "learning_rate": 3.892678494718506e-06, "loss": 1.5136, "step": 29191 }, { "epoch": 2.2877742946708466, "grad_norm": 1.997418999671936, "learning_rate": 3.892377791864671e-06, "loss": 1.1946, "step": 29192 }, { "epoch": 2.2878526645768025, "grad_norm": 4.219374656677246, "learning_rate": 3.892077093223663e-06, "loss": 1.1605, "step": 29193 }, { "epoch": 2.287931034482759, "grad_norm": 5.747287273406982, "learning_rate": 3.891776398796622e-06, "loss": 1.2468, "step": 29194 }, { "epoch": 2.2880094043887147, "grad_norm": 6.359109878540039, "learning_rate": 3.891475708584695e-06, "loss": 1.8435, "step": 29195 }, { "epoch": 2.288087774294671, "grad_norm": 3.985696792602539, "learning_rate": 3.891175022589026e-06, "loss": 1.406, "step": 29196 }, { "epoch": 2.288166144200627, "grad_norm": 5.564922332763672, "learning_rate": 3.890874340810759e-06, "loss": 1.5596, "step": 29197 }, { "epoch": 2.2882445141065832, "grad_norm": 2.6975667476654053, "learning_rate": 3.890573663251035e-06, "loss": 1.2512, "step": 29198 }, { "epoch": 2.288322884012539, "grad_norm": 2.82122802734375, "learning_rate": 3.890272989910997e-06, "loss": 0.7309, "step": 29199 }, { "epoch": 2.2884012539184955, "grad_norm": 2.184617280960083, "learning_rate": 3.889972320791794e-06, "loss": 1.4765, "step": 29200 }, { "epoch": 2.2884796238244514, "grad_norm": 2.510129928588867, "learning_rate": 3.889671655894564e-06, "loss": 1.0981, "step": 29201 }, { "epoch": 2.2885579937304077, "grad_norm": 2.3109066486358643, "learning_rate": 3.8893709952204525e-06, "loss": 1.3249, "step": 29202 }, { "epoch": 2.2886363636363636, "grad_norm": 2.3493125438690186, "learning_rate": 3.889070338770604e-06, "loss": 1.2254, "step": 29203 }, { "epoch": 2.28871473354232, "grad_norm": 6.915862083435059, "learning_rate": 3.888769686546162e-06, "loss": 0.6592, "step": 29204 }, { "epoch": 2.288793103448276, "grad_norm": 2.9626305103302, "learning_rate": 3.888469038548269e-06, "loss": 1.1265, "step": 29205 }, { "epoch": 2.288871473354232, "grad_norm": 6.380308628082275, "learning_rate": 3.8881683947780665e-06, "loss": 1.2739, "step": 29206 }, { "epoch": 2.288949843260188, "grad_norm": 2.4507052898406982, "learning_rate": 3.887867755236703e-06, "loss": 1.292, "step": 29207 }, { "epoch": 2.2890282131661444, "grad_norm": 3.0781595706939697, "learning_rate": 3.88756711992532e-06, "loss": 1.4308, "step": 29208 }, { "epoch": 2.2891065830721002, "grad_norm": 2.290530204772949, "learning_rate": 3.887266488845058e-06, "loss": 1.3947, "step": 29209 }, { "epoch": 2.2891849529780566, "grad_norm": 2.4727354049682617, "learning_rate": 3.8869658619970645e-06, "loss": 1.1651, "step": 29210 }, { "epoch": 2.2892633228840125, "grad_norm": 3.694112777709961, "learning_rate": 3.8866652393824805e-06, "loss": 1.3394, "step": 29211 }, { "epoch": 2.289341692789969, "grad_norm": 4.136892795562744, "learning_rate": 3.88636462100245e-06, "loss": 1.3176, "step": 29212 }, { "epoch": 2.2894200626959247, "grad_norm": 2.642747402191162, "learning_rate": 3.8860640068581156e-06, "loss": 1.3555, "step": 29213 }, { "epoch": 2.289498432601881, "grad_norm": 3.682565212249756, "learning_rate": 3.885763396950622e-06, "loss": 1.4678, "step": 29214 }, { "epoch": 2.289576802507837, "grad_norm": 5.993201732635498, "learning_rate": 3.885462791281114e-06, "loss": 1.8236, "step": 29215 }, { "epoch": 2.2896551724137932, "grad_norm": 2.5667405128479004, "learning_rate": 3.885162189850731e-06, "loss": 1.1911, "step": 29216 }, { "epoch": 2.289733542319749, "grad_norm": 3.660212516784668, "learning_rate": 3.884861592660619e-06, "loss": 1.4364, "step": 29217 }, { "epoch": 2.2898119122257055, "grad_norm": 3.4039387702941895, "learning_rate": 3.884560999711919e-06, "loss": 1.2506, "step": 29218 }, { "epoch": 2.2898902821316613, "grad_norm": 5.244045734405518, "learning_rate": 3.88426041100578e-06, "loss": 1.2111, "step": 29219 }, { "epoch": 2.2899686520376177, "grad_norm": 2.2382102012634277, "learning_rate": 3.883959826543338e-06, "loss": 1.2926, "step": 29220 }, { "epoch": 2.2900470219435736, "grad_norm": 3.747591495513916, "learning_rate": 3.8836592463257404e-06, "loss": 1.5699, "step": 29221 }, { "epoch": 2.29012539184953, "grad_norm": 6.785645484924316, "learning_rate": 3.88335867035413e-06, "loss": 1.7198, "step": 29222 }, { "epoch": 2.290203761755486, "grad_norm": 3.7431957721710205, "learning_rate": 3.883058098629651e-06, "loss": 1.3616, "step": 29223 }, { "epoch": 2.290282131661442, "grad_norm": 4.799402713775635, "learning_rate": 3.882757531153444e-06, "loss": 1.1422, "step": 29224 }, { "epoch": 2.290360501567398, "grad_norm": 4.188717365264893, "learning_rate": 3.882456967926652e-06, "loss": 1.3693, "step": 29225 }, { "epoch": 2.2904388714733543, "grad_norm": 2.813873767852783, "learning_rate": 3.882156408950423e-06, "loss": 1.2812, "step": 29226 }, { "epoch": 2.2905172413793102, "grad_norm": 2.7688448429107666, "learning_rate": 3.881855854225894e-06, "loss": 1.5413, "step": 29227 }, { "epoch": 2.2905956112852666, "grad_norm": 4.733388900756836, "learning_rate": 3.881555303754212e-06, "loss": 1.267, "step": 29228 }, { "epoch": 2.2906739811912225, "grad_norm": 4.182578086853027, "learning_rate": 3.881254757536519e-06, "loss": 1.3753, "step": 29229 }, { "epoch": 2.290752351097179, "grad_norm": 2.645923614501953, "learning_rate": 3.880954215573959e-06, "loss": 1.474, "step": 29230 }, { "epoch": 2.2908307210031347, "grad_norm": 6.477014064788818, "learning_rate": 3.880653677867674e-06, "loss": 1.6292, "step": 29231 }, { "epoch": 2.290909090909091, "grad_norm": 2.532136917114258, "learning_rate": 3.880353144418806e-06, "loss": 1.421, "step": 29232 }, { "epoch": 2.290987460815047, "grad_norm": 6.163817405700684, "learning_rate": 3.880052615228501e-06, "loss": 1.0065, "step": 29233 }, { "epoch": 2.2910658307210032, "grad_norm": 2.8253960609436035, "learning_rate": 3.879752090297901e-06, "loss": 1.377, "step": 29234 }, { "epoch": 2.291144200626959, "grad_norm": 2.008122682571411, "learning_rate": 3.879451569628148e-06, "loss": 1.4376, "step": 29235 }, { "epoch": 2.2912225705329154, "grad_norm": 2.7787060737609863, "learning_rate": 3.8791510532203855e-06, "loss": 1.3201, "step": 29236 }, { "epoch": 2.2913009404388713, "grad_norm": 2.1467442512512207, "learning_rate": 3.878850541075756e-06, "loss": 1.3234, "step": 29237 }, { "epoch": 2.2913793103448277, "grad_norm": 6.911529064178467, "learning_rate": 3.878550033195406e-06, "loss": 1.0685, "step": 29238 }, { "epoch": 2.2914576802507836, "grad_norm": 4.28331995010376, "learning_rate": 3.8782495295804725e-06, "loss": 1.1515, "step": 29239 }, { "epoch": 2.29153605015674, "grad_norm": 3.8777217864990234, "learning_rate": 3.877949030232103e-06, "loss": 1.2868, "step": 29240 }, { "epoch": 2.291614420062696, "grad_norm": 2.8074355125427246, "learning_rate": 3.87764853515144e-06, "loss": 1.1994, "step": 29241 }, { "epoch": 2.291692789968652, "grad_norm": 3.4171507358551025, "learning_rate": 3.877348044339624e-06, "loss": 1.6045, "step": 29242 }, { "epoch": 2.291771159874608, "grad_norm": 2.5504181385040283, "learning_rate": 3.8770475577978e-06, "loss": 1.2948, "step": 29243 }, { "epoch": 2.2918495297805643, "grad_norm": 4.939938068389893, "learning_rate": 3.876747075527109e-06, "loss": 1.8518, "step": 29244 }, { "epoch": 2.29192789968652, "grad_norm": 2.0632212162017822, "learning_rate": 3.876446597528698e-06, "loss": 1.3367, "step": 29245 }, { "epoch": 2.2920062695924766, "grad_norm": 2.95056414604187, "learning_rate": 3.876146123803704e-06, "loss": 1.229, "step": 29246 }, { "epoch": 2.2920846394984324, "grad_norm": 3.005889892578125, "learning_rate": 3.875845654353275e-06, "loss": 1.6826, "step": 29247 }, { "epoch": 2.2921630094043888, "grad_norm": 7.062244415283203, "learning_rate": 3.87554518917855e-06, "loss": 1.7064, "step": 29248 }, { "epoch": 2.2922413793103447, "grad_norm": 3.2832043170928955, "learning_rate": 3.875244728280676e-06, "loss": 1.495, "step": 29249 }, { "epoch": 2.292319749216301, "grad_norm": 3.166353464126587, "learning_rate": 3.874944271660792e-06, "loss": 1.4339, "step": 29250 }, { "epoch": 2.292398119122257, "grad_norm": 2.312217950820923, "learning_rate": 3.874643819320041e-06, "loss": 1.3834, "step": 29251 }, { "epoch": 2.292476489028213, "grad_norm": 2.5906496047973633, "learning_rate": 3.8743433712595675e-06, "loss": 1.9133, "step": 29252 }, { "epoch": 2.292554858934169, "grad_norm": 4.019618988037109, "learning_rate": 3.874042927480515e-06, "loss": 1.1028, "step": 29253 }, { "epoch": 2.2926332288401254, "grad_norm": 2.449284791946411, "learning_rate": 3.873742487984024e-06, "loss": 0.4775, "step": 29254 }, { "epoch": 2.2927115987460813, "grad_norm": 6.495911121368408, "learning_rate": 3.873442052771238e-06, "loss": 1.5414, "step": 29255 }, { "epoch": 2.2927899686520377, "grad_norm": 2.7674307823181152, "learning_rate": 3.873141621843301e-06, "loss": 1.3033, "step": 29256 }, { "epoch": 2.2928683385579935, "grad_norm": 2.9348878860473633, "learning_rate": 3.872841195201353e-06, "loss": 1.4712, "step": 29257 }, { "epoch": 2.29294670846395, "grad_norm": 2.720400094985962, "learning_rate": 3.872540772846537e-06, "loss": 1.4165, "step": 29258 }, { "epoch": 2.2930250783699058, "grad_norm": 2.481924057006836, "learning_rate": 3.872240354779998e-06, "loss": 0.934, "step": 29259 }, { "epoch": 2.293103448275862, "grad_norm": 3.6508076190948486, "learning_rate": 3.87193994100288e-06, "loss": 1.178, "step": 29260 }, { "epoch": 2.293181818181818, "grad_norm": 5.028609275817871, "learning_rate": 3.871639531516321e-06, "loss": 1.5004, "step": 29261 }, { "epoch": 2.2932601880877743, "grad_norm": 2.776496410369873, "learning_rate": 3.871339126321465e-06, "loss": 1.3295, "step": 29262 }, { "epoch": 2.2933385579937307, "grad_norm": 5.335636615753174, "learning_rate": 3.871038725419454e-06, "loss": 1.3319, "step": 29263 }, { "epoch": 2.2934169278996865, "grad_norm": 3.3006763458251953, "learning_rate": 3.870738328811436e-06, "loss": 0.757, "step": 29264 }, { "epoch": 2.2934952978056424, "grad_norm": 5.003976345062256, "learning_rate": 3.870437936498545e-06, "loss": 1.4579, "step": 29265 }, { "epoch": 2.2935736677115988, "grad_norm": 4.273538112640381, "learning_rate": 3.870137548481929e-06, "loss": 1.3872, "step": 29266 }, { "epoch": 2.293652037617555, "grad_norm": 3.4744012355804443, "learning_rate": 3.869837164762732e-06, "loss": 1.1394, "step": 29267 }, { "epoch": 2.293730407523511, "grad_norm": 2.3509247303009033, "learning_rate": 3.869536785342092e-06, "loss": 1.492, "step": 29268 }, { "epoch": 2.293808777429467, "grad_norm": 2.9855029582977295, "learning_rate": 3.869236410221153e-06, "loss": 1.0387, "step": 29269 }, { "epoch": 2.293887147335423, "grad_norm": 4.913828372955322, "learning_rate": 3.868936039401058e-06, "loss": 1.2643, "step": 29270 }, { "epoch": 2.2939655172413795, "grad_norm": 7.323367118835449, "learning_rate": 3.868635672882952e-06, "loss": 1.2479, "step": 29271 }, { "epoch": 2.2940438871473354, "grad_norm": 3.2101218700408936, "learning_rate": 3.868335310667971e-06, "loss": 1.0166, "step": 29272 }, { "epoch": 2.2941222570532913, "grad_norm": 2.3213422298431396, "learning_rate": 3.868034952757263e-06, "loss": 1.1567, "step": 29273 }, { "epoch": 2.2942006269592476, "grad_norm": 4.210059642791748, "learning_rate": 3.867734599151968e-06, "loss": 1.4452, "step": 29274 }, { "epoch": 2.294278996865204, "grad_norm": 3.5123555660247803, "learning_rate": 3.867434249853231e-06, "loss": 1.0407, "step": 29275 }, { "epoch": 2.29435736677116, "grad_norm": 3.1889700889587402, "learning_rate": 3.8671339048621906e-06, "loss": 1.3021, "step": 29276 }, { "epoch": 2.2944357366771158, "grad_norm": 2.926154136657715, "learning_rate": 3.86683356417999e-06, "loss": 1.2947, "step": 29277 }, { "epoch": 2.294514106583072, "grad_norm": 4.1219282150268555, "learning_rate": 3.8665332278077735e-06, "loss": 1.5304, "step": 29278 }, { "epoch": 2.2945924764890284, "grad_norm": 11.389474868774414, "learning_rate": 3.866232895746684e-06, "loss": 2.1761, "step": 29279 }, { "epoch": 2.2946708463949843, "grad_norm": 2.7147088050842285, "learning_rate": 3.8659325679978605e-06, "loss": 1.1802, "step": 29280 }, { "epoch": 2.29474921630094, "grad_norm": 3.0615859031677246, "learning_rate": 3.865632244562447e-06, "loss": 1.4427, "step": 29281 }, { "epoch": 2.2948275862068965, "grad_norm": 3.621983766555786, "learning_rate": 3.865331925441587e-06, "loss": 1.3459, "step": 29282 }, { "epoch": 2.294905956112853, "grad_norm": 5.0879740715026855, "learning_rate": 3.86503161063642e-06, "loss": 0.9528, "step": 29283 }, { "epoch": 2.2949843260188088, "grad_norm": 2.022213935852051, "learning_rate": 3.86473130014809e-06, "loss": 1.4317, "step": 29284 }, { "epoch": 2.295062695924765, "grad_norm": 3.1010329723358154, "learning_rate": 3.864430993977739e-06, "loss": 1.1196, "step": 29285 }, { "epoch": 2.295141065830721, "grad_norm": 3.674844264984131, "learning_rate": 3.864130692126511e-06, "loss": 1.4922, "step": 29286 }, { "epoch": 2.2952194357366773, "grad_norm": 11.833877563476562, "learning_rate": 3.863830394595546e-06, "loss": 0.5692, "step": 29287 }, { "epoch": 2.295297805642633, "grad_norm": 2.1840367317199707, "learning_rate": 3.863530101385985e-06, "loss": 1.0251, "step": 29288 }, { "epoch": 2.2953761755485895, "grad_norm": 2.3348844051361084, "learning_rate": 3.863229812498972e-06, "loss": 1.5219, "step": 29289 }, { "epoch": 2.2954545454545454, "grad_norm": 2.1547446250915527, "learning_rate": 3.8629295279356495e-06, "loss": 1.2416, "step": 29290 }, { "epoch": 2.2955329153605017, "grad_norm": 2.6888880729675293, "learning_rate": 3.86262924769716e-06, "loss": 0.9525, "step": 29291 }, { "epoch": 2.2956112852664576, "grad_norm": 7.800261497497559, "learning_rate": 3.862328971784643e-06, "loss": 1.1492, "step": 29292 }, { "epoch": 2.295689655172414, "grad_norm": 8.420689582824707, "learning_rate": 3.862028700199243e-06, "loss": 1.1265, "step": 29293 }, { "epoch": 2.29576802507837, "grad_norm": 4.442483901977539, "learning_rate": 3.861728432942103e-06, "loss": 1.4462, "step": 29294 }, { "epoch": 2.295846394984326, "grad_norm": 3.824413299560547, "learning_rate": 3.86142817001436e-06, "loss": 1.6564, "step": 29295 }, { "epoch": 2.295924764890282, "grad_norm": 3.343970775604248, "learning_rate": 3.861127911417161e-06, "loss": 1.2351, "step": 29296 }, { "epoch": 2.2960031347962384, "grad_norm": 8.502286911010742, "learning_rate": 3.860827657151647e-06, "loss": 1.3285, "step": 29297 }, { "epoch": 2.2960815047021943, "grad_norm": 2.3252859115600586, "learning_rate": 3.8605274072189594e-06, "loss": 1.5829, "step": 29298 }, { "epoch": 2.2961598746081506, "grad_norm": 2.595858097076416, "learning_rate": 3.8602271616202395e-06, "loss": 1.3641, "step": 29299 }, { "epoch": 2.2962382445141065, "grad_norm": 2.0642716884613037, "learning_rate": 3.859926920356629e-06, "loss": 1.4014, "step": 29300 }, { "epoch": 2.296316614420063, "grad_norm": 2.0355730056762695, "learning_rate": 3.859626683429274e-06, "loss": 1.628, "step": 29301 }, { "epoch": 2.2963949843260187, "grad_norm": 8.085163116455078, "learning_rate": 3.859326450839311e-06, "loss": 1.5264, "step": 29302 }, { "epoch": 2.296473354231975, "grad_norm": 3.3269338607788086, "learning_rate": 3.859026222587885e-06, "loss": 1.1399, "step": 29303 }, { "epoch": 2.296551724137931, "grad_norm": 4.229496479034424, "learning_rate": 3.858725998676137e-06, "loss": 1.3529, "step": 29304 }, { "epoch": 2.2966300940438873, "grad_norm": 5.086696624755859, "learning_rate": 3.858425779105211e-06, "loss": 1.3041, "step": 29305 }, { "epoch": 2.296708463949843, "grad_norm": 3.234905958175659, "learning_rate": 3.858125563876245e-06, "loss": 1.4351, "step": 29306 }, { "epoch": 2.2967868338557995, "grad_norm": 3.7574005126953125, "learning_rate": 3.857825352990382e-06, "loss": 1.6178, "step": 29307 }, { "epoch": 2.2968652037617554, "grad_norm": 8.473013877868652, "learning_rate": 3.857525146448765e-06, "loss": 1.6043, "step": 29308 }, { "epoch": 2.2969435736677117, "grad_norm": 3.2877252101898193, "learning_rate": 3.857224944252538e-06, "loss": 1.4201, "step": 29309 }, { "epoch": 2.2970219435736676, "grad_norm": 2.7548303604125977, "learning_rate": 3.856924746402838e-06, "loss": 1.4345, "step": 29310 }, { "epoch": 2.297100313479624, "grad_norm": 2.4692723751068115, "learning_rate": 3.85662455290081e-06, "loss": 1.1027, "step": 29311 }, { "epoch": 2.29717868338558, "grad_norm": 3.5885534286499023, "learning_rate": 3.856324363747596e-06, "loss": 1.7609, "step": 29312 }, { "epoch": 2.297257053291536, "grad_norm": 7.840715408325195, "learning_rate": 3.856024178944335e-06, "loss": 1.4317, "step": 29313 }, { "epoch": 2.297335423197492, "grad_norm": 3.40795636177063, "learning_rate": 3.85572399849217e-06, "loss": 1.0057, "step": 29314 }, { "epoch": 2.2974137931034484, "grad_norm": 2.210893154144287, "learning_rate": 3.855423822392244e-06, "loss": 1.3618, "step": 29315 }, { "epoch": 2.2974921630094043, "grad_norm": 5.51425838470459, "learning_rate": 3.855123650645699e-06, "loss": 0.8238, "step": 29316 }, { "epoch": 2.2975705329153606, "grad_norm": 2.150930166244507, "learning_rate": 3.854823483253675e-06, "loss": 1.0597, "step": 29317 }, { "epoch": 2.2976489028213165, "grad_norm": 3.387044668197632, "learning_rate": 3.854523320217313e-06, "loss": 1.2676, "step": 29318 }, { "epoch": 2.297727272727273, "grad_norm": 2.9073777198791504, "learning_rate": 3.854223161537756e-06, "loss": 1.734, "step": 29319 }, { "epoch": 2.2978056426332287, "grad_norm": 3.611985921859741, "learning_rate": 3.853923007216148e-06, "loss": 1.2582, "step": 29320 }, { "epoch": 2.297884012539185, "grad_norm": 7.791272163391113, "learning_rate": 3.8536228572536254e-06, "loss": 1.4696, "step": 29321 }, { "epoch": 2.297962382445141, "grad_norm": 6.160621166229248, "learning_rate": 3.8533227116513335e-06, "loss": 1.7725, "step": 29322 }, { "epoch": 2.2980407523510973, "grad_norm": 3.5295567512512207, "learning_rate": 3.853022570410414e-06, "loss": 1.8933, "step": 29323 }, { "epoch": 2.298119122257053, "grad_norm": 4.573531150817871, "learning_rate": 3.852722433532008e-06, "loss": 1.344, "step": 29324 }, { "epoch": 2.2981974921630095, "grad_norm": 3.6741092205047607, "learning_rate": 3.852422301017255e-06, "loss": 1.9331, "step": 29325 }, { "epoch": 2.2982758620689654, "grad_norm": 2.9585700035095215, "learning_rate": 3.852122172867298e-06, "loss": 0.8605, "step": 29326 }, { "epoch": 2.2983542319749217, "grad_norm": 17.435850143432617, "learning_rate": 3.851822049083281e-06, "loss": 1.5816, "step": 29327 }, { "epoch": 2.2984326018808776, "grad_norm": 2.3891518115997314, "learning_rate": 3.851521929666339e-06, "loss": 1.3492, "step": 29328 }, { "epoch": 2.298510971786834, "grad_norm": 2.874521017074585, "learning_rate": 3.851221814617621e-06, "loss": 1.3944, "step": 29329 }, { "epoch": 2.29858934169279, "grad_norm": 2.8840465545654297, "learning_rate": 3.850921703938264e-06, "loss": 0.6769, "step": 29330 }, { "epoch": 2.298667711598746, "grad_norm": 3.24896240234375, "learning_rate": 3.850621597629412e-06, "loss": 1.6008, "step": 29331 }, { "epoch": 2.298746081504702, "grad_norm": 7.739711761474609, "learning_rate": 3.850321495692204e-06, "loss": 0.9613, "step": 29332 }, { "epoch": 2.2988244514106584, "grad_norm": 4.349292755126953, "learning_rate": 3.850021398127781e-06, "loss": 1.1373, "step": 29333 }, { "epoch": 2.2989028213166143, "grad_norm": 3.20739483833313, "learning_rate": 3.849721304937288e-06, "loss": 1.2419, "step": 29334 }, { "epoch": 2.2989811912225706, "grad_norm": 1.790438175201416, "learning_rate": 3.849421216121865e-06, "loss": 1.4218, "step": 29335 }, { "epoch": 2.2990595611285265, "grad_norm": 2.112086534500122, "learning_rate": 3.849121131682651e-06, "loss": 1.2, "step": 29336 }, { "epoch": 2.299137931034483, "grad_norm": 2.585583448410034, "learning_rate": 3.848821051620789e-06, "loss": 1.5364, "step": 29337 }, { "epoch": 2.2992163009404387, "grad_norm": 2.479274034500122, "learning_rate": 3.848520975937422e-06, "loss": 1.3996, "step": 29338 }, { "epoch": 2.299294670846395, "grad_norm": 4.8163743019104, "learning_rate": 3.848220904633689e-06, "loss": 1.4868, "step": 29339 }, { "epoch": 2.299373040752351, "grad_norm": 3.0609328746795654, "learning_rate": 3.847920837710731e-06, "loss": 1.4816, "step": 29340 }, { "epoch": 2.2994514106583073, "grad_norm": 2.681601047515869, "learning_rate": 3.847620775169691e-06, "loss": 1.1949, "step": 29341 }, { "epoch": 2.299529780564263, "grad_norm": 7.122539043426514, "learning_rate": 3.847320717011711e-06, "loss": 1.4981, "step": 29342 }, { "epoch": 2.2996081504702195, "grad_norm": 2.580042839050293, "learning_rate": 3.84702066323793e-06, "loss": 1.2051, "step": 29343 }, { "epoch": 2.2996865203761754, "grad_norm": 2.8073041439056396, "learning_rate": 3.846720613849489e-06, "loss": 0.9907, "step": 29344 }, { "epoch": 2.2997648902821317, "grad_norm": 2.90548038482666, "learning_rate": 3.846420568847531e-06, "loss": 1.5299, "step": 29345 }, { "epoch": 2.2998432601880876, "grad_norm": 2.061889410018921, "learning_rate": 3.8461205282331994e-06, "loss": 1.0169, "step": 29346 }, { "epoch": 2.299921630094044, "grad_norm": 4.068146228790283, "learning_rate": 3.845820492007629e-06, "loss": 1.9749, "step": 29347 }, { "epoch": 2.3, "grad_norm": 3.028918504714966, "learning_rate": 3.845520460171966e-06, "loss": 1.4342, "step": 29348 }, { "epoch": 2.300078369905956, "grad_norm": 2.7562055587768555, "learning_rate": 3.84522043272735e-06, "loss": 1.3132, "step": 29349 }, { "epoch": 2.300156739811912, "grad_norm": 2.922369956970215, "learning_rate": 3.8449204096749235e-06, "loss": 1.7587, "step": 29350 }, { "epoch": 2.3002351097178684, "grad_norm": 2.3542909622192383, "learning_rate": 3.8446203910158255e-06, "loss": 1.4185, "step": 29351 }, { "epoch": 2.3003134796238243, "grad_norm": 2.6661670207977295, "learning_rate": 3.844320376751197e-06, "loss": 1.512, "step": 29352 }, { "epoch": 2.3003918495297806, "grad_norm": 3.0341975688934326, "learning_rate": 3.8440203668821835e-06, "loss": 1.7332, "step": 29353 }, { "epoch": 2.3004702194357365, "grad_norm": 7.736031532287598, "learning_rate": 3.84372036140992e-06, "loss": 1.1481, "step": 29354 }, { "epoch": 2.300548589341693, "grad_norm": 2.471935272216797, "learning_rate": 3.84342036033555e-06, "loss": 1.1038, "step": 29355 }, { "epoch": 2.3006269592476487, "grad_norm": 2.245328187942505, "learning_rate": 3.843120363660216e-06, "loss": 1.2708, "step": 29356 }, { "epoch": 2.300705329153605, "grad_norm": 3.082956075668335, "learning_rate": 3.842820371385059e-06, "loss": 1.5592, "step": 29357 }, { "epoch": 2.3007836990595614, "grad_norm": 3.8758246898651123, "learning_rate": 3.842520383511217e-06, "loss": 1.1457, "step": 29358 }, { "epoch": 2.3008620689655173, "grad_norm": 4.65406608581543, "learning_rate": 3.842220400039833e-06, "loss": 1.4827, "step": 29359 }, { "epoch": 2.300940438871473, "grad_norm": 4.148502349853516, "learning_rate": 3.841920420972049e-06, "loss": 1.5927, "step": 29360 }, { "epoch": 2.3010188087774295, "grad_norm": 3.683255910873413, "learning_rate": 3.841620446309005e-06, "loss": 1.3246, "step": 29361 }, { "epoch": 2.301097178683386, "grad_norm": 2.9803502559661865, "learning_rate": 3.841320476051841e-06, "loss": 1.5332, "step": 29362 }, { "epoch": 2.3011755485893417, "grad_norm": 3.834998846054077, "learning_rate": 3.841020510201698e-06, "loss": 0.9577, "step": 29363 }, { "epoch": 2.3012539184952976, "grad_norm": 7.921751499176025, "learning_rate": 3.840720548759721e-06, "loss": 1.3822, "step": 29364 }, { "epoch": 2.301332288401254, "grad_norm": 3.3518848419189453, "learning_rate": 3.840420591727045e-06, "loss": 1.2869, "step": 29365 }, { "epoch": 2.3014106583072103, "grad_norm": 3.27831768989563, "learning_rate": 3.840120639104813e-06, "loss": 0.8556, "step": 29366 }, { "epoch": 2.301489028213166, "grad_norm": 3.638483762741089, "learning_rate": 3.8398206908941665e-06, "loss": 1.3876, "step": 29367 }, { "epoch": 2.301567398119122, "grad_norm": 1.9592398405075073, "learning_rate": 3.839520747096248e-06, "loss": 1.1496, "step": 29368 }, { "epoch": 2.3016457680250784, "grad_norm": 3.0007123947143555, "learning_rate": 3.839220807712195e-06, "loss": 1.3673, "step": 29369 }, { "epoch": 2.3017241379310347, "grad_norm": 5.046305179595947, "learning_rate": 3.83892087274315e-06, "loss": 1.6778, "step": 29370 }, { "epoch": 2.3018025078369906, "grad_norm": 7.1055216789245605, "learning_rate": 3.838620942190253e-06, "loss": 0.7265, "step": 29371 }, { "epoch": 2.3018808777429465, "grad_norm": 3.111678123474121, "learning_rate": 3.838321016054648e-06, "loss": 1.4383, "step": 29372 }, { "epoch": 2.301959247648903, "grad_norm": 3.0678787231445312, "learning_rate": 3.83802109433747e-06, "loss": 1.2736, "step": 29373 }, { "epoch": 2.302037617554859, "grad_norm": 4.163488388061523, "learning_rate": 3.8377211770398645e-06, "loss": 1.7915, "step": 29374 }, { "epoch": 2.302115987460815, "grad_norm": 3.8322181701660156, "learning_rate": 3.837421264162971e-06, "loss": 1.3988, "step": 29375 }, { "epoch": 2.302194357366771, "grad_norm": 3.12886118888855, "learning_rate": 3.83712135570793e-06, "loss": 1.5402, "step": 29376 }, { "epoch": 2.3022727272727272, "grad_norm": 2.306952714920044, "learning_rate": 3.836821451675881e-06, "loss": 1.0609, "step": 29377 }, { "epoch": 2.3023510971786836, "grad_norm": 3.2899506092071533, "learning_rate": 3.836521552067965e-06, "loss": 1.2929, "step": 29378 }, { "epoch": 2.3024294670846395, "grad_norm": 3.4220542907714844, "learning_rate": 3.836221656885325e-06, "loss": 1.9099, "step": 29379 }, { "epoch": 2.302507836990596, "grad_norm": 6.254464149475098, "learning_rate": 3.8359217661291e-06, "loss": 1.3165, "step": 29380 }, { "epoch": 2.3025862068965517, "grad_norm": 3.4134018421173096, "learning_rate": 3.8356218798004305e-06, "loss": 1.0856, "step": 29381 }, { "epoch": 2.302664576802508, "grad_norm": 4.832852840423584, "learning_rate": 3.8353219979004574e-06, "loss": 0.4777, "step": 29382 }, { "epoch": 2.302742946708464, "grad_norm": 3.1439638137817383, "learning_rate": 3.835022120430322e-06, "loss": 1.3414, "step": 29383 }, { "epoch": 2.3028213166144202, "grad_norm": 5.659517765045166, "learning_rate": 3.834722247391163e-06, "loss": 1.4418, "step": 29384 }, { "epoch": 2.302899686520376, "grad_norm": 4.7555952072143555, "learning_rate": 3.834422378784121e-06, "loss": 0.9244, "step": 29385 }, { "epoch": 2.3029780564263325, "grad_norm": 9.888324737548828, "learning_rate": 3.8341225146103385e-06, "loss": 1.0743, "step": 29386 }, { "epoch": 2.3030564263322884, "grad_norm": 3.433215379714966, "learning_rate": 3.833822654870956e-06, "loss": 1.4218, "step": 29387 }, { "epoch": 2.3031347962382447, "grad_norm": 3.0799026489257812, "learning_rate": 3.833522799567112e-06, "loss": 0.9292, "step": 29388 }, { "epoch": 2.3032131661442006, "grad_norm": 4.53568696975708, "learning_rate": 3.833222948699949e-06, "loss": 1.3706, "step": 29389 }, { "epoch": 2.303291536050157, "grad_norm": 2.409996509552002, "learning_rate": 3.8329231022706045e-06, "loss": 1.0548, "step": 29390 }, { "epoch": 2.303369905956113, "grad_norm": 2.7164459228515625, "learning_rate": 3.8326232602802235e-06, "loss": 1.0525, "step": 29391 }, { "epoch": 2.303448275862069, "grad_norm": 2.267789840698242, "learning_rate": 3.8323234227299435e-06, "loss": 1.0688, "step": 29392 }, { "epoch": 2.303526645768025, "grad_norm": 3.7018401622772217, "learning_rate": 3.8320235896209055e-06, "loss": 1.622, "step": 29393 }, { "epoch": 2.3036050156739813, "grad_norm": 2.413656234741211, "learning_rate": 3.83172376095425e-06, "loss": 1.382, "step": 29394 }, { "epoch": 2.3036833855799372, "grad_norm": 3.410850763320923, "learning_rate": 3.831423936731117e-06, "loss": 0.8456, "step": 29395 }, { "epoch": 2.3037617554858936, "grad_norm": 3.3430674076080322, "learning_rate": 3.8311241169526455e-06, "loss": 1.4889, "step": 29396 }, { "epoch": 2.3038401253918495, "grad_norm": 2.808976888656616, "learning_rate": 3.83082430161998e-06, "loss": 1.3687, "step": 29397 }, { "epoch": 2.303918495297806, "grad_norm": 2.963620901107788, "learning_rate": 3.830524490734258e-06, "loss": 1.4526, "step": 29398 }, { "epoch": 2.3039968652037617, "grad_norm": 3.015202283859253, "learning_rate": 3.830224684296619e-06, "loss": 1.1781, "step": 29399 }, { "epoch": 2.304075235109718, "grad_norm": 2.2598938941955566, "learning_rate": 3.829924882308205e-06, "loss": 1.1336, "step": 29400 }, { "epoch": 2.304153605015674, "grad_norm": 4.7111430168151855, "learning_rate": 3.829625084770156e-06, "loss": 1.3558, "step": 29401 }, { "epoch": 2.3042319749216302, "grad_norm": 4.754241466522217, "learning_rate": 3.829325291683612e-06, "loss": 1.5364, "step": 29402 }, { "epoch": 2.304310344827586, "grad_norm": 3.88271427154541, "learning_rate": 3.8290255030497115e-06, "loss": 0.8777, "step": 29403 }, { "epoch": 2.3043887147335425, "grad_norm": 7.2981462478637695, "learning_rate": 3.828725718869598e-06, "loss": 0.9498, "step": 29404 }, { "epoch": 2.3044670846394983, "grad_norm": 8.02048110961914, "learning_rate": 3.8284259391444105e-06, "loss": 1.1569, "step": 29405 }, { "epoch": 2.3045454545454547, "grad_norm": 5.062358856201172, "learning_rate": 3.82812616387529e-06, "loss": 1.3594, "step": 29406 }, { "epoch": 2.3046238244514106, "grad_norm": 6.944533824920654, "learning_rate": 3.8278263930633745e-06, "loss": 1.5806, "step": 29407 }, { "epoch": 2.304702194357367, "grad_norm": 5.296837329864502, "learning_rate": 3.8275266267098035e-06, "loss": 1.3861, "step": 29408 }, { "epoch": 2.304780564263323, "grad_norm": 3.7073795795440674, "learning_rate": 3.827226864815723e-06, "loss": 1.1077, "step": 29409 }, { "epoch": 2.304858934169279, "grad_norm": 2.729398250579834, "learning_rate": 3.826927107382266e-06, "loss": 1.4964, "step": 29410 }, { "epoch": 2.304937304075235, "grad_norm": 8.790094375610352, "learning_rate": 3.826627354410576e-06, "loss": 0.7864, "step": 29411 }, { "epoch": 2.3050156739811913, "grad_norm": 2.1973321437835693, "learning_rate": 3.826327605901794e-06, "loss": 1.0987, "step": 29412 }, { "epoch": 2.3050940438871472, "grad_norm": 5.498830318450928, "learning_rate": 3.826027861857059e-06, "loss": 1.7638, "step": 29413 }, { "epoch": 2.3051724137931036, "grad_norm": 2.4744973182678223, "learning_rate": 3.82572812227751e-06, "loss": 1.2337, "step": 29414 }, { "epoch": 2.3052507836990594, "grad_norm": 4.027894973754883, "learning_rate": 3.8254283871642875e-06, "loss": 1.4156, "step": 29415 }, { "epoch": 2.305329153605016, "grad_norm": 2.8864455223083496, "learning_rate": 3.825128656518533e-06, "loss": 1.636, "step": 29416 }, { "epoch": 2.3054075235109717, "grad_norm": 6.352977275848389, "learning_rate": 3.824828930341386e-06, "loss": 1.3982, "step": 29417 }, { "epoch": 2.305485893416928, "grad_norm": 2.695586681365967, "learning_rate": 3.824529208633986e-06, "loss": 1.4812, "step": 29418 }, { "epoch": 2.305564263322884, "grad_norm": 2.3307831287384033, "learning_rate": 3.824229491397474e-06, "loss": 1.4439, "step": 29419 }, { "epoch": 2.30564263322884, "grad_norm": 1.9534759521484375, "learning_rate": 3.823929778632989e-06, "loss": 1.4702, "step": 29420 }, { "epoch": 2.305721003134796, "grad_norm": 4.15574312210083, "learning_rate": 3.8236300703416695e-06, "loss": 1.2196, "step": 29421 }, { "epoch": 2.3057993730407524, "grad_norm": 2.7043943405151367, "learning_rate": 3.823330366524657e-06, "loss": 1.386, "step": 29422 }, { "epoch": 2.3058777429467083, "grad_norm": 5.109714031219482, "learning_rate": 3.823030667183092e-06, "loss": 0.8682, "step": 29423 }, { "epoch": 2.3059561128526647, "grad_norm": 3.6240901947021484, "learning_rate": 3.822730972318115e-06, "loss": 1.7472, "step": 29424 }, { "epoch": 2.3060344827586206, "grad_norm": 6.122503280639648, "learning_rate": 3.822431281930863e-06, "loss": 1.2551, "step": 29425 }, { "epoch": 2.306112852664577, "grad_norm": 2.7532124519348145, "learning_rate": 3.822131596022478e-06, "loss": 1.3914, "step": 29426 }, { "epoch": 2.3061912225705328, "grad_norm": 3.429187536239624, "learning_rate": 3.821831914594099e-06, "loss": 1.4038, "step": 29427 }, { "epoch": 2.306269592476489, "grad_norm": 2.9298417568206787, "learning_rate": 3.821532237646869e-06, "loss": 1.6767, "step": 29428 }, { "epoch": 2.306347962382445, "grad_norm": 5.069640636444092, "learning_rate": 3.821232565181922e-06, "loss": 0.6852, "step": 29429 }, { "epoch": 2.3064263322884013, "grad_norm": 2.783139944076538, "learning_rate": 3.8209328972004016e-06, "loss": 1.0097, "step": 29430 }, { "epoch": 2.306504702194357, "grad_norm": 3.257439613342285, "learning_rate": 3.8206332337034465e-06, "loss": 1.4407, "step": 29431 }, { "epoch": 2.3065830721003135, "grad_norm": 5.110963821411133, "learning_rate": 3.820333574692198e-06, "loss": 1.263, "step": 29432 }, { "epoch": 2.3066614420062694, "grad_norm": 3.6490328311920166, "learning_rate": 3.820033920167795e-06, "loss": 1.4081, "step": 29433 }, { "epoch": 2.3067398119122258, "grad_norm": 2.470219850540161, "learning_rate": 3.819734270131374e-06, "loss": 1.1483, "step": 29434 }, { "epoch": 2.3068181818181817, "grad_norm": 5.268052101135254, "learning_rate": 3.819434624584082e-06, "loss": 1.3385, "step": 29435 }, { "epoch": 2.306896551724138, "grad_norm": 2.4967868328094482, "learning_rate": 3.81913498352705e-06, "loss": 1.668, "step": 29436 }, { "epoch": 2.306974921630094, "grad_norm": 3.498102903366089, "learning_rate": 3.818835346961425e-06, "loss": 1.7126, "step": 29437 }, { "epoch": 2.30705329153605, "grad_norm": 4.063322067260742, "learning_rate": 3.818535714888342e-06, "loss": 1.5049, "step": 29438 }, { "epoch": 2.307131661442006, "grad_norm": 2.494617462158203, "learning_rate": 3.818236087308943e-06, "loss": 1.1856, "step": 29439 }, { "epoch": 2.3072100313479624, "grad_norm": 3.2837090492248535, "learning_rate": 3.817936464224367e-06, "loss": 1.433, "step": 29440 }, { "epoch": 2.3072884012539183, "grad_norm": 2.488525629043579, "learning_rate": 3.817636845635752e-06, "loss": 1.22, "step": 29441 }, { "epoch": 2.3073667711598747, "grad_norm": 3.1847856044769287, "learning_rate": 3.817337231544241e-06, "loss": 1.3811, "step": 29442 }, { "epoch": 2.3074451410658305, "grad_norm": 3.648458480834961, "learning_rate": 3.8170376219509716e-06, "loss": 1.0713, "step": 29443 }, { "epoch": 2.307523510971787, "grad_norm": 2.7013444900512695, "learning_rate": 3.816738016857083e-06, "loss": 1.3556, "step": 29444 }, { "epoch": 2.3076018808777428, "grad_norm": 7.101832866668701, "learning_rate": 3.816438416263714e-06, "loss": 1.3178, "step": 29445 }, { "epoch": 2.307680250783699, "grad_norm": 5.104151248931885, "learning_rate": 3.8161388201720065e-06, "loss": 1.8085, "step": 29446 }, { "epoch": 2.307758620689655, "grad_norm": 2.484295129776001, "learning_rate": 3.815839228583101e-06, "loss": 1.2215, "step": 29447 }, { "epoch": 2.3078369905956113, "grad_norm": 3.304934024810791, "learning_rate": 3.815539641498132e-06, "loss": 1.2891, "step": 29448 }, { "epoch": 2.3079153605015676, "grad_norm": 3.1587445735931396, "learning_rate": 3.815240058918242e-06, "loss": 0.888, "step": 29449 }, { "epoch": 2.3079937304075235, "grad_norm": 3.0760364532470703, "learning_rate": 3.814940480844572e-06, "loss": 1.3329, "step": 29450 }, { "epoch": 2.3080721003134794, "grad_norm": 4.173757553100586, "learning_rate": 3.8146409072782585e-06, "loss": 1.2451, "step": 29451 }, { "epoch": 2.3081504702194358, "grad_norm": 4.880183219909668, "learning_rate": 3.814341338220442e-06, "loss": 1.354, "step": 29452 }, { "epoch": 2.308228840125392, "grad_norm": 5.707237720489502, "learning_rate": 3.8140417736722613e-06, "loss": 1.0389, "step": 29453 }, { "epoch": 2.308307210031348, "grad_norm": 4.208095073699951, "learning_rate": 3.8137422136348602e-06, "loss": 1.4552, "step": 29454 }, { "epoch": 2.308385579937304, "grad_norm": 3.8538358211517334, "learning_rate": 3.8134426581093703e-06, "loss": 1.3154, "step": 29455 }, { "epoch": 2.30846394984326, "grad_norm": 2.6128129959106445, "learning_rate": 3.813143107096936e-06, "loss": 1.3928, "step": 29456 }, { "epoch": 2.3085423197492165, "grad_norm": 2.8154022693634033, "learning_rate": 3.812843560598697e-06, "loss": 0.963, "step": 29457 }, { "epoch": 2.3086206896551724, "grad_norm": 2.359236478805542, "learning_rate": 3.8125440186157914e-06, "loss": 1.3091, "step": 29458 }, { "epoch": 2.3086990595611283, "grad_norm": 2.936366558074951, "learning_rate": 3.8122444811493573e-06, "loss": 1.6755, "step": 29459 }, { "epoch": 2.3087774294670846, "grad_norm": 3.4347290992736816, "learning_rate": 3.8119449482005345e-06, "loss": 1.6319, "step": 29460 }, { "epoch": 2.308855799373041, "grad_norm": 3.6946828365325928, "learning_rate": 3.8116454197704635e-06, "loss": 1.4323, "step": 29461 }, { "epoch": 2.308934169278997, "grad_norm": 2.5798380374908447, "learning_rate": 3.8113458958602843e-06, "loss": 1.1002, "step": 29462 }, { "epoch": 2.3090125391849528, "grad_norm": 3.4880430698394775, "learning_rate": 3.811046376471134e-06, "loss": 1.5125, "step": 29463 }, { "epoch": 2.309090909090909, "grad_norm": 20.132442474365234, "learning_rate": 3.810746861604152e-06, "loss": 2.4903, "step": 29464 }, { "epoch": 2.3091692789968654, "grad_norm": 6.096209526062012, "learning_rate": 3.8104473512604795e-06, "loss": 1.8569, "step": 29465 }, { "epoch": 2.3092476489028213, "grad_norm": 3.7069203853607178, "learning_rate": 3.8101478454412533e-06, "loss": 1.5529, "step": 29466 }, { "epoch": 2.309326018808777, "grad_norm": 2.175300359725952, "learning_rate": 3.8098483441476124e-06, "loss": 1.3281, "step": 29467 }, { "epoch": 2.3094043887147335, "grad_norm": 2.9230310916900635, "learning_rate": 3.8095488473806985e-06, "loss": 1.5891, "step": 29468 }, { "epoch": 2.30948275862069, "grad_norm": 6.423851490020752, "learning_rate": 3.80924935514165e-06, "loss": 1.5504, "step": 29469 }, { "epoch": 2.3095611285266457, "grad_norm": 6.0325188636779785, "learning_rate": 3.8089498674316038e-06, "loss": 0.9066, "step": 29470 }, { "epoch": 2.3096394984326016, "grad_norm": 7.110485553741455, "learning_rate": 3.8086503842517014e-06, "loss": 1.5697, "step": 29471 }, { "epoch": 2.309717868338558, "grad_norm": 2.748908042907715, "learning_rate": 3.8083509056030798e-06, "loss": 1.0054, "step": 29472 }, { "epoch": 2.3097962382445143, "grad_norm": 2.8421058654785156, "learning_rate": 3.8080514314868823e-06, "loss": 1.1923, "step": 29473 }, { "epoch": 2.30987460815047, "grad_norm": 2.5649476051330566, "learning_rate": 3.8077519619042415e-06, "loss": 1.4865, "step": 29474 }, { "epoch": 2.3099529780564265, "grad_norm": 5.366912364959717, "learning_rate": 3.8074524968563008e-06, "loss": 1.1064, "step": 29475 }, { "epoch": 2.3100313479623824, "grad_norm": 3.00357723236084, "learning_rate": 3.8071530363442e-06, "loss": 1.3385, "step": 29476 }, { "epoch": 2.3101097178683387, "grad_norm": 2.5526039600372314, "learning_rate": 3.806853580369075e-06, "loss": 1.2557, "step": 29477 }, { "epoch": 2.3101880877742946, "grad_norm": 2.206343650817871, "learning_rate": 3.8065541289320663e-06, "loss": 1.243, "step": 29478 }, { "epoch": 2.310266457680251, "grad_norm": 2.8678534030914307, "learning_rate": 3.8062546820343116e-06, "loss": 1.239, "step": 29479 }, { "epoch": 2.310344827586207, "grad_norm": 2.697207450866699, "learning_rate": 3.805955239676954e-06, "loss": 1.4509, "step": 29480 }, { "epoch": 2.310423197492163, "grad_norm": 4.401321887969971, "learning_rate": 3.805655801861126e-06, "loss": 1.3909, "step": 29481 }, { "epoch": 2.310501567398119, "grad_norm": 3.492783308029175, "learning_rate": 3.8053563685879715e-06, "loss": 0.9761, "step": 29482 }, { "epoch": 2.3105799373040754, "grad_norm": 2.816544771194458, "learning_rate": 3.8050569398586267e-06, "loss": 1.2457, "step": 29483 }, { "epoch": 2.3106583072100313, "grad_norm": 2.9925782680511475, "learning_rate": 3.8047575156742333e-06, "loss": 1.5881, "step": 29484 }, { "epoch": 2.3107366771159876, "grad_norm": 3.754178047180176, "learning_rate": 3.804458096035928e-06, "loss": 1.3288, "step": 29485 }, { "epoch": 2.3108150470219435, "grad_norm": 5.0787739753723145, "learning_rate": 3.804158680944848e-06, "loss": 1.2875, "step": 29486 }, { "epoch": 2.3108934169279, "grad_norm": 2.500282049179077, "learning_rate": 3.8038592704021356e-06, "loss": 1.8233, "step": 29487 }, { "epoch": 2.3109717868338557, "grad_norm": 4.12843132019043, "learning_rate": 3.803559864408929e-06, "loss": 0.1377, "step": 29488 }, { "epoch": 2.311050156739812, "grad_norm": 3.7156217098236084, "learning_rate": 3.803260462966365e-06, "loss": 1.2735, "step": 29489 }, { "epoch": 2.311128526645768, "grad_norm": 3.909802198410034, "learning_rate": 3.802961066075584e-06, "loss": 1.2107, "step": 29490 }, { "epoch": 2.3112068965517243, "grad_norm": 4.836048126220703, "learning_rate": 3.802661673737725e-06, "loss": 1.1046, "step": 29491 }, { "epoch": 2.31128526645768, "grad_norm": 3.040330648422241, "learning_rate": 3.802362285953924e-06, "loss": 0.9121, "step": 29492 }, { "epoch": 2.3113636363636365, "grad_norm": 2.4860496520996094, "learning_rate": 3.8020629027253227e-06, "loss": 1.1502, "step": 29493 }, { "epoch": 2.3114420062695924, "grad_norm": 2.5181527137756348, "learning_rate": 3.8017635240530586e-06, "loss": 1.4112, "step": 29494 }, { "epoch": 2.3115203761755487, "grad_norm": 2.782243251800537, "learning_rate": 3.801464149938272e-06, "loss": 1.3861, "step": 29495 }, { "epoch": 2.3115987460815046, "grad_norm": 2.6406173706054688, "learning_rate": 3.801164780382099e-06, "loss": 1.2738, "step": 29496 }, { "epoch": 2.311677115987461, "grad_norm": 3.690889835357666, "learning_rate": 3.800865415385678e-06, "loss": 1.1506, "step": 29497 }, { "epoch": 2.311755485893417, "grad_norm": 4.1158294677734375, "learning_rate": 3.800566054950151e-06, "loss": 1.533, "step": 29498 }, { "epoch": 2.311833855799373, "grad_norm": 7.673532962799072, "learning_rate": 3.800266699076655e-06, "loss": 1.0786, "step": 29499 }, { "epoch": 2.311912225705329, "grad_norm": 2.767693519592285, "learning_rate": 3.7999673477663275e-06, "loss": 1.0686, "step": 29500 }, { "epoch": 2.3119905956112854, "grad_norm": 3.8393514156341553, "learning_rate": 3.7996680010203073e-06, "loss": 1.5009, "step": 29501 }, { "epoch": 2.3120689655172413, "grad_norm": 2.509287118911743, "learning_rate": 3.7993686588397345e-06, "loss": 1.4471, "step": 29502 }, { "epoch": 2.3121473354231976, "grad_norm": 6.955305099487305, "learning_rate": 3.7990693212257467e-06, "loss": 0.957, "step": 29503 }, { "epoch": 2.3122257053291535, "grad_norm": 4.815376281738281, "learning_rate": 3.798769988179481e-06, "loss": 1.2156, "step": 29504 }, { "epoch": 2.31230407523511, "grad_norm": 3.246711492538452, "learning_rate": 3.7984706597020778e-06, "loss": 1.1437, "step": 29505 }, { "epoch": 2.3123824451410657, "grad_norm": 3.5876052379608154, "learning_rate": 3.7981713357946764e-06, "loss": 1.2195, "step": 29506 }, { "epoch": 2.312460815047022, "grad_norm": 2.382840394973755, "learning_rate": 3.7978720164584128e-06, "loss": 1.0534, "step": 29507 }, { "epoch": 2.312539184952978, "grad_norm": 3.5947868824005127, "learning_rate": 3.7975727016944262e-06, "loss": 1.6973, "step": 29508 }, { "epoch": 2.3126175548589343, "grad_norm": 4.272880554199219, "learning_rate": 3.797273391503855e-06, "loss": 1.3281, "step": 29509 }, { "epoch": 2.31269592476489, "grad_norm": 5.490392684936523, "learning_rate": 3.796974085887841e-06, "loss": 1.1892, "step": 29510 }, { "epoch": 2.3127742946708465, "grad_norm": 4.049859523773193, "learning_rate": 3.7966747848475166e-06, "loss": 1.0847, "step": 29511 }, { "epoch": 2.3128526645768024, "grad_norm": 2.3494515419006348, "learning_rate": 3.796375488384024e-06, "loss": 1.54, "step": 29512 }, { "epoch": 2.3129310344827587, "grad_norm": 2.7772841453552246, "learning_rate": 3.796076196498501e-06, "loss": 1.1611, "step": 29513 }, { "epoch": 2.3130094043887146, "grad_norm": 1.928497314453125, "learning_rate": 3.795776909192087e-06, "loss": 1.193, "step": 29514 }, { "epoch": 2.313087774294671, "grad_norm": 3.129877805709839, "learning_rate": 3.7954776264659175e-06, "loss": 0.9083, "step": 29515 }, { "epoch": 2.313166144200627, "grad_norm": 3.8092427253723145, "learning_rate": 3.7951783483211325e-06, "loss": 1.3589, "step": 29516 }, { "epoch": 2.313244514106583, "grad_norm": 3.9126698970794678, "learning_rate": 3.7948790747588704e-06, "loss": 0.9009, "step": 29517 }, { "epoch": 2.313322884012539, "grad_norm": 2.67427134513855, "learning_rate": 3.79457980578027e-06, "loss": 1.1629, "step": 29518 }, { "epoch": 2.3134012539184954, "grad_norm": 2.4348082542419434, "learning_rate": 3.794280541386469e-06, "loss": 1.2578, "step": 29519 }, { "epoch": 2.3134796238244513, "grad_norm": 1.9188120365142822, "learning_rate": 3.7939812815786046e-06, "loss": 1.4151, "step": 29520 }, { "epoch": 2.3135579937304076, "grad_norm": 2.555246353149414, "learning_rate": 3.7936820263578172e-06, "loss": 0.9898, "step": 29521 }, { "epoch": 2.3136363636363635, "grad_norm": 2.7017788887023926, "learning_rate": 3.793382775725243e-06, "loss": 1.0096, "step": 29522 }, { "epoch": 2.31371473354232, "grad_norm": 2.4316279888153076, "learning_rate": 3.7930835296820195e-06, "loss": 1.3926, "step": 29523 }, { "epoch": 2.3137931034482757, "grad_norm": 2.701281785964966, "learning_rate": 3.7927842882292877e-06, "loss": 1.0616, "step": 29524 }, { "epoch": 2.313871473354232, "grad_norm": 4.2017669677734375, "learning_rate": 3.7924850513681855e-06, "loss": 1.0634, "step": 29525 }, { "epoch": 2.313949843260188, "grad_norm": 3.7000725269317627, "learning_rate": 3.7921858190998485e-06, "loss": 0.971, "step": 29526 }, { "epoch": 2.3140282131661443, "grad_norm": 3.8488781452178955, "learning_rate": 3.7918865914254166e-06, "loss": 0.4696, "step": 29527 }, { "epoch": 2.3141065830721, "grad_norm": 6.046900272369385, "learning_rate": 3.7915873683460263e-06, "loss": 1.8065, "step": 29528 }, { "epoch": 2.3141849529780565, "grad_norm": 3.334913492202759, "learning_rate": 3.7912881498628205e-06, "loss": 1.2648, "step": 29529 }, { "epoch": 2.3142633228840124, "grad_norm": 6.835302352905273, "learning_rate": 3.79098893597693e-06, "loss": 1.4643, "step": 29530 }, { "epoch": 2.3143416927899687, "grad_norm": 3.4973957538604736, "learning_rate": 3.790689726689498e-06, "loss": 1.2703, "step": 29531 }, { "epoch": 2.3144200626959246, "grad_norm": 5.753052711486816, "learning_rate": 3.790390522001662e-06, "loss": 1.6271, "step": 29532 }, { "epoch": 2.314498432601881, "grad_norm": 4.389047622680664, "learning_rate": 3.7900913219145585e-06, "loss": 1.7869, "step": 29533 }, { "epoch": 2.314576802507837, "grad_norm": 2.9146697521209717, "learning_rate": 3.7897921264293264e-06, "loss": 0.9416, "step": 29534 }, { "epoch": 2.314655172413793, "grad_norm": 2.932025671005249, "learning_rate": 3.789492935547102e-06, "loss": 1.5591, "step": 29535 }, { "epoch": 2.314733542319749, "grad_norm": 7.306188583374023, "learning_rate": 3.789193749269028e-06, "loss": 1.7312, "step": 29536 }, { "epoch": 2.3148119122257054, "grad_norm": 3.4293017387390137, "learning_rate": 3.788894567596235e-06, "loss": 1.3905, "step": 29537 }, { "epoch": 2.3148902821316613, "grad_norm": 5.0222039222717285, "learning_rate": 3.7885953905298674e-06, "loss": 0.9004, "step": 29538 }, { "epoch": 2.3149686520376176, "grad_norm": 2.8864614963531494, "learning_rate": 3.7882962180710604e-06, "loss": 1.7726, "step": 29539 }, { "epoch": 2.3150470219435735, "grad_norm": 2.312059164047241, "learning_rate": 3.787997050220953e-06, "loss": 1.4318, "step": 29540 }, { "epoch": 2.31512539184953, "grad_norm": 2.7610671520233154, "learning_rate": 3.7876978869806807e-06, "loss": 1.53, "step": 29541 }, { "epoch": 2.3152037617554857, "grad_norm": 3.7205538749694824, "learning_rate": 3.7873987283513823e-06, "loss": 0.9842, "step": 29542 }, { "epoch": 2.315282131661442, "grad_norm": 2.7563459873199463, "learning_rate": 3.7870995743341978e-06, "loss": 1.3756, "step": 29543 }, { "epoch": 2.3153605015673984, "grad_norm": 3.595745086669922, "learning_rate": 3.7868004249302643e-06, "loss": 1.4751, "step": 29544 }, { "epoch": 2.3154388714733543, "grad_norm": 2.558279037475586, "learning_rate": 3.7865012801407177e-06, "loss": 1.5448, "step": 29545 }, { "epoch": 2.31551724137931, "grad_norm": 3.774955987930298, "learning_rate": 3.786202139966697e-06, "loss": 1.2904, "step": 29546 }, { "epoch": 2.3155956112852665, "grad_norm": 4.3385820388793945, "learning_rate": 3.7859030044093413e-06, "loss": 1.2197, "step": 29547 }, { "epoch": 2.315673981191223, "grad_norm": 3.282834529876709, "learning_rate": 3.7856038734697854e-06, "loss": 1.488, "step": 29548 }, { "epoch": 2.3157523510971787, "grad_norm": 3.538919448852539, "learning_rate": 3.7853047471491675e-06, "loss": 1.1677, "step": 29549 }, { "epoch": 2.3158307210031346, "grad_norm": 3.1168978214263916, "learning_rate": 3.7850056254486278e-06, "loss": 1.3987, "step": 29550 }, { "epoch": 2.315909090909091, "grad_norm": 2.7248871326446533, "learning_rate": 3.784706508369304e-06, "loss": 1.2432, "step": 29551 }, { "epoch": 2.3159874608150472, "grad_norm": 7.5388503074646, "learning_rate": 3.7844073959123307e-06, "loss": 0.7599, "step": 29552 }, { "epoch": 2.316065830721003, "grad_norm": 3.9538869857788086, "learning_rate": 3.784108288078848e-06, "loss": 1.5787, "step": 29553 }, { "epoch": 2.316144200626959, "grad_norm": 2.8819098472595215, "learning_rate": 3.783809184869991e-06, "loss": 1.45, "step": 29554 }, { "epoch": 2.3162225705329154, "grad_norm": 8.502728462219238, "learning_rate": 3.7835100862869034e-06, "loss": 1.3279, "step": 29555 }, { "epoch": 2.3163009404388717, "grad_norm": 4.062148094177246, "learning_rate": 3.783210992330715e-06, "loss": 1.4853, "step": 29556 }, { "epoch": 2.3163793103448276, "grad_norm": 3.855585813522339, "learning_rate": 3.7829119030025673e-06, "loss": 1.3343, "step": 29557 }, { "epoch": 2.3164576802507835, "grad_norm": 2.7317519187927246, "learning_rate": 3.7826128183035983e-06, "loss": 1.2164, "step": 29558 }, { "epoch": 2.31653605015674, "grad_norm": 7.228530406951904, "learning_rate": 3.7823137382349456e-06, "loss": 1.5323, "step": 29559 }, { "epoch": 2.316614420062696, "grad_norm": 4.536812782287598, "learning_rate": 3.782014662797745e-06, "loss": 1.3721, "step": 29560 }, { "epoch": 2.316692789968652, "grad_norm": 2.666557550430298, "learning_rate": 3.7817155919931343e-06, "loss": 1.2869, "step": 29561 }, { "epoch": 2.316771159874608, "grad_norm": 7.437715530395508, "learning_rate": 3.7814165258222547e-06, "loss": 1.7555, "step": 29562 }, { "epoch": 2.3168495297805642, "grad_norm": 5.201425075531006, "learning_rate": 3.7811174642862373e-06, "loss": 2.0755, "step": 29563 }, { "epoch": 2.3169278996865206, "grad_norm": 9.701679229736328, "learning_rate": 3.780818407386224e-06, "loss": 1.609, "step": 29564 }, { "epoch": 2.3170062695924765, "grad_norm": 2.4329357147216797, "learning_rate": 3.780519355123352e-06, "loss": 1.3474, "step": 29565 }, { "epoch": 2.3170846394984324, "grad_norm": 3.1604902744293213, "learning_rate": 3.7802203074987587e-06, "loss": 1.0481, "step": 29566 }, { "epoch": 2.3171630094043887, "grad_norm": 4.165907382965088, "learning_rate": 3.77992126451358e-06, "loss": 1.3872, "step": 29567 }, { "epoch": 2.317241379310345, "grad_norm": 7.477720737457275, "learning_rate": 3.7796222261689523e-06, "loss": 1.6522, "step": 29568 }, { "epoch": 2.317319749216301, "grad_norm": 6.4526472091674805, "learning_rate": 3.7793231924660166e-06, "loss": 1.1017, "step": 29569 }, { "epoch": 2.3173981191222572, "grad_norm": 3.039048433303833, "learning_rate": 3.7790241634059094e-06, "loss": 1.3709, "step": 29570 }, { "epoch": 2.317476489028213, "grad_norm": 8.840239524841309, "learning_rate": 3.7787251389897663e-06, "loss": 0.735, "step": 29571 }, { "epoch": 2.3175548589341695, "grad_norm": 6.541114330291748, "learning_rate": 3.7784261192187248e-06, "loss": 1.5067, "step": 29572 }, { "epoch": 2.3176332288401253, "grad_norm": 2.011814832687378, "learning_rate": 3.7781271040939223e-06, "loss": 1.3285, "step": 29573 }, { "epoch": 2.3177115987460817, "grad_norm": 3.6577515602111816, "learning_rate": 3.7778280936164996e-06, "loss": 1.5366, "step": 29574 }, { "epoch": 2.3177899686520376, "grad_norm": 5.203529357910156, "learning_rate": 3.777529087787588e-06, "loss": 1.6641, "step": 29575 }, { "epoch": 2.317868338557994, "grad_norm": 3.527846574783325, "learning_rate": 3.7772300866083294e-06, "loss": 1.3217, "step": 29576 }, { "epoch": 2.31794670846395, "grad_norm": 7.345709800720215, "learning_rate": 3.77693109007986e-06, "loss": 2.2935, "step": 29577 }, { "epoch": 2.318025078369906, "grad_norm": 5.6235880851745605, "learning_rate": 3.7766320982033156e-06, "loss": 1.1346, "step": 29578 }, { "epoch": 2.318103448275862, "grad_norm": 3.151278257369995, "learning_rate": 3.7763331109798344e-06, "loss": 1.3065, "step": 29579 }, { "epoch": 2.3181818181818183, "grad_norm": 3.7904207706451416, "learning_rate": 3.7760341284105527e-06, "loss": 1.1553, "step": 29580 }, { "epoch": 2.3182601880877742, "grad_norm": 2.8560543060302734, "learning_rate": 3.775735150496611e-06, "loss": 1.3718, "step": 29581 }, { "epoch": 2.3183385579937306, "grad_norm": 3.7300493717193604, "learning_rate": 3.7754361772391417e-06, "loss": 1.2328, "step": 29582 }, { "epoch": 2.3184169278996865, "grad_norm": 2.6784963607788086, "learning_rate": 3.7751372086392845e-06, "loss": 1.2231, "step": 29583 }, { "epoch": 2.318495297805643, "grad_norm": 2.8094537258148193, "learning_rate": 3.7748382446981767e-06, "loss": 1.517, "step": 29584 }, { "epoch": 2.3185736677115987, "grad_norm": 2.9404473304748535, "learning_rate": 3.774539285416956e-06, "loss": 1.4421, "step": 29585 }, { "epoch": 2.318652037617555, "grad_norm": 3.320608377456665, "learning_rate": 3.7742403307967575e-06, "loss": 1.5097, "step": 29586 }, { "epoch": 2.318730407523511, "grad_norm": 2.952195167541504, "learning_rate": 3.773941380838718e-06, "loss": 1.3792, "step": 29587 }, { "epoch": 2.3188087774294672, "grad_norm": 3.913609027862549, "learning_rate": 3.7736424355439773e-06, "loss": 1.9927, "step": 29588 }, { "epoch": 2.318887147335423, "grad_norm": 1.8060939311981201, "learning_rate": 3.77334349491367e-06, "loss": 0.4463, "step": 29589 }, { "epoch": 2.3189655172413794, "grad_norm": 3.243696451187134, "learning_rate": 3.773044558948934e-06, "loss": 1.8198, "step": 29590 }, { "epoch": 2.3190438871473353, "grad_norm": 2.7749266624450684, "learning_rate": 3.7727456276509066e-06, "loss": 1.2132, "step": 29591 }, { "epoch": 2.3191222570532917, "grad_norm": 3.9544966220855713, "learning_rate": 3.7724467010207256e-06, "loss": 1.326, "step": 29592 }, { "epoch": 2.3192006269592476, "grad_norm": 3.40205454826355, "learning_rate": 3.7721477790595244e-06, "loss": 1.1185, "step": 29593 }, { "epoch": 2.319278996865204, "grad_norm": 3.037036180496216, "learning_rate": 3.7718488617684433e-06, "loss": 1.1066, "step": 29594 }, { "epoch": 2.31935736677116, "grad_norm": 3.109020948410034, "learning_rate": 3.771549949148619e-06, "loss": 1.4939, "step": 29595 }, { "epoch": 2.319435736677116, "grad_norm": 3.368680953979492, "learning_rate": 3.771251041201188e-06, "loss": 1.7384, "step": 29596 }, { "epoch": 2.319514106583072, "grad_norm": 4.7671942710876465, "learning_rate": 3.7709521379272867e-06, "loss": 1.2011, "step": 29597 }, { "epoch": 2.3195924764890283, "grad_norm": 4.937356948852539, "learning_rate": 3.7706532393280517e-06, "loss": 1.4088, "step": 29598 }, { "epoch": 2.319670846394984, "grad_norm": 2.2663798332214355, "learning_rate": 3.7703543454046194e-06, "loss": 1.5304, "step": 29599 }, { "epoch": 2.3197492163009406, "grad_norm": 4.141260147094727, "learning_rate": 3.770055456158129e-06, "loss": 1.4043, "step": 29600 }, { "epoch": 2.3198275862068964, "grad_norm": 5.627161979675293, "learning_rate": 3.769756571589716e-06, "loss": 0.965, "step": 29601 }, { "epoch": 2.3199059561128528, "grad_norm": 4.595532417297363, "learning_rate": 3.7694576917005166e-06, "loss": 1.1526, "step": 29602 }, { "epoch": 2.3199843260188087, "grad_norm": 2.416639804840088, "learning_rate": 3.769158816491669e-06, "loss": 1.5678, "step": 29603 }, { "epoch": 2.320062695924765, "grad_norm": 2.57828950881958, "learning_rate": 3.7688599459643084e-06, "loss": 1.2477, "step": 29604 }, { "epoch": 2.320141065830721, "grad_norm": 6.682190418243408, "learning_rate": 3.7685610801195706e-06, "loss": 1.2129, "step": 29605 }, { "epoch": 2.320219435736677, "grad_norm": 2.625124931335449, "learning_rate": 3.7682622189585956e-06, "loss": 1.4354, "step": 29606 }, { "epoch": 2.320297805642633, "grad_norm": 5.490562438964844, "learning_rate": 3.767963362482519e-06, "loss": 1.4517, "step": 29607 }, { "epoch": 2.3203761755485894, "grad_norm": 3.98646879196167, "learning_rate": 3.767664510692476e-06, "loss": 1.5435, "step": 29608 }, { "epoch": 2.3204545454545453, "grad_norm": 2.8938755989074707, "learning_rate": 3.7673656635896046e-06, "loss": 1.2777, "step": 29609 }, { "epoch": 2.3205329153605017, "grad_norm": 3.1811790466308594, "learning_rate": 3.7670668211750395e-06, "loss": 1.3294, "step": 29610 }, { "epoch": 2.3206112852664575, "grad_norm": 6.641334056854248, "learning_rate": 3.766767983449923e-06, "loss": 1.4901, "step": 29611 }, { "epoch": 2.320689655172414, "grad_norm": 2.881986379623413, "learning_rate": 3.7664691504153832e-06, "loss": 1.2345, "step": 29612 }, { "epoch": 2.3207680250783698, "grad_norm": 3.6418299674987793, "learning_rate": 3.7661703220725634e-06, "loss": 1.7512, "step": 29613 }, { "epoch": 2.320846394984326, "grad_norm": 2.6378300189971924, "learning_rate": 3.765871498422597e-06, "loss": 1.1607, "step": 29614 }, { "epoch": 2.320924764890282, "grad_norm": 2.9330177307128906, "learning_rate": 3.765572679466623e-06, "loss": 1.4941, "step": 29615 }, { "epoch": 2.3210031347962383, "grad_norm": 4.02626371383667, "learning_rate": 3.7652738652057758e-06, "loss": 1.2938, "step": 29616 }, { "epoch": 2.321081504702194, "grad_norm": 2.783578395843506, "learning_rate": 3.7649750556411914e-06, "loss": 1.7621, "step": 29617 }, { "epoch": 2.3211598746081505, "grad_norm": 2.7188987731933594, "learning_rate": 3.7646762507740104e-06, "loss": 1.3617, "step": 29618 }, { "epoch": 2.3212382445141064, "grad_norm": 13.084759712219238, "learning_rate": 3.764377450605363e-06, "loss": 2.125, "step": 29619 }, { "epoch": 2.3213166144200628, "grad_norm": 2.504610061645508, "learning_rate": 3.764078655136391e-06, "loss": 1.28, "step": 29620 }, { "epoch": 2.3213949843260187, "grad_norm": 2.2664947509765625, "learning_rate": 3.763779864368228e-06, "loss": 1.0918, "step": 29621 }, { "epoch": 2.321473354231975, "grad_norm": 3.4697353839874268, "learning_rate": 3.7634810783020136e-06, "loss": 1.3627, "step": 29622 }, { "epoch": 2.321551724137931, "grad_norm": 3.6626365184783936, "learning_rate": 3.7631822969388805e-06, "loss": 0.9727, "step": 29623 }, { "epoch": 2.321630094043887, "grad_norm": 4.4861741065979, "learning_rate": 3.762883520279965e-06, "loss": 0.9222, "step": 29624 }, { "epoch": 2.321708463949843, "grad_norm": 2.2370264530181885, "learning_rate": 3.762584748326407e-06, "loss": 1.4038, "step": 29625 }, { "epoch": 2.3217868338557994, "grad_norm": 4.276562690734863, "learning_rate": 3.7622859810793422e-06, "loss": 1.0889, "step": 29626 }, { "epoch": 2.3218652037617553, "grad_norm": 7.172811031341553, "learning_rate": 3.761987218539904e-06, "loss": 0.7763, "step": 29627 }, { "epoch": 2.3219435736677116, "grad_norm": 5.656269073486328, "learning_rate": 3.7616884607092313e-06, "loss": 1.1344, "step": 29628 }, { "epoch": 2.3220219435736675, "grad_norm": 9.1650390625, "learning_rate": 3.7613897075884585e-06, "loss": 1.0808, "step": 29629 }, { "epoch": 2.322100313479624, "grad_norm": 3.7906711101531982, "learning_rate": 3.761090959178726e-06, "loss": 0.8851, "step": 29630 }, { "epoch": 2.3221786833855798, "grad_norm": 2.5808637142181396, "learning_rate": 3.7607922154811637e-06, "loss": 0.9625, "step": 29631 }, { "epoch": 2.322257053291536, "grad_norm": 2.6291959285736084, "learning_rate": 3.760493476496913e-06, "loss": 1.3354, "step": 29632 }, { "epoch": 2.322335423197492, "grad_norm": 4.6252031326293945, "learning_rate": 3.760194742227109e-06, "loss": 1.4797, "step": 29633 }, { "epoch": 2.3224137931034483, "grad_norm": 2.208653211593628, "learning_rate": 3.759896012672887e-06, "loss": 0.9336, "step": 29634 }, { "epoch": 2.322492163009404, "grad_norm": 3.5930047035217285, "learning_rate": 3.7595972878353824e-06, "loss": 1.3361, "step": 29635 }, { "epoch": 2.3225705329153605, "grad_norm": 3.517152786254883, "learning_rate": 3.759298567715733e-06, "loss": 1.3442, "step": 29636 }, { "epoch": 2.3226489028213164, "grad_norm": 2.269594669342041, "learning_rate": 3.7589998523150766e-06, "loss": 1.3556, "step": 29637 }, { "epoch": 2.3227272727272728, "grad_norm": 2.881474733352661, "learning_rate": 3.758701141634545e-06, "loss": 1.3563, "step": 29638 }, { "epoch": 2.322805642633229, "grad_norm": 3.325376510620117, "learning_rate": 3.758402435675277e-06, "loss": 1.1901, "step": 29639 }, { "epoch": 2.322884012539185, "grad_norm": 2.851926326751709, "learning_rate": 3.7581037344384093e-06, "loss": 1.5207, "step": 29640 }, { "epoch": 2.322962382445141, "grad_norm": 3.301819324493408, "learning_rate": 3.7578050379250775e-06, "loss": 1.9637, "step": 29641 }, { "epoch": 2.323040752351097, "grad_norm": 3.153010368347168, "learning_rate": 3.7575063461364163e-06, "loss": 1.2143, "step": 29642 }, { "epoch": 2.3231191222570535, "grad_norm": 2.4003384113311768, "learning_rate": 3.7572076590735616e-06, "loss": 1.0259, "step": 29643 }, { "epoch": 2.3231974921630094, "grad_norm": 3.937537670135498, "learning_rate": 3.7569089767376542e-06, "loss": 1.4572, "step": 29644 }, { "epoch": 2.3232758620689653, "grad_norm": 2.3989133834838867, "learning_rate": 3.756610299129823e-06, "loss": 1.1502, "step": 29645 }, { "epoch": 2.3233542319749216, "grad_norm": 2.9066600799560547, "learning_rate": 3.7563116262512088e-06, "loss": 1.3918, "step": 29646 }, { "epoch": 2.323432601880878, "grad_norm": 7.052791118621826, "learning_rate": 3.7560129581029466e-06, "loss": 1.8204, "step": 29647 }, { "epoch": 2.323510971786834, "grad_norm": 3.6225953102111816, "learning_rate": 3.7557142946861728e-06, "loss": 1.5494, "step": 29648 }, { "epoch": 2.3235893416927897, "grad_norm": 3.1409711837768555, "learning_rate": 3.7554156360020223e-06, "loss": 1.0891, "step": 29649 }, { "epoch": 2.323667711598746, "grad_norm": 5.084255695343018, "learning_rate": 3.75511698205163e-06, "loss": 1.6464, "step": 29650 }, { "epoch": 2.3237460815047024, "grad_norm": 2.809415817260742, "learning_rate": 3.7548183328361342e-06, "loss": 1.3707, "step": 29651 }, { "epoch": 2.3238244514106583, "grad_norm": 3.7374699115753174, "learning_rate": 3.754519688356671e-06, "loss": 1.6829, "step": 29652 }, { "epoch": 2.323902821316614, "grad_norm": 2.412607431411743, "learning_rate": 3.7542210486143743e-06, "loss": 1.4848, "step": 29653 }, { "epoch": 2.3239811912225705, "grad_norm": 3.28056001663208, "learning_rate": 3.753922413610381e-06, "loss": 1.2013, "step": 29654 }, { "epoch": 2.324059561128527, "grad_norm": 3.0942506790161133, "learning_rate": 3.7536237833458257e-06, "loss": 1.4175, "step": 29655 }, { "epoch": 2.3241379310344827, "grad_norm": 2.27675199508667, "learning_rate": 3.7533251578218475e-06, "loss": 1.4657, "step": 29656 }, { "epoch": 2.3242163009404386, "grad_norm": 2.917172431945801, "learning_rate": 3.753026537039578e-06, "loss": 1.2109, "step": 29657 }, { "epoch": 2.324294670846395, "grad_norm": 2.453296184539795, "learning_rate": 3.752727921000156e-06, "loss": 1.3799, "step": 29658 }, { "epoch": 2.3243730407523513, "grad_norm": 2.6742002964019775, "learning_rate": 3.752429309704717e-06, "loss": 1.3622, "step": 29659 }, { "epoch": 2.324451410658307, "grad_norm": 4.265887260437012, "learning_rate": 3.7521307031543956e-06, "loss": 1.1664, "step": 29660 }, { "epoch": 2.3245297805642635, "grad_norm": 5.061811923980713, "learning_rate": 3.751832101350328e-06, "loss": 1.5148, "step": 29661 }, { "epoch": 2.3246081504702194, "grad_norm": 4.424152374267578, "learning_rate": 3.7515335042936485e-06, "loss": 1.705, "step": 29662 }, { "epoch": 2.3246865203761757, "grad_norm": 3.538837194442749, "learning_rate": 3.7512349119854973e-06, "loss": 1.5244, "step": 29663 }, { "epoch": 2.3247648902821316, "grad_norm": 6.872782230377197, "learning_rate": 3.7509363244270046e-06, "loss": 1.3717, "step": 29664 }, { "epoch": 2.324843260188088, "grad_norm": 7.070764064788818, "learning_rate": 3.7506377416193096e-06, "loss": 1.3083, "step": 29665 }, { "epoch": 2.324921630094044, "grad_norm": 2.74186110496521, "learning_rate": 3.7503391635635465e-06, "loss": 1.2378, "step": 29666 }, { "epoch": 2.325, "grad_norm": 4.685931205749512, "learning_rate": 3.750040590260853e-06, "loss": 0.8295, "step": 29667 }, { "epoch": 2.325078369905956, "grad_norm": 3.31058931350708, "learning_rate": 3.7497420217123616e-06, "loss": 1.0964, "step": 29668 }, { "epoch": 2.3251567398119124, "grad_norm": 2.8572447299957275, "learning_rate": 3.7494434579192074e-06, "loss": 1.343, "step": 29669 }, { "epoch": 2.3252351097178683, "grad_norm": 2.319899320602417, "learning_rate": 3.7491448988825302e-06, "loss": 1.1857, "step": 29670 }, { "epoch": 2.3253134796238246, "grad_norm": 5.3003716468811035, "learning_rate": 3.748846344603464e-06, "loss": 1.3689, "step": 29671 }, { "epoch": 2.3253918495297805, "grad_norm": 4.00405216217041, "learning_rate": 3.7485477950831427e-06, "loss": 1.8084, "step": 29672 }, { "epoch": 2.325470219435737, "grad_norm": 2.9200599193573, "learning_rate": 3.748249250322703e-06, "loss": 1.2476, "step": 29673 }, { "epoch": 2.3255485893416927, "grad_norm": 3.3812949657440186, "learning_rate": 3.7479507103232816e-06, "loss": 1.4547, "step": 29674 }, { "epoch": 2.325626959247649, "grad_norm": 2.257249355316162, "learning_rate": 3.7476521750860107e-06, "loss": 1.3138, "step": 29675 }, { "epoch": 2.325705329153605, "grad_norm": 2.6086747646331787, "learning_rate": 3.747353644612027e-06, "loss": 1.5284, "step": 29676 }, { "epoch": 2.3257836990595613, "grad_norm": 10.403487205505371, "learning_rate": 3.747055118902467e-06, "loss": 1.5518, "step": 29677 }, { "epoch": 2.325862068965517, "grad_norm": 2.790760040283203, "learning_rate": 3.746756597958468e-06, "loss": 1.3035, "step": 29678 }, { "epoch": 2.3259404388714735, "grad_norm": 3.4312243461608887, "learning_rate": 3.7464580817811612e-06, "loss": 1.0158, "step": 29679 }, { "epoch": 2.3260188087774294, "grad_norm": 6.646568298339844, "learning_rate": 3.7461595703716847e-06, "loss": 1.4175, "step": 29680 }, { "epoch": 2.3260971786833857, "grad_norm": 2.9323272705078125, "learning_rate": 3.7458610637311723e-06, "loss": 0.8314, "step": 29681 }, { "epoch": 2.3261755485893416, "grad_norm": 2.823214530944824, "learning_rate": 3.7455625618607627e-06, "loss": 1.6992, "step": 29682 }, { "epoch": 2.326253918495298, "grad_norm": 4.371928691864014, "learning_rate": 3.7452640647615856e-06, "loss": 1.5333, "step": 29683 }, { "epoch": 2.326332288401254, "grad_norm": 2.677590847015381, "learning_rate": 3.744965572434781e-06, "loss": 1.4102, "step": 29684 }, { "epoch": 2.32641065830721, "grad_norm": 2.0710854530334473, "learning_rate": 3.744667084881484e-06, "loss": 1.3655, "step": 29685 }, { "epoch": 2.326489028213166, "grad_norm": 4.872617244720459, "learning_rate": 3.7443686021028268e-06, "loss": 1.7125, "step": 29686 }, { "epoch": 2.3265673981191224, "grad_norm": 2.331246852874756, "learning_rate": 3.744070124099947e-06, "loss": 1.2936, "step": 29687 }, { "epoch": 2.3266457680250783, "grad_norm": 9.417994499206543, "learning_rate": 3.743771650873978e-06, "loss": 1.2838, "step": 29688 }, { "epoch": 2.3267241379310346, "grad_norm": 2.6178061962127686, "learning_rate": 3.743473182426059e-06, "loss": 1.4815, "step": 29689 }, { "epoch": 2.3268025078369905, "grad_norm": 2.842071294784546, "learning_rate": 3.7431747187573207e-06, "loss": 1.3151, "step": 29690 }, { "epoch": 2.326880877742947, "grad_norm": 5.161839485168457, "learning_rate": 3.742876259868902e-06, "loss": 1.2114, "step": 29691 }, { "epoch": 2.3269592476489027, "grad_norm": 4.0006489753723145, "learning_rate": 3.7425778057619343e-06, "loss": 1.3146, "step": 29692 }, { "epoch": 2.327037617554859, "grad_norm": 2.3019819259643555, "learning_rate": 3.742279356437557e-06, "loss": 1.3598, "step": 29693 }, { "epoch": 2.327115987460815, "grad_norm": 2.3126940727233887, "learning_rate": 3.7419809118969e-06, "loss": 0.8797, "step": 29694 }, { "epoch": 2.3271943573667713, "grad_norm": 2.686112642288208, "learning_rate": 3.741682472141104e-06, "loss": 1.2197, "step": 29695 }, { "epoch": 2.327272727272727, "grad_norm": 4.846528053283691, "learning_rate": 3.7413840371713e-06, "loss": 1.4486, "step": 29696 }, { "epoch": 2.3273510971786835, "grad_norm": 4.346251487731934, "learning_rate": 3.7410856069886265e-06, "loss": 1.1304, "step": 29697 }, { "epoch": 2.3274294670846394, "grad_norm": 3.010442018508911, "learning_rate": 3.7407871815942154e-06, "loss": 1.6737, "step": 29698 }, { "epoch": 2.3275078369905957, "grad_norm": 2.9546473026275635, "learning_rate": 3.740488760989204e-06, "loss": 1.3647, "step": 29699 }, { "epoch": 2.3275862068965516, "grad_norm": 2.572760820388794, "learning_rate": 3.7401903451747258e-06, "loss": 1.429, "step": 29700 }, { "epoch": 2.327664576802508, "grad_norm": 13.12234115600586, "learning_rate": 3.739891934151916e-06, "loss": 0.2783, "step": 29701 }, { "epoch": 2.327742946708464, "grad_norm": 2.970250368118286, "learning_rate": 3.73959352792191e-06, "loss": 1.3183, "step": 29702 }, { "epoch": 2.32782131661442, "grad_norm": 13.568202018737793, "learning_rate": 3.739295126485843e-06, "loss": 1.5078, "step": 29703 }, { "epoch": 2.327899686520376, "grad_norm": 3.170332431793213, "learning_rate": 3.7389967298448514e-06, "loss": 1.0435, "step": 29704 }, { "epoch": 2.3279780564263324, "grad_norm": 2.194349527359009, "learning_rate": 3.738698338000067e-06, "loss": 1.3487, "step": 29705 }, { "epoch": 2.3280564263322883, "grad_norm": 5.341950416564941, "learning_rate": 3.7383999509526246e-06, "loss": 1.0149, "step": 29706 }, { "epoch": 2.3281347962382446, "grad_norm": 3.275073528289795, "learning_rate": 3.7381015687036633e-06, "loss": 1.4407, "step": 29707 }, { "epoch": 2.3282131661442005, "grad_norm": 2.682025909423828, "learning_rate": 3.7378031912543156e-06, "loss": 1.2601, "step": 29708 }, { "epoch": 2.328291536050157, "grad_norm": 2.4301607608795166, "learning_rate": 3.7375048186057155e-06, "loss": 1.2941, "step": 29709 }, { "epoch": 2.3283699059561127, "grad_norm": 5.480030059814453, "learning_rate": 3.737206450758999e-06, "loss": 1.3105, "step": 29710 }, { "epoch": 2.328448275862069, "grad_norm": 2.977747917175293, "learning_rate": 3.7369080877152985e-06, "loss": 1.4487, "step": 29711 }, { "epoch": 2.328526645768025, "grad_norm": 9.860457420349121, "learning_rate": 3.7366097294757547e-06, "loss": 1.3991, "step": 29712 }, { "epoch": 2.3286050156739813, "grad_norm": 5.44611120223999, "learning_rate": 3.7363113760414956e-06, "loss": 1.2113, "step": 29713 }, { "epoch": 2.328683385579937, "grad_norm": 3.361053705215454, "learning_rate": 3.73601302741366e-06, "loss": 0.8727, "step": 29714 }, { "epoch": 2.3287617554858935, "grad_norm": 7.688446521759033, "learning_rate": 3.735714683593382e-06, "loss": 0.8888, "step": 29715 }, { "epoch": 2.3288401253918494, "grad_norm": 2.3705430030822754, "learning_rate": 3.735416344581795e-06, "loss": 1.1914, "step": 29716 }, { "epoch": 2.3289184952978057, "grad_norm": 2.5391509532928467, "learning_rate": 3.735118010380035e-06, "loss": 1.5061, "step": 29717 }, { "epoch": 2.3289968652037616, "grad_norm": 2.5092575550079346, "learning_rate": 3.734819680989236e-06, "loss": 1.5457, "step": 29718 }, { "epoch": 2.329075235109718, "grad_norm": 2.617490291595459, "learning_rate": 3.734521356410535e-06, "loss": 1.3841, "step": 29719 }, { "epoch": 2.329153605015674, "grad_norm": 9.795166015625, "learning_rate": 3.734223036645062e-06, "loss": 1.1955, "step": 29720 }, { "epoch": 2.32923197492163, "grad_norm": 4.934027671813965, "learning_rate": 3.7339247216939556e-06, "loss": 1.8792, "step": 29721 }, { "epoch": 2.329310344827586, "grad_norm": 3.2910959720611572, "learning_rate": 3.7336264115583496e-06, "loss": 1.6256, "step": 29722 }, { "epoch": 2.3293887147335424, "grad_norm": 4.204746723175049, "learning_rate": 3.7333281062393787e-06, "loss": 1.6071, "step": 29723 }, { "epoch": 2.3294670846394983, "grad_norm": 2.8639700412750244, "learning_rate": 3.7330298057381766e-06, "loss": 1.6635, "step": 29724 }, { "epoch": 2.3295454545454546, "grad_norm": 3.188969850540161, "learning_rate": 3.732731510055877e-06, "loss": 1.3125, "step": 29725 }, { "epoch": 2.3296238244514105, "grad_norm": 3.036248207092285, "learning_rate": 3.732433219193617e-06, "loss": 1.3401, "step": 29726 }, { "epoch": 2.329702194357367, "grad_norm": 3.7998502254486084, "learning_rate": 3.7321349331525314e-06, "loss": 1.1546, "step": 29727 }, { "epoch": 2.3297805642633227, "grad_norm": 3.180720329284668, "learning_rate": 3.7318366519337514e-06, "loss": 1.3583, "step": 29728 }, { "epoch": 2.329858934169279, "grad_norm": 2.4112753868103027, "learning_rate": 3.7315383755384137e-06, "loss": 1.4085, "step": 29729 }, { "epoch": 2.329937304075235, "grad_norm": 3.3924853801727295, "learning_rate": 3.7312401039676537e-06, "loss": 1.1977, "step": 29730 }, { "epoch": 2.3300156739811912, "grad_norm": 3.930016279220581, "learning_rate": 3.7309418372226034e-06, "loss": 0.6701, "step": 29731 }, { "epoch": 2.330094043887147, "grad_norm": 3.032743215560913, "learning_rate": 3.730643575304398e-06, "loss": 0.9469, "step": 29732 }, { "epoch": 2.3301724137931035, "grad_norm": 4.695648670196533, "learning_rate": 3.7303453182141735e-06, "loss": 0.6899, "step": 29733 }, { "epoch": 2.33025078369906, "grad_norm": 4.018378734588623, "learning_rate": 3.730047065953064e-06, "loss": 1.3133, "step": 29734 }, { "epoch": 2.3303291536050157, "grad_norm": 2.4625613689422607, "learning_rate": 3.729748818522202e-06, "loss": 1.2784, "step": 29735 }, { "epoch": 2.3304075235109716, "grad_norm": 2.873209238052368, "learning_rate": 3.729450575922723e-06, "loss": 1.5078, "step": 29736 }, { "epoch": 2.330485893416928, "grad_norm": 2.8545289039611816, "learning_rate": 3.7291523381557604e-06, "loss": 1.6263, "step": 29737 }, { "epoch": 2.3305642633228842, "grad_norm": 2.1383144855499268, "learning_rate": 3.728854105222453e-06, "loss": 1.4136, "step": 29738 }, { "epoch": 2.33064263322884, "grad_norm": 2.9944143295288086, "learning_rate": 3.7285558771239282e-06, "loss": 1.4234, "step": 29739 }, { "epoch": 2.330721003134796, "grad_norm": 3.1701126098632812, "learning_rate": 3.7282576538613257e-06, "loss": 1.4272, "step": 29740 }, { "epoch": 2.3307993730407524, "grad_norm": 3.7761292457580566, "learning_rate": 3.7279594354357778e-06, "loss": 1.2892, "step": 29741 }, { "epoch": 2.3308777429467087, "grad_norm": 5.719310760498047, "learning_rate": 3.727661221848419e-06, "loss": 1.4833, "step": 29742 }, { "epoch": 2.3309561128526646, "grad_norm": 2.7746384143829346, "learning_rate": 3.7273630131003824e-06, "loss": 1.2343, "step": 29743 }, { "epoch": 2.3310344827586205, "grad_norm": 2.227245330810547, "learning_rate": 3.727064809192802e-06, "loss": 1.2197, "step": 29744 }, { "epoch": 2.331112852664577, "grad_norm": 2.6709656715393066, "learning_rate": 3.7267666101268168e-06, "loss": 1.3069, "step": 29745 }, { "epoch": 2.331191222570533, "grad_norm": 4.852446556091309, "learning_rate": 3.7264684159035547e-06, "loss": 1.3231, "step": 29746 }, { "epoch": 2.331269592476489, "grad_norm": 5.631560325622559, "learning_rate": 3.7261702265241527e-06, "loss": 1.1836, "step": 29747 }, { "epoch": 2.331347962382445, "grad_norm": 3.8265607357025146, "learning_rate": 3.7258720419897456e-06, "loss": 1.5892, "step": 29748 }, { "epoch": 2.3314263322884012, "grad_norm": 3.740060567855835, "learning_rate": 3.725573862301468e-06, "loss": 1.343, "step": 29749 }, { "epoch": 2.3315047021943576, "grad_norm": 3.148672103881836, "learning_rate": 3.725275687460451e-06, "loss": 1.3367, "step": 29750 }, { "epoch": 2.3315830721003135, "grad_norm": 3.005222797393799, "learning_rate": 3.7249775174678303e-06, "loss": 1.0382, "step": 29751 }, { "epoch": 2.3316614420062693, "grad_norm": 2.1496188640594482, "learning_rate": 3.724679352324741e-06, "loss": 1.4178, "step": 29752 }, { "epoch": 2.3317398119122257, "grad_norm": 7.4175190925598145, "learning_rate": 3.7243811920323175e-06, "loss": 1.3694, "step": 29753 }, { "epoch": 2.331818181818182, "grad_norm": 3.6030383110046387, "learning_rate": 3.7240830365916914e-06, "loss": 1.6228, "step": 29754 }, { "epoch": 2.331896551724138, "grad_norm": 5.281221389770508, "learning_rate": 3.7237848860039983e-06, "loss": 1.243, "step": 29755 }, { "epoch": 2.3319749216300942, "grad_norm": 2.073195219039917, "learning_rate": 3.7234867402703735e-06, "loss": 1.2716, "step": 29756 }, { "epoch": 2.33205329153605, "grad_norm": 2.8005309104919434, "learning_rate": 3.723188599391948e-06, "loss": 1.9014, "step": 29757 }, { "epoch": 2.3321316614420065, "grad_norm": 1.9050943851470947, "learning_rate": 3.722890463369856e-06, "loss": 1.3446, "step": 29758 }, { "epoch": 2.3322100313479623, "grad_norm": 1.8872318267822266, "learning_rate": 3.722592332205234e-06, "loss": 1.2117, "step": 29759 }, { "epoch": 2.3322884012539187, "grad_norm": 4.67929744720459, "learning_rate": 3.7222942058992157e-06, "loss": 1.5218, "step": 29760 }, { "epoch": 2.3323667711598746, "grad_norm": 2.9488468170166016, "learning_rate": 3.721996084452933e-06, "loss": 1.2025, "step": 29761 }, { "epoch": 2.332445141065831, "grad_norm": 2.935788869857788, "learning_rate": 3.7216979678675204e-06, "loss": 1.3363, "step": 29762 }, { "epoch": 2.332523510971787, "grad_norm": 5.62684965133667, "learning_rate": 3.721399856144111e-06, "loss": 1.1214, "step": 29763 }, { "epoch": 2.332601880877743, "grad_norm": 2.8309953212738037, "learning_rate": 3.7211017492838443e-06, "loss": 1.4609, "step": 29764 }, { "epoch": 2.332680250783699, "grad_norm": 2.482461929321289, "learning_rate": 3.720803647287845e-06, "loss": 1.4979, "step": 29765 }, { "epoch": 2.3327586206896553, "grad_norm": 2.1396141052246094, "learning_rate": 3.7205055501572527e-06, "loss": 1.0039, "step": 29766 }, { "epoch": 2.3328369905956112, "grad_norm": 3.1390371322631836, "learning_rate": 3.720207457893201e-06, "loss": 1.6979, "step": 29767 }, { "epoch": 2.3329153605015676, "grad_norm": 5.620185852050781, "learning_rate": 3.719909370496824e-06, "loss": 1.1902, "step": 29768 }, { "epoch": 2.3329937304075234, "grad_norm": 7.314517498016357, "learning_rate": 3.719611287969252e-06, "loss": 1.2024, "step": 29769 }, { "epoch": 2.33307210031348, "grad_norm": 2.5476481914520264, "learning_rate": 3.7193132103116204e-06, "loss": 0.7385, "step": 29770 }, { "epoch": 2.3331504702194357, "grad_norm": 4.0751543045043945, "learning_rate": 3.7190151375250673e-06, "loss": 1.6464, "step": 29771 }, { "epoch": 2.333228840125392, "grad_norm": 3.0296316146850586, "learning_rate": 3.718717069610719e-06, "loss": 1.6414, "step": 29772 }, { "epoch": 2.333307210031348, "grad_norm": 2.671663999557495, "learning_rate": 3.7184190065697135e-06, "loss": 1.3162, "step": 29773 }, { "epoch": 2.333385579937304, "grad_norm": 4.089316368103027, "learning_rate": 3.7181209484031845e-06, "loss": 0.9338, "step": 29774 }, { "epoch": 2.33346394984326, "grad_norm": 4.965000152587891, "learning_rate": 3.7178228951122665e-06, "loss": 1.3906, "step": 29775 }, { "epoch": 2.3335423197492164, "grad_norm": 4.878421306610107, "learning_rate": 3.7175248466980895e-06, "loss": 1.0556, "step": 29776 }, { "epoch": 2.3336206896551723, "grad_norm": 3.8680615425109863, "learning_rate": 3.7172268031617885e-06, "loss": 1.4246, "step": 29777 }, { "epoch": 2.3336990595611287, "grad_norm": 2.8981823921203613, "learning_rate": 3.7169287645044993e-06, "loss": 1.3371, "step": 29778 }, { "epoch": 2.3337774294670846, "grad_norm": 4.886695861816406, "learning_rate": 3.716630730727355e-06, "loss": 1.424, "step": 29779 }, { "epoch": 2.333855799373041, "grad_norm": 2.396951675415039, "learning_rate": 3.7163327018314863e-06, "loss": 0.9209, "step": 29780 }, { "epoch": 2.3339341692789968, "grad_norm": 6.5145111083984375, "learning_rate": 3.71603467781803e-06, "loss": 1.278, "step": 29781 }, { "epoch": 2.334012539184953, "grad_norm": 2.1686899662017822, "learning_rate": 3.7157366586881165e-06, "loss": 1.245, "step": 29782 }, { "epoch": 2.334090909090909, "grad_norm": 3.1370558738708496, "learning_rate": 3.7154386444428848e-06, "loss": 1.3902, "step": 29783 }, { "epoch": 2.3341692789968653, "grad_norm": 2.1367950439453125, "learning_rate": 3.7151406350834606e-06, "loss": 1.5421, "step": 29784 }, { "epoch": 2.334247648902821, "grad_norm": 2.1133790016174316, "learning_rate": 3.7148426306109838e-06, "loss": 1.0902, "step": 29785 }, { "epoch": 2.3343260188087775, "grad_norm": 3.1628220081329346, "learning_rate": 3.7145446310265854e-06, "loss": 1.2171, "step": 29786 }, { "epoch": 2.3344043887147334, "grad_norm": 5.407131671905518, "learning_rate": 3.714246636331399e-06, "loss": 1.9101, "step": 29787 }, { "epoch": 2.3344827586206898, "grad_norm": 2.9315638542175293, "learning_rate": 3.7139486465265574e-06, "loss": 1.1817, "step": 29788 }, { "epoch": 2.3345611285266457, "grad_norm": 2.5137441158294678, "learning_rate": 3.7136506616131927e-06, "loss": 1.2069, "step": 29789 }, { "epoch": 2.334639498432602, "grad_norm": 3.059765100479126, "learning_rate": 3.7133526815924437e-06, "loss": 0.9676, "step": 29790 }, { "epoch": 2.334717868338558, "grad_norm": 2.021496295928955, "learning_rate": 3.7130547064654386e-06, "loss": 1.1746, "step": 29791 }, { "epoch": 2.334796238244514, "grad_norm": 3.0810773372650146, "learning_rate": 3.7127567362333127e-06, "loss": 1.3069, "step": 29792 }, { "epoch": 2.33487460815047, "grad_norm": 4.735683917999268, "learning_rate": 3.7124587708971982e-06, "loss": 1.4184, "step": 29793 }, { "epoch": 2.3349529780564264, "grad_norm": 2.6347484588623047, "learning_rate": 3.7121608104582306e-06, "loss": 1.4983, "step": 29794 }, { "epoch": 2.3350313479623823, "grad_norm": 29.98040771484375, "learning_rate": 3.711862854917542e-06, "loss": 1.2383, "step": 29795 }, { "epoch": 2.3351097178683387, "grad_norm": 6.396855354309082, "learning_rate": 3.7115649042762627e-06, "loss": 0.7573, "step": 29796 }, { "epoch": 2.3351880877742945, "grad_norm": 19.53403663635254, "learning_rate": 3.7112669585355316e-06, "loss": 1.225, "step": 29797 }, { "epoch": 2.335266457680251, "grad_norm": 2.1477386951446533, "learning_rate": 3.7109690176964774e-06, "loss": 0.9156, "step": 29798 }, { "epoch": 2.3353448275862068, "grad_norm": 6.004379749298096, "learning_rate": 3.7106710817602353e-06, "loss": 1.3138, "step": 29799 }, { "epoch": 2.335423197492163, "grad_norm": 3.0212955474853516, "learning_rate": 3.7103731507279383e-06, "loss": 1.2181, "step": 29800 }, { "epoch": 2.335501567398119, "grad_norm": 3.2098162174224854, "learning_rate": 3.710075224600721e-06, "loss": 1.3759, "step": 29801 }, { "epoch": 2.3355799373040753, "grad_norm": 2.3043112754821777, "learning_rate": 3.7097773033797126e-06, "loss": 1.5493, "step": 29802 }, { "epoch": 2.335658307210031, "grad_norm": 6.731593132019043, "learning_rate": 3.7094793870660496e-06, "loss": 1.406, "step": 29803 }, { "epoch": 2.3357366771159875, "grad_norm": 4.116870880126953, "learning_rate": 3.7091814756608634e-06, "loss": 1.3658, "step": 29804 }, { "epoch": 2.3358150470219434, "grad_norm": 5.0449934005737305, "learning_rate": 3.7088835691652898e-06, "loss": 1.1808, "step": 29805 }, { "epoch": 2.3358934169278998, "grad_norm": 2.5202176570892334, "learning_rate": 3.7085856675804587e-06, "loss": 1.5085, "step": 29806 }, { "epoch": 2.3359717868338556, "grad_norm": 4.436028003692627, "learning_rate": 3.708287770907503e-06, "loss": 1.2926, "step": 29807 }, { "epoch": 2.336050156739812, "grad_norm": 5.310824394226074, "learning_rate": 3.7079898791475588e-06, "loss": 0.7116, "step": 29808 }, { "epoch": 2.336128526645768, "grad_norm": 3.027674436569214, "learning_rate": 3.7076919923017583e-06, "loss": 0.992, "step": 29809 }, { "epoch": 2.336206896551724, "grad_norm": 3.353583812713623, "learning_rate": 3.7073941103712324e-06, "loss": 1.5643, "step": 29810 }, { "epoch": 2.33628526645768, "grad_norm": 6.738961219787598, "learning_rate": 3.7070962333571157e-06, "loss": 1.6029, "step": 29811 }, { "epoch": 2.3363636363636364, "grad_norm": 2.6092329025268555, "learning_rate": 3.706798361260542e-06, "loss": 1.3937, "step": 29812 }, { "epoch": 2.3364420062695923, "grad_norm": 2.660348892211914, "learning_rate": 3.7065004940826415e-06, "loss": 1.2748, "step": 29813 }, { "epoch": 2.3365203761755486, "grad_norm": 2.584789991378784, "learning_rate": 3.706202631824548e-06, "loss": 1.2273, "step": 29814 }, { "epoch": 2.3365987460815045, "grad_norm": 3.3618922233581543, "learning_rate": 3.705904774487396e-06, "loss": 1.4443, "step": 29815 }, { "epoch": 2.336677115987461, "grad_norm": 4.177647113800049, "learning_rate": 3.705606922072319e-06, "loss": 1.3321, "step": 29816 }, { "epoch": 2.3367554858934168, "grad_norm": 3.3888838291168213, "learning_rate": 3.705309074580447e-06, "loss": 1.1527, "step": 29817 }, { "epoch": 2.336833855799373, "grad_norm": 3.6246871948242188, "learning_rate": 3.7050112320129145e-06, "loss": 0.973, "step": 29818 }, { "epoch": 2.336912225705329, "grad_norm": 4.2655348777771, "learning_rate": 3.704713394370853e-06, "loss": 1.6345, "step": 29819 }, { "epoch": 2.3369905956112853, "grad_norm": 2.4383063316345215, "learning_rate": 3.7044155616554e-06, "loss": 1.1971, "step": 29820 }, { "epoch": 2.337068965517241, "grad_norm": 2.2517452239990234, "learning_rate": 3.7041177338676807e-06, "loss": 1.0989, "step": 29821 }, { "epoch": 2.3371473354231975, "grad_norm": 3.1462206840515137, "learning_rate": 3.7038199110088336e-06, "loss": 1.2701, "step": 29822 }, { "epoch": 2.3372257053291534, "grad_norm": 3.6224329471588135, "learning_rate": 3.7035220930799897e-06, "loss": 1.2722, "step": 29823 }, { "epoch": 2.3373040752351097, "grad_norm": 2.9718704223632812, "learning_rate": 3.7032242800822833e-06, "loss": 1.63, "step": 29824 }, { "epoch": 2.337382445141066, "grad_norm": 2.441105604171753, "learning_rate": 3.702926472016844e-06, "loss": 0.9883, "step": 29825 }, { "epoch": 2.337460815047022, "grad_norm": 2.608954906463623, "learning_rate": 3.7026286688848057e-06, "loss": 1.2419, "step": 29826 }, { "epoch": 2.337539184952978, "grad_norm": 4.93059778213501, "learning_rate": 3.702330870687304e-06, "loss": 1.1572, "step": 29827 }, { "epoch": 2.337617554858934, "grad_norm": 3.8333048820495605, "learning_rate": 3.7020330774254667e-06, "loss": 1.463, "step": 29828 }, { "epoch": 2.3376959247648905, "grad_norm": 2.016326427459717, "learning_rate": 3.7017352891004294e-06, "loss": 1.4223, "step": 29829 }, { "epoch": 2.3377742946708464, "grad_norm": 3.9802513122558594, "learning_rate": 3.7014375057133244e-06, "loss": 0.9482, "step": 29830 }, { "epoch": 2.3378526645768023, "grad_norm": 3.0782577991485596, "learning_rate": 3.701139727265285e-06, "loss": 1.3963, "step": 29831 }, { "epoch": 2.3379310344827586, "grad_norm": 2.333092451095581, "learning_rate": 3.7008419537574425e-06, "loss": 1.4768, "step": 29832 }, { "epoch": 2.338009404388715, "grad_norm": 3.8991572856903076, "learning_rate": 3.7005441851909285e-06, "loss": 1.4762, "step": 29833 }, { "epoch": 2.338087774294671, "grad_norm": 3.407313823699951, "learning_rate": 3.7002464215668775e-06, "loss": 1.6289, "step": 29834 }, { "epoch": 2.3381661442006267, "grad_norm": 2.5448625087738037, "learning_rate": 3.6999486628864234e-06, "loss": 1.6052, "step": 29835 }, { "epoch": 2.338244514106583, "grad_norm": 2.5545639991760254, "learning_rate": 3.6996509091506957e-06, "loss": 1.1789, "step": 29836 }, { "epoch": 2.3383228840125394, "grad_norm": 2.5666818618774414, "learning_rate": 3.699353160360828e-06, "loss": 1.0321, "step": 29837 }, { "epoch": 2.3384012539184953, "grad_norm": 2.9964499473571777, "learning_rate": 3.6990554165179517e-06, "loss": 1.2834, "step": 29838 }, { "epoch": 2.338479623824451, "grad_norm": 3.0510215759277344, "learning_rate": 3.6987576776232025e-06, "loss": 1.4128, "step": 29839 }, { "epoch": 2.3385579937304075, "grad_norm": 3.0057365894317627, "learning_rate": 3.6984599436777093e-06, "loss": 2.2001, "step": 29840 }, { "epoch": 2.338636363636364, "grad_norm": 5.534137725830078, "learning_rate": 3.6981622146826056e-06, "loss": 1.9083, "step": 29841 }, { "epoch": 2.3387147335423197, "grad_norm": 4.4911065101623535, "learning_rate": 3.697864490639026e-06, "loss": 1.5275, "step": 29842 }, { "epoch": 2.3387931034482756, "grad_norm": 3.5612661838531494, "learning_rate": 3.6975667715481e-06, "loss": 1.5189, "step": 29843 }, { "epoch": 2.338871473354232, "grad_norm": 3.454130172729492, "learning_rate": 3.697269057410961e-06, "loss": 1.2729, "step": 29844 }, { "epoch": 2.3389498432601883, "grad_norm": 2.1564812660217285, "learning_rate": 3.69697134822874e-06, "loss": 1.0091, "step": 29845 }, { "epoch": 2.339028213166144, "grad_norm": 3.515916585922241, "learning_rate": 3.696673644002574e-06, "loss": 1.1274, "step": 29846 }, { "epoch": 2.3391065830721, "grad_norm": 3.2423956394195557, "learning_rate": 3.696375944733589e-06, "loss": 1.5187, "step": 29847 }, { "epoch": 2.3391849529780564, "grad_norm": 4.281861782073975, "learning_rate": 3.6960782504229216e-06, "loss": 1.3152, "step": 29848 }, { "epoch": 2.3392633228840127, "grad_norm": 8.651993751525879, "learning_rate": 3.6957805610717026e-06, "loss": 2.0696, "step": 29849 }, { "epoch": 2.3393416927899686, "grad_norm": 3.1453723907470703, "learning_rate": 3.695482876681065e-06, "loss": 1.5906, "step": 29850 }, { "epoch": 2.339420062695925, "grad_norm": 2.0183935165405273, "learning_rate": 3.6951851972521403e-06, "loss": 1.2884, "step": 29851 }, { "epoch": 2.339498432601881, "grad_norm": 3.3496809005737305, "learning_rate": 3.694887522786059e-06, "loss": 1.3554, "step": 29852 }, { "epoch": 2.339576802507837, "grad_norm": 4.117028713226318, "learning_rate": 3.694589853283959e-06, "loss": 1.173, "step": 29853 }, { "epoch": 2.339655172413793, "grad_norm": 8.273066520690918, "learning_rate": 3.6942921887469653e-06, "loss": 0.9478, "step": 29854 }, { "epoch": 2.3397335423197494, "grad_norm": 8.086015701293945, "learning_rate": 3.6939945291762147e-06, "loss": 1.2203, "step": 29855 }, { "epoch": 2.3398119122257053, "grad_norm": 7.092281818389893, "learning_rate": 3.693696874572838e-06, "loss": 1.1877, "step": 29856 }, { "epoch": 2.3398902821316616, "grad_norm": 6.147856712341309, "learning_rate": 3.6933992249379685e-06, "loss": 1.6984, "step": 29857 }, { "epoch": 2.3399686520376175, "grad_norm": 2.899022340774536, "learning_rate": 3.693101580272736e-06, "loss": 1.4333, "step": 29858 }, { "epoch": 2.340047021943574, "grad_norm": 5.402469635009766, "learning_rate": 3.6928039405782727e-06, "loss": 1.5756, "step": 29859 }, { "epoch": 2.3401253918495297, "grad_norm": 10.07816219329834, "learning_rate": 3.692506305855713e-06, "loss": 1.0462, "step": 29860 }, { "epoch": 2.340203761755486, "grad_norm": 7.001832962036133, "learning_rate": 3.6922086761061893e-06, "loss": 1.3528, "step": 29861 }, { "epoch": 2.340282131661442, "grad_norm": 2.6254305839538574, "learning_rate": 3.6919110513308305e-06, "loss": 1.2887, "step": 29862 }, { "epoch": 2.3403605015673983, "grad_norm": 3.136232614517212, "learning_rate": 3.6916134315307704e-06, "loss": 1.1719, "step": 29863 }, { "epoch": 2.340438871473354, "grad_norm": 3.8041200637817383, "learning_rate": 3.6913158167071394e-06, "loss": 1.542, "step": 29864 }, { "epoch": 2.3405172413793105, "grad_norm": 2.9644806385040283, "learning_rate": 3.6910182068610743e-06, "loss": 1.5482, "step": 29865 }, { "epoch": 2.3405956112852664, "grad_norm": 10.185924530029297, "learning_rate": 3.6907206019937003e-06, "loss": 1.4183, "step": 29866 }, { "epoch": 2.3406739811912227, "grad_norm": 2.5742571353912354, "learning_rate": 3.6904230021061536e-06, "loss": 1.3734, "step": 29867 }, { "epoch": 2.3407523510971786, "grad_norm": 9.717803955078125, "learning_rate": 3.6901254071995665e-06, "loss": 1.2205, "step": 29868 }, { "epoch": 2.340830721003135, "grad_norm": 3.0029072761535645, "learning_rate": 3.689827817275068e-06, "loss": 1.0938, "step": 29869 }, { "epoch": 2.340909090909091, "grad_norm": 4.255945205688477, "learning_rate": 3.6895302323337924e-06, "loss": 0.8258, "step": 29870 }, { "epoch": 2.340987460815047, "grad_norm": 2.4444775581359863, "learning_rate": 3.6892326523768683e-06, "loss": 1.1608, "step": 29871 }, { "epoch": 2.341065830721003, "grad_norm": 2.6799850463867188, "learning_rate": 3.6889350774054345e-06, "loss": 1.3718, "step": 29872 }, { "epoch": 2.3411442006269594, "grad_norm": 5.054339408874512, "learning_rate": 3.6886375074206145e-06, "loss": 0.9903, "step": 29873 }, { "epoch": 2.3412225705329153, "grad_norm": 2.8814127445220947, "learning_rate": 3.6883399424235446e-06, "loss": 1.4251, "step": 29874 }, { "epoch": 2.3413009404388716, "grad_norm": 3.8252274990081787, "learning_rate": 3.6880423824153565e-06, "loss": 1.674, "step": 29875 }, { "epoch": 2.3413793103448275, "grad_norm": 5.4796295166015625, "learning_rate": 3.6877448273971817e-06, "loss": 1.1267, "step": 29876 }, { "epoch": 2.341457680250784, "grad_norm": 2.5210511684417725, "learning_rate": 3.6874472773701508e-06, "loss": 1.3954, "step": 29877 }, { "epoch": 2.3415360501567397, "grad_norm": 3.626831531524658, "learning_rate": 3.6871497323353956e-06, "loss": 1.7492, "step": 29878 }, { "epoch": 2.341614420062696, "grad_norm": 4.261911869049072, "learning_rate": 3.686852192294049e-06, "loss": 1.4468, "step": 29879 }, { "epoch": 2.341692789968652, "grad_norm": 2.859626531600952, "learning_rate": 3.686554657247243e-06, "loss": 1.3238, "step": 29880 }, { "epoch": 2.3417711598746083, "grad_norm": 2.4705545902252197, "learning_rate": 3.6862571271961082e-06, "loss": 1.4573, "step": 29881 }, { "epoch": 2.341849529780564, "grad_norm": 23.972747802734375, "learning_rate": 3.6859596021417765e-06, "loss": 1.3736, "step": 29882 }, { "epoch": 2.3419278996865205, "grad_norm": 2.465970754623413, "learning_rate": 3.6856620820853805e-06, "loss": 1.2981, "step": 29883 }, { "epoch": 2.3420062695924764, "grad_norm": 11.752401351928711, "learning_rate": 3.6853645670280493e-06, "loss": 1.6267, "step": 29884 }, { "epoch": 2.3420846394984327, "grad_norm": 2.4898509979248047, "learning_rate": 3.6850670569709158e-06, "loss": 1.1598, "step": 29885 }, { "epoch": 2.3421630094043886, "grad_norm": 2.93463134765625, "learning_rate": 3.6847695519151116e-06, "loss": 1.5353, "step": 29886 }, { "epoch": 2.342241379310345, "grad_norm": 2.75148868560791, "learning_rate": 3.6844720518617705e-06, "loss": 1.1762, "step": 29887 }, { "epoch": 2.342319749216301, "grad_norm": 2.604487419128418, "learning_rate": 3.6841745568120207e-06, "loss": 1.2982, "step": 29888 }, { "epoch": 2.342398119122257, "grad_norm": 4.264647006988525, "learning_rate": 3.6838770667669956e-06, "loss": 1.3418, "step": 29889 }, { "epoch": 2.342476489028213, "grad_norm": 2.739664077758789, "learning_rate": 3.683579581727824e-06, "loss": 1.3271, "step": 29890 }, { "epoch": 2.3425548589341694, "grad_norm": 2.865367889404297, "learning_rate": 3.683282101695642e-06, "loss": 1.4599, "step": 29891 }, { "epoch": 2.3426332288401253, "grad_norm": 2.7086336612701416, "learning_rate": 3.6829846266715778e-06, "loss": 1.4554, "step": 29892 }, { "epoch": 2.3427115987460816, "grad_norm": 2.866399049758911, "learning_rate": 3.6826871566567634e-06, "loss": 1.2401, "step": 29893 }, { "epoch": 2.3427899686520375, "grad_norm": 4.82426118850708, "learning_rate": 3.6823896916523306e-06, "loss": 1.049, "step": 29894 }, { "epoch": 2.342868338557994, "grad_norm": 2.6300694942474365, "learning_rate": 3.682092231659412e-06, "loss": 1.0977, "step": 29895 }, { "epoch": 2.3429467084639497, "grad_norm": 2.9833662509918213, "learning_rate": 3.6817947766791364e-06, "loss": 1.3028, "step": 29896 }, { "epoch": 2.343025078369906, "grad_norm": 2.73443341255188, "learning_rate": 3.681497326712635e-06, "loss": 1.3912, "step": 29897 }, { "epoch": 2.343103448275862, "grad_norm": 9.206270217895508, "learning_rate": 3.681199881761043e-06, "loss": 1.1665, "step": 29898 }, { "epoch": 2.3431818181818183, "grad_norm": 3.0489683151245117, "learning_rate": 3.680902441825488e-06, "loss": 1.227, "step": 29899 }, { "epoch": 2.343260188087774, "grad_norm": 2.8828723430633545, "learning_rate": 3.6806050069071025e-06, "loss": 1.0149, "step": 29900 }, { "epoch": 2.3433385579937305, "grad_norm": 2.225100517272949, "learning_rate": 3.680307577007018e-06, "loss": 1.3848, "step": 29901 }, { "epoch": 2.3434169278996864, "grad_norm": 2.9118540287017822, "learning_rate": 3.6800101521263666e-06, "loss": 0.8612, "step": 29902 }, { "epoch": 2.3434952978056427, "grad_norm": 6.539895057678223, "learning_rate": 3.679712732266277e-06, "loss": 1.3978, "step": 29903 }, { "epoch": 2.3435736677115986, "grad_norm": 3.8063321113586426, "learning_rate": 3.6794153174278824e-06, "loss": 1.6149, "step": 29904 }, { "epoch": 2.343652037617555, "grad_norm": 3.269352674484253, "learning_rate": 3.679117907612314e-06, "loss": 1.3548, "step": 29905 }, { "epoch": 2.343730407523511, "grad_norm": 7.792928695678711, "learning_rate": 3.6788205028207036e-06, "loss": 1.109, "step": 29906 }, { "epoch": 2.343808777429467, "grad_norm": 2.474578857421875, "learning_rate": 3.67852310305418e-06, "loss": 1.0883, "step": 29907 }, { "epoch": 2.343887147335423, "grad_norm": 2.48919415473938, "learning_rate": 3.678225708313875e-06, "loss": 1.6014, "step": 29908 }, { "epoch": 2.3439655172413794, "grad_norm": 3.341087818145752, "learning_rate": 3.677928318600924e-06, "loss": 1.9243, "step": 29909 }, { "epoch": 2.3440438871473352, "grad_norm": 2.1029021739959717, "learning_rate": 3.6776309339164507e-06, "loss": 1.3005, "step": 29910 }, { "epoch": 2.3441222570532916, "grad_norm": 3.7235238552093506, "learning_rate": 3.6773335542615916e-06, "loss": 0.9887, "step": 29911 }, { "epoch": 2.3442006269592475, "grad_norm": 2.6664955615997314, "learning_rate": 3.6770361796374763e-06, "loss": 1.4499, "step": 29912 }, { "epoch": 2.344278996865204, "grad_norm": 2.9012811183929443, "learning_rate": 3.6767388100452373e-06, "loss": 1.4988, "step": 29913 }, { "epoch": 2.3443573667711597, "grad_norm": 3.014214515686035, "learning_rate": 3.6764414454860033e-06, "loss": 1.3768, "step": 29914 }, { "epoch": 2.344435736677116, "grad_norm": 3.0158627033233643, "learning_rate": 3.6761440859609042e-06, "loss": 1.5771, "step": 29915 }, { "epoch": 2.344514106583072, "grad_norm": 3.6534156799316406, "learning_rate": 3.6758467314710742e-06, "loss": 0.6113, "step": 29916 }, { "epoch": 2.3445924764890282, "grad_norm": 2.298964738845825, "learning_rate": 3.6755493820176445e-06, "loss": 1.3257, "step": 29917 }, { "epoch": 2.344670846394984, "grad_norm": 3.286975383758545, "learning_rate": 3.675252037601744e-06, "loss": 1.3175, "step": 29918 }, { "epoch": 2.3447492163009405, "grad_norm": 2.448530673980713, "learning_rate": 3.6749546982245037e-06, "loss": 1.1033, "step": 29919 }, { "epoch": 2.344827586206897, "grad_norm": 3.3493010997772217, "learning_rate": 3.674657363887054e-06, "loss": 1.4127, "step": 29920 }, { "epoch": 2.3449059561128527, "grad_norm": 3.696316719055176, "learning_rate": 3.6743600345905306e-06, "loss": 1.3373, "step": 29921 }, { "epoch": 2.3449843260188086, "grad_norm": 4.072622299194336, "learning_rate": 3.674062710336057e-06, "loss": 1.6748, "step": 29922 }, { "epoch": 2.345062695924765, "grad_norm": 7.895579814910889, "learning_rate": 3.6737653911247694e-06, "loss": 1.1225, "step": 29923 }, { "epoch": 2.3451410658307212, "grad_norm": 3.6517364978790283, "learning_rate": 3.673468076957798e-06, "loss": 1.1997, "step": 29924 }, { "epoch": 2.345219435736677, "grad_norm": 3.0409457683563232, "learning_rate": 3.6731707678362716e-06, "loss": 1.0166, "step": 29925 }, { "epoch": 2.345297805642633, "grad_norm": 2.6837503910064697, "learning_rate": 3.672873463761322e-06, "loss": 1.3214, "step": 29926 }, { "epoch": 2.3453761755485893, "grad_norm": 3.3508636951446533, "learning_rate": 3.6725761647340796e-06, "loss": 1.2925, "step": 29927 }, { "epoch": 2.3454545454545457, "grad_norm": 3.0825459957122803, "learning_rate": 3.6722788707556777e-06, "loss": 1.4451, "step": 29928 }, { "epoch": 2.3455329153605016, "grad_norm": 2.2530155181884766, "learning_rate": 3.671981581827243e-06, "loss": 1.2431, "step": 29929 }, { "epoch": 2.3456112852664575, "grad_norm": 9.079004287719727, "learning_rate": 3.671684297949909e-06, "loss": 1.235, "step": 29930 }, { "epoch": 2.345689655172414, "grad_norm": 12.190821647644043, "learning_rate": 3.6713870191248054e-06, "loss": 1.3357, "step": 29931 }, { "epoch": 2.34576802507837, "grad_norm": 3.799923896789551, "learning_rate": 3.6710897453530644e-06, "loss": 1.564, "step": 29932 }, { "epoch": 2.345846394984326, "grad_norm": 2.5843546390533447, "learning_rate": 3.6707924766358137e-06, "loss": 1.4981, "step": 29933 }, { "epoch": 2.345924764890282, "grad_norm": 6.16201639175415, "learning_rate": 3.6704952129741856e-06, "loss": 1.0616, "step": 29934 }, { "epoch": 2.3460031347962382, "grad_norm": 3.866917371749878, "learning_rate": 3.670197954369311e-06, "loss": 1.4386, "step": 29935 }, { "epoch": 2.3460815047021946, "grad_norm": 3.1901700496673584, "learning_rate": 3.669900700822322e-06, "loss": 1.5619, "step": 29936 }, { "epoch": 2.3461598746081505, "grad_norm": 6.909806251525879, "learning_rate": 3.6696034523343466e-06, "loss": 0.9969, "step": 29937 }, { "epoch": 2.3462382445141063, "grad_norm": 6.108874797821045, "learning_rate": 3.6693062089065157e-06, "loss": 1.4476, "step": 29938 }, { "epoch": 2.3463166144200627, "grad_norm": 3.505589008331299, "learning_rate": 3.6690089705399624e-06, "loss": 1.3839, "step": 29939 }, { "epoch": 2.346394984326019, "grad_norm": 3.5141513347625732, "learning_rate": 3.668711737235814e-06, "loss": 1.4928, "step": 29940 }, { "epoch": 2.346473354231975, "grad_norm": 4.643561840057373, "learning_rate": 3.6684145089952006e-06, "loss": 1.5763, "step": 29941 }, { "epoch": 2.3465517241379312, "grad_norm": 2.5692520141601562, "learning_rate": 3.6681172858192558e-06, "loss": 1.1354, "step": 29942 }, { "epoch": 2.346630094043887, "grad_norm": 3.942005157470703, "learning_rate": 3.66782006770911e-06, "loss": 1.3174, "step": 29943 }, { "epoch": 2.3467084639498434, "grad_norm": 2.3622398376464844, "learning_rate": 3.667522854665892e-06, "loss": 1.3328, "step": 29944 }, { "epoch": 2.3467868338557993, "grad_norm": 2.2378504276275635, "learning_rate": 3.667225646690732e-06, "loss": 1.156, "step": 29945 }, { "epoch": 2.3468652037617557, "grad_norm": 3.131587028503418, "learning_rate": 3.6669284437847603e-06, "loss": 1.2493, "step": 29946 }, { "epoch": 2.3469435736677116, "grad_norm": 2.9272384643554688, "learning_rate": 3.666631245949111e-06, "loss": 1.2475, "step": 29947 }, { "epoch": 2.347021943573668, "grad_norm": 20.55886459350586, "learning_rate": 3.6663340531849083e-06, "loss": 1.1165, "step": 29948 }, { "epoch": 2.347100313479624, "grad_norm": 3.3011977672576904, "learning_rate": 3.6660368654932878e-06, "loss": 1.514, "step": 29949 }, { "epoch": 2.34717868338558, "grad_norm": 3.86841082572937, "learning_rate": 3.6657396828753777e-06, "loss": 1.1354, "step": 29950 }, { "epoch": 2.347257053291536, "grad_norm": 2.882725238800049, "learning_rate": 3.6654425053323084e-06, "loss": 1.6498, "step": 29951 }, { "epoch": 2.3473354231974923, "grad_norm": 6.868989944458008, "learning_rate": 3.66514533286521e-06, "loss": 1.1275, "step": 29952 }, { "epoch": 2.347413793103448, "grad_norm": 2.453437328338623, "learning_rate": 3.6648481654752115e-06, "loss": 1.3881, "step": 29953 }, { "epoch": 2.3474921630094046, "grad_norm": 3.317828893661499, "learning_rate": 3.6645510031634484e-06, "loss": 1.2705, "step": 29954 }, { "epoch": 2.3475705329153604, "grad_norm": 3.6282050609588623, "learning_rate": 3.6642538459310446e-06, "loss": 1.1835, "step": 29955 }, { "epoch": 2.3476489028213168, "grad_norm": 4.8131914138793945, "learning_rate": 3.663956693779134e-06, "loss": 1.4113, "step": 29956 }, { "epoch": 2.3477272727272727, "grad_norm": 3.961702585220337, "learning_rate": 3.663659546708845e-06, "loss": 1.354, "step": 29957 }, { "epoch": 2.347805642633229, "grad_norm": 10.182126998901367, "learning_rate": 3.663362404721311e-06, "loss": 1.2615, "step": 29958 }, { "epoch": 2.347884012539185, "grad_norm": 5.054540157318115, "learning_rate": 3.6630652678176583e-06, "loss": 1.4687, "step": 29959 }, { "epoch": 2.347962382445141, "grad_norm": 4.340758323669434, "learning_rate": 3.6627681359990174e-06, "loss": 1.6052, "step": 29960 }, { "epoch": 2.348040752351097, "grad_norm": 3.06575608253479, "learning_rate": 3.6624710092665205e-06, "loss": 1.0811, "step": 29961 }, { "epoch": 2.3481191222570534, "grad_norm": 3.2030527591705322, "learning_rate": 3.6621738876212988e-06, "loss": 1.2608, "step": 29962 }, { "epoch": 2.3481974921630093, "grad_norm": 2.813871383666992, "learning_rate": 3.661876771064479e-06, "loss": 0.9638, "step": 29963 }, { "epoch": 2.3482758620689657, "grad_norm": 6.602829456329346, "learning_rate": 3.6615796595971923e-06, "loss": 1.1018, "step": 29964 }, { "epoch": 2.3483542319749215, "grad_norm": 5.723362922668457, "learning_rate": 3.661282553220571e-06, "loss": 1.3706, "step": 29965 }, { "epoch": 2.348432601880878, "grad_norm": 5.705687999725342, "learning_rate": 3.6609854519357415e-06, "loss": 1.2952, "step": 29966 }, { "epoch": 2.3485109717868338, "grad_norm": 4.3294854164123535, "learning_rate": 3.6606883557438344e-06, "loss": 1.0643, "step": 29967 }, { "epoch": 2.34858934169279, "grad_norm": 4.2677226066589355, "learning_rate": 3.6603912646459823e-06, "loss": 1.1317, "step": 29968 }, { "epoch": 2.348667711598746, "grad_norm": 5.495993614196777, "learning_rate": 3.660094178643315e-06, "loss": 2.003, "step": 29969 }, { "epoch": 2.3487460815047023, "grad_norm": 2.5225813388824463, "learning_rate": 3.6597970977369594e-06, "loss": 1.598, "step": 29970 }, { "epoch": 2.348824451410658, "grad_norm": 2.5634818077087402, "learning_rate": 3.659500021928048e-06, "loss": 1.3829, "step": 29971 }, { "epoch": 2.3489028213166145, "grad_norm": 2.8844616413116455, "learning_rate": 3.6592029512177085e-06, "loss": 0.8977, "step": 29972 }, { "epoch": 2.3489811912225704, "grad_norm": 2.4228274822235107, "learning_rate": 3.658905885607076e-06, "loss": 1.5704, "step": 29973 }, { "epoch": 2.3490595611285268, "grad_norm": 4.616299629211426, "learning_rate": 3.658608825097273e-06, "loss": 1.392, "step": 29974 }, { "epoch": 2.3491379310344827, "grad_norm": 3.576517105102539, "learning_rate": 3.6583117696894344e-06, "loss": 1.2713, "step": 29975 }, { "epoch": 2.349216300940439, "grad_norm": 2.9628310203552246, "learning_rate": 3.6580147193846886e-06, "loss": 1.2312, "step": 29976 }, { "epoch": 2.349294670846395, "grad_norm": 5.220284461975098, "learning_rate": 3.657717674184167e-06, "loss": 1.2737, "step": 29977 }, { "epoch": 2.349373040752351, "grad_norm": 3.080179452896118, "learning_rate": 3.657420634088996e-06, "loss": 1.6856, "step": 29978 }, { "epoch": 2.349451410658307, "grad_norm": 3.1230883598327637, "learning_rate": 3.6571235991003068e-06, "loss": 0.8264, "step": 29979 }, { "epoch": 2.3495297805642634, "grad_norm": 16.391033172607422, "learning_rate": 3.656826569219233e-06, "loss": 2.8222, "step": 29980 }, { "epoch": 2.3496081504702193, "grad_norm": 2.411809206008911, "learning_rate": 3.656529544446897e-06, "loss": 1.301, "step": 29981 }, { "epoch": 2.3496865203761756, "grad_norm": 2.496427297592163, "learning_rate": 3.6562325247844343e-06, "loss": 1.4937, "step": 29982 }, { "epoch": 2.3497648902821315, "grad_norm": 10.297757148742676, "learning_rate": 3.655935510232973e-06, "loss": 1.212, "step": 29983 }, { "epoch": 2.349843260188088, "grad_norm": 3.959552526473999, "learning_rate": 3.6556385007936433e-06, "loss": 0.8245, "step": 29984 }, { "epoch": 2.3499216300940438, "grad_norm": 5.249539852142334, "learning_rate": 3.6553414964675738e-06, "loss": 1.2139, "step": 29985 }, { "epoch": 2.35, "grad_norm": 2.9119417667388916, "learning_rate": 3.655044497255893e-06, "loss": 1.3258, "step": 29986 }, { "epoch": 2.350078369905956, "grad_norm": 3.2886860370635986, "learning_rate": 3.6547475031597334e-06, "loss": 1.3745, "step": 29987 }, { "epoch": 2.3501567398119123, "grad_norm": 2.893555164337158, "learning_rate": 3.6544505141802247e-06, "loss": 1.1581, "step": 29988 }, { "epoch": 2.350235109717868, "grad_norm": 2.9640188217163086, "learning_rate": 3.654153530318494e-06, "loss": 1.1169, "step": 29989 }, { "epoch": 2.3503134796238245, "grad_norm": 4.006764888763428, "learning_rate": 3.6538565515756714e-06, "loss": 1.3941, "step": 29990 }, { "epoch": 2.3503918495297804, "grad_norm": 3.8790969848632812, "learning_rate": 3.653559577952886e-06, "loss": 1.386, "step": 29991 }, { "epoch": 2.3504702194357368, "grad_norm": 4.969630718231201, "learning_rate": 3.653262609451272e-06, "loss": 1.3891, "step": 29992 }, { "epoch": 2.3505485893416926, "grad_norm": 3.268603801727295, "learning_rate": 3.652965646071952e-06, "loss": 1.4076, "step": 29993 }, { "epoch": 2.350626959247649, "grad_norm": 2.235131025314331, "learning_rate": 3.65266868781606e-06, "loss": 1.346, "step": 29994 }, { "epoch": 2.350705329153605, "grad_norm": 2.2809324264526367, "learning_rate": 3.6523717346847246e-06, "loss": 1.3314, "step": 29995 }, { "epoch": 2.350783699059561, "grad_norm": 5.273563385009766, "learning_rate": 3.652074786679075e-06, "loss": 1.4836, "step": 29996 }, { "epoch": 2.350862068965517, "grad_norm": 3.205561637878418, "learning_rate": 3.65177784380024e-06, "loss": 1.6152, "step": 29997 }, { "epoch": 2.3509404388714734, "grad_norm": 3.3406310081481934, "learning_rate": 3.6514809060493482e-06, "loss": 1.7536, "step": 29998 }, { "epoch": 2.3510188087774293, "grad_norm": 2.646238088607788, "learning_rate": 3.6511839734275336e-06, "loss": 1.2948, "step": 29999 }, { "epoch": 2.3510971786833856, "grad_norm": 2.6609323024749756, "learning_rate": 3.65088704593592e-06, "loss": 1.4588, "step": 30000 }, { "epoch": 2.3511755485893415, "grad_norm": 6.544553756713867, "learning_rate": 3.65059012357564e-06, "loss": 0.9549, "step": 30001 }, { "epoch": 2.351253918495298, "grad_norm": 2.316575050354004, "learning_rate": 3.6502932063478215e-06, "loss": 1.3354, "step": 30002 }, { "epoch": 2.3513322884012537, "grad_norm": 16.794784545898438, "learning_rate": 3.649996294253596e-06, "loss": 1.6688, "step": 30003 }, { "epoch": 2.35141065830721, "grad_norm": 11.020087242126465, "learning_rate": 3.649699387294089e-06, "loss": 1.0618, "step": 30004 }, { "epoch": 2.351489028213166, "grad_norm": 3.921236753463745, "learning_rate": 3.6494024854704325e-06, "loss": 1.5313, "step": 30005 }, { "epoch": 2.3515673981191223, "grad_norm": 5.9728264808654785, "learning_rate": 3.649105588783757e-06, "loss": 2.1023, "step": 30006 }, { "epoch": 2.351645768025078, "grad_norm": 7.3484721183776855, "learning_rate": 3.648808697235188e-06, "loss": 0.8668, "step": 30007 }, { "epoch": 2.3517241379310345, "grad_norm": 4.208142280578613, "learning_rate": 3.6485118108258573e-06, "loss": 1.5217, "step": 30008 }, { "epoch": 2.3518025078369904, "grad_norm": 2.509420394897461, "learning_rate": 3.6482149295568936e-06, "loss": 1.5554, "step": 30009 }, { "epoch": 2.3518808777429467, "grad_norm": 4.276337623596191, "learning_rate": 3.6479180534294266e-06, "loss": 1.4953, "step": 30010 }, { "epoch": 2.3519592476489026, "grad_norm": 3.1786277294158936, "learning_rate": 3.647621182444583e-06, "loss": 0.9873, "step": 30011 }, { "epoch": 2.352037617554859, "grad_norm": 7.043867111206055, "learning_rate": 3.6473243166034948e-06, "loss": 1.0753, "step": 30012 }, { "epoch": 2.352115987460815, "grad_norm": 3.1425609588623047, "learning_rate": 3.6470274559072906e-06, "loss": 1.3311, "step": 30013 }, { "epoch": 2.352194357366771, "grad_norm": 2.1262972354888916, "learning_rate": 3.6467306003570996e-06, "loss": 1.413, "step": 30014 }, { "epoch": 2.3522727272727275, "grad_norm": 3.4339303970336914, "learning_rate": 3.646433749954049e-06, "loss": 1.4835, "step": 30015 }, { "epoch": 2.3523510971786834, "grad_norm": 2.818136215209961, "learning_rate": 3.646136904699269e-06, "loss": 1.2438, "step": 30016 }, { "epoch": 2.3524294670846393, "grad_norm": 2.387907028198242, "learning_rate": 3.645840064593889e-06, "loss": 1.6674, "step": 30017 }, { "epoch": 2.3525078369905956, "grad_norm": 2.148715019226074, "learning_rate": 3.645543229639039e-06, "loss": 1.2974, "step": 30018 }, { "epoch": 2.352586206896552, "grad_norm": 2.533078193664551, "learning_rate": 3.6452463998358467e-06, "loss": 0.8614, "step": 30019 }, { "epoch": 2.352664576802508, "grad_norm": 3.7526352405548096, "learning_rate": 3.64494957518544e-06, "loss": 1.076, "step": 30020 }, { "epoch": 2.3527429467084637, "grad_norm": 2.8247461318969727, "learning_rate": 3.644652755688951e-06, "loss": 1.2556, "step": 30021 }, { "epoch": 2.35282131661442, "grad_norm": 3.1443655490875244, "learning_rate": 3.6443559413475054e-06, "loss": 1.6452, "step": 30022 }, { "epoch": 2.3528996865203764, "grad_norm": 3.2052135467529297, "learning_rate": 3.6440591321622326e-06, "loss": 1.4243, "step": 30023 }, { "epoch": 2.3529780564263323, "grad_norm": 2.4574527740478516, "learning_rate": 3.643762328134264e-06, "loss": 1.6797, "step": 30024 }, { "epoch": 2.353056426332288, "grad_norm": 2.679011106491089, "learning_rate": 3.643465529264727e-06, "loss": 1.5115, "step": 30025 }, { "epoch": 2.3531347962382445, "grad_norm": 3.384798288345337, "learning_rate": 3.6431687355547497e-06, "loss": 1.5193, "step": 30026 }, { "epoch": 2.353213166144201, "grad_norm": 6.2799224853515625, "learning_rate": 3.642871947005461e-06, "loss": 1.6549, "step": 30027 }, { "epoch": 2.3532915360501567, "grad_norm": 9.475863456726074, "learning_rate": 3.642575163617991e-06, "loss": 1.3958, "step": 30028 }, { "epoch": 2.3533699059561126, "grad_norm": 2.871459484100342, "learning_rate": 3.6422783853934698e-06, "loss": 1.0746, "step": 30029 }, { "epoch": 2.353448275862069, "grad_norm": 1.9994505643844604, "learning_rate": 3.641981612333021e-06, "loss": 1.2132, "step": 30030 }, { "epoch": 2.3535266457680253, "grad_norm": 5.860929012298584, "learning_rate": 3.6416848444377774e-06, "loss": 1.9393, "step": 30031 }, { "epoch": 2.353605015673981, "grad_norm": 3.7282912731170654, "learning_rate": 3.6413880817088677e-06, "loss": 1.0724, "step": 30032 }, { "epoch": 2.353683385579937, "grad_norm": 2.705589532852173, "learning_rate": 3.6410913241474206e-06, "loss": 1.5059, "step": 30033 }, { "epoch": 2.3537617554858934, "grad_norm": 3.278208017349243, "learning_rate": 3.6407945717545635e-06, "loss": 1.5452, "step": 30034 }, { "epoch": 2.3538401253918497, "grad_norm": 4.016155242919922, "learning_rate": 3.6404978245314237e-06, "loss": 1.293, "step": 30035 }, { "epoch": 2.3539184952978056, "grad_norm": 3.486337661743164, "learning_rate": 3.6402010824791354e-06, "loss": 1.6641, "step": 30036 }, { "epoch": 2.353996865203762, "grad_norm": 3.589465618133545, "learning_rate": 3.6399043455988214e-06, "loss": 1.2564, "step": 30037 }, { "epoch": 2.354075235109718, "grad_norm": 4.434989929199219, "learning_rate": 3.639607613891613e-06, "loss": 1.3603, "step": 30038 }, { "epoch": 2.354153605015674, "grad_norm": 2.9404661655426025, "learning_rate": 3.639310887358639e-06, "loss": 1.7178, "step": 30039 }, { "epoch": 2.35423197492163, "grad_norm": 2.6254329681396484, "learning_rate": 3.639014166001028e-06, "loss": 1.2272, "step": 30040 }, { "epoch": 2.3543103448275864, "grad_norm": 5.0999860763549805, "learning_rate": 3.6387174498199072e-06, "loss": 1.6735, "step": 30041 }, { "epoch": 2.3543887147335423, "grad_norm": 3.04669451713562, "learning_rate": 3.638420738816405e-06, "loss": 1.4553, "step": 30042 }, { "epoch": 2.3544670846394986, "grad_norm": 4.771539688110352, "learning_rate": 3.6381240329916523e-06, "loss": 1.2062, "step": 30043 }, { "epoch": 2.3545454545454545, "grad_norm": 3.868821859359741, "learning_rate": 3.637827332346777e-06, "loss": 1.365, "step": 30044 }, { "epoch": 2.354623824451411, "grad_norm": 5.8449387550354, "learning_rate": 3.637530636882906e-06, "loss": 1.2558, "step": 30045 }, { "epoch": 2.3547021943573667, "grad_norm": 3.870408535003662, "learning_rate": 3.6372339466011687e-06, "loss": 1.5228, "step": 30046 }, { "epoch": 2.354780564263323, "grad_norm": 2.356308698654175, "learning_rate": 3.6369372615026923e-06, "loss": 1.0893, "step": 30047 }, { "epoch": 2.354858934169279, "grad_norm": 2.433699607849121, "learning_rate": 3.6366405815886097e-06, "loss": 1.542, "step": 30048 }, { "epoch": 2.3549373040752353, "grad_norm": 5.931626319885254, "learning_rate": 3.6363439068600426e-06, "loss": 1.3873, "step": 30049 }, { "epoch": 2.355015673981191, "grad_norm": 4.334619522094727, "learning_rate": 3.6360472373181245e-06, "loss": 1.6224, "step": 30050 }, { "epoch": 2.3550940438871475, "grad_norm": 2.9975926876068115, "learning_rate": 3.6357505729639818e-06, "loss": 1.2641, "step": 30051 }, { "epoch": 2.3551724137931034, "grad_norm": 3.066248655319214, "learning_rate": 3.6354539137987434e-06, "loss": 2.0176, "step": 30052 }, { "epoch": 2.3552507836990597, "grad_norm": 4.292989730834961, "learning_rate": 3.6351572598235375e-06, "loss": 1.7135, "step": 30053 }, { "epoch": 2.3553291536050156, "grad_norm": 3.4767954349517822, "learning_rate": 3.6348606110394913e-06, "loss": 0.8723, "step": 30054 }, { "epoch": 2.355407523510972, "grad_norm": 3.339304208755493, "learning_rate": 3.6345639674477367e-06, "loss": 1.6612, "step": 30055 }, { "epoch": 2.355485893416928, "grad_norm": 3.1770310401916504, "learning_rate": 3.634267329049397e-06, "loss": 0.8381, "step": 30056 }, { "epoch": 2.355564263322884, "grad_norm": 4.092693328857422, "learning_rate": 3.6339706958456035e-06, "loss": 1.5507, "step": 30057 }, { "epoch": 2.35564263322884, "grad_norm": 1.9823211431503296, "learning_rate": 3.6336740678374836e-06, "loss": 1.0335, "step": 30058 }, { "epoch": 2.3557210031347964, "grad_norm": 6.028229713439941, "learning_rate": 3.633377445026167e-06, "loss": 1.4287, "step": 30059 }, { "epoch": 2.3557993730407523, "grad_norm": 3.566087007522583, "learning_rate": 3.6330808274127795e-06, "loss": 1.0433, "step": 30060 }, { "epoch": 2.3558777429467086, "grad_norm": 6.647920608520508, "learning_rate": 3.6327842149984493e-06, "loss": 1.222, "step": 30061 }, { "epoch": 2.3559561128526645, "grad_norm": 4.049575328826904, "learning_rate": 3.6324876077843097e-06, "loss": 1.5507, "step": 30062 }, { "epoch": 2.356034482758621, "grad_norm": 3.987943172454834, "learning_rate": 3.6321910057714806e-06, "loss": 1.4096, "step": 30063 }, { "epoch": 2.3561128526645767, "grad_norm": 3.4368579387664795, "learning_rate": 3.6318944089610965e-06, "loss": 1.2371, "step": 30064 }, { "epoch": 2.356191222570533, "grad_norm": 3.9470181465148926, "learning_rate": 3.6315978173542822e-06, "loss": 1.8333, "step": 30065 }, { "epoch": 2.356269592476489, "grad_norm": 5.243988037109375, "learning_rate": 3.631301230952169e-06, "loss": 1.0914, "step": 30066 }, { "epoch": 2.3563479623824453, "grad_norm": 4.1330037117004395, "learning_rate": 3.6310046497558815e-06, "loss": 1.4333, "step": 30067 }, { "epoch": 2.356426332288401, "grad_norm": 2.602477788925171, "learning_rate": 3.630708073766548e-06, "loss": 1.4899, "step": 30068 }, { "epoch": 2.3565047021943575, "grad_norm": 9.65738582611084, "learning_rate": 3.6304115029852984e-06, "loss": 0.903, "step": 30069 }, { "epoch": 2.3565830721003134, "grad_norm": 2.543942451477051, "learning_rate": 3.6301149374132615e-06, "loss": 1.4858, "step": 30070 }, { "epoch": 2.3566614420062697, "grad_norm": 2.995784282684326, "learning_rate": 3.6298183770515627e-06, "loss": 1.3932, "step": 30071 }, { "epoch": 2.3567398119122256, "grad_norm": 2.4191837310791016, "learning_rate": 3.6295218219013307e-06, "loss": 1.1217, "step": 30072 }, { "epoch": 2.356818181818182, "grad_norm": 2.5126585960388184, "learning_rate": 3.6292252719636933e-06, "loss": 1.0993, "step": 30073 }, { "epoch": 2.356896551724138, "grad_norm": 3.355656623840332, "learning_rate": 3.6289287272397817e-06, "loss": 1.237, "step": 30074 }, { "epoch": 2.356974921630094, "grad_norm": 6.867888450622559, "learning_rate": 3.6286321877307174e-06, "loss": 1.3187, "step": 30075 }, { "epoch": 2.35705329153605, "grad_norm": 3.8080883026123047, "learning_rate": 3.6283356534376337e-06, "loss": 1.0372, "step": 30076 }, { "epoch": 2.3571316614420064, "grad_norm": 4.893381118774414, "learning_rate": 3.6280391243616575e-06, "loss": 0.9629, "step": 30077 }, { "epoch": 2.3572100313479623, "grad_norm": 2.7409160137176514, "learning_rate": 3.627742600503914e-06, "loss": 1.5383, "step": 30078 }, { "epoch": 2.3572884012539186, "grad_norm": 2.470884084701538, "learning_rate": 3.627446081865534e-06, "loss": 0.9693, "step": 30079 }, { "epoch": 2.3573667711598745, "grad_norm": 4.268174648284912, "learning_rate": 3.6271495684476423e-06, "loss": 1.3753, "step": 30080 }, { "epoch": 2.357445141065831, "grad_norm": 3.6865336894989014, "learning_rate": 3.6268530602513718e-06, "loss": 1.3662, "step": 30081 }, { "epoch": 2.3575235109717867, "grad_norm": 4.3627214431762695, "learning_rate": 3.6265565572778437e-06, "loss": 1.6692, "step": 30082 }, { "epoch": 2.357601880877743, "grad_norm": 4.443047523498535, "learning_rate": 3.6262600595281907e-06, "loss": 1.0711, "step": 30083 }, { "epoch": 2.357680250783699, "grad_norm": 2.357156276702881, "learning_rate": 3.6259635670035374e-06, "loss": 1.4177, "step": 30084 }, { "epoch": 2.3577586206896552, "grad_norm": 7.741061210632324, "learning_rate": 3.625667079705015e-06, "loss": 1.1626, "step": 30085 }, { "epoch": 2.357836990595611, "grad_norm": 2.9946248531341553, "learning_rate": 3.6253705976337484e-06, "loss": 0.8514, "step": 30086 }, { "epoch": 2.3579153605015675, "grad_norm": 2.4166958332061768, "learning_rate": 3.6250741207908645e-06, "loss": 1.0923, "step": 30087 }, { "epoch": 2.3579937304075234, "grad_norm": 2.7893483638763428, "learning_rate": 3.6247776491774934e-06, "loss": 1.2327, "step": 30088 }, { "epoch": 2.3580721003134797, "grad_norm": 3.803018808364868, "learning_rate": 3.6244811827947633e-06, "loss": 1.1318, "step": 30089 }, { "epoch": 2.3581504702194356, "grad_norm": 7.226773262023926, "learning_rate": 3.6241847216437987e-06, "loss": 1.6358, "step": 30090 }, { "epoch": 2.358228840125392, "grad_norm": 6.211609363555908, "learning_rate": 3.623888265725729e-06, "loss": 1.2394, "step": 30091 }, { "epoch": 2.358307210031348, "grad_norm": 3.1048238277435303, "learning_rate": 3.6235918150416825e-06, "loss": 1.2594, "step": 30092 }, { "epoch": 2.358385579937304, "grad_norm": 3.141342878341675, "learning_rate": 3.6232953695927847e-06, "loss": 1.3023, "step": 30093 }, { "epoch": 2.35846394984326, "grad_norm": 2.6080307960510254, "learning_rate": 3.6229989293801633e-06, "loss": 0.9967, "step": 30094 }, { "epoch": 2.3585423197492164, "grad_norm": 2.5192034244537354, "learning_rate": 3.622702494404947e-06, "loss": 1.3198, "step": 30095 }, { "epoch": 2.3586206896551722, "grad_norm": 2.977236032485962, "learning_rate": 3.622406064668265e-06, "loss": 1.3797, "step": 30096 }, { "epoch": 2.3586990595611286, "grad_norm": 4.051522254943848, "learning_rate": 3.6221096401712415e-06, "loss": 1.4004, "step": 30097 }, { "epoch": 2.3587774294670845, "grad_norm": 3.5552046298980713, "learning_rate": 3.6218132209150047e-06, "loss": 1.2698, "step": 30098 }, { "epoch": 2.358855799373041, "grad_norm": 2.7699766159057617, "learning_rate": 3.6215168069006825e-06, "loss": 1.4788, "step": 30099 }, { "epoch": 2.3589341692789967, "grad_norm": 4.972620487213135, "learning_rate": 3.6212203981294036e-06, "loss": 1.3414, "step": 30100 }, { "epoch": 2.359012539184953, "grad_norm": 2.586975336074829, "learning_rate": 3.620923994602293e-06, "loss": 1.1748, "step": 30101 }, { "epoch": 2.359090909090909, "grad_norm": 8.157524108886719, "learning_rate": 3.6206275963204794e-06, "loss": 1.1901, "step": 30102 }, { "epoch": 2.3591692789968652, "grad_norm": 2.213918924331665, "learning_rate": 3.62033120328509e-06, "loss": 1.3289, "step": 30103 }, { "epoch": 2.359247648902821, "grad_norm": 9.39389705657959, "learning_rate": 3.620034815497253e-06, "loss": 1.5197, "step": 30104 }, { "epoch": 2.3593260188087775, "grad_norm": 2.985011100769043, "learning_rate": 3.6197384329580924e-06, "loss": 1.2713, "step": 30105 }, { "epoch": 2.359404388714734, "grad_norm": 3.480595827102661, "learning_rate": 3.6194420556687395e-06, "loss": 1.2304, "step": 30106 }, { "epoch": 2.3594827586206897, "grad_norm": 3.886850357055664, "learning_rate": 3.619145683630321e-06, "loss": 1.2097, "step": 30107 }, { "epoch": 2.3595611285266456, "grad_norm": 5.904870986938477, "learning_rate": 3.6188493168439613e-06, "loss": 1.3871, "step": 30108 }, { "epoch": 2.359639498432602, "grad_norm": 3.157865524291992, "learning_rate": 3.61855295531079e-06, "loss": 1.5819, "step": 30109 }, { "epoch": 2.3597178683385582, "grad_norm": 2.947880268096924, "learning_rate": 3.6182565990319335e-06, "loss": 1.1951, "step": 30110 }, { "epoch": 2.359796238244514, "grad_norm": 2.441202163696289, "learning_rate": 3.61796024800852e-06, "loss": 1.3033, "step": 30111 }, { "epoch": 2.35987460815047, "grad_norm": 3.5154244899749756, "learning_rate": 3.6176639022416736e-06, "loss": 1.7613, "step": 30112 }, { "epoch": 2.3599529780564263, "grad_norm": 21.531749725341797, "learning_rate": 3.6173675617325254e-06, "loss": 0.973, "step": 30113 }, { "epoch": 2.3600313479623827, "grad_norm": 3.230623722076416, "learning_rate": 3.6170712264822007e-06, "loss": 1.1526, "step": 30114 }, { "epoch": 2.3601097178683386, "grad_norm": 3.064237356185913, "learning_rate": 3.6167748964918282e-06, "loss": 1.4847, "step": 30115 }, { "epoch": 2.3601880877742945, "grad_norm": 2.5661869049072266, "learning_rate": 3.6164785717625316e-06, "loss": 1.3825, "step": 30116 }, { "epoch": 2.360266457680251, "grad_norm": 2.589245557785034, "learning_rate": 3.6161822522954394e-06, "loss": 1.2134, "step": 30117 }, { "epoch": 2.360344827586207, "grad_norm": 2.6409802436828613, "learning_rate": 3.615885938091682e-06, "loss": 1.3644, "step": 30118 }, { "epoch": 2.360423197492163, "grad_norm": 3.5936973094940186, "learning_rate": 3.6155896291523805e-06, "loss": 1.0958, "step": 30119 }, { "epoch": 2.360501567398119, "grad_norm": 5.346283912658691, "learning_rate": 3.615293325478666e-06, "loss": 1.2142, "step": 30120 }, { "epoch": 2.3605799373040752, "grad_norm": 5.512710094451904, "learning_rate": 3.6149970270716645e-06, "loss": 1.4232, "step": 30121 }, { "epoch": 2.3606583072100316, "grad_norm": 3.770122766494751, "learning_rate": 3.6147007339325035e-06, "loss": 1.3463, "step": 30122 }, { "epoch": 2.3607366771159874, "grad_norm": 3.1384899616241455, "learning_rate": 3.614404446062309e-06, "loss": 1.2893, "step": 30123 }, { "epoch": 2.3608150470219433, "grad_norm": 5.456295967102051, "learning_rate": 3.614108163462207e-06, "loss": 1.1911, "step": 30124 }, { "epoch": 2.3608934169278997, "grad_norm": 1.8727667331695557, "learning_rate": 3.613811886133327e-06, "loss": 1.3073, "step": 30125 }, { "epoch": 2.360971786833856, "grad_norm": 3.0202529430389404, "learning_rate": 3.6135156140767957e-06, "loss": 1.4995, "step": 30126 }, { "epoch": 2.361050156739812, "grad_norm": 3.107661485671997, "learning_rate": 3.613219347293737e-06, "loss": 1.8086, "step": 30127 }, { "epoch": 2.3611285266457678, "grad_norm": 3.1996514797210693, "learning_rate": 3.6129230857852804e-06, "loss": 1.665, "step": 30128 }, { "epoch": 2.361206896551724, "grad_norm": 4.050694465637207, "learning_rate": 3.6126268295525503e-06, "loss": 0.8649, "step": 30129 }, { "epoch": 2.3612852664576804, "grad_norm": 2.6881096363067627, "learning_rate": 3.612330578596679e-06, "loss": 1.3562, "step": 30130 }, { "epoch": 2.3613636363636363, "grad_norm": 3.0100185871124268, "learning_rate": 3.612034332918786e-06, "loss": 1.8993, "step": 30131 }, { "epoch": 2.3614420062695927, "grad_norm": 3.319277763366699, "learning_rate": 3.6117380925200018e-06, "loss": 1.1392, "step": 30132 }, { "epoch": 2.3615203761755486, "grad_norm": 3.7347633838653564, "learning_rate": 3.6114418574014555e-06, "loss": 1.4806, "step": 30133 }, { "epoch": 2.361598746081505, "grad_norm": 6.528247356414795, "learning_rate": 3.611145627564269e-06, "loss": 1.2865, "step": 30134 }, { "epoch": 2.3616771159874608, "grad_norm": 4.569393634796143, "learning_rate": 3.610849403009571e-06, "loss": 1.078, "step": 30135 }, { "epoch": 2.361755485893417, "grad_norm": 4.26575231552124, "learning_rate": 3.6105531837384877e-06, "loss": 1.4632, "step": 30136 }, { "epoch": 2.361833855799373, "grad_norm": 1.9072939157485962, "learning_rate": 3.6102569697521496e-06, "loss": 1.3752, "step": 30137 }, { "epoch": 2.3619122257053293, "grad_norm": 3.1613645553588867, "learning_rate": 3.6099607610516762e-06, "loss": 1.2812, "step": 30138 }, { "epoch": 2.361990595611285, "grad_norm": 3.8369321823120117, "learning_rate": 3.6096645576382004e-06, "loss": 1.6252, "step": 30139 }, { "epoch": 2.3620689655172415, "grad_norm": 3.099205493927002, "learning_rate": 3.609368359512846e-06, "loss": 1.6106, "step": 30140 }, { "epoch": 2.3621473354231974, "grad_norm": 2.47822904586792, "learning_rate": 3.6090721666767407e-06, "loss": 1.6023, "step": 30141 }, { "epoch": 2.3622257053291538, "grad_norm": 4.407249927520752, "learning_rate": 3.60877597913101e-06, "loss": 1.5168, "step": 30142 }, { "epoch": 2.3623040752351097, "grad_norm": 2.9782583713531494, "learning_rate": 3.6084797968767797e-06, "loss": 1.2038, "step": 30143 }, { "epoch": 2.362382445141066, "grad_norm": 2.296882152557373, "learning_rate": 3.608183619915179e-06, "loss": 1.5347, "step": 30144 }, { "epoch": 2.362460815047022, "grad_norm": 2.2894797325134277, "learning_rate": 3.607887448247333e-06, "loss": 1.4457, "step": 30145 }, { "epoch": 2.362539184952978, "grad_norm": 3.7932040691375732, "learning_rate": 3.6075912818743676e-06, "loss": 1.2093, "step": 30146 }, { "epoch": 2.362617554858934, "grad_norm": 4.614785194396973, "learning_rate": 3.6072951207974094e-06, "loss": 1.2732, "step": 30147 }, { "epoch": 2.3626959247648904, "grad_norm": 2.7316036224365234, "learning_rate": 3.6069989650175873e-06, "loss": 1.5116, "step": 30148 }, { "epoch": 2.3627742946708463, "grad_norm": 2.4362547397613525, "learning_rate": 3.606702814536024e-06, "loss": 1.1299, "step": 30149 }, { "epoch": 2.3628526645768027, "grad_norm": 3.069512367248535, "learning_rate": 3.6064066693538457e-06, "loss": 1.279, "step": 30150 }, { "epoch": 2.3629310344827585, "grad_norm": 3.197807788848877, "learning_rate": 3.606110529472183e-06, "loss": 1.5299, "step": 30151 }, { "epoch": 2.363009404388715, "grad_norm": 3.303032398223877, "learning_rate": 3.6058143948921597e-06, "loss": 0.9441, "step": 30152 }, { "epoch": 2.3630877742946708, "grad_norm": 3.6842873096466064, "learning_rate": 3.6055182656149024e-06, "loss": 1.4503, "step": 30153 }, { "epoch": 2.363166144200627, "grad_norm": 2.4169046878814697, "learning_rate": 3.605222141641537e-06, "loss": 1.1695, "step": 30154 }, { "epoch": 2.363244514106583, "grad_norm": 2.4991042613983154, "learning_rate": 3.6049260229731885e-06, "loss": 1.4304, "step": 30155 }, { "epoch": 2.3633228840125393, "grad_norm": 2.4103753566741943, "learning_rate": 3.6046299096109887e-06, "loss": 1.3269, "step": 30156 }, { "epoch": 2.363401253918495, "grad_norm": 2.2594127655029297, "learning_rate": 3.604333801556057e-06, "loss": 1.1695, "step": 30157 }, { "epoch": 2.3634796238244515, "grad_norm": 5.031523704528809, "learning_rate": 3.6040376988095232e-06, "loss": 1.4335, "step": 30158 }, { "epoch": 2.3635579937304074, "grad_norm": 2.912208318710327, "learning_rate": 3.6037416013725136e-06, "loss": 1.2303, "step": 30159 }, { "epoch": 2.3636363636363638, "grad_norm": 3.1707208156585693, "learning_rate": 3.603445509246154e-06, "loss": 1.6679, "step": 30160 }, { "epoch": 2.3637147335423196, "grad_norm": 2.946270227432251, "learning_rate": 3.60314942243157e-06, "loss": 1.3427, "step": 30161 }, { "epoch": 2.363793103448276, "grad_norm": 4.526852130889893, "learning_rate": 3.6028533409298867e-06, "loss": 1.5048, "step": 30162 }, { "epoch": 2.363871473354232, "grad_norm": 2.5342793464660645, "learning_rate": 3.6025572647422347e-06, "loss": 1.5981, "step": 30163 }, { "epoch": 2.363949843260188, "grad_norm": 2.8286123275756836, "learning_rate": 3.6022611938697342e-06, "loss": 1.3962, "step": 30164 }, { "epoch": 2.364028213166144, "grad_norm": 3.976271867752075, "learning_rate": 3.6019651283135154e-06, "loss": 1.3528, "step": 30165 }, { "epoch": 2.3641065830721004, "grad_norm": 2.9699347019195557, "learning_rate": 3.601669068074703e-06, "loss": 1.1661, "step": 30166 }, { "epoch": 2.3641849529780563, "grad_norm": 4.923223972320557, "learning_rate": 3.6013730131544245e-06, "loss": 1.4497, "step": 30167 }, { "epoch": 2.3642633228840126, "grad_norm": 2.3516368865966797, "learning_rate": 3.6010769635538034e-06, "loss": 1.4847, "step": 30168 }, { "epoch": 2.3643416927899685, "grad_norm": 3.299591064453125, "learning_rate": 3.6007809192739663e-06, "loss": 1.5363, "step": 30169 }, { "epoch": 2.364420062695925, "grad_norm": 4.947665214538574, "learning_rate": 3.60048488031604e-06, "loss": 1.3368, "step": 30170 }, { "epoch": 2.3644984326018808, "grad_norm": 4.696240425109863, "learning_rate": 3.600188846681152e-06, "loss": 1.5692, "step": 30171 }, { "epoch": 2.364576802507837, "grad_norm": 3.0936226844787598, "learning_rate": 3.5998928183704258e-06, "loss": 1.1392, "step": 30172 }, { "epoch": 2.364655172413793, "grad_norm": 6.669469356536865, "learning_rate": 3.599596795384988e-06, "loss": 1.1789, "step": 30173 }, { "epoch": 2.3647335423197493, "grad_norm": 3.426708936691284, "learning_rate": 3.599300777725966e-06, "loss": 0.9015, "step": 30174 }, { "epoch": 2.364811912225705, "grad_norm": 3.2299466133117676, "learning_rate": 3.5990047653944827e-06, "loss": 1.3855, "step": 30175 }, { "epoch": 2.3648902821316615, "grad_norm": 2.905747652053833, "learning_rate": 3.5987087583916652e-06, "loss": 1.4389, "step": 30176 }, { "epoch": 2.3649686520376174, "grad_norm": 2.1051273345947266, "learning_rate": 3.5984127567186412e-06, "loss": 1.3043, "step": 30177 }, { "epoch": 2.3650470219435737, "grad_norm": 2.6358022689819336, "learning_rate": 3.5981167603765355e-06, "loss": 0.8948, "step": 30178 }, { "epoch": 2.3651253918495296, "grad_norm": 3.090449333190918, "learning_rate": 3.5978207693664726e-06, "loss": 1.2349, "step": 30179 }, { "epoch": 2.365203761755486, "grad_norm": 2.210599422454834, "learning_rate": 3.5975247836895793e-06, "loss": 1.4991, "step": 30180 }, { "epoch": 2.365282131661442, "grad_norm": 2.472646713256836, "learning_rate": 3.5972288033469803e-06, "loss": 1.1943, "step": 30181 }, { "epoch": 2.365360501567398, "grad_norm": 2.824244260787964, "learning_rate": 3.5969328283398062e-06, "loss": 1.1718, "step": 30182 }, { "epoch": 2.365438871473354, "grad_norm": 3.3796749114990234, "learning_rate": 3.596636858669175e-06, "loss": 1.4331, "step": 30183 }, { "epoch": 2.3655172413793104, "grad_norm": 3.3471224308013916, "learning_rate": 3.5963408943362175e-06, "loss": 1.6448, "step": 30184 }, { "epoch": 2.3655956112852663, "grad_norm": 2.7969675064086914, "learning_rate": 3.596044935342059e-06, "loss": 1.3782, "step": 30185 }, { "epoch": 2.3656739811912226, "grad_norm": 4.515508651733398, "learning_rate": 3.5957489816878253e-06, "loss": 1.158, "step": 30186 }, { "epoch": 2.3657523510971785, "grad_norm": 4.704625129699707, "learning_rate": 3.5954530333746396e-06, "loss": 2.2966, "step": 30187 }, { "epoch": 2.365830721003135, "grad_norm": 3.8462252616882324, "learning_rate": 3.595157090403629e-06, "loss": 1.1908, "step": 30188 }, { "epoch": 2.3659090909090907, "grad_norm": 3.4879138469696045, "learning_rate": 3.5948611527759213e-06, "loss": 1.0082, "step": 30189 }, { "epoch": 2.365987460815047, "grad_norm": 3.214947462081909, "learning_rate": 3.5945652204926372e-06, "loss": 1.1647, "step": 30190 }, { "epoch": 2.366065830721003, "grad_norm": 5.507809162139893, "learning_rate": 3.594269293554907e-06, "loss": 0.739, "step": 30191 }, { "epoch": 2.3661442006269593, "grad_norm": 4.212677478790283, "learning_rate": 3.5939733719638535e-06, "loss": 1.4123, "step": 30192 }, { "epoch": 2.366222570532915, "grad_norm": 3.731374502182007, "learning_rate": 3.5936774557206044e-06, "loss": 1.3305, "step": 30193 }, { "epoch": 2.3663009404388715, "grad_norm": 8.787705421447754, "learning_rate": 3.5933815448262833e-06, "loss": 1.5957, "step": 30194 }, { "epoch": 2.3663793103448274, "grad_norm": 3.8942761421203613, "learning_rate": 3.5930856392820146e-06, "loss": 1.624, "step": 30195 }, { "epoch": 2.3664576802507837, "grad_norm": 2.391719341278076, "learning_rate": 3.5927897390889265e-06, "loss": 1.7221, "step": 30196 }, { "epoch": 2.3665360501567396, "grad_norm": 8.01929759979248, "learning_rate": 3.592493844248145e-06, "loss": 0.872, "step": 30197 }, { "epoch": 2.366614420062696, "grad_norm": 5.460190773010254, "learning_rate": 3.592197954760793e-06, "loss": 2.3125, "step": 30198 }, { "epoch": 2.366692789968652, "grad_norm": 3.4738893508911133, "learning_rate": 3.591902070627996e-06, "loss": 1.1525, "step": 30199 }, { "epoch": 2.366771159874608, "grad_norm": 4.965731620788574, "learning_rate": 3.5916061918508803e-06, "loss": 1.5218, "step": 30200 }, { "epoch": 2.3668495297805645, "grad_norm": 2.642432689666748, "learning_rate": 3.5913103184305733e-06, "loss": 1.1667, "step": 30201 }, { "epoch": 2.3669278996865204, "grad_norm": 11.515515327453613, "learning_rate": 3.5910144503681964e-06, "loss": 1.1622, "step": 30202 }, { "epoch": 2.3670062695924763, "grad_norm": 2.73822283744812, "learning_rate": 3.5907185876648776e-06, "loss": 1.3801, "step": 30203 }, { "epoch": 2.3670846394984326, "grad_norm": 2.6266095638275146, "learning_rate": 3.590422730321742e-06, "loss": 1.1532, "step": 30204 }, { "epoch": 2.367163009404389, "grad_norm": 1.872551679611206, "learning_rate": 3.5901268783399135e-06, "loss": 1.0198, "step": 30205 }, { "epoch": 2.367241379310345, "grad_norm": 6.037437438964844, "learning_rate": 3.5898310317205175e-06, "loss": 1.0648, "step": 30206 }, { "epoch": 2.3673197492163007, "grad_norm": 2.607496500015259, "learning_rate": 3.5895351904646802e-06, "loss": 1.4454, "step": 30207 }, { "epoch": 2.367398119122257, "grad_norm": 5.296163558959961, "learning_rate": 3.5892393545735282e-06, "loss": 1.6905, "step": 30208 }, { "epoch": 2.3674764890282134, "grad_norm": 4.448031902313232, "learning_rate": 3.5889435240481834e-06, "loss": 1.3575, "step": 30209 }, { "epoch": 2.3675548589341693, "grad_norm": 3.2282016277313232, "learning_rate": 3.5886476988897732e-06, "loss": 1.166, "step": 30210 }, { "epoch": 2.367633228840125, "grad_norm": 9.005487442016602, "learning_rate": 3.588351879099422e-06, "loss": 1.2465, "step": 30211 }, { "epoch": 2.3677115987460815, "grad_norm": 3.2431352138519287, "learning_rate": 3.588056064678257e-06, "loss": 1.347, "step": 30212 }, { "epoch": 2.367789968652038, "grad_norm": 1.7869899272918701, "learning_rate": 3.5877602556273983e-06, "loss": 1.255, "step": 30213 }, { "epoch": 2.3678683385579937, "grad_norm": 3.5186517238616943, "learning_rate": 3.5874644519479763e-06, "loss": 1.526, "step": 30214 }, { "epoch": 2.3679467084639496, "grad_norm": 2.961972713470459, "learning_rate": 3.5871686536411144e-06, "loss": 1.1934, "step": 30215 }, { "epoch": 2.368025078369906, "grad_norm": 2.975645065307617, "learning_rate": 3.5868728607079363e-06, "loss": 1.2726, "step": 30216 }, { "epoch": 2.3681034482758623, "grad_norm": 3.273362398147583, "learning_rate": 3.5865770731495676e-06, "loss": 1.5889, "step": 30217 }, { "epoch": 2.368181818181818, "grad_norm": 5.875467300415039, "learning_rate": 3.586281290967133e-06, "loss": 1.5639, "step": 30218 }, { "epoch": 2.368260188087774, "grad_norm": 3.5354652404785156, "learning_rate": 3.5859855141617615e-06, "loss": 1.287, "step": 30219 }, { "epoch": 2.3683385579937304, "grad_norm": 2.2815845012664795, "learning_rate": 3.585689742734572e-06, "loss": 1.154, "step": 30220 }, { "epoch": 2.3684169278996867, "grad_norm": 3.2305421829223633, "learning_rate": 3.585393976686693e-06, "loss": 1.2186, "step": 30221 }, { "epoch": 2.3684952978056426, "grad_norm": 4.951864719390869, "learning_rate": 3.585098216019249e-06, "loss": 1.3834, "step": 30222 }, { "epoch": 2.368573667711599, "grad_norm": 4.877384185791016, "learning_rate": 3.5848024607333654e-06, "loss": 1.6257, "step": 30223 }, { "epoch": 2.368652037617555, "grad_norm": 4.8314056396484375, "learning_rate": 3.5845067108301653e-06, "loss": 1.2426, "step": 30224 }, { "epoch": 2.368730407523511, "grad_norm": 4.440887928009033, "learning_rate": 3.584210966310774e-06, "loss": 1.1087, "step": 30225 }, { "epoch": 2.368808777429467, "grad_norm": 5.602461814880371, "learning_rate": 3.583915227176317e-06, "loss": 1.6493, "step": 30226 }, { "epoch": 2.3688871473354234, "grad_norm": 3.5304086208343506, "learning_rate": 3.5836194934279212e-06, "loss": 1.284, "step": 30227 }, { "epoch": 2.3689655172413793, "grad_norm": 3.696277618408203, "learning_rate": 3.5833237650667074e-06, "loss": 0.9641, "step": 30228 }, { "epoch": 2.3690438871473356, "grad_norm": 4.144545078277588, "learning_rate": 3.5830280420938035e-06, "loss": 1.4653, "step": 30229 }, { "epoch": 2.3691222570532915, "grad_norm": 3.3317317962646484, "learning_rate": 3.582732324510333e-06, "loss": 1.5292, "step": 30230 }, { "epoch": 2.369200626959248, "grad_norm": 2.2209651470184326, "learning_rate": 3.5824366123174205e-06, "loss": 1.3523, "step": 30231 }, { "epoch": 2.3692789968652037, "grad_norm": 5.651923656463623, "learning_rate": 3.58214090551619e-06, "loss": 1.5228, "step": 30232 }, { "epoch": 2.36935736677116, "grad_norm": 5.302579879760742, "learning_rate": 3.581845204107768e-06, "loss": 0.9498, "step": 30233 }, { "epoch": 2.369435736677116, "grad_norm": 15.836118698120117, "learning_rate": 3.5815495080932795e-06, "loss": 1.4603, "step": 30234 }, { "epoch": 2.3695141065830723, "grad_norm": 2.575610399246216, "learning_rate": 3.581253817473847e-06, "loss": 1.1168, "step": 30235 }, { "epoch": 2.369592476489028, "grad_norm": 4.586666107177734, "learning_rate": 3.580958132250596e-06, "loss": 1.5625, "step": 30236 }, { "epoch": 2.3696708463949845, "grad_norm": 3.032837390899658, "learning_rate": 3.5806624524246515e-06, "loss": 1.5601, "step": 30237 }, { "epoch": 2.3697492163009404, "grad_norm": 2.8839707374572754, "learning_rate": 3.5803667779971396e-06, "loss": 1.348, "step": 30238 }, { "epoch": 2.3698275862068967, "grad_norm": 4.010706901550293, "learning_rate": 3.5800711089691807e-06, "loss": 0.9795, "step": 30239 }, { "epoch": 2.3699059561128526, "grad_norm": 2.620102643966675, "learning_rate": 3.5797754453419038e-06, "loss": 1.2195, "step": 30240 }, { "epoch": 2.369984326018809, "grad_norm": 4.050404071807861, "learning_rate": 3.579479787116431e-06, "loss": 1.4677, "step": 30241 }, { "epoch": 2.370062695924765, "grad_norm": 2.5285022258758545, "learning_rate": 3.579184134293889e-06, "loss": 1.3817, "step": 30242 }, { "epoch": 2.370141065830721, "grad_norm": 2.80625057220459, "learning_rate": 3.5788884868753993e-06, "loss": 1.3705, "step": 30243 }, { "epoch": 2.370219435736677, "grad_norm": 2.346078395843506, "learning_rate": 3.5785928448620866e-06, "loss": 0.8819, "step": 30244 }, { "epoch": 2.3702978056426334, "grad_norm": 3.0122718811035156, "learning_rate": 3.57829720825508e-06, "loss": 1.4823, "step": 30245 }, { "epoch": 2.3703761755485893, "grad_norm": 5.092467308044434, "learning_rate": 3.5780015770554976e-06, "loss": 1.2711, "step": 30246 }, { "epoch": 2.3704545454545456, "grad_norm": 4.726291656494141, "learning_rate": 3.5777059512644676e-06, "loss": 1.3914, "step": 30247 }, { "epoch": 2.3705329153605015, "grad_norm": 4.068894386291504, "learning_rate": 3.577410330883114e-06, "loss": 1.8483, "step": 30248 }, { "epoch": 2.370611285266458, "grad_norm": 2.8041038513183594, "learning_rate": 3.5771147159125613e-06, "loss": 1.3062, "step": 30249 }, { "epoch": 2.3706896551724137, "grad_norm": 8.247345924377441, "learning_rate": 3.5768191063539326e-06, "loss": 0.8185, "step": 30250 }, { "epoch": 2.37076802507837, "grad_norm": 3.444814443588257, "learning_rate": 3.576523502208352e-06, "loss": 1.2591, "step": 30251 }, { "epoch": 2.370846394984326, "grad_norm": 3.0751287937164307, "learning_rate": 3.576227903476946e-06, "loss": 1.2418, "step": 30252 }, { "epoch": 2.3709247648902823, "grad_norm": 10.967299461364746, "learning_rate": 3.575932310160839e-06, "loss": 1.5371, "step": 30253 }, { "epoch": 2.371003134796238, "grad_norm": 3.89086651802063, "learning_rate": 3.5756367222611527e-06, "loss": 1.5233, "step": 30254 }, { "epoch": 2.3710815047021945, "grad_norm": 2.8261470794677734, "learning_rate": 3.575341139779013e-06, "loss": 1.4602, "step": 30255 }, { "epoch": 2.3711598746081504, "grad_norm": 2.784839630126953, "learning_rate": 3.575045562715542e-06, "loss": 1.401, "step": 30256 }, { "epoch": 2.3712382445141067, "grad_norm": 2.350388526916504, "learning_rate": 3.5747499910718695e-06, "loss": 1.1666, "step": 30257 }, { "epoch": 2.3713166144200626, "grad_norm": 2.4081757068634033, "learning_rate": 3.574454424849113e-06, "loss": 1.3708, "step": 30258 }, { "epoch": 2.371394984326019, "grad_norm": 3.212218761444092, "learning_rate": 3.5741588640484006e-06, "loss": 1.156, "step": 30259 }, { "epoch": 2.371473354231975, "grad_norm": 3.110792398452759, "learning_rate": 3.573863308670856e-06, "loss": 1.0882, "step": 30260 }, { "epoch": 2.371551724137931, "grad_norm": 2.7532362937927246, "learning_rate": 3.5735677587176026e-06, "loss": 1.5683, "step": 30261 }, { "epoch": 2.371630094043887, "grad_norm": 3.0795228481292725, "learning_rate": 3.5732722141897646e-06, "loss": 1.4018, "step": 30262 }, { "epoch": 2.3717084639498434, "grad_norm": 3.5540425777435303, "learning_rate": 3.5729766750884655e-06, "loss": 1.4168, "step": 30263 }, { "epoch": 2.3717868338557992, "grad_norm": 3.373183488845825, "learning_rate": 3.572681141414832e-06, "loss": 1.1006, "step": 30264 }, { "epoch": 2.3718652037617556, "grad_norm": 3.6539113521575928, "learning_rate": 3.572385613169985e-06, "loss": 1.1794, "step": 30265 }, { "epoch": 2.3719435736677115, "grad_norm": 2.801330089569092, "learning_rate": 3.5720900903550494e-06, "loss": 1.4746, "step": 30266 }, { "epoch": 2.372021943573668, "grad_norm": 2.232529640197754, "learning_rate": 3.5717945729711504e-06, "loss": 1.1515, "step": 30267 }, { "epoch": 2.3721003134796237, "grad_norm": 5.512813091278076, "learning_rate": 3.571499061019412e-06, "loss": 1.3727, "step": 30268 }, { "epoch": 2.37217868338558, "grad_norm": 2.3666422367095947, "learning_rate": 3.571203554500957e-06, "loss": 1.6499, "step": 30269 }, { "epoch": 2.372257053291536, "grad_norm": 3.71954345703125, "learning_rate": 3.570908053416908e-06, "loss": 1.2942, "step": 30270 }, { "epoch": 2.3723354231974922, "grad_norm": 6.395730972290039, "learning_rate": 3.5706125577683947e-06, "loss": 1.6429, "step": 30271 }, { "epoch": 2.372413793103448, "grad_norm": 3.5347869396209717, "learning_rate": 3.5703170675565334e-06, "loss": 1.3525, "step": 30272 }, { "epoch": 2.3724921630094045, "grad_norm": 7.864550590515137, "learning_rate": 3.5700215827824525e-06, "loss": 1.2374, "step": 30273 }, { "epoch": 2.3725705329153604, "grad_norm": 4.5935893058776855, "learning_rate": 3.569726103447275e-06, "loss": 1.3914, "step": 30274 }, { "epoch": 2.3726489028213167, "grad_norm": 3.5554261207580566, "learning_rate": 3.5694306295521263e-06, "loss": 1.4381, "step": 30275 }, { "epoch": 2.3727272727272726, "grad_norm": 4.145656585693359, "learning_rate": 3.5691351610981273e-06, "loss": 1.0577, "step": 30276 }, { "epoch": 2.372805642633229, "grad_norm": 2.699007034301758, "learning_rate": 3.5688396980864027e-06, "loss": 1.4201, "step": 30277 }, { "epoch": 2.372884012539185, "grad_norm": 5.169280052185059, "learning_rate": 3.568544240518077e-06, "loss": 1.6745, "step": 30278 }, { "epoch": 2.372962382445141, "grad_norm": 3.0931217670440674, "learning_rate": 3.568248788394275e-06, "loss": 1.4808, "step": 30279 }, { "epoch": 2.373040752351097, "grad_norm": 3.1300318241119385, "learning_rate": 3.567953341716119e-06, "loss": 1.2777, "step": 30280 }, { "epoch": 2.3731191222570533, "grad_norm": 2.549762487411499, "learning_rate": 3.5676579004847323e-06, "loss": 1.3233, "step": 30281 }, { "epoch": 2.3731974921630092, "grad_norm": 3.187978982925415, "learning_rate": 3.5673624647012384e-06, "loss": 1.6944, "step": 30282 }, { "epoch": 2.3732758620689656, "grad_norm": 3.2538702487945557, "learning_rate": 3.5670670343667646e-06, "loss": 1.0103, "step": 30283 }, { "epoch": 2.3733542319749215, "grad_norm": 3.611605167388916, "learning_rate": 3.566771609482429e-06, "loss": 1.5225, "step": 30284 }, { "epoch": 2.373432601880878, "grad_norm": 2.960780620574951, "learning_rate": 3.566476190049359e-06, "loss": 1.5734, "step": 30285 }, { "epoch": 2.3735109717868337, "grad_norm": 4.333707332611084, "learning_rate": 3.5661807760686783e-06, "loss": 1.2526, "step": 30286 }, { "epoch": 2.37358934169279, "grad_norm": 10.848804473876953, "learning_rate": 3.5658853675415086e-06, "loss": 0.9126, "step": 30287 }, { "epoch": 2.373667711598746, "grad_norm": 10.668069839477539, "learning_rate": 3.565589964468974e-06, "loss": 1.9282, "step": 30288 }, { "epoch": 2.3737460815047022, "grad_norm": 2.3800253868103027, "learning_rate": 3.5652945668521977e-06, "loss": 1.2482, "step": 30289 }, { "epoch": 2.373824451410658, "grad_norm": 2.550405740737915, "learning_rate": 3.5649991746923066e-06, "loss": 1.3735, "step": 30290 }, { "epoch": 2.3739028213166145, "grad_norm": 3.630601167678833, "learning_rate": 3.564703787990419e-06, "loss": 1.0097, "step": 30291 }, { "epoch": 2.3739811912225703, "grad_norm": 2.6743109226226807, "learning_rate": 3.5644084067476615e-06, "loss": 1.3754, "step": 30292 }, { "epoch": 2.3740595611285267, "grad_norm": 2.6323752403259277, "learning_rate": 3.564113030965157e-06, "loss": 1.2848, "step": 30293 }, { "epoch": 2.3741379310344826, "grad_norm": 6.540166854858398, "learning_rate": 3.563817660644031e-06, "loss": 2.4614, "step": 30294 }, { "epoch": 2.374216300940439, "grad_norm": 2.6651394367218018, "learning_rate": 3.563522295785403e-06, "loss": 1.3173, "step": 30295 }, { "epoch": 2.3742946708463952, "grad_norm": 3.148015260696411, "learning_rate": 3.5632269363903982e-06, "loss": 1.6623, "step": 30296 }, { "epoch": 2.374373040752351, "grad_norm": 13.00611686706543, "learning_rate": 3.5629315824601407e-06, "loss": 1.4434, "step": 30297 }, { "epoch": 2.374451410658307, "grad_norm": 4.2168426513671875, "learning_rate": 3.562636233995754e-06, "loss": 1.3885, "step": 30298 }, { "epoch": 2.3745297805642633, "grad_norm": 3.8801887035369873, "learning_rate": 3.5623408909983605e-06, "loss": 1.2253, "step": 30299 }, { "epoch": 2.3746081504702197, "grad_norm": 2.4543280601501465, "learning_rate": 3.562045553469083e-06, "loss": 1.2708, "step": 30300 }, { "epoch": 2.3746865203761756, "grad_norm": 3.5806870460510254, "learning_rate": 3.561750221409047e-06, "loss": 1.3079, "step": 30301 }, { "epoch": 2.3747648902821314, "grad_norm": 3.930846929550171, "learning_rate": 3.561454894819373e-06, "loss": 1.0326, "step": 30302 }, { "epoch": 2.374843260188088, "grad_norm": 3.3434042930603027, "learning_rate": 3.5611595737011854e-06, "loss": 1.5345, "step": 30303 }, { "epoch": 2.374921630094044, "grad_norm": 3.680823802947998, "learning_rate": 3.5608642580556084e-06, "loss": 1.88, "step": 30304 }, { "epoch": 2.375, "grad_norm": 3.14203143119812, "learning_rate": 3.5605689478837658e-06, "loss": 1.7092, "step": 30305 }, { "epoch": 2.375078369905956, "grad_norm": 3.618563652038574, "learning_rate": 3.560273643186779e-06, "loss": 1.4775, "step": 30306 }, { "epoch": 2.375156739811912, "grad_norm": 2.8432137966156006, "learning_rate": 3.559978343965771e-06, "loss": 1.2408, "step": 30307 }, { "epoch": 2.3752351097178686, "grad_norm": 5.248673439025879, "learning_rate": 3.5596830502218647e-06, "loss": 0.9364, "step": 30308 }, { "epoch": 2.3753134796238244, "grad_norm": 4.153820991516113, "learning_rate": 3.559387761956187e-06, "loss": 1.2661, "step": 30309 }, { "epoch": 2.3753918495297803, "grad_norm": 3.7317464351654053, "learning_rate": 3.5590924791698567e-06, "loss": 1.1121, "step": 30310 }, { "epoch": 2.3754702194357367, "grad_norm": 2.834488868713379, "learning_rate": 3.558797201863998e-06, "loss": 1.4037, "step": 30311 }, { "epoch": 2.375548589341693, "grad_norm": 2.645073890686035, "learning_rate": 3.558501930039736e-06, "loss": 1.3999, "step": 30312 }, { "epoch": 2.375626959247649, "grad_norm": 2.92138409614563, "learning_rate": 3.558206663698192e-06, "loss": 1.2563, "step": 30313 }, { "epoch": 2.3757053291536048, "grad_norm": 5.070022106170654, "learning_rate": 3.5579114028404883e-06, "loss": 1.934, "step": 30314 }, { "epoch": 2.375783699059561, "grad_norm": 4.370537281036377, "learning_rate": 3.557616147467749e-06, "loss": 1.3105, "step": 30315 }, { "epoch": 2.3758620689655174, "grad_norm": 4.537171363830566, "learning_rate": 3.557320897581098e-06, "loss": 1.2741, "step": 30316 }, { "epoch": 2.3759404388714733, "grad_norm": 3.162235975265503, "learning_rate": 3.557025653181657e-06, "loss": 1.0906, "step": 30317 }, { "epoch": 2.3760188087774297, "grad_norm": 3.3955209255218506, "learning_rate": 3.556730414270549e-06, "loss": 1.361, "step": 30318 }, { "epoch": 2.3760971786833855, "grad_norm": 2.1139087677001953, "learning_rate": 3.556435180848896e-06, "loss": 1.0145, "step": 30319 }, { "epoch": 2.376175548589342, "grad_norm": 8.048105239868164, "learning_rate": 3.5561399529178252e-06, "loss": 1.4932, "step": 30320 }, { "epoch": 2.3762539184952978, "grad_norm": 3.241459846496582, "learning_rate": 3.555844730478454e-06, "loss": 1.6576, "step": 30321 }, { "epoch": 2.376332288401254, "grad_norm": 2.756068706512451, "learning_rate": 3.5555495135319075e-06, "loss": 1.011, "step": 30322 }, { "epoch": 2.37641065830721, "grad_norm": 3.7730371952056885, "learning_rate": 3.5552543020793095e-06, "loss": 1.2624, "step": 30323 }, { "epoch": 2.3764890282131663, "grad_norm": 4.166884422302246, "learning_rate": 3.5549590961217827e-06, "loss": 1.5102, "step": 30324 }, { "epoch": 2.376567398119122, "grad_norm": 4.974363327026367, "learning_rate": 3.5546638956604486e-06, "loss": 1.2008, "step": 30325 }, { "epoch": 2.3766457680250785, "grad_norm": 3.7973837852478027, "learning_rate": 3.5543687006964294e-06, "loss": 0.927, "step": 30326 }, { "epoch": 2.3767241379310344, "grad_norm": 5.208495616912842, "learning_rate": 3.554073511230852e-06, "loss": 1.4349, "step": 30327 }, { "epoch": 2.3768025078369908, "grad_norm": 2.845993995666504, "learning_rate": 3.553778327264834e-06, "loss": 1.4403, "step": 30328 }, { "epoch": 2.3768808777429467, "grad_norm": 3.17629075050354, "learning_rate": 3.5534831487995015e-06, "loss": 1.5085, "step": 30329 }, { "epoch": 2.376959247648903, "grad_norm": 4.225080490112305, "learning_rate": 3.5531879758359756e-06, "loss": 1.3396, "step": 30330 }, { "epoch": 2.377037617554859, "grad_norm": 2.333080768585205, "learning_rate": 3.55289280837538e-06, "loss": 1.5238, "step": 30331 }, { "epoch": 2.377115987460815, "grad_norm": 1.7938003540039062, "learning_rate": 3.552597646418837e-06, "loss": 1.4322, "step": 30332 }, { "epoch": 2.377194357366771, "grad_norm": 6.046757221221924, "learning_rate": 3.552302489967467e-06, "loss": 1.1498, "step": 30333 }, { "epoch": 2.3772727272727274, "grad_norm": 3.514357566833496, "learning_rate": 3.5520073390223965e-06, "loss": 1.665, "step": 30334 }, { "epoch": 2.3773510971786833, "grad_norm": 3.1991443634033203, "learning_rate": 3.551712193584747e-06, "loss": 1.2343, "step": 30335 }, { "epoch": 2.3774294670846396, "grad_norm": 4.409124851226807, "learning_rate": 3.551417053655639e-06, "loss": 0.8185, "step": 30336 }, { "epoch": 2.3775078369905955, "grad_norm": 3.6779818534851074, "learning_rate": 3.551121919236198e-06, "loss": 1.4731, "step": 30337 }, { "epoch": 2.377586206896552, "grad_norm": 5.3002190589904785, "learning_rate": 3.550826790327543e-06, "loss": 1.1833, "step": 30338 }, { "epoch": 2.3776645768025078, "grad_norm": 2.864564895629883, "learning_rate": 3.550531666930801e-06, "loss": 1.3011, "step": 30339 }, { "epoch": 2.377742946708464, "grad_norm": 2.80867600440979, "learning_rate": 3.55023654904709e-06, "loss": 1.3711, "step": 30340 }, { "epoch": 2.37782131661442, "grad_norm": 4.431238174438477, "learning_rate": 3.5499414366775354e-06, "loss": 1.5698, "step": 30341 }, { "epoch": 2.3778996865203763, "grad_norm": 2.4431233406066895, "learning_rate": 3.5496463298232594e-06, "loss": 1.949, "step": 30342 }, { "epoch": 2.377978056426332, "grad_norm": 4.761162281036377, "learning_rate": 3.5493512284853824e-06, "loss": 1.0349, "step": 30343 }, { "epoch": 2.3780564263322885, "grad_norm": 3.126155138015747, "learning_rate": 3.549056132665029e-06, "loss": 1.7612, "step": 30344 }, { "epoch": 2.3781347962382444, "grad_norm": 2.7054638862609863, "learning_rate": 3.54876104236332e-06, "loss": 1.3755, "step": 30345 }, { "epoch": 2.3782131661442008, "grad_norm": 10.489968299865723, "learning_rate": 3.5484659575813806e-06, "loss": 1.6915, "step": 30346 }, { "epoch": 2.3782915360501566, "grad_norm": 5.465768337249756, "learning_rate": 3.5481708783203294e-06, "loss": 1.3257, "step": 30347 }, { "epoch": 2.378369905956113, "grad_norm": 4.7505011558532715, "learning_rate": 3.5478758045812906e-06, "loss": 1.3832, "step": 30348 }, { "epoch": 2.378448275862069, "grad_norm": 6.080373764038086, "learning_rate": 3.547580736365387e-06, "loss": 1.6393, "step": 30349 }, { "epoch": 2.378526645768025, "grad_norm": 3.628112316131592, "learning_rate": 3.5472856736737405e-06, "loss": 0.8557, "step": 30350 }, { "epoch": 2.378605015673981, "grad_norm": 5.347476482391357, "learning_rate": 3.5469906165074726e-06, "loss": 1.6347, "step": 30351 }, { "epoch": 2.3786833855799374, "grad_norm": 7.1130475997924805, "learning_rate": 3.5466955648677046e-06, "loss": 1.4353, "step": 30352 }, { "epoch": 2.3787617554858933, "grad_norm": 3.0696656703948975, "learning_rate": 3.546400518755562e-06, "loss": 1.4898, "step": 30353 }, { "epoch": 2.3788401253918496, "grad_norm": 4.202297687530518, "learning_rate": 3.5461054781721658e-06, "loss": 1.3402, "step": 30354 }, { "epoch": 2.3789184952978055, "grad_norm": 2.639314651489258, "learning_rate": 3.545810443118637e-06, "loss": 1.1948, "step": 30355 }, { "epoch": 2.378996865203762, "grad_norm": 2.5467753410339355, "learning_rate": 3.545515413596098e-06, "loss": 1.4104, "step": 30356 }, { "epoch": 2.3790752351097177, "grad_norm": 6.645265579223633, "learning_rate": 3.5452203896056715e-06, "loss": 1.1727, "step": 30357 }, { "epoch": 2.379153605015674, "grad_norm": 2.1633429527282715, "learning_rate": 3.5449253711484793e-06, "loss": 1.326, "step": 30358 }, { "epoch": 2.37923197492163, "grad_norm": 3.008575677871704, "learning_rate": 3.5446303582256425e-06, "loss": 1.3349, "step": 30359 }, { "epoch": 2.3793103448275863, "grad_norm": 8.206068992614746, "learning_rate": 3.544335350838285e-06, "loss": 1.7912, "step": 30360 }, { "epoch": 2.379388714733542, "grad_norm": 2.240495204925537, "learning_rate": 3.5440403489875296e-06, "loss": 1.3005, "step": 30361 }, { "epoch": 2.3794670846394985, "grad_norm": 2.7197306156158447, "learning_rate": 3.543745352674496e-06, "loss": 1.312, "step": 30362 }, { "epoch": 2.3795454545454544, "grad_norm": 3.2331948280334473, "learning_rate": 3.543450361900307e-06, "loss": 1.0073, "step": 30363 }, { "epoch": 2.3796238244514107, "grad_norm": 2.47572922706604, "learning_rate": 3.543155376666083e-06, "loss": 0.9087, "step": 30364 }, { "epoch": 2.3797021943573666, "grad_norm": 5.747982978820801, "learning_rate": 3.5428603969729515e-06, "loss": 1.3388, "step": 30365 }, { "epoch": 2.379780564263323, "grad_norm": 10.32849407196045, "learning_rate": 3.5425654228220276e-06, "loss": 0.7318, "step": 30366 }, { "epoch": 2.379858934169279, "grad_norm": 4.658801078796387, "learning_rate": 3.5422704542144367e-06, "loss": 1.3152, "step": 30367 }, { "epoch": 2.379937304075235, "grad_norm": 6.994734287261963, "learning_rate": 3.5419754911513003e-06, "loss": 1.4525, "step": 30368 }, { "epoch": 2.380015673981191, "grad_norm": 3.252284049987793, "learning_rate": 3.541680533633741e-06, "loss": 1.2266, "step": 30369 }, { "epoch": 2.3800940438871474, "grad_norm": 6.795454978942871, "learning_rate": 3.5413855816628793e-06, "loss": 1.4822, "step": 30370 }, { "epoch": 2.3801724137931033, "grad_norm": 3.908332586288452, "learning_rate": 3.5410906352398373e-06, "loss": 1.3406, "step": 30371 }, { "epoch": 2.3802507836990596, "grad_norm": 2.523172378540039, "learning_rate": 3.5407956943657394e-06, "loss": 0.9894, "step": 30372 }, { "epoch": 2.3803291536050155, "grad_norm": 3.1376702785491943, "learning_rate": 3.5405007590417025e-06, "loss": 0.9518, "step": 30373 }, { "epoch": 2.380407523510972, "grad_norm": 2.346942186355591, "learning_rate": 3.540205829268852e-06, "loss": 1.0632, "step": 30374 }, { "epoch": 2.3804858934169277, "grad_norm": 3.7269275188446045, "learning_rate": 3.5399109050483082e-06, "loss": 1.0538, "step": 30375 }, { "epoch": 2.380564263322884, "grad_norm": 2.997410535812378, "learning_rate": 3.539615986381195e-06, "loss": 1.2989, "step": 30376 }, { "epoch": 2.38064263322884, "grad_norm": 2.8837246894836426, "learning_rate": 3.5393210732686312e-06, "loss": 1.3774, "step": 30377 }, { "epoch": 2.3807210031347963, "grad_norm": 2.934910535812378, "learning_rate": 3.5390261657117388e-06, "loss": 1.4663, "step": 30378 }, { "epoch": 2.380799373040752, "grad_norm": 5.252644062042236, "learning_rate": 3.5387312637116416e-06, "loss": 0.7316, "step": 30379 }, { "epoch": 2.3808777429467085, "grad_norm": 4.677175045013428, "learning_rate": 3.538436367269461e-06, "loss": 0.9934, "step": 30380 }, { "epoch": 2.3809561128526644, "grad_norm": 2.7355148792266846, "learning_rate": 3.538141476386317e-06, "loss": 1.4259, "step": 30381 }, { "epoch": 2.3810344827586207, "grad_norm": 15.142045974731445, "learning_rate": 3.5378465910633318e-06, "loss": 1.4193, "step": 30382 }, { "epoch": 2.3811128526645766, "grad_norm": 3.0309717655181885, "learning_rate": 3.5375517113016285e-06, "loss": 0.9658, "step": 30383 }, { "epoch": 2.381191222570533, "grad_norm": 3.8679938316345215, "learning_rate": 3.537256837102326e-06, "loss": 1.2792, "step": 30384 }, { "epoch": 2.381269592476489, "grad_norm": 2.296703815460205, "learning_rate": 3.5369619684665456e-06, "loss": 1.2934, "step": 30385 }, { "epoch": 2.381347962382445, "grad_norm": 3.0469720363616943, "learning_rate": 3.5366671053954123e-06, "loss": 1.3458, "step": 30386 }, { "epoch": 2.381426332288401, "grad_norm": 3.123295545578003, "learning_rate": 3.5363722478900465e-06, "loss": 0.9945, "step": 30387 }, { "epoch": 2.3815047021943574, "grad_norm": 2.3974101543426514, "learning_rate": 3.5360773959515676e-06, "loss": 1.5476, "step": 30388 }, { "epoch": 2.3815830721003133, "grad_norm": 4.843762397766113, "learning_rate": 3.5357825495810983e-06, "loss": 1.5664, "step": 30389 }, { "epoch": 2.3816614420062696, "grad_norm": 3.3629565238952637, "learning_rate": 3.5354877087797596e-06, "loss": 1.3659, "step": 30390 }, { "epoch": 2.381739811912226, "grad_norm": 4.198180675506592, "learning_rate": 3.535192873548676e-06, "loss": 1.1523, "step": 30391 }, { "epoch": 2.381818181818182, "grad_norm": 3.5548722743988037, "learning_rate": 3.5348980438889636e-06, "loss": 0.8612, "step": 30392 }, { "epoch": 2.3818965517241377, "grad_norm": 2.114576816558838, "learning_rate": 3.5346032198017475e-06, "loss": 1.3313, "step": 30393 }, { "epoch": 2.381974921630094, "grad_norm": 2.8953981399536133, "learning_rate": 3.534308401288148e-06, "loss": 1.0914, "step": 30394 }, { "epoch": 2.3820532915360504, "grad_norm": 6.608119487762451, "learning_rate": 3.5340135883492877e-06, "loss": 0.4909, "step": 30395 }, { "epoch": 2.3821316614420063, "grad_norm": 6.229064464569092, "learning_rate": 3.5337187809862857e-06, "loss": 1.1737, "step": 30396 }, { "epoch": 2.382210031347962, "grad_norm": 2.968524217605591, "learning_rate": 3.5334239792002633e-06, "loss": 1.6158, "step": 30397 }, { "epoch": 2.3822884012539185, "grad_norm": 2.8589258193969727, "learning_rate": 3.5331291829923443e-06, "loss": 1.1112, "step": 30398 }, { "epoch": 2.382366771159875, "grad_norm": 2.441311836242676, "learning_rate": 3.532834392363648e-06, "loss": 1.4564, "step": 30399 }, { "epoch": 2.3824451410658307, "grad_norm": 9.015707969665527, "learning_rate": 3.5325396073152964e-06, "loss": 1.0543, "step": 30400 }, { "epoch": 2.3825235109717866, "grad_norm": 5.108985424041748, "learning_rate": 3.5322448278484093e-06, "loss": 1.3631, "step": 30401 }, { "epoch": 2.382601880877743, "grad_norm": 3.350905418395996, "learning_rate": 3.5319500539641105e-06, "loss": 1.4722, "step": 30402 }, { "epoch": 2.3826802507836993, "grad_norm": 4.23329496383667, "learning_rate": 3.531655285663519e-06, "loss": 1.2652, "step": 30403 }, { "epoch": 2.382758620689655, "grad_norm": 3.031839370727539, "learning_rate": 3.5313605229477552e-06, "loss": 1.5098, "step": 30404 }, { "epoch": 2.382836990595611, "grad_norm": 3.3859667778015137, "learning_rate": 3.531065765817943e-06, "loss": 1.5356, "step": 30405 }, { "epoch": 2.3829153605015674, "grad_norm": 5.047151565551758, "learning_rate": 3.530771014275203e-06, "loss": 1.1604, "step": 30406 }, { "epoch": 2.3829937304075237, "grad_norm": 2.944812536239624, "learning_rate": 3.5304762683206543e-06, "loss": 0.9554, "step": 30407 }, { "epoch": 2.3830721003134796, "grad_norm": 3.014507532119751, "learning_rate": 3.53018152795542e-06, "loss": 1.5132, "step": 30408 }, { "epoch": 2.3831504702194355, "grad_norm": 2.9497241973876953, "learning_rate": 3.529886793180618e-06, "loss": 1.3764, "step": 30409 }, { "epoch": 2.383228840125392, "grad_norm": 2.156601905822754, "learning_rate": 3.5295920639973734e-06, "loss": 1.4116, "step": 30410 }, { "epoch": 2.383307210031348, "grad_norm": 3.0288541316986084, "learning_rate": 3.5292973404068053e-06, "loss": 1.2821, "step": 30411 }, { "epoch": 2.383385579937304, "grad_norm": 7.73431396484375, "learning_rate": 3.529002622410034e-06, "loss": 1.6149, "step": 30412 }, { "epoch": 2.3834639498432604, "grad_norm": 2.442108154296875, "learning_rate": 3.5287079100081822e-06, "loss": 1.5589, "step": 30413 }, { "epoch": 2.3835423197492163, "grad_norm": 2.9939773082733154, "learning_rate": 3.5284132032023694e-06, "loss": 1.464, "step": 30414 }, { "epoch": 2.3836206896551726, "grad_norm": 12.48214340209961, "learning_rate": 3.528118501993716e-06, "loss": 0.6029, "step": 30415 }, { "epoch": 2.3836990595611285, "grad_norm": 2.5491673946380615, "learning_rate": 3.5278238063833435e-06, "loss": 1.216, "step": 30416 }, { "epoch": 2.383777429467085, "grad_norm": 2.7505693435668945, "learning_rate": 3.527529116372376e-06, "loss": 1.5976, "step": 30417 }, { "epoch": 2.3838557993730407, "grad_norm": 3.055152654647827, "learning_rate": 3.527234431961929e-06, "loss": 1.3896, "step": 30418 }, { "epoch": 2.383934169278997, "grad_norm": 3.028424024581909, "learning_rate": 3.526939753153126e-06, "loss": 1.3088, "step": 30419 }, { "epoch": 2.384012539184953, "grad_norm": 1.9954291582107544, "learning_rate": 3.5266450799470874e-06, "loss": 1.3011, "step": 30420 }, { "epoch": 2.3840909090909093, "grad_norm": 6.284150123596191, "learning_rate": 3.5263504123449355e-06, "loss": 1.7249, "step": 30421 }, { "epoch": 2.384169278996865, "grad_norm": 3.313422918319702, "learning_rate": 3.5260557503477877e-06, "loss": 1.7764, "step": 30422 }, { "epoch": 2.3842476489028215, "grad_norm": 3.4448657035827637, "learning_rate": 3.5257610939567677e-06, "loss": 1.1452, "step": 30423 }, { "epoch": 2.3843260188087774, "grad_norm": 2.814781904220581, "learning_rate": 3.5254664431729958e-06, "loss": 1.3963, "step": 30424 }, { "epoch": 2.3844043887147337, "grad_norm": 5.222474575042725, "learning_rate": 3.525171797997593e-06, "loss": 1.3953, "step": 30425 }, { "epoch": 2.3844827586206896, "grad_norm": 3.9171621799468994, "learning_rate": 3.5248771584316775e-06, "loss": 1.3772, "step": 30426 }, { "epoch": 2.384561128526646, "grad_norm": 2.736062526702881, "learning_rate": 3.524582524476371e-06, "loss": 0.6385, "step": 30427 }, { "epoch": 2.384639498432602, "grad_norm": 2.359292507171631, "learning_rate": 3.5242878961327974e-06, "loss": 1.2201, "step": 30428 }, { "epoch": 2.384717868338558, "grad_norm": 4.467491149902344, "learning_rate": 3.5239932734020723e-06, "loss": 1.5653, "step": 30429 }, { "epoch": 2.384796238244514, "grad_norm": 3.3457090854644775, "learning_rate": 3.5236986562853193e-06, "loss": 1.3047, "step": 30430 }, { "epoch": 2.3848746081504704, "grad_norm": 2.6876471042633057, "learning_rate": 3.5234040447836582e-06, "loss": 1.4303, "step": 30431 }, { "epoch": 2.3849529780564263, "grad_norm": 4.081589221954346, "learning_rate": 3.523109438898211e-06, "loss": 0.9338, "step": 30432 }, { "epoch": 2.3850313479623826, "grad_norm": 2.629962205886841, "learning_rate": 3.5228148386300966e-06, "loss": 1.1421, "step": 30433 }, { "epoch": 2.3851097178683385, "grad_norm": 3.2018775939941406, "learning_rate": 3.5225202439804334e-06, "loss": 1.4789, "step": 30434 }, { "epoch": 2.385188087774295, "grad_norm": 2.8375823497772217, "learning_rate": 3.5222256549503464e-06, "loss": 1.3487, "step": 30435 }, { "epoch": 2.3852664576802507, "grad_norm": 5.900640487670898, "learning_rate": 3.5219310715409545e-06, "loss": 0.6913, "step": 30436 }, { "epoch": 2.385344827586207, "grad_norm": 2.1628005504608154, "learning_rate": 3.5216364937533765e-06, "loss": 1.4023, "step": 30437 }, { "epoch": 2.385423197492163, "grad_norm": 3.0434813499450684, "learning_rate": 3.521341921588734e-06, "loss": 1.5939, "step": 30438 }, { "epoch": 2.3855015673981192, "grad_norm": 3.7436368465423584, "learning_rate": 3.5210473550481485e-06, "loss": 1.5159, "step": 30439 }, { "epoch": 2.385579937304075, "grad_norm": 3.472374677658081, "learning_rate": 3.520752794132738e-06, "loss": 1.5971, "step": 30440 }, { "epoch": 2.3856583072100315, "grad_norm": 2.6863958835601807, "learning_rate": 3.5204582388436236e-06, "loss": 1.1158, "step": 30441 }, { "epoch": 2.3857366771159874, "grad_norm": 2.5705854892730713, "learning_rate": 3.5201636891819267e-06, "loss": 1.4021, "step": 30442 }, { "epoch": 2.3858150470219437, "grad_norm": 5.680936813354492, "learning_rate": 3.5198691451487675e-06, "loss": 1.574, "step": 30443 }, { "epoch": 2.3858934169278996, "grad_norm": 2.429373264312744, "learning_rate": 3.519574606745266e-06, "loss": 1.0099, "step": 30444 }, { "epoch": 2.385971786833856, "grad_norm": 5.473790168762207, "learning_rate": 3.519280073972542e-06, "loss": 1.4983, "step": 30445 }, { "epoch": 2.386050156739812, "grad_norm": 4.196065425872803, "learning_rate": 3.5189855468317148e-06, "loss": 1.107, "step": 30446 }, { "epoch": 2.386128526645768, "grad_norm": 3.362792730331421, "learning_rate": 3.518691025323908e-06, "loss": 1.3641, "step": 30447 }, { "epoch": 2.386206896551724, "grad_norm": 2.010150671005249, "learning_rate": 3.5183965094502382e-06, "loss": 0.8782, "step": 30448 }, { "epoch": 2.3862852664576804, "grad_norm": 2.7538130283355713, "learning_rate": 3.5181019992118273e-06, "loss": 1.2002, "step": 30449 }, { "epoch": 2.3863636363636362, "grad_norm": 3.3806841373443604, "learning_rate": 3.517807494609795e-06, "loss": 1.521, "step": 30450 }, { "epoch": 2.3864420062695926, "grad_norm": 2.3964056968688965, "learning_rate": 3.5175129956452638e-06, "loss": 1.2999, "step": 30451 }, { "epoch": 2.3865203761755485, "grad_norm": 6.616974830627441, "learning_rate": 3.5172185023193496e-06, "loss": 1.498, "step": 30452 }, { "epoch": 2.386598746081505, "grad_norm": 6.731561183929443, "learning_rate": 3.516924014633174e-06, "loss": 1.5672, "step": 30453 }, { "epoch": 2.3866771159874607, "grad_norm": 3.22230863571167, "learning_rate": 3.5166295325878607e-06, "loss": 1.2648, "step": 30454 }, { "epoch": 2.386755485893417, "grad_norm": 3.2004566192626953, "learning_rate": 3.5163350561845234e-06, "loss": 1.4267, "step": 30455 }, { "epoch": 2.386833855799373, "grad_norm": 2.0177059173583984, "learning_rate": 3.5160405854242873e-06, "loss": 1.1475, "step": 30456 }, { "epoch": 2.3869122257053292, "grad_norm": 3.520911931991577, "learning_rate": 3.51574612030827e-06, "loss": 1.5404, "step": 30457 }, { "epoch": 2.386990595611285, "grad_norm": 3.367245674133301, "learning_rate": 3.5154516608375932e-06, "loss": 1.617, "step": 30458 }, { "epoch": 2.3870689655172415, "grad_norm": 4.077630996704102, "learning_rate": 3.5151572070133744e-06, "loss": 0.9024, "step": 30459 }, { "epoch": 2.3871473354231973, "grad_norm": 2.8855533599853516, "learning_rate": 3.5148627588367345e-06, "loss": 1.2976, "step": 30460 }, { "epoch": 2.3872257053291537, "grad_norm": 8.417205810546875, "learning_rate": 3.5145683163087952e-06, "loss": 1.2798, "step": 30461 }, { "epoch": 2.3873040752351096, "grad_norm": 3.4502081871032715, "learning_rate": 3.5142738794306754e-06, "loss": 1.4053, "step": 30462 }, { "epoch": 2.387382445141066, "grad_norm": 3.6239116191864014, "learning_rate": 3.5139794482034935e-06, "loss": 1.4118, "step": 30463 }, { "epoch": 2.387460815047022, "grad_norm": 3.159726142883301, "learning_rate": 3.513685022628371e-06, "loss": 1.1275, "step": 30464 }, { "epoch": 2.387539184952978, "grad_norm": 2.868166208267212, "learning_rate": 3.5133906027064256e-06, "loss": 1.3888, "step": 30465 }, { "epoch": 2.387617554858934, "grad_norm": 2.6658222675323486, "learning_rate": 3.5130961884387826e-06, "loss": 1.1542, "step": 30466 }, { "epoch": 2.3876959247648903, "grad_norm": 2.7211856842041016, "learning_rate": 3.5128017798265545e-06, "loss": 1.2937, "step": 30467 }, { "epoch": 2.3877742946708462, "grad_norm": 2.701043128967285, "learning_rate": 3.512507376870866e-06, "loss": 1.2023, "step": 30468 }, { "epoch": 2.3878526645768026, "grad_norm": 3.1724305152893066, "learning_rate": 3.5122129795728354e-06, "loss": 1.2366, "step": 30469 }, { "epoch": 2.3879310344827585, "grad_norm": 3.499804973602295, "learning_rate": 3.5119185879335827e-06, "loss": 1.5874, "step": 30470 }, { "epoch": 2.388009404388715, "grad_norm": 2.745518684387207, "learning_rate": 3.5116242019542262e-06, "loss": 1.2367, "step": 30471 }, { "epoch": 2.3880877742946707, "grad_norm": 2.8925108909606934, "learning_rate": 3.5113298216358865e-06, "loss": 1.36, "step": 30472 }, { "epoch": 2.388166144200627, "grad_norm": 13.704044342041016, "learning_rate": 3.5110354469796863e-06, "loss": 1.0618, "step": 30473 }, { "epoch": 2.388244514106583, "grad_norm": 3.1906955242156982, "learning_rate": 3.5107410779867395e-06, "loss": 1.3792, "step": 30474 }, { "epoch": 2.3883228840125392, "grad_norm": 3.4632718563079834, "learning_rate": 3.51044671465817e-06, "loss": 1.4345, "step": 30475 }, { "epoch": 2.388401253918495, "grad_norm": 3.9278669357299805, "learning_rate": 3.510152356995096e-06, "loss": 1.1466, "step": 30476 }, { "epoch": 2.3884796238244514, "grad_norm": 3.153942823410034, "learning_rate": 3.5098580049986376e-06, "loss": 1.383, "step": 30477 }, { "epoch": 2.3885579937304073, "grad_norm": 1.803882360458374, "learning_rate": 3.509563658669913e-06, "loss": 1.4089, "step": 30478 }, { "epoch": 2.3886363636363637, "grad_norm": 2.8430051803588867, "learning_rate": 3.5092693180100423e-06, "loss": 1.5374, "step": 30479 }, { "epoch": 2.3887147335423196, "grad_norm": 8.141861915588379, "learning_rate": 3.508974983020149e-06, "loss": 1.0245, "step": 30480 }, { "epoch": 2.388793103448276, "grad_norm": 4.253532409667969, "learning_rate": 3.508680653701345e-06, "loss": 1.4414, "step": 30481 }, { "epoch": 2.3888714733542322, "grad_norm": 2.3123154640197754, "learning_rate": 3.508386330054755e-06, "loss": 1.5421, "step": 30482 }, { "epoch": 2.388949843260188, "grad_norm": 5.153496265411377, "learning_rate": 3.5080920120814974e-06, "loss": 1.3729, "step": 30483 }, { "epoch": 2.389028213166144, "grad_norm": 2.5836613178253174, "learning_rate": 3.5077976997826922e-06, "loss": 1.3681, "step": 30484 }, { "epoch": 2.3891065830721003, "grad_norm": 19.897172927856445, "learning_rate": 3.5075033931594576e-06, "loss": 0.6074, "step": 30485 }, { "epoch": 2.3891849529780567, "grad_norm": 2.5369210243225098, "learning_rate": 3.507209092212912e-06, "loss": 1.2645, "step": 30486 }, { "epoch": 2.3892633228840126, "grad_norm": 3.253000020980835, "learning_rate": 3.5069147969441776e-06, "loss": 1.3141, "step": 30487 }, { "epoch": 2.3893416927899684, "grad_norm": 4.159073829650879, "learning_rate": 3.5066205073543736e-06, "loss": 1.7019, "step": 30488 }, { "epoch": 2.3894200626959248, "grad_norm": 5.265920639038086, "learning_rate": 3.5063262234446166e-06, "loss": 1.3602, "step": 30489 }, { "epoch": 2.389498432601881, "grad_norm": 2.6632840633392334, "learning_rate": 3.506031945216028e-06, "loss": 1.4833, "step": 30490 }, { "epoch": 2.389576802507837, "grad_norm": 3.578923225402832, "learning_rate": 3.5057376726697253e-06, "loss": 1.2171, "step": 30491 }, { "epoch": 2.389655172413793, "grad_norm": 2.605621099472046, "learning_rate": 3.505443405806832e-06, "loss": 1.3954, "step": 30492 }, { "epoch": 2.389733542319749, "grad_norm": 2.8772199153900146, "learning_rate": 3.5051491446284613e-06, "loss": 1.8081, "step": 30493 }, { "epoch": 2.3898119122257055, "grad_norm": 4.714333534240723, "learning_rate": 3.504854889135736e-06, "loss": 1.0892, "step": 30494 }, { "epoch": 2.3898902821316614, "grad_norm": 2.800126314163208, "learning_rate": 3.504560639329776e-06, "loss": 1.5311, "step": 30495 }, { "epoch": 2.3899686520376173, "grad_norm": 2.090994358062744, "learning_rate": 3.5042663952116983e-06, "loss": 1.4565, "step": 30496 }, { "epoch": 2.3900470219435737, "grad_norm": 2.3747408390045166, "learning_rate": 3.5039721567826233e-06, "loss": 1.3704, "step": 30497 }, { "epoch": 2.39012539184953, "grad_norm": 3.46765398979187, "learning_rate": 3.5036779240436688e-06, "loss": 1.238, "step": 30498 }, { "epoch": 2.390203761755486, "grad_norm": 2.8590195178985596, "learning_rate": 3.5033836969959578e-06, "loss": 1.232, "step": 30499 }, { "epoch": 2.3902821316614418, "grad_norm": 2.5935816764831543, "learning_rate": 3.503089475640603e-06, "loss": 1.3364, "step": 30500 }, { "epoch": 2.390360501567398, "grad_norm": 2.272739887237549, "learning_rate": 3.5027952599787284e-06, "loss": 1.2793, "step": 30501 }, { "epoch": 2.3904388714733544, "grad_norm": 3.656294822692871, "learning_rate": 3.502501050011452e-06, "loss": 1.3977, "step": 30502 }, { "epoch": 2.3905172413793103, "grad_norm": 3.8978183269500732, "learning_rate": 3.502206845739893e-06, "loss": 1.2637, "step": 30503 }, { "epoch": 2.390595611285266, "grad_norm": 2.873257637023926, "learning_rate": 3.5019126471651686e-06, "loss": 1.3111, "step": 30504 }, { "epoch": 2.3906739811912225, "grad_norm": 4.058547496795654, "learning_rate": 3.501618454288398e-06, "loss": 1.2239, "step": 30505 }, { "epoch": 2.390752351097179, "grad_norm": 3.072801113128662, "learning_rate": 3.501324267110703e-06, "loss": 1.2116, "step": 30506 }, { "epoch": 2.3908307210031348, "grad_norm": 3.3001980781555176, "learning_rate": 3.5010300856332014e-06, "loss": 1.1759, "step": 30507 }, { "epoch": 2.390909090909091, "grad_norm": 2.538703680038452, "learning_rate": 3.5007359098570096e-06, "loss": 1.3928, "step": 30508 }, { "epoch": 2.390987460815047, "grad_norm": 2.7147812843322754, "learning_rate": 3.500441739783249e-06, "loss": 1.0991, "step": 30509 }, { "epoch": 2.3910658307210033, "grad_norm": 3.034811019897461, "learning_rate": 3.5001475754130387e-06, "loss": 1.2757, "step": 30510 }, { "epoch": 2.391144200626959, "grad_norm": 4.048792362213135, "learning_rate": 3.4998534167474953e-06, "loss": 1.3939, "step": 30511 }, { "epoch": 2.3912225705329155, "grad_norm": 2.7679033279418945, "learning_rate": 3.499559263787739e-06, "loss": 1.3043, "step": 30512 }, { "epoch": 2.3913009404388714, "grad_norm": 9.29765510559082, "learning_rate": 3.499265116534889e-06, "loss": 1.4711, "step": 30513 }, { "epoch": 2.3913793103448278, "grad_norm": 2.422464370727539, "learning_rate": 3.498970974990065e-06, "loss": 1.3333, "step": 30514 }, { "epoch": 2.3914576802507836, "grad_norm": 3.1961889266967773, "learning_rate": 3.4986768391543823e-06, "loss": 1.4674, "step": 30515 }, { "epoch": 2.39153605015674, "grad_norm": 3.5455806255340576, "learning_rate": 3.498382709028962e-06, "loss": 1.2301, "step": 30516 }, { "epoch": 2.391614420062696, "grad_norm": 2.7606866359710693, "learning_rate": 3.4980885846149227e-06, "loss": 1.798, "step": 30517 }, { "epoch": 2.391692789968652, "grad_norm": 3.5595405101776123, "learning_rate": 3.4977944659133844e-06, "loss": 1.458, "step": 30518 }, { "epoch": 2.391771159874608, "grad_norm": 6.091136455535889, "learning_rate": 3.4975003529254635e-06, "loss": 1.7494, "step": 30519 }, { "epoch": 2.3918495297805644, "grad_norm": 10.359809875488281, "learning_rate": 3.49720624565228e-06, "loss": 1.103, "step": 30520 }, { "epoch": 2.3919278996865203, "grad_norm": 2.5551774501800537, "learning_rate": 3.4969121440949515e-06, "loss": 1.3815, "step": 30521 }, { "epoch": 2.3920062695924766, "grad_norm": 5.202072620391846, "learning_rate": 3.496618048254598e-06, "loss": 0.984, "step": 30522 }, { "epoch": 2.3920846394984325, "grad_norm": 4.707502841949463, "learning_rate": 3.496323958132335e-06, "loss": 1.8674, "step": 30523 }, { "epoch": 2.392163009404389, "grad_norm": 3.156677007675171, "learning_rate": 3.496029873729285e-06, "loss": 1.1299, "step": 30524 }, { "epoch": 2.3922413793103448, "grad_norm": 10.552651405334473, "learning_rate": 3.4957357950465654e-06, "loss": 1.3035, "step": 30525 }, { "epoch": 2.392319749216301, "grad_norm": 10.246028900146484, "learning_rate": 3.4954417220852936e-06, "loss": 1.5595, "step": 30526 }, { "epoch": 2.392398119122257, "grad_norm": 3.0339534282684326, "learning_rate": 3.4951476548465883e-06, "loss": 1.3629, "step": 30527 }, { "epoch": 2.3924764890282133, "grad_norm": 3.956808090209961, "learning_rate": 3.4948535933315675e-06, "loss": 1.245, "step": 30528 }, { "epoch": 2.392554858934169, "grad_norm": 2.7080695629119873, "learning_rate": 3.4945595375413545e-06, "loss": 1.4387, "step": 30529 }, { "epoch": 2.3926332288401255, "grad_norm": 6.283206939697266, "learning_rate": 3.494265487477059e-06, "loss": 1.709, "step": 30530 }, { "epoch": 2.3927115987460814, "grad_norm": 5.683994770050049, "learning_rate": 3.4939714431398064e-06, "loss": 1.5696, "step": 30531 }, { "epoch": 2.3927899686520377, "grad_norm": 3.036102294921875, "learning_rate": 3.4936774045307126e-06, "loss": 1.6455, "step": 30532 }, { "epoch": 2.3928683385579936, "grad_norm": 4.502058982849121, "learning_rate": 3.4933833716508976e-06, "loss": 1.4523, "step": 30533 }, { "epoch": 2.39294670846395, "grad_norm": 3.9611809253692627, "learning_rate": 3.493089344501477e-06, "loss": 1.2823, "step": 30534 }, { "epoch": 2.393025078369906, "grad_norm": 5.627278804779053, "learning_rate": 3.4927953230835695e-06, "loss": 1.6646, "step": 30535 }, { "epoch": 2.393103448275862, "grad_norm": 3.4347872734069824, "learning_rate": 3.492501307398297e-06, "loss": 1.6511, "step": 30536 }, { "epoch": 2.393181818181818, "grad_norm": 3.133626937866211, "learning_rate": 3.4922072974467733e-06, "loss": 1.1644, "step": 30537 }, { "epoch": 2.3932601880877744, "grad_norm": 3.1011455059051514, "learning_rate": 3.491913293230119e-06, "loss": 1.2466, "step": 30538 }, { "epoch": 2.3933385579937303, "grad_norm": 2.8992257118225098, "learning_rate": 3.491619294749452e-06, "loss": 1.7618, "step": 30539 }, { "epoch": 2.3934169278996866, "grad_norm": 4.540345191955566, "learning_rate": 3.491325302005891e-06, "loss": 0.7198, "step": 30540 }, { "epoch": 2.3934952978056425, "grad_norm": 2.807821035385132, "learning_rate": 3.491031315000554e-06, "loss": 1.2307, "step": 30541 }, { "epoch": 2.393573667711599, "grad_norm": 2.565784215927124, "learning_rate": 3.4907373337345564e-06, "loss": 1.5195, "step": 30542 }, { "epoch": 2.3936520376175547, "grad_norm": 2.762394905090332, "learning_rate": 3.4904433582090202e-06, "loss": 0.9137, "step": 30543 }, { "epoch": 2.393730407523511, "grad_norm": 9.53217601776123, "learning_rate": 3.490149388425063e-06, "loss": 0.7455, "step": 30544 }, { "epoch": 2.393808777429467, "grad_norm": 3.5821380615234375, "learning_rate": 3.489855424383801e-06, "loss": 1.5519, "step": 30545 }, { "epoch": 2.3938871473354233, "grad_norm": 5.452120780944824, "learning_rate": 3.4895614660863544e-06, "loss": 1.1201, "step": 30546 }, { "epoch": 2.393965517241379, "grad_norm": 2.8147287368774414, "learning_rate": 3.489267513533838e-06, "loss": 1.2364, "step": 30547 }, { "epoch": 2.3940438871473355, "grad_norm": 2.6232504844665527, "learning_rate": 3.488973566727375e-06, "loss": 1.2391, "step": 30548 }, { "epoch": 2.3941222570532914, "grad_norm": 9.194742202758789, "learning_rate": 3.4886796256680776e-06, "loss": 1.7183, "step": 30549 }, { "epoch": 2.3942006269592477, "grad_norm": 2.2436301708221436, "learning_rate": 3.488385690357068e-06, "loss": 1.2986, "step": 30550 }, { "epoch": 2.3942789968652036, "grad_norm": 2.6118013858795166, "learning_rate": 3.4880917607954633e-06, "loss": 1.414, "step": 30551 }, { "epoch": 2.39435736677116, "grad_norm": 6.125433921813965, "learning_rate": 3.4877978369843803e-06, "loss": 1.5243, "step": 30552 }, { "epoch": 2.394435736677116, "grad_norm": 2.1143195629119873, "learning_rate": 3.4875039189249376e-06, "loss": 1.1656, "step": 30553 }, { "epoch": 2.394514106583072, "grad_norm": 3.175441265106201, "learning_rate": 3.4872100066182522e-06, "loss": 1.0651, "step": 30554 }, { "epoch": 2.394592476489028, "grad_norm": 2.261591672897339, "learning_rate": 3.486916100065446e-06, "loss": 1.4863, "step": 30555 }, { "epoch": 2.3946708463949844, "grad_norm": 3.765099048614502, "learning_rate": 3.4866221992676306e-06, "loss": 1.4104, "step": 30556 }, { "epoch": 2.3947492163009403, "grad_norm": 2.572021245956421, "learning_rate": 3.4863283042259276e-06, "loss": 1.0156, "step": 30557 }, { "epoch": 2.3948275862068966, "grad_norm": 13.725730895996094, "learning_rate": 3.486034414941455e-06, "loss": 1.3081, "step": 30558 }, { "epoch": 2.3949059561128525, "grad_norm": 3.360491991043091, "learning_rate": 3.4857405314153304e-06, "loss": 1.0239, "step": 30559 }, { "epoch": 2.394984326018809, "grad_norm": 3.76391863822937, "learning_rate": 3.4854466536486704e-06, "loss": 1.3074, "step": 30560 }, { "epoch": 2.3950626959247647, "grad_norm": 3.2483510971069336, "learning_rate": 3.4851527816425925e-06, "loss": 1.1471, "step": 30561 }, { "epoch": 2.395141065830721, "grad_norm": 3.259035110473633, "learning_rate": 3.484858915398216e-06, "loss": 1.1123, "step": 30562 }, { "epoch": 2.395219435736677, "grad_norm": 5.033419609069824, "learning_rate": 3.4845650549166588e-06, "loss": 0.5604, "step": 30563 }, { "epoch": 2.3952978056426333, "grad_norm": 3.5220091342926025, "learning_rate": 3.484271200199037e-06, "loss": 1.2753, "step": 30564 }, { "epoch": 2.395376175548589, "grad_norm": 2.956146240234375, "learning_rate": 3.4839773512464693e-06, "loss": 1.2694, "step": 30565 }, { "epoch": 2.3954545454545455, "grad_norm": 2.038094997406006, "learning_rate": 3.4836835080600734e-06, "loss": 1.4247, "step": 30566 }, { "epoch": 2.3955329153605014, "grad_norm": 17.059099197387695, "learning_rate": 3.4833896706409663e-06, "loss": 0.8029, "step": 30567 }, { "epoch": 2.3956112852664577, "grad_norm": 10.236665725708008, "learning_rate": 3.4830958389902646e-06, "loss": 1.1148, "step": 30568 }, { "epoch": 2.3956896551724136, "grad_norm": 3.295820951461792, "learning_rate": 3.482802013109088e-06, "loss": 1.7067, "step": 30569 }, { "epoch": 2.39576802507837, "grad_norm": 3.274571418762207, "learning_rate": 3.482508192998554e-06, "loss": 1.1993, "step": 30570 }, { "epoch": 2.395846394984326, "grad_norm": 5.973057270050049, "learning_rate": 3.4822143786597795e-06, "loss": 0.9906, "step": 30571 }, { "epoch": 2.395924764890282, "grad_norm": 3.198793888092041, "learning_rate": 3.481920570093881e-06, "loss": 1.615, "step": 30572 }, { "epoch": 2.396003134796238, "grad_norm": 1.9287538528442383, "learning_rate": 3.4816267673019766e-06, "loss": 1.368, "step": 30573 }, { "epoch": 2.3960815047021944, "grad_norm": 2.8774373531341553, "learning_rate": 3.4813329702851862e-06, "loss": 1.6728, "step": 30574 }, { "epoch": 2.3961598746081503, "grad_norm": 2.1065726280212402, "learning_rate": 3.481039179044623e-06, "loss": 1.2856, "step": 30575 }, { "epoch": 2.3962382445141066, "grad_norm": 3.6353254318237305, "learning_rate": 3.4807453935814074e-06, "loss": 1.7986, "step": 30576 }, { "epoch": 2.396316614420063, "grad_norm": 3.081328868865967, "learning_rate": 3.4804516138966556e-06, "loss": 1.376, "step": 30577 }, { "epoch": 2.396394984326019, "grad_norm": 6.01881742477417, "learning_rate": 3.4801578399914857e-06, "loss": 1.5447, "step": 30578 }, { "epoch": 2.3964733542319747, "grad_norm": 2.8100225925445557, "learning_rate": 3.479864071867015e-06, "loss": 1.3221, "step": 30579 }, { "epoch": 2.396551724137931, "grad_norm": 2.079925775527954, "learning_rate": 3.4795703095243594e-06, "loss": 1.2805, "step": 30580 }, { "epoch": 2.3966300940438874, "grad_norm": 3.6207334995269775, "learning_rate": 3.47927655296464e-06, "loss": 1.6394, "step": 30581 }, { "epoch": 2.3967084639498433, "grad_norm": 5.008162498474121, "learning_rate": 3.478982802188969e-06, "loss": 1.454, "step": 30582 }, { "epoch": 2.396786833855799, "grad_norm": 1.9463406801223755, "learning_rate": 3.4786890571984664e-06, "loss": 1.366, "step": 30583 }, { "epoch": 2.3968652037617555, "grad_norm": 5.709420204162598, "learning_rate": 3.4783953179942497e-06, "loss": 2.0899, "step": 30584 }, { "epoch": 2.396943573667712, "grad_norm": 3.413774013519287, "learning_rate": 3.4781015845774368e-06, "loss": 1.6611, "step": 30585 }, { "epoch": 2.3970219435736677, "grad_norm": 5.445371150970459, "learning_rate": 3.4778078569491424e-06, "loss": 1.5229, "step": 30586 }, { "epoch": 2.3971003134796236, "grad_norm": 2.7182400226593018, "learning_rate": 3.4775141351104847e-06, "loss": 0.5719, "step": 30587 }, { "epoch": 2.39717868338558, "grad_norm": 1.9938963651657104, "learning_rate": 3.4772204190625813e-06, "loss": 1.5925, "step": 30588 }, { "epoch": 2.3972570532915363, "grad_norm": 3.470411777496338, "learning_rate": 3.476926708806551e-06, "loss": 1.0341, "step": 30589 }, { "epoch": 2.397335423197492, "grad_norm": 10.059718132019043, "learning_rate": 3.476633004343508e-06, "loss": 1.9658, "step": 30590 }, { "epoch": 2.397413793103448, "grad_norm": 7.584779739379883, "learning_rate": 3.476339305674571e-06, "loss": 1.7505, "step": 30591 }, { "epoch": 2.3974921630094044, "grad_norm": 2.4490902423858643, "learning_rate": 3.4760456128008566e-06, "loss": 1.4083, "step": 30592 }, { "epoch": 2.3975705329153607, "grad_norm": 7.588868618011475, "learning_rate": 3.4757519257234824e-06, "loss": 1.3943, "step": 30593 }, { "epoch": 2.3976489028213166, "grad_norm": 3.528047561645508, "learning_rate": 3.4754582444435625e-06, "loss": 1.4908, "step": 30594 }, { "epoch": 2.3977272727272725, "grad_norm": 2.3532702922821045, "learning_rate": 3.4751645689622184e-06, "loss": 1.4102, "step": 30595 }, { "epoch": 2.397805642633229, "grad_norm": 4.134979724884033, "learning_rate": 3.4748708992805657e-06, "loss": 1.6024, "step": 30596 }, { "epoch": 2.397884012539185, "grad_norm": 7.6443376541137695, "learning_rate": 3.4745772353997198e-06, "loss": 1.4507, "step": 30597 }, { "epoch": 2.397962382445141, "grad_norm": 2.845791816711426, "learning_rate": 3.4742835773207984e-06, "loss": 1.3539, "step": 30598 }, { "epoch": 2.3980407523510974, "grad_norm": 3.8370509147644043, "learning_rate": 3.4739899250449185e-06, "loss": 1.3055, "step": 30599 }, { "epoch": 2.3981191222570533, "grad_norm": 4.249166488647461, "learning_rate": 3.4736962785731988e-06, "loss": 1.0519, "step": 30600 }, { "epoch": 2.3981974921630096, "grad_norm": 3.5347940921783447, "learning_rate": 3.473402637906752e-06, "loss": 1.4342, "step": 30601 }, { "epoch": 2.3982758620689655, "grad_norm": 3.1644866466522217, "learning_rate": 3.4731090030466986e-06, "loss": 1.6495, "step": 30602 }, { "epoch": 2.398354231974922, "grad_norm": 3.9199025630950928, "learning_rate": 3.4728153739941543e-06, "loss": 1.5021, "step": 30603 }, { "epoch": 2.3984326018808777, "grad_norm": 3.9218037128448486, "learning_rate": 3.4725217507502375e-06, "loss": 1.1904, "step": 30604 }, { "epoch": 2.398510971786834, "grad_norm": 2.34860897064209, "learning_rate": 3.4722281333160613e-06, "loss": 1.3446, "step": 30605 }, { "epoch": 2.39858934169279, "grad_norm": 2.8233449459075928, "learning_rate": 3.471934521692744e-06, "loss": 1.1281, "step": 30606 }, { "epoch": 2.3986677115987463, "grad_norm": 3.7195262908935547, "learning_rate": 3.4716409158814053e-06, "loss": 1.7556, "step": 30607 }, { "epoch": 2.398746081504702, "grad_norm": 2.670222520828247, "learning_rate": 3.471347315883158e-06, "loss": 1.1889, "step": 30608 }, { "epoch": 2.3988244514106585, "grad_norm": 2.480126142501831, "learning_rate": 3.47105372169912e-06, "loss": 1.1247, "step": 30609 }, { "epoch": 2.3989028213166144, "grad_norm": 8.783629417419434, "learning_rate": 3.4707601333304093e-06, "loss": 0.8329, "step": 30610 }, { "epoch": 2.3989811912225707, "grad_norm": 2.6642422676086426, "learning_rate": 3.470466550778142e-06, "loss": 0.8069, "step": 30611 }, { "epoch": 2.3990595611285266, "grad_norm": 3.62349796295166, "learning_rate": 3.4701729740434327e-06, "loss": 1.2152, "step": 30612 }, { "epoch": 2.399137931034483, "grad_norm": 3.0573370456695557, "learning_rate": 3.4698794031274e-06, "loss": 1.7381, "step": 30613 }, { "epoch": 2.399216300940439, "grad_norm": 2.613447666168213, "learning_rate": 3.469585838031161e-06, "loss": 1.5919, "step": 30614 }, { "epoch": 2.399294670846395, "grad_norm": 2.9446022510528564, "learning_rate": 3.4692922787558313e-06, "loss": 1.2225, "step": 30615 }, { "epoch": 2.399373040752351, "grad_norm": 2.9128706455230713, "learning_rate": 3.4689987253025276e-06, "loss": 0.9941, "step": 30616 }, { "epoch": 2.3994514106583074, "grad_norm": 4.469663143157959, "learning_rate": 3.468705177672366e-06, "loss": 1.0688, "step": 30617 }, { "epoch": 2.3995297805642632, "grad_norm": 2.6099250316619873, "learning_rate": 3.4684116358664623e-06, "loss": 0.4905, "step": 30618 }, { "epoch": 2.3996081504702196, "grad_norm": 5.829789638519287, "learning_rate": 3.4681180998859363e-06, "loss": 1.249, "step": 30619 }, { "epoch": 2.3996865203761755, "grad_norm": 2.8704631328582764, "learning_rate": 3.4678245697319006e-06, "loss": 1.3558, "step": 30620 }, { "epoch": 2.399764890282132, "grad_norm": 2.812697410583496, "learning_rate": 3.467531045405474e-06, "loss": 1.6354, "step": 30621 }, { "epoch": 2.3998432601880877, "grad_norm": 3.8568274974823, "learning_rate": 3.467237526907773e-06, "loss": 0.8302, "step": 30622 }, { "epoch": 2.399921630094044, "grad_norm": 4.804623126983643, "learning_rate": 3.466944014239912e-06, "loss": 1.2865, "step": 30623 }, { "epoch": 2.4, "grad_norm": 2.6349456310272217, "learning_rate": 3.4666505074030076e-06, "loss": 1.4684, "step": 30624 }, { "epoch": 2.4000783699059562, "grad_norm": 3.2206215858459473, "learning_rate": 3.4663570063981776e-06, "loss": 1.1042, "step": 30625 }, { "epoch": 2.400156739811912, "grad_norm": 3.238582134246826, "learning_rate": 3.4660635112265394e-06, "loss": 1.1973, "step": 30626 }, { "epoch": 2.4002351097178685, "grad_norm": 4.746129989624023, "learning_rate": 3.4657700218892066e-06, "loss": 1.4202, "step": 30627 }, { "epoch": 2.4003134796238244, "grad_norm": 3.381427764892578, "learning_rate": 3.465476538387297e-06, "loss": 1.3771, "step": 30628 }, { "epoch": 2.4003918495297807, "grad_norm": 2.567744016647339, "learning_rate": 3.465183060721925e-06, "loss": 1.5601, "step": 30629 }, { "epoch": 2.4004702194357366, "grad_norm": 2.152494192123413, "learning_rate": 3.4648895888942114e-06, "loss": 1.3139, "step": 30630 }, { "epoch": 2.400548589341693, "grad_norm": 3.2351999282836914, "learning_rate": 3.464596122905266e-06, "loss": 1.092, "step": 30631 }, { "epoch": 2.400626959247649, "grad_norm": 12.624380111694336, "learning_rate": 3.46430266275621e-06, "loss": 2.1975, "step": 30632 }, { "epoch": 2.400705329153605, "grad_norm": 3.074444532394409, "learning_rate": 3.4640092084481577e-06, "loss": 1.3591, "step": 30633 }, { "epoch": 2.400783699059561, "grad_norm": 1.8159303665161133, "learning_rate": 3.463715759982226e-06, "loss": 1.4342, "step": 30634 }, { "epoch": 2.4008620689655173, "grad_norm": 5.256270885467529, "learning_rate": 3.463422317359529e-06, "loss": 1.2996, "step": 30635 }, { "epoch": 2.4009404388714732, "grad_norm": 3.71583890914917, "learning_rate": 3.4631288805811845e-06, "loss": 1.3851, "step": 30636 }, { "epoch": 2.4010188087774296, "grad_norm": 6.129344463348389, "learning_rate": 3.462835449648311e-06, "loss": 1.0718, "step": 30637 }, { "epoch": 2.4010971786833855, "grad_norm": 2.5319621562957764, "learning_rate": 3.462542024562019e-06, "loss": 1.2807, "step": 30638 }, { "epoch": 2.401175548589342, "grad_norm": 5.059001445770264, "learning_rate": 3.462248605323428e-06, "loss": 1.7242, "step": 30639 }, { "epoch": 2.4012539184952977, "grad_norm": 2.529517889022827, "learning_rate": 3.4619551919336538e-06, "loss": 1.1854, "step": 30640 }, { "epoch": 2.401332288401254, "grad_norm": 3.2357566356658936, "learning_rate": 3.461661784393813e-06, "loss": 1.4745, "step": 30641 }, { "epoch": 2.40141065830721, "grad_norm": 3.9787769317626953, "learning_rate": 3.4613683827050193e-06, "loss": 1.2716, "step": 30642 }, { "epoch": 2.4014890282131662, "grad_norm": 7.588384628295898, "learning_rate": 3.461074986868389e-06, "loss": 1.6889, "step": 30643 }, { "epoch": 2.401567398119122, "grad_norm": 2.3832571506500244, "learning_rate": 3.4607815968850407e-06, "loss": 1.3822, "step": 30644 }, { "epoch": 2.4016457680250785, "grad_norm": 2.868760585784912, "learning_rate": 3.460488212756089e-06, "loss": 1.541, "step": 30645 }, { "epoch": 2.4017241379310343, "grad_norm": 3.106588840484619, "learning_rate": 3.4601948344826487e-06, "loss": 1.5666, "step": 30646 }, { "epoch": 2.4018025078369907, "grad_norm": 2.850477933883667, "learning_rate": 3.459901462065836e-06, "loss": 1.4812, "step": 30647 }, { "epoch": 2.4018808777429466, "grad_norm": 3.3409597873687744, "learning_rate": 3.4596080955067685e-06, "loss": 1.1529, "step": 30648 }, { "epoch": 2.401959247648903, "grad_norm": 20.7558536529541, "learning_rate": 3.459314734806559e-06, "loss": 0.9133, "step": 30649 }, { "epoch": 2.402037617554859, "grad_norm": 2.9558887481689453, "learning_rate": 3.459021379966324e-06, "loss": 1.4011, "step": 30650 }, { "epoch": 2.402115987460815, "grad_norm": 2.514572858810425, "learning_rate": 3.4587280309871816e-06, "loss": 1.2733, "step": 30651 }, { "epoch": 2.402194357366771, "grad_norm": 3.3947298526763916, "learning_rate": 3.4584346878702475e-06, "loss": 1.5091, "step": 30652 }, { "epoch": 2.4022727272727273, "grad_norm": 3.858297109603882, "learning_rate": 3.4581413506166343e-06, "loss": 1.448, "step": 30653 }, { "epoch": 2.4023510971786832, "grad_norm": 4.115458011627197, "learning_rate": 3.45784801922746e-06, "loss": 1.1181, "step": 30654 }, { "epoch": 2.4024294670846396, "grad_norm": 3.383371114730835, "learning_rate": 3.4575546937038387e-06, "loss": 1.2413, "step": 30655 }, { "epoch": 2.4025078369905954, "grad_norm": 7.227263927459717, "learning_rate": 3.45726137404689e-06, "loss": 2.0349, "step": 30656 }, { "epoch": 2.402586206896552, "grad_norm": 3.601524829864502, "learning_rate": 3.4569680602577228e-06, "loss": 1.4086, "step": 30657 }, { "epoch": 2.4026645768025077, "grad_norm": 2.382779836654663, "learning_rate": 3.456674752337458e-06, "loss": 1.0189, "step": 30658 }, { "epoch": 2.402742946708464, "grad_norm": 4.340243816375732, "learning_rate": 3.4563814502872106e-06, "loss": 1.3081, "step": 30659 }, { "epoch": 2.40282131661442, "grad_norm": 15.600217819213867, "learning_rate": 3.4560881541080952e-06, "loss": 1.0817, "step": 30660 }, { "epoch": 2.402899686520376, "grad_norm": 3.9227194786071777, "learning_rate": 3.4557948638012272e-06, "loss": 1.6196, "step": 30661 }, { "epoch": 2.402978056426332, "grad_norm": 2.3083748817443848, "learning_rate": 3.455501579367721e-06, "loss": 1.5248, "step": 30662 }, { "epoch": 2.4030564263322884, "grad_norm": 2.971074104309082, "learning_rate": 3.4552083008086967e-06, "loss": 1.1318, "step": 30663 }, { "epoch": 2.4031347962382443, "grad_norm": 10.56493091583252, "learning_rate": 3.4549150281252635e-06, "loss": 1.2303, "step": 30664 }, { "epoch": 2.4032131661442007, "grad_norm": 3.216447114944458, "learning_rate": 3.4546217613185407e-06, "loss": 1.4924, "step": 30665 }, { "epoch": 2.4032915360501566, "grad_norm": 2.835385322570801, "learning_rate": 3.4543285003896434e-06, "loss": 1.2223, "step": 30666 }, { "epoch": 2.403369905956113, "grad_norm": 3.5389151573181152, "learning_rate": 3.4540352453396875e-06, "loss": 0.7529, "step": 30667 }, { "epoch": 2.4034482758620688, "grad_norm": 5.078334808349609, "learning_rate": 3.453741996169786e-06, "loss": 1.7518, "step": 30668 }, { "epoch": 2.403526645768025, "grad_norm": 3.089487075805664, "learning_rate": 3.453448752881055e-06, "loss": 0.7093, "step": 30669 }, { "epoch": 2.403605015673981, "grad_norm": 2.480135679244995, "learning_rate": 3.453155515474612e-06, "loss": 1.0305, "step": 30670 }, { "epoch": 2.4036833855799373, "grad_norm": 2.859365463256836, "learning_rate": 3.4528622839515713e-06, "loss": 1.2857, "step": 30671 }, { "epoch": 2.4037617554858937, "grad_norm": 3.1164145469665527, "learning_rate": 3.4525690583130473e-06, "loss": 1.1912, "step": 30672 }, { "epoch": 2.4038401253918495, "grad_norm": 4.63417911529541, "learning_rate": 3.452275838560155e-06, "loss": 1.2444, "step": 30673 }, { "epoch": 2.4039184952978054, "grad_norm": 2.159529209136963, "learning_rate": 3.4519826246940103e-06, "loss": 1.1304, "step": 30674 }, { "epoch": 2.4039968652037618, "grad_norm": 3.496882915496826, "learning_rate": 3.451689416715731e-06, "loss": 1.3343, "step": 30675 }, { "epoch": 2.404075235109718, "grad_norm": 3.6785380840301514, "learning_rate": 3.451396214626428e-06, "loss": 1.4508, "step": 30676 }, { "epoch": 2.404153605015674, "grad_norm": 3.4692270755767822, "learning_rate": 3.451103018427218e-06, "loss": 1.3616, "step": 30677 }, { "epoch": 2.40423197492163, "grad_norm": 3.9221420288085938, "learning_rate": 3.4508098281192183e-06, "loss": 1.4419, "step": 30678 }, { "epoch": 2.404310344827586, "grad_norm": 4.7174859046936035, "learning_rate": 3.450516643703541e-06, "loss": 1.2813, "step": 30679 }, { "epoch": 2.4043887147335425, "grad_norm": 3.713167667388916, "learning_rate": 3.4502234651813026e-06, "loss": 1.642, "step": 30680 }, { "epoch": 2.4044670846394984, "grad_norm": 5.03195333480835, "learning_rate": 3.449930292553617e-06, "loss": 1.0867, "step": 30681 }, { "epoch": 2.4045454545454543, "grad_norm": 2.4017670154571533, "learning_rate": 3.4496371258216036e-06, "loss": 1.2982, "step": 30682 }, { "epoch": 2.4046238244514107, "grad_norm": 2.9022748470306396, "learning_rate": 3.4493439649863715e-06, "loss": 1.5363, "step": 30683 }, { "epoch": 2.404702194357367, "grad_norm": 3.5885231494903564, "learning_rate": 3.4490508100490396e-06, "loss": 1.2135, "step": 30684 }, { "epoch": 2.404780564263323, "grad_norm": 10.139809608459473, "learning_rate": 3.448757661010721e-06, "loss": 1.3635, "step": 30685 }, { "epoch": 2.4048589341692788, "grad_norm": 3.2880375385284424, "learning_rate": 3.448464517872533e-06, "loss": 1.2827, "step": 30686 }, { "epoch": 2.404937304075235, "grad_norm": 3.0626304149627686, "learning_rate": 3.4481713806355875e-06, "loss": 0.728, "step": 30687 }, { "epoch": 2.4050156739811914, "grad_norm": 2.0178661346435547, "learning_rate": 3.447878249301001e-06, "loss": 1.3397, "step": 30688 }, { "epoch": 2.4050940438871473, "grad_norm": 3.0802133083343506, "learning_rate": 3.4475851238698886e-06, "loss": 1.4765, "step": 30689 }, { "epoch": 2.405172413793103, "grad_norm": 3.05779767036438, "learning_rate": 3.4472920043433666e-06, "loss": 1.1481, "step": 30690 }, { "epoch": 2.4052507836990595, "grad_norm": 7.460960388183594, "learning_rate": 3.4469988907225473e-06, "loss": 1.3737, "step": 30691 }, { "epoch": 2.405329153605016, "grad_norm": 6.26419734954834, "learning_rate": 3.446705783008546e-06, "loss": 1.3647, "step": 30692 }, { "epoch": 2.4054075235109718, "grad_norm": 2.647366762161255, "learning_rate": 3.446412681202479e-06, "loss": 1.328, "step": 30693 }, { "epoch": 2.405485893416928, "grad_norm": 6.1816840171813965, "learning_rate": 3.4461195853054597e-06, "loss": 1.3044, "step": 30694 }, { "epoch": 2.405564263322884, "grad_norm": 7.365495204925537, "learning_rate": 3.4458264953186015e-06, "loss": 1.6697, "step": 30695 }, { "epoch": 2.4056426332288403, "grad_norm": 2.571690797805786, "learning_rate": 3.4455334112430227e-06, "loss": 1.2049, "step": 30696 }, { "epoch": 2.405721003134796, "grad_norm": 9.468097686767578, "learning_rate": 3.4452403330798377e-06, "loss": 0.7599, "step": 30697 }, { "epoch": 2.4057993730407525, "grad_norm": 3.431896924972534, "learning_rate": 3.444947260830159e-06, "loss": 1.4299, "step": 30698 }, { "epoch": 2.4058777429467084, "grad_norm": 3.1415250301361084, "learning_rate": 3.4446541944951017e-06, "loss": 1.5042, "step": 30699 }, { "epoch": 2.4059561128526648, "grad_norm": 3.9948065280914307, "learning_rate": 3.44436113407578e-06, "loss": 1.4082, "step": 30700 }, { "epoch": 2.4060344827586206, "grad_norm": 3.8253040313720703, "learning_rate": 3.4440680795733127e-06, "loss": 1.3274, "step": 30701 }, { "epoch": 2.406112852664577, "grad_norm": 2.497835874557495, "learning_rate": 3.443775030988808e-06, "loss": 1.2845, "step": 30702 }, { "epoch": 2.406191222570533, "grad_norm": 3.5219554901123047, "learning_rate": 3.443481988323385e-06, "loss": 1.3854, "step": 30703 }, { "epoch": 2.406269592476489, "grad_norm": 2.252514123916626, "learning_rate": 3.443188951578158e-06, "loss": 1.3083, "step": 30704 }, { "epoch": 2.406347962382445, "grad_norm": 3.0961029529571533, "learning_rate": 3.442895920754239e-06, "loss": 1.2772, "step": 30705 }, { "epoch": 2.4064263322884014, "grad_norm": 3.557053804397583, "learning_rate": 3.442602895852745e-06, "loss": 1.6322, "step": 30706 }, { "epoch": 2.4065047021943573, "grad_norm": 2.8037421703338623, "learning_rate": 3.442309876874788e-06, "loss": 1.17, "step": 30707 }, { "epoch": 2.4065830721003136, "grad_norm": 5.964776039123535, "learning_rate": 3.442016863821487e-06, "loss": 1.2664, "step": 30708 }, { "epoch": 2.4066614420062695, "grad_norm": 2.7727272510528564, "learning_rate": 3.4417238566939514e-06, "loss": 1.2544, "step": 30709 }, { "epoch": 2.406739811912226, "grad_norm": 2.8060176372528076, "learning_rate": 3.4414308554932984e-06, "loss": 1.2815, "step": 30710 }, { "epoch": 2.4068181818181817, "grad_norm": 3.8161089420318604, "learning_rate": 3.4411378602206424e-06, "loss": 1.1916, "step": 30711 }, { "epoch": 2.406896551724138, "grad_norm": 2.3797435760498047, "learning_rate": 3.440844870877098e-06, "loss": 1.4483, "step": 30712 }, { "epoch": 2.406974921630094, "grad_norm": 4.195693492889404, "learning_rate": 3.4405518874637766e-06, "loss": 0.7559, "step": 30713 }, { "epoch": 2.4070532915360503, "grad_norm": 3.0666725635528564, "learning_rate": 3.440258909981796e-06, "loss": 1.1751, "step": 30714 }, { "epoch": 2.407131661442006, "grad_norm": 4.278685092926025, "learning_rate": 3.4399659384322697e-06, "loss": 0.8322, "step": 30715 }, { "epoch": 2.4072100313479625, "grad_norm": 3.0088887214660645, "learning_rate": 3.4396729728163124e-06, "loss": 1.5399, "step": 30716 }, { "epoch": 2.4072884012539184, "grad_norm": 4.414648056030273, "learning_rate": 3.4393800131350364e-06, "loss": 0.6175, "step": 30717 }, { "epoch": 2.4073667711598747, "grad_norm": 3.196892738342285, "learning_rate": 3.4390870593895576e-06, "loss": 1.0535, "step": 30718 }, { "epoch": 2.4074451410658306, "grad_norm": 2.9667141437530518, "learning_rate": 3.438794111580991e-06, "loss": 1.1433, "step": 30719 }, { "epoch": 2.407523510971787, "grad_norm": 3.254943370819092, "learning_rate": 3.4385011697104477e-06, "loss": 1.311, "step": 30720 }, { "epoch": 2.407601880877743, "grad_norm": 3.509063243865967, "learning_rate": 3.4382082337790447e-06, "loss": 1.4823, "step": 30721 }, { "epoch": 2.407680250783699, "grad_norm": 6.045408248901367, "learning_rate": 3.4379153037878955e-06, "loss": 0.8032, "step": 30722 }, { "epoch": 2.407758620689655, "grad_norm": 3.7917566299438477, "learning_rate": 3.4376223797381146e-06, "loss": 1.3411, "step": 30723 }, { "epoch": 2.4078369905956114, "grad_norm": 3.825584650039673, "learning_rate": 3.4373294616308157e-06, "loss": 1.5506, "step": 30724 }, { "epoch": 2.4079153605015673, "grad_norm": 2.7862982749938965, "learning_rate": 3.437036549467111e-06, "loss": 1.3612, "step": 30725 }, { "epoch": 2.4079937304075236, "grad_norm": 2.8521816730499268, "learning_rate": 3.436743643248118e-06, "loss": 1.454, "step": 30726 }, { "epoch": 2.4080721003134795, "grad_norm": 2.4342153072357178, "learning_rate": 3.43645074297495e-06, "loss": 1.1621, "step": 30727 }, { "epoch": 2.408150470219436, "grad_norm": 3.0052225589752197, "learning_rate": 3.436157848648719e-06, "loss": 1.274, "step": 30728 }, { "epoch": 2.4082288401253917, "grad_norm": 2.96813702583313, "learning_rate": 3.4358649602705406e-06, "loss": 1.3188, "step": 30729 }, { "epoch": 2.408307210031348, "grad_norm": 2.94558048248291, "learning_rate": 3.435572077841528e-06, "loss": 1.5447, "step": 30730 }, { "epoch": 2.408385579937304, "grad_norm": 2.4411354064941406, "learning_rate": 3.4352792013627984e-06, "loss": 1.2, "step": 30731 }, { "epoch": 2.4084639498432603, "grad_norm": 6.607096195220947, "learning_rate": 3.43498633083546e-06, "loss": 0.9654, "step": 30732 }, { "epoch": 2.408542319749216, "grad_norm": 4.2050323486328125, "learning_rate": 3.43469346626063e-06, "loss": 1.2409, "step": 30733 }, { "epoch": 2.4086206896551725, "grad_norm": 2.3404743671417236, "learning_rate": 3.434400607639424e-06, "loss": 1.2146, "step": 30734 }, { "epoch": 2.4086990595611284, "grad_norm": 3.0960793495178223, "learning_rate": 3.4341077549729527e-06, "loss": 1.0556, "step": 30735 }, { "epoch": 2.4087774294670847, "grad_norm": 4.436664581298828, "learning_rate": 3.433814908262331e-06, "loss": 0.4505, "step": 30736 }, { "epoch": 2.4088557993730406, "grad_norm": 2.177915573120117, "learning_rate": 3.4335220675086724e-06, "loss": 0.9745, "step": 30737 }, { "epoch": 2.408934169278997, "grad_norm": 2.44484281539917, "learning_rate": 3.4332292327130935e-06, "loss": 1.4632, "step": 30738 }, { "epoch": 2.409012539184953, "grad_norm": 5.384920120239258, "learning_rate": 3.432936403876703e-06, "loss": 1.6751, "step": 30739 }, { "epoch": 2.409090909090909, "grad_norm": 3.430968761444092, "learning_rate": 3.432643581000619e-06, "loss": 1.502, "step": 30740 }, { "epoch": 2.409169278996865, "grad_norm": 5.251039505004883, "learning_rate": 3.432350764085954e-06, "loss": 1.4315, "step": 30741 }, { "epoch": 2.4092476489028214, "grad_norm": 3.1383049488067627, "learning_rate": 3.4320579531338223e-06, "loss": 1.6099, "step": 30742 }, { "epoch": 2.4093260188087773, "grad_norm": 6.5472798347473145, "learning_rate": 3.4317651481453353e-06, "loss": 1.1996, "step": 30743 }, { "epoch": 2.4094043887147336, "grad_norm": 5.7030439376831055, "learning_rate": 3.4314723491216073e-06, "loss": 1.4988, "step": 30744 }, { "epoch": 2.4094827586206895, "grad_norm": 3.7203214168548584, "learning_rate": 3.431179556063756e-06, "loss": 1.529, "step": 30745 }, { "epoch": 2.409561128526646, "grad_norm": 3.4543771743774414, "learning_rate": 3.4308867689728892e-06, "loss": 1.2938, "step": 30746 }, { "epoch": 2.4096394984326017, "grad_norm": 3.283449649810791, "learning_rate": 3.430593987850124e-06, "loss": 1.3617, "step": 30747 }, { "epoch": 2.409717868338558, "grad_norm": 4.075410842895508, "learning_rate": 3.430301212696573e-06, "loss": 1.3742, "step": 30748 }, { "epoch": 2.409796238244514, "grad_norm": 3.594966173171997, "learning_rate": 3.4300084435133507e-06, "loss": 1.4571, "step": 30749 }, { "epoch": 2.4098746081504703, "grad_norm": 5.8391432762146, "learning_rate": 3.4297156803015697e-06, "loss": 1.1833, "step": 30750 }, { "epoch": 2.409952978056426, "grad_norm": 3.323137044906616, "learning_rate": 3.4294229230623416e-06, "loss": 1.4706, "step": 30751 }, { "epoch": 2.4100313479623825, "grad_norm": 3.862203359603882, "learning_rate": 3.429130171796784e-06, "loss": 1.2961, "step": 30752 }, { "epoch": 2.4101097178683384, "grad_norm": 4.1032209396362305, "learning_rate": 3.4288374265060087e-06, "loss": 1.0017, "step": 30753 }, { "epoch": 2.4101880877742947, "grad_norm": 8.84711742401123, "learning_rate": 3.4285446871911278e-06, "loss": 1.2889, "step": 30754 }, { "epoch": 2.4102664576802506, "grad_norm": 3.3410377502441406, "learning_rate": 3.428251953853256e-06, "loss": 1.4587, "step": 30755 }, { "epoch": 2.410344827586207, "grad_norm": 4.359347343444824, "learning_rate": 3.4279592264935057e-06, "loss": 1.1531, "step": 30756 }, { "epoch": 2.410423197492163, "grad_norm": 4.067020893096924, "learning_rate": 3.4276665051129944e-06, "loss": 1.4145, "step": 30757 }, { "epoch": 2.410501567398119, "grad_norm": 2.997015953063965, "learning_rate": 3.427373789712828e-06, "loss": 1.2636, "step": 30758 }, { "epoch": 2.410579937304075, "grad_norm": 2.98681902885437, "learning_rate": 3.4270810802941256e-06, "loss": 1.2906, "step": 30759 }, { "epoch": 2.4106583072100314, "grad_norm": 3.025740385055542, "learning_rate": 3.4267883768579996e-06, "loss": 1.4222, "step": 30760 }, { "epoch": 2.4107366771159873, "grad_norm": 3.119539260864258, "learning_rate": 3.4264956794055613e-06, "loss": 1.6081, "step": 30761 }, { "epoch": 2.4108150470219436, "grad_norm": 5.695996284484863, "learning_rate": 3.426202987937926e-06, "loss": 1.4284, "step": 30762 }, { "epoch": 2.4108934169279, "grad_norm": 2.7154922485351562, "learning_rate": 3.4259103024562047e-06, "loss": 0.8468, "step": 30763 }, { "epoch": 2.410971786833856, "grad_norm": 3.5047740936279297, "learning_rate": 3.425617622961516e-06, "loss": 1.1876, "step": 30764 }, { "epoch": 2.4110501567398117, "grad_norm": 2.830888509750366, "learning_rate": 3.425324949454965e-06, "loss": 1.3896, "step": 30765 }, { "epoch": 2.411128526645768, "grad_norm": 2.3887405395507812, "learning_rate": 3.42503228193767e-06, "loss": 0.883, "step": 30766 }, { "epoch": 2.4112068965517244, "grad_norm": 3.964376926422119, "learning_rate": 3.4247396204107442e-06, "loss": 1.0247, "step": 30767 }, { "epoch": 2.4112852664576803, "grad_norm": 6.426671028137207, "learning_rate": 3.424446964875301e-06, "loss": 1.1936, "step": 30768 }, { "epoch": 2.411363636363636, "grad_norm": 3.487912178039551, "learning_rate": 3.4241543153324504e-06, "loss": 1.5589, "step": 30769 }, { "epoch": 2.4114420062695925, "grad_norm": 2.138296365737915, "learning_rate": 3.4238616717833064e-06, "loss": 1.2159, "step": 30770 }, { "epoch": 2.411520376175549, "grad_norm": 2.098083734512329, "learning_rate": 3.423569034228984e-06, "loss": 1.4858, "step": 30771 }, { "epoch": 2.4115987460815047, "grad_norm": 4.318116188049316, "learning_rate": 3.423276402670597e-06, "loss": 1.6202, "step": 30772 }, { "epoch": 2.4116771159874606, "grad_norm": 4.11168098449707, "learning_rate": 3.4229837771092554e-06, "loss": 1.4443, "step": 30773 }, { "epoch": 2.411755485893417, "grad_norm": 2.581969976425171, "learning_rate": 3.422691157546073e-06, "loss": 1.2606, "step": 30774 }, { "epoch": 2.4118338557993733, "grad_norm": 3.1994216442108154, "learning_rate": 3.422398543982165e-06, "loss": 1.3881, "step": 30775 }, { "epoch": 2.411912225705329, "grad_norm": 4.092815399169922, "learning_rate": 3.4221059364186415e-06, "loss": 1.2265, "step": 30776 }, { "epoch": 2.411990595611285, "grad_norm": 3.358314275741577, "learning_rate": 3.4218133348566153e-06, "loss": 1.5928, "step": 30777 }, { "epoch": 2.4120689655172414, "grad_norm": 1.9684138298034668, "learning_rate": 3.4215207392972016e-06, "loss": 1.3181, "step": 30778 }, { "epoch": 2.4121473354231977, "grad_norm": 4.394065856933594, "learning_rate": 3.4212281497415133e-06, "loss": 1.0884, "step": 30779 }, { "epoch": 2.4122257053291536, "grad_norm": 7.561212062835693, "learning_rate": 3.4209355661906605e-06, "loss": 0.9245, "step": 30780 }, { "epoch": 2.4123040752351095, "grad_norm": 2.6548969745635986, "learning_rate": 3.4206429886457593e-06, "loss": 1.3283, "step": 30781 }, { "epoch": 2.412382445141066, "grad_norm": 3.1515955924987793, "learning_rate": 3.4203504171079187e-06, "loss": 1.1805, "step": 30782 }, { "epoch": 2.412460815047022, "grad_norm": 3.741250514984131, "learning_rate": 3.4200578515782573e-06, "loss": 1.0516, "step": 30783 }, { "epoch": 2.412539184952978, "grad_norm": 2.959211826324463, "learning_rate": 3.4197652920578813e-06, "loss": 1.2558, "step": 30784 }, { "epoch": 2.412617554858934, "grad_norm": 2.7309730052948, "learning_rate": 3.419472738547907e-06, "loss": 1.2815, "step": 30785 }, { "epoch": 2.4126959247648903, "grad_norm": 2.572349786758423, "learning_rate": 3.4191801910494473e-06, "loss": 1.1227, "step": 30786 }, { "epoch": 2.4127742946708466, "grad_norm": 3.0415713787078857, "learning_rate": 3.418887649563615e-06, "loss": 1.2366, "step": 30787 }, { "epoch": 2.4128526645768025, "grad_norm": 2.9179396629333496, "learning_rate": 3.4185951140915207e-06, "loss": 0.856, "step": 30788 }, { "epoch": 2.412931034482759, "grad_norm": 3.343076467514038, "learning_rate": 3.418302584634277e-06, "loss": 1.4353, "step": 30789 }, { "epoch": 2.4130094043887147, "grad_norm": 4.860434532165527, "learning_rate": 3.4180100611930012e-06, "loss": 1.8364, "step": 30790 }, { "epoch": 2.413087774294671, "grad_norm": 3.2461819648742676, "learning_rate": 3.4177175437688e-06, "loss": 0.9276, "step": 30791 }, { "epoch": 2.413166144200627, "grad_norm": 4.310271739959717, "learning_rate": 3.4174250323627887e-06, "loss": 1.5417, "step": 30792 }, { "epoch": 2.4132445141065832, "grad_norm": 3.637141227722168, "learning_rate": 3.4171325269760804e-06, "loss": 1.6036, "step": 30793 }, { "epoch": 2.413322884012539, "grad_norm": 2.1555089950561523, "learning_rate": 3.416840027609788e-06, "loss": 1.1905, "step": 30794 }, { "epoch": 2.4134012539184955, "grad_norm": 3.6981022357940674, "learning_rate": 3.416547534265022e-06, "loss": 0.9193, "step": 30795 }, { "epoch": 2.4134796238244514, "grad_norm": 2.2042288780212402, "learning_rate": 3.4162550469428946e-06, "loss": 1.1873, "step": 30796 }, { "epoch": 2.4135579937304077, "grad_norm": 2.050028085708618, "learning_rate": 3.4159625656445205e-06, "loss": 1.3829, "step": 30797 }, { "epoch": 2.4136363636363636, "grad_norm": 4.515031814575195, "learning_rate": 3.415670090371012e-06, "loss": 1.355, "step": 30798 }, { "epoch": 2.41371473354232, "grad_norm": 2.7386209964752197, "learning_rate": 3.41537762112348e-06, "loss": 1.6081, "step": 30799 }, { "epoch": 2.413793103448276, "grad_norm": 5.183434963226318, "learning_rate": 3.4150851579030375e-06, "loss": 1.3799, "step": 30800 }, { "epoch": 2.413871473354232, "grad_norm": 3.719783306121826, "learning_rate": 3.4147927007107984e-06, "loss": 1.3718, "step": 30801 }, { "epoch": 2.413949843260188, "grad_norm": 2.8138744831085205, "learning_rate": 3.414500249547872e-06, "loss": 1.3449, "step": 30802 }, { "epoch": 2.4140282131661444, "grad_norm": 2.7339394092559814, "learning_rate": 3.4142078044153716e-06, "loss": 1.4812, "step": 30803 }, { "epoch": 2.4141065830721002, "grad_norm": 2.6383485794067383, "learning_rate": 3.413915365314411e-06, "loss": 1.2196, "step": 30804 }, { "epoch": 2.4141849529780566, "grad_norm": 2.581613779067993, "learning_rate": 3.4136229322461033e-06, "loss": 1.3232, "step": 30805 }, { "epoch": 2.4142633228840125, "grad_norm": 3.3927597999572754, "learning_rate": 3.413330505211557e-06, "loss": 1.4305, "step": 30806 }, { "epoch": 2.414341692789969, "grad_norm": 3.139937400817871, "learning_rate": 3.413038084211888e-06, "loss": 1.2108, "step": 30807 }, { "epoch": 2.4144200626959247, "grad_norm": 2.433522939682007, "learning_rate": 3.412745669248205e-06, "loss": 1.167, "step": 30808 }, { "epoch": 2.414498432601881, "grad_norm": 3.363628387451172, "learning_rate": 3.412453260321624e-06, "loss": 1.1842, "step": 30809 }, { "epoch": 2.414576802507837, "grad_norm": 3.731630802154541, "learning_rate": 3.4121608574332556e-06, "loss": 1.4028, "step": 30810 }, { "epoch": 2.4146551724137932, "grad_norm": 8.112195014953613, "learning_rate": 3.4118684605842107e-06, "loss": 1.7259, "step": 30811 }, { "epoch": 2.414733542319749, "grad_norm": 2.8465819358825684, "learning_rate": 3.4115760697756026e-06, "loss": 1.356, "step": 30812 }, { "epoch": 2.4148119122257055, "grad_norm": 2.472926378250122, "learning_rate": 3.4112836850085447e-06, "loss": 1.5065, "step": 30813 }, { "epoch": 2.4148902821316613, "grad_norm": 3.0614566802978516, "learning_rate": 3.410991306284146e-06, "loss": 1.1691, "step": 30814 }, { "epoch": 2.4149686520376177, "grad_norm": 2.3248074054718018, "learning_rate": 3.4106989336035195e-06, "loss": 1.4005, "step": 30815 }, { "epoch": 2.4150470219435736, "grad_norm": 5.161861419677734, "learning_rate": 3.41040656696778e-06, "loss": 1.5248, "step": 30816 }, { "epoch": 2.41512539184953, "grad_norm": 2.937446355819702, "learning_rate": 3.410114206378037e-06, "loss": 1.5855, "step": 30817 }, { "epoch": 2.415203761755486, "grad_norm": 3.1202220916748047, "learning_rate": 3.409821851835402e-06, "loss": 1.9765, "step": 30818 }, { "epoch": 2.415282131661442, "grad_norm": 2.513458251953125, "learning_rate": 3.409529503340988e-06, "loss": 1.0799, "step": 30819 }, { "epoch": 2.415360501567398, "grad_norm": 2.848853588104248, "learning_rate": 3.4092371608959085e-06, "loss": 1.2917, "step": 30820 }, { "epoch": 2.4154388714733543, "grad_norm": 3.2609429359436035, "learning_rate": 3.408944824501271e-06, "loss": 1.1669, "step": 30821 }, { "epoch": 2.4155172413793102, "grad_norm": 3.0061795711517334, "learning_rate": 3.408652494158192e-06, "loss": 1.4862, "step": 30822 }, { "epoch": 2.4155956112852666, "grad_norm": 3.5940780639648438, "learning_rate": 3.408360169867781e-06, "loss": 1.2733, "step": 30823 }, { "epoch": 2.4156739811912225, "grad_norm": 2.567406415939331, "learning_rate": 3.4080678516311518e-06, "loss": 1.5495, "step": 30824 }, { "epoch": 2.415752351097179, "grad_norm": 2.29313063621521, "learning_rate": 3.4077755394494133e-06, "loss": 1.3959, "step": 30825 }, { "epoch": 2.4158307210031347, "grad_norm": 4.259774208068848, "learning_rate": 3.407483233323678e-06, "loss": 1.1058, "step": 30826 }, { "epoch": 2.415909090909091, "grad_norm": 4.123864650726318, "learning_rate": 3.4071909332550596e-06, "loss": 1.6024, "step": 30827 }, { "epoch": 2.415987460815047, "grad_norm": 7.145596504211426, "learning_rate": 3.4068986392446703e-06, "loss": 1.331, "step": 30828 }, { "epoch": 2.4160658307210032, "grad_norm": 3.020918369293213, "learning_rate": 3.406606351293618e-06, "loss": 1.2726, "step": 30829 }, { "epoch": 2.416144200626959, "grad_norm": 2.5445642471313477, "learning_rate": 3.406314069403018e-06, "loss": 1.4536, "step": 30830 }, { "epoch": 2.4162225705329154, "grad_norm": 2.4999003410339355, "learning_rate": 3.4060217935739807e-06, "loss": 1.4072, "step": 30831 }, { "epoch": 2.4163009404388713, "grad_norm": 6.536450386047363, "learning_rate": 3.405729523807618e-06, "loss": 1.889, "step": 30832 }, { "epoch": 2.4163793103448277, "grad_norm": 3.3449089527130127, "learning_rate": 3.405437260105039e-06, "loss": 1.1726, "step": 30833 }, { "epoch": 2.4164576802507836, "grad_norm": 2.9622433185577393, "learning_rate": 3.4051450024673595e-06, "loss": 1.5103, "step": 30834 }, { "epoch": 2.41653605015674, "grad_norm": 3.4562289714813232, "learning_rate": 3.4048527508956896e-06, "loss": 1.5327, "step": 30835 }, { "epoch": 2.416614420062696, "grad_norm": 2.9376347064971924, "learning_rate": 3.40456050539114e-06, "loss": 0.7627, "step": 30836 }, { "epoch": 2.416692789968652, "grad_norm": 4.158301830291748, "learning_rate": 3.404268265954823e-06, "loss": 1.3715, "step": 30837 }, { "epoch": 2.416771159874608, "grad_norm": 2.4467148780822754, "learning_rate": 3.403976032587848e-06, "loss": 1.1931, "step": 30838 }, { "epoch": 2.4168495297805643, "grad_norm": 3.354250431060791, "learning_rate": 3.4036838052913325e-06, "loss": 1.2698, "step": 30839 }, { "epoch": 2.41692789968652, "grad_norm": 3.0231165885925293, "learning_rate": 3.40339158406638e-06, "loss": 1.4015, "step": 30840 }, { "epoch": 2.4170062695924766, "grad_norm": 2.814337968826294, "learning_rate": 3.4030993689141073e-06, "loss": 1.2749, "step": 30841 }, { "epoch": 2.4170846394984324, "grad_norm": 6.378390312194824, "learning_rate": 3.4028071598356243e-06, "loss": 1.749, "step": 30842 }, { "epoch": 2.4171630094043888, "grad_norm": 4.022000312805176, "learning_rate": 3.4025149568320435e-06, "loss": 0.9949, "step": 30843 }, { "epoch": 2.4172413793103447, "grad_norm": 3.1710941791534424, "learning_rate": 3.4022227599044744e-06, "loss": 1.2582, "step": 30844 }, { "epoch": 2.417319749216301, "grad_norm": 3.4451491832733154, "learning_rate": 3.4019305690540282e-06, "loss": 1.3794, "step": 30845 }, { "epoch": 2.417398119122257, "grad_norm": 2.093883991241455, "learning_rate": 3.4016383842818202e-06, "loss": 1.4529, "step": 30846 }, { "epoch": 2.417476489028213, "grad_norm": 9.689656257629395, "learning_rate": 3.4013462055889557e-06, "loss": 1.0066, "step": 30847 }, { "epoch": 2.417554858934169, "grad_norm": 3.33205509185791, "learning_rate": 3.4010540329765503e-06, "loss": 1.7935, "step": 30848 }, { "epoch": 2.4176332288401254, "grad_norm": 6.42038631439209, "learning_rate": 3.4007618664457143e-06, "loss": 1.1272, "step": 30849 }, { "epoch": 2.4177115987460813, "grad_norm": 2.423440933227539, "learning_rate": 3.4004697059975587e-06, "loss": 1.6893, "step": 30850 }, { "epoch": 2.4177899686520377, "grad_norm": 3.762683629989624, "learning_rate": 3.400177551633195e-06, "loss": 1.5605, "step": 30851 }, { "epoch": 2.4178683385579935, "grad_norm": 7.512711524963379, "learning_rate": 3.3998854033537328e-06, "loss": 0.5184, "step": 30852 }, { "epoch": 2.41794670846395, "grad_norm": 3.143125534057617, "learning_rate": 3.3995932611602854e-06, "loss": 1.2653, "step": 30853 }, { "epoch": 2.4180250783699058, "grad_norm": 3.0015370845794678, "learning_rate": 3.3993011250539644e-06, "loss": 1.1053, "step": 30854 }, { "epoch": 2.418103448275862, "grad_norm": 3.458087682723999, "learning_rate": 3.3990089950358783e-06, "loss": 0.8399, "step": 30855 }, { "epoch": 2.418181818181818, "grad_norm": 5.49679708480835, "learning_rate": 3.3987168711071404e-06, "loss": 1.708, "step": 30856 }, { "epoch": 2.4182601880877743, "grad_norm": 5.039045810699463, "learning_rate": 3.398424753268862e-06, "loss": 1.9012, "step": 30857 }, { "epoch": 2.4183385579937307, "grad_norm": 7.734216690063477, "learning_rate": 3.3981326415221516e-06, "loss": 1.5053, "step": 30858 }, { "epoch": 2.4184169278996865, "grad_norm": 3.9827685356140137, "learning_rate": 3.397840535868121e-06, "loss": 1.3718, "step": 30859 }, { "epoch": 2.4184952978056424, "grad_norm": 10.7486572265625, "learning_rate": 3.397548436307884e-06, "loss": 2.0824, "step": 30860 }, { "epoch": 2.4185736677115988, "grad_norm": 4.2530975341796875, "learning_rate": 3.39725634284255e-06, "loss": 1.2303, "step": 30861 }, { "epoch": 2.418652037617555, "grad_norm": 4.747603416442871, "learning_rate": 3.3969642554732287e-06, "loss": 1.6136, "step": 30862 }, { "epoch": 2.418730407523511, "grad_norm": 4.840986728668213, "learning_rate": 3.396672174201031e-06, "loss": 1.3404, "step": 30863 }, { "epoch": 2.418808777429467, "grad_norm": 2.7743358612060547, "learning_rate": 3.396380099027069e-06, "loss": 1.3404, "step": 30864 }, { "epoch": 2.418887147335423, "grad_norm": 3.7046844959259033, "learning_rate": 3.3960880299524564e-06, "loss": 1.5055, "step": 30865 }, { "epoch": 2.4189655172413795, "grad_norm": 10.319559097290039, "learning_rate": 3.395795966978298e-06, "loss": 1.0297, "step": 30866 }, { "epoch": 2.4190438871473354, "grad_norm": 2.4781196117401123, "learning_rate": 3.3955039101057084e-06, "loss": 1.16, "step": 30867 }, { "epoch": 2.4191222570532913, "grad_norm": 3.263166904449463, "learning_rate": 3.3952118593357978e-06, "loss": 1.1937, "step": 30868 }, { "epoch": 2.4192006269592476, "grad_norm": 3.8897950649261475, "learning_rate": 3.394919814669678e-06, "loss": 0.8569, "step": 30869 }, { "epoch": 2.419278996865204, "grad_norm": 2.8010900020599365, "learning_rate": 3.394627776108459e-06, "loss": 1.1503, "step": 30870 }, { "epoch": 2.41935736677116, "grad_norm": 2.9963266849517822, "learning_rate": 3.3943357436532488e-06, "loss": 1.521, "step": 30871 }, { "epoch": 2.4194357366771158, "grad_norm": 4.805141448974609, "learning_rate": 3.3940437173051645e-06, "loss": 1.7549, "step": 30872 }, { "epoch": 2.419514106583072, "grad_norm": 5.353156566619873, "learning_rate": 3.39375169706531e-06, "loss": 1.3858, "step": 30873 }, { "epoch": 2.4195924764890284, "grad_norm": 3.720284938812256, "learning_rate": 3.3934596829348e-06, "loss": 1.6994, "step": 30874 }, { "epoch": 2.4196708463949843, "grad_norm": 5.014028549194336, "learning_rate": 3.393167674914744e-06, "loss": 1.3782, "step": 30875 }, { "epoch": 2.41974921630094, "grad_norm": 5.283725261688232, "learning_rate": 3.392875673006254e-06, "loss": 1.1129, "step": 30876 }, { "epoch": 2.4198275862068965, "grad_norm": 3.4987993240356445, "learning_rate": 3.392583677210438e-06, "loss": 1.2316, "step": 30877 }, { "epoch": 2.419905956112853, "grad_norm": 3.3721628189086914, "learning_rate": 3.3922916875284077e-06, "loss": 1.2728, "step": 30878 }, { "epoch": 2.4199843260188088, "grad_norm": 6.855449199676514, "learning_rate": 3.391999703961275e-06, "loss": 1.1876, "step": 30879 }, { "epoch": 2.420062695924765, "grad_norm": 3.5599453449249268, "learning_rate": 3.3917077265101505e-06, "loss": 1.2589, "step": 30880 }, { "epoch": 2.420141065830721, "grad_norm": 3.694791078567505, "learning_rate": 3.391415755176143e-06, "loss": 1.6386, "step": 30881 }, { "epoch": 2.4202194357366773, "grad_norm": 4.954975128173828, "learning_rate": 3.391123789960363e-06, "loss": 1.0742, "step": 30882 }, { "epoch": 2.420297805642633, "grad_norm": 4.708953380584717, "learning_rate": 3.390831830863921e-06, "loss": 1.0428, "step": 30883 }, { "epoch": 2.4203761755485895, "grad_norm": 3.8430395126342773, "learning_rate": 3.390539877887931e-06, "loss": 1.2658, "step": 30884 }, { "epoch": 2.4204545454545454, "grad_norm": 2.9039459228515625, "learning_rate": 3.390247931033498e-06, "loss": 1.6085, "step": 30885 }, { "epoch": 2.4205329153605017, "grad_norm": 2.5275824069976807, "learning_rate": 3.389955990301736e-06, "loss": 1.2062, "step": 30886 }, { "epoch": 2.4206112852664576, "grad_norm": 8.57463264465332, "learning_rate": 3.3896640556937556e-06, "loss": 1.3758, "step": 30887 }, { "epoch": 2.420689655172414, "grad_norm": 4.0559234619140625, "learning_rate": 3.3893721272106646e-06, "loss": 1.446, "step": 30888 }, { "epoch": 2.42076802507837, "grad_norm": 4.032266139984131, "learning_rate": 3.3890802048535746e-06, "loss": 1.5493, "step": 30889 }, { "epoch": 2.420846394984326, "grad_norm": 1.7608637809753418, "learning_rate": 3.3887882886235958e-06, "loss": 1.279, "step": 30890 }, { "epoch": 2.420924764890282, "grad_norm": 2.165104389190674, "learning_rate": 3.3884963785218418e-06, "loss": 1.4697, "step": 30891 }, { "epoch": 2.4210031347962384, "grad_norm": 2.3586151599884033, "learning_rate": 3.388204474549416e-06, "loss": 1.3822, "step": 30892 }, { "epoch": 2.4210815047021943, "grad_norm": 2.103158950805664, "learning_rate": 3.387912576707434e-06, "loss": 1.4639, "step": 30893 }, { "epoch": 2.4211598746081506, "grad_norm": 5.856448650360107, "learning_rate": 3.3876206849970044e-06, "loss": 1.2235, "step": 30894 }, { "epoch": 2.4212382445141065, "grad_norm": 6.292014122009277, "learning_rate": 3.3873287994192382e-06, "loss": 1.2201, "step": 30895 }, { "epoch": 2.421316614420063, "grad_norm": 4.4213080406188965, "learning_rate": 3.387036919975245e-06, "loss": 1.3849, "step": 30896 }, { "epoch": 2.4213949843260187, "grad_norm": 2.1652991771698, "learning_rate": 3.3867450466661332e-06, "loss": 1.125, "step": 30897 }, { "epoch": 2.421473354231975, "grad_norm": 3.256512403488159, "learning_rate": 3.3864531794930157e-06, "loss": 1.5968, "step": 30898 }, { "epoch": 2.421551724137931, "grad_norm": 3.945255994796753, "learning_rate": 3.3861613184570026e-06, "loss": 1.3937, "step": 30899 }, { "epoch": 2.4216300940438873, "grad_norm": 3.1971516609191895, "learning_rate": 3.3858694635592013e-06, "loss": 1.1585, "step": 30900 }, { "epoch": 2.421708463949843, "grad_norm": 3.334054946899414, "learning_rate": 3.3855776148007245e-06, "loss": 1.4034, "step": 30901 }, { "epoch": 2.4217868338557995, "grad_norm": 3.0233263969421387, "learning_rate": 3.3852857721826814e-06, "loss": 1.3011, "step": 30902 }, { "epoch": 2.4218652037617554, "grad_norm": 2.708678722381592, "learning_rate": 3.384993935706181e-06, "loss": 1.1557, "step": 30903 }, { "epoch": 2.4219435736677117, "grad_norm": 3.7706336975097656, "learning_rate": 3.384702105372334e-06, "loss": 1.441, "step": 30904 }, { "epoch": 2.4220219435736676, "grad_norm": 2.498833417892456, "learning_rate": 3.384410281182251e-06, "loss": 1.0871, "step": 30905 }, { "epoch": 2.422100313479624, "grad_norm": 2.6244146823883057, "learning_rate": 3.384118463137042e-06, "loss": 1.0826, "step": 30906 }, { "epoch": 2.42217868338558, "grad_norm": 3.717442750930786, "learning_rate": 3.383826651237816e-06, "loss": 1.1448, "step": 30907 }, { "epoch": 2.422257053291536, "grad_norm": 2.604543924331665, "learning_rate": 3.383534845485683e-06, "loss": 1.4473, "step": 30908 }, { "epoch": 2.422335423197492, "grad_norm": 3.1498160362243652, "learning_rate": 3.3832430458817522e-06, "loss": 1.1594, "step": 30909 }, { "epoch": 2.4224137931034484, "grad_norm": 3.227809190750122, "learning_rate": 3.3829512524271378e-06, "loss": 1.0439, "step": 30910 }, { "epoch": 2.4224921630094043, "grad_norm": 2.383131980895996, "learning_rate": 3.3826594651229427e-06, "loss": 1.2099, "step": 30911 }, { "epoch": 2.4225705329153606, "grad_norm": 2.5032613277435303, "learning_rate": 3.382367683970281e-06, "loss": 1.7027, "step": 30912 }, { "epoch": 2.4226489028213165, "grad_norm": 3.756819248199463, "learning_rate": 3.3820759089702636e-06, "loss": 1.3878, "step": 30913 }, { "epoch": 2.422727272727273, "grad_norm": 2.9419758319854736, "learning_rate": 3.381784140123997e-06, "loss": 1.3905, "step": 30914 }, { "epoch": 2.4228056426332287, "grad_norm": 7.917984485626221, "learning_rate": 3.3814923774325924e-06, "loss": 1.757, "step": 30915 }, { "epoch": 2.422884012539185, "grad_norm": 3.7964420318603516, "learning_rate": 3.381200620897159e-06, "loss": 1.4822, "step": 30916 }, { "epoch": 2.422962382445141, "grad_norm": 2.8295934200286865, "learning_rate": 3.380908870518809e-06, "loss": 1.2673, "step": 30917 }, { "epoch": 2.4230407523510973, "grad_norm": 3.048659324645996, "learning_rate": 3.3806171262986477e-06, "loss": 1.0417, "step": 30918 }, { "epoch": 2.423119122257053, "grad_norm": 2.826841354370117, "learning_rate": 3.3803253882377882e-06, "loss": 1.3428, "step": 30919 }, { "epoch": 2.4231974921630095, "grad_norm": 2.148831605911255, "learning_rate": 3.380033656337339e-06, "loss": 1.0463, "step": 30920 }, { "epoch": 2.4232758620689654, "grad_norm": 4.1080322265625, "learning_rate": 3.3797419305984104e-06, "loss": 1.3117, "step": 30921 }, { "epoch": 2.4233542319749217, "grad_norm": 6.614323616027832, "learning_rate": 3.3794502110221096e-06, "loss": 1.9607, "step": 30922 }, { "epoch": 2.4234326018808776, "grad_norm": 2.5506725311279297, "learning_rate": 3.379158497609548e-06, "loss": 1.3135, "step": 30923 }, { "epoch": 2.423510971786834, "grad_norm": 3.1419637203216553, "learning_rate": 3.3788667903618355e-06, "loss": 0.8823, "step": 30924 }, { "epoch": 2.42358934169279, "grad_norm": 2.7589259147644043, "learning_rate": 3.3785750892800823e-06, "loss": 1.5891, "step": 30925 }, { "epoch": 2.423667711598746, "grad_norm": 2.830587863922119, "learning_rate": 3.3782833943653952e-06, "loss": 1.5177, "step": 30926 }, { "epoch": 2.423746081504702, "grad_norm": 3.8409247398376465, "learning_rate": 3.377991705618884e-06, "loss": 1.7634, "step": 30927 }, { "epoch": 2.4238244514106584, "grad_norm": 13.776883125305176, "learning_rate": 3.3777000230416635e-06, "loss": 0.8535, "step": 30928 }, { "epoch": 2.4239028213166143, "grad_norm": 2.9694900512695312, "learning_rate": 3.3774083466348347e-06, "loss": 1.1257, "step": 30929 }, { "epoch": 2.4239811912225706, "grad_norm": 10.137382507324219, "learning_rate": 3.377116676399512e-06, "loss": 0.8045, "step": 30930 }, { "epoch": 2.4240595611285265, "grad_norm": 9.2037353515625, "learning_rate": 3.376825012336804e-06, "loss": 1.4163, "step": 30931 }, { "epoch": 2.424137931034483, "grad_norm": 4.106427192687988, "learning_rate": 3.376533354447822e-06, "loss": 1.4063, "step": 30932 }, { "epoch": 2.4242163009404387, "grad_norm": 5.270320415496826, "learning_rate": 3.376241702733671e-06, "loss": 1.0738, "step": 30933 }, { "epoch": 2.424294670846395, "grad_norm": 3.4150872230529785, "learning_rate": 3.3759500571954617e-06, "loss": 1.5045, "step": 30934 }, { "epoch": 2.424373040752351, "grad_norm": 3.977591037750244, "learning_rate": 3.3756584178343054e-06, "loss": 1.4814, "step": 30935 }, { "epoch": 2.4244514106583073, "grad_norm": 4.210303783416748, "learning_rate": 3.3753667846513117e-06, "loss": 1.2316, "step": 30936 }, { "epoch": 2.424529780564263, "grad_norm": 3.0748422145843506, "learning_rate": 3.3750751576475873e-06, "loss": 1.4304, "step": 30937 }, { "epoch": 2.4246081504702195, "grad_norm": 2.1428377628326416, "learning_rate": 3.374783536824241e-06, "loss": 1.2825, "step": 30938 }, { "epoch": 2.4246865203761754, "grad_norm": 4.271507263183594, "learning_rate": 3.3744919221823837e-06, "loss": 1.6423, "step": 30939 }, { "epoch": 2.4247648902821317, "grad_norm": 2.647918939590454, "learning_rate": 3.3742003137231273e-06, "loss": 1.1872, "step": 30940 }, { "epoch": 2.4248432601880876, "grad_norm": 2.1397221088409424, "learning_rate": 3.373908711447574e-06, "loss": 1.2008, "step": 30941 }, { "epoch": 2.424921630094044, "grad_norm": 26.886638641357422, "learning_rate": 3.3736171153568377e-06, "loss": 1.8005, "step": 30942 }, { "epoch": 2.425, "grad_norm": 3.4926886558532715, "learning_rate": 3.373325525452028e-06, "loss": 1.1967, "step": 30943 }, { "epoch": 2.425078369905956, "grad_norm": 3.3100147247314453, "learning_rate": 3.3730339417342517e-06, "loss": 1.1108, "step": 30944 }, { "epoch": 2.425156739811912, "grad_norm": 3.1786224842071533, "learning_rate": 3.372742364204618e-06, "loss": 1.3652, "step": 30945 }, { "epoch": 2.4252351097178684, "grad_norm": 3.1696949005126953, "learning_rate": 3.372450792864236e-06, "loss": 1.3114, "step": 30946 }, { "epoch": 2.4253134796238243, "grad_norm": 2.750828504562378, "learning_rate": 3.372159227714218e-06, "loss": 1.2438, "step": 30947 }, { "epoch": 2.4253918495297806, "grad_norm": 2.45154070854187, "learning_rate": 3.371867668755667e-06, "loss": 1.5194, "step": 30948 }, { "epoch": 2.4254702194357365, "grad_norm": 2.193495750427246, "learning_rate": 3.3715761159896965e-06, "loss": 1.2288, "step": 30949 }, { "epoch": 2.425548589341693, "grad_norm": 2.567091703414917, "learning_rate": 3.3712845694174145e-06, "loss": 1.5486, "step": 30950 }, { "epoch": 2.4256269592476487, "grad_norm": 2.5587217807769775, "learning_rate": 3.3709930290399296e-06, "loss": 1.4883, "step": 30951 }, { "epoch": 2.425705329153605, "grad_norm": 8.587589263916016, "learning_rate": 3.37070149485835e-06, "loss": 1.6998, "step": 30952 }, { "epoch": 2.4257836990595614, "grad_norm": 6.799379825592041, "learning_rate": 3.370409966873784e-06, "loss": 1.3087, "step": 30953 }, { "epoch": 2.4258620689655173, "grad_norm": 2.6974315643310547, "learning_rate": 3.370118445087342e-06, "loss": 0.6553, "step": 30954 }, { "epoch": 2.425940438871473, "grad_norm": 2.2742507457733154, "learning_rate": 3.369826929500134e-06, "loss": 1.2549, "step": 30955 }, { "epoch": 2.4260188087774295, "grad_norm": 3.271404504776001, "learning_rate": 3.369535420113266e-06, "loss": 1.3829, "step": 30956 }, { "epoch": 2.426097178683386, "grad_norm": 7.013885974884033, "learning_rate": 3.369243916927847e-06, "loss": 1.3795, "step": 30957 }, { "epoch": 2.4261755485893417, "grad_norm": 2.8160104751586914, "learning_rate": 3.3689524199449874e-06, "loss": 1.3269, "step": 30958 }, { "epoch": 2.4262539184952976, "grad_norm": 2.292832612991333, "learning_rate": 3.368660929165795e-06, "loss": 1.1624, "step": 30959 }, { "epoch": 2.426332288401254, "grad_norm": 2.844123363494873, "learning_rate": 3.368369444591377e-06, "loss": 1.553, "step": 30960 }, { "epoch": 2.4264106583072103, "grad_norm": 6.067513942718506, "learning_rate": 3.3680779662228445e-06, "loss": 0.8067, "step": 30961 }, { "epoch": 2.426489028213166, "grad_norm": 3.0750460624694824, "learning_rate": 3.367786494061307e-06, "loss": 1.8774, "step": 30962 }, { "epoch": 2.426567398119122, "grad_norm": 2.7106094360351562, "learning_rate": 3.3674950281078693e-06, "loss": 1.2961, "step": 30963 }, { "epoch": 2.4266457680250784, "grad_norm": 2.4928348064422607, "learning_rate": 3.3672035683636426e-06, "loss": 0.7444, "step": 30964 }, { "epoch": 2.4267241379310347, "grad_norm": 3.1564972400665283, "learning_rate": 3.3669121148297336e-06, "loss": 1.3464, "step": 30965 }, { "epoch": 2.4268025078369906, "grad_norm": 5.001380920410156, "learning_rate": 3.3666206675072556e-06, "loss": 2.2583, "step": 30966 }, { "epoch": 2.4268808777429465, "grad_norm": 10.22819709777832, "learning_rate": 3.3663292263973103e-06, "loss": 0.7291, "step": 30967 }, { "epoch": 2.426959247648903, "grad_norm": 2.5371360778808594, "learning_rate": 3.3660377915010103e-06, "loss": 1.178, "step": 30968 }, { "epoch": 2.427037617554859, "grad_norm": 3.2773659229278564, "learning_rate": 3.365746362819464e-06, "loss": 1.2594, "step": 30969 }, { "epoch": 2.427115987460815, "grad_norm": 4.280852794647217, "learning_rate": 3.365454940353779e-06, "loss": 1.3185, "step": 30970 }, { "epoch": 2.427194357366771, "grad_norm": 2.5253183841705322, "learning_rate": 3.3651635241050632e-06, "loss": 1.2986, "step": 30971 }, { "epoch": 2.4272727272727272, "grad_norm": 3.7480547428131104, "learning_rate": 3.364872114074425e-06, "loss": 1.5571, "step": 30972 }, { "epoch": 2.4273510971786836, "grad_norm": 5.186378002166748, "learning_rate": 3.3645807102629763e-06, "loss": 1.3605, "step": 30973 }, { "epoch": 2.4274294670846395, "grad_norm": 3.112722158432007, "learning_rate": 3.3642893126718194e-06, "loss": 1.1934, "step": 30974 }, { "epoch": 2.427507836990596, "grad_norm": 7.963117599487305, "learning_rate": 3.3639979213020666e-06, "loss": 1.7122, "step": 30975 }, { "epoch": 2.4275862068965517, "grad_norm": 6.1837239265441895, "learning_rate": 3.363706536154826e-06, "loss": 1.2554, "step": 30976 }, { "epoch": 2.427664576802508, "grad_norm": 4.3899922370910645, "learning_rate": 3.3634151572312066e-06, "loss": 1.8843, "step": 30977 }, { "epoch": 2.427742946708464, "grad_norm": 5.243505001068115, "learning_rate": 3.363123784532314e-06, "loss": 1.6538, "step": 30978 }, { "epoch": 2.4278213166144202, "grad_norm": 2.7500710487365723, "learning_rate": 3.3628324180592566e-06, "loss": 1.2473, "step": 30979 }, { "epoch": 2.427899686520376, "grad_norm": 7.329204559326172, "learning_rate": 3.3625410578131446e-06, "loss": 1.135, "step": 30980 }, { "epoch": 2.4279780564263325, "grad_norm": 2.7060768604278564, "learning_rate": 3.3622497037950863e-06, "loss": 1.4974, "step": 30981 }, { "epoch": 2.4280564263322884, "grad_norm": 3.2271997928619385, "learning_rate": 3.361958356006188e-06, "loss": 1.3148, "step": 30982 }, { "epoch": 2.4281347962382447, "grad_norm": 4.262068271636963, "learning_rate": 3.3616670144475594e-06, "loss": 1.3457, "step": 30983 }, { "epoch": 2.4282131661442006, "grad_norm": 5.0369791984558105, "learning_rate": 3.361375679120308e-06, "loss": 1.5082, "step": 30984 }, { "epoch": 2.428291536050157, "grad_norm": 3.087677240371704, "learning_rate": 3.361084350025542e-06, "loss": 1.3083, "step": 30985 }, { "epoch": 2.428369905956113, "grad_norm": 2.1368143558502197, "learning_rate": 3.3607930271643675e-06, "loss": 1.6227, "step": 30986 }, { "epoch": 2.428448275862069, "grad_norm": 2.0159568786621094, "learning_rate": 3.360501710537895e-06, "loss": 0.9741, "step": 30987 }, { "epoch": 2.428526645768025, "grad_norm": 2.171900510787964, "learning_rate": 3.3602104001472334e-06, "loss": 1.1073, "step": 30988 }, { "epoch": 2.4286050156739813, "grad_norm": 2.4412693977355957, "learning_rate": 3.3599190959934885e-06, "loss": 1.2836, "step": 30989 }, { "epoch": 2.4286833855799372, "grad_norm": 3.8529138565063477, "learning_rate": 3.3596277980777685e-06, "loss": 1.4621, "step": 30990 }, { "epoch": 2.4287617554858936, "grad_norm": 5.006011962890625, "learning_rate": 3.3593365064011813e-06, "loss": 0.8711, "step": 30991 }, { "epoch": 2.4288401253918495, "grad_norm": 4.192854404449463, "learning_rate": 3.3590452209648384e-06, "loss": 1.3007, "step": 30992 }, { "epoch": 2.428918495297806, "grad_norm": 3.200263500213623, "learning_rate": 3.358753941769841e-06, "loss": 1.2913, "step": 30993 }, { "epoch": 2.4289968652037617, "grad_norm": 2.4526097774505615, "learning_rate": 3.3584626688173015e-06, "loss": 1.0796, "step": 30994 }, { "epoch": 2.429075235109718, "grad_norm": 8.58755874633789, "learning_rate": 3.358171402108327e-06, "loss": 1.1383, "step": 30995 }, { "epoch": 2.429153605015674, "grad_norm": 3.3799524307250977, "learning_rate": 3.3578801416440266e-06, "loss": 1.4036, "step": 30996 }, { "epoch": 2.4292319749216302, "grad_norm": 3.1410114765167236, "learning_rate": 3.357588887425506e-06, "loss": 1.6546, "step": 30997 }, { "epoch": 2.429310344827586, "grad_norm": 2.973477840423584, "learning_rate": 3.357297639453872e-06, "loss": 1.1634, "step": 30998 }, { "epoch": 2.4293887147335425, "grad_norm": 4.343410491943359, "learning_rate": 3.357006397730237e-06, "loss": 1.3579, "step": 30999 }, { "epoch": 2.4294670846394983, "grad_norm": 6.743935585021973, "learning_rate": 3.3567151622557033e-06, "loss": 1.2761, "step": 31000 }, { "epoch": 2.4295454545454547, "grad_norm": 6.221729755401611, "learning_rate": 3.3564239330313818e-06, "loss": 1.2383, "step": 31001 }, { "epoch": 2.4296238244514106, "grad_norm": 3.405942678451538, "learning_rate": 3.3561327100583794e-06, "loss": 1.6867, "step": 31002 }, { "epoch": 2.429702194357367, "grad_norm": 3.8841989040374756, "learning_rate": 3.3558414933378045e-06, "loss": 1.4025, "step": 31003 }, { "epoch": 2.429780564263323, "grad_norm": 4.964128494262695, "learning_rate": 3.355550282870764e-06, "loss": 1.1188, "step": 31004 }, { "epoch": 2.429858934169279, "grad_norm": 2.5715577602386475, "learning_rate": 3.355259078658364e-06, "loss": 1.331, "step": 31005 }, { "epoch": 2.429937304075235, "grad_norm": 3.149599552154541, "learning_rate": 3.3549678807017143e-06, "loss": 0.9392, "step": 31006 }, { "epoch": 2.4300156739811913, "grad_norm": 5.218488693237305, "learning_rate": 3.3546766890019235e-06, "loss": 0.9834, "step": 31007 }, { "epoch": 2.4300940438871472, "grad_norm": 2.8764305114746094, "learning_rate": 3.354385503560096e-06, "loss": 1.4177, "step": 31008 }, { "epoch": 2.4301724137931036, "grad_norm": 3.2971112728118896, "learning_rate": 3.3540943243773413e-06, "loss": 1.0919, "step": 31009 }, { "epoch": 2.4302507836990594, "grad_norm": 2.4847404956817627, "learning_rate": 3.353803151454767e-06, "loss": 1.0754, "step": 31010 }, { "epoch": 2.430329153605016, "grad_norm": 3.6030867099761963, "learning_rate": 3.3535119847934793e-06, "loss": 1.1533, "step": 31011 }, { "epoch": 2.4304075235109717, "grad_norm": 5.697464942932129, "learning_rate": 3.3532208243945845e-06, "loss": 1.5094, "step": 31012 }, { "epoch": 2.430485893416928, "grad_norm": 5.758935928344727, "learning_rate": 3.352929670259194e-06, "loss": 1.4731, "step": 31013 }, { "epoch": 2.430564263322884, "grad_norm": 3.0081188678741455, "learning_rate": 3.352638522388413e-06, "loss": 1.1928, "step": 31014 }, { "epoch": 2.43064263322884, "grad_norm": 4.700745105743408, "learning_rate": 3.3523473807833483e-06, "loss": 1.4169, "step": 31015 }, { "epoch": 2.430721003134796, "grad_norm": 3.4831066131591797, "learning_rate": 3.3520562454451082e-06, "loss": 1.328, "step": 31016 }, { "epoch": 2.4307993730407524, "grad_norm": 2.4688289165496826, "learning_rate": 3.351765116374799e-06, "loss": 1.3914, "step": 31017 }, { "epoch": 2.4308777429467083, "grad_norm": 3.6999130249023438, "learning_rate": 3.35147399357353e-06, "loss": 1.238, "step": 31018 }, { "epoch": 2.4309561128526647, "grad_norm": 2.895495891571045, "learning_rate": 3.351182877042406e-06, "loss": 1.2986, "step": 31019 }, { "epoch": 2.4310344827586206, "grad_norm": 3.1296002864837646, "learning_rate": 3.350891766782536e-06, "loss": 0.7695, "step": 31020 }, { "epoch": 2.431112852664577, "grad_norm": 3.857196807861328, "learning_rate": 3.3506006627950266e-06, "loss": 0.7459, "step": 31021 }, { "epoch": 2.4311912225705328, "grad_norm": 3.0423877239227295, "learning_rate": 3.350309565080986e-06, "loss": 1.4187, "step": 31022 }, { "epoch": 2.431269592476489, "grad_norm": 3.3272573947906494, "learning_rate": 3.3500184736415187e-06, "loss": 1.7756, "step": 31023 }, { "epoch": 2.431347962382445, "grad_norm": 5.134145736694336, "learning_rate": 3.3497273884777343e-06, "loss": 0.9045, "step": 31024 }, { "epoch": 2.4314263322884013, "grad_norm": 3.7997376918792725, "learning_rate": 3.3494363095907402e-06, "loss": 1.0915, "step": 31025 }, { "epoch": 2.431504702194357, "grad_norm": 2.5132100582122803, "learning_rate": 3.3491452369816426e-06, "loss": 1.1544, "step": 31026 }, { "epoch": 2.4315830721003135, "grad_norm": 2.1847589015960693, "learning_rate": 3.3488541706515477e-06, "loss": 1.326, "step": 31027 }, { "epoch": 2.4316614420062694, "grad_norm": 3.0364487171173096, "learning_rate": 3.348563110601563e-06, "loss": 1.2113, "step": 31028 }, { "epoch": 2.4317398119122258, "grad_norm": 3.16334867477417, "learning_rate": 3.3482720568327973e-06, "loss": 1.3556, "step": 31029 }, { "epoch": 2.4318181818181817, "grad_norm": 3.7335312366485596, "learning_rate": 3.3479810093463547e-06, "loss": 1.8973, "step": 31030 }, { "epoch": 2.431896551724138, "grad_norm": 16.432764053344727, "learning_rate": 3.347689968143345e-06, "loss": 1.1445, "step": 31031 }, { "epoch": 2.431974921630094, "grad_norm": 3.324352264404297, "learning_rate": 3.3473989332248734e-06, "loss": 1.6015, "step": 31032 }, { "epoch": 2.43205329153605, "grad_norm": 6.600794792175293, "learning_rate": 3.347107904592049e-06, "loss": 1.5218, "step": 31033 }, { "epoch": 2.432131661442006, "grad_norm": 8.145081520080566, "learning_rate": 3.3468168822459757e-06, "loss": 1.4526, "step": 31034 }, { "epoch": 2.4322100313479624, "grad_norm": 7.117922306060791, "learning_rate": 3.346525866187761e-06, "loss": 1.1928, "step": 31035 }, { "epoch": 2.4322884012539183, "grad_norm": 4.47233772277832, "learning_rate": 3.346234856418514e-06, "loss": 1.4178, "step": 31036 }, { "epoch": 2.4323667711598747, "grad_norm": 4.610258102416992, "learning_rate": 3.3459438529393408e-06, "loss": 1.6353, "step": 31037 }, { "epoch": 2.4324451410658305, "grad_norm": 3.4065492153167725, "learning_rate": 3.3456528557513464e-06, "loss": 1.5223, "step": 31038 }, { "epoch": 2.432523510971787, "grad_norm": 3.2823877334594727, "learning_rate": 3.345361864855639e-06, "loss": 1.1335, "step": 31039 }, { "epoch": 2.4326018808777428, "grad_norm": 3.06573748588562, "learning_rate": 3.345070880253326e-06, "loss": 1.389, "step": 31040 }, { "epoch": 2.432680250783699, "grad_norm": 2.888828992843628, "learning_rate": 3.3447799019455128e-06, "loss": 1.2256, "step": 31041 }, { "epoch": 2.432758620689655, "grad_norm": 9.628682136535645, "learning_rate": 3.344488929933305e-06, "loss": 1.4962, "step": 31042 }, { "epoch": 2.4328369905956113, "grad_norm": 6.886195182800293, "learning_rate": 3.3441979642178123e-06, "loss": 1.2009, "step": 31043 }, { "epoch": 2.4329153605015676, "grad_norm": 9.720685958862305, "learning_rate": 3.343907004800142e-06, "loss": 1.2926, "step": 31044 }, { "epoch": 2.4329937304075235, "grad_norm": 3.2988550662994385, "learning_rate": 3.3436160516813964e-06, "loss": 1.6238, "step": 31045 }, { "epoch": 2.4330721003134794, "grad_norm": 3.127330780029297, "learning_rate": 3.343325104862686e-06, "loss": 1.2276, "step": 31046 }, { "epoch": 2.4331504702194358, "grad_norm": 3.7338614463806152, "learning_rate": 3.3430341643451137e-06, "loss": 1.4929, "step": 31047 }, { "epoch": 2.433228840125392, "grad_norm": 3.8518362045288086, "learning_rate": 3.342743230129792e-06, "loss": 1.294, "step": 31048 }, { "epoch": 2.433307210031348, "grad_norm": 3.0043656826019287, "learning_rate": 3.34245230221782e-06, "loss": 1.2723, "step": 31049 }, { "epoch": 2.433385579937304, "grad_norm": 2.9334874153137207, "learning_rate": 3.3421613806103106e-06, "loss": 1.4382, "step": 31050 }, { "epoch": 2.43346394984326, "grad_norm": 3.0595285892486572, "learning_rate": 3.3418704653083666e-06, "loss": 1.8535, "step": 31051 }, { "epoch": 2.4335423197492165, "grad_norm": 2.1112968921661377, "learning_rate": 3.341579556313097e-06, "loss": 1.2408, "step": 31052 }, { "epoch": 2.4336206896551724, "grad_norm": 3.0428340435028076, "learning_rate": 3.341288653625606e-06, "loss": 1.5926, "step": 31053 }, { "epoch": 2.4336990595611283, "grad_norm": 2.439160108566284, "learning_rate": 3.340997757247e-06, "loss": 1.3211, "step": 31054 }, { "epoch": 2.4337774294670846, "grad_norm": 2.452028274536133, "learning_rate": 3.3407068671783895e-06, "loss": 1.4934, "step": 31055 }, { "epoch": 2.433855799373041, "grad_norm": 2.7687201499938965, "learning_rate": 3.3404159834208745e-06, "loss": 1.3674, "step": 31056 }, { "epoch": 2.433934169278997, "grad_norm": 3.4266135692596436, "learning_rate": 3.340125105975566e-06, "loss": 1.2264, "step": 31057 }, { "epoch": 2.4340125391849528, "grad_norm": 2.963757276535034, "learning_rate": 3.3398342348435687e-06, "loss": 1.2716, "step": 31058 }, { "epoch": 2.434090909090909, "grad_norm": 2.9177820682525635, "learning_rate": 3.3395433700259903e-06, "loss": 1.4885, "step": 31059 }, { "epoch": 2.4341692789968654, "grad_norm": 4.623656749725342, "learning_rate": 3.3392525115239353e-06, "loss": 0.9787, "step": 31060 }, { "epoch": 2.4342476489028213, "grad_norm": 2.655229330062866, "learning_rate": 3.33896165933851e-06, "loss": 1.4362, "step": 31061 }, { "epoch": 2.434326018808777, "grad_norm": 5.942313194274902, "learning_rate": 3.338670813470822e-06, "loss": 0.7831, "step": 31062 }, { "epoch": 2.4344043887147335, "grad_norm": 6.843887805938721, "learning_rate": 3.338379973921978e-06, "loss": 1.472, "step": 31063 }, { "epoch": 2.43448275862069, "grad_norm": 3.078160285949707, "learning_rate": 3.3380891406930816e-06, "loss": 1.1757, "step": 31064 }, { "epoch": 2.4345611285266457, "grad_norm": 5.931131362915039, "learning_rate": 3.337798313785241e-06, "loss": 1.211, "step": 31065 }, { "epoch": 2.4346394984326016, "grad_norm": 2.435800313949585, "learning_rate": 3.3375074931995633e-06, "loss": 0.6563, "step": 31066 }, { "epoch": 2.434717868338558, "grad_norm": 2.249626636505127, "learning_rate": 3.3372166789371515e-06, "loss": 1.2261, "step": 31067 }, { "epoch": 2.4347962382445143, "grad_norm": 4.092029571533203, "learning_rate": 3.3369258709991124e-06, "loss": 1.4258, "step": 31068 }, { "epoch": 2.43487460815047, "grad_norm": 2.9211649894714355, "learning_rate": 3.336635069386554e-06, "loss": 1.3045, "step": 31069 }, { "epoch": 2.4349529780564265, "grad_norm": 2.6882741451263428, "learning_rate": 3.3363442741005826e-06, "loss": 1.3795, "step": 31070 }, { "epoch": 2.4350313479623824, "grad_norm": 8.472833633422852, "learning_rate": 3.3360534851423022e-06, "loss": 1.0248, "step": 31071 }, { "epoch": 2.4351097178683387, "grad_norm": 3.649784564971924, "learning_rate": 3.3357627025128203e-06, "loss": 1.0895, "step": 31072 }, { "epoch": 2.4351880877742946, "grad_norm": 2.8287596702575684, "learning_rate": 3.3354719262132402e-06, "loss": 1.8554, "step": 31073 }, { "epoch": 2.435266457680251, "grad_norm": 4.4633026123046875, "learning_rate": 3.3351811562446736e-06, "loss": 1.8539, "step": 31074 }, { "epoch": 2.435344827586207, "grad_norm": 3.5522022247314453, "learning_rate": 3.3348903926082195e-06, "loss": 1.7132, "step": 31075 }, { "epoch": 2.435423197492163, "grad_norm": 2.5635550022125244, "learning_rate": 3.334599635304988e-06, "loss": 1.5112, "step": 31076 }, { "epoch": 2.435501567398119, "grad_norm": 2.8682773113250732, "learning_rate": 3.3343088843360836e-06, "loss": 1.4448, "step": 31077 }, { "epoch": 2.4355799373040754, "grad_norm": 3.8019676208496094, "learning_rate": 3.334018139702615e-06, "loss": 1.4881, "step": 31078 }, { "epoch": 2.4356583072100313, "grad_norm": 3.954756259918213, "learning_rate": 3.333727401405684e-06, "loss": 1.357, "step": 31079 }, { "epoch": 2.4357366771159876, "grad_norm": 2.4768428802490234, "learning_rate": 3.333436669446397e-06, "loss": 1.3614, "step": 31080 }, { "epoch": 2.4358150470219435, "grad_norm": 2.54264760017395, "learning_rate": 3.333145943825864e-06, "loss": 1.5351, "step": 31081 }, { "epoch": 2.4358934169279, "grad_norm": 7.775091648101807, "learning_rate": 3.3328552245451844e-06, "loss": 1.2114, "step": 31082 }, { "epoch": 2.4359717868338557, "grad_norm": 3.4515540599823, "learning_rate": 3.3325645116054682e-06, "loss": 0.9523, "step": 31083 }, { "epoch": 2.436050156739812, "grad_norm": 3.149061679840088, "learning_rate": 3.3322738050078207e-06, "loss": 1.2696, "step": 31084 }, { "epoch": 2.436128526645768, "grad_norm": 5.960155487060547, "learning_rate": 3.331983104753348e-06, "loss": 1.2126, "step": 31085 }, { "epoch": 2.4362068965517243, "grad_norm": 4.226675987243652, "learning_rate": 3.3316924108431535e-06, "loss": 0.6817, "step": 31086 }, { "epoch": 2.43628526645768, "grad_norm": 9.770063400268555, "learning_rate": 3.3314017232783435e-06, "loss": 2.3175, "step": 31087 }, { "epoch": 2.4363636363636365, "grad_norm": 15.44701099395752, "learning_rate": 3.3311110420600244e-06, "loss": 1.3611, "step": 31088 }, { "epoch": 2.4364420062695924, "grad_norm": 3.707242012023926, "learning_rate": 3.3308203671893036e-06, "loss": 1.1241, "step": 31089 }, { "epoch": 2.4365203761755487, "grad_norm": 2.964766502380371, "learning_rate": 3.330529698667284e-06, "loss": 1.3854, "step": 31090 }, { "epoch": 2.4365987460815046, "grad_norm": 3.9799089431762695, "learning_rate": 3.3302390364950713e-06, "loss": 0.9297, "step": 31091 }, { "epoch": 2.436677115987461, "grad_norm": 4.0671491622924805, "learning_rate": 3.329948380673771e-06, "loss": 1.1406, "step": 31092 }, { "epoch": 2.436755485893417, "grad_norm": 2.6924424171447754, "learning_rate": 3.329657731204492e-06, "loss": 1.2289, "step": 31093 }, { "epoch": 2.436833855799373, "grad_norm": 2.770908832550049, "learning_rate": 3.3293670880883344e-06, "loss": 1.4923, "step": 31094 }, { "epoch": 2.436912225705329, "grad_norm": 3.3929367065429688, "learning_rate": 3.329076451326407e-06, "loss": 1.4388, "step": 31095 }, { "epoch": 2.4369905956112854, "grad_norm": 6.947501182556152, "learning_rate": 3.3287858209198155e-06, "loss": 1.4614, "step": 31096 }, { "epoch": 2.4370689655172413, "grad_norm": 4.5987019538879395, "learning_rate": 3.3284951968696635e-06, "loss": 1.7864, "step": 31097 }, { "epoch": 2.4371473354231976, "grad_norm": 4.220631122589111, "learning_rate": 3.3282045791770574e-06, "loss": 1.0245, "step": 31098 }, { "epoch": 2.4372257053291535, "grad_norm": 3.191331624984741, "learning_rate": 3.327913967843101e-06, "loss": 1.5931, "step": 31099 }, { "epoch": 2.43730407523511, "grad_norm": 3.8148252964019775, "learning_rate": 3.327623362868904e-06, "loss": 1.3549, "step": 31100 }, { "epoch": 2.4373824451410657, "grad_norm": 3.964148759841919, "learning_rate": 3.3273327642555665e-06, "loss": 1.5691, "step": 31101 }, { "epoch": 2.437460815047022, "grad_norm": 2.2310590744018555, "learning_rate": 3.3270421720041964e-06, "loss": 1.454, "step": 31102 }, { "epoch": 2.437539184952978, "grad_norm": 3.447871446609497, "learning_rate": 3.3267515861158982e-06, "loss": 1.4998, "step": 31103 }, { "epoch": 2.4376175548589343, "grad_norm": 3.5768697261810303, "learning_rate": 3.326461006591779e-06, "loss": 1.416, "step": 31104 }, { "epoch": 2.43769592476489, "grad_norm": 4.463326930999756, "learning_rate": 3.3261704334329416e-06, "loss": 1.2853, "step": 31105 }, { "epoch": 2.4377742946708465, "grad_norm": 6.303543567657471, "learning_rate": 3.3258798666404914e-06, "loss": 1.0375, "step": 31106 }, { "epoch": 2.4378526645768024, "grad_norm": 6.982903957366943, "learning_rate": 3.3255893062155347e-06, "loss": 1.3647, "step": 31107 }, { "epoch": 2.4379310344827587, "grad_norm": 4.8715901374816895, "learning_rate": 3.3252987521591777e-06, "loss": 1.539, "step": 31108 }, { "epoch": 2.4380094043887146, "grad_norm": 3.2574281692504883, "learning_rate": 3.325008204472523e-06, "loss": 1.2691, "step": 31109 }, { "epoch": 2.438087774294671, "grad_norm": 11.593944549560547, "learning_rate": 3.324717663156677e-06, "loss": 0.9432, "step": 31110 }, { "epoch": 2.438166144200627, "grad_norm": 5.0048627853393555, "learning_rate": 3.324427128212745e-06, "loss": 1.5324, "step": 31111 }, { "epoch": 2.438244514106583, "grad_norm": 2.157341957092285, "learning_rate": 3.3241365996418307e-06, "loss": 1.009, "step": 31112 }, { "epoch": 2.438322884012539, "grad_norm": 3.3560872077941895, "learning_rate": 3.3238460774450387e-06, "loss": 1.3711, "step": 31113 }, { "epoch": 2.4384012539184954, "grad_norm": 2.149451971054077, "learning_rate": 3.323555561623477e-06, "loss": 1.2053, "step": 31114 }, { "epoch": 2.4384796238244513, "grad_norm": 2.7966866493225098, "learning_rate": 3.3232650521782495e-06, "loss": 1.4668, "step": 31115 }, { "epoch": 2.4385579937304076, "grad_norm": 4.045960903167725, "learning_rate": 3.322974549110459e-06, "loss": 1.0429, "step": 31116 }, { "epoch": 2.4386363636363635, "grad_norm": 2.0501694679260254, "learning_rate": 3.3226840524212135e-06, "loss": 1.3158, "step": 31117 }, { "epoch": 2.43871473354232, "grad_norm": 2.879695415496826, "learning_rate": 3.322393562111614e-06, "loss": 1.226, "step": 31118 }, { "epoch": 2.4387931034482757, "grad_norm": 2.9330012798309326, "learning_rate": 3.32210307818277e-06, "loss": 1.129, "step": 31119 }, { "epoch": 2.438871473354232, "grad_norm": 3.1309492588043213, "learning_rate": 3.321812600635783e-06, "loss": 1.3538, "step": 31120 }, { "epoch": 2.438949843260188, "grad_norm": 3.098963975906372, "learning_rate": 3.3215221294717592e-06, "loss": 1.3404, "step": 31121 }, { "epoch": 2.4390282131661443, "grad_norm": 1.9882302284240723, "learning_rate": 3.321231664691804e-06, "loss": 1.4403, "step": 31122 }, { "epoch": 2.4391065830721, "grad_norm": 3.786968231201172, "learning_rate": 3.32094120629702e-06, "loss": 1.4126, "step": 31123 }, { "epoch": 2.4391849529780565, "grad_norm": 4.0387468338012695, "learning_rate": 3.3206507542885127e-06, "loss": 1.6648, "step": 31124 }, { "epoch": 2.4392633228840124, "grad_norm": 4.276678562164307, "learning_rate": 3.3203603086673874e-06, "loss": 1.5885, "step": 31125 }, { "epoch": 2.4393416927899687, "grad_norm": 2.3670802116394043, "learning_rate": 3.320069869434751e-06, "loss": 1.167, "step": 31126 }, { "epoch": 2.4394200626959246, "grad_norm": 2.7681851387023926, "learning_rate": 3.319779436591704e-06, "loss": 1.2966, "step": 31127 }, { "epoch": 2.439498432601881, "grad_norm": 3.994192123413086, "learning_rate": 3.319489010139354e-06, "loss": 1.1175, "step": 31128 }, { "epoch": 2.439576802507837, "grad_norm": 3.2741711139678955, "learning_rate": 3.319198590078804e-06, "loss": 1.3802, "step": 31129 }, { "epoch": 2.439655172413793, "grad_norm": 4.773082733154297, "learning_rate": 3.3189081764111597e-06, "loss": 1.7205, "step": 31130 }, { "epoch": 2.439733542319749, "grad_norm": 2.280320167541504, "learning_rate": 3.318617769137524e-06, "loss": 1.1324, "step": 31131 }, { "epoch": 2.4398119122257054, "grad_norm": 4.415595054626465, "learning_rate": 3.3183273682590034e-06, "loss": 1.2942, "step": 31132 }, { "epoch": 2.4398902821316613, "grad_norm": 5.50306510925293, "learning_rate": 3.318036973776702e-06, "loss": 0.7472, "step": 31133 }, { "epoch": 2.4399686520376176, "grad_norm": 2.7429862022399902, "learning_rate": 3.3177465856917247e-06, "loss": 1.2753, "step": 31134 }, { "epoch": 2.4400470219435735, "grad_norm": 2.343273878097534, "learning_rate": 3.317456204005175e-06, "loss": 1.2883, "step": 31135 }, { "epoch": 2.44012539184953, "grad_norm": 4.699585437774658, "learning_rate": 3.3171658287181564e-06, "loss": 0.524, "step": 31136 }, { "epoch": 2.4402037617554857, "grad_norm": 6.418943881988525, "learning_rate": 3.316875459831777e-06, "loss": 1.334, "step": 31137 }, { "epoch": 2.440282131661442, "grad_norm": 2.825753688812256, "learning_rate": 3.3165850973471363e-06, "loss": 0.9313, "step": 31138 }, { "epoch": 2.4403605015673984, "grad_norm": 2.9799811840057373, "learning_rate": 3.316294741265342e-06, "loss": 1.2667, "step": 31139 }, { "epoch": 2.4404388714733543, "grad_norm": 3.261270523071289, "learning_rate": 3.3160043915874983e-06, "loss": 0.9087, "step": 31140 }, { "epoch": 2.44051724137931, "grad_norm": 2.182360887527466, "learning_rate": 3.315714048314709e-06, "loss": 1.184, "step": 31141 }, { "epoch": 2.4405956112852665, "grad_norm": 4.722425937652588, "learning_rate": 3.315423711448078e-06, "loss": 1.5593, "step": 31142 }, { "epoch": 2.440673981191223, "grad_norm": 7.92122745513916, "learning_rate": 3.315133380988709e-06, "loss": 0.9874, "step": 31143 }, { "epoch": 2.4407523510971787, "grad_norm": 3.818974018096924, "learning_rate": 3.314843056937708e-06, "loss": 1.3565, "step": 31144 }, { "epoch": 2.4408307210031346, "grad_norm": 2.4388253688812256, "learning_rate": 3.3145527392961797e-06, "loss": 1.4354, "step": 31145 }, { "epoch": 2.440909090909091, "grad_norm": 2.710637331008911, "learning_rate": 3.3142624280652256e-06, "loss": 1.3641, "step": 31146 }, { "epoch": 2.4409874608150472, "grad_norm": 7.15130090713501, "learning_rate": 3.3139721232459523e-06, "loss": 1.2246, "step": 31147 }, { "epoch": 2.441065830721003, "grad_norm": 2.8642966747283936, "learning_rate": 3.3136818248394607e-06, "loss": 1.6867, "step": 31148 }, { "epoch": 2.441144200626959, "grad_norm": 2.6638312339782715, "learning_rate": 3.3133915328468615e-06, "loss": 1.2659, "step": 31149 }, { "epoch": 2.4412225705329154, "grad_norm": 2.233396291732788, "learning_rate": 3.3131012472692515e-06, "loss": 1.108, "step": 31150 }, { "epoch": 2.4413009404388717, "grad_norm": 3.737687826156616, "learning_rate": 3.312810968107738e-06, "loss": 1.6253, "step": 31151 }, { "epoch": 2.4413793103448276, "grad_norm": 4.232590198516846, "learning_rate": 3.312520695363427e-06, "loss": 1.6324, "step": 31152 }, { "epoch": 2.4414576802507835, "grad_norm": 3.3430373668670654, "learning_rate": 3.3122304290374185e-06, "loss": 1.2632, "step": 31153 }, { "epoch": 2.44153605015674, "grad_norm": 2.6042747497558594, "learning_rate": 3.3119401691308195e-06, "loss": 1.7375, "step": 31154 }, { "epoch": 2.441614420062696, "grad_norm": 2.8158066272735596, "learning_rate": 3.311649915644731e-06, "loss": 1.2626, "step": 31155 }, { "epoch": 2.441692789968652, "grad_norm": 8.156336784362793, "learning_rate": 3.3113596685802633e-06, "loss": 1.1807, "step": 31156 }, { "epoch": 2.441771159874608, "grad_norm": 7.234924793243408, "learning_rate": 3.3110694279385126e-06, "loss": 1.2656, "step": 31157 }, { "epoch": 2.4418495297805642, "grad_norm": 4.569512844085693, "learning_rate": 3.310779193720587e-06, "loss": 1.3678, "step": 31158 }, { "epoch": 2.4419278996865206, "grad_norm": 4.349566459655762, "learning_rate": 3.31048896592759e-06, "loss": 0.9448, "step": 31159 }, { "epoch": 2.4420062695924765, "grad_norm": 3.3786587715148926, "learning_rate": 3.3101987445606263e-06, "loss": 1.3371, "step": 31160 }, { "epoch": 2.4420846394984324, "grad_norm": 2.878662347793579, "learning_rate": 3.3099085296207974e-06, "loss": 1.3617, "step": 31161 }, { "epoch": 2.4421630094043887, "grad_norm": 4.555577278137207, "learning_rate": 3.309618321109207e-06, "loss": 1.5158, "step": 31162 }, { "epoch": 2.442241379310345, "grad_norm": 4.780226707458496, "learning_rate": 3.309328119026962e-06, "loss": 1.5306, "step": 31163 }, { "epoch": 2.442319749216301, "grad_norm": 3.8577959537506104, "learning_rate": 3.3090379233751653e-06, "loss": 1.4793, "step": 31164 }, { "epoch": 2.4423981191222572, "grad_norm": 3.244053602218628, "learning_rate": 3.3087477341549184e-06, "loss": 1.4155, "step": 31165 }, { "epoch": 2.442476489028213, "grad_norm": 7.968283176422119, "learning_rate": 3.308457551367327e-06, "loss": 1.1463, "step": 31166 }, { "epoch": 2.4425548589341695, "grad_norm": 3.0237205028533936, "learning_rate": 3.3081673750134948e-06, "loss": 1.1522, "step": 31167 }, { "epoch": 2.4426332288401253, "grad_norm": 6.897487640380859, "learning_rate": 3.307877205094524e-06, "loss": 0.9004, "step": 31168 }, { "epoch": 2.4427115987460817, "grad_norm": 2.5139737129211426, "learning_rate": 3.307587041611518e-06, "loss": 1.4874, "step": 31169 }, { "epoch": 2.4427899686520376, "grad_norm": 3.9868452548980713, "learning_rate": 3.3072968845655827e-06, "loss": 1.4225, "step": 31170 }, { "epoch": 2.442868338557994, "grad_norm": 3.1679673194885254, "learning_rate": 3.3070067339578215e-06, "loss": 1.1166, "step": 31171 }, { "epoch": 2.44294670846395, "grad_norm": 3.441610813140869, "learning_rate": 3.3067165897893364e-06, "loss": 1.5794, "step": 31172 }, { "epoch": 2.443025078369906, "grad_norm": 3.2622385025024414, "learning_rate": 3.306426452061232e-06, "loss": 1.4256, "step": 31173 }, { "epoch": 2.443103448275862, "grad_norm": 3.9172909259796143, "learning_rate": 3.30613632077461e-06, "loss": 1.1917, "step": 31174 }, { "epoch": 2.4431818181818183, "grad_norm": 3.559922218322754, "learning_rate": 3.305846195930579e-06, "loss": 1.5417, "step": 31175 }, { "epoch": 2.4432601880877742, "grad_norm": 2.755392074584961, "learning_rate": 3.3055560775302354e-06, "loss": 1.4622, "step": 31176 }, { "epoch": 2.4433385579937306, "grad_norm": 7.313881874084473, "learning_rate": 3.3052659655746865e-06, "loss": 1.1853, "step": 31177 }, { "epoch": 2.4434169278996865, "grad_norm": 5.998321056365967, "learning_rate": 3.304975860065037e-06, "loss": 1.4377, "step": 31178 }, { "epoch": 2.443495297805643, "grad_norm": 3.2264983654022217, "learning_rate": 3.3046857610023884e-06, "loss": 1.3729, "step": 31179 }, { "epoch": 2.4435736677115987, "grad_norm": 2.4686923027038574, "learning_rate": 3.3043956683878437e-06, "loss": 1.2611, "step": 31180 }, { "epoch": 2.443652037617555, "grad_norm": 4.375547409057617, "learning_rate": 3.3041055822225058e-06, "loss": 1.5078, "step": 31181 }, { "epoch": 2.443730407523511, "grad_norm": 4.47651481628418, "learning_rate": 3.3038155025074826e-06, "loss": 1.5207, "step": 31182 }, { "epoch": 2.4438087774294672, "grad_norm": 3.4410359859466553, "learning_rate": 3.3035254292438714e-06, "loss": 1.2733, "step": 31183 }, { "epoch": 2.443887147335423, "grad_norm": 4.243980407714844, "learning_rate": 3.3032353624327785e-06, "loss": 1.517, "step": 31184 }, { "epoch": 2.4439655172413794, "grad_norm": 3.2158892154693604, "learning_rate": 3.3029453020753067e-06, "loss": 1.1945, "step": 31185 }, { "epoch": 2.4440438871473353, "grad_norm": 7.702213764190674, "learning_rate": 3.3026552481725605e-06, "loss": 1.424, "step": 31186 }, { "epoch": 2.4441222570532917, "grad_norm": 7.272906303405762, "learning_rate": 3.3023652007256403e-06, "loss": 1.8042, "step": 31187 }, { "epoch": 2.4442006269592476, "grad_norm": 9.688009262084961, "learning_rate": 3.3020751597356506e-06, "loss": 1.7621, "step": 31188 }, { "epoch": 2.444278996865204, "grad_norm": 2.3850955963134766, "learning_rate": 3.3017851252036958e-06, "loss": 1.331, "step": 31189 }, { "epoch": 2.44435736677116, "grad_norm": 9.038132667541504, "learning_rate": 3.301495097130879e-06, "loss": 1.584, "step": 31190 }, { "epoch": 2.444435736677116, "grad_norm": 4.7791242599487305, "learning_rate": 3.3012050755183017e-06, "loss": 1.4513, "step": 31191 }, { "epoch": 2.444514106583072, "grad_norm": 3.806600570678711, "learning_rate": 3.3009150603670677e-06, "loss": 1.4335, "step": 31192 }, { "epoch": 2.4445924764890283, "grad_norm": 2.847965955734253, "learning_rate": 3.3006250516782812e-06, "loss": 1.1583, "step": 31193 }, { "epoch": 2.444670846394984, "grad_norm": 2.6761088371276855, "learning_rate": 3.3003350494530427e-06, "loss": 1.1625, "step": 31194 }, { "epoch": 2.4447492163009406, "grad_norm": 4.382732391357422, "learning_rate": 3.3000450536924554e-06, "loss": 1.143, "step": 31195 }, { "epoch": 2.4448275862068964, "grad_norm": 6.275578022003174, "learning_rate": 3.2997550643976254e-06, "loss": 1.2412, "step": 31196 }, { "epoch": 2.4449059561128528, "grad_norm": 4.925638675689697, "learning_rate": 3.299465081569655e-06, "loss": 0.7826, "step": 31197 }, { "epoch": 2.4449843260188087, "grad_norm": 3.411402702331543, "learning_rate": 3.2991751052096444e-06, "loss": 1.4432, "step": 31198 }, { "epoch": 2.445062695924765, "grad_norm": 5.991248607635498, "learning_rate": 3.298885135318698e-06, "loss": 0.8128, "step": 31199 }, { "epoch": 2.445141065830721, "grad_norm": 2.4765613079071045, "learning_rate": 3.298595171897918e-06, "loss": 1.1674, "step": 31200 }, { "epoch": 2.445219435736677, "grad_norm": 5.185506343841553, "learning_rate": 3.298305214948411e-06, "loss": 1.5559, "step": 31201 }, { "epoch": 2.445297805642633, "grad_norm": 6.273968696594238, "learning_rate": 3.298015264471274e-06, "loss": 1.2997, "step": 31202 }, { "epoch": 2.4453761755485894, "grad_norm": 2.5584003925323486, "learning_rate": 3.2977253204676136e-06, "loss": 1.301, "step": 31203 }, { "epoch": 2.4454545454545453, "grad_norm": 3.328587770462036, "learning_rate": 3.2974353829385318e-06, "loss": 1.8254, "step": 31204 }, { "epoch": 2.4455329153605017, "grad_norm": 6.54832124710083, "learning_rate": 3.297145451885132e-06, "loss": 1.392, "step": 31205 }, { "epoch": 2.4456112852664575, "grad_norm": 2.6599998474121094, "learning_rate": 3.2968555273085155e-06, "loss": 1.32, "step": 31206 }, { "epoch": 2.445689655172414, "grad_norm": 3.010993003845215, "learning_rate": 3.2965656092097855e-06, "loss": 1.3476, "step": 31207 }, { "epoch": 2.4457680250783698, "grad_norm": 4.135217189788818, "learning_rate": 3.2962756975900473e-06, "loss": 1.3156, "step": 31208 }, { "epoch": 2.445846394984326, "grad_norm": 5.392930030822754, "learning_rate": 3.295985792450398e-06, "loss": 1.5762, "step": 31209 }, { "epoch": 2.445924764890282, "grad_norm": 3.58005952835083, "learning_rate": 3.2956958937919448e-06, "loss": 1.2168, "step": 31210 }, { "epoch": 2.4460031347962383, "grad_norm": 4.237050533294678, "learning_rate": 3.2954060016157895e-06, "loss": 1.4491, "step": 31211 }, { "epoch": 2.446081504702194, "grad_norm": 3.9919931888580322, "learning_rate": 3.295116115923035e-06, "loss": 1.6075, "step": 31212 }, { "epoch": 2.4461598746081505, "grad_norm": 2.392444133758545, "learning_rate": 3.2948262367147815e-06, "loss": 1.3493, "step": 31213 }, { "epoch": 2.4462382445141064, "grad_norm": 3.8046422004699707, "learning_rate": 3.2945363639921325e-06, "loss": 1.3155, "step": 31214 }, { "epoch": 2.4463166144200628, "grad_norm": 14.332439422607422, "learning_rate": 3.2942464977561926e-06, "loss": 1.0999, "step": 31215 }, { "epoch": 2.4463949843260187, "grad_norm": 2.5715835094451904, "learning_rate": 3.2939566380080635e-06, "loss": 1.5679, "step": 31216 }, { "epoch": 2.446473354231975, "grad_norm": 2.152125835418701, "learning_rate": 3.293666784748846e-06, "loss": 1.1794, "step": 31217 }, { "epoch": 2.446551724137931, "grad_norm": 3.5648674964904785, "learning_rate": 3.2933769379796438e-06, "loss": 1.4337, "step": 31218 }, { "epoch": 2.446630094043887, "grad_norm": 5.52197265625, "learning_rate": 3.2930870977015584e-06, "loss": 1.621, "step": 31219 }, { "epoch": 2.446708463949843, "grad_norm": 4.604253768920898, "learning_rate": 3.2927972639156947e-06, "loss": 1.4065, "step": 31220 }, { "epoch": 2.4467868338557994, "grad_norm": 4.666541576385498, "learning_rate": 3.2925074366231525e-06, "loss": 1.2845, "step": 31221 }, { "epoch": 2.4468652037617553, "grad_norm": 3.797027826309204, "learning_rate": 3.292217615825034e-06, "loss": 1.4798, "step": 31222 }, { "epoch": 2.4469435736677116, "grad_norm": 2.4403891563415527, "learning_rate": 3.291927801522444e-06, "loss": 1.0773, "step": 31223 }, { "epoch": 2.4470219435736675, "grad_norm": 3.9021222591400146, "learning_rate": 3.291637993716483e-06, "loss": 1.3935, "step": 31224 }, { "epoch": 2.447100313479624, "grad_norm": 13.177285194396973, "learning_rate": 3.2913481924082534e-06, "loss": 0.8941, "step": 31225 }, { "epoch": 2.4471786833855798, "grad_norm": 4.551107406616211, "learning_rate": 3.2910583975988564e-06, "loss": 1.4034, "step": 31226 }, { "epoch": 2.447257053291536, "grad_norm": 2.5074539184570312, "learning_rate": 3.2907686092893976e-06, "loss": 1.3699, "step": 31227 }, { "epoch": 2.447335423197492, "grad_norm": 2.4696238040924072, "learning_rate": 3.2904788274809763e-06, "loss": 0.9854, "step": 31228 }, { "epoch": 2.4474137931034483, "grad_norm": 4.2483344078063965, "learning_rate": 3.290189052174695e-06, "loss": 1.0231, "step": 31229 }, { "epoch": 2.447492163009404, "grad_norm": 3.326033592224121, "learning_rate": 3.289899283371657e-06, "loss": 1.0955, "step": 31230 }, { "epoch": 2.4475705329153605, "grad_norm": 3.7258646488189697, "learning_rate": 3.2896095210729638e-06, "loss": 1.2713, "step": 31231 }, { "epoch": 2.4476489028213164, "grad_norm": 2.6184730529785156, "learning_rate": 3.289319765279716e-06, "loss": 1.4023, "step": 31232 }, { "epoch": 2.4477272727272728, "grad_norm": 2.4519095420837402, "learning_rate": 3.289030015993018e-06, "loss": 1.3076, "step": 31233 }, { "epoch": 2.447805642633229, "grad_norm": 2.9216787815093994, "learning_rate": 3.2887402732139726e-06, "loss": 1.4063, "step": 31234 }, { "epoch": 2.447884012539185, "grad_norm": 3.071047067642212, "learning_rate": 3.288450536943679e-06, "loss": 1.3338, "step": 31235 }, { "epoch": 2.447962382445141, "grad_norm": 3.6888554096221924, "learning_rate": 3.28816080718324e-06, "loss": 1.1035, "step": 31236 }, { "epoch": 2.448040752351097, "grad_norm": 2.8613264560699463, "learning_rate": 3.2878710839337574e-06, "loss": 1.5759, "step": 31237 }, { "epoch": 2.4481191222570535, "grad_norm": 3.29848575592041, "learning_rate": 3.2875813671963362e-06, "loss": 1.8229, "step": 31238 }, { "epoch": 2.4481974921630094, "grad_norm": 2.312659502029419, "learning_rate": 3.2872916569720737e-06, "loss": 1.4171, "step": 31239 }, { "epoch": 2.4482758620689653, "grad_norm": 3.505908966064453, "learning_rate": 3.2870019532620744e-06, "loss": 1.4182, "step": 31240 }, { "epoch": 2.4483542319749216, "grad_norm": 5.738698482513428, "learning_rate": 3.28671225606744e-06, "loss": 1.0912, "step": 31241 }, { "epoch": 2.448432601880878, "grad_norm": 3.676175594329834, "learning_rate": 3.286422565389274e-06, "loss": 1.0371, "step": 31242 }, { "epoch": 2.448510971786834, "grad_norm": 3.4564473628997803, "learning_rate": 3.2861328812286747e-06, "loss": 1.4802, "step": 31243 }, { "epoch": 2.4485893416927897, "grad_norm": 3.3131930828094482, "learning_rate": 3.285843203586745e-06, "loss": 1.1773, "step": 31244 }, { "epoch": 2.448667711598746, "grad_norm": 3.3791730403900146, "learning_rate": 3.2855535324645878e-06, "loss": 1.3732, "step": 31245 }, { "epoch": 2.4487460815047024, "grad_norm": 2.3987104892730713, "learning_rate": 3.2852638678633053e-06, "loss": 1.4272, "step": 31246 }, { "epoch": 2.4488244514106583, "grad_norm": 2.918431282043457, "learning_rate": 3.2849742097839974e-06, "loss": 0.8114, "step": 31247 }, { "epoch": 2.448902821316614, "grad_norm": 8.555136680603027, "learning_rate": 3.2846845582277678e-06, "loss": 0.9653, "step": 31248 }, { "epoch": 2.4489811912225705, "grad_norm": 2.5973191261291504, "learning_rate": 3.2843949131957166e-06, "loss": 1.5298, "step": 31249 }, { "epoch": 2.449059561128527, "grad_norm": 2.8095548152923584, "learning_rate": 3.2841052746889456e-06, "loss": 1.2334, "step": 31250 }, { "epoch": 2.4491379310344827, "grad_norm": 13.324673652648926, "learning_rate": 3.2838156427085553e-06, "loss": 1.2384, "step": 31251 }, { "epoch": 2.4492163009404386, "grad_norm": 8.718137741088867, "learning_rate": 3.2835260172556505e-06, "loss": 1.2031, "step": 31252 }, { "epoch": 2.449294670846395, "grad_norm": 4.70619535446167, "learning_rate": 3.283236398331332e-06, "loss": 1.2595, "step": 31253 }, { "epoch": 2.4493730407523513, "grad_norm": 14.728657722473145, "learning_rate": 3.282946785936699e-06, "loss": 1.5717, "step": 31254 }, { "epoch": 2.449451410658307, "grad_norm": 7.371335983276367, "learning_rate": 3.2826571800728546e-06, "loss": 0.7676, "step": 31255 }, { "epoch": 2.4495297805642635, "grad_norm": 2.343931198120117, "learning_rate": 3.2823675807408995e-06, "loss": 1.1774, "step": 31256 }, { "epoch": 2.4496081504702194, "grad_norm": 4.40623664855957, "learning_rate": 3.282077987941939e-06, "loss": 1.1608, "step": 31257 }, { "epoch": 2.4496865203761757, "grad_norm": 3.174924612045288, "learning_rate": 3.2817884016770684e-06, "loss": 1.4707, "step": 31258 }, { "epoch": 2.4497648902821316, "grad_norm": 19.055585861206055, "learning_rate": 3.281498821947393e-06, "loss": 1.1433, "step": 31259 }, { "epoch": 2.449843260188088, "grad_norm": 2.502113103866577, "learning_rate": 3.281209248754013e-06, "loss": 1.3044, "step": 31260 }, { "epoch": 2.449921630094044, "grad_norm": 5.605293273925781, "learning_rate": 3.2809196820980316e-06, "loss": 1.2982, "step": 31261 }, { "epoch": 2.45, "grad_norm": 16.34102439880371, "learning_rate": 3.280630121980548e-06, "loss": 0.9895, "step": 31262 }, { "epoch": 2.450078369905956, "grad_norm": 2.043107271194458, "learning_rate": 3.280340568402663e-06, "loss": 1.4711, "step": 31263 }, { "epoch": 2.4501567398119124, "grad_norm": 2.6262290477752686, "learning_rate": 3.2800510213654823e-06, "loss": 1.466, "step": 31264 }, { "epoch": 2.4502351097178683, "grad_norm": 3.7204089164733887, "learning_rate": 3.2797614808701006e-06, "loss": 0.6975, "step": 31265 }, { "epoch": 2.4503134796238246, "grad_norm": 3.4528582096099854, "learning_rate": 3.2794719469176245e-06, "loss": 1.359, "step": 31266 }, { "epoch": 2.4503918495297805, "grad_norm": 2.347090005874634, "learning_rate": 3.2791824195091525e-06, "loss": 1.2742, "step": 31267 }, { "epoch": 2.450470219435737, "grad_norm": 3.0874578952789307, "learning_rate": 3.2788928986457884e-06, "loss": 1.496, "step": 31268 }, { "epoch": 2.4505485893416927, "grad_norm": 3.3373982906341553, "learning_rate": 3.278603384328631e-06, "loss": 1.6288, "step": 31269 }, { "epoch": 2.450626959247649, "grad_norm": 3.5363175868988037, "learning_rate": 3.278313876558781e-06, "loss": 1.522, "step": 31270 }, { "epoch": 2.450705329153605, "grad_norm": 4.734889984130859, "learning_rate": 3.278024375337341e-06, "loss": 1.4302, "step": 31271 }, { "epoch": 2.4507836990595613, "grad_norm": 5.874294281005859, "learning_rate": 3.277734880665413e-06, "loss": 1.4676, "step": 31272 }, { "epoch": 2.450862068965517, "grad_norm": 3.7128422260284424, "learning_rate": 3.2774453925440963e-06, "loss": 1.7572, "step": 31273 }, { "epoch": 2.4509404388714735, "grad_norm": 2.432997703552246, "learning_rate": 3.277155910974492e-06, "loss": 1.3451, "step": 31274 }, { "epoch": 2.4510188087774294, "grad_norm": 5.243110179901123, "learning_rate": 3.276866435957703e-06, "loss": 0.9611, "step": 31275 }, { "epoch": 2.4510971786833857, "grad_norm": 2.3205909729003906, "learning_rate": 3.276576967494828e-06, "loss": 1.3445, "step": 31276 }, { "epoch": 2.4511755485893416, "grad_norm": 5.584974765777588, "learning_rate": 3.276287505586968e-06, "loss": 1.1903, "step": 31277 }, { "epoch": 2.451253918495298, "grad_norm": 2.673478841781616, "learning_rate": 3.275998050235226e-06, "loss": 0.7799, "step": 31278 }, { "epoch": 2.451332288401254, "grad_norm": 2.873206615447998, "learning_rate": 3.2757086014407036e-06, "loss": 1.0426, "step": 31279 }, { "epoch": 2.45141065830721, "grad_norm": 4.0562310218811035, "learning_rate": 3.275419159204498e-06, "loss": 1.1935, "step": 31280 }, { "epoch": 2.451489028213166, "grad_norm": 2.319136142730713, "learning_rate": 3.2751297235277122e-06, "loss": 1.3323, "step": 31281 }, { "epoch": 2.4515673981191224, "grad_norm": 3.202376365661621, "learning_rate": 3.2748402944114465e-06, "loss": 1.6057, "step": 31282 }, { "epoch": 2.4516457680250783, "grad_norm": 2.5956482887268066, "learning_rate": 3.2745508718568053e-06, "loss": 1.3984, "step": 31283 }, { "epoch": 2.4517241379310346, "grad_norm": 2.7261672019958496, "learning_rate": 3.2742614558648835e-06, "loss": 1.4077, "step": 31284 }, { "epoch": 2.4518025078369905, "grad_norm": 4.538415431976318, "learning_rate": 3.2739720464367848e-06, "loss": 1.078, "step": 31285 }, { "epoch": 2.451880877742947, "grad_norm": 3.3486645221710205, "learning_rate": 3.27368264357361e-06, "loss": 1.251, "step": 31286 }, { "epoch": 2.4519592476489027, "grad_norm": 4.0363969802856445, "learning_rate": 3.273393247276462e-06, "loss": 1.367, "step": 31287 }, { "epoch": 2.452037617554859, "grad_norm": 5.3537211418151855, "learning_rate": 3.273103857546437e-06, "loss": 1.4875, "step": 31288 }, { "epoch": 2.452115987460815, "grad_norm": 3.9912538528442383, "learning_rate": 3.272814474384637e-06, "loss": 1.1873, "step": 31289 }, { "epoch": 2.4521943573667713, "grad_norm": 3.197437047958374, "learning_rate": 3.2725250977921667e-06, "loss": 1.3322, "step": 31290 }, { "epoch": 2.452272727272727, "grad_norm": 3.781705856323242, "learning_rate": 3.272235727770121e-06, "loss": 1.4417, "step": 31291 }, { "epoch": 2.4523510971786835, "grad_norm": 3.9853391647338867, "learning_rate": 3.2719463643196038e-06, "loss": 1.1639, "step": 31292 }, { "epoch": 2.4524294670846394, "grad_norm": 5.29800271987915, "learning_rate": 3.2716570074417154e-06, "loss": 1.6189, "step": 31293 }, { "epoch": 2.4525078369905957, "grad_norm": 3.4487452507019043, "learning_rate": 3.271367657137557e-06, "loss": 1.4362, "step": 31294 }, { "epoch": 2.4525862068965516, "grad_norm": 3.1509573459625244, "learning_rate": 3.271078313408227e-06, "loss": 1.6179, "step": 31295 }, { "epoch": 2.452664576802508, "grad_norm": 2.166598081588745, "learning_rate": 3.270788976254826e-06, "loss": 1.2281, "step": 31296 }, { "epoch": 2.452742946708464, "grad_norm": 5.359233856201172, "learning_rate": 3.2704996456784566e-06, "loss": 1.2447, "step": 31297 }, { "epoch": 2.45282131661442, "grad_norm": 3.019442558288574, "learning_rate": 3.2702103216802197e-06, "loss": 1.483, "step": 31298 }, { "epoch": 2.452899686520376, "grad_norm": 3.409893035888672, "learning_rate": 3.2699210042612125e-06, "loss": 1.0801, "step": 31299 }, { "epoch": 2.4529780564263324, "grad_norm": 3.235706090927124, "learning_rate": 3.269631693422537e-06, "loss": 1.3823, "step": 31300 }, { "epoch": 2.4530564263322883, "grad_norm": 6.7535014152526855, "learning_rate": 3.269342389165293e-06, "loss": 1.7251, "step": 31301 }, { "epoch": 2.4531347962382446, "grad_norm": 2.655247211456299, "learning_rate": 3.2690530914905853e-06, "loss": 1.5608, "step": 31302 }, { "epoch": 2.4532131661442005, "grad_norm": 3.3447067737579346, "learning_rate": 3.268763800399507e-06, "loss": 1.276, "step": 31303 }, { "epoch": 2.453291536050157, "grad_norm": 3.4080309867858887, "learning_rate": 3.268474515893163e-06, "loss": 1.5743, "step": 31304 }, { "epoch": 2.4533699059561127, "grad_norm": 3.990842580795288, "learning_rate": 3.2681852379726543e-06, "loss": 1.4194, "step": 31305 }, { "epoch": 2.453448275862069, "grad_norm": 2.8177754878997803, "learning_rate": 3.2678959666390775e-06, "loss": 1.4192, "step": 31306 }, { "epoch": 2.453526645768025, "grad_norm": 2.263011932373047, "learning_rate": 3.2676067018935354e-06, "loss": 1.3795, "step": 31307 }, { "epoch": 2.4536050156739813, "grad_norm": 2.6934077739715576, "learning_rate": 3.2673174437371257e-06, "loss": 1.2702, "step": 31308 }, { "epoch": 2.453683385579937, "grad_norm": 3.068181037902832, "learning_rate": 3.2670281921709545e-06, "loss": 1.1981, "step": 31309 }, { "epoch": 2.4537617554858935, "grad_norm": 9.2473726272583, "learning_rate": 3.2667389471961143e-06, "loss": 1.3994, "step": 31310 }, { "epoch": 2.4538401253918494, "grad_norm": 6.10623025894165, "learning_rate": 3.2664497088137103e-06, "loss": 1.1082, "step": 31311 }, { "epoch": 2.4539184952978057, "grad_norm": 3.002042770385742, "learning_rate": 3.266160477024841e-06, "loss": 1.5461, "step": 31312 }, { "epoch": 2.4539968652037616, "grad_norm": 5.482217788696289, "learning_rate": 3.265871251830608e-06, "loss": 1.3852, "step": 31313 }, { "epoch": 2.454075235109718, "grad_norm": 6.993821620941162, "learning_rate": 3.2655820332321087e-06, "loss": 0.9123, "step": 31314 }, { "epoch": 2.454153605015674, "grad_norm": 3.102592706680298, "learning_rate": 3.2652928212304434e-06, "loss": 1.5116, "step": 31315 }, { "epoch": 2.45423197492163, "grad_norm": 7.7971110343933105, "learning_rate": 3.265003615826714e-06, "loss": 1.5278, "step": 31316 }, { "epoch": 2.454310344827586, "grad_norm": 3.2549498081207275, "learning_rate": 3.264714417022021e-06, "loss": 1.1608, "step": 31317 }, { "epoch": 2.4543887147335424, "grad_norm": 2.8345818519592285, "learning_rate": 3.2644252248174622e-06, "loss": 0.9643, "step": 31318 }, { "epoch": 2.4544670846394983, "grad_norm": 1.976958155632019, "learning_rate": 3.2641360392141386e-06, "loss": 1.2165, "step": 31319 }, { "epoch": 2.4545454545454546, "grad_norm": 3.647366762161255, "learning_rate": 3.26384686021315e-06, "loss": 1.4105, "step": 31320 }, { "epoch": 2.4546238244514105, "grad_norm": 3.961291790008545, "learning_rate": 3.263557687815596e-06, "loss": 1.0681, "step": 31321 }, { "epoch": 2.454702194357367, "grad_norm": 2.263012409210205, "learning_rate": 3.2632685220225757e-06, "loss": 1.324, "step": 31322 }, { "epoch": 2.4547805642633227, "grad_norm": 2.414382219314575, "learning_rate": 3.262979362835191e-06, "loss": 1.2691, "step": 31323 }, { "epoch": 2.454858934169279, "grad_norm": 2.5123677253723145, "learning_rate": 3.262690210254541e-06, "loss": 1.3877, "step": 31324 }, { "epoch": 2.454937304075235, "grad_norm": 4.321882247924805, "learning_rate": 3.2624010642817244e-06, "loss": 1.1783, "step": 31325 }, { "epoch": 2.4550156739811912, "grad_norm": 9.38933277130127, "learning_rate": 3.2621119249178416e-06, "loss": 1.7351, "step": 31326 }, { "epoch": 2.455094043887147, "grad_norm": 10.244894981384277, "learning_rate": 3.261822792163991e-06, "loss": 1.7585, "step": 31327 }, { "epoch": 2.4551724137931035, "grad_norm": 2.8039376735687256, "learning_rate": 3.261533666021277e-06, "loss": 1.3624, "step": 31328 }, { "epoch": 2.45525078369906, "grad_norm": 3.0523793697357178, "learning_rate": 3.261244546490794e-06, "loss": 1.3215, "step": 31329 }, { "epoch": 2.4553291536050157, "grad_norm": 8.909332275390625, "learning_rate": 3.2609554335736435e-06, "loss": 1.7062, "step": 31330 }, { "epoch": 2.4554075235109716, "grad_norm": 2.4284470081329346, "learning_rate": 3.260666327270927e-06, "loss": 1.3681, "step": 31331 }, { "epoch": 2.455485893416928, "grad_norm": 4.57036828994751, "learning_rate": 3.260377227583741e-06, "loss": 1.3498, "step": 31332 }, { "epoch": 2.4555642633228842, "grad_norm": 3.871021032333374, "learning_rate": 3.2600881345131854e-06, "loss": 1.4355, "step": 31333 }, { "epoch": 2.45564263322884, "grad_norm": 3.436164140701294, "learning_rate": 3.259799048060362e-06, "loss": 1.4784, "step": 31334 }, { "epoch": 2.455721003134796, "grad_norm": 3.062729597091675, "learning_rate": 3.2595099682263704e-06, "loss": 1.2115, "step": 31335 }, { "epoch": 2.4557993730407524, "grad_norm": 1.9739421606063843, "learning_rate": 3.259220895012308e-06, "loss": 1.5309, "step": 31336 }, { "epoch": 2.4558777429467087, "grad_norm": 4.90571403503418, "learning_rate": 3.2589318284192746e-06, "loss": 1.6147, "step": 31337 }, { "epoch": 2.4559561128526646, "grad_norm": 3.9583311080932617, "learning_rate": 3.2586427684483694e-06, "loss": 1.2555, "step": 31338 }, { "epoch": 2.4560344827586205, "grad_norm": 3.9356253147125244, "learning_rate": 3.258353715100696e-06, "loss": 1.4817, "step": 31339 }, { "epoch": 2.456112852664577, "grad_norm": 3.151871681213379, "learning_rate": 3.2580646683773476e-06, "loss": 1.6212, "step": 31340 }, { "epoch": 2.456191222570533, "grad_norm": 4.279940605163574, "learning_rate": 3.257775628279427e-06, "loss": 1.4052, "step": 31341 }, { "epoch": 2.456269592476489, "grad_norm": 2.903394937515259, "learning_rate": 3.2574865948080334e-06, "loss": 1.5095, "step": 31342 }, { "epoch": 2.456347962382445, "grad_norm": 6.060739994049072, "learning_rate": 3.2571975679642666e-06, "loss": 2.0378, "step": 31343 }, { "epoch": 2.4564263322884012, "grad_norm": 8.153664588928223, "learning_rate": 3.2569085477492242e-06, "loss": 1.1478, "step": 31344 }, { "epoch": 2.4565047021943576, "grad_norm": 4.08603572845459, "learning_rate": 3.256619534164005e-06, "loss": 1.3561, "step": 31345 }, { "epoch": 2.4565830721003135, "grad_norm": 3.9592535495758057, "learning_rate": 3.2563305272097134e-06, "loss": 1.9238, "step": 31346 }, { "epoch": 2.4566614420062693, "grad_norm": 2.5984156131744385, "learning_rate": 3.2560415268874414e-06, "loss": 1.0534, "step": 31347 }, { "epoch": 2.4567398119122257, "grad_norm": 3.1789019107818604, "learning_rate": 3.2557525331982924e-06, "loss": 1.7105, "step": 31348 }, { "epoch": 2.456818181818182, "grad_norm": 6.018877983093262, "learning_rate": 3.2554635461433655e-06, "loss": 0.9262, "step": 31349 }, { "epoch": 2.456896551724138, "grad_norm": 4.136083126068115, "learning_rate": 3.2551745657237597e-06, "loss": 1.5182, "step": 31350 }, { "epoch": 2.4569749216300942, "grad_norm": 2.8395378589630127, "learning_rate": 3.2548855919405732e-06, "loss": 1.2442, "step": 31351 }, { "epoch": 2.45705329153605, "grad_norm": 3.5250959396362305, "learning_rate": 3.254596624794904e-06, "loss": 1.529, "step": 31352 }, { "epoch": 2.4571316614420065, "grad_norm": 3.524142026901245, "learning_rate": 3.254307664287854e-06, "loss": 1.4773, "step": 31353 }, { "epoch": 2.4572100313479623, "grad_norm": 4.228175640106201, "learning_rate": 3.254018710420522e-06, "loss": 1.1989, "step": 31354 }, { "epoch": 2.4572884012539187, "grad_norm": 11.003763198852539, "learning_rate": 3.2537297631940044e-06, "loss": 1.2205, "step": 31355 }, { "epoch": 2.4573667711598746, "grad_norm": 3.3682777881622314, "learning_rate": 3.2534408226094026e-06, "loss": 1.3116, "step": 31356 }, { "epoch": 2.457445141065831, "grad_norm": 4.158703327178955, "learning_rate": 3.253151888667814e-06, "loss": 1.0889, "step": 31357 }, { "epoch": 2.457523510971787, "grad_norm": 2.8759636878967285, "learning_rate": 3.252862961370341e-06, "loss": 1.2313, "step": 31358 }, { "epoch": 2.457601880877743, "grad_norm": 6.863409042358398, "learning_rate": 3.252574040718077e-06, "loss": 1.3328, "step": 31359 }, { "epoch": 2.457680250783699, "grad_norm": 6.798775672912598, "learning_rate": 3.2522851267121245e-06, "loss": 1.1903, "step": 31360 }, { "epoch": 2.4577586206896553, "grad_norm": 2.4487125873565674, "learning_rate": 3.2519962193535836e-06, "loss": 1.3394, "step": 31361 }, { "epoch": 2.4578369905956112, "grad_norm": 4.776952266693115, "learning_rate": 3.2517073186435496e-06, "loss": 1.4358, "step": 31362 }, { "epoch": 2.4579153605015676, "grad_norm": 4.208643913269043, "learning_rate": 3.2514184245831227e-06, "loss": 1.1503, "step": 31363 }, { "epoch": 2.4579937304075234, "grad_norm": 6.01618766784668, "learning_rate": 3.2511295371734008e-06, "loss": 1.221, "step": 31364 }, { "epoch": 2.45807210031348, "grad_norm": 2.8441519737243652, "learning_rate": 3.2508406564154883e-06, "loss": 1.2205, "step": 31365 }, { "epoch": 2.4581504702194357, "grad_norm": 5.252183437347412, "learning_rate": 3.250551782310475e-06, "loss": 0.9417, "step": 31366 }, { "epoch": 2.458228840125392, "grad_norm": 5.639472007751465, "learning_rate": 3.2502629148594657e-06, "loss": 1.5548, "step": 31367 }, { "epoch": 2.458307210031348, "grad_norm": 3.571256637573242, "learning_rate": 3.249974054063557e-06, "loss": 1.4758, "step": 31368 }, { "epoch": 2.458385579937304, "grad_norm": 2.5360281467437744, "learning_rate": 3.2496851999238497e-06, "loss": 1.3898, "step": 31369 }, { "epoch": 2.45846394984326, "grad_norm": 3.6659536361694336, "learning_rate": 3.2493963524414406e-06, "loss": 1.0261, "step": 31370 }, { "epoch": 2.4585423197492164, "grad_norm": 3.2777106761932373, "learning_rate": 3.2491075116174266e-06, "loss": 1.4774, "step": 31371 }, { "epoch": 2.4586206896551723, "grad_norm": 3.3868463039398193, "learning_rate": 3.2488186774529094e-06, "loss": 1.2165, "step": 31372 }, { "epoch": 2.4586990595611287, "grad_norm": 3.229116678237915, "learning_rate": 3.2485298499489883e-06, "loss": 1.1723, "step": 31373 }, { "epoch": 2.4587774294670846, "grad_norm": 11.72594928741455, "learning_rate": 3.248241029106758e-06, "loss": 1.29, "step": 31374 }, { "epoch": 2.458855799373041, "grad_norm": 3.703169822692871, "learning_rate": 3.2479522149273195e-06, "loss": 1.4466, "step": 31375 }, { "epoch": 2.4589341692789968, "grad_norm": 2.4989044666290283, "learning_rate": 3.2476634074117717e-06, "loss": 1.2812, "step": 31376 }, { "epoch": 2.459012539184953, "grad_norm": 3.8570404052734375, "learning_rate": 3.2473746065612114e-06, "loss": 1.138, "step": 31377 }, { "epoch": 2.459090909090909, "grad_norm": 2.6024796962738037, "learning_rate": 3.247085812376737e-06, "loss": 1.5027, "step": 31378 }, { "epoch": 2.4591692789968653, "grad_norm": 3.496061086654663, "learning_rate": 3.246797024859448e-06, "loss": 1.3178, "step": 31379 }, { "epoch": 2.459247648902821, "grad_norm": 4.541456699371338, "learning_rate": 3.246508244010445e-06, "loss": 1.4651, "step": 31380 }, { "epoch": 2.4593260188087775, "grad_norm": 3.1663265228271484, "learning_rate": 3.2462194698308226e-06, "loss": 1.387, "step": 31381 }, { "epoch": 2.4594043887147334, "grad_norm": 3.4619638919830322, "learning_rate": 3.24593070232168e-06, "loss": 1.1065, "step": 31382 }, { "epoch": 2.4594827586206898, "grad_norm": 2.563227891921997, "learning_rate": 3.2456419414841152e-06, "loss": 1.6285, "step": 31383 }, { "epoch": 2.4595611285266457, "grad_norm": 3.0781874656677246, "learning_rate": 3.24535318731923e-06, "loss": 1.014, "step": 31384 }, { "epoch": 2.459639498432602, "grad_norm": 2.376338481903076, "learning_rate": 3.2450644398281178e-06, "loss": 0.9987, "step": 31385 }, { "epoch": 2.459717868338558, "grad_norm": 3.524416446685791, "learning_rate": 3.2447756990118795e-06, "loss": 1.6437, "step": 31386 }, { "epoch": 2.459796238244514, "grad_norm": 6.830414772033691, "learning_rate": 3.244486964871615e-06, "loss": 1.4127, "step": 31387 }, { "epoch": 2.45987460815047, "grad_norm": 2.7328693866729736, "learning_rate": 3.244198237408418e-06, "loss": 1.2328, "step": 31388 }, { "epoch": 2.4599529780564264, "grad_norm": 4.671082019805908, "learning_rate": 3.2439095166233898e-06, "loss": 1.7685, "step": 31389 }, { "epoch": 2.4600313479623823, "grad_norm": 3.1176514625549316, "learning_rate": 3.2436208025176265e-06, "loss": 1.1816, "step": 31390 }, { "epoch": 2.4601097178683387, "grad_norm": 2.3703479766845703, "learning_rate": 3.243332095092231e-06, "loss": 1.2283, "step": 31391 }, { "epoch": 2.4601880877742945, "grad_norm": 3.187251329421997, "learning_rate": 3.2430433943482942e-06, "loss": 1.5445, "step": 31392 }, { "epoch": 2.460266457680251, "grad_norm": 3.946397542953491, "learning_rate": 3.24275470028692e-06, "loss": 0.8627, "step": 31393 }, { "epoch": 2.4603448275862068, "grad_norm": 3.4932587146759033, "learning_rate": 3.2424660129092033e-06, "loss": 1.258, "step": 31394 }, { "epoch": 2.460423197492163, "grad_norm": 3.1955981254577637, "learning_rate": 3.242177332216244e-06, "loss": 1.4321, "step": 31395 }, { "epoch": 2.460501567398119, "grad_norm": 3.7827796936035156, "learning_rate": 3.241888658209139e-06, "loss": 1.1974, "step": 31396 }, { "epoch": 2.4605799373040753, "grad_norm": 7.208288192749023, "learning_rate": 3.241599990888985e-06, "loss": 1.3916, "step": 31397 }, { "epoch": 2.460658307210031, "grad_norm": 3.9679343700408936, "learning_rate": 3.241311330256883e-06, "loss": 1.1033, "step": 31398 }, { "epoch": 2.4607366771159875, "grad_norm": 3.8567700386047363, "learning_rate": 3.2410226763139304e-06, "loss": 0.8942, "step": 31399 }, { "epoch": 2.4608150470219434, "grad_norm": 2.458117961883545, "learning_rate": 3.240734029061222e-06, "loss": 1.4557, "step": 31400 }, { "epoch": 2.4608934169278998, "grad_norm": 2.8615171909332275, "learning_rate": 3.240445388499859e-06, "loss": 1.1182, "step": 31401 }, { "epoch": 2.4609717868338556, "grad_norm": 3.711510181427002, "learning_rate": 3.2401567546309387e-06, "loss": 1.2799, "step": 31402 }, { "epoch": 2.461050156739812, "grad_norm": 2.046520709991455, "learning_rate": 3.2398681274555565e-06, "loss": 1.3019, "step": 31403 }, { "epoch": 2.461128526645768, "grad_norm": 3.1280126571655273, "learning_rate": 3.2395795069748115e-06, "loss": 1.303, "step": 31404 }, { "epoch": 2.461206896551724, "grad_norm": 2.1661510467529297, "learning_rate": 3.2392908931898027e-06, "loss": 1.0904, "step": 31405 }, { "epoch": 2.46128526645768, "grad_norm": 2.557098865509033, "learning_rate": 3.2390022861016277e-06, "loss": 1.3205, "step": 31406 }, { "epoch": 2.4613636363636364, "grad_norm": 7.527204990386963, "learning_rate": 3.238713685711383e-06, "loss": 0.22, "step": 31407 }, { "epoch": 2.4614420062695923, "grad_norm": 7.148137092590332, "learning_rate": 3.238425092020167e-06, "loss": 1.9538, "step": 31408 }, { "epoch": 2.4615203761755486, "grad_norm": 2.6241414546966553, "learning_rate": 3.238136505029076e-06, "loss": 1.3752, "step": 31409 }, { "epoch": 2.4615987460815045, "grad_norm": 2.195613145828247, "learning_rate": 3.2378479247392115e-06, "loss": 1.5202, "step": 31410 }, { "epoch": 2.461677115987461, "grad_norm": 5.938746452331543, "learning_rate": 3.2375593511516652e-06, "loss": 1.5474, "step": 31411 }, { "epoch": 2.4617554858934168, "grad_norm": 2.7491037845611572, "learning_rate": 3.237270784267539e-06, "loss": 0.9702, "step": 31412 }, { "epoch": 2.461833855799373, "grad_norm": 4.204365253448486, "learning_rate": 3.2369822240879296e-06, "loss": 1.3715, "step": 31413 }, { "epoch": 2.461912225705329, "grad_norm": 2.422794818878174, "learning_rate": 3.236693670613935e-06, "loss": 1.4706, "step": 31414 }, { "epoch": 2.4619905956112853, "grad_norm": 4.728226184844971, "learning_rate": 3.236405123846651e-06, "loss": 1.6653, "step": 31415 }, { "epoch": 2.462068965517241, "grad_norm": 17.331096649169922, "learning_rate": 3.2361165837871756e-06, "loss": 1.0797, "step": 31416 }, { "epoch": 2.4621473354231975, "grad_norm": 3.7817859649658203, "learning_rate": 3.235828050436608e-06, "loss": 1.3966, "step": 31417 }, { "epoch": 2.4622257053291534, "grad_norm": 3.8702332973480225, "learning_rate": 3.235539523796044e-06, "loss": 1.2358, "step": 31418 }, { "epoch": 2.4623040752351097, "grad_norm": 2.8291330337524414, "learning_rate": 3.2352510038665808e-06, "loss": 1.2122, "step": 31419 }, { "epoch": 2.462382445141066, "grad_norm": 3.425778388977051, "learning_rate": 3.2349624906493164e-06, "loss": 1.3144, "step": 31420 }, { "epoch": 2.462460815047022, "grad_norm": 2.391427993774414, "learning_rate": 3.2346739841453494e-06, "loss": 1.4182, "step": 31421 }, { "epoch": 2.462539184952978, "grad_norm": 4.033340930938721, "learning_rate": 3.2343854843557748e-06, "loss": 1.2125, "step": 31422 }, { "epoch": 2.462617554858934, "grad_norm": 3.36098313331604, "learning_rate": 3.23409699128169e-06, "loss": 1.4055, "step": 31423 }, { "epoch": 2.4626959247648905, "grad_norm": 4.613154888153076, "learning_rate": 3.233808504924194e-06, "loss": 1.4076, "step": 31424 }, { "epoch": 2.4627742946708464, "grad_norm": 2.3605246543884277, "learning_rate": 3.2335200252843847e-06, "loss": 1.258, "step": 31425 }, { "epoch": 2.4628526645768023, "grad_norm": 2.3739664554595947, "learning_rate": 3.2332315523633563e-06, "loss": 0.8624, "step": 31426 }, { "epoch": 2.4629310344827586, "grad_norm": 5.159801483154297, "learning_rate": 3.232943086162208e-06, "loss": 0.9288, "step": 31427 }, { "epoch": 2.463009404388715, "grad_norm": 3.0989251136779785, "learning_rate": 3.232654626682035e-06, "loss": 1.0589, "step": 31428 }, { "epoch": 2.463087774294671, "grad_norm": 2.834970712661743, "learning_rate": 3.2323661739239386e-06, "loss": 0.8677, "step": 31429 }, { "epoch": 2.4631661442006267, "grad_norm": 2.7629711627960205, "learning_rate": 3.2320777278890125e-06, "loss": 1.2175, "step": 31430 }, { "epoch": 2.463244514106583, "grad_norm": 3.446352005004883, "learning_rate": 3.2317892885783543e-06, "loss": 1.2366, "step": 31431 }, { "epoch": 2.4633228840125394, "grad_norm": 3.4035885334014893, "learning_rate": 3.231500855993063e-06, "loss": 1.4848, "step": 31432 }, { "epoch": 2.4634012539184953, "grad_norm": 3.0614492893218994, "learning_rate": 3.231212430134232e-06, "loss": 1.456, "step": 31433 }, { "epoch": 2.463479623824451, "grad_norm": 3.1034419536590576, "learning_rate": 3.2309240110029606e-06, "loss": 1.4529, "step": 31434 }, { "epoch": 2.4635579937304075, "grad_norm": 13.875781059265137, "learning_rate": 3.230635598600346e-06, "loss": 1.4332, "step": 31435 }, { "epoch": 2.463636363636364, "grad_norm": 2.9798035621643066, "learning_rate": 3.2303471929274853e-06, "loss": 1.408, "step": 31436 }, { "epoch": 2.4637147335423197, "grad_norm": 3.9403653144836426, "learning_rate": 3.2300587939854743e-06, "loss": 1.2905, "step": 31437 }, { "epoch": 2.4637931034482756, "grad_norm": 3.458287477493286, "learning_rate": 3.2297704017754105e-06, "loss": 1.4002, "step": 31438 }, { "epoch": 2.463871473354232, "grad_norm": 3.2206871509552, "learning_rate": 3.2294820162983912e-06, "loss": 1.5022, "step": 31439 }, { "epoch": 2.4639498432601883, "grad_norm": 2.2806174755096436, "learning_rate": 3.2291936375555143e-06, "loss": 1.3839, "step": 31440 }, { "epoch": 2.464028213166144, "grad_norm": 1.95643150806427, "learning_rate": 3.2289052655478724e-06, "loss": 1.3007, "step": 31441 }, { "epoch": 2.4641065830721, "grad_norm": 3.1075665950775146, "learning_rate": 3.2286169002765666e-06, "loss": 1.2098, "step": 31442 }, { "epoch": 2.4641849529780564, "grad_norm": 3.8641436100006104, "learning_rate": 3.2283285417426935e-06, "loss": 1.537, "step": 31443 }, { "epoch": 2.4642633228840127, "grad_norm": 7.360417366027832, "learning_rate": 3.228040189947347e-06, "loss": 1.5052, "step": 31444 }, { "epoch": 2.4643416927899686, "grad_norm": 6.666266918182373, "learning_rate": 3.2277518448916255e-06, "loss": 1.5156, "step": 31445 }, { "epoch": 2.464420062695925, "grad_norm": 2.9589216709136963, "learning_rate": 3.2274635065766253e-06, "loss": 1.3553, "step": 31446 }, { "epoch": 2.464498432601881, "grad_norm": 3.731483221054077, "learning_rate": 3.227175175003446e-06, "loss": 1.5464, "step": 31447 }, { "epoch": 2.464576802507837, "grad_norm": 2.358541488647461, "learning_rate": 3.2268868501731788e-06, "loss": 1.2846, "step": 31448 }, { "epoch": 2.464655172413793, "grad_norm": 5.343379497528076, "learning_rate": 3.226598532086924e-06, "loss": 0.6999, "step": 31449 }, { "epoch": 2.4647335423197494, "grad_norm": 3.1912951469421387, "learning_rate": 3.2263102207457788e-06, "loss": 1.3767, "step": 31450 }, { "epoch": 2.4648119122257053, "grad_norm": 4.821523189544678, "learning_rate": 3.2260219161508384e-06, "loss": 1.861, "step": 31451 }, { "epoch": 2.4648902821316616, "grad_norm": 3.2156498432159424, "learning_rate": 3.2257336183031985e-06, "loss": 1.1819, "step": 31452 }, { "epoch": 2.4649686520376175, "grad_norm": 2.8285317420959473, "learning_rate": 3.225445327203956e-06, "loss": 1.2245, "step": 31453 }, { "epoch": 2.465047021943574, "grad_norm": 4.919775485992432, "learning_rate": 3.225157042854209e-06, "loss": 1.4921, "step": 31454 }, { "epoch": 2.4651253918495297, "grad_norm": 3.8260910511016846, "learning_rate": 3.224868765255054e-06, "loss": 1.0871, "step": 31455 }, { "epoch": 2.465203761755486, "grad_norm": 12.049991607666016, "learning_rate": 3.224580494407585e-06, "loss": 0.8742, "step": 31456 }, { "epoch": 2.465282131661442, "grad_norm": 5.633899688720703, "learning_rate": 3.2242922303128997e-06, "loss": 1.0604, "step": 31457 }, { "epoch": 2.4653605015673983, "grad_norm": 3.6234233379364014, "learning_rate": 3.224003972972096e-06, "loss": 1.3705, "step": 31458 }, { "epoch": 2.465438871473354, "grad_norm": 3.809744358062744, "learning_rate": 3.223715722386268e-06, "loss": 1.4321, "step": 31459 }, { "epoch": 2.4655172413793105, "grad_norm": 3.182443857192993, "learning_rate": 3.2234274785565124e-06, "loss": 1.2345, "step": 31460 }, { "epoch": 2.4655956112852664, "grad_norm": 3.4711568355560303, "learning_rate": 3.223139241483927e-06, "loss": 1.2069, "step": 31461 }, { "epoch": 2.4656739811912227, "grad_norm": 5.3912458419799805, "learning_rate": 3.2228510111696077e-06, "loss": 0.9897, "step": 31462 }, { "epoch": 2.4657523510971786, "grad_norm": 2.5587122440338135, "learning_rate": 3.2225627876146504e-06, "loss": 1.5244, "step": 31463 }, { "epoch": 2.465830721003135, "grad_norm": 3.1977014541625977, "learning_rate": 3.222274570820151e-06, "loss": 1.2658, "step": 31464 }, { "epoch": 2.465909090909091, "grad_norm": 3.153578042984009, "learning_rate": 3.2219863607872044e-06, "loss": 0.4496, "step": 31465 }, { "epoch": 2.465987460815047, "grad_norm": 3.2432494163513184, "learning_rate": 3.2216981575169116e-06, "loss": 1.6777, "step": 31466 }, { "epoch": 2.466065830721003, "grad_norm": 9.932188987731934, "learning_rate": 3.221409961010363e-06, "loss": 1.1756, "step": 31467 }, { "epoch": 2.4661442006269594, "grad_norm": 3.317884683609009, "learning_rate": 3.2211217712686584e-06, "loss": 1.6536, "step": 31468 }, { "epoch": 2.4662225705329153, "grad_norm": 3.3308746814727783, "learning_rate": 3.220833588292892e-06, "loss": 1.5924, "step": 31469 }, { "epoch": 2.4663009404388716, "grad_norm": 3.6813931465148926, "learning_rate": 3.2205454120841628e-06, "loss": 1.3234, "step": 31470 }, { "epoch": 2.4663793103448275, "grad_norm": 4.828972339630127, "learning_rate": 3.2202572426435634e-06, "loss": 1.2615, "step": 31471 }, { "epoch": 2.466457680250784, "grad_norm": 3.187509775161743, "learning_rate": 3.2199690799721904e-06, "loss": 1.3127, "step": 31472 }, { "epoch": 2.4665360501567397, "grad_norm": 3.131370782852173, "learning_rate": 3.2196809240711437e-06, "loss": 1.5069, "step": 31473 }, { "epoch": 2.466614420062696, "grad_norm": 2.4725067615509033, "learning_rate": 3.2193927749415134e-06, "loss": 1.2466, "step": 31474 }, { "epoch": 2.466692789968652, "grad_norm": 3.677379608154297, "learning_rate": 3.2191046325843995e-06, "loss": 1.4638, "step": 31475 }, { "epoch": 2.4667711598746083, "grad_norm": 3.9146525859832764, "learning_rate": 3.2188164970008966e-06, "loss": 1.2926, "step": 31476 }, { "epoch": 2.466849529780564, "grad_norm": 2.469176769256592, "learning_rate": 3.218528368192102e-06, "loss": 1.5241, "step": 31477 }, { "epoch": 2.4669278996865205, "grad_norm": 5.78660774230957, "learning_rate": 3.21824024615911e-06, "loss": 1.6182, "step": 31478 }, { "epoch": 2.4670062695924764, "grad_norm": 4.517358779907227, "learning_rate": 3.2179521309030148e-06, "loss": 1.3446, "step": 31479 }, { "epoch": 2.4670846394984327, "grad_norm": 3.768803119659424, "learning_rate": 3.217664022424917e-06, "loss": 1.3829, "step": 31480 }, { "epoch": 2.4671630094043886, "grad_norm": 11.976505279541016, "learning_rate": 3.217375920725909e-06, "loss": 1.7218, "step": 31481 }, { "epoch": 2.467241379310345, "grad_norm": 3.9728052616119385, "learning_rate": 3.2170878258070876e-06, "loss": 1.28, "step": 31482 }, { "epoch": 2.467319749216301, "grad_norm": 2.173346996307373, "learning_rate": 3.2167997376695483e-06, "loss": 1.2037, "step": 31483 }, { "epoch": 2.467398119122257, "grad_norm": 2.5260138511657715, "learning_rate": 3.216511656314385e-06, "loss": 1.2183, "step": 31484 }, { "epoch": 2.467476489028213, "grad_norm": 2.4111733436584473, "learning_rate": 3.216223581742699e-06, "loss": 1.4448, "step": 31485 }, { "epoch": 2.4675548589341694, "grad_norm": 8.545120239257812, "learning_rate": 3.2159355139555793e-06, "loss": 1.0653, "step": 31486 }, { "epoch": 2.4676332288401253, "grad_norm": 2.6497278213500977, "learning_rate": 3.215647452954125e-06, "loss": 1.459, "step": 31487 }, { "epoch": 2.4677115987460816, "grad_norm": 3.6881263256073, "learning_rate": 3.2153593987394327e-06, "loss": 1.4275, "step": 31488 }, { "epoch": 2.4677899686520375, "grad_norm": 4.734177589416504, "learning_rate": 3.2150713513125953e-06, "loss": 0.9176, "step": 31489 }, { "epoch": 2.467868338557994, "grad_norm": 3.8756585121154785, "learning_rate": 3.214783310674711e-06, "loss": 1.7175, "step": 31490 }, { "epoch": 2.4679467084639497, "grad_norm": 5.012007713317871, "learning_rate": 3.214495276826871e-06, "loss": 1.5581, "step": 31491 }, { "epoch": 2.468025078369906, "grad_norm": 7.1802544593811035, "learning_rate": 3.2142072497701783e-06, "loss": 2.114, "step": 31492 }, { "epoch": 2.468103448275862, "grad_norm": 1.933577060699463, "learning_rate": 3.21391922950572e-06, "loss": 1.4346, "step": 31493 }, { "epoch": 2.4681818181818183, "grad_norm": 4.621344089508057, "learning_rate": 3.2136312160345972e-06, "loss": 1.2852, "step": 31494 }, { "epoch": 2.468260188087774, "grad_norm": 5.019816875457764, "learning_rate": 3.213343209357904e-06, "loss": 1.1112, "step": 31495 }, { "epoch": 2.4683385579937305, "grad_norm": 2.819854974746704, "learning_rate": 3.2130552094767362e-06, "loss": 1.2728, "step": 31496 }, { "epoch": 2.4684169278996864, "grad_norm": 5.027682304382324, "learning_rate": 3.2127672163921873e-06, "loss": 1.4387, "step": 31497 }, { "epoch": 2.4684952978056427, "grad_norm": 2.64186429977417, "learning_rate": 3.212479230105353e-06, "loss": 1.1349, "step": 31498 }, { "epoch": 2.4685736677115986, "grad_norm": 3.5265398025512695, "learning_rate": 3.212191250617333e-06, "loss": 1.028, "step": 31499 }, { "epoch": 2.468652037617555, "grad_norm": 2.651679754257202, "learning_rate": 3.211903277929216e-06, "loss": 1.1127, "step": 31500 }, { "epoch": 2.468730407523511, "grad_norm": 6.071453094482422, "learning_rate": 3.2116153120421013e-06, "loss": 1.2306, "step": 31501 }, { "epoch": 2.468808777429467, "grad_norm": 4.222989082336426, "learning_rate": 3.211327352957083e-06, "loss": 1.8436, "step": 31502 }, { "epoch": 2.468887147335423, "grad_norm": 2.3977057933807373, "learning_rate": 3.211039400675259e-06, "loss": 1.2289, "step": 31503 }, { "epoch": 2.4689655172413794, "grad_norm": 2.5645217895507812, "learning_rate": 3.2107514551977204e-06, "loss": 1.2669, "step": 31504 }, { "epoch": 2.4690438871473352, "grad_norm": 3.0141944885253906, "learning_rate": 3.2104635165255634e-06, "loss": 1.4469, "step": 31505 }, { "epoch": 2.4691222570532916, "grad_norm": 4.83717155456543, "learning_rate": 3.210175584659885e-06, "loss": 1.0347, "step": 31506 }, { "epoch": 2.4692006269592475, "grad_norm": 3.2743334770202637, "learning_rate": 3.2098876596017803e-06, "loss": 1.387, "step": 31507 }, { "epoch": 2.469278996865204, "grad_norm": 2.878183126449585, "learning_rate": 3.2095997413523424e-06, "loss": 1.5755, "step": 31508 }, { "epoch": 2.4693573667711597, "grad_norm": 3.9043331146240234, "learning_rate": 3.2093118299126674e-06, "loss": 1.3631, "step": 31509 }, { "epoch": 2.469435736677116, "grad_norm": 3.7245495319366455, "learning_rate": 3.2090239252838496e-06, "loss": 1.124, "step": 31510 }, { "epoch": 2.469514106583072, "grad_norm": 2.5697391033172607, "learning_rate": 3.2087360274669877e-06, "loss": 1.7049, "step": 31511 }, { "epoch": 2.4695924764890282, "grad_norm": 3.4154131412506104, "learning_rate": 3.2084481364631714e-06, "loss": 1.4033, "step": 31512 }, { "epoch": 2.469670846394984, "grad_norm": 3.6151955127716064, "learning_rate": 3.2081602522734987e-06, "loss": 1.4022, "step": 31513 }, { "epoch": 2.4697492163009405, "grad_norm": 2.2577476501464844, "learning_rate": 3.207872374899066e-06, "loss": 1.3697, "step": 31514 }, { "epoch": 2.469827586206897, "grad_norm": 3.5038328170776367, "learning_rate": 3.2075845043409635e-06, "loss": 1.5209, "step": 31515 }, { "epoch": 2.4699059561128527, "grad_norm": 3.3914575576782227, "learning_rate": 3.2072966406002904e-06, "loss": 1.412, "step": 31516 }, { "epoch": 2.4699843260188086, "grad_norm": 5.006132125854492, "learning_rate": 3.207008783678138e-06, "loss": 1.3945, "step": 31517 }, { "epoch": 2.470062695924765, "grad_norm": 2.431267023086548, "learning_rate": 3.2067209335756078e-06, "loss": 1.2043, "step": 31518 }, { "epoch": 2.4701410658307212, "grad_norm": 3.0413002967834473, "learning_rate": 3.2064330902937856e-06, "loss": 1.9521, "step": 31519 }, { "epoch": 2.470219435736677, "grad_norm": 2.9341397285461426, "learning_rate": 3.2061452538337725e-06, "loss": 1.5521, "step": 31520 }, { "epoch": 2.470297805642633, "grad_norm": 8.0979642868042, "learning_rate": 3.205857424196661e-06, "loss": 1.7277, "step": 31521 }, { "epoch": 2.4703761755485893, "grad_norm": 3.6836466789245605, "learning_rate": 3.205569601383548e-06, "loss": 1.562, "step": 31522 }, { "epoch": 2.4704545454545457, "grad_norm": 3.3987419605255127, "learning_rate": 3.2052817853955254e-06, "loss": 1.2309, "step": 31523 }, { "epoch": 2.4705329153605016, "grad_norm": 2.547880172729492, "learning_rate": 3.2049939762336875e-06, "loss": 1.4753, "step": 31524 }, { "epoch": 2.4706112852664575, "grad_norm": 4.968616485595703, "learning_rate": 3.2047061738991312e-06, "loss": 1.3593, "step": 31525 }, { "epoch": 2.470689655172414, "grad_norm": 6.768054485321045, "learning_rate": 3.2044183783929527e-06, "loss": 1.3573, "step": 31526 }, { "epoch": 2.47076802507837, "grad_norm": 2.1648507118225098, "learning_rate": 3.204130589716243e-06, "loss": 1.0821, "step": 31527 }, { "epoch": 2.470846394984326, "grad_norm": 12.553813934326172, "learning_rate": 3.2038428078700978e-06, "loss": 0.9022, "step": 31528 }, { "epoch": 2.470924764890282, "grad_norm": 3.4932498931884766, "learning_rate": 3.203555032855613e-06, "loss": 1.3787, "step": 31529 }, { "epoch": 2.4710031347962382, "grad_norm": 4.244822025299072, "learning_rate": 3.2032672646738805e-06, "loss": 1.4972, "step": 31530 }, { "epoch": 2.4710815047021946, "grad_norm": 11.25269603729248, "learning_rate": 3.202979503325997e-06, "loss": 1.5187, "step": 31531 }, { "epoch": 2.4711598746081505, "grad_norm": 2.3696491718292236, "learning_rate": 3.202691748813057e-06, "loss": 1.2781, "step": 31532 }, { "epoch": 2.4712382445141063, "grad_norm": 2.86553692817688, "learning_rate": 3.202404001136155e-06, "loss": 1.7368, "step": 31533 }, { "epoch": 2.4713166144200627, "grad_norm": 2.808443784713745, "learning_rate": 3.202116260296384e-06, "loss": 1.3003, "step": 31534 }, { "epoch": 2.471394984326019, "grad_norm": 3.1162407398223877, "learning_rate": 3.2018285262948385e-06, "loss": 1.6173, "step": 31535 }, { "epoch": 2.471473354231975, "grad_norm": 4.641122817993164, "learning_rate": 3.201540799132614e-06, "loss": 1.3482, "step": 31536 }, { "epoch": 2.4715517241379312, "grad_norm": 2.4500303268432617, "learning_rate": 3.2012530788108066e-06, "loss": 1.2987, "step": 31537 }, { "epoch": 2.471630094043887, "grad_norm": 3.0579609870910645, "learning_rate": 3.200965365330507e-06, "loss": 1.2408, "step": 31538 }, { "epoch": 2.4717084639498434, "grad_norm": 3.242727041244507, "learning_rate": 3.2006776586928113e-06, "loss": 1.2745, "step": 31539 }, { "epoch": 2.4717868338557993, "grad_norm": 2.8820037841796875, "learning_rate": 3.2003899588988143e-06, "loss": 1.3322, "step": 31540 }, { "epoch": 2.4718652037617557, "grad_norm": 3.9134681224823, "learning_rate": 3.2001022659496084e-06, "loss": 1.467, "step": 31541 }, { "epoch": 2.4719435736677116, "grad_norm": 2.6092758178710938, "learning_rate": 3.1998145798462887e-06, "loss": 1.3153, "step": 31542 }, { "epoch": 2.472021943573668, "grad_norm": 2.43479061126709, "learning_rate": 3.1995269005899498e-06, "loss": 0.8748, "step": 31543 }, { "epoch": 2.472100313479624, "grad_norm": 2.2957539558410645, "learning_rate": 3.1992392281816875e-06, "loss": 1.1796, "step": 31544 }, { "epoch": 2.47217868338558, "grad_norm": 2.8853530883789062, "learning_rate": 3.198951562622593e-06, "loss": 1.2822, "step": 31545 }, { "epoch": 2.472257053291536, "grad_norm": 3.338571071624756, "learning_rate": 3.198663903913762e-06, "loss": 1.1457, "step": 31546 }, { "epoch": 2.4723354231974923, "grad_norm": 3.2810182571411133, "learning_rate": 3.1983762520562866e-06, "loss": 1.1267, "step": 31547 }, { "epoch": 2.472413793103448, "grad_norm": 3.388371229171753, "learning_rate": 3.198088607051266e-06, "loss": 1.2527, "step": 31548 }, { "epoch": 2.4724921630094046, "grad_norm": 2.741041421890259, "learning_rate": 3.197800968899788e-06, "loss": 1.0636, "step": 31549 }, { "epoch": 2.4725705329153604, "grad_norm": 5.099606513977051, "learning_rate": 3.1975133376029505e-06, "loss": 1.7204, "step": 31550 }, { "epoch": 2.4726489028213168, "grad_norm": 6.037658214569092, "learning_rate": 3.197225713161846e-06, "loss": 1.3783, "step": 31551 }, { "epoch": 2.4727272727272727, "grad_norm": 3.0757317543029785, "learning_rate": 3.1969380955775702e-06, "loss": 1.1597, "step": 31552 }, { "epoch": 2.472805642633229, "grad_norm": 8.220702171325684, "learning_rate": 3.196650484851215e-06, "loss": 1.1522, "step": 31553 }, { "epoch": 2.472884012539185, "grad_norm": 5.578499794006348, "learning_rate": 3.1963628809838733e-06, "loss": 2.2948, "step": 31554 }, { "epoch": 2.472962382445141, "grad_norm": 4.202807426452637, "learning_rate": 3.1960752839766447e-06, "loss": 1.1767, "step": 31555 }, { "epoch": 2.473040752351097, "grad_norm": 5.162272930145264, "learning_rate": 3.1957876938306154e-06, "loss": 0.9905, "step": 31556 }, { "epoch": 2.4731191222570534, "grad_norm": 3.425614833831787, "learning_rate": 3.195500110546884e-06, "loss": 1.3715, "step": 31557 }, { "epoch": 2.4731974921630093, "grad_norm": 3.604992628097534, "learning_rate": 3.195212534126544e-06, "loss": 1.3241, "step": 31558 }, { "epoch": 2.4732758620689657, "grad_norm": 4.135581016540527, "learning_rate": 3.1949249645706893e-06, "loss": 0.8945, "step": 31559 }, { "epoch": 2.4733542319749215, "grad_norm": 5.963071823120117, "learning_rate": 3.1946374018804115e-06, "loss": 1.4856, "step": 31560 }, { "epoch": 2.473432601880878, "grad_norm": 4.092265605926514, "learning_rate": 3.1943498460568057e-06, "loss": 1.2511, "step": 31561 }, { "epoch": 2.4735109717868338, "grad_norm": 3.2472543716430664, "learning_rate": 3.194062297100966e-06, "loss": 0.8147, "step": 31562 }, { "epoch": 2.47358934169279, "grad_norm": 3.974630832672119, "learning_rate": 3.1937747550139865e-06, "loss": 1.2945, "step": 31563 }, { "epoch": 2.473667711598746, "grad_norm": 14.914551734924316, "learning_rate": 3.1934872197969592e-06, "loss": 1.3106, "step": 31564 }, { "epoch": 2.4737460815047023, "grad_norm": 4.269151210784912, "learning_rate": 3.193199691450979e-06, "loss": 1.199, "step": 31565 }, { "epoch": 2.473824451410658, "grad_norm": 2.4890143871307373, "learning_rate": 3.1929121699771376e-06, "loss": 1.1414, "step": 31566 }, { "epoch": 2.4739028213166145, "grad_norm": 3.55271053314209, "learning_rate": 3.192624655376534e-06, "loss": 1.3751, "step": 31567 }, { "epoch": 2.4739811912225704, "grad_norm": 3.8828508853912354, "learning_rate": 3.192337147650254e-06, "loss": 1.2877, "step": 31568 }, { "epoch": 2.4740595611285268, "grad_norm": 5.253017902374268, "learning_rate": 3.1920496467993962e-06, "loss": 1.8111, "step": 31569 }, { "epoch": 2.4741379310344827, "grad_norm": 3.3802926540374756, "learning_rate": 3.191762152825054e-06, "loss": 1.5214, "step": 31570 }, { "epoch": 2.474216300940439, "grad_norm": 2.159834384918213, "learning_rate": 3.191474665728319e-06, "loss": 1.3971, "step": 31571 }, { "epoch": 2.474294670846395, "grad_norm": 3.283226251602173, "learning_rate": 3.191187185510285e-06, "loss": 1.3942, "step": 31572 }, { "epoch": 2.474373040752351, "grad_norm": 2.858062505722046, "learning_rate": 3.190899712172045e-06, "loss": 0.9177, "step": 31573 }, { "epoch": 2.474451410658307, "grad_norm": 3.0769786834716797, "learning_rate": 3.190612245714697e-06, "loss": 1.5681, "step": 31574 }, { "epoch": 2.4745297805642634, "grad_norm": 3.208432674407959, "learning_rate": 3.1903247861393273e-06, "loss": 1.272, "step": 31575 }, { "epoch": 2.4746081504702193, "grad_norm": 3.1068837642669678, "learning_rate": 3.1900373334470324e-06, "loss": 1.3372, "step": 31576 }, { "epoch": 2.4746865203761756, "grad_norm": 3.015709161758423, "learning_rate": 3.1897498876389076e-06, "loss": 1.3987, "step": 31577 }, { "epoch": 2.4747648902821315, "grad_norm": 3.720716714859009, "learning_rate": 3.1894624487160442e-06, "loss": 0.8862, "step": 31578 }, { "epoch": 2.474843260188088, "grad_norm": 3.2122576236724854, "learning_rate": 3.1891750166795355e-06, "loss": 0.9185, "step": 31579 }, { "epoch": 2.4749216300940438, "grad_norm": 8.543275833129883, "learning_rate": 3.1888875915304735e-06, "loss": 1.0909, "step": 31580 }, { "epoch": 2.475, "grad_norm": 2.88081955909729, "learning_rate": 3.188600173269954e-06, "loss": 1.2452, "step": 31581 }, { "epoch": 2.475078369905956, "grad_norm": 5.708848476409912, "learning_rate": 3.18831276189907e-06, "loss": 1.2875, "step": 31582 }, { "epoch": 2.4751567398119123, "grad_norm": 6.2572503089904785, "learning_rate": 3.188025357418912e-06, "loss": 1.6948, "step": 31583 }, { "epoch": 2.475235109717868, "grad_norm": 7.372143268585205, "learning_rate": 3.1877379598305757e-06, "loss": 0.9483, "step": 31584 }, { "epoch": 2.4753134796238245, "grad_norm": 2.606729745864868, "learning_rate": 3.187450569135155e-06, "loss": 1.1005, "step": 31585 }, { "epoch": 2.4753918495297804, "grad_norm": 3.7668824195861816, "learning_rate": 3.1871631853337394e-06, "loss": 1.6205, "step": 31586 }, { "epoch": 2.4754702194357368, "grad_norm": 3.38641619682312, "learning_rate": 3.1868758084274233e-06, "loss": 1.3778, "step": 31587 }, { "epoch": 2.4755485893416926, "grad_norm": 3.2294232845306396, "learning_rate": 3.186588438417301e-06, "loss": 1.3688, "step": 31588 }, { "epoch": 2.475626959247649, "grad_norm": 2.058833599090576, "learning_rate": 3.1863010753044652e-06, "loss": 1.2688, "step": 31589 }, { "epoch": 2.475705329153605, "grad_norm": 3.7051994800567627, "learning_rate": 3.1860137190900087e-06, "loss": 1.147, "step": 31590 }, { "epoch": 2.475783699059561, "grad_norm": 3.123957633972168, "learning_rate": 3.1857263697750242e-06, "loss": 1.5043, "step": 31591 }, { "epoch": 2.475862068965517, "grad_norm": 3.404240846633911, "learning_rate": 3.1854390273606034e-06, "loss": 1.1827, "step": 31592 }, { "epoch": 2.4759404388714734, "grad_norm": 2.728330373764038, "learning_rate": 3.185151691847843e-06, "loss": 1.2133, "step": 31593 }, { "epoch": 2.4760188087774293, "grad_norm": 3.085188388824463, "learning_rate": 3.184864363237832e-06, "loss": 1.5769, "step": 31594 }, { "epoch": 2.4760971786833856, "grad_norm": 13.537055969238281, "learning_rate": 3.184577041531664e-06, "loss": 1.7722, "step": 31595 }, { "epoch": 2.4761755485893415, "grad_norm": 3.546104669570923, "learning_rate": 3.184289726730434e-06, "loss": 1.0158, "step": 31596 }, { "epoch": 2.476253918495298, "grad_norm": 2.186168670654297, "learning_rate": 3.1840024188352325e-06, "loss": 1.247, "step": 31597 }, { "epoch": 2.4763322884012537, "grad_norm": 10.930468559265137, "learning_rate": 3.1837151178471526e-06, "loss": 0.4245, "step": 31598 }, { "epoch": 2.47641065830721, "grad_norm": 5.322975158691406, "learning_rate": 3.1834278237672866e-06, "loss": 1.4947, "step": 31599 }, { "epoch": 2.476489028213166, "grad_norm": 4.588572025299072, "learning_rate": 3.1831405365967315e-06, "loss": 1.756, "step": 31600 }, { "epoch": 2.4765673981191223, "grad_norm": 4.733300685882568, "learning_rate": 3.182853256336573e-06, "loss": 1.4103, "step": 31601 }, { "epoch": 2.476645768025078, "grad_norm": 2.9454870223999023, "learning_rate": 3.1825659829879096e-06, "loss": 1.4637, "step": 31602 }, { "epoch": 2.4767241379310345, "grad_norm": 6.8021039962768555, "learning_rate": 3.182278716551831e-06, "loss": 1.5624, "step": 31603 }, { "epoch": 2.4768025078369904, "grad_norm": 3.7524096965789795, "learning_rate": 3.1819914570294306e-06, "loss": 1.5267, "step": 31604 }, { "epoch": 2.4768808777429467, "grad_norm": 4.007514476776123, "learning_rate": 3.181704204421801e-06, "loss": 1.0354, "step": 31605 }, { "epoch": 2.4769592476489026, "grad_norm": 3.190441131591797, "learning_rate": 3.1814169587300336e-06, "loss": 0.9458, "step": 31606 }, { "epoch": 2.477037617554859, "grad_norm": 2.4275500774383545, "learning_rate": 3.1811297199552226e-06, "loss": 1.3377, "step": 31607 }, { "epoch": 2.477115987460815, "grad_norm": 6.48137092590332, "learning_rate": 3.1808424880984613e-06, "loss": 2.0626, "step": 31608 }, { "epoch": 2.477194357366771, "grad_norm": 6.0239996910095215, "learning_rate": 3.1805552631608394e-06, "loss": 1.107, "step": 31609 }, { "epoch": 2.4772727272727275, "grad_norm": 7.164832592010498, "learning_rate": 3.180268045143451e-06, "loss": 0.9745, "step": 31610 }, { "epoch": 2.4773510971786834, "grad_norm": 2.460789442062378, "learning_rate": 3.179980834047389e-06, "loss": 1.3138, "step": 31611 }, { "epoch": 2.4774294670846393, "grad_norm": 6.3070759773254395, "learning_rate": 3.1796936298737444e-06, "loss": 1.8556, "step": 31612 }, { "epoch": 2.4775078369905956, "grad_norm": 5.367620468139648, "learning_rate": 3.1794064326236094e-06, "loss": 1.4794, "step": 31613 }, { "epoch": 2.477586206896552, "grad_norm": 3.3007023334503174, "learning_rate": 3.1791192422980776e-06, "loss": 1.078, "step": 31614 }, { "epoch": 2.477664576802508, "grad_norm": 3.628533363342285, "learning_rate": 3.1788320588982424e-06, "loss": 1.0308, "step": 31615 }, { "epoch": 2.4777429467084637, "grad_norm": 2.530945301055908, "learning_rate": 3.178544882425194e-06, "loss": 1.5223, "step": 31616 }, { "epoch": 2.47782131661442, "grad_norm": 2.4092373847961426, "learning_rate": 3.1782577128800245e-06, "loss": 1.4042, "step": 31617 }, { "epoch": 2.4778996865203764, "grad_norm": 2.5423223972320557, "learning_rate": 3.177970550263826e-06, "loss": 1.257, "step": 31618 }, { "epoch": 2.4779780564263323, "grad_norm": 7.042460918426514, "learning_rate": 3.1776833945776953e-06, "loss": 1.1091, "step": 31619 }, { "epoch": 2.478056426332288, "grad_norm": 2.5936665534973145, "learning_rate": 3.177396245822718e-06, "loss": 1.3409, "step": 31620 }, { "epoch": 2.4781347962382445, "grad_norm": 2.7807462215423584, "learning_rate": 3.1771091039999894e-06, "loss": 1.1463, "step": 31621 }, { "epoch": 2.478213166144201, "grad_norm": 3.3937392234802246, "learning_rate": 3.1768219691106024e-06, "loss": 1.2803, "step": 31622 }, { "epoch": 2.4782915360501567, "grad_norm": 3.1914572715759277, "learning_rate": 3.1765348411556483e-06, "loss": 1.2932, "step": 31623 }, { "epoch": 2.4783699059561126, "grad_norm": 3.8414745330810547, "learning_rate": 3.1762477201362186e-06, "loss": 1.569, "step": 31624 }, { "epoch": 2.478448275862069, "grad_norm": 11.113465309143066, "learning_rate": 3.1759606060534048e-06, "loss": 1.6798, "step": 31625 }, { "epoch": 2.4785266457680253, "grad_norm": 4.105170249938965, "learning_rate": 3.175673498908302e-06, "loss": 1.3714, "step": 31626 }, { "epoch": 2.478605015673981, "grad_norm": 4.2671332359313965, "learning_rate": 3.1753863987019994e-06, "loss": 1.0351, "step": 31627 }, { "epoch": 2.478683385579937, "grad_norm": 2.986966848373413, "learning_rate": 3.1750993054355895e-06, "loss": 1.2742, "step": 31628 }, { "epoch": 2.4787617554858934, "grad_norm": 3.4702229499816895, "learning_rate": 3.1748122191101643e-06, "loss": 1.1882, "step": 31629 }, { "epoch": 2.4788401253918497, "grad_norm": 2.4138343334198, "learning_rate": 3.1745251397268175e-06, "loss": 1.4982, "step": 31630 }, { "epoch": 2.4789184952978056, "grad_norm": 11.248315811157227, "learning_rate": 3.174238067286637e-06, "loss": 2.1596, "step": 31631 }, { "epoch": 2.478996865203762, "grad_norm": 2.6036572456359863, "learning_rate": 3.1739510017907176e-06, "loss": 1.3959, "step": 31632 }, { "epoch": 2.479075235109718, "grad_norm": 1.954628825187683, "learning_rate": 3.1736639432401516e-06, "loss": 1.1761, "step": 31633 }, { "epoch": 2.479153605015674, "grad_norm": 3.698554039001465, "learning_rate": 3.173376891636031e-06, "loss": 1.3466, "step": 31634 }, { "epoch": 2.47923197492163, "grad_norm": 2.723989486694336, "learning_rate": 3.1730898469794445e-06, "loss": 1.1208, "step": 31635 }, { "epoch": 2.4793103448275864, "grad_norm": 2.2813515663146973, "learning_rate": 3.172802809271486e-06, "loss": 1.3666, "step": 31636 }, { "epoch": 2.4793887147335423, "grad_norm": 2.0709474086761475, "learning_rate": 3.1725157785132465e-06, "loss": 1.1927, "step": 31637 }, { "epoch": 2.4794670846394986, "grad_norm": 5.391843795776367, "learning_rate": 3.1722287547058205e-06, "loss": 1.1422, "step": 31638 }, { "epoch": 2.4795454545454545, "grad_norm": 1.9729514122009277, "learning_rate": 3.1719417378502963e-06, "loss": 1.3691, "step": 31639 }, { "epoch": 2.479623824451411, "grad_norm": 6.780984878540039, "learning_rate": 3.171654727947767e-06, "loss": 1.5835, "step": 31640 }, { "epoch": 2.4797021943573667, "grad_norm": 3.182861804962158, "learning_rate": 3.171367724999325e-06, "loss": 0.9182, "step": 31641 }, { "epoch": 2.479780564263323, "grad_norm": 3.3114562034606934, "learning_rate": 3.1710807290060596e-06, "loss": 1.3708, "step": 31642 }, { "epoch": 2.479858934169279, "grad_norm": 2.0712950229644775, "learning_rate": 3.170793739969063e-06, "loss": 1.0517, "step": 31643 }, { "epoch": 2.4799373040752353, "grad_norm": 5.066988468170166, "learning_rate": 3.1705067578894284e-06, "loss": 1.4565, "step": 31644 }, { "epoch": 2.480015673981191, "grad_norm": 2.196326971054077, "learning_rate": 3.1702197827682476e-06, "loss": 1.2054, "step": 31645 }, { "epoch": 2.4800940438871475, "grad_norm": 4.1535258293151855, "learning_rate": 3.1699328146066096e-06, "loss": 1.3871, "step": 31646 }, { "epoch": 2.4801724137931034, "grad_norm": 3.55708909034729, "learning_rate": 3.1696458534056076e-06, "loss": 1.5527, "step": 31647 }, { "epoch": 2.4802507836990597, "grad_norm": 3.0456032752990723, "learning_rate": 3.1693588991663304e-06, "loss": 1.1035, "step": 31648 }, { "epoch": 2.4803291536050156, "grad_norm": 2.544940948486328, "learning_rate": 3.169071951889876e-06, "loss": 1.4276, "step": 31649 }, { "epoch": 2.480407523510972, "grad_norm": 4.877967357635498, "learning_rate": 3.168785011577328e-06, "loss": 1.5192, "step": 31650 }, { "epoch": 2.480485893416928, "grad_norm": 15.88271713256836, "learning_rate": 3.1684980782297825e-06, "loss": 1.783, "step": 31651 }, { "epoch": 2.480564263322884, "grad_norm": 2.2938644886016846, "learning_rate": 3.1682111518483305e-06, "loss": 1.543, "step": 31652 }, { "epoch": 2.48064263322884, "grad_norm": 3.6675963401794434, "learning_rate": 3.1679242324340608e-06, "loss": 1.1995, "step": 31653 }, { "epoch": 2.4807210031347964, "grad_norm": 1.9956413507461548, "learning_rate": 3.1676373199880672e-06, "loss": 0.9451, "step": 31654 }, { "epoch": 2.4807993730407523, "grad_norm": 3.036996364593506, "learning_rate": 3.1673504145114388e-06, "loss": 1.2598, "step": 31655 }, { "epoch": 2.4808777429467086, "grad_norm": 2.810720443725586, "learning_rate": 3.167063516005271e-06, "loss": 1.3943, "step": 31656 }, { "epoch": 2.4809561128526645, "grad_norm": 4.869651794433594, "learning_rate": 3.166776624470649e-06, "loss": 1.6476, "step": 31657 }, { "epoch": 2.481034482758621, "grad_norm": 2.364508867263794, "learning_rate": 3.166489739908668e-06, "loss": 0.9978, "step": 31658 }, { "epoch": 2.4811128526645767, "grad_norm": 3.1791160106658936, "learning_rate": 3.166202862320419e-06, "loss": 1.029, "step": 31659 }, { "epoch": 2.481191222570533, "grad_norm": 3.2994346618652344, "learning_rate": 3.1659159917069927e-06, "loss": 1.5541, "step": 31660 }, { "epoch": 2.481269592476489, "grad_norm": 7.420530319213867, "learning_rate": 3.1656291280694796e-06, "loss": 1.6837, "step": 31661 }, { "epoch": 2.4813479623824453, "grad_norm": 3.564688205718994, "learning_rate": 3.165342271408969e-06, "loss": 1.3629, "step": 31662 }, { "epoch": 2.481426332288401, "grad_norm": 3.2125470638275146, "learning_rate": 3.165055421726556e-06, "loss": 1.4249, "step": 31663 }, { "epoch": 2.4815047021943575, "grad_norm": 2.2538254261016846, "learning_rate": 3.1647685790233305e-06, "loss": 1.2534, "step": 31664 }, { "epoch": 2.4815830721003134, "grad_norm": 2.927344799041748, "learning_rate": 3.1644817433003815e-06, "loss": 1.5416, "step": 31665 }, { "epoch": 2.4816614420062697, "grad_norm": 3.192227602005005, "learning_rate": 3.1641949145588013e-06, "loss": 1.3574, "step": 31666 }, { "epoch": 2.4817398119122256, "grad_norm": 11.192174911499023, "learning_rate": 3.1639080927996813e-06, "loss": 1.5506, "step": 31667 }, { "epoch": 2.481818181818182, "grad_norm": 4.159769535064697, "learning_rate": 3.1636212780241116e-06, "loss": 1.5858, "step": 31668 }, { "epoch": 2.481896551724138, "grad_norm": 2.17222261428833, "learning_rate": 3.163334470233182e-06, "loss": 1.4081, "step": 31669 }, { "epoch": 2.481974921630094, "grad_norm": 3.131465435028076, "learning_rate": 3.1630476694279855e-06, "loss": 1.5507, "step": 31670 }, { "epoch": 2.48205329153605, "grad_norm": 3.796107769012451, "learning_rate": 3.1627608756096135e-06, "loss": 1.6649, "step": 31671 }, { "epoch": 2.4821316614420064, "grad_norm": 2.120770215988159, "learning_rate": 3.1624740887791537e-06, "loss": 1.3507, "step": 31672 }, { "epoch": 2.4822100313479623, "grad_norm": 2.643446922302246, "learning_rate": 3.1621873089377e-06, "loss": 1.1941, "step": 31673 }, { "epoch": 2.4822884012539186, "grad_norm": 2.5285377502441406, "learning_rate": 3.1619005360863398e-06, "loss": 0.7054, "step": 31674 }, { "epoch": 2.4823667711598745, "grad_norm": 10.276544570922852, "learning_rate": 3.161613770226169e-06, "loss": 1.0215, "step": 31675 }, { "epoch": 2.482445141065831, "grad_norm": 3.0087053775787354, "learning_rate": 3.1613270113582724e-06, "loss": 1.4819, "step": 31676 }, { "epoch": 2.4825235109717867, "grad_norm": 3.096074342727661, "learning_rate": 3.1610402594837442e-06, "loss": 1.8354, "step": 31677 }, { "epoch": 2.482601880877743, "grad_norm": 2.543566942214966, "learning_rate": 3.1607535146036745e-06, "loss": 1.4547, "step": 31678 }, { "epoch": 2.482680250783699, "grad_norm": 2.2162914276123047, "learning_rate": 3.1604667767191545e-06, "loss": 1.2797, "step": 31679 }, { "epoch": 2.4827586206896552, "grad_norm": 2.350904703140259, "learning_rate": 3.1601800458312732e-06, "loss": 1.2509, "step": 31680 }, { "epoch": 2.482836990595611, "grad_norm": 6.224969863891602, "learning_rate": 3.1598933219411208e-06, "loss": 0.7677, "step": 31681 }, { "epoch": 2.4829153605015675, "grad_norm": 8.2008638381958, "learning_rate": 3.1596066050497928e-06, "loss": 1.495, "step": 31682 }, { "epoch": 2.4829937304075234, "grad_norm": 3.1384530067443848, "learning_rate": 3.1593198951583725e-06, "loss": 1.4306, "step": 31683 }, { "epoch": 2.4830721003134797, "grad_norm": 2.5278453826904297, "learning_rate": 3.1590331922679545e-06, "loss": 1.1911, "step": 31684 }, { "epoch": 2.4831504702194356, "grad_norm": 3.927694082260132, "learning_rate": 3.1587464963796288e-06, "loss": 0.8246, "step": 31685 }, { "epoch": 2.483228840125392, "grad_norm": 2.9032115936279297, "learning_rate": 3.158459807494487e-06, "loss": 1.4157, "step": 31686 }, { "epoch": 2.483307210031348, "grad_norm": 3.8778316974639893, "learning_rate": 3.1581731256136173e-06, "loss": 1.3192, "step": 31687 }, { "epoch": 2.483385579937304, "grad_norm": 4.280791282653809, "learning_rate": 3.1578864507381096e-06, "loss": 1.2922, "step": 31688 }, { "epoch": 2.48346394984326, "grad_norm": 3.968885660171509, "learning_rate": 3.1575997828690574e-06, "loss": 1.1804, "step": 31689 }, { "epoch": 2.4835423197492164, "grad_norm": 3.234163522720337, "learning_rate": 3.1573131220075494e-06, "loss": 1.3587, "step": 31690 }, { "epoch": 2.4836206896551722, "grad_norm": 3.2355289459228516, "learning_rate": 3.1570264681546753e-06, "loss": 1.4315, "step": 31691 }, { "epoch": 2.4836990595611286, "grad_norm": 3.473987340927124, "learning_rate": 3.1567398213115257e-06, "loss": 1.6773, "step": 31692 }, { "epoch": 2.4837774294670845, "grad_norm": 16.98402976989746, "learning_rate": 3.15645318147919e-06, "loss": 0.809, "step": 31693 }, { "epoch": 2.483855799373041, "grad_norm": 3.0347461700439453, "learning_rate": 3.1561665486587635e-06, "loss": 1.4837, "step": 31694 }, { "epoch": 2.4839341692789967, "grad_norm": 3.0963046550750732, "learning_rate": 3.155879922851328e-06, "loss": 0.9448, "step": 31695 }, { "epoch": 2.484012539184953, "grad_norm": 2.410057783126831, "learning_rate": 3.15559330405798e-06, "loss": 1.3954, "step": 31696 }, { "epoch": 2.484090909090909, "grad_norm": 5.839309215545654, "learning_rate": 3.1553066922798082e-06, "loss": 1.0093, "step": 31697 }, { "epoch": 2.4841692789968652, "grad_norm": 2.6274330615997314, "learning_rate": 3.1550200875179018e-06, "loss": 1.0913, "step": 31698 }, { "epoch": 2.484247648902821, "grad_norm": 4.300690650939941, "learning_rate": 3.1547334897733507e-06, "loss": 1.7036, "step": 31699 }, { "epoch": 2.4843260188087775, "grad_norm": 2.7670271396636963, "learning_rate": 3.1544468990472443e-06, "loss": 1.2362, "step": 31700 }, { "epoch": 2.484404388714734, "grad_norm": 2.2948689460754395, "learning_rate": 3.154160315340678e-06, "loss": 1.4767, "step": 31701 }, { "epoch": 2.4844827586206897, "grad_norm": 3.2162106037139893, "learning_rate": 3.1538737386547343e-06, "loss": 0.9164, "step": 31702 }, { "epoch": 2.4845611285266456, "grad_norm": 3.2070295810699463, "learning_rate": 3.1535871689905074e-06, "loss": 1.4614, "step": 31703 }, { "epoch": 2.484639498432602, "grad_norm": 2.414773941040039, "learning_rate": 3.1533006063490867e-06, "loss": 1.7432, "step": 31704 }, { "epoch": 2.4847178683385582, "grad_norm": 2.568636655807495, "learning_rate": 3.1530140507315634e-06, "loss": 1.4406, "step": 31705 }, { "epoch": 2.484796238244514, "grad_norm": 3.2072412967681885, "learning_rate": 3.1527275021390247e-06, "loss": 1.5338, "step": 31706 }, { "epoch": 2.48487460815047, "grad_norm": 4.327556610107422, "learning_rate": 3.1524409605725604e-06, "loss": 1.0791, "step": 31707 }, { "epoch": 2.4849529780564263, "grad_norm": 4.857437610626221, "learning_rate": 3.1521544260332652e-06, "loss": 1.9129, "step": 31708 }, { "epoch": 2.4850313479623827, "grad_norm": 4.29298734664917, "learning_rate": 3.151867898522223e-06, "loss": 1.6349, "step": 31709 }, { "epoch": 2.4851097178683386, "grad_norm": 2.8072776794433594, "learning_rate": 3.151581378040526e-06, "loss": 1.1657, "step": 31710 }, { "epoch": 2.4851880877742945, "grad_norm": 2.7095296382904053, "learning_rate": 3.1512948645892645e-06, "loss": 1.2286, "step": 31711 }, { "epoch": 2.485266457680251, "grad_norm": 5.6959452629089355, "learning_rate": 3.151008358169528e-06, "loss": 1.8114, "step": 31712 }, { "epoch": 2.485344827586207, "grad_norm": 9.745676040649414, "learning_rate": 3.150721858782406e-06, "loss": 0.7588, "step": 31713 }, { "epoch": 2.485423197492163, "grad_norm": 1.8439639806747437, "learning_rate": 3.1504353664289867e-06, "loss": 0.8055, "step": 31714 }, { "epoch": 2.485501567398119, "grad_norm": 3.4735701084136963, "learning_rate": 3.1501488811103623e-06, "loss": 1.1008, "step": 31715 }, { "epoch": 2.4855799373040752, "grad_norm": 3.3916842937469482, "learning_rate": 3.1498624028276227e-06, "loss": 1.5864, "step": 31716 }, { "epoch": 2.4856583072100316, "grad_norm": 2.9762840270996094, "learning_rate": 3.1495759315818552e-06, "loss": 1.3064, "step": 31717 }, { "epoch": 2.4857366771159874, "grad_norm": 10.683650970458984, "learning_rate": 3.14928946737415e-06, "loss": 1.4086, "step": 31718 }, { "epoch": 2.4858150470219433, "grad_norm": 4.404536247253418, "learning_rate": 3.149003010205597e-06, "loss": 1.4137, "step": 31719 }, { "epoch": 2.4858934169278997, "grad_norm": 2.20415997505188, "learning_rate": 3.1487165600772883e-06, "loss": 1.2657, "step": 31720 }, { "epoch": 2.485971786833856, "grad_norm": 2.2094833850860596, "learning_rate": 3.1484301169903075e-06, "loss": 1.5871, "step": 31721 }, { "epoch": 2.486050156739812, "grad_norm": 3.6149003505706787, "learning_rate": 3.1481436809457487e-06, "loss": 1.3772, "step": 31722 }, { "epoch": 2.4861285266457678, "grad_norm": 2.414283514022827, "learning_rate": 3.1478572519447014e-06, "loss": 1.38, "step": 31723 }, { "epoch": 2.486206896551724, "grad_norm": 2.6715400218963623, "learning_rate": 3.1475708299882525e-06, "loss": 1.5156, "step": 31724 }, { "epoch": 2.4862852664576804, "grad_norm": 2.288405656814575, "learning_rate": 3.1472844150774927e-06, "loss": 1.5565, "step": 31725 }, { "epoch": 2.4863636363636363, "grad_norm": 2.5629844665527344, "learning_rate": 3.146998007213511e-06, "loss": 1.4469, "step": 31726 }, { "epoch": 2.4864420062695927, "grad_norm": 2.4917213916778564, "learning_rate": 3.146711606397399e-06, "loss": 1.3508, "step": 31727 }, { "epoch": 2.4865203761755486, "grad_norm": 2.6266753673553467, "learning_rate": 3.1464252126302425e-06, "loss": 1.1114, "step": 31728 }, { "epoch": 2.486598746081505, "grad_norm": 5.496637344360352, "learning_rate": 3.1461388259131324e-06, "loss": 1.6883, "step": 31729 }, { "epoch": 2.4866771159874608, "grad_norm": 21.5977783203125, "learning_rate": 3.1458524462471586e-06, "loss": 1.1516, "step": 31730 }, { "epoch": 2.486755485893417, "grad_norm": 3.323593854904175, "learning_rate": 3.1455660736334105e-06, "loss": 1.3439, "step": 31731 }, { "epoch": 2.486833855799373, "grad_norm": 2.792644500732422, "learning_rate": 3.145279708072976e-06, "loss": 1.292, "step": 31732 }, { "epoch": 2.4869122257053293, "grad_norm": 5.544990539550781, "learning_rate": 3.1449933495669427e-06, "loss": 1.1062, "step": 31733 }, { "epoch": 2.486990595611285, "grad_norm": 3.4036972522735596, "learning_rate": 3.1447069981164034e-06, "loss": 1.4025, "step": 31734 }, { "epoch": 2.4870689655172415, "grad_norm": 2.9196743965148926, "learning_rate": 3.1444206537224477e-06, "loss": 0.8934, "step": 31735 }, { "epoch": 2.4871473354231974, "grad_norm": 3.37678599357605, "learning_rate": 3.1441343163861603e-06, "loss": 1.4766, "step": 31736 }, { "epoch": 2.4872257053291538, "grad_norm": 4.746060371398926, "learning_rate": 3.1438479861086333e-06, "loss": 1.0546, "step": 31737 }, { "epoch": 2.4873040752351097, "grad_norm": 4.974003791809082, "learning_rate": 3.143561662890957e-06, "loss": 1.1006, "step": 31738 }, { "epoch": 2.487382445141066, "grad_norm": 6.515688896179199, "learning_rate": 3.143275346734216e-06, "loss": 1.1967, "step": 31739 }, { "epoch": 2.487460815047022, "grad_norm": 5.783251762390137, "learning_rate": 3.1429890376395024e-06, "loss": 1.5689, "step": 31740 }, { "epoch": 2.487539184952978, "grad_norm": 3.124985456466675, "learning_rate": 3.1427027356079052e-06, "loss": 1.1439, "step": 31741 }, { "epoch": 2.487617554858934, "grad_norm": 2.986419677734375, "learning_rate": 3.1424164406405137e-06, "loss": 1.6315, "step": 31742 }, { "epoch": 2.4876959247648904, "grad_norm": 3.5053536891937256, "learning_rate": 3.142130152738415e-06, "loss": 1.6506, "step": 31743 }, { "epoch": 2.4877742946708463, "grad_norm": 3.112154483795166, "learning_rate": 3.141843871902698e-06, "loss": 1.2288, "step": 31744 }, { "epoch": 2.4878526645768027, "grad_norm": 3.9569344520568848, "learning_rate": 3.1415575981344536e-06, "loss": 1.5103, "step": 31745 }, { "epoch": 2.4879310344827585, "grad_norm": 2.948782444000244, "learning_rate": 3.14127133143477e-06, "loss": 1.5387, "step": 31746 }, { "epoch": 2.488009404388715, "grad_norm": 3.2320311069488525, "learning_rate": 3.1409850718047353e-06, "loss": 1.501, "step": 31747 }, { "epoch": 2.4880877742946708, "grad_norm": 2.475259780883789, "learning_rate": 3.140698819245438e-06, "loss": 1.4235, "step": 31748 }, { "epoch": 2.488166144200627, "grad_norm": 4.676602840423584, "learning_rate": 3.140412573757966e-06, "loss": 1.2667, "step": 31749 }, { "epoch": 2.488244514106583, "grad_norm": 1.9871007204055786, "learning_rate": 3.140126335343413e-06, "loss": 1.3059, "step": 31750 }, { "epoch": 2.4883228840125393, "grad_norm": 2.8326306343078613, "learning_rate": 3.1398401040028615e-06, "loss": 1.2269, "step": 31751 }, { "epoch": 2.488401253918495, "grad_norm": 2.767568826675415, "learning_rate": 3.1395538797374037e-06, "loss": 1.2689, "step": 31752 }, { "epoch": 2.4884796238244515, "grad_norm": 4.194568157196045, "learning_rate": 3.1392676625481277e-06, "loss": 1.0839, "step": 31753 }, { "epoch": 2.4885579937304074, "grad_norm": 4.086074352264404, "learning_rate": 3.138981452436121e-06, "loss": 1.5606, "step": 31754 }, { "epoch": 2.4886363636363638, "grad_norm": 2.7037322521209717, "learning_rate": 3.1386952494024738e-06, "loss": 1.2373, "step": 31755 }, { "epoch": 2.4887147335423196, "grad_norm": 6.125793933868408, "learning_rate": 3.1384090534482714e-06, "loss": 1.3614, "step": 31756 }, { "epoch": 2.488793103448276, "grad_norm": 4.769283771514893, "learning_rate": 3.1381228645746085e-06, "loss": 1.1669, "step": 31757 }, { "epoch": 2.488871473354232, "grad_norm": 12.13432788848877, "learning_rate": 3.1378366827825667e-06, "loss": 1.0311, "step": 31758 }, { "epoch": 2.488949843260188, "grad_norm": 2.7120437622070312, "learning_rate": 3.1375505080732394e-06, "loss": 1.2166, "step": 31759 }, { "epoch": 2.489028213166144, "grad_norm": 4.948678970336914, "learning_rate": 3.137264340447713e-06, "loss": 1.3773, "step": 31760 }, { "epoch": 2.4891065830721004, "grad_norm": 3.991809606552124, "learning_rate": 3.136978179907077e-06, "loss": 1.2757, "step": 31761 }, { "epoch": 2.4891849529780563, "grad_norm": 2.2745025157928467, "learning_rate": 3.136692026452418e-06, "loss": 1.1242, "step": 31762 }, { "epoch": 2.4892633228840126, "grad_norm": 4.561585426330566, "learning_rate": 3.136405880084825e-06, "loss": 1.3676, "step": 31763 }, { "epoch": 2.4893416927899685, "grad_norm": 4.324941635131836, "learning_rate": 3.1361197408053894e-06, "loss": 1.5325, "step": 31764 }, { "epoch": 2.489420062695925, "grad_norm": 2.699955940246582, "learning_rate": 3.135833608615194e-06, "loss": 1.1386, "step": 31765 }, { "epoch": 2.4894984326018808, "grad_norm": 3.428673028945923, "learning_rate": 3.1355474835153317e-06, "loss": 1.5559, "step": 31766 }, { "epoch": 2.489576802507837, "grad_norm": 3.095618963241577, "learning_rate": 3.1352613655068885e-06, "loss": 1.4156, "step": 31767 }, { "epoch": 2.489655172413793, "grad_norm": 7.930776119232178, "learning_rate": 3.134975254590954e-06, "loss": 0.9839, "step": 31768 }, { "epoch": 2.4897335423197493, "grad_norm": 2.6362648010253906, "learning_rate": 3.134689150768615e-06, "loss": 0.9036, "step": 31769 }, { "epoch": 2.489811912225705, "grad_norm": 2.9591522216796875, "learning_rate": 3.13440305404096e-06, "loss": 1.958, "step": 31770 }, { "epoch": 2.4898902821316615, "grad_norm": 7.705245018005371, "learning_rate": 3.1341169644090774e-06, "loss": 1.4494, "step": 31771 }, { "epoch": 2.4899686520376174, "grad_norm": 6.171753406524658, "learning_rate": 3.1338308818740576e-06, "loss": 1.6123, "step": 31772 }, { "epoch": 2.4900470219435737, "grad_norm": 2.9258978366851807, "learning_rate": 3.133544806436985e-06, "loss": 1.0541, "step": 31773 }, { "epoch": 2.4901253918495296, "grad_norm": 6.014229774475098, "learning_rate": 3.1332587380989493e-06, "loss": 1.0137, "step": 31774 }, { "epoch": 2.490203761755486, "grad_norm": 2.831913948059082, "learning_rate": 3.132972676861038e-06, "loss": 1.3922, "step": 31775 }, { "epoch": 2.490282131661442, "grad_norm": 4.239953994750977, "learning_rate": 3.1326866227243424e-06, "loss": 0.849, "step": 31776 }, { "epoch": 2.490360501567398, "grad_norm": 2.29919171333313, "learning_rate": 3.1324005756899447e-06, "loss": 1.4353, "step": 31777 }, { "epoch": 2.490438871473354, "grad_norm": 4.343081951141357, "learning_rate": 3.132114535758937e-06, "loss": 1.5093, "step": 31778 }, { "epoch": 2.4905172413793104, "grad_norm": 3.704648494720459, "learning_rate": 3.1318285029324067e-06, "loss": 1.494, "step": 31779 }, { "epoch": 2.4905956112852663, "grad_norm": 3.006614923477173, "learning_rate": 3.1315424772114404e-06, "loss": 1.3314, "step": 31780 }, { "epoch": 2.4906739811912226, "grad_norm": 11.720291137695312, "learning_rate": 3.131256458597127e-06, "loss": 1.0935, "step": 31781 }, { "epoch": 2.4907523510971785, "grad_norm": 3.434579849243164, "learning_rate": 3.130970447090553e-06, "loss": 1.1754, "step": 31782 }, { "epoch": 2.490830721003135, "grad_norm": 6.402005672454834, "learning_rate": 3.130684442692811e-06, "loss": 1.2358, "step": 31783 }, { "epoch": 2.4909090909090907, "grad_norm": 3.2266738414764404, "learning_rate": 3.130398445404982e-06, "loss": 1.2276, "step": 31784 }, { "epoch": 2.490987460815047, "grad_norm": 5.363894462585449, "learning_rate": 3.130112455228158e-06, "loss": 1.2053, "step": 31785 }, { "epoch": 2.491065830721003, "grad_norm": 2.607727527618408, "learning_rate": 3.1298264721634257e-06, "loss": 1.3635, "step": 31786 }, { "epoch": 2.4911442006269593, "grad_norm": 3.0492634773254395, "learning_rate": 3.1295404962118736e-06, "loss": 1.015, "step": 31787 }, { "epoch": 2.491222570532915, "grad_norm": 3.160935163497925, "learning_rate": 3.129254527374588e-06, "loss": 1.5546, "step": 31788 }, { "epoch": 2.4913009404388715, "grad_norm": 3.160526990890503, "learning_rate": 3.128968565652657e-06, "loss": 1.431, "step": 31789 }, { "epoch": 2.4913793103448274, "grad_norm": 2.7205162048339844, "learning_rate": 3.128682611047169e-06, "loss": 1.5558, "step": 31790 }, { "epoch": 2.4914576802507837, "grad_norm": 21.355146408081055, "learning_rate": 3.1283966635592123e-06, "loss": 0.8314, "step": 31791 }, { "epoch": 2.4915360501567396, "grad_norm": 2.7419238090515137, "learning_rate": 3.1281107231898715e-06, "loss": 1.2533, "step": 31792 }, { "epoch": 2.491614420062696, "grad_norm": 2.128567695617676, "learning_rate": 3.127824789940237e-06, "loss": 1.4152, "step": 31793 }, { "epoch": 2.491692789968652, "grad_norm": 4.488375186920166, "learning_rate": 3.127538863811396e-06, "loss": 1.6569, "step": 31794 }, { "epoch": 2.491771159874608, "grad_norm": 3.45914363861084, "learning_rate": 3.127252944804434e-06, "loss": 1.39, "step": 31795 }, { "epoch": 2.4918495297805645, "grad_norm": 3.87392520904541, "learning_rate": 3.12696703292044e-06, "loss": 1.3835, "step": 31796 }, { "epoch": 2.4919278996865204, "grad_norm": 2.572490930557251, "learning_rate": 3.1266811281605015e-06, "loss": 1.4154, "step": 31797 }, { "epoch": 2.4920062695924763, "grad_norm": 3.8581700325012207, "learning_rate": 3.1263952305257073e-06, "loss": 1.4975, "step": 31798 }, { "epoch": 2.4920846394984326, "grad_norm": 3.1928553581237793, "learning_rate": 3.1261093400171423e-06, "loss": 0.9267, "step": 31799 }, { "epoch": 2.492163009404389, "grad_norm": 3.3450422286987305, "learning_rate": 3.125823456635895e-06, "loss": 1.9115, "step": 31800 }, { "epoch": 2.492241379310345, "grad_norm": 2.503857374191284, "learning_rate": 3.1255375803830507e-06, "loss": 1.3086, "step": 31801 }, { "epoch": 2.4923197492163007, "grad_norm": 4.099802494049072, "learning_rate": 3.1252517112597027e-06, "loss": 1.3569, "step": 31802 }, { "epoch": 2.492398119122257, "grad_norm": 5.140655994415283, "learning_rate": 3.12496584926693e-06, "loss": 1.4296, "step": 31803 }, { "epoch": 2.4924764890282134, "grad_norm": 7.537861347198486, "learning_rate": 3.1246799944058257e-06, "loss": 0.6498, "step": 31804 }, { "epoch": 2.4925548589341693, "grad_norm": 3.536405563354492, "learning_rate": 3.1243941466774775e-06, "loss": 1.6768, "step": 31805 }, { "epoch": 2.492633228840125, "grad_norm": 3.2911181449890137, "learning_rate": 3.1241083060829687e-06, "loss": 1.5071, "step": 31806 }, { "epoch": 2.4927115987460815, "grad_norm": 3.7018656730651855, "learning_rate": 3.1238224726233877e-06, "loss": 1.0521, "step": 31807 }, { "epoch": 2.492789968652038, "grad_norm": 4.335754871368408, "learning_rate": 3.123536646299823e-06, "loss": 1.4209, "step": 31808 }, { "epoch": 2.4928683385579937, "grad_norm": 2.607872724533081, "learning_rate": 3.1232508271133633e-06, "loss": 1.4452, "step": 31809 }, { "epoch": 2.4929467084639496, "grad_norm": 4.601648330688477, "learning_rate": 3.1229650150650905e-06, "loss": 1.2691, "step": 31810 }, { "epoch": 2.493025078369906, "grad_norm": 2.598369598388672, "learning_rate": 3.1226792101560954e-06, "loss": 0.881, "step": 31811 }, { "epoch": 2.4931034482758623, "grad_norm": 3.1695752143859863, "learning_rate": 3.1223934123874644e-06, "loss": 1.6009, "step": 31812 }, { "epoch": 2.493181818181818, "grad_norm": 2.906548261642456, "learning_rate": 3.122107621760285e-06, "loss": 1.208, "step": 31813 }, { "epoch": 2.493260188087774, "grad_norm": 3.6717441082000732, "learning_rate": 3.1218218382756437e-06, "loss": 1.0025, "step": 31814 }, { "epoch": 2.4933385579937304, "grad_norm": 2.391773223876953, "learning_rate": 3.121536061934626e-06, "loss": 1.2413, "step": 31815 }, { "epoch": 2.4934169278996867, "grad_norm": 2.7998223304748535, "learning_rate": 3.121250292738321e-06, "loss": 1.244, "step": 31816 }, { "epoch": 2.4934952978056426, "grad_norm": 2.5253570079803467, "learning_rate": 3.1209645306878163e-06, "loss": 1.1854, "step": 31817 }, { "epoch": 2.493573667711599, "grad_norm": 6.031678199768066, "learning_rate": 3.120678775784196e-06, "loss": 1.4592, "step": 31818 }, { "epoch": 2.493652037617555, "grad_norm": 2.4596593379974365, "learning_rate": 3.1203930280285487e-06, "loss": 1.2056, "step": 31819 }, { "epoch": 2.493730407523511, "grad_norm": 12.124495506286621, "learning_rate": 3.1201072874219616e-06, "loss": 1.7988, "step": 31820 }, { "epoch": 2.493808777429467, "grad_norm": 3.4408299922943115, "learning_rate": 3.11982155396552e-06, "loss": 1.261, "step": 31821 }, { "epoch": 2.4938871473354234, "grad_norm": 4.034191131591797, "learning_rate": 3.1195358276603105e-06, "loss": 0.8572, "step": 31822 }, { "epoch": 2.4939655172413793, "grad_norm": 2.658090353012085, "learning_rate": 3.119250108507422e-06, "loss": 1.1754, "step": 31823 }, { "epoch": 2.4940438871473356, "grad_norm": 2.9766921997070312, "learning_rate": 3.118964396507941e-06, "loss": 1.1287, "step": 31824 }, { "epoch": 2.4941222570532915, "grad_norm": 6.276760101318359, "learning_rate": 3.1186786916629517e-06, "loss": 1.1792, "step": 31825 }, { "epoch": 2.494200626959248, "grad_norm": 4.116575241088867, "learning_rate": 3.1183929939735436e-06, "loss": 1.9064, "step": 31826 }, { "epoch": 2.4942789968652037, "grad_norm": 2.63916015625, "learning_rate": 3.1181073034408e-06, "loss": 1.233, "step": 31827 }, { "epoch": 2.49435736677116, "grad_norm": 6.641112327575684, "learning_rate": 3.1178216200658116e-06, "loss": 1.6302, "step": 31828 }, { "epoch": 2.494435736677116, "grad_norm": 2.2906980514526367, "learning_rate": 3.117535943849662e-06, "loss": 1.2143, "step": 31829 }, { "epoch": 2.4945141065830723, "grad_norm": 2.9974899291992188, "learning_rate": 3.1172502747934395e-06, "loss": 1.2638, "step": 31830 }, { "epoch": 2.494592476489028, "grad_norm": 1.9842392206192017, "learning_rate": 3.11696461289823e-06, "loss": 1.4281, "step": 31831 }, { "epoch": 2.4946708463949845, "grad_norm": 3.2037057876586914, "learning_rate": 3.11667895816512e-06, "loss": 1.7615, "step": 31832 }, { "epoch": 2.4947492163009404, "grad_norm": 2.9356541633605957, "learning_rate": 3.1163933105951957e-06, "loss": 0.9257, "step": 31833 }, { "epoch": 2.4948275862068967, "grad_norm": 2.774976968765259, "learning_rate": 3.116107670189542e-06, "loss": 1.2787, "step": 31834 }, { "epoch": 2.4949059561128526, "grad_norm": 3.3972418308258057, "learning_rate": 3.1158220369492497e-06, "loss": 0.7947, "step": 31835 }, { "epoch": 2.494984326018809, "grad_norm": 3.566380023956299, "learning_rate": 3.1155364108754016e-06, "loss": 1.3311, "step": 31836 }, { "epoch": 2.495062695924765, "grad_norm": 4.851081371307373, "learning_rate": 3.1152507919690844e-06, "loss": 1.4259, "step": 31837 }, { "epoch": 2.495141065830721, "grad_norm": 3.1562600135803223, "learning_rate": 3.1149651802313863e-06, "loss": 1.6119, "step": 31838 }, { "epoch": 2.495219435736677, "grad_norm": 10.108213424682617, "learning_rate": 3.1146795756633925e-06, "loss": 1.3737, "step": 31839 }, { "epoch": 2.4952978056426334, "grad_norm": 6.989110946655273, "learning_rate": 3.1143939782661875e-06, "loss": 1.3702, "step": 31840 }, { "epoch": 2.4953761755485893, "grad_norm": 2.824382781982422, "learning_rate": 3.11410838804086e-06, "loss": 1.2776, "step": 31841 }, { "epoch": 2.4954545454545456, "grad_norm": 10.183913230895996, "learning_rate": 3.113822804988496e-06, "loss": 0.9487, "step": 31842 }, { "epoch": 2.4955329153605015, "grad_norm": 5.577200889587402, "learning_rate": 3.113537229110182e-06, "loss": 1.5116, "step": 31843 }, { "epoch": 2.495611285266458, "grad_norm": 3.9887888431549072, "learning_rate": 3.113251660407003e-06, "loss": 1.5797, "step": 31844 }, { "epoch": 2.4956896551724137, "grad_norm": 13.535346031188965, "learning_rate": 3.1129660988800437e-06, "loss": 0.581, "step": 31845 }, { "epoch": 2.49576802507837, "grad_norm": 2.929976463317871, "learning_rate": 3.112680544530393e-06, "loss": 1.2376, "step": 31846 }, { "epoch": 2.495846394984326, "grad_norm": 8.642398834228516, "learning_rate": 3.112394997359138e-06, "loss": 1.2944, "step": 31847 }, { "epoch": 2.4959247648902823, "grad_norm": 3.9417881965637207, "learning_rate": 3.1121094573673617e-06, "loss": 1.3078, "step": 31848 }, { "epoch": 2.496003134796238, "grad_norm": 2.8380279541015625, "learning_rate": 3.1118239245561512e-06, "loss": 1.2917, "step": 31849 }, { "epoch": 2.4960815047021945, "grad_norm": 3.510017156600952, "learning_rate": 3.1115383989265934e-06, "loss": 1.5066, "step": 31850 }, { "epoch": 2.4961598746081504, "grad_norm": 4.6387739181518555, "learning_rate": 3.1112528804797725e-06, "loss": 1.5278, "step": 31851 }, { "epoch": 2.4962382445141067, "grad_norm": 2.6751978397369385, "learning_rate": 3.1109673692167753e-06, "loss": 0.9255, "step": 31852 }, { "epoch": 2.4963166144200626, "grad_norm": 4.082879543304443, "learning_rate": 3.1106818651386884e-06, "loss": 1.2998, "step": 31853 }, { "epoch": 2.496394984326019, "grad_norm": 2.98199200630188, "learning_rate": 3.1103963682465982e-06, "loss": 1.5633, "step": 31854 }, { "epoch": 2.496473354231975, "grad_norm": 3.670053482055664, "learning_rate": 3.110110878541589e-06, "loss": 1.1061, "step": 31855 }, { "epoch": 2.496551724137931, "grad_norm": 3.8717918395996094, "learning_rate": 3.109825396024747e-06, "loss": 1.0747, "step": 31856 }, { "epoch": 2.496630094043887, "grad_norm": 2.4406516551971436, "learning_rate": 3.1095399206971573e-06, "loss": 1.2332, "step": 31857 }, { "epoch": 2.4967084639498434, "grad_norm": 3.7897422313690186, "learning_rate": 3.10925445255991e-06, "loss": 1.4465, "step": 31858 }, { "epoch": 2.4967868338557992, "grad_norm": 3.0419998168945312, "learning_rate": 3.108968991614084e-06, "loss": 1.3823, "step": 31859 }, { "epoch": 2.4968652037617556, "grad_norm": 3.202526569366455, "learning_rate": 3.1086835378607704e-06, "loss": 1.1785, "step": 31860 }, { "epoch": 2.4969435736677115, "grad_norm": 4.453588485717773, "learning_rate": 3.1083980913010534e-06, "loss": 1.4304, "step": 31861 }, { "epoch": 2.497021943573668, "grad_norm": 2.734890937805176, "learning_rate": 3.108112651936018e-06, "loss": 1.209, "step": 31862 }, { "epoch": 2.4971003134796237, "grad_norm": 3.4730727672576904, "learning_rate": 3.107827219766751e-06, "loss": 1.896, "step": 31863 }, { "epoch": 2.49717868338558, "grad_norm": 2.362154006958008, "learning_rate": 3.1075417947943353e-06, "loss": 1.3494, "step": 31864 }, { "epoch": 2.497257053291536, "grad_norm": 2.6752946376800537, "learning_rate": 3.107256377019863e-06, "loss": 1.5068, "step": 31865 }, { "epoch": 2.4973354231974922, "grad_norm": 14.906418800354004, "learning_rate": 3.1069709664444102e-06, "loss": 1.7165, "step": 31866 }, { "epoch": 2.497413793103448, "grad_norm": 2.794231653213501, "learning_rate": 3.10668556306907e-06, "loss": 1.0853, "step": 31867 }, { "epoch": 2.4974921630094045, "grad_norm": 2.657111644744873, "learning_rate": 3.106400166894925e-06, "loss": 0.9664, "step": 31868 }, { "epoch": 2.4975705329153604, "grad_norm": 2.9306342601776123, "learning_rate": 3.106114777923062e-06, "loss": 1.3659, "step": 31869 }, { "epoch": 2.4976489028213167, "grad_norm": 3.4485297203063965, "learning_rate": 3.1058293961545648e-06, "loss": 1.4841, "step": 31870 }, { "epoch": 2.4977272727272726, "grad_norm": 7.819797515869141, "learning_rate": 3.1055440215905184e-06, "loss": 1.5836, "step": 31871 }, { "epoch": 2.497805642633229, "grad_norm": 6.50588321685791, "learning_rate": 3.105258654232011e-06, "loss": 1.419, "step": 31872 }, { "epoch": 2.497884012539185, "grad_norm": 1.8972727060317993, "learning_rate": 3.1049732940801274e-06, "loss": 1.2263, "step": 31873 }, { "epoch": 2.497962382445141, "grad_norm": 4.711094379425049, "learning_rate": 3.1046879411359508e-06, "loss": 1.5849, "step": 31874 }, { "epoch": 2.498040752351097, "grad_norm": 3.0005478858947754, "learning_rate": 3.104402595400568e-06, "loss": 1.2409, "step": 31875 }, { "epoch": 2.4981191222570533, "grad_norm": 3.173430919647217, "learning_rate": 3.104117256875065e-06, "loss": 0.9921, "step": 31876 }, { "epoch": 2.4981974921630092, "grad_norm": 3.8739542961120605, "learning_rate": 3.1038319255605252e-06, "loss": 1.5561, "step": 31877 }, { "epoch": 2.4982758620689656, "grad_norm": 2.3130478858947754, "learning_rate": 3.103546601458034e-06, "loss": 1.8357, "step": 31878 }, { "epoch": 2.4983542319749215, "grad_norm": 2.656404733657837, "learning_rate": 3.1032612845686778e-06, "loss": 1.2052, "step": 31879 }, { "epoch": 2.498432601880878, "grad_norm": 3.6066572666168213, "learning_rate": 3.1029759748935427e-06, "loss": 1.2652, "step": 31880 }, { "epoch": 2.4985109717868337, "grad_norm": 6.477449417114258, "learning_rate": 3.1026906724337118e-06, "loss": 1.3885, "step": 31881 }, { "epoch": 2.49858934169279, "grad_norm": 2.8825185298919678, "learning_rate": 3.102405377190271e-06, "loss": 1.5484, "step": 31882 }, { "epoch": 2.498667711598746, "grad_norm": 3.9547536373138428, "learning_rate": 3.1021200891643043e-06, "loss": 1.4919, "step": 31883 }, { "epoch": 2.4987460815047022, "grad_norm": 5.961122989654541, "learning_rate": 3.1018348083569017e-06, "loss": 0.9697, "step": 31884 }, { "epoch": 2.498824451410658, "grad_norm": 2.509578227996826, "learning_rate": 3.1015495347691404e-06, "loss": 1.374, "step": 31885 }, { "epoch": 2.4989028213166145, "grad_norm": 2.9127330780029297, "learning_rate": 3.1012642684021118e-06, "loss": 1.6155, "step": 31886 }, { "epoch": 2.4989811912225703, "grad_norm": 3.2059946060180664, "learning_rate": 3.100979009256898e-06, "loss": 1.2593, "step": 31887 }, { "epoch": 2.4990595611285267, "grad_norm": 8.25356674194336, "learning_rate": 3.100693757334585e-06, "loss": 0.8198, "step": 31888 }, { "epoch": 2.4991379310344826, "grad_norm": 2.7600173950195312, "learning_rate": 3.1004085126362576e-06, "loss": 1.5669, "step": 31889 }, { "epoch": 2.499216300940439, "grad_norm": 3.933959722518921, "learning_rate": 3.100123275162999e-06, "loss": 1.5076, "step": 31890 }, { "epoch": 2.4992946708463952, "grad_norm": 3.7330329418182373, "learning_rate": 3.0998380449158983e-06, "loss": 1.4287, "step": 31891 }, { "epoch": 2.499373040752351, "grad_norm": 3.158403158187866, "learning_rate": 3.0995528218960353e-06, "loss": 1.6067, "step": 31892 }, { "epoch": 2.499451410658307, "grad_norm": 7.9822998046875, "learning_rate": 3.0992676061044977e-06, "loss": 1.7744, "step": 31893 }, { "epoch": 2.4995297805642633, "grad_norm": 4.122313499450684, "learning_rate": 3.0989823975423703e-06, "loss": 1.3381, "step": 31894 }, { "epoch": 2.4996081504702197, "grad_norm": 4.300368785858154, "learning_rate": 3.098697196210738e-06, "loss": 1.2557, "step": 31895 }, { "epoch": 2.4996865203761756, "grad_norm": 2.5587308406829834, "learning_rate": 3.0984120021106843e-06, "loss": 1.1743, "step": 31896 }, { "epoch": 2.4997648902821314, "grad_norm": 2.1451239585876465, "learning_rate": 3.098126815243294e-06, "loss": 1.2923, "step": 31897 }, { "epoch": 2.499843260188088, "grad_norm": 2.647369146347046, "learning_rate": 3.097841635609653e-06, "loss": 1.4359, "step": 31898 }, { "epoch": 2.499921630094044, "grad_norm": 3.7281334400177, "learning_rate": 3.0975564632108463e-06, "loss": 1.6294, "step": 31899 }, { "epoch": 2.5, "grad_norm": 5.268177032470703, "learning_rate": 3.0972712980479567e-06, "loss": 1.6589, "step": 31900 }, { "epoch": 2.5, "eval_loss": 1.2570867538452148, "eval_runtime": 110.4632, "eval_samples_per_second": 24.325, "eval_steps_per_second": 6.083, "step": 31900 }, { "epoch": 2.500078369905956, "grad_norm": 2.2085771560668945, "learning_rate": 3.0969861401220696e-06, "loss": 1.0264, "step": 31901 }, { "epoch": 2.500156739811912, "grad_norm": 3.4095802307128906, "learning_rate": 3.096700989434269e-06, "loss": 1.1624, "step": 31902 }, { "epoch": 2.5002351097178686, "grad_norm": 4.266517639160156, "learning_rate": 3.096415845985643e-06, "loss": 1.5761, "step": 31903 }, { "epoch": 2.5003134796238244, "grad_norm": 3.1702957153320312, "learning_rate": 3.0961307097772707e-06, "loss": 1.0696, "step": 31904 }, { "epoch": 2.5003918495297803, "grad_norm": 2.264219045639038, "learning_rate": 3.0958455808102406e-06, "loss": 1.2868, "step": 31905 }, { "epoch": 2.5004702194357367, "grad_norm": 3.1655240058898926, "learning_rate": 3.0955604590856366e-06, "loss": 1.6252, "step": 31906 }, { "epoch": 2.500548589341693, "grad_norm": 4.208052635192871, "learning_rate": 3.0952753446045416e-06, "loss": 1.5624, "step": 31907 }, { "epoch": 2.500626959247649, "grad_norm": 3.046097993850708, "learning_rate": 3.09499023736804e-06, "loss": 1.4932, "step": 31908 }, { "epoch": 2.5007053291536048, "grad_norm": 2.9347286224365234, "learning_rate": 3.094705137377217e-06, "loss": 1.5233, "step": 31909 }, { "epoch": 2.500783699059561, "grad_norm": 7.938658714294434, "learning_rate": 3.0944200446331595e-06, "loss": 0.4524, "step": 31910 }, { "epoch": 2.5008620689655174, "grad_norm": 3.297083854675293, "learning_rate": 3.094134959136946e-06, "loss": 1.334, "step": 31911 }, { "epoch": 2.5009404388714733, "grad_norm": 4.58515739440918, "learning_rate": 3.0938498808896667e-06, "loss": 1.1166, "step": 31912 }, { "epoch": 2.501018808777429, "grad_norm": 3.485346555709839, "learning_rate": 3.093564809892402e-06, "loss": 1.006, "step": 31913 }, { "epoch": 2.5010971786833855, "grad_norm": 3.756159782409668, "learning_rate": 3.0932797461462395e-06, "loss": 1.1152, "step": 31914 }, { "epoch": 2.501175548589342, "grad_norm": 4.611344814300537, "learning_rate": 3.09299468965226e-06, "loss": 1.117, "step": 31915 }, { "epoch": 2.5012539184952978, "grad_norm": 4.306380748748779, "learning_rate": 3.0927096404115473e-06, "loss": 0.9638, "step": 31916 }, { "epoch": 2.5013322884012537, "grad_norm": 2.6934430599212646, "learning_rate": 3.0924245984251922e-06, "loss": 1.6335, "step": 31917 }, { "epoch": 2.50141065830721, "grad_norm": 2.545588731765747, "learning_rate": 3.0921395636942698e-06, "loss": 1.4762, "step": 31918 }, { "epoch": 2.5014890282131663, "grad_norm": 3.609537124633789, "learning_rate": 3.0918545362198705e-06, "loss": 1.2863, "step": 31919 }, { "epoch": 2.501567398119122, "grad_norm": 3.073751926422119, "learning_rate": 3.0915695160030757e-06, "loss": 1.6609, "step": 31920 }, { "epoch": 2.501645768025078, "grad_norm": 2.7299649715423584, "learning_rate": 3.0912845030449717e-06, "loss": 1.3288, "step": 31921 }, { "epoch": 2.5017241379310344, "grad_norm": 2.503286361694336, "learning_rate": 3.09099949734664e-06, "loss": 1.5759, "step": 31922 }, { "epoch": 2.5018025078369908, "grad_norm": 9.034822463989258, "learning_rate": 3.0907144989091647e-06, "loss": 1.2258, "step": 31923 }, { "epoch": 2.5018808777429467, "grad_norm": 4.674158573150635, "learning_rate": 3.0904295077336314e-06, "loss": 1.0149, "step": 31924 }, { "epoch": 2.501959247648903, "grad_norm": 2.379876136779785, "learning_rate": 3.090144523821125e-06, "loss": 1.4717, "step": 31925 }, { "epoch": 2.502037617554859, "grad_norm": 3.5218191146850586, "learning_rate": 3.0898595471727267e-06, "loss": 1.3568, "step": 31926 }, { "epoch": 2.502115987460815, "grad_norm": 2.585407257080078, "learning_rate": 3.0895745777895216e-06, "loss": 1.4502, "step": 31927 }, { "epoch": 2.502194357366771, "grad_norm": 2.605807304382324, "learning_rate": 3.089289615672592e-06, "loss": 1.3303, "step": 31928 }, { "epoch": 2.5022727272727274, "grad_norm": 4.593047142028809, "learning_rate": 3.089004660823027e-06, "loss": 0.8848, "step": 31929 }, { "epoch": 2.5023510971786833, "grad_norm": 2.098740339279175, "learning_rate": 3.0887197132419033e-06, "loss": 1.2312, "step": 31930 }, { "epoch": 2.5024294670846396, "grad_norm": 4.021085262298584, "learning_rate": 3.0884347729303093e-06, "loss": 1.1657, "step": 31931 }, { "epoch": 2.5025078369905955, "grad_norm": 3.2875847816467285, "learning_rate": 3.0881498398893285e-06, "loss": 1.1873, "step": 31932 }, { "epoch": 2.502586206896552, "grad_norm": 12.623003959655762, "learning_rate": 3.087864914120043e-06, "loss": 1.2278, "step": 31933 }, { "epoch": 2.5026645768025078, "grad_norm": 3.520390033721924, "learning_rate": 3.087579995623537e-06, "loss": 1.2466, "step": 31934 }, { "epoch": 2.502742946708464, "grad_norm": 4.508629322052002, "learning_rate": 3.087295084400893e-06, "loss": 0.492, "step": 31935 }, { "epoch": 2.50282131661442, "grad_norm": 5.753121376037598, "learning_rate": 3.087010180453199e-06, "loss": 0.9392, "step": 31936 }, { "epoch": 2.5028996865203763, "grad_norm": 3.905776023864746, "learning_rate": 3.0867252837815344e-06, "loss": 1.6052, "step": 31937 }, { "epoch": 2.502978056426332, "grad_norm": 4.539946556091309, "learning_rate": 3.0864403943869835e-06, "loss": 1.7585, "step": 31938 }, { "epoch": 2.5030564263322885, "grad_norm": 3.028829574584961, "learning_rate": 3.086155512270631e-06, "loss": 1.4909, "step": 31939 }, { "epoch": 2.5031347962382444, "grad_norm": 2.5070393085479736, "learning_rate": 3.0858706374335607e-06, "loss": 1.3301, "step": 31940 }, { "epoch": 2.5032131661442008, "grad_norm": 2.251376152038574, "learning_rate": 3.0855857698768533e-06, "loss": 1.1734, "step": 31941 }, { "epoch": 2.5032915360501566, "grad_norm": 4.766018390655518, "learning_rate": 3.085300909601595e-06, "loss": 1.6168, "step": 31942 }, { "epoch": 2.503369905956113, "grad_norm": 4.852425575256348, "learning_rate": 3.0850160566088692e-06, "loss": 2.0238, "step": 31943 }, { "epoch": 2.503448275862069, "grad_norm": 4.019242763519287, "learning_rate": 3.0847312108997597e-06, "loss": 1.2806, "step": 31944 }, { "epoch": 2.503526645768025, "grad_norm": 7.806796073913574, "learning_rate": 3.0844463724753483e-06, "loss": 1.2881, "step": 31945 }, { "epoch": 2.503605015673981, "grad_norm": 2.835649251937866, "learning_rate": 3.084161541336717e-06, "loss": 1.2068, "step": 31946 }, { "epoch": 2.5036833855799374, "grad_norm": 3.4317808151245117, "learning_rate": 3.0838767174849547e-06, "loss": 1.4164, "step": 31947 }, { "epoch": 2.5037617554858933, "grad_norm": 2.6047799587249756, "learning_rate": 3.083591900921138e-06, "loss": 1.1213, "step": 31948 }, { "epoch": 2.5038401253918496, "grad_norm": 2.6037073135375977, "learning_rate": 3.083307091646355e-06, "loss": 1.4126, "step": 31949 }, { "epoch": 2.5039184952978055, "grad_norm": 3.056880474090576, "learning_rate": 3.0830222896616867e-06, "loss": 1.713, "step": 31950 }, { "epoch": 2.503996865203762, "grad_norm": 3.5632593631744385, "learning_rate": 3.0827374949682184e-06, "loss": 1.4765, "step": 31951 }, { "epoch": 2.5040752351097177, "grad_norm": 2.8461368083953857, "learning_rate": 3.0824527075670307e-06, "loss": 1.051, "step": 31952 }, { "epoch": 2.504153605015674, "grad_norm": 4.490706920623779, "learning_rate": 3.0821679274592063e-06, "loss": 1.7032, "step": 31953 }, { "epoch": 2.50423197492163, "grad_norm": 3.0468223094940186, "learning_rate": 3.0818831546458315e-06, "loss": 1.4828, "step": 31954 }, { "epoch": 2.5043103448275863, "grad_norm": 5.167273044586182, "learning_rate": 3.08159838912799e-06, "loss": 1.1608, "step": 31955 }, { "epoch": 2.504388714733542, "grad_norm": 12.81692886352539, "learning_rate": 3.0813136309067604e-06, "loss": 0.848, "step": 31956 }, { "epoch": 2.5044670846394985, "grad_norm": 3.8800814151763916, "learning_rate": 3.0810288799832287e-06, "loss": 1.1124, "step": 31957 }, { "epoch": 2.5045454545454544, "grad_norm": 5.21614933013916, "learning_rate": 3.080744136358477e-06, "loss": 0.8467, "step": 31958 }, { "epoch": 2.5046238244514107, "grad_norm": 2.597201108932495, "learning_rate": 3.0804594000335917e-06, "loss": 1.4808, "step": 31959 }, { "epoch": 2.5047021943573666, "grad_norm": 2.3538730144500732, "learning_rate": 3.0801746710096497e-06, "loss": 1.456, "step": 31960 }, { "epoch": 2.504780564263323, "grad_norm": 5.010619640350342, "learning_rate": 3.079889949287738e-06, "loss": 0.9015, "step": 31961 }, { "epoch": 2.504858934169279, "grad_norm": 5.051117420196533, "learning_rate": 3.0796052348689396e-06, "loss": 0.548, "step": 31962 }, { "epoch": 2.504937304075235, "grad_norm": 1.8011664152145386, "learning_rate": 3.0793205277543357e-06, "loss": 1.3756, "step": 31963 }, { "epoch": 2.505015673981191, "grad_norm": 3.6797780990600586, "learning_rate": 3.0790358279450106e-06, "loss": 1.1387, "step": 31964 }, { "epoch": 2.5050940438871474, "grad_norm": 2.605440139770508, "learning_rate": 3.0787511354420453e-06, "loss": 1.5978, "step": 31965 }, { "epoch": 2.5051724137931033, "grad_norm": 4.858767032623291, "learning_rate": 3.0784664502465263e-06, "loss": 1.3825, "step": 31966 }, { "epoch": 2.5052507836990596, "grad_norm": 3.0319933891296387, "learning_rate": 3.0781817723595307e-06, "loss": 1.4386, "step": 31967 }, { "epoch": 2.5053291536050155, "grad_norm": 3.400710105895996, "learning_rate": 3.077897101782146e-06, "loss": 1.355, "step": 31968 }, { "epoch": 2.505407523510972, "grad_norm": 3.438699722290039, "learning_rate": 3.077612438515454e-06, "loss": 1.1592, "step": 31969 }, { "epoch": 2.5054858934169277, "grad_norm": 4.4034271240234375, "learning_rate": 3.0773277825605375e-06, "loss": 1.08, "step": 31970 }, { "epoch": 2.505564263322884, "grad_norm": 6.858936309814453, "learning_rate": 3.077043133918477e-06, "loss": 0.9562, "step": 31971 }, { "epoch": 2.50564263322884, "grad_norm": 2.9925122261047363, "learning_rate": 3.0767584925903567e-06, "loss": 1.55, "step": 31972 }, { "epoch": 2.5057210031347963, "grad_norm": 4.634225845336914, "learning_rate": 3.0764738585772614e-06, "loss": 2.1459, "step": 31973 }, { "epoch": 2.5057993730407526, "grad_norm": 2.2646875381469727, "learning_rate": 3.0761892318802687e-06, "loss": 1.0843, "step": 31974 }, { "epoch": 2.5058777429467085, "grad_norm": 2.8966915607452393, "learning_rate": 3.0759046125004655e-06, "loss": 1.0903, "step": 31975 }, { "epoch": 2.5059561128526644, "grad_norm": 3.5683929920196533, "learning_rate": 3.0756200004389324e-06, "loss": 1.0679, "step": 31976 }, { "epoch": 2.5060344827586207, "grad_norm": 2.642573595046997, "learning_rate": 3.0753353956967542e-06, "loss": 1.1904, "step": 31977 }, { "epoch": 2.506112852664577, "grad_norm": 2.634219169616699, "learning_rate": 3.07505079827501e-06, "loss": 1.0889, "step": 31978 }, { "epoch": 2.506191222570533, "grad_norm": 3.7896008491516113, "learning_rate": 3.0747662081747827e-06, "loss": 1.2897, "step": 31979 }, { "epoch": 2.506269592476489, "grad_norm": 6.058268070220947, "learning_rate": 3.0744816253971565e-06, "loss": 1.6406, "step": 31980 }, { "epoch": 2.506347962382445, "grad_norm": 2.6567530632019043, "learning_rate": 3.074197049943215e-06, "loss": 1.1466, "step": 31981 }, { "epoch": 2.5064263322884015, "grad_norm": 3.69223690032959, "learning_rate": 3.0739124818140374e-06, "loss": 1.047, "step": 31982 }, { "epoch": 2.5065047021943574, "grad_norm": 3.417005777359009, "learning_rate": 3.073627921010708e-06, "loss": 1.1981, "step": 31983 }, { "epoch": 2.5065830721003133, "grad_norm": 2.7568721771240234, "learning_rate": 3.0733433675343073e-06, "loss": 0.9289, "step": 31984 }, { "epoch": 2.5066614420062696, "grad_norm": 4.396361827850342, "learning_rate": 3.073058821385921e-06, "loss": 1.1317, "step": 31985 }, { "epoch": 2.506739811912226, "grad_norm": 4.030559062957764, "learning_rate": 3.0727742825666266e-06, "loss": 1.0216, "step": 31986 }, { "epoch": 2.506818181818182, "grad_norm": 2.1007890701293945, "learning_rate": 3.0724897510775103e-06, "loss": 1.5288, "step": 31987 }, { "epoch": 2.5068965517241377, "grad_norm": 3.632629871368408, "learning_rate": 3.072205226919654e-06, "loss": 1.4194, "step": 31988 }, { "epoch": 2.506974921630094, "grad_norm": 3.231649875640869, "learning_rate": 3.0719207100941373e-06, "loss": 1.74, "step": 31989 }, { "epoch": 2.5070532915360504, "grad_norm": 2.6493654251098633, "learning_rate": 3.0716362006020443e-06, "loss": 1.218, "step": 31990 }, { "epoch": 2.5071316614420063, "grad_norm": 4.262214183807373, "learning_rate": 3.071351698444456e-06, "loss": 1.6606, "step": 31991 }, { "epoch": 2.507210031347962, "grad_norm": 3.2409515380859375, "learning_rate": 3.0710672036224577e-06, "loss": 1.9705, "step": 31992 }, { "epoch": 2.5072884012539185, "grad_norm": 3.880039930343628, "learning_rate": 3.070782716137126e-06, "loss": 1.7043, "step": 31993 }, { "epoch": 2.507366771159875, "grad_norm": 3.511874198913574, "learning_rate": 3.0704982359895476e-06, "loss": 1.3408, "step": 31994 }, { "epoch": 2.5074451410658307, "grad_norm": 3.5343315601348877, "learning_rate": 3.0702137631808016e-06, "loss": 1.9432, "step": 31995 }, { "epoch": 2.5075235109717866, "grad_norm": 5.108608245849609, "learning_rate": 3.0699292977119734e-06, "loss": 1.3487, "step": 31996 }, { "epoch": 2.507601880877743, "grad_norm": 2.33388352394104, "learning_rate": 3.069644839584142e-06, "loss": 0.9308, "step": 31997 }, { "epoch": 2.5076802507836993, "grad_norm": 2.3805603981018066, "learning_rate": 3.069360388798388e-06, "loss": 1.3184, "step": 31998 }, { "epoch": 2.507758620689655, "grad_norm": 5.397745132446289, "learning_rate": 3.0690759453557973e-06, "loss": 1.0657, "step": 31999 }, { "epoch": 2.507836990595611, "grad_norm": 4.0055437088012695, "learning_rate": 3.0687915092574503e-06, "loss": 1.3443, "step": 32000 }, { "epoch": 2.5079153605015674, "grad_norm": 2.225768804550171, "learning_rate": 3.0685070805044274e-06, "loss": 1.3614, "step": 32001 }, { "epoch": 2.5079937304075237, "grad_norm": 2.2342121601104736, "learning_rate": 3.0682226590978122e-06, "loss": 1.3755, "step": 32002 }, { "epoch": 2.5080721003134796, "grad_norm": 2.800093173980713, "learning_rate": 3.067938245038686e-06, "loss": 1.2359, "step": 32003 }, { "epoch": 2.5081504702194355, "grad_norm": 3.098271608352661, "learning_rate": 3.0676538383281303e-06, "loss": 1.4515, "step": 32004 }, { "epoch": 2.508228840125392, "grad_norm": 2.265918016433716, "learning_rate": 3.0673694389672247e-06, "loss": 1.3307, "step": 32005 }, { "epoch": 2.508307210031348, "grad_norm": 2.7807178497314453, "learning_rate": 3.067085046957055e-06, "loss": 1.8133, "step": 32006 }, { "epoch": 2.508385579937304, "grad_norm": 2.543172836303711, "learning_rate": 3.0668006622987024e-06, "loss": 1.4304, "step": 32007 }, { "epoch": 2.50846394984326, "grad_norm": 5.073256969451904, "learning_rate": 3.0665162849932452e-06, "loss": 0.9091, "step": 32008 }, { "epoch": 2.5085423197492163, "grad_norm": 3.154614210128784, "learning_rate": 3.066231915041767e-06, "loss": 1.0246, "step": 32009 }, { "epoch": 2.5086206896551726, "grad_norm": 4.206690311431885, "learning_rate": 3.0659475524453486e-06, "loss": 1.246, "step": 32010 }, { "epoch": 2.5086990595611285, "grad_norm": 2.4249258041381836, "learning_rate": 3.0656631972050755e-06, "loss": 1.0657, "step": 32011 }, { "epoch": 2.5087774294670844, "grad_norm": 6.017784118652344, "learning_rate": 3.0653788493220226e-06, "loss": 1.4556, "step": 32012 }, { "epoch": 2.5088557993730407, "grad_norm": 3.069319725036621, "learning_rate": 3.065094508797276e-06, "loss": 1.4656, "step": 32013 }, { "epoch": 2.508934169278997, "grad_norm": 10.795702934265137, "learning_rate": 3.064810175631916e-06, "loss": 0.6863, "step": 32014 }, { "epoch": 2.509012539184953, "grad_norm": 9.04637336730957, "learning_rate": 3.0645258498270257e-06, "loss": 1.0797, "step": 32015 }, { "epoch": 2.509090909090909, "grad_norm": 7.148153781890869, "learning_rate": 3.0642415313836837e-06, "loss": 1.4395, "step": 32016 }, { "epoch": 2.509169278996865, "grad_norm": 4.115879535675049, "learning_rate": 3.063957220302971e-06, "loss": 1.5448, "step": 32017 }, { "epoch": 2.5092476489028215, "grad_norm": 3.469346046447754, "learning_rate": 3.063672916585975e-06, "loss": 1.4239, "step": 32018 }, { "epoch": 2.5093260188087774, "grad_norm": 3.9138998985290527, "learning_rate": 3.063388620233768e-06, "loss": 1.5911, "step": 32019 }, { "epoch": 2.5094043887147337, "grad_norm": 8.600234985351562, "learning_rate": 3.0631043312474375e-06, "loss": 0.7765, "step": 32020 }, { "epoch": 2.5094827586206896, "grad_norm": 8.867956161499023, "learning_rate": 3.0628200496280634e-06, "loss": 1.3768, "step": 32021 }, { "epoch": 2.509561128526646, "grad_norm": 3.2677409648895264, "learning_rate": 3.0625357753767283e-06, "loss": 1.3341, "step": 32022 }, { "epoch": 2.509639498432602, "grad_norm": 22.321434020996094, "learning_rate": 3.0622515084945103e-06, "loss": 1.6075, "step": 32023 }, { "epoch": 2.509717868338558, "grad_norm": 3.349750518798828, "learning_rate": 3.061967248982492e-06, "loss": 1.1316, "step": 32024 }, { "epoch": 2.509796238244514, "grad_norm": 3.923532724380493, "learning_rate": 3.061682996841755e-06, "loss": 1.4274, "step": 32025 }, { "epoch": 2.5098746081504704, "grad_norm": 2.5307934284210205, "learning_rate": 3.0613987520733824e-06, "loss": 1.2583, "step": 32026 }, { "epoch": 2.5099529780564263, "grad_norm": 4.017087459564209, "learning_rate": 3.0611145146784515e-06, "loss": 1.3658, "step": 32027 }, { "epoch": 2.5100313479623826, "grad_norm": 7.8503899574279785, "learning_rate": 3.0608302846580452e-06, "loss": 1.3117, "step": 32028 }, { "epoch": 2.5101097178683385, "grad_norm": 3.3090097904205322, "learning_rate": 3.0605460620132452e-06, "loss": 1.1953, "step": 32029 }, { "epoch": 2.510188087774295, "grad_norm": 2.7099239826202393, "learning_rate": 3.0602618467451307e-06, "loss": 1.2877, "step": 32030 }, { "epoch": 2.5102664576802507, "grad_norm": 3.1522574424743652, "learning_rate": 3.059977638854783e-06, "loss": 1.3446, "step": 32031 }, { "epoch": 2.510344827586207, "grad_norm": 2.955695390701294, "learning_rate": 3.0596934383432853e-06, "loss": 1.1397, "step": 32032 }, { "epoch": 2.510423197492163, "grad_norm": 25.95029067993164, "learning_rate": 3.059409245211718e-06, "loss": 0.8529, "step": 32033 }, { "epoch": 2.5105015673981192, "grad_norm": 2.829819679260254, "learning_rate": 3.05912505946116e-06, "loss": 1.4305, "step": 32034 }, { "epoch": 2.510579937304075, "grad_norm": 7.375959873199463, "learning_rate": 3.058840881092694e-06, "loss": 1.7055, "step": 32035 }, { "epoch": 2.5106583072100315, "grad_norm": 2.3082220554351807, "learning_rate": 3.058556710107399e-06, "loss": 1.1307, "step": 32036 }, { "epoch": 2.5107366771159874, "grad_norm": 3.043564558029175, "learning_rate": 3.0582725465063577e-06, "loss": 1.2659, "step": 32037 }, { "epoch": 2.5108150470219437, "grad_norm": 5.425886154174805, "learning_rate": 3.057988390290651e-06, "loss": 1.5279, "step": 32038 }, { "epoch": 2.5108934169278996, "grad_norm": 3.01389741897583, "learning_rate": 3.0577042414613585e-06, "loss": 1.4073, "step": 32039 }, { "epoch": 2.510971786833856, "grad_norm": 3.4505131244659424, "learning_rate": 3.057420100019561e-06, "loss": 1.6422, "step": 32040 }, { "epoch": 2.511050156739812, "grad_norm": 2.672382354736328, "learning_rate": 3.057135965966341e-06, "loss": 1.2718, "step": 32041 }, { "epoch": 2.511128526645768, "grad_norm": 3.217780351638794, "learning_rate": 3.0568518393027755e-06, "loss": 0.8086, "step": 32042 }, { "epoch": 2.511206896551724, "grad_norm": 3.179833173751831, "learning_rate": 3.0565677200299493e-06, "loss": 1.0747, "step": 32043 }, { "epoch": 2.5112852664576804, "grad_norm": 4.819578170776367, "learning_rate": 3.0562836081489424e-06, "loss": 1.4417, "step": 32044 }, { "epoch": 2.5113636363636362, "grad_norm": 4.334558486938477, "learning_rate": 3.0559995036608327e-06, "loss": 1.4117, "step": 32045 }, { "epoch": 2.5114420062695926, "grad_norm": 2.9006099700927734, "learning_rate": 3.0557154065667023e-06, "loss": 0.9186, "step": 32046 }, { "epoch": 2.5115203761755485, "grad_norm": 3.028923273086548, "learning_rate": 3.055431316867632e-06, "loss": 1.5447, "step": 32047 }, { "epoch": 2.511598746081505, "grad_norm": 4.318953514099121, "learning_rate": 3.055147234564703e-06, "loss": 1.6587, "step": 32048 }, { "epoch": 2.5116771159874607, "grad_norm": 5.670513153076172, "learning_rate": 3.0548631596589934e-06, "loss": 1.4064, "step": 32049 }, { "epoch": 2.511755485893417, "grad_norm": 3.5169544219970703, "learning_rate": 3.054579092151586e-06, "loss": 1.1832, "step": 32050 }, { "epoch": 2.511833855799373, "grad_norm": 4.872060298919678, "learning_rate": 3.0542950320435607e-06, "loss": 1.486, "step": 32051 }, { "epoch": 2.5119122257053292, "grad_norm": 3.0356853008270264, "learning_rate": 3.0540109793359985e-06, "loss": 1.3512, "step": 32052 }, { "epoch": 2.511990595611285, "grad_norm": 2.5907061100006104, "learning_rate": 3.053726934029978e-06, "loss": 1.4643, "step": 32053 }, { "epoch": 2.5120689655172415, "grad_norm": 3.221708297729492, "learning_rate": 3.0534428961265794e-06, "loss": 1.4176, "step": 32054 }, { "epoch": 2.5121473354231973, "grad_norm": 2.636898994445801, "learning_rate": 3.0531588656268857e-06, "loss": 1.6459, "step": 32055 }, { "epoch": 2.5122257053291537, "grad_norm": 4.6232686042785645, "learning_rate": 3.0528748425319765e-06, "loss": 1.3076, "step": 32056 }, { "epoch": 2.5123040752351096, "grad_norm": 5.934967041015625, "learning_rate": 3.0525908268429305e-06, "loss": 1.4753, "step": 32057 }, { "epoch": 2.512382445141066, "grad_norm": 2.630232572555542, "learning_rate": 3.0523068185608285e-06, "loss": 1.3166, "step": 32058 }, { "epoch": 2.512460815047022, "grad_norm": 3.2292051315307617, "learning_rate": 3.0520228176867517e-06, "loss": 1.3333, "step": 32059 }, { "epoch": 2.512539184952978, "grad_norm": 2.3483927249908447, "learning_rate": 3.051738824221779e-06, "loss": 1.1979, "step": 32060 }, { "epoch": 2.512617554858934, "grad_norm": 3.446657180786133, "learning_rate": 3.0514548381669894e-06, "loss": 0.9419, "step": 32061 }, { "epoch": 2.5126959247648903, "grad_norm": 2.024268627166748, "learning_rate": 3.0511708595234667e-06, "loss": 1.2968, "step": 32062 }, { "epoch": 2.5127742946708462, "grad_norm": 2.5863003730773926, "learning_rate": 3.0508868882922896e-06, "loss": 0.9903, "step": 32063 }, { "epoch": 2.5128526645768026, "grad_norm": 7.421188831329346, "learning_rate": 3.050602924474537e-06, "loss": 0.9411, "step": 32064 }, { "epoch": 2.512931034482759, "grad_norm": 4.534669399261475, "learning_rate": 3.0503189680712893e-06, "loss": 1.2218, "step": 32065 }, { "epoch": 2.513009404388715, "grad_norm": 2.596898317337036, "learning_rate": 3.050035019083626e-06, "loss": 1.014, "step": 32066 }, { "epoch": 2.5130877742946707, "grad_norm": 3.691990613937378, "learning_rate": 3.0497510775126306e-06, "loss": 1.5757, "step": 32067 }, { "epoch": 2.513166144200627, "grad_norm": 4.2386980056762695, "learning_rate": 3.0494671433593776e-06, "loss": 1.2922, "step": 32068 }, { "epoch": 2.5132445141065833, "grad_norm": 3.4850900173187256, "learning_rate": 3.0491832166249503e-06, "loss": 1.5568, "step": 32069 }, { "epoch": 2.5133228840125392, "grad_norm": 2.614818811416626, "learning_rate": 3.0488992973104287e-06, "loss": 1.5523, "step": 32070 }, { "epoch": 2.513401253918495, "grad_norm": 2.5288023948669434, "learning_rate": 3.0486153854168932e-06, "loss": 1.4933, "step": 32071 }, { "epoch": 2.5134796238244514, "grad_norm": 3.5224311351776123, "learning_rate": 3.0483314809454204e-06, "loss": 1.5115, "step": 32072 }, { "epoch": 2.513557993730408, "grad_norm": 3.639694929122925, "learning_rate": 3.0480475838970924e-06, "loss": 1.1137, "step": 32073 }, { "epoch": 2.5136363636363637, "grad_norm": 3.5361621379852295, "learning_rate": 3.047763694272991e-06, "loss": 1.5041, "step": 32074 }, { "epoch": 2.5137147335423196, "grad_norm": 4.969458103179932, "learning_rate": 3.0474798120741913e-06, "loss": 1.6437, "step": 32075 }, { "epoch": 2.513793103448276, "grad_norm": 3.623006820678711, "learning_rate": 3.0471959373017756e-06, "loss": 1.3055, "step": 32076 }, { "epoch": 2.5138714733542322, "grad_norm": 3.65030837059021, "learning_rate": 3.0469120699568246e-06, "loss": 1.3651, "step": 32077 }, { "epoch": 2.513949843260188, "grad_norm": 2.128375768661499, "learning_rate": 3.0466282100404177e-06, "loss": 1.3924, "step": 32078 }, { "epoch": 2.514028213166144, "grad_norm": 12.36979866027832, "learning_rate": 3.0463443575536324e-06, "loss": 1.5319, "step": 32079 }, { "epoch": 2.5141065830721003, "grad_norm": 7.566892623901367, "learning_rate": 3.0460605124975483e-06, "loss": 1.4733, "step": 32080 }, { "epoch": 2.5141849529780567, "grad_norm": 2.172043561935425, "learning_rate": 3.0457766748732486e-06, "loss": 1.6684, "step": 32081 }, { "epoch": 2.5142633228840126, "grad_norm": 2.9737918376922607, "learning_rate": 3.045492844681811e-06, "loss": 1.4422, "step": 32082 }, { "epoch": 2.5143416927899684, "grad_norm": 3.168360710144043, "learning_rate": 3.0452090219243134e-06, "loss": 1.0636, "step": 32083 }, { "epoch": 2.5144200626959248, "grad_norm": 3.1269614696502686, "learning_rate": 3.0449252066018377e-06, "loss": 1.1827, "step": 32084 }, { "epoch": 2.514498432601881, "grad_norm": 2.5147721767425537, "learning_rate": 3.044641398715462e-06, "loss": 1.477, "step": 32085 }, { "epoch": 2.514576802507837, "grad_norm": 2.8846166133880615, "learning_rate": 3.044357598266266e-06, "loss": 1.1508, "step": 32086 }, { "epoch": 2.514655172413793, "grad_norm": 6.711733341217041, "learning_rate": 3.0440738052553274e-06, "loss": 0.9173, "step": 32087 }, { "epoch": 2.514733542319749, "grad_norm": 10.236696243286133, "learning_rate": 3.0437900196837295e-06, "loss": 1.2281, "step": 32088 }, { "epoch": 2.5148119122257055, "grad_norm": 3.785645008087158, "learning_rate": 3.0435062415525495e-06, "loss": 1.3876, "step": 32089 }, { "epoch": 2.5148902821316614, "grad_norm": 2.637315273284912, "learning_rate": 3.0432224708628665e-06, "loss": 1.2476, "step": 32090 }, { "epoch": 2.5149686520376173, "grad_norm": 2.3698558807373047, "learning_rate": 3.04293870761576e-06, "loss": 1.2564, "step": 32091 }, { "epoch": 2.5150470219435737, "grad_norm": 2.011768102645874, "learning_rate": 3.0426549518123087e-06, "loss": 1.16, "step": 32092 }, { "epoch": 2.51512539184953, "grad_norm": 2.5986058712005615, "learning_rate": 3.0423712034535956e-06, "loss": 1.4856, "step": 32093 }, { "epoch": 2.515203761755486, "grad_norm": 4.086793899536133, "learning_rate": 3.042087462540694e-06, "loss": 1.4298, "step": 32094 }, { "epoch": 2.5152821316614418, "grad_norm": 6.551150798797607, "learning_rate": 3.041803729074686e-06, "loss": 1.2836, "step": 32095 }, { "epoch": 2.515360501567398, "grad_norm": 2.0211946964263916, "learning_rate": 3.0415200030566515e-06, "loss": 1.1949, "step": 32096 }, { "epoch": 2.5154388714733544, "grad_norm": 5.778332710266113, "learning_rate": 3.041236284487671e-06, "loss": 1.3542, "step": 32097 }, { "epoch": 2.5155172413793103, "grad_norm": 2.1440205574035645, "learning_rate": 3.0409525733688193e-06, "loss": 1.3303, "step": 32098 }, { "epoch": 2.515595611285266, "grad_norm": 3.186142921447754, "learning_rate": 3.0406688697011773e-06, "loss": 0.8684, "step": 32099 }, { "epoch": 2.5156739811912225, "grad_norm": 3.6483685970306396, "learning_rate": 3.040385173485827e-06, "loss": 1.3163, "step": 32100 }, { "epoch": 2.515752351097179, "grad_norm": 5.033327579498291, "learning_rate": 3.0401014847238424e-06, "loss": 1.5848, "step": 32101 }, { "epoch": 2.5158307210031348, "grad_norm": 2.2612390518188477, "learning_rate": 3.039817803416306e-06, "loss": 1.1448, "step": 32102 }, { "epoch": 2.5159090909090907, "grad_norm": 2.3985049724578857, "learning_rate": 3.039534129564296e-06, "loss": 1.015, "step": 32103 }, { "epoch": 2.515987460815047, "grad_norm": 3.469790458679199, "learning_rate": 3.0392504631688923e-06, "loss": 1.3388, "step": 32104 }, { "epoch": 2.5160658307210033, "grad_norm": 3.001946210861206, "learning_rate": 3.0389668042311717e-06, "loss": 1.2797, "step": 32105 }, { "epoch": 2.516144200626959, "grad_norm": 3.8229353427886963, "learning_rate": 3.038683152752213e-06, "loss": 1.0763, "step": 32106 }, { "epoch": 2.516222570532915, "grad_norm": 3.647932529449463, "learning_rate": 3.0383995087330963e-06, "loss": 1.4398, "step": 32107 }, { "epoch": 2.5163009404388714, "grad_norm": 2.880122661590576, "learning_rate": 3.0381158721749027e-06, "loss": 1.1149, "step": 32108 }, { "epoch": 2.5163793103448278, "grad_norm": 3.535662889480591, "learning_rate": 3.0378322430787073e-06, "loss": 1.2041, "step": 32109 }, { "epoch": 2.5164576802507836, "grad_norm": 3.1991117000579834, "learning_rate": 3.0375486214455895e-06, "loss": 1.4137, "step": 32110 }, { "epoch": 2.5165360501567395, "grad_norm": 3.195904493331909, "learning_rate": 3.0372650072766286e-06, "loss": 1.4621, "step": 32111 }, { "epoch": 2.516614420062696, "grad_norm": 6.440532207489014, "learning_rate": 3.0369814005729057e-06, "loss": 1.2814, "step": 32112 }, { "epoch": 2.516692789968652, "grad_norm": 3.4399194717407227, "learning_rate": 3.0366978013354954e-06, "loss": 1.308, "step": 32113 }, { "epoch": 2.516771159874608, "grad_norm": 2.5365397930145264, "learning_rate": 3.036414209565479e-06, "loss": 1.5031, "step": 32114 }, { "epoch": 2.5168495297805644, "grad_norm": 2.1599783897399902, "learning_rate": 3.0361306252639343e-06, "loss": 1.1809, "step": 32115 }, { "epoch": 2.5169278996865203, "grad_norm": 10.801315307617188, "learning_rate": 3.0358470484319404e-06, "loss": 1.2812, "step": 32116 }, { "epoch": 2.5170062695924766, "grad_norm": 3.4383137226104736, "learning_rate": 3.0355634790705748e-06, "loss": 0.1952, "step": 32117 }, { "epoch": 2.5170846394984325, "grad_norm": 2.5788164138793945, "learning_rate": 3.0352799171809155e-06, "loss": 1.6079, "step": 32118 }, { "epoch": 2.517163009404389, "grad_norm": 2.220360279083252, "learning_rate": 3.0349963627640456e-06, "loss": 1.3968, "step": 32119 }, { "epoch": 2.5172413793103448, "grad_norm": 2.607560396194458, "learning_rate": 3.034712815821037e-06, "loss": 1.2121, "step": 32120 }, { "epoch": 2.517319749216301, "grad_norm": 2.6780149936676025, "learning_rate": 3.0344292763529726e-06, "loss": 1.3594, "step": 32121 }, { "epoch": 2.517398119122257, "grad_norm": 4.132845878601074, "learning_rate": 3.034145744360929e-06, "loss": 1.1144, "step": 32122 }, { "epoch": 2.5174764890282133, "grad_norm": 2.593696117401123, "learning_rate": 3.033862219845987e-06, "loss": 1.7052, "step": 32123 }, { "epoch": 2.517554858934169, "grad_norm": 6.472090721130371, "learning_rate": 3.0335787028092223e-06, "loss": 0.9797, "step": 32124 }, { "epoch": 2.5176332288401255, "grad_norm": 3.5198795795440674, "learning_rate": 3.033295193251713e-06, "loss": 1.0644, "step": 32125 }, { "epoch": 2.5177115987460814, "grad_norm": 3.253467082977295, "learning_rate": 3.033011691174541e-06, "loss": 0.9521, "step": 32126 }, { "epoch": 2.5177899686520377, "grad_norm": 5.033398151397705, "learning_rate": 3.03272819657878e-06, "loss": 1.1105, "step": 32127 }, { "epoch": 2.5178683385579936, "grad_norm": 2.448580503463745, "learning_rate": 3.0324447094655118e-06, "loss": 1.1404, "step": 32128 }, { "epoch": 2.51794670846395, "grad_norm": 3.384632110595703, "learning_rate": 3.0321612298358125e-06, "loss": 1.3295, "step": 32129 }, { "epoch": 2.518025078369906, "grad_norm": 2.5603909492492676, "learning_rate": 3.031877757690763e-06, "loss": 0.8591, "step": 32130 }, { "epoch": 2.518103448275862, "grad_norm": 4.080389499664307, "learning_rate": 3.031594293031438e-06, "loss": 1.5641, "step": 32131 }, { "epoch": 2.518181818181818, "grad_norm": 5.172976493835449, "learning_rate": 3.031310835858916e-06, "loss": 1.0649, "step": 32132 }, { "epoch": 2.5182601880877744, "grad_norm": 9.423646926879883, "learning_rate": 3.0310273861742778e-06, "loss": 0.9277, "step": 32133 }, { "epoch": 2.5183385579937303, "grad_norm": 2.1060426235198975, "learning_rate": 3.030743943978601e-06, "loss": 1.0662, "step": 32134 }, { "epoch": 2.5184169278996866, "grad_norm": 2.4210457801818848, "learning_rate": 3.0304605092729617e-06, "loss": 1.501, "step": 32135 }, { "epoch": 2.5184952978056425, "grad_norm": 3.853792190551758, "learning_rate": 3.0301770820584392e-06, "loss": 1.4226, "step": 32136 }, { "epoch": 2.518573667711599, "grad_norm": 3.8363218307495117, "learning_rate": 3.0298936623361104e-06, "loss": 1.5909, "step": 32137 }, { "epoch": 2.5186520376175547, "grad_norm": 3.943988800048828, "learning_rate": 3.0296102501070555e-06, "loss": 1.417, "step": 32138 }, { "epoch": 2.518730407523511, "grad_norm": 3.8038337230682373, "learning_rate": 3.0293268453723513e-06, "loss": 1.4473, "step": 32139 }, { "epoch": 2.518808777429467, "grad_norm": 4.076805591583252, "learning_rate": 3.0290434481330746e-06, "loss": 1.1731, "step": 32140 }, { "epoch": 2.5188871473354233, "grad_norm": 5.339836597442627, "learning_rate": 3.0287600583903054e-06, "loss": 1.4537, "step": 32141 }, { "epoch": 2.518965517241379, "grad_norm": 2.495485305786133, "learning_rate": 3.02847667614512e-06, "loss": 1.1521, "step": 32142 }, { "epoch": 2.5190438871473355, "grad_norm": 3.6867215633392334, "learning_rate": 3.0281933013985953e-06, "loss": 1.6217, "step": 32143 }, { "epoch": 2.5191222570532914, "grad_norm": 3.3902981281280518, "learning_rate": 3.0279099341518113e-06, "loss": 1.3734, "step": 32144 }, { "epoch": 2.5192006269592477, "grad_norm": 2.4795613288879395, "learning_rate": 3.0276265744058466e-06, "loss": 1.2966, "step": 32145 }, { "epoch": 2.5192789968652036, "grad_norm": 3.219362258911133, "learning_rate": 3.027343222161776e-06, "loss": 1.4891, "step": 32146 }, { "epoch": 2.51935736677116, "grad_norm": 3.5140082836151123, "learning_rate": 3.0270598774206783e-06, "loss": 0.99, "step": 32147 }, { "epoch": 2.519435736677116, "grad_norm": 3.7054824829101562, "learning_rate": 3.026776540183632e-06, "loss": 1.1622, "step": 32148 }, { "epoch": 2.519514106583072, "grad_norm": 2.5150840282440186, "learning_rate": 3.026493210451715e-06, "loss": 1.2548, "step": 32149 }, { "epoch": 2.519592476489028, "grad_norm": 4.01362419128418, "learning_rate": 3.0262098882260017e-06, "loss": 1.506, "step": 32150 }, { "epoch": 2.5196708463949844, "grad_norm": 2.2221791744232178, "learning_rate": 3.0259265735075737e-06, "loss": 0.8323, "step": 32151 }, { "epoch": 2.5197492163009403, "grad_norm": 2.726102828979492, "learning_rate": 3.0256432662975067e-06, "loss": 1.1067, "step": 32152 }, { "epoch": 2.5198275862068966, "grad_norm": 3.7201316356658936, "learning_rate": 3.02535996659688e-06, "loss": 1.6226, "step": 32153 }, { "epoch": 2.5199059561128525, "grad_norm": 2.5355148315429688, "learning_rate": 3.0250766744067684e-06, "loss": 1.3398, "step": 32154 }, { "epoch": 2.519984326018809, "grad_norm": 3.6821510791778564, "learning_rate": 3.02479338972825e-06, "loss": 1.6856, "step": 32155 }, { "epoch": 2.5200626959247647, "grad_norm": 5.686016082763672, "learning_rate": 3.0245101125624064e-06, "loss": 1.6496, "step": 32156 }, { "epoch": 2.520141065830721, "grad_norm": 4.153397560119629, "learning_rate": 3.0242268429103084e-06, "loss": 1.116, "step": 32157 }, { "epoch": 2.520219435736677, "grad_norm": 5.398716449737549, "learning_rate": 3.023943580773037e-06, "loss": 1.1596, "step": 32158 }, { "epoch": 2.5202978056426333, "grad_norm": 3.0911853313446045, "learning_rate": 3.02366032615167e-06, "loss": 1.3926, "step": 32159 }, { "epoch": 2.5203761755485896, "grad_norm": 3.0787365436553955, "learning_rate": 3.023377079047285e-06, "loss": 1.2487, "step": 32160 }, { "epoch": 2.5204545454545455, "grad_norm": 3.926905870437622, "learning_rate": 3.0230938394609567e-06, "loss": 1.401, "step": 32161 }, { "epoch": 2.5205329153605014, "grad_norm": 3.6189255714416504, "learning_rate": 3.0228106073937637e-06, "loss": 1.6986, "step": 32162 }, { "epoch": 2.5206112852664577, "grad_norm": 3.0538341999053955, "learning_rate": 3.022527382846784e-06, "loss": 1.3239, "step": 32163 }, { "epoch": 2.520689655172414, "grad_norm": 3.28190016746521, "learning_rate": 3.0222441658210956e-06, "loss": 1.5006, "step": 32164 }, { "epoch": 2.52076802507837, "grad_norm": 9.159123420715332, "learning_rate": 3.0219609563177744e-06, "loss": 0.8635, "step": 32165 }, { "epoch": 2.520846394984326, "grad_norm": 4.205005645751953, "learning_rate": 3.0216777543378965e-06, "loss": 1.4513, "step": 32166 }, { "epoch": 2.520924764890282, "grad_norm": 5.419938087463379, "learning_rate": 3.02139455988254e-06, "loss": 1.2459, "step": 32167 }, { "epoch": 2.5210031347962385, "grad_norm": 2.3414413928985596, "learning_rate": 3.0211113729527854e-06, "loss": 1.4656, "step": 32168 }, { "epoch": 2.5210815047021944, "grad_norm": 3.6580491065979004, "learning_rate": 3.020828193549703e-06, "loss": 0.861, "step": 32169 }, { "epoch": 2.5211598746081503, "grad_norm": 5.96715784072876, "learning_rate": 3.0205450216743753e-06, "loss": 1.4038, "step": 32170 }, { "epoch": 2.5212382445141066, "grad_norm": 3.799978256225586, "learning_rate": 3.0202618573278785e-06, "loss": 1.4089, "step": 32171 }, { "epoch": 2.521316614420063, "grad_norm": 2.1136252880096436, "learning_rate": 3.019978700511288e-06, "loss": 1.1443, "step": 32172 }, { "epoch": 2.521394984326019, "grad_norm": 2.207695960998535, "learning_rate": 3.019695551225681e-06, "loss": 1.1309, "step": 32173 }, { "epoch": 2.5214733542319747, "grad_norm": 3.2859320640563965, "learning_rate": 3.019412409472134e-06, "loss": 1.3961, "step": 32174 }, { "epoch": 2.521551724137931, "grad_norm": 2.770293712615967, "learning_rate": 3.0191292752517276e-06, "loss": 1.1721, "step": 32175 }, { "epoch": 2.5216300940438874, "grad_norm": 2.0028460025787354, "learning_rate": 3.018846148565534e-06, "loss": 1.2637, "step": 32176 }, { "epoch": 2.5217084639498433, "grad_norm": 3.8812613487243652, "learning_rate": 3.0185630294146318e-06, "loss": 0.7177, "step": 32177 }, { "epoch": 2.521786833855799, "grad_norm": 4.159058094024658, "learning_rate": 3.0182799178000987e-06, "loss": 1.3863, "step": 32178 }, { "epoch": 2.5218652037617555, "grad_norm": 3.3245904445648193, "learning_rate": 3.0179968137230124e-06, "loss": 1.2584, "step": 32179 }, { "epoch": 2.521943573667712, "grad_norm": 4.146753787994385, "learning_rate": 3.0177137171844474e-06, "loss": 0.9804, "step": 32180 }, { "epoch": 2.5220219435736677, "grad_norm": 3.3415732383728027, "learning_rate": 3.0174306281854793e-06, "loss": 1.6117, "step": 32181 }, { "epoch": 2.5221003134796236, "grad_norm": 2.572219133377075, "learning_rate": 3.01714754672719e-06, "loss": 1.2938, "step": 32182 }, { "epoch": 2.52217868338558, "grad_norm": 2.9080569744110107, "learning_rate": 3.01686447281065e-06, "loss": 1.5981, "step": 32183 }, { "epoch": 2.5222570532915363, "grad_norm": 7.9772772789001465, "learning_rate": 3.0165814064369404e-06, "loss": 1.318, "step": 32184 }, { "epoch": 2.522335423197492, "grad_norm": 3.887404203414917, "learning_rate": 3.016298347607136e-06, "loss": 0.895, "step": 32185 }, { "epoch": 2.522413793103448, "grad_norm": 3.5038743019104004, "learning_rate": 3.0160152963223143e-06, "loss": 1.5054, "step": 32186 }, { "epoch": 2.5224921630094044, "grad_norm": 2.759686231613159, "learning_rate": 3.015732252583551e-06, "loss": 1.3769, "step": 32187 }, { "epoch": 2.5225705329153607, "grad_norm": 2.378159523010254, "learning_rate": 3.0154492163919223e-06, "loss": 1.7049, "step": 32188 }, { "epoch": 2.5226489028213166, "grad_norm": 5.240152359008789, "learning_rate": 3.0151661877485063e-06, "loss": 0.9607, "step": 32189 }, { "epoch": 2.5227272727272725, "grad_norm": 2.586487054824829, "learning_rate": 3.0148831666543785e-06, "loss": 1.176, "step": 32190 }, { "epoch": 2.522805642633229, "grad_norm": 5.261894702911377, "learning_rate": 3.014600153110616e-06, "loss": 1.2343, "step": 32191 }, { "epoch": 2.522884012539185, "grad_norm": 3.0754916667938232, "learning_rate": 3.0143171471182942e-06, "loss": 1.0529, "step": 32192 }, { "epoch": 2.522962382445141, "grad_norm": 3.6880781650543213, "learning_rate": 3.014034148678489e-06, "loss": 1.1794, "step": 32193 }, { "epoch": 2.523040752351097, "grad_norm": 2.3044168949127197, "learning_rate": 3.0137511577922807e-06, "loss": 0.9758, "step": 32194 }, { "epoch": 2.5231191222570533, "grad_norm": 5.226810455322266, "learning_rate": 3.01346817446074e-06, "loss": 1.3149, "step": 32195 }, { "epoch": 2.5231974921630096, "grad_norm": 2.8718419075012207, "learning_rate": 3.013185198684947e-06, "loss": 1.4758, "step": 32196 }, { "epoch": 2.5232758620689655, "grad_norm": 3.462804079055786, "learning_rate": 3.0129022304659785e-06, "loss": 1.0107, "step": 32197 }, { "epoch": 2.5233542319749214, "grad_norm": 3.602255344390869, "learning_rate": 3.012619269804907e-06, "loss": 1.2371, "step": 32198 }, { "epoch": 2.5234326018808777, "grad_norm": 3.138652801513672, "learning_rate": 3.0123363167028124e-06, "loss": 1.6714, "step": 32199 }, { "epoch": 2.523510971786834, "grad_norm": 24.960479736328125, "learning_rate": 3.012053371160768e-06, "loss": 1.3199, "step": 32200 }, { "epoch": 2.52358934169279, "grad_norm": 3.5645182132720947, "learning_rate": 3.0117704331798546e-06, "loss": 1.1078, "step": 32201 }, { "epoch": 2.523667711598746, "grad_norm": 3.2158560752868652, "learning_rate": 3.0114875027611418e-06, "loss": 0.9155, "step": 32202 }, { "epoch": 2.523746081504702, "grad_norm": 2.231997489929199, "learning_rate": 3.0112045799057106e-06, "loss": 1.2646, "step": 32203 }, { "epoch": 2.5238244514106585, "grad_norm": 3.638421058654785, "learning_rate": 3.0109216646146356e-06, "loss": 1.2805, "step": 32204 }, { "epoch": 2.5239028213166144, "grad_norm": 2.8012309074401855, "learning_rate": 3.0106387568889937e-06, "loss": 1.3406, "step": 32205 }, { "epoch": 2.5239811912225707, "grad_norm": 2.3830838203430176, "learning_rate": 3.0103558567298595e-06, "loss": 1.5426, "step": 32206 }, { "epoch": 2.5240595611285266, "grad_norm": 3.775311231613159, "learning_rate": 3.0100729641383086e-06, "loss": 1.4051, "step": 32207 }, { "epoch": 2.524137931034483, "grad_norm": 3.0485405921936035, "learning_rate": 3.009790079115419e-06, "loss": 1.1396, "step": 32208 }, { "epoch": 2.524216300940439, "grad_norm": 2.331899881362915, "learning_rate": 3.009507201662266e-06, "loss": 1.3148, "step": 32209 }, { "epoch": 2.524294670846395, "grad_norm": 2.327962875366211, "learning_rate": 3.0092243317799253e-06, "loss": 1.6484, "step": 32210 }, { "epoch": 2.524373040752351, "grad_norm": 4.413043022155762, "learning_rate": 3.0089414694694723e-06, "loss": 1.4219, "step": 32211 }, { "epoch": 2.5244514106583074, "grad_norm": 2.6506805419921875, "learning_rate": 3.008658614731984e-06, "loss": 1.0472, "step": 32212 }, { "epoch": 2.5245297805642632, "grad_norm": 3.7893385887145996, "learning_rate": 3.0083757675685354e-06, "loss": 1.5454, "step": 32213 }, { "epoch": 2.5246081504702196, "grad_norm": 10.611237525939941, "learning_rate": 3.008092927980201e-06, "loss": 1.5081, "step": 32214 }, { "epoch": 2.5246865203761755, "grad_norm": 2.8179619312286377, "learning_rate": 3.0078100959680584e-06, "loss": 1.3883, "step": 32215 }, { "epoch": 2.524764890282132, "grad_norm": 3.662741184234619, "learning_rate": 3.0075272715331843e-06, "loss": 1.4237, "step": 32216 }, { "epoch": 2.5248432601880877, "grad_norm": 3.393399238586426, "learning_rate": 3.007244454676652e-06, "loss": 1.0185, "step": 32217 }, { "epoch": 2.524921630094044, "grad_norm": 4.193595886230469, "learning_rate": 3.0069616453995385e-06, "loss": 1.5647, "step": 32218 }, { "epoch": 2.525, "grad_norm": 7.0389404296875, "learning_rate": 3.0066788437029183e-06, "loss": 1.4241, "step": 32219 }, { "epoch": 2.5250783699059562, "grad_norm": 10.455988883972168, "learning_rate": 3.0063960495878706e-06, "loss": 1.1908, "step": 32220 }, { "epoch": 2.525156739811912, "grad_norm": 2.957815170288086, "learning_rate": 3.0061132630554657e-06, "loss": 0.4595, "step": 32221 }, { "epoch": 2.5252351097178685, "grad_norm": 8.997325897216797, "learning_rate": 3.005830484106782e-06, "loss": 1.2236, "step": 32222 }, { "epoch": 2.5253134796238244, "grad_norm": 3.3220648765563965, "learning_rate": 3.0055477127428955e-06, "loss": 1.2663, "step": 32223 }, { "epoch": 2.5253918495297807, "grad_norm": 2.175855875015259, "learning_rate": 3.005264948964882e-06, "loss": 1.3108, "step": 32224 }, { "epoch": 2.5254702194357366, "grad_norm": 7.382209300994873, "learning_rate": 3.0049821927738152e-06, "loss": 0.7811, "step": 32225 }, { "epoch": 2.525548589341693, "grad_norm": 2.901202440261841, "learning_rate": 3.0046994441707706e-06, "loss": 1.2981, "step": 32226 }, { "epoch": 2.525626959247649, "grad_norm": 4.4134979248046875, "learning_rate": 3.0044167031568274e-06, "loss": 1.2056, "step": 32227 }, { "epoch": 2.525705329153605, "grad_norm": 3.0183160305023193, "learning_rate": 3.0041339697330545e-06, "loss": 1.1728, "step": 32228 }, { "epoch": 2.525783699059561, "grad_norm": 3.0053906440734863, "learning_rate": 3.003851243900532e-06, "loss": 1.2244, "step": 32229 }, { "epoch": 2.5258620689655173, "grad_norm": 6.4214887619018555, "learning_rate": 3.003568525660334e-06, "loss": 0.8222, "step": 32230 }, { "epoch": 2.5259404388714732, "grad_norm": 3.544652223587036, "learning_rate": 3.003285815013538e-06, "loss": 1.3175, "step": 32231 }, { "epoch": 2.5260188087774296, "grad_norm": 5.9619035720825195, "learning_rate": 3.0030031119612154e-06, "loss": 1.1672, "step": 32232 }, { "epoch": 2.5260971786833855, "grad_norm": 2.799220323562622, "learning_rate": 3.002720416504442e-06, "loss": 1.5206, "step": 32233 }, { "epoch": 2.526175548589342, "grad_norm": 5.074211597442627, "learning_rate": 3.002437728644295e-06, "loss": 1.3964, "step": 32234 }, { "epoch": 2.5262539184952977, "grad_norm": 3.456981658935547, "learning_rate": 3.0021550483818507e-06, "loss": 1.0812, "step": 32235 }, { "epoch": 2.526332288401254, "grad_norm": 4.237425327301025, "learning_rate": 3.0018723757181813e-06, "loss": 1.4456, "step": 32236 }, { "epoch": 2.52641065830721, "grad_norm": 3.2498812675476074, "learning_rate": 3.0015897106543625e-06, "loss": 1.7598, "step": 32237 }, { "epoch": 2.5264890282131662, "grad_norm": 3.515439510345459, "learning_rate": 3.0013070531914714e-06, "loss": 1.2274, "step": 32238 }, { "epoch": 2.526567398119122, "grad_norm": 6.327769756317139, "learning_rate": 3.00102440333058e-06, "loss": 1.7657, "step": 32239 }, { "epoch": 2.5266457680250785, "grad_norm": 5.423401355743408, "learning_rate": 3.0007417610727653e-06, "loss": 1.1716, "step": 32240 }, { "epoch": 2.5267241379310343, "grad_norm": 8.930721282958984, "learning_rate": 3.0004591264191022e-06, "loss": 1.0746, "step": 32241 }, { "epoch": 2.5268025078369907, "grad_norm": 2.963945150375366, "learning_rate": 3.000176499370667e-06, "loss": 1.255, "step": 32242 }, { "epoch": 2.5268808777429466, "grad_norm": 2.2153992652893066, "learning_rate": 2.999893879928532e-06, "loss": 0.9509, "step": 32243 }, { "epoch": 2.526959247648903, "grad_norm": 4.712728500366211, "learning_rate": 2.999611268093773e-06, "loss": 1.789, "step": 32244 }, { "epoch": 2.527037617554859, "grad_norm": 4.087210178375244, "learning_rate": 2.999328663867465e-06, "loss": 1.4418, "step": 32245 }, { "epoch": 2.527115987460815, "grad_norm": 2.6060988903045654, "learning_rate": 2.999046067250685e-06, "loss": 1.0281, "step": 32246 }, { "epoch": 2.527194357366771, "grad_norm": 5.681641578674316, "learning_rate": 2.998763478244505e-06, "loss": 0.9842, "step": 32247 }, { "epoch": 2.5272727272727273, "grad_norm": 6.894808292388916, "learning_rate": 2.99848089685e-06, "loss": 0.8567, "step": 32248 }, { "epoch": 2.5273510971786832, "grad_norm": 4.609036922454834, "learning_rate": 2.9981983230682466e-06, "loss": 1.329, "step": 32249 }, { "epoch": 2.5274294670846396, "grad_norm": 3.458716869354248, "learning_rate": 2.997915756900319e-06, "loss": 1.3878, "step": 32250 }, { "epoch": 2.5275078369905954, "grad_norm": 3.0193698406219482, "learning_rate": 2.9976331983472897e-06, "loss": 1.6163, "step": 32251 }, { "epoch": 2.527586206896552, "grad_norm": 2.5299689769744873, "learning_rate": 2.9973506474102364e-06, "loss": 1.5437, "step": 32252 }, { "epoch": 2.5276645768025077, "grad_norm": 7.118593215942383, "learning_rate": 2.997068104090234e-06, "loss": 1.1546, "step": 32253 }, { "epoch": 2.527742946708464, "grad_norm": 4.6301374435424805, "learning_rate": 2.996785568388354e-06, "loss": 1.2634, "step": 32254 }, { "epoch": 2.5278213166144203, "grad_norm": 2.8322246074676514, "learning_rate": 2.9965030403056726e-06, "loss": 1.5312, "step": 32255 }, { "epoch": 2.527899686520376, "grad_norm": 2.632498025894165, "learning_rate": 2.9962205198432643e-06, "loss": 1.3715, "step": 32256 }, { "epoch": 2.527978056426332, "grad_norm": 3.8304505348205566, "learning_rate": 2.9959380070022073e-06, "loss": 1.438, "step": 32257 }, { "epoch": 2.5280564263322884, "grad_norm": 3.1453769207000732, "learning_rate": 2.995655501783569e-06, "loss": 1.4882, "step": 32258 }, { "epoch": 2.5281347962382448, "grad_norm": 2.421191692352295, "learning_rate": 2.995373004188428e-06, "loss": 1.5543, "step": 32259 }, { "epoch": 2.5282131661442007, "grad_norm": 9.632342338562012, "learning_rate": 2.9950905142178594e-06, "loss": 1.3609, "step": 32260 }, { "epoch": 2.5282915360501566, "grad_norm": 5.122470855712891, "learning_rate": 2.994808031872938e-06, "loss": 2.1737, "step": 32261 }, { "epoch": 2.528369905956113, "grad_norm": 4.7868781089782715, "learning_rate": 2.994525557154735e-06, "loss": 1.402, "step": 32262 }, { "epoch": 2.528448275862069, "grad_norm": 5.405541896820068, "learning_rate": 2.994243090064326e-06, "loss": 1.7588, "step": 32263 }, { "epoch": 2.528526645768025, "grad_norm": 4.244254112243652, "learning_rate": 2.9939606306027874e-06, "loss": 1.3321, "step": 32264 }, { "epoch": 2.528605015673981, "grad_norm": 3.76426362991333, "learning_rate": 2.993678178771193e-06, "loss": 1.5117, "step": 32265 }, { "epoch": 2.5286833855799373, "grad_norm": 3.812495708465576, "learning_rate": 2.9933957345706156e-06, "loss": 1.2611, "step": 32266 }, { "epoch": 2.5287617554858937, "grad_norm": 2.7705137729644775, "learning_rate": 2.9931132980021294e-06, "loss": 1.2573, "step": 32267 }, { "epoch": 2.5288401253918495, "grad_norm": 3.533735990524292, "learning_rate": 2.992830869066811e-06, "loss": 1.2255, "step": 32268 }, { "epoch": 2.5289184952978054, "grad_norm": 2.7186119556427, "learning_rate": 2.992548447765732e-06, "loss": 1.4281, "step": 32269 }, { "epoch": 2.5289968652037618, "grad_norm": 2.8024909496307373, "learning_rate": 2.9922660340999664e-06, "loss": 1.2036, "step": 32270 }, { "epoch": 2.529075235109718, "grad_norm": 3.261615753173828, "learning_rate": 2.991983628070591e-06, "loss": 1.4901, "step": 32271 }, { "epoch": 2.529153605015674, "grad_norm": 5.415919303894043, "learning_rate": 2.991701229678679e-06, "loss": 1.8089, "step": 32272 }, { "epoch": 2.52923197492163, "grad_norm": 2.4694976806640625, "learning_rate": 2.991418838925303e-06, "loss": 1.2901, "step": 32273 }, { "epoch": 2.529310344827586, "grad_norm": 3.421314001083374, "learning_rate": 2.9911364558115387e-06, "loss": 1.4114, "step": 32274 }, { "epoch": 2.5293887147335425, "grad_norm": 4.156943321228027, "learning_rate": 2.9908540803384577e-06, "loss": 1.2773, "step": 32275 }, { "epoch": 2.5294670846394984, "grad_norm": 2.9115867614746094, "learning_rate": 2.99057171250714e-06, "loss": 1.3982, "step": 32276 }, { "epoch": 2.5295454545454543, "grad_norm": 2.957230806350708, "learning_rate": 2.990289352318651e-06, "loss": 1.2374, "step": 32277 }, { "epoch": 2.5296238244514107, "grad_norm": 4.320314884185791, "learning_rate": 2.990006999774071e-06, "loss": 1.2408, "step": 32278 }, { "epoch": 2.529702194357367, "grad_norm": 2.2167773246765137, "learning_rate": 2.989724654874471e-06, "loss": 1.3459, "step": 32279 }, { "epoch": 2.529780564263323, "grad_norm": 3.5128660202026367, "learning_rate": 2.9894423176209275e-06, "loss": 1.2093, "step": 32280 }, { "epoch": 2.5298589341692788, "grad_norm": 3.1637115478515625, "learning_rate": 2.9891599880145116e-06, "loss": 0.9815, "step": 32281 }, { "epoch": 2.529937304075235, "grad_norm": 7.681966781616211, "learning_rate": 2.988877666056298e-06, "loss": 1.7833, "step": 32282 }, { "epoch": 2.5300156739811914, "grad_norm": 5.309986591339111, "learning_rate": 2.9885953517473633e-06, "loss": 0.9179, "step": 32283 }, { "epoch": 2.5300940438871473, "grad_norm": 2.6520893573760986, "learning_rate": 2.9883130450887754e-06, "loss": 1.1661, "step": 32284 }, { "epoch": 2.530172413793103, "grad_norm": 4.973044395446777, "learning_rate": 2.988030746081613e-06, "loss": 1.0793, "step": 32285 }, { "epoch": 2.5302507836990595, "grad_norm": 2.1571803092956543, "learning_rate": 2.9877484547269474e-06, "loss": 1.3918, "step": 32286 }, { "epoch": 2.530329153605016, "grad_norm": 4.639614582061768, "learning_rate": 2.9874661710258544e-06, "loss": 1.4419, "step": 32287 }, { "epoch": 2.5304075235109718, "grad_norm": 3.770648956298828, "learning_rate": 2.9871838949794058e-06, "loss": 1.1826, "step": 32288 }, { "epoch": 2.5304858934169276, "grad_norm": 3.8738174438476562, "learning_rate": 2.9869016265886743e-06, "loss": 1.3115, "step": 32289 }, { "epoch": 2.530564263322884, "grad_norm": 3.847642421722412, "learning_rate": 2.9866193658547365e-06, "loss": 0.4364, "step": 32290 }, { "epoch": 2.5306426332288403, "grad_norm": 2.3376688957214355, "learning_rate": 2.9863371127786656e-06, "loss": 1.3613, "step": 32291 }, { "epoch": 2.530721003134796, "grad_norm": 3.5933680534362793, "learning_rate": 2.9860548673615327e-06, "loss": 1.5093, "step": 32292 }, { "epoch": 2.530799373040752, "grad_norm": 3.302765130996704, "learning_rate": 2.985772629604412e-06, "loss": 1.8408, "step": 32293 }, { "epoch": 2.5308777429467084, "grad_norm": 3.851046562194824, "learning_rate": 2.9854903995083795e-06, "loss": 1.6353, "step": 32294 }, { "epoch": 2.5309561128526648, "grad_norm": 3.650960922241211, "learning_rate": 2.9852081770745056e-06, "loss": 1.4033, "step": 32295 }, { "epoch": 2.5310344827586206, "grad_norm": 6.850396633148193, "learning_rate": 2.984925962303864e-06, "loss": 1.5426, "step": 32296 }, { "epoch": 2.5311128526645765, "grad_norm": 5.711291313171387, "learning_rate": 2.9846437551975304e-06, "loss": 1.2476, "step": 32297 }, { "epoch": 2.531191222570533, "grad_norm": 2.488537549972534, "learning_rate": 2.9843615557565776e-06, "loss": 1.3525, "step": 32298 }, { "epoch": 2.531269592476489, "grad_norm": 2.196272134780884, "learning_rate": 2.9840793639820764e-06, "loss": 1.0608, "step": 32299 }, { "epoch": 2.531347962382445, "grad_norm": 2.431316375732422, "learning_rate": 2.9837971798751025e-06, "loss": 1.3941, "step": 32300 }, { "epoch": 2.5314263322884014, "grad_norm": 3.008653402328491, "learning_rate": 2.983515003436728e-06, "loss": 1.145, "step": 32301 }, { "epoch": 2.5315047021943573, "grad_norm": 3.62023663520813, "learning_rate": 2.983232834668028e-06, "loss": 1.9478, "step": 32302 }, { "epoch": 2.5315830721003136, "grad_norm": 4.115704536437988, "learning_rate": 2.9829506735700733e-06, "loss": 1.5407, "step": 32303 }, { "epoch": 2.5316614420062695, "grad_norm": 11.523365020751953, "learning_rate": 2.982668520143939e-06, "loss": 0.9978, "step": 32304 }, { "epoch": 2.531739811912226, "grad_norm": 3.0341007709503174, "learning_rate": 2.982386374390697e-06, "loss": 1.2195, "step": 32305 }, { "epoch": 2.5318181818181817, "grad_norm": 2.453712224960327, "learning_rate": 2.9821042363114217e-06, "loss": 1.5118, "step": 32306 }, { "epoch": 2.531896551724138, "grad_norm": 3.0048513412475586, "learning_rate": 2.9818221059071837e-06, "loss": 1.1188, "step": 32307 }, { "epoch": 2.531974921630094, "grad_norm": 3.8971199989318848, "learning_rate": 2.981539983179058e-06, "loss": 1.6368, "step": 32308 }, { "epoch": 2.5320532915360503, "grad_norm": 5.110628604888916, "learning_rate": 2.9812578681281203e-06, "loss": 1.3315, "step": 32309 }, { "epoch": 2.532131661442006, "grad_norm": 2.0956614017486572, "learning_rate": 2.980975760755438e-06, "loss": 1.1781, "step": 32310 }, { "epoch": 2.5322100313479625, "grad_norm": 3.272101640701294, "learning_rate": 2.9806936610620874e-06, "loss": 1.2368, "step": 32311 }, { "epoch": 2.5322884012539184, "grad_norm": 3.2362515926361084, "learning_rate": 2.980411569049141e-06, "loss": 0.7022, "step": 32312 }, { "epoch": 2.5323667711598747, "grad_norm": 2.1593756675720215, "learning_rate": 2.9801294847176727e-06, "loss": 1.1089, "step": 32313 }, { "epoch": 2.5324451410658306, "grad_norm": 2.6417720317840576, "learning_rate": 2.9798474080687535e-06, "loss": 1.0845, "step": 32314 }, { "epoch": 2.532523510971787, "grad_norm": 4.2113261222839355, "learning_rate": 2.979565339103456e-06, "loss": 1.287, "step": 32315 }, { "epoch": 2.532601880877743, "grad_norm": 3.6148955821990967, "learning_rate": 2.9792832778228555e-06, "loss": 1.554, "step": 32316 }, { "epoch": 2.532680250783699, "grad_norm": 4.081969738006592, "learning_rate": 2.979001224228024e-06, "loss": 1.3719, "step": 32317 }, { "epoch": 2.532758620689655, "grad_norm": 2.375652313232422, "learning_rate": 2.978719178320033e-06, "loss": 1.4053, "step": 32318 }, { "epoch": 2.5328369905956114, "grad_norm": 3.0331521034240723, "learning_rate": 2.9784371400999556e-06, "loss": 1.6981, "step": 32319 }, { "epoch": 2.5329153605015673, "grad_norm": 2.595686197280884, "learning_rate": 2.978155109568864e-06, "loss": 1.2673, "step": 32320 }, { "epoch": 2.5329937304075236, "grad_norm": 2.615926504135132, "learning_rate": 2.9778730867278362e-06, "loss": 1.3097, "step": 32321 }, { "epoch": 2.5330721003134795, "grad_norm": 6.238137722015381, "learning_rate": 2.9775910715779365e-06, "loss": 1.0595, "step": 32322 }, { "epoch": 2.533150470219436, "grad_norm": 2.543900966644287, "learning_rate": 2.977309064120243e-06, "loss": 1.3024, "step": 32323 }, { "epoch": 2.5332288401253917, "grad_norm": 3.2135140895843506, "learning_rate": 2.9770270643558275e-06, "loss": 1.2484, "step": 32324 }, { "epoch": 2.533307210031348, "grad_norm": 3.545689105987549, "learning_rate": 2.9767450722857612e-06, "loss": 1.4071, "step": 32325 }, { "epoch": 2.533385579937304, "grad_norm": 3.504127264022827, "learning_rate": 2.976463087911117e-06, "loss": 1.6041, "step": 32326 }, { "epoch": 2.5334639498432603, "grad_norm": 4.041430473327637, "learning_rate": 2.9761811112329677e-06, "loss": 1.3654, "step": 32327 }, { "epoch": 2.533542319749216, "grad_norm": 2.8933889865875244, "learning_rate": 2.9758991422523885e-06, "loss": 1.3633, "step": 32328 }, { "epoch": 2.5336206896551725, "grad_norm": 4.991590976715088, "learning_rate": 2.975617180970446e-06, "loss": 1.114, "step": 32329 }, { "epoch": 2.5336990595611284, "grad_norm": 3.385974168777466, "learning_rate": 2.9753352273882173e-06, "loss": 1.3456, "step": 32330 }, { "epoch": 2.5337774294670847, "grad_norm": 3.1978750228881836, "learning_rate": 2.9750532815067734e-06, "loss": 1.17, "step": 32331 }, { "epoch": 2.5338557993730406, "grad_norm": 5.673486232757568, "learning_rate": 2.974771343327188e-06, "loss": 1.8494, "step": 32332 }, { "epoch": 2.533934169278997, "grad_norm": 4.08282470703125, "learning_rate": 2.974489412850531e-06, "loss": 1.4386, "step": 32333 }, { "epoch": 2.534012539184953, "grad_norm": 5.548247337341309, "learning_rate": 2.974207490077875e-06, "loss": 1.2202, "step": 32334 }, { "epoch": 2.534090909090909, "grad_norm": 3.0321595668792725, "learning_rate": 2.9739255750102937e-06, "loss": 1.2886, "step": 32335 }, { "epoch": 2.534169278996865, "grad_norm": 6.278369903564453, "learning_rate": 2.97364366764886e-06, "loss": 1.1979, "step": 32336 }, { "epoch": 2.5342476489028214, "grad_norm": 2.2749924659729004, "learning_rate": 2.9733617679946446e-06, "loss": 1.3964, "step": 32337 }, { "epoch": 2.5343260188087773, "grad_norm": 4.449681282043457, "learning_rate": 2.9730798760487194e-06, "loss": 1.1472, "step": 32338 }, { "epoch": 2.5344043887147336, "grad_norm": 2.5874221324920654, "learning_rate": 2.9727979918121587e-06, "loss": 1.2406, "step": 32339 }, { "epoch": 2.5344827586206895, "grad_norm": 4.032315731048584, "learning_rate": 2.972516115286032e-06, "loss": 1.0653, "step": 32340 }, { "epoch": 2.534561128526646, "grad_norm": 2.760921001434326, "learning_rate": 2.9722342464714116e-06, "loss": 1.4812, "step": 32341 }, { "epoch": 2.5346394984326017, "grad_norm": 2.881718158721924, "learning_rate": 2.971952385369372e-06, "loss": 1.21, "step": 32342 }, { "epoch": 2.534717868338558, "grad_norm": 2.1581640243530273, "learning_rate": 2.9716705319809846e-06, "loss": 0.9523, "step": 32343 }, { "epoch": 2.534796238244514, "grad_norm": 3.486969232559204, "learning_rate": 2.9713886863073195e-06, "loss": 1.1961, "step": 32344 }, { "epoch": 2.5348746081504703, "grad_norm": 3.3769450187683105, "learning_rate": 2.97110684834945e-06, "loss": 0.825, "step": 32345 }, { "epoch": 2.5349529780564266, "grad_norm": 3.251164436340332, "learning_rate": 2.9708250181084464e-06, "loss": 1.2916, "step": 32346 }, { "epoch": 2.5350313479623825, "grad_norm": 5.078383445739746, "learning_rate": 2.9705431955853848e-06, "loss": 1.4665, "step": 32347 }, { "epoch": 2.5351097178683384, "grad_norm": 5.877395153045654, "learning_rate": 2.9702613807813335e-06, "loss": 1.4355, "step": 32348 }, { "epoch": 2.5351880877742947, "grad_norm": 3.0924525260925293, "learning_rate": 2.969979573697366e-06, "loss": 1.2399, "step": 32349 }, { "epoch": 2.535266457680251, "grad_norm": 3.736023187637329, "learning_rate": 2.9696977743345533e-06, "loss": 1.1672, "step": 32350 }, { "epoch": 2.535344827586207, "grad_norm": 3.1625726222991943, "learning_rate": 2.9694159826939665e-06, "loss": 1.5205, "step": 32351 }, { "epoch": 2.535423197492163, "grad_norm": 2.7231972217559814, "learning_rate": 2.9691341987766782e-06, "loss": 1.5326, "step": 32352 }, { "epoch": 2.535501567398119, "grad_norm": 3.0179669857025146, "learning_rate": 2.9688524225837608e-06, "loss": 1.1941, "step": 32353 }, { "epoch": 2.5355799373040755, "grad_norm": 3.364685297012329, "learning_rate": 2.968570654116286e-06, "loss": 1.7397, "step": 32354 }, { "epoch": 2.5356583072100314, "grad_norm": 3.0945193767547607, "learning_rate": 2.9682888933753252e-06, "loss": 1.2221, "step": 32355 }, { "epoch": 2.5357366771159873, "grad_norm": 6.509010314941406, "learning_rate": 2.9680071403619493e-06, "loss": 1.1666, "step": 32356 }, { "epoch": 2.5358150470219436, "grad_norm": 4.462770462036133, "learning_rate": 2.967725395077229e-06, "loss": 0.6638, "step": 32357 }, { "epoch": 2.5358934169279, "grad_norm": 4.689025402069092, "learning_rate": 2.967443657522241e-06, "loss": 1.1979, "step": 32358 }, { "epoch": 2.535971786833856, "grad_norm": 2.695697784423828, "learning_rate": 2.9671619276980502e-06, "loss": 1.2255, "step": 32359 }, { "epoch": 2.5360501567398117, "grad_norm": 3.472210645675659, "learning_rate": 2.9668802056057327e-06, "loss": 1.4836, "step": 32360 }, { "epoch": 2.536128526645768, "grad_norm": 4.185080051422119, "learning_rate": 2.9665984912463585e-06, "loss": 1.0984, "step": 32361 }, { "epoch": 2.5362068965517244, "grad_norm": 2.91471529006958, "learning_rate": 2.966316784621e-06, "loss": 0.9849, "step": 32362 }, { "epoch": 2.5362852664576803, "grad_norm": 2.934619665145874, "learning_rate": 2.966035085730727e-06, "loss": 0.9165, "step": 32363 }, { "epoch": 2.536363636363636, "grad_norm": 4.396378993988037, "learning_rate": 2.9657533945766102e-06, "loss": 1.342, "step": 32364 }, { "epoch": 2.5364420062695925, "grad_norm": 7.94171667098999, "learning_rate": 2.9654717111597265e-06, "loss": 1.8941, "step": 32365 }, { "epoch": 2.536520376175549, "grad_norm": 2.978515386581421, "learning_rate": 2.96519003548114e-06, "loss": 1.1378, "step": 32366 }, { "epoch": 2.5365987460815047, "grad_norm": 6.517471790313721, "learning_rate": 2.9649083675419268e-06, "loss": 1.4724, "step": 32367 }, { "epoch": 2.5366771159874606, "grad_norm": 2.756406545639038, "learning_rate": 2.9646267073431567e-06, "loss": 1.6075, "step": 32368 }, { "epoch": 2.536755485893417, "grad_norm": 2.249640464782715, "learning_rate": 2.964345054885902e-06, "loss": 1.4147, "step": 32369 }, { "epoch": 2.5368338557993733, "grad_norm": 3.833710193634033, "learning_rate": 2.9640634101712317e-06, "loss": 1.3117, "step": 32370 }, { "epoch": 2.536912225705329, "grad_norm": 3.211688756942749, "learning_rate": 2.9637817732002184e-06, "loss": 1.1897, "step": 32371 }, { "epoch": 2.536990595611285, "grad_norm": 3.3835675716400146, "learning_rate": 2.963500143973934e-06, "loss": 1.4043, "step": 32372 }, { "epoch": 2.5370689655172414, "grad_norm": 2.3142447471618652, "learning_rate": 2.96321852249345e-06, "loss": 1.3917, "step": 32373 }, { "epoch": 2.5371473354231977, "grad_norm": 4.50723934173584, "learning_rate": 2.9629369087598357e-06, "loss": 1.6687, "step": 32374 }, { "epoch": 2.5372257053291536, "grad_norm": 3.744019031524658, "learning_rate": 2.9626553027741634e-06, "loss": 1.8684, "step": 32375 }, { "epoch": 2.5373040752351095, "grad_norm": 2.920872211456299, "learning_rate": 2.9623737045375022e-06, "loss": 1.3323, "step": 32376 }, { "epoch": 2.537382445141066, "grad_norm": 2.917527198791504, "learning_rate": 2.962092114050928e-06, "loss": 1.4712, "step": 32377 }, { "epoch": 2.537460815047022, "grad_norm": 4.678043365478516, "learning_rate": 2.9618105313155053e-06, "loss": 1.659, "step": 32378 }, { "epoch": 2.537539184952978, "grad_norm": 2.760373830795288, "learning_rate": 2.96152895633231e-06, "loss": 0.9858, "step": 32379 }, { "epoch": 2.537617554858934, "grad_norm": 3.32999587059021, "learning_rate": 2.961247389102413e-06, "loss": 1.3784, "step": 32380 }, { "epoch": 2.5376959247648903, "grad_norm": 5.995907306671143, "learning_rate": 2.9609658296268816e-06, "loss": 1.7289, "step": 32381 }, { "epoch": 2.5377742946708466, "grad_norm": 7.3070831298828125, "learning_rate": 2.96068427790679e-06, "loss": 1.4828, "step": 32382 }, { "epoch": 2.5378526645768025, "grad_norm": 4.508825302124023, "learning_rate": 2.960402733943206e-06, "loss": 1.2726, "step": 32383 }, { "epoch": 2.5379310344827584, "grad_norm": 3.791933536529541, "learning_rate": 2.960121197737206e-06, "loss": 1.4984, "step": 32384 }, { "epoch": 2.5380094043887147, "grad_norm": 2.2496402263641357, "learning_rate": 2.9598396692898545e-06, "loss": 1.5046, "step": 32385 }, { "epoch": 2.538087774294671, "grad_norm": 3.702641010284424, "learning_rate": 2.959558148602225e-06, "loss": 1.5323, "step": 32386 }, { "epoch": 2.538166144200627, "grad_norm": 4.138941764831543, "learning_rate": 2.9592766356753887e-06, "loss": 1.6883, "step": 32387 }, { "epoch": 2.538244514106583, "grad_norm": 1.8442423343658447, "learning_rate": 2.9589951305104177e-06, "loss": 1.2552, "step": 32388 }, { "epoch": 2.538322884012539, "grad_norm": 4.523461818695068, "learning_rate": 2.9587136331083787e-06, "loss": 1.3621, "step": 32389 }, { "epoch": 2.5384012539184955, "grad_norm": 2.4700675010681152, "learning_rate": 2.9584321434703444e-06, "loss": 1.4257, "step": 32390 }, { "epoch": 2.5384796238244514, "grad_norm": 3.765612840652466, "learning_rate": 2.958150661597388e-06, "loss": 1.3789, "step": 32391 }, { "epoch": 2.5385579937304072, "grad_norm": 4.271660327911377, "learning_rate": 2.957869187490575e-06, "loss": 1.9876, "step": 32392 }, { "epoch": 2.5386363636363636, "grad_norm": 2.777425527572632, "learning_rate": 2.9575877211509795e-06, "loss": 1.2796, "step": 32393 }, { "epoch": 2.53871473354232, "grad_norm": 4.411571025848389, "learning_rate": 2.957306262579672e-06, "loss": 1.1613, "step": 32394 }, { "epoch": 2.538793103448276, "grad_norm": 3.5530805587768555, "learning_rate": 2.957024811777723e-06, "loss": 1.5263, "step": 32395 }, { "epoch": 2.538871473354232, "grad_norm": 3.1454484462738037, "learning_rate": 2.956743368746201e-06, "loss": 1.2951, "step": 32396 }, { "epoch": 2.538949843260188, "grad_norm": 16.765438079833984, "learning_rate": 2.9564619334861764e-06, "loss": 0.6695, "step": 32397 }, { "epoch": 2.5390282131661444, "grad_norm": 1.9892070293426514, "learning_rate": 2.9561805059987226e-06, "loss": 1.1597, "step": 32398 }, { "epoch": 2.5391065830721002, "grad_norm": 4.044309139251709, "learning_rate": 2.955899086284909e-06, "loss": 1.0145, "step": 32399 }, { "epoch": 2.5391849529780566, "grad_norm": 6.380195617675781, "learning_rate": 2.9556176743458042e-06, "loss": 1.2005, "step": 32400 }, { "epoch": 2.5392633228840125, "grad_norm": 3.5992705821990967, "learning_rate": 2.9553362701824805e-06, "loss": 1.3573, "step": 32401 }, { "epoch": 2.539341692789969, "grad_norm": 2.776702880859375, "learning_rate": 2.955054873796006e-06, "loss": 1.4009, "step": 32402 }, { "epoch": 2.5394200626959247, "grad_norm": 13.353782653808594, "learning_rate": 2.9547734851874555e-06, "loss": 0.7339, "step": 32403 }, { "epoch": 2.539498432601881, "grad_norm": 2.195317268371582, "learning_rate": 2.954492104357892e-06, "loss": 1.5368, "step": 32404 }, { "epoch": 2.539576802507837, "grad_norm": 3.3996424674987793, "learning_rate": 2.9542107313083925e-06, "loss": 0.9378, "step": 32405 }, { "epoch": 2.5396551724137932, "grad_norm": 2.6476733684539795, "learning_rate": 2.9539293660400246e-06, "loss": 1.0884, "step": 32406 }, { "epoch": 2.539733542319749, "grad_norm": 3.81990909576416, "learning_rate": 2.9536480085538584e-06, "loss": 0.9379, "step": 32407 }, { "epoch": 2.5398119122257055, "grad_norm": 3.856297731399536, "learning_rate": 2.9533666588509635e-06, "loss": 1.0497, "step": 32408 }, { "epoch": 2.5398902821316613, "grad_norm": 4.3453779220581055, "learning_rate": 2.95308531693241e-06, "loss": 1.9121, "step": 32409 }, { "epoch": 2.5399686520376177, "grad_norm": 2.384654998779297, "learning_rate": 2.952803982799271e-06, "loss": 1.2057, "step": 32410 }, { "epoch": 2.5400470219435736, "grad_norm": 2.9782025814056396, "learning_rate": 2.952522656452611e-06, "loss": 1.2671, "step": 32411 }, { "epoch": 2.54012539184953, "grad_norm": 3.5413944721221924, "learning_rate": 2.9522413378935048e-06, "loss": 1.4431, "step": 32412 }, { "epoch": 2.540203761755486, "grad_norm": 3.755141496658325, "learning_rate": 2.9519600271230208e-06, "loss": 1.3076, "step": 32413 }, { "epoch": 2.540282131661442, "grad_norm": 2.2150094509124756, "learning_rate": 2.9516787241422295e-06, "loss": 1.2512, "step": 32414 }, { "epoch": 2.540360501567398, "grad_norm": 3.8647186756134033, "learning_rate": 2.9513974289521995e-06, "loss": 1.7874, "step": 32415 }, { "epoch": 2.5404388714733543, "grad_norm": 8.677983283996582, "learning_rate": 2.951116141554e-06, "loss": 0.9551, "step": 32416 }, { "epoch": 2.5405172413793102, "grad_norm": 6.356128692626953, "learning_rate": 2.950834861948704e-06, "loss": 1.2472, "step": 32417 }, { "epoch": 2.5405956112852666, "grad_norm": 4.187661647796631, "learning_rate": 2.9505535901373804e-06, "loss": 1.4428, "step": 32418 }, { "epoch": 2.5406739811912225, "grad_norm": 2.700932502746582, "learning_rate": 2.9502723261210975e-06, "loss": 1.4711, "step": 32419 }, { "epoch": 2.540752351097179, "grad_norm": 3.213409662246704, "learning_rate": 2.949991069900925e-06, "loss": 1.3158, "step": 32420 }, { "epoch": 2.5408307210031347, "grad_norm": 2.333613157272339, "learning_rate": 2.9497098214779352e-06, "loss": 0.9539, "step": 32421 }, { "epoch": 2.540909090909091, "grad_norm": 2.121931791305542, "learning_rate": 2.949428580853195e-06, "loss": 1.054, "step": 32422 }, { "epoch": 2.540987460815047, "grad_norm": 2.4333302974700928, "learning_rate": 2.9491473480277743e-06, "loss": 1.0355, "step": 32423 }, { "epoch": 2.5410658307210032, "grad_norm": 2.817732572555542, "learning_rate": 2.9488661230027447e-06, "loss": 0.9359, "step": 32424 }, { "epoch": 2.541144200626959, "grad_norm": 2.3896358013153076, "learning_rate": 2.948584905779176e-06, "loss": 0.9419, "step": 32425 }, { "epoch": 2.5412225705329154, "grad_norm": 5.543647766113281, "learning_rate": 2.9483036963581358e-06, "loss": 1.3374, "step": 32426 }, { "epoch": 2.5413009404388713, "grad_norm": 14.36474895477295, "learning_rate": 2.948022494740694e-06, "loss": 0.8186, "step": 32427 }, { "epoch": 2.5413793103448277, "grad_norm": 3.9163427352905273, "learning_rate": 2.94774130092792e-06, "loss": 1.8695, "step": 32428 }, { "epoch": 2.5414576802507836, "grad_norm": 2.4290194511413574, "learning_rate": 2.9474601149208863e-06, "loss": 1.0281, "step": 32429 }, { "epoch": 2.54153605015674, "grad_norm": 3.638915538787842, "learning_rate": 2.9471789367206583e-06, "loss": 1.044, "step": 32430 }, { "epoch": 2.541614420062696, "grad_norm": 8.288287162780762, "learning_rate": 2.9468977663283073e-06, "loss": 1.5214, "step": 32431 }, { "epoch": 2.541692789968652, "grad_norm": 2.254248857498169, "learning_rate": 2.946616603744903e-06, "loss": 1.0768, "step": 32432 }, { "epoch": 2.541771159874608, "grad_norm": 3.282759189605713, "learning_rate": 2.946335448971515e-06, "loss": 1.6504, "step": 32433 }, { "epoch": 2.5418495297805643, "grad_norm": 6.14224100112915, "learning_rate": 2.9460543020092116e-06, "loss": 1.462, "step": 32434 }, { "epoch": 2.54192789968652, "grad_norm": 5.620460033416748, "learning_rate": 2.9457731628590613e-06, "loss": 1.2325, "step": 32435 }, { "epoch": 2.5420062695924766, "grad_norm": 2.8225619792938232, "learning_rate": 2.945492031522137e-06, "loss": 1.6086, "step": 32436 }, { "epoch": 2.5420846394984324, "grad_norm": 3.657853841781616, "learning_rate": 2.945210907999505e-06, "loss": 1.6392, "step": 32437 }, { "epoch": 2.5421630094043888, "grad_norm": 3.3470370769500732, "learning_rate": 2.944929792292235e-06, "loss": 1.2589, "step": 32438 }, { "epoch": 2.5422413793103447, "grad_norm": 4.716880798339844, "learning_rate": 2.9446486844013968e-06, "loss": 1.757, "step": 32439 }, { "epoch": 2.542319749216301, "grad_norm": 4.910208702087402, "learning_rate": 2.94436758432806e-06, "loss": 1.4835, "step": 32440 }, { "epoch": 2.5423981191222573, "grad_norm": 4.848446369171143, "learning_rate": 2.9440864920732925e-06, "loss": 1.2475, "step": 32441 }, { "epoch": 2.542476489028213, "grad_norm": 3.5241177082061768, "learning_rate": 2.943805407638162e-06, "loss": 1.4374, "step": 32442 }, { "epoch": 2.542554858934169, "grad_norm": 2.744666576385498, "learning_rate": 2.9435243310237416e-06, "loss": 1.3424, "step": 32443 }, { "epoch": 2.5426332288401254, "grad_norm": 2.6611783504486084, "learning_rate": 2.943243262231099e-06, "loss": 1.2954, "step": 32444 }, { "epoch": 2.5427115987460818, "grad_norm": 2.206240653991699, "learning_rate": 2.9429622012613016e-06, "loss": 1.4117, "step": 32445 }, { "epoch": 2.5427899686520377, "grad_norm": 3.675441265106201, "learning_rate": 2.9426811481154193e-06, "loss": 1.5055, "step": 32446 }, { "epoch": 2.5428683385579935, "grad_norm": 2.8981311321258545, "learning_rate": 2.9424001027945226e-06, "loss": 0.602, "step": 32447 }, { "epoch": 2.54294670846395, "grad_norm": 3.716783285140991, "learning_rate": 2.9421190652996766e-06, "loss": 1.3311, "step": 32448 }, { "epoch": 2.543025078369906, "grad_norm": 3.364483594894409, "learning_rate": 2.9418380356319543e-06, "loss": 1.6114, "step": 32449 }, { "epoch": 2.543103448275862, "grad_norm": 3.1015615463256836, "learning_rate": 2.9415570137924228e-06, "loss": 1.4703, "step": 32450 }, { "epoch": 2.543181818181818, "grad_norm": 3.5165677070617676, "learning_rate": 2.9412759997821515e-06, "loss": 0.9792, "step": 32451 }, { "epoch": 2.5432601880877743, "grad_norm": 2.5729217529296875, "learning_rate": 2.9409949936022085e-06, "loss": 1.4047, "step": 32452 }, { "epoch": 2.5433385579937307, "grad_norm": 7.396430015563965, "learning_rate": 2.940713995253662e-06, "loss": 1.3804, "step": 32453 }, { "epoch": 2.5434169278996865, "grad_norm": 8.421358108520508, "learning_rate": 2.9404330047375824e-06, "loss": 1.0069, "step": 32454 }, { "epoch": 2.5434952978056424, "grad_norm": 3.566145181655884, "learning_rate": 2.940152022055039e-06, "loss": 1.1141, "step": 32455 }, { "epoch": 2.5435736677115988, "grad_norm": 2.227275848388672, "learning_rate": 2.939871047207098e-06, "loss": 1.2695, "step": 32456 }, { "epoch": 2.543652037617555, "grad_norm": 1.8484740257263184, "learning_rate": 2.9395900801948296e-06, "loss": 1.189, "step": 32457 }, { "epoch": 2.543730407523511, "grad_norm": 3.6469507217407227, "learning_rate": 2.9393091210193025e-06, "loss": 1.2441, "step": 32458 }, { "epoch": 2.543808777429467, "grad_norm": 5.942532539367676, "learning_rate": 2.939028169681586e-06, "loss": 1.41, "step": 32459 }, { "epoch": 2.543887147335423, "grad_norm": 3.328181505203247, "learning_rate": 2.9387472261827456e-06, "loss": 1.0291, "step": 32460 }, { "epoch": 2.5439655172413795, "grad_norm": 2.509871244430542, "learning_rate": 2.9384662905238524e-06, "loss": 1.4572, "step": 32461 }, { "epoch": 2.5440438871473354, "grad_norm": 2.704864740371704, "learning_rate": 2.9381853627059763e-06, "loss": 1.4128, "step": 32462 }, { "epoch": 2.5441222570532913, "grad_norm": 3.2312395572662354, "learning_rate": 2.9379044427301827e-06, "loss": 1.1948, "step": 32463 }, { "epoch": 2.5442006269592476, "grad_norm": 3.018974781036377, "learning_rate": 2.937623530597542e-06, "loss": 1.4812, "step": 32464 }, { "epoch": 2.544278996865204, "grad_norm": 4.413910388946533, "learning_rate": 2.937342626309121e-06, "loss": 1.479, "step": 32465 }, { "epoch": 2.54435736677116, "grad_norm": 5.063536167144775, "learning_rate": 2.937061729865992e-06, "loss": 1.1812, "step": 32466 }, { "epoch": 2.5444357366771158, "grad_norm": 2.72358775138855, "learning_rate": 2.9367808412692174e-06, "loss": 1.2665, "step": 32467 }, { "epoch": 2.544514106583072, "grad_norm": 2.409719467163086, "learning_rate": 2.9364999605198695e-06, "loss": 1.4578, "step": 32468 }, { "epoch": 2.5445924764890284, "grad_norm": 3.3298070430755615, "learning_rate": 2.9362190876190166e-06, "loss": 1.3349, "step": 32469 }, { "epoch": 2.5446708463949843, "grad_norm": 3.7491471767425537, "learning_rate": 2.935938222567727e-06, "loss": 1.261, "step": 32470 }, { "epoch": 2.54474921630094, "grad_norm": 1.7767683267593384, "learning_rate": 2.935657365367067e-06, "loss": 1.1531, "step": 32471 }, { "epoch": 2.5448275862068965, "grad_norm": 3.2298471927642822, "learning_rate": 2.935376516018105e-06, "loss": 1.202, "step": 32472 }, { "epoch": 2.544905956112853, "grad_norm": 3.6640982627868652, "learning_rate": 2.935095674521912e-06, "loss": 1.2451, "step": 32473 }, { "epoch": 2.5449843260188088, "grad_norm": 5.173917293548584, "learning_rate": 2.9348148408795547e-06, "loss": 1.7562, "step": 32474 }, { "epoch": 2.5450626959247646, "grad_norm": 2.7843425273895264, "learning_rate": 2.9345340150921e-06, "loss": 1.1656, "step": 32475 }, { "epoch": 2.545141065830721, "grad_norm": 3.350226879119873, "learning_rate": 2.934253197160617e-06, "loss": 1.2401, "step": 32476 }, { "epoch": 2.5452194357366773, "grad_norm": 1.858778476715088, "learning_rate": 2.933972387086176e-06, "loss": 1.4229, "step": 32477 }, { "epoch": 2.545297805642633, "grad_norm": 4.233622074127197, "learning_rate": 2.9336915848698405e-06, "loss": 1.4226, "step": 32478 }, { "epoch": 2.545376175548589, "grad_norm": 2.793855667114258, "learning_rate": 2.9334107905126804e-06, "loss": 1.1654, "step": 32479 }, { "epoch": 2.5454545454545454, "grad_norm": 4.240675926208496, "learning_rate": 2.933130004015765e-06, "loss": 1.2869, "step": 32480 }, { "epoch": 2.5455329153605017, "grad_norm": 3.0880064964294434, "learning_rate": 2.932849225380162e-06, "loss": 1.1334, "step": 32481 }, { "epoch": 2.5456112852664576, "grad_norm": 2.42348575592041, "learning_rate": 2.9325684546069387e-06, "loss": 0.9293, "step": 32482 }, { "epoch": 2.5456896551724135, "grad_norm": 4.530739784240723, "learning_rate": 2.9322876916971634e-06, "loss": 1.5761, "step": 32483 }, { "epoch": 2.54576802507837, "grad_norm": 11.8912992477417, "learning_rate": 2.932006936651901e-06, "loss": 0.7422, "step": 32484 }, { "epoch": 2.545846394984326, "grad_norm": 6.7392144203186035, "learning_rate": 2.9317261894722266e-06, "loss": 1.1553, "step": 32485 }, { "epoch": 2.545924764890282, "grad_norm": 2.4587228298187256, "learning_rate": 2.9314454501591998e-06, "loss": 1.2288, "step": 32486 }, { "epoch": 2.5460031347962384, "grad_norm": 6.415009021759033, "learning_rate": 2.931164718713893e-06, "loss": 1.671, "step": 32487 }, { "epoch": 2.5460815047021943, "grad_norm": 6.222342491149902, "learning_rate": 2.9308839951373737e-06, "loss": 1.7198, "step": 32488 }, { "epoch": 2.5461598746081506, "grad_norm": 2.097378969192505, "learning_rate": 2.9306032794307087e-06, "loss": 1.2605, "step": 32489 }, { "epoch": 2.5462382445141065, "grad_norm": 3.023087739944458, "learning_rate": 2.9303225715949656e-06, "loss": 1.3487, "step": 32490 }, { "epoch": 2.546316614420063, "grad_norm": 4.174313068389893, "learning_rate": 2.9300418716312118e-06, "loss": 1.466, "step": 32491 }, { "epoch": 2.5463949843260187, "grad_norm": 6.007711410522461, "learning_rate": 2.9297611795405178e-06, "loss": 1.6106, "step": 32492 }, { "epoch": 2.546473354231975, "grad_norm": 3.9055371284484863, "learning_rate": 2.929480495323946e-06, "loss": 1.215, "step": 32493 }, { "epoch": 2.546551724137931, "grad_norm": 2.7361810207366943, "learning_rate": 2.929199818982568e-06, "loss": 1.5417, "step": 32494 }, { "epoch": 2.5466300940438873, "grad_norm": 2.1489078998565674, "learning_rate": 2.928919150517451e-06, "loss": 1.1805, "step": 32495 }, { "epoch": 2.546708463949843, "grad_norm": 4.263345718383789, "learning_rate": 2.9286384899296614e-06, "loss": 1.6318, "step": 32496 }, { "epoch": 2.5467868338557995, "grad_norm": 3.689596652984619, "learning_rate": 2.928357837220267e-06, "loss": 1.4202, "step": 32497 }, { "epoch": 2.5468652037617554, "grad_norm": 3.2745516300201416, "learning_rate": 2.9280771923903338e-06, "loss": 1.5102, "step": 32498 }, { "epoch": 2.5469435736677117, "grad_norm": 2.9077532291412354, "learning_rate": 2.927796555440933e-06, "loss": 1.4761, "step": 32499 }, { "epoch": 2.5470219435736676, "grad_norm": 9.566423416137695, "learning_rate": 2.927515926373129e-06, "loss": 0.826, "step": 32500 }, { "epoch": 2.547100313479624, "grad_norm": 2.8121652603149414, "learning_rate": 2.92723530518799e-06, "loss": 1.1144, "step": 32501 }, { "epoch": 2.54717868338558, "grad_norm": 3.8143155574798584, "learning_rate": 2.926954691886583e-06, "loss": 1.3217, "step": 32502 }, { "epoch": 2.547257053291536, "grad_norm": 1.938988208770752, "learning_rate": 2.9266740864699767e-06, "loss": 0.9706, "step": 32503 }, { "epoch": 2.547335423197492, "grad_norm": 3.1902313232421875, "learning_rate": 2.926393488939236e-06, "loss": 0.9112, "step": 32504 }, { "epoch": 2.5474137931034484, "grad_norm": 3.9380104541778564, "learning_rate": 2.9261128992954287e-06, "loss": 1.6739, "step": 32505 }, { "epoch": 2.5474921630094043, "grad_norm": 3.2476871013641357, "learning_rate": 2.925832317539623e-06, "loss": 1.4771, "step": 32506 }, { "epoch": 2.5475705329153606, "grad_norm": 5.322813510894775, "learning_rate": 2.9255517436728876e-06, "loss": 1.2125, "step": 32507 }, { "epoch": 2.5476489028213165, "grad_norm": 6.306060314178467, "learning_rate": 2.9252711776962866e-06, "loss": 1.0863, "step": 32508 }, { "epoch": 2.547727272727273, "grad_norm": 3.7654688358306885, "learning_rate": 2.9249906196108885e-06, "loss": 1.3733, "step": 32509 }, { "epoch": 2.5478056426332287, "grad_norm": 2.6429781913757324, "learning_rate": 2.924710069417759e-06, "loss": 1.3921, "step": 32510 }, { "epoch": 2.547884012539185, "grad_norm": 6.089688301086426, "learning_rate": 2.92442952711797e-06, "loss": 0.9553, "step": 32511 }, { "epoch": 2.547962382445141, "grad_norm": 2.760373830795288, "learning_rate": 2.9241489927125812e-06, "loss": 1.1971, "step": 32512 }, { "epoch": 2.5480407523510973, "grad_norm": 7.039427757263184, "learning_rate": 2.923868466202665e-06, "loss": 0.7505, "step": 32513 }, { "epoch": 2.548119122257053, "grad_norm": 5.2780842781066895, "learning_rate": 2.923587947589287e-06, "loss": 0.7291, "step": 32514 }, { "epoch": 2.5481974921630095, "grad_norm": 3.242272138595581, "learning_rate": 2.923307436873514e-06, "loss": 1.3421, "step": 32515 }, { "epoch": 2.5482758620689654, "grad_norm": 2.9163413047790527, "learning_rate": 2.9230269340564127e-06, "loss": 1.0201, "step": 32516 }, { "epoch": 2.5483542319749217, "grad_norm": 3.8228707313537598, "learning_rate": 2.9227464391390485e-06, "loss": 1.2635, "step": 32517 }, { "epoch": 2.5484326018808776, "grad_norm": 4.478979110717773, "learning_rate": 2.9224659521224934e-06, "loss": 1.2757, "step": 32518 }, { "epoch": 2.548510971786834, "grad_norm": 4.411590099334717, "learning_rate": 2.9221854730078082e-06, "loss": 1.2791, "step": 32519 }, { "epoch": 2.54858934169279, "grad_norm": 4.753863334655762, "learning_rate": 2.921905001796062e-06, "loss": 1.0574, "step": 32520 }, { "epoch": 2.548667711598746, "grad_norm": 6.21331262588501, "learning_rate": 2.9216245384883225e-06, "loss": 1.583, "step": 32521 }, { "epoch": 2.548746081504702, "grad_norm": 2.825897216796875, "learning_rate": 2.9213440830856564e-06, "loss": 0.9131, "step": 32522 }, { "epoch": 2.5488244514106584, "grad_norm": 4.3675642013549805, "learning_rate": 2.9210636355891294e-06, "loss": 1.3704, "step": 32523 }, { "epoch": 2.5489028213166143, "grad_norm": 3.2809395790100098, "learning_rate": 2.920783195999807e-06, "loss": 1.1512, "step": 32524 }, { "epoch": 2.5489811912225706, "grad_norm": 3.3066136837005615, "learning_rate": 2.920502764318759e-06, "loss": 1.1976, "step": 32525 }, { "epoch": 2.5490595611285265, "grad_norm": 4.114068508148193, "learning_rate": 2.9202223405470505e-06, "loss": 0.6557, "step": 32526 }, { "epoch": 2.549137931034483, "grad_norm": 3.470698356628418, "learning_rate": 2.9199419246857475e-06, "loss": 1.3032, "step": 32527 }, { "epoch": 2.5492163009404387, "grad_norm": 3.1515438556671143, "learning_rate": 2.9196615167359165e-06, "loss": 1.3132, "step": 32528 }, { "epoch": 2.549294670846395, "grad_norm": 7.310710906982422, "learning_rate": 2.919381116698624e-06, "loss": 1.272, "step": 32529 }, { "epoch": 2.549373040752351, "grad_norm": 9.909760475158691, "learning_rate": 2.9191007245749404e-06, "loss": 1.0486, "step": 32530 }, { "epoch": 2.5494514106583073, "grad_norm": 3.9531476497650146, "learning_rate": 2.918820340365926e-06, "loss": 1.3154, "step": 32531 }, { "epoch": 2.549529780564263, "grad_norm": 3.2142691612243652, "learning_rate": 2.9185399640726485e-06, "loss": 1.7507, "step": 32532 }, { "epoch": 2.5496081504702195, "grad_norm": 3.9007492065429688, "learning_rate": 2.91825959569618e-06, "loss": 1.8866, "step": 32533 }, { "epoch": 2.5496865203761754, "grad_norm": 7.968874454498291, "learning_rate": 2.91797923523758e-06, "loss": 0.1683, "step": 32534 }, { "epoch": 2.5497648902821317, "grad_norm": 4.497527599334717, "learning_rate": 2.9176988826979187e-06, "loss": 1.2982, "step": 32535 }, { "epoch": 2.549843260188088, "grad_norm": 6.2031636238098145, "learning_rate": 2.9174185380782607e-06, "loss": 1.448, "step": 32536 }, { "epoch": 2.549921630094044, "grad_norm": 2.482239246368408, "learning_rate": 2.9171382013796736e-06, "loss": 1.701, "step": 32537 }, { "epoch": 2.55, "grad_norm": 2.263780355453491, "learning_rate": 2.9168578726032213e-06, "loss": 0.9768, "step": 32538 }, { "epoch": 2.550078369905956, "grad_norm": 3.534355640411377, "learning_rate": 2.9165775517499733e-06, "loss": 1.5661, "step": 32539 }, { "epoch": 2.5501567398119125, "grad_norm": 2.533421754837036, "learning_rate": 2.9162972388209937e-06, "loss": 1.2929, "step": 32540 }, { "epoch": 2.5502351097178684, "grad_norm": 3.5283408164978027, "learning_rate": 2.9160169338173496e-06, "loss": 1.2461, "step": 32541 }, { "epoch": 2.5503134796238243, "grad_norm": 2.5031509399414062, "learning_rate": 2.915736636740106e-06, "loss": 1.4755, "step": 32542 }, { "epoch": 2.5503918495297806, "grad_norm": 3.6527957916259766, "learning_rate": 2.91545634759033e-06, "loss": 1.4928, "step": 32543 }, { "epoch": 2.550470219435737, "grad_norm": 3.7716195583343506, "learning_rate": 2.9151760663690864e-06, "loss": 0.8781, "step": 32544 }, { "epoch": 2.550548589341693, "grad_norm": 4.057424068450928, "learning_rate": 2.9148957930774456e-06, "loss": 1.5871, "step": 32545 }, { "epoch": 2.5506269592476487, "grad_norm": 3.0847792625427246, "learning_rate": 2.9146155277164655e-06, "loss": 1.4252, "step": 32546 }, { "epoch": 2.550705329153605, "grad_norm": 2.3495354652404785, "learning_rate": 2.9143352702872183e-06, "loss": 1.5508, "step": 32547 }, { "epoch": 2.5507836990595614, "grad_norm": 4.924455165863037, "learning_rate": 2.9140550207907714e-06, "loss": 1.3831, "step": 32548 }, { "epoch": 2.5508620689655173, "grad_norm": 3.398836135864258, "learning_rate": 2.9137747792281856e-06, "loss": 1.4043, "step": 32549 }, { "epoch": 2.550940438871473, "grad_norm": 2.6097421646118164, "learning_rate": 2.913494545600529e-06, "loss": 1.4867, "step": 32550 }, { "epoch": 2.5510188087774295, "grad_norm": 6.862488746643066, "learning_rate": 2.913214319908866e-06, "loss": 1.6868, "step": 32551 }, { "epoch": 2.551097178683386, "grad_norm": 3.5655534267425537, "learning_rate": 2.9129341021542666e-06, "loss": 1.4571, "step": 32552 }, { "epoch": 2.5511755485893417, "grad_norm": 2.3203163146972656, "learning_rate": 2.912653892337792e-06, "loss": 1.1767, "step": 32553 }, { "epoch": 2.5512539184952976, "grad_norm": 3.340008497238159, "learning_rate": 2.912373690460511e-06, "loss": 1.3483, "step": 32554 }, { "epoch": 2.551332288401254, "grad_norm": 4.17193603515625, "learning_rate": 2.9120934965234877e-06, "loss": 1.1302, "step": 32555 }, { "epoch": 2.5514106583072103, "grad_norm": 3.8875341415405273, "learning_rate": 2.911813310527788e-06, "loss": 1.7337, "step": 32556 }, { "epoch": 2.551489028213166, "grad_norm": 7.560990333557129, "learning_rate": 2.9115331324744785e-06, "loss": 1.4734, "step": 32557 }, { "epoch": 2.551567398119122, "grad_norm": 2.5431478023529053, "learning_rate": 2.9112529623646234e-06, "loss": 1.4402, "step": 32558 }, { "epoch": 2.5516457680250784, "grad_norm": 3.465157985687256, "learning_rate": 2.910972800199292e-06, "loss": 1.2917, "step": 32559 }, { "epoch": 2.5517241379310347, "grad_norm": 3.2548024654388428, "learning_rate": 2.9106926459795426e-06, "loss": 1.6298, "step": 32560 }, { "epoch": 2.5518025078369906, "grad_norm": 2.672264575958252, "learning_rate": 2.9104124997064478e-06, "loss": 1.3564, "step": 32561 }, { "epoch": 2.5518808777429465, "grad_norm": 4.790502071380615, "learning_rate": 2.91013236138107e-06, "loss": 1.0548, "step": 32562 }, { "epoch": 2.551959247648903, "grad_norm": 1.9275851249694824, "learning_rate": 2.9098522310044773e-06, "loss": 1.4176, "step": 32563 }, { "epoch": 2.552037617554859, "grad_norm": 3.4580345153808594, "learning_rate": 2.9095721085777316e-06, "loss": 1.2366, "step": 32564 }, { "epoch": 2.552115987460815, "grad_norm": 3.101078987121582, "learning_rate": 2.909291994101897e-06, "loss": 1.2331, "step": 32565 }, { "epoch": 2.552194357366771, "grad_norm": 4.196749687194824, "learning_rate": 2.909011887578045e-06, "loss": 1.394, "step": 32566 }, { "epoch": 2.5522727272727272, "grad_norm": 2.7623493671417236, "learning_rate": 2.9087317890072385e-06, "loss": 1.3127, "step": 32567 }, { "epoch": 2.5523510971786836, "grad_norm": 2.340515375137329, "learning_rate": 2.9084516983905407e-06, "loss": 1.2746, "step": 32568 }, { "epoch": 2.5524294670846395, "grad_norm": 11.517826080322266, "learning_rate": 2.9081716157290192e-06, "loss": 1.3223, "step": 32569 }, { "epoch": 2.5525078369905954, "grad_norm": 2.861269235610962, "learning_rate": 2.907891541023735e-06, "loss": 1.096, "step": 32570 }, { "epoch": 2.5525862068965517, "grad_norm": 2.405921697616577, "learning_rate": 2.907611474275762e-06, "loss": 1.2665, "step": 32571 }, { "epoch": 2.552664576802508, "grad_norm": 2.974137306213379, "learning_rate": 2.9073314154861578e-06, "loss": 1.0371, "step": 32572 }, { "epoch": 2.552742946708464, "grad_norm": 3.215120553970337, "learning_rate": 2.9070513646559896e-06, "loss": 1.2484, "step": 32573 }, { "epoch": 2.55282131661442, "grad_norm": 4.450199127197266, "learning_rate": 2.9067713217863235e-06, "loss": 0.7326, "step": 32574 }, { "epoch": 2.552899686520376, "grad_norm": 2.6853747367858887, "learning_rate": 2.906491286878224e-06, "loss": 1.5864, "step": 32575 }, { "epoch": 2.5529780564263325, "grad_norm": 2.4663922786712646, "learning_rate": 2.906211259932756e-06, "loss": 1.2349, "step": 32576 }, { "epoch": 2.5530564263322884, "grad_norm": 2.458050012588501, "learning_rate": 2.9059312409509853e-06, "loss": 1.3155, "step": 32577 }, { "epoch": 2.5531347962382442, "grad_norm": 6.971612453460693, "learning_rate": 2.905651229933979e-06, "loss": 0.8333, "step": 32578 }, { "epoch": 2.5532131661442006, "grad_norm": 6.272542953491211, "learning_rate": 2.905371226882795e-06, "loss": 1.2779, "step": 32579 }, { "epoch": 2.553291536050157, "grad_norm": 3.8562071323394775, "learning_rate": 2.905091231798505e-06, "loss": 1.6943, "step": 32580 }, { "epoch": 2.553369905956113, "grad_norm": 4.080589771270752, "learning_rate": 2.9048112446821726e-06, "loss": 1.1194, "step": 32581 }, { "epoch": 2.553448275862069, "grad_norm": 2.725776195526123, "learning_rate": 2.9045312655348633e-06, "loss": 1.2165, "step": 32582 }, { "epoch": 2.553526645768025, "grad_norm": 2.9660303592681885, "learning_rate": 2.9042512943576393e-06, "loss": 1.3577, "step": 32583 }, { "epoch": 2.5536050156739813, "grad_norm": 3.033215045928955, "learning_rate": 2.903971331151565e-06, "loss": 1.2427, "step": 32584 }, { "epoch": 2.5536833855799372, "grad_norm": 3.3585691452026367, "learning_rate": 2.903691375917709e-06, "loss": 1.1858, "step": 32585 }, { "epoch": 2.5537617554858936, "grad_norm": 2.929226875305176, "learning_rate": 2.9034114286571358e-06, "loss": 1.2461, "step": 32586 }, { "epoch": 2.5538401253918495, "grad_norm": 5.306393146514893, "learning_rate": 2.9031314893709072e-06, "loss": 0.8779, "step": 32587 }, { "epoch": 2.553918495297806, "grad_norm": 3.8816840648651123, "learning_rate": 2.902851558060089e-06, "loss": 1.3576, "step": 32588 }, { "epoch": 2.5539968652037617, "grad_norm": 3.36283540725708, "learning_rate": 2.902571634725747e-06, "loss": 1.0189, "step": 32589 }, { "epoch": 2.554075235109718, "grad_norm": 2.8200056552886963, "learning_rate": 2.902291719368945e-06, "loss": 1.3422, "step": 32590 }, { "epoch": 2.554153605015674, "grad_norm": 3.9393043518066406, "learning_rate": 2.902011811990747e-06, "loss": 1.4604, "step": 32591 }, { "epoch": 2.5542319749216302, "grad_norm": 2.874521017074585, "learning_rate": 2.9017319125922194e-06, "loss": 1.1318, "step": 32592 }, { "epoch": 2.554310344827586, "grad_norm": 3.00042986869812, "learning_rate": 2.9014520211744253e-06, "loss": 1.345, "step": 32593 }, { "epoch": 2.5543887147335425, "grad_norm": 3.099439859390259, "learning_rate": 2.90117213773843e-06, "loss": 1.422, "step": 32594 }, { "epoch": 2.5544670846394983, "grad_norm": 2.6236937046051025, "learning_rate": 2.9008922622852974e-06, "loss": 1.3966, "step": 32595 }, { "epoch": 2.5545454545454547, "grad_norm": 7.749087333679199, "learning_rate": 2.9006123948160935e-06, "loss": 0.9492, "step": 32596 }, { "epoch": 2.5546238244514106, "grad_norm": 1.893588662147522, "learning_rate": 2.900332535331881e-06, "loss": 0.8987, "step": 32597 }, { "epoch": 2.554702194357367, "grad_norm": 3.1695759296417236, "learning_rate": 2.900052683833725e-06, "loss": 1.3276, "step": 32598 }, { "epoch": 2.554780564263323, "grad_norm": 4.405057430267334, "learning_rate": 2.899772840322691e-06, "loss": 1.455, "step": 32599 }, { "epoch": 2.554858934169279, "grad_norm": 3.574550151824951, "learning_rate": 2.8994930047998413e-06, "loss": 1.326, "step": 32600 }, { "epoch": 2.554937304075235, "grad_norm": 3.79620099067688, "learning_rate": 2.8992131772662436e-06, "loss": 1.1256, "step": 32601 }, { "epoch": 2.5550156739811913, "grad_norm": 4.379934787750244, "learning_rate": 2.8989333577229563e-06, "loss": 0.9318, "step": 32602 }, { "epoch": 2.5550940438871472, "grad_norm": 2.897097587585449, "learning_rate": 2.8986535461710487e-06, "loss": 1.6027, "step": 32603 }, { "epoch": 2.5551724137931036, "grad_norm": 4.000670909881592, "learning_rate": 2.8983737426115865e-06, "loss": 1.0239, "step": 32604 }, { "epoch": 2.5552507836990594, "grad_norm": 2.6159684658050537, "learning_rate": 2.8980939470456284e-06, "loss": 1.4885, "step": 32605 }, { "epoch": 2.555329153605016, "grad_norm": 4.206065654754639, "learning_rate": 2.8978141594742413e-06, "loss": 1.1924, "step": 32606 }, { "epoch": 2.5554075235109717, "grad_norm": 4.590467929840088, "learning_rate": 2.89753437989849e-06, "loss": 1.5227, "step": 32607 }, { "epoch": 2.555485893416928, "grad_norm": 3.552490234375, "learning_rate": 2.8972546083194376e-06, "loss": 1.0566, "step": 32608 }, { "epoch": 2.555564263322884, "grad_norm": 2.7704110145568848, "learning_rate": 2.8969748447381483e-06, "loss": 1.1984, "step": 32609 }, { "epoch": 2.55564263322884, "grad_norm": 3.110877275466919, "learning_rate": 2.8966950891556872e-06, "loss": 1.524, "step": 32610 }, { "epoch": 2.555721003134796, "grad_norm": 2.020176410675049, "learning_rate": 2.8964153415731177e-06, "loss": 1.063, "step": 32611 }, { "epoch": 2.5557993730407524, "grad_norm": 4.415172576904297, "learning_rate": 2.8961356019915033e-06, "loss": 1.4543, "step": 32612 }, { "epoch": 2.5558777429467083, "grad_norm": 3.5137321949005127, "learning_rate": 2.895855870411909e-06, "loss": 1.0787, "step": 32613 }, { "epoch": 2.5559561128526647, "grad_norm": 2.796271324157715, "learning_rate": 2.895576146835397e-06, "loss": 1.2408, "step": 32614 }, { "epoch": 2.5560344827586206, "grad_norm": 4.433777332305908, "learning_rate": 2.8952964312630354e-06, "loss": 1.147, "step": 32615 }, { "epoch": 2.556112852664577, "grad_norm": 2.4255294799804688, "learning_rate": 2.8950167236958816e-06, "loss": 1.2648, "step": 32616 }, { "epoch": 2.5561912225705328, "grad_norm": 3.7657434940338135, "learning_rate": 2.8947370241350035e-06, "loss": 1.3458, "step": 32617 }, { "epoch": 2.556269592476489, "grad_norm": 5.411280155181885, "learning_rate": 2.8944573325814655e-06, "loss": 1.47, "step": 32618 }, { "epoch": 2.556347962382445, "grad_norm": 3.439419746398926, "learning_rate": 2.8941776490363325e-06, "loss": 1.5591, "step": 32619 }, { "epoch": 2.5564263322884013, "grad_norm": 6.491252422332764, "learning_rate": 2.8938979735006635e-06, "loss": 1.5338, "step": 32620 }, { "epoch": 2.556504702194357, "grad_norm": 6.627764701843262, "learning_rate": 2.893618305975523e-06, "loss": 0.9236, "step": 32621 }, { "epoch": 2.5565830721003135, "grad_norm": 2.4930977821350098, "learning_rate": 2.893338646461979e-06, "loss": 1.5076, "step": 32622 }, { "epoch": 2.5566614420062694, "grad_norm": 2.5677781105041504, "learning_rate": 2.8930589949610944e-06, "loss": 0.7587, "step": 32623 }, { "epoch": 2.5567398119122258, "grad_norm": 2.905656576156616, "learning_rate": 2.8927793514739287e-06, "loss": 1.449, "step": 32624 }, { "epoch": 2.5568181818181817, "grad_norm": 2.3987698554992676, "learning_rate": 2.892499716001549e-06, "loss": 1.4543, "step": 32625 }, { "epoch": 2.556896551724138, "grad_norm": 4.54376745223999, "learning_rate": 2.8922200885450154e-06, "loss": 1.5751, "step": 32626 }, { "epoch": 2.556974921630094, "grad_norm": 3.647826671600342, "learning_rate": 2.8919404691053975e-06, "loss": 1.2792, "step": 32627 }, { "epoch": 2.55705329153605, "grad_norm": 4.998879432678223, "learning_rate": 2.891660857683754e-06, "loss": 1.5382, "step": 32628 }, { "epoch": 2.557131661442006, "grad_norm": 4.3304314613342285, "learning_rate": 2.8913812542811492e-06, "loss": 1.3897, "step": 32629 }, { "epoch": 2.5572100313479624, "grad_norm": 6.452675819396973, "learning_rate": 2.8911016588986473e-06, "loss": 1.209, "step": 32630 }, { "epoch": 2.5572884012539188, "grad_norm": 3.7321910858154297, "learning_rate": 2.8908220715373116e-06, "loss": 1.1347, "step": 32631 }, { "epoch": 2.5573667711598747, "grad_norm": 4.150782108306885, "learning_rate": 2.890542492198205e-06, "loss": 1.3101, "step": 32632 }, { "epoch": 2.5574451410658305, "grad_norm": 8.441834449768066, "learning_rate": 2.8902629208823918e-06, "loss": 1.0482, "step": 32633 }, { "epoch": 2.557523510971787, "grad_norm": 9.199438095092773, "learning_rate": 2.8899833575909363e-06, "loss": 1.7897, "step": 32634 }, { "epoch": 2.557601880877743, "grad_norm": 6.008922576904297, "learning_rate": 2.889703802324897e-06, "loss": 1.3609, "step": 32635 }, { "epoch": 2.557680250783699, "grad_norm": 8.646493911743164, "learning_rate": 2.8894242550853418e-06, "loss": 1.4947, "step": 32636 }, { "epoch": 2.557758620689655, "grad_norm": 2.696646213531494, "learning_rate": 2.8891447158733334e-06, "loss": 1.273, "step": 32637 }, { "epoch": 2.5578369905956113, "grad_norm": 3.225402593612671, "learning_rate": 2.8888651846899358e-06, "loss": 1.2496, "step": 32638 }, { "epoch": 2.5579153605015676, "grad_norm": 3.42984676361084, "learning_rate": 2.8885856615362084e-06, "loss": 0.7461, "step": 32639 }, { "epoch": 2.5579937304075235, "grad_norm": 2.6297876834869385, "learning_rate": 2.888306146413215e-06, "loss": 0.8717, "step": 32640 }, { "epoch": 2.5580721003134794, "grad_norm": 4.894289016723633, "learning_rate": 2.8880266393220223e-06, "loss": 1.7754, "step": 32641 }, { "epoch": 2.5581504702194358, "grad_norm": 2.799638509750366, "learning_rate": 2.8877471402636934e-06, "loss": 1.4288, "step": 32642 }, { "epoch": 2.558228840125392, "grad_norm": 2.7815299034118652, "learning_rate": 2.8874676492392882e-06, "loss": 1.2011, "step": 32643 }, { "epoch": 2.558307210031348, "grad_norm": 7.133529186248779, "learning_rate": 2.88718816624987e-06, "loss": 0.96, "step": 32644 }, { "epoch": 2.558385579937304, "grad_norm": 3.8840489387512207, "learning_rate": 2.886908691296504e-06, "loss": 1.2032, "step": 32645 }, { "epoch": 2.55846394984326, "grad_norm": 2.3813114166259766, "learning_rate": 2.8866292243802506e-06, "loss": 0.9475, "step": 32646 }, { "epoch": 2.5585423197492165, "grad_norm": 10.111154556274414, "learning_rate": 2.886349765502175e-06, "loss": 0.8758, "step": 32647 }, { "epoch": 2.5586206896551724, "grad_norm": 3.511746883392334, "learning_rate": 2.886070314663339e-06, "loss": 1.374, "step": 32648 }, { "epoch": 2.5586990595611283, "grad_norm": 3.0387332439422607, "learning_rate": 2.885790871864806e-06, "loss": 1.1345, "step": 32649 }, { "epoch": 2.5587774294670846, "grad_norm": 3.388195753097534, "learning_rate": 2.885511437107638e-06, "loss": 1.353, "step": 32650 }, { "epoch": 2.558855799373041, "grad_norm": 2.892188549041748, "learning_rate": 2.885232010392899e-06, "loss": 1.5135, "step": 32651 }, { "epoch": 2.558934169278997, "grad_norm": 6.271298885345459, "learning_rate": 2.884952591721651e-06, "loss": 1.1161, "step": 32652 }, { "epoch": 2.5590125391849528, "grad_norm": 2.113004446029663, "learning_rate": 2.8846731810949593e-06, "loss": 1.2225, "step": 32653 }, { "epoch": 2.559090909090909, "grad_norm": 3.8717474937438965, "learning_rate": 2.88439377851388e-06, "loss": 1.4921, "step": 32654 }, { "epoch": 2.5591692789968654, "grad_norm": 2.7651278972625732, "learning_rate": 2.8841143839794816e-06, "loss": 1.3397, "step": 32655 }, { "epoch": 2.5592476489028213, "grad_norm": 6.128389835357666, "learning_rate": 2.883834997492828e-06, "loss": 1.4182, "step": 32656 }, { "epoch": 2.559326018808777, "grad_norm": 2.870504140853882, "learning_rate": 2.8835556190549774e-06, "loss": 1.2231, "step": 32657 }, { "epoch": 2.5594043887147335, "grad_norm": 3.754672050476074, "learning_rate": 2.883276248666993e-06, "loss": 1.3289, "step": 32658 }, { "epoch": 2.55948275862069, "grad_norm": 2.667088747024536, "learning_rate": 2.8829968863299372e-06, "loss": 1.3402, "step": 32659 }, { "epoch": 2.5595611285266457, "grad_norm": 4.1569037437438965, "learning_rate": 2.8827175320448788e-06, "loss": 1.1571, "step": 32660 }, { "epoch": 2.5596394984326016, "grad_norm": 6.11207389831543, "learning_rate": 2.8824381858128717e-06, "loss": 1.297, "step": 32661 }, { "epoch": 2.559717868338558, "grad_norm": 6.560230731964111, "learning_rate": 2.8821588476349834e-06, "loss": 1.4708, "step": 32662 }, { "epoch": 2.5597962382445143, "grad_norm": 3.8383724689483643, "learning_rate": 2.8818795175122745e-06, "loss": 0.71, "step": 32663 }, { "epoch": 2.55987460815047, "grad_norm": 3.5032856464385986, "learning_rate": 2.8816001954458083e-06, "loss": 1.7709, "step": 32664 }, { "epoch": 2.559952978056426, "grad_norm": 2.4329466819763184, "learning_rate": 2.881320881436647e-06, "loss": 1.3436, "step": 32665 }, { "epoch": 2.5600313479623824, "grad_norm": 2.5323262214660645, "learning_rate": 2.881041575485852e-06, "loss": 1.5524, "step": 32666 }, { "epoch": 2.5601097178683387, "grad_norm": 2.4806673526763916, "learning_rate": 2.8807622775944866e-06, "loss": 1.5019, "step": 32667 }, { "epoch": 2.5601880877742946, "grad_norm": 4.428184986114502, "learning_rate": 2.8804829877636133e-06, "loss": 1.3972, "step": 32668 }, { "epoch": 2.5602664576802505, "grad_norm": 4.6888861656188965, "learning_rate": 2.8802037059942945e-06, "loss": 1.1725, "step": 32669 }, { "epoch": 2.560344827586207, "grad_norm": 2.5810158252716064, "learning_rate": 2.8799244322875912e-06, "loss": 1.2626, "step": 32670 }, { "epoch": 2.560423197492163, "grad_norm": 2.743220329284668, "learning_rate": 2.879645166644569e-06, "loss": 1.162, "step": 32671 }, { "epoch": 2.560501567398119, "grad_norm": 2.2943272590637207, "learning_rate": 2.8793659090662852e-06, "loss": 1.421, "step": 32672 }, { "epoch": 2.560579937304075, "grad_norm": 2.885242223739624, "learning_rate": 2.8790866595538025e-06, "loss": 1.107, "step": 32673 }, { "epoch": 2.5606583072100313, "grad_norm": 4.21483039855957, "learning_rate": 2.878807418108187e-06, "loss": 1.3468, "step": 32674 }, { "epoch": 2.5607366771159876, "grad_norm": 6.224087715148926, "learning_rate": 2.8785281847305e-06, "loss": 1.6426, "step": 32675 }, { "epoch": 2.5608150470219435, "grad_norm": 2.5673437118530273, "learning_rate": 2.8782489594218005e-06, "loss": 1.1694, "step": 32676 }, { "epoch": 2.5608934169279, "grad_norm": 3.9496304988861084, "learning_rate": 2.877969742183152e-06, "loss": 1.2361, "step": 32677 }, { "epoch": 2.5609717868338557, "grad_norm": 2.045605182647705, "learning_rate": 2.8776905330156145e-06, "loss": 1.2525, "step": 32678 }, { "epoch": 2.561050156739812, "grad_norm": 3.764816999435425, "learning_rate": 2.8774113319202567e-06, "loss": 1.7016, "step": 32679 }, { "epoch": 2.561128526645768, "grad_norm": 4.169378280639648, "learning_rate": 2.8771321388981334e-06, "loss": 1.1362, "step": 32680 }, { "epoch": 2.5612068965517243, "grad_norm": 4.5994553565979, "learning_rate": 2.8768529539503094e-06, "loss": 0.9237, "step": 32681 }, { "epoch": 2.56128526645768, "grad_norm": 2.7587170600891113, "learning_rate": 2.8765737770778455e-06, "loss": 1.5261, "step": 32682 }, { "epoch": 2.5613636363636365, "grad_norm": 3.973360300064087, "learning_rate": 2.8762946082818056e-06, "loss": 1.5006, "step": 32683 }, { "epoch": 2.5614420062695924, "grad_norm": 3.8940412998199463, "learning_rate": 2.876015447563249e-06, "loss": 0.7572, "step": 32684 }, { "epoch": 2.5615203761755487, "grad_norm": 5.643977642059326, "learning_rate": 2.8757362949232392e-06, "loss": 2.0036, "step": 32685 }, { "epoch": 2.5615987460815046, "grad_norm": 2.3257548809051514, "learning_rate": 2.8754571503628393e-06, "loss": 1.5452, "step": 32686 }, { "epoch": 2.561677115987461, "grad_norm": 2.5951364040374756, "learning_rate": 2.8751780138831053e-06, "loss": 1.5904, "step": 32687 }, { "epoch": 2.561755485893417, "grad_norm": 3.2404415607452393, "learning_rate": 2.8748988854851045e-06, "loss": 1.3105, "step": 32688 }, { "epoch": 2.561833855799373, "grad_norm": 2.444988965988159, "learning_rate": 2.8746197651698974e-06, "loss": 1.2109, "step": 32689 }, { "epoch": 2.561912225705329, "grad_norm": 3.1039979457855225, "learning_rate": 2.8743406529385466e-06, "loss": 1.066, "step": 32690 }, { "epoch": 2.5619905956112854, "grad_norm": 4.609155178070068, "learning_rate": 2.874061548792111e-06, "loss": 1.104, "step": 32691 }, { "epoch": 2.5620689655172413, "grad_norm": 8.658841133117676, "learning_rate": 2.87378245273165e-06, "loss": 1.1331, "step": 32692 }, { "epoch": 2.5621473354231976, "grad_norm": 3.2944085597991943, "learning_rate": 2.8735033647582313e-06, "loss": 1.4447, "step": 32693 }, { "epoch": 2.5622257053291535, "grad_norm": 2.4119861125946045, "learning_rate": 2.873224284872915e-06, "loss": 0.7433, "step": 32694 }, { "epoch": 2.56230407523511, "grad_norm": 3.211812973022461, "learning_rate": 2.87294521307676e-06, "loss": 1.0853, "step": 32695 }, { "epoch": 2.5623824451410657, "grad_norm": 3.124039888381958, "learning_rate": 2.8726661493708276e-06, "loss": 1.3098, "step": 32696 }, { "epoch": 2.562460815047022, "grad_norm": 5.89728307723999, "learning_rate": 2.8723870937561792e-06, "loss": 1.2992, "step": 32697 }, { "epoch": 2.562539184952978, "grad_norm": 6.071439266204834, "learning_rate": 2.8721080462338816e-06, "loss": 0.9901, "step": 32698 }, { "epoch": 2.5626175548589343, "grad_norm": 2.90455961227417, "learning_rate": 2.87182900680499e-06, "loss": 1.511, "step": 32699 }, { "epoch": 2.56269592476489, "grad_norm": 3.266892433166504, "learning_rate": 2.871549975470568e-06, "loss": 1.4076, "step": 32700 }, { "epoch": 2.5627742946708465, "grad_norm": 2.7172164916992188, "learning_rate": 2.8712709522316762e-06, "loss": 1.5218, "step": 32701 }, { "epoch": 2.5628526645768024, "grad_norm": 12.095536231994629, "learning_rate": 2.8709919370893764e-06, "loss": 1.4937, "step": 32702 }, { "epoch": 2.5629310344827587, "grad_norm": 3.5824697017669678, "learning_rate": 2.87071293004473e-06, "loss": 0.852, "step": 32703 }, { "epoch": 2.5630094043887146, "grad_norm": 3.2903804779052734, "learning_rate": 2.870433931098797e-06, "loss": 1.4745, "step": 32704 }, { "epoch": 2.563087774294671, "grad_norm": 4.516034126281738, "learning_rate": 2.870154940252641e-06, "loss": 1.4057, "step": 32705 }, { "epoch": 2.563166144200627, "grad_norm": 2.5178139209747314, "learning_rate": 2.869875957507321e-06, "loss": 1.3266, "step": 32706 }, { "epoch": 2.563244514106583, "grad_norm": 2.485743284225464, "learning_rate": 2.869596982863898e-06, "loss": 1.3526, "step": 32707 }, { "epoch": 2.563322884012539, "grad_norm": 4.095168113708496, "learning_rate": 2.869318016323435e-06, "loss": 1.4101, "step": 32708 }, { "epoch": 2.5634012539184954, "grad_norm": 2.6887383460998535, "learning_rate": 2.869039057886993e-06, "loss": 1.5266, "step": 32709 }, { "epoch": 2.5634796238244513, "grad_norm": 3.5451908111572266, "learning_rate": 2.868760107555628e-06, "loss": 1.0181, "step": 32710 }, { "epoch": 2.5635579937304076, "grad_norm": 2.5015745162963867, "learning_rate": 2.868481165330408e-06, "loss": 1.5568, "step": 32711 }, { "epoch": 2.5636363636363635, "grad_norm": 2.103638172149658, "learning_rate": 2.8682022312123915e-06, "loss": 1.2872, "step": 32712 }, { "epoch": 2.56371473354232, "grad_norm": 2.664984703063965, "learning_rate": 2.8679233052026376e-06, "loss": 1.2211, "step": 32713 }, { "epoch": 2.5637931034482757, "grad_norm": 2.647099256515503, "learning_rate": 2.8676443873022074e-06, "loss": 1.5542, "step": 32714 }, { "epoch": 2.563871473354232, "grad_norm": 4.097370147705078, "learning_rate": 2.8673654775121617e-06, "loss": 1.3985, "step": 32715 }, { "epoch": 2.563949843260188, "grad_norm": 3.6233391761779785, "learning_rate": 2.8670865758335657e-06, "loss": 1.6362, "step": 32716 }, { "epoch": 2.5640282131661443, "grad_norm": 3.8058393001556396, "learning_rate": 2.866807682267475e-06, "loss": 0.8288, "step": 32717 }, { "epoch": 2.5641065830721, "grad_norm": 4.547895908355713, "learning_rate": 2.8665287968149523e-06, "loss": 1.5975, "step": 32718 }, { "epoch": 2.5641849529780565, "grad_norm": 3.9949376583099365, "learning_rate": 2.866249919477059e-06, "loss": 1.3376, "step": 32719 }, { "epoch": 2.5642633228840124, "grad_norm": 4.424797058105469, "learning_rate": 2.865971050254853e-06, "loss": 0.8673, "step": 32720 }, { "epoch": 2.5643416927899687, "grad_norm": 5.310250282287598, "learning_rate": 2.8656921891493987e-06, "loss": 1.0709, "step": 32721 }, { "epoch": 2.564420062695925, "grad_norm": 7.41273832321167, "learning_rate": 2.8654133361617547e-06, "loss": 1.6078, "step": 32722 }, { "epoch": 2.564498432601881, "grad_norm": 3.0193889141082764, "learning_rate": 2.8651344912929825e-06, "loss": 1.2589, "step": 32723 }, { "epoch": 2.564576802507837, "grad_norm": 4.847040176391602, "learning_rate": 2.864855654544141e-06, "loss": 1.8488, "step": 32724 }, { "epoch": 2.564655172413793, "grad_norm": 2.681243419647217, "learning_rate": 2.8645768259162925e-06, "loss": 1.6786, "step": 32725 }, { "epoch": 2.5647335423197495, "grad_norm": 3.962601900100708, "learning_rate": 2.864298005410497e-06, "loss": 1.3989, "step": 32726 }, { "epoch": 2.5648119122257054, "grad_norm": 4.341956615447998, "learning_rate": 2.8640191930278166e-06, "loss": 1.2781, "step": 32727 }, { "epoch": 2.5648902821316613, "grad_norm": 2.9109373092651367, "learning_rate": 2.863740388769308e-06, "loss": 1.6056, "step": 32728 }, { "epoch": 2.5649686520376176, "grad_norm": 6.677048683166504, "learning_rate": 2.863461592636032e-06, "loss": 0.776, "step": 32729 }, { "epoch": 2.565047021943574, "grad_norm": 4.238286972045898, "learning_rate": 2.863182804629053e-06, "loss": 1.6221, "step": 32730 }, { "epoch": 2.56512539184953, "grad_norm": 2.6070356369018555, "learning_rate": 2.862904024749431e-06, "loss": 1.6633, "step": 32731 }, { "epoch": 2.5652037617554857, "grad_norm": 4.699334621429443, "learning_rate": 2.8626252529982224e-06, "loss": 0.9744, "step": 32732 }, { "epoch": 2.565282131661442, "grad_norm": 2.0085079669952393, "learning_rate": 2.8623464893764884e-06, "loss": 1.2723, "step": 32733 }, { "epoch": 2.5653605015673984, "grad_norm": 3.459855556488037, "learning_rate": 2.86206773388529e-06, "loss": 1.3657, "step": 32734 }, { "epoch": 2.5654388714733543, "grad_norm": 2.3394525051116943, "learning_rate": 2.861788986525691e-06, "loss": 1.8687, "step": 32735 }, { "epoch": 2.56551724137931, "grad_norm": 3.2524590492248535, "learning_rate": 2.8615102472987465e-06, "loss": 1.4147, "step": 32736 }, { "epoch": 2.5655956112852665, "grad_norm": 3.343410015106201, "learning_rate": 2.8612315162055188e-06, "loss": 1.2833, "step": 32737 }, { "epoch": 2.565673981191223, "grad_norm": 2.729154586791992, "learning_rate": 2.8609527932470676e-06, "loss": 1.7158, "step": 32738 }, { "epoch": 2.5657523510971787, "grad_norm": 3.365817070007324, "learning_rate": 2.8606740784244535e-06, "loss": 1.2601, "step": 32739 }, { "epoch": 2.5658307210031346, "grad_norm": 7.78331184387207, "learning_rate": 2.860395371738736e-06, "loss": 0.677, "step": 32740 }, { "epoch": 2.565909090909091, "grad_norm": 7.092960834503174, "learning_rate": 2.8601166731909756e-06, "loss": 0.8267, "step": 32741 }, { "epoch": 2.5659874608150472, "grad_norm": 2.6957571506500244, "learning_rate": 2.8598379827822354e-06, "loss": 1.4441, "step": 32742 }, { "epoch": 2.566065830721003, "grad_norm": 3.490142583847046, "learning_rate": 2.859559300513567e-06, "loss": 1.4356, "step": 32743 }, { "epoch": 2.566144200626959, "grad_norm": 3.289121389389038, "learning_rate": 2.859280626386038e-06, "loss": 1.3222, "step": 32744 }, { "epoch": 2.5662225705329154, "grad_norm": 3.0970962047576904, "learning_rate": 2.859001960400706e-06, "loss": 1.8644, "step": 32745 }, { "epoch": 2.5663009404388717, "grad_norm": 4.7907843589782715, "learning_rate": 2.8587233025586325e-06, "loss": 1.5215, "step": 32746 }, { "epoch": 2.5663793103448276, "grad_norm": 3.282742500305176, "learning_rate": 2.8584446528608745e-06, "loss": 1.1842, "step": 32747 }, { "epoch": 2.5664576802507835, "grad_norm": 2.4496498107910156, "learning_rate": 2.8581660113084903e-06, "loss": 1.2421, "step": 32748 }, { "epoch": 2.56653605015674, "grad_norm": 2.5675833225250244, "learning_rate": 2.857887377902545e-06, "loss": 1.381, "step": 32749 }, { "epoch": 2.566614420062696, "grad_norm": 4.565312385559082, "learning_rate": 2.8576087526440978e-06, "loss": 1.2047, "step": 32750 }, { "epoch": 2.566692789968652, "grad_norm": 4.009696006774902, "learning_rate": 2.857330135534204e-06, "loss": 1.1305, "step": 32751 }, { "epoch": 2.566771159874608, "grad_norm": 3.048546314239502, "learning_rate": 2.857051526573926e-06, "loss": 1.1741, "step": 32752 }, { "epoch": 2.5668495297805642, "grad_norm": 7.039844036102295, "learning_rate": 2.8567729257643217e-06, "loss": 2.0949, "step": 32753 }, { "epoch": 2.5669278996865206, "grad_norm": 2.819948196411133, "learning_rate": 2.856494333106456e-06, "loss": 1.3414, "step": 32754 }, { "epoch": 2.5670062695924765, "grad_norm": 3.9589924812316895, "learning_rate": 2.856215748601383e-06, "loss": 1.392, "step": 32755 }, { "epoch": 2.5670846394984324, "grad_norm": 2.85060977935791, "learning_rate": 2.8559371722501638e-06, "loss": 1.1202, "step": 32756 }, { "epoch": 2.5671630094043887, "grad_norm": 3.237948417663574, "learning_rate": 2.8556586040538592e-06, "loss": 1.6075, "step": 32757 }, { "epoch": 2.567241379310345, "grad_norm": 5.343872547149658, "learning_rate": 2.855380044013527e-06, "loss": 1.8082, "step": 32758 }, { "epoch": 2.567319749216301, "grad_norm": 2.830472946166992, "learning_rate": 2.855101492130228e-06, "loss": 1.1694, "step": 32759 }, { "epoch": 2.567398119122257, "grad_norm": 3.5877602100372314, "learning_rate": 2.8548229484050215e-06, "loss": 1.7552, "step": 32760 }, { "epoch": 2.567476489028213, "grad_norm": 4.2258148193359375, "learning_rate": 2.854544412838968e-06, "loss": 1.291, "step": 32761 }, { "epoch": 2.5675548589341695, "grad_norm": 2.490968942642212, "learning_rate": 2.8542658854331216e-06, "loss": 1.4691, "step": 32762 }, { "epoch": 2.5676332288401253, "grad_norm": 2.7410776615142822, "learning_rate": 2.8539873661885485e-06, "loss": 1.1228, "step": 32763 }, { "epoch": 2.5677115987460812, "grad_norm": 3.3781747817993164, "learning_rate": 2.8537088551063047e-06, "loss": 1.3007, "step": 32764 }, { "epoch": 2.5677899686520376, "grad_norm": 7.37841272354126, "learning_rate": 2.853430352187452e-06, "loss": 0.9945, "step": 32765 }, { "epoch": 2.567868338557994, "grad_norm": 2.7687368392944336, "learning_rate": 2.853151857433046e-06, "loss": 1.1099, "step": 32766 }, { "epoch": 2.56794670846395, "grad_norm": 3.625957489013672, "learning_rate": 2.8528733708441446e-06, "loss": 1.4084, "step": 32767 }, { "epoch": 2.568025078369906, "grad_norm": 7.866910457611084, "learning_rate": 2.852594892421815e-06, "loss": 1.3885, "step": 32768 }, { "epoch": 2.568103448275862, "grad_norm": 2.925554037094116, "learning_rate": 2.8523164221671097e-06, "loss": 1.6343, "step": 32769 }, { "epoch": 2.5681818181818183, "grad_norm": 2.8858611583709717, "learning_rate": 2.8520379600810886e-06, "loss": 1.6637, "step": 32770 }, { "epoch": 2.5682601880877742, "grad_norm": 2.5495071411132812, "learning_rate": 2.851759506164813e-06, "loss": 1.4895, "step": 32771 }, { "epoch": 2.5683385579937306, "grad_norm": 8.173087120056152, "learning_rate": 2.85148106041934e-06, "loss": 1.2048, "step": 32772 }, { "epoch": 2.5684169278996865, "grad_norm": 2.5400922298431396, "learning_rate": 2.8512026228457297e-06, "loss": 1.2704, "step": 32773 }, { "epoch": 2.568495297805643, "grad_norm": 2.5410146713256836, "learning_rate": 2.8509241934450405e-06, "loss": 1.5965, "step": 32774 }, { "epoch": 2.5685736677115987, "grad_norm": 2.8360164165496826, "learning_rate": 2.8506457722183324e-06, "loss": 1.3196, "step": 32775 }, { "epoch": 2.568652037617555, "grad_norm": 3.4562549591064453, "learning_rate": 2.8503673591666636e-06, "loss": 1.2974, "step": 32776 }, { "epoch": 2.568730407523511, "grad_norm": 8.544554710388184, "learning_rate": 2.850088954291094e-06, "loss": 1.3302, "step": 32777 }, { "epoch": 2.5688087774294672, "grad_norm": 2.5456724166870117, "learning_rate": 2.8498105575926803e-06, "loss": 1.3081, "step": 32778 }, { "epoch": 2.568887147335423, "grad_norm": 9.744917869567871, "learning_rate": 2.849532169072484e-06, "loss": 2.2361, "step": 32779 }, { "epoch": 2.5689655172413794, "grad_norm": 5.931995391845703, "learning_rate": 2.849253788731564e-06, "loss": 0.6128, "step": 32780 }, { "epoch": 2.5690438871473353, "grad_norm": 3.7125422954559326, "learning_rate": 2.8489754165709747e-06, "loss": 1.0094, "step": 32781 }, { "epoch": 2.5691222570532917, "grad_norm": 2.6858696937561035, "learning_rate": 2.8486970525917794e-06, "loss": 1.1706, "step": 32782 }, { "epoch": 2.5692006269592476, "grad_norm": 2.6331875324249268, "learning_rate": 2.848418696795038e-06, "loss": 1.341, "step": 32783 }, { "epoch": 2.569278996865204, "grad_norm": 2.415243148803711, "learning_rate": 2.848140349181805e-06, "loss": 1.2237, "step": 32784 }, { "epoch": 2.56935736677116, "grad_norm": 4.2380571365356445, "learning_rate": 2.8478620097531402e-06, "loss": 1.2298, "step": 32785 }, { "epoch": 2.569435736677116, "grad_norm": 1.938812017440796, "learning_rate": 2.847583678510101e-06, "loss": 1.2908, "step": 32786 }, { "epoch": 2.569514106583072, "grad_norm": 4.258813858032227, "learning_rate": 2.847305355453752e-06, "loss": 1.5649, "step": 32787 }, { "epoch": 2.5695924764890283, "grad_norm": 6.557938575744629, "learning_rate": 2.847027040585146e-06, "loss": 1.1069, "step": 32788 }, { "epoch": 2.569670846394984, "grad_norm": 3.9228246212005615, "learning_rate": 2.846748733905343e-06, "loss": 0.4575, "step": 32789 }, { "epoch": 2.5697492163009406, "grad_norm": 2.0673482418060303, "learning_rate": 2.8464704354154017e-06, "loss": 0.9701, "step": 32790 }, { "epoch": 2.5698275862068964, "grad_norm": 4.68625545501709, "learning_rate": 2.846192145116381e-06, "loss": 1.6644, "step": 32791 }, { "epoch": 2.5699059561128528, "grad_norm": 4.16480016708374, "learning_rate": 2.845913863009339e-06, "loss": 1.6211, "step": 32792 }, { "epoch": 2.5699843260188087, "grad_norm": 3.0397109985351562, "learning_rate": 2.8456355890953346e-06, "loss": 1.3736, "step": 32793 }, { "epoch": 2.570062695924765, "grad_norm": 2.560124158859253, "learning_rate": 2.845357323375425e-06, "loss": 1.1947, "step": 32794 }, { "epoch": 2.570141065830721, "grad_norm": 3.47802996635437, "learning_rate": 2.8450790658506704e-06, "loss": 1.302, "step": 32795 }, { "epoch": 2.570219435736677, "grad_norm": 2.5346226692199707, "learning_rate": 2.844800816522128e-06, "loss": 1.3251, "step": 32796 }, { "epoch": 2.570297805642633, "grad_norm": 3.6994669437408447, "learning_rate": 2.844522575390856e-06, "loss": 0.9718, "step": 32797 }, { "epoch": 2.5703761755485894, "grad_norm": 12.377930641174316, "learning_rate": 2.8442443424579153e-06, "loss": 1.4144, "step": 32798 }, { "epoch": 2.5704545454545453, "grad_norm": 3.946416139602661, "learning_rate": 2.8439661177243576e-06, "loss": 1.4433, "step": 32799 }, { "epoch": 2.5705329153605017, "grad_norm": 2.7462825775146484, "learning_rate": 2.843687901191248e-06, "loss": 1.3992, "step": 32800 }, { "epoch": 2.5706112852664575, "grad_norm": 2.923518419265747, "learning_rate": 2.843409692859641e-06, "loss": 1.222, "step": 32801 }, { "epoch": 2.570689655172414, "grad_norm": 3.472318649291992, "learning_rate": 2.8431314927305988e-06, "loss": 1.8269, "step": 32802 }, { "epoch": 2.5707680250783698, "grad_norm": 2.8509521484375, "learning_rate": 2.8428533008051738e-06, "loss": 1.6248, "step": 32803 }, { "epoch": 2.570846394984326, "grad_norm": 3.8671860694885254, "learning_rate": 2.8425751170844275e-06, "loss": 1.32, "step": 32804 }, { "epoch": 2.570924764890282, "grad_norm": 2.9862396717071533, "learning_rate": 2.8422969415694146e-06, "loss": 1.2414, "step": 32805 }, { "epoch": 2.5710031347962383, "grad_norm": 5.869577407836914, "learning_rate": 2.8420187742612006e-06, "loss": 1.315, "step": 32806 }, { "epoch": 2.571081504702194, "grad_norm": 2.6481833457946777, "learning_rate": 2.841740615160836e-06, "loss": 1.3547, "step": 32807 }, { "epoch": 2.5711598746081505, "grad_norm": 3.3798298835754395, "learning_rate": 2.8414624642693823e-06, "loss": 1.1485, "step": 32808 }, { "epoch": 2.5712382445141064, "grad_norm": 11.420775413513184, "learning_rate": 2.8411843215878965e-06, "loss": 0.8166, "step": 32809 }, { "epoch": 2.5713166144200628, "grad_norm": 4.1485981941223145, "learning_rate": 2.840906187117437e-06, "loss": 0.9435, "step": 32810 }, { "epoch": 2.5713949843260187, "grad_norm": 4.842505931854248, "learning_rate": 2.840628060859061e-06, "loss": 1.7475, "step": 32811 }, { "epoch": 2.571473354231975, "grad_norm": 2.7964847087860107, "learning_rate": 2.8403499428138266e-06, "loss": 1.1884, "step": 32812 }, { "epoch": 2.571551724137931, "grad_norm": 3.711475372314453, "learning_rate": 2.8400718329827916e-06, "loss": 1.0255, "step": 32813 }, { "epoch": 2.571630094043887, "grad_norm": 3.3715810775756836, "learning_rate": 2.8397937313670144e-06, "loss": 1.2525, "step": 32814 }, { "epoch": 2.571708463949843, "grad_norm": 3.2671899795532227, "learning_rate": 2.839515637967552e-06, "loss": 1.2347, "step": 32815 }, { "epoch": 2.5717868338557994, "grad_norm": 10.456722259521484, "learning_rate": 2.8392375527854627e-06, "loss": 1.0584, "step": 32816 }, { "epoch": 2.5718652037617558, "grad_norm": 2.6817238330841064, "learning_rate": 2.8389594758218057e-06, "loss": 1.1491, "step": 32817 }, { "epoch": 2.5719435736677116, "grad_norm": 9.704452514648438, "learning_rate": 2.838681407077633e-06, "loss": 0.8836, "step": 32818 }, { "epoch": 2.5720219435736675, "grad_norm": 5.0470757484436035, "learning_rate": 2.8384033465540085e-06, "loss": 1.5468, "step": 32819 }, { "epoch": 2.572100313479624, "grad_norm": 3.2242088317871094, "learning_rate": 2.8381252942519864e-06, "loss": 1.3332, "step": 32820 }, { "epoch": 2.57217868338558, "grad_norm": 2.429324150085449, "learning_rate": 2.8378472501726277e-06, "loss": 1.3966, "step": 32821 }, { "epoch": 2.572257053291536, "grad_norm": 2.9457879066467285, "learning_rate": 2.8375692143169856e-06, "loss": 1.0906, "step": 32822 }, { "epoch": 2.572335423197492, "grad_norm": 3.0785329341888428, "learning_rate": 2.8372911866861174e-06, "loss": 1.228, "step": 32823 }, { "epoch": 2.5724137931034483, "grad_norm": 8.179078102111816, "learning_rate": 2.8370131672810876e-06, "loss": 1.0371, "step": 32824 }, { "epoch": 2.5724921630094046, "grad_norm": 9.124627113342285, "learning_rate": 2.8367351561029453e-06, "loss": 1.4301, "step": 32825 }, { "epoch": 2.5725705329153605, "grad_norm": 3.7006072998046875, "learning_rate": 2.8364571531527524e-06, "loss": 0.6976, "step": 32826 }, { "epoch": 2.5726489028213164, "grad_norm": 3.2312607765197754, "learning_rate": 2.8361791584315646e-06, "loss": 1.4221, "step": 32827 }, { "epoch": 2.5727272727272728, "grad_norm": 2.767957925796509, "learning_rate": 2.8359011719404407e-06, "loss": 1.4945, "step": 32828 }, { "epoch": 2.572805642633229, "grad_norm": 17.946561813354492, "learning_rate": 2.8356231936804357e-06, "loss": 1.4487, "step": 32829 }, { "epoch": 2.572884012539185, "grad_norm": 5.545750617980957, "learning_rate": 2.8353452236526097e-06, "loss": 1.4605, "step": 32830 }, { "epoch": 2.572962382445141, "grad_norm": 2.077765941619873, "learning_rate": 2.8350672618580184e-06, "loss": 1.3337, "step": 32831 }, { "epoch": 2.573040752351097, "grad_norm": 4.97661018371582, "learning_rate": 2.8347893082977187e-06, "loss": 1.5962, "step": 32832 }, { "epoch": 2.5731191222570535, "grad_norm": 5.05047082901001, "learning_rate": 2.8345113629727687e-06, "loss": 1.9147, "step": 32833 }, { "epoch": 2.5731974921630094, "grad_norm": 3.4250097274780273, "learning_rate": 2.8342334258842252e-06, "loss": 0.8703, "step": 32834 }, { "epoch": 2.5732758620689653, "grad_norm": 2.4570701122283936, "learning_rate": 2.833955497033145e-06, "loss": 1.4347, "step": 32835 }, { "epoch": 2.5733542319749216, "grad_norm": 3.006605386734009, "learning_rate": 2.8336775764205883e-06, "loss": 1.569, "step": 32836 }, { "epoch": 2.573432601880878, "grad_norm": 3.671191692352295, "learning_rate": 2.833399664047605e-06, "loss": 0.574, "step": 32837 }, { "epoch": 2.573510971786834, "grad_norm": 5.950355052947998, "learning_rate": 2.8331217599152584e-06, "loss": 1.8325, "step": 32838 }, { "epoch": 2.5735893416927897, "grad_norm": 4.440699577331543, "learning_rate": 2.832843864024606e-06, "loss": 1.3242, "step": 32839 }, { "epoch": 2.573667711598746, "grad_norm": 3.9711358547210693, "learning_rate": 2.8325659763767e-06, "loss": 1.1703, "step": 32840 }, { "epoch": 2.5737460815047024, "grad_norm": 4.258313179016113, "learning_rate": 2.8322880969726e-06, "loss": 1.525, "step": 32841 }, { "epoch": 2.5738244514106583, "grad_norm": 4.4336347579956055, "learning_rate": 2.8320102258133607e-06, "loss": 1.523, "step": 32842 }, { "epoch": 2.573902821316614, "grad_norm": 6.6708807945251465, "learning_rate": 2.8317323629000447e-06, "loss": 1.7302, "step": 32843 }, { "epoch": 2.5739811912225705, "grad_norm": 5.181534767150879, "learning_rate": 2.8314545082337036e-06, "loss": 1.16, "step": 32844 }, { "epoch": 2.574059561128527, "grad_norm": 2.9257209300994873, "learning_rate": 2.831176661815396e-06, "loss": 1.4345, "step": 32845 }, { "epoch": 2.5741379310344827, "grad_norm": 2.5102429389953613, "learning_rate": 2.8308988236461777e-06, "loss": 1.2225, "step": 32846 }, { "epoch": 2.5742163009404386, "grad_norm": 3.0501291751861572, "learning_rate": 2.8306209937271057e-06, "loss": 1.2014, "step": 32847 }, { "epoch": 2.574294670846395, "grad_norm": 2.209261655807495, "learning_rate": 2.830343172059238e-06, "loss": 1.2849, "step": 32848 }, { "epoch": 2.5743730407523513, "grad_norm": 3.477423667907715, "learning_rate": 2.83006535864363e-06, "loss": 1.3184, "step": 32849 }, { "epoch": 2.574451410658307, "grad_norm": 10.061484336853027, "learning_rate": 2.8297875534813392e-06, "loss": 1.5604, "step": 32850 }, { "epoch": 2.574529780564263, "grad_norm": 3.148444652557373, "learning_rate": 2.8295097565734214e-06, "loss": 1.5345, "step": 32851 }, { "epoch": 2.5746081504702194, "grad_norm": 3.291940212249756, "learning_rate": 2.829231967920934e-06, "loss": 1.4815, "step": 32852 }, { "epoch": 2.5746865203761757, "grad_norm": 2.501253366470337, "learning_rate": 2.8289541875249333e-06, "loss": 1.8153, "step": 32853 }, { "epoch": 2.5747648902821316, "grad_norm": 3.5394668579101562, "learning_rate": 2.8286764153864776e-06, "loss": 0.8808, "step": 32854 }, { "epoch": 2.5748432601880875, "grad_norm": 2.677647590637207, "learning_rate": 2.82839865150662e-06, "loss": 1.4927, "step": 32855 }, { "epoch": 2.574921630094044, "grad_norm": 3.1329612731933594, "learning_rate": 2.828120895886416e-06, "loss": 1.2557, "step": 32856 }, { "epoch": 2.575, "grad_norm": 3.4487531185150146, "learning_rate": 2.8278431485269266e-06, "loss": 1.3731, "step": 32857 }, { "epoch": 2.575078369905956, "grad_norm": 3.733898878097534, "learning_rate": 2.827565409429208e-06, "loss": 1.5658, "step": 32858 }, { "epoch": 2.575156739811912, "grad_norm": 1.6647284030914307, "learning_rate": 2.8272876785943126e-06, "loss": 0.8836, "step": 32859 }, { "epoch": 2.5752351097178683, "grad_norm": 4.76026725769043, "learning_rate": 2.8270099560232992e-06, "loss": 1.1621, "step": 32860 }, { "epoch": 2.5753134796238246, "grad_norm": 4.031188488006592, "learning_rate": 2.826732241717222e-06, "loss": 1.3061, "step": 32861 }, { "epoch": 2.5753918495297805, "grad_norm": 3.034848690032959, "learning_rate": 2.8264545356771434e-06, "loss": 1.4762, "step": 32862 }, { "epoch": 2.575470219435737, "grad_norm": 2.484550952911377, "learning_rate": 2.8261768379041127e-06, "loss": 1.2118, "step": 32863 }, { "epoch": 2.5755485893416927, "grad_norm": 2.523183584213257, "learning_rate": 2.8258991483991886e-06, "loss": 1.2665, "step": 32864 }, { "epoch": 2.575626959247649, "grad_norm": 3.748793363571167, "learning_rate": 2.825621467163428e-06, "loss": 1.3577, "step": 32865 }, { "epoch": 2.575705329153605, "grad_norm": 4.689346790313721, "learning_rate": 2.825343794197887e-06, "loss": 1.6276, "step": 32866 }, { "epoch": 2.5757836990595613, "grad_norm": 2.8488423824310303, "learning_rate": 2.8250661295036197e-06, "loss": 1.4909, "step": 32867 }, { "epoch": 2.575862068965517, "grad_norm": 2.3461833000183105, "learning_rate": 2.824788473081685e-06, "loss": 1.3967, "step": 32868 }, { "epoch": 2.5759404388714735, "grad_norm": 2.330007791519165, "learning_rate": 2.8245108249331386e-06, "loss": 1.2287, "step": 32869 }, { "epoch": 2.5760188087774294, "grad_norm": 2.455120086669922, "learning_rate": 2.824233185059033e-06, "loss": 1.2656, "step": 32870 }, { "epoch": 2.5760971786833857, "grad_norm": 3.0857458114624023, "learning_rate": 2.823955553460428e-06, "loss": 1.4074, "step": 32871 }, { "epoch": 2.5761755485893416, "grad_norm": 3.698762893676758, "learning_rate": 2.823677930138379e-06, "loss": 0.997, "step": 32872 }, { "epoch": 2.576253918495298, "grad_norm": 2.954850673675537, "learning_rate": 2.8234003150939426e-06, "loss": 1.3892, "step": 32873 }, { "epoch": 2.576332288401254, "grad_norm": 5.713067054748535, "learning_rate": 2.8231227083281714e-06, "loss": 1.2395, "step": 32874 }, { "epoch": 2.57641065830721, "grad_norm": 3.1635611057281494, "learning_rate": 2.822845109842122e-06, "loss": 1.1943, "step": 32875 }, { "epoch": 2.576489028213166, "grad_norm": 2.5022356510162354, "learning_rate": 2.822567519636853e-06, "loss": 1.3075, "step": 32876 }, { "epoch": 2.5765673981191224, "grad_norm": 2.6379480361938477, "learning_rate": 2.8222899377134206e-06, "loss": 1.3538, "step": 32877 }, { "epoch": 2.5766457680250783, "grad_norm": 4.144709587097168, "learning_rate": 2.822012364072877e-06, "loss": 1.3969, "step": 32878 }, { "epoch": 2.5767241379310346, "grad_norm": 3.9189958572387695, "learning_rate": 2.8217347987162797e-06, "loss": 1.2996, "step": 32879 }, { "epoch": 2.5768025078369905, "grad_norm": 4.699196815490723, "learning_rate": 2.8214572416446838e-06, "loss": 1.2942, "step": 32880 }, { "epoch": 2.576880877742947, "grad_norm": 2.429694414138794, "learning_rate": 2.8211796928591466e-06, "loss": 1.4254, "step": 32881 }, { "epoch": 2.5769592476489027, "grad_norm": 2.798698663711548, "learning_rate": 2.8209021523607223e-06, "loss": 1.1487, "step": 32882 }, { "epoch": 2.577037617554859, "grad_norm": 3.2486722469329834, "learning_rate": 2.820624620150467e-06, "loss": 1.4386, "step": 32883 }, { "epoch": 2.577115987460815, "grad_norm": 2.8824524879455566, "learning_rate": 2.820347096229436e-06, "loss": 1.4249, "step": 32884 }, { "epoch": 2.5771943573667713, "grad_norm": 3.440891742706299, "learning_rate": 2.8200695805986855e-06, "loss": 1.2175, "step": 32885 }, { "epoch": 2.577272727272727, "grad_norm": 2.832371234893799, "learning_rate": 2.8197920732592705e-06, "loss": 1.2258, "step": 32886 }, { "epoch": 2.5773510971786835, "grad_norm": 2.6249310970306396, "learning_rate": 2.8195145742122465e-06, "loss": 1.3898, "step": 32887 }, { "epoch": 2.5774294670846394, "grad_norm": 2.9918346405029297, "learning_rate": 2.8192370834586713e-06, "loss": 1.5645, "step": 32888 }, { "epoch": 2.5775078369905957, "grad_norm": 2.8903088569641113, "learning_rate": 2.8189596009995933e-06, "loss": 1.3116, "step": 32889 }, { "epoch": 2.5775862068965516, "grad_norm": 4.198540210723877, "learning_rate": 2.8186821268360757e-06, "loss": 1.5529, "step": 32890 }, { "epoch": 2.577664576802508, "grad_norm": 3.153660297393799, "learning_rate": 2.818404660969171e-06, "loss": 1.5583, "step": 32891 }, { "epoch": 2.577742946708464, "grad_norm": 2.922135353088379, "learning_rate": 2.8181272033999358e-06, "loss": 1.0503, "step": 32892 }, { "epoch": 2.57782131661442, "grad_norm": 15.521324157714844, "learning_rate": 2.8178497541294226e-06, "loss": 1.7397, "step": 32893 }, { "epoch": 2.577899686520376, "grad_norm": 2.7903852462768555, "learning_rate": 2.817572313158685e-06, "loss": 1.5437, "step": 32894 }, { "epoch": 2.5779780564263324, "grad_norm": 5.68865442276001, "learning_rate": 2.817294880488787e-06, "loss": 1.7347, "step": 32895 }, { "epoch": 2.5780564263322883, "grad_norm": 4.61734676361084, "learning_rate": 2.817017456120775e-06, "loss": 1.5857, "step": 32896 }, { "epoch": 2.5781347962382446, "grad_norm": 2.7024683952331543, "learning_rate": 2.8167400400557077e-06, "loss": 1.2826, "step": 32897 }, { "epoch": 2.5782131661442005, "grad_norm": 2.993701696395874, "learning_rate": 2.8164626322946396e-06, "loss": 1.3353, "step": 32898 }, { "epoch": 2.578291536050157, "grad_norm": 2.4842684268951416, "learning_rate": 2.816185232838626e-06, "loss": 1.1762, "step": 32899 }, { "epoch": 2.5783699059561127, "grad_norm": 4.450223922729492, "learning_rate": 2.8159078416887233e-06, "loss": 1.7124, "step": 32900 }, { "epoch": 2.578448275862069, "grad_norm": 3.9841926097869873, "learning_rate": 2.815630458845984e-06, "loss": 1.12, "step": 32901 }, { "epoch": 2.578526645768025, "grad_norm": 3.4355807304382324, "learning_rate": 2.8153530843114653e-06, "loss": 1.0913, "step": 32902 }, { "epoch": 2.5786050156739813, "grad_norm": 3.9801595211029053, "learning_rate": 2.815075718086221e-06, "loss": 1.5062, "step": 32903 }, { "epoch": 2.578683385579937, "grad_norm": 7.591659069061279, "learning_rate": 2.8147983601713058e-06, "loss": 0.8965, "step": 32904 }, { "epoch": 2.5787617554858935, "grad_norm": 3.1556499004364014, "learning_rate": 2.814521010567777e-06, "loss": 1.0373, "step": 32905 }, { "epoch": 2.5788401253918494, "grad_norm": 2.736341714859009, "learning_rate": 2.8142436692766863e-06, "loss": 1.1523, "step": 32906 }, { "epoch": 2.5789184952978057, "grad_norm": 6.440057277679443, "learning_rate": 2.81396633629909e-06, "loss": 1.0327, "step": 32907 }, { "epoch": 2.5789968652037616, "grad_norm": 2.664109468460083, "learning_rate": 2.8136890116360433e-06, "loss": 1.4278, "step": 32908 }, { "epoch": 2.579075235109718, "grad_norm": 2.1004343032836914, "learning_rate": 2.8134116952886004e-06, "loss": 1.232, "step": 32909 }, { "epoch": 2.579153605015674, "grad_norm": 3.133028030395508, "learning_rate": 2.8131343872578187e-06, "loss": 1.5098, "step": 32910 }, { "epoch": 2.57923197492163, "grad_norm": 2.4841184616088867, "learning_rate": 2.812857087544748e-06, "loss": 0.6909, "step": 32911 }, { "epoch": 2.5793103448275865, "grad_norm": 3.234283685684204, "learning_rate": 2.812579796150444e-06, "loss": 1.2377, "step": 32912 }, { "epoch": 2.5793887147335424, "grad_norm": 2.304266929626465, "learning_rate": 2.812302513075965e-06, "loss": 1.3115, "step": 32913 }, { "epoch": 2.5794670846394983, "grad_norm": 5.124608516693115, "learning_rate": 2.8120252383223657e-06, "loss": 1.0914, "step": 32914 }, { "epoch": 2.5795454545454546, "grad_norm": 3.238053321838379, "learning_rate": 2.811747971890696e-06, "loss": 1.2637, "step": 32915 }, { "epoch": 2.579623824451411, "grad_norm": 3.5036144256591797, "learning_rate": 2.8114707137820135e-06, "loss": 1.5563, "step": 32916 }, { "epoch": 2.579702194357367, "grad_norm": 4.956724166870117, "learning_rate": 2.8111934639973724e-06, "loss": 1.981, "step": 32917 }, { "epoch": 2.5797805642633227, "grad_norm": 2.7633092403411865, "learning_rate": 2.8109162225378268e-06, "loss": 1.376, "step": 32918 }, { "epoch": 2.579858934169279, "grad_norm": 3.4420180320739746, "learning_rate": 2.8106389894044317e-06, "loss": 1.5426, "step": 32919 }, { "epoch": 2.5799373040752354, "grad_norm": 3.003713607788086, "learning_rate": 2.810361764598241e-06, "loss": 1.4036, "step": 32920 }, { "epoch": 2.5800156739811912, "grad_norm": 3.6374948024749756, "learning_rate": 2.81008454812031e-06, "loss": 1.5875, "step": 32921 }, { "epoch": 2.580094043887147, "grad_norm": 3.837249755859375, "learning_rate": 2.809807339971692e-06, "loss": 1.8266, "step": 32922 }, { "epoch": 2.5801724137931035, "grad_norm": 1.84783935546875, "learning_rate": 2.8095301401534426e-06, "loss": 1.0189, "step": 32923 }, { "epoch": 2.58025078369906, "grad_norm": 2.6623005867004395, "learning_rate": 2.809252948666615e-06, "loss": 1.6804, "step": 32924 }, { "epoch": 2.5803291536050157, "grad_norm": 3.443742275238037, "learning_rate": 2.808975765512266e-06, "loss": 1.3091, "step": 32925 }, { "epoch": 2.5804075235109716, "grad_norm": 2.4380645751953125, "learning_rate": 2.8086985906914437e-06, "loss": 1.381, "step": 32926 }, { "epoch": 2.580485893416928, "grad_norm": 9.57231330871582, "learning_rate": 2.808421424205209e-06, "loss": 2.2957, "step": 32927 }, { "epoch": 2.5805642633228842, "grad_norm": 2.3641703128814697, "learning_rate": 2.8081442660546126e-06, "loss": 1.4025, "step": 32928 }, { "epoch": 2.58064263322884, "grad_norm": 5.209578037261963, "learning_rate": 2.8078671162407123e-06, "loss": 1.2798, "step": 32929 }, { "epoch": 2.580721003134796, "grad_norm": 2.791825771331787, "learning_rate": 2.807589974764557e-06, "loss": 1.4887, "step": 32930 }, { "epoch": 2.5807993730407524, "grad_norm": 2.682025671005249, "learning_rate": 2.807312841627202e-06, "loss": 1.4913, "step": 32931 }, { "epoch": 2.5808777429467087, "grad_norm": 4.209768295288086, "learning_rate": 2.8070357168297043e-06, "loss": 1.4193, "step": 32932 }, { "epoch": 2.5809561128526646, "grad_norm": 3.1255578994750977, "learning_rate": 2.8067586003731186e-06, "loss": 1.0163, "step": 32933 }, { "epoch": 2.5810344827586205, "grad_norm": 6.342472553253174, "learning_rate": 2.8064814922584945e-06, "loss": 1.4554, "step": 32934 }, { "epoch": 2.581112852664577, "grad_norm": 2.558314800262451, "learning_rate": 2.8062043924868874e-06, "loss": 1.2201, "step": 32935 }, { "epoch": 2.581191222570533, "grad_norm": 1.9828283786773682, "learning_rate": 2.8059273010593536e-06, "loss": 1.3119, "step": 32936 }, { "epoch": 2.581269592476489, "grad_norm": 2.346442222595215, "learning_rate": 2.805650217976944e-06, "loss": 1.4035, "step": 32937 }, { "epoch": 2.581347962382445, "grad_norm": 7.512507438659668, "learning_rate": 2.805373143240715e-06, "loss": 1.3587, "step": 32938 }, { "epoch": 2.5814263322884012, "grad_norm": 12.158190727233887, "learning_rate": 2.805096076851719e-06, "loss": 1.7481, "step": 32939 }, { "epoch": 2.5815047021943576, "grad_norm": 4.92191743850708, "learning_rate": 2.804819018811009e-06, "loss": 1.3897, "step": 32940 }, { "epoch": 2.5815830721003135, "grad_norm": 3.595233917236328, "learning_rate": 2.8045419691196414e-06, "loss": 1.2318, "step": 32941 }, { "epoch": 2.5816614420062693, "grad_norm": 3.765120029449463, "learning_rate": 2.8042649277786676e-06, "loss": 1.4809, "step": 32942 }, { "epoch": 2.5817398119122257, "grad_norm": 2.7761685848236084, "learning_rate": 2.803987894789143e-06, "loss": 1.3692, "step": 32943 }, { "epoch": 2.581818181818182, "grad_norm": 3.5415210723876953, "learning_rate": 2.8037108701521222e-06, "loss": 1.2607, "step": 32944 }, { "epoch": 2.581896551724138, "grad_norm": 2.706716537475586, "learning_rate": 2.803433853868652e-06, "loss": 1.3339, "step": 32945 }, { "epoch": 2.581974921630094, "grad_norm": 5.451425075531006, "learning_rate": 2.803156845939795e-06, "loss": 0.6914, "step": 32946 }, { "epoch": 2.58205329153605, "grad_norm": 2.867882013320923, "learning_rate": 2.8028798463665995e-06, "loss": 1.4069, "step": 32947 }, { "epoch": 2.5821316614420065, "grad_norm": 2.654716968536377, "learning_rate": 2.802602855150124e-06, "loss": 1.1784, "step": 32948 }, { "epoch": 2.5822100313479623, "grad_norm": 2.3056743144989014, "learning_rate": 2.8023258722914154e-06, "loss": 1.4223, "step": 32949 }, { "epoch": 2.5822884012539182, "grad_norm": 5.037342071533203, "learning_rate": 2.802048897791529e-06, "loss": 1.9027, "step": 32950 }, { "epoch": 2.5823667711598746, "grad_norm": 2.532700300216675, "learning_rate": 2.8017719316515237e-06, "loss": 1.326, "step": 32951 }, { "epoch": 2.582445141065831, "grad_norm": 3.056452989578247, "learning_rate": 2.801494973872446e-06, "loss": 1.3263, "step": 32952 }, { "epoch": 2.582523510971787, "grad_norm": 5.887502193450928, "learning_rate": 2.801218024455353e-06, "loss": 1.1476, "step": 32953 }, { "epoch": 2.5826018808777427, "grad_norm": 2.1128268241882324, "learning_rate": 2.8009410834012973e-06, "loss": 1.3704, "step": 32954 }, { "epoch": 2.582680250783699, "grad_norm": 2.8229804039001465, "learning_rate": 2.800664150711332e-06, "loss": 1.6229, "step": 32955 }, { "epoch": 2.5827586206896553, "grad_norm": 3.558486223220825, "learning_rate": 2.800387226386511e-06, "loss": 1.2319, "step": 32956 }, { "epoch": 2.5828369905956112, "grad_norm": 2.5123789310455322, "learning_rate": 2.8001103104278863e-06, "loss": 1.4368, "step": 32957 }, { "epoch": 2.5829153605015676, "grad_norm": 19.048559188842773, "learning_rate": 2.7998334028365125e-06, "loss": 0.8873, "step": 32958 }, { "epoch": 2.5829937304075234, "grad_norm": 2.279320001602173, "learning_rate": 2.799556503613443e-06, "loss": 1.1562, "step": 32959 }, { "epoch": 2.58307210031348, "grad_norm": 3.1963577270507812, "learning_rate": 2.79927961275973e-06, "loss": 1.545, "step": 32960 }, { "epoch": 2.5831504702194357, "grad_norm": 4.282219886779785, "learning_rate": 2.799002730276427e-06, "loss": 1.4539, "step": 32961 }, { "epoch": 2.583228840125392, "grad_norm": 3.0328197479248047, "learning_rate": 2.7987258561645863e-06, "loss": 1.484, "step": 32962 }, { "epoch": 2.583307210031348, "grad_norm": 2.969308853149414, "learning_rate": 2.798448990425264e-06, "loss": 0.9627, "step": 32963 }, { "epoch": 2.583385579937304, "grad_norm": 2.6792819499969482, "learning_rate": 2.798172133059508e-06, "loss": 1.2141, "step": 32964 }, { "epoch": 2.58346394984326, "grad_norm": 2.721762180328369, "learning_rate": 2.7978952840683752e-06, "loss": 1.4148, "step": 32965 }, { "epoch": 2.5835423197492164, "grad_norm": 3.217111110687256, "learning_rate": 2.79761844345292e-06, "loss": 1.3676, "step": 32966 }, { "epoch": 2.5836206896551723, "grad_norm": 3.6379024982452393, "learning_rate": 2.7973416112141915e-06, "loss": 1.2537, "step": 32967 }, { "epoch": 2.5836990595611287, "grad_norm": 2.8535308837890625, "learning_rate": 2.7970647873532442e-06, "loss": 1.2963, "step": 32968 }, { "epoch": 2.5837774294670846, "grad_norm": 3.7050600051879883, "learning_rate": 2.796787971871128e-06, "loss": 1.855, "step": 32969 }, { "epoch": 2.583855799373041, "grad_norm": 2.574842691421509, "learning_rate": 2.7965111647689036e-06, "loss": 1.4791, "step": 32970 }, { "epoch": 2.5839341692789968, "grad_norm": 3.212238073348999, "learning_rate": 2.7962343660476168e-06, "loss": 1.4071, "step": 32971 }, { "epoch": 2.584012539184953, "grad_norm": 3.6141042709350586, "learning_rate": 2.795957575708322e-06, "loss": 1.4458, "step": 32972 }, { "epoch": 2.584090909090909, "grad_norm": 3.413843870162964, "learning_rate": 2.7956807937520724e-06, "loss": 1.4956, "step": 32973 }, { "epoch": 2.5841692789968653, "grad_norm": 3.2131190299987793, "learning_rate": 2.7954040201799214e-06, "loss": 1.3421, "step": 32974 }, { "epoch": 2.584247648902821, "grad_norm": 4.5633015632629395, "learning_rate": 2.795127254992921e-06, "loss": 1.5243, "step": 32975 }, { "epoch": 2.5843260188087775, "grad_norm": 2.9241702556610107, "learning_rate": 2.7948504981921244e-06, "loss": 1.6263, "step": 32976 }, { "epoch": 2.5844043887147334, "grad_norm": 3.3495399951934814, "learning_rate": 2.7945737497785853e-06, "loss": 1.3891, "step": 32977 }, { "epoch": 2.5844827586206898, "grad_norm": 2.2891926765441895, "learning_rate": 2.7942970097533512e-06, "loss": 1.2702, "step": 32978 }, { "epoch": 2.5845611285266457, "grad_norm": 4.020003795623779, "learning_rate": 2.794020278117481e-06, "loss": 1.3452, "step": 32979 }, { "epoch": 2.584639498432602, "grad_norm": 2.607318162918091, "learning_rate": 2.7937435548720232e-06, "loss": 1.1311, "step": 32980 }, { "epoch": 2.584717868338558, "grad_norm": 3.402862071990967, "learning_rate": 2.793466840018034e-06, "loss": 1.4025, "step": 32981 }, { "epoch": 2.584796238244514, "grad_norm": 2.6078908443450928, "learning_rate": 2.7931901335565624e-06, "loss": 1.361, "step": 32982 }, { "epoch": 2.58487460815047, "grad_norm": 8.6357421875, "learning_rate": 2.79291343548866e-06, "loss": 1.2384, "step": 32983 }, { "epoch": 2.5849529780564264, "grad_norm": 16.4569034576416, "learning_rate": 2.7926367458153827e-06, "loss": 1.1217, "step": 32984 }, { "epoch": 2.5850313479623823, "grad_norm": 6.905070781707764, "learning_rate": 2.7923600645377835e-06, "loss": 1.3852, "step": 32985 }, { "epoch": 2.5851097178683387, "grad_norm": 2.4158895015716553, "learning_rate": 2.7920833916569112e-06, "loss": 1.4482, "step": 32986 }, { "epoch": 2.5851880877742945, "grad_norm": 4.30127477645874, "learning_rate": 2.7918067271738193e-06, "loss": 1.5293, "step": 32987 }, { "epoch": 2.585266457680251, "grad_norm": 3.7831668853759766, "learning_rate": 2.791530071089558e-06, "loss": 1.191, "step": 32988 }, { "epoch": 2.5853448275862068, "grad_norm": 2.6092939376831055, "learning_rate": 2.7912534234051862e-06, "loss": 1.4729, "step": 32989 }, { "epoch": 2.585423197492163, "grad_norm": 6.847400188446045, "learning_rate": 2.79097678412175e-06, "loss": 0.5651, "step": 32990 }, { "epoch": 2.585501567398119, "grad_norm": 3.4164974689483643, "learning_rate": 2.7907001532403035e-06, "loss": 1.3264, "step": 32991 }, { "epoch": 2.5855799373040753, "grad_norm": 3.2106497287750244, "learning_rate": 2.790423530761899e-06, "loss": 1.5505, "step": 32992 }, { "epoch": 2.585658307210031, "grad_norm": 2.5809969902038574, "learning_rate": 2.7901469166875884e-06, "loss": 1.1107, "step": 32993 }, { "epoch": 2.5857366771159875, "grad_norm": 3.4594016075134277, "learning_rate": 2.789870311018423e-06, "loss": 1.3586, "step": 32994 }, { "epoch": 2.5858150470219434, "grad_norm": 3.4403345584869385, "learning_rate": 2.7895937137554576e-06, "loss": 0.9779, "step": 32995 }, { "epoch": 2.5858934169278998, "grad_norm": 3.163231134414673, "learning_rate": 2.789317124899741e-06, "loss": 1.5171, "step": 32996 }, { "epoch": 2.5859717868338556, "grad_norm": 7.593623161315918, "learning_rate": 2.789040544452327e-06, "loss": 1.748, "step": 32997 }, { "epoch": 2.586050156739812, "grad_norm": 2.2545385360717773, "learning_rate": 2.788763972414267e-06, "loss": 1.5629, "step": 32998 }, { "epoch": 2.586128526645768, "grad_norm": 2.603566884994507, "learning_rate": 2.7884874087866134e-06, "loss": 0.9869, "step": 32999 }, { "epoch": 2.586206896551724, "grad_norm": 2.8715367317199707, "learning_rate": 2.788210853570419e-06, "loss": 0.5461, "step": 33000 }, { "epoch": 2.58628526645768, "grad_norm": 6.223435878753662, "learning_rate": 2.787934306766733e-06, "loss": 1.7363, "step": 33001 }, { "epoch": 2.5863636363636364, "grad_norm": 2.255380153656006, "learning_rate": 2.787657768376607e-06, "loss": 1.4115, "step": 33002 }, { "epoch": 2.5864420062695928, "grad_norm": 2.4830362796783447, "learning_rate": 2.787381238401096e-06, "loss": 1.3707, "step": 33003 }, { "epoch": 2.5865203761755486, "grad_norm": 2.5348947048187256, "learning_rate": 2.7871047168412525e-06, "loss": 1.4402, "step": 33004 }, { "epoch": 2.5865987460815045, "grad_norm": 2.9150586128234863, "learning_rate": 2.786828203698124e-06, "loss": 1.2804, "step": 33005 }, { "epoch": 2.586677115987461, "grad_norm": 3.8105688095092773, "learning_rate": 2.786551698972765e-06, "loss": 1.147, "step": 33006 }, { "epoch": 2.586755485893417, "grad_norm": 7.4274821281433105, "learning_rate": 2.7862752026662252e-06, "loss": 1.8872, "step": 33007 }, { "epoch": 2.586833855799373, "grad_norm": 3.0265846252441406, "learning_rate": 2.7859987147795586e-06, "loss": 1.1302, "step": 33008 }, { "epoch": 2.586912225705329, "grad_norm": 5.063901424407959, "learning_rate": 2.785722235313815e-06, "loss": 0.6202, "step": 33009 }, { "epoch": 2.5869905956112853, "grad_norm": 4.499945640563965, "learning_rate": 2.785445764270047e-06, "loss": 1.7014, "step": 33010 }, { "epoch": 2.5870689655172416, "grad_norm": 5.64730167388916, "learning_rate": 2.7851693016493053e-06, "loss": 1.3771, "step": 33011 }, { "epoch": 2.5871473354231975, "grad_norm": 5.616481781005859, "learning_rate": 2.7848928474526428e-06, "loss": 1.2407, "step": 33012 }, { "epoch": 2.5872257053291534, "grad_norm": 3.7639076709747314, "learning_rate": 2.7846164016811094e-06, "loss": 1.2694, "step": 33013 }, { "epoch": 2.5873040752351097, "grad_norm": 2.7361347675323486, "learning_rate": 2.784339964335758e-06, "loss": 1.2767, "step": 33014 }, { "epoch": 2.587382445141066, "grad_norm": 3.7830450534820557, "learning_rate": 2.7840635354176386e-06, "loss": 1.7248, "step": 33015 }, { "epoch": 2.587460815047022, "grad_norm": 2.778463363647461, "learning_rate": 2.783787114927804e-06, "loss": 1.3447, "step": 33016 }, { "epoch": 2.587539184952978, "grad_norm": 3.47662091255188, "learning_rate": 2.7835107028673047e-06, "loss": 1.2118, "step": 33017 }, { "epoch": 2.587617554858934, "grad_norm": 2.8996875286102295, "learning_rate": 2.783234299237192e-06, "loss": 1.3082, "step": 33018 }, { "epoch": 2.5876959247648905, "grad_norm": 3.7727596759796143, "learning_rate": 2.782957904038519e-06, "loss": 1.5715, "step": 33019 }, { "epoch": 2.5877742946708464, "grad_norm": 2.902923107147217, "learning_rate": 2.782681517272332e-06, "loss": 0.9842, "step": 33020 }, { "epoch": 2.5878526645768023, "grad_norm": 5.210294246673584, "learning_rate": 2.782405138939688e-06, "loss": 1.1055, "step": 33021 }, { "epoch": 2.5879310344827586, "grad_norm": 5.006606101989746, "learning_rate": 2.7821287690416377e-06, "loss": 1.429, "step": 33022 }, { "epoch": 2.588009404388715, "grad_norm": 5.040289878845215, "learning_rate": 2.7818524075792276e-06, "loss": 1.5407, "step": 33023 }, { "epoch": 2.588087774294671, "grad_norm": 5.223440170288086, "learning_rate": 2.781576054553512e-06, "loss": 1.3897, "step": 33024 }, { "epoch": 2.5881661442006267, "grad_norm": 4.050417423248291, "learning_rate": 2.7812997099655397e-06, "loss": 1.3427, "step": 33025 }, { "epoch": 2.588244514106583, "grad_norm": 2.2287254333496094, "learning_rate": 2.7810233738163684e-06, "loss": 1.3158, "step": 33026 }, { "epoch": 2.5883228840125394, "grad_norm": 2.2658369541168213, "learning_rate": 2.7807470461070413e-06, "loss": 1.3095, "step": 33027 }, { "epoch": 2.5884012539184953, "grad_norm": 5.019302845001221, "learning_rate": 2.7804707268386132e-06, "loss": 1.4601, "step": 33028 }, { "epoch": 2.588479623824451, "grad_norm": 3.835308313369751, "learning_rate": 2.7801944160121343e-06, "loss": 1.1293, "step": 33029 }, { "epoch": 2.5885579937304075, "grad_norm": 2.110225200653076, "learning_rate": 2.779918113628656e-06, "loss": 1.2386, "step": 33030 }, { "epoch": 2.588636363636364, "grad_norm": 3.2829198837280273, "learning_rate": 2.7796418196892284e-06, "loss": 1.3203, "step": 33031 }, { "epoch": 2.5887147335423197, "grad_norm": 2.1096315383911133, "learning_rate": 2.779365534194903e-06, "loss": 1.1889, "step": 33032 }, { "epoch": 2.5887931034482756, "grad_norm": 3.887213945388794, "learning_rate": 2.779089257146732e-06, "loss": 1.8048, "step": 33033 }, { "epoch": 2.588871473354232, "grad_norm": 3.9178431034088135, "learning_rate": 2.7788129885457616e-06, "loss": 1.4641, "step": 33034 }, { "epoch": 2.5889498432601883, "grad_norm": 2.618182420730591, "learning_rate": 2.7785367283930474e-06, "loss": 1.2227, "step": 33035 }, { "epoch": 2.589028213166144, "grad_norm": 4.015438556671143, "learning_rate": 2.778260476689638e-06, "loss": 1.3971, "step": 33036 }, { "epoch": 2.5891065830721, "grad_norm": 4.343940258026123, "learning_rate": 2.7779842334365877e-06, "loss": 1.5097, "step": 33037 }, { "epoch": 2.5891849529780564, "grad_norm": 3.0799176692962646, "learning_rate": 2.777707998634941e-06, "loss": 1.3379, "step": 33038 }, { "epoch": 2.5892633228840127, "grad_norm": 3.9158315658569336, "learning_rate": 2.7774317722857503e-06, "loss": 1.4379, "step": 33039 }, { "epoch": 2.5893416927899686, "grad_norm": 2.9458606243133545, "learning_rate": 2.77715555439007e-06, "loss": 1.245, "step": 33040 }, { "epoch": 2.5894200626959245, "grad_norm": 3.80133056640625, "learning_rate": 2.776879344948949e-06, "loss": 1.5141, "step": 33041 }, { "epoch": 2.589498432601881, "grad_norm": 4.3214335441589355, "learning_rate": 2.7766031439634355e-06, "loss": 1.0905, "step": 33042 }, { "epoch": 2.589576802507837, "grad_norm": 4.0546159744262695, "learning_rate": 2.7763269514345813e-06, "loss": 0.9542, "step": 33043 }, { "epoch": 2.589655172413793, "grad_norm": 3.0590555667877197, "learning_rate": 2.7760507673634362e-06, "loss": 1.2182, "step": 33044 }, { "epoch": 2.589733542319749, "grad_norm": 2.502959728240967, "learning_rate": 2.7757745917510552e-06, "loss": 1.5246, "step": 33045 }, { "epoch": 2.5898119122257053, "grad_norm": 5.287391185760498, "learning_rate": 2.7754984245984836e-06, "loss": 1.2024, "step": 33046 }, { "epoch": 2.5898902821316616, "grad_norm": 3.2843494415283203, "learning_rate": 2.775222265906773e-06, "loss": 1.4705, "step": 33047 }, { "epoch": 2.5899686520376175, "grad_norm": 2.8875200748443604, "learning_rate": 2.774946115676974e-06, "loss": 1.4913, "step": 33048 }, { "epoch": 2.5900470219435734, "grad_norm": 6.622915744781494, "learning_rate": 2.7746699739101374e-06, "loss": 1.6806, "step": 33049 }, { "epoch": 2.5901253918495297, "grad_norm": 2.1654748916625977, "learning_rate": 2.7743938406073135e-06, "loss": 1.267, "step": 33050 }, { "epoch": 2.590203761755486, "grad_norm": 2.471729278564453, "learning_rate": 2.7741177157695513e-06, "loss": 1.358, "step": 33051 }, { "epoch": 2.590282131661442, "grad_norm": 3.0508391857147217, "learning_rate": 2.773841599397905e-06, "loss": 1.4564, "step": 33052 }, { "epoch": 2.5903605015673983, "grad_norm": 5.121703147888184, "learning_rate": 2.7735654914934173e-06, "loss": 1.1871, "step": 33053 }, { "epoch": 2.590438871473354, "grad_norm": 3.0516679286956787, "learning_rate": 2.7732893920571453e-06, "loss": 1.6593, "step": 33054 }, { "epoch": 2.5905172413793105, "grad_norm": 20.895065307617188, "learning_rate": 2.773013301090136e-06, "loss": 1.1248, "step": 33055 }, { "epoch": 2.5905956112852664, "grad_norm": 6.182298183441162, "learning_rate": 2.772737218593443e-06, "loss": 1.9313, "step": 33056 }, { "epoch": 2.5906739811912227, "grad_norm": 2.7541356086730957, "learning_rate": 2.7724611445681116e-06, "loss": 1.4722, "step": 33057 }, { "epoch": 2.5907523510971786, "grad_norm": 3.5227255821228027, "learning_rate": 2.7721850790151916e-06, "loss": 1.4094, "step": 33058 }, { "epoch": 2.590830721003135, "grad_norm": 2.598078489303589, "learning_rate": 2.7719090219357377e-06, "loss": 1.3034, "step": 33059 }, { "epoch": 2.590909090909091, "grad_norm": 3.488424777984619, "learning_rate": 2.771632973330799e-06, "loss": 1.6434, "step": 33060 }, { "epoch": 2.590987460815047, "grad_norm": 3.5769360065460205, "learning_rate": 2.771356933201422e-06, "loss": 1.2788, "step": 33061 }, { "epoch": 2.591065830721003, "grad_norm": 2.6711015701293945, "learning_rate": 2.7710809015486585e-06, "loss": 1.2038, "step": 33062 }, { "epoch": 2.5911442006269594, "grad_norm": 3.3389945030212402, "learning_rate": 2.770804878373559e-06, "loss": 1.3302, "step": 33063 }, { "epoch": 2.5912225705329153, "grad_norm": 6.848239898681641, "learning_rate": 2.770528863677172e-06, "loss": 1.5235, "step": 33064 }, { "epoch": 2.5913009404388716, "grad_norm": 3.357333183288574, "learning_rate": 2.770252857460548e-06, "loss": 1.4105, "step": 33065 }, { "epoch": 2.5913793103448275, "grad_norm": 4.015051364898682, "learning_rate": 2.7699768597247372e-06, "loss": 0.9421, "step": 33066 }, { "epoch": 2.591457680250784, "grad_norm": 7.940642833709717, "learning_rate": 2.769700870470789e-06, "loss": 1.9781, "step": 33067 }, { "epoch": 2.5915360501567397, "grad_norm": 2.684713840484619, "learning_rate": 2.7694248896997535e-06, "loss": 1.2953, "step": 33068 }, { "epoch": 2.591614420062696, "grad_norm": 2.436688184738159, "learning_rate": 2.7691489174126797e-06, "loss": 1.3939, "step": 33069 }, { "epoch": 2.591692789968652, "grad_norm": 13.935771942138672, "learning_rate": 2.7688729536106175e-06, "loss": 2.0772, "step": 33070 }, { "epoch": 2.5917711598746083, "grad_norm": 4.420040130615234, "learning_rate": 2.7685969982946184e-06, "loss": 1.3111, "step": 33071 }, { "epoch": 2.591849529780564, "grad_norm": 4.6287150382995605, "learning_rate": 2.7683210514657267e-06, "loss": 1.9864, "step": 33072 }, { "epoch": 2.5919278996865205, "grad_norm": 2.8150110244750977, "learning_rate": 2.7680451131249963e-06, "loss": 1.4304, "step": 33073 }, { "epoch": 2.5920062695924764, "grad_norm": 2.3588755130767822, "learning_rate": 2.767769183273477e-06, "loss": 1.0, "step": 33074 }, { "epoch": 2.5920846394984327, "grad_norm": 3.552672863006592, "learning_rate": 2.767493261912219e-06, "loss": 1.6654, "step": 33075 }, { "epoch": 2.5921630094043886, "grad_norm": 2.455925941467285, "learning_rate": 2.7672173490422683e-06, "loss": 1.5209, "step": 33076 }, { "epoch": 2.592241379310345, "grad_norm": 2.723724842071533, "learning_rate": 2.7669414446646737e-06, "loss": 1.2986, "step": 33077 }, { "epoch": 2.592319749216301, "grad_norm": 3.6971163749694824, "learning_rate": 2.7666655487804905e-06, "loss": 1.4085, "step": 33078 }, { "epoch": 2.592398119122257, "grad_norm": 3.359879970550537, "learning_rate": 2.7663896613907625e-06, "loss": 1.2342, "step": 33079 }, { "epoch": 2.592476489028213, "grad_norm": 3.0486767292022705, "learning_rate": 2.766113782496541e-06, "loss": 1.6045, "step": 33080 }, { "epoch": 2.5925548589341694, "grad_norm": 2.5172271728515625, "learning_rate": 2.7658379120988756e-06, "loss": 1.2666, "step": 33081 }, { "epoch": 2.5926332288401253, "grad_norm": 3.0212953090667725, "learning_rate": 2.765562050198815e-06, "loss": 1.2454, "step": 33082 }, { "epoch": 2.5927115987460816, "grad_norm": 3.3005611896514893, "learning_rate": 2.7652861967974086e-06, "loss": 0.9782, "step": 33083 }, { "epoch": 2.5927899686520375, "grad_norm": 6.472497940063477, "learning_rate": 2.765010351895706e-06, "loss": 1.5307, "step": 33084 }, { "epoch": 2.592868338557994, "grad_norm": 3.534104108810425, "learning_rate": 2.7647345154947557e-06, "loss": 1.5455, "step": 33085 }, { "epoch": 2.5929467084639497, "grad_norm": 3.100031852722168, "learning_rate": 2.7644586875956074e-06, "loss": 1.4766, "step": 33086 }, { "epoch": 2.593025078369906, "grad_norm": 4.613798141479492, "learning_rate": 2.76418286819931e-06, "loss": 1.488, "step": 33087 }, { "epoch": 2.593103448275862, "grad_norm": 3.03196382522583, "learning_rate": 2.7639070573069126e-06, "loss": 1.0897, "step": 33088 }, { "epoch": 2.5931818181818183, "grad_norm": 3.6392300128936768, "learning_rate": 2.7636312549194667e-06, "loss": 1.39, "step": 33089 }, { "epoch": 2.593260188087774, "grad_norm": 3.429112195968628, "learning_rate": 2.763355461038016e-06, "loss": 1.1838, "step": 33090 }, { "epoch": 2.5933385579937305, "grad_norm": 1.9792656898498535, "learning_rate": 2.7630796756636104e-06, "loss": 0.844, "step": 33091 }, { "epoch": 2.5934169278996864, "grad_norm": 2.956430435180664, "learning_rate": 2.7628038987973032e-06, "loss": 1.6359, "step": 33092 }, { "epoch": 2.5934952978056427, "grad_norm": 2.2418837547302246, "learning_rate": 2.762528130440143e-06, "loss": 1.2041, "step": 33093 }, { "epoch": 2.5935736677115986, "grad_norm": 4.456027507781982, "learning_rate": 2.762252370593174e-06, "loss": 1.5916, "step": 33094 }, { "epoch": 2.593652037617555, "grad_norm": 4.41325569152832, "learning_rate": 2.7619766192574476e-06, "loss": 1.2174, "step": 33095 }, { "epoch": 2.593730407523511, "grad_norm": 9.86967945098877, "learning_rate": 2.7617008764340116e-06, "loss": 0.9058, "step": 33096 }, { "epoch": 2.593808777429467, "grad_norm": 3.3948402404785156, "learning_rate": 2.7614251421239187e-06, "loss": 1.4862, "step": 33097 }, { "epoch": 2.5938871473354235, "grad_norm": 2.5975353717803955, "learning_rate": 2.761149416328213e-06, "loss": 1.5885, "step": 33098 }, { "epoch": 2.5939655172413794, "grad_norm": 10.289212226867676, "learning_rate": 2.7608736990479447e-06, "loss": 0.7393, "step": 33099 }, { "epoch": 2.5940438871473352, "grad_norm": 3.008932113647461, "learning_rate": 2.7605979902841635e-06, "loss": 0.6112, "step": 33100 }, { "epoch": 2.5941222570532916, "grad_norm": 2.733430862426758, "learning_rate": 2.760322290037917e-06, "loss": 1.1963, "step": 33101 }, { "epoch": 2.594200626959248, "grad_norm": 4.59464693069458, "learning_rate": 2.7600465983102543e-06, "loss": 1.4191, "step": 33102 }, { "epoch": 2.594278996865204, "grad_norm": 5.6663618087768555, "learning_rate": 2.759770915102223e-06, "loss": 1.2386, "step": 33103 }, { "epoch": 2.5943573667711597, "grad_norm": 3.689030170440674, "learning_rate": 2.7594952404148735e-06, "loss": 1.3251, "step": 33104 }, { "epoch": 2.594435736677116, "grad_norm": 3.1718533039093018, "learning_rate": 2.759219574249253e-06, "loss": 1.092, "step": 33105 }, { "epoch": 2.5945141065830724, "grad_norm": 2.680664539337158, "learning_rate": 2.7589439166064104e-06, "loss": 1.2288, "step": 33106 }, { "epoch": 2.5945924764890282, "grad_norm": 3.283439874649048, "learning_rate": 2.758668267487393e-06, "loss": 1.1666, "step": 33107 }, { "epoch": 2.594670846394984, "grad_norm": 5.252228260040283, "learning_rate": 2.7583926268932537e-06, "loss": 2.1992, "step": 33108 }, { "epoch": 2.5947492163009405, "grad_norm": 3.361140727996826, "learning_rate": 2.758116994825033e-06, "loss": 1.4491, "step": 33109 }, { "epoch": 2.594827586206897, "grad_norm": 4.056107997894287, "learning_rate": 2.757841371283786e-06, "loss": 1.5505, "step": 33110 }, { "epoch": 2.5949059561128527, "grad_norm": 2.2820355892181396, "learning_rate": 2.7575657562705584e-06, "loss": 1.1833, "step": 33111 }, { "epoch": 2.5949843260188086, "grad_norm": 2.549649477005005, "learning_rate": 2.7572901497864006e-06, "loss": 1.3618, "step": 33112 }, { "epoch": 2.595062695924765, "grad_norm": 6.995037078857422, "learning_rate": 2.7570145518323577e-06, "loss": 1.8043, "step": 33113 }, { "epoch": 2.5951410658307212, "grad_norm": 10.934309005737305, "learning_rate": 2.7567389624094785e-06, "loss": 1.7788, "step": 33114 }, { "epoch": 2.595219435736677, "grad_norm": 3.0780224800109863, "learning_rate": 2.7564633815188113e-06, "loss": 1.1176, "step": 33115 }, { "epoch": 2.595297805642633, "grad_norm": 3.1455554962158203, "learning_rate": 2.7561878091614086e-06, "loss": 1.247, "step": 33116 }, { "epoch": 2.5953761755485893, "grad_norm": 2.8165273666381836, "learning_rate": 2.7559122453383126e-06, "loss": 1.3295, "step": 33117 }, { "epoch": 2.5954545454545457, "grad_norm": 2.653167247772217, "learning_rate": 2.7556366900505745e-06, "loss": 1.3253, "step": 33118 }, { "epoch": 2.5955329153605016, "grad_norm": 3.7099640369415283, "learning_rate": 2.755361143299241e-06, "loss": 1.2444, "step": 33119 }, { "epoch": 2.5956112852664575, "grad_norm": 3.4636638164520264, "learning_rate": 2.755085605085361e-06, "loss": 1.4311, "step": 33120 }, { "epoch": 2.595689655172414, "grad_norm": 2.9004080295562744, "learning_rate": 2.7548100754099823e-06, "loss": 1.0997, "step": 33121 }, { "epoch": 2.59576802507837, "grad_norm": 3.0555074214935303, "learning_rate": 2.754534554274153e-06, "loss": 1.4198, "step": 33122 }, { "epoch": 2.595846394984326, "grad_norm": 3.374025583267212, "learning_rate": 2.7542590416789206e-06, "loss": 1.3428, "step": 33123 }, { "epoch": 2.595924764890282, "grad_norm": 6.372405052185059, "learning_rate": 2.753983537625333e-06, "loss": 0.512, "step": 33124 }, { "epoch": 2.5960031347962382, "grad_norm": 8.97077751159668, "learning_rate": 2.753708042114439e-06, "loss": 0.6955, "step": 33125 }, { "epoch": 2.5960815047021946, "grad_norm": 2.7671351432800293, "learning_rate": 2.753432555147286e-06, "loss": 1.4462, "step": 33126 }, { "epoch": 2.5961598746081505, "grad_norm": 3.068946361541748, "learning_rate": 2.7531570767249237e-06, "loss": 1.3285, "step": 33127 }, { "epoch": 2.5962382445141063, "grad_norm": 2.4945342540740967, "learning_rate": 2.752881606848393e-06, "loss": 0.9829, "step": 33128 }, { "epoch": 2.5963166144200627, "grad_norm": 3.5691652297973633, "learning_rate": 2.752606145518749e-06, "loss": 2.0106, "step": 33129 }, { "epoch": 2.596394984326019, "grad_norm": 2.828578233718872, "learning_rate": 2.7523306927370375e-06, "loss": 1.3555, "step": 33130 }, { "epoch": 2.596473354231975, "grad_norm": 3.32072114944458, "learning_rate": 2.7520552485043066e-06, "loss": 1.1086, "step": 33131 }, { "epoch": 2.596551724137931, "grad_norm": 2.8805549144744873, "learning_rate": 2.7517798128216013e-06, "loss": 1.2431, "step": 33132 }, { "epoch": 2.596630094043887, "grad_norm": 4.967724800109863, "learning_rate": 2.751504385689969e-06, "loss": 1.4309, "step": 33133 }, { "epoch": 2.5967084639498434, "grad_norm": 3.6611368656158447, "learning_rate": 2.751228967110463e-06, "loss": 1.0335, "step": 33134 }, { "epoch": 2.5967868338557993, "grad_norm": 5.175346374511719, "learning_rate": 2.7509535570841243e-06, "loss": 1.3458, "step": 33135 }, { "epoch": 2.5968652037617552, "grad_norm": 3.2139084339141846, "learning_rate": 2.7506781556120044e-06, "loss": 1.1943, "step": 33136 }, { "epoch": 2.5969435736677116, "grad_norm": 2.6138222217559814, "learning_rate": 2.750402762695148e-06, "loss": 1.5493, "step": 33137 }, { "epoch": 2.597021943573668, "grad_norm": 2.93780255317688, "learning_rate": 2.750127378334605e-06, "loss": 1.6865, "step": 33138 }, { "epoch": 2.597100313479624, "grad_norm": 2.9440131187438965, "learning_rate": 2.749852002531421e-06, "loss": 1.4466, "step": 33139 }, { "epoch": 2.5971786833855797, "grad_norm": 4.218411922454834, "learning_rate": 2.7495766352866452e-06, "loss": 1.275, "step": 33140 }, { "epoch": 2.597257053291536, "grad_norm": 2.7692975997924805, "learning_rate": 2.749301276601324e-06, "loss": 1.6221, "step": 33141 }, { "epoch": 2.5973354231974923, "grad_norm": 6.269169330596924, "learning_rate": 2.749025926476504e-06, "loss": 1.2464, "step": 33142 }, { "epoch": 2.597413793103448, "grad_norm": 10.006290435791016, "learning_rate": 2.7487505849132334e-06, "loss": 1.5288, "step": 33143 }, { "epoch": 2.5974921630094046, "grad_norm": 3.533080816268921, "learning_rate": 2.74847525191256e-06, "loss": 1.0049, "step": 33144 }, { "epoch": 2.5975705329153604, "grad_norm": 3.3393478393554688, "learning_rate": 2.748199927475531e-06, "loss": 1.8101, "step": 33145 }, { "epoch": 2.5976489028213168, "grad_norm": 4.976804733276367, "learning_rate": 2.747924611603192e-06, "loss": 1.6609, "step": 33146 }, { "epoch": 2.5977272727272727, "grad_norm": 3.5531508922576904, "learning_rate": 2.747649304296589e-06, "loss": 1.3886, "step": 33147 }, { "epoch": 2.597805642633229, "grad_norm": 2.8204283714294434, "learning_rate": 2.7473740055567723e-06, "loss": 1.2305, "step": 33148 }, { "epoch": 2.597884012539185, "grad_norm": 9.617612838745117, "learning_rate": 2.7470987153847906e-06, "loss": 1.3415, "step": 33149 }, { "epoch": 2.597962382445141, "grad_norm": 3.1597042083740234, "learning_rate": 2.7468234337816855e-06, "loss": 1.13, "step": 33150 }, { "epoch": 2.598040752351097, "grad_norm": 3.250399112701416, "learning_rate": 2.746548160748508e-06, "loss": 1.4466, "step": 33151 }, { "epoch": 2.5981191222570534, "grad_norm": 4.393336296081543, "learning_rate": 2.7462728962863007e-06, "loss": 0.9998, "step": 33152 }, { "epoch": 2.5981974921630093, "grad_norm": 2.573702812194824, "learning_rate": 2.745997640396118e-06, "loss": 1.6883, "step": 33153 }, { "epoch": 2.5982758620689657, "grad_norm": 2.487933397293091, "learning_rate": 2.7457223930790007e-06, "loss": 1.0496, "step": 33154 }, { "epoch": 2.5983542319749215, "grad_norm": 2.4085476398468018, "learning_rate": 2.7454471543359974e-06, "loss": 1.222, "step": 33155 }, { "epoch": 2.598432601880878, "grad_norm": 2.753264904022217, "learning_rate": 2.7451719241681553e-06, "loss": 1.3985, "step": 33156 }, { "epoch": 2.5985109717868338, "grad_norm": 3.762399911880493, "learning_rate": 2.7448967025765206e-06, "loss": 1.4265, "step": 33157 }, { "epoch": 2.59858934169279, "grad_norm": 4.057318210601807, "learning_rate": 2.7446214895621413e-06, "loss": 1.4676, "step": 33158 }, { "epoch": 2.598667711598746, "grad_norm": 7.484066963195801, "learning_rate": 2.744346285126063e-06, "loss": 1.6409, "step": 33159 }, { "epoch": 2.5987460815047023, "grad_norm": 5.124037742614746, "learning_rate": 2.7440710892693346e-06, "loss": 0.8133, "step": 33160 }, { "epoch": 2.598824451410658, "grad_norm": 3.2955663204193115, "learning_rate": 2.743795901992997e-06, "loss": 1.21, "step": 33161 }, { "epoch": 2.5989028213166145, "grad_norm": 4.152552604675293, "learning_rate": 2.743520723298104e-06, "loss": 1.6568, "step": 33162 }, { "epoch": 2.5989811912225704, "grad_norm": 2.3569185733795166, "learning_rate": 2.7432455531856982e-06, "loss": 1.2349, "step": 33163 }, { "epoch": 2.5990595611285268, "grad_norm": 2.6265456676483154, "learning_rate": 2.7429703916568297e-06, "loss": 1.281, "step": 33164 }, { "epoch": 2.5991379310344827, "grad_norm": 2.416245460510254, "learning_rate": 2.7426952387125393e-06, "loss": 1.4965, "step": 33165 }, { "epoch": 2.599216300940439, "grad_norm": 3.3707151412963867, "learning_rate": 2.742420094353876e-06, "loss": 1.288, "step": 33166 }, { "epoch": 2.599294670846395, "grad_norm": 3.537245988845825, "learning_rate": 2.7421449585818894e-06, "loss": 1.7869, "step": 33167 }, { "epoch": 2.599373040752351, "grad_norm": 2.5159804821014404, "learning_rate": 2.7418698313976243e-06, "loss": 1.1472, "step": 33168 }, { "epoch": 2.599451410658307, "grad_norm": 1.9474233388900757, "learning_rate": 2.741594712802126e-06, "loss": 1.2273, "step": 33169 }, { "epoch": 2.5995297805642634, "grad_norm": 3.770615339279175, "learning_rate": 2.7413196027964406e-06, "loss": 1.2729, "step": 33170 }, { "epoch": 2.5996081504702193, "grad_norm": 3.1930770874023438, "learning_rate": 2.7410445013816134e-06, "loss": 1.7009, "step": 33171 }, { "epoch": 2.5996865203761756, "grad_norm": 9.285940170288086, "learning_rate": 2.7407694085586967e-06, "loss": 1.3425, "step": 33172 }, { "epoch": 2.5997648902821315, "grad_norm": 4.700353622436523, "learning_rate": 2.740494324328731e-06, "loss": 1.6615, "step": 33173 }, { "epoch": 2.599843260188088, "grad_norm": 3.37441086769104, "learning_rate": 2.7402192486927637e-06, "loss": 1.4427, "step": 33174 }, { "epoch": 2.5999216300940438, "grad_norm": 2.2433278560638428, "learning_rate": 2.7399441816518417e-06, "loss": 1.5325, "step": 33175 }, { "epoch": 2.6, "grad_norm": 3.3085696697235107, "learning_rate": 2.7396691232070116e-06, "loss": 1.7249, "step": 33176 }, { "epoch": 2.600078369905956, "grad_norm": 3.5409531593322754, "learning_rate": 2.7393940733593194e-06, "loss": 1.1612, "step": 33177 }, { "epoch": 2.6001567398119123, "grad_norm": 3.3291587829589844, "learning_rate": 2.7391190321098104e-06, "loss": 1.6035, "step": 33178 }, { "epoch": 2.600235109717868, "grad_norm": 4.429830551147461, "learning_rate": 2.738843999459534e-06, "loss": 1.3781, "step": 33179 }, { "epoch": 2.6003134796238245, "grad_norm": 6.12166166305542, "learning_rate": 2.738568975409529e-06, "loss": 0.4676, "step": 33180 }, { "epoch": 2.6003918495297804, "grad_norm": 2.8810389041900635, "learning_rate": 2.7382939599608482e-06, "loss": 1.17, "step": 33181 }, { "epoch": 2.6004702194357368, "grad_norm": 3.8854541778564453, "learning_rate": 2.7380189531145357e-06, "loss": 1.2068, "step": 33182 }, { "epoch": 2.6005485893416926, "grad_norm": 4.087827205657959, "learning_rate": 2.737743954871639e-06, "loss": 1.5046, "step": 33183 }, { "epoch": 2.600626959247649, "grad_norm": 3.06351900100708, "learning_rate": 2.737468965233201e-06, "loss": 1.7554, "step": 33184 }, { "epoch": 2.600705329153605, "grad_norm": 4.252863883972168, "learning_rate": 2.7371939842002666e-06, "loss": 1.3578, "step": 33185 }, { "epoch": 2.600783699059561, "grad_norm": 5.242929458618164, "learning_rate": 2.7369190117738875e-06, "loss": 1.3276, "step": 33186 }, { "epoch": 2.600862068965517, "grad_norm": 3.439089775085449, "learning_rate": 2.736644047955105e-06, "loss": 1.281, "step": 33187 }, { "epoch": 2.6009404388714734, "grad_norm": 3.096461534500122, "learning_rate": 2.736369092744965e-06, "loss": 1.5164, "step": 33188 }, { "epoch": 2.6010188087774293, "grad_norm": 8.20322322845459, "learning_rate": 2.736094146144515e-06, "loss": 0.8102, "step": 33189 }, { "epoch": 2.6010971786833856, "grad_norm": 5.236436367034912, "learning_rate": 2.7358192081547994e-06, "loss": 1.7136, "step": 33190 }, { "epoch": 2.6011755485893415, "grad_norm": 7.615363597869873, "learning_rate": 2.7355442787768662e-06, "loss": 1.493, "step": 33191 }, { "epoch": 2.601253918495298, "grad_norm": 10.4446382522583, "learning_rate": 2.735269358011758e-06, "loss": 0.9153, "step": 33192 }, { "epoch": 2.601332288401254, "grad_norm": 3.199530839920044, "learning_rate": 2.734994445860522e-06, "loss": 1.5187, "step": 33193 }, { "epoch": 2.60141065830721, "grad_norm": 3.055528402328491, "learning_rate": 2.734719542324204e-06, "loss": 1.3916, "step": 33194 }, { "epoch": 2.601489028213166, "grad_norm": 2.7353029251098633, "learning_rate": 2.7344446474038495e-06, "loss": 1.523, "step": 33195 }, { "epoch": 2.6015673981191223, "grad_norm": 4.217339038848877, "learning_rate": 2.7341697611005038e-06, "loss": 1.2593, "step": 33196 }, { "epoch": 2.6016457680250786, "grad_norm": 7.549929618835449, "learning_rate": 2.7338948834152125e-06, "loss": 1.29, "step": 33197 }, { "epoch": 2.6017241379310345, "grad_norm": 2.982431411743164, "learning_rate": 2.733620014349023e-06, "loss": 1.2371, "step": 33198 }, { "epoch": 2.6018025078369904, "grad_norm": 4.7282795906066895, "learning_rate": 2.7333451539029748e-06, "loss": 1.4703, "step": 33199 }, { "epoch": 2.6018808777429467, "grad_norm": 2.9625403881073, "learning_rate": 2.733070302078119e-06, "loss": 1.1802, "step": 33200 }, { "epoch": 2.601959247648903, "grad_norm": 2.2188801765441895, "learning_rate": 2.7327954588755023e-06, "loss": 1.4553, "step": 33201 }, { "epoch": 2.602037617554859, "grad_norm": 3.950364828109741, "learning_rate": 2.732520624296165e-06, "loss": 1.2024, "step": 33202 }, { "epoch": 2.602115987460815, "grad_norm": 4.155874729156494, "learning_rate": 2.732245798341154e-06, "loss": 1.3662, "step": 33203 }, { "epoch": 2.602194357366771, "grad_norm": 12.47464656829834, "learning_rate": 2.7319709810115136e-06, "loss": 1.1171, "step": 33204 }, { "epoch": 2.6022727272727275, "grad_norm": 2.7152199745178223, "learning_rate": 2.7316961723082947e-06, "loss": 1.398, "step": 33205 }, { "epoch": 2.6023510971786834, "grad_norm": 5.27495813369751, "learning_rate": 2.7314213722325366e-06, "loss": 1.0977, "step": 33206 }, { "epoch": 2.6024294670846393, "grad_norm": 4.5323710441589355, "learning_rate": 2.7311465807852868e-06, "loss": 1.8102, "step": 33207 }, { "epoch": 2.6025078369905956, "grad_norm": 2.9631059169769287, "learning_rate": 2.730871797967589e-06, "loss": 1.5505, "step": 33208 }, { "epoch": 2.602586206896552, "grad_norm": 2.5481932163238525, "learning_rate": 2.73059702378049e-06, "loss": 1.0364, "step": 33209 }, { "epoch": 2.602664576802508, "grad_norm": 3.368124485015869, "learning_rate": 2.730322258225034e-06, "loss": 1.3357, "step": 33210 }, { "epoch": 2.6027429467084637, "grad_norm": 3.336229085922241, "learning_rate": 2.7300475013022666e-06, "loss": 1.1105, "step": 33211 }, { "epoch": 2.60282131661442, "grad_norm": 3.9188711643218994, "learning_rate": 2.729772753013232e-06, "loss": 1.1432, "step": 33212 }, { "epoch": 2.6028996865203764, "grad_norm": 2.4044885635375977, "learning_rate": 2.7294980133589766e-06, "loss": 0.9206, "step": 33213 }, { "epoch": 2.6029780564263323, "grad_norm": 3.1309425830841064, "learning_rate": 2.729223282340544e-06, "loss": 1.4562, "step": 33214 }, { "epoch": 2.603056426332288, "grad_norm": 5.90895414352417, "learning_rate": 2.7289485599589805e-06, "loss": 1.4519, "step": 33215 }, { "epoch": 2.6031347962382445, "grad_norm": 6.012948513031006, "learning_rate": 2.728673846215332e-06, "loss": 1.0203, "step": 33216 }, { "epoch": 2.603213166144201, "grad_norm": 3.4485154151916504, "learning_rate": 2.728399141110637e-06, "loss": 1.3229, "step": 33217 }, { "epoch": 2.6032915360501567, "grad_norm": 5.112033843994141, "learning_rate": 2.7281244446459464e-06, "loss": 1.5572, "step": 33218 }, { "epoch": 2.6033699059561126, "grad_norm": 6.1399126052856445, "learning_rate": 2.7278497568223043e-06, "loss": 1.3966, "step": 33219 }, { "epoch": 2.603448275862069, "grad_norm": 5.145654678344727, "learning_rate": 2.7275750776407568e-06, "loss": 1.4793, "step": 33220 }, { "epoch": 2.6035266457680253, "grad_norm": 3.915818691253662, "learning_rate": 2.727300407102344e-06, "loss": 1.3801, "step": 33221 }, { "epoch": 2.603605015673981, "grad_norm": 2.41180682182312, "learning_rate": 2.7270257452081118e-06, "loss": 1.4192, "step": 33222 }, { "epoch": 2.603683385579937, "grad_norm": 7.5426812171936035, "learning_rate": 2.7267510919591077e-06, "loss": 1.3274, "step": 33223 }, { "epoch": 2.6037617554858934, "grad_norm": 3.419312000274658, "learning_rate": 2.7264764473563767e-06, "loss": 1.2983, "step": 33224 }, { "epoch": 2.6038401253918497, "grad_norm": 2.8709716796875, "learning_rate": 2.7262018114009597e-06, "loss": 1.5373, "step": 33225 }, { "epoch": 2.6039184952978056, "grad_norm": 2.1316373348236084, "learning_rate": 2.725927184093903e-06, "loss": 1.3338, "step": 33226 }, { "epoch": 2.6039968652037615, "grad_norm": 4.651147365570068, "learning_rate": 2.7256525654362497e-06, "loss": 1.0881, "step": 33227 }, { "epoch": 2.604075235109718, "grad_norm": 5.569235324859619, "learning_rate": 2.7253779554290492e-06, "loss": 1.4647, "step": 33228 }, { "epoch": 2.604153605015674, "grad_norm": 8.150967597961426, "learning_rate": 2.7251033540733413e-06, "loss": 1.0823, "step": 33229 }, { "epoch": 2.60423197492163, "grad_norm": 2.711934804916382, "learning_rate": 2.724828761370172e-06, "loss": 1.4821, "step": 33230 }, { "epoch": 2.604310344827586, "grad_norm": 3.936176061630249, "learning_rate": 2.724554177320584e-06, "loss": 1.552, "step": 33231 }, { "epoch": 2.6043887147335423, "grad_norm": 2.4303171634674072, "learning_rate": 2.7242796019256244e-06, "loss": 1.3212, "step": 33232 }, { "epoch": 2.6044670846394986, "grad_norm": 2.454929828643799, "learning_rate": 2.724005035186336e-06, "loss": 1.1068, "step": 33233 }, { "epoch": 2.6045454545454545, "grad_norm": 4.911956787109375, "learning_rate": 2.723730477103764e-06, "loss": 1.9849, "step": 33234 }, { "epoch": 2.6046238244514104, "grad_norm": 5.278983116149902, "learning_rate": 2.723455927678953e-06, "loss": 1.3624, "step": 33235 }, { "epoch": 2.6047021943573667, "grad_norm": 2.8900907039642334, "learning_rate": 2.723181386912943e-06, "loss": 1.1355, "step": 33236 }, { "epoch": 2.604780564263323, "grad_norm": 3.3046793937683105, "learning_rate": 2.722906854806784e-06, "loss": 1.1172, "step": 33237 }, { "epoch": 2.604858934169279, "grad_norm": 3.203693389892578, "learning_rate": 2.7226323313615167e-06, "loss": 1.3083, "step": 33238 }, { "epoch": 2.6049373040752353, "grad_norm": 2.5833933353424072, "learning_rate": 2.7223578165781888e-06, "loss": 1.1039, "step": 33239 }, { "epoch": 2.605015673981191, "grad_norm": 9.533966064453125, "learning_rate": 2.7220833104578403e-06, "loss": 0.6636, "step": 33240 }, { "epoch": 2.6050940438871475, "grad_norm": 2.8451433181762695, "learning_rate": 2.721808813001514e-06, "loss": 1.3091, "step": 33241 }, { "epoch": 2.6051724137931034, "grad_norm": 3.962916612625122, "learning_rate": 2.7215343242102616e-06, "loss": 1.3729, "step": 33242 }, { "epoch": 2.6052507836990597, "grad_norm": 2.2836687564849854, "learning_rate": 2.72125984408512e-06, "loss": 1.393, "step": 33243 }, { "epoch": 2.6053291536050156, "grad_norm": 3.5657267570495605, "learning_rate": 2.7209853726271353e-06, "loss": 1.3897, "step": 33244 }, { "epoch": 2.605407523510972, "grad_norm": 3.4109463691711426, "learning_rate": 2.720710909837352e-06, "loss": 1.173, "step": 33245 }, { "epoch": 2.605485893416928, "grad_norm": 2.23421311378479, "learning_rate": 2.720436455716814e-06, "loss": 1.1879, "step": 33246 }, { "epoch": 2.605564263322884, "grad_norm": 4.008098125457764, "learning_rate": 2.7201620102665637e-06, "loss": 1.4348, "step": 33247 }, { "epoch": 2.60564263322884, "grad_norm": 3.847877025604248, "learning_rate": 2.719887573487647e-06, "loss": 1.0858, "step": 33248 }, { "epoch": 2.6057210031347964, "grad_norm": 2.3667209148406982, "learning_rate": 2.719613145381107e-06, "loss": 1.0412, "step": 33249 }, { "epoch": 2.6057993730407523, "grad_norm": 6.9757537841796875, "learning_rate": 2.719338725947987e-06, "loss": 1.7406, "step": 33250 }, { "epoch": 2.6058777429467086, "grad_norm": 3.68853497505188, "learning_rate": 2.7190643151893307e-06, "loss": 1.4208, "step": 33251 }, { "epoch": 2.6059561128526645, "grad_norm": 3.273798942565918, "learning_rate": 2.718789913106183e-06, "loss": 1.3639, "step": 33252 }, { "epoch": 2.606034482758621, "grad_norm": 3.5001399517059326, "learning_rate": 2.718515519699586e-06, "loss": 1.3902, "step": 33253 }, { "epoch": 2.6061128526645767, "grad_norm": 3.7869248390197754, "learning_rate": 2.7182411349705867e-06, "loss": 1.4158, "step": 33254 }, { "epoch": 2.606191222570533, "grad_norm": 3.2008655071258545, "learning_rate": 2.717966758920222e-06, "loss": 1.3555, "step": 33255 }, { "epoch": 2.606269592476489, "grad_norm": 22.381839752197266, "learning_rate": 2.7176923915495413e-06, "loss": 1.975, "step": 33256 }, { "epoch": 2.6063479623824453, "grad_norm": 8.240885734558105, "learning_rate": 2.717418032859589e-06, "loss": 0.9931, "step": 33257 }, { "epoch": 2.606426332288401, "grad_norm": 2.7997167110443115, "learning_rate": 2.7171436828514033e-06, "loss": 1.0696, "step": 33258 }, { "epoch": 2.6065047021943575, "grad_norm": 3.240753650665283, "learning_rate": 2.716869341526031e-06, "loss": 1.2052, "step": 33259 }, { "epoch": 2.6065830721003134, "grad_norm": 2.4846346378326416, "learning_rate": 2.716595008884513e-06, "loss": 1.3149, "step": 33260 }, { "epoch": 2.6066614420062697, "grad_norm": 3.1089508533477783, "learning_rate": 2.716320684927899e-06, "loss": 1.4992, "step": 33261 }, { "epoch": 2.6067398119122256, "grad_norm": 3.283236026763916, "learning_rate": 2.716046369657226e-06, "loss": 1.0543, "step": 33262 }, { "epoch": 2.606818181818182, "grad_norm": 1.9746286869049072, "learning_rate": 2.715772063073539e-06, "loss": 1.4915, "step": 33263 }, { "epoch": 2.606896551724138, "grad_norm": 2.96479868888855, "learning_rate": 2.7154977651778823e-06, "loss": 1.0389, "step": 33264 }, { "epoch": 2.606974921630094, "grad_norm": 3.893434524536133, "learning_rate": 2.7152234759712986e-06, "loss": 1.4114, "step": 33265 }, { "epoch": 2.60705329153605, "grad_norm": 2.4473044872283936, "learning_rate": 2.714949195454831e-06, "loss": 1.5974, "step": 33266 }, { "epoch": 2.6071316614420064, "grad_norm": 3.8266453742980957, "learning_rate": 2.714674923629523e-06, "loss": 1.1385, "step": 33267 }, { "epoch": 2.6072100313479623, "grad_norm": 2.3761165142059326, "learning_rate": 2.714400660496419e-06, "loss": 1.19, "step": 33268 }, { "epoch": 2.6072884012539186, "grad_norm": 1.9856220483779907, "learning_rate": 2.7141264060565596e-06, "loss": 1.1745, "step": 33269 }, { "epoch": 2.6073667711598745, "grad_norm": 5.834471225738525, "learning_rate": 2.7138521603109895e-06, "loss": 1.233, "step": 33270 }, { "epoch": 2.607445141065831, "grad_norm": 3.067185878753662, "learning_rate": 2.7135779232607517e-06, "loss": 1.047, "step": 33271 }, { "epoch": 2.6075235109717867, "grad_norm": 3.2143712043762207, "learning_rate": 2.7133036949068915e-06, "loss": 1.1202, "step": 33272 }, { "epoch": 2.607601880877743, "grad_norm": 3.7449347972869873, "learning_rate": 2.7130294752504465e-06, "loss": 1.317, "step": 33273 }, { "epoch": 2.607680250783699, "grad_norm": 3.4387032985687256, "learning_rate": 2.7127552642924614e-06, "loss": 1.1217, "step": 33274 }, { "epoch": 2.6077586206896552, "grad_norm": 4.194291114807129, "learning_rate": 2.7124810620339825e-06, "loss": 1.2648, "step": 33275 }, { "epoch": 2.607836990595611, "grad_norm": 2.594810724258423, "learning_rate": 2.712206868476053e-06, "loss": 1.3249, "step": 33276 }, { "epoch": 2.6079153605015675, "grad_norm": 5.06349515914917, "learning_rate": 2.7119326836197104e-06, "loss": 1.2792, "step": 33277 }, { "epoch": 2.6079937304075234, "grad_norm": 4.355367660522461, "learning_rate": 2.7116585074660017e-06, "loss": 1.247, "step": 33278 }, { "epoch": 2.6080721003134797, "grad_norm": 2.701417922973633, "learning_rate": 2.7113843400159657e-06, "loss": 1.5352, "step": 33279 }, { "epoch": 2.6081504702194356, "grad_norm": 4.029747486114502, "learning_rate": 2.711110181270653e-06, "loss": 1.5098, "step": 33280 }, { "epoch": 2.608228840125392, "grad_norm": 2.974243640899658, "learning_rate": 2.7108360312310993e-06, "loss": 1.5357, "step": 33281 }, { "epoch": 2.608307210031348, "grad_norm": 14.123933792114258, "learning_rate": 2.7105618898983492e-06, "loss": 1.0173, "step": 33282 }, { "epoch": 2.608385579937304, "grad_norm": 2.1798031330108643, "learning_rate": 2.7102877572734465e-06, "loss": 1.251, "step": 33283 }, { "epoch": 2.6084639498432605, "grad_norm": 2.733372449874878, "learning_rate": 2.7100136333574323e-06, "loss": 0.8006, "step": 33284 }, { "epoch": 2.6085423197492164, "grad_norm": 4.134943962097168, "learning_rate": 2.709739518151351e-06, "loss": 2.047, "step": 33285 }, { "epoch": 2.6086206896551722, "grad_norm": 4.627043724060059, "learning_rate": 2.709465411656243e-06, "loss": 1.4824, "step": 33286 }, { "epoch": 2.6086990595611286, "grad_norm": 10.680887222290039, "learning_rate": 2.709191313873155e-06, "loss": 1.0479, "step": 33287 }, { "epoch": 2.608777429467085, "grad_norm": 6.808894157409668, "learning_rate": 2.708917224803122e-06, "loss": 1.5346, "step": 33288 }, { "epoch": 2.608855799373041, "grad_norm": 8.429337501525879, "learning_rate": 2.708643144447193e-06, "loss": 0.893, "step": 33289 }, { "epoch": 2.6089341692789967, "grad_norm": 2.7351081371307373, "learning_rate": 2.7083690728064095e-06, "loss": 1.4722, "step": 33290 }, { "epoch": 2.609012539184953, "grad_norm": 13.620320320129395, "learning_rate": 2.708095009881814e-06, "loss": 1.6942, "step": 33291 }, { "epoch": 2.6090909090909093, "grad_norm": 3.030946731567383, "learning_rate": 2.7078209556744466e-06, "loss": 1.0569, "step": 33292 }, { "epoch": 2.6091692789968652, "grad_norm": 4.518736362457275, "learning_rate": 2.707546910185349e-06, "loss": 1.4291, "step": 33293 }, { "epoch": 2.609247648902821, "grad_norm": 3.7814507484436035, "learning_rate": 2.7072728734155667e-06, "loss": 1.7189, "step": 33294 }, { "epoch": 2.6093260188087775, "grad_norm": 4.351171970367432, "learning_rate": 2.706998845366143e-06, "loss": 1.2753, "step": 33295 }, { "epoch": 2.609404388714734, "grad_norm": 4.9915595054626465, "learning_rate": 2.7067248260381164e-06, "loss": 1.3713, "step": 33296 }, { "epoch": 2.6094827586206897, "grad_norm": 3.3899803161621094, "learning_rate": 2.706450815432531e-06, "loss": 1.2609, "step": 33297 }, { "epoch": 2.6095611285266456, "grad_norm": 5.495247840881348, "learning_rate": 2.7061768135504275e-06, "loss": 1.9931, "step": 33298 }, { "epoch": 2.609639498432602, "grad_norm": 3.405824899673462, "learning_rate": 2.7059028203928495e-06, "loss": 1.3342, "step": 33299 }, { "epoch": 2.6097178683385582, "grad_norm": 5.861258029937744, "learning_rate": 2.7056288359608395e-06, "loss": 1.1695, "step": 33300 }, { "epoch": 2.609796238244514, "grad_norm": 3.443984031677246, "learning_rate": 2.7053548602554386e-06, "loss": 1.2454, "step": 33301 }, { "epoch": 2.60987460815047, "grad_norm": 4.274213790893555, "learning_rate": 2.7050808932776894e-06, "loss": 1.4382, "step": 33302 }, { "epoch": 2.6099529780564263, "grad_norm": 2.6144416332244873, "learning_rate": 2.7048069350286333e-06, "loss": 1.3385, "step": 33303 }, { "epoch": 2.6100313479623827, "grad_norm": 2.872485876083374, "learning_rate": 2.7045329855093127e-06, "loss": 0.8763, "step": 33304 }, { "epoch": 2.6101097178683386, "grad_norm": 4.639519214630127, "learning_rate": 2.70425904472077e-06, "loss": 1.654, "step": 33305 }, { "epoch": 2.6101880877742945, "grad_norm": 7.787546157836914, "learning_rate": 2.7039851126640464e-06, "loss": 1.2621, "step": 33306 }, { "epoch": 2.610266457680251, "grad_norm": 8.359797477722168, "learning_rate": 2.703711189340185e-06, "loss": 1.1245, "step": 33307 }, { "epoch": 2.610344827586207, "grad_norm": 2.327036142349243, "learning_rate": 2.7034372747502257e-06, "loss": 1.2287, "step": 33308 }, { "epoch": 2.610423197492163, "grad_norm": 2.8696653842926025, "learning_rate": 2.703163368895212e-06, "loss": 1.1067, "step": 33309 }, { "epoch": 2.610501567398119, "grad_norm": 2.1504430770874023, "learning_rate": 2.7028894717761867e-06, "loss": 1.2832, "step": 33310 }, { "epoch": 2.6105799373040752, "grad_norm": 6.392391204833984, "learning_rate": 2.702615583394188e-06, "loss": 1.591, "step": 33311 }, { "epoch": 2.6106583072100316, "grad_norm": 4.050132751464844, "learning_rate": 2.702341703750258e-06, "loss": 1.6409, "step": 33312 }, { "epoch": 2.6107366771159874, "grad_norm": 2.604738712310791, "learning_rate": 2.7020678328454436e-06, "loss": 1.5192, "step": 33313 }, { "epoch": 2.6108150470219433, "grad_norm": 3.9862735271453857, "learning_rate": 2.701793970680781e-06, "loss": 1.1631, "step": 33314 }, { "epoch": 2.6108934169278997, "grad_norm": 2.2598464488983154, "learning_rate": 2.701520117257314e-06, "loss": 1.5685, "step": 33315 }, { "epoch": 2.610971786833856, "grad_norm": 1.9906961917877197, "learning_rate": 2.7012462725760834e-06, "loss": 1.3353, "step": 33316 }, { "epoch": 2.611050156739812, "grad_norm": 3.4442408084869385, "learning_rate": 2.7009724366381318e-06, "loss": 1.2425, "step": 33317 }, { "epoch": 2.6111285266457678, "grad_norm": 2.6526904106140137, "learning_rate": 2.7006986094444998e-06, "loss": 1.3755, "step": 33318 }, { "epoch": 2.611206896551724, "grad_norm": 2.346599817276001, "learning_rate": 2.70042479099623e-06, "loss": 1.3226, "step": 33319 }, { "epoch": 2.6112852664576804, "grad_norm": 3.754042387008667, "learning_rate": 2.700150981294362e-06, "loss": 1.3731, "step": 33320 }, { "epoch": 2.6113636363636363, "grad_norm": 3.9323079586029053, "learning_rate": 2.6998771803399393e-06, "loss": 1.4067, "step": 33321 }, { "epoch": 2.611442006269592, "grad_norm": 3.9539079666137695, "learning_rate": 2.6996033881340023e-06, "loss": 1.6244, "step": 33322 }, { "epoch": 2.6115203761755486, "grad_norm": 3.7949397563934326, "learning_rate": 2.699329604677592e-06, "loss": 1.0352, "step": 33323 }, { "epoch": 2.611598746081505, "grad_norm": 3.493929386138916, "learning_rate": 2.699055829971751e-06, "loss": 1.0042, "step": 33324 }, { "epoch": 2.6116771159874608, "grad_norm": 2.6142897605895996, "learning_rate": 2.698782064017519e-06, "loss": 1.3475, "step": 33325 }, { "epoch": 2.6117554858934167, "grad_norm": 2.9474847316741943, "learning_rate": 2.698508306815938e-06, "loss": 1.1621, "step": 33326 }, { "epoch": 2.611833855799373, "grad_norm": 4.333774566650391, "learning_rate": 2.698234558368049e-06, "loss": 1.5728, "step": 33327 }, { "epoch": 2.6119122257053293, "grad_norm": 4.233912467956543, "learning_rate": 2.697960818674896e-06, "loss": 1.4856, "step": 33328 }, { "epoch": 2.611990595611285, "grad_norm": 3.71917986869812, "learning_rate": 2.697687087737516e-06, "loss": 1.0801, "step": 33329 }, { "epoch": 2.612068965517241, "grad_norm": 3.4195752143859863, "learning_rate": 2.6974133655569495e-06, "loss": 1.4926, "step": 33330 }, { "epoch": 2.6121473354231974, "grad_norm": 5.270871639251709, "learning_rate": 2.697139652134242e-06, "loss": 1.6491, "step": 33331 }, { "epoch": 2.6122257053291538, "grad_norm": 4.86049747467041, "learning_rate": 2.696865947470434e-06, "loss": 1.2318, "step": 33332 }, { "epoch": 2.6123040752351097, "grad_norm": 3.3786027431488037, "learning_rate": 2.696592251566563e-06, "loss": 1.0092, "step": 33333 }, { "epoch": 2.612382445141066, "grad_norm": 5.678836345672607, "learning_rate": 2.6963185644236722e-06, "loss": 1.3066, "step": 33334 }, { "epoch": 2.612460815047022, "grad_norm": 5.163253307342529, "learning_rate": 2.6960448860428002e-06, "loss": 1.4049, "step": 33335 }, { "epoch": 2.612539184952978, "grad_norm": 3.574585437774658, "learning_rate": 2.6957712164249944e-06, "loss": 1.2904, "step": 33336 }, { "epoch": 2.612617554858934, "grad_norm": 2.12378191947937, "learning_rate": 2.6954975555712894e-06, "loss": 0.9586, "step": 33337 }, { "epoch": 2.6126959247648904, "grad_norm": 3.3431708812713623, "learning_rate": 2.6952239034827275e-06, "loss": 1.6258, "step": 33338 }, { "epoch": 2.6127742946708463, "grad_norm": 3.2672667503356934, "learning_rate": 2.6949502601603506e-06, "loss": 1.6041, "step": 33339 }, { "epoch": 2.6128526645768027, "grad_norm": 3.2497622966766357, "learning_rate": 2.6946766256051983e-06, "loss": 1.2788, "step": 33340 }, { "epoch": 2.6129310344827585, "grad_norm": 3.285935401916504, "learning_rate": 2.694402999818313e-06, "loss": 1.4358, "step": 33341 }, { "epoch": 2.613009404388715, "grad_norm": 3.231506586074829, "learning_rate": 2.6941293828007335e-06, "loss": 1.3351, "step": 33342 }, { "epoch": 2.6130877742946708, "grad_norm": 3.9052236080169678, "learning_rate": 2.693855774553504e-06, "loss": 1.4374, "step": 33343 }, { "epoch": 2.613166144200627, "grad_norm": 2.916233539581299, "learning_rate": 2.6935821750776583e-06, "loss": 1.3825, "step": 33344 }, { "epoch": 2.613244514106583, "grad_norm": 5.192075252532959, "learning_rate": 2.6933085843742436e-06, "loss": 1.5404, "step": 33345 }, { "epoch": 2.6133228840125393, "grad_norm": 2.6682791709899902, "learning_rate": 2.6930350024442984e-06, "loss": 1.3695, "step": 33346 }, { "epoch": 2.613401253918495, "grad_norm": 3.153007984161377, "learning_rate": 2.692761429288865e-06, "loss": 1.1517, "step": 33347 }, { "epoch": 2.6134796238244515, "grad_norm": 8.682143211364746, "learning_rate": 2.69248786490898e-06, "loss": 1.3519, "step": 33348 }, { "epoch": 2.6135579937304074, "grad_norm": 4.651479244232178, "learning_rate": 2.6922143093056836e-06, "loss": 1.2714, "step": 33349 }, { "epoch": 2.6136363636363638, "grad_norm": 3.3829452991485596, "learning_rate": 2.691940762480021e-06, "loss": 0.8494, "step": 33350 }, { "epoch": 2.6137147335423196, "grad_norm": 2.919889211654663, "learning_rate": 2.691667224433032e-06, "loss": 1.0827, "step": 33351 }, { "epoch": 2.613793103448276, "grad_norm": 3.9024758338928223, "learning_rate": 2.6913936951657536e-06, "loss": 1.3778, "step": 33352 }, { "epoch": 2.613871473354232, "grad_norm": 3.8627424240112305, "learning_rate": 2.6911201746792277e-06, "loss": 1.2626, "step": 33353 }, { "epoch": 2.613949843260188, "grad_norm": 2.8728249073028564, "learning_rate": 2.6908466629744947e-06, "loss": 1.3042, "step": 33354 }, { "epoch": 2.614028213166144, "grad_norm": 6.8385090827941895, "learning_rate": 2.6905731600525946e-06, "loss": 0.4892, "step": 33355 }, { "epoch": 2.6141065830721004, "grad_norm": 2.6564764976501465, "learning_rate": 2.6902996659145684e-06, "loss": 1.1618, "step": 33356 }, { "epoch": 2.6141849529780563, "grad_norm": 4.540473461151123, "learning_rate": 2.690026180561456e-06, "loss": 1.622, "step": 33357 }, { "epoch": 2.6142633228840126, "grad_norm": 2.49276065826416, "learning_rate": 2.6897527039942974e-06, "loss": 1.1274, "step": 33358 }, { "epoch": 2.6143416927899685, "grad_norm": 3.572744846343994, "learning_rate": 2.6894792362141324e-06, "loss": 1.3418, "step": 33359 }, { "epoch": 2.614420062695925, "grad_norm": 10.777328491210938, "learning_rate": 2.689205777222003e-06, "loss": 1.7512, "step": 33360 }, { "epoch": 2.6144984326018808, "grad_norm": 2.577871322631836, "learning_rate": 2.688932327018946e-06, "loss": 1.2651, "step": 33361 }, { "epoch": 2.614576802507837, "grad_norm": 3.3769705295562744, "learning_rate": 2.6886588856060068e-06, "loss": 1.3336, "step": 33362 }, { "epoch": 2.614655172413793, "grad_norm": 2.509754180908203, "learning_rate": 2.6883854529842173e-06, "loss": 1.3289, "step": 33363 }, { "epoch": 2.6147335423197493, "grad_norm": 10.111955642700195, "learning_rate": 2.6881120291546246e-06, "loss": 1.3288, "step": 33364 }, { "epoch": 2.614811912225705, "grad_norm": 3.933136224746704, "learning_rate": 2.6878386141182666e-06, "loss": 1.1497, "step": 33365 }, { "epoch": 2.6148902821316615, "grad_norm": 3.587449789047241, "learning_rate": 2.687565207876184e-06, "loss": 1.829, "step": 33366 }, { "epoch": 2.6149686520376174, "grad_norm": 3.4522178173065186, "learning_rate": 2.687291810429414e-06, "loss": 1.1355, "step": 33367 }, { "epoch": 2.6150470219435737, "grad_norm": 5.015069484710693, "learning_rate": 2.6870184217789964e-06, "loss": 1.4757, "step": 33368 }, { "epoch": 2.6151253918495296, "grad_norm": 4.017014026641846, "learning_rate": 2.6867450419259767e-06, "loss": 1.2157, "step": 33369 }, { "epoch": 2.615203761755486, "grad_norm": 3.165724277496338, "learning_rate": 2.6864716708713885e-06, "loss": 1.781, "step": 33370 }, { "epoch": 2.615282131661442, "grad_norm": 3.964672327041626, "learning_rate": 2.6861983086162737e-06, "loss": 1.5083, "step": 33371 }, { "epoch": 2.615360501567398, "grad_norm": 3.2890264987945557, "learning_rate": 2.685924955161672e-06, "loss": 0.9385, "step": 33372 }, { "epoch": 2.615438871473354, "grad_norm": 15.708828926086426, "learning_rate": 2.6856516105086238e-06, "loss": 0.8347, "step": 33373 }, { "epoch": 2.6155172413793104, "grad_norm": 2.295531988143921, "learning_rate": 2.6853782746581674e-06, "loss": 1.3935, "step": 33374 }, { "epoch": 2.6155956112852663, "grad_norm": 4.7671613693237305, "learning_rate": 2.6851049476113433e-06, "loss": 1.5113, "step": 33375 }, { "epoch": 2.6156739811912226, "grad_norm": 1.9225401878356934, "learning_rate": 2.684831629369191e-06, "loss": 1.362, "step": 33376 }, { "epoch": 2.6157523510971785, "grad_norm": 2.6989240646362305, "learning_rate": 2.6845583199327508e-06, "loss": 1.3484, "step": 33377 }, { "epoch": 2.615830721003135, "grad_norm": 6.913963794708252, "learning_rate": 2.6842850193030606e-06, "loss": 0.794, "step": 33378 }, { "epoch": 2.615909090909091, "grad_norm": 3.6108224391937256, "learning_rate": 2.6840117274811604e-06, "loss": 1.2245, "step": 33379 }, { "epoch": 2.615987460815047, "grad_norm": 3.4292280673980713, "learning_rate": 2.6837384444680915e-06, "loss": 1.2989, "step": 33380 }, { "epoch": 2.616065830721003, "grad_norm": 5.8765645027160645, "learning_rate": 2.683465170264893e-06, "loss": 1.0947, "step": 33381 }, { "epoch": 2.6161442006269593, "grad_norm": 6.685393333435059, "learning_rate": 2.6831919048725984e-06, "loss": 1.6237, "step": 33382 }, { "epoch": 2.6162225705329156, "grad_norm": 5.637969493865967, "learning_rate": 2.682918648292255e-06, "loss": 1.2062, "step": 33383 }, { "epoch": 2.6163009404388715, "grad_norm": 2.1561875343322754, "learning_rate": 2.6826454005249003e-06, "loss": 1.2043, "step": 33384 }, { "epoch": 2.6163793103448274, "grad_norm": 4.461488246917725, "learning_rate": 2.6823721615715703e-06, "loss": 1.5949, "step": 33385 }, { "epoch": 2.6164576802507837, "grad_norm": 2.8375744819641113, "learning_rate": 2.6820989314333062e-06, "loss": 1.4854, "step": 33386 }, { "epoch": 2.61653605015674, "grad_norm": 4.059154033660889, "learning_rate": 2.6818257101111457e-06, "loss": 1.6436, "step": 33387 }, { "epoch": 2.616614420062696, "grad_norm": 5.561277866363525, "learning_rate": 2.6815524976061324e-06, "loss": 1.5272, "step": 33388 }, { "epoch": 2.616692789968652, "grad_norm": 4.322843074798584, "learning_rate": 2.6812792939193016e-06, "loss": 1.1493, "step": 33389 }, { "epoch": 2.616771159874608, "grad_norm": 4.129417419433594, "learning_rate": 2.6810060990516928e-06, "loss": 1.4685, "step": 33390 }, { "epoch": 2.6168495297805645, "grad_norm": 3.8002822399139404, "learning_rate": 2.6807329130043464e-06, "loss": 1.0781, "step": 33391 }, { "epoch": 2.6169278996865204, "grad_norm": 3.708467721939087, "learning_rate": 2.6804597357783e-06, "loss": 1.5715, "step": 33392 }, { "epoch": 2.6170062695924763, "grad_norm": 3.3373098373413086, "learning_rate": 2.6801865673745942e-06, "loss": 1.5751, "step": 33393 }, { "epoch": 2.6170846394984326, "grad_norm": 4.500013828277588, "learning_rate": 2.679913407794266e-06, "loss": 1.3332, "step": 33394 }, { "epoch": 2.617163009404389, "grad_norm": 3.177647829055786, "learning_rate": 2.679640257038356e-06, "loss": 1.5501, "step": 33395 }, { "epoch": 2.617241379310345, "grad_norm": 10.869637489318848, "learning_rate": 2.679367115107903e-06, "loss": 1.5085, "step": 33396 }, { "epoch": 2.6173197492163007, "grad_norm": 4.560400485992432, "learning_rate": 2.6790939820039458e-06, "loss": 1.3062, "step": 33397 }, { "epoch": 2.617398119122257, "grad_norm": 4.508079528808594, "learning_rate": 2.678820857727522e-06, "loss": 0.7593, "step": 33398 }, { "epoch": 2.6174764890282134, "grad_norm": 8.143563270568848, "learning_rate": 2.6785477422796746e-06, "loss": 1.566, "step": 33399 }, { "epoch": 2.6175548589341693, "grad_norm": 3.5491533279418945, "learning_rate": 2.6782746356614364e-06, "loss": 1.3568, "step": 33400 }, { "epoch": 2.617633228840125, "grad_norm": 3.4200050830841064, "learning_rate": 2.678001537873847e-06, "loss": 1.2744, "step": 33401 }, { "epoch": 2.6177115987460815, "grad_norm": 3.8601176738739014, "learning_rate": 2.6777284489179496e-06, "loss": 1.0416, "step": 33402 }, { "epoch": 2.617789968652038, "grad_norm": 2.624387264251709, "learning_rate": 2.6774553687947817e-06, "loss": 1.6714, "step": 33403 }, { "epoch": 2.6178683385579937, "grad_norm": 4.4911088943481445, "learning_rate": 2.6771822975053793e-06, "loss": 1.4299, "step": 33404 }, { "epoch": 2.6179467084639496, "grad_norm": 3.3418450355529785, "learning_rate": 2.676909235050782e-06, "loss": 1.4258, "step": 33405 }, { "epoch": 2.618025078369906, "grad_norm": 3.3802876472473145, "learning_rate": 2.676636181432027e-06, "loss": 1.5933, "step": 33406 }, { "epoch": 2.6181034482758623, "grad_norm": 3.630136728286743, "learning_rate": 2.676363136650158e-06, "loss": 1.3533, "step": 33407 }, { "epoch": 2.618181818181818, "grad_norm": 3.561622381210327, "learning_rate": 2.676090100706209e-06, "loss": 0.8752, "step": 33408 }, { "epoch": 2.618260188087774, "grad_norm": 11.92353630065918, "learning_rate": 2.675817073601219e-06, "loss": 0.7238, "step": 33409 }, { "epoch": 2.6183385579937304, "grad_norm": 6.387729167938232, "learning_rate": 2.675544055336227e-06, "loss": 1.8172, "step": 33410 }, { "epoch": 2.6184169278996867, "grad_norm": 2.0844507217407227, "learning_rate": 2.675271045912272e-06, "loss": 1.0955, "step": 33411 }, { "epoch": 2.6184952978056426, "grad_norm": 2.993457317352295, "learning_rate": 2.6749980453303914e-06, "loss": 1.4081, "step": 33412 }, { "epoch": 2.6185736677115985, "grad_norm": 3.991187334060669, "learning_rate": 2.6747250535916247e-06, "loss": 1.6289, "step": 33413 }, { "epoch": 2.618652037617555, "grad_norm": 2.0109963417053223, "learning_rate": 2.674452070697009e-06, "loss": 1.367, "step": 33414 }, { "epoch": 2.618730407523511, "grad_norm": 5.247916221618652, "learning_rate": 2.6741790966475835e-06, "loss": 1.6447, "step": 33415 }, { "epoch": 2.618808777429467, "grad_norm": 2.778801202774048, "learning_rate": 2.6739061314443855e-06, "loss": 1.2688, "step": 33416 }, { "epoch": 2.618887147335423, "grad_norm": 3.2821269035339355, "learning_rate": 2.6736331750884542e-06, "loss": 1.3131, "step": 33417 }, { "epoch": 2.6189655172413793, "grad_norm": 7.097739219665527, "learning_rate": 2.673360227580829e-06, "loss": 1.566, "step": 33418 }, { "epoch": 2.6190438871473356, "grad_norm": 3.030529022216797, "learning_rate": 2.6730872889225433e-06, "loss": 1.2263, "step": 33419 }, { "epoch": 2.6191222570532915, "grad_norm": 2.1919219493865967, "learning_rate": 2.67281435911464e-06, "loss": 1.4436, "step": 33420 }, { "epoch": 2.6192006269592474, "grad_norm": 7.360487461090088, "learning_rate": 2.6725414381581555e-06, "loss": 1.592, "step": 33421 }, { "epoch": 2.6192789968652037, "grad_norm": 5.39100980758667, "learning_rate": 2.67226852605413e-06, "loss": 1.543, "step": 33422 }, { "epoch": 2.61935736677116, "grad_norm": 3.4538004398345947, "learning_rate": 2.671995622803597e-06, "loss": 1.3577, "step": 33423 }, { "epoch": 2.619435736677116, "grad_norm": 2.773336172103882, "learning_rate": 2.6717227284075953e-06, "loss": 1.1609, "step": 33424 }, { "epoch": 2.6195141065830723, "grad_norm": 3.7334868907928467, "learning_rate": 2.6714498428671687e-06, "loss": 1.4392, "step": 33425 }, { "epoch": 2.619592476489028, "grad_norm": 3.697650671005249, "learning_rate": 2.6711769661833488e-06, "loss": 1.4436, "step": 33426 }, { "epoch": 2.6196708463949845, "grad_norm": 5.612180709838867, "learning_rate": 2.6709040983571747e-06, "loss": 0.794, "step": 33427 }, { "epoch": 2.6197492163009404, "grad_norm": 4.228527069091797, "learning_rate": 2.6706312393896862e-06, "loss": 1.4284, "step": 33428 }, { "epoch": 2.6198275862068967, "grad_norm": 4.094654083251953, "learning_rate": 2.6703583892819197e-06, "loss": 1.2145, "step": 33429 }, { "epoch": 2.6199059561128526, "grad_norm": 6.1433258056640625, "learning_rate": 2.670085548034913e-06, "loss": 1.2214, "step": 33430 }, { "epoch": 2.619984326018809, "grad_norm": 1.9146391153335571, "learning_rate": 2.6698127156497047e-06, "loss": 1.4144, "step": 33431 }, { "epoch": 2.620062695924765, "grad_norm": 6.348879814147949, "learning_rate": 2.669539892127331e-06, "loss": 1.2957, "step": 33432 }, { "epoch": 2.620141065830721, "grad_norm": 3.8729538917541504, "learning_rate": 2.6692670774688314e-06, "loss": 1.4684, "step": 33433 }, { "epoch": 2.620219435736677, "grad_norm": 4.989248752593994, "learning_rate": 2.6689942716752427e-06, "loss": 1.5979, "step": 33434 }, { "epoch": 2.6202978056426334, "grad_norm": 5.780633449554443, "learning_rate": 2.668721474747602e-06, "loss": 1.6089, "step": 33435 }, { "epoch": 2.6203761755485893, "grad_norm": 2.6364502906799316, "learning_rate": 2.6684486866869475e-06, "loss": 1.2338, "step": 33436 }, { "epoch": 2.6204545454545456, "grad_norm": 3.3138301372528076, "learning_rate": 2.668175907494319e-06, "loss": 1.1874, "step": 33437 }, { "epoch": 2.6205329153605015, "grad_norm": 8.826430320739746, "learning_rate": 2.6679031371707476e-06, "loss": 1.0987, "step": 33438 }, { "epoch": 2.620611285266458, "grad_norm": 3.2264394760131836, "learning_rate": 2.6676303757172762e-06, "loss": 1.3782, "step": 33439 }, { "epoch": 2.6206896551724137, "grad_norm": 6.22031831741333, "learning_rate": 2.6673576231349436e-06, "loss": 1.3598, "step": 33440 }, { "epoch": 2.62076802507837, "grad_norm": 3.1721909046173096, "learning_rate": 2.667084879424782e-06, "loss": 1.3789, "step": 33441 }, { "epoch": 2.620846394984326, "grad_norm": 4.355776786804199, "learning_rate": 2.6668121445878316e-06, "loss": 1.3289, "step": 33442 }, { "epoch": 2.6209247648902823, "grad_norm": 3.8487722873687744, "learning_rate": 2.6665394186251275e-06, "loss": 0.8775, "step": 33443 }, { "epoch": 2.621003134796238, "grad_norm": 5.445781707763672, "learning_rate": 2.6662667015377126e-06, "loss": 1.3679, "step": 33444 }, { "epoch": 2.6210815047021945, "grad_norm": 3.3188772201538086, "learning_rate": 2.665993993326619e-06, "loss": 1.2699, "step": 33445 }, { "epoch": 2.6211598746081504, "grad_norm": 8.355030059814453, "learning_rate": 2.6657212939928846e-06, "loss": 1.7187, "step": 33446 }, { "epoch": 2.6212382445141067, "grad_norm": 2.1410181522369385, "learning_rate": 2.665448603537548e-06, "loss": 1.466, "step": 33447 }, { "epoch": 2.6213166144200626, "grad_norm": 3.3736040592193604, "learning_rate": 2.665175921961646e-06, "loss": 1.3847, "step": 33448 }, { "epoch": 2.621394984326019, "grad_norm": 2.5585384368896484, "learning_rate": 2.6649032492662148e-06, "loss": 1.4693, "step": 33449 }, { "epoch": 2.621473354231975, "grad_norm": 4.977532863616943, "learning_rate": 2.6646305854522925e-06, "loss": 0.8939, "step": 33450 }, { "epoch": 2.621551724137931, "grad_norm": 4.531939506530762, "learning_rate": 2.6643579305209184e-06, "loss": 1.2393, "step": 33451 }, { "epoch": 2.621630094043887, "grad_norm": 2.8798775672912598, "learning_rate": 2.6640852844731226e-06, "loss": 1.4595, "step": 33452 }, { "epoch": 2.6217084639498434, "grad_norm": 4.825433731079102, "learning_rate": 2.663812647309949e-06, "loss": 1.3532, "step": 33453 }, { "epoch": 2.6217868338557992, "grad_norm": 4.215489387512207, "learning_rate": 2.663540019032431e-06, "loss": 1.69, "step": 33454 }, { "epoch": 2.6218652037617556, "grad_norm": 3.4528541564941406, "learning_rate": 2.6632673996416105e-06, "loss": 1.1583, "step": 33455 }, { "epoch": 2.6219435736677115, "grad_norm": 6.285069942474365, "learning_rate": 2.6629947891385176e-06, "loss": 1.1071, "step": 33456 }, { "epoch": 2.622021943573668, "grad_norm": 4.49900484085083, "learning_rate": 2.6627221875241898e-06, "loss": 1.6823, "step": 33457 }, { "epoch": 2.6221003134796237, "grad_norm": 3.6986515522003174, "learning_rate": 2.662449594799669e-06, "loss": 1.1666, "step": 33458 }, { "epoch": 2.62217868338558, "grad_norm": 4.730815410614014, "learning_rate": 2.6621770109659906e-06, "loss": 1.9336, "step": 33459 }, { "epoch": 2.622257053291536, "grad_norm": 2.7705814838409424, "learning_rate": 2.6619044360241886e-06, "loss": 1.293, "step": 33460 }, { "epoch": 2.6223354231974922, "grad_norm": 2.8966012001037598, "learning_rate": 2.661631869975301e-06, "loss": 1.6466, "step": 33461 }, { "epoch": 2.622413793103448, "grad_norm": 3.465273380279541, "learning_rate": 2.661359312820363e-06, "loss": 1.1943, "step": 33462 }, { "epoch": 2.6224921630094045, "grad_norm": 3.879269599914551, "learning_rate": 2.6610867645604166e-06, "loss": 1.076, "step": 33463 }, { "epoch": 2.6225705329153604, "grad_norm": 1.8620457649230957, "learning_rate": 2.660814225196493e-06, "loss": 1.1421, "step": 33464 }, { "epoch": 2.6226489028213167, "grad_norm": 4.6965436935424805, "learning_rate": 2.660541694729631e-06, "loss": 1.3201, "step": 33465 }, { "epoch": 2.6227272727272726, "grad_norm": 3.251627206802368, "learning_rate": 2.6602691731608664e-06, "loss": 1.1433, "step": 33466 }, { "epoch": 2.622805642633229, "grad_norm": 9.813733100891113, "learning_rate": 2.659996660491236e-06, "loss": 1.2319, "step": 33467 }, { "epoch": 2.622884012539185, "grad_norm": 7.485592842102051, "learning_rate": 2.6597241567217763e-06, "loss": 1.538, "step": 33468 }, { "epoch": 2.622962382445141, "grad_norm": 2.4755003452301025, "learning_rate": 2.6594516618535238e-06, "loss": 1.2291, "step": 33469 }, { "epoch": 2.623040752351097, "grad_norm": 3.6783735752105713, "learning_rate": 2.6591791758875176e-06, "loss": 1.2833, "step": 33470 }, { "epoch": 2.6231191222570533, "grad_norm": 5.0850701332092285, "learning_rate": 2.658906698824787e-06, "loss": 1.3143, "step": 33471 }, { "epoch": 2.6231974921630092, "grad_norm": 3.763633966445923, "learning_rate": 2.658634230666375e-06, "loss": 1.7674, "step": 33472 }, { "epoch": 2.6232758620689656, "grad_norm": 4.537064075469971, "learning_rate": 2.658361771413316e-06, "loss": 1.4671, "step": 33473 }, { "epoch": 2.623354231974922, "grad_norm": 3.532642126083374, "learning_rate": 2.6580893210666476e-06, "loss": 1.4317, "step": 33474 }, { "epoch": 2.623432601880878, "grad_norm": 2.8663699626922607, "learning_rate": 2.657816879627403e-06, "loss": 1.6178, "step": 33475 }, { "epoch": 2.6235109717868337, "grad_norm": 5.2260212898254395, "learning_rate": 2.657544447096617e-06, "loss": 1.5112, "step": 33476 }, { "epoch": 2.62358934169279, "grad_norm": 3.549766778945923, "learning_rate": 2.6572720234753322e-06, "loss": 1.0719, "step": 33477 }, { "epoch": 2.6236677115987463, "grad_norm": 3.4501166343688965, "learning_rate": 2.6569996087645823e-06, "loss": 1.4561, "step": 33478 }, { "epoch": 2.6237460815047022, "grad_norm": 12.15068531036377, "learning_rate": 2.6567272029654013e-06, "loss": 1.242, "step": 33479 }, { "epoch": 2.623824451410658, "grad_norm": 3.8586888313293457, "learning_rate": 2.6564548060788254e-06, "loss": 1.4434, "step": 33480 }, { "epoch": 2.6239028213166145, "grad_norm": 5.677080154418945, "learning_rate": 2.6561824181058927e-06, "loss": 1.008, "step": 33481 }, { "epoch": 2.623981191222571, "grad_norm": 3.576144218444824, "learning_rate": 2.655910039047638e-06, "loss": 1.4426, "step": 33482 }, { "epoch": 2.6240595611285267, "grad_norm": 2.3012843132019043, "learning_rate": 2.6556376689050965e-06, "loss": 1.6347, "step": 33483 }, { "epoch": 2.6241379310344826, "grad_norm": 3.6744351387023926, "learning_rate": 2.655365307679306e-06, "loss": 1.3275, "step": 33484 }, { "epoch": 2.624216300940439, "grad_norm": 2.9477086067199707, "learning_rate": 2.6550929553713017e-06, "loss": 1.4495, "step": 33485 }, { "epoch": 2.6242946708463952, "grad_norm": 5.003596305847168, "learning_rate": 2.654820611982119e-06, "loss": 1.3895, "step": 33486 }, { "epoch": 2.624373040752351, "grad_norm": 2.709441661834717, "learning_rate": 2.6545482775127944e-06, "loss": 1.6602, "step": 33487 }, { "epoch": 2.624451410658307, "grad_norm": 3.1931185722351074, "learning_rate": 2.6542759519643634e-06, "loss": 1.3285, "step": 33488 }, { "epoch": 2.6245297805642633, "grad_norm": 3.5209765434265137, "learning_rate": 2.654003635337864e-06, "loss": 1.5737, "step": 33489 }, { "epoch": 2.6246081504702197, "grad_norm": 3.8832366466522217, "learning_rate": 2.6537313276343255e-06, "loss": 1.9781, "step": 33490 }, { "epoch": 2.6246865203761756, "grad_norm": 3.591505289077759, "learning_rate": 2.65345902885479e-06, "loss": 1.5357, "step": 33491 }, { "epoch": 2.6247648902821314, "grad_norm": 3.2825558185577393, "learning_rate": 2.6531867390002907e-06, "loss": 1.355, "step": 33492 }, { "epoch": 2.624843260188088, "grad_norm": 4.367151260375977, "learning_rate": 2.6529144580718663e-06, "loss": 1.2912, "step": 33493 }, { "epoch": 2.624921630094044, "grad_norm": 2.4961347579956055, "learning_rate": 2.6526421860705474e-06, "loss": 1.1902, "step": 33494 }, { "epoch": 2.625, "grad_norm": 4.9733099937438965, "learning_rate": 2.6523699229973705e-06, "loss": 1.3184, "step": 33495 }, { "epoch": 2.625078369905956, "grad_norm": 3.735626220703125, "learning_rate": 2.6520976688533764e-06, "loss": 1.5522, "step": 33496 }, { "epoch": 2.625156739811912, "grad_norm": 2.4300613403320312, "learning_rate": 2.651825423639594e-06, "loss": 0.7362, "step": 33497 }, { "epoch": 2.6252351097178686, "grad_norm": 3.87575101852417, "learning_rate": 2.651553187357062e-06, "loss": 0.9542, "step": 33498 }, { "epoch": 2.6253134796238244, "grad_norm": 3.647412061691284, "learning_rate": 2.6512809600068155e-06, "loss": 0.6175, "step": 33499 }, { "epoch": 2.6253918495297803, "grad_norm": 3.404829978942871, "learning_rate": 2.65100874158989e-06, "loss": 1.5241, "step": 33500 }, { "epoch": 2.6254702194357367, "grad_norm": 4.534574508666992, "learning_rate": 2.650736532107321e-06, "loss": 1.119, "step": 33501 }, { "epoch": 2.625548589341693, "grad_norm": 3.3305563926696777, "learning_rate": 2.650464331560143e-06, "loss": 1.1944, "step": 33502 }, { "epoch": 2.625626959247649, "grad_norm": 5.7527313232421875, "learning_rate": 2.6501921399493922e-06, "loss": 1.542, "step": 33503 }, { "epoch": 2.6257053291536048, "grad_norm": 4.707080364227295, "learning_rate": 2.6499199572761036e-06, "loss": 1.4393, "step": 33504 }, { "epoch": 2.625783699059561, "grad_norm": 2.6595141887664795, "learning_rate": 2.6496477835413127e-06, "loss": 1.1941, "step": 33505 }, { "epoch": 2.6258620689655174, "grad_norm": 3.3318119049072266, "learning_rate": 2.6493756187460533e-06, "loss": 1.2972, "step": 33506 }, { "epoch": 2.6259404388714733, "grad_norm": 3.71675705909729, "learning_rate": 2.6491034628913648e-06, "loss": 1.3062, "step": 33507 }, { "epoch": 2.626018808777429, "grad_norm": 5.4585442543029785, "learning_rate": 2.648831315978276e-06, "loss": 1.1726, "step": 33508 }, { "epoch": 2.6260971786833855, "grad_norm": 3.9944655895233154, "learning_rate": 2.6485591780078248e-06, "loss": 1.1609, "step": 33509 }, { "epoch": 2.626175548589342, "grad_norm": 7.1217546463012695, "learning_rate": 2.6482870489810477e-06, "loss": 1.7033, "step": 33510 }, { "epoch": 2.6262539184952978, "grad_norm": 7.355869293212891, "learning_rate": 2.648014928898981e-06, "loss": 1.45, "step": 33511 }, { "epoch": 2.6263322884012537, "grad_norm": 4.413269519805908, "learning_rate": 2.6477428177626556e-06, "loss": 0.7296, "step": 33512 }, { "epoch": 2.62641065830721, "grad_norm": 2.0962026119232178, "learning_rate": 2.6474707155731082e-06, "loss": 1.3444, "step": 33513 }, { "epoch": 2.6264890282131663, "grad_norm": 6.1034698486328125, "learning_rate": 2.6471986223313717e-06, "loss": 0.8675, "step": 33514 }, { "epoch": 2.626567398119122, "grad_norm": 6.915496826171875, "learning_rate": 2.646926538038488e-06, "loss": 1.271, "step": 33515 }, { "epoch": 2.626645768025078, "grad_norm": 3.6425657272338867, "learning_rate": 2.6466544626954837e-06, "loss": 1.0578, "step": 33516 }, { "epoch": 2.6267241379310344, "grad_norm": 4.854197978973389, "learning_rate": 2.646382396303398e-06, "loss": 0.9916, "step": 33517 }, { "epoch": 2.6268025078369908, "grad_norm": 4.664721488952637, "learning_rate": 2.6461103388632646e-06, "loss": 1.3811, "step": 33518 }, { "epoch": 2.6268808777429467, "grad_norm": 2.986889123916626, "learning_rate": 2.645838290376118e-06, "loss": 1.3803, "step": 33519 }, { "epoch": 2.626959247648903, "grad_norm": 3.5445029735565186, "learning_rate": 2.6455662508429946e-06, "loss": 0.8406, "step": 33520 }, { "epoch": 2.627037617554859, "grad_norm": 3.0807363986968994, "learning_rate": 2.6452942202649266e-06, "loss": 1.5581, "step": 33521 }, { "epoch": 2.627115987460815, "grad_norm": 3.010486125946045, "learning_rate": 2.6450221986429505e-06, "loss": 1.4375, "step": 33522 }, { "epoch": 2.627194357366771, "grad_norm": 4.378155708312988, "learning_rate": 2.6447501859781004e-06, "loss": 1.3277, "step": 33523 }, { "epoch": 2.6272727272727274, "grad_norm": 19.28618049621582, "learning_rate": 2.6444781822714106e-06, "loss": 1.3422, "step": 33524 }, { "epoch": 2.6273510971786833, "grad_norm": 3.970857858657837, "learning_rate": 2.6442061875239163e-06, "loss": 1.0273, "step": 33525 }, { "epoch": 2.6274294670846396, "grad_norm": 2.667543411254883, "learning_rate": 2.643934201736653e-06, "loss": 1.2453, "step": 33526 }, { "epoch": 2.6275078369905955, "grad_norm": 3.652240037918091, "learning_rate": 2.64366222491065e-06, "loss": 1.7056, "step": 33527 }, { "epoch": 2.627586206896552, "grad_norm": 2.260779857635498, "learning_rate": 2.643390257046948e-06, "loss": 1.1352, "step": 33528 }, { "epoch": 2.6276645768025078, "grad_norm": 3.573456287384033, "learning_rate": 2.6431182981465797e-06, "loss": 1.5135, "step": 33529 }, { "epoch": 2.627742946708464, "grad_norm": 2.2489655017852783, "learning_rate": 2.642846348210579e-06, "loss": 1.3564, "step": 33530 }, { "epoch": 2.62782131661442, "grad_norm": 3.3487534523010254, "learning_rate": 2.6425744072399796e-06, "loss": 1.408, "step": 33531 }, { "epoch": 2.6278996865203763, "grad_norm": 8.356022834777832, "learning_rate": 2.642302475235814e-06, "loss": 1.4251, "step": 33532 }, { "epoch": 2.627978056426332, "grad_norm": 3.9975316524505615, "learning_rate": 2.6420305521991207e-06, "loss": 1.786, "step": 33533 }, { "epoch": 2.6280564263322885, "grad_norm": 2.3475534915924072, "learning_rate": 2.6417586381309334e-06, "loss": 1.2877, "step": 33534 }, { "epoch": 2.6281347962382444, "grad_norm": 5.211324691772461, "learning_rate": 2.6414867330322835e-06, "loss": 1.5365, "step": 33535 }, { "epoch": 2.6282131661442008, "grad_norm": 3.9662458896636963, "learning_rate": 2.6412148369042067e-06, "loss": 1.6407, "step": 33536 }, { "epoch": 2.6282915360501566, "grad_norm": 4.2094621658325195, "learning_rate": 2.6409429497477372e-06, "loss": 1.0681, "step": 33537 }, { "epoch": 2.628369905956113, "grad_norm": 3.5822975635528564, "learning_rate": 2.6406710715639094e-06, "loss": 0.7225, "step": 33538 }, { "epoch": 2.628448275862069, "grad_norm": 3.923609972000122, "learning_rate": 2.640399202353756e-06, "loss": 1.4291, "step": 33539 }, { "epoch": 2.628526645768025, "grad_norm": 2.446787118911743, "learning_rate": 2.6401273421183125e-06, "loss": 1.1847, "step": 33540 }, { "epoch": 2.628605015673981, "grad_norm": 3.4532246589660645, "learning_rate": 2.6398554908586128e-06, "loss": 1.5217, "step": 33541 }, { "epoch": 2.6286833855799374, "grad_norm": 4.646673202514648, "learning_rate": 2.6395836485756903e-06, "loss": 1.377, "step": 33542 }, { "epoch": 2.6287617554858933, "grad_norm": 4.441026210784912, "learning_rate": 2.6393118152705792e-06, "loss": 0.9393, "step": 33543 }, { "epoch": 2.6288401253918496, "grad_norm": 13.369095802307129, "learning_rate": 2.639039990944314e-06, "loss": 1.0915, "step": 33544 }, { "epoch": 2.6289184952978055, "grad_norm": 7.582577228546143, "learning_rate": 2.638768175597929e-06, "loss": 1.3803, "step": 33545 }, { "epoch": 2.628996865203762, "grad_norm": 3.2385265827178955, "learning_rate": 2.638496369232454e-06, "loss": 1.173, "step": 33546 }, { "epoch": 2.6290752351097177, "grad_norm": 2.8815369606018066, "learning_rate": 2.6382245718489276e-06, "loss": 1.6066, "step": 33547 }, { "epoch": 2.629153605015674, "grad_norm": 3.5826449394226074, "learning_rate": 2.637952783448381e-06, "loss": 1.2802, "step": 33548 }, { "epoch": 2.62923197492163, "grad_norm": 2.936906099319458, "learning_rate": 2.6376810040318513e-06, "loss": 1.098, "step": 33549 }, { "epoch": 2.6293103448275863, "grad_norm": 5.607538223266602, "learning_rate": 2.6374092336003684e-06, "loss": 1.4914, "step": 33550 }, { "epoch": 2.629388714733542, "grad_norm": 3.722125768661499, "learning_rate": 2.6371374721549648e-06, "loss": 1.3098, "step": 33551 }, { "epoch": 2.6294670846394985, "grad_norm": 5.23659086227417, "learning_rate": 2.6368657196966803e-06, "loss": 0.9379, "step": 33552 }, { "epoch": 2.6295454545454544, "grad_norm": 5.3331618309021, "learning_rate": 2.636593976226543e-06, "loss": 1.2618, "step": 33553 }, { "epoch": 2.6296238244514107, "grad_norm": 4.688635349273682, "learning_rate": 2.6363222417455882e-06, "loss": 1.45, "step": 33554 }, { "epoch": 2.6297021943573666, "grad_norm": 7.350579261779785, "learning_rate": 2.63605051625485e-06, "loss": 1.0574, "step": 33555 }, { "epoch": 2.629780564263323, "grad_norm": 3.876925230026245, "learning_rate": 2.6357787997553608e-06, "loss": 1.4958, "step": 33556 }, { "epoch": 2.629858934169279, "grad_norm": 6.707693099975586, "learning_rate": 2.635507092248155e-06, "loss": 0.9039, "step": 33557 }, { "epoch": 2.629937304075235, "grad_norm": 2.678070068359375, "learning_rate": 2.6352353937342655e-06, "loss": 1.6421, "step": 33558 }, { "epoch": 2.630015673981191, "grad_norm": 3.8212454319000244, "learning_rate": 2.634963704214727e-06, "loss": 1.2864, "step": 33559 }, { "epoch": 2.6300940438871474, "grad_norm": 4.402229309082031, "learning_rate": 2.6346920236905706e-06, "loss": 1.2998, "step": 33560 }, { "epoch": 2.6301724137931033, "grad_norm": 2.2959823608398438, "learning_rate": 2.634420352162831e-06, "loss": 1.1418, "step": 33561 }, { "epoch": 2.6302507836990596, "grad_norm": 5.271723747253418, "learning_rate": 2.634148689632542e-06, "loss": 1.2949, "step": 33562 }, { "epoch": 2.6303291536050155, "grad_norm": 7.198072910308838, "learning_rate": 2.6338770361007376e-06, "loss": 1.7796, "step": 33563 }, { "epoch": 2.630407523510972, "grad_norm": 2.785616874694824, "learning_rate": 2.633605391568448e-06, "loss": 1.3584, "step": 33564 }, { "epoch": 2.6304858934169277, "grad_norm": 3.3822970390319824, "learning_rate": 2.6333337560367056e-06, "loss": 1.5503, "step": 33565 }, { "epoch": 2.630564263322884, "grad_norm": 3.3749029636383057, "learning_rate": 2.6330621295065486e-06, "loss": 1.8976, "step": 33566 }, { "epoch": 2.63064263322884, "grad_norm": 3.2599854469299316, "learning_rate": 2.6327905119790087e-06, "loss": 1.1229, "step": 33567 }, { "epoch": 2.6307210031347963, "grad_norm": 3.16939115524292, "learning_rate": 2.632518903455117e-06, "loss": 1.4089, "step": 33568 }, { "epoch": 2.6307993730407526, "grad_norm": 4.788700580596924, "learning_rate": 2.6322473039359063e-06, "loss": 1.5249, "step": 33569 }, { "epoch": 2.6308777429467085, "grad_norm": 2.7184042930603027, "learning_rate": 2.631975713422409e-06, "loss": 1.4602, "step": 33570 }, { "epoch": 2.6309561128526644, "grad_norm": 5.11642599105835, "learning_rate": 2.6317041319156638e-06, "loss": 1.3937, "step": 33571 }, { "epoch": 2.6310344827586207, "grad_norm": 4.267541885375977, "learning_rate": 2.631432559416698e-06, "loss": 1.3122, "step": 33572 }, { "epoch": 2.631112852664577, "grad_norm": 4.133832931518555, "learning_rate": 2.631160995926546e-06, "loss": 1.4724, "step": 33573 }, { "epoch": 2.631191222570533, "grad_norm": 3.457124710083008, "learning_rate": 2.6308894414462405e-06, "loss": 1.1586, "step": 33574 }, { "epoch": 2.631269592476489, "grad_norm": 5.826341152191162, "learning_rate": 2.6306178959768146e-06, "loss": 1.3605, "step": 33575 }, { "epoch": 2.631347962382445, "grad_norm": 7.393223285675049, "learning_rate": 2.6303463595193017e-06, "loss": 1.0652, "step": 33576 }, { "epoch": 2.6314263322884015, "grad_norm": 4.632722854614258, "learning_rate": 2.6300748320747343e-06, "loss": 1.0257, "step": 33577 }, { "epoch": 2.6315047021943574, "grad_norm": 2.628746509552002, "learning_rate": 2.629803313644147e-06, "loss": 1.197, "step": 33578 }, { "epoch": 2.6315830721003133, "grad_norm": 6.680360794067383, "learning_rate": 2.6295318042285656e-06, "loss": 1.8513, "step": 33579 }, { "epoch": 2.6316614420062696, "grad_norm": 4.566582202911377, "learning_rate": 2.6292603038290306e-06, "loss": 1.0565, "step": 33580 }, { "epoch": 2.631739811912226, "grad_norm": 3.86517596244812, "learning_rate": 2.6289888124465712e-06, "loss": 0.9906, "step": 33581 }, { "epoch": 2.631818181818182, "grad_norm": 7.363284111022949, "learning_rate": 2.6287173300822227e-06, "loss": 1.1881, "step": 33582 }, { "epoch": 2.6318965517241377, "grad_norm": 3.7557907104492188, "learning_rate": 2.628445856737013e-06, "loss": 1.6169, "step": 33583 }, { "epoch": 2.631974921630094, "grad_norm": 3.011235475540161, "learning_rate": 2.6281743924119764e-06, "loss": 0.993, "step": 33584 }, { "epoch": 2.6320532915360504, "grad_norm": 3.6303257942199707, "learning_rate": 2.6279029371081473e-06, "loss": 1.5899, "step": 33585 }, { "epoch": 2.6321316614420063, "grad_norm": 3.019188642501831, "learning_rate": 2.6276314908265586e-06, "loss": 0.9993, "step": 33586 }, { "epoch": 2.632210031347962, "grad_norm": 3.239959955215454, "learning_rate": 2.62736005356824e-06, "loss": 1.1371, "step": 33587 }, { "epoch": 2.6322884012539185, "grad_norm": 3.262882709503174, "learning_rate": 2.6270886253342247e-06, "loss": 0.7916, "step": 33588 }, { "epoch": 2.632366771159875, "grad_norm": 2.991178512573242, "learning_rate": 2.6268172061255438e-06, "loss": 1.4626, "step": 33589 }, { "epoch": 2.6324451410658307, "grad_norm": 2.988920211791992, "learning_rate": 2.626545795943235e-06, "loss": 1.3098, "step": 33590 }, { "epoch": 2.6325235109717866, "grad_norm": 4.401852130889893, "learning_rate": 2.6262743947883253e-06, "loss": 1.5201, "step": 33591 }, { "epoch": 2.632601880877743, "grad_norm": 3.405273914337158, "learning_rate": 2.626003002661848e-06, "loss": 1.214, "step": 33592 }, { "epoch": 2.6326802507836993, "grad_norm": 2.4559264183044434, "learning_rate": 2.6257316195648364e-06, "loss": 1.1425, "step": 33593 }, { "epoch": 2.632758620689655, "grad_norm": 5.37978982925415, "learning_rate": 2.625460245498322e-06, "loss": 1.5609, "step": 33594 }, { "epoch": 2.632836990595611, "grad_norm": 2.6463656425476074, "learning_rate": 2.6251888804633374e-06, "loss": 1.3538, "step": 33595 }, { "epoch": 2.6329153605015674, "grad_norm": 4.900259971618652, "learning_rate": 2.624917524460914e-06, "loss": 1.7429, "step": 33596 }, { "epoch": 2.6329937304075237, "grad_norm": 2.7092323303222656, "learning_rate": 2.6246461774920867e-06, "loss": 1.7078, "step": 33597 }, { "epoch": 2.6330721003134796, "grad_norm": 4.452077865600586, "learning_rate": 2.6243748395578814e-06, "loss": 1.7111, "step": 33598 }, { "epoch": 2.6331504702194355, "grad_norm": 2.659655809402466, "learning_rate": 2.624103510659336e-06, "loss": 1.1358, "step": 33599 }, { "epoch": 2.633228840125392, "grad_norm": 4.550661563873291, "learning_rate": 2.6238321907974805e-06, "loss": 1.3164, "step": 33600 }, { "epoch": 2.633307210031348, "grad_norm": 4.021167278289795, "learning_rate": 2.6235608799733487e-06, "loss": 1.1599, "step": 33601 }, { "epoch": 2.633385579937304, "grad_norm": 5.147218704223633, "learning_rate": 2.6232895781879685e-06, "loss": 1.4956, "step": 33602 }, { "epoch": 2.63346394984326, "grad_norm": 2.9644718170166016, "learning_rate": 2.6230182854423726e-06, "loss": 1.295, "step": 33603 }, { "epoch": 2.6335423197492163, "grad_norm": 2.5242745876312256, "learning_rate": 2.6227470017375955e-06, "loss": 1.4288, "step": 33604 }, { "epoch": 2.6336206896551726, "grad_norm": 2.247901678085327, "learning_rate": 2.6224757270746702e-06, "loss": 0.8024, "step": 33605 }, { "epoch": 2.6336990595611285, "grad_norm": 3.812978982925415, "learning_rate": 2.6222044614546242e-06, "loss": 1.582, "step": 33606 }, { "epoch": 2.6337774294670844, "grad_norm": 10.966913223266602, "learning_rate": 2.6219332048784913e-06, "loss": 1.6977, "step": 33607 }, { "epoch": 2.6338557993730407, "grad_norm": 3.740119695663452, "learning_rate": 2.6216619573473023e-06, "loss": 1.1844, "step": 33608 }, { "epoch": 2.633934169278997, "grad_norm": 2.9660658836364746, "learning_rate": 2.6213907188620902e-06, "loss": 1.4236, "step": 33609 }, { "epoch": 2.634012539184953, "grad_norm": 2.49739408493042, "learning_rate": 2.6211194894238863e-06, "loss": 1.3991, "step": 33610 }, { "epoch": 2.634090909090909, "grad_norm": 2.8197901248931885, "learning_rate": 2.6208482690337213e-06, "loss": 1.2996, "step": 33611 }, { "epoch": 2.634169278996865, "grad_norm": 3.0791053771972656, "learning_rate": 2.620577057692628e-06, "loss": 1.3693, "step": 33612 }, { "epoch": 2.6342476489028215, "grad_norm": 2.508349657058716, "learning_rate": 2.620305855401637e-06, "loss": 1.1417, "step": 33613 }, { "epoch": 2.6343260188087774, "grad_norm": 2.0370187759399414, "learning_rate": 2.6200346621617804e-06, "loss": 0.9825, "step": 33614 }, { "epoch": 2.6344043887147337, "grad_norm": 2.550259828567505, "learning_rate": 2.61976347797409e-06, "loss": 1.1458, "step": 33615 }, { "epoch": 2.6344827586206896, "grad_norm": 2.9334120750427246, "learning_rate": 2.619492302839596e-06, "loss": 1.566, "step": 33616 }, { "epoch": 2.634561128526646, "grad_norm": 4.5494704246521, "learning_rate": 2.619221136759331e-06, "loss": 1.0773, "step": 33617 }, { "epoch": 2.634639498432602, "grad_norm": 4.3774919509887695, "learning_rate": 2.6189499797343252e-06, "loss": 1.1308, "step": 33618 }, { "epoch": 2.634717868338558, "grad_norm": 5.221538066864014, "learning_rate": 2.618678831765614e-06, "loss": 1.2542, "step": 33619 }, { "epoch": 2.634796238244514, "grad_norm": 4.540338516235352, "learning_rate": 2.6184076928542223e-06, "loss": 1.3304, "step": 33620 }, { "epoch": 2.6348746081504704, "grad_norm": 4.9699602127075195, "learning_rate": 2.6181365630011835e-06, "loss": 1.8603, "step": 33621 }, { "epoch": 2.6349529780564263, "grad_norm": 2.7907674312591553, "learning_rate": 2.6178654422075313e-06, "loss": 1.2658, "step": 33622 }, { "epoch": 2.6350313479623826, "grad_norm": 3.305910587310791, "learning_rate": 2.6175943304742966e-06, "loss": 1.3979, "step": 33623 }, { "epoch": 2.6351097178683385, "grad_norm": 2.8577985763549805, "learning_rate": 2.6173232278025087e-06, "loss": 0.9662, "step": 33624 }, { "epoch": 2.635188087774295, "grad_norm": 2.1102519035339355, "learning_rate": 2.6170521341931986e-06, "loss": 1.3014, "step": 33625 }, { "epoch": 2.6352664576802507, "grad_norm": 2.76121187210083, "learning_rate": 2.616781049647399e-06, "loss": 1.1094, "step": 33626 }, { "epoch": 2.635344827586207, "grad_norm": 2.2918312549591064, "learning_rate": 2.6165099741661394e-06, "loss": 1.5498, "step": 33627 }, { "epoch": 2.635423197492163, "grad_norm": 10.960981369018555, "learning_rate": 2.616238907750452e-06, "loss": 1.191, "step": 33628 }, { "epoch": 2.6355015673981192, "grad_norm": 3.4878175258636475, "learning_rate": 2.6159678504013675e-06, "loss": 1.4625, "step": 33629 }, { "epoch": 2.635579937304075, "grad_norm": 2.3472466468811035, "learning_rate": 2.615696802119916e-06, "loss": 1.3972, "step": 33630 }, { "epoch": 2.6356583072100315, "grad_norm": 2.1656980514526367, "learning_rate": 2.61542576290713e-06, "loss": 1.3815, "step": 33631 }, { "epoch": 2.6357366771159874, "grad_norm": 2.8255932331085205, "learning_rate": 2.615154732764039e-06, "loss": 1.3248, "step": 33632 }, { "epoch": 2.6358150470219437, "grad_norm": 5.039945125579834, "learning_rate": 2.6148837116916753e-06, "loss": 1.32, "step": 33633 }, { "epoch": 2.6358934169278996, "grad_norm": 2.6746091842651367, "learning_rate": 2.6146126996910704e-06, "loss": 1.3422, "step": 33634 }, { "epoch": 2.635971786833856, "grad_norm": 2.5270798206329346, "learning_rate": 2.6143416967632497e-06, "loss": 1.4719, "step": 33635 }, { "epoch": 2.636050156739812, "grad_norm": 2.5729527473449707, "learning_rate": 2.6140707029092495e-06, "loss": 1.2705, "step": 33636 }, { "epoch": 2.636128526645768, "grad_norm": 2.7339272499084473, "learning_rate": 2.613799718130099e-06, "loss": 1.5156, "step": 33637 }, { "epoch": 2.636206896551724, "grad_norm": 4.850015163421631, "learning_rate": 2.613528742426831e-06, "loss": 1.6259, "step": 33638 }, { "epoch": 2.6362852664576804, "grad_norm": 2.371397018432617, "learning_rate": 2.613257775800472e-06, "loss": 1.3736, "step": 33639 }, { "epoch": 2.6363636363636362, "grad_norm": 8.1329345703125, "learning_rate": 2.6129868182520525e-06, "loss": 1.7442, "step": 33640 }, { "epoch": 2.6364420062695926, "grad_norm": 6.366182327270508, "learning_rate": 2.6127158697826066e-06, "loss": 1.1092, "step": 33641 }, { "epoch": 2.6365203761755485, "grad_norm": 3.1554481983184814, "learning_rate": 2.6124449303931655e-06, "loss": 1.3586, "step": 33642 }, { "epoch": 2.636598746081505, "grad_norm": 3.667778253555298, "learning_rate": 2.6121740000847557e-06, "loss": 1.1587, "step": 33643 }, { "epoch": 2.6366771159874607, "grad_norm": 3.4619383811950684, "learning_rate": 2.6119030788584093e-06, "loss": 1.5734, "step": 33644 }, { "epoch": 2.636755485893417, "grad_norm": 10.911980628967285, "learning_rate": 2.6116321667151555e-06, "loss": 2.0538, "step": 33645 }, { "epoch": 2.636833855799373, "grad_norm": 4.275414943695068, "learning_rate": 2.61136126365603e-06, "loss": 1.63, "step": 33646 }, { "epoch": 2.6369122257053292, "grad_norm": 2.287065267562866, "learning_rate": 2.6110903696820567e-06, "loss": 1.1532, "step": 33647 }, { "epoch": 2.636990595611285, "grad_norm": 2.4676425457000732, "learning_rate": 2.610819484794269e-06, "loss": 1.5823, "step": 33648 }, { "epoch": 2.6370689655172415, "grad_norm": 2.9541501998901367, "learning_rate": 2.610548608993697e-06, "loss": 1.4036, "step": 33649 }, { "epoch": 2.6371473354231973, "grad_norm": 3.127354383468628, "learning_rate": 2.6102777422813714e-06, "loss": 1.2239, "step": 33650 }, { "epoch": 2.6372257053291537, "grad_norm": 2.9785046577453613, "learning_rate": 2.6100068846583208e-06, "loss": 1.1469, "step": 33651 }, { "epoch": 2.6373040752351096, "grad_norm": 10.825480461120605, "learning_rate": 2.609736036125577e-06, "loss": 0.6986, "step": 33652 }, { "epoch": 2.637382445141066, "grad_norm": 4.112987518310547, "learning_rate": 2.609465196684171e-06, "loss": 1.6465, "step": 33653 }, { "epoch": 2.637460815047022, "grad_norm": 2.8862409591674805, "learning_rate": 2.609194366335128e-06, "loss": 1.3317, "step": 33654 }, { "epoch": 2.637539184952978, "grad_norm": 3.3674263954162598, "learning_rate": 2.608923545079484e-06, "loss": 1.4493, "step": 33655 }, { "epoch": 2.637617554858934, "grad_norm": 3.133091449737549, "learning_rate": 2.6086527329182665e-06, "loss": 1.1317, "step": 33656 }, { "epoch": 2.6376959247648903, "grad_norm": 6.807825088500977, "learning_rate": 2.608381929852507e-06, "loss": 1.4602, "step": 33657 }, { "epoch": 2.6377742946708462, "grad_norm": 2.401242733001709, "learning_rate": 2.6081111358832333e-06, "loss": 1.2539, "step": 33658 }, { "epoch": 2.6378526645768026, "grad_norm": 2.834749221801758, "learning_rate": 2.6078403510114737e-06, "loss": 1.2585, "step": 33659 }, { "epoch": 2.637931034482759, "grad_norm": 2.8761026859283447, "learning_rate": 2.6075695752382624e-06, "loss": 1.4275, "step": 33660 }, { "epoch": 2.638009404388715, "grad_norm": 7.4401373863220215, "learning_rate": 2.60729880856463e-06, "loss": 1.7665, "step": 33661 }, { "epoch": 2.6380877742946707, "grad_norm": 3.4249379634857178, "learning_rate": 2.6070280509916025e-06, "loss": 1.0277, "step": 33662 }, { "epoch": 2.638166144200627, "grad_norm": 3.1989023685455322, "learning_rate": 2.6067573025202105e-06, "loss": 1.2839, "step": 33663 }, { "epoch": 2.6382445141065833, "grad_norm": 6.625656604766846, "learning_rate": 2.6064865631514847e-06, "loss": 1.5764, "step": 33664 }, { "epoch": 2.6383228840125392, "grad_norm": 2.396545648574829, "learning_rate": 2.606215832886454e-06, "loss": 1.1699, "step": 33665 }, { "epoch": 2.638401253918495, "grad_norm": 2.94490385055542, "learning_rate": 2.6059451117261492e-06, "loss": 1.4029, "step": 33666 }, { "epoch": 2.6384796238244514, "grad_norm": 2.4950368404388428, "learning_rate": 2.6056743996716e-06, "loss": 1.1714, "step": 33667 }, { "epoch": 2.638557993730408, "grad_norm": 2.436338424682617, "learning_rate": 2.605403696723835e-06, "loss": 0.9194, "step": 33668 }, { "epoch": 2.6386363636363637, "grad_norm": 4.154644012451172, "learning_rate": 2.605133002883885e-06, "loss": 1.2038, "step": 33669 }, { "epoch": 2.6387147335423196, "grad_norm": 3.296137809753418, "learning_rate": 2.604862318152778e-06, "loss": 2.0278, "step": 33670 }, { "epoch": 2.638793103448276, "grad_norm": 3.2299864292144775, "learning_rate": 2.6045916425315453e-06, "loss": 1.0792, "step": 33671 }, { "epoch": 2.6388714733542322, "grad_norm": 4.677515983581543, "learning_rate": 2.6043209760212174e-06, "loss": 1.5096, "step": 33672 }, { "epoch": 2.638949843260188, "grad_norm": 3.564971446990967, "learning_rate": 2.6040503186228184e-06, "loss": 1.3078, "step": 33673 }, { "epoch": 2.639028213166144, "grad_norm": 10.837750434875488, "learning_rate": 2.603779670337383e-06, "loss": 1.3407, "step": 33674 }, { "epoch": 2.6391065830721003, "grad_norm": 4.072944164276123, "learning_rate": 2.6035090311659405e-06, "loss": 1.1788, "step": 33675 }, { "epoch": 2.6391849529780567, "grad_norm": 3.1629507541656494, "learning_rate": 2.603238401109517e-06, "loss": 1.3052, "step": 33676 }, { "epoch": 2.6392633228840126, "grad_norm": 5.056756973266602, "learning_rate": 2.602967780169144e-06, "loss": 1.4677, "step": 33677 }, { "epoch": 2.6393416927899684, "grad_norm": 4.103440761566162, "learning_rate": 2.6026971683458478e-06, "loss": 1.4708, "step": 33678 }, { "epoch": 2.6394200626959248, "grad_norm": 3.1138854026794434, "learning_rate": 2.6024265656406644e-06, "loss": 1.0539, "step": 33679 }, { "epoch": 2.639498432601881, "grad_norm": 12.217979431152344, "learning_rate": 2.6021559720546173e-06, "loss": 1.8013, "step": 33680 }, { "epoch": 2.639576802507837, "grad_norm": 11.782088279724121, "learning_rate": 2.6018853875887375e-06, "loss": 1.3634, "step": 33681 }, { "epoch": 2.639655172413793, "grad_norm": 3.0773532390594482, "learning_rate": 2.6016148122440532e-06, "loss": 1.343, "step": 33682 }, { "epoch": 2.639733542319749, "grad_norm": 2.4676761627197266, "learning_rate": 2.6013442460215944e-06, "loss": 1.4147, "step": 33683 }, { "epoch": 2.6398119122257055, "grad_norm": 3.088897705078125, "learning_rate": 2.60107368892239e-06, "loss": 1.4727, "step": 33684 }, { "epoch": 2.6398902821316614, "grad_norm": 3.385105848312378, "learning_rate": 2.6008031409474687e-06, "loss": 1.5903, "step": 33685 }, { "epoch": 2.6399686520376173, "grad_norm": 3.9919655323028564, "learning_rate": 2.6005326020978605e-06, "loss": 1.7258, "step": 33686 }, { "epoch": 2.6400470219435737, "grad_norm": 3.3250434398651123, "learning_rate": 2.6002620723745944e-06, "loss": 1.3477, "step": 33687 }, { "epoch": 2.64012539184953, "grad_norm": 6.813052177429199, "learning_rate": 2.5999915517786978e-06, "loss": 0.8526, "step": 33688 }, { "epoch": 2.640203761755486, "grad_norm": 3.6534793376922607, "learning_rate": 2.5997210403112015e-06, "loss": 1.0176, "step": 33689 }, { "epoch": 2.6402821316614418, "grad_norm": 2.41675066947937, "learning_rate": 2.599450537973135e-06, "loss": 1.4267, "step": 33690 }, { "epoch": 2.640360501567398, "grad_norm": 5.112297058105469, "learning_rate": 2.5991800447655235e-06, "loss": 1.3385, "step": 33691 }, { "epoch": 2.6404388714733544, "grad_norm": 4.152761936187744, "learning_rate": 2.5989095606893965e-06, "loss": 1.3713, "step": 33692 }, { "epoch": 2.6405172413793103, "grad_norm": 2.079120397567749, "learning_rate": 2.5986390857457856e-06, "loss": 1.4273, "step": 33693 }, { "epoch": 2.640595611285266, "grad_norm": 4.406339645385742, "learning_rate": 2.5983686199357204e-06, "loss": 1.3962, "step": 33694 }, { "epoch": 2.6406739811912225, "grad_norm": 2.419827699661255, "learning_rate": 2.5980981632602254e-06, "loss": 1.1417, "step": 33695 }, { "epoch": 2.640752351097179, "grad_norm": 3.135453224182129, "learning_rate": 2.597827715720332e-06, "loss": 1.3602, "step": 33696 }, { "epoch": 2.6408307210031348, "grad_norm": 2.41892147064209, "learning_rate": 2.5975572773170654e-06, "loss": 1.5471, "step": 33697 }, { "epoch": 2.6409090909090907, "grad_norm": 2.3662264347076416, "learning_rate": 2.5972868480514613e-06, "loss": 1.6685, "step": 33698 }, { "epoch": 2.640987460815047, "grad_norm": 2.6737782955169678, "learning_rate": 2.5970164279245414e-06, "loss": 1.307, "step": 33699 }, { "epoch": 2.6410658307210033, "grad_norm": 3.254188299179077, "learning_rate": 2.596746016937337e-06, "loss": 1.081, "step": 33700 }, { "epoch": 2.641144200626959, "grad_norm": 4.24038553237915, "learning_rate": 2.5964756150908767e-06, "loss": 1.1457, "step": 33701 }, { "epoch": 2.641222570532915, "grad_norm": 2.0621109008789062, "learning_rate": 2.5962052223861885e-06, "loss": 1.0707, "step": 33702 }, { "epoch": 2.6413009404388714, "grad_norm": 2.9002315998077393, "learning_rate": 2.595934838824301e-06, "loss": 1.5281, "step": 33703 }, { "epoch": 2.6413793103448278, "grad_norm": 2.209754467010498, "learning_rate": 2.5956644644062423e-06, "loss": 1.3132, "step": 33704 }, { "epoch": 2.6414576802507836, "grad_norm": 2.9253807067871094, "learning_rate": 2.5953940991330427e-06, "loss": 1.1181, "step": 33705 }, { "epoch": 2.6415360501567395, "grad_norm": 7.460416793823242, "learning_rate": 2.5951237430057253e-06, "loss": 1.9315, "step": 33706 }, { "epoch": 2.641614420062696, "grad_norm": 3.2044787406921387, "learning_rate": 2.594853396025324e-06, "loss": 1.2544, "step": 33707 }, { "epoch": 2.641692789968652, "grad_norm": 3.220604419708252, "learning_rate": 2.5945830581928654e-06, "loss": 1.1569, "step": 33708 }, { "epoch": 2.641771159874608, "grad_norm": 6.693886756896973, "learning_rate": 2.5943127295093783e-06, "loss": 1.6543, "step": 33709 }, { "epoch": 2.6418495297805644, "grad_norm": 9.154741287231445, "learning_rate": 2.5940424099758887e-06, "loss": 0.7804, "step": 33710 }, { "epoch": 2.6419278996865203, "grad_norm": 2.5071239471435547, "learning_rate": 2.5937720995934247e-06, "loss": 1.3626, "step": 33711 }, { "epoch": 2.6420062695924766, "grad_norm": 4.152017593383789, "learning_rate": 2.593501798363016e-06, "loss": 1.796, "step": 33712 }, { "epoch": 2.6420846394984325, "grad_norm": 5.8423967361450195, "learning_rate": 2.5932315062856938e-06, "loss": 1.0791, "step": 33713 }, { "epoch": 2.642163009404389, "grad_norm": 4.041480541229248, "learning_rate": 2.59296122336248e-06, "loss": 1.7756, "step": 33714 }, { "epoch": 2.6422413793103448, "grad_norm": 2.2162463665008545, "learning_rate": 2.5926909495944056e-06, "loss": 1.5618, "step": 33715 }, { "epoch": 2.642319749216301, "grad_norm": 2.180868148803711, "learning_rate": 2.5924206849824985e-06, "loss": 1.3975, "step": 33716 }, { "epoch": 2.642398119122257, "grad_norm": 3.013573408126831, "learning_rate": 2.5921504295277867e-06, "loss": 0.957, "step": 33717 }, { "epoch": 2.6424764890282133, "grad_norm": 2.2715084552764893, "learning_rate": 2.5918801832312975e-06, "loss": 1.2999, "step": 33718 }, { "epoch": 2.642554858934169, "grad_norm": 10.655933380126953, "learning_rate": 2.591609946094059e-06, "loss": 1.3068, "step": 33719 }, { "epoch": 2.6426332288401255, "grad_norm": 2.8109467029571533, "learning_rate": 2.5913397181170997e-06, "loss": 1.1652, "step": 33720 }, { "epoch": 2.6427115987460814, "grad_norm": 2.8708150386810303, "learning_rate": 2.591069499301447e-06, "loss": 1.0802, "step": 33721 }, { "epoch": 2.6427899686520377, "grad_norm": 3.9694557189941406, "learning_rate": 2.590799289648128e-06, "loss": 1.4352, "step": 33722 }, { "epoch": 2.6428683385579936, "grad_norm": 2.5559606552124023, "learning_rate": 2.5905290891581715e-06, "loss": 1.7097, "step": 33723 }, { "epoch": 2.64294670846395, "grad_norm": 4.336261749267578, "learning_rate": 2.590258897832605e-06, "loss": 1.0681, "step": 33724 }, { "epoch": 2.643025078369906, "grad_norm": 4.111523628234863, "learning_rate": 2.5899887156724557e-06, "loss": 1.19, "step": 33725 }, { "epoch": 2.643103448275862, "grad_norm": 4.479902267456055, "learning_rate": 2.5897185426787513e-06, "loss": 1.6576, "step": 33726 }, { "epoch": 2.643181818181818, "grad_norm": 3.210765838623047, "learning_rate": 2.5894483788525193e-06, "loss": 1.2893, "step": 33727 }, { "epoch": 2.6432601880877744, "grad_norm": 3.9770944118499756, "learning_rate": 2.5891782241947904e-06, "loss": 1.3537, "step": 33728 }, { "epoch": 2.6433385579937303, "grad_norm": 2.7630555629730225, "learning_rate": 2.5889080787065845e-06, "loss": 1.1374, "step": 33729 }, { "epoch": 2.6434169278996866, "grad_norm": 2.810462713241577, "learning_rate": 2.5886379423889362e-06, "loss": 1.5968, "step": 33730 }, { "epoch": 2.6434952978056425, "grad_norm": 6.141519069671631, "learning_rate": 2.5883678152428726e-06, "loss": 2.2851, "step": 33731 }, { "epoch": 2.643573667711599, "grad_norm": 4.793710231781006, "learning_rate": 2.5880976972694173e-06, "loss": 1.3745, "step": 33732 }, { "epoch": 2.6436520376175547, "grad_norm": 3.068204164505005, "learning_rate": 2.587827588469599e-06, "loss": 1.3368, "step": 33733 }, { "epoch": 2.643730407523511, "grad_norm": 3.8425865173339844, "learning_rate": 2.587557488844444e-06, "loss": 1.4109, "step": 33734 }, { "epoch": 2.643808777429467, "grad_norm": 3.1456708908081055, "learning_rate": 2.5872873983949855e-06, "loss": 1.0928, "step": 33735 }, { "epoch": 2.6438871473354233, "grad_norm": 7.290587902069092, "learning_rate": 2.5870173171222434e-06, "loss": 1.3995, "step": 33736 }, { "epoch": 2.643965517241379, "grad_norm": 4.082096576690674, "learning_rate": 2.586747245027249e-06, "loss": 1.2263, "step": 33737 }, { "epoch": 2.6440438871473355, "grad_norm": 3.4366633892059326, "learning_rate": 2.586477182111028e-06, "loss": 1.5081, "step": 33738 }, { "epoch": 2.6441222570532914, "grad_norm": 3.4830732345581055, "learning_rate": 2.5862071283746083e-06, "loss": 1.0263, "step": 33739 }, { "epoch": 2.6442006269592477, "grad_norm": 6.5123372077941895, "learning_rate": 2.5859370838190174e-06, "loss": 0.6849, "step": 33740 }, { "epoch": 2.6442789968652036, "grad_norm": 2.17177414894104, "learning_rate": 2.585667048445282e-06, "loss": 1.1774, "step": 33741 }, { "epoch": 2.64435736677116, "grad_norm": 3.4437074661254883, "learning_rate": 2.5853970222544278e-06, "loss": 1.265, "step": 33742 }, { "epoch": 2.644435736677116, "grad_norm": 15.93336009979248, "learning_rate": 2.585127005247484e-06, "loss": 1.2, "step": 33743 }, { "epoch": 2.644514106583072, "grad_norm": 2.610412836074829, "learning_rate": 2.5848569974254767e-06, "loss": 0.9587, "step": 33744 }, { "epoch": 2.644592476489028, "grad_norm": 2.259921073913574, "learning_rate": 2.5845869987894324e-06, "loss": 1.1044, "step": 33745 }, { "epoch": 2.6446708463949844, "grad_norm": 2.6884028911590576, "learning_rate": 2.584317009340381e-06, "loss": 1.2022, "step": 33746 }, { "epoch": 2.6447492163009403, "grad_norm": 3.310352325439453, "learning_rate": 2.584047029079345e-06, "loss": 1.645, "step": 33747 }, { "epoch": 2.6448275862068966, "grad_norm": 2.2257421016693115, "learning_rate": 2.5837770580073507e-06, "loss": 1.0836, "step": 33748 }, { "epoch": 2.6449059561128525, "grad_norm": 2.0073821544647217, "learning_rate": 2.5835070961254297e-06, "loss": 1.5186, "step": 33749 }, { "epoch": 2.644984326018809, "grad_norm": 8.423587799072266, "learning_rate": 2.5832371434346086e-06, "loss": 1.2508, "step": 33750 }, { "epoch": 2.6450626959247647, "grad_norm": 4.900001525878906, "learning_rate": 2.5829671999359102e-06, "loss": 1.1768, "step": 33751 }, { "epoch": 2.645141065830721, "grad_norm": 3.2929084300994873, "learning_rate": 2.5826972656303628e-06, "loss": 1.1795, "step": 33752 }, { "epoch": 2.645219435736677, "grad_norm": 3.4171159267425537, "learning_rate": 2.5824273405189916e-06, "loss": 1.5664, "step": 33753 }, { "epoch": 2.6452978056426333, "grad_norm": 2.380061626434326, "learning_rate": 2.5821574246028293e-06, "loss": 1.2554, "step": 33754 }, { "epoch": 2.6453761755485896, "grad_norm": 11.260588645935059, "learning_rate": 2.5818875178828956e-06, "loss": 2.1498, "step": 33755 }, { "epoch": 2.6454545454545455, "grad_norm": 4.138291358947754, "learning_rate": 2.58161762036022e-06, "loss": 0.7179, "step": 33756 }, { "epoch": 2.6455329153605014, "grad_norm": 4.707350730895996, "learning_rate": 2.5813477320358287e-06, "loss": 1.426, "step": 33757 }, { "epoch": 2.6456112852664577, "grad_norm": 3.993917226791382, "learning_rate": 2.5810778529107487e-06, "loss": 1.3459, "step": 33758 }, { "epoch": 2.645689655172414, "grad_norm": 4.458212375640869, "learning_rate": 2.580807982986006e-06, "loss": 1.7003, "step": 33759 }, { "epoch": 2.64576802507837, "grad_norm": 2.7051620483398438, "learning_rate": 2.580538122262627e-06, "loss": 1.7237, "step": 33760 }, { "epoch": 2.645846394984326, "grad_norm": 4.3907694816589355, "learning_rate": 2.5802682707416404e-06, "loss": 0.9857, "step": 33761 }, { "epoch": 2.645924764890282, "grad_norm": 6.889855861663818, "learning_rate": 2.5799984284240663e-06, "loss": 1.4707, "step": 33762 }, { "epoch": 2.6460031347962385, "grad_norm": 3.3492138385772705, "learning_rate": 2.5797285953109374e-06, "loss": 0.9964, "step": 33763 }, { "epoch": 2.6460815047021944, "grad_norm": 2.70811128616333, "learning_rate": 2.579458771403278e-06, "loss": 1.3666, "step": 33764 }, { "epoch": 2.6461598746081503, "grad_norm": 8.794136047363281, "learning_rate": 2.5791889567021157e-06, "loss": 1.6765, "step": 33765 }, { "epoch": 2.6462382445141066, "grad_norm": 4.108503818511963, "learning_rate": 2.578919151208473e-06, "loss": 1.3656, "step": 33766 }, { "epoch": 2.646316614420063, "grad_norm": 3.967041492462158, "learning_rate": 2.578649354923377e-06, "loss": 1.4914, "step": 33767 }, { "epoch": 2.646394984326019, "grad_norm": 4.163796424865723, "learning_rate": 2.578379567847857e-06, "loss": 1.1535, "step": 33768 }, { "epoch": 2.6464733542319747, "grad_norm": 3.6890149116516113, "learning_rate": 2.5781097899829387e-06, "loss": 1.5562, "step": 33769 }, { "epoch": 2.646551724137931, "grad_norm": 2.262772798538208, "learning_rate": 2.5778400213296457e-06, "loss": 1.1022, "step": 33770 }, { "epoch": 2.6466300940438874, "grad_norm": 2.5902135372161865, "learning_rate": 2.577570261889004e-06, "loss": 1.6443, "step": 33771 }, { "epoch": 2.6467084639498433, "grad_norm": 5.119804859161377, "learning_rate": 2.577300511662042e-06, "loss": 1.3964, "step": 33772 }, { "epoch": 2.646786833855799, "grad_norm": 2.7505035400390625, "learning_rate": 2.5770307706497845e-06, "loss": 1.5535, "step": 33773 }, { "epoch": 2.6468652037617555, "grad_norm": 3.5729546546936035, "learning_rate": 2.576761038853256e-06, "loss": 1.4852, "step": 33774 }, { "epoch": 2.646943573667712, "grad_norm": 2.336379289627075, "learning_rate": 2.576491316273485e-06, "loss": 1.2698, "step": 33775 }, { "epoch": 2.6470219435736677, "grad_norm": 2.0142955780029297, "learning_rate": 2.576221602911496e-06, "loss": 0.6585, "step": 33776 }, { "epoch": 2.6471003134796236, "grad_norm": 3.9632513523101807, "learning_rate": 2.5759518987683154e-06, "loss": 1.0862, "step": 33777 }, { "epoch": 2.64717868338558, "grad_norm": 2.7962684631347656, "learning_rate": 2.575682203844968e-06, "loss": 1.2574, "step": 33778 }, { "epoch": 2.6472570532915363, "grad_norm": 3.3541667461395264, "learning_rate": 2.5754125181424805e-06, "loss": 1.3281, "step": 33779 }, { "epoch": 2.647335423197492, "grad_norm": 2.7058656215667725, "learning_rate": 2.575142841661881e-06, "loss": 1.5185, "step": 33780 }, { "epoch": 2.647413793103448, "grad_norm": 4.623234272003174, "learning_rate": 2.5748731744041885e-06, "loss": 1.0478, "step": 33781 }, { "epoch": 2.6474921630094044, "grad_norm": 3.693626642227173, "learning_rate": 2.574603516370434e-06, "loss": 0.914, "step": 33782 }, { "epoch": 2.6475705329153607, "grad_norm": 3.8125195503234863, "learning_rate": 2.5743338675616423e-06, "loss": 1.3857, "step": 33783 }, { "epoch": 2.6476489028213166, "grad_norm": 2.2766289710998535, "learning_rate": 2.5740642279788414e-06, "loss": 1.4513, "step": 33784 }, { "epoch": 2.6477272727272725, "grad_norm": 4.263124942779541, "learning_rate": 2.5737945976230516e-06, "loss": 1.761, "step": 33785 }, { "epoch": 2.647805642633229, "grad_norm": 2.6715731620788574, "learning_rate": 2.573524976495299e-06, "loss": 1.7315, "step": 33786 }, { "epoch": 2.647884012539185, "grad_norm": 3.199742317199707, "learning_rate": 2.573255364596615e-06, "loss": 1.3908, "step": 33787 }, { "epoch": 2.647962382445141, "grad_norm": 3.119734525680542, "learning_rate": 2.5729857619280196e-06, "loss": 1.1674, "step": 33788 }, { "epoch": 2.648040752351097, "grad_norm": 5.077907085418701, "learning_rate": 2.572716168490539e-06, "loss": 1.5047, "step": 33789 }, { "epoch": 2.6481191222570533, "grad_norm": 4.726511478424072, "learning_rate": 2.5724465842852e-06, "loss": 1.3194, "step": 33790 }, { "epoch": 2.6481974921630096, "grad_norm": 3.175572633743286, "learning_rate": 2.572177009313027e-06, "loss": 1.1367, "step": 33791 }, { "epoch": 2.6482758620689655, "grad_norm": 3.6679913997650146, "learning_rate": 2.571907443575046e-06, "loss": 1.3163, "step": 33792 }, { "epoch": 2.6483542319749214, "grad_norm": 3.3367860317230225, "learning_rate": 2.5716378870722815e-06, "loss": 1.346, "step": 33793 }, { "epoch": 2.6484326018808777, "grad_norm": 3.0481929779052734, "learning_rate": 2.5713683398057592e-06, "loss": 1.2118, "step": 33794 }, { "epoch": 2.648510971786834, "grad_norm": 2.4506494998931885, "learning_rate": 2.5710988017765048e-06, "loss": 1.4332, "step": 33795 }, { "epoch": 2.64858934169279, "grad_norm": 2.7497615814208984, "learning_rate": 2.5708292729855422e-06, "loss": 1.2901, "step": 33796 }, { "epoch": 2.648667711598746, "grad_norm": 2.3615517616271973, "learning_rate": 2.5705597534338977e-06, "loss": 1.3559, "step": 33797 }, { "epoch": 2.648746081504702, "grad_norm": 3.036468267440796, "learning_rate": 2.5702902431225962e-06, "loss": 1.6399, "step": 33798 }, { "epoch": 2.6488244514106585, "grad_norm": 2.7017061710357666, "learning_rate": 2.5700207420526647e-06, "loss": 1.1285, "step": 33799 }, { "epoch": 2.6489028213166144, "grad_norm": 2.887989044189453, "learning_rate": 2.5697512502251222e-06, "loss": 1.3698, "step": 33800 }, { "epoch": 2.6489811912225707, "grad_norm": 2.3485379219055176, "learning_rate": 2.569481767641e-06, "loss": 1.4152, "step": 33801 }, { "epoch": 2.6490595611285266, "grad_norm": 2.5909390449523926, "learning_rate": 2.569212294301322e-06, "loss": 1.4947, "step": 33802 }, { "epoch": 2.649137931034483, "grad_norm": 4.936563491821289, "learning_rate": 2.56894283020711e-06, "loss": 1.1376, "step": 33803 }, { "epoch": 2.649216300940439, "grad_norm": 2.655123472213745, "learning_rate": 2.568673375359391e-06, "loss": 1.5197, "step": 33804 }, { "epoch": 2.649294670846395, "grad_norm": 2.1923935413360596, "learning_rate": 2.568403929759188e-06, "loss": 1.2484, "step": 33805 }, { "epoch": 2.649373040752351, "grad_norm": 3.7167036533355713, "learning_rate": 2.5681344934075314e-06, "loss": 1.4651, "step": 33806 }, { "epoch": 2.6494514106583074, "grad_norm": 2.5383293628692627, "learning_rate": 2.5678650663054394e-06, "loss": 1.3473, "step": 33807 }, { "epoch": 2.6495297805642632, "grad_norm": 2.290839910507202, "learning_rate": 2.56759564845394e-06, "loss": 1.1635, "step": 33808 }, { "epoch": 2.6496081504702196, "grad_norm": 2.326896905899048, "learning_rate": 2.5673262398540578e-06, "loss": 1.4134, "step": 33809 }, { "epoch": 2.6496865203761755, "grad_norm": 4.456592082977295, "learning_rate": 2.5670568405068163e-06, "loss": 1.1906, "step": 33810 }, { "epoch": 2.649764890282132, "grad_norm": 5.800734996795654, "learning_rate": 2.5667874504132417e-06, "loss": 1.2332, "step": 33811 }, { "epoch": 2.6498432601880877, "grad_norm": 3.426612615585327, "learning_rate": 2.5665180695743575e-06, "loss": 1.5757, "step": 33812 }, { "epoch": 2.649921630094044, "grad_norm": 3.489457368850708, "learning_rate": 2.5662486979911882e-06, "loss": 1.1651, "step": 33813 }, { "epoch": 2.65, "grad_norm": 3.9528722763061523, "learning_rate": 2.5659793356647593e-06, "loss": 1.5769, "step": 33814 }, { "epoch": 2.6500783699059562, "grad_norm": 5.6290364265441895, "learning_rate": 2.565709982596095e-06, "loss": 1.53, "step": 33815 }, { "epoch": 2.650156739811912, "grad_norm": 3.694040060043335, "learning_rate": 2.5654406387862192e-06, "loss": 1.1633, "step": 33816 }, { "epoch": 2.6502351097178685, "grad_norm": 3.6296935081481934, "learning_rate": 2.565171304236159e-06, "loss": 1.5056, "step": 33817 }, { "epoch": 2.6503134796238244, "grad_norm": 2.809255599975586, "learning_rate": 2.564901978946932e-06, "loss": 1.6304, "step": 33818 }, { "epoch": 2.6503918495297807, "grad_norm": 3.4186973571777344, "learning_rate": 2.5646326629195695e-06, "loss": 1.2052, "step": 33819 }, { "epoch": 2.6504702194357366, "grad_norm": 4.025119304656982, "learning_rate": 2.564363356155094e-06, "loss": 1.1034, "step": 33820 }, { "epoch": 2.650548589341693, "grad_norm": 3.0165319442749023, "learning_rate": 2.5640940586545303e-06, "loss": 1.5048, "step": 33821 }, { "epoch": 2.650626959247649, "grad_norm": 3.9864139556884766, "learning_rate": 2.5638247704189e-06, "loss": 1.7686, "step": 33822 }, { "epoch": 2.650705329153605, "grad_norm": 2.9742870330810547, "learning_rate": 2.5635554914492293e-06, "loss": 1.4042, "step": 33823 }, { "epoch": 2.650783699059561, "grad_norm": 3.0666680335998535, "learning_rate": 2.5632862217465405e-06, "loss": 1.5161, "step": 33824 }, { "epoch": 2.6508620689655173, "grad_norm": 4.743138313293457, "learning_rate": 2.563016961311862e-06, "loss": 1.418, "step": 33825 }, { "epoch": 2.6509404388714732, "grad_norm": 3.3558812141418457, "learning_rate": 2.562747710146214e-06, "loss": 1.3481, "step": 33826 }, { "epoch": 2.6510188087774296, "grad_norm": 3.3161675930023193, "learning_rate": 2.562478468250622e-06, "loss": 1.1617, "step": 33827 }, { "epoch": 2.6510971786833855, "grad_norm": 3.4759652614593506, "learning_rate": 2.5622092356261098e-06, "loss": 1.2964, "step": 33828 }, { "epoch": 2.651175548589342, "grad_norm": 2.9944541454315186, "learning_rate": 2.5619400122737016e-06, "loss": 1.5083, "step": 33829 }, { "epoch": 2.6512539184952977, "grad_norm": 2.0752980709075928, "learning_rate": 2.5616707981944213e-06, "loss": 1.3909, "step": 33830 }, { "epoch": 2.651332288401254, "grad_norm": 2.5324947834014893, "learning_rate": 2.561401593389292e-06, "loss": 0.6614, "step": 33831 }, { "epoch": 2.65141065830721, "grad_norm": 2.969510793685913, "learning_rate": 2.56113239785934e-06, "loss": 1.1299, "step": 33832 }, { "epoch": 2.6514890282131662, "grad_norm": 8.736900329589844, "learning_rate": 2.5608632116055866e-06, "loss": 1.2485, "step": 33833 }, { "epoch": 2.651567398119122, "grad_norm": 2.3105640411376953, "learning_rate": 2.560594034629057e-06, "loss": 1.3626, "step": 33834 }, { "epoch": 2.6516457680250785, "grad_norm": 4.348479747772217, "learning_rate": 2.560324866930775e-06, "loss": 1.5953, "step": 33835 }, { "epoch": 2.6517241379310343, "grad_norm": 2.526595115661621, "learning_rate": 2.560055708511766e-06, "loss": 1.2026, "step": 33836 }, { "epoch": 2.6518025078369907, "grad_norm": 3.8532297611236572, "learning_rate": 2.5597865593730474e-06, "loss": 1.3252, "step": 33837 }, { "epoch": 2.6518808777429466, "grad_norm": 3.9984841346740723, "learning_rate": 2.5595174195156504e-06, "loss": 1.5993, "step": 33838 }, { "epoch": 2.651959247648903, "grad_norm": 4.807622909545898, "learning_rate": 2.5592482889405943e-06, "loss": 1.2235, "step": 33839 }, { "epoch": 2.652037617554859, "grad_norm": 4.129284381866455, "learning_rate": 2.558979167648907e-06, "loss": 1.2252, "step": 33840 }, { "epoch": 2.652115987460815, "grad_norm": 3.481520175933838, "learning_rate": 2.5587100556416066e-06, "loss": 1.1342, "step": 33841 }, { "epoch": 2.652194357366771, "grad_norm": 2.5520365238189697, "learning_rate": 2.558440952919718e-06, "loss": 1.275, "step": 33842 }, { "epoch": 2.6522727272727273, "grad_norm": 4.77232027053833, "learning_rate": 2.55817185948427e-06, "loss": 1.1446, "step": 33843 }, { "epoch": 2.6523510971786832, "grad_norm": 7.366043567657471, "learning_rate": 2.557902775336279e-06, "loss": 1.5222, "step": 33844 }, { "epoch": 2.6524294670846396, "grad_norm": 2.8430395126342773, "learning_rate": 2.5576337004767717e-06, "loss": 1.3158, "step": 33845 }, { "epoch": 2.6525078369905954, "grad_norm": 3.217041254043579, "learning_rate": 2.5573646349067717e-06, "loss": 1.4886, "step": 33846 }, { "epoch": 2.652586206896552, "grad_norm": 2.5320448875427246, "learning_rate": 2.5570955786273023e-06, "loss": 1.3246, "step": 33847 }, { "epoch": 2.6526645768025077, "grad_norm": 8.822564125061035, "learning_rate": 2.556826531639386e-06, "loss": 1.6252, "step": 33848 }, { "epoch": 2.652742946708464, "grad_norm": 2.3365962505340576, "learning_rate": 2.5565574939440467e-06, "loss": 1.3016, "step": 33849 }, { "epoch": 2.6528213166144203, "grad_norm": 4.284555435180664, "learning_rate": 2.556288465542308e-06, "loss": 1.4627, "step": 33850 }, { "epoch": 2.652899686520376, "grad_norm": 3.1944358348846436, "learning_rate": 2.556019446435193e-06, "loss": 1.1036, "step": 33851 }, { "epoch": 2.652978056426332, "grad_norm": 4.193277359008789, "learning_rate": 2.555750436623724e-06, "loss": 1.4105, "step": 33852 }, { "epoch": 2.6530564263322884, "grad_norm": 2.4051215648651123, "learning_rate": 2.5554814361089254e-06, "loss": 1.3797, "step": 33853 }, { "epoch": 2.6531347962382448, "grad_norm": 3.624109983444214, "learning_rate": 2.5552124448918193e-06, "loss": 1.3358, "step": 33854 }, { "epoch": 2.6532131661442007, "grad_norm": 6.201800346374512, "learning_rate": 2.5549434629734316e-06, "loss": 2.1062, "step": 33855 }, { "epoch": 2.6532915360501566, "grad_norm": 3.1274197101593018, "learning_rate": 2.554674490354779e-06, "loss": 1.5307, "step": 33856 }, { "epoch": 2.653369905956113, "grad_norm": 2.872190237045288, "learning_rate": 2.5544055270368916e-06, "loss": 1.0665, "step": 33857 }, { "epoch": 2.653448275862069, "grad_norm": 2.4616827964782715, "learning_rate": 2.5541365730207907e-06, "loss": 1.1331, "step": 33858 }, { "epoch": 2.653526645768025, "grad_norm": 6.957147598266602, "learning_rate": 2.553867628307496e-06, "loss": 1.6879, "step": 33859 }, { "epoch": 2.653605015673981, "grad_norm": 6.875126361846924, "learning_rate": 2.5535986928980323e-06, "loss": 1.7399, "step": 33860 }, { "epoch": 2.6536833855799373, "grad_norm": 3.654982089996338, "learning_rate": 2.553329766793421e-06, "loss": 1.483, "step": 33861 }, { "epoch": 2.6537617554858937, "grad_norm": 2.8989250659942627, "learning_rate": 2.55306084999469e-06, "loss": 1.4746, "step": 33862 }, { "epoch": 2.6538401253918495, "grad_norm": 3.055436849594116, "learning_rate": 2.552791942502857e-06, "loss": 1.4208, "step": 33863 }, { "epoch": 2.6539184952978054, "grad_norm": 11.642801284790039, "learning_rate": 2.5525230443189463e-06, "loss": 0.9428, "step": 33864 }, { "epoch": 2.6539968652037618, "grad_norm": 5.7410502433776855, "learning_rate": 2.5522541554439816e-06, "loss": 1.1449, "step": 33865 }, { "epoch": 2.654075235109718, "grad_norm": 2.766624689102173, "learning_rate": 2.5519852758789842e-06, "loss": 1.4289, "step": 33866 }, { "epoch": 2.654153605015674, "grad_norm": 2.6003170013427734, "learning_rate": 2.5517164056249775e-06, "loss": 0.9509, "step": 33867 }, { "epoch": 2.65423197492163, "grad_norm": 4.2575883865356445, "learning_rate": 2.5514475446829846e-06, "loss": 1.3827, "step": 33868 }, { "epoch": 2.654310344827586, "grad_norm": 2.516463041305542, "learning_rate": 2.5511786930540266e-06, "loss": 1.4436, "step": 33869 }, { "epoch": 2.6543887147335425, "grad_norm": 3.0074408054351807, "learning_rate": 2.5509098507391273e-06, "loss": 1.4896, "step": 33870 }, { "epoch": 2.6544670846394984, "grad_norm": 5.9954118728637695, "learning_rate": 2.5506410177393092e-06, "loss": 1.6133, "step": 33871 }, { "epoch": 2.6545454545454543, "grad_norm": 3.3997457027435303, "learning_rate": 2.550372194055595e-06, "loss": 1.0885, "step": 33872 }, { "epoch": 2.6546238244514107, "grad_norm": 2.7552857398986816, "learning_rate": 2.5501033796890084e-06, "loss": 1.6782, "step": 33873 }, { "epoch": 2.654702194357367, "grad_norm": 4.853179454803467, "learning_rate": 2.5498345746405682e-06, "loss": 1.7265, "step": 33874 }, { "epoch": 2.654780564263323, "grad_norm": 4.52899694442749, "learning_rate": 2.549565778911297e-06, "loss": 1.1856, "step": 33875 }, { "epoch": 2.6548589341692788, "grad_norm": 13.72545051574707, "learning_rate": 2.5492969925022202e-06, "loss": 1.1854, "step": 33876 }, { "epoch": 2.654937304075235, "grad_norm": 2.362773895263672, "learning_rate": 2.5490282154143618e-06, "loss": 1.3699, "step": 33877 }, { "epoch": 2.6550156739811914, "grad_norm": 3.3741958141326904, "learning_rate": 2.5487594476487387e-06, "loss": 1.3388, "step": 33878 }, { "epoch": 2.6550940438871473, "grad_norm": 3.974980592727661, "learning_rate": 2.548490689206376e-06, "loss": 0.848, "step": 33879 }, { "epoch": 2.655172413793103, "grad_norm": 2.8506243228912354, "learning_rate": 2.5482219400882934e-06, "loss": 1.1548, "step": 33880 }, { "epoch": 2.6552507836990595, "grad_norm": 2.4565646648406982, "learning_rate": 2.547953200295519e-06, "loss": 1.6018, "step": 33881 }, { "epoch": 2.655329153605016, "grad_norm": 9.63699722290039, "learning_rate": 2.547684469829069e-06, "loss": 1.4514, "step": 33882 }, { "epoch": 2.6554075235109718, "grad_norm": 3.5669240951538086, "learning_rate": 2.547415748689968e-06, "loss": 1.5642, "step": 33883 }, { "epoch": 2.6554858934169276, "grad_norm": 6.167746543884277, "learning_rate": 2.5471470368792385e-06, "loss": 2.4115, "step": 33884 }, { "epoch": 2.655564263322884, "grad_norm": 2.881885051727295, "learning_rate": 2.546878334397901e-06, "loss": 0.9398, "step": 33885 }, { "epoch": 2.6556426332288403, "grad_norm": 3.1567025184631348, "learning_rate": 2.5466096412469775e-06, "loss": 1.3417, "step": 33886 }, { "epoch": 2.655721003134796, "grad_norm": 4.690842628479004, "learning_rate": 2.5463409574274923e-06, "loss": 1.2735, "step": 33887 }, { "epoch": 2.655799373040752, "grad_norm": 3.267505645751953, "learning_rate": 2.546072282940467e-06, "loss": 1.104, "step": 33888 }, { "epoch": 2.6558777429467084, "grad_norm": 3.745600461959839, "learning_rate": 2.5458036177869183e-06, "loss": 1.3722, "step": 33889 }, { "epoch": 2.6559561128526648, "grad_norm": 4.117669582366943, "learning_rate": 2.545534961967874e-06, "loss": 1.0545, "step": 33890 }, { "epoch": 2.6560344827586206, "grad_norm": 2.180917501449585, "learning_rate": 2.545266315484355e-06, "loss": 1.327, "step": 33891 }, { "epoch": 2.6561128526645765, "grad_norm": 3.2590627670288086, "learning_rate": 2.544997678337383e-06, "loss": 1.3744, "step": 33892 }, { "epoch": 2.656191222570533, "grad_norm": 3.202026128768921, "learning_rate": 2.5447290505279778e-06, "loss": 0.7869, "step": 33893 }, { "epoch": 2.656269592476489, "grad_norm": 4.593881607055664, "learning_rate": 2.544460432057159e-06, "loss": 1.4375, "step": 33894 }, { "epoch": 2.656347962382445, "grad_norm": 2.604410171508789, "learning_rate": 2.5441918229259543e-06, "loss": 1.2919, "step": 33895 }, { "epoch": 2.6564263322884014, "grad_norm": 3.9936397075653076, "learning_rate": 2.5439232231353838e-06, "loss": 1.288, "step": 33896 }, { "epoch": 2.6565047021943573, "grad_norm": 3.2021803855895996, "learning_rate": 2.5436546326864665e-06, "loss": 1.2753, "step": 33897 }, { "epoch": 2.6565830721003136, "grad_norm": 3.088165760040283, "learning_rate": 2.5433860515802256e-06, "loss": 0.997, "step": 33898 }, { "epoch": 2.6566614420062695, "grad_norm": 2.9939045906066895, "learning_rate": 2.543117479817682e-06, "loss": 1.5417, "step": 33899 }, { "epoch": 2.656739811912226, "grad_norm": 3.5750200748443604, "learning_rate": 2.542848917399857e-06, "loss": 1.514, "step": 33900 }, { "epoch": 2.6568181818181817, "grad_norm": 3.7472972869873047, "learning_rate": 2.5425803643277737e-06, "loss": 1.3349, "step": 33901 }, { "epoch": 2.656896551724138, "grad_norm": 4.027103900909424, "learning_rate": 2.5423118206024516e-06, "loss": 1.2138, "step": 33902 }, { "epoch": 2.656974921630094, "grad_norm": 6.590956687927246, "learning_rate": 2.542043286224913e-06, "loss": 1.4443, "step": 33903 }, { "epoch": 2.6570532915360503, "grad_norm": 2.8700010776519775, "learning_rate": 2.54177476119618e-06, "loss": 1.3756, "step": 33904 }, { "epoch": 2.657131661442006, "grad_norm": 2.5194711685180664, "learning_rate": 2.5415062455172727e-06, "loss": 1.4117, "step": 33905 }, { "epoch": 2.6572100313479625, "grad_norm": 5.598899841308594, "learning_rate": 2.541237739189213e-06, "loss": 1.5769, "step": 33906 }, { "epoch": 2.6572884012539184, "grad_norm": 5.7467265129089355, "learning_rate": 2.540969242213024e-06, "loss": 1.0059, "step": 33907 }, { "epoch": 2.6573667711598747, "grad_norm": 3.303144931793213, "learning_rate": 2.540700754589721e-06, "loss": 1.3997, "step": 33908 }, { "epoch": 2.6574451410658306, "grad_norm": 3.3347718715667725, "learning_rate": 2.540432276320331e-06, "loss": 1.0924, "step": 33909 }, { "epoch": 2.657523510971787, "grad_norm": 3.707284927368164, "learning_rate": 2.540163807405873e-06, "loss": 1.2605, "step": 33910 }, { "epoch": 2.657601880877743, "grad_norm": 6.32688570022583, "learning_rate": 2.539895347847371e-06, "loss": 0.9476, "step": 33911 }, { "epoch": 2.657680250783699, "grad_norm": 3.9778411388397217, "learning_rate": 2.5396268976458417e-06, "loss": 1.4816, "step": 33912 }, { "epoch": 2.657758620689655, "grad_norm": 4.784805774688721, "learning_rate": 2.5393584568023056e-06, "loss": 1.4701, "step": 33913 }, { "epoch": 2.6578369905956114, "grad_norm": 3.045501470565796, "learning_rate": 2.5390900253177895e-06, "loss": 1.339, "step": 33914 }, { "epoch": 2.6579153605015673, "grad_norm": 3.4823431968688965, "learning_rate": 2.53882160319331e-06, "loss": 1.3839, "step": 33915 }, { "epoch": 2.6579937304075236, "grad_norm": 3.001469612121582, "learning_rate": 2.5385531904298887e-06, "loss": 1.2877, "step": 33916 }, { "epoch": 2.6580721003134795, "grad_norm": 6.976798057556152, "learning_rate": 2.5382847870285466e-06, "loss": 1.7762, "step": 33917 }, { "epoch": 2.658150470219436, "grad_norm": 2.374434232711792, "learning_rate": 2.5380163929903046e-06, "loss": 1.3678, "step": 33918 }, { "epoch": 2.6582288401253917, "grad_norm": 4.864282131195068, "learning_rate": 2.537748008316184e-06, "loss": 1.3576, "step": 33919 }, { "epoch": 2.658307210031348, "grad_norm": 3.0891847610473633, "learning_rate": 2.537479633007205e-06, "loss": 1.2178, "step": 33920 }, { "epoch": 2.658385579937304, "grad_norm": 2.852842092514038, "learning_rate": 2.537211267064389e-06, "loss": 1.1961, "step": 33921 }, { "epoch": 2.6584639498432603, "grad_norm": 3.8090860843658447, "learning_rate": 2.5369429104887556e-06, "loss": 1.2215, "step": 33922 }, { "epoch": 2.658542319749216, "grad_norm": 2.634399175643921, "learning_rate": 2.536674563281327e-06, "loss": 1.2547, "step": 33923 }, { "epoch": 2.6586206896551725, "grad_norm": 3.0917751789093018, "learning_rate": 2.5364062254431227e-06, "loss": 1.6004, "step": 33924 }, { "epoch": 2.6586990595611284, "grad_norm": 3.747117042541504, "learning_rate": 2.5361378969751636e-06, "loss": 1.2571, "step": 33925 }, { "epoch": 2.6587774294670847, "grad_norm": 10.442867279052734, "learning_rate": 2.5358695778784716e-06, "loss": 0.8424, "step": 33926 }, { "epoch": 2.6588557993730406, "grad_norm": 5.195105075836182, "learning_rate": 2.535601268154065e-06, "loss": 1.309, "step": 33927 }, { "epoch": 2.658934169278997, "grad_norm": 3.9517722129821777, "learning_rate": 2.535332967802965e-06, "loss": 1.0725, "step": 33928 }, { "epoch": 2.659012539184953, "grad_norm": 1.9832115173339844, "learning_rate": 2.5350646768261954e-06, "loss": 0.9815, "step": 33929 }, { "epoch": 2.659090909090909, "grad_norm": 3.7935426235198975, "learning_rate": 2.5347963952247712e-06, "loss": 1.046, "step": 33930 }, { "epoch": 2.659169278996865, "grad_norm": 2.87451434135437, "learning_rate": 2.534528122999714e-06, "loss": 1.1193, "step": 33931 }, { "epoch": 2.6592476489028214, "grad_norm": 3.2100305557250977, "learning_rate": 2.5342598601520463e-06, "loss": 1.4774, "step": 33932 }, { "epoch": 2.6593260188087773, "grad_norm": 4.804046630859375, "learning_rate": 2.5339916066827908e-06, "loss": 0.8385, "step": 33933 }, { "epoch": 2.6594043887147336, "grad_norm": 3.9433627128601074, "learning_rate": 2.5337233625929614e-06, "loss": 1.4604, "step": 33934 }, { "epoch": 2.6594827586206895, "grad_norm": 3.7886252403259277, "learning_rate": 2.5334551278835828e-06, "loss": 1.31, "step": 33935 }, { "epoch": 2.659561128526646, "grad_norm": 2.8415379524230957, "learning_rate": 2.5331869025556734e-06, "loss": 0.9933, "step": 33936 }, { "epoch": 2.6596394984326017, "grad_norm": 3.0601918697357178, "learning_rate": 2.532918686610254e-06, "loss": 1.3409, "step": 33937 }, { "epoch": 2.659717868338558, "grad_norm": 2.4349615573883057, "learning_rate": 2.5326504800483443e-06, "loss": 1.2748, "step": 33938 }, { "epoch": 2.659796238244514, "grad_norm": 7.886950969696045, "learning_rate": 2.5323822828709654e-06, "loss": 1.7429, "step": 33939 }, { "epoch": 2.6598746081504703, "grad_norm": 3.860461711883545, "learning_rate": 2.532114095079137e-06, "loss": 1.3139, "step": 33940 }, { "epoch": 2.6599529780564266, "grad_norm": 4.3668599128723145, "learning_rate": 2.5318459166738784e-06, "loss": 1.419, "step": 33941 }, { "epoch": 2.6600313479623825, "grad_norm": 2.892662525177002, "learning_rate": 2.5315777476562107e-06, "loss": 1.2721, "step": 33942 }, { "epoch": 2.6601097178683384, "grad_norm": 3.3675689697265625, "learning_rate": 2.531309588027153e-06, "loss": 1.5174, "step": 33943 }, { "epoch": 2.6601880877742947, "grad_norm": 3.725250482559204, "learning_rate": 2.5310414377877278e-06, "loss": 1.3104, "step": 33944 }, { "epoch": 2.660266457680251, "grad_norm": 3.1008126735687256, "learning_rate": 2.5307732969389492e-06, "loss": 1.2917, "step": 33945 }, { "epoch": 2.660344827586207, "grad_norm": 3.9597296714782715, "learning_rate": 2.5305051654818425e-06, "loss": 1.5781, "step": 33946 }, { "epoch": 2.660423197492163, "grad_norm": 3.3029966354370117, "learning_rate": 2.5302370434174256e-06, "loss": 1.097, "step": 33947 }, { "epoch": 2.660501567398119, "grad_norm": 4.948567867279053, "learning_rate": 2.52996893074672e-06, "loss": 1.6875, "step": 33948 }, { "epoch": 2.6605799373040755, "grad_norm": 3.371575355529785, "learning_rate": 2.5297008274707426e-06, "loss": 1.0852, "step": 33949 }, { "epoch": 2.6606583072100314, "grad_norm": 2.222900152206421, "learning_rate": 2.5294327335905123e-06, "loss": 1.5125, "step": 33950 }, { "epoch": 2.6607366771159873, "grad_norm": 2.0719125270843506, "learning_rate": 2.5291646491070534e-06, "loss": 1.4411, "step": 33951 }, { "epoch": 2.6608150470219436, "grad_norm": 2.9677700996398926, "learning_rate": 2.5288965740213843e-06, "loss": 1.5703, "step": 33952 }, { "epoch": 2.6608934169279, "grad_norm": 3.0422754287719727, "learning_rate": 2.528628508334521e-06, "loss": 1.025, "step": 33953 }, { "epoch": 2.660971786833856, "grad_norm": 3.206927537918091, "learning_rate": 2.5283604520474863e-06, "loss": 1.2653, "step": 33954 }, { "epoch": 2.6610501567398117, "grad_norm": 6.414917469024658, "learning_rate": 2.5280924051612986e-06, "loss": 1.3145, "step": 33955 }, { "epoch": 2.661128526645768, "grad_norm": 5.04430627822876, "learning_rate": 2.527824367676978e-06, "loss": 1.312, "step": 33956 }, { "epoch": 2.6612068965517244, "grad_norm": 2.6644601821899414, "learning_rate": 2.5275563395955436e-06, "loss": 1.5276, "step": 33957 }, { "epoch": 2.6612852664576803, "grad_norm": 3.1886215209960938, "learning_rate": 2.527288320918015e-06, "loss": 1.2834, "step": 33958 }, { "epoch": 2.661363636363636, "grad_norm": 2.887749195098877, "learning_rate": 2.5270203116454118e-06, "loss": 1.4415, "step": 33959 }, { "epoch": 2.6614420062695925, "grad_norm": 5.1351823806762695, "learning_rate": 2.5267523117787524e-06, "loss": 1.3841, "step": 33960 }, { "epoch": 2.661520376175549, "grad_norm": 2.338134527206421, "learning_rate": 2.526484321319057e-06, "loss": 1.2453, "step": 33961 }, { "epoch": 2.6615987460815047, "grad_norm": 7.3885321617126465, "learning_rate": 2.5262163402673457e-06, "loss": 0.5851, "step": 33962 }, { "epoch": 2.6616771159874606, "grad_norm": 3.038151502609253, "learning_rate": 2.5259483686246376e-06, "loss": 1.2649, "step": 33963 }, { "epoch": 2.661755485893417, "grad_norm": 2.8768198490142822, "learning_rate": 2.525680406391948e-06, "loss": 1.2908, "step": 33964 }, { "epoch": 2.6618338557993733, "grad_norm": 3.5010135173797607, "learning_rate": 2.5254124535703007e-06, "loss": 1.5438, "step": 33965 }, { "epoch": 2.661912225705329, "grad_norm": 4.797223091125488, "learning_rate": 2.525144510160714e-06, "loss": 1.782, "step": 33966 }, { "epoch": 2.661990595611285, "grad_norm": 2.6791508197784424, "learning_rate": 2.524876576164208e-06, "loss": 1.4649, "step": 33967 }, { "epoch": 2.6620689655172414, "grad_norm": 2.0523030757904053, "learning_rate": 2.5246086515817982e-06, "loss": 1.311, "step": 33968 }, { "epoch": 2.6621473354231977, "grad_norm": 2.552633047103882, "learning_rate": 2.5243407364145033e-06, "loss": 1.6055, "step": 33969 }, { "epoch": 2.6622257053291536, "grad_norm": 2.123438596725464, "learning_rate": 2.5240728306633492e-06, "loss": 1.3905, "step": 33970 }, { "epoch": 2.6623040752351095, "grad_norm": 2.8502635955810547, "learning_rate": 2.523804934329348e-06, "loss": 1.6386, "step": 33971 }, { "epoch": 2.662382445141066, "grad_norm": 2.8429224491119385, "learning_rate": 2.5235370474135214e-06, "loss": 1.5689, "step": 33972 }, { "epoch": 2.662460815047022, "grad_norm": 2.9900283813476562, "learning_rate": 2.5232691699168877e-06, "loss": 1.3618, "step": 33973 }, { "epoch": 2.662539184952978, "grad_norm": 3.0716631412506104, "learning_rate": 2.5230013018404665e-06, "loss": 1.388, "step": 33974 }, { "epoch": 2.662617554858934, "grad_norm": 3.4731857776641846, "learning_rate": 2.522733443185275e-06, "loss": 1.0373, "step": 33975 }, { "epoch": 2.6626959247648903, "grad_norm": 2.9613029956817627, "learning_rate": 2.5224655939523336e-06, "loss": 1.0942, "step": 33976 }, { "epoch": 2.6627742946708466, "grad_norm": 9.555502891540527, "learning_rate": 2.5221977541426605e-06, "loss": 1.3703, "step": 33977 }, { "epoch": 2.6628526645768025, "grad_norm": 2.4131627082824707, "learning_rate": 2.5219299237572747e-06, "loss": 1.2748, "step": 33978 }, { "epoch": 2.6629310344827584, "grad_norm": 3.5088369846343994, "learning_rate": 2.5216621027971943e-06, "loss": 1.7209, "step": 33979 }, { "epoch": 2.6630094043887147, "grad_norm": 2.6783571243286133, "learning_rate": 2.521394291263438e-06, "loss": 1.6235, "step": 33980 }, { "epoch": 2.663087774294671, "grad_norm": 2.4668543338775635, "learning_rate": 2.5211264891570274e-06, "loss": 1.3898, "step": 33981 }, { "epoch": 2.663166144200627, "grad_norm": 2.2791314125061035, "learning_rate": 2.520858696478976e-06, "loss": 0.9818, "step": 33982 }, { "epoch": 2.663244514106583, "grad_norm": 9.487872123718262, "learning_rate": 2.520590913230303e-06, "loss": 1.2839, "step": 33983 }, { "epoch": 2.663322884012539, "grad_norm": 4.819961071014404, "learning_rate": 2.5203231394120307e-06, "loss": 1.9336, "step": 33984 }, { "epoch": 2.6634012539184955, "grad_norm": 2.9486706256866455, "learning_rate": 2.520055375025178e-06, "loss": 1.4986, "step": 33985 }, { "epoch": 2.6634796238244514, "grad_norm": 6.531561374664307, "learning_rate": 2.5197876200707577e-06, "loss": 0.9515, "step": 33986 }, { "epoch": 2.6635579937304072, "grad_norm": 3.0954930782318115, "learning_rate": 2.519519874549792e-06, "loss": 1.4802, "step": 33987 }, { "epoch": 2.6636363636363636, "grad_norm": 2.207249164581299, "learning_rate": 2.519252138463297e-06, "loss": 1.0747, "step": 33988 }, { "epoch": 2.66371473354232, "grad_norm": 3.0203847885131836, "learning_rate": 2.518984411812296e-06, "loss": 1.3812, "step": 33989 }, { "epoch": 2.663793103448276, "grad_norm": 5.979050159454346, "learning_rate": 2.5187166945978027e-06, "loss": 1.5205, "step": 33990 }, { "epoch": 2.663871473354232, "grad_norm": 2.7213618755340576, "learning_rate": 2.5184489868208354e-06, "loss": 1.2378, "step": 33991 }, { "epoch": 2.663949843260188, "grad_norm": 3.5357470512390137, "learning_rate": 2.5181812884824142e-06, "loss": 1.4217, "step": 33992 }, { "epoch": 2.6640282131661444, "grad_norm": 3.1633541584014893, "learning_rate": 2.517913599583557e-06, "loss": 0.8782, "step": 33993 }, { "epoch": 2.6641065830721002, "grad_norm": 3.0583438873291016, "learning_rate": 2.5176459201252812e-06, "loss": 1.5697, "step": 33994 }, { "epoch": 2.6641849529780566, "grad_norm": 4.054730415344238, "learning_rate": 2.5173782501086056e-06, "loss": 1.6107, "step": 33995 }, { "epoch": 2.6642633228840125, "grad_norm": 5.290954113006592, "learning_rate": 2.51711058953455e-06, "loss": 1.38, "step": 33996 }, { "epoch": 2.664341692789969, "grad_norm": 3.5749218463897705, "learning_rate": 2.516842938404126e-06, "loss": 0.7301, "step": 33997 }, { "epoch": 2.6644200626959247, "grad_norm": 5.865888595581055, "learning_rate": 2.516575296718359e-06, "loss": 1.2196, "step": 33998 }, { "epoch": 2.664498432601881, "grad_norm": 4.193371772766113, "learning_rate": 2.5163076644782625e-06, "loss": 1.5841, "step": 33999 }, { "epoch": 2.664576802507837, "grad_norm": 3.9268107414245605, "learning_rate": 2.5160400416848583e-06, "loss": 1.1282, "step": 34000 }, { "epoch": 2.6646551724137932, "grad_norm": 2.7091128826141357, "learning_rate": 2.515772428339161e-06, "loss": 1.4507, "step": 34001 }, { "epoch": 2.664733542319749, "grad_norm": 4.037781715393066, "learning_rate": 2.5155048244421864e-06, "loss": 1.2847, "step": 34002 }, { "epoch": 2.6648119122257055, "grad_norm": 3.571176290512085, "learning_rate": 2.515237229994958e-06, "loss": 1.1935, "step": 34003 }, { "epoch": 2.6648902821316613, "grad_norm": 3.5000312328338623, "learning_rate": 2.5149696449984927e-06, "loss": 1.4494, "step": 34004 }, { "epoch": 2.6649686520376177, "grad_norm": 2.629664182662964, "learning_rate": 2.5147020694538043e-06, "loss": 1.3843, "step": 34005 }, { "epoch": 2.6650470219435736, "grad_norm": 3.792071580886841, "learning_rate": 2.514434503361913e-06, "loss": 1.4087, "step": 34006 }, { "epoch": 2.66512539184953, "grad_norm": 6.030118465423584, "learning_rate": 2.514166946723834e-06, "loss": 1.4657, "step": 34007 }, { "epoch": 2.665203761755486, "grad_norm": 3.71142840385437, "learning_rate": 2.5138993995405915e-06, "loss": 1.2805, "step": 34008 }, { "epoch": 2.665282131661442, "grad_norm": 4.835896968841553, "learning_rate": 2.513631861813196e-06, "loss": 1.3364, "step": 34009 }, { "epoch": 2.665360501567398, "grad_norm": 2.4591763019561768, "learning_rate": 2.5133643335426682e-06, "loss": 1.4033, "step": 34010 }, { "epoch": 2.6654388714733543, "grad_norm": 5.498668193817139, "learning_rate": 2.513096814730025e-06, "loss": 1.9354, "step": 34011 }, { "epoch": 2.6655172413793102, "grad_norm": 5.507436752319336, "learning_rate": 2.512829305376284e-06, "loss": 1.769, "step": 34012 }, { "epoch": 2.6655956112852666, "grad_norm": 2.426551580429077, "learning_rate": 2.5125618054824633e-06, "loss": 1.3871, "step": 34013 }, { "epoch": 2.6656739811912225, "grad_norm": 3.5493204593658447, "learning_rate": 2.5122943150495805e-06, "loss": 0.8671, "step": 34014 }, { "epoch": 2.665752351097179, "grad_norm": 3.342060089111328, "learning_rate": 2.5120268340786512e-06, "loss": 1.3014, "step": 34015 }, { "epoch": 2.6658307210031347, "grad_norm": 7.892840385437012, "learning_rate": 2.5117593625706944e-06, "loss": 1.0965, "step": 34016 }, { "epoch": 2.665909090909091, "grad_norm": 6.90916109085083, "learning_rate": 2.511491900526727e-06, "loss": 1.1174, "step": 34017 }, { "epoch": 2.665987460815047, "grad_norm": 4.197018146514893, "learning_rate": 2.511224447947766e-06, "loss": 0.8971, "step": 34018 }, { "epoch": 2.6660658307210032, "grad_norm": 3.284435272216797, "learning_rate": 2.5109570048348307e-06, "loss": 1.3331, "step": 34019 }, { "epoch": 2.666144200626959, "grad_norm": 2.8381412029266357, "learning_rate": 2.5106895711889353e-06, "loss": 1.2831, "step": 34020 }, { "epoch": 2.6662225705329154, "grad_norm": 6.337871074676514, "learning_rate": 2.510422147011096e-06, "loss": 1.1375, "step": 34021 }, { "epoch": 2.6663009404388713, "grad_norm": 2.4040400981903076, "learning_rate": 2.5101547323023334e-06, "loss": 1.2744, "step": 34022 }, { "epoch": 2.6663793103448277, "grad_norm": 3.210096597671509, "learning_rate": 2.5098873270636657e-06, "loss": 1.1837, "step": 34023 }, { "epoch": 2.6664576802507836, "grad_norm": 3.2083873748779297, "learning_rate": 2.5096199312961054e-06, "loss": 1.6733, "step": 34024 }, { "epoch": 2.66653605015674, "grad_norm": 4.547173023223877, "learning_rate": 2.5093525450006723e-06, "loss": 1.6305, "step": 34025 }, { "epoch": 2.666614420062696, "grad_norm": 12.66690731048584, "learning_rate": 2.5090851681783824e-06, "loss": 2.3002, "step": 34026 }, { "epoch": 2.666692789968652, "grad_norm": 8.281572341918945, "learning_rate": 2.5088178008302534e-06, "loss": 1.3943, "step": 34027 }, { "epoch": 2.666771159874608, "grad_norm": 2.429002523422241, "learning_rate": 2.5085504429573014e-06, "loss": 1.447, "step": 34028 }, { "epoch": 2.6668495297805643, "grad_norm": 5.755448341369629, "learning_rate": 2.5082830945605444e-06, "loss": 1.5985, "step": 34029 }, { "epoch": 2.66692789968652, "grad_norm": 2.528968572616577, "learning_rate": 2.508015755640999e-06, "loss": 1.3595, "step": 34030 }, { "epoch": 2.6670062695924766, "grad_norm": 4.824566841125488, "learning_rate": 2.507748426199681e-06, "loss": 1.3897, "step": 34031 }, { "epoch": 2.6670846394984324, "grad_norm": 7.715793132781982, "learning_rate": 2.507481106237608e-06, "loss": 1.4882, "step": 34032 }, { "epoch": 2.6671630094043888, "grad_norm": 7.377780437469482, "learning_rate": 2.507213795755797e-06, "loss": 1.7956, "step": 34033 }, { "epoch": 2.6672413793103447, "grad_norm": 2.856445550918579, "learning_rate": 2.5069464947552637e-06, "loss": 1.3221, "step": 34034 }, { "epoch": 2.667319749216301, "grad_norm": 3.158525228500366, "learning_rate": 2.5066792032370258e-06, "loss": 1.266, "step": 34035 }, { "epoch": 2.6673981191222573, "grad_norm": 3.8214168548583984, "learning_rate": 2.5064119212020997e-06, "loss": 1.387, "step": 34036 }, { "epoch": 2.667476489028213, "grad_norm": 4.366873741149902, "learning_rate": 2.506144648651504e-06, "loss": 1.9885, "step": 34037 }, { "epoch": 2.667554858934169, "grad_norm": 4.013669967651367, "learning_rate": 2.5058773855862506e-06, "loss": 1.6015, "step": 34038 }, { "epoch": 2.6676332288401254, "grad_norm": 14.530887603759766, "learning_rate": 2.5056101320073566e-06, "loss": 0.7797, "step": 34039 }, { "epoch": 2.6677115987460818, "grad_norm": 6.734159469604492, "learning_rate": 2.5053428879158427e-06, "loss": 0.79, "step": 34040 }, { "epoch": 2.6677899686520377, "grad_norm": 3.7883260250091553, "learning_rate": 2.505075653312725e-06, "loss": 1.3671, "step": 34041 }, { "epoch": 2.6678683385579935, "grad_norm": 3.4150333404541016, "learning_rate": 2.504808428199016e-06, "loss": 1.4522, "step": 34042 }, { "epoch": 2.66794670846395, "grad_norm": 2.8372318744659424, "learning_rate": 2.504541212575734e-06, "loss": 1.2868, "step": 34043 }, { "epoch": 2.668025078369906, "grad_norm": 4.417975425720215, "learning_rate": 2.5042740064438942e-06, "loss": 1.1444, "step": 34044 }, { "epoch": 2.668103448275862, "grad_norm": 2.537968397140503, "learning_rate": 2.504006809804518e-06, "loss": 1.3956, "step": 34045 }, { "epoch": 2.668181818181818, "grad_norm": 3.5043749809265137, "learning_rate": 2.5037396226586153e-06, "loss": 0.9363, "step": 34046 }, { "epoch": 2.6682601880877743, "grad_norm": 3.3156468868255615, "learning_rate": 2.503472445007205e-06, "loss": 0.8423, "step": 34047 }, { "epoch": 2.6683385579937307, "grad_norm": 2.441645622253418, "learning_rate": 2.503205276851304e-06, "loss": 1.466, "step": 34048 }, { "epoch": 2.6684169278996865, "grad_norm": 5.126535892486572, "learning_rate": 2.5029381181919264e-06, "loss": 1.6749, "step": 34049 }, { "epoch": 2.6684952978056424, "grad_norm": 4.590693950653076, "learning_rate": 2.502670969030091e-06, "loss": 1.2809, "step": 34050 }, { "epoch": 2.6685736677115988, "grad_norm": 5.710817337036133, "learning_rate": 2.5024038293668116e-06, "loss": 1.8921, "step": 34051 }, { "epoch": 2.668652037617555, "grad_norm": 3.0464086532592773, "learning_rate": 2.502136699203107e-06, "loss": 1.2246, "step": 34052 }, { "epoch": 2.668730407523511, "grad_norm": 6.264540672302246, "learning_rate": 2.5018695785399883e-06, "loss": 1.8488, "step": 34053 }, { "epoch": 2.668808777429467, "grad_norm": 2.7032055854797363, "learning_rate": 2.501602467378476e-06, "loss": 1.3559, "step": 34054 }, { "epoch": 2.668887147335423, "grad_norm": 2.263676643371582, "learning_rate": 2.501335365719585e-06, "loss": 1.2447, "step": 34055 }, { "epoch": 2.6689655172413795, "grad_norm": 2.6708719730377197, "learning_rate": 2.501068273564333e-06, "loss": 1.7376, "step": 34056 }, { "epoch": 2.6690438871473354, "grad_norm": 1.929848074913025, "learning_rate": 2.500801190913731e-06, "loss": 1.406, "step": 34057 }, { "epoch": 2.6691222570532913, "grad_norm": 3.600019931793213, "learning_rate": 2.500534117768796e-06, "loss": 1.1059, "step": 34058 }, { "epoch": 2.6692006269592476, "grad_norm": 2.3242740631103516, "learning_rate": 2.5002670541305474e-06, "loss": 1.1139, "step": 34059 }, { "epoch": 2.669278996865204, "grad_norm": 2.4975805282592773, "learning_rate": 2.5000000000000015e-06, "loss": 1.1591, "step": 34060 }, { "epoch": 2.66935736677116, "grad_norm": 4.4446845054626465, "learning_rate": 2.4997329553781686e-06, "loss": 1.4607, "step": 34061 }, { "epoch": 2.6694357366771158, "grad_norm": 3.051893949508667, "learning_rate": 2.4994659202660683e-06, "loss": 1.3182, "step": 34062 }, { "epoch": 2.669514106583072, "grad_norm": 3.1931674480438232, "learning_rate": 2.4991988946647126e-06, "loss": 1.3224, "step": 34063 }, { "epoch": 2.6695924764890284, "grad_norm": 2.5413262844085693, "learning_rate": 2.4989318785751237e-06, "loss": 1.4734, "step": 34064 }, { "epoch": 2.6696708463949843, "grad_norm": 3.346160411834717, "learning_rate": 2.4986648719983114e-06, "loss": 1.4509, "step": 34065 }, { "epoch": 2.66974921630094, "grad_norm": 3.5731968879699707, "learning_rate": 2.498397874935294e-06, "loss": 1.3288, "step": 34066 }, { "epoch": 2.6698275862068965, "grad_norm": 14.52043628692627, "learning_rate": 2.4981308873870854e-06, "loss": 1.5906, "step": 34067 }, { "epoch": 2.669905956112853, "grad_norm": 8.493159294128418, "learning_rate": 2.497863909354702e-06, "loss": 1.7678, "step": 34068 }, { "epoch": 2.6699843260188088, "grad_norm": 2.7060458660125732, "learning_rate": 2.497596940839159e-06, "loss": 1.377, "step": 34069 }, { "epoch": 2.6700626959247646, "grad_norm": 2.88246488571167, "learning_rate": 2.497329981841472e-06, "loss": 1.5053, "step": 34070 }, { "epoch": 2.670141065830721, "grad_norm": 2.050642728805542, "learning_rate": 2.4970630323626584e-06, "loss": 1.3626, "step": 34071 }, { "epoch": 2.6702194357366773, "grad_norm": 2.643815279006958, "learning_rate": 2.496796092403727e-06, "loss": 1.3548, "step": 34072 }, { "epoch": 2.670297805642633, "grad_norm": 2.6261179447174072, "learning_rate": 2.4965291619657e-06, "loss": 1.2554, "step": 34073 }, { "epoch": 2.670376175548589, "grad_norm": 2.5082292556762695, "learning_rate": 2.4962622410495902e-06, "loss": 1.3092, "step": 34074 }, { "epoch": 2.6704545454545454, "grad_norm": 4.199126720428467, "learning_rate": 2.4959953296564143e-06, "loss": 1.3184, "step": 34075 }, { "epoch": 2.6705329153605017, "grad_norm": 3.4803974628448486, "learning_rate": 2.4957284277871845e-06, "loss": 1.0805, "step": 34076 }, { "epoch": 2.6706112852664576, "grad_norm": 4.401509761810303, "learning_rate": 2.4954615354429156e-06, "loss": 1.1709, "step": 34077 }, { "epoch": 2.6706896551724135, "grad_norm": 16.93181800842285, "learning_rate": 2.4951946526246266e-06, "loss": 1.2174, "step": 34078 }, { "epoch": 2.67076802507837, "grad_norm": 3.891519784927368, "learning_rate": 2.4949277793333326e-06, "loss": 1.3531, "step": 34079 }, { "epoch": 2.670846394984326, "grad_norm": 6.888225078582764, "learning_rate": 2.494660915570044e-06, "loss": 1.5821, "step": 34080 }, { "epoch": 2.670924764890282, "grad_norm": 3.0695345401763916, "learning_rate": 2.4943940613357793e-06, "loss": 0.8607, "step": 34081 }, { "epoch": 2.6710031347962384, "grad_norm": 4.580231189727783, "learning_rate": 2.494127216631552e-06, "loss": 1.3999, "step": 34082 }, { "epoch": 2.6710815047021943, "grad_norm": 10.914743423461914, "learning_rate": 2.493860381458378e-06, "loss": 2.4497, "step": 34083 }, { "epoch": 2.6711598746081506, "grad_norm": 3.1145272254943848, "learning_rate": 2.4935935558172718e-06, "loss": 1.5772, "step": 34084 }, { "epoch": 2.6712382445141065, "grad_norm": 6.67447566986084, "learning_rate": 2.4933267397092487e-06, "loss": 1.2741, "step": 34085 }, { "epoch": 2.671316614420063, "grad_norm": 4.363720417022705, "learning_rate": 2.493059933135323e-06, "loss": 1.0229, "step": 34086 }, { "epoch": 2.6713949843260187, "grad_norm": 4.879657745361328, "learning_rate": 2.492793136096509e-06, "loss": 1.2388, "step": 34087 }, { "epoch": 2.671473354231975, "grad_norm": 3.0749173164367676, "learning_rate": 2.4925263485938228e-06, "loss": 1.4413, "step": 34088 }, { "epoch": 2.671551724137931, "grad_norm": 3.063472032546997, "learning_rate": 2.492259570628278e-06, "loss": 1.3331, "step": 34089 }, { "epoch": 2.6716300940438873, "grad_norm": 4.488216400146484, "learning_rate": 2.491992802200892e-06, "loss": 1.1527, "step": 34090 }, { "epoch": 2.671708463949843, "grad_norm": 3.8308639526367188, "learning_rate": 2.4917260433126732e-06, "loss": 1.4276, "step": 34091 }, { "epoch": 2.6717868338557995, "grad_norm": 4.1653666496276855, "learning_rate": 2.4914592939646424e-06, "loss": 0.8968, "step": 34092 }, { "epoch": 2.6718652037617554, "grad_norm": 3.765641450881958, "learning_rate": 2.4911925541578135e-06, "loss": 1.3704, "step": 34093 }, { "epoch": 2.6719435736677117, "grad_norm": 4.147097587585449, "learning_rate": 2.490925823893198e-06, "loss": 1.5269, "step": 34094 }, { "epoch": 2.6720219435736676, "grad_norm": 2.2010390758514404, "learning_rate": 2.490659103171812e-06, "loss": 1.3698, "step": 34095 }, { "epoch": 2.672100313479624, "grad_norm": 3.3081541061401367, "learning_rate": 2.4903923919946678e-06, "loss": 1.2903, "step": 34096 }, { "epoch": 2.67217868338558, "grad_norm": 2.7219083309173584, "learning_rate": 2.4901256903627854e-06, "loss": 1.3787, "step": 34097 }, { "epoch": 2.672257053291536, "grad_norm": 3.964022159576416, "learning_rate": 2.4898589982771735e-06, "loss": 1.2337, "step": 34098 }, { "epoch": 2.672335423197492, "grad_norm": 2.674544095993042, "learning_rate": 2.4895923157388495e-06, "loss": 1.5421, "step": 34099 }, { "epoch": 2.6724137931034484, "grad_norm": 3.1920111179351807, "learning_rate": 2.4893256427488268e-06, "loss": 1.4659, "step": 34100 }, { "epoch": 2.6724921630094043, "grad_norm": 4.149027347564697, "learning_rate": 2.4890589793081192e-06, "loss": 1.5152, "step": 34101 }, { "epoch": 2.6725705329153606, "grad_norm": 2.795711040496826, "learning_rate": 2.4887923254177425e-06, "loss": 1.5679, "step": 34102 }, { "epoch": 2.6726489028213165, "grad_norm": 2.959818124771118, "learning_rate": 2.4885256810787095e-06, "loss": 1.1348, "step": 34103 }, { "epoch": 2.672727272727273, "grad_norm": 2.6143949031829834, "learning_rate": 2.4882590462920348e-06, "loss": 1.3286, "step": 34104 }, { "epoch": 2.6728056426332287, "grad_norm": 3.9553635120391846, "learning_rate": 2.4879924210587325e-06, "loss": 1.2922, "step": 34105 }, { "epoch": 2.672884012539185, "grad_norm": 2.834800958633423, "learning_rate": 2.4877258053798166e-06, "loss": 1.3262, "step": 34106 }, { "epoch": 2.672962382445141, "grad_norm": 3.2626893520355225, "learning_rate": 2.487459199256302e-06, "loss": 1.0959, "step": 34107 }, { "epoch": 2.6730407523510973, "grad_norm": 3.757845640182495, "learning_rate": 2.487192602689203e-06, "loss": 1.1653, "step": 34108 }, { "epoch": 2.673119122257053, "grad_norm": 4.5329461097717285, "learning_rate": 2.4869260156795316e-06, "loss": 1.1903, "step": 34109 }, { "epoch": 2.6731974921630095, "grad_norm": 2.8442533016204834, "learning_rate": 2.4866594382283007e-06, "loss": 1.3101, "step": 34110 }, { "epoch": 2.6732758620689654, "grad_norm": 2.0065107345581055, "learning_rate": 2.486392870336528e-06, "loss": 1.2651, "step": 34111 }, { "epoch": 2.6733542319749217, "grad_norm": 3.2652125358581543, "learning_rate": 2.4861263120052283e-06, "loss": 1.0573, "step": 34112 }, { "epoch": 2.6734326018808776, "grad_norm": 3.6537833213806152, "learning_rate": 2.4858597632354104e-06, "loss": 1.2186, "step": 34113 }, { "epoch": 2.673510971786834, "grad_norm": 3.0128350257873535, "learning_rate": 2.4855932240280905e-06, "loss": 1.1122, "step": 34114 }, { "epoch": 2.67358934169279, "grad_norm": 3.0708744525909424, "learning_rate": 2.4853266943842813e-06, "loss": 1.2462, "step": 34115 }, { "epoch": 2.673667711598746, "grad_norm": 2.923936367034912, "learning_rate": 2.485060174305001e-06, "loss": 1.0964, "step": 34116 }, { "epoch": 2.673746081504702, "grad_norm": 2.715184450149536, "learning_rate": 2.4847936637912583e-06, "loss": 1.5816, "step": 34117 }, { "epoch": 2.6738244514106584, "grad_norm": 3.0789945125579834, "learning_rate": 2.4845271628440685e-06, "loss": 1.2988, "step": 34118 }, { "epoch": 2.6739028213166143, "grad_norm": 3.2444519996643066, "learning_rate": 2.4842606714644458e-06, "loss": 1.2932, "step": 34119 }, { "epoch": 2.6739811912225706, "grad_norm": 4.6745147705078125, "learning_rate": 2.4839941896534027e-06, "loss": 1.2051, "step": 34120 }, { "epoch": 2.6740595611285265, "grad_norm": 4.420029163360596, "learning_rate": 2.4837277174119533e-06, "loss": 1.4412, "step": 34121 }, { "epoch": 2.674137931034483, "grad_norm": 4.78021240234375, "learning_rate": 2.4834612547411118e-06, "loss": 0.7821, "step": 34122 }, { "epoch": 2.6742163009404387, "grad_norm": 2.7388203144073486, "learning_rate": 2.483194801641891e-06, "loss": 1.4759, "step": 34123 }, { "epoch": 2.674294670846395, "grad_norm": 3.151261806488037, "learning_rate": 2.4829283581153046e-06, "loss": 1.3018, "step": 34124 }, { "epoch": 2.674373040752351, "grad_norm": 2.5795481204986572, "learning_rate": 2.4826619241623655e-06, "loss": 1.3485, "step": 34125 }, { "epoch": 2.6744514106583073, "grad_norm": 5.740023136138916, "learning_rate": 2.482395499784088e-06, "loss": 0.9324, "step": 34126 }, { "epoch": 2.674529780564263, "grad_norm": 3.860952138900757, "learning_rate": 2.4821290849814863e-06, "loss": 1.3391, "step": 34127 }, { "epoch": 2.6746081504702195, "grad_norm": 2.836588144302368, "learning_rate": 2.4818626797555684e-06, "loss": 1.5302, "step": 34128 }, { "epoch": 2.6746865203761754, "grad_norm": 2.3972504138946533, "learning_rate": 2.481596284107354e-06, "loss": 1.0382, "step": 34129 }, { "epoch": 2.6747648902821317, "grad_norm": 3.7976691722869873, "learning_rate": 2.481329898037853e-06, "loss": 1.0685, "step": 34130 }, { "epoch": 2.674843260188088, "grad_norm": 4.857819080352783, "learning_rate": 2.4810635215480815e-06, "loss": 0.9836, "step": 34131 }, { "epoch": 2.674921630094044, "grad_norm": 4.153862476348877, "learning_rate": 2.480797154639049e-06, "loss": 1.3671, "step": 34132 }, { "epoch": 2.675, "grad_norm": 4.564699649810791, "learning_rate": 2.48053079731177e-06, "loss": 1.5362, "step": 34133 }, { "epoch": 2.675078369905956, "grad_norm": 2.35308575630188, "learning_rate": 2.4802644495672557e-06, "loss": 1.3358, "step": 34134 }, { "epoch": 2.6751567398119125, "grad_norm": 3.558570146560669, "learning_rate": 2.4799981114065254e-06, "loss": 1.4722, "step": 34135 }, { "epoch": 2.6752351097178684, "grad_norm": 4.875960350036621, "learning_rate": 2.4797317828305857e-06, "loss": 1.6082, "step": 34136 }, { "epoch": 2.6753134796238243, "grad_norm": 2.4177000522613525, "learning_rate": 2.479465463840452e-06, "loss": 1.3725, "step": 34137 }, { "epoch": 2.6753918495297806, "grad_norm": 2.4079384803771973, "learning_rate": 2.4791991544371368e-06, "loss": 1.3579, "step": 34138 }, { "epoch": 2.675470219435737, "grad_norm": 3.3592333793640137, "learning_rate": 2.4789328546216536e-06, "loss": 1.2876, "step": 34139 }, { "epoch": 2.675548589341693, "grad_norm": 4.371740818023682, "learning_rate": 2.4786665643950146e-06, "loss": 0.9473, "step": 34140 }, { "epoch": 2.6756269592476487, "grad_norm": 3.329575300216675, "learning_rate": 2.478400283758233e-06, "loss": 1.362, "step": 34141 }, { "epoch": 2.675705329153605, "grad_norm": 4.957365989685059, "learning_rate": 2.4781340127123214e-06, "loss": 1.4697, "step": 34142 }, { "epoch": 2.6757836990595614, "grad_norm": 3.0381503105163574, "learning_rate": 2.4778677512582934e-06, "loss": 1.5816, "step": 34143 }, { "epoch": 2.6758620689655173, "grad_norm": 2.7969868183135986, "learning_rate": 2.4776014993971604e-06, "loss": 1.084, "step": 34144 }, { "epoch": 2.675940438871473, "grad_norm": 3.259532928466797, "learning_rate": 2.477335257129936e-06, "loss": 0.8329, "step": 34145 }, { "epoch": 2.6760188087774295, "grad_norm": 3.718581438064575, "learning_rate": 2.4770690244576344e-06, "loss": 1.2786, "step": 34146 }, { "epoch": 2.676097178683386, "grad_norm": 2.7248780727386475, "learning_rate": 2.4768028013812623e-06, "loss": 1.1909, "step": 34147 }, { "epoch": 2.6761755485893417, "grad_norm": 3.189455032348633, "learning_rate": 2.4765365879018383e-06, "loss": 1.2961, "step": 34148 }, { "epoch": 2.6762539184952976, "grad_norm": 2.0144360065460205, "learning_rate": 2.476270384020375e-06, "loss": 1.0246, "step": 34149 }, { "epoch": 2.676332288401254, "grad_norm": 12.004168510437012, "learning_rate": 2.4760041897378813e-06, "loss": 1.9483, "step": 34150 }, { "epoch": 2.6764106583072103, "grad_norm": 2.5497071743011475, "learning_rate": 2.475738005055371e-06, "loss": 1.2575, "step": 34151 }, { "epoch": 2.676489028213166, "grad_norm": 2.250471353530884, "learning_rate": 2.475471829973855e-06, "loss": 1.5648, "step": 34152 }, { "epoch": 2.676567398119122, "grad_norm": 3.9330294132232666, "learning_rate": 2.475205664494351e-06, "loss": 1.6838, "step": 34153 }, { "epoch": 2.6766457680250784, "grad_norm": 2.1250455379486084, "learning_rate": 2.4749395086178665e-06, "loss": 1.5206, "step": 34154 }, { "epoch": 2.6767241379310347, "grad_norm": 6.682737350463867, "learning_rate": 2.4746733623454145e-06, "loss": 0.9954, "step": 34155 }, { "epoch": 2.6768025078369906, "grad_norm": 3.0358667373657227, "learning_rate": 2.474407225678008e-06, "loss": 1.0789, "step": 34156 }, { "epoch": 2.6768808777429465, "grad_norm": 3.489563226699829, "learning_rate": 2.474141098616659e-06, "loss": 1.523, "step": 34157 }, { "epoch": 2.676959247648903, "grad_norm": 3.2498972415924072, "learning_rate": 2.473874981162381e-06, "loss": 1.5152, "step": 34158 }, { "epoch": 2.677037617554859, "grad_norm": 2.1810479164123535, "learning_rate": 2.473608873316184e-06, "loss": 1.2859, "step": 34159 }, { "epoch": 2.677115987460815, "grad_norm": 2.347369432449341, "learning_rate": 2.473342775079082e-06, "loss": 1.365, "step": 34160 }, { "epoch": 2.677194357366771, "grad_norm": 4.494950771331787, "learning_rate": 2.4730766864520856e-06, "loss": 1.8713, "step": 34161 }, { "epoch": 2.6772727272727272, "grad_norm": 5.666050910949707, "learning_rate": 2.472810607436208e-06, "loss": 1.4066, "step": 34162 }, { "epoch": 2.6773510971786836, "grad_norm": 4.158441543579102, "learning_rate": 2.4725445380324605e-06, "loss": 1.4508, "step": 34163 }, { "epoch": 2.6774294670846395, "grad_norm": 7.899157524108887, "learning_rate": 2.4722784782418572e-06, "loss": 0.9955, "step": 34164 }, { "epoch": 2.6775078369905954, "grad_norm": 2.7433507442474365, "learning_rate": 2.4720124280654066e-06, "loss": 1.2888, "step": 34165 }, { "epoch": 2.6775862068965517, "grad_norm": 2.744067668914795, "learning_rate": 2.4717463875041204e-06, "loss": 1.3725, "step": 34166 }, { "epoch": 2.677664576802508, "grad_norm": 5.377747058868408, "learning_rate": 2.471480356559014e-06, "loss": 1.489, "step": 34167 }, { "epoch": 2.677742946708464, "grad_norm": 2.7565691471099854, "learning_rate": 2.4712143352310996e-06, "loss": 1.4242, "step": 34168 }, { "epoch": 2.67782131661442, "grad_norm": 3.0462069511413574, "learning_rate": 2.4709483235213856e-06, "loss": 1.7784, "step": 34169 }, { "epoch": 2.677899686520376, "grad_norm": 7.995050430297852, "learning_rate": 2.4706823214308844e-06, "loss": 1.2172, "step": 34170 }, { "epoch": 2.6779780564263325, "grad_norm": 7.782520294189453, "learning_rate": 2.470416328960607e-06, "loss": 1.1405, "step": 34171 }, { "epoch": 2.6780564263322884, "grad_norm": 4.7973809242248535, "learning_rate": 2.4701503461115705e-06, "loss": 1.2573, "step": 34172 }, { "epoch": 2.6781347962382442, "grad_norm": 8.810785293579102, "learning_rate": 2.469884372884781e-06, "loss": 1.344, "step": 34173 }, { "epoch": 2.6782131661442006, "grad_norm": 3.1381707191467285, "learning_rate": 2.4696184092812516e-06, "loss": 1.3786, "step": 34174 }, { "epoch": 2.678291536050157, "grad_norm": 2.962460994720459, "learning_rate": 2.4693524553019935e-06, "loss": 1.1668, "step": 34175 }, { "epoch": 2.678369905956113, "grad_norm": 7.880548477172852, "learning_rate": 2.4690865109480194e-06, "loss": 0.7324, "step": 34176 }, { "epoch": 2.678448275862069, "grad_norm": 11.983207702636719, "learning_rate": 2.46882057622034e-06, "loss": 0.8922, "step": 34177 }, { "epoch": 2.678526645768025, "grad_norm": 5.2425971031188965, "learning_rate": 2.468554651119967e-06, "loss": 1.1644, "step": 34178 }, { "epoch": 2.6786050156739813, "grad_norm": 2.4951796531677246, "learning_rate": 2.468288735647914e-06, "loss": 1.3602, "step": 34179 }, { "epoch": 2.6786833855799372, "grad_norm": 4.769164562225342, "learning_rate": 2.4680228298051866e-06, "loss": 0.7172, "step": 34180 }, { "epoch": 2.6787617554858936, "grad_norm": 3.68341326713562, "learning_rate": 2.4677569335928014e-06, "loss": 0.8214, "step": 34181 }, { "epoch": 2.6788401253918495, "grad_norm": 5.019166946411133, "learning_rate": 2.4674910470117686e-06, "loss": 1.5712, "step": 34182 }, { "epoch": 2.678918495297806, "grad_norm": 3.3172686100006104, "learning_rate": 2.4672251700631005e-06, "loss": 0.9501, "step": 34183 }, { "epoch": 2.6789968652037617, "grad_norm": 2.6210741996765137, "learning_rate": 2.4669593027478054e-06, "loss": 1.5147, "step": 34184 }, { "epoch": 2.679075235109718, "grad_norm": 3.9019715785980225, "learning_rate": 2.466693445066894e-06, "loss": 1.0226, "step": 34185 }, { "epoch": 2.679153605015674, "grad_norm": 2.6839146614074707, "learning_rate": 2.4664275970213815e-06, "loss": 1.5116, "step": 34186 }, { "epoch": 2.6792319749216302, "grad_norm": 3.1757700443267822, "learning_rate": 2.466161758612279e-06, "loss": 0.7642, "step": 34187 }, { "epoch": 2.679310344827586, "grad_norm": 2.995640993118286, "learning_rate": 2.4658959298405937e-06, "loss": 1.5201, "step": 34188 }, { "epoch": 2.6793887147335425, "grad_norm": 4.963179588317871, "learning_rate": 2.465630110707339e-06, "loss": 1.4932, "step": 34189 }, { "epoch": 2.6794670846394983, "grad_norm": 2.402480125427246, "learning_rate": 2.465364301213523e-06, "loss": 1.5443, "step": 34190 }, { "epoch": 2.6795454545454547, "grad_norm": 2.4430644512176514, "learning_rate": 2.4650985013601642e-06, "loss": 1.4108, "step": 34191 }, { "epoch": 2.6796238244514106, "grad_norm": 4.956031322479248, "learning_rate": 2.464832711148266e-06, "loss": 1.3883, "step": 34192 }, { "epoch": 2.679702194357367, "grad_norm": 4.563451290130615, "learning_rate": 2.464566930578842e-06, "loss": 0.9807, "step": 34193 }, { "epoch": 2.679780564263323, "grad_norm": 4.572667598724365, "learning_rate": 2.464301159652904e-06, "loss": 1.4331, "step": 34194 }, { "epoch": 2.679858934169279, "grad_norm": 2.040431261062622, "learning_rate": 2.464035398371461e-06, "loss": 0.9231, "step": 34195 }, { "epoch": 2.679937304075235, "grad_norm": 23.473831176757812, "learning_rate": 2.4637696467355247e-06, "loss": 0.9491, "step": 34196 }, { "epoch": 2.6800156739811913, "grad_norm": 7.220124244689941, "learning_rate": 2.463503904746107e-06, "loss": 1.3074, "step": 34197 }, { "epoch": 2.6800940438871472, "grad_norm": 2.785425901412964, "learning_rate": 2.463238172404219e-06, "loss": 1.248, "step": 34198 }, { "epoch": 2.6801724137931036, "grad_norm": 3.3357670307159424, "learning_rate": 2.4629724497108655e-06, "loss": 1.0304, "step": 34199 }, { "epoch": 2.6802507836990594, "grad_norm": 3.511578321456909, "learning_rate": 2.4627067366670643e-06, "loss": 1.7604, "step": 34200 }, { "epoch": 2.680329153605016, "grad_norm": 2.748716115951538, "learning_rate": 2.4624410332738237e-06, "loss": 1.3574, "step": 34201 }, { "epoch": 2.6804075235109717, "grad_norm": 3.865657329559326, "learning_rate": 2.462175339532155e-06, "loss": 1.2959, "step": 34202 }, { "epoch": 2.680485893416928, "grad_norm": 2.4729952812194824, "learning_rate": 2.4619096554430665e-06, "loss": 1.2337, "step": 34203 }, { "epoch": 2.680564263322884, "grad_norm": 3.285273313522339, "learning_rate": 2.4616439810075683e-06, "loss": 1.4474, "step": 34204 }, { "epoch": 2.68064263322884, "grad_norm": 3.504873275756836, "learning_rate": 2.461378316226676e-06, "loss": 1.3649, "step": 34205 }, { "epoch": 2.680721003134796, "grad_norm": 4.331967353820801, "learning_rate": 2.461112661101395e-06, "loss": 1.5936, "step": 34206 }, { "epoch": 2.6807993730407524, "grad_norm": 3.0988595485687256, "learning_rate": 2.460847015632738e-06, "loss": 1.2312, "step": 34207 }, { "epoch": 2.6808777429467083, "grad_norm": 9.892797470092773, "learning_rate": 2.4605813798217135e-06, "loss": 1.3555, "step": 34208 }, { "epoch": 2.6809561128526647, "grad_norm": 12.147266387939453, "learning_rate": 2.4603157536693337e-06, "loss": 1.9315, "step": 34209 }, { "epoch": 2.6810344827586206, "grad_norm": 4.5467000007629395, "learning_rate": 2.4600501371766087e-06, "loss": 1.5209, "step": 34210 }, { "epoch": 2.681112852664577, "grad_norm": 2.954745054244995, "learning_rate": 2.459784530344548e-06, "loss": 1.4407, "step": 34211 }, { "epoch": 2.6811912225705328, "grad_norm": 3.201429843902588, "learning_rate": 2.4595189331741626e-06, "loss": 1.3174, "step": 34212 }, { "epoch": 2.681269592476489, "grad_norm": 3.2437198162078857, "learning_rate": 2.459253345666462e-06, "loss": 1.5193, "step": 34213 }, { "epoch": 2.681347962382445, "grad_norm": 3.816539764404297, "learning_rate": 2.4589877678224566e-06, "loss": 1.0469, "step": 34214 }, { "epoch": 2.6814263322884013, "grad_norm": 3.843569040298462, "learning_rate": 2.4587221996431564e-06, "loss": 1.0288, "step": 34215 }, { "epoch": 2.681504702194357, "grad_norm": 2.665196418762207, "learning_rate": 2.458456641129572e-06, "loss": 1.459, "step": 34216 }, { "epoch": 2.6815830721003135, "grad_norm": 6.270131587982178, "learning_rate": 2.4581910922827145e-06, "loss": 1.9385, "step": 34217 }, { "epoch": 2.6816614420062694, "grad_norm": 5.510415077209473, "learning_rate": 2.4579255531035896e-06, "loss": 0.8938, "step": 34218 }, { "epoch": 2.6817398119122258, "grad_norm": 3.492672920227051, "learning_rate": 2.457660023593211e-06, "loss": 1.6147, "step": 34219 }, { "epoch": 2.6818181818181817, "grad_norm": 1.9782965183258057, "learning_rate": 2.457394503752591e-06, "loss": 1.0414, "step": 34220 }, { "epoch": 2.681896551724138, "grad_norm": 3.076880931854248, "learning_rate": 2.4571289935827334e-06, "loss": 1.4833, "step": 34221 }, { "epoch": 2.681974921630094, "grad_norm": 3.918246269226074, "learning_rate": 2.456863493084652e-06, "loss": 1.6369, "step": 34222 }, { "epoch": 2.68205329153605, "grad_norm": 2.9524662494659424, "learning_rate": 2.4565980022593527e-06, "loss": 1.6697, "step": 34223 }, { "epoch": 2.682131661442006, "grad_norm": 3.694661855697632, "learning_rate": 2.4563325211078525e-06, "loss": 1.419, "step": 34224 }, { "epoch": 2.6822100313479624, "grad_norm": 3.728193998336792, "learning_rate": 2.456067049631155e-06, "loss": 1.3637, "step": 34225 }, { "epoch": 2.6822884012539188, "grad_norm": 3.314683437347412, "learning_rate": 2.4558015878302716e-06, "loss": 1.1334, "step": 34226 }, { "epoch": 2.6823667711598747, "grad_norm": 6.808771133422852, "learning_rate": 2.4555361357062124e-06, "loss": 1.4952, "step": 34227 }, { "epoch": 2.6824451410658305, "grad_norm": 4.325192451477051, "learning_rate": 2.455270693259987e-06, "loss": 1.2311, "step": 34228 }, { "epoch": 2.682523510971787, "grad_norm": 9.887404441833496, "learning_rate": 2.4550052604926045e-06, "loss": 1.404, "step": 34229 }, { "epoch": 2.682601880877743, "grad_norm": 2.4326255321502686, "learning_rate": 2.4547398374050744e-06, "loss": 0.9526, "step": 34230 }, { "epoch": 2.682680250783699, "grad_norm": 3.0881080627441406, "learning_rate": 2.4544744239984075e-06, "loss": 1.3602, "step": 34231 }, { "epoch": 2.682758620689655, "grad_norm": 3.3334414958953857, "learning_rate": 2.454209020273612e-06, "loss": 1.575, "step": 34232 }, { "epoch": 2.6828369905956113, "grad_norm": 2.902592897415161, "learning_rate": 2.453943626231698e-06, "loss": 1.3024, "step": 34233 }, { "epoch": 2.6829153605015676, "grad_norm": 2.40559983253479, "learning_rate": 2.4536782418736743e-06, "loss": 1.2118, "step": 34234 }, { "epoch": 2.6829937304075235, "grad_norm": 4.268158435821533, "learning_rate": 2.453412867200553e-06, "loss": 1.1616, "step": 34235 }, { "epoch": 2.6830721003134794, "grad_norm": 2.346895217895508, "learning_rate": 2.453147502213337e-06, "loss": 1.0928, "step": 34236 }, { "epoch": 2.6831504702194358, "grad_norm": 4.735517501831055, "learning_rate": 2.452882146913042e-06, "loss": 1.3458, "step": 34237 }, { "epoch": 2.683228840125392, "grad_norm": 12.013557434082031, "learning_rate": 2.4526168013006746e-06, "loss": 1.6142, "step": 34238 }, { "epoch": 2.683307210031348, "grad_norm": 6.045409202575684, "learning_rate": 2.452351465377247e-06, "loss": 1.4088, "step": 34239 }, { "epoch": 2.683385579937304, "grad_norm": 3.260255813598633, "learning_rate": 2.4520861391437635e-06, "loss": 1.4725, "step": 34240 }, { "epoch": 2.68346394984326, "grad_norm": 3.395606279373169, "learning_rate": 2.4518208226012336e-06, "loss": 1.4668, "step": 34241 }, { "epoch": 2.6835423197492165, "grad_norm": 16.068584442138672, "learning_rate": 2.4515555157506703e-06, "loss": 1.7371, "step": 34242 }, { "epoch": 2.6836206896551724, "grad_norm": 22.56771469116211, "learning_rate": 2.4512902185930836e-06, "loss": 1.0305, "step": 34243 }, { "epoch": 2.6836990595611283, "grad_norm": 5.169393539428711, "learning_rate": 2.451024931129477e-06, "loss": 1.2208, "step": 34244 }, { "epoch": 2.6837774294670846, "grad_norm": 3.370004415512085, "learning_rate": 2.450759653360863e-06, "loss": 1.4778, "step": 34245 }, { "epoch": 2.683855799373041, "grad_norm": 2.3269121646881104, "learning_rate": 2.4504943852882495e-06, "loss": 0.9881, "step": 34246 }, { "epoch": 2.683934169278997, "grad_norm": 2.8730342388153076, "learning_rate": 2.450229126912646e-06, "loss": 0.9383, "step": 34247 }, { "epoch": 2.6840125391849528, "grad_norm": 4.669799327850342, "learning_rate": 2.449963878235061e-06, "loss": 1.2676, "step": 34248 }, { "epoch": 2.684090909090909, "grad_norm": 3.4495224952697754, "learning_rate": 2.4496986392565035e-06, "loss": 1.209, "step": 34249 }, { "epoch": 2.6841692789968654, "grad_norm": 4.289383888244629, "learning_rate": 2.449433409977983e-06, "loss": 1.3752, "step": 34250 }, { "epoch": 2.6842476489028213, "grad_norm": 1.9584795236587524, "learning_rate": 2.449168190400508e-06, "loss": 1.2936, "step": 34251 }, { "epoch": 2.684326018808777, "grad_norm": 3.8117010593414307, "learning_rate": 2.4489029805250865e-06, "loss": 1.1907, "step": 34252 }, { "epoch": 2.6844043887147335, "grad_norm": 2.8340036869049072, "learning_rate": 2.4486377803527274e-06, "loss": 1.3614, "step": 34253 }, { "epoch": 2.68448275862069, "grad_norm": 3.8958468437194824, "learning_rate": 2.4483725898844423e-06, "loss": 1.0483, "step": 34254 }, { "epoch": 2.6845611285266457, "grad_norm": 2.721358299255371, "learning_rate": 2.4481074091212333e-06, "loss": 1.3145, "step": 34255 }, { "epoch": 2.6846394984326016, "grad_norm": 5.790725231170654, "learning_rate": 2.4478422380641143e-06, "loss": 1.2805, "step": 34256 }, { "epoch": 2.684717868338558, "grad_norm": 1.9232827425003052, "learning_rate": 2.447577076714093e-06, "loss": 1.3396, "step": 34257 }, { "epoch": 2.6847962382445143, "grad_norm": 3.46425461769104, "learning_rate": 2.4473119250721793e-06, "loss": 1.4835, "step": 34258 }, { "epoch": 2.68487460815047, "grad_norm": 8.887311935424805, "learning_rate": 2.4470467831393784e-06, "loss": 0.9562, "step": 34259 }, { "epoch": 2.684952978056426, "grad_norm": 7.616800308227539, "learning_rate": 2.446781650916698e-06, "loss": 1.2564, "step": 34260 }, { "epoch": 2.6850313479623824, "grad_norm": 14.877036094665527, "learning_rate": 2.446516528405153e-06, "loss": 1.8561, "step": 34261 }, { "epoch": 2.6851097178683387, "grad_norm": 5.493857383728027, "learning_rate": 2.446251415605745e-06, "loss": 1.2034, "step": 34262 }, { "epoch": 2.6851880877742946, "grad_norm": 4.091525077819824, "learning_rate": 2.4459863125194847e-06, "loss": 1.3635, "step": 34263 }, { "epoch": 2.6852664576802505, "grad_norm": 3.2786707878112793, "learning_rate": 2.4457212191473818e-06, "loss": 0.8884, "step": 34264 }, { "epoch": 2.685344827586207, "grad_norm": 2.735278367996216, "learning_rate": 2.4454561354904426e-06, "loss": 1.3032, "step": 34265 }, { "epoch": 2.685423197492163, "grad_norm": 3.8575823307037354, "learning_rate": 2.4451910615496765e-06, "loss": 1.3851, "step": 34266 }, { "epoch": 2.685501567398119, "grad_norm": 2.780513286590576, "learning_rate": 2.4449259973260914e-06, "loss": 1.1022, "step": 34267 }, { "epoch": 2.685579937304075, "grad_norm": 2.1933882236480713, "learning_rate": 2.4446609428206954e-06, "loss": 1.1437, "step": 34268 }, { "epoch": 2.6856583072100313, "grad_norm": 2.1592843532562256, "learning_rate": 2.4443958980344963e-06, "loss": 1.2745, "step": 34269 }, { "epoch": 2.6857366771159876, "grad_norm": 3.2578885555267334, "learning_rate": 2.444130862968503e-06, "loss": 1.3234, "step": 34270 }, { "epoch": 2.6858150470219435, "grad_norm": 4.172233581542969, "learning_rate": 2.4438658376237225e-06, "loss": 1.1486, "step": 34271 }, { "epoch": 2.6858934169279, "grad_norm": 3.7390451431274414, "learning_rate": 2.443600822001164e-06, "loss": 1.2692, "step": 34272 }, { "epoch": 2.6859717868338557, "grad_norm": 2.545114278793335, "learning_rate": 2.4433358161018365e-06, "loss": 1.1529, "step": 34273 }, { "epoch": 2.686050156739812, "grad_norm": 4.066168308258057, "learning_rate": 2.4430708199267426e-06, "loss": 1.2349, "step": 34274 }, { "epoch": 2.686128526645768, "grad_norm": 3.0584867000579834, "learning_rate": 2.442805833476896e-06, "loss": 1.2403, "step": 34275 }, { "epoch": 2.6862068965517243, "grad_norm": 8.266491889953613, "learning_rate": 2.442540856753305e-06, "loss": 1.0006, "step": 34276 }, { "epoch": 2.68628526645768, "grad_norm": 9.181421279907227, "learning_rate": 2.4422758897569726e-06, "loss": 1.5551, "step": 34277 }, { "epoch": 2.6863636363636365, "grad_norm": 3.6467409133911133, "learning_rate": 2.4420109324889086e-06, "loss": 1.3447, "step": 34278 }, { "epoch": 2.6864420062695924, "grad_norm": 3.4182024002075195, "learning_rate": 2.4417459849501196e-06, "loss": 0.9418, "step": 34279 }, { "epoch": 2.6865203761755487, "grad_norm": 3.815110445022583, "learning_rate": 2.441481047141619e-06, "loss": 1.3424, "step": 34280 }, { "epoch": 2.6865987460815046, "grad_norm": 3.1454896926879883, "learning_rate": 2.441216119064408e-06, "loss": 1.1314, "step": 34281 }, { "epoch": 2.686677115987461, "grad_norm": 6.070245742797852, "learning_rate": 2.440951200719497e-06, "loss": 1.2615, "step": 34282 }, { "epoch": 2.686755485893417, "grad_norm": 6.91409969329834, "learning_rate": 2.440686292107893e-06, "loss": 1.3905, "step": 34283 }, { "epoch": 2.686833855799373, "grad_norm": 2.5590009689331055, "learning_rate": 2.440421393230604e-06, "loss": 1.0676, "step": 34284 }, { "epoch": 2.686912225705329, "grad_norm": 2.879166603088379, "learning_rate": 2.4401565040886366e-06, "loss": 1.6355, "step": 34285 }, { "epoch": 2.6869905956112854, "grad_norm": 3.286940574645996, "learning_rate": 2.4398916246829998e-06, "loss": 1.5785, "step": 34286 }, { "epoch": 2.6870689655172413, "grad_norm": 6.316096782684326, "learning_rate": 2.4396267550146995e-06, "loss": 1.1417, "step": 34287 }, { "epoch": 2.6871473354231976, "grad_norm": 3.446685791015625, "learning_rate": 2.4393618950847446e-06, "loss": 1.5358, "step": 34288 }, { "epoch": 2.6872257053291535, "grad_norm": 4.016263484954834, "learning_rate": 2.439097044894142e-06, "loss": 1.4504, "step": 34289 }, { "epoch": 2.68730407523511, "grad_norm": 6.878602981567383, "learning_rate": 2.4388322044438984e-06, "loss": 1.3734, "step": 34290 }, { "epoch": 2.6873824451410657, "grad_norm": 2.8992867469787598, "learning_rate": 2.4385673737350234e-06, "loss": 1.0976, "step": 34291 }, { "epoch": 2.687460815047022, "grad_norm": 4.455820560455322, "learning_rate": 2.438302552768521e-06, "loss": 1.4121, "step": 34292 }, { "epoch": 2.687539184952978, "grad_norm": 2.976963520050049, "learning_rate": 2.4380377415453984e-06, "loss": 1.5168, "step": 34293 }, { "epoch": 2.6876175548589343, "grad_norm": 2.7512645721435547, "learning_rate": 2.437772940066665e-06, "loss": 1.419, "step": 34294 }, { "epoch": 2.68769592476489, "grad_norm": 4.007411003112793, "learning_rate": 2.43750814833333e-06, "loss": 1.3026, "step": 34295 }, { "epoch": 2.6877742946708465, "grad_norm": 3.1763856410980225, "learning_rate": 2.4372433663463955e-06, "loss": 1.4913, "step": 34296 }, { "epoch": 2.6878526645768024, "grad_norm": 3.2195167541503906, "learning_rate": 2.436978594106872e-06, "loss": 1.5743, "step": 34297 }, { "epoch": 2.6879310344827587, "grad_norm": 3.0854651927948, "learning_rate": 2.4367138316157622e-06, "loss": 1.4036, "step": 34298 }, { "epoch": 2.6880094043887146, "grad_norm": 2.6005749702453613, "learning_rate": 2.4364490788740813e-06, "loss": 1.259, "step": 34299 }, { "epoch": 2.688087774294671, "grad_norm": 2.9042017459869385, "learning_rate": 2.4361843358828287e-06, "loss": 1.4507, "step": 34300 }, { "epoch": 2.688166144200627, "grad_norm": 3.095479726791382, "learning_rate": 2.4359196026430145e-06, "loss": 1.0733, "step": 34301 }, { "epoch": 2.688244514106583, "grad_norm": 2.762052536010742, "learning_rate": 2.435654879155645e-06, "loss": 1.2687, "step": 34302 }, { "epoch": 2.688322884012539, "grad_norm": 2.1155731678009033, "learning_rate": 2.435390165421727e-06, "loss": 1.2096, "step": 34303 }, { "epoch": 2.6884012539184954, "grad_norm": 3.87956166267395, "learning_rate": 2.4351254614422674e-06, "loss": 1.5379, "step": 34304 }, { "epoch": 2.6884796238244513, "grad_norm": 4.410641670227051, "learning_rate": 2.434860767218274e-06, "loss": 1.2955, "step": 34305 }, { "epoch": 2.6885579937304076, "grad_norm": 3.393902540206909, "learning_rate": 2.4345960827507536e-06, "loss": 1.2215, "step": 34306 }, { "epoch": 2.6886363636363635, "grad_norm": 2.7953035831451416, "learning_rate": 2.434331408040709e-06, "loss": 1.3674, "step": 34307 }, { "epoch": 2.68871473354232, "grad_norm": 2.805445909500122, "learning_rate": 2.4340667430891514e-06, "loss": 1.2201, "step": 34308 }, { "epoch": 2.6887931034482757, "grad_norm": 6.751040935516357, "learning_rate": 2.4338020878970858e-06, "loss": 1.4908, "step": 34309 }, { "epoch": 2.688871473354232, "grad_norm": 3.984492063522339, "learning_rate": 2.433537442465521e-06, "loss": 1.5636, "step": 34310 }, { "epoch": 2.688949843260188, "grad_norm": 2.381206512451172, "learning_rate": 2.4332728067954604e-06, "loss": 1.1066, "step": 34311 }, { "epoch": 2.6890282131661443, "grad_norm": 3.492288827896118, "learning_rate": 2.4330081808879085e-06, "loss": 1.4262, "step": 34312 }, { "epoch": 2.6891065830721, "grad_norm": 4.057043552398682, "learning_rate": 2.432743564743878e-06, "loss": 2.05, "step": 34313 }, { "epoch": 2.6891849529780565, "grad_norm": 4.1294026374816895, "learning_rate": 2.4324789583643738e-06, "loss": 1.5237, "step": 34314 }, { "epoch": 2.6892633228840124, "grad_norm": 7.1002349853515625, "learning_rate": 2.432214361750399e-06, "loss": 1.0727, "step": 34315 }, { "epoch": 2.6893416927899687, "grad_norm": 2.272266387939453, "learning_rate": 2.431949774902962e-06, "loss": 1.3399, "step": 34316 }, { "epoch": 2.689420062695925, "grad_norm": 2.936406135559082, "learning_rate": 2.4316851978230686e-06, "loss": 1.4403, "step": 34317 }, { "epoch": 2.689498432601881, "grad_norm": 3.4145820140838623, "learning_rate": 2.431420630511726e-06, "loss": 1.3723, "step": 34318 }, { "epoch": 2.689576802507837, "grad_norm": 4.583069801330566, "learning_rate": 2.4311560729699396e-06, "loss": 1.9003, "step": 34319 }, { "epoch": 2.689655172413793, "grad_norm": 8.894065856933594, "learning_rate": 2.4308915251987164e-06, "loss": 1.1191, "step": 34320 }, { "epoch": 2.6897335423197495, "grad_norm": 3.7076759338378906, "learning_rate": 2.4306269871990623e-06, "loss": 1.3439, "step": 34321 }, { "epoch": 2.6898119122257054, "grad_norm": 3.1324379444122314, "learning_rate": 2.430362458971983e-06, "loss": 1.2272, "step": 34322 }, { "epoch": 2.6898902821316613, "grad_norm": 3.822876453399658, "learning_rate": 2.4300979405184854e-06, "loss": 1.3825, "step": 34323 }, { "epoch": 2.6899686520376176, "grad_norm": 3.428267240524292, "learning_rate": 2.429833431839575e-06, "loss": 1.1646, "step": 34324 }, { "epoch": 2.690047021943574, "grad_norm": 2.3240303993225098, "learning_rate": 2.429568932936259e-06, "loss": 1.3141, "step": 34325 }, { "epoch": 2.69012539184953, "grad_norm": 2.706777334213257, "learning_rate": 2.4293044438095415e-06, "loss": 1.2348, "step": 34326 }, { "epoch": 2.6902037617554857, "grad_norm": 4.160970687866211, "learning_rate": 2.4290399644604294e-06, "loss": 1.3588, "step": 34327 }, { "epoch": 2.690282131661442, "grad_norm": 2.643679141998291, "learning_rate": 2.4287754948899294e-06, "loss": 1.4699, "step": 34328 }, { "epoch": 2.6903605015673984, "grad_norm": 3.7666375637054443, "learning_rate": 2.4285110350990482e-06, "loss": 1.4066, "step": 34329 }, { "epoch": 2.6904388714733543, "grad_norm": 3.3042049407958984, "learning_rate": 2.4282465850887887e-06, "loss": 1.567, "step": 34330 }, { "epoch": 2.69051724137931, "grad_norm": 3.2478034496307373, "learning_rate": 2.4279821448601564e-06, "loss": 1.6455, "step": 34331 }, { "epoch": 2.6905956112852665, "grad_norm": 2.468846559524536, "learning_rate": 2.4277177144141624e-06, "loss": 1.5424, "step": 34332 }, { "epoch": 2.690673981191223, "grad_norm": 4.741502285003662, "learning_rate": 2.427453293751807e-06, "loss": 1.3528, "step": 34333 }, { "epoch": 2.6907523510971787, "grad_norm": 5.5314621925354, "learning_rate": 2.4271888828740985e-06, "loss": 2.0263, "step": 34334 }, { "epoch": 2.6908307210031346, "grad_norm": 2.9679131507873535, "learning_rate": 2.4269244817820416e-06, "loss": 1.4691, "step": 34335 }, { "epoch": 2.690909090909091, "grad_norm": 5.494709014892578, "learning_rate": 2.4266600904766422e-06, "loss": 0.9276, "step": 34336 }, { "epoch": 2.6909874608150472, "grad_norm": 3.077315330505371, "learning_rate": 2.4263957089589064e-06, "loss": 0.9776, "step": 34337 }, { "epoch": 2.691065830721003, "grad_norm": 5.4959869384765625, "learning_rate": 2.4261313372298394e-06, "loss": 1.2204, "step": 34338 }, { "epoch": 2.691144200626959, "grad_norm": 4.601139068603516, "learning_rate": 2.425866975290447e-06, "loss": 1.4687, "step": 34339 }, { "epoch": 2.6912225705329154, "grad_norm": 2.626305103302002, "learning_rate": 2.425602623141734e-06, "loss": 1.2792, "step": 34340 }, { "epoch": 2.6913009404388717, "grad_norm": 2.634427070617676, "learning_rate": 2.4253382807847064e-06, "loss": 1.2963, "step": 34341 }, { "epoch": 2.6913793103448276, "grad_norm": 3.0408201217651367, "learning_rate": 2.4250739482203694e-06, "loss": 1.3541, "step": 34342 }, { "epoch": 2.6914576802507835, "grad_norm": 4.262429237365723, "learning_rate": 2.424809625449729e-06, "loss": 1.2204, "step": 34343 }, { "epoch": 2.69153605015674, "grad_norm": 3.1105167865753174, "learning_rate": 2.42454531247379e-06, "loss": 1.1787, "step": 34344 }, { "epoch": 2.691614420062696, "grad_norm": 4.577517032623291, "learning_rate": 2.4242810092935578e-06, "loss": 1.351, "step": 34345 }, { "epoch": 2.691692789968652, "grad_norm": 8.088151931762695, "learning_rate": 2.4240167159100376e-06, "loss": 1.4853, "step": 34346 }, { "epoch": 2.691771159874608, "grad_norm": 3.307783365249634, "learning_rate": 2.4237524323242364e-06, "loss": 1.0956, "step": 34347 }, { "epoch": 2.6918495297805642, "grad_norm": 3.795283794403076, "learning_rate": 2.4234881585371555e-06, "loss": 1.5413, "step": 34348 }, { "epoch": 2.6919278996865206, "grad_norm": 4.2360334396362305, "learning_rate": 2.423223894549801e-06, "loss": 1.334, "step": 34349 }, { "epoch": 2.6920062695924765, "grad_norm": 2.7512600421905518, "learning_rate": 2.422959640363181e-06, "loss": 1.2789, "step": 34350 }, { "epoch": 2.6920846394984324, "grad_norm": 2.1321804523468018, "learning_rate": 2.422695395978301e-06, "loss": 1.2712, "step": 34351 }, { "epoch": 2.6921630094043887, "grad_norm": 3.523186683654785, "learning_rate": 2.4224311613961617e-06, "loss": 1.4942, "step": 34352 }, { "epoch": 2.692241379310345, "grad_norm": 3.6527881622314453, "learning_rate": 2.42216693661777e-06, "loss": 1.2965, "step": 34353 }, { "epoch": 2.692319749216301, "grad_norm": 2.5686352252960205, "learning_rate": 2.4219027216441305e-06, "loss": 1.3538, "step": 34354 }, { "epoch": 2.692398119122257, "grad_norm": 2.795480728149414, "learning_rate": 2.4216385164762517e-06, "loss": 1.3066, "step": 34355 }, { "epoch": 2.692476489028213, "grad_norm": 2.3130908012390137, "learning_rate": 2.421374321115134e-06, "loss": 1.4246, "step": 34356 }, { "epoch": 2.6925548589341695, "grad_norm": 3.1830031871795654, "learning_rate": 2.421110135561784e-06, "loss": 1.3117, "step": 34357 }, { "epoch": 2.6926332288401253, "grad_norm": 2.5399811267852783, "learning_rate": 2.420845959817206e-06, "loss": 1.1873, "step": 34358 }, { "epoch": 2.6927115987460812, "grad_norm": 3.3018457889556885, "learning_rate": 2.4205817938824056e-06, "loss": 1.0634, "step": 34359 }, { "epoch": 2.6927899686520376, "grad_norm": 3.269428014755249, "learning_rate": 2.420317637758387e-06, "loss": 1.2514, "step": 34360 }, { "epoch": 2.692868338557994, "grad_norm": 4.890933513641357, "learning_rate": 2.4200534914461554e-06, "loss": 1.1428, "step": 34361 }, { "epoch": 2.69294670846395, "grad_norm": 2.608321189880371, "learning_rate": 2.4197893549467167e-06, "loss": 1.4226, "step": 34362 }, { "epoch": 2.693025078369906, "grad_norm": 3.7773005962371826, "learning_rate": 2.4195252282610703e-06, "loss": 1.6605, "step": 34363 }, { "epoch": 2.693103448275862, "grad_norm": 2.430762767791748, "learning_rate": 2.419261111390227e-06, "loss": 1.1708, "step": 34364 }, { "epoch": 2.6931818181818183, "grad_norm": 2.620009183883667, "learning_rate": 2.4189970043351885e-06, "loss": 1.7038, "step": 34365 }, { "epoch": 2.6932601880877742, "grad_norm": 4.824291229248047, "learning_rate": 2.418732907096961e-06, "loss": 1.1565, "step": 34366 }, { "epoch": 2.6933385579937306, "grad_norm": 6.8841872215271, "learning_rate": 2.4184688196765467e-06, "loss": 1.2501, "step": 34367 }, { "epoch": 2.6934169278996865, "grad_norm": 3.7418439388275146, "learning_rate": 2.418204742074949e-06, "loss": 0.3156, "step": 34368 }, { "epoch": 2.693495297805643, "grad_norm": 4.371028900146484, "learning_rate": 2.4179406742931756e-06, "loss": 1.0523, "step": 34369 }, { "epoch": 2.6935736677115987, "grad_norm": 4.173037052154541, "learning_rate": 2.417676616332232e-06, "loss": 1.3823, "step": 34370 }, { "epoch": 2.693652037617555, "grad_norm": 6.008193492889404, "learning_rate": 2.417412568193118e-06, "loss": 1.2941, "step": 34371 }, { "epoch": 2.693730407523511, "grad_norm": 2.3764963150024414, "learning_rate": 2.417148529876841e-06, "loss": 1.165, "step": 34372 }, { "epoch": 2.6938087774294672, "grad_norm": 3.0204367637634277, "learning_rate": 2.416884501384403e-06, "loss": 1.5256, "step": 34373 }, { "epoch": 2.693887147335423, "grad_norm": 3.1451165676116943, "learning_rate": 2.416620482716811e-06, "loss": 1.364, "step": 34374 }, { "epoch": 2.6939655172413794, "grad_norm": 4.541666507720947, "learning_rate": 2.4163564738750666e-06, "loss": 1.2671, "step": 34375 }, { "epoch": 2.6940438871473353, "grad_norm": 3.110504150390625, "learning_rate": 2.416092474860176e-06, "loss": 1.1835, "step": 34376 }, { "epoch": 2.6941222570532917, "grad_norm": 3.760451555252075, "learning_rate": 2.415828485673142e-06, "loss": 1.539, "step": 34377 }, { "epoch": 2.6942006269592476, "grad_norm": 7.825617790222168, "learning_rate": 2.415564506314969e-06, "loss": 1.1531, "step": 34378 }, { "epoch": 2.694278996865204, "grad_norm": 3.3249504566192627, "learning_rate": 2.4153005367866618e-06, "loss": 1.1477, "step": 34379 }, { "epoch": 2.69435736677116, "grad_norm": 2.674011468887329, "learning_rate": 2.4150365770892233e-06, "loss": 1.3992, "step": 34380 }, { "epoch": 2.694435736677116, "grad_norm": 2.423396110534668, "learning_rate": 2.4147726272236595e-06, "loss": 1.4636, "step": 34381 }, { "epoch": 2.694514106583072, "grad_norm": 5.956665992736816, "learning_rate": 2.4145086871909697e-06, "loss": 1.3091, "step": 34382 }, { "epoch": 2.6945924764890283, "grad_norm": 2.4031167030334473, "learning_rate": 2.414244756992162e-06, "loss": 1.4284, "step": 34383 }, { "epoch": 2.694670846394984, "grad_norm": 2.9305050373077393, "learning_rate": 2.4139808366282397e-06, "loss": 0.9999, "step": 34384 }, { "epoch": 2.6947492163009406, "grad_norm": 3.0742385387420654, "learning_rate": 2.4137169261002074e-06, "loss": 1.5906, "step": 34385 }, { "epoch": 2.6948275862068964, "grad_norm": 4.931591987609863, "learning_rate": 2.413453025409066e-06, "loss": 0.8157, "step": 34386 }, { "epoch": 2.6949059561128528, "grad_norm": 4.7662034034729, "learning_rate": 2.413189134555819e-06, "loss": 1.1295, "step": 34387 }, { "epoch": 2.6949843260188087, "grad_norm": 2.3247921466827393, "learning_rate": 2.4129252535414754e-06, "loss": 1.134, "step": 34388 }, { "epoch": 2.695062695924765, "grad_norm": 4.703152179718018, "learning_rate": 2.4126613823670334e-06, "loss": 1.6219, "step": 34389 }, { "epoch": 2.695141065830721, "grad_norm": 3.1342241764068604, "learning_rate": 2.4123975210334987e-06, "loss": 1.4502, "step": 34390 }, { "epoch": 2.695219435736677, "grad_norm": 2.5501832962036133, "learning_rate": 2.412133669541875e-06, "loss": 1.3621, "step": 34391 }, { "epoch": 2.695297805642633, "grad_norm": 3.7028465270996094, "learning_rate": 2.411869827893165e-06, "loss": 1.0724, "step": 34392 }, { "epoch": 2.6953761755485894, "grad_norm": 4.685199737548828, "learning_rate": 2.4116059960883736e-06, "loss": 1.2584, "step": 34393 }, { "epoch": 2.6954545454545453, "grad_norm": 5.736542701721191, "learning_rate": 2.411342174128503e-06, "loss": 1.2129, "step": 34394 }, { "epoch": 2.6955329153605017, "grad_norm": 3.416002035140991, "learning_rate": 2.4110783620145573e-06, "loss": 1.6656, "step": 34395 }, { "epoch": 2.6956112852664575, "grad_norm": 6.859062671661377, "learning_rate": 2.4108145597475394e-06, "loss": 1.0694, "step": 34396 }, { "epoch": 2.695689655172414, "grad_norm": 3.108062267303467, "learning_rate": 2.410550767328453e-06, "loss": 0.9794, "step": 34397 }, { "epoch": 2.6957680250783698, "grad_norm": 3.311812400817871, "learning_rate": 2.4102869847583023e-06, "loss": 1.7473, "step": 34398 }, { "epoch": 2.695846394984326, "grad_norm": 2.8184502124786377, "learning_rate": 2.410023212038089e-06, "loss": 1.5465, "step": 34399 }, { "epoch": 2.695924764890282, "grad_norm": 3.693986415863037, "learning_rate": 2.409759449168819e-06, "loss": 1.2174, "step": 34400 }, { "epoch": 2.6960031347962383, "grad_norm": 2.1973564624786377, "learning_rate": 2.40949569615149e-06, "loss": 1.3139, "step": 34401 }, { "epoch": 2.696081504702194, "grad_norm": 2.802405595779419, "learning_rate": 2.409231952987111e-06, "loss": 1.1945, "step": 34402 }, { "epoch": 2.6961598746081505, "grad_norm": 2.402912139892578, "learning_rate": 2.4089682196766846e-06, "loss": 1.4523, "step": 34403 }, { "epoch": 2.6962382445141064, "grad_norm": 4.261831760406494, "learning_rate": 2.4087044962212104e-06, "loss": 1.3586, "step": 34404 }, { "epoch": 2.6963166144200628, "grad_norm": 2.6946232318878174, "learning_rate": 2.4084407826216934e-06, "loss": 1.7014, "step": 34405 }, { "epoch": 2.6963949843260187, "grad_norm": 5.718087196350098, "learning_rate": 2.4081770788791344e-06, "loss": 1.1868, "step": 34406 }, { "epoch": 2.696473354231975, "grad_norm": 3.4548232555389404, "learning_rate": 2.4079133849945426e-06, "loss": 1.3197, "step": 34407 }, { "epoch": 2.696551724137931, "grad_norm": 2.339644432067871, "learning_rate": 2.407649700968915e-06, "loss": 1.4617, "step": 34408 }, { "epoch": 2.696630094043887, "grad_norm": 2.806252956390381, "learning_rate": 2.407386026803256e-06, "loss": 1.1281, "step": 34409 }, { "epoch": 2.696708463949843, "grad_norm": 4.239439487457275, "learning_rate": 2.407122362498569e-06, "loss": 1.2389, "step": 34410 }, { "epoch": 2.6967868338557994, "grad_norm": 5.413064956665039, "learning_rate": 2.406858708055857e-06, "loss": 1.6383, "step": 34411 }, { "epoch": 2.6968652037617558, "grad_norm": 3.895975112915039, "learning_rate": 2.4065950634761227e-06, "loss": 1.4429, "step": 34412 }, { "epoch": 2.6969435736677116, "grad_norm": 3.0181565284729004, "learning_rate": 2.4063314287603686e-06, "loss": 1.7409, "step": 34413 }, { "epoch": 2.6970219435736675, "grad_norm": 3.3997819423675537, "learning_rate": 2.406067803909599e-06, "loss": 1.4112, "step": 34414 }, { "epoch": 2.697100313479624, "grad_norm": 4.062281131744385, "learning_rate": 2.405804188924812e-06, "loss": 1.3123, "step": 34415 }, { "epoch": 2.69717868338558, "grad_norm": 4.373302459716797, "learning_rate": 2.405540583807015e-06, "loss": 1.6493, "step": 34416 }, { "epoch": 2.697257053291536, "grad_norm": 3.257758617401123, "learning_rate": 2.4052769885572092e-06, "loss": 1.9292, "step": 34417 }, { "epoch": 2.697335423197492, "grad_norm": 4.342550754547119, "learning_rate": 2.4050134031763983e-06, "loss": 1.2969, "step": 34418 }, { "epoch": 2.6974137931034483, "grad_norm": 2.484097719192505, "learning_rate": 2.4047498276655824e-06, "loss": 1.6133, "step": 34419 }, { "epoch": 2.6974921630094046, "grad_norm": 4.172735214233398, "learning_rate": 2.404486262025763e-06, "loss": 1.6774, "step": 34420 }, { "epoch": 2.6975705329153605, "grad_norm": 3.4321470260620117, "learning_rate": 2.4042227062579467e-06, "loss": 1.1004, "step": 34421 }, { "epoch": 2.6976489028213164, "grad_norm": 2.7626397609710693, "learning_rate": 2.403959160363135e-06, "loss": 1.3878, "step": 34422 }, { "epoch": 2.6977272727272728, "grad_norm": 3.4511070251464844, "learning_rate": 2.403695624342328e-06, "loss": 1.3414, "step": 34423 }, { "epoch": 2.697805642633229, "grad_norm": 3.4297609329223633, "learning_rate": 2.4034320981965298e-06, "loss": 1.3194, "step": 34424 }, { "epoch": 2.697884012539185, "grad_norm": 4.6551127433776855, "learning_rate": 2.4031685819267396e-06, "loss": 0.9249, "step": 34425 }, { "epoch": 2.697962382445141, "grad_norm": 3.6736347675323486, "learning_rate": 2.402905075533967e-06, "loss": 1.1912, "step": 34426 }, { "epoch": 2.698040752351097, "grad_norm": 4.746435165405273, "learning_rate": 2.4026415790192063e-06, "loss": 1.1967, "step": 34427 }, { "epoch": 2.6981191222570535, "grad_norm": 9.644268035888672, "learning_rate": 2.402378092383464e-06, "loss": 1.6817, "step": 34428 }, { "epoch": 2.6981974921630094, "grad_norm": 2.952751398086548, "learning_rate": 2.402114615627741e-06, "loss": 1.2033, "step": 34429 }, { "epoch": 2.6982758620689653, "grad_norm": 5.305264949798584, "learning_rate": 2.4018511487530395e-06, "loss": 1.0254, "step": 34430 }, { "epoch": 2.6983542319749216, "grad_norm": 2.5914459228515625, "learning_rate": 2.4015876917603615e-06, "loss": 1.1377, "step": 34431 }, { "epoch": 2.698432601880878, "grad_norm": 2.081726551055908, "learning_rate": 2.4013242446507094e-06, "loss": 1.2704, "step": 34432 }, { "epoch": 2.698510971786834, "grad_norm": 2.5438730716705322, "learning_rate": 2.401060807425085e-06, "loss": 1.5072, "step": 34433 }, { "epoch": 2.6985893416927897, "grad_norm": 3.526851177215576, "learning_rate": 2.400797380084491e-06, "loss": 1.3964, "step": 34434 }, { "epoch": 2.698667711598746, "grad_norm": 2.790466547012329, "learning_rate": 2.4005339626299285e-06, "loss": 1.2336, "step": 34435 }, { "epoch": 2.6987460815047024, "grad_norm": 2.440532684326172, "learning_rate": 2.4002705550623994e-06, "loss": 1.3612, "step": 34436 }, { "epoch": 2.6988244514106583, "grad_norm": 3.997084379196167, "learning_rate": 2.400007157382908e-06, "loss": 1.541, "step": 34437 }, { "epoch": 2.698902821316614, "grad_norm": 2.562734603881836, "learning_rate": 2.3997437695924495e-06, "loss": 1.4233, "step": 34438 }, { "epoch": 2.6989811912225705, "grad_norm": 2.397512435913086, "learning_rate": 2.399480391692033e-06, "loss": 1.2814, "step": 34439 }, { "epoch": 2.699059561128527, "grad_norm": 6.062220096588135, "learning_rate": 2.3992170236826567e-06, "loss": 1.0916, "step": 34440 }, { "epoch": 2.6991379310344827, "grad_norm": 6.501625061035156, "learning_rate": 2.3989536655653252e-06, "loss": 0.512, "step": 34441 }, { "epoch": 2.6992163009404386, "grad_norm": 4.195839881896973, "learning_rate": 2.3986903173410357e-06, "loss": 1.1338, "step": 34442 }, { "epoch": 2.699294670846395, "grad_norm": 3.021305799484253, "learning_rate": 2.3984269790107907e-06, "loss": 1.1003, "step": 34443 }, { "epoch": 2.6993730407523513, "grad_norm": 2.774442195892334, "learning_rate": 2.3981636505755964e-06, "loss": 1.3342, "step": 34444 }, { "epoch": 2.699451410658307, "grad_norm": 3.4710264205932617, "learning_rate": 2.39790033203645e-06, "loss": 1.3668, "step": 34445 }, { "epoch": 2.699529780564263, "grad_norm": 3.807326555252075, "learning_rate": 2.3976370233943534e-06, "loss": 1.3268, "step": 34446 }, { "epoch": 2.6996081504702194, "grad_norm": 4.814950466156006, "learning_rate": 2.3973737246503092e-06, "loss": 1.5248, "step": 34447 }, { "epoch": 2.6996865203761757, "grad_norm": 3.192946672439575, "learning_rate": 2.397110435805319e-06, "loss": 1.4969, "step": 34448 }, { "epoch": 2.6997648902821316, "grad_norm": 2.466822862625122, "learning_rate": 2.3968471568603835e-06, "loss": 1.4915, "step": 34449 }, { "epoch": 2.6998432601880875, "grad_norm": 4.43037748336792, "learning_rate": 2.3965838878165043e-06, "loss": 2.2113, "step": 34450 }, { "epoch": 2.699921630094044, "grad_norm": 3.1246089935302734, "learning_rate": 2.3963206286746826e-06, "loss": 1.4534, "step": 34451 }, { "epoch": 2.7, "grad_norm": 2.200071334838867, "learning_rate": 2.3960573794359203e-06, "loss": 1.1214, "step": 34452 }, { "epoch": 2.700078369905956, "grad_norm": 5.777470111846924, "learning_rate": 2.395794140101218e-06, "loss": 1.1075, "step": 34453 }, { "epoch": 2.700156739811912, "grad_norm": 3.9318771362304688, "learning_rate": 2.3955309106715776e-06, "loss": 1.189, "step": 34454 }, { "epoch": 2.7002351097178683, "grad_norm": 4.439810752868652, "learning_rate": 2.3952676911479995e-06, "loss": 1.6742, "step": 34455 }, { "epoch": 2.7003134796238246, "grad_norm": 4.9518208503723145, "learning_rate": 2.3950044815314873e-06, "loss": 1.7109, "step": 34456 }, { "epoch": 2.7003918495297805, "grad_norm": 2.9885547161102295, "learning_rate": 2.394741281823036e-06, "loss": 1.1018, "step": 34457 }, { "epoch": 2.700470219435737, "grad_norm": 2.2016513347625732, "learning_rate": 2.394478092023653e-06, "loss": 1.3494, "step": 34458 }, { "epoch": 2.7005485893416927, "grad_norm": 10.467475891113281, "learning_rate": 2.3942149121343383e-06, "loss": 1.3882, "step": 34459 }, { "epoch": 2.700626959247649, "grad_norm": 2.45761775970459, "learning_rate": 2.39395174215609e-06, "loss": 1.536, "step": 34460 }, { "epoch": 2.700705329153605, "grad_norm": 2.6924545764923096, "learning_rate": 2.3936885820899108e-06, "loss": 1.2838, "step": 34461 }, { "epoch": 2.7007836990595613, "grad_norm": 2.8429174423217773, "learning_rate": 2.3934254319368e-06, "loss": 1.3155, "step": 34462 }, { "epoch": 2.700862068965517, "grad_norm": 12.920897483825684, "learning_rate": 2.3931622916977632e-06, "loss": 1.2021, "step": 34463 }, { "epoch": 2.7009404388714735, "grad_norm": 5.673586368560791, "learning_rate": 2.3928991613737967e-06, "loss": 1.377, "step": 34464 }, { "epoch": 2.7010188087774294, "grad_norm": 3.3137195110321045, "learning_rate": 2.3926360409659022e-06, "loss": 1.335, "step": 34465 }, { "epoch": 2.7010971786833857, "grad_norm": 3.476457357406616, "learning_rate": 2.3923729304750813e-06, "loss": 1.4129, "step": 34466 }, { "epoch": 2.7011755485893416, "grad_norm": 3.5168397426605225, "learning_rate": 2.3921098299023348e-06, "loss": 1.0735, "step": 34467 }, { "epoch": 2.701253918495298, "grad_norm": 3.426976203918457, "learning_rate": 2.3918467392486626e-06, "loss": 0.8927, "step": 34468 }, { "epoch": 2.701332288401254, "grad_norm": 8.458663940429688, "learning_rate": 2.391583658515066e-06, "loss": 1.094, "step": 34469 }, { "epoch": 2.70141065830721, "grad_norm": 3.4598753452301025, "learning_rate": 2.391320587702547e-06, "loss": 1.1686, "step": 34470 }, { "epoch": 2.701489028213166, "grad_norm": 7.565968036651611, "learning_rate": 2.3910575268121006e-06, "loss": 1.2441, "step": 34471 }, { "epoch": 2.7015673981191224, "grad_norm": 8.682022094726562, "learning_rate": 2.3907944758447337e-06, "loss": 1.1882, "step": 34472 }, { "epoch": 2.7016457680250783, "grad_norm": 3.1691715717315674, "learning_rate": 2.390531434801444e-06, "loss": 1.3951, "step": 34473 }, { "epoch": 2.7017241379310346, "grad_norm": 2.035238742828369, "learning_rate": 2.3902684036832346e-06, "loss": 1.2716, "step": 34474 }, { "epoch": 2.7018025078369905, "grad_norm": 7.1837592124938965, "learning_rate": 2.390005382491102e-06, "loss": 2.0843, "step": 34475 }, { "epoch": 2.701880877742947, "grad_norm": 6.509190559387207, "learning_rate": 2.389742371226046e-06, "loss": 1.2324, "step": 34476 }, { "epoch": 2.7019592476489027, "grad_norm": 3.0746614933013916, "learning_rate": 2.389479369889072e-06, "loss": 1.3119, "step": 34477 }, { "epoch": 2.702037617554859, "grad_norm": 2.967559337615967, "learning_rate": 2.3892163784811785e-06, "loss": 1.1326, "step": 34478 }, { "epoch": 2.702115987460815, "grad_norm": 9.037217140197754, "learning_rate": 2.3889533970033633e-06, "loss": 0.7796, "step": 34479 }, { "epoch": 2.7021943573667713, "grad_norm": 3.3872365951538086, "learning_rate": 2.388690425456629e-06, "loss": 1.4519, "step": 34480 }, { "epoch": 2.702272727272727, "grad_norm": 2.983952760696411, "learning_rate": 2.3884274638419724e-06, "loss": 1.4912, "step": 34481 }, { "epoch": 2.7023510971786835, "grad_norm": 4.670076847076416, "learning_rate": 2.3881645121604004e-06, "loss": 1.498, "step": 34482 }, { "epoch": 2.7024294670846394, "grad_norm": 2.7807793617248535, "learning_rate": 2.387901570412907e-06, "loss": 1.3664, "step": 34483 }, { "epoch": 2.7025078369905957, "grad_norm": 5.789765357971191, "learning_rate": 2.3876386386004947e-06, "loss": 1.7343, "step": 34484 }, { "epoch": 2.7025862068965516, "grad_norm": 9.31588077545166, "learning_rate": 2.387375716724163e-06, "loss": 1.1825, "step": 34485 }, { "epoch": 2.702664576802508, "grad_norm": 3.4760143756866455, "learning_rate": 2.3871128047849122e-06, "loss": 1.1405, "step": 34486 }, { "epoch": 2.702742946708464, "grad_norm": 4.03831148147583, "learning_rate": 2.3868499027837424e-06, "loss": 1.3371, "step": 34487 }, { "epoch": 2.70282131661442, "grad_norm": 5.353871822357178, "learning_rate": 2.3865870107216533e-06, "loss": 1.5445, "step": 34488 }, { "epoch": 2.702899686520376, "grad_norm": 11.276369094848633, "learning_rate": 2.386324128599646e-06, "loss": 1.2371, "step": 34489 }, { "epoch": 2.7029780564263324, "grad_norm": 6.3106255531311035, "learning_rate": 2.3860612564187163e-06, "loss": 1.2389, "step": 34490 }, { "epoch": 2.7030564263322883, "grad_norm": 2.6745095252990723, "learning_rate": 2.385798394179869e-06, "loss": 1.0817, "step": 34491 }, { "epoch": 2.7031347962382446, "grad_norm": 8.790925025939941, "learning_rate": 2.3855355418841016e-06, "loss": 0.9429, "step": 34492 }, { "epoch": 2.7032131661442005, "grad_norm": 3.192249298095703, "learning_rate": 2.3852726995324153e-06, "loss": 0.7787, "step": 34493 }, { "epoch": 2.703291536050157, "grad_norm": 6.046824932098389, "learning_rate": 2.3850098671258076e-06, "loss": 1.3935, "step": 34494 }, { "epoch": 2.7033699059561127, "grad_norm": 2.503293514251709, "learning_rate": 2.3847470446652777e-06, "loss": 1.1981, "step": 34495 }, { "epoch": 2.703448275862069, "grad_norm": 2.9048006534576416, "learning_rate": 2.384484232151828e-06, "loss": 1.4224, "step": 34496 }, { "epoch": 2.703526645768025, "grad_norm": 2.5289218425750732, "learning_rate": 2.3842214295864586e-06, "loss": 1.2201, "step": 34497 }, { "epoch": 2.7036050156739813, "grad_norm": 2.6341471672058105, "learning_rate": 2.3839586369701655e-06, "loss": 1.1991, "step": 34498 }, { "epoch": 2.703683385579937, "grad_norm": 3.85929274559021, "learning_rate": 2.3836958543039507e-06, "loss": 1.6217, "step": 34499 }, { "epoch": 2.7037617554858935, "grad_norm": 2.9807605743408203, "learning_rate": 2.3834330815888124e-06, "loss": 1.5734, "step": 34500 }, { "epoch": 2.7038401253918494, "grad_norm": 3.8435962200164795, "learning_rate": 2.383170318825751e-06, "loss": 1.259, "step": 34501 }, { "epoch": 2.7039184952978057, "grad_norm": 2.5092809200286865, "learning_rate": 2.3829075660157656e-06, "loss": 1.6194, "step": 34502 }, { "epoch": 2.7039968652037616, "grad_norm": 2.9119014739990234, "learning_rate": 2.3826448231598554e-06, "loss": 1.2894, "step": 34503 }, { "epoch": 2.704075235109718, "grad_norm": 3.9031429290771484, "learning_rate": 2.38238209025902e-06, "loss": 1.5267, "step": 34504 }, { "epoch": 2.704153605015674, "grad_norm": 3.3226404190063477, "learning_rate": 2.382119367314259e-06, "loss": 0.6834, "step": 34505 }, { "epoch": 2.70423197492163, "grad_norm": 2.272775650024414, "learning_rate": 2.3818566543265707e-06, "loss": 1.4114, "step": 34506 }, { "epoch": 2.7043103448275865, "grad_norm": 2.9055795669555664, "learning_rate": 2.3815939512969553e-06, "loss": 1.0052, "step": 34507 }, { "epoch": 2.7043887147335424, "grad_norm": 3.01132869720459, "learning_rate": 2.3813312582264137e-06, "loss": 1.3388, "step": 34508 }, { "epoch": 2.7044670846394983, "grad_norm": 4.0808844566345215, "learning_rate": 2.3810685751159385e-06, "loss": 1.4772, "step": 34509 }, { "epoch": 2.7045454545454546, "grad_norm": 3.4835691452026367, "learning_rate": 2.380805901966536e-06, "loss": 1.4087, "step": 34510 }, { "epoch": 2.704623824451411, "grad_norm": 3.4531712532043457, "learning_rate": 2.380543238779202e-06, "loss": 1.2606, "step": 34511 }, { "epoch": 2.704702194357367, "grad_norm": 3.341020345687866, "learning_rate": 2.3802805855549384e-06, "loss": 1.407, "step": 34512 }, { "epoch": 2.7047805642633227, "grad_norm": 3.142681837081909, "learning_rate": 2.3800179422947397e-06, "loss": 1.2612, "step": 34513 }, { "epoch": 2.704858934169279, "grad_norm": 2.266000986099243, "learning_rate": 2.3797553089996045e-06, "loss": 1.2804, "step": 34514 }, { "epoch": 2.7049373040752354, "grad_norm": 2.3215737342834473, "learning_rate": 2.3794926856705387e-06, "loss": 1.0413, "step": 34515 }, { "epoch": 2.7050156739811912, "grad_norm": 2.26302433013916, "learning_rate": 2.379230072308535e-06, "loss": 1.0518, "step": 34516 }, { "epoch": 2.705094043887147, "grad_norm": 2.354586124420166, "learning_rate": 2.3789674689145937e-06, "loss": 1.4649, "step": 34517 }, { "epoch": 2.7051724137931035, "grad_norm": 5.739841461181641, "learning_rate": 2.378704875489714e-06, "loss": 1.3338, "step": 34518 }, { "epoch": 2.70525078369906, "grad_norm": 2.3161776065826416, "learning_rate": 2.378442292034895e-06, "loss": 1.3755, "step": 34519 }, { "epoch": 2.7053291536050157, "grad_norm": 3.4921445846557617, "learning_rate": 2.378179718551134e-06, "loss": 1.3832, "step": 34520 }, { "epoch": 2.7054075235109716, "grad_norm": 2.9539549350738525, "learning_rate": 2.3779171550394313e-06, "loss": 1.3852, "step": 34521 }, { "epoch": 2.705485893416928, "grad_norm": 2.9743435382843018, "learning_rate": 2.377654601500785e-06, "loss": 1.3966, "step": 34522 }, { "epoch": 2.7055642633228842, "grad_norm": 6.066790580749512, "learning_rate": 2.3773920579361935e-06, "loss": 1.3127, "step": 34523 }, { "epoch": 2.70564263322884, "grad_norm": 11.677046775817871, "learning_rate": 2.377129524346656e-06, "loss": 1.7618, "step": 34524 }, { "epoch": 2.705721003134796, "grad_norm": 5.263341426849365, "learning_rate": 2.37686700073317e-06, "loss": 1.1139, "step": 34525 }, { "epoch": 2.7057993730407524, "grad_norm": 4.296816825866699, "learning_rate": 2.3766044870967368e-06, "loss": 1.5483, "step": 34526 }, { "epoch": 2.7058777429467087, "grad_norm": 7.928128242492676, "learning_rate": 2.376341983438351e-06, "loss": 1.9551, "step": 34527 }, { "epoch": 2.7059561128526646, "grad_norm": 3.1700289249420166, "learning_rate": 2.3760794897590106e-06, "loss": 1.8357, "step": 34528 }, { "epoch": 2.7060344827586205, "grad_norm": 2.8059425354003906, "learning_rate": 2.375817006059718e-06, "loss": 1.3249, "step": 34529 }, { "epoch": 2.706112852664577, "grad_norm": 10.884718894958496, "learning_rate": 2.3755545323414716e-06, "loss": 1.8416, "step": 34530 }, { "epoch": 2.706191222570533, "grad_norm": 3.2209160327911377, "learning_rate": 2.3752920686052656e-06, "loss": 1.1103, "step": 34531 }, { "epoch": 2.706269592476489, "grad_norm": 3.035433292388916, "learning_rate": 2.3750296148521005e-06, "loss": 1.7343, "step": 34532 }, { "epoch": 2.706347962382445, "grad_norm": 2.3352553844451904, "learning_rate": 2.374767171082973e-06, "loss": 1.0447, "step": 34533 }, { "epoch": 2.7064263322884012, "grad_norm": 4.350378036499023, "learning_rate": 2.374504737298886e-06, "loss": 1.3104, "step": 34534 }, { "epoch": 2.7065047021943576, "grad_norm": 4.050750255584717, "learning_rate": 2.3742423135008325e-06, "loss": 1.487, "step": 34535 }, { "epoch": 2.7065830721003135, "grad_norm": 4.568387985229492, "learning_rate": 2.3739798996898127e-06, "loss": 1.1731, "step": 34536 }, { "epoch": 2.7066614420062693, "grad_norm": 2.5993599891662598, "learning_rate": 2.373717495866825e-06, "loss": 1.3719, "step": 34537 }, { "epoch": 2.7067398119122257, "grad_norm": 4.722971439361572, "learning_rate": 2.3734551020328668e-06, "loss": 1.5313, "step": 34538 }, { "epoch": 2.706818181818182, "grad_norm": 2.3555495738983154, "learning_rate": 2.373192718188936e-06, "loss": 1.0, "step": 34539 }, { "epoch": 2.706896551724138, "grad_norm": 2.6608877182006836, "learning_rate": 2.3729303443360312e-06, "loss": 1.4552, "step": 34540 }, { "epoch": 2.706974921630094, "grad_norm": 4.4429612159729, "learning_rate": 2.3726679804751495e-06, "loss": 1.5686, "step": 34541 }, { "epoch": 2.70705329153605, "grad_norm": 8.203486442565918, "learning_rate": 2.37240562660729e-06, "loss": 1.0545, "step": 34542 }, { "epoch": 2.7071316614420065, "grad_norm": 4.47379732131958, "learning_rate": 2.3721432827334494e-06, "loss": 1.6707, "step": 34543 }, { "epoch": 2.7072100313479623, "grad_norm": 2.7591500282287598, "learning_rate": 2.371880948854626e-06, "loss": 0.9267, "step": 34544 }, { "epoch": 2.7072884012539182, "grad_norm": 2.614126205444336, "learning_rate": 2.3716186249718205e-06, "loss": 1.3647, "step": 34545 }, { "epoch": 2.7073667711598746, "grad_norm": 4.739411354064941, "learning_rate": 2.371356311086023e-06, "loss": 1.1162, "step": 34546 }, { "epoch": 2.707445141065831, "grad_norm": 5.299035549163818, "learning_rate": 2.3710940071982375e-06, "loss": 1.1763, "step": 34547 }, { "epoch": 2.707523510971787, "grad_norm": 3.715489387512207, "learning_rate": 2.3708317133094605e-06, "loss": 1.1691, "step": 34548 }, { "epoch": 2.7076018808777427, "grad_norm": 4.3866963386535645, "learning_rate": 2.3705694294206914e-06, "loss": 1.4766, "step": 34549 }, { "epoch": 2.707680250783699, "grad_norm": 3.4240667819976807, "learning_rate": 2.3703071555329237e-06, "loss": 1.4013, "step": 34550 }, { "epoch": 2.7077586206896553, "grad_norm": 3.2594356536865234, "learning_rate": 2.370044891647155e-06, "loss": 1.3172, "step": 34551 }, { "epoch": 2.7078369905956112, "grad_norm": 4.079066753387451, "learning_rate": 2.3697826377643864e-06, "loss": 1.2965, "step": 34552 }, { "epoch": 2.7079153605015676, "grad_norm": 2.3466434478759766, "learning_rate": 2.3695203938856154e-06, "loss": 1.3702, "step": 34553 }, { "epoch": 2.7079937304075234, "grad_norm": 2.403564691543579, "learning_rate": 2.3692581600118358e-06, "loss": 1.3845, "step": 34554 }, { "epoch": 2.70807210031348, "grad_norm": 6.976996898651123, "learning_rate": 2.3689959361440475e-06, "loss": 1.7012, "step": 34555 }, { "epoch": 2.7081504702194357, "grad_norm": 7.032373905181885, "learning_rate": 2.368733722283247e-06, "loss": 1.4494, "step": 34556 }, { "epoch": 2.708228840125392, "grad_norm": 2.505866289138794, "learning_rate": 2.3684715184304313e-06, "loss": 1.2732, "step": 34557 }, { "epoch": 2.708307210031348, "grad_norm": 2.491574764251709, "learning_rate": 2.368209324586599e-06, "loss": 1.184, "step": 34558 }, { "epoch": 2.708385579937304, "grad_norm": 3.048665761947632, "learning_rate": 2.3679471407527465e-06, "loss": 1.4563, "step": 34559 }, { "epoch": 2.70846394984326, "grad_norm": 9.447243690490723, "learning_rate": 2.367684966929871e-06, "loss": 0.9577, "step": 34560 }, { "epoch": 2.7085423197492164, "grad_norm": 3.6427836418151855, "learning_rate": 2.36742280311897e-06, "loss": 1.175, "step": 34561 }, { "epoch": 2.7086206896551723, "grad_norm": 4.500559329986572, "learning_rate": 2.367160649321041e-06, "loss": 1.5018, "step": 34562 }, { "epoch": 2.7086990595611287, "grad_norm": 7.6588592529296875, "learning_rate": 2.3668985055370796e-06, "loss": 1.0696, "step": 34563 }, { "epoch": 2.7087774294670846, "grad_norm": 4.71364688873291, "learning_rate": 2.3666363717680863e-06, "loss": 1.1119, "step": 34564 }, { "epoch": 2.708855799373041, "grad_norm": 4.10884428024292, "learning_rate": 2.366374248015052e-06, "loss": 1.166, "step": 34565 }, { "epoch": 2.7089341692789968, "grad_norm": 6.173657417297363, "learning_rate": 2.366112134278979e-06, "loss": 1.02, "step": 34566 }, { "epoch": 2.709012539184953, "grad_norm": 2.4781179428100586, "learning_rate": 2.3658500305608644e-06, "loss": 1.3035, "step": 34567 }, { "epoch": 2.709090909090909, "grad_norm": 5.664767742156982, "learning_rate": 2.365587936861701e-06, "loss": 1.7933, "step": 34568 }, { "epoch": 2.7091692789968653, "grad_norm": 3.5415804386138916, "learning_rate": 2.365325853182489e-06, "loss": 1.4836, "step": 34569 }, { "epoch": 2.709247648902821, "grad_norm": 6.505616188049316, "learning_rate": 2.365063779524222e-06, "loss": 1.54, "step": 34570 }, { "epoch": 2.7093260188087775, "grad_norm": 2.9523842334747314, "learning_rate": 2.364801715887902e-06, "loss": 1.2674, "step": 34571 }, { "epoch": 2.7094043887147334, "grad_norm": 4.42698860168457, "learning_rate": 2.364539662274522e-06, "loss": 0.9921, "step": 34572 }, { "epoch": 2.7094827586206898, "grad_norm": 3.196323871612549, "learning_rate": 2.3642776186850784e-06, "loss": 1.2226, "step": 34573 }, { "epoch": 2.7095611285266457, "grad_norm": 3.0727157592773438, "learning_rate": 2.36401558512057e-06, "loss": 1.0855, "step": 34574 }, { "epoch": 2.709639498432602, "grad_norm": 3.1390111446380615, "learning_rate": 2.363753561581992e-06, "loss": 1.2922, "step": 34575 }, { "epoch": 2.709717868338558, "grad_norm": 2.9796812534332275, "learning_rate": 2.3634915480703407e-06, "loss": 1.419, "step": 34576 }, { "epoch": 2.709796238244514, "grad_norm": 3.1070871353149414, "learning_rate": 2.3632295445866136e-06, "loss": 1.2556, "step": 34577 }, { "epoch": 2.70987460815047, "grad_norm": 2.8883183002471924, "learning_rate": 2.3629675511318074e-06, "loss": 1.1961, "step": 34578 }, { "epoch": 2.7099529780564264, "grad_norm": 2.9570205211639404, "learning_rate": 2.3627055677069183e-06, "loss": 1.4121, "step": 34579 }, { "epoch": 2.7100313479623823, "grad_norm": 2.885662317276001, "learning_rate": 2.362443594312942e-06, "loss": 1.3182, "step": 34580 }, { "epoch": 2.7101097178683387, "grad_norm": 8.366321563720703, "learning_rate": 2.362181630950876e-06, "loss": 0.3552, "step": 34581 }, { "epoch": 2.7101880877742945, "grad_norm": 4.609787940979004, "learning_rate": 2.361919677621718e-06, "loss": 1.2789, "step": 34582 }, { "epoch": 2.710266457680251, "grad_norm": 4.760426998138428, "learning_rate": 2.361657734326461e-06, "loss": 1.7892, "step": 34583 }, { "epoch": 2.7103448275862068, "grad_norm": 3.144322395324707, "learning_rate": 2.3613958010661e-06, "loss": 1.5258, "step": 34584 }, { "epoch": 2.710423197492163, "grad_norm": 6.300255298614502, "learning_rate": 2.361133877841637e-06, "loss": 1.0806, "step": 34585 }, { "epoch": 2.710501567398119, "grad_norm": 3.8777503967285156, "learning_rate": 2.3608719646540666e-06, "loss": 1.2719, "step": 34586 }, { "epoch": 2.7105799373040753, "grad_norm": 2.0011215209960938, "learning_rate": 2.360610061504382e-06, "loss": 1.4888, "step": 34587 }, { "epoch": 2.710658307210031, "grad_norm": 2.3895297050476074, "learning_rate": 2.360348168393581e-06, "loss": 1.5092, "step": 34588 }, { "epoch": 2.7107366771159875, "grad_norm": 6.3979973793029785, "learning_rate": 2.360086285322658e-06, "loss": 0.89, "step": 34589 }, { "epoch": 2.7108150470219434, "grad_norm": 2.133995771408081, "learning_rate": 2.3598244122926146e-06, "loss": 1.1381, "step": 34590 }, { "epoch": 2.7108934169278998, "grad_norm": 6.049236297607422, "learning_rate": 2.3595625493044412e-06, "loss": 1.3781, "step": 34591 }, { "epoch": 2.7109717868338556, "grad_norm": 3.943908929824829, "learning_rate": 2.359300696359135e-06, "loss": 0.7284, "step": 34592 }, { "epoch": 2.711050156739812, "grad_norm": 3.006258964538574, "learning_rate": 2.3590388534576934e-06, "loss": 1.6708, "step": 34593 }, { "epoch": 2.711128526645768, "grad_norm": 3.976516008377075, "learning_rate": 2.358777020601111e-06, "loss": 1.546, "step": 34594 }, { "epoch": 2.711206896551724, "grad_norm": 3.8655872344970703, "learning_rate": 2.3585151977903846e-06, "loss": 1.3689, "step": 34595 }, { "epoch": 2.71128526645768, "grad_norm": 6.493919849395752, "learning_rate": 2.3582533850265093e-06, "loss": 1.0916, "step": 34596 }, { "epoch": 2.7113636363636364, "grad_norm": 2.6934828758239746, "learning_rate": 2.3579915823104836e-06, "loss": 1.2197, "step": 34597 }, { "epoch": 2.7114420062695928, "grad_norm": 2.624643325805664, "learning_rate": 2.3577297896432964e-06, "loss": 1.1861, "step": 34598 }, { "epoch": 2.7115203761755486, "grad_norm": 3.578282356262207, "learning_rate": 2.3574680070259505e-06, "loss": 2.0243, "step": 34599 }, { "epoch": 2.7115987460815045, "grad_norm": 3.0095770359039307, "learning_rate": 2.3572062344594387e-06, "loss": 1.1211, "step": 34600 }, { "epoch": 2.711677115987461, "grad_norm": 3.4509952068328857, "learning_rate": 2.356944471944759e-06, "loss": 1.4034, "step": 34601 }, { "epoch": 2.711755485893417, "grad_norm": 2.4629249572753906, "learning_rate": 2.3566827194829027e-06, "loss": 1.3213, "step": 34602 }, { "epoch": 2.711833855799373, "grad_norm": 2.635786294937134, "learning_rate": 2.3564209770748665e-06, "loss": 0.663, "step": 34603 }, { "epoch": 2.711912225705329, "grad_norm": 3.4058635234832764, "learning_rate": 2.356159244721648e-06, "loss": 1.4301, "step": 34604 }, { "epoch": 2.7119905956112853, "grad_norm": 3.374662399291992, "learning_rate": 2.3558975224242444e-06, "loss": 1.1935, "step": 34605 }, { "epoch": 2.7120689655172416, "grad_norm": 3.0610909461975098, "learning_rate": 2.355635810183646e-06, "loss": 1.0055, "step": 34606 }, { "epoch": 2.7121473354231975, "grad_norm": 6.915953159332275, "learning_rate": 2.3553741080008513e-06, "loss": 1.0902, "step": 34607 }, { "epoch": 2.7122257053291534, "grad_norm": 2.6029183864593506, "learning_rate": 2.3551124158768535e-06, "loss": 1.082, "step": 34608 }, { "epoch": 2.7123040752351097, "grad_norm": 4.224459648132324, "learning_rate": 2.3548507338126523e-06, "loss": 1.6293, "step": 34609 }, { "epoch": 2.712382445141066, "grad_norm": 2.786316394805908, "learning_rate": 2.3545890618092395e-06, "loss": 1.4991, "step": 34610 }, { "epoch": 2.712460815047022, "grad_norm": 7.39560604095459, "learning_rate": 2.354327399867611e-06, "loss": 1.6912, "step": 34611 }, { "epoch": 2.712539184952978, "grad_norm": 3.313161611557007, "learning_rate": 2.3540657479887615e-06, "loss": 0.8349, "step": 34612 }, { "epoch": 2.712617554858934, "grad_norm": 3.665522575378418, "learning_rate": 2.353804106173687e-06, "loss": 1.3511, "step": 34613 }, { "epoch": 2.7126959247648905, "grad_norm": 3.044590711593628, "learning_rate": 2.3535424744233828e-06, "loss": 1.5767, "step": 34614 }, { "epoch": 2.7127742946708464, "grad_norm": 8.053328514099121, "learning_rate": 2.3532808527388435e-06, "loss": 1.5079, "step": 34615 }, { "epoch": 2.7128526645768023, "grad_norm": 3.725834369659424, "learning_rate": 2.353019241121067e-06, "loss": 1.4491, "step": 34616 }, { "epoch": 2.7129310344827586, "grad_norm": 2.3903989791870117, "learning_rate": 2.352757639571041e-06, "loss": 1.2013, "step": 34617 }, { "epoch": 2.713009404388715, "grad_norm": 4.049055576324463, "learning_rate": 2.3524960480897673e-06, "loss": 1.1464, "step": 34618 }, { "epoch": 2.713087774294671, "grad_norm": 4.234801769256592, "learning_rate": 2.3522344666782388e-06, "loss": 0.9058, "step": 34619 }, { "epoch": 2.7131661442006267, "grad_norm": 3.0288171768188477, "learning_rate": 2.3519728953374526e-06, "loss": 1.2968, "step": 34620 }, { "epoch": 2.713244514106583, "grad_norm": 4.35807991027832, "learning_rate": 2.3517113340683993e-06, "loss": 1.3601, "step": 34621 }, { "epoch": 2.7133228840125394, "grad_norm": 2.68595290184021, "learning_rate": 2.3514497828720738e-06, "loss": 1.1625, "step": 34622 }, { "epoch": 2.7134012539184953, "grad_norm": 3.766033172607422, "learning_rate": 2.351188241749477e-06, "loss": 1.4045, "step": 34623 }, { "epoch": 2.713479623824451, "grad_norm": 2.619320869445801, "learning_rate": 2.3509267107015977e-06, "loss": 1.0251, "step": 34624 }, { "epoch": 2.7135579937304075, "grad_norm": 8.013434410095215, "learning_rate": 2.3506651897294326e-06, "loss": 0.9265, "step": 34625 }, { "epoch": 2.713636363636364, "grad_norm": 2.386943817138672, "learning_rate": 2.3504036788339763e-06, "loss": 1.2581, "step": 34626 }, { "epoch": 2.7137147335423197, "grad_norm": 5.8235650062561035, "learning_rate": 2.3501421780162233e-06, "loss": 1.6237, "step": 34627 }, { "epoch": 2.7137931034482756, "grad_norm": 3.0248708724975586, "learning_rate": 2.3498806872771683e-06, "loss": 1.3005, "step": 34628 }, { "epoch": 2.713871473354232, "grad_norm": 2.2659718990325928, "learning_rate": 2.3496192066178063e-06, "loss": 1.3842, "step": 34629 }, { "epoch": 2.7139498432601883, "grad_norm": 3.4926958084106445, "learning_rate": 2.3493577360391316e-06, "loss": 1.3082, "step": 34630 }, { "epoch": 2.714028213166144, "grad_norm": 5.537425994873047, "learning_rate": 2.349096275542138e-06, "loss": 1.1889, "step": 34631 }, { "epoch": 2.7141065830721, "grad_norm": 4.095458984375, "learning_rate": 2.3488348251278214e-06, "loss": 1.6281, "step": 34632 }, { "epoch": 2.7141849529780564, "grad_norm": 4.261371612548828, "learning_rate": 2.3485733847971747e-06, "loss": 0.954, "step": 34633 }, { "epoch": 2.7142633228840127, "grad_norm": 2.7281076908111572, "learning_rate": 2.3483119545511935e-06, "loss": 1.0657, "step": 34634 }, { "epoch": 2.7143416927899686, "grad_norm": 3.0236079692840576, "learning_rate": 2.3480505343908718e-06, "loss": 1.146, "step": 34635 }, { "epoch": 2.7144200626959245, "grad_norm": 5.425270080566406, "learning_rate": 2.347789124317203e-06, "loss": 1.167, "step": 34636 }, { "epoch": 2.714498432601881, "grad_norm": 3.5015852451324463, "learning_rate": 2.3475277243311824e-06, "loss": 1.276, "step": 34637 }, { "epoch": 2.714576802507837, "grad_norm": 3.5084447860717773, "learning_rate": 2.347266334433806e-06, "loss": 1.4995, "step": 34638 }, { "epoch": 2.714655172413793, "grad_norm": 2.2931935787200928, "learning_rate": 2.3470049546260638e-06, "loss": 1.2461, "step": 34639 }, { "epoch": 2.714733542319749, "grad_norm": 4.472950458526611, "learning_rate": 2.3467435849089502e-06, "loss": 1.4749, "step": 34640 }, { "epoch": 2.7148119122257053, "grad_norm": 2.7467424869537354, "learning_rate": 2.3464822252834635e-06, "loss": 1.3306, "step": 34641 }, { "epoch": 2.7148902821316616, "grad_norm": 3.0940849781036377, "learning_rate": 2.346220875750597e-06, "loss": 0.7777, "step": 34642 }, { "epoch": 2.7149686520376175, "grad_norm": 3.2248072624206543, "learning_rate": 2.345959536311341e-06, "loss": 1.4698, "step": 34643 }, { "epoch": 2.7150470219435734, "grad_norm": 4.013673782348633, "learning_rate": 2.3456982069666928e-06, "loss": 1.4214, "step": 34644 }, { "epoch": 2.7151253918495297, "grad_norm": 3.0212242603302, "learning_rate": 2.345436887717645e-06, "loss": 1.2997, "step": 34645 }, { "epoch": 2.715203761755486, "grad_norm": 2.365229845046997, "learning_rate": 2.345175578565191e-06, "loss": 1.4256, "step": 34646 }, { "epoch": 2.715282131661442, "grad_norm": 5.945120811462402, "learning_rate": 2.344914279510326e-06, "loss": 1.0727, "step": 34647 }, { "epoch": 2.7153605015673983, "grad_norm": 5.143802642822266, "learning_rate": 2.344652990554043e-06, "loss": 1.8447, "step": 34648 }, { "epoch": 2.715438871473354, "grad_norm": 7.767187595367432, "learning_rate": 2.344391711697337e-06, "loss": 0.8685, "step": 34649 }, { "epoch": 2.7155172413793105, "grad_norm": 2.5485639572143555, "learning_rate": 2.3441304429412e-06, "loss": 1.3048, "step": 34650 }, { "epoch": 2.7155956112852664, "grad_norm": 3.1044352054595947, "learning_rate": 2.3438691842866268e-06, "loss": 1.7631, "step": 34651 }, { "epoch": 2.7156739811912227, "grad_norm": 6.272075653076172, "learning_rate": 2.3436079357346113e-06, "loss": 1.2397, "step": 34652 }, { "epoch": 2.7157523510971786, "grad_norm": 2.9017937183380127, "learning_rate": 2.3433466972861484e-06, "loss": 1.2286, "step": 34653 }, { "epoch": 2.715830721003135, "grad_norm": 2.6350650787353516, "learning_rate": 2.3430854689422266e-06, "loss": 1.6515, "step": 34654 }, { "epoch": 2.715909090909091, "grad_norm": 3.335670232772827, "learning_rate": 2.342824250703845e-06, "loss": 1.7685, "step": 34655 }, { "epoch": 2.715987460815047, "grad_norm": 2.693274974822998, "learning_rate": 2.342563042571995e-06, "loss": 1.2338, "step": 34656 }, { "epoch": 2.716065830721003, "grad_norm": 2.1281039714813232, "learning_rate": 2.342301844547672e-06, "loss": 1.4146, "step": 34657 }, { "epoch": 2.7161442006269594, "grad_norm": 3.0114519596099854, "learning_rate": 2.342040656631866e-06, "loss": 1.3315, "step": 34658 }, { "epoch": 2.7162225705329153, "grad_norm": 2.5351498126983643, "learning_rate": 2.34177947882557e-06, "loss": 1.3406, "step": 34659 }, { "epoch": 2.7163009404388716, "grad_norm": 8.387256622314453, "learning_rate": 2.341518311129781e-06, "loss": 2.0992, "step": 34660 }, { "epoch": 2.7163793103448275, "grad_norm": 2.876049280166626, "learning_rate": 2.3412571535454937e-06, "loss": 1.7342, "step": 34661 }, { "epoch": 2.716457680250784, "grad_norm": 3.1762144565582275, "learning_rate": 2.3409960060736963e-06, "loss": 1.4882, "step": 34662 }, { "epoch": 2.7165360501567397, "grad_norm": 3.1786282062530518, "learning_rate": 2.3407348687153843e-06, "loss": 1.0786, "step": 34663 }, { "epoch": 2.716614420062696, "grad_norm": 2.2582333087921143, "learning_rate": 2.340473741471549e-06, "loss": 1.4375, "step": 34664 }, { "epoch": 2.716692789968652, "grad_norm": 3.257284641265869, "learning_rate": 2.3402126243431895e-06, "loss": 1.4339, "step": 34665 }, { "epoch": 2.7167711598746083, "grad_norm": 4.163980484008789, "learning_rate": 2.339951517331293e-06, "loss": 0.6955, "step": 34666 }, { "epoch": 2.716849529780564, "grad_norm": 2.4161486625671387, "learning_rate": 2.339690420436855e-06, "loss": 1.5336, "step": 34667 }, { "epoch": 2.7169278996865205, "grad_norm": 3.198387622833252, "learning_rate": 2.3394293336608674e-06, "loss": 1.0933, "step": 34668 }, { "epoch": 2.7170062695924764, "grad_norm": 2.880762815475464, "learning_rate": 2.339168257004325e-06, "loss": 1.5435, "step": 34669 }, { "epoch": 2.7170846394984327, "grad_norm": 2.3171913623809814, "learning_rate": 2.3389071904682193e-06, "loss": 1.3182, "step": 34670 }, { "epoch": 2.7171630094043886, "grad_norm": 2.9784178733825684, "learning_rate": 2.3386461340535437e-06, "loss": 0.3254, "step": 34671 }, { "epoch": 2.717241379310345, "grad_norm": 2.5197339057922363, "learning_rate": 2.3383850877612934e-06, "loss": 1.4434, "step": 34672 }, { "epoch": 2.717319749216301, "grad_norm": 2.6210572719573975, "learning_rate": 2.3381240515924557e-06, "loss": 1.387, "step": 34673 }, { "epoch": 2.717398119122257, "grad_norm": 3.8957393169403076, "learning_rate": 2.3378630255480283e-06, "loss": 1.3806, "step": 34674 }, { "epoch": 2.717476489028213, "grad_norm": 2.6093480587005615, "learning_rate": 2.3376020096290027e-06, "loss": 1.1627, "step": 34675 }, { "epoch": 2.7175548589341694, "grad_norm": 4.572939872741699, "learning_rate": 2.3373410038363737e-06, "loss": 1.8617, "step": 34676 }, { "epoch": 2.7176332288401253, "grad_norm": 2.7114169597625732, "learning_rate": 2.33708000817113e-06, "loss": 1.4887, "step": 34677 }, { "epoch": 2.7177115987460816, "grad_norm": 2.6259818077087402, "learning_rate": 2.3368190226342633e-06, "loss": 1.2358, "step": 34678 }, { "epoch": 2.7177899686520375, "grad_norm": 3.0015268325805664, "learning_rate": 2.336558047226774e-06, "loss": 1.6397, "step": 34679 }, { "epoch": 2.717868338557994, "grad_norm": 2.4127776622772217, "learning_rate": 2.336297081949648e-06, "loss": 1.3784, "step": 34680 }, { "epoch": 2.7179467084639497, "grad_norm": 6.73197603225708, "learning_rate": 2.336036126803879e-06, "loss": 1.0853, "step": 34681 }, { "epoch": 2.718025078369906, "grad_norm": 2.6487114429473877, "learning_rate": 2.3357751817904607e-06, "loss": 1.2713, "step": 34682 }, { "epoch": 2.718103448275862, "grad_norm": 3.5123744010925293, "learning_rate": 2.3355142469103852e-06, "loss": 1.3069, "step": 34683 }, { "epoch": 2.7181818181818183, "grad_norm": 2.012383222579956, "learning_rate": 2.335253322164645e-06, "loss": 1.2063, "step": 34684 }, { "epoch": 2.718260188087774, "grad_norm": 2.1334798336029053, "learning_rate": 2.3349924075542325e-06, "loss": 1.3555, "step": 34685 }, { "epoch": 2.7183385579937305, "grad_norm": 7.334455490112305, "learning_rate": 2.33473150308014e-06, "loss": 1.8585, "step": 34686 }, { "epoch": 2.7184169278996864, "grad_norm": 2.6137964725494385, "learning_rate": 2.3344706087433595e-06, "loss": 1.394, "step": 34687 }, { "epoch": 2.7184952978056427, "grad_norm": 3.4287569522857666, "learning_rate": 2.334209724544884e-06, "loss": 0.8953, "step": 34688 }, { "epoch": 2.7185736677115986, "grad_norm": 3.5298402309417725, "learning_rate": 2.3339488504857054e-06, "loss": 1.7747, "step": 34689 }, { "epoch": 2.718652037617555, "grad_norm": 2.811833620071411, "learning_rate": 2.333687986566816e-06, "loss": 1.4318, "step": 34690 }, { "epoch": 2.718730407523511, "grad_norm": 3.95198392868042, "learning_rate": 2.33342713278921e-06, "loss": 1.3359, "step": 34691 }, { "epoch": 2.718808777429467, "grad_norm": 9.025541305541992, "learning_rate": 2.3331662891538737e-06, "loss": 1.3319, "step": 34692 }, { "epoch": 2.7188871473354235, "grad_norm": 2.393907070159912, "learning_rate": 2.3329054556618048e-06, "loss": 1.5139, "step": 34693 }, { "epoch": 2.7189655172413794, "grad_norm": 3.3498077392578125, "learning_rate": 2.3326446323139956e-06, "loss": 1.3883, "step": 34694 }, { "epoch": 2.7190438871473352, "grad_norm": 3.5812458992004395, "learning_rate": 2.3323838191114346e-06, "loss": 1.5268, "step": 34695 }, { "epoch": 2.7191222570532916, "grad_norm": 3.3160505294799805, "learning_rate": 2.332123016055115e-06, "loss": 1.0551, "step": 34696 }, { "epoch": 2.719200626959248, "grad_norm": 3.2298314571380615, "learning_rate": 2.3318622231460276e-06, "loss": 1.497, "step": 34697 }, { "epoch": 2.719278996865204, "grad_norm": 2.8217685222625732, "learning_rate": 2.3316014403851696e-06, "loss": 1.1843, "step": 34698 }, { "epoch": 2.7193573667711597, "grad_norm": 2.136082887649536, "learning_rate": 2.331340667773527e-06, "loss": 1.1677, "step": 34699 }, { "epoch": 2.719435736677116, "grad_norm": 3.114398717880249, "learning_rate": 2.3310799053120942e-06, "loss": 1.5335, "step": 34700 }, { "epoch": 2.7195141065830724, "grad_norm": 4.012972354888916, "learning_rate": 2.330819153001862e-06, "loss": 1.602, "step": 34701 }, { "epoch": 2.7195924764890282, "grad_norm": 3.1893203258514404, "learning_rate": 2.330558410843823e-06, "loss": 1.244, "step": 34702 }, { "epoch": 2.719670846394984, "grad_norm": 12.484435081481934, "learning_rate": 2.3302976788389686e-06, "loss": 1.18, "step": 34703 }, { "epoch": 2.7197492163009405, "grad_norm": 6.478857040405273, "learning_rate": 2.330036956988291e-06, "loss": 2.0946, "step": 34704 }, { "epoch": 2.719827586206897, "grad_norm": 3.2973921298980713, "learning_rate": 2.3297762452927814e-06, "loss": 1.2459, "step": 34705 }, { "epoch": 2.7199059561128527, "grad_norm": 2.5340614318847656, "learning_rate": 2.329515543753431e-06, "loss": 1.1991, "step": 34706 }, { "epoch": 2.7199843260188086, "grad_norm": 2.874748945236206, "learning_rate": 2.329254852371232e-06, "loss": 1.4055, "step": 34707 }, { "epoch": 2.720062695924765, "grad_norm": 2.8865506649017334, "learning_rate": 2.328994171147176e-06, "loss": 1.1554, "step": 34708 }, { "epoch": 2.7201410658307212, "grad_norm": 3.1430740356445312, "learning_rate": 2.328733500082256e-06, "loss": 1.1312, "step": 34709 }, { "epoch": 2.720219435736677, "grad_norm": 3.1996748447418213, "learning_rate": 2.32847283917746e-06, "loss": 1.5526, "step": 34710 }, { "epoch": 2.720297805642633, "grad_norm": 8.22885799407959, "learning_rate": 2.328212188433779e-06, "loss": 1.0023, "step": 34711 }, { "epoch": 2.7203761755485893, "grad_norm": 3.8582139015197754, "learning_rate": 2.327951547852208e-06, "loss": 1.0758, "step": 34712 }, { "epoch": 2.7204545454545457, "grad_norm": 3.3559868335723877, "learning_rate": 2.327690917433739e-06, "loss": 1.0811, "step": 34713 }, { "epoch": 2.7205329153605016, "grad_norm": 7.492234230041504, "learning_rate": 2.3274302971793593e-06, "loss": 1.5348, "step": 34714 }, { "epoch": 2.7206112852664575, "grad_norm": 3.643808126449585, "learning_rate": 2.3271696870900622e-06, "loss": 1.4776, "step": 34715 }, { "epoch": 2.720689655172414, "grad_norm": 2.241858959197998, "learning_rate": 2.3269090871668373e-06, "loss": 1.1427, "step": 34716 }, { "epoch": 2.72076802507837, "grad_norm": 4.2586469650268555, "learning_rate": 2.3266484974106805e-06, "loss": 1.5726, "step": 34717 }, { "epoch": 2.720846394984326, "grad_norm": 4.250324249267578, "learning_rate": 2.326387917822578e-06, "loss": 1.387, "step": 34718 }, { "epoch": 2.720924764890282, "grad_norm": 2.6495327949523926, "learning_rate": 2.3261273484035224e-06, "loss": 1.4592, "step": 34719 }, { "epoch": 2.7210031347962382, "grad_norm": 3.1037232875823975, "learning_rate": 2.325866789154505e-06, "loss": 1.2123, "step": 34720 }, { "epoch": 2.7210815047021946, "grad_norm": 2.945197820663452, "learning_rate": 2.3256062400765167e-06, "loss": 1.5352, "step": 34721 }, { "epoch": 2.7211598746081505, "grad_norm": 2.3513927459716797, "learning_rate": 2.325345701170549e-06, "loss": 1.2806, "step": 34722 }, { "epoch": 2.7212382445141063, "grad_norm": 6.245904445648193, "learning_rate": 2.3250851724375924e-06, "loss": 1.6345, "step": 34723 }, { "epoch": 2.7213166144200627, "grad_norm": 5.327425479888916, "learning_rate": 2.3248246538786397e-06, "loss": 0.7359, "step": 34724 }, { "epoch": 2.721394984326019, "grad_norm": 6.892463684082031, "learning_rate": 2.324564145494676e-06, "loss": 1.0013, "step": 34725 }, { "epoch": 2.721473354231975, "grad_norm": 3.794196605682373, "learning_rate": 2.324303647286698e-06, "loss": 2.3114, "step": 34726 }, { "epoch": 2.721551724137931, "grad_norm": 2.475497245788574, "learning_rate": 2.324043159255695e-06, "loss": 1.2196, "step": 34727 }, { "epoch": 2.721630094043887, "grad_norm": 14.888387680053711, "learning_rate": 2.3237826814026586e-06, "loss": 1.1451, "step": 34728 }, { "epoch": 2.7217084639498434, "grad_norm": 3.235349416732788, "learning_rate": 2.323522213728577e-06, "loss": 1.5196, "step": 34729 }, { "epoch": 2.7217868338557993, "grad_norm": 3.22898006439209, "learning_rate": 2.32326175623444e-06, "loss": 1.3363, "step": 34730 }, { "epoch": 2.7218652037617552, "grad_norm": 4.903692722320557, "learning_rate": 2.323001308921242e-06, "loss": 0.9242, "step": 34731 }, { "epoch": 2.7219435736677116, "grad_norm": 2.7929673194885254, "learning_rate": 2.3227408717899746e-06, "loss": 1.0157, "step": 34732 }, { "epoch": 2.722021943573668, "grad_norm": 4.858251094818115, "learning_rate": 2.3224804448416234e-06, "loss": 1.2975, "step": 34733 }, { "epoch": 2.722100313479624, "grad_norm": 3.7952098846435547, "learning_rate": 2.322220028077182e-06, "loss": 1.199, "step": 34734 }, { "epoch": 2.7221786833855797, "grad_norm": 4.342190265655518, "learning_rate": 2.3219596214976394e-06, "loss": 1.4654, "step": 34735 }, { "epoch": 2.722257053291536, "grad_norm": 5.963059902191162, "learning_rate": 2.321699225103987e-06, "loss": 1.4095, "step": 34736 }, { "epoch": 2.7223354231974923, "grad_norm": 3.5062808990478516, "learning_rate": 2.3214388388972154e-06, "loss": 1.8028, "step": 34737 }, { "epoch": 2.722413793103448, "grad_norm": 2.4626376628875732, "learning_rate": 2.3211784628783146e-06, "loss": 1.305, "step": 34738 }, { "epoch": 2.7224921630094046, "grad_norm": 2.75718355178833, "learning_rate": 2.320918097048275e-06, "loss": 1.5062, "step": 34739 }, { "epoch": 2.7225705329153604, "grad_norm": 5.775511264801025, "learning_rate": 2.3206577414080865e-06, "loss": 1.5528, "step": 34740 }, { "epoch": 2.7226489028213168, "grad_norm": 13.368675231933594, "learning_rate": 2.3203973959587407e-06, "loss": 2.138, "step": 34741 }, { "epoch": 2.7227272727272727, "grad_norm": 2.5515716075897217, "learning_rate": 2.3201370607012258e-06, "loss": 1.3552, "step": 34742 }, { "epoch": 2.722805642633229, "grad_norm": 6.930479526519775, "learning_rate": 2.319876735636534e-06, "loss": 0.8353, "step": 34743 }, { "epoch": 2.722884012539185, "grad_norm": 2.86331844329834, "learning_rate": 2.3196164207656535e-06, "loss": 1.3589, "step": 34744 }, { "epoch": 2.722962382445141, "grad_norm": 4.398262977600098, "learning_rate": 2.319356116089576e-06, "loss": 0.9429, "step": 34745 }, { "epoch": 2.723040752351097, "grad_norm": 6.06805419921875, "learning_rate": 2.3190958216092906e-06, "loss": 0.9324, "step": 34746 }, { "epoch": 2.7231191222570534, "grad_norm": 5.060059070587158, "learning_rate": 2.3188355373257902e-06, "loss": 0.812, "step": 34747 }, { "epoch": 2.7231974921630093, "grad_norm": 3.0388097763061523, "learning_rate": 2.3185752632400583e-06, "loss": 1.4432, "step": 34748 }, { "epoch": 2.7232758620689657, "grad_norm": 2.9740006923675537, "learning_rate": 2.3183149993530905e-06, "loss": 1.161, "step": 34749 }, { "epoch": 2.7233542319749215, "grad_norm": 8.200193405151367, "learning_rate": 2.318054745665877e-06, "loss": 1.5536, "step": 34750 }, { "epoch": 2.723432601880878, "grad_norm": 3.787839651107788, "learning_rate": 2.317794502179404e-06, "loss": 1.2629, "step": 34751 }, { "epoch": 2.7235109717868338, "grad_norm": 7.895936012268066, "learning_rate": 2.317534268894663e-06, "loss": 1.0511, "step": 34752 }, { "epoch": 2.72358934169279, "grad_norm": 3.122112989425659, "learning_rate": 2.3172740458126413e-06, "loss": 1.1255, "step": 34753 }, { "epoch": 2.723667711598746, "grad_norm": 2.586071491241455, "learning_rate": 2.3170138329343356e-06, "loss": 1.5959, "step": 34754 }, { "epoch": 2.7237460815047023, "grad_norm": 2.9841372966766357, "learning_rate": 2.3167536302607296e-06, "loss": 1.4345, "step": 34755 }, { "epoch": 2.723824451410658, "grad_norm": 2.996342182159424, "learning_rate": 2.316493437792814e-06, "loss": 1.2776, "step": 34756 }, { "epoch": 2.7239028213166145, "grad_norm": 4.0533766746521, "learning_rate": 2.3162332555315792e-06, "loss": 1.6546, "step": 34757 }, { "epoch": 2.7239811912225704, "grad_norm": 2.8413596153259277, "learning_rate": 2.315973083478015e-06, "loss": 1.6009, "step": 34758 }, { "epoch": 2.7240595611285268, "grad_norm": 4.168465614318848, "learning_rate": 2.3157129216331104e-06, "loss": 1.0781, "step": 34759 }, { "epoch": 2.7241379310344827, "grad_norm": 6.889924049377441, "learning_rate": 2.3154527699978546e-06, "loss": 1.2487, "step": 34760 }, { "epoch": 2.724216300940439, "grad_norm": 6.1508402824401855, "learning_rate": 2.315192628573238e-06, "loss": 0.8542, "step": 34761 }, { "epoch": 2.724294670846395, "grad_norm": 2.8658957481384277, "learning_rate": 2.3149324973602496e-06, "loss": 1.3809, "step": 34762 }, { "epoch": 2.724373040752351, "grad_norm": 10.03730583190918, "learning_rate": 2.3146723763598787e-06, "loss": 1.4234, "step": 34763 }, { "epoch": 2.724451410658307, "grad_norm": 2.8125312328338623, "learning_rate": 2.314412265573115e-06, "loss": 1.3854, "step": 34764 }, { "epoch": 2.7245297805642634, "grad_norm": 3.0857458114624023, "learning_rate": 2.3141521650009498e-06, "loss": 1.7646, "step": 34765 }, { "epoch": 2.7246081504702193, "grad_norm": 8.671871185302734, "learning_rate": 2.313892074644368e-06, "loss": 1.4855, "step": 34766 }, { "epoch": 2.7246865203761756, "grad_norm": 3.395268678665161, "learning_rate": 2.313631994504359e-06, "loss": 1.2991, "step": 34767 }, { "epoch": 2.7247648902821315, "grad_norm": 2.463045835494995, "learning_rate": 2.3133719245819163e-06, "loss": 1.3964, "step": 34768 }, { "epoch": 2.724843260188088, "grad_norm": 2.865262269973755, "learning_rate": 2.3131118648780286e-06, "loss": 1.4038, "step": 34769 }, { "epoch": 2.7249216300940438, "grad_norm": 2.482126474380493, "learning_rate": 2.312851815393681e-06, "loss": 1.4262, "step": 34770 }, { "epoch": 2.725, "grad_norm": 13.554703712463379, "learning_rate": 2.3125917761298653e-06, "loss": 1.411, "step": 34771 }, { "epoch": 2.725078369905956, "grad_norm": 3.25871205329895, "learning_rate": 2.3123317470875687e-06, "loss": 1.1637, "step": 34772 }, { "epoch": 2.7251567398119123, "grad_norm": 3.03796124458313, "learning_rate": 2.312071728267785e-06, "loss": 1.1834, "step": 34773 }, { "epoch": 2.725235109717868, "grad_norm": 6.329071998596191, "learning_rate": 2.311811719671498e-06, "loss": 1.6719, "step": 34774 }, { "epoch": 2.7253134796238245, "grad_norm": 3.8718385696411133, "learning_rate": 2.3115517212996986e-06, "loss": 1.3027, "step": 34775 }, { "epoch": 2.7253918495297804, "grad_norm": 2.58591365814209, "learning_rate": 2.311291733153375e-06, "loss": 1.1656, "step": 34776 }, { "epoch": 2.7254702194357368, "grad_norm": 2.9685189723968506, "learning_rate": 2.3110317552335177e-06, "loss": 1.1582, "step": 34777 }, { "epoch": 2.7255485893416926, "grad_norm": 2.4804279804229736, "learning_rate": 2.3107717875411132e-06, "loss": 1.1613, "step": 34778 }, { "epoch": 2.725626959247649, "grad_norm": 2.8848016262054443, "learning_rate": 2.3105118300771525e-06, "loss": 1.3104, "step": 34779 }, { "epoch": 2.725705329153605, "grad_norm": 2.667146682739258, "learning_rate": 2.3102518828426253e-06, "loss": 1.2794, "step": 34780 }, { "epoch": 2.725783699059561, "grad_norm": 2.9224884510040283, "learning_rate": 2.309991945838514e-06, "loss": 1.3375, "step": 34781 }, { "epoch": 2.725862068965517, "grad_norm": 4.210520267486572, "learning_rate": 2.3097320190658135e-06, "loss": 1.6851, "step": 34782 }, { "epoch": 2.7259404388714734, "grad_norm": 3.1138134002685547, "learning_rate": 2.3094721025255114e-06, "loss": 1.4068, "step": 34783 }, { "epoch": 2.7260188087774293, "grad_norm": 3.10382080078125, "learning_rate": 2.3092121962185964e-06, "loss": 1.2803, "step": 34784 }, { "epoch": 2.7260971786833856, "grad_norm": 3.4525938034057617, "learning_rate": 2.308952300146055e-06, "loss": 0.9885, "step": 34785 }, { "epoch": 2.7261755485893415, "grad_norm": 2.9502923488616943, "learning_rate": 2.3086924143088746e-06, "loss": 1.3233, "step": 34786 }, { "epoch": 2.726253918495298, "grad_norm": 3.9955027103424072, "learning_rate": 2.3084325387080483e-06, "loss": 1.6651, "step": 34787 }, { "epoch": 2.726332288401254, "grad_norm": 4.22289514541626, "learning_rate": 2.308172673344563e-06, "loss": 1.4501, "step": 34788 }, { "epoch": 2.72641065830721, "grad_norm": 3.2341201305389404, "learning_rate": 2.307912818219405e-06, "loss": 1.5125, "step": 34789 }, { "epoch": 2.726489028213166, "grad_norm": 12.122591972351074, "learning_rate": 2.307652973333564e-06, "loss": 1.3075, "step": 34790 }, { "epoch": 2.7265673981191223, "grad_norm": 3.588857650756836, "learning_rate": 2.3073931386880284e-06, "loss": 1.5509, "step": 34791 }, { "epoch": 2.7266457680250786, "grad_norm": 2.931027412414551, "learning_rate": 2.3071333142837855e-06, "loss": 1.3969, "step": 34792 }, { "epoch": 2.7267241379310345, "grad_norm": 2.851768732070923, "learning_rate": 2.3068735001218253e-06, "loss": 1.2184, "step": 34793 }, { "epoch": 2.7268025078369904, "grad_norm": 4.185490131378174, "learning_rate": 2.3066136962031344e-06, "loss": 1.4398, "step": 34794 }, { "epoch": 2.7268808777429467, "grad_norm": 2.3354904651641846, "learning_rate": 2.3063539025287014e-06, "loss": 1.2314, "step": 34795 }, { "epoch": 2.726959247648903, "grad_norm": 2.665081739425659, "learning_rate": 2.306094119099515e-06, "loss": 1.2014, "step": 34796 }, { "epoch": 2.727037617554859, "grad_norm": 2.106088876724243, "learning_rate": 2.305834345916563e-06, "loss": 1.2867, "step": 34797 }, { "epoch": 2.727115987460815, "grad_norm": 2.65278959274292, "learning_rate": 2.3055745829808327e-06, "loss": 1.0397, "step": 34798 }, { "epoch": 2.727194357366771, "grad_norm": 3.4737019538879395, "learning_rate": 2.3053148302933152e-06, "loss": 1.2244, "step": 34799 }, { "epoch": 2.7272727272727275, "grad_norm": 3.8735921382904053, "learning_rate": 2.305055087854992e-06, "loss": 1.1686, "step": 34800 }, { "epoch": 2.7273510971786834, "grad_norm": 2.835933208465576, "learning_rate": 2.3047953556668573e-06, "loss": 1.335, "step": 34801 }, { "epoch": 2.7274294670846393, "grad_norm": 4.105539321899414, "learning_rate": 2.3045356337298964e-06, "loss": 1.3957, "step": 34802 }, { "epoch": 2.7275078369905956, "grad_norm": 5.866325855255127, "learning_rate": 2.3042759220450992e-06, "loss": 1.036, "step": 34803 }, { "epoch": 2.727586206896552, "grad_norm": 3.544614315032959, "learning_rate": 2.3040162206134503e-06, "loss": 1.942, "step": 34804 }, { "epoch": 2.727664576802508, "grad_norm": 2.748378276824951, "learning_rate": 2.303756529435937e-06, "loss": 1.0928, "step": 34805 }, { "epoch": 2.7277429467084637, "grad_norm": 2.7255074977874756, "learning_rate": 2.3034968485135523e-06, "loss": 1.6048, "step": 34806 }, { "epoch": 2.72782131661442, "grad_norm": 4.340972900390625, "learning_rate": 2.303237177847279e-06, "loss": 1.2448, "step": 34807 }, { "epoch": 2.7278996865203764, "grad_norm": 3.241708517074585, "learning_rate": 2.3029775174381065e-06, "loss": 1.5059, "step": 34808 }, { "epoch": 2.7279780564263323, "grad_norm": 3.260237455368042, "learning_rate": 2.302717867287022e-06, "loss": 1.1084, "step": 34809 }, { "epoch": 2.728056426332288, "grad_norm": 3.4900660514831543, "learning_rate": 2.3024582273950136e-06, "loss": 1.4124, "step": 34810 }, { "epoch": 2.7281347962382445, "grad_norm": 3.243525505065918, "learning_rate": 2.3021985977630683e-06, "loss": 1.256, "step": 34811 }, { "epoch": 2.728213166144201, "grad_norm": 2.596686601638794, "learning_rate": 2.3019389783921742e-06, "loss": 1.1998, "step": 34812 }, { "epoch": 2.7282915360501567, "grad_norm": 2.5081121921539307, "learning_rate": 2.3016793692833176e-06, "loss": 1.1227, "step": 34813 }, { "epoch": 2.7283699059561126, "grad_norm": 10.289205551147461, "learning_rate": 2.3014197704374875e-06, "loss": 1.5067, "step": 34814 }, { "epoch": 2.728448275862069, "grad_norm": 3.4434814453125, "learning_rate": 2.3011601818556705e-06, "loss": 1.398, "step": 34815 }, { "epoch": 2.7285266457680253, "grad_norm": 4.161740779876709, "learning_rate": 2.3009006035388536e-06, "loss": 1.7858, "step": 34816 }, { "epoch": 2.728605015673981, "grad_norm": 3.498417854309082, "learning_rate": 2.3006410354880247e-06, "loss": 1.5263, "step": 34817 }, { "epoch": 2.728683385579937, "grad_norm": 4.035498142242432, "learning_rate": 2.300381477704173e-06, "loss": 1.2859, "step": 34818 }, { "epoch": 2.7287617554858934, "grad_norm": 3.1127877235412598, "learning_rate": 2.300121930188279e-06, "loss": 1.4891, "step": 34819 }, { "epoch": 2.7288401253918497, "grad_norm": 2.2106120586395264, "learning_rate": 2.2998623929413367e-06, "loss": 1.3954, "step": 34820 }, { "epoch": 2.7289184952978056, "grad_norm": 3.6974093914031982, "learning_rate": 2.2996028659643327e-06, "loss": 1.4872, "step": 34821 }, { "epoch": 2.7289968652037615, "grad_norm": 2.905117988586426, "learning_rate": 2.2993433492582508e-06, "loss": 1.0663, "step": 34822 }, { "epoch": 2.729075235109718, "grad_norm": 2.6335935592651367, "learning_rate": 2.29908384282408e-06, "loss": 1.2754, "step": 34823 }, { "epoch": 2.729153605015674, "grad_norm": 3.6713430881500244, "learning_rate": 2.298824346662804e-06, "loss": 1.4043, "step": 34824 }, { "epoch": 2.72923197492163, "grad_norm": 3.5547611713409424, "learning_rate": 2.2985648607754175e-06, "loss": 1.8174, "step": 34825 }, { "epoch": 2.729310344827586, "grad_norm": 3.619485855102539, "learning_rate": 2.2983053851629e-06, "loss": 1.3121, "step": 34826 }, { "epoch": 2.7293887147335423, "grad_norm": 5.142954349517822, "learning_rate": 2.2980459198262418e-06, "loss": 1.4015, "step": 34827 }, { "epoch": 2.7294670846394986, "grad_norm": 10.12546157836914, "learning_rate": 2.2977864647664287e-06, "loss": 0.7195, "step": 34828 }, { "epoch": 2.7295454545454545, "grad_norm": 10.320171356201172, "learning_rate": 2.2975270199844487e-06, "loss": 1.5815, "step": 34829 }, { "epoch": 2.7296238244514104, "grad_norm": 3.38893461227417, "learning_rate": 2.297267585481287e-06, "loss": 1.1872, "step": 34830 }, { "epoch": 2.7297021943573667, "grad_norm": 3.368299722671509, "learning_rate": 2.297008161257932e-06, "loss": 1.4258, "step": 34831 }, { "epoch": 2.729780564263323, "grad_norm": 5.391376972198486, "learning_rate": 2.296748747315369e-06, "loss": 0.9329, "step": 34832 }, { "epoch": 2.729858934169279, "grad_norm": 3.660691976547241, "learning_rate": 2.2964893436545853e-06, "loss": 1.507, "step": 34833 }, { "epoch": 2.7299373040752353, "grad_norm": 3.1387939453125, "learning_rate": 2.296229950276568e-06, "loss": 1.1646, "step": 34834 }, { "epoch": 2.730015673981191, "grad_norm": 2.7322723865509033, "learning_rate": 2.2959705671823024e-06, "loss": 1.4642, "step": 34835 }, { "epoch": 2.7300940438871475, "grad_norm": 4.979499340057373, "learning_rate": 2.2957111943727784e-06, "loss": 1.201, "step": 34836 }, { "epoch": 2.7301724137931034, "grad_norm": 12.017577171325684, "learning_rate": 2.295451831848976e-06, "loss": 1.2112, "step": 34837 }, { "epoch": 2.7302507836990597, "grad_norm": 5.297273635864258, "learning_rate": 2.295192479611888e-06, "loss": 1.0451, "step": 34838 }, { "epoch": 2.7303291536050156, "grad_norm": 3.1131908893585205, "learning_rate": 2.294933137662498e-06, "loss": 1.4908, "step": 34839 }, { "epoch": 2.730407523510972, "grad_norm": 13.513558387756348, "learning_rate": 2.2946738060017947e-06, "loss": 1.7427, "step": 34840 }, { "epoch": 2.730485893416928, "grad_norm": 3.249857187271118, "learning_rate": 2.2944144846307613e-06, "loss": 1.5899, "step": 34841 }, { "epoch": 2.730564263322884, "grad_norm": 6.02555513381958, "learning_rate": 2.2941551735503854e-06, "loss": 1.1396, "step": 34842 }, { "epoch": 2.73064263322884, "grad_norm": 4.831878662109375, "learning_rate": 2.2938958727616516e-06, "loss": 1.4108, "step": 34843 }, { "epoch": 2.7307210031347964, "grad_norm": 2.363461971282959, "learning_rate": 2.2936365822655515e-06, "loss": 1.1662, "step": 34844 }, { "epoch": 2.7307993730407523, "grad_norm": 2.64054274559021, "learning_rate": 2.2933773020630662e-06, "loss": 1.34, "step": 34845 }, { "epoch": 2.7308777429467086, "grad_norm": 3.2156074047088623, "learning_rate": 2.293118032155183e-06, "loss": 1.1293, "step": 34846 }, { "epoch": 2.7309561128526645, "grad_norm": 2.458984851837158, "learning_rate": 2.2928587725428885e-06, "loss": 1.3787, "step": 34847 }, { "epoch": 2.731034482758621, "grad_norm": 7.584465026855469, "learning_rate": 2.292599523227169e-06, "loss": 2.0303, "step": 34848 }, { "epoch": 2.7311128526645767, "grad_norm": 3.753204584121704, "learning_rate": 2.2923402842090105e-06, "loss": 1.044, "step": 34849 }, { "epoch": 2.731191222570533, "grad_norm": 7.420691967010498, "learning_rate": 2.2920810554893987e-06, "loss": 1.2515, "step": 34850 }, { "epoch": 2.731269592476489, "grad_norm": 3.469147205352783, "learning_rate": 2.291821837069319e-06, "loss": 1.2573, "step": 34851 }, { "epoch": 2.7313479623824453, "grad_norm": 2.749525308609009, "learning_rate": 2.2915626289497587e-06, "loss": 1.1096, "step": 34852 }, { "epoch": 2.731426332288401, "grad_norm": 4.270481586456299, "learning_rate": 2.291303431131703e-06, "loss": 1.6657, "step": 34853 }, { "epoch": 2.7315047021943575, "grad_norm": 2.8114235401153564, "learning_rate": 2.291044243616138e-06, "loss": 0.9025, "step": 34854 }, { "epoch": 2.7315830721003134, "grad_norm": 3.7307417392730713, "learning_rate": 2.2907850664040504e-06, "loss": 1.5362, "step": 34855 }, { "epoch": 2.7316614420062697, "grad_norm": 3.03428316116333, "learning_rate": 2.290525899496422e-06, "loss": 1.1408, "step": 34856 }, { "epoch": 2.7317398119122256, "grad_norm": 2.999110698699951, "learning_rate": 2.2902667428942425e-06, "loss": 1.4275, "step": 34857 }, { "epoch": 2.731818181818182, "grad_norm": 2.02323579788208, "learning_rate": 2.290007596598497e-06, "loss": 1.1031, "step": 34858 }, { "epoch": 2.731896551724138, "grad_norm": 4.629331588745117, "learning_rate": 2.2897484606101724e-06, "loss": 1.512, "step": 34859 }, { "epoch": 2.731974921630094, "grad_norm": 3.0078258514404297, "learning_rate": 2.2894893349302505e-06, "loss": 0.8656, "step": 34860 }, { "epoch": 2.73205329153605, "grad_norm": 5.7109222412109375, "learning_rate": 2.2892302195597173e-06, "loss": 1.2344, "step": 34861 }, { "epoch": 2.7321316614420064, "grad_norm": 3.3757212162017822, "learning_rate": 2.2889711144995636e-06, "loss": 1.1875, "step": 34862 }, { "epoch": 2.7322100313479623, "grad_norm": 2.890188694000244, "learning_rate": 2.2887120197507696e-06, "loss": 1.3078, "step": 34863 }, { "epoch": 2.7322884012539186, "grad_norm": 2.3002450466156006, "learning_rate": 2.2884529353143227e-06, "loss": 0.9162, "step": 34864 }, { "epoch": 2.7323667711598745, "grad_norm": 2.7897534370422363, "learning_rate": 2.2881938611912076e-06, "loss": 1.1992, "step": 34865 }, { "epoch": 2.732445141065831, "grad_norm": 2.3403024673461914, "learning_rate": 2.28793479738241e-06, "loss": 1.3867, "step": 34866 }, { "epoch": 2.7325235109717867, "grad_norm": 9.191755294799805, "learning_rate": 2.2876757438889165e-06, "loss": 1.586, "step": 34867 }, { "epoch": 2.732601880877743, "grad_norm": 3.235161542892456, "learning_rate": 2.2874167007117105e-06, "loss": 1.3656, "step": 34868 }, { "epoch": 2.732680250783699, "grad_norm": 3.335245370864868, "learning_rate": 2.287157667851778e-06, "loss": 1.2873, "step": 34869 }, { "epoch": 2.7327586206896552, "grad_norm": 4.488737106323242, "learning_rate": 2.2868986453101044e-06, "loss": 1.4024, "step": 34870 }, { "epoch": 2.732836990595611, "grad_norm": 2.4383156299591064, "learning_rate": 2.286639633087675e-06, "loss": 1.2265, "step": 34871 }, { "epoch": 2.7329153605015675, "grad_norm": 5.002219200134277, "learning_rate": 2.2863806311854748e-06, "loss": 1.6823, "step": 34872 }, { "epoch": 2.7329937304075234, "grad_norm": 3.8204760551452637, "learning_rate": 2.2861216396044887e-06, "loss": 1.6145, "step": 34873 }, { "epoch": 2.7330721003134797, "grad_norm": 3.1017959117889404, "learning_rate": 2.2858626583457045e-06, "loss": 1.4754, "step": 34874 }, { "epoch": 2.7331504702194356, "grad_norm": 3.9596030712127686, "learning_rate": 2.2856036874101e-06, "loss": 1.3716, "step": 34875 }, { "epoch": 2.733228840125392, "grad_norm": 3.2491981983184814, "learning_rate": 2.285344726798667e-06, "loss": 1.3888, "step": 34876 }, { "epoch": 2.733307210031348, "grad_norm": 2.8659474849700928, "learning_rate": 2.2850857765123906e-06, "loss": 1.2749, "step": 34877 }, { "epoch": 2.733385579937304, "grad_norm": 3.0951759815216064, "learning_rate": 2.284826836552251e-06, "loss": 0.9765, "step": 34878 }, { "epoch": 2.7334639498432605, "grad_norm": 2.730835437774658, "learning_rate": 2.284567906919236e-06, "loss": 1.3014, "step": 34879 }, { "epoch": 2.7335423197492164, "grad_norm": 3.433262348175049, "learning_rate": 2.284308987614328e-06, "loss": 1.427, "step": 34880 }, { "epoch": 2.7336206896551722, "grad_norm": 3.611938714981079, "learning_rate": 2.2840500786385174e-06, "loss": 1.4064, "step": 34881 }, { "epoch": 2.7336990595611286, "grad_norm": 1.8125767707824707, "learning_rate": 2.283791179992783e-06, "loss": 1.2555, "step": 34882 }, { "epoch": 2.733777429467085, "grad_norm": 2.3796231746673584, "learning_rate": 2.2835322916781127e-06, "loss": 1.4128, "step": 34883 }, { "epoch": 2.733855799373041, "grad_norm": 3.4101314544677734, "learning_rate": 2.2832734136954897e-06, "loss": 1.4245, "step": 34884 }, { "epoch": 2.7339341692789967, "grad_norm": 3.750908613204956, "learning_rate": 2.283014546045899e-06, "loss": 1.3614, "step": 34885 }, { "epoch": 2.734012539184953, "grad_norm": 2.329888105392456, "learning_rate": 2.282755688730326e-06, "loss": 1.0746, "step": 34886 }, { "epoch": 2.7340909090909093, "grad_norm": 9.617965698242188, "learning_rate": 2.2824968417497546e-06, "loss": 1.238, "step": 34887 }, { "epoch": 2.7341692789968652, "grad_norm": 3.0885705947875977, "learning_rate": 2.282238005105171e-06, "loss": 1.1401, "step": 34888 }, { "epoch": 2.734247648902821, "grad_norm": 4.026962757110596, "learning_rate": 2.2819791787975548e-06, "loss": 1.6707, "step": 34889 }, { "epoch": 2.7343260188087775, "grad_norm": 7.300018310546875, "learning_rate": 2.281720362827895e-06, "loss": 1.0095, "step": 34890 }, { "epoch": 2.734404388714734, "grad_norm": 3.669165849685669, "learning_rate": 2.2814615571971756e-06, "loss": 1.1138, "step": 34891 }, { "epoch": 2.7344827586206897, "grad_norm": 2.4722530841827393, "learning_rate": 2.2812027619063808e-06, "loss": 1.3358, "step": 34892 }, { "epoch": 2.7345611285266456, "grad_norm": 5.3474531173706055, "learning_rate": 2.2809439769564932e-06, "loss": 1.0753, "step": 34893 }, { "epoch": 2.734639498432602, "grad_norm": 4.515731334686279, "learning_rate": 2.280685202348496e-06, "loss": 1.4427, "step": 34894 }, { "epoch": 2.7347178683385582, "grad_norm": 6.562496662139893, "learning_rate": 2.280426438083378e-06, "loss": 1.7226, "step": 34895 }, { "epoch": 2.734796238244514, "grad_norm": 2.6142094135284424, "learning_rate": 2.2801676841621224e-06, "loss": 1.1672, "step": 34896 }, { "epoch": 2.73487460815047, "grad_norm": 3.735985279083252, "learning_rate": 2.2799089405857095e-06, "loss": 1.4858, "step": 34897 }, { "epoch": 2.7349529780564263, "grad_norm": 2.418431520462036, "learning_rate": 2.2796502073551265e-06, "loss": 1.5472, "step": 34898 }, { "epoch": 2.7350313479623827, "grad_norm": 3.0228946208953857, "learning_rate": 2.279391484471355e-06, "loss": 1.0631, "step": 34899 }, { "epoch": 2.7351097178683386, "grad_norm": 2.800539255142212, "learning_rate": 2.2791327719353847e-06, "loss": 1.1852, "step": 34900 }, { "epoch": 2.7351880877742945, "grad_norm": 3.3552780151367188, "learning_rate": 2.2788740697481933e-06, "loss": 1.1987, "step": 34901 }, { "epoch": 2.735266457680251, "grad_norm": 2.55743145942688, "learning_rate": 2.278615377910768e-06, "loss": 1.1655, "step": 34902 }, { "epoch": 2.735344827586207, "grad_norm": 4.603857040405273, "learning_rate": 2.2783566964240917e-06, "loss": 0.6487, "step": 34903 }, { "epoch": 2.735423197492163, "grad_norm": 3.4687705039978027, "learning_rate": 2.2780980252891484e-06, "loss": 1.6053, "step": 34904 }, { "epoch": 2.735501567398119, "grad_norm": 2.6604652404785156, "learning_rate": 2.2778393645069226e-06, "loss": 1.3455, "step": 34905 }, { "epoch": 2.7355799373040752, "grad_norm": 2.7266554832458496, "learning_rate": 2.2775807140783973e-06, "loss": 0.9692, "step": 34906 }, { "epoch": 2.7356583072100316, "grad_norm": 3.884678363800049, "learning_rate": 2.2773220740045588e-06, "loss": 1.0306, "step": 34907 }, { "epoch": 2.7357366771159874, "grad_norm": 2.7059733867645264, "learning_rate": 2.277063444286385e-06, "loss": 1.306, "step": 34908 }, { "epoch": 2.7358150470219433, "grad_norm": 2.8752896785736084, "learning_rate": 2.2768048249248648e-06, "loss": 1.2149, "step": 34909 }, { "epoch": 2.7358934169278997, "grad_norm": 4.019079685211182, "learning_rate": 2.27654621592098e-06, "loss": 1.6882, "step": 34910 }, { "epoch": 2.735971786833856, "grad_norm": 5.267476558685303, "learning_rate": 2.276287617275717e-06, "loss": 1.9319, "step": 34911 }, { "epoch": 2.736050156739812, "grad_norm": 4.453908443450928, "learning_rate": 2.276029028990055e-06, "loss": 1.275, "step": 34912 }, { "epoch": 2.7361285266457678, "grad_norm": 2.8038265705108643, "learning_rate": 2.2757704510649775e-06, "loss": 1.5276, "step": 34913 }, { "epoch": 2.736206896551724, "grad_norm": 2.1522903442382812, "learning_rate": 2.2755118835014715e-06, "loss": 1.3309, "step": 34914 }, { "epoch": 2.7362852664576804, "grad_norm": 2.8142428398132324, "learning_rate": 2.2752533263005213e-06, "loss": 1.2325, "step": 34915 }, { "epoch": 2.7363636363636363, "grad_norm": 3.910489320755005, "learning_rate": 2.274994779463106e-06, "loss": 1.4492, "step": 34916 }, { "epoch": 2.736442006269592, "grad_norm": 3.265964984893799, "learning_rate": 2.2747362429902103e-06, "loss": 1.3289, "step": 34917 }, { "epoch": 2.7365203761755486, "grad_norm": 3.4330034255981445, "learning_rate": 2.2744777168828193e-06, "loss": 1.2605, "step": 34918 }, { "epoch": 2.736598746081505, "grad_norm": 2.5336077213287354, "learning_rate": 2.274219201141914e-06, "loss": 1.3044, "step": 34919 }, { "epoch": 2.7366771159874608, "grad_norm": 3.8426146507263184, "learning_rate": 2.2739606957684795e-06, "loss": 1.3416, "step": 34920 }, { "epoch": 2.7367554858934167, "grad_norm": 3.2926137447357178, "learning_rate": 2.273702200763498e-06, "loss": 0.9866, "step": 34921 }, { "epoch": 2.736833855799373, "grad_norm": 7.254519462585449, "learning_rate": 2.273443716127953e-06, "loss": 0.9445, "step": 34922 }, { "epoch": 2.7369122257053293, "grad_norm": 2.687441110610962, "learning_rate": 2.2731852418628274e-06, "loss": 1.6149, "step": 34923 }, { "epoch": 2.736990595611285, "grad_norm": 2.173412322998047, "learning_rate": 2.2729267779691056e-06, "loss": 1.1431, "step": 34924 }, { "epoch": 2.737068965517241, "grad_norm": 2.529061794281006, "learning_rate": 2.272668324447769e-06, "loss": 1.6232, "step": 34925 }, { "epoch": 2.7371473354231974, "grad_norm": 4.368470191955566, "learning_rate": 2.2724098812998024e-06, "loss": 1.4697, "step": 34926 }, { "epoch": 2.7372257053291538, "grad_norm": 5.025223255157471, "learning_rate": 2.2721514485261846e-06, "loss": 1.3246, "step": 34927 }, { "epoch": 2.7373040752351097, "grad_norm": 5.315410614013672, "learning_rate": 2.271893026127904e-06, "loss": 1.4861, "step": 34928 }, { "epoch": 2.737382445141066, "grad_norm": 11.200032234191895, "learning_rate": 2.2716346141059405e-06, "loss": 1.2052, "step": 34929 }, { "epoch": 2.737460815047022, "grad_norm": 5.6713337898254395, "learning_rate": 2.2713762124612794e-06, "loss": 1.2537, "step": 34930 }, { "epoch": 2.737539184952978, "grad_norm": 2.4617507457733154, "learning_rate": 2.2711178211948992e-06, "loss": 1.3116, "step": 34931 }, { "epoch": 2.737617554858934, "grad_norm": 3.534691572189331, "learning_rate": 2.2708594403077842e-06, "loss": 1.1944, "step": 34932 }, { "epoch": 2.7376959247648904, "grad_norm": 7.169647216796875, "learning_rate": 2.2706010698009216e-06, "loss": 1.7665, "step": 34933 }, { "epoch": 2.7377742946708463, "grad_norm": 6.825963020324707, "learning_rate": 2.270342709675289e-06, "loss": 1.4518, "step": 34934 }, { "epoch": 2.7378526645768027, "grad_norm": 2.597322463989258, "learning_rate": 2.27008435993187e-06, "loss": 1.0684, "step": 34935 }, { "epoch": 2.7379310344827585, "grad_norm": 1.904502034187317, "learning_rate": 2.2698260205716484e-06, "loss": 1.4085, "step": 34936 }, { "epoch": 2.738009404388715, "grad_norm": 4.13007926940918, "learning_rate": 2.269567691595606e-06, "loss": 1.6288, "step": 34937 }, { "epoch": 2.7380877742946708, "grad_norm": 3.048663854598999, "learning_rate": 2.269309373004726e-06, "loss": 1.1851, "step": 34938 }, { "epoch": 2.738166144200627, "grad_norm": 3.073441505432129, "learning_rate": 2.269051064799991e-06, "loss": 1.8128, "step": 34939 }, { "epoch": 2.738244514106583, "grad_norm": 3.4640135765075684, "learning_rate": 2.2687927669823825e-06, "loss": 1.5776, "step": 34940 }, { "epoch": 2.7383228840125393, "grad_norm": 2.94067120552063, "learning_rate": 2.268534479552883e-06, "loss": 1.1581, "step": 34941 }, { "epoch": 2.738401253918495, "grad_norm": 5.714704990386963, "learning_rate": 2.268276202512476e-06, "loss": 1.6021, "step": 34942 }, { "epoch": 2.7384796238244515, "grad_norm": 3.720214366912842, "learning_rate": 2.268017935862143e-06, "loss": 1.4429, "step": 34943 }, { "epoch": 2.7385579937304074, "grad_norm": 3.321057081222534, "learning_rate": 2.267759679602869e-06, "loss": 1.1917, "step": 34944 }, { "epoch": 2.7386363636363638, "grad_norm": 4.271366119384766, "learning_rate": 2.2675014337356294e-06, "loss": 1.612, "step": 34945 }, { "epoch": 2.7387147335423196, "grad_norm": 3.086127281188965, "learning_rate": 2.267243198261413e-06, "loss": 1.0931, "step": 34946 }, { "epoch": 2.738793103448276, "grad_norm": 3.1750431060791016, "learning_rate": 2.2669849731811995e-06, "loss": 1.1906, "step": 34947 }, { "epoch": 2.738871473354232, "grad_norm": 2.845964193344116, "learning_rate": 2.2667267584959735e-06, "loss": 0.7794, "step": 34948 }, { "epoch": 2.738949843260188, "grad_norm": 2.9496190547943115, "learning_rate": 2.266468554206713e-06, "loss": 1.6453, "step": 34949 }, { "epoch": 2.739028213166144, "grad_norm": 4.55495023727417, "learning_rate": 2.2662103603144003e-06, "loss": 1.3288, "step": 34950 }, { "epoch": 2.7391065830721004, "grad_norm": 6.630066394805908, "learning_rate": 2.2659521768200216e-06, "loss": 0.8985, "step": 34951 }, { "epoch": 2.7391849529780563, "grad_norm": 4.106409549713135, "learning_rate": 2.265694003724558e-06, "loss": 1.247, "step": 34952 }, { "epoch": 2.7392633228840126, "grad_norm": 11.63116455078125, "learning_rate": 2.265435841028988e-06, "loss": 1.1275, "step": 34953 }, { "epoch": 2.7393416927899685, "grad_norm": 3.665832042694092, "learning_rate": 2.2651776887342953e-06, "loss": 1.0997, "step": 34954 }, { "epoch": 2.739420062695925, "grad_norm": 2.9505202770233154, "learning_rate": 2.2649195468414625e-06, "loss": 1.0053, "step": 34955 }, { "epoch": 2.7394984326018808, "grad_norm": 3.4154467582702637, "learning_rate": 2.264661415351471e-06, "loss": 1.3051, "step": 34956 }, { "epoch": 2.739576802507837, "grad_norm": 2.924842596054077, "learning_rate": 2.2644032942653015e-06, "loss": 1.2108, "step": 34957 }, { "epoch": 2.739655172413793, "grad_norm": 2.7585315704345703, "learning_rate": 2.2641451835839375e-06, "loss": 1.4806, "step": 34958 }, { "epoch": 2.7397335423197493, "grad_norm": 2.549091100692749, "learning_rate": 2.2638870833083597e-06, "loss": 1.3337, "step": 34959 }, { "epoch": 2.739811912225705, "grad_norm": 5.231932163238525, "learning_rate": 2.2636289934395506e-06, "loss": 1.4566, "step": 34960 }, { "epoch": 2.7398902821316615, "grad_norm": 5.341717720031738, "learning_rate": 2.263370913978491e-06, "loss": 1.1326, "step": 34961 }, { "epoch": 2.7399686520376174, "grad_norm": 4.51024055480957, "learning_rate": 2.263112844926163e-06, "loss": 1.1743, "step": 34962 }, { "epoch": 2.7400470219435737, "grad_norm": 7.367696762084961, "learning_rate": 2.26285478628355e-06, "loss": 1.6736, "step": 34963 }, { "epoch": 2.7401253918495296, "grad_norm": 3.0953946113586426, "learning_rate": 2.2625967380516275e-06, "loss": 1.4181, "step": 34964 }, { "epoch": 2.740203761755486, "grad_norm": 4.795701503753662, "learning_rate": 2.2623387002313824e-06, "loss": 1.3766, "step": 34965 }, { "epoch": 2.740282131661442, "grad_norm": 5.144029140472412, "learning_rate": 2.2620806728237955e-06, "loss": 1.7295, "step": 34966 }, { "epoch": 2.740360501567398, "grad_norm": 3.2574450969696045, "learning_rate": 2.261822655829849e-06, "loss": 0.5206, "step": 34967 }, { "epoch": 2.740438871473354, "grad_norm": 4.354404926300049, "learning_rate": 2.2615646492505206e-06, "loss": 1.1146, "step": 34968 }, { "epoch": 2.7405172413793104, "grad_norm": 3.3928496837615967, "learning_rate": 2.2613066530867922e-06, "loss": 1.314, "step": 34969 }, { "epoch": 2.7405956112852663, "grad_norm": 2.508349657058716, "learning_rate": 2.261048667339648e-06, "loss": 0.5832, "step": 34970 }, { "epoch": 2.7406739811912226, "grad_norm": 3.2493393421173096, "learning_rate": 2.26079069201007e-06, "loss": 1.4924, "step": 34971 }, { "epoch": 2.7407523510971785, "grad_norm": 5.1775665283203125, "learning_rate": 2.2605327270990355e-06, "loss": 1.4925, "step": 34972 }, { "epoch": 2.740830721003135, "grad_norm": 2.4035708904266357, "learning_rate": 2.2602747726075273e-06, "loss": 1.4197, "step": 34973 }, { "epoch": 2.740909090909091, "grad_norm": 2.5111985206604004, "learning_rate": 2.260016828536527e-06, "loss": 1.43, "step": 34974 }, { "epoch": 2.740987460815047, "grad_norm": 2.468867301940918, "learning_rate": 2.259758894887015e-06, "loss": 1.5151, "step": 34975 }, { "epoch": 2.741065830721003, "grad_norm": 3.196234941482544, "learning_rate": 2.259500971659973e-06, "loss": 1.2034, "step": 34976 }, { "epoch": 2.7411442006269593, "grad_norm": 2.5707013607025146, "learning_rate": 2.2592430588563813e-06, "loss": 1.5154, "step": 34977 }, { "epoch": 2.7412225705329156, "grad_norm": 5.037581443786621, "learning_rate": 2.2589851564772216e-06, "loss": 1.269, "step": 34978 }, { "epoch": 2.7413009404388715, "grad_norm": 3.0262348651885986, "learning_rate": 2.2587272645234743e-06, "loss": 1.8134, "step": 34979 }, { "epoch": 2.7413793103448274, "grad_norm": 4.236630439758301, "learning_rate": 2.2584693829961207e-06, "loss": 1.4161, "step": 34980 }, { "epoch": 2.7414576802507837, "grad_norm": 4.84347677230835, "learning_rate": 2.258211511896141e-06, "loss": 1.5362, "step": 34981 }, { "epoch": 2.74153605015674, "grad_norm": 2.338005542755127, "learning_rate": 2.2579536512245187e-06, "loss": 1.2963, "step": 34982 }, { "epoch": 2.741614420062696, "grad_norm": 5.077061176300049, "learning_rate": 2.2576958009822287e-06, "loss": 1.087, "step": 34983 }, { "epoch": 2.741692789968652, "grad_norm": 2.9766557216644287, "learning_rate": 2.2574379611702567e-06, "loss": 0.9649, "step": 34984 }, { "epoch": 2.741771159874608, "grad_norm": 8.772613525390625, "learning_rate": 2.257180131789582e-06, "loss": 1.6298, "step": 34985 }, { "epoch": 2.7418495297805645, "grad_norm": 3.2051539421081543, "learning_rate": 2.2569223128411876e-06, "loss": 0.858, "step": 34986 }, { "epoch": 2.7419278996865204, "grad_norm": 3.905256509780884, "learning_rate": 2.25666450432605e-06, "loss": 1.3536, "step": 34987 }, { "epoch": 2.7420062695924763, "grad_norm": 2.5699808597564697, "learning_rate": 2.2564067062451488e-06, "loss": 1.1832, "step": 34988 }, { "epoch": 2.7420846394984326, "grad_norm": 3.87378191947937, "learning_rate": 2.2561489185994716e-06, "loss": 1.1379, "step": 34989 }, { "epoch": 2.742163009404389, "grad_norm": 2.293647527694702, "learning_rate": 2.2558911413899933e-06, "loss": 1.4484, "step": 34990 }, { "epoch": 2.742241379310345, "grad_norm": 3.3341658115386963, "learning_rate": 2.255633374617695e-06, "loss": 0.9088, "step": 34991 }, { "epoch": 2.7423197492163007, "grad_norm": 3.5836446285247803, "learning_rate": 2.2553756182835573e-06, "loss": 1.419, "step": 34992 }, { "epoch": 2.742398119122257, "grad_norm": 2.831317901611328, "learning_rate": 2.255117872388562e-06, "loss": 1.4577, "step": 34993 }, { "epoch": 2.7424764890282134, "grad_norm": 3.6361045837402344, "learning_rate": 2.2548601369336883e-06, "loss": 1.03, "step": 34994 }, { "epoch": 2.7425548589341693, "grad_norm": 4.541049957275391, "learning_rate": 2.2546024119199163e-06, "loss": 0.9326, "step": 34995 }, { "epoch": 2.742633228840125, "grad_norm": 2.9130096435546875, "learning_rate": 2.2543446973482262e-06, "loss": 0.8564, "step": 34996 }, { "epoch": 2.7427115987460815, "grad_norm": 6.245009422302246, "learning_rate": 2.254086993219599e-06, "loss": 0.8007, "step": 34997 }, { "epoch": 2.742789968652038, "grad_norm": 2.391270160675049, "learning_rate": 2.253829299535015e-06, "loss": 1.3612, "step": 34998 }, { "epoch": 2.7428683385579937, "grad_norm": 8.467070579528809, "learning_rate": 2.253571616295453e-06, "loss": 1.129, "step": 34999 }, { "epoch": 2.7429467084639496, "grad_norm": 2.4588894844055176, "learning_rate": 2.253313943501896e-06, "loss": 1.3436, "step": 35000 }, { "epoch": 2.743025078369906, "grad_norm": 4.722951412200928, "learning_rate": 2.25305628115532e-06, "loss": 1.3113, "step": 35001 }, { "epoch": 2.7431034482758623, "grad_norm": 3.756293296813965, "learning_rate": 2.2527986292567055e-06, "loss": 1.5302, "step": 35002 }, { "epoch": 2.743181818181818, "grad_norm": 3.2405643463134766, "learning_rate": 2.252540987807036e-06, "loss": 1.257, "step": 35003 }, { "epoch": 2.743260188087774, "grad_norm": 5.440540790557861, "learning_rate": 2.2522833568072912e-06, "loss": 1.3983, "step": 35004 }, { "epoch": 2.7433385579937304, "grad_norm": 2.7783164978027344, "learning_rate": 2.252025736258447e-06, "loss": 1.5805, "step": 35005 }, { "epoch": 2.7434169278996867, "grad_norm": 2.4056010246276855, "learning_rate": 2.2517681261614855e-06, "loss": 1.1563, "step": 35006 }, { "epoch": 2.7434952978056426, "grad_norm": 2.2804348468780518, "learning_rate": 2.251510526517385e-06, "loss": 1.2471, "step": 35007 }, { "epoch": 2.7435736677115985, "grad_norm": 2.0656204223632812, "learning_rate": 2.25125293732713e-06, "loss": 1.183, "step": 35008 }, { "epoch": 2.743652037617555, "grad_norm": 2.7567436695098877, "learning_rate": 2.250995358591695e-06, "loss": 1.1338, "step": 35009 }, { "epoch": 2.743730407523511, "grad_norm": 2.84464430809021, "learning_rate": 2.250737790312062e-06, "loss": 1.2818, "step": 35010 }, { "epoch": 2.743808777429467, "grad_norm": 9.331832885742188, "learning_rate": 2.25048023248921e-06, "loss": 1.0226, "step": 35011 }, { "epoch": 2.743887147335423, "grad_norm": 2.454336643218994, "learning_rate": 2.2502226851241192e-06, "loss": 0.7723, "step": 35012 }, { "epoch": 2.7439655172413793, "grad_norm": 4.024803161621094, "learning_rate": 2.249965148217769e-06, "loss": 1.4553, "step": 35013 }, { "epoch": 2.7440438871473356, "grad_norm": 6.511985778808594, "learning_rate": 2.249707621771139e-06, "loss": 1.5476, "step": 35014 }, { "epoch": 2.7441222570532915, "grad_norm": 2.85239839553833, "learning_rate": 2.24945010578521e-06, "loss": 0.9404, "step": 35015 }, { "epoch": 2.7442006269592474, "grad_norm": 3.6950557231903076, "learning_rate": 2.2491926002609567e-06, "loss": 1.243, "step": 35016 }, { "epoch": 2.7442789968652037, "grad_norm": 5.716235160827637, "learning_rate": 2.2489351051993635e-06, "loss": 1.2351, "step": 35017 }, { "epoch": 2.74435736677116, "grad_norm": 2.624006986618042, "learning_rate": 2.248677620601408e-06, "loss": 1.372, "step": 35018 }, { "epoch": 2.744435736677116, "grad_norm": 4.018444061279297, "learning_rate": 2.2484201464680716e-06, "loss": 1.4884, "step": 35019 }, { "epoch": 2.7445141065830723, "grad_norm": 2.5041840076446533, "learning_rate": 2.24816268280033e-06, "loss": 1.4444, "step": 35020 }, { "epoch": 2.744592476489028, "grad_norm": 2.8426055908203125, "learning_rate": 2.2479052295991622e-06, "loss": 0.9824, "step": 35021 }, { "epoch": 2.7446708463949845, "grad_norm": 8.643330574035645, "learning_rate": 2.247647786865551e-06, "loss": 1.56, "step": 35022 }, { "epoch": 2.7447492163009404, "grad_norm": 3.8852524757385254, "learning_rate": 2.247390354600476e-06, "loss": 1.4204, "step": 35023 }, { "epoch": 2.7448275862068967, "grad_norm": 3.2195446491241455, "learning_rate": 2.247132932804912e-06, "loss": 1.2623, "step": 35024 }, { "epoch": 2.7449059561128526, "grad_norm": 11.93088436126709, "learning_rate": 2.246875521479841e-06, "loss": 1.1459, "step": 35025 }, { "epoch": 2.744984326018809, "grad_norm": 3.4860002994537354, "learning_rate": 2.2466181206262393e-06, "loss": 1.3169, "step": 35026 }, { "epoch": 2.745062695924765, "grad_norm": 5.086780548095703, "learning_rate": 2.246360730245092e-06, "loss": 1.6605, "step": 35027 }, { "epoch": 2.745141065830721, "grad_norm": 6.4363555908203125, "learning_rate": 2.246103350337372e-06, "loss": 1.3493, "step": 35028 }, { "epoch": 2.745219435736677, "grad_norm": 3.222153663635254, "learning_rate": 2.24584598090406e-06, "loss": 1.1715, "step": 35029 }, { "epoch": 2.7452978056426334, "grad_norm": 3.6784181594848633, "learning_rate": 2.245588621946136e-06, "loss": 1.0082, "step": 35030 }, { "epoch": 2.7453761755485893, "grad_norm": 9.07214641571045, "learning_rate": 2.245331273464578e-06, "loss": 1.2285, "step": 35031 }, { "epoch": 2.7454545454545456, "grad_norm": 4.699348449707031, "learning_rate": 2.245073935460364e-06, "loss": 0.9334, "step": 35032 }, { "epoch": 2.7455329153605015, "grad_norm": 2.497457265853882, "learning_rate": 2.2448166079344746e-06, "loss": 1.0537, "step": 35033 }, { "epoch": 2.745611285266458, "grad_norm": 2.864417552947998, "learning_rate": 2.244559290887888e-06, "loss": 1.2678, "step": 35034 }, { "epoch": 2.7456896551724137, "grad_norm": 3.743473768234253, "learning_rate": 2.2443019843215814e-06, "loss": 1.5217, "step": 35035 }, { "epoch": 2.74576802507837, "grad_norm": 2.928921699523926, "learning_rate": 2.244044688236535e-06, "loss": 1.3279, "step": 35036 }, { "epoch": 2.745846394984326, "grad_norm": 5.536856651306152, "learning_rate": 2.243787402633728e-06, "loss": 1.6564, "step": 35037 }, { "epoch": 2.7459247648902823, "grad_norm": 8.178258895874023, "learning_rate": 2.243530127514139e-06, "loss": 1.4186, "step": 35038 }, { "epoch": 2.746003134796238, "grad_norm": 2.746082305908203, "learning_rate": 2.2432728628787444e-06, "loss": 1.5909, "step": 35039 }, { "epoch": 2.7460815047021945, "grad_norm": 1.83157217502594, "learning_rate": 2.2430156087285214e-06, "loss": 1.3141, "step": 35040 }, { "epoch": 2.7461598746081504, "grad_norm": 5.2752685546875, "learning_rate": 2.242758365064453e-06, "loss": 1.2395, "step": 35041 }, { "epoch": 2.7462382445141067, "grad_norm": 2.6328630447387695, "learning_rate": 2.2425011318875174e-06, "loss": 1.408, "step": 35042 }, { "epoch": 2.7463166144200626, "grad_norm": 4.63327169418335, "learning_rate": 2.242243909198689e-06, "loss": 1.3356, "step": 35043 }, { "epoch": 2.746394984326019, "grad_norm": 19.787860870361328, "learning_rate": 2.241986696998949e-06, "loss": 1.5024, "step": 35044 }, { "epoch": 2.746473354231975, "grad_norm": 7.6333537101745605, "learning_rate": 2.2417294952892745e-06, "loss": 1.3332, "step": 35045 }, { "epoch": 2.746551724137931, "grad_norm": 4.331971645355225, "learning_rate": 2.241472304070644e-06, "loss": 1.2886, "step": 35046 }, { "epoch": 2.746630094043887, "grad_norm": 2.364764928817749, "learning_rate": 2.241215123344037e-06, "loss": 1.2801, "step": 35047 }, { "epoch": 2.7467084639498434, "grad_norm": 3.1731042861938477, "learning_rate": 2.24095795311043e-06, "loss": 1.2682, "step": 35048 }, { "epoch": 2.7467868338557992, "grad_norm": 2.2021918296813965, "learning_rate": 2.240700793370802e-06, "loss": 0.7555, "step": 35049 }, { "epoch": 2.7468652037617556, "grad_norm": 4.118297100067139, "learning_rate": 2.2404436441261305e-06, "loss": 0.9134, "step": 35050 }, { "epoch": 2.7469435736677115, "grad_norm": 3.1529524326324463, "learning_rate": 2.240186505377394e-06, "loss": 1.489, "step": 35051 }, { "epoch": 2.747021943573668, "grad_norm": 3.263296127319336, "learning_rate": 2.239929377125571e-06, "loss": 1.7263, "step": 35052 }, { "epoch": 2.7471003134796237, "grad_norm": 2.8041858673095703, "learning_rate": 2.2396722593716393e-06, "loss": 1.5105, "step": 35053 }, { "epoch": 2.74717868338558, "grad_norm": 9.44167423248291, "learning_rate": 2.239415152116576e-06, "loss": 1.6939, "step": 35054 }, { "epoch": 2.747257053291536, "grad_norm": 3.490190267562866, "learning_rate": 2.239158055361359e-06, "loss": 1.2811, "step": 35055 }, { "epoch": 2.7473354231974922, "grad_norm": 3.5698859691619873, "learning_rate": 2.238900969106969e-06, "loss": 1.5711, "step": 35056 }, { "epoch": 2.747413793103448, "grad_norm": 3.885725259780884, "learning_rate": 2.23864389335438e-06, "loss": 1.4684, "step": 35057 }, { "epoch": 2.7474921630094045, "grad_norm": 3.745256185531616, "learning_rate": 2.238386828104569e-06, "loss": 1.6846, "step": 35058 }, { "epoch": 2.7475705329153604, "grad_norm": 2.2230629920959473, "learning_rate": 2.2381297733585183e-06, "loss": 1.1772, "step": 35059 }, { "epoch": 2.7476489028213167, "grad_norm": 2.4403927326202393, "learning_rate": 2.2378727291172047e-06, "loss": 0.9894, "step": 35060 }, { "epoch": 2.7477272727272726, "grad_norm": 6.184692859649658, "learning_rate": 2.237615695381603e-06, "loss": 1.8284, "step": 35061 }, { "epoch": 2.747805642633229, "grad_norm": 2.558990478515625, "learning_rate": 2.2373586721526925e-06, "loss": 1.3151, "step": 35062 }, { "epoch": 2.747884012539185, "grad_norm": 9.101106643676758, "learning_rate": 2.237101659431449e-06, "loss": 0.7106, "step": 35063 }, { "epoch": 2.747962382445141, "grad_norm": 3.3839168548583984, "learning_rate": 2.2368446572188547e-06, "loss": 1.323, "step": 35064 }, { "epoch": 2.748040752351097, "grad_norm": 2.4794206619262695, "learning_rate": 2.2365876655158824e-06, "loss": 1.1322, "step": 35065 }, { "epoch": 2.7481191222570533, "grad_norm": 2.5476043224334717, "learning_rate": 2.236330684323511e-06, "loss": 1.4563, "step": 35066 }, { "epoch": 2.7481974921630092, "grad_norm": 11.239192962646484, "learning_rate": 2.236073713642719e-06, "loss": 0.7828, "step": 35067 }, { "epoch": 2.7482758620689656, "grad_norm": 2.173978328704834, "learning_rate": 2.2358167534744824e-06, "loss": 1.3244, "step": 35068 }, { "epoch": 2.748354231974922, "grad_norm": 2.869353771209717, "learning_rate": 2.235559803819779e-06, "loss": 1.2218, "step": 35069 }, { "epoch": 2.748432601880878, "grad_norm": 4.8305745124816895, "learning_rate": 2.2353028646795866e-06, "loss": 1.5486, "step": 35070 }, { "epoch": 2.7485109717868337, "grad_norm": 4.069547176361084, "learning_rate": 2.235045936054884e-06, "loss": 1.331, "step": 35071 }, { "epoch": 2.74858934169279, "grad_norm": 2.846452474594116, "learning_rate": 2.234789017946642e-06, "loss": 1.5474, "step": 35072 }, { "epoch": 2.7486677115987463, "grad_norm": 2.118088483810425, "learning_rate": 2.2345321103558446e-06, "loss": 1.1919, "step": 35073 }, { "epoch": 2.7487460815047022, "grad_norm": 2.827355146408081, "learning_rate": 2.2342752132834665e-06, "loss": 1.3328, "step": 35074 }, { "epoch": 2.748824451410658, "grad_norm": 3.543478012084961, "learning_rate": 2.234018326730487e-06, "loss": 1.1504, "step": 35075 }, { "epoch": 2.7489028213166145, "grad_norm": 4.976641654968262, "learning_rate": 2.2337614506978787e-06, "loss": 1.2077, "step": 35076 }, { "epoch": 2.748981191222571, "grad_norm": 3.73235821723938, "learning_rate": 2.23350458518662e-06, "loss": 0.9802, "step": 35077 }, { "epoch": 2.7490595611285267, "grad_norm": 2.178445339202881, "learning_rate": 2.23324773019769e-06, "loss": 1.1486, "step": 35078 }, { "epoch": 2.7491379310344826, "grad_norm": 2.6228857040405273, "learning_rate": 2.232990885732067e-06, "loss": 0.9938, "step": 35079 }, { "epoch": 2.749216300940439, "grad_norm": 6.625154972076416, "learning_rate": 2.2327340517907232e-06, "loss": 1.5704, "step": 35080 }, { "epoch": 2.7492946708463952, "grad_norm": 3.5972883701324463, "learning_rate": 2.232477228374638e-06, "loss": 1.3426, "step": 35081 }, { "epoch": 2.749373040752351, "grad_norm": 4.506078243255615, "learning_rate": 2.232220415484786e-06, "loss": 0.9977, "step": 35082 }, { "epoch": 2.749451410658307, "grad_norm": 4.5343475341796875, "learning_rate": 2.2319636131221494e-06, "loss": 1.0789, "step": 35083 }, { "epoch": 2.7495297805642633, "grad_norm": 2.3143882751464844, "learning_rate": 2.2317068212876996e-06, "loss": 1.3222, "step": 35084 }, { "epoch": 2.7496081504702197, "grad_norm": 4.929437160491943, "learning_rate": 2.2314500399824154e-06, "loss": 1.4142, "step": 35085 }, { "epoch": 2.7496865203761756, "grad_norm": 2.593268394470215, "learning_rate": 2.231193269207273e-06, "loss": 1.3459, "step": 35086 }, { "epoch": 2.7497648902821314, "grad_norm": 5.370439052581787, "learning_rate": 2.2309365089632497e-06, "loss": 1.602, "step": 35087 }, { "epoch": 2.749843260188088, "grad_norm": 2.77463436126709, "learning_rate": 2.2306797592513215e-06, "loss": 1.3861, "step": 35088 }, { "epoch": 2.749921630094044, "grad_norm": 3.382974147796631, "learning_rate": 2.230423020072465e-06, "loss": 1.2232, "step": 35089 }, { "epoch": 2.75, "grad_norm": 2.7370107173919678, "learning_rate": 2.2301662914276577e-06, "loss": 1.1538, "step": 35090 }, { "epoch": 2.75, "eval_loss": 1.2529820203781128, "eval_runtime": 109.2956, "eval_samples_per_second": 24.585, "eval_steps_per_second": 6.148, "step": 35090 }, { "epoch": 2.750078369905956, "grad_norm": 3.585789442062378, "learning_rate": 2.2299095733178728e-06, "loss": 0.9931, "step": 35091 }, { "epoch": 2.750156739811912, "grad_norm": 2.587127208709717, "learning_rate": 2.22965286574409e-06, "loss": 1.4839, "step": 35092 }, { "epoch": 2.7502351097178686, "grad_norm": 4.57494592666626, "learning_rate": 2.2293961687072855e-06, "loss": 1.1371, "step": 35093 }, { "epoch": 2.7503134796238244, "grad_norm": 2.8472800254821777, "learning_rate": 2.229139482208436e-06, "loss": 1.0666, "step": 35094 }, { "epoch": 2.7503918495297803, "grad_norm": 2.635749578475952, "learning_rate": 2.2288828062485157e-06, "loss": 1.2207, "step": 35095 }, { "epoch": 2.7504702194357367, "grad_norm": 4.565820217132568, "learning_rate": 2.2286261408285e-06, "loss": 0.8939, "step": 35096 }, { "epoch": 2.750548589341693, "grad_norm": 2.8781538009643555, "learning_rate": 2.22836948594937e-06, "loss": 1.3168, "step": 35097 }, { "epoch": 2.750626959247649, "grad_norm": 3.916638135910034, "learning_rate": 2.228112841612097e-06, "loss": 1.3718, "step": 35098 }, { "epoch": 2.7507053291536048, "grad_norm": 2.437497138977051, "learning_rate": 2.2278562078176595e-06, "loss": 1.5258, "step": 35099 }, { "epoch": 2.750783699059561, "grad_norm": 4.984011650085449, "learning_rate": 2.227599584567033e-06, "loss": 1.605, "step": 35100 }, { "epoch": 2.7508620689655174, "grad_norm": 2.3477871417999268, "learning_rate": 2.227342971861194e-06, "loss": 1.2439, "step": 35101 }, { "epoch": 2.7509404388714733, "grad_norm": 5.417826175689697, "learning_rate": 2.2270863697011174e-06, "loss": 1.0898, "step": 35102 }, { "epoch": 2.751018808777429, "grad_norm": 3.1841366291046143, "learning_rate": 2.226829778087781e-06, "loss": 1.5111, "step": 35103 }, { "epoch": 2.7510971786833855, "grad_norm": 2.759422779083252, "learning_rate": 2.226573197022159e-06, "loss": 0.7933, "step": 35104 }, { "epoch": 2.751175548589342, "grad_norm": 8.008984565734863, "learning_rate": 2.2263166265052278e-06, "loss": 1.5162, "step": 35105 }, { "epoch": 2.7512539184952978, "grad_norm": 3.6136670112609863, "learning_rate": 2.226060066537964e-06, "loss": 1.3122, "step": 35106 }, { "epoch": 2.7513322884012537, "grad_norm": 3.559072971343994, "learning_rate": 2.225803517121343e-06, "loss": 1.225, "step": 35107 }, { "epoch": 2.75141065830721, "grad_norm": 2.410787582397461, "learning_rate": 2.2255469782563406e-06, "loss": 1.3849, "step": 35108 }, { "epoch": 2.7514890282131663, "grad_norm": 3.584700345993042, "learning_rate": 2.2252904499439336e-06, "loss": 1.7223, "step": 35109 }, { "epoch": 2.751567398119122, "grad_norm": 2.911952257156372, "learning_rate": 2.2250339321850934e-06, "loss": 1.3499, "step": 35110 }, { "epoch": 2.751645768025078, "grad_norm": 5.7062835693359375, "learning_rate": 2.2247774249808008e-06, "loss": 1.2188, "step": 35111 }, { "epoch": 2.7517241379310344, "grad_norm": 3.023514747619629, "learning_rate": 2.2245209283320307e-06, "loss": 1.6321, "step": 35112 }, { "epoch": 2.7518025078369908, "grad_norm": 4.157970428466797, "learning_rate": 2.2242644422397556e-06, "loss": 1.0695, "step": 35113 }, { "epoch": 2.7518808777429467, "grad_norm": 3.780057907104492, "learning_rate": 2.2240079667049534e-06, "loss": 1.6928, "step": 35114 }, { "epoch": 2.751959247648903, "grad_norm": 4.386176109313965, "learning_rate": 2.223751501728597e-06, "loss": 1.3294, "step": 35115 }, { "epoch": 2.752037617554859, "grad_norm": 3.8793013095855713, "learning_rate": 2.223495047311668e-06, "loss": 1.2848, "step": 35116 }, { "epoch": 2.752115987460815, "grad_norm": 3.4925968647003174, "learning_rate": 2.223238603455135e-06, "loss": 1.6182, "step": 35117 }, { "epoch": 2.752194357366771, "grad_norm": 3.334148406982422, "learning_rate": 2.222982170159977e-06, "loss": 1.4519, "step": 35118 }, { "epoch": 2.7522727272727274, "grad_norm": 3.423300266265869, "learning_rate": 2.222725747427168e-06, "loss": 1.4638, "step": 35119 }, { "epoch": 2.7523510971786833, "grad_norm": 3.9833240509033203, "learning_rate": 2.2224693352576837e-06, "loss": 1.0313, "step": 35120 }, { "epoch": 2.7524294670846396, "grad_norm": 2.0093493461608887, "learning_rate": 2.2222129336524994e-06, "loss": 1.2941, "step": 35121 }, { "epoch": 2.7525078369905955, "grad_norm": 5.439508438110352, "learning_rate": 2.2219565426125905e-06, "loss": 1.5803, "step": 35122 }, { "epoch": 2.752586206896552, "grad_norm": 2.16109037399292, "learning_rate": 2.221700162138932e-06, "loss": 1.4854, "step": 35123 }, { "epoch": 2.7526645768025078, "grad_norm": 11.536651611328125, "learning_rate": 2.2214437922324987e-06, "loss": 0.8877, "step": 35124 }, { "epoch": 2.752742946708464, "grad_norm": 2.4408230781555176, "learning_rate": 2.2211874328942666e-06, "loss": 1.3205, "step": 35125 }, { "epoch": 2.75282131661442, "grad_norm": 2.9521517753601074, "learning_rate": 2.2209310841252096e-06, "loss": 1.5723, "step": 35126 }, { "epoch": 2.7528996865203763, "grad_norm": 3.809886932373047, "learning_rate": 2.2206747459263054e-06, "loss": 0.9921, "step": 35127 }, { "epoch": 2.752978056426332, "grad_norm": 3.0446386337280273, "learning_rate": 2.2204184182985254e-06, "loss": 1.3508, "step": 35128 }, { "epoch": 2.7530564263322885, "grad_norm": 2.151796579360962, "learning_rate": 2.2201621012428443e-06, "loss": 1.1647, "step": 35129 }, { "epoch": 2.7531347962382444, "grad_norm": 2.83363938331604, "learning_rate": 2.21990579476024e-06, "loss": 1.4887, "step": 35130 }, { "epoch": 2.7532131661442008, "grad_norm": 3.1977906227111816, "learning_rate": 2.2196494988516887e-06, "loss": 1.3275, "step": 35131 }, { "epoch": 2.7532915360501566, "grad_norm": 6.515908718109131, "learning_rate": 2.2193932135181607e-06, "loss": 1.6284, "step": 35132 }, { "epoch": 2.753369905956113, "grad_norm": 2.7576746940612793, "learning_rate": 2.2191369387606322e-06, "loss": 1.1844, "step": 35133 }, { "epoch": 2.753448275862069, "grad_norm": 6.151268482208252, "learning_rate": 2.2188806745800766e-06, "loss": 1.4235, "step": 35134 }, { "epoch": 2.753526645768025, "grad_norm": 2.391524314880371, "learning_rate": 2.218624420977474e-06, "loss": 1.2457, "step": 35135 }, { "epoch": 2.753605015673981, "grad_norm": 5.985590934753418, "learning_rate": 2.2183681779537946e-06, "loss": 1.622, "step": 35136 }, { "epoch": 2.7536833855799374, "grad_norm": 2.576716184616089, "learning_rate": 2.218111945510013e-06, "loss": 1.4243, "step": 35137 }, { "epoch": 2.7537617554858933, "grad_norm": 2.2759954929351807, "learning_rate": 2.2178557236471048e-06, "loss": 1.2119, "step": 35138 }, { "epoch": 2.7538401253918496, "grad_norm": 3.804107666015625, "learning_rate": 2.217599512366045e-06, "loss": 1.3565, "step": 35139 }, { "epoch": 2.7539184952978055, "grad_norm": 3.8005640506744385, "learning_rate": 2.217343311667807e-06, "loss": 1.1383, "step": 35140 }, { "epoch": 2.753996865203762, "grad_norm": 2.8253307342529297, "learning_rate": 2.2170871215533663e-06, "loss": 1.2926, "step": 35141 }, { "epoch": 2.7540752351097177, "grad_norm": 5.442862510681152, "learning_rate": 2.216830942023696e-06, "loss": 1.194, "step": 35142 }, { "epoch": 2.754153605015674, "grad_norm": 3.9010355472564697, "learning_rate": 2.2165747730797714e-06, "loss": 1.3752, "step": 35143 }, { "epoch": 2.75423197492163, "grad_norm": 2.2930476665496826, "learning_rate": 2.216318614722567e-06, "loss": 1.1949, "step": 35144 }, { "epoch": 2.7543103448275863, "grad_norm": 9.095196723937988, "learning_rate": 2.216062466953057e-06, "loss": 2.0174, "step": 35145 }, { "epoch": 2.754388714733542, "grad_norm": 4.010598182678223, "learning_rate": 2.215806329772217e-06, "loss": 1.6474, "step": 35146 }, { "epoch": 2.7544670846394985, "grad_norm": 2.549877643585205, "learning_rate": 2.215550203181016e-06, "loss": 1.4473, "step": 35147 }, { "epoch": 2.7545454545454544, "grad_norm": 3.126643657684326, "learning_rate": 2.215294087180434e-06, "loss": 1.2109, "step": 35148 }, { "epoch": 2.7546238244514107, "grad_norm": 3.5635406970977783, "learning_rate": 2.215037981771443e-06, "loss": 1.0411, "step": 35149 }, { "epoch": 2.7547021943573666, "grad_norm": 2.31026291847229, "learning_rate": 2.2147818869550187e-06, "loss": 1.2356, "step": 35150 }, { "epoch": 2.754780564263323, "grad_norm": 2.9493539333343506, "learning_rate": 2.2145258027321316e-06, "loss": 1.3666, "step": 35151 }, { "epoch": 2.754858934169279, "grad_norm": 15.188570022583008, "learning_rate": 2.214269729103757e-06, "loss": 0.6493, "step": 35152 }, { "epoch": 2.754937304075235, "grad_norm": 3.8393046855926514, "learning_rate": 2.2140136660708705e-06, "loss": 1.6762, "step": 35153 }, { "epoch": 2.755015673981191, "grad_norm": 4.388744354248047, "learning_rate": 2.213757613634445e-06, "loss": 1.2203, "step": 35154 }, { "epoch": 2.7550940438871474, "grad_norm": 2.925493001937866, "learning_rate": 2.213501571795454e-06, "loss": 1.6078, "step": 35155 }, { "epoch": 2.7551724137931033, "grad_norm": 7.311062335968018, "learning_rate": 2.213245540554872e-06, "loss": 0.7192, "step": 35156 }, { "epoch": 2.7552507836990596, "grad_norm": 9.098739624023438, "learning_rate": 2.2129895199136724e-06, "loss": 1.5771, "step": 35157 }, { "epoch": 2.7553291536050155, "grad_norm": 4.874051570892334, "learning_rate": 2.2127335098728292e-06, "loss": 1.3858, "step": 35158 }, { "epoch": 2.755407523510972, "grad_norm": 4.583592414855957, "learning_rate": 2.212477510433316e-06, "loss": 1.397, "step": 35159 }, { "epoch": 2.7554858934169277, "grad_norm": 2.7446792125701904, "learning_rate": 2.2122215215961073e-06, "loss": 0.6228, "step": 35160 }, { "epoch": 2.755564263322884, "grad_norm": 3.0579211711883545, "learning_rate": 2.211965543362175e-06, "loss": 1.2041, "step": 35161 }, { "epoch": 2.75564263322884, "grad_norm": 3.610995054244995, "learning_rate": 2.2117095757324947e-06, "loss": 1.5769, "step": 35162 }, { "epoch": 2.7557210031347963, "grad_norm": 4.484619140625, "learning_rate": 2.2114536187080382e-06, "loss": 1.4815, "step": 35163 }, { "epoch": 2.7557993730407526, "grad_norm": 3.6985888481140137, "learning_rate": 2.2111976722897803e-06, "loss": 1.3636, "step": 35164 }, { "epoch": 2.7558777429467085, "grad_norm": 5.704204559326172, "learning_rate": 2.210941736478696e-06, "loss": 1.5938, "step": 35165 }, { "epoch": 2.7559561128526644, "grad_norm": 4.140050411224365, "learning_rate": 2.210685811275753e-06, "loss": 1.2932, "step": 35166 }, { "epoch": 2.7560344827586207, "grad_norm": 2.2318477630615234, "learning_rate": 2.2104298966819306e-06, "loss": 1.1495, "step": 35167 }, { "epoch": 2.756112852664577, "grad_norm": 3.8185911178588867, "learning_rate": 2.2101739926982013e-06, "loss": 1.7325, "step": 35168 }, { "epoch": 2.756191222570533, "grad_norm": 11.054170608520508, "learning_rate": 2.2099180993255354e-06, "loss": 1.2722, "step": 35169 }, { "epoch": 2.756269592476489, "grad_norm": 2.8790102005004883, "learning_rate": 2.2096622165649082e-06, "loss": 1.5177, "step": 35170 }, { "epoch": 2.756347962382445, "grad_norm": 2.4529690742492676, "learning_rate": 2.209406344417291e-06, "loss": 1.4271, "step": 35171 }, { "epoch": 2.7564263322884015, "grad_norm": 2.724076986312866, "learning_rate": 2.2091504828836617e-06, "loss": 1.5278, "step": 35172 }, { "epoch": 2.7565047021943574, "grad_norm": 3.3056299686431885, "learning_rate": 2.2088946319649897e-06, "loss": 0.8859, "step": 35173 }, { "epoch": 2.7565830721003133, "grad_norm": 3.9327585697174072, "learning_rate": 2.208638791662248e-06, "loss": 1.1577, "step": 35174 }, { "epoch": 2.7566614420062696, "grad_norm": 3.6985795497894287, "learning_rate": 2.2083829619764115e-06, "loss": 1.7214, "step": 35175 }, { "epoch": 2.756739811912226, "grad_norm": 2.529639959335327, "learning_rate": 2.208127142908452e-06, "loss": 1.1716, "step": 35176 }, { "epoch": 2.756818181818182, "grad_norm": 4.60554313659668, "learning_rate": 2.2078713344593422e-06, "loss": 1.3475, "step": 35177 }, { "epoch": 2.7568965517241377, "grad_norm": 4.341932773590088, "learning_rate": 2.207615536630057e-06, "loss": 1.5481, "step": 35178 }, { "epoch": 2.756974921630094, "grad_norm": 6.926056385040283, "learning_rate": 2.2073597494215664e-06, "loss": 1.3954, "step": 35179 }, { "epoch": 2.7570532915360504, "grad_norm": 1.7956761121749878, "learning_rate": 2.207103972834846e-06, "loss": 1.3609, "step": 35180 }, { "epoch": 2.7571316614420063, "grad_norm": 6.40924072265625, "learning_rate": 2.2068482068708675e-06, "loss": 1.2676, "step": 35181 }, { "epoch": 2.757210031347962, "grad_norm": 7.514235973358154, "learning_rate": 2.2065924515306035e-06, "loss": 1.3164, "step": 35182 }, { "epoch": 2.7572884012539185, "grad_norm": 10.034401893615723, "learning_rate": 2.2063367068150294e-06, "loss": 0.7842, "step": 35183 }, { "epoch": 2.757366771159875, "grad_norm": 4.381589412689209, "learning_rate": 2.206080972725113e-06, "loss": 1.5581, "step": 35184 }, { "epoch": 2.7574451410658307, "grad_norm": 3.218275785446167, "learning_rate": 2.205825249261828e-06, "loss": 1.3924, "step": 35185 }, { "epoch": 2.7575235109717866, "grad_norm": 4.609912395477295, "learning_rate": 2.2055695364261505e-06, "loss": 1.1389, "step": 35186 }, { "epoch": 2.757601880877743, "grad_norm": 2.8852593898773193, "learning_rate": 2.2053138342190528e-06, "loss": 1.1257, "step": 35187 }, { "epoch": 2.7576802507836993, "grad_norm": 3.1356441974639893, "learning_rate": 2.205058142641504e-06, "loss": 1.0087, "step": 35188 }, { "epoch": 2.757758620689655, "grad_norm": 8.796723365783691, "learning_rate": 2.2048024616944785e-06, "loss": 1.5994, "step": 35189 }, { "epoch": 2.757836990595611, "grad_norm": 12.776665687561035, "learning_rate": 2.204546791378947e-06, "loss": 1.2974, "step": 35190 }, { "epoch": 2.7579153605015674, "grad_norm": 7.365643501281738, "learning_rate": 2.2042911316958875e-06, "loss": 1.7184, "step": 35191 }, { "epoch": 2.7579937304075237, "grad_norm": 2.230771541595459, "learning_rate": 2.204035482646267e-06, "loss": 1.3183, "step": 35192 }, { "epoch": 2.7580721003134796, "grad_norm": 2.942331314086914, "learning_rate": 2.203779844231059e-06, "loss": 1.2534, "step": 35193 }, { "epoch": 2.7581504702194355, "grad_norm": 5.647455215454102, "learning_rate": 2.203524216451236e-06, "loss": 0.8578, "step": 35194 }, { "epoch": 2.758228840125392, "grad_norm": 3.613912343978882, "learning_rate": 2.2032685993077705e-06, "loss": 1.3946, "step": 35195 }, { "epoch": 2.758307210031348, "grad_norm": 2.705092668533325, "learning_rate": 2.2030129928016346e-06, "loss": 1.128, "step": 35196 }, { "epoch": 2.758385579937304, "grad_norm": 2.6949923038482666, "learning_rate": 2.2027573969338012e-06, "loss": 0.8657, "step": 35197 }, { "epoch": 2.75846394984326, "grad_norm": 2.8362367153167725, "learning_rate": 2.202501811705244e-06, "loss": 1.4024, "step": 35198 }, { "epoch": 2.7585423197492163, "grad_norm": 2.855926036834717, "learning_rate": 2.202246237116929e-06, "loss": 1.3028, "step": 35199 }, { "epoch": 2.7586206896551726, "grad_norm": 4.248404502868652, "learning_rate": 2.2019906731698337e-06, "loss": 1.0671, "step": 35200 }, { "epoch": 2.7586990595611285, "grad_norm": 6.44842004776001, "learning_rate": 2.2017351198649295e-06, "loss": 1.5141, "step": 35201 }, { "epoch": 2.7587774294670844, "grad_norm": 2.776843309402466, "learning_rate": 2.2014795772031887e-06, "loss": 1.3408, "step": 35202 }, { "epoch": 2.7588557993730407, "grad_norm": 3.8784217834472656, "learning_rate": 2.201224045185581e-06, "loss": 1.4523, "step": 35203 }, { "epoch": 2.758934169278997, "grad_norm": 3.9133405685424805, "learning_rate": 2.2009685238130776e-06, "loss": 1.1931, "step": 35204 }, { "epoch": 2.759012539184953, "grad_norm": 5.137957572937012, "learning_rate": 2.200713013086653e-06, "loss": 1.0472, "step": 35205 }, { "epoch": 2.759090909090909, "grad_norm": 7.405311584472656, "learning_rate": 2.2004575130072814e-06, "loss": 1.8163, "step": 35206 }, { "epoch": 2.759169278996865, "grad_norm": 3.207118272781372, "learning_rate": 2.200202023575929e-06, "loss": 1.3631, "step": 35207 }, { "epoch": 2.7592476489028215, "grad_norm": 3.4022226333618164, "learning_rate": 2.19994654479357e-06, "loss": 1.4765, "step": 35208 }, { "epoch": 2.7593260188087774, "grad_norm": 4.9344482421875, "learning_rate": 2.1996910766611767e-06, "loss": 1.1424, "step": 35209 }, { "epoch": 2.7594043887147337, "grad_norm": 4.514271259307861, "learning_rate": 2.19943561917972e-06, "loss": 1.0405, "step": 35210 }, { "epoch": 2.7594827586206896, "grad_norm": 2.8186020851135254, "learning_rate": 2.1991801723501716e-06, "loss": 1.3726, "step": 35211 }, { "epoch": 2.759561128526646, "grad_norm": 3.718576669692993, "learning_rate": 2.1989247361735034e-06, "loss": 1.4934, "step": 35212 }, { "epoch": 2.759639498432602, "grad_norm": 2.3334739208221436, "learning_rate": 2.198669310650687e-06, "loss": 1.3635, "step": 35213 }, { "epoch": 2.759717868338558, "grad_norm": 4.315317153930664, "learning_rate": 2.1984138957826933e-06, "loss": 1.5414, "step": 35214 }, { "epoch": 2.759796238244514, "grad_norm": 5.082200527191162, "learning_rate": 2.198158491570494e-06, "loss": 0.6484, "step": 35215 }, { "epoch": 2.7598746081504704, "grad_norm": 2.420944929122925, "learning_rate": 2.1979030980150613e-06, "loss": 1.2215, "step": 35216 }, { "epoch": 2.7599529780564263, "grad_norm": 4.469784259796143, "learning_rate": 2.197647715117367e-06, "loss": 1.4831, "step": 35217 }, { "epoch": 2.7600313479623826, "grad_norm": 6.5160231590271, "learning_rate": 2.197392342878378e-06, "loss": 1.6832, "step": 35218 }, { "epoch": 2.7601097178683385, "grad_norm": 3.651076555252075, "learning_rate": 2.197136981299071e-06, "loss": 0.9231, "step": 35219 }, { "epoch": 2.760188087774295, "grad_norm": 2.910980224609375, "learning_rate": 2.1968816303804158e-06, "loss": 1.2555, "step": 35220 }, { "epoch": 2.7602664576802507, "grad_norm": 2.551807403564453, "learning_rate": 2.196626290123384e-06, "loss": 1.2767, "step": 35221 }, { "epoch": 2.760344827586207, "grad_norm": 2.977900743484497, "learning_rate": 2.1963709605289445e-06, "loss": 1.3, "step": 35222 }, { "epoch": 2.760423197492163, "grad_norm": 3.073002815246582, "learning_rate": 2.1961156415980683e-06, "loss": 1.4516, "step": 35223 }, { "epoch": 2.7605015673981192, "grad_norm": 4.080654621124268, "learning_rate": 2.195860333331731e-06, "loss": 1.598, "step": 35224 }, { "epoch": 2.760579937304075, "grad_norm": 5.366726398468018, "learning_rate": 2.1956050357308996e-06, "loss": 1.4627, "step": 35225 }, { "epoch": 2.7606583072100315, "grad_norm": 2.582892417907715, "learning_rate": 2.195349748796546e-06, "loss": 1.2798, "step": 35226 }, { "epoch": 2.7607366771159874, "grad_norm": 2.8986356258392334, "learning_rate": 2.1950944725296406e-06, "loss": 1.2754, "step": 35227 }, { "epoch": 2.7608150470219437, "grad_norm": 3.8236303329467773, "learning_rate": 2.1948392069311563e-06, "loss": 1.5168, "step": 35228 }, { "epoch": 2.7608934169278996, "grad_norm": 4.382819175720215, "learning_rate": 2.1945839520020625e-06, "loss": 1.5265, "step": 35229 }, { "epoch": 2.760971786833856, "grad_norm": 4.930877208709717, "learning_rate": 2.1943287077433302e-06, "loss": 1.6358, "step": 35230 }, { "epoch": 2.761050156739812, "grad_norm": 5.434347629547119, "learning_rate": 2.19407347415593e-06, "loss": 1.1886, "step": 35231 }, { "epoch": 2.761128526645768, "grad_norm": 4.963319778442383, "learning_rate": 2.1938182512408334e-06, "loss": 1.3864, "step": 35232 }, { "epoch": 2.761206896551724, "grad_norm": 3.044041633605957, "learning_rate": 2.1935630389990114e-06, "loss": 1.508, "step": 35233 }, { "epoch": 2.7612852664576804, "grad_norm": 2.5532681941986084, "learning_rate": 2.193307837431434e-06, "loss": 1.2363, "step": 35234 }, { "epoch": 2.7613636363636362, "grad_norm": 2.179145097732544, "learning_rate": 2.193052646539071e-06, "loss": 1.1736, "step": 35235 }, { "epoch": 2.7614420062695926, "grad_norm": 4.045075416564941, "learning_rate": 2.192797466322897e-06, "loss": 1.1821, "step": 35236 }, { "epoch": 2.7615203761755485, "grad_norm": 2.078341484069824, "learning_rate": 2.192542296783875e-06, "loss": 1.4585, "step": 35237 }, { "epoch": 2.761598746081505, "grad_norm": 3.5626583099365234, "learning_rate": 2.192287137922982e-06, "loss": 1.7391, "step": 35238 }, { "epoch": 2.7616771159874607, "grad_norm": 3.314002275466919, "learning_rate": 2.1920319897411892e-06, "loss": 1.2327, "step": 35239 }, { "epoch": 2.761755485893417, "grad_norm": 6.790310859680176, "learning_rate": 2.1917768522394624e-06, "loss": 1.9729, "step": 35240 }, { "epoch": 2.761833855799373, "grad_norm": 4.660285949707031, "learning_rate": 2.191521725418773e-06, "loss": 1.3686, "step": 35241 }, { "epoch": 2.7619122257053292, "grad_norm": 3.323302984237671, "learning_rate": 2.1912666092800918e-06, "loss": 1.0594, "step": 35242 }, { "epoch": 2.761990595611285, "grad_norm": 2.497439384460449, "learning_rate": 2.191011503824393e-06, "loss": 1.5116, "step": 35243 }, { "epoch": 2.7620689655172415, "grad_norm": 5.028755187988281, "learning_rate": 2.190756409052642e-06, "loss": 1.6863, "step": 35244 }, { "epoch": 2.7621473354231973, "grad_norm": 3.639463186264038, "learning_rate": 2.1905013249658104e-06, "loss": 1.5237, "step": 35245 }, { "epoch": 2.7622257053291537, "grad_norm": 3.461791515350342, "learning_rate": 2.190246251564869e-06, "loss": 1.6186, "step": 35246 }, { "epoch": 2.7623040752351096, "grad_norm": 3.7082719802856445, "learning_rate": 2.1899911888507874e-06, "loss": 1.5179, "step": 35247 }, { "epoch": 2.762382445141066, "grad_norm": 2.888979434967041, "learning_rate": 2.1897361368245364e-06, "loss": 1.0813, "step": 35248 }, { "epoch": 2.762460815047022, "grad_norm": 5.905411720275879, "learning_rate": 2.1894810954870855e-06, "loss": 1.4879, "step": 35249 }, { "epoch": 2.762539184952978, "grad_norm": 3.3000237941741943, "learning_rate": 2.189226064839405e-06, "loss": 1.6259, "step": 35250 }, { "epoch": 2.762617554858934, "grad_norm": 2.184828281402588, "learning_rate": 2.188971044882464e-06, "loss": 1.1158, "step": 35251 }, { "epoch": 2.7626959247648903, "grad_norm": 13.951406478881836, "learning_rate": 2.1887160356172343e-06, "loss": 1.4698, "step": 35252 }, { "epoch": 2.7627742946708462, "grad_norm": 2.3456249237060547, "learning_rate": 2.1884610370446847e-06, "loss": 1.2008, "step": 35253 }, { "epoch": 2.7628526645768026, "grad_norm": 2.3579654693603516, "learning_rate": 2.188206049165787e-06, "loss": 1.2706, "step": 35254 }, { "epoch": 2.762931034482759, "grad_norm": 3.211202621459961, "learning_rate": 2.1879510719815055e-06, "loss": 1.1723, "step": 35255 }, { "epoch": 2.763009404388715, "grad_norm": 3.263921022415161, "learning_rate": 2.187696105492815e-06, "loss": 1.7647, "step": 35256 }, { "epoch": 2.7630877742946707, "grad_norm": 3.540213108062744, "learning_rate": 2.1874411497006852e-06, "loss": 1.2871, "step": 35257 }, { "epoch": 2.763166144200627, "grad_norm": 3.2100255489349365, "learning_rate": 2.187186204606086e-06, "loss": 1.5458, "step": 35258 }, { "epoch": 2.7632445141065833, "grad_norm": 3.907177209854126, "learning_rate": 2.1869312702099836e-06, "loss": 1.4099, "step": 35259 }, { "epoch": 2.7633228840125392, "grad_norm": 2.6325926780700684, "learning_rate": 2.1866763465133483e-06, "loss": 1.3631, "step": 35260 }, { "epoch": 2.763401253918495, "grad_norm": 4.759891033172607, "learning_rate": 2.186421433517153e-06, "loss": 1.4028, "step": 35261 }, { "epoch": 2.7634796238244514, "grad_norm": 5.169126510620117, "learning_rate": 2.1861665312223666e-06, "loss": 1.0681, "step": 35262 }, { "epoch": 2.763557993730408, "grad_norm": 2.243342876434326, "learning_rate": 2.1859116396299563e-06, "loss": 1.5133, "step": 35263 }, { "epoch": 2.7636363636363637, "grad_norm": 3.1355161666870117, "learning_rate": 2.1856567587408927e-06, "loss": 1.4752, "step": 35264 }, { "epoch": 2.7637147335423196, "grad_norm": 5.7806830406188965, "learning_rate": 2.1854018885561453e-06, "loss": 1.3479, "step": 35265 }, { "epoch": 2.763793103448276, "grad_norm": 4.203629493713379, "learning_rate": 2.185147029076683e-06, "loss": 1.0805, "step": 35266 }, { "epoch": 2.7638714733542322, "grad_norm": 2.612055540084839, "learning_rate": 2.1848921803034754e-06, "loss": 1.1175, "step": 35267 }, { "epoch": 2.763949843260188, "grad_norm": 2.9656600952148438, "learning_rate": 2.1846373422374927e-06, "loss": 1.356, "step": 35268 }, { "epoch": 2.764028213166144, "grad_norm": 2.893399715423584, "learning_rate": 2.184382514879703e-06, "loss": 1.192, "step": 35269 }, { "epoch": 2.7641065830721003, "grad_norm": 2.6751608848571777, "learning_rate": 2.184127698231077e-06, "loss": 1.3677, "step": 35270 }, { "epoch": 2.7641849529780567, "grad_norm": 2.86242938041687, "learning_rate": 2.1838728922925816e-06, "loss": 1.4676, "step": 35271 }, { "epoch": 2.7642633228840126, "grad_norm": 3.167890787124634, "learning_rate": 2.1836180970651877e-06, "loss": 1.3094, "step": 35272 }, { "epoch": 2.7643416927899684, "grad_norm": 12.827630996704102, "learning_rate": 2.183363312549866e-06, "loss": 1.8231, "step": 35273 }, { "epoch": 2.7644200626959248, "grad_norm": 10.929369926452637, "learning_rate": 2.1831085387475796e-06, "loss": 1.1448, "step": 35274 }, { "epoch": 2.764498432601881, "grad_norm": 2.863924741744995, "learning_rate": 2.182853775659304e-06, "loss": 1.2534, "step": 35275 }, { "epoch": 2.764576802507837, "grad_norm": 2.5389976501464844, "learning_rate": 2.182599023286005e-06, "loss": 1.0006, "step": 35276 }, { "epoch": 2.764655172413793, "grad_norm": 2.725470542907715, "learning_rate": 2.1823442816286537e-06, "loss": 1.4965, "step": 35277 }, { "epoch": 2.764733542319749, "grad_norm": 2.8774681091308594, "learning_rate": 2.1820895506882166e-06, "loss": 1.4029, "step": 35278 }, { "epoch": 2.7648119122257055, "grad_norm": 3.5470643043518066, "learning_rate": 2.181834830465661e-06, "loss": 0.803, "step": 35279 }, { "epoch": 2.7648902821316614, "grad_norm": 4.522098541259766, "learning_rate": 2.1815801209619625e-06, "loss": 1.1239, "step": 35280 }, { "epoch": 2.7649686520376173, "grad_norm": 2.751012086868286, "learning_rate": 2.1813254221780835e-06, "loss": 1.4905, "step": 35281 }, { "epoch": 2.7650470219435737, "grad_norm": 7.056790828704834, "learning_rate": 2.1810707341149946e-06, "loss": 1.3241, "step": 35282 }, { "epoch": 2.76512539184953, "grad_norm": 2.2178995609283447, "learning_rate": 2.180816056773665e-06, "loss": 1.1841, "step": 35283 }, { "epoch": 2.765203761755486, "grad_norm": 2.6207807064056396, "learning_rate": 2.1805613901550628e-06, "loss": 1.4333, "step": 35284 }, { "epoch": 2.7652821316614418, "grad_norm": 2.8372349739074707, "learning_rate": 2.180306734260156e-06, "loss": 1.6011, "step": 35285 }, { "epoch": 2.765360501567398, "grad_norm": 2.7546474933624268, "learning_rate": 2.1800520890899158e-06, "loss": 1.6333, "step": 35286 }, { "epoch": 2.7654388714733544, "grad_norm": 2.841960906982422, "learning_rate": 2.1797974546453084e-06, "loss": 1.189, "step": 35287 }, { "epoch": 2.7655172413793103, "grad_norm": 3.6752679347991943, "learning_rate": 2.1795428309273026e-06, "loss": 1.6712, "step": 35288 }, { "epoch": 2.765595611285266, "grad_norm": 3.0654468536376953, "learning_rate": 2.179288217936868e-06, "loss": 1.5611, "step": 35289 }, { "epoch": 2.7656739811912225, "grad_norm": 2.7498111724853516, "learning_rate": 2.179033615674971e-06, "loss": 1.3805, "step": 35290 }, { "epoch": 2.765752351097179, "grad_norm": 3.5292444229125977, "learning_rate": 2.1787790241425823e-06, "loss": 1.0605, "step": 35291 }, { "epoch": 2.7658307210031348, "grad_norm": 2.997124195098877, "learning_rate": 2.1785244433406705e-06, "loss": 1.1063, "step": 35292 }, { "epoch": 2.7659090909090907, "grad_norm": 2.7807016372680664, "learning_rate": 2.178269873270199e-06, "loss": 1.2044, "step": 35293 }, { "epoch": 2.765987460815047, "grad_norm": 3.102839946746826, "learning_rate": 2.1780153139321412e-06, "loss": 1.3746, "step": 35294 }, { "epoch": 2.7660658307210033, "grad_norm": 5.637359619140625, "learning_rate": 2.177760765327466e-06, "loss": 0.9466, "step": 35295 }, { "epoch": 2.766144200626959, "grad_norm": 3.9579906463623047, "learning_rate": 2.1775062274571366e-06, "loss": 1.2856, "step": 35296 }, { "epoch": 2.766222570532915, "grad_norm": 5.088862419128418, "learning_rate": 2.1772517003221245e-06, "loss": 1.6784, "step": 35297 }, { "epoch": 2.7663009404388714, "grad_norm": 3.6148154735565186, "learning_rate": 2.176997183923395e-06, "loss": 1.214, "step": 35298 }, { "epoch": 2.7663793103448278, "grad_norm": 3.570566415786743, "learning_rate": 2.1767426782619215e-06, "loss": 1.0312, "step": 35299 }, { "epoch": 2.7664576802507836, "grad_norm": 5.941490173339844, "learning_rate": 2.1764881833386673e-06, "loss": 1.4075, "step": 35300 }, { "epoch": 2.7665360501567395, "grad_norm": 2.599558115005493, "learning_rate": 2.1762336991546017e-06, "loss": 1.3916, "step": 35301 }, { "epoch": 2.766614420062696, "grad_norm": 3.103445291519165, "learning_rate": 2.1759792257106928e-06, "loss": 0.8603, "step": 35302 }, { "epoch": 2.766692789968652, "grad_norm": 3.4306719303131104, "learning_rate": 2.1757247630079087e-06, "loss": 1.2155, "step": 35303 }, { "epoch": 2.766771159874608, "grad_norm": 3.7903475761413574, "learning_rate": 2.1754703110472163e-06, "loss": 1.3657, "step": 35304 }, { "epoch": 2.7668495297805644, "grad_norm": 4.2511887550354, "learning_rate": 2.1752158698295843e-06, "loss": 1.4246, "step": 35305 }, { "epoch": 2.7669278996865203, "grad_norm": 2.730438709259033, "learning_rate": 2.1749614393559808e-06, "loss": 1.3835, "step": 35306 }, { "epoch": 2.7670062695924766, "grad_norm": 2.225264310836792, "learning_rate": 2.174707019627372e-06, "loss": 1.1702, "step": 35307 }, { "epoch": 2.7670846394984325, "grad_norm": 3.2085280418395996, "learning_rate": 2.174452610644727e-06, "loss": 1.3253, "step": 35308 }, { "epoch": 2.767163009404389, "grad_norm": 5.241133689880371, "learning_rate": 2.1741982124090133e-06, "loss": 1.5043, "step": 35309 }, { "epoch": 2.7672413793103448, "grad_norm": 2.3619072437286377, "learning_rate": 2.1739438249211998e-06, "loss": 1.3136, "step": 35310 }, { "epoch": 2.767319749216301, "grad_norm": 3.6577978134155273, "learning_rate": 2.173689448182251e-06, "loss": 1.2774, "step": 35311 }, { "epoch": 2.767398119122257, "grad_norm": 7.100600719451904, "learning_rate": 2.173435082193133e-06, "loss": 1.1708, "step": 35312 }, { "epoch": 2.7674764890282133, "grad_norm": 6.665003299713135, "learning_rate": 2.1731807269548187e-06, "loss": 2.2484, "step": 35313 }, { "epoch": 2.767554858934169, "grad_norm": 5.237941741943359, "learning_rate": 2.1729263824682743e-06, "loss": 1.625, "step": 35314 }, { "epoch": 2.7676332288401255, "grad_norm": 3.091451406478882, "learning_rate": 2.172672048734465e-06, "loss": 1.3506, "step": 35315 }, { "epoch": 2.7677115987460814, "grad_norm": 3.0569534301757812, "learning_rate": 2.172417725754358e-06, "loss": 1.3292, "step": 35316 }, { "epoch": 2.7677899686520377, "grad_norm": 3.377808094024658, "learning_rate": 2.1721634135289203e-06, "loss": 1.3152, "step": 35317 }, { "epoch": 2.7678683385579936, "grad_norm": 3.694709539413452, "learning_rate": 2.171909112059124e-06, "loss": 1.3947, "step": 35318 }, { "epoch": 2.76794670846395, "grad_norm": 5.220724105834961, "learning_rate": 2.171654821345931e-06, "loss": 1.3236, "step": 35319 }, { "epoch": 2.768025078369906, "grad_norm": 7.790185451507568, "learning_rate": 2.1714005413903105e-06, "loss": 1.5751, "step": 35320 }, { "epoch": 2.768103448275862, "grad_norm": 3.648101568222046, "learning_rate": 2.171146272193229e-06, "loss": 1.1496, "step": 35321 }, { "epoch": 2.768181818181818, "grad_norm": 2.9025230407714844, "learning_rate": 2.170892013755655e-06, "loss": 1.4326, "step": 35322 }, { "epoch": 2.7682601880877744, "grad_norm": 3.3929083347320557, "learning_rate": 2.170637766078554e-06, "loss": 1.1576, "step": 35323 }, { "epoch": 2.7683385579937303, "grad_norm": 3.586078643798828, "learning_rate": 2.170383529162894e-06, "loss": 1.1598, "step": 35324 }, { "epoch": 2.7684169278996866, "grad_norm": 4.282280921936035, "learning_rate": 2.1701293030096433e-06, "loss": 1.5634, "step": 35325 }, { "epoch": 2.7684952978056425, "grad_norm": 2.223036050796509, "learning_rate": 2.169875087619763e-06, "loss": 1.447, "step": 35326 }, { "epoch": 2.768573667711599, "grad_norm": 5.667943954467773, "learning_rate": 2.1696208829942266e-06, "loss": 1.7154, "step": 35327 }, { "epoch": 2.7686520376175547, "grad_norm": 4.600802421569824, "learning_rate": 2.1693666891339986e-06, "loss": 1.2898, "step": 35328 }, { "epoch": 2.768730407523511, "grad_norm": 5.16556978225708, "learning_rate": 2.169112506040048e-06, "loss": 1.242, "step": 35329 }, { "epoch": 2.768808777429467, "grad_norm": 2.3530972003936768, "learning_rate": 2.1688583337133373e-06, "loss": 1.4758, "step": 35330 }, { "epoch": 2.7688871473354233, "grad_norm": 5.19020414352417, "learning_rate": 2.1686041721548336e-06, "loss": 1.8662, "step": 35331 }, { "epoch": 2.768965517241379, "grad_norm": 4.167149066925049, "learning_rate": 2.1683500213655067e-06, "loss": 0.9402, "step": 35332 }, { "epoch": 2.7690438871473355, "grad_norm": 2.859431743621826, "learning_rate": 2.1680958813463236e-06, "loss": 1.2504, "step": 35333 }, { "epoch": 2.7691222570532914, "grad_norm": 4.091923236846924, "learning_rate": 2.1678417520982483e-06, "loss": 0.8715, "step": 35334 }, { "epoch": 2.7692006269592477, "grad_norm": 5.3069562911987305, "learning_rate": 2.1675876336222475e-06, "loss": 1.4226, "step": 35335 }, { "epoch": 2.7692789968652036, "grad_norm": 2.2617342472076416, "learning_rate": 2.167333525919289e-06, "loss": 1.2784, "step": 35336 }, { "epoch": 2.76935736677116, "grad_norm": 7.0579423904418945, "learning_rate": 2.1670794289903385e-06, "loss": 1.2246, "step": 35337 }, { "epoch": 2.769435736677116, "grad_norm": 2.306786298751831, "learning_rate": 2.166825342836363e-06, "loss": 1.2508, "step": 35338 }, { "epoch": 2.769514106583072, "grad_norm": 28.09069061279297, "learning_rate": 2.1665712674583285e-06, "loss": 0.9197, "step": 35339 }, { "epoch": 2.769592476489028, "grad_norm": 2.3808765411376953, "learning_rate": 2.1663172028572014e-06, "loss": 1.4228, "step": 35340 }, { "epoch": 2.7696708463949844, "grad_norm": 2.4238715171813965, "learning_rate": 2.1660631490339486e-06, "loss": 1.5398, "step": 35341 }, { "epoch": 2.7697492163009403, "grad_norm": 3.4205517768859863, "learning_rate": 2.165809105989535e-06, "loss": 1.3508, "step": 35342 }, { "epoch": 2.7698275862068966, "grad_norm": 4.239323616027832, "learning_rate": 2.1655550737249294e-06, "loss": 1.2781, "step": 35343 }, { "epoch": 2.7699059561128525, "grad_norm": 3.372004270553589, "learning_rate": 2.1653010522410955e-06, "loss": 1.2908, "step": 35344 }, { "epoch": 2.769984326018809, "grad_norm": 7.763681411743164, "learning_rate": 2.165047041539001e-06, "loss": 0.5068, "step": 35345 }, { "epoch": 2.7700626959247647, "grad_norm": 3.20048451423645, "learning_rate": 2.1647930416196106e-06, "loss": 1.293, "step": 35346 }, { "epoch": 2.770141065830721, "grad_norm": 2.918466567993164, "learning_rate": 2.1645390524838917e-06, "loss": 0.8529, "step": 35347 }, { "epoch": 2.770219435736677, "grad_norm": 2.8901305198669434, "learning_rate": 2.1642850741328114e-06, "loss": 1.3952, "step": 35348 }, { "epoch": 2.7702978056426333, "grad_norm": 12.656747817993164, "learning_rate": 2.1640311065673324e-06, "loss": 1.2366, "step": 35349 }, { "epoch": 2.7703761755485896, "grad_norm": 2.87900972366333, "learning_rate": 2.1637771497884208e-06, "loss": 1.4006, "step": 35350 }, { "epoch": 2.7704545454545455, "grad_norm": 4.0307769775390625, "learning_rate": 2.1635232037970475e-06, "loss": 1.7307, "step": 35351 }, { "epoch": 2.7705329153605014, "grad_norm": 1.9151791334152222, "learning_rate": 2.1632692685941737e-06, "loss": 1.5277, "step": 35352 }, { "epoch": 2.7706112852664577, "grad_norm": 2.3968870639801025, "learning_rate": 2.163015344180766e-06, "loss": 1.5396, "step": 35353 }, { "epoch": 2.770689655172414, "grad_norm": 3.9515655040740967, "learning_rate": 2.162761430557791e-06, "loss": 1.4694, "step": 35354 }, { "epoch": 2.77076802507837, "grad_norm": 9.402790069580078, "learning_rate": 2.1625075277262142e-06, "loss": 1.3478, "step": 35355 }, { "epoch": 2.770846394984326, "grad_norm": 3.117917060852051, "learning_rate": 2.162253635687002e-06, "loss": 0.6671, "step": 35356 }, { "epoch": 2.770924764890282, "grad_norm": 5.76762580871582, "learning_rate": 2.1619997544411186e-06, "loss": 1.1769, "step": 35357 }, { "epoch": 2.7710031347962385, "grad_norm": 4.26074743270874, "learning_rate": 2.1617458839895307e-06, "loss": 1.0795, "step": 35358 }, { "epoch": 2.7710815047021944, "grad_norm": 3.4296350479125977, "learning_rate": 2.161492024333203e-06, "loss": 0.8951, "step": 35359 }, { "epoch": 2.7711598746081503, "grad_norm": 4.619661808013916, "learning_rate": 2.1612381754731025e-06, "loss": 0.7791, "step": 35360 }, { "epoch": 2.7712382445141066, "grad_norm": 3.9302127361297607, "learning_rate": 2.160984337410194e-06, "loss": 1.3857, "step": 35361 }, { "epoch": 2.771316614420063, "grad_norm": 2.890763759613037, "learning_rate": 2.160730510145444e-06, "loss": 1.0145, "step": 35362 }, { "epoch": 2.771394984326019, "grad_norm": 2.442861795425415, "learning_rate": 2.1604766936798138e-06, "loss": 1.388, "step": 35363 }, { "epoch": 2.7714733542319747, "grad_norm": 2.6885201930999756, "learning_rate": 2.1602228880142735e-06, "loss": 1.5924, "step": 35364 }, { "epoch": 2.771551724137931, "grad_norm": 4.216989517211914, "learning_rate": 2.1599690931497864e-06, "loss": 0.9059, "step": 35365 }, { "epoch": 2.7716300940438874, "grad_norm": 4.809619903564453, "learning_rate": 2.15971530908732e-06, "loss": 0.757, "step": 35366 }, { "epoch": 2.7717084639498433, "grad_norm": 3.52778959274292, "learning_rate": 2.159461535827836e-06, "loss": 1.7163, "step": 35367 }, { "epoch": 2.771786833855799, "grad_norm": 4.202566623687744, "learning_rate": 2.1592077733722995e-06, "loss": 1.4813, "step": 35368 }, { "epoch": 2.7718652037617555, "grad_norm": 7.429927349090576, "learning_rate": 2.1589540217216792e-06, "loss": 2.4411, "step": 35369 }, { "epoch": 2.771943573667712, "grad_norm": 6.760058879852295, "learning_rate": 2.1587002808769408e-06, "loss": 1.2153, "step": 35370 }, { "epoch": 2.7720219435736677, "grad_norm": 2.3740320205688477, "learning_rate": 2.158446550839045e-06, "loss": 1.1648, "step": 35371 }, { "epoch": 2.7721003134796236, "grad_norm": 4.855659008026123, "learning_rate": 2.158192831608959e-06, "loss": 1.2315, "step": 35372 }, { "epoch": 2.77217868338558, "grad_norm": 6.728518962860107, "learning_rate": 2.157939123187646e-06, "loss": 1.3247, "step": 35373 }, { "epoch": 2.7722570532915363, "grad_norm": 7.994133472442627, "learning_rate": 2.1576854255760765e-06, "loss": 1.2698, "step": 35374 }, { "epoch": 2.772335423197492, "grad_norm": 3.1919004917144775, "learning_rate": 2.1574317387752097e-06, "loss": 1.0578, "step": 35375 }, { "epoch": 2.772413793103448, "grad_norm": 4.808001518249512, "learning_rate": 2.157178062786013e-06, "loss": 1.0363, "step": 35376 }, { "epoch": 2.7724921630094044, "grad_norm": 3.227001190185547, "learning_rate": 2.1569243976094507e-06, "loss": 1.4573, "step": 35377 }, { "epoch": 2.7725705329153607, "grad_norm": 4.335495471954346, "learning_rate": 2.1566707432464873e-06, "loss": 1.6767, "step": 35378 }, { "epoch": 2.7726489028213166, "grad_norm": 3.145357847213745, "learning_rate": 2.1564170996980882e-06, "loss": 1.3562, "step": 35379 }, { "epoch": 2.7727272727272725, "grad_norm": 3.458596706390381, "learning_rate": 2.156163466965218e-06, "loss": 1.4032, "step": 35380 }, { "epoch": 2.772805642633229, "grad_norm": 2.7682981491088867, "learning_rate": 2.155909845048843e-06, "loss": 1.3196, "step": 35381 }, { "epoch": 2.772884012539185, "grad_norm": 5.949130058288574, "learning_rate": 2.155656233949922e-06, "loss": 1.1672, "step": 35382 }, { "epoch": 2.772962382445141, "grad_norm": 5.119151592254639, "learning_rate": 2.155402633669426e-06, "loss": 1.3602, "step": 35383 }, { "epoch": 2.773040752351097, "grad_norm": 2.0020439624786377, "learning_rate": 2.155149044208318e-06, "loss": 1.5616, "step": 35384 }, { "epoch": 2.7731191222570533, "grad_norm": 6.579444885253906, "learning_rate": 2.1548954655675623e-06, "loss": 1.0326, "step": 35385 }, { "epoch": 2.7731974921630096, "grad_norm": 3.4563889503479004, "learning_rate": 2.154641897748122e-06, "loss": 1.2229, "step": 35386 }, { "epoch": 2.7732758620689655, "grad_norm": 3.5923118591308594, "learning_rate": 2.15438834075096e-06, "loss": 1.2029, "step": 35387 }, { "epoch": 2.7733542319749214, "grad_norm": 2.6287577152252197, "learning_rate": 2.154134794577046e-06, "loss": 1.4975, "step": 35388 }, { "epoch": 2.7734326018808777, "grad_norm": 4.3222503662109375, "learning_rate": 2.153881259227343e-06, "loss": 1.2827, "step": 35389 }, { "epoch": 2.773510971786834, "grad_norm": 3.6620752811431885, "learning_rate": 2.1536277347028112e-06, "loss": 1.4414, "step": 35390 }, { "epoch": 2.77358934169279, "grad_norm": 4.432066917419434, "learning_rate": 2.1533742210044185e-06, "loss": 1.4762, "step": 35391 }, { "epoch": 2.773667711598746, "grad_norm": 3.4981491565704346, "learning_rate": 2.1531207181331275e-06, "loss": 1.3525, "step": 35392 }, { "epoch": 2.773746081504702, "grad_norm": 3.8151564598083496, "learning_rate": 2.152867226089903e-06, "loss": 1.4415, "step": 35393 }, { "epoch": 2.7738244514106585, "grad_norm": 2.309718370437622, "learning_rate": 2.1526137448757095e-06, "loss": 1.6872, "step": 35394 }, { "epoch": 2.7739028213166144, "grad_norm": 3.555422306060791, "learning_rate": 2.152360274491511e-06, "loss": 1.0069, "step": 35395 }, { "epoch": 2.7739811912225707, "grad_norm": 2.164241075515747, "learning_rate": 2.1521068149382707e-06, "loss": 1.2397, "step": 35396 }, { "epoch": 2.7740595611285266, "grad_norm": 5.496230602264404, "learning_rate": 2.1518533662169536e-06, "loss": 1.5932, "step": 35397 }, { "epoch": 2.774137931034483, "grad_norm": 3.733092784881592, "learning_rate": 2.151599928328524e-06, "loss": 1.3572, "step": 35398 }, { "epoch": 2.774216300940439, "grad_norm": 3.7657999992370605, "learning_rate": 2.151346501273944e-06, "loss": 1.5277, "step": 35399 }, { "epoch": 2.774294670846395, "grad_norm": 3.553718328475952, "learning_rate": 2.1510930850541816e-06, "loss": 1.2046, "step": 35400 }, { "epoch": 2.774373040752351, "grad_norm": 4.126408576965332, "learning_rate": 2.1508396796701943e-06, "loss": 1.6448, "step": 35401 }, { "epoch": 2.7744514106583074, "grad_norm": 2.6500003337860107, "learning_rate": 2.15058628512295e-06, "loss": 1.2075, "step": 35402 }, { "epoch": 2.7745297805642632, "grad_norm": 2.699070453643799, "learning_rate": 2.1503329014134133e-06, "loss": 1.266, "step": 35403 }, { "epoch": 2.7746081504702196, "grad_norm": 7.563024997711182, "learning_rate": 2.150079528542548e-06, "loss": 1.1856, "step": 35404 }, { "epoch": 2.7746865203761755, "grad_norm": 3.192371129989624, "learning_rate": 2.149826166511314e-06, "loss": 0.8411, "step": 35405 }, { "epoch": 2.774764890282132, "grad_norm": 4.5403008460998535, "learning_rate": 2.1495728153206755e-06, "loss": 1.5561, "step": 35406 }, { "epoch": 2.7748432601880877, "grad_norm": 5.723933219909668, "learning_rate": 2.1493194749716023e-06, "loss": 1.0946, "step": 35407 }, { "epoch": 2.774921630094044, "grad_norm": 2.26466965675354, "learning_rate": 2.1490661454650514e-06, "loss": 1.2426, "step": 35408 }, { "epoch": 2.775, "grad_norm": 2.609368324279785, "learning_rate": 2.148812826801989e-06, "loss": 0.8463, "step": 35409 }, { "epoch": 2.7750783699059562, "grad_norm": 3.107823371887207, "learning_rate": 2.1485595189833773e-06, "loss": 1.8413, "step": 35410 }, { "epoch": 2.775156739811912, "grad_norm": 13.20896053314209, "learning_rate": 2.1483062220101815e-06, "loss": 1.2874, "step": 35411 }, { "epoch": 2.7752351097178685, "grad_norm": 2.579554319381714, "learning_rate": 2.1480529358833633e-06, "loss": 0.8601, "step": 35412 }, { "epoch": 2.7753134796238244, "grad_norm": 2.6194050312042236, "learning_rate": 2.1477996606038875e-06, "loss": 1.3554, "step": 35413 }, { "epoch": 2.7753918495297807, "grad_norm": 3.3618085384368896, "learning_rate": 2.1475463961727163e-06, "loss": 1.1186, "step": 35414 }, { "epoch": 2.7754702194357366, "grad_norm": 4.0579514503479, "learning_rate": 2.147293142590814e-06, "loss": 1.0127, "step": 35415 }, { "epoch": 2.775548589341693, "grad_norm": 2.9522786140441895, "learning_rate": 2.147039899859143e-06, "loss": 1.4293, "step": 35416 }, { "epoch": 2.775626959247649, "grad_norm": 2.592259407043457, "learning_rate": 2.146786667978667e-06, "loss": 1.3791, "step": 35417 }, { "epoch": 2.775705329153605, "grad_norm": 5.312378883361816, "learning_rate": 2.1465334469503507e-06, "loss": 1.2865, "step": 35418 }, { "epoch": 2.775783699059561, "grad_norm": 5.897095203399658, "learning_rate": 2.1462802367751533e-06, "loss": 1.1047, "step": 35419 }, { "epoch": 2.7758620689655173, "grad_norm": 2.1740548610687256, "learning_rate": 2.1460270374540393e-06, "loss": 1.1369, "step": 35420 }, { "epoch": 2.7759404388714732, "grad_norm": 7.187408924102783, "learning_rate": 2.1457738489879736e-06, "loss": 1.4567, "step": 35421 }, { "epoch": 2.7760188087774296, "grad_norm": 2.3254406452178955, "learning_rate": 2.1455206713779204e-06, "loss": 1.3121, "step": 35422 }, { "epoch": 2.7760971786833855, "grad_norm": 3.8839523792266846, "learning_rate": 2.1452675046248383e-06, "loss": 1.5423, "step": 35423 }, { "epoch": 2.776175548589342, "grad_norm": 2.439514398574829, "learning_rate": 2.145014348729693e-06, "loss": 1.3599, "step": 35424 }, { "epoch": 2.7762539184952977, "grad_norm": 3.4417574405670166, "learning_rate": 2.1447612036934436e-06, "loss": 1.1624, "step": 35425 }, { "epoch": 2.776332288401254, "grad_norm": 2.731572151184082, "learning_rate": 2.1445080695170606e-06, "loss": 1.3186, "step": 35426 }, { "epoch": 2.77641065830721, "grad_norm": 4.609523773193359, "learning_rate": 2.1442549462015e-06, "loss": 1.01, "step": 35427 }, { "epoch": 2.7764890282131662, "grad_norm": 3.0607333183288574, "learning_rate": 2.144001833747727e-06, "loss": 1.5683, "step": 35428 }, { "epoch": 2.776567398119122, "grad_norm": 2.8185858726501465, "learning_rate": 2.143748732156704e-06, "loss": 1.692, "step": 35429 }, { "epoch": 2.7766457680250785, "grad_norm": 5.106828689575195, "learning_rate": 2.143495641429394e-06, "loss": 1.7088, "step": 35430 }, { "epoch": 2.7767241379310343, "grad_norm": 7.254696369171143, "learning_rate": 2.143242561566759e-06, "loss": 1.6087, "step": 35431 }, { "epoch": 2.7768025078369907, "grad_norm": 5.94503927230835, "learning_rate": 2.1429894925697627e-06, "loss": 1.5801, "step": 35432 }, { "epoch": 2.7768808777429466, "grad_norm": 4.291013717651367, "learning_rate": 2.142736434439368e-06, "loss": 0.7362, "step": 35433 }, { "epoch": 2.776959247648903, "grad_norm": 4.671154499053955, "learning_rate": 2.142483387176533e-06, "loss": 0.9566, "step": 35434 }, { "epoch": 2.777037617554859, "grad_norm": 3.454233407974243, "learning_rate": 2.142230350782225e-06, "loss": 1.3823, "step": 35435 }, { "epoch": 2.777115987460815, "grad_norm": 6.5848588943481445, "learning_rate": 2.141977325257405e-06, "loss": 1.5256, "step": 35436 }, { "epoch": 2.777194357366771, "grad_norm": 3.34397554397583, "learning_rate": 2.141724310603037e-06, "loss": 1.2443, "step": 35437 }, { "epoch": 2.7772727272727273, "grad_norm": 5.188340187072754, "learning_rate": 2.1414713068200804e-06, "loss": 1.8265, "step": 35438 }, { "epoch": 2.7773510971786832, "grad_norm": 3.6528141498565674, "learning_rate": 2.1412183139094967e-06, "loss": 1.4749, "step": 35439 }, { "epoch": 2.7774294670846396, "grad_norm": 3.276663064956665, "learning_rate": 2.1409653318722517e-06, "loss": 1.0616, "step": 35440 }, { "epoch": 2.7775078369905954, "grad_norm": 5.785126686096191, "learning_rate": 2.140712360709308e-06, "loss": 1.4292, "step": 35441 }, { "epoch": 2.777586206896552, "grad_norm": 4.175024032592773, "learning_rate": 2.140459400421624e-06, "loss": 1.1934, "step": 35442 }, { "epoch": 2.7776645768025077, "grad_norm": 3.122652053833008, "learning_rate": 2.140206451010164e-06, "loss": 0.9718, "step": 35443 }, { "epoch": 2.777742946708464, "grad_norm": 5.992027759552002, "learning_rate": 2.1399535124758874e-06, "loss": 1.349, "step": 35444 }, { "epoch": 2.7778213166144203, "grad_norm": 2.1633121967315674, "learning_rate": 2.1397005848197628e-06, "loss": 1.0208, "step": 35445 }, { "epoch": 2.777899686520376, "grad_norm": 2.26851749420166, "learning_rate": 2.1394476680427454e-06, "loss": 1.4611, "step": 35446 }, { "epoch": 2.777978056426332, "grad_norm": 2.9972355365753174, "learning_rate": 2.139194762145801e-06, "loss": 1.2484, "step": 35447 }, { "epoch": 2.7780564263322884, "grad_norm": 5.0177812576293945, "learning_rate": 2.1389418671298894e-06, "loss": 1.4216, "step": 35448 }, { "epoch": 2.7781347962382448, "grad_norm": 3.908289670944214, "learning_rate": 2.138688982995974e-06, "loss": 2.0519, "step": 35449 }, { "epoch": 2.7782131661442007, "grad_norm": 4.301943302154541, "learning_rate": 2.1384361097450156e-06, "loss": 1.2526, "step": 35450 }, { "epoch": 2.7782915360501566, "grad_norm": 3.146665573120117, "learning_rate": 2.1381832473779766e-06, "loss": 0.8942, "step": 35451 }, { "epoch": 2.778369905956113, "grad_norm": 2.703206777572632, "learning_rate": 2.137930395895819e-06, "loss": 1.318, "step": 35452 }, { "epoch": 2.778448275862069, "grad_norm": 2.6469078063964844, "learning_rate": 2.137677555299503e-06, "loss": 1.488, "step": 35453 }, { "epoch": 2.778526645768025, "grad_norm": 3.500436305999756, "learning_rate": 2.137424725589992e-06, "loss": 1.6033, "step": 35454 }, { "epoch": 2.778605015673981, "grad_norm": 3.5859591960906982, "learning_rate": 2.1371719067682474e-06, "loss": 1.3315, "step": 35455 }, { "epoch": 2.7786833855799373, "grad_norm": 3.6634368896484375, "learning_rate": 2.1369190988352324e-06, "loss": 1.1111, "step": 35456 }, { "epoch": 2.7787617554858937, "grad_norm": 4.334597110748291, "learning_rate": 2.136666301791902e-06, "loss": 1.2408, "step": 35457 }, { "epoch": 2.7788401253918495, "grad_norm": 2.237741708755493, "learning_rate": 2.1364135156392253e-06, "loss": 1.3634, "step": 35458 }, { "epoch": 2.7789184952978054, "grad_norm": 3.9352433681488037, "learning_rate": 2.13616074037816e-06, "loss": 1.0328, "step": 35459 }, { "epoch": 2.7789968652037618, "grad_norm": 3.0357768535614014, "learning_rate": 2.13590797600967e-06, "loss": 1.2566, "step": 35460 }, { "epoch": 2.779075235109718, "grad_norm": 4.010420799255371, "learning_rate": 2.1356552225347134e-06, "loss": 1.1164, "step": 35461 }, { "epoch": 2.779153605015674, "grad_norm": 3.084925651550293, "learning_rate": 2.135402479954251e-06, "loss": 1.3859, "step": 35462 }, { "epoch": 2.77923197492163, "grad_norm": 2.4881627559661865, "learning_rate": 2.1351497482692503e-06, "loss": 1.4498, "step": 35463 }, { "epoch": 2.779310344827586, "grad_norm": 3.0979456901550293, "learning_rate": 2.134897027480666e-06, "loss": 1.3877, "step": 35464 }, { "epoch": 2.7793887147335425, "grad_norm": 2.924811601638794, "learning_rate": 2.1346443175894625e-06, "loss": 1.7467, "step": 35465 }, { "epoch": 2.7794670846394984, "grad_norm": 3.36212158203125, "learning_rate": 2.1343916185966e-06, "loss": 1.1185, "step": 35466 }, { "epoch": 2.7795454545454543, "grad_norm": 3.1742634773254395, "learning_rate": 2.13413893050304e-06, "loss": 0.9945, "step": 35467 }, { "epoch": 2.7796238244514107, "grad_norm": 10.436044692993164, "learning_rate": 2.133886253309744e-06, "loss": 0.7493, "step": 35468 }, { "epoch": 2.779702194357367, "grad_norm": 3.9699666500091553, "learning_rate": 2.133633587017672e-06, "loss": 1.7485, "step": 35469 }, { "epoch": 2.779780564263323, "grad_norm": 3.238445281982422, "learning_rate": 2.1333809316277854e-06, "loss": 1.199, "step": 35470 }, { "epoch": 2.7798589341692788, "grad_norm": 2.4262757301330566, "learning_rate": 2.133128287141045e-06, "loss": 0.8741, "step": 35471 }, { "epoch": 2.779937304075235, "grad_norm": 3.5710413455963135, "learning_rate": 2.132875653558413e-06, "loss": 1.4156, "step": 35472 }, { "epoch": 2.7800156739811914, "grad_norm": 5.526268482208252, "learning_rate": 2.1326230308808486e-06, "loss": 0.5899, "step": 35473 }, { "epoch": 2.7800940438871473, "grad_norm": 6.110230445861816, "learning_rate": 2.132370419109316e-06, "loss": 1.7227, "step": 35474 }, { "epoch": 2.780172413793103, "grad_norm": 2.5426318645477295, "learning_rate": 2.132117818244771e-06, "loss": 1.3535, "step": 35475 }, { "epoch": 2.7802507836990595, "grad_norm": 7.596373558044434, "learning_rate": 2.1318652282881754e-06, "loss": 1.8167, "step": 35476 }, { "epoch": 2.780329153605016, "grad_norm": 3.260511636734009, "learning_rate": 2.1316126492404927e-06, "loss": 0.9094, "step": 35477 }, { "epoch": 2.7804075235109718, "grad_norm": 1.9138725996017456, "learning_rate": 2.131360081102684e-06, "loss": 1.2548, "step": 35478 }, { "epoch": 2.7804858934169276, "grad_norm": 2.947888135910034, "learning_rate": 2.131107523875707e-06, "loss": 1.1824, "step": 35479 }, { "epoch": 2.780564263322884, "grad_norm": 2.567308187484741, "learning_rate": 2.1308549775605226e-06, "loss": 1.2196, "step": 35480 }, { "epoch": 2.7806426332288403, "grad_norm": 2.9585916996002197, "learning_rate": 2.130602442158091e-06, "loss": 1.1582, "step": 35481 }, { "epoch": 2.780721003134796, "grad_norm": 2.4879207611083984, "learning_rate": 2.130349917669377e-06, "loss": 1.2047, "step": 35482 }, { "epoch": 2.780799373040752, "grad_norm": 4.495181083679199, "learning_rate": 2.1300974040953366e-06, "loss": 1.5459, "step": 35483 }, { "epoch": 2.7808777429467084, "grad_norm": 3.780242443084717, "learning_rate": 2.129844901436931e-06, "loss": 1.2124, "step": 35484 }, { "epoch": 2.7809561128526648, "grad_norm": 3.526557445526123, "learning_rate": 2.1295924096951216e-06, "loss": 1.1687, "step": 35485 }, { "epoch": 2.7810344827586206, "grad_norm": 2.6873559951782227, "learning_rate": 2.129339928870868e-06, "loss": 1.2641, "step": 35486 }, { "epoch": 2.7811128526645765, "grad_norm": 2.6394693851470947, "learning_rate": 2.129087458965132e-06, "loss": 1.2694, "step": 35487 }, { "epoch": 2.781191222570533, "grad_norm": 5.5760908126831055, "learning_rate": 2.1288349999788714e-06, "loss": 1.5882, "step": 35488 }, { "epoch": 2.781269592476489, "grad_norm": 3.9076285362243652, "learning_rate": 2.12858255191305e-06, "loss": 1.174, "step": 35489 }, { "epoch": 2.781347962382445, "grad_norm": 4.311245918273926, "learning_rate": 2.128330114768622e-06, "loss": 1.5571, "step": 35490 }, { "epoch": 2.7814263322884014, "grad_norm": 3.9595654010772705, "learning_rate": 2.128077688546553e-06, "loss": 1.2122, "step": 35491 }, { "epoch": 2.7815047021943573, "grad_norm": 3.5105485916137695, "learning_rate": 2.1278252732478015e-06, "loss": 1.5254, "step": 35492 }, { "epoch": 2.7815830721003136, "grad_norm": 4.284889221191406, "learning_rate": 2.1275728688733295e-06, "loss": 1.468, "step": 35493 }, { "epoch": 2.7816614420062695, "grad_norm": 3.4710731506347656, "learning_rate": 2.1273204754240927e-06, "loss": 1.5575, "step": 35494 }, { "epoch": 2.781739811912226, "grad_norm": 3.8006856441497803, "learning_rate": 2.127068092901052e-06, "loss": 1.2662, "step": 35495 }, { "epoch": 2.7818181818181817, "grad_norm": 3.2010080814361572, "learning_rate": 2.1268157213051697e-06, "loss": 1.2321, "step": 35496 }, { "epoch": 2.781896551724138, "grad_norm": 2.4167861938476562, "learning_rate": 2.1265633606374077e-06, "loss": 1.4139, "step": 35497 }, { "epoch": 2.781974921630094, "grad_norm": 3.3759121894836426, "learning_rate": 2.1263110108987195e-06, "loss": 1.088, "step": 35498 }, { "epoch": 2.7820532915360503, "grad_norm": 3.4688611030578613, "learning_rate": 2.1260586720900687e-06, "loss": 1.4274, "step": 35499 }, { "epoch": 2.782131661442006, "grad_norm": 4.720592498779297, "learning_rate": 2.125806344212413e-06, "loss": 1.1782, "step": 35500 }, { "epoch": 2.7822100313479625, "grad_norm": 2.3657751083374023, "learning_rate": 2.1255540272667174e-06, "loss": 1.4903, "step": 35501 }, { "epoch": 2.7822884012539184, "grad_norm": 6.924500465393066, "learning_rate": 2.1253017212539362e-06, "loss": 1.9275, "step": 35502 }, { "epoch": 2.7823667711598747, "grad_norm": 3.2039687633514404, "learning_rate": 2.1250494261750304e-06, "loss": 1.4515, "step": 35503 }, { "epoch": 2.7824451410658306, "grad_norm": 3.3103413581848145, "learning_rate": 2.12479714203096e-06, "loss": 1.0739, "step": 35504 }, { "epoch": 2.782523510971787, "grad_norm": 3.5999209880828857, "learning_rate": 2.124544868822684e-06, "loss": 1.5044, "step": 35505 }, { "epoch": 2.782601880877743, "grad_norm": 2.8785653114318848, "learning_rate": 2.124292606551163e-06, "loss": 0.999, "step": 35506 }, { "epoch": 2.782680250783699, "grad_norm": 3.9047458171844482, "learning_rate": 2.1240403552173557e-06, "loss": 1.4771, "step": 35507 }, { "epoch": 2.782758620689655, "grad_norm": 3.0487873554229736, "learning_rate": 2.1237881148222235e-06, "loss": 1.4771, "step": 35508 }, { "epoch": 2.7828369905956114, "grad_norm": 4.171306133270264, "learning_rate": 2.1235358853667197e-06, "loss": 0.8598, "step": 35509 }, { "epoch": 2.7829153605015673, "grad_norm": 3.337019681930542, "learning_rate": 2.123283666851811e-06, "loss": 1.7245, "step": 35510 }, { "epoch": 2.7829937304075236, "grad_norm": 3.2167587280273438, "learning_rate": 2.123031459278453e-06, "loss": 1.4883, "step": 35511 }, { "epoch": 2.7830721003134795, "grad_norm": 3.1835827827453613, "learning_rate": 2.1227792626476073e-06, "loss": 1.1765, "step": 35512 }, { "epoch": 2.783150470219436, "grad_norm": 2.7944986820220947, "learning_rate": 2.1225270769602296e-06, "loss": 1.3331, "step": 35513 }, { "epoch": 2.7832288401253917, "grad_norm": 3.7709968090057373, "learning_rate": 2.1222749022172794e-06, "loss": 1.4419, "step": 35514 }, { "epoch": 2.783307210031348, "grad_norm": 5.3112874031066895, "learning_rate": 2.1220227384197195e-06, "loss": 1.5399, "step": 35515 }, { "epoch": 2.783385579937304, "grad_norm": 10.428930282592773, "learning_rate": 2.1217705855685083e-06, "loss": 0.8514, "step": 35516 }, { "epoch": 2.7834639498432603, "grad_norm": 3.0872464179992676, "learning_rate": 2.1215184436646016e-06, "loss": 1.2711, "step": 35517 }, { "epoch": 2.783542319749216, "grad_norm": 2.570329189300537, "learning_rate": 2.12126631270896e-06, "loss": 1.494, "step": 35518 }, { "epoch": 2.7836206896551725, "grad_norm": 4.578702926635742, "learning_rate": 2.1210141927025434e-06, "loss": 1.8386, "step": 35519 }, { "epoch": 2.7836990595611284, "grad_norm": 3.0322511196136475, "learning_rate": 2.1207620836463093e-06, "loss": 1.1678, "step": 35520 }, { "epoch": 2.7837774294670847, "grad_norm": 3.8900492191314697, "learning_rate": 2.120509985541218e-06, "loss": 1.4156, "step": 35521 }, { "epoch": 2.7838557993730406, "grad_norm": 4.723455905914307, "learning_rate": 2.1202578983882276e-06, "loss": 1.3498, "step": 35522 }, { "epoch": 2.783934169278997, "grad_norm": 5.451512813568115, "learning_rate": 2.120005822188297e-06, "loss": 1.8277, "step": 35523 }, { "epoch": 2.784012539184953, "grad_norm": 2.416006326675415, "learning_rate": 2.1197537569423847e-06, "loss": 1.5107, "step": 35524 }, { "epoch": 2.784090909090909, "grad_norm": 3.0679616928100586, "learning_rate": 2.1195017026514503e-06, "loss": 1.3218, "step": 35525 }, { "epoch": 2.784169278996865, "grad_norm": 4.79241418838501, "learning_rate": 2.1192496593164513e-06, "loss": 1.1997, "step": 35526 }, { "epoch": 2.7842476489028214, "grad_norm": 5.425882339477539, "learning_rate": 2.118997626938349e-06, "loss": 1.3618, "step": 35527 }, { "epoch": 2.7843260188087773, "grad_norm": 2.228088140487671, "learning_rate": 2.118745605518096e-06, "loss": 1.3488, "step": 35528 }, { "epoch": 2.7844043887147336, "grad_norm": 3.7645888328552246, "learning_rate": 2.118493595056657e-06, "loss": 1.0682, "step": 35529 }, { "epoch": 2.7844827586206895, "grad_norm": 2.9414052963256836, "learning_rate": 2.1182415955549905e-06, "loss": 1.3592, "step": 35530 }, { "epoch": 2.784561128526646, "grad_norm": 3.7788302898406982, "learning_rate": 2.1179896070140506e-06, "loss": 1.5025, "step": 35531 }, { "epoch": 2.7846394984326017, "grad_norm": 2.2387354373931885, "learning_rate": 2.1177376294347978e-06, "loss": 1.3956, "step": 35532 }, { "epoch": 2.784717868338558, "grad_norm": 6.4481587409973145, "learning_rate": 2.117485662818189e-06, "loss": 1.0955, "step": 35533 }, { "epoch": 2.784796238244514, "grad_norm": 7.319488525390625, "learning_rate": 2.1172337071651877e-06, "loss": 1.5868, "step": 35534 }, { "epoch": 2.7848746081504703, "grad_norm": 5.681260585784912, "learning_rate": 2.116981762476747e-06, "loss": 0.8917, "step": 35535 }, { "epoch": 2.7849529780564266, "grad_norm": 3.8731956481933594, "learning_rate": 2.116729828753827e-06, "loss": 1.4349, "step": 35536 }, { "epoch": 2.7850313479623825, "grad_norm": 2.817009449005127, "learning_rate": 2.1164779059973857e-06, "loss": 1.3278, "step": 35537 }, { "epoch": 2.7851097178683384, "grad_norm": 3.4564759731292725, "learning_rate": 2.1162259942083805e-06, "loss": 1.4452, "step": 35538 }, { "epoch": 2.7851880877742947, "grad_norm": 2.400019407272339, "learning_rate": 2.1159740933877714e-06, "loss": 1.4951, "step": 35539 }, { "epoch": 2.785266457680251, "grad_norm": 2.126084804534912, "learning_rate": 2.1157222035365154e-06, "loss": 1.1114, "step": 35540 }, { "epoch": 2.785344827586207, "grad_norm": 2.9306530952453613, "learning_rate": 2.1154703246555707e-06, "loss": 1.2294, "step": 35541 }, { "epoch": 2.785423197492163, "grad_norm": 3.5620553493499756, "learning_rate": 2.115218456745895e-06, "loss": 1.2207, "step": 35542 }, { "epoch": 2.785501567398119, "grad_norm": 2.8184709548950195, "learning_rate": 2.114966599808447e-06, "loss": 1.0931, "step": 35543 }, { "epoch": 2.7855799373040755, "grad_norm": 3.6215968132019043, "learning_rate": 2.1147147538441843e-06, "loss": 1.461, "step": 35544 }, { "epoch": 2.7856583072100314, "grad_norm": 3.0772793292999268, "learning_rate": 2.114462918854066e-06, "loss": 1.109, "step": 35545 }, { "epoch": 2.7857366771159873, "grad_norm": 3.4494218826293945, "learning_rate": 2.114211094839047e-06, "loss": 1.4995, "step": 35546 }, { "epoch": 2.7858150470219436, "grad_norm": 2.671710252761841, "learning_rate": 2.113959281800085e-06, "loss": 1.2612, "step": 35547 }, { "epoch": 2.7858934169279, "grad_norm": 2.3924901485443115, "learning_rate": 2.1137074797381413e-06, "loss": 1.2848, "step": 35548 }, { "epoch": 2.785971786833856, "grad_norm": 3.119072675704956, "learning_rate": 2.1134556886541737e-06, "loss": 1.4893, "step": 35549 }, { "epoch": 2.7860501567398117, "grad_norm": 4.134799003601074, "learning_rate": 2.1132039085491363e-06, "loss": 1.3298, "step": 35550 }, { "epoch": 2.786128526645768, "grad_norm": 3.342655658721924, "learning_rate": 2.1129521394239883e-06, "loss": 1.1729, "step": 35551 }, { "epoch": 2.7862068965517244, "grad_norm": 14.416519165039062, "learning_rate": 2.1127003812796853e-06, "loss": 1.3869, "step": 35552 }, { "epoch": 2.7862852664576803, "grad_norm": 4.706927299499512, "learning_rate": 2.112448634117191e-06, "loss": 1.4528, "step": 35553 }, { "epoch": 2.786363636363636, "grad_norm": 5.292647838592529, "learning_rate": 2.1121968979374575e-06, "loss": 1.1857, "step": 35554 }, { "epoch": 2.7864420062695925, "grad_norm": 4.720642566680908, "learning_rate": 2.1119451727414435e-06, "loss": 1.0845, "step": 35555 }, { "epoch": 2.786520376175549, "grad_norm": 3.834174394607544, "learning_rate": 2.111693458530107e-06, "loss": 1.8101, "step": 35556 }, { "epoch": 2.7865987460815047, "grad_norm": 4.546109676361084, "learning_rate": 2.1114417553044043e-06, "loss": 1.3197, "step": 35557 }, { "epoch": 2.7866771159874606, "grad_norm": 2.062196731567383, "learning_rate": 2.1111900630652942e-06, "loss": 1.3397, "step": 35558 }, { "epoch": 2.786755485893417, "grad_norm": 5.769120693206787, "learning_rate": 2.110938381813733e-06, "loss": 1.4393, "step": 35559 }, { "epoch": 2.7868338557993733, "grad_norm": 6.325442790985107, "learning_rate": 2.110686711550678e-06, "loss": 1.3853, "step": 35560 }, { "epoch": 2.786912225705329, "grad_norm": 2.11250901222229, "learning_rate": 2.110435052277087e-06, "loss": 1.3862, "step": 35561 }, { "epoch": 2.786990595611285, "grad_norm": 3.5068187713623047, "learning_rate": 2.110183403993917e-06, "loss": 1.1281, "step": 35562 }, { "epoch": 2.7870689655172414, "grad_norm": 2.4715309143066406, "learning_rate": 2.1099317667021252e-06, "loss": 1.2471, "step": 35563 }, { "epoch": 2.7871473354231977, "grad_norm": 2.4518630504608154, "learning_rate": 2.10968014040267e-06, "loss": 1.3925, "step": 35564 }, { "epoch": 2.7872257053291536, "grad_norm": 2.887272834777832, "learning_rate": 2.1094285250965034e-06, "loss": 1.3778, "step": 35565 }, { "epoch": 2.7873040752351095, "grad_norm": 2.8831944465637207, "learning_rate": 2.1091769207845873e-06, "loss": 1.3766, "step": 35566 }, { "epoch": 2.787382445141066, "grad_norm": 2.8184797763824463, "learning_rate": 2.1089253274678785e-06, "loss": 1.3947, "step": 35567 }, { "epoch": 2.787460815047022, "grad_norm": 1.9187185764312744, "learning_rate": 2.108673745147334e-06, "loss": 1.3025, "step": 35568 }, { "epoch": 2.787539184952978, "grad_norm": 4.923786640167236, "learning_rate": 2.1084221738239075e-06, "loss": 1.4671, "step": 35569 }, { "epoch": 2.787617554858934, "grad_norm": 2.8405954837799072, "learning_rate": 2.1081706134985564e-06, "loss": 1.2295, "step": 35570 }, { "epoch": 2.7876959247648903, "grad_norm": 2.854640007019043, "learning_rate": 2.1079190641722402e-06, "loss": 1.4778, "step": 35571 }, { "epoch": 2.7877742946708466, "grad_norm": 2.9845008850097656, "learning_rate": 2.1076675258459166e-06, "loss": 1.2828, "step": 35572 }, { "epoch": 2.7878526645768025, "grad_norm": 3.2724859714508057, "learning_rate": 2.1074159985205387e-06, "loss": 1.1041, "step": 35573 }, { "epoch": 2.7879310344827584, "grad_norm": 3.0707507133483887, "learning_rate": 2.1071644821970644e-06, "loss": 1.5168, "step": 35574 }, { "epoch": 2.7880094043887147, "grad_norm": 2.791857957839966, "learning_rate": 2.10691297687645e-06, "loss": 1.2583, "step": 35575 }, { "epoch": 2.788087774294671, "grad_norm": 10.565272331237793, "learning_rate": 2.1066614825596532e-06, "loss": 1.3089, "step": 35576 }, { "epoch": 2.788166144200627, "grad_norm": 2.989633321762085, "learning_rate": 2.10640999924763e-06, "loss": 1.6582, "step": 35577 }, { "epoch": 2.788244514106583, "grad_norm": 2.9855313301086426, "learning_rate": 2.106158526941337e-06, "loss": 1.5561, "step": 35578 }, { "epoch": 2.788322884012539, "grad_norm": 2.6885933876037598, "learning_rate": 2.10590706564173e-06, "loss": 1.3699, "step": 35579 }, { "epoch": 2.7884012539184955, "grad_norm": 2.3209550380706787, "learning_rate": 2.1056556153497666e-06, "loss": 1.5204, "step": 35580 }, { "epoch": 2.7884796238244514, "grad_norm": 3.9936342239379883, "learning_rate": 2.1054041760664024e-06, "loss": 1.5819, "step": 35581 }, { "epoch": 2.7885579937304072, "grad_norm": 2.6250176429748535, "learning_rate": 2.105152747792594e-06, "loss": 1.0421, "step": 35582 }, { "epoch": 2.7886363636363636, "grad_norm": 4.456374645233154, "learning_rate": 2.1049013305292994e-06, "loss": 1.0346, "step": 35583 }, { "epoch": 2.78871473354232, "grad_norm": 3.8244004249572754, "learning_rate": 2.1046499242774698e-06, "loss": 1.2843, "step": 35584 }, { "epoch": 2.788793103448276, "grad_norm": 2.626877784729004, "learning_rate": 2.104398529038066e-06, "loss": 1.4299, "step": 35585 }, { "epoch": 2.788871473354232, "grad_norm": 4.403378486633301, "learning_rate": 2.1041471448120452e-06, "loss": 1.3587, "step": 35586 }, { "epoch": 2.788949843260188, "grad_norm": 5.866201877593994, "learning_rate": 2.1038957716003595e-06, "loss": 2.1905, "step": 35587 }, { "epoch": 2.7890282131661444, "grad_norm": 3.367074728012085, "learning_rate": 2.1036444094039667e-06, "loss": 1.2269, "step": 35588 }, { "epoch": 2.7891065830721002, "grad_norm": 7.463663101196289, "learning_rate": 2.1033930582238205e-06, "loss": 1.1089, "step": 35589 }, { "epoch": 2.7891849529780566, "grad_norm": 6.714097023010254, "learning_rate": 2.103141718060883e-06, "loss": 1.0963, "step": 35590 }, { "epoch": 2.7892633228840125, "grad_norm": 5.0639262199401855, "learning_rate": 2.1028903889161056e-06, "loss": 1.2578, "step": 35591 }, { "epoch": 2.789341692789969, "grad_norm": 2.5829367637634277, "learning_rate": 2.1026390707904444e-06, "loss": 1.2933, "step": 35592 }, { "epoch": 2.7894200626959247, "grad_norm": 2.396874189376831, "learning_rate": 2.102387763684856e-06, "loss": 1.02, "step": 35593 }, { "epoch": 2.789498432601881, "grad_norm": 2.9758188724517822, "learning_rate": 2.1021364676002965e-06, "loss": 1.1234, "step": 35594 }, { "epoch": 2.789576802507837, "grad_norm": 3.8973166942596436, "learning_rate": 2.101885182537721e-06, "loss": 1.3905, "step": 35595 }, { "epoch": 2.7896551724137932, "grad_norm": 3.8191397190093994, "learning_rate": 2.1016339084980864e-06, "loss": 1.735, "step": 35596 }, { "epoch": 2.789733542319749, "grad_norm": 2.582012891769409, "learning_rate": 2.101382645482347e-06, "loss": 1.2884, "step": 35597 }, { "epoch": 2.7898119122257055, "grad_norm": 2.5207691192626953, "learning_rate": 2.1011313934914595e-06, "loss": 1.4404, "step": 35598 }, { "epoch": 2.7898902821316613, "grad_norm": 3.5064048767089844, "learning_rate": 2.1008801525263794e-06, "loss": 1.5167, "step": 35599 }, { "epoch": 2.7899686520376177, "grad_norm": 3.301492214202881, "learning_rate": 2.1006289225880623e-06, "loss": 1.5826, "step": 35600 }, { "epoch": 2.7900470219435736, "grad_norm": 5.114290237426758, "learning_rate": 2.100377703677465e-06, "loss": 0.9388, "step": 35601 }, { "epoch": 2.79012539184953, "grad_norm": 5.822072505950928, "learning_rate": 2.10012649579554e-06, "loss": 1.3881, "step": 35602 }, { "epoch": 2.790203761755486, "grad_norm": 3.1002988815307617, "learning_rate": 2.099875298943242e-06, "loss": 1.4157, "step": 35603 }, { "epoch": 2.790282131661442, "grad_norm": 4.090396881103516, "learning_rate": 2.0996241131215313e-06, "loss": 1.1121, "step": 35604 }, { "epoch": 2.790360501567398, "grad_norm": 2.4423909187316895, "learning_rate": 2.0993729383313615e-06, "loss": 1.2959, "step": 35605 }, { "epoch": 2.7904388714733543, "grad_norm": 2.81744384765625, "learning_rate": 2.0991217745736865e-06, "loss": 1.4538, "step": 35606 }, { "epoch": 2.7905172413793102, "grad_norm": 5.3830671310424805, "learning_rate": 2.098870621849462e-06, "loss": 1.1247, "step": 35607 }, { "epoch": 2.7905956112852666, "grad_norm": 3.417515277862549, "learning_rate": 2.0986194801596416e-06, "loss": 1.561, "step": 35608 }, { "epoch": 2.7906739811912225, "grad_norm": 2.4588019847869873, "learning_rate": 2.098368349505186e-06, "loss": 1.4109, "step": 35609 }, { "epoch": 2.790752351097179, "grad_norm": 3.116185426712036, "learning_rate": 2.098117229887045e-06, "loss": 1.3359, "step": 35610 }, { "epoch": 2.7908307210031347, "grad_norm": 2.6687111854553223, "learning_rate": 2.097866121306175e-06, "loss": 1.3696, "step": 35611 }, { "epoch": 2.790909090909091, "grad_norm": 2.653294324874878, "learning_rate": 2.097615023763533e-06, "loss": 1.2976, "step": 35612 }, { "epoch": 2.790987460815047, "grad_norm": 2.5741336345672607, "learning_rate": 2.0973639372600717e-06, "loss": 1.0056, "step": 35613 }, { "epoch": 2.7910658307210032, "grad_norm": 3.3132359981536865, "learning_rate": 2.0971128617967473e-06, "loss": 1.4974, "step": 35614 }, { "epoch": 2.791144200626959, "grad_norm": 6.722320556640625, "learning_rate": 2.0968617973745148e-06, "loss": 1.7528, "step": 35615 }, { "epoch": 2.7912225705329154, "grad_norm": 3.0952799320220947, "learning_rate": 2.09661074399433e-06, "loss": 1.3376, "step": 35616 }, { "epoch": 2.7913009404388713, "grad_norm": 4.292423725128174, "learning_rate": 2.096359701657144e-06, "loss": 1.6996, "step": 35617 }, { "epoch": 2.7913793103448277, "grad_norm": 4.849011421203613, "learning_rate": 2.0961086703639155e-06, "loss": 1.7317, "step": 35618 }, { "epoch": 2.7914576802507836, "grad_norm": 3.665119171142578, "learning_rate": 2.095857650115598e-06, "loss": 1.4888, "step": 35619 }, { "epoch": 2.79153605015674, "grad_norm": 3.8786182403564453, "learning_rate": 2.095606640913149e-06, "loss": 1.1347, "step": 35620 }, { "epoch": 2.791614420062696, "grad_norm": 2.775682210922241, "learning_rate": 2.0953556427575178e-06, "loss": 1.2707, "step": 35621 }, { "epoch": 2.791692789968652, "grad_norm": 5.391411304473877, "learning_rate": 2.09510465564966e-06, "loss": 1.129, "step": 35622 }, { "epoch": 2.791771159874608, "grad_norm": 2.398660898208618, "learning_rate": 2.0948536795905343e-06, "loss": 1.1598, "step": 35623 }, { "epoch": 2.7918495297805643, "grad_norm": 2.2751972675323486, "learning_rate": 2.094602714581095e-06, "loss": 1.2049, "step": 35624 }, { "epoch": 2.79192789968652, "grad_norm": 3.193943977355957, "learning_rate": 2.0943517606222924e-06, "loss": 1.3868, "step": 35625 }, { "epoch": 2.7920062695924766, "grad_norm": 2.8637773990631104, "learning_rate": 2.0941008177150833e-06, "loss": 1.1552, "step": 35626 }, { "epoch": 2.7920846394984324, "grad_norm": 3.3977396488189697, "learning_rate": 2.093849885860422e-06, "loss": 1.2006, "step": 35627 }, { "epoch": 2.7921630094043888, "grad_norm": 3.1685147285461426, "learning_rate": 2.0935989650592636e-06, "loss": 1.4631, "step": 35628 }, { "epoch": 2.7922413793103447, "grad_norm": 3.564852476119995, "learning_rate": 2.0933480553125614e-06, "loss": 1.2454, "step": 35629 }, { "epoch": 2.792319749216301, "grad_norm": 7.519019603729248, "learning_rate": 2.0930971566212705e-06, "loss": 0.5708, "step": 35630 }, { "epoch": 2.7923981191222573, "grad_norm": 10.826872825622559, "learning_rate": 2.092846268986345e-06, "loss": 1.6141, "step": 35631 }, { "epoch": 2.792476489028213, "grad_norm": 3.068711757659912, "learning_rate": 2.092595392408739e-06, "loss": 1.2204, "step": 35632 }, { "epoch": 2.792554858934169, "grad_norm": 3.59561824798584, "learning_rate": 2.0923445268894064e-06, "loss": 1.4544, "step": 35633 }, { "epoch": 2.7926332288401254, "grad_norm": 6.8372297286987305, "learning_rate": 2.0920936724293023e-06, "loss": 1.3287, "step": 35634 }, { "epoch": 2.7927115987460818, "grad_norm": 3.3146259784698486, "learning_rate": 2.0918428290293817e-06, "loss": 1.5894, "step": 35635 }, { "epoch": 2.7927899686520377, "grad_norm": 3.0364935398101807, "learning_rate": 2.091591996690594e-06, "loss": 1.5551, "step": 35636 }, { "epoch": 2.7928683385579935, "grad_norm": 2.578510046005249, "learning_rate": 2.0913411754138977e-06, "loss": 1.2735, "step": 35637 }, { "epoch": 2.79294670846395, "grad_norm": 4.4327521324157715, "learning_rate": 2.091090365200246e-06, "loss": 1.8774, "step": 35638 }, { "epoch": 2.793025078369906, "grad_norm": 3.4668312072753906, "learning_rate": 2.0908395660505947e-06, "loss": 0.9054, "step": 35639 }, { "epoch": 2.793103448275862, "grad_norm": 4.148682594299316, "learning_rate": 2.0905887779658935e-06, "loss": 1.5493, "step": 35640 }, { "epoch": 2.793181818181818, "grad_norm": 3.006535768508911, "learning_rate": 2.090338000947096e-06, "loss": 1.1443, "step": 35641 }, { "epoch": 2.7932601880877743, "grad_norm": 3.3324685096740723, "learning_rate": 2.090087234995162e-06, "loss": 1.4944, "step": 35642 }, { "epoch": 2.7933385579937307, "grad_norm": 4.394645690917969, "learning_rate": 2.0898364801110393e-06, "loss": 1.1469, "step": 35643 }, { "epoch": 2.7934169278996865, "grad_norm": 3.957318067550659, "learning_rate": 2.0895857362956846e-06, "loss": 1.6589, "step": 35644 }, { "epoch": 2.7934952978056424, "grad_norm": 3.668762683868408, "learning_rate": 2.0893350035500506e-06, "loss": 0.7739, "step": 35645 }, { "epoch": 2.7935736677115988, "grad_norm": 3.260294198989868, "learning_rate": 2.089084281875091e-06, "loss": 1.4545, "step": 35646 }, { "epoch": 2.793652037617555, "grad_norm": 19.30327033996582, "learning_rate": 2.08883357127176e-06, "loss": 0.9927, "step": 35647 }, { "epoch": 2.793730407523511, "grad_norm": 3.218071460723877, "learning_rate": 2.0885828717410107e-06, "loss": 1.7159, "step": 35648 }, { "epoch": 2.793808777429467, "grad_norm": 5.513207912445068, "learning_rate": 2.0883321832837966e-06, "loss": 0.6388, "step": 35649 }, { "epoch": 2.793887147335423, "grad_norm": 2.9679243564605713, "learning_rate": 2.0880815059010716e-06, "loss": 1.4594, "step": 35650 }, { "epoch": 2.7939655172413795, "grad_norm": 4.055312633514404, "learning_rate": 2.0878308395937886e-06, "loss": 1.5713, "step": 35651 }, { "epoch": 2.7940438871473354, "grad_norm": 4.169712066650391, "learning_rate": 2.0875801843629018e-06, "loss": 1.5296, "step": 35652 }, { "epoch": 2.7941222570532913, "grad_norm": 2.6376161575317383, "learning_rate": 2.0873295402093636e-06, "loss": 1.0734, "step": 35653 }, { "epoch": 2.7942006269592476, "grad_norm": 2.1500046253204346, "learning_rate": 2.087078907134128e-06, "loss": 1.3225, "step": 35654 }, { "epoch": 2.794278996865204, "grad_norm": 2.7256996631622314, "learning_rate": 2.0868282851381478e-06, "loss": 1.8165, "step": 35655 }, { "epoch": 2.79435736677116, "grad_norm": 2.301969528198242, "learning_rate": 2.0865776742223764e-06, "loss": 1.2457, "step": 35656 }, { "epoch": 2.7944357366771158, "grad_norm": 2.614147186279297, "learning_rate": 2.0863270743877696e-06, "loss": 1.3704, "step": 35657 }, { "epoch": 2.794514106583072, "grad_norm": 4.7411208152771, "learning_rate": 2.0860764856352755e-06, "loss": 1.102, "step": 35658 }, { "epoch": 2.7945924764890284, "grad_norm": 9.963691711425781, "learning_rate": 2.085825907965849e-06, "loss": 1.234, "step": 35659 }, { "epoch": 2.7946708463949843, "grad_norm": 4.324191093444824, "learning_rate": 2.0855753413804446e-06, "loss": 1.3632, "step": 35660 }, { "epoch": 2.79474921630094, "grad_norm": 5.600715160369873, "learning_rate": 2.085324785880017e-06, "loss": 1.1338, "step": 35661 }, { "epoch": 2.7948275862068965, "grad_norm": 3.473708152770996, "learning_rate": 2.085074241465515e-06, "loss": 1.1625, "step": 35662 }, { "epoch": 2.794905956112853, "grad_norm": 5.9463276863098145, "learning_rate": 2.084823708137893e-06, "loss": 1.6263, "step": 35663 }, { "epoch": 2.7949843260188088, "grad_norm": 2.834772825241089, "learning_rate": 2.084573185898105e-06, "loss": 0.9582, "step": 35664 }, { "epoch": 2.7950626959247646, "grad_norm": 4.8239312171936035, "learning_rate": 2.0843226747471025e-06, "loss": 1.9398, "step": 35665 }, { "epoch": 2.795141065830721, "grad_norm": 3.38590407371521, "learning_rate": 2.0840721746858393e-06, "loss": 1.2755, "step": 35666 }, { "epoch": 2.7952194357366773, "grad_norm": 6.907811164855957, "learning_rate": 2.0838216857152677e-06, "loss": 1.5085, "step": 35667 }, { "epoch": 2.795297805642633, "grad_norm": 3.0046539306640625, "learning_rate": 2.0835712078363405e-06, "loss": 1.6225, "step": 35668 }, { "epoch": 2.795376175548589, "grad_norm": 4.4415602684021, "learning_rate": 2.0833207410500107e-06, "loss": 1.4165, "step": 35669 }, { "epoch": 2.7954545454545454, "grad_norm": 3.448223352432251, "learning_rate": 2.0830702853572303e-06, "loss": 1.2662, "step": 35670 }, { "epoch": 2.7955329153605017, "grad_norm": 2.94134521484375, "learning_rate": 2.0828198407589527e-06, "loss": 1.4481, "step": 35671 }, { "epoch": 2.7956112852664576, "grad_norm": 6.681233882904053, "learning_rate": 2.082569407256132e-06, "loss": 0.6886, "step": 35672 }, { "epoch": 2.7956896551724135, "grad_norm": 2.6201395988464355, "learning_rate": 2.082318984849715e-06, "loss": 0.9794, "step": 35673 }, { "epoch": 2.79576802507837, "grad_norm": 3.43942928314209, "learning_rate": 2.082068573540659e-06, "loss": 1.3075, "step": 35674 }, { "epoch": 2.795846394984326, "grad_norm": 3.8303112983703613, "learning_rate": 2.0818181733299162e-06, "loss": 1.1053, "step": 35675 }, { "epoch": 2.795924764890282, "grad_norm": 5.454083442687988, "learning_rate": 2.08156778421844e-06, "loss": 2.0275, "step": 35676 }, { "epoch": 2.7960031347962384, "grad_norm": 2.6967508792877197, "learning_rate": 2.0813174062071793e-06, "loss": 1.4426, "step": 35677 }, { "epoch": 2.7960815047021943, "grad_norm": 5.061869144439697, "learning_rate": 2.081067039297086e-06, "loss": 1.4667, "step": 35678 }, { "epoch": 2.7961598746081506, "grad_norm": 4.437573432922363, "learning_rate": 2.0808166834891163e-06, "loss": 1.6505, "step": 35679 }, { "epoch": 2.7962382445141065, "grad_norm": 4.372379302978516, "learning_rate": 2.080566338784222e-06, "loss": 1.4172, "step": 35680 }, { "epoch": 2.796316614420063, "grad_norm": 6.852684497833252, "learning_rate": 2.0803160051833527e-06, "loss": 1.2728, "step": 35681 }, { "epoch": 2.7963949843260187, "grad_norm": 2.578077554702759, "learning_rate": 2.080065682687461e-06, "loss": 1.3618, "step": 35682 }, { "epoch": 2.796473354231975, "grad_norm": 3.835023880004883, "learning_rate": 2.0798153712975e-06, "loss": 1.2937, "step": 35683 }, { "epoch": 2.796551724137931, "grad_norm": 6.120706081390381, "learning_rate": 2.079565071014421e-06, "loss": 1.4177, "step": 35684 }, { "epoch": 2.7966300940438873, "grad_norm": 5.220058441162109, "learning_rate": 2.0793147818391774e-06, "loss": 1.1945, "step": 35685 }, { "epoch": 2.796708463949843, "grad_norm": 3.9191672801971436, "learning_rate": 2.079064503772719e-06, "loss": 1.2957, "step": 35686 }, { "epoch": 2.7967868338557995, "grad_norm": 3.530447006225586, "learning_rate": 2.0788142368159997e-06, "loss": 1.6871, "step": 35687 }, { "epoch": 2.7968652037617554, "grad_norm": 3.6385443210601807, "learning_rate": 2.0785639809699706e-06, "loss": 1.2712, "step": 35688 }, { "epoch": 2.7969435736677117, "grad_norm": 2.765712261199951, "learning_rate": 2.078313736235584e-06, "loss": 1.2255, "step": 35689 }, { "epoch": 2.7970219435736676, "grad_norm": 18.627399444580078, "learning_rate": 2.07806350261379e-06, "loss": 0.9302, "step": 35690 }, { "epoch": 2.797100313479624, "grad_norm": 2.9810047149658203, "learning_rate": 2.0778132801055447e-06, "loss": 1.5207, "step": 35691 }, { "epoch": 2.79717868338558, "grad_norm": 3.801682233810425, "learning_rate": 2.077563068711792e-06, "loss": 0.8146, "step": 35692 }, { "epoch": 2.797257053291536, "grad_norm": 2.976414680480957, "learning_rate": 2.077312868433491e-06, "loss": 1.4611, "step": 35693 }, { "epoch": 2.797335423197492, "grad_norm": 3.132289409637451, "learning_rate": 2.0770626792715903e-06, "loss": 1.2056, "step": 35694 }, { "epoch": 2.7974137931034484, "grad_norm": 3.7758901119232178, "learning_rate": 2.076812501227044e-06, "loss": 1.554, "step": 35695 }, { "epoch": 2.7974921630094043, "grad_norm": 3.7943830490112305, "learning_rate": 2.0765623343007997e-06, "loss": 1.2287, "step": 35696 }, { "epoch": 2.7975705329153606, "grad_norm": 4.353226661682129, "learning_rate": 2.076312178493808e-06, "loss": 1.713, "step": 35697 }, { "epoch": 2.7976489028213165, "grad_norm": 2.8162248134613037, "learning_rate": 2.0760620338070277e-06, "loss": 1.1431, "step": 35698 }, { "epoch": 2.797727272727273, "grad_norm": 2.708136796951294, "learning_rate": 2.0758119002414036e-06, "loss": 1.1842, "step": 35699 }, { "epoch": 2.7978056426332287, "grad_norm": 2.344447374343872, "learning_rate": 2.0755617777978888e-06, "loss": 1.2399, "step": 35700 }, { "epoch": 2.797884012539185, "grad_norm": 3.3670296669006348, "learning_rate": 2.075311666477435e-06, "loss": 1.5391, "step": 35701 }, { "epoch": 2.797962382445141, "grad_norm": 7.137719631195068, "learning_rate": 2.075061566280993e-06, "loss": 1.0625, "step": 35702 }, { "epoch": 2.7980407523510973, "grad_norm": 3.731344223022461, "learning_rate": 2.0748114772095153e-06, "loss": 1.5088, "step": 35703 }, { "epoch": 2.798119122257053, "grad_norm": 3.6286323070526123, "learning_rate": 2.074561399263952e-06, "loss": 1.2409, "step": 35704 }, { "epoch": 2.7981974921630095, "grad_norm": 2.6430952548980713, "learning_rate": 2.074311332445254e-06, "loss": 1.1734, "step": 35705 }, { "epoch": 2.7982758620689654, "grad_norm": 2.4186980724334717, "learning_rate": 2.0740612767543734e-06, "loss": 1.4203, "step": 35706 }, { "epoch": 2.7983542319749217, "grad_norm": 5.2230095863342285, "learning_rate": 2.073811232192261e-06, "loss": 1.4621, "step": 35707 }, { "epoch": 2.7984326018808776, "grad_norm": 2.7927160263061523, "learning_rate": 2.0735611987598673e-06, "loss": 1.5508, "step": 35708 }, { "epoch": 2.798510971786834, "grad_norm": 2.619058132171631, "learning_rate": 2.0733111764581436e-06, "loss": 1.5461, "step": 35709 }, { "epoch": 2.79858934169279, "grad_norm": 3.4182605743408203, "learning_rate": 2.0730611652880435e-06, "loss": 1.4654, "step": 35710 }, { "epoch": 2.798667711598746, "grad_norm": 2.5663414001464844, "learning_rate": 2.0728111652505113e-06, "loss": 1.6318, "step": 35711 }, { "epoch": 2.798746081504702, "grad_norm": 2.501401424407959, "learning_rate": 2.072561176346504e-06, "loss": 1.1125, "step": 35712 }, { "epoch": 2.7988244514106584, "grad_norm": 4.718419551849365, "learning_rate": 2.072311198576971e-06, "loss": 1.1827, "step": 35713 }, { "epoch": 2.7989028213166143, "grad_norm": 2.7854225635528564, "learning_rate": 2.0720612319428617e-06, "loss": 1.3179, "step": 35714 }, { "epoch": 2.7989811912225706, "grad_norm": 5.24101448059082, "learning_rate": 2.0718112764451277e-06, "loss": 1.4858, "step": 35715 }, { "epoch": 2.7990595611285265, "grad_norm": 4.314253807067871, "learning_rate": 2.0715613320847174e-06, "loss": 1.2463, "step": 35716 }, { "epoch": 2.799137931034483, "grad_norm": 2.2197160720825195, "learning_rate": 2.071311398862587e-06, "loss": 0.9843, "step": 35717 }, { "epoch": 2.7992163009404387, "grad_norm": 5.2508368492126465, "learning_rate": 2.071061476779683e-06, "loss": 1.2056, "step": 35718 }, { "epoch": 2.799294670846395, "grad_norm": 3.2038753032684326, "learning_rate": 2.0708115658369554e-06, "loss": 1.4844, "step": 35719 }, { "epoch": 2.799373040752351, "grad_norm": 2.5120797157287598, "learning_rate": 2.0705616660353566e-06, "loss": 1.011, "step": 35720 }, { "epoch": 2.7994514106583073, "grad_norm": 5.224311351776123, "learning_rate": 2.0703117773758367e-06, "loss": 1.1803, "step": 35721 }, { "epoch": 2.799529780564263, "grad_norm": 5.1932501792907715, "learning_rate": 2.0700618998593465e-06, "loss": 1.7906, "step": 35722 }, { "epoch": 2.7996081504702195, "grad_norm": 4.4902215003967285, "learning_rate": 2.0698120334868356e-06, "loss": 1.2896, "step": 35723 }, { "epoch": 2.7996865203761754, "grad_norm": 2.7329118251800537, "learning_rate": 2.0695621782592538e-06, "loss": 1.326, "step": 35724 }, { "epoch": 2.7997648902821317, "grad_norm": 8.242574691772461, "learning_rate": 2.0693123341775532e-06, "loss": 1.6512, "step": 35725 }, { "epoch": 2.799843260188088, "grad_norm": 6.2719926834106445, "learning_rate": 2.069062501242683e-06, "loss": 0.4306, "step": 35726 }, { "epoch": 2.799921630094044, "grad_norm": 2.773435115814209, "learning_rate": 2.068812679455594e-06, "loss": 1.2958, "step": 35727 }, { "epoch": 2.8, "grad_norm": 2.5491480827331543, "learning_rate": 2.0685628688172373e-06, "loss": 1.6052, "step": 35728 }, { "epoch": 2.800078369905956, "grad_norm": 2.9026589393615723, "learning_rate": 2.0683130693285603e-06, "loss": 1.1795, "step": 35729 }, { "epoch": 2.8001567398119125, "grad_norm": 2.697984218597412, "learning_rate": 2.0680632809905122e-06, "loss": 1.1498, "step": 35730 }, { "epoch": 2.8002351097178684, "grad_norm": 3.1545820236206055, "learning_rate": 2.0678135038040476e-06, "loss": 1.1036, "step": 35731 }, { "epoch": 2.8003134796238243, "grad_norm": 6.038059234619141, "learning_rate": 2.0675637377701164e-06, "loss": 1.4677, "step": 35732 }, { "epoch": 2.8003918495297806, "grad_norm": 3.300326108932495, "learning_rate": 2.0673139828896643e-06, "loss": 0.9272, "step": 35733 }, { "epoch": 2.800470219435737, "grad_norm": 2.3090591430664062, "learning_rate": 2.0670642391636434e-06, "loss": 0.4008, "step": 35734 }, { "epoch": 2.800548589341693, "grad_norm": 3.763489246368408, "learning_rate": 2.066814506593003e-06, "loss": 1.2446, "step": 35735 }, { "epoch": 2.8006269592476487, "grad_norm": 3.8994557857513428, "learning_rate": 2.0665647851786963e-06, "loss": 1.0999, "step": 35736 }, { "epoch": 2.800705329153605, "grad_norm": 3.74688720703125, "learning_rate": 2.0663150749216687e-06, "loss": 1.5296, "step": 35737 }, { "epoch": 2.8007836990595614, "grad_norm": 2.352590560913086, "learning_rate": 2.0660653758228716e-06, "loss": 1.3712, "step": 35738 }, { "epoch": 2.8008620689655173, "grad_norm": 3.527371406555176, "learning_rate": 2.065815687883255e-06, "loss": 1.2198, "step": 35739 }, { "epoch": 2.800940438871473, "grad_norm": 3.7371487617492676, "learning_rate": 2.0655660111037685e-06, "loss": 0.9102, "step": 35740 }, { "epoch": 2.8010188087774295, "grad_norm": 3.0318191051483154, "learning_rate": 2.065316345485362e-06, "loss": 1.011, "step": 35741 }, { "epoch": 2.801097178683386, "grad_norm": 2.312694549560547, "learning_rate": 2.065066691028984e-06, "loss": 1.3468, "step": 35742 }, { "epoch": 2.8011755485893417, "grad_norm": 5.208120346069336, "learning_rate": 2.0648170477355874e-06, "loss": 1.6292, "step": 35743 }, { "epoch": 2.8012539184952976, "grad_norm": 6.1453094482421875, "learning_rate": 2.0645674156061153e-06, "loss": 1.2149, "step": 35744 }, { "epoch": 2.801332288401254, "grad_norm": 5.736386775970459, "learning_rate": 2.0643177946415227e-06, "loss": 1.1905, "step": 35745 }, { "epoch": 2.8014106583072103, "grad_norm": 2.729789972305298, "learning_rate": 2.064068184842757e-06, "loss": 1.0392, "step": 35746 }, { "epoch": 2.801489028213166, "grad_norm": 3.453047752380371, "learning_rate": 2.06381858621077e-06, "loss": 1.2961, "step": 35747 }, { "epoch": 2.801567398119122, "grad_norm": 4.381157875061035, "learning_rate": 2.0635689987465075e-06, "loss": 1.5883, "step": 35748 }, { "epoch": 2.8016457680250784, "grad_norm": 2.652120590209961, "learning_rate": 2.0633194224509186e-06, "loss": 1.4942, "step": 35749 }, { "epoch": 2.8017241379310347, "grad_norm": 2.701268196105957, "learning_rate": 2.0630698573249553e-06, "loss": 1.4621, "step": 35750 }, { "epoch": 2.8018025078369906, "grad_norm": 3.3069558143615723, "learning_rate": 2.062820303369568e-06, "loss": 1.4091, "step": 35751 }, { "epoch": 2.8018808777429465, "grad_norm": 3.1606996059417725, "learning_rate": 2.062570760585702e-06, "loss": 1.4723, "step": 35752 }, { "epoch": 2.801959247648903, "grad_norm": 7.719593048095703, "learning_rate": 2.0623212289743078e-06, "loss": 0.8103, "step": 35753 }, { "epoch": 2.802037617554859, "grad_norm": 4.007246971130371, "learning_rate": 2.062071708536335e-06, "loss": 1.3446, "step": 35754 }, { "epoch": 2.802115987460815, "grad_norm": 3.3871304988861084, "learning_rate": 2.0618221992727324e-06, "loss": 1.4203, "step": 35755 }, { "epoch": 2.802194357366771, "grad_norm": 2.568784713745117, "learning_rate": 2.0615727011844485e-06, "loss": 1.0284, "step": 35756 }, { "epoch": 2.8022727272727272, "grad_norm": 2.7971341609954834, "learning_rate": 2.061323214272433e-06, "loss": 1.3292, "step": 35757 }, { "epoch": 2.8023510971786836, "grad_norm": 5.312570095062256, "learning_rate": 2.061073738537635e-06, "loss": 1.3658, "step": 35758 }, { "epoch": 2.8024294670846395, "grad_norm": 8.622823715209961, "learning_rate": 2.0608242739810023e-06, "loss": 0.5674, "step": 35759 }, { "epoch": 2.8025078369905954, "grad_norm": 3.785923957824707, "learning_rate": 2.0605748206034847e-06, "loss": 1.6347, "step": 35760 }, { "epoch": 2.8025862068965517, "grad_norm": 2.9502406120300293, "learning_rate": 2.060325378406031e-06, "loss": 1.095, "step": 35761 }, { "epoch": 2.802664576802508, "grad_norm": 2.5666911602020264, "learning_rate": 2.06007594738959e-06, "loss": 1.3261, "step": 35762 }, { "epoch": 2.802742946708464, "grad_norm": 5.580362319946289, "learning_rate": 2.059826527555109e-06, "loss": 1.2231, "step": 35763 }, { "epoch": 2.80282131661442, "grad_norm": 2.600811243057251, "learning_rate": 2.0595771189035385e-06, "loss": 1.4797, "step": 35764 }, { "epoch": 2.802899686520376, "grad_norm": 2.7413604259490967, "learning_rate": 2.059327721435826e-06, "loss": 1.2711, "step": 35765 }, { "epoch": 2.8029780564263325, "grad_norm": 8.780158996582031, "learning_rate": 2.059078335152922e-06, "loss": 0.8287, "step": 35766 }, { "epoch": 2.8030564263322884, "grad_norm": 4.176778316497803, "learning_rate": 2.0588289600557705e-06, "loss": 1.0733, "step": 35767 }, { "epoch": 2.8031347962382442, "grad_norm": 5.432452201843262, "learning_rate": 2.0585795961453243e-06, "loss": 1.4103, "step": 35768 }, { "epoch": 2.8032131661442006, "grad_norm": 3.812739849090576, "learning_rate": 2.058330243422532e-06, "loss": 1.3656, "step": 35769 }, { "epoch": 2.803291536050157, "grad_norm": 3.6809561252593994, "learning_rate": 2.0580809018883397e-06, "loss": 1.5054, "step": 35770 }, { "epoch": 2.803369905956113, "grad_norm": 3.6332602500915527, "learning_rate": 2.057831571543696e-06, "loss": 1.0906, "step": 35771 }, { "epoch": 2.803448275862069, "grad_norm": 7.08450174331665, "learning_rate": 2.057582252389548e-06, "loss": 1.3132, "step": 35772 }, { "epoch": 2.803526645768025, "grad_norm": 7.828245639801025, "learning_rate": 2.057332944426849e-06, "loss": 1.087, "step": 35773 }, { "epoch": 2.8036050156739813, "grad_norm": 4.9465131759643555, "learning_rate": 2.0570836476565427e-06, "loss": 1.8605, "step": 35774 }, { "epoch": 2.8036833855799372, "grad_norm": 4.503066062927246, "learning_rate": 2.056834362079578e-06, "loss": 1.4334, "step": 35775 }, { "epoch": 2.8037617554858936, "grad_norm": 4.399874687194824, "learning_rate": 2.0565850876969045e-06, "loss": 1.2067, "step": 35776 }, { "epoch": 2.8038401253918495, "grad_norm": 4.3106889724731445, "learning_rate": 2.0563358245094696e-06, "loss": 1.3481, "step": 35777 }, { "epoch": 2.803918495297806, "grad_norm": 2.4205219745635986, "learning_rate": 2.056086572518221e-06, "loss": 1.6297, "step": 35778 }, { "epoch": 2.8039968652037617, "grad_norm": 4.430028438568115, "learning_rate": 2.0558373317241063e-06, "loss": 2.0247, "step": 35779 }, { "epoch": 2.804075235109718, "grad_norm": 3.2743752002716064, "learning_rate": 2.0555881021280753e-06, "loss": 1.2584, "step": 35780 }, { "epoch": 2.804153605015674, "grad_norm": 7.579367637634277, "learning_rate": 2.055338883731074e-06, "loss": 1.7239, "step": 35781 }, { "epoch": 2.8042319749216302, "grad_norm": 3.5384199619293213, "learning_rate": 2.0550896765340516e-06, "loss": 1.3828, "step": 35782 }, { "epoch": 2.804310344827586, "grad_norm": 4.4917755126953125, "learning_rate": 2.054840480537955e-06, "loss": 1.4626, "step": 35783 }, { "epoch": 2.8043887147335425, "grad_norm": 6.627810955047607, "learning_rate": 2.054591295743735e-06, "loss": 1.0196, "step": 35784 }, { "epoch": 2.8044670846394983, "grad_norm": 3.200141429901123, "learning_rate": 2.054342122152334e-06, "loss": 1.2266, "step": 35785 }, { "epoch": 2.8045454545454547, "grad_norm": 11.120122909545898, "learning_rate": 2.054092959764702e-06, "loss": 0.7736, "step": 35786 }, { "epoch": 2.8046238244514106, "grad_norm": 4.2158050537109375, "learning_rate": 2.053843808581788e-06, "loss": 1.2814, "step": 35787 }, { "epoch": 2.804702194357367, "grad_norm": 5.6538519859313965, "learning_rate": 2.0535946686045412e-06, "loss": 1.4077, "step": 35788 }, { "epoch": 2.804780564263323, "grad_norm": 3.325730323791504, "learning_rate": 2.0533455398339052e-06, "loss": 1.2367, "step": 35789 }, { "epoch": 2.804858934169279, "grad_norm": 3.2599639892578125, "learning_rate": 2.053096422270829e-06, "loss": 1.2911, "step": 35790 }, { "epoch": 2.804937304075235, "grad_norm": 2.7658872604370117, "learning_rate": 2.0528473159162587e-06, "loss": 1.3465, "step": 35791 }, { "epoch": 2.8050156739811913, "grad_norm": 3.0725924968719482, "learning_rate": 2.052598220771147e-06, "loss": 0.8616, "step": 35792 }, { "epoch": 2.8050940438871472, "grad_norm": 3.060359239578247, "learning_rate": 2.0523491368364363e-06, "loss": 1.6394, "step": 35793 }, { "epoch": 2.8051724137931036, "grad_norm": 4.804060935974121, "learning_rate": 2.052100064113075e-06, "loss": 1.3479, "step": 35794 }, { "epoch": 2.8052507836990594, "grad_norm": 4.592226028442383, "learning_rate": 2.051851002602011e-06, "loss": 0.8234, "step": 35795 }, { "epoch": 2.805329153605016, "grad_norm": 2.570134401321411, "learning_rate": 2.051601952304192e-06, "loss": 1.4484, "step": 35796 }, { "epoch": 2.8054075235109717, "grad_norm": 4.180131435394287, "learning_rate": 2.0513529132205646e-06, "loss": 1.4423, "step": 35797 }, { "epoch": 2.805485893416928, "grad_norm": 9.118671417236328, "learning_rate": 2.0511038853520758e-06, "loss": 1.1063, "step": 35798 }, { "epoch": 2.805564263322884, "grad_norm": 3.4100286960601807, "learning_rate": 2.0508548686996754e-06, "loss": 1.6351, "step": 35799 }, { "epoch": 2.80564263322884, "grad_norm": 3.336721658706665, "learning_rate": 2.0506058632643044e-06, "loss": 1.249, "step": 35800 }, { "epoch": 2.805721003134796, "grad_norm": 3.906510353088379, "learning_rate": 2.0503568690469155e-06, "loss": 1.5698, "step": 35801 }, { "epoch": 2.8057993730407524, "grad_norm": 4.187350273132324, "learning_rate": 2.0501078860484546e-06, "loss": 1.2926, "step": 35802 }, { "epoch": 2.8058777429467083, "grad_norm": 3.6577868461608887, "learning_rate": 2.0498589142698692e-06, "loss": 1.2234, "step": 35803 }, { "epoch": 2.8059561128526647, "grad_norm": 2.5390241146087646, "learning_rate": 2.049609953712104e-06, "loss": 1.2094, "step": 35804 }, { "epoch": 2.8060344827586206, "grad_norm": 3.07140851020813, "learning_rate": 2.0493610043761053e-06, "loss": 1.115, "step": 35805 }, { "epoch": 2.806112852664577, "grad_norm": 3.7036049365997314, "learning_rate": 2.0491120662628232e-06, "loss": 1.7706, "step": 35806 }, { "epoch": 2.8061912225705328, "grad_norm": 4.378645420074463, "learning_rate": 2.048863139373205e-06, "loss": 1.1236, "step": 35807 }, { "epoch": 2.806269592476489, "grad_norm": 3.9016051292419434, "learning_rate": 2.048614223708194e-06, "loss": 1.3454, "step": 35808 }, { "epoch": 2.806347962382445, "grad_norm": 7.798415660858154, "learning_rate": 2.048365319268739e-06, "loss": 1.7507, "step": 35809 }, { "epoch": 2.8064263322884013, "grad_norm": 2.4661171436309814, "learning_rate": 2.0481164260557856e-06, "loss": 1.1824, "step": 35810 }, { "epoch": 2.806504702194357, "grad_norm": 5.202683448791504, "learning_rate": 2.047867544070281e-06, "loss": 1.5767, "step": 35811 }, { "epoch": 2.8065830721003135, "grad_norm": 2.599769353866577, "learning_rate": 2.0476186733131726e-06, "loss": 1.0064, "step": 35812 }, { "epoch": 2.8066614420062694, "grad_norm": 2.5128324031829834, "learning_rate": 2.047369813785406e-06, "loss": 1.3514, "step": 35813 }, { "epoch": 2.8067398119122258, "grad_norm": 2.554274320602417, "learning_rate": 2.0471209654879282e-06, "loss": 1.1119, "step": 35814 }, { "epoch": 2.8068181818181817, "grad_norm": 3.800806760787964, "learning_rate": 2.046872128421685e-06, "loss": 1.5224, "step": 35815 }, { "epoch": 2.806896551724138, "grad_norm": 2.6339497566223145, "learning_rate": 2.046623302587624e-06, "loss": 1.4845, "step": 35816 }, { "epoch": 2.806974921630094, "grad_norm": 4.737443923950195, "learning_rate": 2.0463744879866908e-06, "loss": 2.0988, "step": 35817 }, { "epoch": 2.80705329153605, "grad_norm": 2.318403482437134, "learning_rate": 2.046125684619834e-06, "loss": 1.2683, "step": 35818 }, { "epoch": 2.807131661442006, "grad_norm": 3.0133538246154785, "learning_rate": 2.045876892487994e-06, "loss": 1.2933, "step": 35819 }, { "epoch": 2.8072100313479624, "grad_norm": 3.051560401916504, "learning_rate": 2.045628111592123e-06, "loss": 1.5226, "step": 35820 }, { "epoch": 2.8072884012539188, "grad_norm": 3.1001787185668945, "learning_rate": 2.0453793419331646e-06, "loss": 1.42, "step": 35821 }, { "epoch": 2.8073667711598747, "grad_norm": 3.4335317611694336, "learning_rate": 2.0451305835120685e-06, "loss": 1.4446, "step": 35822 }, { "epoch": 2.8074451410658305, "grad_norm": 3.6331987380981445, "learning_rate": 2.044881836329775e-06, "loss": 1.2551, "step": 35823 }, { "epoch": 2.807523510971787, "grad_norm": 5.008825302124023, "learning_rate": 2.044633100387232e-06, "loss": 1.5084, "step": 35824 }, { "epoch": 2.807601880877743, "grad_norm": 3.8821048736572266, "learning_rate": 2.04438437568539e-06, "loss": 1.2502, "step": 35825 }, { "epoch": 2.807680250783699, "grad_norm": 2.083552122116089, "learning_rate": 2.04413566222519e-06, "loss": 1.1655, "step": 35826 }, { "epoch": 2.807758620689655, "grad_norm": 3.1969504356384277, "learning_rate": 2.04388696000758e-06, "loss": 1.3617, "step": 35827 }, { "epoch": 2.8078369905956113, "grad_norm": 7.398013591766357, "learning_rate": 2.0436382690335057e-06, "loss": 0.6062, "step": 35828 }, { "epoch": 2.8079153605015676, "grad_norm": 6.537206649780273, "learning_rate": 2.0433895893039125e-06, "loss": 1.6047, "step": 35829 }, { "epoch": 2.8079937304075235, "grad_norm": 4.673101902008057, "learning_rate": 2.043140920819747e-06, "loss": 1.3248, "step": 35830 }, { "epoch": 2.8080721003134794, "grad_norm": 8.256233215332031, "learning_rate": 2.0428922635819544e-06, "loss": 1.796, "step": 35831 }, { "epoch": 2.8081504702194358, "grad_norm": 3.286447525024414, "learning_rate": 2.042643617591481e-06, "loss": 1.218, "step": 35832 }, { "epoch": 2.808228840125392, "grad_norm": 4.221883773803711, "learning_rate": 2.0423949828492724e-06, "loss": 1.4886, "step": 35833 }, { "epoch": 2.808307210031348, "grad_norm": 2.5984292030334473, "learning_rate": 2.042146359356274e-06, "loss": 1.5222, "step": 35834 }, { "epoch": 2.808385579937304, "grad_norm": 9.581242561340332, "learning_rate": 2.0418977471134315e-06, "loss": 1.3317, "step": 35835 }, { "epoch": 2.80846394984326, "grad_norm": 2.763908863067627, "learning_rate": 2.041649146121691e-06, "loss": 1.3152, "step": 35836 }, { "epoch": 2.8085423197492165, "grad_norm": 3.2546777725219727, "learning_rate": 2.0414005563819987e-06, "loss": 1.3563, "step": 35837 }, { "epoch": 2.8086206896551724, "grad_norm": 4.438094615936279, "learning_rate": 2.041151977895296e-06, "loss": 1.2098, "step": 35838 }, { "epoch": 2.8086990595611283, "grad_norm": 2.3892927169799805, "learning_rate": 2.040903410662533e-06, "loss": 1.13, "step": 35839 }, { "epoch": 2.8087774294670846, "grad_norm": 3.200807809829712, "learning_rate": 2.0406548546846556e-06, "loss": 1.348, "step": 35840 }, { "epoch": 2.808855799373041, "grad_norm": 3.3465681076049805, "learning_rate": 2.040406309962605e-06, "loss": 1.431, "step": 35841 }, { "epoch": 2.808934169278997, "grad_norm": 2.75236439704895, "learning_rate": 2.040157776497329e-06, "loss": 1.2244, "step": 35842 }, { "epoch": 2.8090125391849528, "grad_norm": 3.9080638885498047, "learning_rate": 2.0399092542897704e-06, "loss": 1.2779, "step": 35843 }, { "epoch": 2.809090909090909, "grad_norm": 3.6129400730133057, "learning_rate": 2.0396607433408805e-06, "loss": 1.34, "step": 35844 }, { "epoch": 2.8091692789968654, "grad_norm": 3.612950086593628, "learning_rate": 2.0394122436515983e-06, "loss": 1.5386, "step": 35845 }, { "epoch": 2.8092476489028213, "grad_norm": 2.9700844287872314, "learning_rate": 2.0391637552228715e-06, "loss": 1.2333, "step": 35846 }, { "epoch": 2.809326018808777, "grad_norm": 12.17266845703125, "learning_rate": 2.0389152780556452e-06, "loss": 2.5128, "step": 35847 }, { "epoch": 2.8094043887147335, "grad_norm": 3.6210272312164307, "learning_rate": 2.0386668121508646e-06, "loss": 1.6454, "step": 35848 }, { "epoch": 2.80948275862069, "grad_norm": 2.5707881450653076, "learning_rate": 2.038418357509474e-06, "loss": 1.4179, "step": 35849 }, { "epoch": 2.8095611285266457, "grad_norm": 3.2425477504730225, "learning_rate": 2.0381699141324185e-06, "loss": 1.3102, "step": 35850 }, { "epoch": 2.8096394984326016, "grad_norm": 2.993568181991577, "learning_rate": 2.0379214820206437e-06, "loss": 1.294, "step": 35851 }, { "epoch": 2.809717868338558, "grad_norm": 2.5672249794006348, "learning_rate": 2.0376730611750944e-06, "loss": 1.3664, "step": 35852 }, { "epoch": 2.8097962382445143, "grad_norm": 3.774237871170044, "learning_rate": 2.037424651596715e-06, "loss": 1.3015, "step": 35853 }, { "epoch": 2.80987460815047, "grad_norm": 8.550630569458008, "learning_rate": 2.0371762532864503e-06, "loss": 2.6064, "step": 35854 }, { "epoch": 2.809952978056426, "grad_norm": 3.191347599029541, "learning_rate": 2.0369278662452474e-06, "loss": 1.2517, "step": 35855 }, { "epoch": 2.8100313479623824, "grad_norm": 3.975881814956665, "learning_rate": 2.0366794904740452e-06, "loss": 1.416, "step": 35856 }, { "epoch": 2.8101097178683387, "grad_norm": 4.030043601989746, "learning_rate": 2.036431125973794e-06, "loss": 1.1393, "step": 35857 }, { "epoch": 2.8101880877742946, "grad_norm": 4.299619197845459, "learning_rate": 2.036182772745437e-06, "loss": 1.4651, "step": 35858 }, { "epoch": 2.8102664576802505, "grad_norm": 2.6856584548950195, "learning_rate": 2.0359344307899197e-06, "loss": 1.3, "step": 35859 }, { "epoch": 2.810344827586207, "grad_norm": 3.469203233718872, "learning_rate": 2.0356861001081833e-06, "loss": 1.6858, "step": 35860 }, { "epoch": 2.810423197492163, "grad_norm": 4.381213188171387, "learning_rate": 2.0354377807011754e-06, "loss": 1.2764, "step": 35861 }, { "epoch": 2.810501567398119, "grad_norm": 3.4465744495391846, "learning_rate": 2.0351894725698367e-06, "loss": 1.4515, "step": 35862 }, { "epoch": 2.810579937304075, "grad_norm": 2.775151014328003, "learning_rate": 2.0349411757151187e-06, "loss": 1.0945, "step": 35863 }, { "epoch": 2.8106583072100313, "grad_norm": 3.4449069499969482, "learning_rate": 2.0346928901379597e-06, "loss": 1.3747, "step": 35864 }, { "epoch": 2.8107366771159876, "grad_norm": 1.9773008823394775, "learning_rate": 2.0344446158393056e-06, "loss": 0.7149, "step": 35865 }, { "epoch": 2.8108150470219435, "grad_norm": 3.5364744663238525, "learning_rate": 2.034196352820101e-06, "loss": 1.0485, "step": 35866 }, { "epoch": 2.8108934169279, "grad_norm": 4.710826873779297, "learning_rate": 2.03394810108129e-06, "loss": 1.2771, "step": 35867 }, { "epoch": 2.8109717868338557, "grad_norm": 4.426129341125488, "learning_rate": 2.0336998606238173e-06, "loss": 0.8852, "step": 35868 }, { "epoch": 2.811050156739812, "grad_norm": 4.863001823425293, "learning_rate": 2.033451631448627e-06, "loss": 1.8966, "step": 35869 }, { "epoch": 2.811128526645768, "grad_norm": 2.642439603805542, "learning_rate": 2.0332034135566627e-06, "loss": 1.4129, "step": 35870 }, { "epoch": 2.8112068965517243, "grad_norm": 8.98935317993164, "learning_rate": 2.0329552069488685e-06, "loss": 0.8352, "step": 35871 }, { "epoch": 2.81128526645768, "grad_norm": 2.4976367950439453, "learning_rate": 2.0327070116261888e-06, "loss": 1.234, "step": 35872 }, { "epoch": 2.8113636363636365, "grad_norm": 6.077325820922852, "learning_rate": 2.0324588275895674e-06, "loss": 1.9867, "step": 35873 }, { "epoch": 2.8114420062695924, "grad_norm": 3.498853921890259, "learning_rate": 2.0322106548399506e-06, "loss": 1.1937, "step": 35874 }, { "epoch": 2.8115203761755487, "grad_norm": 5.325125694274902, "learning_rate": 2.031962493378276e-06, "loss": 1.7007, "step": 35875 }, { "epoch": 2.8115987460815046, "grad_norm": 4.622349262237549, "learning_rate": 2.0317143432054936e-06, "loss": 1.5346, "step": 35876 }, { "epoch": 2.811677115987461, "grad_norm": 3.237180471420288, "learning_rate": 2.0314662043225457e-06, "loss": 1.7783, "step": 35877 }, { "epoch": 2.811755485893417, "grad_norm": 4.144398212432861, "learning_rate": 2.0312180767303767e-06, "loss": 1.7071, "step": 35878 }, { "epoch": 2.811833855799373, "grad_norm": 3.6608834266662598, "learning_rate": 2.030969960429928e-06, "loss": 1.2406, "step": 35879 }, { "epoch": 2.811912225705329, "grad_norm": 4.03613805770874, "learning_rate": 2.0307218554221427e-06, "loss": 1.3446, "step": 35880 }, { "epoch": 2.8119905956112854, "grad_norm": 8.399530410766602, "learning_rate": 2.03047376170797e-06, "loss": 0.3167, "step": 35881 }, { "epoch": 2.8120689655172413, "grad_norm": 4.923458576202393, "learning_rate": 2.030225679288348e-06, "loss": 1.3466, "step": 35882 }, { "epoch": 2.8121473354231976, "grad_norm": 3.7859582901000977, "learning_rate": 2.029977608164222e-06, "loss": 0.9814, "step": 35883 }, { "epoch": 2.8122257053291535, "grad_norm": 8.319791793823242, "learning_rate": 2.0297295483365358e-06, "loss": 1.6673, "step": 35884 }, { "epoch": 2.81230407523511, "grad_norm": 3.097815752029419, "learning_rate": 2.0294814998062323e-06, "loss": 1.4773, "step": 35885 }, { "epoch": 2.8123824451410657, "grad_norm": 3.4199869632720947, "learning_rate": 2.029233462574256e-06, "loss": 1.4869, "step": 35886 }, { "epoch": 2.812460815047022, "grad_norm": 7.095843315124512, "learning_rate": 2.0289854366415496e-06, "loss": 1.3003, "step": 35887 }, { "epoch": 2.812539184952978, "grad_norm": 2.5680389404296875, "learning_rate": 2.0287374220090566e-06, "loss": 1.3599, "step": 35888 }, { "epoch": 2.8126175548589343, "grad_norm": 10.92296028137207, "learning_rate": 2.0284894186777205e-06, "loss": 1.6005, "step": 35889 }, { "epoch": 2.81269592476489, "grad_norm": 8.107821464538574, "learning_rate": 2.028241426648484e-06, "loss": 0.5375, "step": 35890 }, { "epoch": 2.8127742946708465, "grad_norm": 2.6381983757019043, "learning_rate": 2.027993445922291e-06, "loss": 1.4811, "step": 35891 }, { "epoch": 2.8128526645768024, "grad_norm": 2.9273841381073, "learning_rate": 2.027745476500086e-06, "loss": 1.2152, "step": 35892 }, { "epoch": 2.8129310344827587, "grad_norm": 2.563676118850708, "learning_rate": 2.0274975183828087e-06, "loss": 1.4118, "step": 35893 }, { "epoch": 2.8130094043887146, "grad_norm": 2.656017780303955, "learning_rate": 2.0272495715714023e-06, "loss": 1.2219, "step": 35894 }, { "epoch": 2.813087774294671, "grad_norm": 3.932142734527588, "learning_rate": 2.0270016360668138e-06, "loss": 1.6688, "step": 35895 }, { "epoch": 2.813166144200627, "grad_norm": 4.029994487762451, "learning_rate": 2.0267537118699856e-06, "loss": 1.3825, "step": 35896 }, { "epoch": 2.813244514106583, "grad_norm": 2.8232691287994385, "learning_rate": 2.026505798981857e-06, "loss": 1.1203, "step": 35897 }, { "epoch": 2.813322884012539, "grad_norm": 2.6360158920288086, "learning_rate": 2.0262578974033736e-06, "loss": 1.3656, "step": 35898 }, { "epoch": 2.8134012539184954, "grad_norm": 3.7527618408203125, "learning_rate": 2.0260100071354754e-06, "loss": 1.3628, "step": 35899 }, { "epoch": 2.8134796238244513, "grad_norm": 3.430487871170044, "learning_rate": 2.025762128179112e-06, "loss": 0.9948, "step": 35900 }, { "epoch": 2.8135579937304076, "grad_norm": 4.3660430908203125, "learning_rate": 2.0255142605352198e-06, "loss": 1.3634, "step": 35901 }, { "epoch": 2.8136363636363635, "grad_norm": 4.288480281829834, "learning_rate": 2.0252664042047427e-06, "loss": 1.319, "step": 35902 }, { "epoch": 2.81371473354232, "grad_norm": 5.6045660972595215, "learning_rate": 2.025018559188625e-06, "loss": 1.1121, "step": 35903 }, { "epoch": 2.8137931034482757, "grad_norm": 3.5163536071777344, "learning_rate": 2.0247707254878085e-06, "loss": 1.366, "step": 35904 }, { "epoch": 2.813871473354232, "grad_norm": 6.222588062286377, "learning_rate": 2.024522903103236e-06, "loss": 1.7206, "step": 35905 }, { "epoch": 2.813949843260188, "grad_norm": 3.6927528381347656, "learning_rate": 2.02427509203585e-06, "loss": 1.2741, "step": 35906 }, { "epoch": 2.8140282131661443, "grad_norm": 2.3813729286193848, "learning_rate": 2.024027292286595e-06, "loss": 1.2249, "step": 35907 }, { "epoch": 2.8141065830721, "grad_norm": 3.5494773387908936, "learning_rate": 2.0237795038564078e-06, "loss": 1.2944, "step": 35908 }, { "epoch": 2.8141849529780565, "grad_norm": 4.127806663513184, "learning_rate": 2.0235317267462367e-06, "loss": 1.3277, "step": 35909 }, { "epoch": 2.8142633228840124, "grad_norm": 3.313713788986206, "learning_rate": 2.0232839609570227e-06, "loss": 1.5506, "step": 35910 }, { "epoch": 2.8143416927899687, "grad_norm": 2.9488513469696045, "learning_rate": 2.0230362064897085e-06, "loss": 0.999, "step": 35911 }, { "epoch": 2.814420062695925, "grad_norm": 9.262208938598633, "learning_rate": 2.022788463345234e-06, "loss": 1.7199, "step": 35912 }, { "epoch": 2.814498432601881, "grad_norm": 2.6032238006591797, "learning_rate": 2.022540731524541e-06, "loss": 1.3029, "step": 35913 }, { "epoch": 2.814576802507837, "grad_norm": 3.3699722290039062, "learning_rate": 2.022293011028576e-06, "loss": 1.257, "step": 35914 }, { "epoch": 2.814655172413793, "grad_norm": 2.3744301795959473, "learning_rate": 2.0220453018582805e-06, "loss": 1.3276, "step": 35915 }, { "epoch": 2.8147335423197495, "grad_norm": 2.7984349727630615, "learning_rate": 2.021797604014593e-06, "loss": 1.3969, "step": 35916 }, { "epoch": 2.8148119122257054, "grad_norm": 3.4092953205108643, "learning_rate": 2.0215499174984583e-06, "loss": 0.6778, "step": 35917 }, { "epoch": 2.8148902821316613, "grad_norm": 2.415403127670288, "learning_rate": 2.0213022423108153e-06, "loss": 0.9611, "step": 35918 }, { "epoch": 2.8149686520376176, "grad_norm": 2.347489833831787, "learning_rate": 2.0210545784526127e-06, "loss": 1.3652, "step": 35919 }, { "epoch": 2.815047021943574, "grad_norm": 2.5033652782440186, "learning_rate": 2.0208069259247866e-06, "loss": 0.9169, "step": 35920 }, { "epoch": 2.81512539184953, "grad_norm": 3.756389856338501, "learning_rate": 2.0205592847282805e-06, "loss": 1.2001, "step": 35921 }, { "epoch": 2.8152037617554857, "grad_norm": 2.2938241958618164, "learning_rate": 2.020311654864037e-06, "loss": 1.3785, "step": 35922 }, { "epoch": 2.815282131661442, "grad_norm": 2.561599016189575, "learning_rate": 2.020064036332997e-06, "loss": 1.3476, "step": 35923 }, { "epoch": 2.8153605015673984, "grad_norm": 5.425716876983643, "learning_rate": 2.0198164291361033e-06, "loss": 0.9911, "step": 35924 }, { "epoch": 2.8154388714733543, "grad_norm": 2.5099260807037354, "learning_rate": 2.019568833274297e-06, "loss": 1.3505, "step": 35925 }, { "epoch": 2.81551724137931, "grad_norm": 4.842301845550537, "learning_rate": 2.0193212487485216e-06, "loss": 1.4237, "step": 35926 }, { "epoch": 2.8155956112852665, "grad_norm": 3.3168272972106934, "learning_rate": 2.019073675559714e-06, "loss": 1.7337, "step": 35927 }, { "epoch": 2.815673981191223, "grad_norm": 5.487635612487793, "learning_rate": 2.0188261137088207e-06, "loss": 0.8935, "step": 35928 }, { "epoch": 2.8157523510971787, "grad_norm": 10.967875480651855, "learning_rate": 2.0185785631967813e-06, "loss": 1.2951, "step": 35929 }, { "epoch": 2.8158307210031346, "grad_norm": 2.890937328338623, "learning_rate": 2.0183310240245395e-06, "loss": 1.1141, "step": 35930 }, { "epoch": 2.815909090909091, "grad_norm": 5.727811336517334, "learning_rate": 2.0180834961930334e-06, "loss": 1.1839, "step": 35931 }, { "epoch": 2.8159874608150472, "grad_norm": 4.528716564178467, "learning_rate": 2.0178359797032044e-06, "loss": 1.6288, "step": 35932 }, { "epoch": 2.816065830721003, "grad_norm": 2.3404810428619385, "learning_rate": 2.017588474555998e-06, "loss": 1.5924, "step": 35933 }, { "epoch": 2.816144200626959, "grad_norm": 2.2033164501190186, "learning_rate": 2.0173409807523534e-06, "loss": 1.0208, "step": 35934 }, { "epoch": 2.8162225705329154, "grad_norm": 3.191818952560425, "learning_rate": 2.017093498293211e-06, "loss": 1.3386, "step": 35935 }, { "epoch": 2.8163009404388717, "grad_norm": 4.421388626098633, "learning_rate": 2.0168460271795126e-06, "loss": 1.6523, "step": 35936 }, { "epoch": 2.8163793103448276, "grad_norm": 4.495467662811279, "learning_rate": 2.0165985674121996e-06, "loss": 1.3579, "step": 35937 }, { "epoch": 2.8164576802507835, "grad_norm": 2.9770925045013428, "learning_rate": 2.0163511189922138e-06, "loss": 1.329, "step": 35938 }, { "epoch": 2.81653605015674, "grad_norm": 3.8420188426971436, "learning_rate": 2.016103681920495e-06, "loss": 1.3201, "step": 35939 }, { "epoch": 2.816614420062696, "grad_norm": 4.010021686553955, "learning_rate": 2.0158562561979857e-06, "loss": 1.298, "step": 35940 }, { "epoch": 2.816692789968652, "grad_norm": 2.7110493183135986, "learning_rate": 2.0156088418256263e-06, "loss": 1.3759, "step": 35941 }, { "epoch": 2.816771159874608, "grad_norm": 3.1793031692504883, "learning_rate": 2.015361438804358e-06, "loss": 1.8703, "step": 35942 }, { "epoch": 2.8168495297805642, "grad_norm": 4.031754016876221, "learning_rate": 2.0151140471351215e-06, "loss": 1.2529, "step": 35943 }, { "epoch": 2.8169278996865206, "grad_norm": 6.5215163230896, "learning_rate": 2.014866666818858e-06, "loss": 1.2801, "step": 35944 }, { "epoch": 2.8170062695924765, "grad_norm": 2.918968439102173, "learning_rate": 2.014619297856511e-06, "loss": 1.3286, "step": 35945 }, { "epoch": 2.8170846394984324, "grad_norm": 2.8295395374298096, "learning_rate": 2.0143719402490146e-06, "loss": 0.9208, "step": 35946 }, { "epoch": 2.8171630094043887, "grad_norm": 2.788059711456299, "learning_rate": 2.0141245939973157e-06, "loss": 1.4816, "step": 35947 }, { "epoch": 2.817241379310345, "grad_norm": 4.386406421661377, "learning_rate": 2.013877259102355e-06, "loss": 1.4217, "step": 35948 }, { "epoch": 2.817319749216301, "grad_norm": 3.3860344886779785, "learning_rate": 2.01362993556507e-06, "loss": 1.4725, "step": 35949 }, { "epoch": 2.817398119122257, "grad_norm": 3.705455780029297, "learning_rate": 2.0133826233864023e-06, "loss": 1.7325, "step": 35950 }, { "epoch": 2.817476489028213, "grad_norm": 4.169682025909424, "learning_rate": 2.0131353225672917e-06, "loss": 1.7467, "step": 35951 }, { "epoch": 2.8175548589341695, "grad_norm": 5.93269157409668, "learning_rate": 2.0128880331086836e-06, "loss": 1.551, "step": 35952 }, { "epoch": 2.8176332288401253, "grad_norm": 2.6753196716308594, "learning_rate": 2.0126407550115134e-06, "loss": 1.2594, "step": 35953 }, { "epoch": 2.8177115987460812, "grad_norm": 4.510082244873047, "learning_rate": 2.012393488276723e-06, "loss": 1.1974, "step": 35954 }, { "epoch": 2.8177899686520376, "grad_norm": 3.397491693496704, "learning_rate": 2.012146232905254e-06, "loss": 1.4903, "step": 35955 }, { "epoch": 2.817868338557994, "grad_norm": 2.6660211086273193, "learning_rate": 2.0118989888980454e-06, "loss": 1.5983, "step": 35956 }, { "epoch": 2.81794670846395, "grad_norm": 3.4798243045806885, "learning_rate": 2.0116517562560382e-06, "loss": 1.4315, "step": 35957 }, { "epoch": 2.818025078369906, "grad_norm": 2.443406343460083, "learning_rate": 2.011404534980173e-06, "loss": 1.4183, "step": 35958 }, { "epoch": 2.818103448275862, "grad_norm": 4.2936625480651855, "learning_rate": 2.0111573250713905e-06, "loss": 1.7328, "step": 35959 }, { "epoch": 2.8181818181818183, "grad_norm": 3.2104907035827637, "learning_rate": 2.0109101265306295e-06, "loss": 1.6112, "step": 35960 }, { "epoch": 2.8182601880877742, "grad_norm": 3.328335762023926, "learning_rate": 2.0106629393588315e-06, "loss": 0.5995, "step": 35961 }, { "epoch": 2.8183385579937306, "grad_norm": 3.093562364578247, "learning_rate": 2.0104157635569364e-06, "loss": 1.2938, "step": 35962 }, { "epoch": 2.8184169278996865, "grad_norm": 3.6878507137298584, "learning_rate": 2.0101685991258856e-06, "loss": 1.0188, "step": 35963 }, { "epoch": 2.818495297805643, "grad_norm": 3.9259774684906006, "learning_rate": 2.0099214460666146e-06, "loss": 1.1189, "step": 35964 }, { "epoch": 2.8185736677115987, "grad_norm": 4.1127028465271, "learning_rate": 2.0096743043800683e-06, "loss": 1.1457, "step": 35965 }, { "epoch": 2.818652037617555, "grad_norm": 3.598336935043335, "learning_rate": 2.0094271740671844e-06, "loss": 1.1429, "step": 35966 }, { "epoch": 2.818730407523511, "grad_norm": 2.4997124671936035, "learning_rate": 2.009180055128906e-06, "loss": 1.0748, "step": 35967 }, { "epoch": 2.8188087774294672, "grad_norm": 2.6547224521636963, "learning_rate": 2.0089329475661684e-06, "loss": 1.571, "step": 35968 }, { "epoch": 2.818887147335423, "grad_norm": 2.5304973125457764, "learning_rate": 2.0086858513799116e-06, "loss": 1.0934, "step": 35969 }, { "epoch": 2.8189655172413794, "grad_norm": 2.775956153869629, "learning_rate": 2.0084387665710797e-06, "loss": 1.3909, "step": 35970 }, { "epoch": 2.8190438871473353, "grad_norm": 2.8132386207580566, "learning_rate": 2.0081916931406116e-06, "loss": 1.1431, "step": 35971 }, { "epoch": 2.8191222570532917, "grad_norm": 2.816862106323242, "learning_rate": 2.0079446310894442e-06, "loss": 1.6028, "step": 35972 }, { "epoch": 2.8192006269592476, "grad_norm": 4.491332054138184, "learning_rate": 2.0076975804185177e-06, "loss": 1.135, "step": 35973 }, { "epoch": 2.819278996865204, "grad_norm": 5.194606304168701, "learning_rate": 2.007450541128774e-06, "loss": 0.9648, "step": 35974 }, { "epoch": 2.81935736677116, "grad_norm": 4.5769243240356445, "learning_rate": 2.0072035132211505e-06, "loss": 1.3698, "step": 35975 }, { "epoch": 2.819435736677116, "grad_norm": 4.295253753662109, "learning_rate": 2.0069564966965878e-06, "loss": 1.6012, "step": 35976 }, { "epoch": 2.819514106583072, "grad_norm": 4.321954250335693, "learning_rate": 2.0067094915560246e-06, "loss": 1.4448, "step": 35977 }, { "epoch": 2.8195924764890283, "grad_norm": 2.2958738803863525, "learning_rate": 2.006462497800402e-06, "loss": 1.3361, "step": 35978 }, { "epoch": 2.819670846394984, "grad_norm": 5.140566349029541, "learning_rate": 2.0062155154306582e-06, "loss": 1.2892, "step": 35979 }, { "epoch": 2.8197492163009406, "grad_norm": 12.376976013183594, "learning_rate": 2.005968544447733e-06, "loss": 1.0367, "step": 35980 }, { "epoch": 2.8198275862068964, "grad_norm": 2.8747687339782715, "learning_rate": 2.005721584852565e-06, "loss": 1.2062, "step": 35981 }, { "epoch": 2.8199059561128528, "grad_norm": 2.1242282390594482, "learning_rate": 2.0054746366460964e-06, "loss": 1.2433, "step": 35982 }, { "epoch": 2.8199843260188087, "grad_norm": 4.549534320831299, "learning_rate": 2.0052276998292606e-06, "loss": 1.7553, "step": 35983 }, { "epoch": 2.820062695924765, "grad_norm": 2.285226583480835, "learning_rate": 2.0049807744030024e-06, "loss": 1.194, "step": 35984 }, { "epoch": 2.820141065830721, "grad_norm": 4.277343273162842, "learning_rate": 2.0047338603682586e-06, "loss": 1.3562, "step": 35985 }, { "epoch": 2.820219435736677, "grad_norm": 3.9532368183135986, "learning_rate": 2.004486957725971e-06, "loss": 1.504, "step": 35986 }, { "epoch": 2.820297805642633, "grad_norm": 3.35196590423584, "learning_rate": 2.004240066477074e-06, "loss": 1.3945, "step": 35987 }, { "epoch": 2.8203761755485894, "grad_norm": 3.039602756500244, "learning_rate": 2.0039931866225077e-06, "loss": 1.403, "step": 35988 }, { "epoch": 2.8204545454545453, "grad_norm": 2.4375767707824707, "learning_rate": 2.0037463181632144e-06, "loss": 1.1795, "step": 35989 }, { "epoch": 2.8205329153605017, "grad_norm": 4.770888328552246, "learning_rate": 2.0034994611001324e-06, "loss": 1.2753, "step": 35990 }, { "epoch": 2.8206112852664575, "grad_norm": 4.0094313621521, "learning_rate": 2.003252615434198e-06, "loss": 1.4738, "step": 35991 }, { "epoch": 2.820689655172414, "grad_norm": 5.297491550445557, "learning_rate": 2.003005781166351e-06, "loss": 1.4036, "step": 35992 }, { "epoch": 2.8207680250783698, "grad_norm": 2.5075392723083496, "learning_rate": 2.002758958297531e-06, "loss": 1.281, "step": 35993 }, { "epoch": 2.820846394984326, "grad_norm": 4.460519313812256, "learning_rate": 2.0025121468286766e-06, "loss": 1.5898, "step": 35994 }, { "epoch": 2.820924764890282, "grad_norm": 3.3011794090270996, "learning_rate": 2.002265346760726e-06, "loss": 1.3366, "step": 35995 }, { "epoch": 2.8210031347962383, "grad_norm": 2.786163806915283, "learning_rate": 2.002018558094619e-06, "loss": 1.4353, "step": 35996 }, { "epoch": 2.821081504702194, "grad_norm": 8.546666145324707, "learning_rate": 2.0017717808312927e-06, "loss": 1.4742, "step": 35997 }, { "epoch": 2.8211598746081505, "grad_norm": 2.8728513717651367, "learning_rate": 2.001525014971687e-06, "loss": 1.2035, "step": 35998 }, { "epoch": 2.8212382445141064, "grad_norm": 8.168036460876465, "learning_rate": 2.0012782605167398e-06, "loss": 1.5287, "step": 35999 }, { "epoch": 2.8213166144200628, "grad_norm": 3.9338572025299072, "learning_rate": 2.00103151746739e-06, "loss": 1.7553, "step": 36000 }, { "epoch": 2.8213949843260187, "grad_norm": 2.321455955505371, "learning_rate": 2.000784785824578e-06, "loss": 1.4202, "step": 36001 }, { "epoch": 2.821473354231975, "grad_norm": 5.933071136474609, "learning_rate": 2.000538065589236e-06, "loss": 1.1522, "step": 36002 }, { "epoch": 2.821551724137931, "grad_norm": 2.244711399078369, "learning_rate": 2.000291356762308e-06, "loss": 1.5746, "step": 36003 }, { "epoch": 2.821630094043887, "grad_norm": 2.9500250816345215, "learning_rate": 2.0000446593447337e-06, "loss": 1.4413, "step": 36004 }, { "epoch": 2.821708463949843, "grad_norm": 3.9744374752044678, "learning_rate": 1.9997979733374466e-06, "loss": 1.6723, "step": 36005 }, { "epoch": 2.8217868338557994, "grad_norm": 2.3265910148620605, "learning_rate": 1.999551298741387e-06, "loss": 0.9318, "step": 36006 }, { "epoch": 2.8218652037617558, "grad_norm": 2.2148079872131348, "learning_rate": 1.9993046355574913e-06, "loss": 1.1786, "step": 36007 }, { "epoch": 2.8219435736677116, "grad_norm": 2.6599535942077637, "learning_rate": 1.999057983786703e-06, "loss": 1.3463, "step": 36008 }, { "epoch": 2.8220219435736675, "grad_norm": 9.265419006347656, "learning_rate": 1.998811343429955e-06, "loss": 1.923, "step": 36009 }, { "epoch": 2.822100313479624, "grad_norm": 3.561694383621216, "learning_rate": 1.998564714488187e-06, "loss": 1.5766, "step": 36010 }, { "epoch": 2.82217868338558, "grad_norm": 5.693062782287598, "learning_rate": 1.9983180969623374e-06, "loss": 1.6973, "step": 36011 }, { "epoch": 2.822257053291536, "grad_norm": 4.727969169616699, "learning_rate": 1.998071490853344e-06, "loss": 1.4921, "step": 36012 }, { "epoch": 2.822335423197492, "grad_norm": 3.2027673721313477, "learning_rate": 1.997824896162145e-06, "loss": 1.3036, "step": 36013 }, { "epoch": 2.8224137931034483, "grad_norm": 7.970448970794678, "learning_rate": 1.9975783128896775e-06, "loss": 1.1083, "step": 36014 }, { "epoch": 2.8224921630094046, "grad_norm": 3.3488047122955322, "learning_rate": 1.9973317410368804e-06, "loss": 0.7757, "step": 36015 }, { "epoch": 2.8225705329153605, "grad_norm": 2.9133920669555664, "learning_rate": 1.997085180604691e-06, "loss": 1.4196, "step": 36016 }, { "epoch": 2.8226489028213164, "grad_norm": 4.68671178817749, "learning_rate": 1.996838631594047e-06, "loss": 1.5336, "step": 36017 }, { "epoch": 2.8227272727272728, "grad_norm": 3.237962484359741, "learning_rate": 1.9965920940058864e-06, "loss": 2.033, "step": 36018 }, { "epoch": 2.822805642633229, "grad_norm": 3.8898162841796875, "learning_rate": 1.9963455678411487e-06, "loss": 1.3604, "step": 36019 }, { "epoch": 2.822884012539185, "grad_norm": 7.343111991882324, "learning_rate": 1.996099053100768e-06, "loss": 1.234, "step": 36020 }, { "epoch": 2.822962382445141, "grad_norm": 2.779005527496338, "learning_rate": 1.995852549785681e-06, "loss": 1.4931, "step": 36021 }, { "epoch": 2.823040752351097, "grad_norm": 3.8266384601593018, "learning_rate": 1.9956060578968307e-06, "loss": 1.4788, "step": 36022 }, { "epoch": 2.8231191222570535, "grad_norm": 7.7792582511901855, "learning_rate": 1.9953595774351532e-06, "loss": 1.5934, "step": 36023 }, { "epoch": 2.8231974921630094, "grad_norm": 3.7099344730377197, "learning_rate": 1.995113108401583e-06, "loss": 1.4691, "step": 36024 }, { "epoch": 2.8232758620689653, "grad_norm": 3.4132003784179688, "learning_rate": 1.994866650797059e-06, "loss": 1.2675, "step": 36025 }, { "epoch": 2.8233542319749216, "grad_norm": 4.511594295501709, "learning_rate": 1.9946202046225166e-06, "loss": 1.5429, "step": 36026 }, { "epoch": 2.823432601880878, "grad_norm": 2.7203216552734375, "learning_rate": 1.994373769878899e-06, "loss": 1.4762, "step": 36027 }, { "epoch": 2.823510971786834, "grad_norm": 5.208275318145752, "learning_rate": 1.9941273465671385e-06, "loss": 1.1206, "step": 36028 }, { "epoch": 2.8235893416927897, "grad_norm": 3.8605785369873047, "learning_rate": 1.9938809346881734e-06, "loss": 1.1921, "step": 36029 }, { "epoch": 2.823667711598746, "grad_norm": 13.499332427978516, "learning_rate": 1.993634534242941e-06, "loss": 1.6691, "step": 36030 }, { "epoch": 2.8237460815047024, "grad_norm": 2.1029882431030273, "learning_rate": 1.9933881452323783e-06, "loss": 1.2107, "step": 36031 }, { "epoch": 2.8238244514106583, "grad_norm": 4.269386291503906, "learning_rate": 1.9931417676574234e-06, "loss": 1.1165, "step": 36032 }, { "epoch": 2.823902821316614, "grad_norm": 4.644377708435059, "learning_rate": 1.992895401519013e-06, "loss": 1.3758, "step": 36033 }, { "epoch": 2.8239811912225705, "grad_norm": 4.2913498878479, "learning_rate": 1.9926490468180848e-06, "loss": 1.4706, "step": 36034 }, { "epoch": 2.824059561128527, "grad_norm": 2.798217296600342, "learning_rate": 1.9924027035555722e-06, "loss": 1.2575, "step": 36035 }, { "epoch": 2.8241379310344827, "grad_norm": 2.3958635330200195, "learning_rate": 1.9921563717324167e-06, "loss": 1.3306, "step": 36036 }, { "epoch": 2.8242163009404386, "grad_norm": 2.3555471897125244, "learning_rate": 1.9919100513495532e-06, "loss": 1.1301, "step": 36037 }, { "epoch": 2.824294670846395, "grad_norm": 4.920503616333008, "learning_rate": 1.9916637424079204e-06, "loss": 1.1632, "step": 36038 }, { "epoch": 2.8243730407523513, "grad_norm": 2.9750335216522217, "learning_rate": 1.9914174449084516e-06, "loss": 1.4225, "step": 36039 }, { "epoch": 2.824451410658307, "grad_norm": 4.512276649475098, "learning_rate": 1.9911711588520845e-06, "loss": 1.1666, "step": 36040 }, { "epoch": 2.824529780564263, "grad_norm": 2.815457582473755, "learning_rate": 1.990924884239758e-06, "loss": 1.4573, "step": 36041 }, { "epoch": 2.8246081504702194, "grad_norm": 4.509828567504883, "learning_rate": 1.99067862107241e-06, "loss": 1.4929, "step": 36042 }, { "epoch": 2.8246865203761757, "grad_norm": 2.1818368434906006, "learning_rate": 1.9904323693509724e-06, "loss": 1.5902, "step": 36043 }, { "epoch": 2.8247648902821316, "grad_norm": 2.241783380508423, "learning_rate": 1.9901861290763846e-06, "loss": 1.4175, "step": 36044 }, { "epoch": 2.8248432601880875, "grad_norm": 3.6565794944763184, "learning_rate": 1.9899399002495813e-06, "loss": 1.3047, "step": 36045 }, { "epoch": 2.824921630094044, "grad_norm": 3.2749037742614746, "learning_rate": 1.989693682871504e-06, "loss": 1.4212, "step": 36046 }, { "epoch": 2.825, "grad_norm": 3.8267855644226074, "learning_rate": 1.9894474769430838e-06, "loss": 1.5293, "step": 36047 }, { "epoch": 2.825078369905956, "grad_norm": 3.1963186264038086, "learning_rate": 1.989201282465259e-06, "loss": 1.559, "step": 36048 }, { "epoch": 2.825156739811912, "grad_norm": 3.5377633571624756, "learning_rate": 1.988955099438966e-06, "loss": 1.6363, "step": 36049 }, { "epoch": 2.8252351097178683, "grad_norm": 2.592582941055298, "learning_rate": 1.9887089278651417e-06, "loss": 1.0972, "step": 36050 }, { "epoch": 2.8253134796238246, "grad_norm": 4.234835147857666, "learning_rate": 1.9884627677447218e-06, "loss": 0.6896, "step": 36051 }, { "epoch": 2.8253918495297805, "grad_norm": 4.023046016693115, "learning_rate": 1.988216619078643e-06, "loss": 0.6529, "step": 36052 }, { "epoch": 2.825470219435737, "grad_norm": 4.680334091186523, "learning_rate": 1.9879704818678427e-06, "loss": 0.4131, "step": 36053 }, { "epoch": 2.8255485893416927, "grad_norm": 5.371922969818115, "learning_rate": 1.9877243561132516e-06, "loss": 1.0672, "step": 36054 }, { "epoch": 2.825626959247649, "grad_norm": 2.4888176918029785, "learning_rate": 1.987478241815812e-06, "loss": 1.2897, "step": 36055 }, { "epoch": 2.825705329153605, "grad_norm": 5.17650842666626, "learning_rate": 1.9872321389764576e-06, "loss": 1.4603, "step": 36056 }, { "epoch": 2.8257836990595613, "grad_norm": 4.639580249786377, "learning_rate": 1.9869860475961266e-06, "loss": 1.5245, "step": 36057 }, { "epoch": 2.825862068965517, "grad_norm": 2.7871034145355225, "learning_rate": 1.9867399676757517e-06, "loss": 1.4372, "step": 36058 }, { "epoch": 2.8259404388714735, "grad_norm": 2.3744349479675293, "learning_rate": 1.9864938992162682e-06, "loss": 1.3219, "step": 36059 }, { "epoch": 2.8260188087774294, "grad_norm": 7.1202473640441895, "learning_rate": 1.9862478422186178e-06, "loss": 1.5032, "step": 36060 }, { "epoch": 2.8260971786833857, "grad_norm": 2.3860719203948975, "learning_rate": 1.9860017966837304e-06, "loss": 1.6349, "step": 36061 }, { "epoch": 2.8261755485893416, "grad_norm": 2.5028228759765625, "learning_rate": 1.9857557626125445e-06, "loss": 1.2819, "step": 36062 }, { "epoch": 2.826253918495298, "grad_norm": 3.86360502243042, "learning_rate": 1.9855097400059954e-06, "loss": 0.9365, "step": 36063 }, { "epoch": 2.826332288401254, "grad_norm": 2.7873780727386475, "learning_rate": 1.9852637288650184e-06, "loss": 1.1672, "step": 36064 }, { "epoch": 2.82641065830721, "grad_norm": 4.336167812347412, "learning_rate": 1.9850177291905505e-06, "loss": 1.8151, "step": 36065 }, { "epoch": 2.826489028213166, "grad_norm": 2.559373378753662, "learning_rate": 1.984771740983526e-06, "loss": 1.3299, "step": 36066 }, { "epoch": 2.8265673981191224, "grad_norm": 4.472956657409668, "learning_rate": 1.984525764244881e-06, "loss": 1.5518, "step": 36067 }, { "epoch": 2.8266457680250783, "grad_norm": 2.413567543029785, "learning_rate": 1.984279798975552e-06, "loss": 1.236, "step": 36068 }, { "epoch": 2.8267241379310346, "grad_norm": 2.5964577198028564, "learning_rate": 1.9840338451764728e-06, "loss": 1.3027, "step": 36069 }, { "epoch": 2.8268025078369905, "grad_norm": 4.536169052124023, "learning_rate": 1.98378790284858e-06, "loss": 1.316, "step": 36070 }, { "epoch": 2.826880877742947, "grad_norm": 4.871654033660889, "learning_rate": 1.983541971992809e-06, "loss": 1.6076, "step": 36071 }, { "epoch": 2.8269592476489027, "grad_norm": 5.5583319664001465, "learning_rate": 1.9832960526100948e-06, "loss": 1.6085, "step": 36072 }, { "epoch": 2.827037617554859, "grad_norm": 4.735121250152588, "learning_rate": 1.9830501447013735e-06, "loss": 1.3594, "step": 36073 }, { "epoch": 2.827115987460815, "grad_norm": 2.7118005752563477, "learning_rate": 1.982804248267579e-06, "loss": 1.3841, "step": 36074 }, { "epoch": 2.8271943573667713, "grad_norm": 4.176810264587402, "learning_rate": 1.9825583633096495e-06, "loss": 1.4853, "step": 36075 }, { "epoch": 2.827272727272727, "grad_norm": 5.324422836303711, "learning_rate": 1.9823124898285163e-06, "loss": 1.2977, "step": 36076 }, { "epoch": 2.8273510971786835, "grad_norm": 2.680650472640991, "learning_rate": 1.982066627825115e-06, "loss": 1.2924, "step": 36077 }, { "epoch": 2.8274294670846394, "grad_norm": 2.5301363468170166, "learning_rate": 1.9818207773003834e-06, "loss": 1.0885, "step": 36078 }, { "epoch": 2.8275078369905957, "grad_norm": 4.36944055557251, "learning_rate": 1.9815749382552573e-06, "loss": 1.2402, "step": 36079 }, { "epoch": 2.8275862068965516, "grad_norm": 3.7243428230285645, "learning_rate": 1.981329110690668e-06, "loss": 1.1147, "step": 36080 }, { "epoch": 2.827664576802508, "grad_norm": 2.742260456085205, "learning_rate": 1.9810832946075527e-06, "loss": 1.5314, "step": 36081 }, { "epoch": 2.827742946708464, "grad_norm": 3.882082462310791, "learning_rate": 1.980837490006843e-06, "loss": 1.6571, "step": 36082 }, { "epoch": 2.82782131661442, "grad_norm": 1.9748972654342651, "learning_rate": 1.9805916968894812e-06, "loss": 0.8704, "step": 36083 }, { "epoch": 2.827899686520376, "grad_norm": 4.458461284637451, "learning_rate": 1.9803459152563957e-06, "loss": 0.9576, "step": 36084 }, { "epoch": 2.8279780564263324, "grad_norm": 3.3166465759277344, "learning_rate": 1.980100145108524e-06, "loss": 1.2845, "step": 36085 }, { "epoch": 2.8280564263322883, "grad_norm": 4.388299942016602, "learning_rate": 1.9798543864467995e-06, "loss": 1.6914, "step": 36086 }, { "epoch": 2.8281347962382446, "grad_norm": 9.678997993469238, "learning_rate": 1.9796086392721574e-06, "loss": 2.7489, "step": 36087 }, { "epoch": 2.8282131661442005, "grad_norm": 3.299731492996216, "learning_rate": 1.9793629035855334e-06, "loss": 1.5055, "step": 36088 }, { "epoch": 2.828291536050157, "grad_norm": 2.878546714782715, "learning_rate": 1.979117179387861e-06, "loss": 0.9306, "step": 36089 }, { "epoch": 2.8283699059561127, "grad_norm": 4.661230087280273, "learning_rate": 1.9788714666800775e-06, "loss": 1.3759, "step": 36090 }, { "epoch": 2.828448275862069, "grad_norm": 6.742547512054443, "learning_rate": 1.9786257654631108e-06, "loss": 1.173, "step": 36091 }, { "epoch": 2.828526645768025, "grad_norm": 2.2028026580810547, "learning_rate": 1.9783800757379023e-06, "loss": 1.4485, "step": 36092 }, { "epoch": 2.8286050156739813, "grad_norm": 4.836180210113525, "learning_rate": 1.9781343975053834e-06, "loss": 1.1845, "step": 36093 }, { "epoch": 2.828683385579937, "grad_norm": 2.8965442180633545, "learning_rate": 1.9778887307664917e-06, "loss": 1.3018, "step": 36094 }, { "epoch": 2.8287617554858935, "grad_norm": 3.706970691680908, "learning_rate": 1.9776430755221564e-06, "loss": 1.188, "step": 36095 }, { "epoch": 2.8288401253918494, "grad_norm": 7.432524681091309, "learning_rate": 1.9773974317733136e-06, "loss": 0.5249, "step": 36096 }, { "epoch": 2.8289184952978057, "grad_norm": 3.9299540519714355, "learning_rate": 1.9771517995208995e-06, "loss": 1.5342, "step": 36097 }, { "epoch": 2.8289968652037616, "grad_norm": 2.5485129356384277, "learning_rate": 1.9769061787658493e-06, "loss": 1.3625, "step": 36098 }, { "epoch": 2.829075235109718, "grad_norm": 2.979168653488159, "learning_rate": 1.9766605695090936e-06, "loss": 1.2047, "step": 36099 }, { "epoch": 2.829153605015674, "grad_norm": 4.383316993713379, "learning_rate": 1.976414971751568e-06, "loss": 0.7537, "step": 36100 }, { "epoch": 2.82923197492163, "grad_norm": 2.61582088470459, "learning_rate": 1.976169385494205e-06, "loss": 1.1887, "step": 36101 }, { "epoch": 2.8293103448275865, "grad_norm": 7.6207404136657715, "learning_rate": 1.975923810737944e-06, "loss": 1.3601, "step": 36102 }, { "epoch": 2.8293887147335424, "grad_norm": 3.9704580307006836, "learning_rate": 1.9756782474837142e-06, "loss": 1.7582, "step": 36103 }, { "epoch": 2.8294670846394983, "grad_norm": 3.0495283603668213, "learning_rate": 1.975432695732451e-06, "loss": 1.4128, "step": 36104 }, { "epoch": 2.8295454545454546, "grad_norm": 3.6950674057006836, "learning_rate": 1.9751871554850876e-06, "loss": 1.3079, "step": 36105 }, { "epoch": 2.829623824451411, "grad_norm": 4.487335681915283, "learning_rate": 1.974941626742559e-06, "loss": 1.7068, "step": 36106 }, { "epoch": 2.829702194357367, "grad_norm": 3.1063315868377686, "learning_rate": 1.9746961095057995e-06, "loss": 1.5451, "step": 36107 }, { "epoch": 2.8297805642633227, "grad_norm": 3.4268927574157715, "learning_rate": 1.974450603775741e-06, "loss": 1.5215, "step": 36108 }, { "epoch": 2.829858934169279, "grad_norm": 2.405565023422241, "learning_rate": 1.9742051095533203e-06, "loss": 1.4554, "step": 36109 }, { "epoch": 2.8299373040752354, "grad_norm": 3.7451581954956055, "learning_rate": 1.9739596268394663e-06, "loss": 1.6052, "step": 36110 }, { "epoch": 2.8300156739811912, "grad_norm": 3.588808059692383, "learning_rate": 1.9737141556351164e-06, "loss": 1.6253, "step": 36111 }, { "epoch": 2.830094043887147, "grad_norm": 3.438267469406128, "learning_rate": 1.9734686959412043e-06, "loss": 1.5152, "step": 36112 }, { "epoch": 2.8301724137931035, "grad_norm": 2.3661770820617676, "learning_rate": 1.9732232477586643e-06, "loss": 0.9539, "step": 36113 }, { "epoch": 2.83025078369906, "grad_norm": 2.6424355506896973, "learning_rate": 1.972977811088426e-06, "loss": 1.3632, "step": 36114 }, { "epoch": 2.8303291536050157, "grad_norm": 3.1205639839172363, "learning_rate": 1.9727323859314248e-06, "loss": 1.3595, "step": 36115 }, { "epoch": 2.8304075235109716, "grad_norm": 3.10329270362854, "learning_rate": 1.9724869722885976e-06, "loss": 1.2345, "step": 36116 }, { "epoch": 2.830485893416928, "grad_norm": 2.1004467010498047, "learning_rate": 1.9722415701608733e-06, "loss": 1.2749, "step": 36117 }, { "epoch": 2.8305642633228842, "grad_norm": 3.6112682819366455, "learning_rate": 1.9719961795491867e-06, "loss": 1.6431, "step": 36118 }, { "epoch": 2.83064263322884, "grad_norm": 2.854534387588501, "learning_rate": 1.971750800454471e-06, "loss": 0.9245, "step": 36119 }, { "epoch": 2.830721003134796, "grad_norm": 1.9104509353637695, "learning_rate": 1.9715054328776607e-06, "loss": 1.1405, "step": 36120 }, { "epoch": 2.8307993730407524, "grad_norm": 3.249361753463745, "learning_rate": 1.971260076819688e-06, "loss": 1.2844, "step": 36121 }, { "epoch": 2.8308777429467087, "grad_norm": 10.750308990478516, "learning_rate": 1.9710147322814855e-06, "loss": 1.5922, "step": 36122 }, { "epoch": 2.8309561128526646, "grad_norm": 3.3371164798736572, "learning_rate": 1.9707693992639875e-06, "loss": 1.137, "step": 36123 }, { "epoch": 2.8310344827586205, "grad_norm": 4.4824042320251465, "learning_rate": 1.970524077768127e-06, "loss": 1.1374, "step": 36124 }, { "epoch": 2.831112852664577, "grad_norm": 4.936127185821533, "learning_rate": 1.9702787677948373e-06, "loss": 1.0788, "step": 36125 }, { "epoch": 2.831191222570533, "grad_norm": 3.3873398303985596, "learning_rate": 1.9700334693450507e-06, "loss": 1.3204, "step": 36126 }, { "epoch": 2.831269592476489, "grad_norm": 3.2344970703125, "learning_rate": 1.9697881824197e-06, "loss": 1.3837, "step": 36127 }, { "epoch": 2.831347962382445, "grad_norm": 3.554372549057007, "learning_rate": 1.969542907019721e-06, "loss": 1.2029, "step": 36128 }, { "epoch": 2.8314263322884012, "grad_norm": 3.1928534507751465, "learning_rate": 1.9692976431460403e-06, "loss": 0.8279, "step": 36129 }, { "epoch": 2.8315047021943576, "grad_norm": 2.270841598510742, "learning_rate": 1.9690523907995968e-06, "loss": 1.3784, "step": 36130 }, { "epoch": 2.8315830721003135, "grad_norm": 3.647836923599243, "learning_rate": 1.9688071499813233e-06, "loss": 1.3049, "step": 36131 }, { "epoch": 2.8316614420062693, "grad_norm": 11.667287826538086, "learning_rate": 1.968561920692148e-06, "loss": 2.22, "step": 36132 }, { "epoch": 2.8317398119122257, "grad_norm": 2.6890645027160645, "learning_rate": 1.9683167029330065e-06, "loss": 1.3883, "step": 36133 }, { "epoch": 2.831818181818182, "grad_norm": 3.8666534423828125, "learning_rate": 1.968071496704829e-06, "loss": 1.3845, "step": 36134 }, { "epoch": 2.831896551724138, "grad_norm": 2.615009069442749, "learning_rate": 1.967826302008554e-06, "loss": 1.4197, "step": 36135 }, { "epoch": 2.831974921630094, "grad_norm": 16.788772583007812, "learning_rate": 1.967581118845109e-06, "loss": 1.3143, "step": 36136 }, { "epoch": 2.83205329153605, "grad_norm": 3.7632617950439453, "learning_rate": 1.967335947215427e-06, "loss": 1.2548, "step": 36137 }, { "epoch": 2.8321316614420065, "grad_norm": 4.7223944664001465, "learning_rate": 1.967090787120442e-06, "loss": 1.8611, "step": 36138 }, { "epoch": 2.8322100313479623, "grad_norm": 6.243790626525879, "learning_rate": 1.9668456385610858e-06, "loss": 0.9933, "step": 36139 }, { "epoch": 2.8322884012539182, "grad_norm": 3.3883252143859863, "learning_rate": 1.9666005015382907e-06, "loss": 1.5879, "step": 36140 }, { "epoch": 2.8323667711598746, "grad_norm": 2.342454433441162, "learning_rate": 1.966355376052989e-06, "loss": 1.3686, "step": 36141 }, { "epoch": 2.832445141065831, "grad_norm": 2.734313726425171, "learning_rate": 1.9661102621061136e-06, "loss": 1.3841, "step": 36142 }, { "epoch": 2.832523510971787, "grad_norm": 3.182565212249756, "learning_rate": 1.965865159698596e-06, "loss": 1.508, "step": 36143 }, { "epoch": 2.8326018808777427, "grad_norm": 3.6734986305236816, "learning_rate": 1.9656200688313694e-06, "loss": 1.1842, "step": 36144 }, { "epoch": 2.832680250783699, "grad_norm": 3.1247615814208984, "learning_rate": 1.9653749895053653e-06, "loss": 1.0342, "step": 36145 }, { "epoch": 2.8327586206896553, "grad_norm": 3.6296417713165283, "learning_rate": 1.965129921721518e-06, "loss": 1.3618, "step": 36146 }, { "epoch": 2.8328369905956112, "grad_norm": 4.097437381744385, "learning_rate": 1.9648848654807556e-06, "loss": 1.0644, "step": 36147 }, { "epoch": 2.8329153605015676, "grad_norm": 4.9497480392456055, "learning_rate": 1.964639820784011e-06, "loss": 1.0414, "step": 36148 }, { "epoch": 2.8329937304075234, "grad_norm": 4.656455039978027, "learning_rate": 1.964394787632219e-06, "loss": 1.0051, "step": 36149 }, { "epoch": 2.83307210031348, "grad_norm": 4.653467178344727, "learning_rate": 1.9641497660263117e-06, "loss": 1.1882, "step": 36150 }, { "epoch": 2.8331504702194357, "grad_norm": 3.626304864883423, "learning_rate": 1.9639047559672175e-06, "loss": 1.6972, "step": 36151 }, { "epoch": 2.833228840125392, "grad_norm": 2.5979268550872803, "learning_rate": 1.96365975745587e-06, "loss": 1.0873, "step": 36152 }, { "epoch": 2.833307210031348, "grad_norm": 2.213562250137329, "learning_rate": 1.9634147704932e-06, "loss": 1.3424, "step": 36153 }, { "epoch": 2.833385579937304, "grad_norm": 6.825122356414795, "learning_rate": 1.9631697950801436e-06, "loss": 1.1586, "step": 36154 }, { "epoch": 2.83346394984326, "grad_norm": 3.287276029586792, "learning_rate": 1.962924831217628e-06, "loss": 1.2347, "step": 36155 }, { "epoch": 2.8335423197492164, "grad_norm": 2.479665994644165, "learning_rate": 1.9626798789065865e-06, "loss": 1.3484, "step": 36156 }, { "epoch": 2.8336206896551723, "grad_norm": 3.00728178024292, "learning_rate": 1.9624349381479503e-06, "loss": 1.5271, "step": 36157 }, { "epoch": 2.8336990595611287, "grad_norm": 2.2923243045806885, "learning_rate": 1.9621900089426517e-06, "loss": 1.5358, "step": 36158 }, { "epoch": 2.8337774294670846, "grad_norm": 2.8408427238464355, "learning_rate": 1.961945091291622e-06, "loss": 1.5493, "step": 36159 }, { "epoch": 2.833855799373041, "grad_norm": 2.90643310546875, "learning_rate": 1.9617001851957924e-06, "loss": 1.3254, "step": 36160 }, { "epoch": 2.8339341692789968, "grad_norm": 4.481573104858398, "learning_rate": 1.961455290656095e-06, "loss": 1.0145, "step": 36161 }, { "epoch": 2.834012539184953, "grad_norm": 3.9471702575683594, "learning_rate": 1.961210407673461e-06, "loss": 1.7711, "step": 36162 }, { "epoch": 2.834090909090909, "grad_norm": 3.1890645027160645, "learning_rate": 1.9609655362488217e-06, "loss": 1.2664, "step": 36163 }, { "epoch": 2.8341692789968653, "grad_norm": 5.318288326263428, "learning_rate": 1.960720676383109e-06, "loss": 0.9651, "step": 36164 }, { "epoch": 2.834247648902821, "grad_norm": 2.52685546875, "learning_rate": 1.9604758280772546e-06, "loss": 1.1626, "step": 36165 }, { "epoch": 2.8343260188087775, "grad_norm": 3.1076924800872803, "learning_rate": 1.960230991332186e-06, "loss": 1.3033, "step": 36166 }, { "epoch": 2.8344043887147334, "grad_norm": 2.982548475265503, "learning_rate": 1.959986166148839e-06, "loss": 1.0534, "step": 36167 }, { "epoch": 2.8344827586206898, "grad_norm": 4.053296089172363, "learning_rate": 1.9597413525281434e-06, "loss": 1.5696, "step": 36168 }, { "epoch": 2.8345611285266457, "grad_norm": 3.5465869903564453, "learning_rate": 1.9594965504710317e-06, "loss": 1.076, "step": 36169 }, { "epoch": 2.834639498432602, "grad_norm": 5.538002014160156, "learning_rate": 1.9592517599784315e-06, "loss": 1.3746, "step": 36170 }, { "epoch": 2.834717868338558, "grad_norm": 2.252612352371216, "learning_rate": 1.9590069810512764e-06, "loss": 1.1561, "step": 36171 }, { "epoch": 2.834796238244514, "grad_norm": 2.7359654903411865, "learning_rate": 1.958762213690496e-06, "loss": 1.2538, "step": 36172 }, { "epoch": 2.83487460815047, "grad_norm": 3.6588056087493896, "learning_rate": 1.9585174578970224e-06, "loss": 1.3899, "step": 36173 }, { "epoch": 2.8349529780564264, "grad_norm": 3.5204384326934814, "learning_rate": 1.9582727136717865e-06, "loss": 1.309, "step": 36174 }, { "epoch": 2.8350313479623823, "grad_norm": 3.4776337146759033, "learning_rate": 1.9580279810157183e-06, "loss": 1.3042, "step": 36175 }, { "epoch": 2.8351097178683387, "grad_norm": 3.6283390522003174, "learning_rate": 1.9577832599297496e-06, "loss": 1.2357, "step": 36176 }, { "epoch": 2.8351880877742945, "grad_norm": 2.7567291259765625, "learning_rate": 1.957538550414811e-06, "loss": 1.4004, "step": 36177 }, { "epoch": 2.835266457680251, "grad_norm": 3.407583475112915, "learning_rate": 1.9572938524718325e-06, "loss": 1.4446, "step": 36178 }, { "epoch": 2.8353448275862068, "grad_norm": 10.738956451416016, "learning_rate": 1.9570491661017455e-06, "loss": 1.9509, "step": 36179 }, { "epoch": 2.835423197492163, "grad_norm": 2.086500406265259, "learning_rate": 1.9568044913054807e-06, "loss": 0.9386, "step": 36180 }, { "epoch": 2.835501567398119, "grad_norm": 3.093257427215576, "learning_rate": 1.956559828083968e-06, "loss": 0.9493, "step": 36181 }, { "epoch": 2.8355799373040753, "grad_norm": 3.5161921977996826, "learning_rate": 1.9563151764381394e-06, "loss": 1.7509, "step": 36182 }, { "epoch": 2.835658307210031, "grad_norm": 4.109861850738525, "learning_rate": 1.956070536368924e-06, "loss": 1.5096, "step": 36183 }, { "epoch": 2.8357366771159875, "grad_norm": 2.2867212295532227, "learning_rate": 1.9558259078772547e-06, "loss": 1.355, "step": 36184 }, { "epoch": 2.8358150470219434, "grad_norm": 2.8195977210998535, "learning_rate": 1.9555812909640565e-06, "loss": 1.2003, "step": 36185 }, { "epoch": 2.8358934169278998, "grad_norm": 3.0683889389038086, "learning_rate": 1.9553366856302646e-06, "loss": 1.2999, "step": 36186 }, { "epoch": 2.8359717868338556, "grad_norm": 3.4466636180877686, "learning_rate": 1.9550920918768107e-06, "loss": 1.652, "step": 36187 }, { "epoch": 2.836050156739812, "grad_norm": 2.8060967922210693, "learning_rate": 1.9548475097046206e-06, "loss": 1.4661, "step": 36188 }, { "epoch": 2.836128526645768, "grad_norm": 2.963229179382324, "learning_rate": 1.954602939114626e-06, "loss": 1.4055, "step": 36189 }, { "epoch": 2.836206896551724, "grad_norm": 2.267899751663208, "learning_rate": 1.9543583801077567e-06, "loss": 1.3862, "step": 36190 }, { "epoch": 2.83628526645768, "grad_norm": 11.729881286621094, "learning_rate": 1.9541138326849463e-06, "loss": 1.1422, "step": 36191 }, { "epoch": 2.8363636363636364, "grad_norm": 3.711479663848877, "learning_rate": 1.953869296847121e-06, "loss": 1.2544, "step": 36192 }, { "epoch": 2.8364420062695928, "grad_norm": 2.4593465328216553, "learning_rate": 1.9536247725952125e-06, "loss": 1.424, "step": 36193 }, { "epoch": 2.8365203761755486, "grad_norm": 3.295104742050171, "learning_rate": 1.9533802599301505e-06, "loss": 1.0257, "step": 36194 }, { "epoch": 2.8365987460815045, "grad_norm": 5.952584266662598, "learning_rate": 1.9531357588528655e-06, "loss": 1.0402, "step": 36195 }, { "epoch": 2.836677115987461, "grad_norm": 2.91198992729187, "learning_rate": 1.9528912693642864e-06, "loss": 1.6195, "step": 36196 }, { "epoch": 2.836755485893417, "grad_norm": 2.1539549827575684, "learning_rate": 1.952646791465345e-06, "loss": 1.3114, "step": 36197 }, { "epoch": 2.836833855799373, "grad_norm": 4.393792152404785, "learning_rate": 1.9524023251569683e-06, "loss": 1.6957, "step": 36198 }, { "epoch": 2.836912225705329, "grad_norm": 4.13090181350708, "learning_rate": 1.952157870440089e-06, "loss": 1.3781, "step": 36199 }, { "epoch": 2.8369905956112853, "grad_norm": 4.825320720672607, "learning_rate": 1.9519134273156354e-06, "loss": 0.9839, "step": 36200 }, { "epoch": 2.8370689655172416, "grad_norm": 2.1303508281707764, "learning_rate": 1.9516689957845373e-06, "loss": 1.5152, "step": 36201 }, { "epoch": 2.8371473354231975, "grad_norm": 2.2153055667877197, "learning_rate": 1.9514245758477266e-06, "loss": 1.6373, "step": 36202 }, { "epoch": 2.8372257053291534, "grad_norm": 3.297433614730835, "learning_rate": 1.95118016750613e-06, "loss": 1.0553, "step": 36203 }, { "epoch": 2.8373040752351097, "grad_norm": 2.6705636978149414, "learning_rate": 1.950935770760675e-06, "loss": 1.5545, "step": 36204 }, { "epoch": 2.837382445141066, "grad_norm": 10.034841537475586, "learning_rate": 1.9506913856122966e-06, "loss": 0.6388, "step": 36205 }, { "epoch": 2.837460815047022, "grad_norm": 2.43520188331604, "learning_rate": 1.950447012061924e-06, "loss": 1.5833, "step": 36206 }, { "epoch": 2.837539184952978, "grad_norm": 3.371295213699341, "learning_rate": 1.950202650110482e-06, "loss": 1.2856, "step": 36207 }, { "epoch": 2.837617554858934, "grad_norm": 3.268141031265259, "learning_rate": 1.9499582997589034e-06, "loss": 1.1784, "step": 36208 }, { "epoch": 2.8376959247648905, "grad_norm": 2.0268137454986572, "learning_rate": 1.949713961008115e-06, "loss": 1.2948, "step": 36209 }, { "epoch": 2.8377742946708464, "grad_norm": 3.5529751777648926, "learning_rate": 1.9494696338590515e-06, "loss": 1.3088, "step": 36210 }, { "epoch": 2.8378526645768023, "grad_norm": 2.364757776260376, "learning_rate": 1.9492253183126365e-06, "loss": 0.92, "step": 36211 }, { "epoch": 2.8379310344827586, "grad_norm": 3.0915377140045166, "learning_rate": 1.9489810143698013e-06, "loss": 1.7565, "step": 36212 }, { "epoch": 2.838009404388715, "grad_norm": 7.315420627593994, "learning_rate": 1.948736722031476e-06, "loss": 1.4205, "step": 36213 }, { "epoch": 2.838087774294671, "grad_norm": 3.774566650390625, "learning_rate": 1.9484924412985885e-06, "loss": 1.6653, "step": 36214 }, { "epoch": 2.8381661442006267, "grad_norm": 3.7885303497314453, "learning_rate": 1.948248172172068e-06, "loss": 1.5946, "step": 36215 }, { "epoch": 2.838244514106583, "grad_norm": 3.815361499786377, "learning_rate": 1.9480039146528445e-06, "loss": 1.2825, "step": 36216 }, { "epoch": 2.8383228840125394, "grad_norm": 2.1848161220550537, "learning_rate": 1.947759668741848e-06, "loss": 0.9115, "step": 36217 }, { "epoch": 2.8384012539184953, "grad_norm": 4.962275981903076, "learning_rate": 1.9475154344400026e-06, "loss": 1.5705, "step": 36218 }, { "epoch": 2.838479623824451, "grad_norm": 3.55648136138916, "learning_rate": 1.9472712117482425e-06, "loss": 1.3569, "step": 36219 }, { "epoch": 2.8385579937304075, "grad_norm": 7.678310394287109, "learning_rate": 1.9470270006674944e-06, "loss": 1.1714, "step": 36220 }, { "epoch": 2.838636363636364, "grad_norm": 3.118773937225342, "learning_rate": 1.9467828011986893e-06, "loss": 1.6248, "step": 36221 }, { "epoch": 2.8387147335423197, "grad_norm": 3.487859010696411, "learning_rate": 1.946538613342753e-06, "loss": 1.1322, "step": 36222 }, { "epoch": 2.8387931034482756, "grad_norm": 1.637344479560852, "learning_rate": 1.946294437100613e-06, "loss": 1.2108, "step": 36223 }, { "epoch": 2.838871473354232, "grad_norm": 4.90041446685791, "learning_rate": 1.946050272473203e-06, "loss": 1.3292, "step": 36224 }, { "epoch": 2.8389498432601883, "grad_norm": 3.5179483890533447, "learning_rate": 1.9458061194614503e-06, "loss": 1.3488, "step": 36225 }, { "epoch": 2.839028213166144, "grad_norm": 3.033555269241333, "learning_rate": 1.945561978066281e-06, "loss": 1.174, "step": 36226 }, { "epoch": 2.8391065830721, "grad_norm": 3.791294574737549, "learning_rate": 1.945317848288625e-06, "loss": 1.5992, "step": 36227 }, { "epoch": 2.8391849529780564, "grad_norm": 2.555410385131836, "learning_rate": 1.9450737301294106e-06, "loss": 1.0602, "step": 36228 }, { "epoch": 2.8392633228840127, "grad_norm": 2.620109796524048, "learning_rate": 1.9448296235895663e-06, "loss": 1.3279, "step": 36229 }, { "epoch": 2.8393416927899686, "grad_norm": 4.466801166534424, "learning_rate": 1.9445855286700217e-06, "loss": 1.3929, "step": 36230 }, { "epoch": 2.8394200626959245, "grad_norm": 3.518475294113159, "learning_rate": 1.944341445371704e-06, "loss": 1.5025, "step": 36231 }, { "epoch": 2.839498432601881, "grad_norm": 3.386996030807495, "learning_rate": 1.9440973736955422e-06, "loss": 1.2861, "step": 36232 }, { "epoch": 2.839576802507837, "grad_norm": 2.62162446975708, "learning_rate": 1.9438533136424644e-06, "loss": 1.1787, "step": 36233 }, { "epoch": 2.839655172413793, "grad_norm": 2.0345215797424316, "learning_rate": 1.9436092652133985e-06, "loss": 1.299, "step": 36234 }, { "epoch": 2.839733542319749, "grad_norm": 2.5840353965759277, "learning_rate": 1.943365228409273e-06, "loss": 1.4466, "step": 36235 }, { "epoch": 2.8398119122257053, "grad_norm": 3.631401538848877, "learning_rate": 1.943121203231018e-06, "loss": 1.0396, "step": 36236 }, { "epoch": 2.8398902821316616, "grad_norm": 3.7692577838897705, "learning_rate": 1.9428771896795565e-06, "loss": 1.331, "step": 36237 }, { "epoch": 2.8399686520376175, "grad_norm": 4.319551467895508, "learning_rate": 1.9426331877558224e-06, "loss": 1.3084, "step": 36238 }, { "epoch": 2.8400470219435734, "grad_norm": 2.7104578018188477, "learning_rate": 1.9423891974607405e-06, "loss": 1.0089, "step": 36239 }, { "epoch": 2.8401253918495297, "grad_norm": 2.938182830810547, "learning_rate": 1.942145218795241e-06, "loss": 1.3753, "step": 36240 }, { "epoch": 2.840203761755486, "grad_norm": 2.3852779865264893, "learning_rate": 1.9419012517602497e-06, "loss": 1.5163, "step": 36241 }, { "epoch": 2.840282131661442, "grad_norm": 12.080599784851074, "learning_rate": 1.941657296356693e-06, "loss": 1.3108, "step": 36242 }, { "epoch": 2.8403605015673983, "grad_norm": 3.615586996078491, "learning_rate": 1.9414133525855046e-06, "loss": 1.3565, "step": 36243 }, { "epoch": 2.840438871473354, "grad_norm": 3.0236542224884033, "learning_rate": 1.9411694204476067e-06, "loss": 1.3461, "step": 36244 }, { "epoch": 2.8405172413793105, "grad_norm": 3.978386878967285, "learning_rate": 1.94092549994393e-06, "loss": 1.4864, "step": 36245 }, { "epoch": 2.8405956112852664, "grad_norm": 2.818744659423828, "learning_rate": 1.9406815910754005e-06, "loss": 0.9728, "step": 36246 }, { "epoch": 2.8406739811912227, "grad_norm": 3.2731430530548096, "learning_rate": 1.9404376938429476e-06, "loss": 1.3734, "step": 36247 }, { "epoch": 2.8407523510971786, "grad_norm": 3.1347968578338623, "learning_rate": 1.940193808247498e-06, "loss": 1.3542, "step": 36248 }, { "epoch": 2.840830721003135, "grad_norm": 11.264204025268555, "learning_rate": 1.9399499342899795e-06, "loss": 1.115, "step": 36249 }, { "epoch": 2.840909090909091, "grad_norm": 3.2873337268829346, "learning_rate": 1.93970607197132e-06, "loss": 1.6065, "step": 36250 }, { "epoch": 2.840987460815047, "grad_norm": 3.6180717945098877, "learning_rate": 1.939462221292447e-06, "loss": 1.6424, "step": 36251 }, { "epoch": 2.841065830721003, "grad_norm": 2.93843412399292, "learning_rate": 1.939218382254287e-06, "loss": 1.5507, "step": 36252 }, { "epoch": 2.8411442006269594, "grad_norm": 7.2376389503479, "learning_rate": 1.938974554857768e-06, "loss": 1.2635, "step": 36253 }, { "epoch": 2.8412225705329153, "grad_norm": 3.9044806957244873, "learning_rate": 1.938730739103818e-06, "loss": 1.5329, "step": 36254 }, { "epoch": 2.8413009404388716, "grad_norm": 3.044018507003784, "learning_rate": 1.938486934993366e-06, "loss": 1.2707, "step": 36255 }, { "epoch": 2.8413793103448275, "grad_norm": 3.3318898677825928, "learning_rate": 1.938243142527333e-06, "loss": 1.23, "step": 36256 }, { "epoch": 2.841457680250784, "grad_norm": 4.736120700836182, "learning_rate": 1.937999361706653e-06, "loss": 0.3144, "step": 36257 }, { "epoch": 2.8415360501567397, "grad_norm": 4.078431129455566, "learning_rate": 1.9377555925322515e-06, "loss": 1.5994, "step": 36258 }, { "epoch": 2.841614420062696, "grad_norm": 2.595158100128174, "learning_rate": 1.9375118350050537e-06, "loss": 1.5561, "step": 36259 }, { "epoch": 2.841692789968652, "grad_norm": 5.431044578552246, "learning_rate": 1.9372680891259875e-06, "loss": 1.6346, "step": 36260 }, { "epoch": 2.8417711598746083, "grad_norm": 14.018353462219238, "learning_rate": 1.937024354895979e-06, "loss": 1.3356, "step": 36261 }, { "epoch": 2.841849529780564, "grad_norm": 2.9847829341888428, "learning_rate": 1.93678063231596e-06, "loss": 1.364, "step": 36262 }, { "epoch": 2.8419278996865205, "grad_norm": 3.7596921920776367, "learning_rate": 1.9365369213868522e-06, "loss": 1.3498, "step": 36263 }, { "epoch": 2.8420062695924764, "grad_norm": 3.350740909576416, "learning_rate": 1.9362932221095846e-06, "loss": 1.2567, "step": 36264 }, { "epoch": 2.8420846394984327, "grad_norm": 2.7420034408569336, "learning_rate": 1.9360495344850844e-06, "loss": 1.3982, "step": 36265 }, { "epoch": 2.8421630094043886, "grad_norm": 3.5440917015075684, "learning_rate": 1.9358058585142774e-06, "loss": 1.345, "step": 36266 }, { "epoch": 2.842241379310345, "grad_norm": 2.829920768737793, "learning_rate": 1.9355621941980914e-06, "loss": 1.2557, "step": 36267 }, { "epoch": 2.842319749216301, "grad_norm": 2.7701470851898193, "learning_rate": 1.9353185415374527e-06, "loss": 1.2215, "step": 36268 }, { "epoch": 2.842398119122257, "grad_norm": 2.775817632675171, "learning_rate": 1.9350749005332876e-06, "loss": 1.1916, "step": 36269 }, { "epoch": 2.842476489028213, "grad_norm": 3.408578634262085, "learning_rate": 1.934831271186524e-06, "loss": 1.1558, "step": 36270 }, { "epoch": 2.8425548589341694, "grad_norm": 2.3892157077789307, "learning_rate": 1.934587653498088e-06, "loss": 1.57, "step": 36271 }, { "epoch": 2.8426332288401253, "grad_norm": 3.212938070297241, "learning_rate": 1.9343440474689056e-06, "loss": 1.232, "step": 36272 }, { "epoch": 2.8427115987460816, "grad_norm": 4.38344669342041, "learning_rate": 1.934100453099906e-06, "loss": 1.5301, "step": 36273 }, { "epoch": 2.8427899686520375, "grad_norm": 3.520874500274658, "learning_rate": 1.9338568703920097e-06, "loss": 1.3997, "step": 36274 }, { "epoch": 2.842868338557994, "grad_norm": 3.546787977218628, "learning_rate": 1.9336132993461486e-06, "loss": 1.4322, "step": 36275 }, { "epoch": 2.8429467084639497, "grad_norm": 7.266629219055176, "learning_rate": 1.933369739963248e-06, "loss": 1.8233, "step": 36276 }, { "epoch": 2.843025078369906, "grad_norm": 8.7808837890625, "learning_rate": 1.933126192244235e-06, "loss": 1.4218, "step": 36277 }, { "epoch": 2.843103448275862, "grad_norm": 4.628717422485352, "learning_rate": 1.932882656190033e-06, "loss": 1.1413, "step": 36278 }, { "epoch": 2.8431818181818183, "grad_norm": 7.154890060424805, "learning_rate": 1.9326391318015686e-06, "loss": 1.2117, "step": 36279 }, { "epoch": 2.843260188087774, "grad_norm": 6.628971099853516, "learning_rate": 1.932395619079771e-06, "loss": 1.7696, "step": 36280 }, { "epoch": 2.8433385579937305, "grad_norm": 2.9964497089385986, "learning_rate": 1.932152118025567e-06, "loss": 1.338, "step": 36281 }, { "epoch": 2.8434169278996864, "grad_norm": 3.047940969467163, "learning_rate": 1.931908628639879e-06, "loss": 1.3426, "step": 36282 }, { "epoch": 2.8434952978056427, "grad_norm": 2.588191270828247, "learning_rate": 1.931665150923633e-06, "loss": 1.3848, "step": 36283 }, { "epoch": 2.8435736677115986, "grad_norm": 2.4501826763153076, "learning_rate": 1.9314216848777584e-06, "loss": 1.3806, "step": 36284 }, { "epoch": 2.843652037617555, "grad_norm": 2.623163938522339, "learning_rate": 1.9311782305031793e-06, "loss": 1.4489, "step": 36285 }, { "epoch": 2.843730407523511, "grad_norm": 3.9957640171051025, "learning_rate": 1.9309347878008223e-06, "loss": 0.9569, "step": 36286 }, { "epoch": 2.843808777429467, "grad_norm": 3.202727794647217, "learning_rate": 1.9306913567716123e-06, "loss": 1.3717, "step": 36287 }, { "epoch": 2.8438871473354235, "grad_norm": 5.6927995681762695, "learning_rate": 1.930447937416476e-06, "loss": 1.3491, "step": 36288 }, { "epoch": 2.8439655172413794, "grad_norm": 2.870826244354248, "learning_rate": 1.9302045297363393e-06, "loss": 1.1713, "step": 36289 }, { "epoch": 2.8440438871473352, "grad_norm": 3.8572587966918945, "learning_rate": 1.929961133732128e-06, "loss": 1.182, "step": 36290 }, { "epoch": 2.8441222570532916, "grad_norm": 2.32967472076416, "learning_rate": 1.9297177494047676e-06, "loss": 1.5026, "step": 36291 }, { "epoch": 2.844200626959248, "grad_norm": 4.541433334350586, "learning_rate": 1.9294743767551855e-06, "loss": 0.7846, "step": 36292 }, { "epoch": 2.844278996865204, "grad_norm": 2.240588903427124, "learning_rate": 1.9292310157843017e-06, "loss": 1.3652, "step": 36293 }, { "epoch": 2.8443573667711597, "grad_norm": 5.748195171356201, "learning_rate": 1.9289876664930486e-06, "loss": 1.1993, "step": 36294 }, { "epoch": 2.844435736677116, "grad_norm": 4.0119147300720215, "learning_rate": 1.928744328882348e-06, "loss": 1.4237, "step": 36295 }, { "epoch": 2.8445141065830724, "grad_norm": 4.936406135559082, "learning_rate": 1.928501002953129e-06, "loss": 1.0816, "step": 36296 }, { "epoch": 2.8445924764890282, "grad_norm": 2.53765606880188, "learning_rate": 1.9282576887063133e-06, "loss": 1.2017, "step": 36297 }, { "epoch": 2.844670846394984, "grad_norm": 2.775113582611084, "learning_rate": 1.9280143861428245e-06, "loss": 1.4158, "step": 36298 }, { "epoch": 2.8447492163009405, "grad_norm": 2.646956205368042, "learning_rate": 1.9277710952635957e-06, "loss": 1.36, "step": 36299 }, { "epoch": 2.844827586206897, "grad_norm": 3.839231252670288, "learning_rate": 1.9275278160695455e-06, "loss": 1.5414, "step": 36300 }, { "epoch": 2.8449059561128527, "grad_norm": 3.183180332183838, "learning_rate": 1.9272845485616015e-06, "loss": 1.002, "step": 36301 }, { "epoch": 2.8449843260188086, "grad_norm": 2.3380281925201416, "learning_rate": 1.9270412927406886e-06, "loss": 1.3741, "step": 36302 }, { "epoch": 2.845062695924765, "grad_norm": 2.596895694732666, "learning_rate": 1.9267980486077326e-06, "loss": 0.9211, "step": 36303 }, { "epoch": 2.8451410658307212, "grad_norm": 5.019472599029541, "learning_rate": 1.9265548161636577e-06, "loss": 1.1979, "step": 36304 }, { "epoch": 2.845219435736677, "grad_norm": 6.189867973327637, "learning_rate": 1.9263115954093904e-06, "loss": 1.5672, "step": 36305 }, { "epoch": 2.845297805642633, "grad_norm": 4.8820343017578125, "learning_rate": 1.9260683863458545e-06, "loss": 0.5913, "step": 36306 }, { "epoch": 2.8453761755485893, "grad_norm": 4.706686973571777, "learning_rate": 1.925825188973976e-06, "loss": 1.4628, "step": 36307 }, { "epoch": 2.8454545454545457, "grad_norm": 2.501217842102051, "learning_rate": 1.9255820032946792e-06, "loss": 1.7372, "step": 36308 }, { "epoch": 2.8455329153605016, "grad_norm": 4.345736980438232, "learning_rate": 1.9253388293088894e-06, "loss": 2.0207, "step": 36309 }, { "epoch": 2.8456112852664575, "grad_norm": 6.933176517486572, "learning_rate": 1.9250956670175315e-06, "loss": 1.2275, "step": 36310 }, { "epoch": 2.845689655172414, "grad_norm": 4.07982063293457, "learning_rate": 1.924852516421532e-06, "loss": 1.4394, "step": 36311 }, { "epoch": 2.84576802507837, "grad_norm": 4.517035961151123, "learning_rate": 1.9246093775218106e-06, "loss": 1.1975, "step": 36312 }, { "epoch": 2.845846394984326, "grad_norm": 2.2394299507141113, "learning_rate": 1.9243662503192966e-06, "loss": 1.1804, "step": 36313 }, { "epoch": 2.845924764890282, "grad_norm": 10.091574668884277, "learning_rate": 1.9241231348149166e-06, "loss": 1.1294, "step": 36314 }, { "epoch": 2.8460031347962382, "grad_norm": 4.221145153045654, "learning_rate": 1.9238800310095897e-06, "loss": 1.2161, "step": 36315 }, { "epoch": 2.8460815047021946, "grad_norm": 3.9147825241088867, "learning_rate": 1.9236369389042437e-06, "loss": 1.0699, "step": 36316 }, { "epoch": 2.8461598746081505, "grad_norm": 2.5358409881591797, "learning_rate": 1.9233938584998003e-06, "loss": 1.1514, "step": 36317 }, { "epoch": 2.8462382445141063, "grad_norm": 6.698034763336182, "learning_rate": 1.9231507897971903e-06, "loss": 1.5244, "step": 36318 }, { "epoch": 2.8463166144200627, "grad_norm": 5.282711982727051, "learning_rate": 1.922907732797333e-06, "loss": 1.4957, "step": 36319 }, { "epoch": 2.846394984326019, "grad_norm": 3.4139091968536377, "learning_rate": 1.9226646875011535e-06, "loss": 1.5536, "step": 36320 }, { "epoch": 2.846473354231975, "grad_norm": 2.0495223999023438, "learning_rate": 1.922421653909577e-06, "loss": 1.4667, "step": 36321 }, { "epoch": 2.846551724137931, "grad_norm": 4.678350448608398, "learning_rate": 1.922178632023528e-06, "loss": 1.6099, "step": 36322 }, { "epoch": 2.846630094043887, "grad_norm": 14.508028984069824, "learning_rate": 1.9219356218439304e-06, "loss": 1.574, "step": 36323 }, { "epoch": 2.8467084639498434, "grad_norm": 7.294650077819824, "learning_rate": 1.9216926233717087e-06, "loss": 1.2509, "step": 36324 }, { "epoch": 2.8467868338557993, "grad_norm": 3.195598602294922, "learning_rate": 1.9214496366077893e-06, "loss": 1.332, "step": 36325 }, { "epoch": 2.8468652037617552, "grad_norm": 3.050978422164917, "learning_rate": 1.9212066615530908e-06, "loss": 1.1058, "step": 36326 }, { "epoch": 2.8469435736677116, "grad_norm": 5.546404838562012, "learning_rate": 1.9209636982085418e-06, "loss": 1.0626, "step": 36327 }, { "epoch": 2.847021943573668, "grad_norm": 4.24238920211792, "learning_rate": 1.920720746575066e-06, "loss": 1.4705, "step": 36328 }, { "epoch": 2.847100313479624, "grad_norm": 3.1129281520843506, "learning_rate": 1.9204778066535882e-06, "loss": 1.4664, "step": 36329 }, { "epoch": 2.8471786833855797, "grad_norm": 3.469330310821533, "learning_rate": 1.9202348784450298e-06, "loss": 1.1551, "step": 36330 }, { "epoch": 2.847257053291536, "grad_norm": 11.267663955688477, "learning_rate": 1.9199919619503136e-06, "loss": 0.6436, "step": 36331 }, { "epoch": 2.8473354231974923, "grad_norm": 6.103605270385742, "learning_rate": 1.919749057170369e-06, "loss": 1.3351, "step": 36332 }, { "epoch": 2.847413793103448, "grad_norm": 5.311374664306641, "learning_rate": 1.9195061641061175e-06, "loss": 1.615, "step": 36333 }, { "epoch": 2.8474921630094046, "grad_norm": 4.6126708984375, "learning_rate": 1.9192632827584807e-06, "loss": 1.4877, "step": 36334 }, { "epoch": 2.8475705329153604, "grad_norm": 4.38617467880249, "learning_rate": 1.919020413128384e-06, "loss": 1.597, "step": 36335 }, { "epoch": 2.8476489028213168, "grad_norm": 3.6382625102996826, "learning_rate": 1.91877755521675e-06, "loss": 1.3771, "step": 36336 }, { "epoch": 2.8477272727272727, "grad_norm": 2.809046983718872, "learning_rate": 1.9185347090245067e-06, "loss": 1.2936, "step": 36337 }, { "epoch": 2.847805642633229, "grad_norm": 3.42331862449646, "learning_rate": 1.9182918745525724e-06, "loss": 1.4497, "step": 36338 }, { "epoch": 2.847884012539185, "grad_norm": 3.9502267837524414, "learning_rate": 1.9180490518018736e-06, "loss": 1.6513, "step": 36339 }, { "epoch": 2.847962382445141, "grad_norm": 2.16217303276062, "learning_rate": 1.917806240773333e-06, "loss": 1.3299, "step": 36340 }, { "epoch": 2.848040752351097, "grad_norm": 9.014312744140625, "learning_rate": 1.917563441467874e-06, "loss": 1.69, "step": 36341 }, { "epoch": 2.8481191222570534, "grad_norm": 7.34281587600708, "learning_rate": 1.917320653886421e-06, "loss": 1.4215, "step": 36342 }, { "epoch": 2.8481974921630093, "grad_norm": 3.938708543777466, "learning_rate": 1.9170778780298967e-06, "loss": 1.7199, "step": 36343 }, { "epoch": 2.8482758620689657, "grad_norm": 2.693392276763916, "learning_rate": 1.9168351138992265e-06, "loss": 1.1462, "step": 36344 }, { "epoch": 2.8483542319749215, "grad_norm": 2.471630334854126, "learning_rate": 1.916592361495328e-06, "loss": 1.241, "step": 36345 }, { "epoch": 2.848432601880878, "grad_norm": 2.3264780044555664, "learning_rate": 1.9163496208191306e-06, "loss": 1.5491, "step": 36346 }, { "epoch": 2.8485109717868338, "grad_norm": 3.829787254333496, "learning_rate": 1.9161068918715553e-06, "loss": 1.5187, "step": 36347 }, { "epoch": 2.84858934169279, "grad_norm": 3.698556661605835, "learning_rate": 1.9158641746535266e-06, "loss": 0.8325, "step": 36348 }, { "epoch": 2.848667711598746, "grad_norm": 8.383718490600586, "learning_rate": 1.9156214691659653e-06, "loss": 0.9074, "step": 36349 }, { "epoch": 2.8487460815047023, "grad_norm": 3.109570026397705, "learning_rate": 1.9153787754097942e-06, "loss": 1.237, "step": 36350 }, { "epoch": 2.848824451410658, "grad_norm": 7.275446891784668, "learning_rate": 1.915136093385939e-06, "loss": 1.6691, "step": 36351 }, { "epoch": 2.8489028213166145, "grad_norm": 2.1736855506896973, "learning_rate": 1.914893423095324e-06, "loss": 1.2707, "step": 36352 }, { "epoch": 2.8489811912225704, "grad_norm": 2.984933614730835, "learning_rate": 1.9146507645388673e-06, "loss": 1.2598, "step": 36353 }, { "epoch": 2.8490595611285268, "grad_norm": 5.520021915435791, "learning_rate": 1.9144081177174945e-06, "loss": 1.4644, "step": 36354 }, { "epoch": 2.8491379310344827, "grad_norm": 2.846813201904297, "learning_rate": 1.9141654826321286e-06, "loss": 1.0736, "step": 36355 }, { "epoch": 2.849216300940439, "grad_norm": 4.803053855895996, "learning_rate": 1.913922859283692e-06, "loss": 1.3632, "step": 36356 }, { "epoch": 2.849294670846395, "grad_norm": 3.915580987930298, "learning_rate": 1.9136802476731083e-06, "loss": 1.1219, "step": 36357 }, { "epoch": 2.849373040752351, "grad_norm": 4.681150436401367, "learning_rate": 1.9134376478012996e-06, "loss": 1.8117, "step": 36358 }, { "epoch": 2.849451410658307, "grad_norm": 2.740000009536743, "learning_rate": 1.9131950596691884e-06, "loss": 1.3561, "step": 36359 }, { "epoch": 2.8495297805642634, "grad_norm": 3.767220973968506, "learning_rate": 1.9129524832776975e-06, "loss": 1.6811, "step": 36360 }, { "epoch": 2.8496081504702193, "grad_norm": 3.1273345947265625, "learning_rate": 1.91270991862775e-06, "loss": 1.0689, "step": 36361 }, { "epoch": 2.8496865203761756, "grad_norm": 3.784022331237793, "learning_rate": 1.9124673657202685e-06, "loss": 1.632, "step": 36362 }, { "epoch": 2.8497648902821315, "grad_norm": 2.6577744483947754, "learning_rate": 1.912224824556175e-06, "loss": 1.5047, "step": 36363 }, { "epoch": 2.849843260188088, "grad_norm": 6.690896511077881, "learning_rate": 1.911982295136392e-06, "loss": 0.6878, "step": 36364 }, { "epoch": 2.8499216300940438, "grad_norm": 5.589537620544434, "learning_rate": 1.911739777461843e-06, "loss": 1.2751, "step": 36365 }, { "epoch": 2.85, "grad_norm": 3.6290409564971924, "learning_rate": 1.911497271533449e-06, "loss": 1.236, "step": 36366 }, { "epoch": 2.850078369905956, "grad_norm": 3.376312017440796, "learning_rate": 1.911254777352135e-06, "loss": 1.121, "step": 36367 }, { "epoch": 2.8501567398119123, "grad_norm": 2.9135961532592773, "learning_rate": 1.9110122949188198e-06, "loss": 1.1781, "step": 36368 }, { "epoch": 2.850235109717868, "grad_norm": 3.1572861671447754, "learning_rate": 1.910769824234425e-06, "loss": 1.3032, "step": 36369 }, { "epoch": 2.8503134796238245, "grad_norm": 3.0124385356903076, "learning_rate": 1.910527365299879e-06, "loss": 1.0903, "step": 36370 }, { "epoch": 2.8503918495297804, "grad_norm": 4.287689685821533, "learning_rate": 1.910284918116098e-06, "loss": 1.7195, "step": 36371 }, { "epoch": 2.8504702194357368, "grad_norm": 5.446481704711914, "learning_rate": 1.9100424826840063e-06, "loss": 1.2782, "step": 36372 }, { "epoch": 2.8505485893416926, "grad_norm": 3.0865254402160645, "learning_rate": 1.9098000590045263e-06, "loss": 1.5962, "step": 36373 }, { "epoch": 2.850626959247649, "grad_norm": 2.7779910564422607, "learning_rate": 1.9095576470785793e-06, "loss": 1.526, "step": 36374 }, { "epoch": 2.850705329153605, "grad_norm": 3.8277437686920166, "learning_rate": 1.9093152469070873e-06, "loss": 1.275, "step": 36375 }, { "epoch": 2.850783699059561, "grad_norm": 14.79847240447998, "learning_rate": 1.909072858490973e-06, "loss": 1.0123, "step": 36376 }, { "epoch": 2.850862068965517, "grad_norm": 4.2801079750061035, "learning_rate": 1.908830481831158e-06, "loss": 1.2232, "step": 36377 }, { "epoch": 2.8509404388714734, "grad_norm": 5.567538738250732, "learning_rate": 1.908588116928564e-06, "loss": 1.5278, "step": 36378 }, { "epoch": 2.8510188087774293, "grad_norm": 2.811945915222168, "learning_rate": 1.908345763784113e-06, "loss": 1.3401, "step": 36379 }, { "epoch": 2.8510971786833856, "grad_norm": 3.6966941356658936, "learning_rate": 1.9081034223987267e-06, "loss": 1.6847, "step": 36380 }, { "epoch": 2.8511755485893415, "grad_norm": 2.5537221431732178, "learning_rate": 1.9078610927733287e-06, "loss": 1.5211, "step": 36381 }, { "epoch": 2.851253918495298, "grad_norm": 2.7663633823394775, "learning_rate": 1.9076187749088353e-06, "loss": 1.4067, "step": 36382 }, { "epoch": 2.851332288401254, "grad_norm": 2.7995805740356445, "learning_rate": 1.9073764688061737e-06, "loss": 1.4642, "step": 36383 }, { "epoch": 2.85141065830721, "grad_norm": 4.758383274078369, "learning_rate": 1.907134174466263e-06, "loss": 1.1049, "step": 36384 }, { "epoch": 2.851489028213166, "grad_norm": 3.701068878173828, "learning_rate": 1.9068918918900275e-06, "loss": 1.3563, "step": 36385 }, { "epoch": 2.8515673981191223, "grad_norm": 2.650463104248047, "learning_rate": 1.9066496210783846e-06, "loss": 1.454, "step": 36386 }, { "epoch": 2.8516457680250786, "grad_norm": 7.238039493560791, "learning_rate": 1.906407362032256e-06, "loss": 1.3868, "step": 36387 }, { "epoch": 2.8517241379310345, "grad_norm": 3.963477611541748, "learning_rate": 1.9061651147525667e-06, "loss": 1.5598, "step": 36388 }, { "epoch": 2.8518025078369904, "grad_norm": 4.919084548950195, "learning_rate": 1.905922879240238e-06, "loss": 1.4903, "step": 36389 }, { "epoch": 2.8518808777429467, "grad_norm": 2.9033446311950684, "learning_rate": 1.9056806554961877e-06, "loss": 0.6694, "step": 36390 }, { "epoch": 2.851959247648903, "grad_norm": 4.004477500915527, "learning_rate": 1.9054384435213386e-06, "loss": 1.7723, "step": 36391 }, { "epoch": 2.852037617554859, "grad_norm": 2.833153486251831, "learning_rate": 1.9051962433166106e-06, "loss": 1.1956, "step": 36392 }, { "epoch": 2.852115987460815, "grad_norm": 3.3762500286102295, "learning_rate": 1.9049540548829297e-06, "loss": 1.1106, "step": 36393 }, { "epoch": 2.852194357366771, "grad_norm": 2.894901990890503, "learning_rate": 1.9047118782212127e-06, "loss": 1.0905, "step": 36394 }, { "epoch": 2.8522727272727275, "grad_norm": 3.840111017227173, "learning_rate": 1.9044697133323813e-06, "loss": 1.3016, "step": 36395 }, { "epoch": 2.8523510971786834, "grad_norm": 5.258129596710205, "learning_rate": 1.904227560217357e-06, "loss": 1.0849, "step": 36396 }, { "epoch": 2.8524294670846393, "grad_norm": 3.133638381958008, "learning_rate": 1.9039854188770617e-06, "loss": 1.3375, "step": 36397 }, { "epoch": 2.8525078369905956, "grad_norm": 2.3654210567474365, "learning_rate": 1.9037432893124153e-06, "loss": 1.453, "step": 36398 }, { "epoch": 2.852586206896552, "grad_norm": 9.168608665466309, "learning_rate": 1.903501171524339e-06, "loss": 1.0883, "step": 36399 }, { "epoch": 2.852664576802508, "grad_norm": 3.887883424758911, "learning_rate": 1.9032590655137557e-06, "loss": 1.2795, "step": 36400 }, { "epoch": 2.8527429467084637, "grad_norm": 9.652997970581055, "learning_rate": 1.9030169712815805e-06, "loss": 0.834, "step": 36401 }, { "epoch": 2.85282131661442, "grad_norm": 4.129860877990723, "learning_rate": 1.90277488882874e-06, "loss": 1.6257, "step": 36402 }, { "epoch": 2.8528996865203764, "grad_norm": 3.975276470184326, "learning_rate": 1.902532818156153e-06, "loss": 1.2824, "step": 36403 }, { "epoch": 2.8529780564263323, "grad_norm": 3.0985958576202393, "learning_rate": 1.902290759264742e-06, "loss": 1.1251, "step": 36404 }, { "epoch": 2.853056426332288, "grad_norm": 4.50383186340332, "learning_rate": 1.9020487121554242e-06, "loss": 1.1969, "step": 36405 }, { "epoch": 2.8531347962382445, "grad_norm": 4.508534908294678, "learning_rate": 1.90180667682912e-06, "loss": 1.1015, "step": 36406 }, { "epoch": 2.853213166144201, "grad_norm": 1.8595802783966064, "learning_rate": 1.9015646532867538e-06, "loss": 1.1631, "step": 36407 }, { "epoch": 2.8532915360501567, "grad_norm": 4.479672431945801, "learning_rate": 1.9013226415292456e-06, "loss": 1.2903, "step": 36408 }, { "epoch": 2.8533699059561126, "grad_norm": 3.8353676795959473, "learning_rate": 1.9010806415575133e-06, "loss": 1.5108, "step": 36409 }, { "epoch": 2.853448275862069, "grad_norm": 3.1961681842803955, "learning_rate": 1.9008386533724783e-06, "loss": 0.8997, "step": 36410 }, { "epoch": 2.8535266457680253, "grad_norm": 2.498164653778076, "learning_rate": 1.9005966769750606e-06, "loss": 1.2541, "step": 36411 }, { "epoch": 2.853605015673981, "grad_norm": 3.705106258392334, "learning_rate": 1.900354712366182e-06, "loss": 1.3137, "step": 36412 }, { "epoch": 2.853683385579937, "grad_norm": 2.677966356277466, "learning_rate": 1.9001127595467617e-06, "loss": 1.2427, "step": 36413 }, { "epoch": 2.8537617554858934, "grad_norm": 3.6353893280029297, "learning_rate": 1.8998708185177201e-06, "loss": 1.1381, "step": 36414 }, { "epoch": 2.8538401253918497, "grad_norm": 2.9371557235717773, "learning_rate": 1.8996288892799775e-06, "loss": 1.2139, "step": 36415 }, { "epoch": 2.8539184952978056, "grad_norm": 2.7220377922058105, "learning_rate": 1.8993869718344548e-06, "loss": 1.1459, "step": 36416 }, { "epoch": 2.8539968652037615, "grad_norm": 2.1519572734832764, "learning_rate": 1.899145066182071e-06, "loss": 1.5482, "step": 36417 }, { "epoch": 2.854075235109718, "grad_norm": 3.590867519378662, "learning_rate": 1.8989031723237466e-06, "loss": 1.0808, "step": 36418 }, { "epoch": 2.854153605015674, "grad_norm": 2.8882577419281006, "learning_rate": 1.8986612902604034e-06, "loss": 1.4785, "step": 36419 }, { "epoch": 2.85423197492163, "grad_norm": 2.881591558456421, "learning_rate": 1.8984194199929568e-06, "loss": 1.2802, "step": 36420 }, { "epoch": 2.854310344827586, "grad_norm": 2.54115629196167, "learning_rate": 1.8981775615223308e-06, "loss": 1.2757, "step": 36421 }, { "epoch": 2.8543887147335423, "grad_norm": 2.494316816329956, "learning_rate": 1.897935714849446e-06, "loss": 1.3768, "step": 36422 }, { "epoch": 2.8544670846394986, "grad_norm": 2.1954615116119385, "learning_rate": 1.8976938799752187e-06, "loss": 1.127, "step": 36423 }, { "epoch": 2.8545454545454545, "grad_norm": 4.735429763793945, "learning_rate": 1.8974520569005705e-06, "loss": 1.1472, "step": 36424 }, { "epoch": 2.8546238244514104, "grad_norm": 3.322472095489502, "learning_rate": 1.8972102456264196e-06, "loss": 0.923, "step": 36425 }, { "epoch": 2.8547021943573667, "grad_norm": 2.9637610912323, "learning_rate": 1.8969684461536902e-06, "loss": 1.4075, "step": 36426 }, { "epoch": 2.854780564263323, "grad_norm": 3.799149513244629, "learning_rate": 1.8967266584832976e-06, "loss": 1.3432, "step": 36427 }, { "epoch": 2.854858934169279, "grad_norm": 3.5579802989959717, "learning_rate": 1.896484882616163e-06, "loss": 1.4596, "step": 36428 }, { "epoch": 2.8549373040752353, "grad_norm": 5.895802021026611, "learning_rate": 1.8962431185532053e-06, "loss": 1.152, "step": 36429 }, { "epoch": 2.855015673981191, "grad_norm": 5.9361891746521, "learning_rate": 1.8960013662953452e-06, "loss": 0.9575, "step": 36430 }, { "epoch": 2.8550940438871475, "grad_norm": 5.603152751922607, "learning_rate": 1.8957596258435013e-06, "loss": 1.0482, "step": 36431 }, { "epoch": 2.8551724137931034, "grad_norm": 4.070849418640137, "learning_rate": 1.8955178971985933e-06, "loss": 1.7962, "step": 36432 }, { "epoch": 2.8552507836990597, "grad_norm": 6.100856304168701, "learning_rate": 1.895276180361541e-06, "loss": 1.7888, "step": 36433 }, { "epoch": 2.8553291536050156, "grad_norm": 2.9810116291046143, "learning_rate": 1.8950344753332633e-06, "loss": 1.5777, "step": 36434 }, { "epoch": 2.855407523510972, "grad_norm": 2.3055338859558105, "learning_rate": 1.8947927821146795e-06, "loss": 1.1947, "step": 36435 }, { "epoch": 2.855485893416928, "grad_norm": 2.4818153381347656, "learning_rate": 1.8945511007067085e-06, "loss": 1.1612, "step": 36436 }, { "epoch": 2.855564263322884, "grad_norm": 4.409136772155762, "learning_rate": 1.8943094311102727e-06, "loss": 1.2602, "step": 36437 }, { "epoch": 2.85564263322884, "grad_norm": 9.559934616088867, "learning_rate": 1.8940677733262863e-06, "loss": 2.2006, "step": 36438 }, { "epoch": 2.8557210031347964, "grad_norm": 3.57798433303833, "learning_rate": 1.8938261273556695e-06, "loss": 0.9829, "step": 36439 }, { "epoch": 2.8557993730407523, "grad_norm": 2.648780345916748, "learning_rate": 1.893584493199344e-06, "loss": 1.2525, "step": 36440 }, { "epoch": 2.8558777429467086, "grad_norm": 3.7865631580352783, "learning_rate": 1.8933428708582297e-06, "loss": 1.1842, "step": 36441 }, { "epoch": 2.8559561128526645, "grad_norm": 4.693609237670898, "learning_rate": 1.893101260333241e-06, "loss": 1.2962, "step": 36442 }, { "epoch": 2.856034482758621, "grad_norm": 3.80717134475708, "learning_rate": 1.8928596616252999e-06, "loss": 1.5327, "step": 36443 }, { "epoch": 2.8561128526645767, "grad_norm": 3.4390015602111816, "learning_rate": 1.8926180747353229e-06, "loss": 1.0229, "step": 36444 }, { "epoch": 2.856191222570533, "grad_norm": 2.665045738220215, "learning_rate": 1.892376499664234e-06, "loss": 1.4136, "step": 36445 }, { "epoch": 2.856269592476489, "grad_norm": 4.343697547912598, "learning_rate": 1.8921349364129471e-06, "loss": 1.5719, "step": 36446 }, { "epoch": 2.8563479623824453, "grad_norm": 3.1330795288085938, "learning_rate": 1.8918933849823829e-06, "loss": 1.5954, "step": 36447 }, { "epoch": 2.856426332288401, "grad_norm": 2.2891643047332764, "learning_rate": 1.8916518453734595e-06, "loss": 1.1978, "step": 36448 }, { "epoch": 2.8565047021943575, "grad_norm": 2.620673179626465, "learning_rate": 1.8914103175870957e-06, "loss": 1.1087, "step": 36449 }, { "epoch": 2.8565830721003134, "grad_norm": 3.5603199005126953, "learning_rate": 1.8911688016242108e-06, "loss": 1.3922, "step": 36450 }, { "epoch": 2.8566614420062697, "grad_norm": 2.8554677963256836, "learning_rate": 1.890927297485723e-06, "loss": 1.5183, "step": 36451 }, { "epoch": 2.8567398119122256, "grad_norm": 2.8401849269866943, "learning_rate": 1.8906858051725519e-06, "loss": 1.5285, "step": 36452 }, { "epoch": 2.856818181818182, "grad_norm": 2.8790831565856934, "learning_rate": 1.890444324685612e-06, "loss": 1.8035, "step": 36453 }, { "epoch": 2.856896551724138, "grad_norm": 4.056947708129883, "learning_rate": 1.8902028560258268e-06, "loss": 1.4093, "step": 36454 }, { "epoch": 2.856974921630094, "grad_norm": 2.9542295932769775, "learning_rate": 1.8899613991941117e-06, "loss": 1.5922, "step": 36455 }, { "epoch": 2.85705329153605, "grad_norm": 5.833892822265625, "learning_rate": 1.8897199541913886e-06, "loss": 1.5817, "step": 36456 }, { "epoch": 2.8571316614420064, "grad_norm": 6.33287239074707, "learning_rate": 1.8894785210185712e-06, "loss": 1.5177, "step": 36457 }, { "epoch": 2.8572100313479623, "grad_norm": 2.8024914264678955, "learning_rate": 1.8892370996765786e-06, "loss": 1.5242, "step": 36458 }, { "epoch": 2.8572884012539186, "grad_norm": 4.6977458000183105, "learning_rate": 1.8889956901663315e-06, "loss": 1.0191, "step": 36459 }, { "epoch": 2.8573667711598745, "grad_norm": 5.433083534240723, "learning_rate": 1.8887542924887486e-06, "loss": 2.0442, "step": 36460 }, { "epoch": 2.857445141065831, "grad_norm": 3.5007436275482178, "learning_rate": 1.8885129066447443e-06, "loss": 1.3933, "step": 36461 }, { "epoch": 2.8575235109717867, "grad_norm": 3.0701169967651367, "learning_rate": 1.8882715326352392e-06, "loss": 1.2319, "step": 36462 }, { "epoch": 2.857601880877743, "grad_norm": 4.534866809844971, "learning_rate": 1.8880301704611492e-06, "loss": 1.4428, "step": 36463 }, { "epoch": 2.857680250783699, "grad_norm": 2.9903950691223145, "learning_rate": 1.8877888201233973e-06, "loss": 1.1814, "step": 36464 }, { "epoch": 2.8577586206896552, "grad_norm": 8.264372825622559, "learning_rate": 1.8875474816228967e-06, "loss": 0.9089, "step": 36465 }, { "epoch": 2.857836990595611, "grad_norm": 5.329712390899658, "learning_rate": 1.8873061549605665e-06, "loss": 1.4133, "step": 36466 }, { "epoch": 2.8579153605015675, "grad_norm": 3.011704444885254, "learning_rate": 1.8870648401373255e-06, "loss": 1.3601, "step": 36467 }, { "epoch": 2.8579937304075234, "grad_norm": 5.129711627960205, "learning_rate": 1.8868235371540906e-06, "loss": 1.7112, "step": 36468 }, { "epoch": 2.8580721003134797, "grad_norm": 2.8123581409454346, "learning_rate": 1.8865822460117795e-06, "loss": 1.332, "step": 36469 }, { "epoch": 2.8581504702194356, "grad_norm": 3.056212902069092, "learning_rate": 1.8863409667113107e-06, "loss": 1.34, "step": 36470 }, { "epoch": 2.858228840125392, "grad_norm": 4.283818244934082, "learning_rate": 1.8860996992536018e-06, "loss": 1.2964, "step": 36471 }, { "epoch": 2.858307210031348, "grad_norm": 3.1920177936553955, "learning_rate": 1.88585844363957e-06, "loss": 1.5379, "step": 36472 }, { "epoch": 2.858385579937304, "grad_norm": 3.822779893875122, "learning_rate": 1.8856171998701335e-06, "loss": 1.3351, "step": 36473 }, { "epoch": 2.8584639498432605, "grad_norm": 5.1995134353637695, "learning_rate": 1.8853759679462092e-06, "loss": 1.2316, "step": 36474 }, { "epoch": 2.8585423197492164, "grad_norm": 2.4398293495178223, "learning_rate": 1.8851347478687165e-06, "loss": 1.1142, "step": 36475 }, { "epoch": 2.8586206896551722, "grad_norm": 3.3547325134277344, "learning_rate": 1.884893539638568e-06, "loss": 1.271, "step": 36476 }, { "epoch": 2.8586990595611286, "grad_norm": 3.297090768814087, "learning_rate": 1.8846523432566865e-06, "loss": 1.4338, "step": 36477 }, { "epoch": 2.858777429467085, "grad_norm": 2.9070143699645996, "learning_rate": 1.8844111587239889e-06, "loss": 1.1411, "step": 36478 }, { "epoch": 2.858855799373041, "grad_norm": 4.675008773803711, "learning_rate": 1.8841699860413893e-06, "loss": 0.58, "step": 36479 }, { "epoch": 2.8589341692789967, "grad_norm": 6.730550765991211, "learning_rate": 1.8839288252098064e-06, "loss": 1.1583, "step": 36480 }, { "epoch": 2.859012539184953, "grad_norm": 2.265702486038208, "learning_rate": 1.8836876762301558e-06, "loss": 1.1907, "step": 36481 }, { "epoch": 2.8590909090909093, "grad_norm": 2.980264186859131, "learning_rate": 1.8834465391033607e-06, "loss": 1.0809, "step": 36482 }, { "epoch": 2.8591692789968652, "grad_norm": 3.5791194438934326, "learning_rate": 1.8832054138303325e-06, "loss": 1.4965, "step": 36483 }, { "epoch": 2.859247648902821, "grad_norm": 3.4684345722198486, "learning_rate": 1.8829643004119897e-06, "loss": 1.3184, "step": 36484 }, { "epoch": 2.8593260188087775, "grad_norm": 3.7170426845550537, "learning_rate": 1.8827231988492495e-06, "loss": 0.9356, "step": 36485 }, { "epoch": 2.859404388714734, "grad_norm": 3.3402621746063232, "learning_rate": 1.8824821091430295e-06, "loss": 1.3839, "step": 36486 }, { "epoch": 2.8594827586206897, "grad_norm": 2.6557719707489014, "learning_rate": 1.8822410312942464e-06, "loss": 1.0856, "step": 36487 }, { "epoch": 2.8595611285266456, "grad_norm": 3.934581756591797, "learning_rate": 1.8819999653038169e-06, "loss": 1.3866, "step": 36488 }, { "epoch": 2.859639498432602, "grad_norm": 3.628840923309326, "learning_rate": 1.8817589111726576e-06, "loss": 1.4827, "step": 36489 }, { "epoch": 2.8597178683385582, "grad_norm": 3.8115670680999756, "learning_rate": 1.8815178689016862e-06, "loss": 1.2858, "step": 36490 }, { "epoch": 2.859796238244514, "grad_norm": 11.992700576782227, "learning_rate": 1.881276838491819e-06, "loss": 1.0244, "step": 36491 }, { "epoch": 2.85987460815047, "grad_norm": 2.36911940574646, "learning_rate": 1.8810358199439726e-06, "loss": 1.2008, "step": 36492 }, { "epoch": 2.8599529780564263, "grad_norm": 6.074010848999023, "learning_rate": 1.8807948132590659e-06, "loss": 1.594, "step": 36493 }, { "epoch": 2.8600313479623827, "grad_norm": 4.65071439743042, "learning_rate": 1.8805538184380124e-06, "loss": 1.4477, "step": 36494 }, { "epoch": 2.8601097178683386, "grad_norm": 2.9502341747283936, "learning_rate": 1.8803128354817274e-06, "loss": 1.2529, "step": 36495 }, { "epoch": 2.8601880877742945, "grad_norm": 5.1168317794799805, "learning_rate": 1.880071864391132e-06, "loss": 0.976, "step": 36496 }, { "epoch": 2.860266457680251, "grad_norm": 3.058577060699463, "learning_rate": 1.8798309051671426e-06, "loss": 1.0075, "step": 36497 }, { "epoch": 2.860344827586207, "grad_norm": 3.907151222229004, "learning_rate": 1.8795899578106718e-06, "loss": 1.4674, "step": 36498 }, { "epoch": 2.860423197492163, "grad_norm": 15.247297286987305, "learning_rate": 1.879349022322638e-06, "loss": 2.2708, "step": 36499 }, { "epoch": 2.860501567398119, "grad_norm": 3.239915609359741, "learning_rate": 1.8791080987039563e-06, "loss": 1.3152, "step": 36500 }, { "epoch": 2.8605799373040752, "grad_norm": 2.6156461238861084, "learning_rate": 1.8788671869555475e-06, "loss": 1.7117, "step": 36501 }, { "epoch": 2.8606583072100316, "grad_norm": 2.850358724594116, "learning_rate": 1.8786262870783233e-06, "loss": 1.3865, "step": 36502 }, { "epoch": 2.8607366771159874, "grad_norm": 6.014634609222412, "learning_rate": 1.8783853990732014e-06, "loss": 1.2345, "step": 36503 }, { "epoch": 2.8608150470219433, "grad_norm": 3.991361141204834, "learning_rate": 1.8781445229410983e-06, "loss": 1.4182, "step": 36504 }, { "epoch": 2.8608934169278997, "grad_norm": 4.01522970199585, "learning_rate": 1.8779036586829296e-06, "loss": 2.1434, "step": 36505 }, { "epoch": 2.860971786833856, "grad_norm": 3.7317471504211426, "learning_rate": 1.877662806299612e-06, "loss": 1.2225, "step": 36506 }, { "epoch": 2.861050156739812, "grad_norm": 9.059572219848633, "learning_rate": 1.8774219657920611e-06, "loss": 0.5833, "step": 36507 }, { "epoch": 2.8611285266457678, "grad_norm": 3.7723283767700195, "learning_rate": 1.8771811371611953e-06, "loss": 1.3577, "step": 36508 }, { "epoch": 2.861206896551724, "grad_norm": 5.1501312255859375, "learning_rate": 1.8769403204079244e-06, "loss": 1.4427, "step": 36509 }, { "epoch": 2.8612852664576804, "grad_norm": 3.99133038520813, "learning_rate": 1.8766995155331708e-06, "loss": 1.6123, "step": 36510 }, { "epoch": 2.8613636363636363, "grad_norm": 3.7399678230285645, "learning_rate": 1.8764587225378478e-06, "loss": 1.4336, "step": 36511 }, { "epoch": 2.861442006269592, "grad_norm": 2.5374536514282227, "learning_rate": 1.8762179414228733e-06, "loss": 0.8651, "step": 36512 }, { "epoch": 2.8615203761755486, "grad_norm": 3.724410057067871, "learning_rate": 1.8759771721891595e-06, "loss": 1.2855, "step": 36513 }, { "epoch": 2.861598746081505, "grad_norm": 2.7896828651428223, "learning_rate": 1.8757364148376216e-06, "loss": 1.2187, "step": 36514 }, { "epoch": 2.8616771159874608, "grad_norm": 3.3289568424224854, "learning_rate": 1.87549566936918e-06, "loss": 1.058, "step": 36515 }, { "epoch": 2.8617554858934167, "grad_norm": 2.400466203689575, "learning_rate": 1.87525493578475e-06, "loss": 1.2478, "step": 36516 }, { "epoch": 2.861833855799373, "grad_norm": 3.975632429122925, "learning_rate": 1.8750142140852428e-06, "loss": 1.3442, "step": 36517 }, { "epoch": 2.8619122257053293, "grad_norm": 4.179839134216309, "learning_rate": 1.874773504271577e-06, "loss": 1.5482, "step": 36518 }, { "epoch": 2.861990595611285, "grad_norm": 2.2516376972198486, "learning_rate": 1.8745328063446655e-06, "loss": 1.0394, "step": 36519 }, { "epoch": 2.862068965517241, "grad_norm": 3.705573558807373, "learning_rate": 1.87429212030543e-06, "loss": 1.3784, "step": 36520 }, { "epoch": 2.8621473354231974, "grad_norm": 2.7332375049591064, "learning_rate": 1.8740514461547792e-06, "loss": 1.4285, "step": 36521 }, { "epoch": 2.8622257053291538, "grad_norm": 5.3927836418151855, "learning_rate": 1.8738107838936314e-06, "loss": 1.2233, "step": 36522 }, { "epoch": 2.8623040752351097, "grad_norm": 2.724214792251587, "learning_rate": 1.8735701335229017e-06, "loss": 1.4943, "step": 36523 }, { "epoch": 2.862382445141066, "grad_norm": 4.515649795532227, "learning_rate": 1.8733294950435054e-06, "loss": 1.097, "step": 36524 }, { "epoch": 2.862460815047022, "grad_norm": 2.6218514442443848, "learning_rate": 1.873088868456358e-06, "loss": 1.1583, "step": 36525 }, { "epoch": 2.862539184952978, "grad_norm": 3.292776107788086, "learning_rate": 1.8728482537623742e-06, "loss": 1.6619, "step": 36526 }, { "epoch": 2.862617554858934, "grad_norm": 2.2911558151245117, "learning_rate": 1.8726076509624713e-06, "loss": 1.5769, "step": 36527 }, { "epoch": 2.8626959247648904, "grad_norm": 4.3127593994140625, "learning_rate": 1.8723670600575589e-06, "loss": 1.8463, "step": 36528 }, { "epoch": 2.8627742946708463, "grad_norm": 3.3533759117126465, "learning_rate": 1.8721264810485573e-06, "loss": 1.2952, "step": 36529 }, { "epoch": 2.8628526645768027, "grad_norm": 2.8631155490875244, "learning_rate": 1.8718859139363804e-06, "loss": 1.4742, "step": 36530 }, { "epoch": 2.8629310344827585, "grad_norm": 5.49349308013916, "learning_rate": 1.8716453587219442e-06, "loss": 1.2963, "step": 36531 }, { "epoch": 2.863009404388715, "grad_norm": 7.116938591003418, "learning_rate": 1.871404815406161e-06, "loss": 1.9768, "step": 36532 }, { "epoch": 2.8630877742946708, "grad_norm": 4.08711051940918, "learning_rate": 1.8711642839899447e-06, "loss": 1.3057, "step": 36533 }, { "epoch": 2.863166144200627, "grad_norm": 2.632596492767334, "learning_rate": 1.8709237644742162e-06, "loss": 1.3234, "step": 36534 }, { "epoch": 2.863244514106583, "grad_norm": 3.8278183937072754, "learning_rate": 1.8706832568598843e-06, "loss": 1.167, "step": 36535 }, { "epoch": 2.8633228840125393, "grad_norm": 2.8250174522399902, "learning_rate": 1.8704427611478665e-06, "loss": 1.3213, "step": 36536 }, { "epoch": 2.863401253918495, "grad_norm": 3.7276148796081543, "learning_rate": 1.8702022773390766e-06, "loss": 0.5931, "step": 36537 }, { "epoch": 2.8634796238244515, "grad_norm": 3.364508867263794, "learning_rate": 1.8699618054344298e-06, "loss": 1.3787, "step": 36538 }, { "epoch": 2.8635579937304074, "grad_norm": 8.561005592346191, "learning_rate": 1.8697213454348406e-06, "loss": 1.2973, "step": 36539 }, { "epoch": 2.8636363636363638, "grad_norm": 3.1808295249938965, "learning_rate": 1.8694808973412237e-06, "loss": 1.4237, "step": 36540 }, { "epoch": 2.8637147335423196, "grad_norm": 3.1606199741363525, "learning_rate": 1.8692404611544935e-06, "loss": 1.5895, "step": 36541 }, { "epoch": 2.863793103448276, "grad_norm": 3.084484338760376, "learning_rate": 1.8690000368755646e-06, "loss": 1.5285, "step": 36542 }, { "epoch": 2.863871473354232, "grad_norm": 4.277162551879883, "learning_rate": 1.8687596245053514e-06, "loss": 1.4804, "step": 36543 }, { "epoch": 2.863949843260188, "grad_norm": 11.243241310119629, "learning_rate": 1.868519224044768e-06, "loss": 1.4342, "step": 36544 }, { "epoch": 2.864028213166144, "grad_norm": 11.147107124328613, "learning_rate": 1.868278835494729e-06, "loss": 1.544, "step": 36545 }, { "epoch": 2.8641065830721004, "grad_norm": 3.3194923400878906, "learning_rate": 1.8680384588561507e-06, "loss": 1.5496, "step": 36546 }, { "epoch": 2.8641849529780563, "grad_norm": 6.53208589553833, "learning_rate": 1.8677980941299416e-06, "loss": 1.0085, "step": 36547 }, { "epoch": 2.8642633228840126, "grad_norm": 3.4300568103790283, "learning_rate": 1.867557741317022e-06, "loss": 1.7269, "step": 36548 }, { "epoch": 2.8643416927899685, "grad_norm": 9.63489818572998, "learning_rate": 1.8673174004183054e-06, "loss": 1.3161, "step": 36549 }, { "epoch": 2.864420062695925, "grad_norm": 6.739866256713867, "learning_rate": 1.8670770714347024e-06, "loss": 0.6356, "step": 36550 }, { "epoch": 2.8644984326018808, "grad_norm": 10.713606834411621, "learning_rate": 1.8668367543671295e-06, "loss": 2.2134, "step": 36551 }, { "epoch": 2.864576802507837, "grad_norm": 2.2667057514190674, "learning_rate": 1.8665964492164984e-06, "loss": 1.262, "step": 36552 }, { "epoch": 2.864655172413793, "grad_norm": 3.6890573501586914, "learning_rate": 1.8663561559837285e-06, "loss": 1.3588, "step": 36553 }, { "epoch": 2.8647335423197493, "grad_norm": 2.915984630584717, "learning_rate": 1.8661158746697284e-06, "loss": 1.7665, "step": 36554 }, { "epoch": 2.864811912225705, "grad_norm": 2.7936460971832275, "learning_rate": 1.8658756052754135e-06, "loss": 1.4774, "step": 36555 }, { "epoch": 2.8648902821316615, "grad_norm": 7.916283130645752, "learning_rate": 1.8656353478016987e-06, "loss": 1.4427, "step": 36556 }, { "epoch": 2.8649686520376174, "grad_norm": 2.9396538734436035, "learning_rate": 1.8653951022494965e-06, "loss": 1.8732, "step": 36557 }, { "epoch": 2.8650470219435737, "grad_norm": 2.5343809127807617, "learning_rate": 1.8651548686197218e-06, "loss": 1.1739, "step": 36558 }, { "epoch": 2.8651253918495296, "grad_norm": 3.1428301334381104, "learning_rate": 1.8649146469132873e-06, "loss": 1.4997, "step": 36559 }, { "epoch": 2.865203761755486, "grad_norm": 2.2640469074249268, "learning_rate": 1.8646744371311076e-06, "loss": 1.1271, "step": 36560 }, { "epoch": 2.865282131661442, "grad_norm": 3.8595452308654785, "learning_rate": 1.8644342392740962e-06, "loss": 1.5888, "step": 36561 }, { "epoch": 2.865360501567398, "grad_norm": 3.0228071212768555, "learning_rate": 1.8641940533431657e-06, "loss": 1.2005, "step": 36562 }, { "epoch": 2.865438871473354, "grad_norm": 4.217039108276367, "learning_rate": 1.863953879339231e-06, "loss": 1.1626, "step": 36563 }, { "epoch": 2.8655172413793104, "grad_norm": 4.324820041656494, "learning_rate": 1.8637137172632059e-06, "loss": 0.8027, "step": 36564 }, { "epoch": 2.8655956112852663, "grad_norm": 1.9862436056137085, "learning_rate": 1.863473567116002e-06, "loss": 1.5077, "step": 36565 }, { "epoch": 2.8656739811912226, "grad_norm": 3.381986379623413, "learning_rate": 1.8632334288985316e-06, "loss": 1.2305, "step": 36566 }, { "epoch": 2.8657523510971785, "grad_norm": 3.0014052391052246, "learning_rate": 1.8629933026117114e-06, "loss": 1.0853, "step": 36567 }, { "epoch": 2.865830721003135, "grad_norm": 2.0740513801574707, "learning_rate": 1.862753188256456e-06, "loss": 1.0527, "step": 36568 }, { "epoch": 2.865909090909091, "grad_norm": 4.281207084655762, "learning_rate": 1.8625130858336736e-06, "loss": 0.8497, "step": 36569 }, { "epoch": 2.865987460815047, "grad_norm": 4.4274516105651855, "learning_rate": 1.8622729953442803e-06, "loss": 1.0166, "step": 36570 }, { "epoch": 2.866065830721003, "grad_norm": 6.075578212738037, "learning_rate": 1.862032916789187e-06, "loss": 0.6929, "step": 36571 }, { "epoch": 2.8661442006269593, "grad_norm": 2.330808639526367, "learning_rate": 1.8617928501693122e-06, "loss": 1.2566, "step": 36572 }, { "epoch": 2.8662225705329156, "grad_norm": 2.9156253337860107, "learning_rate": 1.8615527954855638e-06, "loss": 1.3164, "step": 36573 }, { "epoch": 2.8663009404388715, "grad_norm": 4.239295482635498, "learning_rate": 1.8613127527388568e-06, "loss": 1.5655, "step": 36574 }, { "epoch": 2.8663793103448274, "grad_norm": 3.689183235168457, "learning_rate": 1.861072721930104e-06, "loss": 1.3062, "step": 36575 }, { "epoch": 2.8664576802507837, "grad_norm": 2.674198865890503, "learning_rate": 1.860832703060218e-06, "loss": 1.4548, "step": 36576 }, { "epoch": 2.86653605015674, "grad_norm": 8.000907897949219, "learning_rate": 1.8605926961301119e-06, "loss": 0.8721, "step": 36577 }, { "epoch": 2.866614420062696, "grad_norm": 3.8956706523895264, "learning_rate": 1.860352701140699e-06, "loss": 0.9322, "step": 36578 }, { "epoch": 2.866692789968652, "grad_norm": 4.163868427276611, "learning_rate": 1.8601127180928913e-06, "loss": 1.6698, "step": 36579 }, { "epoch": 2.866771159874608, "grad_norm": 2.497680902481079, "learning_rate": 1.8598727469876027e-06, "loss": 1.418, "step": 36580 }, { "epoch": 2.8668495297805645, "grad_norm": 3.0863707065582275, "learning_rate": 1.8596327878257447e-06, "loss": 1.4972, "step": 36581 }, { "epoch": 2.8669278996865204, "grad_norm": 3.1138370037078857, "learning_rate": 1.8593928406082313e-06, "loss": 1.3251, "step": 36582 }, { "epoch": 2.8670062695924763, "grad_norm": 4.390643119812012, "learning_rate": 1.859152905335976e-06, "loss": 1.382, "step": 36583 }, { "epoch": 2.8670846394984326, "grad_norm": 2.704995632171631, "learning_rate": 1.8589129820098855e-06, "loss": 0.9776, "step": 36584 }, { "epoch": 2.867163009404389, "grad_norm": 4.313614368438721, "learning_rate": 1.8586730706308793e-06, "loss": 1.8213, "step": 36585 }, { "epoch": 2.867241379310345, "grad_norm": 4.86577033996582, "learning_rate": 1.8584331711998672e-06, "loss": 1.3838, "step": 36586 }, { "epoch": 2.8673197492163007, "grad_norm": 3.060105085372925, "learning_rate": 1.858193283717763e-06, "loss": 1.2622, "step": 36587 }, { "epoch": 2.867398119122257, "grad_norm": 5.741509437561035, "learning_rate": 1.857953408185476e-06, "loss": 1.4749, "step": 36588 }, { "epoch": 2.8674764890282134, "grad_norm": 6.51408052444458, "learning_rate": 1.857713544603919e-06, "loss": 0.9857, "step": 36589 }, { "epoch": 2.8675548589341693, "grad_norm": 4.928686618804932, "learning_rate": 1.857473692974009e-06, "loss": 1.2702, "step": 36590 }, { "epoch": 2.867633228840125, "grad_norm": 2.550206422805786, "learning_rate": 1.8572338532966532e-06, "loss": 1.4864, "step": 36591 }, { "epoch": 2.8677115987460815, "grad_norm": 3.520829916000366, "learning_rate": 1.8569940255727658e-06, "loss": 1.3795, "step": 36592 }, { "epoch": 2.867789968652038, "grad_norm": 3.588423490524292, "learning_rate": 1.8567542098032587e-06, "loss": 1.8175, "step": 36593 }, { "epoch": 2.8678683385579937, "grad_norm": 2.855090856552124, "learning_rate": 1.8565144059890444e-06, "loss": 1.1782, "step": 36594 }, { "epoch": 2.8679467084639496, "grad_norm": 4.002984523773193, "learning_rate": 1.856274614131034e-06, "loss": 1.0172, "step": 36595 }, { "epoch": 2.868025078369906, "grad_norm": 3.8727312088012695, "learning_rate": 1.856034834230141e-06, "loss": 1.5844, "step": 36596 }, { "epoch": 2.8681034482758623, "grad_norm": 2.6716649532318115, "learning_rate": 1.8557950662872759e-06, "loss": 1.157, "step": 36597 }, { "epoch": 2.868181818181818, "grad_norm": 6.0205559730529785, "learning_rate": 1.8555553103033519e-06, "loss": 1.6205, "step": 36598 }, { "epoch": 2.868260188087774, "grad_norm": 4.481168270111084, "learning_rate": 1.8553155662792805e-06, "loss": 2.0246, "step": 36599 }, { "epoch": 2.8683385579937304, "grad_norm": 3.0760130882263184, "learning_rate": 1.8550758342159731e-06, "loss": 1.6346, "step": 36600 }, { "epoch": 2.8684169278996867, "grad_norm": 4.2238688468933105, "learning_rate": 1.8548361141143422e-06, "loss": 1.3614, "step": 36601 }, { "epoch": 2.8684952978056426, "grad_norm": 2.87654447555542, "learning_rate": 1.8545964059753008e-06, "loss": 1.1178, "step": 36602 }, { "epoch": 2.8685736677115985, "grad_norm": 2.209705352783203, "learning_rate": 1.8543567097997556e-06, "loss": 1.2937, "step": 36603 }, { "epoch": 2.868652037617555, "grad_norm": 3.4983222484588623, "learning_rate": 1.8541170255886237e-06, "loss": 1.277, "step": 36604 }, { "epoch": 2.868730407523511, "grad_norm": 3.0149989128112793, "learning_rate": 1.8538773533428167e-06, "loss": 0.769, "step": 36605 }, { "epoch": 2.868808777429467, "grad_norm": 4.143492221832275, "learning_rate": 1.8536376930632422e-06, "loss": 1.4807, "step": 36606 }, { "epoch": 2.868887147335423, "grad_norm": 2.581537961959839, "learning_rate": 1.8533980447508138e-06, "loss": 1.3039, "step": 36607 }, { "epoch": 2.8689655172413793, "grad_norm": 3.0345256328582764, "learning_rate": 1.8531584084064413e-06, "loss": 1.0504, "step": 36608 }, { "epoch": 2.8690438871473356, "grad_norm": 5.300792694091797, "learning_rate": 1.8529187840310415e-06, "loss": 1.3422, "step": 36609 }, { "epoch": 2.8691222570532915, "grad_norm": 2.7968523502349854, "learning_rate": 1.8526791716255205e-06, "loss": 1.3255, "step": 36610 }, { "epoch": 2.8692006269592474, "grad_norm": 5.894394874572754, "learning_rate": 1.8524395711907918e-06, "loss": 1.1715, "step": 36611 }, { "epoch": 2.8692789968652037, "grad_norm": 2.8064422607421875, "learning_rate": 1.8521999827277654e-06, "loss": 1.5337, "step": 36612 }, { "epoch": 2.86935736677116, "grad_norm": 4.708746910095215, "learning_rate": 1.8519604062373543e-06, "loss": 1.2318, "step": 36613 }, { "epoch": 2.869435736677116, "grad_norm": 3.011408805847168, "learning_rate": 1.851720841720468e-06, "loss": 1.0924, "step": 36614 }, { "epoch": 2.8695141065830723, "grad_norm": 2.671830177307129, "learning_rate": 1.8514812891780192e-06, "loss": 1.1992, "step": 36615 }, { "epoch": 2.869592476489028, "grad_norm": 2.733031988143921, "learning_rate": 1.851241748610918e-06, "loss": 1.1936, "step": 36616 }, { "epoch": 2.8696708463949845, "grad_norm": 2.7632205486297607, "learning_rate": 1.8510022200200761e-06, "loss": 1.093, "step": 36617 }, { "epoch": 2.8697492163009404, "grad_norm": 2.697157859802246, "learning_rate": 1.8507627034064046e-06, "loss": 1.5525, "step": 36618 }, { "epoch": 2.8698275862068967, "grad_norm": 4.041130542755127, "learning_rate": 1.850523198770814e-06, "loss": 0.9833, "step": 36619 }, { "epoch": 2.8699059561128526, "grad_norm": 3.553481101989746, "learning_rate": 1.850283706114217e-06, "loss": 1.4158, "step": 36620 }, { "epoch": 2.869984326018809, "grad_norm": 2.951577663421631, "learning_rate": 1.8500442254375212e-06, "loss": 1.4436, "step": 36621 }, { "epoch": 2.870062695924765, "grad_norm": 4.0314555168151855, "learning_rate": 1.8498047567416376e-06, "loss": 1.4175, "step": 36622 }, { "epoch": 2.870141065830721, "grad_norm": 4.5072784423828125, "learning_rate": 1.849565300027481e-06, "loss": 1.3473, "step": 36623 }, { "epoch": 2.870219435736677, "grad_norm": 2.55747652053833, "learning_rate": 1.8493258552959608e-06, "loss": 1.1989, "step": 36624 }, { "epoch": 2.8702978056426334, "grad_norm": 3.0077953338623047, "learning_rate": 1.8490864225479855e-06, "loss": 0.6164, "step": 36625 }, { "epoch": 2.8703761755485893, "grad_norm": 3.2357072830200195, "learning_rate": 1.8488470017844668e-06, "loss": 1.2014, "step": 36626 }, { "epoch": 2.8704545454545456, "grad_norm": 2.6772589683532715, "learning_rate": 1.848607593006314e-06, "loss": 1.6949, "step": 36627 }, { "epoch": 2.8705329153605015, "grad_norm": 2.3184280395507812, "learning_rate": 1.8483681962144423e-06, "loss": 1.3974, "step": 36628 }, { "epoch": 2.870611285266458, "grad_norm": 3.7879133224487305, "learning_rate": 1.8481288114097578e-06, "loss": 1.5818, "step": 36629 }, { "epoch": 2.8706896551724137, "grad_norm": 4.130122661590576, "learning_rate": 1.8478894385931729e-06, "loss": 1.339, "step": 36630 }, { "epoch": 2.87076802507837, "grad_norm": 3.392345666885376, "learning_rate": 1.8476500777655964e-06, "loss": 1.1494, "step": 36631 }, { "epoch": 2.870846394984326, "grad_norm": 4.8020100593566895, "learning_rate": 1.8474107289279407e-06, "loss": 1.4295, "step": 36632 }, { "epoch": 2.8709247648902823, "grad_norm": 3.7810182571411133, "learning_rate": 1.8471713920811151e-06, "loss": 1.0887, "step": 36633 }, { "epoch": 2.871003134796238, "grad_norm": 3.524726390838623, "learning_rate": 1.8469320672260305e-06, "loss": 0.9607, "step": 36634 }, { "epoch": 2.8710815047021945, "grad_norm": 2.289168357849121, "learning_rate": 1.8466927543635976e-06, "loss": 1.3121, "step": 36635 }, { "epoch": 2.8711598746081504, "grad_norm": 8.80986213684082, "learning_rate": 1.8464534534947232e-06, "loss": 0.818, "step": 36636 }, { "epoch": 2.8712382445141067, "grad_norm": 3.8273720741271973, "learning_rate": 1.8462141646203213e-06, "loss": 1.401, "step": 36637 }, { "epoch": 2.8713166144200626, "grad_norm": 4.801496982574463, "learning_rate": 1.8459748877413015e-06, "loss": 1.6338, "step": 36638 }, { "epoch": 2.871394984326019, "grad_norm": 3.2434966564178467, "learning_rate": 1.8457356228585744e-06, "loss": 1.291, "step": 36639 }, { "epoch": 2.871473354231975, "grad_norm": 3.2526438236236572, "learning_rate": 1.8454963699730471e-06, "loss": 1.1945, "step": 36640 }, { "epoch": 2.871551724137931, "grad_norm": 2.7754244804382324, "learning_rate": 1.8452571290856297e-06, "loss": 1.2805, "step": 36641 }, { "epoch": 2.871630094043887, "grad_norm": 1.7073428630828857, "learning_rate": 1.8450179001972357e-06, "loss": 1.2743, "step": 36642 }, { "epoch": 2.8717084639498434, "grad_norm": 6.2873687744140625, "learning_rate": 1.844778683308775e-06, "loss": 2.1049, "step": 36643 }, { "epoch": 2.8717868338557992, "grad_norm": 2.38700008392334, "learning_rate": 1.8445394784211534e-06, "loss": 1.3557, "step": 36644 }, { "epoch": 2.8718652037617556, "grad_norm": 2.9386675357818604, "learning_rate": 1.8443002855352833e-06, "loss": 1.2875, "step": 36645 }, { "epoch": 2.8719435736677115, "grad_norm": 3.1261329650878906, "learning_rate": 1.8440611046520735e-06, "loss": 1.7643, "step": 36646 }, { "epoch": 2.872021943573668, "grad_norm": 14.745927810668945, "learning_rate": 1.8438219357724346e-06, "loss": 1.728, "step": 36647 }, { "epoch": 2.8721003134796237, "grad_norm": 2.786674737930298, "learning_rate": 1.843582778897276e-06, "loss": 1.4171, "step": 36648 }, { "epoch": 2.87217868338558, "grad_norm": 2.9809179306030273, "learning_rate": 1.8433436340275073e-06, "loss": 1.6089, "step": 36649 }, { "epoch": 2.872257053291536, "grad_norm": 2.880985736846924, "learning_rate": 1.843104501164038e-06, "loss": 1.6148, "step": 36650 }, { "epoch": 2.8723354231974922, "grad_norm": 3.4168686866760254, "learning_rate": 1.8428653803077772e-06, "loss": 0.8618, "step": 36651 }, { "epoch": 2.872413793103448, "grad_norm": 3.123767375946045, "learning_rate": 1.8426262714596355e-06, "loss": 1.2462, "step": 36652 }, { "epoch": 2.8724921630094045, "grad_norm": 3.0496835708618164, "learning_rate": 1.842387174620522e-06, "loss": 1.4885, "step": 36653 }, { "epoch": 2.8725705329153604, "grad_norm": 3.0825557708740234, "learning_rate": 1.8421480897913468e-06, "loss": 1.5385, "step": 36654 }, { "epoch": 2.8726489028213167, "grad_norm": 7.450226306915283, "learning_rate": 1.8419090169730153e-06, "loss": 1.3489, "step": 36655 }, { "epoch": 2.8727272727272726, "grad_norm": 2.5105180740356445, "learning_rate": 1.8416699561664415e-06, "loss": 1.1325, "step": 36656 }, { "epoch": 2.872805642633229, "grad_norm": 2.5253820419311523, "learning_rate": 1.8414309073725323e-06, "loss": 1.3216, "step": 36657 }, { "epoch": 2.872884012539185, "grad_norm": 2.876786947250366, "learning_rate": 1.8411918705922e-06, "loss": 1.707, "step": 36658 }, { "epoch": 2.872962382445141, "grad_norm": 4.282082557678223, "learning_rate": 1.8409528458263492e-06, "loss": 1.5451, "step": 36659 }, { "epoch": 2.873040752351097, "grad_norm": 17.36520767211914, "learning_rate": 1.84071383307589e-06, "loss": 1.3969, "step": 36660 }, { "epoch": 2.8731191222570533, "grad_norm": 2.9687039852142334, "learning_rate": 1.840474832341736e-06, "loss": 1.4764, "step": 36661 }, { "epoch": 2.8731974921630092, "grad_norm": 4.491177082061768, "learning_rate": 1.8402358436247909e-06, "loss": 1.4929, "step": 36662 }, { "epoch": 2.8732758620689656, "grad_norm": 2.6739401817321777, "learning_rate": 1.839996866925966e-06, "loss": 1.4509, "step": 36663 }, { "epoch": 2.873354231974922, "grad_norm": 2.536104917526245, "learning_rate": 1.8397579022461697e-06, "loss": 1.3095, "step": 36664 }, { "epoch": 2.873432601880878, "grad_norm": 3.7708957195281982, "learning_rate": 1.8395189495863115e-06, "loss": 1.0973, "step": 36665 }, { "epoch": 2.8735109717868337, "grad_norm": 2.9728617668151855, "learning_rate": 1.8392800089472994e-06, "loss": 1.1192, "step": 36666 }, { "epoch": 2.87358934169279, "grad_norm": 3.8356211185455322, "learning_rate": 1.8390410803300428e-06, "loss": 0.9817, "step": 36667 }, { "epoch": 2.8736677115987463, "grad_norm": 3.5444719791412354, "learning_rate": 1.8388021637354503e-06, "loss": 1.2635, "step": 36668 }, { "epoch": 2.8737460815047022, "grad_norm": 3.8283629417419434, "learning_rate": 1.838563259164431e-06, "loss": 1.2415, "step": 36669 }, { "epoch": 2.873824451410658, "grad_norm": 3.062166452407837, "learning_rate": 1.8383243666178929e-06, "loss": 1.5697, "step": 36670 }, { "epoch": 2.8739028213166145, "grad_norm": 4.021528720855713, "learning_rate": 1.838085486096745e-06, "loss": 1.09, "step": 36671 }, { "epoch": 2.873981191222571, "grad_norm": 3.409916639328003, "learning_rate": 1.8378466176018961e-06, "loss": 1.1404, "step": 36672 }, { "epoch": 2.8740595611285267, "grad_norm": 7.00140905380249, "learning_rate": 1.8376077611342546e-06, "loss": 1.8968, "step": 36673 }, { "epoch": 2.8741379310344826, "grad_norm": 2.7424821853637695, "learning_rate": 1.8373689166947285e-06, "loss": 1.4947, "step": 36674 }, { "epoch": 2.874216300940439, "grad_norm": 3.5460703372955322, "learning_rate": 1.8371300842842272e-06, "loss": 1.2346, "step": 36675 }, { "epoch": 2.8742946708463952, "grad_norm": 2.816478967666626, "learning_rate": 1.8368912639036595e-06, "loss": 0.9596, "step": 36676 }, { "epoch": 2.874373040752351, "grad_norm": 3.1210832595825195, "learning_rate": 1.836652455553931e-06, "loss": 1.2907, "step": 36677 }, { "epoch": 2.874451410658307, "grad_norm": 2.596233367919922, "learning_rate": 1.8364136592359506e-06, "loss": 1.0269, "step": 36678 }, { "epoch": 2.8745297805642633, "grad_norm": 4.014586925506592, "learning_rate": 1.8361748749506291e-06, "loss": 1.5873, "step": 36679 }, { "epoch": 2.8746081504702197, "grad_norm": 3.3763554096221924, "learning_rate": 1.8359361026988754e-06, "loss": 1.0518, "step": 36680 }, { "epoch": 2.8746865203761756, "grad_norm": 2.7141499519348145, "learning_rate": 1.8356973424815932e-06, "loss": 1.1322, "step": 36681 }, { "epoch": 2.8747648902821314, "grad_norm": 2.8531885147094727, "learning_rate": 1.8354585942996938e-06, "loss": 1.5094, "step": 36682 }, { "epoch": 2.874843260188088, "grad_norm": 3.1712238788604736, "learning_rate": 1.8352198581540842e-06, "loss": 1.3746, "step": 36683 }, { "epoch": 2.874921630094044, "grad_norm": 3.0560355186462402, "learning_rate": 1.8349811340456724e-06, "loss": 0.9668, "step": 36684 }, { "epoch": 2.875, "grad_norm": 3.3910560607910156, "learning_rate": 1.8347424219753674e-06, "loss": 1.6752, "step": 36685 }, { "epoch": 2.875078369905956, "grad_norm": 3.2619032859802246, "learning_rate": 1.8345037219440758e-06, "loss": 1.3145, "step": 36686 }, { "epoch": 2.875156739811912, "grad_norm": 2.9967949390411377, "learning_rate": 1.8342650339527063e-06, "loss": 1.1092, "step": 36687 }, { "epoch": 2.8752351097178686, "grad_norm": 3.979374885559082, "learning_rate": 1.8340263580021668e-06, "loss": 1.6966, "step": 36688 }, { "epoch": 2.8753134796238244, "grad_norm": 3.345207452774048, "learning_rate": 1.8337876940933646e-06, "loss": 1.0893, "step": 36689 }, { "epoch": 2.8753918495297803, "grad_norm": 5.1642866134643555, "learning_rate": 1.8335490422272078e-06, "loss": 1.4881, "step": 36690 }, { "epoch": 2.8754702194357367, "grad_norm": 2.35437273979187, "learning_rate": 1.8333104024046062e-06, "loss": 1.3516, "step": 36691 }, { "epoch": 2.875548589341693, "grad_norm": 3.689617872238159, "learning_rate": 1.8330717746264615e-06, "loss": 1.3251, "step": 36692 }, { "epoch": 2.875626959247649, "grad_norm": 3.1748063564300537, "learning_rate": 1.8328331588936872e-06, "loss": 0.9426, "step": 36693 }, { "epoch": 2.8757053291536048, "grad_norm": 3.7166056632995605, "learning_rate": 1.8325945552071884e-06, "loss": 1.4381, "step": 36694 }, { "epoch": 2.875783699059561, "grad_norm": 3.404978036880493, "learning_rate": 1.8323559635678744e-06, "loss": 1.462, "step": 36695 }, { "epoch": 2.8758620689655174, "grad_norm": 2.1397781372070312, "learning_rate": 1.8321173839766504e-06, "loss": 1.5137, "step": 36696 }, { "epoch": 2.8759404388714733, "grad_norm": 2.533193826675415, "learning_rate": 1.8318788164344225e-06, "loss": 1.3254, "step": 36697 }, { "epoch": 2.876018808777429, "grad_norm": 3.494802951812744, "learning_rate": 1.831640260942102e-06, "loss": 1.2753, "step": 36698 }, { "epoch": 2.8760971786833855, "grad_norm": 5.66323184967041, "learning_rate": 1.8314017175005966e-06, "loss": 1.1507, "step": 36699 }, { "epoch": 2.876175548589342, "grad_norm": 5.63712739944458, "learning_rate": 1.8311631861108097e-06, "loss": 1.587, "step": 36700 }, { "epoch": 2.8762539184952978, "grad_norm": 3.108464479446411, "learning_rate": 1.8309246667736507e-06, "loss": 1.4776, "step": 36701 }, { "epoch": 2.8763322884012537, "grad_norm": 3.891519546508789, "learning_rate": 1.8306861594900261e-06, "loss": 1.2848, "step": 36702 }, { "epoch": 2.87641065830721, "grad_norm": 3.1155219078063965, "learning_rate": 1.8304476642608438e-06, "loss": 1.3203, "step": 36703 }, { "epoch": 2.8764890282131663, "grad_norm": 4.204837322235107, "learning_rate": 1.8302091810870104e-06, "loss": 1.2953, "step": 36704 }, { "epoch": 2.876567398119122, "grad_norm": 9.575753211975098, "learning_rate": 1.8299707099694337e-06, "loss": 1.3813, "step": 36705 }, { "epoch": 2.876645768025078, "grad_norm": 4.372086524963379, "learning_rate": 1.8297322509090194e-06, "loss": 1.5809, "step": 36706 }, { "epoch": 2.8767241379310344, "grad_norm": 5.804941177368164, "learning_rate": 1.8294938039066756e-06, "loss": 1.1841, "step": 36707 }, { "epoch": 2.8768025078369908, "grad_norm": 2.9933130741119385, "learning_rate": 1.829255368963309e-06, "loss": 1.1848, "step": 36708 }, { "epoch": 2.8768808777429467, "grad_norm": 5.5960259437561035, "learning_rate": 1.8290169460798263e-06, "loss": 0.7294, "step": 36709 }, { "epoch": 2.876959247648903, "grad_norm": 2.4716784954071045, "learning_rate": 1.8287785352571358e-06, "loss": 1.3513, "step": 36710 }, { "epoch": 2.877037617554859, "grad_norm": 2.705096483230591, "learning_rate": 1.8285401364961403e-06, "loss": 1.4073, "step": 36711 }, { "epoch": 2.877115987460815, "grad_norm": 3.317702293395996, "learning_rate": 1.82830174979775e-06, "loss": 1.5272, "step": 36712 }, { "epoch": 2.877194357366771, "grad_norm": 3.401353120803833, "learning_rate": 1.8280633751628712e-06, "loss": 1.4565, "step": 36713 }, { "epoch": 2.8772727272727274, "grad_norm": 3.0805506706237793, "learning_rate": 1.8278250125924119e-06, "loss": 1.1693, "step": 36714 }, { "epoch": 2.8773510971786833, "grad_norm": 4.279000282287598, "learning_rate": 1.827586662087275e-06, "loss": 1.4997, "step": 36715 }, { "epoch": 2.8774294670846396, "grad_norm": 2.176492214202881, "learning_rate": 1.8273483236483674e-06, "loss": 1.22, "step": 36716 }, { "epoch": 2.8775078369905955, "grad_norm": 3.6524581909179688, "learning_rate": 1.8271099972766005e-06, "loss": 0.9829, "step": 36717 }, { "epoch": 2.877586206896552, "grad_norm": 2.5643808841705322, "learning_rate": 1.8268716829728756e-06, "loss": 1.6271, "step": 36718 }, { "epoch": 2.8776645768025078, "grad_norm": 4.8555073738098145, "learning_rate": 1.8266333807381015e-06, "loss": 1.1732, "step": 36719 }, { "epoch": 2.877742946708464, "grad_norm": 6.082702159881592, "learning_rate": 1.8263950905731836e-06, "loss": 0.8173, "step": 36720 }, { "epoch": 2.87782131661442, "grad_norm": 7.181314945220947, "learning_rate": 1.8261568124790291e-06, "loss": 1.8155, "step": 36721 }, { "epoch": 2.8778996865203763, "grad_norm": 3.5672078132629395, "learning_rate": 1.8259185464565437e-06, "loss": 1.2942, "step": 36722 }, { "epoch": 2.877978056426332, "grad_norm": 2.8504679203033447, "learning_rate": 1.8256802925066335e-06, "loss": 1.3394, "step": 36723 }, { "epoch": 2.8780564263322885, "grad_norm": 2.872380018234253, "learning_rate": 1.8254420506302056e-06, "loss": 1.2828, "step": 36724 }, { "epoch": 2.8781347962382444, "grad_norm": 3.3669302463531494, "learning_rate": 1.8252038208281653e-06, "loss": 1.2961, "step": 36725 }, { "epoch": 2.8782131661442008, "grad_norm": 6.840297222137451, "learning_rate": 1.824965603101419e-06, "loss": 1.3107, "step": 36726 }, { "epoch": 2.8782915360501566, "grad_norm": 4.724040508270264, "learning_rate": 1.8247273974508727e-06, "loss": 1.5133, "step": 36727 }, { "epoch": 2.878369905956113, "grad_norm": 5.554474830627441, "learning_rate": 1.8244892038774325e-06, "loss": 1.6446, "step": 36728 }, { "epoch": 2.878448275862069, "grad_norm": 2.5285234451293945, "learning_rate": 1.8242510223820064e-06, "loss": 1.1557, "step": 36729 }, { "epoch": 2.878526645768025, "grad_norm": 6.333286762237549, "learning_rate": 1.8240128529654944e-06, "loss": 1.2882, "step": 36730 }, { "epoch": 2.878605015673981, "grad_norm": 3.445303440093994, "learning_rate": 1.8237746956288076e-06, "loss": 1.2447, "step": 36731 }, { "epoch": 2.8786833855799374, "grad_norm": 3.6638379096984863, "learning_rate": 1.8235365503728531e-06, "loss": 1.4461, "step": 36732 }, { "epoch": 2.8787617554858933, "grad_norm": 2.9468839168548584, "learning_rate": 1.8232984171985319e-06, "loss": 1.4932, "step": 36733 }, { "epoch": 2.8788401253918496, "grad_norm": 3.85097336769104, "learning_rate": 1.823060296106751e-06, "loss": 0.8185, "step": 36734 }, { "epoch": 2.8789184952978055, "grad_norm": 3.0418202877044678, "learning_rate": 1.8228221870984164e-06, "loss": 1.0456, "step": 36735 }, { "epoch": 2.878996865203762, "grad_norm": 2.529385566711426, "learning_rate": 1.8225840901744373e-06, "loss": 1.1508, "step": 36736 }, { "epoch": 2.8790752351097177, "grad_norm": 4.442322254180908, "learning_rate": 1.8223460053357145e-06, "loss": 0.7347, "step": 36737 }, { "epoch": 2.879153605015674, "grad_norm": 4.17185640335083, "learning_rate": 1.8221079325831558e-06, "loss": 1.5198, "step": 36738 }, { "epoch": 2.87923197492163, "grad_norm": 5.678986072540283, "learning_rate": 1.821869871917666e-06, "loss": 1.7095, "step": 36739 }, { "epoch": 2.8793103448275863, "grad_norm": 2.207629919052124, "learning_rate": 1.8216318233401504e-06, "loss": 1.3365, "step": 36740 }, { "epoch": 2.879388714733542, "grad_norm": 3.740750551223755, "learning_rate": 1.8213937868515152e-06, "loss": 1.4566, "step": 36741 }, { "epoch": 2.8794670846394985, "grad_norm": 2.960862398147583, "learning_rate": 1.8211557624526655e-06, "loss": 1.4017, "step": 36742 }, { "epoch": 2.8795454545454544, "grad_norm": 2.45393705368042, "learning_rate": 1.820917750144508e-06, "loss": 1.3388, "step": 36743 }, { "epoch": 2.8796238244514107, "grad_norm": 4.000968933105469, "learning_rate": 1.8206797499279433e-06, "loss": 1.3089, "step": 36744 }, { "epoch": 2.8797021943573666, "grad_norm": 4.686366081237793, "learning_rate": 1.820441761803881e-06, "loss": 1.3042, "step": 36745 }, { "epoch": 2.879780564263323, "grad_norm": 3.3968441486358643, "learning_rate": 1.8202037857732257e-06, "loss": 1.1966, "step": 36746 }, { "epoch": 2.879858934169279, "grad_norm": 2.327007532119751, "learning_rate": 1.8199658218368837e-06, "loss": 1.2316, "step": 36747 }, { "epoch": 2.879937304075235, "grad_norm": 3.876342296600342, "learning_rate": 1.819727869995756e-06, "loss": 1.2659, "step": 36748 }, { "epoch": 2.880015673981191, "grad_norm": 2.970207929611206, "learning_rate": 1.819489930250749e-06, "loss": 1.1121, "step": 36749 }, { "epoch": 2.8800940438871474, "grad_norm": 3.3841421604156494, "learning_rate": 1.8192520026027699e-06, "loss": 1.319, "step": 36750 }, { "epoch": 2.8801724137931033, "grad_norm": 4.470633506774902, "learning_rate": 1.8190140870527244e-06, "loss": 1.7203, "step": 36751 }, { "epoch": 2.8802507836990596, "grad_norm": 3.2569546699523926, "learning_rate": 1.8187761836015138e-06, "loss": 1.323, "step": 36752 }, { "epoch": 2.8803291536050155, "grad_norm": 2.888399124145508, "learning_rate": 1.8185382922500444e-06, "loss": 1.5856, "step": 36753 }, { "epoch": 2.880407523510972, "grad_norm": 3.2603886127471924, "learning_rate": 1.8183004129992198e-06, "loss": 1.3661, "step": 36754 }, { "epoch": 2.8804858934169277, "grad_norm": 3.0292370319366455, "learning_rate": 1.8180625458499495e-06, "loss": 1.2828, "step": 36755 }, { "epoch": 2.880564263322884, "grad_norm": 7.418952941894531, "learning_rate": 1.8178246908031333e-06, "loss": 1.2616, "step": 36756 }, { "epoch": 2.88064263322884, "grad_norm": 4.966981410980225, "learning_rate": 1.8175868478596775e-06, "loss": 1.185, "step": 36757 }, { "epoch": 2.8807210031347963, "grad_norm": 3.353532075881958, "learning_rate": 1.8173490170204865e-06, "loss": 1.769, "step": 36758 }, { "epoch": 2.8807993730407526, "grad_norm": 6.778845310211182, "learning_rate": 1.8171111982864654e-06, "loss": 1.5178, "step": 36759 }, { "epoch": 2.8808777429467085, "grad_norm": 3.970104217529297, "learning_rate": 1.816873391658518e-06, "loss": 1.3498, "step": 36760 }, { "epoch": 2.8809561128526644, "grad_norm": 5.212502956390381, "learning_rate": 1.8166355971375498e-06, "loss": 1.2314, "step": 36761 }, { "epoch": 2.8810344827586207, "grad_norm": 4.9060516357421875, "learning_rate": 1.816397814724466e-06, "loss": 1.9074, "step": 36762 }, { "epoch": 2.881112852664577, "grad_norm": 15.46206283569336, "learning_rate": 1.8161600444201666e-06, "loss": 0.5554, "step": 36763 }, { "epoch": 2.881191222570533, "grad_norm": 3.740516424179077, "learning_rate": 1.815922286225561e-06, "loss": 1.1193, "step": 36764 }, { "epoch": 2.881269592476489, "grad_norm": 3.2345685958862305, "learning_rate": 1.8156845401415507e-06, "loss": 1.346, "step": 36765 }, { "epoch": 2.881347962382445, "grad_norm": 8.705702781677246, "learning_rate": 1.8154468061690434e-06, "loss": 1.6008, "step": 36766 }, { "epoch": 2.8814263322884015, "grad_norm": 2.591820478439331, "learning_rate": 1.8152090843089387e-06, "loss": 1.2943, "step": 36767 }, { "epoch": 2.8815047021943574, "grad_norm": 4.372646331787109, "learning_rate": 1.8149713745621412e-06, "loss": 1.4991, "step": 36768 }, { "epoch": 2.8815830721003133, "grad_norm": 3.33536434173584, "learning_rate": 1.8147336769295582e-06, "loss": 1.4583, "step": 36769 }, { "epoch": 2.8816614420062696, "grad_norm": 2.400437116622925, "learning_rate": 1.8144959914120941e-06, "loss": 1.4192, "step": 36770 }, { "epoch": 2.881739811912226, "grad_norm": 2.2400879859924316, "learning_rate": 1.8142583180106493e-06, "loss": 1.1568, "step": 36771 }, { "epoch": 2.881818181818182, "grad_norm": 2.6563034057617188, "learning_rate": 1.8140206567261299e-06, "loss": 1.3482, "step": 36772 }, { "epoch": 2.8818965517241377, "grad_norm": 1.9813789129257202, "learning_rate": 1.8137830075594392e-06, "loss": 1.3607, "step": 36773 }, { "epoch": 2.881974921630094, "grad_norm": 3.002361536026001, "learning_rate": 1.8135453705114814e-06, "loss": 1.3043, "step": 36774 }, { "epoch": 2.8820532915360504, "grad_norm": 3.9352855682373047, "learning_rate": 1.8133077455831605e-06, "loss": 1.421, "step": 36775 }, { "epoch": 2.8821316614420063, "grad_norm": 9.931852340698242, "learning_rate": 1.8130701327753796e-06, "loss": 0.6857, "step": 36776 }, { "epoch": 2.882210031347962, "grad_norm": 3.4282784461975098, "learning_rate": 1.8128325320890438e-06, "loss": 1.5148, "step": 36777 }, { "epoch": 2.8822884012539185, "grad_norm": 7.9617085456848145, "learning_rate": 1.8125949435250556e-06, "loss": 1.4551, "step": 36778 }, { "epoch": 2.882366771159875, "grad_norm": 3.502096652984619, "learning_rate": 1.8123573670843192e-06, "loss": 0.9861, "step": 36779 }, { "epoch": 2.8824451410658307, "grad_norm": 2.614377498626709, "learning_rate": 1.8121198027677378e-06, "loss": 0.9862, "step": 36780 }, { "epoch": 2.8825235109717866, "grad_norm": 9.514739036560059, "learning_rate": 1.811882250576215e-06, "loss": 1.5329, "step": 36781 }, { "epoch": 2.882601880877743, "grad_norm": 4.134336471557617, "learning_rate": 1.8116447105106555e-06, "loss": 1.0803, "step": 36782 }, { "epoch": 2.8826802507836993, "grad_norm": 4.02630615234375, "learning_rate": 1.8114071825719615e-06, "loss": 1.2963, "step": 36783 }, { "epoch": 2.882758620689655, "grad_norm": 3.3066351413726807, "learning_rate": 1.8111696667610369e-06, "loss": 1.4677, "step": 36784 }, { "epoch": 2.882836990595611, "grad_norm": 7.523190498352051, "learning_rate": 1.8109321630787863e-06, "loss": 1.2332, "step": 36785 }, { "epoch": 2.8829153605015674, "grad_norm": 2.367612600326538, "learning_rate": 1.8106946715261086e-06, "loss": 1.4621, "step": 36786 }, { "epoch": 2.8829937304075237, "grad_norm": 2.645948648452759, "learning_rate": 1.810457192103912e-06, "loss": 1.2897, "step": 36787 }, { "epoch": 2.8830721003134796, "grad_norm": 4.890191555023193, "learning_rate": 1.8102197248130998e-06, "loss": 1.2967, "step": 36788 }, { "epoch": 2.8831504702194355, "grad_norm": 3.7207958698272705, "learning_rate": 1.8099822696545716e-06, "loss": 1.5664, "step": 36789 }, { "epoch": 2.883228840125392, "grad_norm": 10.182168006896973, "learning_rate": 1.8097448266292322e-06, "loss": 0.9639, "step": 36790 }, { "epoch": 2.883307210031348, "grad_norm": 3.771024465560913, "learning_rate": 1.8095073957379832e-06, "loss": 1.3025, "step": 36791 }, { "epoch": 2.883385579937304, "grad_norm": 2.501478433609009, "learning_rate": 1.8092699769817323e-06, "loss": 0.8336, "step": 36792 }, { "epoch": 2.88346394984326, "grad_norm": 11.267271995544434, "learning_rate": 1.8090325703613787e-06, "loss": 1.4773, "step": 36793 }, { "epoch": 2.8835423197492163, "grad_norm": 2.540382146835327, "learning_rate": 1.8087951758778255e-06, "loss": 0.9153, "step": 36794 }, { "epoch": 2.8836206896551726, "grad_norm": 7.369420051574707, "learning_rate": 1.808557793531977e-06, "loss": 1.1515, "step": 36795 }, { "epoch": 2.8836990595611285, "grad_norm": 3.360452175140381, "learning_rate": 1.8083204233247348e-06, "loss": 1.1638, "step": 36796 }, { "epoch": 2.8837774294670844, "grad_norm": 4.03178071975708, "learning_rate": 1.8080830652570031e-06, "loss": 1.1591, "step": 36797 }, { "epoch": 2.8838557993730407, "grad_norm": 5.044626235961914, "learning_rate": 1.8078457193296832e-06, "loss": 1.1759, "step": 36798 }, { "epoch": 2.883934169278997, "grad_norm": 4.384092807769775, "learning_rate": 1.8076083855436805e-06, "loss": 1.2084, "step": 36799 }, { "epoch": 2.884012539184953, "grad_norm": 2.5165867805480957, "learning_rate": 1.8073710638998927e-06, "loss": 1.1125, "step": 36800 }, { "epoch": 2.884090909090909, "grad_norm": 2.7527880668640137, "learning_rate": 1.8071337543992268e-06, "loss": 1.3241, "step": 36801 }, { "epoch": 2.884169278996865, "grad_norm": 2.499732255935669, "learning_rate": 1.8068964570425846e-06, "loss": 1.1179, "step": 36802 }, { "epoch": 2.8842476489028215, "grad_norm": 3.172650098800659, "learning_rate": 1.8066591718308697e-06, "loss": 1.0849, "step": 36803 }, { "epoch": 2.8843260188087774, "grad_norm": 2.3985869884490967, "learning_rate": 1.8064218987649813e-06, "loss": 1.4993, "step": 36804 }, { "epoch": 2.8844043887147337, "grad_norm": 3.18273663520813, "learning_rate": 1.806184637845822e-06, "loss": 1.6811, "step": 36805 }, { "epoch": 2.8844827586206896, "grad_norm": 5.308200836181641, "learning_rate": 1.8059473890742973e-06, "loss": 1.005, "step": 36806 }, { "epoch": 2.884561128526646, "grad_norm": 3.3953118324279785, "learning_rate": 1.8057101524513105e-06, "loss": 1.0315, "step": 36807 }, { "epoch": 2.884639498432602, "grad_norm": 4.450319766998291, "learning_rate": 1.8054729279777594e-06, "loss": 1.2065, "step": 36808 }, { "epoch": 2.884717868338558, "grad_norm": 3.6022653579711914, "learning_rate": 1.8052357156545487e-06, "loss": 1.3197, "step": 36809 }, { "epoch": 2.884796238244514, "grad_norm": 3.410717010498047, "learning_rate": 1.804998515482579e-06, "loss": 0.7283, "step": 36810 }, { "epoch": 2.8848746081504704, "grad_norm": 2.8549418449401855, "learning_rate": 1.8047613274627568e-06, "loss": 1.3228, "step": 36811 }, { "epoch": 2.8849529780564263, "grad_norm": 2.303724765777588, "learning_rate": 1.8045241515959795e-06, "loss": 1.2175, "step": 36812 }, { "epoch": 2.8850313479623826, "grad_norm": 6.758631706237793, "learning_rate": 1.8042869878831514e-06, "loss": 1.4763, "step": 36813 }, { "epoch": 2.8851097178683385, "grad_norm": 4.864902019500732, "learning_rate": 1.804049836325174e-06, "loss": 1.1142, "step": 36814 }, { "epoch": 2.885188087774295, "grad_norm": 3.0812931060791016, "learning_rate": 1.8038126969229496e-06, "loss": 1.2624, "step": 36815 }, { "epoch": 2.8852664576802507, "grad_norm": 5.62933349609375, "learning_rate": 1.8035755696773798e-06, "loss": 1.2156, "step": 36816 }, { "epoch": 2.885344827586207, "grad_norm": 3.2106659412384033, "learning_rate": 1.8033384545893667e-06, "loss": 1.0737, "step": 36817 }, { "epoch": 2.885423197492163, "grad_norm": 2.4179558753967285, "learning_rate": 1.8031013516598144e-06, "loss": 1.3823, "step": 36818 }, { "epoch": 2.8855015673981192, "grad_norm": 2.0901012420654297, "learning_rate": 1.802864260889619e-06, "loss": 1.2702, "step": 36819 }, { "epoch": 2.885579937304075, "grad_norm": 3.5011072158813477, "learning_rate": 1.802627182279687e-06, "loss": 1.6656, "step": 36820 }, { "epoch": 2.8856583072100315, "grad_norm": 5.0430707931518555, "learning_rate": 1.8023901158309193e-06, "loss": 1.3464, "step": 36821 }, { "epoch": 2.8857366771159874, "grad_norm": 2.8868439197540283, "learning_rate": 1.802153061544219e-06, "loss": 1.3846, "step": 36822 }, { "epoch": 2.8858150470219437, "grad_norm": 3.3252832889556885, "learning_rate": 1.8019160194204843e-06, "loss": 1.7764, "step": 36823 }, { "epoch": 2.8858934169278996, "grad_norm": 5.7944655418396, "learning_rate": 1.8016789894606163e-06, "loss": 1.3795, "step": 36824 }, { "epoch": 2.885971786833856, "grad_norm": 2.5064189434051514, "learning_rate": 1.8014419716655207e-06, "loss": 1.5065, "step": 36825 }, { "epoch": 2.886050156739812, "grad_norm": 3.2673842906951904, "learning_rate": 1.8012049660360987e-06, "loss": 1.4458, "step": 36826 }, { "epoch": 2.886128526645768, "grad_norm": 4.030299663543701, "learning_rate": 1.8009679725732482e-06, "loss": 1.6048, "step": 36827 }, { "epoch": 2.886206896551724, "grad_norm": 3.525214672088623, "learning_rate": 1.8007309912778726e-06, "loss": 0.9936, "step": 36828 }, { "epoch": 2.8862852664576804, "grad_norm": 2.674746513366699, "learning_rate": 1.8004940221508727e-06, "loss": 1.1184, "step": 36829 }, { "epoch": 2.8863636363636362, "grad_norm": 3.1520144939422607, "learning_rate": 1.80025706519315e-06, "loss": 1.1689, "step": 36830 }, { "epoch": 2.8864420062695926, "grad_norm": 7.650313854217529, "learning_rate": 1.8000201204056062e-06, "loss": 1.862, "step": 36831 }, { "epoch": 2.8865203761755485, "grad_norm": 8.470247268676758, "learning_rate": 1.7997831877891425e-06, "loss": 1.0868, "step": 36832 }, { "epoch": 2.886598746081505, "grad_norm": 4.401637077331543, "learning_rate": 1.7995462673446595e-06, "loss": 1.471, "step": 36833 }, { "epoch": 2.8866771159874607, "grad_norm": 2.62034010887146, "learning_rate": 1.7993093590730587e-06, "loss": 1.3905, "step": 36834 }, { "epoch": 2.886755485893417, "grad_norm": 3.340894937515259, "learning_rate": 1.7990724629752415e-06, "loss": 1.5669, "step": 36835 }, { "epoch": 2.886833855799373, "grad_norm": 2.622515916824341, "learning_rate": 1.7988355790521078e-06, "loss": 1.1623, "step": 36836 }, { "epoch": 2.8869122257053292, "grad_norm": 2.7506682872772217, "learning_rate": 1.7985987073045618e-06, "loss": 1.1912, "step": 36837 }, { "epoch": 2.886990595611285, "grad_norm": 2.602816104888916, "learning_rate": 1.798361847733498e-06, "loss": 1.1102, "step": 36838 }, { "epoch": 2.8870689655172415, "grad_norm": 3.643401861190796, "learning_rate": 1.798125000339823e-06, "loss": 1.1498, "step": 36839 }, { "epoch": 2.8871473354231973, "grad_norm": 3.2251715660095215, "learning_rate": 1.7978881651244363e-06, "loss": 0.9325, "step": 36840 }, { "epoch": 2.8872257053291537, "grad_norm": 6.455480098724365, "learning_rate": 1.7976513420882397e-06, "loss": 1.0572, "step": 36841 }, { "epoch": 2.8873040752351096, "grad_norm": 3.6364188194274902, "learning_rate": 1.797414531232131e-06, "loss": 0.8006, "step": 36842 }, { "epoch": 2.887382445141066, "grad_norm": 2.313945770263672, "learning_rate": 1.797177732557011e-06, "loss": 1.5594, "step": 36843 }, { "epoch": 2.887460815047022, "grad_norm": 3.819047212600708, "learning_rate": 1.796940946063785e-06, "loss": 1.6606, "step": 36844 }, { "epoch": 2.887539184952978, "grad_norm": 3.541245937347412, "learning_rate": 1.7967041717533484e-06, "loss": 1.0638, "step": 36845 }, { "epoch": 2.887617554858934, "grad_norm": 5.27311897277832, "learning_rate": 1.7964674096266044e-06, "loss": 1.194, "step": 36846 }, { "epoch": 2.8876959247648903, "grad_norm": 4.780573844909668, "learning_rate": 1.7962306596844526e-06, "loss": 1.6505, "step": 36847 }, { "epoch": 2.8877742946708462, "grad_norm": 4.674326419830322, "learning_rate": 1.7959939219277938e-06, "loss": 1.5751, "step": 36848 }, { "epoch": 2.8878526645768026, "grad_norm": 15.910138130187988, "learning_rate": 1.7957571963575288e-06, "loss": 1.8404, "step": 36849 }, { "epoch": 2.887931034482759, "grad_norm": 3.3559160232543945, "learning_rate": 1.7955204829745571e-06, "loss": 1.2397, "step": 36850 }, { "epoch": 2.888009404388715, "grad_norm": 2.3003268241882324, "learning_rate": 1.7952837817797797e-06, "loss": 1.7797, "step": 36851 }, { "epoch": 2.8880877742946707, "grad_norm": 2.806964159011841, "learning_rate": 1.795047092774097e-06, "loss": 1.0726, "step": 36852 }, { "epoch": 2.888166144200627, "grad_norm": 4.29008150100708, "learning_rate": 1.794810415958409e-06, "loss": 1.189, "step": 36853 }, { "epoch": 2.8882445141065833, "grad_norm": 6.755007743835449, "learning_rate": 1.7945737513336153e-06, "loss": 1.3002, "step": 36854 }, { "epoch": 2.8883228840125392, "grad_norm": 2.8105292320251465, "learning_rate": 1.7943370989006192e-06, "loss": 1.0412, "step": 36855 }, { "epoch": 2.888401253918495, "grad_norm": 4.5240678787231445, "learning_rate": 1.794100458660316e-06, "loss": 1.7985, "step": 36856 }, { "epoch": 2.8884796238244514, "grad_norm": 2.8826067447662354, "learning_rate": 1.7938638306136063e-06, "loss": 1.251, "step": 36857 }, { "epoch": 2.888557993730408, "grad_norm": 2.126934766769409, "learning_rate": 1.7936272147613936e-06, "loss": 1.4015, "step": 36858 }, { "epoch": 2.8886363636363637, "grad_norm": 3.065377950668335, "learning_rate": 1.7933906111045779e-06, "loss": 1.4588, "step": 36859 }, { "epoch": 2.8887147335423196, "grad_norm": 4.374575138092041, "learning_rate": 1.7931540196440555e-06, "loss": 1.3399, "step": 36860 }, { "epoch": 2.888793103448276, "grad_norm": 3.0675885677337646, "learning_rate": 1.7929174403807281e-06, "loss": 1.5262, "step": 36861 }, { "epoch": 2.8888714733542322, "grad_norm": 3.80812931060791, "learning_rate": 1.7926808733154937e-06, "loss": 0.8677, "step": 36862 }, { "epoch": 2.888949843260188, "grad_norm": 2.412029266357422, "learning_rate": 1.7924443184492574e-06, "loss": 1.2766, "step": 36863 }, { "epoch": 2.889028213166144, "grad_norm": 2.4204330444335938, "learning_rate": 1.7922077757829137e-06, "loss": 1.2193, "step": 36864 }, { "epoch": 2.8891065830721003, "grad_norm": 4.166790962219238, "learning_rate": 1.7919712453173637e-06, "loss": 1.238, "step": 36865 }, { "epoch": 2.8891849529780567, "grad_norm": 12.061558723449707, "learning_rate": 1.791734727053508e-06, "loss": 0.9911, "step": 36866 }, { "epoch": 2.8892633228840126, "grad_norm": 3.2984724044799805, "learning_rate": 1.791498220992245e-06, "loss": 1.0008, "step": 36867 }, { "epoch": 2.8893416927899684, "grad_norm": 4.566408634185791, "learning_rate": 1.7912617271344746e-06, "loss": 1.1885, "step": 36868 }, { "epoch": 2.8894200626959248, "grad_norm": 2.5034868717193604, "learning_rate": 1.7910252454810972e-06, "loss": 1.2312, "step": 36869 }, { "epoch": 2.889498432601881, "grad_norm": 1.967262625694275, "learning_rate": 1.7907887760330106e-06, "loss": 1.4474, "step": 36870 }, { "epoch": 2.889576802507837, "grad_norm": 2.5673558712005615, "learning_rate": 1.7905523187911162e-06, "loss": 1.1883, "step": 36871 }, { "epoch": 2.889655172413793, "grad_norm": 2.9543144702911377, "learning_rate": 1.7903158737563114e-06, "loss": 1.1081, "step": 36872 }, { "epoch": 2.889733542319749, "grad_norm": 3.831881284713745, "learning_rate": 1.7900794409294969e-06, "loss": 1.7538, "step": 36873 }, { "epoch": 2.8898119122257055, "grad_norm": 2.512594699859619, "learning_rate": 1.789843020311573e-06, "loss": 1.4507, "step": 36874 }, { "epoch": 2.8898902821316614, "grad_norm": 2.790449619293213, "learning_rate": 1.7896066119034338e-06, "loss": 1.5822, "step": 36875 }, { "epoch": 2.8899686520376173, "grad_norm": 2.273822069168091, "learning_rate": 1.7893702157059844e-06, "loss": 1.4386, "step": 36876 }, { "epoch": 2.8900470219435737, "grad_norm": 3.1578314304351807, "learning_rate": 1.7891338317201213e-06, "loss": 1.2839, "step": 36877 }, { "epoch": 2.89012539184953, "grad_norm": 5.757509231567383, "learning_rate": 1.7888974599467452e-06, "loss": 1.476, "step": 36878 }, { "epoch": 2.890203761755486, "grad_norm": 4.908112525939941, "learning_rate": 1.7886611003867528e-06, "loss": 1.2046, "step": 36879 }, { "epoch": 2.8902821316614418, "grad_norm": 2.8181300163269043, "learning_rate": 1.7884247530410436e-06, "loss": 1.5687, "step": 36880 }, { "epoch": 2.890360501567398, "grad_norm": 3.4280829429626465, "learning_rate": 1.7881884179105158e-06, "loss": 1.2529, "step": 36881 }, { "epoch": 2.8904388714733544, "grad_norm": 5.222691059112549, "learning_rate": 1.7879520949960728e-06, "loss": 1.2851, "step": 36882 }, { "epoch": 2.8905172413793103, "grad_norm": 2.154790163040161, "learning_rate": 1.787715784298608e-06, "loss": 1.085, "step": 36883 }, { "epoch": 2.890595611285266, "grad_norm": 4.385047912597656, "learning_rate": 1.787479485819023e-06, "loss": 1.0827, "step": 36884 }, { "epoch": 2.8906739811912225, "grad_norm": 4.836544513702393, "learning_rate": 1.7872431995582162e-06, "loss": 0.6575, "step": 36885 }, { "epoch": 2.890752351097179, "grad_norm": 3.1141700744628906, "learning_rate": 1.7870069255170857e-06, "loss": 1.5217, "step": 36886 }, { "epoch": 2.8908307210031348, "grad_norm": 2.709298610687256, "learning_rate": 1.7867706636965304e-06, "loss": 1.2999, "step": 36887 }, { "epoch": 2.8909090909090907, "grad_norm": 5.704329490661621, "learning_rate": 1.7865344140974495e-06, "loss": 1.3629, "step": 36888 }, { "epoch": 2.890987460815047, "grad_norm": 6.037484645843506, "learning_rate": 1.786298176720741e-06, "loss": 1.7669, "step": 36889 }, { "epoch": 2.8910658307210033, "grad_norm": 2.4798007011413574, "learning_rate": 1.7860619515673034e-06, "loss": 1.2644, "step": 36890 }, { "epoch": 2.891144200626959, "grad_norm": 1.9266595840454102, "learning_rate": 1.7858257386380357e-06, "loss": 1.1127, "step": 36891 }, { "epoch": 2.891222570532915, "grad_norm": 4.02025032043457, "learning_rate": 1.7855895379338356e-06, "loss": 1.6377, "step": 36892 }, { "epoch": 2.8913009404388714, "grad_norm": 4.492194175720215, "learning_rate": 1.7853533494556035e-06, "loss": 1.2899, "step": 36893 }, { "epoch": 2.8913793103448278, "grad_norm": 6.137948989868164, "learning_rate": 1.7851171732042332e-06, "loss": 1.0637, "step": 36894 }, { "epoch": 2.8914576802507836, "grad_norm": 3.5949645042419434, "learning_rate": 1.7848810091806273e-06, "loss": 1.7211, "step": 36895 }, { "epoch": 2.8915360501567395, "grad_norm": 2.027203321456909, "learning_rate": 1.784644857385683e-06, "loss": 1.2882, "step": 36896 }, { "epoch": 2.891614420062696, "grad_norm": 3.04451322555542, "learning_rate": 1.7844087178202995e-06, "loss": 1.4583, "step": 36897 }, { "epoch": 2.891692789968652, "grad_norm": 5.1199541091918945, "learning_rate": 1.7841725904853718e-06, "loss": 1.1241, "step": 36898 }, { "epoch": 2.891771159874608, "grad_norm": 3.624300956726074, "learning_rate": 1.7839364753817985e-06, "loss": 1.3232, "step": 36899 }, { "epoch": 2.8918495297805644, "grad_norm": 4.824703216552734, "learning_rate": 1.7837003725104818e-06, "loss": 0.9713, "step": 36900 }, { "epoch": 2.8919278996865203, "grad_norm": 3.3705880641937256, "learning_rate": 1.7834642818723157e-06, "loss": 1.4131, "step": 36901 }, { "epoch": 2.8920062695924766, "grad_norm": 2.563396453857422, "learning_rate": 1.7832282034681992e-06, "loss": 1.5797, "step": 36902 }, { "epoch": 2.8920846394984325, "grad_norm": 6.285439491271973, "learning_rate": 1.7829921372990305e-06, "loss": 1.8737, "step": 36903 }, { "epoch": 2.892163009404389, "grad_norm": 4.3361358642578125, "learning_rate": 1.782756083365707e-06, "loss": 1.7277, "step": 36904 }, { "epoch": 2.8922413793103448, "grad_norm": 2.054673194885254, "learning_rate": 1.7825200416691275e-06, "loss": 1.2441, "step": 36905 }, { "epoch": 2.892319749216301, "grad_norm": 2.3130009174346924, "learning_rate": 1.7822840122101886e-06, "loss": 0.9573, "step": 36906 }, { "epoch": 2.892398119122257, "grad_norm": 8.332545280456543, "learning_rate": 1.7820479949897884e-06, "loss": 1.4297, "step": 36907 }, { "epoch": 2.8924764890282133, "grad_norm": 4.567744255065918, "learning_rate": 1.7818119900088254e-06, "loss": 1.5518, "step": 36908 }, { "epoch": 2.892554858934169, "grad_norm": 2.8046469688415527, "learning_rate": 1.7815759972681961e-06, "loss": 1.3316, "step": 36909 }, { "epoch": 2.8926332288401255, "grad_norm": 2.64011812210083, "learning_rate": 1.781340016768799e-06, "loss": 1.1738, "step": 36910 }, { "epoch": 2.8927115987460814, "grad_norm": 3.8753960132598877, "learning_rate": 1.7811040485115328e-06, "loss": 1.2334, "step": 36911 }, { "epoch": 2.8927899686520377, "grad_norm": 6.597634792327881, "learning_rate": 1.780868092497292e-06, "loss": 1.4401, "step": 36912 }, { "epoch": 2.8928683385579936, "grad_norm": 3.0323100090026855, "learning_rate": 1.7806321487269735e-06, "loss": 1.2089, "step": 36913 }, { "epoch": 2.89294670846395, "grad_norm": 3.5283031463623047, "learning_rate": 1.7803962172014787e-06, "loss": 1.4026, "step": 36914 }, { "epoch": 2.893025078369906, "grad_norm": 4.541159629821777, "learning_rate": 1.7801602979217048e-06, "loss": 1.1923, "step": 36915 }, { "epoch": 2.893103448275862, "grad_norm": 2.4364428520202637, "learning_rate": 1.779924390888545e-06, "loss": 1.2789, "step": 36916 }, { "epoch": 2.893181818181818, "grad_norm": 3.68414044380188, "learning_rate": 1.7796884961028993e-06, "loss": 1.2005, "step": 36917 }, { "epoch": 2.8932601880877744, "grad_norm": 7.980481147766113, "learning_rate": 1.7794526135656626e-06, "loss": 0.3955, "step": 36918 }, { "epoch": 2.8933385579937303, "grad_norm": 4.851686477661133, "learning_rate": 1.779216743277738e-06, "loss": 1.4639, "step": 36919 }, { "epoch": 2.8934169278996866, "grad_norm": 6.6540422439575195, "learning_rate": 1.7789808852400164e-06, "loss": 1.3837, "step": 36920 }, { "epoch": 2.8934952978056425, "grad_norm": 2.006510019302368, "learning_rate": 1.7787450394533966e-06, "loss": 1.4055, "step": 36921 }, { "epoch": 2.893573667711599, "grad_norm": 2.5316081047058105, "learning_rate": 1.7785092059187769e-06, "loss": 1.3867, "step": 36922 }, { "epoch": 2.8936520376175547, "grad_norm": 2.6392173767089844, "learning_rate": 1.7782733846370537e-06, "loss": 1.496, "step": 36923 }, { "epoch": 2.893730407523511, "grad_norm": 2.778728485107422, "learning_rate": 1.7780375756091234e-06, "loss": 1.3194, "step": 36924 }, { "epoch": 2.893808777429467, "grad_norm": 3.3053271770477295, "learning_rate": 1.777801778835883e-06, "loss": 1.2712, "step": 36925 }, { "epoch": 2.8938871473354233, "grad_norm": 2.5479111671447754, "learning_rate": 1.7775659943182316e-06, "loss": 1.16, "step": 36926 }, { "epoch": 2.893965517241379, "grad_norm": 2.8679447174072266, "learning_rate": 1.7773302220570604e-06, "loss": 1.4426, "step": 36927 }, { "epoch": 2.8940438871473355, "grad_norm": 4.778721332550049, "learning_rate": 1.7770944620532715e-06, "loss": 1.8278, "step": 36928 }, { "epoch": 2.8941222570532914, "grad_norm": 12.525684356689453, "learning_rate": 1.7768587143077604e-06, "loss": 1.0373, "step": 36929 }, { "epoch": 2.8942006269592477, "grad_norm": 3.166313409805298, "learning_rate": 1.776622978821424e-06, "loss": 1.6306, "step": 36930 }, { "epoch": 2.8942789968652036, "grad_norm": 3.300856590270996, "learning_rate": 1.7763872555951572e-06, "loss": 1.2541, "step": 36931 }, { "epoch": 2.89435736677116, "grad_norm": 4.22599458694458, "learning_rate": 1.7761515446298554e-06, "loss": 1.2962, "step": 36932 }, { "epoch": 2.894435736677116, "grad_norm": 4.804427146911621, "learning_rate": 1.7759158459264192e-06, "loss": 1.7872, "step": 36933 }, { "epoch": 2.894514106583072, "grad_norm": 2.20078182220459, "learning_rate": 1.7756801594857447e-06, "loss": 1.1309, "step": 36934 }, { "epoch": 2.894592476489028, "grad_norm": 4.483153343200684, "learning_rate": 1.7754444853087254e-06, "loss": 1.2563, "step": 36935 }, { "epoch": 2.8946708463949844, "grad_norm": 5.458908557891846, "learning_rate": 1.7752088233962588e-06, "loss": 1.2348, "step": 36936 }, { "epoch": 2.8947492163009403, "grad_norm": 2.953558921813965, "learning_rate": 1.7749731737492398e-06, "loss": 1.5299, "step": 36937 }, { "epoch": 2.8948275862068966, "grad_norm": 2.7505404949188232, "learning_rate": 1.7747375363685698e-06, "loss": 1.2566, "step": 36938 }, { "epoch": 2.8949059561128525, "grad_norm": 8.659849166870117, "learning_rate": 1.7745019112551398e-06, "loss": 0.8124, "step": 36939 }, { "epoch": 2.894984326018809, "grad_norm": 10.477851867675781, "learning_rate": 1.774266298409848e-06, "loss": 0.9874, "step": 36940 }, { "epoch": 2.8950626959247647, "grad_norm": 4.002865791320801, "learning_rate": 1.7740306978335903e-06, "loss": 1.678, "step": 36941 }, { "epoch": 2.895141065830721, "grad_norm": 3.804882764816284, "learning_rate": 1.7737951095272626e-06, "loss": 1.2236, "step": 36942 }, { "epoch": 2.895219435736677, "grad_norm": 3.5321717262268066, "learning_rate": 1.7735595334917616e-06, "loss": 1.3297, "step": 36943 }, { "epoch": 2.8952978056426333, "grad_norm": 5.108442783355713, "learning_rate": 1.7733239697279825e-06, "loss": 1.1161, "step": 36944 }, { "epoch": 2.8953761755485896, "grad_norm": 15.414985656738281, "learning_rate": 1.7730884182368236e-06, "loss": 0.7677, "step": 36945 }, { "epoch": 2.8954545454545455, "grad_norm": 2.5098297595977783, "learning_rate": 1.7728528790191756e-06, "loss": 1.5037, "step": 36946 }, { "epoch": 2.8955329153605014, "grad_norm": 2.5424482822418213, "learning_rate": 1.7726173520759393e-06, "loss": 1.0256, "step": 36947 }, { "epoch": 2.8956112852664577, "grad_norm": 2.459413766860962, "learning_rate": 1.7723818374080088e-06, "loss": 1.316, "step": 36948 }, { "epoch": 2.895689655172414, "grad_norm": 4.11761474609375, "learning_rate": 1.772146335016282e-06, "loss": 1.4643, "step": 36949 }, { "epoch": 2.89576802507837, "grad_norm": 2.710393190383911, "learning_rate": 1.7719108449016504e-06, "loss": 1.0202, "step": 36950 }, { "epoch": 2.895846394984326, "grad_norm": 3.103865146636963, "learning_rate": 1.7716753670650104e-06, "loss": 1.324, "step": 36951 }, { "epoch": 2.895924764890282, "grad_norm": 5.202303409576416, "learning_rate": 1.7714399015072631e-06, "loss": 1.6754, "step": 36952 }, { "epoch": 2.8960031347962385, "grad_norm": 2.7045013904571533, "learning_rate": 1.7712044482292979e-06, "loss": 1.3559, "step": 36953 }, { "epoch": 2.8960815047021944, "grad_norm": 4.699398994445801, "learning_rate": 1.7709690072320125e-06, "loss": 1.8821, "step": 36954 }, { "epoch": 2.8961598746081503, "grad_norm": 1.6754094362258911, "learning_rate": 1.7707335785163027e-06, "loss": 1.2759, "step": 36955 }, { "epoch": 2.8962382445141066, "grad_norm": 5.513200759887695, "learning_rate": 1.7704981620830636e-06, "loss": 1.3407, "step": 36956 }, { "epoch": 2.896316614420063, "grad_norm": 2.5349462032318115, "learning_rate": 1.7702627579331905e-06, "loss": 1.14, "step": 36957 }, { "epoch": 2.896394984326019, "grad_norm": 2.9423820972442627, "learning_rate": 1.7700273660675792e-06, "loss": 1.3884, "step": 36958 }, { "epoch": 2.8964733542319747, "grad_norm": 10.115976333618164, "learning_rate": 1.769791986487125e-06, "loss": 1.0087, "step": 36959 }, { "epoch": 2.896551724137931, "grad_norm": 2.2657206058502197, "learning_rate": 1.7695566191927227e-06, "loss": 1.4508, "step": 36960 }, { "epoch": 2.8966300940438874, "grad_norm": 3.887873649597168, "learning_rate": 1.7693212641852675e-06, "loss": 0.9867, "step": 36961 }, { "epoch": 2.8967084639498433, "grad_norm": 6.927462577819824, "learning_rate": 1.7690859214656553e-06, "loss": 1.5965, "step": 36962 }, { "epoch": 2.896786833855799, "grad_norm": 2.6174771785736084, "learning_rate": 1.7688505910347804e-06, "loss": 1.0824, "step": 36963 }, { "epoch": 2.8968652037617555, "grad_norm": 7.5348711013793945, "learning_rate": 1.7686152728935402e-06, "loss": 1.3115, "step": 36964 }, { "epoch": 2.896943573667712, "grad_norm": 2.2666327953338623, "learning_rate": 1.7683799670428242e-06, "loss": 1.1253, "step": 36965 }, { "epoch": 2.8970219435736677, "grad_norm": 4.471559524536133, "learning_rate": 1.7681446734835328e-06, "loss": 1.2696, "step": 36966 }, { "epoch": 2.8971003134796236, "grad_norm": 3.620370626449585, "learning_rate": 1.7679093922165603e-06, "loss": 1.2245, "step": 36967 }, { "epoch": 2.89717868338558, "grad_norm": 2.282665491104126, "learning_rate": 1.7676741232427991e-06, "loss": 1.5358, "step": 36968 }, { "epoch": 2.8972570532915363, "grad_norm": 3.5940990447998047, "learning_rate": 1.7674388665631454e-06, "loss": 1.3483, "step": 36969 }, { "epoch": 2.897335423197492, "grad_norm": 4.331545352935791, "learning_rate": 1.7672036221784917e-06, "loss": 1.2539, "step": 36970 }, { "epoch": 2.897413793103448, "grad_norm": 3.5079221725463867, "learning_rate": 1.7669683900897388e-06, "loss": 1.2305, "step": 36971 }, { "epoch": 2.8974921630094044, "grad_norm": 2.8688912391662598, "learning_rate": 1.7667331702977757e-06, "loss": 1.4713, "step": 36972 }, { "epoch": 2.8975705329153607, "grad_norm": 11.611522674560547, "learning_rate": 1.766497962803499e-06, "loss": 1.1944, "step": 36973 }, { "epoch": 2.8976489028213166, "grad_norm": 8.891216278076172, "learning_rate": 1.7662627676078032e-06, "loss": 1.8612, "step": 36974 }, { "epoch": 2.8977272727272725, "grad_norm": 4.28737211227417, "learning_rate": 1.7660275847115827e-06, "loss": 1.7302, "step": 36975 }, { "epoch": 2.897805642633229, "grad_norm": 6.240845203399658, "learning_rate": 1.7657924141157323e-06, "loss": 1.1553, "step": 36976 }, { "epoch": 2.897884012539185, "grad_norm": 2.921078681945801, "learning_rate": 1.7655572558211465e-06, "loss": 1.2916, "step": 36977 }, { "epoch": 2.897962382445141, "grad_norm": 12.434174537658691, "learning_rate": 1.7653221098287194e-06, "loss": 0.3664, "step": 36978 }, { "epoch": 2.898040752351097, "grad_norm": 3.560572862625122, "learning_rate": 1.7650869761393458e-06, "loss": 1.3563, "step": 36979 }, { "epoch": 2.8981191222570533, "grad_norm": 3.038273572921753, "learning_rate": 1.7648518547539195e-06, "loss": 1.4946, "step": 36980 }, { "epoch": 2.8981974921630096, "grad_norm": 2.6811587810516357, "learning_rate": 1.7646167456733354e-06, "loss": 1.2949, "step": 36981 }, { "epoch": 2.8982758620689655, "grad_norm": 2.0792160034179688, "learning_rate": 1.7643816488984888e-06, "loss": 1.5961, "step": 36982 }, { "epoch": 2.8983542319749214, "grad_norm": 2.6068973541259766, "learning_rate": 1.7641465644302691e-06, "loss": 1.325, "step": 36983 }, { "epoch": 2.8984326018808777, "grad_norm": 12.991153717041016, "learning_rate": 1.7639114922695756e-06, "loss": 1.1423, "step": 36984 }, { "epoch": 2.898510971786834, "grad_norm": 4.180874347686768, "learning_rate": 1.7636764324173005e-06, "loss": 1.5219, "step": 36985 }, { "epoch": 2.89858934169279, "grad_norm": 1.9769623279571533, "learning_rate": 1.7634413848743397e-06, "loss": 1.2696, "step": 36986 }, { "epoch": 2.898667711598746, "grad_norm": 2.4133858680725098, "learning_rate": 1.763206349641584e-06, "loss": 1.3335, "step": 36987 }, { "epoch": 2.898746081504702, "grad_norm": 3.0018534660339355, "learning_rate": 1.7629713267199272e-06, "loss": 1.4636, "step": 36988 }, { "epoch": 2.8988244514106585, "grad_norm": 4.9534196853637695, "learning_rate": 1.7627363161102662e-06, "loss": 1.1964, "step": 36989 }, { "epoch": 2.8989028213166144, "grad_norm": 4.729295253753662, "learning_rate": 1.7625013178134954e-06, "loss": 1.4295, "step": 36990 }, { "epoch": 2.8989811912225707, "grad_norm": 3.191343307495117, "learning_rate": 1.762266331830505e-06, "loss": 1.5128, "step": 36991 }, { "epoch": 2.8990595611285266, "grad_norm": 7.339513301849365, "learning_rate": 1.7620313581621906e-06, "loss": 1.9444, "step": 36992 }, { "epoch": 2.899137931034483, "grad_norm": 11.278679847717285, "learning_rate": 1.7617963968094453e-06, "loss": 1.2134, "step": 36993 }, { "epoch": 2.899216300940439, "grad_norm": 5.061222076416016, "learning_rate": 1.7615614477731634e-06, "loss": 1.1649, "step": 36994 }, { "epoch": 2.899294670846395, "grad_norm": 5.108108997344971, "learning_rate": 1.7613265110542388e-06, "loss": 1.3677, "step": 36995 }, { "epoch": 2.899373040752351, "grad_norm": 2.478041887283325, "learning_rate": 1.7610915866535644e-06, "loss": 1.2258, "step": 36996 }, { "epoch": 2.8994514106583074, "grad_norm": 7.199084758758545, "learning_rate": 1.760856674572034e-06, "loss": 1.6163, "step": 36997 }, { "epoch": 2.8995297805642632, "grad_norm": 3.0227982997894287, "learning_rate": 1.7606217748105408e-06, "loss": 0.9857, "step": 36998 }, { "epoch": 2.8996081504702196, "grad_norm": 4.0493364334106445, "learning_rate": 1.7603868873699786e-06, "loss": 1.3043, "step": 36999 }, { "epoch": 2.8996865203761755, "grad_norm": 5.205815315246582, "learning_rate": 1.760152012251241e-06, "loss": 1.1333, "step": 37000 }, { "epoch": 2.899764890282132, "grad_norm": 4.200166702270508, "learning_rate": 1.7599171494552232e-06, "loss": 1.4544, "step": 37001 }, { "epoch": 2.8998432601880877, "grad_norm": 2.812346935272217, "learning_rate": 1.7596822989828121e-06, "loss": 1.2601, "step": 37002 }, { "epoch": 2.899921630094044, "grad_norm": 5.890719890594482, "learning_rate": 1.759447460834907e-06, "loss": 1.3362, "step": 37003 }, { "epoch": 2.9, "grad_norm": 6.137092113494873, "learning_rate": 1.7592126350123995e-06, "loss": 1.9479, "step": 37004 }, { "epoch": 2.9000783699059562, "grad_norm": 4.470274925231934, "learning_rate": 1.7589778215161846e-06, "loss": 1.2185, "step": 37005 }, { "epoch": 2.900156739811912, "grad_norm": 2.4068996906280518, "learning_rate": 1.7587430203471511e-06, "loss": 0.7097, "step": 37006 }, { "epoch": 2.9002351097178685, "grad_norm": 2.6329843997955322, "learning_rate": 1.758508231506193e-06, "loss": 1.0789, "step": 37007 }, { "epoch": 2.9003134796238244, "grad_norm": 3.1212916374206543, "learning_rate": 1.7582734549942077e-06, "loss": 1.2067, "step": 37008 }, { "epoch": 2.9003918495297807, "grad_norm": 10.524539947509766, "learning_rate": 1.7580386908120838e-06, "loss": 1.0076, "step": 37009 }, { "epoch": 2.9004702194357366, "grad_norm": 5.681945323944092, "learning_rate": 1.7578039389607154e-06, "loss": 1.4145, "step": 37010 }, { "epoch": 2.900548589341693, "grad_norm": 2.873353958129883, "learning_rate": 1.7575691994409955e-06, "loss": 1.443, "step": 37011 }, { "epoch": 2.900626959247649, "grad_norm": 3.137275457382202, "learning_rate": 1.7573344722538172e-06, "loss": 1.2847, "step": 37012 }, { "epoch": 2.900705329153605, "grad_norm": 2.838641405105591, "learning_rate": 1.7570997574000725e-06, "loss": 1.6101, "step": 37013 }, { "epoch": 2.900783699059561, "grad_norm": 2.527390241622925, "learning_rate": 1.7568650548806548e-06, "loss": 1.3539, "step": 37014 }, { "epoch": 2.9008620689655173, "grad_norm": 2.6242549419403076, "learning_rate": 1.7566303646964572e-06, "loss": 1.2581, "step": 37015 }, { "epoch": 2.9009404388714732, "grad_norm": 3.5175180435180664, "learning_rate": 1.756395686848371e-06, "loss": 1.4964, "step": 37016 }, { "epoch": 2.9010188087774296, "grad_norm": 3.173208713531494, "learning_rate": 1.7561610213372904e-06, "loss": 1.7193, "step": 37017 }, { "epoch": 2.9010971786833855, "grad_norm": 2.6261837482452393, "learning_rate": 1.7559263681641064e-06, "loss": 1.1405, "step": 37018 }, { "epoch": 2.901175548589342, "grad_norm": 7.555102348327637, "learning_rate": 1.7556917273297126e-06, "loss": 1.5876, "step": 37019 }, { "epoch": 2.9012539184952977, "grad_norm": 3.0874550342559814, "learning_rate": 1.7554570988350027e-06, "loss": 1.3304, "step": 37020 }, { "epoch": 2.901332288401254, "grad_norm": 5.765296936035156, "learning_rate": 1.7552224826808645e-06, "loss": 1.5089, "step": 37021 }, { "epoch": 2.90141065830721, "grad_norm": 3.07597279548645, "learning_rate": 1.7549878788681946e-06, "loss": 1.5876, "step": 37022 }, { "epoch": 2.9014890282131662, "grad_norm": 4.561652183532715, "learning_rate": 1.7547532873978857e-06, "loss": 0.8979, "step": 37023 }, { "epoch": 2.901567398119122, "grad_norm": 10.4248685836792, "learning_rate": 1.7545187082708271e-06, "loss": 1.4909, "step": 37024 }, { "epoch": 2.9016457680250785, "grad_norm": 4.051590442657471, "learning_rate": 1.754284141487912e-06, "loss": 1.3399, "step": 37025 }, { "epoch": 2.9017241379310343, "grad_norm": 2.5991461277008057, "learning_rate": 1.7540495870500313e-06, "loss": 1.0819, "step": 37026 }, { "epoch": 2.9018025078369907, "grad_norm": 3.2176620960235596, "learning_rate": 1.7538150449580821e-06, "loss": 1.4478, "step": 37027 }, { "epoch": 2.9018808777429466, "grad_norm": 6.29940938949585, "learning_rate": 1.7535805152129515e-06, "loss": 1.616, "step": 37028 }, { "epoch": 2.901959247648903, "grad_norm": 2.361680030822754, "learning_rate": 1.7533459978155332e-06, "loss": 1.4277, "step": 37029 }, { "epoch": 2.902037617554859, "grad_norm": 3.5868935585021973, "learning_rate": 1.7531114927667192e-06, "loss": 1.4485, "step": 37030 }, { "epoch": 2.902115987460815, "grad_norm": 2.513489246368408, "learning_rate": 1.7528770000674011e-06, "loss": 1.2874, "step": 37031 }, { "epoch": 2.902194357366771, "grad_norm": 8.888319969177246, "learning_rate": 1.7526425197184715e-06, "loss": 1.527, "step": 37032 }, { "epoch": 2.9022727272727273, "grad_norm": 9.495932579040527, "learning_rate": 1.7524080517208215e-06, "loss": 0.9078, "step": 37033 }, { "epoch": 2.9023510971786832, "grad_norm": 9.046107292175293, "learning_rate": 1.7521735960753432e-06, "loss": 1.0603, "step": 37034 }, { "epoch": 2.9024294670846396, "grad_norm": 5.70452880859375, "learning_rate": 1.751939152782929e-06, "loss": 1.6332, "step": 37035 }, { "epoch": 2.9025078369905954, "grad_norm": 3.185177803039551, "learning_rate": 1.7517047218444687e-06, "loss": 1.2794, "step": 37036 }, { "epoch": 2.902586206896552, "grad_norm": 5.922824859619141, "learning_rate": 1.751470303260856e-06, "loss": 1.387, "step": 37037 }, { "epoch": 2.9026645768025077, "grad_norm": 2.8610892295837402, "learning_rate": 1.7512358970329834e-06, "loss": 1.287, "step": 37038 }, { "epoch": 2.902742946708464, "grad_norm": 13.253486633300781, "learning_rate": 1.751001503161739e-06, "loss": 1.4237, "step": 37039 }, { "epoch": 2.9028213166144203, "grad_norm": 2.49098539352417, "learning_rate": 1.7507671216480144e-06, "loss": 1.1238, "step": 37040 }, { "epoch": 2.902899686520376, "grad_norm": 2.8651862144470215, "learning_rate": 1.7505327524927047e-06, "loss": 1.4555, "step": 37041 }, { "epoch": 2.902978056426332, "grad_norm": 2.7459676265716553, "learning_rate": 1.7502983956967007e-06, "loss": 1.207, "step": 37042 }, { "epoch": 2.9030564263322884, "grad_norm": 3.5706963539123535, "learning_rate": 1.7500640512608907e-06, "loss": 1.1597, "step": 37043 }, { "epoch": 2.9031347962382448, "grad_norm": 2.039987802505493, "learning_rate": 1.749829719186168e-06, "loss": 1.0464, "step": 37044 }, { "epoch": 2.9032131661442007, "grad_norm": 4.101956844329834, "learning_rate": 1.7495953994734216e-06, "loss": 1.445, "step": 37045 }, { "epoch": 2.9032915360501566, "grad_norm": 2.7770934104919434, "learning_rate": 1.7493610921235487e-06, "loss": 1.4807, "step": 37046 }, { "epoch": 2.903369905956113, "grad_norm": 3.234294891357422, "learning_rate": 1.7491267971374343e-06, "loss": 1.228, "step": 37047 }, { "epoch": 2.903448275862069, "grad_norm": 12.055123329162598, "learning_rate": 1.7488925145159725e-06, "loss": 1.1941, "step": 37048 }, { "epoch": 2.903526645768025, "grad_norm": 2.4078729152679443, "learning_rate": 1.7486582442600531e-06, "loss": 1.2153, "step": 37049 }, { "epoch": 2.903605015673981, "grad_norm": 3.9361019134521484, "learning_rate": 1.7484239863705683e-06, "loss": 1.3943, "step": 37050 }, { "epoch": 2.9036833855799373, "grad_norm": 4.218883991241455, "learning_rate": 1.7481897408484078e-06, "loss": 1.8313, "step": 37051 }, { "epoch": 2.9037617554858937, "grad_norm": 6.278263568878174, "learning_rate": 1.7479555076944638e-06, "loss": 1.1356, "step": 37052 }, { "epoch": 2.9038401253918495, "grad_norm": 4.004465103149414, "learning_rate": 1.7477212869096282e-06, "loss": 1.1335, "step": 37053 }, { "epoch": 2.9039184952978054, "grad_norm": 3.327192783355713, "learning_rate": 1.7474870784947872e-06, "loss": 1.3149, "step": 37054 }, { "epoch": 2.9039968652037618, "grad_norm": 9.370384216308594, "learning_rate": 1.7472528824508366e-06, "loss": 1.3714, "step": 37055 }, { "epoch": 2.904075235109718, "grad_norm": 2.91634202003479, "learning_rate": 1.7470186987786647e-06, "loss": 1.3729, "step": 37056 }, { "epoch": 2.904153605015674, "grad_norm": 6.308865070343018, "learning_rate": 1.746784527479165e-06, "loss": 1.4223, "step": 37057 }, { "epoch": 2.90423197492163, "grad_norm": 2.405075788497925, "learning_rate": 1.7465503685532242e-06, "loss": 1.1535, "step": 37058 }, { "epoch": 2.904310344827586, "grad_norm": 4.289332866668701, "learning_rate": 1.7463162220017331e-06, "loss": 1.2909, "step": 37059 }, { "epoch": 2.9043887147335425, "grad_norm": 2.818077802658081, "learning_rate": 1.7460820878255853e-06, "loss": 1.3993, "step": 37060 }, { "epoch": 2.9044670846394984, "grad_norm": 5.142958164215088, "learning_rate": 1.7458479660256717e-06, "loss": 0.7144, "step": 37061 }, { "epoch": 2.9045454545454543, "grad_norm": 3.933725357055664, "learning_rate": 1.7456138566028797e-06, "loss": 1.8366, "step": 37062 }, { "epoch": 2.9046238244514107, "grad_norm": 4.0342183113098145, "learning_rate": 1.7453797595581007e-06, "loss": 1.5751, "step": 37063 }, { "epoch": 2.904702194357367, "grad_norm": 3.611337661743164, "learning_rate": 1.7451456748922252e-06, "loss": 1.1434, "step": 37064 }, { "epoch": 2.904780564263323, "grad_norm": 7.996330261230469, "learning_rate": 1.7449116026061441e-06, "loss": 1.6045, "step": 37065 }, { "epoch": 2.9048589341692788, "grad_norm": 3.607304573059082, "learning_rate": 1.7446775427007472e-06, "loss": 1.7745, "step": 37066 }, { "epoch": 2.904937304075235, "grad_norm": 5.052306652069092, "learning_rate": 1.7444434951769252e-06, "loss": 1.4537, "step": 37067 }, { "epoch": 2.9050156739811914, "grad_norm": 3.0199544429779053, "learning_rate": 1.7442094600355675e-06, "loss": 1.357, "step": 37068 }, { "epoch": 2.9050940438871473, "grad_norm": 3.664069890975952, "learning_rate": 1.7439754372775652e-06, "loss": 1.4303, "step": 37069 }, { "epoch": 2.905172413793103, "grad_norm": 2.8291566371917725, "learning_rate": 1.7437414269038077e-06, "loss": 0.813, "step": 37070 }, { "epoch": 2.9052507836990595, "grad_norm": 2.6484577655792236, "learning_rate": 1.7435074289151854e-06, "loss": 1.5206, "step": 37071 }, { "epoch": 2.905329153605016, "grad_norm": 3.051215171813965, "learning_rate": 1.7432734433125897e-06, "loss": 1.5498, "step": 37072 }, { "epoch": 2.9054075235109718, "grad_norm": 3.158568859100342, "learning_rate": 1.7430394700969056e-06, "loss": 1.4334, "step": 37073 }, { "epoch": 2.9054858934169276, "grad_norm": 2.2718613147735596, "learning_rate": 1.7428055092690283e-06, "loss": 1.1079, "step": 37074 }, { "epoch": 2.905564263322884, "grad_norm": 2.7533297538757324, "learning_rate": 1.742571560829846e-06, "loss": 1.1145, "step": 37075 }, { "epoch": 2.9056426332288403, "grad_norm": 2.962930917739868, "learning_rate": 1.7423376247802498e-06, "loss": 1.4762, "step": 37076 }, { "epoch": 2.905721003134796, "grad_norm": 2.81811785697937, "learning_rate": 1.7421037011211268e-06, "loss": 1.2604, "step": 37077 }, { "epoch": 2.905799373040752, "grad_norm": 4.010458469390869, "learning_rate": 1.7418697898533654e-06, "loss": 1.5184, "step": 37078 }, { "epoch": 2.9058777429467084, "grad_norm": 2.5940892696380615, "learning_rate": 1.741635890977862e-06, "loss": 1.0029, "step": 37079 }, { "epoch": 2.9059561128526648, "grad_norm": 4.480726718902588, "learning_rate": 1.7414020044955e-06, "loss": 1.2309, "step": 37080 }, { "epoch": 2.9060344827586206, "grad_norm": 4.746582508087158, "learning_rate": 1.7411681304071715e-06, "loss": 1.8896, "step": 37081 }, { "epoch": 2.9061128526645765, "grad_norm": 2.24242901802063, "learning_rate": 1.7409342687137647e-06, "loss": 1.1311, "step": 37082 }, { "epoch": 2.906191222570533, "grad_norm": 4.2916483879089355, "learning_rate": 1.7407004194161708e-06, "loss": 1.0601, "step": 37083 }, { "epoch": 2.906269592476489, "grad_norm": 6.081701755523682, "learning_rate": 1.7404665825152783e-06, "loss": 1.1908, "step": 37084 }, { "epoch": 2.906347962382445, "grad_norm": 3.886197328567505, "learning_rate": 1.7402327580119765e-06, "loss": 1.3447, "step": 37085 }, { "epoch": 2.9064263322884014, "grad_norm": 6.791630268096924, "learning_rate": 1.7399989459071553e-06, "loss": 1.3957, "step": 37086 }, { "epoch": 2.9065047021943573, "grad_norm": 8.072400093078613, "learning_rate": 1.7397651462017034e-06, "loss": 1.6134, "step": 37087 }, { "epoch": 2.9065830721003136, "grad_norm": 5.0106635093688965, "learning_rate": 1.7395313588965101e-06, "loss": 0.9997, "step": 37088 }, { "epoch": 2.9066614420062695, "grad_norm": 2.0990982055664062, "learning_rate": 1.7392975839924654e-06, "loss": 1.0157, "step": 37089 }, { "epoch": 2.906739811912226, "grad_norm": 2.8699989318847656, "learning_rate": 1.7390638214904576e-06, "loss": 1.5515, "step": 37090 }, { "epoch": 2.9068181818181817, "grad_norm": 6.780229568481445, "learning_rate": 1.738830071391376e-06, "loss": 1.474, "step": 37091 }, { "epoch": 2.906896551724138, "grad_norm": 3.404639482498169, "learning_rate": 1.7385963336961098e-06, "loss": 1.6606, "step": 37092 }, { "epoch": 2.906974921630094, "grad_norm": 5.9841389656066895, "learning_rate": 1.7383626084055478e-06, "loss": 1.2694, "step": 37093 }, { "epoch": 2.9070532915360503, "grad_norm": 2.7675366401672363, "learning_rate": 1.738128895520581e-06, "loss": 0.934, "step": 37094 }, { "epoch": 2.907131661442006, "grad_norm": 4.733996868133545, "learning_rate": 1.7378951950420952e-06, "loss": 1.2509, "step": 37095 }, { "epoch": 2.9072100313479625, "grad_norm": 3.830620765686035, "learning_rate": 1.7376615069709785e-06, "loss": 1.2305, "step": 37096 }, { "epoch": 2.9072884012539184, "grad_norm": 4.815067768096924, "learning_rate": 1.7374278313081234e-06, "loss": 1.4235, "step": 37097 }, { "epoch": 2.9073667711598747, "grad_norm": 2.3972418308258057, "learning_rate": 1.7371941680544186e-06, "loss": 1.1832, "step": 37098 }, { "epoch": 2.9074451410658306, "grad_norm": 3.796046733856201, "learning_rate": 1.7369605172107501e-06, "loss": 1.151, "step": 37099 }, { "epoch": 2.907523510971787, "grad_norm": 3.6580052375793457, "learning_rate": 1.7367268787780073e-06, "loss": 1.4037, "step": 37100 }, { "epoch": 2.907601880877743, "grad_norm": 3.2741713523864746, "learning_rate": 1.7364932527570782e-06, "loss": 1.5451, "step": 37101 }, { "epoch": 2.907680250783699, "grad_norm": 3.345656633377075, "learning_rate": 1.7362596391488557e-06, "loss": 1.8217, "step": 37102 }, { "epoch": 2.907758620689655, "grad_norm": 6.31674337387085, "learning_rate": 1.7360260379542232e-06, "loss": 0.7723, "step": 37103 }, { "epoch": 2.9078369905956114, "grad_norm": 5.78922176361084, "learning_rate": 1.7357924491740714e-06, "loss": 0.8143, "step": 37104 }, { "epoch": 2.9079153605015673, "grad_norm": 3.714266538619995, "learning_rate": 1.7355588728092881e-06, "loss": 1.3712, "step": 37105 }, { "epoch": 2.9079937304075236, "grad_norm": 4.570394515991211, "learning_rate": 1.7353253088607625e-06, "loss": 1.5171, "step": 37106 }, { "epoch": 2.9080721003134795, "grad_norm": 4.037352561950684, "learning_rate": 1.735091757329383e-06, "loss": 1.6446, "step": 37107 }, { "epoch": 2.908150470219436, "grad_norm": 3.089695692062378, "learning_rate": 1.7348582182160368e-06, "loss": 1.2525, "step": 37108 }, { "epoch": 2.9082288401253917, "grad_norm": 3.6019580364227295, "learning_rate": 1.7346246915216147e-06, "loss": 1.599, "step": 37109 }, { "epoch": 2.908307210031348, "grad_norm": 4.008512020111084, "learning_rate": 1.734391177247e-06, "loss": 1.5605, "step": 37110 }, { "epoch": 2.908385579937304, "grad_norm": 2.850388288497925, "learning_rate": 1.7341576753930855e-06, "loss": 1.3298, "step": 37111 }, { "epoch": 2.9084639498432603, "grad_norm": 3.435098886489868, "learning_rate": 1.7339241859607575e-06, "loss": 1.2204, "step": 37112 }, { "epoch": 2.908542319749216, "grad_norm": 3.575068712234497, "learning_rate": 1.7336907089509065e-06, "loss": 1.4188, "step": 37113 }, { "epoch": 2.9086206896551725, "grad_norm": 2.8199689388275146, "learning_rate": 1.733457244364416e-06, "loss": 0.9456, "step": 37114 }, { "epoch": 2.9086990595611284, "grad_norm": 11.562005043029785, "learning_rate": 1.733223792202175e-06, "loss": 1.48, "step": 37115 }, { "epoch": 2.9087774294670847, "grad_norm": 3.2984659671783447, "learning_rate": 1.7329903524650748e-06, "loss": 1.0898, "step": 37116 }, { "epoch": 2.9088557993730406, "grad_norm": 3.529742479324341, "learning_rate": 1.7327569251540027e-06, "loss": 1.1991, "step": 37117 }, { "epoch": 2.908934169278997, "grad_norm": 3.558537721633911, "learning_rate": 1.732523510269843e-06, "loss": 1.3801, "step": 37118 }, { "epoch": 2.909012539184953, "grad_norm": 2.8702456951141357, "learning_rate": 1.7322901078134863e-06, "loss": 1.1893, "step": 37119 }, { "epoch": 2.909090909090909, "grad_norm": 3.3576905727386475, "learning_rate": 1.7320567177858188e-06, "loss": 0.9472, "step": 37120 }, { "epoch": 2.909169278996865, "grad_norm": 2.257854461669922, "learning_rate": 1.7318233401877299e-06, "loss": 1.3737, "step": 37121 }, { "epoch": 2.9092476489028214, "grad_norm": 3.9024720191955566, "learning_rate": 1.731589975020106e-06, "loss": 1.2399, "step": 37122 }, { "epoch": 2.9093260188087773, "grad_norm": 8.186575889587402, "learning_rate": 1.7313566222838346e-06, "loss": 0.5912, "step": 37123 }, { "epoch": 2.9094043887147336, "grad_norm": 2.37408709526062, "learning_rate": 1.731123281979804e-06, "loss": 1.284, "step": 37124 }, { "epoch": 2.9094827586206895, "grad_norm": 3.594357490539551, "learning_rate": 1.730889954108902e-06, "loss": 1.6156, "step": 37125 }, { "epoch": 2.909561128526646, "grad_norm": 2.939535617828369, "learning_rate": 1.7306566386720148e-06, "loss": 1.3086, "step": 37126 }, { "epoch": 2.9096394984326017, "grad_norm": 5.39757776260376, "learning_rate": 1.7304233356700313e-06, "loss": 1.0029, "step": 37127 }, { "epoch": 2.909717868338558, "grad_norm": 3.308603048324585, "learning_rate": 1.730190045103839e-06, "loss": 1.4466, "step": 37128 }, { "epoch": 2.909796238244514, "grad_norm": 2.758821487426758, "learning_rate": 1.729956766974321e-06, "loss": 1.1952, "step": 37129 }, { "epoch": 2.9098746081504703, "grad_norm": 3.4225175380706787, "learning_rate": 1.7297235012823698e-06, "loss": 1.0714, "step": 37130 }, { "epoch": 2.9099529780564266, "grad_norm": 3.853292942047119, "learning_rate": 1.7294902480288706e-06, "loss": 1.8552, "step": 37131 }, { "epoch": 2.9100313479623825, "grad_norm": 3.502523422241211, "learning_rate": 1.7292570072147124e-06, "loss": 0.565, "step": 37132 }, { "epoch": 2.9101097178683384, "grad_norm": 3.4062623977661133, "learning_rate": 1.7290237788407788e-06, "loss": 1.2229, "step": 37133 }, { "epoch": 2.9101880877742947, "grad_norm": 2.9477646350860596, "learning_rate": 1.7287905629079572e-06, "loss": 1.1753, "step": 37134 }, { "epoch": 2.910266457680251, "grad_norm": 3.030761957168579, "learning_rate": 1.728557359417139e-06, "loss": 1.4845, "step": 37135 }, { "epoch": 2.910344827586207, "grad_norm": 2.435314655303955, "learning_rate": 1.728324168369207e-06, "loss": 1.2966, "step": 37136 }, { "epoch": 2.910423197492163, "grad_norm": 3.490466594696045, "learning_rate": 1.7280909897650493e-06, "loss": 1.4873, "step": 37137 }, { "epoch": 2.910501567398119, "grad_norm": 2.7933406829833984, "learning_rate": 1.7278578236055528e-06, "loss": 1.2645, "step": 37138 }, { "epoch": 2.9105799373040755, "grad_norm": 15.086167335510254, "learning_rate": 1.7276246698916045e-06, "loss": 0.8341, "step": 37139 }, { "epoch": 2.9106583072100314, "grad_norm": 5.710793972015381, "learning_rate": 1.727391528624091e-06, "loss": 1.5866, "step": 37140 }, { "epoch": 2.9107366771159873, "grad_norm": 4.142698287963867, "learning_rate": 1.7271583998038998e-06, "loss": 1.6963, "step": 37141 }, { "epoch": 2.9108150470219436, "grad_norm": 2.0068681240081787, "learning_rate": 1.7269252834319168e-06, "loss": 1.2431, "step": 37142 }, { "epoch": 2.9108934169279, "grad_norm": 3.9515485763549805, "learning_rate": 1.7266921795090285e-06, "loss": 1.3874, "step": 37143 }, { "epoch": 2.910971786833856, "grad_norm": 3.852407932281494, "learning_rate": 1.726459088036122e-06, "loss": 1.4994, "step": 37144 }, { "epoch": 2.9110501567398117, "grad_norm": 2.5158658027648926, "learning_rate": 1.7262260090140837e-06, "loss": 1.4464, "step": 37145 }, { "epoch": 2.911128526645768, "grad_norm": 3.09031081199646, "learning_rate": 1.7259929424438005e-06, "loss": 1.3792, "step": 37146 }, { "epoch": 2.9112068965517244, "grad_norm": 4.305634498596191, "learning_rate": 1.72575988832616e-06, "loss": 1.1731, "step": 37147 }, { "epoch": 2.9112852664576803, "grad_norm": 3.1563549041748047, "learning_rate": 1.7255268466620433e-06, "loss": 1.3487, "step": 37148 }, { "epoch": 2.911363636363636, "grad_norm": 6.757184028625488, "learning_rate": 1.725293817452343e-06, "loss": 1.3878, "step": 37149 }, { "epoch": 2.9114420062695925, "grad_norm": 5.607272624969482, "learning_rate": 1.7250608006979447e-06, "loss": 2.0203, "step": 37150 }, { "epoch": 2.911520376175549, "grad_norm": 4.423411846160889, "learning_rate": 1.7248277963997305e-06, "loss": 0.977, "step": 37151 }, { "epoch": 2.9115987460815047, "grad_norm": 2.8218672275543213, "learning_rate": 1.7245948045585898e-06, "loss": 1.2436, "step": 37152 }, { "epoch": 2.9116771159874606, "grad_norm": 2.624465227127075, "learning_rate": 1.7243618251754062e-06, "loss": 1.1186, "step": 37153 }, { "epoch": 2.911755485893417, "grad_norm": 4.288261890411377, "learning_rate": 1.7241288582510712e-06, "loss": 1.2075, "step": 37154 }, { "epoch": 2.9118338557993733, "grad_norm": 2.4024248123168945, "learning_rate": 1.723895903786466e-06, "loss": 1.2631, "step": 37155 }, { "epoch": 2.911912225705329, "grad_norm": 3.1576523780822754, "learning_rate": 1.723662961782478e-06, "loss": 1.4381, "step": 37156 }, { "epoch": 2.911990595611285, "grad_norm": 9.514036178588867, "learning_rate": 1.7234300322399928e-06, "loss": 0.8061, "step": 37157 }, { "epoch": 2.9120689655172414, "grad_norm": 3.3519458770751953, "learning_rate": 1.723197115159897e-06, "loss": 1.3032, "step": 37158 }, { "epoch": 2.9121473354231977, "grad_norm": 3.938838005065918, "learning_rate": 1.7229642105430767e-06, "loss": 1.3946, "step": 37159 }, { "epoch": 2.9122257053291536, "grad_norm": 2.171114444732666, "learning_rate": 1.7227313183904165e-06, "loss": 1.304, "step": 37160 }, { "epoch": 2.9123040752351095, "grad_norm": 3.4028677940368652, "learning_rate": 1.7224984387028039e-06, "loss": 1.2028, "step": 37161 }, { "epoch": 2.912382445141066, "grad_norm": 3.0730717182159424, "learning_rate": 1.7222655714811231e-06, "loss": 1.6025, "step": 37162 }, { "epoch": 2.912460815047022, "grad_norm": 2.426764726638794, "learning_rate": 1.7220327167262612e-06, "loss": 0.8765, "step": 37163 }, { "epoch": 2.912539184952978, "grad_norm": 2.454972982406616, "learning_rate": 1.7217998744391029e-06, "loss": 1.3794, "step": 37164 }, { "epoch": 2.912617554858934, "grad_norm": 4.407966136932373, "learning_rate": 1.7215670446205358e-06, "loss": 1.4647, "step": 37165 }, { "epoch": 2.9126959247648903, "grad_norm": 6.166140079498291, "learning_rate": 1.7213342272714418e-06, "loss": 1.7248, "step": 37166 }, { "epoch": 2.9127742946708466, "grad_norm": 6.332196235656738, "learning_rate": 1.721101422392707e-06, "loss": 0.9166, "step": 37167 }, { "epoch": 2.9128526645768025, "grad_norm": 13.53805160522461, "learning_rate": 1.7208686299852202e-06, "loss": 1.5542, "step": 37168 }, { "epoch": 2.9129310344827584, "grad_norm": 2.87137508392334, "learning_rate": 1.720635850049866e-06, "loss": 1.2741, "step": 37169 }, { "epoch": 2.9130094043887147, "grad_norm": 2.0538041591644287, "learning_rate": 1.7204030825875267e-06, "loss": 1.3063, "step": 37170 }, { "epoch": 2.913087774294671, "grad_norm": 3.789616107940674, "learning_rate": 1.7201703275990906e-06, "loss": 1.3305, "step": 37171 }, { "epoch": 2.913166144200627, "grad_norm": 2.4909043312072754, "learning_rate": 1.7199375850854395e-06, "loss": 1.2748, "step": 37172 }, { "epoch": 2.913244514106583, "grad_norm": 2.8911125659942627, "learning_rate": 1.7197048550474643e-06, "loss": 1.2849, "step": 37173 }, { "epoch": 2.913322884012539, "grad_norm": 2.7890522480010986, "learning_rate": 1.7194721374860458e-06, "loss": 1.2591, "step": 37174 }, { "epoch": 2.9134012539184955, "grad_norm": 16.356834411621094, "learning_rate": 1.7192394324020701e-06, "loss": 1.3388, "step": 37175 }, { "epoch": 2.9134796238244514, "grad_norm": 3.015781879425049, "learning_rate": 1.7190067397964223e-06, "loss": 1.1997, "step": 37176 }, { "epoch": 2.9135579937304072, "grad_norm": 4.544364929199219, "learning_rate": 1.718774059669988e-06, "loss": 1.5162, "step": 37177 }, { "epoch": 2.9136363636363636, "grad_norm": 2.3345015048980713, "learning_rate": 1.7185413920236517e-06, "loss": 1.6204, "step": 37178 }, { "epoch": 2.91371473354232, "grad_norm": 3.72998309135437, "learning_rate": 1.7183087368582985e-06, "loss": 1.3022, "step": 37179 }, { "epoch": 2.913793103448276, "grad_norm": 2.4432265758514404, "learning_rate": 1.7180760941748132e-06, "loss": 1.2211, "step": 37180 }, { "epoch": 2.913871473354232, "grad_norm": 3.0906312465667725, "learning_rate": 1.7178434639740804e-06, "loss": 1.6368, "step": 37181 }, { "epoch": 2.913949843260188, "grad_norm": 2.06648850440979, "learning_rate": 1.7176108462569858e-06, "loss": 1.3543, "step": 37182 }, { "epoch": 2.9140282131661444, "grad_norm": 4.479336738586426, "learning_rate": 1.7173782410244134e-06, "loss": 1.7954, "step": 37183 }, { "epoch": 2.9141065830721002, "grad_norm": 3.108055591583252, "learning_rate": 1.7171456482772498e-06, "loss": 1.3845, "step": 37184 }, { "epoch": 2.9141849529780566, "grad_norm": 2.934523344039917, "learning_rate": 1.7169130680163742e-06, "loss": 1.3407, "step": 37185 }, { "epoch": 2.9142633228840125, "grad_norm": 3.4184231758117676, "learning_rate": 1.7166805002426773e-06, "loss": 1.3959, "step": 37186 }, { "epoch": 2.914341692789969, "grad_norm": 2.566218376159668, "learning_rate": 1.716447944957041e-06, "loss": 1.5887, "step": 37187 }, { "epoch": 2.9144200626959247, "grad_norm": 3.791553020477295, "learning_rate": 1.7162154021603522e-06, "loss": 1.9325, "step": 37188 }, { "epoch": 2.914498432601881, "grad_norm": 2.2865610122680664, "learning_rate": 1.7159828718534915e-06, "loss": 1.2843, "step": 37189 }, { "epoch": 2.914576802507837, "grad_norm": 5.182239532470703, "learning_rate": 1.715750354037345e-06, "loss": 1.2558, "step": 37190 }, { "epoch": 2.9146551724137932, "grad_norm": 4.956449508666992, "learning_rate": 1.7155178487127978e-06, "loss": 1.2985, "step": 37191 }, { "epoch": 2.914733542319749, "grad_norm": 2.343961477279663, "learning_rate": 1.715285355880733e-06, "loss": 1.202, "step": 37192 }, { "epoch": 2.9148119122257055, "grad_norm": 8.293872833251953, "learning_rate": 1.715052875542036e-06, "loss": 1.0801, "step": 37193 }, { "epoch": 2.9148902821316613, "grad_norm": 4.588250637054443, "learning_rate": 1.71482040769759e-06, "loss": 1.1614, "step": 37194 }, { "epoch": 2.9149686520376177, "grad_norm": 8.752490997314453, "learning_rate": 1.7145879523482807e-06, "loss": 1.1611, "step": 37195 }, { "epoch": 2.9150470219435736, "grad_norm": 2.8076422214508057, "learning_rate": 1.7143555094949905e-06, "loss": 1.3982, "step": 37196 }, { "epoch": 2.91512539184953, "grad_norm": 2.5363376140594482, "learning_rate": 1.7141230791386049e-06, "loss": 1.291, "step": 37197 }, { "epoch": 2.915203761755486, "grad_norm": 3.625049114227295, "learning_rate": 1.7138906612800072e-06, "loss": 1.3788, "step": 37198 }, { "epoch": 2.915282131661442, "grad_norm": 3.89082932472229, "learning_rate": 1.7136582559200816e-06, "loss": 1.3046, "step": 37199 }, { "epoch": 2.915360501567398, "grad_norm": 2.778383493423462, "learning_rate": 1.7134258630597117e-06, "loss": 1.286, "step": 37200 }, { "epoch": 2.9154388714733543, "grad_norm": 5.275629997253418, "learning_rate": 1.713193482699782e-06, "loss": 1.1615, "step": 37201 }, { "epoch": 2.9155172413793102, "grad_norm": 4.262384414672852, "learning_rate": 1.7129611148411763e-06, "loss": 1.0005, "step": 37202 }, { "epoch": 2.9155956112852666, "grad_norm": 3.399578332901001, "learning_rate": 1.7127287594847798e-06, "loss": 1.0363, "step": 37203 }, { "epoch": 2.9156739811912225, "grad_norm": 5.234593868255615, "learning_rate": 1.7124964166314711e-06, "loss": 1.4227, "step": 37204 }, { "epoch": 2.915752351097179, "grad_norm": 2.403308868408203, "learning_rate": 1.712264086282139e-06, "loss": 0.6742, "step": 37205 }, { "epoch": 2.9158307210031347, "grad_norm": 2.7243003845214844, "learning_rate": 1.7120317684376676e-06, "loss": 1.7686, "step": 37206 }, { "epoch": 2.915909090909091, "grad_norm": 6.463204860687256, "learning_rate": 1.7117994630989365e-06, "loss": 1.2356, "step": 37207 }, { "epoch": 2.915987460815047, "grad_norm": 4.678611755371094, "learning_rate": 1.7115671702668319e-06, "loss": 1.4335, "step": 37208 }, { "epoch": 2.9160658307210032, "grad_norm": 2.7577435970306396, "learning_rate": 1.7113348899422344e-06, "loss": 1.0235, "step": 37209 }, { "epoch": 2.916144200626959, "grad_norm": 2.3193247318267822, "learning_rate": 1.7111026221260334e-06, "loss": 1.3487, "step": 37210 }, { "epoch": 2.9162225705329154, "grad_norm": 2.915043354034424, "learning_rate": 1.7108703668191074e-06, "loss": 1.2585, "step": 37211 }, { "epoch": 2.9163009404388713, "grad_norm": 3.160982131958008, "learning_rate": 1.7106381240223407e-06, "loss": 1.2958, "step": 37212 }, { "epoch": 2.9163793103448277, "grad_norm": 2.1124351024627686, "learning_rate": 1.7104058937366174e-06, "loss": 1.1844, "step": 37213 }, { "epoch": 2.9164576802507836, "grad_norm": 3.531907558441162, "learning_rate": 1.71017367596282e-06, "loss": 0.9183, "step": 37214 }, { "epoch": 2.91653605015674, "grad_norm": 3.541846752166748, "learning_rate": 1.709941470701833e-06, "loss": 1.2569, "step": 37215 }, { "epoch": 2.916614420062696, "grad_norm": 4.595255374908447, "learning_rate": 1.7097092779545382e-06, "loss": 0.9037, "step": 37216 }, { "epoch": 2.916692789968652, "grad_norm": 2.706839084625244, "learning_rate": 1.709477097721819e-06, "loss": 1.033, "step": 37217 }, { "epoch": 2.916771159874608, "grad_norm": 2.63623309135437, "learning_rate": 1.7092449300045594e-06, "loss": 1.2108, "step": 37218 }, { "epoch": 2.9168495297805643, "grad_norm": 5.0385260581970215, "learning_rate": 1.7090127748036412e-06, "loss": 1.3126, "step": 37219 }, { "epoch": 2.91692789968652, "grad_norm": 2.8699698448181152, "learning_rate": 1.7087806321199484e-06, "loss": 1.1512, "step": 37220 }, { "epoch": 2.9170062695924766, "grad_norm": 4.13134241104126, "learning_rate": 1.7085485019543657e-06, "loss": 1.7109, "step": 37221 }, { "epoch": 2.9170846394984324, "grad_norm": 3.445420503616333, "learning_rate": 1.7083163843077715e-06, "loss": 1.2362, "step": 37222 }, { "epoch": 2.9171630094043888, "grad_norm": 2.853672742843628, "learning_rate": 1.7080842791810497e-06, "loss": 1.873, "step": 37223 }, { "epoch": 2.9172413793103447, "grad_norm": 2.199097156524658, "learning_rate": 1.7078521865750864e-06, "loss": 1.4867, "step": 37224 }, { "epoch": 2.917319749216301, "grad_norm": 3.051084518432617, "learning_rate": 1.7076201064907637e-06, "loss": 1.6015, "step": 37225 }, { "epoch": 2.9173981191222573, "grad_norm": 2.9364638328552246, "learning_rate": 1.7073880389289615e-06, "loss": 1.1768, "step": 37226 }, { "epoch": 2.917476489028213, "grad_norm": 17.41619110107422, "learning_rate": 1.707155983890564e-06, "loss": 1.9336, "step": 37227 }, { "epoch": 2.917554858934169, "grad_norm": 3.042184829711914, "learning_rate": 1.7069239413764522e-06, "loss": 1.298, "step": 37228 }, { "epoch": 2.9176332288401254, "grad_norm": 3.6846706867218018, "learning_rate": 1.7066919113875136e-06, "loss": 1.5979, "step": 37229 }, { "epoch": 2.9177115987460818, "grad_norm": 3.096344470977783, "learning_rate": 1.7064598939246258e-06, "loss": 0.9544, "step": 37230 }, { "epoch": 2.9177899686520377, "grad_norm": 4.681224822998047, "learning_rate": 1.706227888988673e-06, "loss": 1.188, "step": 37231 }, { "epoch": 2.9178683385579935, "grad_norm": 3.18188738822937, "learning_rate": 1.7059958965805373e-06, "loss": 1.326, "step": 37232 }, { "epoch": 2.91794670846395, "grad_norm": 4.2799835205078125, "learning_rate": 1.7057639167011015e-06, "loss": 1.4627, "step": 37233 }, { "epoch": 2.918025078369906, "grad_norm": 3.76557993888855, "learning_rate": 1.7055319493512474e-06, "loss": 1.5096, "step": 37234 }, { "epoch": 2.918103448275862, "grad_norm": 4.443014621734619, "learning_rate": 1.7052999945318577e-06, "loss": 1.6911, "step": 37235 }, { "epoch": 2.918181818181818, "grad_norm": 2.9561572074890137, "learning_rate": 1.7050680522438167e-06, "loss": 1.0186, "step": 37236 }, { "epoch": 2.9182601880877743, "grad_norm": 8.966179847717285, "learning_rate": 1.7048361224880006e-06, "loss": 1.3061, "step": 37237 }, { "epoch": 2.9183385579937307, "grad_norm": 4.442173480987549, "learning_rate": 1.7046042052652971e-06, "loss": 1.5029, "step": 37238 }, { "epoch": 2.9184169278996865, "grad_norm": 4.254333019256592, "learning_rate": 1.7043723005765866e-06, "loss": 1.7995, "step": 37239 }, { "epoch": 2.9184952978056424, "grad_norm": 4.98553466796875, "learning_rate": 1.704140408422753e-06, "loss": 1.8918, "step": 37240 }, { "epoch": 2.9185736677115988, "grad_norm": 1.9485751390457153, "learning_rate": 1.7039085288046747e-06, "loss": 1.1772, "step": 37241 }, { "epoch": 2.918652037617555, "grad_norm": 3.4485435485839844, "learning_rate": 1.7036766617232336e-06, "loss": 1.4593, "step": 37242 }, { "epoch": 2.918730407523511, "grad_norm": 2.6333673000335693, "learning_rate": 1.7034448071793153e-06, "loss": 1.2652, "step": 37243 }, { "epoch": 2.918808777429467, "grad_norm": 4.026763439178467, "learning_rate": 1.7032129651738012e-06, "loss": 1.4464, "step": 37244 }, { "epoch": 2.918887147335423, "grad_norm": 4.259469032287598, "learning_rate": 1.70298113570757e-06, "loss": 1.5641, "step": 37245 }, { "epoch": 2.9189655172413795, "grad_norm": 3.467979669570923, "learning_rate": 1.7027493187815052e-06, "loss": 0.9178, "step": 37246 }, { "epoch": 2.9190438871473354, "grad_norm": 2.9966814517974854, "learning_rate": 1.7025175143964883e-06, "loss": 1.4154, "step": 37247 }, { "epoch": 2.9191222570532913, "grad_norm": 2.5309979915618896, "learning_rate": 1.7022857225534012e-06, "loss": 1.2594, "step": 37248 }, { "epoch": 2.9192006269592476, "grad_norm": 4.092851161956787, "learning_rate": 1.7020539432531253e-06, "loss": 0.9801, "step": 37249 }, { "epoch": 2.919278996865204, "grad_norm": 3.0887866020202637, "learning_rate": 1.701822176496542e-06, "loss": 1.3215, "step": 37250 }, { "epoch": 2.91935736677116, "grad_norm": 2.6848976612091064, "learning_rate": 1.7015904222845335e-06, "loss": 1.4064, "step": 37251 }, { "epoch": 2.9194357366771158, "grad_norm": 3.210543155670166, "learning_rate": 1.7013586806179806e-06, "loss": 1.3387, "step": 37252 }, { "epoch": 2.919514106583072, "grad_norm": 1.8839000463485718, "learning_rate": 1.7011269514977652e-06, "loss": 1.3184, "step": 37253 }, { "epoch": 2.9195924764890284, "grad_norm": 2.349837064743042, "learning_rate": 1.700895234924768e-06, "loss": 1.2157, "step": 37254 }, { "epoch": 2.9196708463949843, "grad_norm": 2.8626532554626465, "learning_rate": 1.7006635308998731e-06, "loss": 1.335, "step": 37255 }, { "epoch": 2.91974921630094, "grad_norm": 7.063401222229004, "learning_rate": 1.7004318394239561e-06, "loss": 1.626, "step": 37256 }, { "epoch": 2.9198275862068965, "grad_norm": 2.7591989040374756, "learning_rate": 1.700200160497903e-06, "loss": 1.5944, "step": 37257 }, { "epoch": 2.919905956112853, "grad_norm": 3.7918155193328857, "learning_rate": 1.6999684941225935e-06, "loss": 1.4117, "step": 37258 }, { "epoch": 2.9199843260188088, "grad_norm": 4.1316070556640625, "learning_rate": 1.6997368402989105e-06, "loss": 1.6105, "step": 37259 }, { "epoch": 2.9200626959247646, "grad_norm": 4.297885894775391, "learning_rate": 1.6995051990277322e-06, "loss": 1.562, "step": 37260 }, { "epoch": 2.920141065830721, "grad_norm": 3.7532620429992676, "learning_rate": 1.699273570309939e-06, "loss": 1.3008, "step": 37261 }, { "epoch": 2.9202194357366773, "grad_norm": 3.1202828884124756, "learning_rate": 1.699041954146417e-06, "loss": 0.7556, "step": 37262 }, { "epoch": 2.920297805642633, "grad_norm": 10.20703411102295, "learning_rate": 1.6988103505380427e-06, "loss": 1.0177, "step": 37263 }, { "epoch": 2.920376175548589, "grad_norm": 5.9371018409729, "learning_rate": 1.6985787594856979e-06, "loss": 0.9071, "step": 37264 }, { "epoch": 2.9204545454545454, "grad_norm": 2.4183709621429443, "learning_rate": 1.6983471809902636e-06, "loss": 1.5715, "step": 37265 }, { "epoch": 2.9205329153605017, "grad_norm": 3.130635976791382, "learning_rate": 1.6981156150526207e-06, "loss": 1.4592, "step": 37266 }, { "epoch": 2.9206112852664576, "grad_norm": 3.4559760093688965, "learning_rate": 1.6978840616736504e-06, "loss": 1.2306, "step": 37267 }, { "epoch": 2.9206896551724135, "grad_norm": 8.058408737182617, "learning_rate": 1.6976525208542327e-06, "loss": 1.2012, "step": 37268 }, { "epoch": 2.92076802507837, "grad_norm": 3.676098585128784, "learning_rate": 1.697420992595249e-06, "loss": 1.2684, "step": 37269 }, { "epoch": 2.920846394984326, "grad_norm": 3.341991424560547, "learning_rate": 1.6971894768975794e-06, "loss": 1.2282, "step": 37270 }, { "epoch": 2.920924764890282, "grad_norm": 7.2696614265441895, "learning_rate": 1.6969579737621045e-06, "loss": 0.7641, "step": 37271 }, { "epoch": 2.9210031347962384, "grad_norm": 3.6786510944366455, "learning_rate": 1.6967264831897051e-06, "loss": 1.6037, "step": 37272 }, { "epoch": 2.9210815047021943, "grad_norm": 2.6212127208709717, "learning_rate": 1.6964950051812633e-06, "loss": 1.3252, "step": 37273 }, { "epoch": 2.9211598746081506, "grad_norm": 8.784696578979492, "learning_rate": 1.696263539737656e-06, "loss": 0.7691, "step": 37274 }, { "epoch": 2.9212382445141065, "grad_norm": 3.0165252685546875, "learning_rate": 1.6960320868597629e-06, "loss": 1.5585, "step": 37275 }, { "epoch": 2.921316614420063, "grad_norm": 3.492586135864258, "learning_rate": 1.6958006465484694e-06, "loss": 1.3294, "step": 37276 }, { "epoch": 2.9213949843260187, "grad_norm": 3.7847256660461426, "learning_rate": 1.695569218804654e-06, "loss": 1.764, "step": 37277 }, { "epoch": 2.921473354231975, "grad_norm": 2.386528730392456, "learning_rate": 1.6953378036291946e-06, "loss": 1.1987, "step": 37278 }, { "epoch": 2.921551724137931, "grad_norm": 4.0782928466796875, "learning_rate": 1.695106401022973e-06, "loss": 0.9111, "step": 37279 }, { "epoch": 2.9216300940438873, "grad_norm": 11.491697311401367, "learning_rate": 1.6948750109868673e-06, "loss": 2.2326, "step": 37280 }, { "epoch": 2.921708463949843, "grad_norm": 2.578153371810913, "learning_rate": 1.694643633521763e-06, "loss": 1.6756, "step": 37281 }, { "epoch": 2.9217868338557995, "grad_norm": 4.111958026885986, "learning_rate": 1.6944122686285347e-06, "loss": 1.1969, "step": 37282 }, { "epoch": 2.9218652037617554, "grad_norm": 2.260047435760498, "learning_rate": 1.6941809163080642e-06, "loss": 1.0597, "step": 37283 }, { "epoch": 2.9219435736677117, "grad_norm": 2.6847848892211914, "learning_rate": 1.6939495765612318e-06, "loss": 1.531, "step": 37284 }, { "epoch": 2.9220219435736676, "grad_norm": 3.0788450241088867, "learning_rate": 1.6937182493889165e-06, "loss": 0.8371, "step": 37285 }, { "epoch": 2.922100313479624, "grad_norm": 2.6586318016052246, "learning_rate": 1.6934869347919997e-06, "loss": 0.5709, "step": 37286 }, { "epoch": 2.92217868338558, "grad_norm": 2.858015298843384, "learning_rate": 1.6932556327713596e-06, "loss": 1.2386, "step": 37287 }, { "epoch": 2.922257053291536, "grad_norm": 2.697404146194458, "learning_rate": 1.6930243433278765e-06, "loss": 1.0955, "step": 37288 }, { "epoch": 2.922335423197492, "grad_norm": 3.1119046211242676, "learning_rate": 1.6927930664624308e-06, "loss": 0.7336, "step": 37289 }, { "epoch": 2.9224137931034484, "grad_norm": 4.777828216552734, "learning_rate": 1.692561802175901e-06, "loss": 1.4694, "step": 37290 }, { "epoch": 2.9224921630094043, "grad_norm": 2.8023483753204346, "learning_rate": 1.6923305504691679e-06, "loss": 1.2308, "step": 37291 }, { "epoch": 2.9225705329153606, "grad_norm": 3.232950448989868, "learning_rate": 1.6920993113431117e-06, "loss": 1.8554, "step": 37292 }, { "epoch": 2.9226489028213165, "grad_norm": 6.541749000549316, "learning_rate": 1.6918680847986073e-06, "loss": 1.5065, "step": 37293 }, { "epoch": 2.922727272727273, "grad_norm": 4.299927711486816, "learning_rate": 1.6916368708365394e-06, "loss": 1.5651, "step": 37294 }, { "epoch": 2.9228056426332287, "grad_norm": 7.775384426116943, "learning_rate": 1.6914056694577857e-06, "loss": 1.0201, "step": 37295 }, { "epoch": 2.922884012539185, "grad_norm": 4.833037853240967, "learning_rate": 1.6911744806632268e-06, "loss": 0.8953, "step": 37296 }, { "epoch": 2.922962382445141, "grad_norm": 4.240053176879883, "learning_rate": 1.6909433044537394e-06, "loss": 1.2693, "step": 37297 }, { "epoch": 2.9230407523510973, "grad_norm": 8.084943771362305, "learning_rate": 1.690712140830202e-06, "loss": 1.0537, "step": 37298 }, { "epoch": 2.923119122257053, "grad_norm": 2.7418198585510254, "learning_rate": 1.6904809897934976e-06, "loss": 1.3771, "step": 37299 }, { "epoch": 2.9231974921630095, "grad_norm": 4.278358459472656, "learning_rate": 1.6902498513445053e-06, "loss": 1.646, "step": 37300 }, { "epoch": 2.9232758620689654, "grad_norm": 3.234502077102661, "learning_rate": 1.6900187254841011e-06, "loss": 1.7176, "step": 37301 }, { "epoch": 2.9233542319749217, "grad_norm": 2.898780584335327, "learning_rate": 1.6897876122131652e-06, "loss": 1.3804, "step": 37302 }, { "epoch": 2.9234326018808776, "grad_norm": 2.930661678314209, "learning_rate": 1.6895565115325775e-06, "loss": 1.4218, "step": 37303 }, { "epoch": 2.923510971786834, "grad_norm": 4.436877250671387, "learning_rate": 1.689325423443216e-06, "loss": 1.0596, "step": 37304 }, { "epoch": 2.92358934169279, "grad_norm": 3.1983418464660645, "learning_rate": 1.6890943479459598e-06, "loss": 1.2124, "step": 37305 }, { "epoch": 2.923667711598746, "grad_norm": 2.8105766773223877, "learning_rate": 1.6888632850416886e-06, "loss": 1.145, "step": 37306 }, { "epoch": 2.923746081504702, "grad_norm": 4.079148292541504, "learning_rate": 1.688632234731281e-06, "loss": 1.247, "step": 37307 }, { "epoch": 2.9238244514106584, "grad_norm": 3.4110682010650635, "learning_rate": 1.688401197015615e-06, "loss": 0.9363, "step": 37308 }, { "epoch": 2.9239028213166143, "grad_norm": 4.14746618270874, "learning_rate": 1.6881701718955706e-06, "loss": 1.5726, "step": 37309 }, { "epoch": 2.9239811912225706, "grad_norm": 3.18489146232605, "learning_rate": 1.6879391593720252e-06, "loss": 1.0926, "step": 37310 }, { "epoch": 2.9240595611285265, "grad_norm": 3.224987030029297, "learning_rate": 1.68770815944586e-06, "loss": 1.25, "step": 37311 }, { "epoch": 2.924137931034483, "grad_norm": 3.80768084526062, "learning_rate": 1.6874771721179484e-06, "loss": 1.3199, "step": 37312 }, { "epoch": 2.9242163009404387, "grad_norm": 6.288384437561035, "learning_rate": 1.6872461973891734e-06, "loss": 1.4479, "step": 37313 }, { "epoch": 2.924294670846395, "grad_norm": 3.2294135093688965, "learning_rate": 1.687015235260413e-06, "loss": 1.3203, "step": 37314 }, { "epoch": 2.924373040752351, "grad_norm": 2.8186941146850586, "learning_rate": 1.6867842857325462e-06, "loss": 1.4829, "step": 37315 }, { "epoch": 2.9244514106583073, "grad_norm": 3.975635528564453, "learning_rate": 1.6865533488064485e-06, "loss": 1.9077, "step": 37316 }, { "epoch": 2.924529780564263, "grad_norm": 2.2664759159088135, "learning_rate": 1.6863224244829984e-06, "loss": 1.1561, "step": 37317 }, { "epoch": 2.9246081504702195, "grad_norm": 10.348824501037598, "learning_rate": 1.6860915127630794e-06, "loss": 1.5589, "step": 37318 }, { "epoch": 2.9246865203761754, "grad_norm": 2.35436749458313, "learning_rate": 1.685860613647564e-06, "loss": 1.3702, "step": 37319 }, { "epoch": 2.9247648902821317, "grad_norm": 5.694375514984131, "learning_rate": 1.6856297271373328e-06, "loss": 1.7696, "step": 37320 }, { "epoch": 2.924843260188088, "grad_norm": 2.8140127658843994, "learning_rate": 1.6853988532332637e-06, "loss": 1.1777, "step": 37321 }, { "epoch": 2.924921630094044, "grad_norm": 3.207139015197754, "learning_rate": 1.685167991936235e-06, "loss": 1.2327, "step": 37322 }, { "epoch": 2.925, "grad_norm": 2.6727523803710938, "learning_rate": 1.6849371432471245e-06, "loss": 1.2522, "step": 37323 }, { "epoch": 2.925078369905956, "grad_norm": 3.446927785873413, "learning_rate": 1.6847063071668107e-06, "loss": 1.6207, "step": 37324 }, { "epoch": 2.9251567398119125, "grad_norm": 4.009300708770752, "learning_rate": 1.6844754836961713e-06, "loss": 1.093, "step": 37325 }, { "epoch": 2.9252351097178684, "grad_norm": 2.9370880126953125, "learning_rate": 1.684244672836084e-06, "loss": 1.6205, "step": 37326 }, { "epoch": 2.9253134796238243, "grad_norm": 2.337343692779541, "learning_rate": 1.684013874587427e-06, "loss": 1.3032, "step": 37327 }, { "epoch": 2.9253918495297806, "grad_norm": 3.622422456741333, "learning_rate": 1.6837830889510775e-06, "loss": 1.643, "step": 37328 }, { "epoch": 2.925470219435737, "grad_norm": 2.687213897705078, "learning_rate": 1.6835523159279161e-06, "loss": 1.2038, "step": 37329 }, { "epoch": 2.925548589341693, "grad_norm": 2.85097074508667, "learning_rate": 1.683321555518816e-06, "loss": 1.3616, "step": 37330 }, { "epoch": 2.9256269592476487, "grad_norm": 2.6605257987976074, "learning_rate": 1.683090807724656e-06, "loss": 1.2604, "step": 37331 }, { "epoch": 2.925705329153605, "grad_norm": 2.552428960800171, "learning_rate": 1.6828600725463162e-06, "loss": 1.2046, "step": 37332 }, { "epoch": 2.9257836990595614, "grad_norm": 2.854435443878174, "learning_rate": 1.6826293499846746e-06, "loss": 1.5478, "step": 37333 }, { "epoch": 2.9258620689655173, "grad_norm": 2.5383739471435547, "learning_rate": 1.6823986400406056e-06, "loss": 1.4804, "step": 37334 }, { "epoch": 2.925940438871473, "grad_norm": 3.589790105819702, "learning_rate": 1.6821679427149878e-06, "loss": 1.2227, "step": 37335 }, { "epoch": 2.9260188087774295, "grad_norm": 4.140042304992676, "learning_rate": 1.681937258008698e-06, "loss": 1.0071, "step": 37336 }, { "epoch": 2.926097178683386, "grad_norm": 2.5466043949127197, "learning_rate": 1.6817065859226173e-06, "loss": 1.175, "step": 37337 }, { "epoch": 2.9261755485893417, "grad_norm": 2.514409065246582, "learning_rate": 1.6814759264576191e-06, "loss": 1.5139, "step": 37338 }, { "epoch": 2.9262539184952976, "grad_norm": 2.0987708568573, "learning_rate": 1.6812452796145818e-06, "loss": 1.5704, "step": 37339 }, { "epoch": 2.926332288401254, "grad_norm": 2.258892297744751, "learning_rate": 1.6810146453943827e-06, "loss": 1.4063, "step": 37340 }, { "epoch": 2.9264106583072103, "grad_norm": 4.0685014724731445, "learning_rate": 1.6807840237978996e-06, "loss": 1.4494, "step": 37341 }, { "epoch": 2.926489028213166, "grad_norm": 3.839841842651367, "learning_rate": 1.6805534148260088e-06, "loss": 1.4656, "step": 37342 }, { "epoch": 2.926567398119122, "grad_norm": 5.4914422035217285, "learning_rate": 1.6803228184795877e-06, "loss": 1.485, "step": 37343 }, { "epoch": 2.9266457680250784, "grad_norm": 2.9918816089630127, "learning_rate": 1.6800922347595156e-06, "loss": 1.2979, "step": 37344 }, { "epoch": 2.9267241379310347, "grad_norm": 2.4447104930877686, "learning_rate": 1.6798616636666638e-06, "loss": 1.2519, "step": 37345 }, { "epoch": 2.9268025078369906, "grad_norm": 2.7311785221099854, "learning_rate": 1.679631105201915e-06, "loss": 1.2951, "step": 37346 }, { "epoch": 2.9268808777429465, "grad_norm": 2.270493745803833, "learning_rate": 1.6794005593661433e-06, "loss": 1.3972, "step": 37347 }, { "epoch": 2.926959247648903, "grad_norm": 3.5139706134796143, "learning_rate": 1.6791700261602284e-06, "loss": 1.6125, "step": 37348 }, { "epoch": 2.927037617554859, "grad_norm": 2.6012866497039795, "learning_rate": 1.6789395055850433e-06, "loss": 1.5634, "step": 37349 }, { "epoch": 2.927115987460815, "grad_norm": 2.6241025924682617, "learning_rate": 1.678708997641465e-06, "loss": 1.3252, "step": 37350 }, { "epoch": 2.927194357366771, "grad_norm": 2.492175340652466, "learning_rate": 1.6784785023303734e-06, "loss": 1.177, "step": 37351 }, { "epoch": 2.9272727272727272, "grad_norm": 2.3702266216278076, "learning_rate": 1.6782480196526451e-06, "loss": 1.4883, "step": 37352 }, { "epoch": 2.9273510971786836, "grad_norm": 2.3135323524475098, "learning_rate": 1.6780175496091534e-06, "loss": 0.9035, "step": 37353 }, { "epoch": 2.9274294670846395, "grad_norm": 2.991182565689087, "learning_rate": 1.6777870922007767e-06, "loss": 1.2065, "step": 37354 }, { "epoch": 2.9275078369905954, "grad_norm": 3.6815004348754883, "learning_rate": 1.6775566474283899e-06, "loss": 1.2348, "step": 37355 }, { "epoch": 2.9275862068965517, "grad_norm": 2.2656311988830566, "learning_rate": 1.6773262152928743e-06, "loss": 1.2466, "step": 37356 }, { "epoch": 2.927664576802508, "grad_norm": 3.255106210708618, "learning_rate": 1.6770957957951017e-06, "loss": 0.9263, "step": 37357 }, { "epoch": 2.927742946708464, "grad_norm": 2.6124327182769775, "learning_rate": 1.6768653889359498e-06, "loss": 1.3368, "step": 37358 }, { "epoch": 2.92782131661442, "grad_norm": 2.2094109058380127, "learning_rate": 1.6766349947162952e-06, "loss": 1.2419, "step": 37359 }, { "epoch": 2.927899686520376, "grad_norm": 5.973074436187744, "learning_rate": 1.6764046131370142e-06, "loss": 0.9857, "step": 37360 }, { "epoch": 2.9279780564263325, "grad_norm": 3.3905091285705566, "learning_rate": 1.6761742441989825e-06, "loss": 1.3665, "step": 37361 }, { "epoch": 2.9280564263322884, "grad_norm": 2.8708643913269043, "learning_rate": 1.6759438879030777e-06, "loss": 1.5617, "step": 37362 }, { "epoch": 2.9281347962382442, "grad_norm": 3.4057517051696777, "learning_rate": 1.6757135442501753e-06, "loss": 1.4478, "step": 37363 }, { "epoch": 2.9282131661442006, "grad_norm": 2.85115385055542, "learning_rate": 1.6754832132411487e-06, "loss": 1.1272, "step": 37364 }, { "epoch": 2.928291536050157, "grad_norm": 6.022391319274902, "learning_rate": 1.6752528948768782e-06, "loss": 1.5757, "step": 37365 }, { "epoch": 2.928369905956113, "grad_norm": 3.882870674133301, "learning_rate": 1.6750225891582371e-06, "loss": 1.1733, "step": 37366 }, { "epoch": 2.928448275862069, "grad_norm": 2.46844220161438, "learning_rate": 1.6747922960861045e-06, "loss": 0.9647, "step": 37367 }, { "epoch": 2.928526645768025, "grad_norm": 2.2502810955047607, "learning_rate": 1.674562015661353e-06, "loss": 1.7695, "step": 37368 }, { "epoch": 2.9286050156739813, "grad_norm": 4.844027996063232, "learning_rate": 1.674331747884857e-06, "loss": 1.31, "step": 37369 }, { "epoch": 2.9286833855799372, "grad_norm": 12.486551284790039, "learning_rate": 1.6741014927574966e-06, "loss": 0.9655, "step": 37370 }, { "epoch": 2.9287617554858936, "grad_norm": 2.407062292098999, "learning_rate": 1.6738712502801473e-06, "loss": 1.4617, "step": 37371 }, { "epoch": 2.9288401253918495, "grad_norm": 2.855977773666382, "learning_rate": 1.6736410204536824e-06, "loss": 1.4358, "step": 37372 }, { "epoch": 2.928918495297806, "grad_norm": 6.863988876342773, "learning_rate": 1.6734108032789782e-06, "loss": 0.9916, "step": 37373 }, { "epoch": 2.9289968652037617, "grad_norm": 2.901055335998535, "learning_rate": 1.6731805987569106e-06, "loss": 1.4125, "step": 37374 }, { "epoch": 2.929075235109718, "grad_norm": 4.573303699493408, "learning_rate": 1.672950406888355e-06, "loss": 2.4771, "step": 37375 }, { "epoch": 2.929153605015674, "grad_norm": 3.471649408340454, "learning_rate": 1.6727202276741877e-06, "loss": 1.5756, "step": 37376 }, { "epoch": 2.9292319749216302, "grad_norm": 3.10752272605896, "learning_rate": 1.6724900611152834e-06, "loss": 1.6303, "step": 37377 }, { "epoch": 2.929310344827586, "grad_norm": 3.593846082687378, "learning_rate": 1.6722599072125178e-06, "loss": 1.0576, "step": 37378 }, { "epoch": 2.9293887147335425, "grad_norm": 3.582454204559326, "learning_rate": 1.672029765966766e-06, "loss": 1.3665, "step": 37379 }, { "epoch": 2.9294670846394983, "grad_norm": 3.2745957374572754, "learning_rate": 1.6717996373789042e-06, "loss": 1.3756, "step": 37380 }, { "epoch": 2.9295454545454547, "grad_norm": 3.656080722808838, "learning_rate": 1.6715695214498068e-06, "loss": 1.0236, "step": 37381 }, { "epoch": 2.9296238244514106, "grad_norm": 2.5394084453582764, "learning_rate": 1.6713394181803495e-06, "loss": 1.1969, "step": 37382 }, { "epoch": 2.929702194357367, "grad_norm": 4.4976701736450195, "learning_rate": 1.6711093275714068e-06, "loss": 1.556, "step": 37383 }, { "epoch": 2.929780564263323, "grad_norm": 4.942013263702393, "learning_rate": 1.6708792496238552e-06, "loss": 0.3445, "step": 37384 }, { "epoch": 2.929858934169279, "grad_norm": 2.7973878383636475, "learning_rate": 1.6706491843385703e-06, "loss": 1.2182, "step": 37385 }, { "epoch": 2.929937304075235, "grad_norm": 2.23734450340271, "learning_rate": 1.6704191317164236e-06, "loss": 0.9692, "step": 37386 }, { "epoch": 2.9300156739811913, "grad_norm": 2.214184522628784, "learning_rate": 1.670189091758293e-06, "loss": 1.172, "step": 37387 }, { "epoch": 2.9300940438871472, "grad_norm": 15.812883377075195, "learning_rate": 1.669959064465051e-06, "loss": 1.3856, "step": 37388 }, { "epoch": 2.9301724137931036, "grad_norm": 2.586294174194336, "learning_rate": 1.6697290498375778e-06, "loss": 1.3937, "step": 37389 }, { "epoch": 2.9302507836990594, "grad_norm": 2.6400153636932373, "learning_rate": 1.6694990478767432e-06, "loss": 1.029, "step": 37390 }, { "epoch": 2.930329153605016, "grad_norm": 3.239783763885498, "learning_rate": 1.6692690585834232e-06, "loss": 1.554, "step": 37391 }, { "epoch": 2.9304075235109717, "grad_norm": 4.321605682373047, "learning_rate": 1.6690390819584928e-06, "loss": 1.4858, "step": 37392 }, { "epoch": 2.930485893416928, "grad_norm": 8.67381477355957, "learning_rate": 1.668809118002827e-06, "loss": 1.3528, "step": 37393 }, { "epoch": 2.930564263322884, "grad_norm": 3.0551459789276123, "learning_rate": 1.6685791667173001e-06, "loss": 1.204, "step": 37394 }, { "epoch": 2.93064263322884, "grad_norm": 3.3142073154449463, "learning_rate": 1.6683492281027875e-06, "loss": 1.2378, "step": 37395 }, { "epoch": 2.930721003134796, "grad_norm": 4.480724334716797, "learning_rate": 1.6681193021601628e-06, "loss": 1.266, "step": 37396 }, { "epoch": 2.9307993730407524, "grad_norm": 3.278681516647339, "learning_rate": 1.6678893888903009e-06, "loss": 1.5483, "step": 37397 }, { "epoch": 2.9308777429467083, "grad_norm": 3.559938669204712, "learning_rate": 1.6676594882940766e-06, "loss": 1.5958, "step": 37398 }, { "epoch": 2.9309561128526647, "grad_norm": 3.7952723503112793, "learning_rate": 1.6674296003723639e-06, "loss": 1.4589, "step": 37399 }, { "epoch": 2.9310344827586206, "grad_norm": 2.3985376358032227, "learning_rate": 1.6671997251260386e-06, "loss": 1.1421, "step": 37400 }, { "epoch": 2.931112852664577, "grad_norm": 3.2760636806488037, "learning_rate": 1.6669698625559704e-06, "loss": 1.0408, "step": 37401 }, { "epoch": 2.9311912225705328, "grad_norm": 4.155755996704102, "learning_rate": 1.666740012663039e-06, "loss": 1.2006, "step": 37402 }, { "epoch": 2.931269592476489, "grad_norm": 2.198586940765381, "learning_rate": 1.666510175448116e-06, "loss": 0.9835, "step": 37403 }, { "epoch": 2.931347962382445, "grad_norm": 4.218274116516113, "learning_rate": 1.6662803509120784e-06, "loss": 1.2781, "step": 37404 }, { "epoch": 2.9314263322884013, "grad_norm": 2.557987928390503, "learning_rate": 1.6660505390557958e-06, "loss": 1.4568, "step": 37405 }, { "epoch": 2.931504702194357, "grad_norm": 3.2645034790039062, "learning_rate": 1.6658207398801434e-06, "loss": 1.058, "step": 37406 }, { "epoch": 2.9315830721003135, "grad_norm": 3.8568456172943115, "learning_rate": 1.6655909533859977e-06, "loss": 1.0692, "step": 37407 }, { "epoch": 2.9316614420062694, "grad_norm": 4.1050124168396, "learning_rate": 1.6653611795742331e-06, "loss": 1.2382, "step": 37408 }, { "epoch": 2.9317398119122258, "grad_norm": 4.295456886291504, "learning_rate": 1.6651314184457202e-06, "loss": 0.8317, "step": 37409 }, { "epoch": 2.9318181818181817, "grad_norm": 3.515639305114746, "learning_rate": 1.6649016700013337e-06, "loss": 1.5672, "step": 37410 }, { "epoch": 2.931896551724138, "grad_norm": 2.5971505641937256, "learning_rate": 1.6646719342419475e-06, "loss": 1.7007, "step": 37411 }, { "epoch": 2.931974921630094, "grad_norm": 3.4590651988983154, "learning_rate": 1.6644422111684393e-06, "loss": 1.0183, "step": 37412 }, { "epoch": 2.93205329153605, "grad_norm": 3.826972007751465, "learning_rate": 1.664212500781678e-06, "loss": 1.2214, "step": 37413 }, { "epoch": 2.932131661442006, "grad_norm": 2.6726837158203125, "learning_rate": 1.663982803082539e-06, "loss": 1.445, "step": 37414 }, { "epoch": 2.9322100313479624, "grad_norm": 3.1202926635742188, "learning_rate": 1.6637531180718951e-06, "loss": 1.4663, "step": 37415 }, { "epoch": 2.9322884012539188, "grad_norm": 3.5345654487609863, "learning_rate": 1.6635234457506216e-06, "loss": 1.5768, "step": 37416 }, { "epoch": 2.9323667711598747, "grad_norm": 11.019415855407715, "learning_rate": 1.6632937861195915e-06, "loss": 1.4789, "step": 37417 }, { "epoch": 2.9324451410658305, "grad_norm": 6.638132095336914, "learning_rate": 1.663064139179677e-06, "loss": 1.3575, "step": 37418 }, { "epoch": 2.932523510971787, "grad_norm": 3.418137550354004, "learning_rate": 1.662834504931755e-06, "loss": 1.6502, "step": 37419 }, { "epoch": 2.932601880877743, "grad_norm": 5.256282806396484, "learning_rate": 1.6626048833766927e-06, "loss": 1.3507, "step": 37420 }, { "epoch": 2.932680250783699, "grad_norm": 3.9787521362304688, "learning_rate": 1.6623752745153692e-06, "loss": 1.465, "step": 37421 }, { "epoch": 2.932758620689655, "grad_norm": 5.217483997344971, "learning_rate": 1.6621456783486555e-06, "loss": 0.994, "step": 37422 }, { "epoch": 2.9328369905956113, "grad_norm": 3.048743963241577, "learning_rate": 1.661916094877427e-06, "loss": 1.2601, "step": 37423 }, { "epoch": 2.9329153605015676, "grad_norm": 2.7473034858703613, "learning_rate": 1.6616865241025538e-06, "loss": 1.5093, "step": 37424 }, { "epoch": 2.9329937304075235, "grad_norm": 3.14546275138855, "learning_rate": 1.6614569660249085e-06, "loss": 1.3166, "step": 37425 }, { "epoch": 2.9330721003134794, "grad_norm": 3.1880362033843994, "learning_rate": 1.6612274206453678e-06, "loss": 0.9664, "step": 37426 }, { "epoch": 2.9331504702194358, "grad_norm": 3.011699676513672, "learning_rate": 1.6609978879648047e-06, "loss": 1.3525, "step": 37427 }, { "epoch": 2.933228840125392, "grad_norm": 4.161494255065918, "learning_rate": 1.6607683679840892e-06, "loss": 1.3313, "step": 37428 }, { "epoch": 2.933307210031348, "grad_norm": 2.6481614112854004, "learning_rate": 1.660538860704095e-06, "loss": 1.0395, "step": 37429 }, { "epoch": 2.933385579937304, "grad_norm": 3.0577473640441895, "learning_rate": 1.6603093661256964e-06, "loss": 1.4422, "step": 37430 }, { "epoch": 2.93346394984326, "grad_norm": 5.510390281677246, "learning_rate": 1.6600798842497655e-06, "loss": 1.4265, "step": 37431 }, { "epoch": 2.9335423197492165, "grad_norm": 2.7903835773468018, "learning_rate": 1.6598504150771749e-06, "loss": 1.4689, "step": 37432 }, { "epoch": 2.9336206896551724, "grad_norm": 3.624391555786133, "learning_rate": 1.6596209586087981e-06, "loss": 1.3986, "step": 37433 }, { "epoch": 2.9336990595611283, "grad_norm": 3.282493829727173, "learning_rate": 1.6593915148455075e-06, "loss": 1.3831, "step": 37434 }, { "epoch": 2.9337774294670846, "grad_norm": 3.4707891941070557, "learning_rate": 1.659162083788175e-06, "loss": 1.3897, "step": 37435 }, { "epoch": 2.933855799373041, "grad_norm": 3.5385375022888184, "learning_rate": 1.6589326654376748e-06, "loss": 1.4859, "step": 37436 }, { "epoch": 2.933934169278997, "grad_norm": 3.1545650959014893, "learning_rate": 1.6587032597948781e-06, "loss": 1.2333, "step": 37437 }, { "epoch": 2.9340125391849528, "grad_norm": 2.7100257873535156, "learning_rate": 1.6584738668606597e-06, "loss": 1.4378, "step": 37438 }, { "epoch": 2.934090909090909, "grad_norm": 4.498293876647949, "learning_rate": 1.6582444866358876e-06, "loss": 1.3755, "step": 37439 }, { "epoch": 2.9341692789968654, "grad_norm": 2.223078727722168, "learning_rate": 1.6580151191214382e-06, "loss": 1.6047, "step": 37440 }, { "epoch": 2.9342476489028213, "grad_norm": 2.3652470111846924, "learning_rate": 1.657785764318185e-06, "loss": 1.2829, "step": 37441 }, { "epoch": 2.934326018808777, "grad_norm": 2.732656717300415, "learning_rate": 1.6575564222269958e-06, "loss": 1.1108, "step": 37442 }, { "epoch": 2.9344043887147335, "grad_norm": 4.244478702545166, "learning_rate": 1.657327092848745e-06, "loss": 1.3927, "step": 37443 }, { "epoch": 2.93448275862069, "grad_norm": 8.139344215393066, "learning_rate": 1.6570977761843037e-06, "loss": 1.4354, "step": 37444 }, { "epoch": 2.9345611285266457, "grad_norm": 4.970917224884033, "learning_rate": 1.656868472234549e-06, "loss": 1.1749, "step": 37445 }, { "epoch": 2.9346394984326016, "grad_norm": 9.350483894348145, "learning_rate": 1.656639181000348e-06, "loss": 1.7435, "step": 37446 }, { "epoch": 2.934717868338558, "grad_norm": 3.8036048412323, "learning_rate": 1.6564099024825736e-06, "loss": 1.2772, "step": 37447 }, { "epoch": 2.9347962382445143, "grad_norm": 2.5854945182800293, "learning_rate": 1.6561806366820987e-06, "loss": 1.4513, "step": 37448 }, { "epoch": 2.93487460815047, "grad_norm": 5.089188098907471, "learning_rate": 1.6559513835997953e-06, "loss": 0.6742, "step": 37449 }, { "epoch": 2.934952978056426, "grad_norm": 7.213987350463867, "learning_rate": 1.6557221432365355e-06, "loss": 1.7677, "step": 37450 }, { "epoch": 2.9350313479623824, "grad_norm": 3.026948928833008, "learning_rate": 1.6554929155931908e-06, "loss": 1.3277, "step": 37451 }, { "epoch": 2.9351097178683387, "grad_norm": 3.3254969120025635, "learning_rate": 1.655263700670633e-06, "loss": 1.6524, "step": 37452 }, { "epoch": 2.9351880877742946, "grad_norm": 3.199005603790283, "learning_rate": 1.6550344984697342e-06, "loss": 1.4508, "step": 37453 }, { "epoch": 2.9352664576802505, "grad_norm": 3.4833197593688965, "learning_rate": 1.6548053089913663e-06, "loss": 1.4931, "step": 37454 }, { "epoch": 2.935344827586207, "grad_norm": 3.9766414165496826, "learning_rate": 1.6545761322364006e-06, "loss": 1.2761, "step": 37455 }, { "epoch": 2.935423197492163, "grad_norm": 3.2530078887939453, "learning_rate": 1.6543469682057105e-06, "loss": 0.9948, "step": 37456 }, { "epoch": 2.935501567398119, "grad_norm": 3.216801404953003, "learning_rate": 1.6541178169001648e-06, "loss": 1.4139, "step": 37457 }, { "epoch": 2.935579937304075, "grad_norm": 2.753014326095581, "learning_rate": 1.6538886783206343e-06, "loss": 1.3138, "step": 37458 }, { "epoch": 2.9356583072100313, "grad_norm": 2.813815116882324, "learning_rate": 1.6536595524679943e-06, "loss": 1.4574, "step": 37459 }, { "epoch": 2.9357366771159876, "grad_norm": 2.3372154235839844, "learning_rate": 1.6534304393431166e-06, "loss": 1.2828, "step": 37460 }, { "epoch": 2.9358150470219435, "grad_norm": 6.362898349761963, "learning_rate": 1.6532013389468688e-06, "loss": 0.8458, "step": 37461 }, { "epoch": 2.9358934169279, "grad_norm": 3.5836634635925293, "learning_rate": 1.6529722512801239e-06, "loss": 1.2503, "step": 37462 }, { "epoch": 2.9359717868338557, "grad_norm": 5.049005031585693, "learning_rate": 1.6527431763437518e-06, "loss": 1.3417, "step": 37463 }, { "epoch": 2.936050156739812, "grad_norm": 4.224188804626465, "learning_rate": 1.6525141141386286e-06, "loss": 1.4176, "step": 37464 }, { "epoch": 2.936128526645768, "grad_norm": 3.1269192695617676, "learning_rate": 1.6522850646656208e-06, "loss": 1.6914, "step": 37465 }, { "epoch": 2.9362068965517243, "grad_norm": 3.2557148933410645, "learning_rate": 1.6520560279256015e-06, "loss": 1.7188, "step": 37466 }, { "epoch": 2.93628526645768, "grad_norm": 16.827512741088867, "learning_rate": 1.6518270039194412e-06, "loss": 1.575, "step": 37467 }, { "epoch": 2.9363636363636365, "grad_norm": 4.121273040771484, "learning_rate": 1.6515979926480108e-06, "loss": 1.4967, "step": 37468 }, { "epoch": 2.9364420062695924, "grad_norm": 3.052827835083008, "learning_rate": 1.6513689941121825e-06, "loss": 1.1756, "step": 37469 }, { "epoch": 2.9365203761755487, "grad_norm": 1.9415565729141235, "learning_rate": 1.651140008312826e-06, "loss": 1.217, "step": 37470 }, { "epoch": 2.9365987460815046, "grad_norm": 3.6312811374664307, "learning_rate": 1.6509110352508146e-06, "loss": 1.1882, "step": 37471 }, { "epoch": 2.936677115987461, "grad_norm": 11.046438217163086, "learning_rate": 1.6506820749270148e-06, "loss": 1.371, "step": 37472 }, { "epoch": 2.936755485893417, "grad_norm": 4.047389984130859, "learning_rate": 1.6504531273423009e-06, "loss": 1.3317, "step": 37473 }, { "epoch": 2.936833855799373, "grad_norm": 3.690542459487915, "learning_rate": 1.650224192497543e-06, "loss": 0.8252, "step": 37474 }, { "epoch": 2.936912225705329, "grad_norm": 3.483151912689209, "learning_rate": 1.6499952703936133e-06, "loss": 1.4156, "step": 37475 }, { "epoch": 2.9369905956112854, "grad_norm": 8.198203086853027, "learning_rate": 1.6497663610313797e-06, "loss": 0.7233, "step": 37476 }, { "epoch": 2.9370689655172413, "grad_norm": 8.096839904785156, "learning_rate": 1.6495374644117124e-06, "loss": 1.6641, "step": 37477 }, { "epoch": 2.9371473354231976, "grad_norm": 1.9775863885879517, "learning_rate": 1.6493085805354847e-06, "loss": 1.2524, "step": 37478 }, { "epoch": 2.9372257053291535, "grad_norm": 8.710074424743652, "learning_rate": 1.6490797094035682e-06, "loss": 1.8451, "step": 37479 }, { "epoch": 2.93730407523511, "grad_norm": 3.89603590965271, "learning_rate": 1.64885085101683e-06, "loss": 0.889, "step": 37480 }, { "epoch": 2.9373824451410657, "grad_norm": 3.0640881061553955, "learning_rate": 1.6486220053761416e-06, "loss": 1.5968, "step": 37481 }, { "epoch": 2.937460815047022, "grad_norm": 4.164112091064453, "learning_rate": 1.648393172482372e-06, "loss": 1.574, "step": 37482 }, { "epoch": 2.937539184952978, "grad_norm": 2.9982333183288574, "learning_rate": 1.6481643523363966e-06, "loss": 1.4632, "step": 37483 }, { "epoch": 2.9376175548589343, "grad_norm": 5.850210189819336, "learning_rate": 1.647935544939081e-06, "loss": 1.1238, "step": 37484 }, { "epoch": 2.93769592476489, "grad_norm": 2.7501261234283447, "learning_rate": 1.6477067502912964e-06, "loss": 1.2546, "step": 37485 }, { "epoch": 2.9377742946708465, "grad_norm": 2.1292407512664795, "learning_rate": 1.6474779683939135e-06, "loss": 1.2416, "step": 37486 }, { "epoch": 2.9378526645768024, "grad_norm": 3.6034417152404785, "learning_rate": 1.6472491992478028e-06, "loss": 1.3816, "step": 37487 }, { "epoch": 2.9379310344827587, "grad_norm": 2.6842854022979736, "learning_rate": 1.6470204428538339e-06, "loss": 1.237, "step": 37488 }, { "epoch": 2.9380094043887146, "grad_norm": 9.95057487487793, "learning_rate": 1.6467916992128775e-06, "loss": 1.0067, "step": 37489 }, { "epoch": 2.938087774294671, "grad_norm": 6.75637674331665, "learning_rate": 1.6465629683258023e-06, "loss": 1.672, "step": 37490 }, { "epoch": 2.938166144200627, "grad_norm": 2.9638776779174805, "learning_rate": 1.6463342501934799e-06, "loss": 1.3477, "step": 37491 }, { "epoch": 2.938244514106583, "grad_norm": 10.161806106567383, "learning_rate": 1.646105544816779e-06, "loss": 1.5847, "step": 37492 }, { "epoch": 2.938322884012539, "grad_norm": 2.2324490547180176, "learning_rate": 1.6458768521965702e-06, "loss": 1.4059, "step": 37493 }, { "epoch": 2.9384012539184954, "grad_norm": 4.21634578704834, "learning_rate": 1.6456481723337242e-06, "loss": 1.5889, "step": 37494 }, { "epoch": 2.9384796238244513, "grad_norm": 3.949807643890381, "learning_rate": 1.6454195052291066e-06, "loss": 1.8279, "step": 37495 }, { "epoch": 2.9385579937304076, "grad_norm": 3.9868035316467285, "learning_rate": 1.6451908508835924e-06, "loss": 1.3902, "step": 37496 }, { "epoch": 2.9386363636363635, "grad_norm": 6.489319324493408, "learning_rate": 1.6449622092980499e-06, "loss": 1.0409, "step": 37497 }, { "epoch": 2.93871473354232, "grad_norm": 2.9489853382110596, "learning_rate": 1.6447335804733467e-06, "loss": 1.0929, "step": 37498 }, { "epoch": 2.9387931034482757, "grad_norm": 24.406417846679688, "learning_rate": 1.6445049644103533e-06, "loss": 0.9188, "step": 37499 }, { "epoch": 2.938871473354232, "grad_norm": 3.963009834289551, "learning_rate": 1.6442763611099383e-06, "loss": 1.4713, "step": 37500 }, { "epoch": 2.938949843260188, "grad_norm": 3.620798349380493, "learning_rate": 1.6440477705729758e-06, "loss": 1.3373, "step": 37501 }, { "epoch": 2.9390282131661443, "grad_norm": 3.4911699295043945, "learning_rate": 1.6438191928003294e-06, "loss": 1.1959, "step": 37502 }, { "epoch": 2.9391065830721, "grad_norm": 2.4889113903045654, "learning_rate": 1.6435906277928716e-06, "loss": 1.4067, "step": 37503 }, { "epoch": 2.9391849529780565, "grad_norm": 3.744530200958252, "learning_rate": 1.6433620755514712e-06, "loss": 1.3452, "step": 37504 }, { "epoch": 2.9392633228840124, "grad_norm": 4.458374500274658, "learning_rate": 1.6431335360769969e-06, "loss": 1.4595, "step": 37505 }, { "epoch": 2.9393416927899687, "grad_norm": 2.20375657081604, "learning_rate": 1.6429050093703186e-06, "loss": 1.3152, "step": 37506 }, { "epoch": 2.939420062695925, "grad_norm": 2.561387538909912, "learning_rate": 1.6426764954323054e-06, "loss": 1.0002, "step": 37507 }, { "epoch": 2.939498432601881, "grad_norm": 2.653949499130249, "learning_rate": 1.6424479942638266e-06, "loss": 1.3292, "step": 37508 }, { "epoch": 2.939576802507837, "grad_norm": 2.7641685009002686, "learning_rate": 1.642219505865751e-06, "loss": 1.4567, "step": 37509 }, { "epoch": 2.939655172413793, "grad_norm": 2.518109083175659, "learning_rate": 1.6419910302389475e-06, "loss": 1.2805, "step": 37510 }, { "epoch": 2.9397335423197495, "grad_norm": 15.388848304748535, "learning_rate": 1.6417625673842852e-06, "loss": 1.8254, "step": 37511 }, { "epoch": 2.9398119122257054, "grad_norm": 2.57601261138916, "learning_rate": 1.641534117302635e-06, "loss": 1.2469, "step": 37512 }, { "epoch": 2.9398902821316613, "grad_norm": 2.6662068367004395, "learning_rate": 1.6413056799948622e-06, "loss": 0.7446, "step": 37513 }, { "epoch": 2.9399686520376176, "grad_norm": 3.7293436527252197, "learning_rate": 1.6410772554618365e-06, "loss": 1.4711, "step": 37514 }, { "epoch": 2.940047021943574, "grad_norm": 4.374139308929443, "learning_rate": 1.6408488437044285e-06, "loss": 1.3632, "step": 37515 }, { "epoch": 2.94012539184953, "grad_norm": 2.3586814403533936, "learning_rate": 1.640620444723508e-06, "loss": 1.1806, "step": 37516 }, { "epoch": 2.9402037617554857, "grad_norm": 2.6677908897399902, "learning_rate": 1.6403920585199406e-06, "loss": 1.1764, "step": 37517 }, { "epoch": 2.940282131661442, "grad_norm": 4.69210147857666, "learning_rate": 1.6401636850945957e-06, "loss": 1.3255, "step": 37518 }, { "epoch": 2.9403605015673984, "grad_norm": 2.3579046726226807, "learning_rate": 1.6399353244483413e-06, "loss": 1.2856, "step": 37519 }, { "epoch": 2.9404388714733543, "grad_norm": 4.064826965332031, "learning_rate": 1.63970697658205e-06, "loss": 1.0706, "step": 37520 }, { "epoch": 2.94051724137931, "grad_norm": 5.2663679122924805, "learning_rate": 1.6394786414965858e-06, "loss": 1.2447, "step": 37521 }, { "epoch": 2.9405956112852665, "grad_norm": 3.618222951889038, "learning_rate": 1.6392503191928188e-06, "loss": 1.4566, "step": 37522 }, { "epoch": 2.940673981191223, "grad_norm": 2.672589063644409, "learning_rate": 1.6390220096716175e-06, "loss": 1.3771, "step": 37523 }, { "epoch": 2.9407523510971787, "grad_norm": 3.5983495712280273, "learning_rate": 1.63879371293385e-06, "loss": 1.2522, "step": 37524 }, { "epoch": 2.9408307210031346, "grad_norm": 2.495321035385132, "learning_rate": 1.6385654289803848e-06, "loss": 1.3216, "step": 37525 }, { "epoch": 2.940909090909091, "grad_norm": 3.721912384033203, "learning_rate": 1.6383371578120906e-06, "loss": 1.4628, "step": 37526 }, { "epoch": 2.9409874608150472, "grad_norm": 3.557081460952759, "learning_rate": 1.638108899429836e-06, "loss": 0.6431, "step": 37527 }, { "epoch": 2.941065830721003, "grad_norm": 3.270824670791626, "learning_rate": 1.6378806538344855e-06, "loss": 1.5281, "step": 37528 }, { "epoch": 2.941144200626959, "grad_norm": 2.291745662689209, "learning_rate": 1.6376524210269118e-06, "loss": 1.2409, "step": 37529 }, { "epoch": 2.9412225705329154, "grad_norm": 2.784715175628662, "learning_rate": 1.6374242010079816e-06, "loss": 1.4672, "step": 37530 }, { "epoch": 2.9413009404388717, "grad_norm": 7.454466819763184, "learning_rate": 1.6371959937785636e-06, "loss": 1.2797, "step": 37531 }, { "epoch": 2.9413793103448276, "grad_norm": 4.499866008758545, "learning_rate": 1.636967799339524e-06, "loss": 1.0557, "step": 37532 }, { "epoch": 2.9414576802507835, "grad_norm": 2.5555927753448486, "learning_rate": 1.6367396176917288e-06, "loss": 1.2459, "step": 37533 }, { "epoch": 2.94153605015674, "grad_norm": 5.5904412269592285, "learning_rate": 1.636511448836051e-06, "loss": 1.7196, "step": 37534 }, { "epoch": 2.941614420062696, "grad_norm": 2.766866683959961, "learning_rate": 1.6362832927733575e-06, "loss": 1.0397, "step": 37535 }, { "epoch": 2.941692789968652, "grad_norm": 2.4765024185180664, "learning_rate": 1.6360551495045129e-06, "loss": 0.9904, "step": 37536 }, { "epoch": 2.941771159874608, "grad_norm": 3.1488866806030273, "learning_rate": 1.635827019030387e-06, "loss": 0.8629, "step": 37537 }, { "epoch": 2.9418495297805642, "grad_norm": 2.786283254623413, "learning_rate": 1.6355989013518469e-06, "loss": 1.0407, "step": 37538 }, { "epoch": 2.9419278996865206, "grad_norm": 2.3599655628204346, "learning_rate": 1.6353707964697606e-06, "loss": 1.3254, "step": 37539 }, { "epoch": 2.9420062695924765, "grad_norm": 3.0451393127441406, "learning_rate": 1.6351427043849955e-06, "loss": 1.3132, "step": 37540 }, { "epoch": 2.9420846394984324, "grad_norm": 2.231574058532715, "learning_rate": 1.6349146250984193e-06, "loss": 1.2458, "step": 37541 }, { "epoch": 2.9421630094043887, "grad_norm": 5.058129787445068, "learning_rate": 1.6346865586108996e-06, "loss": 1.4716, "step": 37542 }, { "epoch": 2.942241379310345, "grad_norm": 3.2262978553771973, "learning_rate": 1.6344585049233036e-06, "loss": 1.0874, "step": 37543 }, { "epoch": 2.942319749216301, "grad_norm": 3.573645830154419, "learning_rate": 1.6342304640364986e-06, "loss": 1.3451, "step": 37544 }, { "epoch": 2.942398119122257, "grad_norm": 2.6100828647613525, "learning_rate": 1.6340024359513522e-06, "loss": 1.4904, "step": 37545 }, { "epoch": 2.942476489028213, "grad_norm": 3.184178113937378, "learning_rate": 1.6337744206687339e-06, "loss": 1.6605, "step": 37546 }, { "epoch": 2.9425548589341695, "grad_norm": 7.103452682495117, "learning_rate": 1.6335464181895045e-06, "loss": 1.7429, "step": 37547 }, { "epoch": 2.9426332288401253, "grad_norm": 2.8533692359924316, "learning_rate": 1.6333184285145376e-06, "loss": 1.2193, "step": 37548 }, { "epoch": 2.9427115987460812, "grad_norm": 18.0147647857666, "learning_rate": 1.6330904516446982e-06, "loss": 0.5087, "step": 37549 }, { "epoch": 2.9427899686520376, "grad_norm": 4.6810197830200195, "learning_rate": 1.6328624875808546e-06, "loss": 1.6083, "step": 37550 }, { "epoch": 2.942868338557994, "grad_norm": 3.3631081581115723, "learning_rate": 1.6326345363238715e-06, "loss": 1.2508, "step": 37551 }, { "epoch": 2.94294670846395, "grad_norm": 4.124885082244873, "learning_rate": 1.632406597874615e-06, "loss": 1.3608, "step": 37552 }, { "epoch": 2.943025078369906, "grad_norm": 5.086734771728516, "learning_rate": 1.6321786722339579e-06, "loss": 1.3383, "step": 37553 }, { "epoch": 2.943103448275862, "grad_norm": 2.1778197288513184, "learning_rate": 1.6319507594027612e-06, "loss": 1.3825, "step": 37554 }, { "epoch": 2.9431818181818183, "grad_norm": 5.252904891967773, "learning_rate": 1.6317228593818935e-06, "loss": 1.2457, "step": 37555 }, { "epoch": 2.9432601880877742, "grad_norm": 4.389743328094482, "learning_rate": 1.631494972172223e-06, "loss": 1.4578, "step": 37556 }, { "epoch": 2.9433385579937306, "grad_norm": 4.186524868011475, "learning_rate": 1.6312670977746143e-06, "loss": 0.7308, "step": 37557 }, { "epoch": 2.9434169278996865, "grad_norm": 3.242499351501465, "learning_rate": 1.6310392361899363e-06, "loss": 1.293, "step": 37558 }, { "epoch": 2.943495297805643, "grad_norm": 2.68843936920166, "learning_rate": 1.630811387419054e-06, "loss": 1.1636, "step": 37559 }, { "epoch": 2.9435736677115987, "grad_norm": 8.171244621276855, "learning_rate": 1.630583551462835e-06, "loss": 0.7541, "step": 37560 }, { "epoch": 2.943652037617555, "grad_norm": 8.761748313903809, "learning_rate": 1.6303557283221449e-06, "loss": 0.9222, "step": 37561 }, { "epoch": 2.943730407523511, "grad_norm": 6.736028671264648, "learning_rate": 1.6301279179978518e-06, "loss": 1.672, "step": 37562 }, { "epoch": 2.9438087774294672, "grad_norm": 6.743807792663574, "learning_rate": 1.6299001204908205e-06, "loss": 1.1714, "step": 37563 }, { "epoch": 2.943887147335423, "grad_norm": 3.850264310836792, "learning_rate": 1.6296723358019184e-06, "loss": 1.4878, "step": 37564 }, { "epoch": 2.9439655172413794, "grad_norm": 3.2451021671295166, "learning_rate": 1.6294445639320134e-06, "loss": 1.3333, "step": 37565 }, { "epoch": 2.9440438871473353, "grad_norm": 5.430039882659912, "learning_rate": 1.6292168048819667e-06, "loss": 1.1554, "step": 37566 }, { "epoch": 2.9441222570532917, "grad_norm": 2.5979318618774414, "learning_rate": 1.6289890586526492e-06, "loss": 1.6024, "step": 37567 }, { "epoch": 2.9442006269592476, "grad_norm": 16.509475708007812, "learning_rate": 1.628761325244928e-06, "loss": 2.2662, "step": 37568 }, { "epoch": 2.944278996865204, "grad_norm": 5.558356285095215, "learning_rate": 1.6285336046596651e-06, "loss": 1.1845, "step": 37569 }, { "epoch": 2.94435736677116, "grad_norm": 3.533716917037964, "learning_rate": 1.6283058968977289e-06, "loss": 1.5478, "step": 37570 }, { "epoch": 2.944435736677116, "grad_norm": 4.658875465393066, "learning_rate": 1.6280782019599839e-06, "loss": 1.3495, "step": 37571 }, { "epoch": 2.944514106583072, "grad_norm": 6.499171733856201, "learning_rate": 1.6278505198473005e-06, "loss": 1.0639, "step": 37572 }, { "epoch": 2.9445924764890283, "grad_norm": 5.4571428298950195, "learning_rate": 1.6276228505605401e-06, "loss": 1.6849, "step": 37573 }, { "epoch": 2.944670846394984, "grad_norm": 2.441241502761841, "learning_rate": 1.62739519410057e-06, "loss": 1.4488, "step": 37574 }, { "epoch": 2.9447492163009406, "grad_norm": 4.1176228523254395, "learning_rate": 1.6271675504682566e-06, "loss": 1.2876, "step": 37575 }, { "epoch": 2.9448275862068964, "grad_norm": 3.5847814083099365, "learning_rate": 1.6269399196644646e-06, "loss": 1.4526, "step": 37576 }, { "epoch": 2.9449059561128528, "grad_norm": 2.9730148315429688, "learning_rate": 1.6267123016900615e-06, "loss": 1.4389, "step": 37577 }, { "epoch": 2.9449843260188087, "grad_norm": 6.901882648468018, "learning_rate": 1.6264846965459125e-06, "loss": 1.7297, "step": 37578 }, { "epoch": 2.945062695924765, "grad_norm": 4.015230178833008, "learning_rate": 1.6262571042328822e-06, "loss": 1.3955, "step": 37579 }, { "epoch": 2.945141065830721, "grad_norm": 3.627516746520996, "learning_rate": 1.6260295247518372e-06, "loss": 1.0816, "step": 37580 }, { "epoch": 2.945219435736677, "grad_norm": 3.837040901184082, "learning_rate": 1.6258019581036428e-06, "loss": 1.1175, "step": 37581 }, { "epoch": 2.945297805642633, "grad_norm": 2.0942139625549316, "learning_rate": 1.6255744042891646e-06, "loss": 0.7279, "step": 37582 }, { "epoch": 2.9453761755485894, "grad_norm": 5.014852523803711, "learning_rate": 1.6253468633092706e-06, "loss": 1.7637, "step": 37583 }, { "epoch": 2.9454545454545453, "grad_norm": 3.960874557495117, "learning_rate": 1.6251193351648214e-06, "loss": 1.4578, "step": 37584 }, { "epoch": 2.9455329153605017, "grad_norm": 6.909992218017578, "learning_rate": 1.6248918198566832e-06, "loss": 1.4258, "step": 37585 }, { "epoch": 2.9456112852664575, "grad_norm": 2.2054505348205566, "learning_rate": 1.6246643173857245e-06, "loss": 1.2621, "step": 37586 }, { "epoch": 2.945689655172414, "grad_norm": 2.9868597984313965, "learning_rate": 1.624436827752811e-06, "loss": 1.351, "step": 37587 }, { "epoch": 2.9457680250783698, "grad_norm": 3.831202745437622, "learning_rate": 1.6242093509588047e-06, "loss": 1.2667, "step": 37588 }, { "epoch": 2.945846394984326, "grad_norm": 4.280874252319336, "learning_rate": 1.6239818870045715e-06, "loss": 1.0107, "step": 37589 }, { "epoch": 2.945924764890282, "grad_norm": 4.389677047729492, "learning_rate": 1.6237544358909758e-06, "loss": 1.027, "step": 37590 }, { "epoch": 2.9460031347962383, "grad_norm": 3.0672965049743652, "learning_rate": 1.6235269976188872e-06, "loss": 1.201, "step": 37591 }, { "epoch": 2.946081504702194, "grad_norm": 2.547607183456421, "learning_rate": 1.6232995721891659e-06, "loss": 1.2721, "step": 37592 }, { "epoch": 2.9461598746081505, "grad_norm": 5.058443546295166, "learning_rate": 1.6230721596026789e-06, "loss": 1.1382, "step": 37593 }, { "epoch": 2.9462382445141064, "grad_norm": 5.638699531555176, "learning_rate": 1.622844759860291e-06, "loss": 1.3607, "step": 37594 }, { "epoch": 2.9463166144200628, "grad_norm": 5.292801856994629, "learning_rate": 1.6226173729628669e-06, "loss": 1.5348, "step": 37595 }, { "epoch": 2.9463949843260187, "grad_norm": 3.979384660720825, "learning_rate": 1.6223899989112713e-06, "loss": 1.2546, "step": 37596 }, { "epoch": 2.946473354231975, "grad_norm": 2.4016034603118896, "learning_rate": 1.6221626377063698e-06, "loss": 1.3717, "step": 37597 }, { "epoch": 2.946551724137931, "grad_norm": 2.710923194885254, "learning_rate": 1.6219352893490264e-06, "loss": 1.1858, "step": 37598 }, { "epoch": 2.946630094043887, "grad_norm": 4.0756964683532715, "learning_rate": 1.621707953840106e-06, "loss": 1.6617, "step": 37599 }, { "epoch": 2.946708463949843, "grad_norm": 2.574733018875122, "learning_rate": 1.621480631180473e-06, "loss": 1.4379, "step": 37600 }, { "epoch": 2.9467868338557994, "grad_norm": 2.5971314907073975, "learning_rate": 1.621253321370993e-06, "loss": 1.4583, "step": 37601 }, { "epoch": 2.9468652037617558, "grad_norm": 3.8441872596740723, "learning_rate": 1.6210260244125315e-06, "loss": 1.6369, "step": 37602 }, { "epoch": 2.9469435736677116, "grad_norm": 2.05220365524292, "learning_rate": 1.6207987403059478e-06, "loss": 1.1907, "step": 37603 }, { "epoch": 2.9470219435736675, "grad_norm": 3.3173646926879883, "learning_rate": 1.6205714690521117e-06, "loss": 1.278, "step": 37604 }, { "epoch": 2.947100313479624, "grad_norm": 4.768283367156982, "learning_rate": 1.6203442106518863e-06, "loss": 1.3146, "step": 37605 }, { "epoch": 2.94717868338558, "grad_norm": 2.731863498687744, "learning_rate": 1.6201169651061372e-06, "loss": 1.459, "step": 37606 }, { "epoch": 2.947257053291536, "grad_norm": 3.39523983001709, "learning_rate": 1.6198897324157254e-06, "loss": 1.1648, "step": 37607 }, { "epoch": 2.947335423197492, "grad_norm": 3.4699089527130127, "learning_rate": 1.6196625125815151e-06, "loss": 0.8541, "step": 37608 }, { "epoch": 2.9474137931034483, "grad_norm": 3.0842432975769043, "learning_rate": 1.619435305604376e-06, "loss": 1.2325, "step": 37609 }, { "epoch": 2.9474921630094046, "grad_norm": 3.569394111633301, "learning_rate": 1.6192081114851665e-06, "loss": 1.2045, "step": 37610 }, { "epoch": 2.9475705329153605, "grad_norm": 3.373232364654541, "learning_rate": 1.6189809302247534e-06, "loss": 1.1704, "step": 37611 }, { "epoch": 2.9476489028213164, "grad_norm": 3.844780683517456, "learning_rate": 1.6187537618240002e-06, "loss": 1.0826, "step": 37612 }, { "epoch": 2.9477272727272728, "grad_norm": 2.708494186401367, "learning_rate": 1.6185266062837706e-06, "loss": 0.9549, "step": 37613 }, { "epoch": 2.947805642633229, "grad_norm": 5.216606616973877, "learning_rate": 1.6182994636049288e-06, "loss": 1.1095, "step": 37614 }, { "epoch": 2.947884012539185, "grad_norm": 7.189995765686035, "learning_rate": 1.6180723337883392e-06, "loss": 1.2365, "step": 37615 }, { "epoch": 2.947962382445141, "grad_norm": 2.7371909618377686, "learning_rate": 1.6178452168348656e-06, "loss": 1.4025, "step": 37616 }, { "epoch": 2.948040752351097, "grad_norm": 1.9836549758911133, "learning_rate": 1.617618112745371e-06, "loss": 1.4255, "step": 37617 }, { "epoch": 2.9481191222570535, "grad_norm": 3.800215482711792, "learning_rate": 1.6173910215207206e-06, "loss": 1.1856, "step": 37618 }, { "epoch": 2.9481974921630094, "grad_norm": 3.2063639163970947, "learning_rate": 1.6171639431617769e-06, "loss": 0.9791, "step": 37619 }, { "epoch": 2.9482758620689653, "grad_norm": 2.7449114322662354, "learning_rate": 1.616936877669404e-06, "loss": 1.2345, "step": 37620 }, { "epoch": 2.9483542319749216, "grad_norm": 3.4881253242492676, "learning_rate": 1.6167098250444674e-06, "loss": 1.3938, "step": 37621 }, { "epoch": 2.948432601880878, "grad_norm": 4.10349178314209, "learning_rate": 1.616482785287825e-06, "loss": 1.4514, "step": 37622 }, { "epoch": 2.948510971786834, "grad_norm": 3.616015672683716, "learning_rate": 1.6162557584003462e-06, "loss": 1.8998, "step": 37623 }, { "epoch": 2.9485893416927897, "grad_norm": 3.8978793621063232, "learning_rate": 1.6160287443828948e-06, "loss": 1.2337, "step": 37624 }, { "epoch": 2.948667711598746, "grad_norm": 4.212775707244873, "learning_rate": 1.6158017432363293e-06, "loss": 1.4159, "step": 37625 }, { "epoch": 2.9487460815047024, "grad_norm": 5.56925630569458, "learning_rate": 1.6155747549615164e-06, "loss": 1.5279, "step": 37626 }, { "epoch": 2.9488244514106583, "grad_norm": 3.512667179107666, "learning_rate": 1.615347779559317e-06, "loss": 1.1117, "step": 37627 }, { "epoch": 2.948902821316614, "grad_norm": 5.093119144439697, "learning_rate": 1.6151208170305993e-06, "loss": 1.386, "step": 37628 }, { "epoch": 2.9489811912225705, "grad_norm": 3.5156021118164062, "learning_rate": 1.6148938673762216e-06, "loss": 1.3185, "step": 37629 }, { "epoch": 2.949059561128527, "grad_norm": 3.6018080711364746, "learning_rate": 1.6146669305970493e-06, "loss": 1.4629, "step": 37630 }, { "epoch": 2.9491379310344827, "grad_norm": 3.6245720386505127, "learning_rate": 1.6144400066939447e-06, "loss": 1.2494, "step": 37631 }, { "epoch": 2.9492163009404386, "grad_norm": 2.9457197189331055, "learning_rate": 1.614213095667772e-06, "loss": 0.9327, "step": 37632 }, { "epoch": 2.949294670846395, "grad_norm": 6.055238723754883, "learning_rate": 1.6139861975193931e-06, "loss": 1.3863, "step": 37633 }, { "epoch": 2.9493730407523513, "grad_norm": 3.365631580352783, "learning_rate": 1.613759312249672e-06, "loss": 1.4646, "step": 37634 }, { "epoch": 2.949451410658307, "grad_norm": 3.3350467681884766, "learning_rate": 1.613532439859471e-06, "loss": 1.3399, "step": 37635 }, { "epoch": 2.949529780564263, "grad_norm": 2.2707886695861816, "learning_rate": 1.6133055803496533e-06, "loss": 1.2401, "step": 37636 }, { "epoch": 2.9496081504702194, "grad_norm": 6.3597283363342285, "learning_rate": 1.6130787337210812e-06, "loss": 0.9827, "step": 37637 }, { "epoch": 2.9496865203761757, "grad_norm": 2.8863112926483154, "learning_rate": 1.6128518999746185e-06, "loss": 1.3022, "step": 37638 }, { "epoch": 2.9497648902821316, "grad_norm": 3.0760750770568848, "learning_rate": 1.6126250791111287e-06, "loss": 1.3014, "step": 37639 }, { "epoch": 2.9498432601880875, "grad_norm": 2.7705941200256348, "learning_rate": 1.6123982711314718e-06, "loss": 1.3477, "step": 37640 }, { "epoch": 2.949921630094044, "grad_norm": 12.625094413757324, "learning_rate": 1.6121714760365104e-06, "loss": 1.3378, "step": 37641 }, { "epoch": 2.95, "grad_norm": 2.2588961124420166, "learning_rate": 1.6119446938271099e-06, "loss": 1.2065, "step": 37642 }, { "epoch": 2.950078369905956, "grad_norm": 2.865981340408325, "learning_rate": 1.6117179245041337e-06, "loss": 1.2044, "step": 37643 }, { "epoch": 2.950156739811912, "grad_norm": 3.5224661827087402, "learning_rate": 1.6114911680684397e-06, "loss": 1.4684, "step": 37644 }, { "epoch": 2.9502351097178683, "grad_norm": 4.48433256149292, "learning_rate": 1.6112644245208935e-06, "loss": 1.3471, "step": 37645 }, { "epoch": 2.9503134796238246, "grad_norm": 3.3133175373077393, "learning_rate": 1.611037693862355e-06, "loss": 1.472, "step": 37646 }, { "epoch": 2.9503918495297805, "grad_norm": 3.468623161315918, "learning_rate": 1.610810976093692e-06, "loss": 1.4124, "step": 37647 }, { "epoch": 2.950470219435737, "grad_norm": 5.412425518035889, "learning_rate": 1.6105842712157616e-06, "loss": 1.4685, "step": 37648 }, { "epoch": 2.9505485893416927, "grad_norm": 5.407719135284424, "learning_rate": 1.6103575792294274e-06, "loss": 1.7525, "step": 37649 }, { "epoch": 2.950626959247649, "grad_norm": 2.8507978916168213, "learning_rate": 1.6101309001355519e-06, "loss": 1.3132, "step": 37650 }, { "epoch": 2.950705329153605, "grad_norm": 3.804274559020996, "learning_rate": 1.6099042339349974e-06, "loss": 1.2937, "step": 37651 }, { "epoch": 2.9507836990595613, "grad_norm": 3.1434547901153564, "learning_rate": 1.6096775806286264e-06, "loss": 1.2626, "step": 37652 }, { "epoch": 2.950862068965517, "grad_norm": 2.363602876663208, "learning_rate": 1.6094509402172997e-06, "loss": 1.4024, "step": 37653 }, { "epoch": 2.9509404388714735, "grad_norm": 3.1384522914886475, "learning_rate": 1.6092243127018825e-06, "loss": 1.5533, "step": 37654 }, { "epoch": 2.9510188087774294, "grad_norm": 2.8188905715942383, "learning_rate": 1.6089976980832306e-06, "loss": 1.3108, "step": 37655 }, { "epoch": 2.9510971786833857, "grad_norm": 2.4414451122283936, "learning_rate": 1.6087710963622115e-06, "loss": 1.3562, "step": 37656 }, { "epoch": 2.9511755485893416, "grad_norm": 4.719689846038818, "learning_rate": 1.6085445075396855e-06, "loss": 1.2679, "step": 37657 }, { "epoch": 2.951253918495298, "grad_norm": 2.1664845943450928, "learning_rate": 1.608317931616516e-06, "loss": 1.3674, "step": 37658 }, { "epoch": 2.951332288401254, "grad_norm": 2.5840611457824707, "learning_rate": 1.6080913685935612e-06, "loss": 0.7817, "step": 37659 }, { "epoch": 2.95141065830721, "grad_norm": 3.3028242588043213, "learning_rate": 1.6078648184716827e-06, "loss": 0.8017, "step": 37660 }, { "epoch": 2.951489028213166, "grad_norm": 3.445629358291626, "learning_rate": 1.607638281251746e-06, "loss": 1.4347, "step": 37661 }, { "epoch": 2.9515673981191224, "grad_norm": 6.3024420738220215, "learning_rate": 1.6074117569346127e-06, "loss": 1.914, "step": 37662 }, { "epoch": 2.9516457680250783, "grad_norm": 5.632287979125977, "learning_rate": 1.60718524552114e-06, "loss": 1.4968, "step": 37663 }, { "epoch": 2.9517241379310346, "grad_norm": 3.613185405731201, "learning_rate": 1.6069587470121928e-06, "loss": 1.3109, "step": 37664 }, { "epoch": 2.9518025078369905, "grad_norm": 2.9183475971221924, "learning_rate": 1.6067322614086316e-06, "loss": 1.1089, "step": 37665 }, { "epoch": 2.951880877742947, "grad_norm": 3.6758313179016113, "learning_rate": 1.6065057887113184e-06, "loss": 1.253, "step": 37666 }, { "epoch": 2.9519592476489027, "grad_norm": 4.244545936584473, "learning_rate": 1.6062793289211131e-06, "loss": 1.257, "step": 37667 }, { "epoch": 2.952037617554859, "grad_norm": 2.5850908756256104, "learning_rate": 1.6060528820388794e-06, "loss": 1.3625, "step": 37668 }, { "epoch": 2.952115987460815, "grad_norm": 10.57862377166748, "learning_rate": 1.6058264480654762e-06, "loss": 0.9466, "step": 37669 }, { "epoch": 2.9521943573667713, "grad_norm": 3.12351393699646, "learning_rate": 1.6056000270017664e-06, "loss": 1.385, "step": 37670 }, { "epoch": 2.952272727272727, "grad_norm": 3.95296311378479, "learning_rate": 1.6053736188486103e-06, "loss": 1.6507, "step": 37671 }, { "epoch": 2.9523510971786835, "grad_norm": 3.0677831172943115, "learning_rate": 1.6051472236068699e-06, "loss": 1.0335, "step": 37672 }, { "epoch": 2.9524294670846394, "grad_norm": 4.320259094238281, "learning_rate": 1.604920841277407e-06, "loss": 1.373, "step": 37673 }, { "epoch": 2.9525078369905957, "grad_norm": 17.33960723876953, "learning_rate": 1.604694471861078e-06, "loss": 1.1956, "step": 37674 }, { "epoch": 2.9525862068965516, "grad_norm": 4.795328617095947, "learning_rate": 1.6044681153587493e-06, "loss": 1.5704, "step": 37675 }, { "epoch": 2.952664576802508, "grad_norm": 3.7879867553710938, "learning_rate": 1.6042417717712794e-06, "loss": 1.2112, "step": 37676 }, { "epoch": 2.952742946708464, "grad_norm": 3.3563103675842285, "learning_rate": 1.6040154410995317e-06, "loss": 1.5057, "step": 37677 }, { "epoch": 2.95282131661442, "grad_norm": 3.7222278118133545, "learning_rate": 1.6037891233443632e-06, "loss": 1.3431, "step": 37678 }, { "epoch": 2.952899686520376, "grad_norm": 3.7229034900665283, "learning_rate": 1.6035628185066343e-06, "loss": 1.2372, "step": 37679 }, { "epoch": 2.9529780564263324, "grad_norm": 2.0708422660827637, "learning_rate": 1.6033365265872118e-06, "loss": 1.3831, "step": 37680 }, { "epoch": 2.9530564263322883, "grad_norm": 3.0715255737304688, "learning_rate": 1.6031102475869508e-06, "loss": 1.6998, "step": 37681 }, { "epoch": 2.9531347962382446, "grad_norm": 2.8172640800476074, "learning_rate": 1.6028839815067137e-06, "loss": 1.5582, "step": 37682 }, { "epoch": 2.9532131661442005, "grad_norm": 4.888399600982666, "learning_rate": 1.602657728347361e-06, "loss": 1.2626, "step": 37683 }, { "epoch": 2.953291536050157, "grad_norm": 2.6992125511169434, "learning_rate": 1.602431488109753e-06, "loss": 1.1192, "step": 37684 }, { "epoch": 2.9533699059561127, "grad_norm": 2.0173702239990234, "learning_rate": 1.6022052607947508e-06, "loss": 1.4787, "step": 37685 }, { "epoch": 2.953448275862069, "grad_norm": 3.444289445877075, "learning_rate": 1.6019790464032149e-06, "loss": 1.4059, "step": 37686 }, { "epoch": 2.953526645768025, "grad_norm": 2.4859721660614014, "learning_rate": 1.6017528449360048e-06, "loss": 0.9627, "step": 37687 }, { "epoch": 2.9536050156739813, "grad_norm": 6.306868553161621, "learning_rate": 1.6015266563939819e-06, "loss": 1.131, "step": 37688 }, { "epoch": 2.953683385579937, "grad_norm": 5.943675994873047, "learning_rate": 1.6013004807780058e-06, "loss": 1.3637, "step": 37689 }, { "epoch": 2.9537617554858935, "grad_norm": 6.614683151245117, "learning_rate": 1.601074318088937e-06, "loss": 1.1436, "step": 37690 }, { "epoch": 2.9538401253918494, "grad_norm": 2.4840917587280273, "learning_rate": 1.6008481683276355e-06, "loss": 1.6051, "step": 37691 }, { "epoch": 2.9539184952978057, "grad_norm": 5.71760892868042, "learning_rate": 1.6006220314949617e-06, "loss": 1.3873, "step": 37692 }, { "epoch": 2.9539968652037616, "grad_norm": 2.5227479934692383, "learning_rate": 1.6003959075917758e-06, "loss": 1.4068, "step": 37693 }, { "epoch": 2.954075235109718, "grad_norm": 2.779695510864258, "learning_rate": 1.6001697966189378e-06, "loss": 1.4111, "step": 37694 }, { "epoch": 2.954153605015674, "grad_norm": 2.941605806350708, "learning_rate": 1.5999436985773093e-06, "loss": 1.4722, "step": 37695 }, { "epoch": 2.95423197492163, "grad_norm": 2.698127508163452, "learning_rate": 1.5997176134677466e-06, "loss": 1.3852, "step": 37696 }, { "epoch": 2.9543103448275865, "grad_norm": 6.321450710296631, "learning_rate": 1.5994915412911104e-06, "loss": 1.4783, "step": 37697 }, { "epoch": 2.9543887147335424, "grad_norm": 4.457973957061768, "learning_rate": 1.5992654820482633e-06, "loss": 1.1781, "step": 37698 }, { "epoch": 2.9544670846394983, "grad_norm": 3.0352890491485596, "learning_rate": 1.5990394357400653e-06, "loss": 1.2227, "step": 37699 }, { "epoch": 2.9545454545454546, "grad_norm": 4.824745178222656, "learning_rate": 1.5988134023673723e-06, "loss": 1.2426, "step": 37700 }, { "epoch": 2.954623824451411, "grad_norm": 9.461304664611816, "learning_rate": 1.5985873819310466e-06, "loss": 1.5349, "step": 37701 }, { "epoch": 2.954702194357367, "grad_norm": 4.18074369430542, "learning_rate": 1.598361374431947e-06, "loss": 1.4393, "step": 37702 }, { "epoch": 2.9547805642633227, "grad_norm": 2.943164348602295, "learning_rate": 1.5981353798709342e-06, "loss": 1.6066, "step": 37703 }, { "epoch": 2.954858934169279, "grad_norm": 4.944517135620117, "learning_rate": 1.597909398248867e-06, "loss": 1.8774, "step": 37704 }, { "epoch": 2.9549373040752354, "grad_norm": 3.383040189743042, "learning_rate": 1.5976834295666044e-06, "loss": 0.888, "step": 37705 }, { "epoch": 2.9550156739811912, "grad_norm": 3.5123496055603027, "learning_rate": 1.597457473825007e-06, "loss": 1.8196, "step": 37706 }, { "epoch": 2.955094043887147, "grad_norm": 3.9735121726989746, "learning_rate": 1.5972315310249337e-06, "loss": 1.4102, "step": 37707 }, { "epoch": 2.9551724137931035, "grad_norm": 2.896193027496338, "learning_rate": 1.5970056011672435e-06, "loss": 1.5484, "step": 37708 }, { "epoch": 2.95525078369906, "grad_norm": 2.4648277759552, "learning_rate": 1.5967796842527967e-06, "loss": 1.0644, "step": 37709 }, { "epoch": 2.9553291536050157, "grad_norm": 3.5719552040100098, "learning_rate": 1.5965537802824532e-06, "loss": 1.2533, "step": 37710 }, { "epoch": 2.9554075235109716, "grad_norm": 3.506141424179077, "learning_rate": 1.5963278892570676e-06, "loss": 1.5542, "step": 37711 }, { "epoch": 2.955485893416928, "grad_norm": 3.0981385707855225, "learning_rate": 1.5961020111775045e-06, "loss": 1.4418, "step": 37712 }, { "epoch": 2.9555642633228842, "grad_norm": 2.156480073928833, "learning_rate": 1.595876146044621e-06, "loss": 1.4877, "step": 37713 }, { "epoch": 2.95564263322884, "grad_norm": 3.8197989463806152, "learning_rate": 1.5956502938592777e-06, "loss": 1.5548, "step": 37714 }, { "epoch": 2.955721003134796, "grad_norm": 9.28240966796875, "learning_rate": 1.5954244546223301e-06, "loss": 1.713, "step": 37715 }, { "epoch": 2.9557993730407524, "grad_norm": 10.811896324157715, "learning_rate": 1.5951986283346376e-06, "loss": 0.7865, "step": 37716 }, { "epoch": 2.9558777429467087, "grad_norm": 3.4769985675811768, "learning_rate": 1.5949728149970628e-06, "loss": 1.3802, "step": 37717 }, { "epoch": 2.9559561128526646, "grad_norm": 5.819784641265869, "learning_rate": 1.5947470146104639e-06, "loss": 1.7564, "step": 37718 }, { "epoch": 2.9560344827586205, "grad_norm": 3.085388422012329, "learning_rate": 1.5945212271756966e-06, "loss": 1.527, "step": 37719 }, { "epoch": 2.956112852664577, "grad_norm": 4.086789608001709, "learning_rate": 1.5942954526936217e-06, "loss": 0.7125, "step": 37720 }, { "epoch": 2.956191222570533, "grad_norm": 3.177385091781616, "learning_rate": 1.594069691165097e-06, "loss": 1.0786, "step": 37721 }, { "epoch": 2.956269592476489, "grad_norm": 2.8090901374816895, "learning_rate": 1.5938439425909823e-06, "loss": 1.3632, "step": 37722 }, { "epoch": 2.956347962382445, "grad_norm": 3.5848610401153564, "learning_rate": 1.5936182069721357e-06, "loss": 1.1657, "step": 37723 }, { "epoch": 2.9564263322884012, "grad_norm": 9.433104515075684, "learning_rate": 1.5933924843094155e-06, "loss": 2.1196, "step": 37724 }, { "epoch": 2.9565047021943576, "grad_norm": 8.349287033081055, "learning_rate": 1.593166774603681e-06, "loss": 1.9457, "step": 37725 }, { "epoch": 2.9565830721003135, "grad_norm": 2.590447187423706, "learning_rate": 1.5929410778557903e-06, "loss": 1.0337, "step": 37726 }, { "epoch": 2.9566614420062693, "grad_norm": 3.7228152751922607, "learning_rate": 1.5927153940666013e-06, "loss": 0.7794, "step": 37727 }, { "epoch": 2.9567398119122257, "grad_norm": 2.410025119781494, "learning_rate": 1.592489723236973e-06, "loss": 1.4278, "step": 37728 }, { "epoch": 2.956818181818182, "grad_norm": 3.643686294555664, "learning_rate": 1.5922640653677652e-06, "loss": 1.9126, "step": 37729 }, { "epoch": 2.956896551724138, "grad_norm": 2.9043540954589844, "learning_rate": 1.592038420459831e-06, "loss": 1.3807, "step": 37730 }, { "epoch": 2.956974921630094, "grad_norm": 2.4605705738067627, "learning_rate": 1.5918127885140344e-06, "loss": 1.3568, "step": 37731 }, { "epoch": 2.95705329153605, "grad_norm": 2.9232096672058105, "learning_rate": 1.5915871695312312e-06, "loss": 1.4041, "step": 37732 }, { "epoch": 2.9571316614420065, "grad_norm": 2.977001905441284, "learning_rate": 1.5913615635122814e-06, "loss": 1.3159, "step": 37733 }, { "epoch": 2.9572100313479623, "grad_norm": 2.3392419815063477, "learning_rate": 1.5911359704580394e-06, "loss": 1.2496, "step": 37734 }, { "epoch": 2.9572884012539182, "grad_norm": 8.510382652282715, "learning_rate": 1.590910390369364e-06, "loss": 1.0627, "step": 37735 }, { "epoch": 2.9573667711598746, "grad_norm": 3.4067869186401367, "learning_rate": 1.5906848232471178e-06, "loss": 1.2289, "step": 37736 }, { "epoch": 2.957445141065831, "grad_norm": 2.8141400814056396, "learning_rate": 1.5904592690921534e-06, "loss": 1.3094, "step": 37737 }, { "epoch": 2.957523510971787, "grad_norm": 3.090580940246582, "learning_rate": 1.590233727905331e-06, "loss": 1.2982, "step": 37738 }, { "epoch": 2.9576018808777427, "grad_norm": 2.5894691944122314, "learning_rate": 1.5900081996875083e-06, "loss": 1.367, "step": 37739 }, { "epoch": 2.957680250783699, "grad_norm": 3.707918643951416, "learning_rate": 1.5897826844395426e-06, "loss": 1.2977, "step": 37740 }, { "epoch": 2.9577586206896553, "grad_norm": 3.98936128616333, "learning_rate": 1.5895571821622918e-06, "loss": 1.3669, "step": 37741 }, { "epoch": 2.9578369905956112, "grad_norm": 3.318087100982666, "learning_rate": 1.5893316928566143e-06, "loss": 1.1475, "step": 37742 }, { "epoch": 2.9579153605015676, "grad_norm": 3.060832977294922, "learning_rate": 1.5891062165233667e-06, "loss": 1.042, "step": 37743 }, { "epoch": 2.9579937304075234, "grad_norm": 8.529153823852539, "learning_rate": 1.5888807531634076e-06, "loss": 0.5044, "step": 37744 }, { "epoch": 2.95807210031348, "grad_norm": 2.026923656463623, "learning_rate": 1.5886553027775936e-06, "loss": 1.0141, "step": 37745 }, { "epoch": 2.9581504702194357, "grad_norm": 3.523268222808838, "learning_rate": 1.588429865366783e-06, "loss": 1.2472, "step": 37746 }, { "epoch": 2.958228840125392, "grad_norm": 3.845977544784546, "learning_rate": 1.5882044409318325e-06, "loss": 1.2396, "step": 37747 }, { "epoch": 2.958307210031348, "grad_norm": 4.855645656585693, "learning_rate": 1.5879790294736019e-06, "loss": 1.5026, "step": 37748 }, { "epoch": 2.958385579937304, "grad_norm": 13.735885620117188, "learning_rate": 1.587753630992943e-06, "loss": 1.5925, "step": 37749 }, { "epoch": 2.95846394984326, "grad_norm": 5.168706893920898, "learning_rate": 1.5875282454907187e-06, "loss": 1.3716, "step": 37750 }, { "epoch": 2.9585423197492164, "grad_norm": 2.068291425704956, "learning_rate": 1.5873028729677852e-06, "loss": 0.9192, "step": 37751 }, { "epoch": 2.9586206896551723, "grad_norm": 5.856706619262695, "learning_rate": 1.5870775134249978e-06, "loss": 1.8712, "step": 37752 }, { "epoch": 2.9586990595611287, "grad_norm": 2.5915472507476807, "learning_rate": 1.5868521668632142e-06, "loss": 1.2945, "step": 37753 }, { "epoch": 2.9587774294670846, "grad_norm": 3.442411184310913, "learning_rate": 1.5866268332832902e-06, "loss": 1.7309, "step": 37754 }, { "epoch": 2.958855799373041, "grad_norm": 9.73812198638916, "learning_rate": 1.5864015126860877e-06, "loss": 1.5311, "step": 37755 }, { "epoch": 2.9589341692789968, "grad_norm": 3.6366286277770996, "learning_rate": 1.586176205072459e-06, "loss": 1.2995, "step": 37756 }, { "epoch": 2.959012539184953, "grad_norm": 3.2531795501708984, "learning_rate": 1.5859509104432625e-06, "loss": 1.3732, "step": 37757 }, { "epoch": 2.959090909090909, "grad_norm": 3.1342265605926514, "learning_rate": 1.585725628799355e-06, "loss": 1.142, "step": 37758 }, { "epoch": 2.9591692789968653, "grad_norm": 3.2102396488189697, "learning_rate": 1.5855003601415936e-06, "loss": 1.5438, "step": 37759 }, { "epoch": 2.959247648902821, "grad_norm": 8.146430015563965, "learning_rate": 1.5852751044708353e-06, "loss": 1.3548, "step": 37760 }, { "epoch": 2.9593260188087775, "grad_norm": 4.199692249298096, "learning_rate": 1.5850498617879367e-06, "loss": 1.3105, "step": 37761 }, { "epoch": 2.9594043887147334, "grad_norm": 2.2711575031280518, "learning_rate": 1.5848246320937556e-06, "loss": 1.297, "step": 37762 }, { "epoch": 2.9594827586206898, "grad_norm": 3.81229567527771, "learning_rate": 1.5845994153891442e-06, "loss": 1.053, "step": 37763 }, { "epoch": 2.9595611285266457, "grad_norm": 4.16622257232666, "learning_rate": 1.5843742116749643e-06, "loss": 1.1132, "step": 37764 }, { "epoch": 2.959639498432602, "grad_norm": 2.1178951263427734, "learning_rate": 1.5841490209520705e-06, "loss": 1.3264, "step": 37765 }, { "epoch": 2.959717868338558, "grad_norm": 5.583779811859131, "learning_rate": 1.5839238432213206e-06, "loss": 1.2745, "step": 37766 }, { "epoch": 2.959796238244514, "grad_norm": 2.8191943168640137, "learning_rate": 1.5836986784835684e-06, "loss": 0.9572, "step": 37767 }, { "epoch": 2.95987460815047, "grad_norm": 3.1697680950164795, "learning_rate": 1.58347352673967e-06, "loss": 1.3037, "step": 37768 }, { "epoch": 2.9599529780564264, "grad_norm": 2.3736329078674316, "learning_rate": 1.5832483879904854e-06, "loss": 1.2194, "step": 37769 }, { "epoch": 2.9600313479623823, "grad_norm": 3.472595453262329, "learning_rate": 1.5830232622368702e-06, "loss": 0.6691, "step": 37770 }, { "epoch": 2.9601097178683387, "grad_norm": 3.507215976715088, "learning_rate": 1.582798149479678e-06, "loss": 1.9108, "step": 37771 }, { "epoch": 2.9601880877742945, "grad_norm": 2.7095701694488525, "learning_rate": 1.5825730497197667e-06, "loss": 1.1934, "step": 37772 }, { "epoch": 2.960266457680251, "grad_norm": 4.278995037078857, "learning_rate": 1.5823479629579902e-06, "loss": 1.5842, "step": 37773 }, { "epoch": 2.9603448275862068, "grad_norm": 2.5500781536102295, "learning_rate": 1.5821228891952102e-06, "loss": 1.4119, "step": 37774 }, { "epoch": 2.960423197492163, "grad_norm": 5.524224758148193, "learning_rate": 1.5818978284322773e-06, "loss": 2.1781, "step": 37775 }, { "epoch": 2.960501567398119, "grad_norm": 7.375519275665283, "learning_rate": 1.5816727806700494e-06, "loss": 1.0093, "step": 37776 }, { "epoch": 2.9605799373040753, "grad_norm": 2.522261381149292, "learning_rate": 1.5814477459093825e-06, "loss": 1.1504, "step": 37777 }, { "epoch": 2.960658307210031, "grad_norm": 3.7651779651641846, "learning_rate": 1.5812227241511323e-06, "loss": 1.451, "step": 37778 }, { "epoch": 2.9607366771159875, "grad_norm": 2.915369749069214, "learning_rate": 1.5809977153961552e-06, "loss": 1.2219, "step": 37779 }, { "epoch": 2.9608150470219434, "grad_norm": 2.747364044189453, "learning_rate": 1.5807727196453065e-06, "loss": 1.3333, "step": 37780 }, { "epoch": 2.9608934169278998, "grad_norm": 6.698976039886475, "learning_rate": 1.5805477368994443e-06, "loss": 1.1933, "step": 37781 }, { "epoch": 2.9609717868338556, "grad_norm": 2.83227801322937, "learning_rate": 1.5803227671594184e-06, "loss": 1.3843, "step": 37782 }, { "epoch": 2.961050156739812, "grad_norm": 3.442965269088745, "learning_rate": 1.58009781042609e-06, "loss": 1.5638, "step": 37783 }, { "epoch": 2.961128526645768, "grad_norm": 2.727224111557007, "learning_rate": 1.5798728667003128e-06, "loss": 1.3359, "step": 37784 }, { "epoch": 2.961206896551724, "grad_norm": 2.6844117641448975, "learning_rate": 1.5796479359829443e-06, "loss": 1.176, "step": 37785 }, { "epoch": 2.96128526645768, "grad_norm": 5.40888786315918, "learning_rate": 1.5794230182748365e-06, "loss": 1.7376, "step": 37786 }, { "epoch": 2.9613636363636364, "grad_norm": 3.366051197052002, "learning_rate": 1.5791981135768452e-06, "loss": 0.6652, "step": 37787 }, { "epoch": 2.9614420062695928, "grad_norm": 5.88855504989624, "learning_rate": 1.578973221889829e-06, "loss": 1.2144, "step": 37788 }, { "epoch": 2.9615203761755486, "grad_norm": 2.56929087638855, "learning_rate": 1.5787483432146428e-06, "loss": 1.2062, "step": 37789 }, { "epoch": 2.9615987460815045, "grad_norm": 6.614963054656982, "learning_rate": 1.578523477552139e-06, "loss": 1.2935, "step": 37790 }, { "epoch": 2.961677115987461, "grad_norm": 3.7390360832214355, "learning_rate": 1.5782986249031746e-06, "loss": 1.2032, "step": 37791 }, { "epoch": 2.961755485893417, "grad_norm": 3.6444289684295654, "learning_rate": 1.5780737852686046e-06, "loss": 1.489, "step": 37792 }, { "epoch": 2.961833855799373, "grad_norm": 4.41389799118042, "learning_rate": 1.5778489586492845e-06, "loss": 1.3452, "step": 37793 }, { "epoch": 2.961912225705329, "grad_norm": 7.617102146148682, "learning_rate": 1.5776241450460694e-06, "loss": 1.1523, "step": 37794 }, { "epoch": 2.9619905956112853, "grad_norm": 7.315264701843262, "learning_rate": 1.577399344459814e-06, "loss": 1.1888, "step": 37795 }, { "epoch": 2.9620689655172416, "grad_norm": 2.6747078895568848, "learning_rate": 1.5771745568913732e-06, "loss": 1.3444, "step": 37796 }, { "epoch": 2.9621473354231975, "grad_norm": 2.309000015258789, "learning_rate": 1.576949782341603e-06, "loss": 1.0989, "step": 37797 }, { "epoch": 2.9622257053291534, "grad_norm": 6.025986194610596, "learning_rate": 1.5767250208113572e-06, "loss": 1.2491, "step": 37798 }, { "epoch": 2.9623040752351097, "grad_norm": 3.1289026737213135, "learning_rate": 1.576500272301491e-06, "loss": 1.4595, "step": 37799 }, { "epoch": 2.962382445141066, "grad_norm": 4.827458381652832, "learning_rate": 1.5762755368128597e-06, "loss": 1.667, "step": 37800 }, { "epoch": 2.962460815047022, "grad_norm": 2.965651035308838, "learning_rate": 1.5760508143463176e-06, "loss": 1.2606, "step": 37801 }, { "epoch": 2.962539184952978, "grad_norm": 3.7215020656585693, "learning_rate": 1.5758261049027195e-06, "loss": 1.4415, "step": 37802 }, { "epoch": 2.962617554858934, "grad_norm": 2.5281007289886475, "learning_rate": 1.5756014084829219e-06, "loss": 1.1565, "step": 37803 }, { "epoch": 2.9626959247648905, "grad_norm": 3.1927881240844727, "learning_rate": 1.5753767250877754e-06, "loss": 1.109, "step": 37804 }, { "epoch": 2.9627742946708464, "grad_norm": 3.4630513191223145, "learning_rate": 1.5751520547181358e-06, "loss": 1.1588, "step": 37805 }, { "epoch": 2.9628526645768023, "grad_norm": 3.4226083755493164, "learning_rate": 1.5749273973748603e-06, "loss": 1.1898, "step": 37806 }, { "epoch": 2.9629310344827586, "grad_norm": 4.437085151672363, "learning_rate": 1.5747027530588032e-06, "loss": 1.2859, "step": 37807 }, { "epoch": 2.963009404388715, "grad_norm": 3.09092116355896, "learning_rate": 1.5744781217708161e-06, "loss": 1.4045, "step": 37808 }, { "epoch": 2.963087774294671, "grad_norm": 9.1687593460083, "learning_rate": 1.5742535035117545e-06, "loss": 0.8205, "step": 37809 }, { "epoch": 2.9631661442006267, "grad_norm": 2.5076746940612793, "learning_rate": 1.574028898282472e-06, "loss": 1.2927, "step": 37810 }, { "epoch": 2.963244514106583, "grad_norm": 4.404662609100342, "learning_rate": 1.5738043060838264e-06, "loss": 1.661, "step": 37811 }, { "epoch": 2.9633228840125394, "grad_norm": 3.785686731338501, "learning_rate": 1.573579726916668e-06, "loss": 1.3946, "step": 37812 }, { "epoch": 2.9634012539184953, "grad_norm": 2.887129783630371, "learning_rate": 1.5733551607818525e-06, "loss": 1.3624, "step": 37813 }, { "epoch": 2.963479623824451, "grad_norm": 4.31727409362793, "learning_rate": 1.573130607680234e-06, "loss": 1.5319, "step": 37814 }, { "epoch": 2.9635579937304075, "grad_norm": 2.640681743621826, "learning_rate": 1.5729060676126662e-06, "loss": 1.5313, "step": 37815 }, { "epoch": 2.963636363636364, "grad_norm": 3.20497989654541, "learning_rate": 1.5726815405800034e-06, "loss": 1.485, "step": 37816 }, { "epoch": 2.9637147335423197, "grad_norm": 2.217975378036499, "learning_rate": 1.5724570265831002e-06, "loss": 1.3442, "step": 37817 }, { "epoch": 2.9637931034482756, "grad_norm": 4.7493462562561035, "learning_rate": 1.572232525622811e-06, "loss": 1.2376, "step": 37818 }, { "epoch": 2.963871473354232, "grad_norm": 14.139453887939453, "learning_rate": 1.5720080376999857e-06, "loss": 0.8957, "step": 37819 }, { "epoch": 2.9639498432601883, "grad_norm": 1.9026548862457275, "learning_rate": 1.5717835628154827e-06, "loss": 1.341, "step": 37820 }, { "epoch": 2.964028213166144, "grad_norm": 6.1967315673828125, "learning_rate": 1.5715591009701543e-06, "loss": 1.6632, "step": 37821 }, { "epoch": 2.9641065830721, "grad_norm": 2.2368690967559814, "learning_rate": 1.5713346521648553e-06, "loss": 1.4667, "step": 37822 }, { "epoch": 2.9641849529780564, "grad_norm": 3.9240670204162598, "learning_rate": 1.5711102164004372e-06, "loss": 0.6046, "step": 37823 }, { "epoch": 2.9642633228840127, "grad_norm": 5.022161960601807, "learning_rate": 1.5708857936777527e-06, "loss": 1.5189, "step": 37824 }, { "epoch": 2.9643416927899686, "grad_norm": 4.453867435455322, "learning_rate": 1.5706613839976587e-06, "loss": 1.6617, "step": 37825 }, { "epoch": 2.9644200626959245, "grad_norm": 5.500635623931885, "learning_rate": 1.5704369873610093e-06, "loss": 1.6593, "step": 37826 }, { "epoch": 2.964498432601881, "grad_norm": 3.6743736267089844, "learning_rate": 1.5702126037686543e-06, "loss": 1.5289, "step": 37827 }, { "epoch": 2.964576802507837, "grad_norm": 3.6354687213897705, "learning_rate": 1.5699882332214493e-06, "loss": 1.0565, "step": 37828 }, { "epoch": 2.964655172413793, "grad_norm": 3.298236131668091, "learning_rate": 1.5697638757202454e-06, "loss": 1.639, "step": 37829 }, { "epoch": 2.964733542319749, "grad_norm": 2.809962272644043, "learning_rate": 1.5695395312659011e-06, "loss": 1.0867, "step": 37830 }, { "epoch": 2.9648119122257053, "grad_norm": 3.8690407276153564, "learning_rate": 1.5693151998592649e-06, "loss": 1.0139, "step": 37831 }, { "epoch": 2.9648902821316616, "grad_norm": 4.426859378814697, "learning_rate": 1.5690908815011913e-06, "loss": 0.7889, "step": 37832 }, { "epoch": 2.9649686520376175, "grad_norm": 2.4777190685272217, "learning_rate": 1.5688665761925338e-06, "loss": 1.5944, "step": 37833 }, { "epoch": 2.9650470219435734, "grad_norm": 2.778351306915283, "learning_rate": 1.568642283934146e-06, "loss": 1.2887, "step": 37834 }, { "epoch": 2.9651253918495297, "grad_norm": 2.3677144050598145, "learning_rate": 1.5684180047268798e-06, "loss": 1.6284, "step": 37835 }, { "epoch": 2.965203761755486, "grad_norm": 2.58241605758667, "learning_rate": 1.5681937385715889e-06, "loss": 1.1027, "step": 37836 }, { "epoch": 2.965282131661442, "grad_norm": 3.982800006866455, "learning_rate": 1.5679694854691286e-06, "loss": 1.4512, "step": 37837 }, { "epoch": 2.9653605015673983, "grad_norm": 3.8064632415771484, "learning_rate": 1.5677452454203462e-06, "loss": 1.2046, "step": 37838 }, { "epoch": 2.965438871473354, "grad_norm": 4.603826999664307, "learning_rate": 1.5675210184260991e-06, "loss": 1.4387, "step": 37839 }, { "epoch": 2.9655172413793105, "grad_norm": 6.4785943031311035, "learning_rate": 1.5672968044872395e-06, "loss": 1.2162, "step": 37840 }, { "epoch": 2.9655956112852664, "grad_norm": 4.18551778793335, "learning_rate": 1.5670726036046213e-06, "loss": 1.5829, "step": 37841 }, { "epoch": 2.9656739811912227, "grad_norm": 3.064638614654541, "learning_rate": 1.5668484157790937e-06, "loss": 1.3459, "step": 37842 }, { "epoch": 2.9657523510971786, "grad_norm": 3.8274576663970947, "learning_rate": 1.56662424101151e-06, "loss": 1.2152, "step": 37843 }, { "epoch": 2.965830721003135, "grad_norm": 4.761483669281006, "learning_rate": 1.5664000793027257e-06, "loss": 1.2112, "step": 37844 }, { "epoch": 2.965909090909091, "grad_norm": 3.612941026687622, "learning_rate": 1.5661759306535934e-06, "loss": 1.1952, "step": 37845 }, { "epoch": 2.965987460815047, "grad_norm": 3.1045286655426025, "learning_rate": 1.5659517950649628e-06, "loss": 1.4899, "step": 37846 }, { "epoch": 2.966065830721003, "grad_norm": 8.87326431274414, "learning_rate": 1.5657276725376875e-06, "loss": 1.7139, "step": 37847 }, { "epoch": 2.9661442006269594, "grad_norm": 2.92584490776062, "learning_rate": 1.5655035630726206e-06, "loss": 1.2932, "step": 37848 }, { "epoch": 2.9662225705329153, "grad_norm": 4.235652923583984, "learning_rate": 1.5652794666706133e-06, "loss": 1.233, "step": 37849 }, { "epoch": 2.9663009404388716, "grad_norm": 4.163507461547852, "learning_rate": 1.5650553833325194e-06, "loss": 1.2113, "step": 37850 }, { "epoch": 2.9663793103448275, "grad_norm": 5.8614821434021, "learning_rate": 1.5648313130591903e-06, "loss": 1.2678, "step": 37851 }, { "epoch": 2.966457680250784, "grad_norm": 3.0638959407806396, "learning_rate": 1.5646072558514786e-06, "loss": 1.6216, "step": 37852 }, { "epoch": 2.9665360501567397, "grad_norm": 3.249253511428833, "learning_rate": 1.5643832117102365e-06, "loss": 1.4526, "step": 37853 }, { "epoch": 2.966614420062696, "grad_norm": 3.132201671600342, "learning_rate": 1.5641591806363154e-06, "loss": 0.8644, "step": 37854 }, { "epoch": 2.966692789968652, "grad_norm": 2.1937859058380127, "learning_rate": 1.5639351626305682e-06, "loss": 1.5799, "step": 37855 }, { "epoch": 2.9667711598746083, "grad_norm": 2.881187915802002, "learning_rate": 1.563711157693849e-06, "loss": 0.6585, "step": 37856 }, { "epoch": 2.966849529780564, "grad_norm": 2.575946807861328, "learning_rate": 1.5634871658270034e-06, "loss": 1.2586, "step": 37857 }, { "epoch": 2.9669278996865205, "grad_norm": 4.3802971839904785, "learning_rate": 1.5632631870308901e-06, "loss": 1.6115, "step": 37858 }, { "epoch": 2.9670062695924764, "grad_norm": 3.676273822784424, "learning_rate": 1.5630392213063595e-06, "loss": 1.3839, "step": 37859 }, { "epoch": 2.9670846394984327, "grad_norm": 3.2843379974365234, "learning_rate": 1.5628152686542607e-06, "loss": 1.0454, "step": 37860 }, { "epoch": 2.9671630094043886, "grad_norm": 2.5186686515808105, "learning_rate": 1.5625913290754474e-06, "loss": 1.3837, "step": 37861 }, { "epoch": 2.967241379310345, "grad_norm": 2.9969213008880615, "learning_rate": 1.5623674025707691e-06, "loss": 1.5324, "step": 37862 }, { "epoch": 2.967319749216301, "grad_norm": 5.366166114807129, "learning_rate": 1.562143489141083e-06, "loss": 1.2417, "step": 37863 }, { "epoch": 2.967398119122257, "grad_norm": 3.79318904876709, "learning_rate": 1.5619195887872357e-06, "loss": 1.5794, "step": 37864 }, { "epoch": 2.967476489028213, "grad_norm": 4.334095478057861, "learning_rate": 1.561695701510081e-06, "loss": 0.8791, "step": 37865 }, { "epoch": 2.9675548589341694, "grad_norm": 4.1077494621276855, "learning_rate": 1.5614718273104689e-06, "loss": 1.2292, "step": 37866 }, { "epoch": 2.9676332288401253, "grad_norm": 5.447553634643555, "learning_rate": 1.561247966189252e-06, "loss": 1.1554, "step": 37867 }, { "epoch": 2.9677115987460816, "grad_norm": 4.072476387023926, "learning_rate": 1.5610241181472819e-06, "loss": 1.1459, "step": 37868 }, { "epoch": 2.9677899686520375, "grad_norm": 3.9256465435028076, "learning_rate": 1.5608002831854092e-06, "loss": 2.0825, "step": 37869 }, { "epoch": 2.967868338557994, "grad_norm": 5.248345375061035, "learning_rate": 1.560576461304486e-06, "loss": 1.4334, "step": 37870 }, { "epoch": 2.9679467084639497, "grad_norm": 2.457624673843384, "learning_rate": 1.5603526525053637e-06, "loss": 1.3437, "step": 37871 }, { "epoch": 2.968025078369906, "grad_norm": 9.769134521484375, "learning_rate": 1.5601288567888928e-06, "loss": 1.3402, "step": 37872 }, { "epoch": 2.968103448275862, "grad_norm": 2.5305700302124023, "learning_rate": 1.5599050741559247e-06, "loss": 1.3543, "step": 37873 }, { "epoch": 2.9681818181818183, "grad_norm": 2.3917486667633057, "learning_rate": 1.559681304607313e-06, "loss": 1.1754, "step": 37874 }, { "epoch": 2.968260188087774, "grad_norm": 2.734193801879883, "learning_rate": 1.5594575481439046e-06, "loss": 1.2983, "step": 37875 }, { "epoch": 2.9683385579937305, "grad_norm": 4.0127668380737305, "learning_rate": 1.5592338047665507e-06, "loss": 1.0872, "step": 37876 }, { "epoch": 2.9684169278996864, "grad_norm": 2.22635817527771, "learning_rate": 1.5590100744761061e-06, "loss": 1.289, "step": 37877 }, { "epoch": 2.9684952978056427, "grad_norm": 2.9266059398651123, "learning_rate": 1.5587863572734207e-06, "loss": 1.7124, "step": 37878 }, { "epoch": 2.9685736677115986, "grad_norm": 2.7798960208892822, "learning_rate": 1.5585626531593434e-06, "loss": 1.0524, "step": 37879 }, { "epoch": 2.968652037617555, "grad_norm": 2.4264719486236572, "learning_rate": 1.5583389621347262e-06, "loss": 1.1621, "step": 37880 }, { "epoch": 2.968730407523511, "grad_norm": 5.0800251960754395, "learning_rate": 1.5581152842004183e-06, "loss": 1.1843, "step": 37881 }, { "epoch": 2.968808777429467, "grad_norm": 5.635780334472656, "learning_rate": 1.5578916193572752e-06, "loss": 1.7442, "step": 37882 }, { "epoch": 2.9688871473354235, "grad_norm": 2.3386189937591553, "learning_rate": 1.5576679676061424e-06, "loss": 1.3893, "step": 37883 }, { "epoch": 2.9689655172413794, "grad_norm": 2.9337074756622314, "learning_rate": 1.557444328947873e-06, "loss": 1.1549, "step": 37884 }, { "epoch": 2.9690438871473352, "grad_norm": 3.991363048553467, "learning_rate": 1.557220703383317e-06, "loss": 1.0449, "step": 37885 }, { "epoch": 2.9691222570532916, "grad_norm": 2.8554348945617676, "learning_rate": 1.5569970909133248e-06, "loss": 1.1941, "step": 37886 }, { "epoch": 2.969200626959248, "grad_norm": 3.280942440032959, "learning_rate": 1.5567734915387478e-06, "loss": 1.5706, "step": 37887 }, { "epoch": 2.969278996865204, "grad_norm": 2.5665059089660645, "learning_rate": 1.5565499052604354e-06, "loss": 1.501, "step": 37888 }, { "epoch": 2.9693573667711597, "grad_norm": 3.1595163345336914, "learning_rate": 1.5563263320792387e-06, "loss": 1.3996, "step": 37889 }, { "epoch": 2.969435736677116, "grad_norm": 2.9365358352661133, "learning_rate": 1.5561027719960076e-06, "loss": 1.515, "step": 37890 }, { "epoch": 2.9695141065830724, "grad_norm": 2.2404661178588867, "learning_rate": 1.5558792250115928e-06, "loss": 1.3165, "step": 37891 }, { "epoch": 2.9695924764890282, "grad_norm": 3.0801146030426025, "learning_rate": 1.555655691126845e-06, "loss": 1.382, "step": 37892 }, { "epoch": 2.969670846394984, "grad_norm": 7.083639144897461, "learning_rate": 1.5554321703426144e-06, "loss": 1.4164, "step": 37893 }, { "epoch": 2.9697492163009405, "grad_norm": 3.118098497390747, "learning_rate": 1.5552086626597495e-06, "loss": 1.3022, "step": 37894 }, { "epoch": 2.969827586206897, "grad_norm": 2.488022804260254, "learning_rate": 1.5549851680791e-06, "loss": 1.2915, "step": 37895 }, { "epoch": 2.9699059561128527, "grad_norm": 3.2790913581848145, "learning_rate": 1.5547616866015192e-06, "loss": 1.4475, "step": 37896 }, { "epoch": 2.9699843260188086, "grad_norm": 2.5357775688171387, "learning_rate": 1.5545382182278568e-06, "loss": 1.1116, "step": 37897 }, { "epoch": 2.970062695924765, "grad_norm": 3.8276326656341553, "learning_rate": 1.5543147629589595e-06, "loss": 1.5567, "step": 37898 }, { "epoch": 2.9701410658307212, "grad_norm": 2.3602192401885986, "learning_rate": 1.5540913207956792e-06, "loss": 1.0368, "step": 37899 }, { "epoch": 2.970219435736677, "grad_norm": 7.428220272064209, "learning_rate": 1.5538678917388638e-06, "loss": 1.1742, "step": 37900 }, { "epoch": 2.970297805642633, "grad_norm": 5.067237854003906, "learning_rate": 1.5536444757893681e-06, "loss": 1.4137, "step": 37901 }, { "epoch": 2.9703761755485893, "grad_norm": 2.6832692623138428, "learning_rate": 1.5534210729480376e-06, "loss": 1.3255, "step": 37902 }, { "epoch": 2.9704545454545457, "grad_norm": 2.5663013458251953, "learning_rate": 1.5531976832157225e-06, "loss": 1.1962, "step": 37903 }, { "epoch": 2.9705329153605016, "grad_norm": 7.775760650634766, "learning_rate": 1.5529743065932729e-06, "loss": 1.1911, "step": 37904 }, { "epoch": 2.9706112852664575, "grad_norm": 3.3049733638763428, "learning_rate": 1.552750943081539e-06, "loss": 1.0677, "step": 37905 }, { "epoch": 2.970689655172414, "grad_norm": 12.198859214782715, "learning_rate": 1.55252759268137e-06, "loss": 1.2412, "step": 37906 }, { "epoch": 2.97076802507837, "grad_norm": 4.9694647789001465, "learning_rate": 1.5523042553936145e-06, "loss": 1.4799, "step": 37907 }, { "epoch": 2.970846394984326, "grad_norm": 13.905288696289062, "learning_rate": 1.552080931219123e-06, "loss": 1.075, "step": 37908 }, { "epoch": 2.970924764890282, "grad_norm": 2.6528992652893066, "learning_rate": 1.5518576201587453e-06, "loss": 1.2515, "step": 37909 }, { "epoch": 2.9710031347962382, "grad_norm": 5.572894096374512, "learning_rate": 1.55163432221333e-06, "loss": 1.2169, "step": 37910 }, { "epoch": 2.9710815047021946, "grad_norm": 2.6953189373016357, "learning_rate": 1.551411037383726e-06, "loss": 1.3159, "step": 37911 }, { "epoch": 2.9711598746081505, "grad_norm": 2.786794424057007, "learning_rate": 1.5511877656707851e-06, "loss": 1.4969, "step": 37912 }, { "epoch": 2.9712382445141063, "grad_norm": 4.449642658233643, "learning_rate": 1.5509645070753516e-06, "loss": 1.2219, "step": 37913 }, { "epoch": 2.9713166144200627, "grad_norm": 3.1131222248077393, "learning_rate": 1.5507412615982787e-06, "loss": 1.2248, "step": 37914 }, { "epoch": 2.971394984326019, "grad_norm": 4.272220134735107, "learning_rate": 1.550518029240416e-06, "loss": 1.2044, "step": 37915 }, { "epoch": 2.971473354231975, "grad_norm": 3.859182357788086, "learning_rate": 1.5502948100026093e-06, "loss": 1.4795, "step": 37916 }, { "epoch": 2.971551724137931, "grad_norm": 3.980888605117798, "learning_rate": 1.5500716038857095e-06, "loss": 1.4379, "step": 37917 }, { "epoch": 2.971630094043887, "grad_norm": 3.4003710746765137, "learning_rate": 1.5498484108905637e-06, "loss": 1.4438, "step": 37918 }, { "epoch": 2.9717084639498434, "grad_norm": 5.172918796539307, "learning_rate": 1.5496252310180254e-06, "loss": 1.7551, "step": 37919 }, { "epoch": 2.9717868338557993, "grad_norm": 3.114950180053711, "learning_rate": 1.5494020642689395e-06, "loss": 1.4802, "step": 37920 }, { "epoch": 2.9718652037617552, "grad_norm": 5.274097919464111, "learning_rate": 1.549178910644155e-06, "loss": 1.4645, "step": 37921 }, { "epoch": 2.9719435736677116, "grad_norm": 3.4175519943237305, "learning_rate": 1.5489557701445218e-06, "loss": 1.3952, "step": 37922 }, { "epoch": 2.972021943573668, "grad_norm": 2.214040994644165, "learning_rate": 1.5487326427708888e-06, "loss": 1.3433, "step": 37923 }, { "epoch": 2.972100313479624, "grad_norm": 3.4605257511138916, "learning_rate": 1.5485095285241037e-06, "loss": 2.0077, "step": 37924 }, { "epoch": 2.9721786833855797, "grad_norm": 2.520998954772949, "learning_rate": 1.5482864274050152e-06, "loss": 1.5731, "step": 37925 }, { "epoch": 2.972257053291536, "grad_norm": 3.3483645915985107, "learning_rate": 1.5480633394144729e-06, "loss": 1.1314, "step": 37926 }, { "epoch": 2.9723354231974923, "grad_norm": 2.652989625930786, "learning_rate": 1.5478402645533242e-06, "loss": 1.4543, "step": 37927 }, { "epoch": 2.972413793103448, "grad_norm": 7.220274925231934, "learning_rate": 1.547617202822418e-06, "loss": 0.8563, "step": 37928 }, { "epoch": 2.9724921630094046, "grad_norm": 3.4615111351013184, "learning_rate": 1.5473941542226028e-06, "loss": 1.393, "step": 37929 }, { "epoch": 2.9725705329153604, "grad_norm": 3.4545207023620605, "learning_rate": 1.5471711187547284e-06, "loss": 1.3195, "step": 37930 }, { "epoch": 2.9726489028213168, "grad_norm": 6.787374496459961, "learning_rate": 1.5469480964196398e-06, "loss": 2.1125, "step": 37931 }, { "epoch": 2.9727272727272727, "grad_norm": 3.3984458446502686, "learning_rate": 1.546725087218186e-06, "loss": 1.4703, "step": 37932 }, { "epoch": 2.972805642633229, "grad_norm": 4.072822093963623, "learning_rate": 1.5465020911512174e-06, "loss": 1.7493, "step": 37933 }, { "epoch": 2.972884012539185, "grad_norm": 3.816073179244995, "learning_rate": 1.5462791082195826e-06, "loss": 1.1749, "step": 37934 }, { "epoch": 2.972962382445141, "grad_norm": 5.7125935554504395, "learning_rate": 1.5460561384241268e-06, "loss": 1.1122, "step": 37935 }, { "epoch": 2.973040752351097, "grad_norm": 2.343977689743042, "learning_rate": 1.5458331817656996e-06, "loss": 1.6503, "step": 37936 }, { "epoch": 2.9731191222570534, "grad_norm": 2.651763439178467, "learning_rate": 1.5456102382451465e-06, "loss": 1.2417, "step": 37937 }, { "epoch": 2.9731974921630093, "grad_norm": 4.546618461608887, "learning_rate": 1.5453873078633215e-06, "loss": 1.7427, "step": 37938 }, { "epoch": 2.9732758620689657, "grad_norm": 4.666528701782227, "learning_rate": 1.5451643906210673e-06, "loss": 1.1594, "step": 37939 }, { "epoch": 2.9733542319749215, "grad_norm": 3.002572536468506, "learning_rate": 1.5449414865192324e-06, "loss": 1.0687, "step": 37940 }, { "epoch": 2.973432601880878, "grad_norm": 2.823490619659424, "learning_rate": 1.5447185955586663e-06, "loss": 1.0757, "step": 37941 }, { "epoch": 2.9735109717868338, "grad_norm": 2.957547426223755, "learning_rate": 1.5444957177402153e-06, "loss": 1.5701, "step": 37942 }, { "epoch": 2.97358934169279, "grad_norm": 2.157355308532715, "learning_rate": 1.5442728530647282e-06, "loss": 1.2265, "step": 37943 }, { "epoch": 2.973667711598746, "grad_norm": 8.441225051879883, "learning_rate": 1.5440500015330522e-06, "loss": 1.3068, "step": 37944 }, { "epoch": 2.9737460815047023, "grad_norm": 3.7372887134552, "learning_rate": 1.5438271631460366e-06, "loss": 1.4948, "step": 37945 }, { "epoch": 2.973824451410658, "grad_norm": 2.4633381366729736, "learning_rate": 1.5436043379045235e-06, "loss": 1.054, "step": 37946 }, { "epoch": 2.9739028213166145, "grad_norm": 3.0653350353240967, "learning_rate": 1.5433815258093666e-06, "loss": 1.3235, "step": 37947 }, { "epoch": 2.9739811912225704, "grad_norm": 2.193228244781494, "learning_rate": 1.5431587268614102e-06, "loss": 1.311, "step": 37948 }, { "epoch": 2.9740595611285268, "grad_norm": 2.8659026622772217, "learning_rate": 1.542935941061504e-06, "loss": 1.3859, "step": 37949 }, { "epoch": 2.9741379310344827, "grad_norm": 3.8524937629699707, "learning_rate": 1.542713168410493e-06, "loss": 0.9299, "step": 37950 }, { "epoch": 2.974216300940439, "grad_norm": 2.7374343872070312, "learning_rate": 1.542490408909223e-06, "loss": 0.5055, "step": 37951 }, { "epoch": 2.974294670846395, "grad_norm": 2.868985414505005, "learning_rate": 1.5422676625585452e-06, "loss": 1.4848, "step": 37952 }, { "epoch": 2.974373040752351, "grad_norm": 5.256910800933838, "learning_rate": 1.5420449293593076e-06, "loss": 1.2021, "step": 37953 }, { "epoch": 2.974451410658307, "grad_norm": 2.7477123737335205, "learning_rate": 1.5418222093123525e-06, "loss": 1.2361, "step": 37954 }, { "epoch": 2.9745297805642634, "grad_norm": 10.110374450683594, "learning_rate": 1.5415995024185299e-06, "loss": 1.5788, "step": 37955 }, { "epoch": 2.9746081504702193, "grad_norm": 2.7883756160736084, "learning_rate": 1.5413768086786846e-06, "loss": 1.2287, "step": 37956 }, { "epoch": 2.9746865203761756, "grad_norm": 3.4843974113464355, "learning_rate": 1.5411541280936693e-06, "loss": 1.4173, "step": 37957 }, { "epoch": 2.9747648902821315, "grad_norm": 6.258711338043213, "learning_rate": 1.5409314606643249e-06, "loss": 1.6941, "step": 37958 }, { "epoch": 2.974843260188088, "grad_norm": 3.024395227432251, "learning_rate": 1.5407088063915004e-06, "loss": 1.286, "step": 37959 }, { "epoch": 2.9749216300940438, "grad_norm": 2.556774139404297, "learning_rate": 1.5404861652760434e-06, "loss": 1.303, "step": 37960 }, { "epoch": 2.975, "grad_norm": 3.4934263229370117, "learning_rate": 1.5402635373187996e-06, "loss": 1.4307, "step": 37961 }, { "epoch": 2.975078369905956, "grad_norm": 3.5539681911468506, "learning_rate": 1.5400409225206164e-06, "loss": 1.2615, "step": 37962 }, { "epoch": 2.9751567398119123, "grad_norm": 2.3928632736206055, "learning_rate": 1.5398183208823403e-06, "loss": 1.281, "step": 37963 }, { "epoch": 2.975235109717868, "grad_norm": 3.218728542327881, "learning_rate": 1.53959573240482e-06, "loss": 1.4488, "step": 37964 }, { "epoch": 2.9753134796238245, "grad_norm": 1.8820829391479492, "learning_rate": 1.539373157088897e-06, "loss": 1.1646, "step": 37965 }, { "epoch": 2.9753918495297804, "grad_norm": 2.910675525665283, "learning_rate": 1.5391505949354225e-06, "loss": 1.4551, "step": 37966 }, { "epoch": 2.9754702194357368, "grad_norm": 4.545895576477051, "learning_rate": 1.538928045945241e-06, "loss": 0.9674, "step": 37967 }, { "epoch": 2.9755485893416926, "grad_norm": 3.124749183654785, "learning_rate": 1.5387055101192017e-06, "loss": 1.8462, "step": 37968 }, { "epoch": 2.975626959247649, "grad_norm": 5.625943660736084, "learning_rate": 1.538482987458147e-06, "loss": 1.333, "step": 37969 }, { "epoch": 2.975705329153605, "grad_norm": 4.528103828430176, "learning_rate": 1.5382604779629234e-06, "loss": 1.2653, "step": 37970 }, { "epoch": 2.975783699059561, "grad_norm": 3.7353014945983887, "learning_rate": 1.538037981634382e-06, "loss": 1.3886, "step": 37971 }, { "epoch": 2.975862068965517, "grad_norm": 2.3494606018066406, "learning_rate": 1.5378154984733646e-06, "loss": 1.2656, "step": 37972 }, { "epoch": 2.9759404388714734, "grad_norm": 2.2204225063323975, "learning_rate": 1.537593028480719e-06, "loss": 1.0848, "step": 37973 }, { "epoch": 2.9760188087774293, "grad_norm": 3.8764939308166504, "learning_rate": 1.5373705716572905e-06, "loss": 1.4289, "step": 37974 }, { "epoch": 2.9760971786833856, "grad_norm": 2.601463794708252, "learning_rate": 1.5371481280039257e-06, "loss": 1.1481, "step": 37975 }, { "epoch": 2.9761755485893415, "grad_norm": 3.4285025596618652, "learning_rate": 1.5369256975214714e-06, "loss": 1.1875, "step": 37976 }, { "epoch": 2.976253918495298, "grad_norm": 3.129629373550415, "learning_rate": 1.5367032802107728e-06, "loss": 1.3672, "step": 37977 }, { "epoch": 2.976332288401254, "grad_norm": 4.294017314910889, "learning_rate": 1.5364808760726762e-06, "loss": 1.3274, "step": 37978 }, { "epoch": 2.97641065830721, "grad_norm": 2.343787908554077, "learning_rate": 1.536258485108027e-06, "loss": 0.9648, "step": 37979 }, { "epoch": 2.976489028213166, "grad_norm": 3.3813953399658203, "learning_rate": 1.5360361073176717e-06, "loss": 1.3952, "step": 37980 }, { "epoch": 2.9765673981191223, "grad_norm": 6.486560821533203, "learning_rate": 1.5358137427024555e-06, "loss": 0.9793, "step": 37981 }, { "epoch": 2.9766457680250786, "grad_norm": 3.3253016471862793, "learning_rate": 1.5355913912632254e-06, "loss": 1.2459, "step": 37982 }, { "epoch": 2.9767241379310345, "grad_norm": 3.9018073081970215, "learning_rate": 1.535369053000827e-06, "loss": 1.4336, "step": 37983 }, { "epoch": 2.9768025078369904, "grad_norm": 2.664062023162842, "learning_rate": 1.535146727916102e-06, "loss": 1.2182, "step": 37984 }, { "epoch": 2.9768808777429467, "grad_norm": 4.7740159034729, "learning_rate": 1.534924416009901e-06, "loss": 1.0432, "step": 37985 }, { "epoch": 2.976959247648903, "grad_norm": 2.60547137260437, "learning_rate": 1.5347021172830696e-06, "loss": 1.2564, "step": 37986 }, { "epoch": 2.977037617554859, "grad_norm": 3.2686281204223633, "learning_rate": 1.5344798317364495e-06, "loss": 1.0417, "step": 37987 }, { "epoch": 2.977115987460815, "grad_norm": 4.2614264488220215, "learning_rate": 1.534257559370888e-06, "loss": 1.2968, "step": 37988 }, { "epoch": 2.977194357366771, "grad_norm": 4.546886444091797, "learning_rate": 1.5340353001872294e-06, "loss": 1.2017, "step": 37989 }, { "epoch": 2.9772727272727275, "grad_norm": 3.082780599594116, "learning_rate": 1.5338130541863233e-06, "loss": 1.1539, "step": 37990 }, { "epoch": 2.9773510971786834, "grad_norm": 2.481536865234375, "learning_rate": 1.5335908213690104e-06, "loss": 1.1569, "step": 37991 }, { "epoch": 2.9774294670846393, "grad_norm": 2.2698280811309814, "learning_rate": 1.5333686017361372e-06, "loss": 1.2545, "step": 37992 }, { "epoch": 2.9775078369905956, "grad_norm": 2.7833168506622314, "learning_rate": 1.5331463952885494e-06, "loss": 1.1092, "step": 37993 }, { "epoch": 2.977586206896552, "grad_norm": 8.682404518127441, "learning_rate": 1.5329242020270924e-06, "loss": 0.7029, "step": 37994 }, { "epoch": 2.977664576802508, "grad_norm": 2.4584546089172363, "learning_rate": 1.5327020219526111e-06, "loss": 1.3638, "step": 37995 }, { "epoch": 2.9777429467084637, "grad_norm": 4.950236797332764, "learning_rate": 1.5324798550659498e-06, "loss": 1.3727, "step": 37996 }, { "epoch": 2.97782131661442, "grad_norm": 2.747107744216919, "learning_rate": 1.5322577013679546e-06, "loss": 1.2409, "step": 37997 }, { "epoch": 2.9778996865203764, "grad_norm": 17.356531143188477, "learning_rate": 1.5320355608594694e-06, "loss": 1.1113, "step": 37998 }, { "epoch": 2.9779780564263323, "grad_norm": 3.641087532043457, "learning_rate": 1.5318134335413405e-06, "loss": 1.4741, "step": 37999 }, { "epoch": 2.978056426332288, "grad_norm": 5.9857282638549805, "learning_rate": 1.5315913194144111e-06, "loss": 1.6466, "step": 38000 }, { "epoch": 2.9781347962382445, "grad_norm": 3.7412638664245605, "learning_rate": 1.5313692184795286e-06, "loss": 1.3442, "step": 38001 }, { "epoch": 2.978213166144201, "grad_norm": 4.972698211669922, "learning_rate": 1.531147130737533e-06, "loss": 1.1907, "step": 38002 }, { "epoch": 2.9782915360501567, "grad_norm": 4.2459211349487305, "learning_rate": 1.5309250561892736e-06, "loss": 1.4377, "step": 38003 }, { "epoch": 2.9783699059561126, "grad_norm": 2.0389835834503174, "learning_rate": 1.5307029948355934e-06, "loss": 1.0988, "step": 38004 }, { "epoch": 2.978448275862069, "grad_norm": 5.331708908081055, "learning_rate": 1.5304809466773385e-06, "loss": 1.4027, "step": 38005 }, { "epoch": 2.9785266457680253, "grad_norm": 2.7896816730499268, "learning_rate": 1.5302589117153504e-06, "loss": 1.6423, "step": 38006 }, { "epoch": 2.978605015673981, "grad_norm": 2.902742862701416, "learning_rate": 1.5300368899504741e-06, "loss": 1.1581, "step": 38007 }, { "epoch": 2.978683385579937, "grad_norm": 2.9174506664276123, "learning_rate": 1.5298148813835567e-06, "loss": 0.9227, "step": 38008 }, { "epoch": 2.9787617554858934, "grad_norm": 3.341902494430542, "learning_rate": 1.5295928860154424e-06, "loss": 1.2465, "step": 38009 }, { "epoch": 2.9788401253918497, "grad_norm": 4.980244159698486, "learning_rate": 1.5293709038469727e-06, "loss": 1.0221, "step": 38010 }, { "epoch": 2.9789184952978056, "grad_norm": 2.805542469024658, "learning_rate": 1.5291489348789934e-06, "loss": 1.3008, "step": 38011 }, { "epoch": 2.9789968652037615, "grad_norm": 2.13512921333313, "learning_rate": 1.528926979112349e-06, "loss": 1.2883, "step": 38012 }, { "epoch": 2.979075235109718, "grad_norm": 3.182041645050049, "learning_rate": 1.5287050365478833e-06, "loss": 1.4397, "step": 38013 }, { "epoch": 2.979153605015674, "grad_norm": 4.847725868225098, "learning_rate": 1.528483107186441e-06, "loss": 1.1577, "step": 38014 }, { "epoch": 2.97923197492163, "grad_norm": 2.574216365814209, "learning_rate": 1.5282611910288654e-06, "loss": 1.4721, "step": 38015 }, { "epoch": 2.979310344827586, "grad_norm": 2.3761770725250244, "learning_rate": 1.5280392880760015e-06, "loss": 0.8754, "step": 38016 }, { "epoch": 2.9793887147335423, "grad_norm": 8.068585395812988, "learning_rate": 1.5278173983286926e-06, "loss": 0.8327, "step": 38017 }, { "epoch": 2.9794670846394986, "grad_norm": 2.5057320594787598, "learning_rate": 1.527595521787783e-06, "loss": 1.6612, "step": 38018 }, { "epoch": 2.9795454545454545, "grad_norm": 3.5262253284454346, "learning_rate": 1.527373658454116e-06, "loss": 1.554, "step": 38019 }, { "epoch": 2.9796238244514104, "grad_norm": 3.3416247367858887, "learning_rate": 1.527151808328538e-06, "loss": 1.2766, "step": 38020 }, { "epoch": 2.9797021943573667, "grad_norm": 2.9634268283843994, "learning_rate": 1.526929971411888e-06, "loss": 1.071, "step": 38021 }, { "epoch": 2.979780564263323, "grad_norm": 3.08335542678833, "learning_rate": 1.5267081477050132e-06, "loss": 1.4858, "step": 38022 }, { "epoch": 2.979858934169279, "grad_norm": 3.1142497062683105, "learning_rate": 1.5264863372087574e-06, "loss": 1.2686, "step": 38023 }, { "epoch": 2.9799373040752353, "grad_norm": 4.775906085968018, "learning_rate": 1.5262645399239644e-06, "loss": 1.3665, "step": 38024 }, { "epoch": 2.980015673981191, "grad_norm": 2.618648052215576, "learning_rate": 1.5260427558514757e-06, "loss": 1.3381, "step": 38025 }, { "epoch": 2.9800940438871475, "grad_norm": 3.4746274948120117, "learning_rate": 1.525820984992134e-06, "loss": 1.5432, "step": 38026 }, { "epoch": 2.9801724137931034, "grad_norm": 2.507505416870117, "learning_rate": 1.5255992273467885e-06, "loss": 1.3779, "step": 38027 }, { "epoch": 2.9802507836990597, "grad_norm": 2.747645139694214, "learning_rate": 1.5253774829162771e-06, "loss": 1.1998, "step": 38028 }, { "epoch": 2.9803291536050156, "grad_norm": 6.667727947235107, "learning_rate": 1.5251557517014453e-06, "loss": 1.1381, "step": 38029 }, { "epoch": 2.980407523510972, "grad_norm": 3.6865525245666504, "learning_rate": 1.524934033703136e-06, "loss": 1.2986, "step": 38030 }, { "epoch": 2.980485893416928, "grad_norm": 2.815016508102417, "learning_rate": 1.5247123289221927e-06, "loss": 1.0114, "step": 38031 }, { "epoch": 2.980564263322884, "grad_norm": 2.6919734477996826, "learning_rate": 1.5244906373594588e-06, "loss": 1.1771, "step": 38032 }, { "epoch": 2.98064263322884, "grad_norm": 3.167360782623291, "learning_rate": 1.5242689590157778e-06, "loss": 1.2514, "step": 38033 }, { "epoch": 2.9807210031347964, "grad_norm": 3.461461067199707, "learning_rate": 1.5240472938919914e-06, "loss": 1.2763, "step": 38034 }, { "epoch": 2.9807993730407523, "grad_norm": 2.3088808059692383, "learning_rate": 1.5238256419889447e-06, "loss": 1.2946, "step": 38035 }, { "epoch": 2.9808777429467086, "grad_norm": 2.872934579849243, "learning_rate": 1.5236040033074794e-06, "loss": 1.1601, "step": 38036 }, { "epoch": 2.9809561128526645, "grad_norm": 4.175245761871338, "learning_rate": 1.523382377848439e-06, "loss": 0.9038, "step": 38037 }, { "epoch": 2.981034482758621, "grad_norm": 2.9365971088409424, "learning_rate": 1.5231607656126663e-06, "loss": 1.4018, "step": 38038 }, { "epoch": 2.9811128526645767, "grad_norm": 3.1968929767608643, "learning_rate": 1.5229391666010057e-06, "loss": 1.512, "step": 38039 }, { "epoch": 2.981191222570533, "grad_norm": 3.6829993724823, "learning_rate": 1.5227175808142952e-06, "loss": 1.6151, "step": 38040 }, { "epoch": 2.981269592476489, "grad_norm": 2.589311361312866, "learning_rate": 1.5224960082533829e-06, "loss": 1.3402, "step": 38041 }, { "epoch": 2.9813479623824453, "grad_norm": 3.054356336593628, "learning_rate": 1.522274448919111e-06, "loss": 1.1115, "step": 38042 }, { "epoch": 2.981426332288401, "grad_norm": 4.492208003997803, "learning_rate": 1.5220529028123193e-06, "loss": 1.5767, "step": 38043 }, { "epoch": 2.9815047021943575, "grad_norm": 3.990460157394409, "learning_rate": 1.521831369933852e-06, "loss": 1.434, "step": 38044 }, { "epoch": 2.9815830721003134, "grad_norm": 4.933966636657715, "learning_rate": 1.5216098502845505e-06, "loss": 1.0829, "step": 38045 }, { "epoch": 2.9816614420062697, "grad_norm": 2.7204692363739014, "learning_rate": 1.521388343865261e-06, "loss": 1.3113, "step": 38046 }, { "epoch": 2.9817398119122256, "grad_norm": 8.287081718444824, "learning_rate": 1.5211668506768223e-06, "loss": 1.4046, "step": 38047 }, { "epoch": 2.981818181818182, "grad_norm": 3.1316874027252197, "learning_rate": 1.5209453707200783e-06, "loss": 1.2536, "step": 38048 }, { "epoch": 2.981896551724138, "grad_norm": 2.4992599487304688, "learning_rate": 1.5207239039958705e-06, "loss": 1.2419, "step": 38049 }, { "epoch": 2.981974921630094, "grad_norm": 4.889383792877197, "learning_rate": 1.5205024505050424e-06, "loss": 1.3826, "step": 38050 }, { "epoch": 2.98205329153605, "grad_norm": 3.5593693256378174, "learning_rate": 1.5202810102484356e-06, "loss": 1.1455, "step": 38051 }, { "epoch": 2.9821316614420064, "grad_norm": 2.4876060485839844, "learning_rate": 1.5200595832268927e-06, "loss": 1.3692, "step": 38052 }, { "epoch": 2.9822100313479623, "grad_norm": 4.83960485458374, "learning_rate": 1.5198381694412556e-06, "loss": 1.1922, "step": 38053 }, { "epoch": 2.9822884012539186, "grad_norm": 8.553661346435547, "learning_rate": 1.5196167688923662e-06, "loss": 1.3371, "step": 38054 }, { "epoch": 2.9823667711598745, "grad_norm": 2.867732524871826, "learning_rate": 1.519395381581068e-06, "loss": 1.5646, "step": 38055 }, { "epoch": 2.982445141065831, "grad_norm": 2.3835151195526123, "learning_rate": 1.5191740075082012e-06, "loss": 1.1695, "step": 38056 }, { "epoch": 2.9825235109717867, "grad_norm": 2.6298038959503174, "learning_rate": 1.5189526466746102e-06, "loss": 1.7167, "step": 38057 }, { "epoch": 2.982601880877743, "grad_norm": 3.5622942447662354, "learning_rate": 1.5187312990811342e-06, "loss": 1.0162, "step": 38058 }, { "epoch": 2.982680250783699, "grad_norm": 2.649437189102173, "learning_rate": 1.5185099647286144e-06, "loss": 1.0945, "step": 38059 }, { "epoch": 2.9827586206896552, "grad_norm": 4.005218505859375, "learning_rate": 1.5182886436178963e-06, "loss": 0.5537, "step": 38060 }, { "epoch": 2.982836990595611, "grad_norm": 4.724791526794434, "learning_rate": 1.518067335749821e-06, "loss": 1.0495, "step": 38061 }, { "epoch": 2.9829153605015675, "grad_norm": 2.3614909648895264, "learning_rate": 1.5178460411252276e-06, "loss": 1.3883, "step": 38062 }, { "epoch": 2.9829937304075234, "grad_norm": 3.580899477005005, "learning_rate": 1.5176247597449596e-06, "loss": 1.1933, "step": 38063 }, { "epoch": 2.9830721003134797, "grad_norm": 3.5909717082977295, "learning_rate": 1.5174034916098569e-06, "loss": 1.6725, "step": 38064 }, { "epoch": 2.9831504702194356, "grad_norm": 13.520462989807129, "learning_rate": 1.5171822367207651e-06, "loss": 1.0806, "step": 38065 }, { "epoch": 2.983228840125392, "grad_norm": 4.1019744873046875, "learning_rate": 1.516960995078522e-06, "loss": 1.3368, "step": 38066 }, { "epoch": 2.983307210031348, "grad_norm": 4.719737529754639, "learning_rate": 1.5167397666839707e-06, "loss": 1.4338, "step": 38067 }, { "epoch": 2.983385579937304, "grad_norm": 2.8658151626586914, "learning_rate": 1.5165185515379515e-06, "loss": 1.5418, "step": 38068 }, { "epoch": 2.9834639498432605, "grad_norm": 3.263143301010132, "learning_rate": 1.5162973496413069e-06, "loss": 1.3265, "step": 38069 }, { "epoch": 2.9835423197492164, "grad_norm": 2.821674346923828, "learning_rate": 1.5160761609948776e-06, "loss": 1.2775, "step": 38070 }, { "epoch": 2.9836206896551722, "grad_norm": 2.3420064449310303, "learning_rate": 1.5158549855995052e-06, "loss": 1.0867, "step": 38071 }, { "epoch": 2.9836990595611286, "grad_norm": 3.589148759841919, "learning_rate": 1.5156338234560325e-06, "loss": 1.6094, "step": 38072 }, { "epoch": 2.983777429467085, "grad_norm": 3.3611667156219482, "learning_rate": 1.5154126745652958e-06, "loss": 1.6713, "step": 38073 }, { "epoch": 2.983855799373041, "grad_norm": 5.618391513824463, "learning_rate": 1.5151915389281414e-06, "loss": 1.3935, "step": 38074 }, { "epoch": 2.9839341692789967, "grad_norm": 2.8943655490875244, "learning_rate": 1.5149704165454083e-06, "loss": 1.6298, "step": 38075 }, { "epoch": 2.984012539184953, "grad_norm": 2.5701544284820557, "learning_rate": 1.5147493074179387e-06, "loss": 1.0778, "step": 38076 }, { "epoch": 2.9840909090909093, "grad_norm": 3.0434563159942627, "learning_rate": 1.5145282115465715e-06, "loss": 1.1395, "step": 38077 }, { "epoch": 2.9841692789968652, "grad_norm": 2.499079465866089, "learning_rate": 1.5143071289321476e-06, "loss": 1.4339, "step": 38078 }, { "epoch": 2.984247648902821, "grad_norm": 3.513875961303711, "learning_rate": 1.5140860595755097e-06, "loss": 1.0886, "step": 38079 }, { "epoch": 2.9843260188087775, "grad_norm": 2.5657596588134766, "learning_rate": 1.5138650034775004e-06, "loss": 1.3529, "step": 38080 }, { "epoch": 2.984404388714734, "grad_norm": 2.511406421661377, "learning_rate": 1.5136439606389557e-06, "loss": 1.661, "step": 38081 }, { "epoch": 2.9844827586206897, "grad_norm": 3.720679521560669, "learning_rate": 1.513422931060719e-06, "loss": 1.5284, "step": 38082 }, { "epoch": 2.9845611285266456, "grad_norm": 3.5804431438446045, "learning_rate": 1.5132019147436311e-06, "loss": 0.987, "step": 38083 }, { "epoch": 2.984639498432602, "grad_norm": 13.226292610168457, "learning_rate": 1.5129809116885313e-06, "loss": 1.0469, "step": 38084 }, { "epoch": 2.9847178683385582, "grad_norm": 6.665774822235107, "learning_rate": 1.5127599218962618e-06, "loss": 1.4978, "step": 38085 }, { "epoch": 2.984796238244514, "grad_norm": 3.1349728107452393, "learning_rate": 1.512538945367662e-06, "loss": 1.5056, "step": 38086 }, { "epoch": 2.98487460815047, "grad_norm": 2.8766520023345947, "learning_rate": 1.5123179821035728e-06, "loss": 1.41, "step": 38087 }, { "epoch": 2.9849529780564263, "grad_norm": 11.931944847106934, "learning_rate": 1.512097032104835e-06, "loss": 1.5368, "step": 38088 }, { "epoch": 2.9850313479623827, "grad_norm": 4.055322170257568, "learning_rate": 1.5118760953722882e-06, "loss": 1.5137, "step": 38089 }, { "epoch": 2.9851097178683386, "grad_norm": 4.65578031539917, "learning_rate": 1.5116551719067735e-06, "loss": 1.4908, "step": 38090 }, { "epoch": 2.9851880877742945, "grad_norm": 8.774991035461426, "learning_rate": 1.5114342617091315e-06, "loss": 1.3047, "step": 38091 }, { "epoch": 2.985266457680251, "grad_norm": 3.1765480041503906, "learning_rate": 1.5112133647801997e-06, "loss": 0.8638, "step": 38092 }, { "epoch": 2.985344827586207, "grad_norm": 3.5626280307769775, "learning_rate": 1.510992481120821e-06, "loss": 1.1375, "step": 38093 }, { "epoch": 2.985423197492163, "grad_norm": 6.833293914794922, "learning_rate": 1.5107716107318353e-06, "loss": 1.6923, "step": 38094 }, { "epoch": 2.985501567398119, "grad_norm": 3.1027355194091797, "learning_rate": 1.5105507536140834e-06, "loss": 1.3304, "step": 38095 }, { "epoch": 2.9855799373040752, "grad_norm": 4.344313144683838, "learning_rate": 1.5103299097684032e-06, "loss": 1.533, "step": 38096 }, { "epoch": 2.9856583072100316, "grad_norm": 4.4004974365234375, "learning_rate": 1.5101090791956334e-06, "loss": 1.4913, "step": 38097 }, { "epoch": 2.9857366771159874, "grad_norm": 2.007371664047241, "learning_rate": 1.5098882618966198e-06, "loss": 1.0677, "step": 38098 }, { "epoch": 2.9858150470219433, "grad_norm": 2.5792901515960693, "learning_rate": 1.509667457872197e-06, "loss": 1.184, "step": 38099 }, { "epoch": 2.9858934169278997, "grad_norm": 2.593546152114868, "learning_rate": 1.5094466671232061e-06, "loss": 1.3396, "step": 38100 }, { "epoch": 2.985971786833856, "grad_norm": 5.809886932373047, "learning_rate": 1.5092258896504875e-06, "loss": 0.9394, "step": 38101 }, { "epoch": 2.986050156739812, "grad_norm": 5.633169174194336, "learning_rate": 1.5090051254548803e-06, "loss": 0.9564, "step": 38102 }, { "epoch": 2.9861285266457678, "grad_norm": 4.817589282989502, "learning_rate": 1.5087843745372248e-06, "loss": 1.441, "step": 38103 }, { "epoch": 2.986206896551724, "grad_norm": 3.5263123512268066, "learning_rate": 1.5085636368983602e-06, "loss": 1.3182, "step": 38104 }, { "epoch": 2.9862852664576804, "grad_norm": 5.818052291870117, "learning_rate": 1.5083429125391264e-06, "loss": 1.3702, "step": 38105 }, { "epoch": 2.9863636363636363, "grad_norm": 3.446112632751465, "learning_rate": 1.5081222014603624e-06, "loss": 1.5508, "step": 38106 }, { "epoch": 2.986442006269592, "grad_norm": 5.297454357147217, "learning_rate": 1.5079015036629086e-06, "loss": 1.4493, "step": 38107 }, { "epoch": 2.9865203761755486, "grad_norm": 3.689304828643799, "learning_rate": 1.5076808191476033e-06, "loss": 1.6411, "step": 38108 }, { "epoch": 2.986598746081505, "grad_norm": 3.1974472999572754, "learning_rate": 1.5074601479152862e-06, "loss": 1.302, "step": 38109 }, { "epoch": 2.9866771159874608, "grad_norm": 4.197535991668701, "learning_rate": 1.5072394899667974e-06, "loss": 1.16, "step": 38110 }, { "epoch": 2.9867554858934167, "grad_norm": 3.8431379795074463, "learning_rate": 1.5070188453029754e-06, "loss": 1.8174, "step": 38111 }, { "epoch": 2.986833855799373, "grad_norm": 3.082752227783203, "learning_rate": 1.5067982139246596e-06, "loss": 1.6609, "step": 38112 }, { "epoch": 2.9869122257053293, "grad_norm": 4.0122809410095215, "learning_rate": 1.5065775958326906e-06, "loss": 1.4443, "step": 38113 }, { "epoch": 2.986990595611285, "grad_norm": 1.8201268911361694, "learning_rate": 1.5063569910279047e-06, "loss": 1.3627, "step": 38114 }, { "epoch": 2.987068965517241, "grad_norm": 8.270423889160156, "learning_rate": 1.5061363995111406e-06, "loss": 1.121, "step": 38115 }, { "epoch": 2.9871473354231974, "grad_norm": 4.3698601722717285, "learning_rate": 1.5059158212832404e-06, "loss": 1.1652, "step": 38116 }, { "epoch": 2.9872257053291538, "grad_norm": 4.22066068649292, "learning_rate": 1.505695256345044e-06, "loss": 1.6602, "step": 38117 }, { "epoch": 2.9873040752351097, "grad_norm": 2.7541792392730713, "learning_rate": 1.5054747046973861e-06, "loss": 1.3289, "step": 38118 }, { "epoch": 2.987382445141066, "grad_norm": 3.331965208053589, "learning_rate": 1.5052541663411073e-06, "loss": 1.4173, "step": 38119 }, { "epoch": 2.987460815047022, "grad_norm": 2.546386957168579, "learning_rate": 1.5050336412770456e-06, "loss": 1.2236, "step": 38120 }, { "epoch": 2.987539184952978, "grad_norm": 3.822535991668701, "learning_rate": 1.5048131295060436e-06, "loss": 1.3734, "step": 38121 }, { "epoch": 2.987617554858934, "grad_norm": 3.93754506111145, "learning_rate": 1.5045926310289355e-06, "loss": 1.3355, "step": 38122 }, { "epoch": 2.9876959247648904, "grad_norm": 3.380248546600342, "learning_rate": 1.504372145846562e-06, "loss": 1.3661, "step": 38123 }, { "epoch": 2.9877742946708463, "grad_norm": 3.568695545196533, "learning_rate": 1.5041516739597612e-06, "loss": 1.5501, "step": 38124 }, { "epoch": 2.9878526645768027, "grad_norm": 2.4739108085632324, "learning_rate": 1.503931215369372e-06, "loss": 1.6237, "step": 38125 }, { "epoch": 2.9879310344827585, "grad_norm": 2.541921854019165, "learning_rate": 1.5037107700762328e-06, "loss": 1.1414, "step": 38126 }, { "epoch": 2.988009404388715, "grad_norm": 3.3677797317504883, "learning_rate": 1.5034903380811815e-06, "loss": 1.3349, "step": 38127 }, { "epoch": 2.9880877742946708, "grad_norm": 5.0737810134887695, "learning_rate": 1.5032699193850592e-06, "loss": 1.3459, "step": 38128 }, { "epoch": 2.988166144200627, "grad_norm": 5.154611587524414, "learning_rate": 1.5030495139886986e-06, "loss": 1.4178, "step": 38129 }, { "epoch": 2.988244514106583, "grad_norm": 3.3470089435577393, "learning_rate": 1.5028291218929435e-06, "loss": 0.9956, "step": 38130 }, { "epoch": 2.9883228840125393, "grad_norm": 7.829816818237305, "learning_rate": 1.5026087430986297e-06, "loss": 1.3578, "step": 38131 }, { "epoch": 2.988401253918495, "grad_norm": 4.371522903442383, "learning_rate": 1.5023883776065979e-06, "loss": 1.3067, "step": 38132 }, { "epoch": 2.9884796238244515, "grad_norm": 2.856313705444336, "learning_rate": 1.5021680254176823e-06, "loss": 1.1954, "step": 38133 }, { "epoch": 2.9885579937304074, "grad_norm": 3.8394858837127686, "learning_rate": 1.5019476865327214e-06, "loss": 0.868, "step": 38134 }, { "epoch": 2.9886363636363638, "grad_norm": 5.091721534729004, "learning_rate": 1.5017273609525562e-06, "loss": 1.1836, "step": 38135 }, { "epoch": 2.9887147335423196, "grad_norm": 3.048013925552368, "learning_rate": 1.5015070486780254e-06, "loss": 1.4691, "step": 38136 }, { "epoch": 2.988793103448276, "grad_norm": 4.834438800811768, "learning_rate": 1.5012867497099625e-06, "loss": 1.5179, "step": 38137 }, { "epoch": 2.988871473354232, "grad_norm": 5.151291847229004, "learning_rate": 1.5010664640492084e-06, "loss": 1.0661, "step": 38138 }, { "epoch": 2.988949843260188, "grad_norm": 3.8770511150360107, "learning_rate": 1.5008461916966e-06, "loss": 1.4723, "step": 38139 }, { "epoch": 2.989028213166144, "grad_norm": 4.660502910614014, "learning_rate": 1.5006259326529755e-06, "loss": 1.1298, "step": 38140 }, { "epoch": 2.9891065830721004, "grad_norm": 21.176288604736328, "learning_rate": 1.5004056869191724e-06, "loss": 1.0408, "step": 38141 }, { "epoch": 2.9891849529780563, "grad_norm": 2.5508499145507812, "learning_rate": 1.5001854544960287e-06, "loss": 1.1779, "step": 38142 }, { "epoch": 2.9892633228840126, "grad_norm": 2.9238247871398926, "learning_rate": 1.4999652353843825e-06, "loss": 1.4066, "step": 38143 }, { "epoch": 2.9893416927899685, "grad_norm": 2.538001298904419, "learning_rate": 1.4997450295850702e-06, "loss": 1.3584, "step": 38144 }, { "epoch": 2.989420062695925, "grad_norm": 4.246107578277588, "learning_rate": 1.4995248370989296e-06, "loss": 1.8629, "step": 38145 }, { "epoch": 2.9894984326018808, "grad_norm": 6.312795639038086, "learning_rate": 1.499304657926799e-06, "loss": 1.7159, "step": 38146 }, { "epoch": 2.989576802507837, "grad_norm": 3.2742230892181396, "learning_rate": 1.4990844920695169e-06, "loss": 1.0446, "step": 38147 }, { "epoch": 2.989655172413793, "grad_norm": 2.652573585510254, "learning_rate": 1.4988643395279161e-06, "loss": 1.4448, "step": 38148 }, { "epoch": 2.9897335423197493, "grad_norm": 2.696606159210205, "learning_rate": 1.498644200302839e-06, "loss": 1.2399, "step": 38149 }, { "epoch": 2.989811912225705, "grad_norm": 3.1668992042541504, "learning_rate": 1.4984240743951205e-06, "loss": 1.4559, "step": 38150 }, { "epoch": 2.9898902821316615, "grad_norm": 5.7943034172058105, "learning_rate": 1.4982039618055999e-06, "loss": 1.1859, "step": 38151 }, { "epoch": 2.9899686520376174, "grad_norm": 2.8206143379211426, "learning_rate": 1.4979838625351112e-06, "loss": 1.3177, "step": 38152 }, { "epoch": 2.9900470219435737, "grad_norm": 9.29182243347168, "learning_rate": 1.4977637765844917e-06, "loss": 0.7555, "step": 38153 }, { "epoch": 2.9901253918495296, "grad_norm": 8.897500038146973, "learning_rate": 1.4975437039545831e-06, "loss": 0.8531, "step": 38154 }, { "epoch": 2.990203761755486, "grad_norm": 3.8331429958343506, "learning_rate": 1.4973236446462176e-06, "loss": 1.0492, "step": 38155 }, { "epoch": 2.990282131661442, "grad_norm": 3.434093713760376, "learning_rate": 1.497103598660234e-06, "loss": 1.527, "step": 38156 }, { "epoch": 2.990360501567398, "grad_norm": 7.012852668762207, "learning_rate": 1.496883565997469e-06, "loss": 1.2845, "step": 38157 }, { "epoch": 2.990438871473354, "grad_norm": 3.079484462738037, "learning_rate": 1.4966635466587599e-06, "loss": 1.5819, "step": 38158 }, { "epoch": 2.9905172413793104, "grad_norm": 3.137087345123291, "learning_rate": 1.496443540644943e-06, "loss": 1.3707, "step": 38159 }, { "epoch": 2.9905956112852663, "grad_norm": 3.684784173965454, "learning_rate": 1.496223547956855e-06, "loss": 1.5793, "step": 38160 }, { "epoch": 2.9906739811912226, "grad_norm": 2.6550345420837402, "learning_rate": 1.4960035685953333e-06, "loss": 1.5307, "step": 38161 }, { "epoch": 2.9907523510971785, "grad_norm": 2.6684324741363525, "learning_rate": 1.4957836025612143e-06, "loss": 1.2446, "step": 38162 }, { "epoch": 2.990830721003135, "grad_norm": 2.2162883281707764, "learning_rate": 1.4955636498553344e-06, "loss": 1.1846, "step": 38163 }, { "epoch": 2.990909090909091, "grad_norm": 2.3534622192382812, "learning_rate": 1.4953437104785307e-06, "loss": 1.3995, "step": 38164 }, { "epoch": 2.990987460815047, "grad_norm": 4.466858863830566, "learning_rate": 1.495123784431639e-06, "loss": 1.2544, "step": 38165 }, { "epoch": 2.991065830721003, "grad_norm": 2.6578681468963623, "learning_rate": 1.4949038717154979e-06, "loss": 1.4854, "step": 38166 }, { "epoch": 2.9911442006269593, "grad_norm": 3.283842086791992, "learning_rate": 1.494683972330939e-06, "loss": 1.4, "step": 38167 }, { "epoch": 2.9912225705329156, "grad_norm": 6.110470294952393, "learning_rate": 1.4944640862788034e-06, "loss": 2.0074, "step": 38168 }, { "epoch": 2.9913009404388715, "grad_norm": 3.471508026123047, "learning_rate": 1.4942442135599277e-06, "loss": 1.3603, "step": 38169 }, { "epoch": 2.9913793103448274, "grad_norm": 3.5304648876190186, "learning_rate": 1.4940243541751449e-06, "loss": 1.1628, "step": 38170 }, { "epoch": 2.9914576802507837, "grad_norm": 4.5274739265441895, "learning_rate": 1.4938045081252923e-06, "loss": 1.8881, "step": 38171 }, { "epoch": 2.99153605015674, "grad_norm": 3.9378318786621094, "learning_rate": 1.4935846754112048e-06, "loss": 1.1752, "step": 38172 }, { "epoch": 2.991614420062696, "grad_norm": 8.858149528503418, "learning_rate": 1.4933648560337233e-06, "loss": 1.1686, "step": 38173 }, { "epoch": 2.991692789968652, "grad_norm": 2.716485023498535, "learning_rate": 1.4931450499936795e-06, "loss": 1.4456, "step": 38174 }, { "epoch": 2.991771159874608, "grad_norm": 4.734024524688721, "learning_rate": 1.4929252572919106e-06, "loss": 1.3034, "step": 38175 }, { "epoch": 2.9918495297805645, "grad_norm": 4.697481632232666, "learning_rate": 1.4927054779292527e-06, "loss": 1.0504, "step": 38176 }, { "epoch": 2.9919278996865204, "grad_norm": 4.657777309417725, "learning_rate": 1.4924857119065417e-06, "loss": 1.2206, "step": 38177 }, { "epoch": 2.9920062695924763, "grad_norm": 9.488373756408691, "learning_rate": 1.492265959224613e-06, "loss": 1.0476, "step": 38178 }, { "epoch": 2.9920846394984326, "grad_norm": 2.2302639484405518, "learning_rate": 1.4920462198843034e-06, "loss": 1.3216, "step": 38179 }, { "epoch": 2.992163009404389, "grad_norm": 6.863969802856445, "learning_rate": 1.4918264938864495e-06, "loss": 0.8551, "step": 38180 }, { "epoch": 2.992241379310345, "grad_norm": 2.856574296951294, "learning_rate": 1.4916067812318829e-06, "loss": 1.4991, "step": 38181 }, { "epoch": 2.9923197492163007, "grad_norm": 2.8377935886383057, "learning_rate": 1.4913870819214438e-06, "loss": 1.2595, "step": 38182 }, { "epoch": 2.992398119122257, "grad_norm": 3.350767135620117, "learning_rate": 1.4911673959559658e-06, "loss": 1.507, "step": 38183 }, { "epoch": 2.9924764890282134, "grad_norm": 3.253711462020874, "learning_rate": 1.4909477233362863e-06, "loss": 1.4115, "step": 38184 }, { "epoch": 2.9925548589341693, "grad_norm": 3.5612053871154785, "learning_rate": 1.4907280640632377e-06, "loss": 1.2186, "step": 38185 }, { "epoch": 2.992633228840125, "grad_norm": 2.7219431400299072, "learning_rate": 1.4905084181376556e-06, "loss": 1.5609, "step": 38186 }, { "epoch": 2.9927115987460815, "grad_norm": 3.3227121829986572, "learning_rate": 1.4902887855603782e-06, "loss": 1.037, "step": 38187 }, { "epoch": 2.992789968652038, "grad_norm": 4.221363067626953, "learning_rate": 1.4900691663322414e-06, "loss": 1.1755, "step": 38188 }, { "epoch": 2.9928683385579937, "grad_norm": 3.2765114307403564, "learning_rate": 1.489849560454077e-06, "loss": 1.2164, "step": 38189 }, { "epoch": 2.9929467084639496, "grad_norm": 4.445977210998535, "learning_rate": 1.4896299679267223e-06, "loss": 1.8206, "step": 38190 }, { "epoch": 2.993025078369906, "grad_norm": 3.477600574493408, "learning_rate": 1.4894103887510102e-06, "loss": 1.0296, "step": 38191 }, { "epoch": 2.9931034482758623, "grad_norm": 3.0592567920684814, "learning_rate": 1.4891908229277813e-06, "loss": 1.4284, "step": 38192 }, { "epoch": 2.993181818181818, "grad_norm": 3.302680492401123, "learning_rate": 1.4889712704578652e-06, "loss": 1.0903, "step": 38193 }, { "epoch": 2.993260188087774, "grad_norm": 4.238071918487549, "learning_rate": 1.488751731342099e-06, "loss": 1.2472, "step": 38194 }, { "epoch": 2.9933385579937304, "grad_norm": 3.0670626163482666, "learning_rate": 1.4885322055813183e-06, "loss": 0.884, "step": 38195 }, { "epoch": 2.9934169278996867, "grad_norm": 2.6932921409606934, "learning_rate": 1.488312693176357e-06, "loss": 1.0755, "step": 38196 }, { "epoch": 2.9934952978056426, "grad_norm": 3.730293035507202, "learning_rate": 1.4880931941280507e-06, "loss": 1.0857, "step": 38197 }, { "epoch": 2.9935736677115985, "grad_norm": 2.270296573638916, "learning_rate": 1.487873708437234e-06, "loss": 1.4162, "step": 38198 }, { "epoch": 2.993652037617555, "grad_norm": 5.759944915771484, "learning_rate": 1.487654236104742e-06, "loss": 1.6005, "step": 38199 }, { "epoch": 2.993730407523511, "grad_norm": 3.5237293243408203, "learning_rate": 1.487434777131409e-06, "loss": 1.1618, "step": 38200 }, { "epoch": 2.993808777429467, "grad_norm": 3.46787428855896, "learning_rate": 1.4872153315180699e-06, "loss": 1.3928, "step": 38201 }, { "epoch": 2.993887147335423, "grad_norm": 4.564770221710205, "learning_rate": 1.4869958992655592e-06, "loss": 1.4602, "step": 38202 }, { "epoch": 2.9939655172413793, "grad_norm": 4.269495487213135, "learning_rate": 1.4867764803747137e-06, "loss": 0.8888, "step": 38203 }, { "epoch": 2.9940438871473356, "grad_norm": 3.111563205718994, "learning_rate": 1.4865570748463625e-06, "loss": 1.4738, "step": 38204 }, { "epoch": 2.9941222570532915, "grad_norm": 2.4982173442840576, "learning_rate": 1.4863376826813452e-06, "loss": 1.2134, "step": 38205 }, { "epoch": 2.9942006269592474, "grad_norm": 3.566791296005249, "learning_rate": 1.486118303880495e-06, "loss": 1.0218, "step": 38206 }, { "epoch": 2.9942789968652037, "grad_norm": 2.4292051792144775, "learning_rate": 1.485898938444647e-06, "loss": 1.0432, "step": 38207 }, { "epoch": 2.99435736677116, "grad_norm": 2.7041335105895996, "learning_rate": 1.4856795863746332e-06, "loss": 1.3755, "step": 38208 }, { "epoch": 2.994435736677116, "grad_norm": 2.0243113040924072, "learning_rate": 1.4854602476712892e-06, "loss": 1.3321, "step": 38209 }, { "epoch": 2.9945141065830723, "grad_norm": 3.777409315109253, "learning_rate": 1.4852409223354492e-06, "loss": 1.4009, "step": 38210 }, { "epoch": 2.994592476489028, "grad_norm": 3.537376642227173, "learning_rate": 1.485021610367947e-06, "loss": 1.305, "step": 38211 }, { "epoch": 2.9946708463949845, "grad_norm": 4.818332672119141, "learning_rate": 1.4848023117696181e-06, "loss": 1.5023, "step": 38212 }, { "epoch": 2.9947492163009404, "grad_norm": 3.0737643241882324, "learning_rate": 1.4845830265412947e-06, "loss": 1.1383, "step": 38213 }, { "epoch": 2.9948275862068967, "grad_norm": 4.406199932098389, "learning_rate": 1.4843637546838124e-06, "loss": 1.4797, "step": 38214 }, { "epoch": 2.9949059561128526, "grad_norm": 2.746060371398926, "learning_rate": 1.4841444961980045e-06, "loss": 1.4665, "step": 38215 }, { "epoch": 2.994984326018809, "grad_norm": 12.74088191986084, "learning_rate": 1.483925251084705e-06, "loss": 1.2368, "step": 38216 }, { "epoch": 2.995062695924765, "grad_norm": 2.3540515899658203, "learning_rate": 1.4837060193447483e-06, "loss": 1.1543, "step": 38217 }, { "epoch": 2.995141065830721, "grad_norm": 5.004845142364502, "learning_rate": 1.4834868009789677e-06, "loss": 1.0193, "step": 38218 }, { "epoch": 2.995219435736677, "grad_norm": 3.451606035232544, "learning_rate": 1.4832675959881964e-06, "loss": 1.1607, "step": 38219 }, { "epoch": 2.9952978056426334, "grad_norm": 4.668771743774414, "learning_rate": 1.4830484043732695e-06, "loss": 1.1628, "step": 38220 }, { "epoch": 2.9953761755485893, "grad_norm": 5.418859481811523, "learning_rate": 1.4828292261350202e-06, "loss": 1.986, "step": 38221 }, { "epoch": 2.9954545454545456, "grad_norm": 5.3069658279418945, "learning_rate": 1.4826100612742832e-06, "loss": 1.4803, "step": 38222 }, { "epoch": 2.9955329153605015, "grad_norm": 3.6531014442443848, "learning_rate": 1.4823909097918877e-06, "loss": 1.6324, "step": 38223 }, { "epoch": 2.995611285266458, "grad_norm": 2.407116174697876, "learning_rate": 1.4821717716886718e-06, "loss": 1.333, "step": 38224 }, { "epoch": 2.9956896551724137, "grad_norm": 3.1134755611419678, "learning_rate": 1.4819526469654699e-06, "loss": 1.3635, "step": 38225 }, { "epoch": 2.99576802507837, "grad_norm": 3.640566110610962, "learning_rate": 1.481733535623111e-06, "loss": 1.3378, "step": 38226 }, { "epoch": 2.995846394984326, "grad_norm": 6.454837322235107, "learning_rate": 1.4815144376624307e-06, "loss": 1.8742, "step": 38227 }, { "epoch": 2.9959247648902823, "grad_norm": 6.552382469177246, "learning_rate": 1.481295353084261e-06, "loss": 1.5453, "step": 38228 }, { "epoch": 2.996003134796238, "grad_norm": 3.0562963485717773, "learning_rate": 1.4810762818894392e-06, "loss": 1.0981, "step": 38229 }, { "epoch": 2.9960815047021945, "grad_norm": 2.3621761798858643, "learning_rate": 1.4808572240787943e-06, "loss": 1.4342, "step": 38230 }, { "epoch": 2.9961598746081504, "grad_norm": 3.352299213409424, "learning_rate": 1.4806381796531617e-06, "loss": 1.2694, "step": 38231 }, { "epoch": 2.9962382445141067, "grad_norm": 5.290422439575195, "learning_rate": 1.4804191486133728e-06, "loss": 1.3111, "step": 38232 }, { "epoch": 2.9963166144200626, "grad_norm": 4.196145057678223, "learning_rate": 1.480200130960262e-06, "loss": 1.2353, "step": 38233 }, { "epoch": 2.996394984326019, "grad_norm": 2.610767364501953, "learning_rate": 1.4799811266946623e-06, "loss": 1.1386, "step": 38234 }, { "epoch": 2.996473354231975, "grad_norm": 3.655179977416992, "learning_rate": 1.4797621358174062e-06, "loss": 1.4322, "step": 38235 }, { "epoch": 2.996551724137931, "grad_norm": 3.65887451171875, "learning_rate": 1.4795431583293286e-06, "loss": 1.2097, "step": 38236 }, { "epoch": 2.996630094043887, "grad_norm": 6.681095600128174, "learning_rate": 1.479324194231257e-06, "loss": 1.0734, "step": 38237 }, { "epoch": 2.9967084639498434, "grad_norm": 2.226719379425049, "learning_rate": 1.4791052435240299e-06, "loss": 1.5446, "step": 38238 }, { "epoch": 2.9967868338557992, "grad_norm": 3.6953585147857666, "learning_rate": 1.4788863062084775e-06, "loss": 1.3474, "step": 38239 }, { "epoch": 2.9968652037617556, "grad_norm": 13.511762619018555, "learning_rate": 1.4786673822854353e-06, "loss": 1.8305, "step": 38240 }, { "epoch": 2.9969435736677115, "grad_norm": 3.4495797157287598, "learning_rate": 1.4784484717557313e-06, "loss": 1.2617, "step": 38241 }, { "epoch": 2.997021943573668, "grad_norm": 3.2196543216705322, "learning_rate": 1.4782295746201997e-06, "loss": 1.0405, "step": 38242 }, { "epoch": 2.9971003134796237, "grad_norm": 2.3842763900756836, "learning_rate": 1.478010690879675e-06, "loss": 1.2014, "step": 38243 }, { "epoch": 2.99717868338558, "grad_norm": 3.8736824989318848, "learning_rate": 1.4777918205349905e-06, "loss": 1.231, "step": 38244 }, { "epoch": 2.997257053291536, "grad_norm": 4.368835926055908, "learning_rate": 1.4775729635869751e-06, "loss": 1.5593, "step": 38245 }, { "epoch": 2.9973354231974922, "grad_norm": 3.0046815872192383, "learning_rate": 1.4773541200364627e-06, "loss": 1.1593, "step": 38246 }, { "epoch": 2.997413793103448, "grad_norm": 9.21158504486084, "learning_rate": 1.4771352898842844e-06, "loss": 0.5604, "step": 38247 }, { "epoch": 2.9974921630094045, "grad_norm": 6.120553016662598, "learning_rate": 1.4769164731312773e-06, "loss": 1.3322, "step": 38248 }, { "epoch": 2.9975705329153604, "grad_norm": 5.0226149559021, "learning_rate": 1.4766976697782686e-06, "loss": 1.059, "step": 38249 }, { "epoch": 2.9976489028213167, "grad_norm": 2.584817886352539, "learning_rate": 1.4764788798260927e-06, "loss": 1.233, "step": 38250 }, { "epoch": 2.9977272727272726, "grad_norm": 3.399893045425415, "learning_rate": 1.4762601032755809e-06, "loss": 1.3977, "step": 38251 }, { "epoch": 2.997805642633229, "grad_norm": 3.1624391078948975, "learning_rate": 1.4760413401275659e-06, "loss": 1.3887, "step": 38252 }, { "epoch": 2.997884012539185, "grad_norm": 2.528818130493164, "learning_rate": 1.47582259038288e-06, "loss": 1.4713, "step": 38253 }, { "epoch": 2.997962382445141, "grad_norm": 9.07651424407959, "learning_rate": 1.4756038540423539e-06, "loss": 1.8588, "step": 38254 }, { "epoch": 2.998040752351097, "grad_norm": 6.366568088531494, "learning_rate": 1.475385131106823e-06, "loss": 0.9149, "step": 38255 }, { "epoch": 2.9981191222570533, "grad_norm": 2.571967601776123, "learning_rate": 1.4751664215771133e-06, "loss": 1.133, "step": 38256 }, { "epoch": 2.9981974921630092, "grad_norm": 2.457364320755005, "learning_rate": 1.4749477254540612e-06, "loss": 1.2528, "step": 38257 }, { "epoch": 2.9982758620689656, "grad_norm": 4.679173946380615, "learning_rate": 1.4747290427384981e-06, "loss": 1.2657, "step": 38258 }, { "epoch": 2.998354231974922, "grad_norm": 2.445025682449341, "learning_rate": 1.474510373431256e-06, "loss": 1.1962, "step": 38259 }, { "epoch": 2.998432601880878, "grad_norm": 3.8760156631469727, "learning_rate": 1.4742917175331644e-06, "loss": 1.109, "step": 38260 }, { "epoch": 2.9985109717868337, "grad_norm": 2.8649699687957764, "learning_rate": 1.4740730750450544e-06, "loss": 1.2854, "step": 38261 }, { "epoch": 2.99858934169279, "grad_norm": 2.416975498199463, "learning_rate": 1.4738544459677612e-06, "loss": 1.2112, "step": 38262 }, { "epoch": 2.9986677115987463, "grad_norm": 2.2574405670166016, "learning_rate": 1.4736358303021158e-06, "loss": 1.4383, "step": 38263 }, { "epoch": 2.9987460815047022, "grad_norm": 3.1488866806030273, "learning_rate": 1.4734172280489466e-06, "loss": 1.2467, "step": 38264 }, { "epoch": 2.998824451410658, "grad_norm": 4.545670032501221, "learning_rate": 1.4731986392090875e-06, "loss": 1.1689, "step": 38265 }, { "epoch": 2.9989028213166145, "grad_norm": 4.213961124420166, "learning_rate": 1.472980063783369e-06, "loss": 1.4919, "step": 38266 }, { "epoch": 2.998981191222571, "grad_norm": 4.037112236022949, "learning_rate": 1.4727615017726231e-06, "loss": 1.648, "step": 38267 }, { "epoch": 2.9990595611285267, "grad_norm": 5.151053428649902, "learning_rate": 1.4725429531776808e-06, "loss": 0.6542, "step": 38268 }, { "epoch": 2.9991379310344826, "grad_norm": 2.8591747283935547, "learning_rate": 1.472324417999373e-06, "loss": 0.874, "step": 38269 }, { "epoch": 2.999216300940439, "grad_norm": 2.724159002304077, "learning_rate": 1.4721058962385309e-06, "loss": 1.0025, "step": 38270 }, { "epoch": 2.9992946708463952, "grad_norm": 4.394558429718018, "learning_rate": 1.4718873878959867e-06, "loss": 0.8992, "step": 38271 }, { "epoch": 2.999373040752351, "grad_norm": 6.203212738037109, "learning_rate": 1.4716688929725698e-06, "loss": 1.3569, "step": 38272 }, { "epoch": 2.999451410658307, "grad_norm": 4.041024208068848, "learning_rate": 1.4714504114691125e-06, "loss": 1.4453, "step": 38273 }, { "epoch": 2.9995297805642633, "grad_norm": 1.9274592399597168, "learning_rate": 1.4712319433864475e-06, "loss": 1.0939, "step": 38274 }, { "epoch": 2.9996081504702197, "grad_norm": 5.466106414794922, "learning_rate": 1.4710134887254003e-06, "loss": 2.0075, "step": 38275 }, { "epoch": 2.9996865203761756, "grad_norm": 3.0337493419647217, "learning_rate": 1.4707950474868067e-06, "loss": 1.3779, "step": 38276 }, { "epoch": 2.9997648902821314, "grad_norm": 3.126574993133545, "learning_rate": 1.4705766196714955e-06, "loss": 1.1783, "step": 38277 }, { "epoch": 2.999843260188088, "grad_norm": 2.745734453201294, "learning_rate": 1.4703582052803e-06, "loss": 1.381, "step": 38278 }, { "epoch": 2.999921630094044, "grad_norm": 2.720890998840332, "learning_rate": 1.4701398043140475e-06, "loss": 1.6554, "step": 38279 }, { "epoch": 3.0, "grad_norm": 3.8691301345825195, "learning_rate": 1.4699214167735688e-06, "loss": 1.5866, "step": 38280 }, { "epoch": 3.0, "eval_loss": 1.247266173362732, "eval_runtime": 110.6741, "eval_samples_per_second": 24.278, "eval_steps_per_second": 6.072, "step": 38280 }, { "epoch": 3.0000783699059563, "grad_norm": 4.13778829574585, "learning_rate": 1.4697030426596986e-06, "loss": 1.4804, "step": 38281 }, { "epoch": 3.000156739811912, "grad_norm": 2.690119743347168, "learning_rate": 1.4694846819732634e-06, "loss": 1.2277, "step": 38282 }, { "epoch": 3.0002351097178686, "grad_norm": 2.924550771713257, "learning_rate": 1.4692663347150948e-06, "loss": 1.5634, "step": 38283 }, { "epoch": 3.0003134796238244, "grad_norm": 2.52730655670166, "learning_rate": 1.4690480008860237e-06, "loss": 1.5758, "step": 38284 }, { "epoch": 3.0003918495297808, "grad_norm": 4.17989444732666, "learning_rate": 1.4688296804868806e-06, "loss": 1.3481, "step": 38285 }, { "epoch": 3.0004702194357367, "grad_norm": 3.5046372413635254, "learning_rate": 1.4686113735184954e-06, "loss": 1.2015, "step": 38286 }, { "epoch": 3.000548589341693, "grad_norm": 3.8433685302734375, "learning_rate": 1.4683930799816987e-06, "loss": 1.431, "step": 38287 }, { "epoch": 3.000626959247649, "grad_norm": 4.830215930938721, "learning_rate": 1.468174799877321e-06, "loss": 0.9417, "step": 38288 }, { "epoch": 3.000705329153605, "grad_norm": 2.7270703315734863, "learning_rate": 1.4679565332061923e-06, "loss": 1.2169, "step": 38289 }, { "epoch": 3.000783699059561, "grad_norm": 5.72693395614624, "learning_rate": 1.4677382799691425e-06, "loss": 1.5989, "step": 38290 }, { "epoch": 3.0008620689655174, "grad_norm": 3.9625704288482666, "learning_rate": 1.4675200401670025e-06, "loss": 1.1607, "step": 38291 }, { "epoch": 3.0009404388714733, "grad_norm": 3.062894105911255, "learning_rate": 1.4673018138006029e-06, "loss": 1.486, "step": 38292 }, { "epoch": 3.0010188087774297, "grad_norm": 5.421639919281006, "learning_rate": 1.4670836008707712e-06, "loss": 0.8954, "step": 38293 }, { "epoch": 3.0010971786833855, "grad_norm": 2.8432674407958984, "learning_rate": 1.4668654013783373e-06, "loss": 1.5553, "step": 38294 }, { "epoch": 3.001175548589342, "grad_norm": 2.901475191116333, "learning_rate": 1.4666472153241347e-06, "loss": 1.2917, "step": 38295 }, { "epoch": 3.0012539184952978, "grad_norm": 7.394567489624023, "learning_rate": 1.4664290427089921e-06, "loss": 1.4339, "step": 38296 }, { "epoch": 3.001332288401254, "grad_norm": 2.642749547958374, "learning_rate": 1.4662108835337375e-06, "loss": 1.4836, "step": 38297 }, { "epoch": 3.00141065830721, "grad_norm": 2.3659958839416504, "learning_rate": 1.4659927377992017e-06, "loss": 1.1919, "step": 38298 }, { "epoch": 3.0014890282131663, "grad_norm": 3.631821870803833, "learning_rate": 1.4657746055062122e-06, "loss": 0.7806, "step": 38299 }, { "epoch": 3.001567398119122, "grad_norm": 9.338506698608398, "learning_rate": 1.4655564866556043e-06, "loss": 1.266, "step": 38300 }, { "epoch": 3.0016457680250785, "grad_norm": 2.167398691177368, "learning_rate": 1.4653383812482024e-06, "loss": 0.9907, "step": 38301 }, { "epoch": 3.0017241379310344, "grad_norm": 2.739227771759033, "learning_rate": 1.465120289284838e-06, "loss": 1.1907, "step": 38302 }, { "epoch": 3.0018025078369908, "grad_norm": 4.135947227478027, "learning_rate": 1.4649022107663403e-06, "loss": 0.8689, "step": 38303 }, { "epoch": 3.0018808777429467, "grad_norm": 2.5544798374176025, "learning_rate": 1.4646841456935384e-06, "loss": 1.2332, "step": 38304 }, { "epoch": 3.001959247648903, "grad_norm": 3.7600579261779785, "learning_rate": 1.4644660940672628e-06, "loss": 1.423, "step": 38305 }, { "epoch": 3.002037617554859, "grad_norm": 3.338406562805176, "learning_rate": 1.4642480558883421e-06, "loss": 1.1055, "step": 38306 }, { "epoch": 3.002115987460815, "grad_norm": 12.786100387573242, "learning_rate": 1.4640300311576056e-06, "loss": 1.1497, "step": 38307 }, { "epoch": 3.002194357366771, "grad_norm": 14.429582595825195, "learning_rate": 1.4638120198758826e-06, "loss": 1.8546, "step": 38308 }, { "epoch": 3.0022727272727274, "grad_norm": 4.157233715057373, "learning_rate": 1.4635940220440025e-06, "loss": 1.3491, "step": 38309 }, { "epoch": 3.0023510971786833, "grad_norm": 3.1092891693115234, "learning_rate": 1.4633760376627942e-06, "loss": 1.4859, "step": 38310 }, { "epoch": 3.0024294670846396, "grad_norm": 8.47167682647705, "learning_rate": 1.4631580667330885e-06, "loss": 1.5389, "step": 38311 }, { "epoch": 3.0025078369905955, "grad_norm": 2.8327364921569824, "learning_rate": 1.46294010925571e-06, "loss": 1.5355, "step": 38312 }, { "epoch": 3.002586206896552, "grad_norm": 3.6893069744110107, "learning_rate": 1.4627221652314916e-06, "loss": 1.5751, "step": 38313 }, { "epoch": 3.0026645768025078, "grad_norm": 3.3682572841644287, "learning_rate": 1.4625042346612618e-06, "loss": 1.4141, "step": 38314 }, { "epoch": 3.002742946708464, "grad_norm": 2.4073944091796875, "learning_rate": 1.4622863175458502e-06, "loss": 1.4055, "step": 38315 }, { "epoch": 3.00282131661442, "grad_norm": 3.242295503616333, "learning_rate": 1.462068413886083e-06, "loss": 1.4332, "step": 38316 }, { "epoch": 3.0028996865203763, "grad_norm": 3.9027528762817383, "learning_rate": 1.4618505236827885e-06, "loss": 1.5391, "step": 38317 }, { "epoch": 3.002978056426332, "grad_norm": 4.472753524780273, "learning_rate": 1.4616326469367991e-06, "loss": 1.5044, "step": 38318 }, { "epoch": 3.0030564263322885, "grad_norm": 3.75012469291687, "learning_rate": 1.461414783648943e-06, "loss": 1.2134, "step": 38319 }, { "epoch": 3.0031347962382444, "grad_norm": 2.5080041885375977, "learning_rate": 1.461196933820046e-06, "loss": 1.3171, "step": 38320 }, { "epoch": 3.0032131661442008, "grad_norm": 3.8213489055633545, "learning_rate": 1.4609790974509385e-06, "loss": 1.4683, "step": 38321 }, { "epoch": 3.0032915360501566, "grad_norm": 3.354158639907837, "learning_rate": 1.4607612745424488e-06, "loss": 1.2486, "step": 38322 }, { "epoch": 3.003369905956113, "grad_norm": 4.1460981369018555, "learning_rate": 1.4605434650954048e-06, "loss": 1.4987, "step": 38323 }, { "epoch": 3.003448275862069, "grad_norm": 2.451085090637207, "learning_rate": 1.460325669110636e-06, "loss": 0.9247, "step": 38324 }, { "epoch": 3.003526645768025, "grad_norm": 5.6413350105285645, "learning_rate": 1.46010788658897e-06, "loss": 1.1908, "step": 38325 }, { "epoch": 3.003605015673981, "grad_norm": 2.6863620281219482, "learning_rate": 1.4598901175312358e-06, "loss": 1.4491, "step": 38326 }, { "epoch": 3.0036833855799374, "grad_norm": 4.0342183113098145, "learning_rate": 1.4596723619382607e-06, "loss": 1.3602, "step": 38327 }, { "epoch": 3.0037617554858933, "grad_norm": 3.8801894187927246, "learning_rate": 1.459454619810874e-06, "loss": 1.4375, "step": 38328 }, { "epoch": 3.0038401253918496, "grad_norm": 3.134408712387085, "learning_rate": 1.4592368911499028e-06, "loss": 1.4364, "step": 38329 }, { "epoch": 3.0039184952978055, "grad_norm": 2.8025736808776855, "learning_rate": 1.4590191759561777e-06, "loss": 1.4481, "step": 38330 }, { "epoch": 3.003996865203762, "grad_norm": 2.820039987564087, "learning_rate": 1.4588014742305217e-06, "loss": 1.3963, "step": 38331 }, { "epoch": 3.0040752351097177, "grad_norm": 5.3241496086120605, "learning_rate": 1.4585837859737678e-06, "loss": 1.2742, "step": 38332 }, { "epoch": 3.004153605015674, "grad_norm": 6.8171000480651855, "learning_rate": 1.4583661111867437e-06, "loss": 1.967, "step": 38333 }, { "epoch": 3.00423197492163, "grad_norm": 5.474855422973633, "learning_rate": 1.4581484498702742e-06, "loss": 0.8224, "step": 38334 }, { "epoch": 3.0043103448275863, "grad_norm": 4.586028575897217, "learning_rate": 1.4579308020251886e-06, "loss": 1.301, "step": 38335 }, { "epoch": 3.004388714733542, "grad_norm": 5.14980411529541, "learning_rate": 1.4577131676523137e-06, "loss": 1.5505, "step": 38336 }, { "epoch": 3.0044670846394985, "grad_norm": 2.0274465084075928, "learning_rate": 1.4574955467524816e-06, "loss": 1.3352, "step": 38337 }, { "epoch": 3.0045454545454544, "grad_norm": 2.2780957221984863, "learning_rate": 1.4572779393265152e-06, "loss": 1.3963, "step": 38338 }, { "epoch": 3.0046238244514107, "grad_norm": 2.770592212677002, "learning_rate": 1.4570603453752435e-06, "loss": 0.9898, "step": 38339 }, { "epoch": 3.0047021943573666, "grad_norm": 4.476892948150635, "learning_rate": 1.4568427648994949e-06, "loss": 1.2873, "step": 38340 }, { "epoch": 3.004780564263323, "grad_norm": 2.3399531841278076, "learning_rate": 1.4566251979000968e-06, "loss": 1.4904, "step": 38341 }, { "epoch": 3.004858934169279, "grad_norm": 2.3488736152648926, "learning_rate": 1.4564076443778757e-06, "loss": 1.4181, "step": 38342 }, { "epoch": 3.004937304075235, "grad_norm": 2.737730026245117, "learning_rate": 1.45619010433366e-06, "loss": 1.3748, "step": 38343 }, { "epoch": 3.005015673981191, "grad_norm": 2.237055778503418, "learning_rate": 1.4559725777682776e-06, "loss": 1.1352, "step": 38344 }, { "epoch": 3.0050940438871474, "grad_norm": 10.292824745178223, "learning_rate": 1.4557550646825548e-06, "loss": 1.2564, "step": 38345 }, { "epoch": 3.0051724137931033, "grad_norm": 3.6970019340515137, "learning_rate": 1.455537565077319e-06, "loss": 1.2873, "step": 38346 }, { "epoch": 3.0052507836990596, "grad_norm": 3.176271677017212, "learning_rate": 1.4553200789533977e-06, "loss": 1.4963, "step": 38347 }, { "epoch": 3.0053291536050155, "grad_norm": 2.7118940353393555, "learning_rate": 1.4551026063116202e-06, "loss": 1.3829, "step": 38348 }, { "epoch": 3.005407523510972, "grad_norm": 2.8447425365448, "learning_rate": 1.4548851471528097e-06, "loss": 1.109, "step": 38349 }, { "epoch": 3.0054858934169277, "grad_norm": 5.739332675933838, "learning_rate": 1.4546677014777938e-06, "loss": 1.4708, "step": 38350 }, { "epoch": 3.005564263322884, "grad_norm": 6.859142303466797, "learning_rate": 1.4544502692874029e-06, "loss": 1.3324, "step": 38351 }, { "epoch": 3.00564263322884, "grad_norm": 3.978178024291992, "learning_rate": 1.4542328505824632e-06, "loss": 1.1212, "step": 38352 }, { "epoch": 3.0057210031347963, "grad_norm": 3.71481990814209, "learning_rate": 1.454015445363799e-06, "loss": 1.4178, "step": 38353 }, { "epoch": 3.005799373040752, "grad_norm": 2.7977030277252197, "learning_rate": 1.4537980536322387e-06, "loss": 1.3423, "step": 38354 }, { "epoch": 3.0058777429467085, "grad_norm": 3.30212140083313, "learning_rate": 1.4535806753886073e-06, "loss": 1.2983, "step": 38355 }, { "epoch": 3.0059561128526644, "grad_norm": 2.9778225421905518, "learning_rate": 1.453363310633737e-06, "loss": 1.165, "step": 38356 }, { "epoch": 3.0060344827586207, "grad_norm": 2.91902756690979, "learning_rate": 1.4531459593684495e-06, "loss": 1.2591, "step": 38357 }, { "epoch": 3.0061128526645766, "grad_norm": 2.5214529037475586, "learning_rate": 1.4529286215935728e-06, "loss": 1.2975, "step": 38358 }, { "epoch": 3.006191222570533, "grad_norm": 3.3195724487304688, "learning_rate": 1.4527112973099339e-06, "loss": 1.4104, "step": 38359 }, { "epoch": 3.006269592476489, "grad_norm": 4.86546516418457, "learning_rate": 1.452493986518359e-06, "loss": 1.4964, "step": 38360 }, { "epoch": 3.006347962382445, "grad_norm": 3.052358627319336, "learning_rate": 1.452276689219675e-06, "loss": 0.9842, "step": 38361 }, { "epoch": 3.006426332288401, "grad_norm": 3.0154058933258057, "learning_rate": 1.452059405414708e-06, "loss": 1.2451, "step": 38362 }, { "epoch": 3.0065047021943574, "grad_norm": 7.065099239349365, "learning_rate": 1.4518421351042866e-06, "loss": 1.6564, "step": 38363 }, { "epoch": 3.0065830721003133, "grad_norm": 3.051208972930908, "learning_rate": 1.4516248782892317e-06, "loss": 1.2884, "step": 38364 }, { "epoch": 3.0066614420062696, "grad_norm": 4.699494361877441, "learning_rate": 1.4514076349703749e-06, "loss": 1.4216, "step": 38365 }, { "epoch": 3.0067398119122255, "grad_norm": 5.988713264465332, "learning_rate": 1.451190405148541e-06, "loss": 0.5683, "step": 38366 }, { "epoch": 3.006818181818182, "grad_norm": 2.5134365558624268, "learning_rate": 1.4509731888245577e-06, "loss": 1.1927, "step": 38367 }, { "epoch": 3.0068965517241377, "grad_norm": 3.235818386077881, "learning_rate": 1.4507559859992476e-06, "loss": 1.5941, "step": 38368 }, { "epoch": 3.006974921630094, "grad_norm": 7.702361106872559, "learning_rate": 1.4505387966734374e-06, "loss": 1.0916, "step": 38369 }, { "epoch": 3.00705329153605, "grad_norm": 3.010676383972168, "learning_rate": 1.450321620847956e-06, "loss": 1.2053, "step": 38370 }, { "epoch": 3.0071316614420063, "grad_norm": 4.004420280456543, "learning_rate": 1.450104458523629e-06, "loss": 1.1706, "step": 38371 }, { "epoch": 3.007210031347962, "grad_norm": 4.024418354034424, "learning_rate": 1.4498873097012805e-06, "loss": 1.304, "step": 38372 }, { "epoch": 3.0072884012539185, "grad_norm": 4.1613078117370605, "learning_rate": 1.449670174381737e-06, "loss": 1.3907, "step": 38373 }, { "epoch": 3.007366771159875, "grad_norm": 3.164293050765991, "learning_rate": 1.4494530525658229e-06, "loss": 1.7316, "step": 38374 }, { "epoch": 3.0074451410658307, "grad_norm": 3.590470790863037, "learning_rate": 1.4492359442543686e-06, "loss": 1.3209, "step": 38375 }, { "epoch": 3.007523510971787, "grad_norm": 3.4139904975891113, "learning_rate": 1.449018849448196e-06, "loss": 1.4636, "step": 38376 }, { "epoch": 3.007601880877743, "grad_norm": 3.4203336238861084, "learning_rate": 1.4488017681481315e-06, "loss": 1.0274, "step": 38377 }, { "epoch": 3.0076802507836993, "grad_norm": 3.6218602657318115, "learning_rate": 1.4485847003550008e-06, "loss": 1.2854, "step": 38378 }, { "epoch": 3.007758620689655, "grad_norm": 4.054335117340088, "learning_rate": 1.44836764606963e-06, "loss": 1.4468, "step": 38379 }, { "epoch": 3.0078369905956115, "grad_norm": 2.5470921993255615, "learning_rate": 1.4481506052928445e-06, "loss": 1.4288, "step": 38380 }, { "epoch": 3.0079153605015674, "grad_norm": 11.218632698059082, "learning_rate": 1.44793357802547e-06, "loss": 0.8707, "step": 38381 }, { "epoch": 3.0079937304075237, "grad_norm": 3.5915074348449707, "learning_rate": 1.4477165642683327e-06, "loss": 1.1849, "step": 38382 }, { "epoch": 3.0080721003134796, "grad_norm": 2.9984536170959473, "learning_rate": 1.4474995640222545e-06, "loss": 1.1336, "step": 38383 }, { "epoch": 3.008150470219436, "grad_norm": 7.477191925048828, "learning_rate": 1.4472825772880644e-06, "loss": 1.3924, "step": 38384 }, { "epoch": 3.008228840125392, "grad_norm": 4.1209001541137695, "learning_rate": 1.4470656040665866e-06, "loss": 1.2883, "step": 38385 }, { "epoch": 3.008307210031348, "grad_norm": 3.5073349475860596, "learning_rate": 1.4468486443586476e-06, "loss": 1.0647, "step": 38386 }, { "epoch": 3.008385579937304, "grad_norm": 3.593862771987915, "learning_rate": 1.4466316981650707e-06, "loss": 1.1237, "step": 38387 }, { "epoch": 3.0084639498432604, "grad_norm": 4.745782852172852, "learning_rate": 1.4464147654866794e-06, "loss": 1.4684, "step": 38388 }, { "epoch": 3.0085423197492163, "grad_norm": 2.4504714012145996, "learning_rate": 1.4461978463243042e-06, "loss": 1.1735, "step": 38389 }, { "epoch": 3.0086206896551726, "grad_norm": 3.4271559715270996, "learning_rate": 1.4459809406787661e-06, "loss": 1.2115, "step": 38390 }, { "epoch": 3.0086990595611285, "grad_norm": 3.1292998790740967, "learning_rate": 1.4457640485508906e-06, "loss": 1.6228, "step": 38391 }, { "epoch": 3.008777429467085, "grad_norm": 4.925736427307129, "learning_rate": 1.445547169941503e-06, "loss": 1.7004, "step": 38392 }, { "epoch": 3.0088557993730407, "grad_norm": 4.7433671951293945, "learning_rate": 1.4453303048514284e-06, "loss": 1.6313, "step": 38393 }, { "epoch": 3.008934169278997, "grad_norm": 2.882164478302002, "learning_rate": 1.445113453281492e-06, "loss": 1.4424, "step": 38394 }, { "epoch": 3.009012539184953, "grad_norm": 3.6789791584014893, "learning_rate": 1.4448966152325171e-06, "loss": 0.8199, "step": 38395 }, { "epoch": 3.0090909090909093, "grad_norm": 4.003334045410156, "learning_rate": 1.4446797907053306e-06, "loss": 1.4978, "step": 38396 }, { "epoch": 3.009169278996865, "grad_norm": 3.712733745574951, "learning_rate": 1.4444629797007554e-06, "loss": 1.6636, "step": 38397 }, { "epoch": 3.0092476489028215, "grad_norm": 2.683835744857788, "learning_rate": 1.444246182219617e-06, "loss": 1.5031, "step": 38398 }, { "epoch": 3.0093260188087774, "grad_norm": 2.463279962539673, "learning_rate": 1.4440293982627396e-06, "loss": 1.5261, "step": 38399 }, { "epoch": 3.0094043887147337, "grad_norm": 4.937563896179199, "learning_rate": 1.4438126278309478e-06, "loss": 1.5254, "step": 38400 }, { "epoch": 3.0094827586206896, "grad_norm": 4.222713470458984, "learning_rate": 1.4435958709250664e-06, "loss": 1.596, "step": 38401 }, { "epoch": 3.009561128526646, "grad_norm": 2.5015578269958496, "learning_rate": 1.4433791275459197e-06, "loss": 1.253, "step": 38402 }, { "epoch": 3.009639498432602, "grad_norm": 2.780754327774048, "learning_rate": 1.4431623976943316e-06, "loss": 1.1864, "step": 38403 }, { "epoch": 3.009717868338558, "grad_norm": 3.1453280448913574, "learning_rate": 1.4429456813711285e-06, "loss": 0.9826, "step": 38404 }, { "epoch": 3.009796238244514, "grad_norm": 3.1175010204315186, "learning_rate": 1.442728978577132e-06, "loss": 1.5812, "step": 38405 }, { "epoch": 3.0098746081504704, "grad_norm": 5.191271781921387, "learning_rate": 1.4425122893131665e-06, "loss": 1.2116, "step": 38406 }, { "epoch": 3.0099529780564263, "grad_norm": 6.268819808959961, "learning_rate": 1.4422956135800564e-06, "loss": 1.4131, "step": 38407 }, { "epoch": 3.0100313479623826, "grad_norm": 2.805016040802002, "learning_rate": 1.4420789513786287e-06, "loss": 0.9905, "step": 38408 }, { "epoch": 3.0101097178683385, "grad_norm": 6.325456142425537, "learning_rate": 1.4418623027097039e-06, "loss": 1.1302, "step": 38409 }, { "epoch": 3.010188087774295, "grad_norm": 2.7477798461914062, "learning_rate": 1.4416456675741076e-06, "loss": 1.2217, "step": 38410 }, { "epoch": 3.0102664576802507, "grad_norm": 8.135969161987305, "learning_rate": 1.4414290459726633e-06, "loss": 1.7376, "step": 38411 }, { "epoch": 3.010344827586207, "grad_norm": 3.2099156379699707, "learning_rate": 1.4412124379061954e-06, "loss": 1.5374, "step": 38412 }, { "epoch": 3.010423197492163, "grad_norm": 3.198242664337158, "learning_rate": 1.4409958433755273e-06, "loss": 1.2247, "step": 38413 }, { "epoch": 3.0105015673981192, "grad_norm": 3.444098949432373, "learning_rate": 1.4407792623814827e-06, "loss": 1.1328, "step": 38414 }, { "epoch": 3.010579937304075, "grad_norm": 2.8674368858337402, "learning_rate": 1.4405626949248857e-06, "loss": 1.4447, "step": 38415 }, { "epoch": 3.0106583072100315, "grad_norm": 3.0301198959350586, "learning_rate": 1.44034614100656e-06, "loss": 1.4921, "step": 38416 }, { "epoch": 3.0107366771159874, "grad_norm": 4.325455188751221, "learning_rate": 1.4401296006273298e-06, "loss": 1.4523, "step": 38417 }, { "epoch": 3.0108150470219437, "grad_norm": 3.349825382232666, "learning_rate": 1.4399130737880175e-06, "loss": 1.2355, "step": 38418 }, { "epoch": 3.0108934169278996, "grad_norm": 3.863074541091919, "learning_rate": 1.439696560489449e-06, "loss": 1.3803, "step": 38419 }, { "epoch": 3.010971786833856, "grad_norm": 3.9316980838775635, "learning_rate": 1.4394800607324428e-06, "loss": 1.2914, "step": 38420 }, { "epoch": 3.011050156739812, "grad_norm": 2.868875026702881, "learning_rate": 1.4392635745178274e-06, "loss": 1.3356, "step": 38421 }, { "epoch": 3.011128526645768, "grad_norm": 4.80954647064209, "learning_rate": 1.4390471018464248e-06, "loss": 1.7141, "step": 38422 }, { "epoch": 3.011206896551724, "grad_norm": 4.36609411239624, "learning_rate": 1.438830642719059e-06, "loss": 1.2181, "step": 38423 }, { "epoch": 3.0112852664576804, "grad_norm": 9.97518253326416, "learning_rate": 1.4386141971365508e-06, "loss": 0.7851, "step": 38424 }, { "epoch": 3.0113636363636362, "grad_norm": 3.774664878845215, "learning_rate": 1.4383977650997238e-06, "loss": 1.5084, "step": 38425 }, { "epoch": 3.0114420062695926, "grad_norm": 2.637176275253296, "learning_rate": 1.4381813466094035e-06, "loss": 1.4954, "step": 38426 }, { "epoch": 3.0115203761755485, "grad_norm": 17.033523559570312, "learning_rate": 1.4379649416664138e-06, "loss": 1.3408, "step": 38427 }, { "epoch": 3.011598746081505, "grad_norm": 7.212040424346924, "learning_rate": 1.4377485502715744e-06, "loss": 1.3813, "step": 38428 }, { "epoch": 3.0116771159874607, "grad_norm": 4.117311477661133, "learning_rate": 1.4375321724257097e-06, "loss": 1.6844, "step": 38429 }, { "epoch": 3.011755485893417, "grad_norm": 2.8498291969299316, "learning_rate": 1.4373158081296407e-06, "loss": 1.6917, "step": 38430 }, { "epoch": 3.011833855799373, "grad_norm": 4.257906913757324, "learning_rate": 1.4370994573841962e-06, "loss": 1.4839, "step": 38431 }, { "epoch": 3.0119122257053292, "grad_norm": 3.041363477706909, "learning_rate": 1.4368831201901933e-06, "loss": 1.4204, "step": 38432 }, { "epoch": 3.011990595611285, "grad_norm": 3.5261318683624268, "learning_rate": 1.4366667965484576e-06, "loss": 1.352, "step": 38433 }, { "epoch": 3.0120689655172415, "grad_norm": 8.113057136535645, "learning_rate": 1.4364504864598105e-06, "loss": 1.2983, "step": 38434 }, { "epoch": 3.0121473354231973, "grad_norm": 2.5804483890533447, "learning_rate": 1.436234189925076e-06, "loss": 1.2224, "step": 38435 }, { "epoch": 3.0122257053291537, "grad_norm": 2.744332790374756, "learning_rate": 1.4360179069450754e-06, "loss": 1.0115, "step": 38436 }, { "epoch": 3.0123040752351096, "grad_norm": 6.847692966461182, "learning_rate": 1.4358016375206324e-06, "loss": 1.7463, "step": 38437 }, { "epoch": 3.012382445141066, "grad_norm": 2.9170918464660645, "learning_rate": 1.4355853816525712e-06, "loss": 1.2948, "step": 38438 }, { "epoch": 3.012460815047022, "grad_norm": 2.528252124786377, "learning_rate": 1.435369139341709e-06, "loss": 1.0947, "step": 38439 }, { "epoch": 3.012539184952978, "grad_norm": 3.1666765213012695, "learning_rate": 1.4351529105888735e-06, "loss": 1.1219, "step": 38440 }, { "epoch": 3.012617554858934, "grad_norm": 2.504695415496826, "learning_rate": 1.434936695394885e-06, "loss": 1.374, "step": 38441 }, { "epoch": 3.0126959247648903, "grad_norm": 2.941046953201294, "learning_rate": 1.4347204937605675e-06, "loss": 1.47, "step": 38442 }, { "epoch": 3.0127742946708462, "grad_norm": 3.464674234390259, "learning_rate": 1.434504305686741e-06, "loss": 1.5638, "step": 38443 }, { "epoch": 3.0128526645768026, "grad_norm": 2.5876970291137695, "learning_rate": 1.4342881311742269e-06, "loss": 1.6902, "step": 38444 }, { "epoch": 3.0129310344827585, "grad_norm": 2.8445498943328857, "learning_rate": 1.4340719702238525e-06, "loss": 1.6672, "step": 38445 }, { "epoch": 3.013009404388715, "grad_norm": 2.385897397994995, "learning_rate": 1.4338558228364353e-06, "loss": 1.321, "step": 38446 }, { "epoch": 3.0130877742946707, "grad_norm": 3.231821298599243, "learning_rate": 1.4336396890127985e-06, "loss": 0.7951, "step": 38447 }, { "epoch": 3.013166144200627, "grad_norm": 3.5642502307891846, "learning_rate": 1.4334235687537651e-06, "loss": 1.653, "step": 38448 }, { "epoch": 3.013244514106583, "grad_norm": 4.77666711807251, "learning_rate": 1.4332074620601561e-06, "loss": 1.2534, "step": 38449 }, { "epoch": 3.0133228840125392, "grad_norm": 3.6858105659484863, "learning_rate": 1.432991368932794e-06, "loss": 1.4695, "step": 38450 }, { "epoch": 3.013401253918495, "grad_norm": 2.4676706790924072, "learning_rate": 1.4327752893725006e-06, "loss": 1.218, "step": 38451 }, { "epoch": 3.0134796238244514, "grad_norm": 2.590571165084839, "learning_rate": 1.4325592233800978e-06, "loss": 1.13, "step": 38452 }, { "epoch": 3.0135579937304073, "grad_norm": 3.2809512615203857, "learning_rate": 1.4323431709564073e-06, "loss": 1.2455, "step": 38453 }, { "epoch": 3.0136363636363637, "grad_norm": 2.856887102127075, "learning_rate": 1.4321271321022506e-06, "loss": 1.5126, "step": 38454 }, { "epoch": 3.0137147335423196, "grad_norm": 4.521632194519043, "learning_rate": 1.4319111068184499e-06, "loss": 1.1244, "step": 38455 }, { "epoch": 3.013793103448276, "grad_norm": 2.9111175537109375, "learning_rate": 1.4316950951058273e-06, "loss": 1.4509, "step": 38456 }, { "epoch": 3.013871473354232, "grad_norm": 3.959979295730591, "learning_rate": 1.4314790969652042e-06, "loss": 1.2383, "step": 38457 }, { "epoch": 3.013949843260188, "grad_norm": 4.068245887756348, "learning_rate": 1.4312631123973997e-06, "loss": 0.743, "step": 38458 }, { "epoch": 3.014028213166144, "grad_norm": 4.1516642570495605, "learning_rate": 1.4310471414032385e-06, "loss": 1.1157, "step": 38459 }, { "epoch": 3.0141065830721003, "grad_norm": 2.6526260375976562, "learning_rate": 1.4308311839835421e-06, "loss": 1.3614, "step": 38460 }, { "epoch": 3.014184952978056, "grad_norm": 3.3449997901916504, "learning_rate": 1.4306152401391293e-06, "loss": 1.7008, "step": 38461 }, { "epoch": 3.0142633228840126, "grad_norm": 2.504420280456543, "learning_rate": 1.4303993098708223e-06, "loss": 1.2391, "step": 38462 }, { "epoch": 3.0143416927899684, "grad_norm": 2.6143476963043213, "learning_rate": 1.4301833931794417e-06, "loss": 1.7761, "step": 38463 }, { "epoch": 3.0144200626959248, "grad_norm": 2.9439849853515625, "learning_rate": 1.4299674900658133e-06, "loss": 1.4863, "step": 38464 }, { "epoch": 3.0144984326018807, "grad_norm": 3.3278350830078125, "learning_rate": 1.4297516005307532e-06, "loss": 1.3881, "step": 38465 }, { "epoch": 3.014576802507837, "grad_norm": 3.861191749572754, "learning_rate": 1.4295357245750842e-06, "loss": 1.5845, "step": 38466 }, { "epoch": 3.014655172413793, "grad_norm": 6.542092800140381, "learning_rate": 1.4293198621996269e-06, "loss": 1.0085, "step": 38467 }, { "epoch": 3.014733542319749, "grad_norm": 2.654223680496216, "learning_rate": 1.429104013405203e-06, "loss": 1.3126, "step": 38468 }, { "epoch": 3.0148119122257055, "grad_norm": 3.655726909637451, "learning_rate": 1.428888178192634e-06, "loss": 1.2684, "step": 38469 }, { "epoch": 3.0148902821316614, "grad_norm": 3.451584815979004, "learning_rate": 1.4286723565627397e-06, "loss": 1.1384, "step": 38470 }, { "epoch": 3.0149686520376178, "grad_norm": 4.307087421417236, "learning_rate": 1.428456548516341e-06, "loss": 1.7457, "step": 38471 }, { "epoch": 3.0150470219435737, "grad_norm": 3.8196260929107666, "learning_rate": 1.4282407540542598e-06, "loss": 1.2955, "step": 38472 }, { "epoch": 3.01512539184953, "grad_norm": 2.8453876972198486, "learning_rate": 1.4280249731773155e-06, "loss": 1.4207, "step": 38473 }, { "epoch": 3.015203761755486, "grad_norm": 3.8031013011932373, "learning_rate": 1.4278092058863301e-06, "loss": 1.6188, "step": 38474 }, { "epoch": 3.015282131661442, "grad_norm": 2.9865829944610596, "learning_rate": 1.427593452182125e-06, "loss": 1.1998, "step": 38475 }, { "epoch": 3.015360501567398, "grad_norm": 2.790079355239868, "learning_rate": 1.427377712065518e-06, "loss": 1.2292, "step": 38476 }, { "epoch": 3.0154388714733544, "grad_norm": 3.154557943344116, "learning_rate": 1.4271619855373297e-06, "loss": 1.0781, "step": 38477 }, { "epoch": 3.0155172413793103, "grad_norm": 3.3049376010894775, "learning_rate": 1.4269462725983834e-06, "loss": 1.6163, "step": 38478 }, { "epoch": 3.0155956112852667, "grad_norm": 2.356496810913086, "learning_rate": 1.4267305732495002e-06, "loss": 1.1962, "step": 38479 }, { "epoch": 3.0156739811912225, "grad_norm": 2.696045398712158, "learning_rate": 1.4265148874914963e-06, "loss": 1.4208, "step": 38480 }, { "epoch": 3.015752351097179, "grad_norm": 7.7882890701293945, "learning_rate": 1.4262992153251948e-06, "loss": 0.8848, "step": 38481 }, { "epoch": 3.0158307210031348, "grad_norm": 2.608762264251709, "learning_rate": 1.4260835567514142e-06, "loss": 1.2671, "step": 38482 }, { "epoch": 3.015909090909091, "grad_norm": 6.236859321594238, "learning_rate": 1.4258679117709789e-06, "loss": 1.3192, "step": 38483 }, { "epoch": 3.015987460815047, "grad_norm": 2.552441358566284, "learning_rate": 1.4256522803847044e-06, "loss": 1.317, "step": 38484 }, { "epoch": 3.0160658307210033, "grad_norm": 14.28771686553955, "learning_rate": 1.425436662593413e-06, "loss": 1.7819, "step": 38485 }, { "epoch": 3.016144200626959, "grad_norm": 2.7597320079803467, "learning_rate": 1.4252210583979247e-06, "loss": 1.4455, "step": 38486 }, { "epoch": 3.0162225705329155, "grad_norm": 3.9724068641662598, "learning_rate": 1.4250054677990593e-06, "loss": 1.2326, "step": 38487 }, { "epoch": 3.0163009404388714, "grad_norm": 4.167212009429932, "learning_rate": 1.4247898907976366e-06, "loss": 1.5513, "step": 38488 }, { "epoch": 3.0163793103448278, "grad_norm": 2.6060585975646973, "learning_rate": 1.4245743273944772e-06, "loss": 1.5519, "step": 38489 }, { "epoch": 3.0164576802507836, "grad_norm": 3.6224536895751953, "learning_rate": 1.4243587775904016e-06, "loss": 1.3095, "step": 38490 }, { "epoch": 3.01653605015674, "grad_norm": 5.67549467086792, "learning_rate": 1.4241432413862262e-06, "loss": 1.2142, "step": 38491 }, { "epoch": 3.016614420062696, "grad_norm": 3.3440630435943604, "learning_rate": 1.4239277187827743e-06, "loss": 1.3441, "step": 38492 }, { "epoch": 3.016692789968652, "grad_norm": 7.271711826324463, "learning_rate": 1.4237122097808648e-06, "loss": 1.4085, "step": 38493 }, { "epoch": 3.016771159874608, "grad_norm": 3.2935502529144287, "learning_rate": 1.423496714381319e-06, "loss": 1.1671, "step": 38494 }, { "epoch": 3.0168495297805644, "grad_norm": 6.382594585418701, "learning_rate": 1.4232812325849526e-06, "loss": 1.771, "step": 38495 }, { "epoch": 3.0169278996865203, "grad_norm": 4.587118625640869, "learning_rate": 1.423065764392586e-06, "loss": 1.5383, "step": 38496 }, { "epoch": 3.0170062695924766, "grad_norm": 2.596325159072876, "learning_rate": 1.422850309805041e-06, "loss": 1.3318, "step": 38497 }, { "epoch": 3.0170846394984325, "grad_norm": 2.396010398864746, "learning_rate": 1.4226348688231384e-06, "loss": 1.3031, "step": 38498 }, { "epoch": 3.017163009404389, "grad_norm": 8.212175369262695, "learning_rate": 1.4224194414476934e-06, "loss": 0.9954, "step": 38499 }, { "epoch": 3.0172413793103448, "grad_norm": 2.175245761871338, "learning_rate": 1.4222040276795273e-06, "loss": 1.3716, "step": 38500 }, { "epoch": 3.017319749216301, "grad_norm": 7.189816474914551, "learning_rate": 1.4219886275194595e-06, "loss": 1.0688, "step": 38501 }, { "epoch": 3.017398119122257, "grad_norm": 2.300771474838257, "learning_rate": 1.4217732409683094e-06, "loss": 1.3669, "step": 38502 }, { "epoch": 3.0174764890282133, "grad_norm": 14.390541076660156, "learning_rate": 1.4215578680268953e-06, "loss": 1.4676, "step": 38503 }, { "epoch": 3.017554858934169, "grad_norm": 3.6425650119781494, "learning_rate": 1.421342508696038e-06, "loss": 1.3335, "step": 38504 }, { "epoch": 3.0176332288401255, "grad_norm": 1.78481924533844, "learning_rate": 1.421127162976555e-06, "loss": 0.6229, "step": 38505 }, { "epoch": 3.0177115987460814, "grad_norm": 2.707127332687378, "learning_rate": 1.420911830869266e-06, "loss": 1.4842, "step": 38506 }, { "epoch": 3.0177899686520377, "grad_norm": 4.629559516906738, "learning_rate": 1.4206965123749904e-06, "loss": 0.9369, "step": 38507 }, { "epoch": 3.0178683385579936, "grad_norm": 3.7127487659454346, "learning_rate": 1.4204812074945469e-06, "loss": 1.9157, "step": 38508 }, { "epoch": 3.01794670846395, "grad_norm": 4.1006550788879395, "learning_rate": 1.4202659162287542e-06, "loss": 1.1315, "step": 38509 }, { "epoch": 3.018025078369906, "grad_norm": 2.6191911697387695, "learning_rate": 1.4200506385784312e-06, "loss": 1.1734, "step": 38510 }, { "epoch": 3.018103448275862, "grad_norm": 2.7886438369750977, "learning_rate": 1.4198353745443965e-06, "loss": 1.4091, "step": 38511 }, { "epoch": 3.018181818181818, "grad_norm": 4.129982948303223, "learning_rate": 1.4196201241274699e-06, "loss": 1.2964, "step": 38512 }, { "epoch": 3.0182601880877744, "grad_norm": 2.2366185188293457, "learning_rate": 1.41940488732847e-06, "loss": 1.149, "step": 38513 }, { "epoch": 3.0183385579937303, "grad_norm": 3.772674560546875, "learning_rate": 1.4191896641482122e-06, "loss": 0.8989, "step": 38514 }, { "epoch": 3.0184169278996866, "grad_norm": 3.252265691757202, "learning_rate": 1.4189744545875194e-06, "loss": 1.7519, "step": 38515 }, { "epoch": 3.0184952978056425, "grad_norm": 3.3530936241149902, "learning_rate": 1.41875925864721e-06, "loss": 1.5328, "step": 38516 }, { "epoch": 3.018573667711599, "grad_norm": 3.258789539337158, "learning_rate": 1.418544076328099e-06, "loss": 1.1279, "step": 38517 }, { "epoch": 3.0186520376175547, "grad_norm": 4.125288486480713, "learning_rate": 1.418328907631007e-06, "loss": 1.218, "step": 38518 }, { "epoch": 3.018730407523511, "grad_norm": 2.3947432041168213, "learning_rate": 1.4181137525567507e-06, "loss": 1.4678, "step": 38519 }, { "epoch": 3.018808777429467, "grad_norm": 7.425023555755615, "learning_rate": 1.4178986111061527e-06, "loss": 1.4579, "step": 38520 }, { "epoch": 3.0188871473354233, "grad_norm": 2.3785548210144043, "learning_rate": 1.4176834832800273e-06, "loss": 1.2523, "step": 38521 }, { "epoch": 3.018965517241379, "grad_norm": 4.097402095794678, "learning_rate": 1.4174683690791935e-06, "loss": 1.0026, "step": 38522 }, { "epoch": 3.0190438871473355, "grad_norm": 1.8657712936401367, "learning_rate": 1.4172532685044704e-06, "loss": 1.3262, "step": 38523 }, { "epoch": 3.0191222570532914, "grad_norm": 5.503879070281982, "learning_rate": 1.4170381815566752e-06, "loss": 1.4679, "step": 38524 }, { "epoch": 3.0192006269592477, "grad_norm": 4.456681251525879, "learning_rate": 1.4168231082366268e-06, "loss": 1.375, "step": 38525 }, { "epoch": 3.0192789968652036, "grad_norm": 4.041552543640137, "learning_rate": 1.4166080485451429e-06, "loss": 1.3213, "step": 38526 }, { "epoch": 3.01935736677116, "grad_norm": 4.347289562225342, "learning_rate": 1.4163930024830414e-06, "loss": 1.2789, "step": 38527 }, { "epoch": 3.019435736677116, "grad_norm": 3.9092659950256348, "learning_rate": 1.4161779700511403e-06, "loss": 1.4974, "step": 38528 }, { "epoch": 3.019514106583072, "grad_norm": 6.613556861877441, "learning_rate": 1.415962951250257e-06, "loss": 1.7022, "step": 38529 }, { "epoch": 3.019592476489028, "grad_norm": 8.085476875305176, "learning_rate": 1.41574794608121e-06, "loss": 0.7302, "step": 38530 }, { "epoch": 3.0196708463949844, "grad_norm": 6.088321208953857, "learning_rate": 1.4155329545448183e-06, "loss": 1.1273, "step": 38531 }, { "epoch": 3.0197492163009403, "grad_norm": 3.264228582382202, "learning_rate": 1.4153179766418968e-06, "loss": 1.3763, "step": 38532 }, { "epoch": 3.0198275862068966, "grad_norm": 3.3355813026428223, "learning_rate": 1.4151030123732628e-06, "loss": 1.2741, "step": 38533 }, { "epoch": 3.0199059561128525, "grad_norm": 2.546480655670166, "learning_rate": 1.4148880617397376e-06, "loss": 1.574, "step": 38534 }, { "epoch": 3.019984326018809, "grad_norm": 2.9696717262268066, "learning_rate": 1.414673124742138e-06, "loss": 1.4354, "step": 38535 }, { "epoch": 3.0200626959247647, "grad_norm": 4.648674964904785, "learning_rate": 1.4144582013812786e-06, "loss": 1.4895, "step": 38536 }, { "epoch": 3.020141065830721, "grad_norm": 2.890734910964966, "learning_rate": 1.4142432916579785e-06, "loss": 1.4419, "step": 38537 }, { "epoch": 3.020219435736677, "grad_norm": 3.3257627487182617, "learning_rate": 1.414028395573054e-06, "loss": 1.2106, "step": 38538 }, { "epoch": 3.0202978056426333, "grad_norm": 2.657144069671631, "learning_rate": 1.4138135131273261e-06, "loss": 1.4299, "step": 38539 }, { "epoch": 3.020376175548589, "grad_norm": 6.0091552734375, "learning_rate": 1.4135986443216087e-06, "loss": 1.4293, "step": 38540 }, { "epoch": 3.0204545454545455, "grad_norm": 3.731656074523926, "learning_rate": 1.4133837891567198e-06, "loss": 1.5228, "step": 38541 }, { "epoch": 3.0205329153605014, "grad_norm": 7.262885570526123, "learning_rate": 1.4131689476334764e-06, "loss": 1.475, "step": 38542 }, { "epoch": 3.0206112852664577, "grad_norm": 2.9455785751342773, "learning_rate": 1.4129541197526963e-06, "loss": 1.212, "step": 38543 }, { "epoch": 3.0206896551724136, "grad_norm": 3.2410888671875, "learning_rate": 1.4127393055151966e-06, "loss": 1.4698, "step": 38544 }, { "epoch": 3.02076802507837, "grad_norm": 2.8792226314544678, "learning_rate": 1.4125245049217933e-06, "loss": 1.4271, "step": 38545 }, { "epoch": 3.020846394984326, "grad_norm": 3.905588150024414, "learning_rate": 1.4123097179733065e-06, "loss": 0.9259, "step": 38546 }, { "epoch": 3.020924764890282, "grad_norm": 3.5873498916625977, "learning_rate": 1.4120949446705474e-06, "loss": 1.117, "step": 38547 }, { "epoch": 3.021003134796238, "grad_norm": 4.258333683013916, "learning_rate": 1.4118801850143377e-06, "loss": 1.2203, "step": 38548 }, { "epoch": 3.0210815047021944, "grad_norm": 2.7630865573883057, "learning_rate": 1.4116654390054929e-06, "loss": 1.0746, "step": 38549 }, { "epoch": 3.0211598746081503, "grad_norm": 17.965288162231445, "learning_rate": 1.4114507066448308e-06, "loss": 1.5895, "step": 38550 }, { "epoch": 3.0212382445141066, "grad_norm": 2.30476450920105, "learning_rate": 1.4112359879331662e-06, "loss": 1.1523, "step": 38551 }, { "epoch": 3.0213166144200625, "grad_norm": 2.564857244491577, "learning_rate": 1.4110212828713142e-06, "loss": 1.4025, "step": 38552 }, { "epoch": 3.021394984326019, "grad_norm": 2.5811984539031982, "learning_rate": 1.410806591460096e-06, "loss": 1.4653, "step": 38553 }, { "epoch": 3.0214733542319747, "grad_norm": 9.599411964416504, "learning_rate": 1.4105919137003276e-06, "loss": 1.7327, "step": 38554 }, { "epoch": 3.021551724137931, "grad_norm": 3.3476576805114746, "learning_rate": 1.4103772495928219e-06, "loss": 1.1849, "step": 38555 }, { "epoch": 3.021630094043887, "grad_norm": 3.889671564102173, "learning_rate": 1.4101625991383978e-06, "loss": 1.4205, "step": 38556 }, { "epoch": 3.0217084639498433, "grad_norm": 7.66135835647583, "learning_rate": 1.4099479623378715e-06, "loss": 1.6877, "step": 38557 }, { "epoch": 3.021786833855799, "grad_norm": 3.7751424312591553, "learning_rate": 1.4097333391920587e-06, "loss": 1.3626, "step": 38558 }, { "epoch": 3.0218652037617555, "grad_norm": 3.9021923542022705, "learning_rate": 1.4095187297017764e-06, "loss": 1.4133, "step": 38559 }, { "epoch": 3.0219435736677114, "grad_norm": 4.168057441711426, "learning_rate": 1.4093041338678404e-06, "loss": 1.1568, "step": 38560 }, { "epoch": 3.0220219435736677, "grad_norm": 4.745033264160156, "learning_rate": 1.4090895516910674e-06, "loss": 1.5675, "step": 38561 }, { "epoch": 3.0221003134796236, "grad_norm": 4.054201602935791, "learning_rate": 1.4088749831722738e-06, "loss": 1.0512, "step": 38562 }, { "epoch": 3.02217868338558, "grad_norm": 4.875940799713135, "learning_rate": 1.4086604283122746e-06, "loss": 0.6361, "step": 38563 }, { "epoch": 3.0222570532915363, "grad_norm": 4.4551825523376465, "learning_rate": 1.4084458871118868e-06, "loss": 1.6852, "step": 38564 }, { "epoch": 3.022335423197492, "grad_norm": 3.8878283500671387, "learning_rate": 1.408231359571927e-06, "loss": 1.2311, "step": 38565 }, { "epoch": 3.0224137931034485, "grad_norm": 3.7104969024658203, "learning_rate": 1.408016845693208e-06, "loss": 1.5482, "step": 38566 }, { "epoch": 3.0224921630094044, "grad_norm": 3.062326192855835, "learning_rate": 1.4078023454765494e-06, "loss": 1.3563, "step": 38567 }, { "epoch": 3.0225705329153607, "grad_norm": 3.487452745437622, "learning_rate": 1.4075878589227654e-06, "loss": 1.4807, "step": 38568 }, { "epoch": 3.0226489028213166, "grad_norm": 3.409613847732544, "learning_rate": 1.4073733860326733e-06, "loss": 1.6846, "step": 38569 }, { "epoch": 3.022727272727273, "grad_norm": 5.540656566619873, "learning_rate": 1.4071589268070862e-06, "loss": 1.4553, "step": 38570 }, { "epoch": 3.022805642633229, "grad_norm": 10.616539001464844, "learning_rate": 1.4069444812468196e-06, "loss": 1.015, "step": 38571 }, { "epoch": 3.022884012539185, "grad_norm": 3.3427517414093018, "learning_rate": 1.4067300493526942e-06, "loss": 1.1043, "step": 38572 }, { "epoch": 3.022962382445141, "grad_norm": 6.064774513244629, "learning_rate": 1.4065156311255202e-06, "loss": 1.6515, "step": 38573 }, { "epoch": 3.0230407523510974, "grad_norm": 3.629755973815918, "learning_rate": 1.4063012265661153e-06, "loss": 1.2671, "step": 38574 }, { "epoch": 3.0231191222570533, "grad_norm": 4.051573276519775, "learning_rate": 1.4060868356752938e-06, "loss": 1.2294, "step": 38575 }, { "epoch": 3.0231974921630096, "grad_norm": 4.947788238525391, "learning_rate": 1.405872458453873e-06, "loss": 1.0897, "step": 38576 }, { "epoch": 3.0232758620689655, "grad_norm": 6.367440700531006, "learning_rate": 1.4056580949026667e-06, "loss": 1.7644, "step": 38577 }, { "epoch": 3.023354231974922, "grad_norm": 4.669384479522705, "learning_rate": 1.4054437450224912e-06, "loss": 1.3295, "step": 38578 }, { "epoch": 3.0234326018808777, "grad_norm": 3.1029350757598877, "learning_rate": 1.405229408814161e-06, "loss": 1.2679, "step": 38579 }, { "epoch": 3.023510971786834, "grad_norm": 2.4034805297851562, "learning_rate": 1.4050150862784923e-06, "loss": 1.2557, "step": 38580 }, { "epoch": 3.02358934169279, "grad_norm": 5.5973381996154785, "learning_rate": 1.4048007774162992e-06, "loss": 1.384, "step": 38581 }, { "epoch": 3.0236677115987463, "grad_norm": 3.6182234287261963, "learning_rate": 1.4045864822283977e-06, "loss": 1.2455, "step": 38582 }, { "epoch": 3.023746081504702, "grad_norm": 2.68676495552063, "learning_rate": 1.4043722007156025e-06, "loss": 1.6364, "step": 38583 }, { "epoch": 3.0238244514106585, "grad_norm": 4.389754295349121, "learning_rate": 1.4041579328787297e-06, "loss": 1.5799, "step": 38584 }, { "epoch": 3.0239028213166144, "grad_norm": 6.112680912017822, "learning_rate": 1.4039436787185906e-06, "loss": 1.5472, "step": 38585 }, { "epoch": 3.0239811912225707, "grad_norm": 8.781745910644531, "learning_rate": 1.403729438236004e-06, "loss": 1.6249, "step": 38586 }, { "epoch": 3.0240595611285266, "grad_norm": 3.6631970405578613, "learning_rate": 1.4035152114317852e-06, "loss": 1.5588, "step": 38587 }, { "epoch": 3.024137931034483, "grad_norm": 3.1402361392974854, "learning_rate": 1.4033009983067454e-06, "loss": 1.4171, "step": 38588 }, { "epoch": 3.024216300940439, "grad_norm": 3.840813398361206, "learning_rate": 1.4030867988617014e-06, "loss": 1.4627, "step": 38589 }, { "epoch": 3.024294670846395, "grad_norm": 1.9741313457489014, "learning_rate": 1.4028726130974662e-06, "loss": 1.1385, "step": 38590 }, { "epoch": 3.024373040752351, "grad_norm": 3.039600372314453, "learning_rate": 1.4026584410148591e-06, "loss": 1.408, "step": 38591 }, { "epoch": 3.0244514106583074, "grad_norm": 3.8371360301971436, "learning_rate": 1.40244428261469e-06, "loss": 1.6096, "step": 38592 }, { "epoch": 3.0245297805642632, "grad_norm": 3.6513328552246094, "learning_rate": 1.402230137897775e-06, "loss": 1.4453, "step": 38593 }, { "epoch": 3.0246081504702196, "grad_norm": 3.310147523880005, "learning_rate": 1.4020160068649286e-06, "loss": 1.3976, "step": 38594 }, { "epoch": 3.0246865203761755, "grad_norm": 4.333510398864746, "learning_rate": 1.401801889516965e-06, "loss": 0.8241, "step": 38595 }, { "epoch": 3.024764890282132, "grad_norm": 3.2836384773254395, "learning_rate": 1.4015877858546996e-06, "loss": 1.4578, "step": 38596 }, { "epoch": 3.0248432601880877, "grad_norm": 3.331996440887451, "learning_rate": 1.4013736958789454e-06, "loss": 1.4727, "step": 38597 }, { "epoch": 3.024921630094044, "grad_norm": 5.049764156341553, "learning_rate": 1.4011596195905175e-06, "loss": 1.4783, "step": 38598 }, { "epoch": 3.025, "grad_norm": 7.293085098266602, "learning_rate": 1.4009455569902297e-06, "loss": 1.2995, "step": 38599 }, { "epoch": 3.0250783699059562, "grad_norm": 3.7663686275482178, "learning_rate": 1.4007315080788964e-06, "loss": 1.3928, "step": 38600 }, { "epoch": 3.025156739811912, "grad_norm": 3.614386558532715, "learning_rate": 1.4005174728573323e-06, "loss": 1.7099, "step": 38601 }, { "epoch": 3.0252351097178685, "grad_norm": 3.1168267726898193, "learning_rate": 1.400303451326352e-06, "loss": 1.0963, "step": 38602 }, { "epoch": 3.0253134796238244, "grad_norm": 3.260359764099121, "learning_rate": 1.4000894434867668e-06, "loss": 1.2728, "step": 38603 }, { "epoch": 3.0253918495297807, "grad_norm": 13.914042472839355, "learning_rate": 1.3998754493393912e-06, "loss": 1.4475, "step": 38604 }, { "epoch": 3.0254702194357366, "grad_norm": 3.1881415843963623, "learning_rate": 1.3996614688850413e-06, "loss": 1.8247, "step": 38605 }, { "epoch": 3.025548589341693, "grad_norm": 3.759976625442505, "learning_rate": 1.3994475021245319e-06, "loss": 1.2132, "step": 38606 }, { "epoch": 3.025626959247649, "grad_norm": 4.267693042755127, "learning_rate": 1.3992335490586735e-06, "loss": 1.4508, "step": 38607 }, { "epoch": 3.025705329153605, "grad_norm": 2.141312837600708, "learning_rate": 1.3990196096882803e-06, "loss": 1.409, "step": 38608 }, { "epoch": 3.025783699059561, "grad_norm": 11.18025016784668, "learning_rate": 1.3988056840141661e-06, "loss": 1.1394, "step": 38609 }, { "epoch": 3.0258620689655173, "grad_norm": 3.1623494625091553, "learning_rate": 1.3985917720371484e-06, "loss": 1.1461, "step": 38610 }, { "epoch": 3.0259404388714732, "grad_norm": 2.5314416885375977, "learning_rate": 1.3983778737580362e-06, "loss": 1.3704, "step": 38611 }, { "epoch": 3.0260188087774296, "grad_norm": 6.729043960571289, "learning_rate": 1.3981639891776443e-06, "loss": 0.9743, "step": 38612 }, { "epoch": 3.0260971786833855, "grad_norm": 4.404256820678711, "learning_rate": 1.397950118296787e-06, "loss": 1.9489, "step": 38613 }, { "epoch": 3.026175548589342, "grad_norm": 3.1529197692871094, "learning_rate": 1.397736261116277e-06, "loss": 0.9759, "step": 38614 }, { "epoch": 3.0262539184952977, "grad_norm": 3.657336473464966, "learning_rate": 1.3975224176369284e-06, "loss": 0.9859, "step": 38615 }, { "epoch": 3.026332288401254, "grad_norm": 5.678646564483643, "learning_rate": 1.3973085878595539e-06, "loss": 1.2188, "step": 38616 }, { "epoch": 3.02641065830721, "grad_norm": 4.407100677490234, "learning_rate": 1.397094771784967e-06, "loss": 1.1297, "step": 38617 }, { "epoch": 3.0264890282131662, "grad_norm": 3.7606029510498047, "learning_rate": 1.3968809694139812e-06, "loss": 1.2309, "step": 38618 }, { "epoch": 3.026567398119122, "grad_norm": 6.8219170570373535, "learning_rate": 1.3966671807474097e-06, "loss": 1.7851, "step": 38619 }, { "epoch": 3.0266457680250785, "grad_norm": 5.4744672775268555, "learning_rate": 1.3964534057860652e-06, "loss": 1.2016, "step": 38620 }, { "epoch": 3.0267241379310343, "grad_norm": 3.660996675491333, "learning_rate": 1.396239644530763e-06, "loss": 1.3507, "step": 38621 }, { "epoch": 3.0268025078369907, "grad_norm": 2.906776189804077, "learning_rate": 1.3960258969823104e-06, "loss": 1.5586, "step": 38622 }, { "epoch": 3.0268808777429466, "grad_norm": 5.444530963897705, "learning_rate": 1.395812163141526e-06, "loss": 1.4764, "step": 38623 }, { "epoch": 3.026959247648903, "grad_norm": 4.491590976715088, "learning_rate": 1.3955984430092212e-06, "loss": 0.9949, "step": 38624 }, { "epoch": 3.027037617554859, "grad_norm": 3.7536399364471436, "learning_rate": 1.39538473658621e-06, "loss": 1.3983, "step": 38625 }, { "epoch": 3.027115987460815, "grad_norm": 3.3348424434661865, "learning_rate": 1.3951710438733019e-06, "loss": 1.735, "step": 38626 }, { "epoch": 3.027194357366771, "grad_norm": 3.513805389404297, "learning_rate": 1.3949573648713105e-06, "loss": 1.2823, "step": 38627 }, { "epoch": 3.0272727272727273, "grad_norm": 4.893709182739258, "learning_rate": 1.3947436995810525e-06, "loss": 1.3032, "step": 38628 }, { "epoch": 3.0273510971786832, "grad_norm": 3.34631085395813, "learning_rate": 1.3945300480033357e-06, "loss": 1.2444, "step": 38629 }, { "epoch": 3.0274294670846396, "grad_norm": 3.8730764389038086, "learning_rate": 1.3943164101389756e-06, "loss": 1.0132, "step": 38630 }, { "epoch": 3.0275078369905954, "grad_norm": 2.837756872177124, "learning_rate": 1.3941027859887834e-06, "loss": 1.2909, "step": 38631 }, { "epoch": 3.027586206896552, "grad_norm": 13.898518562316895, "learning_rate": 1.3938891755535722e-06, "loss": 0.9476, "step": 38632 }, { "epoch": 3.0276645768025077, "grad_norm": 2.675485610961914, "learning_rate": 1.3936755788341545e-06, "loss": 1.3309, "step": 38633 }, { "epoch": 3.027742946708464, "grad_norm": 5.174831390380859, "learning_rate": 1.3934619958313423e-06, "loss": 1.2181, "step": 38634 }, { "epoch": 3.02782131661442, "grad_norm": 11.62459659576416, "learning_rate": 1.3932484265459479e-06, "loss": 1.6393, "step": 38635 }, { "epoch": 3.027899686520376, "grad_norm": 2.518817186355591, "learning_rate": 1.3930348709787844e-06, "loss": 1.5489, "step": 38636 }, { "epoch": 3.027978056426332, "grad_norm": 4.2154340744018555, "learning_rate": 1.3928213291306635e-06, "loss": 1.157, "step": 38637 }, { "epoch": 3.0280564263322884, "grad_norm": 2.6927599906921387, "learning_rate": 1.3926078010023974e-06, "loss": 1.0966, "step": 38638 }, { "epoch": 3.0281347962382443, "grad_norm": 4.072132587432861, "learning_rate": 1.3923942865947986e-06, "loss": 1.1396, "step": 38639 }, { "epoch": 3.0282131661442007, "grad_norm": 10.230803489685059, "learning_rate": 1.3921807859086805e-06, "loss": 0.9324, "step": 38640 }, { "epoch": 3.0282915360501566, "grad_norm": 4.435755252838135, "learning_rate": 1.3919672989448508e-06, "loss": 1.3969, "step": 38641 }, { "epoch": 3.028369905956113, "grad_norm": 2.817694902420044, "learning_rate": 1.3917538257041252e-06, "loss": 1.4577, "step": 38642 }, { "epoch": 3.0284482758620688, "grad_norm": 1.9432222843170166, "learning_rate": 1.3915403661873167e-06, "loss": 1.1169, "step": 38643 }, { "epoch": 3.028526645768025, "grad_norm": 4.965132236480713, "learning_rate": 1.3913269203952334e-06, "loss": 1.6339, "step": 38644 }, { "epoch": 3.028605015673981, "grad_norm": 4.674820899963379, "learning_rate": 1.3911134883286887e-06, "loss": 1.3258, "step": 38645 }, { "epoch": 3.0286833855799373, "grad_norm": 3.173708438873291, "learning_rate": 1.3909000699884938e-06, "loss": 1.3807, "step": 38646 }, { "epoch": 3.028761755485893, "grad_norm": 8.746845245361328, "learning_rate": 1.3906866653754642e-06, "loss": 0.8091, "step": 38647 }, { "epoch": 3.0288401253918495, "grad_norm": 4.0482025146484375, "learning_rate": 1.3904732744904066e-06, "loss": 1.2755, "step": 38648 }, { "epoch": 3.0289184952978054, "grad_norm": 3.6103386878967285, "learning_rate": 1.3902598973341352e-06, "loss": 1.1225, "step": 38649 }, { "epoch": 3.0289968652037618, "grad_norm": 2.538309097290039, "learning_rate": 1.3900465339074609e-06, "loss": 1.6516, "step": 38650 }, { "epoch": 3.0290752351097177, "grad_norm": 5.4959516525268555, "learning_rate": 1.389833184211195e-06, "loss": 1.1197, "step": 38651 }, { "epoch": 3.029153605015674, "grad_norm": 5.149352550506592, "learning_rate": 1.3896198482461503e-06, "loss": 1.757, "step": 38652 }, { "epoch": 3.02923197492163, "grad_norm": 3.3614141941070557, "learning_rate": 1.3894065260131368e-06, "loss": 0.9652, "step": 38653 }, { "epoch": 3.029310344827586, "grad_norm": 6.122039794921875, "learning_rate": 1.3891932175129674e-06, "loss": 1.3479, "step": 38654 }, { "epoch": 3.0293887147335425, "grad_norm": 3.8625497817993164, "learning_rate": 1.38897992274645e-06, "loss": 0.8192, "step": 38655 }, { "epoch": 3.0294670846394984, "grad_norm": 4.663957595825195, "learning_rate": 1.3887666417143996e-06, "loss": 1.2345, "step": 38656 }, { "epoch": 3.0295454545454548, "grad_norm": 3.140367031097412, "learning_rate": 1.388553374417626e-06, "loss": 1.5468, "step": 38657 }, { "epoch": 3.0296238244514107, "grad_norm": 3.285416603088379, "learning_rate": 1.388340120856942e-06, "loss": 1.6645, "step": 38658 }, { "epoch": 3.029702194357367, "grad_norm": 6.315323829650879, "learning_rate": 1.3881268810331555e-06, "loss": 1.1713, "step": 38659 }, { "epoch": 3.029780564263323, "grad_norm": 3.7022266387939453, "learning_rate": 1.3879136549470773e-06, "loss": 1.6339, "step": 38660 }, { "epoch": 3.029858934169279, "grad_norm": 4.169234752655029, "learning_rate": 1.3877004425995223e-06, "loss": 1.0479, "step": 38661 }, { "epoch": 3.029937304075235, "grad_norm": 2.7500829696655273, "learning_rate": 1.3874872439913011e-06, "loss": 1.3589, "step": 38662 }, { "epoch": 3.0300156739811914, "grad_norm": 5.695961952209473, "learning_rate": 1.387274059123221e-06, "loss": 1.3227, "step": 38663 }, { "epoch": 3.0300940438871473, "grad_norm": 9.15270709991455, "learning_rate": 1.3870608879960946e-06, "loss": 0.9495, "step": 38664 }, { "epoch": 3.0301724137931036, "grad_norm": 2.576397657394409, "learning_rate": 1.386847730610732e-06, "loss": 1.563, "step": 38665 }, { "epoch": 3.0302507836990595, "grad_norm": 3.486292600631714, "learning_rate": 1.386634586967947e-06, "loss": 1.3979, "step": 38666 }, { "epoch": 3.030329153605016, "grad_norm": 4.1845526695251465, "learning_rate": 1.3864214570685474e-06, "loss": 1.3673, "step": 38667 }, { "epoch": 3.0304075235109718, "grad_norm": 3.828441619873047, "learning_rate": 1.3862083409133441e-06, "loss": 1.3602, "step": 38668 }, { "epoch": 3.030485893416928, "grad_norm": 2.7333333492279053, "learning_rate": 1.3859952385031478e-06, "loss": 1.6131, "step": 38669 }, { "epoch": 3.030564263322884, "grad_norm": 3.3001394271850586, "learning_rate": 1.3857821498387696e-06, "loss": 1.3834, "step": 38670 }, { "epoch": 3.0306426332288403, "grad_norm": 2.9219818115234375, "learning_rate": 1.38556907492102e-06, "loss": 0.943, "step": 38671 }, { "epoch": 3.030721003134796, "grad_norm": 9.156232833862305, "learning_rate": 1.3853560137507089e-06, "loss": 1.7944, "step": 38672 }, { "epoch": 3.0307993730407525, "grad_norm": 2.506997585296631, "learning_rate": 1.3851429663286487e-06, "loss": 1.2836, "step": 38673 }, { "epoch": 3.0308777429467084, "grad_norm": 3.0628130435943604, "learning_rate": 1.3849299326556448e-06, "loss": 1.2385, "step": 38674 }, { "epoch": 3.0309561128526648, "grad_norm": 4.6556830406188965, "learning_rate": 1.3847169127325122e-06, "loss": 1.1255, "step": 38675 }, { "epoch": 3.0310344827586206, "grad_norm": 2.9611055850982666, "learning_rate": 1.384503906560059e-06, "loss": 0.8648, "step": 38676 }, { "epoch": 3.031112852664577, "grad_norm": 13.28278923034668, "learning_rate": 1.3842909141390982e-06, "loss": 0.7915, "step": 38677 }, { "epoch": 3.031191222570533, "grad_norm": 4.419806003570557, "learning_rate": 1.3840779354704354e-06, "loss": 1.2649, "step": 38678 }, { "epoch": 3.031269592476489, "grad_norm": 3.9751522541046143, "learning_rate": 1.3838649705548817e-06, "loss": 1.6306, "step": 38679 }, { "epoch": 3.031347962382445, "grad_norm": 3.533752918243408, "learning_rate": 1.3836520193932495e-06, "loss": 1.2172, "step": 38680 }, { "epoch": 3.0314263322884014, "grad_norm": 5.8588104248046875, "learning_rate": 1.3834390819863497e-06, "loss": 1.1758, "step": 38681 }, { "epoch": 3.0315047021943573, "grad_norm": 4.687588214874268, "learning_rate": 1.3832261583349882e-06, "loss": 1.2856, "step": 38682 }, { "epoch": 3.0315830721003136, "grad_norm": 4.151060581207275, "learning_rate": 1.3830132484399766e-06, "loss": 1.2974, "step": 38683 }, { "epoch": 3.0316614420062695, "grad_norm": 8.140498161315918, "learning_rate": 1.382800352302125e-06, "loss": 0.9245, "step": 38684 }, { "epoch": 3.031739811912226, "grad_norm": 3.18485426902771, "learning_rate": 1.3825874699222426e-06, "loss": 1.2675, "step": 38685 }, { "epoch": 3.0318181818181817, "grad_norm": 3.0490338802337646, "learning_rate": 1.3823746013011397e-06, "loss": 1.3424, "step": 38686 }, { "epoch": 3.031896551724138, "grad_norm": 10.39171314239502, "learning_rate": 1.3821617464396258e-06, "loss": 1.2886, "step": 38687 }, { "epoch": 3.031974921630094, "grad_norm": 4.328906059265137, "learning_rate": 1.3819489053385098e-06, "loss": 1.0078, "step": 38688 }, { "epoch": 3.0320532915360503, "grad_norm": 3.037269115447998, "learning_rate": 1.3817360779986029e-06, "loss": 1.4198, "step": 38689 }, { "epoch": 3.032131661442006, "grad_norm": 4.43409538269043, "learning_rate": 1.3815232644207127e-06, "loss": 1.4566, "step": 38690 }, { "epoch": 3.0322100313479625, "grad_norm": 3.1528913974761963, "learning_rate": 1.3813104646056496e-06, "loss": 1.2612, "step": 38691 }, { "epoch": 3.0322884012539184, "grad_norm": 2.050675392150879, "learning_rate": 1.381097678554224e-06, "loss": 1.2448, "step": 38692 }, { "epoch": 3.0323667711598747, "grad_norm": 8.439611434936523, "learning_rate": 1.3808849062672413e-06, "loss": 1.3126, "step": 38693 }, { "epoch": 3.0324451410658306, "grad_norm": 3.4414305686950684, "learning_rate": 1.3806721477455155e-06, "loss": 1.2423, "step": 38694 }, { "epoch": 3.032523510971787, "grad_norm": 3.4805502891540527, "learning_rate": 1.3804594029898534e-06, "loss": 1.1313, "step": 38695 }, { "epoch": 3.032601880877743, "grad_norm": 7.557694435119629, "learning_rate": 1.380246672001066e-06, "loss": 1.2298, "step": 38696 }, { "epoch": 3.032680250783699, "grad_norm": 4.6902570724487305, "learning_rate": 1.3800339547799596e-06, "loss": 1.2954, "step": 38697 }, { "epoch": 3.032758620689655, "grad_norm": 2.6055068969726562, "learning_rate": 1.3798212513273435e-06, "loss": 1.1207, "step": 38698 }, { "epoch": 3.0328369905956114, "grad_norm": 2.5132036209106445, "learning_rate": 1.3796085616440302e-06, "loss": 1.234, "step": 38699 }, { "epoch": 3.0329153605015673, "grad_norm": 3.4424355030059814, "learning_rate": 1.3793958857308254e-06, "loss": 1.6135, "step": 38700 }, { "epoch": 3.0329937304075236, "grad_norm": 3.595703601837158, "learning_rate": 1.3791832235885393e-06, "loss": 1.3325, "step": 38701 }, { "epoch": 3.0330721003134795, "grad_norm": 2.3075530529022217, "learning_rate": 1.3789705752179798e-06, "loss": 1.3409, "step": 38702 }, { "epoch": 3.033150470219436, "grad_norm": 4.575556755065918, "learning_rate": 1.378757940619957e-06, "loss": 0.5974, "step": 38703 }, { "epoch": 3.0332288401253917, "grad_norm": 3.698113203048706, "learning_rate": 1.3785453197952786e-06, "loss": 1.1123, "step": 38704 }, { "epoch": 3.033307210031348, "grad_norm": 4.292427062988281, "learning_rate": 1.3783327127447538e-06, "loss": 1.2424, "step": 38705 }, { "epoch": 3.033385579937304, "grad_norm": 2.8281755447387695, "learning_rate": 1.3781201194691913e-06, "loss": 1.3679, "step": 38706 }, { "epoch": 3.0334639498432603, "grad_norm": 3.7316250801086426, "learning_rate": 1.3779075399693992e-06, "loss": 1.2807, "step": 38707 }, { "epoch": 3.033542319749216, "grad_norm": 3.4334890842437744, "learning_rate": 1.3776949742461865e-06, "loss": 1.163, "step": 38708 }, { "epoch": 3.0336206896551725, "grad_norm": 7.235237121582031, "learning_rate": 1.3774824223003618e-06, "loss": 1.3657, "step": 38709 }, { "epoch": 3.0336990595611284, "grad_norm": 3.4402945041656494, "learning_rate": 1.3772698841327347e-06, "loss": 1.613, "step": 38710 }, { "epoch": 3.0337774294670847, "grad_norm": 3.3556649684906006, "learning_rate": 1.3770573597441088e-06, "loss": 1.573, "step": 38711 }, { "epoch": 3.0338557993730406, "grad_norm": 4.9940505027771, "learning_rate": 1.3768448491352981e-06, "loss": 0.8559, "step": 38712 }, { "epoch": 3.033934169278997, "grad_norm": 8.19803524017334, "learning_rate": 1.3766323523071078e-06, "loss": 0.6176, "step": 38713 }, { "epoch": 3.034012539184953, "grad_norm": 3.061664342880249, "learning_rate": 1.3764198692603492e-06, "loss": 1.3475, "step": 38714 }, { "epoch": 3.034090909090909, "grad_norm": 6.278822898864746, "learning_rate": 1.3762073999958265e-06, "loss": 1.1998, "step": 38715 }, { "epoch": 3.034169278996865, "grad_norm": 2.092273235321045, "learning_rate": 1.3759949445143477e-06, "loss": 1.2655, "step": 38716 }, { "epoch": 3.0342476489028214, "grad_norm": 4.497890472412109, "learning_rate": 1.375782502816725e-06, "loss": 1.0832, "step": 38717 }, { "epoch": 3.0343260188087773, "grad_norm": 2.854407548904419, "learning_rate": 1.3755700749037648e-06, "loss": 1.2351, "step": 38718 }, { "epoch": 3.0344043887147336, "grad_norm": 6.27389669418335, "learning_rate": 1.3753576607762737e-06, "loss": 0.9739, "step": 38719 }, { "epoch": 3.0344827586206895, "grad_norm": 3.5928940773010254, "learning_rate": 1.3751452604350597e-06, "loss": 1.337, "step": 38720 }, { "epoch": 3.034561128526646, "grad_norm": 3.198031187057495, "learning_rate": 1.3749328738809315e-06, "loss": 1.4167, "step": 38721 }, { "epoch": 3.0346394984326017, "grad_norm": 6.972599506378174, "learning_rate": 1.3747205011146964e-06, "loss": 1.2438, "step": 38722 }, { "epoch": 3.034717868338558, "grad_norm": 2.9913625717163086, "learning_rate": 1.374508142137163e-06, "loss": 1.3313, "step": 38723 }, { "epoch": 3.034796238244514, "grad_norm": 2.9242608547210693, "learning_rate": 1.3742957969491382e-06, "loss": 1.4631, "step": 38724 }, { "epoch": 3.0348746081504703, "grad_norm": 3.2751195430755615, "learning_rate": 1.3740834655514296e-06, "loss": 1.1047, "step": 38725 }, { "epoch": 3.034952978056426, "grad_norm": 2.186607837677002, "learning_rate": 1.3738711479448457e-06, "loss": 1.1985, "step": 38726 }, { "epoch": 3.0350313479623825, "grad_norm": 2.669229745864868, "learning_rate": 1.3736588441301929e-06, "loss": 0.9368, "step": 38727 }, { "epoch": 3.0351097178683384, "grad_norm": 5.771290302276611, "learning_rate": 1.3734465541082798e-06, "loss": 1.3708, "step": 38728 }, { "epoch": 3.0351880877742947, "grad_norm": 2.8466713428497314, "learning_rate": 1.3732342778799145e-06, "loss": 1.1886, "step": 38729 }, { "epoch": 3.0352664576802506, "grad_norm": 6.286237716674805, "learning_rate": 1.3730220154459001e-06, "loss": 1.5418, "step": 38730 }, { "epoch": 3.035344827586207, "grad_norm": 2.190781831741333, "learning_rate": 1.3728097668070484e-06, "loss": 1.5667, "step": 38731 }, { "epoch": 3.035423197492163, "grad_norm": 4.4409589767456055, "learning_rate": 1.3725975319641654e-06, "loss": 1.3892, "step": 38732 }, { "epoch": 3.035501567398119, "grad_norm": 2.5859243869781494, "learning_rate": 1.37238531091806e-06, "loss": 1.62, "step": 38733 }, { "epoch": 3.035579937304075, "grad_norm": 3.317861318588257, "learning_rate": 1.3721731036695363e-06, "loss": 1.2751, "step": 38734 }, { "epoch": 3.0356583072100314, "grad_norm": 3.64920711517334, "learning_rate": 1.3719609102194008e-06, "loss": 1.2805, "step": 38735 }, { "epoch": 3.0357366771159873, "grad_norm": 5.440231800079346, "learning_rate": 1.371748730568464e-06, "loss": 1.5737, "step": 38736 }, { "epoch": 3.0358150470219436, "grad_norm": 20.707366943359375, "learning_rate": 1.3715365647175332e-06, "loss": 2.451, "step": 38737 }, { "epoch": 3.0358934169278995, "grad_norm": 2.44624662399292, "learning_rate": 1.371324412667412e-06, "loss": 1.1664, "step": 38738 }, { "epoch": 3.035971786833856, "grad_norm": 2.8829305171966553, "learning_rate": 1.3711122744189088e-06, "loss": 1.2834, "step": 38739 }, { "epoch": 3.0360501567398117, "grad_norm": 2.3653206825256348, "learning_rate": 1.3709001499728308e-06, "loss": 1.3834, "step": 38740 }, { "epoch": 3.036128526645768, "grad_norm": 3.602973461151123, "learning_rate": 1.3706880393299843e-06, "loss": 1.2512, "step": 38741 }, { "epoch": 3.036206896551724, "grad_norm": 4.006237506866455, "learning_rate": 1.3704759424911769e-06, "loss": 1.3238, "step": 38742 }, { "epoch": 3.0362852664576803, "grad_norm": 3.3859095573425293, "learning_rate": 1.3702638594572143e-06, "loss": 1.4541, "step": 38743 }, { "epoch": 3.036363636363636, "grad_norm": 4.7600016593933105, "learning_rate": 1.3700517902289035e-06, "loss": 1.0305, "step": 38744 }, { "epoch": 3.0364420062695925, "grad_norm": 3.998075008392334, "learning_rate": 1.3698397348070519e-06, "loss": 1.0571, "step": 38745 }, { "epoch": 3.0365203761755484, "grad_norm": 3.4903883934020996, "learning_rate": 1.3696276931924645e-06, "loss": 0.806, "step": 38746 }, { "epoch": 3.0365987460815047, "grad_norm": 3.2905502319335938, "learning_rate": 1.3694156653859492e-06, "loss": 1.0795, "step": 38747 }, { "epoch": 3.0366771159874606, "grad_norm": 3.222684383392334, "learning_rate": 1.369203651388314e-06, "loss": 1.685, "step": 38748 }, { "epoch": 3.036755485893417, "grad_norm": 4.5152716636657715, "learning_rate": 1.3689916512003593e-06, "loss": 0.7633, "step": 38749 }, { "epoch": 3.0368338557993733, "grad_norm": 5.511035442352295, "learning_rate": 1.3687796648228968e-06, "loss": 1.1477, "step": 38750 }, { "epoch": 3.036912225705329, "grad_norm": 5.019566059112549, "learning_rate": 1.3685676922567321e-06, "loss": 1.1945, "step": 38751 }, { "epoch": 3.0369905956112855, "grad_norm": 6.117905616760254, "learning_rate": 1.3683557335026714e-06, "loss": 1.5124, "step": 38752 }, { "epoch": 3.0370689655172414, "grad_norm": 3.644432783126831, "learning_rate": 1.3681437885615196e-06, "loss": 1.6834, "step": 38753 }, { "epoch": 3.0371473354231977, "grad_norm": 2.5460872650146484, "learning_rate": 1.367931857434081e-06, "loss": 0.9931, "step": 38754 }, { "epoch": 3.0372257053291536, "grad_norm": 2.610053777694702, "learning_rate": 1.3677199401211672e-06, "loss": 1.2756, "step": 38755 }, { "epoch": 3.03730407523511, "grad_norm": 2.40556263923645, "learning_rate": 1.36750803662358e-06, "loss": 1.2788, "step": 38756 }, { "epoch": 3.037382445141066, "grad_norm": 3.8756942749023438, "learning_rate": 1.3672961469421264e-06, "loss": 0.8559, "step": 38757 }, { "epoch": 3.037460815047022, "grad_norm": 2.3320932388305664, "learning_rate": 1.3670842710776122e-06, "loss": 1.354, "step": 38758 }, { "epoch": 3.037539184952978, "grad_norm": 2.184845209121704, "learning_rate": 1.3668724090308432e-06, "loss": 1.3227, "step": 38759 }, { "epoch": 3.0376175548589344, "grad_norm": 2.9853410720825195, "learning_rate": 1.3666605608026256e-06, "loss": 1.1996, "step": 38760 }, { "epoch": 3.0376959247648903, "grad_norm": 2.621767520904541, "learning_rate": 1.3664487263937654e-06, "loss": 1.3256, "step": 38761 }, { "epoch": 3.0377742946708466, "grad_norm": 2.379059076309204, "learning_rate": 1.3662369058050678e-06, "loss": 1.4019, "step": 38762 }, { "epoch": 3.0378526645768025, "grad_norm": 4.538116931915283, "learning_rate": 1.3660250990373385e-06, "loss": 1.5314, "step": 38763 }, { "epoch": 3.037931034482759, "grad_norm": 3.4803781509399414, "learning_rate": 1.365813306091383e-06, "loss": 1.1261, "step": 38764 }, { "epoch": 3.0380094043887147, "grad_norm": 2.949777364730835, "learning_rate": 1.3656015269680074e-06, "loss": 0.9574, "step": 38765 }, { "epoch": 3.038087774294671, "grad_norm": 2.9774622917175293, "learning_rate": 1.3653897616680184e-06, "loss": 1.414, "step": 38766 }, { "epoch": 3.038166144200627, "grad_norm": 2.3895998001098633, "learning_rate": 1.365178010192218e-06, "loss": 1.4009, "step": 38767 }, { "epoch": 3.0382445141065832, "grad_norm": 2.93367075920105, "learning_rate": 1.364966272541412e-06, "loss": 1.408, "step": 38768 }, { "epoch": 3.038322884012539, "grad_norm": 2.5328280925750732, "learning_rate": 1.3647545487164087e-06, "loss": 1.1289, "step": 38769 }, { "epoch": 3.0384012539184955, "grad_norm": 4.105257034301758, "learning_rate": 1.3645428387180137e-06, "loss": 1.2466, "step": 38770 }, { "epoch": 3.0384796238244514, "grad_norm": 3.7384586334228516, "learning_rate": 1.364331142547029e-06, "loss": 1.4135, "step": 38771 }, { "epoch": 3.0385579937304077, "grad_norm": 5.047712326049805, "learning_rate": 1.3641194602042612e-06, "loss": 1.3426, "step": 38772 }, { "epoch": 3.0386363636363636, "grad_norm": 2.8110973834991455, "learning_rate": 1.3639077916905135e-06, "loss": 1.5867, "step": 38773 }, { "epoch": 3.03871473354232, "grad_norm": 2.6926982402801514, "learning_rate": 1.3636961370065965e-06, "loss": 1.3683, "step": 38774 }, { "epoch": 3.038793103448276, "grad_norm": 5.035759449005127, "learning_rate": 1.36348449615331e-06, "loss": 1.2861, "step": 38775 }, { "epoch": 3.038871473354232, "grad_norm": 3.4608678817749023, "learning_rate": 1.3632728691314607e-06, "loss": 1.5434, "step": 38776 }, { "epoch": 3.038949843260188, "grad_norm": 5.729115962982178, "learning_rate": 1.3630612559418533e-06, "loss": 1.5076, "step": 38777 }, { "epoch": 3.0390282131661444, "grad_norm": 3.218184232711792, "learning_rate": 1.3628496565852928e-06, "loss": 1.1227, "step": 38778 }, { "epoch": 3.0391065830721002, "grad_norm": 2.9186506271362305, "learning_rate": 1.3626380710625841e-06, "loss": 1.1286, "step": 38779 }, { "epoch": 3.0391849529780566, "grad_norm": 3.6410329341888428, "learning_rate": 1.3624264993745323e-06, "loss": 1.4394, "step": 38780 }, { "epoch": 3.0392633228840125, "grad_norm": 2.884131669998169, "learning_rate": 1.3622149415219427e-06, "loss": 1.4705, "step": 38781 }, { "epoch": 3.039341692789969, "grad_norm": 3.6169166564941406, "learning_rate": 1.3620033975056162e-06, "loss": 1.2229, "step": 38782 }, { "epoch": 3.0394200626959247, "grad_norm": 3.53533935546875, "learning_rate": 1.3617918673263619e-06, "loss": 1.6129, "step": 38783 }, { "epoch": 3.039498432601881, "grad_norm": 3.1178812980651855, "learning_rate": 1.3615803509849823e-06, "loss": 1.8939, "step": 38784 }, { "epoch": 3.039576802507837, "grad_norm": 3.4905457496643066, "learning_rate": 1.361368848482284e-06, "loss": 1.3812, "step": 38785 }, { "epoch": 3.0396551724137932, "grad_norm": 2.91571044921875, "learning_rate": 1.3611573598190682e-06, "loss": 0.9689, "step": 38786 }, { "epoch": 3.039733542319749, "grad_norm": 3.760775566101074, "learning_rate": 1.3609458849961389e-06, "loss": 1.2503, "step": 38787 }, { "epoch": 3.0398119122257055, "grad_norm": 2.627131938934326, "learning_rate": 1.360734424014304e-06, "loss": 1.4628, "step": 38788 }, { "epoch": 3.0398902821316613, "grad_norm": 3.5452849864959717, "learning_rate": 1.3605229768743678e-06, "loss": 1.3651, "step": 38789 }, { "epoch": 3.0399686520376177, "grad_norm": 2.807541847229004, "learning_rate": 1.3603115435771307e-06, "loss": 1.6487, "step": 38790 }, { "epoch": 3.0400470219435736, "grad_norm": 2.6920101642608643, "learning_rate": 1.3601001241233997e-06, "loss": 1.3766, "step": 38791 }, { "epoch": 3.04012539184953, "grad_norm": 3.586545467376709, "learning_rate": 1.3598887185139765e-06, "loss": 1.5029, "step": 38792 }, { "epoch": 3.040203761755486, "grad_norm": 6.137172698974609, "learning_rate": 1.35967732674967e-06, "loss": 1.4533, "step": 38793 }, { "epoch": 3.040282131661442, "grad_norm": 3.846226930618286, "learning_rate": 1.3594659488312794e-06, "loss": 1.358, "step": 38794 }, { "epoch": 3.040360501567398, "grad_norm": 6.32581901550293, "learning_rate": 1.3592545847596105e-06, "loss": 1.596, "step": 38795 }, { "epoch": 3.0404388714733543, "grad_norm": 3.407026529312134, "learning_rate": 1.3590432345354675e-06, "loss": 1.018, "step": 38796 }, { "epoch": 3.0405172413793102, "grad_norm": 10.866556167602539, "learning_rate": 1.3588318981596537e-06, "loss": 1.321, "step": 38797 }, { "epoch": 3.0405956112852666, "grad_norm": 3.9511337280273438, "learning_rate": 1.3586205756329735e-06, "loss": 1.5125, "step": 38798 }, { "epoch": 3.0406739811912225, "grad_norm": 3.7996914386749268, "learning_rate": 1.3584092669562298e-06, "loss": 1.6243, "step": 38799 }, { "epoch": 3.040752351097179, "grad_norm": 3.1230404376983643, "learning_rate": 1.3581979721302286e-06, "loss": 1.4869, "step": 38800 }, { "epoch": 3.0408307210031347, "grad_norm": 4.509349346160889, "learning_rate": 1.3579866911557688e-06, "loss": 1.6263, "step": 38801 }, { "epoch": 3.040909090909091, "grad_norm": 3.445953607559204, "learning_rate": 1.3577754240336583e-06, "loss": 1.9774, "step": 38802 }, { "epoch": 3.040987460815047, "grad_norm": 2.5056347846984863, "learning_rate": 1.3575641707646992e-06, "loss": 1.3693, "step": 38803 }, { "epoch": 3.0410658307210032, "grad_norm": 3.1975314617156982, "learning_rate": 1.3573529313496969e-06, "loss": 1.434, "step": 38804 }, { "epoch": 3.041144200626959, "grad_norm": 3.224015712738037, "learning_rate": 1.3571417057894515e-06, "loss": 1.2934, "step": 38805 }, { "epoch": 3.0412225705329154, "grad_norm": 2.6902408599853516, "learning_rate": 1.3569304940847661e-06, "loss": 1.4446, "step": 38806 }, { "epoch": 3.0413009404388713, "grad_norm": 3.43009352684021, "learning_rate": 1.3567192962364477e-06, "loss": 1.6614, "step": 38807 }, { "epoch": 3.0413793103448277, "grad_norm": 4.761958599090576, "learning_rate": 1.3565081122452995e-06, "loss": 0.7271, "step": 38808 }, { "epoch": 3.0414576802507836, "grad_norm": 4.7086076736450195, "learning_rate": 1.3562969421121213e-06, "loss": 1.5803, "step": 38809 }, { "epoch": 3.04153605015674, "grad_norm": 3.5413331985473633, "learning_rate": 1.356085785837718e-06, "loss": 1.3491, "step": 38810 }, { "epoch": 3.041614420062696, "grad_norm": 2.799802541732788, "learning_rate": 1.3558746434228925e-06, "loss": 1.3366, "step": 38811 }, { "epoch": 3.041692789968652, "grad_norm": 3.529663562774658, "learning_rate": 1.355663514868449e-06, "loss": 0.9187, "step": 38812 }, { "epoch": 3.041771159874608, "grad_norm": 3.6824426651000977, "learning_rate": 1.3554524001751883e-06, "loss": 1.5923, "step": 38813 }, { "epoch": 3.0418495297805643, "grad_norm": 3.890662670135498, "learning_rate": 1.3552412993439157e-06, "loss": 1.2103, "step": 38814 }, { "epoch": 3.04192789968652, "grad_norm": 2.3485493659973145, "learning_rate": 1.3550302123754328e-06, "loss": 1.3706, "step": 38815 }, { "epoch": 3.0420062695924766, "grad_norm": 2.498493194580078, "learning_rate": 1.354819139270543e-06, "loss": 1.2797, "step": 38816 }, { "epoch": 3.0420846394984324, "grad_norm": 3.2762651443481445, "learning_rate": 1.3546080800300487e-06, "loss": 1.8941, "step": 38817 }, { "epoch": 3.0421630094043888, "grad_norm": 5.983969211578369, "learning_rate": 1.354397034654753e-06, "loss": 1.2708, "step": 38818 }, { "epoch": 3.0422413793103447, "grad_norm": 2.671170711517334, "learning_rate": 1.3541860031454585e-06, "loss": 1.4278, "step": 38819 }, { "epoch": 3.042319749216301, "grad_norm": 15.314608573913574, "learning_rate": 1.3539749855029676e-06, "loss": 0.8558, "step": 38820 }, { "epoch": 3.042398119122257, "grad_norm": 5.405744552612305, "learning_rate": 1.3537639817280834e-06, "loss": 1.0251, "step": 38821 }, { "epoch": 3.042476489028213, "grad_norm": 3.001406669616699, "learning_rate": 1.3535529918216096e-06, "loss": 1.3698, "step": 38822 }, { "epoch": 3.042554858934169, "grad_norm": 5.479452610015869, "learning_rate": 1.3533420157843458e-06, "loss": 1.2403, "step": 38823 }, { "epoch": 3.0426332288401254, "grad_norm": 3.6196365356445312, "learning_rate": 1.3531310536170943e-06, "loss": 1.7012, "step": 38824 }, { "epoch": 3.0427115987460813, "grad_norm": 3.3002326488494873, "learning_rate": 1.3529201053206608e-06, "loss": 1.5251, "step": 38825 }, { "epoch": 3.0427899686520377, "grad_norm": 2.992696523666382, "learning_rate": 1.3527091708958472e-06, "loss": 1.339, "step": 38826 }, { "epoch": 3.0428683385579935, "grad_norm": 3.660109281539917, "learning_rate": 1.352498250343453e-06, "loss": 1.59, "step": 38827 }, { "epoch": 3.04294670846395, "grad_norm": 2.490713596343994, "learning_rate": 1.3522873436642825e-06, "loss": 1.1627, "step": 38828 }, { "epoch": 3.0430250783699058, "grad_norm": 3.2555272579193115, "learning_rate": 1.3520764508591356e-06, "loss": 1.3692, "step": 38829 }, { "epoch": 3.043103448275862, "grad_norm": 4.086956977844238, "learning_rate": 1.3518655719288193e-06, "loss": 1.9687, "step": 38830 }, { "epoch": 3.043181818181818, "grad_norm": 3.1463723182678223, "learning_rate": 1.351654706874131e-06, "loss": 1.3764, "step": 38831 }, { "epoch": 3.0432601880877743, "grad_norm": 3.526782989501953, "learning_rate": 1.3514438556958742e-06, "loss": 1.553, "step": 38832 }, { "epoch": 3.04333855799373, "grad_norm": 3.807952642440796, "learning_rate": 1.3512330183948508e-06, "loss": 2.093, "step": 38833 }, { "epoch": 3.0434169278996865, "grad_norm": 3.5168044567108154, "learning_rate": 1.3510221949718626e-06, "loss": 1.3841, "step": 38834 }, { "epoch": 3.0434952978056424, "grad_norm": 4.119033336639404, "learning_rate": 1.3508113854277126e-06, "loss": 1.199, "step": 38835 }, { "epoch": 3.0435736677115988, "grad_norm": 3.130349636077881, "learning_rate": 1.3506005897632012e-06, "loss": 1.1554, "step": 38836 }, { "epoch": 3.0436520376175547, "grad_norm": 7.872837066650391, "learning_rate": 1.3503898079791322e-06, "loss": 1.3631, "step": 38837 }, { "epoch": 3.043730407523511, "grad_norm": 3.0587828159332275, "learning_rate": 1.3501790400763033e-06, "loss": 1.4061, "step": 38838 }, { "epoch": 3.043808777429467, "grad_norm": 7.118302345275879, "learning_rate": 1.3499682860555196e-06, "loss": 0.8421, "step": 38839 }, { "epoch": 3.043887147335423, "grad_norm": 3.176311492919922, "learning_rate": 1.3497575459175821e-06, "loss": 1.1546, "step": 38840 }, { "epoch": 3.043965517241379, "grad_norm": 2.9177708625793457, "learning_rate": 1.3495468196632934e-06, "loss": 1.354, "step": 38841 }, { "epoch": 3.0440438871473354, "grad_norm": 2.8776328563690186, "learning_rate": 1.349336107293452e-06, "loss": 1.5301, "step": 38842 }, { "epoch": 3.0441222570532913, "grad_norm": 2.5140740871429443, "learning_rate": 1.3491254088088595e-06, "loss": 1.1104, "step": 38843 }, { "epoch": 3.0442006269592476, "grad_norm": 1.9300929307937622, "learning_rate": 1.3489147242103201e-06, "loss": 1.3998, "step": 38844 }, { "epoch": 3.044278996865204, "grad_norm": 5.2260260581970215, "learning_rate": 1.3487040534986357e-06, "loss": 1.5162, "step": 38845 }, { "epoch": 3.04435736677116, "grad_norm": 4.066991806030273, "learning_rate": 1.3484933966746038e-06, "loss": 1.1162, "step": 38846 }, { "epoch": 3.044435736677116, "grad_norm": 2.70509672164917, "learning_rate": 1.3482827537390274e-06, "loss": 1.1421, "step": 38847 }, { "epoch": 3.044514106583072, "grad_norm": 3.5886363983154297, "learning_rate": 1.3480721246927069e-06, "loss": 0.9354, "step": 38848 }, { "epoch": 3.0445924764890284, "grad_norm": 2.8686563968658447, "learning_rate": 1.3478615095364466e-06, "loss": 1.3515, "step": 38849 }, { "epoch": 3.0446708463949843, "grad_norm": 2.7566184997558594, "learning_rate": 1.3476509082710442e-06, "loss": 1.0964, "step": 38850 }, { "epoch": 3.0447492163009406, "grad_norm": 2.889288902282715, "learning_rate": 1.3474403208973013e-06, "loss": 1.0312, "step": 38851 }, { "epoch": 3.0448275862068965, "grad_norm": 5.417389392852783, "learning_rate": 1.3472297474160196e-06, "loss": 1.3925, "step": 38852 }, { "epoch": 3.044905956112853, "grad_norm": 2.716409206390381, "learning_rate": 1.3470191878279997e-06, "loss": 1.3952, "step": 38853 }, { "epoch": 3.0449843260188088, "grad_norm": 4.055235385894775, "learning_rate": 1.3468086421340431e-06, "loss": 1.6938, "step": 38854 }, { "epoch": 3.045062695924765, "grad_norm": 4.381535053253174, "learning_rate": 1.3465981103349495e-06, "loss": 1.8814, "step": 38855 }, { "epoch": 3.045141065830721, "grad_norm": 2.9840588569641113, "learning_rate": 1.3463875924315218e-06, "loss": 1.3653, "step": 38856 }, { "epoch": 3.0452194357366773, "grad_norm": 7.376872539520264, "learning_rate": 1.3461770884245556e-06, "loss": 1.0372, "step": 38857 }, { "epoch": 3.045297805642633, "grad_norm": 5.038435459136963, "learning_rate": 1.3459665983148573e-06, "loss": 1.6866, "step": 38858 }, { "epoch": 3.0453761755485895, "grad_norm": 6.0730156898498535, "learning_rate": 1.3457561221032245e-06, "loss": 1.3715, "step": 38859 }, { "epoch": 3.0454545454545454, "grad_norm": 5.2396674156188965, "learning_rate": 1.3455456597904605e-06, "loss": 1.6253, "step": 38860 }, { "epoch": 3.0455329153605017, "grad_norm": 2.564324140548706, "learning_rate": 1.345335211377362e-06, "loss": 1.1548, "step": 38861 }, { "epoch": 3.0456112852664576, "grad_norm": 2.858433961868286, "learning_rate": 1.3451247768647301e-06, "loss": 1.1103, "step": 38862 }, { "epoch": 3.045689655172414, "grad_norm": 3.6362240314483643, "learning_rate": 1.344914356253369e-06, "loss": 1.0567, "step": 38863 }, { "epoch": 3.04576802507837, "grad_norm": 3.596134901046753, "learning_rate": 1.3447039495440745e-06, "loss": 1.6789, "step": 38864 }, { "epoch": 3.045846394984326, "grad_norm": 8.918315887451172, "learning_rate": 1.3444935567376494e-06, "loss": 1.1534, "step": 38865 }, { "epoch": 3.045924764890282, "grad_norm": 4.748157501220703, "learning_rate": 1.3442831778348925e-06, "loss": 1.026, "step": 38866 }, { "epoch": 3.0460031347962384, "grad_norm": 3.587193012237549, "learning_rate": 1.3440728128366054e-06, "loss": 1.2421, "step": 38867 }, { "epoch": 3.0460815047021943, "grad_norm": 11.495738983154297, "learning_rate": 1.343862461743587e-06, "loss": 0.3818, "step": 38868 }, { "epoch": 3.0461598746081506, "grad_norm": 2.428713321685791, "learning_rate": 1.343652124556638e-06, "loss": 1.3054, "step": 38869 }, { "epoch": 3.0462382445141065, "grad_norm": 2.4803807735443115, "learning_rate": 1.3434418012765582e-06, "loss": 1.4168, "step": 38870 }, { "epoch": 3.046316614420063, "grad_norm": 3.816333532333374, "learning_rate": 1.3432314919041478e-06, "loss": 1.4959, "step": 38871 }, { "epoch": 3.0463949843260187, "grad_norm": 4.506686687469482, "learning_rate": 1.3430211964402067e-06, "loss": 1.2631, "step": 38872 }, { "epoch": 3.046473354231975, "grad_norm": 3.7174973487854004, "learning_rate": 1.3428109148855345e-06, "loss": 1.3576, "step": 38873 }, { "epoch": 3.046551724137931, "grad_norm": 7.836298942565918, "learning_rate": 1.3426006472409309e-06, "loss": 1.0591, "step": 38874 }, { "epoch": 3.0466300940438873, "grad_norm": 11.697149276733398, "learning_rate": 1.3423903935071974e-06, "loss": 1.2822, "step": 38875 }, { "epoch": 3.046708463949843, "grad_norm": 2.452944278717041, "learning_rate": 1.3421801536851297e-06, "loss": 1.3108, "step": 38876 }, { "epoch": 3.0467868338557995, "grad_norm": 3.2778260707855225, "learning_rate": 1.341969927775531e-06, "loss": 1.1869, "step": 38877 }, { "epoch": 3.0468652037617554, "grad_norm": 2.7643702030181885, "learning_rate": 1.3417597157792018e-06, "loss": 1.0334, "step": 38878 }, { "epoch": 3.0469435736677117, "grad_norm": 8.712617874145508, "learning_rate": 1.3415495176969374e-06, "loss": 1.4585, "step": 38879 }, { "epoch": 3.0470219435736676, "grad_norm": 2.940417766571045, "learning_rate": 1.3413393335295398e-06, "loss": 1.4197, "step": 38880 }, { "epoch": 3.047100313479624, "grad_norm": 2.8113908767700195, "learning_rate": 1.3411291632778067e-06, "loss": 0.9955, "step": 38881 }, { "epoch": 3.04717868338558, "grad_norm": 3.2615034580230713, "learning_rate": 1.3409190069425421e-06, "loss": 1.4248, "step": 38882 }, { "epoch": 3.047257053291536, "grad_norm": 10.563611030578613, "learning_rate": 1.340708864524541e-06, "loss": 1.3395, "step": 38883 }, { "epoch": 3.047335423197492, "grad_norm": 3.2336714267730713, "learning_rate": 1.340498736024603e-06, "loss": 1.376, "step": 38884 }, { "epoch": 3.0474137931034484, "grad_norm": 2.7492270469665527, "learning_rate": 1.3402886214435285e-06, "loss": 1.4926, "step": 38885 }, { "epoch": 3.0474921630094043, "grad_norm": 2.4448983669281006, "learning_rate": 1.3400785207821164e-06, "loss": 1.1189, "step": 38886 }, { "epoch": 3.0475705329153606, "grad_norm": 3.6960272789001465, "learning_rate": 1.3398684340411661e-06, "loss": 0.9105, "step": 38887 }, { "epoch": 3.0476489028213165, "grad_norm": 5.348999977111816, "learning_rate": 1.339658361221476e-06, "loss": 1.6909, "step": 38888 }, { "epoch": 3.047727272727273, "grad_norm": 2.925910711288452, "learning_rate": 1.3394483023238452e-06, "loss": 1.5887, "step": 38889 }, { "epoch": 3.0478056426332287, "grad_norm": 5.98129940032959, "learning_rate": 1.339238257349073e-06, "loss": 1.2652, "step": 38890 }, { "epoch": 3.047884012539185, "grad_norm": 2.84798264503479, "learning_rate": 1.3390282262979588e-06, "loss": 1.5348, "step": 38891 }, { "epoch": 3.047962382445141, "grad_norm": 5.731499195098877, "learning_rate": 1.3388182091713004e-06, "loss": 1.7006, "step": 38892 }, { "epoch": 3.0480407523510973, "grad_norm": 4.158463001251221, "learning_rate": 1.338608205969898e-06, "loss": 1.5836, "step": 38893 }, { "epoch": 3.048119122257053, "grad_norm": 4.661098480224609, "learning_rate": 1.3383982166945487e-06, "loss": 1.6211, "step": 38894 }, { "epoch": 3.0481974921630095, "grad_norm": 3.345440626144409, "learning_rate": 1.3381882413460496e-06, "loss": 1.3082, "step": 38895 }, { "epoch": 3.0482758620689654, "grad_norm": 7.883763313293457, "learning_rate": 1.337978279925204e-06, "loss": 1.5731, "step": 38896 }, { "epoch": 3.0483542319749217, "grad_norm": 3.3106906414031982, "learning_rate": 1.3377683324328089e-06, "loss": 1.2789, "step": 38897 }, { "epoch": 3.0484326018808776, "grad_norm": 10.876072883605957, "learning_rate": 1.3375583988696606e-06, "loss": 0.6452, "step": 38898 }, { "epoch": 3.048510971786834, "grad_norm": 8.035514831542969, "learning_rate": 1.3373484792365592e-06, "loss": 1.3548, "step": 38899 }, { "epoch": 3.04858934169279, "grad_norm": 3.089341163635254, "learning_rate": 1.3371385735343011e-06, "loss": 1.244, "step": 38900 }, { "epoch": 3.048667711598746, "grad_norm": 4.011122226715088, "learning_rate": 1.33692868176369e-06, "loss": 1.1923, "step": 38901 }, { "epoch": 3.048746081504702, "grad_norm": 4.8431925773620605, "learning_rate": 1.3367188039255192e-06, "loss": 1.3878, "step": 38902 }, { "epoch": 3.0488244514106584, "grad_norm": 2.90019154548645, "learning_rate": 1.3365089400205876e-06, "loss": 1.4108, "step": 38903 }, { "epoch": 3.0489028213166143, "grad_norm": 2.9888875484466553, "learning_rate": 1.3362990900496953e-06, "loss": 1.5975, "step": 38904 }, { "epoch": 3.0489811912225706, "grad_norm": 2.766838312149048, "learning_rate": 1.336089254013639e-06, "loss": 1.6378, "step": 38905 }, { "epoch": 3.0490595611285265, "grad_norm": 2.2181243896484375, "learning_rate": 1.3358794319132174e-06, "loss": 1.1645, "step": 38906 }, { "epoch": 3.049137931034483, "grad_norm": 6.720587730407715, "learning_rate": 1.3356696237492284e-06, "loss": 1.4275, "step": 38907 }, { "epoch": 3.0492163009404387, "grad_norm": 2.5330543518066406, "learning_rate": 1.3354598295224703e-06, "loss": 1.1742, "step": 38908 }, { "epoch": 3.049294670846395, "grad_norm": 2.75895619392395, "learning_rate": 1.3352500492337406e-06, "loss": 1.0228, "step": 38909 }, { "epoch": 3.049373040752351, "grad_norm": 3.6620163917541504, "learning_rate": 1.335040282883837e-06, "loss": 1.372, "step": 38910 }, { "epoch": 3.0494514106583073, "grad_norm": 2.6652376651763916, "learning_rate": 1.3348305304735582e-06, "loss": 1.1198, "step": 38911 }, { "epoch": 3.049529780564263, "grad_norm": 6.061656475067139, "learning_rate": 1.334620792003703e-06, "loss": 1.1775, "step": 38912 }, { "epoch": 3.0496081504702195, "grad_norm": 3.9972620010375977, "learning_rate": 1.334411067475066e-06, "loss": 2.0558, "step": 38913 }, { "epoch": 3.0496865203761754, "grad_norm": 5.447263717651367, "learning_rate": 1.334201356888445e-06, "loss": 0.9911, "step": 38914 }, { "epoch": 3.0497648902821317, "grad_norm": 2.1472394466400146, "learning_rate": 1.3339916602446406e-06, "loss": 1.2735, "step": 38915 }, { "epoch": 3.0498432601880876, "grad_norm": 12.07962417602539, "learning_rate": 1.3337819775444505e-06, "loss": 0.7222, "step": 38916 }, { "epoch": 3.049921630094044, "grad_norm": 4.2379608154296875, "learning_rate": 1.3335723087886694e-06, "loss": 1.5775, "step": 38917 }, { "epoch": 3.05, "grad_norm": 2.5873327255249023, "learning_rate": 1.3333626539780958e-06, "loss": 0.9945, "step": 38918 }, { "epoch": 3.050078369905956, "grad_norm": 3.568065643310547, "learning_rate": 1.3331530131135273e-06, "loss": 1.4861, "step": 38919 }, { "epoch": 3.050156739811912, "grad_norm": 4.028417110443115, "learning_rate": 1.3329433861957614e-06, "loss": 1.4322, "step": 38920 }, { "epoch": 3.0502351097178684, "grad_norm": 2.219019651412964, "learning_rate": 1.3327337732255958e-06, "loss": 1.0579, "step": 38921 }, { "epoch": 3.0503134796238243, "grad_norm": 9.033930778503418, "learning_rate": 1.3325241742038264e-06, "loss": 0.7411, "step": 38922 }, { "epoch": 3.0503918495297806, "grad_norm": 3.211714029312134, "learning_rate": 1.3323145891312522e-06, "loss": 1.2185, "step": 38923 }, { "epoch": 3.0504702194357365, "grad_norm": 4.4553704261779785, "learning_rate": 1.3321050180086687e-06, "loss": 1.178, "step": 38924 }, { "epoch": 3.050548589341693, "grad_norm": 4.979179382324219, "learning_rate": 1.3318954608368744e-06, "loss": 1.2328, "step": 38925 }, { "epoch": 3.0506269592476487, "grad_norm": 3.262218952178955, "learning_rate": 1.3316859176166651e-06, "loss": 1.6311, "step": 38926 }, { "epoch": 3.050705329153605, "grad_norm": 3.546588659286499, "learning_rate": 1.331476388348839e-06, "loss": 1.2012, "step": 38927 }, { "epoch": 3.050783699059561, "grad_norm": 11.0381441116333, "learning_rate": 1.331266873034192e-06, "loss": 1.5267, "step": 38928 }, { "epoch": 3.0508620689655173, "grad_norm": 4.216057300567627, "learning_rate": 1.3310573716735215e-06, "loss": 1.6015, "step": 38929 }, { "epoch": 3.050940438871473, "grad_norm": 6.150317192077637, "learning_rate": 1.3308478842676248e-06, "loss": 1.5166, "step": 38930 }, { "epoch": 3.0510188087774295, "grad_norm": 5.579030990600586, "learning_rate": 1.3306384108172987e-06, "loss": 1.1869, "step": 38931 }, { "epoch": 3.0510971786833854, "grad_norm": 4.343162536621094, "learning_rate": 1.3304289513233371e-06, "loss": 0.9376, "step": 38932 }, { "epoch": 3.0511755485893417, "grad_norm": 4.031234264373779, "learning_rate": 1.3302195057865403e-06, "loss": 1.3439, "step": 38933 }, { "epoch": 3.0512539184952976, "grad_norm": 3.586362838745117, "learning_rate": 1.3300100742077049e-06, "loss": 1.4003, "step": 38934 }, { "epoch": 3.051332288401254, "grad_norm": 3.926041603088379, "learning_rate": 1.3298006565876249e-06, "loss": 1.1339, "step": 38935 }, { "epoch": 3.0514106583072103, "grad_norm": 3.9351232051849365, "learning_rate": 1.3295912529270983e-06, "loss": 1.6543, "step": 38936 }, { "epoch": 3.051489028213166, "grad_norm": 3.9967823028564453, "learning_rate": 1.3293818632269196e-06, "loss": 1.3817, "step": 38937 }, { "epoch": 3.051567398119122, "grad_norm": 3.4561867713928223, "learning_rate": 1.3291724874878897e-06, "loss": 1.0519, "step": 38938 }, { "epoch": 3.0516457680250784, "grad_norm": 4.2867255210876465, "learning_rate": 1.3289631257108005e-06, "loss": 1.7718, "step": 38939 }, { "epoch": 3.0517241379310347, "grad_norm": 4.231784820556641, "learning_rate": 1.3287537778964504e-06, "loss": 1.3555, "step": 38940 }, { "epoch": 3.0518025078369906, "grad_norm": 3.7419376373291016, "learning_rate": 1.328544444045635e-06, "loss": 0.5858, "step": 38941 }, { "epoch": 3.051880877742947, "grad_norm": 2.534916877746582, "learning_rate": 1.3283351241591513e-06, "loss": 1.1043, "step": 38942 }, { "epoch": 3.051959247648903, "grad_norm": 3.2160775661468506, "learning_rate": 1.3281258182377949e-06, "loss": 1.2341, "step": 38943 }, { "epoch": 3.052037617554859, "grad_norm": 4.299324035644531, "learning_rate": 1.3279165262823612e-06, "loss": 1.2662, "step": 38944 }, { "epoch": 3.052115987460815, "grad_norm": 3.7035889625549316, "learning_rate": 1.3277072482936475e-06, "loss": 1.237, "step": 38945 }, { "epoch": 3.0521943573667714, "grad_norm": 3.230889081954956, "learning_rate": 1.3274979842724489e-06, "loss": 1.0154, "step": 38946 }, { "epoch": 3.0522727272727272, "grad_norm": 3.148098945617676, "learning_rate": 1.3272887342195617e-06, "loss": 1.5615, "step": 38947 }, { "epoch": 3.0523510971786836, "grad_norm": 6.003515720367432, "learning_rate": 1.327079498135782e-06, "loss": 1.521, "step": 38948 }, { "epoch": 3.0524294670846395, "grad_norm": 3.5811541080474854, "learning_rate": 1.3268702760219065e-06, "loss": 1.3233, "step": 38949 }, { "epoch": 3.052507836990596, "grad_norm": 4.928853511810303, "learning_rate": 1.3266610678787283e-06, "loss": 1.8189, "step": 38950 }, { "epoch": 3.0525862068965517, "grad_norm": 3.607207775115967, "learning_rate": 1.3264518737070437e-06, "loss": 1.5648, "step": 38951 }, { "epoch": 3.052664576802508, "grad_norm": 2.374274253845215, "learning_rate": 1.3262426935076506e-06, "loss": 1.213, "step": 38952 }, { "epoch": 3.052742946708464, "grad_norm": 2.9565534591674805, "learning_rate": 1.3260335272813446e-06, "loss": 0.9781, "step": 38953 }, { "epoch": 3.0528213166144202, "grad_norm": 3.4209389686584473, "learning_rate": 1.3258243750289184e-06, "loss": 1.1759, "step": 38954 }, { "epoch": 3.052899686520376, "grad_norm": 3.0408637523651123, "learning_rate": 1.3256152367511694e-06, "loss": 1.2254, "step": 38955 }, { "epoch": 3.0529780564263325, "grad_norm": 3.402099847793579, "learning_rate": 1.325406112448891e-06, "loss": 1.2436, "step": 38956 }, { "epoch": 3.0530564263322884, "grad_norm": 3.4105384349823, "learning_rate": 1.3251970021228833e-06, "loss": 1.0119, "step": 38957 }, { "epoch": 3.0531347962382447, "grad_norm": 3.1986286640167236, "learning_rate": 1.324987905773938e-06, "loss": 1.1997, "step": 38958 }, { "epoch": 3.0532131661442006, "grad_norm": 2.849175214767456, "learning_rate": 1.3247788234028503e-06, "loss": 1.3395, "step": 38959 }, { "epoch": 3.053291536050157, "grad_norm": 2.7874279022216797, "learning_rate": 1.3245697550104165e-06, "loss": 1.4064, "step": 38960 }, { "epoch": 3.053369905956113, "grad_norm": 8.359657287597656, "learning_rate": 1.3243607005974313e-06, "loss": 1.2715, "step": 38961 }, { "epoch": 3.053448275862069, "grad_norm": 4.252283096313477, "learning_rate": 1.3241516601646903e-06, "loss": 1.2204, "step": 38962 }, { "epoch": 3.053526645768025, "grad_norm": 2.7110941410064697, "learning_rate": 1.3239426337129884e-06, "loss": 1.0882, "step": 38963 }, { "epoch": 3.0536050156739813, "grad_norm": 16.46233558654785, "learning_rate": 1.3237336212431218e-06, "loss": 1.7445, "step": 38964 }, { "epoch": 3.0536833855799372, "grad_norm": 5.017269134521484, "learning_rate": 1.3235246227558813e-06, "loss": 1.653, "step": 38965 }, { "epoch": 3.0537617554858936, "grad_norm": 4.79447078704834, "learning_rate": 1.3233156382520667e-06, "loss": 1.5017, "step": 38966 }, { "epoch": 3.0538401253918495, "grad_norm": 2.332026243209839, "learning_rate": 1.3231066677324705e-06, "loss": 1.3025, "step": 38967 }, { "epoch": 3.053918495297806, "grad_norm": 2.963251829147339, "learning_rate": 1.3228977111978897e-06, "loss": 1.3622, "step": 38968 }, { "epoch": 3.0539968652037617, "grad_norm": 2.7211508750915527, "learning_rate": 1.3226887686491158e-06, "loss": 0.9351, "step": 38969 }, { "epoch": 3.054075235109718, "grad_norm": 4.23582124710083, "learning_rate": 1.3224798400869432e-06, "loss": 1.1094, "step": 38970 }, { "epoch": 3.054153605015674, "grad_norm": 2.5267932415008545, "learning_rate": 1.3222709255121702e-06, "loss": 1.447, "step": 38971 }, { "epoch": 3.0542319749216302, "grad_norm": 4.395055294036865, "learning_rate": 1.3220620249255911e-06, "loss": 1.7867, "step": 38972 }, { "epoch": 3.054310344827586, "grad_norm": 5.881052017211914, "learning_rate": 1.3218531383279975e-06, "loss": 1.7434, "step": 38973 }, { "epoch": 3.0543887147335425, "grad_norm": 6.744062423706055, "learning_rate": 1.3216442657201855e-06, "loss": 1.0421, "step": 38974 }, { "epoch": 3.0544670846394983, "grad_norm": 2.422459602355957, "learning_rate": 1.321435407102949e-06, "loss": 1.2479, "step": 38975 }, { "epoch": 3.0545454545454547, "grad_norm": 2.3477625846862793, "learning_rate": 1.3212265624770832e-06, "loss": 1.4441, "step": 38976 }, { "epoch": 3.0546238244514106, "grad_norm": 2.4889719486236572, "learning_rate": 1.3210177318433814e-06, "loss": 1.502, "step": 38977 }, { "epoch": 3.054702194357367, "grad_norm": 7.304123878479004, "learning_rate": 1.3208089152026388e-06, "loss": 0.9314, "step": 38978 }, { "epoch": 3.054780564263323, "grad_norm": 4.62675142288208, "learning_rate": 1.3206001125556495e-06, "loss": 1.3862, "step": 38979 }, { "epoch": 3.054858934169279, "grad_norm": 3.8469913005828857, "learning_rate": 1.3203913239032074e-06, "loss": 1.2749, "step": 38980 }, { "epoch": 3.054937304075235, "grad_norm": 8.849257469177246, "learning_rate": 1.3201825492461067e-06, "loss": 0.936, "step": 38981 }, { "epoch": 3.0550156739811913, "grad_norm": 2.882535219192505, "learning_rate": 1.3199737885851416e-06, "loss": 1.2746, "step": 38982 }, { "epoch": 3.0550940438871472, "grad_norm": 3.5529439449310303, "learning_rate": 1.3197650419211079e-06, "loss": 1.8427, "step": 38983 }, { "epoch": 3.0551724137931036, "grad_norm": 3.455880880355835, "learning_rate": 1.3195563092547941e-06, "loss": 1.0416, "step": 38984 }, { "epoch": 3.0552507836990594, "grad_norm": 5.482598304748535, "learning_rate": 1.3193475905869996e-06, "loss": 1.3291, "step": 38985 }, { "epoch": 3.055329153605016, "grad_norm": 5.312206745147705, "learning_rate": 1.319138885918516e-06, "loss": 1.4383, "step": 38986 }, { "epoch": 3.0554075235109717, "grad_norm": 3.547525405883789, "learning_rate": 1.3189301952501387e-06, "loss": 1.416, "step": 38987 }, { "epoch": 3.055485893416928, "grad_norm": 4.490411758422852, "learning_rate": 1.3187215185826597e-06, "loss": 1.6149, "step": 38988 }, { "epoch": 3.055564263322884, "grad_norm": 6.315470218658447, "learning_rate": 1.318512855916871e-06, "loss": 1.335, "step": 38989 }, { "epoch": 3.05564263322884, "grad_norm": 3.0008163452148438, "learning_rate": 1.3183042072535719e-06, "loss": 1.61, "step": 38990 }, { "epoch": 3.055721003134796, "grad_norm": 2.8848729133605957, "learning_rate": 1.3180955725935502e-06, "loss": 1.6215, "step": 38991 }, { "epoch": 3.0557993730407524, "grad_norm": 5.868881702423096, "learning_rate": 1.3178869519376025e-06, "loss": 1.3715, "step": 38992 }, { "epoch": 3.0558777429467083, "grad_norm": 3.4374983310699463, "learning_rate": 1.3176783452865216e-06, "loss": 1.2866, "step": 38993 }, { "epoch": 3.0559561128526647, "grad_norm": 3.6689064502716064, "learning_rate": 1.3174697526411007e-06, "loss": 1.3705, "step": 38994 }, { "epoch": 3.0560344827586206, "grad_norm": 2.498055934906006, "learning_rate": 1.3172611740021335e-06, "loss": 1.0712, "step": 38995 }, { "epoch": 3.056112852664577, "grad_norm": 2.585742712020874, "learning_rate": 1.3170526093704134e-06, "loss": 1.6715, "step": 38996 }, { "epoch": 3.0561912225705328, "grad_norm": 4.542436599731445, "learning_rate": 1.316844058746733e-06, "loss": 1.189, "step": 38997 }, { "epoch": 3.056269592476489, "grad_norm": 5.628189563751221, "learning_rate": 1.3166355221318865e-06, "loss": 1.2829, "step": 38998 }, { "epoch": 3.056347962382445, "grad_norm": 4.12959623336792, "learning_rate": 1.316426999526666e-06, "loss": 1.47, "step": 38999 }, { "epoch": 3.0564263322884013, "grad_norm": 3.9962451457977295, "learning_rate": 1.316218490931866e-06, "loss": 1.5657, "step": 39000 }, { "epoch": 3.056504702194357, "grad_norm": 5.970259189605713, "learning_rate": 1.3160099963482782e-06, "loss": 1.309, "step": 39001 }, { "epoch": 3.0565830721003135, "grad_norm": 4.714405059814453, "learning_rate": 1.315801515776698e-06, "loss": 1.2936, "step": 39002 }, { "epoch": 3.0566614420062694, "grad_norm": 3.307018995285034, "learning_rate": 1.315593049217913e-06, "loss": 1.5686, "step": 39003 }, { "epoch": 3.0567398119122258, "grad_norm": 12.842610359191895, "learning_rate": 1.3153845966727214e-06, "loss": 1.1937, "step": 39004 }, { "epoch": 3.0568181818181817, "grad_norm": 11.26172924041748, "learning_rate": 1.315176158141916e-06, "loss": 1.533, "step": 39005 }, { "epoch": 3.056896551724138, "grad_norm": 3.0000181198120117, "learning_rate": 1.3149677336262856e-06, "loss": 1.2532, "step": 39006 }, { "epoch": 3.056974921630094, "grad_norm": 3.6542234420776367, "learning_rate": 1.314759323126626e-06, "loss": 1.4569, "step": 39007 }, { "epoch": 3.05705329153605, "grad_norm": 2.3642728328704834, "learning_rate": 1.3145509266437272e-06, "loss": 1.1693, "step": 39008 }, { "epoch": 3.057131661442006, "grad_norm": 2.429802656173706, "learning_rate": 1.3143425441783869e-06, "loss": 1.2253, "step": 39009 }, { "epoch": 3.0572100313479624, "grad_norm": 2.558532476425171, "learning_rate": 1.3141341757313924e-06, "loss": 0.9759, "step": 39010 }, { "epoch": 3.0572884012539183, "grad_norm": 4.319552898406982, "learning_rate": 1.313925821303539e-06, "loss": 1.5957, "step": 39011 }, { "epoch": 3.0573667711598747, "grad_norm": 3.0302979946136475, "learning_rate": 1.3137174808956182e-06, "loss": 1.5938, "step": 39012 }, { "epoch": 3.0574451410658305, "grad_norm": 3.337576150894165, "learning_rate": 1.3135091545084228e-06, "loss": 1.7095, "step": 39013 }, { "epoch": 3.057523510971787, "grad_norm": 3.1360440254211426, "learning_rate": 1.313300842142745e-06, "loss": 0.9711, "step": 39014 }, { "epoch": 3.0576018808777428, "grad_norm": 6.124320030212402, "learning_rate": 1.3130925437993775e-06, "loss": 1.3191, "step": 39015 }, { "epoch": 3.057680250783699, "grad_norm": 3.1984024047851562, "learning_rate": 1.3128842594791118e-06, "loss": 1.1159, "step": 39016 }, { "epoch": 3.057758620689655, "grad_norm": 3.7988035678863525, "learning_rate": 1.3126759891827406e-06, "loss": 1.5602, "step": 39017 }, { "epoch": 3.0578369905956113, "grad_norm": 3.029839515686035, "learning_rate": 1.3124677329110563e-06, "loss": 1.1429, "step": 39018 }, { "epoch": 3.057915360501567, "grad_norm": 3.3831803798675537, "learning_rate": 1.3122594906648505e-06, "loss": 1.3899, "step": 39019 }, { "epoch": 3.0579937304075235, "grad_norm": 2.8075318336486816, "learning_rate": 1.312051262444917e-06, "loss": 1.4376, "step": 39020 }, { "epoch": 3.0580721003134794, "grad_norm": 2.787594795227051, "learning_rate": 1.3118430482520433e-06, "loss": 1.3404, "step": 39021 }, { "epoch": 3.0581504702194358, "grad_norm": 6.840291500091553, "learning_rate": 1.3116348480870255e-06, "loss": 1.3174, "step": 39022 }, { "epoch": 3.0582288401253916, "grad_norm": 2.971238136291504, "learning_rate": 1.311426661950655e-06, "loss": 1.4909, "step": 39023 }, { "epoch": 3.058307210031348, "grad_norm": 6.894536018371582, "learning_rate": 1.3112184898437237e-06, "loss": 1.5872, "step": 39024 }, { "epoch": 3.058385579937304, "grad_norm": 4.389256954193115, "learning_rate": 1.3110103317670215e-06, "loss": 1.5968, "step": 39025 }, { "epoch": 3.05846394984326, "grad_norm": 4.022115230560303, "learning_rate": 1.3108021877213394e-06, "loss": 2.1874, "step": 39026 }, { "epoch": 3.058542319749216, "grad_norm": 5.5653157234191895, "learning_rate": 1.3105940577074721e-06, "loss": 1.3333, "step": 39027 }, { "epoch": 3.0586206896551724, "grad_norm": 5.5703654289245605, "learning_rate": 1.310385941726212e-06, "loss": 1.6037, "step": 39028 }, { "epoch": 3.0586990595611283, "grad_norm": 3.848421573638916, "learning_rate": 1.3101778397783472e-06, "loss": 1.1452, "step": 39029 }, { "epoch": 3.0587774294670846, "grad_norm": 5.307425022125244, "learning_rate": 1.309969751864671e-06, "loss": 1.0457, "step": 39030 }, { "epoch": 3.058855799373041, "grad_norm": 3.574075222015381, "learning_rate": 1.309761677985974e-06, "loss": 1.4153, "step": 39031 }, { "epoch": 3.058934169278997, "grad_norm": 4.197956085205078, "learning_rate": 1.309553618143048e-06, "loss": 1.4752, "step": 39032 }, { "epoch": 3.059012539184953, "grad_norm": 5.347195148468018, "learning_rate": 1.3093455723366848e-06, "loss": 1.2889, "step": 39033 }, { "epoch": 3.059090909090909, "grad_norm": 2.7759835720062256, "learning_rate": 1.3091375405676755e-06, "loss": 0.8239, "step": 39034 }, { "epoch": 3.0591692789968654, "grad_norm": 3.3066978454589844, "learning_rate": 1.308929522836811e-06, "loss": 1.4921, "step": 39035 }, { "epoch": 3.0592476489028213, "grad_norm": 4.204065799713135, "learning_rate": 1.3087215191448827e-06, "loss": 1.8367, "step": 39036 }, { "epoch": 3.0593260188087776, "grad_norm": 2.7478673458099365, "learning_rate": 1.3085135294926821e-06, "loss": 1.224, "step": 39037 }, { "epoch": 3.0594043887147335, "grad_norm": 5.100580215454102, "learning_rate": 1.3083055538809998e-06, "loss": 0.8037, "step": 39038 }, { "epoch": 3.05948275862069, "grad_norm": 2.8094985485076904, "learning_rate": 1.3080975923106288e-06, "loss": 1.234, "step": 39039 }, { "epoch": 3.0595611285266457, "grad_norm": 2.5811004638671875, "learning_rate": 1.3078896447823547e-06, "loss": 1.2209, "step": 39040 }, { "epoch": 3.059639498432602, "grad_norm": 3.61552095413208, "learning_rate": 1.3076817112969736e-06, "loss": 1.3773, "step": 39041 }, { "epoch": 3.059717868338558, "grad_norm": 3.7440245151519775, "learning_rate": 1.3074737918552755e-06, "loss": 1.3644, "step": 39042 }, { "epoch": 3.0597962382445143, "grad_norm": 5.989102840423584, "learning_rate": 1.3072658864580512e-06, "loss": 1.3951, "step": 39043 }, { "epoch": 3.05987460815047, "grad_norm": 3.050187110900879, "learning_rate": 1.3070579951060898e-06, "loss": 1.3058, "step": 39044 }, { "epoch": 3.0599529780564265, "grad_norm": 4.025197505950928, "learning_rate": 1.3068501178001807e-06, "loss": 1.1999, "step": 39045 }, { "epoch": 3.0600313479623824, "grad_norm": 3.66821026802063, "learning_rate": 1.3066422545411205e-06, "loss": 1.3747, "step": 39046 }, { "epoch": 3.0601097178683387, "grad_norm": 2.7220139503479004, "learning_rate": 1.3064344053296945e-06, "loss": 1.9719, "step": 39047 }, { "epoch": 3.0601880877742946, "grad_norm": 4.950676918029785, "learning_rate": 1.306226570166695e-06, "loss": 1.3307, "step": 39048 }, { "epoch": 3.060266457680251, "grad_norm": 2.368756055831909, "learning_rate": 1.3060187490529118e-06, "loss": 1.4888, "step": 39049 }, { "epoch": 3.060344827586207, "grad_norm": 5.713038444519043, "learning_rate": 1.3058109419891367e-06, "loss": 0.8534, "step": 39050 }, { "epoch": 3.060423197492163, "grad_norm": 3.118304967880249, "learning_rate": 1.305603148976159e-06, "loss": 1.5411, "step": 39051 }, { "epoch": 3.060501567398119, "grad_norm": 2.7067453861236572, "learning_rate": 1.3053953700147693e-06, "loss": 1.6277, "step": 39052 }, { "epoch": 3.0605799373040754, "grad_norm": 2.9176201820373535, "learning_rate": 1.3051876051057576e-06, "loss": 1.6643, "step": 39053 }, { "epoch": 3.0606583072100313, "grad_norm": 2.905653715133667, "learning_rate": 1.3049798542499154e-06, "loss": 1.536, "step": 39054 }, { "epoch": 3.0607366771159876, "grad_norm": 2.947725772857666, "learning_rate": 1.304772117448031e-06, "loss": 1.2712, "step": 39055 }, { "epoch": 3.0608150470219435, "grad_norm": 3.1099748611450195, "learning_rate": 1.3045643947008962e-06, "loss": 1.6671, "step": 39056 }, { "epoch": 3.0608934169279, "grad_norm": 3.3294432163238525, "learning_rate": 1.3043566860093e-06, "loss": 1.1509, "step": 39057 }, { "epoch": 3.0609717868338557, "grad_norm": 2.824218273162842, "learning_rate": 1.304148991374034e-06, "loss": 1.3928, "step": 39058 }, { "epoch": 3.061050156739812, "grad_norm": 3.4546701908111572, "learning_rate": 1.3039413107958843e-06, "loss": 0.9144, "step": 39059 }, { "epoch": 3.061128526645768, "grad_norm": 3.151118040084839, "learning_rate": 1.3037336442756448e-06, "loss": 1.2807, "step": 39060 }, { "epoch": 3.0612068965517243, "grad_norm": 4.57086706161499, "learning_rate": 1.3035259918141059e-06, "loss": 1.3385, "step": 39061 }, { "epoch": 3.06128526645768, "grad_norm": 2.743929862976074, "learning_rate": 1.3033183534120536e-06, "loss": 1.548, "step": 39062 }, { "epoch": 3.0613636363636365, "grad_norm": 3.8155641555786133, "learning_rate": 1.3031107290702793e-06, "loss": 1.6313, "step": 39063 }, { "epoch": 3.0614420062695924, "grad_norm": 2.9510512351989746, "learning_rate": 1.3029031187895719e-06, "loss": 1.1405, "step": 39064 }, { "epoch": 3.0615203761755487, "grad_norm": 2.9055371284484863, "learning_rate": 1.302695522570725e-06, "loss": 0.8092, "step": 39065 }, { "epoch": 3.0615987460815046, "grad_norm": 2.1898608207702637, "learning_rate": 1.3024879404145236e-06, "loss": 1.4261, "step": 39066 }, { "epoch": 3.061677115987461, "grad_norm": 3.1565968990325928, "learning_rate": 1.3022803723217592e-06, "loss": 1.1998, "step": 39067 }, { "epoch": 3.061755485893417, "grad_norm": 3.721231460571289, "learning_rate": 1.3020728182932206e-06, "loss": 1.3244, "step": 39068 }, { "epoch": 3.061833855799373, "grad_norm": 6.058733940124512, "learning_rate": 1.301865278329698e-06, "loss": 1.2375, "step": 39069 }, { "epoch": 3.061912225705329, "grad_norm": 4.141164302825928, "learning_rate": 1.30165775243198e-06, "loss": 1.5641, "step": 39070 }, { "epoch": 3.0619905956112854, "grad_norm": 3.4529991149902344, "learning_rate": 1.3014502406008567e-06, "loss": 1.3953, "step": 39071 }, { "epoch": 3.0620689655172413, "grad_norm": 2.311886787414551, "learning_rate": 1.3012427428371171e-06, "loss": 1.3732, "step": 39072 }, { "epoch": 3.0621473354231976, "grad_norm": 19.946622848510742, "learning_rate": 1.3010352591415498e-06, "loss": 0.5599, "step": 39073 }, { "epoch": 3.0622257053291535, "grad_norm": 2.7530181407928467, "learning_rate": 1.3008277895149446e-06, "loss": 0.8125, "step": 39074 }, { "epoch": 3.06230407523511, "grad_norm": 3.106562376022339, "learning_rate": 1.3006203339580909e-06, "loss": 0.8393, "step": 39075 }, { "epoch": 3.0623824451410657, "grad_norm": 3.417616128921509, "learning_rate": 1.3004128924717785e-06, "loss": 1.4302, "step": 39076 }, { "epoch": 3.062460815047022, "grad_norm": 2.885406017303467, "learning_rate": 1.3002054650567936e-06, "loss": 1.2728, "step": 39077 }, { "epoch": 3.062539184952978, "grad_norm": 3.5175106525421143, "learning_rate": 1.2999980517139254e-06, "loss": 1.2506, "step": 39078 }, { "epoch": 3.0626175548589343, "grad_norm": 4.997549057006836, "learning_rate": 1.2997906524439658e-06, "loss": 1.1393, "step": 39079 }, { "epoch": 3.06269592476489, "grad_norm": 2.74300479888916, "learning_rate": 1.2995832672477031e-06, "loss": 1.4471, "step": 39080 }, { "epoch": 3.0627742946708465, "grad_norm": 2.725421905517578, "learning_rate": 1.299375896125924e-06, "loss": 1.2452, "step": 39081 }, { "epoch": 3.0628526645768024, "grad_norm": 6.918847560882568, "learning_rate": 1.2991685390794183e-06, "loss": 1.0167, "step": 39082 }, { "epoch": 3.0629310344827587, "grad_norm": 4.3447370529174805, "learning_rate": 1.2989611961089727e-06, "loss": 1.5519, "step": 39083 }, { "epoch": 3.0630094043887146, "grad_norm": 2.3936824798583984, "learning_rate": 1.298753867215381e-06, "loss": 1.056, "step": 39084 }, { "epoch": 3.063087774294671, "grad_norm": 4.256768226623535, "learning_rate": 1.2985465523994267e-06, "loss": 1.2594, "step": 39085 }, { "epoch": 3.063166144200627, "grad_norm": 2.869960069656372, "learning_rate": 1.2983392516619002e-06, "loss": 1.3697, "step": 39086 }, { "epoch": 3.063244514106583, "grad_norm": 4.692203521728516, "learning_rate": 1.29813196500359e-06, "loss": 0.8387, "step": 39087 }, { "epoch": 3.063322884012539, "grad_norm": 3.4292523860931396, "learning_rate": 1.2979246924252848e-06, "loss": 1.3966, "step": 39088 }, { "epoch": 3.0634012539184954, "grad_norm": 2.9658074378967285, "learning_rate": 1.297717433927772e-06, "loss": 1.2208, "step": 39089 }, { "epoch": 3.0634796238244513, "grad_norm": 3.0898959636688232, "learning_rate": 1.297510189511841e-06, "loss": 1.0201, "step": 39090 }, { "epoch": 3.0635579937304076, "grad_norm": 2.575716733932495, "learning_rate": 1.2973029591782804e-06, "loss": 1.2793, "step": 39091 }, { "epoch": 3.0636363636363635, "grad_norm": 2.7343289852142334, "learning_rate": 1.2970957429278747e-06, "loss": 1.5694, "step": 39092 }, { "epoch": 3.06371473354232, "grad_norm": 3.220877170562744, "learning_rate": 1.2968885407614167e-06, "loss": 1.3543, "step": 39093 }, { "epoch": 3.0637931034482757, "grad_norm": 6.584787845611572, "learning_rate": 1.2966813526796924e-06, "loss": 1.5113, "step": 39094 }, { "epoch": 3.063871473354232, "grad_norm": 2.256427526473999, "learning_rate": 1.2964741786834918e-06, "loss": 1.0976, "step": 39095 }, { "epoch": 3.063949843260188, "grad_norm": 4.259160041809082, "learning_rate": 1.2962670187735992e-06, "loss": 1.3346, "step": 39096 }, { "epoch": 3.0640282131661443, "grad_norm": 2.710179567337036, "learning_rate": 1.2960598729508035e-06, "loss": 1.3602, "step": 39097 }, { "epoch": 3.0641065830721, "grad_norm": 3.575516700744629, "learning_rate": 1.2958527412158944e-06, "loss": 1.1939, "step": 39098 }, { "epoch": 3.0641849529780565, "grad_norm": 3.409965991973877, "learning_rate": 1.2956456235696607e-06, "loss": 1.6055, "step": 39099 }, { "epoch": 3.0642633228840124, "grad_norm": 5.929725646972656, "learning_rate": 1.295438520012887e-06, "loss": 0.769, "step": 39100 }, { "epoch": 3.0643416927899687, "grad_norm": 4.723587989807129, "learning_rate": 1.2952314305463625e-06, "loss": 1.4659, "step": 39101 }, { "epoch": 3.0644200626959246, "grad_norm": 2.7787840366363525, "learning_rate": 1.2950243551708747e-06, "loss": 1.4741, "step": 39102 }, { "epoch": 3.064498432601881, "grad_norm": 3.1598494052886963, "learning_rate": 1.2948172938872118e-06, "loss": 1.4148, "step": 39103 }, { "epoch": 3.064576802507837, "grad_norm": 4.790929794311523, "learning_rate": 1.2946102466961596e-06, "loss": 1.3453, "step": 39104 }, { "epoch": 3.064655172413793, "grad_norm": 3.172541856765747, "learning_rate": 1.294403213598508e-06, "loss": 1.1408, "step": 39105 }, { "epoch": 3.064733542319749, "grad_norm": 2.030351400375366, "learning_rate": 1.2941961945950425e-06, "loss": 1.3486, "step": 39106 }, { "epoch": 3.0648119122257054, "grad_norm": 4.33827543258667, "learning_rate": 1.2939891896865514e-06, "loss": 1.8184, "step": 39107 }, { "epoch": 3.0648902821316613, "grad_norm": 3.4422197341918945, "learning_rate": 1.2937821988738219e-06, "loss": 1.202, "step": 39108 }, { "epoch": 3.0649686520376176, "grad_norm": 2.3267743587493896, "learning_rate": 1.2935752221576414e-06, "loss": 0.9268, "step": 39109 }, { "epoch": 3.0650470219435735, "grad_norm": 2.94064998626709, "learning_rate": 1.2933682595387986e-06, "loss": 1.4265, "step": 39110 }, { "epoch": 3.06512539184953, "grad_norm": 10.58859920501709, "learning_rate": 1.2931613110180758e-06, "loss": 1.6064, "step": 39111 }, { "epoch": 3.0652037617554857, "grad_norm": 2.199531078338623, "learning_rate": 1.2929543765962649e-06, "loss": 1.2683, "step": 39112 }, { "epoch": 3.065282131661442, "grad_norm": 2.957498073577881, "learning_rate": 1.2927474562741517e-06, "loss": 1.4577, "step": 39113 }, { "epoch": 3.065360501567398, "grad_norm": 3.2331576347351074, "learning_rate": 1.2925405500525246e-06, "loss": 1.7958, "step": 39114 }, { "epoch": 3.0654388714733543, "grad_norm": 5.810371398925781, "learning_rate": 1.292333657932167e-06, "loss": 1.2779, "step": 39115 }, { "epoch": 3.06551724137931, "grad_norm": 5.385998249053955, "learning_rate": 1.292126779913867e-06, "loss": 1.3019, "step": 39116 }, { "epoch": 3.0655956112852665, "grad_norm": 4.193897247314453, "learning_rate": 1.2919199159984147e-06, "loss": 1.208, "step": 39117 }, { "epoch": 3.0656739811912224, "grad_norm": 2.91768479347229, "learning_rate": 1.2917130661865933e-06, "loss": 1.471, "step": 39118 }, { "epoch": 3.0657523510971787, "grad_norm": 3.1193835735321045, "learning_rate": 1.2915062304791902e-06, "loss": 1.3871, "step": 39119 }, { "epoch": 3.0658307210031346, "grad_norm": 3.9549920558929443, "learning_rate": 1.2912994088769931e-06, "loss": 1.4333, "step": 39120 }, { "epoch": 3.065909090909091, "grad_norm": 2.812056303024292, "learning_rate": 1.2910926013807879e-06, "loss": 1.5046, "step": 39121 }, { "epoch": 3.0659874608150472, "grad_norm": 2.8659253120422363, "learning_rate": 1.2908858079913617e-06, "loss": 1.3445, "step": 39122 }, { "epoch": 3.066065830721003, "grad_norm": 4.3752923011779785, "learning_rate": 1.2906790287095006e-06, "loss": 1.7724, "step": 39123 }, { "epoch": 3.066144200626959, "grad_norm": 3.0897419452667236, "learning_rate": 1.2904722635359913e-06, "loss": 1.2801, "step": 39124 }, { "epoch": 3.0662225705329154, "grad_norm": 2.294705867767334, "learning_rate": 1.2902655124716201e-06, "loss": 0.662, "step": 39125 }, { "epoch": 3.0663009404388717, "grad_norm": 2.28125, "learning_rate": 1.2900587755171733e-06, "loss": 1.3072, "step": 39126 }, { "epoch": 3.0663793103448276, "grad_norm": 2.61108660697937, "learning_rate": 1.2898520526734376e-06, "loss": 0.9356, "step": 39127 }, { "epoch": 3.066457680250784, "grad_norm": 4.099897861480713, "learning_rate": 1.2896453439412005e-06, "loss": 1.5586, "step": 39128 }, { "epoch": 3.06653605015674, "grad_norm": 3.8472301959991455, "learning_rate": 1.2894386493212436e-06, "loss": 1.1823, "step": 39129 }, { "epoch": 3.066614420062696, "grad_norm": 2.527372121810913, "learning_rate": 1.2892319688143578e-06, "loss": 1.1994, "step": 39130 }, { "epoch": 3.066692789968652, "grad_norm": 2.677621841430664, "learning_rate": 1.2890253024213273e-06, "loss": 1.3862, "step": 39131 }, { "epoch": 3.0667711598746084, "grad_norm": 3.536100387573242, "learning_rate": 1.2888186501429406e-06, "loss": 1.4212, "step": 39132 }, { "epoch": 3.0668495297805642, "grad_norm": 3.2239997386932373, "learning_rate": 1.2886120119799795e-06, "loss": 1.3497, "step": 39133 }, { "epoch": 3.0669278996865206, "grad_norm": 2.914506435394287, "learning_rate": 1.2884053879332308e-06, "loss": 1.3569, "step": 39134 }, { "epoch": 3.0670062695924765, "grad_norm": 3.293931484222412, "learning_rate": 1.2881987780034832e-06, "loss": 0.766, "step": 39135 }, { "epoch": 3.067084639498433, "grad_norm": 3.8042659759521484, "learning_rate": 1.2879921821915226e-06, "loss": 0.9693, "step": 39136 }, { "epoch": 3.0671630094043887, "grad_norm": 3.288893699645996, "learning_rate": 1.287785600498131e-06, "loss": 1.2172, "step": 39137 }, { "epoch": 3.067241379310345, "grad_norm": 6.738462448120117, "learning_rate": 1.2875790329240968e-06, "loss": 0.7694, "step": 39138 }, { "epoch": 3.067319749216301, "grad_norm": 4.402028560638428, "learning_rate": 1.2873724794702036e-06, "loss": 1.3676, "step": 39139 }, { "epoch": 3.0673981191222572, "grad_norm": 3.0747148990631104, "learning_rate": 1.287165940137241e-06, "loss": 1.5535, "step": 39140 }, { "epoch": 3.067476489028213, "grad_norm": 3.3386073112487793, "learning_rate": 1.286959414925991e-06, "loss": 0.763, "step": 39141 }, { "epoch": 3.0675548589341695, "grad_norm": 2.729494571685791, "learning_rate": 1.2867529038372407e-06, "loss": 1.376, "step": 39142 }, { "epoch": 3.0676332288401253, "grad_norm": 3.5521390438079834, "learning_rate": 1.2865464068717749e-06, "loss": 1.33, "step": 39143 }, { "epoch": 3.0677115987460817, "grad_norm": 2.298807382583618, "learning_rate": 1.2863399240303787e-06, "loss": 1.2606, "step": 39144 }, { "epoch": 3.0677899686520376, "grad_norm": 7.335634708404541, "learning_rate": 1.2861334553138382e-06, "loss": 0.5882, "step": 39145 }, { "epoch": 3.067868338557994, "grad_norm": 9.790999412536621, "learning_rate": 1.2859270007229392e-06, "loss": 1.8815, "step": 39146 }, { "epoch": 3.06794670846395, "grad_norm": 2.5454256534576416, "learning_rate": 1.2857205602584667e-06, "loss": 1.57, "step": 39147 }, { "epoch": 3.068025078369906, "grad_norm": 5.997561454772949, "learning_rate": 1.285514133921203e-06, "loss": 0.8219, "step": 39148 }, { "epoch": 3.068103448275862, "grad_norm": 2.4923460483551025, "learning_rate": 1.2853077217119369e-06, "loss": 0.9951, "step": 39149 }, { "epoch": 3.0681818181818183, "grad_norm": 8.735283851623535, "learning_rate": 1.2851013236314525e-06, "loss": 1.7024, "step": 39150 }, { "epoch": 3.0682601880877742, "grad_norm": 3.4725899696350098, "learning_rate": 1.2848949396805354e-06, "loss": 1.3635, "step": 39151 }, { "epoch": 3.0683385579937306, "grad_norm": 3.1351048946380615, "learning_rate": 1.2846885698599694e-06, "loss": 1.2465, "step": 39152 }, { "epoch": 3.0684169278996865, "grad_norm": 2.804097890853882, "learning_rate": 1.2844822141705372e-06, "loss": 0.9927, "step": 39153 }, { "epoch": 3.068495297805643, "grad_norm": 3.3666186332702637, "learning_rate": 1.2842758726130283e-06, "loss": 1.2417, "step": 39154 }, { "epoch": 3.0685736677115987, "grad_norm": 4.374941349029541, "learning_rate": 1.284069545188227e-06, "loss": 1.2836, "step": 39155 }, { "epoch": 3.068652037617555, "grad_norm": 3.184574604034424, "learning_rate": 1.2838632318969145e-06, "loss": 1.528, "step": 39156 }, { "epoch": 3.068730407523511, "grad_norm": 3.4039573669433594, "learning_rate": 1.283656932739878e-06, "loss": 1.2127, "step": 39157 }, { "epoch": 3.0688087774294672, "grad_norm": 3.377713441848755, "learning_rate": 1.2834506477179016e-06, "loss": 1.5354, "step": 39158 }, { "epoch": 3.068887147335423, "grad_norm": 13.738970756530762, "learning_rate": 1.28324437683177e-06, "loss": 1.7771, "step": 39159 }, { "epoch": 3.0689655172413794, "grad_norm": 5.146871566772461, "learning_rate": 1.283038120082268e-06, "loss": 1.1801, "step": 39160 }, { "epoch": 3.0690438871473353, "grad_norm": 4.294777870178223, "learning_rate": 1.2828318774701792e-06, "loss": 1.4309, "step": 39161 }, { "epoch": 3.0691222570532917, "grad_norm": 3.37697434425354, "learning_rate": 1.2826256489962891e-06, "loss": 0.717, "step": 39162 }, { "epoch": 3.0692006269592476, "grad_norm": 2.8855268955230713, "learning_rate": 1.2824194346613817e-06, "loss": 1.5016, "step": 39163 }, { "epoch": 3.069278996865204, "grad_norm": 2.531398057937622, "learning_rate": 1.2822132344662414e-06, "loss": 1.3597, "step": 39164 }, { "epoch": 3.06935736677116, "grad_norm": 6.627399444580078, "learning_rate": 1.282007048411652e-06, "loss": 1.5792, "step": 39165 }, { "epoch": 3.069435736677116, "grad_norm": 7.348434925079346, "learning_rate": 1.2818008764983996e-06, "loss": 1.0202, "step": 39166 }, { "epoch": 3.069514106583072, "grad_norm": 3.318589448928833, "learning_rate": 1.2815947187272642e-06, "loss": 1.3627, "step": 39167 }, { "epoch": 3.0695924764890283, "grad_norm": 2.614452362060547, "learning_rate": 1.2813885750990341e-06, "loss": 1.1413, "step": 39168 }, { "epoch": 3.069670846394984, "grad_norm": 3.240478038787842, "learning_rate": 1.2811824456144918e-06, "loss": 1.6758, "step": 39169 }, { "epoch": 3.0697492163009406, "grad_norm": 4.247640609741211, "learning_rate": 1.2809763302744228e-06, "loss": 1.5788, "step": 39170 }, { "epoch": 3.0698275862068964, "grad_norm": 6.841688632965088, "learning_rate": 1.2807702290796082e-06, "loss": 1.3403, "step": 39171 }, { "epoch": 3.0699059561128528, "grad_norm": 8.660572052001953, "learning_rate": 1.2805641420308318e-06, "loss": 1.2895, "step": 39172 }, { "epoch": 3.0699843260188087, "grad_norm": 2.3213295936584473, "learning_rate": 1.2803580691288819e-06, "loss": 1.1329, "step": 39173 }, { "epoch": 3.070062695924765, "grad_norm": 8.859128952026367, "learning_rate": 1.2801520103745379e-06, "loss": 0.9182, "step": 39174 }, { "epoch": 3.070141065830721, "grad_norm": 5.520473480224609, "learning_rate": 1.2799459657685848e-06, "loss": 1.6701, "step": 39175 }, { "epoch": 3.070219435736677, "grad_norm": 3.2943274974823, "learning_rate": 1.2797399353118067e-06, "loss": 1.3597, "step": 39176 }, { "epoch": 3.070297805642633, "grad_norm": 4.196145534515381, "learning_rate": 1.2795339190049872e-06, "loss": 1.4449, "step": 39177 }, { "epoch": 3.0703761755485894, "grad_norm": 3.063908576965332, "learning_rate": 1.2793279168489098e-06, "loss": 1.3384, "step": 39178 }, { "epoch": 3.0704545454545453, "grad_norm": 8.781956672668457, "learning_rate": 1.2791219288443575e-06, "loss": 1.3585, "step": 39179 }, { "epoch": 3.0705329153605017, "grad_norm": 5.522148132324219, "learning_rate": 1.2789159549921149e-06, "loss": 1.5473, "step": 39180 }, { "epoch": 3.0706112852664575, "grad_norm": 2.1162211894989014, "learning_rate": 1.2787099952929638e-06, "loss": 1.4095, "step": 39181 }, { "epoch": 3.070689655172414, "grad_norm": 6.029783248901367, "learning_rate": 1.2785040497476892e-06, "loss": 1.1572, "step": 39182 }, { "epoch": 3.0707680250783698, "grad_norm": 3.7975964546203613, "learning_rate": 1.2782981183570736e-06, "loss": 1.5439, "step": 39183 }, { "epoch": 3.070846394984326, "grad_norm": 3.835658550262451, "learning_rate": 1.2780922011219017e-06, "loss": 1.6679, "step": 39184 }, { "epoch": 3.070924764890282, "grad_norm": 2.116600751876831, "learning_rate": 1.277886298042954e-06, "loss": 1.47, "step": 39185 }, { "epoch": 3.0710031347962383, "grad_norm": 2.8727591037750244, "learning_rate": 1.2776804091210137e-06, "loss": 1.4119, "step": 39186 }, { "epoch": 3.071081504702194, "grad_norm": 7.392218589782715, "learning_rate": 1.2774745343568662e-06, "loss": 2.2327, "step": 39187 }, { "epoch": 3.0711598746081505, "grad_norm": 4.8276801109313965, "learning_rate": 1.2772686737512958e-06, "loss": 1.2637, "step": 39188 }, { "epoch": 3.0712382445141064, "grad_norm": 3.7379090785980225, "learning_rate": 1.2770628273050811e-06, "loss": 1.3755, "step": 39189 }, { "epoch": 3.0713166144200628, "grad_norm": 3.3191795349121094, "learning_rate": 1.2768569950190074e-06, "loss": 1.5233, "step": 39190 }, { "epoch": 3.0713949843260187, "grad_norm": 2.310138463973999, "learning_rate": 1.276651176893856e-06, "loss": 1.0692, "step": 39191 }, { "epoch": 3.071473354231975, "grad_norm": 4.616400241851807, "learning_rate": 1.2764453729304138e-06, "loss": 1.688, "step": 39192 }, { "epoch": 3.071551724137931, "grad_norm": 2.4956259727478027, "learning_rate": 1.2762395831294593e-06, "loss": 1.4161, "step": 39193 }, { "epoch": 3.071630094043887, "grad_norm": 2.990194797515869, "learning_rate": 1.2760338074917772e-06, "loss": 1.3568, "step": 39194 }, { "epoch": 3.071708463949843, "grad_norm": 3.4290237426757812, "learning_rate": 1.2758280460181494e-06, "loss": 1.038, "step": 39195 }, { "epoch": 3.0717868338557994, "grad_norm": 2.8687829971313477, "learning_rate": 1.275622298709359e-06, "loss": 1.2003, "step": 39196 }, { "epoch": 3.0718652037617553, "grad_norm": 3.2470946311950684, "learning_rate": 1.2754165655661887e-06, "loss": 1.511, "step": 39197 }, { "epoch": 3.0719435736677116, "grad_norm": 4.332860946655273, "learning_rate": 1.2752108465894202e-06, "loss": 1.339, "step": 39198 }, { "epoch": 3.0720219435736675, "grad_norm": 8.511249542236328, "learning_rate": 1.2750051417798382e-06, "loss": 1.1623, "step": 39199 }, { "epoch": 3.072100313479624, "grad_norm": 4.517153263092041, "learning_rate": 1.2747994511382206e-06, "loss": 1.7306, "step": 39200 }, { "epoch": 3.0721786833855798, "grad_norm": 4.589385986328125, "learning_rate": 1.2745937746653536e-06, "loss": 1.5361, "step": 39201 }, { "epoch": 3.072257053291536, "grad_norm": 6.3062052726745605, "learning_rate": 1.2743881123620188e-06, "loss": 1.4744, "step": 39202 }, { "epoch": 3.072335423197492, "grad_norm": 3.2367467880249023, "learning_rate": 1.274182464228999e-06, "loss": 1.2334, "step": 39203 }, { "epoch": 3.0724137931034483, "grad_norm": 4.574675559997559, "learning_rate": 1.2739768302670746e-06, "loss": 1.397, "step": 39204 }, { "epoch": 3.072492163009404, "grad_norm": 7.028929710388184, "learning_rate": 1.2737712104770267e-06, "loss": 1.4132, "step": 39205 }, { "epoch": 3.0725705329153605, "grad_norm": 4.743207931518555, "learning_rate": 1.2735656048596401e-06, "loss": 1.2314, "step": 39206 }, { "epoch": 3.0726489028213164, "grad_norm": 3.0981485843658447, "learning_rate": 1.2733600134156982e-06, "loss": 1.1785, "step": 39207 }, { "epoch": 3.0727272727272728, "grad_norm": 3.1441659927368164, "learning_rate": 1.2731544361459792e-06, "loss": 1.4444, "step": 39208 }, { "epoch": 3.0728056426332286, "grad_norm": 2.4264087677001953, "learning_rate": 1.2729488730512663e-06, "loss": 1.2216, "step": 39209 }, { "epoch": 3.072884012539185, "grad_norm": 6.586785316467285, "learning_rate": 1.2727433241323395e-06, "loss": 1.4533, "step": 39210 }, { "epoch": 3.072962382445141, "grad_norm": 5.8080668449401855, "learning_rate": 1.2725377893899865e-06, "loss": 2.2279, "step": 39211 }, { "epoch": 3.073040752351097, "grad_norm": 2.554126501083374, "learning_rate": 1.2723322688249834e-06, "loss": 1.5217, "step": 39212 }, { "epoch": 3.073119122257053, "grad_norm": 3.811727523803711, "learning_rate": 1.2721267624381133e-06, "loss": 1.1354, "step": 39213 }, { "epoch": 3.0731974921630094, "grad_norm": 5.173627853393555, "learning_rate": 1.2719212702301587e-06, "loss": 1.6305, "step": 39214 }, { "epoch": 3.0732758620689653, "grad_norm": 3.209935426712036, "learning_rate": 1.2717157922019008e-06, "loss": 1.2207, "step": 39215 }, { "epoch": 3.0733542319749216, "grad_norm": 4.124800682067871, "learning_rate": 1.2715103283541208e-06, "loss": 1.653, "step": 39216 }, { "epoch": 3.073432601880878, "grad_norm": 16.76500701904297, "learning_rate": 1.2713048786876003e-06, "loss": 1.6783, "step": 39217 }, { "epoch": 3.073510971786834, "grad_norm": 2.1910505294799805, "learning_rate": 1.271099443203121e-06, "loss": 1.4685, "step": 39218 }, { "epoch": 3.0735893416927897, "grad_norm": 2.5709385871887207, "learning_rate": 1.2708940219014643e-06, "loss": 1.3862, "step": 39219 }, { "epoch": 3.073667711598746, "grad_norm": 2.6655447483062744, "learning_rate": 1.2706886147834114e-06, "loss": 1.211, "step": 39220 }, { "epoch": 3.0737460815047024, "grad_norm": 5.797098636627197, "learning_rate": 1.2704832218497432e-06, "loss": 1.2125, "step": 39221 }, { "epoch": 3.0738244514106583, "grad_norm": 4.084399700164795, "learning_rate": 1.270277843101243e-06, "loss": 1.5555, "step": 39222 }, { "epoch": 3.0739028213166146, "grad_norm": 2.762155055999756, "learning_rate": 1.270072478538687e-06, "loss": 1.8743, "step": 39223 }, { "epoch": 3.0739811912225705, "grad_norm": 3.251497507095337, "learning_rate": 1.269867128162861e-06, "loss": 1.1992, "step": 39224 }, { "epoch": 3.074059561128527, "grad_norm": 3.391000270843506, "learning_rate": 1.2696617919745447e-06, "loss": 1.5845, "step": 39225 }, { "epoch": 3.0741379310344827, "grad_norm": 6.737506866455078, "learning_rate": 1.2694564699745203e-06, "loss": 0.6594, "step": 39226 }, { "epoch": 3.074216300940439, "grad_norm": 4.386610507965088, "learning_rate": 1.2692511621635656e-06, "loss": 1.7017, "step": 39227 }, { "epoch": 3.074294670846395, "grad_norm": 3.2828867435455322, "learning_rate": 1.2690458685424638e-06, "loss": 1.67, "step": 39228 }, { "epoch": 3.0743730407523513, "grad_norm": 13.303333282470703, "learning_rate": 1.268840589111995e-06, "loss": 1.3975, "step": 39229 }, { "epoch": 3.074451410658307, "grad_norm": 3.658797264099121, "learning_rate": 1.2686353238729398e-06, "loss": 1.3651, "step": 39230 }, { "epoch": 3.0745297805642635, "grad_norm": 3.130551338195801, "learning_rate": 1.2684300728260795e-06, "loss": 0.8742, "step": 39231 }, { "epoch": 3.0746081504702194, "grad_norm": 4.1813836097717285, "learning_rate": 1.268224835972195e-06, "loss": 1.5966, "step": 39232 }, { "epoch": 3.0746865203761757, "grad_norm": 2.2483091354370117, "learning_rate": 1.2680196133120659e-06, "loss": 1.5353, "step": 39233 }, { "epoch": 3.0747648902821316, "grad_norm": 2.5650501251220703, "learning_rate": 1.267814404846473e-06, "loss": 1.5849, "step": 39234 }, { "epoch": 3.074843260188088, "grad_norm": 5.450230598449707, "learning_rate": 1.2676092105761979e-06, "loss": 1.8971, "step": 39235 }, { "epoch": 3.074921630094044, "grad_norm": 3.545316457748413, "learning_rate": 1.2674040305020196e-06, "loss": 1.2434, "step": 39236 }, { "epoch": 3.075, "grad_norm": 2.619410276412964, "learning_rate": 1.2671988646247196e-06, "loss": 1.1914, "step": 39237 }, { "epoch": 3.075078369905956, "grad_norm": 4.761558532714844, "learning_rate": 1.2669937129450782e-06, "loss": 1.3494, "step": 39238 }, { "epoch": 3.0751567398119124, "grad_norm": 2.670034646987915, "learning_rate": 1.2667885754638749e-06, "loss": 1.0001, "step": 39239 }, { "epoch": 3.0752351097178683, "grad_norm": 3.031543254852295, "learning_rate": 1.2665834521818914e-06, "loss": 1.2044, "step": 39240 }, { "epoch": 3.0753134796238246, "grad_norm": 2.357943534851074, "learning_rate": 1.266378343099906e-06, "loss": 1.2469, "step": 39241 }, { "epoch": 3.0753918495297805, "grad_norm": 9.121537208557129, "learning_rate": 1.2661732482186984e-06, "loss": 1.2462, "step": 39242 }, { "epoch": 3.075470219435737, "grad_norm": 2.7246716022491455, "learning_rate": 1.2659681675390506e-06, "loss": 1.3577, "step": 39243 }, { "epoch": 3.0755485893416927, "grad_norm": 2.402653217315674, "learning_rate": 1.2657631010617437e-06, "loss": 1.2811, "step": 39244 }, { "epoch": 3.075626959247649, "grad_norm": 3.8425893783569336, "learning_rate": 1.2655580487875552e-06, "loss": 0.9865, "step": 39245 }, { "epoch": 3.075705329153605, "grad_norm": 3.0317535400390625, "learning_rate": 1.2653530107172652e-06, "loss": 1.2476, "step": 39246 }, { "epoch": 3.0757836990595613, "grad_norm": 2.13592791557312, "learning_rate": 1.2651479868516524e-06, "loss": 1.4882, "step": 39247 }, { "epoch": 3.075862068965517, "grad_norm": 2.8516361713409424, "learning_rate": 1.2649429771915018e-06, "loss": 1.0516, "step": 39248 }, { "epoch": 3.0759404388714735, "grad_norm": 2.8057005405426025, "learning_rate": 1.2647379817375876e-06, "loss": 1.5967, "step": 39249 }, { "epoch": 3.0760188087774294, "grad_norm": 6.0185933113098145, "learning_rate": 1.2645330004906919e-06, "loss": 1.4224, "step": 39250 }, { "epoch": 3.0760971786833857, "grad_norm": 8.60781192779541, "learning_rate": 1.264328033451594e-06, "loss": 1.4021, "step": 39251 }, { "epoch": 3.0761755485893416, "grad_norm": 2.4246504306793213, "learning_rate": 1.2641230806210735e-06, "loss": 1.3854, "step": 39252 }, { "epoch": 3.076253918495298, "grad_norm": 5.693977355957031, "learning_rate": 1.2639181419999096e-06, "loss": 1.8504, "step": 39253 }, { "epoch": 3.076332288401254, "grad_norm": 2.966850996017456, "learning_rate": 1.2637132175888828e-06, "loss": 1.3997, "step": 39254 }, { "epoch": 3.07641065830721, "grad_norm": 2.7593064308166504, "learning_rate": 1.2635083073887727e-06, "loss": 1.1951, "step": 39255 }, { "epoch": 3.076489028213166, "grad_norm": 5.879617691040039, "learning_rate": 1.263303411400355e-06, "loss": 1.4009, "step": 39256 }, { "epoch": 3.0765673981191224, "grad_norm": 4.072602272033691, "learning_rate": 1.2630985296244135e-06, "loss": 1.3172, "step": 39257 }, { "epoch": 3.0766457680250783, "grad_norm": 2.43691086769104, "learning_rate": 1.2628936620617255e-06, "loss": 1.4752, "step": 39258 }, { "epoch": 3.0767241379310346, "grad_norm": 2.5903100967407227, "learning_rate": 1.2626888087130724e-06, "loss": 1.2351, "step": 39259 }, { "epoch": 3.0768025078369905, "grad_norm": 5.023110866546631, "learning_rate": 1.2624839695792297e-06, "loss": 1.2868, "step": 39260 }, { "epoch": 3.076880877742947, "grad_norm": 2.6867613792419434, "learning_rate": 1.2622791446609772e-06, "loss": 1.5085, "step": 39261 }, { "epoch": 3.0769592476489027, "grad_norm": 2.017378807067871, "learning_rate": 1.2620743339590957e-06, "loss": 1.2979, "step": 39262 }, { "epoch": 3.077037617554859, "grad_norm": 2.7205963134765625, "learning_rate": 1.2618695374743656e-06, "loss": 1.2288, "step": 39263 }, { "epoch": 3.077115987460815, "grad_norm": 2.5353379249572754, "learning_rate": 1.2616647552075623e-06, "loss": 1.0517, "step": 39264 }, { "epoch": 3.0771943573667713, "grad_norm": 3.7064974308013916, "learning_rate": 1.2614599871594657e-06, "loss": 1.534, "step": 39265 }, { "epoch": 3.077272727272727, "grad_norm": 2.890442132949829, "learning_rate": 1.2612552333308543e-06, "loss": 0.9804, "step": 39266 }, { "epoch": 3.0773510971786835, "grad_norm": 6.2011919021606445, "learning_rate": 1.26105049372251e-06, "loss": 1.5764, "step": 39267 }, { "epoch": 3.0774294670846394, "grad_norm": 2.7330586910247803, "learning_rate": 1.2608457683352076e-06, "loss": 1.053, "step": 39268 }, { "epoch": 3.0775078369905957, "grad_norm": 3.390599489212036, "learning_rate": 1.260641057169728e-06, "loss": 1.1828, "step": 39269 }, { "epoch": 3.0775862068965516, "grad_norm": 2.523085832595825, "learning_rate": 1.2604363602268488e-06, "loss": 1.3683, "step": 39270 }, { "epoch": 3.077664576802508, "grad_norm": 5.309132099151611, "learning_rate": 1.2602316775073486e-06, "loss": 1.8112, "step": 39271 }, { "epoch": 3.077742946708464, "grad_norm": 4.992584228515625, "learning_rate": 1.260027009012007e-06, "loss": 1.1759, "step": 39272 }, { "epoch": 3.07782131661442, "grad_norm": 4.591814041137695, "learning_rate": 1.259822354741601e-06, "loss": 0.835, "step": 39273 }, { "epoch": 3.077899686520376, "grad_norm": 5.611784934997559, "learning_rate": 1.2596177146969113e-06, "loss": 1.3324, "step": 39274 }, { "epoch": 3.0779780564263324, "grad_norm": 3.8764824867248535, "learning_rate": 1.259413088878712e-06, "loss": 1.2321, "step": 39275 }, { "epoch": 3.0780564263322883, "grad_norm": 4.2423996925354, "learning_rate": 1.259208477287785e-06, "loss": 1.3573, "step": 39276 }, { "epoch": 3.0781347962382446, "grad_norm": 7.782683372497559, "learning_rate": 1.2590038799249082e-06, "loss": 1.3825, "step": 39277 }, { "epoch": 3.0782131661442005, "grad_norm": 6.388368129730225, "learning_rate": 1.2587992967908602e-06, "loss": 0.8645, "step": 39278 }, { "epoch": 3.078291536050157, "grad_norm": 6.298194408416748, "learning_rate": 1.2585947278864163e-06, "loss": 1.187, "step": 39279 }, { "epoch": 3.0783699059561127, "grad_norm": 6.61259126663208, "learning_rate": 1.2583901732123555e-06, "loss": 1.4152, "step": 39280 }, { "epoch": 3.078448275862069, "grad_norm": 2.3695485591888428, "learning_rate": 1.2581856327694581e-06, "loss": 1.1056, "step": 39281 }, { "epoch": 3.078526645768025, "grad_norm": 3.5938873291015625, "learning_rate": 1.257981106558502e-06, "loss": 1.4751, "step": 39282 }, { "epoch": 3.0786050156739813, "grad_norm": 3.269446849822998, "learning_rate": 1.2577765945802617e-06, "loss": 1.2525, "step": 39283 }, { "epoch": 3.078683385579937, "grad_norm": 2.480576515197754, "learning_rate": 1.2575720968355176e-06, "loss": 0.5615, "step": 39284 }, { "epoch": 3.0787617554858935, "grad_norm": 2.899167776107788, "learning_rate": 1.2573676133250467e-06, "loss": 1.4416, "step": 39285 }, { "epoch": 3.0788401253918494, "grad_norm": 2.9871435165405273, "learning_rate": 1.2571631440496274e-06, "loss": 1.1931, "step": 39286 }, { "epoch": 3.0789184952978057, "grad_norm": 2.4681384563446045, "learning_rate": 1.256958689010037e-06, "loss": 1.0144, "step": 39287 }, { "epoch": 3.0789968652037616, "grad_norm": 5.023195743560791, "learning_rate": 1.2567542482070532e-06, "loss": 1.5773, "step": 39288 }, { "epoch": 3.079075235109718, "grad_norm": 2.780613899230957, "learning_rate": 1.256549821641454e-06, "loss": 1.7268, "step": 39289 }, { "epoch": 3.079153605015674, "grad_norm": 5.334515571594238, "learning_rate": 1.2563454093140159e-06, "loss": 1.4548, "step": 39290 }, { "epoch": 3.07923197492163, "grad_norm": 4.530367851257324, "learning_rate": 1.2561410112255174e-06, "loss": 1.2205, "step": 39291 }, { "epoch": 3.079310344827586, "grad_norm": 4.233277797698975, "learning_rate": 1.2559366273767347e-06, "loss": 0.7296, "step": 39292 }, { "epoch": 3.0793887147335424, "grad_norm": 4.052683353424072, "learning_rate": 1.2557322577684483e-06, "loss": 1.1343, "step": 39293 }, { "epoch": 3.0794670846394983, "grad_norm": 4.500323295593262, "learning_rate": 1.25552790240143e-06, "loss": 1.3206, "step": 39294 }, { "epoch": 3.0795454545454546, "grad_norm": 2.781301259994507, "learning_rate": 1.255323561276462e-06, "loss": 1.0335, "step": 39295 }, { "epoch": 3.0796238244514105, "grad_norm": 3.1881461143493652, "learning_rate": 1.2551192343943213e-06, "loss": 1.4067, "step": 39296 }, { "epoch": 3.079702194357367, "grad_norm": 7.040390968322754, "learning_rate": 1.2549149217557815e-06, "loss": 1.5926, "step": 39297 }, { "epoch": 3.0797805642633227, "grad_norm": 2.8549671173095703, "learning_rate": 1.2547106233616224e-06, "loss": 1.3733, "step": 39298 }, { "epoch": 3.079858934169279, "grad_norm": 3.515685796737671, "learning_rate": 1.2545063392126184e-06, "loss": 1.0315, "step": 39299 }, { "epoch": 3.079937304075235, "grad_norm": 6.745458602905273, "learning_rate": 1.2543020693095514e-06, "loss": 1.5008, "step": 39300 }, { "epoch": 3.0800156739811912, "grad_norm": 3.784550428390503, "learning_rate": 1.2540978136531945e-06, "loss": 1.5315, "step": 39301 }, { "epoch": 3.080094043887147, "grad_norm": 3.2964563369750977, "learning_rate": 1.2538935722443251e-06, "loss": 1.7681, "step": 39302 }, { "epoch": 3.0801724137931035, "grad_norm": 2.587921619415283, "learning_rate": 1.2536893450837207e-06, "loss": 1.031, "step": 39303 }, { "epoch": 3.0802507836990594, "grad_norm": 3.135402202606201, "learning_rate": 1.2534851321721575e-06, "loss": 1.3743, "step": 39304 }, { "epoch": 3.0803291536050157, "grad_norm": 3.296874761581421, "learning_rate": 1.2532809335104124e-06, "loss": 1.6758, "step": 39305 }, { "epoch": 3.0804075235109716, "grad_norm": 4.713821887969971, "learning_rate": 1.2530767490992623e-06, "loss": 1.008, "step": 39306 }, { "epoch": 3.080485893416928, "grad_norm": 3.250526189804077, "learning_rate": 1.252872578939484e-06, "loss": 1.1527, "step": 39307 }, { "epoch": 3.080564263322884, "grad_norm": 2.536500930786133, "learning_rate": 1.2526684230318537e-06, "loss": 1.7155, "step": 39308 }, { "epoch": 3.08064263322884, "grad_norm": 5.750361442565918, "learning_rate": 1.2524642813771476e-06, "loss": 1.6295, "step": 39309 }, { "epoch": 3.080721003134796, "grad_norm": 4.102429389953613, "learning_rate": 1.252260153976143e-06, "loss": 1.5183, "step": 39310 }, { "epoch": 3.0807993730407524, "grad_norm": 5.841999530792236, "learning_rate": 1.2520560408296167e-06, "loss": 1.9561, "step": 39311 }, { "epoch": 3.0808777429467087, "grad_norm": 3.587801218032837, "learning_rate": 1.2518519419383434e-06, "loss": 1.4994, "step": 39312 }, { "epoch": 3.0809561128526646, "grad_norm": 3.142886161804199, "learning_rate": 1.2516478573030983e-06, "loss": 1.3007, "step": 39313 }, { "epoch": 3.0810344827586205, "grad_norm": 3.146267890930176, "learning_rate": 1.251443786924661e-06, "loss": 1.4987, "step": 39314 }, { "epoch": 3.081112852664577, "grad_norm": 3.3775882720947266, "learning_rate": 1.251239730803807e-06, "loss": 1.546, "step": 39315 }, { "epoch": 3.081191222570533, "grad_norm": 3.254196882247925, "learning_rate": 1.2510356889413111e-06, "loss": 1.3599, "step": 39316 }, { "epoch": 3.081269592476489, "grad_norm": 2.2372829914093018, "learning_rate": 1.2508316613379496e-06, "loss": 1.5036, "step": 39317 }, { "epoch": 3.0813479623824453, "grad_norm": 6.437839984893799, "learning_rate": 1.2506276479944967e-06, "loss": 1.757, "step": 39318 }, { "epoch": 3.0814263322884012, "grad_norm": 5.94962739944458, "learning_rate": 1.250423648911734e-06, "loss": 1.6914, "step": 39319 }, { "epoch": 3.0815047021943576, "grad_norm": 7.206260681152344, "learning_rate": 1.2502196640904323e-06, "loss": 1.4501, "step": 39320 }, { "epoch": 3.0815830721003135, "grad_norm": 7.493871212005615, "learning_rate": 1.250015693531369e-06, "loss": 0.276, "step": 39321 }, { "epoch": 3.08166144200627, "grad_norm": 3.941051959991455, "learning_rate": 1.2498117372353197e-06, "loss": 1.095, "step": 39322 }, { "epoch": 3.0817398119122257, "grad_norm": 4.6437811851501465, "learning_rate": 1.24960779520306e-06, "loss": 2.0755, "step": 39323 }, { "epoch": 3.081818181818182, "grad_norm": 4.161489009857178, "learning_rate": 1.2494038674353665e-06, "loss": 0.8783, "step": 39324 }, { "epoch": 3.081896551724138, "grad_norm": 3.1625187397003174, "learning_rate": 1.2491999539330142e-06, "loss": 1.4799, "step": 39325 }, { "epoch": 3.0819749216300942, "grad_norm": 3.9364101886749268, "learning_rate": 1.2489960546967783e-06, "loss": 1.2197, "step": 39326 }, { "epoch": 3.08205329153605, "grad_norm": 5.251888275146484, "learning_rate": 1.2487921697274352e-06, "loss": 1.192, "step": 39327 }, { "epoch": 3.0821316614420065, "grad_norm": 2.975923776626587, "learning_rate": 1.2485882990257598e-06, "loss": 1.6133, "step": 39328 }, { "epoch": 3.0822100313479623, "grad_norm": 1.8688685894012451, "learning_rate": 1.2483844425925278e-06, "loss": 1.3982, "step": 39329 }, { "epoch": 3.0822884012539187, "grad_norm": 7.679153919219971, "learning_rate": 1.2481806004285158e-06, "loss": 1.5757, "step": 39330 }, { "epoch": 3.0823667711598746, "grad_norm": 6.916134357452393, "learning_rate": 1.247976772534495e-06, "loss": 1.3393, "step": 39331 }, { "epoch": 3.082445141065831, "grad_norm": 6.0299973487854, "learning_rate": 1.247772958911244e-06, "loss": 1.4482, "step": 39332 }, { "epoch": 3.082523510971787, "grad_norm": 8.500699043273926, "learning_rate": 1.2475691595595385e-06, "loss": 1.2559, "step": 39333 }, { "epoch": 3.082601880877743, "grad_norm": 5.362154006958008, "learning_rate": 1.2473653744801534e-06, "loss": 1.5393, "step": 39334 }, { "epoch": 3.082680250783699, "grad_norm": 3.419602155685425, "learning_rate": 1.2471616036738616e-06, "loss": 1.2758, "step": 39335 }, { "epoch": 3.0827586206896553, "grad_norm": 3.763672351837158, "learning_rate": 1.246957847141438e-06, "loss": 1.4728, "step": 39336 }, { "epoch": 3.0828369905956112, "grad_norm": 4.871546745300293, "learning_rate": 1.2467541048836606e-06, "loss": 1.2701, "step": 39337 }, { "epoch": 3.0829153605015676, "grad_norm": 2.3573129177093506, "learning_rate": 1.2465503769013043e-06, "loss": 1.4594, "step": 39338 }, { "epoch": 3.0829937304075234, "grad_norm": 2.157006025314331, "learning_rate": 1.246346663195141e-06, "loss": 1.5387, "step": 39339 }, { "epoch": 3.08307210031348, "grad_norm": 2.695176601409912, "learning_rate": 1.2461429637659466e-06, "loss": 1.0307, "step": 39340 }, { "epoch": 3.0831504702194357, "grad_norm": 2.8793463706970215, "learning_rate": 1.245939278614497e-06, "loss": 1.3934, "step": 39341 }, { "epoch": 3.083228840125392, "grad_norm": 3.5931832790374756, "learning_rate": 1.2457356077415656e-06, "loss": 1.5075, "step": 39342 }, { "epoch": 3.083307210031348, "grad_norm": 4.179096698760986, "learning_rate": 1.245531951147928e-06, "loss": 1.34, "step": 39343 }, { "epoch": 3.083385579937304, "grad_norm": 6.065669059753418, "learning_rate": 1.245328308834358e-06, "loss": 1.4403, "step": 39344 }, { "epoch": 3.08346394984326, "grad_norm": 4.149660587310791, "learning_rate": 1.245124680801631e-06, "loss": 1.3736, "step": 39345 }, { "epoch": 3.0835423197492164, "grad_norm": 2.552022933959961, "learning_rate": 1.2449210670505207e-06, "loss": 1.5213, "step": 39346 }, { "epoch": 3.0836206896551723, "grad_norm": 10.40302848815918, "learning_rate": 1.2447174675818024e-06, "loss": 2.0967, "step": 39347 }, { "epoch": 3.0836990595611287, "grad_norm": 2.713224411010742, "learning_rate": 1.24451388239625e-06, "loss": 1.4525, "step": 39348 }, { "epoch": 3.0837774294670846, "grad_norm": 2.826606035232544, "learning_rate": 1.2443103114946386e-06, "loss": 1.5112, "step": 39349 }, { "epoch": 3.083855799373041, "grad_norm": 2.394531011581421, "learning_rate": 1.2441067548777396e-06, "loss": 1.2758, "step": 39350 }, { "epoch": 3.0839341692789968, "grad_norm": 6.5438551902771, "learning_rate": 1.2439032125463308e-06, "loss": 1.1535, "step": 39351 }, { "epoch": 3.084012539184953, "grad_norm": 2.0537939071655273, "learning_rate": 1.2436996845011855e-06, "loss": 1.2096, "step": 39352 }, { "epoch": 3.084090909090909, "grad_norm": 2.572399377822876, "learning_rate": 1.2434961707430765e-06, "loss": 1.2854, "step": 39353 }, { "epoch": 3.0841692789968653, "grad_norm": 2.9348225593566895, "learning_rate": 1.243292671272779e-06, "loss": 1.2915, "step": 39354 }, { "epoch": 3.084247648902821, "grad_norm": 7.362522125244141, "learning_rate": 1.2430891860910643e-06, "loss": 0.3937, "step": 39355 }, { "epoch": 3.0843260188087775, "grad_norm": 2.852945327758789, "learning_rate": 1.2428857151987118e-06, "loss": 1.3048, "step": 39356 }, { "epoch": 3.0844043887147334, "grad_norm": 3.018353223800659, "learning_rate": 1.242682258596491e-06, "loss": 1.4301, "step": 39357 }, { "epoch": 3.0844827586206898, "grad_norm": 3.368823766708374, "learning_rate": 1.242478816285177e-06, "loss": 1.175, "step": 39358 }, { "epoch": 3.0845611285266457, "grad_norm": 3.38187313079834, "learning_rate": 1.2422753882655436e-06, "loss": 1.3214, "step": 39359 }, { "epoch": 3.084639498432602, "grad_norm": 2.543025016784668, "learning_rate": 1.2420719745383647e-06, "loss": 0.871, "step": 39360 }, { "epoch": 3.084717868338558, "grad_norm": 3.262434959411621, "learning_rate": 1.2418685751044135e-06, "loss": 0.8024, "step": 39361 }, { "epoch": 3.084796238244514, "grad_norm": 10.085968971252441, "learning_rate": 1.2416651899644638e-06, "loss": 1.113, "step": 39362 }, { "epoch": 3.08487460815047, "grad_norm": 4.064980506896973, "learning_rate": 1.2414618191192895e-06, "loss": 1.3786, "step": 39363 }, { "epoch": 3.0849529780564264, "grad_norm": 6.865512371063232, "learning_rate": 1.241258462569664e-06, "loss": 1.1407, "step": 39364 }, { "epoch": 3.0850313479623823, "grad_norm": 3.1416923999786377, "learning_rate": 1.2410551203163606e-06, "loss": 1.0367, "step": 39365 }, { "epoch": 3.0851097178683387, "grad_norm": 2.666438579559326, "learning_rate": 1.2408517923601532e-06, "loss": 1.0301, "step": 39366 }, { "epoch": 3.0851880877742945, "grad_norm": 8.093111991882324, "learning_rate": 1.2406484787018158e-06, "loss": 1.4809, "step": 39367 }, { "epoch": 3.085266457680251, "grad_norm": 2.5257911682128906, "learning_rate": 1.240445179342119e-06, "loss": 1.5855, "step": 39368 }, { "epoch": 3.0853448275862068, "grad_norm": 1.8781682252883911, "learning_rate": 1.2402418942818368e-06, "loss": 1.2064, "step": 39369 }, { "epoch": 3.085423197492163, "grad_norm": 3.5672500133514404, "learning_rate": 1.2400386235217444e-06, "loss": 1.2341, "step": 39370 }, { "epoch": 3.085501567398119, "grad_norm": 4.52379846572876, "learning_rate": 1.2398353670626157e-06, "loss": 1.2256, "step": 39371 }, { "epoch": 3.0855799373040753, "grad_norm": 5.694296836853027, "learning_rate": 1.2396321249052201e-06, "loss": 1.1091, "step": 39372 }, { "epoch": 3.085658307210031, "grad_norm": 13.92740249633789, "learning_rate": 1.2394288970503327e-06, "loss": 0.617, "step": 39373 }, { "epoch": 3.0857366771159875, "grad_norm": 4.961357593536377, "learning_rate": 1.2392256834987243e-06, "loss": 2.1166, "step": 39374 }, { "epoch": 3.0858150470219434, "grad_norm": 5.103566646575928, "learning_rate": 1.2390224842511727e-06, "loss": 1.3769, "step": 39375 }, { "epoch": 3.0858934169278998, "grad_norm": 5.392154693603516, "learning_rate": 1.238819299308447e-06, "loss": 1.3539, "step": 39376 }, { "epoch": 3.0859717868338556, "grad_norm": 4.425144672393799, "learning_rate": 1.2386161286713206e-06, "loss": 1.5314, "step": 39377 }, { "epoch": 3.086050156739812, "grad_norm": 3.1085543632507324, "learning_rate": 1.2384129723405658e-06, "loss": 1.6052, "step": 39378 }, { "epoch": 3.086128526645768, "grad_norm": 3.6579036712646484, "learning_rate": 1.2382098303169565e-06, "loss": 1.3271, "step": 39379 }, { "epoch": 3.086206896551724, "grad_norm": 6.982803821563721, "learning_rate": 1.2380067026012649e-06, "loss": 0.7632, "step": 39380 }, { "epoch": 3.08628526645768, "grad_norm": 2.40382719039917, "learning_rate": 1.2378035891942636e-06, "loss": 1.3643, "step": 39381 }, { "epoch": 3.0863636363636364, "grad_norm": 5.956123352050781, "learning_rate": 1.2376004900967264e-06, "loss": 1.5804, "step": 39382 }, { "epoch": 3.0864420062695923, "grad_norm": 2.7979044914245605, "learning_rate": 1.2373974053094211e-06, "loss": 1.241, "step": 39383 }, { "epoch": 3.0865203761755486, "grad_norm": 3.039458751678467, "learning_rate": 1.2371943348331255e-06, "loss": 1.2746, "step": 39384 }, { "epoch": 3.0865987460815045, "grad_norm": 3.606990098953247, "learning_rate": 1.23699127866861e-06, "loss": 1.4321, "step": 39385 }, { "epoch": 3.086677115987461, "grad_norm": 3.0198464393615723, "learning_rate": 1.2367882368166478e-06, "loss": 1.2358, "step": 39386 }, { "epoch": 3.0867554858934168, "grad_norm": 5.353627681732178, "learning_rate": 1.2365852092780089e-06, "loss": 1.1367, "step": 39387 }, { "epoch": 3.086833855799373, "grad_norm": 2.4714934825897217, "learning_rate": 1.2363821960534661e-06, "loss": 1.3341, "step": 39388 }, { "epoch": 3.086912225705329, "grad_norm": 2.492675304412842, "learning_rate": 1.236179197143793e-06, "loss": 1.1714, "step": 39389 }, { "epoch": 3.0869905956112853, "grad_norm": 3.4053359031677246, "learning_rate": 1.2359762125497627e-06, "loss": 1.358, "step": 39390 }, { "epoch": 3.087068965517241, "grad_norm": 4.963254451751709, "learning_rate": 1.235773242272144e-06, "loss": 1.216, "step": 39391 }, { "epoch": 3.0871473354231975, "grad_norm": 2.300079107284546, "learning_rate": 1.2355702863117107e-06, "loss": 1.2736, "step": 39392 }, { "epoch": 3.0872257053291534, "grad_norm": 2.862891435623169, "learning_rate": 1.2353673446692344e-06, "loss": 1.3496, "step": 39393 }, { "epoch": 3.0873040752351097, "grad_norm": 2.9281363487243652, "learning_rate": 1.2351644173454868e-06, "loss": 1.3605, "step": 39394 }, { "epoch": 3.0873824451410656, "grad_norm": 2.9538419246673584, "learning_rate": 1.2349615043412405e-06, "loss": 1.4922, "step": 39395 }, { "epoch": 3.087460815047022, "grad_norm": 3.0881357192993164, "learning_rate": 1.2347586056572663e-06, "loss": 1.1702, "step": 39396 }, { "epoch": 3.087539184952978, "grad_norm": 2.9261770248413086, "learning_rate": 1.2345557212943372e-06, "loss": 1.4759, "step": 39397 }, { "epoch": 3.087617554858934, "grad_norm": 2.7947161197662354, "learning_rate": 1.2343528512532232e-06, "loss": 1.5584, "step": 39398 }, { "epoch": 3.08769592476489, "grad_norm": 8.827898025512695, "learning_rate": 1.2341499955346975e-06, "loss": 1.6694, "step": 39399 }, { "epoch": 3.0877742946708464, "grad_norm": 3.084599733352661, "learning_rate": 1.2339471541395304e-06, "loss": 1.3742, "step": 39400 }, { "epoch": 3.0878526645768023, "grad_norm": 4.025038719177246, "learning_rate": 1.2337443270684956e-06, "loss": 1.5561, "step": 39401 }, { "epoch": 3.0879310344827586, "grad_norm": 2.7838313579559326, "learning_rate": 1.23354151432236e-06, "loss": 1.4617, "step": 39402 }, { "epoch": 3.088009404388715, "grad_norm": 5.093153476715088, "learning_rate": 1.2333387159018996e-06, "loss": 0.8005, "step": 39403 }, { "epoch": 3.088087774294671, "grad_norm": 3.4054415225982666, "learning_rate": 1.2331359318078834e-06, "loss": 1.3871, "step": 39404 }, { "epoch": 3.0881661442006267, "grad_norm": 4.110586166381836, "learning_rate": 1.2329331620410856e-06, "loss": 1.5858, "step": 39405 }, { "epoch": 3.088244514106583, "grad_norm": 18.39291000366211, "learning_rate": 1.2327304066022733e-06, "loss": 1.2056, "step": 39406 }, { "epoch": 3.0883228840125394, "grad_norm": 2.6933465003967285, "learning_rate": 1.232527665492218e-06, "loss": 1.2264, "step": 39407 }, { "epoch": 3.0884012539184953, "grad_norm": 3.1612789630889893, "learning_rate": 1.2323249387116954e-06, "loss": 1.3801, "step": 39408 }, { "epoch": 3.0884796238244516, "grad_norm": 3.0803110599517822, "learning_rate": 1.2321222262614718e-06, "loss": 1.3243, "step": 39409 }, { "epoch": 3.0885579937304075, "grad_norm": 4.210455417633057, "learning_rate": 1.23191952814232e-06, "loss": 1.3742, "step": 39410 }, { "epoch": 3.088636363636364, "grad_norm": 2.529334306716919, "learning_rate": 1.2317168443550115e-06, "loss": 0.7593, "step": 39411 }, { "epoch": 3.0887147335423197, "grad_norm": 4.565941333770752, "learning_rate": 1.2315141749003161e-06, "loss": 1.0006, "step": 39412 }, { "epoch": 3.088793103448276, "grad_norm": 3.3089165687561035, "learning_rate": 1.2313115197790053e-06, "loss": 1.3854, "step": 39413 }, { "epoch": 3.088871473354232, "grad_norm": 3.0012948513031006, "learning_rate": 1.2311088789918495e-06, "loss": 1.5979, "step": 39414 }, { "epoch": 3.0889498432601883, "grad_norm": 3.36114239692688, "learning_rate": 1.2309062525396203e-06, "loss": 1.5156, "step": 39415 }, { "epoch": 3.089028213166144, "grad_norm": 4.148387432098389, "learning_rate": 1.230703640423087e-06, "loss": 1.3036, "step": 39416 }, { "epoch": 3.0891065830721005, "grad_norm": 4.125736713409424, "learning_rate": 1.230501042643022e-06, "loss": 1.3984, "step": 39417 }, { "epoch": 3.0891849529780564, "grad_norm": 4.3196258544921875, "learning_rate": 1.2302984592001942e-06, "loss": 0.763, "step": 39418 }, { "epoch": 3.0892633228840127, "grad_norm": 3.291317939758301, "learning_rate": 1.2300958900953746e-06, "loss": 0.9231, "step": 39419 }, { "epoch": 3.0893416927899686, "grad_norm": 2.1162002086639404, "learning_rate": 1.2298933353293346e-06, "loss": 1.4999, "step": 39420 }, { "epoch": 3.089420062695925, "grad_norm": 2.6135551929473877, "learning_rate": 1.2296907949028436e-06, "loss": 1.4543, "step": 39421 }, { "epoch": 3.089498432601881, "grad_norm": 6.535680770874023, "learning_rate": 1.229488268816672e-06, "loss": 1.5077, "step": 39422 }, { "epoch": 3.089576802507837, "grad_norm": 7.324211120605469, "learning_rate": 1.2292857570715922e-06, "loss": 1.2232, "step": 39423 }, { "epoch": 3.089655172413793, "grad_norm": 2.616312026977539, "learning_rate": 1.229083259668371e-06, "loss": 1.2699, "step": 39424 }, { "epoch": 3.0897335423197494, "grad_norm": 3.9756295680999756, "learning_rate": 1.2288807766077804e-06, "loss": 1.4902, "step": 39425 }, { "epoch": 3.0898119122257053, "grad_norm": 2.569880723953247, "learning_rate": 1.2286783078905884e-06, "loss": 1.0399, "step": 39426 }, { "epoch": 3.0898902821316616, "grad_norm": 5.479255676269531, "learning_rate": 1.2284758535175705e-06, "loss": 1.16, "step": 39427 }, { "epoch": 3.0899686520376175, "grad_norm": 2.4280223846435547, "learning_rate": 1.2282734134894914e-06, "loss": 1.2268, "step": 39428 }, { "epoch": 3.090047021943574, "grad_norm": 8.230729103088379, "learning_rate": 1.2280709878071233e-06, "loss": 1.3615, "step": 39429 }, { "epoch": 3.0901253918495297, "grad_norm": 4.2384467124938965, "learning_rate": 1.2278685764712356e-06, "loss": 1.5842, "step": 39430 }, { "epoch": 3.090203761755486, "grad_norm": 2.9072179794311523, "learning_rate": 1.2276661794825984e-06, "loss": 1.1159, "step": 39431 }, { "epoch": 3.090282131661442, "grad_norm": 3.6019580364227295, "learning_rate": 1.227463796841981e-06, "loss": 1.4966, "step": 39432 }, { "epoch": 3.0903605015673983, "grad_norm": 3.1045377254486084, "learning_rate": 1.2272614285501544e-06, "loss": 1.3816, "step": 39433 }, { "epoch": 3.090438871473354, "grad_norm": 10.198248863220215, "learning_rate": 1.227059074607887e-06, "loss": 0.7881, "step": 39434 }, { "epoch": 3.0905172413793105, "grad_norm": 2.404609441757202, "learning_rate": 1.2268567350159493e-06, "loss": 1.1798, "step": 39435 }, { "epoch": 3.0905956112852664, "grad_norm": 3.6925079822540283, "learning_rate": 1.2266544097751105e-06, "loss": 1.5889, "step": 39436 }, { "epoch": 3.0906739811912227, "grad_norm": 3.9650094509124756, "learning_rate": 1.22645209888614e-06, "loss": 1.4795, "step": 39437 }, { "epoch": 3.0907523510971786, "grad_norm": 4.146900177001953, "learning_rate": 1.2262498023498092e-06, "loss": 1.3536, "step": 39438 }, { "epoch": 3.090830721003135, "grad_norm": 3.9138834476470947, "learning_rate": 1.2260475201668826e-06, "loss": 0.7558, "step": 39439 }, { "epoch": 3.090909090909091, "grad_norm": 9.03994369506836, "learning_rate": 1.2258452523381347e-06, "loss": 1.2485, "step": 39440 }, { "epoch": 3.090987460815047, "grad_norm": 4.9305853843688965, "learning_rate": 1.2256429988643326e-06, "loss": 1.4232, "step": 39441 }, { "epoch": 3.091065830721003, "grad_norm": 5.266137599945068, "learning_rate": 1.2254407597462476e-06, "loss": 1.026, "step": 39442 }, { "epoch": 3.0911442006269594, "grad_norm": 4.980712413787842, "learning_rate": 1.2252385349846457e-06, "loss": 1.0956, "step": 39443 }, { "epoch": 3.0912225705329153, "grad_norm": 3.7554314136505127, "learning_rate": 1.2250363245802959e-06, "loss": 1.3644, "step": 39444 }, { "epoch": 3.0913009404388716, "grad_norm": 2.8970813751220703, "learning_rate": 1.2248341285339705e-06, "loss": 0.9959, "step": 39445 }, { "epoch": 3.0913793103448275, "grad_norm": 4.357223033905029, "learning_rate": 1.2246319468464385e-06, "loss": 1.3092, "step": 39446 }, { "epoch": 3.091457680250784, "grad_norm": 2.3936853408813477, "learning_rate": 1.2244297795184661e-06, "loss": 1.1535, "step": 39447 }, { "epoch": 3.0915360501567397, "grad_norm": 2.9974277019500732, "learning_rate": 1.2242276265508235e-06, "loss": 0.852, "step": 39448 }, { "epoch": 3.091614420062696, "grad_norm": 4.2415618896484375, "learning_rate": 1.2240254879442798e-06, "loss": 1.4335, "step": 39449 }, { "epoch": 3.091692789968652, "grad_norm": 3.50553560256958, "learning_rate": 1.2238233636996032e-06, "loss": 1.2439, "step": 39450 }, { "epoch": 3.0917711598746083, "grad_norm": 3.872537851333618, "learning_rate": 1.2236212538175635e-06, "loss": 1.0802, "step": 39451 }, { "epoch": 3.091849529780564, "grad_norm": 3.17527437210083, "learning_rate": 1.2234191582989286e-06, "loss": 1.5134, "step": 39452 }, { "epoch": 3.0919278996865205, "grad_norm": 2.5572071075439453, "learning_rate": 1.2232170771444674e-06, "loss": 1.4056, "step": 39453 }, { "epoch": 3.0920062695924764, "grad_norm": 3.9553887844085693, "learning_rate": 1.2230150103549487e-06, "loss": 1.654, "step": 39454 }, { "epoch": 3.0920846394984327, "grad_norm": 3.828852653503418, "learning_rate": 1.222812957931141e-06, "loss": 1.2146, "step": 39455 }, { "epoch": 3.0921630094043886, "grad_norm": 3.230252981185913, "learning_rate": 1.2226109198738124e-06, "loss": 1.5595, "step": 39456 }, { "epoch": 3.092241379310345, "grad_norm": 5.319361686706543, "learning_rate": 1.2224088961837333e-06, "loss": 1.4192, "step": 39457 }, { "epoch": 3.092319749216301, "grad_norm": 2.9658191204071045, "learning_rate": 1.2222068868616677e-06, "loss": 0.9586, "step": 39458 }, { "epoch": 3.092398119122257, "grad_norm": 3.20432710647583, "learning_rate": 1.2220048919083882e-06, "loss": 1.4749, "step": 39459 }, { "epoch": 3.092476489028213, "grad_norm": 2.3303487300872803, "learning_rate": 1.2218029113246616e-06, "loss": 1.2938, "step": 39460 }, { "epoch": 3.0925548589341694, "grad_norm": 2.837493658065796, "learning_rate": 1.221600945111257e-06, "loss": 1.5459, "step": 39461 }, { "epoch": 3.0926332288401253, "grad_norm": 3.8679006099700928, "learning_rate": 1.2213989932689407e-06, "loss": 0.8949, "step": 39462 }, { "epoch": 3.0927115987460816, "grad_norm": 6.293816566467285, "learning_rate": 1.2211970557984805e-06, "loss": 1.4415, "step": 39463 }, { "epoch": 3.0927899686520375, "grad_norm": 2.799952268600464, "learning_rate": 1.2209951327006487e-06, "loss": 1.2473, "step": 39464 }, { "epoch": 3.092868338557994, "grad_norm": 5.709914207458496, "learning_rate": 1.2207932239762083e-06, "loss": 1.2935, "step": 39465 }, { "epoch": 3.0929467084639497, "grad_norm": 1.992883563041687, "learning_rate": 1.2205913296259298e-06, "loss": 1.1398, "step": 39466 }, { "epoch": 3.093025078369906, "grad_norm": 4.023117542266846, "learning_rate": 1.2203894496505808e-06, "loss": 1.1084, "step": 39467 }, { "epoch": 3.093103448275862, "grad_norm": 2.9992222785949707, "learning_rate": 1.2201875840509285e-06, "loss": 1.3512, "step": 39468 }, { "epoch": 3.0931818181818183, "grad_norm": 2.570887565612793, "learning_rate": 1.219985732827742e-06, "loss": 1.2631, "step": 39469 }, { "epoch": 3.093260188087774, "grad_norm": 2.727041721343994, "learning_rate": 1.2197838959817875e-06, "loss": 1.3265, "step": 39470 }, { "epoch": 3.0933385579937305, "grad_norm": 7.936000347137451, "learning_rate": 1.2195820735138337e-06, "loss": 1.7893, "step": 39471 }, { "epoch": 3.0934169278996864, "grad_norm": 2.362332344055176, "learning_rate": 1.2193802654246483e-06, "loss": 1.5515, "step": 39472 }, { "epoch": 3.0934952978056427, "grad_norm": 2.796302318572998, "learning_rate": 1.219178471714998e-06, "loss": 0.9248, "step": 39473 }, { "epoch": 3.0935736677115986, "grad_norm": 5.548139572143555, "learning_rate": 1.2189766923856512e-06, "loss": 1.4819, "step": 39474 }, { "epoch": 3.093652037617555, "grad_norm": 4.589635372161865, "learning_rate": 1.2187749274373744e-06, "loss": 1.1841, "step": 39475 }, { "epoch": 3.093730407523511, "grad_norm": 3.81459903717041, "learning_rate": 1.218573176870938e-06, "loss": 1.6346, "step": 39476 }, { "epoch": 3.093808777429467, "grad_norm": 3.6925690174102783, "learning_rate": 1.2183714406871034e-06, "loss": 1.5238, "step": 39477 }, { "epoch": 3.093887147335423, "grad_norm": 9.142646789550781, "learning_rate": 1.218169718886643e-06, "loss": 1.0522, "step": 39478 }, { "epoch": 3.0939655172413794, "grad_norm": 3.9009974002838135, "learning_rate": 1.217968011470324e-06, "loss": 1.3641, "step": 39479 }, { "epoch": 3.0940438871473352, "grad_norm": 3.6847939491271973, "learning_rate": 1.2177663184389105e-06, "loss": 1.6229, "step": 39480 }, { "epoch": 3.0941222570532916, "grad_norm": 6.460052013397217, "learning_rate": 1.2175646397931718e-06, "loss": 0.9463, "step": 39481 }, { "epoch": 3.0942006269592475, "grad_norm": 2.349583387374878, "learning_rate": 1.2173629755338722e-06, "loss": 1.2027, "step": 39482 }, { "epoch": 3.094278996865204, "grad_norm": 4.346763610839844, "learning_rate": 1.217161325661785e-06, "loss": 0.8548, "step": 39483 }, { "epoch": 3.0943573667711597, "grad_norm": 4.096101760864258, "learning_rate": 1.2169596901776703e-06, "loss": 0.9199, "step": 39484 }, { "epoch": 3.094435736677116, "grad_norm": 7.310217380523682, "learning_rate": 1.216758069082299e-06, "loss": 1.6042, "step": 39485 }, { "epoch": 3.094514106583072, "grad_norm": 3.0683391094207764, "learning_rate": 1.2165564623764358e-06, "loss": 1.0169, "step": 39486 }, { "epoch": 3.0945924764890282, "grad_norm": 5.34798002243042, "learning_rate": 1.2163548700608497e-06, "loss": 0.7749, "step": 39487 }, { "epoch": 3.094670846394984, "grad_norm": 2.4893479347229004, "learning_rate": 1.216153292136305e-06, "loss": 1.4591, "step": 39488 }, { "epoch": 3.0947492163009405, "grad_norm": 2.5308308601379395, "learning_rate": 1.2159517286035704e-06, "loss": 1.159, "step": 39489 }, { "epoch": 3.0948275862068964, "grad_norm": 3.4188952445983887, "learning_rate": 1.2157501794634118e-06, "loss": 1.0395, "step": 39490 }, { "epoch": 3.0949059561128527, "grad_norm": 2.80135440826416, "learning_rate": 1.215548644716596e-06, "loss": 1.4295, "step": 39491 }, { "epoch": 3.0949843260188086, "grad_norm": 3.0774407386779785, "learning_rate": 1.2153471243638886e-06, "loss": 1.306, "step": 39492 }, { "epoch": 3.095062695924765, "grad_norm": 6.263967037200928, "learning_rate": 1.2151456184060573e-06, "loss": 0.7625, "step": 39493 }, { "epoch": 3.095141065830721, "grad_norm": 4.588445663452148, "learning_rate": 1.2149441268438695e-06, "loss": 1.2559, "step": 39494 }, { "epoch": 3.095219435736677, "grad_norm": 2.798243284225464, "learning_rate": 1.214742649678089e-06, "loss": 1.4368, "step": 39495 }, { "epoch": 3.095297805642633, "grad_norm": 4.139716625213623, "learning_rate": 1.214541186909481e-06, "loss": 1.4969, "step": 39496 }, { "epoch": 3.0953761755485893, "grad_norm": 2.653777599334717, "learning_rate": 1.2143397385388162e-06, "loss": 1.4212, "step": 39497 }, { "epoch": 3.0954545454545457, "grad_norm": 7.586429119110107, "learning_rate": 1.2141383045668592e-06, "loss": 1.8945, "step": 39498 }, { "epoch": 3.0955329153605016, "grad_norm": 2.633967161178589, "learning_rate": 1.213936884994375e-06, "loss": 1.1272, "step": 39499 }, { "epoch": 3.0956112852664575, "grad_norm": 4.068081378936768, "learning_rate": 1.2137354798221295e-06, "loss": 1.1662, "step": 39500 }, { "epoch": 3.095689655172414, "grad_norm": 2.5020599365234375, "learning_rate": 1.2135340890508885e-06, "loss": 1.3356, "step": 39501 }, { "epoch": 3.09576802507837, "grad_norm": 3.0953164100646973, "learning_rate": 1.2133327126814215e-06, "loss": 1.3164, "step": 39502 }, { "epoch": 3.095846394984326, "grad_norm": 2.5056302547454834, "learning_rate": 1.213131350714491e-06, "loss": 1.1403, "step": 39503 }, { "epoch": 3.0959247648902823, "grad_norm": 2.3271524906158447, "learning_rate": 1.2129300031508633e-06, "loss": 1.5241, "step": 39504 }, { "epoch": 3.0960031347962382, "grad_norm": 5.976475238800049, "learning_rate": 1.2127286699913049e-06, "loss": 1.9291, "step": 39505 }, { "epoch": 3.0960815047021946, "grad_norm": 10.318315505981445, "learning_rate": 1.2125273512365814e-06, "loss": 1.0826, "step": 39506 }, { "epoch": 3.0961598746081505, "grad_norm": 3.415213108062744, "learning_rate": 1.2123260468874587e-06, "loss": 1.1615, "step": 39507 }, { "epoch": 3.096238244514107, "grad_norm": 7.385183334350586, "learning_rate": 1.212124756944702e-06, "loss": 0.9112, "step": 39508 }, { "epoch": 3.0963166144200627, "grad_norm": 2.8105382919311523, "learning_rate": 1.2119234814090785e-06, "loss": 1.3357, "step": 39509 }, { "epoch": 3.096394984326019, "grad_norm": 4.932544231414795, "learning_rate": 1.2117222202813495e-06, "loss": 1.1429, "step": 39510 }, { "epoch": 3.096473354231975, "grad_norm": 2.818641185760498, "learning_rate": 1.2115209735622852e-06, "loss": 1.0703, "step": 39511 }, { "epoch": 3.0965517241379312, "grad_norm": 4.991213798522949, "learning_rate": 1.2113197412526484e-06, "loss": 1.3439, "step": 39512 }, { "epoch": 3.096630094043887, "grad_norm": 2.9122629165649414, "learning_rate": 1.2111185233532075e-06, "loss": 1.3366, "step": 39513 }, { "epoch": 3.0967084639498434, "grad_norm": 2.9093706607818604, "learning_rate": 1.210917319864724e-06, "loss": 1.2741, "step": 39514 }, { "epoch": 3.0967868338557993, "grad_norm": 3.9643218517303467, "learning_rate": 1.2107161307879634e-06, "loss": 1.5304, "step": 39515 }, { "epoch": 3.0968652037617557, "grad_norm": 2.555148124694824, "learning_rate": 1.210514956123694e-06, "loss": 1.1906, "step": 39516 }, { "epoch": 3.0969435736677116, "grad_norm": 3.503011465072632, "learning_rate": 1.21031379587268e-06, "loss": 1.5605, "step": 39517 }, { "epoch": 3.097021943573668, "grad_norm": 3.243062973022461, "learning_rate": 1.2101126500356853e-06, "loss": 1.3773, "step": 39518 }, { "epoch": 3.097100313479624, "grad_norm": 6.278973579406738, "learning_rate": 1.2099115186134752e-06, "loss": 1.741, "step": 39519 }, { "epoch": 3.09717868338558, "grad_norm": 2.639953851699829, "learning_rate": 1.2097104016068146e-06, "loss": 1.3516, "step": 39520 }, { "epoch": 3.097257053291536, "grad_norm": 2.5292012691497803, "learning_rate": 1.2095092990164693e-06, "loss": 1.445, "step": 39521 }, { "epoch": 3.0973354231974923, "grad_norm": 11.882704734802246, "learning_rate": 1.209308210843203e-06, "loss": 1.384, "step": 39522 }, { "epoch": 3.097413793103448, "grad_norm": 2.756904363632202, "learning_rate": 1.2091071370877822e-06, "loss": 1.4558, "step": 39523 }, { "epoch": 3.0974921630094046, "grad_norm": 5.279545783996582, "learning_rate": 1.2089060777509698e-06, "loss": 1.3677, "step": 39524 }, { "epoch": 3.0975705329153604, "grad_norm": 2.45678973197937, "learning_rate": 1.208705032833532e-06, "loss": 1.3832, "step": 39525 }, { "epoch": 3.0976489028213168, "grad_norm": 4.2983503341674805, "learning_rate": 1.2085040023362328e-06, "loss": 1.6451, "step": 39526 }, { "epoch": 3.0977272727272727, "grad_norm": 3.2953572273254395, "learning_rate": 1.2083029862598373e-06, "loss": 1.3749, "step": 39527 }, { "epoch": 3.097805642633229, "grad_norm": 3.039661169052124, "learning_rate": 1.208101984605109e-06, "loss": 1.2246, "step": 39528 }, { "epoch": 3.097884012539185, "grad_norm": 3.0616626739501953, "learning_rate": 1.2079009973728133e-06, "loss": 1.2271, "step": 39529 }, { "epoch": 3.097962382445141, "grad_norm": 2.3416695594787598, "learning_rate": 1.207700024563715e-06, "loss": 1.347, "step": 39530 }, { "epoch": 3.098040752351097, "grad_norm": 2.8481266498565674, "learning_rate": 1.2074990661785773e-06, "loss": 1.3919, "step": 39531 }, { "epoch": 3.0981191222570534, "grad_norm": 3.816601037979126, "learning_rate": 1.2072981222181667e-06, "loss": 1.1108, "step": 39532 }, { "epoch": 3.0981974921630093, "grad_norm": 9.130375862121582, "learning_rate": 1.2070971926832437e-06, "loss": 1.5474, "step": 39533 }, { "epoch": 3.0982758620689657, "grad_norm": 2.507215976715088, "learning_rate": 1.2068962775745757e-06, "loss": 1.6373, "step": 39534 }, { "epoch": 3.0983542319749215, "grad_norm": 14.463350296020508, "learning_rate": 1.206695376892928e-06, "loss": 1.3417, "step": 39535 }, { "epoch": 3.098432601880878, "grad_norm": 2.385521411895752, "learning_rate": 1.2064944906390607e-06, "loss": 1.2845, "step": 39536 }, { "epoch": 3.0985109717868338, "grad_norm": 10.794769287109375, "learning_rate": 1.2062936188137398e-06, "loss": 1.9191, "step": 39537 }, { "epoch": 3.09858934169279, "grad_norm": 4.405050754547119, "learning_rate": 1.2060927614177288e-06, "loss": 1.5202, "step": 39538 }, { "epoch": 3.098667711598746, "grad_norm": 2.9377851486206055, "learning_rate": 1.2058919184517941e-06, "loss": 1.1964, "step": 39539 }, { "epoch": 3.0987460815047023, "grad_norm": 3.4251344203948975, "learning_rate": 1.2056910899166969e-06, "loss": 1.4338, "step": 39540 }, { "epoch": 3.098824451410658, "grad_norm": 3.6297223567962646, "learning_rate": 1.205490275813202e-06, "loss": 1.5246, "step": 39541 }, { "epoch": 3.0989028213166145, "grad_norm": 2.772697925567627, "learning_rate": 1.2052894761420731e-06, "loss": 1.4605, "step": 39542 }, { "epoch": 3.0989811912225704, "grad_norm": 18.422039031982422, "learning_rate": 1.2050886909040738e-06, "loss": 1.587, "step": 39543 }, { "epoch": 3.0990595611285268, "grad_norm": 2.6513912677764893, "learning_rate": 1.2048879200999674e-06, "loss": 1.3113, "step": 39544 }, { "epoch": 3.0991379310344827, "grad_norm": 2.8059465885162354, "learning_rate": 1.2046871637305191e-06, "loss": 1.2733, "step": 39545 }, { "epoch": 3.099216300940439, "grad_norm": 11.911735534667969, "learning_rate": 1.2044864217964908e-06, "loss": 1.8526, "step": 39546 }, { "epoch": 3.099294670846395, "grad_norm": 2.5747461318969727, "learning_rate": 1.2042856942986465e-06, "loss": 1.0716, "step": 39547 }, { "epoch": 3.099373040752351, "grad_norm": 2.8970940113067627, "learning_rate": 1.2040849812377504e-06, "loss": 0.9354, "step": 39548 }, { "epoch": 3.099451410658307, "grad_norm": 4.86224365234375, "learning_rate": 1.2038842826145647e-06, "loss": 1.5173, "step": 39549 }, { "epoch": 3.0995297805642634, "grad_norm": 2.358381748199463, "learning_rate": 1.203683598429855e-06, "loss": 1.3395, "step": 39550 }, { "epoch": 3.0996081504702193, "grad_norm": 2.937842607498169, "learning_rate": 1.203482928684382e-06, "loss": 1.1265, "step": 39551 }, { "epoch": 3.0996865203761756, "grad_norm": 3.2848117351531982, "learning_rate": 1.2032822733789079e-06, "loss": 1.3329, "step": 39552 }, { "epoch": 3.0997648902821315, "grad_norm": 5.909494400024414, "learning_rate": 1.2030816325141997e-06, "loss": 0.9764, "step": 39553 }, { "epoch": 3.099843260188088, "grad_norm": 5.899084091186523, "learning_rate": 1.2028810060910201e-06, "loss": 1.3195, "step": 39554 }, { "epoch": 3.0999216300940438, "grad_norm": 2.7821271419525146, "learning_rate": 1.2026803941101295e-06, "loss": 1.2262, "step": 39555 }, { "epoch": 3.1, "grad_norm": 2.801602602005005, "learning_rate": 1.2024797965722922e-06, "loss": 1.2873, "step": 39556 }, { "epoch": 3.100078369905956, "grad_norm": 4.153244972229004, "learning_rate": 1.2022792134782695e-06, "loss": 2.0711, "step": 39557 }, { "epoch": 3.1001567398119123, "grad_norm": 3.376445770263672, "learning_rate": 1.202078644828829e-06, "loss": 1.9141, "step": 39558 }, { "epoch": 3.100235109717868, "grad_norm": 3.518017053604126, "learning_rate": 1.2018780906247296e-06, "loss": 1.3911, "step": 39559 }, { "epoch": 3.1003134796238245, "grad_norm": 2.741809368133545, "learning_rate": 1.2016775508667349e-06, "loss": 1.3963, "step": 39560 }, { "epoch": 3.1003918495297804, "grad_norm": 3.7825324535369873, "learning_rate": 1.2014770255556075e-06, "loss": 1.4135, "step": 39561 }, { "epoch": 3.1004702194357368, "grad_norm": 5.356918811798096, "learning_rate": 1.2012765146921108e-06, "loss": 1.3421, "step": 39562 }, { "epoch": 3.1005485893416926, "grad_norm": 3.329328775405884, "learning_rate": 1.201076018277007e-06, "loss": 1.1772, "step": 39563 }, { "epoch": 3.100626959247649, "grad_norm": 3.263906478881836, "learning_rate": 1.2008755363110587e-06, "loss": 1.5208, "step": 39564 }, { "epoch": 3.100705329153605, "grad_norm": 3.459428310394287, "learning_rate": 1.2006750687950302e-06, "loss": 1.6908, "step": 39565 }, { "epoch": 3.100783699059561, "grad_norm": 2.3793911933898926, "learning_rate": 1.2004746157296792e-06, "loss": 1.2947, "step": 39566 }, { "epoch": 3.100862068965517, "grad_norm": 3.5781309604644775, "learning_rate": 1.2002741771157727e-06, "loss": 0.7334, "step": 39567 }, { "epoch": 3.1009404388714734, "grad_norm": 2.404907703399658, "learning_rate": 1.2000737529540713e-06, "loss": 1.3275, "step": 39568 }, { "epoch": 3.1010188087774293, "grad_norm": 5.122693061828613, "learning_rate": 1.1998733432453396e-06, "loss": 1.6768, "step": 39569 }, { "epoch": 3.1010971786833856, "grad_norm": 3.359632968902588, "learning_rate": 1.1996729479903358e-06, "loss": 1.5892, "step": 39570 }, { "epoch": 3.1011755485893415, "grad_norm": 9.19925594329834, "learning_rate": 1.1994725671898233e-06, "loss": 1.8096, "step": 39571 }, { "epoch": 3.101253918495298, "grad_norm": 4.183670520782471, "learning_rate": 1.1992722008445656e-06, "loss": 1.8, "step": 39572 }, { "epoch": 3.1013322884012537, "grad_norm": 13.146258354187012, "learning_rate": 1.1990718489553266e-06, "loss": 1.0587, "step": 39573 }, { "epoch": 3.10141065830721, "grad_norm": 2.868784189224243, "learning_rate": 1.1988715115228645e-06, "loss": 1.1924, "step": 39574 }, { "epoch": 3.101489028213166, "grad_norm": 18.511762619018555, "learning_rate": 1.1986711885479424e-06, "loss": 1.3798, "step": 39575 }, { "epoch": 3.1015673981191223, "grad_norm": 2.536494016647339, "learning_rate": 1.1984708800313227e-06, "loss": 1.4175, "step": 39576 }, { "epoch": 3.101645768025078, "grad_norm": 2.8329131603240967, "learning_rate": 1.198270585973767e-06, "loss": 1.5059, "step": 39577 }, { "epoch": 3.1017241379310345, "grad_norm": 5.706270217895508, "learning_rate": 1.1980703063760374e-06, "loss": 1.2144, "step": 39578 }, { "epoch": 3.1018025078369904, "grad_norm": 2.5998828411102295, "learning_rate": 1.197870041238896e-06, "loss": 1.5072, "step": 39579 }, { "epoch": 3.1018808777429467, "grad_norm": 3.6657254695892334, "learning_rate": 1.1976697905631036e-06, "loss": 1.1119, "step": 39580 }, { "epoch": 3.1019592476489026, "grad_norm": 2.897138833999634, "learning_rate": 1.1974695543494225e-06, "loss": 1.454, "step": 39581 }, { "epoch": 3.102037617554859, "grad_norm": 4.956265449523926, "learning_rate": 1.197269332598614e-06, "loss": 1.3937, "step": 39582 }, { "epoch": 3.102115987460815, "grad_norm": 4.5871076583862305, "learning_rate": 1.1970691253114397e-06, "loss": 0.7362, "step": 39583 }, { "epoch": 3.102194357366771, "grad_norm": 2.946131706237793, "learning_rate": 1.196868932488663e-06, "loss": 1.4799, "step": 39584 }, { "epoch": 3.102272727272727, "grad_norm": 3.6745266914367676, "learning_rate": 1.19666875413104e-06, "loss": 1.1141, "step": 39585 }, { "epoch": 3.1023510971786834, "grad_norm": 2.5499863624572754, "learning_rate": 1.1964685902393375e-06, "loss": 0.8927, "step": 39586 }, { "epoch": 3.1024294670846393, "grad_norm": 6.22837495803833, "learning_rate": 1.196268440814315e-06, "loss": 1.3597, "step": 39587 }, { "epoch": 3.1025078369905956, "grad_norm": 7.594659328460693, "learning_rate": 1.1960683058567346e-06, "loss": 0.9292, "step": 39588 }, { "epoch": 3.1025862068965515, "grad_norm": 3.515279769897461, "learning_rate": 1.1958681853673548e-06, "loss": 1.3841, "step": 39589 }, { "epoch": 3.102664576802508, "grad_norm": 2.391390085220337, "learning_rate": 1.1956680793469377e-06, "loss": 0.8835, "step": 39590 }, { "epoch": 3.1027429467084637, "grad_norm": 2.407763957977295, "learning_rate": 1.1954679877962477e-06, "loss": 1.1888, "step": 39591 }, { "epoch": 3.10282131661442, "grad_norm": 3.349998712539673, "learning_rate": 1.195267910716042e-06, "loss": 0.7699, "step": 39592 }, { "epoch": 3.1028996865203764, "grad_norm": 2.832357168197632, "learning_rate": 1.1950678481070832e-06, "loss": 1.2855, "step": 39593 }, { "epoch": 3.1029780564263323, "grad_norm": 3.2548396587371826, "learning_rate": 1.194867799970132e-06, "loss": 1.2518, "step": 39594 }, { "epoch": 3.103056426332288, "grad_norm": 2.118093490600586, "learning_rate": 1.1946677663059487e-06, "loss": 1.2236, "step": 39595 }, { "epoch": 3.1031347962382445, "grad_norm": 3.292874336242676, "learning_rate": 1.194467747115295e-06, "loss": 1.3396, "step": 39596 }, { "epoch": 3.103213166144201, "grad_norm": 2.33026123046875, "learning_rate": 1.1942677423989314e-06, "loss": 1.1868, "step": 39597 }, { "epoch": 3.1032915360501567, "grad_norm": 2.4806904792785645, "learning_rate": 1.1940677521576183e-06, "loss": 1.3632, "step": 39598 }, { "epoch": 3.103369905956113, "grad_norm": 3.706779956817627, "learning_rate": 1.1938677763921169e-06, "loss": 1.6272, "step": 39599 }, { "epoch": 3.103448275862069, "grad_norm": 3.0904083251953125, "learning_rate": 1.1936678151031878e-06, "loss": 1.3437, "step": 39600 }, { "epoch": 3.1035266457680253, "grad_norm": 2.787592887878418, "learning_rate": 1.1934678682915907e-06, "loss": 1.3127, "step": 39601 }, { "epoch": 3.103605015673981, "grad_norm": 4.133814811706543, "learning_rate": 1.1932679359580867e-06, "loss": 1.5638, "step": 39602 }, { "epoch": 3.1036833855799375, "grad_norm": 2.659252643585205, "learning_rate": 1.1930680181034383e-06, "loss": 1.2688, "step": 39603 }, { "epoch": 3.1037617554858934, "grad_norm": 6.028817653656006, "learning_rate": 1.1928681147283998e-06, "loss": 1.4745, "step": 39604 }, { "epoch": 3.1038401253918497, "grad_norm": 5.096016883850098, "learning_rate": 1.192668225833738e-06, "loss": 1.1528, "step": 39605 }, { "epoch": 3.1039184952978056, "grad_norm": 3.504002571105957, "learning_rate": 1.1924683514202119e-06, "loss": 0.8762, "step": 39606 }, { "epoch": 3.103996865203762, "grad_norm": 2.9526612758636475, "learning_rate": 1.1922684914885791e-06, "loss": 1.6605, "step": 39607 }, { "epoch": 3.104075235109718, "grad_norm": 5.175381183624268, "learning_rate": 1.1920686460396009e-06, "loss": 1.5277, "step": 39608 }, { "epoch": 3.104153605015674, "grad_norm": 5.426723480224609, "learning_rate": 1.191868815074036e-06, "loss": 1.6896, "step": 39609 }, { "epoch": 3.10423197492163, "grad_norm": 5.878785610198975, "learning_rate": 1.1916689985926494e-06, "loss": 1.354, "step": 39610 }, { "epoch": 3.1043103448275864, "grad_norm": 3.007474422454834, "learning_rate": 1.1914691965961966e-06, "loss": 1.666, "step": 39611 }, { "epoch": 3.1043887147335423, "grad_norm": 2.555169105529785, "learning_rate": 1.1912694090854388e-06, "loss": 1.2865, "step": 39612 }, { "epoch": 3.1044670846394986, "grad_norm": 4.621379852294922, "learning_rate": 1.1910696360611358e-06, "loss": 1.6202, "step": 39613 }, { "epoch": 3.1045454545454545, "grad_norm": 4.527894973754883, "learning_rate": 1.1908698775240468e-06, "loss": 1.3258, "step": 39614 }, { "epoch": 3.104623824451411, "grad_norm": 5.367187976837158, "learning_rate": 1.1906701334749331e-06, "loss": 1.2283, "step": 39615 }, { "epoch": 3.1047021943573667, "grad_norm": 8.686735153198242, "learning_rate": 1.1904704039145537e-06, "loss": 1.0015, "step": 39616 }, { "epoch": 3.104780564263323, "grad_norm": 5.627355098724365, "learning_rate": 1.1902706888436677e-06, "loss": 1.2649, "step": 39617 }, { "epoch": 3.104858934169279, "grad_norm": 3.204343795776367, "learning_rate": 1.1900709882630351e-06, "loss": 1.4685, "step": 39618 }, { "epoch": 3.1049373040752353, "grad_norm": 2.4090144634246826, "learning_rate": 1.189871302173416e-06, "loss": 1.3239, "step": 39619 }, { "epoch": 3.105015673981191, "grad_norm": 3.8858439922332764, "learning_rate": 1.1896716305755695e-06, "loss": 1.4213, "step": 39620 }, { "epoch": 3.1050940438871475, "grad_norm": 3.811218738555908, "learning_rate": 1.1894719734702558e-06, "loss": 1.2302, "step": 39621 }, { "epoch": 3.1051724137931034, "grad_norm": 4.718319892883301, "learning_rate": 1.1892723308582326e-06, "loss": 1.328, "step": 39622 }, { "epoch": 3.1052507836990597, "grad_norm": 2.932990312576294, "learning_rate": 1.1890727027402588e-06, "loss": 1.5249, "step": 39623 }, { "epoch": 3.1053291536050156, "grad_norm": 5.258087158203125, "learning_rate": 1.1888730891170962e-06, "loss": 0.9094, "step": 39624 }, { "epoch": 3.105407523510972, "grad_norm": 4.16802978515625, "learning_rate": 1.1886734899895048e-06, "loss": 1.6553, "step": 39625 }, { "epoch": 3.105485893416928, "grad_norm": 2.6386375427246094, "learning_rate": 1.1884739053582395e-06, "loss": 1.4194, "step": 39626 }, { "epoch": 3.105564263322884, "grad_norm": 6.2881269454956055, "learning_rate": 1.1882743352240621e-06, "loss": 1.357, "step": 39627 }, { "epoch": 3.10564263322884, "grad_norm": 1.93186354637146, "learning_rate": 1.1880747795877301e-06, "loss": 1.6684, "step": 39628 }, { "epoch": 3.1057210031347964, "grad_norm": 4.853786945343018, "learning_rate": 1.1878752384500064e-06, "loss": 1.1359, "step": 39629 }, { "epoch": 3.1057993730407523, "grad_norm": 3.576464891433716, "learning_rate": 1.1876757118116462e-06, "loss": 1.2757, "step": 39630 }, { "epoch": 3.1058777429467086, "grad_norm": 2.7000083923339844, "learning_rate": 1.187476199673409e-06, "loss": 1.3961, "step": 39631 }, { "epoch": 3.1059561128526645, "grad_norm": 2.895728349685669, "learning_rate": 1.1872767020360542e-06, "loss": 1.5837, "step": 39632 }, { "epoch": 3.106034482758621, "grad_norm": 11.721488952636719, "learning_rate": 1.1870772189003405e-06, "loss": 1.6471, "step": 39633 }, { "epoch": 3.1061128526645767, "grad_norm": 4.151122570037842, "learning_rate": 1.1868777502670265e-06, "loss": 1.3072, "step": 39634 }, { "epoch": 3.106191222570533, "grad_norm": 8.428642272949219, "learning_rate": 1.186678296136871e-06, "loss": 1.1221, "step": 39635 }, { "epoch": 3.106269592476489, "grad_norm": 5.78267765045166, "learning_rate": 1.1864788565106327e-06, "loss": 1.5518, "step": 39636 }, { "epoch": 3.1063479623824453, "grad_norm": 3.8226122856140137, "learning_rate": 1.1862794313890702e-06, "loss": 0.8964, "step": 39637 }, { "epoch": 3.106426332288401, "grad_norm": 5.446777820587158, "learning_rate": 1.1860800207729422e-06, "loss": 1.2209, "step": 39638 }, { "epoch": 3.1065047021943575, "grad_norm": 9.051793098449707, "learning_rate": 1.1858806246630061e-06, "loss": 1.0335, "step": 39639 }, { "epoch": 3.1065830721003134, "grad_norm": 3.777419090270996, "learning_rate": 1.1856812430600228e-06, "loss": 1.3913, "step": 39640 }, { "epoch": 3.1066614420062697, "grad_norm": 11.513535499572754, "learning_rate": 1.1854818759647463e-06, "loss": 1.6782, "step": 39641 }, { "epoch": 3.1067398119122256, "grad_norm": 3.7370824813842773, "learning_rate": 1.1852825233779386e-06, "loss": 1.4183, "step": 39642 }, { "epoch": 3.106818181818182, "grad_norm": 6.811735153198242, "learning_rate": 1.185083185300357e-06, "loss": 1.0182, "step": 39643 }, { "epoch": 3.106896551724138, "grad_norm": 3.293823480606079, "learning_rate": 1.1848838617327607e-06, "loss": 1.4162, "step": 39644 }, { "epoch": 3.106974921630094, "grad_norm": 3.4550278186798096, "learning_rate": 1.1846845526759053e-06, "loss": 1.0563, "step": 39645 }, { "epoch": 3.10705329153605, "grad_norm": 3.4665794372558594, "learning_rate": 1.1844852581305489e-06, "loss": 0.9175, "step": 39646 }, { "epoch": 3.1071316614420064, "grad_norm": 4.324298858642578, "learning_rate": 1.1842859780974537e-06, "loss": 1.1864, "step": 39647 }, { "epoch": 3.1072100313479623, "grad_norm": 2.9379003047943115, "learning_rate": 1.1840867125773731e-06, "loss": 1.3785, "step": 39648 }, { "epoch": 3.1072884012539186, "grad_norm": 4.681837558746338, "learning_rate": 1.1838874615710666e-06, "loss": 1.2687, "step": 39649 }, { "epoch": 3.1073667711598745, "grad_norm": 2.760129451751709, "learning_rate": 1.1836882250792924e-06, "loss": 1.4644, "step": 39650 }, { "epoch": 3.107445141065831, "grad_norm": 9.888917922973633, "learning_rate": 1.183489003102808e-06, "loss": 0.8186, "step": 39651 }, { "epoch": 3.1075235109717867, "grad_norm": 2.4484522342681885, "learning_rate": 1.183289795642371e-06, "loss": 1.1667, "step": 39652 }, { "epoch": 3.107601880877743, "grad_norm": 4.590292453765869, "learning_rate": 1.1830906026987392e-06, "loss": 1.3724, "step": 39653 }, { "epoch": 3.107680250783699, "grad_norm": 3.1958467960357666, "learning_rate": 1.1828914242726702e-06, "loss": 1.4388, "step": 39654 }, { "epoch": 3.1077586206896552, "grad_norm": 2.0863325595855713, "learning_rate": 1.1826922603649221e-06, "loss": 1.2468, "step": 39655 }, { "epoch": 3.107836990595611, "grad_norm": 2.1037094593048096, "learning_rate": 1.1824931109762515e-06, "loss": 1.4246, "step": 39656 }, { "epoch": 3.1079153605015675, "grad_norm": 4.241610527038574, "learning_rate": 1.1822939761074165e-06, "loss": 1.2401, "step": 39657 }, { "epoch": 3.1079937304075234, "grad_norm": 3.96533203125, "learning_rate": 1.1820948557591744e-06, "loss": 1.5484, "step": 39658 }, { "epoch": 3.1080721003134797, "grad_norm": 3.597019672393799, "learning_rate": 1.1818957499322836e-06, "loss": 1.3738, "step": 39659 }, { "epoch": 3.1081504702194356, "grad_norm": 4.006280899047852, "learning_rate": 1.1816966586274975e-06, "loss": 1.1667, "step": 39660 }, { "epoch": 3.108228840125392, "grad_norm": 3.873833656311035, "learning_rate": 1.1814975818455776e-06, "loss": 1.4297, "step": 39661 }, { "epoch": 3.108307210031348, "grad_norm": 4.181265354156494, "learning_rate": 1.1812985195872812e-06, "loss": 1.3521, "step": 39662 }, { "epoch": 3.108385579937304, "grad_norm": 5.733443260192871, "learning_rate": 1.1810994718533619e-06, "loss": 0.7183, "step": 39663 }, { "epoch": 3.10846394984326, "grad_norm": 4.54868745803833, "learning_rate": 1.1809004386445788e-06, "loss": 1.2864, "step": 39664 }, { "epoch": 3.1085423197492164, "grad_norm": 2.82002592086792, "learning_rate": 1.180701419961688e-06, "loss": 1.4373, "step": 39665 }, { "epoch": 3.1086206896551722, "grad_norm": 4.1673502922058105, "learning_rate": 1.1805024158054495e-06, "loss": 0.943, "step": 39666 }, { "epoch": 3.1086990595611286, "grad_norm": 2.52877140045166, "learning_rate": 1.1803034261766166e-06, "loss": 1.2134, "step": 39667 }, { "epoch": 3.1087774294670845, "grad_norm": 4.417820453643799, "learning_rate": 1.180104451075948e-06, "loss": 1.3662, "step": 39668 }, { "epoch": 3.108855799373041, "grad_norm": 2.133538007736206, "learning_rate": 1.1799054905041995e-06, "loss": 1.2781, "step": 39669 }, { "epoch": 3.1089341692789967, "grad_norm": 2.924281597137451, "learning_rate": 1.1797065444621286e-06, "loss": 1.3803, "step": 39670 }, { "epoch": 3.109012539184953, "grad_norm": 3.3470492362976074, "learning_rate": 1.1795076129504919e-06, "loss": 1.1127, "step": 39671 }, { "epoch": 3.109090909090909, "grad_norm": 3.6252291202545166, "learning_rate": 1.1793086959700462e-06, "loss": 0.769, "step": 39672 }, { "epoch": 3.1091692789968652, "grad_norm": 2.3828938007354736, "learning_rate": 1.1791097935215484e-06, "loss": 1.1642, "step": 39673 }, { "epoch": 3.109247648902821, "grad_norm": 3.511080741882324, "learning_rate": 1.1789109056057523e-06, "loss": 1.5897, "step": 39674 }, { "epoch": 3.1093260188087775, "grad_norm": 2.3244338035583496, "learning_rate": 1.1787120322234175e-06, "loss": 1.2587, "step": 39675 }, { "epoch": 3.1094043887147333, "grad_norm": 3.232142925262451, "learning_rate": 1.1785131733752996e-06, "loss": 1.4226, "step": 39676 }, { "epoch": 3.1094827586206897, "grad_norm": 3.987990379333496, "learning_rate": 1.1783143290621563e-06, "loss": 1.1379, "step": 39677 }, { "epoch": 3.1095611285266456, "grad_norm": 2.327578067779541, "learning_rate": 1.1781154992847405e-06, "loss": 1.3186, "step": 39678 }, { "epoch": 3.109639498432602, "grad_norm": 4.112969398498535, "learning_rate": 1.177916684043809e-06, "loss": 1.5782, "step": 39679 }, { "epoch": 3.109717868338558, "grad_norm": 2.3982837200164795, "learning_rate": 1.177717883340121e-06, "loss": 1.4896, "step": 39680 }, { "epoch": 3.109796238244514, "grad_norm": 3.5273995399475098, "learning_rate": 1.1775190971744321e-06, "loss": 1.4513, "step": 39681 }, { "epoch": 3.10987460815047, "grad_norm": 6.762207508087158, "learning_rate": 1.1773203255474957e-06, "loss": 0.9447, "step": 39682 }, { "epoch": 3.1099529780564263, "grad_norm": 2.6705474853515625, "learning_rate": 1.1771215684600696e-06, "loss": 1.2557, "step": 39683 }, { "epoch": 3.1100313479623822, "grad_norm": 4.363012313842773, "learning_rate": 1.1769228259129078e-06, "loss": 1.4421, "step": 39684 }, { "epoch": 3.1101097178683386, "grad_norm": 2.5966577529907227, "learning_rate": 1.1767240979067712e-06, "loss": 1.4491, "step": 39685 }, { "epoch": 3.1101880877742945, "grad_norm": 5.8413496017456055, "learning_rate": 1.1765253844424107e-06, "loss": 1.2534, "step": 39686 }, { "epoch": 3.110266457680251, "grad_norm": 3.784712791442871, "learning_rate": 1.1763266855205835e-06, "loss": 1.5927, "step": 39687 }, { "epoch": 3.110344827586207, "grad_norm": 6.997175693511963, "learning_rate": 1.1761280011420456e-06, "loss": 1.3356, "step": 39688 }, { "epoch": 3.110423197492163, "grad_norm": 2.9290552139282227, "learning_rate": 1.1759293313075532e-06, "loss": 1.2022, "step": 39689 }, { "epoch": 3.110501567398119, "grad_norm": 6.064873695373535, "learning_rate": 1.175730676017861e-06, "loss": 1.5134, "step": 39690 }, { "epoch": 3.1105799373040752, "grad_norm": 2.7427847385406494, "learning_rate": 1.1755320352737253e-06, "loss": 1.4152, "step": 39691 }, { "epoch": 3.1106583072100316, "grad_norm": 5.088552951812744, "learning_rate": 1.1753334090759028e-06, "loss": 1.3547, "step": 39692 }, { "epoch": 3.1107366771159874, "grad_norm": 4.723084926605225, "learning_rate": 1.1751347974251448e-06, "loss": 0.7997, "step": 39693 }, { "epoch": 3.110815047021944, "grad_norm": 4.035284996032715, "learning_rate": 1.174936200322211e-06, "loss": 1.2461, "step": 39694 }, { "epoch": 3.1108934169278997, "grad_norm": 14.734099388122559, "learning_rate": 1.174737617767855e-06, "loss": 0.6965, "step": 39695 }, { "epoch": 3.110971786833856, "grad_norm": 4.400027275085449, "learning_rate": 1.1745390497628335e-06, "loss": 1.4427, "step": 39696 }, { "epoch": 3.111050156739812, "grad_norm": 4.972009181976318, "learning_rate": 1.174340496307899e-06, "loss": 1.3837, "step": 39697 }, { "epoch": 3.111128526645768, "grad_norm": 5.5750412940979, "learning_rate": 1.1741419574038077e-06, "loss": 1.3733, "step": 39698 }, { "epoch": 3.111206896551724, "grad_norm": 7.759016990661621, "learning_rate": 1.173943433051316e-06, "loss": 1.3246, "step": 39699 }, { "epoch": 3.1112852664576804, "grad_norm": 3.957632064819336, "learning_rate": 1.1737449232511799e-06, "loss": 1.4015, "step": 39700 }, { "epoch": 3.1113636363636363, "grad_norm": 3.0770387649536133, "learning_rate": 1.1735464280041515e-06, "loss": 1.2892, "step": 39701 }, { "epoch": 3.1114420062695927, "grad_norm": 10.463512420654297, "learning_rate": 1.1733479473109871e-06, "loss": 1.125, "step": 39702 }, { "epoch": 3.1115203761755486, "grad_norm": 3.9428751468658447, "learning_rate": 1.1731494811724414e-06, "loss": 1.6425, "step": 39703 }, { "epoch": 3.111598746081505, "grad_norm": 4.074771404266357, "learning_rate": 1.17295102958927e-06, "loss": 0.4931, "step": 39704 }, { "epoch": 3.1116771159874608, "grad_norm": 3.279982566833496, "learning_rate": 1.1727525925622264e-06, "loss": 1.2391, "step": 39705 }, { "epoch": 3.111755485893417, "grad_norm": 9.98273754119873, "learning_rate": 1.1725541700920668e-06, "loss": 1.2636, "step": 39706 }, { "epoch": 3.111833855799373, "grad_norm": 3.3722410202026367, "learning_rate": 1.172355762179545e-06, "loss": 1.3848, "step": 39707 }, { "epoch": 3.1119122257053293, "grad_norm": 4.312453746795654, "learning_rate": 1.1721573688254157e-06, "loss": 1.1977, "step": 39708 }, { "epoch": 3.111990595611285, "grad_norm": 5.16834020614624, "learning_rate": 1.171958990030434e-06, "loss": 0.8175, "step": 39709 }, { "epoch": 3.1120689655172415, "grad_norm": 4.248335838317871, "learning_rate": 1.171760625795354e-06, "loss": 0.9084, "step": 39710 }, { "epoch": 3.1121473354231974, "grad_norm": 2.3355915546417236, "learning_rate": 1.1715622761209316e-06, "loss": 1.3147, "step": 39711 }, { "epoch": 3.1122257053291538, "grad_norm": 16.59461212158203, "learning_rate": 1.1713639410079169e-06, "loss": 1.3656, "step": 39712 }, { "epoch": 3.1123040752351097, "grad_norm": 3.563796043395996, "learning_rate": 1.1711656204570686e-06, "loss": 1.4068, "step": 39713 }, { "epoch": 3.112382445141066, "grad_norm": 2.9508166313171387, "learning_rate": 1.1709673144691414e-06, "loss": 1.2413, "step": 39714 }, { "epoch": 3.112460815047022, "grad_norm": 2.4484505653381348, "learning_rate": 1.1707690230448864e-06, "loss": 1.3355, "step": 39715 }, { "epoch": 3.112539184952978, "grad_norm": 2.0963215827941895, "learning_rate": 1.1705707461850592e-06, "loss": 1.2759, "step": 39716 }, { "epoch": 3.112617554858934, "grad_norm": 4.898629188537598, "learning_rate": 1.1703724838904123e-06, "loss": 1.1175, "step": 39717 }, { "epoch": 3.1126959247648904, "grad_norm": 4.340194225311279, "learning_rate": 1.1701742361617047e-06, "loss": 1.3803, "step": 39718 }, { "epoch": 3.1127742946708463, "grad_norm": 2.7434446811676025, "learning_rate": 1.1699760029996854e-06, "loss": 1.2286, "step": 39719 }, { "epoch": 3.1128526645768027, "grad_norm": 4.102241516113281, "learning_rate": 1.1697777844051105e-06, "loss": 1.0337, "step": 39720 }, { "epoch": 3.1129310344827585, "grad_norm": 6.597087383270264, "learning_rate": 1.1695795803787335e-06, "loss": 1.5065, "step": 39721 }, { "epoch": 3.113009404388715, "grad_norm": 2.795708417892456, "learning_rate": 1.169381390921308e-06, "loss": 1.16, "step": 39722 }, { "epoch": 3.1130877742946708, "grad_norm": 3.415651798248291, "learning_rate": 1.1691832160335887e-06, "loss": 1.2751, "step": 39723 }, { "epoch": 3.113166144200627, "grad_norm": 6.788586616516113, "learning_rate": 1.1689850557163284e-06, "loss": 1.3208, "step": 39724 }, { "epoch": 3.113244514106583, "grad_norm": 7.5919013023376465, "learning_rate": 1.1687869099702815e-06, "loss": 1.0322, "step": 39725 }, { "epoch": 3.1133228840125393, "grad_norm": 2.412893295288086, "learning_rate": 1.1685887787962009e-06, "loss": 1.4255, "step": 39726 }, { "epoch": 3.113401253918495, "grad_norm": 4.989678382873535, "learning_rate": 1.168390662194841e-06, "loss": 1.0017, "step": 39727 }, { "epoch": 3.1134796238244515, "grad_norm": 8.542778968811035, "learning_rate": 1.1681925601669546e-06, "loss": 0.8703, "step": 39728 }, { "epoch": 3.1135579937304074, "grad_norm": 3.3061254024505615, "learning_rate": 1.1679944727132974e-06, "loss": 1.5151, "step": 39729 }, { "epoch": 3.1136363636363638, "grad_norm": 2.769064426422119, "learning_rate": 1.1677963998346182e-06, "loss": 1.1808, "step": 39730 }, { "epoch": 3.1137147335423196, "grad_norm": 2.7122654914855957, "learning_rate": 1.1675983415316744e-06, "loss": 1.0016, "step": 39731 }, { "epoch": 3.113793103448276, "grad_norm": 7.906895160675049, "learning_rate": 1.1674002978052185e-06, "loss": 0.8104, "step": 39732 }, { "epoch": 3.113871473354232, "grad_norm": 5.921292781829834, "learning_rate": 1.1672022686560041e-06, "loss": 1.1634, "step": 39733 }, { "epoch": 3.113949843260188, "grad_norm": 8.752437591552734, "learning_rate": 1.1670042540847825e-06, "loss": 1.0261, "step": 39734 }, { "epoch": 3.114028213166144, "grad_norm": 5.259548664093018, "learning_rate": 1.1668062540923076e-06, "loss": 1.3186, "step": 39735 }, { "epoch": 3.1141065830721004, "grad_norm": 2.7954819202423096, "learning_rate": 1.1666082686793318e-06, "loss": 1.422, "step": 39736 }, { "epoch": 3.1141849529780563, "grad_norm": 3.916177272796631, "learning_rate": 1.166410297846612e-06, "loss": 1.1791, "step": 39737 }, { "epoch": 3.1142633228840126, "grad_norm": 6.805528163909912, "learning_rate": 1.1662123415948968e-06, "loss": 1.4775, "step": 39738 }, { "epoch": 3.1143416927899685, "grad_norm": 3.706587553024292, "learning_rate": 1.1660143999249407e-06, "loss": 1.3721, "step": 39739 }, { "epoch": 3.114420062695925, "grad_norm": 3.5525104999542236, "learning_rate": 1.165816472837497e-06, "loss": 1.4282, "step": 39740 }, { "epoch": 3.1144984326018808, "grad_norm": 5.757652759552002, "learning_rate": 1.1656185603333175e-06, "loss": 1.4839, "step": 39741 }, { "epoch": 3.114576802507837, "grad_norm": 4.593517303466797, "learning_rate": 1.1654206624131559e-06, "loss": 1.7425, "step": 39742 }, { "epoch": 3.114655172413793, "grad_norm": 2.8130860328674316, "learning_rate": 1.1652227790777642e-06, "loss": 1.1469, "step": 39743 }, { "epoch": 3.1147335423197493, "grad_norm": 6.267756938934326, "learning_rate": 1.165024910327896e-06, "loss": 1.7142, "step": 39744 }, { "epoch": 3.114811912225705, "grad_norm": 2.4522109031677246, "learning_rate": 1.1648270561643027e-06, "loss": 1.5152, "step": 39745 }, { "epoch": 3.1148902821316615, "grad_norm": 2.3999321460723877, "learning_rate": 1.1646292165877377e-06, "loss": 1.5891, "step": 39746 }, { "epoch": 3.1149686520376174, "grad_norm": 4.01059103012085, "learning_rate": 1.164431391598953e-06, "loss": 0.6372, "step": 39747 }, { "epoch": 3.1150470219435737, "grad_norm": 4.221987724304199, "learning_rate": 1.1642335811987027e-06, "loss": 1.2672, "step": 39748 }, { "epoch": 3.1151253918495296, "grad_norm": 3.108243227005005, "learning_rate": 1.164035785387735e-06, "loss": 1.1742, "step": 39749 }, { "epoch": 3.115203761755486, "grad_norm": 2.6576175689697266, "learning_rate": 1.1638380041668058e-06, "loss": 1.3929, "step": 39750 }, { "epoch": 3.115282131661442, "grad_norm": 2.8642563819885254, "learning_rate": 1.1636402375366669e-06, "loss": 1.3758, "step": 39751 }, { "epoch": 3.115360501567398, "grad_norm": 4.156680583953857, "learning_rate": 1.1634424854980715e-06, "loss": 1.0742, "step": 39752 }, { "epoch": 3.115438871473354, "grad_norm": 3.233126401901245, "learning_rate": 1.1632447480517688e-06, "loss": 1.1715, "step": 39753 }, { "epoch": 3.1155172413793104, "grad_norm": 2.796614646911621, "learning_rate": 1.163047025198511e-06, "loss": 1.0078, "step": 39754 }, { "epoch": 3.1155956112852663, "grad_norm": 2.536206007003784, "learning_rate": 1.1628493169390526e-06, "loss": 1.1075, "step": 39755 }, { "epoch": 3.1156739811912226, "grad_norm": 15.981030464172363, "learning_rate": 1.1626516232741463e-06, "loss": 2.2389, "step": 39756 }, { "epoch": 3.1157523510971785, "grad_norm": 7.058164596557617, "learning_rate": 1.1624539442045408e-06, "loss": 1.5781, "step": 39757 }, { "epoch": 3.115830721003135, "grad_norm": 4.356619834899902, "learning_rate": 1.1622562797309894e-06, "loss": 1.2295, "step": 39758 }, { "epoch": 3.1159090909090907, "grad_norm": 5.455636978149414, "learning_rate": 1.1620586298542435e-06, "loss": 1.3916, "step": 39759 }, { "epoch": 3.115987460815047, "grad_norm": 2.7513537406921387, "learning_rate": 1.1618609945750558e-06, "loss": 1.1918, "step": 39760 }, { "epoch": 3.116065830721003, "grad_norm": 2.8463551998138428, "learning_rate": 1.1616633738941774e-06, "loss": 1.1589, "step": 39761 }, { "epoch": 3.1161442006269593, "grad_norm": 2.9223921298980713, "learning_rate": 1.1614657678123603e-06, "loss": 1.3665, "step": 39762 }, { "epoch": 3.116222570532915, "grad_norm": 3.1060047149658203, "learning_rate": 1.161268176330355e-06, "loss": 1.2249, "step": 39763 }, { "epoch": 3.1163009404388715, "grad_norm": 5.9803242683410645, "learning_rate": 1.1610705994489147e-06, "loss": 0.8845, "step": 39764 }, { "epoch": 3.1163793103448274, "grad_norm": 3.3552839756011963, "learning_rate": 1.1608730371687898e-06, "loss": 1.474, "step": 39765 }, { "epoch": 3.1164576802507837, "grad_norm": 10.15909194946289, "learning_rate": 1.1606754894907318e-06, "loss": 1.2686, "step": 39766 }, { "epoch": 3.1165360501567396, "grad_norm": 3.275376558303833, "learning_rate": 1.1604779564154938e-06, "loss": 1.0861, "step": 39767 }, { "epoch": 3.116614420062696, "grad_norm": 5.234377861022949, "learning_rate": 1.1602804379438226e-06, "loss": 1.1776, "step": 39768 }, { "epoch": 3.116692789968652, "grad_norm": 2.5451791286468506, "learning_rate": 1.1600829340764736e-06, "loss": 1.1749, "step": 39769 }, { "epoch": 3.116771159874608, "grad_norm": 3.3887476921081543, "learning_rate": 1.1598854448141988e-06, "loss": 0.9882, "step": 39770 }, { "epoch": 3.116849529780564, "grad_norm": 2.653164863586426, "learning_rate": 1.1596879701577458e-06, "loss": 1.3479, "step": 39771 }, { "epoch": 3.1169278996865204, "grad_norm": 3.583404064178467, "learning_rate": 1.1594905101078669e-06, "loss": 1.2785, "step": 39772 }, { "epoch": 3.1170062695924763, "grad_norm": 2.4870922565460205, "learning_rate": 1.1592930646653123e-06, "loss": 0.5821, "step": 39773 }, { "epoch": 3.1170846394984326, "grad_norm": 3.818019390106201, "learning_rate": 1.1590956338308368e-06, "loss": 1.1308, "step": 39774 }, { "epoch": 3.1171630094043885, "grad_norm": 3.192013740539551, "learning_rate": 1.1588982176051877e-06, "loss": 1.258, "step": 39775 }, { "epoch": 3.117241379310345, "grad_norm": 3.0194058418273926, "learning_rate": 1.1587008159891167e-06, "loss": 0.9704, "step": 39776 }, { "epoch": 3.1173197492163007, "grad_norm": 5.867697715759277, "learning_rate": 1.1585034289833747e-06, "loss": 0.7939, "step": 39777 }, { "epoch": 3.117398119122257, "grad_norm": 2.138009548187256, "learning_rate": 1.1583060565887122e-06, "loss": 1.354, "step": 39778 }, { "epoch": 3.1174764890282134, "grad_norm": 2.368586778640747, "learning_rate": 1.1581086988058805e-06, "loss": 1.3748, "step": 39779 }, { "epoch": 3.1175548589341693, "grad_norm": 3.8909318447113037, "learning_rate": 1.1579113556356302e-06, "loss": 1.239, "step": 39780 }, { "epoch": 3.117633228840125, "grad_norm": 5.042906761169434, "learning_rate": 1.1577140270787113e-06, "loss": 1.3171, "step": 39781 }, { "epoch": 3.1177115987460815, "grad_norm": 2.6427347660064697, "learning_rate": 1.157516713135875e-06, "loss": 1.26, "step": 39782 }, { "epoch": 3.117789968652038, "grad_norm": 10.001977920532227, "learning_rate": 1.1573194138078714e-06, "loss": 2.0701, "step": 39783 }, { "epoch": 3.1178683385579937, "grad_norm": 27.38474464416504, "learning_rate": 1.1571221290954515e-06, "loss": 1.4521, "step": 39784 }, { "epoch": 3.11794670846395, "grad_norm": 3.2455270290374756, "learning_rate": 1.156924858999366e-06, "loss": 1.3639, "step": 39785 }, { "epoch": 3.118025078369906, "grad_norm": 2.6900665760040283, "learning_rate": 1.1567276035203634e-06, "loss": 1.3582, "step": 39786 }, { "epoch": 3.1181034482758623, "grad_norm": 3.451005220413208, "learning_rate": 1.1565303626591933e-06, "loss": 1.0508, "step": 39787 }, { "epoch": 3.118181818181818, "grad_norm": 6.596742153167725, "learning_rate": 1.156333136416609e-06, "loss": 1.7111, "step": 39788 }, { "epoch": 3.1182601880877745, "grad_norm": 2.7447702884674072, "learning_rate": 1.1561359247933606e-06, "loss": 1.5378, "step": 39789 }, { "epoch": 3.1183385579937304, "grad_norm": 2.4519529342651367, "learning_rate": 1.1559387277901958e-06, "loss": 1.3357, "step": 39790 }, { "epoch": 3.1184169278996867, "grad_norm": 4.440295696258545, "learning_rate": 1.1557415454078646e-06, "loss": 1.1184, "step": 39791 }, { "epoch": 3.1184952978056426, "grad_norm": 3.5778939723968506, "learning_rate": 1.1555443776471176e-06, "loss": 1.2696, "step": 39792 }, { "epoch": 3.118573667711599, "grad_norm": 3.005030870437622, "learning_rate": 1.1553472245087076e-06, "loss": 1.1691, "step": 39793 }, { "epoch": 3.118652037617555, "grad_norm": 4.731088161468506, "learning_rate": 1.1551500859933806e-06, "loss": 1.3002, "step": 39794 }, { "epoch": 3.118730407523511, "grad_norm": 2.6185569763183594, "learning_rate": 1.154952962101888e-06, "loss": 1.2205, "step": 39795 }, { "epoch": 3.118808777429467, "grad_norm": 2.714690685272217, "learning_rate": 1.1547558528349795e-06, "loss": 1.3193, "step": 39796 }, { "epoch": 3.1188871473354234, "grad_norm": 4.0366997718811035, "learning_rate": 1.1545587581934043e-06, "loss": 1.3475, "step": 39797 }, { "epoch": 3.1189655172413793, "grad_norm": 3.2433977127075195, "learning_rate": 1.1543616781779127e-06, "loss": 1.2757, "step": 39798 }, { "epoch": 3.1190438871473356, "grad_norm": 3.6462788581848145, "learning_rate": 1.1541646127892541e-06, "loss": 1.1471, "step": 39799 }, { "epoch": 3.1191222570532915, "grad_norm": 3.826308012008667, "learning_rate": 1.1539675620281793e-06, "loss": 1.4424, "step": 39800 }, { "epoch": 3.119200626959248, "grad_norm": 2.8688101768493652, "learning_rate": 1.1537705258954335e-06, "loss": 1.1837, "step": 39801 }, { "epoch": 3.1192789968652037, "grad_norm": 2.7011704444885254, "learning_rate": 1.1535735043917706e-06, "loss": 1.4117, "step": 39802 }, { "epoch": 3.11935736677116, "grad_norm": 3.6094212532043457, "learning_rate": 1.1533764975179384e-06, "loss": 1.0968, "step": 39803 }, { "epoch": 3.119435736677116, "grad_norm": 3.2604258060455322, "learning_rate": 1.1531795052746874e-06, "loss": 1.1617, "step": 39804 }, { "epoch": 3.1195141065830723, "grad_norm": 8.847784996032715, "learning_rate": 1.1529825276627648e-06, "loss": 0.9909, "step": 39805 }, { "epoch": 3.119592476489028, "grad_norm": 2.9791715145111084, "learning_rate": 1.152785564682919e-06, "loss": 1.3819, "step": 39806 }, { "epoch": 3.1196708463949845, "grad_norm": 2.7784831523895264, "learning_rate": 1.1525886163359023e-06, "loss": 1.2372, "step": 39807 }, { "epoch": 3.1197492163009404, "grad_norm": 3.1570684909820557, "learning_rate": 1.152391682622464e-06, "loss": 1.4463, "step": 39808 }, { "epoch": 3.1198275862068967, "grad_norm": 6.697591781616211, "learning_rate": 1.15219476354335e-06, "loss": 0.4391, "step": 39809 }, { "epoch": 3.1199059561128526, "grad_norm": 4.454008102416992, "learning_rate": 1.1519978590993103e-06, "loss": 1.1931, "step": 39810 }, { "epoch": 3.119984326018809, "grad_norm": 3.1288063526153564, "learning_rate": 1.151800969291093e-06, "loss": 1.2342, "step": 39811 }, { "epoch": 3.120062695924765, "grad_norm": 2.7784862518310547, "learning_rate": 1.1516040941194517e-06, "loss": 1.3299, "step": 39812 }, { "epoch": 3.120141065830721, "grad_norm": 3.2125866413116455, "learning_rate": 1.1514072335851301e-06, "loss": 1.4047, "step": 39813 }, { "epoch": 3.120219435736677, "grad_norm": 25.76721954345703, "learning_rate": 1.151210387688878e-06, "loss": 1.7956, "step": 39814 }, { "epoch": 3.1202978056426334, "grad_norm": 5.206148624420166, "learning_rate": 1.1510135564314457e-06, "loss": 0.9528, "step": 39815 }, { "epoch": 3.1203761755485893, "grad_norm": 3.6791632175445557, "learning_rate": 1.15081673981358e-06, "loss": 1.1229, "step": 39816 }, { "epoch": 3.1204545454545456, "grad_norm": 2.505788564682007, "learning_rate": 1.1506199378360305e-06, "loss": 1.2146, "step": 39817 }, { "epoch": 3.1205329153605015, "grad_norm": 10.653765678405762, "learning_rate": 1.150423150499546e-06, "loss": 1.7101, "step": 39818 }, { "epoch": 3.120611285266458, "grad_norm": 3.9086694717407227, "learning_rate": 1.1502263778048756e-06, "loss": 1.9882, "step": 39819 }, { "epoch": 3.1206896551724137, "grad_norm": 3.4564735889434814, "learning_rate": 1.1500296197527643e-06, "loss": 1.4152, "step": 39820 }, { "epoch": 3.12076802507837, "grad_norm": 3.3268580436706543, "learning_rate": 1.1498328763439642e-06, "loss": 1.7164, "step": 39821 }, { "epoch": 3.120846394984326, "grad_norm": 2.0151519775390625, "learning_rate": 1.1496361475792218e-06, "loss": 1.0506, "step": 39822 }, { "epoch": 3.1209247648902823, "grad_norm": 5.0023651123046875, "learning_rate": 1.1494394334592873e-06, "loss": 1.0966, "step": 39823 }, { "epoch": 3.121003134796238, "grad_norm": 4.149803638458252, "learning_rate": 1.1492427339849067e-06, "loss": 1.9624, "step": 39824 }, { "epoch": 3.1210815047021945, "grad_norm": 2.940460205078125, "learning_rate": 1.1490460491568267e-06, "loss": 1.0217, "step": 39825 }, { "epoch": 3.1211598746081504, "grad_norm": 2.3512375354766846, "learning_rate": 1.148849378975801e-06, "loss": 1.3388, "step": 39826 }, { "epoch": 3.1212382445141067, "grad_norm": 3.582073211669922, "learning_rate": 1.148652723442572e-06, "loss": 1.1959, "step": 39827 }, { "epoch": 3.1213166144200626, "grad_norm": 3.1677777767181396, "learning_rate": 1.1484560825578904e-06, "loss": 1.1952, "step": 39828 }, { "epoch": 3.121394984326019, "grad_norm": 2.63934063911438, "learning_rate": 1.1482594563225035e-06, "loss": 1.595, "step": 39829 }, { "epoch": 3.121473354231975, "grad_norm": 3.3542463779449463, "learning_rate": 1.1480628447371594e-06, "loss": 1.1953, "step": 39830 }, { "epoch": 3.121551724137931, "grad_norm": 5.460233688354492, "learning_rate": 1.1478662478026053e-06, "loss": 1.9351, "step": 39831 }, { "epoch": 3.121630094043887, "grad_norm": 4.78480863571167, "learning_rate": 1.1476696655195895e-06, "loss": 1.4215, "step": 39832 }, { "epoch": 3.1217084639498434, "grad_norm": 4.834259033203125, "learning_rate": 1.1474730978888598e-06, "loss": 1.4145, "step": 39833 }, { "epoch": 3.1217868338557992, "grad_norm": 5.642115592956543, "learning_rate": 1.147276544911164e-06, "loss": 1.6408, "step": 39834 }, { "epoch": 3.1218652037617556, "grad_norm": 2.784856081008911, "learning_rate": 1.1470800065872489e-06, "loss": 1.3219, "step": 39835 }, { "epoch": 3.1219435736677115, "grad_norm": 2.927879810333252, "learning_rate": 1.1468834829178627e-06, "loss": 1.0498, "step": 39836 }, { "epoch": 3.122021943573668, "grad_norm": 3.3741304874420166, "learning_rate": 1.1466869739037522e-06, "loss": 1.2804, "step": 39837 }, { "epoch": 3.1221003134796237, "grad_norm": 2.805927276611328, "learning_rate": 1.146490479545666e-06, "loss": 1.3716, "step": 39838 }, { "epoch": 3.12217868338558, "grad_norm": 2.0964863300323486, "learning_rate": 1.14629399984435e-06, "loss": 1.1295, "step": 39839 }, { "epoch": 3.122257053291536, "grad_norm": 4.392054080963135, "learning_rate": 1.1460975348005527e-06, "loss": 1.405, "step": 39840 }, { "epoch": 3.1223354231974922, "grad_norm": 9.478605270385742, "learning_rate": 1.145901084415022e-06, "loss": 1.1311, "step": 39841 }, { "epoch": 3.122413793103448, "grad_norm": 3.0706262588500977, "learning_rate": 1.1457046486885027e-06, "loss": 1.0543, "step": 39842 }, { "epoch": 3.1224921630094045, "grad_norm": 3.1757683753967285, "learning_rate": 1.1455082276217416e-06, "loss": 1.1541, "step": 39843 }, { "epoch": 3.1225705329153604, "grad_norm": 2.7931742668151855, "learning_rate": 1.1453118212154895e-06, "loss": 1.3944, "step": 39844 }, { "epoch": 3.1226489028213167, "grad_norm": 2.953683853149414, "learning_rate": 1.1451154294704919e-06, "loss": 1.2861, "step": 39845 }, { "epoch": 3.1227272727272726, "grad_norm": 3.3101041316986084, "learning_rate": 1.1449190523874942e-06, "loss": 1.3102, "step": 39846 }, { "epoch": 3.122805642633229, "grad_norm": 2.8571665287017822, "learning_rate": 1.1447226899672448e-06, "loss": 1.4217, "step": 39847 }, { "epoch": 3.122884012539185, "grad_norm": 10.875085830688477, "learning_rate": 1.144526342210488e-06, "loss": 0.9393, "step": 39848 }, { "epoch": 3.122962382445141, "grad_norm": 4.375975608825684, "learning_rate": 1.144330009117976e-06, "loss": 0.9573, "step": 39849 }, { "epoch": 3.123040752351097, "grad_norm": 3.626368284225464, "learning_rate": 1.1441336906904504e-06, "loss": 1.2365, "step": 39850 }, { "epoch": 3.1231191222570533, "grad_norm": 4.003573894500732, "learning_rate": 1.1439373869286597e-06, "loss": 1.2158, "step": 39851 }, { "epoch": 3.1231974921630092, "grad_norm": 3.963547945022583, "learning_rate": 1.1437410978333507e-06, "loss": 1.1913, "step": 39852 }, { "epoch": 3.1232758620689656, "grad_norm": 4.445033550262451, "learning_rate": 1.1435448234052703e-06, "loss": 1.496, "step": 39853 }, { "epoch": 3.1233542319749215, "grad_norm": 3.3648855686187744, "learning_rate": 1.143348563645164e-06, "loss": 1.6858, "step": 39854 }, { "epoch": 3.123432601880878, "grad_norm": 4.586942195892334, "learning_rate": 1.143152318553779e-06, "loss": 1.468, "step": 39855 }, { "epoch": 3.1235109717868337, "grad_norm": 4.658688068389893, "learning_rate": 1.1429560881318636e-06, "loss": 1.8307, "step": 39856 }, { "epoch": 3.12358934169279, "grad_norm": 5.125243663787842, "learning_rate": 1.1427598723801586e-06, "loss": 1.5833, "step": 39857 }, { "epoch": 3.123667711598746, "grad_norm": 3.751542329788208, "learning_rate": 1.1425636712994155e-06, "loss": 1.4216, "step": 39858 }, { "epoch": 3.1237460815047022, "grad_norm": 3.625732898712158, "learning_rate": 1.1423674848903787e-06, "loss": 1.2308, "step": 39859 }, { "epoch": 3.123824451410658, "grad_norm": 4.739847660064697, "learning_rate": 1.1421713131537958e-06, "loss": 1.2871, "step": 39860 }, { "epoch": 3.1239028213166145, "grad_norm": 4.557687759399414, "learning_rate": 1.1419751560904114e-06, "loss": 0.8258, "step": 39861 }, { "epoch": 3.1239811912225703, "grad_norm": 2.3922243118286133, "learning_rate": 1.1417790137009693e-06, "loss": 1.1447, "step": 39862 }, { "epoch": 3.1240595611285267, "grad_norm": 7.040955066680908, "learning_rate": 1.14158288598622e-06, "loss": 1.4792, "step": 39863 }, { "epoch": 3.1241379310344826, "grad_norm": 3.3127315044403076, "learning_rate": 1.1413867729469092e-06, "loss": 1.312, "step": 39864 }, { "epoch": 3.124216300940439, "grad_norm": 6.400704860687256, "learning_rate": 1.1411906745837792e-06, "loss": 1.5936, "step": 39865 }, { "epoch": 3.124294670846395, "grad_norm": 2.8585097789764404, "learning_rate": 1.1409945908975784e-06, "loss": 1.2641, "step": 39866 }, { "epoch": 3.124373040752351, "grad_norm": 3.762887477874756, "learning_rate": 1.1407985218890504e-06, "loss": 1.5264, "step": 39867 }, { "epoch": 3.124451410658307, "grad_norm": 5.0733513832092285, "learning_rate": 1.1406024675589455e-06, "loss": 1.1705, "step": 39868 }, { "epoch": 3.1245297805642633, "grad_norm": 2.157480001449585, "learning_rate": 1.1404064279080052e-06, "loss": 1.2699, "step": 39869 }, { "epoch": 3.1246081504702192, "grad_norm": 3.155792474746704, "learning_rate": 1.1402104029369766e-06, "loss": 1.0379, "step": 39870 }, { "epoch": 3.1246865203761756, "grad_norm": 5.238938808441162, "learning_rate": 1.1400143926466046e-06, "loss": 1.5964, "step": 39871 }, { "epoch": 3.1247648902821314, "grad_norm": 11.117377281188965, "learning_rate": 1.1398183970376358e-06, "loss": 1.2891, "step": 39872 }, { "epoch": 3.124843260188088, "grad_norm": 3.8889784812927246, "learning_rate": 1.1396224161108148e-06, "loss": 1.041, "step": 39873 }, { "epoch": 3.124921630094044, "grad_norm": 3.559166193008423, "learning_rate": 1.1394264498668872e-06, "loss": 1.2865, "step": 39874 }, { "epoch": 3.125, "grad_norm": 3.264512538909912, "learning_rate": 1.1392304983066e-06, "loss": 1.3941, "step": 39875 }, { "epoch": 3.125078369905956, "grad_norm": 3.43560528755188, "learning_rate": 1.1390345614306942e-06, "loss": 1.1476, "step": 39876 }, { "epoch": 3.125156739811912, "grad_norm": 3.3990917205810547, "learning_rate": 1.1388386392399192e-06, "loss": 1.4847, "step": 39877 }, { "epoch": 3.1252351097178686, "grad_norm": 3.996028184890747, "learning_rate": 1.1386427317350184e-06, "loss": 1.4641, "step": 39878 }, { "epoch": 3.1253134796238244, "grad_norm": 2.5110647678375244, "learning_rate": 1.1384468389167398e-06, "loss": 1.3571, "step": 39879 }, { "epoch": 3.1253918495297803, "grad_norm": 2.3354833126068115, "learning_rate": 1.1382509607858233e-06, "loss": 1.3631, "step": 39880 }, { "epoch": 3.1254702194357367, "grad_norm": 2.5119364261627197, "learning_rate": 1.138055097343016e-06, "loss": 1.2629, "step": 39881 }, { "epoch": 3.125548589341693, "grad_norm": 4.316011428833008, "learning_rate": 1.137859248589066e-06, "loss": 1.6718, "step": 39882 }, { "epoch": 3.125626959247649, "grad_norm": 5.277353286743164, "learning_rate": 1.1376634145247144e-06, "loss": 1.6928, "step": 39883 }, { "epoch": 3.125705329153605, "grad_norm": 3.010498523712158, "learning_rate": 1.1374675951507075e-06, "loss": 1.4071, "step": 39884 }, { "epoch": 3.125783699059561, "grad_norm": 3.230315685272217, "learning_rate": 1.1372717904677899e-06, "loss": 1.1767, "step": 39885 }, { "epoch": 3.1258620689655174, "grad_norm": 2.98795485496521, "learning_rate": 1.1370760004767062e-06, "loss": 1.1771, "step": 39886 }, { "epoch": 3.1259404388714733, "grad_norm": 3.064603090286255, "learning_rate": 1.1368802251782014e-06, "loss": 1.4124, "step": 39887 }, { "epoch": 3.1260188087774297, "grad_norm": 3.908963203430176, "learning_rate": 1.1366844645730201e-06, "loss": 1.3675, "step": 39888 }, { "epoch": 3.1260971786833855, "grad_norm": 11.618534088134766, "learning_rate": 1.1364887186619067e-06, "loss": 1.1124, "step": 39889 }, { "epoch": 3.126175548589342, "grad_norm": 5.8506760597229, "learning_rate": 1.1362929874456057e-06, "loss": 0.7049, "step": 39890 }, { "epoch": 3.1262539184952978, "grad_norm": 4.0185441970825195, "learning_rate": 1.1360972709248618e-06, "loss": 1.5058, "step": 39891 }, { "epoch": 3.126332288401254, "grad_norm": 4.626134395599365, "learning_rate": 1.1359015691004188e-06, "loss": 1.509, "step": 39892 }, { "epoch": 3.12641065830721, "grad_norm": 3.951183319091797, "learning_rate": 1.1357058819730221e-06, "loss": 1.1987, "step": 39893 }, { "epoch": 3.1264890282131663, "grad_norm": 3.65537428855896, "learning_rate": 1.1355102095434168e-06, "loss": 1.3891, "step": 39894 }, { "epoch": 3.126567398119122, "grad_norm": 6.719956874847412, "learning_rate": 1.1353145518123426e-06, "loss": 1.2402, "step": 39895 }, { "epoch": 3.1266457680250785, "grad_norm": 4.939021587371826, "learning_rate": 1.1351189087805482e-06, "loss": 1.1263, "step": 39896 }, { "epoch": 3.1267241379310344, "grad_norm": 4.156402587890625, "learning_rate": 1.1349232804487776e-06, "loss": 1.2849, "step": 39897 }, { "epoch": 3.1268025078369908, "grad_norm": 3.548379421234131, "learning_rate": 1.1347276668177725e-06, "loss": 1.2527, "step": 39898 }, { "epoch": 3.1268808777429467, "grad_norm": 3.4086103439331055, "learning_rate": 1.1345320678882782e-06, "loss": 1.1236, "step": 39899 }, { "epoch": 3.126959247648903, "grad_norm": 4.486826419830322, "learning_rate": 1.1343364836610366e-06, "loss": 1.6807, "step": 39900 }, { "epoch": 3.127037617554859, "grad_norm": 10.467306137084961, "learning_rate": 1.1341409141367965e-06, "loss": 0.911, "step": 39901 }, { "epoch": 3.127115987460815, "grad_norm": 5.459743499755859, "learning_rate": 1.1339453593162975e-06, "loss": 1.3998, "step": 39902 }, { "epoch": 3.127194357366771, "grad_norm": 3.73227596282959, "learning_rate": 1.1337498192002844e-06, "loss": 1.6192, "step": 39903 }, { "epoch": 3.1272727272727274, "grad_norm": 2.7275731563568115, "learning_rate": 1.1335542937895016e-06, "loss": 1.2876, "step": 39904 }, { "epoch": 3.1273510971786833, "grad_norm": 4.300987243652344, "learning_rate": 1.133358783084692e-06, "loss": 1.022, "step": 39905 }, { "epoch": 3.1274294670846396, "grad_norm": 3.1213574409484863, "learning_rate": 1.1331632870865993e-06, "loss": 1.1603, "step": 39906 }, { "epoch": 3.1275078369905955, "grad_norm": 4.001786231994629, "learning_rate": 1.1329678057959674e-06, "loss": 1.2962, "step": 39907 }, { "epoch": 3.127586206896552, "grad_norm": 10.969775199890137, "learning_rate": 1.1327723392135403e-06, "loss": 1.0314, "step": 39908 }, { "epoch": 3.1276645768025078, "grad_norm": 3.2308297157287598, "learning_rate": 1.1325768873400605e-06, "loss": 1.3422, "step": 39909 }, { "epoch": 3.127742946708464, "grad_norm": 3.1471872329711914, "learning_rate": 1.1323814501762714e-06, "loss": 1.4849, "step": 39910 }, { "epoch": 3.12782131661442, "grad_norm": 3.2039222717285156, "learning_rate": 1.1321860277229174e-06, "loss": 1.362, "step": 39911 }, { "epoch": 3.1278996865203763, "grad_norm": 3.1467678546905518, "learning_rate": 1.1319906199807423e-06, "loss": 1.3097, "step": 39912 }, { "epoch": 3.127978056426332, "grad_norm": 2.523205041885376, "learning_rate": 1.1317952269504867e-06, "loss": 1.0702, "step": 39913 }, { "epoch": 3.1280564263322885, "grad_norm": 6.237528324127197, "learning_rate": 1.1315998486328939e-06, "loss": 1.1935, "step": 39914 }, { "epoch": 3.1281347962382444, "grad_norm": 3.766083002090454, "learning_rate": 1.1314044850287093e-06, "loss": 1.3149, "step": 39915 }, { "epoch": 3.1282131661442008, "grad_norm": 3.933899402618408, "learning_rate": 1.1312091361386768e-06, "loss": 1.7622, "step": 39916 }, { "epoch": 3.1282915360501566, "grad_norm": 7.168773174285889, "learning_rate": 1.1310138019635364e-06, "loss": 1.8799, "step": 39917 }, { "epoch": 3.128369905956113, "grad_norm": 2.4869308471679688, "learning_rate": 1.130818482504032e-06, "loss": 1.4148, "step": 39918 }, { "epoch": 3.128448275862069, "grad_norm": 2.4458224773406982, "learning_rate": 1.130623177760905e-06, "loss": 1.2675, "step": 39919 }, { "epoch": 3.128526645768025, "grad_norm": 3.7327637672424316, "learning_rate": 1.1304278877349034e-06, "loss": 1.1365, "step": 39920 }, { "epoch": 3.128605015673981, "grad_norm": 2.6724236011505127, "learning_rate": 1.1302326124267648e-06, "loss": 1.3061, "step": 39921 }, { "epoch": 3.1286833855799374, "grad_norm": 3.689680814743042, "learning_rate": 1.1300373518372337e-06, "loss": 1.5601, "step": 39922 }, { "epoch": 3.1287617554858933, "grad_norm": 2.7644901275634766, "learning_rate": 1.1298421059670534e-06, "loss": 1.1078, "step": 39923 }, { "epoch": 3.1288401253918496, "grad_norm": 6.8548197746276855, "learning_rate": 1.1296468748169653e-06, "loss": 0.7185, "step": 39924 }, { "epoch": 3.1289184952978055, "grad_norm": 3.03623366355896, "learning_rate": 1.1294516583877125e-06, "loss": 1.4206, "step": 39925 }, { "epoch": 3.128996865203762, "grad_norm": 4.933391094207764, "learning_rate": 1.129256456680038e-06, "loss": 0.9469, "step": 39926 }, { "epoch": 3.1290752351097177, "grad_norm": 3.6062850952148438, "learning_rate": 1.129061269694683e-06, "loss": 1.297, "step": 39927 }, { "epoch": 3.129153605015674, "grad_norm": 3.6311304569244385, "learning_rate": 1.128866097432391e-06, "loss": 1.4703, "step": 39928 }, { "epoch": 3.12923197492163, "grad_norm": 4.390038967132568, "learning_rate": 1.1286709398939044e-06, "loss": 1.2557, "step": 39929 }, { "epoch": 3.1293103448275863, "grad_norm": 3.0481371879577637, "learning_rate": 1.1284757970799653e-06, "loss": 1.5091, "step": 39930 }, { "epoch": 3.129388714733542, "grad_norm": 7.240025043487549, "learning_rate": 1.1282806689913168e-06, "loss": 1.1372, "step": 39931 }, { "epoch": 3.1294670846394985, "grad_norm": 4.078948974609375, "learning_rate": 1.1280855556286984e-06, "loss": 1.2551, "step": 39932 }, { "epoch": 3.1295454545454544, "grad_norm": 2.917957067489624, "learning_rate": 1.1278904569928524e-06, "loss": 1.2898, "step": 39933 }, { "epoch": 3.1296238244514107, "grad_norm": 2.51086688041687, "learning_rate": 1.1276953730845236e-06, "loss": 1.2546, "step": 39934 }, { "epoch": 3.1297021943573666, "grad_norm": 4.013666152954102, "learning_rate": 1.1275003039044547e-06, "loss": 1.1326, "step": 39935 }, { "epoch": 3.129780564263323, "grad_norm": 3.713592767715454, "learning_rate": 1.1273052494533838e-06, "loss": 1.6312, "step": 39936 }, { "epoch": 3.129858934169279, "grad_norm": 3.4525439739227295, "learning_rate": 1.1271102097320546e-06, "loss": 1.1523, "step": 39937 }, { "epoch": 3.129937304075235, "grad_norm": 3.186563014984131, "learning_rate": 1.1269151847412084e-06, "loss": 1.4231, "step": 39938 }, { "epoch": 3.130015673981191, "grad_norm": 3.028879404067993, "learning_rate": 1.126720174481588e-06, "loss": 1.5935, "step": 39939 }, { "epoch": 3.1300940438871474, "grad_norm": 3.4511303901672363, "learning_rate": 1.126525178953935e-06, "loss": 1.5011, "step": 39940 }, { "epoch": 3.1301724137931033, "grad_norm": 2.8863484859466553, "learning_rate": 1.1263301981589902e-06, "loss": 1.0363, "step": 39941 }, { "epoch": 3.1302507836990596, "grad_norm": 5.692923545837402, "learning_rate": 1.1261352320974956e-06, "loss": 1.7512, "step": 39942 }, { "epoch": 3.1303291536050155, "grad_norm": 2.6477444171905518, "learning_rate": 1.125940280770193e-06, "loss": 1.1267, "step": 39943 }, { "epoch": 3.130407523510972, "grad_norm": 3.0858771800994873, "learning_rate": 1.1257453441778238e-06, "loss": 1.4572, "step": 39944 }, { "epoch": 3.1304858934169277, "grad_norm": 2.677401065826416, "learning_rate": 1.1255504223211295e-06, "loss": 1.2499, "step": 39945 }, { "epoch": 3.130564263322884, "grad_norm": 4.388213634490967, "learning_rate": 1.1253555152008511e-06, "loss": 1.8776, "step": 39946 }, { "epoch": 3.13064263322884, "grad_norm": 2.808114528656006, "learning_rate": 1.1251606228177298e-06, "loss": 1.3333, "step": 39947 }, { "epoch": 3.1307210031347963, "grad_norm": 4.027437210083008, "learning_rate": 1.1249657451725078e-06, "loss": 1.3158, "step": 39948 }, { "epoch": 3.130799373040752, "grad_norm": 2.856436252593994, "learning_rate": 1.1247708822659258e-06, "loss": 1.2032, "step": 39949 }, { "epoch": 3.1308777429467085, "grad_norm": 2.234156608581543, "learning_rate": 1.124576034098726e-06, "loss": 1.3264, "step": 39950 }, { "epoch": 3.1309561128526644, "grad_norm": 2.583829641342163, "learning_rate": 1.1243812006716459e-06, "loss": 1.3655, "step": 39951 }, { "epoch": 3.1310344827586207, "grad_norm": 2.9035284519195557, "learning_rate": 1.1241863819854303e-06, "loss": 1.0931, "step": 39952 }, { "epoch": 3.1311128526645766, "grad_norm": 2.1495015621185303, "learning_rate": 1.1239915780408201e-06, "loss": 1.4032, "step": 39953 }, { "epoch": 3.131191222570533, "grad_norm": 3.0619843006134033, "learning_rate": 1.1237967888385543e-06, "loss": 1.6955, "step": 39954 }, { "epoch": 3.131269592476489, "grad_norm": 2.978215217590332, "learning_rate": 1.123602014379374e-06, "loss": 1.4181, "step": 39955 }, { "epoch": 3.131347962382445, "grad_norm": 3.1691734790802, "learning_rate": 1.1234072546640196e-06, "loss": 1.3611, "step": 39956 }, { "epoch": 3.131426332288401, "grad_norm": 4.477632522583008, "learning_rate": 1.1232125096932355e-06, "loss": 1.8957, "step": 39957 }, { "epoch": 3.1315047021943574, "grad_norm": 6.196046352386475, "learning_rate": 1.1230177794677582e-06, "loss": 0.2552, "step": 39958 }, { "epoch": 3.1315830721003133, "grad_norm": 2.799454689025879, "learning_rate": 1.1228230639883303e-06, "loss": 1.4797, "step": 39959 }, { "epoch": 3.1316614420062696, "grad_norm": 2.4532153606414795, "learning_rate": 1.1226283632556916e-06, "loss": 1.0441, "step": 39960 }, { "epoch": 3.1317398119122255, "grad_norm": 4.831851482391357, "learning_rate": 1.1224336772705835e-06, "loss": 2.0649, "step": 39961 }, { "epoch": 3.131818181818182, "grad_norm": 2.6772468090057373, "learning_rate": 1.1222390060337457e-06, "loss": 1.2587, "step": 39962 }, { "epoch": 3.1318965517241377, "grad_norm": 5.204830169677734, "learning_rate": 1.1220443495459194e-06, "loss": 1.1647, "step": 39963 }, { "epoch": 3.131974921630094, "grad_norm": 3.6676740646362305, "learning_rate": 1.121849707807844e-06, "loss": 1.0201, "step": 39964 }, { "epoch": 3.1320532915360504, "grad_norm": 4.7562479972839355, "learning_rate": 1.1216550808202609e-06, "loss": 1.2156, "step": 39965 }, { "epoch": 3.1321316614420063, "grad_norm": 3.418536424636841, "learning_rate": 1.1214604685839093e-06, "loss": 1.3769, "step": 39966 }, { "epoch": 3.132210031347962, "grad_norm": 4.4277472496032715, "learning_rate": 1.1212658710995305e-06, "loss": 1.5647, "step": 39967 }, { "epoch": 3.1322884012539185, "grad_norm": 4.937773704528809, "learning_rate": 1.121071288367865e-06, "loss": 1.0645, "step": 39968 }, { "epoch": 3.132366771159875, "grad_norm": 3.230769634246826, "learning_rate": 1.1208767203896508e-06, "loss": 1.5156, "step": 39969 }, { "epoch": 3.1324451410658307, "grad_norm": 3.0488743782043457, "learning_rate": 1.1206821671656277e-06, "loss": 1.4028, "step": 39970 }, { "epoch": 3.1325235109717866, "grad_norm": 2.773879051208496, "learning_rate": 1.1204876286965382e-06, "loss": 0.6283, "step": 39971 }, { "epoch": 3.132601880877743, "grad_norm": 1.805749535560608, "learning_rate": 1.1202931049831228e-06, "loss": 1.2453, "step": 39972 }, { "epoch": 3.1326802507836993, "grad_norm": 2.778064250946045, "learning_rate": 1.1200985960261185e-06, "loss": 1.0514, "step": 39973 }, { "epoch": 3.132758620689655, "grad_norm": 5.82986307144165, "learning_rate": 1.1199041018262658e-06, "loss": 1.4023, "step": 39974 }, { "epoch": 3.1328369905956115, "grad_norm": 3.044945001602173, "learning_rate": 1.119709622384304e-06, "loss": 1.1388, "step": 39975 }, { "epoch": 3.1329153605015674, "grad_norm": 2.7857167720794678, "learning_rate": 1.1195151577009762e-06, "loss": 1.5269, "step": 39976 }, { "epoch": 3.1329937304075237, "grad_norm": 2.668240785598755, "learning_rate": 1.1193207077770185e-06, "loss": 1.1464, "step": 39977 }, { "epoch": 3.1330721003134796, "grad_norm": 2.54524827003479, "learning_rate": 1.1191262726131714e-06, "loss": 1.1352, "step": 39978 }, { "epoch": 3.133150470219436, "grad_norm": 4.515730857849121, "learning_rate": 1.1189318522101744e-06, "loss": 1.7862, "step": 39979 }, { "epoch": 3.133228840125392, "grad_norm": 4.033969402313232, "learning_rate": 1.1187374465687679e-06, "loss": 1.39, "step": 39980 }, { "epoch": 3.133307210031348, "grad_norm": 3.7522146701812744, "learning_rate": 1.1185430556896898e-06, "loss": 1.6258, "step": 39981 }, { "epoch": 3.133385579937304, "grad_norm": 3.3900816440582275, "learning_rate": 1.1183486795736808e-06, "loss": 1.4772, "step": 39982 }, { "epoch": 3.1334639498432604, "grad_norm": 4.605664253234863, "learning_rate": 1.118154318221481e-06, "loss": 1.5731, "step": 39983 }, { "epoch": 3.1335423197492163, "grad_norm": 2.864962100982666, "learning_rate": 1.1179599716338251e-06, "loss": 1.2113, "step": 39984 }, { "epoch": 3.1336206896551726, "grad_norm": 4.114645481109619, "learning_rate": 1.1177656398114579e-06, "loss": 1.0483, "step": 39985 }, { "epoch": 3.1336990595611285, "grad_norm": 3.402803421020508, "learning_rate": 1.1175713227551149e-06, "loss": 1.3271, "step": 39986 }, { "epoch": 3.133777429467085, "grad_norm": 4.071551322937012, "learning_rate": 1.1173770204655388e-06, "loss": 1.9095, "step": 39987 }, { "epoch": 3.1338557993730407, "grad_norm": 3.150913953781128, "learning_rate": 1.1171827329434648e-06, "loss": 1.2724, "step": 39988 }, { "epoch": 3.133934169278997, "grad_norm": 2.5065176486968994, "learning_rate": 1.1169884601896313e-06, "loss": 1.2804, "step": 39989 }, { "epoch": 3.134012539184953, "grad_norm": 2.843244791030884, "learning_rate": 1.1167942022047812e-06, "loss": 1.4024, "step": 39990 }, { "epoch": 3.1340909090909093, "grad_norm": 7.572970390319824, "learning_rate": 1.1165999589896526e-06, "loss": 1.3124, "step": 39991 }, { "epoch": 3.134169278996865, "grad_norm": 4.786386966705322, "learning_rate": 1.1164057305449817e-06, "loss": 1.8667, "step": 39992 }, { "epoch": 3.1342476489028215, "grad_norm": 4.016425132751465, "learning_rate": 1.1162115168715087e-06, "loss": 1.2245, "step": 39993 }, { "epoch": 3.1343260188087774, "grad_norm": 7.830002307891846, "learning_rate": 1.1160173179699724e-06, "loss": 1.2164, "step": 39994 }, { "epoch": 3.1344043887147337, "grad_norm": 2.199373960494995, "learning_rate": 1.1158231338411113e-06, "loss": 1.3613, "step": 39995 }, { "epoch": 3.1344827586206896, "grad_norm": 2.520411729812622, "learning_rate": 1.115628964485664e-06, "loss": 1.348, "step": 39996 }, { "epoch": 3.134561128526646, "grad_norm": 3.886211633682251, "learning_rate": 1.1154348099043687e-06, "loss": 1.0032, "step": 39997 }, { "epoch": 3.134639498432602, "grad_norm": 3.426363229751587, "learning_rate": 1.115240670097964e-06, "loss": 1.1223, "step": 39998 }, { "epoch": 3.134717868338558, "grad_norm": 2.822052001953125, "learning_rate": 1.1150465450671889e-06, "loss": 1.5093, "step": 39999 }, { "epoch": 3.134796238244514, "grad_norm": 4.2581562995910645, "learning_rate": 1.114852434812781e-06, "loss": 2.0004, "step": 40000 }, { "epoch": 3.1348746081504704, "grad_norm": 2.9357662200927734, "learning_rate": 1.1146583393354793e-06, "loss": 1.314, "step": 40001 }, { "epoch": 3.1349529780564263, "grad_norm": 3.832167387008667, "learning_rate": 1.1144642586360227e-06, "loss": 1.5771, "step": 40002 }, { "epoch": 3.1350313479623826, "grad_norm": 9.315933227539062, "learning_rate": 1.1142701927151456e-06, "loss": 1.1967, "step": 40003 }, { "epoch": 3.1351097178683385, "grad_norm": 4.771037578582764, "learning_rate": 1.1140761415735902e-06, "loss": 1.1289, "step": 40004 }, { "epoch": 3.135188087774295, "grad_norm": 4.0000834465026855, "learning_rate": 1.1138821052120928e-06, "loss": 1.1758, "step": 40005 }, { "epoch": 3.1352664576802507, "grad_norm": 2.76403546333313, "learning_rate": 1.113688083631394e-06, "loss": 1.3105, "step": 40006 }, { "epoch": 3.135344827586207, "grad_norm": 2.3692448139190674, "learning_rate": 1.1134940768322278e-06, "loss": 1.4015, "step": 40007 }, { "epoch": 3.135423197492163, "grad_norm": 3.939910650253296, "learning_rate": 1.113300084815333e-06, "loss": 1.6486, "step": 40008 }, { "epoch": 3.1355015673981192, "grad_norm": 3.2741997241973877, "learning_rate": 1.1131061075814502e-06, "loss": 0.9431, "step": 40009 }, { "epoch": 3.135579937304075, "grad_norm": 6.89335298538208, "learning_rate": 1.1129121451313145e-06, "loss": 1.5509, "step": 40010 }, { "epoch": 3.1356583072100315, "grad_norm": 2.031399726867676, "learning_rate": 1.1127181974656648e-06, "loss": 1.1494, "step": 40011 }, { "epoch": 3.1357366771159874, "grad_norm": 11.342639923095703, "learning_rate": 1.112524264585238e-06, "loss": 1.6062, "step": 40012 }, { "epoch": 3.1358150470219437, "grad_norm": 5.647346496582031, "learning_rate": 1.112330346490772e-06, "loss": 0.9373, "step": 40013 }, { "epoch": 3.1358934169278996, "grad_norm": 3.4258599281311035, "learning_rate": 1.1121364431830051e-06, "loss": 1.266, "step": 40014 }, { "epoch": 3.135971786833856, "grad_norm": 4.723948001861572, "learning_rate": 1.1119425546626738e-06, "loss": 1.2843, "step": 40015 }, { "epoch": 3.136050156739812, "grad_norm": 2.4762332439422607, "learning_rate": 1.1117486809305162e-06, "loss": 1.2732, "step": 40016 }, { "epoch": 3.136128526645768, "grad_norm": 2.8473217487335205, "learning_rate": 1.111554821987269e-06, "loss": 0.7358, "step": 40017 }, { "epoch": 3.136206896551724, "grad_norm": 2.2047462463378906, "learning_rate": 1.1113609778336704e-06, "loss": 1.2599, "step": 40018 }, { "epoch": 3.1362852664576804, "grad_norm": 2.830764055252075, "learning_rate": 1.1111671484704572e-06, "loss": 1.2912, "step": 40019 }, { "epoch": 3.1363636363636362, "grad_norm": 4.271641731262207, "learning_rate": 1.110973333898367e-06, "loss": 1.5787, "step": 40020 }, { "epoch": 3.1364420062695926, "grad_norm": 2.926474094390869, "learning_rate": 1.1107795341181381e-06, "loss": 1.3316, "step": 40021 }, { "epoch": 3.1365203761755485, "grad_norm": 2.0202064514160156, "learning_rate": 1.1105857491305033e-06, "loss": 1.3642, "step": 40022 }, { "epoch": 3.136598746081505, "grad_norm": 3.9329960346221924, "learning_rate": 1.1103919789362038e-06, "loss": 1.8467, "step": 40023 }, { "epoch": 3.1366771159874607, "grad_norm": 3.1483426094055176, "learning_rate": 1.110198223535977e-06, "loss": 1.156, "step": 40024 }, { "epoch": 3.136755485893417, "grad_norm": 3.5760724544525146, "learning_rate": 1.1100044829305567e-06, "loss": 1.4842, "step": 40025 }, { "epoch": 3.136833855799373, "grad_norm": 3.8223514556884766, "learning_rate": 1.1098107571206811e-06, "loss": 1.2071, "step": 40026 }, { "epoch": 3.1369122257053292, "grad_norm": 3.6517019271850586, "learning_rate": 1.1096170461070865e-06, "loss": 1.3337, "step": 40027 }, { "epoch": 3.136990595611285, "grad_norm": 10.320205688476562, "learning_rate": 1.1094233498905126e-06, "loss": 0.3169, "step": 40028 }, { "epoch": 3.1370689655172415, "grad_norm": 3.5211904048919678, "learning_rate": 1.109229668471693e-06, "loss": 1.3498, "step": 40029 }, { "epoch": 3.1371473354231973, "grad_norm": 3.004453659057617, "learning_rate": 1.1090360018513652e-06, "loss": 1.3753, "step": 40030 }, { "epoch": 3.1372257053291537, "grad_norm": 3.4573895931243896, "learning_rate": 1.108842350030266e-06, "loss": 1.3032, "step": 40031 }, { "epoch": 3.1373040752351096, "grad_norm": 3.4123733043670654, "learning_rate": 1.1086487130091317e-06, "loss": 1.4042, "step": 40032 }, { "epoch": 3.137382445141066, "grad_norm": 5.522991180419922, "learning_rate": 1.1084550907886986e-06, "loss": 1.3414, "step": 40033 }, { "epoch": 3.137460815047022, "grad_norm": 2.9053776264190674, "learning_rate": 1.1082614833697042e-06, "loss": 1.3151, "step": 40034 }, { "epoch": 3.137539184952978, "grad_norm": 3.24406099319458, "learning_rate": 1.1080678907528842e-06, "loss": 1.4238, "step": 40035 }, { "epoch": 3.137617554858934, "grad_norm": 3.043006181716919, "learning_rate": 1.1078743129389741e-06, "loss": 1.39, "step": 40036 }, { "epoch": 3.1376959247648903, "grad_norm": 3.1765241622924805, "learning_rate": 1.1076807499287119e-06, "loss": 1.4359, "step": 40037 }, { "epoch": 3.1377742946708462, "grad_norm": 2.5666489601135254, "learning_rate": 1.1074872017228321e-06, "loss": 1.7213, "step": 40038 }, { "epoch": 3.1378526645768026, "grad_norm": 2.5683765411376953, "learning_rate": 1.1072936683220736e-06, "loss": 1.316, "step": 40039 }, { "epoch": 3.1379310344827585, "grad_norm": 3.824018955230713, "learning_rate": 1.1071001497271678e-06, "loss": 1.3088, "step": 40040 }, { "epoch": 3.138009404388715, "grad_norm": 3.058363437652588, "learning_rate": 1.106906645938855e-06, "loss": 1.1546, "step": 40041 }, { "epoch": 3.1380877742946707, "grad_norm": 3.9322328567504883, "learning_rate": 1.1067131569578693e-06, "loss": 0.7171, "step": 40042 }, { "epoch": 3.138166144200627, "grad_norm": 6.479659080505371, "learning_rate": 1.1065196827849483e-06, "loss": 1.2376, "step": 40043 }, { "epoch": 3.138244514106583, "grad_norm": 2.37955379486084, "learning_rate": 1.1063262234208255e-06, "loss": 1.5134, "step": 40044 }, { "epoch": 3.1383228840125392, "grad_norm": 2.6545627117156982, "learning_rate": 1.1061327788662369e-06, "loss": 0.897, "step": 40045 }, { "epoch": 3.138401253918495, "grad_norm": 3.4014081954956055, "learning_rate": 1.10593934912192e-06, "loss": 1.1574, "step": 40046 }, { "epoch": 3.1384796238244514, "grad_norm": 4.2885823249816895, "learning_rate": 1.1057459341886112e-06, "loss": 1.2275, "step": 40047 }, { "epoch": 3.1385579937304073, "grad_norm": 3.5755481719970703, "learning_rate": 1.1055525340670431e-06, "loss": 1.5454, "step": 40048 }, { "epoch": 3.1386363636363637, "grad_norm": 2.575566053390503, "learning_rate": 1.105359148757953e-06, "loss": 1.5044, "step": 40049 }, { "epoch": 3.1387147335423196, "grad_norm": 2.624666213989258, "learning_rate": 1.1051657782620762e-06, "loss": 0.6655, "step": 40050 }, { "epoch": 3.138793103448276, "grad_norm": 3.549795627593994, "learning_rate": 1.1049724225801485e-06, "loss": 1.1824, "step": 40051 }, { "epoch": 3.138871473354232, "grad_norm": 2.573505163192749, "learning_rate": 1.104779081712905e-06, "loss": 1.1745, "step": 40052 }, { "epoch": 3.138949843260188, "grad_norm": 2.989894151687622, "learning_rate": 1.1045857556610811e-06, "loss": 0.8788, "step": 40053 }, { "epoch": 3.139028213166144, "grad_norm": 4.131856918334961, "learning_rate": 1.1043924444254122e-06, "loss": 1.0958, "step": 40054 }, { "epoch": 3.1391065830721003, "grad_norm": 3.3726656436920166, "learning_rate": 1.1041991480066338e-06, "loss": 1.2788, "step": 40055 }, { "epoch": 3.139184952978056, "grad_norm": 2.717179298400879, "learning_rate": 1.1040058664054804e-06, "loss": 1.2748, "step": 40056 }, { "epoch": 3.1392633228840126, "grad_norm": 5.233165264129639, "learning_rate": 1.1038125996226878e-06, "loss": 1.5479, "step": 40057 }, { "epoch": 3.1393416927899684, "grad_norm": 7.4729413986206055, "learning_rate": 1.1036193476589924e-06, "loss": 1.5333, "step": 40058 }, { "epoch": 3.1394200626959248, "grad_norm": 3.753117322921753, "learning_rate": 1.1034261105151245e-06, "loss": 1.6435, "step": 40059 }, { "epoch": 3.139498432601881, "grad_norm": 3.6999640464782715, "learning_rate": 1.1032328881918237e-06, "loss": 1.6401, "step": 40060 }, { "epoch": 3.139576802507837, "grad_norm": 3.094520330429077, "learning_rate": 1.1030396806898236e-06, "loss": 1.4413, "step": 40061 }, { "epoch": 3.139655172413793, "grad_norm": 3.2886288166046143, "learning_rate": 1.1028464880098604e-06, "loss": 1.2255, "step": 40062 }, { "epoch": 3.139733542319749, "grad_norm": 2.7171783447265625, "learning_rate": 1.102653310152666e-06, "loss": 0.6893, "step": 40063 }, { "epoch": 3.1398119122257055, "grad_norm": 3.1384711265563965, "learning_rate": 1.1024601471189755e-06, "loss": 1.0864, "step": 40064 }, { "epoch": 3.1398902821316614, "grad_norm": 6.378608703613281, "learning_rate": 1.102266998909527e-06, "loss": 1.8309, "step": 40065 }, { "epoch": 3.1399686520376173, "grad_norm": 2.2651889324188232, "learning_rate": 1.1020738655250518e-06, "loss": 1.4126, "step": 40066 }, { "epoch": 3.1400470219435737, "grad_norm": 6.679487705230713, "learning_rate": 1.1018807469662857e-06, "loss": 1.3657, "step": 40067 }, { "epoch": 3.14012539184953, "grad_norm": 3.6862540245056152, "learning_rate": 1.1016876432339623e-06, "loss": 1.2259, "step": 40068 }, { "epoch": 3.140203761755486, "grad_norm": 3.208397150039673, "learning_rate": 1.1014945543288175e-06, "loss": 1.6247, "step": 40069 }, { "epoch": 3.140282131661442, "grad_norm": 3.4510881900787354, "learning_rate": 1.1013014802515847e-06, "loss": 1.1245, "step": 40070 }, { "epoch": 3.140360501567398, "grad_norm": 6.278585433959961, "learning_rate": 1.1011084210029988e-06, "loss": 1.7436, "step": 40071 }, { "epoch": 3.1404388714733544, "grad_norm": 2.4682984352111816, "learning_rate": 1.1009153765837937e-06, "loss": 1.2081, "step": 40072 }, { "epoch": 3.1405172413793103, "grad_norm": 2.582937002182007, "learning_rate": 1.1007223469947037e-06, "loss": 1.1408, "step": 40073 }, { "epoch": 3.1405956112852667, "grad_norm": 2.5823793411254883, "learning_rate": 1.100529332236463e-06, "loss": 1.2208, "step": 40074 }, { "epoch": 3.1406739811912225, "grad_norm": 2.4650094509124756, "learning_rate": 1.1003363323098065e-06, "loss": 1.3419, "step": 40075 }, { "epoch": 3.140752351097179, "grad_norm": 2.178959369659424, "learning_rate": 1.100143347215467e-06, "loss": 1.2161, "step": 40076 }, { "epoch": 3.1408307210031348, "grad_norm": 3.0848071575164795, "learning_rate": 1.0999503769541803e-06, "loss": 1.405, "step": 40077 }, { "epoch": 3.140909090909091, "grad_norm": 2.8706376552581787, "learning_rate": 1.099757421526677e-06, "loss": 0.8046, "step": 40078 }, { "epoch": 3.140987460815047, "grad_norm": 4.698956489562988, "learning_rate": 1.099564480933694e-06, "loss": 1.3506, "step": 40079 }, { "epoch": 3.1410658307210033, "grad_norm": 6.409096717834473, "learning_rate": 1.0993715551759665e-06, "loss": 1.2043, "step": 40080 }, { "epoch": 3.141144200626959, "grad_norm": 3.346780776977539, "learning_rate": 1.099178644254224e-06, "loss": 1.2566, "step": 40081 }, { "epoch": 3.1412225705329155, "grad_norm": 10.250361442565918, "learning_rate": 1.0989857481692028e-06, "loss": 1.207, "step": 40082 }, { "epoch": 3.1413009404388714, "grad_norm": 6.66541051864624, "learning_rate": 1.098792866921634e-06, "loss": 1.6172, "step": 40083 }, { "epoch": 3.1413793103448278, "grad_norm": 2.3499562740325928, "learning_rate": 1.098600000512257e-06, "loss": 1.4126, "step": 40084 }, { "epoch": 3.1414576802507836, "grad_norm": 3.284170150756836, "learning_rate": 1.0984071489418003e-06, "loss": 1.2051, "step": 40085 }, { "epoch": 3.14153605015674, "grad_norm": 2.9700355529785156, "learning_rate": 1.0982143122109984e-06, "loss": 1.3679, "step": 40086 }, { "epoch": 3.141614420062696, "grad_norm": 8.71632194519043, "learning_rate": 1.098021490320585e-06, "loss": 0.7498, "step": 40087 }, { "epoch": 3.141692789968652, "grad_norm": 2.5739262104034424, "learning_rate": 1.0978286832712941e-06, "loss": 1.7811, "step": 40088 }, { "epoch": 3.141771159874608, "grad_norm": 3.6264867782592773, "learning_rate": 1.0976358910638584e-06, "loss": 1.4182, "step": 40089 }, { "epoch": 3.1418495297805644, "grad_norm": 5.269096851348877, "learning_rate": 1.0974431136990115e-06, "loss": 1.6013, "step": 40090 }, { "epoch": 3.1419278996865203, "grad_norm": 3.5094053745269775, "learning_rate": 1.097250351177488e-06, "loss": 1.2779, "step": 40091 }, { "epoch": 3.1420062695924766, "grad_norm": 5.116195201873779, "learning_rate": 1.0970576035000164e-06, "loss": 1.2129, "step": 40092 }, { "epoch": 3.1420846394984325, "grad_norm": 3.0745434761047363, "learning_rate": 1.0968648706673346e-06, "loss": 1.6285, "step": 40093 }, { "epoch": 3.142163009404389, "grad_norm": 6.656648635864258, "learning_rate": 1.0966721526801745e-06, "loss": 1.4311, "step": 40094 }, { "epoch": 3.1422413793103448, "grad_norm": 3.1665236949920654, "learning_rate": 1.0964794495392693e-06, "loss": 1.2563, "step": 40095 }, { "epoch": 3.142319749216301, "grad_norm": 4.740617275238037, "learning_rate": 1.0962867612453504e-06, "loss": 1.443, "step": 40096 }, { "epoch": 3.142398119122257, "grad_norm": 4.6106462478637695, "learning_rate": 1.09609408779915e-06, "loss": 1.2769, "step": 40097 }, { "epoch": 3.1424764890282133, "grad_norm": 5.27947998046875, "learning_rate": 1.0959014292014036e-06, "loss": 2.012, "step": 40098 }, { "epoch": 3.142554858934169, "grad_norm": 3.596174955368042, "learning_rate": 1.0957087854528447e-06, "loss": 1.5693, "step": 40099 }, { "epoch": 3.1426332288401255, "grad_norm": 5.324108123779297, "learning_rate": 1.0955161565542027e-06, "loss": 1.5514, "step": 40100 }, { "epoch": 3.1427115987460814, "grad_norm": 4.607724666595459, "learning_rate": 1.095323542506212e-06, "loss": 1.4101, "step": 40101 }, { "epoch": 3.1427899686520377, "grad_norm": 5.591935634613037, "learning_rate": 1.0951309433096035e-06, "loss": 1.574, "step": 40102 }, { "epoch": 3.1428683385579936, "grad_norm": 7.723315238952637, "learning_rate": 1.0949383589651135e-06, "loss": 1.0743, "step": 40103 }, { "epoch": 3.14294670846395, "grad_norm": 3.8290483951568604, "learning_rate": 1.0947457894734714e-06, "loss": 1.1943, "step": 40104 }, { "epoch": 3.143025078369906, "grad_norm": 6.143765926361084, "learning_rate": 1.0945532348354104e-06, "loss": 1.4604, "step": 40105 }, { "epoch": 3.143103448275862, "grad_norm": 3.612549066543579, "learning_rate": 1.0943606950516628e-06, "loss": 1.5462, "step": 40106 }, { "epoch": 3.143181818181818, "grad_norm": 4.389336585998535, "learning_rate": 1.0941681701229612e-06, "loss": 1.3199, "step": 40107 }, { "epoch": 3.1432601880877744, "grad_norm": 4.281391620635986, "learning_rate": 1.0939756600500372e-06, "loss": 0.8654, "step": 40108 }, { "epoch": 3.1433385579937303, "grad_norm": 4.762369632720947, "learning_rate": 1.093783164833624e-06, "loss": 1.2995, "step": 40109 }, { "epoch": 3.1434169278996866, "grad_norm": 3.2160966396331787, "learning_rate": 1.0935906844744548e-06, "loss": 1.4923, "step": 40110 }, { "epoch": 3.1434952978056425, "grad_norm": 4.486854553222656, "learning_rate": 1.0933982189732572e-06, "loss": 1.1983, "step": 40111 }, { "epoch": 3.143573667711599, "grad_norm": 9.517045021057129, "learning_rate": 1.0932057683307674e-06, "loss": 2.1683, "step": 40112 }, { "epoch": 3.1436520376175547, "grad_norm": 6.235776901245117, "learning_rate": 1.0930133325477165e-06, "loss": 1.793, "step": 40113 }, { "epoch": 3.143730407523511, "grad_norm": 4.785834312438965, "learning_rate": 1.0928209116248372e-06, "loss": 1.7047, "step": 40114 }, { "epoch": 3.143808777429467, "grad_norm": 5.326831817626953, "learning_rate": 1.0926285055628594e-06, "loss": 1.4964, "step": 40115 }, { "epoch": 3.1438871473354233, "grad_norm": 11.106167793273926, "learning_rate": 1.092436114362514e-06, "loss": 1.5616, "step": 40116 }, { "epoch": 3.143965517241379, "grad_norm": 5.302072525024414, "learning_rate": 1.092243738024536e-06, "loss": 1.1988, "step": 40117 }, { "epoch": 3.1440438871473355, "grad_norm": 2.970216989517212, "learning_rate": 1.0920513765496572e-06, "loss": 1.1459, "step": 40118 }, { "epoch": 3.1441222570532914, "grad_norm": 2.744185209274292, "learning_rate": 1.0918590299386067e-06, "loss": 1.3867, "step": 40119 }, { "epoch": 3.1442006269592477, "grad_norm": 2.6392621994018555, "learning_rate": 1.0916666981921164e-06, "loss": 1.4198, "step": 40120 }, { "epoch": 3.1442789968652036, "grad_norm": 8.187054634094238, "learning_rate": 1.091474381310919e-06, "loss": 1.2969, "step": 40121 }, { "epoch": 3.14435736677116, "grad_norm": 2.7616655826568604, "learning_rate": 1.091282079295745e-06, "loss": 1.4789, "step": 40122 }, { "epoch": 3.144435736677116, "grad_norm": 5.6687822341918945, "learning_rate": 1.0910897921473268e-06, "loss": 1.5547, "step": 40123 }, { "epoch": 3.144514106583072, "grad_norm": 6.303807735443115, "learning_rate": 1.090897519866395e-06, "loss": 1.8561, "step": 40124 }, { "epoch": 3.144592476489028, "grad_norm": 2.7445764541625977, "learning_rate": 1.0907052624536819e-06, "loss": 1.5368, "step": 40125 }, { "epoch": 3.1446708463949844, "grad_norm": 3.644911289215088, "learning_rate": 1.0905130199099174e-06, "loss": 1.7008, "step": 40126 }, { "epoch": 3.1447492163009403, "grad_norm": 5.716315746307373, "learning_rate": 1.0903207922358332e-06, "loss": 1.2243, "step": 40127 }, { "epoch": 3.1448275862068966, "grad_norm": 3.0418365001678467, "learning_rate": 1.0901285794321608e-06, "loss": 1.2499, "step": 40128 }, { "epoch": 3.1449059561128525, "grad_norm": 3.5185112953186035, "learning_rate": 1.0899363814996322e-06, "loss": 1.3678, "step": 40129 }, { "epoch": 3.144984326018809, "grad_norm": 7.288854598999023, "learning_rate": 1.089744198438975e-06, "loss": 0.9424, "step": 40130 }, { "epoch": 3.1450626959247647, "grad_norm": 2.5204038619995117, "learning_rate": 1.0895520302509233e-06, "loss": 1.1354, "step": 40131 }, { "epoch": 3.145141065830721, "grad_norm": 2.9153027534484863, "learning_rate": 1.0893598769362073e-06, "loss": 1.1789, "step": 40132 }, { "epoch": 3.145219435736677, "grad_norm": 4.605032920837402, "learning_rate": 1.089167738495559e-06, "loss": 1.5966, "step": 40133 }, { "epoch": 3.1452978056426333, "grad_norm": 2.7250852584838867, "learning_rate": 1.088975614929707e-06, "loss": 1.0847, "step": 40134 }, { "epoch": 3.145376175548589, "grad_norm": 2.5899064540863037, "learning_rate": 1.088783506239381e-06, "loss": 1.2587, "step": 40135 }, { "epoch": 3.1454545454545455, "grad_norm": 2.8924036026000977, "learning_rate": 1.0885914124253167e-06, "loss": 1.5888, "step": 40136 }, { "epoch": 3.1455329153605014, "grad_norm": 2.9350874423980713, "learning_rate": 1.0883993334882404e-06, "loss": 1.1111, "step": 40137 }, { "epoch": 3.1456112852664577, "grad_norm": 3.459808111190796, "learning_rate": 1.0882072694288837e-06, "loss": 1.2238, "step": 40138 }, { "epoch": 3.1456896551724136, "grad_norm": 3.1251981258392334, "learning_rate": 1.088015220247977e-06, "loss": 1.5206, "step": 40139 }, { "epoch": 3.14576802507837, "grad_norm": 10.405558586120605, "learning_rate": 1.0878231859462513e-06, "loss": 1.3577, "step": 40140 }, { "epoch": 3.145846394984326, "grad_norm": 12.744263648986816, "learning_rate": 1.0876311665244372e-06, "loss": 1.7631, "step": 40141 }, { "epoch": 3.145924764890282, "grad_norm": 3.083583354949951, "learning_rate": 1.0874391619832642e-06, "loss": 1.3601, "step": 40142 }, { "epoch": 3.146003134796238, "grad_norm": 7.228616237640381, "learning_rate": 1.0872471723234628e-06, "loss": 1.2463, "step": 40143 }, { "epoch": 3.1460815047021944, "grad_norm": 2.9126806259155273, "learning_rate": 1.0870551975457638e-06, "loss": 1.1661, "step": 40144 }, { "epoch": 3.1461598746081503, "grad_norm": 3.877347469329834, "learning_rate": 1.086863237650897e-06, "loss": 1.3382, "step": 40145 }, { "epoch": 3.1462382445141066, "grad_norm": 2.7956156730651855, "learning_rate": 1.0866712926395922e-06, "loss": 0.7724, "step": 40146 }, { "epoch": 3.1463166144200625, "grad_norm": 3.460810422897339, "learning_rate": 1.0864793625125813e-06, "loss": 1.2365, "step": 40147 }, { "epoch": 3.146394984326019, "grad_norm": 4.972141265869141, "learning_rate": 1.0862874472705904e-06, "loss": 1.3606, "step": 40148 }, { "epoch": 3.1464733542319747, "grad_norm": 8.398697853088379, "learning_rate": 1.086095546914353e-06, "loss": 1.1967, "step": 40149 }, { "epoch": 3.146551724137931, "grad_norm": 11.055198669433594, "learning_rate": 1.0859036614445977e-06, "loss": 1.5128, "step": 40150 }, { "epoch": 3.1466300940438874, "grad_norm": 2.4071333408355713, "learning_rate": 1.0857117908620562e-06, "loss": 1.1846, "step": 40151 }, { "epoch": 3.1467084639498433, "grad_norm": 10.695908546447754, "learning_rate": 1.085519935167455e-06, "loss": 1.1382, "step": 40152 }, { "epoch": 3.146786833855799, "grad_norm": 6.286434173583984, "learning_rate": 1.0853280943615235e-06, "loss": 2.0811, "step": 40153 }, { "epoch": 3.1468652037617555, "grad_norm": 2.4443678855895996, "learning_rate": 1.0851362684449952e-06, "loss": 1.3635, "step": 40154 }, { "epoch": 3.146943573667712, "grad_norm": 3.302816867828369, "learning_rate": 1.084944457418599e-06, "loss": 0.845, "step": 40155 }, { "epoch": 3.1470219435736677, "grad_norm": 2.570739269256592, "learning_rate": 1.0847526612830618e-06, "loss": 0.9381, "step": 40156 }, { "epoch": 3.1471003134796236, "grad_norm": 3.4841208457946777, "learning_rate": 1.0845608800391144e-06, "loss": 1.0531, "step": 40157 }, { "epoch": 3.14717868338558, "grad_norm": 3.0371530055999756, "learning_rate": 1.0843691136874846e-06, "loss": 0.7942, "step": 40158 }, { "epoch": 3.1472570532915363, "grad_norm": 9.808782577514648, "learning_rate": 1.084177362228907e-06, "loss": 0.5951, "step": 40159 }, { "epoch": 3.147335423197492, "grad_norm": 7.119457721710205, "learning_rate": 1.0839856256641058e-06, "loss": 1.2588, "step": 40160 }, { "epoch": 3.147413793103448, "grad_norm": 3.113806962966919, "learning_rate": 1.0837939039938112e-06, "loss": 1.3487, "step": 40161 }, { "epoch": 3.1474921630094044, "grad_norm": 2.705181360244751, "learning_rate": 1.0836021972187538e-06, "loss": 1.397, "step": 40162 }, { "epoch": 3.1475705329153607, "grad_norm": 5.574914455413818, "learning_rate": 1.083410505339662e-06, "loss": 1.1093, "step": 40163 }, { "epoch": 3.1476489028213166, "grad_norm": 4.436680316925049, "learning_rate": 1.0832188283572647e-06, "loss": 1.2451, "step": 40164 }, { "epoch": 3.147727272727273, "grad_norm": 4.10951042175293, "learning_rate": 1.0830271662722914e-06, "loss": 1.386, "step": 40165 }, { "epoch": 3.147805642633229, "grad_norm": 5.8634257316589355, "learning_rate": 1.0828355190854722e-06, "loss": 1.3215, "step": 40166 }, { "epoch": 3.147884012539185, "grad_norm": 2.7624080181121826, "learning_rate": 1.0826438867975319e-06, "loss": 1.4376, "step": 40167 }, { "epoch": 3.147962382445141, "grad_norm": 2.7535853385925293, "learning_rate": 1.0824522694092033e-06, "loss": 1.3478, "step": 40168 }, { "epoch": 3.1480407523510974, "grad_norm": 3.734663724899292, "learning_rate": 1.0822606669212144e-06, "loss": 1.5701, "step": 40169 }, { "epoch": 3.1481191222570533, "grad_norm": 4.466121196746826, "learning_rate": 1.0820690793342953e-06, "loss": 1.5139, "step": 40170 }, { "epoch": 3.1481974921630096, "grad_norm": 2.2531652450561523, "learning_rate": 1.081877506649171e-06, "loss": 1.2953, "step": 40171 }, { "epoch": 3.1482758620689655, "grad_norm": 2.4542884826660156, "learning_rate": 1.0816859488665715e-06, "loss": 1.123, "step": 40172 }, { "epoch": 3.148354231974922, "grad_norm": 3.1174681186676025, "learning_rate": 1.0814944059872267e-06, "loss": 1.2154, "step": 40173 }, { "epoch": 3.1484326018808777, "grad_norm": 2.473153829574585, "learning_rate": 1.0813028780118667e-06, "loss": 1.3611, "step": 40174 }, { "epoch": 3.148510971786834, "grad_norm": 5.516052722930908, "learning_rate": 1.0811113649412159e-06, "loss": 0.9971, "step": 40175 }, { "epoch": 3.14858934169279, "grad_norm": 4.508073806762695, "learning_rate": 1.0809198667760045e-06, "loss": 1.6518, "step": 40176 }, { "epoch": 3.1486677115987463, "grad_norm": 2.248701333999634, "learning_rate": 1.0807283835169614e-06, "loss": 0.9272, "step": 40177 }, { "epoch": 3.148746081504702, "grad_norm": 4.2002105712890625, "learning_rate": 1.0805369151648142e-06, "loss": 1.1775, "step": 40178 }, { "epoch": 3.1488244514106585, "grad_norm": 9.202118873596191, "learning_rate": 1.0803454617202913e-06, "loss": 1.1478, "step": 40179 }, { "epoch": 3.1489028213166144, "grad_norm": 3.0500781536102295, "learning_rate": 1.0801540231841213e-06, "loss": 1.13, "step": 40180 }, { "epoch": 3.1489811912225707, "grad_norm": 1.8886404037475586, "learning_rate": 1.0799625995570317e-06, "loss": 1.4149, "step": 40181 }, { "epoch": 3.1490595611285266, "grad_norm": 4.857156276702881, "learning_rate": 1.0797711908397513e-06, "loss": 1.293, "step": 40182 }, { "epoch": 3.149137931034483, "grad_norm": 3.220566987991333, "learning_rate": 1.0795797970330075e-06, "loss": 1.2889, "step": 40183 }, { "epoch": 3.149216300940439, "grad_norm": 5.8204755783081055, "learning_rate": 1.0793884181375286e-06, "loss": 1.2802, "step": 40184 }, { "epoch": 3.149294670846395, "grad_norm": 3.341952323913574, "learning_rate": 1.0791970541540436e-06, "loss": 1.2345, "step": 40185 }, { "epoch": 3.149373040752351, "grad_norm": 3.9978761672973633, "learning_rate": 1.0790057050832765e-06, "loss": 1.5785, "step": 40186 }, { "epoch": 3.1494514106583074, "grad_norm": 2.7343268394470215, "learning_rate": 1.078814370925959e-06, "loss": 1.3727, "step": 40187 }, { "epoch": 3.1495297805642632, "grad_norm": 3.567502737045288, "learning_rate": 1.0786230516828178e-06, "loss": 1.6162, "step": 40188 }, { "epoch": 3.1496081504702196, "grad_norm": 3.636066436767578, "learning_rate": 1.078431747354582e-06, "loss": 1.2368, "step": 40189 }, { "epoch": 3.1496865203761755, "grad_norm": 4.776369571685791, "learning_rate": 1.0782404579419752e-06, "loss": 1.2355, "step": 40190 }, { "epoch": 3.149764890282132, "grad_norm": 5.922375679016113, "learning_rate": 1.0780491834457268e-06, "loss": 1.1881, "step": 40191 }, { "epoch": 3.1498432601880877, "grad_norm": 2.960266590118408, "learning_rate": 1.0778579238665677e-06, "loss": 1.2291, "step": 40192 }, { "epoch": 3.149921630094044, "grad_norm": 2.4757745265960693, "learning_rate": 1.0776666792052208e-06, "loss": 1.2691, "step": 40193 }, { "epoch": 3.15, "grad_norm": 3.4196550846099854, "learning_rate": 1.077475449462415e-06, "loss": 1.3516, "step": 40194 }, { "epoch": 3.1500783699059562, "grad_norm": 2.864250898361206, "learning_rate": 1.0772842346388784e-06, "loss": 0.7937, "step": 40195 }, { "epoch": 3.150156739811912, "grad_norm": 2.9735894203186035, "learning_rate": 1.0770930347353376e-06, "loss": 1.1857, "step": 40196 }, { "epoch": 3.1502351097178685, "grad_norm": 6.903688907623291, "learning_rate": 1.0769018497525196e-06, "loss": 1.4265, "step": 40197 }, { "epoch": 3.1503134796238244, "grad_norm": 4.817074775695801, "learning_rate": 1.0767106796911525e-06, "loss": 1.6792, "step": 40198 }, { "epoch": 3.1503918495297807, "grad_norm": 2.5984702110290527, "learning_rate": 1.0765195245519627e-06, "loss": 1.6499, "step": 40199 }, { "epoch": 3.1504702194357366, "grad_norm": 1.8806475400924683, "learning_rate": 1.076328384335677e-06, "loss": 1.0843, "step": 40200 }, { "epoch": 3.150548589341693, "grad_norm": 2.858001470565796, "learning_rate": 1.076137259043023e-06, "loss": 1.2983, "step": 40201 }, { "epoch": 3.150626959247649, "grad_norm": 2.0347959995269775, "learning_rate": 1.0759461486747274e-06, "loss": 1.4127, "step": 40202 }, { "epoch": 3.150705329153605, "grad_norm": 4.612868309020996, "learning_rate": 1.0757550532315187e-06, "loss": 1.5129, "step": 40203 }, { "epoch": 3.150783699059561, "grad_norm": 3.8276636600494385, "learning_rate": 1.0755639727141209e-06, "loss": 1.1255, "step": 40204 }, { "epoch": 3.1508620689655173, "grad_norm": 3.8799853324890137, "learning_rate": 1.0753729071232604e-06, "loss": 1.4716, "step": 40205 }, { "epoch": 3.1509404388714732, "grad_norm": 7.095425128936768, "learning_rate": 1.0751818564596667e-06, "loss": 1.351, "step": 40206 }, { "epoch": 3.1510188087774296, "grad_norm": 2.613997220993042, "learning_rate": 1.0749908207240673e-06, "loss": 1.3413, "step": 40207 }, { "epoch": 3.1510971786833855, "grad_norm": 4.09414005279541, "learning_rate": 1.074799799917185e-06, "loss": 1.3242, "step": 40208 }, { "epoch": 3.151175548589342, "grad_norm": 3.99277400970459, "learning_rate": 1.074608794039748e-06, "loss": 1.3434, "step": 40209 }, { "epoch": 3.1512539184952977, "grad_norm": 2.8576064109802246, "learning_rate": 1.0744178030924817e-06, "loss": 1.2486, "step": 40210 }, { "epoch": 3.151332288401254, "grad_norm": 4.371355056762695, "learning_rate": 1.074226827076117e-06, "loss": 1.4535, "step": 40211 }, { "epoch": 3.15141065830721, "grad_norm": 3.971925973892212, "learning_rate": 1.0740358659913752e-06, "loss": 1.1361, "step": 40212 }, { "epoch": 3.1514890282131662, "grad_norm": 4.728686332702637, "learning_rate": 1.0738449198389844e-06, "loss": 1.2045, "step": 40213 }, { "epoch": 3.151567398119122, "grad_norm": 3.0024192333221436, "learning_rate": 1.0736539886196712e-06, "loss": 1.4727, "step": 40214 }, { "epoch": 3.1516457680250785, "grad_norm": 4.728550910949707, "learning_rate": 1.0734630723341611e-06, "loss": 1.1165, "step": 40215 }, { "epoch": 3.1517241379310343, "grad_norm": 3.5039865970611572, "learning_rate": 1.0732721709831812e-06, "loss": 1.3135, "step": 40216 }, { "epoch": 3.1518025078369907, "grad_norm": 2.3406779766082764, "learning_rate": 1.0730812845674566e-06, "loss": 1.3733, "step": 40217 }, { "epoch": 3.1518808777429466, "grad_norm": 9.288585662841797, "learning_rate": 1.0728904130877155e-06, "loss": 1.777, "step": 40218 }, { "epoch": 3.151959247648903, "grad_norm": 12.303837776184082, "learning_rate": 1.0726995565446797e-06, "loss": 0.6697, "step": 40219 }, { "epoch": 3.152037617554859, "grad_norm": 3.545684814453125, "learning_rate": 1.0725087149390783e-06, "loss": 1.3392, "step": 40220 }, { "epoch": 3.152115987460815, "grad_norm": 7.92611837387085, "learning_rate": 1.0723178882716367e-06, "loss": 1.1335, "step": 40221 }, { "epoch": 3.152194357366771, "grad_norm": 4.3925347328186035, "learning_rate": 1.072127076543082e-06, "loss": 1.2799, "step": 40222 }, { "epoch": 3.1522727272727273, "grad_norm": 5.899825096130371, "learning_rate": 1.0719362797541372e-06, "loss": 0.8211, "step": 40223 }, { "epoch": 3.1523510971786832, "grad_norm": 4.317780494689941, "learning_rate": 1.0717454979055276e-06, "loss": 1.6235, "step": 40224 }, { "epoch": 3.1524294670846396, "grad_norm": 2.8852241039276123, "learning_rate": 1.0715547309979818e-06, "loss": 1.3677, "step": 40225 }, { "epoch": 3.1525078369905954, "grad_norm": 3.3260486125946045, "learning_rate": 1.071363979032225e-06, "loss": 1.1278, "step": 40226 }, { "epoch": 3.152586206896552, "grad_norm": 4.713263511657715, "learning_rate": 1.071173242008981e-06, "loss": 1.1463, "step": 40227 }, { "epoch": 3.1526645768025077, "grad_norm": 13.571237564086914, "learning_rate": 1.0709825199289758e-06, "loss": 1.7602, "step": 40228 }, { "epoch": 3.152742946708464, "grad_norm": 2.4996626377105713, "learning_rate": 1.0707918127929335e-06, "loss": 0.9699, "step": 40229 }, { "epoch": 3.15282131661442, "grad_norm": 2.8742740154266357, "learning_rate": 1.0706011206015842e-06, "loss": 1.0523, "step": 40230 }, { "epoch": 3.152899686520376, "grad_norm": 3.957271099090576, "learning_rate": 1.0704104433556484e-06, "loss": 0.9613, "step": 40231 }, { "epoch": 3.152978056426332, "grad_norm": 2.909846305847168, "learning_rate": 1.0702197810558523e-06, "loss": 1.099, "step": 40232 }, { "epoch": 3.1530564263322884, "grad_norm": 3.0417354106903076, "learning_rate": 1.0700291337029222e-06, "loss": 1.3859, "step": 40233 }, { "epoch": 3.1531347962382443, "grad_norm": 4.146823883056641, "learning_rate": 1.0698385012975827e-06, "loss": 1.163, "step": 40234 }, { "epoch": 3.1532131661442007, "grad_norm": 6.813239574432373, "learning_rate": 1.069647883840559e-06, "loss": 1.498, "step": 40235 }, { "epoch": 3.1532915360501566, "grad_norm": 3.2321417331695557, "learning_rate": 1.069457281332576e-06, "loss": 1.5931, "step": 40236 }, { "epoch": 3.153369905956113, "grad_norm": 2.7333664894104004, "learning_rate": 1.0692666937743584e-06, "loss": 1.1904, "step": 40237 }, { "epoch": 3.1534482758620688, "grad_norm": 3.7431442737579346, "learning_rate": 1.0690761211666317e-06, "loss": 0.901, "step": 40238 }, { "epoch": 3.153526645768025, "grad_norm": 3.5571231842041016, "learning_rate": 1.0688855635101198e-06, "loss": 1.8932, "step": 40239 }, { "epoch": 3.153605015673981, "grad_norm": 3.4073822498321533, "learning_rate": 1.0686950208055486e-06, "loss": 1.3605, "step": 40240 }, { "epoch": 3.1536833855799373, "grad_norm": 2.469093084335327, "learning_rate": 1.0685044930536438e-06, "loss": 1.1857, "step": 40241 }, { "epoch": 3.153761755485893, "grad_norm": 3.479433298110962, "learning_rate": 1.0683139802551252e-06, "loss": 1.5729, "step": 40242 }, { "epoch": 3.1538401253918495, "grad_norm": 4.476428031921387, "learning_rate": 1.0681234824107223e-06, "loss": 1.4757, "step": 40243 }, { "epoch": 3.1539184952978054, "grad_norm": 2.932816982269287, "learning_rate": 1.0679329995211596e-06, "loss": 1.2952, "step": 40244 }, { "epoch": 3.1539968652037618, "grad_norm": 4.008965969085693, "learning_rate": 1.0677425315871588e-06, "loss": 1.7308, "step": 40245 }, { "epoch": 3.154075235109718, "grad_norm": 2.7395777702331543, "learning_rate": 1.0675520786094457e-06, "loss": 1.1637, "step": 40246 }, { "epoch": 3.154153605015674, "grad_norm": 2.5626773834228516, "learning_rate": 1.067361640588745e-06, "loss": 1.177, "step": 40247 }, { "epoch": 3.15423197492163, "grad_norm": 3.4606106281280518, "learning_rate": 1.0671712175257803e-06, "loss": 1.4588, "step": 40248 }, { "epoch": 3.154310344827586, "grad_norm": 6.259932994842529, "learning_rate": 1.0669808094212763e-06, "loss": 1.2237, "step": 40249 }, { "epoch": 3.1543887147335425, "grad_norm": 2.2000842094421387, "learning_rate": 1.0667904162759568e-06, "loss": 1.5339, "step": 40250 }, { "epoch": 3.1544670846394984, "grad_norm": 3.125746488571167, "learning_rate": 1.0666000380905468e-06, "loss": 1.2336, "step": 40251 }, { "epoch": 3.1545454545454543, "grad_norm": 4.456552505493164, "learning_rate": 1.0664096748657698e-06, "loss": 1.4794, "step": 40252 }, { "epoch": 3.1546238244514107, "grad_norm": 2.985232353210449, "learning_rate": 1.0662193266023502e-06, "loss": 1.0318, "step": 40253 }, { "epoch": 3.154702194357367, "grad_norm": 2.633866548538208, "learning_rate": 1.0660289933010114e-06, "loss": 1.344, "step": 40254 }, { "epoch": 3.154780564263323, "grad_norm": 10.967275619506836, "learning_rate": 1.0658386749624777e-06, "loss": 0.492, "step": 40255 }, { "epoch": 3.1548589341692788, "grad_norm": 2.9660630226135254, "learning_rate": 1.0656483715874733e-06, "loss": 1.0382, "step": 40256 }, { "epoch": 3.154937304075235, "grad_norm": 3.428264617919922, "learning_rate": 1.0654580831767215e-06, "loss": 1.0666, "step": 40257 }, { "epoch": 3.1550156739811914, "grad_norm": 2.7496516704559326, "learning_rate": 1.0652678097309467e-06, "loss": 1.3416, "step": 40258 }, { "epoch": 3.1550940438871473, "grad_norm": 6.394022464752197, "learning_rate": 1.0650775512508727e-06, "loss": 1.3808, "step": 40259 }, { "epoch": 3.1551724137931036, "grad_norm": 15.091361999511719, "learning_rate": 1.0648873077372219e-06, "loss": 1.8677, "step": 40260 }, { "epoch": 3.1552507836990595, "grad_norm": 3.8762640953063965, "learning_rate": 1.064697079190717e-06, "loss": 1.4319, "step": 40261 }, { "epoch": 3.155329153605016, "grad_norm": 2.5084078311920166, "learning_rate": 1.0645068656120837e-06, "loss": 1.3458, "step": 40262 }, { "epoch": 3.1554075235109718, "grad_norm": 3.87093448638916, "learning_rate": 1.0643166670020472e-06, "loss": 1.1824, "step": 40263 }, { "epoch": 3.155485893416928, "grad_norm": 3.1214146614074707, "learning_rate": 1.0641264833613268e-06, "loss": 1.059, "step": 40264 }, { "epoch": 3.155564263322884, "grad_norm": 9.117423057556152, "learning_rate": 1.0639363146906483e-06, "loss": 0.7096, "step": 40265 }, { "epoch": 3.1556426332288403, "grad_norm": 7.1379499435424805, "learning_rate": 1.0637461609907323e-06, "loss": 1.5336, "step": 40266 }, { "epoch": 3.155721003134796, "grad_norm": 5.618801116943359, "learning_rate": 1.063556022262307e-06, "loss": 1.5964, "step": 40267 }, { "epoch": 3.1557993730407525, "grad_norm": 6.421687126159668, "learning_rate": 1.0633658985060918e-06, "loss": 1.1382, "step": 40268 }, { "epoch": 3.1558777429467084, "grad_norm": 2.9633374214172363, "learning_rate": 1.06317578972281e-06, "loss": 1.531, "step": 40269 }, { "epoch": 3.1559561128526648, "grad_norm": 4.080637454986572, "learning_rate": 1.0629856959131861e-06, "loss": 1.1503, "step": 40270 }, { "epoch": 3.1560344827586206, "grad_norm": 6.2452192306518555, "learning_rate": 1.062795617077942e-06, "loss": 1.1379, "step": 40271 }, { "epoch": 3.156112852664577, "grad_norm": 3.680178165435791, "learning_rate": 1.0626055532178015e-06, "loss": 1.5903, "step": 40272 }, { "epoch": 3.156191222570533, "grad_norm": 2.6283133029937744, "learning_rate": 1.0624155043334865e-06, "loss": 1.6017, "step": 40273 }, { "epoch": 3.156269592476489, "grad_norm": 2.277891159057617, "learning_rate": 1.0622254704257224e-06, "loss": 1.4858, "step": 40274 }, { "epoch": 3.156347962382445, "grad_norm": 4.956921100616455, "learning_rate": 1.0620354514952268e-06, "loss": 1.3339, "step": 40275 }, { "epoch": 3.1564263322884014, "grad_norm": 2.648510694503784, "learning_rate": 1.0618454475427276e-06, "loss": 1.51, "step": 40276 }, { "epoch": 3.1565047021943573, "grad_norm": 6.528627872467041, "learning_rate": 1.0616554585689448e-06, "loss": 1.15, "step": 40277 }, { "epoch": 3.1565830721003136, "grad_norm": 3.5580697059631348, "learning_rate": 1.0614654845746037e-06, "loss": 1.1246, "step": 40278 }, { "epoch": 3.1566614420062695, "grad_norm": 2.4534900188446045, "learning_rate": 1.0612755255604234e-06, "loss": 1.0056, "step": 40279 }, { "epoch": 3.156739811912226, "grad_norm": 4.1547322273254395, "learning_rate": 1.061085581527127e-06, "loss": 1.3479, "step": 40280 }, { "epoch": 3.1568181818181817, "grad_norm": 2.4045615196228027, "learning_rate": 1.0608956524754387e-06, "loss": 1.3651, "step": 40281 }, { "epoch": 3.156896551724138, "grad_norm": 11.0392427444458, "learning_rate": 1.0607057384060815e-06, "loss": 1.286, "step": 40282 }, { "epoch": 3.156974921630094, "grad_norm": 3.019834518432617, "learning_rate": 1.060515839319775e-06, "loss": 0.8244, "step": 40283 }, { "epoch": 3.1570532915360503, "grad_norm": 3.476821184158325, "learning_rate": 1.0603259552172435e-06, "loss": 1.3956, "step": 40284 }, { "epoch": 3.157131661442006, "grad_norm": 2.0764830112457275, "learning_rate": 1.060136086099206e-06, "loss": 1.3705, "step": 40285 }, { "epoch": 3.1572100313479625, "grad_norm": 3.968013286590576, "learning_rate": 1.0599462319663906e-06, "loss": 1.2857, "step": 40286 }, { "epoch": 3.1572884012539184, "grad_norm": 3.5551297664642334, "learning_rate": 1.0597563928195147e-06, "loss": 1.2599, "step": 40287 }, { "epoch": 3.1573667711598747, "grad_norm": 2.1191725730895996, "learning_rate": 1.0595665686593015e-06, "loss": 1.3718, "step": 40288 }, { "epoch": 3.1574451410658306, "grad_norm": 4.226146697998047, "learning_rate": 1.0593767594864729e-06, "loss": 1.5874, "step": 40289 }, { "epoch": 3.157523510971787, "grad_norm": 2.8315927982330322, "learning_rate": 1.059186965301751e-06, "loss": 1.1872, "step": 40290 }, { "epoch": 3.157601880877743, "grad_norm": 2.0362071990966797, "learning_rate": 1.0589971861058585e-06, "loss": 1.3501, "step": 40291 }, { "epoch": 3.157680250783699, "grad_norm": 2.408869504928589, "learning_rate": 1.058807421899516e-06, "loss": 1.3388, "step": 40292 }, { "epoch": 3.157758620689655, "grad_norm": 3.7170896530151367, "learning_rate": 1.0586176726834468e-06, "loss": 1.4501, "step": 40293 }, { "epoch": 3.1578369905956114, "grad_norm": 3.231278896331787, "learning_rate": 1.0584279384583696e-06, "loss": 1.488, "step": 40294 }, { "epoch": 3.1579153605015673, "grad_norm": 6.7796101570129395, "learning_rate": 1.0582382192250085e-06, "loss": 0.8572, "step": 40295 }, { "epoch": 3.1579937304075236, "grad_norm": 4.307601451873779, "learning_rate": 1.0580485149840853e-06, "loss": 1.5132, "step": 40296 }, { "epoch": 3.1580721003134795, "grad_norm": 2.939662218093872, "learning_rate": 1.0578588257363214e-06, "loss": 1.4261, "step": 40297 }, { "epoch": 3.158150470219436, "grad_norm": 3.3046157360076904, "learning_rate": 1.0576691514824372e-06, "loss": 1.1609, "step": 40298 }, { "epoch": 3.1582288401253917, "grad_norm": 3.529569149017334, "learning_rate": 1.057479492223153e-06, "loss": 1.1214, "step": 40299 }, { "epoch": 3.158307210031348, "grad_norm": 2.6930110454559326, "learning_rate": 1.0572898479591942e-06, "loss": 1.3059, "step": 40300 }, { "epoch": 3.158385579937304, "grad_norm": 3.048443078994751, "learning_rate": 1.0571002186912788e-06, "loss": 1.357, "step": 40301 }, { "epoch": 3.1584639498432603, "grad_norm": 3.8936479091644287, "learning_rate": 1.056910604420129e-06, "loss": 1.1082, "step": 40302 }, { "epoch": 3.158542319749216, "grad_norm": 5.227937698364258, "learning_rate": 1.0567210051464654e-06, "loss": 0.9738, "step": 40303 }, { "epoch": 3.1586206896551725, "grad_norm": 2.4819321632385254, "learning_rate": 1.0565314208710097e-06, "loss": 1.2685, "step": 40304 }, { "epoch": 3.1586990595611284, "grad_norm": 3.1945881843566895, "learning_rate": 1.0563418515944833e-06, "loss": 1.4145, "step": 40305 }, { "epoch": 3.1587774294670847, "grad_norm": 2.9416844844818115, "learning_rate": 1.0561522973176075e-06, "loss": 1.1082, "step": 40306 }, { "epoch": 3.1588557993730406, "grad_norm": 4.924911975860596, "learning_rate": 1.055962758041102e-06, "loss": 1.2598, "step": 40307 }, { "epoch": 3.158934169278997, "grad_norm": 5.979518890380859, "learning_rate": 1.0557732337656885e-06, "loss": 2.2876, "step": 40308 }, { "epoch": 3.159012539184953, "grad_norm": 6.486942291259766, "learning_rate": 1.0555837244920875e-06, "loss": 1.4179, "step": 40309 }, { "epoch": 3.159090909090909, "grad_norm": 3.353475332260132, "learning_rate": 1.0553942302210202e-06, "loss": 1.2913, "step": 40310 }, { "epoch": 3.159169278996865, "grad_norm": 3.8078083992004395, "learning_rate": 1.055204750953207e-06, "loss": 1.3415, "step": 40311 }, { "epoch": 3.1592476489028214, "grad_norm": 3.61205792427063, "learning_rate": 1.0550152866893703e-06, "loss": 1.1297, "step": 40312 }, { "epoch": 3.1593260188087773, "grad_norm": 4.775285720825195, "learning_rate": 1.0548258374302267e-06, "loss": 1.8361, "step": 40313 }, { "epoch": 3.1594043887147336, "grad_norm": 5.021810531616211, "learning_rate": 1.0546364031765005e-06, "loss": 1.3545, "step": 40314 }, { "epoch": 3.1594827586206895, "grad_norm": 5.739391803741455, "learning_rate": 1.0544469839289123e-06, "loss": 0.9225, "step": 40315 }, { "epoch": 3.159561128526646, "grad_norm": 5.103228569030762, "learning_rate": 1.05425757968818e-06, "loss": 1.3042, "step": 40316 }, { "epoch": 3.1596394984326017, "grad_norm": 3.979994773864746, "learning_rate": 1.0540681904550248e-06, "loss": 1.5538, "step": 40317 }, { "epoch": 3.159717868338558, "grad_norm": 9.664887428283691, "learning_rate": 1.0538788162301662e-06, "loss": 1.4388, "step": 40318 }, { "epoch": 3.159796238244514, "grad_norm": 3.0944266319274902, "learning_rate": 1.0536894570143286e-06, "loss": 1.1245, "step": 40319 }, { "epoch": 3.1598746081504703, "grad_norm": 7.934520244598389, "learning_rate": 1.0535001128082278e-06, "loss": 1.2519, "step": 40320 }, { "epoch": 3.159952978056426, "grad_norm": 3.583231210708618, "learning_rate": 1.0533107836125856e-06, "loss": 1.534, "step": 40321 }, { "epoch": 3.1600313479623825, "grad_norm": 4.857594013214111, "learning_rate": 1.0531214694281216e-06, "loss": 1.37, "step": 40322 }, { "epoch": 3.1601097178683384, "grad_norm": 3.212697744369507, "learning_rate": 1.0529321702555568e-06, "loss": 1.3282, "step": 40323 }, { "epoch": 3.1601880877742947, "grad_norm": 3.159151554107666, "learning_rate": 1.0527428860956108e-06, "loss": 1.4123, "step": 40324 }, { "epoch": 3.1602664576802506, "grad_norm": 3.5638492107391357, "learning_rate": 1.052553616949003e-06, "loss": 1.1982, "step": 40325 }, { "epoch": 3.160344827586207, "grad_norm": 2.515950918197632, "learning_rate": 1.0523643628164542e-06, "loss": 1.0165, "step": 40326 }, { "epoch": 3.160423197492163, "grad_norm": 4.942446231842041, "learning_rate": 1.0521751236986833e-06, "loss": 0.9541, "step": 40327 }, { "epoch": 3.160501567398119, "grad_norm": 5.392895221710205, "learning_rate": 1.0519858995964105e-06, "loss": 1.5869, "step": 40328 }, { "epoch": 3.160579937304075, "grad_norm": 4.128900051116943, "learning_rate": 1.051796690510356e-06, "loss": 1.5568, "step": 40329 }, { "epoch": 3.1606583072100314, "grad_norm": 3.4315993785858154, "learning_rate": 1.05160749644124e-06, "loss": 1.4148, "step": 40330 }, { "epoch": 3.1607366771159873, "grad_norm": 2.9030728340148926, "learning_rate": 1.0514183173897796e-06, "loss": 1.2719, "step": 40331 }, { "epoch": 3.1608150470219436, "grad_norm": 2.604822874069214, "learning_rate": 1.0512291533566942e-06, "loss": 1.0636, "step": 40332 }, { "epoch": 3.1608934169278995, "grad_norm": 4.922613143920898, "learning_rate": 1.0510400043427072e-06, "loss": 0.8791, "step": 40333 }, { "epoch": 3.160971786833856, "grad_norm": 2.804112195968628, "learning_rate": 1.050850870348536e-06, "loss": 1.0884, "step": 40334 }, { "epoch": 3.1610501567398117, "grad_norm": 2.6814205646514893, "learning_rate": 1.050661751374899e-06, "loss": 1.2598, "step": 40335 }, { "epoch": 3.161128526645768, "grad_norm": 3.0390474796295166, "learning_rate": 1.0504726474225158e-06, "loss": 1.4084, "step": 40336 }, { "epoch": 3.161206896551724, "grad_norm": 3.168780565261841, "learning_rate": 1.050283558492105e-06, "loss": 0.9242, "step": 40337 }, { "epoch": 3.1612852664576803, "grad_norm": 2.3386952877044678, "learning_rate": 1.0500944845843892e-06, "loss": 1.1496, "step": 40338 }, { "epoch": 3.161363636363636, "grad_norm": 3.7035136222839355, "learning_rate": 1.049905425700084e-06, "loss": 1.1536, "step": 40339 }, { "epoch": 3.1614420062695925, "grad_norm": 4.228540897369385, "learning_rate": 1.0497163818399097e-06, "loss": 1.57, "step": 40340 }, { "epoch": 3.161520376175549, "grad_norm": 2.648841619491577, "learning_rate": 1.0495273530045853e-06, "loss": 1.325, "step": 40341 }, { "epoch": 3.1615987460815047, "grad_norm": 2.7703049182891846, "learning_rate": 1.0493383391948298e-06, "loss": 1.2926, "step": 40342 }, { "epoch": 3.1616771159874606, "grad_norm": 4.599079132080078, "learning_rate": 1.0491493404113616e-06, "loss": 1.2547, "step": 40343 }, { "epoch": 3.161755485893417, "grad_norm": 5.846985816955566, "learning_rate": 1.0489603566549007e-06, "loss": 1.3549, "step": 40344 }, { "epoch": 3.1618338557993733, "grad_norm": 3.473066806793213, "learning_rate": 1.0487713879261652e-06, "loss": 1.5605, "step": 40345 }, { "epoch": 3.161912225705329, "grad_norm": 3.412107229232788, "learning_rate": 1.0485824342258732e-06, "loss": 1.1419, "step": 40346 }, { "epoch": 3.161990595611285, "grad_norm": 3.8356127738952637, "learning_rate": 1.0483934955547448e-06, "loss": 1.2579, "step": 40347 }, { "epoch": 3.1620689655172414, "grad_norm": 2.7722110748291016, "learning_rate": 1.0482045719134977e-06, "loss": 1.4676, "step": 40348 }, { "epoch": 3.1621473354231977, "grad_norm": 6.403844356536865, "learning_rate": 1.0480156633028515e-06, "loss": 0.9369, "step": 40349 }, { "epoch": 3.1622257053291536, "grad_norm": 2.5125691890716553, "learning_rate": 1.0478267697235217e-06, "loss": 1.0115, "step": 40350 }, { "epoch": 3.16230407523511, "grad_norm": 2.754545211791992, "learning_rate": 1.0476378911762297e-06, "loss": 1.4471, "step": 40351 }, { "epoch": 3.162382445141066, "grad_norm": 5.694187641143799, "learning_rate": 1.0474490276616932e-06, "loss": 1.4541, "step": 40352 }, { "epoch": 3.162460815047022, "grad_norm": 3.2800967693328857, "learning_rate": 1.0472601791806325e-06, "loss": 1.1818, "step": 40353 }, { "epoch": 3.162539184952978, "grad_norm": 2.9030613899230957, "learning_rate": 1.0470713457337616e-06, "loss": 1.3234, "step": 40354 }, { "epoch": 3.1626175548589344, "grad_norm": 3.0768938064575195, "learning_rate": 1.0468825273218002e-06, "loss": 1.3607, "step": 40355 }, { "epoch": 3.1626959247648903, "grad_norm": 2.2172489166259766, "learning_rate": 1.0466937239454695e-06, "loss": 1.3446, "step": 40356 }, { "epoch": 3.1627742946708466, "grad_norm": 4.069037437438965, "learning_rate": 1.046504935605484e-06, "loss": 1.3355, "step": 40357 }, { "epoch": 3.1628526645768025, "grad_norm": 3.2912135124206543, "learning_rate": 1.0463161623025636e-06, "loss": 1.8186, "step": 40358 }, { "epoch": 3.162931034482759, "grad_norm": 3.2215116024017334, "learning_rate": 1.0461274040374252e-06, "loss": 1.0059, "step": 40359 }, { "epoch": 3.1630094043887147, "grad_norm": 4.183558940887451, "learning_rate": 1.045938660810788e-06, "loss": 1.1723, "step": 40360 }, { "epoch": 3.163087774294671, "grad_norm": 4.550976753234863, "learning_rate": 1.0457499326233683e-06, "loss": 0.695, "step": 40361 }, { "epoch": 3.163166144200627, "grad_norm": 2.7992281913757324, "learning_rate": 1.0455612194758852e-06, "loss": 1.1455, "step": 40362 }, { "epoch": 3.1632445141065832, "grad_norm": 5.666451454162598, "learning_rate": 1.0453725213690564e-06, "loss": 1.3925, "step": 40363 }, { "epoch": 3.163322884012539, "grad_norm": 3.359342575073242, "learning_rate": 1.0451838383035988e-06, "loss": 1.2003, "step": 40364 }, { "epoch": 3.1634012539184955, "grad_norm": 2.624373197555542, "learning_rate": 1.0449951702802313e-06, "loss": 1.4571, "step": 40365 }, { "epoch": 3.1634796238244514, "grad_norm": 3.0674490928649902, "learning_rate": 1.0448065172996703e-06, "loss": 1.3541, "step": 40366 }, { "epoch": 3.1635579937304077, "grad_norm": 2.6378724575042725, "learning_rate": 1.0446178793626333e-06, "loss": 1.2847, "step": 40367 }, { "epoch": 3.1636363636363636, "grad_norm": 4.155269622802734, "learning_rate": 1.0444292564698404e-06, "loss": 0.7266, "step": 40368 }, { "epoch": 3.16371473354232, "grad_norm": 3.3814239501953125, "learning_rate": 1.0442406486220036e-06, "loss": 1.2576, "step": 40369 }, { "epoch": 3.163793103448276, "grad_norm": 2.6720361709594727, "learning_rate": 1.0440520558198452e-06, "loss": 1.3569, "step": 40370 }, { "epoch": 3.163871473354232, "grad_norm": 2.494640350341797, "learning_rate": 1.0438634780640826e-06, "loss": 1.3067, "step": 40371 }, { "epoch": 3.163949843260188, "grad_norm": 2.323213577270508, "learning_rate": 1.0436749153554292e-06, "loss": 1.1905, "step": 40372 }, { "epoch": 3.1640282131661444, "grad_norm": 4.958292007446289, "learning_rate": 1.043486367694605e-06, "loss": 1.2483, "step": 40373 }, { "epoch": 3.1641065830721002, "grad_norm": 2.777318239212036, "learning_rate": 1.0432978350823247e-06, "loss": 1.5707, "step": 40374 }, { "epoch": 3.1641849529780566, "grad_norm": 4.118167877197266, "learning_rate": 1.0431093175193102e-06, "loss": 1.2542, "step": 40375 }, { "epoch": 3.1642633228840125, "grad_norm": 3.9411816596984863, "learning_rate": 1.0429208150062736e-06, "loss": 0.9823, "step": 40376 }, { "epoch": 3.164341692789969, "grad_norm": 5.729244709014893, "learning_rate": 1.0427323275439343e-06, "loss": 0.957, "step": 40377 }, { "epoch": 3.1644200626959247, "grad_norm": 4.533440589904785, "learning_rate": 1.0425438551330087e-06, "loss": 1.3592, "step": 40378 }, { "epoch": 3.164498432601881, "grad_norm": 2.3733057975769043, "learning_rate": 1.042355397774213e-06, "loss": 1.1252, "step": 40379 }, { "epoch": 3.164576802507837, "grad_norm": 3.9269332885742188, "learning_rate": 1.042166955468265e-06, "loss": 1.267, "step": 40380 }, { "epoch": 3.1646551724137932, "grad_norm": 3.763378143310547, "learning_rate": 1.0419785282158807e-06, "loss": 1.452, "step": 40381 }, { "epoch": 3.164733542319749, "grad_norm": 3.117485284805298, "learning_rate": 1.0417901160177774e-06, "loss": 1.4425, "step": 40382 }, { "epoch": 3.1648119122257055, "grad_norm": 5.557454586029053, "learning_rate": 1.0416017188746714e-06, "loss": 1.2957, "step": 40383 }, { "epoch": 3.1648902821316613, "grad_norm": 5.958818435668945, "learning_rate": 1.0414133367872792e-06, "loss": 1.1044, "step": 40384 }, { "epoch": 3.1649686520376177, "grad_norm": 6.032312393188477, "learning_rate": 1.0412249697563175e-06, "loss": 1.43, "step": 40385 }, { "epoch": 3.1650470219435736, "grad_norm": 7.421419620513916, "learning_rate": 1.0410366177825037e-06, "loss": 1.5742, "step": 40386 }, { "epoch": 3.16512539184953, "grad_norm": 3.0330758094787598, "learning_rate": 1.0408482808665527e-06, "loss": 1.2341, "step": 40387 }, { "epoch": 3.165203761755486, "grad_norm": 11.112208366394043, "learning_rate": 1.0406599590091793e-06, "loss": 1.3388, "step": 40388 }, { "epoch": 3.165282131661442, "grad_norm": 1.9286611080169678, "learning_rate": 1.040471652211103e-06, "loss": 1.2249, "step": 40389 }, { "epoch": 3.165360501567398, "grad_norm": 3.36657977104187, "learning_rate": 1.04028336047304e-06, "loss": 1.1886, "step": 40390 }, { "epoch": 3.1654388714733543, "grad_norm": 2.3277833461761475, "learning_rate": 1.040095083795704e-06, "loss": 1.2102, "step": 40391 }, { "epoch": 3.1655172413793102, "grad_norm": 4.757965087890625, "learning_rate": 1.0399068221798132e-06, "loss": 1.683, "step": 40392 }, { "epoch": 3.1655956112852666, "grad_norm": 2.8202974796295166, "learning_rate": 1.0397185756260803e-06, "loss": 1.5629, "step": 40393 }, { "epoch": 3.1656739811912225, "grad_norm": 12.394523620605469, "learning_rate": 1.0395303441352272e-06, "loss": 1.3934, "step": 40394 }, { "epoch": 3.165752351097179, "grad_norm": 2.777942657470703, "learning_rate": 1.0393421277079651e-06, "loss": 1.3082, "step": 40395 }, { "epoch": 3.1658307210031347, "grad_norm": 4.031477451324463, "learning_rate": 1.0391539263450106e-06, "loss": 0.8058, "step": 40396 }, { "epoch": 3.165909090909091, "grad_norm": 3.587594747543335, "learning_rate": 1.038965740047081e-06, "loss": 1.285, "step": 40397 }, { "epoch": 3.165987460815047, "grad_norm": 3.374868631362915, "learning_rate": 1.0387775688148904e-06, "loss": 1.7789, "step": 40398 }, { "epoch": 3.1660658307210032, "grad_norm": 3.2336525917053223, "learning_rate": 1.038589412649156e-06, "loss": 1.3241, "step": 40399 }, { "epoch": 3.166144200626959, "grad_norm": 3.057682991027832, "learning_rate": 1.0384012715505925e-06, "loss": 1.4046, "step": 40400 }, { "epoch": 3.1662225705329154, "grad_norm": 7.661880016326904, "learning_rate": 1.0382131455199173e-06, "loss": 1.6916, "step": 40401 }, { "epoch": 3.1663009404388713, "grad_norm": 3.9941701889038086, "learning_rate": 1.0380250345578413e-06, "loss": 0.653, "step": 40402 }, { "epoch": 3.1663793103448277, "grad_norm": 3.7275478839874268, "learning_rate": 1.0378369386650843e-06, "loss": 0.9283, "step": 40403 }, { "epoch": 3.1664576802507836, "grad_norm": 3.813397169113159, "learning_rate": 1.037648857842361e-06, "loss": 0.8154, "step": 40404 }, { "epoch": 3.16653605015674, "grad_norm": 2.702606439590454, "learning_rate": 1.0374607920903874e-06, "loss": 1.3502, "step": 40405 }, { "epoch": 3.166614420062696, "grad_norm": 3.2632617950439453, "learning_rate": 1.0372727414098766e-06, "loss": 1.2062, "step": 40406 }, { "epoch": 3.166692789968652, "grad_norm": 2.5231235027313232, "learning_rate": 1.0370847058015431e-06, "loss": 1.3016, "step": 40407 }, { "epoch": 3.166771159874608, "grad_norm": 3.4568166732788086, "learning_rate": 1.0368966852661055e-06, "loss": 1.4384, "step": 40408 }, { "epoch": 3.1668495297805643, "grad_norm": 3.0120973587036133, "learning_rate": 1.0367086798042785e-06, "loss": 1.3487, "step": 40409 }, { "epoch": 3.16692789968652, "grad_norm": 3.5155134201049805, "learning_rate": 1.0365206894167744e-06, "loss": 1.3614, "step": 40410 }, { "epoch": 3.1670062695924766, "grad_norm": 3.1766977310180664, "learning_rate": 1.0363327141043106e-06, "loss": 1.4825, "step": 40411 }, { "epoch": 3.1670846394984324, "grad_norm": 5.14375638961792, "learning_rate": 1.0361447538676005e-06, "loss": 1.2514, "step": 40412 }, { "epoch": 3.1671630094043888, "grad_norm": 3.879774332046509, "learning_rate": 1.0359568087073601e-06, "loss": 1.4714, "step": 40413 }, { "epoch": 3.1672413793103447, "grad_norm": 4.463043689727783, "learning_rate": 1.0357688786243042e-06, "loss": 1.5648, "step": 40414 }, { "epoch": 3.167319749216301, "grad_norm": 4.2958831787109375, "learning_rate": 1.0355809636191472e-06, "loss": 1.3687, "step": 40415 }, { "epoch": 3.167398119122257, "grad_norm": 3.3725392818450928, "learning_rate": 1.035393063692604e-06, "loss": 1.1445, "step": 40416 }, { "epoch": 3.167476489028213, "grad_norm": 7.12199592590332, "learning_rate": 1.0352051788453893e-06, "loss": 2.0393, "step": 40417 }, { "epoch": 3.167554858934169, "grad_norm": 3.4850480556488037, "learning_rate": 1.0350173090782174e-06, "loss": 1.2564, "step": 40418 }, { "epoch": 3.1676332288401254, "grad_norm": 2.6741344928741455, "learning_rate": 1.034829454391803e-06, "loss": 1.2842, "step": 40419 }, { "epoch": 3.1677115987460813, "grad_norm": 3.986346483230591, "learning_rate": 1.034641614786862e-06, "loss": 1.6959, "step": 40420 }, { "epoch": 3.1677899686520377, "grad_norm": 2.131462335586548, "learning_rate": 1.0344537902641056e-06, "loss": 1.4186, "step": 40421 }, { "epoch": 3.1678683385579935, "grad_norm": 4.103757381439209, "learning_rate": 1.0342659808242506e-06, "loss": 1.0209, "step": 40422 }, { "epoch": 3.16794670846395, "grad_norm": 5.323506832122803, "learning_rate": 1.0340781864680116e-06, "loss": 1.6646, "step": 40423 }, { "epoch": 3.1680250783699058, "grad_norm": 3.2956881523132324, "learning_rate": 1.0338904071961032e-06, "loss": 1.09, "step": 40424 }, { "epoch": 3.168103448275862, "grad_norm": 3.8368091583251953, "learning_rate": 1.0337026430092368e-06, "loss": 1.3385, "step": 40425 }, { "epoch": 3.168181818181818, "grad_norm": 3.585374116897583, "learning_rate": 1.0335148939081274e-06, "loss": 1.7426, "step": 40426 }, { "epoch": 3.1682601880877743, "grad_norm": 3.588819980621338, "learning_rate": 1.033327159893493e-06, "loss": 1.0021, "step": 40427 }, { "epoch": 3.16833855799373, "grad_norm": 3.758188247680664, "learning_rate": 1.0331394409660428e-06, "loss": 1.5603, "step": 40428 }, { "epoch": 3.1684169278996865, "grad_norm": 4.395136833190918, "learning_rate": 1.032951737126493e-06, "loss": 1.6561, "step": 40429 }, { "epoch": 3.1684952978056424, "grad_norm": 2.3540380001068115, "learning_rate": 1.032764048375557e-06, "loss": 1.3672, "step": 40430 }, { "epoch": 3.1685736677115988, "grad_norm": 4.012157917022705, "learning_rate": 1.032576374713949e-06, "loss": 1.5381, "step": 40431 }, { "epoch": 3.1686520376175547, "grad_norm": 4.087151050567627, "learning_rate": 1.0323887161423824e-06, "loss": 2.0664, "step": 40432 }, { "epoch": 3.168730407523511, "grad_norm": 4.484863758087158, "learning_rate": 1.0322010726615716e-06, "loss": 0.9654, "step": 40433 }, { "epoch": 3.168808777429467, "grad_norm": 2.397770643234253, "learning_rate": 1.0320134442722297e-06, "loss": 1.3348, "step": 40434 }, { "epoch": 3.168887147335423, "grad_norm": 6.123551368713379, "learning_rate": 1.0318258309750706e-06, "loss": 0.9108, "step": 40435 }, { "epoch": 3.1689655172413795, "grad_norm": 9.398177146911621, "learning_rate": 1.031638232770808e-06, "loss": 1.28, "step": 40436 }, { "epoch": 3.1690438871473354, "grad_norm": 5.128931522369385, "learning_rate": 1.031450649660155e-06, "loss": 1.5218, "step": 40437 }, { "epoch": 3.1691222570532913, "grad_norm": 4.205221176147461, "learning_rate": 1.0312630816438257e-06, "loss": 1.4294, "step": 40438 }, { "epoch": 3.1692006269592476, "grad_norm": 3.184906482696533, "learning_rate": 1.0310755287225332e-06, "loss": 1.3829, "step": 40439 }, { "epoch": 3.169278996865204, "grad_norm": 2.8104169368743896, "learning_rate": 1.0308879908969905e-06, "loss": 1.4903, "step": 40440 }, { "epoch": 3.16935736677116, "grad_norm": 3.7878730297088623, "learning_rate": 1.030700468167911e-06, "loss": 1.1608, "step": 40441 }, { "epoch": 3.1694357366771158, "grad_norm": 2.924720287322998, "learning_rate": 1.0305129605360104e-06, "loss": 1.0543, "step": 40442 }, { "epoch": 3.169514106583072, "grad_norm": 4.245306968688965, "learning_rate": 1.0303254680019975e-06, "loss": 1.465, "step": 40443 }, { "epoch": 3.1695924764890284, "grad_norm": 3.2550265789031982, "learning_rate": 1.0301379905665876e-06, "loss": 1.1615, "step": 40444 }, { "epoch": 3.1696708463949843, "grad_norm": 2.736661195755005, "learning_rate": 1.0299505282304927e-06, "loss": 1.1981, "step": 40445 }, { "epoch": 3.1697492163009406, "grad_norm": 4.262425899505615, "learning_rate": 1.0297630809944292e-06, "loss": 1.5512, "step": 40446 }, { "epoch": 3.1698275862068965, "grad_norm": 3.4254307746887207, "learning_rate": 1.0295756488591068e-06, "loss": 1.2159, "step": 40447 }, { "epoch": 3.169905956112853, "grad_norm": 3.375570774078369, "learning_rate": 1.029388231825239e-06, "loss": 1.3528, "step": 40448 }, { "epoch": 3.1699843260188088, "grad_norm": 2.2764813899993896, "learning_rate": 1.029200829893539e-06, "loss": 1.4317, "step": 40449 }, { "epoch": 3.170062695924765, "grad_norm": 3.376253366470337, "learning_rate": 1.0290134430647196e-06, "loss": 1.1654, "step": 40450 }, { "epoch": 3.170141065830721, "grad_norm": 3.4294936656951904, "learning_rate": 1.0288260713394932e-06, "loss": 1.4922, "step": 40451 }, { "epoch": 3.1702194357366773, "grad_norm": 2.1514439582824707, "learning_rate": 1.0286387147185728e-06, "loss": 1.2843, "step": 40452 }, { "epoch": 3.170297805642633, "grad_norm": 1.9566980600357056, "learning_rate": 1.0284513732026714e-06, "loss": 1.2169, "step": 40453 }, { "epoch": 3.1703761755485895, "grad_norm": 2.5593888759613037, "learning_rate": 1.0282640467925003e-06, "loss": 1.2532, "step": 40454 }, { "epoch": 3.1704545454545454, "grad_norm": 5.215372085571289, "learning_rate": 1.0280767354887733e-06, "loss": 1.4658, "step": 40455 }, { "epoch": 3.1705329153605017, "grad_norm": 20.39191436767578, "learning_rate": 1.0278894392922016e-06, "loss": 0.7731, "step": 40456 }, { "epoch": 3.1706112852664576, "grad_norm": 4.207407474517822, "learning_rate": 1.0277021582035002e-06, "loss": 1.3796, "step": 40457 }, { "epoch": 3.170689655172414, "grad_norm": 2.8892548084259033, "learning_rate": 1.027514892223377e-06, "loss": 1.5621, "step": 40458 }, { "epoch": 3.17076802507837, "grad_norm": 4.609220504760742, "learning_rate": 1.0273276413525473e-06, "loss": 1.6309, "step": 40459 }, { "epoch": 3.170846394984326, "grad_norm": 9.420246124267578, "learning_rate": 1.0271404055917233e-06, "loss": 0.6799, "step": 40460 }, { "epoch": 3.170924764890282, "grad_norm": 2.5748820304870605, "learning_rate": 1.026953184941618e-06, "loss": 1.3914, "step": 40461 }, { "epoch": 3.1710031347962384, "grad_norm": 3.193399667739868, "learning_rate": 1.0267659794029406e-06, "loss": 1.3025, "step": 40462 }, { "epoch": 3.1710815047021943, "grad_norm": 3.23748779296875, "learning_rate": 1.0265787889764034e-06, "loss": 1.6406, "step": 40463 }, { "epoch": 3.1711598746081506, "grad_norm": 3.6822807788848877, "learning_rate": 1.0263916136627211e-06, "loss": 1.6281, "step": 40464 }, { "epoch": 3.1712382445141065, "grad_norm": 2.6653945446014404, "learning_rate": 1.0262044534626053e-06, "loss": 1.4325, "step": 40465 }, { "epoch": 3.171316614420063, "grad_norm": 2.4985201358795166, "learning_rate": 1.026017308376765e-06, "loss": 1.4368, "step": 40466 }, { "epoch": 3.1713949843260187, "grad_norm": 2.1743252277374268, "learning_rate": 1.0258301784059138e-06, "loss": 1.5383, "step": 40467 }, { "epoch": 3.171473354231975, "grad_norm": 2.8016815185546875, "learning_rate": 1.0256430635507635e-06, "loss": 1.2621, "step": 40468 }, { "epoch": 3.171551724137931, "grad_norm": 3.174513101577759, "learning_rate": 1.025455963812026e-06, "loss": 1.6215, "step": 40469 }, { "epoch": 3.1716300940438873, "grad_norm": 2.0763633251190186, "learning_rate": 1.025268879190412e-06, "loss": 0.7315, "step": 40470 }, { "epoch": 3.171708463949843, "grad_norm": 4.486451148986816, "learning_rate": 1.0250818096866338e-06, "loss": 1.1652, "step": 40471 }, { "epoch": 3.1717868338557995, "grad_norm": 2.8964810371398926, "learning_rate": 1.0248947553014032e-06, "loss": 1.4431, "step": 40472 }, { "epoch": 3.1718652037617554, "grad_norm": 4.873043537139893, "learning_rate": 1.0247077160354307e-06, "loss": 1.137, "step": 40473 }, { "epoch": 3.1719435736677117, "grad_norm": 3.827636480331421, "learning_rate": 1.0245206918894279e-06, "loss": 1.4077, "step": 40474 }, { "epoch": 3.1720219435736676, "grad_norm": 3.413363456726074, "learning_rate": 1.024333682864107e-06, "loss": 1.0322, "step": 40475 }, { "epoch": 3.172100313479624, "grad_norm": 2.6590733528137207, "learning_rate": 1.0241466889601803e-06, "loss": 1.2971, "step": 40476 }, { "epoch": 3.17217868338558, "grad_norm": 2.8935763835906982, "learning_rate": 1.0239597101783545e-06, "loss": 1.0316, "step": 40477 }, { "epoch": 3.172257053291536, "grad_norm": 2.9021084308624268, "learning_rate": 1.0237727465193448e-06, "loss": 1.3821, "step": 40478 }, { "epoch": 3.172335423197492, "grad_norm": 4.595056056976318, "learning_rate": 1.0235857979838615e-06, "loss": 1.6746, "step": 40479 }, { "epoch": 3.1724137931034484, "grad_norm": 6.284038066864014, "learning_rate": 1.0233988645726166e-06, "loss": 1.65, "step": 40480 }, { "epoch": 3.1724921630094043, "grad_norm": 3.588101387023926, "learning_rate": 1.0232119462863189e-06, "loss": 1.5752, "step": 40481 }, { "epoch": 3.1725705329153606, "grad_norm": 3.6665024757385254, "learning_rate": 1.0230250431256784e-06, "loss": 1.1723, "step": 40482 }, { "epoch": 3.1726489028213165, "grad_norm": 7.161147117614746, "learning_rate": 1.0228381550914114e-06, "loss": 1.6391, "step": 40483 }, { "epoch": 3.172727272727273, "grad_norm": 3.877089023590088, "learning_rate": 1.0226512821842234e-06, "loss": 1.4872, "step": 40484 }, { "epoch": 3.1728056426332287, "grad_norm": 3.0504143238067627, "learning_rate": 1.022464424404827e-06, "loss": 1.3801, "step": 40485 }, { "epoch": 3.172884012539185, "grad_norm": 5.745842933654785, "learning_rate": 1.0222775817539333e-06, "loss": 1.0984, "step": 40486 }, { "epoch": 3.172962382445141, "grad_norm": 2.97463059425354, "learning_rate": 1.0220907542322522e-06, "loss": 1.2623, "step": 40487 }, { "epoch": 3.1730407523510973, "grad_norm": 7.032838344573975, "learning_rate": 1.0219039418404946e-06, "loss": 1.8222, "step": 40488 }, { "epoch": 3.173119122257053, "grad_norm": 4.081233501434326, "learning_rate": 1.0217171445793716e-06, "loss": 1.5993, "step": 40489 }, { "epoch": 3.1731974921630095, "grad_norm": 3.605454683303833, "learning_rate": 1.0215303624495931e-06, "loss": 1.4861, "step": 40490 }, { "epoch": 3.1732758620689654, "grad_norm": 7.323841571807861, "learning_rate": 1.0213435954518698e-06, "loss": 1.1989, "step": 40491 }, { "epoch": 3.1733542319749217, "grad_norm": 4.690621852874756, "learning_rate": 1.0211568435869114e-06, "loss": 0.7272, "step": 40492 }, { "epoch": 3.1734326018808776, "grad_norm": 3.2264046669006348, "learning_rate": 1.020970106855429e-06, "loss": 1.2937, "step": 40493 }, { "epoch": 3.173510971786834, "grad_norm": 2.797238349914551, "learning_rate": 1.0207833852581322e-06, "loss": 0.9932, "step": 40494 }, { "epoch": 3.17358934169279, "grad_norm": 3.145174980163574, "learning_rate": 1.0205966787957333e-06, "loss": 0.8091, "step": 40495 }, { "epoch": 3.173667711598746, "grad_norm": 2.9826197624206543, "learning_rate": 1.0204099874689377e-06, "loss": 1.5096, "step": 40496 }, { "epoch": 3.173746081504702, "grad_norm": 19.11865997314453, "learning_rate": 1.02022331127846e-06, "loss": 0.8116, "step": 40497 }, { "epoch": 3.1738244514106584, "grad_norm": 3.4559524059295654, "learning_rate": 1.0200366502250098e-06, "loss": 1.1878, "step": 40498 }, { "epoch": 3.1739028213166143, "grad_norm": 2.5105063915252686, "learning_rate": 1.0198500043092952e-06, "loss": 1.2355, "step": 40499 }, { "epoch": 3.1739811912225706, "grad_norm": 4.9138994216918945, "learning_rate": 1.019663373532026e-06, "loss": 1.3025, "step": 40500 }, { "epoch": 3.1740595611285265, "grad_norm": 2.42958402633667, "learning_rate": 1.0194767578939124e-06, "loss": 1.3338, "step": 40501 }, { "epoch": 3.174137931034483, "grad_norm": 2.5399842262268066, "learning_rate": 1.019290157395667e-06, "loss": 1.2642, "step": 40502 }, { "epoch": 3.1742163009404387, "grad_norm": 3.4476590156555176, "learning_rate": 1.0191035720379955e-06, "loss": 1.5995, "step": 40503 }, { "epoch": 3.174294670846395, "grad_norm": 9.440513610839844, "learning_rate": 1.01891700182161e-06, "loss": 0.8263, "step": 40504 }, { "epoch": 3.174373040752351, "grad_norm": 3.9245340824127197, "learning_rate": 1.0187304467472186e-06, "loss": 1.2251, "step": 40505 }, { "epoch": 3.1744514106583073, "grad_norm": 2.286468505859375, "learning_rate": 1.0185439068155323e-06, "loss": 1.2178, "step": 40506 }, { "epoch": 3.174529780564263, "grad_norm": 3.6985576152801514, "learning_rate": 1.0183573820272597e-06, "loss": 1.6777, "step": 40507 }, { "epoch": 3.1746081504702195, "grad_norm": 2.4091057777404785, "learning_rate": 1.0181708723831107e-06, "loss": 1.3938, "step": 40508 }, { "epoch": 3.1746865203761754, "grad_norm": 4.820573806762695, "learning_rate": 1.017984377883796e-06, "loss": 1.743, "step": 40509 }, { "epoch": 3.1747648902821317, "grad_norm": 3.37485933303833, "learning_rate": 1.0177978985300203e-06, "loss": 1.3944, "step": 40510 }, { "epoch": 3.1748432601880876, "grad_norm": 8.943513870239258, "learning_rate": 1.0176114343224974e-06, "loss": 1.0998, "step": 40511 }, { "epoch": 3.174921630094044, "grad_norm": 2.828975200653076, "learning_rate": 1.0174249852619355e-06, "loss": 1.3387, "step": 40512 }, { "epoch": 3.175, "grad_norm": 3.778201103210449, "learning_rate": 1.0172385513490447e-06, "loss": 1.382, "step": 40513 }, { "epoch": 3.175078369905956, "grad_norm": 2.9962198734283447, "learning_rate": 1.0170521325845313e-06, "loss": 1.554, "step": 40514 }, { "epoch": 3.175156739811912, "grad_norm": 3.319762945175171, "learning_rate": 1.016865728969104e-06, "loss": 1.4306, "step": 40515 }, { "epoch": 3.1752351097178684, "grad_norm": 7.907347679138184, "learning_rate": 1.0166793405034753e-06, "loss": 1.7116, "step": 40516 }, { "epoch": 3.1753134796238243, "grad_norm": 2.9617555141448975, "learning_rate": 1.0164929671883538e-06, "loss": 1.4701, "step": 40517 }, { "epoch": 3.1753918495297806, "grad_norm": 39.23566436767578, "learning_rate": 1.016306609024446e-06, "loss": 1.5202, "step": 40518 }, { "epoch": 3.1754702194357365, "grad_norm": 4.035568714141846, "learning_rate": 1.0161202660124615e-06, "loss": 1.2422, "step": 40519 }, { "epoch": 3.175548589341693, "grad_norm": 4.567208290100098, "learning_rate": 1.0159339381531075e-06, "loss": 1.3151, "step": 40520 }, { "epoch": 3.1756269592476487, "grad_norm": 3.3563954830169678, "learning_rate": 1.0157476254470978e-06, "loss": 1.1296, "step": 40521 }, { "epoch": 3.175705329153605, "grad_norm": 5.978551864624023, "learning_rate": 1.0155613278951354e-06, "loss": 1.6707, "step": 40522 }, { "epoch": 3.175783699059561, "grad_norm": 3.949901819229126, "learning_rate": 1.0153750454979322e-06, "loss": 1.4032, "step": 40523 }, { "epoch": 3.1758620689655173, "grad_norm": 9.28901195526123, "learning_rate": 1.0151887782561953e-06, "loss": 1.1507, "step": 40524 }, { "epoch": 3.175940438871473, "grad_norm": 5.583372116088867, "learning_rate": 1.0150025261706332e-06, "loss": 1.4216, "step": 40525 }, { "epoch": 3.1760188087774295, "grad_norm": 5.561842441558838, "learning_rate": 1.0148162892419549e-06, "loss": 1.4512, "step": 40526 }, { "epoch": 3.176097178683386, "grad_norm": 4.0156683921813965, "learning_rate": 1.0146300674708687e-06, "loss": 1.2868, "step": 40527 }, { "epoch": 3.1761755485893417, "grad_norm": 3.9311439990997314, "learning_rate": 1.0144438608580842e-06, "loss": 1.6446, "step": 40528 }, { "epoch": 3.1762539184952976, "grad_norm": 2.4123566150665283, "learning_rate": 1.014257669404305e-06, "loss": 1.4395, "step": 40529 }, { "epoch": 3.176332288401254, "grad_norm": 2.9674270153045654, "learning_rate": 1.014071493110244e-06, "loss": 1.3789, "step": 40530 }, { "epoch": 3.1764106583072103, "grad_norm": 5.706209659576416, "learning_rate": 1.0138853319766074e-06, "loss": 0.949, "step": 40531 }, { "epoch": 3.176489028213166, "grad_norm": 11.05409049987793, "learning_rate": 1.0136991860041057e-06, "loss": 0.6498, "step": 40532 }, { "epoch": 3.176567398119122, "grad_norm": 3.0718605518341064, "learning_rate": 1.0135130551934424e-06, "loss": 1.5954, "step": 40533 }, { "epoch": 3.1766457680250784, "grad_norm": 7.895241737365723, "learning_rate": 1.0133269395453277e-06, "loss": 0.9876, "step": 40534 }, { "epoch": 3.1767241379310347, "grad_norm": 3.916962146759033, "learning_rate": 1.0131408390604702e-06, "loss": 1.4152, "step": 40535 }, { "epoch": 3.1768025078369906, "grad_norm": 3.078824758529663, "learning_rate": 1.0129547537395785e-06, "loss": 1.27, "step": 40536 }, { "epoch": 3.1768808777429465, "grad_norm": 2.6712889671325684, "learning_rate": 1.0127686835833578e-06, "loss": 1.2841, "step": 40537 }, { "epoch": 3.176959247648903, "grad_norm": 4.8386969566345215, "learning_rate": 1.0125826285925173e-06, "loss": 1.1713, "step": 40538 }, { "epoch": 3.177037617554859, "grad_norm": 3.041390895843506, "learning_rate": 1.0123965887677644e-06, "loss": 1.3949, "step": 40539 }, { "epoch": 3.177115987460815, "grad_norm": 2.52297043800354, "learning_rate": 1.0122105641098062e-06, "loss": 1.2013, "step": 40540 }, { "epoch": 3.1771943573667714, "grad_norm": 2.0607380867004395, "learning_rate": 1.0120245546193513e-06, "loss": 1.3523, "step": 40541 }, { "epoch": 3.1772727272727272, "grad_norm": 4.066008567810059, "learning_rate": 1.0118385602971064e-06, "loss": 1.2734, "step": 40542 }, { "epoch": 3.1773510971786836, "grad_norm": 3.070502996444702, "learning_rate": 1.0116525811437794e-06, "loss": 1.0863, "step": 40543 }, { "epoch": 3.1774294670846395, "grad_norm": 6.213055610656738, "learning_rate": 1.0114666171600773e-06, "loss": 0.6908, "step": 40544 }, { "epoch": 3.177507836990596, "grad_norm": 2.6212356090545654, "learning_rate": 1.0112806683467074e-06, "loss": 1.2693, "step": 40545 }, { "epoch": 3.1775862068965517, "grad_norm": 1.9764161109924316, "learning_rate": 1.011094734704377e-06, "loss": 1.1352, "step": 40546 }, { "epoch": 3.177664576802508, "grad_norm": 4.178501605987549, "learning_rate": 1.0109088162337937e-06, "loss": 0.9337, "step": 40547 }, { "epoch": 3.177742946708464, "grad_norm": 3.2316319942474365, "learning_rate": 1.0107229129356644e-06, "loss": 1.4951, "step": 40548 }, { "epoch": 3.1778213166144202, "grad_norm": 7.696281433105469, "learning_rate": 1.010537024810696e-06, "loss": 1.5934, "step": 40549 }, { "epoch": 3.177899686520376, "grad_norm": 2.901637315750122, "learning_rate": 1.0103511518595953e-06, "loss": 1.2104, "step": 40550 }, { "epoch": 3.1779780564263325, "grad_norm": 2.6149039268493652, "learning_rate": 1.0101652940830715e-06, "loss": 1.4873, "step": 40551 }, { "epoch": 3.1780564263322884, "grad_norm": 3.745382070541382, "learning_rate": 1.0099794514818268e-06, "loss": 1.2093, "step": 40552 }, { "epoch": 3.1781347962382447, "grad_norm": 3.830242395401001, "learning_rate": 1.009793624056572e-06, "loss": 1.4498, "step": 40553 }, { "epoch": 3.1782131661442006, "grad_norm": 3.336667060852051, "learning_rate": 1.0096078118080143e-06, "loss": 1.5445, "step": 40554 }, { "epoch": 3.178291536050157, "grad_norm": 3.8794126510620117, "learning_rate": 1.009422014736857e-06, "loss": 1.257, "step": 40555 }, { "epoch": 3.178369905956113, "grad_norm": 5.42453670501709, "learning_rate": 1.0092362328438093e-06, "loss": 2.0008, "step": 40556 }, { "epoch": 3.178448275862069, "grad_norm": 3.2123875617980957, "learning_rate": 1.0090504661295758e-06, "loss": 1.4032, "step": 40557 }, { "epoch": 3.178526645768025, "grad_norm": 3.177419900894165, "learning_rate": 1.008864714594867e-06, "loss": 1.2122, "step": 40558 }, { "epoch": 3.1786050156739813, "grad_norm": 4.1346330642700195, "learning_rate": 1.0086789782403856e-06, "loss": 1.1437, "step": 40559 }, { "epoch": 3.1786833855799372, "grad_norm": 3.0539183616638184, "learning_rate": 1.0084932570668393e-06, "loss": 1.5957, "step": 40560 }, { "epoch": 3.1787617554858936, "grad_norm": 3.9911344051361084, "learning_rate": 1.008307551074934e-06, "loss": 1.3697, "step": 40561 }, { "epoch": 3.1788401253918495, "grad_norm": 5.935473918914795, "learning_rate": 1.008121860265377e-06, "loss": 0.9807, "step": 40562 }, { "epoch": 3.178918495297806, "grad_norm": 10.859354972839355, "learning_rate": 1.007936184638874e-06, "loss": 0.9417, "step": 40563 }, { "epoch": 3.1789968652037617, "grad_norm": 3.7023746967315674, "learning_rate": 1.0077505241961311e-06, "loss": 1.2942, "step": 40564 }, { "epoch": 3.179075235109718, "grad_norm": 5.834690093994141, "learning_rate": 1.0075648789378562e-06, "loss": 1.4251, "step": 40565 }, { "epoch": 3.179153605015674, "grad_norm": 14.200746536254883, "learning_rate": 1.007379248864751e-06, "loss": 1.0371, "step": 40566 }, { "epoch": 3.1792319749216302, "grad_norm": 5.831258296966553, "learning_rate": 1.0071936339775262e-06, "loss": 1.6088, "step": 40567 }, { "epoch": 3.179310344827586, "grad_norm": 3.2280373573303223, "learning_rate": 1.0070080342768857e-06, "loss": 1.2915, "step": 40568 }, { "epoch": 3.1793887147335425, "grad_norm": 8.583817481994629, "learning_rate": 1.006822449763537e-06, "loss": 0.8437, "step": 40569 }, { "epoch": 3.1794670846394983, "grad_norm": 3.2189488410949707, "learning_rate": 1.0066368804381833e-06, "loss": 1.1327, "step": 40570 }, { "epoch": 3.1795454545454547, "grad_norm": 3.2671396732330322, "learning_rate": 1.0064513263015302e-06, "loss": 1.2829, "step": 40571 }, { "epoch": 3.1796238244514106, "grad_norm": 2.9202351570129395, "learning_rate": 1.0062657873542869e-06, "loss": 1.6551, "step": 40572 }, { "epoch": 3.179702194357367, "grad_norm": 3.6573944091796875, "learning_rate": 1.0060802635971583e-06, "loss": 1.1476, "step": 40573 }, { "epoch": 3.179780564263323, "grad_norm": 4.127517223358154, "learning_rate": 1.0058947550308478e-06, "loss": 0.8281, "step": 40574 }, { "epoch": 3.179858934169279, "grad_norm": 3.785181999206543, "learning_rate": 1.0057092616560622e-06, "loss": 1.4279, "step": 40575 }, { "epoch": 3.179937304075235, "grad_norm": 2.9499523639678955, "learning_rate": 1.005523783473505e-06, "loss": 1.335, "step": 40576 }, { "epoch": 3.1800156739811913, "grad_norm": 3.0200841426849365, "learning_rate": 1.0053383204838869e-06, "loss": 1.3681, "step": 40577 }, { "epoch": 3.1800940438871472, "grad_norm": 5.267801761627197, "learning_rate": 1.0051528726879083e-06, "loss": 1.4754, "step": 40578 }, { "epoch": 3.1801724137931036, "grad_norm": 9.801761627197266, "learning_rate": 1.0049674400862768e-06, "loss": 0.958, "step": 40579 }, { "epoch": 3.1802507836990594, "grad_norm": 3.4890806674957275, "learning_rate": 1.0047820226796966e-06, "loss": 1.4883, "step": 40580 }, { "epoch": 3.180329153605016, "grad_norm": 3.049241304397583, "learning_rate": 1.004596620468874e-06, "loss": 1.4838, "step": 40581 }, { "epoch": 3.1804075235109717, "grad_norm": 9.19800853729248, "learning_rate": 1.0044112334545137e-06, "loss": 1.8132, "step": 40582 }, { "epoch": 3.180485893416928, "grad_norm": 3.894986152648926, "learning_rate": 1.0042258616373207e-06, "loss": 1.4401, "step": 40583 }, { "epoch": 3.180564263322884, "grad_norm": 5.238790035247803, "learning_rate": 1.0040405050180013e-06, "loss": 1.767, "step": 40584 }, { "epoch": 3.18064263322884, "grad_norm": 3.071162700653076, "learning_rate": 1.003855163597257e-06, "loss": 1.4241, "step": 40585 }, { "epoch": 3.180721003134796, "grad_norm": 3.215291738510132, "learning_rate": 1.0036698373757963e-06, "loss": 1.6029, "step": 40586 }, { "epoch": 3.1807993730407524, "grad_norm": 5.909171104431152, "learning_rate": 1.0034845263543226e-06, "loss": 1.3844, "step": 40587 }, { "epoch": 3.1808777429467083, "grad_norm": 2.985687494277954, "learning_rate": 1.0032992305335426e-06, "loss": 0.9921, "step": 40588 }, { "epoch": 3.1809561128526647, "grad_norm": 4.217599868774414, "learning_rate": 1.0031139499141585e-06, "loss": 1.6129, "step": 40589 }, { "epoch": 3.1810344827586206, "grad_norm": 6.324328899383545, "learning_rate": 1.0029286844968738e-06, "loss": 1.0418, "step": 40590 }, { "epoch": 3.181112852664577, "grad_norm": 2.8114259243011475, "learning_rate": 1.0027434342823978e-06, "loss": 1.388, "step": 40591 }, { "epoch": 3.1811912225705328, "grad_norm": 9.890658378601074, "learning_rate": 1.0025581992714334e-06, "loss": 1.2632, "step": 40592 }, { "epoch": 3.181269592476489, "grad_norm": 2.545436382293701, "learning_rate": 1.002372979464683e-06, "loss": 1.1995, "step": 40593 }, { "epoch": 3.181347962382445, "grad_norm": 3.0037620067596436, "learning_rate": 1.0021877748628522e-06, "loss": 1.2777, "step": 40594 }, { "epoch": 3.1814263322884013, "grad_norm": 6.296749114990234, "learning_rate": 1.0020025854666466e-06, "loss": 1.6398, "step": 40595 }, { "epoch": 3.181504702194357, "grad_norm": 5.105883598327637, "learning_rate": 1.0018174112767687e-06, "loss": 1.4707, "step": 40596 }, { "epoch": 3.1815830721003135, "grad_norm": 2.4665396213531494, "learning_rate": 1.0016322522939242e-06, "loss": 1.0462, "step": 40597 }, { "epoch": 3.1816614420062694, "grad_norm": 2.954221725463867, "learning_rate": 1.001447108518817e-06, "loss": 1.3007, "step": 40598 }, { "epoch": 3.1817398119122258, "grad_norm": 2.7270545959472656, "learning_rate": 1.001261979952151e-06, "loss": 1.5742, "step": 40599 }, { "epoch": 3.1818181818181817, "grad_norm": 2.2465484142303467, "learning_rate": 1.0010768665946309e-06, "loss": 1.1368, "step": 40600 }, { "epoch": 3.181896551724138, "grad_norm": 2.6986095905303955, "learning_rate": 1.0008917684469598e-06, "loss": 1.3557, "step": 40601 }, { "epoch": 3.181974921630094, "grad_norm": 3.122520923614502, "learning_rate": 1.000706685509843e-06, "loss": 1.4307, "step": 40602 }, { "epoch": 3.18205329153605, "grad_norm": 2.522892951965332, "learning_rate": 1.0005216177839848e-06, "loss": 1.4161, "step": 40603 }, { "epoch": 3.182131661442006, "grad_norm": 4.065420150756836, "learning_rate": 1.0003365652700857e-06, "loss": 0.8834, "step": 40604 }, { "epoch": 3.1822100313479624, "grad_norm": 4.157252311706543, "learning_rate": 1.0001515279688528e-06, "loss": 1.1328, "step": 40605 }, { "epoch": 3.1822884012539183, "grad_norm": 2.783752918243408, "learning_rate": 9.999665058809893e-07, "loss": 1.283, "step": 40606 }, { "epoch": 3.1823667711598747, "grad_norm": 3.221709966659546, "learning_rate": 9.997814990072003e-07, "loss": 1.275, "step": 40607 }, { "epoch": 3.1824451410658305, "grad_norm": 3.569014072418213, "learning_rate": 9.99596507348186e-07, "loss": 1.4952, "step": 40608 }, { "epoch": 3.182523510971787, "grad_norm": 2.881246328353882, "learning_rate": 9.99411530904651e-07, "loss": 1.0719, "step": 40609 }, { "epoch": 3.1826018808777428, "grad_norm": 6.549111366271973, "learning_rate": 9.992265696773024e-07, "loss": 1.2977, "step": 40610 }, { "epoch": 3.182680250783699, "grad_norm": 5.467233180999756, "learning_rate": 9.990416236668394e-07, "loss": 1.3667, "step": 40611 }, { "epoch": 3.182758620689655, "grad_norm": 4.257142066955566, "learning_rate": 9.988566928739668e-07, "loss": 0.9981, "step": 40612 }, { "epoch": 3.1828369905956113, "grad_norm": 4.180258274078369, "learning_rate": 9.98671777299388e-07, "loss": 1.7366, "step": 40613 }, { "epoch": 3.182915360501567, "grad_norm": 5.22079610824585, "learning_rate": 9.984868769438072e-07, "loss": 1.3821, "step": 40614 }, { "epoch": 3.1829937304075235, "grad_norm": 5.988280773162842, "learning_rate": 9.983019918079267e-07, "loss": 1.2276, "step": 40615 }, { "epoch": 3.1830721003134794, "grad_norm": 2.1288974285125732, "learning_rate": 9.981171218924502e-07, "loss": 1.287, "step": 40616 }, { "epoch": 3.1831504702194358, "grad_norm": 4.79960298538208, "learning_rate": 9.979322671980806e-07, "loss": 1.92, "step": 40617 }, { "epoch": 3.1832288401253916, "grad_norm": 4.189884662628174, "learning_rate": 9.977474277255212e-07, "loss": 1.5817, "step": 40618 }, { "epoch": 3.183307210031348, "grad_norm": 3.1462690830230713, "learning_rate": 9.97562603475475e-07, "loss": 1.6163, "step": 40619 }, { "epoch": 3.183385579937304, "grad_norm": 7.198812484741211, "learning_rate": 9.973777944486447e-07, "loss": 1.1606, "step": 40620 }, { "epoch": 3.18346394984326, "grad_norm": 2.8266913890838623, "learning_rate": 9.971930006457343e-07, "loss": 1.5093, "step": 40621 }, { "epoch": 3.1835423197492165, "grad_norm": 4.478581428527832, "learning_rate": 9.97008222067445e-07, "loss": 1.3994, "step": 40622 }, { "epoch": 3.1836206896551724, "grad_norm": 3.8755362033843994, "learning_rate": 9.968234587144787e-07, "loss": 1.2816, "step": 40623 }, { "epoch": 3.1836990595611283, "grad_norm": 2.9642131328582764, "learning_rate": 9.966387105875414e-07, "loss": 1.2934, "step": 40624 }, { "epoch": 3.1837774294670846, "grad_norm": 2.3788552284240723, "learning_rate": 9.96453977687335e-07, "loss": 1.4857, "step": 40625 }, { "epoch": 3.183855799373041, "grad_norm": 3.630556106567383, "learning_rate": 9.962692600145607e-07, "loss": 1.6033, "step": 40626 }, { "epoch": 3.183934169278997, "grad_norm": 5.25580358505249, "learning_rate": 9.960845575699213e-07, "loss": 1.1853, "step": 40627 }, { "epoch": 3.1840125391849528, "grad_norm": 2.8019487857818604, "learning_rate": 9.958998703541185e-07, "loss": 1.1329, "step": 40628 }, { "epoch": 3.184090909090909, "grad_norm": 2.1910243034362793, "learning_rate": 9.95715198367858e-07, "loss": 1.0094, "step": 40629 }, { "epoch": 3.1841692789968654, "grad_norm": 3.1352202892303467, "learning_rate": 9.9553054161184e-07, "loss": 1.4381, "step": 40630 }, { "epoch": 3.1842476489028213, "grad_norm": 4.064004898071289, "learning_rate": 9.953459000867661e-07, "loss": 1.4281, "step": 40631 }, { "epoch": 3.1843260188087776, "grad_norm": 3.2592058181762695, "learning_rate": 9.951612737933398e-07, "loss": 1.3787, "step": 40632 }, { "epoch": 3.1844043887147335, "grad_norm": 2.5515825748443604, "learning_rate": 9.949766627322626e-07, "loss": 1.5843, "step": 40633 }, { "epoch": 3.18448275862069, "grad_norm": 5.929932117462158, "learning_rate": 9.947920669042377e-07, "loss": 1.2165, "step": 40634 }, { "epoch": 3.1845611285266457, "grad_norm": 7.831128120422363, "learning_rate": 9.946074863099664e-07, "loss": 1.5703, "step": 40635 }, { "epoch": 3.184639498432602, "grad_norm": 3.1986496448516846, "learning_rate": 9.944229209501504e-07, "loss": 1.2228, "step": 40636 }, { "epoch": 3.184717868338558, "grad_norm": 3.2766449451446533, "learning_rate": 9.942383708254927e-07, "loss": 1.4402, "step": 40637 }, { "epoch": 3.1847962382445143, "grad_norm": 3.915029764175415, "learning_rate": 9.940538359366946e-07, "loss": 1.127, "step": 40638 }, { "epoch": 3.18487460815047, "grad_norm": 4.268222808837891, "learning_rate": 9.938693162844582e-07, "loss": 1.2649, "step": 40639 }, { "epoch": 3.1849529780564265, "grad_norm": 8.515499114990234, "learning_rate": 9.936848118694865e-07, "loss": 1.9022, "step": 40640 }, { "epoch": 3.1850313479623824, "grad_norm": 4.942101001739502, "learning_rate": 9.935003226924789e-07, "loss": 1.2828, "step": 40641 }, { "epoch": 3.1851097178683387, "grad_norm": 2.60033917427063, "learning_rate": 9.933158487541367e-07, "loss": 1.4015, "step": 40642 }, { "epoch": 3.1851880877742946, "grad_norm": 2.8205738067626953, "learning_rate": 9.931313900551643e-07, "loss": 1.3423, "step": 40643 }, { "epoch": 3.185266457680251, "grad_norm": 4.0569257736206055, "learning_rate": 9.929469465962632e-07, "loss": 1.2607, "step": 40644 }, { "epoch": 3.185344827586207, "grad_norm": 4.277392864227295, "learning_rate": 9.927625183781331e-07, "loss": 1.3633, "step": 40645 }, { "epoch": 3.185423197492163, "grad_norm": 2.5273818969726562, "learning_rate": 9.925781054014755e-07, "loss": 1.0041, "step": 40646 }, { "epoch": 3.185501567398119, "grad_norm": 3.7500767707824707, "learning_rate": 9.923937076669916e-07, "loss": 1.4128, "step": 40647 }, { "epoch": 3.1855799373040754, "grad_norm": 5.022342205047607, "learning_rate": 9.922093251753856e-07, "loss": 1.0075, "step": 40648 }, { "epoch": 3.1856583072100313, "grad_norm": 5.19766902923584, "learning_rate": 9.92024957927356e-07, "loss": 1.3788, "step": 40649 }, { "epoch": 3.1857366771159876, "grad_norm": 2.75019907951355, "learning_rate": 9.918406059236047e-07, "loss": 1.2698, "step": 40650 }, { "epoch": 3.1858150470219435, "grad_norm": 5.38434362411499, "learning_rate": 9.916562691648329e-07, "loss": 1.6432, "step": 40651 }, { "epoch": 3.1858934169279, "grad_norm": 4.279138565063477, "learning_rate": 9.91471947651742e-07, "loss": 0.8395, "step": 40652 }, { "epoch": 3.1859717868338557, "grad_norm": 2.995732545852661, "learning_rate": 9.912876413850325e-07, "loss": 1.1204, "step": 40653 }, { "epoch": 3.186050156739812, "grad_norm": 3.8231611251831055, "learning_rate": 9.91103350365406e-07, "loss": 1.5915, "step": 40654 }, { "epoch": 3.186128526645768, "grad_norm": 3.523833990097046, "learning_rate": 9.909190745935637e-07, "loss": 1.2129, "step": 40655 }, { "epoch": 3.1862068965517243, "grad_norm": 3.127609968185425, "learning_rate": 9.907348140702054e-07, "loss": 1.317, "step": 40656 }, { "epoch": 3.18628526645768, "grad_norm": 4.346719741821289, "learning_rate": 9.905505687960332e-07, "loss": 0.9242, "step": 40657 }, { "epoch": 3.1863636363636365, "grad_norm": 2.154428482055664, "learning_rate": 9.903663387717471e-07, "loss": 1.3204, "step": 40658 }, { "epoch": 3.1864420062695924, "grad_norm": 5.725396156311035, "learning_rate": 9.90182123998049e-07, "loss": 1.2778, "step": 40659 }, { "epoch": 3.1865203761755487, "grad_norm": 2.8816184997558594, "learning_rate": 9.899979244756358e-07, "loss": 1.2166, "step": 40660 }, { "epoch": 3.1865987460815046, "grad_norm": 3.3580760955810547, "learning_rate": 9.898137402052127e-07, "loss": 1.1213, "step": 40661 }, { "epoch": 3.186677115987461, "grad_norm": 2.30595326423645, "learning_rate": 9.89629571187478e-07, "loss": 0.9902, "step": 40662 }, { "epoch": 3.186755485893417, "grad_norm": 3.794267177581787, "learning_rate": 9.894454174231337e-07, "loss": 1.3305, "step": 40663 }, { "epoch": 3.186833855799373, "grad_norm": 4.364669322967529, "learning_rate": 9.89261278912878e-07, "loss": 1.8997, "step": 40664 }, { "epoch": 3.186912225705329, "grad_norm": 3.036407470703125, "learning_rate": 9.890771556574113e-07, "loss": 1.2893, "step": 40665 }, { "epoch": 3.1869905956112854, "grad_norm": 3.2814996242523193, "learning_rate": 9.888930476574372e-07, "loss": 0.8329, "step": 40666 }, { "epoch": 3.1870689655172413, "grad_norm": 2.6079330444335938, "learning_rate": 9.887089549136525e-07, "loss": 1.1178, "step": 40667 }, { "epoch": 3.1871473354231976, "grad_norm": 5.436903953552246, "learning_rate": 9.885248774267586e-07, "loss": 1.3425, "step": 40668 }, { "epoch": 3.1872257053291535, "grad_norm": 4.266282081604004, "learning_rate": 9.883408151974562e-07, "loss": 1.2733, "step": 40669 }, { "epoch": 3.18730407523511, "grad_norm": 3.1032423973083496, "learning_rate": 9.881567682264437e-07, "loss": 1.0964, "step": 40670 }, { "epoch": 3.1873824451410657, "grad_norm": 6.359376907348633, "learning_rate": 9.879727365144231e-07, "loss": 1.022, "step": 40671 }, { "epoch": 3.187460815047022, "grad_norm": 2.4923527240753174, "learning_rate": 9.877887200620933e-07, "loss": 1.0132, "step": 40672 }, { "epoch": 3.187539184952978, "grad_norm": 3.0003366470336914, "learning_rate": 9.876047188701542e-07, "loss": 0.9942, "step": 40673 }, { "epoch": 3.1876175548589343, "grad_norm": 3.84747052192688, "learning_rate": 9.87420732939306e-07, "loss": 1.3317, "step": 40674 }, { "epoch": 3.18769592476489, "grad_norm": 2.6777374744415283, "learning_rate": 9.872367622702484e-07, "loss": 1.2703, "step": 40675 }, { "epoch": 3.1877742946708465, "grad_norm": 3.0266501903533936, "learning_rate": 9.87052806863681e-07, "loss": 1.4051, "step": 40676 }, { "epoch": 3.1878526645768024, "grad_norm": 2.6509153842926025, "learning_rate": 9.868688667203047e-07, "loss": 0.9461, "step": 40677 }, { "epoch": 3.1879310344827587, "grad_norm": 3.8571715354919434, "learning_rate": 9.86684941840817e-07, "loss": 1.5633, "step": 40678 }, { "epoch": 3.1880094043887146, "grad_norm": 2.7922706604003906, "learning_rate": 9.865010322259167e-07, "loss": 1.2796, "step": 40679 }, { "epoch": 3.188087774294671, "grad_norm": 4.635634899139404, "learning_rate": 9.863171378763064e-07, "loss": 1.4785, "step": 40680 }, { "epoch": 3.188166144200627, "grad_norm": 2.9427073001861572, "learning_rate": 9.861332587926857e-07, "loss": 1.4927, "step": 40681 }, { "epoch": 3.188244514106583, "grad_norm": 2.9662108421325684, "learning_rate": 9.85949394975751e-07, "loss": 1.2213, "step": 40682 }, { "epoch": 3.188322884012539, "grad_norm": 2.6265687942504883, "learning_rate": 9.857655464262034e-07, "loss": 1.3598, "step": 40683 }, { "epoch": 3.1884012539184954, "grad_norm": 2.9354794025421143, "learning_rate": 9.8558171314474e-07, "loss": 1.3373, "step": 40684 }, { "epoch": 3.1884796238244513, "grad_norm": 2.733644485473633, "learning_rate": 9.853978951320647e-07, "loss": 1.2424, "step": 40685 }, { "epoch": 3.1885579937304076, "grad_norm": 2.860914707183838, "learning_rate": 9.85214092388872e-07, "loss": 1.49, "step": 40686 }, { "epoch": 3.1886363636363635, "grad_norm": 3.362910032272339, "learning_rate": 9.850303049158632e-07, "loss": 1.4016, "step": 40687 }, { "epoch": 3.18871473354232, "grad_norm": 11.04102897644043, "learning_rate": 9.84846532713737e-07, "loss": 0.9808, "step": 40688 }, { "epoch": 3.1887931034482757, "grad_norm": 6.34730339050293, "learning_rate": 9.846627757831923e-07, "loss": 1.1603, "step": 40689 }, { "epoch": 3.188871473354232, "grad_norm": 3.620114326477051, "learning_rate": 9.844790341249276e-07, "loss": 0.9802, "step": 40690 }, { "epoch": 3.188949843260188, "grad_norm": 2.333235502243042, "learning_rate": 9.842953077396422e-07, "loss": 1.18, "step": 40691 }, { "epoch": 3.1890282131661443, "grad_norm": 4.240981101989746, "learning_rate": 9.841115966280368e-07, "loss": 1.1006, "step": 40692 }, { "epoch": 3.1891065830721, "grad_norm": 4.958477020263672, "learning_rate": 9.839279007908048e-07, "loss": 1.2593, "step": 40693 }, { "epoch": 3.1891849529780565, "grad_norm": 4.099371433258057, "learning_rate": 9.8374422022865e-07, "loss": 1.4931, "step": 40694 }, { "epoch": 3.1892633228840124, "grad_norm": 5.560638427734375, "learning_rate": 9.835605549422694e-07, "loss": 1.375, "step": 40695 }, { "epoch": 3.1893416927899687, "grad_norm": 6.634140491485596, "learning_rate": 9.833769049323623e-07, "loss": 1.2731, "step": 40696 }, { "epoch": 3.1894200626959246, "grad_norm": 3.5583243370056152, "learning_rate": 9.831932701996255e-07, "loss": 1.4846, "step": 40697 }, { "epoch": 3.189498432601881, "grad_norm": 3.5782930850982666, "learning_rate": 9.830096507447567e-07, "loss": 1.2543, "step": 40698 }, { "epoch": 3.189576802507837, "grad_norm": 4.633767604827881, "learning_rate": 9.828260465684575e-07, "loss": 1.5841, "step": 40699 }, { "epoch": 3.189655172413793, "grad_norm": 2.9425411224365234, "learning_rate": 9.826424576714262e-07, "loss": 1.2462, "step": 40700 }, { "epoch": 3.189733542319749, "grad_norm": 2.0503342151641846, "learning_rate": 9.824588840543575e-07, "loss": 1.2514, "step": 40701 }, { "epoch": 3.1898119122257054, "grad_norm": 3.375028371810913, "learning_rate": 9.822753257179524e-07, "loss": 0.7792, "step": 40702 }, { "epoch": 3.1898902821316613, "grad_norm": 8.35017204284668, "learning_rate": 9.820917826629068e-07, "loss": 1.0318, "step": 40703 }, { "epoch": 3.1899686520376176, "grad_norm": 5.831389427185059, "learning_rate": 9.819082548899228e-07, "loss": 1.5573, "step": 40704 }, { "epoch": 3.1900470219435735, "grad_norm": 3.6503243446350098, "learning_rate": 9.81724742399695e-07, "loss": 1.1866, "step": 40705 }, { "epoch": 3.19012539184953, "grad_norm": 2.956864833831787, "learning_rate": 9.815412451929218e-07, "loss": 1.4032, "step": 40706 }, { "epoch": 3.1902037617554857, "grad_norm": 4.422419548034668, "learning_rate": 9.81357763270302e-07, "loss": 1.5118, "step": 40707 }, { "epoch": 3.190282131661442, "grad_norm": 5.123709678649902, "learning_rate": 9.811742966325332e-07, "loss": 1.2786, "step": 40708 }, { "epoch": 3.190360501567398, "grad_norm": 3.317117929458618, "learning_rate": 9.80990845280313e-07, "loss": 1.2168, "step": 40709 }, { "epoch": 3.1904388714733543, "grad_norm": 3.0991883277893066, "learning_rate": 9.808074092143399e-07, "loss": 1.2874, "step": 40710 }, { "epoch": 3.19051724137931, "grad_norm": 5.156348705291748, "learning_rate": 9.80623988435312e-07, "loss": 1.9322, "step": 40711 }, { "epoch": 3.1905956112852665, "grad_norm": 2.739058494567871, "learning_rate": 9.80440582943923e-07, "loss": 0.9978, "step": 40712 }, { "epoch": 3.1906739811912224, "grad_norm": 4.305971622467041, "learning_rate": 9.80257192740875e-07, "loss": 0.8155, "step": 40713 }, { "epoch": 3.1907523510971787, "grad_norm": 2.737220287322998, "learning_rate": 9.800738178268642e-07, "loss": 1.2056, "step": 40714 }, { "epoch": 3.1908307210031346, "grad_norm": 3.733262777328491, "learning_rate": 9.798904582025887e-07, "loss": 1.535, "step": 40715 }, { "epoch": 3.190909090909091, "grad_norm": 3.4615883827209473, "learning_rate": 9.79707113868744e-07, "loss": 1.2552, "step": 40716 }, { "epoch": 3.1909874608150472, "grad_norm": 2.7760415077209473, "learning_rate": 9.79523784826027e-07, "loss": 1.2284, "step": 40717 }, { "epoch": 3.191065830721003, "grad_norm": 3.199066400527954, "learning_rate": 9.79340471075138e-07, "loss": 1.4265, "step": 40718 }, { "epoch": 3.191144200626959, "grad_norm": 2.685675621032715, "learning_rate": 9.791571726167742e-07, "loss": 1.3815, "step": 40719 }, { "epoch": 3.1912225705329154, "grad_norm": 7.2448811531066895, "learning_rate": 9.789738894516294e-07, "loss": 1.2033, "step": 40720 }, { "epoch": 3.1913009404388717, "grad_norm": 3.72546124458313, "learning_rate": 9.787906215804033e-07, "loss": 1.1331, "step": 40721 }, { "epoch": 3.1913793103448276, "grad_norm": 5.219292163848877, "learning_rate": 9.786073690037917e-07, "loss": 1.0247, "step": 40722 }, { "epoch": 3.1914576802507835, "grad_norm": 4.087313652038574, "learning_rate": 9.784241317224924e-07, "loss": 0.9647, "step": 40723 }, { "epoch": 3.19153605015674, "grad_norm": 4.9368672370910645, "learning_rate": 9.78240909737202e-07, "loss": 1.623, "step": 40724 }, { "epoch": 3.191614420062696, "grad_norm": 2.3221402168273926, "learning_rate": 9.780577030486171e-07, "loss": 1.3166, "step": 40725 }, { "epoch": 3.191692789968652, "grad_norm": 3.764564037322998, "learning_rate": 9.778745116574356e-07, "loss": 1.131, "step": 40726 }, { "epoch": 3.1917711598746084, "grad_norm": 9.542375564575195, "learning_rate": 9.77691335564353e-07, "loss": 0.6979, "step": 40727 }, { "epoch": 3.1918495297805642, "grad_norm": 2.5472970008850098, "learning_rate": 9.77508174770067e-07, "loss": 1.3637, "step": 40728 }, { "epoch": 3.1919278996865206, "grad_norm": 7.248903751373291, "learning_rate": 9.773250292752734e-07, "loss": 1.1907, "step": 40729 }, { "epoch": 3.1920062695924765, "grad_norm": 4.890104293823242, "learning_rate": 9.771418990806708e-07, "loss": 1.9409, "step": 40730 }, { "epoch": 3.192084639498433, "grad_norm": 4.930848121643066, "learning_rate": 9.769587841869515e-07, "loss": 1.4051, "step": 40731 }, { "epoch": 3.1921630094043887, "grad_norm": 3.047755241394043, "learning_rate": 9.767756845948155e-07, "loss": 1.4012, "step": 40732 }, { "epoch": 3.192241379310345, "grad_norm": 8.401549339294434, "learning_rate": 9.765926003049602e-07, "loss": 1.0213, "step": 40733 }, { "epoch": 3.192319749216301, "grad_norm": 3.4851207733154297, "learning_rate": 9.76409531318079e-07, "loss": 1.4989, "step": 40734 }, { "epoch": 3.1923981191222572, "grad_norm": 2.4779531955718994, "learning_rate": 9.76226477634869e-07, "loss": 1.026, "step": 40735 }, { "epoch": 3.192476489028213, "grad_norm": 3.384172201156616, "learning_rate": 9.760434392560252e-07, "loss": 1.1506, "step": 40736 }, { "epoch": 3.1925548589341695, "grad_norm": 2.4785263538360596, "learning_rate": 9.758604161822483e-07, "loss": 1.3633, "step": 40737 }, { "epoch": 3.1926332288401253, "grad_norm": 3.454319477081299, "learning_rate": 9.756774084142296e-07, "loss": 1.3945, "step": 40738 }, { "epoch": 3.1927115987460817, "grad_norm": 3.6286630630493164, "learning_rate": 9.754944159526675e-07, "loss": 1.3031, "step": 40739 }, { "epoch": 3.1927899686520376, "grad_norm": 5.372372150421143, "learning_rate": 9.753114387982576e-07, "loss": 0.8267, "step": 40740 }, { "epoch": 3.192868338557994, "grad_norm": 4.22259521484375, "learning_rate": 9.751284769516955e-07, "loss": 1.5462, "step": 40741 }, { "epoch": 3.19294670846395, "grad_norm": 5.459222793579102, "learning_rate": 9.749455304136773e-07, "loss": 1.444, "step": 40742 }, { "epoch": 3.193025078369906, "grad_norm": 2.99031662940979, "learning_rate": 9.747625991848992e-07, "loss": 1.4595, "step": 40743 }, { "epoch": 3.193103448275862, "grad_norm": 3.467428207397461, "learning_rate": 9.745796832660565e-07, "loss": 0.9045, "step": 40744 }, { "epoch": 3.1931818181818183, "grad_norm": 2.1032044887542725, "learning_rate": 9.743967826578449e-07, "loss": 1.3027, "step": 40745 }, { "epoch": 3.1932601880877742, "grad_norm": 2.9281837940216064, "learning_rate": 9.742138973609606e-07, "loss": 1.0945, "step": 40746 }, { "epoch": 3.1933385579937306, "grad_norm": 24.61701011657715, "learning_rate": 9.740310273760984e-07, "loss": 2.0535, "step": 40747 }, { "epoch": 3.1934169278996865, "grad_norm": 10.778501510620117, "learning_rate": 9.73848172703956e-07, "loss": 1.177, "step": 40748 }, { "epoch": 3.193495297805643, "grad_norm": 2.8742835521698, "learning_rate": 9.736653333452246e-07, "loss": 1.1799, "step": 40749 }, { "epoch": 3.1935736677115987, "grad_norm": 5.7130961418151855, "learning_rate": 9.734825093006034e-07, "loss": 0.9956, "step": 40750 }, { "epoch": 3.193652037617555, "grad_norm": 5.49473237991333, "learning_rate": 9.732997005707867e-07, "loss": 0.8816, "step": 40751 }, { "epoch": 3.193730407523511, "grad_norm": 3.9718613624572754, "learning_rate": 9.731169071564712e-07, "loss": 1.0697, "step": 40752 }, { "epoch": 3.1938087774294672, "grad_norm": 6.935365200042725, "learning_rate": 9.729341290583488e-07, "loss": 1.3597, "step": 40753 }, { "epoch": 3.193887147335423, "grad_norm": 3.43226957321167, "learning_rate": 9.727513662771166e-07, "loss": 0.9853, "step": 40754 }, { "epoch": 3.1939655172413794, "grad_norm": 2.999786376953125, "learning_rate": 9.725686188134687e-07, "loss": 1.1497, "step": 40755 }, { "epoch": 3.1940438871473353, "grad_norm": 2.3840274810791016, "learning_rate": 9.723858866681034e-07, "loss": 1.4995, "step": 40756 }, { "epoch": 3.1941222570532917, "grad_norm": 4.22274112701416, "learning_rate": 9.722031698417129e-07, "loss": 1.4354, "step": 40757 }, { "epoch": 3.1942006269592476, "grad_norm": 9.102712631225586, "learning_rate": 9.72020468334992e-07, "loss": 0.9554, "step": 40758 }, { "epoch": 3.194278996865204, "grad_norm": 2.3464858531951904, "learning_rate": 9.71837782148637e-07, "loss": 1.2407, "step": 40759 }, { "epoch": 3.19435736677116, "grad_norm": 2.7932310104370117, "learning_rate": 9.71655111283341e-07, "loss": 1.3491, "step": 40760 }, { "epoch": 3.194435736677116, "grad_norm": 3.8031678199768066, "learning_rate": 9.71472455739801e-07, "loss": 1.178, "step": 40761 }, { "epoch": 3.194514106583072, "grad_norm": 5.25673770904541, "learning_rate": 9.7128981551871e-07, "loss": 1.7158, "step": 40762 }, { "epoch": 3.1945924764890283, "grad_norm": 2.3481574058532715, "learning_rate": 9.71107190620763e-07, "loss": 1.061, "step": 40763 }, { "epoch": 3.194670846394984, "grad_norm": 5.177066802978516, "learning_rate": 9.709245810466556e-07, "loss": 1.4433, "step": 40764 }, { "epoch": 3.1947492163009406, "grad_norm": 3.656283140182495, "learning_rate": 9.707419867970812e-07, "loss": 1.5395, "step": 40765 }, { "epoch": 3.1948275862068964, "grad_norm": 3.2690184116363525, "learning_rate": 9.705594078727348e-07, "loss": 1.5072, "step": 40766 }, { "epoch": 3.1949059561128528, "grad_norm": 4.493466377258301, "learning_rate": 9.703768442743122e-07, "loss": 1.0801, "step": 40767 }, { "epoch": 3.1949843260188087, "grad_norm": 2.1831343173980713, "learning_rate": 9.701942960025035e-07, "loss": 1.4941, "step": 40768 }, { "epoch": 3.195062695924765, "grad_norm": 3.2219817638397217, "learning_rate": 9.700117630580074e-07, "loss": 1.1259, "step": 40769 }, { "epoch": 3.195141065830721, "grad_norm": 5.3410964012146, "learning_rate": 9.698292454415165e-07, "loss": 1.2094, "step": 40770 }, { "epoch": 3.195219435736677, "grad_norm": 2.7436091899871826, "learning_rate": 9.696467431537266e-07, "loss": 1.1899, "step": 40771 }, { "epoch": 3.195297805642633, "grad_norm": 4.80560827255249, "learning_rate": 9.69464256195329e-07, "loss": 1.2688, "step": 40772 }, { "epoch": 3.1953761755485894, "grad_norm": 3.5675177574157715, "learning_rate": 9.692817845670178e-07, "loss": 1.7846, "step": 40773 }, { "epoch": 3.1954545454545453, "grad_norm": 7.278171539306641, "learning_rate": 9.690993282694906e-07, "loss": 0.7107, "step": 40774 }, { "epoch": 3.1955329153605017, "grad_norm": 2.560988187789917, "learning_rate": 9.689168873034382e-07, "loss": 1.4972, "step": 40775 }, { "epoch": 3.1956112852664575, "grad_norm": 4.179409027099609, "learning_rate": 9.687344616695555e-07, "loss": 1.3972, "step": 40776 }, { "epoch": 3.195689655172414, "grad_norm": 4.27206563949585, "learning_rate": 9.685520513685366e-07, "loss": 1.6829, "step": 40777 }, { "epoch": 3.1957680250783698, "grad_norm": 2.6942684650421143, "learning_rate": 9.68369656401074e-07, "loss": 1.275, "step": 40778 }, { "epoch": 3.195846394984326, "grad_norm": 4.191238880157471, "learning_rate": 9.681872767678634e-07, "loss": 1.191, "step": 40779 }, { "epoch": 3.195924764890282, "grad_norm": 3.5630650520324707, "learning_rate": 9.680049124695973e-07, "loss": 1.4259, "step": 40780 }, { "epoch": 3.1960031347962383, "grad_norm": 4.121580600738525, "learning_rate": 9.678225635069694e-07, "loss": 1.5722, "step": 40781 }, { "epoch": 3.196081504702194, "grad_norm": 5.8253374099731445, "learning_rate": 9.676402298806737e-07, "loss": 1.267, "step": 40782 }, { "epoch": 3.1961598746081505, "grad_norm": 4.496553421020508, "learning_rate": 9.674579115914034e-07, "loss": 1.268, "step": 40783 }, { "epoch": 3.1962382445141064, "grad_norm": 3.1949148178100586, "learning_rate": 9.672756086398515e-07, "loss": 1.4691, "step": 40784 }, { "epoch": 3.1963166144200628, "grad_norm": 4.150741100311279, "learning_rate": 9.670933210267125e-07, "loss": 1.1866, "step": 40785 }, { "epoch": 3.1963949843260187, "grad_norm": 3.55779767036438, "learning_rate": 9.6691104875268e-07, "loss": 0.8382, "step": 40786 }, { "epoch": 3.196473354231975, "grad_norm": 2.3099875450134277, "learning_rate": 9.667287918184436e-07, "loss": 1.1713, "step": 40787 }, { "epoch": 3.196551724137931, "grad_norm": 2.8507297039031982, "learning_rate": 9.665465502247008e-07, "loss": 1.2372, "step": 40788 }, { "epoch": 3.196630094043887, "grad_norm": 2.6994104385375977, "learning_rate": 9.663643239721442e-07, "loss": 1.1712, "step": 40789 }, { "epoch": 3.196708463949843, "grad_norm": 2.7868237495422363, "learning_rate": 9.661821130614651e-07, "loss": 1.3328, "step": 40790 }, { "epoch": 3.1967868338557994, "grad_norm": 3.189244031906128, "learning_rate": 9.65999917493357e-07, "loss": 1.1373, "step": 40791 }, { "epoch": 3.1968652037617553, "grad_norm": 3.288417339324951, "learning_rate": 9.658177372685118e-07, "loss": 1.212, "step": 40792 }, { "epoch": 3.1969435736677116, "grad_norm": 3.9272191524505615, "learning_rate": 9.65635572387627e-07, "loss": 1.0569, "step": 40793 }, { "epoch": 3.1970219435736675, "grad_norm": 3.760591506958008, "learning_rate": 9.654534228513906e-07, "loss": 1.2833, "step": 40794 }, { "epoch": 3.197100313479624, "grad_norm": 3.5977542400360107, "learning_rate": 9.65271288660497e-07, "loss": 1.3128, "step": 40795 }, { "epoch": 3.1971786833855798, "grad_norm": 2.357020139694214, "learning_rate": 9.650891698156394e-07, "loss": 1.2353, "step": 40796 }, { "epoch": 3.197257053291536, "grad_norm": 2.618251323699951, "learning_rate": 9.649070663175097e-07, "loss": 1.2558, "step": 40797 }, { "epoch": 3.197335423197492, "grad_norm": 2.897270917892456, "learning_rate": 9.647249781668011e-07, "loss": 1.3356, "step": 40798 }, { "epoch": 3.1974137931034483, "grad_norm": 7.863814353942871, "learning_rate": 9.645429053642063e-07, "loss": 1.4545, "step": 40799 }, { "epoch": 3.197492163009404, "grad_norm": 3.8635060787200928, "learning_rate": 9.643608479104173e-07, "loss": 1.772, "step": 40800 }, { "epoch": 3.1975705329153605, "grad_norm": 3.38456130027771, "learning_rate": 9.641788058061269e-07, "loss": 1.2916, "step": 40801 }, { "epoch": 3.1976489028213164, "grad_norm": 5.381677150726318, "learning_rate": 9.639967790520278e-07, "loss": 1.0738, "step": 40802 }, { "epoch": 3.1977272727272728, "grad_norm": 2.747256278991699, "learning_rate": 9.638147676488112e-07, "loss": 1.3996, "step": 40803 }, { "epoch": 3.1978056426332286, "grad_norm": 8.866294860839844, "learning_rate": 9.636327715971715e-07, "loss": 1.3525, "step": 40804 }, { "epoch": 3.197884012539185, "grad_norm": 2.3309717178344727, "learning_rate": 9.634507908977986e-07, "loss": 1.1008, "step": 40805 }, { "epoch": 3.197962382445141, "grad_norm": 5.110066890716553, "learning_rate": 9.63268825551384e-07, "loss": 1.5036, "step": 40806 }, { "epoch": 3.198040752351097, "grad_norm": 12.522397994995117, "learning_rate": 9.630868755586232e-07, "loss": 0.7008, "step": 40807 }, { "epoch": 3.1981191222570535, "grad_norm": 4.135100364685059, "learning_rate": 9.629049409202073e-07, "loss": 1.0732, "step": 40808 }, { "epoch": 3.1981974921630094, "grad_norm": 3.403839588165283, "learning_rate": 9.627230216368261e-07, "loss": 1.4598, "step": 40809 }, { "epoch": 3.1982758620689653, "grad_norm": 2.7450380325317383, "learning_rate": 9.625411177091731e-07, "loss": 1.0451, "step": 40810 }, { "epoch": 3.1983542319749216, "grad_norm": 3.140228509902954, "learning_rate": 9.62359229137938e-07, "loss": 1.2088, "step": 40811 }, { "epoch": 3.198432601880878, "grad_norm": 10.51999282836914, "learning_rate": 9.62177355923818e-07, "loss": 1.2478, "step": 40812 }, { "epoch": 3.198510971786834, "grad_norm": 3.445951461791992, "learning_rate": 9.619954980674995e-07, "loss": 1.4125, "step": 40813 }, { "epoch": 3.1985893416927897, "grad_norm": 4.199424743652344, "learning_rate": 9.618136555696762e-07, "loss": 1.5188, "step": 40814 }, { "epoch": 3.198667711598746, "grad_norm": 2.935147285461426, "learning_rate": 9.616318284310395e-07, "loss": 1.2326, "step": 40815 }, { "epoch": 3.1987460815047024, "grad_norm": 2.432058811187744, "learning_rate": 9.614500166522812e-07, "loss": 1.3668, "step": 40816 }, { "epoch": 3.1988244514106583, "grad_norm": 4.463296413421631, "learning_rate": 9.612682202340927e-07, "loss": 1.3602, "step": 40817 }, { "epoch": 3.198902821316614, "grad_norm": 3.594135046005249, "learning_rate": 9.610864391771657e-07, "loss": 1.3863, "step": 40818 }, { "epoch": 3.1989811912225705, "grad_norm": 3.283513307571411, "learning_rate": 9.609046734821925e-07, "loss": 1.5921, "step": 40819 }, { "epoch": 3.199059561128527, "grad_norm": 3.557913303375244, "learning_rate": 9.607229231498605e-07, "loss": 1.4382, "step": 40820 }, { "epoch": 3.1991379310344827, "grad_norm": 2.618196964263916, "learning_rate": 9.605411881808657e-07, "loss": 1.1571, "step": 40821 }, { "epoch": 3.199216300940439, "grad_norm": 3.277801513671875, "learning_rate": 9.603594685758965e-07, "loss": 1.3393, "step": 40822 }, { "epoch": 3.199294670846395, "grad_norm": 5.039342880249023, "learning_rate": 9.60177764335647e-07, "loss": 1.1727, "step": 40823 }, { "epoch": 3.1993730407523513, "grad_norm": 6.837944030761719, "learning_rate": 9.599960754608052e-07, "loss": 0.5356, "step": 40824 }, { "epoch": 3.199451410658307, "grad_norm": 3.3581180572509766, "learning_rate": 9.598144019520612e-07, "loss": 1.0474, "step": 40825 }, { "epoch": 3.1995297805642635, "grad_norm": 3.920517683029175, "learning_rate": 9.5963274381011e-07, "loss": 1.4059, "step": 40826 }, { "epoch": 3.1996081504702194, "grad_norm": 3.5597000122070312, "learning_rate": 9.59451101035641e-07, "loss": 1.442, "step": 40827 }, { "epoch": 3.1996865203761757, "grad_norm": 5.190817832946777, "learning_rate": 9.59269473629344e-07, "loss": 1.4295, "step": 40828 }, { "epoch": 3.1997648902821316, "grad_norm": 5.498406410217285, "learning_rate": 9.590878615919108e-07, "loss": 0.9542, "step": 40829 }, { "epoch": 3.199843260188088, "grad_norm": 6.573281764984131, "learning_rate": 9.58906264924031e-07, "loss": 1.2749, "step": 40830 }, { "epoch": 3.199921630094044, "grad_norm": 2.9260997772216797, "learning_rate": 9.587246836263968e-07, "loss": 1.522, "step": 40831 }, { "epoch": 3.2, "grad_norm": 5.115076541900635, "learning_rate": 9.58543117699698e-07, "loss": 1.4107, "step": 40832 }, { "epoch": 3.200078369905956, "grad_norm": 4.434340476989746, "learning_rate": 9.58361567144625e-07, "loss": 1.8459, "step": 40833 }, { "epoch": 3.2001567398119124, "grad_norm": 2.3441689014434814, "learning_rate": 9.581800319618695e-07, "loss": 1.4055, "step": 40834 }, { "epoch": 3.2002351097178683, "grad_norm": 3.8709118366241455, "learning_rate": 9.579985121521202e-07, "loss": 1.833, "step": 40835 }, { "epoch": 3.2003134796238246, "grad_norm": 4.2642412185668945, "learning_rate": 9.578170077160693e-07, "loss": 1.5353, "step": 40836 }, { "epoch": 3.2003918495297805, "grad_norm": 2.9500396251678467, "learning_rate": 9.576355186544056e-07, "loss": 1.1732, "step": 40837 }, { "epoch": 3.200470219435737, "grad_norm": 3.7041211128234863, "learning_rate": 9.574540449678216e-07, "loss": 1.6217, "step": 40838 }, { "epoch": 3.2005485893416927, "grad_norm": 3.1327319145202637, "learning_rate": 9.572725866570037e-07, "loss": 1.4915, "step": 40839 }, { "epoch": 3.200626959247649, "grad_norm": 2.7212252616882324, "learning_rate": 9.570911437226454e-07, "loss": 1.0125, "step": 40840 }, { "epoch": 3.200705329153605, "grad_norm": 3.4629178047180176, "learning_rate": 9.569097161654361e-07, "loss": 1.4381, "step": 40841 }, { "epoch": 3.2007836990595613, "grad_norm": 3.6807498931884766, "learning_rate": 9.567283039860664e-07, "loss": 1.1253, "step": 40842 }, { "epoch": 3.200862068965517, "grad_norm": 3.868473529815674, "learning_rate": 9.565469071852246e-07, "loss": 1.5948, "step": 40843 }, { "epoch": 3.2009404388714735, "grad_norm": 2.676603078842163, "learning_rate": 9.563655257636e-07, "loss": 1.14, "step": 40844 }, { "epoch": 3.2010188087774294, "grad_norm": 6.096485137939453, "learning_rate": 9.561841597218869e-07, "loss": 1.9145, "step": 40845 }, { "epoch": 3.2010971786833857, "grad_norm": 4.078240871429443, "learning_rate": 9.56002809060771e-07, "loss": 1.6568, "step": 40846 }, { "epoch": 3.2011755485893416, "grad_norm": 2.7220962047576904, "learning_rate": 9.558214737809429e-07, "loss": 1.5083, "step": 40847 }, { "epoch": 3.201253918495298, "grad_norm": 9.699532508850098, "learning_rate": 9.55640153883093e-07, "loss": 1.5514, "step": 40848 }, { "epoch": 3.201332288401254, "grad_norm": 3.0554940700531006, "learning_rate": 9.554588493679102e-07, "loss": 1.0566, "step": 40849 }, { "epoch": 3.20141065830721, "grad_norm": 2.728282928466797, "learning_rate": 9.552775602360847e-07, "loss": 1.379, "step": 40850 }, { "epoch": 3.201489028213166, "grad_norm": 7.913081645965576, "learning_rate": 9.550962864883056e-07, "loss": 0.8021, "step": 40851 }, { "epoch": 3.2015673981191224, "grad_norm": 2.8345282077789307, "learning_rate": 9.549150281252633e-07, "loss": 1.45, "step": 40852 }, { "epoch": 3.2016457680250783, "grad_norm": 4.665539741516113, "learning_rate": 9.547337851476463e-07, "loss": 1.3905, "step": 40853 }, { "epoch": 3.2017241379310346, "grad_norm": 2.734241247177124, "learning_rate": 9.545525575561437e-07, "loss": 1.4946, "step": 40854 }, { "epoch": 3.2018025078369905, "grad_norm": 3.5950403213500977, "learning_rate": 9.543713453514459e-07, "loss": 1.3286, "step": 40855 }, { "epoch": 3.201880877742947, "grad_norm": 2.756789207458496, "learning_rate": 9.541901485342414e-07, "loss": 1.0958, "step": 40856 }, { "epoch": 3.2019592476489027, "grad_norm": 3.5134236812591553, "learning_rate": 9.540089671052194e-07, "loss": 1.2789, "step": 40857 }, { "epoch": 3.202037617554859, "grad_norm": 2.4183292388916016, "learning_rate": 9.538278010650693e-07, "loss": 1.297, "step": 40858 }, { "epoch": 3.202115987460815, "grad_norm": 2.8278465270996094, "learning_rate": 9.536466504144798e-07, "loss": 1.3227, "step": 40859 }, { "epoch": 3.2021943573667713, "grad_norm": 2.4569201469421387, "learning_rate": 9.534655151541417e-07, "loss": 1.4892, "step": 40860 }, { "epoch": 3.202272727272727, "grad_norm": 3.0815389156341553, "learning_rate": 9.532843952847409e-07, "loss": 1.3112, "step": 40861 }, { "epoch": 3.2023510971786835, "grad_norm": 4.164834499359131, "learning_rate": 9.531032908069665e-07, "loss": 1.3044, "step": 40862 }, { "epoch": 3.2024294670846394, "grad_norm": 5.360384464263916, "learning_rate": 9.529222017215095e-07, "loss": 1.4165, "step": 40863 }, { "epoch": 3.2025078369905957, "grad_norm": 2.3480570316314697, "learning_rate": 9.527411280290594e-07, "loss": 1.2661, "step": 40864 }, { "epoch": 3.2025862068965516, "grad_norm": 14.240938186645508, "learning_rate": 9.525600697303017e-07, "loss": 1.2588, "step": 40865 }, { "epoch": 3.202664576802508, "grad_norm": 3.611112356185913, "learning_rate": 9.52379026825927e-07, "loss": 1.283, "step": 40866 }, { "epoch": 3.202742946708464, "grad_norm": 3.3149280548095703, "learning_rate": 9.52197999316622e-07, "loss": 1.2663, "step": 40867 }, { "epoch": 3.20282131661442, "grad_norm": 3.1640336513519287, "learning_rate": 9.520169872030787e-07, "loss": 0.8502, "step": 40868 }, { "epoch": 3.202899686520376, "grad_norm": 4.046717643737793, "learning_rate": 9.518359904859831e-07, "loss": 0.9982, "step": 40869 }, { "epoch": 3.2029780564263324, "grad_norm": 5.332705020904541, "learning_rate": 9.516550091660237e-07, "loss": 1.6188, "step": 40870 }, { "epoch": 3.2030564263322883, "grad_norm": 3.641735315322876, "learning_rate": 9.514740432438896e-07, "loss": 1.072, "step": 40871 }, { "epoch": 3.2031347962382446, "grad_norm": 4.113489627838135, "learning_rate": 9.512930927202685e-07, "loss": 1.4036, "step": 40872 }, { "epoch": 3.2032131661442005, "grad_norm": 2.538867473602295, "learning_rate": 9.511121575958493e-07, "loss": 1.342, "step": 40873 }, { "epoch": 3.203291536050157, "grad_norm": 2.446362257003784, "learning_rate": 9.509312378713198e-07, "loss": 1.3479, "step": 40874 }, { "epoch": 3.2033699059561127, "grad_norm": 3.0318057537078857, "learning_rate": 9.507503335473689e-07, "loss": 1.3774, "step": 40875 }, { "epoch": 3.203448275862069, "grad_norm": 5.508615016937256, "learning_rate": 9.505694446246816e-07, "loss": 1.5245, "step": 40876 }, { "epoch": 3.203526645768025, "grad_norm": 4.616070747375488, "learning_rate": 9.503885711039496e-07, "loss": 1.0993, "step": 40877 }, { "epoch": 3.2036050156739813, "grad_norm": 6.779660701751709, "learning_rate": 9.502077129858595e-07, "loss": 1.5854, "step": 40878 }, { "epoch": 3.203683385579937, "grad_norm": 3.1512231826782227, "learning_rate": 9.500268702711002e-07, "loss": 1.332, "step": 40879 }, { "epoch": 3.2037617554858935, "grad_norm": 4.892230987548828, "learning_rate": 9.498460429603573e-07, "loss": 1.1198, "step": 40880 }, { "epoch": 3.2038401253918494, "grad_norm": 3.5834362506866455, "learning_rate": 9.496652310543186e-07, "loss": 1.1126, "step": 40881 }, { "epoch": 3.2039184952978057, "grad_norm": 3.707789897918701, "learning_rate": 9.494844345536741e-07, "loss": 0.6507, "step": 40882 }, { "epoch": 3.2039968652037616, "grad_norm": 10.219310760498047, "learning_rate": 9.493036534591122e-07, "loss": 0.3735, "step": 40883 }, { "epoch": 3.204075235109718, "grad_norm": 3.4301741123199463, "learning_rate": 9.491228877713165e-07, "loss": 1.0211, "step": 40884 }, { "epoch": 3.204153605015674, "grad_norm": 2.4911861419677734, "learning_rate": 9.489421374909768e-07, "loss": 1.4014, "step": 40885 }, { "epoch": 3.20423197492163, "grad_norm": 2.527708053588867, "learning_rate": 9.487614026187808e-07, "loss": 1.2696, "step": 40886 }, { "epoch": 3.204310344827586, "grad_norm": 3.2987184524536133, "learning_rate": 9.485806831554156e-07, "loss": 1.5766, "step": 40887 }, { "epoch": 3.2043887147335424, "grad_norm": 3.7879574298858643, "learning_rate": 9.483999791015686e-07, "loss": 1.6618, "step": 40888 }, { "epoch": 3.2044670846394983, "grad_norm": 4.646810531616211, "learning_rate": 9.482192904579268e-07, "loss": 0.8622, "step": 40889 }, { "epoch": 3.2045454545454546, "grad_norm": 2.541550397872925, "learning_rate": 9.480386172251776e-07, "loss": 0.7699, "step": 40890 }, { "epoch": 3.2046238244514105, "grad_norm": 3.5163302421569824, "learning_rate": 9.478579594040083e-07, "loss": 1.5412, "step": 40891 }, { "epoch": 3.204702194357367, "grad_norm": 3.19817852973938, "learning_rate": 9.476773169951064e-07, "loss": 1.3633, "step": 40892 }, { "epoch": 3.2047805642633227, "grad_norm": 3.2580416202545166, "learning_rate": 9.474966899991584e-07, "loss": 0.9329, "step": 40893 }, { "epoch": 3.204858934169279, "grad_norm": 2.845153570175171, "learning_rate": 9.473160784168528e-07, "loss": 0.9567, "step": 40894 }, { "epoch": 3.204937304075235, "grad_norm": 6.769885063171387, "learning_rate": 9.471354822488721e-07, "loss": 1.2258, "step": 40895 }, { "epoch": 3.2050156739811912, "grad_norm": 2.7429683208465576, "learning_rate": 9.469549014959084e-07, "loss": 0.9031, "step": 40896 }, { "epoch": 3.205094043887147, "grad_norm": 2.721115827560425, "learning_rate": 9.467743361586463e-07, "loss": 1.3839, "step": 40897 }, { "epoch": 3.2051724137931035, "grad_norm": 2.2535154819488525, "learning_rate": 9.465937862377739e-07, "loss": 1.2887, "step": 40898 }, { "epoch": 3.2052507836990594, "grad_norm": 5.683831691741943, "learning_rate": 9.464132517339753e-07, "loss": 0.5576, "step": 40899 }, { "epoch": 3.2053291536050157, "grad_norm": 2.93715238571167, "learning_rate": 9.462327326479376e-07, "loss": 1.6632, "step": 40900 }, { "epoch": 3.2054075235109716, "grad_norm": 4.300097942352295, "learning_rate": 9.460522289803509e-07, "loss": 1.186, "step": 40901 }, { "epoch": 3.205485893416928, "grad_norm": 4.586425304412842, "learning_rate": 9.458717407318979e-07, "loss": 1.67, "step": 40902 }, { "epoch": 3.2055642633228842, "grad_norm": 2.980490207672119, "learning_rate": 9.45691267903267e-07, "loss": 1.4222, "step": 40903 }, { "epoch": 3.20564263322884, "grad_norm": 2.6182963848114014, "learning_rate": 9.455108104951438e-07, "loss": 1.3833, "step": 40904 }, { "epoch": 3.205721003134796, "grad_norm": 39.704689025878906, "learning_rate": 9.453303685082149e-07, "loss": 1.8596, "step": 40905 }, { "epoch": 3.2057993730407524, "grad_norm": 4.556707859039307, "learning_rate": 9.451499419431665e-07, "loss": 1.3235, "step": 40906 }, { "epoch": 3.2058777429467087, "grad_norm": 3.200073719024658, "learning_rate": 9.449695308006851e-07, "loss": 1.1875, "step": 40907 }, { "epoch": 3.2059561128526646, "grad_norm": 7.7685370445251465, "learning_rate": 9.44789135081457e-07, "loss": 1.1542, "step": 40908 }, { "epoch": 3.2060344827586205, "grad_norm": 7.195158958435059, "learning_rate": 9.446087547861682e-07, "loss": 1.5694, "step": 40909 }, { "epoch": 3.206112852664577, "grad_norm": 2.47990345954895, "learning_rate": 9.444283899155043e-07, "loss": 1.313, "step": 40910 }, { "epoch": 3.206191222570533, "grad_norm": 2.5761466026306152, "learning_rate": 9.442480404701521e-07, "loss": 1.2038, "step": 40911 }, { "epoch": 3.206269592476489, "grad_norm": 2.6604151725769043, "learning_rate": 9.440677064507969e-07, "loss": 1.3571, "step": 40912 }, { "epoch": 3.2063479623824453, "grad_norm": 5.516537189483643, "learning_rate": 9.438873878581262e-07, "loss": 1.1443, "step": 40913 }, { "epoch": 3.2064263322884012, "grad_norm": 2.8884992599487305, "learning_rate": 9.437070846928221e-07, "loss": 1.4051, "step": 40914 }, { "epoch": 3.2065047021943576, "grad_norm": 2.535231351852417, "learning_rate": 9.435267969555745e-07, "loss": 1.5001, "step": 40915 }, { "epoch": 3.2065830721003135, "grad_norm": 3.778017520904541, "learning_rate": 9.43346524647068e-07, "loss": 1.2964, "step": 40916 }, { "epoch": 3.20666144200627, "grad_norm": 5.373080253601074, "learning_rate": 9.43166267767987e-07, "loss": 0.9169, "step": 40917 }, { "epoch": 3.2067398119122257, "grad_norm": 3.0450479984283447, "learning_rate": 9.429860263190171e-07, "loss": 1.0409, "step": 40918 }, { "epoch": 3.206818181818182, "grad_norm": 3.3804373741149902, "learning_rate": 9.428058003008439e-07, "loss": 1.4889, "step": 40919 }, { "epoch": 3.206896551724138, "grad_norm": 3.5351412296295166, "learning_rate": 9.426255897141557e-07, "loss": 1.2518, "step": 40920 }, { "epoch": 3.2069749216300942, "grad_norm": 3.5197272300720215, "learning_rate": 9.424453945596346e-07, "loss": 1.2345, "step": 40921 }, { "epoch": 3.20705329153605, "grad_norm": 5.319479942321777, "learning_rate": 9.422652148379675e-07, "loss": 0.4244, "step": 40922 }, { "epoch": 3.2071316614420065, "grad_norm": 7.425815105438232, "learning_rate": 9.420850505498386e-07, "loss": 1.5663, "step": 40923 }, { "epoch": 3.2072100313479623, "grad_norm": 4.02728796005249, "learning_rate": 9.419049016959342e-07, "loss": 1.46, "step": 40924 }, { "epoch": 3.2072884012539187, "grad_norm": 4.806111812591553, "learning_rate": 9.417247682769393e-07, "loss": 1.8341, "step": 40925 }, { "epoch": 3.2073667711598746, "grad_norm": 3.810516357421875, "learning_rate": 9.415446502935394e-07, "loss": 1.2429, "step": 40926 }, { "epoch": 3.207445141065831, "grad_norm": 2.651956796646118, "learning_rate": 9.413645477464184e-07, "loss": 0.9564, "step": 40927 }, { "epoch": 3.207523510971787, "grad_norm": 2.9178898334503174, "learning_rate": 9.411844606362619e-07, "loss": 1.4458, "step": 40928 }, { "epoch": 3.207601880877743, "grad_norm": 2.9970788955688477, "learning_rate": 9.410043889637555e-07, "loss": 1.4146, "step": 40929 }, { "epoch": 3.207680250783699, "grad_norm": 3.6608388423919678, "learning_rate": 9.408243327295835e-07, "loss": 1.1116, "step": 40930 }, { "epoch": 3.2077586206896553, "grad_norm": 2.5135669708251953, "learning_rate": 9.406442919344316e-07, "loss": 1.4255, "step": 40931 }, { "epoch": 3.2078369905956112, "grad_norm": 7.10870885848999, "learning_rate": 9.404642665789831e-07, "loss": 1.2972, "step": 40932 }, { "epoch": 3.2079153605015676, "grad_norm": 3.309056520462036, "learning_rate": 9.40284256663922e-07, "loss": 1.2903, "step": 40933 }, { "epoch": 3.2079937304075234, "grad_norm": 2.851562261581421, "learning_rate": 9.401042621899353e-07, "loss": 1.4967, "step": 40934 }, { "epoch": 3.20807210031348, "grad_norm": 3.030599355697632, "learning_rate": 9.399242831577082e-07, "loss": 1.5891, "step": 40935 }, { "epoch": 3.2081504702194357, "grad_norm": 2.746471405029297, "learning_rate": 9.397443195679224e-07, "loss": 1.1418, "step": 40936 }, { "epoch": 3.208228840125392, "grad_norm": 3.5449604988098145, "learning_rate": 9.395643714212638e-07, "loss": 1.079, "step": 40937 }, { "epoch": 3.208307210031348, "grad_norm": 2.875535249710083, "learning_rate": 9.393844387184159e-07, "loss": 1.1756, "step": 40938 }, { "epoch": 3.208385579937304, "grad_norm": 2.583122491836548, "learning_rate": 9.392045214600659e-07, "loss": 1.4062, "step": 40939 }, { "epoch": 3.20846394984326, "grad_norm": 2.855334758758545, "learning_rate": 9.390246196468949e-07, "loss": 1.3, "step": 40940 }, { "epoch": 3.2085423197492164, "grad_norm": 5.6631011962890625, "learning_rate": 9.388447332795886e-07, "loss": 1.2021, "step": 40941 }, { "epoch": 3.2086206896551723, "grad_norm": 7.229127883911133, "learning_rate": 9.386648623588312e-07, "loss": 1.2713, "step": 40942 }, { "epoch": 3.2086990595611287, "grad_norm": 3.8916237354278564, "learning_rate": 9.384850068853063e-07, "loss": 1.5671, "step": 40943 }, { "epoch": 3.2087774294670846, "grad_norm": 2.0651767253875732, "learning_rate": 9.383051668596982e-07, "loss": 1.0798, "step": 40944 }, { "epoch": 3.208855799373041, "grad_norm": 4.429615020751953, "learning_rate": 9.381253422826914e-07, "loss": 1.2181, "step": 40945 }, { "epoch": 3.2089341692789968, "grad_norm": 2.7088661193847656, "learning_rate": 9.379455331549692e-07, "loss": 1.3683, "step": 40946 }, { "epoch": 3.209012539184953, "grad_norm": 3.255280017852783, "learning_rate": 9.377657394772161e-07, "loss": 1.329, "step": 40947 }, { "epoch": 3.209090909090909, "grad_norm": 5.52875280380249, "learning_rate": 9.375859612501153e-07, "loss": 1.7309, "step": 40948 }, { "epoch": 3.2091692789968653, "grad_norm": 2.21429705619812, "learning_rate": 9.37406198474351e-07, "loss": 1.3343, "step": 40949 }, { "epoch": 3.209247648902821, "grad_norm": 12.40885066986084, "learning_rate": 9.37226451150608e-07, "loss": 0.7759, "step": 40950 }, { "epoch": 3.2093260188087775, "grad_norm": 14.064053535461426, "learning_rate": 9.370467192795679e-07, "loss": 1.1292, "step": 40951 }, { "epoch": 3.2094043887147334, "grad_norm": 2.6309237480163574, "learning_rate": 9.368670028619136e-07, "loss": 0.9918, "step": 40952 }, { "epoch": 3.2094827586206898, "grad_norm": 4.138621807098389, "learning_rate": 9.366873018983319e-07, "loss": 1.9331, "step": 40953 }, { "epoch": 3.2095611285266457, "grad_norm": 3.4667375087738037, "learning_rate": 9.365076163895054e-07, "loss": 1.1779, "step": 40954 }, { "epoch": 3.209639498432602, "grad_norm": 8.856796264648438, "learning_rate": 9.363279463361158e-07, "loss": 1.8209, "step": 40955 }, { "epoch": 3.209717868338558, "grad_norm": 2.33536434173584, "learning_rate": 9.361482917388475e-07, "loss": 1.8578, "step": 40956 }, { "epoch": 3.209796238244514, "grad_norm": 10.117867469787598, "learning_rate": 9.359686525983836e-07, "loss": 1.2088, "step": 40957 }, { "epoch": 3.20987460815047, "grad_norm": 3.2352511882781982, "learning_rate": 9.357890289154076e-07, "loss": 1.513, "step": 40958 }, { "epoch": 3.2099529780564264, "grad_norm": 4.249900817871094, "learning_rate": 9.35609420690603e-07, "loss": 1.0566, "step": 40959 }, { "epoch": 3.2100313479623823, "grad_norm": 3.984653949737549, "learning_rate": 9.35429827924652e-07, "loss": 1.1869, "step": 40960 }, { "epoch": 3.2101097178683387, "grad_norm": 2.941091299057007, "learning_rate": 9.352502506182387e-07, "loss": 1.1956, "step": 40961 }, { "epoch": 3.2101880877742945, "grad_norm": 3.5798537731170654, "learning_rate": 9.350706887720457e-07, "loss": 0.9925, "step": 40962 }, { "epoch": 3.210266457680251, "grad_norm": 3.5806057453155518, "learning_rate": 9.348911423867551e-07, "loss": 1.1882, "step": 40963 }, { "epoch": 3.2103448275862068, "grad_norm": 6.777584552764893, "learning_rate": 9.347116114630516e-07, "loss": 1.2659, "step": 40964 }, { "epoch": 3.210423197492163, "grad_norm": 2.264404296875, "learning_rate": 9.345320960016163e-07, "loss": 1.175, "step": 40965 }, { "epoch": 3.210501567398119, "grad_norm": 7.605413436889648, "learning_rate": 9.343525960031336e-07, "loss": 1.4501, "step": 40966 }, { "epoch": 3.2105799373040753, "grad_norm": 2.7077362537384033, "learning_rate": 9.341731114682845e-07, "loss": 1.1487, "step": 40967 }, { "epoch": 3.210658307210031, "grad_norm": 3.069427728652954, "learning_rate": 9.33993642397753e-07, "loss": 1.4488, "step": 40968 }, { "epoch": 3.2107366771159875, "grad_norm": 3.258147716522217, "learning_rate": 9.338141887922225e-07, "loss": 1.1963, "step": 40969 }, { "epoch": 3.2108150470219434, "grad_norm": 7.026670455932617, "learning_rate": 9.336347506523719e-07, "loss": 1.4273, "step": 40970 }, { "epoch": 3.2108934169278998, "grad_norm": 3.6231043338775635, "learning_rate": 9.334553279788872e-07, "loss": 1.326, "step": 40971 }, { "epoch": 3.2109717868338556, "grad_norm": 3.630866765975952, "learning_rate": 9.332759207724507e-07, "loss": 1.3218, "step": 40972 }, { "epoch": 3.211050156739812, "grad_norm": 2.8974452018737793, "learning_rate": 9.33096529033743e-07, "loss": 1.2367, "step": 40973 }, { "epoch": 3.211128526645768, "grad_norm": 6.344569206237793, "learning_rate": 9.329171527634467e-07, "loss": 0.945, "step": 40974 }, { "epoch": 3.211206896551724, "grad_norm": 4.206646919250488, "learning_rate": 9.327377919622438e-07, "loss": 1.3057, "step": 40975 }, { "epoch": 3.21128526645768, "grad_norm": 5.547720909118652, "learning_rate": 9.325584466308196e-07, "loss": 1.5181, "step": 40976 }, { "epoch": 3.2113636363636364, "grad_norm": 4.778224468231201, "learning_rate": 9.323791167698526e-07, "loss": 1.9869, "step": 40977 }, { "epoch": 3.2114420062695923, "grad_norm": 7.135971546173096, "learning_rate": 9.321998023800261e-07, "loss": 1.0187, "step": 40978 }, { "epoch": 3.2115203761755486, "grad_norm": 4.711702823638916, "learning_rate": 9.320205034620222e-07, "loss": 1.004, "step": 40979 }, { "epoch": 3.2115987460815045, "grad_norm": 3.6979870796203613, "learning_rate": 9.318412200165228e-07, "loss": 1.6436, "step": 40980 }, { "epoch": 3.211677115987461, "grad_norm": 6.670243740081787, "learning_rate": 9.316619520442105e-07, "loss": 1.1926, "step": 40981 }, { "epoch": 3.2117554858934168, "grad_norm": 3.3012731075286865, "learning_rate": 9.31482699545766e-07, "loss": 1.4346, "step": 40982 }, { "epoch": 3.211833855799373, "grad_norm": 5.331993103027344, "learning_rate": 9.313034625218714e-07, "loss": 0.8516, "step": 40983 }, { "epoch": 3.211912225705329, "grad_norm": 2.9389710426330566, "learning_rate": 9.311242409732085e-07, "loss": 1.3422, "step": 40984 }, { "epoch": 3.2119905956112853, "grad_norm": 4.178891658782959, "learning_rate": 9.309450349004595e-07, "loss": 1.5167, "step": 40985 }, { "epoch": 3.212068965517241, "grad_norm": 3.966008186340332, "learning_rate": 9.307658443043055e-07, "loss": 1.1452, "step": 40986 }, { "epoch": 3.2121473354231975, "grad_norm": 3.8467040061950684, "learning_rate": 9.305866691854299e-07, "loss": 1.3865, "step": 40987 }, { "epoch": 3.2122257053291534, "grad_norm": 3.7265865802764893, "learning_rate": 9.304075095445103e-07, "loss": 1.1494, "step": 40988 }, { "epoch": 3.2123040752351097, "grad_norm": 4.180886268615723, "learning_rate": 9.302283653822291e-07, "loss": 1.4753, "step": 40989 }, { "epoch": 3.2123824451410656, "grad_norm": 3.2723138332366943, "learning_rate": 9.300492366992708e-07, "loss": 1.3206, "step": 40990 }, { "epoch": 3.212460815047022, "grad_norm": 4.13305139541626, "learning_rate": 9.298701234963154e-07, "loss": 1.1754, "step": 40991 }, { "epoch": 3.212539184952978, "grad_norm": 3.3602406978607178, "learning_rate": 9.296910257740421e-07, "loss": 1.0887, "step": 40992 }, { "epoch": 3.212617554858934, "grad_norm": 2.665933609008789, "learning_rate": 9.29511943533134e-07, "loss": 1.3486, "step": 40993 }, { "epoch": 3.21269592476489, "grad_norm": 4.510025978088379, "learning_rate": 9.293328767742705e-07, "loss": 1.0103, "step": 40994 }, { "epoch": 3.2127742946708464, "grad_norm": 5.374619960784912, "learning_rate": 9.291538254981358e-07, "loss": 1.0601, "step": 40995 }, { "epoch": 3.2128526645768023, "grad_norm": 3.2213289737701416, "learning_rate": 9.289747897054086e-07, "loss": 1.534, "step": 40996 }, { "epoch": 3.2129310344827586, "grad_norm": 9.171723365783691, "learning_rate": 9.287957693967697e-07, "loss": 1.0757, "step": 40997 }, { "epoch": 3.213009404388715, "grad_norm": 3.20307993888855, "learning_rate": 9.286167645729005e-07, "loss": 1.1461, "step": 40998 }, { "epoch": 3.213087774294671, "grad_norm": 2.3876211643218994, "learning_rate": 9.284377752344825e-07, "loss": 1.1263, "step": 40999 }, { "epoch": 3.2131661442006267, "grad_norm": 3.35124135017395, "learning_rate": 9.282588013821953e-07, "loss": 1.5155, "step": 41000 }, { "epoch": 3.213244514106583, "grad_norm": 4.361406326293945, "learning_rate": 9.280798430167204e-07, "loss": 1.0614, "step": 41001 }, { "epoch": 3.2133228840125394, "grad_norm": 9.081550598144531, "learning_rate": 9.279009001387401e-07, "loss": 1.0594, "step": 41002 }, { "epoch": 3.2134012539184953, "grad_norm": 2.9108612537384033, "learning_rate": 9.2772197274893e-07, "loss": 1.3912, "step": 41003 }, { "epoch": 3.213479623824451, "grad_norm": 5.7705912590026855, "learning_rate": 9.275430608479757e-07, "loss": 1.4431, "step": 41004 }, { "epoch": 3.2135579937304075, "grad_norm": 7.389566898345947, "learning_rate": 9.273641644365555e-07, "loss": 1.0686, "step": 41005 }, { "epoch": 3.213636363636364, "grad_norm": 3.922990560531616, "learning_rate": 9.271852835153516e-07, "loss": 1.0992, "step": 41006 }, { "epoch": 3.2137147335423197, "grad_norm": 2.394113063812256, "learning_rate": 9.270064180850413e-07, "loss": 1.2699, "step": 41007 }, { "epoch": 3.213793103448276, "grad_norm": 2.6798787117004395, "learning_rate": 9.268275681463057e-07, "loss": 1.2121, "step": 41008 }, { "epoch": 3.213871473354232, "grad_norm": 3.619319200515747, "learning_rate": 9.26648733699827e-07, "loss": 1.2274, "step": 41009 }, { "epoch": 3.2139498432601883, "grad_norm": 2.9562363624572754, "learning_rate": 9.264699147462858e-07, "loss": 1.5748, "step": 41010 }, { "epoch": 3.214028213166144, "grad_norm": 2.3215625286102295, "learning_rate": 9.262911112863593e-07, "loss": 0.9911, "step": 41011 }, { "epoch": 3.2141065830721005, "grad_norm": 4.011744022369385, "learning_rate": 9.261123233207292e-07, "loss": 1.3217, "step": 41012 }, { "epoch": 3.2141849529780564, "grad_norm": 3.458292245864868, "learning_rate": 9.25933550850075e-07, "loss": 1.3068, "step": 41013 }, { "epoch": 3.2142633228840127, "grad_norm": 2.200650215148926, "learning_rate": 9.257547938750772e-07, "loss": 1.0606, "step": 41014 }, { "epoch": 3.2143416927899686, "grad_norm": 4.486688613891602, "learning_rate": 9.255760523964158e-07, "loss": 1.4173, "step": 41015 }, { "epoch": 3.214420062695925, "grad_norm": 5.890989780426025, "learning_rate": 9.253973264147698e-07, "loss": 1.3362, "step": 41016 }, { "epoch": 3.214498432601881, "grad_norm": 4.792747497558594, "learning_rate": 9.252186159308197e-07, "loss": 1.6301, "step": 41017 }, { "epoch": 3.214576802507837, "grad_norm": 10.018237113952637, "learning_rate": 9.250399209452448e-07, "loss": 1.4247, "step": 41018 }, { "epoch": 3.214655172413793, "grad_norm": 4.1235880851745605, "learning_rate": 9.248612414587254e-07, "loss": 1.4985, "step": 41019 }, { "epoch": 3.2147335423197494, "grad_norm": 3.37383770942688, "learning_rate": 9.246825774719409e-07, "loss": 1.8232, "step": 41020 }, { "epoch": 3.2148119122257053, "grad_norm": 3.6476378440856934, "learning_rate": 9.245039289855717e-07, "loss": 0.9208, "step": 41021 }, { "epoch": 3.2148902821316616, "grad_norm": 5.337749004364014, "learning_rate": 9.243252960002936e-07, "loss": 1.68, "step": 41022 }, { "epoch": 3.2149686520376175, "grad_norm": 2.677783966064453, "learning_rate": 9.241466785167901e-07, "loss": 1.3846, "step": 41023 }, { "epoch": 3.215047021943574, "grad_norm": 3.335141181945801, "learning_rate": 9.239680765357389e-07, "loss": 0.8998, "step": 41024 }, { "epoch": 3.2151253918495297, "grad_norm": 2.205568313598633, "learning_rate": 9.23789490057821e-07, "loss": 1.1444, "step": 41025 }, { "epoch": 3.215203761755486, "grad_norm": 3.069910764694214, "learning_rate": 9.236109190837128e-07, "loss": 1.258, "step": 41026 }, { "epoch": 3.215282131661442, "grad_norm": 3.971611499786377, "learning_rate": 9.23432363614094e-07, "loss": 1.4921, "step": 41027 }, { "epoch": 3.2153605015673983, "grad_norm": 2.538311004638672, "learning_rate": 9.232538236496469e-07, "loss": 1.4919, "step": 41028 }, { "epoch": 3.215438871473354, "grad_norm": 3.942718505859375, "learning_rate": 9.230752991910469e-07, "loss": 1.2784, "step": 41029 }, { "epoch": 3.2155172413793105, "grad_norm": 2.9739997386932373, "learning_rate": 9.228967902389746e-07, "loss": 1.314, "step": 41030 }, { "epoch": 3.2155956112852664, "grad_norm": 5.306634902954102, "learning_rate": 9.22718296794109e-07, "loss": 1.3165, "step": 41031 }, { "epoch": 3.2156739811912227, "grad_norm": 9.967297554016113, "learning_rate": 9.225398188571283e-07, "loss": 0.7235, "step": 41032 }, { "epoch": 3.2157523510971786, "grad_norm": 5.184664249420166, "learning_rate": 9.22361356428712e-07, "loss": 1.5829, "step": 41033 }, { "epoch": 3.215830721003135, "grad_norm": 3.3220086097717285, "learning_rate": 9.221829095095386e-07, "loss": 1.372, "step": 41034 }, { "epoch": 3.215909090909091, "grad_norm": 4.2437744140625, "learning_rate": 9.220044781002874e-07, "loss": 1.4879, "step": 41035 }, { "epoch": 3.215987460815047, "grad_norm": 7.011501789093018, "learning_rate": 9.218260622016362e-07, "loss": 1.5136, "step": 41036 }, { "epoch": 3.216065830721003, "grad_norm": 3.486212730407715, "learning_rate": 9.216476618142639e-07, "loss": 1.2196, "step": 41037 }, { "epoch": 3.2161442006269594, "grad_norm": 4.912768840789795, "learning_rate": 9.214692769388489e-07, "loss": 1.2243, "step": 41038 }, { "epoch": 3.2162225705329153, "grad_norm": 2.7416725158691406, "learning_rate": 9.212909075760718e-07, "loss": 1.295, "step": 41039 }, { "epoch": 3.2163009404388716, "grad_norm": 3.738711357116699, "learning_rate": 9.211125537266075e-07, "loss": 1.4098, "step": 41040 }, { "epoch": 3.2163793103448275, "grad_norm": 4.398000717163086, "learning_rate": 9.209342153911349e-07, "loss": 0.9547, "step": 41041 }, { "epoch": 3.216457680250784, "grad_norm": 2.8200197219848633, "learning_rate": 9.207558925703342e-07, "loss": 1.1351, "step": 41042 }, { "epoch": 3.2165360501567397, "grad_norm": 3.946620464324951, "learning_rate": 9.205775852648846e-07, "loss": 1.5235, "step": 41043 }, { "epoch": 3.216614420062696, "grad_norm": 2.3226709365844727, "learning_rate": 9.203992934754613e-07, "loss": 1.4537, "step": 41044 }, { "epoch": 3.216692789968652, "grad_norm": 3.781114339828491, "learning_rate": 9.202210172027432e-07, "loss": 1.689, "step": 41045 }, { "epoch": 3.2167711598746083, "grad_norm": 3.6885972023010254, "learning_rate": 9.20042756447408e-07, "loss": 1.2165, "step": 41046 }, { "epoch": 3.216849529780564, "grad_norm": 2.3761138916015625, "learning_rate": 9.198645112101373e-07, "loss": 1.6293, "step": 41047 }, { "epoch": 3.2169278996865205, "grad_norm": 3.0461807250976562, "learning_rate": 9.196862814916046e-07, "loss": 1.325, "step": 41048 }, { "epoch": 3.2170062695924764, "grad_norm": 2.3419811725616455, "learning_rate": 9.195080672924895e-07, "loss": 1.0299, "step": 41049 }, { "epoch": 3.2170846394984327, "grad_norm": 3.9985899925231934, "learning_rate": 9.193298686134699e-07, "loss": 1.3457, "step": 41050 }, { "epoch": 3.2171630094043886, "grad_norm": 3.509800910949707, "learning_rate": 9.191516854552235e-07, "loss": 1.1544, "step": 41051 }, { "epoch": 3.217241379310345, "grad_norm": 7.205834865570068, "learning_rate": 9.189735178184284e-07, "loss": 0.9272, "step": 41052 }, { "epoch": 3.217319749216301, "grad_norm": 5.125765323638916, "learning_rate": 9.187953657037613e-07, "loss": 1.1585, "step": 41053 }, { "epoch": 3.217398119122257, "grad_norm": 3.277557611465454, "learning_rate": 9.186172291119011e-07, "loss": 1.2933, "step": 41054 }, { "epoch": 3.217476489028213, "grad_norm": 2.623758316040039, "learning_rate": 9.18439108043524e-07, "loss": 1.1313, "step": 41055 }, { "epoch": 3.2175548589341694, "grad_norm": 3.1630218029022217, "learning_rate": 9.182610024993083e-07, "loss": 1.628, "step": 41056 }, { "epoch": 3.2176332288401253, "grad_norm": 6.265436172485352, "learning_rate": 9.180829124799312e-07, "loss": 1.6034, "step": 41057 }, { "epoch": 3.2177115987460816, "grad_norm": 2.7064125537872314, "learning_rate": 9.179048379860716e-07, "loss": 1.2328, "step": 41058 }, { "epoch": 3.2177899686520375, "grad_norm": 3.0451223850250244, "learning_rate": 9.177267790184025e-07, "loss": 1.4785, "step": 41059 }, { "epoch": 3.217868338557994, "grad_norm": 2.4639029502868652, "learning_rate": 9.175487355776053e-07, "loss": 1.561, "step": 41060 }, { "epoch": 3.2179467084639497, "grad_norm": 5.8323798179626465, "learning_rate": 9.173707076643556e-07, "loss": 1.462, "step": 41061 }, { "epoch": 3.218025078369906, "grad_norm": 2.772259473800659, "learning_rate": 9.171926952793326e-07, "loss": 1.3615, "step": 41062 }, { "epoch": 3.218103448275862, "grad_norm": 3.114373207092285, "learning_rate": 9.170146984232098e-07, "loss": 1.4825, "step": 41063 }, { "epoch": 3.2181818181818183, "grad_norm": 3.683122158050537, "learning_rate": 9.168367170966646e-07, "loss": 1.5095, "step": 41064 }, { "epoch": 3.218260188087774, "grad_norm": 3.123413562774658, "learning_rate": 9.166587513003772e-07, "loss": 1.6254, "step": 41065 }, { "epoch": 3.2183385579937305, "grad_norm": 4.21044397354126, "learning_rate": 9.164808010350229e-07, "loss": 1.2754, "step": 41066 }, { "epoch": 3.2184169278996864, "grad_norm": 2.58652925491333, "learning_rate": 9.163028663012774e-07, "loss": 1.0341, "step": 41067 }, { "epoch": 3.2184952978056427, "grad_norm": 3.9688682556152344, "learning_rate": 9.161249470998185e-07, "loss": 1.2226, "step": 41068 }, { "epoch": 3.2185736677115986, "grad_norm": 2.786630630493164, "learning_rate": 9.159470434313228e-07, "loss": 1.4307, "step": 41069 }, { "epoch": 3.218652037617555, "grad_norm": 2.2010903358459473, "learning_rate": 9.157691552964665e-07, "loss": 1.0136, "step": 41070 }, { "epoch": 3.218730407523511, "grad_norm": 5.289573669433594, "learning_rate": 9.155912826959262e-07, "loss": 1.6787, "step": 41071 }, { "epoch": 3.218808777429467, "grad_norm": 2.360917329788208, "learning_rate": 9.154134256303792e-07, "loss": 1.3917, "step": 41072 }, { "epoch": 3.218887147335423, "grad_norm": 2.941929578781128, "learning_rate": 9.152355841005017e-07, "loss": 1.1278, "step": 41073 }, { "epoch": 3.2189655172413794, "grad_norm": 8.917585372924805, "learning_rate": 9.150577581069697e-07, "loss": 1.3833, "step": 41074 }, { "epoch": 3.2190438871473352, "grad_norm": 6.2538743019104, "learning_rate": 9.148799476504599e-07, "loss": 1.2722, "step": 41075 }, { "epoch": 3.2191222570532916, "grad_norm": 3.5991950035095215, "learning_rate": 9.147021527316485e-07, "loss": 1.6148, "step": 41076 }, { "epoch": 3.2192006269592475, "grad_norm": 3.2485146522521973, "learning_rate": 9.145243733512132e-07, "loss": 1.5403, "step": 41077 }, { "epoch": 3.219278996865204, "grad_norm": 3.328495502471924, "learning_rate": 9.143466095098258e-07, "loss": 1.531, "step": 41078 }, { "epoch": 3.2193573667711597, "grad_norm": 3.025895595550537, "learning_rate": 9.141688612081673e-07, "loss": 1.422, "step": 41079 }, { "epoch": 3.219435736677116, "grad_norm": 2.2112956047058105, "learning_rate": 9.139911284469111e-07, "loss": 1.0724, "step": 41080 }, { "epoch": 3.219514106583072, "grad_norm": 3.011547327041626, "learning_rate": 9.138134112267355e-07, "loss": 1.2172, "step": 41081 }, { "epoch": 3.2195924764890282, "grad_norm": 13.263806343078613, "learning_rate": 9.136357095483139e-07, "loss": 1.1207, "step": 41082 }, { "epoch": 3.219670846394984, "grad_norm": 4.239315509796143, "learning_rate": 9.134580234123213e-07, "loss": 1.5576, "step": 41083 }, { "epoch": 3.2197492163009405, "grad_norm": 2.742893934249878, "learning_rate": 9.132803528194384e-07, "loss": 1.3038, "step": 41084 }, { "epoch": 3.2198275862068964, "grad_norm": 2.5402650833129883, "learning_rate": 9.131026977703367e-07, "loss": 1.2244, "step": 41085 }, { "epoch": 3.2199059561128527, "grad_norm": 2.8022892475128174, "learning_rate": 9.129250582656934e-07, "loss": 1.2643, "step": 41086 }, { "epoch": 3.2199843260188086, "grad_norm": 3.251969575881958, "learning_rate": 9.127474343061832e-07, "loss": 1.3982, "step": 41087 }, { "epoch": 3.220062695924765, "grad_norm": 2.4769270420074463, "learning_rate": 9.125698258924831e-07, "loss": 1.4231, "step": 41088 }, { "epoch": 3.2201410658307212, "grad_norm": 3.0802714824676514, "learning_rate": 9.123922330252677e-07, "loss": 1.5474, "step": 41089 }, { "epoch": 3.220219435736677, "grad_norm": 3.633582592010498, "learning_rate": 9.122146557052125e-07, "loss": 1.3118, "step": 41090 }, { "epoch": 3.220297805642633, "grad_norm": 4.048192024230957, "learning_rate": 9.120370939329937e-07, "loss": 1.3885, "step": 41091 }, { "epoch": 3.2203761755485893, "grad_norm": 2.108556032180786, "learning_rate": 9.118595477092856e-07, "loss": 1.4368, "step": 41092 }, { "epoch": 3.2204545454545457, "grad_norm": 3.3433456420898438, "learning_rate": 9.11682017034764e-07, "loss": 1.0991, "step": 41093 }, { "epoch": 3.2205329153605016, "grad_norm": 3.061891794204712, "learning_rate": 9.115045019101043e-07, "loss": 1.3712, "step": 41094 }, { "epoch": 3.2206112852664575, "grad_norm": 6.399927616119385, "learning_rate": 9.113270023359827e-07, "loss": 1.3379, "step": 41095 }, { "epoch": 3.220689655172414, "grad_norm": 2.9631755352020264, "learning_rate": 9.111495183130714e-07, "loss": 1.6016, "step": 41096 }, { "epoch": 3.22076802507837, "grad_norm": 6.277077674865723, "learning_rate": 9.109720498420465e-07, "loss": 1.6864, "step": 41097 }, { "epoch": 3.220846394984326, "grad_norm": 3.5895943641662598, "learning_rate": 9.107945969235848e-07, "loss": 1.2401, "step": 41098 }, { "epoch": 3.220924764890282, "grad_norm": 3.913452625274658, "learning_rate": 9.106171595583613e-07, "loss": 1.1576, "step": 41099 }, { "epoch": 3.2210031347962382, "grad_norm": 2.3959689140319824, "learning_rate": 9.104397377470486e-07, "loss": 1.4252, "step": 41100 }, { "epoch": 3.2210815047021946, "grad_norm": 12.339519500732422, "learning_rate": 9.10262331490322e-07, "loss": 0.9767, "step": 41101 }, { "epoch": 3.2211598746081505, "grad_norm": 3.4154083728790283, "learning_rate": 9.100849407888563e-07, "loss": 1.3114, "step": 41102 }, { "epoch": 3.221238244514107, "grad_norm": 3.476529121398926, "learning_rate": 9.099075656433287e-07, "loss": 1.4091, "step": 41103 }, { "epoch": 3.2213166144200627, "grad_norm": 4.795661449432373, "learning_rate": 9.097302060544111e-07, "loss": 1.5165, "step": 41104 }, { "epoch": 3.221394984326019, "grad_norm": 8.355303764343262, "learning_rate": 9.095528620227789e-07, "loss": 2.0833, "step": 41105 }, { "epoch": 3.221473354231975, "grad_norm": 3.2476279735565186, "learning_rate": 9.093755335491067e-07, "loss": 1.3601, "step": 41106 }, { "epoch": 3.2215517241379312, "grad_norm": 4.3989644050598145, "learning_rate": 9.091982206340682e-07, "loss": 1.2122, "step": 41107 }, { "epoch": 3.221630094043887, "grad_norm": 4.066089153289795, "learning_rate": 9.090209232783392e-07, "loss": 1.359, "step": 41108 }, { "epoch": 3.2217084639498434, "grad_norm": 5.1317267417907715, "learning_rate": 9.088436414825935e-07, "loss": 1.6553, "step": 41109 }, { "epoch": 3.2217868338557993, "grad_norm": 4.239708423614502, "learning_rate": 9.086663752475061e-07, "loss": 1.712, "step": 41110 }, { "epoch": 3.2218652037617557, "grad_norm": 3.315213918685913, "learning_rate": 9.084891245737476e-07, "loss": 1.1563, "step": 41111 }, { "epoch": 3.2219435736677116, "grad_norm": 3.586602210998535, "learning_rate": 9.083118894619968e-07, "loss": 1.304, "step": 41112 }, { "epoch": 3.222021943573668, "grad_norm": 4.414312362670898, "learning_rate": 9.081346699129257e-07, "loss": 1.5382, "step": 41113 }, { "epoch": 3.222100313479624, "grad_norm": 5.697346210479736, "learning_rate": 9.079574659272095e-07, "loss": 1.3292, "step": 41114 }, { "epoch": 3.22217868338558, "grad_norm": 4.430464744567871, "learning_rate": 9.077802775055205e-07, "loss": 1.1871, "step": 41115 }, { "epoch": 3.222257053291536, "grad_norm": 4.48390531539917, "learning_rate": 9.076031046485318e-07, "loss": 1.3067, "step": 41116 }, { "epoch": 3.2223354231974923, "grad_norm": 4.792320251464844, "learning_rate": 9.074259473569203e-07, "loss": 1.1314, "step": 41117 }, { "epoch": 3.222413793103448, "grad_norm": 2.977400302886963, "learning_rate": 9.0724880563136e-07, "loss": 1.2232, "step": 41118 }, { "epoch": 3.2224921630094046, "grad_norm": 12.652787208557129, "learning_rate": 9.070716794725215e-07, "loss": 2.2694, "step": 41119 }, { "epoch": 3.2225705329153604, "grad_norm": 5.386817932128906, "learning_rate": 9.068945688810799e-07, "loss": 1.2954, "step": 41120 }, { "epoch": 3.2226489028213168, "grad_norm": 5.420428276062012, "learning_rate": 9.067174738577083e-07, "loss": 1.1932, "step": 41121 }, { "epoch": 3.2227272727272727, "grad_norm": 2.8440725803375244, "learning_rate": 9.065403944030832e-07, "loss": 1.3008, "step": 41122 }, { "epoch": 3.222805642633229, "grad_norm": 4.608367443084717, "learning_rate": 9.063633305178743e-07, "loss": 1.3695, "step": 41123 }, { "epoch": 3.222884012539185, "grad_norm": 3.4198074340820312, "learning_rate": 9.061862822027572e-07, "loss": 1.0715, "step": 41124 }, { "epoch": 3.222962382445141, "grad_norm": 2.6341559886932373, "learning_rate": 9.060092494584044e-07, "loss": 1.3831, "step": 41125 }, { "epoch": 3.223040752351097, "grad_norm": 2.7988083362579346, "learning_rate": 9.058322322854901e-07, "loss": 1.1901, "step": 41126 }, { "epoch": 3.2231191222570534, "grad_norm": 3.1249916553497314, "learning_rate": 9.056552306846867e-07, "loss": 1.5094, "step": 41127 }, { "epoch": 3.2231974921630093, "grad_norm": 3.896700620651245, "learning_rate": 9.054782446566679e-07, "loss": 1.2771, "step": 41128 }, { "epoch": 3.2232758620689657, "grad_norm": 2.8208978176116943, "learning_rate": 9.053012742021078e-07, "loss": 1.4556, "step": 41129 }, { "epoch": 3.2233542319749215, "grad_norm": 5.607213497161865, "learning_rate": 9.051243193216764e-07, "loss": 1.0208, "step": 41130 }, { "epoch": 3.223432601880878, "grad_norm": 4.985348701477051, "learning_rate": 9.049473800160496e-07, "loss": 1.0263, "step": 41131 }, { "epoch": 3.2235109717868338, "grad_norm": 8.6368408203125, "learning_rate": 9.047704562859e-07, "loss": 1.7761, "step": 41132 }, { "epoch": 3.22358934169279, "grad_norm": 4.438934326171875, "learning_rate": 9.045935481319012e-07, "loss": 1.1373, "step": 41133 }, { "epoch": 3.223667711598746, "grad_norm": 2.1137735843658447, "learning_rate": 9.04416655554724e-07, "loss": 1.4493, "step": 41134 }, { "epoch": 3.2237460815047023, "grad_norm": 6.121195316314697, "learning_rate": 9.042397785550405e-07, "loss": 1.3712, "step": 41135 }, { "epoch": 3.223824451410658, "grad_norm": 6.827805042266846, "learning_rate": 9.040629171335269e-07, "loss": 1.3209, "step": 41136 }, { "epoch": 3.2239028213166145, "grad_norm": 2.9950406551361084, "learning_rate": 9.038860712908554e-07, "loss": 1.434, "step": 41137 }, { "epoch": 3.2239811912225704, "grad_norm": 5.472174644470215, "learning_rate": 9.03709241027696e-07, "loss": 1.1654, "step": 41138 }, { "epoch": 3.2240595611285268, "grad_norm": 6.2565999031066895, "learning_rate": 9.035324263447226e-07, "loss": 1.0288, "step": 41139 }, { "epoch": 3.2241379310344827, "grad_norm": 7.788813591003418, "learning_rate": 9.033556272426075e-07, "loss": 0.9209, "step": 41140 }, { "epoch": 3.224216300940439, "grad_norm": 4.255237102508545, "learning_rate": 9.03178843722024e-07, "loss": 1.4048, "step": 41141 }, { "epoch": 3.224294670846395, "grad_norm": 5.714826583862305, "learning_rate": 9.030020757836433e-07, "loss": 1.2356, "step": 41142 }, { "epoch": 3.224373040752351, "grad_norm": 4.526648998260498, "learning_rate": 9.028253234281386e-07, "loss": 1.0838, "step": 41143 }, { "epoch": 3.224451410658307, "grad_norm": 11.556733131408691, "learning_rate": 9.026485866561823e-07, "loss": 1.0726, "step": 41144 }, { "epoch": 3.2245297805642634, "grad_norm": 3.267362594604492, "learning_rate": 9.024718654684461e-07, "loss": 1.2616, "step": 41145 }, { "epoch": 3.2246081504702193, "grad_norm": 3.318615674972534, "learning_rate": 9.022951598656022e-07, "loss": 1.2116, "step": 41146 }, { "epoch": 3.2246865203761756, "grad_norm": 8.25076961517334, "learning_rate": 9.021184698483226e-07, "loss": 1.2607, "step": 41147 }, { "epoch": 3.2247648902821315, "grad_norm": 4.076715469360352, "learning_rate": 9.01941795417281e-07, "loss": 1.2133, "step": 41148 }, { "epoch": 3.224843260188088, "grad_norm": 3.8853659629821777, "learning_rate": 9.01765136573145e-07, "loss": 1.1715, "step": 41149 }, { "epoch": 3.2249216300940438, "grad_norm": 3.259282350540161, "learning_rate": 9.015884933165913e-07, "loss": 1.3887, "step": 41150 }, { "epoch": 3.225, "grad_norm": 1.991424322128296, "learning_rate": 9.014118656482907e-07, "loss": 1.2259, "step": 41151 }, { "epoch": 3.225078369905956, "grad_norm": 4.253683090209961, "learning_rate": 9.012352535689134e-07, "loss": 1.047, "step": 41152 }, { "epoch": 3.2251567398119123, "grad_norm": 3.9762518405914307, "learning_rate": 9.010586570791313e-07, "loss": 1.8689, "step": 41153 }, { "epoch": 3.225235109717868, "grad_norm": 9.850369453430176, "learning_rate": 9.008820761796161e-07, "loss": 1.2432, "step": 41154 }, { "epoch": 3.2253134796238245, "grad_norm": 3.815096378326416, "learning_rate": 9.007055108710422e-07, "loss": 1.6471, "step": 41155 }, { "epoch": 3.2253918495297804, "grad_norm": 2.938319683074951, "learning_rate": 9.00528961154078e-07, "loss": 1.2415, "step": 41156 }, { "epoch": 3.2254702194357368, "grad_norm": 2.708155632019043, "learning_rate": 9.003524270293956e-07, "loss": 1.4045, "step": 41157 }, { "epoch": 3.2255485893416926, "grad_norm": 3.5292792320251465, "learning_rate": 9.001759084976675e-07, "loss": 1.0136, "step": 41158 }, { "epoch": 3.225626959247649, "grad_norm": 4.771613121032715, "learning_rate": 8.99999405559564e-07, "loss": 1.4511, "step": 41159 }, { "epoch": 3.225705329153605, "grad_norm": 6.55148983001709, "learning_rate": 8.998229182157569e-07, "loss": 1.1171, "step": 41160 }, { "epoch": 3.225783699059561, "grad_norm": 3.404043674468994, "learning_rate": 8.996464464669174e-07, "loss": 1.1541, "step": 41161 }, { "epoch": 3.225862068965517, "grad_norm": 2.8747828006744385, "learning_rate": 8.994699903137167e-07, "loss": 1.4063, "step": 41162 }, { "epoch": 3.2259404388714734, "grad_norm": 2.7200942039489746, "learning_rate": 8.992935497568262e-07, "loss": 1.3229, "step": 41163 }, { "epoch": 3.2260188087774293, "grad_norm": 5.905673980712891, "learning_rate": 8.991171247969166e-07, "loss": 1.3084, "step": 41164 }, { "epoch": 3.2260971786833856, "grad_norm": 3.0157346725463867, "learning_rate": 8.989407154346591e-07, "loss": 1.1521, "step": 41165 }, { "epoch": 3.2261755485893415, "grad_norm": 2.6728498935699463, "learning_rate": 8.987643216707259e-07, "loss": 1.3983, "step": 41166 }, { "epoch": 3.226253918495298, "grad_norm": 3.945770025253296, "learning_rate": 8.985879435057843e-07, "loss": 1.6599, "step": 41167 }, { "epoch": 3.2263322884012537, "grad_norm": 4.220223426818848, "learning_rate": 8.984115809405087e-07, "loss": 1.1696, "step": 41168 }, { "epoch": 3.22641065830721, "grad_norm": 2.9216201305389404, "learning_rate": 8.98235233975569e-07, "loss": 1.6878, "step": 41169 }, { "epoch": 3.226489028213166, "grad_norm": 5.218815803527832, "learning_rate": 8.980589026116365e-07, "loss": 1.4332, "step": 41170 }, { "epoch": 3.2265673981191223, "grad_norm": 2.7016279697418213, "learning_rate": 8.978825868493801e-07, "loss": 1.1119, "step": 41171 }, { "epoch": 3.226645768025078, "grad_norm": 3.3582189083099365, "learning_rate": 8.977062866894704e-07, "loss": 1.3025, "step": 41172 }, { "epoch": 3.2267241379310345, "grad_norm": 4.5958147048950195, "learning_rate": 8.975300021325795e-07, "loss": 0.7422, "step": 41173 }, { "epoch": 3.2268025078369904, "grad_norm": 5.311827182769775, "learning_rate": 8.973537331793791e-07, "loss": 1.5167, "step": 41174 }, { "epoch": 3.2268808777429467, "grad_norm": 3.3162364959716797, "learning_rate": 8.97177479830536e-07, "loss": 1.3594, "step": 41175 }, { "epoch": 3.2269592476489026, "grad_norm": 2.9812278747558594, "learning_rate": 8.97001242086723e-07, "loss": 1.2731, "step": 41176 }, { "epoch": 3.227037617554859, "grad_norm": 5.559677600860596, "learning_rate": 8.968250199486083e-07, "loss": 0.7055, "step": 41177 }, { "epoch": 3.227115987460815, "grad_norm": 8.85136604309082, "learning_rate": 8.966488134168666e-07, "loss": 0.9971, "step": 41178 }, { "epoch": 3.227194357366771, "grad_norm": 6.712930202484131, "learning_rate": 8.964726224921633e-07, "loss": 2.0022, "step": 41179 }, { "epoch": 3.227272727272727, "grad_norm": 8.871367454528809, "learning_rate": 8.962964471751712e-07, "loss": 0.9957, "step": 41180 }, { "epoch": 3.2273510971786834, "grad_norm": 17.2023983001709, "learning_rate": 8.961202874665587e-07, "loss": 1.9669, "step": 41181 }, { "epoch": 3.2274294670846393, "grad_norm": 3.855131149291992, "learning_rate": 8.959441433669975e-07, "loss": 1.1415, "step": 41182 }, { "epoch": 3.2275078369905956, "grad_norm": 3.261291980743408, "learning_rate": 8.957680148771564e-07, "loss": 1.3658, "step": 41183 }, { "epoch": 3.227586206896552, "grad_norm": 3.2842979431152344, "learning_rate": 8.95591901997706e-07, "loss": 0.9052, "step": 41184 }, { "epoch": 3.227664576802508, "grad_norm": 6.1355180740356445, "learning_rate": 8.954158047293172e-07, "loss": 1.2546, "step": 41185 }, { "epoch": 3.2277429467084637, "grad_norm": 2.470062017440796, "learning_rate": 8.95239723072655e-07, "loss": 1.2662, "step": 41186 }, { "epoch": 3.22782131661442, "grad_norm": 2.4597442150115967, "learning_rate": 8.95063657028395e-07, "loss": 1.6182, "step": 41187 }, { "epoch": 3.2278996865203764, "grad_norm": 3.3647987842559814, "learning_rate": 8.948876065972034e-07, "loss": 0.9908, "step": 41188 }, { "epoch": 3.2279780564263323, "grad_norm": 3.0091545581817627, "learning_rate": 8.947115717797527e-07, "loss": 1.4258, "step": 41189 }, { "epoch": 3.228056426332288, "grad_norm": 15.2758207321167, "learning_rate": 8.945355525767091e-07, "loss": 1.5031, "step": 41190 }, { "epoch": 3.2281347962382445, "grad_norm": 4.256289958953857, "learning_rate": 8.943595489887425e-07, "loss": 1.3375, "step": 41191 }, { "epoch": 3.228213166144201, "grad_norm": 2.266162157058716, "learning_rate": 8.941835610165245e-07, "loss": 1.2794, "step": 41192 }, { "epoch": 3.2282915360501567, "grad_norm": 4.568036079406738, "learning_rate": 8.940075886607247e-07, "loss": 1.4782, "step": 41193 }, { "epoch": 3.2283699059561126, "grad_norm": 2.361823081970215, "learning_rate": 8.938316319220098e-07, "loss": 1.2705, "step": 41194 }, { "epoch": 3.228448275862069, "grad_norm": 4.0277323722839355, "learning_rate": 8.936556908010502e-07, "loss": 1.2193, "step": 41195 }, { "epoch": 3.2285266457680253, "grad_norm": 2.4280872344970703, "learning_rate": 8.934797652985156e-07, "loss": 1.2662, "step": 41196 }, { "epoch": 3.228605015673981, "grad_norm": 4.3244147300720215, "learning_rate": 8.933038554150741e-07, "loss": 1.5034, "step": 41197 }, { "epoch": 3.2286833855799375, "grad_norm": 3.2769341468811035, "learning_rate": 8.93127961151396e-07, "loss": 1.1704, "step": 41198 }, { "epoch": 3.2287617554858934, "grad_norm": 3.464945077896118, "learning_rate": 8.929520825081495e-07, "loss": 1.2443, "step": 41199 }, { "epoch": 3.2288401253918497, "grad_norm": 3.766890525817871, "learning_rate": 8.927762194860034e-07, "loss": 0.9028, "step": 41200 }, { "epoch": 3.2289184952978056, "grad_norm": 4.428264617919922, "learning_rate": 8.926003720856274e-07, "loss": 1.133, "step": 41201 }, { "epoch": 3.228996865203762, "grad_norm": 4.4634881019592285, "learning_rate": 8.924245403076898e-07, "loss": 1.7219, "step": 41202 }, { "epoch": 3.229075235109718, "grad_norm": 4.012232303619385, "learning_rate": 8.922487241528593e-07, "loss": 1.302, "step": 41203 }, { "epoch": 3.229153605015674, "grad_norm": 16.590116500854492, "learning_rate": 8.920729236218062e-07, "loss": 1.5578, "step": 41204 }, { "epoch": 3.22923197492163, "grad_norm": 3.6607589721679688, "learning_rate": 8.918971387151948e-07, "loss": 1.6775, "step": 41205 }, { "epoch": 3.2293103448275864, "grad_norm": 4.988041877746582, "learning_rate": 8.917213694336984e-07, "loss": 1.5777, "step": 41206 }, { "epoch": 3.2293887147335423, "grad_norm": 3.651303291320801, "learning_rate": 8.915456157779851e-07, "loss": 1.1676, "step": 41207 }, { "epoch": 3.2294670846394986, "grad_norm": 5.054327487945557, "learning_rate": 8.913698777487201e-07, "loss": 1.1789, "step": 41208 }, { "epoch": 3.2295454545454545, "grad_norm": 4.350964069366455, "learning_rate": 8.911941553465741e-07, "loss": 1.4141, "step": 41209 }, { "epoch": 3.229623824451411, "grad_norm": 3.5832738876342773, "learning_rate": 8.91018448572214e-07, "loss": 1.0548, "step": 41210 }, { "epoch": 3.2297021943573667, "grad_norm": 5.897925853729248, "learning_rate": 8.908427574263118e-07, "loss": 1.2142, "step": 41211 }, { "epoch": 3.229780564263323, "grad_norm": 14.238675117492676, "learning_rate": 8.90667081909532e-07, "loss": 1.1566, "step": 41212 }, { "epoch": 3.229858934169279, "grad_norm": 5.052753448486328, "learning_rate": 8.904914220225436e-07, "loss": 1.1061, "step": 41213 }, { "epoch": 3.2299373040752353, "grad_norm": 5.906373977661133, "learning_rate": 8.903157777660148e-07, "loss": 1.3905, "step": 41214 }, { "epoch": 3.230015673981191, "grad_norm": 2.5179738998413086, "learning_rate": 8.901401491406148e-07, "loss": 1.3747, "step": 41215 }, { "epoch": 3.2300940438871475, "grad_norm": 4.4601640701293945, "learning_rate": 8.899645361470099e-07, "loss": 1.5127, "step": 41216 }, { "epoch": 3.2301724137931034, "grad_norm": 5.795354843139648, "learning_rate": 8.897889387858694e-07, "loss": 1.3947, "step": 41217 }, { "epoch": 3.2302507836990597, "grad_norm": 2.7412497997283936, "learning_rate": 8.896133570578603e-07, "loss": 1.2855, "step": 41218 }, { "epoch": 3.2303291536050156, "grad_norm": 3.309356927871704, "learning_rate": 8.894377909636509e-07, "loss": 1.6465, "step": 41219 }, { "epoch": 3.230407523510972, "grad_norm": 3.8872828483581543, "learning_rate": 8.892622405039086e-07, "loss": 1.2839, "step": 41220 }, { "epoch": 3.230485893416928, "grad_norm": 3.092107057571411, "learning_rate": 8.890867056793017e-07, "loss": 1.0473, "step": 41221 }, { "epoch": 3.230564263322884, "grad_norm": 2.51129412651062, "learning_rate": 8.889111864904987e-07, "loss": 1.3615, "step": 41222 }, { "epoch": 3.23064263322884, "grad_norm": 4.7951555252075195, "learning_rate": 8.887356829381644e-07, "loss": 1.3976, "step": 41223 }, { "epoch": 3.2307210031347964, "grad_norm": 3.0984716415405273, "learning_rate": 8.885601950229666e-07, "loss": 0.9834, "step": 41224 }, { "epoch": 3.2307993730407523, "grad_norm": 3.262423276901245, "learning_rate": 8.883847227455755e-07, "loss": 1.036, "step": 41225 }, { "epoch": 3.2308777429467086, "grad_norm": 2.9723148345947266, "learning_rate": 8.882092661066583e-07, "loss": 1.4268, "step": 41226 }, { "epoch": 3.2309561128526645, "grad_norm": 5.095651626586914, "learning_rate": 8.880338251068799e-07, "loss": 0.6194, "step": 41227 }, { "epoch": 3.231034482758621, "grad_norm": 8.598615646362305, "learning_rate": 8.878583997469087e-07, "loss": 1.095, "step": 41228 }, { "epoch": 3.2311128526645767, "grad_norm": 4.201709747314453, "learning_rate": 8.876829900274108e-07, "loss": 1.4139, "step": 41229 }, { "epoch": 3.231191222570533, "grad_norm": 3.35113263130188, "learning_rate": 8.87507595949057e-07, "loss": 1.3447, "step": 41230 }, { "epoch": 3.231269592476489, "grad_norm": 2.7873337268829346, "learning_rate": 8.87332217512511e-07, "loss": 1.6356, "step": 41231 }, { "epoch": 3.2313479623824453, "grad_norm": 2.7845921516418457, "learning_rate": 8.871568547184406e-07, "loss": 1.4541, "step": 41232 }, { "epoch": 3.231426332288401, "grad_norm": 4.228922367095947, "learning_rate": 8.869815075675131e-07, "loss": 1.4238, "step": 41233 }, { "epoch": 3.2315047021943575, "grad_norm": 3.032623052597046, "learning_rate": 8.868061760603952e-07, "loss": 1.3754, "step": 41234 }, { "epoch": 3.2315830721003134, "grad_norm": 5.3439154624938965, "learning_rate": 8.866308601977542e-07, "loss": 1.4682, "step": 41235 }, { "epoch": 3.2316614420062697, "grad_norm": 4.474998474121094, "learning_rate": 8.864555599802566e-07, "loss": 1.3982, "step": 41236 }, { "epoch": 3.2317398119122256, "grad_norm": 5.404296398162842, "learning_rate": 8.862802754085703e-07, "loss": 1.3511, "step": 41237 }, { "epoch": 3.231818181818182, "grad_norm": 4.210250377655029, "learning_rate": 8.861050064833587e-07, "loss": 1.0618, "step": 41238 }, { "epoch": 3.231896551724138, "grad_norm": 3.0354089736938477, "learning_rate": 8.859297532052912e-07, "loss": 1.3847, "step": 41239 }, { "epoch": 3.231974921630094, "grad_norm": 10.38683032989502, "learning_rate": 8.857545155750347e-07, "loss": 1.3111, "step": 41240 }, { "epoch": 3.23205329153605, "grad_norm": 2.805621862411499, "learning_rate": 8.855792935932555e-07, "loss": 1.4342, "step": 41241 }, { "epoch": 3.2321316614420064, "grad_norm": 4.378922462463379, "learning_rate": 8.854040872606184e-07, "loss": 1.4545, "step": 41242 }, { "epoch": 3.2322100313479623, "grad_norm": 3.185807943344116, "learning_rate": 8.852288965777895e-07, "loss": 1.4758, "step": 41243 }, { "epoch": 3.2322884012539186, "grad_norm": 4.540682315826416, "learning_rate": 8.850537215454375e-07, "loss": 1.5788, "step": 41244 }, { "epoch": 3.2323667711598745, "grad_norm": 2.720160484313965, "learning_rate": 8.848785621642286e-07, "loss": 1.3229, "step": 41245 }, { "epoch": 3.232445141065831, "grad_norm": 3.240156888961792, "learning_rate": 8.847034184348269e-07, "loss": 1.5372, "step": 41246 }, { "epoch": 3.2325235109717867, "grad_norm": 9.068742752075195, "learning_rate": 8.845282903578995e-07, "loss": 1.1919, "step": 41247 }, { "epoch": 3.232601880877743, "grad_norm": 3.7849221229553223, "learning_rate": 8.843531779341119e-07, "loss": 1.3119, "step": 41248 }, { "epoch": 3.232680250783699, "grad_norm": 4.207243919372559, "learning_rate": 8.841780811641332e-07, "loss": 0.8108, "step": 41249 }, { "epoch": 3.2327586206896552, "grad_norm": 5.306785583496094, "learning_rate": 8.840030000486255e-07, "loss": 1.0892, "step": 41250 }, { "epoch": 3.232836990595611, "grad_norm": 2.8520848751068115, "learning_rate": 8.838279345882567e-07, "loss": 1.2117, "step": 41251 }, { "epoch": 3.2329153605015675, "grad_norm": 2.6602885723114014, "learning_rate": 8.836528847836917e-07, "loss": 1.2089, "step": 41252 }, { "epoch": 3.2329937304075234, "grad_norm": 3.307830572128296, "learning_rate": 8.834778506355973e-07, "loss": 1.2181, "step": 41253 }, { "epoch": 3.2330721003134797, "grad_norm": 2.793959856033325, "learning_rate": 8.833028321446385e-07, "loss": 1.3677, "step": 41254 }, { "epoch": 3.2331504702194356, "grad_norm": 6.239797115325928, "learning_rate": 8.831278293114814e-07, "loss": 1.6437, "step": 41255 }, { "epoch": 3.233228840125392, "grad_norm": 2.861208438873291, "learning_rate": 8.829528421367917e-07, "loss": 1.3813, "step": 41256 }, { "epoch": 3.233307210031348, "grad_norm": 2.928129196166992, "learning_rate": 8.827778706212342e-07, "loss": 1.2411, "step": 41257 }, { "epoch": 3.233385579937304, "grad_norm": 8.651007652282715, "learning_rate": 8.826029147654752e-07, "loss": 1.1481, "step": 41258 }, { "epoch": 3.23346394984326, "grad_norm": 3.686856746673584, "learning_rate": 8.824279745701803e-07, "loss": 1.2302, "step": 41259 }, { "epoch": 3.2335423197492164, "grad_norm": 6.691349983215332, "learning_rate": 8.822530500360149e-07, "loss": 1.1449, "step": 41260 }, { "epoch": 3.2336206896551722, "grad_norm": 2.572690486907959, "learning_rate": 8.82078141163642e-07, "loss": 1.2323, "step": 41261 }, { "epoch": 3.2336990595611286, "grad_norm": 4.834831714630127, "learning_rate": 8.819032479537299e-07, "loss": 1.4107, "step": 41262 }, { "epoch": 3.2337774294670845, "grad_norm": 2.526853084564209, "learning_rate": 8.81728370406944e-07, "loss": 1.2328, "step": 41263 }, { "epoch": 3.233855799373041, "grad_norm": 4.313535213470459, "learning_rate": 8.815535085239468e-07, "loss": 1.2147, "step": 41264 }, { "epoch": 3.2339341692789967, "grad_norm": 5.6756463050842285, "learning_rate": 8.813786623054043e-07, "loss": 1.5998, "step": 41265 }, { "epoch": 3.234012539184953, "grad_norm": 3.3448126316070557, "learning_rate": 8.812038317519822e-07, "loss": 1.4303, "step": 41266 }, { "epoch": 3.234090909090909, "grad_norm": 2.867309331893921, "learning_rate": 8.810290168643454e-07, "loss": 1.1298, "step": 41267 }, { "epoch": 3.2341692789968652, "grad_norm": 4.751033782958984, "learning_rate": 8.808542176431584e-07, "loss": 1.3142, "step": 41268 }, { "epoch": 3.234247648902821, "grad_norm": 8.392682075500488, "learning_rate": 8.806794340890862e-07, "loss": 1.4136, "step": 41269 }, { "epoch": 3.2343260188087775, "grad_norm": 4.465888023376465, "learning_rate": 8.805046662027933e-07, "loss": 1.5201, "step": 41270 }, { "epoch": 3.2344043887147333, "grad_norm": 4.346224784851074, "learning_rate": 8.803299139849453e-07, "loss": 1.1407, "step": 41271 }, { "epoch": 3.2344827586206897, "grad_norm": 2.568875312805176, "learning_rate": 8.801551774362061e-07, "loss": 1.3471, "step": 41272 }, { "epoch": 3.2345611285266456, "grad_norm": 3.724278211593628, "learning_rate": 8.799804565572401e-07, "loss": 1.4795, "step": 41273 }, { "epoch": 3.234639498432602, "grad_norm": 2.2109386920928955, "learning_rate": 8.798057513487129e-07, "loss": 0.9507, "step": 41274 }, { "epoch": 3.234717868338558, "grad_norm": 3.8099894523620605, "learning_rate": 8.79631061811288e-07, "loss": 1.3914, "step": 41275 }, { "epoch": 3.234796238244514, "grad_norm": 4.74019193649292, "learning_rate": 8.794563879456302e-07, "loss": 1.2437, "step": 41276 }, { "epoch": 3.23487460815047, "grad_norm": 3.289047956466675, "learning_rate": 8.792817297524037e-07, "loss": 1.3275, "step": 41277 }, { "epoch": 3.2349529780564263, "grad_norm": 2.763164520263672, "learning_rate": 8.791070872322749e-07, "loss": 0.8888, "step": 41278 }, { "epoch": 3.2350313479623827, "grad_norm": 2.393716335296631, "learning_rate": 8.78932460385904e-07, "loss": 1.5866, "step": 41279 }, { "epoch": 3.2351097178683386, "grad_norm": 2.6487958431243896, "learning_rate": 8.787578492139564e-07, "loss": 1.2301, "step": 41280 }, { "epoch": 3.2351880877742945, "grad_norm": 3.904411792755127, "learning_rate": 8.785832537170985e-07, "loss": 1.2018, "step": 41281 }, { "epoch": 3.235266457680251, "grad_norm": 3.7358038425445557, "learning_rate": 8.784086738959946e-07, "loss": 0.8547, "step": 41282 }, { "epoch": 3.235344827586207, "grad_norm": 2.447061777114868, "learning_rate": 8.782341097513053e-07, "loss": 1.2607, "step": 41283 }, { "epoch": 3.235423197492163, "grad_norm": 2.7068850994110107, "learning_rate": 8.780595612836967e-07, "loss": 0.8775, "step": 41284 }, { "epoch": 3.235501567398119, "grad_norm": 3.9991960525512695, "learning_rate": 8.778850284938312e-07, "loss": 1.4371, "step": 41285 }, { "epoch": 3.2355799373040752, "grad_norm": 4.238609790802002, "learning_rate": 8.777105113823759e-07, "loss": 1.5671, "step": 41286 }, { "epoch": 3.2356583072100316, "grad_norm": 4.597497463226318, "learning_rate": 8.775360099499913e-07, "loss": 1.414, "step": 41287 }, { "epoch": 3.2357366771159874, "grad_norm": 3.287612199783325, "learning_rate": 8.773615241973421e-07, "loss": 1.4207, "step": 41288 }, { "epoch": 3.235815047021944, "grad_norm": 2.8036768436431885, "learning_rate": 8.771870541250927e-07, "loss": 1.1935, "step": 41289 }, { "epoch": 3.2358934169278997, "grad_norm": 5.532831192016602, "learning_rate": 8.770125997339058e-07, "loss": 1.3984, "step": 41290 }, { "epoch": 3.235971786833856, "grad_norm": 7.513184547424316, "learning_rate": 8.768381610244453e-07, "loss": 1.3021, "step": 41291 }, { "epoch": 3.236050156739812, "grad_norm": 3.5806832313537598, "learning_rate": 8.766637379973747e-07, "loss": 1.4741, "step": 41292 }, { "epoch": 3.236128526645768, "grad_norm": 2.8431973457336426, "learning_rate": 8.764893306533589e-07, "loss": 1.1678, "step": 41293 }, { "epoch": 3.236206896551724, "grad_norm": 4.397761344909668, "learning_rate": 8.763149389930569e-07, "loss": 1.3771, "step": 41294 }, { "epoch": 3.2362852664576804, "grad_norm": 3.4808661937713623, "learning_rate": 8.76140563017136e-07, "loss": 1.4376, "step": 41295 }, { "epoch": 3.2363636363636363, "grad_norm": 2.995742082595825, "learning_rate": 8.759662027262583e-07, "loss": 1.4922, "step": 41296 }, { "epoch": 3.2364420062695927, "grad_norm": 3.1298320293426514, "learning_rate": 8.757918581210884e-07, "loss": 1.2636, "step": 41297 }, { "epoch": 3.2365203761755486, "grad_norm": 2.65299654006958, "learning_rate": 8.756175292022862e-07, "loss": 1.2264, "step": 41298 }, { "epoch": 3.236598746081505, "grad_norm": 6.057955741882324, "learning_rate": 8.754432159705156e-07, "loss": 1.8089, "step": 41299 }, { "epoch": 3.2366771159874608, "grad_norm": 14.473550796508789, "learning_rate": 8.752689184264418e-07, "loss": 1.0699, "step": 41300 }, { "epoch": 3.236755485893417, "grad_norm": 2.747772216796875, "learning_rate": 8.750946365707274e-07, "loss": 1.6944, "step": 41301 }, { "epoch": 3.236833855799373, "grad_norm": 5.168368339538574, "learning_rate": 8.749203704040332e-07, "loss": 1.5833, "step": 41302 }, { "epoch": 3.2369122257053293, "grad_norm": 7.397318363189697, "learning_rate": 8.747461199270235e-07, "loss": 1.6429, "step": 41303 }, { "epoch": 3.236990595611285, "grad_norm": 3.102288007736206, "learning_rate": 8.745718851403601e-07, "loss": 1.4632, "step": 41304 }, { "epoch": 3.2370689655172415, "grad_norm": 3.174048900604248, "learning_rate": 8.743976660447068e-07, "loss": 1.0704, "step": 41305 }, { "epoch": 3.2371473354231974, "grad_norm": 2.388902187347412, "learning_rate": 8.742234626407254e-07, "loss": 1.5231, "step": 41306 }, { "epoch": 3.2372257053291538, "grad_norm": 2.473695755004883, "learning_rate": 8.740492749290791e-07, "loss": 1.369, "step": 41307 }, { "epoch": 3.2373040752351097, "grad_norm": 5.584177494049072, "learning_rate": 8.738751029104298e-07, "loss": 1.3729, "step": 41308 }, { "epoch": 3.237382445141066, "grad_norm": 4.577347755432129, "learning_rate": 8.737009465854407e-07, "loss": 1.2074, "step": 41309 }, { "epoch": 3.237460815047022, "grad_norm": 5.496901988983154, "learning_rate": 8.735268059547735e-07, "loss": 0.631, "step": 41310 }, { "epoch": 3.237539184952978, "grad_norm": 4.581480503082275, "learning_rate": 8.733526810190907e-07, "loss": 1.7002, "step": 41311 }, { "epoch": 3.237617554858934, "grad_norm": 3.4131033420562744, "learning_rate": 8.731785717790564e-07, "loss": 1.1073, "step": 41312 }, { "epoch": 3.2376959247648904, "grad_norm": 4.489859104156494, "learning_rate": 8.730044782353281e-07, "loss": 1.2811, "step": 41313 }, { "epoch": 3.2377742946708463, "grad_norm": 3.248406171798706, "learning_rate": 8.72830400388573e-07, "loss": 1.2348, "step": 41314 }, { "epoch": 3.2378526645768027, "grad_norm": 3.444061040878296, "learning_rate": 8.726563382394504e-07, "loss": 1.1096, "step": 41315 }, { "epoch": 3.2379310344827585, "grad_norm": 4.140066623687744, "learning_rate": 8.72482291788625e-07, "loss": 1.1227, "step": 41316 }, { "epoch": 3.238009404388715, "grad_norm": 3.7140729427337646, "learning_rate": 8.723082610367556e-07, "loss": 1.2591, "step": 41317 }, { "epoch": 3.2380877742946708, "grad_norm": 2.8825902938842773, "learning_rate": 8.721342459845039e-07, "loss": 1.1769, "step": 41318 }, { "epoch": 3.238166144200627, "grad_norm": 3.4409289360046387, "learning_rate": 8.719602466325361e-07, "loss": 1.445, "step": 41319 }, { "epoch": 3.238244514106583, "grad_norm": 4.331148147583008, "learning_rate": 8.717862629815099e-07, "loss": 1.2163, "step": 41320 }, { "epoch": 3.2383228840125393, "grad_norm": 2.9468204975128174, "learning_rate": 8.716122950320882e-07, "loss": 1.4422, "step": 41321 }, { "epoch": 3.238401253918495, "grad_norm": 3.4161744117736816, "learning_rate": 8.714383427849326e-07, "loss": 1.1809, "step": 41322 }, { "epoch": 3.2384796238244515, "grad_norm": 3.4995367527008057, "learning_rate": 8.712644062407056e-07, "loss": 1.3473, "step": 41323 }, { "epoch": 3.2385579937304074, "grad_norm": 14.0427827835083, "learning_rate": 8.710904854000679e-07, "loss": 1.7603, "step": 41324 }, { "epoch": 3.2386363636363638, "grad_norm": 1.7698496580123901, "learning_rate": 8.709165802636816e-07, "loss": 0.8239, "step": 41325 }, { "epoch": 3.2387147335423196, "grad_norm": 2.4525692462921143, "learning_rate": 8.707426908322075e-07, "loss": 1.2308, "step": 41326 }, { "epoch": 3.238793103448276, "grad_norm": 4.700986385345459, "learning_rate": 8.70568817106307e-07, "loss": 1.3286, "step": 41327 }, { "epoch": 3.238871473354232, "grad_norm": 2.691544532775879, "learning_rate": 8.703949590866422e-07, "loss": 1.0784, "step": 41328 }, { "epoch": 3.238949843260188, "grad_norm": 2.900848388671875, "learning_rate": 8.702211167738733e-07, "loss": 1.3754, "step": 41329 }, { "epoch": 3.239028213166144, "grad_norm": 4.299808502197266, "learning_rate": 8.700472901686624e-07, "loss": 1.092, "step": 41330 }, { "epoch": 3.2391065830721004, "grad_norm": 3.608368396759033, "learning_rate": 8.698734792716718e-07, "loss": 1.1847, "step": 41331 }, { "epoch": 3.2391849529780563, "grad_norm": 13.206733703613281, "learning_rate": 8.696996840835587e-07, "loss": 1.3695, "step": 41332 }, { "epoch": 3.2392633228840126, "grad_norm": 4.601912021636963, "learning_rate": 8.695259046049875e-07, "loss": 0.9554, "step": 41333 }, { "epoch": 3.2393416927899685, "grad_norm": 2.6654655933380127, "learning_rate": 8.69352140836619e-07, "loss": 1.2402, "step": 41334 }, { "epoch": 3.239420062695925, "grad_norm": 3.3711984157562256, "learning_rate": 8.691783927791131e-07, "loss": 1.2609, "step": 41335 }, { "epoch": 3.2394984326018808, "grad_norm": 3.689833879470825, "learning_rate": 8.690046604331298e-07, "loss": 1.1048, "step": 41336 }, { "epoch": 3.239576802507837, "grad_norm": 6.208470344543457, "learning_rate": 8.688309437993303e-07, "loss": 1.132, "step": 41337 }, { "epoch": 3.239655172413793, "grad_norm": 3.753422737121582, "learning_rate": 8.686572428783785e-07, "loss": 1.4787, "step": 41338 }, { "epoch": 3.2397335423197493, "grad_norm": 5.001079082489014, "learning_rate": 8.684835576709316e-07, "loss": 1.2929, "step": 41339 }, { "epoch": 3.239811912225705, "grad_norm": 2.307387351989746, "learning_rate": 8.683098881776509e-07, "loss": 1.3834, "step": 41340 }, { "epoch": 3.2398902821316615, "grad_norm": 3.4517345428466797, "learning_rate": 8.681362343991967e-07, "loss": 1.2141, "step": 41341 }, { "epoch": 3.2399686520376174, "grad_norm": 2.545976161956787, "learning_rate": 8.679625963362304e-07, "loss": 0.8752, "step": 41342 }, { "epoch": 3.2400470219435737, "grad_norm": 3.606405019760132, "learning_rate": 8.677889739894124e-07, "loss": 1.1688, "step": 41343 }, { "epoch": 3.2401253918495296, "grad_norm": 5.597850322723389, "learning_rate": 8.676153673594023e-07, "loss": 1.3285, "step": 41344 }, { "epoch": 3.240203761755486, "grad_norm": 2.305779218673706, "learning_rate": 8.674417764468607e-07, "loss": 0.9538, "step": 41345 }, { "epoch": 3.240282131661442, "grad_norm": 3.414921522140503, "learning_rate": 8.67268201252448e-07, "loss": 1.4248, "step": 41346 }, { "epoch": 3.240360501567398, "grad_norm": 2.9328806400299072, "learning_rate": 8.670946417768245e-07, "loss": 1.5202, "step": 41347 }, { "epoch": 3.240438871473354, "grad_norm": 6.275356769561768, "learning_rate": 8.669210980206504e-07, "loss": 1.6335, "step": 41348 }, { "epoch": 3.2405172413793104, "grad_norm": 3.0723109245300293, "learning_rate": 8.667475699845867e-07, "loss": 1.8277, "step": 41349 }, { "epoch": 3.2405956112852663, "grad_norm": 2.687734365463257, "learning_rate": 8.665740576692905e-07, "loss": 1.2248, "step": 41350 }, { "epoch": 3.2406739811912226, "grad_norm": 10.144632339477539, "learning_rate": 8.664005610754228e-07, "loss": 2.0786, "step": 41351 }, { "epoch": 3.2407523510971785, "grad_norm": 4.110479354858398, "learning_rate": 8.662270802036454e-07, "loss": 1.3155, "step": 41352 }, { "epoch": 3.240830721003135, "grad_norm": 3.721378803253174, "learning_rate": 8.660536150546184e-07, "loss": 1.1476, "step": 41353 }, { "epoch": 3.2409090909090907, "grad_norm": 3.220144033432007, "learning_rate": 8.658801656289984e-07, "loss": 1.1456, "step": 41354 }, { "epoch": 3.240987460815047, "grad_norm": 3.148297071456909, "learning_rate": 8.657067319274471e-07, "loss": 1.1509, "step": 41355 }, { "epoch": 3.241065830721003, "grad_norm": 2.716618776321411, "learning_rate": 8.655333139506228e-07, "loss": 1.241, "step": 41356 }, { "epoch": 3.2411442006269593, "grad_norm": 3.4368512630462646, "learning_rate": 8.653599116991884e-07, "loss": 0.9023, "step": 41357 }, { "epoch": 3.241222570532915, "grad_norm": 8.62835693359375, "learning_rate": 8.651865251738001e-07, "loss": 1.724, "step": 41358 }, { "epoch": 3.2413009404388715, "grad_norm": 4.381057262420654, "learning_rate": 8.650131543751183e-07, "loss": 1.6502, "step": 41359 }, { "epoch": 3.2413793103448274, "grad_norm": 3.8559041023254395, "learning_rate": 8.648397993038027e-07, "loss": 0.6549, "step": 41360 }, { "epoch": 3.2414576802507837, "grad_norm": 3.50443434715271, "learning_rate": 8.646664599605125e-07, "loss": 1.4214, "step": 41361 }, { "epoch": 3.2415360501567396, "grad_norm": 3.862722396850586, "learning_rate": 8.644931363459069e-07, "loss": 0.7319, "step": 41362 }, { "epoch": 3.241614420062696, "grad_norm": 3.206005573272705, "learning_rate": 8.643198284606458e-07, "loss": 1.3137, "step": 41363 }, { "epoch": 3.241692789968652, "grad_norm": 2.823535442352295, "learning_rate": 8.641465363053875e-07, "loss": 1.2884, "step": 41364 }, { "epoch": 3.241771159874608, "grad_norm": 2.8947153091430664, "learning_rate": 8.639732598807915e-07, "loss": 1.2429, "step": 41365 }, { "epoch": 3.241849529780564, "grad_norm": 6.5321550369262695, "learning_rate": 8.637999991875168e-07, "loss": 0.7982, "step": 41366 }, { "epoch": 3.2419278996865204, "grad_norm": 3.928494930267334, "learning_rate": 8.636267542262222e-07, "loss": 1.0833, "step": 41367 }, { "epoch": 3.2420062695924763, "grad_norm": 7.074544429779053, "learning_rate": 8.634535249975684e-07, "loss": 1.2982, "step": 41368 }, { "epoch": 3.2420846394984326, "grad_norm": 3.917433023452759, "learning_rate": 8.632803115022104e-07, "loss": 1.352, "step": 41369 }, { "epoch": 3.2421630094043885, "grad_norm": 3.11871337890625, "learning_rate": 8.631071137408104e-07, "loss": 1.1919, "step": 41370 }, { "epoch": 3.242241379310345, "grad_norm": 3.251795530319214, "learning_rate": 8.629339317140257e-07, "loss": 1.36, "step": 41371 }, { "epoch": 3.2423197492163007, "grad_norm": 3.5381102561950684, "learning_rate": 8.627607654225173e-07, "loss": 1.5541, "step": 41372 }, { "epoch": 3.242398119122257, "grad_norm": 5.946868419647217, "learning_rate": 8.625876148669404e-07, "loss": 1.1836, "step": 41373 }, { "epoch": 3.2424764890282134, "grad_norm": 3.067643404006958, "learning_rate": 8.624144800479539e-07, "loss": 1.0979, "step": 41374 }, { "epoch": 3.2425548589341693, "grad_norm": 2.942819595336914, "learning_rate": 8.622413609662195e-07, "loss": 1.5751, "step": 41375 }, { "epoch": 3.242633228840125, "grad_norm": 2.827666759490967, "learning_rate": 8.620682576223933e-07, "loss": 1.0032, "step": 41376 }, { "epoch": 3.2427115987460815, "grad_norm": 3.7306809425354004, "learning_rate": 8.618951700171335e-07, "loss": 1.2557, "step": 41377 }, { "epoch": 3.242789968652038, "grad_norm": 3.677050828933716, "learning_rate": 8.617220981510993e-07, "loss": 1.4636, "step": 41378 }, { "epoch": 3.2428683385579937, "grad_norm": 3.6056809425354004, "learning_rate": 8.615490420249484e-07, "loss": 1.2503, "step": 41379 }, { "epoch": 3.2429467084639496, "grad_norm": 4.170260906219482, "learning_rate": 8.613760016393396e-07, "loss": 1.6488, "step": 41380 }, { "epoch": 3.243025078369906, "grad_norm": 3.274275541305542, "learning_rate": 8.612029769949304e-07, "loss": 1.4791, "step": 41381 }, { "epoch": 3.2431034482758623, "grad_norm": 2.719728708267212, "learning_rate": 8.610299680923795e-07, "loss": 1.2687, "step": 41382 }, { "epoch": 3.243181818181818, "grad_norm": 7.029157638549805, "learning_rate": 8.608569749323448e-07, "loss": 1.1157, "step": 41383 }, { "epoch": 3.2432601880877745, "grad_norm": 10.534733772277832, "learning_rate": 8.606839975154841e-07, "loss": 0.9504, "step": 41384 }, { "epoch": 3.2433385579937304, "grad_norm": 3.4423837661743164, "learning_rate": 8.605110358424551e-07, "loss": 1.2925, "step": 41385 }, { "epoch": 3.2434169278996867, "grad_norm": 6.669499397277832, "learning_rate": 8.603380899139163e-07, "loss": 1.3106, "step": 41386 }, { "epoch": 3.2434952978056426, "grad_norm": 2.948925495147705, "learning_rate": 8.601651597305261e-07, "loss": 1.2631, "step": 41387 }, { "epoch": 3.243573667711599, "grad_norm": 5.148098468780518, "learning_rate": 8.599922452929388e-07, "loss": 1.4651, "step": 41388 }, { "epoch": 3.243652037617555, "grad_norm": 8.071271896362305, "learning_rate": 8.598193466018162e-07, "loss": 1.0513, "step": 41389 }, { "epoch": 3.243730407523511, "grad_norm": 3.66947340965271, "learning_rate": 8.59646463657815e-07, "loss": 1.5069, "step": 41390 }, { "epoch": 3.243808777429467, "grad_norm": 2.3747265338897705, "learning_rate": 8.594735964615907e-07, "loss": 1.2106, "step": 41391 }, { "epoch": 3.2438871473354234, "grad_norm": 3.4748215675354004, "learning_rate": 8.59300745013803e-07, "loss": 1.2078, "step": 41392 }, { "epoch": 3.2439655172413793, "grad_norm": 3.0107204914093018, "learning_rate": 8.591279093151067e-07, "loss": 1.2146, "step": 41393 }, { "epoch": 3.2440438871473356, "grad_norm": 3.636246681213379, "learning_rate": 8.58955089366163e-07, "loss": 1.2113, "step": 41394 }, { "epoch": 3.2441222570532915, "grad_norm": 4.595267295837402, "learning_rate": 8.587822851676264e-07, "loss": 1.2552, "step": 41395 }, { "epoch": 3.244200626959248, "grad_norm": 6.755312442779541, "learning_rate": 8.586094967201553e-07, "loss": 1.2256, "step": 41396 }, { "epoch": 3.2442789968652037, "grad_norm": 6.469718933105469, "learning_rate": 8.584367240244057e-07, "loss": 1.5365, "step": 41397 }, { "epoch": 3.24435736677116, "grad_norm": 3.667689800262451, "learning_rate": 8.582639670810361e-07, "loss": 1.0659, "step": 41398 }, { "epoch": 3.244435736677116, "grad_norm": 3.7093517780303955, "learning_rate": 8.580912258907026e-07, "loss": 1.2923, "step": 41399 }, { "epoch": 3.2445141065830723, "grad_norm": 5.342020034790039, "learning_rate": 8.57918500454063e-07, "loss": 1.1785, "step": 41400 }, { "epoch": 3.244592476489028, "grad_norm": 3.2104883193969727, "learning_rate": 8.57745790771774e-07, "loss": 1.3665, "step": 41401 }, { "epoch": 3.2446708463949845, "grad_norm": 2.376216173171997, "learning_rate": 8.575730968444924e-07, "loss": 1.2537, "step": 41402 }, { "epoch": 3.2447492163009404, "grad_norm": 3.2819204330444336, "learning_rate": 8.574004186728746e-07, "loss": 1.4006, "step": 41403 }, { "epoch": 3.2448275862068967, "grad_norm": 4.823513984680176, "learning_rate": 8.572277562575782e-07, "loss": 1.3714, "step": 41404 }, { "epoch": 3.2449059561128526, "grad_norm": 3.40222430229187, "learning_rate": 8.570551095992608e-07, "loss": 1.0557, "step": 41405 }, { "epoch": 3.244984326018809, "grad_norm": 3.490255355834961, "learning_rate": 8.568824786985768e-07, "loss": 1.4193, "step": 41406 }, { "epoch": 3.245062695924765, "grad_norm": 3.0362191200256348, "learning_rate": 8.567098635561821e-07, "loss": 1.4965, "step": 41407 }, { "epoch": 3.245141065830721, "grad_norm": 2.950406789779663, "learning_rate": 8.565372641727365e-07, "loss": 1.3709, "step": 41408 }, { "epoch": 3.245219435736677, "grad_norm": 2.892228364944458, "learning_rate": 8.563646805488963e-07, "loss": 1.5201, "step": 41409 }, { "epoch": 3.2452978056426334, "grad_norm": 3.9357361793518066, "learning_rate": 8.561921126853151e-07, "loss": 1.3573, "step": 41410 }, { "epoch": 3.2453761755485893, "grad_norm": 2.760647773742676, "learning_rate": 8.560195605826504e-07, "loss": 1.314, "step": 41411 }, { "epoch": 3.2454545454545456, "grad_norm": 12.380393981933594, "learning_rate": 8.558470242415578e-07, "loss": 1.613, "step": 41412 }, { "epoch": 3.2455329153605015, "grad_norm": 3.3599817752838135, "learning_rate": 8.556745036626974e-07, "loss": 1.4494, "step": 41413 }, { "epoch": 3.245611285266458, "grad_norm": 9.489574432373047, "learning_rate": 8.555019988467206e-07, "loss": 1.391, "step": 41414 }, { "epoch": 3.2456896551724137, "grad_norm": 3.2321319580078125, "learning_rate": 8.55329509794286e-07, "loss": 1.2425, "step": 41415 }, { "epoch": 3.24576802507837, "grad_norm": 2.461118221282959, "learning_rate": 8.551570365060491e-07, "loss": 0.8574, "step": 41416 }, { "epoch": 3.245846394984326, "grad_norm": 5.454012393951416, "learning_rate": 8.549845789826655e-07, "loss": 1.2687, "step": 41417 }, { "epoch": 3.2459247648902823, "grad_norm": 3.7533955574035645, "learning_rate": 8.54812137224792e-07, "loss": 1.32, "step": 41418 }, { "epoch": 3.246003134796238, "grad_norm": 4.688175678253174, "learning_rate": 8.546397112330834e-07, "loss": 1.5206, "step": 41419 }, { "epoch": 3.2460815047021945, "grad_norm": 4.906912803649902, "learning_rate": 8.544673010081977e-07, "loss": 1.4169, "step": 41420 }, { "epoch": 3.2461598746081504, "grad_norm": 3.0441536903381348, "learning_rate": 8.542949065507872e-07, "loss": 1.8364, "step": 41421 }, { "epoch": 3.2462382445141067, "grad_norm": 3.637573719024658, "learning_rate": 8.541225278615101e-07, "loss": 1.1104, "step": 41422 }, { "epoch": 3.2463166144200626, "grad_norm": 3.5966098308563232, "learning_rate": 8.539501649410215e-07, "loss": 1.378, "step": 41423 }, { "epoch": 3.246394984326019, "grad_norm": 3.0169970989227295, "learning_rate": 8.537778177899781e-07, "loss": 1.3331, "step": 41424 }, { "epoch": 3.246473354231975, "grad_norm": 5.51178503036499, "learning_rate": 8.536054864090332e-07, "loss": 1.4609, "step": 41425 }, { "epoch": 3.246551724137931, "grad_norm": 4.723750114440918, "learning_rate": 8.534331707988425e-07, "loss": 1.4131, "step": 41426 }, { "epoch": 3.246630094043887, "grad_norm": 3.0955140590667725, "learning_rate": 8.532608709600626e-07, "loss": 1.5126, "step": 41427 }, { "epoch": 3.2467084639498434, "grad_norm": 3.2051641941070557, "learning_rate": 8.530885868933503e-07, "loss": 1.1591, "step": 41428 }, { "epoch": 3.2467868338557992, "grad_norm": 2.8038692474365234, "learning_rate": 8.529163185993578e-07, "loss": 1.1792, "step": 41429 }, { "epoch": 3.2468652037617556, "grad_norm": 5.037703037261963, "learning_rate": 8.527440660787411e-07, "loss": 1.2566, "step": 41430 }, { "epoch": 3.2469435736677115, "grad_norm": 9.27628231048584, "learning_rate": 8.525718293321561e-07, "loss": 1.095, "step": 41431 }, { "epoch": 3.247021943573668, "grad_norm": 2.464540958404541, "learning_rate": 8.523996083602576e-07, "loss": 1.4394, "step": 41432 }, { "epoch": 3.2471003134796237, "grad_norm": 7.003040790557861, "learning_rate": 8.522274031637001e-07, "loss": 0.8818, "step": 41433 }, { "epoch": 3.24717868338558, "grad_norm": 2.6883463859558105, "learning_rate": 8.520552137431398e-07, "loss": 1.4268, "step": 41434 }, { "epoch": 3.247257053291536, "grad_norm": 8.742790222167969, "learning_rate": 8.518830400992306e-07, "loss": 1.2152, "step": 41435 }, { "epoch": 3.2473354231974922, "grad_norm": 2.7220818996429443, "learning_rate": 8.517108822326281e-07, "loss": 1.2244, "step": 41436 }, { "epoch": 3.247413793103448, "grad_norm": 3.591512441635132, "learning_rate": 8.515387401439868e-07, "loss": 1.1752, "step": 41437 }, { "epoch": 3.2474921630094045, "grad_norm": 3.2291553020477295, "learning_rate": 8.513666138339605e-07, "loss": 0.9736, "step": 41438 }, { "epoch": 3.2475705329153604, "grad_norm": 2.3827500343322754, "learning_rate": 8.511945033032065e-07, "loss": 0.8379, "step": 41439 }, { "epoch": 3.2476489028213167, "grad_norm": 4.300304412841797, "learning_rate": 8.510224085523755e-07, "loss": 1.7771, "step": 41440 }, { "epoch": 3.2477272727272726, "grad_norm": 4.32107400894165, "learning_rate": 8.508503295821252e-07, "loss": 1.232, "step": 41441 }, { "epoch": 3.247805642633229, "grad_norm": 2.750434637069702, "learning_rate": 8.50678266393109e-07, "loss": 1.3872, "step": 41442 }, { "epoch": 3.247884012539185, "grad_norm": 3.7163004875183105, "learning_rate": 8.505062189859825e-07, "loss": 0.7171, "step": 41443 }, { "epoch": 3.247962382445141, "grad_norm": 2.5759947299957275, "learning_rate": 8.503341873613979e-07, "loss": 1.3929, "step": 41444 }, { "epoch": 3.248040752351097, "grad_norm": 9.690705299377441, "learning_rate": 8.501621715200098e-07, "loss": 0.9838, "step": 41445 }, { "epoch": 3.2481191222570533, "grad_norm": 7.123963356018066, "learning_rate": 8.499901714624753e-07, "loss": 1.251, "step": 41446 }, { "epoch": 3.2481974921630092, "grad_norm": 4.475435256958008, "learning_rate": 8.498181871894456e-07, "loss": 1.3537, "step": 41447 }, { "epoch": 3.2482758620689656, "grad_norm": 2.854362964630127, "learning_rate": 8.496462187015758e-07, "loss": 1.4483, "step": 41448 }, { "epoch": 3.2483542319749215, "grad_norm": 3.1130294799804688, "learning_rate": 8.494742659995197e-07, "loss": 1.3384, "step": 41449 }, { "epoch": 3.248432601880878, "grad_norm": 2.842679262161255, "learning_rate": 8.493023290839319e-07, "loss": 1.2659, "step": 41450 }, { "epoch": 3.2485109717868337, "grad_norm": 6.023807525634766, "learning_rate": 8.491304079554663e-07, "loss": 1.3505, "step": 41451 }, { "epoch": 3.24858934169279, "grad_norm": 2.347334861755371, "learning_rate": 8.489585026147768e-07, "loss": 1.3372, "step": 41452 }, { "epoch": 3.248667711598746, "grad_norm": 5.046338081359863, "learning_rate": 8.487866130625166e-07, "loss": 1.0496, "step": 41453 }, { "epoch": 3.2487460815047022, "grad_norm": 12.707441329956055, "learning_rate": 8.486147392993399e-07, "loss": 1.0097, "step": 41454 }, { "epoch": 3.248824451410658, "grad_norm": 3.4451568126678467, "learning_rate": 8.48442881325901e-07, "loss": 1.2963, "step": 41455 }, { "epoch": 3.2489028213166145, "grad_norm": 4.44529390335083, "learning_rate": 8.482710391428523e-07, "loss": 1.7029, "step": 41456 }, { "epoch": 3.2489811912225703, "grad_norm": 2.712796926498413, "learning_rate": 8.480992127508486e-07, "loss": 1.3297, "step": 41457 }, { "epoch": 3.2490595611285267, "grad_norm": 5.251709938049316, "learning_rate": 8.479274021505429e-07, "loss": 1.3939, "step": 41458 }, { "epoch": 3.2491379310344826, "grad_norm": 3.2367944717407227, "learning_rate": 8.477556073425886e-07, "loss": 1.2262, "step": 41459 }, { "epoch": 3.249216300940439, "grad_norm": 2.3924412727355957, "learning_rate": 8.475838283276388e-07, "loss": 1.3619, "step": 41460 }, { "epoch": 3.249294670846395, "grad_norm": 3.9496004581451416, "learning_rate": 8.474120651063494e-07, "loss": 1.2404, "step": 41461 }, { "epoch": 3.249373040752351, "grad_norm": 8.316466331481934, "learning_rate": 8.472403176793698e-07, "loss": 1.009, "step": 41462 }, { "epoch": 3.249451410658307, "grad_norm": 3.0356550216674805, "learning_rate": 8.470685860473548e-07, "loss": 1.4157, "step": 41463 }, { "epoch": 3.2495297805642633, "grad_norm": 3.9705910682678223, "learning_rate": 8.46896870210957e-07, "loss": 1.0735, "step": 41464 }, { "epoch": 3.2496081504702197, "grad_norm": 4.054068565368652, "learning_rate": 8.467251701708329e-07, "loss": 1.192, "step": 41465 }, { "epoch": 3.2496865203761756, "grad_norm": 5.631704330444336, "learning_rate": 8.46553485927632e-07, "loss": 0.8694, "step": 41466 }, { "epoch": 3.2497648902821314, "grad_norm": 2.2593719959259033, "learning_rate": 8.463818174820077e-07, "loss": 1.063, "step": 41467 }, { "epoch": 3.249843260188088, "grad_norm": 2.949951171875, "learning_rate": 8.462101648346138e-07, "loss": 1.4295, "step": 41468 }, { "epoch": 3.249921630094044, "grad_norm": 10.064537048339844, "learning_rate": 8.460385279861028e-07, "loss": 1.3533, "step": 41469 }, { "epoch": 3.25, "grad_norm": 4.2973504066467285, "learning_rate": 8.458669069371278e-07, "loss": 1.0768, "step": 41470 }, { "epoch": 3.25, "eval_loss": 1.2463722229003906, "eval_runtime": 110.8095, "eval_samples_per_second": 24.249, "eval_steps_per_second": 6.064, "step": 41470 }, { "epoch": 3.250078369905956, "grad_norm": 3.5039713382720947, "learning_rate": 8.456953016883412e-07, "loss": 1.3456, "step": 41471 }, { "epoch": 3.250156739811912, "grad_norm": 4.807149887084961, "learning_rate": 8.455237122403958e-07, "loss": 1.341, "step": 41472 }, { "epoch": 3.2502351097178686, "grad_norm": 2.812013864517212, "learning_rate": 8.453521385939445e-07, "loss": 1.4666, "step": 41473 }, { "epoch": 3.2503134796238244, "grad_norm": 4.020503520965576, "learning_rate": 8.4518058074964e-07, "loss": 1.702, "step": 41474 }, { "epoch": 3.2503918495297803, "grad_norm": 5.934309482574463, "learning_rate": 8.450090387081339e-07, "loss": 1.5344, "step": 41475 }, { "epoch": 3.2504702194357367, "grad_norm": 3.934544563293457, "learning_rate": 8.448375124700808e-07, "loss": 1.2575, "step": 41476 }, { "epoch": 3.250548589341693, "grad_norm": 4.4236979484558105, "learning_rate": 8.446660020361292e-07, "loss": 1.6637, "step": 41477 }, { "epoch": 3.250626959247649, "grad_norm": 4.916393280029297, "learning_rate": 8.44494507406935e-07, "loss": 1.5362, "step": 41478 }, { "epoch": 3.2507053291536048, "grad_norm": 4.270809173583984, "learning_rate": 8.443230285831494e-07, "loss": 0.9435, "step": 41479 }, { "epoch": 3.250783699059561, "grad_norm": 8.840828895568848, "learning_rate": 8.441515655654254e-07, "loss": 1.2686, "step": 41480 }, { "epoch": 3.2508620689655174, "grad_norm": 3.0178260803222656, "learning_rate": 8.439801183544133e-07, "loss": 1.3216, "step": 41481 }, { "epoch": 3.2509404388714733, "grad_norm": 3.7913999557495117, "learning_rate": 8.43808686950765e-07, "loss": 1.4421, "step": 41482 }, { "epoch": 3.2510188087774297, "grad_norm": 5.126572608947754, "learning_rate": 8.436372713551349e-07, "loss": 1.5401, "step": 41483 }, { "epoch": 3.2510971786833855, "grad_norm": 3.585728883743286, "learning_rate": 8.434658715681748e-07, "loss": 1.4073, "step": 41484 }, { "epoch": 3.251175548589342, "grad_norm": 3.281343936920166, "learning_rate": 8.432944875905347e-07, "loss": 1.4832, "step": 41485 }, { "epoch": 3.2512539184952978, "grad_norm": 4.352609157562256, "learning_rate": 8.431231194228667e-07, "loss": 1.2747, "step": 41486 }, { "epoch": 3.251332288401254, "grad_norm": 3.8019001483917236, "learning_rate": 8.429517670658238e-07, "loss": 1.3673, "step": 41487 }, { "epoch": 3.25141065830721, "grad_norm": 3.6648025512695312, "learning_rate": 8.427804305200571e-07, "loss": 1.3707, "step": 41488 }, { "epoch": 3.2514890282131663, "grad_norm": 7.071093559265137, "learning_rate": 8.426091097862182e-07, "loss": 1.79, "step": 41489 }, { "epoch": 3.251567398119122, "grad_norm": 2.5890121459960938, "learning_rate": 8.42437804864959e-07, "loss": 1.2683, "step": 41490 }, { "epoch": 3.2516457680250785, "grad_norm": 6.096312522888184, "learning_rate": 8.422665157569309e-07, "loss": 1.5405, "step": 41491 }, { "epoch": 3.2517241379310344, "grad_norm": 3.661212205886841, "learning_rate": 8.420952424627854e-07, "loss": 1.0713, "step": 41492 }, { "epoch": 3.2518025078369908, "grad_norm": 3.047433853149414, "learning_rate": 8.419239849831746e-07, "loss": 1.4136, "step": 41493 }, { "epoch": 3.2518808777429467, "grad_norm": 2.718179941177368, "learning_rate": 8.417527433187484e-07, "loss": 1.4525, "step": 41494 }, { "epoch": 3.251959247648903, "grad_norm": 3.019322156906128, "learning_rate": 8.415815174701602e-07, "loss": 1.4446, "step": 41495 }, { "epoch": 3.252037617554859, "grad_norm": 2.462116241455078, "learning_rate": 8.41410307438058e-07, "loss": 1.2847, "step": 41496 }, { "epoch": 3.252115987460815, "grad_norm": 2.5107545852661133, "learning_rate": 8.412391132230957e-07, "loss": 1.3668, "step": 41497 }, { "epoch": 3.252194357366771, "grad_norm": 3.97407603263855, "learning_rate": 8.410679348259243e-07, "loss": 1.6522, "step": 41498 }, { "epoch": 3.2522727272727274, "grad_norm": 2.1717641353607178, "learning_rate": 8.408967722471956e-07, "loss": 1.2205, "step": 41499 }, { "epoch": 3.2523510971786833, "grad_norm": 5.808356761932373, "learning_rate": 8.407256254875573e-07, "loss": 1.2502, "step": 41500 }, { "epoch": 3.2524294670846396, "grad_norm": 2.537519931793213, "learning_rate": 8.405544945476618e-07, "loss": 0.9834, "step": 41501 }, { "epoch": 3.2525078369905955, "grad_norm": 2.9377663135528564, "learning_rate": 8.40383379428163e-07, "loss": 1.4279, "step": 41502 }, { "epoch": 3.252586206896552, "grad_norm": 3.4755420684814453, "learning_rate": 8.40212280129708e-07, "loss": 1.433, "step": 41503 }, { "epoch": 3.2526645768025078, "grad_norm": 14.932591438293457, "learning_rate": 8.400411966529487e-07, "loss": 2.0793, "step": 41504 }, { "epoch": 3.252742946708464, "grad_norm": 1.9875894784927368, "learning_rate": 8.398701289985362e-07, "loss": 1.2851, "step": 41505 }, { "epoch": 3.25282131661442, "grad_norm": 3.9629299640655518, "learning_rate": 8.39699077167121e-07, "loss": 1.6333, "step": 41506 }, { "epoch": 3.2528996865203763, "grad_norm": 11.034729957580566, "learning_rate": 8.395280411593537e-07, "loss": 0.9753, "step": 41507 }, { "epoch": 3.252978056426332, "grad_norm": 2.823004961013794, "learning_rate": 8.393570209758845e-07, "loss": 1.1278, "step": 41508 }, { "epoch": 3.2530564263322885, "grad_norm": 4.863690376281738, "learning_rate": 8.391860166173643e-07, "loss": 1.161, "step": 41509 }, { "epoch": 3.2531347962382444, "grad_norm": 3.780414581298828, "learning_rate": 8.390150280844434e-07, "loss": 1.3407, "step": 41510 }, { "epoch": 3.2532131661442008, "grad_norm": 5.5218505859375, "learning_rate": 8.388440553777721e-07, "loss": 1.0907, "step": 41511 }, { "epoch": 3.2532915360501566, "grad_norm": 6.450698375701904, "learning_rate": 8.386730984980007e-07, "loss": 2.6351, "step": 41512 }, { "epoch": 3.253369905956113, "grad_norm": 4.303902626037598, "learning_rate": 8.385021574457791e-07, "loss": 1.1183, "step": 41513 }, { "epoch": 3.253448275862069, "grad_norm": 4.802694797515869, "learning_rate": 8.383312322217596e-07, "loss": 1.3826, "step": 41514 }, { "epoch": 3.253526645768025, "grad_norm": 11.952310562133789, "learning_rate": 8.381603228265877e-07, "loss": 1.2895, "step": 41515 }, { "epoch": 3.253605015673981, "grad_norm": 5.4934821128845215, "learning_rate": 8.379894292609175e-07, "loss": 1.3374, "step": 41516 }, { "epoch": 3.2536833855799374, "grad_norm": 3.7621049880981445, "learning_rate": 8.378185515253995e-07, "loss": 1.1787, "step": 41517 }, { "epoch": 3.2537617554858933, "grad_norm": 3.0905721187591553, "learning_rate": 8.376476896206798e-07, "loss": 1.3786, "step": 41518 }, { "epoch": 3.2538401253918496, "grad_norm": 2.9845175743103027, "learning_rate": 8.374768435474112e-07, "loss": 1.2616, "step": 41519 }, { "epoch": 3.2539184952978055, "grad_norm": 2.1734657287597656, "learning_rate": 8.373060133062411e-07, "loss": 1.5449, "step": 41520 }, { "epoch": 3.253996865203762, "grad_norm": 12.586102485656738, "learning_rate": 8.371351988978233e-07, "loss": 0.6543, "step": 41521 }, { "epoch": 3.2540752351097177, "grad_norm": 2.6472764015197754, "learning_rate": 8.369644003228039e-07, "loss": 1.3789, "step": 41522 }, { "epoch": 3.254153605015674, "grad_norm": 6.079135894775391, "learning_rate": 8.367936175818331e-07, "loss": 2.4155, "step": 41523 }, { "epoch": 3.25423197492163, "grad_norm": 2.718371629714966, "learning_rate": 8.36622850675562e-07, "loss": 1.0833, "step": 41524 }, { "epoch": 3.2543103448275863, "grad_norm": 3.1838109493255615, "learning_rate": 8.364520996046382e-07, "loss": 1.0809, "step": 41525 }, { "epoch": 3.254388714733542, "grad_norm": 3.9140465259552, "learning_rate": 8.362813643697126e-07, "loss": 1.7998, "step": 41526 }, { "epoch": 3.2544670846394985, "grad_norm": 9.577409744262695, "learning_rate": 8.36110644971434e-07, "loss": 0.9522, "step": 41527 }, { "epoch": 3.2545454545454544, "grad_norm": 3.747231960296631, "learning_rate": 8.35939941410453e-07, "loss": 1.1911, "step": 41528 }, { "epoch": 3.2546238244514107, "grad_norm": 3.163639783859253, "learning_rate": 8.357692536874152e-07, "loss": 1.5037, "step": 41529 }, { "epoch": 3.2547021943573666, "grad_norm": 4.3911261558532715, "learning_rate": 8.355985818029733e-07, "loss": 1.2386, "step": 41530 }, { "epoch": 3.254780564263323, "grad_norm": 2.7039692401885986, "learning_rate": 8.354279257577757e-07, "loss": 1.3538, "step": 41531 }, { "epoch": 3.254858934169279, "grad_norm": 18.673137664794922, "learning_rate": 8.352572855524721e-07, "loss": 2.0114, "step": 41532 }, { "epoch": 3.254937304075235, "grad_norm": 3.178900957107544, "learning_rate": 8.350866611877101e-07, "loss": 1.1764, "step": 41533 }, { "epoch": 3.255015673981191, "grad_norm": 5.90456485748291, "learning_rate": 8.349160526641375e-07, "loss": 1.2091, "step": 41534 }, { "epoch": 3.2550940438871474, "grad_norm": 2.841858148574829, "learning_rate": 8.347454599824061e-07, "loss": 0.9968, "step": 41535 }, { "epoch": 3.2551724137931033, "grad_norm": 3.2485783100128174, "learning_rate": 8.345748831431649e-07, "loss": 1.471, "step": 41536 }, { "epoch": 3.2552507836990596, "grad_norm": 8.816896438598633, "learning_rate": 8.344043221470599e-07, "loss": 0.6574, "step": 41537 }, { "epoch": 3.2553291536050155, "grad_norm": 5.352146625518799, "learning_rate": 8.342337769947412e-07, "loss": 1.4173, "step": 41538 }, { "epoch": 3.255407523510972, "grad_norm": 2.8228683471679688, "learning_rate": 8.340632476868571e-07, "loss": 1.5847, "step": 41539 }, { "epoch": 3.2554858934169277, "grad_norm": 3.4810001850128174, "learning_rate": 8.338927342240583e-07, "loss": 1.1208, "step": 41540 }, { "epoch": 3.255564263322884, "grad_norm": 7.403620719909668, "learning_rate": 8.337222366069908e-07, "loss": 1.311, "step": 41541 }, { "epoch": 3.25564263322884, "grad_norm": 4.5518479347229, "learning_rate": 8.335517548363036e-07, "loss": 1.4509, "step": 41542 }, { "epoch": 3.2557210031347963, "grad_norm": 2.90030574798584, "learning_rate": 8.333812889126458e-07, "loss": 1.4578, "step": 41543 }, { "epoch": 3.255799373040752, "grad_norm": 3.4855573177337646, "learning_rate": 8.332108388366655e-07, "loss": 1.609, "step": 41544 }, { "epoch": 3.2558777429467085, "grad_norm": 4.2196221351623535, "learning_rate": 8.330404046090112e-07, "loss": 1.5376, "step": 41545 }, { "epoch": 3.2559561128526644, "grad_norm": 2.916487216949463, "learning_rate": 8.328699862303308e-07, "loss": 1.0576, "step": 41546 }, { "epoch": 3.2560344827586207, "grad_norm": 7.712173938751221, "learning_rate": 8.326995837012736e-07, "loss": 1.1176, "step": 41547 }, { "epoch": 3.2561128526645766, "grad_norm": 4.967308521270752, "learning_rate": 8.325291970224842e-07, "loss": 0.5987, "step": 41548 }, { "epoch": 3.256191222570533, "grad_norm": 11.0081787109375, "learning_rate": 8.323588261946147e-07, "loss": 1.226, "step": 41549 }, { "epoch": 3.256269592476489, "grad_norm": 9.433812141418457, "learning_rate": 8.321884712183115e-07, "loss": 1.4193, "step": 41550 }, { "epoch": 3.256347962382445, "grad_norm": 5.361051559448242, "learning_rate": 8.320181320942239e-07, "loss": 1.2844, "step": 41551 }, { "epoch": 3.256426332288401, "grad_norm": 2.7862131595611572, "learning_rate": 8.31847808822997e-07, "loss": 1.3, "step": 41552 }, { "epoch": 3.2565047021943574, "grad_norm": 2.4847571849823, "learning_rate": 8.316775014052791e-07, "loss": 1.1182, "step": 41553 }, { "epoch": 3.2565830721003133, "grad_norm": 2.3422186374664307, "learning_rate": 8.315072098417204e-07, "loss": 1.3364, "step": 41554 }, { "epoch": 3.2566614420062696, "grad_norm": 12.134040832519531, "learning_rate": 8.31336934132968e-07, "loss": 1.4995, "step": 41555 }, { "epoch": 3.256739811912226, "grad_norm": 3.405346155166626, "learning_rate": 8.311666742796682e-07, "loss": 1.6596, "step": 41556 }, { "epoch": 3.256818181818182, "grad_norm": 4.717437267303467, "learning_rate": 8.309964302824686e-07, "loss": 1.2514, "step": 41557 }, { "epoch": 3.2568965517241377, "grad_norm": 3.85262393951416, "learning_rate": 8.308262021420172e-07, "loss": 1.6045, "step": 41558 }, { "epoch": 3.256974921630094, "grad_norm": 3.5366296768188477, "learning_rate": 8.306559898589617e-07, "loss": 1.569, "step": 41559 }, { "epoch": 3.2570532915360504, "grad_norm": 3.107734441757202, "learning_rate": 8.304857934339494e-07, "loss": 1.1485, "step": 41560 }, { "epoch": 3.2571316614420063, "grad_norm": 3.513589859008789, "learning_rate": 8.303156128676271e-07, "loss": 1.0698, "step": 41561 }, { "epoch": 3.257210031347962, "grad_norm": 3.347127676010132, "learning_rate": 8.30145448160643e-07, "loss": 1.5147, "step": 41562 }, { "epoch": 3.2572884012539185, "grad_norm": 4.959341049194336, "learning_rate": 8.299752993136433e-07, "loss": 1.2789, "step": 41563 }, { "epoch": 3.257366771159875, "grad_norm": 6.640567779541016, "learning_rate": 8.298051663272766e-07, "loss": 1.0969, "step": 41564 }, { "epoch": 3.2574451410658307, "grad_norm": 3.107738971710205, "learning_rate": 8.296350492021882e-07, "loss": 1.3525, "step": 41565 }, { "epoch": 3.2575235109717866, "grad_norm": 5.295907020568848, "learning_rate": 8.294649479390265e-07, "loss": 1.2298, "step": 41566 }, { "epoch": 3.257601880877743, "grad_norm": 2.626403331756592, "learning_rate": 8.292948625384378e-07, "loss": 0.7999, "step": 41567 }, { "epoch": 3.2576802507836993, "grad_norm": 5.600849628448486, "learning_rate": 8.291247930010692e-07, "loss": 1.3352, "step": 41568 }, { "epoch": 3.257758620689655, "grad_norm": 4.3074870109558105, "learning_rate": 8.289547393275693e-07, "loss": 1.5561, "step": 41569 }, { "epoch": 3.257836990595611, "grad_norm": 2.830796718597412, "learning_rate": 8.287847015185818e-07, "loss": 1.188, "step": 41570 }, { "epoch": 3.2579153605015674, "grad_norm": 7.550502300262451, "learning_rate": 8.286146795747535e-07, "loss": 1.0156, "step": 41571 }, { "epoch": 3.2579937304075237, "grad_norm": 2.149061679840088, "learning_rate": 8.284446734967333e-07, "loss": 1.4044, "step": 41572 }, { "epoch": 3.2580721003134796, "grad_norm": 2.3830063343048096, "learning_rate": 8.282746832851685e-07, "loss": 1.6788, "step": 41573 }, { "epoch": 3.2581504702194355, "grad_norm": 2.581125497817993, "learning_rate": 8.281047089407024e-07, "loss": 1.5796, "step": 41574 }, { "epoch": 3.258228840125392, "grad_norm": 3.1518774032592773, "learning_rate": 8.27934750463984e-07, "loss": 1.2704, "step": 41575 }, { "epoch": 3.258307210031348, "grad_norm": 4.083459377288818, "learning_rate": 8.277648078556582e-07, "loss": 1.5389, "step": 41576 }, { "epoch": 3.258385579937304, "grad_norm": 10.693007469177246, "learning_rate": 8.275948811163726e-07, "loss": 2.1885, "step": 41577 }, { "epoch": 3.2584639498432604, "grad_norm": 2.810959577560425, "learning_rate": 8.274249702467724e-07, "loss": 1.3062, "step": 41578 }, { "epoch": 3.2585423197492163, "grad_norm": 2.591498374938965, "learning_rate": 8.272550752475045e-07, "loss": 1.4988, "step": 41579 }, { "epoch": 3.2586206896551726, "grad_norm": 2.90091872215271, "learning_rate": 8.270851961192155e-07, "loss": 1.2218, "step": 41580 }, { "epoch": 3.2586990595611285, "grad_norm": 3.1077773571014404, "learning_rate": 8.269153328625507e-07, "loss": 1.5323, "step": 41581 }, { "epoch": 3.258777429467085, "grad_norm": 8.681018829345703, "learning_rate": 8.267454854781564e-07, "loss": 1.0947, "step": 41582 }, { "epoch": 3.2588557993730407, "grad_norm": 4.512613296508789, "learning_rate": 8.265756539666791e-07, "loss": 1.6718, "step": 41583 }, { "epoch": 3.258934169278997, "grad_norm": 2.6093738079071045, "learning_rate": 8.264058383287659e-07, "loss": 1.1371, "step": 41584 }, { "epoch": 3.259012539184953, "grad_norm": 2.2495808601379395, "learning_rate": 8.262360385650581e-07, "loss": 1.2015, "step": 41585 }, { "epoch": 3.2590909090909093, "grad_norm": 4.200381278991699, "learning_rate": 8.260662546762066e-07, "loss": 1.4848, "step": 41586 }, { "epoch": 3.259169278996865, "grad_norm": 3.01076602935791, "learning_rate": 8.258964866628544e-07, "loss": 1.5972, "step": 41587 }, { "epoch": 3.2592476489028215, "grad_norm": 2.6878175735473633, "learning_rate": 8.257267345256498e-07, "loss": 1.1486, "step": 41588 }, { "epoch": 3.2593260188087774, "grad_norm": 10.427816390991211, "learning_rate": 8.255569982652351e-07, "loss": 1.4289, "step": 41589 }, { "epoch": 3.2594043887147337, "grad_norm": 4.964766502380371, "learning_rate": 8.253872778822564e-07, "loss": 1.2712, "step": 41590 }, { "epoch": 3.2594827586206896, "grad_norm": 3.027832508087158, "learning_rate": 8.252175733773615e-07, "loss": 1.334, "step": 41591 }, { "epoch": 3.259561128526646, "grad_norm": 12.139510154724121, "learning_rate": 8.250478847511956e-07, "loss": 2.1434, "step": 41592 }, { "epoch": 3.259639498432602, "grad_norm": 2.334791898727417, "learning_rate": 8.248782120044024e-07, "loss": 0.9985, "step": 41593 }, { "epoch": 3.259717868338558, "grad_norm": 5.800750732421875, "learning_rate": 8.247085551376277e-07, "loss": 1.2961, "step": 41594 }, { "epoch": 3.259796238244514, "grad_norm": 2.782893419265747, "learning_rate": 8.245389141515159e-07, "loss": 1.4323, "step": 41595 }, { "epoch": 3.2598746081504704, "grad_norm": 3.277909278869629, "learning_rate": 8.243692890467159e-07, "loss": 1.3879, "step": 41596 }, { "epoch": 3.2599529780564263, "grad_norm": 4.436770439147949, "learning_rate": 8.241996798238694e-07, "loss": 1.5165, "step": 41597 }, { "epoch": 3.2600313479623826, "grad_norm": 3.991737127304077, "learning_rate": 8.240300864836221e-07, "loss": 1.379, "step": 41598 }, { "epoch": 3.2601097178683385, "grad_norm": 2.226724863052368, "learning_rate": 8.238605090266195e-07, "loss": 1.3814, "step": 41599 }, { "epoch": 3.260188087774295, "grad_norm": 3.411851644515991, "learning_rate": 8.236909474535066e-07, "loss": 1.2589, "step": 41600 }, { "epoch": 3.2602664576802507, "grad_norm": 2.2524821758270264, "learning_rate": 8.235214017649284e-07, "loss": 1.162, "step": 41601 }, { "epoch": 3.260344827586207, "grad_norm": 3.574415445327759, "learning_rate": 8.233518719615291e-07, "loss": 0.9293, "step": 41602 }, { "epoch": 3.260423197492163, "grad_norm": 2.7165191173553467, "learning_rate": 8.231823580439558e-07, "loss": 1.3026, "step": 41603 }, { "epoch": 3.2605015673981192, "grad_norm": 3.8864214420318604, "learning_rate": 8.23012860012849e-07, "loss": 1.58, "step": 41604 }, { "epoch": 3.260579937304075, "grad_norm": 2.385896682739258, "learning_rate": 8.228433778688566e-07, "loss": 1.2808, "step": 41605 }, { "epoch": 3.2606583072100315, "grad_norm": 3.414930820465088, "learning_rate": 8.226739116126231e-07, "loss": 1.2005, "step": 41606 }, { "epoch": 3.2607366771159874, "grad_norm": 3.1499269008636475, "learning_rate": 8.22504461244793e-07, "loss": 1.4646, "step": 41607 }, { "epoch": 3.2608150470219437, "grad_norm": 2.3410756587982178, "learning_rate": 8.223350267660096e-07, "loss": 1.0541, "step": 41608 }, { "epoch": 3.2608934169278996, "grad_norm": 2.9372811317443848, "learning_rate": 8.221656081769164e-07, "loss": 1.0655, "step": 41609 }, { "epoch": 3.260971786833856, "grad_norm": 2.7913100719451904, "learning_rate": 8.219962054781605e-07, "loss": 1.2929, "step": 41610 }, { "epoch": 3.261050156739812, "grad_norm": 3.3760266304016113, "learning_rate": 8.218268186703864e-07, "loss": 1.2277, "step": 41611 }, { "epoch": 3.261128526645768, "grad_norm": 3.1451737880706787, "learning_rate": 8.216574477542361e-07, "loss": 0.7662, "step": 41612 }, { "epoch": 3.261206896551724, "grad_norm": 3.6039376258850098, "learning_rate": 8.214880927303543e-07, "loss": 1.2101, "step": 41613 }, { "epoch": 3.2612852664576804, "grad_norm": 5.307114124298096, "learning_rate": 8.213187535993855e-07, "loss": 1.4967, "step": 41614 }, { "epoch": 3.2613636363636362, "grad_norm": 4.071875095367432, "learning_rate": 8.211494303619744e-07, "loss": 1.4944, "step": 41615 }, { "epoch": 3.2614420062695926, "grad_norm": 2.8442270755767822, "learning_rate": 8.209801230187642e-07, "loss": 1.4011, "step": 41616 }, { "epoch": 3.2615203761755485, "grad_norm": 4.476260185241699, "learning_rate": 8.208108315703989e-07, "loss": 1.3518, "step": 41617 }, { "epoch": 3.261598746081505, "grad_norm": 2.583707809448242, "learning_rate": 8.206415560175229e-07, "loss": 1.4168, "step": 41618 }, { "epoch": 3.2616771159874607, "grad_norm": 5.4630303382873535, "learning_rate": 8.204722963607797e-07, "loss": 1.1807, "step": 41619 }, { "epoch": 3.261755485893417, "grad_norm": 3.18245530128479, "learning_rate": 8.203030526008132e-07, "loss": 1.0987, "step": 41620 }, { "epoch": 3.261833855799373, "grad_norm": 4.055471897125244, "learning_rate": 8.201338247382673e-07, "loss": 1.5891, "step": 41621 }, { "epoch": 3.2619122257053292, "grad_norm": 4.777977466583252, "learning_rate": 8.199646127737865e-07, "loss": 1.4162, "step": 41622 }, { "epoch": 3.261990595611285, "grad_norm": 2.884631633758545, "learning_rate": 8.197954167080108e-07, "loss": 1.1605, "step": 41623 }, { "epoch": 3.2620689655172415, "grad_norm": 6.573493003845215, "learning_rate": 8.196262365415869e-07, "loss": 1.0799, "step": 41624 }, { "epoch": 3.2621473354231973, "grad_norm": 4.95137882232666, "learning_rate": 8.194570722751594e-07, "loss": 1.753, "step": 41625 }, { "epoch": 3.2622257053291537, "grad_norm": 3.1287856101989746, "learning_rate": 8.192879239093687e-07, "loss": 1.1161, "step": 41626 }, { "epoch": 3.2623040752351096, "grad_norm": 2.630619764328003, "learning_rate": 8.191187914448595e-07, "loss": 1.0535, "step": 41627 }, { "epoch": 3.262382445141066, "grad_norm": 2.7788379192352295, "learning_rate": 8.189496748822734e-07, "loss": 1.6275, "step": 41628 }, { "epoch": 3.262460815047022, "grad_norm": 3.110196113586426, "learning_rate": 8.187805742222582e-07, "loss": 1.2753, "step": 41629 }, { "epoch": 3.262539184952978, "grad_norm": 3.3007707595825195, "learning_rate": 8.186114894654523e-07, "loss": 0.941, "step": 41630 }, { "epoch": 3.262617554858934, "grad_norm": 2.3038108348846436, "learning_rate": 8.184424206125008e-07, "loss": 1.3114, "step": 41631 }, { "epoch": 3.2626959247648903, "grad_norm": 3.2764971256256104, "learning_rate": 8.18273367664047e-07, "loss": 1.2845, "step": 41632 }, { "epoch": 3.2627742946708462, "grad_norm": 3.2484288215637207, "learning_rate": 8.181043306207326e-07, "loss": 1.2712, "step": 41633 }, { "epoch": 3.2628526645768026, "grad_norm": 3.3665835857391357, "learning_rate": 8.179353094832021e-07, "loss": 1.3983, "step": 41634 }, { "epoch": 3.2629310344827585, "grad_norm": 3.2640960216522217, "learning_rate": 8.177663042520972e-07, "loss": 1.6369, "step": 41635 }, { "epoch": 3.263009404388715, "grad_norm": 2.5462827682495117, "learning_rate": 8.175973149280614e-07, "loss": 1.108, "step": 41636 }, { "epoch": 3.2630877742946707, "grad_norm": 3.3768155574798584, "learning_rate": 8.174283415117368e-07, "loss": 1.2524, "step": 41637 }, { "epoch": 3.263166144200627, "grad_norm": 2.843156099319458, "learning_rate": 8.172593840037668e-07, "loss": 1.2366, "step": 41638 }, { "epoch": 3.263244514106583, "grad_norm": 5.1459269523620605, "learning_rate": 8.170904424047937e-07, "loss": 1.486, "step": 41639 }, { "epoch": 3.2633228840125392, "grad_norm": 8.040572166442871, "learning_rate": 8.16921516715461e-07, "loss": 1.5113, "step": 41640 }, { "epoch": 3.263401253918495, "grad_norm": 4.198203086853027, "learning_rate": 8.167526069364095e-07, "loss": 1.2931, "step": 41641 }, { "epoch": 3.2634796238244514, "grad_norm": 5.05604887008667, "learning_rate": 8.165837130682813e-07, "loss": 1.3722, "step": 41642 }, { "epoch": 3.2635579937304073, "grad_norm": 5.812724590301514, "learning_rate": 8.164148351117207e-07, "loss": 1.6075, "step": 41643 }, { "epoch": 3.2636363636363637, "grad_norm": 2.2205498218536377, "learning_rate": 8.162459730673705e-07, "loss": 1.3871, "step": 41644 }, { "epoch": 3.2637147335423196, "grad_norm": 7.041932106018066, "learning_rate": 8.160771269358709e-07, "loss": 1.0999, "step": 41645 }, { "epoch": 3.263793103448276, "grad_norm": 13.63883113861084, "learning_rate": 8.159082967178645e-07, "loss": 1.0704, "step": 41646 }, { "epoch": 3.2638714733542322, "grad_norm": 3.138612985610962, "learning_rate": 8.157394824139925e-07, "loss": 1.1903, "step": 41647 }, { "epoch": 3.263949843260188, "grad_norm": 5.095546245574951, "learning_rate": 8.155706840249011e-07, "loss": 1.1914, "step": 41648 }, { "epoch": 3.264028213166144, "grad_norm": 2.8864963054656982, "learning_rate": 8.154019015512283e-07, "loss": 0.9526, "step": 41649 }, { "epoch": 3.2641065830721003, "grad_norm": 2.0991899967193604, "learning_rate": 8.152331349936177e-07, "loss": 1.3719, "step": 41650 }, { "epoch": 3.2641849529780567, "grad_norm": 3.9936623573303223, "learning_rate": 8.1506438435271e-07, "loss": 1.3422, "step": 41651 }, { "epoch": 3.2642633228840126, "grad_norm": 2.3411672115325928, "learning_rate": 8.148956496291482e-07, "loss": 1.1073, "step": 41652 }, { "epoch": 3.2643416927899684, "grad_norm": 4.739139080047607, "learning_rate": 8.147269308235739e-07, "loss": 1.3738, "step": 41653 }, { "epoch": 3.2644200626959248, "grad_norm": 3.53604793548584, "learning_rate": 8.145582279366288e-07, "loss": 1.496, "step": 41654 }, { "epoch": 3.264498432601881, "grad_norm": 3.0813398361206055, "learning_rate": 8.14389540968954e-07, "loss": 1.2856, "step": 41655 }, { "epoch": 3.264576802507837, "grad_norm": 4.123015403747559, "learning_rate": 8.142208699211917e-07, "loss": 1.3271, "step": 41656 }, { "epoch": 3.264655172413793, "grad_norm": 5.733005046844482, "learning_rate": 8.140522147939833e-07, "loss": 1.4466, "step": 41657 }, { "epoch": 3.264733542319749, "grad_norm": 4.094655513763428, "learning_rate": 8.138835755879703e-07, "loss": 1.2473, "step": 41658 }, { "epoch": 3.2648119122257055, "grad_norm": 2.4929232597351074, "learning_rate": 8.137149523037951e-07, "loss": 1.491, "step": 41659 }, { "epoch": 3.2648902821316614, "grad_norm": 2.784817934036255, "learning_rate": 8.135463449420972e-07, "loss": 1.0939, "step": 41660 }, { "epoch": 3.2649686520376173, "grad_norm": 3.6018686294555664, "learning_rate": 8.13377753503517e-07, "loss": 1.0888, "step": 41661 }, { "epoch": 3.2650470219435737, "grad_norm": 4.6720099449157715, "learning_rate": 8.132091779886991e-07, "loss": 1.7076, "step": 41662 }, { "epoch": 3.26512539184953, "grad_norm": 2.564042806625366, "learning_rate": 8.130406183982836e-07, "loss": 1.5931, "step": 41663 }, { "epoch": 3.265203761755486, "grad_norm": 2.7311394214630127, "learning_rate": 8.128720747329105e-07, "loss": 1.3519, "step": 41664 }, { "epoch": 3.2652821316614418, "grad_norm": 3.7607028484344482, "learning_rate": 8.12703546993221e-07, "loss": 1.1783, "step": 41665 }, { "epoch": 3.265360501567398, "grad_norm": 2.430039405822754, "learning_rate": 8.125350351798556e-07, "loss": 1.2361, "step": 41666 }, { "epoch": 3.2654388714733544, "grad_norm": 3.452399730682373, "learning_rate": 8.123665392934582e-07, "loss": 1.3044, "step": 41667 }, { "epoch": 3.2655172413793103, "grad_norm": 3.270068407058716, "learning_rate": 8.121980593346662e-07, "loss": 1.0989, "step": 41668 }, { "epoch": 3.265595611285266, "grad_norm": 6.6786065101623535, "learning_rate": 8.120295953041218e-07, "loss": 1.9773, "step": 41669 }, { "epoch": 3.2656739811912225, "grad_norm": 4.131755828857422, "learning_rate": 8.118611472024662e-07, "loss": 1.9357, "step": 41670 }, { "epoch": 3.265752351097179, "grad_norm": 5.977836608886719, "learning_rate": 8.116927150303394e-07, "loss": 1.4583, "step": 41671 }, { "epoch": 3.2658307210031348, "grad_norm": 2.5291213989257812, "learning_rate": 8.115242987883826e-07, "loss": 1.2953, "step": 41672 }, { "epoch": 3.265909090909091, "grad_norm": 2.3112030029296875, "learning_rate": 8.113558984772352e-07, "loss": 1.2969, "step": 41673 }, { "epoch": 3.265987460815047, "grad_norm": 4.065903186798096, "learning_rate": 8.111875140975395e-07, "loss": 1.4322, "step": 41674 }, { "epoch": 3.2660658307210033, "grad_norm": 9.809466361999512, "learning_rate": 8.110191456499345e-07, "loss": 1.8767, "step": 41675 }, { "epoch": 3.266144200626959, "grad_norm": 2.0659146308898926, "learning_rate": 8.108507931350606e-07, "loss": 1.0201, "step": 41676 }, { "epoch": 3.2662225705329155, "grad_norm": 8.180648803710938, "learning_rate": 8.106824565535592e-07, "loss": 1.4794, "step": 41677 }, { "epoch": 3.2663009404388714, "grad_norm": 3.0423035621643066, "learning_rate": 8.105141359060714e-07, "loss": 1.2423, "step": 41678 }, { "epoch": 3.2663793103448278, "grad_norm": 2.9391162395477295, "learning_rate": 8.103458311932327e-07, "loss": 1.4286, "step": 41679 }, { "epoch": 3.2664576802507836, "grad_norm": 3.3075990676879883, "learning_rate": 8.101775424156888e-07, "loss": 1.1538, "step": 41680 }, { "epoch": 3.26653605015674, "grad_norm": 3.427021026611328, "learning_rate": 8.10009269574078e-07, "loss": 1.2926, "step": 41681 }, { "epoch": 3.266614420062696, "grad_norm": 4.060444355010986, "learning_rate": 8.098410126690387e-07, "loss": 1.3497, "step": 41682 }, { "epoch": 3.266692789968652, "grad_norm": 5.413744926452637, "learning_rate": 8.096727717012121e-07, "loss": 0.9084, "step": 41683 }, { "epoch": 3.266771159874608, "grad_norm": 7.325051307678223, "learning_rate": 8.095045466712365e-07, "loss": 1.6229, "step": 41684 }, { "epoch": 3.2668495297805644, "grad_norm": 6.612044811248779, "learning_rate": 8.093363375797558e-07, "loss": 1.4512, "step": 41685 }, { "epoch": 3.2669278996865203, "grad_norm": 3.2471675872802734, "learning_rate": 8.091681444274063e-07, "loss": 1.2761, "step": 41686 }, { "epoch": 3.2670062695924766, "grad_norm": 3.508920431137085, "learning_rate": 8.089999672148285e-07, "loss": 1.3395, "step": 41687 }, { "epoch": 3.2670846394984325, "grad_norm": 4.0897626876831055, "learning_rate": 8.08831805942662e-07, "loss": 0.6203, "step": 41688 }, { "epoch": 3.267163009404389, "grad_norm": 4.535943031311035, "learning_rate": 8.086636606115466e-07, "loss": 1.3145, "step": 41689 }, { "epoch": 3.2672413793103448, "grad_norm": 5.634037971496582, "learning_rate": 8.084955312221221e-07, "loss": 1.4444, "step": 41690 }, { "epoch": 3.267319749216301, "grad_norm": 3.1432013511657715, "learning_rate": 8.083274177750272e-07, "loss": 1.3525, "step": 41691 }, { "epoch": 3.267398119122257, "grad_norm": 2.6986536979675293, "learning_rate": 8.081593202709026e-07, "loss": 1.2113, "step": 41692 }, { "epoch": 3.2674764890282133, "grad_norm": 3.0221035480499268, "learning_rate": 8.079912387103861e-07, "loss": 1.1066, "step": 41693 }, { "epoch": 3.267554858934169, "grad_norm": 4.041258335113525, "learning_rate": 8.078231730941183e-07, "loss": 1.0662, "step": 41694 }, { "epoch": 3.2676332288401255, "grad_norm": 2.9202282428741455, "learning_rate": 8.076551234227376e-07, "loss": 1.7247, "step": 41695 }, { "epoch": 3.2677115987460814, "grad_norm": 2.7082061767578125, "learning_rate": 8.074870896968851e-07, "loss": 1.2916, "step": 41696 }, { "epoch": 3.2677899686520377, "grad_norm": 3.1867618560791016, "learning_rate": 8.073190719171964e-07, "loss": 1.4013, "step": 41697 }, { "epoch": 3.2678683385579936, "grad_norm": 3.0788440704345703, "learning_rate": 8.071510700843116e-07, "loss": 1.3248, "step": 41698 }, { "epoch": 3.26794670846395, "grad_norm": 2.8995256423950195, "learning_rate": 8.069830841988718e-07, "loss": 1.3962, "step": 41699 }, { "epoch": 3.268025078369906, "grad_norm": 6.093199253082275, "learning_rate": 8.068151142615161e-07, "loss": 1.674, "step": 41700 }, { "epoch": 3.268103448275862, "grad_norm": 8.65073299407959, "learning_rate": 8.066471602728804e-07, "loss": 1.298, "step": 41701 }, { "epoch": 3.268181818181818, "grad_norm": 3.0188233852386475, "learning_rate": 8.064792222336054e-07, "loss": 1.3947, "step": 41702 }, { "epoch": 3.2682601880877744, "grad_norm": 3.1510839462280273, "learning_rate": 8.063113001443279e-07, "loss": 1.3485, "step": 41703 }, { "epoch": 3.2683385579937303, "grad_norm": 3.3251800537109375, "learning_rate": 8.061433940056912e-07, "loss": 1.3674, "step": 41704 }, { "epoch": 3.2684169278996866, "grad_norm": 3.520282745361328, "learning_rate": 8.059755038183292e-07, "loss": 1.487, "step": 41705 }, { "epoch": 3.2684952978056425, "grad_norm": 2.878089666366577, "learning_rate": 8.05807629582882e-07, "loss": 0.9382, "step": 41706 }, { "epoch": 3.268573667711599, "grad_norm": 4.070228576660156, "learning_rate": 8.056397712999891e-07, "loss": 1.171, "step": 41707 }, { "epoch": 3.2686520376175547, "grad_norm": 3.7730464935302734, "learning_rate": 8.054719289702872e-07, "loss": 1.3164, "step": 41708 }, { "epoch": 3.268730407523511, "grad_norm": 4.050468444824219, "learning_rate": 8.053041025944164e-07, "loss": 1.7734, "step": 41709 }, { "epoch": 3.268808777429467, "grad_norm": 3.818108081817627, "learning_rate": 8.051362921730139e-07, "loss": 1.4638, "step": 41710 }, { "epoch": 3.2688871473354233, "grad_norm": 6.489114761352539, "learning_rate": 8.049684977067196e-07, "loss": 0.7507, "step": 41711 }, { "epoch": 3.268965517241379, "grad_norm": 4.7799553871154785, "learning_rate": 8.048007191961682e-07, "loss": 1.0765, "step": 41712 }, { "epoch": 3.2690438871473355, "grad_norm": 2.0689315795898438, "learning_rate": 8.046329566420008e-07, "loss": 1.3669, "step": 41713 }, { "epoch": 3.2691222570532914, "grad_norm": 3.281135082244873, "learning_rate": 8.044652100448553e-07, "loss": 1.409, "step": 41714 }, { "epoch": 3.2692006269592477, "grad_norm": 3.7602078914642334, "learning_rate": 8.042974794053704e-07, "loss": 1.0815, "step": 41715 }, { "epoch": 3.2692789968652036, "grad_norm": 3.716352939605713, "learning_rate": 8.041297647241813e-07, "loss": 1.3214, "step": 41716 }, { "epoch": 3.26935736677116, "grad_norm": 9.22485065460205, "learning_rate": 8.039620660019259e-07, "loss": 1.467, "step": 41717 }, { "epoch": 3.269435736677116, "grad_norm": 4.902912616729736, "learning_rate": 8.037943832392453e-07, "loss": 1.3814, "step": 41718 }, { "epoch": 3.269514106583072, "grad_norm": 5.581411361694336, "learning_rate": 8.036267164367767e-07, "loss": 1.1839, "step": 41719 }, { "epoch": 3.269592476489028, "grad_norm": 3.3245880603790283, "learning_rate": 8.034590655951557e-07, "loss": 1.6702, "step": 41720 }, { "epoch": 3.2696708463949844, "grad_norm": 4.644313812255859, "learning_rate": 8.032914307150208e-07, "loss": 0.8515, "step": 41721 }, { "epoch": 3.2697492163009403, "grad_norm": 4.83219051361084, "learning_rate": 8.031238117970081e-07, "loss": 0.8451, "step": 41722 }, { "epoch": 3.2698275862068966, "grad_norm": 3.0503807067871094, "learning_rate": 8.029562088417597e-07, "loss": 1.5255, "step": 41723 }, { "epoch": 3.2699059561128525, "grad_norm": 2.734964609146118, "learning_rate": 8.027886218499087e-07, "loss": 1.2274, "step": 41724 }, { "epoch": 3.269984326018809, "grad_norm": 3.3605659008026123, "learning_rate": 8.026210508220938e-07, "loss": 1.0475, "step": 41725 }, { "epoch": 3.2700626959247647, "grad_norm": 3.779470205307007, "learning_rate": 8.024534957589524e-07, "loss": 1.501, "step": 41726 }, { "epoch": 3.270141065830721, "grad_norm": 4.750398635864258, "learning_rate": 8.022859566611225e-07, "loss": 1.0647, "step": 41727 }, { "epoch": 3.270219435736677, "grad_norm": 3.2254741191864014, "learning_rate": 8.021184335292403e-07, "loss": 1.0918, "step": 41728 }, { "epoch": 3.2702978056426333, "grad_norm": 3.0052123069763184, "learning_rate": 8.019509263639436e-07, "loss": 0.897, "step": 41729 }, { "epoch": 3.270376175548589, "grad_norm": 4.4424309730529785, "learning_rate": 8.017834351658704e-07, "loss": 1.7501, "step": 41730 }, { "epoch": 3.2704545454545455, "grad_norm": 3.297128200531006, "learning_rate": 8.016159599356544e-07, "loss": 1.4564, "step": 41731 }, { "epoch": 3.2705329153605014, "grad_norm": 3.079782485961914, "learning_rate": 8.014485006739359e-07, "loss": 1.6054, "step": 41732 }, { "epoch": 3.2706112852664577, "grad_norm": 4.080974578857422, "learning_rate": 8.012810573813512e-07, "loss": 1.091, "step": 41733 }, { "epoch": 3.2706896551724136, "grad_norm": 3.5049266815185547, "learning_rate": 8.011136300585376e-07, "loss": 0.893, "step": 41734 }, { "epoch": 3.27076802507837, "grad_norm": 2.7377305030822754, "learning_rate": 8.0094621870613e-07, "loss": 1.5124, "step": 41735 }, { "epoch": 3.270846394984326, "grad_norm": 2.5710511207580566, "learning_rate": 8.007788233247649e-07, "loss": 1.2263, "step": 41736 }, { "epoch": 3.270924764890282, "grad_norm": 13.177457809448242, "learning_rate": 8.006114439150825e-07, "loss": 1.7952, "step": 41737 }, { "epoch": 3.271003134796238, "grad_norm": 3.198563814163208, "learning_rate": 8.004440804777163e-07, "loss": 1.2606, "step": 41738 }, { "epoch": 3.2710815047021944, "grad_norm": 11.260222434997559, "learning_rate": 8.002767330133038e-07, "loss": 1.3281, "step": 41739 }, { "epoch": 3.2711598746081503, "grad_norm": 2.317960739135742, "learning_rate": 8.001094015224813e-07, "loss": 1.403, "step": 41740 }, { "epoch": 3.2712382445141066, "grad_norm": 3.843719005584717, "learning_rate": 7.99942086005886e-07, "loss": 1.0619, "step": 41741 }, { "epoch": 3.271316614420063, "grad_norm": 4.275298118591309, "learning_rate": 7.997747864641531e-07, "loss": 1.4946, "step": 41742 }, { "epoch": 3.271394984326019, "grad_norm": 3.9594790935516357, "learning_rate": 7.996075028979194e-07, "loss": 1.315, "step": 41743 }, { "epoch": 3.2714733542319747, "grad_norm": 4.670668601989746, "learning_rate": 7.994402353078218e-07, "loss": 1.5711, "step": 41744 }, { "epoch": 3.271551724137931, "grad_norm": 3.8686702251434326, "learning_rate": 7.992729836944962e-07, "loss": 0.8182, "step": 41745 }, { "epoch": 3.2716300940438874, "grad_norm": 4.035281658172607, "learning_rate": 7.991057480585778e-07, "loss": 1.4923, "step": 41746 }, { "epoch": 3.2717084639498433, "grad_norm": 3.9494266510009766, "learning_rate": 7.98938528400704e-07, "loss": 1.5279, "step": 41747 }, { "epoch": 3.271786833855799, "grad_norm": 3.306894302368164, "learning_rate": 7.987713247215101e-07, "loss": 1.2032, "step": 41748 }, { "epoch": 3.2718652037617555, "grad_norm": 3.534379243850708, "learning_rate": 7.986041370216329e-07, "loss": 1.4229, "step": 41749 }, { "epoch": 3.271943573667712, "grad_norm": 2.45957088470459, "learning_rate": 7.98436965301706e-07, "loss": 1.2347, "step": 41750 }, { "epoch": 3.2720219435736677, "grad_norm": 3.2553138732910156, "learning_rate": 7.982698095623675e-07, "loss": 1.3841, "step": 41751 }, { "epoch": 3.2721003134796236, "grad_norm": 6.51092529296875, "learning_rate": 7.981026698042537e-07, "loss": 1.039, "step": 41752 }, { "epoch": 3.27217868338558, "grad_norm": 2.1913247108459473, "learning_rate": 7.97935546027998e-07, "loss": 0.9959, "step": 41753 }, { "epoch": 3.2722570532915363, "grad_norm": 10.177393913269043, "learning_rate": 7.977684382342376e-07, "loss": 0.7576, "step": 41754 }, { "epoch": 3.272335423197492, "grad_norm": 2.9422812461853027, "learning_rate": 7.97601346423606e-07, "loss": 1.343, "step": 41755 }, { "epoch": 3.272413793103448, "grad_norm": 2.931844711303711, "learning_rate": 7.974342705967431e-07, "loss": 1.2301, "step": 41756 }, { "epoch": 3.2724921630094044, "grad_norm": 5.2201337814331055, "learning_rate": 7.9726721075428e-07, "loss": 0.934, "step": 41757 }, { "epoch": 3.2725705329153607, "grad_norm": 15.094820976257324, "learning_rate": 7.971001668968542e-07, "loss": 1.4747, "step": 41758 }, { "epoch": 3.2726489028213166, "grad_norm": 2.16843581199646, "learning_rate": 7.969331390251e-07, "loss": 1.3198, "step": 41759 }, { "epoch": 3.2727272727272725, "grad_norm": 3.758188009262085, "learning_rate": 7.967661271396537e-07, "loss": 1.1799, "step": 41760 }, { "epoch": 3.272805642633229, "grad_norm": 6.235858917236328, "learning_rate": 7.965991312411503e-07, "loss": 1.7108, "step": 41761 }, { "epoch": 3.272884012539185, "grad_norm": 3.2573142051696777, "learning_rate": 7.964321513302248e-07, "loss": 1.5705, "step": 41762 }, { "epoch": 3.272962382445141, "grad_norm": 2.5417072772979736, "learning_rate": 7.962651874075122e-07, "loss": 1.4293, "step": 41763 }, { "epoch": 3.2730407523510974, "grad_norm": 3.170853853225708, "learning_rate": 7.960982394736478e-07, "loss": 1.53, "step": 41764 }, { "epoch": 3.2731191222570533, "grad_norm": 3.175323009490967, "learning_rate": 7.959313075292657e-07, "loss": 1.1597, "step": 41765 }, { "epoch": 3.2731974921630096, "grad_norm": 28.936656951904297, "learning_rate": 7.957643915750024e-07, "loss": 0.657, "step": 41766 }, { "epoch": 3.2732758620689655, "grad_norm": 4.1741862297058105, "learning_rate": 7.955974916114928e-07, "loss": 1.4236, "step": 41767 }, { "epoch": 3.273354231974922, "grad_norm": 2.8520002365112305, "learning_rate": 7.954306076393686e-07, "loss": 1.2386, "step": 41768 }, { "epoch": 3.2734326018808777, "grad_norm": 3.7991604804992676, "learning_rate": 7.952637396592678e-07, "loss": 1.409, "step": 41769 }, { "epoch": 3.273510971786834, "grad_norm": 2.8896918296813965, "learning_rate": 7.95096887671824e-07, "loss": 1.5349, "step": 41770 }, { "epoch": 3.27358934169279, "grad_norm": 4.542104244232178, "learning_rate": 7.949300516776731e-07, "loss": 1.1898, "step": 41771 }, { "epoch": 3.2736677115987463, "grad_norm": 3.8653838634490967, "learning_rate": 7.947632316774473e-07, "loss": 0.8451, "step": 41772 }, { "epoch": 3.273746081504702, "grad_norm": 5.636852741241455, "learning_rate": 7.945964276717821e-07, "loss": 1.3845, "step": 41773 }, { "epoch": 3.2738244514106585, "grad_norm": 3.822554111480713, "learning_rate": 7.94429639661311e-07, "loss": 1.4775, "step": 41774 }, { "epoch": 3.2739028213166144, "grad_norm": 2.6157338619232178, "learning_rate": 7.942628676466718e-07, "loss": 1.0524, "step": 41775 }, { "epoch": 3.2739811912225707, "grad_norm": 2.8278095722198486, "learning_rate": 7.940961116284951e-07, "loss": 1.3769, "step": 41776 }, { "epoch": 3.2740595611285266, "grad_norm": 2.477435827255249, "learning_rate": 7.93929371607417e-07, "loss": 1.144, "step": 41777 }, { "epoch": 3.274137931034483, "grad_norm": 2.5927882194519043, "learning_rate": 7.937626475840704e-07, "loss": 1.061, "step": 41778 }, { "epoch": 3.274216300940439, "grad_norm": 5.215548515319824, "learning_rate": 7.935959395590909e-07, "loss": 1.509, "step": 41779 }, { "epoch": 3.274294670846395, "grad_norm": 2.932288646697998, "learning_rate": 7.934292475331118e-07, "loss": 1.5361, "step": 41780 }, { "epoch": 3.274373040752351, "grad_norm": 5.741690158843994, "learning_rate": 7.932625715067671e-07, "loss": 1.586, "step": 41781 }, { "epoch": 3.2744514106583074, "grad_norm": 3.512139081954956, "learning_rate": 7.930959114806907e-07, "loss": 1.3832, "step": 41782 }, { "epoch": 3.2745297805642632, "grad_norm": 2.841671943664551, "learning_rate": 7.929292674555167e-07, "loss": 1.2162, "step": 41783 }, { "epoch": 3.2746081504702196, "grad_norm": 2.2812397480010986, "learning_rate": 7.927626394318789e-07, "loss": 1.2854, "step": 41784 }, { "epoch": 3.2746865203761755, "grad_norm": 3.905042886734009, "learning_rate": 7.925960274104111e-07, "loss": 1.6578, "step": 41785 }, { "epoch": 3.274764890282132, "grad_norm": 3.038022041320801, "learning_rate": 7.92429431391748e-07, "loss": 1.1755, "step": 41786 }, { "epoch": 3.2748432601880877, "grad_norm": 6.862236976623535, "learning_rate": 7.922628513765202e-07, "loss": 0.3198, "step": 41787 }, { "epoch": 3.274921630094044, "grad_norm": 2.520702838897705, "learning_rate": 7.920962873653643e-07, "loss": 1.2043, "step": 41788 }, { "epoch": 3.275, "grad_norm": 3.2110867500305176, "learning_rate": 7.919297393589126e-07, "loss": 1.5732, "step": 41789 }, { "epoch": 3.2750783699059562, "grad_norm": 2.2621302604675293, "learning_rate": 7.917632073578002e-07, "loss": 1.363, "step": 41790 }, { "epoch": 3.275156739811912, "grad_norm": 2.7686638832092285, "learning_rate": 7.915966913626577e-07, "loss": 1.1328, "step": 41791 }, { "epoch": 3.2752351097178685, "grad_norm": 3.4472739696502686, "learning_rate": 7.914301913741185e-07, "loss": 1.2353, "step": 41792 }, { "epoch": 3.2753134796238244, "grad_norm": 3.4194023609161377, "learning_rate": 7.912637073928198e-07, "loss": 1.5751, "step": 41793 }, { "epoch": 3.2753918495297807, "grad_norm": 3.968268871307373, "learning_rate": 7.910972394193911e-07, "loss": 1.0546, "step": 41794 }, { "epoch": 3.2754702194357366, "grad_norm": 2.992401599884033, "learning_rate": 7.90930787454467e-07, "loss": 1.1392, "step": 41795 }, { "epoch": 3.275548589341693, "grad_norm": 3.074877977371216, "learning_rate": 7.907643514986796e-07, "loss": 1.2342, "step": 41796 }, { "epoch": 3.275626959247649, "grad_norm": 6.842636585235596, "learning_rate": 7.90597931552663e-07, "loss": 1.5272, "step": 41797 }, { "epoch": 3.275705329153605, "grad_norm": 3.195587635040283, "learning_rate": 7.904315276170493e-07, "loss": 1.5292, "step": 41798 }, { "epoch": 3.275783699059561, "grad_norm": 2.52164888381958, "learning_rate": 7.902651396924721e-07, "loss": 1.1418, "step": 41799 }, { "epoch": 3.2758620689655173, "grad_norm": 2.841083288192749, "learning_rate": 7.900987677795646e-07, "loss": 1.3708, "step": 41800 }, { "epoch": 3.2759404388714732, "grad_norm": 4.198544502258301, "learning_rate": 7.899324118789587e-07, "loss": 1.3785, "step": 41801 }, { "epoch": 3.2760188087774296, "grad_norm": 3.883941888809204, "learning_rate": 7.89766071991287e-07, "loss": 1.2491, "step": 41802 }, { "epoch": 3.2760971786833855, "grad_norm": 3.3925533294677734, "learning_rate": 7.89599748117183e-07, "loss": 1.3049, "step": 41803 }, { "epoch": 3.276175548589342, "grad_norm": 6.444717884063721, "learning_rate": 7.894334402572796e-07, "loss": 1.2978, "step": 41804 }, { "epoch": 3.2762539184952977, "grad_norm": 4.669622421264648, "learning_rate": 7.892671484122089e-07, "loss": 1.3425, "step": 41805 }, { "epoch": 3.276332288401254, "grad_norm": 2.356672763824463, "learning_rate": 7.89100872582601e-07, "loss": 1.0136, "step": 41806 }, { "epoch": 3.27641065830721, "grad_norm": 13.610650062561035, "learning_rate": 7.889346127690922e-07, "loss": 1.4355, "step": 41807 }, { "epoch": 3.2764890282131662, "grad_norm": 5.023743629455566, "learning_rate": 7.887683689723141e-07, "loss": 1.501, "step": 41808 }, { "epoch": 3.276567398119122, "grad_norm": 17.449216842651367, "learning_rate": 7.886021411928968e-07, "loss": 1.521, "step": 41809 }, { "epoch": 3.2766457680250785, "grad_norm": 4.522799015045166, "learning_rate": 7.884359294314742e-07, "loss": 1.7329, "step": 41810 }, { "epoch": 3.2767241379310343, "grad_norm": 2.790881872177124, "learning_rate": 7.882697336886769e-07, "loss": 0.9163, "step": 41811 }, { "epoch": 3.2768025078369907, "grad_norm": 3.1751327514648438, "learning_rate": 7.881035539651405e-07, "loss": 1.1563, "step": 41812 }, { "epoch": 3.2768808777429466, "grad_norm": 2.9702396392822266, "learning_rate": 7.879373902614934e-07, "loss": 1.0868, "step": 41813 }, { "epoch": 3.276959247648903, "grad_norm": 2.682821035385132, "learning_rate": 7.87771242578369e-07, "loss": 1.5673, "step": 41814 }, { "epoch": 3.277037617554859, "grad_norm": 11.282590866088867, "learning_rate": 7.876051109163995e-07, "loss": 1.5951, "step": 41815 }, { "epoch": 3.277115987460815, "grad_norm": 2.5923287868499756, "learning_rate": 7.874389952762168e-07, "loss": 1.2993, "step": 41816 }, { "epoch": 3.277194357366771, "grad_norm": 4.63189172744751, "learning_rate": 7.872728956584519e-07, "loss": 1.6242, "step": 41817 }, { "epoch": 3.2772727272727273, "grad_norm": 5.252551555633545, "learning_rate": 7.871068120637376e-07, "loss": 1.4594, "step": 41818 }, { "epoch": 3.2773510971786832, "grad_norm": 2.3125991821289062, "learning_rate": 7.869407444927046e-07, "loss": 0.3383, "step": 41819 }, { "epoch": 3.2774294670846396, "grad_norm": 3.0388453006744385, "learning_rate": 7.867746929459852e-07, "loss": 1.2101, "step": 41820 }, { "epoch": 3.2775078369905954, "grad_norm": 3.361607789993286, "learning_rate": 7.86608657424211e-07, "loss": 1.4856, "step": 41821 }, { "epoch": 3.277586206896552, "grad_norm": 2.561685800552368, "learning_rate": 7.864426379280132e-07, "loss": 1.3544, "step": 41822 }, { "epoch": 3.2776645768025077, "grad_norm": 9.731929779052734, "learning_rate": 7.862766344580247e-07, "loss": 1.3183, "step": 41823 }, { "epoch": 3.277742946708464, "grad_norm": 3.017716646194458, "learning_rate": 7.861106470148744e-07, "loss": 1.0436, "step": 41824 }, { "epoch": 3.27782131661442, "grad_norm": 3.1472771167755127, "learning_rate": 7.859446755991934e-07, "loss": 1.4548, "step": 41825 }, { "epoch": 3.277899686520376, "grad_norm": 4.576779842376709, "learning_rate": 7.857787202116157e-07, "loss": 1.2148, "step": 41826 }, { "epoch": 3.277978056426332, "grad_norm": 2.5506579875946045, "learning_rate": 7.856127808527725e-07, "loss": 1.4397, "step": 41827 }, { "epoch": 3.2780564263322884, "grad_norm": 2.0181002616882324, "learning_rate": 7.854468575232927e-07, "loss": 1.1003, "step": 41828 }, { "epoch": 3.2781347962382443, "grad_norm": 3.9382526874542236, "learning_rate": 7.852809502238079e-07, "loss": 1.6829, "step": 41829 }, { "epoch": 3.2782131661442007, "grad_norm": 3.309643507003784, "learning_rate": 7.851150589549483e-07, "loss": 1.4382, "step": 41830 }, { "epoch": 3.2782915360501566, "grad_norm": 6.247401714324951, "learning_rate": 7.849491837173484e-07, "loss": 1.2321, "step": 41831 }, { "epoch": 3.278369905956113, "grad_norm": 2.366522789001465, "learning_rate": 7.84783324511636e-07, "loss": 1.4732, "step": 41832 }, { "epoch": 3.2784482758620688, "grad_norm": 3.7878544330596924, "learning_rate": 7.846174813384428e-07, "loss": 1.2969, "step": 41833 }, { "epoch": 3.278526645768025, "grad_norm": 3.0413246154785156, "learning_rate": 7.844516541983993e-07, "loss": 1.2763, "step": 41834 }, { "epoch": 3.278605015673981, "grad_norm": 3.58408260345459, "learning_rate": 7.842858430921363e-07, "loss": 1.5408, "step": 41835 }, { "epoch": 3.2786833855799373, "grad_norm": 3.3177804946899414, "learning_rate": 7.841200480202849e-07, "loss": 1.399, "step": 41836 }, { "epoch": 3.2787617554858937, "grad_norm": 3.0134260654449463, "learning_rate": 7.839542689834751e-07, "loss": 1.3851, "step": 41837 }, { "epoch": 3.2788401253918495, "grad_norm": 2.927640438079834, "learning_rate": 7.837885059823391e-07, "loss": 1.0479, "step": 41838 }, { "epoch": 3.2789184952978054, "grad_norm": 3.460357666015625, "learning_rate": 7.836227590175038e-07, "loss": 1.7063, "step": 41839 }, { "epoch": 3.2789968652037618, "grad_norm": 3.50355863571167, "learning_rate": 7.834570280896031e-07, "loss": 0.9057, "step": 41840 }, { "epoch": 3.279075235109718, "grad_norm": 5.67465877532959, "learning_rate": 7.832913131992664e-07, "loss": 0.6897, "step": 41841 }, { "epoch": 3.279153605015674, "grad_norm": 3.784224271774292, "learning_rate": 7.831256143471244e-07, "loss": 1.3035, "step": 41842 }, { "epoch": 3.27923197492163, "grad_norm": 5.769038677215576, "learning_rate": 7.829599315338054e-07, "loss": 1.5602, "step": 41843 }, { "epoch": 3.279310344827586, "grad_norm": 2.6902129650115967, "learning_rate": 7.827942647599402e-07, "loss": 1.0026, "step": 41844 }, { "epoch": 3.2793887147335425, "grad_norm": 24.24971580505371, "learning_rate": 7.826286140261602e-07, "loss": 1.1149, "step": 41845 }, { "epoch": 3.2794670846394984, "grad_norm": 3.967076539993286, "learning_rate": 7.824629793330957e-07, "loss": 0.9772, "step": 41846 }, { "epoch": 3.2795454545454543, "grad_norm": 4.736400127410889, "learning_rate": 7.82297360681375e-07, "loss": 0.7682, "step": 41847 }, { "epoch": 3.2796238244514107, "grad_norm": 5.345719337463379, "learning_rate": 7.821317580716287e-07, "loss": 0.7628, "step": 41848 }, { "epoch": 3.279702194357367, "grad_norm": 2.5842654705047607, "learning_rate": 7.819661715044869e-07, "loss": 1.34, "step": 41849 }, { "epoch": 3.279780564263323, "grad_norm": 3.3459746837615967, "learning_rate": 7.818006009805789e-07, "loss": 1.1559, "step": 41850 }, { "epoch": 3.2798589341692788, "grad_norm": 3.5866823196411133, "learning_rate": 7.816350465005351e-07, "loss": 1.3548, "step": 41851 }, { "epoch": 3.279937304075235, "grad_norm": 2.8593177795410156, "learning_rate": 7.814695080649853e-07, "loss": 1.4681, "step": 41852 }, { "epoch": 3.2800156739811914, "grad_norm": 8.930438041687012, "learning_rate": 7.81303985674558e-07, "loss": 0.8673, "step": 41853 }, { "epoch": 3.2800940438871473, "grad_norm": 4.098429203033447, "learning_rate": 7.811384793298837e-07, "loss": 0.7149, "step": 41854 }, { "epoch": 3.280172413793103, "grad_norm": 3.331829071044922, "learning_rate": 7.80972989031592e-07, "loss": 1.0809, "step": 41855 }, { "epoch": 3.2802507836990595, "grad_norm": 6.598948955535889, "learning_rate": 7.808075147803118e-07, "loss": 1.5966, "step": 41856 }, { "epoch": 3.280329153605016, "grad_norm": 3.820042610168457, "learning_rate": 7.806420565766737e-07, "loss": 1.1883, "step": 41857 }, { "epoch": 3.2804075235109718, "grad_norm": 3.664787769317627, "learning_rate": 7.804766144213039e-07, "loss": 0.7858, "step": 41858 }, { "epoch": 3.280485893416928, "grad_norm": 3.1136155128479004, "learning_rate": 7.803111883148345e-07, "loss": 1.1818, "step": 41859 }, { "epoch": 3.280564263322884, "grad_norm": 4.080559730529785, "learning_rate": 7.801457782578947e-07, "loss": 1.2302, "step": 41860 }, { "epoch": 3.2806426332288403, "grad_norm": 3.2202978134155273, "learning_rate": 7.799803842511133e-07, "loss": 1.4927, "step": 41861 }, { "epoch": 3.280721003134796, "grad_norm": 3.4164187908172607, "learning_rate": 7.798150062951187e-07, "loss": 1.5385, "step": 41862 }, { "epoch": 3.2807993730407525, "grad_norm": 2.8342819213867188, "learning_rate": 7.796496443905382e-07, "loss": 1.1753, "step": 41863 }, { "epoch": 3.2808777429467084, "grad_norm": 2.2188689708709717, "learning_rate": 7.794842985380058e-07, "loss": 1.2577, "step": 41864 }, { "epoch": 3.2809561128526648, "grad_norm": 2.55281400680542, "learning_rate": 7.793189687381453e-07, "loss": 1.2325, "step": 41865 }, { "epoch": 3.2810344827586206, "grad_norm": 4.602710247039795, "learning_rate": 7.791536549915884e-07, "loss": 1.0853, "step": 41866 }, { "epoch": 3.281112852664577, "grad_norm": 2.5725748538970947, "learning_rate": 7.789883572989626e-07, "loss": 1.3324, "step": 41867 }, { "epoch": 3.281191222570533, "grad_norm": 11.940597534179688, "learning_rate": 7.788230756608972e-07, "loss": 1.4226, "step": 41868 }, { "epoch": 3.281269592476489, "grad_norm": 3.9239025115966797, "learning_rate": 7.786578100780206e-07, "loss": 1.3017, "step": 41869 }, { "epoch": 3.281347962382445, "grad_norm": 3.594111204147339, "learning_rate": 7.784925605509613e-07, "loss": 1.4546, "step": 41870 }, { "epoch": 3.2814263322884014, "grad_norm": 2.312131881713867, "learning_rate": 7.783273270803482e-07, "loss": 1.3593, "step": 41871 }, { "epoch": 3.2815047021943573, "grad_norm": 3.328317880630493, "learning_rate": 7.781621096668096e-07, "loss": 1.3819, "step": 41872 }, { "epoch": 3.2815830721003136, "grad_norm": 4.075932502746582, "learning_rate": 7.779969083109734e-07, "loss": 1.445, "step": 41873 }, { "epoch": 3.2816614420062695, "grad_norm": 3.452963352203369, "learning_rate": 7.778317230134691e-07, "loss": 1.3561, "step": 41874 }, { "epoch": 3.281739811912226, "grad_norm": 2.9920341968536377, "learning_rate": 7.77666553774924e-07, "loss": 1.4774, "step": 41875 }, { "epoch": 3.2818181818181817, "grad_norm": 8.597947120666504, "learning_rate": 7.775014005959663e-07, "loss": 2.2833, "step": 41876 }, { "epoch": 3.281896551724138, "grad_norm": 4.3367719650268555, "learning_rate": 7.773362634772252e-07, "loss": 1.4127, "step": 41877 }, { "epoch": 3.281974921630094, "grad_norm": 2.4949913024902344, "learning_rate": 7.771711424193273e-07, "loss": 1.183, "step": 41878 }, { "epoch": 3.2820532915360503, "grad_norm": 10.487630844116211, "learning_rate": 7.770060374229027e-07, "loss": 1.399, "step": 41879 }, { "epoch": 3.282131661442006, "grad_norm": 2.6219542026519775, "learning_rate": 7.768409484885775e-07, "loss": 1.0654, "step": 41880 }, { "epoch": 3.2822100313479625, "grad_norm": 2.7418625354766846, "learning_rate": 7.766758756169784e-07, "loss": 0.9097, "step": 41881 }, { "epoch": 3.2822884012539184, "grad_norm": 5.501707077026367, "learning_rate": 7.765108188087361e-07, "loss": 1.1426, "step": 41882 }, { "epoch": 3.2823667711598747, "grad_norm": 2.166116714477539, "learning_rate": 7.763457780644789e-07, "loss": 1.5543, "step": 41883 }, { "epoch": 3.2824451410658306, "grad_norm": 2.4687578678131104, "learning_rate": 7.761807533848315e-07, "loss": 1.3335, "step": 41884 }, { "epoch": 3.282523510971787, "grad_norm": 2.8186585903167725, "learning_rate": 7.760157447704236e-07, "loss": 1.5188, "step": 41885 }, { "epoch": 3.282601880877743, "grad_norm": 2.4591822624206543, "learning_rate": 7.758507522218805e-07, "loss": 1.1361, "step": 41886 }, { "epoch": 3.282680250783699, "grad_norm": 4.847532749176025, "learning_rate": 7.756857757398339e-07, "loss": 1.2394, "step": 41887 }, { "epoch": 3.282758620689655, "grad_norm": 4.6866984367370605, "learning_rate": 7.755208153249078e-07, "loss": 1.123, "step": 41888 }, { "epoch": 3.2828369905956114, "grad_norm": 3.821868896484375, "learning_rate": 7.753558709777304e-07, "loss": 1.3839, "step": 41889 }, { "epoch": 3.2829153605015673, "grad_norm": 2.7435190677642822, "learning_rate": 7.751909426989296e-07, "loss": 1.3906, "step": 41890 }, { "epoch": 3.2829937304075236, "grad_norm": 2.5447559356689453, "learning_rate": 7.750260304891327e-07, "loss": 1.0314, "step": 41891 }, { "epoch": 3.2830721003134795, "grad_norm": 2.929863929748535, "learning_rate": 7.748611343489665e-07, "loss": 1.3019, "step": 41892 }, { "epoch": 3.283150470219436, "grad_norm": 4.1537699699401855, "learning_rate": 7.746962542790581e-07, "loss": 0.7454, "step": 41893 }, { "epoch": 3.2832288401253917, "grad_norm": 3.314617156982422, "learning_rate": 7.745313902800361e-07, "loss": 1.2866, "step": 41894 }, { "epoch": 3.283307210031348, "grad_norm": 5.248823165893555, "learning_rate": 7.743665423525243e-07, "loss": 1.2527, "step": 41895 }, { "epoch": 3.283385579937304, "grad_norm": 4.595693111419678, "learning_rate": 7.742017104971527e-07, "loss": 1.4088, "step": 41896 }, { "epoch": 3.2834639498432603, "grad_norm": 3.3190269470214844, "learning_rate": 7.740368947145477e-07, "loss": 1.1009, "step": 41897 }, { "epoch": 3.283542319749216, "grad_norm": 3.0564229488372803, "learning_rate": 7.738720950053363e-07, "loss": 1.534, "step": 41898 }, { "epoch": 3.2836206896551725, "grad_norm": 2.340831756591797, "learning_rate": 7.737073113701443e-07, "loss": 1.406, "step": 41899 }, { "epoch": 3.2836990595611284, "grad_norm": 3.574035167694092, "learning_rate": 7.735425438095972e-07, "loss": 1.3927, "step": 41900 }, { "epoch": 3.2837774294670847, "grad_norm": 3.2807016372680664, "learning_rate": 7.733777923243247e-07, "loss": 1.6087, "step": 41901 }, { "epoch": 3.2838557993730406, "grad_norm": 4.196867942810059, "learning_rate": 7.732130569149538e-07, "loss": 1.7182, "step": 41902 }, { "epoch": 3.283934169278997, "grad_norm": 4.83383846282959, "learning_rate": 7.730483375821074e-07, "loss": 1.3217, "step": 41903 }, { "epoch": 3.284012539184953, "grad_norm": 5.435274600982666, "learning_rate": 7.728836343264146e-07, "loss": 1.7199, "step": 41904 }, { "epoch": 3.284090909090909, "grad_norm": 6.453385829925537, "learning_rate": 7.727189471485013e-07, "loss": 1.2101, "step": 41905 }, { "epoch": 3.284169278996865, "grad_norm": 3.9886319637298584, "learning_rate": 7.725542760489935e-07, "loss": 1.2865, "step": 41906 }, { "epoch": 3.2842476489028214, "grad_norm": 3.6759212017059326, "learning_rate": 7.723896210285181e-07, "loss": 1.3342, "step": 41907 }, { "epoch": 3.2843260188087773, "grad_norm": 2.6296722888946533, "learning_rate": 7.722249820877009e-07, "loss": 1.3299, "step": 41908 }, { "epoch": 3.2844043887147336, "grad_norm": 2.437591314315796, "learning_rate": 7.720603592271686e-07, "loss": 1.1996, "step": 41909 }, { "epoch": 3.2844827586206895, "grad_norm": 3.7261478900909424, "learning_rate": 7.718957524475473e-07, "loss": 1.2308, "step": 41910 }, { "epoch": 3.284561128526646, "grad_norm": 3.4347264766693115, "learning_rate": 7.717311617494627e-07, "loss": 1.4377, "step": 41911 }, { "epoch": 3.2846394984326017, "grad_norm": 3.6928863525390625, "learning_rate": 7.71566587133541e-07, "loss": 1.7066, "step": 41912 }, { "epoch": 3.284717868338558, "grad_norm": 4.642910003662109, "learning_rate": 7.714020286004092e-07, "loss": 1.5891, "step": 41913 }, { "epoch": 3.284796238244514, "grad_norm": 5.111918926239014, "learning_rate": 7.712374861506893e-07, "loss": 0.9298, "step": 41914 }, { "epoch": 3.2848746081504703, "grad_norm": 6.5483293533325195, "learning_rate": 7.71072959785012e-07, "loss": 1.2853, "step": 41915 }, { "epoch": 3.284952978056426, "grad_norm": 3.127070665359497, "learning_rate": 7.709084495040009e-07, "loss": 1.781, "step": 41916 }, { "epoch": 3.2850313479623825, "grad_norm": 3.428738594055176, "learning_rate": 7.707439553082824e-07, "loss": 0.9335, "step": 41917 }, { "epoch": 3.2851097178683384, "grad_norm": 5.259005069732666, "learning_rate": 7.70579477198481e-07, "loss": 1.3895, "step": 41918 }, { "epoch": 3.2851880877742947, "grad_norm": 2.1377170085906982, "learning_rate": 7.704150151752216e-07, "loss": 1.0714, "step": 41919 }, { "epoch": 3.2852664576802506, "grad_norm": 4.84486722946167, "learning_rate": 7.702505692391332e-07, "loss": 0.9551, "step": 41920 }, { "epoch": 3.285344827586207, "grad_norm": 3.1750588417053223, "learning_rate": 7.700861393908382e-07, "loss": 0.6289, "step": 41921 }, { "epoch": 3.285423197492163, "grad_norm": 2.439819097518921, "learning_rate": 7.699217256309626e-07, "loss": 0.9184, "step": 41922 }, { "epoch": 3.285501567398119, "grad_norm": 3.286609172821045, "learning_rate": 7.697573279601328e-07, "loss": 1.3521, "step": 41923 }, { "epoch": 3.285579937304075, "grad_norm": 2.2091264724731445, "learning_rate": 7.695929463789725e-07, "loss": 0.8806, "step": 41924 }, { "epoch": 3.2856583072100314, "grad_norm": 3.9396870136260986, "learning_rate": 7.69428580888108e-07, "loss": 1.073, "step": 41925 }, { "epoch": 3.2857366771159873, "grad_norm": 3.3499021530151367, "learning_rate": 7.692642314881643e-07, "loss": 1.224, "step": 41926 }, { "epoch": 3.2858150470219436, "grad_norm": 2.3951542377471924, "learning_rate": 7.690998981797665e-07, "loss": 1.5189, "step": 41927 }, { "epoch": 3.2858934169279, "grad_norm": 5.271578311920166, "learning_rate": 7.689355809635396e-07, "loss": 1.3437, "step": 41928 }, { "epoch": 3.285971786833856, "grad_norm": 2.2606120109558105, "learning_rate": 7.687712798401081e-07, "loss": 1.3383, "step": 41929 }, { "epoch": 3.2860501567398117, "grad_norm": 2.8157365322113037, "learning_rate": 7.686069948100978e-07, "loss": 1.3713, "step": 41930 }, { "epoch": 3.286128526645768, "grad_norm": 3.609992742538452, "learning_rate": 7.684427258741328e-07, "loss": 1.4796, "step": 41931 }, { "epoch": 3.2862068965517244, "grad_norm": 2.925727367401123, "learning_rate": 7.682784730328396e-07, "loss": 1.2122, "step": 41932 }, { "epoch": 3.2862852664576803, "grad_norm": 2.524388313293457, "learning_rate": 7.681142362868393e-07, "loss": 0.9475, "step": 41933 }, { "epoch": 3.286363636363636, "grad_norm": 2.694261312484741, "learning_rate": 7.679500156367597e-07, "loss": 1.3797, "step": 41934 }, { "epoch": 3.2864420062695925, "grad_norm": 3.472752809524536, "learning_rate": 7.677858110832259e-07, "loss": 1.45, "step": 41935 }, { "epoch": 3.286520376175549, "grad_norm": 2.4559333324432373, "learning_rate": 7.676216226268596e-07, "loss": 1.6005, "step": 41936 }, { "epoch": 3.2865987460815047, "grad_norm": 2.3045222759246826, "learning_rate": 7.674574502682869e-07, "loss": 1.2293, "step": 41937 }, { "epoch": 3.2866771159874606, "grad_norm": 3.5245521068573, "learning_rate": 7.672932940081313e-07, "loss": 1.482, "step": 41938 }, { "epoch": 3.286755485893417, "grad_norm": 2.5391414165496826, "learning_rate": 7.671291538470199e-07, "loss": 1.4415, "step": 41939 }, { "epoch": 3.2868338557993733, "grad_norm": 3.230898857116699, "learning_rate": 7.669650297855735e-07, "loss": 0.9487, "step": 41940 }, { "epoch": 3.286912225705329, "grad_norm": 3.0733344554901123, "learning_rate": 7.668009218244188e-07, "loss": 1.4391, "step": 41941 }, { "epoch": 3.286990595611285, "grad_norm": 7.095649242401123, "learning_rate": 7.666368299641785e-07, "loss": 1.1698, "step": 41942 }, { "epoch": 3.2870689655172414, "grad_norm": 4.3728837966918945, "learning_rate": 7.664727542054773e-07, "loss": 1.3295, "step": 41943 }, { "epoch": 3.2871473354231977, "grad_norm": 12.183097839355469, "learning_rate": 7.663086945489396e-07, "loss": 2.046, "step": 41944 }, { "epoch": 3.2872257053291536, "grad_norm": 3.594444751739502, "learning_rate": 7.661446509951887e-07, "loss": 1.1781, "step": 41945 }, { "epoch": 3.2873040752351095, "grad_norm": 9.147128105163574, "learning_rate": 7.659806235448502e-07, "loss": 1.236, "step": 41946 }, { "epoch": 3.287382445141066, "grad_norm": 4.8199782371521, "learning_rate": 7.658166121985444e-07, "loss": 2.1085, "step": 41947 }, { "epoch": 3.287460815047022, "grad_norm": 4.221709728240967, "learning_rate": 7.656526169568989e-07, "loss": 1.536, "step": 41948 }, { "epoch": 3.287539184952978, "grad_norm": 2.978985548019409, "learning_rate": 7.654886378205356e-07, "loss": 1.2508, "step": 41949 }, { "epoch": 3.287617554858934, "grad_norm": 3.661106824874878, "learning_rate": 7.653246747900794e-07, "loss": 1.26, "step": 41950 }, { "epoch": 3.2876959247648903, "grad_norm": 4.991006374359131, "learning_rate": 7.651607278661526e-07, "loss": 1.9436, "step": 41951 }, { "epoch": 3.2877742946708466, "grad_norm": 3.690957546234131, "learning_rate": 7.649967970493777e-07, "loss": 1.2646, "step": 41952 }, { "epoch": 3.2878526645768025, "grad_norm": 5.494324207305908, "learning_rate": 7.648328823403805e-07, "loss": 1.4791, "step": 41953 }, { "epoch": 3.287931034482759, "grad_norm": 2.8369128704071045, "learning_rate": 7.646689837397852e-07, "loss": 1.4015, "step": 41954 }, { "epoch": 3.2880094043887147, "grad_norm": 3.1115756034851074, "learning_rate": 7.64505101248213e-07, "loss": 1.5358, "step": 41955 }, { "epoch": 3.288087774294671, "grad_norm": 2.7459516525268555, "learning_rate": 7.643412348662871e-07, "loss": 1.3154, "step": 41956 }, { "epoch": 3.288166144200627, "grad_norm": 3.704817533493042, "learning_rate": 7.641773845946309e-07, "loss": 1.3779, "step": 41957 }, { "epoch": 3.2882445141065832, "grad_norm": 24.08644676208496, "learning_rate": 7.640135504338709e-07, "loss": 1.0041, "step": 41958 }, { "epoch": 3.288322884012539, "grad_norm": 2.432560443878174, "learning_rate": 7.638497323846256e-07, "loss": 1.2486, "step": 41959 }, { "epoch": 3.2884012539184955, "grad_norm": 3.714625120162964, "learning_rate": 7.636859304475208e-07, "loss": 1.4457, "step": 41960 }, { "epoch": 3.2884796238244514, "grad_norm": 3.2134759426116943, "learning_rate": 7.635221446231783e-07, "loss": 1.2949, "step": 41961 }, { "epoch": 3.2885579937304077, "grad_norm": 3.297150135040283, "learning_rate": 7.633583749122219e-07, "loss": 1.4937, "step": 41962 }, { "epoch": 3.2886363636363636, "grad_norm": 3.8003005981445312, "learning_rate": 7.631946213152742e-07, "loss": 1.4791, "step": 41963 }, { "epoch": 3.28871473354232, "grad_norm": 6.718493461608887, "learning_rate": 7.630308838329581e-07, "loss": 1.0279, "step": 41964 }, { "epoch": 3.288793103448276, "grad_norm": 3.466996669769287, "learning_rate": 7.628671624658957e-07, "loss": 1.8636, "step": 41965 }, { "epoch": 3.288871473354232, "grad_norm": 2.988407611846924, "learning_rate": 7.627034572147107e-07, "loss": 1.4574, "step": 41966 }, { "epoch": 3.288949843260188, "grad_norm": 4.753353118896484, "learning_rate": 7.625397680800256e-07, "loss": 1.7526, "step": 41967 }, { "epoch": 3.2890282131661444, "grad_norm": 3.8774752616882324, "learning_rate": 7.62376095062462e-07, "loss": 1.5106, "step": 41968 }, { "epoch": 3.2891065830721002, "grad_norm": 3.860321044921875, "learning_rate": 7.62212438162645e-07, "loss": 1.2416, "step": 41969 }, { "epoch": 3.2891849529780566, "grad_norm": 3.2267796993255615, "learning_rate": 7.620487973811936e-07, "loss": 1.2556, "step": 41970 }, { "epoch": 3.2892633228840125, "grad_norm": 2.476505756378174, "learning_rate": 7.618851727187309e-07, "loss": 1.249, "step": 41971 }, { "epoch": 3.289341692789969, "grad_norm": 5.291979789733887, "learning_rate": 7.617215641758812e-07, "loss": 1.2399, "step": 41972 }, { "epoch": 3.2894200626959247, "grad_norm": 7.319622039794922, "learning_rate": 7.615579717532673e-07, "loss": 1.0415, "step": 41973 }, { "epoch": 3.289498432601881, "grad_norm": 3.094092845916748, "learning_rate": 7.61394395451508e-07, "loss": 0.7512, "step": 41974 }, { "epoch": 3.289576802507837, "grad_norm": 2.6018264293670654, "learning_rate": 7.612308352712277e-07, "loss": 1.267, "step": 41975 }, { "epoch": 3.2896551724137932, "grad_norm": 2.785771608352661, "learning_rate": 7.610672912130484e-07, "loss": 0.9742, "step": 41976 }, { "epoch": 3.289733542319749, "grad_norm": 19.818769454956055, "learning_rate": 7.609037632775917e-07, "loss": 2.2614, "step": 41977 }, { "epoch": 3.2898119122257055, "grad_norm": 4.871673583984375, "learning_rate": 7.607402514654799e-07, "loss": 1.2386, "step": 41978 }, { "epoch": 3.2898902821316613, "grad_norm": 5.28737211227417, "learning_rate": 7.605767557773341e-07, "loss": 1.5829, "step": 41979 }, { "epoch": 3.2899686520376177, "grad_norm": 3.8907032012939453, "learning_rate": 7.604132762137773e-07, "loss": 1.2476, "step": 41980 }, { "epoch": 3.2900470219435736, "grad_norm": 3.6412928104400635, "learning_rate": 7.602498127754309e-07, "loss": 1.4797, "step": 41981 }, { "epoch": 3.29012539184953, "grad_norm": 9.96768569946289, "learning_rate": 7.600863654629164e-07, "loss": 1.1863, "step": 41982 }, { "epoch": 3.290203761755486, "grad_norm": 3.1917402744293213, "learning_rate": 7.599229342768555e-07, "loss": 1.4624, "step": 41983 }, { "epoch": 3.290282131661442, "grad_norm": 2.568467617034912, "learning_rate": 7.597595192178702e-07, "loss": 1.3481, "step": 41984 }, { "epoch": 3.290360501567398, "grad_norm": 4.085702419281006, "learning_rate": 7.595961202865814e-07, "loss": 1.2027, "step": 41985 }, { "epoch": 3.2904388714733543, "grad_norm": 8.184416770935059, "learning_rate": 7.594327374836114e-07, "loss": 1.8664, "step": 41986 }, { "epoch": 3.2905172413793102, "grad_norm": 4.18576192855835, "learning_rate": 7.592693708095811e-07, "loss": 1.2878, "step": 41987 }, { "epoch": 3.2905956112852666, "grad_norm": 13.258537292480469, "learning_rate": 7.591060202651129e-07, "loss": 1.6525, "step": 41988 }, { "epoch": 3.2906739811912225, "grad_norm": 2.373169422149658, "learning_rate": 7.589426858508248e-07, "loss": 1.3979, "step": 41989 }, { "epoch": 3.290752351097179, "grad_norm": 12.212859153747559, "learning_rate": 7.587793675673416e-07, "loss": 1.8439, "step": 41990 }, { "epoch": 3.2908307210031347, "grad_norm": 2.414940595626831, "learning_rate": 7.586160654152841e-07, "loss": 0.6946, "step": 41991 }, { "epoch": 3.290909090909091, "grad_norm": 3.9502639770507812, "learning_rate": 7.584527793952718e-07, "loss": 1.3924, "step": 41992 }, { "epoch": 3.290987460815047, "grad_norm": 3.7012810707092285, "learning_rate": 7.582895095079262e-07, "loss": 0.8936, "step": 41993 }, { "epoch": 3.2910658307210032, "grad_norm": 2.6924586296081543, "learning_rate": 7.581262557538683e-07, "loss": 1.3446, "step": 41994 }, { "epoch": 3.291144200626959, "grad_norm": 4.013758659362793, "learning_rate": 7.579630181337205e-07, "loss": 1.6194, "step": 41995 }, { "epoch": 3.2912225705329154, "grad_norm": 6.427102088928223, "learning_rate": 7.577997966481021e-07, "loss": 1.7085, "step": 41996 }, { "epoch": 3.2913009404388713, "grad_norm": 2.5462396144866943, "learning_rate": 7.576365912976342e-07, "loss": 1.2392, "step": 41997 }, { "epoch": 3.2913793103448277, "grad_norm": 4.749229431152344, "learning_rate": 7.574734020829378e-07, "loss": 1.4258, "step": 41998 }, { "epoch": 3.2914576802507836, "grad_norm": 3.5411458015441895, "learning_rate": 7.573102290046336e-07, "loss": 1.0198, "step": 41999 }, { "epoch": 3.29153605015674, "grad_norm": 4.755752086639404, "learning_rate": 7.571470720633418e-07, "loss": 1.1075, "step": 42000 }, { "epoch": 3.291614420062696, "grad_norm": 3.4572529792785645, "learning_rate": 7.569839312596833e-07, "loss": 1.0355, "step": 42001 }, { "epoch": 3.291692789968652, "grad_norm": 2.542285919189453, "learning_rate": 7.568208065942795e-07, "loss": 1.0457, "step": 42002 }, { "epoch": 3.291771159874608, "grad_norm": 3.8176071643829346, "learning_rate": 7.566576980677481e-07, "loss": 1.101, "step": 42003 }, { "epoch": 3.2918495297805643, "grad_norm": 4.143123626708984, "learning_rate": 7.564946056807121e-07, "loss": 1.0321, "step": 42004 }, { "epoch": 3.29192789968652, "grad_norm": 5.572765827178955, "learning_rate": 7.563315294337914e-07, "loss": 1.235, "step": 42005 }, { "epoch": 3.2920062695924766, "grad_norm": 14.228365898132324, "learning_rate": 7.561684693276066e-07, "loss": 0.8278, "step": 42006 }, { "epoch": 3.2920846394984324, "grad_norm": 4.403049945831299, "learning_rate": 7.560054253627763e-07, "loss": 1.4478, "step": 42007 }, { "epoch": 3.2921630094043888, "grad_norm": 2.4850594997406006, "learning_rate": 7.558423975399204e-07, "loss": 1.6162, "step": 42008 }, { "epoch": 3.2922413793103447, "grad_norm": 5.720689296722412, "learning_rate": 7.556793858596606e-07, "loss": 1.3868, "step": 42009 }, { "epoch": 3.292319749216301, "grad_norm": 2.7217679023742676, "learning_rate": 7.555163903226182e-07, "loss": 1.4353, "step": 42010 }, { "epoch": 3.292398119122257, "grad_norm": 4.565883636474609, "learning_rate": 7.553534109294103e-07, "loss": 1.173, "step": 42011 }, { "epoch": 3.292476489028213, "grad_norm": 3.0873849391937256, "learning_rate": 7.551904476806576e-07, "loss": 1.3084, "step": 42012 }, { "epoch": 3.292554858934169, "grad_norm": 7.263002395629883, "learning_rate": 7.550275005769786e-07, "loss": 1.4668, "step": 42013 }, { "epoch": 3.2926332288401254, "grad_norm": 2.2432303428649902, "learning_rate": 7.548645696189977e-07, "loss": 1.1843, "step": 42014 }, { "epoch": 3.2927115987460813, "grad_norm": 2.55085825920105, "learning_rate": 7.547016548073299e-07, "loss": 1.2908, "step": 42015 }, { "epoch": 3.2927899686520377, "grad_norm": 3.0398313999176025, "learning_rate": 7.545387561425965e-07, "loss": 1.225, "step": 42016 }, { "epoch": 3.2928683385579935, "grad_norm": 12.153975486755371, "learning_rate": 7.54375873625417e-07, "loss": 0.8979, "step": 42017 }, { "epoch": 3.29294670846395, "grad_norm": 8.858171463012695, "learning_rate": 7.542130072564114e-07, "loss": 0.5657, "step": 42018 }, { "epoch": 3.2930250783699058, "grad_norm": 5.148072242736816, "learning_rate": 7.540501570361985e-07, "loss": 0.8839, "step": 42019 }, { "epoch": 3.293103448275862, "grad_norm": 5.60209846496582, "learning_rate": 7.538873229653981e-07, "loss": 0.9944, "step": 42020 }, { "epoch": 3.293181818181818, "grad_norm": 15.375716209411621, "learning_rate": 7.537245050446301e-07, "loss": 1.49, "step": 42021 }, { "epoch": 3.2932601880877743, "grad_norm": 3.542067289352417, "learning_rate": 7.535617032745113e-07, "loss": 1.2615, "step": 42022 }, { "epoch": 3.2933385579937307, "grad_norm": 5.668873310089111, "learning_rate": 7.533989176556633e-07, "loss": 1.4999, "step": 42023 }, { "epoch": 3.2934169278996865, "grad_norm": 3.230161428451538, "learning_rate": 7.532361481887046e-07, "loss": 1.3093, "step": 42024 }, { "epoch": 3.2934952978056424, "grad_norm": 3.886613368988037, "learning_rate": 7.530733948742558e-07, "loss": 1.4943, "step": 42025 }, { "epoch": 3.2935736677115988, "grad_norm": 2.973607063293457, "learning_rate": 7.529106577129336e-07, "loss": 0.8212, "step": 42026 }, { "epoch": 3.293652037617555, "grad_norm": 2.1140334606170654, "learning_rate": 7.527479367053564e-07, "loss": 1.0481, "step": 42027 }, { "epoch": 3.293730407523511, "grad_norm": 3.1735236644744873, "learning_rate": 7.525852318521454e-07, "loss": 1.1599, "step": 42028 }, { "epoch": 3.293808777429467, "grad_norm": 3.00303316116333, "learning_rate": 7.524225431539201e-07, "loss": 1.2706, "step": 42029 }, { "epoch": 3.293887147335423, "grad_norm": 2.7330758571624756, "learning_rate": 7.52259870611296e-07, "loss": 0.8612, "step": 42030 }, { "epoch": 3.2939655172413795, "grad_norm": 2.7554147243499756, "learning_rate": 7.520972142248939e-07, "loss": 1.3023, "step": 42031 }, { "epoch": 3.2940438871473354, "grad_norm": 4.249457836151123, "learning_rate": 7.519345739953321e-07, "loss": 0.8589, "step": 42032 }, { "epoch": 3.2941222570532913, "grad_norm": 3.9512441158294678, "learning_rate": 7.517719499232295e-07, "loss": 1.6558, "step": 42033 }, { "epoch": 3.2942006269592476, "grad_norm": 2.4512038230895996, "learning_rate": 7.516093420092041e-07, "loss": 1.3118, "step": 42034 }, { "epoch": 3.294278996865204, "grad_norm": 2.4750170707702637, "learning_rate": 7.514467502538747e-07, "loss": 1.3559, "step": 42035 }, { "epoch": 3.29435736677116, "grad_norm": 4.8352952003479, "learning_rate": 7.512841746578598e-07, "loss": 1.3789, "step": 42036 }, { "epoch": 3.2944357366771158, "grad_norm": 5.7267045974731445, "learning_rate": 7.511216152217776e-07, "loss": 0.8457, "step": 42037 }, { "epoch": 3.294514106583072, "grad_norm": 2.9395079612731934, "learning_rate": 7.509590719462467e-07, "loss": 1.0995, "step": 42038 }, { "epoch": 3.2945924764890284, "grad_norm": 3.7843177318573, "learning_rate": 7.507965448318843e-07, "loss": 1.5211, "step": 42039 }, { "epoch": 3.2946708463949843, "grad_norm": 3.3258259296417236, "learning_rate": 7.506340338793111e-07, "loss": 1.2013, "step": 42040 }, { "epoch": 3.29474921630094, "grad_norm": 2.1464955806732178, "learning_rate": 7.504715390891409e-07, "loss": 1.2802, "step": 42041 }, { "epoch": 3.2948275862068965, "grad_norm": 3.2023701667785645, "learning_rate": 7.503090604619956e-07, "loss": 1.2855, "step": 42042 }, { "epoch": 3.294905956112853, "grad_norm": 3.781560182571411, "learning_rate": 7.501465979984917e-07, "loss": 1.4567, "step": 42043 }, { "epoch": 3.2949843260188088, "grad_norm": 3.008962392807007, "learning_rate": 7.499841516992485e-07, "loss": 1.9342, "step": 42044 }, { "epoch": 3.295062695924765, "grad_norm": 3.9296157360076904, "learning_rate": 7.498217215648812e-07, "loss": 1.7168, "step": 42045 }, { "epoch": 3.295141065830721, "grad_norm": 8.360983848571777, "learning_rate": 7.496593075960079e-07, "loss": 0.7581, "step": 42046 }, { "epoch": 3.2952194357366773, "grad_norm": 3.264030933380127, "learning_rate": 7.494969097932503e-07, "loss": 1.3569, "step": 42047 }, { "epoch": 3.295297805642633, "grad_norm": 3.768862724304199, "learning_rate": 7.493345281572218e-07, "loss": 1.2922, "step": 42048 }, { "epoch": 3.2953761755485895, "grad_norm": 2.7909646034240723, "learning_rate": 7.491721626885412e-07, "loss": 1.5036, "step": 42049 }, { "epoch": 3.2954545454545454, "grad_norm": 3.5082850456237793, "learning_rate": 7.490098133878265e-07, "loss": 1.1651, "step": 42050 }, { "epoch": 3.2955329153605017, "grad_norm": 2.5607621669769287, "learning_rate": 7.48847480255695e-07, "loss": 1.5116, "step": 42051 }, { "epoch": 3.2956112852664576, "grad_norm": 9.90036678314209, "learning_rate": 7.486851632927644e-07, "loss": 1.5459, "step": 42052 }, { "epoch": 3.295689655172414, "grad_norm": 2.8417296409606934, "learning_rate": 7.485228624996515e-07, "loss": 1.3272, "step": 42053 }, { "epoch": 3.29576802507837, "grad_norm": 2.125511646270752, "learning_rate": 7.483605778769742e-07, "loss": 1.3418, "step": 42054 }, { "epoch": 3.295846394984326, "grad_norm": 2.450963258743286, "learning_rate": 7.481983094253498e-07, "loss": 1.4125, "step": 42055 }, { "epoch": 3.295924764890282, "grad_norm": 3.5588366985321045, "learning_rate": 7.480360571453943e-07, "loss": 1.5721, "step": 42056 }, { "epoch": 3.2960031347962384, "grad_norm": 3.792987823486328, "learning_rate": 7.478738210377262e-07, "loss": 1.3538, "step": 42057 }, { "epoch": 3.2960815047021943, "grad_norm": 8.157456398010254, "learning_rate": 7.477116011029634e-07, "loss": 1.105, "step": 42058 }, { "epoch": 3.2961598746081506, "grad_norm": 2.9485340118408203, "learning_rate": 7.475493973417203e-07, "loss": 1.2163, "step": 42059 }, { "epoch": 3.2962382445141065, "grad_norm": 2.4590697288513184, "learning_rate": 7.473872097546136e-07, "loss": 1.5584, "step": 42060 }, { "epoch": 3.296316614420063, "grad_norm": 3.2437596321105957, "learning_rate": 7.47225038342263e-07, "loss": 1.5699, "step": 42061 }, { "epoch": 3.2963949843260187, "grad_norm": 2.44022536277771, "learning_rate": 7.470628831052851e-07, "loss": 1.1922, "step": 42062 }, { "epoch": 3.296473354231975, "grad_norm": 2.833272933959961, "learning_rate": 7.469007440442944e-07, "loss": 1.0985, "step": 42063 }, { "epoch": 3.296551724137931, "grad_norm": 4.599606037139893, "learning_rate": 7.467386211599092e-07, "loss": 1.2426, "step": 42064 }, { "epoch": 3.2966300940438873, "grad_norm": 2.238399028778076, "learning_rate": 7.465765144527437e-07, "loss": 1.1961, "step": 42065 }, { "epoch": 3.296708463949843, "grad_norm": 4.479377269744873, "learning_rate": 7.464144239234194e-07, "loss": 1.394, "step": 42066 }, { "epoch": 3.2967868338557995, "grad_norm": 3.6430277824401855, "learning_rate": 7.462523495725487e-07, "loss": 1.5918, "step": 42067 }, { "epoch": 3.2968652037617554, "grad_norm": 3.1293022632598877, "learning_rate": 7.460902914007489e-07, "loss": 1.2357, "step": 42068 }, { "epoch": 3.2969435736677117, "grad_norm": 3.567357301712036, "learning_rate": 7.459282494086368e-07, "loss": 1.2405, "step": 42069 }, { "epoch": 3.2970219435736676, "grad_norm": 4.551837921142578, "learning_rate": 7.457662235968283e-07, "loss": 1.4058, "step": 42070 }, { "epoch": 3.297100313479624, "grad_norm": 9.553112030029297, "learning_rate": 7.456042139659403e-07, "loss": 2.0331, "step": 42071 }, { "epoch": 3.29717868338558, "grad_norm": 2.931016206741333, "learning_rate": 7.45442220516589e-07, "loss": 1.355, "step": 42072 }, { "epoch": 3.297257053291536, "grad_norm": 6.239739418029785, "learning_rate": 7.452802432493894e-07, "loss": 1.2605, "step": 42073 }, { "epoch": 3.297335423197492, "grad_norm": 6.793613910675049, "learning_rate": 7.451182821649588e-07, "loss": 1.272, "step": 42074 }, { "epoch": 3.2974137931034484, "grad_norm": 3.270127058029175, "learning_rate": 7.449563372639129e-07, "loss": 1.6167, "step": 42075 }, { "epoch": 3.2974921630094043, "grad_norm": 3.9019505977630615, "learning_rate": 7.447944085468678e-07, "loss": 1.2404, "step": 42076 }, { "epoch": 3.2975705329153606, "grad_norm": 2.685584545135498, "learning_rate": 7.446324960144397e-07, "loss": 1.4349, "step": 42077 }, { "epoch": 3.2976489028213165, "grad_norm": 2.271059036254883, "learning_rate": 7.444705996672418e-07, "loss": 1.3416, "step": 42078 }, { "epoch": 3.297727272727273, "grad_norm": 3.726184129714966, "learning_rate": 7.443087195058935e-07, "loss": 1.1986, "step": 42079 }, { "epoch": 3.2978056426332287, "grad_norm": 2.4845988750457764, "learning_rate": 7.441468555310089e-07, "loss": 1.3098, "step": 42080 }, { "epoch": 3.297884012539185, "grad_norm": 4.989226341247559, "learning_rate": 7.439850077432042e-07, "loss": 1.9606, "step": 42081 }, { "epoch": 3.297962382445141, "grad_norm": 3.6455235481262207, "learning_rate": 7.438231761430942e-07, "loss": 1.4805, "step": 42082 }, { "epoch": 3.2980407523510973, "grad_norm": 3.6757655143737793, "learning_rate": 7.436613607312932e-07, "loss": 1.4084, "step": 42083 }, { "epoch": 3.298119122257053, "grad_norm": 3.356109619140625, "learning_rate": 7.43499561508419e-07, "loss": 1.3389, "step": 42084 }, { "epoch": 3.2981974921630095, "grad_norm": 2.2666361331939697, "learning_rate": 7.433377784750878e-07, "loss": 1.2151, "step": 42085 }, { "epoch": 3.2982758620689654, "grad_norm": 2.607241630554199, "learning_rate": 7.43176011631912e-07, "loss": 1.1994, "step": 42086 }, { "epoch": 3.2983542319749217, "grad_norm": 7.371274948120117, "learning_rate": 7.430142609795082e-07, "loss": 1.3148, "step": 42087 }, { "epoch": 3.2984326018808776, "grad_norm": 3.501194953918457, "learning_rate": 7.428525265184916e-07, "loss": 1.3751, "step": 42088 }, { "epoch": 3.298510971786834, "grad_norm": 4.126646041870117, "learning_rate": 7.426908082494777e-07, "loss": 1.5755, "step": 42089 }, { "epoch": 3.29858934169279, "grad_norm": 5.408682823181152, "learning_rate": 7.42529106173081e-07, "loss": 1.4753, "step": 42090 }, { "epoch": 3.298667711598746, "grad_norm": 4.382092475891113, "learning_rate": 7.423674202899167e-07, "loss": 1.2232, "step": 42091 }, { "epoch": 3.298746081504702, "grad_norm": 4.128396034240723, "learning_rate": 7.422057506006003e-07, "loss": 0.9927, "step": 42092 }, { "epoch": 3.2988244514106584, "grad_norm": 2.287468671798706, "learning_rate": 7.420440971057457e-07, "loss": 1.4248, "step": 42093 }, { "epoch": 3.2989028213166143, "grad_norm": 2.5331809520721436, "learning_rate": 7.418824598059687e-07, "loss": 1.4472, "step": 42094 }, { "epoch": 3.2989811912225706, "grad_norm": 2.599344491958618, "learning_rate": 7.417208387018837e-07, "loss": 1.2233, "step": 42095 }, { "epoch": 3.2990595611285265, "grad_norm": 11.899763107299805, "learning_rate": 7.415592337941064e-07, "loss": 0.6684, "step": 42096 }, { "epoch": 3.299137931034483, "grad_norm": 3.2705976963043213, "learning_rate": 7.413976450832483e-07, "loss": 1.5675, "step": 42097 }, { "epoch": 3.2992163009404387, "grad_norm": 3.5770390033721924, "learning_rate": 7.412360725699275e-07, "loss": 1.5405, "step": 42098 }, { "epoch": 3.299294670846395, "grad_norm": 7.6292829513549805, "learning_rate": 7.410745162547584e-07, "loss": 1.1359, "step": 42099 }, { "epoch": 3.299373040752351, "grad_norm": 4.847360610961914, "learning_rate": 7.409129761383527e-07, "loss": 1.2537, "step": 42100 }, { "epoch": 3.2994514106583073, "grad_norm": 3.3927104473114014, "learning_rate": 7.407514522213272e-07, "loss": 1.3141, "step": 42101 }, { "epoch": 3.299529780564263, "grad_norm": 2.5838849544525146, "learning_rate": 7.405899445042936e-07, "loss": 1.2286, "step": 42102 }, { "epoch": 3.2996081504702195, "grad_norm": 8.256556510925293, "learning_rate": 7.404284529878708e-07, "loss": 0.5916, "step": 42103 }, { "epoch": 3.2996865203761754, "grad_norm": 3.017148971557617, "learning_rate": 7.40266977672669e-07, "loss": 1.3743, "step": 42104 }, { "epoch": 3.2997648902821317, "grad_norm": 4.909350395202637, "learning_rate": 7.401055185593037e-07, "loss": 1.5506, "step": 42105 }, { "epoch": 3.2998432601880876, "grad_norm": 3.0507171154022217, "learning_rate": 7.399440756483889e-07, "loss": 1.3288, "step": 42106 }, { "epoch": 3.299921630094044, "grad_norm": 2.9170594215393066, "learning_rate": 7.397826489405391e-07, "loss": 1.2969, "step": 42107 }, { "epoch": 3.3, "grad_norm": 3.4833741188049316, "learning_rate": 7.396212384363677e-07, "loss": 1.3982, "step": 42108 }, { "epoch": 3.300078369905956, "grad_norm": 3.850149154663086, "learning_rate": 7.394598441364887e-07, "loss": 1.3494, "step": 42109 }, { "epoch": 3.300156739811912, "grad_norm": 4.0724639892578125, "learning_rate": 7.392984660415165e-07, "loss": 1.1437, "step": 42110 }, { "epoch": 3.3002351097178684, "grad_norm": 3.547304153442383, "learning_rate": 7.39137104152064e-07, "loss": 1.4717, "step": 42111 }, { "epoch": 3.3003134796238243, "grad_norm": 2.971829652786255, "learning_rate": 7.389757584687457e-07, "loss": 1.6642, "step": 42112 }, { "epoch": 3.3003918495297806, "grad_norm": 2.6225287914276123, "learning_rate": 7.388144289921751e-07, "loss": 1.6199, "step": 42113 }, { "epoch": 3.3004702194357365, "grad_norm": 5.867392063140869, "learning_rate": 7.386531157229665e-07, "loss": 1.2673, "step": 42114 }, { "epoch": 3.300548589341693, "grad_norm": 2.8338611125946045, "learning_rate": 7.384918186617319e-07, "loss": 0.9991, "step": 42115 }, { "epoch": 3.3006269592476487, "grad_norm": 3.078573703765869, "learning_rate": 7.383305378090844e-07, "loss": 1.3586, "step": 42116 }, { "epoch": 3.300705329153605, "grad_norm": 2.666165351867676, "learning_rate": 7.381692731656392e-07, "loss": 1.1704, "step": 42117 }, { "epoch": 3.3007836990595614, "grad_norm": 2.6967735290527344, "learning_rate": 7.380080247320109e-07, "loss": 1.4441, "step": 42118 }, { "epoch": 3.3008620689655173, "grad_norm": 5.422562599182129, "learning_rate": 7.378467925088095e-07, "loss": 1.2665, "step": 42119 }, { "epoch": 3.300940438871473, "grad_norm": 3.4170923233032227, "learning_rate": 7.376855764966501e-07, "loss": 1.3476, "step": 42120 }, { "epoch": 3.3010188087774295, "grad_norm": 2.9145634174346924, "learning_rate": 7.375243766961437e-07, "loss": 1.1212, "step": 42121 }, { "epoch": 3.301097178683386, "grad_norm": 4.537694931030273, "learning_rate": 7.373631931079084e-07, "loss": 1.5492, "step": 42122 }, { "epoch": 3.3011755485893417, "grad_norm": 3.2862815856933594, "learning_rate": 7.372020257325524e-07, "loss": 1.0859, "step": 42123 }, { "epoch": 3.3012539184952976, "grad_norm": 3.0286378860473633, "learning_rate": 7.370408745706903e-07, "loss": 1.5073, "step": 42124 }, { "epoch": 3.301332288401254, "grad_norm": 2.129680871963501, "learning_rate": 7.368797396229349e-07, "loss": 1.0587, "step": 42125 }, { "epoch": 3.3014106583072103, "grad_norm": 5.970065593719482, "learning_rate": 7.367186208898997e-07, "loss": 0.9419, "step": 42126 }, { "epoch": 3.301489028213166, "grad_norm": 5.957390308380127, "learning_rate": 7.365575183721973e-07, "loss": 1.6986, "step": 42127 }, { "epoch": 3.301567398119122, "grad_norm": 5.163851737976074, "learning_rate": 7.363964320704397e-07, "loss": 0.9732, "step": 42128 }, { "epoch": 3.3016457680250784, "grad_norm": 2.125778913497925, "learning_rate": 7.362353619852419e-07, "loss": 1.1477, "step": 42129 }, { "epoch": 3.3017241379310347, "grad_norm": 3.288346767425537, "learning_rate": 7.360743081172122e-07, "loss": 1.5435, "step": 42130 }, { "epoch": 3.3018025078369906, "grad_norm": 6.224820613861084, "learning_rate": 7.359132704669669e-07, "loss": 0.7268, "step": 42131 }, { "epoch": 3.3018808777429465, "grad_norm": 3.9083642959594727, "learning_rate": 7.357522490351172e-07, "loss": 1.4208, "step": 42132 }, { "epoch": 3.301959247648903, "grad_norm": 2.2758193016052246, "learning_rate": 7.355912438222767e-07, "loss": 1.4787, "step": 42133 }, { "epoch": 3.302037617554859, "grad_norm": 5.1967597007751465, "learning_rate": 7.35430254829056e-07, "loss": 1.1879, "step": 42134 }, { "epoch": 3.302115987460815, "grad_norm": 2.343400716781616, "learning_rate": 7.352692820560669e-07, "loss": 1.2701, "step": 42135 }, { "epoch": 3.302194357366771, "grad_norm": 5.597986221313477, "learning_rate": 7.351083255039243e-07, "loss": 1.7238, "step": 42136 }, { "epoch": 3.3022727272727272, "grad_norm": 5.562885284423828, "learning_rate": 7.349473851732397e-07, "loss": 1.0101, "step": 42137 }, { "epoch": 3.3023510971786836, "grad_norm": 3.2132909297943115, "learning_rate": 7.347864610646238e-07, "loss": 1.1423, "step": 42138 }, { "epoch": 3.3024294670846395, "grad_norm": 2.826007604598999, "learning_rate": 7.346255531786894e-07, "loss": 1.2912, "step": 42139 }, { "epoch": 3.302507836990596, "grad_norm": 2.507746696472168, "learning_rate": 7.344646615160472e-07, "loss": 1.1334, "step": 42140 }, { "epoch": 3.3025862068965517, "grad_norm": 4.135439872741699, "learning_rate": 7.343037860773133e-07, "loss": 1.1436, "step": 42141 }, { "epoch": 3.302664576802508, "grad_norm": 2.149609088897705, "learning_rate": 7.341429268630951e-07, "loss": 1.2944, "step": 42142 }, { "epoch": 3.302742946708464, "grad_norm": 5.322289943695068, "learning_rate": 7.339820838740064e-07, "loss": 1.3845, "step": 42143 }, { "epoch": 3.3028213166144202, "grad_norm": 5.3016839027404785, "learning_rate": 7.338212571106584e-07, "loss": 1.7319, "step": 42144 }, { "epoch": 3.302899686520376, "grad_norm": 2.2502753734588623, "learning_rate": 7.336604465736635e-07, "loss": 1.2607, "step": 42145 }, { "epoch": 3.3029780564263325, "grad_norm": 2.960078477859497, "learning_rate": 7.334996522636328e-07, "loss": 1.0364, "step": 42146 }, { "epoch": 3.3030564263322884, "grad_norm": 4.189159393310547, "learning_rate": 7.333388741811781e-07, "loss": 1.0561, "step": 42147 }, { "epoch": 3.3031347962382447, "grad_norm": 3.9124915599823, "learning_rate": 7.331781123269116e-07, "loss": 1.4179, "step": 42148 }, { "epoch": 3.3032131661442006, "grad_norm": 2.5388693809509277, "learning_rate": 7.330173667014423e-07, "loss": 1.3682, "step": 42149 }, { "epoch": 3.303291536050157, "grad_norm": 7.40139627456665, "learning_rate": 7.328566373053841e-07, "loss": 1.1054, "step": 42150 }, { "epoch": 3.303369905956113, "grad_norm": 3.8628151416778564, "learning_rate": 7.326959241393472e-07, "loss": 1.3926, "step": 42151 }, { "epoch": 3.303448275862069, "grad_norm": 3.052907705307007, "learning_rate": 7.325352272039449e-07, "loss": 1.5575, "step": 42152 }, { "epoch": 3.303526645768025, "grad_norm": 4.115146636962891, "learning_rate": 7.32374546499785e-07, "loss": 1.2886, "step": 42153 }, { "epoch": 3.3036050156739813, "grad_norm": 2.4435105323791504, "learning_rate": 7.322138820274794e-07, "loss": 1.232, "step": 42154 }, { "epoch": 3.3036833855799372, "grad_norm": 3.176421642303467, "learning_rate": 7.320532337876424e-07, "loss": 1.676, "step": 42155 }, { "epoch": 3.3037617554858936, "grad_norm": 3.960205554962158, "learning_rate": 7.31892601780882e-07, "loss": 1.7069, "step": 42156 }, { "epoch": 3.3038401253918495, "grad_norm": 4.816593170166016, "learning_rate": 7.317319860078092e-07, "loss": 0.932, "step": 42157 }, { "epoch": 3.303918495297806, "grad_norm": 3.7003908157348633, "learning_rate": 7.31571386469036e-07, "loss": 1.4284, "step": 42158 }, { "epoch": 3.3039968652037617, "grad_norm": 2.5884287357330322, "learning_rate": 7.314108031651729e-07, "loss": 1.3226, "step": 42159 }, { "epoch": 3.304075235109718, "grad_norm": 2.2794976234436035, "learning_rate": 7.312502360968305e-07, "loss": 1.3402, "step": 42160 }, { "epoch": 3.304153605015674, "grad_norm": 7.919301986694336, "learning_rate": 7.310896852646193e-07, "loss": 1.1118, "step": 42161 }, { "epoch": 3.3042319749216302, "grad_norm": 4.275275707244873, "learning_rate": 7.309291506691507e-07, "loss": 1.093, "step": 42162 }, { "epoch": 3.304310344827586, "grad_norm": 4.224796295166016, "learning_rate": 7.307686323110352e-07, "loss": 1.5578, "step": 42163 }, { "epoch": 3.3043887147335425, "grad_norm": 3.319300413131714, "learning_rate": 7.306081301908824e-07, "loss": 1.4421, "step": 42164 }, { "epoch": 3.3044670846394983, "grad_norm": 12.79989242553711, "learning_rate": 7.304476443093034e-07, "loss": 1.2115, "step": 42165 }, { "epoch": 3.3045454545454547, "grad_norm": 3.094158411026001, "learning_rate": 7.302871746669088e-07, "loss": 1.2558, "step": 42166 }, { "epoch": 3.3046238244514106, "grad_norm": 2.442901849746704, "learning_rate": 7.301267212643098e-07, "loss": 1.4495, "step": 42167 }, { "epoch": 3.304702194357367, "grad_norm": 3.360975980758667, "learning_rate": 7.299662841021132e-07, "loss": 1.2063, "step": 42168 }, { "epoch": 3.304780564263323, "grad_norm": 3.42008900642395, "learning_rate": 7.298058631809335e-07, "loss": 1.6404, "step": 42169 }, { "epoch": 3.304858934169279, "grad_norm": 3.3534021377563477, "learning_rate": 7.29645458501379e-07, "loss": 1.1969, "step": 42170 }, { "epoch": 3.304937304075235, "grad_norm": 4.658295154571533, "learning_rate": 7.294850700640593e-07, "loss": 1.1894, "step": 42171 }, { "epoch": 3.3050156739811913, "grad_norm": 6.158330917358398, "learning_rate": 7.293246978695851e-07, "loss": 1.7086, "step": 42172 }, { "epoch": 3.3050940438871472, "grad_norm": 4.507549285888672, "learning_rate": 7.291643419185645e-07, "loss": 1.655, "step": 42173 }, { "epoch": 3.3051724137931036, "grad_norm": 4.591429233551025, "learning_rate": 7.290040022116118e-07, "loss": 1.5337, "step": 42174 }, { "epoch": 3.3052507836990594, "grad_norm": 3.8551414012908936, "learning_rate": 7.28843678749333e-07, "loss": 2.0837, "step": 42175 }, { "epoch": 3.305329153605016, "grad_norm": 5.864713191986084, "learning_rate": 7.286833715323389e-07, "loss": 0.581, "step": 42176 }, { "epoch": 3.3054075235109717, "grad_norm": 5.931416988372803, "learning_rate": 7.285230805612392e-07, "loss": 1.3335, "step": 42177 }, { "epoch": 3.305485893416928, "grad_norm": 5.944540977478027, "learning_rate": 7.283628058366437e-07, "loss": 1.2405, "step": 42178 }, { "epoch": 3.305564263322884, "grad_norm": 4.368188381195068, "learning_rate": 7.282025473591625e-07, "loss": 1.6592, "step": 42179 }, { "epoch": 3.30564263322884, "grad_norm": 3.9673798084259033, "learning_rate": 7.280423051294044e-07, "loss": 1.0551, "step": 42180 }, { "epoch": 3.305721003134796, "grad_norm": 2.5826265811920166, "learning_rate": 7.27882079147979e-07, "loss": 1.0973, "step": 42181 }, { "epoch": 3.3057993730407524, "grad_norm": 3.0057172775268555, "learning_rate": 7.277218694154964e-07, "loss": 0.9613, "step": 42182 }, { "epoch": 3.3058777429467083, "grad_norm": 9.964561462402344, "learning_rate": 7.275616759325649e-07, "loss": 1.2427, "step": 42183 }, { "epoch": 3.3059561128526647, "grad_norm": 2.815274238586426, "learning_rate": 7.274014986997946e-07, "loss": 1.2878, "step": 42184 }, { "epoch": 3.3060344827586206, "grad_norm": 3.135850191116333, "learning_rate": 7.272413377177955e-07, "loss": 1.2062, "step": 42185 }, { "epoch": 3.306112852664577, "grad_norm": 2.929368734359741, "learning_rate": 7.270811929871735e-07, "loss": 1.096, "step": 42186 }, { "epoch": 3.3061912225705328, "grad_norm": 4.792143821716309, "learning_rate": 7.269210645085406e-07, "loss": 0.9312, "step": 42187 }, { "epoch": 3.306269592476489, "grad_norm": 3.2405028343200684, "learning_rate": 7.267609522825059e-07, "loss": 1.4636, "step": 42188 }, { "epoch": 3.306347962382445, "grad_norm": 2.851743698120117, "learning_rate": 7.266008563096782e-07, "loss": 1.3416, "step": 42189 }, { "epoch": 3.3064263322884013, "grad_norm": 4.270773887634277, "learning_rate": 7.26440776590665e-07, "loss": 1.303, "step": 42190 }, { "epoch": 3.306504702194357, "grad_norm": 7.201094150543213, "learning_rate": 7.262807131260752e-07, "loss": 1.7431, "step": 42191 }, { "epoch": 3.3065830721003135, "grad_norm": 4.818587779998779, "learning_rate": 7.261206659165193e-07, "loss": 0.9461, "step": 42192 }, { "epoch": 3.3066614420062694, "grad_norm": 4.772420883178711, "learning_rate": 7.259606349626064e-07, "loss": 0.9824, "step": 42193 }, { "epoch": 3.3067398119122258, "grad_norm": 4.094379425048828, "learning_rate": 7.258006202649426e-07, "loss": 1.8782, "step": 42194 }, { "epoch": 3.3068181818181817, "grad_norm": 3.326061725616455, "learning_rate": 7.256406218241379e-07, "loss": 1.5961, "step": 42195 }, { "epoch": 3.306896551724138, "grad_norm": 3.201282501220703, "learning_rate": 7.254806396407999e-07, "loss": 1.0001, "step": 42196 }, { "epoch": 3.306974921630094, "grad_norm": 3.9881207942962646, "learning_rate": 7.253206737155405e-07, "loss": 1.2027, "step": 42197 }, { "epoch": 3.30705329153605, "grad_norm": 4.805484771728516, "learning_rate": 7.251607240489639e-07, "loss": 1.1278, "step": 42198 }, { "epoch": 3.307131661442006, "grad_norm": 3.043255567550659, "learning_rate": 7.250007906416806e-07, "loss": 1.0334, "step": 42199 }, { "epoch": 3.3072100313479624, "grad_norm": 3.1022157669067383, "learning_rate": 7.248408734942985e-07, "loss": 0.8597, "step": 42200 }, { "epoch": 3.3072884012539183, "grad_norm": 4.372138500213623, "learning_rate": 7.246809726074256e-07, "loss": 1.0676, "step": 42201 }, { "epoch": 3.3073667711598747, "grad_norm": 2.9957435131073, "learning_rate": 7.245210879816706e-07, "loss": 1.3023, "step": 42202 }, { "epoch": 3.3074451410658305, "grad_norm": 4.803404808044434, "learning_rate": 7.243612196176414e-07, "loss": 1.2567, "step": 42203 }, { "epoch": 3.307523510971787, "grad_norm": 3.071532964706421, "learning_rate": 7.24201367515947e-07, "loss": 1.3667, "step": 42204 }, { "epoch": 3.3076018808777428, "grad_norm": 2.5749218463897705, "learning_rate": 7.24041531677192e-07, "loss": 1.2533, "step": 42205 }, { "epoch": 3.307680250783699, "grad_norm": 2.195704460144043, "learning_rate": 7.238817121019881e-07, "loss": 1.1947, "step": 42206 }, { "epoch": 3.307758620689655, "grad_norm": 5.199868679046631, "learning_rate": 7.237219087909419e-07, "loss": 1.8085, "step": 42207 }, { "epoch": 3.3078369905956113, "grad_norm": 47.23325729370117, "learning_rate": 7.23562121744662e-07, "loss": 1.6132, "step": 42208 }, { "epoch": 3.3079153605015676, "grad_norm": 4.478671073913574, "learning_rate": 7.234023509637545e-07, "loss": 1.3065, "step": 42209 }, { "epoch": 3.3079937304075235, "grad_norm": 4.5665388107299805, "learning_rate": 7.232425964488265e-07, "loss": 1.8256, "step": 42210 }, { "epoch": 3.3080721003134794, "grad_norm": 3.780015230178833, "learning_rate": 7.23082858200489e-07, "loss": 1.5084, "step": 42211 }, { "epoch": 3.3081504702194358, "grad_norm": 8.921663284301758, "learning_rate": 7.229231362193468e-07, "loss": 1.3063, "step": 42212 }, { "epoch": 3.308228840125392, "grad_norm": 3.458040475845337, "learning_rate": 7.227634305060083e-07, "loss": 1.3277, "step": 42213 }, { "epoch": 3.308307210031348, "grad_norm": 4.429052352905273, "learning_rate": 7.226037410610804e-07, "loss": 1.1135, "step": 42214 }, { "epoch": 3.308385579937304, "grad_norm": 3.777477741241455, "learning_rate": 7.224440678851707e-07, "loss": 1.2422, "step": 42215 }, { "epoch": 3.30846394984326, "grad_norm": 3.3762199878692627, "learning_rate": 7.222844109788868e-07, "loss": 1.3565, "step": 42216 }, { "epoch": 3.3085423197492165, "grad_norm": 5.877608776092529, "learning_rate": 7.221247703428364e-07, "loss": 1.3641, "step": 42217 }, { "epoch": 3.3086206896551724, "grad_norm": 7.555000305175781, "learning_rate": 7.219651459776255e-07, "loss": 1.3612, "step": 42218 }, { "epoch": 3.3086990595611283, "grad_norm": 6.809240818023682, "learning_rate": 7.218055378838617e-07, "loss": 1.2444, "step": 42219 }, { "epoch": 3.3087774294670846, "grad_norm": 4.823801040649414, "learning_rate": 7.216459460621528e-07, "loss": 1.5714, "step": 42220 }, { "epoch": 3.308855799373041, "grad_norm": 2.902294874191284, "learning_rate": 7.214863705131048e-07, "loss": 1.0094, "step": 42221 }, { "epoch": 3.308934169278997, "grad_norm": 5.533011436462402, "learning_rate": 7.21326811237325e-07, "loss": 1.2259, "step": 42222 }, { "epoch": 3.3090125391849528, "grad_norm": 2.750629425048828, "learning_rate": 7.21167268235422e-07, "loss": 1.2562, "step": 42223 }, { "epoch": 3.309090909090909, "grad_norm": 9.852866172790527, "learning_rate": 7.21007741507998e-07, "loss": 1.0933, "step": 42224 }, { "epoch": 3.3091692789968654, "grad_norm": 4.292074680328369, "learning_rate": 7.208482310556647e-07, "loss": 0.9206, "step": 42225 }, { "epoch": 3.3092476489028213, "grad_norm": 7.154077529907227, "learning_rate": 7.206887368790272e-07, "loss": 0.9455, "step": 42226 }, { "epoch": 3.309326018808777, "grad_norm": 3.32401442527771, "learning_rate": 7.20529258978691e-07, "loss": 0.7281, "step": 42227 }, { "epoch": 3.3094043887147335, "grad_norm": 5.723996162414551, "learning_rate": 7.203697973552631e-07, "loss": 1.7824, "step": 42228 }, { "epoch": 3.30948275862069, "grad_norm": 4.872201919555664, "learning_rate": 7.202103520093495e-07, "loss": 1.8901, "step": 42229 }, { "epoch": 3.3095611285266457, "grad_norm": 3.3168599605560303, "learning_rate": 7.200509229415598e-07, "loss": 1.3802, "step": 42230 }, { "epoch": 3.3096394984326016, "grad_norm": 3.0171186923980713, "learning_rate": 7.198915101524967e-07, "loss": 1.3972, "step": 42231 }, { "epoch": 3.309717868338558, "grad_norm": 2.2099859714508057, "learning_rate": 7.197321136427682e-07, "loss": 1.3265, "step": 42232 }, { "epoch": 3.3097962382445143, "grad_norm": 2.6655960083007812, "learning_rate": 7.195727334129799e-07, "loss": 1.0106, "step": 42233 }, { "epoch": 3.30987460815047, "grad_norm": 4.438262939453125, "learning_rate": 7.194133694637384e-07, "loss": 1.2159, "step": 42234 }, { "epoch": 3.3099529780564265, "grad_norm": 5.703305244445801, "learning_rate": 7.1925402179565e-07, "loss": 1.1297, "step": 42235 }, { "epoch": 3.3100313479623824, "grad_norm": 3.158798933029175, "learning_rate": 7.190946904093204e-07, "loss": 1.4455, "step": 42236 }, { "epoch": 3.3101097178683387, "grad_norm": 3.735361337661743, "learning_rate": 7.189353753053563e-07, "loss": 1.0254, "step": 42237 }, { "epoch": 3.3101880877742946, "grad_norm": 3.391838788986206, "learning_rate": 7.187760764843627e-07, "loss": 1.6388, "step": 42238 }, { "epoch": 3.310266457680251, "grad_norm": 3.033392906188965, "learning_rate": 7.186167939469457e-07, "loss": 1.6566, "step": 42239 }, { "epoch": 3.310344827586207, "grad_norm": 2.6365227699279785, "learning_rate": 7.184575276937117e-07, "loss": 0.98, "step": 42240 }, { "epoch": 3.310423197492163, "grad_norm": 2.9979212284088135, "learning_rate": 7.182982777252673e-07, "loss": 1.2763, "step": 42241 }, { "epoch": 3.310501567398119, "grad_norm": 6.277098655700684, "learning_rate": 7.181390440422153e-07, "loss": 1.2301, "step": 42242 }, { "epoch": 3.3105799373040754, "grad_norm": 2.9754374027252197, "learning_rate": 7.179798266451627e-07, "loss": 1.1747, "step": 42243 }, { "epoch": 3.3106583072100313, "grad_norm": 3.0700995922088623, "learning_rate": 7.178206255347159e-07, "loss": 1.6353, "step": 42244 }, { "epoch": 3.3107366771159876, "grad_norm": 2.56260085105896, "learning_rate": 7.176614407114818e-07, "loss": 1.3007, "step": 42245 }, { "epoch": 3.3108150470219435, "grad_norm": 3.8961057662963867, "learning_rate": 7.175022721760627e-07, "loss": 1.3272, "step": 42246 }, { "epoch": 3.3108934169279, "grad_norm": 2.2097949981689453, "learning_rate": 7.173431199290648e-07, "loss": 0.9542, "step": 42247 }, { "epoch": 3.3109717868338557, "grad_norm": 4.236388206481934, "learning_rate": 7.171839839710936e-07, "loss": 1.3967, "step": 42248 }, { "epoch": 3.311050156739812, "grad_norm": 2.5578207969665527, "learning_rate": 7.170248643027566e-07, "loss": 0.87, "step": 42249 }, { "epoch": 3.311128526645768, "grad_norm": 4.216842174530029, "learning_rate": 7.16865760924656e-07, "loss": 1.2802, "step": 42250 }, { "epoch": 3.3112068965517243, "grad_norm": 2.257124423980713, "learning_rate": 7.167066738373979e-07, "loss": 1.3089, "step": 42251 }, { "epoch": 3.31128526645768, "grad_norm": 2.5579910278320312, "learning_rate": 7.16547603041588e-07, "loss": 1.317, "step": 42252 }, { "epoch": 3.3113636363636365, "grad_norm": 3.2744579315185547, "learning_rate": 7.163885485378308e-07, "loss": 1.3297, "step": 42253 }, { "epoch": 3.3114420062695924, "grad_norm": 3.4958338737487793, "learning_rate": 7.162295103267313e-07, "loss": 1.0781, "step": 42254 }, { "epoch": 3.3115203761755487, "grad_norm": 2.4785854816436768, "learning_rate": 7.160704884088943e-07, "loss": 1.2956, "step": 42255 }, { "epoch": 3.3115987460815046, "grad_norm": 2.925959587097168, "learning_rate": 7.159114827849262e-07, "loss": 1.3325, "step": 42256 }, { "epoch": 3.311677115987461, "grad_norm": 4.200613021850586, "learning_rate": 7.157524934554277e-07, "loss": 1.7589, "step": 42257 }, { "epoch": 3.311755485893417, "grad_norm": 5.1030964851379395, "learning_rate": 7.155935204210074e-07, "loss": 1.5118, "step": 42258 }, { "epoch": 3.311833855799373, "grad_norm": 4.866025447845459, "learning_rate": 7.15434563682269e-07, "loss": 1.5221, "step": 42259 }, { "epoch": 3.311912225705329, "grad_norm": 3.1373302936553955, "learning_rate": 7.152756232398178e-07, "loss": 1.6006, "step": 42260 }, { "epoch": 3.3119905956112854, "grad_norm": 3.481658935546875, "learning_rate": 7.151166990942565e-07, "loss": 1.2318, "step": 42261 }, { "epoch": 3.3120689655172413, "grad_norm": 5.209012031555176, "learning_rate": 7.149577912461891e-07, "loss": 1.2744, "step": 42262 }, { "epoch": 3.3121473354231976, "grad_norm": 4.013604164123535, "learning_rate": 7.147988996962224e-07, "loss": 1.4016, "step": 42263 }, { "epoch": 3.3122257053291535, "grad_norm": 3.4177727699279785, "learning_rate": 7.146400244449608e-07, "loss": 1.5443, "step": 42264 }, { "epoch": 3.31230407523511, "grad_norm": 5.978862285614014, "learning_rate": 7.144811654930067e-07, "loss": 1.4419, "step": 42265 }, { "epoch": 3.3123824451410657, "grad_norm": 7.781739711761475, "learning_rate": 7.143223228409646e-07, "loss": 1.3442, "step": 42266 }, { "epoch": 3.312460815047022, "grad_norm": 3.7400522232055664, "learning_rate": 7.141634964894389e-07, "loss": 1.1628, "step": 42267 }, { "epoch": 3.312539184952978, "grad_norm": 3.4289584159851074, "learning_rate": 7.140046864390343e-07, "loss": 1.4385, "step": 42268 }, { "epoch": 3.3126175548589343, "grad_norm": 5.962927341461182, "learning_rate": 7.138458926903547e-07, "loss": 1.2543, "step": 42269 }, { "epoch": 3.31269592476489, "grad_norm": 3.6666109561920166, "learning_rate": 7.136871152440034e-07, "loss": 0.9909, "step": 42270 }, { "epoch": 3.3127742946708465, "grad_norm": 5.7437357902526855, "learning_rate": 7.135283541005849e-07, "loss": 1.53, "step": 42271 }, { "epoch": 3.3128526645768024, "grad_norm": 3.652418375015259, "learning_rate": 7.133696092607029e-07, "loss": 0.9884, "step": 42272 }, { "epoch": 3.3129310344827587, "grad_norm": 4.006312847137451, "learning_rate": 7.132108807249611e-07, "loss": 1.2501, "step": 42273 }, { "epoch": 3.3130094043887146, "grad_norm": 4.101929664611816, "learning_rate": 7.130521684939629e-07, "loss": 1.8598, "step": 42274 }, { "epoch": 3.313087774294671, "grad_norm": 5.650014877319336, "learning_rate": 7.128934725683129e-07, "loss": 1.118, "step": 42275 }, { "epoch": 3.313166144200627, "grad_norm": 3.497722625732422, "learning_rate": 7.127347929486139e-07, "loss": 1.3483, "step": 42276 }, { "epoch": 3.313244514106583, "grad_norm": 3.452463388442993, "learning_rate": 7.125761296354699e-07, "loss": 1.0446, "step": 42277 }, { "epoch": 3.313322884012539, "grad_norm": 5.392167568206787, "learning_rate": 7.124174826294838e-07, "loss": 1.4863, "step": 42278 }, { "epoch": 3.3134012539184954, "grad_norm": 2.319267511367798, "learning_rate": 7.122588519312607e-07, "loss": 1.1924, "step": 42279 }, { "epoch": 3.3134796238244513, "grad_norm": 2.254072427749634, "learning_rate": 7.121002375413999e-07, "loss": 1.1457, "step": 42280 }, { "epoch": 3.3135579937304076, "grad_norm": 2.3249075412750244, "learning_rate": 7.119416394605089e-07, "loss": 1.3583, "step": 42281 }, { "epoch": 3.3136363636363635, "grad_norm": 2.5664114952087402, "learning_rate": 7.117830576891904e-07, "loss": 1.2043, "step": 42282 }, { "epoch": 3.31371473354232, "grad_norm": 3.1989755630493164, "learning_rate": 7.116244922280457e-07, "loss": 1.441, "step": 42283 }, { "epoch": 3.3137931034482757, "grad_norm": 2.737440586090088, "learning_rate": 7.114659430776787e-07, "loss": 1.2616, "step": 42284 }, { "epoch": 3.313871473354232, "grad_norm": 4.255044460296631, "learning_rate": 7.113074102386925e-07, "loss": 1.2642, "step": 42285 }, { "epoch": 3.313949843260188, "grad_norm": 3.8903706073760986, "learning_rate": 7.111488937116901e-07, "loss": 1.2975, "step": 42286 }, { "epoch": 3.3140282131661443, "grad_norm": 9.00082778930664, "learning_rate": 7.109903934972746e-07, "loss": 1.552, "step": 42287 }, { "epoch": 3.3141065830721, "grad_norm": 2.446946620941162, "learning_rate": 7.10831909596048e-07, "loss": 1.5959, "step": 42288 }, { "epoch": 3.3141849529780565, "grad_norm": 5.64229679107666, "learning_rate": 7.106734420086148e-07, "loss": 1.5801, "step": 42289 }, { "epoch": 3.3142633228840124, "grad_norm": 5.225001335144043, "learning_rate": 7.105149907355758e-07, "loss": 1.5252, "step": 42290 }, { "epoch": 3.3143416927899687, "grad_norm": 6.722758769989014, "learning_rate": 7.103565557775349e-07, "loss": 1.1745, "step": 42291 }, { "epoch": 3.3144200626959246, "grad_norm": 4.777093410491943, "learning_rate": 7.101981371350947e-07, "loss": 1.3034, "step": 42292 }, { "epoch": 3.314498432601881, "grad_norm": 4.465956687927246, "learning_rate": 7.100397348088567e-07, "loss": 1.4926, "step": 42293 }, { "epoch": 3.314576802507837, "grad_norm": 2.2369275093078613, "learning_rate": 7.098813487994244e-07, "loss": 1.5192, "step": 42294 }, { "epoch": 3.314655172413793, "grad_norm": 4.159911155700684, "learning_rate": 7.097229791074001e-07, "loss": 1.188, "step": 42295 }, { "epoch": 3.314733542319749, "grad_norm": 4.095093250274658, "learning_rate": 7.095646257333854e-07, "loss": 1.4633, "step": 42296 }, { "epoch": 3.3148119122257054, "grad_norm": 2.6144626140594482, "learning_rate": 7.094062886779852e-07, "loss": 1.615, "step": 42297 }, { "epoch": 3.3148902821316613, "grad_norm": 2.995384693145752, "learning_rate": 7.09247967941798e-07, "loss": 0.8444, "step": 42298 }, { "epoch": 3.3149686520376176, "grad_norm": 6.413072109222412, "learning_rate": 7.090896635254263e-07, "loss": 0.9581, "step": 42299 }, { "epoch": 3.3150470219435735, "grad_norm": 6.906055927276611, "learning_rate": 7.089313754294752e-07, "loss": 1.7382, "step": 42300 }, { "epoch": 3.31512539184953, "grad_norm": 2.212005376815796, "learning_rate": 7.087731036545459e-07, "loss": 1.0322, "step": 42301 }, { "epoch": 3.3152037617554857, "grad_norm": 7.3841447830200195, "learning_rate": 7.086148482012384e-07, "loss": 1.4681, "step": 42302 }, { "epoch": 3.315282131661442, "grad_norm": 4.4336748123168945, "learning_rate": 7.084566090701562e-07, "loss": 1.6573, "step": 42303 }, { "epoch": 3.3153605015673984, "grad_norm": 2.201761245727539, "learning_rate": 7.082983862618991e-07, "loss": 1.1832, "step": 42304 }, { "epoch": 3.3154388714733543, "grad_norm": 2.4308619499206543, "learning_rate": 7.081401797770737e-07, "loss": 1.5485, "step": 42305 }, { "epoch": 3.31551724137931, "grad_norm": 2.8887827396392822, "learning_rate": 7.079819896162765e-07, "loss": 1.3462, "step": 42306 }, { "epoch": 3.3155956112852665, "grad_norm": 4.085719108581543, "learning_rate": 7.078238157801121e-07, "loss": 1.4926, "step": 42307 }, { "epoch": 3.315673981191223, "grad_norm": 4.923254489898682, "learning_rate": 7.07665658269181e-07, "loss": 2.0591, "step": 42308 }, { "epoch": 3.3157523510971787, "grad_norm": 5.000910758972168, "learning_rate": 7.07507517084085e-07, "loss": 1.5101, "step": 42309 }, { "epoch": 3.3158307210031346, "grad_norm": 4.5213398933410645, "learning_rate": 7.073493922254254e-07, "loss": 1.2992, "step": 42310 }, { "epoch": 3.315909090909091, "grad_norm": 4.0935750007629395, "learning_rate": 7.071912836938044e-07, "loss": 1.4458, "step": 42311 }, { "epoch": 3.3159874608150472, "grad_norm": 3.145702600479126, "learning_rate": 7.070331914898238e-07, "loss": 1.033, "step": 42312 }, { "epoch": 3.316065830721003, "grad_norm": 6.198614597320557, "learning_rate": 7.068751156140813e-07, "loss": 1.0105, "step": 42313 }, { "epoch": 3.316144200626959, "grad_norm": 4.197291374206543, "learning_rate": 7.067170560671821e-07, "loss": 1.4479, "step": 42314 }, { "epoch": 3.3162225705329154, "grad_norm": 6.663177490234375, "learning_rate": 7.065590128497263e-07, "loss": 1.4388, "step": 42315 }, { "epoch": 3.3163009404388717, "grad_norm": 9.570926666259766, "learning_rate": 7.064009859623155e-07, "loss": 0.8125, "step": 42316 }, { "epoch": 3.3163793103448276, "grad_norm": 4.098531246185303, "learning_rate": 7.062429754055495e-07, "loss": 1.4937, "step": 42317 }, { "epoch": 3.3164576802507835, "grad_norm": 3.468001365661621, "learning_rate": 7.06084981180028e-07, "loss": 1.2814, "step": 42318 }, { "epoch": 3.31653605015674, "grad_norm": 2.302595853805542, "learning_rate": 7.059270032863553e-07, "loss": 1.3017, "step": 42319 }, { "epoch": 3.316614420062696, "grad_norm": 2.159323215484619, "learning_rate": 7.057690417251317e-07, "loss": 1.4483, "step": 42320 }, { "epoch": 3.316692789968652, "grad_norm": 3.0760998725891113, "learning_rate": 7.056110964969559e-07, "loss": 1.31, "step": 42321 }, { "epoch": 3.316771159874608, "grad_norm": 5.128720283508301, "learning_rate": 7.054531676024295e-07, "loss": 1.2017, "step": 42322 }, { "epoch": 3.3168495297805642, "grad_norm": 6.87245512008667, "learning_rate": 7.052952550421537e-07, "loss": 1.4221, "step": 42323 }, { "epoch": 3.3169278996865206, "grad_norm": 2.5537514686584473, "learning_rate": 7.051373588167293e-07, "loss": 0.8837, "step": 42324 }, { "epoch": 3.3170062695924765, "grad_norm": 7.080942153930664, "learning_rate": 7.049794789267556e-07, "loss": 1.3984, "step": 42325 }, { "epoch": 3.3170846394984324, "grad_norm": 4.823235511779785, "learning_rate": 7.048216153728343e-07, "loss": 1.0336, "step": 42326 }, { "epoch": 3.3171630094043887, "grad_norm": 2.637148857116699, "learning_rate": 7.046637681555652e-07, "loss": 1.2278, "step": 42327 }, { "epoch": 3.317241379310345, "grad_norm": 4.555231094360352, "learning_rate": 7.045059372755492e-07, "loss": 1.1467, "step": 42328 }, { "epoch": 3.317319749216301, "grad_norm": 2.6839535236358643, "learning_rate": 7.04348122733386e-07, "loss": 1.5216, "step": 42329 }, { "epoch": 3.3173981191222572, "grad_norm": 2.984689474105835, "learning_rate": 7.041903245296766e-07, "loss": 1.1428, "step": 42330 }, { "epoch": 3.317476489028213, "grad_norm": 4.301746845245361, "learning_rate": 7.040325426650213e-07, "loss": 1.6863, "step": 42331 }, { "epoch": 3.3175548589341695, "grad_norm": 3.3319828510284424, "learning_rate": 7.038747771400179e-07, "loss": 1.0517, "step": 42332 }, { "epoch": 3.3176332288401253, "grad_norm": 7.724211692810059, "learning_rate": 7.037170279552691e-07, "loss": 0.9962, "step": 42333 }, { "epoch": 3.3177115987460817, "grad_norm": 5.200742244720459, "learning_rate": 7.03559295111374e-07, "loss": 0.672, "step": 42334 }, { "epoch": 3.3177899686520376, "grad_norm": 4.00314998626709, "learning_rate": 7.034015786089338e-07, "loss": 1.4703, "step": 42335 }, { "epoch": 3.317868338557994, "grad_norm": 3.838386297225952, "learning_rate": 7.032438784485457e-07, "loss": 1.4519, "step": 42336 }, { "epoch": 3.31794670846395, "grad_norm": 3.136922597885132, "learning_rate": 7.030861946308099e-07, "loss": 1.3065, "step": 42337 }, { "epoch": 3.318025078369906, "grad_norm": 2.735610008239746, "learning_rate": 7.029285271563297e-07, "loss": 1.2597, "step": 42338 }, { "epoch": 3.318103448275862, "grad_norm": 2.471198320388794, "learning_rate": 7.027708760257013e-07, "loss": 1.4194, "step": 42339 }, { "epoch": 3.3181818181818183, "grad_norm": 4.414283752441406, "learning_rate": 7.026132412395247e-07, "loss": 1.0958, "step": 42340 }, { "epoch": 3.3182601880877742, "grad_norm": 2.9850354194641113, "learning_rate": 7.024556227984003e-07, "loss": 1.2528, "step": 42341 }, { "epoch": 3.3183385579937306, "grad_norm": 2.357623338699341, "learning_rate": 7.022980207029273e-07, "loss": 1.2621, "step": 42342 }, { "epoch": 3.3184169278996865, "grad_norm": 3.8660473823547363, "learning_rate": 7.021404349537054e-07, "loss": 1.182, "step": 42343 }, { "epoch": 3.318495297805643, "grad_norm": 3.959902763366699, "learning_rate": 7.019828655513339e-07, "loss": 0.9478, "step": 42344 }, { "epoch": 3.3185736677115987, "grad_norm": 2.6120409965515137, "learning_rate": 7.018253124964114e-07, "loss": 1.2412, "step": 42345 }, { "epoch": 3.318652037617555, "grad_norm": 2.092665433883667, "learning_rate": 7.016677757895379e-07, "loss": 1.1796, "step": 42346 }, { "epoch": 3.318730407523511, "grad_norm": 2.874265193939209, "learning_rate": 7.01510255431313e-07, "loss": 1.0512, "step": 42347 }, { "epoch": 3.3188087774294672, "grad_norm": 9.236608505249023, "learning_rate": 7.013527514223345e-07, "loss": 1.3198, "step": 42348 }, { "epoch": 3.318887147335423, "grad_norm": 4.181456565856934, "learning_rate": 7.011952637632024e-07, "loss": 1.2444, "step": 42349 }, { "epoch": 3.3189655172413794, "grad_norm": 2.8986034393310547, "learning_rate": 7.010377924545169e-07, "loss": 1.1475, "step": 42350 }, { "epoch": 3.3190438871473353, "grad_norm": 5.0807952880859375, "learning_rate": 7.008803374968731e-07, "loss": 1.3269, "step": 42351 }, { "epoch": 3.3191222570532917, "grad_norm": 3.0496909618377686, "learning_rate": 7.007228988908738e-07, "loss": 1.3419, "step": 42352 }, { "epoch": 3.3192006269592476, "grad_norm": 3.681224822998047, "learning_rate": 7.005654766371173e-07, "loss": 1.0762, "step": 42353 }, { "epoch": 3.319278996865204, "grad_norm": 6.471193313598633, "learning_rate": 7.004080707361999e-07, "loss": 1.4393, "step": 42354 }, { "epoch": 3.31935736677116, "grad_norm": 3.2945353984832764, "learning_rate": 7.002506811887222e-07, "loss": 1.5233, "step": 42355 }, { "epoch": 3.319435736677116, "grad_norm": 2.6576449871063232, "learning_rate": 7.00093307995281e-07, "loss": 1.164, "step": 42356 }, { "epoch": 3.319514106583072, "grad_norm": 3.529977798461914, "learning_rate": 6.999359511564785e-07, "loss": 1.2766, "step": 42357 }, { "epoch": 3.3195924764890283, "grad_norm": 3.5580880641937256, "learning_rate": 6.997786106729104e-07, "loss": 1.329, "step": 42358 }, { "epoch": 3.319670846394984, "grad_norm": 2.2746822834014893, "learning_rate": 6.996212865451751e-07, "loss": 1.3224, "step": 42359 }, { "epoch": 3.3197492163009406, "grad_norm": 2.121677875518799, "learning_rate": 6.994639787738722e-07, "loss": 1.3376, "step": 42360 }, { "epoch": 3.3198275862068964, "grad_norm": 6.237607479095459, "learning_rate": 6.993066873595989e-07, "loss": 1.2539, "step": 42361 }, { "epoch": 3.3199059561128528, "grad_norm": 4.439428806304932, "learning_rate": 6.99149412302954e-07, "loss": 1.0126, "step": 42362 }, { "epoch": 3.3199843260188087, "grad_norm": 3.161076784133911, "learning_rate": 6.989921536045363e-07, "loss": 1.0876, "step": 42363 }, { "epoch": 3.320062695924765, "grad_norm": 2.753335952758789, "learning_rate": 6.988349112649429e-07, "loss": 1.1331, "step": 42364 }, { "epoch": 3.320141065830721, "grad_norm": 2.794053077697754, "learning_rate": 6.986776852847721e-07, "loss": 1.4976, "step": 42365 }, { "epoch": 3.320219435736677, "grad_norm": 3.344923973083496, "learning_rate": 6.985204756646224e-07, "loss": 1.1899, "step": 42366 }, { "epoch": 3.320297805642633, "grad_norm": 2.8498547077178955, "learning_rate": 6.983632824050912e-07, "loss": 1.6785, "step": 42367 }, { "epoch": 3.3203761755485894, "grad_norm": 5.335378170013428, "learning_rate": 6.98206105506778e-07, "loss": 0.8236, "step": 42368 }, { "epoch": 3.3204545454545453, "grad_norm": 8.110081672668457, "learning_rate": 6.980489449702776e-07, "loss": 1.1922, "step": 42369 }, { "epoch": 3.3205329153605017, "grad_norm": 3.6790692806243896, "learning_rate": 6.978918007961888e-07, "loss": 1.399, "step": 42370 }, { "epoch": 3.3206112852664575, "grad_norm": 2.672616958618164, "learning_rate": 6.977346729851109e-07, "loss": 1.3674, "step": 42371 }, { "epoch": 3.320689655172414, "grad_norm": 3.421186923980713, "learning_rate": 6.975775615376413e-07, "loss": 1.4439, "step": 42372 }, { "epoch": 3.3207680250783698, "grad_norm": 5.645891189575195, "learning_rate": 6.97420466454376e-07, "loss": 0.9399, "step": 42373 }, { "epoch": 3.320846394984326, "grad_norm": 3.735666036605835, "learning_rate": 6.972633877359131e-07, "loss": 1.207, "step": 42374 }, { "epoch": 3.320924764890282, "grad_norm": 2.875532627105713, "learning_rate": 6.971063253828492e-07, "loss": 1.3731, "step": 42375 }, { "epoch": 3.3210031347962383, "grad_norm": 3.4950063228607178, "learning_rate": 6.969492793957849e-07, "loss": 1.4959, "step": 42376 }, { "epoch": 3.321081504702194, "grad_norm": 3.7095413208007812, "learning_rate": 6.967922497753143e-07, "loss": 1.0987, "step": 42377 }, { "epoch": 3.3211598746081505, "grad_norm": 4.38762092590332, "learning_rate": 6.966352365220363e-07, "loss": 1.3587, "step": 42378 }, { "epoch": 3.3212382445141064, "grad_norm": 3.6508758068084717, "learning_rate": 6.964782396365466e-07, "loss": 1.5235, "step": 42379 }, { "epoch": 3.3213166144200628, "grad_norm": 7.353717803955078, "learning_rate": 6.963212591194441e-07, "loss": 1.2772, "step": 42380 }, { "epoch": 3.3213949843260187, "grad_norm": 3.829807758331299, "learning_rate": 6.961642949713243e-07, "loss": 1.1981, "step": 42381 }, { "epoch": 3.321473354231975, "grad_norm": 3.6442747116088867, "learning_rate": 6.960073471927853e-07, "loss": 1.3541, "step": 42382 }, { "epoch": 3.321551724137931, "grad_norm": 6.727984428405762, "learning_rate": 6.958504157844237e-07, "loss": 0.9455, "step": 42383 }, { "epoch": 3.321630094043887, "grad_norm": 2.4783782958984375, "learning_rate": 6.956935007468357e-07, "loss": 1.0933, "step": 42384 }, { "epoch": 3.321708463949843, "grad_norm": 2.1124398708343506, "learning_rate": 6.955366020806198e-07, "loss": 1.268, "step": 42385 }, { "epoch": 3.3217868338557994, "grad_norm": 5.617339134216309, "learning_rate": 6.953797197863709e-07, "loss": 1.4007, "step": 42386 }, { "epoch": 3.3218652037617553, "grad_norm": 3.7735910415649414, "learning_rate": 6.952228538646882e-07, "loss": 1.3687, "step": 42387 }, { "epoch": 3.3219435736677116, "grad_norm": 4.535029888153076, "learning_rate": 6.950660043161644e-07, "loss": 1.3439, "step": 42388 }, { "epoch": 3.3220219435736675, "grad_norm": 5.205885410308838, "learning_rate": 6.949091711413992e-07, "loss": 1.3142, "step": 42389 }, { "epoch": 3.322100313479624, "grad_norm": 5.220641613006592, "learning_rate": 6.947523543409884e-07, "loss": 1.4144, "step": 42390 }, { "epoch": 3.3221786833855798, "grad_norm": 3.9745020866394043, "learning_rate": 6.945955539155297e-07, "loss": 1.5227, "step": 42391 }, { "epoch": 3.322257053291536, "grad_norm": 2.7322804927825928, "learning_rate": 6.944387698656168e-07, "loss": 1.3743, "step": 42392 }, { "epoch": 3.322335423197492, "grad_norm": 3.966843366622925, "learning_rate": 6.942820021918462e-07, "loss": 1.2711, "step": 42393 }, { "epoch": 3.3224137931034483, "grad_norm": 3.5564560890197754, "learning_rate": 6.941252508948176e-07, "loss": 1.1392, "step": 42394 }, { "epoch": 3.322492163009404, "grad_norm": 6.4499735832214355, "learning_rate": 6.939685159751236e-07, "loss": 1.4523, "step": 42395 }, { "epoch": 3.3225705329153605, "grad_norm": 10.307731628417969, "learning_rate": 6.938117974333619e-07, "loss": 1.7323, "step": 42396 }, { "epoch": 3.3226489028213164, "grad_norm": 2.4517667293548584, "learning_rate": 6.936550952701282e-07, "loss": 1.1842, "step": 42397 }, { "epoch": 3.3227272727272728, "grad_norm": 2.7002835273742676, "learning_rate": 6.934984094860187e-07, "loss": 1.4424, "step": 42398 }, { "epoch": 3.322805642633229, "grad_norm": 2.650657892227173, "learning_rate": 6.933417400816295e-07, "loss": 1.1577, "step": 42399 }, { "epoch": 3.322884012539185, "grad_norm": 6.967220783233643, "learning_rate": 6.931850870575563e-07, "loss": 0.8944, "step": 42400 }, { "epoch": 3.322962382445141, "grad_norm": 3.3518831729888916, "learning_rate": 6.930284504143947e-07, "loss": 1.3933, "step": 42401 }, { "epoch": 3.323040752351097, "grad_norm": 2.6195077896118164, "learning_rate": 6.928718301527404e-07, "loss": 1.6421, "step": 42402 }, { "epoch": 3.3231191222570535, "grad_norm": 4.055180549621582, "learning_rate": 6.927152262731895e-07, "loss": 1.0585, "step": 42403 }, { "epoch": 3.3231974921630094, "grad_norm": 3.17673659324646, "learning_rate": 6.925586387763378e-07, "loss": 1.3226, "step": 42404 }, { "epoch": 3.3232758620689653, "grad_norm": 2.7875943183898926, "learning_rate": 6.924020676627802e-07, "loss": 1.4115, "step": 42405 }, { "epoch": 3.3233542319749216, "grad_norm": 2.7763619422912598, "learning_rate": 6.92245512933114e-07, "loss": 1.2876, "step": 42406 }, { "epoch": 3.323432601880878, "grad_norm": 2.931875705718994, "learning_rate": 6.920889745879311e-07, "loss": 1.3191, "step": 42407 }, { "epoch": 3.323510971786834, "grad_norm": 3.340212345123291, "learning_rate": 6.919324526278303e-07, "loss": 1.1269, "step": 42408 }, { "epoch": 3.3235893416927897, "grad_norm": 3.0364785194396973, "learning_rate": 6.917759470534064e-07, "loss": 1.3099, "step": 42409 }, { "epoch": 3.323667711598746, "grad_norm": 2.7349681854248047, "learning_rate": 6.91619457865253e-07, "loss": 1.4276, "step": 42410 }, { "epoch": 3.3237460815047024, "grad_norm": 2.83134126663208, "learning_rate": 6.914629850639665e-07, "loss": 1.5026, "step": 42411 }, { "epoch": 3.3238244514106583, "grad_norm": 4.859096050262451, "learning_rate": 6.913065286501403e-07, "loss": 1.3756, "step": 42412 }, { "epoch": 3.323902821316614, "grad_norm": 3.6378912925720215, "learning_rate": 6.911500886243732e-07, "loss": 1.4104, "step": 42413 }, { "epoch": 3.3239811912225705, "grad_norm": 2.5832130908966064, "learning_rate": 6.909936649872573e-07, "loss": 1.476, "step": 42414 }, { "epoch": 3.324059561128527, "grad_norm": 4.22271203994751, "learning_rate": 6.908372577393879e-07, "loss": 1.2352, "step": 42415 }, { "epoch": 3.3241379310344827, "grad_norm": 3.399127960205078, "learning_rate": 6.906808668813603e-07, "loss": 1.1275, "step": 42416 }, { "epoch": 3.3242163009404386, "grad_norm": 4.0477166175842285, "learning_rate": 6.905244924137694e-07, "loss": 1.7018, "step": 42417 }, { "epoch": 3.324294670846395, "grad_norm": 3.8651654720306396, "learning_rate": 6.903681343372099e-07, "loss": 1.5014, "step": 42418 }, { "epoch": 3.3243730407523513, "grad_norm": 4.801342010498047, "learning_rate": 6.902117926522761e-07, "loss": 1.1756, "step": 42419 }, { "epoch": 3.324451410658307, "grad_norm": 3.935887575149536, "learning_rate": 6.900554673595644e-07, "loss": 1.1404, "step": 42420 }, { "epoch": 3.3245297805642635, "grad_norm": 5.774978160858154, "learning_rate": 6.898991584596654e-07, "loss": 1.1187, "step": 42421 }, { "epoch": 3.3246081504702194, "grad_norm": 6.583437442779541, "learning_rate": 6.897428659531779e-07, "loss": 0.5798, "step": 42422 }, { "epoch": 3.3246865203761757, "grad_norm": 2.4491934776306152, "learning_rate": 6.895865898406945e-07, "loss": 1.3261, "step": 42423 }, { "epoch": 3.3247648902821316, "grad_norm": 2.8772530555725098, "learning_rate": 6.894303301228106e-07, "loss": 1.1549, "step": 42424 }, { "epoch": 3.324843260188088, "grad_norm": 4.286547660827637, "learning_rate": 6.892740868001186e-07, "loss": 1.1759, "step": 42425 }, { "epoch": 3.324921630094044, "grad_norm": 5.776730060577393, "learning_rate": 6.891178598732129e-07, "loss": 1.425, "step": 42426 }, { "epoch": 3.325, "grad_norm": 7.824066162109375, "learning_rate": 6.889616493426893e-07, "loss": 0.4863, "step": 42427 }, { "epoch": 3.325078369905956, "grad_norm": 3.4376537799835205, "learning_rate": 6.888054552091428e-07, "loss": 1.5179, "step": 42428 }, { "epoch": 3.3251567398119124, "grad_norm": 3.3908607959747314, "learning_rate": 6.886492774731651e-07, "loss": 1.3875, "step": 42429 }, { "epoch": 3.3252351097178683, "grad_norm": 15.662099838256836, "learning_rate": 6.884931161353509e-07, "loss": 1.781, "step": 42430 }, { "epoch": 3.3253134796238246, "grad_norm": 4.696346282958984, "learning_rate": 6.883369711962928e-07, "loss": 1.5428, "step": 42431 }, { "epoch": 3.3253918495297805, "grad_norm": 5.927261829376221, "learning_rate": 6.881808426565889e-07, "loss": 1.3249, "step": 42432 }, { "epoch": 3.325470219435737, "grad_norm": 4.11881160736084, "learning_rate": 6.880247305168292e-07, "loss": 1.2268, "step": 42433 }, { "epoch": 3.3255485893416927, "grad_norm": 3.0906755924224854, "learning_rate": 6.878686347776087e-07, "loss": 1.2493, "step": 42434 }, { "epoch": 3.325626959247649, "grad_norm": 3.2395989894866943, "learning_rate": 6.877125554395209e-07, "loss": 1.4433, "step": 42435 }, { "epoch": 3.325705329153605, "grad_norm": 3.9148178100585938, "learning_rate": 6.875564925031597e-07, "loss": 1.4097, "step": 42436 }, { "epoch": 3.3257836990595613, "grad_norm": 4.1010518074035645, "learning_rate": 6.874004459691186e-07, "loss": 1.2979, "step": 42437 }, { "epoch": 3.325862068965517, "grad_norm": 5.415030479431152, "learning_rate": 6.872444158379915e-07, "loss": 0.7009, "step": 42438 }, { "epoch": 3.3259404388714735, "grad_norm": 2.439347505569458, "learning_rate": 6.870884021103719e-07, "loss": 1.493, "step": 42439 }, { "epoch": 3.3260188087774294, "grad_norm": 2.888413429260254, "learning_rate": 6.869324047868509e-07, "loss": 1.6192, "step": 42440 }, { "epoch": 3.3260971786833857, "grad_norm": 2.7142412662506104, "learning_rate": 6.867764238680247e-07, "loss": 1.2272, "step": 42441 }, { "epoch": 3.3261755485893416, "grad_norm": 4.217763900756836, "learning_rate": 6.866204593544856e-07, "loss": 1.5924, "step": 42442 }, { "epoch": 3.326253918495298, "grad_norm": 3.125938653945923, "learning_rate": 6.864645112468282e-07, "loss": 1.1381, "step": 42443 }, { "epoch": 3.326332288401254, "grad_norm": 4.541082382202148, "learning_rate": 6.863085795456432e-07, "loss": 1.5397, "step": 42444 }, { "epoch": 3.32641065830721, "grad_norm": 2.9509592056274414, "learning_rate": 6.861526642515232e-07, "loss": 1.2575, "step": 42445 }, { "epoch": 3.326489028213166, "grad_norm": 3.4876694679260254, "learning_rate": 6.859967653650635e-07, "loss": 1.5609, "step": 42446 }, { "epoch": 3.3265673981191224, "grad_norm": 2.862983465194702, "learning_rate": 6.858408828868579e-07, "loss": 1.1254, "step": 42447 }, { "epoch": 3.3266457680250783, "grad_norm": 4.775559902191162, "learning_rate": 6.856850168174966e-07, "loss": 0.5885, "step": 42448 }, { "epoch": 3.3267241379310346, "grad_norm": 3.7436399459838867, "learning_rate": 6.855291671575731e-07, "loss": 1.1522, "step": 42449 }, { "epoch": 3.3268025078369905, "grad_norm": 3.6559534072875977, "learning_rate": 6.853733339076812e-07, "loss": 1.4471, "step": 42450 }, { "epoch": 3.326880877742947, "grad_norm": 3.807742118835449, "learning_rate": 6.852175170684128e-07, "loss": 1.4015, "step": 42451 }, { "epoch": 3.3269592476489027, "grad_norm": 4.711103439331055, "learning_rate": 6.850617166403606e-07, "loss": 1.4671, "step": 42452 }, { "epoch": 3.327037617554859, "grad_norm": 2.925126075744629, "learning_rate": 6.849059326241175e-07, "loss": 1.7856, "step": 42453 }, { "epoch": 3.327115987460815, "grad_norm": 8.772655487060547, "learning_rate": 6.847501650202759e-07, "loss": 2.2413, "step": 42454 }, { "epoch": 3.3271943573667713, "grad_norm": 2.7549641132354736, "learning_rate": 6.84594413829428e-07, "loss": 1.2583, "step": 42455 }, { "epoch": 3.327272727272727, "grad_norm": 4.16159725189209, "learning_rate": 6.844386790521662e-07, "loss": 2.1526, "step": 42456 }, { "epoch": 3.3273510971786835, "grad_norm": 4.166127681732178, "learning_rate": 6.842829606890833e-07, "loss": 1.2196, "step": 42457 }, { "epoch": 3.3274294670846394, "grad_norm": 4.465532302856445, "learning_rate": 6.841272587407727e-07, "loss": 1.785, "step": 42458 }, { "epoch": 3.3275078369905957, "grad_norm": 5.99052619934082, "learning_rate": 6.839715732078229e-07, "loss": 1.0716, "step": 42459 }, { "epoch": 3.3275862068965516, "grad_norm": 2.8168444633483887, "learning_rate": 6.838159040908294e-07, "loss": 1.2431, "step": 42460 }, { "epoch": 3.327664576802508, "grad_norm": 3.71366024017334, "learning_rate": 6.836602513903834e-07, "loss": 1.2197, "step": 42461 }, { "epoch": 3.327742946708464, "grad_norm": 5.028420448303223, "learning_rate": 6.835046151070774e-07, "loss": 1.4588, "step": 42462 }, { "epoch": 3.32782131661442, "grad_norm": 2.5729525089263916, "learning_rate": 6.833489952415024e-07, "loss": 1.2554, "step": 42463 }, { "epoch": 3.327899686520376, "grad_norm": 4.926794528961182, "learning_rate": 6.831933917942491e-07, "loss": 1.2369, "step": 42464 }, { "epoch": 3.3279780564263324, "grad_norm": 3.084214210510254, "learning_rate": 6.830378047659131e-07, "loss": 1.5486, "step": 42465 }, { "epoch": 3.3280564263322883, "grad_norm": 2.3447659015655518, "learning_rate": 6.828822341570829e-07, "loss": 1.3755, "step": 42466 }, { "epoch": 3.3281347962382446, "grad_norm": 4.388881683349609, "learning_rate": 6.827266799683513e-07, "loss": 1.4334, "step": 42467 }, { "epoch": 3.3282131661442005, "grad_norm": 3.720767021179199, "learning_rate": 6.825711422003095e-07, "loss": 1.449, "step": 42468 }, { "epoch": 3.328291536050157, "grad_norm": 5.813684463500977, "learning_rate": 6.824156208535498e-07, "loss": 0.7375, "step": 42469 }, { "epoch": 3.3283699059561127, "grad_norm": 2.863662004470825, "learning_rate": 6.82260115928664e-07, "loss": 1.4153, "step": 42470 }, { "epoch": 3.328448275862069, "grad_norm": 9.488120079040527, "learning_rate": 6.821046274262422e-07, "loss": 1.667, "step": 42471 }, { "epoch": 3.328526645768025, "grad_norm": 5.020186424255371, "learning_rate": 6.819491553468771e-07, "loss": 1.0477, "step": 42472 }, { "epoch": 3.3286050156739813, "grad_norm": 3.207538366317749, "learning_rate": 6.817936996911595e-07, "loss": 1.387, "step": 42473 }, { "epoch": 3.328683385579937, "grad_norm": 3.3294224739074707, "learning_rate": 6.816382604596805e-07, "loss": 1.2941, "step": 42474 }, { "epoch": 3.3287617554858935, "grad_norm": 4.742967128753662, "learning_rate": 6.814828376530313e-07, "loss": 0.9446, "step": 42475 }, { "epoch": 3.3288401253918494, "grad_norm": 5.348574638366699, "learning_rate": 6.813274312718043e-07, "loss": 1.162, "step": 42476 }, { "epoch": 3.3289184952978057, "grad_norm": 8.381461143493652, "learning_rate": 6.811720413165878e-07, "loss": 1.2003, "step": 42477 }, { "epoch": 3.3289968652037616, "grad_norm": 2.7938292026519775, "learning_rate": 6.810166677879754e-07, "loss": 1.4811, "step": 42478 }, { "epoch": 3.329075235109718, "grad_norm": 2.8360815048217773, "learning_rate": 6.808613106865574e-07, "loss": 1.4472, "step": 42479 }, { "epoch": 3.329153605015674, "grad_norm": 3.2687010765075684, "learning_rate": 6.807059700129254e-07, "loss": 0.7866, "step": 42480 }, { "epoch": 3.32923197492163, "grad_norm": 3.884955883026123, "learning_rate": 6.805506457676686e-07, "loss": 0.9057, "step": 42481 }, { "epoch": 3.329310344827586, "grad_norm": 7.457735538482666, "learning_rate": 6.803953379513783e-07, "loss": 1.2171, "step": 42482 }, { "epoch": 3.3293887147335424, "grad_norm": 4.711280822753906, "learning_rate": 6.802400465646442e-07, "loss": 1.255, "step": 42483 }, { "epoch": 3.3294670846394983, "grad_norm": 2.4006693363189697, "learning_rate": 6.800847716080605e-07, "loss": 1.0641, "step": 42484 }, { "epoch": 3.3295454545454546, "grad_norm": 2.7797210216522217, "learning_rate": 6.799295130822148e-07, "loss": 1.3456, "step": 42485 }, { "epoch": 3.3296238244514105, "grad_norm": 2.367959976196289, "learning_rate": 6.797742709876981e-07, "loss": 1.3613, "step": 42486 }, { "epoch": 3.329702194357367, "grad_norm": 3.4636850357055664, "learning_rate": 6.796190453251006e-07, "loss": 1.8725, "step": 42487 }, { "epoch": 3.3297805642633227, "grad_norm": 2.247659206390381, "learning_rate": 6.794638360950134e-07, "loss": 1.2665, "step": 42488 }, { "epoch": 3.329858934169279, "grad_norm": 4.127347946166992, "learning_rate": 6.793086432980268e-07, "loss": 1.2619, "step": 42489 }, { "epoch": 3.329937304075235, "grad_norm": 2.7188057899475098, "learning_rate": 6.791534669347311e-07, "loss": 1.315, "step": 42490 }, { "epoch": 3.3300156739811912, "grad_norm": 4.081213474273682, "learning_rate": 6.789983070057154e-07, "loss": 1.0984, "step": 42491 }, { "epoch": 3.330094043887147, "grad_norm": 2.6220290660858154, "learning_rate": 6.788431635115716e-07, "loss": 1.3327, "step": 42492 }, { "epoch": 3.3301724137931035, "grad_norm": 4.450235366821289, "learning_rate": 6.786880364528886e-07, "loss": 1.4197, "step": 42493 }, { "epoch": 3.33025078369906, "grad_norm": 2.317201852798462, "learning_rate": 6.785329258302565e-07, "loss": 1.0141, "step": 42494 }, { "epoch": 3.3303291536050157, "grad_norm": 3.1164329051971436, "learning_rate": 6.783778316442668e-07, "loss": 1.2464, "step": 42495 }, { "epoch": 3.3304075235109716, "grad_norm": 4.050881862640381, "learning_rate": 6.782227538955061e-07, "loss": 1.3474, "step": 42496 }, { "epoch": 3.330485893416928, "grad_norm": 12.089218139648438, "learning_rate": 6.780676925845669e-07, "loss": 0.888, "step": 42497 }, { "epoch": 3.3305642633228842, "grad_norm": 5.035217761993408, "learning_rate": 6.77912647712039e-07, "loss": 1.3191, "step": 42498 }, { "epoch": 3.33064263322884, "grad_norm": 4.749563694000244, "learning_rate": 6.777576192785118e-07, "loss": 1.1551, "step": 42499 }, { "epoch": 3.330721003134796, "grad_norm": 2.6274707317352295, "learning_rate": 6.776026072845738e-07, "loss": 1.5312, "step": 42500 }, { "epoch": 3.3307993730407524, "grad_norm": 2.933450222015381, "learning_rate": 6.774476117308143e-07, "loss": 1.4475, "step": 42501 }, { "epoch": 3.3308777429467087, "grad_norm": 3.2491354942321777, "learning_rate": 6.772926326178253e-07, "loss": 1.226, "step": 42502 }, { "epoch": 3.3309561128526646, "grad_norm": 2.325871706008911, "learning_rate": 6.771376699461951e-07, "loss": 0.9296, "step": 42503 }, { "epoch": 3.3310344827586205, "grad_norm": 3.0640780925750732, "learning_rate": 6.769827237165122e-07, "loss": 1.146, "step": 42504 }, { "epoch": 3.331112852664577, "grad_norm": 6.09327507019043, "learning_rate": 6.768277939293666e-07, "loss": 1.4516, "step": 42505 }, { "epoch": 3.331191222570533, "grad_norm": 3.3360626697540283, "learning_rate": 6.766728805853478e-07, "loss": 0.9839, "step": 42506 }, { "epoch": 3.331269592476489, "grad_norm": 5.006463050842285, "learning_rate": 6.765179836850444e-07, "loss": 1.576, "step": 42507 }, { "epoch": 3.331347962382445, "grad_norm": 4.696588039398193, "learning_rate": 6.763631032290463e-07, "loss": 0.3123, "step": 42508 }, { "epoch": 3.3314263322884012, "grad_norm": 3.837362051010132, "learning_rate": 6.762082392179415e-07, "loss": 1.3362, "step": 42509 }, { "epoch": 3.3315047021943576, "grad_norm": 2.87515926361084, "learning_rate": 6.760533916523204e-07, "loss": 1.1378, "step": 42510 }, { "epoch": 3.3315830721003135, "grad_norm": 3.6090304851531982, "learning_rate": 6.758985605327712e-07, "loss": 1.9245, "step": 42511 }, { "epoch": 3.3316614420062693, "grad_norm": 13.407135009765625, "learning_rate": 6.757437458598826e-07, "loss": 0.8459, "step": 42512 }, { "epoch": 3.3317398119122257, "grad_norm": 4.441155433654785, "learning_rate": 6.755889476342436e-07, "loss": 1.1028, "step": 42513 }, { "epoch": 3.331818181818182, "grad_norm": 2.456805467605591, "learning_rate": 6.754341658564445e-07, "loss": 1.3335, "step": 42514 }, { "epoch": 3.331896551724138, "grad_norm": 3.8437108993530273, "learning_rate": 6.752794005270702e-07, "loss": 1.2902, "step": 42515 }, { "epoch": 3.3319749216300942, "grad_norm": 5.132384300231934, "learning_rate": 6.75124651646713e-07, "loss": 1.4304, "step": 42516 }, { "epoch": 3.33205329153605, "grad_norm": 2.3589553833007812, "learning_rate": 6.749699192159598e-07, "loss": 0.9407, "step": 42517 }, { "epoch": 3.3321316614420065, "grad_norm": 5.9631123542785645, "learning_rate": 6.74815203235401e-07, "loss": 1.4707, "step": 42518 }, { "epoch": 3.3322100313479623, "grad_norm": 7.473028659820557, "learning_rate": 6.746605037056225e-07, "loss": 1.2308, "step": 42519 }, { "epoch": 3.3322884012539187, "grad_norm": 2.8931429386138916, "learning_rate": 6.745058206272132e-07, "loss": 1.3144, "step": 42520 }, { "epoch": 3.3323667711598746, "grad_norm": 3.1615371704101562, "learning_rate": 6.743511540007636e-07, "loss": 1.2694, "step": 42521 }, { "epoch": 3.332445141065831, "grad_norm": 4.297525882720947, "learning_rate": 6.741965038268594e-07, "loss": 1.3029, "step": 42522 }, { "epoch": 3.332523510971787, "grad_norm": 3.063105583190918, "learning_rate": 6.740418701060897e-07, "loss": 1.2547, "step": 42523 }, { "epoch": 3.332601880877743, "grad_norm": 9.305713653564453, "learning_rate": 6.738872528390434e-07, "loss": 1.5276, "step": 42524 }, { "epoch": 3.332680250783699, "grad_norm": 7.246814727783203, "learning_rate": 6.737326520263071e-07, "loss": 1.6288, "step": 42525 }, { "epoch": 3.3327586206896553, "grad_norm": 3.651355743408203, "learning_rate": 6.7357806766847e-07, "loss": 1.5416, "step": 42526 }, { "epoch": 3.3328369905956112, "grad_norm": 3.2622458934783936, "learning_rate": 6.734234997661199e-07, "loss": 1.0093, "step": 42527 }, { "epoch": 3.3329153605015676, "grad_norm": 2.453167200088501, "learning_rate": 6.732689483198446e-07, "loss": 0.9125, "step": 42528 }, { "epoch": 3.3329937304075234, "grad_norm": 6.496792316436768, "learning_rate": 6.731144133302319e-07, "loss": 1.0935, "step": 42529 }, { "epoch": 3.33307210031348, "grad_norm": 9.444607734680176, "learning_rate": 6.729598947978694e-07, "loss": 0.655, "step": 42530 }, { "epoch": 3.3331504702194357, "grad_norm": 4.125358581542969, "learning_rate": 6.728053927233447e-07, "loss": 1.4945, "step": 42531 }, { "epoch": 3.333228840125392, "grad_norm": 3.110604763031006, "learning_rate": 6.726509071072473e-07, "loss": 1.5144, "step": 42532 }, { "epoch": 3.333307210031348, "grad_norm": 2.8158082962036133, "learning_rate": 6.72496437950162e-07, "loss": 1.2853, "step": 42533 }, { "epoch": 3.333385579937304, "grad_norm": 2.458350658416748, "learning_rate": 6.723419852526758e-07, "loss": 1.3622, "step": 42534 }, { "epoch": 3.33346394984326, "grad_norm": 4.230419635772705, "learning_rate": 6.721875490153795e-07, "loss": 1.1929, "step": 42535 }, { "epoch": 3.3335423197492164, "grad_norm": 3.977226972579956, "learning_rate": 6.720331292388599e-07, "loss": 0.8984, "step": 42536 }, { "epoch": 3.3336206896551723, "grad_norm": 3.0823123455047607, "learning_rate": 6.718787259237025e-07, "loss": 1.3488, "step": 42537 }, { "epoch": 3.3336990595611287, "grad_norm": 3.2081925868988037, "learning_rate": 6.717243390704947e-07, "loss": 1.1636, "step": 42538 }, { "epoch": 3.3337774294670846, "grad_norm": 2.122114419937134, "learning_rate": 6.715699686798238e-07, "loss": 0.7062, "step": 42539 }, { "epoch": 3.333855799373041, "grad_norm": 2.262495279312134, "learning_rate": 6.714156147522793e-07, "loss": 1.2235, "step": 42540 }, { "epoch": 3.3339341692789968, "grad_norm": 4.979361057281494, "learning_rate": 6.712612772884458e-07, "loss": 1.6251, "step": 42541 }, { "epoch": 3.334012539184953, "grad_norm": 2.6309967041015625, "learning_rate": 6.711069562889105e-07, "loss": 1.4428, "step": 42542 }, { "epoch": 3.334090909090909, "grad_norm": 2.7991669178009033, "learning_rate": 6.709526517542613e-07, "loss": 1.2941, "step": 42543 }, { "epoch": 3.3341692789968653, "grad_norm": 2.3436272144317627, "learning_rate": 6.707983636850846e-07, "loss": 1.0879, "step": 42544 }, { "epoch": 3.334247648902821, "grad_norm": 2.659822702407837, "learning_rate": 6.706440920819673e-07, "loss": 1.3089, "step": 42545 }, { "epoch": 3.3343260188087775, "grad_norm": 3.945236921310425, "learning_rate": 6.704898369454966e-07, "loss": 1.0929, "step": 42546 }, { "epoch": 3.3344043887147334, "grad_norm": 2.5150063037872314, "learning_rate": 6.703355982762594e-07, "loss": 1.2133, "step": 42547 }, { "epoch": 3.3344827586206898, "grad_norm": 4.458286285400391, "learning_rate": 6.701813760748393e-07, "loss": 1.2378, "step": 42548 }, { "epoch": 3.3345611285266457, "grad_norm": 5.3904924392700195, "learning_rate": 6.700271703418265e-07, "loss": 1.686, "step": 42549 }, { "epoch": 3.334639498432602, "grad_norm": 4.033977031707764, "learning_rate": 6.698729810778065e-07, "loss": 1.2136, "step": 42550 }, { "epoch": 3.334717868338558, "grad_norm": 4.520963191986084, "learning_rate": 6.697188082833667e-07, "loss": 1.7353, "step": 42551 }, { "epoch": 3.334796238244514, "grad_norm": 4.163466453552246, "learning_rate": 6.695646519590915e-07, "loss": 1.1, "step": 42552 }, { "epoch": 3.33487460815047, "grad_norm": 2.8075296878814697, "learning_rate": 6.694105121055666e-07, "loss": 1.1289, "step": 42553 }, { "epoch": 3.3349529780564264, "grad_norm": 5.014371395111084, "learning_rate": 6.69256388723381e-07, "loss": 1.2689, "step": 42554 }, { "epoch": 3.3350313479623823, "grad_norm": 2.8679873943328857, "learning_rate": 6.69102281813121e-07, "loss": 1.391, "step": 42555 }, { "epoch": 3.3351097178683387, "grad_norm": 2.6026453971862793, "learning_rate": 6.689481913753703e-07, "loss": 1.2943, "step": 42556 }, { "epoch": 3.3351880877742945, "grad_norm": 19.504364013671875, "learning_rate": 6.68794117410716e-07, "loss": 1.2687, "step": 42557 }, { "epoch": 3.335266457680251, "grad_norm": 2.0693960189819336, "learning_rate": 6.686400599197429e-07, "loss": 1.4331, "step": 42558 }, { "epoch": 3.3353448275862068, "grad_norm": 3.064474105834961, "learning_rate": 6.684860189030412e-07, "loss": 1.4292, "step": 42559 }, { "epoch": 3.335423197492163, "grad_norm": 3.1647491455078125, "learning_rate": 6.683319943611921e-07, "loss": 1.2873, "step": 42560 }, { "epoch": 3.335501567398119, "grad_norm": 24.434228897094727, "learning_rate": 6.681779862947834e-07, "loss": 0.8895, "step": 42561 }, { "epoch": 3.3355799373040753, "grad_norm": 4.201045989990234, "learning_rate": 6.680239947044003e-07, "loss": 0.8918, "step": 42562 }, { "epoch": 3.335658307210031, "grad_norm": 2.452220916748047, "learning_rate": 6.678700195906296e-07, "loss": 1.4167, "step": 42563 }, { "epoch": 3.3357366771159875, "grad_norm": 2.144486904144287, "learning_rate": 6.677160609540556e-07, "loss": 1.3021, "step": 42564 }, { "epoch": 3.3358150470219434, "grad_norm": 2.745530128479004, "learning_rate": 6.675621187952652e-07, "loss": 1.5052, "step": 42565 }, { "epoch": 3.3358934169278998, "grad_norm": 2.8191819190979004, "learning_rate": 6.674081931148435e-07, "loss": 1.3085, "step": 42566 }, { "epoch": 3.3359717868338556, "grad_norm": 2.4391608238220215, "learning_rate": 6.672542839133739e-07, "loss": 1.124, "step": 42567 }, { "epoch": 3.336050156739812, "grad_norm": 4.967154026031494, "learning_rate": 6.671003911914448e-07, "loss": 1.2556, "step": 42568 }, { "epoch": 3.336128526645768, "grad_norm": 3.7780990600585938, "learning_rate": 6.669465149496402e-07, "loss": 1.7073, "step": 42569 }, { "epoch": 3.336206896551724, "grad_norm": 4.005194187164307, "learning_rate": 6.667926551885467e-07, "loss": 1.4537, "step": 42570 }, { "epoch": 3.33628526645768, "grad_norm": 4.149490833282471, "learning_rate": 6.666388119087469e-07, "loss": 0.9407, "step": 42571 }, { "epoch": 3.3363636363636364, "grad_norm": 3.069329023361206, "learning_rate": 6.664849851108263e-07, "loss": 1.3675, "step": 42572 }, { "epoch": 3.3364420062695923, "grad_norm": 4.452537536621094, "learning_rate": 6.663311747953715e-07, "loss": 1.3311, "step": 42573 }, { "epoch": 3.3365203761755486, "grad_norm": 4.028634071350098, "learning_rate": 6.661773809629691e-07, "loss": 1.412, "step": 42574 }, { "epoch": 3.3365987460815045, "grad_norm": 3.999396324157715, "learning_rate": 6.660236036142003e-07, "loss": 0.693, "step": 42575 }, { "epoch": 3.336677115987461, "grad_norm": 7.189159870147705, "learning_rate": 6.658698427496518e-07, "loss": 1.2097, "step": 42576 }, { "epoch": 3.3367554858934168, "grad_norm": 3.841127872467041, "learning_rate": 6.65716098369908e-07, "loss": 1.539, "step": 42577 }, { "epoch": 3.336833855799373, "grad_norm": 20.15812873840332, "learning_rate": 6.655623704755543e-07, "loss": 0.8089, "step": 42578 }, { "epoch": 3.336912225705329, "grad_norm": 3.2302732467651367, "learning_rate": 6.654086590671749e-07, "loss": 1.1677, "step": 42579 }, { "epoch": 3.3369905956112853, "grad_norm": 3.117156505584717, "learning_rate": 6.652549641453543e-07, "loss": 0.7136, "step": 42580 }, { "epoch": 3.337068965517241, "grad_norm": 4.539432525634766, "learning_rate": 6.651012857106776e-07, "loss": 1.1052, "step": 42581 }, { "epoch": 3.3371473354231975, "grad_norm": 3.287716865539551, "learning_rate": 6.649476237637292e-07, "loss": 1.05, "step": 42582 }, { "epoch": 3.3372257053291534, "grad_norm": 3.117516279220581, "learning_rate": 6.647939783050933e-07, "loss": 1.2416, "step": 42583 }, { "epoch": 3.3373040752351097, "grad_norm": 2.6688966751098633, "learning_rate": 6.646403493353542e-07, "loss": 1.5385, "step": 42584 }, { "epoch": 3.337382445141066, "grad_norm": 2.6551923751831055, "learning_rate": 6.644867368550967e-07, "loss": 1.1985, "step": 42585 }, { "epoch": 3.337460815047022, "grad_norm": 4.862443447113037, "learning_rate": 6.643331408649051e-07, "loss": 1.1504, "step": 42586 }, { "epoch": 3.337539184952978, "grad_norm": 12.46793270111084, "learning_rate": 6.641795613653629e-07, "loss": 2.0266, "step": 42587 }, { "epoch": 3.337617554858934, "grad_norm": 2.7846596240997314, "learning_rate": 6.640259983570552e-07, "loss": 1.1353, "step": 42588 }, { "epoch": 3.3376959247648905, "grad_norm": 3.516279935836792, "learning_rate": 6.638724518405653e-07, "loss": 0.8024, "step": 42589 }, { "epoch": 3.3377742946708464, "grad_norm": 2.925297260284424, "learning_rate": 6.637189218164758e-07, "loss": 1.1631, "step": 42590 }, { "epoch": 3.3378526645768023, "grad_norm": 4.497131824493408, "learning_rate": 6.635654082853737e-07, "loss": 1.3595, "step": 42591 }, { "epoch": 3.3379310344827586, "grad_norm": 2.8133959770202637, "learning_rate": 6.634119112478421e-07, "loss": 1.6651, "step": 42592 }, { "epoch": 3.338009404388715, "grad_norm": 2.5055856704711914, "learning_rate": 6.632584307044637e-07, "loss": 1.4306, "step": 42593 }, { "epoch": 3.338087774294671, "grad_norm": 10.623431205749512, "learning_rate": 6.63104966655822e-07, "loss": 0.7917, "step": 42594 }, { "epoch": 3.3381661442006267, "grad_norm": 3.788073778152466, "learning_rate": 6.629515191025022e-07, "loss": 0.345, "step": 42595 }, { "epoch": 3.338244514106583, "grad_norm": 4.4806623458862305, "learning_rate": 6.627980880450868e-07, "loss": 1.2989, "step": 42596 }, { "epoch": 3.3383228840125394, "grad_norm": 4.0194549560546875, "learning_rate": 6.626446734841597e-07, "loss": 1.6532, "step": 42597 }, { "epoch": 3.3384012539184953, "grad_norm": 2.6591436862945557, "learning_rate": 6.624912754203044e-07, "loss": 1.2475, "step": 42598 }, { "epoch": 3.338479623824451, "grad_norm": 2.7109901905059814, "learning_rate": 6.623378938541042e-07, "loss": 1.3714, "step": 42599 }, { "epoch": 3.3385579937304075, "grad_norm": 2.995915651321411, "learning_rate": 6.621845287861433e-07, "loss": 1.1099, "step": 42600 }, { "epoch": 3.338636363636364, "grad_norm": 3.119340658187866, "learning_rate": 6.620311802170037e-07, "loss": 0.9893, "step": 42601 }, { "epoch": 3.3387147335423197, "grad_norm": 2.532045364379883, "learning_rate": 6.618778481472699e-07, "loss": 1.5184, "step": 42602 }, { "epoch": 3.3387931034482756, "grad_norm": 12.562732696533203, "learning_rate": 6.617245325775257e-07, "loss": 1.6943, "step": 42603 }, { "epoch": 3.338871473354232, "grad_norm": 2.5895304679870605, "learning_rate": 6.615712335083507e-07, "loss": 1.1608, "step": 42604 }, { "epoch": 3.3389498432601883, "grad_norm": 4.245048522949219, "learning_rate": 6.614179509403312e-07, "loss": 1.5077, "step": 42605 }, { "epoch": 3.339028213166144, "grad_norm": 2.0032403469085693, "learning_rate": 6.6126468487405e-07, "loss": 0.54, "step": 42606 }, { "epoch": 3.3391065830721, "grad_norm": 7.343755722045898, "learning_rate": 6.611114353100901e-07, "loss": 1.1906, "step": 42607 }, { "epoch": 3.3391849529780564, "grad_norm": 3.6700406074523926, "learning_rate": 6.60958202249033e-07, "loss": 1.3939, "step": 42608 }, { "epoch": 3.3392633228840127, "grad_norm": 2.81931209564209, "learning_rate": 6.608049856914606e-07, "loss": 1.3749, "step": 42609 }, { "epoch": 3.3393416927899686, "grad_norm": 4.303833961486816, "learning_rate": 6.606517856379585e-07, "loss": 1.1499, "step": 42610 }, { "epoch": 3.339420062695925, "grad_norm": 9.759918212890625, "learning_rate": 6.604986020891096e-07, "loss": 1.3563, "step": 42611 }, { "epoch": 3.339498432601881, "grad_norm": 3.608538866043091, "learning_rate": 6.603454350454935e-07, "loss": 1.0476, "step": 42612 }, { "epoch": 3.339576802507837, "grad_norm": 3.017970323562622, "learning_rate": 6.601922845076947e-07, "loss": 1.7649, "step": 42613 }, { "epoch": 3.339655172413793, "grad_norm": 4.002851486206055, "learning_rate": 6.600391504762943e-07, "loss": 1.7184, "step": 42614 }, { "epoch": 3.3397335423197494, "grad_norm": 7.710230350494385, "learning_rate": 6.59886032951878e-07, "loss": 1.1207, "step": 42615 }, { "epoch": 3.3398119122257053, "grad_norm": 2.9527437686920166, "learning_rate": 6.597329319350249e-07, "loss": 1.2888, "step": 42616 }, { "epoch": 3.3398902821316616, "grad_norm": 3.3879640102386475, "learning_rate": 6.595798474263187e-07, "loss": 1.3353, "step": 42617 }, { "epoch": 3.3399686520376175, "grad_norm": 3.4373791217803955, "learning_rate": 6.594267794263409e-07, "loss": 1.349, "step": 42618 }, { "epoch": 3.340047021943574, "grad_norm": 3.7934412956237793, "learning_rate": 6.592737279356743e-07, "loss": 1.3319, "step": 42619 }, { "epoch": 3.3401253918495297, "grad_norm": 7.160111427307129, "learning_rate": 6.59120692954901e-07, "loss": 0.9524, "step": 42620 }, { "epoch": 3.340203761755486, "grad_norm": 5.006476402282715, "learning_rate": 6.58967674484603e-07, "loss": 1.5449, "step": 42621 }, { "epoch": 3.340282131661442, "grad_norm": 2.581735610961914, "learning_rate": 6.588146725253636e-07, "loss": 1.4979, "step": 42622 }, { "epoch": 3.3403605015673983, "grad_norm": 8.071136474609375, "learning_rate": 6.586616870777607e-07, "loss": 0.4586, "step": 42623 }, { "epoch": 3.340438871473354, "grad_norm": 7.831239700317383, "learning_rate": 6.5850871814238e-07, "loss": 1.7836, "step": 42624 }, { "epoch": 3.3405172413793105, "grad_norm": 3.129319190979004, "learning_rate": 6.583557657198025e-07, "loss": 0.9946, "step": 42625 }, { "epoch": 3.3405956112852664, "grad_norm": 2.8435089588165283, "learning_rate": 6.582028298106107e-07, "loss": 1.4749, "step": 42626 }, { "epoch": 3.3406739811912227, "grad_norm": 4.146947383880615, "learning_rate": 6.580499104153842e-07, "loss": 1.3443, "step": 42627 }, { "epoch": 3.3407523510971786, "grad_norm": 2.5206141471862793, "learning_rate": 6.578970075347046e-07, "loss": 1.3359, "step": 42628 }, { "epoch": 3.340830721003135, "grad_norm": 3.4984371662139893, "learning_rate": 6.577441211691555e-07, "loss": 1.3099, "step": 42629 }, { "epoch": 3.340909090909091, "grad_norm": 3.2653543949127197, "learning_rate": 6.575912513193184e-07, "loss": 1.3858, "step": 42630 }, { "epoch": 3.340987460815047, "grad_norm": 3.385460615158081, "learning_rate": 6.574383979857729e-07, "loss": 1.466, "step": 42631 }, { "epoch": 3.341065830721003, "grad_norm": 3.631891965866089, "learning_rate": 6.572855611691015e-07, "loss": 0.9728, "step": 42632 }, { "epoch": 3.3411442006269594, "grad_norm": 3.2056682109832764, "learning_rate": 6.571327408698846e-07, "loss": 1.3517, "step": 42633 }, { "epoch": 3.3412225705329153, "grad_norm": 5.394867897033691, "learning_rate": 6.569799370887048e-07, "loss": 1.5604, "step": 42634 }, { "epoch": 3.3413009404388716, "grad_norm": 6.917916774749756, "learning_rate": 6.56827149826142e-07, "loss": 0.9681, "step": 42635 }, { "epoch": 3.3413793103448275, "grad_norm": 7.291393280029297, "learning_rate": 6.566743790827784e-07, "loss": 1.1091, "step": 42636 }, { "epoch": 3.341457680250784, "grad_norm": 4.136322498321533, "learning_rate": 6.565216248591944e-07, "loss": 1.5467, "step": 42637 }, { "epoch": 3.3415360501567397, "grad_norm": 3.151923418045044, "learning_rate": 6.563688871559715e-07, "loss": 1.5404, "step": 42638 }, { "epoch": 3.341614420062696, "grad_norm": 23.768232345581055, "learning_rate": 6.562161659736898e-07, "loss": 1.3139, "step": 42639 }, { "epoch": 3.341692789968652, "grad_norm": 4.092545986175537, "learning_rate": 6.560634613129308e-07, "loss": 1.3237, "step": 42640 }, { "epoch": 3.3417711598746083, "grad_norm": 4.498673915863037, "learning_rate": 6.559107731742764e-07, "loss": 1.179, "step": 42641 }, { "epoch": 3.341849529780564, "grad_norm": 4.918572902679443, "learning_rate": 6.557581015583042e-07, "loss": 1.6454, "step": 42642 }, { "epoch": 3.3419278996865205, "grad_norm": 2.2403900623321533, "learning_rate": 6.556054464655976e-07, "loss": 1.2512, "step": 42643 }, { "epoch": 3.3420062695924764, "grad_norm": 3.2361011505126953, "learning_rate": 6.554528078967376e-07, "loss": 1.1115, "step": 42644 }, { "epoch": 3.3420846394984327, "grad_norm": 2.806243658065796, "learning_rate": 6.553001858523028e-07, "loss": 1.3771, "step": 42645 }, { "epoch": 3.3421630094043886, "grad_norm": 4.084733486175537, "learning_rate": 6.551475803328744e-07, "loss": 1.4613, "step": 42646 }, { "epoch": 3.342241379310345, "grad_norm": 3.2177534103393555, "learning_rate": 6.549949913390313e-07, "loss": 1.3867, "step": 42647 }, { "epoch": 3.342319749216301, "grad_norm": 3.8702032566070557, "learning_rate": 6.548424188713581e-07, "loss": 1.3955, "step": 42648 }, { "epoch": 3.342398119122257, "grad_norm": 2.2015254497528076, "learning_rate": 6.546898629304316e-07, "loss": 1.1843, "step": 42649 }, { "epoch": 3.342476489028213, "grad_norm": 4.132594585418701, "learning_rate": 6.545373235168328e-07, "loss": 1.5166, "step": 42650 }, { "epoch": 3.3425548589341694, "grad_norm": 7.85633659362793, "learning_rate": 6.543848006311426e-07, "loss": 2.0265, "step": 42651 }, { "epoch": 3.3426332288401253, "grad_norm": 2.969186782836914, "learning_rate": 6.542322942739398e-07, "loss": 1.0457, "step": 42652 }, { "epoch": 3.3427115987460816, "grad_norm": 4.525846004486084, "learning_rate": 6.540798044458057e-07, "loss": 1.218, "step": 42653 }, { "epoch": 3.3427899686520375, "grad_norm": 4.517303466796875, "learning_rate": 6.539273311473193e-07, "loss": 1.5874, "step": 42654 }, { "epoch": 3.342868338557994, "grad_norm": 3.0123229026794434, "learning_rate": 6.537748743790618e-07, "loss": 1.452, "step": 42655 }, { "epoch": 3.3429467084639497, "grad_norm": 2.9186582565307617, "learning_rate": 6.536224341416115e-07, "loss": 1.2621, "step": 42656 }, { "epoch": 3.343025078369906, "grad_norm": 3.4280104637145996, "learning_rate": 6.534700104355496e-07, "loss": 1.4618, "step": 42657 }, { "epoch": 3.343103448275862, "grad_norm": 2.4744253158569336, "learning_rate": 6.533176032614552e-07, "loss": 1.4433, "step": 42658 }, { "epoch": 3.3431818181818183, "grad_norm": 5.3990325927734375, "learning_rate": 6.53165212619909e-07, "loss": 1.7218, "step": 42659 }, { "epoch": 3.343260188087774, "grad_norm": 3.951624870300293, "learning_rate": 6.530128385114887e-07, "loss": 1.2417, "step": 42660 }, { "epoch": 3.3433385579937305, "grad_norm": 2.8131651878356934, "learning_rate": 6.528604809367734e-07, "loss": 1.3429, "step": 42661 }, { "epoch": 3.3434169278996864, "grad_norm": 5.12061882019043, "learning_rate": 6.527081398963453e-07, "loss": 1.5427, "step": 42662 }, { "epoch": 3.3434952978056427, "grad_norm": 3.8281407356262207, "learning_rate": 6.525558153907835e-07, "loss": 1.2569, "step": 42663 }, { "epoch": 3.3435736677115986, "grad_norm": 2.5712368488311768, "learning_rate": 6.524035074206653e-07, "loss": 1.0827, "step": 42664 }, { "epoch": 3.343652037617555, "grad_norm": 2.262946605682373, "learning_rate": 6.522512159865713e-07, "loss": 1.1828, "step": 42665 }, { "epoch": 3.343730407523511, "grad_norm": 3.5692548751831055, "learning_rate": 6.520989410890793e-07, "loss": 1.2339, "step": 42666 }, { "epoch": 3.343808777429467, "grad_norm": 2.7359986305236816, "learning_rate": 6.519466827287718e-07, "loss": 1.3864, "step": 42667 }, { "epoch": 3.343887147335423, "grad_norm": 3.6756792068481445, "learning_rate": 6.517944409062249e-07, "loss": 1.344, "step": 42668 }, { "epoch": 3.3439655172413794, "grad_norm": 5.043404579162598, "learning_rate": 6.516422156220187e-07, "loss": 1.5809, "step": 42669 }, { "epoch": 3.3440438871473352, "grad_norm": 6.749942779541016, "learning_rate": 6.514900068767316e-07, "loss": 1.2189, "step": 42670 }, { "epoch": 3.3441222570532916, "grad_norm": 7.040539264678955, "learning_rate": 6.513378146709431e-07, "loss": 0.8703, "step": 42671 }, { "epoch": 3.3442006269592475, "grad_norm": 3.7767958641052246, "learning_rate": 6.511856390052323e-07, "loss": 1.4123, "step": 42672 }, { "epoch": 3.344278996865204, "grad_norm": 7.822164058685303, "learning_rate": 6.510334798801777e-07, "loss": 1.2509, "step": 42673 }, { "epoch": 3.3443573667711597, "grad_norm": 3.439711332321167, "learning_rate": 6.508813372963574e-07, "loss": 1.1079, "step": 42674 }, { "epoch": 3.344435736677116, "grad_norm": 6.509673118591309, "learning_rate": 6.507292112543511e-07, "loss": 0.9854, "step": 42675 }, { "epoch": 3.344514106583072, "grad_norm": 4.156750679016113, "learning_rate": 6.505771017547369e-07, "loss": 1.456, "step": 42676 }, { "epoch": 3.3445924764890282, "grad_norm": 5.276834011077881, "learning_rate": 6.504250087980929e-07, "loss": 1.0797, "step": 42677 }, { "epoch": 3.344670846394984, "grad_norm": 2.8988006114959717, "learning_rate": 6.502729323849999e-07, "loss": 0.9593, "step": 42678 }, { "epoch": 3.3447492163009405, "grad_norm": 3.595668315887451, "learning_rate": 6.501208725160329e-07, "loss": 1.649, "step": 42679 }, { "epoch": 3.344827586206897, "grad_norm": 2.632723808288574, "learning_rate": 6.49968829191771e-07, "loss": 1.2555, "step": 42680 }, { "epoch": 3.3449059561128527, "grad_norm": 11.830032348632812, "learning_rate": 6.498168024127943e-07, "loss": 1.9198, "step": 42681 }, { "epoch": 3.3449843260188086, "grad_norm": 3.446972131729126, "learning_rate": 6.496647921796812e-07, "loss": 1.1215, "step": 42682 }, { "epoch": 3.345062695924765, "grad_norm": 3.2277300357818604, "learning_rate": 6.495127984930077e-07, "loss": 1.3673, "step": 42683 }, { "epoch": 3.3451410658307212, "grad_norm": 5.4947285652160645, "learning_rate": 6.493608213533531e-07, "loss": 1.3959, "step": 42684 }, { "epoch": 3.345219435736677, "grad_norm": 6.000600814819336, "learning_rate": 6.492088607612951e-07, "loss": 1.0727, "step": 42685 }, { "epoch": 3.345297805642633, "grad_norm": 2.91464900970459, "learning_rate": 6.490569167174121e-07, "loss": 1.6097, "step": 42686 }, { "epoch": 3.3453761755485893, "grad_norm": 5.462491035461426, "learning_rate": 6.489049892222815e-07, "loss": 1.1364, "step": 42687 }, { "epoch": 3.3454545454545457, "grad_norm": 3.4140892028808594, "learning_rate": 6.487530782764817e-07, "loss": 1.6453, "step": 42688 }, { "epoch": 3.3455329153605016, "grad_norm": 3.183716058731079, "learning_rate": 6.486011838805906e-07, "loss": 1.7103, "step": 42689 }, { "epoch": 3.3456112852664575, "grad_norm": 5.9306535720825195, "learning_rate": 6.484493060351848e-07, "loss": 1.1334, "step": 42690 }, { "epoch": 3.345689655172414, "grad_norm": 3.6395068168640137, "learning_rate": 6.48297444740843e-07, "loss": 1.5067, "step": 42691 }, { "epoch": 3.34576802507837, "grad_norm": 2.420663833618164, "learning_rate": 6.481455999981429e-07, "loss": 1.453, "step": 42692 }, { "epoch": 3.345846394984326, "grad_norm": 2.503908634185791, "learning_rate": 6.479937718076612e-07, "loss": 1.6897, "step": 42693 }, { "epoch": 3.345924764890282, "grad_norm": 6.392592906951904, "learning_rate": 6.478419601699765e-07, "loss": 1.2814, "step": 42694 }, { "epoch": 3.3460031347962382, "grad_norm": 2.440908432006836, "learning_rate": 6.476901650856649e-07, "loss": 1.3301, "step": 42695 }, { "epoch": 3.3460815047021946, "grad_norm": 3.7054085731506348, "learning_rate": 6.475383865553048e-07, "loss": 1.5013, "step": 42696 }, { "epoch": 3.3461598746081505, "grad_norm": 4.697775840759277, "learning_rate": 6.473866245794741e-07, "loss": 1.8665, "step": 42697 }, { "epoch": 3.3462382445141063, "grad_norm": 4.379617691040039, "learning_rate": 6.472348791587468e-07, "loss": 1.099, "step": 42698 }, { "epoch": 3.3463166144200627, "grad_norm": 3.850329875946045, "learning_rate": 6.470831502937036e-07, "loss": 1.1651, "step": 42699 }, { "epoch": 3.346394984326019, "grad_norm": 6.0023722648620605, "learning_rate": 6.469314379849212e-07, "loss": 1.1009, "step": 42700 }, { "epoch": 3.346473354231975, "grad_norm": 3.6722376346588135, "learning_rate": 6.467797422329747e-07, "loss": 1.3684, "step": 42701 }, { "epoch": 3.3465517241379312, "grad_norm": 13.545554161071777, "learning_rate": 6.46628063038442e-07, "loss": 1.1812, "step": 42702 }, { "epoch": 3.346630094043887, "grad_norm": 2.6099789142608643, "learning_rate": 6.464764004018992e-07, "loss": 1.1137, "step": 42703 }, { "epoch": 3.3467084639498434, "grad_norm": 3.364633321762085, "learning_rate": 6.463247543239259e-07, "loss": 1.5496, "step": 42704 }, { "epoch": 3.3467868338557993, "grad_norm": 2.6540298461914062, "learning_rate": 6.461731248050962e-07, "loss": 1.2232, "step": 42705 }, { "epoch": 3.3468652037617557, "grad_norm": 5.303674221038818, "learning_rate": 6.460215118459879e-07, "loss": 1.2338, "step": 42706 }, { "epoch": 3.3469435736677116, "grad_norm": 17.531543731689453, "learning_rate": 6.45869915447177e-07, "loss": 1.0463, "step": 42707 }, { "epoch": 3.347021943573668, "grad_norm": 2.6081268787384033, "learning_rate": 6.457183356092411e-07, "loss": 1.5075, "step": 42708 }, { "epoch": 3.347100313479624, "grad_norm": 2.4124927520751953, "learning_rate": 6.455667723327558e-07, "loss": 1.278, "step": 42709 }, { "epoch": 3.34717868338558, "grad_norm": 2.248699903488159, "learning_rate": 6.454152256182977e-07, "loss": 1.4007, "step": 42710 }, { "epoch": 3.347257053291536, "grad_norm": 4.279336929321289, "learning_rate": 6.452636954664438e-07, "loss": 1.0859, "step": 42711 }, { "epoch": 3.3473354231974923, "grad_norm": 3.5584418773651123, "learning_rate": 6.451121818777695e-07, "loss": 1.3037, "step": 42712 }, { "epoch": 3.347413793103448, "grad_norm": 6.123719215393066, "learning_rate": 6.449606848528522e-07, "loss": 1.4772, "step": 42713 }, { "epoch": 3.3474921630094046, "grad_norm": 11.25218391418457, "learning_rate": 6.448092043922677e-07, "loss": 0.5001, "step": 42714 }, { "epoch": 3.3475705329153604, "grad_norm": 3.112355947494507, "learning_rate": 6.446577404965926e-07, "loss": 1.2948, "step": 42715 }, { "epoch": 3.3476489028213168, "grad_norm": 2.2950196266174316, "learning_rate": 6.445062931664014e-07, "loss": 1.0851, "step": 42716 }, { "epoch": 3.3477272727272727, "grad_norm": 25.255496978759766, "learning_rate": 6.4435486240227e-07, "loss": 2.7537, "step": 42717 }, { "epoch": 3.347805642633229, "grad_norm": 3.0821168422698975, "learning_rate": 6.442034482047771e-07, "loss": 1.41, "step": 42718 }, { "epoch": 3.347884012539185, "grad_norm": 2.9189069271087646, "learning_rate": 6.440520505744979e-07, "loss": 1.1193, "step": 42719 }, { "epoch": 3.347962382445141, "grad_norm": 2.559418201446533, "learning_rate": 6.439006695120059e-07, "loss": 1.3989, "step": 42720 }, { "epoch": 3.348040752351097, "grad_norm": 3.3854122161865234, "learning_rate": 6.437493050178783e-07, "loss": 1.4144, "step": 42721 }, { "epoch": 3.3481191222570534, "grad_norm": 4.366861820220947, "learning_rate": 6.435979570926904e-07, "loss": 1.5123, "step": 42722 }, { "epoch": 3.3481974921630093, "grad_norm": 2.9507784843444824, "learning_rate": 6.434466257370198e-07, "loss": 1.2323, "step": 42723 }, { "epoch": 3.3482758620689657, "grad_norm": 2.6267900466918945, "learning_rate": 6.432953109514395e-07, "loss": 1.1125, "step": 42724 }, { "epoch": 3.3483542319749215, "grad_norm": 3.1323397159576416, "learning_rate": 6.431440127365268e-07, "loss": 1.7528, "step": 42725 }, { "epoch": 3.348432601880878, "grad_norm": 4.357877731323242, "learning_rate": 6.429927310928558e-07, "loss": 1.5352, "step": 42726 }, { "epoch": 3.3485109717868338, "grad_norm": 2.738300323486328, "learning_rate": 6.428414660210025e-07, "loss": 1.0282, "step": 42727 }, { "epoch": 3.34858934169279, "grad_norm": 4.0921173095703125, "learning_rate": 6.426902175215427e-07, "loss": 1.3393, "step": 42728 }, { "epoch": 3.348667711598746, "grad_norm": 4.54763126373291, "learning_rate": 6.425389855950509e-07, "loss": 1.2618, "step": 42729 }, { "epoch": 3.3487460815047023, "grad_norm": 4.3435516357421875, "learning_rate": 6.423877702421038e-07, "loss": 0.9417, "step": 42730 }, { "epoch": 3.348824451410658, "grad_norm": 2.438596248626709, "learning_rate": 6.422365714632733e-07, "loss": 1.3105, "step": 42731 }, { "epoch": 3.3489028213166145, "grad_norm": 2.1572420597076416, "learning_rate": 6.420853892591378e-07, "loss": 1.4912, "step": 42732 }, { "epoch": 3.3489811912225704, "grad_norm": 2.499722957611084, "learning_rate": 6.419342236302712e-07, "loss": 1.3177, "step": 42733 }, { "epoch": 3.3490595611285268, "grad_norm": 3.1330981254577637, "learning_rate": 6.417830745772491e-07, "loss": 1.1581, "step": 42734 }, { "epoch": 3.3491379310344827, "grad_norm": 3.737673759460449, "learning_rate": 6.41631942100645e-07, "loss": 1.255, "step": 42735 }, { "epoch": 3.349216300940439, "grad_norm": 3.4712486267089844, "learning_rate": 6.41480826201033e-07, "loss": 1.548, "step": 42736 }, { "epoch": 3.349294670846395, "grad_norm": 2.1329452991485596, "learning_rate": 6.413297268789903e-07, "loss": 1.391, "step": 42737 }, { "epoch": 3.349373040752351, "grad_norm": 3.553483009338379, "learning_rate": 6.41178644135092e-07, "loss": 1.0049, "step": 42738 }, { "epoch": 3.349451410658307, "grad_norm": 2.8646671772003174, "learning_rate": 6.410275779699099e-07, "loss": 1.3446, "step": 42739 }, { "epoch": 3.3495297805642634, "grad_norm": 2.70780873298645, "learning_rate": 6.408765283840201e-07, "loss": 1.4171, "step": 42740 }, { "epoch": 3.3496081504702193, "grad_norm": 4.3358964920043945, "learning_rate": 6.407254953779968e-07, "loss": 0.9113, "step": 42741 }, { "epoch": 3.3496865203761756, "grad_norm": 3.600956439971924, "learning_rate": 6.405744789524149e-07, "loss": 1.5194, "step": 42742 }, { "epoch": 3.3497648902821315, "grad_norm": 2.4098451137542725, "learning_rate": 6.404234791078485e-07, "loss": 1.5852, "step": 42743 }, { "epoch": 3.349843260188088, "grad_norm": 2.893548011779785, "learning_rate": 6.40272495844872e-07, "loss": 1.3274, "step": 42744 }, { "epoch": 3.3499216300940438, "grad_norm": 2.894047260284424, "learning_rate": 6.401215291640595e-07, "loss": 1.1621, "step": 42745 }, { "epoch": 3.35, "grad_norm": 8.10767650604248, "learning_rate": 6.399705790659855e-07, "loss": 1.2191, "step": 42746 }, { "epoch": 3.350078369905956, "grad_norm": 4.045685768127441, "learning_rate": 6.39819645551224e-07, "loss": 1.3423, "step": 42747 }, { "epoch": 3.3501567398119123, "grad_norm": 4.236265659332275, "learning_rate": 6.396687286203485e-07, "loss": 1.1929, "step": 42748 }, { "epoch": 3.350235109717868, "grad_norm": 4.057574272155762, "learning_rate": 6.395178282739351e-07, "loss": 1.5731, "step": 42749 }, { "epoch": 3.3503134796238245, "grad_norm": 2.4143402576446533, "learning_rate": 6.393669445125538e-07, "loss": 1.3763, "step": 42750 }, { "epoch": 3.3503918495297804, "grad_norm": 2.8895275592803955, "learning_rate": 6.392160773367822e-07, "loss": 1.4183, "step": 42751 }, { "epoch": 3.3504702194357368, "grad_norm": 3.820328950881958, "learning_rate": 6.390652267471931e-07, "loss": 1.7018, "step": 42752 }, { "epoch": 3.3505485893416926, "grad_norm": 6.886996269226074, "learning_rate": 6.389143927443603e-07, "loss": 1.4709, "step": 42753 }, { "epoch": 3.350626959247649, "grad_norm": 3.6443209648132324, "learning_rate": 6.387635753288568e-07, "loss": 1.4526, "step": 42754 }, { "epoch": 3.350705329153605, "grad_norm": 3.326308250427246, "learning_rate": 6.386127745012549e-07, "loss": 1.3844, "step": 42755 }, { "epoch": 3.350783699059561, "grad_norm": 4.428954601287842, "learning_rate": 6.384619902621325e-07, "loss": 1.3473, "step": 42756 }, { "epoch": 3.350862068965517, "grad_norm": 3.289475679397583, "learning_rate": 6.383112226120592e-07, "loss": 1.3687, "step": 42757 }, { "epoch": 3.3509404388714734, "grad_norm": 2.8591153621673584, "learning_rate": 6.381604715516093e-07, "loss": 1.5923, "step": 42758 }, { "epoch": 3.3510188087774293, "grad_norm": 7.515578269958496, "learning_rate": 6.38009737081357e-07, "loss": 1.332, "step": 42759 }, { "epoch": 3.3510971786833856, "grad_norm": 3.649739980697632, "learning_rate": 6.378590192018752e-07, "loss": 1.2716, "step": 42760 }, { "epoch": 3.3511755485893415, "grad_norm": 4.150207996368408, "learning_rate": 6.377083179137373e-07, "loss": 1.4634, "step": 42761 }, { "epoch": 3.351253918495298, "grad_norm": 3.205144166946411, "learning_rate": 6.375576332175159e-07, "loss": 1.5242, "step": 42762 }, { "epoch": 3.3513322884012537, "grad_norm": 3.877652883529663, "learning_rate": 6.374069651137849e-07, "loss": 1.0485, "step": 42763 }, { "epoch": 3.35141065830721, "grad_norm": 3.081529378890991, "learning_rate": 6.372563136031168e-07, "loss": 1.2681, "step": 42764 }, { "epoch": 3.351489028213166, "grad_norm": 3.4479739665985107, "learning_rate": 6.371056786860852e-07, "loss": 1.4999, "step": 42765 }, { "epoch": 3.3515673981191223, "grad_norm": 3.773206949234009, "learning_rate": 6.369550603632624e-07, "loss": 1.3047, "step": 42766 }, { "epoch": 3.351645768025078, "grad_norm": 2.8505489826202393, "learning_rate": 6.368044586352212e-07, "loss": 1.2288, "step": 42767 }, { "epoch": 3.3517241379310345, "grad_norm": 2.310795307159424, "learning_rate": 6.366538735025362e-07, "loss": 1.4545, "step": 42768 }, { "epoch": 3.3518025078369904, "grad_norm": 2.918761730194092, "learning_rate": 6.365033049657765e-07, "loss": 0.9872, "step": 42769 }, { "epoch": 3.3518808777429467, "grad_norm": 5.3226799964904785, "learning_rate": 6.363527530255187e-07, "loss": 1.4914, "step": 42770 }, { "epoch": 3.3519592476489026, "grad_norm": 4.234410285949707, "learning_rate": 6.36202217682334e-07, "loss": 1.0137, "step": 42771 }, { "epoch": 3.352037617554859, "grad_norm": 3.43524169921875, "learning_rate": 6.360516989367937e-07, "loss": 0.8974, "step": 42772 }, { "epoch": 3.352115987460815, "grad_norm": 2.406245708465576, "learning_rate": 6.359011967894713e-07, "loss": 1.0893, "step": 42773 }, { "epoch": 3.352194357366771, "grad_norm": 4.5238776206970215, "learning_rate": 6.357507112409384e-07, "loss": 1.1281, "step": 42774 }, { "epoch": 3.3522727272727275, "grad_norm": 2.7578532695770264, "learning_rate": 6.356002422917701e-07, "loss": 1.1273, "step": 42775 }, { "epoch": 3.3523510971786834, "grad_norm": 2.6439733505249023, "learning_rate": 6.35449789942536e-07, "loss": 1.5182, "step": 42776 }, { "epoch": 3.3524294670846393, "grad_norm": 3.8284354209899902, "learning_rate": 6.35299354193809e-07, "loss": 1.3966, "step": 42777 }, { "epoch": 3.3525078369905956, "grad_norm": 2.678805351257324, "learning_rate": 6.351489350461609e-07, "loss": 1.3816, "step": 42778 }, { "epoch": 3.352586206896552, "grad_norm": 3.25007963180542, "learning_rate": 6.34998532500165e-07, "loss": 1.1982, "step": 42779 }, { "epoch": 3.352664576802508, "grad_norm": 4.387510776519775, "learning_rate": 6.348481465563921e-07, "loss": 1.5715, "step": 42780 }, { "epoch": 3.3527429467084637, "grad_norm": 2.3106017112731934, "learning_rate": 6.346977772154156e-07, "loss": 1.4632, "step": 42781 }, { "epoch": 3.35282131661442, "grad_norm": 4.372908115386963, "learning_rate": 6.34547424477806e-07, "loss": 1.154, "step": 42782 }, { "epoch": 3.3528996865203764, "grad_norm": 6.6212544441223145, "learning_rate": 6.343970883441359e-07, "loss": 1.4024, "step": 42783 }, { "epoch": 3.3529780564263323, "grad_norm": 3.8379945755004883, "learning_rate": 6.342467688149767e-07, "loss": 1.1402, "step": 42784 }, { "epoch": 3.353056426332288, "grad_norm": 5.233582019805908, "learning_rate": 6.340964658909004e-07, "loss": 1.5175, "step": 42785 }, { "epoch": 3.3531347962382445, "grad_norm": 2.698509931564331, "learning_rate": 6.339461795724799e-07, "loss": 1.1288, "step": 42786 }, { "epoch": 3.353213166144201, "grad_norm": 3.5259292125701904, "learning_rate": 6.337959098602836e-07, "loss": 1.2195, "step": 42787 }, { "epoch": 3.3532915360501567, "grad_norm": 13.421799659729004, "learning_rate": 6.33645656754886e-07, "loss": 1.1064, "step": 42788 }, { "epoch": 3.3533699059561126, "grad_norm": 4.43543815612793, "learning_rate": 6.334954202568577e-07, "loss": 1.5287, "step": 42789 }, { "epoch": 3.353448275862069, "grad_norm": 5.2334723472595215, "learning_rate": 6.333452003667712e-07, "loss": 0.985, "step": 42790 }, { "epoch": 3.3535266457680253, "grad_norm": 2.9725682735443115, "learning_rate": 6.331949970851953e-07, "loss": 1.4402, "step": 42791 }, { "epoch": 3.353605015673981, "grad_norm": 4.551353454589844, "learning_rate": 6.33044810412703e-07, "loss": 0.5995, "step": 42792 }, { "epoch": 3.353683385579937, "grad_norm": 3.583810329437256, "learning_rate": 6.328946403498637e-07, "loss": 0.9844, "step": 42793 }, { "epoch": 3.3537617554858934, "grad_norm": 15.924469947814941, "learning_rate": 6.327444868972527e-07, "loss": 1.3409, "step": 42794 }, { "epoch": 3.3538401253918497, "grad_norm": 2.767033576965332, "learning_rate": 6.325943500554372e-07, "loss": 1.1502, "step": 42795 }, { "epoch": 3.3539184952978056, "grad_norm": 5.5046610832214355, "learning_rate": 6.324442298249894e-07, "loss": 0.758, "step": 42796 }, { "epoch": 3.353996865203762, "grad_norm": 3.698664665222168, "learning_rate": 6.322941262064807e-07, "loss": 1.5097, "step": 42797 }, { "epoch": 3.354075235109718, "grad_norm": 4.416265487670898, "learning_rate": 6.321440392004818e-07, "loss": 1.0921, "step": 42798 }, { "epoch": 3.354153605015674, "grad_norm": 10.877628326416016, "learning_rate": 6.319939688075632e-07, "loss": 1.1787, "step": 42799 }, { "epoch": 3.35423197492163, "grad_norm": 3.7592530250549316, "learning_rate": 6.318439150282962e-07, "loss": 1.0654, "step": 42800 }, { "epoch": 3.3543103448275864, "grad_norm": 6.965152740478516, "learning_rate": 6.316938778632515e-07, "loss": 0.8158, "step": 42801 }, { "epoch": 3.3543887147335423, "grad_norm": 4.0098958015441895, "learning_rate": 6.315438573129995e-07, "loss": 1.2942, "step": 42802 }, { "epoch": 3.3544670846394986, "grad_norm": 6.152727127075195, "learning_rate": 6.313938533781106e-07, "loss": 1.3809, "step": 42803 }, { "epoch": 3.3545454545454545, "grad_norm": 4.9575066566467285, "learning_rate": 6.312438660591558e-07, "loss": 1.5085, "step": 42804 }, { "epoch": 3.354623824451411, "grad_norm": 3.795262336730957, "learning_rate": 6.310938953567069e-07, "loss": 1.5863, "step": 42805 }, { "epoch": 3.3547021943573667, "grad_norm": 4.8299102783203125, "learning_rate": 6.309439412713303e-07, "loss": 1.0322, "step": 42806 }, { "epoch": 3.354780564263323, "grad_norm": 2.6298282146453857, "learning_rate": 6.307940038035998e-07, "loss": 1.2495, "step": 42807 }, { "epoch": 3.354858934169279, "grad_norm": 5.326729774475098, "learning_rate": 6.306440829540844e-07, "loss": 1.5728, "step": 42808 }, { "epoch": 3.3549373040752353, "grad_norm": 2.5549240112304688, "learning_rate": 6.304941787233565e-07, "loss": 1.1238, "step": 42809 }, { "epoch": 3.355015673981191, "grad_norm": 4.749596118927002, "learning_rate": 6.303442911119828e-07, "loss": 1.281, "step": 42810 }, { "epoch": 3.3550940438871475, "grad_norm": 6.496145248413086, "learning_rate": 6.301944201205346e-07, "loss": 1.4556, "step": 42811 }, { "epoch": 3.3551724137931034, "grad_norm": 3.3297107219696045, "learning_rate": 6.30044565749584e-07, "loss": 1.5901, "step": 42812 }, { "epoch": 3.3552507836990597, "grad_norm": 3.943070650100708, "learning_rate": 6.298947279996981e-07, "loss": 1.6765, "step": 42813 }, { "epoch": 3.3553291536050156, "grad_norm": 2.4190735816955566, "learning_rate": 6.297449068714484e-07, "loss": 1.6517, "step": 42814 }, { "epoch": 3.355407523510972, "grad_norm": 2.4565377235412598, "learning_rate": 6.29595102365404e-07, "loss": 1.253, "step": 42815 }, { "epoch": 3.355485893416928, "grad_norm": 3.7507503032684326, "learning_rate": 6.294453144821355e-07, "loss": 1.3647, "step": 42816 }, { "epoch": 3.355564263322884, "grad_norm": 3.501945972442627, "learning_rate": 6.29295543222212e-07, "loss": 0.9478, "step": 42817 }, { "epoch": 3.35564263322884, "grad_norm": 2.8528687953948975, "learning_rate": 6.291457885862029e-07, "loss": 1.6604, "step": 42818 }, { "epoch": 3.3557210031347964, "grad_norm": 3.392704963684082, "learning_rate": 6.289960505746789e-07, "loss": 1.2202, "step": 42819 }, { "epoch": 3.3557993730407523, "grad_norm": 4.519265174865723, "learning_rate": 6.288463291882085e-07, "loss": 1.2485, "step": 42820 }, { "epoch": 3.3558777429467086, "grad_norm": 2.1214473247528076, "learning_rate": 6.286966244273612e-07, "loss": 0.6113, "step": 42821 }, { "epoch": 3.3559561128526645, "grad_norm": 2.5567405223846436, "learning_rate": 6.285469362927072e-07, "loss": 1.2681, "step": 42822 }, { "epoch": 3.356034482758621, "grad_norm": 5.420192718505859, "learning_rate": 6.283972647848152e-07, "loss": 1.34, "step": 42823 }, { "epoch": 3.3561128526645767, "grad_norm": 2.8892574310302734, "learning_rate": 6.282476099042556e-07, "loss": 1.5099, "step": 42824 }, { "epoch": 3.356191222570533, "grad_norm": 12.087923049926758, "learning_rate": 6.280979716515945e-07, "loss": 1.2571, "step": 42825 }, { "epoch": 3.356269592476489, "grad_norm": 4.0486741065979, "learning_rate": 6.279483500274042e-07, "loss": 1.4886, "step": 42826 }, { "epoch": 3.3563479623824453, "grad_norm": 2.6488921642303467, "learning_rate": 6.277987450322537e-07, "loss": 1.4581, "step": 42827 }, { "epoch": 3.356426332288401, "grad_norm": 3.460256814956665, "learning_rate": 6.276491566667097e-07, "loss": 1.3558, "step": 42828 }, { "epoch": 3.3565047021943575, "grad_norm": 2.9331183433532715, "learning_rate": 6.274995849313431e-07, "loss": 1.6513, "step": 42829 }, { "epoch": 3.3565830721003134, "grad_norm": 3.1467480659484863, "learning_rate": 6.27350029826721e-07, "loss": 0.9885, "step": 42830 }, { "epoch": 3.3566614420062697, "grad_norm": 5.504819393157959, "learning_rate": 6.272004913534158e-07, "loss": 1.5132, "step": 42831 }, { "epoch": 3.3567398119122256, "grad_norm": 3.5587809085845947, "learning_rate": 6.270509695119925e-07, "loss": 1.392, "step": 42832 }, { "epoch": 3.356818181818182, "grad_norm": 4.241469860076904, "learning_rate": 6.269014643030214e-07, "loss": 1.3696, "step": 42833 }, { "epoch": 3.356896551724138, "grad_norm": 2.7704524993896484, "learning_rate": 6.267519757270707e-07, "loss": 1.5134, "step": 42834 }, { "epoch": 3.356974921630094, "grad_norm": 3.7849838733673096, "learning_rate": 6.266025037847096e-07, "loss": 1.1103, "step": 42835 }, { "epoch": 3.35705329153605, "grad_norm": 3.300412654876709, "learning_rate": 6.264530484765063e-07, "loss": 1.111, "step": 42836 }, { "epoch": 3.3571316614420064, "grad_norm": 2.633345365524292, "learning_rate": 6.26303609803029e-07, "loss": 1.5303, "step": 42837 }, { "epoch": 3.3572100313479623, "grad_norm": 2.481813669204712, "learning_rate": 6.261541877648464e-07, "loss": 1.1939, "step": 42838 }, { "epoch": 3.3572884012539186, "grad_norm": 4.816120147705078, "learning_rate": 6.260047823625265e-07, "loss": 1.5682, "step": 42839 }, { "epoch": 3.3573667711598745, "grad_norm": 2.2826461791992188, "learning_rate": 6.258553935966382e-07, "loss": 1.2932, "step": 42840 }, { "epoch": 3.357445141065831, "grad_norm": 4.008690357208252, "learning_rate": 6.257060214677491e-07, "loss": 1.134, "step": 42841 }, { "epoch": 3.3575235109717867, "grad_norm": 3.1108274459838867, "learning_rate": 6.255566659764284e-07, "loss": 1.1724, "step": 42842 }, { "epoch": 3.357601880877743, "grad_norm": 2.4383771419525146, "learning_rate": 6.254073271232419e-07, "loss": 1.0742, "step": 42843 }, { "epoch": 3.357680250783699, "grad_norm": 2.869830369949341, "learning_rate": 6.252580049087587e-07, "loss": 1.0911, "step": 42844 }, { "epoch": 3.3577586206896552, "grad_norm": 5.0936055183410645, "learning_rate": 6.251086993335476e-07, "loss": 1.2605, "step": 42845 }, { "epoch": 3.357836990595611, "grad_norm": 3.063732385635376, "learning_rate": 6.24959410398177e-07, "loss": 1.5139, "step": 42846 }, { "epoch": 3.3579153605015675, "grad_norm": 5.157701015472412, "learning_rate": 6.248101381032129e-07, "loss": 1.1818, "step": 42847 }, { "epoch": 3.3579937304075234, "grad_norm": 3.812507152557373, "learning_rate": 6.246608824492234e-07, "loss": 1.8009, "step": 42848 }, { "epoch": 3.3580721003134797, "grad_norm": 2.28425931930542, "learning_rate": 6.245116434367759e-07, "loss": 1.401, "step": 42849 }, { "epoch": 3.3581504702194356, "grad_norm": 3.734100818634033, "learning_rate": 6.243624210664406e-07, "loss": 1.2189, "step": 42850 }, { "epoch": 3.358228840125392, "grad_norm": 4.701943397521973, "learning_rate": 6.242132153387825e-07, "loss": 1.2982, "step": 42851 }, { "epoch": 3.358307210031348, "grad_norm": 4.983834266662598, "learning_rate": 6.24064026254369e-07, "loss": 1.5409, "step": 42852 }, { "epoch": 3.358385579937304, "grad_norm": 4.312386512756348, "learning_rate": 6.239148538137691e-07, "loss": 0.9378, "step": 42853 }, { "epoch": 3.35846394984326, "grad_norm": 2.7252182960510254, "learning_rate": 6.237656980175489e-07, "loss": 1.2106, "step": 42854 }, { "epoch": 3.3585423197492164, "grad_norm": 6.322122097015381, "learning_rate": 6.236165588662768e-07, "loss": 1.3251, "step": 42855 }, { "epoch": 3.3586206896551722, "grad_norm": 8.593087196350098, "learning_rate": 6.234674363605187e-07, "loss": 0.9337, "step": 42856 }, { "epoch": 3.3586990595611286, "grad_norm": 3.560126543045044, "learning_rate": 6.233183305008439e-07, "loss": 1.3299, "step": 42857 }, { "epoch": 3.3587774294670845, "grad_norm": 3.0852181911468506, "learning_rate": 6.231692412878165e-07, "loss": 1.2103, "step": 42858 }, { "epoch": 3.358855799373041, "grad_norm": 2.6267096996307373, "learning_rate": 6.230201687220056e-07, "loss": 1.3859, "step": 42859 }, { "epoch": 3.3589341692789967, "grad_norm": 2.7280006408691406, "learning_rate": 6.228711128039783e-07, "loss": 1.2586, "step": 42860 }, { "epoch": 3.359012539184953, "grad_norm": 6.901788234710693, "learning_rate": 6.22722073534302e-07, "loss": 1.2975, "step": 42861 }, { "epoch": 3.359090909090909, "grad_norm": 2.663888692855835, "learning_rate": 6.225730509135414e-07, "loss": 1.0654, "step": 42862 }, { "epoch": 3.3591692789968652, "grad_norm": 3.469728946685791, "learning_rate": 6.224240449422636e-07, "loss": 1.4221, "step": 42863 }, { "epoch": 3.359247648902821, "grad_norm": 4.4477338790893555, "learning_rate": 6.222750556210371e-07, "loss": 0.9772, "step": 42864 }, { "epoch": 3.3593260188087775, "grad_norm": 3.1568098068237305, "learning_rate": 6.221260829504289e-07, "loss": 1.5711, "step": 42865 }, { "epoch": 3.359404388714734, "grad_norm": 2.3583595752716064, "learning_rate": 6.219771269310037e-07, "loss": 0.9657, "step": 42866 }, { "epoch": 3.3594827586206897, "grad_norm": 5.364260196685791, "learning_rate": 6.218281875633287e-07, "loss": 1.109, "step": 42867 }, { "epoch": 3.3595611285266456, "grad_norm": 2.8337504863739014, "learning_rate": 6.216792648479703e-07, "loss": 1.2358, "step": 42868 }, { "epoch": 3.359639498432602, "grad_norm": 2.329338312149048, "learning_rate": 6.215303587854954e-07, "loss": 1.4352, "step": 42869 }, { "epoch": 3.3597178683385582, "grad_norm": 2.857442855834961, "learning_rate": 6.213814693764697e-07, "loss": 1.2864, "step": 42870 }, { "epoch": 3.359796238244514, "grad_norm": 3.7986745834350586, "learning_rate": 6.2123259662146e-07, "loss": 1.3593, "step": 42871 }, { "epoch": 3.35987460815047, "grad_norm": 2.859548807144165, "learning_rate": 6.210837405210324e-07, "loss": 1.3961, "step": 42872 }, { "epoch": 3.3599529780564263, "grad_norm": 4.612856388092041, "learning_rate": 6.209349010757532e-07, "loss": 1.117, "step": 42873 }, { "epoch": 3.3600313479623827, "grad_norm": 3.5460824966430664, "learning_rate": 6.207860782861886e-07, "loss": 1.1375, "step": 42874 }, { "epoch": 3.3601097178683386, "grad_norm": 4.564394950866699, "learning_rate": 6.206372721529042e-07, "loss": 1.2343, "step": 42875 }, { "epoch": 3.3601880877742945, "grad_norm": 2.841761350631714, "learning_rate": 6.204884826764674e-07, "loss": 0.6722, "step": 42876 }, { "epoch": 3.360266457680251, "grad_norm": 11.105098724365234, "learning_rate": 6.203397098574409e-07, "loss": 0.9757, "step": 42877 }, { "epoch": 3.360344827586207, "grad_norm": 5.152698516845703, "learning_rate": 6.201909536963934e-07, "loss": 1.2976, "step": 42878 }, { "epoch": 3.360423197492163, "grad_norm": 4.919619560241699, "learning_rate": 6.200422141938905e-07, "loss": 1.5225, "step": 42879 }, { "epoch": 3.360501567398119, "grad_norm": 3.09462308883667, "learning_rate": 6.198934913504978e-07, "loss": 1.4168, "step": 42880 }, { "epoch": 3.3605799373040752, "grad_norm": 5.005454063415527, "learning_rate": 6.197447851667798e-07, "loss": 1.3815, "step": 42881 }, { "epoch": 3.3606583072100316, "grad_norm": 3.302294969558716, "learning_rate": 6.195960956433012e-07, "loss": 1.218, "step": 42882 }, { "epoch": 3.3607366771159874, "grad_norm": 5.171322822570801, "learning_rate": 6.194474227806318e-07, "loss": 0.6576, "step": 42883 }, { "epoch": 3.3608150470219433, "grad_norm": 2.5842463970184326, "learning_rate": 6.192987665793332e-07, "loss": 1.2625, "step": 42884 }, { "epoch": 3.3608934169278997, "grad_norm": 3.5311739444732666, "learning_rate": 6.191501270399719e-07, "loss": 1.1478, "step": 42885 }, { "epoch": 3.360971786833856, "grad_norm": 3.4524002075195312, "learning_rate": 6.190015041631136e-07, "loss": 0.9929, "step": 42886 }, { "epoch": 3.361050156739812, "grad_norm": 2.9758434295654297, "learning_rate": 6.18852897949323e-07, "loss": 0.7155, "step": 42887 }, { "epoch": 3.3611285266457678, "grad_norm": 2.9683244228363037, "learning_rate": 6.187043083991661e-07, "loss": 1.2468, "step": 42888 }, { "epoch": 3.361206896551724, "grad_norm": 5.618768215179443, "learning_rate": 6.185557355132071e-07, "loss": 0.8563, "step": 42889 }, { "epoch": 3.3612852664576804, "grad_norm": 4.149430274963379, "learning_rate": 6.184071792920121e-07, "loss": 1.4058, "step": 42890 }, { "epoch": 3.3613636363636363, "grad_norm": 4.022895336151123, "learning_rate": 6.182586397361456e-07, "loss": 1.3208, "step": 42891 }, { "epoch": 3.3614420062695927, "grad_norm": 3.9860222339630127, "learning_rate": 6.181101168461723e-07, "loss": 1.4918, "step": 42892 }, { "epoch": 3.3615203761755486, "grad_norm": 2.4336462020874023, "learning_rate": 6.179616106226577e-07, "loss": 1.3073, "step": 42893 }, { "epoch": 3.361598746081505, "grad_norm": 3.320232629776001, "learning_rate": 6.178131210661664e-07, "loss": 1.448, "step": 42894 }, { "epoch": 3.3616771159874608, "grad_norm": 4.2804155349731445, "learning_rate": 6.17664648177263e-07, "loss": 1.176, "step": 42895 }, { "epoch": 3.361755485893417, "grad_norm": 4.586466312408447, "learning_rate": 6.175161919565126e-07, "loss": 0.9784, "step": 42896 }, { "epoch": 3.361833855799373, "grad_norm": 4.1533331871032715, "learning_rate": 6.173677524044796e-07, "loss": 1.4348, "step": 42897 }, { "epoch": 3.3619122257053293, "grad_norm": 14.900910377502441, "learning_rate": 6.172193295217293e-07, "loss": 1.7185, "step": 42898 }, { "epoch": 3.361990595611285, "grad_norm": 3.47959566116333, "learning_rate": 6.170709233088251e-07, "loss": 1.3579, "step": 42899 }, { "epoch": 3.3620689655172415, "grad_norm": 3.3400611877441406, "learning_rate": 6.169225337663304e-07, "loss": 1.288, "step": 42900 }, { "epoch": 3.3621473354231974, "grad_norm": 2.486250400543213, "learning_rate": 6.167741608948119e-07, "loss": 0.8152, "step": 42901 }, { "epoch": 3.3622257053291538, "grad_norm": 2.6821980476379395, "learning_rate": 6.166258046948348e-07, "loss": 1.3287, "step": 42902 }, { "epoch": 3.3623040752351097, "grad_norm": 2.7779009342193604, "learning_rate": 6.164774651669603e-07, "loss": 1.3704, "step": 42903 }, { "epoch": 3.362382445141066, "grad_norm": 6.64284610748291, "learning_rate": 6.16329142311754e-07, "loss": 1.1423, "step": 42904 }, { "epoch": 3.362460815047022, "grad_norm": 2.4644925594329834, "learning_rate": 6.161808361297789e-07, "loss": 0.9502, "step": 42905 }, { "epoch": 3.362539184952978, "grad_norm": 2.3756062984466553, "learning_rate": 6.160325466216028e-07, "loss": 1.1665, "step": 42906 }, { "epoch": 3.362617554858934, "grad_norm": 4.0024027824401855, "learning_rate": 6.158842737877851e-07, "loss": 1.529, "step": 42907 }, { "epoch": 3.3626959247648904, "grad_norm": 2.0505504608154297, "learning_rate": 6.157360176288929e-07, "loss": 1.2167, "step": 42908 }, { "epoch": 3.3627742946708463, "grad_norm": 3.0636274814605713, "learning_rate": 6.155877781454883e-07, "loss": 1.3535, "step": 42909 }, { "epoch": 3.3628526645768027, "grad_norm": 4.0469970703125, "learning_rate": 6.15439555338136e-07, "loss": 1.9297, "step": 42910 }, { "epoch": 3.3629310344827585, "grad_norm": 2.1488378047943115, "learning_rate": 6.152913492073997e-07, "loss": 1.1917, "step": 42911 }, { "epoch": 3.363009404388715, "grad_norm": 2.940073251724243, "learning_rate": 6.151431597538432e-07, "loss": 1.079, "step": 42912 }, { "epoch": 3.3630877742946708, "grad_norm": 2.0972461700439453, "learning_rate": 6.149949869780303e-07, "loss": 1.4421, "step": 42913 }, { "epoch": 3.363166144200627, "grad_norm": 6.689727306365967, "learning_rate": 6.148468308805228e-07, "loss": 0.7978, "step": 42914 }, { "epoch": 3.363244514106583, "grad_norm": 3.8005175590515137, "learning_rate": 6.146986914618864e-07, "loss": 1.5795, "step": 42915 }, { "epoch": 3.3633228840125393, "grad_norm": 4.2661309242248535, "learning_rate": 6.145505687226833e-07, "loss": 1.7281, "step": 42916 }, { "epoch": 3.363401253918495, "grad_norm": 3.392341136932373, "learning_rate": 6.144024626634787e-07, "loss": 1.1585, "step": 42917 }, { "epoch": 3.3634796238244515, "grad_norm": 13.859919548034668, "learning_rate": 6.142543732848339e-07, "loss": 2.1293, "step": 42918 }, { "epoch": 3.3635579937304074, "grad_norm": 2.901207208633423, "learning_rate": 6.14106300587311e-07, "loss": 0.9778, "step": 42919 }, { "epoch": 3.3636363636363638, "grad_norm": 3.0165750980377197, "learning_rate": 6.139582445714764e-07, "loss": 1.119, "step": 42920 }, { "epoch": 3.3637147335423196, "grad_norm": 3.720654010772705, "learning_rate": 6.13810205237893e-07, "loss": 1.0891, "step": 42921 }, { "epoch": 3.363793103448276, "grad_norm": 3.0969908237457275, "learning_rate": 6.136621825871214e-07, "loss": 1.3278, "step": 42922 }, { "epoch": 3.363871473354232, "grad_norm": 6.966848850250244, "learning_rate": 6.135141766197262e-07, "loss": 1.9518, "step": 42923 }, { "epoch": 3.363949843260188, "grad_norm": 5.751452445983887, "learning_rate": 6.1336618733627e-07, "loss": 0.8193, "step": 42924 }, { "epoch": 3.364028213166144, "grad_norm": 2.7691774368286133, "learning_rate": 6.132182147373156e-07, "loss": 1.2306, "step": 42925 }, { "epoch": 3.3641065830721004, "grad_norm": 3.575312852859497, "learning_rate": 6.13070258823426e-07, "loss": 1.2822, "step": 42926 }, { "epoch": 3.3641849529780563, "grad_norm": 3.4736406803131104, "learning_rate": 6.129223195951633e-07, "loss": 1.3309, "step": 42927 }, { "epoch": 3.3642633228840126, "grad_norm": 6.437507152557373, "learning_rate": 6.127743970530914e-07, "loss": 1.6244, "step": 42928 }, { "epoch": 3.3643416927899685, "grad_norm": 2.268406867980957, "learning_rate": 6.12626491197772e-07, "loss": 1.2737, "step": 42929 }, { "epoch": 3.364420062695925, "grad_norm": 6.548564910888672, "learning_rate": 6.12478602029768e-07, "loss": 1.4654, "step": 42930 }, { "epoch": 3.3644984326018808, "grad_norm": 3.0332062244415283, "learning_rate": 6.123307295496422e-07, "loss": 1.0922, "step": 42931 }, { "epoch": 3.364576802507837, "grad_norm": 4.660112380981445, "learning_rate": 6.121828737579571e-07, "loss": 1.9234, "step": 42932 }, { "epoch": 3.364655172413793, "grad_norm": 4.990353584289551, "learning_rate": 6.120350346552728e-07, "loss": 1.5026, "step": 42933 }, { "epoch": 3.3647335423197493, "grad_norm": 5.8777899742126465, "learning_rate": 6.118872122421543e-07, "loss": 1.7121, "step": 42934 }, { "epoch": 3.364811912225705, "grad_norm": 2.66373872756958, "learning_rate": 6.117394065191634e-07, "loss": 1.261, "step": 42935 }, { "epoch": 3.3648902821316615, "grad_norm": 20.07610511779785, "learning_rate": 6.115916174868624e-07, "loss": 1.5216, "step": 42936 }, { "epoch": 3.3649686520376174, "grad_norm": 5.961938381195068, "learning_rate": 6.114438451458122e-07, "loss": 1.472, "step": 42937 }, { "epoch": 3.3650470219435737, "grad_norm": 6.65069055557251, "learning_rate": 6.112960894965741e-07, "loss": 0.8702, "step": 42938 }, { "epoch": 3.3651253918495296, "grad_norm": 2.5271973609924316, "learning_rate": 6.111483505397136e-07, "loss": 0.8721, "step": 42939 }, { "epoch": 3.365203761755486, "grad_norm": 3.686298370361328, "learning_rate": 6.110006282757897e-07, "loss": 1.0381, "step": 42940 }, { "epoch": 3.365282131661442, "grad_norm": 2.4139442443847656, "learning_rate": 6.10852922705365e-07, "loss": 0.7847, "step": 42941 }, { "epoch": 3.365360501567398, "grad_norm": 2.6276566982269287, "learning_rate": 6.107052338290015e-07, "loss": 1.0749, "step": 42942 }, { "epoch": 3.365438871473354, "grad_norm": 5.984017372131348, "learning_rate": 6.105575616472609e-07, "loss": 1.2059, "step": 42943 }, { "epoch": 3.3655172413793104, "grad_norm": 5.583359718322754, "learning_rate": 6.104099061607044e-07, "loss": 1.2881, "step": 42944 }, { "epoch": 3.3655956112852663, "grad_norm": 3.3296523094177246, "learning_rate": 6.102622673698943e-07, "loss": 1.2281, "step": 42945 }, { "epoch": 3.3656739811912226, "grad_norm": 4.782031059265137, "learning_rate": 6.101146452753914e-07, "loss": 1.4634, "step": 42946 }, { "epoch": 3.3657523510971785, "grad_norm": 3.137458324432373, "learning_rate": 6.09967039877758e-07, "loss": 1.4584, "step": 42947 }, { "epoch": 3.365830721003135, "grad_norm": 4.696215629577637, "learning_rate": 6.098194511775552e-07, "loss": 1.1613, "step": 42948 }, { "epoch": 3.3659090909090907, "grad_norm": 5.649737358093262, "learning_rate": 6.096718791753437e-07, "loss": 1.1663, "step": 42949 }, { "epoch": 3.365987460815047, "grad_norm": 8.238219261169434, "learning_rate": 6.095243238716869e-07, "loss": 1.7125, "step": 42950 }, { "epoch": 3.366065830721003, "grad_norm": 2.0118048191070557, "learning_rate": 6.093767852671434e-07, "loss": 1.22, "step": 42951 }, { "epoch": 3.3661442006269593, "grad_norm": 2.767134428024292, "learning_rate": 6.092292633622743e-07, "loss": 1.3629, "step": 42952 }, { "epoch": 3.366222570532915, "grad_norm": 2.4867799282073975, "learning_rate": 6.090817581576425e-07, "loss": 1.078, "step": 42953 }, { "epoch": 3.3663009404388715, "grad_norm": 3.327704906463623, "learning_rate": 6.089342696538103e-07, "loss": 1.4691, "step": 42954 }, { "epoch": 3.3663793103448274, "grad_norm": 4.026520729064941, "learning_rate": 6.087867978513351e-07, "loss": 1.4605, "step": 42955 }, { "epoch": 3.3664576802507837, "grad_norm": 2.7804012298583984, "learning_rate": 6.086393427507792e-07, "loss": 1.075, "step": 42956 }, { "epoch": 3.3665360501567396, "grad_norm": 2.7250325679779053, "learning_rate": 6.084919043527032e-07, "loss": 1.5051, "step": 42957 }, { "epoch": 3.366614420062696, "grad_norm": 15.651318550109863, "learning_rate": 6.083444826576701e-07, "loss": 1.3409, "step": 42958 }, { "epoch": 3.366692789968652, "grad_norm": 5.753413677215576, "learning_rate": 6.081970776662383e-07, "loss": 1.0215, "step": 42959 }, { "epoch": 3.366771159874608, "grad_norm": 4.264568328857422, "learning_rate": 6.080496893789683e-07, "loss": 1.6059, "step": 42960 }, { "epoch": 3.3668495297805645, "grad_norm": 3.2171735763549805, "learning_rate": 6.079023177964222e-07, "loss": 0.9416, "step": 42961 }, { "epoch": 3.3669278996865204, "grad_norm": 4.128721237182617, "learning_rate": 6.077549629191592e-07, "loss": 1.5722, "step": 42962 }, { "epoch": 3.3670062695924763, "grad_norm": 2.934885025024414, "learning_rate": 6.076076247477408e-07, "loss": 1.4981, "step": 42963 }, { "epoch": 3.3670846394984326, "grad_norm": 2.6887733936309814, "learning_rate": 6.074603032827265e-07, "loss": 1.3326, "step": 42964 }, { "epoch": 3.367163009404389, "grad_norm": 5.022424697875977, "learning_rate": 6.073129985246784e-07, "loss": 1.4229, "step": 42965 }, { "epoch": 3.367241379310345, "grad_norm": 8.416577339172363, "learning_rate": 6.071657104741535e-07, "loss": 1.0487, "step": 42966 }, { "epoch": 3.3673197492163007, "grad_norm": 4.067591667175293, "learning_rate": 6.070184391317146e-07, "loss": 1.1151, "step": 42967 }, { "epoch": 3.367398119122257, "grad_norm": 2.9899933338165283, "learning_rate": 6.068711844979214e-07, "loss": 1.4683, "step": 42968 }, { "epoch": 3.3674764890282134, "grad_norm": 4.445436000823975, "learning_rate": 6.067239465733344e-07, "loss": 1.3933, "step": 42969 }, { "epoch": 3.3675548589341693, "grad_norm": 2.6271250247955322, "learning_rate": 6.065767253585125e-07, "loss": 1.3582, "step": 42970 }, { "epoch": 3.367633228840125, "grad_norm": 3.4552414417266846, "learning_rate": 6.064295208540144e-07, "loss": 1.4578, "step": 42971 }, { "epoch": 3.3677115987460815, "grad_norm": 2.868319511413574, "learning_rate": 6.062823330604034e-07, "loss": 1.3584, "step": 42972 }, { "epoch": 3.367789968652038, "grad_norm": 2.413433790206909, "learning_rate": 6.061351619782385e-07, "loss": 1.3881, "step": 42973 }, { "epoch": 3.3678683385579937, "grad_norm": 4.347942352294922, "learning_rate": 6.059880076080777e-07, "loss": 1.5137, "step": 42974 }, { "epoch": 3.3679467084639496, "grad_norm": 3.962409257888794, "learning_rate": 6.058408699504814e-07, "loss": 1.3177, "step": 42975 }, { "epoch": 3.368025078369906, "grad_norm": 2.6160407066345215, "learning_rate": 6.056937490060083e-07, "loss": 1.2821, "step": 42976 }, { "epoch": 3.3681034482758623, "grad_norm": 3.55989146232605, "learning_rate": 6.055466447752211e-07, "loss": 1.1364, "step": 42977 }, { "epoch": 3.368181818181818, "grad_norm": 4.55501127243042, "learning_rate": 6.053995572586768e-07, "loss": 1.6403, "step": 42978 }, { "epoch": 3.368260188087774, "grad_norm": 9.73528003692627, "learning_rate": 6.052524864569354e-07, "loss": 1.2447, "step": 42979 }, { "epoch": 3.3683385579937304, "grad_norm": 2.3616952896118164, "learning_rate": 6.051054323705558e-07, "loss": 1.3529, "step": 42980 }, { "epoch": 3.3684169278996867, "grad_norm": 3.1006269454956055, "learning_rate": 6.049583950000976e-07, "loss": 1.3617, "step": 42981 }, { "epoch": 3.3684952978056426, "grad_norm": 4.144549369812012, "learning_rate": 6.048113743461209e-07, "loss": 1.3149, "step": 42982 }, { "epoch": 3.368573667711599, "grad_norm": 3.5172038078308105, "learning_rate": 6.046643704091837e-07, "loss": 1.3023, "step": 42983 }, { "epoch": 3.368652037617555, "grad_norm": 3.525261878967285, "learning_rate": 6.04517383189846e-07, "loss": 0.6595, "step": 42984 }, { "epoch": 3.368730407523511, "grad_norm": 6.373071193695068, "learning_rate": 6.043704126886663e-07, "loss": 1.4979, "step": 42985 }, { "epoch": 3.368808777429467, "grad_norm": 5.592298984527588, "learning_rate": 6.042234589062041e-07, "loss": 1.4045, "step": 42986 }, { "epoch": 3.3688871473354234, "grad_norm": 6.359563827514648, "learning_rate": 6.040765218430178e-07, "loss": 1.4184, "step": 42987 }, { "epoch": 3.3689655172413793, "grad_norm": 7.235567569732666, "learning_rate": 6.039296014996677e-07, "loss": 0.7163, "step": 42988 }, { "epoch": 3.3690438871473356, "grad_norm": 3.860809803009033, "learning_rate": 6.037826978767098e-07, "loss": 1.2571, "step": 42989 }, { "epoch": 3.3691222570532915, "grad_norm": 3.2524001598358154, "learning_rate": 6.036358109747042e-07, "loss": 1.6167, "step": 42990 }, { "epoch": 3.369200626959248, "grad_norm": 3.2790753841400146, "learning_rate": 6.034889407942107e-07, "loss": 1.3733, "step": 42991 }, { "epoch": 3.3692789968652037, "grad_norm": 4.235622406005859, "learning_rate": 6.033420873357876e-07, "loss": 1.4286, "step": 42992 }, { "epoch": 3.36935736677116, "grad_norm": 3.46504282951355, "learning_rate": 6.031952505999927e-07, "loss": 1.397, "step": 42993 }, { "epoch": 3.369435736677116, "grad_norm": 4.319262981414795, "learning_rate": 6.03048430587384e-07, "loss": 1.4675, "step": 42994 }, { "epoch": 3.3695141065830723, "grad_norm": 4.459314346313477, "learning_rate": 6.02901627298521e-07, "loss": 1.0917, "step": 42995 }, { "epoch": 3.369592476489028, "grad_norm": 2.735258102416992, "learning_rate": 6.027548407339617e-07, "loss": 1.3667, "step": 42996 }, { "epoch": 3.3696708463949845, "grad_norm": 3.569920301437378, "learning_rate": 6.026080708942645e-07, "loss": 1.2822, "step": 42997 }, { "epoch": 3.3697492163009404, "grad_norm": 7.0429606437683105, "learning_rate": 6.024613177799877e-07, "loss": 1.2276, "step": 42998 }, { "epoch": 3.3698275862068967, "grad_norm": 6.0345845222473145, "learning_rate": 6.023145813916892e-07, "loss": 1.3653, "step": 42999 }, { "epoch": 3.3699059561128526, "grad_norm": 11.867241859436035, "learning_rate": 6.021678617299271e-07, "loss": 1.2059, "step": 43000 }, { "epoch": 3.369984326018809, "grad_norm": 3.885258674621582, "learning_rate": 6.020211587952595e-07, "loss": 1.4092, "step": 43001 }, { "epoch": 3.370062695924765, "grad_norm": 3.4860072135925293, "learning_rate": 6.018744725882447e-07, "loss": 1.4683, "step": 43002 }, { "epoch": 3.370141065830721, "grad_norm": 9.090187072753906, "learning_rate": 6.017278031094403e-07, "loss": 1.1171, "step": 43003 }, { "epoch": 3.370219435736677, "grad_norm": 2.395160436630249, "learning_rate": 6.015811503594043e-07, "loss": 1.2154, "step": 43004 }, { "epoch": 3.3702978056426334, "grad_norm": 7.261306285858154, "learning_rate": 6.014345143386941e-07, "loss": 1.1961, "step": 43005 }, { "epoch": 3.3703761755485893, "grad_norm": 5.936720371246338, "learning_rate": 6.012878950478695e-07, "loss": 2.0834, "step": 43006 }, { "epoch": 3.3704545454545456, "grad_norm": 5.779353141784668, "learning_rate": 6.011412924874848e-07, "loss": 1.0818, "step": 43007 }, { "epoch": 3.3705329153605015, "grad_norm": 7.212975978851318, "learning_rate": 6.009947066580984e-07, "loss": 0.7468, "step": 43008 }, { "epoch": 3.370611285266458, "grad_norm": 3.0217514038085938, "learning_rate": 6.008481375602693e-07, "loss": 1.0896, "step": 43009 }, { "epoch": 3.3706896551724137, "grad_norm": 3.0089523792266846, "learning_rate": 6.007015851945558e-07, "loss": 1.3475, "step": 43010 }, { "epoch": 3.37076802507837, "grad_norm": 4.748866081237793, "learning_rate": 6.005550495615131e-07, "loss": 1.5499, "step": 43011 }, { "epoch": 3.370846394984326, "grad_norm": 13.256717681884766, "learning_rate": 6.004085306616986e-07, "loss": 0.6871, "step": 43012 }, { "epoch": 3.3709247648902823, "grad_norm": 7.671933174133301, "learning_rate": 6.002620284956695e-07, "loss": 1.0541, "step": 43013 }, { "epoch": 3.371003134796238, "grad_norm": 5.971037864685059, "learning_rate": 6.00115543063986e-07, "loss": 0.7464, "step": 43014 }, { "epoch": 3.3710815047021945, "grad_norm": 6.323080062866211, "learning_rate": 5.99969074367202e-07, "loss": 1.6418, "step": 43015 }, { "epoch": 3.3711598746081504, "grad_norm": 2.771544933319092, "learning_rate": 5.998226224058756e-07, "loss": 1.3719, "step": 43016 }, { "epoch": 3.3712382445141067, "grad_norm": 6.047177791595459, "learning_rate": 5.99676187180564e-07, "loss": 0.8606, "step": 43017 }, { "epoch": 3.3713166144200626, "grad_norm": 3.1482388973236084, "learning_rate": 5.995297686918239e-07, "loss": 1.0585, "step": 43018 }, { "epoch": 3.371394984326019, "grad_norm": 4.3744096755981445, "learning_rate": 5.993833669402121e-07, "loss": 1.3132, "step": 43019 }, { "epoch": 3.371473354231975, "grad_norm": 3.7420177459716797, "learning_rate": 5.992369819262861e-07, "loss": 1.5394, "step": 43020 }, { "epoch": 3.371551724137931, "grad_norm": 3.411668539047241, "learning_rate": 5.990906136506031e-07, "loss": 1.1436, "step": 43021 }, { "epoch": 3.371630094043887, "grad_norm": 3.003025531768799, "learning_rate": 5.989442621137171e-07, "loss": 1.6349, "step": 43022 }, { "epoch": 3.3717084639498434, "grad_norm": 2.5170881748199463, "learning_rate": 5.98797927316187e-07, "loss": 1.4034, "step": 43023 }, { "epoch": 3.3717868338557992, "grad_norm": 6.338194847106934, "learning_rate": 5.986516092585698e-07, "loss": 0.6587, "step": 43024 }, { "epoch": 3.3718652037617556, "grad_norm": 2.5059516429901123, "learning_rate": 5.985053079414216e-07, "loss": 1.2055, "step": 43025 }, { "epoch": 3.3719435736677115, "grad_norm": 2.6291661262512207, "learning_rate": 5.983590233652975e-07, "loss": 1.2291, "step": 43026 }, { "epoch": 3.372021943573668, "grad_norm": 5.304281234741211, "learning_rate": 5.982127555307543e-07, "loss": 0.9724, "step": 43027 }, { "epoch": 3.3721003134796237, "grad_norm": 4.282003879547119, "learning_rate": 5.980665044383494e-07, "loss": 1.52, "step": 43028 }, { "epoch": 3.37217868338558, "grad_norm": 3.6775286197662354, "learning_rate": 5.9792027008864e-07, "loss": 1.5737, "step": 43029 }, { "epoch": 3.372257053291536, "grad_norm": 3.063612937927246, "learning_rate": 5.977740524821796e-07, "loss": 1.4436, "step": 43030 }, { "epoch": 3.3723354231974922, "grad_norm": 3.7739222049713135, "learning_rate": 5.976278516195255e-07, "loss": 1.4053, "step": 43031 }, { "epoch": 3.372413793103448, "grad_norm": 6.268230438232422, "learning_rate": 5.974816675012329e-07, "loss": 1.0948, "step": 43032 }, { "epoch": 3.3724921630094045, "grad_norm": 2.938288927078247, "learning_rate": 5.973355001278608e-07, "loss": 1.6307, "step": 43033 }, { "epoch": 3.3725705329153604, "grad_norm": 4.106359481811523, "learning_rate": 5.97189349499962e-07, "loss": 1.1742, "step": 43034 }, { "epoch": 3.3726489028213167, "grad_norm": 2.8091588020324707, "learning_rate": 5.970432156180933e-07, "loss": 1.0527, "step": 43035 }, { "epoch": 3.3727272727272726, "grad_norm": 2.813035726547241, "learning_rate": 5.96897098482811e-07, "loss": 1.2658, "step": 43036 }, { "epoch": 3.372805642633229, "grad_norm": 3.4023475646972656, "learning_rate": 5.967509980946706e-07, "loss": 1.3643, "step": 43037 }, { "epoch": 3.372884012539185, "grad_norm": 2.7017009258270264, "learning_rate": 5.966049144542274e-07, "loss": 1.5103, "step": 43038 }, { "epoch": 3.372962382445141, "grad_norm": 2.745309352874756, "learning_rate": 5.964588475620381e-07, "loss": 1.1941, "step": 43039 }, { "epoch": 3.373040752351097, "grad_norm": 2.7022602558135986, "learning_rate": 5.963127974186577e-07, "loss": 1.2333, "step": 43040 }, { "epoch": 3.3731191222570533, "grad_norm": 1.8065664768218994, "learning_rate": 5.961667640246399e-07, "loss": 1.2805, "step": 43041 }, { "epoch": 3.3731974921630092, "grad_norm": 4.528514862060547, "learning_rate": 5.96020747380543e-07, "loss": 0.8621, "step": 43042 }, { "epoch": 3.3732758620689656, "grad_norm": 3.340733528137207, "learning_rate": 5.958747474869209e-07, "loss": 1.5856, "step": 43043 }, { "epoch": 3.3733542319749215, "grad_norm": 4.136139392852783, "learning_rate": 5.957287643443305e-07, "loss": 1.5457, "step": 43044 }, { "epoch": 3.373432601880878, "grad_norm": 4.65685510635376, "learning_rate": 5.955827979533241e-07, "loss": 1.303, "step": 43045 }, { "epoch": 3.3735109717868337, "grad_norm": 4.233944416046143, "learning_rate": 5.954368483144579e-07, "loss": 1.6517, "step": 43046 }, { "epoch": 3.37358934169279, "grad_norm": 2.4144420623779297, "learning_rate": 5.952909154282888e-07, "loss": 1.2949, "step": 43047 }, { "epoch": 3.373667711598746, "grad_norm": 3.7247815132141113, "learning_rate": 5.951449992953712e-07, "loss": 1.1146, "step": 43048 }, { "epoch": 3.3737460815047022, "grad_norm": 3.8078675270080566, "learning_rate": 5.949990999162585e-07, "loss": 1.6005, "step": 43049 }, { "epoch": 3.373824451410658, "grad_norm": 5.097650527954102, "learning_rate": 5.948532172915073e-07, "loss": 1.3193, "step": 43050 }, { "epoch": 3.3739028213166145, "grad_norm": 9.758166313171387, "learning_rate": 5.947073514216711e-07, "loss": 1.7957, "step": 43051 }, { "epoch": 3.3739811912225703, "grad_norm": 3.185528516769409, "learning_rate": 5.945615023073059e-07, "loss": 1.0023, "step": 43052 }, { "epoch": 3.3740595611285267, "grad_norm": 2.2020950317382812, "learning_rate": 5.944156699489656e-07, "loss": 1.2722, "step": 43053 }, { "epoch": 3.3741379310344826, "grad_norm": 3.1297101974487305, "learning_rate": 5.942698543472059e-07, "loss": 0.9351, "step": 43054 }, { "epoch": 3.374216300940439, "grad_norm": 3.450375556945801, "learning_rate": 5.941240555025801e-07, "loss": 1.3555, "step": 43055 }, { "epoch": 3.3742946708463952, "grad_norm": 4.570550918579102, "learning_rate": 5.939782734156435e-07, "loss": 1.8594, "step": 43056 }, { "epoch": 3.374373040752351, "grad_norm": 2.807844877243042, "learning_rate": 5.938325080869506e-07, "loss": 1.4151, "step": 43057 }, { "epoch": 3.374451410658307, "grad_norm": 3.4212701320648193, "learning_rate": 5.936867595170559e-07, "loss": 1.3839, "step": 43058 }, { "epoch": 3.3745297805642633, "grad_norm": 4.094933032989502, "learning_rate": 5.93541027706514e-07, "loss": 0.6776, "step": 43059 }, { "epoch": 3.3746081504702197, "grad_norm": 2.486355781555176, "learning_rate": 5.933953126558772e-07, "loss": 1.4568, "step": 43060 }, { "epoch": 3.3746865203761756, "grad_norm": 3.1860759258270264, "learning_rate": 5.932496143657018e-07, "loss": 0.6564, "step": 43061 }, { "epoch": 3.3747648902821314, "grad_norm": 13.11225414276123, "learning_rate": 5.931039328365429e-07, "loss": 1.6279, "step": 43062 }, { "epoch": 3.374843260188088, "grad_norm": 3.689417600631714, "learning_rate": 5.929582680689522e-07, "loss": 1.0233, "step": 43063 }, { "epoch": 3.374921630094044, "grad_norm": 3.8349671363830566, "learning_rate": 5.928126200634843e-07, "loss": 1.3412, "step": 43064 }, { "epoch": 3.375, "grad_norm": 4.325695037841797, "learning_rate": 5.926669888206926e-07, "loss": 1.1743, "step": 43065 }, { "epoch": 3.375078369905956, "grad_norm": 3.310771942138672, "learning_rate": 5.92521374341134e-07, "loss": 1.3524, "step": 43066 }, { "epoch": 3.375156739811912, "grad_norm": 3.9333159923553467, "learning_rate": 5.923757766253591e-07, "loss": 1.5254, "step": 43067 }, { "epoch": 3.3752351097178686, "grad_norm": 2.2105345726013184, "learning_rate": 5.922301956739229e-07, "loss": 1.288, "step": 43068 }, { "epoch": 3.3753134796238244, "grad_norm": 7.0883026123046875, "learning_rate": 5.92084631487379e-07, "loss": 1.5748, "step": 43069 }, { "epoch": 3.3753918495297803, "grad_norm": 2.0546860694885254, "learning_rate": 5.919390840662814e-07, "loss": 1.4193, "step": 43070 }, { "epoch": 3.3754702194357367, "grad_norm": 3.110198736190796, "learning_rate": 5.917935534111829e-07, "loss": 1.3155, "step": 43071 }, { "epoch": 3.375548589341693, "grad_norm": 10.568603515625, "learning_rate": 5.916480395226381e-07, "loss": 1.2243, "step": 43072 }, { "epoch": 3.375626959247649, "grad_norm": 3.7425553798675537, "learning_rate": 5.915025424011995e-07, "loss": 1.4221, "step": 43073 }, { "epoch": 3.3757053291536048, "grad_norm": 7.918835639953613, "learning_rate": 5.913570620474207e-07, "loss": 0.6362, "step": 43074 }, { "epoch": 3.375783699059561, "grad_norm": 4.608124732971191, "learning_rate": 5.912115984618555e-07, "loss": 1.3129, "step": 43075 }, { "epoch": 3.3758620689655174, "grad_norm": 15.256759643554688, "learning_rate": 5.910661516450572e-07, "loss": 1.2272, "step": 43076 }, { "epoch": 3.3759404388714733, "grad_norm": 3.4605443477630615, "learning_rate": 5.909207215975793e-07, "loss": 0.9807, "step": 43077 }, { "epoch": 3.3760188087774297, "grad_norm": 7.354426383972168, "learning_rate": 5.907753083199735e-07, "loss": 1.2882, "step": 43078 }, { "epoch": 3.3760971786833855, "grad_norm": 3.543820381164551, "learning_rate": 5.906299118127923e-07, "loss": 1.0842, "step": 43079 }, { "epoch": 3.376175548589342, "grad_norm": 4.409649848937988, "learning_rate": 5.904845320765917e-07, "loss": 1.2423, "step": 43080 }, { "epoch": 3.3762539184952978, "grad_norm": 2.729663848876953, "learning_rate": 5.903391691119236e-07, "loss": 1.0599, "step": 43081 }, { "epoch": 3.376332288401254, "grad_norm": 3.526811122894287, "learning_rate": 5.901938229193399e-07, "loss": 1.6538, "step": 43082 }, { "epoch": 3.37641065830721, "grad_norm": 8.50029468536377, "learning_rate": 5.900484934993933e-07, "loss": 1.2768, "step": 43083 }, { "epoch": 3.3764890282131663, "grad_norm": 2.118067502975464, "learning_rate": 5.899031808526362e-07, "loss": 1.1414, "step": 43084 }, { "epoch": 3.376567398119122, "grad_norm": 6.4845685958862305, "learning_rate": 5.897578849796248e-07, "loss": 1.095, "step": 43085 }, { "epoch": 3.3766457680250785, "grad_norm": 3.018723487854004, "learning_rate": 5.896126058809076e-07, "loss": 1.1524, "step": 43086 }, { "epoch": 3.3767241379310344, "grad_norm": 3.1117234230041504, "learning_rate": 5.894673435570392e-07, "loss": 1.2061, "step": 43087 }, { "epoch": 3.3768025078369908, "grad_norm": 2.708347797393799, "learning_rate": 5.893220980085713e-07, "loss": 1.5259, "step": 43088 }, { "epoch": 3.3768808777429467, "grad_norm": 5.365903377532959, "learning_rate": 5.891768692360567e-07, "loss": 1.1791, "step": 43089 }, { "epoch": 3.376959247648903, "grad_norm": 3.316833257675171, "learning_rate": 5.890316572400478e-07, "loss": 0.5682, "step": 43090 }, { "epoch": 3.377037617554859, "grad_norm": 4.498588562011719, "learning_rate": 5.888864620210965e-07, "loss": 1.1224, "step": 43091 }, { "epoch": 3.377115987460815, "grad_norm": 2.922172784805298, "learning_rate": 5.887412835797557e-07, "loss": 1.1526, "step": 43092 }, { "epoch": 3.377194357366771, "grad_norm": 3.2547709941864014, "learning_rate": 5.885961219165776e-07, "loss": 1.1906, "step": 43093 }, { "epoch": 3.3772727272727274, "grad_norm": 3.5882351398468018, "learning_rate": 5.884509770321134e-07, "loss": 0.9773, "step": 43094 }, { "epoch": 3.3773510971786833, "grad_norm": 3.875713586807251, "learning_rate": 5.883058489269161e-07, "loss": 1.5542, "step": 43095 }, { "epoch": 3.3774294670846396, "grad_norm": 4.167635440826416, "learning_rate": 5.881607376015386e-07, "loss": 1.3693, "step": 43096 }, { "epoch": 3.3775078369905955, "grad_norm": 2.5220541954040527, "learning_rate": 5.880156430565293e-07, "loss": 1.1004, "step": 43097 }, { "epoch": 3.377586206896552, "grad_norm": 2.0300636291503906, "learning_rate": 5.878705652924433e-07, "loss": 1.1094, "step": 43098 }, { "epoch": 3.3776645768025078, "grad_norm": 3.747685194015503, "learning_rate": 5.877255043098318e-07, "loss": 1.3898, "step": 43099 }, { "epoch": 3.377742946708464, "grad_norm": 3.4401628971099854, "learning_rate": 5.875804601092467e-07, "loss": 0.9077, "step": 43100 }, { "epoch": 3.37782131661442, "grad_norm": 3.8270351886749268, "learning_rate": 5.874354326912385e-07, "loss": 1.2651, "step": 43101 }, { "epoch": 3.3778996865203763, "grad_norm": 3.196937322616577, "learning_rate": 5.872904220563586e-07, "loss": 1.1267, "step": 43102 }, { "epoch": 3.377978056426332, "grad_norm": 7.742112636566162, "learning_rate": 5.871454282051597e-07, "loss": 1.4564, "step": 43103 }, { "epoch": 3.3780564263322885, "grad_norm": 3.1213576793670654, "learning_rate": 5.870004511381949e-07, "loss": 1.2967, "step": 43104 }, { "epoch": 3.3781347962382444, "grad_norm": 2.90838623046875, "learning_rate": 5.86855490856012e-07, "loss": 0.9001, "step": 43105 }, { "epoch": 3.3782131661442008, "grad_norm": 3.933138847351074, "learning_rate": 5.867105473591645e-07, "loss": 0.9729, "step": 43106 }, { "epoch": 3.3782915360501566, "grad_norm": 3.4217209815979004, "learning_rate": 5.865656206482034e-07, "loss": 1.2408, "step": 43107 }, { "epoch": 3.378369905956113, "grad_norm": 5.910355567932129, "learning_rate": 5.864207107236797e-07, "loss": 1.6795, "step": 43108 }, { "epoch": 3.378448275862069, "grad_norm": 3.4973511695861816, "learning_rate": 5.862758175861444e-07, "loss": 1.1174, "step": 43109 }, { "epoch": 3.378526645768025, "grad_norm": 3.3729348182678223, "learning_rate": 5.861309412361488e-07, "loss": 1.3205, "step": 43110 }, { "epoch": 3.378605015673981, "grad_norm": 4.529863357543945, "learning_rate": 5.859860816742447e-07, "loss": 1.5066, "step": 43111 }, { "epoch": 3.3786833855799374, "grad_norm": 3.9258127212524414, "learning_rate": 5.858412389009816e-07, "loss": 1.3637, "step": 43112 }, { "epoch": 3.3787617554858933, "grad_norm": 2.8382391929626465, "learning_rate": 5.856964129169112e-07, "loss": 0.9781, "step": 43113 }, { "epoch": 3.3788401253918496, "grad_norm": 3.1883747577667236, "learning_rate": 5.85551603722585e-07, "loss": 1.6518, "step": 43114 }, { "epoch": 3.3789184952978055, "grad_norm": 5.9064788818359375, "learning_rate": 5.854068113185535e-07, "loss": 1.2051, "step": 43115 }, { "epoch": 3.378996865203762, "grad_norm": 3.0169408321380615, "learning_rate": 5.852620357053651e-07, "loss": 1.5128, "step": 43116 }, { "epoch": 3.3790752351097177, "grad_norm": 5.232921123504639, "learning_rate": 5.851172768835733e-07, "loss": 0.8735, "step": 43117 }, { "epoch": 3.379153605015674, "grad_norm": 3.4111251831054688, "learning_rate": 5.849725348537289e-07, "loss": 1.4372, "step": 43118 }, { "epoch": 3.37923197492163, "grad_norm": 3.1763641834259033, "learning_rate": 5.848278096163807e-07, "loss": 1.3966, "step": 43119 }, { "epoch": 3.3793103448275863, "grad_norm": 6.694531440734863, "learning_rate": 5.846831011720789e-07, "loss": 1.4172, "step": 43120 }, { "epoch": 3.379388714733542, "grad_norm": 4.819771766662598, "learning_rate": 5.84538409521374e-07, "loss": 1.1787, "step": 43121 }, { "epoch": 3.3794670846394985, "grad_norm": 2.5787110328674316, "learning_rate": 5.843937346648193e-07, "loss": 1.3644, "step": 43122 }, { "epoch": 3.3795454545454544, "grad_norm": 2.3597252368927, "learning_rate": 5.842490766029618e-07, "loss": 1.2933, "step": 43123 }, { "epoch": 3.3796238244514107, "grad_norm": 3.4720358848571777, "learning_rate": 5.841044353363529e-07, "loss": 1.2566, "step": 43124 }, { "epoch": 3.3797021943573666, "grad_norm": 2.630420684814453, "learning_rate": 5.839598108655419e-07, "loss": 1.1796, "step": 43125 }, { "epoch": 3.379780564263323, "grad_norm": 2.8000972270965576, "learning_rate": 5.838152031910804e-07, "loss": 1.2266, "step": 43126 }, { "epoch": 3.379858934169279, "grad_norm": 4.635356426239014, "learning_rate": 5.836706123135172e-07, "loss": 1.1376, "step": 43127 }, { "epoch": 3.379937304075235, "grad_norm": 2.371964693069458, "learning_rate": 5.835260382334024e-07, "loss": 1.213, "step": 43128 }, { "epoch": 3.380015673981191, "grad_norm": 2.720306634902954, "learning_rate": 5.833814809512861e-07, "loss": 0.8111, "step": 43129 }, { "epoch": 3.3800940438871474, "grad_norm": 4.391451835632324, "learning_rate": 5.832369404677185e-07, "loss": 0.8398, "step": 43130 }, { "epoch": 3.3801724137931033, "grad_norm": 2.3909988403320312, "learning_rate": 5.830924167832486e-07, "loss": 1.2274, "step": 43131 }, { "epoch": 3.3802507836990596, "grad_norm": 9.476167678833008, "learning_rate": 5.82947909898427e-07, "loss": 1.0316, "step": 43132 }, { "epoch": 3.3803291536050155, "grad_norm": 4.316798210144043, "learning_rate": 5.828034198138033e-07, "loss": 1.4959, "step": 43133 }, { "epoch": 3.380407523510972, "grad_norm": 3.7618660926818848, "learning_rate": 5.826589465299259e-07, "loss": 1.5617, "step": 43134 }, { "epoch": 3.3804858934169277, "grad_norm": 3.6997697353363037, "learning_rate": 5.825144900473439e-07, "loss": 1.3427, "step": 43135 }, { "epoch": 3.380564263322884, "grad_norm": 3.6842310428619385, "learning_rate": 5.823700503666086e-07, "loss": 1.1393, "step": 43136 }, { "epoch": 3.38064263322884, "grad_norm": 3.9816863536834717, "learning_rate": 5.822256274882698e-07, "loss": 1.6824, "step": 43137 }, { "epoch": 3.3807210031347963, "grad_norm": 5.731081008911133, "learning_rate": 5.820812214128746e-07, "loss": 1.7622, "step": 43138 }, { "epoch": 3.380799373040752, "grad_norm": 2.529297351837158, "learning_rate": 5.819368321409735e-07, "loss": 1.3645, "step": 43139 }, { "epoch": 3.3808777429467085, "grad_norm": 3.878521203994751, "learning_rate": 5.817924596731139e-07, "loss": 1.1451, "step": 43140 }, { "epoch": 3.3809561128526644, "grad_norm": 9.225149154663086, "learning_rate": 5.816481040098487e-07, "loss": 1.3657, "step": 43141 }, { "epoch": 3.3810344827586207, "grad_norm": 5.111140727996826, "learning_rate": 5.815037651517236e-07, "loss": 1.7602, "step": 43142 }, { "epoch": 3.3811128526645766, "grad_norm": 3.0151875019073486, "learning_rate": 5.813594430992892e-07, "loss": 1.4718, "step": 43143 }, { "epoch": 3.381191222570533, "grad_norm": 3.7889132499694824, "learning_rate": 5.812151378530934e-07, "loss": 1.3473, "step": 43144 }, { "epoch": 3.381269592476489, "grad_norm": 4.19467306137085, "learning_rate": 5.810708494136858e-07, "loss": 1.1422, "step": 43145 }, { "epoch": 3.381347962382445, "grad_norm": 4.307193279266357, "learning_rate": 5.809265777816148e-07, "loss": 1.1733, "step": 43146 }, { "epoch": 3.381426332288401, "grad_norm": 7.6745781898498535, "learning_rate": 5.807823229574289e-07, "loss": 0.8568, "step": 43147 }, { "epoch": 3.3815047021943574, "grad_norm": 2.509159564971924, "learning_rate": 5.806380849416787e-07, "loss": 1.0543, "step": 43148 }, { "epoch": 3.3815830721003133, "grad_norm": 2.8944056034088135, "learning_rate": 5.804938637349095e-07, "loss": 1.0717, "step": 43149 }, { "epoch": 3.3816614420062696, "grad_norm": 2.5902278423309326, "learning_rate": 5.803496593376722e-07, "loss": 1.3689, "step": 43150 }, { "epoch": 3.381739811912226, "grad_norm": 3.4525749683380127, "learning_rate": 5.802054717505151e-07, "loss": 1.4423, "step": 43151 }, { "epoch": 3.381818181818182, "grad_norm": 4.649678707122803, "learning_rate": 5.80061300973987e-07, "loss": 1.7692, "step": 43152 }, { "epoch": 3.3818965517241377, "grad_norm": 3.7207651138305664, "learning_rate": 5.799171470086345e-07, "loss": 1.5672, "step": 43153 }, { "epoch": 3.381974921630094, "grad_norm": 3.625152826309204, "learning_rate": 5.797730098550064e-07, "loss": 1.2583, "step": 43154 }, { "epoch": 3.3820532915360504, "grad_norm": 2.872265338897705, "learning_rate": 5.796288895136515e-07, "loss": 1.3238, "step": 43155 }, { "epoch": 3.3821316614420063, "grad_norm": 5.381523132324219, "learning_rate": 5.794847859851193e-07, "loss": 1.4895, "step": 43156 }, { "epoch": 3.382210031347962, "grad_norm": 3.8227553367614746, "learning_rate": 5.793406992699557e-07, "loss": 1.3037, "step": 43157 }, { "epoch": 3.3822884012539185, "grad_norm": 2.739759683609009, "learning_rate": 5.79196629368709e-07, "loss": 1.769, "step": 43158 }, { "epoch": 3.382366771159875, "grad_norm": 3.6046431064605713, "learning_rate": 5.790525762819271e-07, "loss": 1.448, "step": 43159 }, { "epoch": 3.3824451410658307, "grad_norm": 2.5610601902008057, "learning_rate": 5.789085400101602e-07, "loss": 1.0182, "step": 43160 }, { "epoch": 3.3825235109717866, "grad_norm": 3.468780279159546, "learning_rate": 5.787645205539533e-07, "loss": 1.2255, "step": 43161 }, { "epoch": 3.382601880877743, "grad_norm": 2.5496420860290527, "learning_rate": 5.786205179138554e-07, "loss": 1.6914, "step": 43162 }, { "epoch": 3.3826802507836993, "grad_norm": 4.088240623474121, "learning_rate": 5.784765320904146e-07, "loss": 1.4099, "step": 43163 }, { "epoch": 3.382758620689655, "grad_norm": 2.9244165420532227, "learning_rate": 5.783325630841774e-07, "loss": 1.3225, "step": 43164 }, { "epoch": 3.382836990595611, "grad_norm": 2.5161008834838867, "learning_rate": 5.781886108956919e-07, "loss": 0.8649, "step": 43165 }, { "epoch": 3.3829153605015674, "grad_norm": 3.466718912124634, "learning_rate": 5.780446755255059e-07, "loss": 1.445, "step": 43166 }, { "epoch": 3.3829937304075237, "grad_norm": 3.724487066268921, "learning_rate": 5.779007569741679e-07, "loss": 0.9856, "step": 43167 }, { "epoch": 3.3830721003134796, "grad_norm": 3.6222386360168457, "learning_rate": 5.777568552422219e-07, "loss": 1.4146, "step": 43168 }, { "epoch": 3.3831504702194355, "grad_norm": 3.609636068344116, "learning_rate": 5.776129703302186e-07, "loss": 1.4029, "step": 43169 }, { "epoch": 3.383228840125392, "grad_norm": 2.7496542930603027, "learning_rate": 5.774691022387041e-07, "loss": 1.3052, "step": 43170 }, { "epoch": 3.383307210031348, "grad_norm": 7.254340648651123, "learning_rate": 5.773252509682265e-07, "loss": 1.6508, "step": 43171 }, { "epoch": 3.383385579937304, "grad_norm": 3.000392198562622, "learning_rate": 5.771814165193307e-07, "loss": 1.236, "step": 43172 }, { "epoch": 3.3834639498432604, "grad_norm": 3.2337496280670166, "learning_rate": 5.770375988925642e-07, "loss": 1.7315, "step": 43173 }, { "epoch": 3.3835423197492163, "grad_norm": 5.278057098388672, "learning_rate": 5.768937980884776e-07, "loss": 1.2726, "step": 43174 }, { "epoch": 3.3836206896551726, "grad_norm": 2.925128698348999, "learning_rate": 5.767500141076132e-07, "loss": 1.2643, "step": 43175 }, { "epoch": 3.3836990595611285, "grad_norm": 6.867885589599609, "learning_rate": 5.766062469505202e-07, "loss": 1.7954, "step": 43176 }, { "epoch": 3.383777429467085, "grad_norm": 2.3280720710754395, "learning_rate": 5.764624966177452e-07, "loss": 1.1283, "step": 43177 }, { "epoch": 3.3838557993730407, "grad_norm": 4.601885795593262, "learning_rate": 5.763187631098344e-07, "loss": 1.0326, "step": 43178 }, { "epoch": 3.383934169278997, "grad_norm": 14.680992126464844, "learning_rate": 5.761750464273352e-07, "loss": 1.2793, "step": 43179 }, { "epoch": 3.384012539184953, "grad_norm": 3.0381369590759277, "learning_rate": 5.76031346570794e-07, "loss": 0.9161, "step": 43180 }, { "epoch": 3.3840909090909093, "grad_norm": 4.2495927810668945, "learning_rate": 5.758876635407573e-07, "loss": 1.3806, "step": 43181 }, { "epoch": 3.384169278996865, "grad_norm": 3.232640504837036, "learning_rate": 5.75743997337771e-07, "loss": 1.7687, "step": 43182 }, { "epoch": 3.3842476489028215, "grad_norm": 2.378119945526123, "learning_rate": 5.756003479623822e-07, "loss": 1.2946, "step": 43183 }, { "epoch": 3.3843260188087774, "grad_norm": 3.2423906326293945, "learning_rate": 5.754567154151375e-07, "loss": 1.0288, "step": 43184 }, { "epoch": 3.3844043887147337, "grad_norm": 3.3690409660339355, "learning_rate": 5.753130996965827e-07, "loss": 1.6584, "step": 43185 }, { "epoch": 3.3844827586206896, "grad_norm": 2.534635305404663, "learning_rate": 5.751695008072655e-07, "loss": 1.3277, "step": 43186 }, { "epoch": 3.384561128526646, "grad_norm": 3.088735818862915, "learning_rate": 5.750259187477286e-07, "loss": 1.5786, "step": 43187 }, { "epoch": 3.384639498432602, "grad_norm": 8.672615051269531, "learning_rate": 5.748823535185211e-07, "loss": 1.8947, "step": 43188 }, { "epoch": 3.384717868338558, "grad_norm": 2.664212465286255, "learning_rate": 5.747388051201897e-07, "loss": 1.4961, "step": 43189 }, { "epoch": 3.384796238244514, "grad_norm": 2.249366044998169, "learning_rate": 5.745952735532772e-07, "loss": 1.4709, "step": 43190 }, { "epoch": 3.3848746081504704, "grad_norm": 3.214512586593628, "learning_rate": 5.744517588183319e-07, "loss": 1.0796, "step": 43191 }, { "epoch": 3.3849529780564263, "grad_norm": 2.9601759910583496, "learning_rate": 5.743082609158978e-07, "loss": 1.413, "step": 43192 }, { "epoch": 3.3850313479623826, "grad_norm": 3.043423891067505, "learning_rate": 5.741647798465239e-07, "loss": 1.3196, "step": 43193 }, { "epoch": 3.3851097178683385, "grad_norm": 5.966470718383789, "learning_rate": 5.740213156107527e-07, "loss": 1.3271, "step": 43194 }, { "epoch": 3.385188087774295, "grad_norm": 4.140600681304932, "learning_rate": 5.738778682091317e-07, "loss": 1.5224, "step": 43195 }, { "epoch": 3.3852664576802507, "grad_norm": 3.1559970378875732, "learning_rate": 5.737344376422054e-07, "loss": 1.3571, "step": 43196 }, { "epoch": 3.385344827586207, "grad_norm": 7.214629173278809, "learning_rate": 5.735910239105197e-07, "loss": 1.2795, "step": 43197 }, { "epoch": 3.385423197492163, "grad_norm": 8.865418434143066, "learning_rate": 5.734476270146205e-07, "loss": 1.2813, "step": 43198 }, { "epoch": 3.3855015673981192, "grad_norm": 3.0476627349853516, "learning_rate": 5.733042469550527e-07, "loss": 1.0828, "step": 43199 }, { "epoch": 3.385579937304075, "grad_norm": 2.7524092197418213, "learning_rate": 5.731608837323621e-07, "loss": 0.9599, "step": 43200 }, { "epoch": 3.3856583072100315, "grad_norm": 9.552632331848145, "learning_rate": 5.730175373470936e-07, "loss": 1.1027, "step": 43201 }, { "epoch": 3.3857366771159874, "grad_norm": 8.078550338745117, "learning_rate": 5.728742077997923e-07, "loss": 1.2778, "step": 43202 }, { "epoch": 3.3858150470219437, "grad_norm": 3.2382137775421143, "learning_rate": 5.727308950910037e-07, "loss": 1.1163, "step": 43203 }, { "epoch": 3.3858934169278996, "grad_norm": 2.5141963958740234, "learning_rate": 5.725875992212743e-07, "loss": 1.4171, "step": 43204 }, { "epoch": 3.385971786833856, "grad_norm": 8.12321949005127, "learning_rate": 5.72444320191145e-07, "loss": 0.9839, "step": 43205 }, { "epoch": 3.386050156739812, "grad_norm": 3.1674134731292725, "learning_rate": 5.723010580011645e-07, "loss": 1.2583, "step": 43206 }, { "epoch": 3.386128526645768, "grad_norm": 2.1014821529388428, "learning_rate": 5.721578126518767e-07, "loss": 1.2397, "step": 43207 }, { "epoch": 3.386206896551724, "grad_norm": 3.1137332916259766, "learning_rate": 5.720145841438274e-07, "loss": 1.4626, "step": 43208 }, { "epoch": 3.3862852664576804, "grad_norm": 2.981578826904297, "learning_rate": 5.718713724775593e-07, "loss": 1.4239, "step": 43209 }, { "epoch": 3.3863636363636362, "grad_norm": 6.114529609680176, "learning_rate": 5.717281776536166e-07, "loss": 1.9226, "step": 43210 }, { "epoch": 3.3864420062695926, "grad_norm": 5.129144668579102, "learning_rate": 5.71584999672547e-07, "loss": 1.7927, "step": 43211 }, { "epoch": 3.3865203761755485, "grad_norm": 3.029327392578125, "learning_rate": 5.714418385348941e-07, "loss": 1.4061, "step": 43212 }, { "epoch": 3.386598746081505, "grad_norm": 3.539815664291382, "learning_rate": 5.712986942412007e-07, "loss": 1.5923, "step": 43213 }, { "epoch": 3.3866771159874607, "grad_norm": 3.8168020248413086, "learning_rate": 5.711555667920126e-07, "loss": 0.8203, "step": 43214 }, { "epoch": 3.386755485893417, "grad_norm": 11.890758514404297, "learning_rate": 5.710124561878738e-07, "loss": 0.6077, "step": 43215 }, { "epoch": 3.386833855799373, "grad_norm": 2.651020050048828, "learning_rate": 5.708693624293288e-07, "loss": 1.3, "step": 43216 }, { "epoch": 3.3869122257053292, "grad_norm": 2.3386731147766113, "learning_rate": 5.707262855169216e-07, "loss": 1.39, "step": 43217 }, { "epoch": 3.386990595611285, "grad_norm": 3.410130739212036, "learning_rate": 5.705832254511967e-07, "loss": 1.0986, "step": 43218 }, { "epoch": 3.3870689655172415, "grad_norm": 2.1251981258392334, "learning_rate": 5.704401822326983e-07, "loss": 1.2491, "step": 43219 }, { "epoch": 3.3871473354231973, "grad_norm": 3.172441005706787, "learning_rate": 5.702971558619702e-07, "loss": 1.3898, "step": 43220 }, { "epoch": 3.3872257053291537, "grad_norm": 3.062114715576172, "learning_rate": 5.701541463395566e-07, "loss": 0.7692, "step": 43221 }, { "epoch": 3.3873040752351096, "grad_norm": 3.436711072921753, "learning_rate": 5.700111536660013e-07, "loss": 0.9924, "step": 43222 }, { "epoch": 3.387382445141066, "grad_norm": 2.5745110511779785, "learning_rate": 5.698681778418491e-07, "loss": 1.2118, "step": 43223 }, { "epoch": 3.387460815047022, "grad_norm": 3.380300760269165, "learning_rate": 5.697252188676405e-07, "loss": 0.8847, "step": 43224 }, { "epoch": 3.387539184952978, "grad_norm": 4.120588302612305, "learning_rate": 5.69582276743923e-07, "loss": 1.5225, "step": 43225 }, { "epoch": 3.387617554858934, "grad_norm": 3.454583168029785, "learning_rate": 5.694393514712393e-07, "loss": 1.6325, "step": 43226 }, { "epoch": 3.3876959247648903, "grad_norm": 3.1579418182373047, "learning_rate": 5.69296443050133e-07, "loss": 1.7865, "step": 43227 }, { "epoch": 3.3877742946708462, "grad_norm": 3.8763601779937744, "learning_rate": 5.691535514811463e-07, "loss": 1.0937, "step": 43228 }, { "epoch": 3.3878526645768026, "grad_norm": 7.128001689910889, "learning_rate": 5.69010676764823e-07, "loss": 1.5232, "step": 43229 }, { "epoch": 3.3879310344827585, "grad_norm": 4.624022960662842, "learning_rate": 5.688678189017094e-07, "loss": 1.2964, "step": 43230 }, { "epoch": 3.388009404388715, "grad_norm": 2.6980373859405518, "learning_rate": 5.687249778923448e-07, "loss": 1.3092, "step": 43231 }, { "epoch": 3.3880877742946707, "grad_norm": 5.328600883483887, "learning_rate": 5.685821537372748e-07, "loss": 0.9582, "step": 43232 }, { "epoch": 3.388166144200627, "grad_norm": 2.6453747749328613, "learning_rate": 5.684393464370425e-07, "loss": 1.2879, "step": 43233 }, { "epoch": 3.388244514106583, "grad_norm": 2.8134303092956543, "learning_rate": 5.682965559921905e-07, "loss": 1.2593, "step": 43234 }, { "epoch": 3.3883228840125392, "grad_norm": 2.918250560760498, "learning_rate": 5.681537824032618e-07, "loss": 1.5975, "step": 43235 }, { "epoch": 3.388401253918495, "grad_norm": 4.305028915405273, "learning_rate": 5.680110256707999e-07, "loss": 1.3372, "step": 43236 }, { "epoch": 3.3884796238244514, "grad_norm": 5.123480796813965, "learning_rate": 5.678682857953482e-07, "loss": 1.5236, "step": 43237 }, { "epoch": 3.3885579937304073, "grad_norm": 3.337501049041748, "learning_rate": 5.677255627774486e-07, "loss": 1.5419, "step": 43238 }, { "epoch": 3.3886363636363637, "grad_norm": 9.741554260253906, "learning_rate": 5.675828566176445e-07, "loss": 0.5583, "step": 43239 }, { "epoch": 3.3887147335423196, "grad_norm": 2.250922679901123, "learning_rate": 5.674401673164781e-07, "loss": 0.9651, "step": 43240 }, { "epoch": 3.388793103448276, "grad_norm": 4.421445846557617, "learning_rate": 5.672974948744931e-07, "loss": 1.6227, "step": 43241 }, { "epoch": 3.3888714733542322, "grad_norm": 3.879002571105957, "learning_rate": 5.671548392922327e-07, "loss": 1.4481, "step": 43242 }, { "epoch": 3.388949843260188, "grad_norm": 5.070301055908203, "learning_rate": 5.670122005702361e-07, "loss": 1.5334, "step": 43243 }, { "epoch": 3.389028213166144, "grad_norm": 3.9534547328948975, "learning_rate": 5.668695787090495e-07, "loss": 1.5027, "step": 43244 }, { "epoch": 3.3891065830721003, "grad_norm": 11.598530769348145, "learning_rate": 5.667269737092151e-07, "loss": 1.2648, "step": 43245 }, { "epoch": 3.3891849529780567, "grad_norm": 3.6277871131896973, "learning_rate": 5.665843855712727e-07, "loss": 1.5378, "step": 43246 }, { "epoch": 3.3892633228840126, "grad_norm": 2.3398332595825195, "learning_rate": 5.664418142957667e-07, "loss": 1.0722, "step": 43247 }, { "epoch": 3.3893416927899684, "grad_norm": 5.070123672485352, "learning_rate": 5.662992598832373e-07, "loss": 1.3752, "step": 43248 }, { "epoch": 3.3894200626959248, "grad_norm": 7.495486259460449, "learning_rate": 5.661567223342307e-07, "loss": 1.3923, "step": 43249 }, { "epoch": 3.389498432601881, "grad_norm": 2.405895948410034, "learning_rate": 5.660142016492853e-07, "loss": 1.3558, "step": 43250 }, { "epoch": 3.389576802507837, "grad_norm": 3.8287084102630615, "learning_rate": 5.658716978289447e-07, "loss": 1.4018, "step": 43251 }, { "epoch": 3.389655172413793, "grad_norm": 3.5779192447662354, "learning_rate": 5.657292108737505e-07, "loss": 1.4042, "step": 43252 }, { "epoch": 3.389733542319749, "grad_norm": 2.406452178955078, "learning_rate": 5.655867407842447e-07, "loss": 1.409, "step": 43253 }, { "epoch": 3.3898119122257055, "grad_norm": 4.27503776550293, "learning_rate": 5.654442875609694e-07, "loss": 1.7194, "step": 43254 }, { "epoch": 3.3898902821316614, "grad_norm": 3.1988155841827393, "learning_rate": 5.653018512044661e-07, "loss": 1.5928, "step": 43255 }, { "epoch": 3.3899686520376173, "grad_norm": 3.1603548526763916, "learning_rate": 5.65159431715277e-07, "loss": 1.359, "step": 43256 }, { "epoch": 3.3900470219435737, "grad_norm": 4.6446967124938965, "learning_rate": 5.650170290939433e-07, "loss": 1.5901, "step": 43257 }, { "epoch": 3.39012539184953, "grad_norm": 2.481917142868042, "learning_rate": 5.648746433410068e-07, "loss": 1.3151, "step": 43258 }, { "epoch": 3.390203761755486, "grad_norm": 5.302651405334473, "learning_rate": 5.647322744570094e-07, "loss": 1.5272, "step": 43259 }, { "epoch": 3.3902821316614418, "grad_norm": 2.4522578716278076, "learning_rate": 5.645899224424933e-07, "loss": 1.3919, "step": 43260 }, { "epoch": 3.390360501567398, "grad_norm": 5.686126708984375, "learning_rate": 5.644475872979976e-07, "loss": 2.0202, "step": 43261 }, { "epoch": 3.3904388714733544, "grad_norm": 2.3991801738739014, "learning_rate": 5.643052690240647e-07, "loss": 1.4296, "step": 43262 }, { "epoch": 3.3905172413793103, "grad_norm": 3.3641645908355713, "learning_rate": 5.641629676212368e-07, "loss": 0.6328, "step": 43263 }, { "epoch": 3.390595611285266, "grad_norm": 4.233086585998535, "learning_rate": 5.640206830900558e-07, "loss": 1.4923, "step": 43264 }, { "epoch": 3.3906739811912225, "grad_norm": 4.026820659637451, "learning_rate": 5.638784154310606e-07, "loss": 1.1447, "step": 43265 }, { "epoch": 3.390752351097179, "grad_norm": 6.446739196777344, "learning_rate": 5.63736164644793e-07, "loss": 1.4403, "step": 43266 }, { "epoch": 3.3908307210031348, "grad_norm": 3.188608407974243, "learning_rate": 5.635939307317939e-07, "loss": 1.207, "step": 43267 }, { "epoch": 3.390909090909091, "grad_norm": 5.124127388000488, "learning_rate": 5.634517136926065e-07, "loss": 1.0489, "step": 43268 }, { "epoch": 3.390987460815047, "grad_norm": 4.057099342346191, "learning_rate": 5.633095135277695e-07, "loss": 1.2749, "step": 43269 }, { "epoch": 3.3910658307210033, "grad_norm": 2.6735618114471436, "learning_rate": 5.631673302378238e-07, "loss": 1.6098, "step": 43270 }, { "epoch": 3.391144200626959, "grad_norm": 2.8557870388031006, "learning_rate": 5.630251638233109e-07, "loss": 1.3731, "step": 43271 }, { "epoch": 3.3912225705329155, "grad_norm": 6.1743903160095215, "learning_rate": 5.628830142847707e-07, "loss": 1.0253, "step": 43272 }, { "epoch": 3.3913009404388714, "grad_norm": 2.9727320671081543, "learning_rate": 5.627408816227453e-07, "loss": 1.5042, "step": 43273 }, { "epoch": 3.3913793103448278, "grad_norm": 3.0912153720855713, "learning_rate": 5.62598765837774e-07, "loss": 1.4017, "step": 43274 }, { "epoch": 3.3914576802507836, "grad_norm": 3.328199863433838, "learning_rate": 5.624566669303994e-07, "loss": 1.8051, "step": 43275 }, { "epoch": 3.39153605015674, "grad_norm": 3.2677178382873535, "learning_rate": 5.623145849011575e-07, "loss": 1.0905, "step": 43276 }, { "epoch": 3.391614420062696, "grad_norm": 4.298624038696289, "learning_rate": 5.621725197505934e-07, "loss": 1.0951, "step": 43277 }, { "epoch": 3.391692789968652, "grad_norm": 5.132793426513672, "learning_rate": 5.620304714792452e-07, "loss": 2.3741, "step": 43278 }, { "epoch": 3.391771159874608, "grad_norm": 5.063642501831055, "learning_rate": 5.618884400876545e-07, "loss": 1.2497, "step": 43279 }, { "epoch": 3.3918495297805644, "grad_norm": 6.332230567932129, "learning_rate": 5.617464255763593e-07, "loss": 1.0764, "step": 43280 }, { "epoch": 3.3919278996865203, "grad_norm": 4.436394691467285, "learning_rate": 5.61604427945901e-07, "loss": 1.8633, "step": 43281 }, { "epoch": 3.3920062695924766, "grad_norm": 4.269631862640381, "learning_rate": 5.614624471968199e-07, "loss": 1.0695, "step": 43282 }, { "epoch": 3.3920846394984325, "grad_norm": 3.19649600982666, "learning_rate": 5.613204833296576e-07, "loss": 1.2594, "step": 43283 }, { "epoch": 3.392163009404389, "grad_norm": 5.5173821449279785, "learning_rate": 5.611785363449507e-07, "loss": 1.1254, "step": 43284 }, { "epoch": 3.3922413793103448, "grad_norm": 4.3604326248168945, "learning_rate": 5.61036606243241e-07, "loss": 1.378, "step": 43285 }, { "epoch": 3.392319749216301, "grad_norm": 5.184786796569824, "learning_rate": 5.608946930250681e-07, "loss": 1.6026, "step": 43286 }, { "epoch": 3.392398119122257, "grad_norm": 2.449573278427124, "learning_rate": 5.607527966909715e-07, "loss": 1.3718, "step": 43287 }, { "epoch": 3.3924764890282133, "grad_norm": 2.9545106887817383, "learning_rate": 5.606109172414914e-07, "loss": 1.423, "step": 43288 }, { "epoch": 3.392554858934169, "grad_norm": 3.7736761569976807, "learning_rate": 5.604690546771668e-07, "loss": 1.3686, "step": 43289 }, { "epoch": 3.3926332288401255, "grad_norm": 3.8940389156341553, "learning_rate": 5.603272089985384e-07, "loss": 1.2696, "step": 43290 }, { "epoch": 3.3927115987460814, "grad_norm": 5.533756256103516, "learning_rate": 5.601853802061441e-07, "loss": 1.4288, "step": 43291 }, { "epoch": 3.3927899686520377, "grad_norm": 5.2138590812683105, "learning_rate": 5.600435683005245e-07, "loss": 1.5535, "step": 43292 }, { "epoch": 3.3928683385579936, "grad_norm": 3.2193491458892822, "learning_rate": 5.599017732822188e-07, "loss": 1.4778, "step": 43293 }, { "epoch": 3.39294670846395, "grad_norm": 4.171922206878662, "learning_rate": 5.597599951517657e-07, "loss": 1.5959, "step": 43294 }, { "epoch": 3.393025078369906, "grad_norm": 3.520768165588379, "learning_rate": 5.596182339097056e-07, "loss": 1.2924, "step": 43295 }, { "epoch": 3.393103448275862, "grad_norm": 7.0777907371521, "learning_rate": 5.594764895565763e-07, "loss": 1.8374, "step": 43296 }, { "epoch": 3.393181818181818, "grad_norm": 3.3965587615966797, "learning_rate": 5.593347620929184e-07, "loss": 1.0999, "step": 43297 }, { "epoch": 3.3932601880877744, "grad_norm": 4.765865325927734, "learning_rate": 5.591930515192707e-07, "loss": 0.8279, "step": 43298 }, { "epoch": 3.3933385579937303, "grad_norm": 2.6870980262756348, "learning_rate": 5.590513578361695e-07, "loss": 1.0997, "step": 43299 }, { "epoch": 3.3934169278996866, "grad_norm": 4.906308174133301, "learning_rate": 5.589096810441574e-07, "loss": 1.7769, "step": 43300 }, { "epoch": 3.3934952978056425, "grad_norm": 2.7107841968536377, "learning_rate": 5.587680211437724e-07, "loss": 1.1954, "step": 43301 }, { "epoch": 3.393573667711599, "grad_norm": 3.7243616580963135, "learning_rate": 5.586263781355522e-07, "loss": 0.9698, "step": 43302 }, { "epoch": 3.3936520376175547, "grad_norm": 3.6214284896850586, "learning_rate": 5.584847520200354e-07, "loss": 1.2797, "step": 43303 }, { "epoch": 3.393730407523511, "grad_norm": 2.184281349182129, "learning_rate": 5.58343142797762e-07, "loss": 1.2721, "step": 43304 }, { "epoch": 3.393808777429467, "grad_norm": 6.4852190017700195, "learning_rate": 5.582015504692695e-07, "loss": 1.9165, "step": 43305 }, { "epoch": 3.3938871473354233, "grad_norm": 7.628109931945801, "learning_rate": 5.580599750350968e-07, "loss": 0.9746, "step": 43306 }, { "epoch": 3.393965517241379, "grad_norm": 4.297758102416992, "learning_rate": 5.579184164957823e-07, "loss": 1.227, "step": 43307 }, { "epoch": 3.3940438871473355, "grad_norm": 5.34268045425415, "learning_rate": 5.577768748518647e-07, "loss": 1.4576, "step": 43308 }, { "epoch": 3.3941222570532914, "grad_norm": 4.13754940032959, "learning_rate": 5.576353501038822e-07, "loss": 1.3367, "step": 43309 }, { "epoch": 3.3942006269592477, "grad_norm": 7.272706508636475, "learning_rate": 5.574938422523729e-07, "loss": 1.7559, "step": 43310 }, { "epoch": 3.3942789968652036, "grad_norm": 2.8302042484283447, "learning_rate": 5.573523512978757e-07, "loss": 1.3223, "step": 43311 }, { "epoch": 3.39435736677116, "grad_norm": 4.644885063171387, "learning_rate": 5.572108772409279e-07, "loss": 1.6679, "step": 43312 }, { "epoch": 3.394435736677116, "grad_norm": 2.7475109100341797, "learning_rate": 5.570694200820681e-07, "loss": 1.3358, "step": 43313 }, { "epoch": 3.394514106583072, "grad_norm": 6.462557792663574, "learning_rate": 5.569279798218341e-07, "loss": 1.4736, "step": 43314 }, { "epoch": 3.394592476489028, "grad_norm": 2.510806083679199, "learning_rate": 5.567865564607638e-07, "loss": 1.2346, "step": 43315 }, { "epoch": 3.3946708463949844, "grad_norm": 4.070143699645996, "learning_rate": 5.566451499993969e-07, "loss": 1.6035, "step": 43316 }, { "epoch": 3.3947492163009403, "grad_norm": 2.2693114280700684, "learning_rate": 5.565037604382678e-07, "loss": 0.9602, "step": 43317 }, { "epoch": 3.3948275862068966, "grad_norm": 2.6643970012664795, "learning_rate": 5.563623877779156e-07, "loss": 1.5874, "step": 43318 }, { "epoch": 3.3949059561128525, "grad_norm": 10.100936889648438, "learning_rate": 5.562210320188793e-07, "loss": 1.0416, "step": 43319 }, { "epoch": 3.394984326018809, "grad_norm": 2.416327476501465, "learning_rate": 5.560796931616969e-07, "loss": 1.1919, "step": 43320 }, { "epoch": 3.3950626959247647, "grad_norm": 3.5022048950195312, "learning_rate": 5.559383712069033e-07, "loss": 1.1238, "step": 43321 }, { "epoch": 3.395141065830721, "grad_norm": 5.188354015350342, "learning_rate": 5.557970661550382e-07, "loss": 0.9593, "step": 43322 }, { "epoch": 3.395219435736677, "grad_norm": 6.020363807678223, "learning_rate": 5.556557780066368e-07, "loss": 1.4685, "step": 43323 }, { "epoch": 3.3952978056426333, "grad_norm": 9.629030227661133, "learning_rate": 5.555145067622403e-07, "loss": 0.9416, "step": 43324 }, { "epoch": 3.395376175548589, "grad_norm": 2.5241470336914062, "learning_rate": 5.553732524223831e-07, "loss": 1.4755, "step": 43325 }, { "epoch": 3.3954545454545455, "grad_norm": 3.6968889236450195, "learning_rate": 5.552320149876028e-07, "loss": 1.5863, "step": 43326 }, { "epoch": 3.3955329153605014, "grad_norm": 2.6095409393310547, "learning_rate": 5.550907944584366e-07, "loss": 1.0092, "step": 43327 }, { "epoch": 3.3956112852664577, "grad_norm": 3.1637725830078125, "learning_rate": 5.549495908354224e-07, "loss": 0.9204, "step": 43328 }, { "epoch": 3.3956896551724136, "grad_norm": 4.5929107666015625, "learning_rate": 5.548084041190971e-07, "loss": 1.5891, "step": 43329 }, { "epoch": 3.39576802507837, "grad_norm": 4.094923496246338, "learning_rate": 5.546672343099968e-07, "loss": 1.6219, "step": 43330 }, { "epoch": 3.395846394984326, "grad_norm": 4.218477249145508, "learning_rate": 5.545260814086606e-07, "loss": 1.7456, "step": 43331 }, { "epoch": 3.395924764890282, "grad_norm": 2.571481227874756, "learning_rate": 5.543849454156213e-07, "loss": 1.5703, "step": 43332 }, { "epoch": 3.396003134796238, "grad_norm": 2.848266363143921, "learning_rate": 5.542438263314198e-07, "loss": 1.3651, "step": 43333 }, { "epoch": 3.3960815047021944, "grad_norm": 2.3327114582061768, "learning_rate": 5.54102724156591e-07, "loss": 1.5491, "step": 43334 }, { "epoch": 3.3961598746081503, "grad_norm": 3.451328992843628, "learning_rate": 5.539616388916729e-07, "loss": 1.3637, "step": 43335 }, { "epoch": 3.3962382445141066, "grad_norm": 3.9101343154907227, "learning_rate": 5.538205705372001e-07, "loss": 1.3019, "step": 43336 }, { "epoch": 3.396316614420063, "grad_norm": 2.715623140335083, "learning_rate": 5.536795190937094e-07, "loss": 1.2663, "step": 43337 }, { "epoch": 3.396394984326019, "grad_norm": 3.267500400543213, "learning_rate": 5.535384845617386e-07, "loss": 1.0565, "step": 43338 }, { "epoch": 3.3964733542319747, "grad_norm": 4.500006675720215, "learning_rate": 5.533974669418251e-07, "loss": 1.6145, "step": 43339 }, { "epoch": 3.396551724137931, "grad_norm": 5.672717571258545, "learning_rate": 5.532564662345025e-07, "loss": 0.9039, "step": 43340 }, { "epoch": 3.3966300940438874, "grad_norm": 3.8492467403411865, "learning_rate": 5.531154824403078e-07, "loss": 1.1973, "step": 43341 }, { "epoch": 3.3967084639498433, "grad_norm": 2.507899045944214, "learning_rate": 5.529745155597782e-07, "loss": 0.7737, "step": 43342 }, { "epoch": 3.396786833855799, "grad_norm": 3.177220344543457, "learning_rate": 5.528335655934497e-07, "loss": 1.1072, "step": 43343 }, { "epoch": 3.3968652037617555, "grad_norm": 3.827846050262451, "learning_rate": 5.526926325418574e-07, "loss": 1.8615, "step": 43344 }, { "epoch": 3.396943573667712, "grad_norm": 7.15916633605957, "learning_rate": 5.525517164055388e-07, "loss": 1.2259, "step": 43345 }, { "epoch": 3.3970219435736677, "grad_norm": 3.791224479675293, "learning_rate": 5.524108171850285e-07, "loss": 0.8836, "step": 43346 }, { "epoch": 3.3971003134796236, "grad_norm": 5.476400852203369, "learning_rate": 5.522699348808635e-07, "loss": 1.4467, "step": 43347 }, { "epoch": 3.39717868338558, "grad_norm": 3.202373504638672, "learning_rate": 5.521290694935788e-07, "loss": 1.1907, "step": 43348 }, { "epoch": 3.3972570532915363, "grad_norm": 3.3770883083343506, "learning_rate": 5.51988221023711e-07, "loss": 1.4939, "step": 43349 }, { "epoch": 3.397335423197492, "grad_norm": 2.887857675552368, "learning_rate": 5.518473894717958e-07, "loss": 1.5221, "step": 43350 }, { "epoch": 3.397413793103448, "grad_norm": 2.3661718368530273, "learning_rate": 5.517065748383665e-07, "loss": 1.2317, "step": 43351 }, { "epoch": 3.3974921630094044, "grad_norm": 3.9596378803253174, "learning_rate": 5.515657771239619e-07, "loss": 1.7363, "step": 43352 }, { "epoch": 3.3975705329153607, "grad_norm": 3.567572832107544, "learning_rate": 5.514249963291163e-07, "loss": 1.1444, "step": 43353 }, { "epoch": 3.3976489028213166, "grad_norm": 4.322274684906006, "learning_rate": 5.512842324543655e-07, "loss": 1.3504, "step": 43354 }, { "epoch": 3.3977272727272725, "grad_norm": 6.646087169647217, "learning_rate": 5.51143485500244e-07, "loss": 1.3387, "step": 43355 }, { "epoch": 3.397805642633229, "grad_norm": 4.667496204376221, "learning_rate": 5.510027554672864e-07, "loss": 1.0143, "step": 43356 }, { "epoch": 3.397884012539185, "grad_norm": 5.627015590667725, "learning_rate": 5.508620423560313e-07, "loss": 1.8976, "step": 43357 }, { "epoch": 3.397962382445141, "grad_norm": 4.315707683563232, "learning_rate": 5.507213461670108e-07, "loss": 0.8611, "step": 43358 }, { "epoch": 3.3980407523510974, "grad_norm": 11.484498977661133, "learning_rate": 5.505806669007607e-07, "loss": 1.5503, "step": 43359 }, { "epoch": 3.3981191222570533, "grad_norm": 2.259397268295288, "learning_rate": 5.504400045578167e-07, "loss": 1.1627, "step": 43360 }, { "epoch": 3.3981974921630096, "grad_norm": 3.2083327770233154, "learning_rate": 5.502993591387135e-07, "loss": 1.4229, "step": 43361 }, { "epoch": 3.3982758620689655, "grad_norm": 2.919421911239624, "learning_rate": 5.501587306439854e-07, "loss": 1.2873, "step": 43362 }, { "epoch": 3.398354231974922, "grad_norm": 4.985390663146973, "learning_rate": 5.500181190741687e-07, "loss": 1.0618, "step": 43363 }, { "epoch": 3.3984326018808777, "grad_norm": 3.731739044189453, "learning_rate": 5.498775244297971e-07, "loss": 1.3272, "step": 43364 }, { "epoch": 3.398510971786834, "grad_norm": 3.626279592514038, "learning_rate": 5.497369467114055e-07, "loss": 1.3878, "step": 43365 }, { "epoch": 3.39858934169279, "grad_norm": 6.480888843536377, "learning_rate": 5.495963859195291e-07, "loss": 1.7193, "step": 43366 }, { "epoch": 3.3986677115987463, "grad_norm": 2.584473133087158, "learning_rate": 5.494558420547019e-07, "loss": 1.438, "step": 43367 }, { "epoch": 3.398746081504702, "grad_norm": 5.8995208740234375, "learning_rate": 5.493153151174591e-07, "loss": 1.4477, "step": 43368 }, { "epoch": 3.3988244514106585, "grad_norm": 8.424440383911133, "learning_rate": 5.491748051083351e-07, "loss": 0.8564, "step": 43369 }, { "epoch": 3.3989028213166144, "grad_norm": 8.389220237731934, "learning_rate": 5.49034312027863e-07, "loss": 0.7822, "step": 43370 }, { "epoch": 3.3989811912225707, "grad_norm": 2.940303325653076, "learning_rate": 5.488938358765783e-07, "loss": 1.3796, "step": 43371 }, { "epoch": 3.3990595611285266, "grad_norm": 4.278672218322754, "learning_rate": 5.487533766550168e-07, "loss": 1.231, "step": 43372 }, { "epoch": 3.399137931034483, "grad_norm": 2.7238869667053223, "learning_rate": 5.486129343637097e-07, "loss": 1.3223, "step": 43373 }, { "epoch": 3.399216300940439, "grad_norm": 3.4578354358673096, "learning_rate": 5.484725090031934e-07, "loss": 0.782, "step": 43374 }, { "epoch": 3.399294670846395, "grad_norm": 3.2183971405029297, "learning_rate": 5.483321005739995e-07, "loss": 1.3656, "step": 43375 }, { "epoch": 3.399373040752351, "grad_norm": 2.6923251152038574, "learning_rate": 5.481917090766659e-07, "loss": 1.2627, "step": 43376 }, { "epoch": 3.3994514106583074, "grad_norm": 3.7509264945983887, "learning_rate": 5.480513345117239e-07, "loss": 1.4637, "step": 43377 }, { "epoch": 3.3995297805642632, "grad_norm": 2.946814775466919, "learning_rate": 5.479109768797075e-07, "loss": 1.6068, "step": 43378 }, { "epoch": 3.3996081504702196, "grad_norm": 3.0040740966796875, "learning_rate": 5.477706361811508e-07, "loss": 1.2396, "step": 43379 }, { "epoch": 3.3996865203761755, "grad_norm": 2.192148447036743, "learning_rate": 5.476303124165883e-07, "loss": 1.3609, "step": 43380 }, { "epoch": 3.399764890282132, "grad_norm": 2.241136074066162, "learning_rate": 5.474900055865529e-07, "loss": 1.3838, "step": 43381 }, { "epoch": 3.3998432601880877, "grad_norm": 8.550996780395508, "learning_rate": 5.473497156915792e-07, "loss": 1.3978, "step": 43382 }, { "epoch": 3.399921630094044, "grad_norm": 2.2725188732147217, "learning_rate": 5.472094427322006e-07, "loss": 1.4107, "step": 43383 }, { "epoch": 3.4, "grad_norm": 2.1781821250915527, "learning_rate": 5.470691867089478e-07, "loss": 1.2767, "step": 43384 }, { "epoch": 3.4000783699059562, "grad_norm": 2.1377906799316406, "learning_rate": 5.469289476223583e-07, "loss": 1.5611, "step": 43385 }, { "epoch": 3.400156739811912, "grad_norm": 3.503141403198242, "learning_rate": 5.467887254729637e-07, "loss": 1.4601, "step": 43386 }, { "epoch": 3.4002351097178685, "grad_norm": 2.875033140182495, "learning_rate": 5.466485202612981e-07, "loss": 1.6921, "step": 43387 }, { "epoch": 3.4003134796238244, "grad_norm": 4.098357677459717, "learning_rate": 5.465083319878933e-07, "loss": 1.704, "step": 43388 }, { "epoch": 3.4003918495297807, "grad_norm": 3.26672101020813, "learning_rate": 5.463681606532828e-07, "loss": 1.5477, "step": 43389 }, { "epoch": 3.4004702194357366, "grad_norm": 4.626906394958496, "learning_rate": 5.462280062580011e-07, "loss": 1.1969, "step": 43390 }, { "epoch": 3.400548589341693, "grad_norm": 2.54341197013855, "learning_rate": 5.460878688025812e-07, "loss": 1.549, "step": 43391 }, { "epoch": 3.400626959247649, "grad_norm": 3.644989490509033, "learning_rate": 5.459477482875547e-07, "loss": 1.1549, "step": 43392 }, { "epoch": 3.400705329153605, "grad_norm": 5.8413238525390625, "learning_rate": 5.458076447134547e-07, "loss": 1.435, "step": 43393 }, { "epoch": 3.400783699059561, "grad_norm": 5.4988908767700195, "learning_rate": 5.456675580808141e-07, "loss": 1.4586, "step": 43394 }, { "epoch": 3.4008620689655173, "grad_norm": 3.2964930534362793, "learning_rate": 5.455274883901679e-07, "loss": 1.4683, "step": 43395 }, { "epoch": 3.4009404388714732, "grad_norm": 3.4104108810424805, "learning_rate": 5.453874356420464e-07, "loss": 1.15, "step": 43396 }, { "epoch": 3.4010188087774296, "grad_norm": 3.446115493774414, "learning_rate": 5.452473998369829e-07, "loss": 1.0669, "step": 43397 }, { "epoch": 3.4010971786833855, "grad_norm": 3.2920260429382324, "learning_rate": 5.451073809755103e-07, "loss": 1.4351, "step": 43398 }, { "epoch": 3.401175548589342, "grad_norm": 2.949037551879883, "learning_rate": 5.449673790581611e-07, "loss": 1.543, "step": 43399 }, { "epoch": 3.4012539184952977, "grad_norm": 5.050143241882324, "learning_rate": 5.448273940854675e-07, "loss": 1.7169, "step": 43400 }, { "epoch": 3.401332288401254, "grad_norm": 4.153913974761963, "learning_rate": 5.446874260579621e-07, "loss": 1.2775, "step": 43401 }, { "epoch": 3.40141065830721, "grad_norm": 3.25546932220459, "learning_rate": 5.445474749761775e-07, "loss": 1.3364, "step": 43402 }, { "epoch": 3.4014890282131662, "grad_norm": 6.143742561340332, "learning_rate": 5.444075408406463e-07, "loss": 1.8069, "step": 43403 }, { "epoch": 3.401567398119122, "grad_norm": 5.451999664306641, "learning_rate": 5.442676236518995e-07, "loss": 1.3595, "step": 43404 }, { "epoch": 3.4016457680250785, "grad_norm": 5.793561935424805, "learning_rate": 5.441277234104708e-07, "loss": 1.1945, "step": 43405 }, { "epoch": 3.4017241379310343, "grad_norm": 5.47047233581543, "learning_rate": 5.439878401168919e-07, "loss": 0.9853, "step": 43406 }, { "epoch": 3.4018025078369907, "grad_norm": 2.794088840484619, "learning_rate": 5.438479737716928e-07, "loss": 1.3212, "step": 43407 }, { "epoch": 3.4018808777429466, "grad_norm": 2.8770954608917236, "learning_rate": 5.437081243754083e-07, "loss": 1.4429, "step": 43408 }, { "epoch": 3.401959247648903, "grad_norm": 3.02358341217041, "learning_rate": 5.43568291928569e-07, "loss": 1.4067, "step": 43409 }, { "epoch": 3.402037617554859, "grad_norm": 3.2877509593963623, "learning_rate": 5.434284764317083e-07, "loss": 1.2148, "step": 43410 }, { "epoch": 3.402115987460815, "grad_norm": 4.942004203796387, "learning_rate": 5.432886778853552e-07, "loss": 1.709, "step": 43411 }, { "epoch": 3.402194357366771, "grad_norm": 12.626312255859375, "learning_rate": 5.431488962900421e-07, "loss": 2.0434, "step": 43412 }, { "epoch": 3.4022727272727273, "grad_norm": 2.1313183307647705, "learning_rate": 5.430091316463032e-07, "loss": 1.3932, "step": 43413 }, { "epoch": 3.4023510971786832, "grad_norm": 3.5316109657287598, "learning_rate": 5.428693839546673e-07, "loss": 1.5348, "step": 43414 }, { "epoch": 3.4024294670846396, "grad_norm": 2.5393359661102295, "learning_rate": 5.427296532156673e-07, "loss": 1.1817, "step": 43415 }, { "epoch": 3.4025078369905954, "grad_norm": 3.9798786640167236, "learning_rate": 5.42589939429834e-07, "loss": 1.9086, "step": 43416 }, { "epoch": 3.402586206896552, "grad_norm": 2.806772232055664, "learning_rate": 5.424502425976991e-07, "loss": 1.1581, "step": 43417 }, { "epoch": 3.4026645768025077, "grad_norm": 3.790555477142334, "learning_rate": 5.423105627197938e-07, "loss": 0.9544, "step": 43418 }, { "epoch": 3.402742946708464, "grad_norm": 4.441989421844482, "learning_rate": 5.421708997966501e-07, "loss": 1.0435, "step": 43419 }, { "epoch": 3.40282131661442, "grad_norm": 3.8600871562957764, "learning_rate": 5.420312538287981e-07, "loss": 1.6549, "step": 43420 }, { "epoch": 3.402899686520376, "grad_norm": 4.106057167053223, "learning_rate": 5.418916248167694e-07, "loss": 1.3355, "step": 43421 }, { "epoch": 3.402978056426332, "grad_norm": 2.9309535026550293, "learning_rate": 5.417520127610954e-07, "loss": 1.1899, "step": 43422 }, { "epoch": 3.4030564263322884, "grad_norm": 2.6958518028259277, "learning_rate": 5.416124176623067e-07, "loss": 1.3343, "step": 43423 }, { "epoch": 3.4031347962382443, "grad_norm": 2.373884916305542, "learning_rate": 5.414728395209345e-07, "loss": 1.1051, "step": 43424 }, { "epoch": 3.4032131661442007, "grad_norm": 3.966862916946411, "learning_rate": 5.413332783375108e-07, "loss": 0.9066, "step": 43425 }, { "epoch": 3.4032915360501566, "grad_norm": 9.772819519042969, "learning_rate": 5.411937341125628e-07, "loss": 2.7923, "step": 43426 }, { "epoch": 3.403369905956113, "grad_norm": 4.389063835144043, "learning_rate": 5.410542068466246e-07, "loss": 1.4781, "step": 43427 }, { "epoch": 3.4034482758620688, "grad_norm": 3.677337884902954, "learning_rate": 5.409146965402274e-07, "loss": 0.7877, "step": 43428 }, { "epoch": 3.403526645768025, "grad_norm": 2.650454521179199, "learning_rate": 5.40775203193899e-07, "loss": 1.3533, "step": 43429 }, { "epoch": 3.403605015673981, "grad_norm": 4.069940567016602, "learning_rate": 5.406357268081714e-07, "loss": 1.2824, "step": 43430 }, { "epoch": 3.4036833855799373, "grad_norm": 2.543635845184326, "learning_rate": 5.40496267383574e-07, "loss": 1.2317, "step": 43431 }, { "epoch": 3.4037617554858937, "grad_norm": 3.2418603897094727, "learning_rate": 5.403568249206403e-07, "loss": 1.1918, "step": 43432 }, { "epoch": 3.4038401253918495, "grad_norm": 4.938792705535889, "learning_rate": 5.402173994198973e-07, "loss": 1.1076, "step": 43433 }, { "epoch": 3.4039184952978054, "grad_norm": 2.640681028366089, "learning_rate": 5.40077990881877e-07, "loss": 1.253, "step": 43434 }, { "epoch": 3.4039968652037618, "grad_norm": 4.2195024490356445, "learning_rate": 5.399385993071093e-07, "loss": 1.4169, "step": 43435 }, { "epoch": 3.404075235109718, "grad_norm": 3.6377806663513184, "learning_rate": 5.397992246961242e-07, "loss": 1.5355, "step": 43436 }, { "epoch": 3.404153605015674, "grad_norm": 2.7289621829986572, "learning_rate": 5.396598670494518e-07, "loss": 1.1368, "step": 43437 }, { "epoch": 3.40423197492163, "grad_norm": 3.141575574874878, "learning_rate": 5.395205263676224e-07, "loss": 1.1031, "step": 43438 }, { "epoch": 3.404310344827586, "grad_norm": 4.131486415863037, "learning_rate": 5.393812026511669e-07, "loss": 1.3964, "step": 43439 }, { "epoch": 3.4043887147335425, "grad_norm": 5.750921726226807, "learning_rate": 5.392418959006123e-07, "loss": 1.5284, "step": 43440 }, { "epoch": 3.4044670846394984, "grad_norm": 3.5516746044158936, "learning_rate": 5.391026061164912e-07, "loss": 0.8576, "step": 43441 }, { "epoch": 3.4045454545454543, "grad_norm": 6.841161727905273, "learning_rate": 5.389633332993327e-07, "loss": 1.4442, "step": 43442 }, { "epoch": 3.4046238244514107, "grad_norm": 3.2496345043182373, "learning_rate": 5.388240774496673e-07, "loss": 1.6295, "step": 43443 }, { "epoch": 3.404702194357367, "grad_norm": 3.258082866668701, "learning_rate": 5.386848385680221e-07, "loss": 1.0562, "step": 43444 }, { "epoch": 3.404780564263323, "grad_norm": 2.918826103210449, "learning_rate": 5.38545616654928e-07, "loss": 1.2274, "step": 43445 }, { "epoch": 3.4048589341692788, "grad_norm": 4.348113059997559, "learning_rate": 5.384064117109156e-07, "loss": 1.9196, "step": 43446 }, { "epoch": 3.404937304075235, "grad_norm": 6.725711822509766, "learning_rate": 5.382672237365144e-07, "loss": 1.2282, "step": 43447 }, { "epoch": 3.4050156739811914, "grad_norm": 3.8132293224334717, "learning_rate": 5.381280527322525e-07, "loss": 1.5557, "step": 43448 }, { "epoch": 3.4050940438871473, "grad_norm": 3.5423412322998047, "learning_rate": 5.379888986986592e-07, "loss": 1.7571, "step": 43449 }, { "epoch": 3.405172413793103, "grad_norm": 2.971015691757202, "learning_rate": 5.378497616362638e-07, "loss": 1.4141, "step": 43450 }, { "epoch": 3.4052507836990595, "grad_norm": 1.8640930652618408, "learning_rate": 5.377106415455973e-07, "loss": 1.0811, "step": 43451 }, { "epoch": 3.405329153605016, "grad_norm": 2.9771928787231445, "learning_rate": 5.375715384271873e-07, "loss": 1.1947, "step": 43452 }, { "epoch": 3.4054075235109718, "grad_norm": 2.452147960662842, "learning_rate": 5.374324522815627e-07, "loss": 1.3838, "step": 43453 }, { "epoch": 3.405485893416928, "grad_norm": 4.875783443450928, "learning_rate": 5.37293383109253e-07, "loss": 2.1347, "step": 43454 }, { "epoch": 3.405564263322884, "grad_norm": 7.133375644683838, "learning_rate": 5.371543309107874e-07, "loss": 1.0314, "step": 43455 }, { "epoch": 3.4056426332288403, "grad_norm": 3.6417362689971924, "learning_rate": 5.370152956866942e-07, "loss": 1.4797, "step": 43456 }, { "epoch": 3.405721003134796, "grad_norm": 3.3906919956207275, "learning_rate": 5.368762774375025e-07, "loss": 1.2217, "step": 43457 }, { "epoch": 3.4057993730407525, "grad_norm": 3.7987401485443115, "learning_rate": 5.367372761637418e-07, "loss": 1.2382, "step": 43458 }, { "epoch": 3.4058777429467084, "grad_norm": 1.8320058584213257, "learning_rate": 5.365982918659379e-07, "loss": 1.2985, "step": 43459 }, { "epoch": 3.4059561128526648, "grad_norm": 3.7520883083343506, "learning_rate": 5.364593245446231e-07, "loss": 1.1872, "step": 43460 }, { "epoch": 3.4060344827586206, "grad_norm": 6.938948154449463, "learning_rate": 5.363203742003242e-07, "loss": 0.5966, "step": 43461 }, { "epoch": 3.406112852664577, "grad_norm": 2.4961249828338623, "learning_rate": 5.361814408335708e-07, "loss": 1.3359, "step": 43462 }, { "epoch": 3.406191222570533, "grad_norm": 4.747950553894043, "learning_rate": 5.360425244448897e-07, "loss": 1.5726, "step": 43463 }, { "epoch": 3.406269592476489, "grad_norm": 4.813303470611572, "learning_rate": 5.359036250348088e-07, "loss": 1.2277, "step": 43464 }, { "epoch": 3.406347962382445, "grad_norm": 2.3083038330078125, "learning_rate": 5.357647426038587e-07, "loss": 1.4341, "step": 43465 }, { "epoch": 3.4064263322884014, "grad_norm": 3.8556246757507324, "learning_rate": 5.356258771525674e-07, "loss": 1.1882, "step": 43466 }, { "epoch": 3.4065047021943573, "grad_norm": 11.759367942810059, "learning_rate": 5.354870286814611e-07, "loss": 0.9956, "step": 43467 }, { "epoch": 3.4065830721003136, "grad_norm": 3.716979503631592, "learning_rate": 5.353481971910696e-07, "loss": 1.7187, "step": 43468 }, { "epoch": 3.4066614420062695, "grad_norm": 3.090115547180176, "learning_rate": 5.352093826819199e-07, "loss": 1.2439, "step": 43469 }, { "epoch": 3.406739811912226, "grad_norm": 3.3178305625915527, "learning_rate": 5.350705851545401e-07, "loss": 1.2231, "step": 43470 }, { "epoch": 3.4068181818181817, "grad_norm": 7.979538917541504, "learning_rate": 5.349318046094593e-07, "loss": 1.5328, "step": 43471 }, { "epoch": 3.406896551724138, "grad_norm": 2.8673274517059326, "learning_rate": 5.347930410472041e-07, "loss": 1.3171, "step": 43472 }, { "epoch": 3.406974921630094, "grad_norm": 2.9045028686523438, "learning_rate": 5.346542944683025e-07, "loss": 1.3178, "step": 43473 }, { "epoch": 3.4070532915360503, "grad_norm": 2.7690465450286865, "learning_rate": 5.345155648732824e-07, "loss": 1.5851, "step": 43474 }, { "epoch": 3.407131661442006, "grad_norm": 6.360660552978516, "learning_rate": 5.343768522626719e-07, "loss": 1.1967, "step": 43475 }, { "epoch": 3.4072100313479625, "grad_norm": 3.1648452281951904, "learning_rate": 5.34238156636998e-07, "loss": 1.511, "step": 43476 }, { "epoch": 3.4072884012539184, "grad_norm": 4.811052322387695, "learning_rate": 5.34099477996789e-07, "loss": 1.2992, "step": 43477 }, { "epoch": 3.4073667711598747, "grad_norm": 2.4219353199005127, "learning_rate": 5.339608163425702e-07, "loss": 1.3055, "step": 43478 }, { "epoch": 3.4074451410658306, "grad_norm": 3.9226973056793213, "learning_rate": 5.338221716748715e-07, "loss": 1.2827, "step": 43479 }, { "epoch": 3.407523510971787, "grad_norm": 3.158881664276123, "learning_rate": 5.3368354399422e-07, "loss": 1.3652, "step": 43480 }, { "epoch": 3.407601880877743, "grad_norm": 3.107393264770508, "learning_rate": 5.335449333011417e-07, "loss": 1.3545, "step": 43481 }, { "epoch": 3.407680250783699, "grad_norm": 5.020218372344971, "learning_rate": 5.334063395961642e-07, "loss": 1.2158, "step": 43482 }, { "epoch": 3.407758620689655, "grad_norm": 5.684537887573242, "learning_rate": 5.332677628798139e-07, "loss": 0.9941, "step": 43483 }, { "epoch": 3.4078369905956114, "grad_norm": 2.804311513900757, "learning_rate": 5.331292031526203e-07, "loss": 1.6166, "step": 43484 }, { "epoch": 3.4079153605015673, "grad_norm": 4.097561359405518, "learning_rate": 5.329906604151086e-07, "loss": 0.7301, "step": 43485 }, { "epoch": 3.4079937304075236, "grad_norm": 5.137888431549072, "learning_rate": 5.328521346678056e-07, "loss": 1.6155, "step": 43486 }, { "epoch": 3.4080721003134795, "grad_norm": 3.088635206222534, "learning_rate": 5.327136259112387e-07, "loss": 1.0817, "step": 43487 }, { "epoch": 3.408150470219436, "grad_norm": 6.941689491271973, "learning_rate": 5.325751341459339e-07, "loss": 0.8957, "step": 43488 }, { "epoch": 3.4082288401253917, "grad_norm": 3.129774570465088, "learning_rate": 5.324366593724195e-07, "loss": 1.2685, "step": 43489 }, { "epoch": 3.408307210031348, "grad_norm": 2.530089855194092, "learning_rate": 5.322982015912215e-07, "loss": 1.4163, "step": 43490 }, { "epoch": 3.408385579937304, "grad_norm": 10.549518585205078, "learning_rate": 5.321597608028661e-07, "loss": 0.7478, "step": 43491 }, { "epoch": 3.4084639498432603, "grad_norm": 3.3167378902435303, "learning_rate": 5.320213370078797e-07, "loss": 1.0834, "step": 43492 }, { "epoch": 3.408542319749216, "grad_norm": 3.964354991912842, "learning_rate": 5.318829302067901e-07, "loss": 1.5562, "step": 43493 }, { "epoch": 3.4086206896551725, "grad_norm": 6.043534278869629, "learning_rate": 5.31744540400122e-07, "loss": 1.1737, "step": 43494 }, { "epoch": 3.4086990595611284, "grad_norm": 5.14445686340332, "learning_rate": 5.316061675884043e-07, "loss": 0.9645, "step": 43495 }, { "epoch": 3.4087774294670847, "grad_norm": 2.779296636581421, "learning_rate": 5.314678117721594e-07, "loss": 1.198, "step": 43496 }, { "epoch": 3.4088557993730406, "grad_norm": 3.4783825874328613, "learning_rate": 5.313294729519163e-07, "loss": 1.7974, "step": 43497 }, { "epoch": 3.408934169278997, "grad_norm": 3.550570249557495, "learning_rate": 5.311911511282014e-07, "loss": 0.9013, "step": 43498 }, { "epoch": 3.409012539184953, "grad_norm": 4.2038445472717285, "learning_rate": 5.310528463015408e-07, "loss": 1.2753, "step": 43499 }, { "epoch": 3.409090909090909, "grad_norm": 4.5109686851501465, "learning_rate": 5.309145584724584e-07, "loss": 1.7502, "step": 43500 }, { "epoch": 3.409169278996865, "grad_norm": 5.553399562835693, "learning_rate": 5.307762876414818e-07, "loss": 1.9031, "step": 43501 }, { "epoch": 3.4092476489028214, "grad_norm": 5.214653968811035, "learning_rate": 5.30638033809136e-07, "loss": 1.5827, "step": 43502 }, { "epoch": 3.4093260188087773, "grad_norm": 4.570562362670898, "learning_rate": 5.30499796975949e-07, "loss": 0.858, "step": 43503 }, { "epoch": 3.4094043887147336, "grad_norm": 3.3750226497650146, "learning_rate": 5.303615771424441e-07, "loss": 1.5719, "step": 43504 }, { "epoch": 3.4094827586206895, "grad_norm": 4.34306526184082, "learning_rate": 5.302233743091484e-07, "loss": 1.1826, "step": 43505 }, { "epoch": 3.409561128526646, "grad_norm": 3.1400232315063477, "learning_rate": 5.300851884765868e-07, "loss": 1.4442, "step": 43506 }, { "epoch": 3.4096394984326017, "grad_norm": 3.884706974029541, "learning_rate": 5.299470196452855e-07, "loss": 1.4146, "step": 43507 }, { "epoch": 3.409717868338558, "grad_norm": 5.971269130706787, "learning_rate": 5.298088678157698e-07, "loss": 1.4458, "step": 43508 }, { "epoch": 3.409796238244514, "grad_norm": 7.983255863189697, "learning_rate": 5.296707329885648e-07, "loss": 1.0812, "step": 43509 }, { "epoch": 3.4098746081504703, "grad_norm": 3.3856921195983887, "learning_rate": 5.295326151641966e-07, "loss": 1.8243, "step": 43510 }, { "epoch": 3.409952978056426, "grad_norm": 8.141495704650879, "learning_rate": 5.2939451434319e-07, "loss": 1.7562, "step": 43511 }, { "epoch": 3.4100313479623825, "grad_norm": 2.9467740058898926, "learning_rate": 5.292564305260705e-07, "loss": 1.332, "step": 43512 }, { "epoch": 3.4101097178683384, "grad_norm": 3.976344585418701, "learning_rate": 5.291183637133628e-07, "loss": 1.1655, "step": 43513 }, { "epoch": 3.4101880877742947, "grad_norm": 2.842048168182373, "learning_rate": 5.289803139055938e-07, "loss": 1.1995, "step": 43514 }, { "epoch": 3.4102664576802506, "grad_norm": 2.56254506111145, "learning_rate": 5.288422811032856e-07, "loss": 1.6001, "step": 43515 }, { "epoch": 3.410344827586207, "grad_norm": 3.6207809448242188, "learning_rate": 5.287042653069652e-07, "loss": 1.3709, "step": 43516 }, { "epoch": 3.410423197492163, "grad_norm": 4.601563930511475, "learning_rate": 5.285662665171576e-07, "loss": 0.9321, "step": 43517 }, { "epoch": 3.410501567398119, "grad_norm": 2.9645984172821045, "learning_rate": 5.28428284734388e-07, "loss": 1.2262, "step": 43518 }, { "epoch": 3.410579937304075, "grad_norm": 3.826093912124634, "learning_rate": 5.282903199591794e-07, "loss": 1.1497, "step": 43519 }, { "epoch": 3.4106583072100314, "grad_norm": 3.503493547439575, "learning_rate": 5.281523721920567e-07, "loss": 1.5593, "step": 43520 }, { "epoch": 3.4107366771159873, "grad_norm": 2.560563564300537, "learning_rate": 5.280144414335464e-07, "loss": 1.2632, "step": 43521 }, { "epoch": 3.4108150470219436, "grad_norm": 4.459620952606201, "learning_rate": 5.278765276841735e-07, "loss": 1.1787, "step": 43522 }, { "epoch": 3.4108934169279, "grad_norm": 3.459506034851074, "learning_rate": 5.277386309444605e-07, "loss": 1.5143, "step": 43523 }, { "epoch": 3.410971786833856, "grad_norm": 2.587355613708496, "learning_rate": 5.276007512149317e-07, "loss": 1.3231, "step": 43524 }, { "epoch": 3.4110501567398117, "grad_norm": 13.599176406860352, "learning_rate": 5.27462888496113e-07, "loss": 1.0113, "step": 43525 }, { "epoch": 3.411128526645768, "grad_norm": 4.06707239151001, "learning_rate": 5.273250427885285e-07, "loss": 1.6442, "step": 43526 }, { "epoch": 3.4112068965517244, "grad_norm": 5.826169013977051, "learning_rate": 5.27187214092702e-07, "loss": 1.202, "step": 43527 }, { "epoch": 3.4112852664576803, "grad_norm": 3.6410250663757324, "learning_rate": 5.270494024091577e-07, "loss": 1.4449, "step": 43528 }, { "epoch": 3.411363636363636, "grad_norm": 3.3179008960723877, "learning_rate": 5.269116077384206e-07, "loss": 1.2264, "step": 43529 }, { "epoch": 3.4114420062695925, "grad_norm": 2.0221471786499023, "learning_rate": 5.267738300810137e-07, "loss": 1.061, "step": 43530 }, { "epoch": 3.411520376175549, "grad_norm": 24.159921646118164, "learning_rate": 5.266360694374617e-07, "loss": 0.9749, "step": 43531 }, { "epoch": 3.4115987460815047, "grad_norm": 3.96956729888916, "learning_rate": 5.264983258082884e-07, "loss": 1.1529, "step": 43532 }, { "epoch": 3.4116771159874606, "grad_norm": 2.633335590362549, "learning_rate": 5.263605991940191e-07, "loss": 1.0378, "step": 43533 }, { "epoch": 3.411755485893417, "grad_norm": 1.959987998008728, "learning_rate": 5.262228895951738e-07, "loss": 1.1288, "step": 43534 }, { "epoch": 3.4118338557993733, "grad_norm": 4.173346519470215, "learning_rate": 5.260851970122799e-07, "loss": 1.6033, "step": 43535 }, { "epoch": 3.411912225705329, "grad_norm": 2.6078641414642334, "learning_rate": 5.259475214458609e-07, "loss": 1.3829, "step": 43536 }, { "epoch": 3.411990595611285, "grad_norm": 5.126321315765381, "learning_rate": 5.258098628964387e-07, "loss": 1.6893, "step": 43537 }, { "epoch": 3.4120689655172414, "grad_norm": 7.251439094543457, "learning_rate": 5.256722213645377e-07, "loss": 1.7297, "step": 43538 }, { "epoch": 3.4121473354231977, "grad_norm": 3.2699122428894043, "learning_rate": 5.255345968506798e-07, "loss": 1.3022, "step": 43539 }, { "epoch": 3.4122257053291536, "grad_norm": 2.687157154083252, "learning_rate": 5.253969893553929e-07, "loss": 1.4218, "step": 43540 }, { "epoch": 3.4123040752351095, "grad_norm": 3.535280704498291, "learning_rate": 5.252593988791959e-07, "loss": 1.3986, "step": 43541 }, { "epoch": 3.412382445141066, "grad_norm": 2.314744472503662, "learning_rate": 5.251218254226142e-07, "loss": 1.3289, "step": 43542 }, { "epoch": 3.412460815047022, "grad_norm": 2.3615212440490723, "learning_rate": 5.249842689861707e-07, "loss": 1.3331, "step": 43543 }, { "epoch": 3.412539184952978, "grad_norm": 2.471933603286743, "learning_rate": 5.24846729570388e-07, "loss": 1.5105, "step": 43544 }, { "epoch": 3.412617554858934, "grad_norm": 4.020484447479248, "learning_rate": 5.247092071757903e-07, "loss": 1.0527, "step": 43545 }, { "epoch": 3.4126959247648903, "grad_norm": 3.195111036300659, "learning_rate": 5.245717018029001e-07, "loss": 1.095, "step": 43546 }, { "epoch": 3.4127742946708466, "grad_norm": 4.423580169677734, "learning_rate": 5.244342134522401e-07, "loss": 1.4059, "step": 43547 }, { "epoch": 3.4128526645768025, "grad_norm": 2.6522555351257324, "learning_rate": 5.242967421243334e-07, "loss": 1.1737, "step": 43548 }, { "epoch": 3.412931034482759, "grad_norm": 6.066758155822754, "learning_rate": 5.241592878197032e-07, "loss": 1.3119, "step": 43549 }, { "epoch": 3.4130094043887147, "grad_norm": 2.529881000518799, "learning_rate": 5.240218505388728e-07, "loss": 1.2454, "step": 43550 }, { "epoch": 3.413087774294671, "grad_norm": 5.9224700927734375, "learning_rate": 5.238844302823643e-07, "loss": 1.7063, "step": 43551 }, { "epoch": 3.413166144200627, "grad_norm": 4.055401802062988, "learning_rate": 5.237470270507e-07, "loss": 1.8138, "step": 43552 }, { "epoch": 3.4132445141065832, "grad_norm": 3.0623867511749268, "learning_rate": 5.236096408444014e-07, "loss": 1.3973, "step": 43553 }, { "epoch": 3.413322884012539, "grad_norm": 3.3759586811065674, "learning_rate": 5.234722716639939e-07, "loss": 1.2879, "step": 43554 }, { "epoch": 3.4134012539184955, "grad_norm": 3.9728410243988037, "learning_rate": 5.233349195099996e-07, "loss": 1.1585, "step": 43555 }, { "epoch": 3.4134796238244514, "grad_norm": 3.2518022060394287, "learning_rate": 5.231975843829385e-07, "loss": 1.4935, "step": 43556 }, { "epoch": 3.4135579937304077, "grad_norm": 5.888128757476807, "learning_rate": 5.230602662833345e-07, "loss": 1.7239, "step": 43557 }, { "epoch": 3.4136363636363636, "grad_norm": 2.667405843734741, "learning_rate": 5.229229652117091e-07, "loss": 1.2683, "step": 43558 }, { "epoch": 3.41371473354232, "grad_norm": 3.4526896476745605, "learning_rate": 5.227856811685866e-07, "loss": 1.4663, "step": 43559 }, { "epoch": 3.413793103448276, "grad_norm": 3.6442251205444336, "learning_rate": 5.226484141544874e-07, "loss": 1.3548, "step": 43560 }, { "epoch": 3.413871473354232, "grad_norm": 8.768033027648926, "learning_rate": 5.225111641699332e-07, "loss": 1.2738, "step": 43561 }, { "epoch": 3.413949843260188, "grad_norm": 7.2702555656433105, "learning_rate": 5.223739312154474e-07, "loss": 1.4123, "step": 43562 }, { "epoch": 3.4140282131661444, "grad_norm": 4.534743785858154, "learning_rate": 5.222367152915508e-07, "loss": 1.377, "step": 43563 }, { "epoch": 3.4141065830721002, "grad_norm": 3.9242753982543945, "learning_rate": 5.220995163987657e-07, "loss": 1.5776, "step": 43564 }, { "epoch": 3.4141849529780566, "grad_norm": 2.638784408569336, "learning_rate": 5.219623345376146e-07, "loss": 1.4092, "step": 43565 }, { "epoch": 3.4142633228840125, "grad_norm": 5.279797554016113, "learning_rate": 5.218251697086197e-07, "loss": 1.6055, "step": 43566 }, { "epoch": 3.414341692789969, "grad_norm": 3.6785647869110107, "learning_rate": 5.216880219122994e-07, "loss": 1.3148, "step": 43567 }, { "epoch": 3.4144200626959247, "grad_norm": 4.012095928192139, "learning_rate": 5.215508911491784e-07, "loss": 1.5126, "step": 43568 }, { "epoch": 3.414498432601881, "grad_norm": 3.1269619464874268, "learning_rate": 5.214137774197781e-07, "loss": 1.4774, "step": 43569 }, { "epoch": 3.414576802507837, "grad_norm": 6.164781093597412, "learning_rate": 5.212766807246206e-07, "loss": 1.434, "step": 43570 }, { "epoch": 3.4146551724137932, "grad_norm": 5.2428154945373535, "learning_rate": 5.211396010642244e-07, "loss": 0.9618, "step": 43571 }, { "epoch": 3.414733542319749, "grad_norm": 3.598553419113159, "learning_rate": 5.210025384391121e-07, "loss": 1.3294, "step": 43572 }, { "epoch": 3.4148119122257055, "grad_norm": 3.2560315132141113, "learning_rate": 5.208654928498069e-07, "loss": 0.6659, "step": 43573 }, { "epoch": 3.4148902821316613, "grad_norm": 2.7341859340667725, "learning_rate": 5.207284642968291e-07, "loss": 1.3427, "step": 43574 }, { "epoch": 3.4149686520376177, "grad_norm": 2.437335968017578, "learning_rate": 5.205914527806988e-07, "loss": 1.076, "step": 43575 }, { "epoch": 3.4150470219435736, "grad_norm": 4.546374320983887, "learning_rate": 5.204544583019377e-07, "loss": 1.2962, "step": 43576 }, { "epoch": 3.41512539184953, "grad_norm": 3.7394607067108154, "learning_rate": 5.203174808610667e-07, "loss": 1.1223, "step": 43577 }, { "epoch": 3.415203761755486, "grad_norm": 3.117750644683838, "learning_rate": 5.201805204586086e-07, "loss": 1.4815, "step": 43578 }, { "epoch": 3.415282131661442, "grad_norm": 3.1182053089141846, "learning_rate": 5.200435770950818e-07, "loss": 1.3606, "step": 43579 }, { "epoch": 3.415360501567398, "grad_norm": 2.7041172981262207, "learning_rate": 5.199066507710077e-07, "loss": 1.4466, "step": 43580 }, { "epoch": 3.4154388714733543, "grad_norm": 3.6298606395721436, "learning_rate": 5.197697414869085e-07, "loss": 1.2283, "step": 43581 }, { "epoch": 3.4155172413793102, "grad_norm": 3.6584980487823486, "learning_rate": 5.196328492433033e-07, "loss": 1.551, "step": 43582 }, { "epoch": 3.4155956112852666, "grad_norm": 4.194177150726318, "learning_rate": 5.194959740407135e-07, "loss": 0.9592, "step": 43583 }, { "epoch": 3.4156739811912225, "grad_norm": 2.334458827972412, "learning_rate": 5.193591158796601e-07, "loss": 1.0736, "step": 43584 }, { "epoch": 3.415752351097179, "grad_norm": 4.28802490234375, "learning_rate": 5.19222274760664e-07, "loss": 1.4226, "step": 43585 }, { "epoch": 3.4158307210031347, "grad_norm": 3.094815969467163, "learning_rate": 5.190854506842429e-07, "loss": 1.2294, "step": 43586 }, { "epoch": 3.415909090909091, "grad_norm": 2.779261827468872, "learning_rate": 5.189486436509205e-07, "loss": 1.0219, "step": 43587 }, { "epoch": 3.415987460815047, "grad_norm": 3.0127549171447754, "learning_rate": 5.188118536612152e-07, "loss": 1.3564, "step": 43588 }, { "epoch": 3.4160658307210032, "grad_norm": 3.077706813812256, "learning_rate": 5.186750807156493e-07, "loss": 1.148, "step": 43589 }, { "epoch": 3.416144200626959, "grad_norm": 3.078697919845581, "learning_rate": 5.185383248147408e-07, "loss": 1.3302, "step": 43590 }, { "epoch": 3.4162225705329154, "grad_norm": 6.019638538360596, "learning_rate": 5.184015859590097e-07, "loss": 1.1547, "step": 43591 }, { "epoch": 3.4163009404388713, "grad_norm": 5.437633991241455, "learning_rate": 5.18264864148979e-07, "loss": 1.6078, "step": 43592 }, { "epoch": 3.4163793103448277, "grad_norm": 3.4946799278259277, "learning_rate": 5.181281593851656e-07, "loss": 1.4939, "step": 43593 }, { "epoch": 3.4164576802507836, "grad_norm": 5.528792858123779, "learning_rate": 5.179914716680906e-07, "loss": 1.6514, "step": 43594 }, { "epoch": 3.41653605015674, "grad_norm": 6.976243495941162, "learning_rate": 5.178548009982743e-07, "loss": 0.9706, "step": 43595 }, { "epoch": 3.416614420062696, "grad_norm": 2.7872049808502197, "learning_rate": 5.177181473762361e-07, "loss": 1.2721, "step": 43596 }, { "epoch": 3.416692789968652, "grad_norm": 3.1387553215026855, "learning_rate": 5.175815108024956e-07, "loss": 1.1824, "step": 43597 }, { "epoch": 3.416771159874608, "grad_norm": 4.729670524597168, "learning_rate": 5.174448912775726e-07, "loss": 1.6493, "step": 43598 }, { "epoch": 3.4168495297805643, "grad_norm": 3.5896236896514893, "learning_rate": 5.17308288801987e-07, "loss": 1.3081, "step": 43599 }, { "epoch": 3.41692789968652, "grad_norm": 4.266331672668457, "learning_rate": 5.171717033762585e-07, "loss": 0.9322, "step": 43600 }, { "epoch": 3.4170062695924766, "grad_norm": 3.1290788650512695, "learning_rate": 5.170351350009056e-07, "loss": 1.2839, "step": 43601 }, { "epoch": 3.4170846394984324, "grad_norm": 3.476968765258789, "learning_rate": 5.16898583676449e-07, "loss": 1.3802, "step": 43602 }, { "epoch": 3.4171630094043888, "grad_norm": 2.373915910720825, "learning_rate": 5.167620494034075e-07, "loss": 1.4334, "step": 43603 }, { "epoch": 3.4172413793103447, "grad_norm": 2.4989843368530273, "learning_rate": 5.166255321823005e-07, "loss": 1.2728, "step": 43604 }, { "epoch": 3.417319749216301, "grad_norm": 4.578743934631348, "learning_rate": 5.164890320136467e-07, "loss": 1.3614, "step": 43605 }, { "epoch": 3.417398119122257, "grad_norm": 3.9471499919891357, "learning_rate": 5.163525488979665e-07, "loss": 1.4808, "step": 43606 }, { "epoch": 3.417476489028213, "grad_norm": 3.0462779998779297, "learning_rate": 5.162160828357788e-07, "loss": 0.8634, "step": 43607 }, { "epoch": 3.417554858934169, "grad_norm": 2.996671676635742, "learning_rate": 5.160796338276009e-07, "loss": 1.3901, "step": 43608 }, { "epoch": 3.4176332288401254, "grad_norm": 4.788058757781982, "learning_rate": 5.159432018739524e-07, "loss": 1.2279, "step": 43609 }, { "epoch": 3.4177115987460813, "grad_norm": 2.2894911766052246, "learning_rate": 5.158067869753536e-07, "loss": 1.0655, "step": 43610 }, { "epoch": 3.4177899686520377, "grad_norm": 3.6085569858551025, "learning_rate": 5.156703891323233e-07, "loss": 1.1101, "step": 43611 }, { "epoch": 3.4178683385579935, "grad_norm": 2.8290927410125732, "learning_rate": 5.155340083453792e-07, "loss": 1.4799, "step": 43612 }, { "epoch": 3.41794670846395, "grad_norm": 3.89454984664917, "learning_rate": 5.153976446150399e-07, "loss": 1.0181, "step": 43613 }, { "epoch": 3.4180250783699058, "grad_norm": 3.91731858253479, "learning_rate": 5.152612979418248e-07, "loss": 1.2267, "step": 43614 }, { "epoch": 3.418103448275862, "grad_norm": 2.484384775161743, "learning_rate": 5.151249683262516e-07, "loss": 1.4119, "step": 43615 }, { "epoch": 3.418181818181818, "grad_norm": 5.542044639587402, "learning_rate": 5.149886557688399e-07, "loss": 1.2689, "step": 43616 }, { "epoch": 3.4182601880877743, "grad_norm": 2.960561513900757, "learning_rate": 5.148523602701078e-07, "loss": 1.4595, "step": 43617 }, { "epoch": 3.4183385579937307, "grad_norm": 2.6247291564941406, "learning_rate": 5.147160818305735e-07, "loss": 1.325, "step": 43618 }, { "epoch": 3.4184169278996865, "grad_norm": 2.6548311710357666, "learning_rate": 5.145798204507558e-07, "loss": 1.3119, "step": 43619 }, { "epoch": 3.4184952978056424, "grad_norm": 3.3133938312530518, "learning_rate": 5.144435761311717e-07, "loss": 1.4862, "step": 43620 }, { "epoch": 3.4185736677115988, "grad_norm": 3.6806998252868652, "learning_rate": 5.14307348872341e-07, "loss": 1.3365, "step": 43621 }, { "epoch": 3.418652037617555, "grad_norm": 4.575119495391846, "learning_rate": 5.141711386747822e-07, "loss": 0.7582, "step": 43622 }, { "epoch": 3.418730407523511, "grad_norm": 5.232002258300781, "learning_rate": 5.140349455390098e-07, "loss": 1.3113, "step": 43623 }, { "epoch": 3.418808777429467, "grad_norm": 3.117107629776001, "learning_rate": 5.138987694655457e-07, "loss": 1.2429, "step": 43624 }, { "epoch": 3.418887147335423, "grad_norm": 3.517120122909546, "learning_rate": 5.137626104549065e-07, "loss": 0.9768, "step": 43625 }, { "epoch": 3.4189655172413795, "grad_norm": 4.93919563293457, "learning_rate": 5.136264685076109e-07, "loss": 1.3531, "step": 43626 }, { "epoch": 3.4190438871473354, "grad_norm": 4.591804027557373, "learning_rate": 5.13490343624175e-07, "loss": 1.0398, "step": 43627 }, { "epoch": 3.4191222570532913, "grad_norm": 6.229659080505371, "learning_rate": 5.133542358051163e-07, "loss": 1.4813, "step": 43628 }, { "epoch": 3.4192006269592476, "grad_norm": 2.4326059818267822, "learning_rate": 5.132181450509543e-07, "loss": 1.4267, "step": 43629 }, { "epoch": 3.419278996865204, "grad_norm": 3.1102960109710693, "learning_rate": 5.130820713622076e-07, "loss": 1.6514, "step": 43630 }, { "epoch": 3.41935736677116, "grad_norm": 9.822836875915527, "learning_rate": 5.129460147393906e-07, "loss": 1.0993, "step": 43631 }, { "epoch": 3.4194357366771158, "grad_norm": 2.7131431102752686, "learning_rate": 5.128099751830223e-07, "loss": 1.3526, "step": 43632 }, { "epoch": 3.419514106583072, "grad_norm": 3.642226219177246, "learning_rate": 5.126739526936192e-07, "loss": 0.8645, "step": 43633 }, { "epoch": 3.4195924764890284, "grad_norm": 4.43110466003418, "learning_rate": 5.125379472717018e-07, "loss": 1.3079, "step": 43634 }, { "epoch": 3.4196708463949843, "grad_norm": 4.371575832366943, "learning_rate": 5.124019589177837e-07, "loss": 1.9337, "step": 43635 }, { "epoch": 3.41974921630094, "grad_norm": 2.610426425933838, "learning_rate": 5.122659876323838e-07, "loss": 1.3595, "step": 43636 }, { "epoch": 3.4198275862068965, "grad_norm": 3.6745405197143555, "learning_rate": 5.12130033416019e-07, "loss": 1.0458, "step": 43637 }, { "epoch": 3.419905956112853, "grad_norm": 6.764857769012451, "learning_rate": 5.119940962692066e-07, "loss": 1.4038, "step": 43638 }, { "epoch": 3.4199843260188088, "grad_norm": 2.790443181991577, "learning_rate": 5.11858176192463e-07, "loss": 0.7589, "step": 43639 }, { "epoch": 3.420062695924765, "grad_norm": 4.975246429443359, "learning_rate": 5.117222731863059e-07, "loss": 1.2825, "step": 43640 }, { "epoch": 3.420141065830721, "grad_norm": 6.134016990661621, "learning_rate": 5.115863872512533e-07, "loss": 1.4572, "step": 43641 }, { "epoch": 3.4202194357366773, "grad_norm": 6.6513285636901855, "learning_rate": 5.114505183878182e-07, "loss": 1.3562, "step": 43642 }, { "epoch": 3.420297805642633, "grad_norm": 4.505222320556641, "learning_rate": 5.113146665965213e-07, "loss": 1.5586, "step": 43643 }, { "epoch": 3.4203761755485895, "grad_norm": 3.995882987976074, "learning_rate": 5.111788318778771e-07, "loss": 1.4697, "step": 43644 }, { "epoch": 3.4204545454545454, "grad_norm": 3.9506475925445557, "learning_rate": 5.110430142324047e-07, "loss": 1.3663, "step": 43645 }, { "epoch": 3.4205329153605017, "grad_norm": 7.508758544921875, "learning_rate": 5.10907213660618e-07, "loss": 1.0629, "step": 43646 }, { "epoch": 3.4206112852664576, "grad_norm": 3.3571903705596924, "learning_rate": 5.107714301630335e-07, "loss": 0.9543, "step": 43647 }, { "epoch": 3.420689655172414, "grad_norm": 8.555688858032227, "learning_rate": 5.106356637401704e-07, "loss": 1.0249, "step": 43648 }, { "epoch": 3.42076802507837, "grad_norm": 4.597532749176025, "learning_rate": 5.104999143925421e-07, "loss": 1.5209, "step": 43649 }, { "epoch": 3.420846394984326, "grad_norm": 8.37413501739502, "learning_rate": 5.103641821206667e-07, "loss": 0.9446, "step": 43650 }, { "epoch": 3.420924764890282, "grad_norm": 3.389591693878174, "learning_rate": 5.102284669250595e-07, "loss": 1.6074, "step": 43651 }, { "epoch": 3.4210031347962384, "grad_norm": 4.93125581741333, "learning_rate": 5.10092768806238e-07, "loss": 1.2056, "step": 43652 }, { "epoch": 3.4210815047021943, "grad_norm": 7.590440273284912, "learning_rate": 5.099570877647164e-07, "loss": 1.416, "step": 43653 }, { "epoch": 3.4211598746081506, "grad_norm": 2.481051445007324, "learning_rate": 5.098214238010125e-07, "loss": 1.1741, "step": 43654 }, { "epoch": 3.4212382445141065, "grad_norm": 7.176419734954834, "learning_rate": 5.096857769156411e-07, "loss": 1.5995, "step": 43655 }, { "epoch": 3.421316614420063, "grad_norm": 3.164475917816162, "learning_rate": 5.095501471091191e-07, "loss": 1.3177, "step": 43656 }, { "epoch": 3.4213949843260187, "grad_norm": 2.793386936187744, "learning_rate": 5.094145343819617e-07, "loss": 1.3272, "step": 43657 }, { "epoch": 3.421473354231975, "grad_norm": 4.331988334655762, "learning_rate": 5.092789387346847e-07, "loss": 1.1873, "step": 43658 }, { "epoch": 3.421551724137931, "grad_norm": 4.164708137512207, "learning_rate": 5.091433601678042e-07, "loss": 1.3457, "step": 43659 }, { "epoch": 3.4216300940438873, "grad_norm": 2.5634803771972656, "learning_rate": 5.090077986818365e-07, "loss": 1.1733, "step": 43660 }, { "epoch": 3.421708463949843, "grad_norm": 2.959848403930664, "learning_rate": 5.08872254277295e-07, "loss": 1.165, "step": 43661 }, { "epoch": 3.4217868338557995, "grad_norm": 6.135373592376709, "learning_rate": 5.087367269546972e-07, "loss": 1.7762, "step": 43662 }, { "epoch": 3.4218652037617554, "grad_norm": 3.7361950874328613, "learning_rate": 5.086012167145593e-07, "loss": 1.2789, "step": 43663 }, { "epoch": 3.4219435736677117, "grad_norm": 2.576568841934204, "learning_rate": 5.084657235573942e-07, "loss": 1.2668, "step": 43664 }, { "epoch": 3.4220219435736676, "grad_norm": 3.3340656757354736, "learning_rate": 5.083302474837182e-07, "loss": 1.1829, "step": 43665 }, { "epoch": 3.422100313479624, "grad_norm": 3.6021645069122314, "learning_rate": 5.081947884940464e-07, "loss": 1.7943, "step": 43666 }, { "epoch": 3.42217868338558, "grad_norm": 3.450683355331421, "learning_rate": 5.080593465888961e-07, "loss": 1.7483, "step": 43667 }, { "epoch": 3.422257053291536, "grad_norm": 10.092912673950195, "learning_rate": 5.079239217687803e-07, "loss": 1.1458, "step": 43668 }, { "epoch": 3.422335423197492, "grad_norm": 2.9043798446655273, "learning_rate": 5.077885140342143e-07, "loss": 1.3727, "step": 43669 }, { "epoch": 3.4224137931034484, "grad_norm": 4.690389633178711, "learning_rate": 5.076531233857135e-07, "loss": 1.135, "step": 43670 }, { "epoch": 3.4224921630094043, "grad_norm": 4.950840473175049, "learning_rate": 5.075177498237926e-07, "loss": 1.5419, "step": 43671 }, { "epoch": 3.4225705329153606, "grad_norm": 3.5181643962860107, "learning_rate": 5.073823933489674e-07, "loss": 1.367, "step": 43672 }, { "epoch": 3.4226489028213165, "grad_norm": 3.991006374359131, "learning_rate": 5.072470539617514e-07, "loss": 1.4102, "step": 43673 }, { "epoch": 3.422727272727273, "grad_norm": 3.011895179748535, "learning_rate": 5.0711173166266e-07, "loss": 1.1767, "step": 43674 }, { "epoch": 3.4228056426332287, "grad_norm": 4.49736213684082, "learning_rate": 5.069764264522081e-07, "loss": 1.4177, "step": 43675 }, { "epoch": 3.422884012539185, "grad_norm": 4.470441818237305, "learning_rate": 5.068411383309102e-07, "loss": 1.2688, "step": 43676 }, { "epoch": 3.422962382445141, "grad_norm": 5.119357585906982, "learning_rate": 5.067058672992803e-07, "loss": 1.1109, "step": 43677 }, { "epoch": 3.4230407523510973, "grad_norm": 2.8936192989349365, "learning_rate": 5.065706133578346e-07, "loss": 0.8571, "step": 43678 }, { "epoch": 3.423119122257053, "grad_norm": 2.872710704803467, "learning_rate": 5.064353765070856e-07, "loss": 1.1612, "step": 43679 }, { "epoch": 3.4231974921630095, "grad_norm": 7.294283390045166, "learning_rate": 5.063001567475473e-07, "loss": 0.9023, "step": 43680 }, { "epoch": 3.4232758620689654, "grad_norm": 3.6376380920410156, "learning_rate": 5.061649540797359e-07, "loss": 1.5528, "step": 43681 }, { "epoch": 3.4233542319749217, "grad_norm": 2.782561779022217, "learning_rate": 5.06029768504166e-07, "loss": 1.435, "step": 43682 }, { "epoch": 3.4234326018808776, "grad_norm": 14.269316673278809, "learning_rate": 5.058946000213494e-07, "loss": 0.8315, "step": 43683 }, { "epoch": 3.423510971786834, "grad_norm": 2.9884397983551025, "learning_rate": 5.057594486318013e-07, "loss": 1.1592, "step": 43684 }, { "epoch": 3.42358934169279, "grad_norm": 3.993551254272461, "learning_rate": 5.056243143360351e-07, "loss": 1.1432, "step": 43685 }, { "epoch": 3.423667711598746, "grad_norm": 4.605165481567383, "learning_rate": 5.054891971345677e-07, "loss": 1.9998, "step": 43686 }, { "epoch": 3.423746081504702, "grad_norm": 2.3088176250457764, "learning_rate": 5.053540970279097e-07, "loss": 1.2254, "step": 43687 }, { "epoch": 3.4238244514106584, "grad_norm": 3.7596776485443115, "learning_rate": 5.052190140165763e-07, "loss": 1.701, "step": 43688 }, { "epoch": 3.4239028213166143, "grad_norm": 2.979478597640991, "learning_rate": 5.050839481010805e-07, "loss": 1.1936, "step": 43689 }, { "epoch": 3.4239811912225706, "grad_norm": 3.4279356002807617, "learning_rate": 5.049488992819373e-07, "loss": 1.4024, "step": 43690 }, { "epoch": 3.4240595611285265, "grad_norm": 2.5272927284240723, "learning_rate": 5.048138675596598e-07, "loss": 1.1763, "step": 43691 }, { "epoch": 3.424137931034483, "grad_norm": 3.208627939224243, "learning_rate": 5.046788529347607e-07, "loss": 1.5263, "step": 43692 }, { "epoch": 3.4242163009404387, "grad_norm": 4.048004150390625, "learning_rate": 5.045438554077547e-07, "loss": 1.693, "step": 43693 }, { "epoch": 3.424294670846395, "grad_norm": 2.7291338443756104, "learning_rate": 5.044088749791554e-07, "loss": 1.3895, "step": 43694 }, { "epoch": 3.424373040752351, "grad_norm": 3.075958013534546, "learning_rate": 5.042739116494749e-07, "loss": 1.3366, "step": 43695 }, { "epoch": 3.4244514106583073, "grad_norm": 3.299710512161255, "learning_rate": 5.041389654192275e-07, "loss": 1.409, "step": 43696 }, { "epoch": 3.424529780564263, "grad_norm": 3.2994773387908936, "learning_rate": 5.040040362889271e-07, "loss": 1.0874, "step": 43697 }, { "epoch": 3.4246081504702195, "grad_norm": 3.583111047744751, "learning_rate": 5.038691242590854e-07, "loss": 1.3595, "step": 43698 }, { "epoch": 3.4246865203761754, "grad_norm": 3.4216389656066895, "learning_rate": 5.03734229330215e-07, "loss": 1.2558, "step": 43699 }, { "epoch": 3.4247648902821317, "grad_norm": 7.600898742675781, "learning_rate": 5.035993515028315e-07, "loss": 1.4691, "step": 43700 }, { "epoch": 3.4248432601880876, "grad_norm": 4.623706817626953, "learning_rate": 5.034644907774471e-07, "loss": 1.2111, "step": 43701 }, { "epoch": 3.424921630094044, "grad_norm": 4.96605110168457, "learning_rate": 5.033296471545734e-07, "loss": 1.0848, "step": 43702 }, { "epoch": 3.425, "grad_norm": 3.535092830657959, "learning_rate": 5.031948206347242e-07, "loss": 1.4578, "step": 43703 }, { "epoch": 3.425078369905956, "grad_norm": 2.48789381980896, "learning_rate": 5.03060011218412e-07, "loss": 1.1255, "step": 43704 }, { "epoch": 3.425156739811912, "grad_norm": 2.886390447616577, "learning_rate": 5.029252189061496e-07, "loss": 1.161, "step": 43705 }, { "epoch": 3.4252351097178684, "grad_norm": 4.713228225708008, "learning_rate": 5.027904436984504e-07, "loss": 1.8596, "step": 43706 }, { "epoch": 3.4253134796238243, "grad_norm": 2.5945804119110107, "learning_rate": 5.026556855958259e-07, "loss": 0.8023, "step": 43707 }, { "epoch": 3.4253918495297806, "grad_norm": 4.30439567565918, "learning_rate": 5.025209445987894e-07, "loss": 1.1207, "step": 43708 }, { "epoch": 3.4254702194357365, "grad_norm": 3.4553654193878174, "learning_rate": 5.023862207078528e-07, "loss": 1.0284, "step": 43709 }, { "epoch": 3.425548589341693, "grad_norm": 3.7884061336517334, "learning_rate": 5.022515139235296e-07, "loss": 1.3129, "step": 43710 }, { "epoch": 3.4256269592476487, "grad_norm": 2.247476816177368, "learning_rate": 5.021168242463309e-07, "loss": 1.3523, "step": 43711 }, { "epoch": 3.425705329153605, "grad_norm": 4.199777126312256, "learning_rate": 5.019821516767698e-07, "loss": 1.3388, "step": 43712 }, { "epoch": 3.4257836990595614, "grad_norm": 3.770811080932617, "learning_rate": 5.01847496215358e-07, "loss": 1.2902, "step": 43713 }, { "epoch": 3.4258620689655173, "grad_norm": 3.545217990875244, "learning_rate": 5.017128578626085e-07, "loss": 1.4272, "step": 43714 }, { "epoch": 3.425940438871473, "grad_norm": 3.537069082260132, "learning_rate": 5.015782366190325e-07, "loss": 1.375, "step": 43715 }, { "epoch": 3.4260188087774295, "grad_norm": 5.592393398284912, "learning_rate": 5.01443632485143e-07, "loss": 1.8874, "step": 43716 }, { "epoch": 3.426097178683386, "grad_norm": 3.2489678859710693, "learning_rate": 5.0130904546145e-07, "loss": 1.3521, "step": 43717 }, { "epoch": 3.4261755485893417, "grad_norm": 2.8664886951446533, "learning_rate": 5.011744755484671e-07, "loss": 1.5332, "step": 43718 }, { "epoch": 3.4262539184952976, "grad_norm": 2.5209686756134033, "learning_rate": 5.010399227467072e-07, "loss": 1.3506, "step": 43719 }, { "epoch": 3.426332288401254, "grad_norm": 2.380293846130371, "learning_rate": 5.009053870566793e-07, "loss": 1.3672, "step": 43720 }, { "epoch": 3.4264106583072103, "grad_norm": 3.7631430625915527, "learning_rate": 5.007708684788965e-07, "loss": 1.37, "step": 43721 }, { "epoch": 3.426489028213166, "grad_norm": 3.108276844024658, "learning_rate": 5.006363670138692e-07, "loss": 1.1852, "step": 43722 }, { "epoch": 3.426567398119122, "grad_norm": 2.9378981590270996, "learning_rate": 5.005018826621127e-07, "loss": 1.0897, "step": 43723 }, { "epoch": 3.4266457680250784, "grad_norm": 3.8241820335388184, "learning_rate": 5.003674154241345e-07, "loss": 1.3883, "step": 43724 }, { "epoch": 3.4267241379310347, "grad_norm": 2.94468092918396, "learning_rate": 5.002329653004473e-07, "loss": 1.536, "step": 43725 }, { "epoch": 3.4268025078369906, "grad_norm": 2.7992053031921387, "learning_rate": 5.00098532291563e-07, "loss": 1.6944, "step": 43726 }, { "epoch": 3.4268808777429465, "grad_norm": 6.486928462982178, "learning_rate": 4.999641163979929e-07, "loss": 1.1941, "step": 43727 }, { "epoch": 3.426959247648903, "grad_norm": 4.962701797485352, "learning_rate": 4.998297176202472e-07, "loss": 1.3121, "step": 43728 }, { "epoch": 3.427037617554859, "grad_norm": 2.916558027267456, "learning_rate": 4.996953359588385e-07, "loss": 1.6301, "step": 43729 }, { "epoch": 3.427115987460815, "grad_norm": 3.4733943939208984, "learning_rate": 4.995609714142769e-07, "loss": 1.5111, "step": 43730 }, { "epoch": 3.427194357366771, "grad_norm": 4.1291937828063965, "learning_rate": 4.994266239870738e-07, "loss": 1.6555, "step": 43731 }, { "epoch": 3.4272727272727272, "grad_norm": 7.1419525146484375, "learning_rate": 4.9929229367774e-07, "loss": 1.7274, "step": 43732 }, { "epoch": 3.4273510971786836, "grad_norm": 2.757319688796997, "learning_rate": 4.991579804867868e-07, "loss": 1.3225, "step": 43733 }, { "epoch": 3.4274294670846395, "grad_norm": 4.081567287445068, "learning_rate": 4.990236844147261e-07, "loss": 1.5242, "step": 43734 }, { "epoch": 3.427507836990596, "grad_norm": 3.4703288078308105, "learning_rate": 4.988894054620663e-07, "loss": 1.2505, "step": 43735 }, { "epoch": 3.4275862068965517, "grad_norm": 2.891456365585327, "learning_rate": 4.987551436293181e-07, "loss": 1.1854, "step": 43736 }, { "epoch": 3.427664576802508, "grad_norm": 2.920302629470825, "learning_rate": 4.986208989169944e-07, "loss": 0.7225, "step": 43737 }, { "epoch": 3.427742946708464, "grad_norm": 5.827083587646484, "learning_rate": 4.984866713256059e-07, "loss": 1.2488, "step": 43738 }, { "epoch": 3.4278213166144202, "grad_norm": 4.3548970222473145, "learning_rate": 4.983524608556611e-07, "loss": 1.4425, "step": 43739 }, { "epoch": 3.427899686520376, "grad_norm": 3.1916654109954834, "learning_rate": 4.982182675076714e-07, "loss": 1.2834, "step": 43740 }, { "epoch": 3.4279780564263325, "grad_norm": 4.650519847869873, "learning_rate": 4.980840912821456e-07, "loss": 1.4801, "step": 43741 }, { "epoch": 3.4280564263322884, "grad_norm": 6.870148658752441, "learning_rate": 4.979499321795978e-07, "loss": 1.3347, "step": 43742 }, { "epoch": 3.4281347962382447, "grad_norm": 2.736701726913452, "learning_rate": 4.978157902005354e-07, "loss": 1.4607, "step": 43743 }, { "epoch": 3.4282131661442006, "grad_norm": 3.2688074111938477, "learning_rate": 4.976816653454692e-07, "loss": 1.2889, "step": 43744 }, { "epoch": 3.428291536050157, "grad_norm": 2.6847219467163086, "learning_rate": 4.975475576149091e-07, "loss": 0.9867, "step": 43745 }, { "epoch": 3.428369905956113, "grad_norm": 2.3711047172546387, "learning_rate": 4.974134670093656e-07, "loss": 1.4696, "step": 43746 }, { "epoch": 3.428448275862069, "grad_norm": 3.077955484390259, "learning_rate": 4.972793935293485e-07, "loss": 1.2267, "step": 43747 }, { "epoch": 3.428526645768025, "grad_norm": 11.915953636169434, "learning_rate": 4.971453371753676e-07, "loss": 1.4152, "step": 43748 }, { "epoch": 3.4286050156739813, "grad_norm": 2.4368913173675537, "learning_rate": 4.970112979479347e-07, "loss": 1.163, "step": 43749 }, { "epoch": 3.4286833855799372, "grad_norm": 2.8125836849212646, "learning_rate": 4.968772758475554e-07, "loss": 1.2793, "step": 43750 }, { "epoch": 3.4287617554858936, "grad_norm": 3.107539653778076, "learning_rate": 4.967432708747433e-07, "loss": 1.2438, "step": 43751 }, { "epoch": 3.4288401253918495, "grad_norm": 6.347934246063232, "learning_rate": 4.966092830300068e-07, "loss": 1.3209, "step": 43752 }, { "epoch": 3.428918495297806, "grad_norm": 2.6275322437286377, "learning_rate": 4.964753123138566e-07, "loss": 1.5675, "step": 43753 }, { "epoch": 3.4289968652037617, "grad_norm": 2.57991361618042, "learning_rate": 4.963413587267996e-07, "loss": 1.2751, "step": 43754 }, { "epoch": 3.429075235109718, "grad_norm": 3.464524745941162, "learning_rate": 4.962074222693464e-07, "loss": 1.2782, "step": 43755 }, { "epoch": 3.429153605015674, "grad_norm": 2.870734930038452, "learning_rate": 4.960735029420077e-07, "loss": 1.4367, "step": 43756 }, { "epoch": 3.4292319749216302, "grad_norm": 2.6566519737243652, "learning_rate": 4.959396007452927e-07, "loss": 1.2897, "step": 43757 }, { "epoch": 3.429310344827586, "grad_norm": 2.4029855728149414, "learning_rate": 4.958057156797097e-07, "loss": 1.1481, "step": 43758 }, { "epoch": 3.4293887147335425, "grad_norm": 3.1223247051239014, "learning_rate": 4.956718477457678e-07, "loss": 1.3708, "step": 43759 }, { "epoch": 3.4294670846394983, "grad_norm": 2.5713629722595215, "learning_rate": 4.955379969439766e-07, "loss": 1.3464, "step": 43760 }, { "epoch": 3.4295454545454547, "grad_norm": 10.211869239807129, "learning_rate": 4.954041632748452e-07, "loss": 1.7468, "step": 43761 }, { "epoch": 3.4296238244514106, "grad_norm": 2.934471607208252, "learning_rate": 4.952703467388831e-07, "loss": 1.3116, "step": 43762 }, { "epoch": 3.429702194357367, "grad_norm": 3.914588451385498, "learning_rate": 4.951365473365982e-07, "loss": 1.6431, "step": 43763 }, { "epoch": 3.429780564263323, "grad_norm": 2.2239136695861816, "learning_rate": 4.950027650685002e-07, "loss": 1.5763, "step": 43764 }, { "epoch": 3.429858934169279, "grad_norm": 3.5922188758850098, "learning_rate": 4.948689999350981e-07, "loss": 1.6284, "step": 43765 }, { "epoch": 3.429937304075235, "grad_norm": 6.3804144859313965, "learning_rate": 4.947352519369003e-07, "loss": 1.2948, "step": 43766 }, { "epoch": 3.4300156739811913, "grad_norm": 2.649104356765747, "learning_rate": 4.946015210744153e-07, "loss": 1.526, "step": 43767 }, { "epoch": 3.4300940438871472, "grad_norm": 3.7737278938293457, "learning_rate": 4.944678073481529e-07, "loss": 1.3634, "step": 43768 }, { "epoch": 3.4301724137931036, "grad_norm": 4.200927734375, "learning_rate": 4.943341107586192e-07, "loss": 1.924, "step": 43769 }, { "epoch": 3.4302507836990594, "grad_norm": 5.626830577850342, "learning_rate": 4.942004313063247e-07, "loss": 0.5506, "step": 43770 }, { "epoch": 3.430329153605016, "grad_norm": 5.112163543701172, "learning_rate": 4.940667689917783e-07, "loss": 1.196, "step": 43771 }, { "epoch": 3.4304075235109717, "grad_norm": 4.881195545196533, "learning_rate": 4.939331238154877e-07, "loss": 1.0005, "step": 43772 }, { "epoch": 3.430485893416928, "grad_norm": 2.405569314956665, "learning_rate": 4.937994957779602e-07, "loss": 1.4649, "step": 43773 }, { "epoch": 3.430564263322884, "grad_norm": 2.201627492904663, "learning_rate": 4.936658848797043e-07, "loss": 1.2696, "step": 43774 }, { "epoch": 3.43064263322884, "grad_norm": 6.041024684906006, "learning_rate": 4.935322911212304e-07, "loss": 1.6123, "step": 43775 }, { "epoch": 3.430721003134796, "grad_norm": 3.5399763584136963, "learning_rate": 4.933987145030439e-07, "loss": 1.279, "step": 43776 }, { "epoch": 3.4307993730407524, "grad_norm": 4.238463878631592, "learning_rate": 4.932651550256545e-07, "loss": 1.3001, "step": 43777 }, { "epoch": 3.4308777429467083, "grad_norm": 3.079846143722534, "learning_rate": 4.931316126895691e-07, "loss": 0.897, "step": 43778 }, { "epoch": 3.4309561128526647, "grad_norm": 3.837238073348999, "learning_rate": 4.929980874952961e-07, "loss": 1.3103, "step": 43779 }, { "epoch": 3.4310344827586206, "grad_norm": 3.472825527191162, "learning_rate": 4.92864579443344e-07, "loss": 1.4567, "step": 43780 }, { "epoch": 3.431112852664577, "grad_norm": 6.492618083953857, "learning_rate": 4.927310885342196e-07, "loss": 1.1119, "step": 43781 }, { "epoch": 3.4311912225705328, "grad_norm": 4.586558818817139, "learning_rate": 4.925976147684313e-07, "loss": 1.3049, "step": 43782 }, { "epoch": 3.431269592476489, "grad_norm": 4.248967170715332, "learning_rate": 4.924641581464861e-07, "loss": 0.8678, "step": 43783 }, { "epoch": 3.431347962382445, "grad_norm": 4.819744110107422, "learning_rate": 4.923307186688925e-07, "loss": 2.1335, "step": 43784 }, { "epoch": 3.4314263322884013, "grad_norm": 2.897156238555908, "learning_rate": 4.921972963361577e-07, "loss": 1.0265, "step": 43785 }, { "epoch": 3.431504702194357, "grad_norm": 3.683992624282837, "learning_rate": 4.920638911487885e-07, "loss": 1.363, "step": 43786 }, { "epoch": 3.4315830721003135, "grad_norm": 3.4031639099121094, "learning_rate": 4.919305031072941e-07, "loss": 1.3272, "step": 43787 }, { "epoch": 3.4316614420062694, "grad_norm": 4.554845809936523, "learning_rate": 4.917971322121784e-07, "loss": 1.3873, "step": 43788 }, { "epoch": 3.4317398119122258, "grad_norm": 2.4421558380126953, "learning_rate": 4.916637784639522e-07, "loss": 1.3702, "step": 43789 }, { "epoch": 3.4318181818181817, "grad_norm": 3.837150812149048, "learning_rate": 4.915304418631223e-07, "loss": 1.4776, "step": 43790 }, { "epoch": 3.431896551724138, "grad_norm": 3.5436346530914307, "learning_rate": 4.913971224101938e-07, "loss": 1.3426, "step": 43791 }, { "epoch": 3.431974921630094, "grad_norm": 5.022534370422363, "learning_rate": 4.912638201056747e-07, "loss": 1.1913, "step": 43792 }, { "epoch": 3.43205329153605, "grad_norm": 3.5960421562194824, "learning_rate": 4.911305349500711e-07, "loss": 1.4137, "step": 43793 }, { "epoch": 3.432131661442006, "grad_norm": 2.8706932067871094, "learning_rate": 4.909972669438934e-07, "loss": 1.5146, "step": 43794 }, { "epoch": 3.4322100313479624, "grad_norm": 4.476129531860352, "learning_rate": 4.908640160876449e-07, "loss": 1.5471, "step": 43795 }, { "epoch": 3.4322884012539183, "grad_norm": 5.160364151000977, "learning_rate": 4.907307823818335e-07, "loss": 1.4798, "step": 43796 }, { "epoch": 3.4323667711598747, "grad_norm": 15.758030891418457, "learning_rate": 4.905975658269668e-07, "loss": 0.8539, "step": 43797 }, { "epoch": 3.4324451410658305, "grad_norm": 4.323237419128418, "learning_rate": 4.904643664235504e-07, "loss": 1.6462, "step": 43798 }, { "epoch": 3.432523510971787, "grad_norm": 3.065035581588745, "learning_rate": 4.903311841720909e-07, "loss": 1.3498, "step": 43799 }, { "epoch": 3.4326018808777428, "grad_norm": 2.650162935256958, "learning_rate": 4.90198019073096e-07, "loss": 1.592, "step": 43800 }, { "epoch": 3.432680250783699, "grad_norm": 3.0904901027679443, "learning_rate": 4.900648711270705e-07, "loss": 1.1994, "step": 43801 }, { "epoch": 3.432758620689655, "grad_norm": 3.2810420989990234, "learning_rate": 4.899317403345228e-07, "loss": 1.4176, "step": 43802 }, { "epoch": 3.4328369905956113, "grad_norm": 5.206850528717041, "learning_rate": 4.897986266959576e-07, "loss": 1.3267, "step": 43803 }, { "epoch": 3.4329153605015676, "grad_norm": 6.033459663391113, "learning_rate": 4.896655302118819e-07, "loss": 1.5733, "step": 43804 }, { "epoch": 3.4329937304075235, "grad_norm": 2.5908138751983643, "learning_rate": 4.89532450882803e-07, "loss": 1.2359, "step": 43805 }, { "epoch": 3.4330721003134794, "grad_norm": 3.0854954719543457, "learning_rate": 4.893993887092241e-07, "loss": 0.8903, "step": 43806 }, { "epoch": 3.4331504702194358, "grad_norm": 2.2023825645446777, "learning_rate": 4.89266343691654e-07, "loss": 0.714, "step": 43807 }, { "epoch": 3.433228840125392, "grad_norm": 2.4157214164733887, "learning_rate": 4.891333158305978e-07, "loss": 0.6296, "step": 43808 }, { "epoch": 3.433307210031348, "grad_norm": 2.858961343765259, "learning_rate": 4.890003051265623e-07, "loss": 1.6295, "step": 43809 }, { "epoch": 3.433385579937304, "grad_norm": 4.371206283569336, "learning_rate": 4.888673115800519e-07, "loss": 0.6928, "step": 43810 }, { "epoch": 3.43346394984326, "grad_norm": 3.4286048412323, "learning_rate": 4.88734335191573e-07, "loss": 1.2235, "step": 43811 }, { "epoch": 3.4335423197492165, "grad_norm": 3.72070050239563, "learning_rate": 4.88601375961631e-07, "loss": 1.5957, "step": 43812 }, { "epoch": 3.4336206896551724, "grad_norm": 2.6341607570648193, "learning_rate": 4.884684338907341e-07, "loss": 1.3933, "step": 43813 }, { "epoch": 3.4336990595611283, "grad_norm": 3.0568296909332275, "learning_rate": 4.883355089793845e-07, "loss": 0.9964, "step": 43814 }, { "epoch": 3.4337774294670846, "grad_norm": 5.559995174407959, "learning_rate": 4.882026012280899e-07, "loss": 1.7005, "step": 43815 }, { "epoch": 3.433855799373041, "grad_norm": 6.213812828063965, "learning_rate": 4.88069710637355e-07, "loss": 1.1388, "step": 43816 }, { "epoch": 3.433934169278997, "grad_norm": 4.989406108856201, "learning_rate": 4.879368372076848e-07, "loss": 0.9646, "step": 43817 }, { "epoch": 3.4340125391849528, "grad_norm": 2.398951768875122, "learning_rate": 4.878039809395863e-07, "loss": 1.3815, "step": 43818 }, { "epoch": 3.434090909090909, "grad_norm": 4.133391857147217, "learning_rate": 4.876711418335634e-07, "loss": 1.5683, "step": 43819 }, { "epoch": 3.4341692789968654, "grad_norm": 7.7863450050354, "learning_rate": 4.875383198901218e-07, "loss": 1.2302, "step": 43820 }, { "epoch": 3.4342476489028213, "grad_norm": 3.3695032596588135, "learning_rate": 4.874055151097668e-07, "loss": 1.2298, "step": 43821 }, { "epoch": 3.434326018808777, "grad_norm": 4.321250915527344, "learning_rate": 4.872727274930034e-07, "loss": 1.5623, "step": 43822 }, { "epoch": 3.4344043887147335, "grad_norm": 2.9064717292785645, "learning_rate": 4.871399570403362e-07, "loss": 1.3563, "step": 43823 }, { "epoch": 3.43448275862069, "grad_norm": 2.9087536334991455, "learning_rate": 4.870072037522721e-07, "loss": 1.0536, "step": 43824 }, { "epoch": 3.4345611285266457, "grad_norm": 4.47636604309082, "learning_rate": 4.868744676293124e-07, "loss": 0.857, "step": 43825 }, { "epoch": 3.4346394984326016, "grad_norm": 3.214315891265869, "learning_rate": 4.867417486719656e-07, "loss": 1.1952, "step": 43826 }, { "epoch": 3.434717868338558, "grad_norm": 6.329776287078857, "learning_rate": 4.866090468807344e-07, "loss": 1.5996, "step": 43827 }, { "epoch": 3.4347962382445143, "grad_norm": 5.450428485870361, "learning_rate": 4.864763622561252e-07, "loss": 0.9084, "step": 43828 }, { "epoch": 3.43487460815047, "grad_norm": 3.9830238819122314, "learning_rate": 4.863436947986411e-07, "loss": 1.7735, "step": 43829 }, { "epoch": 3.4349529780564265, "grad_norm": 2.758234739303589, "learning_rate": 4.862110445087858e-07, "loss": 1.2462, "step": 43830 }, { "epoch": 3.4350313479623824, "grad_norm": 4.521805763244629, "learning_rate": 4.860784113870676e-07, "loss": 1.1218, "step": 43831 }, { "epoch": 3.4351097178683387, "grad_norm": 2.0634100437164307, "learning_rate": 4.859457954339875e-07, "loss": 0.8344, "step": 43832 }, { "epoch": 3.4351880877742946, "grad_norm": 4.265125751495361, "learning_rate": 4.858131966500507e-07, "loss": 1.1032, "step": 43833 }, { "epoch": 3.435266457680251, "grad_norm": 7.500858783721924, "learning_rate": 4.856806150357623e-07, "loss": 1.1964, "step": 43834 }, { "epoch": 3.435344827586207, "grad_norm": 3.7661514282226562, "learning_rate": 4.85548050591626e-07, "loss": 1.6194, "step": 43835 }, { "epoch": 3.435423197492163, "grad_norm": 4.421350479125977, "learning_rate": 4.854155033181463e-07, "loss": 1.5481, "step": 43836 }, { "epoch": 3.435501567398119, "grad_norm": 2.652721643447876, "learning_rate": 4.852829732158271e-07, "loss": 1.5851, "step": 43837 }, { "epoch": 3.4355799373040754, "grad_norm": 3.937896490097046, "learning_rate": 4.851504602851725e-07, "loss": 0.9788, "step": 43838 }, { "epoch": 3.4356583072100313, "grad_norm": 3.479121208190918, "learning_rate": 4.850179645266867e-07, "loss": 0.9347, "step": 43839 }, { "epoch": 3.4357366771159876, "grad_norm": 6.9049072265625, "learning_rate": 4.848854859408731e-07, "loss": 0.8955, "step": 43840 }, { "epoch": 3.4358150470219435, "grad_norm": 5.407714366912842, "learning_rate": 4.847530245282362e-07, "loss": 1.1014, "step": 43841 }, { "epoch": 3.4358934169279, "grad_norm": 2.8543283939361572, "learning_rate": 4.846205802892801e-07, "loss": 1.1241, "step": 43842 }, { "epoch": 3.4359717868338557, "grad_norm": 4.133944988250732, "learning_rate": 4.844881532245083e-07, "loss": 1.5541, "step": 43843 }, { "epoch": 3.436050156739812, "grad_norm": 3.839611768722534, "learning_rate": 4.843557433344226e-07, "loss": 1.5209, "step": 43844 }, { "epoch": 3.436128526645768, "grad_norm": 2.7495219707489014, "learning_rate": 4.842233506195293e-07, "loss": 1.5018, "step": 43845 }, { "epoch": 3.4362068965517243, "grad_norm": 3.361950635910034, "learning_rate": 4.840909750803319e-07, "loss": 1.491, "step": 43846 }, { "epoch": 3.43628526645768, "grad_norm": 3.4377923011779785, "learning_rate": 4.839586167173321e-07, "loss": 1.0603, "step": 43847 }, { "epoch": 3.4363636363636365, "grad_norm": 3.4524855613708496, "learning_rate": 4.838262755310341e-07, "loss": 1.3524, "step": 43848 }, { "epoch": 3.4364420062695924, "grad_norm": 3.193492889404297, "learning_rate": 4.836939515219402e-07, "loss": 1.2322, "step": 43849 }, { "epoch": 3.4365203761755487, "grad_norm": 3.013042449951172, "learning_rate": 4.835616446905566e-07, "loss": 1.0057, "step": 43850 }, { "epoch": 3.4365987460815046, "grad_norm": 4.524013996124268, "learning_rate": 4.834293550373842e-07, "loss": 1.5119, "step": 43851 }, { "epoch": 3.436677115987461, "grad_norm": 5.260529518127441, "learning_rate": 4.832970825629263e-07, "loss": 1.5823, "step": 43852 }, { "epoch": 3.436755485893417, "grad_norm": 3.53108549118042, "learning_rate": 4.831648272676864e-07, "loss": 1.3297, "step": 43853 }, { "epoch": 3.436833855799373, "grad_norm": 3.4972612857818604, "learning_rate": 4.830325891521676e-07, "loss": 1.1767, "step": 43854 }, { "epoch": 3.436912225705329, "grad_norm": 5.052865982055664, "learning_rate": 4.829003682168725e-07, "loss": 1.0253, "step": 43855 }, { "epoch": 3.4369905956112854, "grad_norm": 3.179152488708496, "learning_rate": 4.827681644623044e-07, "loss": 1.1089, "step": 43856 }, { "epoch": 3.4370689655172413, "grad_norm": 6.55452299118042, "learning_rate": 4.826359778889672e-07, "loss": 1.0323, "step": 43857 }, { "epoch": 3.4371473354231976, "grad_norm": 3.262279510498047, "learning_rate": 4.825038084973605e-07, "loss": 1.1476, "step": 43858 }, { "epoch": 3.4372257053291535, "grad_norm": 2.368210792541504, "learning_rate": 4.823716562879898e-07, "loss": 1.2391, "step": 43859 }, { "epoch": 3.43730407523511, "grad_norm": 3.9247055053710938, "learning_rate": 4.822395212613568e-07, "loss": 1.4366, "step": 43860 }, { "epoch": 3.4373824451410657, "grad_norm": 3.7505745887756348, "learning_rate": 4.821074034179651e-07, "loss": 1.178, "step": 43861 }, { "epoch": 3.437460815047022, "grad_norm": 5.136816024780273, "learning_rate": 4.819753027583158e-07, "loss": 1.0488, "step": 43862 }, { "epoch": 3.437539184952978, "grad_norm": 3.1046175956726074, "learning_rate": 4.818432192829103e-07, "loss": 1.3048, "step": 43863 }, { "epoch": 3.4376175548589343, "grad_norm": 2.135951042175293, "learning_rate": 4.817111529922541e-07, "loss": 0.7888, "step": 43864 }, { "epoch": 3.43769592476489, "grad_norm": 5.866610527038574, "learning_rate": 4.815791038868483e-07, "loss": 1.2261, "step": 43865 }, { "epoch": 3.4377742946708465, "grad_norm": 8.610651016235352, "learning_rate": 4.814470719671943e-07, "loss": 1.6663, "step": 43866 }, { "epoch": 3.4378526645768024, "grad_norm": 2.909789800643921, "learning_rate": 4.813150572337949e-07, "loss": 1.4379, "step": 43867 }, { "epoch": 3.4379310344827587, "grad_norm": 2.785212755203247, "learning_rate": 4.811830596871508e-07, "loss": 1.4464, "step": 43868 }, { "epoch": 3.4380094043887146, "grad_norm": 3.3792266845703125, "learning_rate": 4.810510793277673e-07, "loss": 1.3234, "step": 43869 }, { "epoch": 3.438087774294671, "grad_norm": 3.0514333248138428, "learning_rate": 4.809191161561432e-07, "loss": 1.268, "step": 43870 }, { "epoch": 3.438166144200627, "grad_norm": 2.693936824798584, "learning_rate": 4.807871701727818e-07, "loss": 1.6642, "step": 43871 }, { "epoch": 3.438244514106583, "grad_norm": 4.671785831451416, "learning_rate": 4.806552413781851e-07, "loss": 0.6594, "step": 43872 }, { "epoch": 3.438322884012539, "grad_norm": 4.415796279907227, "learning_rate": 4.80523329772854e-07, "loss": 1.3463, "step": 43873 }, { "epoch": 3.4384012539184954, "grad_norm": 3.8134067058563232, "learning_rate": 4.803914353572914e-07, "loss": 1.4063, "step": 43874 }, { "epoch": 3.4384796238244513, "grad_norm": 4.678676605224609, "learning_rate": 4.802595581319979e-07, "loss": 1.1063, "step": 43875 }, { "epoch": 3.4385579937304076, "grad_norm": 4.173871994018555, "learning_rate": 4.801276980974767e-07, "loss": 1.1734, "step": 43876 }, { "epoch": 3.4386363636363635, "grad_norm": 2.4951281547546387, "learning_rate": 4.799958552542261e-07, "loss": 1.2153, "step": 43877 }, { "epoch": 3.43871473354232, "grad_norm": 4.841977119445801, "learning_rate": 4.798640296027507e-07, "loss": 1.0188, "step": 43878 }, { "epoch": 3.4387931034482757, "grad_norm": 2.8429834842681885, "learning_rate": 4.797322211435513e-07, "loss": 1.2115, "step": 43879 }, { "epoch": 3.438871473354232, "grad_norm": 5.554028511047363, "learning_rate": 4.79600429877129e-07, "loss": 1.5998, "step": 43880 }, { "epoch": 3.438949843260188, "grad_norm": 3.8414275646209717, "learning_rate": 4.794686558039847e-07, "loss": 1.19, "step": 43881 }, { "epoch": 3.4390282131661443, "grad_norm": 2.3821845054626465, "learning_rate": 4.793368989246178e-07, "loss": 1.3372, "step": 43882 }, { "epoch": 3.4391065830721, "grad_norm": 7.3011298179626465, "learning_rate": 4.792051592395331e-07, "loss": 1.5743, "step": 43883 }, { "epoch": 3.4391849529780565, "grad_norm": 3.158107042312622, "learning_rate": 4.790734367492306e-07, "loss": 1.5743, "step": 43884 }, { "epoch": 3.4392633228840124, "grad_norm": 7.21506929397583, "learning_rate": 4.789417314542099e-07, "loss": 1.7764, "step": 43885 }, { "epoch": 3.4393416927899687, "grad_norm": 3.3981621265411377, "learning_rate": 4.788100433549725e-07, "loss": 1.3924, "step": 43886 }, { "epoch": 3.4394200626959246, "grad_norm": 2.7218446731567383, "learning_rate": 4.786783724520194e-07, "loss": 1.486, "step": 43887 }, { "epoch": 3.439498432601881, "grad_norm": 4.2745537757873535, "learning_rate": 4.785467187458515e-07, "loss": 1.0321, "step": 43888 }, { "epoch": 3.439576802507837, "grad_norm": 5.3048882484436035, "learning_rate": 4.784150822369699e-07, "loss": 1.7661, "step": 43889 }, { "epoch": 3.439655172413793, "grad_norm": 3.6124789714813232, "learning_rate": 4.782834629258742e-07, "loss": 1.1313, "step": 43890 }, { "epoch": 3.439733542319749, "grad_norm": 4.124989986419678, "learning_rate": 4.781518608130664e-07, "loss": 1.6295, "step": 43891 }, { "epoch": 3.4398119122257054, "grad_norm": 2.7124063968658447, "learning_rate": 4.780202758990459e-07, "loss": 0.8169, "step": 43892 }, { "epoch": 3.4398902821316613, "grad_norm": 5.284334659576416, "learning_rate": 4.778887081843137e-07, "loss": 1.2777, "step": 43893 }, { "epoch": 3.4399686520376176, "grad_norm": 3.491152763366699, "learning_rate": 4.777571576693701e-07, "loss": 0.8934, "step": 43894 }, { "epoch": 3.4400470219435735, "grad_norm": 4.948622226715088, "learning_rate": 4.776256243547167e-07, "loss": 1.397, "step": 43895 }, { "epoch": 3.44012539184953, "grad_norm": 5.553864479064941, "learning_rate": 4.774941082408507e-07, "loss": 1.3051, "step": 43896 }, { "epoch": 3.4402037617554857, "grad_norm": 6.392352104187012, "learning_rate": 4.773626093282752e-07, "loss": 1.3887, "step": 43897 }, { "epoch": 3.440282131661442, "grad_norm": 3.998112678527832, "learning_rate": 4.77231127617489e-07, "loss": 1.4162, "step": 43898 }, { "epoch": 3.4403605015673984, "grad_norm": 4.330088138580322, "learning_rate": 4.770996631089936e-07, "loss": 1.1307, "step": 43899 }, { "epoch": 3.4404388714733543, "grad_norm": 2.345716714859009, "learning_rate": 4.769682158032873e-07, "loss": 1.3815, "step": 43900 }, { "epoch": 3.44051724137931, "grad_norm": 3.465186595916748, "learning_rate": 4.768367857008693e-07, "loss": 1.2993, "step": 43901 }, { "epoch": 3.4405956112852665, "grad_norm": 4.929846286773682, "learning_rate": 4.7670537280224304e-07, "loss": 1.6606, "step": 43902 }, { "epoch": 3.440673981191223, "grad_norm": 8.714282989501953, "learning_rate": 4.7657397710790587e-07, "loss": 1.6711, "step": 43903 }, { "epoch": 3.4407523510971787, "grad_norm": 3.5828747749328613, "learning_rate": 4.764425986183574e-07, "loss": 1.3009, "step": 43904 }, { "epoch": 3.4408307210031346, "grad_norm": 2.738982677459717, "learning_rate": 4.763112373340978e-07, "loss": 1.3904, "step": 43905 }, { "epoch": 3.440909090909091, "grad_norm": 4.166016578674316, "learning_rate": 4.7617989325562666e-07, "loss": 1.055, "step": 43906 }, { "epoch": 3.4409874608150472, "grad_norm": 5.024325847625732, "learning_rate": 4.760485663834435e-07, "loss": 1.6776, "step": 43907 }, { "epoch": 3.441065830721003, "grad_norm": 10.06646728515625, "learning_rate": 4.7591725671804813e-07, "loss": 0.8573, "step": 43908 }, { "epoch": 3.441144200626959, "grad_norm": 4.393306255340576, "learning_rate": 4.7578596425993993e-07, "loss": 0.8924, "step": 43909 }, { "epoch": 3.4412225705329154, "grad_norm": 16.01524543762207, "learning_rate": 4.756546890096181e-07, "loss": 1.6151, "step": 43910 }, { "epoch": 3.4413009404388717, "grad_norm": 5.192572593688965, "learning_rate": 4.755234309675821e-07, "loss": 1.8612, "step": 43911 }, { "epoch": 3.4413793103448276, "grad_norm": 2.7886691093444824, "learning_rate": 4.7539219013433114e-07, "loss": 1.4286, "step": 43912 }, { "epoch": 3.4414576802507835, "grad_norm": 5.945863723754883, "learning_rate": 4.7526096651036466e-07, "loss": 1.4867, "step": 43913 }, { "epoch": 3.44153605015674, "grad_norm": 3.1385927200317383, "learning_rate": 4.751297600961802e-07, "loss": 1.0337, "step": 43914 }, { "epoch": 3.441614420062696, "grad_norm": 3.2499771118164062, "learning_rate": 4.749985708922783e-07, "loss": 1.403, "step": 43915 }, { "epoch": 3.441692789968652, "grad_norm": 4.950897693634033, "learning_rate": 4.7486739889915814e-07, "loss": 1.3411, "step": 43916 }, { "epoch": 3.441771159874608, "grad_norm": 6.486137866973877, "learning_rate": 4.747362441173187e-07, "loss": 1.3653, "step": 43917 }, { "epoch": 3.4418495297805642, "grad_norm": 3.1882529258728027, "learning_rate": 4.746051065472579e-07, "loss": 0.8272, "step": 43918 }, { "epoch": 3.4419278996865206, "grad_norm": 2.943087339401245, "learning_rate": 4.7447398618947316e-07, "loss": 1.1225, "step": 43919 }, { "epoch": 3.4420062695924765, "grad_norm": 3.721133232116699, "learning_rate": 4.743428830444663e-07, "loss": 1.4798, "step": 43920 }, { "epoch": 3.4420846394984324, "grad_norm": 5.894255638122559, "learning_rate": 4.742117971127358e-07, "loss": 1.6223, "step": 43921 }, { "epoch": 3.4421630094043887, "grad_norm": 10.21922779083252, "learning_rate": 4.7408072839477793e-07, "loss": 1.3934, "step": 43922 }, { "epoch": 3.442241379310345, "grad_norm": 3.9594802856445312, "learning_rate": 4.739496768910917e-07, "loss": 1.2229, "step": 43923 }, { "epoch": 3.442319749216301, "grad_norm": 2.262720823287964, "learning_rate": 4.7381864260217625e-07, "loss": 1.3254, "step": 43924 }, { "epoch": 3.4423981191222572, "grad_norm": 4.2637434005737305, "learning_rate": 4.736876255285311e-07, "loss": 1.6416, "step": 43925 }, { "epoch": 3.442476489028213, "grad_norm": 9.740328788757324, "learning_rate": 4.735566256706525e-07, "loss": 1.5515, "step": 43926 }, { "epoch": 3.4425548589341695, "grad_norm": 3.158612012863159, "learning_rate": 4.734256430290396e-07, "loss": 1.2678, "step": 43927 }, { "epoch": 3.4426332288401253, "grad_norm": 2.4942667484283447, "learning_rate": 4.7329467760419077e-07, "loss": 1.4106, "step": 43928 }, { "epoch": 3.4427115987460817, "grad_norm": 4.475394248962402, "learning_rate": 4.731637293966035e-07, "loss": 1.1815, "step": 43929 }, { "epoch": 3.4427899686520376, "grad_norm": 3.2254416942596436, "learning_rate": 4.7303279840677675e-07, "loss": 1.3996, "step": 43930 }, { "epoch": 3.442868338557994, "grad_norm": 3.6958394050598145, "learning_rate": 4.729018846352074e-07, "loss": 1.6282, "step": 43931 }, { "epoch": 3.44294670846395, "grad_norm": 6.734408855438232, "learning_rate": 4.727709880823955e-07, "loss": 0.8498, "step": 43932 }, { "epoch": 3.443025078369906, "grad_norm": 2.897510051727295, "learning_rate": 4.7264010874883525e-07, "loss": 1.3299, "step": 43933 }, { "epoch": 3.443103448275862, "grad_norm": 3.517308235168457, "learning_rate": 4.7250924663502786e-07, "loss": 1.8473, "step": 43934 }, { "epoch": 3.4431818181818183, "grad_norm": 2.4505441188812256, "learning_rate": 4.7237840174146953e-07, "loss": 1.4329, "step": 43935 }, { "epoch": 3.4432601880877742, "grad_norm": 2.4485745429992676, "learning_rate": 4.722475740686594e-07, "loss": 1.4846, "step": 43936 }, { "epoch": 3.4433385579937306, "grad_norm": 3.5471301078796387, "learning_rate": 4.7211676361709313e-07, "loss": 1.2871, "step": 43937 }, { "epoch": 3.4434169278996865, "grad_norm": 4.231844902038574, "learning_rate": 4.7198597038726757e-07, "loss": 1.2368, "step": 43938 }, { "epoch": 3.443495297805643, "grad_norm": 3.160712718963623, "learning_rate": 4.7185519437968286e-07, "loss": 0.719, "step": 43939 }, { "epoch": 3.4435736677115987, "grad_norm": 3.398927927017212, "learning_rate": 4.717244355948364e-07, "loss": 1.0062, "step": 43940 }, { "epoch": 3.443652037617555, "grad_norm": 4.154277801513672, "learning_rate": 4.7159369403322276e-07, "loss": 1.2116, "step": 43941 }, { "epoch": 3.443730407523511, "grad_norm": 3.094378709793091, "learning_rate": 4.7146296969534156e-07, "loss": 1.2524, "step": 43942 }, { "epoch": 3.4438087774294672, "grad_norm": 3.237485885620117, "learning_rate": 4.713322625816885e-07, "loss": 1.4536, "step": 43943 }, { "epoch": 3.443887147335423, "grad_norm": 3.3908531665802, "learning_rate": 4.712015726927616e-07, "loss": 1.047, "step": 43944 }, { "epoch": 3.4439655172413794, "grad_norm": 3.314744234085083, "learning_rate": 4.710709000290581e-07, "loss": 1.9597, "step": 43945 }, { "epoch": 3.4440438871473353, "grad_norm": 4.283830165863037, "learning_rate": 4.7094024459107435e-07, "loss": 1.1741, "step": 43946 }, { "epoch": 3.4441222570532917, "grad_norm": 24.218883514404297, "learning_rate": 4.708096063793077e-07, "loss": 1.3472, "step": 43947 }, { "epoch": 3.4442006269592476, "grad_norm": 2.373418092727661, "learning_rate": 4.7067898539425503e-07, "loss": 1.3369, "step": 43948 }, { "epoch": 3.444278996865204, "grad_norm": 4.737827301025391, "learning_rate": 4.7054838163641317e-07, "loss": 0.9383, "step": 43949 }, { "epoch": 3.44435736677116, "grad_norm": 3.598468065261841, "learning_rate": 4.704177951062788e-07, "loss": 1.5819, "step": 43950 }, { "epoch": 3.444435736677116, "grad_norm": 4.009315013885498, "learning_rate": 4.702872258043495e-07, "loss": 1.8007, "step": 43951 }, { "epoch": 3.444514106583072, "grad_norm": 4.828452110290527, "learning_rate": 4.7015667373111916e-07, "loss": 1.7048, "step": 43952 }, { "epoch": 3.4445924764890283, "grad_norm": 6.709441661834717, "learning_rate": 4.700261388870869e-07, "loss": 1.6675, "step": 43953 }, { "epoch": 3.444670846394984, "grad_norm": 2.5443131923675537, "learning_rate": 4.6989562127274846e-07, "loss": 1.2683, "step": 43954 }, { "epoch": 3.4447492163009406, "grad_norm": 2.8542189598083496, "learning_rate": 4.697651208886017e-07, "loss": 1.4317, "step": 43955 }, { "epoch": 3.4448275862068964, "grad_norm": 2.695340394973755, "learning_rate": 4.6963463773514015e-07, "loss": 1.454, "step": 43956 }, { "epoch": 3.4449059561128528, "grad_norm": 2.4646482467651367, "learning_rate": 4.695041718128607e-07, "loss": 1.7447, "step": 43957 }, { "epoch": 3.4449843260188087, "grad_norm": 7.37885856628418, "learning_rate": 4.6937372312226225e-07, "loss": 1.4558, "step": 43958 }, { "epoch": 3.445062695924765, "grad_norm": 7.225437641143799, "learning_rate": 4.6924329166383784e-07, "loss": 0.9409, "step": 43959 }, { "epoch": 3.445141065830721, "grad_norm": 4.3776679039001465, "learning_rate": 4.6911287743808486e-07, "loss": 1.518, "step": 43960 }, { "epoch": 3.445219435736677, "grad_norm": 2.5459349155426025, "learning_rate": 4.689824804454995e-07, "loss": 0.8928, "step": 43961 }, { "epoch": 3.445297805642633, "grad_norm": 6.5181965827941895, "learning_rate": 4.688521006865776e-07, "loss": 1.1291, "step": 43962 }, { "epoch": 3.4453761755485894, "grad_norm": 2.194089889526367, "learning_rate": 4.687217381618148e-07, "loss": 1.0337, "step": 43963 }, { "epoch": 3.4454545454545453, "grad_norm": 6.834397792816162, "learning_rate": 4.685913928717073e-07, "loss": 1.1945, "step": 43964 }, { "epoch": 3.4455329153605017, "grad_norm": 4.0492377281188965, "learning_rate": 4.6846106481675035e-07, "loss": 1.4564, "step": 43965 }, { "epoch": 3.4456112852664575, "grad_norm": 2.3059303760528564, "learning_rate": 4.683307539974402e-07, "loss": 0.9699, "step": 43966 }, { "epoch": 3.445689655172414, "grad_norm": 3.656463623046875, "learning_rate": 4.68200460414272e-07, "loss": 1.179, "step": 43967 }, { "epoch": 3.4457680250783698, "grad_norm": 6.557322025299072, "learning_rate": 4.6807018406774153e-07, "loss": 1.2088, "step": 43968 }, { "epoch": 3.445846394984326, "grad_norm": 8.078330993652344, "learning_rate": 4.679399249583455e-07, "loss": 1.6911, "step": 43969 }, { "epoch": 3.445924764890282, "grad_norm": 2.83034610748291, "learning_rate": 4.678096830865775e-07, "loss": 1.3414, "step": 43970 }, { "epoch": 3.4460031347962383, "grad_norm": 2.9581658840179443, "learning_rate": 4.6767945845293203e-07, "loss": 1.3881, "step": 43971 }, { "epoch": 3.446081504702194, "grad_norm": 4.001224994659424, "learning_rate": 4.6754925105790774e-07, "loss": 1.7381, "step": 43972 }, { "epoch": 3.4461598746081505, "grad_norm": 8.234052658081055, "learning_rate": 4.674190609019985e-07, "loss": 1.7475, "step": 43973 }, { "epoch": 3.4462382445141064, "grad_norm": 4.117392539978027, "learning_rate": 4.6728888798569794e-07, "loss": 1.304, "step": 43974 }, { "epoch": 3.4463166144200628, "grad_norm": 3.5045719146728516, "learning_rate": 4.6715873230950225e-07, "loss": 1.3856, "step": 43975 }, { "epoch": 3.4463949843260187, "grad_norm": 6.783419132232666, "learning_rate": 4.6702859387390607e-07, "loss": 1.9666, "step": 43976 }, { "epoch": 3.446473354231975, "grad_norm": 2.424131393432617, "learning_rate": 4.6689847267940623e-07, "loss": 1.4705, "step": 43977 }, { "epoch": 3.446551724137931, "grad_norm": 4.152816295623779, "learning_rate": 4.6676836872649566e-07, "loss": 1.261, "step": 43978 }, { "epoch": 3.446630094043887, "grad_norm": 2.7988643646240234, "learning_rate": 4.6663828201567007e-07, "loss": 1.1355, "step": 43979 }, { "epoch": 3.446708463949843, "grad_norm": 5.092649936676025, "learning_rate": 4.665082125474235e-07, "loss": 1.0364, "step": 43980 }, { "epoch": 3.4467868338557994, "grad_norm": 3.4046390056610107, "learning_rate": 4.6637816032225116e-07, "loss": 1.3528, "step": 43981 }, { "epoch": 3.4468652037617553, "grad_norm": 4.7708964347839355, "learning_rate": 4.6624812534064755e-07, "loss": 1.2491, "step": 43982 }, { "epoch": 3.4469435736677116, "grad_norm": 2.7230896949768066, "learning_rate": 4.6611810760310796e-07, "loss": 0.8311, "step": 43983 }, { "epoch": 3.4470219435736675, "grad_norm": 3.151311159133911, "learning_rate": 4.6598810711012685e-07, "loss": 1.0121, "step": 43984 }, { "epoch": 3.447100313479624, "grad_norm": 3.382976531982422, "learning_rate": 4.658581238621962e-07, "loss": 1.6368, "step": 43985 }, { "epoch": 3.4471786833855798, "grad_norm": 4.014570236206055, "learning_rate": 4.657281578598133e-07, "loss": 1.5173, "step": 43986 }, { "epoch": 3.447257053291536, "grad_norm": 3.1301095485687256, "learning_rate": 4.655982091034722e-07, "loss": 1.1109, "step": 43987 }, { "epoch": 3.447335423197492, "grad_norm": 3.8932230472564697, "learning_rate": 4.6546827759366696e-07, "loss": 1.6302, "step": 43988 }, { "epoch": 3.4474137931034483, "grad_norm": 17.02330780029297, "learning_rate": 4.6533836333089e-07, "loss": 1.5395, "step": 43989 }, { "epoch": 3.447492163009404, "grad_norm": 2.7118899822235107, "learning_rate": 4.652084663156364e-07, "loss": 1.3492, "step": 43990 }, { "epoch": 3.4475705329153605, "grad_norm": 3.100053310394287, "learning_rate": 4.650785865484014e-07, "loss": 1.2791, "step": 43991 }, { "epoch": 3.4476489028213164, "grad_norm": 5.335214138031006, "learning_rate": 4.6494872402967905e-07, "loss": 1.488, "step": 43992 }, { "epoch": 3.4477272727272728, "grad_norm": 2.955533742904663, "learning_rate": 4.6481887875996115e-07, "loss": 1.3843, "step": 43993 }, { "epoch": 3.447805642633229, "grad_norm": 4.8953375816345215, "learning_rate": 4.6468905073974337e-07, "loss": 1.7263, "step": 43994 }, { "epoch": 3.447884012539185, "grad_norm": 4.072354793548584, "learning_rate": 4.645592399695176e-07, "loss": 1.3204, "step": 43995 }, { "epoch": 3.447962382445141, "grad_norm": 5.042360305786133, "learning_rate": 4.644294464497806e-07, "loss": 1.7924, "step": 43996 }, { "epoch": 3.448040752351097, "grad_norm": 4.0112223625183105, "learning_rate": 4.642996701810237e-07, "loss": 0.8982, "step": 43997 }, { "epoch": 3.4481191222570535, "grad_norm": 6.453192234039307, "learning_rate": 4.6416991116374153e-07, "loss": 1.1014, "step": 43998 }, { "epoch": 3.4481974921630094, "grad_norm": 4.083588123321533, "learning_rate": 4.64040169398427e-07, "loss": 1.5444, "step": 43999 }, { "epoch": 3.4482758620689653, "grad_norm": 2.4624500274658203, "learning_rate": 4.639104448855736e-07, "loss": 1.2302, "step": 44000 }, { "epoch": 3.4483542319749216, "grad_norm": 6.25860595703125, "learning_rate": 4.637807376256753e-07, "loss": 1.6724, "step": 44001 }, { "epoch": 3.448432601880878, "grad_norm": 3.58479642868042, "learning_rate": 4.6365104761922517e-07, "loss": 1.4037, "step": 44002 }, { "epoch": 3.448510971786834, "grad_norm": 4.279057502746582, "learning_rate": 4.635213748667161e-07, "loss": 1.618, "step": 44003 }, { "epoch": 3.4485893416927897, "grad_norm": 2.9113996028900146, "learning_rate": 4.6339171936864214e-07, "loss": 1.2681, "step": 44004 }, { "epoch": 3.448667711598746, "grad_norm": 9.157269477844238, "learning_rate": 4.632620811254956e-07, "loss": 1.7159, "step": 44005 }, { "epoch": 3.4487460815047024, "grad_norm": 3.1896719932556152, "learning_rate": 4.6313246013777014e-07, "loss": 1.1483, "step": 44006 }, { "epoch": 3.4488244514106583, "grad_norm": 3.2371349334716797, "learning_rate": 4.630028564059591e-07, "loss": 1.4749, "step": 44007 }, { "epoch": 3.448902821316614, "grad_norm": 4.252913475036621, "learning_rate": 4.628732699305538e-07, "loss": 1.0654, "step": 44008 }, { "epoch": 3.4489811912225705, "grad_norm": 3.4454710483551025, "learning_rate": 4.6274370071204777e-07, "loss": 1.1337, "step": 44009 }, { "epoch": 3.449059561128527, "grad_norm": 1.931801676750183, "learning_rate": 4.626141487509356e-07, "loss": 1.2016, "step": 44010 }, { "epoch": 3.4491379310344827, "grad_norm": 3.1696367263793945, "learning_rate": 4.6248461404770797e-07, "loss": 1.4018, "step": 44011 }, { "epoch": 3.4492163009404386, "grad_norm": 5.495288848876953, "learning_rate": 4.6235509660285835e-07, "loss": 0.8877, "step": 44012 }, { "epoch": 3.449294670846395, "grad_norm": 7.949513912200928, "learning_rate": 4.6222559641687927e-07, "loss": 0.8474, "step": 44013 }, { "epoch": 3.4493730407523513, "grad_norm": 5.159499168395996, "learning_rate": 4.6209611349026296e-07, "loss": 1.5136, "step": 44014 }, { "epoch": 3.449451410658307, "grad_norm": 3.2626891136169434, "learning_rate": 4.6196664782350244e-07, "loss": 1.1106, "step": 44015 }, { "epoch": 3.4495297805642635, "grad_norm": 3.3914124965667725, "learning_rate": 4.6183719941709005e-07, "loss": 1.3045, "step": 44016 }, { "epoch": 3.4496081504702194, "grad_norm": 2.853989362716675, "learning_rate": 4.6170776827151765e-07, "loss": 0.6438, "step": 44017 }, { "epoch": 3.4496865203761757, "grad_norm": 2.1611545085906982, "learning_rate": 4.6157835438727817e-07, "loss": 1.4591, "step": 44018 }, { "epoch": 3.4497648902821316, "grad_norm": 7.8467488288879395, "learning_rate": 4.614489577648634e-07, "loss": 1.265, "step": 44019 }, { "epoch": 3.449843260188088, "grad_norm": 3.8995361328125, "learning_rate": 4.613195784047653e-07, "loss": 1.0846, "step": 44020 }, { "epoch": 3.449921630094044, "grad_norm": 4.328240871429443, "learning_rate": 4.611902163074766e-07, "loss": 0.4747, "step": 44021 }, { "epoch": 3.45, "grad_norm": 3.729158878326416, "learning_rate": 4.610608714734893e-07, "loss": 1.3666, "step": 44022 }, { "epoch": 3.450078369905956, "grad_norm": 2.7165896892547607, "learning_rate": 4.609315439032952e-07, "loss": 1.4353, "step": 44023 }, { "epoch": 3.4501567398119124, "grad_norm": 6.633108615875244, "learning_rate": 4.608022335973855e-07, "loss": 0.8633, "step": 44024 }, { "epoch": 3.4502351097178683, "grad_norm": 3.1837360858917236, "learning_rate": 4.6067294055625377e-07, "loss": 1.5223, "step": 44025 }, { "epoch": 3.4503134796238246, "grad_norm": 4.693160057067871, "learning_rate": 4.6054366478039017e-07, "loss": 1.538, "step": 44026 }, { "epoch": 3.4503918495297805, "grad_norm": 2.8747780323028564, "learning_rate": 4.6041440627028533e-07, "loss": 1.224, "step": 44027 }, { "epoch": 3.450470219435737, "grad_norm": 3.0585858821868896, "learning_rate": 4.602851650264334e-07, "loss": 1.6349, "step": 44028 }, { "epoch": 3.4505485893416927, "grad_norm": 13.039108276367188, "learning_rate": 4.601559410493262e-07, "loss": 1.8914, "step": 44029 }, { "epoch": 3.450626959247649, "grad_norm": 4.014522075653076, "learning_rate": 4.600267343394532e-07, "loss": 1.4149, "step": 44030 }, { "epoch": 3.450705329153605, "grad_norm": 3.431042194366455, "learning_rate": 4.5989754489730585e-07, "loss": 1.3097, "step": 44031 }, { "epoch": 3.4507836990595613, "grad_norm": 7.49299430847168, "learning_rate": 4.5976837272337595e-07, "loss": 0.6335, "step": 44032 }, { "epoch": 3.450862068965517, "grad_norm": 3.2554430961608887, "learning_rate": 4.596392178181569e-07, "loss": 1.2661, "step": 44033 }, { "epoch": 3.4509404388714735, "grad_norm": 4.308168888092041, "learning_rate": 4.5951008018213727e-07, "loss": 1.3093, "step": 44034 }, { "epoch": 3.4510188087774294, "grad_norm": 3.814556121826172, "learning_rate": 4.593809598158094e-07, "loss": 0.9822, "step": 44035 }, { "epoch": 3.4510971786833857, "grad_norm": 5.371602535247803, "learning_rate": 4.592518567196641e-07, "loss": 0.803, "step": 44036 }, { "epoch": 3.4511755485893416, "grad_norm": 2.5572237968444824, "learning_rate": 4.5912277089419197e-07, "loss": 1.288, "step": 44037 }, { "epoch": 3.451253918495298, "grad_norm": 2.4333655834198, "learning_rate": 4.5899370233988493e-07, "loss": 1.283, "step": 44038 }, { "epoch": 3.451332288401254, "grad_norm": 3.7854158878326416, "learning_rate": 4.588646510572331e-07, "loss": 1.3717, "step": 44039 }, { "epoch": 3.45141065830721, "grad_norm": 2.6664984226226807, "learning_rate": 4.587356170467289e-07, "loss": 1.4215, "step": 44040 }, { "epoch": 3.451489028213166, "grad_norm": 2.8808720111846924, "learning_rate": 4.586066003088596e-07, "loss": 1.097, "step": 44041 }, { "epoch": 3.4515673981191224, "grad_norm": 2.873021364212036, "learning_rate": 4.584776008441194e-07, "loss": 1.4945, "step": 44042 }, { "epoch": 3.4516457680250783, "grad_norm": 2.6981639862060547, "learning_rate": 4.5834861865299786e-07, "loss": 1.6633, "step": 44043 }, { "epoch": 3.4517241379310346, "grad_norm": 6.900644302368164, "learning_rate": 4.582196537359862e-07, "loss": 1.4913, "step": 44044 }, { "epoch": 3.4518025078369905, "grad_norm": 8.49429702758789, "learning_rate": 4.5809070609357354e-07, "loss": 1.1897, "step": 44045 }, { "epoch": 3.451880877742947, "grad_norm": 3.5286757946014404, "learning_rate": 4.579617757262494e-07, "loss": 1.0574, "step": 44046 }, { "epoch": 3.4519592476489027, "grad_norm": 4.3279876708984375, "learning_rate": 4.5783286263450743e-07, "loss": 1.2516, "step": 44047 }, { "epoch": 3.452037617554859, "grad_norm": 3.2559499740600586, "learning_rate": 4.5770396681883656e-07, "loss": 1.254, "step": 44048 }, { "epoch": 3.452115987460815, "grad_norm": 3.8884897232055664, "learning_rate": 4.575750882797259e-07, "loss": 1.3656, "step": 44049 }, { "epoch": 3.4521943573667713, "grad_norm": 3.2471370697021484, "learning_rate": 4.574462270176666e-07, "loss": 1.3219, "step": 44050 }, { "epoch": 3.452272727272727, "grad_norm": 2.5695924758911133, "learning_rate": 4.573173830331479e-07, "loss": 1.0673, "step": 44051 }, { "epoch": 3.4523510971786835, "grad_norm": 3.0626087188720703, "learning_rate": 4.5718855632666205e-07, "loss": 1.0983, "step": 44052 }, { "epoch": 3.4524294670846394, "grad_norm": 3.0912632942199707, "learning_rate": 4.5705974689869706e-07, "loss": 1.2294, "step": 44053 }, { "epoch": 3.4525078369905957, "grad_norm": 4.970440864562988, "learning_rate": 4.5693095474974303e-07, "loss": 1.4849, "step": 44054 }, { "epoch": 3.4525862068965516, "grad_norm": 6.322262763977051, "learning_rate": 4.568021798802907e-07, "loss": 1.3875, "step": 44055 }, { "epoch": 3.452664576802508, "grad_norm": 4.235594749450684, "learning_rate": 4.5667342229082865e-07, "loss": 1.4908, "step": 44056 }, { "epoch": 3.452742946708464, "grad_norm": 6.2207841873168945, "learning_rate": 4.565446819818481e-07, "loss": 1.3795, "step": 44057 }, { "epoch": 3.45282131661442, "grad_norm": 3.334566354751587, "learning_rate": 4.564159589538375e-07, "loss": 1.2761, "step": 44058 }, { "epoch": 3.452899686520376, "grad_norm": 3.389287233352661, "learning_rate": 4.562872532072876e-07, "loss": 1.3335, "step": 44059 }, { "epoch": 3.4529780564263324, "grad_norm": 3.4830846786499023, "learning_rate": 4.5615856474268527e-07, "loss": 1.1657, "step": 44060 }, { "epoch": 3.4530564263322883, "grad_norm": 9.576042175292969, "learning_rate": 4.560298935605234e-07, "loss": 1.4114, "step": 44061 }, { "epoch": 3.4531347962382446, "grad_norm": 3.7425408363342285, "learning_rate": 4.559012396612894e-07, "loss": 1.2755, "step": 44062 }, { "epoch": 3.4532131661442005, "grad_norm": 4.201177597045898, "learning_rate": 4.55772603045474e-07, "loss": 1.125, "step": 44063 }, { "epoch": 3.453291536050157, "grad_norm": 3.4934675693511963, "learning_rate": 4.556439837135651e-07, "loss": 1.2045, "step": 44064 }, { "epoch": 3.4533699059561127, "grad_norm": 3.689401865005493, "learning_rate": 4.555153816660507e-07, "loss": 1.2683, "step": 44065 }, { "epoch": 3.453448275862069, "grad_norm": 3.1976876258850098, "learning_rate": 4.5538679690342425e-07, "loss": 1.1892, "step": 44066 }, { "epoch": 3.453526645768025, "grad_norm": 6.4123334884643555, "learning_rate": 4.552582294261704e-07, "loss": 1.5684, "step": 44067 }, { "epoch": 3.4536050156739813, "grad_norm": 14.476238250732422, "learning_rate": 4.5512967923478035e-07, "loss": 0.6007, "step": 44068 }, { "epoch": 3.453683385579937, "grad_norm": 4.7437520027160645, "learning_rate": 4.550011463297427e-07, "loss": 1.1473, "step": 44069 }, { "epoch": 3.4537617554858935, "grad_norm": 4.30369234085083, "learning_rate": 4.5487263071154643e-07, "loss": 1.4018, "step": 44070 }, { "epoch": 3.4538401253918494, "grad_norm": 3.0331978797912598, "learning_rate": 4.547441323806795e-07, "loss": 1.3923, "step": 44071 }, { "epoch": 3.4539184952978057, "grad_norm": 3.086986541748047, "learning_rate": 4.546156513376321e-07, "loss": 1.136, "step": 44072 }, { "epoch": 3.4539968652037616, "grad_norm": 4.7343244552612305, "learning_rate": 4.5448718758289155e-07, "loss": 1.7973, "step": 44073 }, { "epoch": 3.454075235109718, "grad_norm": 2.7443840503692627, "learning_rate": 4.54358741116947e-07, "loss": 1.1211, "step": 44074 }, { "epoch": 3.454153605015674, "grad_norm": 5.203269958496094, "learning_rate": 4.542303119402869e-07, "loss": 1.3029, "step": 44075 }, { "epoch": 3.45423197492163, "grad_norm": 2.479703903198242, "learning_rate": 4.5410190005340027e-07, "loss": 1.11, "step": 44076 }, { "epoch": 3.454310344827586, "grad_norm": 7.858926773071289, "learning_rate": 4.539735054567751e-07, "loss": 1.7412, "step": 44077 }, { "epoch": 3.4543887147335424, "grad_norm": 3.3811957836151123, "learning_rate": 4.538451281509004e-07, "loss": 1.327, "step": 44078 }, { "epoch": 3.4544670846394983, "grad_norm": 3.8337090015411377, "learning_rate": 4.5371676813626197e-07, "loss": 1.1683, "step": 44079 }, { "epoch": 3.4545454545454546, "grad_norm": 2.810492992401123, "learning_rate": 4.5358842541335047e-07, "loss": 1.2276, "step": 44080 }, { "epoch": 3.4546238244514105, "grad_norm": 3.347107410430908, "learning_rate": 4.5346009998265497e-07, "loss": 1.4206, "step": 44081 }, { "epoch": 3.454702194357367, "grad_norm": 11.738880157470703, "learning_rate": 4.533317918446606e-07, "loss": 1.1634, "step": 44082 }, { "epoch": 3.4547805642633227, "grad_norm": 5.080091953277588, "learning_rate": 4.53203500999857e-07, "loss": 1.5607, "step": 44083 }, { "epoch": 3.454858934169279, "grad_norm": 5.006289958953857, "learning_rate": 4.53075227448731e-07, "loss": 1.3936, "step": 44084 }, { "epoch": 3.454937304075235, "grad_norm": 3.133269786834717, "learning_rate": 4.5294697119177335e-07, "loss": 0.9614, "step": 44085 }, { "epoch": 3.4550156739811912, "grad_norm": 4.545159339904785, "learning_rate": 4.528187322294686e-07, "loss": 1.5093, "step": 44086 }, { "epoch": 3.455094043887147, "grad_norm": 4.844374656677246, "learning_rate": 4.5269051056230585e-07, "loss": 1.2905, "step": 44087 }, { "epoch": 3.4551724137931035, "grad_norm": 4.420055866241455, "learning_rate": 4.52562306190773e-07, "loss": 1.4259, "step": 44088 }, { "epoch": 3.45525078369906, "grad_norm": 2.8722121715545654, "learning_rate": 4.524341191153575e-07, "loss": 1.2163, "step": 44089 }, { "epoch": 3.4553291536050157, "grad_norm": 2.707735300064087, "learning_rate": 4.523059493365461e-07, "loss": 1.3928, "step": 44090 }, { "epoch": 3.4554075235109716, "grad_norm": 2.2404661178588867, "learning_rate": 4.52177796854828e-07, "loss": 1.1565, "step": 44091 }, { "epoch": 3.455485893416928, "grad_norm": 3.42234206199646, "learning_rate": 4.5204966167068933e-07, "loss": 1.3143, "step": 44092 }, { "epoch": 3.4555642633228842, "grad_norm": 7.545477867126465, "learning_rate": 4.519215437846175e-07, "loss": 1.3787, "step": 44093 }, { "epoch": 3.45564263322884, "grad_norm": 8.888681411743164, "learning_rate": 4.5179344319709995e-07, "loss": 1.7451, "step": 44094 }, { "epoch": 3.455721003134796, "grad_norm": 2.088681697845459, "learning_rate": 4.516653599086246e-07, "loss": 1.4102, "step": 44095 }, { "epoch": 3.4557993730407524, "grad_norm": 5.341796875, "learning_rate": 4.5153729391967884e-07, "loss": 1.2803, "step": 44096 }, { "epoch": 3.4558777429467087, "grad_norm": 8.96133804321289, "learning_rate": 4.514092452307478e-07, "loss": 1.7822, "step": 44097 }, { "epoch": 3.4559561128526646, "grad_norm": 3.7447915077209473, "learning_rate": 4.5128121384231895e-07, "loss": 1.4573, "step": 44098 }, { "epoch": 3.4560344827586205, "grad_norm": 5.2509846687316895, "learning_rate": 4.511531997548807e-07, "loss": 1.328, "step": 44099 }, { "epoch": 3.456112852664577, "grad_norm": 3.042945623397827, "learning_rate": 4.510252029689205e-07, "loss": 1.4078, "step": 44100 }, { "epoch": 3.456191222570533, "grad_norm": 4.556769847869873, "learning_rate": 4.5089722348492237e-07, "loss": 0.8699, "step": 44101 }, { "epoch": 3.456269592476489, "grad_norm": 3.7526824474334717, "learning_rate": 4.507692613033754e-07, "loss": 1.1036, "step": 44102 }, { "epoch": 3.456347962382445, "grad_norm": 3.15055775642395, "learning_rate": 4.506413164247642e-07, "loss": 1.2659, "step": 44103 }, { "epoch": 3.4564263322884012, "grad_norm": 3.528465509414673, "learning_rate": 4.5051338884957885e-07, "loss": 1.4605, "step": 44104 }, { "epoch": 3.4565047021943576, "grad_norm": 4.797366619110107, "learning_rate": 4.503854785783024e-07, "loss": 1.8026, "step": 44105 }, { "epoch": 3.4565830721003135, "grad_norm": 3.6681671142578125, "learning_rate": 4.502575856114233e-07, "loss": 0.9935, "step": 44106 }, { "epoch": 3.4566614420062693, "grad_norm": 3.4118223190307617, "learning_rate": 4.501297099494273e-07, "loss": 1.1831, "step": 44107 }, { "epoch": 3.4567398119122257, "grad_norm": 3.1292872428894043, "learning_rate": 4.5000185159280064e-07, "loss": 1.4717, "step": 44108 }, { "epoch": 3.456818181818182, "grad_norm": 6.920293807983398, "learning_rate": 4.498740105420302e-07, "loss": 1.2587, "step": 44109 }, { "epoch": 3.456896551724138, "grad_norm": 3.6563496589660645, "learning_rate": 4.4974618679760164e-07, "loss": 1.278, "step": 44110 }, { "epoch": 3.4569749216300942, "grad_norm": 3.0069167613983154, "learning_rate": 4.4961838036000184e-07, "loss": 1.3133, "step": 44111 }, { "epoch": 3.45705329153605, "grad_norm": 2.323215961456299, "learning_rate": 4.4949059122971593e-07, "loss": 1.5071, "step": 44112 }, { "epoch": 3.4571316614420065, "grad_norm": 2.3723697662353516, "learning_rate": 4.493628194072308e-07, "loss": 1.3642, "step": 44113 }, { "epoch": 3.4572100313479623, "grad_norm": 6.187418460845947, "learning_rate": 4.492350648930327e-07, "loss": 1.3991, "step": 44114 }, { "epoch": 3.4572884012539187, "grad_norm": 5.462642669677734, "learning_rate": 4.4910732768760734e-07, "loss": 1.371, "step": 44115 }, { "epoch": 3.4573667711598746, "grad_norm": 3.605325222015381, "learning_rate": 4.4897960779143825e-07, "loss": 1.2752, "step": 44116 }, { "epoch": 3.457445141065831, "grad_norm": 4.004642486572266, "learning_rate": 4.4885190520501445e-07, "loss": 1.4145, "step": 44117 }, { "epoch": 3.457523510971787, "grad_norm": 4.403250694274902, "learning_rate": 4.4872421992882e-07, "loss": 1.2156, "step": 44118 }, { "epoch": 3.457601880877743, "grad_norm": 6.554774284362793, "learning_rate": 4.485965519633423e-07, "loss": 1.2537, "step": 44119 }, { "epoch": 3.457680250783699, "grad_norm": 2.2448368072509766, "learning_rate": 4.484689013090643e-07, "loss": 1.286, "step": 44120 }, { "epoch": 3.4577586206896553, "grad_norm": 3.9643874168395996, "learning_rate": 4.4834126796647226e-07, "loss": 1.1053, "step": 44121 }, { "epoch": 3.4578369905956112, "grad_norm": 2.4442601203918457, "learning_rate": 4.482136519360536e-07, "loss": 1.4072, "step": 44122 }, { "epoch": 3.4579153605015676, "grad_norm": 3.381770610809326, "learning_rate": 4.480860532182918e-07, "loss": 1.3219, "step": 44123 }, { "epoch": 3.4579937304075234, "grad_norm": 2.8670153617858887, "learning_rate": 4.4795847181367203e-07, "loss": 1.1966, "step": 44124 }, { "epoch": 3.45807210031348, "grad_norm": 2.3377294540405273, "learning_rate": 4.478309077226806e-07, "loss": 1.1813, "step": 44125 }, { "epoch": 3.4581504702194357, "grad_norm": 5.1327738761901855, "learning_rate": 4.4770336094580205e-07, "loss": 1.4992, "step": 44126 }, { "epoch": 3.458228840125392, "grad_norm": 2.136262893676758, "learning_rate": 4.4757583148352156e-07, "loss": 1.4705, "step": 44127 }, { "epoch": 3.458307210031348, "grad_norm": 3.156874418258667, "learning_rate": 4.4744831933632496e-07, "loss": 1.1469, "step": 44128 }, { "epoch": 3.458385579937304, "grad_norm": 4.059018611907959, "learning_rate": 4.4732082450469615e-07, "loss": 1.4269, "step": 44129 }, { "epoch": 3.45846394984326, "grad_norm": 3.9427685737609863, "learning_rate": 4.4719334698912043e-07, "loss": 1.3603, "step": 44130 }, { "epoch": 3.4585423197492164, "grad_norm": 3.7141666412353516, "learning_rate": 4.470658867900829e-07, "loss": 1.1513, "step": 44131 }, { "epoch": 3.4586206896551723, "grad_norm": 3.228069305419922, "learning_rate": 4.4693844390806815e-07, "loss": 1.4823, "step": 44132 }, { "epoch": 3.4586990595611287, "grad_norm": 3.4208552837371826, "learning_rate": 4.468110183435609e-07, "loss": 1.1873, "step": 44133 }, { "epoch": 3.4587774294670846, "grad_norm": 3.0627636909484863, "learning_rate": 4.466836100970473e-07, "loss": 1.3693, "step": 44134 }, { "epoch": 3.458855799373041, "grad_norm": 2.850937604904175, "learning_rate": 4.4655621916900816e-07, "loss": 1.1331, "step": 44135 }, { "epoch": 3.4589341692789968, "grad_norm": 3.1978821754455566, "learning_rate": 4.464288455599314e-07, "loss": 0.8752, "step": 44136 }, { "epoch": 3.459012539184953, "grad_norm": 2.8193788528442383, "learning_rate": 4.4630148927030105e-07, "loss": 1.2632, "step": 44137 }, { "epoch": 3.459090909090909, "grad_norm": 4.745883941650391, "learning_rate": 4.461741503006006e-07, "loss": 1.2684, "step": 44138 }, { "epoch": 3.4591692789968653, "grad_norm": 14.655515670776367, "learning_rate": 4.460468286513142e-07, "loss": 1.2041, "step": 44139 }, { "epoch": 3.459247648902821, "grad_norm": 2.6354267597198486, "learning_rate": 4.4591952432292584e-07, "loss": 1.1499, "step": 44140 }, { "epoch": 3.4593260188087775, "grad_norm": 7.952148914337158, "learning_rate": 4.4579223731592236e-07, "loss": 1.4517, "step": 44141 }, { "epoch": 3.4594043887147334, "grad_norm": 5.377695560455322, "learning_rate": 4.4566496763078507e-07, "loss": 1.0547, "step": 44142 }, { "epoch": 3.4594827586206898, "grad_norm": 3.1105763912200928, "learning_rate": 4.4553771526799903e-07, "loss": 1.6178, "step": 44143 }, { "epoch": 3.4595611285266457, "grad_norm": 5.128968238830566, "learning_rate": 4.4541048022804844e-07, "loss": 1.6882, "step": 44144 }, { "epoch": 3.459639498432602, "grad_norm": 3.470374584197998, "learning_rate": 4.4528326251141674e-07, "loss": 1.4681, "step": 44145 }, { "epoch": 3.459717868338558, "grad_norm": 6.861307621002197, "learning_rate": 4.4515606211858795e-07, "loss": 1.7004, "step": 44146 }, { "epoch": 3.459796238244514, "grad_norm": 3.116849899291992, "learning_rate": 4.450288790500462e-07, "loss": 1.3205, "step": 44147 }, { "epoch": 3.45987460815047, "grad_norm": 3.6245298385620117, "learning_rate": 4.44901713306275e-07, "loss": 1.5139, "step": 44148 }, { "epoch": 3.4599529780564264, "grad_norm": 3.3549885749816895, "learning_rate": 4.447745648877577e-07, "loss": 1.3861, "step": 44149 }, { "epoch": 3.4600313479623823, "grad_norm": 3.5325393676757812, "learning_rate": 4.4464743379497854e-07, "loss": 1.2977, "step": 44150 }, { "epoch": 3.4601097178683387, "grad_norm": 7.604761600494385, "learning_rate": 4.4452032002842093e-07, "loss": 1.5107, "step": 44151 }, { "epoch": 3.4601880877742945, "grad_norm": 6.06752872467041, "learning_rate": 4.4439322358856896e-07, "loss": 1.235, "step": 44152 }, { "epoch": 3.460266457680251, "grad_norm": 3.5122413635253906, "learning_rate": 4.4426614447590444e-07, "loss": 1.5508, "step": 44153 }, { "epoch": 3.4603448275862068, "grad_norm": 3.7760379314422607, "learning_rate": 4.4413908269091033e-07, "loss": 1.0415, "step": 44154 }, { "epoch": 3.460423197492163, "grad_norm": 2.122734546661377, "learning_rate": 4.4401203823407236e-07, "loss": 1.1326, "step": 44155 }, { "epoch": 3.460501567398119, "grad_norm": 3.837437868118286, "learning_rate": 4.438850111058729e-07, "loss": 1.1758, "step": 44156 }, { "epoch": 3.4605799373040753, "grad_norm": 10.041769027709961, "learning_rate": 4.437580013067944e-07, "loss": 0.9683, "step": 44157 }, { "epoch": 3.460658307210031, "grad_norm": 3.2442970275878906, "learning_rate": 4.4363100883731966e-07, "loss": 0.9487, "step": 44158 }, { "epoch": 3.4607366771159875, "grad_norm": 2.9460813999176025, "learning_rate": 4.4350403369793184e-07, "loss": 1.367, "step": 44159 }, { "epoch": 3.4608150470219434, "grad_norm": 3.159594774246216, "learning_rate": 4.4337707588911536e-07, "loss": 1.1333, "step": 44160 }, { "epoch": 3.4608934169278998, "grad_norm": 2.3823180198669434, "learning_rate": 4.432501354113522e-07, "loss": 1.1777, "step": 44161 }, { "epoch": 3.4609717868338556, "grad_norm": 2.846597909927368, "learning_rate": 4.4312321226512414e-07, "loss": 1.2895, "step": 44162 }, { "epoch": 3.461050156739812, "grad_norm": 3.1796345710754395, "learning_rate": 4.4299630645091516e-07, "loss": 1.3358, "step": 44163 }, { "epoch": 3.461128526645768, "grad_norm": 4.135578155517578, "learning_rate": 4.428694179692078e-07, "loss": 1.7093, "step": 44164 }, { "epoch": 3.461206896551724, "grad_norm": 2.8712761402130127, "learning_rate": 4.427425468204843e-07, "loss": 1.317, "step": 44165 }, { "epoch": 3.46128526645768, "grad_norm": 3.011760950088501, "learning_rate": 4.426156930052272e-07, "loss": 1.5112, "step": 44166 }, { "epoch": 3.4613636363636364, "grad_norm": 3.554414749145508, "learning_rate": 4.424888565239205e-07, "loss": 1.5888, "step": 44167 }, { "epoch": 3.4614420062695923, "grad_norm": 3.3232882022857666, "learning_rate": 4.423620373770432e-07, "loss": 0.8935, "step": 44168 }, { "epoch": 3.4615203761755486, "grad_norm": 7.3962931632995605, "learning_rate": 4.422352355650811e-07, "loss": 1.3143, "step": 44169 }, { "epoch": 3.4615987460815045, "grad_norm": 3.453388214111328, "learning_rate": 4.421084510885143e-07, "loss": 1.5801, "step": 44170 }, { "epoch": 3.461677115987461, "grad_norm": 4.405956268310547, "learning_rate": 4.419816839478275e-07, "loss": 1.1486, "step": 44171 }, { "epoch": 3.4617554858934168, "grad_norm": 2.667682647705078, "learning_rate": 4.418549341435002e-07, "loss": 1.1537, "step": 44172 }, { "epoch": 3.461833855799373, "grad_norm": 8.31603717803955, "learning_rate": 4.4172820167601436e-07, "loss": 1.1925, "step": 44173 }, { "epoch": 3.461912225705329, "grad_norm": 3.808070182800293, "learning_rate": 4.4160148654585453e-07, "loss": 1.4948, "step": 44174 }, { "epoch": 3.4619905956112853, "grad_norm": 4.958265781402588, "learning_rate": 4.4147478875350145e-07, "loss": 1.1481, "step": 44175 }, { "epoch": 3.462068965517241, "grad_norm": 3.7335004806518555, "learning_rate": 4.41348108299437e-07, "loss": 1.1155, "step": 44176 }, { "epoch": 3.4621473354231975, "grad_norm": 3.80055570602417, "learning_rate": 4.4122144518414176e-07, "loss": 1.8107, "step": 44177 }, { "epoch": 3.4622257053291534, "grad_norm": 3.3926992416381836, "learning_rate": 4.410947994080994e-07, "loss": 1.2212, "step": 44178 }, { "epoch": 3.4623040752351097, "grad_norm": 3.6721231937408447, "learning_rate": 4.4096817097179057e-07, "loss": 1.3281, "step": 44179 }, { "epoch": 3.462382445141066, "grad_norm": 6.08268404006958, "learning_rate": 4.408415598756971e-07, "loss": 0.8726, "step": 44180 }, { "epoch": 3.462460815047022, "grad_norm": 8.331145286560059, "learning_rate": 4.4071496612030086e-07, "loss": 1.3258, "step": 44181 }, { "epoch": 3.462539184952978, "grad_norm": 2.956432580947876, "learning_rate": 4.405883897060825e-07, "loss": 1.4314, "step": 44182 }, { "epoch": 3.462617554858934, "grad_norm": 3.622343063354492, "learning_rate": 4.4046183063352456e-07, "loss": 1.0583, "step": 44183 }, { "epoch": 3.4626959247648905, "grad_norm": 3.6026880741119385, "learning_rate": 4.4033528890310763e-07, "loss": 1.0574, "step": 44184 }, { "epoch": 3.4627742946708464, "grad_norm": 4.427609443664551, "learning_rate": 4.402087645153136e-07, "loss": 1.5294, "step": 44185 }, { "epoch": 3.4628526645768023, "grad_norm": 3.685288190841675, "learning_rate": 4.4008225747062425e-07, "loss": 1.2299, "step": 44186 }, { "epoch": 3.4629310344827586, "grad_norm": 4.23026704788208, "learning_rate": 4.3995576776951765e-07, "loss": 1.4551, "step": 44187 }, { "epoch": 3.463009404388715, "grad_norm": 3.241405487060547, "learning_rate": 4.3982929541247834e-07, "loss": 1.6937, "step": 44188 }, { "epoch": 3.463087774294671, "grad_norm": 2.773977756500244, "learning_rate": 4.397028403999859e-07, "loss": 1.4666, "step": 44189 }, { "epoch": 3.4631661442006267, "grad_norm": 3.095548391342163, "learning_rate": 4.3957640273252277e-07, "loss": 1.8745, "step": 44190 }, { "epoch": 3.463244514106583, "grad_norm": 7.570763111114502, "learning_rate": 4.394499824105675e-07, "loss": 1.4513, "step": 44191 }, { "epoch": 3.4633228840125394, "grad_norm": 4.599562644958496, "learning_rate": 4.3932357943460123e-07, "loss": 0.8984, "step": 44192 }, { "epoch": 3.4634012539184953, "grad_norm": 3.038100481033325, "learning_rate": 4.3919719380510704e-07, "loss": 1.4208, "step": 44193 }, { "epoch": 3.463479623824451, "grad_norm": 6.380554676055908, "learning_rate": 4.3907082552256396e-07, "loss": 0.8331, "step": 44194 }, { "epoch": 3.4635579937304075, "grad_norm": 4.430575370788574, "learning_rate": 4.3894447458745215e-07, "loss": 1.4195, "step": 44195 }, { "epoch": 3.463636363636364, "grad_norm": 3.523488759994507, "learning_rate": 4.3881814100025346e-07, "loss": 1.4799, "step": 44196 }, { "epoch": 3.4637147335423197, "grad_norm": 2.9848239421844482, "learning_rate": 4.386918247614469e-07, "loss": 1.4916, "step": 44197 }, { "epoch": 3.4637931034482756, "grad_norm": 4.227071762084961, "learning_rate": 4.385655258715144e-07, "loss": 1.6097, "step": 44198 }, { "epoch": 3.463871473354232, "grad_norm": 3.20369815826416, "learning_rate": 4.38439244330936e-07, "loss": 1.7015, "step": 44199 }, { "epoch": 3.4639498432601883, "grad_norm": 3.840233325958252, "learning_rate": 4.3831298014019144e-07, "loss": 1.3794, "step": 44200 }, { "epoch": 3.464028213166144, "grad_norm": 4.2568464279174805, "learning_rate": 4.381867332997608e-07, "loss": 1.5235, "step": 44201 }, { "epoch": 3.4641065830721, "grad_norm": 13.867241859436035, "learning_rate": 4.3806050381012546e-07, "loss": 1.3304, "step": 44202 }, { "epoch": 3.4641849529780564, "grad_norm": 3.6866519451141357, "learning_rate": 4.3793429167176437e-07, "loss": 1.3925, "step": 44203 }, { "epoch": 3.4642633228840127, "grad_norm": 3.5083792209625244, "learning_rate": 4.378080968851583e-07, "loss": 1.4748, "step": 44204 }, { "epoch": 3.4643416927899686, "grad_norm": 4.2381134033203125, "learning_rate": 4.376819194507875e-07, "loss": 1.418, "step": 44205 }, { "epoch": 3.464420062695925, "grad_norm": 3.2828359603881836, "learning_rate": 4.3755575936912984e-07, "loss": 1.3636, "step": 44206 }, { "epoch": 3.464498432601881, "grad_norm": 5.719148635864258, "learning_rate": 4.374296166406677e-07, "loss": 1.497, "step": 44207 }, { "epoch": 3.464576802507837, "grad_norm": 2.928571939468384, "learning_rate": 4.3730349126588024e-07, "loss": 1.7739, "step": 44208 }, { "epoch": 3.464655172413793, "grad_norm": 3.4543368816375732, "learning_rate": 4.3717738324524583e-07, "loss": 1.0123, "step": 44209 }, { "epoch": 3.4647335423197494, "grad_norm": 7.655034065246582, "learning_rate": 4.3705129257924527e-07, "loss": 1.4144, "step": 44210 }, { "epoch": 3.4648119122257053, "grad_norm": 8.567672729492188, "learning_rate": 4.369252192683571e-07, "loss": 1.346, "step": 44211 }, { "epoch": 3.4648902821316616, "grad_norm": 3.4017937183380127, "learning_rate": 4.367991633130636e-07, "loss": 0.7674, "step": 44212 }, { "epoch": 3.4649686520376175, "grad_norm": 5.0101423263549805, "learning_rate": 4.3667312471384115e-07, "loss": 1.8372, "step": 44213 }, { "epoch": 3.465047021943574, "grad_norm": 4.423115253448486, "learning_rate": 4.365471034711699e-07, "loss": 1.8702, "step": 44214 }, { "epoch": 3.4651253918495297, "grad_norm": 3.288459300994873, "learning_rate": 4.3642109958553e-07, "loss": 1.0903, "step": 44215 }, { "epoch": 3.465203761755486, "grad_norm": 4.198138236999512, "learning_rate": 4.362951130574e-07, "loss": 1.1904, "step": 44216 }, { "epoch": 3.465282131661442, "grad_norm": 2.647761821746826, "learning_rate": 4.3616914388725953e-07, "loss": 1.409, "step": 44217 }, { "epoch": 3.4653605015673983, "grad_norm": 2.298473834991455, "learning_rate": 4.3604319207558754e-07, "loss": 1.4397, "step": 44218 }, { "epoch": 3.465438871473354, "grad_norm": 2.3961853981018066, "learning_rate": 4.359172576228632e-07, "loss": 1.3909, "step": 44219 }, { "epoch": 3.4655172413793105, "grad_norm": 2.8582425117492676, "learning_rate": 4.3579134052956495e-07, "loss": 1.177, "step": 44220 }, { "epoch": 3.4655956112852664, "grad_norm": 2.7399141788482666, "learning_rate": 4.356654407961719e-07, "loss": 1.1961, "step": 44221 }, { "epoch": 3.4656739811912227, "grad_norm": 6.1172709465026855, "learning_rate": 4.3553955842316366e-07, "loss": 1.1994, "step": 44222 }, { "epoch": 3.4657523510971786, "grad_norm": 5.679360866546631, "learning_rate": 4.3541369341101925e-07, "loss": 1.2155, "step": 44223 }, { "epoch": 3.465830721003135, "grad_norm": 3.016627073287964, "learning_rate": 4.35287845760215e-07, "loss": 1.6613, "step": 44224 }, { "epoch": 3.465909090909091, "grad_norm": 2.8068530559539795, "learning_rate": 4.351620154712316e-07, "loss": 1.6269, "step": 44225 }, { "epoch": 3.465987460815047, "grad_norm": 3.5918970108032227, "learning_rate": 4.350362025445476e-07, "loss": 1.4573, "step": 44226 }, { "epoch": 3.466065830721003, "grad_norm": 4.019998073577881, "learning_rate": 4.3491040698064203e-07, "loss": 1.048, "step": 44227 }, { "epoch": 3.4661442006269594, "grad_norm": 9.110112190246582, "learning_rate": 4.347846287799917e-07, "loss": 0.5223, "step": 44228 }, { "epoch": 3.4662225705329153, "grad_norm": 3.4176669120788574, "learning_rate": 4.3465886794307465e-07, "loss": 1.422, "step": 44229 }, { "epoch": 3.4663009404388716, "grad_norm": 3.95672345161438, "learning_rate": 4.34533124470371e-07, "loss": 1.465, "step": 44230 }, { "epoch": 3.4663793103448275, "grad_norm": 3.295182228088379, "learning_rate": 4.3440739836235976e-07, "loss": 1.4945, "step": 44231 }, { "epoch": 3.466457680250784, "grad_norm": 5.051031112670898, "learning_rate": 4.3428168961951676e-07, "loss": 1.5476, "step": 44232 }, { "epoch": 3.4665360501567397, "grad_norm": 2.606975793838501, "learning_rate": 4.3415599824232104e-07, "loss": 0.8426, "step": 44233 }, { "epoch": 3.466614420062696, "grad_norm": 4.610721588134766, "learning_rate": 4.3403032423125105e-07, "loss": 1.0263, "step": 44234 }, { "epoch": 3.466692789968652, "grad_norm": 3.2806248664855957, "learning_rate": 4.3390466758678417e-07, "loss": 1.6999, "step": 44235 }, { "epoch": 3.4667711598746083, "grad_norm": 5.924670696258545, "learning_rate": 4.3377902830939846e-07, "loss": 1.1338, "step": 44236 }, { "epoch": 3.466849529780564, "grad_norm": 3.2259695529937744, "learning_rate": 4.336534063995723e-07, "loss": 1.0871, "step": 44237 }, { "epoch": 3.4669278996865205, "grad_norm": 2.4898295402526855, "learning_rate": 4.3352780185778265e-07, "loss": 1.1647, "step": 44238 }, { "epoch": 3.4670062695924764, "grad_norm": 8.095120429992676, "learning_rate": 4.334022146845085e-07, "loss": 0.651, "step": 44239 }, { "epoch": 3.4670846394984327, "grad_norm": 4.621775150299072, "learning_rate": 4.3327664488022613e-07, "loss": 0.7131, "step": 44240 }, { "epoch": 3.4671630094043886, "grad_norm": 5.621998310089111, "learning_rate": 4.331510924454141e-07, "loss": 0.9557, "step": 44241 }, { "epoch": 3.467241379310345, "grad_norm": 2.488607883453369, "learning_rate": 4.3302555738055087e-07, "loss": 1.0103, "step": 44242 }, { "epoch": 3.467319749216301, "grad_norm": 3.11407470703125, "learning_rate": 4.329000396861099e-07, "loss": 1.3998, "step": 44243 }, { "epoch": 3.467398119122257, "grad_norm": 8.026659965515137, "learning_rate": 4.3277453936257317e-07, "loss": 1.5717, "step": 44244 }, { "epoch": 3.467476489028213, "grad_norm": 4.033123016357422, "learning_rate": 4.3264905641041575e-07, "loss": 0.7704, "step": 44245 }, { "epoch": 3.4675548589341694, "grad_norm": 3.059630870819092, "learning_rate": 4.3252359083011676e-07, "loss": 1.0127, "step": 44246 }, { "epoch": 3.4676332288401253, "grad_norm": 3.513028860092163, "learning_rate": 4.3239814262215075e-07, "loss": 1.3099, "step": 44247 }, { "epoch": 3.4677115987460816, "grad_norm": 2.829270601272583, "learning_rate": 4.322727117869951e-07, "loss": 1.248, "step": 44248 }, { "epoch": 3.4677899686520375, "grad_norm": 8.52397632598877, "learning_rate": 4.3214729832513016e-07, "loss": 1.1351, "step": 44249 }, { "epoch": 3.467868338557994, "grad_norm": 2.65293025970459, "learning_rate": 4.3202190223702924e-07, "loss": 1.2407, "step": 44250 }, { "epoch": 3.4679467084639497, "grad_norm": 3.54842209815979, "learning_rate": 4.3189652352317147e-07, "loss": 1.6276, "step": 44251 }, { "epoch": 3.468025078369906, "grad_norm": 2.4986021518707275, "learning_rate": 4.3177116218403203e-07, "loss": 1.2529, "step": 44252 }, { "epoch": 3.468103448275862, "grad_norm": 4.725840091705322, "learning_rate": 4.3164581822008944e-07, "loss": 1.4801, "step": 44253 }, { "epoch": 3.4681818181818183, "grad_norm": 3.753753900527954, "learning_rate": 4.315204916318194e-07, "loss": 0.6982, "step": 44254 }, { "epoch": 3.468260188087774, "grad_norm": 3.3521034717559814, "learning_rate": 4.313951824196994e-07, "loss": 1.6373, "step": 44255 }, { "epoch": 3.4683385579937305, "grad_norm": 4.60062313079834, "learning_rate": 4.3126989058420497e-07, "loss": 1.377, "step": 44256 }, { "epoch": 3.4684169278996864, "grad_norm": 2.839582920074463, "learning_rate": 4.3114461612581314e-07, "loss": 1.1529, "step": 44257 }, { "epoch": 3.4684952978056427, "grad_norm": 5.136048316955566, "learning_rate": 4.310193590450007e-07, "loss": 1.7837, "step": 44258 }, { "epoch": 3.4685736677115986, "grad_norm": 9.146296501159668, "learning_rate": 4.308941193422439e-07, "loss": 1.6019, "step": 44259 }, { "epoch": 3.468652037617555, "grad_norm": 2.510199546813965, "learning_rate": 4.3076889701801905e-07, "loss": 1.098, "step": 44260 }, { "epoch": 3.468730407523511, "grad_norm": 4.380392551422119, "learning_rate": 4.3064369207280354e-07, "loss": 1.4455, "step": 44261 }, { "epoch": 3.468808777429467, "grad_norm": 5.276100158691406, "learning_rate": 4.3051850450707034e-07, "loss": 1.5841, "step": 44262 }, { "epoch": 3.468887147335423, "grad_norm": 8.4456205368042, "learning_rate": 4.303933343212985e-07, "loss": 1.9447, "step": 44263 }, { "epoch": 3.4689655172413794, "grad_norm": 7.7184576988220215, "learning_rate": 4.3026818151596426e-07, "loss": 0.9134, "step": 44264 }, { "epoch": 3.4690438871473352, "grad_norm": 4.107913494110107, "learning_rate": 4.3014304609154234e-07, "loss": 1.4194, "step": 44265 }, { "epoch": 3.4691222570532916, "grad_norm": 4.6136474609375, "learning_rate": 4.300179280485084e-07, "loss": 1.4129, "step": 44266 }, { "epoch": 3.4692006269592475, "grad_norm": 3.67995548248291, "learning_rate": 4.298928273873382e-07, "loss": 1.6384, "step": 44267 }, { "epoch": 3.469278996865204, "grad_norm": 5.551589488983154, "learning_rate": 4.297677441085102e-07, "loss": 1.6401, "step": 44268 }, { "epoch": 3.4693573667711597, "grad_norm": 4.4744343757629395, "learning_rate": 4.296426782124974e-07, "loss": 0.9439, "step": 44269 }, { "epoch": 3.469435736677116, "grad_norm": 2.4824838638305664, "learning_rate": 4.295176296997766e-07, "loss": 1.6826, "step": 44270 }, { "epoch": 3.469514106583072, "grad_norm": 3.304859161376953, "learning_rate": 4.2939259857082303e-07, "loss": 1.2722, "step": 44271 }, { "epoch": 3.4695924764890282, "grad_norm": 9.357982635498047, "learning_rate": 4.292675848261124e-07, "loss": 1.3946, "step": 44272 }, { "epoch": 3.469670846394984, "grad_norm": 2.2686939239501953, "learning_rate": 4.291425884661199e-07, "loss": 1.354, "step": 44273 }, { "epoch": 3.4697492163009405, "grad_norm": 3.5583019256591797, "learning_rate": 4.2901760949132175e-07, "loss": 1.1517, "step": 44274 }, { "epoch": 3.469827586206897, "grad_norm": 2.9883923530578613, "learning_rate": 4.288926479021932e-07, "loss": 1.5955, "step": 44275 }, { "epoch": 3.4699059561128527, "grad_norm": 2.4714338779449463, "learning_rate": 4.287677036992077e-07, "loss": 0.9381, "step": 44276 }, { "epoch": 3.4699843260188086, "grad_norm": 7.8706560134887695, "learning_rate": 4.286427768828427e-07, "loss": 0.9293, "step": 44277 }, { "epoch": 3.470062695924765, "grad_norm": 2.985872268676758, "learning_rate": 4.285178674535723e-07, "loss": 1.3589, "step": 44278 }, { "epoch": 3.4701410658307212, "grad_norm": 3.6315085887908936, "learning_rate": 4.283929754118732e-07, "loss": 1.4141, "step": 44279 }, { "epoch": 3.470219435736677, "grad_norm": 8.460233688354492, "learning_rate": 4.282681007582179e-07, "loss": 1.4606, "step": 44280 }, { "epoch": 3.470297805642633, "grad_norm": 3.7346935272216797, "learning_rate": 4.281432434930821e-07, "loss": 1.7557, "step": 44281 }, { "epoch": 3.4703761755485893, "grad_norm": 2.3661341667175293, "learning_rate": 4.280184036169416e-07, "loss": 1.2207, "step": 44282 }, { "epoch": 3.4704545454545457, "grad_norm": 6.196263313293457, "learning_rate": 4.27893581130272e-07, "loss": 1.085, "step": 44283 }, { "epoch": 3.4705329153605016, "grad_norm": 3.500020742416382, "learning_rate": 4.277687760335458e-07, "loss": 1.4837, "step": 44284 }, { "epoch": 3.4706112852664575, "grad_norm": 3.488203525543213, "learning_rate": 4.2764398832723874e-07, "loss": 1.1425, "step": 44285 }, { "epoch": 3.470689655172414, "grad_norm": 2.987196683883667, "learning_rate": 4.275192180118243e-07, "loss": 1.2154, "step": 44286 }, { "epoch": 3.47076802507837, "grad_norm": 3.235358715057373, "learning_rate": 4.2739446508778037e-07, "loss": 1.2482, "step": 44287 }, { "epoch": 3.470846394984326, "grad_norm": 1.936299443244934, "learning_rate": 4.2726972955557776e-07, "loss": 1.2014, "step": 44288 }, { "epoch": 3.470924764890282, "grad_norm": 3.0729146003723145, "learning_rate": 4.271450114156933e-07, "loss": 1.2637, "step": 44289 }, { "epoch": 3.4710031347962382, "grad_norm": 3.519019365310669, "learning_rate": 4.2702031066859993e-07, "loss": 1.4133, "step": 44290 }, { "epoch": 3.4710815047021946, "grad_norm": 4.757055759429932, "learning_rate": 4.268956273147723e-07, "loss": 1.8579, "step": 44291 }, { "epoch": 3.4711598746081505, "grad_norm": 2.3446619510650635, "learning_rate": 4.26770961354685e-07, "loss": 0.8924, "step": 44292 }, { "epoch": 3.4712382445141063, "grad_norm": 3.701017379760742, "learning_rate": 4.2664631278881207e-07, "loss": 1.2839, "step": 44293 }, { "epoch": 3.4713166144200627, "grad_norm": 3.3912224769592285, "learning_rate": 4.265216816176282e-07, "loss": 1.2483, "step": 44294 }, { "epoch": 3.471394984326019, "grad_norm": 4.192635536193848, "learning_rate": 4.2639706784160517e-07, "loss": 1.313, "step": 44295 }, { "epoch": 3.471473354231975, "grad_norm": 5.868838310241699, "learning_rate": 4.262724714612193e-07, "loss": 1.2056, "step": 44296 }, { "epoch": 3.4715517241379312, "grad_norm": 2.634312391281128, "learning_rate": 4.261478924769441e-07, "loss": 1.5927, "step": 44297 }, { "epoch": 3.471630094043887, "grad_norm": 7.687469005584717, "learning_rate": 4.2602333088925363e-07, "loss": 1.8997, "step": 44298 }, { "epoch": 3.4717084639498434, "grad_norm": 6.273512363433838, "learning_rate": 4.258987866986203e-07, "loss": 1.5881, "step": 44299 }, { "epoch": 3.4717868338557993, "grad_norm": 4.672452449798584, "learning_rate": 4.257742599055176e-07, "loss": 1.1967, "step": 44300 }, { "epoch": 3.4718652037617557, "grad_norm": 3.52315092086792, "learning_rate": 4.2564975051042126e-07, "loss": 1.2725, "step": 44301 }, { "epoch": 3.4719435736677116, "grad_norm": 4.2267985343933105, "learning_rate": 4.2552525851380423e-07, "loss": 0.7291, "step": 44302 }, { "epoch": 3.472021943573668, "grad_norm": 4.037499904632568, "learning_rate": 4.2540078391613894e-07, "loss": 1.2345, "step": 44303 }, { "epoch": 3.472100313479624, "grad_norm": 3.1232450008392334, "learning_rate": 4.252763267178989e-07, "loss": 1.2334, "step": 44304 }, { "epoch": 3.47217868338558, "grad_norm": 8.454214096069336, "learning_rate": 4.2515188691955864e-07, "loss": 1.2055, "step": 44305 }, { "epoch": 3.472257053291536, "grad_norm": 2.6308953762054443, "learning_rate": 4.250274645215902e-07, "loss": 1.3165, "step": 44306 }, { "epoch": 3.4723354231974923, "grad_norm": 4.617182731628418, "learning_rate": 4.2490305952446796e-07, "loss": 1.1808, "step": 44307 }, { "epoch": 3.472413793103448, "grad_norm": 4.381947994232178, "learning_rate": 4.247786719286645e-07, "loss": 1.3553, "step": 44308 }, { "epoch": 3.4724921630094046, "grad_norm": 8.78112506866455, "learning_rate": 4.246543017346527e-07, "loss": 0.6925, "step": 44309 }, { "epoch": 3.4725705329153604, "grad_norm": 4.193553924560547, "learning_rate": 4.245299489429061e-07, "loss": 1.6036, "step": 44310 }, { "epoch": 3.4726489028213168, "grad_norm": 3.361588478088379, "learning_rate": 4.2440561355389707e-07, "loss": 1.2463, "step": 44311 }, { "epoch": 3.4727272727272727, "grad_norm": 3.925401210784912, "learning_rate": 4.2428129556809973e-07, "loss": 1.8025, "step": 44312 }, { "epoch": 3.472805642633229, "grad_norm": 2.6503260135650635, "learning_rate": 4.2415699498598527e-07, "loss": 1.2505, "step": 44313 }, { "epoch": 3.472884012539185, "grad_norm": 3.069774866104126, "learning_rate": 4.2403271180802783e-07, "loss": 1.4533, "step": 44314 }, { "epoch": 3.472962382445141, "grad_norm": 4.450634479522705, "learning_rate": 4.2390844603469925e-07, "loss": 1.5655, "step": 44315 }, { "epoch": 3.473040752351097, "grad_norm": 5.151935577392578, "learning_rate": 4.2378419766647303e-07, "loss": 1.3281, "step": 44316 }, { "epoch": 3.4731191222570534, "grad_norm": 3.3245677947998047, "learning_rate": 4.236599667038216e-07, "loss": 1.3988, "step": 44317 }, { "epoch": 3.4731974921630093, "grad_norm": 3.3652327060699463, "learning_rate": 4.2353575314721564e-07, "loss": 1.3819, "step": 44318 }, { "epoch": 3.4732758620689657, "grad_norm": 4.692791938781738, "learning_rate": 4.2341155699712976e-07, "loss": 1.8, "step": 44319 }, { "epoch": 3.4733542319749215, "grad_norm": 3.29292368888855, "learning_rate": 4.2328737825403645e-07, "loss": 1.4345, "step": 44320 }, { "epoch": 3.473432601880878, "grad_norm": 5.262336730957031, "learning_rate": 4.2316321691840633e-07, "loss": 1.0831, "step": 44321 }, { "epoch": 3.4735109717868338, "grad_norm": 2.3485660552978516, "learning_rate": 4.230390729907125e-07, "loss": 1.1758, "step": 44322 }, { "epoch": 3.47358934169279, "grad_norm": 3.5352213382720947, "learning_rate": 4.2291494647142716e-07, "loss": 1.3935, "step": 44323 }, { "epoch": 3.473667711598746, "grad_norm": 3.524796485900879, "learning_rate": 4.2279083736102287e-07, "loss": 1.3176, "step": 44324 }, { "epoch": 3.4737460815047023, "grad_norm": 2.8421897888183594, "learning_rate": 4.226667456599709e-07, "loss": 1.2087, "step": 44325 }, { "epoch": 3.473824451410658, "grad_norm": 5.832832336425781, "learning_rate": 4.225426713687436e-07, "loss": 1.2476, "step": 44326 }, { "epoch": 3.4739028213166145, "grad_norm": 2.97741961479187, "learning_rate": 4.2241861448781283e-07, "loss": 1.5463, "step": 44327 }, { "epoch": 3.4739811912225704, "grad_norm": 3.0251057147979736, "learning_rate": 4.2229457501765045e-07, "loss": 0.9432, "step": 44328 }, { "epoch": 3.4740595611285268, "grad_norm": 3.3675038814544678, "learning_rate": 4.2217055295872834e-07, "loss": 1.0461, "step": 44329 }, { "epoch": 3.4741379310344827, "grad_norm": 6.523416519165039, "learning_rate": 4.220465483115177e-07, "loss": 1.4977, "step": 44330 }, { "epoch": 3.474216300940439, "grad_norm": 4.161031723022461, "learning_rate": 4.219225610764921e-07, "loss": 1.516, "step": 44331 }, { "epoch": 3.474294670846395, "grad_norm": 2.528876543045044, "learning_rate": 4.2179859125411947e-07, "loss": 0.7739, "step": 44332 }, { "epoch": 3.474373040752351, "grad_norm": 4.420321941375732, "learning_rate": 4.216746388448745e-07, "loss": 1.3374, "step": 44333 }, { "epoch": 3.474451410658307, "grad_norm": 4.987961769104004, "learning_rate": 4.2155070384922726e-07, "loss": 1.3939, "step": 44334 }, { "epoch": 3.4745297805642634, "grad_norm": 2.915865659713745, "learning_rate": 4.214267862676508e-07, "loss": 1.1872, "step": 44335 }, { "epoch": 3.4746081504702193, "grad_norm": 3.901895046234131, "learning_rate": 4.213028861006141e-07, "loss": 1.1669, "step": 44336 }, { "epoch": 3.4746865203761756, "grad_norm": 3.4143786430358887, "learning_rate": 4.211790033485885e-07, "loss": 1.1721, "step": 44337 }, { "epoch": 3.4747648902821315, "grad_norm": 3.8196823596954346, "learning_rate": 4.2105513801204754e-07, "loss": 0.8059, "step": 44338 }, { "epoch": 3.474843260188088, "grad_norm": 6.714043617248535, "learning_rate": 4.2093129009146137e-07, "loss": 1.3644, "step": 44339 }, { "epoch": 3.4749216300940438, "grad_norm": 4.121038436889648, "learning_rate": 4.208074595872996e-07, "loss": 1.5716, "step": 44340 }, { "epoch": 3.475, "grad_norm": 4.613077640533447, "learning_rate": 4.206836465000347e-07, "loss": 1.6323, "step": 44341 }, { "epoch": 3.475078369905956, "grad_norm": 2.7800817489624023, "learning_rate": 4.2055985083013616e-07, "loss": 1.466, "step": 44342 }, { "epoch": 3.4751567398119123, "grad_norm": 3.5422780513763428, "learning_rate": 4.204360725780771e-07, "loss": 1.7918, "step": 44343 }, { "epoch": 3.475235109717868, "grad_norm": 10.859424591064453, "learning_rate": 4.203123117443264e-07, "loss": 0.7828, "step": 44344 }, { "epoch": 3.4753134796238245, "grad_norm": 2.6207454204559326, "learning_rate": 4.201885683293555e-07, "loss": 0.9415, "step": 44345 }, { "epoch": 3.4753918495297804, "grad_norm": 3.0324594974517822, "learning_rate": 4.200648423336351e-07, "loss": 1.5563, "step": 44346 }, { "epoch": 3.4754702194357368, "grad_norm": 3.535660982131958, "learning_rate": 4.1994113375763536e-07, "loss": 0.9795, "step": 44347 }, { "epoch": 3.4755485893416926, "grad_norm": 3.07727313041687, "learning_rate": 4.19817442601827e-07, "loss": 1.1424, "step": 44348 }, { "epoch": 3.475626959247649, "grad_norm": 3.4132800102233887, "learning_rate": 4.196937688666808e-07, "loss": 1.6488, "step": 44349 }, { "epoch": 3.475705329153605, "grad_norm": 4.098738670349121, "learning_rate": 4.195701125526674e-07, "loss": 1.3389, "step": 44350 }, { "epoch": 3.475783699059561, "grad_norm": 3.4631669521331787, "learning_rate": 4.194464736602549e-07, "loss": 1.1419, "step": 44351 }, { "epoch": 3.475862068965517, "grad_norm": 2.6996374130249023, "learning_rate": 4.1932285218991675e-07, "loss": 1.1785, "step": 44352 }, { "epoch": 3.4759404388714734, "grad_norm": 2.3626773357391357, "learning_rate": 4.1919924814212145e-07, "loss": 1.2943, "step": 44353 }, { "epoch": 3.4760188087774293, "grad_norm": 2.5387074947357178, "learning_rate": 4.1907566151733967e-07, "loss": 1.1895, "step": 44354 }, { "epoch": 3.4760971786833856, "grad_norm": 2.723500967025757, "learning_rate": 4.1895209231604117e-07, "loss": 1.2264, "step": 44355 }, { "epoch": 3.4761755485893415, "grad_norm": 2.9472291469573975, "learning_rate": 4.1882854053869437e-07, "loss": 1.4633, "step": 44356 }, { "epoch": 3.476253918495298, "grad_norm": 3.3206989765167236, "learning_rate": 4.187050061857717e-07, "loss": 1.2845, "step": 44357 }, { "epoch": 3.4763322884012537, "grad_norm": 9.432717323303223, "learning_rate": 4.185814892577433e-07, "loss": 1.3195, "step": 44358 }, { "epoch": 3.47641065830721, "grad_norm": 2.5749895572662354, "learning_rate": 4.184579897550767e-07, "loss": 1.3792, "step": 44359 }, { "epoch": 3.476489028213166, "grad_norm": 2.567164182662964, "learning_rate": 4.18334507678243e-07, "loss": 1.2997, "step": 44360 }, { "epoch": 3.4765673981191223, "grad_norm": 3.616680145263672, "learning_rate": 4.1821104302771085e-07, "loss": 1.2177, "step": 44361 }, { "epoch": 3.476645768025078, "grad_norm": 2.7434189319610596, "learning_rate": 4.1808759580395096e-07, "loss": 1.5472, "step": 44362 }, { "epoch": 3.4767241379310345, "grad_norm": 2.764139413833618, "learning_rate": 4.179641660074324e-07, "loss": 1.2559, "step": 44363 }, { "epoch": 3.4768025078369904, "grad_norm": 5.430861949920654, "learning_rate": 4.1784075363862475e-07, "loss": 1.5089, "step": 44364 }, { "epoch": 3.4768808777429467, "grad_norm": 3.680595636367798, "learning_rate": 4.1771735869799767e-07, "loss": 1.2196, "step": 44365 }, { "epoch": 3.4769592476489026, "grad_norm": 4.69271183013916, "learning_rate": 4.1759398118601967e-07, "loss": 1.6132, "step": 44366 }, { "epoch": 3.477037617554859, "grad_norm": 3.5445327758789062, "learning_rate": 4.174706211031604e-07, "loss": 1.5351, "step": 44367 }, { "epoch": 3.477115987460815, "grad_norm": 3.9901015758514404, "learning_rate": 4.173472784498894e-07, "loss": 1.5875, "step": 44368 }, { "epoch": 3.477194357366771, "grad_norm": 11.105281829833984, "learning_rate": 4.172239532266764e-07, "loss": 0.8156, "step": 44369 }, { "epoch": 3.4772727272727275, "grad_norm": 3.579894781112671, "learning_rate": 4.1710064543398766e-07, "loss": 1.0689, "step": 44370 }, { "epoch": 3.4773510971786834, "grad_norm": 2.3040337562561035, "learning_rate": 4.1697735507229496e-07, "loss": 1.1547, "step": 44371 }, { "epoch": 3.4774294670846393, "grad_norm": 2.50917911529541, "learning_rate": 4.168540821420669e-07, "loss": 1.1783, "step": 44372 }, { "epoch": 3.4775078369905956, "grad_norm": 3.137779951095581, "learning_rate": 4.167308266437725e-07, "loss": 0.9426, "step": 44373 }, { "epoch": 3.477586206896552, "grad_norm": 3.3966710567474365, "learning_rate": 4.166075885778792e-07, "loss": 1.7091, "step": 44374 }, { "epoch": 3.477664576802508, "grad_norm": 3.177569627761841, "learning_rate": 4.164843679448549e-07, "loss": 1.5747, "step": 44375 }, { "epoch": 3.4777429467084637, "grad_norm": 2.9691007137298584, "learning_rate": 4.163611647451721e-07, "loss": 1.7257, "step": 44376 }, { "epoch": 3.47782131661442, "grad_norm": 2.3102097511291504, "learning_rate": 4.162379789792964e-07, "loss": 1.5981, "step": 44377 }, { "epoch": 3.4778996865203764, "grad_norm": 3.029019594192505, "learning_rate": 4.16114810647697e-07, "loss": 1.1695, "step": 44378 }, { "epoch": 3.4779780564263323, "grad_norm": 3.3552794456481934, "learning_rate": 4.1599165975084234e-07, "loss": 1.5956, "step": 44379 }, { "epoch": 3.478056426332288, "grad_norm": 3.0077316761016846, "learning_rate": 4.1586852628920095e-07, "loss": 1.3136, "step": 44380 }, { "epoch": 3.4781347962382445, "grad_norm": 4.299736499786377, "learning_rate": 4.1574541026324077e-07, "loss": 0.7248, "step": 44381 }, { "epoch": 3.478213166144201, "grad_norm": 3.945038080215454, "learning_rate": 4.1562231167343093e-07, "loss": 1.0733, "step": 44382 }, { "epoch": 3.4782915360501567, "grad_norm": 3.225649356842041, "learning_rate": 4.154992305202388e-07, "loss": 1.3176, "step": 44383 }, { "epoch": 3.4783699059561126, "grad_norm": 2.709815502166748, "learning_rate": 4.1537616680413287e-07, "loss": 1.2707, "step": 44384 }, { "epoch": 3.478448275862069, "grad_norm": 10.012740135192871, "learning_rate": 4.152531205255811e-07, "loss": 2.0146, "step": 44385 }, { "epoch": 3.4785266457680253, "grad_norm": 4.071288108825684, "learning_rate": 4.1513009168505203e-07, "loss": 1.8604, "step": 44386 }, { "epoch": 3.478605015673981, "grad_norm": 3.195796012878418, "learning_rate": 4.150070802830136e-07, "loss": 1.0701, "step": 44387 }, { "epoch": 3.478683385579937, "grad_norm": 2.2999863624572754, "learning_rate": 4.1488408631993273e-07, "loss": 1.6303, "step": 44388 }, { "epoch": 3.4787617554858934, "grad_norm": 3.5394606590270996, "learning_rate": 4.1476110979627613e-07, "loss": 1.542, "step": 44389 }, { "epoch": 3.4788401253918497, "grad_norm": 2.6969830989837646, "learning_rate": 4.1463815071251466e-07, "loss": 1.533, "step": 44390 }, { "epoch": 3.4789184952978056, "grad_norm": 5.902307033538818, "learning_rate": 4.1451520906911516e-07, "loss": 1.7397, "step": 44391 }, { "epoch": 3.478996865203762, "grad_norm": 11.374917030334473, "learning_rate": 4.143922848665438e-07, "loss": 0.8094, "step": 44392 }, { "epoch": 3.479075235109718, "grad_norm": 5.075275421142578, "learning_rate": 4.142693781052687e-07, "loss": 1.2724, "step": 44393 }, { "epoch": 3.479153605015674, "grad_norm": 4.304635524749756, "learning_rate": 4.1414648878575716e-07, "loss": 1.0222, "step": 44394 }, { "epoch": 3.47923197492163, "grad_norm": 3.501492500305176, "learning_rate": 4.140236169084777e-07, "loss": 1.434, "step": 44395 }, { "epoch": 3.4793103448275864, "grad_norm": 4.3617424964904785, "learning_rate": 4.1390076247389665e-07, "loss": 0.9112, "step": 44396 }, { "epoch": 3.4793887147335423, "grad_norm": 3.5437588691711426, "learning_rate": 4.1377792548248144e-07, "loss": 1.291, "step": 44397 }, { "epoch": 3.4794670846394986, "grad_norm": 3.606757879257202, "learning_rate": 4.1365510593469937e-07, "loss": 1.3313, "step": 44398 }, { "epoch": 3.4795454545454545, "grad_norm": 2.6645305156707764, "learning_rate": 4.1353230383101795e-07, "loss": 1.6108, "step": 44399 }, { "epoch": 3.479623824451411, "grad_norm": 3.172271251678467, "learning_rate": 4.1340951917190343e-07, "loss": 1.4187, "step": 44400 }, { "epoch": 3.4797021943573667, "grad_norm": 5.169336318969727, "learning_rate": 4.132867519578232e-07, "loss": 1.0476, "step": 44401 }, { "epoch": 3.479780564263323, "grad_norm": 3.587608575820923, "learning_rate": 4.1316400218924525e-07, "loss": 1.2214, "step": 44402 }, { "epoch": 3.479858934169279, "grad_norm": 3.3939311504364014, "learning_rate": 4.130412698666342e-07, "loss": 1.4845, "step": 44403 }, { "epoch": 3.4799373040752353, "grad_norm": 12.930977821350098, "learning_rate": 4.129185549904585e-07, "loss": 0.958, "step": 44404 }, { "epoch": 3.480015673981191, "grad_norm": 2.2009317874908447, "learning_rate": 4.127958575611846e-07, "loss": 1.4218, "step": 44405 }, { "epoch": 3.4800940438871475, "grad_norm": 2.3219921588897705, "learning_rate": 4.1267317757928026e-07, "loss": 1.2631, "step": 44406 }, { "epoch": 3.4801724137931034, "grad_norm": 4.703863620758057, "learning_rate": 4.1255051504520974e-07, "loss": 1.6351, "step": 44407 }, { "epoch": 3.4802507836990597, "grad_norm": 4.001689434051514, "learning_rate": 4.124278699594403e-07, "loss": 1.4014, "step": 44408 }, { "epoch": 3.4803291536050156, "grad_norm": 6.120872497558594, "learning_rate": 4.123052423224394e-07, "loss": 1.6549, "step": 44409 }, { "epoch": 3.480407523510972, "grad_norm": 3.1706595420837402, "learning_rate": 4.121826321346739e-07, "loss": 1.687, "step": 44410 }, { "epoch": 3.480485893416928, "grad_norm": 3.098140001296997, "learning_rate": 4.1206003939660844e-07, "loss": 1.5258, "step": 44411 }, { "epoch": 3.480564263322884, "grad_norm": 8.476705551147461, "learning_rate": 4.1193746410871036e-07, "loss": 1.1569, "step": 44412 }, { "epoch": 3.48064263322884, "grad_norm": 2.7105376720428467, "learning_rate": 4.118149062714444e-07, "loss": 1.3734, "step": 44413 }, { "epoch": 3.4807210031347964, "grad_norm": 2.901611328125, "learning_rate": 4.116923658852795e-07, "loss": 1.4652, "step": 44414 }, { "epoch": 3.4807993730407523, "grad_norm": 3.7820816040039062, "learning_rate": 4.115698429506798e-07, "loss": 1.5394, "step": 44415 }, { "epoch": 3.4808777429467086, "grad_norm": 10.022314071655273, "learning_rate": 4.1144733746811103e-07, "loss": 1.0745, "step": 44416 }, { "epoch": 3.4809561128526645, "grad_norm": 3.249567747116089, "learning_rate": 4.113248494380401e-07, "loss": 1.4396, "step": 44417 }, { "epoch": 3.481034482758621, "grad_norm": 3.053703546524048, "learning_rate": 4.112023788609326e-07, "loss": 1.1911, "step": 44418 }, { "epoch": 3.4811128526645767, "grad_norm": 3.2349135875701904, "learning_rate": 4.1107992573725387e-07, "loss": 1.2559, "step": 44419 }, { "epoch": 3.481191222570533, "grad_norm": 2.802868366241455, "learning_rate": 4.1095749006747064e-07, "loss": 1.154, "step": 44420 }, { "epoch": 3.481269592476489, "grad_norm": 6.49708366394043, "learning_rate": 4.1083507185204765e-07, "loss": 1.4843, "step": 44421 }, { "epoch": 3.4813479623824453, "grad_norm": 2.142141103744507, "learning_rate": 4.107126710914511e-07, "loss": 1.3759, "step": 44422 }, { "epoch": 3.481426332288401, "grad_norm": 3.373048782348633, "learning_rate": 4.1059028778614616e-07, "loss": 1.0257, "step": 44423 }, { "epoch": 3.4815047021943575, "grad_norm": 2.7866032123565674, "learning_rate": 4.1046792193659866e-07, "loss": 1.1723, "step": 44424 }, { "epoch": 3.4815830721003134, "grad_norm": 2.8858301639556885, "learning_rate": 4.1034557354327486e-07, "loss": 1.8636, "step": 44425 }, { "epoch": 3.4816614420062697, "grad_norm": 4.482527732849121, "learning_rate": 4.1022324260663715e-07, "loss": 1.3524, "step": 44426 }, { "epoch": 3.4817398119122256, "grad_norm": 2.5788094997406006, "learning_rate": 4.101009291271535e-07, "loss": 1.1623, "step": 44427 }, { "epoch": 3.481818181818182, "grad_norm": 4.798338413238525, "learning_rate": 4.099786331052885e-07, "loss": 1.3705, "step": 44428 }, { "epoch": 3.481896551724138, "grad_norm": 6.593638896942139, "learning_rate": 4.0985635454150853e-07, "loss": 1.7822, "step": 44429 }, { "epoch": 3.481974921630094, "grad_norm": 3.8786497116088867, "learning_rate": 4.097340934362759e-07, "loss": 1.4639, "step": 44430 }, { "epoch": 3.48205329153605, "grad_norm": 2.9360361099243164, "learning_rate": 4.0961184979005644e-07, "loss": 1.4218, "step": 44431 }, { "epoch": 3.4821316614420064, "grad_norm": 5.466203689575195, "learning_rate": 4.0948962360331743e-07, "loss": 0.6144, "step": 44432 }, { "epoch": 3.4822100313479623, "grad_norm": 2.4276020526885986, "learning_rate": 4.0936741487652087e-07, "loss": 1.2895, "step": 44433 }, { "epoch": 3.4822884012539186, "grad_norm": 2.6834120750427246, "learning_rate": 4.0924522361013294e-07, "loss": 1.1116, "step": 44434 }, { "epoch": 3.4823667711598745, "grad_norm": 2.691523790359497, "learning_rate": 4.0912304980461835e-07, "loss": 1.6384, "step": 44435 }, { "epoch": 3.482445141065831, "grad_norm": 3.3034896850585938, "learning_rate": 4.090008934604417e-07, "loss": 1.4956, "step": 44436 }, { "epoch": 3.4825235109717867, "grad_norm": 2.9330878257751465, "learning_rate": 4.0887875457806756e-07, "loss": 1.095, "step": 44437 }, { "epoch": 3.482601880877743, "grad_norm": 9.91880989074707, "learning_rate": 4.0875663315796066e-07, "loss": 1.1975, "step": 44438 }, { "epoch": 3.482680250783699, "grad_norm": 2.585775852203369, "learning_rate": 4.08634529200585e-07, "loss": 1.6866, "step": 44439 }, { "epoch": 3.4827586206896552, "grad_norm": 3.0305027961730957, "learning_rate": 4.085124427064052e-07, "loss": 1.138, "step": 44440 }, { "epoch": 3.482836990595611, "grad_norm": 6.118109703063965, "learning_rate": 4.0839037367588597e-07, "loss": 1.9255, "step": 44441 }, { "epoch": 3.4829153605015675, "grad_norm": 2.741246223449707, "learning_rate": 4.082683221094913e-07, "loss": 1.4601, "step": 44442 }, { "epoch": 3.4829937304075234, "grad_norm": 2.304917573928833, "learning_rate": 4.081462880076864e-07, "loss": 1.3154, "step": 44443 }, { "epoch": 3.4830721003134797, "grad_norm": 4.455516338348389, "learning_rate": 4.080242713709337e-07, "loss": 1.4401, "step": 44444 }, { "epoch": 3.4831504702194356, "grad_norm": 2.9350545406341553, "learning_rate": 4.0790227219969726e-07, "loss": 1.4246, "step": 44445 }, { "epoch": 3.483228840125392, "grad_norm": 2.384328603744507, "learning_rate": 4.077802904944428e-07, "loss": 1.081, "step": 44446 }, { "epoch": 3.483307210031348, "grad_norm": 8.150753021240234, "learning_rate": 4.0765832625563386e-07, "loss": 1.4051, "step": 44447 }, { "epoch": 3.483385579937304, "grad_norm": 2.8747851848602295, "learning_rate": 4.0753637948373337e-07, "loss": 1.1742, "step": 44448 }, { "epoch": 3.48346394984326, "grad_norm": 4.971201419830322, "learning_rate": 4.0741445017920545e-07, "loss": 1.2238, "step": 44449 }, { "epoch": 3.4835423197492164, "grad_norm": 7.413704872131348, "learning_rate": 4.0729253834251306e-07, "loss": 1.3696, "step": 44450 }, { "epoch": 3.4836206896551722, "grad_norm": 2.446882724761963, "learning_rate": 4.07170643974123e-07, "loss": 1.1155, "step": 44451 }, { "epoch": 3.4836990595611286, "grad_norm": 5.481801509857178, "learning_rate": 4.0704876707449603e-07, "loss": 1.388, "step": 44452 }, { "epoch": 3.4837774294670845, "grad_norm": 3.2085776329040527, "learning_rate": 4.069269076440957e-07, "loss": 1.5394, "step": 44453 }, { "epoch": 3.483855799373041, "grad_norm": 3.9787025451660156, "learning_rate": 4.0680506568338715e-07, "loss": 0.9118, "step": 44454 }, { "epoch": 3.4839341692789967, "grad_norm": 3.7143666744232178, "learning_rate": 4.0668324119283233e-07, "loss": 1.5092, "step": 44455 }, { "epoch": 3.484012539184953, "grad_norm": 2.5282654762268066, "learning_rate": 4.0656143417289517e-07, "loss": 1.2548, "step": 44456 }, { "epoch": 3.484090909090909, "grad_norm": 3.920833110809326, "learning_rate": 4.064396446240393e-07, "loss": 1.4174, "step": 44457 }, { "epoch": 3.4841692789968652, "grad_norm": 2.7469048500061035, "learning_rate": 4.063178725467287e-07, "loss": 1.0187, "step": 44458 }, { "epoch": 3.484247648902821, "grad_norm": 4.039836883544922, "learning_rate": 4.061961179414231e-07, "loss": 1.3877, "step": 44459 }, { "epoch": 3.4843260188087775, "grad_norm": 3.578477621078491, "learning_rate": 4.060743808085893e-07, "loss": 1.3313, "step": 44460 }, { "epoch": 3.484404388714734, "grad_norm": 3.344728708267212, "learning_rate": 4.059526611486886e-07, "loss": 1.7267, "step": 44461 }, { "epoch": 3.4844827586206897, "grad_norm": 3.116772413253784, "learning_rate": 4.058309589621856e-07, "loss": 1.1053, "step": 44462 }, { "epoch": 3.4845611285266456, "grad_norm": 2.8393492698669434, "learning_rate": 4.057092742495405e-07, "loss": 1.2267, "step": 44463 }, { "epoch": 3.484639498432602, "grad_norm": 2.63149356842041, "learning_rate": 4.055876070112169e-07, "loss": 1.2737, "step": 44464 }, { "epoch": 3.4847178683385582, "grad_norm": 4.837772846221924, "learning_rate": 4.0546595724767936e-07, "loss": 1.2934, "step": 44465 }, { "epoch": 3.484796238244514, "grad_norm": 5.419539928436279, "learning_rate": 4.0534432495939024e-07, "loss": 1.3355, "step": 44466 }, { "epoch": 3.48487460815047, "grad_norm": 2.646559000015259, "learning_rate": 4.0522271014681035e-07, "loss": 1.5256, "step": 44467 }, { "epoch": 3.4849529780564263, "grad_norm": 3.0781190395355225, "learning_rate": 4.051011128104032e-07, "loss": 1.2259, "step": 44468 }, { "epoch": 3.4850313479623827, "grad_norm": 4.645244598388672, "learning_rate": 4.0497953295063007e-07, "loss": 1.2482, "step": 44469 }, { "epoch": 3.4851097178683386, "grad_norm": 4.155318737030029, "learning_rate": 4.0485797056795675e-07, "loss": 1.1849, "step": 44470 }, { "epoch": 3.4851880877742945, "grad_norm": 25.399906158447266, "learning_rate": 4.047364256628428e-07, "loss": 1.4346, "step": 44471 }, { "epoch": 3.485266457680251, "grad_norm": 2.915726661682129, "learning_rate": 4.046148982357512e-07, "loss": 1.3162, "step": 44472 }, { "epoch": 3.485344827586207, "grad_norm": 2.8174808025360107, "learning_rate": 4.044933882871438e-07, "loss": 0.9975, "step": 44473 }, { "epoch": 3.485423197492163, "grad_norm": 5.753925800323486, "learning_rate": 4.0437189581748357e-07, "loss": 0.8884, "step": 44474 }, { "epoch": 3.485501567398119, "grad_norm": 3.805272340774536, "learning_rate": 4.042504208272319e-07, "loss": 1.7734, "step": 44475 }, { "epoch": 3.4855799373040752, "grad_norm": 8.34754467010498, "learning_rate": 4.041289633168516e-07, "loss": 1.0404, "step": 44476 }, { "epoch": 3.4856583072100316, "grad_norm": 16.195589065551758, "learning_rate": 4.040075232868046e-07, "loss": 1.0155, "step": 44477 }, { "epoch": 3.4857366771159874, "grad_norm": 4.614210605621338, "learning_rate": 4.0388610073755054e-07, "loss": 1.4747, "step": 44478 }, { "epoch": 3.4858150470219433, "grad_norm": 5.259486675262451, "learning_rate": 4.03764695669554e-07, "loss": 1.1691, "step": 44479 }, { "epoch": 3.4858934169278997, "grad_norm": 2.9223403930664062, "learning_rate": 4.036433080832758e-07, "loss": 1.5686, "step": 44480 }, { "epoch": 3.485971786833856, "grad_norm": 5.572432994842529, "learning_rate": 4.035219379791783e-07, "loss": 0.9163, "step": 44481 }, { "epoch": 3.486050156739812, "grad_norm": 3.252638339996338, "learning_rate": 4.0340058535772163e-07, "loss": 1.3792, "step": 44482 }, { "epoch": 3.4861285266457678, "grad_norm": 3.0277886390686035, "learning_rate": 4.032792502193672e-07, "loss": 1.4881, "step": 44483 }, { "epoch": 3.486206896551724, "grad_norm": 5.4519782066345215, "learning_rate": 4.0315793256457846e-07, "loss": 1.8773, "step": 44484 }, { "epoch": 3.4862852664576804, "grad_norm": 5.871988773345947, "learning_rate": 4.030366323938173e-07, "loss": 1.0031, "step": 44485 }, { "epoch": 3.4863636363636363, "grad_norm": 3.5691287517547607, "learning_rate": 4.029153497075422e-07, "loss": 1.3444, "step": 44486 }, { "epoch": 3.4864420062695927, "grad_norm": 5.371336460113525, "learning_rate": 4.027940845062156e-07, "loss": 1.5707, "step": 44487 }, { "epoch": 3.4865203761755486, "grad_norm": 2.9523606300354004, "learning_rate": 4.026728367902999e-07, "loss": 1.3144, "step": 44488 }, { "epoch": 3.486598746081505, "grad_norm": 5.6657304763793945, "learning_rate": 4.025516065602547e-07, "loss": 0.9466, "step": 44489 }, { "epoch": 3.4866771159874608, "grad_norm": 2.721747636795044, "learning_rate": 4.024303938165419e-07, "loss": 1.42, "step": 44490 }, { "epoch": 3.486755485893417, "grad_norm": 5.950657367706299, "learning_rate": 4.023091985596228e-07, "loss": 1.3204, "step": 44491 }, { "epoch": 3.486833855799373, "grad_norm": 7.124434471130371, "learning_rate": 4.0218802078995753e-07, "loss": 1.6802, "step": 44492 }, { "epoch": 3.4869122257053293, "grad_norm": 4.297311305999756, "learning_rate": 4.0206686050800747e-07, "loss": 1.1457, "step": 44493 }, { "epoch": 3.486990595611285, "grad_norm": 2.6067276000976562, "learning_rate": 4.0194571771423385e-07, "loss": 1.3719, "step": 44494 }, { "epoch": 3.4870689655172415, "grad_norm": 2.4379208087921143, "learning_rate": 4.0182459240909634e-07, "loss": 1.4197, "step": 44495 }, { "epoch": 3.4871473354231974, "grad_norm": 3.1147022247314453, "learning_rate": 4.0170348459305784e-07, "loss": 1.0126, "step": 44496 }, { "epoch": 3.4872257053291538, "grad_norm": 4.0884857177734375, "learning_rate": 4.015823942665753e-07, "loss": 1.4667, "step": 44497 }, { "epoch": 3.4873040752351097, "grad_norm": 2.423480987548828, "learning_rate": 4.014613214301127e-07, "loss": 1.0889, "step": 44498 }, { "epoch": 3.487382445141066, "grad_norm": 2.932446241378784, "learning_rate": 4.013402660841298e-07, "loss": 1.4607, "step": 44499 }, { "epoch": 3.487460815047022, "grad_norm": 13.134488105773926, "learning_rate": 4.0121922822908556e-07, "loss": 1.6592, "step": 44500 }, { "epoch": 3.487539184952978, "grad_norm": 2.907397985458374, "learning_rate": 4.0109820786544127e-07, "loss": 1.4286, "step": 44501 }, { "epoch": 3.487617554858934, "grad_norm": 2.9369449615478516, "learning_rate": 4.009772049936567e-07, "loss": 1.2616, "step": 44502 }, { "epoch": 3.4876959247648904, "grad_norm": 3.589224100112915, "learning_rate": 4.0085621961419465e-07, "loss": 1.4337, "step": 44503 }, { "epoch": 3.4877742946708463, "grad_norm": 5.548171043395996, "learning_rate": 4.0073525172751147e-07, "loss": 1.636, "step": 44504 }, { "epoch": 3.4878526645768027, "grad_norm": 3.222196102142334, "learning_rate": 4.0061430133406966e-07, "loss": 1.0646, "step": 44505 }, { "epoch": 3.4879310344827585, "grad_norm": 2.893002986907959, "learning_rate": 4.0049336843432874e-07, "loss": 1.4745, "step": 44506 }, { "epoch": 3.488009404388715, "grad_norm": 2.987736225128174, "learning_rate": 4.003724530287484e-07, "loss": 1.3147, "step": 44507 }, { "epoch": 3.4880877742946708, "grad_norm": 3.3616726398468018, "learning_rate": 4.002515551177888e-07, "loss": 1.3565, "step": 44508 }, { "epoch": 3.488166144200627, "grad_norm": 4.073883056640625, "learning_rate": 4.001306747019096e-07, "loss": 1.2306, "step": 44509 }, { "epoch": 3.488244514106583, "grad_norm": 3.5808446407318115, "learning_rate": 4.0000981178157095e-07, "loss": 1.3445, "step": 44510 }, { "epoch": 3.4883228840125393, "grad_norm": 4.357408046722412, "learning_rate": 3.9988896635723193e-07, "loss": 1.1914, "step": 44511 }, { "epoch": 3.488401253918495, "grad_norm": 3.394190549850464, "learning_rate": 3.997681384293528e-07, "loss": 0.9362, "step": 44512 }, { "epoch": 3.4884796238244515, "grad_norm": 5.3594231605529785, "learning_rate": 3.9964732799839255e-07, "loss": 1.4025, "step": 44513 }, { "epoch": 3.4885579937304074, "grad_norm": 3.978030204772949, "learning_rate": 3.99526535064812e-07, "loss": 1.2885, "step": 44514 }, { "epoch": 3.4886363636363638, "grad_norm": 6.673266410827637, "learning_rate": 3.994057596290679e-07, "loss": 0.7191, "step": 44515 }, { "epoch": 3.4887147335423196, "grad_norm": 2.8208253383636475, "learning_rate": 3.9928500169162167e-07, "loss": 0.8584, "step": 44516 }, { "epoch": 3.488793103448276, "grad_norm": 12.9616117477417, "learning_rate": 3.9916426125293286e-07, "loss": 2.277, "step": 44517 }, { "epoch": 3.488871473354232, "grad_norm": 3.4565811157226562, "learning_rate": 3.9904353831346056e-07, "loss": 1.4794, "step": 44518 }, { "epoch": 3.488949843260188, "grad_norm": 11.000204086303711, "learning_rate": 3.9892283287366274e-07, "loss": 1.1605, "step": 44519 }, { "epoch": 3.489028213166144, "grad_norm": 5.036452293395996, "learning_rate": 3.9880214493399903e-07, "loss": 2.0908, "step": 44520 }, { "epoch": 3.4891065830721004, "grad_norm": 2.7285146713256836, "learning_rate": 3.98681474494928e-07, "loss": 1.0445, "step": 44521 }, { "epoch": 3.4891849529780563, "grad_norm": 2.737545967102051, "learning_rate": 3.985608215569109e-07, "loss": 1.5334, "step": 44522 }, { "epoch": 3.4892633228840126, "grad_norm": 1.9319556951522827, "learning_rate": 3.9844018612040404e-07, "loss": 1.2871, "step": 44523 }, { "epoch": 3.4893416927899685, "grad_norm": 3.6564831733703613, "learning_rate": 3.9831956818586704e-07, "loss": 1.3334, "step": 44524 }, { "epoch": 3.489420062695925, "grad_norm": 4.071985721588135, "learning_rate": 3.9819896775375897e-07, "loss": 1.5617, "step": 44525 }, { "epoch": 3.4894984326018808, "grad_norm": 2.912266254425049, "learning_rate": 3.9807838482453844e-07, "loss": 1.3225, "step": 44526 }, { "epoch": 3.489576802507837, "grad_norm": 3.28409481048584, "learning_rate": 3.979578193986644e-07, "loss": 1.4948, "step": 44527 }, { "epoch": 3.489655172413793, "grad_norm": 3.079286575317383, "learning_rate": 3.9783727147659433e-07, "loss": 1.3434, "step": 44528 }, { "epoch": 3.4897335423197493, "grad_norm": 4.655503749847412, "learning_rate": 3.9771674105878787e-07, "loss": 0.9719, "step": 44529 }, { "epoch": 3.489811912225705, "grad_norm": 5.656576633453369, "learning_rate": 3.975962281457035e-07, "loss": 0.8782, "step": 44530 }, { "epoch": 3.4898902821316615, "grad_norm": 3.9715685844421387, "learning_rate": 3.9747573273779816e-07, "loss": 0.7896, "step": 44531 }, { "epoch": 3.4899686520376174, "grad_norm": 5.869069576263428, "learning_rate": 3.9735525483553193e-07, "loss": 1.6781, "step": 44532 }, { "epoch": 3.4900470219435737, "grad_norm": 3.8780481815338135, "learning_rate": 3.9723479443936286e-07, "loss": 1.6706, "step": 44533 }, { "epoch": 3.4901253918495296, "grad_norm": 2.8302128314971924, "learning_rate": 3.9711435154974666e-07, "loss": 1.273, "step": 44534 }, { "epoch": 3.490203761755486, "grad_norm": 4.475041389465332, "learning_rate": 3.9699392616714407e-07, "loss": 1.3842, "step": 44535 }, { "epoch": 3.490282131661442, "grad_norm": 5.194573879241943, "learning_rate": 3.968735182920125e-07, "loss": 1.5541, "step": 44536 }, { "epoch": 3.490360501567398, "grad_norm": 10.34485149383545, "learning_rate": 3.9675312792481044e-07, "loss": 1.3393, "step": 44537 }, { "epoch": 3.490438871473354, "grad_norm": 2.6861560344696045, "learning_rate": 3.966327550659943e-07, "loss": 1.0876, "step": 44538 }, { "epoch": 3.4905172413793104, "grad_norm": 4.424667835235596, "learning_rate": 3.9651239971602195e-07, "loss": 1.341, "step": 44539 }, { "epoch": 3.4905956112852663, "grad_norm": 3.5990383625030518, "learning_rate": 3.9639206187535363e-07, "loss": 1.0906, "step": 44540 }, { "epoch": 3.4906739811912226, "grad_norm": 2.761122226715088, "learning_rate": 3.9627174154444446e-07, "loss": 1.6489, "step": 44541 }, { "epoch": 3.4907523510971785, "grad_norm": 8.258445739746094, "learning_rate": 3.9615143872375306e-07, "loss": 0.7332, "step": 44542 }, { "epoch": 3.490830721003135, "grad_norm": 9.331975936889648, "learning_rate": 3.9603115341373677e-07, "loss": 1.2556, "step": 44543 }, { "epoch": 3.4909090909090907, "grad_norm": 16.288637161254883, "learning_rate": 3.959108856148536e-07, "loss": 1.5264, "step": 44544 }, { "epoch": 3.490987460815047, "grad_norm": 3.3187403678894043, "learning_rate": 3.9579063532756035e-07, "loss": 1.3613, "step": 44545 }, { "epoch": 3.491065830721003, "grad_norm": 3.4662818908691406, "learning_rate": 3.956704025523145e-07, "loss": 1.2816, "step": 44546 }, { "epoch": 3.4911442006269593, "grad_norm": 4.788219928741455, "learning_rate": 3.955501872895734e-07, "loss": 1.1197, "step": 44547 }, { "epoch": 3.491222570532915, "grad_norm": 3.4861457347869873, "learning_rate": 3.954299895397945e-07, "loss": 1.0622, "step": 44548 }, { "epoch": 3.4913009404388715, "grad_norm": 5.281210899353027, "learning_rate": 3.953098093034352e-07, "loss": 1.3094, "step": 44549 }, { "epoch": 3.4913793103448274, "grad_norm": 4.662204265594482, "learning_rate": 3.951896465809518e-07, "loss": 1.4134, "step": 44550 }, { "epoch": 3.4914576802507837, "grad_norm": 6.539426326751709, "learning_rate": 3.9506950137280233e-07, "loss": 1.1422, "step": 44551 }, { "epoch": 3.4915360501567396, "grad_norm": 2.6697635650634766, "learning_rate": 3.9494937367944354e-07, "loss": 1.3141, "step": 44552 }, { "epoch": 3.491614420062696, "grad_norm": 4.529217720031738, "learning_rate": 3.9482926350133066e-07, "loss": 1.5018, "step": 44553 }, { "epoch": 3.491692789968652, "grad_norm": 5.0155253410339355, "learning_rate": 3.947091708389228e-07, "loss": 1.2119, "step": 44554 }, { "epoch": 3.491771159874608, "grad_norm": 2.470336675643921, "learning_rate": 3.945890956926762e-07, "loss": 0.9124, "step": 44555 }, { "epoch": 3.4918495297805645, "grad_norm": 2.482469081878662, "learning_rate": 3.944690380630467e-07, "loss": 1.0774, "step": 44556 }, { "epoch": 3.4919278996865204, "grad_norm": 3.6251637935638428, "learning_rate": 3.9434899795049166e-07, "loss": 1.2531, "step": 44557 }, { "epoch": 3.4920062695924763, "grad_norm": 8.51772689819336, "learning_rate": 3.942289753554662e-07, "loss": 0.5159, "step": 44558 }, { "epoch": 3.4920846394984326, "grad_norm": 5.005080223083496, "learning_rate": 3.941089702784295e-07, "loss": 1.6319, "step": 44559 }, { "epoch": 3.492163009404389, "grad_norm": 3.422481060028076, "learning_rate": 3.939889827198362e-07, "loss": 0.7299, "step": 44560 }, { "epoch": 3.492241379310345, "grad_norm": 3.1366357803344727, "learning_rate": 3.938690126801431e-07, "loss": 1.3074, "step": 44561 }, { "epoch": 3.4923197492163007, "grad_norm": 5.928165912628174, "learning_rate": 3.9374906015980597e-07, "loss": 1.5262, "step": 44562 }, { "epoch": 3.492398119122257, "grad_norm": 8.530935287475586, "learning_rate": 3.9362912515928165e-07, "loss": 1.6696, "step": 44563 }, { "epoch": 3.4924764890282134, "grad_norm": 8.03000259399414, "learning_rate": 3.9350920767902647e-07, "loss": 1.5885, "step": 44564 }, { "epoch": 3.4925548589341693, "grad_norm": 3.561068296432495, "learning_rate": 3.933893077194956e-07, "loss": 1.1829, "step": 44565 }, { "epoch": 3.492633228840125, "grad_norm": 2.8352439403533936, "learning_rate": 3.9326942528114644e-07, "loss": 1.2367, "step": 44566 }, { "epoch": 3.4927115987460815, "grad_norm": 5.021385669708252, "learning_rate": 3.9314956036443364e-07, "loss": 1.6716, "step": 44567 }, { "epoch": 3.492789968652038, "grad_norm": 4.8324737548828125, "learning_rate": 3.9302971296981406e-07, "loss": 1.3576, "step": 44568 }, { "epoch": 3.4928683385579937, "grad_norm": 3.360809564590454, "learning_rate": 3.9290988309774293e-07, "loss": 1.8131, "step": 44569 }, { "epoch": 3.4929467084639496, "grad_norm": 4.397858619689941, "learning_rate": 3.92790070748677e-07, "loss": 1.2819, "step": 44570 }, { "epoch": 3.493025078369906, "grad_norm": 3.354909658432007, "learning_rate": 3.926702759230705e-07, "loss": 1.2083, "step": 44571 }, { "epoch": 3.4931034482758623, "grad_norm": 2.716677188873291, "learning_rate": 3.9255049862137904e-07, "loss": 1.2499, "step": 44572 }, { "epoch": 3.493181818181818, "grad_norm": 2.9077539443969727, "learning_rate": 3.924307388440596e-07, "loss": 1.5679, "step": 44573 }, { "epoch": 3.493260188087774, "grad_norm": 5.262303829193115, "learning_rate": 3.9231099659156835e-07, "loss": 1.1398, "step": 44574 }, { "epoch": 3.4933385579937304, "grad_norm": 3.5330705642700195, "learning_rate": 3.9219127186435834e-07, "loss": 1.0914, "step": 44575 }, { "epoch": 3.4934169278996867, "grad_norm": 3.314434766769409, "learning_rate": 3.9207156466288587e-07, "loss": 1.0578, "step": 44576 }, { "epoch": 3.4934952978056426, "grad_norm": 4.317862033843994, "learning_rate": 3.9195187498760556e-07, "loss": 1.8093, "step": 44577 }, { "epoch": 3.493573667711599, "grad_norm": 2.318450927734375, "learning_rate": 3.9183220283897483e-07, "loss": 1.1255, "step": 44578 }, { "epoch": 3.493652037617555, "grad_norm": 2.759896993637085, "learning_rate": 3.917125482174472e-07, "loss": 1.3162, "step": 44579 }, { "epoch": 3.493730407523511, "grad_norm": 2.769275188446045, "learning_rate": 3.9159291112347787e-07, "loss": 1.2743, "step": 44580 }, { "epoch": 3.493808777429467, "grad_norm": 5.929393291473389, "learning_rate": 3.91473291557522e-07, "loss": 1.1431, "step": 44581 }, { "epoch": 3.4938871473354234, "grad_norm": 2.268686294555664, "learning_rate": 3.9135368952003483e-07, "loss": 1.1483, "step": 44582 }, { "epoch": 3.4939655172413793, "grad_norm": 4.116063594818115, "learning_rate": 3.9123410501147095e-07, "loss": 1.1119, "step": 44583 }, { "epoch": 3.4940438871473356, "grad_norm": 3.3862743377685547, "learning_rate": 3.9111453803228504e-07, "loss": 1.2078, "step": 44584 }, { "epoch": 3.4941222570532915, "grad_norm": 5.557584285736084, "learning_rate": 3.909949885829334e-07, "loss": 1.8283, "step": 44585 }, { "epoch": 3.494200626959248, "grad_norm": 4.95035457611084, "learning_rate": 3.9087545666386784e-07, "loss": 1.6787, "step": 44586 }, { "epoch": 3.4942789968652037, "grad_norm": 4.663400650024414, "learning_rate": 3.9075594227554525e-07, "loss": 1.2693, "step": 44587 }, { "epoch": 3.49435736677116, "grad_norm": 3.442667245864868, "learning_rate": 3.906364454184197e-07, "loss": 0.7228, "step": 44588 }, { "epoch": 3.494435736677116, "grad_norm": 3.155784845352173, "learning_rate": 3.9051696609294645e-07, "loss": 1.1585, "step": 44589 }, { "epoch": 3.4945141065830723, "grad_norm": 3.968778610229492, "learning_rate": 3.9039750429957835e-07, "loss": 1.17, "step": 44590 }, { "epoch": 3.494592476489028, "grad_norm": 6.001358985900879, "learning_rate": 3.9027806003876957e-07, "loss": 0.9875, "step": 44591 }, { "epoch": 3.4946708463949845, "grad_norm": 3.7869420051574707, "learning_rate": 3.901586333109758e-07, "loss": 1.5698, "step": 44592 }, { "epoch": 3.4947492163009404, "grad_norm": 4.302072525024414, "learning_rate": 3.900392241166523e-07, "loss": 0.8982, "step": 44593 }, { "epoch": 3.4948275862068967, "grad_norm": 7.979471206665039, "learning_rate": 3.8991983245625087e-07, "loss": 1.1968, "step": 44594 }, { "epoch": 3.4949059561128526, "grad_norm": 4.867201328277588, "learning_rate": 3.8980045833022615e-07, "loss": 1.3933, "step": 44595 }, { "epoch": 3.494984326018809, "grad_norm": 3.4541566371917725, "learning_rate": 3.8968110173903286e-07, "loss": 1.4809, "step": 44596 }, { "epoch": 3.495062695924765, "grad_norm": 3.200495481491089, "learning_rate": 3.895617626831244e-07, "loss": 1.3659, "step": 44597 }, { "epoch": 3.495141065830721, "grad_norm": 3.279703378677368, "learning_rate": 3.8944244116295494e-07, "loss": 1.1332, "step": 44598 }, { "epoch": 3.495219435736677, "grad_norm": 5.615241527557373, "learning_rate": 3.8932313717897795e-07, "loss": 1.1538, "step": 44599 }, { "epoch": 3.4952978056426334, "grad_norm": 3.6545674800872803, "learning_rate": 3.892038507316481e-07, "loss": 0.9725, "step": 44600 }, { "epoch": 3.4953761755485893, "grad_norm": 2.274148464202881, "learning_rate": 3.890845818214184e-07, "loss": 1.3009, "step": 44601 }, { "epoch": 3.4954545454545456, "grad_norm": 3.0177478790283203, "learning_rate": 3.8896533044874283e-07, "loss": 1.5093, "step": 44602 }, { "epoch": 3.4955329153605015, "grad_norm": 2.128840684890747, "learning_rate": 3.8884609661407445e-07, "loss": 1.2221, "step": 44603 }, { "epoch": 3.495611285266458, "grad_norm": 2.7950117588043213, "learning_rate": 3.887268803178679e-07, "loss": 1.1309, "step": 44604 }, { "epoch": 3.4956896551724137, "grad_norm": 4.972599983215332, "learning_rate": 3.886076815605744e-07, "loss": 1.0092, "step": 44605 }, { "epoch": 3.49576802507837, "grad_norm": 6.584280014038086, "learning_rate": 3.884885003426492e-07, "loss": 1.6265, "step": 44606 }, { "epoch": 3.495846394984326, "grad_norm": 3.0298845767974854, "learning_rate": 3.8836933666454526e-07, "loss": 1.1732, "step": 44607 }, { "epoch": 3.4959247648902823, "grad_norm": 2.716562509536743, "learning_rate": 3.8825019052671664e-07, "loss": 1.2494, "step": 44608 }, { "epoch": 3.496003134796238, "grad_norm": 6.981986999511719, "learning_rate": 3.881310619296147e-07, "loss": 1.6395, "step": 44609 }, { "epoch": 3.4960815047021945, "grad_norm": 4.133040428161621, "learning_rate": 3.880119508736929e-07, "loss": 1.0475, "step": 44610 }, { "epoch": 3.4961598746081504, "grad_norm": 2.0778911113739014, "learning_rate": 3.87892857359406e-07, "loss": 1.0223, "step": 44611 }, { "epoch": 3.4962382445141067, "grad_norm": 6.341180801391602, "learning_rate": 3.877737813872046e-07, "loss": 1.7573, "step": 44612 }, { "epoch": 3.4963166144200626, "grad_norm": 2.2043039798736572, "learning_rate": 3.8765472295754336e-07, "loss": 1.2645, "step": 44613 }, { "epoch": 3.496394984326019, "grad_norm": 2.7521207332611084, "learning_rate": 3.875356820708742e-07, "loss": 1.2067, "step": 44614 }, { "epoch": 3.496473354231975, "grad_norm": 3.6264445781707764, "learning_rate": 3.874166587276501e-07, "loss": 1.5987, "step": 44615 }, { "epoch": 3.496551724137931, "grad_norm": 3.558584690093994, "learning_rate": 3.87297652928324e-07, "loss": 1.2472, "step": 44616 }, { "epoch": 3.496630094043887, "grad_norm": 3.088986396789551, "learning_rate": 3.8717866467334887e-07, "loss": 1.2795, "step": 44617 }, { "epoch": 3.4967084639498434, "grad_norm": 2.582014560699463, "learning_rate": 3.87059693963176e-07, "loss": 1.2512, "step": 44618 }, { "epoch": 3.4967868338557992, "grad_norm": 3.16158127784729, "learning_rate": 3.869407407982589e-07, "loss": 1.3019, "step": 44619 }, { "epoch": 3.4968652037617556, "grad_norm": 4.520990371704102, "learning_rate": 3.868218051790501e-07, "loss": 1.1949, "step": 44620 }, { "epoch": 3.4969435736677115, "grad_norm": 6.935237407684326, "learning_rate": 3.867028871060019e-07, "loss": 1.4816, "step": 44621 }, { "epoch": 3.497021943573668, "grad_norm": 3.3914270401000977, "learning_rate": 3.8658398657956564e-07, "loss": 1.4384, "step": 44622 }, { "epoch": 3.4971003134796237, "grad_norm": 5.010814666748047, "learning_rate": 3.8646510360019486e-07, "loss": 1.2671, "step": 44623 }, { "epoch": 3.49717868338558, "grad_norm": 4.171885013580322, "learning_rate": 3.8634623816834084e-07, "loss": 1.1769, "step": 44624 }, { "epoch": 3.497257053291536, "grad_norm": 4.546619415283203, "learning_rate": 3.86227390284456e-07, "loss": 1.3415, "step": 44625 }, { "epoch": 3.4973354231974922, "grad_norm": 7.837762355804443, "learning_rate": 3.8610855994899334e-07, "loss": 1.0777, "step": 44626 }, { "epoch": 3.497413793103448, "grad_norm": 3.5117404460906982, "learning_rate": 3.85989747162403e-07, "loss": 0.9346, "step": 44627 }, { "epoch": 3.4974921630094045, "grad_norm": 3.3118972778320312, "learning_rate": 3.858709519251369e-07, "loss": 1.2374, "step": 44628 }, { "epoch": 3.4975705329153604, "grad_norm": 7.998197078704834, "learning_rate": 3.857521742376491e-07, "loss": 0.6536, "step": 44629 }, { "epoch": 3.4976489028213167, "grad_norm": 4.913949012756348, "learning_rate": 3.8563341410039033e-07, "loss": 1.1251, "step": 44630 }, { "epoch": 3.4977272727272726, "grad_norm": 6.049412250518799, "learning_rate": 3.8551467151381084e-07, "loss": 1.7144, "step": 44631 }, { "epoch": 3.497805642633229, "grad_norm": 2.9788506031036377, "learning_rate": 3.853959464783641e-07, "loss": 1.3417, "step": 44632 }, { "epoch": 3.497884012539185, "grad_norm": 2.6526684761047363, "learning_rate": 3.8527723899450033e-07, "loss": 0.8951, "step": 44633 }, { "epoch": 3.497962382445141, "grad_norm": 2.85532808303833, "learning_rate": 3.8515854906267193e-07, "loss": 1.3477, "step": 44634 }, { "epoch": 3.498040752351097, "grad_norm": 2.6254358291625977, "learning_rate": 3.850398766833302e-07, "loss": 1.3307, "step": 44635 }, { "epoch": 3.4981191222570533, "grad_norm": 8.527053833007812, "learning_rate": 3.8492122185692647e-07, "loss": 1.6106, "step": 44636 }, { "epoch": 3.4981974921630092, "grad_norm": 2.5429205894470215, "learning_rate": 3.848025845839115e-07, "loss": 1.1531, "step": 44637 }, { "epoch": 3.4982758620689656, "grad_norm": 2.0192301273345947, "learning_rate": 3.8468396486473716e-07, "loss": 1.5588, "step": 44638 }, { "epoch": 3.4983542319749215, "grad_norm": 4.7935686111450195, "learning_rate": 3.845653626998547e-07, "loss": 1.5183, "step": 44639 }, { "epoch": 3.498432601880878, "grad_norm": 2.985985517501831, "learning_rate": 3.8444677808971486e-07, "loss": 1.4582, "step": 44640 }, { "epoch": 3.4985109717868337, "grad_norm": 7.7015838623046875, "learning_rate": 3.8432821103476956e-07, "loss": 0.8532, "step": 44641 }, { "epoch": 3.49858934169279, "grad_norm": 2.5979526042938232, "learning_rate": 3.8420966153546735e-07, "loss": 1.1504, "step": 44642 }, { "epoch": 3.498667711598746, "grad_norm": 3.0608155727386475, "learning_rate": 3.840911295922611e-07, "loss": 1.3147, "step": 44643 }, { "epoch": 3.4987460815047022, "grad_norm": 4.409737586975098, "learning_rate": 3.839726152056017e-07, "loss": 1.0795, "step": 44644 }, { "epoch": 3.498824451410658, "grad_norm": 2.773991823196411, "learning_rate": 3.838541183759409e-07, "loss": 1.197, "step": 44645 }, { "epoch": 3.4989028213166145, "grad_norm": 3.112642288208008, "learning_rate": 3.837356391037267e-07, "loss": 1.6295, "step": 44646 }, { "epoch": 3.4989811912225703, "grad_norm": 6.180700778961182, "learning_rate": 3.8361717738940986e-07, "loss": 1.7658, "step": 44647 }, { "epoch": 3.4990595611285267, "grad_norm": 2.8715193271636963, "learning_rate": 3.834987332334433e-07, "loss": 1.2511, "step": 44648 }, { "epoch": 3.4991379310344826, "grad_norm": 3.133089303970337, "learning_rate": 3.8338030663627737e-07, "loss": 1.2512, "step": 44649 }, { "epoch": 3.499216300940439, "grad_norm": 3.690768241882324, "learning_rate": 3.8326189759836097e-07, "loss": 1.6589, "step": 44650 }, { "epoch": 3.4992946708463952, "grad_norm": 4.002934455871582, "learning_rate": 3.831435061201444e-07, "loss": 1.1731, "step": 44651 }, { "epoch": 3.499373040752351, "grad_norm": 2.626511335372925, "learning_rate": 3.830251322020789e-07, "loss": 1.1044, "step": 44652 }, { "epoch": 3.499451410658307, "grad_norm": 3.015646457672119, "learning_rate": 3.829067758446142e-07, "loss": 1.2081, "step": 44653 }, { "epoch": 3.4995297805642633, "grad_norm": 6.68888521194458, "learning_rate": 3.8278843704820046e-07, "loss": 1.299, "step": 44654 }, { "epoch": 3.4996081504702197, "grad_norm": 4.915459632873535, "learning_rate": 3.826701158132884e-07, "loss": 1.4798, "step": 44655 }, { "epoch": 3.4996865203761756, "grad_norm": 2.6101996898651123, "learning_rate": 3.825518121403277e-07, "loss": 1.2885, "step": 44656 }, { "epoch": 3.4997648902821314, "grad_norm": 5.179037570953369, "learning_rate": 3.824335260297679e-07, "loss": 0.4498, "step": 44657 }, { "epoch": 3.499843260188088, "grad_norm": 14.028088569641113, "learning_rate": 3.823152574820588e-07, "loss": 1.1956, "step": 44658 }, { "epoch": 3.499921630094044, "grad_norm": 2.488110065460205, "learning_rate": 3.8219700649765156e-07, "loss": 1.1345, "step": 44659 }, { "epoch": 3.5, "grad_norm": 2.867997407913208, "learning_rate": 3.820787730769954e-07, "loss": 1.4019, "step": 44660 }, { "epoch": 3.5, "eval_loss": 1.2451926469802856, "eval_runtime": 110.0278, "eval_samples_per_second": 24.421, "eval_steps_per_second": 6.108, "step": 44660 }, { "epoch": 3.500078369905956, "grad_norm": 4.094343185424805, "learning_rate": 3.819605572205382e-07, "loss": 1.1802, "step": 44661 }, { "epoch": 3.500156739811912, "grad_norm": 6.02619743347168, "learning_rate": 3.8184235892873125e-07, "loss": 1.1568, "step": 44662 }, { "epoch": 3.5002351097178686, "grad_norm": 2.7336630821228027, "learning_rate": 3.817241782020248e-07, "loss": 1.7041, "step": 44663 }, { "epoch": 3.5003134796238244, "grad_norm": 4.073590278625488, "learning_rate": 3.816060150408679e-07, "loss": 1.2712, "step": 44664 }, { "epoch": 3.5003918495297803, "grad_norm": 4.987733840942383, "learning_rate": 3.814878694457086e-07, "loss": 0.9179, "step": 44665 }, { "epoch": 3.5004702194357367, "grad_norm": 1.887874960899353, "learning_rate": 3.81369741416997e-07, "loss": 0.8861, "step": 44666 }, { "epoch": 3.500548589341693, "grad_norm": 4.08311653137207, "learning_rate": 3.812516309551839e-07, "loss": 1.1381, "step": 44667 }, { "epoch": 3.500626959247649, "grad_norm": 2.652073383331299, "learning_rate": 3.8113353806071616e-07, "loss": 1.2603, "step": 44668 }, { "epoch": 3.5007053291536048, "grad_norm": 5.897490501403809, "learning_rate": 3.81015462734044e-07, "loss": 1.5555, "step": 44669 }, { "epoch": 3.500783699059561, "grad_norm": 7.109838485717773, "learning_rate": 3.80897404975617e-07, "loss": 1.7764, "step": 44670 }, { "epoch": 3.5008620689655174, "grad_norm": 3.3480918407440186, "learning_rate": 3.8077936478588375e-07, "loss": 1.4135, "step": 44671 }, { "epoch": 3.5009404388714733, "grad_norm": 3.005091667175293, "learning_rate": 3.8066134216529326e-07, "loss": 1.215, "step": 44672 }, { "epoch": 3.501018808777429, "grad_norm": 2.8846585750579834, "learning_rate": 3.8054333711429415e-07, "loss": 1.3855, "step": 44673 }, { "epoch": 3.5010971786833855, "grad_norm": 4.506755352020264, "learning_rate": 3.804253496333354e-07, "loss": 1.4685, "step": 44674 }, { "epoch": 3.501175548589342, "grad_norm": 3.083998680114746, "learning_rate": 3.8030737972286566e-07, "loss": 1.5165, "step": 44675 }, { "epoch": 3.5012539184952978, "grad_norm": 3.007051706314087, "learning_rate": 3.801894273833345e-07, "loss": 1.4074, "step": 44676 }, { "epoch": 3.5013322884012537, "grad_norm": 3.7825887203216553, "learning_rate": 3.800714926151894e-07, "loss": 1.0269, "step": 44677 }, { "epoch": 3.50141065830721, "grad_norm": 3.112339973449707, "learning_rate": 3.7995357541887935e-07, "loss": 1.2902, "step": 44678 }, { "epoch": 3.5014890282131663, "grad_norm": 14.114261627197266, "learning_rate": 3.7983567579485405e-07, "loss": 0.8548, "step": 44679 }, { "epoch": 3.501567398119122, "grad_norm": 3.4729695320129395, "learning_rate": 3.7971779374355866e-07, "loss": 1.2099, "step": 44680 }, { "epoch": 3.501645768025078, "grad_norm": 3.868474006652832, "learning_rate": 3.795999292654451e-07, "loss": 1.2019, "step": 44681 }, { "epoch": 3.5017241379310344, "grad_norm": 3.572537422180176, "learning_rate": 3.7948208236096075e-07, "loss": 0.9178, "step": 44682 }, { "epoch": 3.5018025078369908, "grad_norm": 3.979989767074585, "learning_rate": 3.793642530305525e-07, "loss": 1.2531, "step": 44683 }, { "epoch": 3.5018808777429467, "grad_norm": 4.4600114822387695, "learning_rate": 3.792464412746688e-07, "loss": 1.6913, "step": 44684 }, { "epoch": 3.501959247648903, "grad_norm": 4.663455963134766, "learning_rate": 3.791286470937583e-07, "loss": 1.5135, "step": 44685 }, { "epoch": 3.502037617554859, "grad_norm": 2.9941203594207764, "learning_rate": 3.7901087048827e-07, "loss": 1.7621, "step": 44686 }, { "epoch": 3.502115987460815, "grad_norm": 3.1655871868133545, "learning_rate": 3.7889311145865027e-07, "loss": 1.5782, "step": 44687 }, { "epoch": 3.502194357366771, "grad_norm": 2.7254083156585693, "learning_rate": 3.787753700053476e-07, "loss": 1.3263, "step": 44688 }, { "epoch": 3.5022727272727274, "grad_norm": 3.0064034461975098, "learning_rate": 3.7865764612880996e-07, "loss": 0.649, "step": 44689 }, { "epoch": 3.5023510971786833, "grad_norm": 3.838021993637085, "learning_rate": 3.785399398294848e-07, "loss": 1.4032, "step": 44690 }, { "epoch": 3.5024294670846396, "grad_norm": 3.354059934616089, "learning_rate": 3.7842225110782e-07, "loss": 1.5168, "step": 44691 }, { "epoch": 3.5025078369905955, "grad_norm": 2.8368704319000244, "learning_rate": 3.783045799642632e-07, "loss": 1.6882, "step": 44692 }, { "epoch": 3.502586206896552, "grad_norm": 2.9473748207092285, "learning_rate": 3.7818692639926215e-07, "loss": 1.5419, "step": 44693 }, { "epoch": 3.5026645768025078, "grad_norm": 3.954660177230835, "learning_rate": 3.780692904132638e-07, "loss": 1.3773, "step": 44694 }, { "epoch": 3.502742946708464, "grad_norm": 2.730849266052246, "learning_rate": 3.7795167200671554e-07, "loss": 1.3596, "step": 44695 }, { "epoch": 3.50282131661442, "grad_norm": 4.263368606567383, "learning_rate": 3.77834071180066e-07, "loss": 1.7786, "step": 44696 }, { "epoch": 3.5028996865203763, "grad_norm": 4.233678340911865, "learning_rate": 3.777164879337619e-07, "loss": 1.1857, "step": 44697 }, { "epoch": 3.502978056426332, "grad_norm": 4.676933288574219, "learning_rate": 3.7759892226824914e-07, "loss": 1.2258, "step": 44698 }, { "epoch": 3.5030564263322885, "grad_norm": 4.491218566894531, "learning_rate": 3.77481374183975e-07, "loss": 1.4514, "step": 44699 }, { "epoch": 3.5031347962382444, "grad_norm": 8.701326370239258, "learning_rate": 3.7736384368138867e-07, "loss": 1.4622, "step": 44700 }, { "epoch": 3.5032131661442008, "grad_norm": 2.8583271503448486, "learning_rate": 3.7724633076093644e-07, "loss": 1.3628, "step": 44701 }, { "epoch": 3.5032915360501566, "grad_norm": 2.506798028945923, "learning_rate": 3.77128835423064e-07, "loss": 1.3309, "step": 44702 }, { "epoch": 3.503369905956113, "grad_norm": 2.458258867263794, "learning_rate": 3.770113576682194e-07, "loss": 1.3053, "step": 44703 }, { "epoch": 3.503448275862069, "grad_norm": 3.8006505966186523, "learning_rate": 3.768938974968478e-07, "loss": 1.3427, "step": 44704 }, { "epoch": 3.503526645768025, "grad_norm": 3.34572172164917, "learning_rate": 3.7677645490939884e-07, "loss": 1.6149, "step": 44705 }, { "epoch": 3.503605015673981, "grad_norm": 3.1103832721710205, "learning_rate": 3.766590299063172e-07, "loss": 1.3476, "step": 44706 }, { "epoch": 3.5036833855799374, "grad_norm": 5.234623908996582, "learning_rate": 3.765416224880497e-07, "loss": 1.3198, "step": 44707 }, { "epoch": 3.5037617554858933, "grad_norm": 3.062509298324585, "learning_rate": 3.764242326550427e-07, "loss": 1.3171, "step": 44708 }, { "epoch": 3.5038401253918496, "grad_norm": 3.4369757175445557, "learning_rate": 3.7630686040774356e-07, "loss": 1.0611, "step": 44709 }, { "epoch": 3.5039184952978055, "grad_norm": 2.1851484775543213, "learning_rate": 3.7618950574659807e-07, "loss": 1.3598, "step": 44710 }, { "epoch": 3.503996865203762, "grad_norm": 5.491819858551025, "learning_rate": 3.760721686720531e-07, "loss": 2.184, "step": 44711 }, { "epoch": 3.5040752351097177, "grad_norm": 3.3707196712493896, "learning_rate": 3.759548491845544e-07, "loss": 1.514, "step": 44712 }, { "epoch": 3.504153605015674, "grad_norm": 3.6990914344787598, "learning_rate": 3.7583754728454823e-07, "loss": 1.3461, "step": 44713 }, { "epoch": 3.50423197492163, "grad_norm": 13.719939231872559, "learning_rate": 3.75720262972481e-07, "loss": 0.7497, "step": 44714 }, { "epoch": 3.5043103448275863, "grad_norm": 2.4456088542938232, "learning_rate": 3.756029962487989e-07, "loss": 1.1573, "step": 44715 }, { "epoch": 3.504388714733542, "grad_norm": 2.9100852012634277, "learning_rate": 3.7548574711394834e-07, "loss": 1.147, "step": 44716 }, { "epoch": 3.5044670846394985, "grad_norm": 3.407533645629883, "learning_rate": 3.7536851556837395e-07, "loss": 1.3854, "step": 44717 }, { "epoch": 3.5045454545454544, "grad_norm": 2.779534339904785, "learning_rate": 3.7525130161252143e-07, "loss": 1.2519, "step": 44718 }, { "epoch": 3.5046238244514107, "grad_norm": 6.938882350921631, "learning_rate": 3.751341052468382e-07, "loss": 1.7015, "step": 44719 }, { "epoch": 3.5047021943573666, "grad_norm": 3.191718101501465, "learning_rate": 3.7501692647177067e-07, "loss": 1.3789, "step": 44720 }, { "epoch": 3.504780564263323, "grad_norm": 7.083643436431885, "learning_rate": 3.748997652877623e-07, "loss": 1.5014, "step": 44721 }, { "epoch": 3.504858934169279, "grad_norm": 8.560836791992188, "learning_rate": 3.7478262169525936e-07, "loss": 1.2037, "step": 44722 }, { "epoch": 3.504937304075235, "grad_norm": 3.4222307205200195, "learning_rate": 3.7466549569470765e-07, "loss": 1.78, "step": 44723 }, { "epoch": 3.505015673981191, "grad_norm": 4.2984232902526855, "learning_rate": 3.7454838728655294e-07, "loss": 1.6222, "step": 44724 }, { "epoch": 3.5050940438871474, "grad_norm": 3.753229856491089, "learning_rate": 3.744312964712404e-07, "loss": 1.2381, "step": 44725 }, { "epoch": 3.5051724137931033, "grad_norm": 3.4704151153564453, "learning_rate": 3.743142232492153e-07, "loss": 1.5085, "step": 44726 }, { "epoch": 3.5052507836990596, "grad_norm": 2.575658082962036, "learning_rate": 3.7419716762092274e-07, "loss": 1.3669, "step": 44727 }, { "epoch": 3.5053291536050155, "grad_norm": 3.480421543121338, "learning_rate": 3.740801295868085e-07, "loss": 1.3445, "step": 44728 }, { "epoch": 3.505407523510972, "grad_norm": 4.058382511138916, "learning_rate": 3.7396310914731724e-07, "loss": 1.2885, "step": 44729 }, { "epoch": 3.5054858934169277, "grad_norm": 1.9547420740127563, "learning_rate": 3.738461063028942e-07, "loss": 1.3103, "step": 44730 }, { "epoch": 3.505564263322884, "grad_norm": 4.097104072570801, "learning_rate": 3.737291210539845e-07, "loss": 1.3868, "step": 44731 }, { "epoch": 3.50564263322884, "grad_norm": 2.331820011138916, "learning_rate": 3.736121534010334e-07, "loss": 1.2031, "step": 44732 }, { "epoch": 3.5057210031347963, "grad_norm": 2.1115400791168213, "learning_rate": 3.73495203344485e-07, "loss": 1.2285, "step": 44733 }, { "epoch": 3.5057993730407526, "grad_norm": 4.393328666687012, "learning_rate": 3.733782708847844e-07, "loss": 1.3122, "step": 44734 }, { "epoch": 3.5058777429467085, "grad_norm": 2.9840314388275146, "learning_rate": 3.7326135602237744e-07, "loss": 1.4378, "step": 44735 }, { "epoch": 3.5059561128526644, "grad_norm": 4.095124244689941, "learning_rate": 3.73144458757706e-07, "loss": 1.3939, "step": 44736 }, { "epoch": 3.5060344827586207, "grad_norm": 12.129833221435547, "learning_rate": 3.73027579091218e-07, "loss": 0.9466, "step": 44737 }, { "epoch": 3.506112852664577, "grad_norm": 3.935795545578003, "learning_rate": 3.72910717023357e-07, "loss": 1.3526, "step": 44738 }, { "epoch": 3.506191222570533, "grad_norm": 4.21826171875, "learning_rate": 3.7279387255456603e-07, "loss": 1.524, "step": 44739 }, { "epoch": 3.506269592476489, "grad_norm": 10.397575378417969, "learning_rate": 3.7267704568529015e-07, "loss": 0.3053, "step": 44740 }, { "epoch": 3.506347962382445, "grad_norm": 2.55047869682312, "learning_rate": 3.725602364159736e-07, "loss": 1.1099, "step": 44741 }, { "epoch": 3.5064263322884015, "grad_norm": 3.0916340351104736, "learning_rate": 3.7244344474706317e-07, "loss": 0.7871, "step": 44742 }, { "epoch": 3.5065047021943574, "grad_norm": 3.516481876373291, "learning_rate": 3.7232667067899965e-07, "loss": 1.6687, "step": 44743 }, { "epoch": 3.5065830721003133, "grad_norm": 6.254789352416992, "learning_rate": 3.7220991421222874e-07, "loss": 1.2974, "step": 44744 }, { "epoch": 3.5066614420062696, "grad_norm": 2.5609495639801025, "learning_rate": 3.720931753471946e-07, "loss": 1.3053, "step": 44745 }, { "epoch": 3.506739811912226, "grad_norm": 2.601895809173584, "learning_rate": 3.719764540843407e-07, "loss": 1.0487, "step": 44746 }, { "epoch": 3.506818181818182, "grad_norm": 3.5617568492889404, "learning_rate": 3.7185975042411113e-07, "loss": 1.4812, "step": 44747 }, { "epoch": 3.5068965517241377, "grad_norm": 3.3836488723754883, "learning_rate": 3.717430643669501e-07, "loss": 1.514, "step": 44748 }, { "epoch": 3.506974921630094, "grad_norm": 8.088687896728516, "learning_rate": 3.7162639591330094e-07, "loss": 1.5814, "step": 44749 }, { "epoch": 3.5070532915360504, "grad_norm": 4.305602550506592, "learning_rate": 3.7150974506360794e-07, "loss": 1.0252, "step": 44750 }, { "epoch": 3.5071316614420063, "grad_norm": 3.6917836666107178, "learning_rate": 3.713931118183145e-07, "loss": 1.4313, "step": 44751 }, { "epoch": 3.507210031347962, "grad_norm": 5.074151515960693, "learning_rate": 3.7127649617786423e-07, "loss": 1.4075, "step": 44752 }, { "epoch": 3.5072884012539185, "grad_norm": 12.09631633758545, "learning_rate": 3.711598981427017e-07, "loss": 1.3301, "step": 44753 }, { "epoch": 3.507366771159875, "grad_norm": 7.452512264251709, "learning_rate": 3.7104331771326827e-07, "loss": 1.3715, "step": 44754 }, { "epoch": 3.5074451410658307, "grad_norm": 3.378504514694214, "learning_rate": 3.709267548900075e-07, "loss": 1.4016, "step": 44755 }, { "epoch": 3.5075235109717866, "grad_norm": 4.261514186859131, "learning_rate": 3.7081020967336513e-07, "loss": 1.4752, "step": 44756 }, { "epoch": 3.507601880877743, "grad_norm": 2.6901772022247314, "learning_rate": 3.7069368206378355e-07, "loss": 1.5595, "step": 44757 }, { "epoch": 3.5076802507836993, "grad_norm": 3.098737955093384, "learning_rate": 3.7057717206170464e-07, "loss": 1.3799, "step": 44758 }, { "epoch": 3.507758620689655, "grad_norm": 6.167295932769775, "learning_rate": 3.7046067966757247e-07, "loss": 1.3317, "step": 44759 }, { "epoch": 3.507836990595611, "grad_norm": 3.9384851455688477, "learning_rate": 3.70344204881829e-07, "loss": 1.3106, "step": 44760 }, { "epoch": 3.5079153605015674, "grad_norm": 2.393388271331787, "learning_rate": 3.7022774770491986e-07, "loss": 1.4083, "step": 44761 }, { "epoch": 3.5079937304075237, "grad_norm": 2.86669921875, "learning_rate": 3.7011130813728536e-07, "loss": 1.5235, "step": 44762 }, { "epoch": 3.5080721003134796, "grad_norm": 3.9125595092773438, "learning_rate": 3.6999488617936953e-07, "loss": 1.7056, "step": 44763 }, { "epoch": 3.5081504702194355, "grad_norm": 2.7947072982788086, "learning_rate": 3.6987848183161533e-07, "loss": 1.1426, "step": 44764 }, { "epoch": 3.508228840125392, "grad_norm": 2.7165591716766357, "learning_rate": 3.6976209509446527e-07, "loss": 1.5266, "step": 44765 }, { "epoch": 3.508307210031348, "grad_norm": 4.55197811126709, "learning_rate": 3.6964572596836166e-07, "loss": 1.1658, "step": 44766 }, { "epoch": 3.508385579937304, "grad_norm": 3.9189341068267822, "learning_rate": 3.6952937445374703e-07, "loss": 1.2532, "step": 44767 }, { "epoch": 3.50846394984326, "grad_norm": 4.20259952545166, "learning_rate": 3.6941304055106543e-07, "loss": 1.4874, "step": 44768 }, { "epoch": 3.5085423197492163, "grad_norm": 5.358364582061768, "learning_rate": 3.69296724260757e-07, "loss": 1.4851, "step": 44769 }, { "epoch": 3.5086206896551726, "grad_norm": 4.099673748016357, "learning_rate": 3.6918042558326597e-07, "loss": 1.6089, "step": 44770 }, { "epoch": 3.5086990595611285, "grad_norm": 4.436187744140625, "learning_rate": 3.690641445190335e-07, "loss": 1.8494, "step": 44771 }, { "epoch": 3.5087774294670844, "grad_norm": 4.232778549194336, "learning_rate": 3.6894788106850376e-07, "loss": 1.4859, "step": 44772 }, { "epoch": 3.5088557993730407, "grad_norm": 6.382140636444092, "learning_rate": 3.6883163523211694e-07, "loss": 1.1582, "step": 44773 }, { "epoch": 3.508934169278997, "grad_norm": 4.452763557434082, "learning_rate": 3.6871540701031437e-07, "loss": 0.9811, "step": 44774 }, { "epoch": 3.509012539184953, "grad_norm": 2.494091272354126, "learning_rate": 3.6859919640354127e-07, "loss": 1.0987, "step": 44775 }, { "epoch": 3.509090909090909, "grad_norm": 2.4813175201416016, "learning_rate": 3.684830034122383e-07, "loss": 1.1635, "step": 44776 }, { "epoch": 3.509169278996865, "grad_norm": 4.156776428222656, "learning_rate": 3.683668280368463e-07, "loss": 1.5207, "step": 44777 }, { "epoch": 3.5092476489028215, "grad_norm": 5.95125150680542, "learning_rate": 3.6825067027780825e-07, "loss": 1.2334, "step": 44778 }, { "epoch": 3.5093260188087774, "grad_norm": 5.641439437866211, "learning_rate": 3.681345301355654e-07, "loss": 1.479, "step": 44779 }, { "epoch": 3.5094043887147337, "grad_norm": 2.901510238647461, "learning_rate": 3.6801840761055975e-07, "loss": 1.2405, "step": 44780 }, { "epoch": 3.5094827586206896, "grad_norm": 5.137747287750244, "learning_rate": 3.6790230270323247e-07, "loss": 0.9917, "step": 44781 }, { "epoch": 3.509561128526646, "grad_norm": 7.317625045776367, "learning_rate": 3.67786215414026e-07, "loss": 0.9552, "step": 44782 }, { "epoch": 3.509639498432602, "grad_norm": 8.226799964904785, "learning_rate": 3.676701457433818e-07, "loss": 1.3739, "step": 44783 }, { "epoch": 3.509717868338558, "grad_norm": 3.95093035697937, "learning_rate": 3.67554093691741e-07, "loss": 1.1229, "step": 44784 }, { "epoch": 3.509796238244514, "grad_norm": 3.491581439971924, "learning_rate": 3.6743805925954446e-07, "loss": 1.5886, "step": 44785 }, { "epoch": 3.5098746081504704, "grad_norm": 4.360413074493408, "learning_rate": 3.673220424472346e-07, "loss": 1.4763, "step": 44786 }, { "epoch": 3.5099529780564263, "grad_norm": 6.960367679595947, "learning_rate": 3.6720604325525323e-07, "loss": 1.4314, "step": 44787 }, { "epoch": 3.5100313479623826, "grad_norm": 3.664055824279785, "learning_rate": 3.670900616840384e-07, "loss": 1.2644, "step": 44788 }, { "epoch": 3.5101097178683385, "grad_norm": 2.9760982990264893, "learning_rate": 3.6697409773403467e-07, "loss": 1.2362, "step": 44789 }, { "epoch": 3.510188087774295, "grad_norm": 2.843684434890747, "learning_rate": 3.6685815140568125e-07, "loss": 1.4199, "step": 44790 }, { "epoch": 3.5102664576802507, "grad_norm": 3.350966691970825, "learning_rate": 3.6674222269942107e-07, "loss": 1.1585, "step": 44791 }, { "epoch": 3.510344827586207, "grad_norm": 3.3123083114624023, "learning_rate": 3.6662631161569206e-07, "loss": 1.0649, "step": 44792 }, { "epoch": 3.510423197492163, "grad_norm": 2.4324612617492676, "learning_rate": 3.665104181549367e-07, "loss": 1.1977, "step": 44793 }, { "epoch": 3.5105015673981192, "grad_norm": 3.811809539794922, "learning_rate": 3.6639454231759686e-07, "loss": 1.2555, "step": 44794 }, { "epoch": 3.510579937304075, "grad_norm": 3.4896693229675293, "learning_rate": 3.662786841041116e-07, "loss": 1.4572, "step": 44795 }, { "epoch": 3.5106583072100315, "grad_norm": 2.5798022747039795, "learning_rate": 3.6616284351492227e-07, "loss": 1.4408, "step": 44796 }, { "epoch": 3.5107366771159874, "grad_norm": 2.6258866786956787, "learning_rate": 3.660470205504696e-07, "loss": 0.9074, "step": 44797 }, { "epoch": 3.5108150470219437, "grad_norm": 4.602601528167725, "learning_rate": 3.659312152111938e-07, "loss": 0.7362, "step": 44798 }, { "epoch": 3.5108934169278996, "grad_norm": 2.032914638519287, "learning_rate": 3.658154274975351e-07, "loss": 1.3752, "step": 44799 }, { "epoch": 3.510971786833856, "grad_norm": 3.0230536460876465, "learning_rate": 3.6569965740993475e-07, "loss": 1.4008, "step": 44800 }, { "epoch": 3.511050156739812, "grad_norm": 3.6608614921569824, "learning_rate": 3.655839049488319e-07, "loss": 1.4789, "step": 44801 }, { "epoch": 3.511128526645768, "grad_norm": 4.066905498504639, "learning_rate": 3.654681701146684e-07, "loss": 1.2901, "step": 44802 }, { "epoch": 3.511206896551724, "grad_norm": 3.191617488861084, "learning_rate": 3.653524529078828e-07, "loss": 1.0934, "step": 44803 }, { "epoch": 3.5112852664576804, "grad_norm": 8.805214881896973, "learning_rate": 3.652367533289164e-07, "loss": 1.2322, "step": 44804 }, { "epoch": 3.5113636363636362, "grad_norm": 2.3895390033721924, "learning_rate": 3.651210713782083e-07, "loss": 1.3172, "step": 44805 }, { "epoch": 3.5114420062695926, "grad_norm": 2.884521245956421, "learning_rate": 3.6500540705620034e-07, "loss": 1.3227, "step": 44806 }, { "epoch": 3.5115203761755485, "grad_norm": 5.1156792640686035, "learning_rate": 3.6488976036332945e-07, "loss": 0.9644, "step": 44807 }, { "epoch": 3.511598746081505, "grad_norm": 4.77121114730835, "learning_rate": 3.6477413130003745e-07, "loss": 1.2706, "step": 44808 }, { "epoch": 3.5116771159874607, "grad_norm": 11.518176078796387, "learning_rate": 3.646585198667652e-07, "loss": 0.4344, "step": 44809 }, { "epoch": 3.511755485893417, "grad_norm": 7.151540756225586, "learning_rate": 3.645429260639505e-07, "loss": 1.213, "step": 44810 }, { "epoch": 3.511833855799373, "grad_norm": 2.949662446975708, "learning_rate": 3.644273498920331e-07, "loss": 1.014, "step": 44811 }, { "epoch": 3.5119122257053292, "grad_norm": 9.76579475402832, "learning_rate": 3.643117913514521e-07, "loss": 1.3464, "step": 44812 }, { "epoch": 3.511990595611285, "grad_norm": 3.4090049266815186, "learning_rate": 3.6419625044264994e-07, "loss": 1.7367, "step": 44813 }, { "epoch": 3.5120689655172415, "grad_norm": 3.0005476474761963, "learning_rate": 3.6408072716606346e-07, "loss": 1.2887, "step": 44814 }, { "epoch": 3.5121473354231973, "grad_norm": 2.846602439880371, "learning_rate": 3.639652215221323e-07, "loss": 1.035, "step": 44815 }, { "epoch": 3.5122257053291537, "grad_norm": 6.374341011047363, "learning_rate": 3.638497335112967e-07, "loss": 1.4082, "step": 44816 }, { "epoch": 3.5123040752351096, "grad_norm": 3.1131160259246826, "learning_rate": 3.6373426313399515e-07, "loss": 1.371, "step": 44817 }, { "epoch": 3.512382445141066, "grad_norm": 3.811584949493408, "learning_rate": 3.6361881039066737e-07, "loss": 1.3778, "step": 44818 }, { "epoch": 3.512460815047022, "grad_norm": 4.468415260314941, "learning_rate": 3.6350337528175186e-07, "loss": 1.5171, "step": 44819 }, { "epoch": 3.512539184952978, "grad_norm": 2.61037015914917, "learning_rate": 3.6338795780768877e-07, "loss": 1.107, "step": 44820 }, { "epoch": 3.512617554858934, "grad_norm": 3.1536147594451904, "learning_rate": 3.632725579689156e-07, "loss": 1.6937, "step": 44821 }, { "epoch": 3.5126959247648903, "grad_norm": 3.607821226119995, "learning_rate": 3.6315717576587253e-07, "loss": 1.2945, "step": 44822 }, { "epoch": 3.5127742946708462, "grad_norm": 10.087181091308594, "learning_rate": 3.6304181119899753e-07, "loss": 1.461, "step": 44823 }, { "epoch": 3.5128526645768026, "grad_norm": 3.8476202487945557, "learning_rate": 3.629264642687308e-07, "loss": 1.4129, "step": 44824 }, { "epoch": 3.512931034482759, "grad_norm": 2.1533217430114746, "learning_rate": 3.6281113497550867e-07, "loss": 1.1508, "step": 44825 }, { "epoch": 3.513009404388715, "grad_norm": 3.0075864791870117, "learning_rate": 3.626958233197714e-07, "loss": 1.1964, "step": 44826 }, { "epoch": 3.5130877742946707, "grad_norm": 4.073448657989502, "learning_rate": 3.6258052930195795e-07, "loss": 0.8435, "step": 44827 }, { "epoch": 3.513166144200627, "grad_norm": 10.371474266052246, "learning_rate": 3.6246525292250647e-07, "loss": 0.7228, "step": 44828 }, { "epoch": 3.5132445141065833, "grad_norm": 2.578007698059082, "learning_rate": 3.623499941818548e-07, "loss": 1.1409, "step": 44829 }, { "epoch": 3.5133228840125392, "grad_norm": 2.807673454284668, "learning_rate": 3.622347530804415e-07, "loss": 1.414, "step": 44830 }, { "epoch": 3.513401253918495, "grad_norm": 3.620293378829956, "learning_rate": 3.621195296187047e-07, "loss": 1.1161, "step": 44831 }, { "epoch": 3.5134796238244514, "grad_norm": 3.7730424404144287, "learning_rate": 3.6200432379708436e-07, "loss": 0.9858, "step": 44832 }, { "epoch": 3.513557993730408, "grad_norm": 5.8694047927856445, "learning_rate": 3.6188913561601647e-07, "loss": 1.0978, "step": 44833 }, { "epoch": 3.5136363636363637, "grad_norm": 2.726060390472412, "learning_rate": 3.6177396507593997e-07, "loss": 1.3354, "step": 44834 }, { "epoch": 3.5137147335423196, "grad_norm": 3.6806366443634033, "learning_rate": 3.6165881217729295e-07, "loss": 1.2719, "step": 44835 }, { "epoch": 3.513793103448276, "grad_norm": 2.8892812728881836, "learning_rate": 3.6154367692051384e-07, "loss": 1.2956, "step": 44836 }, { "epoch": 3.5138714733542322, "grad_norm": 2.9672892093658447, "learning_rate": 3.614285593060396e-07, "loss": 1.3435, "step": 44837 }, { "epoch": 3.513949843260188, "grad_norm": 3.5637381076812744, "learning_rate": 3.613134593343087e-07, "loss": 1.1849, "step": 44838 }, { "epoch": 3.514028213166144, "grad_norm": 5.241381645202637, "learning_rate": 3.611983770057592e-07, "loss": 1.7488, "step": 44839 }, { "epoch": 3.5141065830721003, "grad_norm": 5.130593299865723, "learning_rate": 3.610833123208279e-07, "loss": 1.6121, "step": 44840 }, { "epoch": 3.5141849529780567, "grad_norm": 2.9460480213165283, "learning_rate": 3.6096826527995343e-07, "loss": 1.6645, "step": 44841 }, { "epoch": 3.5142633228840126, "grad_norm": 2.2133052349090576, "learning_rate": 3.6085323588357256e-07, "loss": 1.4208, "step": 44842 }, { "epoch": 3.5143416927899684, "grad_norm": 4.905781269073486, "learning_rate": 3.607382241321239e-07, "loss": 1.6275, "step": 44843 }, { "epoch": 3.5144200626959248, "grad_norm": 2.839012384414673, "learning_rate": 3.6062323002604317e-07, "loss": 1.0778, "step": 44844 }, { "epoch": 3.514498432601881, "grad_norm": 5.202126502990723, "learning_rate": 3.605082535657689e-07, "loss": 1.6117, "step": 44845 }, { "epoch": 3.514576802507837, "grad_norm": 3.730768918991089, "learning_rate": 3.6039329475173856e-07, "loss": 1.5722, "step": 44846 }, { "epoch": 3.514655172413793, "grad_norm": 3.9758849143981934, "learning_rate": 3.602783535843901e-07, "loss": 1.0265, "step": 44847 }, { "epoch": 3.514733542319749, "grad_norm": 2.92995285987854, "learning_rate": 3.601634300641582e-07, "loss": 1.4799, "step": 44848 }, { "epoch": 3.5148119122257055, "grad_norm": 4.189634799957275, "learning_rate": 3.6004852419148137e-07, "loss": 1.2391, "step": 44849 }, { "epoch": 3.5148902821316614, "grad_norm": 7.623475074768066, "learning_rate": 3.5993363596679766e-07, "loss": 1.4896, "step": 44850 }, { "epoch": 3.5149686520376173, "grad_norm": 3.4404168128967285, "learning_rate": 3.598187653905427e-07, "loss": 1.7742, "step": 44851 }, { "epoch": 3.5150470219435737, "grad_norm": 2.418107748031616, "learning_rate": 3.5970391246315404e-07, "loss": 0.9711, "step": 44852 }, { "epoch": 3.51512539184953, "grad_norm": 2.5894837379455566, "learning_rate": 3.5958907718506796e-07, "loss": 1.3023, "step": 44853 }, { "epoch": 3.515203761755486, "grad_norm": 3.553788423538208, "learning_rate": 3.594742595567213e-07, "loss": 1.3159, "step": 44854 }, { "epoch": 3.5152821316614418, "grad_norm": 2.591592788696289, "learning_rate": 3.5935945957855157e-07, "loss": 1.3074, "step": 44855 }, { "epoch": 3.515360501567398, "grad_norm": 2.7382290363311768, "learning_rate": 3.5924467725099445e-07, "loss": 1.1861, "step": 44856 }, { "epoch": 3.5154388714733544, "grad_norm": 4.356534004211426, "learning_rate": 3.5912991257448735e-07, "loss": 1.6118, "step": 44857 }, { "epoch": 3.5155172413793103, "grad_norm": 2.8037071228027344, "learning_rate": 3.590151655494661e-07, "loss": 1.5079, "step": 44858 }, { "epoch": 3.515595611285266, "grad_norm": 3.0742084980010986, "learning_rate": 3.5890043617636704e-07, "loss": 1.4078, "step": 44859 }, { "epoch": 3.5156739811912225, "grad_norm": 2.9180638790130615, "learning_rate": 3.5878572445562754e-07, "loss": 1.3673, "step": 44860 }, { "epoch": 3.515752351097179, "grad_norm": 3.033867359161377, "learning_rate": 3.586710303876839e-07, "loss": 1.2818, "step": 44861 }, { "epoch": 3.5158307210031348, "grad_norm": 2.628196954727173, "learning_rate": 3.585563539729708e-07, "loss": 1.2644, "step": 44862 }, { "epoch": 3.5159090909090907, "grad_norm": 2.9775049686431885, "learning_rate": 3.584416952119246e-07, "loss": 1.1756, "step": 44863 }, { "epoch": 3.515987460815047, "grad_norm": 3.979396104812622, "learning_rate": 3.583270541049827e-07, "loss": 1.5775, "step": 44864 }, { "epoch": 3.5160658307210033, "grad_norm": 4.354123115539551, "learning_rate": 3.582124306525819e-07, "loss": 1.275, "step": 44865 }, { "epoch": 3.516144200626959, "grad_norm": 2.870206117630005, "learning_rate": 3.580978248551553e-07, "loss": 1.4269, "step": 44866 }, { "epoch": 3.516222570532915, "grad_norm": 2.3360326290130615, "learning_rate": 3.5798323671314083e-07, "loss": 1.3162, "step": 44867 }, { "epoch": 3.5163009404388714, "grad_norm": 4.814781188964844, "learning_rate": 3.578686662269731e-07, "loss": 1.9019, "step": 44868 }, { "epoch": 3.5163793103448278, "grad_norm": 2.8861489295959473, "learning_rate": 3.577541133970902e-07, "loss": 1.2839, "step": 44869 }, { "epoch": 3.5164576802507836, "grad_norm": 2.444120407104492, "learning_rate": 3.5763957822392504e-07, "loss": 1.3297, "step": 44870 }, { "epoch": 3.5165360501567395, "grad_norm": 3.985210418701172, "learning_rate": 3.5752506070791506e-07, "loss": 1.6321, "step": 44871 }, { "epoch": 3.516614420062696, "grad_norm": 2.88193941116333, "learning_rate": 3.574105608494949e-07, "loss": 1.5721, "step": 44872 }, { "epoch": 3.516692789968652, "grad_norm": 3.823122978210449, "learning_rate": 3.572960786491003e-07, "loss": 0.8084, "step": 44873 }, { "epoch": 3.516771159874608, "grad_norm": 2.56020450592041, "learning_rate": 3.5718161410716656e-07, "loss": 1.2971, "step": 44874 }, { "epoch": 3.5168495297805644, "grad_norm": 2.7461283206939697, "learning_rate": 3.5706716722412995e-07, "loss": 0.8884, "step": 44875 }, { "epoch": 3.5169278996865203, "grad_norm": 4.704575538635254, "learning_rate": 3.5695273800042564e-07, "loss": 1.2715, "step": 44876 }, { "epoch": 3.5170062695924766, "grad_norm": 4.259923458099365, "learning_rate": 3.5683832643648606e-07, "loss": 1.1807, "step": 44877 }, { "epoch": 3.5170846394984325, "grad_norm": 9.992080688476562, "learning_rate": 3.5672393253274985e-07, "loss": 1.1295, "step": 44878 }, { "epoch": 3.517163009404389, "grad_norm": 4.27957010269165, "learning_rate": 3.56609556289651e-07, "loss": 1.5245, "step": 44879 }, { "epoch": 3.5172413793103448, "grad_norm": 2.4175994396209717, "learning_rate": 3.564951977076253e-07, "loss": 1.2978, "step": 44880 }, { "epoch": 3.517319749216301, "grad_norm": 3.8632972240448, "learning_rate": 3.563808567871058e-07, "loss": 1.4692, "step": 44881 }, { "epoch": 3.517398119122257, "grad_norm": 3.526211738586426, "learning_rate": 3.5626653352852757e-07, "loss": 1.4791, "step": 44882 }, { "epoch": 3.5174764890282133, "grad_norm": 3.3267860412597656, "learning_rate": 3.5615222793232705e-07, "loss": 0.7099, "step": 44883 }, { "epoch": 3.517554858934169, "grad_norm": 8.577744483947754, "learning_rate": 3.560379399989394e-07, "loss": 0.8144, "step": 44884 }, { "epoch": 3.5176332288401255, "grad_norm": 4.097585678100586, "learning_rate": 3.559236697287971e-07, "loss": 1.0124, "step": 44885 }, { "epoch": 3.5177115987460814, "grad_norm": 2.641620397567749, "learning_rate": 3.558094171223353e-07, "loss": 1.1755, "step": 44886 }, { "epoch": 3.5177899686520377, "grad_norm": 3.9750473499298096, "learning_rate": 3.5569518217998866e-07, "loss": 1.3563, "step": 44887 }, { "epoch": 3.5178683385579936, "grad_norm": 3.3271872997283936, "learning_rate": 3.555809649021941e-07, "loss": 1.6041, "step": 44888 }, { "epoch": 3.51794670846395, "grad_norm": 23.3251953125, "learning_rate": 3.5546676528938283e-07, "loss": 1.0016, "step": 44889 }, { "epoch": 3.518025078369906, "grad_norm": 7.198277950286865, "learning_rate": 3.553525833419902e-07, "loss": 1.3741, "step": 44890 }, { "epoch": 3.518103448275862, "grad_norm": 5.076682090759277, "learning_rate": 3.552384190604513e-07, "loss": 1.1556, "step": 44891 }, { "epoch": 3.518181818181818, "grad_norm": 8.106006622314453, "learning_rate": 3.551242724451992e-07, "loss": 1.5913, "step": 44892 }, { "epoch": 3.5182601880877744, "grad_norm": 3.6723082065582275, "learning_rate": 3.550101434966691e-07, "loss": 1.2143, "step": 44893 }, { "epoch": 3.5183385579937303, "grad_norm": 2.2197341918945312, "learning_rate": 3.548960322152939e-07, "loss": 1.3473, "step": 44894 }, { "epoch": 3.5184169278996866, "grad_norm": 4.119584083557129, "learning_rate": 3.5478193860150943e-07, "loss": 1.1568, "step": 44895 }, { "epoch": 3.5184952978056425, "grad_norm": 2.5588364601135254, "learning_rate": 3.5466786265574703e-07, "loss": 1.3561, "step": 44896 }, { "epoch": 3.518573667711599, "grad_norm": 3.3585243225097656, "learning_rate": 3.54553804378443e-07, "loss": 1.4682, "step": 44897 }, { "epoch": 3.5186520376175547, "grad_norm": 11.29590129852295, "learning_rate": 3.544397637700303e-07, "loss": 2.0738, "step": 44898 }, { "epoch": 3.518730407523511, "grad_norm": 4.8592095375061035, "learning_rate": 3.5432574083094306e-07, "loss": 1.5057, "step": 44899 }, { "epoch": 3.518808777429467, "grad_norm": 4.969663143157959, "learning_rate": 3.542117355616143e-07, "loss": 1.3292, "step": 44900 }, { "epoch": 3.5188871473354233, "grad_norm": 3.4698891639709473, "learning_rate": 3.540977479624763e-07, "loss": 1.5277, "step": 44901 }, { "epoch": 3.518965517241379, "grad_norm": 3.1455366611480713, "learning_rate": 3.5398377803396554e-07, "loss": 1.1576, "step": 44902 }, { "epoch": 3.5190438871473355, "grad_norm": 2.639622688293457, "learning_rate": 3.5386982577651495e-07, "loss": 1.0442, "step": 44903 }, { "epoch": 3.5191222570532914, "grad_norm": 3.4415481090545654, "learning_rate": 3.537558911905559e-07, "loss": 1.5885, "step": 44904 }, { "epoch": 3.5192006269592477, "grad_norm": 5.597236156463623, "learning_rate": 3.536419742765235e-07, "loss": 1.4789, "step": 44905 }, { "epoch": 3.5192789968652036, "grad_norm": 2.2961273193359375, "learning_rate": 3.535280750348502e-07, "loss": 1.4682, "step": 44906 }, { "epoch": 3.51935736677116, "grad_norm": 2.6357572078704834, "learning_rate": 3.534141934659696e-07, "loss": 1.1136, "step": 44907 }, { "epoch": 3.519435736677116, "grad_norm": 2.723238229751587, "learning_rate": 3.533003295703147e-07, "loss": 1.5477, "step": 44908 }, { "epoch": 3.519514106583072, "grad_norm": 2.574632167816162, "learning_rate": 3.5318648334831837e-07, "loss": 1.1347, "step": 44909 }, { "epoch": 3.519592476489028, "grad_norm": 2.3890931606292725, "learning_rate": 3.530726548004143e-07, "loss": 1.1252, "step": 44910 }, { "epoch": 3.5196708463949844, "grad_norm": 5.088102340698242, "learning_rate": 3.5295884392703483e-07, "loss": 1.4137, "step": 44911 }, { "epoch": 3.5197492163009403, "grad_norm": 3.8387386798858643, "learning_rate": 3.5284505072861354e-07, "loss": 1.0082, "step": 44912 }, { "epoch": 3.5198275862068966, "grad_norm": 2.9630656242370605, "learning_rate": 3.527312752055823e-07, "loss": 0.9233, "step": 44913 }, { "epoch": 3.5199059561128525, "grad_norm": 3.4529318809509277, "learning_rate": 3.526175173583751e-07, "loss": 1.2934, "step": 44914 }, { "epoch": 3.519984326018809, "grad_norm": 4.0189714431762695, "learning_rate": 3.5250377718742235e-07, "loss": 1.4474, "step": 44915 }, { "epoch": 3.5200626959247647, "grad_norm": 4.507115364074707, "learning_rate": 3.523900546931591e-07, "loss": 0.9041, "step": 44916 }, { "epoch": 3.520141065830721, "grad_norm": 7.2257819175720215, "learning_rate": 3.522763498760173e-07, "loss": 1.8129, "step": 44917 }, { "epoch": 3.520219435736677, "grad_norm": 5.17437219619751, "learning_rate": 3.521626627364283e-07, "loss": 1.0997, "step": 44918 }, { "epoch": 3.5202978056426333, "grad_norm": 4.870340824127197, "learning_rate": 3.5204899327482554e-07, "loss": 1.0754, "step": 44919 }, { "epoch": 3.5203761755485896, "grad_norm": 2.805422782897949, "learning_rate": 3.519353414916404e-07, "loss": 1.3205, "step": 44920 }, { "epoch": 3.5204545454545455, "grad_norm": 3.764353036880493, "learning_rate": 3.51821707387307e-07, "loss": 1.3073, "step": 44921 }, { "epoch": 3.5205329153605014, "grad_norm": 3.796156883239746, "learning_rate": 3.5170809096225555e-07, "loss": 0.7647, "step": 44922 }, { "epoch": 3.5206112852664577, "grad_norm": 3.1728012561798096, "learning_rate": 3.5159449221691956e-07, "loss": 1.4896, "step": 44923 }, { "epoch": 3.520689655172414, "grad_norm": 8.747281074523926, "learning_rate": 3.514809111517298e-07, "loss": 0.8071, "step": 44924 }, { "epoch": 3.52076802507837, "grad_norm": 11.029987335205078, "learning_rate": 3.513673477671198e-07, "loss": 1.6415, "step": 44925 }, { "epoch": 3.520846394984326, "grad_norm": 5.656183242797852, "learning_rate": 3.512538020635203e-07, "loss": 1.1075, "step": 44926 }, { "epoch": 3.520924764890282, "grad_norm": 3.0555710792541504, "learning_rate": 3.511402740413633e-07, "loss": 1.6105, "step": 44927 }, { "epoch": 3.5210031347962385, "grad_norm": 2.9944539070129395, "learning_rate": 3.510267637010811e-07, "loss": 1.3834, "step": 44928 }, { "epoch": 3.5210815047021944, "grad_norm": 7.955448150634766, "learning_rate": 3.509132710431057e-07, "loss": 0.9914, "step": 44929 }, { "epoch": 3.5211598746081503, "grad_norm": 2.4555206298828125, "learning_rate": 3.507997960678677e-07, "loss": 1.4903, "step": 44930 }, { "epoch": 3.5212382445141066, "grad_norm": 3.1071741580963135, "learning_rate": 3.5068633877579973e-07, "loss": 1.3088, "step": 44931 }, { "epoch": 3.521316614420063, "grad_norm": 4.019757270812988, "learning_rate": 3.505728991673335e-07, "loss": 1.4244, "step": 44932 }, { "epoch": 3.521394984326019, "grad_norm": 3.4686784744262695, "learning_rate": 3.5045947724289817e-07, "loss": 1.25, "step": 44933 }, { "epoch": 3.5214733542319747, "grad_norm": 3.1671555042266846, "learning_rate": 3.503460730029279e-07, "loss": 1.141, "step": 44934 }, { "epoch": 3.521551724137931, "grad_norm": 2.2455408573150635, "learning_rate": 3.5023268644785224e-07, "loss": 1.2206, "step": 44935 }, { "epoch": 3.5216300940438874, "grad_norm": 4.465085983276367, "learning_rate": 3.501193175781048e-07, "loss": 0.6589, "step": 44936 }, { "epoch": 3.5217084639498433, "grad_norm": 8.022317886352539, "learning_rate": 3.500059663941141e-07, "loss": 0.5454, "step": 44937 }, { "epoch": 3.521786833855799, "grad_norm": 5.371290683746338, "learning_rate": 3.4989263289631145e-07, "loss": 1.1664, "step": 44938 }, { "epoch": 3.5218652037617555, "grad_norm": 4.436868667602539, "learning_rate": 3.4977931708512936e-07, "loss": 1.591, "step": 44939 }, { "epoch": 3.521943573667712, "grad_norm": 4.067848205566406, "learning_rate": 3.4966601896099963e-07, "loss": 1.3336, "step": 44940 }, { "epoch": 3.5220219435736677, "grad_norm": 3.6990208625793457, "learning_rate": 3.495527385243502e-07, "loss": 1.1232, "step": 44941 }, { "epoch": 3.5221003134796236, "grad_norm": 2.6313695907592773, "learning_rate": 3.4943947577561424e-07, "loss": 1.0338, "step": 44942 }, { "epoch": 3.52217868338558, "grad_norm": 4.037350177764893, "learning_rate": 3.4932623071522065e-07, "loss": 0.8741, "step": 44943 }, { "epoch": 3.5222570532915363, "grad_norm": 3.028404712677002, "learning_rate": 3.4921300334360307e-07, "loss": 1.2413, "step": 44944 }, { "epoch": 3.522335423197492, "grad_norm": 5.361045837402344, "learning_rate": 3.4909979366118954e-07, "loss": 0.8791, "step": 44945 }, { "epoch": 3.522413793103448, "grad_norm": 5.949859142303467, "learning_rate": 3.489866016684118e-07, "loss": 1.3432, "step": 44946 }, { "epoch": 3.5224921630094044, "grad_norm": 4.553907871246338, "learning_rate": 3.488734273656996e-07, "loss": 1.2185, "step": 44947 }, { "epoch": 3.5225705329153607, "grad_norm": 3.0577452182769775, "learning_rate": 3.487602707534843e-07, "loss": 1.5419, "step": 44948 }, { "epoch": 3.5226489028213166, "grad_norm": 2.6204564571380615, "learning_rate": 3.48647131832196e-07, "loss": 1.2713, "step": 44949 }, { "epoch": 3.5227272727272725, "grad_norm": 5.693353176116943, "learning_rate": 3.48534010602265e-07, "loss": 1.402, "step": 44950 }, { "epoch": 3.522805642633229, "grad_norm": 4.833290100097656, "learning_rate": 3.4842090706412204e-07, "loss": 1.5436, "step": 44951 }, { "epoch": 3.522884012539185, "grad_norm": 2.567295789718628, "learning_rate": 3.4830782121819506e-07, "loss": 1.4488, "step": 44952 }, { "epoch": 3.522962382445141, "grad_norm": 2.729257822036743, "learning_rate": 3.481947530649166e-07, "loss": 0.9868, "step": 44953 }, { "epoch": 3.523040752351097, "grad_norm": 6.868177890777588, "learning_rate": 3.480817026047162e-07, "loss": 1.3904, "step": 44954 }, { "epoch": 3.5231191222570533, "grad_norm": 6.831542491912842, "learning_rate": 3.479686698380247e-07, "loss": 1.1359, "step": 44955 }, { "epoch": 3.5231974921630096, "grad_norm": 2.826002359390259, "learning_rate": 3.478556547652695e-07, "loss": 1.3533, "step": 44956 }, { "epoch": 3.5232758620689655, "grad_norm": 3.0940487384796143, "learning_rate": 3.477426573868814e-07, "loss": 1.3584, "step": 44957 }, { "epoch": 3.5233542319749214, "grad_norm": 3.872903347015381, "learning_rate": 3.476296777032917e-07, "loss": 1.33, "step": 44958 }, { "epoch": 3.5234326018808777, "grad_norm": 2.667881965637207, "learning_rate": 3.475167157149295e-07, "loss": 1.315, "step": 44959 }, { "epoch": 3.523510971786834, "grad_norm": 3.131579875946045, "learning_rate": 3.474037714222234e-07, "loss": 1.218, "step": 44960 }, { "epoch": 3.52358934169279, "grad_norm": 3.7492594718933105, "learning_rate": 3.4729084482560405e-07, "loss": 1.3811, "step": 44961 }, { "epoch": 3.523667711598746, "grad_norm": 3.4031007289886475, "learning_rate": 3.471779359255001e-07, "loss": 1.2444, "step": 44962 }, { "epoch": 3.523746081504702, "grad_norm": 8.499896049499512, "learning_rate": 3.4706504472234114e-07, "loss": 1.6443, "step": 44963 }, { "epoch": 3.5238244514106585, "grad_norm": 4.777261734008789, "learning_rate": 3.4695217121655743e-07, "loss": 2.0141, "step": 44964 }, { "epoch": 3.5239028213166144, "grad_norm": 2.3268837928771973, "learning_rate": 3.468393154085775e-07, "loss": 1.1701, "step": 44965 }, { "epoch": 3.5239811912225707, "grad_norm": 2.7995989322662354, "learning_rate": 3.46726477298831e-07, "loss": 1.4971, "step": 44966 }, { "epoch": 3.5240595611285266, "grad_norm": 3.378499746322632, "learning_rate": 3.466136568877465e-07, "loss": 1.7626, "step": 44967 }, { "epoch": 3.524137931034483, "grad_norm": 3.5020534992218018, "learning_rate": 3.4650085417575363e-07, "loss": 1.3079, "step": 44968 }, { "epoch": 3.524216300940439, "grad_norm": 3.1843128204345703, "learning_rate": 3.463880691632815e-07, "loss": 1.1774, "step": 44969 }, { "epoch": 3.524294670846395, "grad_norm": 4.043651580810547, "learning_rate": 3.4627530185075974e-07, "loss": 1.3452, "step": 44970 }, { "epoch": 3.524373040752351, "grad_norm": 3.326045513153076, "learning_rate": 3.461625522386147e-07, "loss": 1.4847, "step": 44971 }, { "epoch": 3.5244514106583074, "grad_norm": 3.359466075897217, "learning_rate": 3.4604982032727774e-07, "loss": 1.3757, "step": 44972 }, { "epoch": 3.5245297805642632, "grad_norm": 3.933342456817627, "learning_rate": 3.4593710611717734e-07, "loss": 0.9037, "step": 44973 }, { "epoch": 3.5246081504702196, "grad_norm": 3.318037986755371, "learning_rate": 3.458244096087415e-07, "loss": 1.4654, "step": 44974 }, { "epoch": 3.5246865203761755, "grad_norm": 2.8327796459198, "learning_rate": 3.4571173080239883e-07, "loss": 1.5814, "step": 44975 }, { "epoch": 3.524764890282132, "grad_norm": 2.7021830081939697, "learning_rate": 3.455990696985773e-07, "loss": 1.2043, "step": 44976 }, { "epoch": 3.5248432601880877, "grad_norm": 2.9000535011291504, "learning_rate": 3.454864262977081e-07, "loss": 1.2547, "step": 44977 }, { "epoch": 3.524921630094044, "grad_norm": 2.6125941276550293, "learning_rate": 3.4537380060021665e-07, "loss": 1.2198, "step": 44978 }, { "epoch": 3.525, "grad_norm": 9.2074613571167, "learning_rate": 3.4526119260653244e-07, "loss": 1.6345, "step": 44979 }, { "epoch": 3.5250783699059562, "grad_norm": 2.424464702606201, "learning_rate": 3.4514860231708414e-07, "loss": 1.3187, "step": 44980 }, { "epoch": 3.525156739811912, "grad_norm": 3.4446117877960205, "learning_rate": 3.4503602973229965e-07, "loss": 1.0991, "step": 44981 }, { "epoch": 3.5252351097178685, "grad_norm": 4.375706195831299, "learning_rate": 3.44923474852607e-07, "loss": 1.0603, "step": 44982 }, { "epoch": 3.5253134796238244, "grad_norm": 2.869684934616089, "learning_rate": 3.448109376784348e-07, "loss": 1.4244, "step": 44983 }, { "epoch": 3.5253918495297807, "grad_norm": 6.105940341949463, "learning_rate": 3.446984182102103e-07, "loss": 1.7005, "step": 44984 }, { "epoch": 3.5254702194357366, "grad_norm": 2.451357841491699, "learning_rate": 3.445859164483623e-07, "loss": 1.4594, "step": 44985 }, { "epoch": 3.525548589341693, "grad_norm": 2.4088592529296875, "learning_rate": 3.4447343239331797e-07, "loss": 1.3764, "step": 44986 }, { "epoch": 3.525626959247649, "grad_norm": 2.358675003051758, "learning_rate": 3.443609660455055e-07, "loss": 1.193, "step": 44987 }, { "epoch": 3.525705329153605, "grad_norm": 3.6235947608947754, "learning_rate": 3.442485174053534e-07, "loss": 0.9818, "step": 44988 }, { "epoch": 3.525783699059561, "grad_norm": 6.6422438621521, "learning_rate": 3.4413608647328786e-07, "loss": 1.2918, "step": 44989 }, { "epoch": 3.5258620689655173, "grad_norm": 2.402750015258789, "learning_rate": 3.440236732497365e-07, "loss": 1.4422, "step": 44990 }, { "epoch": 3.5259404388714732, "grad_norm": 3.9333860874176025, "learning_rate": 3.4391127773512834e-07, "loss": 1.363, "step": 44991 }, { "epoch": 3.5260188087774296, "grad_norm": 11.681182861328125, "learning_rate": 3.437988999298914e-07, "loss": 1.4938, "step": 44992 }, { "epoch": 3.5260971786833855, "grad_norm": 2.7828469276428223, "learning_rate": 3.4368653983445024e-07, "loss": 1.2226, "step": 44993 }, { "epoch": 3.526175548589342, "grad_norm": 2.725165367126465, "learning_rate": 3.435741974492346e-07, "loss": 1.3593, "step": 44994 }, { "epoch": 3.5262539184952977, "grad_norm": 3.9754250049591064, "learning_rate": 3.434618727746697e-07, "loss": 1.3717, "step": 44995 }, { "epoch": 3.526332288401254, "grad_norm": 5.6829938888549805, "learning_rate": 3.4334956581118627e-07, "loss": 0.8404, "step": 44996 }, { "epoch": 3.52641065830721, "grad_norm": 2.4812347888946533, "learning_rate": 3.4323727655920794e-07, "loss": 1.1666, "step": 44997 }, { "epoch": 3.5264890282131662, "grad_norm": 2.963552951812744, "learning_rate": 3.4312500501916313e-07, "loss": 1.5285, "step": 44998 }, { "epoch": 3.526567398119122, "grad_norm": 3.0984225273132324, "learning_rate": 3.4301275119147934e-07, "loss": 1.3728, "step": 44999 }, { "epoch": 3.5266457680250785, "grad_norm": 6.526635646820068, "learning_rate": 3.4290051507658294e-07, "loss": 1.3112, "step": 45000 }, { "epoch": 3.5267241379310343, "grad_norm": 6.785057067871094, "learning_rate": 3.427882966749008e-07, "loss": 1.0651, "step": 45001 }, { "epoch": 3.5268025078369907, "grad_norm": 3.668325185775757, "learning_rate": 3.426760959868597e-07, "loss": 1.1013, "step": 45002 }, { "epoch": 3.5268808777429466, "grad_norm": 3.431171417236328, "learning_rate": 3.4256391301288774e-07, "loss": 1.5467, "step": 45003 }, { "epoch": 3.526959247648903, "grad_norm": 2.743060350418091, "learning_rate": 3.4245174775340837e-07, "loss": 1.4007, "step": 45004 }, { "epoch": 3.527037617554859, "grad_norm": 2.798295259475708, "learning_rate": 3.423396002088514e-07, "loss": 1.2441, "step": 45005 }, { "epoch": 3.527115987460815, "grad_norm": 7.436317443847656, "learning_rate": 3.4222747037964244e-07, "loss": 1.454, "step": 45006 }, { "epoch": 3.527194357366771, "grad_norm": 4.344583988189697, "learning_rate": 3.4211535826620847e-07, "loss": 1.4998, "step": 45007 }, { "epoch": 3.5272727272727273, "grad_norm": 3.7143325805664062, "learning_rate": 3.420032638689741e-07, "loss": 1.2697, "step": 45008 }, { "epoch": 3.5273510971786832, "grad_norm": 6.314131259918213, "learning_rate": 3.418911871883668e-07, "loss": 1.7069, "step": 45009 }, { "epoch": 3.5274294670846396, "grad_norm": 2.1805644035339355, "learning_rate": 3.4177912822481286e-07, "loss": 1.3234, "step": 45010 }, { "epoch": 3.5275078369905954, "grad_norm": 3.530550718307495, "learning_rate": 3.416670869787397e-07, "loss": 1.5437, "step": 45011 }, { "epoch": 3.527586206896552, "grad_norm": 13.278440475463867, "learning_rate": 3.4155506345057153e-07, "loss": 2.0931, "step": 45012 }, { "epoch": 3.5276645768025077, "grad_norm": 2.8262834548950195, "learning_rate": 3.414430576407346e-07, "loss": 1.2932, "step": 45013 }, { "epoch": 3.527742946708464, "grad_norm": 3.6275930404663086, "learning_rate": 3.413310695496552e-07, "loss": 0.4843, "step": 45014 }, { "epoch": 3.5278213166144203, "grad_norm": 5.162683963775635, "learning_rate": 3.412190991777614e-07, "loss": 1.6728, "step": 45015 }, { "epoch": 3.527899686520376, "grad_norm": 4.460048675537109, "learning_rate": 3.4110714652547614e-07, "loss": 0.7962, "step": 45016 }, { "epoch": 3.527978056426332, "grad_norm": 5.7687530517578125, "learning_rate": 3.4099521159322634e-07, "loss": 1.5274, "step": 45017 }, { "epoch": 3.5280564263322884, "grad_norm": 12.361352920532227, "learning_rate": 3.4088329438143773e-07, "loss": 1.6272, "step": 45018 }, { "epoch": 3.5281347962382448, "grad_norm": 3.1124236583709717, "learning_rate": 3.407713948905356e-07, "loss": 1.187, "step": 45019 }, { "epoch": 3.5282131661442007, "grad_norm": 2.4524941444396973, "learning_rate": 3.406595131209467e-07, "loss": 1.2234, "step": 45020 }, { "epoch": 3.5282915360501566, "grad_norm": 2.750962018966675, "learning_rate": 3.4054764907309525e-07, "loss": 1.171, "step": 45021 }, { "epoch": 3.528369905956113, "grad_norm": 9.894577026367188, "learning_rate": 3.404358027474075e-07, "loss": 0.6529, "step": 45022 }, { "epoch": 3.528448275862069, "grad_norm": 5.8395891189575195, "learning_rate": 3.403239741443087e-07, "loss": 1.3881, "step": 45023 }, { "epoch": 3.528526645768025, "grad_norm": 3.3434338569641113, "learning_rate": 3.402121632642241e-07, "loss": 0.8969, "step": 45024 }, { "epoch": 3.528605015673981, "grad_norm": 4.488088607788086, "learning_rate": 3.401003701075789e-07, "loss": 1.347, "step": 45025 }, { "epoch": 3.5286833855799373, "grad_norm": 6.590862274169922, "learning_rate": 3.399885946747994e-07, "loss": 0.9742, "step": 45026 }, { "epoch": 3.5287617554858937, "grad_norm": 3.1522297859191895, "learning_rate": 3.398768369663091e-07, "loss": 1.5945, "step": 45027 }, { "epoch": 3.5288401253918495, "grad_norm": 2.5415172576904297, "learning_rate": 3.397650969825328e-07, "loss": 1.6846, "step": 45028 }, { "epoch": 3.5289184952978054, "grad_norm": 18.084918975830078, "learning_rate": 3.396533747238978e-07, "loss": 1.9776, "step": 45029 }, { "epoch": 3.5289968652037618, "grad_norm": 8.470366477966309, "learning_rate": 3.3954167019082716e-07, "loss": 0.5414, "step": 45030 }, { "epoch": 3.529075235109718, "grad_norm": 3.521015167236328, "learning_rate": 3.394299833837461e-07, "loss": 2.0391, "step": 45031 }, { "epoch": 3.529153605015674, "grad_norm": 3.068098306655884, "learning_rate": 3.3931831430307926e-07, "loss": 1.0647, "step": 45032 }, { "epoch": 3.52923197492163, "grad_norm": 3.068188428878784, "learning_rate": 3.3920666294925187e-07, "loss": 1.45, "step": 45033 }, { "epoch": 3.529310344827586, "grad_norm": 3.4796650409698486, "learning_rate": 3.3909502932268856e-07, "loss": 1.8032, "step": 45034 }, { "epoch": 3.5293887147335425, "grad_norm": 5.461909770965576, "learning_rate": 3.38983413423814e-07, "loss": 1.2518, "step": 45035 }, { "epoch": 3.5294670846394984, "grad_norm": 3.5657033920288086, "learning_rate": 3.388718152530518e-07, "loss": 1.0002, "step": 45036 }, { "epoch": 3.5295454545454543, "grad_norm": 2.994544267654419, "learning_rate": 3.3876023481082763e-07, "loss": 1.1708, "step": 45037 }, { "epoch": 3.5296238244514107, "grad_norm": 3.916093111038208, "learning_rate": 3.3864867209756515e-07, "loss": 1.669, "step": 45038 }, { "epoch": 3.529702194357367, "grad_norm": 4.838544845581055, "learning_rate": 3.385371271136889e-07, "loss": 1.1733, "step": 45039 }, { "epoch": 3.529780564263323, "grad_norm": 5.548696041107178, "learning_rate": 3.3842559985962363e-07, "loss": 1.4815, "step": 45040 }, { "epoch": 3.5298589341692788, "grad_norm": 12.669560432434082, "learning_rate": 3.383140903357923e-07, "loss": 0.9979, "step": 45041 }, { "epoch": 3.529937304075235, "grad_norm": 4.154323577880859, "learning_rate": 3.3820259854262016e-07, "loss": 0.4428, "step": 45042 }, { "epoch": 3.5300156739811914, "grad_norm": 8.774290084838867, "learning_rate": 3.380911244805307e-07, "loss": 0.938, "step": 45043 }, { "epoch": 3.5300940438871473, "grad_norm": 2.616480827331543, "learning_rate": 3.379796681499492e-07, "loss": 1.4818, "step": 45044 }, { "epoch": 3.530172413793103, "grad_norm": 5.684689521789551, "learning_rate": 3.378682295512975e-07, "loss": 1.5946, "step": 45045 }, { "epoch": 3.5302507836990595, "grad_norm": 3.6856977939605713, "learning_rate": 3.377568086849997e-07, "loss": 1.3317, "step": 45046 }, { "epoch": 3.530329153605016, "grad_norm": 4.09885835647583, "learning_rate": 3.3764540555148104e-07, "loss": 1.4173, "step": 45047 }, { "epoch": 3.5304075235109718, "grad_norm": 8.155447959899902, "learning_rate": 3.375340201511651e-07, "loss": 1.006, "step": 45048 }, { "epoch": 3.5304858934169276, "grad_norm": 3.6369285583496094, "learning_rate": 3.3742265248447426e-07, "loss": 1.3559, "step": 45049 }, { "epoch": 3.530564263322884, "grad_norm": 5.363347053527832, "learning_rate": 3.373113025518332e-07, "loss": 1.4805, "step": 45050 }, { "epoch": 3.5306426332288403, "grad_norm": 2.954928159713745, "learning_rate": 3.371999703536638e-07, "loss": 1.4172, "step": 45051 }, { "epoch": 3.530721003134796, "grad_norm": 3.696592092514038, "learning_rate": 3.370886558903924e-07, "loss": 1.5866, "step": 45052 }, { "epoch": 3.530799373040752, "grad_norm": 4.05772590637207, "learning_rate": 3.369773591624403e-07, "loss": 1.42, "step": 45053 }, { "epoch": 3.5308777429467084, "grad_norm": 2.9285154342651367, "learning_rate": 3.3686608017023117e-07, "loss": 1.2751, "step": 45054 }, { "epoch": 3.5309561128526648, "grad_norm": 2.475466728210449, "learning_rate": 3.3675481891418784e-07, "loss": 1.1267, "step": 45055 }, { "epoch": 3.5310344827586206, "grad_norm": 3.791396141052246, "learning_rate": 3.366435753947345e-07, "loss": 1.5808, "step": 45056 }, { "epoch": 3.5311128526645765, "grad_norm": 3.6892683506011963, "learning_rate": 3.3653234961229355e-07, "loss": 1.4153, "step": 45057 }, { "epoch": 3.531191222570533, "grad_norm": 2.867906093597412, "learning_rate": 3.36421141567288e-07, "loss": 1.2566, "step": 45058 }, { "epoch": 3.531269592476489, "grad_norm": 4.901699542999268, "learning_rate": 3.36309951260142e-07, "loss": 2.0122, "step": 45059 }, { "epoch": 3.531347962382445, "grad_norm": 3.1368250846862793, "learning_rate": 3.3619877869127626e-07, "loss": 1.3381, "step": 45060 }, { "epoch": 3.5314263322884014, "grad_norm": 3.9980006217956543, "learning_rate": 3.360876238611155e-07, "loss": 0.3927, "step": 45061 }, { "epoch": 3.5315047021943573, "grad_norm": 3.1660733222961426, "learning_rate": 3.359764867700815e-07, "loss": 1.7036, "step": 45062 }, { "epoch": 3.5315830721003136, "grad_norm": 4.9755377769470215, "learning_rate": 3.358653674185985e-07, "loss": 1.4652, "step": 45063 }, { "epoch": 3.5316614420062695, "grad_norm": 4.680377006530762, "learning_rate": 3.3575426580708723e-07, "loss": 1.2154, "step": 45064 }, { "epoch": 3.531739811912226, "grad_norm": 3.609837532043457, "learning_rate": 3.3564318193597014e-07, "loss": 1.1982, "step": 45065 }, { "epoch": 3.5318181818181817, "grad_norm": 3.386258363723755, "learning_rate": 3.3553211580567125e-07, "loss": 1.2577, "step": 45066 }, { "epoch": 3.531896551724138, "grad_norm": 2.8626275062561035, "learning_rate": 3.354210674166136e-07, "loss": 1.4435, "step": 45067 }, { "epoch": 3.531974921630094, "grad_norm": 4.085061073303223, "learning_rate": 3.353100367692175e-07, "loss": 1.3815, "step": 45068 }, { "epoch": 3.5320532915360503, "grad_norm": 2.64756441116333, "learning_rate": 3.351990238639058e-07, "loss": 1.1197, "step": 45069 }, { "epoch": 3.532131661442006, "grad_norm": 4.767975330352783, "learning_rate": 3.3508802870109993e-07, "loss": 1.5082, "step": 45070 }, { "epoch": 3.5322100313479625, "grad_norm": 5.9675703048706055, "learning_rate": 3.3497705128122505e-07, "loss": 1.4012, "step": 45071 }, { "epoch": 3.5322884012539184, "grad_norm": 4.419992446899414, "learning_rate": 3.3486609160470083e-07, "loss": 1.4032, "step": 45072 }, { "epoch": 3.5323667711598747, "grad_norm": 2.7201881408691406, "learning_rate": 3.347551496719498e-07, "loss": 0.9958, "step": 45073 }, { "epoch": 3.5324451410658306, "grad_norm": 5.77666711807251, "learning_rate": 3.346442254833937e-07, "loss": 1.2799, "step": 45074 }, { "epoch": 3.532523510971787, "grad_norm": 3.135845422744751, "learning_rate": 3.345333190394545e-07, "loss": 1.4578, "step": 45075 }, { "epoch": 3.532601880877743, "grad_norm": 3.030616521835327, "learning_rate": 3.3442243034055465e-07, "loss": 1.1801, "step": 45076 }, { "epoch": 3.532680250783699, "grad_norm": 3.031611919403076, "learning_rate": 3.3431155938711547e-07, "loss": 1.4852, "step": 45077 }, { "epoch": 3.532758620689655, "grad_norm": 2.6767497062683105, "learning_rate": 3.3420070617955937e-07, "loss": 1.3632, "step": 45078 }, { "epoch": 3.5328369905956114, "grad_norm": 3.6682956218719482, "learning_rate": 3.3408987071830543e-07, "loss": 0.8246, "step": 45079 }, { "epoch": 3.5329153605015673, "grad_norm": 2.8371877670288086, "learning_rate": 3.3397905300377786e-07, "loss": 1.3558, "step": 45080 }, { "epoch": 3.5329937304075236, "grad_norm": 2.688493490219116, "learning_rate": 3.3386825303639793e-07, "loss": 1.4751, "step": 45081 }, { "epoch": 3.5330721003134795, "grad_norm": 2.5105948448181152, "learning_rate": 3.3375747081658696e-07, "loss": 1.2991, "step": 45082 }, { "epoch": 3.533150470219436, "grad_norm": 6.013301849365234, "learning_rate": 3.336467063447646e-07, "loss": 1.4006, "step": 45083 }, { "epoch": 3.5332288401253917, "grad_norm": 5.543315410614014, "learning_rate": 3.335359596213533e-07, "loss": 0.5853, "step": 45084 }, { "epoch": 3.533307210031348, "grad_norm": 2.498538017272949, "learning_rate": 3.3342523064677556e-07, "loss": 1.3821, "step": 45085 }, { "epoch": 3.533385579937304, "grad_norm": 8.331563949584961, "learning_rate": 3.3331451942145045e-07, "loss": 1.0784, "step": 45086 }, { "epoch": 3.5334639498432603, "grad_norm": 3.1741726398468018, "learning_rate": 3.332038259458004e-07, "loss": 1.1769, "step": 45087 }, { "epoch": 3.533542319749216, "grad_norm": 4.006962776184082, "learning_rate": 3.330931502202456e-07, "loss": 1.2711, "step": 45088 }, { "epoch": 3.5336206896551725, "grad_norm": 5.8434929847717285, "learning_rate": 3.329824922452074e-07, "loss": 1.1897, "step": 45089 }, { "epoch": 3.5336990595611284, "grad_norm": 3.3492841720581055, "learning_rate": 3.328718520211066e-07, "loss": 1.4348, "step": 45090 }, { "epoch": 3.5337774294670847, "grad_norm": 4.711234092712402, "learning_rate": 3.32761229548364e-07, "loss": 1.298, "step": 45091 }, { "epoch": 3.5338557993730406, "grad_norm": 2.2964529991149902, "learning_rate": 3.3265062482740083e-07, "loss": 1.2593, "step": 45092 }, { "epoch": 3.533934169278997, "grad_norm": 7.8372883796691895, "learning_rate": 3.325400378586374e-07, "loss": 1.2373, "step": 45093 }, { "epoch": 3.534012539184953, "grad_norm": 3.3064000606536865, "learning_rate": 3.324294686424939e-07, "loss": 1.3195, "step": 45094 }, { "epoch": 3.534090909090909, "grad_norm": 2.086702346801758, "learning_rate": 3.3231891717939167e-07, "loss": 1.1675, "step": 45095 }, { "epoch": 3.534169278996865, "grad_norm": 3.0843029022216797, "learning_rate": 3.322083834697504e-07, "loss": 1.1995, "step": 45096 }, { "epoch": 3.5342476489028214, "grad_norm": 2.700828790664673, "learning_rate": 3.320978675139919e-07, "loss": 1.2184, "step": 45097 }, { "epoch": 3.5343260188087773, "grad_norm": 3.3059356212615967, "learning_rate": 3.319873693125342e-07, "loss": 1.7575, "step": 45098 }, { "epoch": 3.5344043887147336, "grad_norm": 4.597807884216309, "learning_rate": 3.3187688886579914e-07, "loss": 1.0998, "step": 45099 }, { "epoch": 3.5344827586206895, "grad_norm": 7.199812889099121, "learning_rate": 3.3176642617420817e-07, "loss": 1.1121, "step": 45100 }, { "epoch": 3.534561128526646, "grad_norm": 5.352498531341553, "learning_rate": 3.3165598123817865e-07, "loss": 1.7414, "step": 45101 }, { "epoch": 3.5346394984326017, "grad_norm": 2.1088151931762695, "learning_rate": 3.315455540581319e-07, "loss": 1.1391, "step": 45102 }, { "epoch": 3.534717868338558, "grad_norm": 2.871234178543091, "learning_rate": 3.314351446344877e-07, "loss": 1.5026, "step": 45103 }, { "epoch": 3.534796238244514, "grad_norm": 3.302685022354126, "learning_rate": 3.3132475296766776e-07, "loss": 1.5037, "step": 45104 }, { "epoch": 3.5348746081504703, "grad_norm": 2.873446226119995, "learning_rate": 3.3121437905808906e-07, "loss": 1.228, "step": 45105 }, { "epoch": 3.5349529780564266, "grad_norm": 3.892031669616699, "learning_rate": 3.311040229061735e-07, "loss": 1.4382, "step": 45106 }, { "epoch": 3.5350313479623825, "grad_norm": 4.464818000793457, "learning_rate": 3.3099368451234016e-07, "loss": 0.8379, "step": 45107 }, { "epoch": 3.5351097178683384, "grad_norm": 3.5060484409332275, "learning_rate": 3.3088336387700815e-07, "loss": 1.47, "step": 45108 }, { "epoch": 3.5351880877742947, "grad_norm": 2.2168188095092773, "learning_rate": 3.3077306100059825e-07, "loss": 1.411, "step": 45109 }, { "epoch": 3.535266457680251, "grad_norm": 33.18061828613281, "learning_rate": 3.3066277588352903e-07, "loss": 0.9147, "step": 45110 }, { "epoch": 3.535344827586207, "grad_norm": 2.622478723526001, "learning_rate": 3.305525085262201e-07, "loss": 1.0738, "step": 45111 }, { "epoch": 3.535423197492163, "grad_norm": 4.190285682678223, "learning_rate": 3.304422589290912e-07, "loss": 1.404, "step": 45112 }, { "epoch": 3.535501567398119, "grad_norm": 8.217195510864258, "learning_rate": 3.303320270925614e-07, "loss": 1.6835, "step": 45113 }, { "epoch": 3.5355799373040755, "grad_norm": 3.011319398880005, "learning_rate": 3.3022181301705036e-07, "loss": 1.0759, "step": 45114 }, { "epoch": 3.5356583072100314, "grad_norm": 3.1602957248687744, "learning_rate": 3.3011161670297776e-07, "loss": 1.3813, "step": 45115 }, { "epoch": 3.5357366771159873, "grad_norm": 4.922844886779785, "learning_rate": 3.30001438150761e-07, "loss": 1.5702, "step": 45116 }, { "epoch": 3.5358150470219436, "grad_norm": 3.899658441543579, "learning_rate": 3.298912773608193e-07, "loss": 1.2719, "step": 45117 }, { "epoch": 3.5358934169279, "grad_norm": 3.7621188163757324, "learning_rate": 3.297811343335733e-07, "loss": 0.968, "step": 45118 }, { "epoch": 3.535971786833856, "grad_norm": 2.80361008644104, "learning_rate": 3.2967100906944216e-07, "loss": 1.2893, "step": 45119 }, { "epoch": 3.5360501567398117, "grad_norm": 2.1497623920440674, "learning_rate": 3.295609015688422e-07, "loss": 1.2776, "step": 45120 }, { "epoch": 3.536128526645768, "grad_norm": 3.6284801959991455, "learning_rate": 3.294508118321943e-07, "loss": 0.963, "step": 45121 }, { "epoch": 3.5362068965517244, "grad_norm": 4.145996570587158, "learning_rate": 3.293407398599158e-07, "loss": 1.4266, "step": 45122 }, { "epoch": 3.5362852664576803, "grad_norm": 3.223280429840088, "learning_rate": 3.2923068565242743e-07, "loss": 1.1064, "step": 45123 }, { "epoch": 3.536363636363636, "grad_norm": 6.766166687011719, "learning_rate": 3.2912064921014564e-07, "loss": 1.7382, "step": 45124 }, { "epoch": 3.5364420062695925, "grad_norm": 3.044344902038574, "learning_rate": 3.2901063053349e-07, "loss": 1.3738, "step": 45125 }, { "epoch": 3.536520376175549, "grad_norm": 3.4758753776550293, "learning_rate": 3.289006296228786e-07, "loss": 1.7141, "step": 45126 }, { "epoch": 3.5365987460815047, "grad_norm": 2.27966046333313, "learning_rate": 3.2879064647872993e-07, "loss": 1.4668, "step": 45127 }, { "epoch": 3.5366771159874606, "grad_norm": 7.18875789642334, "learning_rate": 3.2868068110146257e-07, "loss": 1.245, "step": 45128 }, { "epoch": 3.536755485893417, "grad_norm": 3.2655229568481445, "learning_rate": 3.2857073349149446e-07, "loss": 1.6527, "step": 45129 }, { "epoch": 3.5368338557993733, "grad_norm": 3.431506395339966, "learning_rate": 3.2846080364924373e-07, "loss": 1.073, "step": 45130 }, { "epoch": 3.536912225705329, "grad_norm": 2.894240617752075, "learning_rate": 3.283508915751288e-07, "loss": 1.2467, "step": 45131 }, { "epoch": 3.536990595611285, "grad_norm": 3.381429433822632, "learning_rate": 3.2824099726956717e-07, "loss": 1.3103, "step": 45132 }, { "epoch": 3.5370689655172414, "grad_norm": 4.009801864624023, "learning_rate": 3.2813112073297795e-07, "loss": 1.3592, "step": 45133 }, { "epoch": 3.5371473354231977, "grad_norm": 3.586308002471924, "learning_rate": 3.2802126196577857e-07, "loss": 1.4397, "step": 45134 }, { "epoch": 3.5372257053291536, "grad_norm": 2.4653549194335938, "learning_rate": 3.2791142096838534e-07, "loss": 1.3774, "step": 45135 }, { "epoch": 3.5373040752351095, "grad_norm": 3.2670047283172607, "learning_rate": 3.27801597741218e-07, "loss": 1.7194, "step": 45136 }, { "epoch": 3.537382445141066, "grad_norm": 14.436022758483887, "learning_rate": 3.2769179228469337e-07, "loss": 1.1401, "step": 45137 }, { "epoch": 3.537460815047022, "grad_norm": 2.9731361865997314, "learning_rate": 3.2758200459923064e-07, "loss": 0.9931, "step": 45138 }, { "epoch": 3.537539184952978, "grad_norm": 3.06247878074646, "learning_rate": 3.274722346852449e-07, "loss": 1.3478, "step": 45139 }, { "epoch": 3.537617554858934, "grad_norm": 4.742161750793457, "learning_rate": 3.2736248254315373e-07, "loss": 1.1265, "step": 45140 }, { "epoch": 3.5376959247648903, "grad_norm": 2.0147931575775146, "learning_rate": 3.2725274817337784e-07, "loss": 1.1632, "step": 45141 }, { "epoch": 3.5377742946708466, "grad_norm": 6.906355857849121, "learning_rate": 3.271430315763313e-07, "loss": 2.3032, "step": 45142 }, { "epoch": 3.5378526645768025, "grad_norm": 9.700284957885742, "learning_rate": 3.2703333275243276e-07, "loss": 1.5835, "step": 45143 }, { "epoch": 3.5379310344827584, "grad_norm": 2.876715898513794, "learning_rate": 3.2692365170209907e-07, "loss": 1.0811, "step": 45144 }, { "epoch": 3.5380094043887147, "grad_norm": 5.945509910583496, "learning_rate": 3.268139884257476e-07, "loss": 1.3546, "step": 45145 }, { "epoch": 3.538087774294671, "grad_norm": 4.320316314697266, "learning_rate": 3.2670434292379537e-07, "loss": 1.2372, "step": 45146 }, { "epoch": 3.538166144200627, "grad_norm": 3.237767219543457, "learning_rate": 3.265947151966592e-07, "loss": 0.8151, "step": 45147 }, { "epoch": 3.538244514106583, "grad_norm": 2.568145751953125, "learning_rate": 3.264851052447571e-07, "loss": 1.0995, "step": 45148 }, { "epoch": 3.538322884012539, "grad_norm": 6.509252548217773, "learning_rate": 3.2637551306850423e-07, "loss": 1.5945, "step": 45149 }, { "epoch": 3.5384012539184955, "grad_norm": 2.473222255706787, "learning_rate": 3.262659386683192e-07, "loss": 1.24, "step": 45150 }, { "epoch": 3.5384796238244514, "grad_norm": 3.518258810043335, "learning_rate": 3.261563820446173e-07, "loss": 1.2205, "step": 45151 }, { "epoch": 3.5385579937304072, "grad_norm": 3.0490996837615967, "learning_rate": 3.260468431978164e-07, "loss": 1.3061, "step": 45152 }, { "epoch": 3.5386363636363636, "grad_norm": 4.188325881958008, "learning_rate": 3.2593732212833353e-07, "loss": 1.2221, "step": 45153 }, { "epoch": 3.53871473354232, "grad_norm": 7.510469436645508, "learning_rate": 3.258278188365821e-07, "loss": 0.7768, "step": 45154 }, { "epoch": 3.538793103448276, "grad_norm": 2.8657708168029785, "learning_rate": 3.2571833332298184e-07, "loss": 1.4608, "step": 45155 }, { "epoch": 3.538871473354232, "grad_norm": 3.1797330379486084, "learning_rate": 3.2560886558794915e-07, "loss": 1.6369, "step": 45156 }, { "epoch": 3.538949843260188, "grad_norm": 4.064119815826416, "learning_rate": 3.254994156318986e-07, "loss": 1.3417, "step": 45157 }, { "epoch": 3.5390282131661444, "grad_norm": 1.9972046613693237, "learning_rate": 3.253899834552476e-07, "loss": 1.3371, "step": 45158 }, { "epoch": 3.5391065830721002, "grad_norm": 4.61084508895874, "learning_rate": 3.25280569058411e-07, "loss": 1.284, "step": 45159 }, { "epoch": 3.5391849529780566, "grad_norm": 3.0251104831695557, "learning_rate": 3.251711724418072e-07, "loss": 1.5901, "step": 45160 }, { "epoch": 3.5392633228840125, "grad_norm": 2.641597032546997, "learning_rate": 3.2506179360585087e-07, "loss": 1.6198, "step": 45161 }, { "epoch": 3.539341692789969, "grad_norm": 2.635239839553833, "learning_rate": 3.2495243255095786e-07, "loss": 1.2617, "step": 45162 }, { "epoch": 3.5394200626959247, "grad_norm": 3.350090265274048, "learning_rate": 3.2484308927754503e-07, "loss": 1.31, "step": 45163 }, { "epoch": 3.539498432601881, "grad_norm": 21.59554672241211, "learning_rate": 3.2473376378602764e-07, "loss": 3.1115, "step": 45164 }, { "epoch": 3.539576802507837, "grad_norm": 2.3172056674957275, "learning_rate": 3.2462445607682136e-07, "loss": 1.0348, "step": 45165 }, { "epoch": 3.5396551724137932, "grad_norm": 5.846957206726074, "learning_rate": 3.245151661503421e-07, "loss": 1.0854, "step": 45166 }, { "epoch": 3.539733542319749, "grad_norm": 2.7241179943084717, "learning_rate": 3.244058940070061e-07, "loss": 1.1975, "step": 45167 }, { "epoch": 3.5398119122257055, "grad_norm": 3.0743861198425293, "learning_rate": 3.2429663964722856e-07, "loss": 1.449, "step": 45168 }, { "epoch": 3.5398902821316613, "grad_norm": 2.8233370780944824, "learning_rate": 3.2418740307142483e-07, "loss": 1.1953, "step": 45169 }, { "epoch": 3.5399686520376177, "grad_norm": 4.081424236297607, "learning_rate": 3.2407818428001057e-07, "loss": 1.6968, "step": 45170 }, { "epoch": 3.5400470219435736, "grad_norm": 3.2101261615753174, "learning_rate": 3.2396898327340167e-07, "loss": 1.1807, "step": 45171 }, { "epoch": 3.54012539184953, "grad_norm": 3.434145927429199, "learning_rate": 3.238598000520127e-07, "loss": 1.2299, "step": 45172 }, { "epoch": 3.540203761755486, "grad_norm": 3.040938138961792, "learning_rate": 3.2375063461625834e-07, "loss": 1.2141, "step": 45173 }, { "epoch": 3.540282131661442, "grad_norm": 2.404313325881958, "learning_rate": 3.236414869665555e-07, "loss": 1.1002, "step": 45174 }, { "epoch": 3.540360501567398, "grad_norm": 4.485227584838867, "learning_rate": 3.2353235710331887e-07, "loss": 1.0304, "step": 45175 }, { "epoch": 3.5404388714733543, "grad_norm": 7.9305419921875, "learning_rate": 3.234232450269631e-07, "loss": 2.0467, "step": 45176 }, { "epoch": 3.5405172413793102, "grad_norm": 3.3391289710998535, "learning_rate": 3.2331415073790283e-07, "loss": 1.3632, "step": 45177 }, { "epoch": 3.5405956112852666, "grad_norm": 3.1562795639038086, "learning_rate": 3.2320507423655276e-07, "loss": 1.7111, "step": 45178 }, { "epoch": 3.5406739811912225, "grad_norm": 3.6857221126556396, "learning_rate": 3.230960155233298e-07, "loss": 1.5391, "step": 45179 }, { "epoch": 3.540752351097179, "grad_norm": 2.4029080867767334, "learning_rate": 3.2298697459864693e-07, "loss": 1.016, "step": 45180 }, { "epoch": 3.5408307210031347, "grad_norm": 3.3578269481658936, "learning_rate": 3.2287795146291936e-07, "loss": 1.3936, "step": 45181 }, { "epoch": 3.540909090909091, "grad_norm": 4.8564534187316895, "learning_rate": 3.2276894611656117e-07, "loss": 1.3407, "step": 45182 }, { "epoch": 3.540987460815047, "grad_norm": 3.5168490409851074, "learning_rate": 3.226599585599882e-07, "loss": 1.6537, "step": 45183 }, { "epoch": 3.5410658307210032, "grad_norm": 4.143545627593994, "learning_rate": 3.22550988793614e-07, "loss": 1.3933, "step": 45184 }, { "epoch": 3.541144200626959, "grad_norm": 2.934180498123169, "learning_rate": 3.2244203681785324e-07, "loss": 1.4732, "step": 45185 }, { "epoch": 3.5412225705329154, "grad_norm": 3.403773784637451, "learning_rate": 3.223331026331211e-07, "loss": 1.6475, "step": 45186 }, { "epoch": 3.5413009404388713, "grad_norm": 4.341822624206543, "learning_rate": 3.222241862398301e-07, "loss": 1.3264, "step": 45187 }, { "epoch": 3.5413793103448277, "grad_norm": 2.7452547550201416, "learning_rate": 3.221152876383965e-07, "loss": 1.1218, "step": 45188 }, { "epoch": 3.5414576802507836, "grad_norm": 5.954352855682373, "learning_rate": 3.2200640682923334e-07, "loss": 0.8346, "step": 45189 }, { "epoch": 3.54153605015674, "grad_norm": 5.488594055175781, "learning_rate": 3.218975438127558e-07, "loss": 1.2407, "step": 45190 }, { "epoch": 3.541614420062696, "grad_norm": 3.721743583679199, "learning_rate": 3.2178869858937644e-07, "loss": 1.4829, "step": 45191 }, { "epoch": 3.541692789968652, "grad_norm": 3.5146970748901367, "learning_rate": 3.216798711595093e-07, "loss": 1.4936, "step": 45192 }, { "epoch": 3.541771159874608, "grad_norm": 4.10217809677124, "learning_rate": 3.2157106152356956e-07, "loss": 1.3764, "step": 45193 }, { "epoch": 3.5418495297805643, "grad_norm": 4.037165641784668, "learning_rate": 3.2146226968197137e-07, "loss": 1.4741, "step": 45194 }, { "epoch": 3.54192789968652, "grad_norm": 3.07649827003479, "learning_rate": 3.213534956351266e-07, "loss": 0.9399, "step": 45195 }, { "epoch": 3.5420062695924766, "grad_norm": 8.096989631652832, "learning_rate": 3.2124473938345057e-07, "loss": 1.205, "step": 45196 }, { "epoch": 3.5420846394984324, "grad_norm": 3.0493335723876953, "learning_rate": 3.211360009273562e-07, "loss": 1.6429, "step": 45197 }, { "epoch": 3.5421630094043888, "grad_norm": 5.020536422729492, "learning_rate": 3.2102728026725704e-07, "loss": 1.1988, "step": 45198 }, { "epoch": 3.5422413793103447, "grad_norm": 2.5133285522460938, "learning_rate": 3.209185774035672e-07, "loss": 1.3223, "step": 45199 }, { "epoch": 3.542319749216301, "grad_norm": 3.516559362411499, "learning_rate": 3.208098923366998e-07, "loss": 1.459, "step": 45200 }, { "epoch": 3.5423981191222573, "grad_norm": 6.706310272216797, "learning_rate": 3.207012250670677e-07, "loss": 1.3371, "step": 45201 }, { "epoch": 3.542476489028213, "grad_norm": 2.8588171005249023, "learning_rate": 3.2059257559508507e-07, "loss": 1.3867, "step": 45202 }, { "epoch": 3.542554858934169, "grad_norm": 3.7619779109954834, "learning_rate": 3.204839439211649e-07, "loss": 1.3591, "step": 45203 }, { "epoch": 3.5426332288401254, "grad_norm": 2.6820690631866455, "learning_rate": 3.203753300457202e-07, "loss": 1.1732, "step": 45204 }, { "epoch": 3.5427115987460818, "grad_norm": 3.6042637825012207, "learning_rate": 3.202667339691645e-07, "loss": 1.1873, "step": 45205 }, { "epoch": 3.5427899686520377, "grad_norm": 5.786388397216797, "learning_rate": 3.2015815569190925e-07, "loss": 0.9902, "step": 45206 }, { "epoch": 3.5428683385579935, "grad_norm": 5.273884296417236, "learning_rate": 3.200495952143695e-07, "loss": 1.9954, "step": 45207 }, { "epoch": 3.54294670846395, "grad_norm": 3.3618946075439453, "learning_rate": 3.1994105253695726e-07, "loss": 1.6796, "step": 45208 }, { "epoch": 3.543025078369906, "grad_norm": 5.758064270019531, "learning_rate": 3.198325276600861e-07, "loss": 1.9198, "step": 45209 }, { "epoch": 3.543103448275862, "grad_norm": 4.486389636993408, "learning_rate": 3.197240205841673e-07, "loss": 1.8672, "step": 45210 }, { "epoch": 3.543181818181818, "grad_norm": 4.122279167175293, "learning_rate": 3.196155313096139e-07, "loss": 1.2868, "step": 45211 }, { "epoch": 3.5432601880877743, "grad_norm": 2.850752592086792, "learning_rate": 3.195070598368399e-07, "loss": 1.3207, "step": 45212 }, { "epoch": 3.5433385579937307, "grad_norm": 2.651669502258301, "learning_rate": 3.1939860616625683e-07, "loss": 1.2607, "step": 45213 }, { "epoch": 3.5434169278996865, "grad_norm": 3.441152334213257, "learning_rate": 3.19290170298277e-07, "loss": 1.1495, "step": 45214 }, { "epoch": 3.5434952978056424, "grad_norm": 3.777353048324585, "learning_rate": 3.191817522333135e-07, "loss": 1.5651, "step": 45215 }, { "epoch": 3.5435736677115988, "grad_norm": 4.131320476531982, "learning_rate": 3.190733519717787e-07, "loss": 1.367, "step": 45216 }, { "epoch": 3.543652037617555, "grad_norm": 4.274089813232422, "learning_rate": 3.1896496951408395e-07, "loss": 1.3186, "step": 45217 }, { "epoch": 3.543730407523511, "grad_norm": 2.267327308654785, "learning_rate": 3.188566048606423e-07, "loss": 1.5896, "step": 45218 }, { "epoch": 3.543808777429467, "grad_norm": 2.0980775356292725, "learning_rate": 3.187482580118656e-07, "loss": 1.0399, "step": 45219 }, { "epoch": 3.543887147335423, "grad_norm": 3.5707778930664062, "learning_rate": 3.1863992896816634e-07, "loss": 1.6486, "step": 45220 }, { "epoch": 3.5439655172413795, "grad_norm": 2.440293788909912, "learning_rate": 3.1853161772995634e-07, "loss": 1.2016, "step": 45221 }, { "epoch": 3.5440438871473354, "grad_norm": 4.408242702484131, "learning_rate": 3.1842332429764754e-07, "loss": 1.1798, "step": 45222 }, { "epoch": 3.5441222570532913, "grad_norm": 4.118158340454102, "learning_rate": 3.1831504867165185e-07, "loss": 0.7008, "step": 45223 }, { "epoch": 3.5442006269592476, "grad_norm": 8.344866752624512, "learning_rate": 3.1820679085238116e-07, "loss": 0.9566, "step": 45224 }, { "epoch": 3.544278996865204, "grad_norm": 3.621549367904663, "learning_rate": 3.180985508402462e-07, "loss": 1.4007, "step": 45225 }, { "epoch": 3.54435736677116, "grad_norm": 2.788565158843994, "learning_rate": 3.1799032863566003e-07, "loss": 1.3095, "step": 45226 }, { "epoch": 3.5444357366771158, "grad_norm": 2.645139455795288, "learning_rate": 3.178821242390351e-07, "loss": 1.2704, "step": 45227 }, { "epoch": 3.544514106583072, "grad_norm": 7.025701522827148, "learning_rate": 3.1777393765078045e-07, "loss": 1.1478, "step": 45228 }, { "epoch": 3.5445924764890284, "grad_norm": 5.583510875701904, "learning_rate": 3.176657688713092e-07, "loss": 1.1802, "step": 45229 }, { "epoch": 3.5446708463949843, "grad_norm": 4.436340808868408, "learning_rate": 3.175576179010309e-07, "loss": 1.1688, "step": 45230 }, { "epoch": 3.54474921630094, "grad_norm": 2.685746431350708, "learning_rate": 3.174494847403603e-07, "loss": 1.2994, "step": 45231 }, { "epoch": 3.5448275862068965, "grad_norm": 6.678665637969971, "learning_rate": 3.173413693897065e-07, "loss": 1.5526, "step": 45232 }, { "epoch": 3.544905956112853, "grad_norm": 3.0575692653656006, "learning_rate": 3.172332718494803e-07, "loss": 1.5117, "step": 45233 }, { "epoch": 3.5449843260188088, "grad_norm": 3.18205189704895, "learning_rate": 3.171251921200935e-07, "loss": 1.4667, "step": 45234 }, { "epoch": 3.5450626959247646, "grad_norm": 2.6591553688049316, "learning_rate": 3.1701713020195757e-07, "loss": 1.3263, "step": 45235 }, { "epoch": 3.545141065830721, "grad_norm": 3.042268753051758, "learning_rate": 3.1690908609548265e-07, "loss": 1.4328, "step": 45236 }, { "epoch": 3.5452194357366773, "grad_norm": 2.7387962341308594, "learning_rate": 3.168010598010807e-07, "loss": 1.4721, "step": 45237 }, { "epoch": 3.545297805642633, "grad_norm": 8.31630802154541, "learning_rate": 3.166930513191618e-07, "loss": 1.5189, "step": 45238 }, { "epoch": 3.545376175548589, "grad_norm": 3.734830379486084, "learning_rate": 3.1658506065013694e-07, "loss": 1.009, "step": 45239 }, { "epoch": 3.5454545454545454, "grad_norm": 8.631817817687988, "learning_rate": 3.1647708779441733e-07, "loss": 1.2068, "step": 45240 }, { "epoch": 3.5455329153605017, "grad_norm": 13.053955078125, "learning_rate": 3.163691327524132e-07, "loss": 1.9137, "step": 45241 }, { "epoch": 3.5456112852664576, "grad_norm": 2.427412748336792, "learning_rate": 3.1626119552453595e-07, "loss": 1.1271, "step": 45242 }, { "epoch": 3.5456896551724135, "grad_norm": 2.073166847229004, "learning_rate": 3.161532761111935e-07, "loss": 0.8565, "step": 45243 }, { "epoch": 3.54576802507837, "grad_norm": 4.088028907775879, "learning_rate": 3.160453745127989e-07, "loss": 1.2682, "step": 45244 }, { "epoch": 3.545846394984326, "grad_norm": 4.1602864265441895, "learning_rate": 3.159374907297624e-07, "loss": 1.9743, "step": 45245 }, { "epoch": 3.545924764890282, "grad_norm": 2.8894708156585693, "learning_rate": 3.1582962476249424e-07, "loss": 0.9065, "step": 45246 }, { "epoch": 3.5460031347962384, "grad_norm": 3.1078245639801025, "learning_rate": 3.15721776611404e-07, "loss": 0.8744, "step": 45247 }, { "epoch": 3.5460815047021943, "grad_norm": 4.962673664093018, "learning_rate": 3.156139462769009e-07, "loss": 1.4477, "step": 45248 }, { "epoch": 3.5461598746081506, "grad_norm": 3.259230613708496, "learning_rate": 3.1550613375939675e-07, "loss": 1.2444, "step": 45249 }, { "epoch": 3.5462382445141065, "grad_norm": 3.033395528793335, "learning_rate": 3.153983390593024e-07, "loss": 1.4935, "step": 45250 }, { "epoch": 3.546316614420063, "grad_norm": 5.2198662757873535, "learning_rate": 3.152905621770258e-07, "loss": 1.2643, "step": 45251 }, { "epoch": 3.5463949843260187, "grad_norm": 5.865916728973389, "learning_rate": 3.151828031129778e-07, "loss": 1.0297, "step": 45252 }, { "epoch": 3.546473354231975, "grad_norm": 8.339921951293945, "learning_rate": 3.1507506186756795e-07, "loss": 1.2149, "step": 45253 }, { "epoch": 3.546551724137931, "grad_norm": 3.011263132095337, "learning_rate": 3.1496733844120607e-07, "loss": 0.8985, "step": 45254 }, { "epoch": 3.5466300940438873, "grad_norm": 3.6254756450653076, "learning_rate": 3.148596328343023e-07, "loss": 1.4611, "step": 45255 }, { "epoch": 3.546708463949843, "grad_norm": 9.274682998657227, "learning_rate": 3.147519450472658e-07, "loss": 2.4314, "step": 45256 }, { "epoch": 3.5467868338557995, "grad_norm": 4.179418087005615, "learning_rate": 3.146442750805068e-07, "loss": 1.1114, "step": 45257 }, { "epoch": 3.5468652037617554, "grad_norm": 3.193855047225952, "learning_rate": 3.1453662293443376e-07, "loss": 1.1128, "step": 45258 }, { "epoch": 3.5469435736677117, "grad_norm": 8.92947006225586, "learning_rate": 3.144289886094576e-07, "loss": 2.0356, "step": 45259 }, { "epoch": 3.5470219435736676, "grad_norm": 6.658332347869873, "learning_rate": 3.1432137210598624e-07, "loss": 1.5516, "step": 45260 }, { "epoch": 3.547100313479624, "grad_norm": 3.5268397331237793, "learning_rate": 3.1421377342443104e-07, "loss": 1.5832, "step": 45261 }, { "epoch": 3.54717868338558, "grad_norm": 2.964094400405884, "learning_rate": 3.141061925651978e-07, "loss": 1.3391, "step": 45262 }, { "epoch": 3.547257053291536, "grad_norm": 3.482224225997925, "learning_rate": 3.13998629528699e-07, "loss": 1.305, "step": 45263 }, { "epoch": 3.547335423197492, "grad_norm": 5.586470127105713, "learning_rate": 3.13891084315342e-07, "loss": 2.1644, "step": 45264 }, { "epoch": 3.5474137931034484, "grad_norm": 2.756549119949341, "learning_rate": 3.1378355692553763e-07, "loss": 1.3346, "step": 45265 }, { "epoch": 3.5474921630094043, "grad_norm": 3.9549193382263184, "learning_rate": 3.136760473596928e-07, "loss": 1.4977, "step": 45266 }, { "epoch": 3.5475705329153606, "grad_norm": 4.763926982879639, "learning_rate": 3.1356855561821607e-07, "loss": 1.4928, "step": 45267 }, { "epoch": 3.5476489028213165, "grad_norm": 2.941481351852417, "learning_rate": 3.1346108170151936e-07, "loss": 1.2642, "step": 45268 }, { "epoch": 3.547727272727273, "grad_norm": 2.1118392944335938, "learning_rate": 3.1335362561000894e-07, "loss": 1.497, "step": 45269 }, { "epoch": 3.5478056426332287, "grad_norm": 3.331711530685425, "learning_rate": 3.132461873440934e-07, "loss": 1.6501, "step": 45270 }, { "epoch": 3.547884012539185, "grad_norm": 4.616305351257324, "learning_rate": 3.13138766904183e-07, "loss": 1.6994, "step": 45271 }, { "epoch": 3.547962382445141, "grad_norm": 3.8541617393493652, "learning_rate": 3.1303136429068457e-07, "loss": 1.6122, "step": 45272 }, { "epoch": 3.5480407523510973, "grad_norm": 3.4597153663635254, "learning_rate": 3.129239795040079e-07, "loss": 1.3996, "step": 45273 }, { "epoch": 3.548119122257053, "grad_norm": 2.948922634124756, "learning_rate": 3.128166125445609e-07, "loss": 1.439, "step": 45274 }, { "epoch": 3.5481974921630095, "grad_norm": 4.107510566711426, "learning_rate": 3.127092634127521e-07, "loss": 0.947, "step": 45275 }, { "epoch": 3.5482758620689654, "grad_norm": 3.0484509468078613, "learning_rate": 3.126019321089896e-07, "loss": 1.4019, "step": 45276 }, { "epoch": 3.5483542319749217, "grad_norm": 3.9447991847991943, "learning_rate": 3.124946186336819e-07, "loss": 1.2485, "step": 45277 }, { "epoch": 3.5484326018808776, "grad_norm": 4.873709201812744, "learning_rate": 3.123873229872365e-07, "loss": 1.2004, "step": 45278 }, { "epoch": 3.548510971786834, "grad_norm": 3.3707668781280518, "learning_rate": 3.122800451700625e-07, "loss": 0.9512, "step": 45279 }, { "epoch": 3.54858934169279, "grad_norm": 2.560211658477783, "learning_rate": 3.1217278518256844e-07, "loss": 1.5865, "step": 45280 }, { "epoch": 3.548667711598746, "grad_norm": 5.558130264282227, "learning_rate": 3.1206554302515957e-07, "loss": 1.6345, "step": 45281 }, { "epoch": 3.548746081504702, "grad_norm": 10.308248519897461, "learning_rate": 3.119583186982461e-07, "loss": 1.8888, "step": 45282 }, { "epoch": 3.5488244514106584, "grad_norm": 4.811300754547119, "learning_rate": 3.118511122022361e-07, "loss": 1.4137, "step": 45283 }, { "epoch": 3.5489028213166143, "grad_norm": 4.514666557312012, "learning_rate": 3.1174392353753526e-07, "loss": 1.363, "step": 45284 }, { "epoch": 3.5489811912225706, "grad_norm": 5.5983710289001465, "learning_rate": 3.116367527045533e-07, "loss": 1.5856, "step": 45285 }, { "epoch": 3.5490595611285265, "grad_norm": 2.7011163234710693, "learning_rate": 3.1152959970369545e-07, "loss": 1.2891, "step": 45286 }, { "epoch": 3.549137931034483, "grad_norm": 3.1776175498962402, "learning_rate": 3.1142246453537306e-07, "loss": 0.9413, "step": 45287 }, { "epoch": 3.5492163009404387, "grad_norm": 3.317495822906494, "learning_rate": 3.113153471999902e-07, "loss": 1.3894, "step": 45288 }, { "epoch": 3.549294670846395, "grad_norm": 2.8459949493408203, "learning_rate": 3.1120824769795556e-07, "loss": 1.6548, "step": 45289 }, { "epoch": 3.549373040752351, "grad_norm": 6.880824565887451, "learning_rate": 3.11101166029677e-07, "loss": 0.7723, "step": 45290 }, { "epoch": 3.5494514106583073, "grad_norm": 2.6951160430908203, "learning_rate": 3.109941021955604e-07, "loss": 1.5953, "step": 45291 }, { "epoch": 3.549529780564263, "grad_norm": 2.849175453186035, "learning_rate": 3.1088705619601424e-07, "loss": 0.8427, "step": 45292 }, { "epoch": 3.5496081504702195, "grad_norm": 5.627843379974365, "learning_rate": 3.107800280314455e-07, "loss": 0.8273, "step": 45293 }, { "epoch": 3.5496865203761754, "grad_norm": 12.675212860107422, "learning_rate": 3.1067301770226157e-07, "loss": 1.1797, "step": 45294 }, { "epoch": 3.5497648902821317, "grad_norm": 4.096186637878418, "learning_rate": 3.1056602520886714e-07, "loss": 1.053, "step": 45295 }, { "epoch": 3.549843260188088, "grad_norm": 2.7238831520080566, "learning_rate": 3.104590505516719e-07, "loss": 1.3309, "step": 45296 }, { "epoch": 3.549921630094044, "grad_norm": 10.428666114807129, "learning_rate": 3.103520937310811e-07, "loss": 1.2154, "step": 45297 }, { "epoch": 3.55, "grad_norm": 2.337219715118408, "learning_rate": 3.1024515474750384e-07, "loss": 1.148, "step": 45298 }, { "epoch": 3.550078369905956, "grad_norm": 2.042973279953003, "learning_rate": 3.101382336013442e-07, "loss": 1.2447, "step": 45299 }, { "epoch": 3.5501567398119125, "grad_norm": 11.562960624694824, "learning_rate": 3.100313302930086e-07, "loss": 1.419, "step": 45300 }, { "epoch": 3.5502351097178684, "grad_norm": 2.6602532863616943, "learning_rate": 3.0992444482290605e-07, "loss": 1.4908, "step": 45301 }, { "epoch": 3.5503134796238243, "grad_norm": 2.7739224433898926, "learning_rate": 3.0981757719144247e-07, "loss": 1.317, "step": 45302 }, { "epoch": 3.5503918495297806, "grad_norm": 4.082329750061035, "learning_rate": 3.0971072739902354e-07, "loss": 1.2717, "step": 45303 }, { "epoch": 3.550470219435737, "grad_norm": 3.2887070178985596, "learning_rate": 3.096038954460556e-07, "loss": 1.258, "step": 45304 }, { "epoch": 3.550548589341693, "grad_norm": 3.113027334213257, "learning_rate": 3.094970813329445e-07, "loss": 1.1635, "step": 45305 }, { "epoch": 3.5506269592476487, "grad_norm": 2.4270529747009277, "learning_rate": 3.093902850600988e-07, "loss": 1.453, "step": 45306 }, { "epoch": 3.550705329153605, "grad_norm": 17.646472930908203, "learning_rate": 3.092835066279226e-07, "loss": 3.1294, "step": 45307 }, { "epoch": 3.5507836990595614, "grad_norm": 5.047422409057617, "learning_rate": 3.091767460368228e-07, "loss": 1.3459, "step": 45308 }, { "epoch": 3.5508620689655173, "grad_norm": 3.4372589588165283, "learning_rate": 3.090700032872046e-07, "loss": 1.2965, "step": 45309 }, { "epoch": 3.550940438871473, "grad_norm": 7.590747833251953, "learning_rate": 3.089632783794755e-07, "loss": 1.257, "step": 45310 }, { "epoch": 3.5510188087774295, "grad_norm": 3.9493086338043213, "learning_rate": 3.0885657131404014e-07, "loss": 1.3718, "step": 45311 }, { "epoch": 3.551097178683386, "grad_norm": 2.3694770336151123, "learning_rate": 3.0874988209130486e-07, "loss": 1.2038, "step": 45312 }, { "epoch": 3.5511755485893417, "grad_norm": 3.907193183898926, "learning_rate": 3.086432107116766e-07, "loss": 1.3774, "step": 45313 }, { "epoch": 3.5512539184952976, "grad_norm": 3.061567544937134, "learning_rate": 3.0853655717555773e-07, "loss": 0.9552, "step": 45314 }, { "epoch": 3.551332288401254, "grad_norm": 3.420872211456299, "learning_rate": 3.084299214833575e-07, "loss": 1.8019, "step": 45315 }, { "epoch": 3.5514106583072103, "grad_norm": 2.8641984462738037, "learning_rate": 3.083233036354799e-07, "loss": 1.626, "step": 45316 }, { "epoch": 3.551489028213166, "grad_norm": 3.979757070541382, "learning_rate": 3.0821670363233137e-07, "loss": 1.489, "step": 45317 }, { "epoch": 3.551567398119122, "grad_norm": 2.3248958587646484, "learning_rate": 3.0811012147431594e-07, "loss": 1.5131, "step": 45318 }, { "epoch": 3.5516457680250784, "grad_norm": 2.8689005374908447, "learning_rate": 3.080035571618389e-07, "loss": 0.7459, "step": 45319 }, { "epoch": 3.5517241379310347, "grad_norm": 2.7619566917419434, "learning_rate": 3.078970106953072e-07, "loss": 1.1704, "step": 45320 }, { "epoch": 3.5518025078369906, "grad_norm": 3.6271703243255615, "learning_rate": 3.07790482075126e-07, "loss": 1.4482, "step": 45321 }, { "epoch": 3.5518808777429465, "grad_norm": 2.8907041549682617, "learning_rate": 3.076839713016988e-07, "loss": 1.0072, "step": 45322 }, { "epoch": 3.551959247648903, "grad_norm": 3.5483102798461914, "learning_rate": 3.0757747837543207e-07, "loss": 1.785, "step": 45323 }, { "epoch": 3.552037617554859, "grad_norm": 3.6638076305389404, "learning_rate": 3.0747100329672985e-07, "loss": 1.333, "step": 45324 }, { "epoch": 3.552115987460815, "grad_norm": 3.4733083248138428, "learning_rate": 3.07364546065998e-07, "loss": 1.2571, "step": 45325 }, { "epoch": 3.552194357366771, "grad_norm": 2.7336978912353516, "learning_rate": 3.072581066836411e-07, "loss": 1.5309, "step": 45326 }, { "epoch": 3.5522727272727272, "grad_norm": 4.374774932861328, "learning_rate": 3.0715168515006445e-07, "loss": 1.4708, "step": 45327 }, { "epoch": 3.5523510971786836, "grad_norm": 3.8527534008026123, "learning_rate": 3.070452814656716e-07, "loss": 1.3053, "step": 45328 }, { "epoch": 3.5524294670846395, "grad_norm": 2.9219393730163574, "learning_rate": 3.0693889563086885e-07, "loss": 1.3788, "step": 45329 }, { "epoch": 3.5525078369905954, "grad_norm": 3.2798726558685303, "learning_rate": 3.068325276460593e-07, "loss": 1.1073, "step": 45330 }, { "epoch": 3.5525862068965517, "grad_norm": 2.962413787841797, "learning_rate": 3.0672617751164866e-07, "loss": 1.3012, "step": 45331 }, { "epoch": 3.552664576802508, "grad_norm": 13.353821754455566, "learning_rate": 3.066198452280411e-07, "loss": 1.7673, "step": 45332 }, { "epoch": 3.552742946708464, "grad_norm": 3.618468761444092, "learning_rate": 3.065135307956407e-07, "loss": 1.3187, "step": 45333 }, { "epoch": 3.55282131661442, "grad_norm": 4.807529449462891, "learning_rate": 3.064072342148522e-07, "loss": 1.5708, "step": 45334 }, { "epoch": 3.552899686520376, "grad_norm": 2.341665506362915, "learning_rate": 3.0630095548607966e-07, "loss": 1.672, "step": 45335 }, { "epoch": 3.5529780564263325, "grad_norm": 3.3005595207214355, "learning_rate": 3.061946946097283e-07, "loss": 1.3518, "step": 45336 }, { "epoch": 3.5530564263322884, "grad_norm": 7.286617755889893, "learning_rate": 3.0608845158620006e-07, "loss": 1.4406, "step": 45337 }, { "epoch": 3.5531347962382442, "grad_norm": 6.5932297706604, "learning_rate": 3.0598222641590124e-07, "loss": 1.4481, "step": 45338 }, { "epoch": 3.5532131661442006, "grad_norm": 5.380032062530518, "learning_rate": 3.0587601909923546e-07, "loss": 1.3462, "step": 45339 }, { "epoch": 3.553291536050157, "grad_norm": 2.588254690170288, "learning_rate": 3.0576982963660575e-07, "loss": 1.0823, "step": 45340 }, { "epoch": 3.553369905956113, "grad_norm": 4.842529296875, "learning_rate": 3.0566365802841667e-07, "loss": 1.5334, "step": 45341 }, { "epoch": 3.553448275862069, "grad_norm": 3.421712875366211, "learning_rate": 3.0555750427507135e-07, "loss": 0.5961, "step": 45342 }, { "epoch": 3.553526645768025, "grad_norm": 3.5020389556884766, "learning_rate": 3.054513683769744e-07, "loss": 1.111, "step": 45343 }, { "epoch": 3.5536050156739813, "grad_norm": 2.7740838527679443, "learning_rate": 3.053452503345289e-07, "loss": 1.4067, "step": 45344 }, { "epoch": 3.5536833855799372, "grad_norm": 7.493526458740234, "learning_rate": 3.0523915014813885e-07, "loss": 1.0923, "step": 45345 }, { "epoch": 3.5537617554858936, "grad_norm": 3.9144949913024902, "learning_rate": 3.0513306781820793e-07, "loss": 1.8335, "step": 45346 }, { "epoch": 3.5538401253918495, "grad_norm": 2.8026750087738037, "learning_rate": 3.0502700334513915e-07, "loss": 1.5746, "step": 45347 }, { "epoch": 3.553918495297806, "grad_norm": 3.8392114639282227, "learning_rate": 3.0492095672933654e-07, "loss": 1.5135, "step": 45348 }, { "epoch": 3.5539968652037617, "grad_norm": 2.783942937850952, "learning_rate": 3.048149279712026e-07, "loss": 1.2806, "step": 45349 }, { "epoch": 3.554075235109718, "grad_norm": 3.3756916522979736, "learning_rate": 3.0470891707114205e-07, "loss": 1.1745, "step": 45350 }, { "epoch": 3.554153605015674, "grad_norm": 2.528186082839966, "learning_rate": 3.046029240295556e-07, "loss": 1.3561, "step": 45351 }, { "epoch": 3.5542319749216302, "grad_norm": 3.033468246459961, "learning_rate": 3.0449694884684856e-07, "loss": 1.2914, "step": 45352 }, { "epoch": 3.554310344827586, "grad_norm": 6.109613418579102, "learning_rate": 3.043909915234233e-07, "loss": 1.2053, "step": 45353 }, { "epoch": 3.5543887147335425, "grad_norm": 2.6811113357543945, "learning_rate": 3.0428505205968404e-07, "loss": 0.811, "step": 45354 }, { "epoch": 3.5544670846394983, "grad_norm": 2.3274526596069336, "learning_rate": 3.041791304560315e-07, "loss": 1.2356, "step": 45355 }, { "epoch": 3.5545454545454547, "grad_norm": 3.8666207790374756, "learning_rate": 3.0407322671286875e-07, "loss": 0.8497, "step": 45356 }, { "epoch": 3.5546238244514106, "grad_norm": 2.724705934524536, "learning_rate": 3.0396734083060043e-07, "loss": 1.3058, "step": 45357 }, { "epoch": 3.554702194357367, "grad_norm": 4.406623840332031, "learning_rate": 3.0386147280962894e-07, "loss": 1.229, "step": 45358 }, { "epoch": 3.554780564263323, "grad_norm": 2.959562301635742, "learning_rate": 3.037556226503557e-07, "loss": 1.4049, "step": 45359 }, { "epoch": 3.554858934169279, "grad_norm": 3.432793617248535, "learning_rate": 3.0364979035318376e-07, "loss": 1.3937, "step": 45360 }, { "epoch": 3.554937304075235, "grad_norm": 3.1012368202209473, "learning_rate": 3.035439759185155e-07, "loss": 1.4294, "step": 45361 }, { "epoch": 3.5550156739811913, "grad_norm": 3.7615160942077637, "learning_rate": 3.03438179346755e-07, "loss": 1.4775, "step": 45362 }, { "epoch": 3.5550940438871472, "grad_norm": 3.0904412269592285, "learning_rate": 3.0333240063830203e-07, "loss": 1.4381, "step": 45363 }, { "epoch": 3.5551724137931036, "grad_norm": 3.041782855987549, "learning_rate": 3.032266397935607e-07, "loss": 1.2221, "step": 45364 }, { "epoch": 3.5552507836990594, "grad_norm": 2.9333407878875732, "learning_rate": 3.031208968129329e-07, "loss": 1.5684, "step": 45365 }, { "epoch": 3.555329153605016, "grad_norm": 4.641325950622559, "learning_rate": 3.0301517169682105e-07, "loss": 1.1337, "step": 45366 }, { "epoch": 3.5554075235109717, "grad_norm": 2.4976141452789307, "learning_rate": 3.0290946444562654e-07, "loss": 1.763, "step": 45367 }, { "epoch": 3.555485893416928, "grad_norm": 3.522817611694336, "learning_rate": 3.028037750597518e-07, "loss": 1.3409, "step": 45368 }, { "epoch": 3.555564263322884, "grad_norm": 2.3548762798309326, "learning_rate": 3.0269810353959983e-07, "loss": 1.2257, "step": 45369 }, { "epoch": 3.55564263322884, "grad_norm": 3.0640320777893066, "learning_rate": 3.0259244988556977e-07, "loss": 1.2044, "step": 45370 }, { "epoch": 3.555721003134796, "grad_norm": 2.455376625061035, "learning_rate": 3.0248681409806634e-07, "loss": 1.065, "step": 45371 }, { "epoch": 3.5557993730407524, "grad_norm": 3.5183215141296387, "learning_rate": 3.023811961774903e-07, "loss": 1.5498, "step": 45372 }, { "epoch": 3.5558777429467083, "grad_norm": 11.705978393554688, "learning_rate": 3.0227559612424405e-07, "loss": 1.158, "step": 45373 }, { "epoch": 3.5559561128526647, "grad_norm": 6.212865829467773, "learning_rate": 3.0217001393872734e-07, "loss": 1.6286, "step": 45374 }, { "epoch": 3.5560344827586206, "grad_norm": 4.366468906402588, "learning_rate": 3.0206444962134264e-07, "loss": 1.3871, "step": 45375 }, { "epoch": 3.556112852664577, "grad_norm": 4.054891109466553, "learning_rate": 3.019589031724923e-07, "loss": 1.5178, "step": 45376 }, { "epoch": 3.5561912225705328, "grad_norm": 2.9110095500946045, "learning_rate": 3.0185337459257835e-07, "loss": 1.6092, "step": 45377 }, { "epoch": 3.556269592476489, "grad_norm": 3.726714849472046, "learning_rate": 3.0174786388199983e-07, "loss": 1.994, "step": 45378 }, { "epoch": 3.556347962382445, "grad_norm": 3.8784191608428955, "learning_rate": 3.016423710411598e-07, "loss": 1.4944, "step": 45379 }, { "epoch": 3.5564263322884013, "grad_norm": 5.830557823181152, "learning_rate": 3.015368960704584e-07, "loss": 1.5379, "step": 45380 }, { "epoch": 3.556504702194357, "grad_norm": 6.817470073699951, "learning_rate": 3.014314389702977e-07, "loss": 0.9314, "step": 45381 }, { "epoch": 3.5565830721003135, "grad_norm": 5.507091522216797, "learning_rate": 3.0132599974107834e-07, "loss": 1.674, "step": 45382 }, { "epoch": 3.5566614420062694, "grad_norm": 1.8962458372116089, "learning_rate": 3.0122057838320117e-07, "loss": 1.2988, "step": 45383 }, { "epoch": 3.5567398119122258, "grad_norm": 2.9623935222625732, "learning_rate": 3.011151748970681e-07, "loss": 0.7454, "step": 45384 }, { "epoch": 3.5568181818181817, "grad_norm": 2.206106424331665, "learning_rate": 3.010097892830788e-07, "loss": 0.7971, "step": 45385 }, { "epoch": 3.556896551724138, "grad_norm": 2.2275798320770264, "learning_rate": 3.009044215416351e-07, "loss": 1.0393, "step": 45386 }, { "epoch": 3.556974921630094, "grad_norm": 3.177866220474243, "learning_rate": 3.0079907167313736e-07, "loss": 1.2332, "step": 45387 }, { "epoch": 3.55705329153605, "grad_norm": 6.09829044342041, "learning_rate": 3.0069373967798686e-07, "loss": 1.2217, "step": 45388 }, { "epoch": 3.557131661442006, "grad_norm": 3.6374781131744385, "learning_rate": 3.0058842555658273e-07, "loss": 1.1585, "step": 45389 }, { "epoch": 3.5572100313479624, "grad_norm": 4.362119197845459, "learning_rate": 3.0048312930932634e-07, "loss": 1.0696, "step": 45390 }, { "epoch": 3.5572884012539188, "grad_norm": 5.852144718170166, "learning_rate": 3.003778509366195e-07, "loss": 1.2956, "step": 45391 }, { "epoch": 3.5573667711598747, "grad_norm": 3.6345200538635254, "learning_rate": 3.002725904388609e-07, "loss": 1.4621, "step": 45392 }, { "epoch": 3.5574451410658305, "grad_norm": 4.0125813484191895, "learning_rate": 3.001673478164513e-07, "loss": 1.7256, "step": 45393 }, { "epoch": 3.557523510971787, "grad_norm": 6.1391191482543945, "learning_rate": 3.0006212306978977e-07, "loss": 1.5485, "step": 45394 }, { "epoch": 3.557601880877743, "grad_norm": 5.9522385597229, "learning_rate": 2.9995691619927994e-07, "loss": 1.2342, "step": 45395 }, { "epoch": 3.557680250783699, "grad_norm": 3.4346656799316406, "learning_rate": 2.998517272053192e-07, "loss": 1.5499, "step": 45396 }, { "epoch": 3.557758620689655, "grad_norm": 5.925712585449219, "learning_rate": 2.997465560883084e-07, "loss": 1.7557, "step": 45397 }, { "epoch": 3.5578369905956113, "grad_norm": 3.0018138885498047, "learning_rate": 2.996414028486472e-07, "loss": 1.129, "step": 45398 }, { "epoch": 3.5579153605015676, "grad_norm": 4.56493616104126, "learning_rate": 2.995362674867358e-07, "loss": 1.2157, "step": 45399 }, { "epoch": 3.5579937304075235, "grad_norm": 3.6248934268951416, "learning_rate": 2.9943115000297453e-07, "loss": 1.1041, "step": 45400 }, { "epoch": 3.5580721003134794, "grad_norm": 4.12454891204834, "learning_rate": 2.993260503977624e-07, "loss": 1.5968, "step": 45401 }, { "epoch": 3.5581504702194358, "grad_norm": 3.5557215213775635, "learning_rate": 2.9922096867149973e-07, "loss": 1.5372, "step": 45402 }, { "epoch": 3.558228840125392, "grad_norm": 4.059266567230225, "learning_rate": 2.991159048245862e-07, "loss": 1.3342, "step": 45403 }, { "epoch": 3.558307210031348, "grad_norm": 6.63552713394165, "learning_rate": 2.9901085885742087e-07, "loss": 1.3073, "step": 45404 }, { "epoch": 3.558385579937304, "grad_norm": 5.94993257522583, "learning_rate": 2.9890583077040403e-07, "loss": 1.4668, "step": 45405 }, { "epoch": 3.55846394984326, "grad_norm": 2.765773296356201, "learning_rate": 2.9880082056393533e-07, "loss": 1.3428, "step": 45406 }, { "epoch": 3.5585423197492165, "grad_norm": 2.8991994857788086, "learning_rate": 2.9869582823841283e-07, "loss": 1.3098, "step": 45407 }, { "epoch": 3.5586206896551724, "grad_norm": 5.909121513366699, "learning_rate": 2.985908537942361e-07, "loss": 1.7042, "step": 45408 }, { "epoch": 3.5586990595611283, "grad_norm": 3.5582385063171387, "learning_rate": 2.9848589723180556e-07, "loss": 1.285, "step": 45409 }, { "epoch": 3.5587774294670846, "grad_norm": 6.515000820159912, "learning_rate": 2.983809585515202e-07, "loss": 1.2798, "step": 45410 }, { "epoch": 3.558855799373041, "grad_norm": 4.445518970489502, "learning_rate": 2.9827603775377856e-07, "loss": 1.4146, "step": 45411 }, { "epoch": 3.558934169278997, "grad_norm": 3.4986233711242676, "learning_rate": 2.9817113483897985e-07, "loss": 1.2229, "step": 45412 }, { "epoch": 3.5590125391849528, "grad_norm": 2.713650941848755, "learning_rate": 2.9806624980752206e-07, "loss": 1.4345, "step": 45413 }, { "epoch": 3.559090909090909, "grad_norm": 3.643754243850708, "learning_rate": 2.979613826598071e-07, "loss": 1.6661, "step": 45414 }, { "epoch": 3.5591692789968654, "grad_norm": 2.9163012504577637, "learning_rate": 2.9785653339623075e-07, "loss": 1.3821, "step": 45415 }, { "epoch": 3.5592476489028213, "grad_norm": 4.884432315826416, "learning_rate": 2.977517020171933e-07, "loss": 0.9528, "step": 45416 }, { "epoch": 3.559326018808777, "grad_norm": 3.8480818271636963, "learning_rate": 2.976468885230932e-07, "loss": 1.5827, "step": 45417 }, { "epoch": 3.5594043887147335, "grad_norm": 2.678712844848633, "learning_rate": 2.975420929143291e-07, "loss": 1.3484, "step": 45418 }, { "epoch": 3.55948275862069, "grad_norm": 3.2367196083068848, "learning_rate": 2.9743731519129905e-07, "loss": 1.2122, "step": 45419 }, { "epoch": 3.5595611285266457, "grad_norm": 4.186896324157715, "learning_rate": 2.9733255535440264e-07, "loss": 1.6011, "step": 45420 }, { "epoch": 3.5596394984326016, "grad_norm": 2.9302618503570557, "learning_rate": 2.972278134040379e-07, "loss": 1.3759, "step": 45421 }, { "epoch": 3.559717868338558, "grad_norm": 2.0989580154418945, "learning_rate": 2.9712308934060183e-07, "loss": 1.3536, "step": 45422 }, { "epoch": 3.5597962382445143, "grad_norm": 3.6936423778533936, "learning_rate": 2.970183831644952e-07, "loss": 1.1927, "step": 45423 }, { "epoch": 3.55987460815047, "grad_norm": 2.5286805629730225, "learning_rate": 2.969136948761142e-07, "loss": 1.3733, "step": 45424 }, { "epoch": 3.559952978056426, "grad_norm": 6.076056480407715, "learning_rate": 2.968090244758592e-07, "loss": 1.2396, "step": 45425 }, { "epoch": 3.5600313479623824, "grad_norm": 2.4189140796661377, "learning_rate": 2.9670437196412607e-07, "loss": 1.1542, "step": 45426 }, { "epoch": 3.5601097178683387, "grad_norm": 2.935065746307373, "learning_rate": 2.9659973734131266e-07, "loss": 1.5545, "step": 45427 }, { "epoch": 3.5601880877742946, "grad_norm": 5.737148284912109, "learning_rate": 2.9649512060781927e-07, "loss": 0.6535, "step": 45428 }, { "epoch": 3.5602664576802505, "grad_norm": 2.97430682182312, "learning_rate": 2.9639052176404337e-07, "loss": 1.1334, "step": 45429 }, { "epoch": 3.560344827586207, "grad_norm": 2.4451794624328613, "learning_rate": 2.962859408103808e-07, "loss": 1.5056, "step": 45430 }, { "epoch": 3.560423197492163, "grad_norm": 2.5587081909179688, "learning_rate": 2.9618137774723056e-07, "loss": 1.2468, "step": 45431 }, { "epoch": 3.560501567398119, "grad_norm": 3.794625759124756, "learning_rate": 2.9607683257498967e-07, "loss": 1.2004, "step": 45432 }, { "epoch": 3.560579937304075, "grad_norm": 4.039191246032715, "learning_rate": 2.959723052940583e-07, "loss": 1.6773, "step": 45433 }, { "epoch": 3.5606583072100313, "grad_norm": 3.2310357093811035, "learning_rate": 2.9586779590483115e-07, "loss": 1.6647, "step": 45434 }, { "epoch": 3.5607366771159876, "grad_norm": 2.642979145050049, "learning_rate": 2.957633044077063e-07, "loss": 1.4163, "step": 45435 }, { "epoch": 3.5608150470219435, "grad_norm": 4.229982376098633, "learning_rate": 2.9565883080308224e-07, "loss": 1.3678, "step": 45436 }, { "epoch": 3.5608934169279, "grad_norm": 3.2789735794067383, "learning_rate": 2.955543750913553e-07, "loss": 1.3154, "step": 45437 }, { "epoch": 3.5609717868338557, "grad_norm": 5.299867630004883, "learning_rate": 2.95449937272923e-07, "loss": 1.2066, "step": 45438 }, { "epoch": 3.561050156739812, "grad_norm": 12.23051643371582, "learning_rate": 2.9534551734818337e-07, "loss": 0.3145, "step": 45439 }, { "epoch": 3.561128526645768, "grad_norm": 9.868215560913086, "learning_rate": 2.952411153175322e-07, "loss": 1.6842, "step": 45440 }, { "epoch": 3.5612068965517243, "grad_norm": 4.722557544708252, "learning_rate": 2.95136731181368e-07, "loss": 1.3956, "step": 45441 }, { "epoch": 3.56128526645768, "grad_norm": 3.411792516708374, "learning_rate": 2.9503236494008657e-07, "loss": 1.3528, "step": 45442 }, { "epoch": 3.5613636363636365, "grad_norm": 4.261519432067871, "learning_rate": 2.949280165940854e-07, "loss": 1.5049, "step": 45443 }, { "epoch": 3.5614420062695924, "grad_norm": 3.270859956741333, "learning_rate": 2.948236861437626e-07, "loss": 1.3382, "step": 45444 }, { "epoch": 3.5615203761755487, "grad_norm": 2.4980568885803223, "learning_rate": 2.947193735895115e-07, "loss": 0.8662, "step": 45445 }, { "epoch": 3.5615987460815046, "grad_norm": 3.7527365684509277, "learning_rate": 2.946150789317326e-07, "loss": 1.0258, "step": 45446 }, { "epoch": 3.561677115987461, "grad_norm": 4.75184440612793, "learning_rate": 2.945108021708215e-07, "loss": 1.4562, "step": 45447 }, { "epoch": 3.561755485893417, "grad_norm": 4.0692949295043945, "learning_rate": 2.9440654330717354e-07, "loss": 1.3297, "step": 45448 }, { "epoch": 3.561833855799373, "grad_norm": 4.6774821281433105, "learning_rate": 2.9430230234118674e-07, "loss": 1.5425, "step": 45449 }, { "epoch": 3.561912225705329, "grad_norm": 2.566434144973755, "learning_rate": 2.9419807927325573e-07, "loss": 1.3021, "step": 45450 }, { "epoch": 3.5619905956112854, "grad_norm": 4.302579402923584, "learning_rate": 2.940938741037797e-07, "loss": 1.7337, "step": 45451 }, { "epoch": 3.5620689655172413, "grad_norm": 5.265718460083008, "learning_rate": 2.9398968683315264e-07, "loss": 1.2685, "step": 45452 }, { "epoch": 3.5621473354231976, "grad_norm": 4.809969902038574, "learning_rate": 2.9388551746177217e-07, "loss": 1.2147, "step": 45453 }, { "epoch": 3.5622257053291535, "grad_norm": 2.260457754135132, "learning_rate": 2.937813659900335e-07, "loss": 1.2504, "step": 45454 }, { "epoch": 3.56230407523511, "grad_norm": 6.573205471038818, "learning_rate": 2.9367723241833346e-07, "loss": 1.4555, "step": 45455 }, { "epoch": 3.5623824451410657, "grad_norm": 5.457258224487305, "learning_rate": 2.935731167470679e-07, "loss": 1.4361, "step": 45456 }, { "epoch": 3.562460815047022, "grad_norm": 5.72974967956543, "learning_rate": 2.934690189766326e-07, "loss": 1.4619, "step": 45457 }, { "epoch": 3.562539184952978, "grad_norm": 3.029201030731201, "learning_rate": 2.933649391074239e-07, "loss": 1.3526, "step": 45458 }, { "epoch": 3.5626175548589343, "grad_norm": 3.2480528354644775, "learning_rate": 2.9326087713983765e-07, "loss": 1.1528, "step": 45459 }, { "epoch": 3.56269592476489, "grad_norm": 2.842674732208252, "learning_rate": 2.93156833074269e-07, "loss": 1.5197, "step": 45460 }, { "epoch": 3.5627742946708465, "grad_norm": 3.266176700592041, "learning_rate": 2.930528069111144e-07, "loss": 1.0881, "step": 45461 }, { "epoch": 3.5628526645768024, "grad_norm": 2.7245564460754395, "learning_rate": 2.929487986507701e-07, "loss": 1.3049, "step": 45462 }, { "epoch": 3.5629310344827587, "grad_norm": 3.338512420654297, "learning_rate": 2.9284480829363026e-07, "loss": 1.1958, "step": 45463 }, { "epoch": 3.5630094043887146, "grad_norm": 3.1408283710479736, "learning_rate": 2.9274083584009016e-07, "loss": 1.561, "step": 45464 }, { "epoch": 3.563087774294671, "grad_norm": 4.120311737060547, "learning_rate": 2.9263688129054667e-07, "loss": 1.1198, "step": 45465 }, { "epoch": 3.563166144200627, "grad_norm": 2.8505020141601562, "learning_rate": 2.925329446453956e-07, "loss": 1.2531, "step": 45466 }, { "epoch": 3.563244514106583, "grad_norm": 6.260836124420166, "learning_rate": 2.9242902590503055e-07, "loss": 1.4654, "step": 45467 }, { "epoch": 3.563322884012539, "grad_norm": 2.593409776687622, "learning_rate": 2.923251250698478e-07, "loss": 1.7746, "step": 45468 }, { "epoch": 3.5634012539184954, "grad_norm": 2.293590545654297, "learning_rate": 2.9222124214024093e-07, "loss": 1.4736, "step": 45469 }, { "epoch": 3.5634796238244513, "grad_norm": 9.191863059997559, "learning_rate": 2.92117377116608e-07, "loss": 0.9151, "step": 45470 }, { "epoch": 3.5635579937304076, "grad_norm": 2.572200298309326, "learning_rate": 2.920135299993421e-07, "loss": 1.4935, "step": 45471 }, { "epoch": 3.5636363636363635, "grad_norm": 2.535689353942871, "learning_rate": 2.919097007888383e-07, "loss": 0.7718, "step": 45472 }, { "epoch": 3.56371473354232, "grad_norm": 2.9667000770568848, "learning_rate": 2.9180588948549195e-07, "loss": 1.3487, "step": 45473 }, { "epoch": 3.5637931034482757, "grad_norm": 3.4368627071380615, "learning_rate": 2.917020960896977e-07, "loss": 1.4431, "step": 45474 }, { "epoch": 3.563871473354232, "grad_norm": 2.82570481300354, "learning_rate": 2.9159832060184966e-07, "loss": 1.2574, "step": 45475 }, { "epoch": 3.563949843260188, "grad_norm": 3.5939760208129883, "learning_rate": 2.9149456302234426e-07, "loss": 0.872, "step": 45476 }, { "epoch": 3.5640282131661443, "grad_norm": 2.899388313293457, "learning_rate": 2.91390823351575e-07, "loss": 1.0123, "step": 45477 }, { "epoch": 3.5641065830721, "grad_norm": 5.969104290008545, "learning_rate": 2.9128710158993545e-07, "loss": 1.3779, "step": 45478 }, { "epoch": 3.5641849529780565, "grad_norm": 8.388004302978516, "learning_rate": 2.91183397737822e-07, "loss": 1.5678, "step": 45479 }, { "epoch": 3.5642633228840124, "grad_norm": 3.5832152366638184, "learning_rate": 2.910797117956282e-07, "loss": 1.432, "step": 45480 }, { "epoch": 3.5643416927899687, "grad_norm": 3.7959814071655273, "learning_rate": 2.909760437637493e-07, "loss": 1.1366, "step": 45481 }, { "epoch": 3.564420062695925, "grad_norm": 4.875190734863281, "learning_rate": 2.908723936425778e-07, "loss": 0.8815, "step": 45482 }, { "epoch": 3.564498432601881, "grad_norm": 2.70481276512146, "learning_rate": 2.9076876143250885e-07, "loss": 1.4528, "step": 45483 }, { "epoch": 3.564576802507837, "grad_norm": 7.580806732177734, "learning_rate": 2.906651471339367e-07, "loss": 1.3897, "step": 45484 }, { "epoch": 3.564655172413793, "grad_norm": 3.727327346801758, "learning_rate": 2.90561550747257e-07, "loss": 1.6316, "step": 45485 }, { "epoch": 3.5647335423197495, "grad_norm": 2.740844964981079, "learning_rate": 2.904579722728607e-07, "loss": 1.2878, "step": 45486 }, { "epoch": 3.5648119122257054, "grad_norm": 2.9298102855682373, "learning_rate": 2.90354411711144e-07, "loss": 1.2072, "step": 45487 }, { "epoch": 3.5648902821316613, "grad_norm": 5.482611656188965, "learning_rate": 2.902508690624989e-07, "loss": 1.3077, "step": 45488 }, { "epoch": 3.5649686520376176, "grad_norm": 3.052316188812256, "learning_rate": 2.9014734432732226e-07, "loss": 1.4011, "step": 45489 }, { "epoch": 3.565047021943574, "grad_norm": 3.2821261882781982, "learning_rate": 2.9004383750600495e-07, "loss": 1.1805, "step": 45490 }, { "epoch": 3.56512539184953, "grad_norm": 2.1967077255249023, "learning_rate": 2.899403485989416e-07, "loss": 1.3523, "step": 45491 }, { "epoch": 3.5652037617554857, "grad_norm": 3.490105628967285, "learning_rate": 2.8983687760652637e-07, "loss": 1.525, "step": 45492 }, { "epoch": 3.565282131661442, "grad_norm": 4.301681041717529, "learning_rate": 2.8973342452915167e-07, "loss": 1.8351, "step": 45493 }, { "epoch": 3.5653605015673984, "grad_norm": 4.930333137512207, "learning_rate": 2.8962998936721223e-07, "loss": 1.3296, "step": 45494 }, { "epoch": 3.5654388714733543, "grad_norm": 7.935943603515625, "learning_rate": 2.895265721211005e-07, "loss": 1.781, "step": 45495 }, { "epoch": 3.56551724137931, "grad_norm": 3.6775951385498047, "learning_rate": 2.894231727912117e-07, "loss": 1.0464, "step": 45496 }, { "epoch": 3.5655956112852665, "grad_norm": 2.8732900619506836, "learning_rate": 2.8931979137793555e-07, "loss": 1.1991, "step": 45497 }, { "epoch": 3.565673981191223, "grad_norm": 2.7865428924560547, "learning_rate": 2.892164278816684e-07, "loss": 1.5014, "step": 45498 }, { "epoch": 3.5657523510971787, "grad_norm": 7.784970760345459, "learning_rate": 2.891130823028021e-07, "loss": 0.9246, "step": 45499 }, { "epoch": 3.5658307210031346, "grad_norm": 4.210665702819824, "learning_rate": 2.890097546417309e-07, "loss": 1.5833, "step": 45500 }, { "epoch": 3.565909090909091, "grad_norm": 2.3454527854919434, "learning_rate": 2.88906444898846e-07, "loss": 1.5655, "step": 45501 }, { "epoch": 3.5659874608150472, "grad_norm": 4.530806064605713, "learning_rate": 2.8880315307454054e-07, "loss": 1.6442, "step": 45502 }, { "epoch": 3.566065830721003, "grad_norm": 3.6016647815704346, "learning_rate": 2.8869987916920915e-07, "loss": 1.3433, "step": 45503 }, { "epoch": 3.566144200626959, "grad_norm": 3.5827646255493164, "learning_rate": 2.885966231832432e-07, "loss": 1.2431, "step": 45504 }, { "epoch": 3.5662225705329154, "grad_norm": 6.93105936050415, "learning_rate": 2.884933851170357e-07, "loss": 1.8202, "step": 45505 }, { "epoch": 3.5663009404388717, "grad_norm": 5.062239646911621, "learning_rate": 2.883901649709797e-07, "loss": 1.0907, "step": 45506 }, { "epoch": 3.5663793103448276, "grad_norm": 4.7693023681640625, "learning_rate": 2.8828696274546707e-07, "loss": 1.2419, "step": 45507 }, { "epoch": 3.5664576802507835, "grad_norm": 4.156249046325684, "learning_rate": 2.8818377844089084e-07, "loss": 0.9747, "step": 45508 }, { "epoch": 3.56653605015674, "grad_norm": 2.823519229888916, "learning_rate": 2.8808061205764347e-07, "loss": 1.3594, "step": 45509 }, { "epoch": 3.566614420062696, "grad_norm": 3.4368059635162354, "learning_rate": 2.8797746359611746e-07, "loss": 1.4299, "step": 45510 }, { "epoch": 3.566692789968652, "grad_norm": 2.8343687057495117, "learning_rate": 2.8787433305670466e-07, "loss": 1.3872, "step": 45511 }, { "epoch": 3.566771159874608, "grad_norm": 2.0971102714538574, "learning_rate": 2.877712204397975e-07, "loss": 1.2482, "step": 45512 }, { "epoch": 3.5668495297805642, "grad_norm": 2.8339836597442627, "learning_rate": 2.8766812574578864e-07, "loss": 1.2704, "step": 45513 }, { "epoch": 3.5669278996865206, "grad_norm": 4.876237869262695, "learning_rate": 2.875650489750692e-07, "loss": 1.6186, "step": 45514 }, { "epoch": 3.5670062695924765, "grad_norm": 4.258709907531738, "learning_rate": 2.8746199012803287e-07, "loss": 1.3332, "step": 45515 }, { "epoch": 3.5670846394984324, "grad_norm": 2.7445530891418457, "learning_rate": 2.873589492050693e-07, "loss": 1.2427, "step": 45516 }, { "epoch": 3.5671630094043887, "grad_norm": 16.257041931152344, "learning_rate": 2.8725592620657274e-07, "loss": 0.9024, "step": 45517 }, { "epoch": 3.567241379310345, "grad_norm": 3.8172268867492676, "learning_rate": 2.8715292113293434e-07, "loss": 1.3446, "step": 45518 }, { "epoch": 3.567319749216301, "grad_norm": 2.548241376876831, "learning_rate": 2.870499339845445e-07, "loss": 1.3335, "step": 45519 }, { "epoch": 3.567398119122257, "grad_norm": 3.629481792449951, "learning_rate": 2.869469647617967e-07, "loss": 1.0766, "step": 45520 }, { "epoch": 3.567476489028213, "grad_norm": 6.35253381729126, "learning_rate": 2.8684401346508074e-07, "loss": 1.0826, "step": 45521 }, { "epoch": 3.5675548589341695, "grad_norm": 20.45747947692871, "learning_rate": 2.8674108009479065e-07, "loss": 1.0853, "step": 45522 }, { "epoch": 3.5676332288401253, "grad_norm": 3.3834381103515625, "learning_rate": 2.866381646513161e-07, "loss": 1.2931, "step": 45523 }, { "epoch": 3.5677115987460812, "grad_norm": 8.880813598632812, "learning_rate": 2.865352671350491e-07, "loss": 1.521, "step": 45524 }, { "epoch": 3.5677899686520376, "grad_norm": 4.148098468780518, "learning_rate": 2.864323875463809e-07, "loss": 1.5546, "step": 45525 }, { "epoch": 3.567868338557994, "grad_norm": 3.4801695346832275, "learning_rate": 2.8632952588570293e-07, "loss": 1.7187, "step": 45526 }, { "epoch": 3.56794670846395, "grad_norm": 2.4906771183013916, "learning_rate": 2.86226682153406e-07, "loss": 1.5465, "step": 45527 }, { "epoch": 3.568025078369906, "grad_norm": 3.268846273422241, "learning_rate": 2.861238563498819e-07, "loss": 1.4404, "step": 45528 }, { "epoch": 3.568103448275862, "grad_norm": 3.3154256343841553, "learning_rate": 2.860210484755216e-07, "loss": 1.4384, "step": 45529 }, { "epoch": 3.5681818181818183, "grad_norm": 2.3756325244903564, "learning_rate": 2.8591825853071574e-07, "loss": 1.1223, "step": 45530 }, { "epoch": 3.5682601880877742, "grad_norm": 3.6633236408233643, "learning_rate": 2.858154865158552e-07, "loss": 1.4448, "step": 45531 }, { "epoch": 3.5683385579937306, "grad_norm": 5.064113140106201, "learning_rate": 2.8571273243133193e-07, "loss": 1.206, "step": 45532 }, { "epoch": 3.5684169278996865, "grad_norm": 5.0505499839782715, "learning_rate": 2.856099962775361e-07, "loss": 1.2301, "step": 45533 }, { "epoch": 3.568495297805643, "grad_norm": 3.369608163833618, "learning_rate": 2.8550727805485734e-07, "loss": 1.5303, "step": 45534 }, { "epoch": 3.5685736677115987, "grad_norm": 4.213177680969238, "learning_rate": 2.854045777636877e-07, "loss": 0.5225, "step": 45535 }, { "epoch": 3.568652037617555, "grad_norm": 4.4984564781188965, "learning_rate": 2.853018954044179e-07, "loss": 1.2905, "step": 45536 }, { "epoch": 3.568730407523511, "grad_norm": 3.0068812370300293, "learning_rate": 2.8519923097743875e-07, "loss": 1.1465, "step": 45537 }, { "epoch": 3.5688087774294672, "grad_norm": 2.6217498779296875, "learning_rate": 2.8509658448313936e-07, "loss": 1.5647, "step": 45538 }, { "epoch": 3.568887147335423, "grad_norm": 3.4303195476531982, "learning_rate": 2.8499395592191057e-07, "loss": 1.5828, "step": 45539 }, { "epoch": 3.5689655172413794, "grad_norm": 3.043581008911133, "learning_rate": 2.848913452941421e-07, "loss": 1.6495, "step": 45540 }, { "epoch": 3.5690438871473353, "grad_norm": 3.0929224491119385, "learning_rate": 2.847887526002269e-07, "loss": 1.3891, "step": 45541 }, { "epoch": 3.5691222570532917, "grad_norm": 3.7005109786987305, "learning_rate": 2.84686177840553e-07, "loss": 1.3644, "step": 45542 }, { "epoch": 3.5692006269592476, "grad_norm": 5.750275135040283, "learning_rate": 2.8458362101551063e-07, "loss": 1.48, "step": 45543 }, { "epoch": 3.569278996865204, "grad_norm": 4.5589776039123535, "learning_rate": 2.8448108212549007e-07, "loss": 1.1146, "step": 45544 }, { "epoch": 3.56935736677116, "grad_norm": 2.5753557682037354, "learning_rate": 2.8437856117088104e-07, "loss": 1.1096, "step": 45545 }, { "epoch": 3.569435736677116, "grad_norm": 2.810148000717163, "learning_rate": 2.842760581520743e-07, "loss": 1.3743, "step": 45546 }, { "epoch": 3.569514106583072, "grad_norm": 2.9397079944610596, "learning_rate": 2.841735730694595e-07, "loss": 1.4327, "step": 45547 }, { "epoch": 3.5695924764890283, "grad_norm": 14.246491432189941, "learning_rate": 2.840711059234258e-07, "loss": 0.9678, "step": 45548 }, { "epoch": 3.569670846394984, "grad_norm": 2.4654150009155273, "learning_rate": 2.8396865671436405e-07, "loss": 1.2303, "step": 45549 }, { "epoch": 3.5697492163009406, "grad_norm": 3.0524497032165527, "learning_rate": 2.8386622544266273e-07, "loss": 0.9431, "step": 45550 }, { "epoch": 3.5698275862068964, "grad_norm": 4.238097667694092, "learning_rate": 2.837638121087116e-07, "loss": 0.9417, "step": 45551 }, { "epoch": 3.5699059561128528, "grad_norm": 3.4088187217712402, "learning_rate": 2.8366141671290146e-07, "loss": 1.4601, "step": 45552 }, { "epoch": 3.5699843260188087, "grad_norm": 3.1820406913757324, "learning_rate": 2.8355903925561976e-07, "loss": 0.6088, "step": 45553 }, { "epoch": 3.570062695924765, "grad_norm": 9.098515510559082, "learning_rate": 2.834566797372573e-07, "loss": 1.01, "step": 45554 }, { "epoch": 3.570141065830721, "grad_norm": 4.567959785461426, "learning_rate": 2.833543381582032e-07, "loss": 1.3371, "step": 45555 }, { "epoch": 3.570219435736677, "grad_norm": 5.165533542633057, "learning_rate": 2.832520145188472e-07, "loss": 1.5455, "step": 45556 }, { "epoch": 3.570297805642633, "grad_norm": 12.515936851501465, "learning_rate": 2.8314970881957726e-07, "loss": 1.1228, "step": 45557 }, { "epoch": 3.5703761755485894, "grad_norm": 3.4309895038604736, "learning_rate": 2.8304742106078255e-07, "loss": 1.3116, "step": 45558 }, { "epoch": 3.5704545454545453, "grad_norm": 2.763488292694092, "learning_rate": 2.8294515124285384e-07, "loss": 1.3074, "step": 45559 }, { "epoch": 3.5705329153605017, "grad_norm": 3.4250190258026123, "learning_rate": 2.8284289936617805e-07, "loss": 1.2704, "step": 45560 }, { "epoch": 3.5706112852664575, "grad_norm": 8.556685447692871, "learning_rate": 2.827406654311454e-07, "loss": 1.0221, "step": 45561 }, { "epoch": 3.570689655172414, "grad_norm": 7.535243511199951, "learning_rate": 2.82638449438144e-07, "loss": 1.412, "step": 45562 }, { "epoch": 3.5707680250783698, "grad_norm": 4.877080917358398, "learning_rate": 2.8253625138756347e-07, "loss": 1.8337, "step": 45563 }, { "epoch": 3.570846394984326, "grad_norm": 3.671607494354248, "learning_rate": 2.8243407127979125e-07, "loss": 1.3872, "step": 45564 }, { "epoch": 3.570924764890282, "grad_norm": 3.47693133354187, "learning_rate": 2.8233190911521713e-07, "loss": 0.8893, "step": 45565 }, { "epoch": 3.5710031347962383, "grad_norm": 3.423755407333374, "learning_rate": 2.822297648942296e-07, "loss": 1.5252, "step": 45566 }, { "epoch": 3.571081504702194, "grad_norm": 3.270341157913208, "learning_rate": 2.8212763861721613e-07, "loss": 1.3494, "step": 45567 }, { "epoch": 3.5711598746081505, "grad_norm": 3.224608898162842, "learning_rate": 2.8202553028456647e-07, "loss": 1.1589, "step": 45568 }, { "epoch": 3.5712382445141064, "grad_norm": 3.079768180847168, "learning_rate": 2.81923439896668e-07, "loss": 1.2962, "step": 45569 }, { "epoch": 3.5713166144200628, "grad_norm": 4.249144077301025, "learning_rate": 2.8182136745390945e-07, "loss": 1.6969, "step": 45570 }, { "epoch": 3.5713949843260187, "grad_norm": 5.579084873199463, "learning_rate": 2.817193129566798e-07, "loss": 1.4741, "step": 45571 }, { "epoch": 3.571473354231975, "grad_norm": 3.119107246398926, "learning_rate": 2.816172764053654e-07, "loss": 1.741, "step": 45572 }, { "epoch": 3.571551724137931, "grad_norm": 4.454538822174072, "learning_rate": 2.815152578003555e-07, "loss": 1.0125, "step": 45573 }, { "epoch": 3.571630094043887, "grad_norm": 3.0217177867889404, "learning_rate": 2.8141325714203915e-07, "loss": 1.399, "step": 45574 }, { "epoch": 3.571708463949843, "grad_norm": 4.525537967681885, "learning_rate": 2.813112744308022e-07, "loss": 1.3605, "step": 45575 }, { "epoch": 3.5717868338557994, "grad_norm": 2.699068784713745, "learning_rate": 2.812093096670332e-07, "loss": 1.4925, "step": 45576 }, { "epoch": 3.5718652037617558, "grad_norm": 2.9592525959014893, "learning_rate": 2.8110736285111964e-07, "loss": 1.1243, "step": 45577 }, { "epoch": 3.5719435736677116, "grad_norm": 3.282823085784912, "learning_rate": 2.810054339834517e-07, "loss": 1.1582, "step": 45578 }, { "epoch": 3.5720219435736675, "grad_norm": 4.280979156494141, "learning_rate": 2.809035230644147e-07, "loss": 1.3548, "step": 45579 }, { "epoch": 3.572100313479624, "grad_norm": 4.608616828918457, "learning_rate": 2.808016300943961e-07, "loss": 1.5942, "step": 45580 }, { "epoch": 3.57217868338558, "grad_norm": 12.963237762451172, "learning_rate": 2.8069975507378443e-07, "loss": 0.7657, "step": 45581 }, { "epoch": 3.572257053291536, "grad_norm": 4.433958530426025, "learning_rate": 2.805978980029672e-07, "loss": 1.2257, "step": 45582 }, { "epoch": 3.572335423197492, "grad_norm": 3.004051446914673, "learning_rate": 2.8049605888233135e-07, "loss": 1.522, "step": 45583 }, { "epoch": 3.5724137931034483, "grad_norm": 2.5155234336853027, "learning_rate": 2.8039423771226424e-07, "loss": 1.164, "step": 45584 }, { "epoch": 3.5724921630094046, "grad_norm": 2.746370792388916, "learning_rate": 2.8029243449315345e-07, "loss": 1.4294, "step": 45585 }, { "epoch": 3.5725705329153605, "grad_norm": 2.8130056858062744, "learning_rate": 2.801906492253864e-07, "loss": 1.4236, "step": 45586 }, { "epoch": 3.5726489028213164, "grad_norm": 3.9579343795776367, "learning_rate": 2.8008888190934946e-07, "loss": 1.6783, "step": 45587 }, { "epoch": 3.5727272727272728, "grad_norm": 8.453957557678223, "learning_rate": 2.7998713254543006e-07, "loss": 1.3097, "step": 45588 }, { "epoch": 3.572805642633229, "grad_norm": 3.959111452102661, "learning_rate": 2.798854011340163e-07, "loss": 1.4556, "step": 45589 }, { "epoch": 3.572884012539185, "grad_norm": 5.7219038009643555, "learning_rate": 2.797836876754928e-07, "loss": 0.755, "step": 45590 }, { "epoch": 3.572962382445141, "grad_norm": 2.803342580795288, "learning_rate": 2.796819921702476e-07, "loss": 1.4572, "step": 45591 }, { "epoch": 3.573040752351097, "grad_norm": 3.229585886001587, "learning_rate": 2.795803146186682e-07, "loss": 1.378, "step": 45592 }, { "epoch": 3.5731191222570535, "grad_norm": 9.339126586914062, "learning_rate": 2.794786550211409e-07, "loss": 1.1449, "step": 45593 }, { "epoch": 3.5731974921630094, "grad_norm": 6.456017017364502, "learning_rate": 2.793770133780521e-07, "loss": 1.0832, "step": 45594 }, { "epoch": 3.5732758620689653, "grad_norm": 3.7124311923980713, "learning_rate": 2.7927538968978816e-07, "loss": 1.5418, "step": 45595 }, { "epoch": 3.5733542319749216, "grad_norm": 4.743582725524902, "learning_rate": 2.7917378395673546e-07, "loss": 0.9037, "step": 45596 }, { "epoch": 3.573432601880878, "grad_norm": 4.884554386138916, "learning_rate": 2.7907219617928193e-07, "loss": 1.2792, "step": 45597 }, { "epoch": 3.573510971786834, "grad_norm": 3.6497414112091064, "learning_rate": 2.7897062635781235e-07, "loss": 1.4164, "step": 45598 }, { "epoch": 3.5735893416927897, "grad_norm": 3.0885753631591797, "learning_rate": 2.788690744927142e-07, "loss": 1.4847, "step": 45599 }, { "epoch": 3.573667711598746, "grad_norm": 4.295164108276367, "learning_rate": 2.787675405843726e-07, "loss": 1.487, "step": 45600 }, { "epoch": 3.5737460815047024, "grad_norm": 3.0897088050842285, "learning_rate": 2.7866602463317404e-07, "loss": 1.4896, "step": 45601 }, { "epoch": 3.5738244514106583, "grad_norm": 5.07119607925415, "learning_rate": 2.785645266395054e-07, "loss": 1.3992, "step": 45602 }, { "epoch": 3.573902821316614, "grad_norm": 2.195946455001831, "learning_rate": 2.784630466037519e-07, "loss": 1.1832, "step": 45603 }, { "epoch": 3.5739811912225705, "grad_norm": 2.4971721172332764, "learning_rate": 2.78361584526301e-07, "loss": 1.2031, "step": 45604 }, { "epoch": 3.574059561128527, "grad_norm": 2.5933189392089844, "learning_rate": 2.782601404075358e-07, "loss": 1.6085, "step": 45605 }, { "epoch": 3.5741379310344827, "grad_norm": 2.3813419342041016, "learning_rate": 2.7815871424784427e-07, "loss": 1.4203, "step": 45606 }, { "epoch": 3.5742163009404386, "grad_norm": 5.1798014640808105, "learning_rate": 2.7805730604761216e-07, "loss": 1.4279, "step": 45607 }, { "epoch": 3.574294670846395, "grad_norm": 5.608471870422363, "learning_rate": 2.779559158072248e-07, "loss": 1.1171, "step": 45608 }, { "epoch": 3.5743730407523513, "grad_norm": 5.772709369659424, "learning_rate": 2.7785454352706745e-07, "loss": 1.5531, "step": 45609 }, { "epoch": 3.574451410658307, "grad_norm": 5.486364364624023, "learning_rate": 2.777531892075253e-07, "loss": 2.2195, "step": 45610 }, { "epoch": 3.574529780564263, "grad_norm": 10.3454008102417, "learning_rate": 2.776518528489852e-07, "loss": 1.0648, "step": 45611 }, { "epoch": 3.5746081504702194, "grad_norm": 9.37698745727539, "learning_rate": 2.7755053445183255e-07, "loss": 0.6437, "step": 45612 }, { "epoch": 3.5746865203761757, "grad_norm": 6.814279079437256, "learning_rate": 2.774492340164509e-07, "loss": 1.4693, "step": 45613 }, { "epoch": 3.5747648902821316, "grad_norm": 2.0792787075042725, "learning_rate": 2.773479515432276e-07, "loss": 1.4209, "step": 45614 }, { "epoch": 3.5748432601880875, "grad_norm": 2.2695770263671875, "learning_rate": 2.772466870325463e-07, "loss": 1.517, "step": 45615 }, { "epoch": 3.574921630094044, "grad_norm": 2.6708977222442627, "learning_rate": 2.7714544048479284e-07, "loss": 1.3406, "step": 45616 }, { "epoch": 3.575, "grad_norm": 3.0839457511901855, "learning_rate": 2.770442119003525e-07, "loss": 1.2056, "step": 45617 }, { "epoch": 3.575078369905956, "grad_norm": 3.001633405685425, "learning_rate": 2.7694300127961036e-07, "loss": 1.525, "step": 45618 }, { "epoch": 3.575156739811912, "grad_norm": 6.970234394073486, "learning_rate": 2.7684180862295073e-07, "loss": 1.169, "step": 45619 }, { "epoch": 3.5752351097178683, "grad_norm": 2.98219633102417, "learning_rate": 2.7674063393075934e-07, "loss": 1.0023, "step": 45620 }, { "epoch": 3.5753134796238246, "grad_norm": 3.1362264156341553, "learning_rate": 2.7663947720342034e-07, "loss": 0.9278, "step": 45621 }, { "epoch": 3.5753918495297805, "grad_norm": 2.9581081867218018, "learning_rate": 2.7653833844131837e-07, "loss": 1.523, "step": 45622 }, { "epoch": 3.575470219435737, "grad_norm": 3.5154571533203125, "learning_rate": 2.764372176448393e-07, "loss": 1.7503, "step": 45623 }, { "epoch": 3.5755485893416927, "grad_norm": 3.283226251602173, "learning_rate": 2.763361148143656e-07, "loss": 1.2242, "step": 45624 }, { "epoch": 3.575626959247649, "grad_norm": 3.5982699394226074, "learning_rate": 2.7623502995028416e-07, "loss": 1.3005, "step": 45625 }, { "epoch": 3.575705329153605, "grad_norm": 3.4289820194244385, "learning_rate": 2.76133963052978e-07, "loss": 1.3439, "step": 45626 }, { "epoch": 3.5757836990595613, "grad_norm": 3.2320165634155273, "learning_rate": 2.760329141228324e-07, "loss": 1.2528, "step": 45627 }, { "epoch": 3.575862068965517, "grad_norm": 3.953608274459839, "learning_rate": 2.759318831602309e-07, "loss": 1.3112, "step": 45628 }, { "epoch": 3.5759404388714735, "grad_norm": 6.241430282592773, "learning_rate": 2.7583087016555707e-07, "loss": 1.9245, "step": 45629 }, { "epoch": 3.5760188087774294, "grad_norm": 2.5545709133148193, "learning_rate": 2.757298751391979e-07, "loss": 1.3003, "step": 45630 }, { "epoch": 3.5760971786833857, "grad_norm": 3.3259212970733643, "learning_rate": 2.756288980815347e-07, "loss": 0.953, "step": 45631 }, { "epoch": 3.5761755485893416, "grad_norm": 6.389627933502197, "learning_rate": 2.755279389929527e-07, "loss": 1.6302, "step": 45632 }, { "epoch": 3.576253918495298, "grad_norm": 4.992300987243652, "learning_rate": 2.7542699787383554e-07, "loss": 1.5429, "step": 45633 }, { "epoch": 3.576332288401254, "grad_norm": 6.9153828620910645, "learning_rate": 2.7532607472456783e-07, "loss": 0.8601, "step": 45634 }, { "epoch": 3.57641065830721, "grad_norm": 5.828633785247803, "learning_rate": 2.752251695455327e-07, "loss": 1.4018, "step": 45635 }, { "epoch": 3.576489028213166, "grad_norm": 4.371723175048828, "learning_rate": 2.751242823371142e-07, "loss": 0.767, "step": 45636 }, { "epoch": 3.5765673981191224, "grad_norm": 3.2153513431549072, "learning_rate": 2.750234130996965e-07, "loss": 1.306, "step": 45637 }, { "epoch": 3.5766457680250783, "grad_norm": 2.7831082344055176, "learning_rate": 2.7492256183366264e-07, "loss": 1.2225, "step": 45638 }, { "epoch": 3.5767241379310346, "grad_norm": 2.963273525238037, "learning_rate": 2.7482172853939613e-07, "loss": 1.5292, "step": 45639 }, { "epoch": 3.5768025078369905, "grad_norm": 3.8180902004241943, "learning_rate": 2.7472091321728067e-07, "loss": 1.3611, "step": 45640 }, { "epoch": 3.576880877742947, "grad_norm": 3.2100491523742676, "learning_rate": 2.7462011586770034e-07, "loss": 1.2438, "step": 45641 }, { "epoch": 3.5769592476489027, "grad_norm": 3.1704819202423096, "learning_rate": 2.7451933649103757e-07, "loss": 1.3246, "step": 45642 }, { "epoch": 3.577037617554859, "grad_norm": 4.269728660583496, "learning_rate": 2.74418575087676e-07, "loss": 1.7186, "step": 45643 }, { "epoch": 3.577115987460815, "grad_norm": 2.5087249279022217, "learning_rate": 2.743178316579992e-07, "loss": 1.3148, "step": 45644 }, { "epoch": 3.5771943573667713, "grad_norm": 2.74882173538208, "learning_rate": 2.742171062023913e-07, "loss": 1.202, "step": 45645 }, { "epoch": 3.577272727272727, "grad_norm": 2.79484486579895, "learning_rate": 2.7411639872123255e-07, "loss": 1.3263, "step": 45646 }, { "epoch": 3.5773510971786835, "grad_norm": 3.4990248680114746, "learning_rate": 2.740157092149076e-07, "loss": 0.9596, "step": 45647 }, { "epoch": 3.5774294670846394, "grad_norm": 3.4902501106262207, "learning_rate": 2.739150376838001e-07, "loss": 1.8677, "step": 45648 }, { "epoch": 3.5775078369905957, "grad_norm": 2.9330530166625977, "learning_rate": 2.738143841282925e-07, "loss": 1.5254, "step": 45649 }, { "epoch": 3.5775862068965516, "grad_norm": 2.664754629135132, "learning_rate": 2.737137485487673e-07, "loss": 1.3561, "step": 45650 }, { "epoch": 3.577664576802508, "grad_norm": 2.964113712310791, "learning_rate": 2.7361313094560746e-07, "loss": 1.159, "step": 45651 }, { "epoch": 3.577742946708464, "grad_norm": 5.5371880531311035, "learning_rate": 2.73512531319195e-07, "loss": 1.7184, "step": 45652 }, { "epoch": 3.57782131661442, "grad_norm": 4.138975143432617, "learning_rate": 2.734119496699139e-07, "loss": 1.0821, "step": 45653 }, { "epoch": 3.577899686520376, "grad_norm": 4.426476955413818, "learning_rate": 2.733113859981457e-07, "loss": 1.2966, "step": 45654 }, { "epoch": 3.5779780564263324, "grad_norm": 2.241115093231201, "learning_rate": 2.7321084030427327e-07, "loss": 1.2026, "step": 45655 }, { "epoch": 3.5780564263322883, "grad_norm": 2.3614754676818848, "learning_rate": 2.7311031258867917e-07, "loss": 1.3151, "step": 45656 }, { "epoch": 3.5781347962382446, "grad_norm": 4.742771625518799, "learning_rate": 2.7300980285174526e-07, "loss": 1.8284, "step": 45657 }, { "epoch": 3.5782131661442005, "grad_norm": 4.530303001403809, "learning_rate": 2.729093110938541e-07, "loss": 1.7303, "step": 45658 }, { "epoch": 3.578291536050157, "grad_norm": 3.285597801208496, "learning_rate": 2.728088373153881e-07, "loss": 1.2452, "step": 45659 }, { "epoch": 3.5783699059561127, "grad_norm": 2.765043258666992, "learning_rate": 2.727083815167297e-07, "loss": 1.1965, "step": 45660 }, { "epoch": 3.578448275862069, "grad_norm": 5.5764851570129395, "learning_rate": 2.726079436982593e-07, "loss": 1.1414, "step": 45661 }, { "epoch": 3.578526645768025, "grad_norm": 4.882185935974121, "learning_rate": 2.725075238603608e-07, "loss": 1.3959, "step": 45662 }, { "epoch": 3.5786050156739813, "grad_norm": 3.2353084087371826, "learning_rate": 2.724071220034158e-07, "loss": 1.4479, "step": 45663 }, { "epoch": 3.578683385579937, "grad_norm": 3.309628963470459, "learning_rate": 2.7230673812780664e-07, "loss": 1.224, "step": 45664 }, { "epoch": 3.5787617554858935, "grad_norm": 3.979059934616089, "learning_rate": 2.72206372233913e-07, "loss": 1.3282, "step": 45665 }, { "epoch": 3.5788401253918494, "grad_norm": 4.127364635467529, "learning_rate": 2.721060243221174e-07, "loss": 1.1501, "step": 45666 }, { "epoch": 3.5789184952978057, "grad_norm": 3.6153252124786377, "learning_rate": 2.7200569439280344e-07, "loss": 1.3065, "step": 45667 }, { "epoch": 3.5789968652037616, "grad_norm": 5.489668369293213, "learning_rate": 2.7190538244635135e-07, "loss": 1.3726, "step": 45668 }, { "epoch": 3.579075235109718, "grad_norm": 3.5481059551239014, "learning_rate": 2.718050884831425e-07, "loss": 1.3445, "step": 45669 }, { "epoch": 3.579153605015674, "grad_norm": 2.611455202102661, "learning_rate": 2.717048125035582e-07, "loss": 1.0468, "step": 45670 }, { "epoch": 3.57923197492163, "grad_norm": 3.3559110164642334, "learning_rate": 2.716045545079798e-07, "loss": 1.5118, "step": 45671 }, { "epoch": 3.5793103448275865, "grad_norm": 3.70766282081604, "learning_rate": 2.7150431449678995e-07, "loss": 1.0411, "step": 45672 }, { "epoch": 3.5793887147335424, "grad_norm": 3.1042768955230713, "learning_rate": 2.714040924703681e-07, "loss": 1.0938, "step": 45673 }, { "epoch": 3.5794670846394983, "grad_norm": 3.1502864360809326, "learning_rate": 2.7130388842909696e-07, "loss": 0.8232, "step": 45674 }, { "epoch": 3.5795454545454546, "grad_norm": 9.412516593933105, "learning_rate": 2.712037023733566e-07, "loss": 0.9052, "step": 45675 }, { "epoch": 3.579623824451411, "grad_norm": 2.889939069747925, "learning_rate": 2.7110353430352896e-07, "loss": 1.568, "step": 45676 }, { "epoch": 3.579702194357367, "grad_norm": 5.62795352935791, "learning_rate": 2.710033842199944e-07, "loss": 1.2981, "step": 45677 }, { "epoch": 3.5797805642633227, "grad_norm": 2.423478364944458, "learning_rate": 2.709032521231336e-07, "loss": 1.1501, "step": 45678 }, { "epoch": 3.579858934169279, "grad_norm": 2.6393423080444336, "learning_rate": 2.7080313801332914e-07, "loss": 1.4754, "step": 45679 }, { "epoch": 3.5799373040752354, "grad_norm": 3.1425023078918457, "learning_rate": 2.707030418909584e-07, "loss": 0.8687, "step": 45680 }, { "epoch": 3.5800156739811912, "grad_norm": 3.0919759273529053, "learning_rate": 2.7060296375640503e-07, "loss": 1.2432, "step": 45681 }, { "epoch": 3.580094043887147, "grad_norm": 2.899812698364258, "learning_rate": 2.7050290361004926e-07, "loss": 1.3025, "step": 45682 }, { "epoch": 3.5801724137931035, "grad_norm": 7.178849697113037, "learning_rate": 2.704028614522713e-07, "loss": 1.4784, "step": 45683 }, { "epoch": 3.58025078369906, "grad_norm": 3.77878475189209, "learning_rate": 2.7030283728345153e-07, "loss": 1.3104, "step": 45684 }, { "epoch": 3.5803291536050157, "grad_norm": 3.738060474395752, "learning_rate": 2.702028311039695e-07, "loss": 1.072, "step": 45685 }, { "epoch": 3.5804075235109716, "grad_norm": 30.129310607910156, "learning_rate": 2.7010284291420774e-07, "loss": 0.8926, "step": 45686 }, { "epoch": 3.580485893416928, "grad_norm": 7.4899001121521, "learning_rate": 2.700028727145448e-07, "loss": 1.4847, "step": 45687 }, { "epoch": 3.5805642633228842, "grad_norm": 3.4333252906799316, "learning_rate": 2.699029205053616e-07, "loss": 1.3358, "step": 45688 }, { "epoch": 3.58064263322884, "grad_norm": 3.3118488788604736, "learning_rate": 2.6980298628703773e-07, "loss": 1.806, "step": 45689 }, { "epoch": 3.580721003134796, "grad_norm": 3.5876519680023193, "learning_rate": 2.69703070059954e-07, "loss": 1.3137, "step": 45690 }, { "epoch": 3.5807993730407524, "grad_norm": 6.8430681228637695, "learning_rate": 2.6960317182449017e-07, "loss": 1.6212, "step": 45691 }, { "epoch": 3.5808777429467087, "grad_norm": 1.9621093273162842, "learning_rate": 2.6950329158102585e-07, "loss": 1.2299, "step": 45692 }, { "epoch": 3.5809561128526646, "grad_norm": 2.780000686645508, "learning_rate": 2.694034293299419e-07, "loss": 1.1744, "step": 45693 }, { "epoch": 3.5810344827586205, "grad_norm": 9.352429389953613, "learning_rate": 2.693035850716169e-07, "loss": 1.0031, "step": 45694 }, { "epoch": 3.581112852664577, "grad_norm": 7.0020341873168945, "learning_rate": 2.6920375880643167e-07, "loss": 1.3007, "step": 45695 }, { "epoch": 3.581191222570533, "grad_norm": 3.0737154483795166, "learning_rate": 2.6910395053476533e-07, "loss": 1.3974, "step": 45696 }, { "epoch": 3.581269592476489, "grad_norm": 4.228686809539795, "learning_rate": 2.690041602569976e-07, "loss": 1.614, "step": 45697 }, { "epoch": 3.581347962382445, "grad_norm": 2.4051263332366943, "learning_rate": 2.689043879735087e-07, "loss": 1.2371, "step": 45698 }, { "epoch": 3.5814263322884012, "grad_norm": 5.305868625640869, "learning_rate": 2.6880463368467616e-07, "loss": 2.4352, "step": 45699 }, { "epoch": 3.5815047021943576, "grad_norm": 3.791370391845703, "learning_rate": 2.6870489739088124e-07, "loss": 1.6204, "step": 45700 }, { "epoch": 3.5815830721003135, "grad_norm": 2.3036086559295654, "learning_rate": 2.6860517909250315e-07, "loss": 1.2811, "step": 45701 }, { "epoch": 3.5816614420062693, "grad_norm": 2.601168155670166, "learning_rate": 2.685054787899205e-07, "loss": 0.957, "step": 45702 }, { "epoch": 3.5817398119122257, "grad_norm": 3.136291027069092, "learning_rate": 2.6840579648351293e-07, "loss": 1.3877, "step": 45703 }, { "epoch": 3.581818181818182, "grad_norm": 4.104156970977783, "learning_rate": 2.6830613217365853e-07, "loss": 1.7838, "step": 45704 }, { "epoch": 3.581896551724138, "grad_norm": 2.744701862335205, "learning_rate": 2.6820648586073863e-07, "loss": 1.3438, "step": 45705 }, { "epoch": 3.581974921630094, "grad_norm": 2.6272170543670654, "learning_rate": 2.681068575451301e-07, "loss": 1.1893, "step": 45706 }, { "epoch": 3.58205329153605, "grad_norm": 2.414187431335449, "learning_rate": 2.680072472272127e-07, "loss": 1.3295, "step": 45707 }, { "epoch": 3.5821316614420065, "grad_norm": 3.213714122772217, "learning_rate": 2.67907654907365e-07, "loss": 1.3106, "step": 45708 }, { "epoch": 3.5822100313479623, "grad_norm": 2.3768608570098877, "learning_rate": 2.678080805859662e-07, "loss": 1.4142, "step": 45709 }, { "epoch": 3.5822884012539182, "grad_norm": 6.069741249084473, "learning_rate": 2.677085242633948e-07, "loss": 1.6063, "step": 45710 }, { "epoch": 3.5823667711598746, "grad_norm": 4.068946361541748, "learning_rate": 2.6760898594002935e-07, "loss": 1.6511, "step": 45711 }, { "epoch": 3.582445141065831, "grad_norm": 3.187286853790283, "learning_rate": 2.6750946561624967e-07, "loss": 1.4049, "step": 45712 }, { "epoch": 3.582523510971787, "grad_norm": 2.4281296730041504, "learning_rate": 2.674099632924315e-07, "loss": 1.233, "step": 45713 }, { "epoch": 3.5826018808777427, "grad_norm": 3.0910418033599854, "learning_rate": 2.673104789689562e-07, "loss": 1.2905, "step": 45714 }, { "epoch": 3.582680250783699, "grad_norm": 3.844376802444458, "learning_rate": 2.6721101264620073e-07, "loss": 1.1164, "step": 45715 }, { "epoch": 3.5827586206896553, "grad_norm": 6.984130382537842, "learning_rate": 2.671115643245442e-07, "loss": 1.2045, "step": 45716 }, { "epoch": 3.5828369905956112, "grad_norm": 3.2915072441101074, "learning_rate": 2.6701213400436356e-07, "loss": 1.6261, "step": 45717 }, { "epoch": 3.5829153605015676, "grad_norm": 2.962895154953003, "learning_rate": 2.669127216860373e-07, "loss": 1.4708, "step": 45718 }, { "epoch": 3.5829937304075234, "grad_norm": 3.1493430137634277, "learning_rate": 2.668133273699447e-07, "loss": 1.4418, "step": 45719 }, { "epoch": 3.58307210031348, "grad_norm": 8.187899589538574, "learning_rate": 2.667139510564637e-07, "loss": 1.2812, "step": 45720 }, { "epoch": 3.5831504702194357, "grad_norm": 6.152522087097168, "learning_rate": 2.666145927459707e-07, "loss": 1.5301, "step": 45721 }, { "epoch": 3.583228840125392, "grad_norm": 2.897024393081665, "learning_rate": 2.665152524388448e-07, "loss": 1.2221, "step": 45722 }, { "epoch": 3.583307210031348, "grad_norm": 2.9076504707336426, "learning_rate": 2.6641593013546294e-07, "loss": 1.3698, "step": 45723 }, { "epoch": 3.583385579937304, "grad_norm": 3.2218315601348877, "learning_rate": 2.6631662583620485e-07, "loss": 1.3958, "step": 45724 }, { "epoch": 3.58346394984326, "grad_norm": 2.610948085784912, "learning_rate": 2.662173395414464e-07, "loss": 1.5132, "step": 45725 }, { "epoch": 3.5835423197492164, "grad_norm": 8.528678894042969, "learning_rate": 2.661180712515654e-07, "loss": 1.4763, "step": 45726 }, { "epoch": 3.5836206896551723, "grad_norm": 3.6375133991241455, "learning_rate": 2.660188209669401e-07, "loss": 0.838, "step": 45727 }, { "epoch": 3.5836990595611287, "grad_norm": 2.219088554382324, "learning_rate": 2.6591958868794785e-07, "loss": 1.0106, "step": 45728 }, { "epoch": 3.5837774294670846, "grad_norm": 2.6736791133880615, "learning_rate": 2.658203744149657e-07, "loss": 1.2359, "step": 45729 }, { "epoch": 3.583855799373041, "grad_norm": 4.183745384216309, "learning_rate": 2.6572117814837096e-07, "loss": 1.4364, "step": 45730 }, { "epoch": 3.5839341692789968, "grad_norm": 4.196970462799072, "learning_rate": 2.6562199988854176e-07, "loss": 1.2666, "step": 45731 }, { "epoch": 3.584012539184953, "grad_norm": 2.5324814319610596, "learning_rate": 2.6552283963585444e-07, "loss": 1.0953, "step": 45732 }, { "epoch": 3.584090909090909, "grad_norm": 6.473834037780762, "learning_rate": 2.654236973906865e-07, "loss": 1.5141, "step": 45733 }, { "epoch": 3.5841692789968653, "grad_norm": 2.7741920948028564, "learning_rate": 2.653245731534149e-07, "loss": 1.2112, "step": 45734 }, { "epoch": 3.584247648902821, "grad_norm": 2.6085777282714844, "learning_rate": 2.65225466924417e-07, "loss": 1.332, "step": 45735 }, { "epoch": 3.5843260188087775, "grad_norm": 3.0231380462646484, "learning_rate": 2.6512637870406923e-07, "loss": 1.5185, "step": 45736 }, { "epoch": 3.5844043887147334, "grad_norm": 3.851850986480713, "learning_rate": 2.650273084927485e-07, "loss": 1.3102, "step": 45737 }, { "epoch": 3.5844827586206898, "grad_norm": 8.536178588867188, "learning_rate": 2.6492825629083173e-07, "loss": 1.0745, "step": 45738 }, { "epoch": 3.5845611285266457, "grad_norm": 4.309355735778809, "learning_rate": 2.64829222098697e-07, "loss": 1.2018, "step": 45739 }, { "epoch": 3.584639498432602, "grad_norm": 3.7921226024627686, "learning_rate": 2.6473020591671895e-07, "loss": 1.1538, "step": 45740 }, { "epoch": 3.584717868338558, "grad_norm": 3.644876718521118, "learning_rate": 2.6463120774527506e-07, "loss": 1.3127, "step": 45741 }, { "epoch": 3.584796238244514, "grad_norm": 12.115601539611816, "learning_rate": 2.645322275847417e-07, "loss": 1.5586, "step": 45742 }, { "epoch": 3.58487460815047, "grad_norm": 3.54464054107666, "learning_rate": 2.644332654354953e-07, "loss": 1.6629, "step": 45743 }, { "epoch": 3.5849529780564264, "grad_norm": 2.671900749206543, "learning_rate": 2.643343212979127e-07, "loss": 1.2501, "step": 45744 }, { "epoch": 3.5850313479623823, "grad_norm": 2.9264862537384033, "learning_rate": 2.6423539517236974e-07, "loss": 1.5519, "step": 45745 }, { "epoch": 3.5851097178683387, "grad_norm": 7.733621597290039, "learning_rate": 2.6413648705924333e-07, "loss": 0.8945, "step": 45746 }, { "epoch": 3.5851880877742945, "grad_norm": 3.93515944480896, "learning_rate": 2.6403759695890873e-07, "loss": 1.0901, "step": 45747 }, { "epoch": 3.585266457680251, "grad_norm": 4.108344554901123, "learning_rate": 2.6393872487174285e-07, "loss": 1.6128, "step": 45748 }, { "epoch": 3.5853448275862068, "grad_norm": 2.6611826419830322, "learning_rate": 2.63839870798121e-07, "loss": 0.7588, "step": 45749 }, { "epoch": 3.585423197492163, "grad_norm": 2.396949291229248, "learning_rate": 2.637410347384201e-07, "loss": 1.6064, "step": 45750 }, { "epoch": 3.585501567398119, "grad_norm": 2.43764066696167, "learning_rate": 2.636422166930158e-07, "loss": 1.1754, "step": 45751 }, { "epoch": 3.5855799373040753, "grad_norm": 2.9392447471618652, "learning_rate": 2.6354341666228357e-07, "loss": 1.6656, "step": 45752 }, { "epoch": 3.585658307210031, "grad_norm": 2.4026365280151367, "learning_rate": 2.634446346465991e-07, "loss": 1.265, "step": 45753 }, { "epoch": 3.5857366771159875, "grad_norm": 2.3746392726898193, "learning_rate": 2.6334587064633933e-07, "loss": 0.9667, "step": 45754 }, { "epoch": 3.5858150470219434, "grad_norm": 4.714215278625488, "learning_rate": 2.632471246618773e-07, "loss": 1.5118, "step": 45755 }, { "epoch": 3.5858934169278998, "grad_norm": 2.462066411972046, "learning_rate": 2.6314839669359105e-07, "loss": 1.1112, "step": 45756 }, { "epoch": 3.5859717868338556, "grad_norm": 4.0549468994140625, "learning_rate": 2.630496867418564e-07, "loss": 1.1064, "step": 45757 }, { "epoch": 3.586050156739812, "grad_norm": 15.742814064025879, "learning_rate": 2.6295099480704687e-07, "loss": 1.0812, "step": 45758 }, { "epoch": 3.586128526645768, "grad_norm": 2.7063148021698, "learning_rate": 2.628523208895384e-07, "loss": 1.4743, "step": 45759 }, { "epoch": 3.586206896551724, "grad_norm": 2.994882106781006, "learning_rate": 2.6275366498970553e-07, "loss": 1.0409, "step": 45760 }, { "epoch": 3.58628526645768, "grad_norm": 4.007625102996826, "learning_rate": 2.6265502710792646e-07, "loss": 0.6568, "step": 45761 }, { "epoch": 3.5863636363636364, "grad_norm": 4.7258782386779785, "learning_rate": 2.6255640724457354e-07, "loss": 1.4375, "step": 45762 }, { "epoch": 3.5864420062695928, "grad_norm": 2.8949151039123535, "learning_rate": 2.624578054000232e-07, "loss": 1.4223, "step": 45763 }, { "epoch": 3.5865203761755486, "grad_norm": 3.069117784500122, "learning_rate": 2.6235922157464955e-07, "loss": 1.4076, "step": 45764 }, { "epoch": 3.5865987460815045, "grad_norm": 3.04972505569458, "learning_rate": 2.622606557688279e-07, "loss": 1.5308, "step": 45765 }, { "epoch": 3.586677115987461, "grad_norm": 5.8177924156188965, "learning_rate": 2.621621079829334e-07, "loss": 1.0744, "step": 45766 }, { "epoch": 3.586755485893417, "grad_norm": 3.8429999351501465, "learning_rate": 2.6206357821734083e-07, "loss": 1.3752, "step": 45767 }, { "epoch": 3.586833855799373, "grad_norm": 3.427067756652832, "learning_rate": 2.619650664724255e-07, "loss": 1.4674, "step": 45768 }, { "epoch": 3.586912225705329, "grad_norm": 3.569505214691162, "learning_rate": 2.6186657274855976e-07, "loss": 1.2641, "step": 45769 }, { "epoch": 3.5869905956112853, "grad_norm": 3.011387586593628, "learning_rate": 2.617680970461212e-07, "loss": 1.4456, "step": 45770 }, { "epoch": 3.5870689655172416, "grad_norm": 2.332911968231201, "learning_rate": 2.6166963936548275e-07, "loss": 1.1972, "step": 45771 }, { "epoch": 3.5871473354231975, "grad_norm": 5.368443012237549, "learning_rate": 2.615711997070197e-07, "loss": 1.7271, "step": 45772 }, { "epoch": 3.5872257053291534, "grad_norm": 4.057391166687012, "learning_rate": 2.614727780711057e-07, "loss": 1.0764, "step": 45773 }, { "epoch": 3.5873040752351097, "grad_norm": 5.046483516693115, "learning_rate": 2.6137437445811433e-07, "loss": 0.9001, "step": 45774 }, { "epoch": 3.587382445141066, "grad_norm": 3.572953701019287, "learning_rate": 2.612759888684219e-07, "loss": 1.2876, "step": 45775 }, { "epoch": 3.587460815047022, "grad_norm": 6.092368125915527, "learning_rate": 2.6117762130240253e-07, "loss": 1.2492, "step": 45776 }, { "epoch": 3.587539184952978, "grad_norm": 4.264428615570068, "learning_rate": 2.6107927176042823e-07, "loss": 1.105, "step": 45777 }, { "epoch": 3.587617554858934, "grad_norm": 2.908935546875, "learning_rate": 2.609809402428748e-07, "loss": 1.026, "step": 45778 }, { "epoch": 3.5876959247648905, "grad_norm": 3.039679527282715, "learning_rate": 2.608826267501147e-07, "loss": 1.2857, "step": 45779 }, { "epoch": 3.5877742946708464, "grad_norm": 4.218843936920166, "learning_rate": 2.607843312825248e-07, "loss": 0.988, "step": 45780 }, { "epoch": 3.5878526645768023, "grad_norm": 3.0116896629333496, "learning_rate": 2.60686053840476e-07, "loss": 1.4935, "step": 45781 }, { "epoch": 3.5879310344827586, "grad_norm": 5.006309986114502, "learning_rate": 2.6058779442434347e-07, "loss": 1.3758, "step": 45782 }, { "epoch": 3.588009404388715, "grad_norm": 3.642829418182373, "learning_rate": 2.604895530345003e-07, "loss": 1.4135, "step": 45783 }, { "epoch": 3.588087774294671, "grad_norm": 3.3395655155181885, "learning_rate": 2.6039132967132064e-07, "loss": 1.2715, "step": 45784 }, { "epoch": 3.5881661442006267, "grad_norm": 7.884938716888428, "learning_rate": 2.6029312433517804e-07, "loss": 0.6888, "step": 45785 }, { "epoch": 3.588244514106583, "grad_norm": 4.054621696472168, "learning_rate": 2.601949370264456e-07, "loss": 1.8466, "step": 45786 }, { "epoch": 3.5883228840125394, "grad_norm": 4.449995040893555, "learning_rate": 2.60096767745498e-07, "loss": 1.2869, "step": 45787 }, { "epoch": 3.5884012539184953, "grad_norm": 4.592594623565674, "learning_rate": 2.599986164927065e-07, "loss": 1.0608, "step": 45788 }, { "epoch": 3.588479623824451, "grad_norm": 5.0639238357543945, "learning_rate": 2.5990048326844655e-07, "loss": 1.3484, "step": 45789 }, { "epoch": 3.5885579937304075, "grad_norm": 6.50766134262085, "learning_rate": 2.598023680730899e-07, "loss": 1.0941, "step": 45790 }, { "epoch": 3.588636363636364, "grad_norm": 3.9908440113067627, "learning_rate": 2.5970427090701136e-07, "loss": 1.4822, "step": 45791 }, { "epoch": 3.5887147335423197, "grad_norm": 3.495098114013672, "learning_rate": 2.596061917705822e-07, "loss": 1.1875, "step": 45792 }, { "epoch": 3.5887931034482756, "grad_norm": 5.528858184814453, "learning_rate": 2.595081306641756e-07, "loss": 1.8535, "step": 45793 }, { "epoch": 3.588871473354232, "grad_norm": 3.0873427391052246, "learning_rate": 2.594100875881661e-07, "loss": 1.4438, "step": 45794 }, { "epoch": 3.5889498432601883, "grad_norm": 2.6852171421051025, "learning_rate": 2.5931206254292575e-07, "loss": 1.4291, "step": 45795 }, { "epoch": 3.589028213166144, "grad_norm": 3.8055496215820312, "learning_rate": 2.592140555288275e-07, "loss": 1.0975, "step": 45796 }, { "epoch": 3.5891065830721, "grad_norm": 7.581970691680908, "learning_rate": 2.591160665462433e-07, "loss": 1.1756, "step": 45797 }, { "epoch": 3.5891849529780564, "grad_norm": 6.721639156341553, "learning_rate": 2.590180955955468e-07, "loss": 1.666, "step": 45798 }, { "epoch": 3.5892633228840127, "grad_norm": 4.054866313934326, "learning_rate": 2.589201426771104e-07, "loss": 1.4193, "step": 45799 }, { "epoch": 3.5893416927899686, "grad_norm": 5.681127071380615, "learning_rate": 2.5882220779130665e-07, "loss": 1.2406, "step": 45800 }, { "epoch": 3.5894200626959245, "grad_norm": 3.9263699054718018, "learning_rate": 2.5872429093850793e-07, "loss": 1.3754, "step": 45801 }, { "epoch": 3.589498432601881, "grad_norm": 2.2634730339050293, "learning_rate": 2.5862639211908624e-07, "loss": 1.2964, "step": 45802 }, { "epoch": 3.589576802507837, "grad_norm": 3.746565818786621, "learning_rate": 2.5852851133341517e-07, "loss": 1.5473, "step": 45803 }, { "epoch": 3.589655172413793, "grad_norm": 3.1473326683044434, "learning_rate": 2.5843064858186604e-07, "loss": 1.4388, "step": 45804 }, { "epoch": 3.589733542319749, "grad_norm": 3.0539519786834717, "learning_rate": 2.5833280386481086e-07, "loss": 1.2743, "step": 45805 }, { "epoch": 3.5898119122257053, "grad_norm": 6.313288688659668, "learning_rate": 2.582349771826237e-07, "loss": 1.7861, "step": 45806 }, { "epoch": 3.5898902821316616, "grad_norm": 3.9279274940490723, "learning_rate": 2.5813716853567317e-07, "loss": 1.2602, "step": 45807 }, { "epoch": 3.5899686520376175, "grad_norm": 3.5047988891601562, "learning_rate": 2.58039377924334e-07, "loss": 1.4495, "step": 45808 }, { "epoch": 3.5900470219435734, "grad_norm": 3.0344491004943848, "learning_rate": 2.579416053489775e-07, "loss": 1.2308, "step": 45809 }, { "epoch": 3.5901253918495297, "grad_norm": 2.2686588764190674, "learning_rate": 2.578438508099762e-07, "loss": 1.0932, "step": 45810 }, { "epoch": 3.590203761755486, "grad_norm": 4.5290446281433105, "learning_rate": 2.577461143077009e-07, "loss": 1.8231, "step": 45811 }, { "epoch": 3.590282131661442, "grad_norm": 2.6311542987823486, "learning_rate": 2.576483958425224e-07, "loss": 1.1552, "step": 45812 }, { "epoch": 3.5903605015673983, "grad_norm": 2.9137041568756104, "learning_rate": 2.575506954148149e-07, "loss": 0.9788, "step": 45813 }, { "epoch": 3.590438871473354, "grad_norm": 4.492980003356934, "learning_rate": 2.574530130249481e-07, "loss": 1.4722, "step": 45814 }, { "epoch": 3.5905172413793105, "grad_norm": 3.088059186935425, "learning_rate": 2.5735534867329435e-07, "loss": 1.3607, "step": 45815 }, { "epoch": 3.5905956112852664, "grad_norm": 2.579383611679077, "learning_rate": 2.5725770236022463e-07, "loss": 1.1304, "step": 45816 }, { "epoch": 3.5906739811912227, "grad_norm": 3.11849308013916, "learning_rate": 2.5716007408611076e-07, "loss": 1.3581, "step": 45817 }, { "epoch": 3.5907523510971786, "grad_norm": 2.7936556339263916, "learning_rate": 2.570624638513236e-07, "loss": 1.4112, "step": 45818 }, { "epoch": 3.590830721003135, "grad_norm": 2.8297274112701416, "learning_rate": 2.5696487165623505e-07, "loss": 1.4538, "step": 45819 }, { "epoch": 3.590909090909091, "grad_norm": 3.1904540061950684, "learning_rate": 2.568672975012154e-07, "loss": 1.4166, "step": 45820 }, { "epoch": 3.590987460815047, "grad_norm": 6.229311466217041, "learning_rate": 2.5676974138663714e-07, "loss": 1.3121, "step": 45821 }, { "epoch": 3.591065830721003, "grad_norm": 2.4098703861236572, "learning_rate": 2.566722033128699e-07, "loss": 1.3295, "step": 45822 }, { "epoch": 3.5911442006269594, "grad_norm": 4.773094177246094, "learning_rate": 2.5657468328028513e-07, "loss": 1.2038, "step": 45823 }, { "epoch": 3.5912225705329153, "grad_norm": 3.0772366523742676, "learning_rate": 2.5647718128925526e-07, "loss": 1.2872, "step": 45824 }, { "epoch": 3.5913009404388716, "grad_norm": 2.3726649284362793, "learning_rate": 2.5637969734014835e-07, "loss": 1.2939, "step": 45825 }, { "epoch": 3.5913793103448275, "grad_norm": 3.3274919986724854, "learning_rate": 2.562822314333363e-07, "loss": 1.3873, "step": 45826 }, { "epoch": 3.591457680250784, "grad_norm": 4.679479598999023, "learning_rate": 2.561847835691911e-07, "loss": 1.4286, "step": 45827 }, { "epoch": 3.5915360501567397, "grad_norm": 3.4901506900787354, "learning_rate": 2.5608735374808234e-07, "loss": 1.3951, "step": 45828 }, { "epoch": 3.591614420062696, "grad_norm": 2.3940553665161133, "learning_rate": 2.5598994197038043e-07, "loss": 1.3553, "step": 45829 }, { "epoch": 3.591692789968652, "grad_norm": 6.885160446166992, "learning_rate": 2.5589254823645605e-07, "loss": 1.8131, "step": 45830 }, { "epoch": 3.5917711598746083, "grad_norm": 5.000585556030273, "learning_rate": 2.5579517254667894e-07, "loss": 0.8087, "step": 45831 }, { "epoch": 3.591849529780564, "grad_norm": 3.014859199523926, "learning_rate": 2.556978149014216e-07, "loss": 1.2218, "step": 45832 }, { "epoch": 3.5919278996865205, "grad_norm": 2.775150775909424, "learning_rate": 2.556004753010527e-07, "loss": 1.2812, "step": 45833 }, { "epoch": 3.5920062695924764, "grad_norm": 2.791062355041504, "learning_rate": 2.5550315374594235e-07, "loss": 1.4076, "step": 45834 }, { "epoch": 3.5920846394984327, "grad_norm": 14.571612358093262, "learning_rate": 2.554058502364609e-07, "loss": 1.9697, "step": 45835 }, { "epoch": 3.5921630094043886, "grad_norm": 5.982489109039307, "learning_rate": 2.553085647729786e-07, "loss": 1.0947, "step": 45836 }, { "epoch": 3.592241379310345, "grad_norm": 2.399585008621216, "learning_rate": 2.552112973558657e-07, "loss": 1.4075, "step": 45837 }, { "epoch": 3.592319749216301, "grad_norm": 3.260012149810791, "learning_rate": 2.551140479854919e-07, "loss": 1.4273, "step": 45838 }, { "epoch": 3.592398119122257, "grad_norm": 2.1247141361236572, "learning_rate": 2.5501681666222746e-07, "loss": 1.3978, "step": 45839 }, { "epoch": 3.592476489028213, "grad_norm": 2.3042054176330566, "learning_rate": 2.549196033864415e-07, "loss": 1.2578, "step": 45840 }, { "epoch": 3.5925548589341694, "grad_norm": 2.6524362564086914, "learning_rate": 2.548224081585043e-07, "loss": 1.4944, "step": 45841 }, { "epoch": 3.5926332288401253, "grad_norm": 3.4698681831359863, "learning_rate": 2.547252309787851e-07, "loss": 1.2999, "step": 45842 }, { "epoch": 3.5927115987460816, "grad_norm": 6.433428764343262, "learning_rate": 2.546280718476546e-07, "loss": 1.6796, "step": 45843 }, { "epoch": 3.5927899686520375, "grad_norm": 4.816202640533447, "learning_rate": 2.545309307654803e-07, "loss": 1.0005, "step": 45844 }, { "epoch": 3.592868338557994, "grad_norm": 3.5920159816741943, "learning_rate": 2.544338077326336e-07, "loss": 1.3178, "step": 45845 }, { "epoch": 3.5929467084639497, "grad_norm": 3.211987018585205, "learning_rate": 2.5433670274948306e-07, "loss": 1.6158, "step": 45846 }, { "epoch": 3.593025078369906, "grad_norm": 2.801877975463867, "learning_rate": 2.54239615816399e-07, "loss": 1.4238, "step": 45847 }, { "epoch": 3.593103448275862, "grad_norm": 6.938296318054199, "learning_rate": 2.541425469337494e-07, "loss": 0.9015, "step": 45848 }, { "epoch": 3.5931818181818183, "grad_norm": 2.552783966064453, "learning_rate": 2.5404549610190346e-07, "loss": 1.568, "step": 45849 }, { "epoch": 3.593260188087774, "grad_norm": 2.688608407974243, "learning_rate": 2.5394846332123026e-07, "loss": 1.3063, "step": 45850 }, { "epoch": 3.5933385579937305, "grad_norm": 2.690609931945801, "learning_rate": 2.5385144859210067e-07, "loss": 1.3154, "step": 45851 }, { "epoch": 3.5934169278996864, "grad_norm": 2.776500940322876, "learning_rate": 2.5375445191488214e-07, "loss": 1.39, "step": 45852 }, { "epoch": 3.5934952978056427, "grad_norm": 5.460997104644775, "learning_rate": 2.536574732899433e-07, "loss": 0.8221, "step": 45853 }, { "epoch": 3.5935736677115986, "grad_norm": 4.985800743103027, "learning_rate": 2.535605127176538e-07, "loss": 1.4673, "step": 45854 }, { "epoch": 3.593652037617555, "grad_norm": 3.83766508102417, "learning_rate": 2.534635701983823e-07, "loss": 1.1368, "step": 45855 }, { "epoch": 3.593730407523511, "grad_norm": 3.1010115146636963, "learning_rate": 2.533666457324968e-07, "loss": 1.2913, "step": 45856 }, { "epoch": 3.593808777429467, "grad_norm": 3.4669549465179443, "learning_rate": 2.53269739320367e-07, "loss": 1.33, "step": 45857 }, { "epoch": 3.5938871473354235, "grad_norm": 3.246889352798462, "learning_rate": 2.5317285096236156e-07, "loss": 1.1634, "step": 45858 }, { "epoch": 3.5939655172413794, "grad_norm": 3.9813973903656006, "learning_rate": 2.5307598065884787e-07, "loss": 1.4915, "step": 45859 }, { "epoch": 3.5940438871473352, "grad_norm": 3.8373308181762695, "learning_rate": 2.529791284101951e-07, "loss": 1.5119, "step": 45860 }, { "epoch": 3.5941222570532916, "grad_norm": 10.066828727722168, "learning_rate": 2.5288229421677133e-07, "loss": 1.1766, "step": 45861 }, { "epoch": 3.594200626959248, "grad_norm": 3.6118669509887695, "learning_rate": 2.5278547807894573e-07, "loss": 1.3593, "step": 45862 }, { "epoch": 3.594278996865204, "grad_norm": 5.410638332366943, "learning_rate": 2.5268867999708456e-07, "loss": 1.204, "step": 45863 }, { "epoch": 3.5943573667711597, "grad_norm": 2.711019515991211, "learning_rate": 2.525918999715582e-07, "loss": 1.4318, "step": 45864 }, { "epoch": 3.594435736677116, "grad_norm": 3.8869733810424805, "learning_rate": 2.524951380027335e-07, "loss": 1.1977, "step": 45865 }, { "epoch": 3.5945141065830724, "grad_norm": 3.2051069736480713, "learning_rate": 2.5239839409097966e-07, "loss": 0.7291, "step": 45866 }, { "epoch": 3.5945924764890282, "grad_norm": 2.4745843410491943, "learning_rate": 2.5230166823666304e-07, "loss": 1.2067, "step": 45867 }, { "epoch": 3.594670846394984, "grad_norm": 3.198946952819824, "learning_rate": 2.5220496044015166e-07, "loss": 1.6729, "step": 45868 }, { "epoch": 3.5947492163009405, "grad_norm": 2.3937721252441406, "learning_rate": 2.5210827070181574e-07, "loss": 1.2875, "step": 45869 }, { "epoch": 3.594827586206897, "grad_norm": 2.8484537601470947, "learning_rate": 2.520115990220201e-07, "loss": 1.2296, "step": 45870 }, { "epoch": 3.5949059561128527, "grad_norm": 2.9064414501190186, "learning_rate": 2.519149454011338e-07, "loss": 1.0923, "step": 45871 }, { "epoch": 3.5949843260188086, "grad_norm": 4.748464107513428, "learning_rate": 2.5181830983952437e-07, "loss": 1.4637, "step": 45872 }, { "epoch": 3.595062695924765, "grad_norm": 11.536466598510742, "learning_rate": 2.517216923375593e-07, "loss": 1.6424, "step": 45873 }, { "epoch": 3.5951410658307212, "grad_norm": 5.557794094085693, "learning_rate": 2.5162509289560544e-07, "loss": 1.0481, "step": 45874 }, { "epoch": 3.595219435736677, "grad_norm": 4.1996636390686035, "learning_rate": 2.5152851151403145e-07, "loss": 1.0368, "step": 45875 }, { "epoch": 3.595297805642633, "grad_norm": 9.830514907836914, "learning_rate": 2.514319481932037e-07, "loss": 1.2428, "step": 45876 }, { "epoch": 3.5953761755485893, "grad_norm": 4.615844249725342, "learning_rate": 2.5133540293349026e-07, "loss": 1.6566, "step": 45877 }, { "epoch": 3.5954545454545457, "grad_norm": 3.43612003326416, "learning_rate": 2.512388757352574e-07, "loss": 1.3715, "step": 45878 }, { "epoch": 3.5955329153605016, "grad_norm": 2.4810068607330322, "learning_rate": 2.511423665988727e-07, "loss": 1.3554, "step": 45879 }, { "epoch": 3.5956112852664575, "grad_norm": 3.590883731842041, "learning_rate": 2.510458755247042e-07, "loss": 1.4434, "step": 45880 }, { "epoch": 3.595689655172414, "grad_norm": 3.3450815677642822, "learning_rate": 2.5094940251311764e-07, "loss": 1.308, "step": 45881 }, { "epoch": 3.59576802507837, "grad_norm": 3.7512619495391846, "learning_rate": 2.5085294756447885e-07, "loss": 1.7821, "step": 45882 }, { "epoch": 3.595846394984326, "grad_norm": 3.975616455078125, "learning_rate": 2.5075651067915705e-07, "loss": 1.0474, "step": 45883 }, { "epoch": 3.595924764890282, "grad_norm": 3.0967185497283936, "learning_rate": 2.5066009185751917e-07, "loss": 1.3393, "step": 45884 }, { "epoch": 3.5960031347962382, "grad_norm": 3.198115587234497, "learning_rate": 2.5056369109992986e-07, "loss": 1.3478, "step": 45885 }, { "epoch": 3.5960815047021946, "grad_norm": 2.2830088138580322, "learning_rate": 2.5046730840675715e-07, "loss": 1.1188, "step": 45886 }, { "epoch": 3.5961598746081505, "grad_norm": 2.2312912940979004, "learning_rate": 2.5037094377836633e-07, "loss": 1.2967, "step": 45887 }, { "epoch": 3.5962382445141063, "grad_norm": 4.96881628036499, "learning_rate": 2.50274597215126e-07, "loss": 1.3112, "step": 45888 }, { "epoch": 3.5963166144200627, "grad_norm": 2.699338912963867, "learning_rate": 2.501782687174009e-07, "loss": 1.2929, "step": 45889 }, { "epoch": 3.596394984326019, "grad_norm": 2.5846216678619385, "learning_rate": 2.500819582855585e-07, "loss": 1.3519, "step": 45890 }, { "epoch": 3.596473354231975, "grad_norm": 3.351623773574829, "learning_rate": 2.49985665919964e-07, "loss": 1.1152, "step": 45891 }, { "epoch": 3.596551724137931, "grad_norm": 3.9910788536071777, "learning_rate": 2.4988939162098493e-07, "loss": 1.2222, "step": 45892 }, { "epoch": 3.596630094043887, "grad_norm": 3.0676114559173584, "learning_rate": 2.4979313538898596e-07, "loss": 1.2121, "step": 45893 }, { "epoch": 3.5967084639498434, "grad_norm": 2.293539524078369, "learning_rate": 2.4969689722433465e-07, "loss": 1.2295, "step": 45894 }, { "epoch": 3.5967868338557993, "grad_norm": 6.985561847686768, "learning_rate": 2.4960067712739733e-07, "loss": 0.4746, "step": 45895 }, { "epoch": 3.5968652037617552, "grad_norm": 3.3916726112365723, "learning_rate": 2.49504475098537e-07, "loss": 1.4052, "step": 45896 }, { "epoch": 3.5969435736677116, "grad_norm": 3.7322723865509033, "learning_rate": 2.494082911381229e-07, "loss": 1.1029, "step": 45897 }, { "epoch": 3.597021943573668, "grad_norm": 3.8757126331329346, "learning_rate": 2.493121252465192e-07, "loss": 1.2799, "step": 45898 }, { "epoch": 3.597100313479624, "grad_norm": 3.622689962387085, "learning_rate": 2.492159774240932e-07, "loss": 1.1037, "step": 45899 }, { "epoch": 3.5971786833855797, "grad_norm": 4.4468560218811035, "learning_rate": 2.491198476712081e-07, "loss": 1.2777, "step": 45900 }, { "epoch": 3.597257053291536, "grad_norm": 3.1981699466705322, "learning_rate": 2.4902373598823084e-07, "loss": 1.4279, "step": 45901 }, { "epoch": 3.5973354231974923, "grad_norm": 3.068392038345337, "learning_rate": 2.489276423755277e-07, "loss": 1.1519, "step": 45902 }, { "epoch": 3.597413793103448, "grad_norm": 4.48899507522583, "learning_rate": 2.48831566833464e-07, "loss": 1.6902, "step": 45903 }, { "epoch": 3.5974921630094046, "grad_norm": 2.5778348445892334, "learning_rate": 2.487355093624039e-07, "loss": 1.2452, "step": 45904 }, { "epoch": 3.5975705329153604, "grad_norm": 2.1879124641418457, "learning_rate": 2.4863946996271373e-07, "loss": 1.406, "step": 45905 }, { "epoch": 3.5976489028213168, "grad_norm": 2.840176820755005, "learning_rate": 2.4854344863475764e-07, "loss": 1.6171, "step": 45906 }, { "epoch": 3.5977272727272727, "grad_norm": 2.9925148487091064, "learning_rate": 2.4844744537890317e-07, "loss": 1.3539, "step": 45907 }, { "epoch": 3.597805642633229, "grad_norm": 3.571793794631958, "learning_rate": 2.483514601955134e-07, "loss": 1.5105, "step": 45908 }, { "epoch": 3.597884012539185, "grad_norm": 5.191120147705078, "learning_rate": 2.48255493084954e-07, "loss": 1.1245, "step": 45909 }, { "epoch": 3.597962382445141, "grad_norm": 3.011422634124756, "learning_rate": 2.4815954404759034e-07, "loss": 1.4239, "step": 45910 }, { "epoch": 3.598040752351097, "grad_norm": 3.637343645095825, "learning_rate": 2.480636130837866e-07, "loss": 1.3671, "step": 45911 }, { "epoch": 3.5981191222570534, "grad_norm": 4.788143634796143, "learning_rate": 2.479677001939085e-07, "loss": 1.384, "step": 45912 }, { "epoch": 3.5981974921630093, "grad_norm": 5.269737243652344, "learning_rate": 2.478718053783208e-07, "loss": 1.1235, "step": 45913 }, { "epoch": 3.5982758620689657, "grad_norm": 4.192681789398193, "learning_rate": 2.4777592863738765e-07, "loss": 1.5994, "step": 45914 }, { "epoch": 3.5983542319749215, "grad_norm": 4.316800594329834, "learning_rate": 2.476800699714732e-07, "loss": 1.3164, "step": 45915 }, { "epoch": 3.598432601880878, "grad_norm": 3.7360267639160156, "learning_rate": 2.4758422938094384e-07, "loss": 1.4352, "step": 45916 }, { "epoch": 3.5985109717868338, "grad_norm": 3.0679783821105957, "learning_rate": 2.474884068661626e-07, "loss": 1.3922, "step": 45917 }, { "epoch": 3.59858934169279, "grad_norm": 3.0585427284240723, "learning_rate": 2.473926024274953e-07, "loss": 0.67, "step": 45918 }, { "epoch": 3.598667711598746, "grad_norm": 2.7723388671875, "learning_rate": 2.4729681606530443e-07, "loss": 1.3289, "step": 45919 }, { "epoch": 3.5987460815047023, "grad_norm": 5.359841823577881, "learning_rate": 2.472010477799547e-07, "loss": 2.0046, "step": 45920 }, { "epoch": 3.598824451410658, "grad_norm": 6.518680572509766, "learning_rate": 2.4710529757181244e-07, "loss": 1.1403, "step": 45921 }, { "epoch": 3.5989028213166145, "grad_norm": 2.1501331329345703, "learning_rate": 2.470095654412402e-07, "loss": 1.3749, "step": 45922 }, { "epoch": 3.5989811912225704, "grad_norm": 11.28353214263916, "learning_rate": 2.469138513886016e-07, "loss": 0.8644, "step": 45923 }, { "epoch": 3.5990595611285268, "grad_norm": 2.7538702487945557, "learning_rate": 2.468181554142618e-07, "loss": 1.1771, "step": 45924 }, { "epoch": 3.5991379310344827, "grad_norm": 3.4643399715423584, "learning_rate": 2.467224775185839e-07, "loss": 1.5623, "step": 45925 }, { "epoch": 3.599216300940439, "grad_norm": 2.612788200378418, "learning_rate": 2.4662681770193263e-07, "loss": 1.3944, "step": 45926 }, { "epoch": 3.599294670846395, "grad_norm": 2.308781385421753, "learning_rate": 2.465311759646716e-07, "loss": 1.0213, "step": 45927 }, { "epoch": 3.599373040752351, "grad_norm": 3.252167224884033, "learning_rate": 2.464355523071643e-07, "loss": 1.5056, "step": 45928 }, { "epoch": 3.599451410658307, "grad_norm": 2.3038716316223145, "learning_rate": 2.4633994672977445e-07, "loss": 0.8461, "step": 45929 }, { "epoch": 3.5995297805642634, "grad_norm": 5.785068511962891, "learning_rate": 2.462443592328656e-07, "loss": 1.6862, "step": 45930 }, { "epoch": 3.5996081504702193, "grad_norm": 2.7039601802825928, "learning_rate": 2.461487898168019e-07, "loss": 1.5939, "step": 45931 }, { "epoch": 3.5996865203761756, "grad_norm": 5.2148637771606445, "learning_rate": 2.460532384819464e-07, "loss": 1.2101, "step": 45932 }, { "epoch": 3.5997648902821315, "grad_norm": 3.2673451900482178, "learning_rate": 2.4595770522866327e-07, "loss": 0.9085, "step": 45933 }, { "epoch": 3.599843260188088, "grad_norm": 3.965886116027832, "learning_rate": 2.458621900573138e-07, "loss": 1.3308, "step": 45934 }, { "epoch": 3.5999216300940438, "grad_norm": 3.4788877964019775, "learning_rate": 2.457666929682634e-07, "loss": 1.3404, "step": 45935 }, { "epoch": 3.6, "grad_norm": 3.278257369995117, "learning_rate": 2.45671213961875e-07, "loss": 1.3905, "step": 45936 }, { "epoch": 3.600078369905956, "grad_norm": 4.255572319030762, "learning_rate": 2.455757530385111e-07, "loss": 1.4898, "step": 45937 }, { "epoch": 3.6001567398119123, "grad_norm": 3.850205183029175, "learning_rate": 2.454803101985348e-07, "loss": 1.3827, "step": 45938 }, { "epoch": 3.600235109717868, "grad_norm": 3.481218099594116, "learning_rate": 2.4538488544230856e-07, "loss": 0.7543, "step": 45939 }, { "epoch": 3.6003134796238245, "grad_norm": 7.128936767578125, "learning_rate": 2.4528947877019706e-07, "loss": 1.4382, "step": 45940 }, { "epoch": 3.6003918495297804, "grad_norm": 3.491511106491089, "learning_rate": 2.4519409018256223e-07, "loss": 1.762, "step": 45941 }, { "epoch": 3.6004702194357368, "grad_norm": 5.2975640296936035, "learning_rate": 2.450987196797666e-07, "loss": 2.245, "step": 45942 }, { "epoch": 3.6005485893416926, "grad_norm": 10.512856483459473, "learning_rate": 2.4500336726217324e-07, "loss": 1.3229, "step": 45943 }, { "epoch": 3.600626959247649, "grad_norm": 2.7093021869659424, "learning_rate": 2.4490803293014456e-07, "loss": 1.8465, "step": 45944 }, { "epoch": 3.600705329153605, "grad_norm": 3.4016387462615967, "learning_rate": 2.448127166840436e-07, "loss": 1.3903, "step": 45945 }, { "epoch": 3.600783699059561, "grad_norm": 2.9541683197021484, "learning_rate": 2.447174185242324e-07, "loss": 1.1803, "step": 45946 }, { "epoch": 3.600862068965517, "grad_norm": 16.863229751586914, "learning_rate": 2.4462213845107385e-07, "loss": 0.9498, "step": 45947 }, { "epoch": 3.6009404388714734, "grad_norm": 3.848493814468384, "learning_rate": 2.4452687646493e-07, "loss": 1.2697, "step": 45948 }, { "epoch": 3.6010188087774293, "grad_norm": 3.9241714477539062, "learning_rate": 2.444316325661633e-07, "loss": 1.3349, "step": 45949 }, { "epoch": 3.6010971786833856, "grad_norm": 17.29003143310547, "learning_rate": 2.4433640675513624e-07, "loss": 1.3162, "step": 45950 }, { "epoch": 3.6011755485893415, "grad_norm": 2.3612985610961914, "learning_rate": 2.442411990322113e-07, "loss": 1.3729, "step": 45951 }, { "epoch": 3.601253918495298, "grad_norm": 3.210052013397217, "learning_rate": 2.4414600939774936e-07, "loss": 1.3036, "step": 45952 }, { "epoch": 3.601332288401254, "grad_norm": 10.459832191467285, "learning_rate": 2.440508378521134e-07, "loss": 1.436, "step": 45953 }, { "epoch": 3.60141065830721, "grad_norm": 4.171624660491943, "learning_rate": 2.439556843956653e-07, "loss": 1.7194, "step": 45954 }, { "epoch": 3.601489028213166, "grad_norm": 10.654579162597656, "learning_rate": 2.4386054902876767e-07, "loss": 1.0683, "step": 45955 }, { "epoch": 3.6015673981191223, "grad_norm": 4.050869464874268, "learning_rate": 2.4376543175178067e-07, "loss": 1.2099, "step": 45956 }, { "epoch": 3.6016457680250786, "grad_norm": 2.8747129440307617, "learning_rate": 2.436703325650669e-07, "loss": 1.586, "step": 45957 }, { "epoch": 3.6017241379310345, "grad_norm": 3.5813844203948975, "learning_rate": 2.435752514689882e-07, "loss": 1.4134, "step": 45958 }, { "epoch": 3.6018025078369904, "grad_norm": 4.202512264251709, "learning_rate": 2.434801884639071e-07, "loss": 1.1617, "step": 45959 }, { "epoch": 3.6018808777429467, "grad_norm": 2.836533784866333, "learning_rate": 2.433851435501838e-07, "loss": 1.4006, "step": 45960 }, { "epoch": 3.601959247648903, "grad_norm": 2.8138535022735596, "learning_rate": 2.4329011672818036e-07, "loss": 1.4075, "step": 45961 }, { "epoch": 3.602037617554859, "grad_norm": 4.3409929275512695, "learning_rate": 2.4319510799825696e-07, "loss": 1.2131, "step": 45962 }, { "epoch": 3.602115987460815, "grad_norm": 2.2916440963745117, "learning_rate": 2.4310011736077775e-07, "loss": 1.4034, "step": 45963 }, { "epoch": 3.602194357366771, "grad_norm": 2.962552309036255, "learning_rate": 2.430051448161019e-07, "loss": 1.2674, "step": 45964 }, { "epoch": 3.6022727272727275, "grad_norm": 3.6313631534576416, "learning_rate": 2.429101903645908e-07, "loss": 1.0023, "step": 45965 }, { "epoch": 3.6023510971786834, "grad_norm": 3.372063636779785, "learning_rate": 2.428152540066059e-07, "loss": 1.5557, "step": 45966 }, { "epoch": 3.6024294670846393, "grad_norm": 3.6304209232330322, "learning_rate": 2.4272033574250844e-07, "loss": 1.4315, "step": 45967 }, { "epoch": 3.6025078369905956, "grad_norm": 4.434387683868408, "learning_rate": 2.426254355726593e-07, "loss": 1.4638, "step": 45968 }, { "epoch": 3.602586206896552, "grad_norm": 3.6056511402130127, "learning_rate": 2.425305534974198e-07, "loss": 1.0953, "step": 45969 }, { "epoch": 3.602664576802508, "grad_norm": 2.848991632461548, "learning_rate": 2.424356895171509e-07, "loss": 1.2446, "step": 45970 }, { "epoch": 3.6027429467084637, "grad_norm": 1.9962196350097656, "learning_rate": 2.4234084363221165e-07, "loss": 0.9199, "step": 45971 }, { "epoch": 3.60282131661442, "grad_norm": 3.8902370929718018, "learning_rate": 2.422460158429646e-07, "loss": 1.2521, "step": 45972 }, { "epoch": 3.6028996865203764, "grad_norm": 6.335582733154297, "learning_rate": 2.421512061497705e-07, "loss": 1.328, "step": 45973 }, { "epoch": 3.6029780564263323, "grad_norm": 2.8949334621429443, "learning_rate": 2.420564145529897e-07, "loss": 1.3061, "step": 45974 }, { "epoch": 3.603056426332288, "grad_norm": 2.744046688079834, "learning_rate": 2.419616410529818e-07, "loss": 1.1994, "step": 45975 }, { "epoch": 3.6031347962382445, "grad_norm": 5.8502421379089355, "learning_rate": 2.4186688565010717e-07, "loss": 1.4838, "step": 45976 }, { "epoch": 3.603213166144201, "grad_norm": 2.6231110095977783, "learning_rate": 2.417721483447283e-07, "loss": 1.2321, "step": 45977 }, { "epoch": 3.6032915360501567, "grad_norm": 3.5196895599365234, "learning_rate": 2.416774291372037e-07, "loss": 0.8101, "step": 45978 }, { "epoch": 3.6033699059561126, "grad_norm": 10.410696983337402, "learning_rate": 2.415827280278943e-07, "loss": 1.6934, "step": 45979 }, { "epoch": 3.603448275862069, "grad_norm": 4.1831583976745605, "learning_rate": 2.414880450171603e-07, "loss": 1.3849, "step": 45980 }, { "epoch": 3.6035266457680253, "grad_norm": 5.170665740966797, "learning_rate": 2.4139338010536084e-07, "loss": 1.2392, "step": 45981 }, { "epoch": 3.603605015673981, "grad_norm": 3.817678928375244, "learning_rate": 2.4129873329285733e-07, "loss": 1.5805, "step": 45982 }, { "epoch": 3.603683385579937, "grad_norm": 6.657965183258057, "learning_rate": 2.4120410458000954e-07, "loss": 1.1451, "step": 45983 }, { "epoch": 3.6037617554858934, "grad_norm": 2.467473268508911, "learning_rate": 2.4110949396717656e-07, "loss": 1.1925, "step": 45984 }, { "epoch": 3.6038401253918497, "grad_norm": 3.0887351036071777, "learning_rate": 2.4101490145471863e-07, "loss": 1.3719, "step": 45985 }, { "epoch": 3.6039184952978056, "grad_norm": 2.1133079528808594, "learning_rate": 2.4092032704299607e-07, "loss": 1.2266, "step": 45986 }, { "epoch": 3.6039968652037615, "grad_norm": 4.0863213539123535, "learning_rate": 2.40825770732368e-07, "loss": 1.2613, "step": 45987 }, { "epoch": 3.604075235109718, "grad_norm": 2.2322072982788086, "learning_rate": 2.407312325231942e-07, "loss": 1.2471, "step": 45988 }, { "epoch": 3.604153605015674, "grad_norm": 4.389138698577881, "learning_rate": 2.4063671241583484e-07, "loss": 1.5363, "step": 45989 }, { "epoch": 3.60423197492163, "grad_norm": 3.1950247287750244, "learning_rate": 2.4054221041064753e-07, "loss": 1.5113, "step": 45990 }, { "epoch": 3.604310344827586, "grad_norm": 4.157998085021973, "learning_rate": 2.404477265079941e-07, "loss": 1.2789, "step": 45991 }, { "epoch": 3.6043887147335423, "grad_norm": 6.408341407775879, "learning_rate": 2.403532607082332e-07, "loss": 0.938, "step": 45992 }, { "epoch": 3.6044670846394986, "grad_norm": 3.3466570377349854, "learning_rate": 2.402588130117228e-07, "loss": 1.342, "step": 45993 }, { "epoch": 3.6045454545454545, "grad_norm": 2.2088711261749268, "learning_rate": 2.401643834188239e-07, "loss": 1.118, "step": 45994 }, { "epoch": 3.6046238244514104, "grad_norm": 3.4665894508361816, "learning_rate": 2.400699719298932e-07, "loss": 1.1086, "step": 45995 }, { "epoch": 3.6047021943573667, "grad_norm": 12.679139137268066, "learning_rate": 2.399755785452934e-07, "loss": 1.5228, "step": 45996 }, { "epoch": 3.604780564263323, "grad_norm": 2.630100965499878, "learning_rate": 2.398812032653813e-07, "loss": 1.6392, "step": 45997 }, { "epoch": 3.604858934169279, "grad_norm": 2.746497631072998, "learning_rate": 2.3978684609051563e-07, "loss": 1.4137, "step": 45998 }, { "epoch": 3.6049373040752353, "grad_norm": 3.22133469581604, "learning_rate": 2.3969250702105595e-07, "loss": 1.7373, "step": 45999 }, { "epoch": 3.605015673981191, "grad_norm": 5.081943035125732, "learning_rate": 2.3959818605736095e-07, "loss": 1.7252, "step": 46000 }, { "epoch": 3.6050940438871475, "grad_norm": 3.446423292160034, "learning_rate": 2.3950388319978924e-07, "loss": 1.2545, "step": 46001 }, { "epoch": 3.6051724137931034, "grad_norm": 3.3346545696258545, "learning_rate": 2.3940959844869995e-07, "loss": 1.304, "step": 46002 }, { "epoch": 3.6052507836990597, "grad_norm": 3.090336799621582, "learning_rate": 2.393153318044511e-07, "loss": 1.5801, "step": 46003 }, { "epoch": 3.6053291536050156, "grad_norm": 2.3441810607910156, "learning_rate": 2.392210832674013e-07, "loss": 1.3657, "step": 46004 }, { "epoch": 3.605407523510972, "grad_norm": 2.726067543029785, "learning_rate": 2.3912685283790925e-07, "loss": 1.3928, "step": 46005 }, { "epoch": 3.605485893416928, "grad_norm": 4.765727519989014, "learning_rate": 2.390326405163335e-07, "loss": 1.3468, "step": 46006 }, { "epoch": 3.605564263322884, "grad_norm": 4.095029354095459, "learning_rate": 2.389384463030331e-07, "loss": 1.2407, "step": 46007 }, { "epoch": 3.60564263322884, "grad_norm": 8.01506519317627, "learning_rate": 2.3884427019836396e-07, "loss": 1.4773, "step": 46008 }, { "epoch": 3.6057210031347964, "grad_norm": 3.211808681488037, "learning_rate": 2.3875011220268585e-07, "loss": 0.7269, "step": 46009 }, { "epoch": 3.6057993730407523, "grad_norm": 5.5665202140808105, "learning_rate": 2.3865597231635727e-07, "loss": 1.5503, "step": 46010 }, { "epoch": 3.6058777429467086, "grad_norm": 7.232393741607666, "learning_rate": 2.3856185053973636e-07, "loss": 1.585, "step": 46011 }, { "epoch": 3.6059561128526645, "grad_norm": 6.473598957061768, "learning_rate": 2.3846774687318e-07, "loss": 1.2905, "step": 46012 }, { "epoch": 3.606034482758621, "grad_norm": 3.145275831222534, "learning_rate": 2.383736613170462e-07, "loss": 1.3211, "step": 46013 }, { "epoch": 3.6061128526645767, "grad_norm": 3.8913676738739014, "learning_rate": 2.382795938716931e-07, "loss": 1.4328, "step": 46014 }, { "epoch": 3.606191222570533, "grad_norm": 6.681745529174805, "learning_rate": 2.3818554453747978e-07, "loss": 1.6339, "step": 46015 }, { "epoch": 3.606269592476489, "grad_norm": 10.101913452148438, "learning_rate": 2.380915133147621e-07, "loss": 2.0236, "step": 46016 }, { "epoch": 3.6063479623824453, "grad_norm": 12.159869194030762, "learning_rate": 2.3799750020389867e-07, "loss": 0.6138, "step": 46017 }, { "epoch": 3.606426332288401, "grad_norm": 3.605769395828247, "learning_rate": 2.3790350520524642e-07, "loss": 1.5831, "step": 46018 }, { "epoch": 3.6065047021943575, "grad_norm": 6.1577019691467285, "learning_rate": 2.3780952831916338e-07, "loss": 1.3004, "step": 46019 }, { "epoch": 3.6065830721003134, "grad_norm": 3.0668981075286865, "learning_rate": 2.3771556954600705e-07, "loss": 1.2776, "step": 46020 }, { "epoch": 3.6066614420062697, "grad_norm": 2.3466439247131348, "learning_rate": 2.3762162888613494e-07, "loss": 1.2866, "step": 46021 }, { "epoch": 3.6067398119122256, "grad_norm": 3.8970348834991455, "learning_rate": 2.3752770633990397e-07, "loss": 1.2994, "step": 46022 }, { "epoch": 3.606818181818182, "grad_norm": 3.451324224472046, "learning_rate": 2.374338019076705e-07, "loss": 1.3252, "step": 46023 }, { "epoch": 3.606896551724138, "grad_norm": 2.687159299850464, "learning_rate": 2.3733991558979374e-07, "loss": 1.4939, "step": 46024 }, { "epoch": 3.606974921630094, "grad_norm": 3.604156732559204, "learning_rate": 2.3724604738662892e-07, "loss": 1.404, "step": 46025 }, { "epoch": 3.60705329153605, "grad_norm": 2.7670681476593018, "learning_rate": 2.3715219729853523e-07, "loss": 1.6955, "step": 46026 }, { "epoch": 3.6071316614420064, "grad_norm": 3.842297077178955, "learning_rate": 2.370583653258668e-07, "loss": 1.3364, "step": 46027 }, { "epoch": 3.6072100313479623, "grad_norm": 6.618363857269287, "learning_rate": 2.3696455146898167e-07, "loss": 1.4247, "step": 46028 }, { "epoch": 3.6072884012539186, "grad_norm": 3.7010247707366943, "learning_rate": 2.3687075572823792e-07, "loss": 1.765, "step": 46029 }, { "epoch": 3.6073667711598745, "grad_norm": 3.7009174823760986, "learning_rate": 2.3677697810399135e-07, "loss": 1.1951, "step": 46030 }, { "epoch": 3.607445141065831, "grad_norm": 6.309090614318848, "learning_rate": 2.3668321859659836e-07, "loss": 1.5652, "step": 46031 }, { "epoch": 3.6075235109717867, "grad_norm": 3.6862783432006836, "learning_rate": 2.365894772064159e-07, "loss": 1.515, "step": 46032 }, { "epoch": 3.607601880877743, "grad_norm": 3.1164071559906006, "learning_rate": 2.3649575393379975e-07, "loss": 1.2489, "step": 46033 }, { "epoch": 3.607680250783699, "grad_norm": 3.2341232299804688, "learning_rate": 2.36402048779108e-07, "loss": 1.4578, "step": 46034 }, { "epoch": 3.6077586206896552, "grad_norm": 2.549600839614868, "learning_rate": 2.3630836174269534e-07, "loss": 1.3028, "step": 46035 }, { "epoch": 3.607836990595611, "grad_norm": 2.882904052734375, "learning_rate": 2.3621469282491927e-07, "loss": 1.3429, "step": 46036 }, { "epoch": 3.6079153605015675, "grad_norm": 5.512709617614746, "learning_rate": 2.3612104202613562e-07, "loss": 0.904, "step": 46037 }, { "epoch": 3.6079937304075234, "grad_norm": 3.2116363048553467, "learning_rate": 2.3602740934670078e-07, "loss": 0.9477, "step": 46038 }, { "epoch": 3.6080721003134797, "grad_norm": 3.0897092819213867, "learning_rate": 2.3593379478697055e-07, "loss": 1.2067, "step": 46039 }, { "epoch": 3.6081504702194356, "grad_norm": 11.142014503479004, "learning_rate": 2.3584019834730133e-07, "loss": 0.7314, "step": 46040 }, { "epoch": 3.608228840125392, "grad_norm": 2.916644811630249, "learning_rate": 2.3574662002804838e-07, "loss": 1.3331, "step": 46041 }, { "epoch": 3.608307210031348, "grad_norm": 2.407773971557617, "learning_rate": 2.3565305982956866e-07, "loss": 1.1837, "step": 46042 }, { "epoch": 3.608385579937304, "grad_norm": 3.3451194763183594, "learning_rate": 2.3555951775221798e-07, "loss": 1.2286, "step": 46043 }, { "epoch": 3.6084639498432605, "grad_norm": 3.064086437225342, "learning_rate": 2.3546599379635105e-07, "loss": 1.15, "step": 46044 }, { "epoch": 3.6085423197492164, "grad_norm": 5.6287617683410645, "learning_rate": 2.3537248796232537e-07, "loss": 1.2125, "step": 46045 }, { "epoch": 3.6086206896551722, "grad_norm": 3.6182844638824463, "learning_rate": 2.3527900025049456e-07, "loss": 1.6148, "step": 46046 }, { "epoch": 3.6086990595611286, "grad_norm": 3.1628966331481934, "learning_rate": 2.3518553066121498e-07, "loss": 1.4115, "step": 46047 }, { "epoch": 3.608777429467085, "grad_norm": 2.6754231452941895, "learning_rate": 2.35092079194843e-07, "loss": 1.0762, "step": 46048 }, { "epoch": 3.608855799373041, "grad_norm": 3.6523520946502686, "learning_rate": 2.3499864585173282e-07, "loss": 0.9995, "step": 46049 }, { "epoch": 3.6089341692789967, "grad_norm": 4.964539051055908, "learning_rate": 2.3490523063224025e-07, "loss": 1.3086, "step": 46050 }, { "epoch": 3.609012539184953, "grad_norm": 2.995849370956421, "learning_rate": 2.348118335367211e-07, "loss": 1.2768, "step": 46051 }, { "epoch": 3.6090909090909093, "grad_norm": 3.5482943058013916, "learning_rate": 2.3471845456552954e-07, "loss": 0.9879, "step": 46052 }, { "epoch": 3.6091692789968652, "grad_norm": 4.125287055969238, "learning_rate": 2.34625093719022e-07, "loss": 1.6197, "step": 46053 }, { "epoch": 3.609247648902821, "grad_norm": 3.2766196727752686, "learning_rate": 2.3453175099755255e-07, "loss": 0.8738, "step": 46054 }, { "epoch": 3.6093260188087775, "grad_norm": 2.4613449573516846, "learning_rate": 2.3443842640147653e-07, "loss": 1.5005, "step": 46055 }, { "epoch": 3.609404388714734, "grad_norm": 3.8295040130615234, "learning_rate": 2.3434511993114918e-07, "loss": 1.0498, "step": 46056 }, { "epoch": 3.6094827586206897, "grad_norm": 4.0959882736206055, "learning_rate": 2.3425183158692522e-07, "loss": 1.0671, "step": 46057 }, { "epoch": 3.6095611285266456, "grad_norm": 3.7343690395355225, "learning_rate": 2.3415856136915937e-07, "loss": 1.1296, "step": 46058 }, { "epoch": 3.609639498432602, "grad_norm": 2.2943925857543945, "learning_rate": 2.3406530927820692e-07, "loss": 1.3238, "step": 46059 }, { "epoch": 3.6097178683385582, "grad_norm": 4.434348106384277, "learning_rate": 2.3397207531442144e-07, "loss": 1.2809, "step": 46060 }, { "epoch": 3.609796238244514, "grad_norm": 2.5572445392608643, "learning_rate": 2.338788594781588e-07, "loss": 1.2982, "step": 46061 }, { "epoch": 3.60987460815047, "grad_norm": 2.9380886554718018, "learning_rate": 2.3378566176977257e-07, "loss": 1.3243, "step": 46062 }, { "epoch": 3.6099529780564263, "grad_norm": 2.7560553550720215, "learning_rate": 2.3369248218961804e-07, "loss": 1.6551, "step": 46063 }, { "epoch": 3.6100313479623827, "grad_norm": 2.7064104080200195, "learning_rate": 2.3359932073804881e-07, "loss": 1.2813, "step": 46064 }, { "epoch": 3.6101097178683386, "grad_norm": 4.3896050453186035, "learning_rate": 2.3350617741541903e-07, "loss": 1.2357, "step": 46065 }, { "epoch": 3.6101880877742945, "grad_norm": 3.8514089584350586, "learning_rate": 2.33413052222084e-07, "loss": 1.9404, "step": 46066 }, { "epoch": 3.610266457680251, "grad_norm": 3.9401473999023438, "learning_rate": 2.333199451583984e-07, "loss": 1.0645, "step": 46067 }, { "epoch": 3.610344827586207, "grad_norm": 2.8840842247009277, "learning_rate": 2.332268562247142e-07, "loss": 1.3481, "step": 46068 }, { "epoch": 3.610423197492163, "grad_norm": 3.1158297061920166, "learning_rate": 2.331337854213872e-07, "loss": 1.2517, "step": 46069 }, { "epoch": 3.610501567398119, "grad_norm": 3.576735496520996, "learning_rate": 2.3304073274876993e-07, "loss": 1.4107, "step": 46070 }, { "epoch": 3.6105799373040752, "grad_norm": 3.170383930206299, "learning_rate": 2.3294769820721874e-07, "loss": 1.3977, "step": 46071 }, { "epoch": 3.6106583072100316, "grad_norm": 10.28132152557373, "learning_rate": 2.3285468179708504e-07, "loss": 0.8902, "step": 46072 }, { "epoch": 3.6107366771159874, "grad_norm": 3.503138780593872, "learning_rate": 2.3276168351872352e-07, "loss": 1.4675, "step": 46073 }, { "epoch": 3.6108150470219433, "grad_norm": 3.0957822799682617, "learning_rate": 2.3266870337248836e-07, "loss": 1.3079, "step": 46074 }, { "epoch": 3.6108934169278997, "grad_norm": 3.843039035797119, "learning_rate": 2.3257574135873263e-07, "loss": 1.643, "step": 46075 }, { "epoch": 3.610971786833856, "grad_norm": 2.613478183746338, "learning_rate": 2.324827974778099e-07, "loss": 1.1402, "step": 46076 }, { "epoch": 3.611050156739812, "grad_norm": 3.9352729320526123, "learning_rate": 2.3238987173007433e-07, "loss": 1.4789, "step": 46077 }, { "epoch": 3.6111285266457678, "grad_norm": 2.719559669494629, "learning_rate": 2.3229696411587898e-07, "loss": 1.5055, "step": 46078 }, { "epoch": 3.611206896551724, "grad_norm": 2.167078733444214, "learning_rate": 2.3220407463557637e-07, "loss": 1.2367, "step": 46079 }, { "epoch": 3.6112852664576804, "grad_norm": 4.1492109298706055, "learning_rate": 2.3211120328952063e-07, "loss": 1.3894, "step": 46080 }, { "epoch": 3.6113636363636363, "grad_norm": 3.3410136699676514, "learning_rate": 2.320183500780654e-07, "loss": 1.1732, "step": 46081 }, { "epoch": 3.611442006269592, "grad_norm": 3.246821641921997, "learning_rate": 2.3192551500156367e-07, "loss": 1.3375, "step": 46082 }, { "epoch": 3.6115203761755486, "grad_norm": 3.1541736125946045, "learning_rate": 2.31832698060368e-07, "loss": 1.2056, "step": 46083 }, { "epoch": 3.611598746081505, "grad_norm": 4.007569789886475, "learning_rate": 2.3173989925483088e-07, "loss": 1.1989, "step": 46084 }, { "epoch": 3.6116771159874608, "grad_norm": 5.367711067199707, "learning_rate": 2.3164711858530642e-07, "loss": 1.4636, "step": 46085 }, { "epoch": 3.6117554858934167, "grad_norm": 3.199366569519043, "learning_rate": 2.3155435605214772e-07, "loss": 0.8355, "step": 46086 }, { "epoch": 3.611833855799373, "grad_norm": 3.0699880123138428, "learning_rate": 2.3146161165570668e-07, "loss": 1.3088, "step": 46087 }, { "epoch": 3.6119122257053293, "grad_norm": 4.938163757324219, "learning_rate": 2.313688853963364e-07, "loss": 1.2436, "step": 46088 }, { "epoch": 3.611990595611285, "grad_norm": 3.198272943496704, "learning_rate": 2.3127617727438933e-07, "loss": 1.2561, "step": 46089 }, { "epoch": 3.612068965517241, "grad_norm": 4.319615840911865, "learning_rate": 2.3118348729021856e-07, "loss": 1.0226, "step": 46090 }, { "epoch": 3.6121473354231974, "grad_norm": 2.5048511028289795, "learning_rate": 2.3109081544417657e-07, "loss": 1.3228, "step": 46091 }, { "epoch": 3.6122257053291538, "grad_norm": 3.317237138748169, "learning_rate": 2.3099816173661526e-07, "loss": 1.2319, "step": 46092 }, { "epoch": 3.6123040752351097, "grad_norm": 4.52500057220459, "learning_rate": 2.3090552616788774e-07, "loss": 1.287, "step": 46093 }, { "epoch": 3.612382445141066, "grad_norm": 2.706397294998169, "learning_rate": 2.3081290873834539e-07, "loss": 1.2906, "step": 46094 }, { "epoch": 3.612460815047022, "grad_norm": 3.645033597946167, "learning_rate": 2.307203094483418e-07, "loss": 1.017, "step": 46095 }, { "epoch": 3.612539184952978, "grad_norm": 4.09822940826416, "learning_rate": 2.3062772829822832e-07, "loss": 1.2658, "step": 46096 }, { "epoch": 3.612617554858934, "grad_norm": 2.8292737007141113, "learning_rate": 2.305351652883575e-07, "loss": 1.3665, "step": 46097 }, { "epoch": 3.6126959247648904, "grad_norm": 4.349719524383545, "learning_rate": 2.304426204190807e-07, "loss": 1.1087, "step": 46098 }, { "epoch": 3.6127742946708463, "grad_norm": 2.524447441101074, "learning_rate": 2.3035009369075046e-07, "loss": 1.3728, "step": 46099 }, { "epoch": 3.6128526645768027, "grad_norm": 3.025508403778076, "learning_rate": 2.3025758510371865e-07, "loss": 1.0008, "step": 46100 }, { "epoch": 3.6129310344827585, "grad_norm": 4.90463924407959, "learning_rate": 2.301650946583378e-07, "loss": 1.522, "step": 46101 }, { "epoch": 3.613009404388715, "grad_norm": 2.3725008964538574, "learning_rate": 2.3007262235495819e-07, "loss": 1.2039, "step": 46102 }, { "epoch": 3.6130877742946708, "grad_norm": 5.089916229248047, "learning_rate": 2.2998016819393177e-07, "loss": 1.4074, "step": 46103 }, { "epoch": 3.613166144200627, "grad_norm": 2.9051668643951416, "learning_rate": 2.2988773217561212e-07, "loss": 1.2913, "step": 46104 }, { "epoch": 3.613244514106583, "grad_norm": 3.0936248302459717, "learning_rate": 2.2979531430034896e-07, "loss": 1.199, "step": 46105 }, { "epoch": 3.6133228840125393, "grad_norm": 4.536638259887695, "learning_rate": 2.2970291456849424e-07, "loss": 1.6423, "step": 46106 }, { "epoch": 3.613401253918495, "grad_norm": 4.933422565460205, "learning_rate": 2.2961053298039937e-07, "loss": 0.5076, "step": 46107 }, { "epoch": 3.6134796238244515, "grad_norm": 4.78391170501709, "learning_rate": 2.295181695364157e-07, "loss": 1.4374, "step": 46108 }, { "epoch": 3.6135579937304074, "grad_norm": 4.542562007904053, "learning_rate": 2.2942582423689463e-07, "loss": 1.4515, "step": 46109 }, { "epoch": 3.6136363636363638, "grad_norm": 3.3385508060455322, "learning_rate": 2.2933349708218756e-07, "loss": 1.2509, "step": 46110 }, { "epoch": 3.6137147335423196, "grad_norm": 13.04096794128418, "learning_rate": 2.2924118807264528e-07, "loss": 1.2963, "step": 46111 }, { "epoch": 3.613793103448276, "grad_norm": 4.740536212921143, "learning_rate": 2.2914889720861922e-07, "loss": 1.3754, "step": 46112 }, { "epoch": 3.613871473354232, "grad_norm": 2.783649444580078, "learning_rate": 2.2905662449046018e-07, "loss": 1.355, "step": 46113 }, { "epoch": 3.613949843260188, "grad_norm": 3.718646287918091, "learning_rate": 2.2896436991851956e-07, "loss": 1.2506, "step": 46114 }, { "epoch": 3.614028213166144, "grad_norm": 2.076364278793335, "learning_rate": 2.2887213349314763e-07, "loss": 1.3365, "step": 46115 }, { "epoch": 3.6141065830721004, "grad_norm": 7.562777996063232, "learning_rate": 2.2877991521469633e-07, "loss": 0.9287, "step": 46116 }, { "epoch": 3.6141849529780563, "grad_norm": 2.412649631500244, "learning_rate": 2.2868771508351429e-07, "loss": 1.2444, "step": 46117 }, { "epoch": 3.6142633228840126, "grad_norm": 2.6109766960144043, "learning_rate": 2.2859553309995398e-07, "loss": 1.1246, "step": 46118 }, { "epoch": 3.6143416927899685, "grad_norm": 4.007666110992432, "learning_rate": 2.2850336926436622e-07, "loss": 1.4745, "step": 46119 }, { "epoch": 3.614420062695925, "grad_norm": 3.8917605876922607, "learning_rate": 2.284112235771002e-07, "loss": 1.3647, "step": 46120 }, { "epoch": 3.6144984326018808, "grad_norm": 4.379809379577637, "learning_rate": 2.2831909603850678e-07, "loss": 1.6687, "step": 46121 }, { "epoch": 3.614576802507837, "grad_norm": 3.849323034286499, "learning_rate": 2.2822698664893616e-07, "loss": 1.3703, "step": 46122 }, { "epoch": 3.614655172413793, "grad_norm": 12.612275123596191, "learning_rate": 2.281348954087409e-07, "loss": 1.2564, "step": 46123 }, { "epoch": 3.6147335423197493, "grad_norm": 3.8647332191467285, "learning_rate": 2.2804282231826846e-07, "loss": 1.198, "step": 46124 }, { "epoch": 3.614811912225705, "grad_norm": 2.9792206287384033, "learning_rate": 2.2795076737786969e-07, "loss": 1.2045, "step": 46125 }, { "epoch": 3.6148902821316615, "grad_norm": 6.67997932434082, "learning_rate": 2.2785873058789598e-07, "loss": 1.9224, "step": 46126 }, { "epoch": 3.6149686520376174, "grad_norm": 3.652022361755371, "learning_rate": 2.2776671194869592e-07, "loss": 1.2386, "step": 46127 }, { "epoch": 3.6150470219435737, "grad_norm": 9.659135818481445, "learning_rate": 2.2767471146062037e-07, "loss": 1.1002, "step": 46128 }, { "epoch": 3.6151253918495296, "grad_norm": 3.1682658195495605, "learning_rate": 2.2758272912401902e-07, "loss": 1.4497, "step": 46129 }, { "epoch": 3.615203761755486, "grad_norm": 3.012446403503418, "learning_rate": 2.2749076493924162e-07, "loss": 1.2378, "step": 46130 }, { "epoch": 3.615282131661442, "grad_norm": 2.796138286590576, "learning_rate": 2.2739881890663785e-07, "loss": 1.272, "step": 46131 }, { "epoch": 3.615360501567398, "grad_norm": 3.270970344543457, "learning_rate": 2.2730689102655746e-07, "loss": 1.4713, "step": 46132 }, { "epoch": 3.615438871473354, "grad_norm": 3.9976751804351807, "learning_rate": 2.272149812993507e-07, "loss": 0.7215, "step": 46133 }, { "epoch": 3.6155172413793104, "grad_norm": 2.3275177478790283, "learning_rate": 2.2712308972536735e-07, "loss": 1.3335, "step": 46134 }, { "epoch": 3.6155956112852663, "grad_norm": 4.8343825340271, "learning_rate": 2.270312163049554e-07, "loss": 1.4172, "step": 46135 }, { "epoch": 3.6156739811912226, "grad_norm": 4.7939629554748535, "learning_rate": 2.269393610384646e-07, "loss": 1.16, "step": 46136 }, { "epoch": 3.6157523510971785, "grad_norm": 3.2282848358154297, "learning_rate": 2.268475239262452e-07, "loss": 1.2244, "step": 46137 }, { "epoch": 3.615830721003135, "grad_norm": 2.79447340965271, "learning_rate": 2.267557049686475e-07, "loss": 1.3866, "step": 46138 }, { "epoch": 3.615909090909091, "grad_norm": 3.6386756896972656, "learning_rate": 2.2666390416601848e-07, "loss": 1.0816, "step": 46139 }, { "epoch": 3.615987460815047, "grad_norm": 2.7489309310913086, "learning_rate": 2.2657212151870777e-07, "loss": 1.2803, "step": 46140 }, { "epoch": 3.616065830721003, "grad_norm": 13.519668579101562, "learning_rate": 2.264803570270646e-07, "loss": 0.8891, "step": 46141 }, { "epoch": 3.6161442006269593, "grad_norm": 13.785670280456543, "learning_rate": 2.2638861069143982e-07, "loss": 1.5708, "step": 46142 }, { "epoch": 3.6162225705329156, "grad_norm": 2.923659086227417, "learning_rate": 2.2629688251217973e-07, "loss": 1.5873, "step": 46143 }, { "epoch": 3.6163009404388715, "grad_norm": 2.7450497150421143, "learning_rate": 2.262051724896347e-07, "loss": 1.1633, "step": 46144 }, { "epoch": 3.6163793103448274, "grad_norm": 2.5456924438476562, "learning_rate": 2.2611348062415328e-07, "loss": 0.9882, "step": 46145 }, { "epoch": 3.6164576802507837, "grad_norm": 5.943333148956299, "learning_rate": 2.260218069160841e-07, "loss": 1.638, "step": 46146 }, { "epoch": 3.61653605015674, "grad_norm": 3.0662925243377686, "learning_rate": 2.2593015136577578e-07, "loss": 1.0145, "step": 46147 }, { "epoch": 3.616614420062696, "grad_norm": 4.516340732574463, "learning_rate": 2.2583851397357747e-07, "loss": 1.6873, "step": 46148 }, { "epoch": 3.616692789968652, "grad_norm": 6.085378170013428, "learning_rate": 2.257468947398367e-07, "loss": 1.4953, "step": 46149 }, { "epoch": 3.616771159874608, "grad_norm": 2.274630069732666, "learning_rate": 2.2565529366490312e-07, "loss": 1.2239, "step": 46150 }, { "epoch": 3.6168495297805645, "grad_norm": 2.98370099067688, "learning_rate": 2.2556371074912375e-07, "loss": 0.4725, "step": 46151 }, { "epoch": 3.6169278996865204, "grad_norm": 4.695104598999023, "learning_rate": 2.2547214599284828e-07, "loss": 1.1683, "step": 46152 }, { "epoch": 3.6170062695924763, "grad_norm": 3.2409675121307373, "learning_rate": 2.2538059939642532e-07, "loss": 1.3668, "step": 46153 }, { "epoch": 3.6170846394984326, "grad_norm": 4.253515720367432, "learning_rate": 2.2528907096020014e-07, "loss": 1.1328, "step": 46154 }, { "epoch": 3.617163009404389, "grad_norm": 3.9004812240600586, "learning_rate": 2.2519756068452414e-07, "loss": 1.1652, "step": 46155 }, { "epoch": 3.617241379310345, "grad_norm": 5.839242935180664, "learning_rate": 2.251060685697437e-07, "loss": 1.2915, "step": 46156 }, { "epoch": 3.6173197492163007, "grad_norm": 4.1189751625061035, "learning_rate": 2.2501459461620802e-07, "loss": 1.5818, "step": 46157 }, { "epoch": 3.617398119122257, "grad_norm": 4.387969017028809, "learning_rate": 2.2492313882426397e-07, "loss": 0.9563, "step": 46158 }, { "epoch": 3.6174764890282134, "grad_norm": 4.61578893661499, "learning_rate": 2.2483170119425857e-07, "loss": 1.148, "step": 46159 }, { "epoch": 3.6175548589341693, "grad_norm": 3.5111448764801025, "learning_rate": 2.2474028172654206e-07, "loss": 1.2923, "step": 46160 }, { "epoch": 3.617633228840125, "grad_norm": 3.768563747406006, "learning_rate": 2.2464888042146028e-07, "loss": 0.954, "step": 46161 }, { "epoch": 3.6177115987460815, "grad_norm": 4.02871036529541, "learning_rate": 2.2455749727936128e-07, "loss": 1.3553, "step": 46162 }, { "epoch": 3.617789968652038, "grad_norm": 2.735520601272583, "learning_rate": 2.2446613230059256e-07, "loss": 1.5686, "step": 46163 }, { "epoch": 3.6178683385579937, "grad_norm": 3.55936861038208, "learning_rate": 2.2437478548550217e-07, "loss": 1.6935, "step": 46164 }, { "epoch": 3.6179467084639496, "grad_norm": 7.87298583984375, "learning_rate": 2.2428345683443707e-07, "loss": 0.992, "step": 46165 }, { "epoch": 3.618025078369906, "grad_norm": 2.4658212661743164, "learning_rate": 2.2419214634774477e-07, "loss": 1.0682, "step": 46166 }, { "epoch": 3.6181034482758623, "grad_norm": 2.4988648891448975, "learning_rate": 2.241008540257722e-07, "loss": 1.5614, "step": 46167 }, { "epoch": 3.618181818181818, "grad_norm": 5.368619918823242, "learning_rate": 2.240095798688674e-07, "loss": 1.965, "step": 46168 }, { "epoch": 3.618260188087774, "grad_norm": 4.641422271728516, "learning_rate": 2.2391832387737678e-07, "loss": 0.7808, "step": 46169 }, { "epoch": 3.6183385579937304, "grad_norm": 11.246370315551758, "learning_rate": 2.238270860516478e-07, "loss": 1.6197, "step": 46170 }, { "epoch": 3.6184169278996867, "grad_norm": 4.321836471557617, "learning_rate": 2.2373586639202694e-07, "loss": 1.2053, "step": 46171 }, { "epoch": 3.6184952978056426, "grad_norm": 4.8740153312683105, "learning_rate": 2.2364466489886272e-07, "loss": 0.9999, "step": 46172 }, { "epoch": 3.6185736677115985, "grad_norm": 4.779038429260254, "learning_rate": 2.235534815724999e-07, "loss": 1.6139, "step": 46173 }, { "epoch": 3.618652037617555, "grad_norm": 2.391411542892456, "learning_rate": 2.2346231641328654e-07, "loss": 1.5368, "step": 46174 }, { "epoch": 3.618730407523511, "grad_norm": 11.457253456115723, "learning_rate": 2.2337116942156957e-07, "loss": 0.8845, "step": 46175 }, { "epoch": 3.618808777429467, "grad_norm": 3.025912284851074, "learning_rate": 2.232800405976948e-07, "loss": 1.0686, "step": 46176 }, { "epoch": 3.618887147335423, "grad_norm": 2.716456174850464, "learning_rate": 2.2318892994200925e-07, "loss": 1.1609, "step": 46177 }, { "epoch": 3.6189655172413793, "grad_norm": 3.8498423099517822, "learning_rate": 2.2309783745485925e-07, "loss": 1.5923, "step": 46178 }, { "epoch": 3.6190438871473356, "grad_norm": 4.293030738830566, "learning_rate": 2.2300676313659231e-07, "loss": 1.4862, "step": 46179 }, { "epoch": 3.6191222570532915, "grad_norm": 5.434839248657227, "learning_rate": 2.229157069875537e-07, "loss": 1.3051, "step": 46180 }, { "epoch": 3.6192006269592474, "grad_norm": 2.9708144664764404, "learning_rate": 2.2282466900808986e-07, "loss": 1.2396, "step": 46181 }, { "epoch": 3.6192789968652037, "grad_norm": 4.841757774353027, "learning_rate": 2.2273364919854713e-07, "loss": 1.5202, "step": 46182 }, { "epoch": 3.61935736677116, "grad_norm": 3.673539400100708, "learning_rate": 2.226426475592719e-07, "loss": 1.1928, "step": 46183 }, { "epoch": 3.619435736677116, "grad_norm": 8.80463695526123, "learning_rate": 2.2255166409061057e-07, "loss": 1.1804, "step": 46184 }, { "epoch": 3.6195141065830723, "grad_norm": 2.7296149730682373, "learning_rate": 2.2246069879290843e-07, "loss": 1.0204, "step": 46185 }, { "epoch": 3.619592476489028, "grad_norm": 3.443511486053467, "learning_rate": 2.22369751666513e-07, "loss": 1.3609, "step": 46186 }, { "epoch": 3.6196708463949845, "grad_norm": 2.6930301189422607, "learning_rate": 2.2227882271176726e-07, "loss": 0.854, "step": 46187 }, { "epoch": 3.6197492163009404, "grad_norm": 3.224848985671997, "learning_rate": 2.2218791192901989e-07, "loss": 1.3792, "step": 46188 }, { "epoch": 3.6198275862068967, "grad_norm": 4.680380821228027, "learning_rate": 2.220970193186156e-07, "loss": 1.3559, "step": 46189 }, { "epoch": 3.6199059561128526, "grad_norm": 2.7391724586486816, "learning_rate": 2.2200614488090077e-07, "loss": 1.2702, "step": 46190 }, { "epoch": 3.619984326018809, "grad_norm": 7.081830024719238, "learning_rate": 2.2191528861622015e-07, "loss": 1.045, "step": 46191 }, { "epoch": 3.620062695924765, "grad_norm": 5.07822847366333, "learning_rate": 2.2182445052491842e-07, "loss": 1.3117, "step": 46192 }, { "epoch": 3.620141065830721, "grad_norm": 12.559188842773438, "learning_rate": 2.217336306073431e-07, "loss": 0.5994, "step": 46193 }, { "epoch": 3.620219435736677, "grad_norm": 5.805581569671631, "learning_rate": 2.2164282886383946e-07, "loss": 1.0279, "step": 46194 }, { "epoch": 3.6202978056426334, "grad_norm": 10.910165786743164, "learning_rate": 2.2155204529475172e-07, "loss": 1.3275, "step": 46195 }, { "epoch": 3.6203761755485893, "grad_norm": 3.684652328491211, "learning_rate": 2.214612799004251e-07, "loss": 0.8354, "step": 46196 }, { "epoch": 3.6204545454545456, "grad_norm": 3.9541244506835938, "learning_rate": 2.213705326812049e-07, "loss": 1.3317, "step": 46197 }, { "epoch": 3.6205329153605015, "grad_norm": 3.354153633117676, "learning_rate": 2.2127980363743862e-07, "loss": 1.8847, "step": 46198 }, { "epoch": 3.620611285266458, "grad_norm": 4.251291275024414, "learning_rate": 2.211890927694682e-07, "loss": 1.1425, "step": 46199 }, { "epoch": 3.6206896551724137, "grad_norm": 2.0324621200561523, "learning_rate": 2.2109840007764005e-07, "loss": 1.5216, "step": 46200 }, { "epoch": 3.62076802507837, "grad_norm": 2.7538914680480957, "learning_rate": 2.2100772556229888e-07, "loss": 1.0852, "step": 46201 }, { "epoch": 3.620846394984326, "grad_norm": 2.514439344406128, "learning_rate": 2.2091706922378997e-07, "loss": 1.2707, "step": 46202 }, { "epoch": 3.6209247648902823, "grad_norm": 5.284080505371094, "learning_rate": 2.2082643106245748e-07, "loss": 0.6745, "step": 46203 }, { "epoch": 3.621003134796238, "grad_norm": 4.539114952087402, "learning_rate": 2.2073581107864672e-07, "loss": 1.3723, "step": 46204 }, { "epoch": 3.6210815047021945, "grad_norm": 2.6772234439849854, "learning_rate": 2.2064520927270294e-07, "loss": 1.3297, "step": 46205 }, { "epoch": 3.6211598746081504, "grad_norm": 2.8947253227233887, "learning_rate": 2.205546256449681e-07, "loss": 1.3567, "step": 46206 }, { "epoch": 3.6212382445141067, "grad_norm": 3.233273506164551, "learning_rate": 2.204640601957897e-07, "loss": 1.5122, "step": 46207 }, { "epoch": 3.6213166144200626, "grad_norm": 4.3074235916137695, "learning_rate": 2.2037351292551136e-07, "loss": 1.8074, "step": 46208 }, { "epoch": 3.621394984326019, "grad_norm": 4.36964750289917, "learning_rate": 2.2028298383447722e-07, "loss": 1.3537, "step": 46209 }, { "epoch": 3.621473354231975, "grad_norm": 7.114173412322998, "learning_rate": 2.2019247292303148e-07, "loss": 1.134, "step": 46210 }, { "epoch": 3.621551724137931, "grad_norm": 2.481667995452881, "learning_rate": 2.201019801915172e-07, "loss": 1.3245, "step": 46211 }, { "epoch": 3.621630094043887, "grad_norm": 3.2032864093780518, "learning_rate": 2.2001150564028129e-07, "loss": 1.4009, "step": 46212 }, { "epoch": 3.6217084639498434, "grad_norm": 6.074521541595459, "learning_rate": 2.199210492696663e-07, "loss": 1.5955, "step": 46213 }, { "epoch": 3.6217868338557992, "grad_norm": 5.055317401885986, "learning_rate": 2.1983061108001637e-07, "loss": 1.1639, "step": 46214 }, { "epoch": 3.6218652037617556, "grad_norm": 4.301307201385498, "learning_rate": 2.1974019107167565e-07, "loss": 1.4227, "step": 46215 }, { "epoch": 3.6219435736677115, "grad_norm": 5.518097400665283, "learning_rate": 2.1964978924498725e-07, "loss": 1.3117, "step": 46216 }, { "epoch": 3.622021943573668, "grad_norm": 4.276698589324951, "learning_rate": 2.1955940560029643e-07, "loss": 1.2784, "step": 46217 }, { "epoch": 3.6221003134796237, "grad_norm": 2.915156126022339, "learning_rate": 2.1946904013794623e-07, "loss": 1.4169, "step": 46218 }, { "epoch": 3.62217868338558, "grad_norm": 3.7413735389709473, "learning_rate": 2.1937869285827973e-07, "loss": 0.7875, "step": 46219 }, { "epoch": 3.622257053291536, "grad_norm": 4.8699564933776855, "learning_rate": 2.1928836376164164e-07, "loss": 1.342, "step": 46220 }, { "epoch": 3.6223354231974922, "grad_norm": 2.929361343383789, "learning_rate": 2.1919805284837502e-07, "loss": 1.3722, "step": 46221 }, { "epoch": 3.622413793103448, "grad_norm": 4.06155252456665, "learning_rate": 2.191077601188235e-07, "loss": 1.2652, "step": 46222 }, { "epoch": 3.6224921630094045, "grad_norm": 2.6677160263061523, "learning_rate": 2.1901748557333013e-07, "loss": 1.2628, "step": 46223 }, { "epoch": 3.6225705329153604, "grad_norm": 3.5080292224884033, "learning_rate": 2.1892722921223964e-07, "loss": 1.9914, "step": 46224 }, { "epoch": 3.6226489028213167, "grad_norm": 3.3373100757598877, "learning_rate": 2.1883699103589285e-07, "loss": 1.2271, "step": 46225 }, { "epoch": 3.6227272727272726, "grad_norm": 3.7030677795410156, "learning_rate": 2.1874677104463505e-07, "loss": 1.2621, "step": 46226 }, { "epoch": 3.622805642633229, "grad_norm": 2.9113330841064453, "learning_rate": 2.1865656923880873e-07, "loss": 1.2066, "step": 46227 }, { "epoch": 3.622884012539185, "grad_norm": 8.55373764038086, "learning_rate": 2.1856638561875754e-07, "loss": 1.5405, "step": 46228 }, { "epoch": 3.622962382445141, "grad_norm": 3.6175079345703125, "learning_rate": 2.1847622018482283e-07, "loss": 1.2245, "step": 46229 }, { "epoch": 3.623040752351097, "grad_norm": 6.642817497253418, "learning_rate": 2.1838607293734882e-07, "loss": 1.7696, "step": 46230 }, { "epoch": 3.6231191222570533, "grad_norm": 6.0894646644592285, "learning_rate": 2.1829594387667906e-07, "loss": 0.8834, "step": 46231 }, { "epoch": 3.6231974921630092, "grad_norm": 3.3637852668762207, "learning_rate": 2.1820583300315445e-07, "loss": 1.3193, "step": 46232 }, { "epoch": 3.6232758620689656, "grad_norm": 2.1859164237976074, "learning_rate": 2.181157403171197e-07, "loss": 1.2505, "step": 46233 }, { "epoch": 3.623354231974922, "grad_norm": 3.063624382019043, "learning_rate": 2.180256658189156e-07, "loss": 1.5309, "step": 46234 }, { "epoch": 3.623432601880878, "grad_norm": 4.713334560394287, "learning_rate": 2.1793560950888637e-07, "loss": 1.1209, "step": 46235 }, { "epoch": 3.6235109717868337, "grad_norm": 3.8460609912872314, "learning_rate": 2.178455713873734e-07, "loss": 1.4707, "step": 46236 }, { "epoch": 3.62358934169279, "grad_norm": 3.4406120777130127, "learning_rate": 2.177555514547197e-07, "loss": 1.1944, "step": 46237 }, { "epoch": 3.6236677115987463, "grad_norm": 3.5112295150756836, "learning_rate": 2.1766554971126785e-07, "loss": 1.1036, "step": 46238 }, { "epoch": 3.6237460815047022, "grad_norm": 3.301137685775757, "learning_rate": 2.1757556615735975e-07, "loss": 1.0137, "step": 46239 }, { "epoch": 3.623824451410658, "grad_norm": 14.010926246643066, "learning_rate": 2.174856007933379e-07, "loss": 1.9596, "step": 46240 }, { "epoch": 3.6239028213166145, "grad_norm": 6.576657772064209, "learning_rate": 2.1739565361954428e-07, "loss": 0.9713, "step": 46241 }, { "epoch": 3.623981191222571, "grad_norm": 4.255260944366455, "learning_rate": 2.1730572463632138e-07, "loss": 1.4293, "step": 46242 }, { "epoch": 3.6240595611285267, "grad_norm": 2.76900315284729, "learning_rate": 2.172158138440106e-07, "loss": 1.3547, "step": 46243 }, { "epoch": 3.6241379310344826, "grad_norm": 3.650088310241699, "learning_rate": 2.1712592124295384e-07, "loss": 1.2929, "step": 46244 }, { "epoch": 3.624216300940439, "grad_norm": 3.5441935062408447, "learning_rate": 2.1703604683349367e-07, "loss": 1.3444, "step": 46245 }, { "epoch": 3.6242946708463952, "grad_norm": 3.3847100734710693, "learning_rate": 2.1694619061597254e-07, "loss": 1.3228, "step": 46246 }, { "epoch": 3.624373040752351, "grad_norm": 3.1665399074554443, "learning_rate": 2.1685635259073078e-07, "loss": 1.3864, "step": 46247 }, { "epoch": 3.624451410658307, "grad_norm": 11.834527969360352, "learning_rate": 2.167665327581109e-07, "loss": 1.2765, "step": 46248 }, { "epoch": 3.6245297805642633, "grad_norm": 4.45916223526001, "learning_rate": 2.1667673111845312e-07, "loss": 1.1781, "step": 46249 }, { "epoch": 3.6246081504702197, "grad_norm": 5.812427043914795, "learning_rate": 2.165869476721022e-07, "loss": 1.7143, "step": 46250 }, { "epoch": 3.6246865203761756, "grad_norm": 3.2543320655822754, "learning_rate": 2.1649718241939622e-07, "loss": 1.4687, "step": 46251 }, { "epoch": 3.6247648902821314, "grad_norm": 3.4425079822540283, "learning_rate": 2.164074353606782e-07, "loss": 1.4666, "step": 46252 }, { "epoch": 3.624843260188088, "grad_norm": 5.931040287017822, "learning_rate": 2.1631770649628958e-07, "loss": 1.5636, "step": 46253 }, { "epoch": 3.624921630094044, "grad_norm": 2.880774974822998, "learning_rate": 2.1622799582657118e-07, "loss": 0.8378, "step": 46254 }, { "epoch": 3.625, "grad_norm": 2.1598258018493652, "learning_rate": 2.1613830335186436e-07, "loss": 1.3963, "step": 46255 }, { "epoch": 3.625078369905956, "grad_norm": 3.746879816055298, "learning_rate": 2.1604862907251e-07, "loss": 1.5881, "step": 46256 }, { "epoch": 3.625156739811912, "grad_norm": 2.3340389728546143, "learning_rate": 2.1595897298885005e-07, "loss": 1.25, "step": 46257 }, { "epoch": 3.6252351097178686, "grad_norm": 2.4930384159088135, "learning_rate": 2.158693351012242e-07, "loss": 1.0786, "step": 46258 }, { "epoch": 3.6253134796238244, "grad_norm": 2.8130388259887695, "learning_rate": 2.15779715409975e-07, "loss": 1.2219, "step": 46259 }, { "epoch": 3.6253918495297803, "grad_norm": 4.702229022979736, "learning_rate": 2.1569011391544159e-07, "loss": 1.4438, "step": 46260 }, { "epoch": 3.6254702194357367, "grad_norm": 5.797143459320068, "learning_rate": 2.1560053061796704e-07, "loss": 2.0874, "step": 46261 }, { "epoch": 3.625548589341693, "grad_norm": 5.593206882476807, "learning_rate": 2.1551096551788887e-07, "loss": 1.316, "step": 46262 }, { "epoch": 3.625626959247649, "grad_norm": 5.252601146697998, "learning_rate": 2.154214186155501e-07, "loss": 1.7081, "step": 46263 }, { "epoch": 3.6257053291536048, "grad_norm": 2.273728847503662, "learning_rate": 2.1533188991129107e-07, "loss": 1.4164, "step": 46264 }, { "epoch": 3.625783699059561, "grad_norm": 2.3895151615142822, "learning_rate": 2.15242379405452e-07, "loss": 0.6919, "step": 46265 }, { "epoch": 3.6258620689655174, "grad_norm": 10.941308975219727, "learning_rate": 2.1515288709837322e-07, "loss": 2.1569, "step": 46266 }, { "epoch": 3.6259404388714733, "grad_norm": 5.455001354217529, "learning_rate": 2.1506341299039445e-07, "loss": 1.384, "step": 46267 }, { "epoch": 3.626018808777429, "grad_norm": 2.8679182529449463, "learning_rate": 2.1497395708185764e-07, "loss": 1.2405, "step": 46268 }, { "epoch": 3.6260971786833855, "grad_norm": 3.4850707054138184, "learning_rate": 2.148845193731025e-07, "loss": 1.5268, "step": 46269 }, { "epoch": 3.626175548589342, "grad_norm": 2.512852907180786, "learning_rate": 2.1479509986446822e-07, "loss": 1.3092, "step": 46270 }, { "epoch": 3.6262539184952978, "grad_norm": 2.3616058826446533, "learning_rate": 2.1470569855629564e-07, "loss": 1.2101, "step": 46271 }, { "epoch": 3.6263322884012537, "grad_norm": 2.947120428085327, "learning_rate": 2.1461631544892447e-07, "loss": 1.4188, "step": 46272 }, { "epoch": 3.62641065830721, "grad_norm": 4.33591890335083, "learning_rate": 2.1452695054269502e-07, "loss": 1.2707, "step": 46273 }, { "epoch": 3.6264890282131663, "grad_norm": 4.616139888763428, "learning_rate": 2.1443760383794698e-07, "loss": 1.3669, "step": 46274 }, { "epoch": 3.626567398119122, "grad_norm": 4.667892932891846, "learning_rate": 2.1434827533502067e-07, "loss": 1.3369, "step": 46275 }, { "epoch": 3.626645768025078, "grad_norm": 2.2631030082702637, "learning_rate": 2.1425896503425525e-07, "loss": 1.1994, "step": 46276 }, { "epoch": 3.6267241379310344, "grad_norm": 3.599813938140869, "learning_rate": 2.1416967293599046e-07, "loss": 1.1204, "step": 46277 }, { "epoch": 3.6268025078369908, "grad_norm": 3.5307719707489014, "learning_rate": 2.1408039904056654e-07, "loss": 1.2598, "step": 46278 }, { "epoch": 3.6268808777429467, "grad_norm": 3.871710777282715, "learning_rate": 2.1399114334832217e-07, "loss": 1.2344, "step": 46279 }, { "epoch": 3.626959247648903, "grad_norm": 6.0466413497924805, "learning_rate": 2.1390190585959757e-07, "loss": 1.9399, "step": 46280 }, { "epoch": 3.627037617554859, "grad_norm": 2.386920928955078, "learning_rate": 2.1381268657473087e-07, "loss": 1.2472, "step": 46281 }, { "epoch": 3.627115987460815, "grad_norm": 5.203794956207275, "learning_rate": 2.137234854940634e-07, "loss": 0.8659, "step": 46282 }, { "epoch": 3.627194357366771, "grad_norm": 10.045039176940918, "learning_rate": 2.1363430261793272e-07, "loss": 0.9021, "step": 46283 }, { "epoch": 3.6272727272727274, "grad_norm": 2.5298237800598145, "learning_rate": 2.135451379466802e-07, "loss": 1.4267, "step": 46284 }, { "epoch": 3.6273510971786833, "grad_norm": 4.949272155761719, "learning_rate": 2.1345599148064222e-07, "loss": 1.7332, "step": 46285 }, { "epoch": 3.6274294670846396, "grad_norm": 2.871100664138794, "learning_rate": 2.1336686322015854e-07, "loss": 1.5554, "step": 46286 }, { "epoch": 3.6275078369905955, "grad_norm": 3.47623348236084, "learning_rate": 2.1327775316556997e-07, "loss": 1.2221, "step": 46287 }, { "epoch": 3.627586206896552, "grad_norm": 6.494564056396484, "learning_rate": 2.1318866131721406e-07, "loss": 0.9653, "step": 46288 }, { "epoch": 3.6276645768025078, "grad_norm": 3.144219398498535, "learning_rate": 2.1309958767542937e-07, "loss": 1.2418, "step": 46289 }, { "epoch": 3.627742946708464, "grad_norm": 5.045856475830078, "learning_rate": 2.1301053224055513e-07, "loss": 1.2242, "step": 46290 }, { "epoch": 3.62782131661442, "grad_norm": 3.596679210662842, "learning_rate": 2.129214950129299e-07, "loss": 1.4541, "step": 46291 }, { "epoch": 3.6278996865203763, "grad_norm": 3.2166545391082764, "learning_rate": 2.128324759928929e-07, "loss": 2.208, "step": 46292 }, { "epoch": 3.627978056426332, "grad_norm": 2.3589491844177246, "learning_rate": 2.127434751807822e-07, "loss": 1.4259, "step": 46293 }, { "epoch": 3.6280564263322885, "grad_norm": 4.3839335441589355, "learning_rate": 2.1265449257693638e-07, "loss": 1.0097, "step": 46294 }, { "epoch": 3.6281347962382444, "grad_norm": 4.0572404861450195, "learning_rate": 2.1256552818169408e-07, "loss": 1.5618, "step": 46295 }, { "epoch": 3.6282131661442008, "grad_norm": 2.9132916927337646, "learning_rate": 2.1247658199539333e-07, "loss": 0.9654, "step": 46296 }, { "epoch": 3.6282915360501566, "grad_norm": 5.931872367858887, "learning_rate": 2.123876540183728e-07, "loss": 1.4632, "step": 46297 }, { "epoch": 3.628369905956113, "grad_norm": 4.299595355987549, "learning_rate": 2.1229874425097108e-07, "loss": 1.3806, "step": 46298 }, { "epoch": 3.628448275862069, "grad_norm": 6.418562889099121, "learning_rate": 2.1220985269352568e-07, "loss": 1.4202, "step": 46299 }, { "epoch": 3.628526645768025, "grad_norm": 3.0521609783172607, "learning_rate": 2.1212097934637356e-07, "loss": 1.2621, "step": 46300 }, { "epoch": 3.628605015673981, "grad_norm": 2.5619397163391113, "learning_rate": 2.1203212420985498e-07, "loss": 1.0998, "step": 46301 }, { "epoch": 3.6286833855799374, "grad_norm": 4.537271976470947, "learning_rate": 2.1194328728430747e-07, "loss": 1.7886, "step": 46302 }, { "epoch": 3.6287617554858933, "grad_norm": 4.2244672775268555, "learning_rate": 2.1185446857006743e-07, "loss": 1.1604, "step": 46303 }, { "epoch": 3.6288401253918496, "grad_norm": 6.592991828918457, "learning_rate": 2.1176566806747401e-07, "loss": 1.3527, "step": 46304 }, { "epoch": 3.6289184952978055, "grad_norm": 4.470416069030762, "learning_rate": 2.1167688577686419e-07, "loss": 0.9219, "step": 46305 }, { "epoch": 3.628996865203762, "grad_norm": 3.4276814460754395, "learning_rate": 2.1158812169857713e-07, "loss": 1.3689, "step": 46306 }, { "epoch": 3.6290752351097177, "grad_norm": 10.35318374633789, "learning_rate": 2.1149937583294866e-07, "loss": 1.2979, "step": 46307 }, { "epoch": 3.629153605015674, "grad_norm": 5.950626850128174, "learning_rate": 2.1141064818031686e-07, "loss": 1.6049, "step": 46308 }, { "epoch": 3.62923197492163, "grad_norm": 3.6109461784362793, "learning_rate": 2.1132193874101926e-07, "loss": 1.4323, "step": 46309 }, { "epoch": 3.6293103448275863, "grad_norm": 3.4269137382507324, "learning_rate": 2.1123324751539387e-07, "loss": 1.213, "step": 46310 }, { "epoch": 3.629388714733542, "grad_norm": 4.046189308166504, "learning_rate": 2.1114457450377713e-07, "loss": 1.5034, "step": 46311 }, { "epoch": 3.6294670846394985, "grad_norm": 6.3802571296691895, "learning_rate": 2.1105591970650706e-07, "loss": 1.5832, "step": 46312 }, { "epoch": 3.6295454545454544, "grad_norm": 2.9081318378448486, "learning_rate": 2.1096728312392122e-07, "loss": 0.9443, "step": 46313 }, { "epoch": 3.6296238244514107, "grad_norm": 4.994563102722168, "learning_rate": 2.1087866475635433e-07, "loss": 1.1192, "step": 46314 }, { "epoch": 3.6297021943573666, "grad_norm": 3.0572688579559326, "learning_rate": 2.1079006460414608e-07, "loss": 1.2291, "step": 46315 }, { "epoch": 3.629780564263323, "grad_norm": 3.5057127475738525, "learning_rate": 2.1070148266763235e-07, "loss": 1.3109, "step": 46316 }, { "epoch": 3.629858934169279, "grad_norm": 3.9190990924835205, "learning_rate": 2.1061291894715118e-07, "loss": 1.0347, "step": 46317 }, { "epoch": 3.629937304075235, "grad_norm": 3.681410312652588, "learning_rate": 2.105243734430379e-07, "loss": 1.4425, "step": 46318 }, { "epoch": 3.630015673981191, "grad_norm": 6.219930648803711, "learning_rate": 2.1043584615562885e-07, "loss": 1.3661, "step": 46319 }, { "epoch": 3.6300940438871474, "grad_norm": 5.61414098739624, "learning_rate": 2.1034733708526268e-07, "loss": 1.7257, "step": 46320 }, { "epoch": 3.6301724137931033, "grad_norm": 3.8344833850860596, "learning_rate": 2.1025884623227523e-07, "loss": 1.3073, "step": 46321 }, { "epoch": 3.6302507836990596, "grad_norm": 2.912839412689209, "learning_rate": 2.1017037359700286e-07, "loss": 1.2914, "step": 46322 }, { "epoch": 3.6303291536050155, "grad_norm": 7.113033771514893, "learning_rate": 2.1008191917978204e-07, "loss": 1.4652, "step": 46323 }, { "epoch": 3.630407523510972, "grad_norm": 4.512977123260498, "learning_rate": 2.0999348298094856e-07, "loss": 1.4361, "step": 46324 }, { "epoch": 3.6304858934169277, "grad_norm": 2.456125259399414, "learning_rate": 2.0990506500084107e-07, "loss": 1.229, "step": 46325 }, { "epoch": 3.630564263322884, "grad_norm": 3.245591878890991, "learning_rate": 2.098166652397937e-07, "loss": 1.39, "step": 46326 }, { "epoch": 3.63064263322884, "grad_norm": 4.099145412445068, "learning_rate": 2.0972828369814346e-07, "loss": 1.7813, "step": 46327 }, { "epoch": 3.6307210031347963, "grad_norm": 3.7917048931121826, "learning_rate": 2.0963992037622617e-07, "loss": 1.4313, "step": 46328 }, { "epoch": 3.6307993730407526, "grad_norm": 2.3756589889526367, "learning_rate": 2.0955157527437818e-07, "loss": 1.4276, "step": 46329 }, { "epoch": 3.6308777429467085, "grad_norm": 3.7633464336395264, "learning_rate": 2.094632483929354e-07, "loss": 1.7607, "step": 46330 }, { "epoch": 3.6309561128526644, "grad_norm": 5.507438659667969, "learning_rate": 2.0937493973223423e-07, "loss": 1.2899, "step": 46331 }, { "epoch": 3.6310344827586207, "grad_norm": 15.888110160827637, "learning_rate": 2.0928664929261045e-07, "loss": 1.9977, "step": 46332 }, { "epoch": 3.631112852664577, "grad_norm": 2.351950168609619, "learning_rate": 2.0919837707439827e-07, "loss": 1.2573, "step": 46333 }, { "epoch": 3.631191222570533, "grad_norm": 2.7251296043395996, "learning_rate": 2.091101230779352e-07, "loss": 1.2891, "step": 46334 }, { "epoch": 3.631269592476489, "grad_norm": 2.0287270545959473, "learning_rate": 2.090218873035571e-07, "loss": 1.2301, "step": 46335 }, { "epoch": 3.631347962382445, "grad_norm": 5.731096267700195, "learning_rate": 2.0893366975159923e-07, "loss": 1.4349, "step": 46336 }, { "epoch": 3.6314263322884015, "grad_norm": 3.974449872970581, "learning_rate": 2.0884547042239577e-07, "loss": 1.0002, "step": 46337 }, { "epoch": 3.6315047021943574, "grad_norm": 5.523385524749756, "learning_rate": 2.0875728931628315e-07, "loss": 1.4397, "step": 46338 }, { "epoch": 3.6315830721003133, "grad_norm": 3.1165568828582764, "learning_rate": 2.086691264335977e-07, "loss": 1.3291, "step": 46339 }, { "epoch": 3.6316614420062696, "grad_norm": 3.398772716522217, "learning_rate": 2.0858098177467422e-07, "loss": 1.456, "step": 46340 }, { "epoch": 3.631739811912226, "grad_norm": 3.133762836456299, "learning_rate": 2.0849285533984687e-07, "loss": 0.9721, "step": 46341 }, { "epoch": 3.631818181818182, "grad_norm": 5.574019908905029, "learning_rate": 2.0840474712945202e-07, "loss": 1.647, "step": 46342 }, { "epoch": 3.6318965517241377, "grad_norm": 2.9047138690948486, "learning_rate": 2.0831665714382386e-07, "loss": 0.957, "step": 46343 }, { "epoch": 3.631974921630094, "grad_norm": 3.2348549365997314, "learning_rate": 2.0822858538329827e-07, "loss": 1.0869, "step": 46344 }, { "epoch": 3.6320532915360504, "grad_norm": 6.099460124969482, "learning_rate": 2.0814053184820993e-07, "loss": 1.4655, "step": 46345 }, { "epoch": 3.6321316614420063, "grad_norm": 3.1554691791534424, "learning_rate": 2.080524965388936e-07, "loss": 1.1994, "step": 46346 }, { "epoch": 3.632210031347962, "grad_norm": 3.6236493587493896, "learning_rate": 2.0796447945568455e-07, "loss": 2.1586, "step": 46347 }, { "epoch": 3.6322884012539185, "grad_norm": 2.9587080478668213, "learning_rate": 2.078764805989175e-07, "loss": 1.0885, "step": 46348 }, { "epoch": 3.632366771159875, "grad_norm": 4.075364589691162, "learning_rate": 2.0778849996892614e-07, "loss": 1.5676, "step": 46349 }, { "epoch": 3.6324451410658307, "grad_norm": 2.7711167335510254, "learning_rate": 2.077005375660468e-07, "loss": 1.0703, "step": 46350 }, { "epoch": 3.6325235109717866, "grad_norm": 2.901876211166382, "learning_rate": 2.0761259339061258e-07, "loss": 1.262, "step": 46351 }, { "epoch": 3.632601880877743, "grad_norm": 2.4590518474578857, "learning_rate": 2.075246674429582e-07, "loss": 1.1871, "step": 46352 }, { "epoch": 3.6326802507836993, "grad_norm": 3.175431728363037, "learning_rate": 2.0743675972341892e-07, "loss": 1.1318, "step": 46353 }, { "epoch": 3.632758620689655, "grad_norm": 2.62226939201355, "learning_rate": 2.0734887023232953e-07, "loss": 1.2867, "step": 46354 }, { "epoch": 3.632836990595611, "grad_norm": 12.854676246643066, "learning_rate": 2.0726099897002193e-07, "loss": 0.649, "step": 46355 }, { "epoch": 3.6329153605015674, "grad_norm": 8.685775756835938, "learning_rate": 2.0717314593683146e-07, "loss": 1.1372, "step": 46356 }, { "epoch": 3.6329937304075237, "grad_norm": 3.550276041030884, "learning_rate": 2.0708531113309338e-07, "loss": 1.3451, "step": 46357 }, { "epoch": 3.6330721003134796, "grad_norm": 5.8109235763549805, "learning_rate": 2.0699749455914186e-07, "loss": 1.6653, "step": 46358 }, { "epoch": 3.6331504702194355, "grad_norm": 2.5603482723236084, "learning_rate": 2.0690969621530887e-07, "loss": 1.4777, "step": 46359 }, { "epoch": 3.633228840125392, "grad_norm": 3.0147488117218018, "learning_rate": 2.068219161019297e-07, "loss": 1.4443, "step": 46360 }, { "epoch": 3.633307210031348, "grad_norm": 2.888045310974121, "learning_rate": 2.067341542193385e-07, "loss": 1.587, "step": 46361 }, { "epoch": 3.633385579937304, "grad_norm": 5.025472640991211, "learning_rate": 2.0664641056786782e-07, "loss": 1.5147, "step": 46362 }, { "epoch": 3.63346394984326, "grad_norm": 11.735944747924805, "learning_rate": 2.0655868514785237e-07, "loss": 1.9065, "step": 46363 }, { "epoch": 3.6335423197492163, "grad_norm": 2.586073637008667, "learning_rate": 2.0647097795962577e-07, "loss": 1.1393, "step": 46364 }, { "epoch": 3.6336206896551726, "grad_norm": 2.1844708919525146, "learning_rate": 2.0638328900352166e-07, "loss": 1.2643, "step": 46365 }, { "epoch": 3.6336990595611285, "grad_norm": 3.1854236125946045, "learning_rate": 2.062956182798731e-07, "loss": 1.115, "step": 46366 }, { "epoch": 3.6337774294670844, "grad_norm": 7.451691627502441, "learning_rate": 2.062079657890137e-07, "loss": 1.8294, "step": 46367 }, { "epoch": 3.6338557993730407, "grad_norm": 9.730194091796875, "learning_rate": 2.0612033153127765e-07, "loss": 0.5027, "step": 46368 }, { "epoch": 3.633934169278997, "grad_norm": 3.9667844772338867, "learning_rate": 2.0603271550699745e-07, "loss": 1.2231, "step": 46369 }, { "epoch": 3.634012539184953, "grad_norm": 3.925389051437378, "learning_rate": 2.059451177165056e-07, "loss": 1.4628, "step": 46370 }, { "epoch": 3.634090909090909, "grad_norm": 5.495543956756592, "learning_rate": 2.0585753816013632e-07, "loss": 0.8098, "step": 46371 }, { "epoch": 3.634169278996865, "grad_norm": 2.4736475944519043, "learning_rate": 2.0576997683822263e-07, "loss": 1.5665, "step": 46372 }, { "epoch": 3.6342476489028215, "grad_norm": 4.49712610244751, "learning_rate": 2.056824337510982e-07, "loss": 1.167, "step": 46373 }, { "epoch": 3.6343260188087774, "grad_norm": 2.6195790767669678, "learning_rate": 2.0559490889909495e-07, "loss": 1.0309, "step": 46374 }, { "epoch": 3.6344043887147337, "grad_norm": 14.764748573303223, "learning_rate": 2.055074022825454e-07, "loss": 1.4609, "step": 46375 }, { "epoch": 3.6344827586206896, "grad_norm": 3.4743406772613525, "learning_rate": 2.0541991390178317e-07, "loss": 1.0551, "step": 46376 }, { "epoch": 3.634561128526646, "grad_norm": 4.3297038078308105, "learning_rate": 2.0533244375714244e-07, "loss": 1.3038, "step": 46377 }, { "epoch": 3.634639498432602, "grad_norm": 2.428755044937134, "learning_rate": 2.0524499184895296e-07, "loss": 1.1415, "step": 46378 }, { "epoch": 3.634717868338558, "grad_norm": 4.307071685791016, "learning_rate": 2.0515755817754946e-07, "loss": 1.1731, "step": 46379 }, { "epoch": 3.634796238244514, "grad_norm": 3.140472173690796, "learning_rate": 2.0507014274326276e-07, "loss": 1.497, "step": 46380 }, { "epoch": 3.6348746081504704, "grad_norm": 7.213133335113525, "learning_rate": 2.049827455464276e-07, "loss": 1.5189, "step": 46381 }, { "epoch": 3.6349529780564263, "grad_norm": 8.415064811706543, "learning_rate": 2.0489536658737485e-07, "loss": 2.5847, "step": 46382 }, { "epoch": 3.6350313479623826, "grad_norm": 2.3581137657165527, "learning_rate": 2.04808005866437e-07, "loss": 1.1492, "step": 46383 }, { "epoch": 3.6351097178683385, "grad_norm": 4.704179286956787, "learning_rate": 2.0472066338394658e-07, "loss": 1.3748, "step": 46384 }, { "epoch": 3.635188087774295, "grad_norm": 5.38224458694458, "learning_rate": 2.0463333914023553e-07, "loss": 1.158, "step": 46385 }, { "epoch": 3.6352664576802507, "grad_norm": 5.497053146362305, "learning_rate": 2.0454603313563636e-07, "loss": 1.6586, "step": 46386 }, { "epoch": 3.635344827586207, "grad_norm": 5.373569965362549, "learning_rate": 2.0445874537048106e-07, "loss": 1.3218, "step": 46387 }, { "epoch": 3.635423197492163, "grad_norm": 9.42001724243164, "learning_rate": 2.0437147584510264e-07, "loss": 1.2764, "step": 46388 }, { "epoch": 3.6355015673981192, "grad_norm": 4.79054594039917, "learning_rate": 2.0428422455982977e-07, "loss": 1.151, "step": 46389 }, { "epoch": 3.635579937304075, "grad_norm": 3.1188037395477295, "learning_rate": 2.0419699151499773e-07, "loss": 0.9134, "step": 46390 }, { "epoch": 3.6356583072100315, "grad_norm": 4.198507785797119, "learning_rate": 2.041097767109368e-07, "loss": 1.4506, "step": 46391 }, { "epoch": 3.6357366771159874, "grad_norm": 3.80586576461792, "learning_rate": 2.040225801479795e-07, "loss": 0.9365, "step": 46392 }, { "epoch": 3.6358150470219437, "grad_norm": 7.990296363830566, "learning_rate": 2.0393540182645667e-07, "loss": 0.9901, "step": 46393 }, { "epoch": 3.6358934169278996, "grad_norm": 3.365262746810913, "learning_rate": 2.0384824174669916e-07, "loss": 1.0153, "step": 46394 }, { "epoch": 3.635971786833856, "grad_norm": 3.581780433654785, "learning_rate": 2.0376109990904059e-07, "loss": 1.4988, "step": 46395 }, { "epoch": 3.636050156739812, "grad_norm": 7.008018493652344, "learning_rate": 2.0367397631381125e-07, "loss": 1.6454, "step": 46396 }, { "epoch": 3.636128526645768, "grad_norm": 1.816285490989685, "learning_rate": 2.0358687096134254e-07, "loss": 1.0799, "step": 46397 }, { "epoch": 3.636206896551724, "grad_norm": 2.9851460456848145, "learning_rate": 2.0349978385196535e-07, "loss": 1.4523, "step": 46398 }, { "epoch": 3.6362852664576804, "grad_norm": 3.693344831466675, "learning_rate": 2.0341271498601156e-07, "loss": 1.5149, "step": 46399 }, { "epoch": 3.6363636363636362, "grad_norm": 3.164825201034546, "learning_rate": 2.0332566436381206e-07, "loss": 1.5588, "step": 46400 }, { "epoch": 3.6364420062695926, "grad_norm": 2.7171428203582764, "learning_rate": 2.0323863198569827e-07, "loss": 1.484, "step": 46401 }, { "epoch": 3.6365203761755485, "grad_norm": 2.5883142948150635, "learning_rate": 2.0315161785200043e-07, "loss": 1.2087, "step": 46402 }, { "epoch": 3.636598746081505, "grad_norm": 6.408504486083984, "learning_rate": 2.0306462196305055e-07, "loss": 1.7764, "step": 46403 }, { "epoch": 3.6366771159874607, "grad_norm": 4.37035608291626, "learning_rate": 2.0297764431917833e-07, "loss": 0.9462, "step": 46404 }, { "epoch": 3.636755485893417, "grad_norm": 2.464921474456787, "learning_rate": 2.0289068492071574e-07, "loss": 1.4996, "step": 46405 }, { "epoch": 3.636833855799373, "grad_norm": 3.6390798091888428, "learning_rate": 2.0280374376799306e-07, "loss": 1.5056, "step": 46406 }, { "epoch": 3.6369122257053292, "grad_norm": 4.60564661026001, "learning_rate": 2.0271682086134114e-07, "loss": 1.2936, "step": 46407 }, { "epoch": 3.636990595611285, "grad_norm": 4.518680572509766, "learning_rate": 2.0262991620108975e-07, "loss": 1.456, "step": 46408 }, { "epoch": 3.6370689655172415, "grad_norm": 2.884381055831909, "learning_rate": 2.0254302978757024e-07, "loss": 1.3774, "step": 46409 }, { "epoch": 3.6371473354231973, "grad_norm": 3.7097909450531006, "learning_rate": 2.0245616162111348e-07, "loss": 1.1506, "step": 46410 }, { "epoch": 3.6372257053291537, "grad_norm": 7.660733699798584, "learning_rate": 2.023693117020492e-07, "loss": 1.6145, "step": 46411 }, { "epoch": 3.6373040752351096, "grad_norm": 5.7848381996154785, "learning_rate": 2.0228248003070773e-07, "loss": 1.3155, "step": 46412 }, { "epoch": 3.637382445141066, "grad_norm": 2.9362049102783203, "learning_rate": 2.0219566660741818e-07, "loss": 1.0989, "step": 46413 }, { "epoch": 3.637460815047022, "grad_norm": 5.486907482147217, "learning_rate": 2.0210887143251367e-07, "loss": 1.608, "step": 46414 }, { "epoch": 3.637539184952978, "grad_norm": 5.035316467285156, "learning_rate": 2.0202209450632226e-07, "loss": 1.0829, "step": 46415 }, { "epoch": 3.637617554858934, "grad_norm": 3.4325592517852783, "learning_rate": 2.0193533582917425e-07, "loss": 1.6968, "step": 46416 }, { "epoch": 3.6376959247648903, "grad_norm": 5.008236885070801, "learning_rate": 2.018485954013999e-07, "loss": 1.3285, "step": 46417 }, { "epoch": 3.6377742946708462, "grad_norm": 2.3695318698883057, "learning_rate": 2.017618732233284e-07, "loss": 1.3768, "step": 46418 }, { "epoch": 3.6378526645768026, "grad_norm": 3.5836334228515625, "learning_rate": 2.0167516929529063e-07, "loss": 1.3367, "step": 46419 }, { "epoch": 3.637931034482759, "grad_norm": 6.166439533233643, "learning_rate": 2.015884836176163e-07, "loss": 1.4985, "step": 46420 }, { "epoch": 3.638009404388715, "grad_norm": 2.496377468109131, "learning_rate": 2.0150181619063402e-07, "loss": 1.2447, "step": 46421 }, { "epoch": 3.6380877742946707, "grad_norm": 7.790894031524658, "learning_rate": 2.0141516701467466e-07, "loss": 1.0369, "step": 46422 }, { "epoch": 3.638166144200627, "grad_norm": 5.021114826202393, "learning_rate": 2.0132853609006742e-07, "loss": 1.2879, "step": 46423 }, { "epoch": 3.6382445141065833, "grad_norm": 3.6090030670166016, "learning_rate": 2.0124192341714145e-07, "loss": 1.4277, "step": 46424 }, { "epoch": 3.6383228840125392, "grad_norm": 7.698555946350098, "learning_rate": 2.0115532899622703e-07, "loss": 1.1539, "step": 46425 }, { "epoch": 3.638401253918495, "grad_norm": 2.314221143722534, "learning_rate": 2.0106875282765226e-07, "loss": 0.8165, "step": 46426 }, { "epoch": 3.6384796238244514, "grad_norm": 2.5620882511138916, "learning_rate": 2.009821949117463e-07, "loss": 1.3166, "step": 46427 }, { "epoch": 3.638557993730408, "grad_norm": 2.553187131881714, "learning_rate": 2.0089565524884004e-07, "loss": 1.4617, "step": 46428 }, { "epoch": 3.6386363636363637, "grad_norm": 4.150965690612793, "learning_rate": 2.0080913383926204e-07, "loss": 1.286, "step": 46429 }, { "epoch": 3.6387147335423196, "grad_norm": 3.8066208362579346, "learning_rate": 2.0072263068334097e-07, "loss": 1.3688, "step": 46430 }, { "epoch": 3.638793103448276, "grad_norm": 3.0562472343444824, "learning_rate": 2.0063614578140544e-07, "loss": 1.4905, "step": 46431 }, { "epoch": 3.6388714733542322, "grad_norm": 3.3088321685791016, "learning_rate": 2.0054967913378465e-07, "loss": 1.2545, "step": 46432 }, { "epoch": 3.638949843260188, "grad_norm": 5.010359287261963, "learning_rate": 2.0046323074080887e-07, "loss": 1.4209, "step": 46433 }, { "epoch": 3.639028213166144, "grad_norm": 4.04481840133667, "learning_rate": 2.0037680060280506e-07, "loss": 1.0689, "step": 46434 }, { "epoch": 3.6391065830721003, "grad_norm": 3.899568796157837, "learning_rate": 2.0029038872010242e-07, "loss": 1.1662, "step": 46435 }, { "epoch": 3.6391849529780567, "grad_norm": 5.6042160987854, "learning_rate": 2.0020399509303012e-07, "loss": 1.1896, "step": 46436 }, { "epoch": 3.6392633228840126, "grad_norm": 6.177253723144531, "learning_rate": 2.0011761972191624e-07, "loss": 1.5241, "step": 46437 }, { "epoch": 3.6393416927899684, "grad_norm": 8.174256324768066, "learning_rate": 2.0003126260708995e-07, "loss": 1.8643, "step": 46438 }, { "epoch": 3.6394200626959248, "grad_norm": 3.307961940765381, "learning_rate": 1.999449237488793e-07, "loss": 1.3859, "step": 46439 }, { "epoch": 3.639498432601881, "grad_norm": 2.6955885887145996, "learning_rate": 1.9985860314761297e-07, "loss": 1.5062, "step": 46440 }, { "epoch": 3.639576802507837, "grad_norm": 4.160593032836914, "learning_rate": 1.9977230080361788e-07, "loss": 1.2808, "step": 46441 }, { "epoch": 3.639655172413793, "grad_norm": 2.728257179260254, "learning_rate": 1.996860167172243e-07, "loss": 1.2796, "step": 46442 }, { "epoch": 3.639733542319749, "grad_norm": 2.2521207332611084, "learning_rate": 1.9959975088875928e-07, "loss": 1.1576, "step": 46443 }, { "epoch": 3.6398119122257055, "grad_norm": 2.928633689880371, "learning_rate": 1.995135033185519e-07, "loss": 1.2573, "step": 46444 }, { "epoch": 3.6398902821316614, "grad_norm": 2.389497995376587, "learning_rate": 1.9942727400692864e-07, "loss": 1.3832, "step": 46445 }, { "epoch": 3.6399686520376173, "grad_norm": 3.313598394393921, "learning_rate": 1.9934106295421806e-07, "loss": 1.094, "step": 46446 }, { "epoch": 3.6400470219435737, "grad_norm": 5.117307662963867, "learning_rate": 1.992548701607483e-07, "loss": 1.2799, "step": 46447 }, { "epoch": 3.64012539184953, "grad_norm": 2.594580888748169, "learning_rate": 1.9916869562684849e-07, "loss": 1.5564, "step": 46448 }, { "epoch": 3.640203761755486, "grad_norm": 3.7016119956970215, "learning_rate": 1.9908253935284394e-07, "loss": 1.6028, "step": 46449 }, { "epoch": 3.6402821316614418, "grad_norm": 2.910520553588867, "learning_rate": 1.9899640133906384e-07, "loss": 1.4501, "step": 46450 }, { "epoch": 3.640360501567398, "grad_norm": 3.174217939376831, "learning_rate": 1.9891028158583513e-07, "loss": 1.215, "step": 46451 }, { "epoch": 3.6404388714733544, "grad_norm": 4.523339748382568, "learning_rate": 1.9882418009348591e-07, "loss": 1.564, "step": 46452 }, { "epoch": 3.6405172413793103, "grad_norm": 2.5309085845947266, "learning_rate": 1.987380968623437e-07, "loss": 1.261, "step": 46453 }, { "epoch": 3.640595611285266, "grad_norm": 2.3878557682037354, "learning_rate": 1.986520318927354e-07, "loss": 1.4147, "step": 46454 }, { "epoch": 3.6406739811912225, "grad_norm": 3.1452157497406006, "learning_rate": 1.9856598518498914e-07, "loss": 1.1787, "step": 46455 }, { "epoch": 3.640752351097179, "grad_norm": 3.8137662410736084, "learning_rate": 1.9847995673943132e-07, "loss": 1.3733, "step": 46456 }, { "epoch": 3.6408307210031348, "grad_norm": 7.185812473297119, "learning_rate": 1.9839394655638943e-07, "loss": 1.7929, "step": 46457 }, { "epoch": 3.6409090909090907, "grad_norm": 3.1280975341796875, "learning_rate": 1.9830795463619046e-07, "loss": 1.2441, "step": 46458 }, { "epoch": 3.640987460815047, "grad_norm": 2.153841495513916, "learning_rate": 1.9822198097916246e-07, "loss": 1.1629, "step": 46459 }, { "epoch": 3.6410658307210033, "grad_norm": 3.667717933654785, "learning_rate": 1.9813602558563128e-07, "loss": 1.4081, "step": 46460 }, { "epoch": 3.641144200626959, "grad_norm": 2.794839382171631, "learning_rate": 1.9805008845592388e-07, "loss": 1.187, "step": 46461 }, { "epoch": 3.641222570532915, "grad_norm": 4.217648506164551, "learning_rate": 1.9796416959036835e-07, "loss": 1.281, "step": 46462 }, { "epoch": 3.6413009404388714, "grad_norm": 4.248904705047607, "learning_rate": 1.9787826898929053e-07, "loss": 1.4973, "step": 46463 }, { "epoch": 3.6413793103448278, "grad_norm": 6.143896579742432, "learning_rate": 1.9779238665301626e-07, "loss": 1.2679, "step": 46464 }, { "epoch": 3.6414576802507836, "grad_norm": 2.744633436203003, "learning_rate": 1.9770652258187418e-07, "loss": 1.4468, "step": 46465 }, { "epoch": 3.6415360501567395, "grad_norm": 5.752012729644775, "learning_rate": 1.976206767761901e-07, "loss": 1.5018, "step": 46466 }, { "epoch": 3.641614420062696, "grad_norm": 3.355825901031494, "learning_rate": 1.975348492362894e-07, "loss": 1.1691, "step": 46467 }, { "epoch": 3.641692789968652, "grad_norm": 2.71667218208313, "learning_rate": 1.974490399625001e-07, "loss": 1.3752, "step": 46468 }, { "epoch": 3.641771159874608, "grad_norm": 5.248157024383545, "learning_rate": 1.973632489551469e-07, "loss": 1.7094, "step": 46469 }, { "epoch": 3.6418495297805644, "grad_norm": 2.8061273097991943, "learning_rate": 1.972774762145585e-07, "loss": 1.1482, "step": 46470 }, { "epoch": 3.6419278996865203, "grad_norm": 5.159267902374268, "learning_rate": 1.9719172174105904e-07, "loss": 1.6475, "step": 46471 }, { "epoch": 3.6420062695924766, "grad_norm": 4.488068103790283, "learning_rate": 1.9710598553497496e-07, "loss": 1.1864, "step": 46472 }, { "epoch": 3.6420846394984325, "grad_norm": 2.3554162979125977, "learning_rate": 1.9702026759663316e-07, "loss": 1.4993, "step": 46473 }, { "epoch": 3.642163009404389, "grad_norm": 3.188202381134033, "learning_rate": 1.9693456792635957e-07, "loss": 1.2195, "step": 46474 }, { "epoch": 3.6422413793103448, "grad_norm": 4.854696750640869, "learning_rate": 1.9684888652447941e-07, "loss": 0.9452, "step": 46475 }, { "epoch": 3.642319749216301, "grad_norm": 2.53678297996521, "learning_rate": 1.967632233913197e-07, "loss": 1.1505, "step": 46476 }, { "epoch": 3.642398119122257, "grad_norm": 3.0716328620910645, "learning_rate": 1.9667757852720515e-07, "loss": 1.3063, "step": 46477 }, { "epoch": 3.6424764890282133, "grad_norm": 4.607032299041748, "learning_rate": 1.9659195193246162e-07, "loss": 1.3695, "step": 46478 }, { "epoch": 3.642554858934169, "grad_norm": 3.750955820083618, "learning_rate": 1.965063436074155e-07, "loss": 1.5845, "step": 46479 }, { "epoch": 3.6426332288401255, "grad_norm": 2.6656177043914795, "learning_rate": 1.964207535523921e-07, "loss": 1.106, "step": 46480 }, { "epoch": 3.6427115987460814, "grad_norm": 2.684476613998413, "learning_rate": 1.9633518176771727e-07, "loss": 1.0063, "step": 46481 }, { "epoch": 3.6427899686520377, "grad_norm": 2.6115148067474365, "learning_rate": 1.9624962825371575e-07, "loss": 0.9997, "step": 46482 }, { "epoch": 3.6428683385579936, "grad_norm": 4.612882614135742, "learning_rate": 1.9616409301071227e-07, "loss": 1.3492, "step": 46483 }, { "epoch": 3.64294670846395, "grad_norm": 2.1524951457977295, "learning_rate": 1.9607857603903436e-07, "loss": 1.3538, "step": 46484 }, { "epoch": 3.643025078369906, "grad_norm": 3.8938443660736084, "learning_rate": 1.9599307733900618e-07, "loss": 1.5832, "step": 46485 }, { "epoch": 3.643103448275862, "grad_norm": 4.635468006134033, "learning_rate": 1.9590759691095252e-07, "loss": 1.4586, "step": 46486 }, { "epoch": 3.643181818181818, "grad_norm": 3.4199037551879883, "learning_rate": 1.9582213475519862e-07, "loss": 1.4893, "step": 46487 }, { "epoch": 3.6432601880877744, "grad_norm": 2.4222443103790283, "learning_rate": 1.9573669087206924e-07, "loss": 1.6466, "step": 46488 }, { "epoch": 3.6433385579937303, "grad_norm": 2.788767099380493, "learning_rate": 1.9565126526189083e-07, "loss": 0.9911, "step": 46489 }, { "epoch": 3.6434169278996866, "grad_norm": 3.6936373710632324, "learning_rate": 1.9556585792498695e-07, "loss": 1.2054, "step": 46490 }, { "epoch": 3.6434952978056425, "grad_norm": 2.695742130279541, "learning_rate": 1.9548046886168292e-07, "loss": 1.1435, "step": 46491 }, { "epoch": 3.643573667711599, "grad_norm": 3.3326637744903564, "learning_rate": 1.953950980723035e-07, "loss": 1.4742, "step": 46492 }, { "epoch": 3.6436520376175547, "grad_norm": 2.874854803085327, "learning_rate": 1.9530974555717342e-07, "loss": 1.5567, "step": 46493 }, { "epoch": 3.643730407523511, "grad_norm": 4.028011322021484, "learning_rate": 1.9522441131661685e-07, "loss": 1.8062, "step": 46494 }, { "epoch": 3.643808777429467, "grad_norm": 2.367534637451172, "learning_rate": 1.9513909535095853e-07, "loss": 1.2711, "step": 46495 }, { "epoch": 3.6438871473354233, "grad_norm": 3.753659725189209, "learning_rate": 1.9505379766052434e-07, "loss": 1.1814, "step": 46496 }, { "epoch": 3.643965517241379, "grad_norm": 5.3459296226501465, "learning_rate": 1.9496851824563622e-07, "loss": 1.3489, "step": 46497 }, { "epoch": 3.6440438871473355, "grad_norm": 2.571904420852661, "learning_rate": 1.9488325710662003e-07, "loss": 1.0525, "step": 46498 }, { "epoch": 3.6441222570532914, "grad_norm": 2.127913236618042, "learning_rate": 1.9479801424379995e-07, "loss": 1.3021, "step": 46499 }, { "epoch": 3.6442006269592477, "grad_norm": 4.13390588760376, "learning_rate": 1.9471278965750129e-07, "loss": 1.9336, "step": 46500 }, { "epoch": 3.6442789968652036, "grad_norm": 2.585782051086426, "learning_rate": 1.9462758334804598e-07, "loss": 1.1651, "step": 46501 }, { "epoch": 3.64435736677116, "grad_norm": 2.774015426635742, "learning_rate": 1.945423953157588e-07, "loss": 1.4945, "step": 46502 }, { "epoch": 3.644435736677116, "grad_norm": 2.8967490196228027, "learning_rate": 1.9445722556096446e-07, "loss": 0.9065, "step": 46503 }, { "epoch": 3.644514106583072, "grad_norm": 2.646810293197632, "learning_rate": 1.9437207408398717e-07, "loss": 1.734, "step": 46504 }, { "epoch": 3.644592476489028, "grad_norm": 3.9067623615264893, "learning_rate": 1.9428694088514942e-07, "loss": 1.4197, "step": 46505 }, { "epoch": 3.6446708463949844, "grad_norm": 2.991546392440796, "learning_rate": 1.94201825964776e-07, "loss": 0.9701, "step": 46506 }, { "epoch": 3.6447492163009403, "grad_norm": 7.546860218048096, "learning_rate": 1.941167293231905e-07, "loss": 1.0156, "step": 46507 }, { "epoch": 3.6448275862068966, "grad_norm": 3.316885232925415, "learning_rate": 1.9403165096071652e-07, "loss": 0.6554, "step": 46508 }, { "epoch": 3.6449059561128525, "grad_norm": 7.935393333435059, "learning_rate": 1.939465908776772e-07, "loss": 0.9615, "step": 46509 }, { "epoch": 3.644984326018809, "grad_norm": 3.0055458545684814, "learning_rate": 1.938615490743967e-07, "loss": 1.3047, "step": 46510 }, { "epoch": 3.6450626959247647, "grad_norm": 2.771270751953125, "learning_rate": 1.9377652555119864e-07, "loss": 1.5938, "step": 46511 }, { "epoch": 3.645141065830721, "grad_norm": 3.5207910537719727, "learning_rate": 1.9369152030840553e-07, "loss": 1.1748, "step": 46512 }, { "epoch": 3.645219435736677, "grad_norm": 3.0566225051879883, "learning_rate": 1.936065333463416e-07, "loss": 1.4604, "step": 46513 }, { "epoch": 3.6452978056426333, "grad_norm": 2.6765365600585938, "learning_rate": 1.9352156466532989e-07, "loss": 1.2179, "step": 46514 }, { "epoch": 3.6453761755485896, "grad_norm": 3.3416619300842285, "learning_rate": 1.934366142656935e-07, "loss": 1.1409, "step": 46515 }, { "epoch": 3.6454545454545455, "grad_norm": 3.7838222980499268, "learning_rate": 1.9335168214775434e-07, "loss": 1.5345, "step": 46516 }, { "epoch": 3.6455329153605014, "grad_norm": 3.46855092048645, "learning_rate": 1.9326676831183723e-07, "loss": 1.521, "step": 46517 }, { "epoch": 3.6456112852664577, "grad_norm": 3.322185516357422, "learning_rate": 1.931818727582646e-07, "loss": 1.1227, "step": 46518 }, { "epoch": 3.645689655172414, "grad_norm": 2.885267734527588, "learning_rate": 1.930969954873596e-07, "loss": 0.8521, "step": 46519 }, { "epoch": 3.64576802507837, "grad_norm": 3.697753667831421, "learning_rate": 1.930121364994436e-07, "loss": 1.438, "step": 46520 }, { "epoch": 3.645846394984326, "grad_norm": 2.410142183303833, "learning_rate": 1.9292729579484026e-07, "loss": 1.3141, "step": 46521 }, { "epoch": 3.645924764890282, "grad_norm": 7.661183834075928, "learning_rate": 1.9284247337387375e-07, "loss": 1.2684, "step": 46522 }, { "epoch": 3.6460031347962385, "grad_norm": 5.506336212158203, "learning_rate": 1.9275766923686435e-07, "loss": 1.6037, "step": 46523 }, { "epoch": 3.6460815047021944, "grad_norm": 4.1830010414123535, "learning_rate": 1.9267288338413626e-07, "loss": 1.5023, "step": 46524 }, { "epoch": 3.6461598746081503, "grad_norm": 7.319272041320801, "learning_rate": 1.925881158160109e-07, "loss": 1.5303, "step": 46525 }, { "epoch": 3.6462382445141066, "grad_norm": 3.534914493560791, "learning_rate": 1.9250336653281077e-07, "loss": 1.7229, "step": 46526 }, { "epoch": 3.646316614420063, "grad_norm": 3.5251896381378174, "learning_rate": 1.9241863553485895e-07, "loss": 1.4519, "step": 46527 }, { "epoch": 3.646394984326019, "grad_norm": 4.230941295623779, "learning_rate": 1.9233392282247742e-07, "loss": 0.9435, "step": 46528 }, { "epoch": 3.6464733542319747, "grad_norm": 3.3477394580841064, "learning_rate": 1.9224922839598814e-07, "loss": 1.1563, "step": 46529 }, { "epoch": 3.646551724137931, "grad_norm": 3.3925135135650635, "learning_rate": 1.9216455225571363e-07, "loss": 1.6957, "step": 46530 }, { "epoch": 3.6466300940438874, "grad_norm": 3.0990421772003174, "learning_rate": 1.9207989440197528e-07, "loss": 1.3694, "step": 46531 }, { "epoch": 3.6467084639498433, "grad_norm": 4.362722396850586, "learning_rate": 1.9199525483509562e-07, "loss": 1.0158, "step": 46532 }, { "epoch": 3.646786833855799, "grad_norm": 3.3479983806610107, "learning_rate": 1.9191063355539664e-07, "loss": 0.7786, "step": 46533 }, { "epoch": 3.6468652037617555, "grad_norm": 2.9020793437957764, "learning_rate": 1.9182603056320027e-07, "loss": 1.2324, "step": 46534 }, { "epoch": 3.646943573667712, "grad_norm": 9.234967231750488, "learning_rate": 1.9174144585882737e-07, "loss": 1.5059, "step": 46535 }, { "epoch": 3.6470219435736677, "grad_norm": 9.69854736328125, "learning_rate": 1.9165687944260046e-07, "loss": 1.3218, "step": 46536 }, { "epoch": 3.6471003134796236, "grad_norm": 3.2856674194335938, "learning_rate": 1.9157233131484154e-07, "loss": 1.2358, "step": 46537 }, { "epoch": 3.64717868338558, "grad_norm": 3.4728713035583496, "learning_rate": 1.9148780147587141e-07, "loss": 1.2766, "step": 46538 }, { "epoch": 3.6472570532915363, "grad_norm": 5.506679058074951, "learning_rate": 1.914032899260121e-07, "loss": 1.4746, "step": 46539 }, { "epoch": 3.647335423197492, "grad_norm": 3.5620837211608887, "learning_rate": 1.9131879666558385e-07, "loss": 1.4788, "step": 46540 }, { "epoch": 3.647413793103448, "grad_norm": 5.868343830108643, "learning_rate": 1.912343216949103e-07, "loss": 1.2573, "step": 46541 }, { "epoch": 3.6474921630094044, "grad_norm": 2.7250680923461914, "learning_rate": 1.911498650143112e-07, "loss": 1.2087, "step": 46542 }, { "epoch": 3.6475705329153607, "grad_norm": 2.836425304412842, "learning_rate": 1.9106542662410744e-07, "loss": 1.3131, "step": 46543 }, { "epoch": 3.6476489028213166, "grad_norm": 4.100310802459717, "learning_rate": 1.9098100652462092e-07, "loss": 1.4744, "step": 46544 }, { "epoch": 3.6477272727272725, "grad_norm": 3.1225554943084717, "learning_rate": 1.9089660471617312e-07, "loss": 1.5322, "step": 46545 }, { "epoch": 3.647805642633229, "grad_norm": 2.5199458599090576, "learning_rate": 1.9081222119908428e-07, "loss": 1.0767, "step": 46546 }, { "epoch": 3.647884012539185, "grad_norm": 2.9180588722229004, "learning_rate": 1.9072785597367527e-07, "loss": 0.7603, "step": 46547 }, { "epoch": 3.647962382445141, "grad_norm": 6.573885917663574, "learning_rate": 1.906435090402675e-07, "loss": 1.3104, "step": 46548 }, { "epoch": 3.648040752351097, "grad_norm": 2.2384989261627197, "learning_rate": 1.9055918039918132e-07, "loss": 1.3126, "step": 46549 }, { "epoch": 3.6481191222570533, "grad_norm": 2.563774824142456, "learning_rate": 1.9047487005073806e-07, "loss": 1.3006, "step": 46550 }, { "epoch": 3.6481974921630096, "grad_norm": 4.015052795410156, "learning_rate": 1.9039057799525808e-07, "loss": 1.3963, "step": 46551 }, { "epoch": 3.6482758620689655, "grad_norm": 6.534207820892334, "learning_rate": 1.9030630423306218e-07, "loss": 1.7829, "step": 46552 }, { "epoch": 3.6483542319749214, "grad_norm": 3.898646354675293, "learning_rate": 1.9022204876447014e-07, "loss": 1.3265, "step": 46553 }, { "epoch": 3.6484326018808777, "grad_norm": 3.3969967365264893, "learning_rate": 1.901378115898028e-07, "loss": 1.4778, "step": 46554 }, { "epoch": 3.648510971786834, "grad_norm": 3.892904758453369, "learning_rate": 1.9005359270938106e-07, "loss": 1.2671, "step": 46555 }, { "epoch": 3.64858934169279, "grad_norm": 3.967841625213623, "learning_rate": 1.8996939212352572e-07, "loss": 1.3859, "step": 46556 }, { "epoch": 3.648667711598746, "grad_norm": 3.697718858718872, "learning_rate": 1.8988520983255543e-07, "loss": 1.2138, "step": 46557 }, { "epoch": 3.648746081504702, "grad_norm": 2.5289363861083984, "learning_rate": 1.8980104583679105e-07, "loss": 1.2384, "step": 46558 }, { "epoch": 3.6488244514106585, "grad_norm": 2.257012128829956, "learning_rate": 1.897169001365523e-07, "loss": 0.9735, "step": 46559 }, { "epoch": 3.6489028213166144, "grad_norm": 6.989476680755615, "learning_rate": 1.8963277273216118e-07, "loss": 0.4935, "step": 46560 }, { "epoch": 3.6489811912225707, "grad_norm": 2.8156583309173584, "learning_rate": 1.8954866362393577e-07, "loss": 1.5738, "step": 46561 }, { "epoch": 3.6490595611285266, "grad_norm": 11.192975044250488, "learning_rate": 1.8946457281219577e-07, "loss": 1.5184, "step": 46562 }, { "epoch": 3.649137931034483, "grad_norm": 5.202812194824219, "learning_rate": 1.8938050029726206e-07, "loss": 1.3617, "step": 46563 }, { "epoch": 3.649216300940439, "grad_norm": 1.9211971759796143, "learning_rate": 1.8929644607945442e-07, "loss": 1.4548, "step": 46564 }, { "epoch": 3.649294670846395, "grad_norm": 5.6590447425842285, "learning_rate": 1.8921241015909141e-07, "loss": 1.3701, "step": 46565 }, { "epoch": 3.649373040752351, "grad_norm": 4.156264305114746, "learning_rate": 1.8912839253649396e-07, "loss": 1.2868, "step": 46566 }, { "epoch": 3.6494514106583074, "grad_norm": 3.3882644176483154, "learning_rate": 1.8904439321198064e-07, "loss": 1.4209, "step": 46567 }, { "epoch": 3.6495297805642632, "grad_norm": 2.7228939533233643, "learning_rate": 1.889604121858718e-07, "loss": 1.4036, "step": 46568 }, { "epoch": 3.6496081504702196, "grad_norm": 3.841160297393799, "learning_rate": 1.888764494584866e-07, "loss": 1.0519, "step": 46569 }, { "epoch": 3.6496865203761755, "grad_norm": 2.935122013092041, "learning_rate": 1.8879250503014367e-07, "loss": 1.1678, "step": 46570 }, { "epoch": 3.649764890282132, "grad_norm": 4.011915683746338, "learning_rate": 1.887085789011639e-07, "loss": 1.0472, "step": 46571 }, { "epoch": 3.6498432601880877, "grad_norm": 2.952690362930298, "learning_rate": 1.8862467107186422e-07, "loss": 1.2766, "step": 46572 }, { "epoch": 3.649921630094044, "grad_norm": 2.9260897636413574, "learning_rate": 1.8854078154256494e-07, "loss": 1.6287, "step": 46573 }, { "epoch": 3.65, "grad_norm": 2.909536600112915, "learning_rate": 1.884569103135858e-07, "loss": 1.1614, "step": 46574 }, { "epoch": 3.6500783699059562, "grad_norm": 4.49907922744751, "learning_rate": 1.8837305738524603e-07, "loss": 1.6704, "step": 46575 }, { "epoch": 3.650156739811912, "grad_norm": 3.340864658355713, "learning_rate": 1.8828922275786254e-07, "loss": 1.1709, "step": 46576 }, { "epoch": 3.6502351097178685, "grad_norm": 2.4076032638549805, "learning_rate": 1.8820540643175511e-07, "loss": 1.1407, "step": 46577 }, { "epoch": 3.6503134796238244, "grad_norm": 4.730109691619873, "learning_rate": 1.8812160840724402e-07, "loss": 1.3253, "step": 46578 }, { "epoch": 3.6503918495297807, "grad_norm": 3.690622091293335, "learning_rate": 1.8803782868464627e-07, "loss": 1.1641, "step": 46579 }, { "epoch": 3.6504702194357366, "grad_norm": 4.95335578918457, "learning_rate": 1.87954067264281e-07, "loss": 1.3361, "step": 46580 }, { "epoch": 3.650548589341693, "grad_norm": 3.2393085956573486, "learning_rate": 1.8787032414646688e-07, "loss": 1.2986, "step": 46581 }, { "epoch": 3.650626959247649, "grad_norm": 2.2742021083831787, "learning_rate": 1.8778659933152254e-07, "loss": 0.9781, "step": 46582 }, { "epoch": 3.650705329153605, "grad_norm": 19.265670776367188, "learning_rate": 1.8770289281976606e-07, "loss": 2.1524, "step": 46583 }, { "epoch": 3.650783699059561, "grad_norm": 8.546577453613281, "learning_rate": 1.8761920461151605e-07, "loss": 1.2401, "step": 46584 }, { "epoch": 3.6508620689655173, "grad_norm": 4.460048675537109, "learning_rate": 1.8753553470709063e-07, "loss": 1.559, "step": 46585 }, { "epoch": 3.6509404388714732, "grad_norm": 3.848008394241333, "learning_rate": 1.874518831068084e-07, "loss": 1.2493, "step": 46586 }, { "epoch": 3.6510188087774296, "grad_norm": 4.301901340484619, "learning_rate": 1.873682498109869e-07, "loss": 1.5801, "step": 46587 }, { "epoch": 3.6510971786833855, "grad_norm": 3.5041897296905518, "learning_rate": 1.872846348199453e-07, "loss": 1.2914, "step": 46588 }, { "epoch": 3.651175548589342, "grad_norm": 7.3269453048706055, "learning_rate": 1.872010381340006e-07, "loss": 1.5394, "step": 46589 }, { "epoch": 3.6512539184952977, "grad_norm": 2.822227716445923, "learning_rate": 1.8711745975347196e-07, "loss": 1.2959, "step": 46590 }, { "epoch": 3.651332288401254, "grad_norm": 4.16018009185791, "learning_rate": 1.870338996786747e-07, "loss": 1.4022, "step": 46591 }, { "epoch": 3.65141065830721, "grad_norm": 11.74030876159668, "learning_rate": 1.8695035790992966e-07, "loss": 1.7096, "step": 46592 }, { "epoch": 3.6514890282131662, "grad_norm": 2.626950263977051, "learning_rate": 1.8686683444755327e-07, "loss": 0.9144, "step": 46593 }, { "epoch": 3.651567398119122, "grad_norm": 3.998202323913574, "learning_rate": 1.8678332929186306e-07, "loss": 1.5045, "step": 46594 }, { "epoch": 3.6516457680250785, "grad_norm": 2.0347514152526855, "learning_rate": 1.8669984244317707e-07, "loss": 1.3431, "step": 46595 }, { "epoch": 3.6517241379310343, "grad_norm": 3.819956064224243, "learning_rate": 1.8661637390181174e-07, "loss": 1.6266, "step": 46596 }, { "epoch": 3.6518025078369907, "grad_norm": 8.241055488586426, "learning_rate": 1.8653292366808684e-07, "loss": 1.6418, "step": 46597 }, { "epoch": 3.6518808777429466, "grad_norm": 2.344729423522949, "learning_rate": 1.8644949174231764e-07, "loss": 1.3329, "step": 46598 }, { "epoch": 3.651959247648903, "grad_norm": 5.741220951080322, "learning_rate": 1.8636607812482167e-07, "loss": 0.6821, "step": 46599 }, { "epoch": 3.652037617554859, "grad_norm": 2.8984460830688477, "learning_rate": 1.86282682815917e-07, "loss": 1.3396, "step": 46600 }, { "epoch": 3.652115987460815, "grad_norm": 2.226433515548706, "learning_rate": 1.861993058159206e-07, "loss": 1.1899, "step": 46601 }, { "epoch": 3.652194357366771, "grad_norm": 4.284746170043945, "learning_rate": 1.8611594712515e-07, "loss": 1.6399, "step": 46602 }, { "epoch": 3.6522727272727273, "grad_norm": 3.455641746520996, "learning_rate": 1.860326067439211e-07, "loss": 1.0936, "step": 46603 }, { "epoch": 3.6523510971786832, "grad_norm": 3.6352107524871826, "learning_rate": 1.859492846725519e-07, "loss": 1.2767, "step": 46604 }, { "epoch": 3.6524294670846396, "grad_norm": 2.0607988834381104, "learning_rate": 1.858659809113589e-07, "loss": 1.2905, "step": 46605 }, { "epoch": 3.6525078369905954, "grad_norm": 8.546048164367676, "learning_rate": 1.8578269546065897e-07, "loss": 1.5994, "step": 46606 }, { "epoch": 3.652586206896552, "grad_norm": 4.515672206878662, "learning_rate": 1.856994283207686e-07, "loss": 1.6243, "step": 46607 }, { "epoch": 3.6526645768025077, "grad_norm": 2.3039119243621826, "learning_rate": 1.8561617949200584e-07, "loss": 0.9531, "step": 46608 }, { "epoch": 3.652742946708464, "grad_norm": 4.171536445617676, "learning_rate": 1.8553294897468544e-07, "loss": 1.2095, "step": 46609 }, { "epoch": 3.6528213166144203, "grad_norm": 2.5566351413726807, "learning_rate": 1.8544973676912493e-07, "loss": 1.1422, "step": 46610 }, { "epoch": 3.652899686520376, "grad_norm": 3.1051251888275146, "learning_rate": 1.8536654287564072e-07, "loss": 1.8153, "step": 46611 }, { "epoch": 3.652978056426332, "grad_norm": 2.560302495956421, "learning_rate": 1.8528336729454976e-07, "loss": 1.1146, "step": 46612 }, { "epoch": 3.6530564263322884, "grad_norm": 2.861191511154175, "learning_rate": 1.852002100261674e-07, "loss": 1.4106, "step": 46613 }, { "epoch": 3.6531347962382448, "grad_norm": 6.039234161376953, "learning_rate": 1.8511707107081057e-07, "loss": 1.336, "step": 46614 }, { "epoch": 3.6532131661442007, "grad_norm": 3.6648190021514893, "learning_rate": 1.8503395042879457e-07, "loss": 1.4018, "step": 46615 }, { "epoch": 3.6532915360501566, "grad_norm": 2.634828805923462, "learning_rate": 1.8495084810043752e-07, "loss": 1.3279, "step": 46616 }, { "epoch": 3.653369905956113, "grad_norm": 2.583204984664917, "learning_rate": 1.8486776408605355e-07, "loss": 1.2563, "step": 46617 }, { "epoch": 3.653448275862069, "grad_norm": 5.082291126251221, "learning_rate": 1.8478469838595913e-07, "loss": 1.146, "step": 46618 }, { "epoch": 3.653526645768025, "grad_norm": 2.511539936065674, "learning_rate": 1.8470165100047123e-07, "loss": 1.4585, "step": 46619 }, { "epoch": 3.653605015673981, "grad_norm": 3.2467589378356934, "learning_rate": 1.8461862192990453e-07, "loss": 1.3948, "step": 46620 }, { "epoch": 3.6536833855799373, "grad_norm": 4.454633712768555, "learning_rate": 1.8453561117457497e-07, "loss": 1.3804, "step": 46621 }, { "epoch": 3.6537617554858937, "grad_norm": 3.6598825454711914, "learning_rate": 1.8445261873479892e-07, "loss": 1.1087, "step": 46622 }, { "epoch": 3.6538401253918495, "grad_norm": 3.5506317615509033, "learning_rate": 1.8436964461089224e-07, "loss": 1.4521, "step": 46623 }, { "epoch": 3.6539184952978054, "grad_norm": 5.333252906799316, "learning_rate": 1.8428668880316913e-07, "loss": 1.1096, "step": 46624 }, { "epoch": 3.6539968652037618, "grad_norm": 2.5757980346679688, "learning_rate": 1.84203751311946e-07, "loss": 1.3897, "step": 46625 }, { "epoch": 3.654075235109718, "grad_norm": 6.652276992797852, "learning_rate": 1.8412083213753873e-07, "loss": 1.4845, "step": 46626 }, { "epoch": 3.654153605015674, "grad_norm": 7.937706470489502, "learning_rate": 1.8403793128026204e-07, "loss": 1.4915, "step": 46627 }, { "epoch": 3.65423197492163, "grad_norm": 2.475067615509033, "learning_rate": 1.8395504874043178e-07, "loss": 1.6485, "step": 46628 }, { "epoch": 3.654310344827586, "grad_norm": 4.141712665557861, "learning_rate": 1.8387218451836164e-07, "loss": 1.5018, "step": 46629 }, { "epoch": 3.6543887147335425, "grad_norm": 3.9280107021331787, "learning_rate": 1.8378933861436855e-07, "loss": 1.0838, "step": 46630 }, { "epoch": 3.6544670846394984, "grad_norm": 3.6091415882110596, "learning_rate": 1.837065110287678e-07, "loss": 1.1247, "step": 46631 }, { "epoch": 3.6545454545454543, "grad_norm": 6.651922702789307, "learning_rate": 1.8362370176187305e-07, "loss": 1.3296, "step": 46632 }, { "epoch": 3.6546238244514107, "grad_norm": 3.0296387672424316, "learning_rate": 1.8354091081400015e-07, "loss": 0.9864, "step": 46633 }, { "epoch": 3.654702194357367, "grad_norm": 3.4554803371429443, "learning_rate": 1.8345813818546331e-07, "loss": 1.1214, "step": 46634 }, { "epoch": 3.654780564263323, "grad_norm": 2.6657416820526123, "learning_rate": 1.8337538387657838e-07, "loss": 1.3594, "step": 46635 }, { "epoch": 3.6548589341692788, "grad_norm": 4.26055383682251, "learning_rate": 1.83292647887659e-07, "loss": 1.3163, "step": 46636 }, { "epoch": 3.654937304075235, "grad_norm": 3.117070436477661, "learning_rate": 1.8320993021902045e-07, "loss": 1.1675, "step": 46637 }, { "epoch": 3.6550156739811914, "grad_norm": 3.286134719848633, "learning_rate": 1.831272308709775e-07, "loss": 1.4078, "step": 46638 }, { "epoch": 3.6550940438871473, "grad_norm": 2.5820043087005615, "learning_rate": 1.8304454984384433e-07, "loss": 1.4772, "step": 46639 }, { "epoch": 3.655172413793103, "grad_norm": 4.472587585449219, "learning_rate": 1.8296188713793517e-07, "loss": 1.4605, "step": 46640 }, { "epoch": 3.6552507836990595, "grad_norm": 4.968594551086426, "learning_rate": 1.8287924275356528e-07, "loss": 1.6416, "step": 46641 }, { "epoch": 3.655329153605016, "grad_norm": 4.052314758300781, "learning_rate": 1.8279661669104887e-07, "loss": 1.5835, "step": 46642 }, { "epoch": 3.6554075235109718, "grad_norm": 2.593247890472412, "learning_rate": 1.8271400895069901e-07, "loss": 1.1856, "step": 46643 }, { "epoch": 3.6554858934169276, "grad_norm": 2.1316065788269043, "learning_rate": 1.8263141953283103e-07, "loss": 1.2674, "step": 46644 }, { "epoch": 3.655564263322884, "grad_norm": 3.2223761081695557, "learning_rate": 1.825488484377591e-07, "loss": 1.2796, "step": 46645 }, { "epoch": 3.6556426332288403, "grad_norm": 3.3465640544891357, "learning_rate": 1.8246629566579744e-07, "loss": 1.3365, "step": 46646 }, { "epoch": 3.655721003134796, "grad_norm": 3.4104702472686768, "learning_rate": 1.8238376121725853e-07, "loss": 1.2973, "step": 46647 }, { "epoch": 3.655799373040752, "grad_norm": 3.246572494506836, "learning_rate": 1.8230124509245717e-07, "loss": 1.5028, "step": 46648 }, { "epoch": 3.6558777429467084, "grad_norm": 4.441342353820801, "learning_rate": 1.8221874729170807e-07, "loss": 1.2073, "step": 46649 }, { "epoch": 3.6559561128526648, "grad_norm": 3.1063671112060547, "learning_rate": 1.8213626781532435e-07, "loss": 1.5347, "step": 46650 }, { "epoch": 3.6560344827586206, "grad_norm": 3.579139471054077, "learning_rate": 1.8205380666361906e-07, "loss": 1.5279, "step": 46651 }, { "epoch": 3.6561128526645765, "grad_norm": 2.5544867515563965, "learning_rate": 1.8197136383690694e-07, "loss": 1.222, "step": 46652 }, { "epoch": 3.656191222570533, "grad_norm": 4.608893394470215, "learning_rate": 1.8188893933550055e-07, "loss": 1.0526, "step": 46653 }, { "epoch": 3.656269592476489, "grad_norm": 5.0595598220825195, "learning_rate": 1.8180653315971408e-07, "loss": 1.6154, "step": 46654 }, { "epoch": 3.656347962382445, "grad_norm": 2.894524335861206, "learning_rate": 1.817241453098606e-07, "loss": 1.5976, "step": 46655 }, { "epoch": 3.6564263322884014, "grad_norm": 6.590143203735352, "learning_rate": 1.816417757862543e-07, "loss": 1.4849, "step": 46656 }, { "epoch": 3.6565047021943573, "grad_norm": 2.7605321407318115, "learning_rate": 1.8155942458920718e-07, "loss": 1.2856, "step": 46657 }, { "epoch": 3.6565830721003136, "grad_norm": 4.89241361618042, "learning_rate": 1.814770917190334e-07, "loss": 1.3638, "step": 46658 }, { "epoch": 3.6566614420062695, "grad_norm": 8.922255516052246, "learning_rate": 1.8139477717604548e-07, "loss": 0.9217, "step": 46659 }, { "epoch": 3.656739811912226, "grad_norm": 2.866285562515259, "learning_rate": 1.813124809605571e-07, "loss": 1.1336, "step": 46660 }, { "epoch": 3.6568181818181817, "grad_norm": 2.2569258213043213, "learning_rate": 1.812302030728813e-07, "loss": 1.3984, "step": 46661 }, { "epoch": 3.656896551724138, "grad_norm": 4.678234577178955, "learning_rate": 1.8114794351333064e-07, "loss": 1.2878, "step": 46662 }, { "epoch": 3.656974921630094, "grad_norm": 3.6292500495910645, "learning_rate": 1.8106570228221766e-07, "loss": 1.5182, "step": 46663 }, { "epoch": 3.6570532915360503, "grad_norm": 9.847553253173828, "learning_rate": 1.809834793798565e-07, "loss": 0.8847, "step": 46664 }, { "epoch": 3.657131661442006, "grad_norm": 2.9969048500061035, "learning_rate": 1.8090127480655806e-07, "loss": 1.4837, "step": 46665 }, { "epoch": 3.6572100313479625, "grad_norm": 2.569819927215576, "learning_rate": 1.8081908856263542e-07, "loss": 1.1072, "step": 46666 }, { "epoch": 3.6572884012539184, "grad_norm": 3.8794643878936768, "learning_rate": 1.8073692064840275e-07, "loss": 0.8629, "step": 46667 }, { "epoch": 3.6573667711598747, "grad_norm": 3.2272355556488037, "learning_rate": 1.806547710641715e-07, "loss": 1.2945, "step": 46668 }, { "epoch": 3.6574451410658306, "grad_norm": 3.7412807941436768, "learning_rate": 1.8057263981025364e-07, "loss": 1.4671, "step": 46669 }, { "epoch": 3.657523510971787, "grad_norm": 4.10279655456543, "learning_rate": 1.8049052688696224e-07, "loss": 1.6774, "step": 46670 }, { "epoch": 3.657601880877743, "grad_norm": 4.332193851470947, "learning_rate": 1.804084322946098e-07, "loss": 1.1905, "step": 46671 }, { "epoch": 3.657680250783699, "grad_norm": 3.757296085357666, "learning_rate": 1.803263560335078e-07, "loss": 1.5186, "step": 46672 }, { "epoch": 3.657758620689655, "grad_norm": 3.4536571502685547, "learning_rate": 1.8024429810396872e-07, "loss": 1.5468, "step": 46673 }, { "epoch": 3.6578369905956114, "grad_norm": 2.5234289169311523, "learning_rate": 1.8016225850630453e-07, "loss": 1.4266, "step": 46674 }, { "epoch": 3.6579153605015673, "grad_norm": 2.8741300106048584, "learning_rate": 1.800802372408278e-07, "loss": 1.3807, "step": 46675 }, { "epoch": 3.6579937304075236, "grad_norm": 2.2647619247436523, "learning_rate": 1.799982343078499e-07, "loss": 1.2667, "step": 46676 }, { "epoch": 3.6580721003134795, "grad_norm": 3.129629135131836, "learning_rate": 1.7991624970768339e-07, "loss": 1.2144, "step": 46677 }, { "epoch": 3.658150470219436, "grad_norm": 3.03599214553833, "learning_rate": 1.7983428344063968e-07, "loss": 1.3328, "step": 46678 }, { "epoch": 3.6582288401253917, "grad_norm": 5.658792018890381, "learning_rate": 1.7975233550703075e-07, "loss": 1.9439, "step": 46679 }, { "epoch": 3.658307210031348, "grad_norm": 2.7903449535369873, "learning_rate": 1.7967040590716745e-07, "loss": 0.7379, "step": 46680 }, { "epoch": 3.658385579937304, "grad_norm": 8.238222122192383, "learning_rate": 1.795884946413623e-07, "loss": 0.673, "step": 46681 }, { "epoch": 3.6584639498432603, "grad_norm": 2.4772582054138184, "learning_rate": 1.7950660170992674e-07, "loss": 1.4799, "step": 46682 }, { "epoch": 3.658542319749216, "grad_norm": 3.484299421310425, "learning_rate": 1.794247271131727e-07, "loss": 1.4597, "step": 46683 }, { "epoch": 3.6586206896551725, "grad_norm": 4.299526691436768, "learning_rate": 1.7934287085141e-07, "loss": 1.3014, "step": 46684 }, { "epoch": 3.6586990595611284, "grad_norm": 4.456565856933594, "learning_rate": 1.7926103292495113e-07, "loss": 1.0013, "step": 46685 }, { "epoch": 3.6587774294670847, "grad_norm": 4.323871612548828, "learning_rate": 1.7917921333410693e-07, "loss": 1.6991, "step": 46686 }, { "epoch": 3.6588557993730406, "grad_norm": 4.0123186111450195, "learning_rate": 1.7909741207918995e-07, "loss": 1.4603, "step": 46687 }, { "epoch": 3.658934169278997, "grad_norm": 3.612471342086792, "learning_rate": 1.7901562916050996e-07, "loss": 1.5375, "step": 46688 }, { "epoch": 3.659012539184953, "grad_norm": 3.27371883392334, "learning_rate": 1.789338645783778e-07, "loss": 1.1965, "step": 46689 }, { "epoch": 3.659090909090909, "grad_norm": 2.677446126937866, "learning_rate": 1.788521183331049e-07, "loss": 1.2041, "step": 46690 }, { "epoch": 3.659169278996865, "grad_norm": 5.683537483215332, "learning_rate": 1.7877039042500266e-07, "loss": 1.9483, "step": 46691 }, { "epoch": 3.6592476489028214, "grad_norm": 3.731013298034668, "learning_rate": 1.786886808543814e-07, "loss": 1.5442, "step": 46692 }, { "epoch": 3.6593260188087773, "grad_norm": 2.0695741176605225, "learning_rate": 1.7860698962155197e-07, "loss": 1.1859, "step": 46693 }, { "epoch": 3.6594043887147336, "grad_norm": 5.301535129547119, "learning_rate": 1.7852531672682526e-07, "loss": 1.2346, "step": 46694 }, { "epoch": 3.6594827586206895, "grad_norm": 8.047179222106934, "learning_rate": 1.7844366217051157e-07, "loss": 0.3559, "step": 46695 }, { "epoch": 3.659561128526646, "grad_norm": 3.230989933013916, "learning_rate": 1.7836202595292173e-07, "loss": 1.223, "step": 46696 }, { "epoch": 3.6596394984326017, "grad_norm": 9.861932754516602, "learning_rate": 1.7828040807436608e-07, "loss": 0.9438, "step": 46697 }, { "epoch": 3.659717868338558, "grad_norm": 2.9626147747039795, "learning_rate": 1.7819880853515604e-07, "loss": 1.185, "step": 46698 }, { "epoch": 3.659796238244514, "grad_norm": 3.815704345703125, "learning_rate": 1.7811722733559965e-07, "loss": 1.2074, "step": 46699 }, { "epoch": 3.6598746081504703, "grad_norm": 14.963894844055176, "learning_rate": 1.7803566447600895e-07, "loss": 0.7955, "step": 46700 }, { "epoch": 3.6599529780564266, "grad_norm": 3.072040557861328, "learning_rate": 1.7795411995669477e-07, "loss": 1.5707, "step": 46701 }, { "epoch": 3.6600313479623825, "grad_norm": 2.9691665172576904, "learning_rate": 1.7787259377796628e-07, "loss": 1.5364, "step": 46702 }, { "epoch": 3.6601097178683384, "grad_norm": 4.607242584228516, "learning_rate": 1.7779108594013327e-07, "loss": 1.1729, "step": 46703 }, { "epoch": 3.6601880877742947, "grad_norm": 4.295454502105713, "learning_rate": 1.7770959644350548e-07, "loss": 1.5714, "step": 46704 }, { "epoch": 3.660266457680251, "grad_norm": 3.6937942504882812, "learning_rate": 1.7762812528839435e-07, "loss": 1.5066, "step": 46705 }, { "epoch": 3.660344827586207, "grad_norm": 3.9666402339935303, "learning_rate": 1.7754667247510903e-07, "loss": 1.2199, "step": 46706 }, { "epoch": 3.660423197492163, "grad_norm": 2.99983286857605, "learning_rate": 1.7746523800395877e-07, "loss": 1.3217, "step": 46707 }, { "epoch": 3.660501567398119, "grad_norm": 6.597776889801025, "learning_rate": 1.7738382187525326e-07, "loss": 1.186, "step": 46708 }, { "epoch": 3.6605799373040755, "grad_norm": 8.27688980102539, "learning_rate": 1.7730242408930344e-07, "loss": 1.867, "step": 46709 }, { "epoch": 3.6606583072100314, "grad_norm": 3.747051477432251, "learning_rate": 1.7722104464641788e-07, "loss": 0.977, "step": 46710 }, { "epoch": 3.6607366771159873, "grad_norm": 2.927150011062622, "learning_rate": 1.7713968354690636e-07, "loss": 1.6266, "step": 46711 }, { "epoch": 3.6608150470219436, "grad_norm": 14.079005241394043, "learning_rate": 1.770583407910781e-07, "loss": 0.6958, "step": 46712 }, { "epoch": 3.6608934169279, "grad_norm": 3.5361056327819824, "learning_rate": 1.7697701637924282e-07, "loss": 1.3254, "step": 46713 }, { "epoch": 3.660971786833856, "grad_norm": 2.68550968170166, "learning_rate": 1.7689571031170972e-07, "loss": 1.4252, "step": 46714 }, { "epoch": 3.6610501567398117, "grad_norm": 3.979095935821533, "learning_rate": 1.7681442258878801e-07, "loss": 1.0921, "step": 46715 }, { "epoch": 3.661128526645768, "grad_norm": 3.104247570037842, "learning_rate": 1.7673315321078743e-07, "loss": 1.2046, "step": 46716 }, { "epoch": 3.6612068965517244, "grad_norm": 4.196219444274902, "learning_rate": 1.766519021780161e-07, "loss": 1.4567, "step": 46717 }, { "epoch": 3.6612852664576803, "grad_norm": 2.6867058277130127, "learning_rate": 1.765706694907826e-07, "loss": 1.3057, "step": 46718 }, { "epoch": 3.661363636363636, "grad_norm": 4.900984764099121, "learning_rate": 1.7648945514939787e-07, "loss": 1.039, "step": 46719 }, { "epoch": 3.6614420062695925, "grad_norm": 8.503230094909668, "learning_rate": 1.7640825915416994e-07, "loss": 2.1011, "step": 46720 }, { "epoch": 3.661520376175549, "grad_norm": 4.396987438201904, "learning_rate": 1.763270815054069e-07, "loss": 1.4456, "step": 46721 }, { "epoch": 3.6615987460815047, "grad_norm": 2.9459638595581055, "learning_rate": 1.7624592220341796e-07, "loss": 1.3663, "step": 46722 }, { "epoch": 3.6616771159874606, "grad_norm": 3.364811897277832, "learning_rate": 1.7616478124851178e-07, "loss": 1.167, "step": 46723 }, { "epoch": 3.661755485893417, "grad_norm": 10.562739372253418, "learning_rate": 1.7608365864099807e-07, "loss": 1.5466, "step": 46724 }, { "epoch": 3.6618338557993733, "grad_norm": 4.020376682281494, "learning_rate": 1.760025543811833e-07, "loss": 0.9184, "step": 46725 }, { "epoch": 3.661912225705329, "grad_norm": 4.645951271057129, "learning_rate": 1.7592146846937773e-07, "loss": 1.0914, "step": 46726 }, { "epoch": 3.661990595611285, "grad_norm": 4.512115478515625, "learning_rate": 1.7584040090588894e-07, "loss": 1.6431, "step": 46727 }, { "epoch": 3.6620689655172414, "grad_norm": 2.704586982727051, "learning_rate": 1.7575935169102497e-07, "loss": 1.1806, "step": 46728 }, { "epoch": 3.6621473354231977, "grad_norm": 2.960585355758667, "learning_rate": 1.7567832082509505e-07, "loss": 1.4512, "step": 46729 }, { "epoch": 3.6622257053291536, "grad_norm": 2.678159475326538, "learning_rate": 1.7559730830840672e-07, "loss": 1.4805, "step": 46730 }, { "epoch": 3.6623040752351095, "grad_norm": 4.3622307777404785, "learning_rate": 1.7551631414126857e-07, "loss": 1.2139, "step": 46731 }, { "epoch": 3.662382445141066, "grad_norm": 10.329702377319336, "learning_rate": 1.7543533832398763e-07, "loss": 1.5257, "step": 46732 }, { "epoch": 3.662460815047022, "grad_norm": 9.309257507324219, "learning_rate": 1.7535438085687306e-07, "loss": 1.6429, "step": 46733 }, { "epoch": 3.662539184952978, "grad_norm": 4.131141662597656, "learning_rate": 1.752734417402324e-07, "loss": 2.1081, "step": 46734 }, { "epoch": 3.662617554858934, "grad_norm": 2.867158889770508, "learning_rate": 1.7519252097437377e-07, "loss": 1.4709, "step": 46735 }, { "epoch": 3.6626959247648903, "grad_norm": 4.787044048309326, "learning_rate": 1.7511161855960413e-07, "loss": 1.5088, "step": 46736 }, { "epoch": 3.6627742946708466, "grad_norm": 8.172210693359375, "learning_rate": 1.7503073449623099e-07, "loss": 1.4994, "step": 46737 }, { "epoch": 3.6628526645768025, "grad_norm": 5.840814113616943, "learning_rate": 1.7494986878456354e-07, "loss": 1.2943, "step": 46738 }, { "epoch": 3.6629310344827584, "grad_norm": 3.022778034210205, "learning_rate": 1.748690214249088e-07, "loss": 1.4236, "step": 46739 }, { "epoch": 3.6630094043887147, "grad_norm": 3.055431604385376, "learning_rate": 1.7478819241757316e-07, "loss": 1.3735, "step": 46740 }, { "epoch": 3.663087774294671, "grad_norm": 2.5719473361968994, "learning_rate": 1.7470738176286528e-07, "loss": 1.1173, "step": 46741 }, { "epoch": 3.663166144200627, "grad_norm": 29.277542114257812, "learning_rate": 1.7462658946109155e-07, "loss": 0.8317, "step": 46742 }, { "epoch": 3.663244514106583, "grad_norm": 3.2168667316436768, "learning_rate": 1.7454581551256068e-07, "loss": 1.4678, "step": 46743 }, { "epoch": 3.663322884012539, "grad_norm": 3.7513270378112793, "learning_rate": 1.7446505991757845e-07, "loss": 1.5836, "step": 46744 }, { "epoch": 3.6634012539184955, "grad_norm": 2.641953945159912, "learning_rate": 1.74384322676453e-07, "loss": 1.2741, "step": 46745 }, { "epoch": 3.6634796238244514, "grad_norm": 2.5048747062683105, "learning_rate": 1.7430360378949073e-07, "loss": 1.5381, "step": 46746 }, { "epoch": 3.6635579937304072, "grad_norm": 4.089677810668945, "learning_rate": 1.7422290325699864e-07, "loss": 1.4257, "step": 46747 }, { "epoch": 3.6636363636363636, "grad_norm": 4.14626932144165, "learning_rate": 1.7414222107928425e-07, "loss": 1.5752, "step": 46748 }, { "epoch": 3.66371473354232, "grad_norm": 3.838193416595459, "learning_rate": 1.7406155725665397e-07, "loss": 1.546, "step": 46749 }, { "epoch": 3.663793103448276, "grad_norm": 4.628373622894287, "learning_rate": 1.739809117894148e-07, "loss": 1.4499, "step": 46750 }, { "epoch": 3.663871473354232, "grad_norm": 2.775134325027466, "learning_rate": 1.7390028467787368e-07, "loss": 1.3184, "step": 46751 }, { "epoch": 3.663949843260188, "grad_norm": 3.1099588871002197, "learning_rate": 1.7381967592233705e-07, "loss": 0.8271, "step": 46752 }, { "epoch": 3.6640282131661444, "grad_norm": 4.003339767456055, "learning_rate": 1.7373908552311136e-07, "loss": 1.3937, "step": 46753 }, { "epoch": 3.6641065830721002, "grad_norm": 6.915427207946777, "learning_rate": 1.7365851348050357e-07, "loss": 1.2122, "step": 46754 }, { "epoch": 3.6641849529780566, "grad_norm": 2.5354769229888916, "learning_rate": 1.735779597948195e-07, "loss": 1.4051, "step": 46755 }, { "epoch": 3.6642633228840125, "grad_norm": 3.5179572105407715, "learning_rate": 1.7349742446636508e-07, "loss": 1.2682, "step": 46756 }, { "epoch": 3.664341692789969, "grad_norm": 3.7054567337036133, "learning_rate": 1.7341690749544837e-07, "loss": 1.6514, "step": 46757 }, { "epoch": 3.6644200626959247, "grad_norm": 3.0400924682617188, "learning_rate": 1.7333640888237525e-07, "loss": 1.5091, "step": 46758 }, { "epoch": 3.664498432601881, "grad_norm": 3.6644234657287598, "learning_rate": 1.7325592862745043e-07, "loss": 1.123, "step": 46759 }, { "epoch": 3.664576802507837, "grad_norm": 2.3688621520996094, "learning_rate": 1.7317546673098095e-07, "loss": 0.9746, "step": 46760 }, { "epoch": 3.6646551724137932, "grad_norm": 5.894595623016357, "learning_rate": 1.7309502319327265e-07, "loss": 1.5012, "step": 46761 }, { "epoch": 3.664733542319749, "grad_norm": 4.507801055908203, "learning_rate": 1.7301459801463195e-07, "loss": 1.499, "step": 46762 }, { "epoch": 3.6648119122257055, "grad_norm": 29.065689086914062, "learning_rate": 1.7293419119536414e-07, "loss": 1.6371, "step": 46763 }, { "epoch": 3.6648902821316613, "grad_norm": 4.842043399810791, "learning_rate": 1.7285380273577513e-07, "loss": 1.4468, "step": 46764 }, { "epoch": 3.6649686520376177, "grad_norm": 2.3454177379608154, "learning_rate": 1.727734326361713e-07, "loss": 1.1189, "step": 46765 }, { "epoch": 3.6650470219435736, "grad_norm": 3.027724027633667, "learning_rate": 1.72693080896858e-07, "loss": 1.6467, "step": 46766 }, { "epoch": 3.66512539184953, "grad_norm": 4.410182476043701, "learning_rate": 1.726127475181405e-07, "loss": 1.306, "step": 46767 }, { "epoch": 3.665203761755486, "grad_norm": 5.092247486114502, "learning_rate": 1.7253243250032468e-07, "loss": 1.4683, "step": 46768 }, { "epoch": 3.665282131661442, "grad_norm": 8.497800827026367, "learning_rate": 1.724521358437159e-07, "loss": 1.6588, "step": 46769 }, { "epoch": 3.665360501567398, "grad_norm": 6.280685901641846, "learning_rate": 1.7237185754861942e-07, "loss": 1.2934, "step": 46770 }, { "epoch": 3.6654388714733543, "grad_norm": 2.8356893062591553, "learning_rate": 1.722915976153411e-07, "loss": 1.3892, "step": 46771 }, { "epoch": 3.6655172413793102, "grad_norm": 3.688584804534912, "learning_rate": 1.7221135604418627e-07, "loss": 1.0731, "step": 46772 }, { "epoch": 3.6655956112852666, "grad_norm": 5.084505081176758, "learning_rate": 1.721311328354591e-07, "loss": 1.4904, "step": 46773 }, { "epoch": 3.6656739811912225, "grad_norm": 5.540501117706299, "learning_rate": 1.72050927989465e-07, "loss": 2.4903, "step": 46774 }, { "epoch": 3.665752351097179, "grad_norm": 4.8791303634643555, "learning_rate": 1.7197074150650973e-07, "loss": 1.8613, "step": 46775 }, { "epoch": 3.6658307210031347, "grad_norm": 3.049499750137329, "learning_rate": 1.7189057338689863e-07, "loss": 1.0652, "step": 46776 }, { "epoch": 3.665909090909091, "grad_norm": 4.787376403808594, "learning_rate": 1.7181042363093536e-07, "loss": 1.2044, "step": 46777 }, { "epoch": 3.665987460815047, "grad_norm": 2.3699162006378174, "learning_rate": 1.7173029223892522e-07, "loss": 1.3407, "step": 46778 }, { "epoch": 3.6660658307210032, "grad_norm": 3.895742654800415, "learning_rate": 1.7165017921117245e-07, "loss": 1.5357, "step": 46779 }, { "epoch": 3.666144200626959, "grad_norm": 3.7121245861053467, "learning_rate": 1.7157008454798395e-07, "loss": 1.4727, "step": 46780 }, { "epoch": 3.6662225705329154, "grad_norm": 2.833491325378418, "learning_rate": 1.7149000824966178e-07, "loss": 1.3225, "step": 46781 }, { "epoch": 3.6663009404388713, "grad_norm": 3.7393393516540527, "learning_rate": 1.7140995031651176e-07, "loss": 1.1542, "step": 46782 }, { "epoch": 3.6663793103448277, "grad_norm": 3.034541606903076, "learning_rate": 1.713299107488381e-07, "loss": 1.578, "step": 46783 }, { "epoch": 3.6664576802507836, "grad_norm": 4.326314926147461, "learning_rate": 1.7124988954694555e-07, "loss": 1.4648, "step": 46784 }, { "epoch": 3.66653605015674, "grad_norm": 11.331803321838379, "learning_rate": 1.7116988671113776e-07, "loss": 1.0874, "step": 46785 }, { "epoch": 3.666614420062696, "grad_norm": 3.9929885864257812, "learning_rate": 1.710899022417195e-07, "loss": 1.5313, "step": 46786 }, { "epoch": 3.666692789968652, "grad_norm": 3.4133074283599854, "learning_rate": 1.7100993613899607e-07, "loss": 1.5525, "step": 46787 }, { "epoch": 3.666771159874608, "grad_norm": 2.909531354904175, "learning_rate": 1.709299884032689e-07, "loss": 1.5037, "step": 46788 }, { "epoch": 3.6668495297805643, "grad_norm": 5.303790092468262, "learning_rate": 1.7085005903484442e-07, "loss": 1.3199, "step": 46789 }, { "epoch": 3.66692789968652, "grad_norm": 3.9224143028259277, "learning_rate": 1.707701480340257e-07, "loss": 2.0785, "step": 46790 }, { "epoch": 3.6670062695924766, "grad_norm": 7.488043308258057, "learning_rate": 1.7069025540111806e-07, "loss": 1.0003, "step": 46791 }, { "epoch": 3.6670846394984324, "grad_norm": 3.4198803901672363, "learning_rate": 1.706103811364229e-07, "loss": 1.7933, "step": 46792 }, { "epoch": 3.6671630094043888, "grad_norm": 3.3128209114074707, "learning_rate": 1.7053052524024504e-07, "loss": 1.4225, "step": 46793 }, { "epoch": 3.6672413793103447, "grad_norm": 3.5753133296966553, "learning_rate": 1.7045068771288918e-07, "loss": 1.2656, "step": 46794 }, { "epoch": 3.667319749216301, "grad_norm": 4.3570756912231445, "learning_rate": 1.7037086855465902e-07, "loss": 1.1857, "step": 46795 }, { "epoch": 3.6673981191222573, "grad_norm": 3.090272903442383, "learning_rate": 1.7029106776585646e-07, "loss": 1.5315, "step": 46796 }, { "epoch": 3.667476489028213, "grad_norm": 3.339327096939087, "learning_rate": 1.7021128534678632e-07, "loss": 1.7237, "step": 46797 }, { "epoch": 3.667554858934169, "grad_norm": 2.899078369140625, "learning_rate": 1.7013152129775057e-07, "loss": 1.0555, "step": 46798 }, { "epoch": 3.6676332288401254, "grad_norm": 4.20499849319458, "learning_rate": 1.7005177561905562e-07, "loss": 1.5583, "step": 46799 }, { "epoch": 3.6677115987460818, "grad_norm": 3.486682653427124, "learning_rate": 1.6997204831100178e-07, "loss": 1.2449, "step": 46800 }, { "epoch": 3.6677899686520377, "grad_norm": 6.466569423675537, "learning_rate": 1.6989233937389383e-07, "loss": 1.4772, "step": 46801 }, { "epoch": 3.6678683385579935, "grad_norm": 2.5763328075408936, "learning_rate": 1.6981264880803429e-07, "loss": 1.3704, "step": 46802 }, { "epoch": 3.66794670846395, "grad_norm": 5.646088123321533, "learning_rate": 1.6973297661372624e-07, "loss": 1.5962, "step": 46803 }, { "epoch": 3.668025078369906, "grad_norm": 2.451446533203125, "learning_rate": 1.6965332279127334e-07, "loss": 1.1304, "step": 46804 }, { "epoch": 3.668103448275862, "grad_norm": 4.595903396606445, "learning_rate": 1.6957368734097813e-07, "loss": 1.318, "step": 46805 }, { "epoch": 3.668181818181818, "grad_norm": 3.914003372192383, "learning_rate": 1.694940702631437e-07, "loss": 1.27, "step": 46806 }, { "epoch": 3.6682601880877743, "grad_norm": 5.018250465393066, "learning_rate": 1.69414471558072e-07, "loss": 1.5658, "step": 46807 }, { "epoch": 3.6683385579937307, "grad_norm": 3.2253541946411133, "learning_rate": 1.6933489122606727e-07, "loss": 1.3063, "step": 46808 }, { "epoch": 3.6684169278996865, "grad_norm": 2.609217882156372, "learning_rate": 1.6925532926743092e-07, "loss": 1.3634, "step": 46809 }, { "epoch": 3.6684952978056424, "grad_norm": 2.891134738922119, "learning_rate": 1.6917578568246717e-07, "loss": 1.5244, "step": 46810 }, { "epoch": 3.6685736677115988, "grad_norm": 5.933414459228516, "learning_rate": 1.690962604714763e-07, "loss": 1.6929, "step": 46811 }, { "epoch": 3.668652037617555, "grad_norm": 2.4628098011016846, "learning_rate": 1.69016753634762e-07, "loss": 1.3495, "step": 46812 }, { "epoch": 3.668730407523511, "grad_norm": 3.4648032188415527, "learning_rate": 1.6893726517262675e-07, "loss": 1.421, "step": 46813 }, { "epoch": 3.668808777429467, "grad_norm": 4.839691162109375, "learning_rate": 1.6885779508537315e-07, "loss": 1.4152, "step": 46814 }, { "epoch": 3.668887147335423, "grad_norm": 7.512823581695557, "learning_rate": 1.6877834337330312e-07, "loss": 0.8796, "step": 46815 }, { "epoch": 3.6689655172413795, "grad_norm": 2.4443135261535645, "learning_rate": 1.6869891003671812e-07, "loss": 1.3252, "step": 46816 }, { "epoch": 3.6690438871473354, "grad_norm": 3.515000820159912, "learning_rate": 1.6861949507592124e-07, "loss": 1.2511, "step": 46817 }, { "epoch": 3.6691222570532913, "grad_norm": 4.06938362121582, "learning_rate": 1.6854009849121445e-07, "loss": 1.3549, "step": 46818 }, { "epoch": 3.6692006269592476, "grad_norm": 4.036377906799316, "learning_rate": 1.6846072028289972e-07, "loss": 1.3364, "step": 46819 }, { "epoch": 3.669278996865204, "grad_norm": 4.822487831115723, "learning_rate": 1.6838136045127796e-07, "loss": 0.9419, "step": 46820 }, { "epoch": 3.66935736677116, "grad_norm": 2.4497740268707275, "learning_rate": 1.6830201899665277e-07, "loss": 1.3449, "step": 46821 }, { "epoch": 3.6694357366771158, "grad_norm": 2.6251964569091797, "learning_rate": 1.682226959193245e-07, "loss": 1.1102, "step": 46822 }, { "epoch": 3.669514106583072, "grad_norm": 3.316906213760376, "learning_rate": 1.681433912195951e-07, "loss": 1.3138, "step": 46823 }, { "epoch": 3.6695924764890284, "grad_norm": 3.7536966800689697, "learning_rate": 1.680641048977666e-07, "loss": 1.0264, "step": 46824 }, { "epoch": 3.6696708463949843, "grad_norm": 5.228165149688721, "learning_rate": 1.6798483695414147e-07, "loss": 1.4362, "step": 46825 }, { "epoch": 3.66974921630094, "grad_norm": 3.4689173698425293, "learning_rate": 1.6790558738901842e-07, "loss": 1.0749, "step": 46826 }, { "epoch": 3.6698275862068965, "grad_norm": 3.596909523010254, "learning_rate": 1.6782635620270105e-07, "loss": 1.1457, "step": 46827 }, { "epoch": 3.669905956112853, "grad_norm": 3.9554712772369385, "learning_rate": 1.6774714339549137e-07, "loss": 1.2527, "step": 46828 }, { "epoch": 3.6699843260188088, "grad_norm": 2.918985366821289, "learning_rate": 1.6766794896768857e-07, "loss": 1.3112, "step": 46829 }, { "epoch": 3.6700626959247646, "grad_norm": 5.4037041664123535, "learning_rate": 1.675887729195952e-07, "loss": 1.2648, "step": 46830 }, { "epoch": 3.670141065830721, "grad_norm": 3.3884358406066895, "learning_rate": 1.6750961525151156e-07, "loss": 1.0629, "step": 46831 }, { "epoch": 3.6702194357366773, "grad_norm": 4.809649467468262, "learning_rate": 1.6743047596373962e-07, "loss": 1.1831, "step": 46832 }, { "epoch": 3.670297805642633, "grad_norm": 13.909817695617676, "learning_rate": 1.6735135505658028e-07, "loss": 1.4034, "step": 46833 }, { "epoch": 3.670376175548589, "grad_norm": 2.360656261444092, "learning_rate": 1.6727225253033384e-07, "loss": 1.3608, "step": 46834 }, { "epoch": 3.6704545454545454, "grad_norm": 2.8292605876922607, "learning_rate": 1.671931683853012e-07, "loss": 1.3653, "step": 46835 }, { "epoch": 3.6705329153605017, "grad_norm": 4.076836585998535, "learning_rate": 1.6711410262178375e-07, "loss": 1.9251, "step": 46836 }, { "epoch": 3.6706112852664576, "grad_norm": 2.7563092708587646, "learning_rate": 1.670350552400818e-07, "loss": 1.314, "step": 46837 }, { "epoch": 3.6706896551724135, "grad_norm": 2.141521692276001, "learning_rate": 1.6695602624049623e-07, "loss": 1.3498, "step": 46838 }, { "epoch": 3.67076802507837, "grad_norm": 2.9047441482543945, "learning_rate": 1.6687701562332791e-07, "loss": 1.1107, "step": 46839 }, { "epoch": 3.670846394984326, "grad_norm": 4.996796607971191, "learning_rate": 1.6679802338887662e-07, "loss": 1.5235, "step": 46840 }, { "epoch": 3.670924764890282, "grad_norm": 7.074965000152588, "learning_rate": 1.6671904953744266e-07, "loss": 1.9082, "step": 46841 }, { "epoch": 3.6710031347962384, "grad_norm": 4.1462626457214355, "learning_rate": 1.6664009406932746e-07, "loss": 1.8579, "step": 46842 }, { "epoch": 3.6710815047021943, "grad_norm": 3.8729248046875, "learning_rate": 1.6656115698483133e-07, "loss": 1.4073, "step": 46843 }, { "epoch": 3.6711598746081506, "grad_norm": 2.4877357482910156, "learning_rate": 1.6648223828425347e-07, "loss": 1.551, "step": 46844 }, { "epoch": 3.6712382445141065, "grad_norm": 5.515848159790039, "learning_rate": 1.6640333796789366e-07, "loss": 0.7761, "step": 46845 }, { "epoch": 3.671316614420063, "grad_norm": 3.8664519786834717, "learning_rate": 1.6632445603605384e-07, "loss": 1.1772, "step": 46846 }, { "epoch": 3.6713949843260187, "grad_norm": 2.454399824142456, "learning_rate": 1.6624559248903328e-07, "loss": 1.2618, "step": 46847 }, { "epoch": 3.671473354231975, "grad_norm": 4.566474914550781, "learning_rate": 1.6616674732713113e-07, "loss": 1.4017, "step": 46848 }, { "epoch": 3.671551724137931, "grad_norm": 4.818544864654541, "learning_rate": 1.660879205506477e-07, "loss": 1.8747, "step": 46849 }, { "epoch": 3.6716300940438873, "grad_norm": 5.527749061584473, "learning_rate": 1.660091121598828e-07, "loss": 1.5113, "step": 46850 }, { "epoch": 3.671708463949843, "grad_norm": 3.2072649002075195, "learning_rate": 1.6593032215513727e-07, "loss": 1.7739, "step": 46851 }, { "epoch": 3.6717868338557995, "grad_norm": 2.247206449508667, "learning_rate": 1.6585155053670975e-07, "loss": 1.1628, "step": 46852 }, { "epoch": 3.6718652037617554, "grad_norm": 4.820836067199707, "learning_rate": 1.6577279730489948e-07, "loss": 1.3084, "step": 46853 }, { "epoch": 3.6719435736677117, "grad_norm": 3.099883556365967, "learning_rate": 1.656940624600062e-07, "loss": 1.7415, "step": 46854 }, { "epoch": 3.6720219435736676, "grad_norm": 3.167142152786255, "learning_rate": 1.656153460023302e-07, "loss": 1.0207, "step": 46855 }, { "epoch": 3.672100313479624, "grad_norm": 3.002176284790039, "learning_rate": 1.6553664793217016e-07, "loss": 0.9563, "step": 46856 }, { "epoch": 3.67217868338558, "grad_norm": 2.883575439453125, "learning_rate": 1.654579682498253e-07, "loss": 1.2085, "step": 46857 }, { "epoch": 3.672257053291536, "grad_norm": 7.378574371337891, "learning_rate": 1.6537930695559535e-07, "loss": 1.3918, "step": 46858 }, { "epoch": 3.672335423197492, "grad_norm": 3.5551185607910156, "learning_rate": 1.6530066404977952e-07, "loss": 1.2117, "step": 46859 }, { "epoch": 3.6724137931034484, "grad_norm": 2.9772961139678955, "learning_rate": 1.6522203953267645e-07, "loss": 1.1203, "step": 46860 }, { "epoch": 3.6724921630094043, "grad_norm": 2.154120683670044, "learning_rate": 1.651434334045854e-07, "loss": 1.2504, "step": 46861 }, { "epoch": 3.6725705329153606, "grad_norm": 3.5565531253814697, "learning_rate": 1.6506484566580606e-07, "loss": 1.2687, "step": 46862 }, { "epoch": 3.6726489028213165, "grad_norm": 3.45896315574646, "learning_rate": 1.64986276316636e-07, "loss": 1.5489, "step": 46863 }, { "epoch": 3.672727272727273, "grad_norm": 3.2966995239257812, "learning_rate": 1.6490772535737442e-07, "loss": 1.288, "step": 46864 }, { "epoch": 3.6728056426332287, "grad_norm": 3.0909841060638428, "learning_rate": 1.648291927883211e-07, "loss": 1.4362, "step": 46865 }, { "epoch": 3.672884012539185, "grad_norm": 2.8153584003448486, "learning_rate": 1.647506786097741e-07, "loss": 1.3942, "step": 46866 }, { "epoch": 3.672962382445141, "grad_norm": 3.5077357292175293, "learning_rate": 1.6467218282203156e-07, "loss": 1.2618, "step": 46867 }, { "epoch": 3.6730407523510973, "grad_norm": 6.813819408416748, "learning_rate": 1.6459370542539266e-07, "loss": 1.1375, "step": 46868 }, { "epoch": 3.673119122257053, "grad_norm": 2.759010076522827, "learning_rate": 1.6451524642015492e-07, "loss": 1.3087, "step": 46869 }, { "epoch": 3.6731974921630095, "grad_norm": 2.8516407012939453, "learning_rate": 1.644368058066187e-07, "loss": 1.1677, "step": 46870 }, { "epoch": 3.6732758620689654, "grad_norm": 3.861074447631836, "learning_rate": 1.643583835850804e-07, "loss": 1.3752, "step": 46871 }, { "epoch": 3.6733542319749217, "grad_norm": 3.9400248527526855, "learning_rate": 1.6427997975583975e-07, "loss": 1.2244, "step": 46872 }, { "epoch": 3.6734326018808776, "grad_norm": 3.387183904647827, "learning_rate": 1.642015943191938e-07, "loss": 1.0701, "step": 46873 }, { "epoch": 3.673510971786834, "grad_norm": 3.271409273147583, "learning_rate": 1.6412322727544116e-07, "loss": 1.2289, "step": 46874 }, { "epoch": 3.67358934169279, "grad_norm": 3.0030479431152344, "learning_rate": 1.6404487862487938e-07, "loss": 1.0597, "step": 46875 }, { "epoch": 3.673667711598746, "grad_norm": 2.2528228759765625, "learning_rate": 1.6396654836780768e-07, "loss": 1.1581, "step": 46876 }, { "epoch": 3.673746081504702, "grad_norm": 9.193635940551758, "learning_rate": 1.6388823650452303e-07, "loss": 0.4455, "step": 46877 }, { "epoch": 3.6738244514106584, "grad_norm": 5.404205799102783, "learning_rate": 1.6380994303532348e-07, "loss": 1.0127, "step": 46878 }, { "epoch": 3.6739028213166143, "grad_norm": 5.279469966888428, "learning_rate": 1.637316679605072e-07, "loss": 1.659, "step": 46879 }, { "epoch": 3.6739811912225706, "grad_norm": 5.618749141693115, "learning_rate": 1.6365341128037172e-07, "loss": 1.6686, "step": 46880 }, { "epoch": 3.6740595611285265, "grad_norm": 2.4050612449645996, "learning_rate": 1.6357517299521453e-07, "loss": 1.4209, "step": 46881 }, { "epoch": 3.674137931034483, "grad_norm": 4.314533233642578, "learning_rate": 1.6349695310533264e-07, "loss": 1.2204, "step": 46882 }, { "epoch": 3.6742163009404387, "grad_norm": 2.2825872898101807, "learning_rate": 1.6341875161102417e-07, "loss": 1.3997, "step": 46883 }, { "epoch": 3.674294670846395, "grad_norm": 2.881967782974243, "learning_rate": 1.6334056851258773e-07, "loss": 1.509, "step": 46884 }, { "epoch": 3.674373040752351, "grad_norm": 4.007972240447998, "learning_rate": 1.6326240381031865e-07, "loss": 1.61, "step": 46885 }, { "epoch": 3.6744514106583073, "grad_norm": 10.57668685913086, "learning_rate": 1.6318425750451505e-07, "loss": 1.0747, "step": 46886 }, { "epoch": 3.674529780564263, "grad_norm": 2.9962878227233887, "learning_rate": 1.6310612959547334e-07, "loss": 1.0948, "step": 46887 }, { "epoch": 3.6746081504702195, "grad_norm": 2.904160976409912, "learning_rate": 1.6302802008349271e-07, "loss": 1.0724, "step": 46888 }, { "epoch": 3.6746865203761754, "grad_norm": 6.207220077514648, "learning_rate": 1.6294992896886907e-07, "loss": 1.5607, "step": 46889 }, { "epoch": 3.6747648902821317, "grad_norm": 2.907334327697754, "learning_rate": 1.6287185625189882e-07, "loss": 1.5863, "step": 46890 }, { "epoch": 3.674843260188088, "grad_norm": 9.56380558013916, "learning_rate": 1.6279380193288007e-07, "loss": 1.5947, "step": 46891 }, { "epoch": 3.674921630094044, "grad_norm": 3.1328365802764893, "learning_rate": 1.627157660121087e-07, "loss": 1.2832, "step": 46892 }, { "epoch": 3.675, "grad_norm": 6.4969282150268555, "learning_rate": 1.6263774848988224e-07, "loss": 1.3265, "step": 46893 }, { "epoch": 3.675078369905956, "grad_norm": 3.995180606842041, "learning_rate": 1.625597493664971e-07, "loss": 1.499, "step": 46894 }, { "epoch": 3.6751567398119125, "grad_norm": 4.149903774261475, "learning_rate": 1.6248176864224973e-07, "loss": 0.9939, "step": 46895 }, { "epoch": 3.6752351097178684, "grad_norm": 6.917270183563232, "learning_rate": 1.6240380631743768e-07, "loss": 1.4657, "step": 46896 }, { "epoch": 3.6753134796238243, "grad_norm": 3.0031661987304688, "learning_rate": 1.6232586239235627e-07, "loss": 0.7705, "step": 46897 }, { "epoch": 3.6753918495297806, "grad_norm": 3.669348955154419, "learning_rate": 1.6224793686730244e-07, "loss": 1.1744, "step": 46898 }, { "epoch": 3.675470219435737, "grad_norm": 16.986906051635742, "learning_rate": 1.6217002974257322e-07, "loss": 0.4804, "step": 46899 }, { "epoch": 3.675548589341693, "grad_norm": 3.845008373260498, "learning_rate": 1.6209214101846394e-07, "loss": 1.727, "step": 46900 }, { "epoch": 3.6756269592476487, "grad_norm": 3.4765090942382812, "learning_rate": 1.62014270695271e-07, "loss": 1.0196, "step": 46901 }, { "epoch": 3.675705329153605, "grad_norm": 3.2289810180664062, "learning_rate": 1.619364187732908e-07, "loss": 1.399, "step": 46902 }, { "epoch": 3.6757836990595614, "grad_norm": 5.1696271896362305, "learning_rate": 1.6185858525282038e-07, "loss": 1.9495, "step": 46903 }, { "epoch": 3.6758620689655173, "grad_norm": 4.208210468292236, "learning_rate": 1.6178077013415395e-07, "loss": 1.5669, "step": 46904 }, { "epoch": 3.675940438871473, "grad_norm": 5.671465873718262, "learning_rate": 1.61702973417589e-07, "loss": 0.8613, "step": 46905 }, { "epoch": 3.6760188087774295, "grad_norm": 4.320860385894775, "learning_rate": 1.616251951034198e-07, "loss": 1.857, "step": 46906 }, { "epoch": 3.676097178683386, "grad_norm": 5.5654425621032715, "learning_rate": 1.615474351919444e-07, "loss": 1.1886, "step": 46907 }, { "epoch": 3.6761755485893417, "grad_norm": 3.146145820617676, "learning_rate": 1.61469693683457e-07, "loss": 1.1321, "step": 46908 }, { "epoch": 3.6762539184952976, "grad_norm": 3.17814040184021, "learning_rate": 1.613919705782535e-07, "loss": 1.1243, "step": 46909 }, { "epoch": 3.676332288401254, "grad_norm": 2.4222652912139893, "learning_rate": 1.613142658766298e-07, "loss": 1.3002, "step": 46910 }, { "epoch": 3.6764106583072103, "grad_norm": 2.4790775775909424, "learning_rate": 1.6123657957888117e-07, "loss": 1.5256, "step": 46911 }, { "epoch": 3.676489028213166, "grad_norm": 4.449570655822754, "learning_rate": 1.611589116853035e-07, "loss": 1.4152, "step": 46912 }, { "epoch": 3.676567398119122, "grad_norm": 3.7914812564849854, "learning_rate": 1.6108126219619158e-07, "loss": 1.6492, "step": 46913 }, { "epoch": 3.6766457680250784, "grad_norm": 2.794820547103882, "learning_rate": 1.610036311118418e-07, "loss": 1.7899, "step": 46914 }, { "epoch": 3.6767241379310347, "grad_norm": 2.859685182571411, "learning_rate": 1.6092601843254784e-07, "loss": 1.6606, "step": 46915 }, { "epoch": 3.6768025078369906, "grad_norm": 2.4603326320648193, "learning_rate": 1.6084842415860614e-07, "loss": 1.2865, "step": 46916 }, { "epoch": 3.6768808777429465, "grad_norm": 3.7128853797912598, "learning_rate": 1.6077084829031142e-07, "loss": 1.2369, "step": 46917 }, { "epoch": 3.676959247648903, "grad_norm": 4.342759132385254, "learning_rate": 1.6069329082795958e-07, "loss": 1.1694, "step": 46918 }, { "epoch": 3.677037617554859, "grad_norm": 2.974534034729004, "learning_rate": 1.606157517718443e-07, "loss": 1.5144, "step": 46919 }, { "epoch": 3.677115987460815, "grad_norm": 9.44972038269043, "learning_rate": 1.605382311222603e-07, "loss": 1.8263, "step": 46920 }, { "epoch": 3.677194357366771, "grad_norm": 4.24967098236084, "learning_rate": 1.6046072887950402e-07, "loss": 0.9824, "step": 46921 }, { "epoch": 3.6772727272727272, "grad_norm": 3.7046706676483154, "learning_rate": 1.6038324504386972e-07, "loss": 1.3896, "step": 46922 }, { "epoch": 3.6773510971786836, "grad_norm": 5.7753448486328125, "learning_rate": 1.6030577961565097e-07, "loss": 1.1815, "step": 46923 }, { "epoch": 3.6774294670846395, "grad_norm": 3.4597558975219727, "learning_rate": 1.602283325951437e-07, "loss": 1.0414, "step": 46924 }, { "epoch": 3.6775078369905954, "grad_norm": 4.077023983001709, "learning_rate": 1.60150903982641e-07, "loss": 1.5788, "step": 46925 }, { "epoch": 3.6775862068965517, "grad_norm": 4.5853729248046875, "learning_rate": 1.6007349377843984e-07, "loss": 1.1312, "step": 46926 }, { "epoch": 3.677664576802508, "grad_norm": 4.105429172515869, "learning_rate": 1.5999610198283223e-07, "loss": 1.2326, "step": 46927 }, { "epoch": 3.677742946708464, "grad_norm": 3.956730842590332, "learning_rate": 1.5991872859611345e-07, "loss": 1.5791, "step": 46928 }, { "epoch": 3.67782131661442, "grad_norm": 3.0693306922912598, "learning_rate": 1.5984137361857833e-07, "loss": 1.4147, "step": 46929 }, { "epoch": 3.677899686520376, "grad_norm": 7.702747821807861, "learning_rate": 1.597640370505199e-07, "loss": 1.3208, "step": 46930 }, { "epoch": 3.6779780564263325, "grad_norm": 3.904555082321167, "learning_rate": 1.5968671889223354e-07, "loss": 1.5589, "step": 46931 }, { "epoch": 3.6780564263322884, "grad_norm": 4.226717472076416, "learning_rate": 1.596094191440123e-07, "loss": 1.1545, "step": 46932 }, { "epoch": 3.6781347962382442, "grad_norm": 3.1203131675720215, "learning_rate": 1.595321378061515e-07, "loss": 1.3429, "step": 46933 }, { "epoch": 3.6782131661442006, "grad_norm": 4.0864577293396, "learning_rate": 1.5945487487894318e-07, "loss": 1.3011, "step": 46934 }, { "epoch": 3.678291536050157, "grad_norm": 3.049635648727417, "learning_rate": 1.5937763036268261e-07, "loss": 1.1176, "step": 46935 }, { "epoch": 3.678369905956113, "grad_norm": 9.526469230651855, "learning_rate": 1.5930040425766347e-07, "loss": 1.1423, "step": 46936 }, { "epoch": 3.678448275862069, "grad_norm": 5.70151424407959, "learning_rate": 1.592231965641794e-07, "loss": 1.5192, "step": 46937 }, { "epoch": 3.678526645768025, "grad_norm": 4.097010135650635, "learning_rate": 1.5914600728252404e-07, "loss": 1.7833, "step": 46938 }, { "epoch": 3.6786050156739813, "grad_norm": 3.804774045944214, "learning_rate": 1.5906883641298944e-07, "loss": 1.4038, "step": 46939 }, { "epoch": 3.6786833855799372, "grad_norm": 2.952707529067993, "learning_rate": 1.589916839558725e-07, "loss": 1.215, "step": 46940 }, { "epoch": 3.6787617554858936, "grad_norm": 2.4935708045959473, "learning_rate": 1.589145499114636e-07, "loss": 1.3699, "step": 46941 }, { "epoch": 3.6788401253918495, "grad_norm": 4.696075916290283, "learning_rate": 1.5883743428005693e-07, "loss": 1.6191, "step": 46942 }, { "epoch": 3.678918495297806, "grad_norm": 4.468361854553223, "learning_rate": 1.587603370619467e-07, "loss": 1.4581, "step": 46943 }, { "epoch": 3.6789968652037617, "grad_norm": 2.5320093631744385, "learning_rate": 1.5868325825742546e-07, "loss": 1.3346, "step": 46944 }, { "epoch": 3.679075235109718, "grad_norm": 6.125237941741943, "learning_rate": 1.5860619786678632e-07, "loss": 1.4375, "step": 46945 }, { "epoch": 3.679153605015674, "grad_norm": 2.6386921405792236, "learning_rate": 1.5852915589032235e-07, "loss": 1.6169, "step": 46946 }, { "epoch": 3.6792319749216302, "grad_norm": 2.922819137573242, "learning_rate": 1.5845213232832667e-07, "loss": 1.4432, "step": 46947 }, { "epoch": 3.679310344827586, "grad_norm": 2.2687015533447266, "learning_rate": 1.5837512718109239e-07, "loss": 1.439, "step": 46948 }, { "epoch": 3.6793887147335425, "grad_norm": 5.676629543304443, "learning_rate": 1.5829814044891257e-07, "loss": 1.8534, "step": 46949 }, { "epoch": 3.6794670846394983, "grad_norm": 3.3062994480133057, "learning_rate": 1.5822117213207977e-07, "loss": 1.6009, "step": 46950 }, { "epoch": 3.6795454545454547, "grad_norm": 3.29598331451416, "learning_rate": 1.5814422223088654e-07, "loss": 1.2202, "step": 46951 }, { "epoch": 3.6796238244514106, "grad_norm": 2.7616214752197266, "learning_rate": 1.5806729074562598e-07, "loss": 1.3531, "step": 46952 }, { "epoch": 3.679702194357367, "grad_norm": 3.5186331272125244, "learning_rate": 1.5799037767659008e-07, "loss": 1.5092, "step": 46953 }, { "epoch": 3.679780564263323, "grad_norm": 5.592020511627197, "learning_rate": 1.5791348302407193e-07, "loss": 2.4955, "step": 46954 }, { "epoch": 3.679858934169279, "grad_norm": 2.8958775997161865, "learning_rate": 1.578366067883641e-07, "loss": 1.3649, "step": 46955 }, { "epoch": 3.679937304075235, "grad_norm": 7.384547233581543, "learning_rate": 1.5775974896975854e-07, "loss": 1.1014, "step": 46956 }, { "epoch": 3.6800156739811913, "grad_norm": 3.9543347358703613, "learning_rate": 1.5768290956854782e-07, "loss": 1.5047, "step": 46957 }, { "epoch": 3.6800940438871472, "grad_norm": 2.95235013961792, "learning_rate": 1.5760608858502336e-07, "loss": 1.3636, "step": 46958 }, { "epoch": 3.6801724137931036, "grad_norm": 6.113468647003174, "learning_rate": 1.5752928601947936e-07, "loss": 1.3986, "step": 46959 }, { "epoch": 3.6802507836990594, "grad_norm": 2.1936709880828857, "learning_rate": 1.5745250187220617e-07, "loss": 1.243, "step": 46960 }, { "epoch": 3.680329153605016, "grad_norm": 4.423846244812012, "learning_rate": 1.5737573614349632e-07, "loss": 1.6902, "step": 46961 }, { "epoch": 3.6804075235109717, "grad_norm": 5.077454090118408, "learning_rate": 1.572989888336418e-07, "loss": 1.8959, "step": 46962 }, { "epoch": 3.680485893416928, "grad_norm": 7.498741626739502, "learning_rate": 1.5722225994293462e-07, "loss": 1.2975, "step": 46963 }, { "epoch": 3.680564263322884, "grad_norm": 4.107464790344238, "learning_rate": 1.5714554947166672e-07, "loss": 1.6172, "step": 46964 }, { "epoch": 3.68064263322884, "grad_norm": 2.403515338897705, "learning_rate": 1.5706885742012956e-07, "loss": 1.3288, "step": 46965 }, { "epoch": 3.680721003134796, "grad_norm": 2.443784475326538, "learning_rate": 1.5699218378861513e-07, "loss": 1.1637, "step": 46966 }, { "epoch": 3.6807993730407524, "grad_norm": 2.333176374435425, "learning_rate": 1.5691552857741488e-07, "loss": 1.4645, "step": 46967 }, { "epoch": 3.6808777429467083, "grad_norm": 3.4155662059783936, "learning_rate": 1.568388917868202e-07, "loss": 1.0722, "step": 46968 }, { "epoch": 3.6809561128526647, "grad_norm": 3.134538173675537, "learning_rate": 1.567622734171226e-07, "loss": 0.7053, "step": 46969 }, { "epoch": 3.6810344827586206, "grad_norm": 2.985971450805664, "learning_rate": 1.5668567346861453e-07, "loss": 1.52, "step": 46970 }, { "epoch": 3.681112852664577, "grad_norm": 4.671530723571777, "learning_rate": 1.5660909194158523e-07, "loss": 1.3994, "step": 46971 }, { "epoch": 3.6811912225705328, "grad_norm": 3.597801685333252, "learning_rate": 1.5653252883632785e-07, "loss": 1.2501, "step": 46972 }, { "epoch": 3.681269592476489, "grad_norm": 7.995782852172852, "learning_rate": 1.5645598415313324e-07, "loss": 1.6662, "step": 46973 }, { "epoch": 3.681347962382445, "grad_norm": 2.756584882736206, "learning_rate": 1.5637945789229226e-07, "loss": 1.4634, "step": 46974 }, { "epoch": 3.6814263322884013, "grad_norm": 2.879753828048706, "learning_rate": 1.5630295005409578e-07, "loss": 1.493, "step": 46975 }, { "epoch": 3.681504702194357, "grad_norm": 3.1544830799102783, "learning_rate": 1.5622646063883417e-07, "loss": 1.2186, "step": 46976 }, { "epoch": 3.6815830721003135, "grad_norm": 2.895702362060547, "learning_rate": 1.5614998964679996e-07, "loss": 1.1102, "step": 46977 }, { "epoch": 3.6816614420062694, "grad_norm": 2.7669429779052734, "learning_rate": 1.56073537078284e-07, "loss": 1.2246, "step": 46978 }, { "epoch": 3.6817398119122258, "grad_norm": 3.658752679824829, "learning_rate": 1.559971029335755e-07, "loss": 1.2171, "step": 46979 }, { "epoch": 3.6818181818181817, "grad_norm": 3.8115787506103516, "learning_rate": 1.5592068721296593e-07, "loss": 1.3998, "step": 46980 }, { "epoch": 3.681896551724138, "grad_norm": 12.2485933303833, "learning_rate": 1.5584428991674616e-07, "loss": 1.1462, "step": 46981 }, { "epoch": 3.681974921630094, "grad_norm": 3.102046489715576, "learning_rate": 1.557679110452065e-07, "loss": 1.3647, "step": 46982 }, { "epoch": 3.68205329153605, "grad_norm": 4.616186141967773, "learning_rate": 1.5569155059863728e-07, "loss": 1.4901, "step": 46983 }, { "epoch": 3.682131661442006, "grad_norm": 2.6911909580230713, "learning_rate": 1.5561520857732937e-07, "loss": 1.3257, "step": 46984 }, { "epoch": 3.6822100313479624, "grad_norm": 4.59246301651001, "learning_rate": 1.555388849815731e-07, "loss": 1.3206, "step": 46985 }, { "epoch": 3.6822884012539188, "grad_norm": 3.3247480392456055, "learning_rate": 1.554625798116588e-07, "loss": 1.3089, "step": 46986 }, { "epoch": 3.6823667711598747, "grad_norm": 3.8643274307250977, "learning_rate": 1.5538629306787623e-07, "loss": 1.3245, "step": 46987 }, { "epoch": 3.6824451410658305, "grad_norm": 11.778555870056152, "learning_rate": 1.5531002475051626e-07, "loss": 0.912, "step": 46988 }, { "epoch": 3.682523510971787, "grad_norm": 2.165783166885376, "learning_rate": 1.5523377485986867e-07, "loss": 1.1852, "step": 46989 }, { "epoch": 3.682601880877743, "grad_norm": 2.0718510150909424, "learning_rate": 1.5515754339622214e-07, "loss": 1.1746, "step": 46990 }, { "epoch": 3.682680250783699, "grad_norm": 2.4310057163238525, "learning_rate": 1.550813303598686e-07, "loss": 1.3815, "step": 46991 }, { "epoch": 3.682758620689655, "grad_norm": 3.8567962646484375, "learning_rate": 1.5500513575109732e-07, "loss": 1.4619, "step": 46992 }, { "epoch": 3.6828369905956113, "grad_norm": 4.657164573669434, "learning_rate": 1.549289595701986e-07, "loss": 1.4292, "step": 46993 }, { "epoch": 3.6829153605015676, "grad_norm": 3.5439493656158447, "learning_rate": 1.548528018174611e-07, "loss": 1.3507, "step": 46994 }, { "epoch": 3.6829937304075235, "grad_norm": 3.30159068107605, "learning_rate": 1.5477666249317403e-07, "loss": 1.6895, "step": 46995 }, { "epoch": 3.6830721003134794, "grad_norm": 5.2770843505859375, "learning_rate": 1.5470054159762881e-07, "loss": 1.564, "step": 46996 }, { "epoch": 3.6831504702194358, "grad_norm": 2.9863052368164062, "learning_rate": 1.5462443913111358e-07, "loss": 1.4077, "step": 46997 }, { "epoch": 3.683228840125392, "grad_norm": 3.9005942344665527, "learning_rate": 1.5454835509391862e-07, "loss": 1.4196, "step": 46998 }, { "epoch": 3.683307210031348, "grad_norm": 4.327674388885498, "learning_rate": 1.5447228948633265e-07, "loss": 1.2678, "step": 46999 }, { "epoch": 3.683385579937304, "grad_norm": 3.9569482803344727, "learning_rate": 1.5439624230864537e-07, "loss": 1.1995, "step": 47000 }, { "epoch": 3.68346394984326, "grad_norm": 3.74060320854187, "learning_rate": 1.5432021356114546e-07, "loss": 1.2712, "step": 47001 }, { "epoch": 3.6835423197492165, "grad_norm": 4.476016044616699, "learning_rate": 1.5424420324412327e-07, "loss": 1.8407, "step": 47002 }, { "epoch": 3.6836206896551724, "grad_norm": 4.163771152496338, "learning_rate": 1.5416821135786685e-07, "loss": 1.4399, "step": 47003 }, { "epoch": 3.6836990595611283, "grad_norm": 3.3266403675079346, "learning_rate": 1.5409223790266548e-07, "loss": 1.3307, "step": 47004 }, { "epoch": 3.6837774294670846, "grad_norm": 3.7137513160705566, "learning_rate": 1.5401628287880831e-07, "loss": 1.3837, "step": 47005 }, { "epoch": 3.683855799373041, "grad_norm": 8.615673065185547, "learning_rate": 1.5394034628658404e-07, "loss": 1.2597, "step": 47006 }, { "epoch": 3.683934169278997, "grad_norm": 3.1652278900146484, "learning_rate": 1.538644281262819e-07, "loss": 1.5162, "step": 47007 }, { "epoch": 3.6840125391849528, "grad_norm": 3.1699187755584717, "learning_rate": 1.5378852839819102e-07, "loss": 1.4846, "step": 47008 }, { "epoch": 3.684090909090909, "grad_norm": 3.9267938137054443, "learning_rate": 1.5371264710259794e-07, "loss": 1.4932, "step": 47009 }, { "epoch": 3.6841692789968654, "grad_norm": 1.9913573265075684, "learning_rate": 1.5363678423979344e-07, "loss": 1.0116, "step": 47010 }, { "epoch": 3.6842476489028213, "grad_norm": 3.8067150115966797, "learning_rate": 1.535609398100657e-07, "loss": 1.1614, "step": 47011 }, { "epoch": 3.684326018808777, "grad_norm": 3.3714213371276855, "learning_rate": 1.5348511381370224e-07, "loss": 1.5394, "step": 47012 }, { "epoch": 3.6844043887147335, "grad_norm": 6.102453231811523, "learning_rate": 1.5340930625099227e-07, "loss": 1.483, "step": 47013 }, { "epoch": 3.68448275862069, "grad_norm": 3.4539310932159424, "learning_rate": 1.5333351712222332e-07, "loss": 1.4433, "step": 47014 }, { "epoch": 3.6845611285266457, "grad_norm": 5.011256694793701, "learning_rate": 1.5325774642768577e-07, "loss": 1.029, "step": 47015 }, { "epoch": 3.6846394984326016, "grad_norm": 3.1922545433044434, "learning_rate": 1.5318199416766544e-07, "loss": 1.3341, "step": 47016 }, { "epoch": 3.684717868338558, "grad_norm": 2.195723295211792, "learning_rate": 1.53106260342451e-07, "loss": 1.355, "step": 47017 }, { "epoch": 3.6847962382445143, "grad_norm": 2.6228528022766113, "learning_rate": 1.5303054495233116e-07, "loss": 1.1929, "step": 47018 }, { "epoch": 3.68487460815047, "grad_norm": 2.778244972229004, "learning_rate": 1.529548479975934e-07, "loss": 1.4217, "step": 47019 }, { "epoch": 3.684952978056426, "grad_norm": 3.471552610397339, "learning_rate": 1.5287916947852643e-07, "loss": 1.5389, "step": 47020 }, { "epoch": 3.6850313479623824, "grad_norm": 3.7038064002990723, "learning_rate": 1.5280350939541665e-07, "loss": 1.3949, "step": 47021 }, { "epoch": 3.6851097178683387, "grad_norm": 4.086723327636719, "learning_rate": 1.527278677485533e-07, "loss": 1.4925, "step": 47022 }, { "epoch": 3.6851880877742946, "grad_norm": 4.166996479034424, "learning_rate": 1.526522445382228e-07, "loss": 1.5778, "step": 47023 }, { "epoch": 3.6852664576802505, "grad_norm": 8.771808624267578, "learning_rate": 1.5257663976471383e-07, "loss": 0.7935, "step": 47024 }, { "epoch": 3.685344827586207, "grad_norm": 3.207202434539795, "learning_rate": 1.5250105342831333e-07, "loss": 1.1711, "step": 47025 }, { "epoch": 3.685423197492163, "grad_norm": 2.4088287353515625, "learning_rate": 1.5242548552930946e-07, "loss": 1.3585, "step": 47026 }, { "epoch": 3.685501567398119, "grad_norm": 5.650463104248047, "learning_rate": 1.5234993606798865e-07, "loss": 2.0425, "step": 47027 }, { "epoch": 3.685579937304075, "grad_norm": 3.970829725265503, "learning_rate": 1.5227440504463842e-07, "loss": 0.9448, "step": 47028 }, { "epoch": 3.6856583072100313, "grad_norm": 6.27440071105957, "learning_rate": 1.5219889245954688e-07, "loss": 1.2076, "step": 47029 }, { "epoch": 3.6857366771159876, "grad_norm": 4.25735330581665, "learning_rate": 1.521233983130016e-07, "loss": 1.6094, "step": 47030 }, { "epoch": 3.6858150470219435, "grad_norm": 5.8135085105896, "learning_rate": 1.5204792260528789e-07, "loss": 1.8882, "step": 47031 }, { "epoch": 3.6858934169279, "grad_norm": 2.948744058609009, "learning_rate": 1.5197246533669385e-07, "loss": 1.4741, "step": 47032 }, { "epoch": 3.6859717868338557, "grad_norm": 3.7199599742889404, "learning_rate": 1.5189702650750592e-07, "loss": 1.4687, "step": 47033 }, { "epoch": 3.686050156739812, "grad_norm": 3.182485580444336, "learning_rate": 1.5182160611801278e-07, "loss": 1.0045, "step": 47034 }, { "epoch": 3.686128526645768, "grad_norm": 3.702254295349121, "learning_rate": 1.5174620416849972e-07, "loss": 1.5358, "step": 47035 }, { "epoch": 3.6862068965517243, "grad_norm": 3.6561760902404785, "learning_rate": 1.5167082065925375e-07, "loss": 1.5343, "step": 47036 }, { "epoch": 3.68628526645768, "grad_norm": 5.277612209320068, "learning_rate": 1.5159545559056133e-07, "loss": 1.6331, "step": 47037 }, { "epoch": 3.6863636363636365, "grad_norm": 3.040733575820923, "learning_rate": 1.5152010896270998e-07, "loss": 1.2691, "step": 47038 }, { "epoch": 3.6864420062695924, "grad_norm": 7.820143699645996, "learning_rate": 1.5144478077598556e-07, "loss": 1.4116, "step": 47039 }, { "epoch": 3.6865203761755487, "grad_norm": 2.8812155723571777, "learning_rate": 1.513694710306751e-07, "loss": 1.2447, "step": 47040 }, { "epoch": 3.6865987460815046, "grad_norm": 3.0667266845703125, "learning_rate": 1.5129417972706507e-07, "loss": 1.1354, "step": 47041 }, { "epoch": 3.686677115987461, "grad_norm": 2.679983615875244, "learning_rate": 1.5121890686544017e-07, "loss": 1.466, "step": 47042 }, { "epoch": 3.686755485893417, "grad_norm": 2.4986910820007324, "learning_rate": 1.5114365244608908e-07, "loss": 1.1352, "step": 47043 }, { "epoch": 3.686833855799373, "grad_norm": 2.8819191455841064, "learning_rate": 1.510684164692966e-07, "loss": 1.374, "step": 47044 }, { "epoch": 3.686912225705329, "grad_norm": 2.1382832527160645, "learning_rate": 1.5099319893535025e-07, "loss": 1.1994, "step": 47045 }, { "epoch": 3.6869905956112854, "grad_norm": 3.6951022148132324, "learning_rate": 1.5091799984453425e-07, "loss": 1.3272, "step": 47046 }, { "epoch": 3.6870689655172413, "grad_norm": 3.415043354034424, "learning_rate": 1.5084281919713506e-07, "loss": 1.2295, "step": 47047 }, { "epoch": 3.6871473354231976, "grad_norm": 2.784313678741455, "learning_rate": 1.5076765699343964e-07, "loss": 1.0364, "step": 47048 }, { "epoch": 3.6872257053291535, "grad_norm": 8.610312461853027, "learning_rate": 1.5069251323373392e-07, "loss": 1.3491, "step": 47049 }, { "epoch": 3.68730407523511, "grad_norm": 4.982176303863525, "learning_rate": 1.506173879183026e-07, "loss": 1.4275, "step": 47050 }, { "epoch": 3.6873824451410657, "grad_norm": 2.393005609512329, "learning_rate": 1.505422810474322e-07, "loss": 1.2659, "step": 47051 }, { "epoch": 3.687460815047022, "grad_norm": 3.739342451095581, "learning_rate": 1.5046719262140796e-07, "loss": 1.1952, "step": 47052 }, { "epoch": 3.687539184952978, "grad_norm": 3.945986270904541, "learning_rate": 1.5039212264051528e-07, "loss": 1.5402, "step": 47053 }, { "epoch": 3.6876175548589343, "grad_norm": 3.1782031059265137, "learning_rate": 1.5031707110504056e-07, "loss": 1.6545, "step": 47054 }, { "epoch": 3.68769592476489, "grad_norm": 5.00339412689209, "learning_rate": 1.502420380152686e-07, "loss": 1.1149, "step": 47055 }, { "epoch": 3.6877742946708465, "grad_norm": 2.932526111602783, "learning_rate": 1.501670233714847e-07, "loss": 1.3955, "step": 47056 }, { "epoch": 3.6878526645768024, "grad_norm": 4.0795063972473145, "learning_rate": 1.5009202717397475e-07, "loss": 1.6054, "step": 47057 }, { "epoch": 3.6879310344827587, "grad_norm": 2.859586477279663, "learning_rate": 1.5001704942302353e-07, "loss": 1.5367, "step": 47058 }, { "epoch": 3.6880094043887146, "grad_norm": 4.094902992248535, "learning_rate": 1.4994209011891637e-07, "loss": 1.0522, "step": 47059 }, { "epoch": 3.688087774294671, "grad_norm": 4.191391468048096, "learning_rate": 1.4986714926193802e-07, "loss": 1.6609, "step": 47060 }, { "epoch": 3.688166144200627, "grad_norm": 2.6281702518463135, "learning_rate": 1.4979222685237383e-07, "loss": 1.3517, "step": 47061 }, { "epoch": 3.688244514106583, "grad_norm": 3.5740580558776855, "learning_rate": 1.497173228905091e-07, "loss": 1.2587, "step": 47062 }, { "epoch": 3.688322884012539, "grad_norm": 2.541966676712036, "learning_rate": 1.4964243737662866e-07, "loss": 1.0553, "step": 47063 }, { "epoch": 3.6884012539184954, "grad_norm": 2.818373203277588, "learning_rate": 1.4956757031101722e-07, "loss": 1.4029, "step": 47064 }, { "epoch": 3.6884796238244513, "grad_norm": 2.861454725265503, "learning_rate": 1.49492721693959e-07, "loss": 1.4317, "step": 47065 }, { "epoch": 3.6885579937304076, "grad_norm": 2.2436439990997314, "learning_rate": 1.4941789152573826e-07, "loss": 0.9517, "step": 47066 }, { "epoch": 3.6886363636363635, "grad_norm": 3.231797933578491, "learning_rate": 1.4934307980664198e-07, "loss": 1.3922, "step": 47067 }, { "epoch": 3.68871473354232, "grad_norm": 3.0223231315612793, "learning_rate": 1.4926828653695268e-07, "loss": 1.2482, "step": 47068 }, { "epoch": 3.6887931034482757, "grad_norm": 3.965533494949341, "learning_rate": 1.4919351171695463e-07, "loss": 1.1619, "step": 47069 }, { "epoch": 3.688871473354232, "grad_norm": 3.6348254680633545, "learning_rate": 1.4911875534693364e-07, "loss": 0.9157, "step": 47070 }, { "epoch": 3.688949843260188, "grad_norm": 2.9205470085144043, "learning_rate": 1.490440174271729e-07, "loss": 1.0548, "step": 47071 }, { "epoch": 3.6890282131661443, "grad_norm": 3.4508516788482666, "learning_rate": 1.4896929795795766e-07, "loss": 1.3361, "step": 47072 }, { "epoch": 3.6891065830721, "grad_norm": 4.092142105102539, "learning_rate": 1.4889459693957109e-07, "loss": 1.1939, "step": 47073 }, { "epoch": 3.6891849529780565, "grad_norm": 4.73255729675293, "learning_rate": 1.4881991437229794e-07, "loss": 1.3421, "step": 47074 }, { "epoch": 3.6892633228840124, "grad_norm": 4.916449069976807, "learning_rate": 1.4874525025642184e-07, "loss": 1.3419, "step": 47075 }, { "epoch": 3.6893416927899687, "grad_norm": 2.824249744415283, "learning_rate": 1.4867060459222703e-07, "loss": 1.28, "step": 47076 }, { "epoch": 3.689420062695925, "grad_norm": 3.2840147018432617, "learning_rate": 1.4859597737999776e-07, "loss": 1.3694, "step": 47077 }, { "epoch": 3.689498432601881, "grad_norm": 5.067223072052002, "learning_rate": 1.4852136862001766e-07, "loss": 1.2133, "step": 47078 }, { "epoch": 3.689576802507837, "grad_norm": 3.2496347427368164, "learning_rate": 1.4844677831257037e-07, "loss": 1.2271, "step": 47079 }, { "epoch": 3.689655172413793, "grad_norm": 4.3992791175842285, "learning_rate": 1.4837220645793905e-07, "loss": 1.5918, "step": 47080 }, { "epoch": 3.6897335423197495, "grad_norm": 3.45609974861145, "learning_rate": 1.4829765305640843e-07, "loss": 1.3586, "step": 47081 }, { "epoch": 3.6898119122257054, "grad_norm": 4.256232261657715, "learning_rate": 1.482231181082622e-07, "loss": 1.2529, "step": 47082 }, { "epoch": 3.6898902821316613, "grad_norm": 21.611560821533203, "learning_rate": 1.481486016137823e-07, "loss": 1.0311, "step": 47083 }, { "epoch": 3.6899686520376176, "grad_norm": 9.311278343200684, "learning_rate": 1.4807410357325248e-07, "loss": 1.684, "step": 47084 }, { "epoch": 3.690047021943574, "grad_norm": 3.0300989151000977, "learning_rate": 1.4799962398695745e-07, "loss": 1.3011, "step": 47085 }, { "epoch": 3.69012539184953, "grad_norm": 4.592874050140381, "learning_rate": 1.4792516285518032e-07, "loss": 1.7701, "step": 47086 }, { "epoch": 3.6902037617554857, "grad_norm": 2.4413952827453613, "learning_rate": 1.4785072017820311e-07, "loss": 1.372, "step": 47087 }, { "epoch": 3.690282131661442, "grad_norm": 3.60032320022583, "learning_rate": 1.4777629595630893e-07, "loss": 1.4033, "step": 47088 }, { "epoch": 3.6903605015673984, "grad_norm": 4.150279998779297, "learning_rate": 1.4770189018978142e-07, "loss": 0.9979, "step": 47089 }, { "epoch": 3.6904388714733543, "grad_norm": 3.2004001140594482, "learning_rate": 1.476275028789048e-07, "loss": 1.3008, "step": 47090 }, { "epoch": 3.69051724137931, "grad_norm": 5.596563339233398, "learning_rate": 1.4755313402396053e-07, "loss": 1.1203, "step": 47091 }, { "epoch": 3.6905956112852665, "grad_norm": 3.429328680038452, "learning_rate": 1.4747878362523115e-07, "loss": 1.5348, "step": 47092 }, { "epoch": 3.690673981191223, "grad_norm": 4.822610855102539, "learning_rate": 1.4740445168300031e-07, "loss": 1.1141, "step": 47093 }, { "epoch": 3.6907523510971787, "grad_norm": 5.664069175720215, "learning_rate": 1.473301381975506e-07, "loss": 1.1792, "step": 47094 }, { "epoch": 3.6908307210031346, "grad_norm": 3.522427797317505, "learning_rate": 1.4725584316916451e-07, "loss": 1.4572, "step": 47095 }, { "epoch": 3.690909090909091, "grad_norm": 2.5149853229522705, "learning_rate": 1.4718156659812465e-07, "loss": 1.3989, "step": 47096 }, { "epoch": 3.6909874608150472, "grad_norm": 4.171532154083252, "learning_rate": 1.4710730848471355e-07, "loss": 1.2567, "step": 47097 }, { "epoch": 3.691065830721003, "grad_norm": 3.8957502841949463, "learning_rate": 1.470330688292132e-07, "loss": 1.5893, "step": 47098 }, { "epoch": 3.691144200626959, "grad_norm": 6.44256591796875, "learning_rate": 1.469588476319067e-07, "loss": 1.705, "step": 47099 }, { "epoch": 3.6912225705329154, "grad_norm": 3.8531181812286377, "learning_rate": 1.4688464489307664e-07, "loss": 0.759, "step": 47100 }, { "epoch": 3.6913009404388717, "grad_norm": 4.508036136627197, "learning_rate": 1.4681046061300442e-07, "loss": 1.1664, "step": 47101 }, { "epoch": 3.6913793103448276, "grad_norm": 3.5823814868927, "learning_rate": 1.4673629479197205e-07, "loss": 1.4551, "step": 47102 }, { "epoch": 3.6914576802507835, "grad_norm": 3.3233540058135986, "learning_rate": 1.4666214743026208e-07, "loss": 1.4317, "step": 47103 }, { "epoch": 3.69153605015674, "grad_norm": 4.1410369873046875, "learning_rate": 1.4658801852815652e-07, "loss": 1.3138, "step": 47104 }, { "epoch": 3.691614420062696, "grad_norm": 2.6202619075775146, "learning_rate": 1.465139080859379e-07, "loss": 1.4232, "step": 47105 }, { "epoch": 3.691692789968652, "grad_norm": 9.074078559875488, "learning_rate": 1.4643981610388658e-07, "loss": 1.5786, "step": 47106 }, { "epoch": 3.691771159874608, "grad_norm": 4.771933078765869, "learning_rate": 1.4636574258228565e-07, "loss": 1.4509, "step": 47107 }, { "epoch": 3.6918495297805642, "grad_norm": 8.627083778381348, "learning_rate": 1.4629168752141654e-07, "loss": 1.5727, "step": 47108 }, { "epoch": 3.6919278996865206, "grad_norm": 3.3979926109313965, "learning_rate": 1.462176509215607e-07, "loss": 1.4449, "step": 47109 }, { "epoch": 3.6920062695924765, "grad_norm": 2.8291900157928467, "learning_rate": 1.461436327829996e-07, "loss": 1.492, "step": 47110 }, { "epoch": 3.6920846394984324, "grad_norm": 2.9441590309143066, "learning_rate": 1.4606963310601518e-07, "loss": 1.408, "step": 47111 }, { "epoch": 3.6921630094043887, "grad_norm": 2.6805951595306396, "learning_rate": 1.4599565189088894e-07, "loss": 1.3609, "step": 47112 }, { "epoch": 3.692241379310345, "grad_norm": 3.5838046073913574, "learning_rate": 1.4592168913790227e-07, "loss": 1.3323, "step": 47113 }, { "epoch": 3.692319749216301, "grad_norm": 2.4575343132019043, "learning_rate": 1.4584774484733554e-07, "loss": 1.5163, "step": 47114 }, { "epoch": 3.692398119122257, "grad_norm": 3.045652389526367, "learning_rate": 1.457738190194713e-07, "loss": 1.4191, "step": 47115 }, { "epoch": 3.692476489028213, "grad_norm": 9.139993667602539, "learning_rate": 1.456999116545904e-07, "loss": 1.7096, "step": 47116 }, { "epoch": 3.6925548589341695, "grad_norm": 5.017668724060059, "learning_rate": 1.456260227529732e-07, "loss": 1.1162, "step": 47117 }, { "epoch": 3.6926332288401253, "grad_norm": 2.823939323425293, "learning_rate": 1.4555215231490117e-07, "loss": 1.3673, "step": 47118 }, { "epoch": 3.6927115987460812, "grad_norm": 4.785548686981201, "learning_rate": 1.454783003406557e-07, "loss": 1.3188, "step": 47119 }, { "epoch": 3.6927899686520376, "grad_norm": 4.139585494995117, "learning_rate": 1.4540446683051768e-07, "loss": 1.4586, "step": 47120 }, { "epoch": 3.692868338557994, "grad_norm": 3.9533910751342773, "learning_rate": 1.4533065178476692e-07, "loss": 1.4215, "step": 47121 }, { "epoch": 3.69294670846395, "grad_norm": 2.4857418537139893, "learning_rate": 1.4525685520368426e-07, "loss": 1.4567, "step": 47122 }, { "epoch": 3.693025078369906, "grad_norm": 2.8223464488983154, "learning_rate": 1.4518307708755232e-07, "loss": 1.4477, "step": 47123 }, { "epoch": 3.693103448275862, "grad_norm": 2.3357930183410645, "learning_rate": 1.4510931743665024e-07, "loss": 1.1538, "step": 47124 }, { "epoch": 3.6931818181818183, "grad_norm": 3.9680418968200684, "learning_rate": 1.4503557625125787e-07, "loss": 1.2129, "step": 47125 }, { "epoch": 3.6932601880877742, "grad_norm": 3.3810408115386963, "learning_rate": 1.4496185353165714e-07, "loss": 1.4097, "step": 47126 }, { "epoch": 3.6933385579937306, "grad_norm": 2.2380285263061523, "learning_rate": 1.4488814927812732e-07, "loss": 1.4539, "step": 47127 }, { "epoch": 3.6934169278996865, "grad_norm": 3.835265874862671, "learning_rate": 1.448144634909493e-07, "loss": 1.1426, "step": 47128 }, { "epoch": 3.693495297805643, "grad_norm": 3.1115448474884033, "learning_rate": 1.4474079617040338e-07, "loss": 1.4363, "step": 47129 }, { "epoch": 3.6935736677115987, "grad_norm": 6.255398750305176, "learning_rate": 1.446671473167699e-07, "loss": 1.0123, "step": 47130 }, { "epoch": 3.693652037617555, "grad_norm": 4.626480579376221, "learning_rate": 1.4459351693032863e-07, "loss": 1.3096, "step": 47131 }, { "epoch": 3.693730407523511, "grad_norm": 5.9969706535339355, "learning_rate": 1.4451990501135937e-07, "loss": 1.6215, "step": 47132 }, { "epoch": 3.6938087774294672, "grad_norm": 3.3207993507385254, "learning_rate": 1.44446311560143e-07, "loss": 1.4423, "step": 47133 }, { "epoch": 3.693887147335423, "grad_norm": 4.2354817390441895, "learning_rate": 1.4437273657695817e-07, "loss": 1.3606, "step": 47134 }, { "epoch": 3.6939655172413794, "grad_norm": 3.384251356124878, "learning_rate": 1.4429918006208688e-07, "loss": 1.4319, "step": 47135 }, { "epoch": 3.6940438871473353, "grad_norm": 3.638338804244995, "learning_rate": 1.442256420158056e-07, "loss": 1.2079, "step": 47136 }, { "epoch": 3.6941222570532917, "grad_norm": 4.885060787200928, "learning_rate": 1.4415212243839683e-07, "loss": 1.5863, "step": 47137 }, { "epoch": 3.6942006269592476, "grad_norm": 3.1715924739837646, "learning_rate": 1.4407862133013983e-07, "loss": 1.2932, "step": 47138 }, { "epoch": 3.694278996865204, "grad_norm": 13.414239883422852, "learning_rate": 1.440051386913127e-07, "loss": 0.7839, "step": 47139 }, { "epoch": 3.69435736677116, "grad_norm": 3.316194772720337, "learning_rate": 1.4393167452219637e-07, "loss": 1.3682, "step": 47140 }, { "epoch": 3.694435736677116, "grad_norm": 3.1989855766296387, "learning_rate": 1.4385822882306887e-07, "loss": 1.2641, "step": 47141 }, { "epoch": 3.694514106583072, "grad_norm": 2.9537529945373535, "learning_rate": 1.4378480159421114e-07, "loss": 1.2353, "step": 47142 }, { "epoch": 3.6945924764890283, "grad_norm": 3.6659581661224365, "learning_rate": 1.4371139283590185e-07, "loss": 1.2421, "step": 47143 }, { "epoch": 3.694670846394984, "grad_norm": 2.647639274597168, "learning_rate": 1.4363800254841965e-07, "loss": 1.1977, "step": 47144 }, { "epoch": 3.6947492163009406, "grad_norm": 3.9110777378082275, "learning_rate": 1.435646307320443e-07, "loss": 1.2802, "step": 47145 }, { "epoch": 3.6948275862068964, "grad_norm": 4.19038724899292, "learning_rate": 1.4349127738705448e-07, "loss": 1.4893, "step": 47146 }, { "epoch": 3.6949059561128528, "grad_norm": 4.168659210205078, "learning_rate": 1.4341794251372942e-07, "loss": 1.2998, "step": 47147 }, { "epoch": 3.6949843260188087, "grad_norm": 4.086639404296875, "learning_rate": 1.4334462611234834e-07, "loss": 1.4298, "step": 47148 }, { "epoch": 3.695062695924765, "grad_norm": 2.5689761638641357, "learning_rate": 1.432713281831899e-07, "loss": 1.2612, "step": 47149 }, { "epoch": 3.695141065830721, "grad_norm": 5.220306873321533, "learning_rate": 1.4319804872653166e-07, "loss": 1.441, "step": 47150 }, { "epoch": 3.695219435736677, "grad_norm": 3.62900447845459, "learning_rate": 1.4312478774265448e-07, "loss": 1.2083, "step": 47151 }, { "epoch": 3.695297805642633, "grad_norm": 3.075038194656372, "learning_rate": 1.430515452318354e-07, "loss": 1.3445, "step": 47152 }, { "epoch": 3.6953761755485894, "grad_norm": 2.4701173305511475, "learning_rate": 1.4297832119435417e-07, "loss": 1.4191, "step": 47153 }, { "epoch": 3.6954545454545453, "grad_norm": 3.382550001144409, "learning_rate": 1.4290511563048835e-07, "loss": 1.0639, "step": 47154 }, { "epoch": 3.6955329153605017, "grad_norm": 4.444145202636719, "learning_rate": 1.428319285405161e-07, "loss": 1.4273, "step": 47155 }, { "epoch": 3.6956112852664575, "grad_norm": 2.9241578578948975, "learning_rate": 1.427587599247171e-07, "loss": 1.2832, "step": 47156 }, { "epoch": 3.695689655172414, "grad_norm": 2.7145774364471436, "learning_rate": 1.42685609783369e-07, "loss": 1.17, "step": 47157 }, { "epoch": 3.6957680250783698, "grad_norm": 3.5989527702331543, "learning_rate": 1.4261247811674983e-07, "loss": 1.198, "step": 47158 }, { "epoch": 3.695846394984326, "grad_norm": 3.987435817718506, "learning_rate": 1.4253936492513775e-07, "loss": 1.5137, "step": 47159 }, { "epoch": 3.695924764890282, "grad_norm": 3.016903877258301, "learning_rate": 1.4246627020881032e-07, "loss": 1.3733, "step": 47160 }, { "epoch": 3.6960031347962383, "grad_norm": 2.4283287525177, "learning_rate": 1.4239319396804728e-07, "loss": 0.813, "step": 47161 }, { "epoch": 3.696081504702194, "grad_norm": 6.080491542816162, "learning_rate": 1.4232013620312513e-07, "loss": 0.9445, "step": 47162 }, { "epoch": 3.6961598746081505, "grad_norm": 3.4059245586395264, "learning_rate": 1.4224709691432248e-07, "loss": 1.4677, "step": 47163 }, { "epoch": 3.6962382445141064, "grad_norm": 2.556006669998169, "learning_rate": 1.4217407610191635e-07, "loss": 1.2659, "step": 47164 }, { "epoch": 3.6963166144200628, "grad_norm": 3.3663575649261475, "learning_rate": 1.4210107376618488e-07, "loss": 1.5435, "step": 47165 }, { "epoch": 3.6963949843260187, "grad_norm": 2.4694254398345947, "learning_rate": 1.4202808990740558e-07, "loss": 1.2203, "step": 47166 }, { "epoch": 3.696473354231975, "grad_norm": 4.320659160614014, "learning_rate": 1.4195512452585659e-07, "loss": 0.8246, "step": 47167 }, { "epoch": 3.696551724137931, "grad_norm": 2.843715190887451, "learning_rate": 1.418821776218149e-07, "loss": 1.0999, "step": 47168 }, { "epoch": 3.696630094043887, "grad_norm": 4.676968097686768, "learning_rate": 1.4180924919555804e-07, "loss": 1.4295, "step": 47169 }, { "epoch": 3.696708463949843, "grad_norm": 3.787614345550537, "learning_rate": 1.4173633924736364e-07, "loss": 1.0998, "step": 47170 }, { "epoch": 3.6967868338557994, "grad_norm": 2.8376781940460205, "learning_rate": 1.4166344777750917e-07, "loss": 1.3252, "step": 47171 }, { "epoch": 3.6968652037617558, "grad_norm": 2.8516483306884766, "learning_rate": 1.415905747862717e-07, "loss": 1.2392, "step": 47172 }, { "epoch": 3.6969435736677116, "grad_norm": 2.714709520339966, "learning_rate": 1.4151772027392707e-07, "loss": 1.5702, "step": 47173 }, { "epoch": 3.6970219435736675, "grad_norm": 3.4863953590393066, "learning_rate": 1.4144488424075398e-07, "loss": 1.0105, "step": 47174 }, { "epoch": 3.697100313479624, "grad_norm": 2.762646198272705, "learning_rate": 1.413720666870294e-07, "loss": 1.2648, "step": 47175 }, { "epoch": 3.69717868338558, "grad_norm": 3.4205312728881836, "learning_rate": 1.4129926761303037e-07, "loss": 1.1701, "step": 47176 }, { "epoch": 3.697257053291536, "grad_norm": 2.430018424987793, "learning_rate": 1.4122648701903274e-07, "loss": 1.4413, "step": 47177 }, { "epoch": 3.697335423197492, "grad_norm": 2.3339080810546875, "learning_rate": 1.41153724905313e-07, "loss": 1.3365, "step": 47178 }, { "epoch": 3.6974137931034483, "grad_norm": 2.8411412239074707, "learning_rate": 1.410809812721503e-07, "loss": 1.4186, "step": 47179 }, { "epoch": 3.6974921630094046, "grad_norm": 4.3787841796875, "learning_rate": 1.410082561198195e-07, "loss": 1.5526, "step": 47180 }, { "epoch": 3.6975705329153605, "grad_norm": 2.966461181640625, "learning_rate": 1.4093554944859756e-07, "loss": 1.4998, "step": 47181 }, { "epoch": 3.6976489028213164, "grad_norm": 3.2662460803985596, "learning_rate": 1.4086286125876092e-07, "loss": 1.2775, "step": 47182 }, { "epoch": 3.6977272727272728, "grad_norm": 4.945240497589111, "learning_rate": 1.4079019155058604e-07, "loss": 1.0634, "step": 47183 }, { "epoch": 3.697805642633229, "grad_norm": 3.0824570655822754, "learning_rate": 1.4071754032434938e-07, "loss": 1.0632, "step": 47184 }, { "epoch": 3.697884012539185, "grad_norm": 2.549027442932129, "learning_rate": 1.4064490758032735e-07, "loss": 1.4639, "step": 47185 }, { "epoch": 3.697962382445141, "grad_norm": 4.462919235229492, "learning_rate": 1.4057229331879586e-07, "loss": 1.1733, "step": 47186 }, { "epoch": 3.698040752351097, "grad_norm": 3.688481569290161, "learning_rate": 1.404996975400319e-07, "loss": 1.1151, "step": 47187 }, { "epoch": 3.6981191222570535, "grad_norm": 2.5352017879486084, "learning_rate": 1.4042712024431083e-07, "loss": 1.225, "step": 47188 }, { "epoch": 3.6981974921630094, "grad_norm": 2.3342294692993164, "learning_rate": 1.4035456143190907e-07, "loss": 1.2209, "step": 47189 }, { "epoch": 3.6982758620689653, "grad_norm": 4.50840425491333, "learning_rate": 1.4028202110310253e-07, "loss": 1.1599, "step": 47190 }, { "epoch": 3.6983542319749216, "grad_norm": 3.898195743560791, "learning_rate": 1.4020949925816761e-07, "loss": 0.9407, "step": 47191 }, { "epoch": 3.698432601880878, "grad_norm": 4.203786849975586, "learning_rate": 1.4013699589737861e-07, "loss": 1.3457, "step": 47192 }, { "epoch": 3.698510971786834, "grad_norm": 3.7912464141845703, "learning_rate": 1.4006451102101248e-07, "loss": 1.4798, "step": 47193 }, { "epoch": 3.6985893416927897, "grad_norm": 19.58697509765625, "learning_rate": 1.399920446293457e-07, "loss": 1.6825, "step": 47194 }, { "epoch": 3.698667711598746, "grad_norm": 2.6793649196624756, "learning_rate": 1.3991959672265186e-07, "loss": 1.3701, "step": 47195 }, { "epoch": 3.6987460815047024, "grad_norm": 5.422623157501221, "learning_rate": 1.3984716730120806e-07, "loss": 1.2379, "step": 47196 }, { "epoch": 3.6988244514106583, "grad_norm": 3.935474157333374, "learning_rate": 1.3977475636528847e-07, "loss": 1.3923, "step": 47197 }, { "epoch": 3.698902821316614, "grad_norm": 11.005056381225586, "learning_rate": 1.3970236391517067e-07, "loss": 0.6684, "step": 47198 }, { "epoch": 3.6989811912225705, "grad_norm": 4.630237579345703, "learning_rate": 1.3962998995112776e-07, "loss": 1.8137, "step": 47199 }, { "epoch": 3.699059561128527, "grad_norm": 3.1910202503204346, "learning_rate": 1.3955763447343618e-07, "loss": 1.4376, "step": 47200 }, { "epoch": 3.6991379310344827, "grad_norm": 3.855886697769165, "learning_rate": 1.3948529748237126e-07, "loss": 1.3104, "step": 47201 }, { "epoch": 3.6992163009404386, "grad_norm": 3.507448673248291, "learning_rate": 1.3941297897820728e-07, "loss": 1.5181, "step": 47202 }, { "epoch": 3.699294670846395, "grad_norm": 4.500576019287109, "learning_rate": 1.393406789612195e-07, "loss": 1.0775, "step": 47203 }, { "epoch": 3.6993730407523513, "grad_norm": 2.657986879348755, "learning_rate": 1.3926839743168387e-07, "loss": 1.1255, "step": 47204 }, { "epoch": 3.699451410658307, "grad_norm": 2.7321207523345947, "learning_rate": 1.3919613438987456e-07, "loss": 1.1378, "step": 47205 }, { "epoch": 3.699529780564263, "grad_norm": 3.8829259872436523, "learning_rate": 1.3912388983606585e-07, "loss": 1.211, "step": 47206 }, { "epoch": 3.6996081504702194, "grad_norm": 4.59284782409668, "learning_rate": 1.3905166377053358e-07, "loss": 1.4284, "step": 47207 }, { "epoch": 3.6996865203761757, "grad_norm": 2.8136146068573, "learning_rate": 1.38979456193552e-07, "loss": 1.4124, "step": 47208 }, { "epoch": 3.6997648902821316, "grad_norm": 4.907778739929199, "learning_rate": 1.3890726710539592e-07, "loss": 1.4186, "step": 47209 }, { "epoch": 3.6998432601880875, "grad_norm": 3.885674238204956, "learning_rate": 1.3883509650634009e-07, "loss": 1.2686, "step": 47210 }, { "epoch": 3.699921630094044, "grad_norm": 4.142806053161621, "learning_rate": 1.387629443966576e-07, "loss": 1.3413, "step": 47211 }, { "epoch": 3.7, "grad_norm": 2.859121561050415, "learning_rate": 1.3869081077662437e-07, "loss": 1.1937, "step": 47212 }, { "epoch": 3.700078369905956, "grad_norm": 2.9216954708099365, "learning_rate": 1.386186956465152e-07, "loss": 1.2537, "step": 47213 }, { "epoch": 3.700156739811912, "grad_norm": 3.32385516166687, "learning_rate": 1.3854659900660316e-07, "loss": 1.5231, "step": 47214 }, { "epoch": 3.7002351097178683, "grad_norm": 2.812856435775757, "learning_rate": 1.3847452085716252e-07, "loss": 1.4109, "step": 47215 }, { "epoch": 3.7003134796238246, "grad_norm": 3.497130870819092, "learning_rate": 1.3840246119846745e-07, "loss": 1.0433, "step": 47216 }, { "epoch": 3.7003918495297805, "grad_norm": 4.246815204620361, "learning_rate": 1.3833042003079334e-07, "loss": 1.0697, "step": 47217 }, { "epoch": 3.700470219435737, "grad_norm": 5.093089580535889, "learning_rate": 1.382583973544127e-07, "loss": 1.4308, "step": 47218 }, { "epoch": 3.7005485893416927, "grad_norm": 3.148350954055786, "learning_rate": 1.381863931695998e-07, "loss": 1.4316, "step": 47219 }, { "epoch": 3.700626959247649, "grad_norm": 7.674506664276123, "learning_rate": 1.3811440747662886e-07, "loss": 0.9263, "step": 47220 }, { "epoch": 3.700705329153605, "grad_norm": 3.0119259357452393, "learning_rate": 1.3804244027577351e-07, "loss": 1.0461, "step": 47221 }, { "epoch": 3.7007836990595613, "grad_norm": 2.681116819381714, "learning_rate": 1.3797049156730745e-07, "loss": 1.2635, "step": 47222 }, { "epoch": 3.700862068965517, "grad_norm": 3.147592782974243, "learning_rate": 1.3789856135150436e-07, "loss": 1.0513, "step": 47223 }, { "epoch": 3.7009404388714735, "grad_norm": 4.348380088806152, "learning_rate": 1.3782664962863845e-07, "loss": 0.591, "step": 47224 }, { "epoch": 3.7010188087774294, "grad_norm": 3.3315188884735107, "learning_rate": 1.3775475639898118e-07, "loss": 1.2585, "step": 47225 }, { "epoch": 3.7010971786833857, "grad_norm": 3.059985876083374, "learning_rate": 1.3768288166280786e-07, "loss": 1.4785, "step": 47226 }, { "epoch": 3.7011755485893416, "grad_norm": 3.2380177974700928, "learning_rate": 1.3761102542039162e-07, "loss": 1.4283, "step": 47227 }, { "epoch": 3.701253918495298, "grad_norm": 8.51262092590332, "learning_rate": 1.3753918767200612e-07, "loss": 1.4168, "step": 47228 }, { "epoch": 3.701332288401254, "grad_norm": 5.299081802368164, "learning_rate": 1.3746736841792397e-07, "loss": 1.2617, "step": 47229 }, { "epoch": 3.70141065830721, "grad_norm": 2.2645952701568604, "learning_rate": 1.3739556765841712e-07, "loss": 0.9544, "step": 47230 }, { "epoch": 3.701489028213166, "grad_norm": 8.092904090881348, "learning_rate": 1.373237853937609e-07, "loss": 1.4689, "step": 47231 }, { "epoch": 3.7015673981191224, "grad_norm": 7.770438194274902, "learning_rate": 1.372520216242279e-07, "loss": 1.2075, "step": 47232 }, { "epoch": 3.7016457680250783, "grad_norm": 2.839085817337036, "learning_rate": 1.3718027635008958e-07, "loss": 1.2666, "step": 47233 }, { "epoch": 3.7017241379310346, "grad_norm": 3.1635026931762695, "learning_rate": 1.3710854957162012e-07, "loss": 1.2596, "step": 47234 }, { "epoch": 3.7018025078369905, "grad_norm": 4.427730560302734, "learning_rate": 1.3703684128909212e-07, "loss": 1.4349, "step": 47235 }, { "epoch": 3.701880877742947, "grad_norm": 5.348193168640137, "learning_rate": 1.369651515027781e-07, "loss": 1.3548, "step": 47236 }, { "epoch": 3.7019592476489027, "grad_norm": 4.900251388549805, "learning_rate": 1.3689348021295068e-07, "loss": 1.6776, "step": 47237 }, { "epoch": 3.702037617554859, "grad_norm": 4.340909957885742, "learning_rate": 1.3682182741988293e-07, "loss": 1.28, "step": 47238 }, { "epoch": 3.702115987460815, "grad_norm": 3.0464718341827393, "learning_rate": 1.3675019312384685e-07, "loss": 1.7746, "step": 47239 }, { "epoch": 3.7021943573667713, "grad_norm": 3.7552542686462402, "learning_rate": 1.3667857732511502e-07, "loss": 0.8798, "step": 47240 }, { "epoch": 3.702272727272727, "grad_norm": 20.945770263671875, "learning_rate": 1.3660698002396e-07, "loss": 1.1478, "step": 47241 }, { "epoch": 3.7023510971786835, "grad_norm": 5.449261665344238, "learning_rate": 1.3653540122065377e-07, "loss": 1.373, "step": 47242 }, { "epoch": 3.7024294670846394, "grad_norm": 2.436415433883667, "learning_rate": 1.364638409154695e-07, "loss": 1.2461, "step": 47243 }, { "epoch": 3.7025078369905957, "grad_norm": 2.3298587799072266, "learning_rate": 1.3639229910867803e-07, "loss": 1.0979, "step": 47244 }, { "epoch": 3.7025862068965516, "grad_norm": 2.272838830947876, "learning_rate": 1.3632077580055246e-07, "loss": 1.2998, "step": 47245 }, { "epoch": 3.702664576802508, "grad_norm": 2.644540309906006, "learning_rate": 1.362492709913643e-07, "loss": 1.3561, "step": 47246 }, { "epoch": 3.702742946708464, "grad_norm": 9.6244535446167, "learning_rate": 1.3617778468138666e-07, "loss": 1.6749, "step": 47247 }, { "epoch": 3.70282131661442, "grad_norm": 3.520326614379883, "learning_rate": 1.361063168708898e-07, "loss": 1.491, "step": 47248 }, { "epoch": 3.702899686520376, "grad_norm": 2.4606776237487793, "learning_rate": 1.3603486756014583e-07, "loss": 1.4201, "step": 47249 }, { "epoch": 3.7029780564263324, "grad_norm": 3.8926188945770264, "learning_rate": 1.359634367494278e-07, "loss": 1.3339, "step": 47250 }, { "epoch": 3.7030564263322883, "grad_norm": 2.8610706329345703, "learning_rate": 1.3589202443900608e-07, "loss": 1.4167, "step": 47251 }, { "epoch": 3.7031347962382446, "grad_norm": 8.933734893798828, "learning_rate": 1.3582063062915264e-07, "loss": 0.5981, "step": 47252 }, { "epoch": 3.7032131661442005, "grad_norm": 2.522296667098999, "learning_rate": 1.3574925532013894e-07, "loss": 1.2056, "step": 47253 }, { "epoch": 3.703291536050157, "grad_norm": 10.467838287353516, "learning_rate": 1.3567789851223701e-07, "loss": 1.9817, "step": 47254 }, { "epoch": 3.7033699059561127, "grad_norm": 2.5619728565216064, "learning_rate": 1.3560656020571772e-07, "loss": 1.1922, "step": 47255 }, { "epoch": 3.703448275862069, "grad_norm": 3.2815487384796143, "learning_rate": 1.3553524040085252e-07, "loss": 1.3006, "step": 47256 }, { "epoch": 3.703526645768025, "grad_norm": 2.2595009803771973, "learning_rate": 1.354639390979129e-07, "loss": 1.369, "step": 47257 }, { "epoch": 3.7036050156739813, "grad_norm": 2.8606045246124268, "learning_rate": 1.353926562971697e-07, "loss": 1.4058, "step": 47258 }, { "epoch": 3.703683385579937, "grad_norm": 2.437283515930176, "learning_rate": 1.3532139199889437e-07, "loss": 1.143, "step": 47259 }, { "epoch": 3.7037617554858935, "grad_norm": 6.422535419464111, "learning_rate": 1.3525014620335786e-07, "loss": 1.5483, "step": 47260 }, { "epoch": 3.7038401253918494, "grad_norm": 2.3168442249298096, "learning_rate": 1.3517891891083156e-07, "loss": 1.2245, "step": 47261 }, { "epoch": 3.7039184952978057, "grad_norm": 2.7404539585113525, "learning_rate": 1.351077101215853e-07, "loss": 1.3047, "step": 47262 }, { "epoch": 3.7039968652037616, "grad_norm": 2.4434025287628174, "learning_rate": 1.3503651983588993e-07, "loss": 1.0425, "step": 47263 }, { "epoch": 3.704075235109718, "grad_norm": 2.8460607528686523, "learning_rate": 1.349653480540175e-07, "loss": 1.1817, "step": 47264 }, { "epoch": 3.704153605015674, "grad_norm": 2.730398416519165, "learning_rate": 1.348941947762389e-07, "loss": 1.0854, "step": 47265 }, { "epoch": 3.70423197492163, "grad_norm": 4.816825866699219, "learning_rate": 1.3482306000282276e-07, "loss": 1.4849, "step": 47266 }, { "epoch": 3.7043103448275865, "grad_norm": 3.0838239192962646, "learning_rate": 1.347519437340411e-07, "loss": 1.2816, "step": 47267 }, { "epoch": 3.7043887147335424, "grad_norm": 3.6349635124206543, "learning_rate": 1.3468084597016372e-07, "loss": 0.9159, "step": 47268 }, { "epoch": 3.7044670846394983, "grad_norm": 8.002447128295898, "learning_rate": 1.3460976671146208e-07, "loss": 1.9602, "step": 47269 }, { "epoch": 3.7045454545454546, "grad_norm": 7.555739879608154, "learning_rate": 1.3453870595820539e-07, "loss": 0.9812, "step": 47270 }, { "epoch": 3.704623824451411, "grad_norm": 3.4298367500305176, "learning_rate": 1.3446766371066457e-07, "loss": 1.364, "step": 47271 }, { "epoch": 3.704702194357367, "grad_norm": 2.975614547729492, "learning_rate": 1.343966399691099e-07, "loss": 1.0768, "step": 47272 }, { "epoch": 3.7047805642633227, "grad_norm": 13.27267837524414, "learning_rate": 1.3432563473381067e-07, "loss": 0.9344, "step": 47273 }, { "epoch": 3.704858934169279, "grad_norm": 2.6727817058563232, "learning_rate": 1.3425464800503828e-07, "loss": 0.8033, "step": 47274 }, { "epoch": 3.7049373040752354, "grad_norm": 2.4503536224365234, "learning_rate": 1.3418367978306145e-07, "loss": 1.276, "step": 47275 }, { "epoch": 3.7050156739811912, "grad_norm": 4.916049957275391, "learning_rate": 1.3411273006815106e-07, "loss": 1.2624, "step": 47276 }, { "epoch": 3.705094043887147, "grad_norm": 3.5441253185272217, "learning_rate": 1.3404179886057633e-07, "loss": 1.371, "step": 47277 }, { "epoch": 3.7051724137931035, "grad_norm": 2.9739174842834473, "learning_rate": 1.3397088616060704e-07, "loss": 1.4334, "step": 47278 }, { "epoch": 3.70525078369906, "grad_norm": 2.532092809677124, "learning_rate": 1.3389999196851356e-07, "loss": 1.5757, "step": 47279 }, { "epoch": 3.7053291536050157, "grad_norm": 7.316729545593262, "learning_rate": 1.3382911628456564e-07, "loss": 1.1543, "step": 47280 }, { "epoch": 3.7054075235109716, "grad_norm": 4.847686290740967, "learning_rate": 1.3375825910903083e-07, "loss": 1.3736, "step": 47281 }, { "epoch": 3.705485893416928, "grad_norm": 4.202596664428711, "learning_rate": 1.336874204421812e-07, "loss": 1.4229, "step": 47282 }, { "epoch": 3.7055642633228842, "grad_norm": 3.242858409881592, "learning_rate": 1.336166002842848e-07, "loss": 1.1798, "step": 47283 }, { "epoch": 3.70564263322884, "grad_norm": 3.238574504852295, "learning_rate": 1.3354579863561202e-07, "loss": 0.5288, "step": 47284 }, { "epoch": 3.705721003134796, "grad_norm": 3.8988735675811768, "learning_rate": 1.3347501549643095e-07, "loss": 1.361, "step": 47285 }, { "epoch": 3.7057993730407524, "grad_norm": 2.9030749797821045, "learning_rate": 1.334042508670108e-07, "loss": 1.3046, "step": 47286 }, { "epoch": 3.7058777429467087, "grad_norm": 2.8513295650482178, "learning_rate": 1.3333350474762142e-07, "loss": 1.5013, "step": 47287 }, { "epoch": 3.7059561128526646, "grad_norm": 4.17275857925415, "learning_rate": 1.3326277713853253e-07, "loss": 1.1953, "step": 47288 }, { "epoch": 3.7060344827586205, "grad_norm": 3.70392107963562, "learning_rate": 1.3319206804001172e-07, "loss": 1.485, "step": 47289 }, { "epoch": 3.706112852664577, "grad_norm": 3.4866342544555664, "learning_rate": 1.3312137745232878e-07, "loss": 1.6111, "step": 47290 }, { "epoch": 3.706191222570533, "grad_norm": 5.172627925872803, "learning_rate": 1.3305070537575237e-07, "loss": 1.2217, "step": 47291 }, { "epoch": 3.706269592476489, "grad_norm": 2.9502322673797607, "learning_rate": 1.3298005181055062e-07, "loss": 1.1625, "step": 47292 }, { "epoch": 3.706347962382445, "grad_norm": 2.6810946464538574, "learning_rate": 1.329094167569933e-07, "loss": 1.41, "step": 47293 }, { "epoch": 3.7064263322884012, "grad_norm": 3.9381792545318604, "learning_rate": 1.3283880021534855e-07, "loss": 0.9297, "step": 47294 }, { "epoch": 3.7065047021943576, "grad_norm": 3.8842215538024902, "learning_rate": 1.327682021858856e-07, "loss": 1.1214, "step": 47295 }, { "epoch": 3.7065830721003135, "grad_norm": 2.4076123237609863, "learning_rate": 1.3269762266887197e-07, "loss": 1.2925, "step": 47296 }, { "epoch": 3.7066614420062693, "grad_norm": 2.5415141582489014, "learning_rate": 1.3262706166457638e-07, "loss": 1.179, "step": 47297 }, { "epoch": 3.7067398119122257, "grad_norm": 3.9603662490844727, "learning_rate": 1.3255651917326751e-07, "loss": 1.6244, "step": 47298 }, { "epoch": 3.706818181818182, "grad_norm": 3.0231902599334717, "learning_rate": 1.3248599519521454e-07, "loss": 1.2946, "step": 47299 }, { "epoch": 3.706896551724138, "grad_norm": 3.6575584411621094, "learning_rate": 1.324154897306834e-07, "loss": 1.5076, "step": 47300 }, { "epoch": 3.706974921630094, "grad_norm": 2.824733257293701, "learning_rate": 1.3234500277994388e-07, "loss": 1.3783, "step": 47301 }, { "epoch": 3.70705329153605, "grad_norm": 4.554073810577393, "learning_rate": 1.3227453434326464e-07, "loss": 1.0375, "step": 47302 }, { "epoch": 3.7071316614420065, "grad_norm": 3.2514114379882812, "learning_rate": 1.3220408442091214e-07, "loss": 1.3613, "step": 47303 }, { "epoch": 3.7072100313479623, "grad_norm": 2.9275145530700684, "learning_rate": 1.3213365301315505e-07, "loss": 1.7237, "step": 47304 }, { "epoch": 3.7072884012539182, "grad_norm": 2.7804160118103027, "learning_rate": 1.3206324012026095e-07, "loss": 1.4292, "step": 47305 }, { "epoch": 3.7073667711598746, "grad_norm": 4.728845596313477, "learning_rate": 1.319928457424985e-07, "loss": 1.3107, "step": 47306 }, { "epoch": 3.707445141065831, "grad_norm": 2.3362739086151123, "learning_rate": 1.319224698801347e-07, "loss": 1.1518, "step": 47307 }, { "epoch": 3.707523510971787, "grad_norm": 3.931034803390503, "learning_rate": 1.3185211253343765e-07, "loss": 1.2356, "step": 47308 }, { "epoch": 3.7076018808777427, "grad_norm": 3.3661789894104004, "learning_rate": 1.3178177370267442e-07, "loss": 1.2245, "step": 47309 }, { "epoch": 3.707680250783699, "grad_norm": 4.246784687042236, "learning_rate": 1.3171145338811307e-07, "loss": 1.3879, "step": 47310 }, { "epoch": 3.7077586206896553, "grad_norm": 11.156990051269531, "learning_rate": 1.3164115159002068e-07, "loss": 1.8599, "step": 47311 }, { "epoch": 3.7078369905956112, "grad_norm": 5.228463649749756, "learning_rate": 1.3157086830866471e-07, "loss": 1.8004, "step": 47312 }, { "epoch": 3.7079153605015676, "grad_norm": 4.565684795379639, "learning_rate": 1.3150060354431282e-07, "loss": 0.8618, "step": 47313 }, { "epoch": 3.7079937304075234, "grad_norm": 6.1221604347229, "learning_rate": 1.3143035729723197e-07, "loss": 1.2203, "step": 47314 }, { "epoch": 3.70807210031348, "grad_norm": 5.6542840003967285, "learning_rate": 1.3136012956768918e-07, "loss": 1.4689, "step": 47315 }, { "epoch": 3.7081504702194357, "grad_norm": 4.000443935394287, "learning_rate": 1.31289920355952e-07, "loss": 1.3203, "step": 47316 }, { "epoch": 3.708228840125392, "grad_norm": 2.9043474197387695, "learning_rate": 1.3121972966228747e-07, "loss": 1.3886, "step": 47317 }, { "epoch": 3.708307210031348, "grad_norm": 2.3087542057037354, "learning_rate": 1.31149557486962e-07, "loss": 0.9231, "step": 47318 }, { "epoch": 3.708385579937304, "grad_norm": 3.210958242416382, "learning_rate": 1.310794038302421e-07, "loss": 1.0984, "step": 47319 }, { "epoch": 3.70846394984326, "grad_norm": 1.9032613039016724, "learning_rate": 1.3100926869239583e-07, "loss": 1.312, "step": 47320 }, { "epoch": 3.7085423197492164, "grad_norm": 2.162919759750366, "learning_rate": 1.3093915207369023e-07, "loss": 1.3526, "step": 47321 }, { "epoch": 3.7086206896551723, "grad_norm": 4.035548210144043, "learning_rate": 1.3086905397439064e-07, "loss": 1.173, "step": 47322 }, { "epoch": 3.7086990595611287, "grad_norm": 2.3870279788970947, "learning_rate": 1.307989743947641e-07, "loss": 1.4766, "step": 47323 }, { "epoch": 3.7087774294670846, "grad_norm": 4.515749454498291, "learning_rate": 1.3072891333507698e-07, "loss": 1.5806, "step": 47324 }, { "epoch": 3.708855799373041, "grad_norm": 4.830541133880615, "learning_rate": 1.306588707955969e-07, "loss": 1.2958, "step": 47325 }, { "epoch": 3.7089341692789968, "grad_norm": 2.5773611068725586, "learning_rate": 1.3058884677658866e-07, "loss": 1.1686, "step": 47326 }, { "epoch": 3.709012539184953, "grad_norm": 8.166913032531738, "learning_rate": 1.3051884127831982e-07, "loss": 0.8424, "step": 47327 }, { "epoch": 3.709090909090909, "grad_norm": 3.2770192623138428, "learning_rate": 1.3044885430105568e-07, "loss": 1.3922, "step": 47328 }, { "epoch": 3.7091692789968653, "grad_norm": 3.368213653564453, "learning_rate": 1.3037888584506332e-07, "loss": 1.0954, "step": 47329 }, { "epoch": 3.709247648902821, "grad_norm": 11.37448787689209, "learning_rate": 1.3030893591060855e-07, "loss": 1.1701, "step": 47330 }, { "epoch": 3.7093260188087775, "grad_norm": 3.2651126384735107, "learning_rate": 1.3023900449795735e-07, "loss": 1.3196, "step": 47331 }, { "epoch": 3.7094043887147334, "grad_norm": 4.620545387268066, "learning_rate": 1.301690916073761e-07, "loss": 1.595, "step": 47332 }, { "epoch": 3.7094827586206898, "grad_norm": 5.604731559753418, "learning_rate": 1.300991972391291e-07, "loss": 1.6164, "step": 47333 }, { "epoch": 3.7095611285266457, "grad_norm": 3.210404396057129, "learning_rate": 1.300293213934839e-07, "loss": 1.3498, "step": 47334 }, { "epoch": 3.709639498432602, "grad_norm": 4.412880897521973, "learning_rate": 1.2995946407070636e-07, "loss": 1.285, "step": 47335 }, { "epoch": 3.709717868338558, "grad_norm": 2.773951292037964, "learning_rate": 1.2988962527106187e-07, "loss": 1.2931, "step": 47336 }, { "epoch": 3.709796238244514, "grad_norm": 4.317274570465088, "learning_rate": 1.2981980499481517e-07, "loss": 1.9965, "step": 47337 }, { "epoch": 3.70987460815047, "grad_norm": 4.834315776824951, "learning_rate": 1.2975000324223164e-07, "loss": 1.2915, "step": 47338 }, { "epoch": 3.7099529780564264, "grad_norm": 8.2017183303833, "learning_rate": 1.2968022001357827e-07, "loss": 0.9912, "step": 47339 }, { "epoch": 3.7100313479623823, "grad_norm": 3.57235050201416, "learning_rate": 1.2961045530912043e-07, "loss": 1.2662, "step": 47340 }, { "epoch": 3.7101097178683387, "grad_norm": 3.0308303833007812, "learning_rate": 1.2954070912912232e-07, "loss": 1.2582, "step": 47341 }, { "epoch": 3.7101880877742945, "grad_norm": 4.01386833190918, "learning_rate": 1.2947098147384984e-07, "loss": 1.7537, "step": 47342 }, { "epoch": 3.710266457680251, "grad_norm": 9.291435241699219, "learning_rate": 1.2940127234356725e-07, "loss": 0.744, "step": 47343 }, { "epoch": 3.7103448275862068, "grad_norm": 2.530027389526367, "learning_rate": 1.293315817385421e-07, "loss": 1.2899, "step": 47344 }, { "epoch": 3.710423197492163, "grad_norm": 3.7981996536254883, "learning_rate": 1.2926190965903695e-07, "loss": 1.295, "step": 47345 }, { "epoch": 3.710501567398119, "grad_norm": 3.607661247253418, "learning_rate": 1.291922561053177e-07, "loss": 1.3352, "step": 47346 }, { "epoch": 3.7105799373040753, "grad_norm": 2.337494373321533, "learning_rate": 1.2912262107764916e-07, "loss": 1.458, "step": 47347 }, { "epoch": 3.710658307210031, "grad_norm": 4.108240604400635, "learning_rate": 1.2905300457629665e-07, "loss": 1.2419, "step": 47348 }, { "epoch": 3.7107366771159875, "grad_norm": 2.375278949737549, "learning_rate": 1.2898340660152442e-07, "loss": 1.4144, "step": 47349 }, { "epoch": 3.7108150470219434, "grad_norm": 2.8207662105560303, "learning_rate": 1.289138271535978e-07, "loss": 1.4967, "step": 47350 }, { "epoch": 3.7108934169278998, "grad_norm": 3.7070553302764893, "learning_rate": 1.28844266232781e-07, "loss": 1.092, "step": 47351 }, { "epoch": 3.7109717868338556, "grad_norm": 3.4670915603637695, "learning_rate": 1.287747238393383e-07, "loss": 1.1449, "step": 47352 }, { "epoch": 3.711050156739812, "grad_norm": 4.005547523498535, "learning_rate": 1.2870519997353448e-07, "loss": 1.505, "step": 47353 }, { "epoch": 3.711128526645768, "grad_norm": 2.7338523864746094, "learning_rate": 1.286356946356343e-07, "loss": 1.0799, "step": 47354 }, { "epoch": 3.711206896551724, "grad_norm": 3.226330041885376, "learning_rate": 1.2856620782590201e-07, "loss": 0.7875, "step": 47355 }, { "epoch": 3.71128526645768, "grad_norm": 4.9562788009643555, "learning_rate": 1.2849673954460128e-07, "loss": 1.6022, "step": 47356 }, { "epoch": 3.7113636363636364, "grad_norm": 3.11336088180542, "learning_rate": 1.2842728979199638e-07, "loss": 1.4645, "step": 47357 }, { "epoch": 3.7114420062695928, "grad_norm": 3.2315237522125244, "learning_rate": 1.2835785856835316e-07, "loss": 1.1922, "step": 47358 }, { "epoch": 3.7115203761755486, "grad_norm": 4.0653204917907715, "learning_rate": 1.2828844587393362e-07, "loss": 1.0009, "step": 47359 }, { "epoch": 3.7115987460815045, "grad_norm": 3.6275079250335693, "learning_rate": 1.2821905170900262e-07, "loss": 1.4068, "step": 47360 }, { "epoch": 3.711677115987461, "grad_norm": 2.845240592956543, "learning_rate": 1.2814967607382433e-07, "loss": 1.2621, "step": 47361 }, { "epoch": 3.711755485893417, "grad_norm": 3.490396738052368, "learning_rate": 1.280803189686619e-07, "loss": 1.8717, "step": 47362 }, { "epoch": 3.711833855799373, "grad_norm": 3.4229536056518555, "learning_rate": 1.2801098039377958e-07, "loss": 1.5654, "step": 47363 }, { "epoch": 3.711912225705329, "grad_norm": 4.87824010848999, "learning_rate": 1.2794166034944154e-07, "loss": 1.188, "step": 47364 }, { "epoch": 3.7119905956112853, "grad_norm": 5.5774617195129395, "learning_rate": 1.2787235883591042e-07, "loss": 1.4929, "step": 47365 }, { "epoch": 3.7120689655172416, "grad_norm": 3.1807334423065186, "learning_rate": 1.2780307585345042e-07, "loss": 1.472, "step": 47366 }, { "epoch": 3.7121473354231975, "grad_norm": 4.052004337310791, "learning_rate": 1.2773381140232522e-07, "loss": 1.6154, "step": 47367 }, { "epoch": 3.7122257053291534, "grad_norm": 4.026782512664795, "learning_rate": 1.2766456548279794e-07, "loss": 1.0461, "step": 47368 }, { "epoch": 3.7123040752351097, "grad_norm": 6.030054092407227, "learning_rate": 1.2759533809513224e-07, "loss": 1.3419, "step": 47369 }, { "epoch": 3.712382445141066, "grad_norm": 2.8672218322753906, "learning_rate": 1.2752612923959128e-07, "loss": 1.6174, "step": 47370 }, { "epoch": 3.712460815047022, "grad_norm": 2.7638447284698486, "learning_rate": 1.2745693891643818e-07, "loss": 1.4264, "step": 47371 }, { "epoch": 3.712539184952978, "grad_norm": 2.688632011413574, "learning_rate": 1.2738776712593604e-07, "loss": 1.3809, "step": 47372 }, { "epoch": 3.712617554858934, "grad_norm": 3.4330999851226807, "learning_rate": 1.2731861386834853e-07, "loss": 1.5071, "step": 47373 }, { "epoch": 3.7126959247648905, "grad_norm": 4.513942718505859, "learning_rate": 1.2724947914393826e-07, "loss": 1.9646, "step": 47374 }, { "epoch": 3.7127742946708464, "grad_norm": 5.494284152984619, "learning_rate": 1.271803629529672e-07, "loss": 0.8949, "step": 47375 }, { "epoch": 3.7128526645768023, "grad_norm": 7.437290668487549, "learning_rate": 1.271112652956996e-07, "loss": 1.1768, "step": 47376 }, { "epoch": 3.7129310344827586, "grad_norm": 5.693682670593262, "learning_rate": 1.270421861723986e-07, "loss": 1.6005, "step": 47377 }, { "epoch": 3.713009404388715, "grad_norm": 3.3802194595336914, "learning_rate": 1.2697312558332564e-07, "loss": 1.4835, "step": 47378 }, { "epoch": 3.713087774294671, "grad_norm": 9.125922203063965, "learning_rate": 1.269040835287444e-07, "loss": 1.4359, "step": 47379 }, { "epoch": 3.7131661442006267, "grad_norm": 2.4615824222564697, "learning_rate": 1.2683506000891634e-07, "loss": 1.3037, "step": 47380 }, { "epoch": 3.713244514106583, "grad_norm": 2.4717516899108887, "learning_rate": 1.267660550241051e-07, "loss": 1.439, "step": 47381 }, { "epoch": 3.7133228840125394, "grad_norm": 8.007955551147461, "learning_rate": 1.2669706857457275e-07, "loss": 1.3376, "step": 47382 }, { "epoch": 3.7134012539184953, "grad_norm": 3.2389705181121826, "learning_rate": 1.2662810066058128e-07, "loss": 1.4274, "step": 47383 }, { "epoch": 3.713479623824451, "grad_norm": 8.47054386138916, "learning_rate": 1.2655915128239383e-07, "loss": 1.9444, "step": 47384 }, { "epoch": 3.7135579937304075, "grad_norm": 6.494203567504883, "learning_rate": 1.2649022044027237e-07, "loss": 1.2948, "step": 47385 }, { "epoch": 3.713636363636364, "grad_norm": 3.3090980052948, "learning_rate": 1.264213081344784e-07, "loss": 1.2313, "step": 47386 }, { "epoch": 3.7137147335423197, "grad_norm": 3.555239677429199, "learning_rate": 1.26352414365275e-07, "loss": 1.194, "step": 47387 }, { "epoch": 3.7137931034482756, "grad_norm": 2.720954656600952, "learning_rate": 1.2628353913292368e-07, "loss": 1.3236, "step": 47388 }, { "epoch": 3.713871473354232, "grad_norm": 3.7080140113830566, "learning_rate": 1.2621468243768586e-07, "loss": 1.3407, "step": 47389 }, { "epoch": 3.7139498432601883, "grad_norm": 2.502537727355957, "learning_rate": 1.2614584427982468e-07, "loss": 1.2676, "step": 47390 }, { "epoch": 3.714028213166144, "grad_norm": 3.39052677154541, "learning_rate": 1.2607702465960103e-07, "loss": 1.3697, "step": 47391 }, { "epoch": 3.7141065830721, "grad_norm": 2.848628520965576, "learning_rate": 1.2600822357727748e-07, "loss": 1.2359, "step": 47392 }, { "epoch": 3.7141849529780564, "grad_norm": 3.9182493686676025, "learning_rate": 1.2593944103311495e-07, "loss": 1.2457, "step": 47393 }, { "epoch": 3.7142633228840127, "grad_norm": 3.7574362754821777, "learning_rate": 1.2587067702737488e-07, "loss": 1.5594, "step": 47394 }, { "epoch": 3.7143416927899686, "grad_norm": 4.350224018096924, "learning_rate": 1.2580193156031927e-07, "loss": 1.2135, "step": 47395 }, { "epoch": 3.7144200626959245, "grad_norm": 2.868447780609131, "learning_rate": 1.2573320463221073e-07, "loss": 1.3619, "step": 47396 }, { "epoch": 3.714498432601881, "grad_norm": 3.802927017211914, "learning_rate": 1.2566449624330845e-07, "loss": 1.3005, "step": 47397 }, { "epoch": 3.714576802507837, "grad_norm": 2.477961301803589, "learning_rate": 1.2559580639387502e-07, "loss": 1.6, "step": 47398 }, { "epoch": 3.714655172413793, "grad_norm": 5.056219100952148, "learning_rate": 1.255271350841708e-07, "loss": 0.899, "step": 47399 }, { "epoch": 3.714733542319749, "grad_norm": 7.383131504058838, "learning_rate": 1.254584823144589e-07, "loss": 1.4722, "step": 47400 }, { "epoch": 3.7148119122257053, "grad_norm": 2.3619191646575928, "learning_rate": 1.2538984808499856e-07, "loss": 1.0288, "step": 47401 }, { "epoch": 3.7148902821316616, "grad_norm": 4.4729084968566895, "learning_rate": 1.2532123239605176e-07, "loss": 1.1887, "step": 47402 }, { "epoch": 3.7149686520376175, "grad_norm": 3.5644524097442627, "learning_rate": 1.252526352478789e-07, "loss": 1.6131, "step": 47403 }, { "epoch": 3.7150470219435734, "grad_norm": 3.4232797622680664, "learning_rate": 1.251840566407414e-07, "loss": 1.3685, "step": 47404 }, { "epoch": 3.7151253918495297, "grad_norm": 3.3461050987243652, "learning_rate": 1.2511549657489962e-07, "loss": 1.4696, "step": 47405 }, { "epoch": 3.715203761755486, "grad_norm": 4.280793190002441, "learning_rate": 1.2504695505061503e-07, "loss": 1.703, "step": 47406 }, { "epoch": 3.715282131661442, "grad_norm": 2.8320138454437256, "learning_rate": 1.2497843206814796e-07, "loss": 1.5911, "step": 47407 }, { "epoch": 3.7153605015673983, "grad_norm": 5.848541259765625, "learning_rate": 1.2490992762775822e-07, "loss": 1.4781, "step": 47408 }, { "epoch": 3.715438871473354, "grad_norm": 5.127102375030518, "learning_rate": 1.2484144172970725e-07, "loss": 1.6627, "step": 47409 }, { "epoch": 3.7155172413793105, "grad_norm": 2.8338723182678223, "learning_rate": 1.2477297437425596e-07, "loss": 1.6355, "step": 47410 }, { "epoch": 3.7155956112852664, "grad_norm": 7.5278778076171875, "learning_rate": 1.2470452556166413e-07, "loss": 1.2364, "step": 47411 }, { "epoch": 3.7156739811912227, "grad_norm": 2.4136404991149902, "learning_rate": 1.2463609529219212e-07, "loss": 1.2669, "step": 47412 }, { "epoch": 3.7157523510971786, "grad_norm": 2.3215155601501465, "learning_rate": 1.2456768356609917e-07, "loss": 1.3548, "step": 47413 }, { "epoch": 3.715830721003135, "grad_norm": 9.562687873840332, "learning_rate": 1.2449929038364838e-07, "loss": 1.5423, "step": 47414 }, { "epoch": 3.715909090909091, "grad_norm": 2.0065202713012695, "learning_rate": 1.2443091574509681e-07, "loss": 1.0376, "step": 47415 }, { "epoch": 3.715987460815047, "grad_norm": 3.5556581020355225, "learning_rate": 1.2436255965070643e-07, "loss": 0.9596, "step": 47416 }, { "epoch": 3.716065830721003, "grad_norm": 2.787886619567871, "learning_rate": 1.2429422210073593e-07, "loss": 1.3865, "step": 47417 }, { "epoch": 3.7161442006269594, "grad_norm": 2.7844398021698, "learning_rate": 1.2422590309544625e-07, "loss": 1.0378, "step": 47418 }, { "epoch": 3.7162225705329153, "grad_norm": 2.91841983795166, "learning_rate": 1.2415760263509713e-07, "loss": 1.8211, "step": 47419 }, { "epoch": 3.7163009404388716, "grad_norm": 2.4822986125946045, "learning_rate": 1.2408932071994785e-07, "loss": 1.1793, "step": 47420 }, { "epoch": 3.7163793103448275, "grad_norm": 4.725641250610352, "learning_rate": 1.2402105735025815e-07, "loss": 1.9392, "step": 47421 }, { "epoch": 3.716457680250784, "grad_norm": 14.971273422241211, "learning_rate": 1.2395281252628844e-07, "loss": 0.8001, "step": 47422 }, { "epoch": 3.7165360501567397, "grad_norm": 7.103313446044922, "learning_rate": 1.2388458624829737e-07, "loss": 1.4176, "step": 47423 }, { "epoch": 3.716614420062696, "grad_norm": 6.0508341789245605, "learning_rate": 1.2381637851654472e-07, "loss": 1.1366, "step": 47424 }, { "epoch": 3.716692789968652, "grad_norm": 6.976726531982422, "learning_rate": 1.2374818933129028e-07, "loss": 1.5002, "step": 47425 }, { "epoch": 3.7167711598746083, "grad_norm": 9.389474868774414, "learning_rate": 1.2368001869279335e-07, "loss": 1.1391, "step": 47426 }, { "epoch": 3.716849529780564, "grad_norm": 4.947335243225098, "learning_rate": 1.2361186660131197e-07, "loss": 1.4405, "step": 47427 }, { "epoch": 3.7169278996865205, "grad_norm": 2.7327041625976562, "learning_rate": 1.235437330571071e-07, "loss": 1.1091, "step": 47428 }, { "epoch": 3.7170062695924764, "grad_norm": 4.4199042320251465, "learning_rate": 1.234756180604374e-07, "loss": 1.4539, "step": 47429 }, { "epoch": 3.7170846394984327, "grad_norm": 2.707634687423706, "learning_rate": 1.2340752161156155e-07, "loss": 1.3567, "step": 47430 }, { "epoch": 3.7171630094043886, "grad_norm": 11.572479248046875, "learning_rate": 1.2333944371073826e-07, "loss": 0.7333, "step": 47431 }, { "epoch": 3.717241379310345, "grad_norm": 4.118728160858154, "learning_rate": 1.2327138435822673e-07, "loss": 1.2448, "step": 47432 }, { "epoch": 3.717319749216301, "grad_norm": 2.6327977180480957, "learning_rate": 1.2320334355428732e-07, "loss": 1.1334, "step": 47433 }, { "epoch": 3.717398119122257, "grad_norm": 2.93677020072937, "learning_rate": 1.2313532129917648e-07, "loss": 1.7909, "step": 47434 }, { "epoch": 3.717476489028213, "grad_norm": 2.247443675994873, "learning_rate": 1.2306731759315405e-07, "loss": 1.4341, "step": 47435 }, { "epoch": 3.7175548589341694, "grad_norm": 3.204357147216797, "learning_rate": 1.2299933243647866e-07, "loss": 1.3592, "step": 47436 }, { "epoch": 3.7176332288401253, "grad_norm": 2.8347764015197754, "learning_rate": 1.22931365829409e-07, "loss": 1.1515, "step": 47437 }, { "epoch": 3.7177115987460816, "grad_norm": 2.439302682876587, "learning_rate": 1.2286341777220323e-07, "loss": 1.1738, "step": 47438 }, { "epoch": 3.7177899686520375, "grad_norm": 3.037489891052246, "learning_rate": 1.2279548826512e-07, "loss": 1.2288, "step": 47439 }, { "epoch": 3.717868338557994, "grad_norm": 5.850217342376709, "learning_rate": 1.2272757730841744e-07, "loss": 1.5782, "step": 47440 }, { "epoch": 3.7179467084639497, "grad_norm": 3.4936540126800537, "learning_rate": 1.226596849023548e-07, "loss": 1.528, "step": 47441 }, { "epoch": 3.718025078369906, "grad_norm": 4.718517303466797, "learning_rate": 1.2259181104718854e-07, "loss": 1.6563, "step": 47442 }, { "epoch": 3.718103448275862, "grad_norm": 6.4940409660339355, "learning_rate": 1.2252395574317844e-07, "loss": 1.4359, "step": 47443 }, { "epoch": 3.7181818181818183, "grad_norm": 3.282078266143799, "learning_rate": 1.2245611899058264e-07, "loss": 1.3685, "step": 47444 }, { "epoch": 3.718260188087774, "grad_norm": 2.269576072692871, "learning_rate": 1.2238830078965758e-07, "loss": 1.2589, "step": 47445 }, { "epoch": 3.7183385579937305, "grad_norm": 3.211508274078369, "learning_rate": 1.223205011406614e-07, "loss": 1.3968, "step": 47446 }, { "epoch": 3.7184169278996864, "grad_norm": 2.5505096912384033, "learning_rate": 1.222527200438539e-07, "loss": 1.0389, "step": 47447 }, { "epoch": 3.7184952978056427, "grad_norm": 3.120178699493408, "learning_rate": 1.2218495749949155e-07, "loss": 1.05, "step": 47448 }, { "epoch": 3.7185736677115986, "grad_norm": 5.635792255401611, "learning_rate": 1.221172135078319e-07, "loss": 1.5139, "step": 47449 }, { "epoch": 3.718652037617555, "grad_norm": 3.2448389530181885, "learning_rate": 1.220494880691331e-07, "loss": 1.2875, "step": 47450 }, { "epoch": 3.718730407523511, "grad_norm": 3.5075197219848633, "learning_rate": 1.219817811836521e-07, "loss": 1.2714, "step": 47451 }, { "epoch": 3.718808777429467, "grad_norm": 2.4038984775543213, "learning_rate": 1.219140928516477e-07, "loss": 1.2408, "step": 47452 }, { "epoch": 3.7188871473354235, "grad_norm": 3.7677645683288574, "learning_rate": 1.2184642307337623e-07, "loss": 1.41, "step": 47453 }, { "epoch": 3.7189655172413794, "grad_norm": 3.764357089996338, "learning_rate": 1.2177877184909538e-07, "loss": 1.0716, "step": 47454 }, { "epoch": 3.7190438871473352, "grad_norm": 2.978081464767456, "learning_rate": 1.2171113917906263e-07, "loss": 1.6869, "step": 47455 }, { "epoch": 3.7191222570532916, "grad_norm": 7.163517951965332, "learning_rate": 1.2164352506353448e-07, "loss": 1.4976, "step": 47456 }, { "epoch": 3.719200626959248, "grad_norm": 4.537080764770508, "learning_rate": 1.215759295027691e-07, "loss": 1.74, "step": 47457 }, { "epoch": 3.719278996865204, "grad_norm": 3.483905076980591, "learning_rate": 1.2150835249702285e-07, "loss": 1.764, "step": 47458 }, { "epoch": 3.7193573667711597, "grad_norm": 4.408936023712158, "learning_rate": 1.2144079404655397e-07, "loss": 1.7427, "step": 47459 }, { "epoch": 3.719435736677116, "grad_norm": 2.871342897415161, "learning_rate": 1.2137325415161716e-07, "loss": 1.4281, "step": 47460 }, { "epoch": 3.7195141065830724, "grad_norm": 3.8419556617736816, "learning_rate": 1.2130573281247172e-07, "loss": 1.0165, "step": 47461 }, { "epoch": 3.7195924764890282, "grad_norm": 3.7119131088256836, "learning_rate": 1.2123823002937296e-07, "loss": 1.0765, "step": 47462 }, { "epoch": 3.719670846394984, "grad_norm": 6.1678056716918945, "learning_rate": 1.2117074580257848e-07, "loss": 1.4365, "step": 47463 }, { "epoch": 3.7197492163009405, "grad_norm": 2.0110859870910645, "learning_rate": 1.2110328013234475e-07, "loss": 1.0669, "step": 47464 }, { "epoch": 3.719827586206897, "grad_norm": 4.504632949829102, "learning_rate": 1.2103583301892709e-07, "loss": 0.6909, "step": 47465 }, { "epoch": 3.7199059561128527, "grad_norm": 3.025712013244629, "learning_rate": 1.2096840446258363e-07, "loss": 1.3456, "step": 47466 }, { "epoch": 3.7199843260188086, "grad_norm": 3.74405837059021, "learning_rate": 1.2090099446357085e-07, "loss": 1.2634, "step": 47467 }, { "epoch": 3.720062695924765, "grad_norm": 3.309278964996338, "learning_rate": 1.2083360302214409e-07, "loss": 1.3066, "step": 47468 }, { "epoch": 3.7201410658307212, "grad_norm": 5.414889335632324, "learning_rate": 1.2076623013856037e-07, "loss": 1.5092, "step": 47469 }, { "epoch": 3.720219435736677, "grad_norm": 3.218090772628784, "learning_rate": 1.2069887581307615e-07, "loss": 1.4386, "step": 47470 }, { "epoch": 3.720297805642633, "grad_norm": 6.286698341369629, "learning_rate": 1.206315400459468e-07, "loss": 1.0247, "step": 47471 }, { "epoch": 3.7203761755485893, "grad_norm": 4.401178359985352, "learning_rate": 1.2056422283742874e-07, "loss": 1.439, "step": 47472 }, { "epoch": 3.7204545454545457, "grad_norm": 3.3048360347747803, "learning_rate": 1.2049692418777848e-07, "loss": 1.0992, "step": 47473 }, { "epoch": 3.7205329153605016, "grad_norm": 3.1959915161132812, "learning_rate": 1.2042964409725133e-07, "loss": 1.6177, "step": 47474 }, { "epoch": 3.7206112852664575, "grad_norm": 3.219560384750366, "learning_rate": 1.2036238256610376e-07, "loss": 1.777, "step": 47475 }, { "epoch": 3.720689655172414, "grad_norm": 3.7753238677978516, "learning_rate": 1.2029513959459117e-07, "loss": 1.1891, "step": 47476 }, { "epoch": 3.72076802507837, "grad_norm": 3.5117990970611572, "learning_rate": 1.2022791518296938e-07, "loss": 1.2355, "step": 47477 }, { "epoch": 3.720846394984326, "grad_norm": 4.837193965911865, "learning_rate": 1.2016070933149438e-07, "loss": 1.1599, "step": 47478 }, { "epoch": 3.720924764890282, "grad_norm": 10.30347728729248, "learning_rate": 1.2009352204042147e-07, "loss": 1.3209, "step": 47479 }, { "epoch": 3.7210031347962382, "grad_norm": 3.642636775970459, "learning_rate": 1.2002635331000657e-07, "loss": 1.6092, "step": 47480 }, { "epoch": 3.7210815047021946, "grad_norm": 2.232764720916748, "learning_rate": 1.1995920314050502e-07, "loss": 1.3397, "step": 47481 }, { "epoch": 3.7211598746081505, "grad_norm": 3.1361026763916016, "learning_rate": 1.1989207153217218e-07, "loss": 0.8241, "step": 47482 }, { "epoch": 3.7212382445141063, "grad_norm": 3.1520836353302, "learning_rate": 1.1982495848526233e-07, "loss": 1.1741, "step": 47483 }, { "epoch": 3.7213166144200627, "grad_norm": 3.831752300262451, "learning_rate": 1.1975786400003241e-07, "loss": 1.2689, "step": 47484 }, { "epoch": 3.721394984326019, "grad_norm": 6.623476982116699, "learning_rate": 1.1969078807673728e-07, "loss": 1.6368, "step": 47485 }, { "epoch": 3.721473354231975, "grad_norm": 2.479032278060913, "learning_rate": 1.196237307156317e-07, "loss": 1.1441, "step": 47486 }, { "epoch": 3.721551724137931, "grad_norm": 6.090997695922852, "learning_rate": 1.1955669191696995e-07, "loss": 1.5399, "step": 47487 }, { "epoch": 3.721630094043887, "grad_norm": 2.8794732093811035, "learning_rate": 1.194896716810079e-07, "loss": 1.5297, "step": 47488 }, { "epoch": 3.7217084639498434, "grad_norm": 2.7096145153045654, "learning_rate": 1.1942267000800146e-07, "loss": 1.5557, "step": 47489 }, { "epoch": 3.7217868338557993, "grad_norm": 3.0747733116149902, "learning_rate": 1.193556868982032e-07, "loss": 1.5095, "step": 47490 }, { "epoch": 3.7218652037617552, "grad_norm": 10.323486328125, "learning_rate": 1.192887223518696e-07, "loss": 1.298, "step": 47491 }, { "epoch": 3.7219435736677116, "grad_norm": 3.254389762878418, "learning_rate": 1.1922177636925437e-07, "loss": 1.4263, "step": 47492 }, { "epoch": 3.722021943573668, "grad_norm": 3.2119810581207275, "learning_rate": 1.191548489506128e-07, "loss": 1.454, "step": 47493 }, { "epoch": 3.722100313479624, "grad_norm": 3.3740110397338867, "learning_rate": 1.1908794009619972e-07, "loss": 1.534, "step": 47494 }, { "epoch": 3.7221786833855797, "grad_norm": 7.178184986114502, "learning_rate": 1.1902104980626828e-07, "loss": 2.1148, "step": 47495 }, { "epoch": 3.722257053291536, "grad_norm": 4.474947929382324, "learning_rate": 1.1895417808107435e-07, "loss": 1.4404, "step": 47496 }, { "epoch": 3.7223354231974923, "grad_norm": 8.506802558898926, "learning_rate": 1.1888732492087107e-07, "loss": 0.4501, "step": 47497 }, { "epoch": 3.722413793103448, "grad_norm": 5.915032386779785, "learning_rate": 1.1882049032591381e-07, "loss": 1.1928, "step": 47498 }, { "epoch": 3.7224921630094046, "grad_norm": 8.744540214538574, "learning_rate": 1.1875367429645624e-07, "loss": 1.1355, "step": 47499 }, { "epoch": 3.7225705329153604, "grad_norm": 3.1912732124328613, "learning_rate": 1.1868687683275259e-07, "loss": 1.403, "step": 47500 }, { "epoch": 3.7226489028213168, "grad_norm": 4.389920711517334, "learning_rate": 1.1862009793505713e-07, "loss": 1.2351, "step": 47501 }, { "epoch": 3.7227272727272727, "grad_norm": 3.428731918334961, "learning_rate": 1.1855333760362242e-07, "loss": 1.4424, "step": 47502 }, { "epoch": 3.722805642633229, "grad_norm": 6.18564510345459, "learning_rate": 1.1848659583870437e-07, "loss": 1.5853, "step": 47503 }, { "epoch": 3.722884012539185, "grad_norm": 4.310731410980225, "learning_rate": 1.184198726405561e-07, "loss": 1.1141, "step": 47504 }, { "epoch": 3.722962382445141, "grad_norm": 3.3112618923187256, "learning_rate": 1.183531680094313e-07, "loss": 0.7252, "step": 47505 }, { "epoch": 3.723040752351097, "grad_norm": 3.091900587081909, "learning_rate": 1.1828648194558367e-07, "loss": 1.5047, "step": 47506 }, { "epoch": 3.7231191222570534, "grad_norm": 4.4370551109313965, "learning_rate": 1.1821981444926578e-07, "loss": 1.1588, "step": 47507 }, { "epoch": 3.7231974921630093, "grad_norm": 2.5020079612731934, "learning_rate": 1.1815316552073353e-07, "loss": 1.3297, "step": 47508 }, { "epoch": 3.7232758620689657, "grad_norm": 5.28539514541626, "learning_rate": 1.1808653516023894e-07, "loss": 1.8031, "step": 47509 }, { "epoch": 3.7233542319749215, "grad_norm": 3.006258726119995, "learning_rate": 1.1801992336803569e-07, "loss": 1.4562, "step": 47510 }, { "epoch": 3.723432601880878, "grad_norm": 9.46231746673584, "learning_rate": 1.1795333014437693e-07, "loss": 1.2042, "step": 47511 }, { "epoch": 3.7235109717868338, "grad_norm": 3.61687970161438, "learning_rate": 1.1788675548951579e-07, "loss": 1.1794, "step": 47512 }, { "epoch": 3.72358934169279, "grad_norm": 5.323662281036377, "learning_rate": 1.1782019940370648e-07, "loss": 1.4269, "step": 47513 }, { "epoch": 3.723667711598746, "grad_norm": 2.5954060554504395, "learning_rate": 1.1775366188720105e-07, "loss": 1.2355, "step": 47514 }, { "epoch": 3.7237460815047023, "grad_norm": 2.2369532585144043, "learning_rate": 1.1768714294025374e-07, "loss": 1.0894, "step": 47515 }, { "epoch": 3.723824451410658, "grad_norm": 18.329938888549805, "learning_rate": 1.1762064256311545e-07, "loss": 1.2196, "step": 47516 }, { "epoch": 3.7239028213166145, "grad_norm": 3.674541711807251, "learning_rate": 1.1755416075604154e-07, "loss": 1.4436, "step": 47517 }, { "epoch": 3.7239811912225704, "grad_norm": 2.954672336578369, "learning_rate": 1.1748769751928346e-07, "loss": 1.3605, "step": 47518 }, { "epoch": 3.7240595611285268, "grad_norm": 3.0162038803100586, "learning_rate": 1.1742125285309491e-07, "loss": 1.2939, "step": 47519 }, { "epoch": 3.7241379310344827, "grad_norm": 2.8314778804779053, "learning_rate": 1.1735482675772736e-07, "loss": 1.1855, "step": 47520 }, { "epoch": 3.724216300940439, "grad_norm": 3.782456874847412, "learning_rate": 1.1728841923343392e-07, "loss": 1.2519, "step": 47521 }, { "epoch": 3.724294670846395, "grad_norm": 12.954163551330566, "learning_rate": 1.1722203028046774e-07, "loss": 1.6002, "step": 47522 }, { "epoch": 3.724373040752351, "grad_norm": 9.677799224853516, "learning_rate": 1.1715565989908196e-07, "loss": 1.7316, "step": 47523 }, { "epoch": 3.724451410658307, "grad_norm": 2.817013740539551, "learning_rate": 1.170893080895269e-07, "loss": 1.0, "step": 47524 }, { "epoch": 3.7245297805642634, "grad_norm": 3.6084611415863037, "learning_rate": 1.1702297485205628e-07, "loss": 1.1204, "step": 47525 }, { "epoch": 3.7246081504702193, "grad_norm": 2.5749030113220215, "learning_rate": 1.1695666018692264e-07, "loss": 1.3554, "step": 47526 }, { "epoch": 3.7246865203761756, "grad_norm": 2.8704800605773926, "learning_rate": 1.1689036409437749e-07, "loss": 1.3374, "step": 47527 }, { "epoch": 3.7247648902821315, "grad_norm": 2.2713088989257812, "learning_rate": 1.1682408657467336e-07, "loss": 1.3875, "step": 47528 }, { "epoch": 3.724843260188088, "grad_norm": 4.937456130981445, "learning_rate": 1.1675782762806176e-07, "loss": 1.2025, "step": 47529 }, { "epoch": 3.7249216300940438, "grad_norm": 5.680593967437744, "learning_rate": 1.1669158725479579e-07, "loss": 1.3956, "step": 47530 }, { "epoch": 3.725, "grad_norm": 4.879886150360107, "learning_rate": 1.1662536545512637e-07, "loss": 1.229, "step": 47531 }, { "epoch": 3.725078369905956, "grad_norm": 2.1644487380981445, "learning_rate": 1.1655916222930608e-07, "loss": 1.4648, "step": 47532 }, { "epoch": 3.7251567398119123, "grad_norm": 5.92627477645874, "learning_rate": 1.1649297757758637e-07, "loss": 0.9792, "step": 47533 }, { "epoch": 3.725235109717868, "grad_norm": 2.135035991668701, "learning_rate": 1.1642681150021984e-07, "loss": 0.8583, "step": 47534 }, { "epoch": 3.7253134796238245, "grad_norm": 2.841909885406494, "learning_rate": 1.1636066399745572e-07, "loss": 1.2376, "step": 47535 }, { "epoch": 3.7253918495297804, "grad_norm": 2.1865622997283936, "learning_rate": 1.1629453506954824e-07, "loss": 1.2562, "step": 47536 }, { "epoch": 3.7254702194357368, "grad_norm": 5.13662052154541, "learning_rate": 1.1622842471674778e-07, "loss": 1.6164, "step": 47537 }, { "epoch": 3.7255485893416926, "grad_norm": 2.5242531299591064, "learning_rate": 1.1616233293930634e-07, "loss": 1.1359, "step": 47538 }, { "epoch": 3.725626959247649, "grad_norm": 3.258488178253174, "learning_rate": 1.1609625973747484e-07, "loss": 1.2425, "step": 47539 }, { "epoch": 3.725705329153605, "grad_norm": 2.4637975692749023, "learning_rate": 1.1603020511150365e-07, "loss": 1.3275, "step": 47540 }, { "epoch": 3.725783699059561, "grad_norm": 10.087869644165039, "learning_rate": 1.1596416906164588e-07, "loss": 0.6657, "step": 47541 }, { "epoch": 3.725862068965517, "grad_norm": 7.302631378173828, "learning_rate": 1.1589815158815187e-07, "loss": 1.0287, "step": 47542 }, { "epoch": 3.7259404388714734, "grad_norm": 2.221649408340454, "learning_rate": 1.1583215269127202e-07, "loss": 1.4159, "step": 47543 }, { "epoch": 3.7260188087774293, "grad_norm": 3.06040096282959, "learning_rate": 1.1576617237125831e-07, "loss": 1.2219, "step": 47544 }, { "epoch": 3.7260971786833856, "grad_norm": 3.069944381713867, "learning_rate": 1.1570021062836168e-07, "loss": 1.1789, "step": 47545 }, { "epoch": 3.7261755485893415, "grad_norm": 3.3290677070617676, "learning_rate": 1.1563426746283246e-07, "loss": 1.1656, "step": 47546 }, { "epoch": 3.726253918495298, "grad_norm": 3.3018016815185547, "learning_rate": 1.1556834287492158e-07, "loss": 1.2842, "step": 47547 }, { "epoch": 3.726332288401254, "grad_norm": 5.80511474609375, "learning_rate": 1.1550243686487994e-07, "loss": 1.9051, "step": 47548 }, { "epoch": 3.72641065830721, "grad_norm": 3.067207098007202, "learning_rate": 1.154365494329579e-07, "loss": 1.3093, "step": 47549 }, { "epoch": 3.726489028213166, "grad_norm": 3.3191699981689453, "learning_rate": 1.1537068057940693e-07, "loss": 1.2408, "step": 47550 }, { "epoch": 3.7265673981191223, "grad_norm": 5.001453399658203, "learning_rate": 1.1530483030447681e-07, "loss": 1.2958, "step": 47551 }, { "epoch": 3.7266457680250786, "grad_norm": 2.784039258956909, "learning_rate": 1.1523899860841791e-07, "loss": 1.2018, "step": 47552 }, { "epoch": 3.7267241379310345, "grad_norm": 3.400897264480591, "learning_rate": 1.1517318549148171e-07, "loss": 1.4795, "step": 47553 }, { "epoch": 3.7268025078369904, "grad_norm": 5.002458572387695, "learning_rate": 1.1510739095391687e-07, "loss": 1.4639, "step": 47554 }, { "epoch": 3.7268808777429467, "grad_norm": 2.9868924617767334, "learning_rate": 1.1504161499597433e-07, "loss": 1.2265, "step": 47555 }, { "epoch": 3.726959247648903, "grad_norm": 9.144777297973633, "learning_rate": 1.1497585761790553e-07, "loss": 1.4358, "step": 47556 }, { "epoch": 3.727037617554859, "grad_norm": 3.3598618507385254, "learning_rate": 1.1491011881995862e-07, "loss": 1.7346, "step": 47557 }, { "epoch": 3.727115987460815, "grad_norm": 3.1366114616394043, "learning_rate": 1.148443986023845e-07, "loss": 1.2486, "step": 47558 }, { "epoch": 3.727194357366771, "grad_norm": 4.2337327003479, "learning_rate": 1.1477869696543298e-07, "loss": 0.8418, "step": 47559 }, { "epoch": 3.7272727272727275, "grad_norm": 10.669195175170898, "learning_rate": 1.1471301390935497e-07, "loss": 1.7977, "step": 47560 }, { "epoch": 3.7273510971786834, "grad_norm": 5.6337151527404785, "learning_rate": 1.1464734943439859e-07, "loss": 1.506, "step": 47561 }, { "epoch": 3.7274294670846393, "grad_norm": 2.501677989959717, "learning_rate": 1.1458170354081477e-07, "loss": 1.4995, "step": 47562 }, { "epoch": 3.7275078369905956, "grad_norm": 3.248779773712158, "learning_rate": 1.1451607622885218e-07, "loss": 1.4174, "step": 47563 }, { "epoch": 3.727586206896552, "grad_norm": 3.9715371131896973, "learning_rate": 1.1445046749876177e-07, "loss": 1.0698, "step": 47564 }, { "epoch": 3.727664576802508, "grad_norm": 4.50262451171875, "learning_rate": 1.1438487735079218e-07, "loss": 1.4928, "step": 47565 }, { "epoch": 3.7277429467084637, "grad_norm": 4.9997711181640625, "learning_rate": 1.1431930578519323e-07, "loss": 1.5887, "step": 47566 }, { "epoch": 3.72782131661442, "grad_norm": 3.19566011428833, "learning_rate": 1.1425375280221418e-07, "loss": 1.3234, "step": 47567 }, { "epoch": 3.7278996865203764, "grad_norm": 3.772317409515381, "learning_rate": 1.1418821840210426e-07, "loss": 1.0454, "step": 47568 }, { "epoch": 3.7279780564263323, "grad_norm": 4.106782913208008, "learning_rate": 1.141227025851127e-07, "loss": 1.4847, "step": 47569 }, { "epoch": 3.728056426332288, "grad_norm": 2.6002211570739746, "learning_rate": 1.1405720535148935e-07, "loss": 1.4605, "step": 47570 }, { "epoch": 3.7281347962382445, "grad_norm": 3.915621042251587, "learning_rate": 1.1399172670148285e-07, "loss": 1.3523, "step": 47571 }, { "epoch": 3.728213166144201, "grad_norm": 4.710450649261475, "learning_rate": 1.1392626663534134e-07, "loss": 1.4634, "step": 47572 }, { "epoch": 3.7282915360501567, "grad_norm": 15.330711364746094, "learning_rate": 1.138608251533152e-07, "loss": 1.2613, "step": 47573 }, { "epoch": 3.7283699059561126, "grad_norm": 2.627809524536133, "learning_rate": 1.1379540225565255e-07, "loss": 1.4676, "step": 47574 }, { "epoch": 3.728448275862069, "grad_norm": 3.934116840362549, "learning_rate": 1.137299979426032e-07, "loss": 1.5224, "step": 47575 }, { "epoch": 3.7285266457680253, "grad_norm": 4.897432804107666, "learning_rate": 1.1366461221441472e-07, "loss": 1.3446, "step": 47576 }, { "epoch": 3.728605015673981, "grad_norm": 3.58166766166687, "learning_rate": 1.1359924507133635e-07, "loss": 1.2938, "step": 47577 }, { "epoch": 3.728683385579937, "grad_norm": 4.1863603591918945, "learning_rate": 1.1353389651361568e-07, "loss": 1.384, "step": 47578 }, { "epoch": 3.7287617554858934, "grad_norm": 5.462508678436279, "learning_rate": 1.1346856654150362e-07, "loss": 1.242, "step": 47579 }, { "epoch": 3.7288401253918497, "grad_norm": 2.398324728012085, "learning_rate": 1.1340325515524665e-07, "loss": 1.221, "step": 47580 }, { "epoch": 3.7289184952978056, "grad_norm": 7.625650882720947, "learning_rate": 1.1333796235509397e-07, "loss": 1.6802, "step": 47581 }, { "epoch": 3.7289968652037615, "grad_norm": 3.579129219055176, "learning_rate": 1.1327268814129377e-07, "loss": 1.2328, "step": 47582 }, { "epoch": 3.729075235109718, "grad_norm": 3.4270870685577393, "learning_rate": 1.1320743251409416e-07, "loss": 0.9516, "step": 47583 }, { "epoch": 3.729153605015674, "grad_norm": 2.9496819972991943, "learning_rate": 1.1314219547374328e-07, "loss": 1.5565, "step": 47584 }, { "epoch": 3.72923197492163, "grad_norm": 2.485060691833496, "learning_rate": 1.1307697702048925e-07, "loss": 1.2003, "step": 47585 }, { "epoch": 3.729310344827586, "grad_norm": 4.140800952911377, "learning_rate": 1.1301177715458078e-07, "loss": 1.3526, "step": 47586 }, { "epoch": 3.7293887147335423, "grad_norm": 5.675005912780762, "learning_rate": 1.1294659587626544e-07, "loss": 0.7902, "step": 47587 }, { "epoch": 3.7294670846394986, "grad_norm": 3.128981113433838, "learning_rate": 1.1288143318579081e-07, "loss": 1.1805, "step": 47588 }, { "epoch": 3.7295454545454545, "grad_norm": 2.909552574157715, "learning_rate": 1.1281628908340503e-07, "loss": 1.2041, "step": 47589 }, { "epoch": 3.7296238244514104, "grad_norm": 2.6543703079223633, "learning_rate": 1.1275116356935622e-07, "loss": 1.1318, "step": 47590 }, { "epoch": 3.7297021943573667, "grad_norm": 2.723353147506714, "learning_rate": 1.1268605664389143e-07, "loss": 1.3665, "step": 47591 }, { "epoch": 3.729780564263323, "grad_norm": 2.411083698272705, "learning_rate": 1.1262096830725821e-07, "loss": 1.4966, "step": 47592 }, { "epoch": 3.729858934169279, "grad_norm": 4.007959365844727, "learning_rate": 1.1255589855970528e-07, "loss": 1.1134, "step": 47593 }, { "epoch": 3.7299373040752353, "grad_norm": 3.0276432037353516, "learning_rate": 1.1249084740147908e-07, "loss": 1.7768, "step": 47594 }, { "epoch": 3.730015673981191, "grad_norm": 3.1554722785949707, "learning_rate": 1.1242581483282777e-07, "loss": 1.1478, "step": 47595 }, { "epoch": 3.7300940438871475, "grad_norm": 7.219456195831299, "learning_rate": 1.1236080085399725e-07, "loss": 1.2221, "step": 47596 }, { "epoch": 3.7301724137931034, "grad_norm": 3.334995746612549, "learning_rate": 1.1229580546523622e-07, "loss": 1.4014, "step": 47597 }, { "epoch": 3.7302507836990597, "grad_norm": 4.573612213134766, "learning_rate": 1.1223082866679169e-07, "loss": 1.5197, "step": 47598 }, { "epoch": 3.7303291536050156, "grad_norm": 2.934969663619995, "learning_rate": 1.121658704589107e-07, "loss": 1.184, "step": 47599 }, { "epoch": 3.730407523510972, "grad_norm": 4.018508434295654, "learning_rate": 1.1210093084183971e-07, "loss": 1.2665, "step": 47600 }, { "epoch": 3.730485893416928, "grad_norm": 7.997124195098877, "learning_rate": 1.120360098158263e-07, "loss": 0.9752, "step": 47601 }, { "epoch": 3.730564263322884, "grad_norm": 3.7378053665161133, "learning_rate": 1.1197110738111749e-07, "loss": 1.3768, "step": 47602 }, { "epoch": 3.73064263322884, "grad_norm": 4.329922676086426, "learning_rate": 1.1190622353795977e-07, "loss": 1.2124, "step": 47603 }, { "epoch": 3.7307210031347964, "grad_norm": 2.1223628520965576, "learning_rate": 1.1184135828660014e-07, "loss": 1.1351, "step": 47604 }, { "epoch": 3.7307993730407523, "grad_norm": 3.713773488998413, "learning_rate": 1.117765116272851e-07, "loss": 1.3681, "step": 47605 }, { "epoch": 3.7308777429467086, "grad_norm": 3.3912668228149414, "learning_rate": 1.1171168356026163e-07, "loss": 1.011, "step": 47606 }, { "epoch": 3.7309561128526645, "grad_norm": 4.441689491271973, "learning_rate": 1.1164687408577623e-07, "loss": 0.9998, "step": 47607 }, { "epoch": 3.731034482758621, "grad_norm": 5.780837059020996, "learning_rate": 1.1158208320407538e-07, "loss": 1.7373, "step": 47608 }, { "epoch": 3.7311128526645767, "grad_norm": 5.305759906768799, "learning_rate": 1.1151731091540552e-07, "loss": 0.5238, "step": 47609 }, { "epoch": 3.731191222570533, "grad_norm": 3.7531185150146484, "learning_rate": 1.1145255722001202e-07, "loss": 1.6847, "step": 47610 }, { "epoch": 3.731269592476489, "grad_norm": 3.1694624423980713, "learning_rate": 1.1138782211814302e-07, "loss": 1.3426, "step": 47611 }, { "epoch": 3.7313479623824453, "grad_norm": 2.2897543907165527, "learning_rate": 1.113231056100439e-07, "loss": 1.3787, "step": 47612 }, { "epoch": 3.731426332288401, "grad_norm": 3.8745639324188232, "learning_rate": 1.1125840769595997e-07, "loss": 1.3329, "step": 47613 }, { "epoch": 3.7315047021943575, "grad_norm": 4.395549297332764, "learning_rate": 1.1119372837613828e-07, "loss": 1.4726, "step": 47614 }, { "epoch": 3.7315830721003134, "grad_norm": 3.121358633041382, "learning_rate": 1.111290676508242e-07, "loss": 0.9423, "step": 47615 }, { "epoch": 3.7316614420062697, "grad_norm": 2.864100217819214, "learning_rate": 1.110644255202653e-07, "loss": 1.2045, "step": 47616 }, { "epoch": 3.7317398119122256, "grad_norm": 3.596268892288208, "learning_rate": 1.1099980198470528e-07, "loss": 1.3758, "step": 47617 }, { "epoch": 3.731818181818182, "grad_norm": 3.390713691711426, "learning_rate": 1.1093519704439115e-07, "loss": 1.5387, "step": 47618 }, { "epoch": 3.731896551724138, "grad_norm": 7.559104919433594, "learning_rate": 1.1087061069956829e-07, "loss": 1.2539, "step": 47619 }, { "epoch": 3.731974921630094, "grad_norm": 6.13083028793335, "learning_rate": 1.1080604295048203e-07, "loss": 1.7378, "step": 47620 }, { "epoch": 3.73205329153605, "grad_norm": 4.631078720092773, "learning_rate": 1.1074149379737886e-07, "loss": 1.3292, "step": 47621 }, { "epoch": 3.7321316614420064, "grad_norm": 3.8765389919281006, "learning_rate": 1.1067696324050359e-07, "loss": 0.6873, "step": 47622 }, { "epoch": 3.7322100313479623, "grad_norm": 2.7127132415771484, "learning_rate": 1.1061245128010212e-07, "loss": 1.1265, "step": 47623 }, { "epoch": 3.7322884012539186, "grad_norm": 3.7519445419311523, "learning_rate": 1.105479579164187e-07, "loss": 1.2904, "step": 47624 }, { "epoch": 3.7323667711598745, "grad_norm": 3.4861366748809814, "learning_rate": 1.1048348314970037e-07, "loss": 1.2197, "step": 47625 }, { "epoch": 3.732445141065831, "grad_norm": 2.2862415313720703, "learning_rate": 1.1041902698019081e-07, "loss": 1.507, "step": 47626 }, { "epoch": 3.7325235109717867, "grad_norm": 5.3470282554626465, "learning_rate": 1.1035458940813648e-07, "loss": 1.2034, "step": 47627 }, { "epoch": 3.732601880877743, "grad_norm": 9.51655101776123, "learning_rate": 1.1029017043378165e-07, "loss": 1.122, "step": 47628 }, { "epoch": 3.732680250783699, "grad_norm": 2.5943517684936523, "learning_rate": 1.1022577005737056e-07, "loss": 1.6094, "step": 47629 }, { "epoch": 3.7327586206896552, "grad_norm": 5.534881114959717, "learning_rate": 1.1016138827915024e-07, "loss": 0.9685, "step": 47630 }, { "epoch": 3.732836990595611, "grad_norm": 3.1449925899505615, "learning_rate": 1.1009702509936437e-07, "loss": 1.3653, "step": 47631 }, { "epoch": 3.7329153605015675, "grad_norm": 4.393145561218262, "learning_rate": 1.1003268051825722e-07, "loss": 1.4919, "step": 47632 }, { "epoch": 3.7329937304075234, "grad_norm": 3.0137667655944824, "learning_rate": 1.0996835453607413e-07, "loss": 1.2323, "step": 47633 }, { "epoch": 3.7330721003134797, "grad_norm": 4.317798614501953, "learning_rate": 1.0990404715305991e-07, "loss": 1.937, "step": 47634 }, { "epoch": 3.7331504702194356, "grad_norm": 4.497717380523682, "learning_rate": 1.0983975836945937e-07, "loss": 0.7603, "step": 47635 }, { "epoch": 3.733228840125392, "grad_norm": 4.235239028930664, "learning_rate": 1.097754881855162e-07, "loss": 1.17, "step": 47636 }, { "epoch": 3.733307210031348, "grad_norm": 4.593652248382568, "learning_rate": 1.097112366014752e-07, "loss": 1.4575, "step": 47637 }, { "epoch": 3.733385579937304, "grad_norm": 2.569683790206909, "learning_rate": 1.0964700361758118e-07, "loss": 0.8525, "step": 47638 }, { "epoch": 3.7334639498432605, "grad_norm": 4.256230354309082, "learning_rate": 1.0958278923407783e-07, "loss": 1.4401, "step": 47639 }, { "epoch": 3.7335423197492164, "grad_norm": 2.773097515106201, "learning_rate": 1.0951859345120996e-07, "loss": 1.2276, "step": 47640 }, { "epoch": 3.7336206896551722, "grad_norm": 3.9898509979248047, "learning_rate": 1.0945441626922127e-07, "loss": 1.235, "step": 47641 }, { "epoch": 3.7336990595611286, "grad_norm": 2.857142448425293, "learning_rate": 1.0939025768835654e-07, "loss": 1.4627, "step": 47642 }, { "epoch": 3.733777429467085, "grad_norm": 5.046521186828613, "learning_rate": 1.0932611770885837e-07, "loss": 1.3475, "step": 47643 }, { "epoch": 3.733855799373041, "grad_norm": 2.2920258045196533, "learning_rate": 1.0926199633097156e-07, "loss": 1.2617, "step": 47644 }, { "epoch": 3.7339341692789967, "grad_norm": 3.2365152835845947, "learning_rate": 1.0919789355494092e-07, "loss": 1.3155, "step": 47645 }, { "epoch": 3.734012539184953, "grad_norm": 2.402482748031616, "learning_rate": 1.0913380938100959e-07, "loss": 1.1925, "step": 47646 }, { "epoch": 3.7340909090909093, "grad_norm": 2.5961716175079346, "learning_rate": 1.090697438094207e-07, "loss": 1.2328, "step": 47647 }, { "epoch": 3.7341692789968652, "grad_norm": 8.489633560180664, "learning_rate": 1.090056968404174e-07, "loss": 0.8623, "step": 47648 }, { "epoch": 3.734247648902821, "grad_norm": 4.069591999053955, "learning_rate": 1.0894166847424559e-07, "loss": 1.2496, "step": 47649 }, { "epoch": 3.7343260188087775, "grad_norm": 3.991755962371826, "learning_rate": 1.0887765871114731e-07, "loss": 1.4984, "step": 47650 }, { "epoch": 3.734404388714734, "grad_norm": 3.1151132583618164, "learning_rate": 1.0881366755136569e-07, "loss": 1.3164, "step": 47651 }, { "epoch": 3.7344827586206897, "grad_norm": 2.893051862716675, "learning_rate": 1.0874969499514499e-07, "loss": 1.2415, "step": 47652 }, { "epoch": 3.7345611285266456, "grad_norm": 3.1313095092773438, "learning_rate": 1.0868574104272778e-07, "loss": 1.4107, "step": 47653 }, { "epoch": 3.734639498432602, "grad_norm": 2.1295578479766846, "learning_rate": 1.0862180569435776e-07, "loss": 1.1115, "step": 47654 }, { "epoch": 3.7347178683385582, "grad_norm": 3.671300172805786, "learning_rate": 1.0855788895027807e-07, "loss": 1.1605, "step": 47655 }, { "epoch": 3.734796238244514, "grad_norm": 10.458133697509766, "learning_rate": 1.0849399081073186e-07, "loss": 1.1471, "step": 47656 }, { "epoch": 3.73487460815047, "grad_norm": 2.6445906162261963, "learning_rate": 1.0843011127596225e-07, "loss": 1.504, "step": 47657 }, { "epoch": 3.7349529780564263, "grad_norm": 2.895369529724121, "learning_rate": 1.0836625034621184e-07, "loss": 1.6472, "step": 47658 }, { "epoch": 3.7350313479623827, "grad_norm": 2.482952833175659, "learning_rate": 1.0830240802172376e-07, "loss": 1.6493, "step": 47659 }, { "epoch": 3.7351097178683386, "grad_norm": 11.123885154724121, "learning_rate": 1.082385843027406e-07, "loss": 1.2858, "step": 47660 }, { "epoch": 3.7351880877742945, "grad_norm": 2.5433290004730225, "learning_rate": 1.0817477918950603e-07, "loss": 1.2491, "step": 47661 }, { "epoch": 3.735266457680251, "grad_norm": 3.3528952598571777, "learning_rate": 1.0811099268226044e-07, "loss": 0.9863, "step": 47662 }, { "epoch": 3.735344827586207, "grad_norm": 3.9071109294891357, "learning_rate": 1.0804722478124919e-07, "loss": 1.478, "step": 47663 }, { "epoch": 3.735423197492163, "grad_norm": 6.454999923706055, "learning_rate": 1.0798347548671318e-07, "loss": 2.2315, "step": 47664 }, { "epoch": 3.735501567398119, "grad_norm": 3.115365505218506, "learning_rate": 1.079197447988961e-07, "loss": 1.6869, "step": 47665 }, { "epoch": 3.7355799373040752, "grad_norm": 4.609677791595459, "learning_rate": 1.078560327180389e-07, "loss": 1.3791, "step": 47666 }, { "epoch": 3.7356583072100316, "grad_norm": 3.9367268085479736, "learning_rate": 1.0779233924438415e-07, "loss": 1.5688, "step": 47667 }, { "epoch": 3.7357366771159874, "grad_norm": 2.1249213218688965, "learning_rate": 1.0772866437817552e-07, "loss": 1.3023, "step": 47668 }, { "epoch": 3.7358150470219433, "grad_norm": 3.4170334339141846, "learning_rate": 1.0766500811965397e-07, "loss": 1.4183, "step": 47669 }, { "epoch": 3.7358934169278997, "grad_norm": 8.312361717224121, "learning_rate": 1.0760137046906149e-07, "loss": 1.1998, "step": 47670 }, { "epoch": 3.735971786833856, "grad_norm": 2.9027328491210938, "learning_rate": 1.0753775142664069e-07, "loss": 1.2976, "step": 47671 }, { "epoch": 3.736050156739812, "grad_norm": 6.654986381530762, "learning_rate": 1.0747415099263358e-07, "loss": 1.7097, "step": 47672 }, { "epoch": 3.7361285266457678, "grad_norm": 2.4368879795074463, "learning_rate": 1.0741056916728165e-07, "loss": 1.5726, "step": 47673 }, { "epoch": 3.736206896551724, "grad_norm": 4.921407699584961, "learning_rate": 1.0734700595082692e-07, "loss": 1.5912, "step": 47674 }, { "epoch": 3.7362852664576804, "grad_norm": 3.1312930583953857, "learning_rate": 1.0728346134351086e-07, "loss": 0.8153, "step": 47675 }, { "epoch": 3.7363636363636363, "grad_norm": 4.0224609375, "learning_rate": 1.0721993534557607e-07, "loss": 1.7219, "step": 47676 }, { "epoch": 3.736442006269592, "grad_norm": 4.878393173217773, "learning_rate": 1.0715642795726289e-07, "loss": 1.1785, "step": 47677 }, { "epoch": 3.7365203761755486, "grad_norm": 2.624959945678711, "learning_rate": 1.0709293917881392e-07, "loss": 1.2717, "step": 47678 }, { "epoch": 3.736598746081505, "grad_norm": 3.336055040359497, "learning_rate": 1.0702946901047062e-07, "loss": 1.3121, "step": 47679 }, { "epoch": 3.7366771159874608, "grad_norm": 2.9034230709075928, "learning_rate": 1.0696601745247337e-07, "loss": 1.2878, "step": 47680 }, { "epoch": 3.7367554858934167, "grad_norm": 2.698153495788574, "learning_rate": 1.0690258450506419e-07, "loss": 1.5878, "step": 47681 }, { "epoch": 3.736833855799373, "grad_norm": 5.955268859863281, "learning_rate": 1.0683917016848456e-07, "loss": 1.221, "step": 47682 }, { "epoch": 3.7369122257053293, "grad_norm": 4.716330528259277, "learning_rate": 1.0677577444297593e-07, "loss": 1.5562, "step": 47683 }, { "epoch": 3.736990595611285, "grad_norm": 3.279343366622925, "learning_rate": 1.0671239732877814e-07, "loss": 1.268, "step": 47684 }, { "epoch": 3.737068965517241, "grad_norm": 1.8860024213790894, "learning_rate": 1.066490388261332e-07, "loss": 1.1725, "step": 47685 }, { "epoch": 3.7371473354231974, "grad_norm": 3.3007864952087402, "learning_rate": 1.0658569893528203e-07, "loss": 1.736, "step": 47686 }, { "epoch": 3.7372257053291538, "grad_norm": 2.635136127471924, "learning_rate": 1.0652237765646555e-07, "loss": 1.5244, "step": 47687 }, { "epoch": 3.7373040752351097, "grad_norm": 4.786369800567627, "learning_rate": 1.064590749899247e-07, "loss": 1.7895, "step": 47688 }, { "epoch": 3.737382445141066, "grad_norm": 5.650163173675537, "learning_rate": 1.0639579093589981e-07, "loss": 1.2058, "step": 47689 }, { "epoch": 3.737460815047022, "grad_norm": 4.636154651641846, "learning_rate": 1.063325254946318e-07, "loss": 1.6706, "step": 47690 }, { "epoch": 3.737539184952978, "grad_norm": 3.800672769546509, "learning_rate": 1.0626927866636106e-07, "loss": 1.7834, "step": 47691 }, { "epoch": 3.737617554858934, "grad_norm": 14.846185684204102, "learning_rate": 1.0620605045132848e-07, "loss": 1.9559, "step": 47692 }, { "epoch": 3.7376959247648904, "grad_norm": 7.8488874435424805, "learning_rate": 1.0614284084977499e-07, "loss": 0.9394, "step": 47693 }, { "epoch": 3.7377742946708463, "grad_norm": 2.736356496810913, "learning_rate": 1.0607964986193986e-07, "loss": 1.1401, "step": 47694 }, { "epoch": 3.7378526645768027, "grad_norm": 3.7856767177581787, "learning_rate": 1.0601647748806454e-07, "loss": 1.0439, "step": 47695 }, { "epoch": 3.7379310344827585, "grad_norm": 2.9683163166046143, "learning_rate": 1.0595332372838829e-07, "loss": 1.6049, "step": 47696 }, { "epoch": 3.738009404388715, "grad_norm": 4.851001262664795, "learning_rate": 1.0589018858315259e-07, "loss": 0.8382, "step": 47697 }, { "epoch": 3.7380877742946708, "grad_norm": 3.082618236541748, "learning_rate": 1.0582707205259668e-07, "loss": 1.4343, "step": 47698 }, { "epoch": 3.738166144200627, "grad_norm": 3.4075276851654053, "learning_rate": 1.0576397413695982e-07, "loss": 1.4731, "step": 47699 }, { "epoch": 3.738244514106583, "grad_norm": 2.2198641300201416, "learning_rate": 1.0570089483648405e-07, "loss": 1.1149, "step": 47700 }, { "epoch": 3.7383228840125393, "grad_norm": 2.353616714477539, "learning_rate": 1.0563783415140749e-07, "loss": 1.2716, "step": 47701 }, { "epoch": 3.738401253918495, "grad_norm": 8.123100280761719, "learning_rate": 1.0557479208197163e-07, "loss": 1.391, "step": 47702 }, { "epoch": 3.7384796238244515, "grad_norm": 2.8230624198913574, "learning_rate": 1.0551176862841461e-07, "loss": 1.4387, "step": 47703 }, { "epoch": 3.7385579937304074, "grad_norm": 6.229251384735107, "learning_rate": 1.0544876379097679e-07, "loss": 1.3723, "step": 47704 }, { "epoch": 3.7386363636363638, "grad_norm": 4.041306018829346, "learning_rate": 1.0538577756989798e-07, "loss": 1.2417, "step": 47705 }, { "epoch": 3.7387147335423196, "grad_norm": 6.382256031036377, "learning_rate": 1.05322809965418e-07, "loss": 1.2496, "step": 47706 }, { "epoch": 3.738793103448276, "grad_norm": 3.289201259613037, "learning_rate": 1.0525986097777607e-07, "loss": 1.5292, "step": 47707 }, { "epoch": 3.738871473354232, "grad_norm": 4.921347618103027, "learning_rate": 1.0519693060721092e-07, "loss": 1.3774, "step": 47708 }, { "epoch": 3.738949843260188, "grad_norm": 3.0817983150482178, "learning_rate": 1.051340188539629e-07, "loss": 1.3784, "step": 47709 }, { "epoch": 3.739028213166144, "grad_norm": 3.0248992443084717, "learning_rate": 1.0507112571827072e-07, "loss": 1.1844, "step": 47710 }, { "epoch": 3.7391065830721004, "grad_norm": 4.604699611663818, "learning_rate": 1.0500825120037362e-07, "loss": 0.903, "step": 47711 }, { "epoch": 3.7391849529780563, "grad_norm": 3.7467544078826904, "learning_rate": 1.049453953005114e-07, "loss": 1.3193, "step": 47712 }, { "epoch": 3.7392633228840126, "grad_norm": 12.258084297180176, "learning_rate": 1.0488255801892222e-07, "loss": 1.5454, "step": 47713 }, { "epoch": 3.7393416927899685, "grad_norm": 2.3254146575927734, "learning_rate": 1.0481973935584589e-07, "loss": 1.1179, "step": 47714 }, { "epoch": 3.739420062695925, "grad_norm": 5.50673770904541, "learning_rate": 1.0475693931152054e-07, "loss": 1.2046, "step": 47715 }, { "epoch": 3.7394984326018808, "grad_norm": 4.263731002807617, "learning_rate": 1.0469415788618542e-07, "loss": 1.4765, "step": 47716 }, { "epoch": 3.739576802507837, "grad_norm": 2.2220945358276367, "learning_rate": 1.0463139508008036e-07, "loss": 1.2961, "step": 47717 }, { "epoch": 3.739655172413793, "grad_norm": 5.008803367614746, "learning_rate": 1.0456865089344126e-07, "loss": 1.2986, "step": 47718 }, { "epoch": 3.7397335423197493, "grad_norm": 4.087337017059326, "learning_rate": 1.045059253265096e-07, "loss": 1.6503, "step": 47719 }, { "epoch": 3.739811912225705, "grad_norm": 6.848243713378906, "learning_rate": 1.0444321837952299e-07, "loss": 1.1339, "step": 47720 }, { "epoch": 3.7398902821316615, "grad_norm": 3.8176262378692627, "learning_rate": 1.0438053005272009e-07, "loss": 1.3982, "step": 47721 }, { "epoch": 3.7399686520376174, "grad_norm": 2.736762046813965, "learning_rate": 1.043178603463385e-07, "loss": 1.3909, "step": 47722 }, { "epoch": 3.7400470219435737, "grad_norm": 4.027814865112305, "learning_rate": 1.0425520926061694e-07, "loss": 1.6161, "step": 47723 }, { "epoch": 3.7401253918495296, "grad_norm": 2.8397650718688965, "learning_rate": 1.0419257679579519e-07, "loss": 1.3442, "step": 47724 }, { "epoch": 3.740203761755486, "grad_norm": 2.3934710025787354, "learning_rate": 1.041299629521092e-07, "loss": 1.0897, "step": 47725 }, { "epoch": 3.740282131661442, "grad_norm": 2.8327956199645996, "learning_rate": 1.0406736772979875e-07, "loss": 1.4942, "step": 47726 }, { "epoch": 3.740360501567398, "grad_norm": 2.7283318042755127, "learning_rate": 1.0400479112910089e-07, "loss": 1.4208, "step": 47727 }, { "epoch": 3.740438871473354, "grad_norm": 3.494051933288574, "learning_rate": 1.0394223315025376e-07, "loss": 1.348, "step": 47728 }, { "epoch": 3.7405172413793104, "grad_norm": 4.8161187171936035, "learning_rate": 1.0387969379349604e-07, "loss": 1.2918, "step": 47729 }, { "epoch": 3.7405956112852663, "grad_norm": 2.130657434463501, "learning_rate": 1.0381717305906536e-07, "loss": 1.2217, "step": 47730 }, { "epoch": 3.7406739811912226, "grad_norm": 3.4074056148529053, "learning_rate": 1.0375467094719871e-07, "loss": 1.484, "step": 47731 }, { "epoch": 3.7407523510971785, "grad_norm": 3.0049421787261963, "learning_rate": 1.036921874581348e-07, "loss": 1.1967, "step": 47732 }, { "epoch": 3.740830721003135, "grad_norm": 8.179917335510254, "learning_rate": 1.0362972259211124e-07, "loss": 1.8993, "step": 47733 }, { "epoch": 3.740909090909091, "grad_norm": 2.36376690864563, "learning_rate": 1.0356727634936504e-07, "loss": 0.832, "step": 47734 }, { "epoch": 3.740987460815047, "grad_norm": 2.9765419960021973, "learning_rate": 1.0350484873013378e-07, "loss": 1.1007, "step": 47735 }, { "epoch": 3.741065830721003, "grad_norm": 2.5856595039367676, "learning_rate": 1.0344243973465562e-07, "loss": 1.5541, "step": 47736 }, { "epoch": 3.7411442006269593, "grad_norm": 2.61057448387146, "learning_rate": 1.0338004936316648e-07, "loss": 1.1662, "step": 47737 }, { "epoch": 3.7412225705329156, "grad_norm": 3.4015414714813232, "learning_rate": 1.033176776159045e-07, "loss": 1.3568, "step": 47738 }, { "epoch": 3.7413009404388715, "grad_norm": 3.8507537841796875, "learning_rate": 1.0325532449310837e-07, "loss": 1.2904, "step": 47739 }, { "epoch": 3.7413793103448274, "grad_norm": 2.590512990951538, "learning_rate": 1.0319298999501293e-07, "loss": 1.3511, "step": 47740 }, { "epoch": 3.7414576802507837, "grad_norm": 5.953139305114746, "learning_rate": 1.0313067412185573e-07, "loss": 1.8155, "step": 47741 }, { "epoch": 3.74153605015674, "grad_norm": 7.146529674530029, "learning_rate": 1.0306837687387438e-07, "loss": 0.8821, "step": 47742 }, { "epoch": 3.741614420062696, "grad_norm": 5.379749298095703, "learning_rate": 1.0300609825130647e-07, "loss": 1.3802, "step": 47743 }, { "epoch": 3.741692789968652, "grad_norm": 5.784861087799072, "learning_rate": 1.0294383825438792e-07, "loss": 1.1296, "step": 47744 }, { "epoch": 3.741771159874608, "grad_norm": 3.5287930965423584, "learning_rate": 1.0288159688335576e-07, "loss": 1.0995, "step": 47745 }, { "epoch": 3.7418495297805645, "grad_norm": 7.316222667694092, "learning_rate": 1.0281937413844645e-07, "loss": 1.1285, "step": 47746 }, { "epoch": 3.7419278996865204, "grad_norm": 3.823925256729126, "learning_rate": 1.0275717001989649e-07, "loss": 1.2129, "step": 47747 }, { "epoch": 3.7420062695924763, "grad_norm": 2.853118658065796, "learning_rate": 1.0269498452794346e-07, "loss": 1.3249, "step": 47748 }, { "epoch": 3.7420846394984326, "grad_norm": 4.9483466148376465, "learning_rate": 1.0263281766282274e-07, "loss": 1.532, "step": 47749 }, { "epoch": 3.742163009404389, "grad_norm": 5.3191657066345215, "learning_rate": 1.0257066942477245e-07, "loss": 1.4111, "step": 47750 }, { "epoch": 3.742241379310345, "grad_norm": 3.55022931098938, "learning_rate": 1.0250853981402631e-07, "loss": 1.466, "step": 47751 }, { "epoch": 3.7423197492163007, "grad_norm": 2.8960862159729004, "learning_rate": 1.0244642883082301e-07, "loss": 1.2331, "step": 47752 }, { "epoch": 3.742398119122257, "grad_norm": 2.9040355682373047, "learning_rate": 1.0238433647539792e-07, "loss": 1.3756, "step": 47753 }, { "epoch": 3.7424764890282134, "grad_norm": 6.081613063812256, "learning_rate": 1.0232226274798807e-07, "loss": 0.9917, "step": 47754 }, { "epoch": 3.7425548589341693, "grad_norm": 4.9042534828186035, "learning_rate": 1.0226020764882771e-07, "loss": 1.6957, "step": 47755 }, { "epoch": 3.742633228840125, "grad_norm": 3.8135926723480225, "learning_rate": 1.0219817117815334e-07, "loss": 1.3551, "step": 47756 }, { "epoch": 3.7427115987460815, "grad_norm": 3.5454211235046387, "learning_rate": 1.0213615333620253e-07, "loss": 1.1955, "step": 47757 }, { "epoch": 3.742789968652038, "grad_norm": 5.065151214599609, "learning_rate": 1.020741541232101e-07, "loss": 1.4159, "step": 47758 }, { "epoch": 3.7428683385579937, "grad_norm": 4.155544281005859, "learning_rate": 1.0201217353941139e-07, "loss": 1.7057, "step": 47759 }, { "epoch": 3.7429467084639496, "grad_norm": 4.4233317375183105, "learning_rate": 1.0195021158504292e-07, "loss": 1.4398, "step": 47760 }, { "epoch": 3.743025078369906, "grad_norm": 4.954626083374023, "learning_rate": 1.0188826826033892e-07, "loss": 1.2139, "step": 47761 }, { "epoch": 3.7431034482758623, "grad_norm": 4.586382865905762, "learning_rate": 1.0182634356553755e-07, "loss": 1.2028, "step": 47762 }, { "epoch": 3.743181818181818, "grad_norm": 4.81251335144043, "learning_rate": 1.017644375008725e-07, "loss": 0.8792, "step": 47763 }, { "epoch": 3.743260188087774, "grad_norm": 4.42629337310791, "learning_rate": 1.0170255006657969e-07, "loss": 1.2852, "step": 47764 }, { "epoch": 3.7433385579937304, "grad_norm": 3.6373653411865234, "learning_rate": 1.0164068126289394e-07, "loss": 1.1518, "step": 47765 }, { "epoch": 3.7434169278996867, "grad_norm": 9.574603080749512, "learning_rate": 1.0157883109005173e-07, "loss": 1.6153, "step": 47766 }, { "epoch": 3.7434952978056426, "grad_norm": 2.9161620140075684, "learning_rate": 1.015169995482873e-07, "loss": 1.4838, "step": 47767 }, { "epoch": 3.7435736677115985, "grad_norm": 7.442214488983154, "learning_rate": 1.0145518663783605e-07, "loss": 1.3795, "step": 47768 }, { "epoch": 3.743652037617555, "grad_norm": 7.5151143074035645, "learning_rate": 1.013933923589333e-07, "loss": 1.7264, "step": 47769 }, { "epoch": 3.743730407523511, "grad_norm": 3.3651304244995117, "learning_rate": 1.0133161671181447e-07, "loss": 1.7091, "step": 47770 }, { "epoch": 3.743808777429467, "grad_norm": 2.3181183338165283, "learning_rate": 1.0126985969671321e-07, "loss": 0.918, "step": 47771 }, { "epoch": 3.743887147335423, "grad_norm": 4.717535495758057, "learning_rate": 1.0120812131386604e-07, "loss": 1.5552, "step": 47772 }, { "epoch": 3.7439655172413793, "grad_norm": 3.814426898956299, "learning_rate": 1.011464015635072e-07, "loss": 1.4343, "step": 47773 }, { "epoch": 3.7440438871473356, "grad_norm": 3.4302420616149902, "learning_rate": 1.0108470044587038e-07, "loss": 1.5144, "step": 47774 }, { "epoch": 3.7441222570532915, "grad_norm": 3.5230419635772705, "learning_rate": 1.0102301796119152e-07, "loss": 1.3683, "step": 47775 }, { "epoch": 3.7442006269592474, "grad_norm": 3.955925941467285, "learning_rate": 1.009613541097043e-07, "loss": 0.7929, "step": 47776 }, { "epoch": 3.7442789968652037, "grad_norm": 6.187939167022705, "learning_rate": 1.0089970889164469e-07, "loss": 1.6845, "step": 47777 }, { "epoch": 3.74435736677116, "grad_norm": 3.890007257461548, "learning_rate": 1.0083808230724578e-07, "loss": 1.2232, "step": 47778 }, { "epoch": 3.744435736677116, "grad_norm": 3.528132200241089, "learning_rate": 1.0077647435674299e-07, "loss": 1.0328, "step": 47779 }, { "epoch": 3.7445141065830723, "grad_norm": 3.3122687339782715, "learning_rate": 1.0071488504036941e-07, "loss": 1.3399, "step": 47780 }, { "epoch": 3.744592476489028, "grad_norm": 6.821609973907471, "learning_rate": 1.006533143583599e-07, "loss": 1.3575, "step": 47781 }, { "epoch": 3.7446708463949845, "grad_norm": 6.1533002853393555, "learning_rate": 1.0059176231094925e-07, "loss": 0.8357, "step": 47782 }, { "epoch": 3.7447492163009404, "grad_norm": 2.491950035095215, "learning_rate": 1.005302288983706e-07, "loss": 1.4877, "step": 47783 }, { "epoch": 3.7448275862068967, "grad_norm": 2.843045473098755, "learning_rate": 1.004687141208588e-07, "loss": 1.3665, "step": 47784 }, { "epoch": 3.7449059561128526, "grad_norm": 2.444758892059326, "learning_rate": 1.0040721797864694e-07, "loss": 1.3902, "step": 47785 }, { "epoch": 3.744984326018809, "grad_norm": 3.2528905868530273, "learning_rate": 1.0034574047196988e-07, "loss": 0.8369, "step": 47786 }, { "epoch": 3.745062695924765, "grad_norm": 3.4151737689971924, "learning_rate": 1.0028428160106073e-07, "loss": 1.5571, "step": 47787 }, { "epoch": 3.745141065830721, "grad_norm": 11.28531551361084, "learning_rate": 1.0022284136615379e-07, "loss": 1.4957, "step": 47788 }, { "epoch": 3.745219435736677, "grad_norm": 2.9004645347595215, "learning_rate": 1.0016141976748217e-07, "loss": 1.1786, "step": 47789 }, { "epoch": 3.7452978056426334, "grad_norm": 3.2863500118255615, "learning_rate": 1.0010001680528015e-07, "loss": 1.1535, "step": 47790 }, { "epoch": 3.7453761755485893, "grad_norm": 4.639658451080322, "learning_rate": 1.0003863247978085e-07, "loss": 1.8387, "step": 47791 }, { "epoch": 3.7454545454545456, "grad_norm": 9.589577674865723, "learning_rate": 9.9977266791218e-08, "loss": 0.8522, "step": 47792 }, { "epoch": 3.7455329153605015, "grad_norm": 2.6814398765563965, "learning_rate": 9.991591973982418e-08, "loss": 1.3718, "step": 47793 }, { "epoch": 3.745611285266458, "grad_norm": 2.962348461151123, "learning_rate": 9.985459132583364e-08, "loss": 1.1768, "step": 47794 }, { "epoch": 3.7456896551724137, "grad_norm": 4.27032470703125, "learning_rate": 9.979328154947954e-08, "loss": 1.217, "step": 47795 }, { "epoch": 3.74576802507837, "grad_norm": 2.7537333965301514, "learning_rate": 9.973199041099501e-08, "loss": 0.9766, "step": 47796 }, { "epoch": 3.745846394984326, "grad_norm": 5.971887111663818, "learning_rate": 9.967071791061267e-08, "loss": 1.2739, "step": 47797 }, { "epoch": 3.7459247648902823, "grad_norm": 2.989961624145508, "learning_rate": 9.960946404856564e-08, "loss": 1.3444, "step": 47798 }, { "epoch": 3.746003134796238, "grad_norm": 5.46232795715332, "learning_rate": 9.95482288250882e-08, "loss": 0.7421, "step": 47799 }, { "epoch": 3.7460815047021945, "grad_norm": 3.4852066040039062, "learning_rate": 9.948701224041124e-08, "loss": 1.1727, "step": 47800 }, { "epoch": 3.7461598746081504, "grad_norm": 2.8846471309661865, "learning_rate": 9.94258142947696e-08, "loss": 0.6274, "step": 47801 }, { "epoch": 3.7462382445141067, "grad_norm": 5.943608283996582, "learning_rate": 9.936463498839421e-08, "loss": 1.3819, "step": 47802 }, { "epoch": 3.7463166144200626, "grad_norm": 2.893430709838867, "learning_rate": 9.930347432151932e-08, "loss": 1.5889, "step": 47803 }, { "epoch": 3.746394984326019, "grad_norm": 3.192629337310791, "learning_rate": 9.924233229437641e-08, "loss": 1.5335, "step": 47804 }, { "epoch": 3.746473354231975, "grad_norm": 3.0727427005767822, "learning_rate": 9.918120890719862e-08, "loss": 1.3706, "step": 47805 }, { "epoch": 3.746551724137931, "grad_norm": 2.4314420223236084, "learning_rate": 9.912010416021856e-08, "loss": 1.2519, "step": 47806 }, { "epoch": 3.746630094043887, "grad_norm": 3.410715341567993, "learning_rate": 9.905901805366769e-08, "loss": 1.4063, "step": 47807 }, { "epoch": 3.7467084639498434, "grad_norm": 4.841366291046143, "learning_rate": 9.899795058777917e-08, "loss": 1.3512, "step": 47808 }, { "epoch": 3.7467868338557992, "grad_norm": 3.01796293258667, "learning_rate": 9.893690176278559e-08, "loss": 1.3614, "step": 47809 }, { "epoch": 3.7468652037617556, "grad_norm": 3.6323015689849854, "learning_rate": 9.887587157891897e-08, "loss": 1.1778, "step": 47810 }, { "epoch": 3.7469435736677115, "grad_norm": 3.514552593231201, "learning_rate": 9.881486003641028e-08, "loss": 1.3993, "step": 47811 }, { "epoch": 3.747021943573668, "grad_norm": 2.8528988361358643, "learning_rate": 9.875386713549261e-08, "loss": 1.2423, "step": 47812 }, { "epoch": 3.7471003134796237, "grad_norm": 2.9112322330474854, "learning_rate": 9.869289287639805e-08, "loss": 1.0972, "step": 47813 }, { "epoch": 3.74717868338558, "grad_norm": 3.1582000255584717, "learning_rate": 9.86319372593586e-08, "loss": 1.3052, "step": 47814 }, { "epoch": 3.747257053291536, "grad_norm": 5.312618732452393, "learning_rate": 9.85710002846052e-08, "loss": 1.2636, "step": 47815 }, { "epoch": 3.7473354231974922, "grad_norm": 7.423459053039551, "learning_rate": 9.851008195237044e-08, "loss": 1.7417, "step": 47816 }, { "epoch": 3.747413793103448, "grad_norm": 11.130770683288574, "learning_rate": 9.844918226288525e-08, "loss": 1.464, "step": 47817 }, { "epoch": 3.7474921630094045, "grad_norm": 3.823530435562134, "learning_rate": 9.838830121638276e-08, "loss": 1.6053, "step": 47818 }, { "epoch": 3.7475705329153604, "grad_norm": 2.638352394104004, "learning_rate": 9.832743881309282e-08, "loss": 1.0862, "step": 47819 }, { "epoch": 3.7476489028213167, "grad_norm": 3.168239116668701, "learning_rate": 9.826659505324799e-08, "loss": 1.3113, "step": 47820 }, { "epoch": 3.7477272727272726, "grad_norm": 3.8006129264831543, "learning_rate": 9.82057699370792e-08, "loss": 1.1515, "step": 47821 }, { "epoch": 3.747805642633229, "grad_norm": 3.0529978275299072, "learning_rate": 9.814496346481794e-08, "loss": 1.2843, "step": 47822 }, { "epoch": 3.747884012539185, "grad_norm": 4.497582912445068, "learning_rate": 9.808417563669515e-08, "loss": 1.0956, "step": 47823 }, { "epoch": 3.747962382445141, "grad_norm": 3.4650979042053223, "learning_rate": 9.802340645294284e-08, "loss": 1.3262, "step": 47824 }, { "epoch": 3.748040752351097, "grad_norm": 2.8433871269226074, "learning_rate": 9.796265591379195e-08, "loss": 1.4309, "step": 47825 }, { "epoch": 3.7481191222570533, "grad_norm": 3.0051169395446777, "learning_rate": 9.79019240194723e-08, "loss": 1.4056, "step": 47826 }, { "epoch": 3.7481974921630092, "grad_norm": 3.3555080890655518, "learning_rate": 9.784121077021647e-08, "loss": 1.1795, "step": 47827 }, { "epoch": 3.7482758620689656, "grad_norm": 2.9081063270568848, "learning_rate": 9.778051616625483e-08, "loss": 1.3547, "step": 47828 }, { "epoch": 3.748354231974922, "grad_norm": 5.860491752624512, "learning_rate": 9.771984020781832e-08, "loss": 1.6532, "step": 47829 }, { "epoch": 3.748432601880878, "grad_norm": 4.0417351722717285, "learning_rate": 9.765918289513731e-08, "loss": 2.2118, "step": 47830 }, { "epoch": 3.7485109717868337, "grad_norm": 3.2132949829101562, "learning_rate": 9.759854422844216e-08, "loss": 1.5517, "step": 47831 }, { "epoch": 3.74858934169279, "grad_norm": 4.374732494354248, "learning_rate": 9.753792420796549e-08, "loss": 1.4184, "step": 47832 }, { "epoch": 3.7486677115987463, "grad_norm": 2.4689862728118896, "learning_rate": 9.74773228339354e-08, "loss": 1.2879, "step": 47833 }, { "epoch": 3.7487460815047022, "grad_norm": 2.760352373123169, "learning_rate": 9.741674010658397e-08, "loss": 1.2969, "step": 47834 }, { "epoch": 3.748824451410658, "grad_norm": 3.627707004547119, "learning_rate": 9.7356176026141e-08, "loss": 1.4081, "step": 47835 }, { "epoch": 3.7489028213166145, "grad_norm": 7.444633960723877, "learning_rate": 9.729563059283686e-08, "loss": 0.8015, "step": 47836 }, { "epoch": 3.748981191222571, "grad_norm": 2.9960920810699463, "learning_rate": 9.723510380690193e-08, "loss": 1.3204, "step": 47837 }, { "epoch": 3.7490595611285267, "grad_norm": 2.166109561920166, "learning_rate": 9.717459566856657e-08, "loss": 1.5275, "step": 47838 }, { "epoch": 3.7491379310344826, "grad_norm": 4.374780654907227, "learning_rate": 9.711410617806061e-08, "loss": 0.9837, "step": 47839 }, { "epoch": 3.749216300940439, "grad_norm": 3.462965726852417, "learning_rate": 9.70536353356144e-08, "loss": 1.1979, "step": 47840 }, { "epoch": 3.7492946708463952, "grad_norm": 4.9811506271362305, "learning_rate": 9.699318314145833e-08, "loss": 1.5942, "step": 47841 }, { "epoch": 3.749373040752351, "grad_norm": 5.274604320526123, "learning_rate": 9.693274959582111e-08, "loss": 1.5689, "step": 47842 }, { "epoch": 3.749451410658307, "grad_norm": 7.495216369628906, "learning_rate": 9.687233469893365e-08, "loss": 1.44, "step": 47843 }, { "epoch": 3.7495297805642633, "grad_norm": 6.966709136962891, "learning_rate": 9.681193845102577e-08, "loss": 1.2179, "step": 47844 }, { "epoch": 3.7496081504702197, "grad_norm": 4.995823860168457, "learning_rate": 9.675156085232617e-08, "loss": 1.2397, "step": 47845 }, { "epoch": 3.7496865203761756, "grad_norm": 5.47482967376709, "learning_rate": 9.669120190306525e-08, "loss": 1.4901, "step": 47846 }, { "epoch": 3.7497648902821314, "grad_norm": 2.8094892501831055, "learning_rate": 9.663086160347335e-08, "loss": 1.5518, "step": 47847 }, { "epoch": 3.749843260188088, "grad_norm": 3.3675999641418457, "learning_rate": 9.657053995377863e-08, "loss": 1.4714, "step": 47848 }, { "epoch": 3.749921630094044, "grad_norm": 2.1167473793029785, "learning_rate": 9.651023695421092e-08, "loss": 1.5139, "step": 47849 }, { "epoch": 3.75, "grad_norm": 6.531112194061279, "learning_rate": 9.64499526049989e-08, "loss": 1.1724, "step": 47850 }, { "epoch": 3.75, "eval_loss": 1.2434347867965698, "eval_runtime": 108.0569, "eval_samples_per_second": 24.867, "eval_steps_per_second": 6.219, "step": 47850 }, { "epoch": 3.750078369905956, "grad_norm": 13.42862606048584, "learning_rate": 9.638968690637407e-08, "loss": 0.9284, "step": 47851 }, { "epoch": 3.750156739811912, "grad_norm": 3.1022088527679443, "learning_rate": 9.632943985856347e-08, "loss": 1.3806, "step": 47852 }, { "epoch": 3.7502351097178686, "grad_norm": 3.5012893676757812, "learning_rate": 9.626921146179691e-08, "loss": 1.3569, "step": 47853 }, { "epoch": 3.7503134796238244, "grad_norm": 2.5130672454833984, "learning_rate": 9.620900171630366e-08, "loss": 1.3245, "step": 47854 }, { "epoch": 3.7503918495297803, "grad_norm": 7.912840366363525, "learning_rate": 9.614881062231296e-08, "loss": 1.4391, "step": 47855 }, { "epoch": 3.7504702194357367, "grad_norm": 3.0804474353790283, "learning_rate": 9.608863818005298e-08, "loss": 1.3187, "step": 47856 }, { "epoch": 3.750548589341693, "grad_norm": 3.0735130310058594, "learning_rate": 9.602848438975299e-08, "loss": 1.424, "step": 47857 }, { "epoch": 3.750626959247649, "grad_norm": 3.7129838466644287, "learning_rate": 9.596834925164222e-08, "loss": 1.3815, "step": 47858 }, { "epoch": 3.7507053291536048, "grad_norm": 5.551504611968994, "learning_rate": 9.590823276594885e-08, "loss": 1.5406, "step": 47859 }, { "epoch": 3.750783699059561, "grad_norm": 4.7815728187561035, "learning_rate": 9.584813493290157e-08, "loss": 1.1817, "step": 47860 }, { "epoch": 3.7508620689655174, "grad_norm": 2.419990301132202, "learning_rate": 9.578805575272853e-08, "loss": 1.4594, "step": 47861 }, { "epoch": 3.7509404388714733, "grad_norm": 7.289647102355957, "learning_rate": 9.572799522566012e-08, "loss": 1.3957, "step": 47862 }, { "epoch": 3.751018808777429, "grad_norm": 3.8548736572265625, "learning_rate": 9.566795335192225e-08, "loss": 1.2772, "step": 47863 }, { "epoch": 3.7510971786833855, "grad_norm": 4.381824016571045, "learning_rate": 9.560793013174419e-08, "loss": 1.1859, "step": 47864 }, { "epoch": 3.751175548589342, "grad_norm": 5.8027496337890625, "learning_rate": 9.55479255653552e-08, "loss": 1.5773, "step": 47865 }, { "epoch": 3.7512539184952978, "grad_norm": 2.7691502571105957, "learning_rate": 9.548793965298342e-08, "loss": 1.3253, "step": 47866 }, { "epoch": 3.7513322884012537, "grad_norm": 8.384860038757324, "learning_rate": 9.542797239485534e-08, "loss": 1.4615, "step": 47867 }, { "epoch": 3.75141065830721, "grad_norm": 3.1946680545806885, "learning_rate": 9.53680237912008e-08, "loss": 1.5163, "step": 47868 }, { "epoch": 3.7514890282131663, "grad_norm": 3.248300790786743, "learning_rate": 9.530809384224626e-08, "loss": 0.8729, "step": 47869 }, { "epoch": 3.751567398119122, "grad_norm": 3.527061700820923, "learning_rate": 9.524818254822154e-08, "loss": 1.3532, "step": 47870 }, { "epoch": 3.751645768025078, "grad_norm": 2.3446924686431885, "learning_rate": 9.518828990935258e-08, "loss": 1.2131, "step": 47871 }, { "epoch": 3.7517241379310344, "grad_norm": 6.488008975982666, "learning_rate": 9.512841592586863e-08, "loss": 1.5097, "step": 47872 }, { "epoch": 3.7518025078369908, "grad_norm": 3.559910535812378, "learning_rate": 9.506856059799675e-08, "loss": 1.2272, "step": 47873 }, { "epoch": 3.7518808777429467, "grad_norm": 4.156826972961426, "learning_rate": 9.500872392596505e-08, "loss": 1.3185, "step": 47874 }, { "epoch": 3.751959247648903, "grad_norm": 3.4535303115844727, "learning_rate": 9.494890591000005e-08, "loss": 1.3479, "step": 47875 }, { "epoch": 3.752037617554859, "grad_norm": 2.7376954555511475, "learning_rate": 9.488910655033046e-08, "loss": 1.0695, "step": 47876 }, { "epoch": 3.752115987460815, "grad_norm": 3.524376392364502, "learning_rate": 9.482932584718329e-08, "loss": 1.4275, "step": 47877 }, { "epoch": 3.752194357366771, "grad_norm": 3.444161891937256, "learning_rate": 9.476956380078616e-08, "loss": 1.6938, "step": 47878 }, { "epoch": 3.7522727272727274, "grad_norm": 5.407094478607178, "learning_rate": 9.470982041136556e-08, "loss": 1.5417, "step": 47879 }, { "epoch": 3.7523510971786833, "grad_norm": 2.407918691635132, "learning_rate": 9.465009567914962e-08, "loss": 1.5118, "step": 47880 }, { "epoch": 3.7524294670846396, "grad_norm": 3.946608066558838, "learning_rate": 9.459038960436595e-08, "loss": 1.5284, "step": 47881 }, { "epoch": 3.7525078369905955, "grad_norm": 2.9337103366851807, "learning_rate": 9.453070218723936e-08, "loss": 1.4681, "step": 47882 }, { "epoch": 3.752586206896552, "grad_norm": 2.78253436088562, "learning_rate": 9.447103342799913e-08, "loss": 1.3659, "step": 47883 }, { "epoch": 3.7526645768025078, "grad_norm": 2.8987884521484375, "learning_rate": 9.441138332687116e-08, "loss": 1.5377, "step": 47884 }, { "epoch": 3.752742946708464, "grad_norm": 4.681456565856934, "learning_rate": 9.435175188408363e-08, "loss": 1.2283, "step": 47885 }, { "epoch": 3.75282131661442, "grad_norm": 11.31863021850586, "learning_rate": 9.429213909986134e-08, "loss": 1.0667, "step": 47886 }, { "epoch": 3.7528996865203763, "grad_norm": 2.3901944160461426, "learning_rate": 9.423254497443245e-08, "loss": 1.0689, "step": 47887 }, { "epoch": 3.752978056426332, "grad_norm": 3.4158527851104736, "learning_rate": 9.41729695080229e-08, "loss": 1.0165, "step": 47888 }, { "epoch": 3.7530564263322885, "grad_norm": 4.353450298309326, "learning_rate": 9.411341270085972e-08, "loss": 1.5664, "step": 47889 }, { "epoch": 3.7531347962382444, "grad_norm": 4.463903903961182, "learning_rate": 9.405387455316884e-08, "loss": 1.6887, "step": 47890 }, { "epoch": 3.7532131661442008, "grad_norm": 3.57019305229187, "learning_rate": 9.399435506517784e-08, "loss": 1.3547, "step": 47891 }, { "epoch": 3.7532915360501566, "grad_norm": 2.1922762393951416, "learning_rate": 9.393485423711213e-08, "loss": 1.3175, "step": 47892 }, { "epoch": 3.753369905956113, "grad_norm": 4.189955234527588, "learning_rate": 9.387537206919817e-08, "loss": 1.6427, "step": 47893 }, { "epoch": 3.753448275862069, "grad_norm": 3.1987500190734863, "learning_rate": 9.381590856166245e-08, "loss": 1.1119, "step": 47894 }, { "epoch": 3.753526645768025, "grad_norm": 4.6053948402404785, "learning_rate": 9.375646371473146e-08, "loss": 0.7772, "step": 47895 }, { "epoch": 3.753605015673981, "grad_norm": 4.058484077453613, "learning_rate": 9.369703752863002e-08, "loss": 1.7451, "step": 47896 }, { "epoch": 3.7536833855799374, "grad_norm": 3.167837381362915, "learning_rate": 9.363763000358572e-08, "loss": 1.3031, "step": 47897 }, { "epoch": 3.7537617554858933, "grad_norm": 5.383408069610596, "learning_rate": 9.357824113982339e-08, "loss": 1.8862, "step": 47898 }, { "epoch": 3.7538401253918496, "grad_norm": 4.665653705596924, "learning_rate": 9.351887093757006e-08, "loss": 1.2841, "step": 47899 }, { "epoch": 3.7539184952978055, "grad_norm": 5.37632942199707, "learning_rate": 9.345951939705055e-08, "loss": 0.8901, "step": 47900 }, { "epoch": 3.753996865203762, "grad_norm": 5.230102062225342, "learning_rate": 9.340018651849026e-08, "loss": 0.8853, "step": 47901 }, { "epoch": 3.7540752351097177, "grad_norm": 5.331053733825684, "learning_rate": 9.334087230211564e-08, "loss": 1.4439, "step": 47902 }, { "epoch": 3.754153605015674, "grad_norm": 3.9157862663269043, "learning_rate": 9.328157674815264e-08, "loss": 1.2001, "step": 47903 }, { "epoch": 3.75423197492163, "grad_norm": 4.1059088706970215, "learning_rate": 9.322229985682608e-08, "loss": 1.3071, "step": 47904 }, { "epoch": 3.7543103448275863, "grad_norm": 3.9731497764587402, "learning_rate": 9.316304162836132e-08, "loss": 1.3294, "step": 47905 }, { "epoch": 3.754388714733542, "grad_norm": 14.031875610351562, "learning_rate": 9.310380206298375e-08, "loss": 0.965, "step": 47906 }, { "epoch": 3.7544670846394985, "grad_norm": 3.7774510383605957, "learning_rate": 9.304458116091986e-08, "loss": 1.357, "step": 47907 }, { "epoch": 3.7545454545454544, "grad_norm": 2.564613103866577, "learning_rate": 9.29853789223939e-08, "loss": 0.7538, "step": 47908 }, { "epoch": 3.7546238244514107, "grad_norm": 4.6245293617248535, "learning_rate": 9.29261953476307e-08, "loss": 1.6872, "step": 47909 }, { "epoch": 3.7547021943573666, "grad_norm": 2.69205379486084, "learning_rate": 9.286703043685618e-08, "loss": 0.9737, "step": 47910 }, { "epoch": 3.754780564263323, "grad_norm": 4.003427028656006, "learning_rate": 9.280788419029463e-08, "loss": 1.0197, "step": 47911 }, { "epoch": 3.754858934169279, "grad_norm": 3.7904093265533447, "learning_rate": 9.27487566081714e-08, "loss": 1.1966, "step": 47912 }, { "epoch": 3.754937304075235, "grad_norm": 4.224815845489502, "learning_rate": 9.268964769071132e-08, "loss": 0.4955, "step": 47913 }, { "epoch": 3.755015673981191, "grad_norm": 2.976513385772705, "learning_rate": 9.263055743813975e-08, "loss": 1.0603, "step": 47914 }, { "epoch": 3.7550940438871474, "grad_norm": 4.802239894866943, "learning_rate": 9.257148585068043e-08, "loss": 1.0968, "step": 47915 }, { "epoch": 3.7551724137931033, "grad_norm": 4.027517318725586, "learning_rate": 9.251243292855927e-08, "loss": 1.6311, "step": 47916 }, { "epoch": 3.7552507836990596, "grad_norm": 2.637367010116577, "learning_rate": 9.245339867199943e-08, "loss": 1.204, "step": 47917 }, { "epoch": 3.7553291536050155, "grad_norm": 2.881492853164673, "learning_rate": 9.239438308122683e-08, "loss": 1.5739, "step": 47918 }, { "epoch": 3.755407523510972, "grad_norm": 3.055863380432129, "learning_rate": 9.23353861564652e-08, "loss": 1.4814, "step": 47919 }, { "epoch": 3.7554858934169277, "grad_norm": 3.0078136920928955, "learning_rate": 9.227640789793823e-08, "loss": 1.3872, "step": 47920 }, { "epoch": 3.755564263322884, "grad_norm": 6.721156597137451, "learning_rate": 9.221744830587132e-08, "loss": 1.8894, "step": 47921 }, { "epoch": 3.75564263322884, "grad_norm": 2.434831380844116, "learning_rate": 9.215850738048926e-08, "loss": 1.2358, "step": 47922 }, { "epoch": 3.7557210031347963, "grad_norm": 3.5628232955932617, "learning_rate": 9.209958512201467e-08, "loss": 1.4035, "step": 47923 }, { "epoch": 3.7557993730407526, "grad_norm": 2.3950414657592773, "learning_rate": 9.204068153067291e-08, "loss": 1.3143, "step": 47924 }, { "epoch": 3.7558777429467085, "grad_norm": 3.670238733291626, "learning_rate": 9.19817966066866e-08, "loss": 1.2233, "step": 47925 }, { "epoch": 3.7559561128526644, "grad_norm": 3.7007694244384766, "learning_rate": 9.192293035028165e-08, "loss": 1.2324, "step": 47926 }, { "epoch": 3.7560344827586207, "grad_norm": 3.1888527870178223, "learning_rate": 9.186408276168012e-08, "loss": 1.3599, "step": 47927 }, { "epoch": 3.756112852664577, "grad_norm": 9.73666763305664, "learning_rate": 9.18052538411074e-08, "loss": 1.5042, "step": 47928 }, { "epoch": 3.756191222570533, "grad_norm": 3.520052909851074, "learning_rate": 9.174644358878603e-08, "loss": 1.0925, "step": 47929 }, { "epoch": 3.756269592476489, "grad_norm": 9.413445472717285, "learning_rate": 9.168765200494035e-08, "loss": 1.7593, "step": 47930 }, { "epoch": 3.756347962382445, "grad_norm": 2.482102870941162, "learning_rate": 9.162887908979345e-08, "loss": 1.2279, "step": 47931 }, { "epoch": 3.7564263322884015, "grad_norm": 4.663079738616943, "learning_rate": 9.157012484356909e-08, "loss": 1.3492, "step": 47932 }, { "epoch": 3.7565047021943574, "grad_norm": 3.456759452819824, "learning_rate": 9.151138926649151e-08, "loss": 1.1334, "step": 47933 }, { "epoch": 3.7565830721003133, "grad_norm": 3.5672175884246826, "learning_rate": 9.145267235878274e-08, "loss": 0.8184, "step": 47934 }, { "epoch": 3.7566614420062696, "grad_norm": 3.0972869396209717, "learning_rate": 9.139397412066709e-08, "loss": 1.3914, "step": 47935 }, { "epoch": 3.756739811912226, "grad_norm": 3.2993884086608887, "learning_rate": 9.133529455236767e-08, "loss": 1.467, "step": 47936 }, { "epoch": 3.756818181818182, "grad_norm": 2.2891554832458496, "learning_rate": 9.127663365410766e-08, "loss": 1.474, "step": 47937 }, { "epoch": 3.7568965517241377, "grad_norm": 5.440744400024414, "learning_rate": 9.12179914261102e-08, "loss": 1.5966, "step": 47938 }, { "epoch": 3.756974921630094, "grad_norm": 4.75407600402832, "learning_rate": 9.115936786859736e-08, "loss": 1.6261, "step": 47939 }, { "epoch": 3.7570532915360504, "grad_norm": 2.8859615325927734, "learning_rate": 9.110076298179337e-08, "loss": 1.4147, "step": 47940 }, { "epoch": 3.7571316614420063, "grad_norm": 2.7975997924804688, "learning_rate": 9.104217676592087e-08, "loss": 0.9351, "step": 47941 }, { "epoch": 3.757210031347962, "grad_norm": 3.9249486923217773, "learning_rate": 9.098360922120242e-08, "loss": 1.6536, "step": 47942 }, { "epoch": 3.7572884012539185, "grad_norm": 5.353425979614258, "learning_rate": 9.092506034786119e-08, "loss": 1.2886, "step": 47943 }, { "epoch": 3.757366771159875, "grad_norm": 3.6254115104675293, "learning_rate": 9.086653014611924e-08, "loss": 1.5074, "step": 47944 }, { "epoch": 3.7574451410658307, "grad_norm": 2.1818089485168457, "learning_rate": 9.08080186161997e-08, "loss": 1.3719, "step": 47945 }, { "epoch": 3.7575235109717866, "grad_norm": 3.054649591445923, "learning_rate": 9.074952575832463e-08, "loss": 1.1653, "step": 47946 }, { "epoch": 3.757601880877743, "grad_norm": 3.6195807456970215, "learning_rate": 9.069105157271719e-08, "loss": 1.4727, "step": 47947 }, { "epoch": 3.7576802507836993, "grad_norm": 3.161820650100708, "learning_rate": 9.063259605959884e-08, "loss": 1.2107, "step": 47948 }, { "epoch": 3.757758620689655, "grad_norm": 2.35888671875, "learning_rate": 9.057415921919277e-08, "loss": 1.309, "step": 47949 }, { "epoch": 3.757836990595611, "grad_norm": 6.384377479553223, "learning_rate": 9.051574105172101e-08, "loss": 1.0591, "step": 47950 }, { "epoch": 3.7579153605015674, "grad_norm": 2.706282377243042, "learning_rate": 9.04573415574056e-08, "loss": 1.4514, "step": 47951 }, { "epoch": 3.7579937304075237, "grad_norm": 9.080814361572266, "learning_rate": 9.03989607364697e-08, "loss": 1.3545, "step": 47952 }, { "epoch": 3.7580721003134796, "grad_norm": 5.98280668258667, "learning_rate": 9.034059858913313e-08, "loss": 1.1447, "step": 47953 }, { "epoch": 3.7581504702194355, "grad_norm": 3.4883880615234375, "learning_rate": 9.028225511561961e-08, "loss": 1.4321, "step": 47954 }, { "epoch": 3.758228840125392, "grad_norm": 6.900389671325684, "learning_rate": 9.022393031615063e-08, "loss": 0.9993, "step": 47955 }, { "epoch": 3.758307210031348, "grad_norm": 2.2098376750946045, "learning_rate": 9.016562419094821e-08, "loss": 1.3665, "step": 47956 }, { "epoch": 3.758385579937304, "grad_norm": 2.969512939453125, "learning_rate": 9.010733674023386e-08, "loss": 1.2411, "step": 47957 }, { "epoch": 3.75846394984326, "grad_norm": 2.5883452892303467, "learning_rate": 9.004906796422852e-08, "loss": 1.407, "step": 47958 }, { "epoch": 3.7585423197492163, "grad_norm": 3.906743288040161, "learning_rate": 8.999081786315589e-08, "loss": 1.3077, "step": 47959 }, { "epoch": 3.7586206896551726, "grad_norm": 7.556210041046143, "learning_rate": 8.993258643723524e-08, "loss": 1.0228, "step": 47960 }, { "epoch": 3.7586990595611285, "grad_norm": 3.087301015853882, "learning_rate": 8.987437368668972e-08, "loss": 1.311, "step": 47961 }, { "epoch": 3.7587774294670844, "grad_norm": 5.30816125869751, "learning_rate": 8.981617961173972e-08, "loss": 1.1483, "step": 47962 }, { "epoch": 3.7588557993730407, "grad_norm": 3.1568098068237305, "learning_rate": 8.975800421260672e-08, "loss": 1.3937, "step": 47963 }, { "epoch": 3.758934169278997, "grad_norm": 4.0090107917785645, "learning_rate": 8.969984748951277e-08, "loss": 1.5116, "step": 47964 }, { "epoch": 3.759012539184953, "grad_norm": 3.513357639312744, "learning_rate": 8.964170944267824e-08, "loss": 1.4855, "step": 47965 }, { "epoch": 3.759090909090909, "grad_norm": 4.226602077484131, "learning_rate": 8.958359007232465e-08, "loss": 1.7453, "step": 47966 }, { "epoch": 3.759169278996865, "grad_norm": 4.186618328094482, "learning_rate": 8.95254893786729e-08, "loss": 1.336, "step": 47967 }, { "epoch": 3.7592476489028215, "grad_norm": 3.561237335205078, "learning_rate": 8.946740736194393e-08, "loss": 1.3772, "step": 47968 }, { "epoch": 3.7593260188087774, "grad_norm": 2.248762607574463, "learning_rate": 8.940934402235869e-08, "loss": 1.227, "step": 47969 }, { "epoch": 3.7594043887147337, "grad_norm": 2.4615461826324463, "learning_rate": 8.935129936013808e-08, "loss": 1.2111, "step": 47970 }, { "epoch": 3.7594827586206896, "grad_norm": 3.2275633811950684, "learning_rate": 8.929327337550309e-08, "loss": 1.2705, "step": 47971 }, { "epoch": 3.759561128526646, "grad_norm": 3.744607448577881, "learning_rate": 8.92352660686735e-08, "loss": 1.3164, "step": 47972 }, { "epoch": 3.759639498432602, "grad_norm": 2.949927806854248, "learning_rate": 8.917727743987137e-08, "loss": 0.9046, "step": 47973 }, { "epoch": 3.759717868338558, "grad_norm": 2.362359046936035, "learning_rate": 8.911930748931651e-08, "loss": 1.1525, "step": 47974 }, { "epoch": 3.759796238244514, "grad_norm": 3.237149477005005, "learning_rate": 8.906135621722933e-08, "loss": 1.7467, "step": 47975 }, { "epoch": 3.7598746081504704, "grad_norm": 5.093850135803223, "learning_rate": 8.900342362383019e-08, "loss": 1.2901, "step": 47976 }, { "epoch": 3.7599529780564263, "grad_norm": 3.257929801940918, "learning_rate": 8.89455097093389e-08, "loss": 0.9939, "step": 47977 }, { "epoch": 3.7600313479623826, "grad_norm": 7.200143814086914, "learning_rate": 8.888761447397753e-08, "loss": 1.4343, "step": 47978 }, { "epoch": 3.7601097178683385, "grad_norm": 2.9569997787475586, "learning_rate": 8.882973791796479e-08, "loss": 1.3378, "step": 47979 }, { "epoch": 3.760188087774295, "grad_norm": 4.844759941101074, "learning_rate": 8.877188004152104e-08, "loss": 1.8791, "step": 47980 }, { "epoch": 3.7602664576802507, "grad_norm": 5.833033084869385, "learning_rate": 8.871404084486668e-08, "loss": 1.4231, "step": 47981 }, { "epoch": 3.760344827586207, "grad_norm": 3.174947500228882, "learning_rate": 8.865622032822097e-08, "loss": 1.7242, "step": 47982 }, { "epoch": 3.760423197492163, "grad_norm": 14.625962257385254, "learning_rate": 8.859841849180484e-08, "loss": 1.6794, "step": 47983 }, { "epoch": 3.7605015673981192, "grad_norm": 3.6443612575531006, "learning_rate": 8.854063533583756e-08, "loss": 1.2552, "step": 47984 }, { "epoch": 3.760579937304075, "grad_norm": 1.9028223752975464, "learning_rate": 8.848287086053953e-08, "loss": 1.2106, "step": 47985 }, { "epoch": 3.7606583072100315, "grad_norm": 3.186663866043091, "learning_rate": 8.842512506612944e-08, "loss": 1.1274, "step": 47986 }, { "epoch": 3.7607366771159874, "grad_norm": 4.051862716674805, "learning_rate": 8.836739795282712e-08, "loss": 1.7252, "step": 47987 }, { "epoch": 3.7608150470219437, "grad_norm": 4.589631080627441, "learning_rate": 8.830968952085295e-08, "loss": 1.3944, "step": 47988 }, { "epoch": 3.7608934169278996, "grad_norm": 3.862531900405884, "learning_rate": 8.82519997704262e-08, "loss": 1.3512, "step": 47989 }, { "epoch": 3.760971786833856, "grad_norm": 4.129426002502441, "learning_rate": 8.819432870176502e-08, "loss": 1.4652, "step": 47990 }, { "epoch": 3.761050156739812, "grad_norm": 6.692866325378418, "learning_rate": 8.813667631509037e-08, "loss": 1.2192, "step": 47991 }, { "epoch": 3.761128526645768, "grad_norm": 3.82867431640625, "learning_rate": 8.807904261062094e-08, "loss": 1.251, "step": 47992 }, { "epoch": 3.761206896551724, "grad_norm": 8.791221618652344, "learning_rate": 8.802142758857602e-08, "loss": 2.1425, "step": 47993 }, { "epoch": 3.7612852664576804, "grad_norm": 3.2423317432403564, "learning_rate": 8.796383124917485e-08, "loss": 1.1374, "step": 47994 }, { "epoch": 3.7613636363636362, "grad_norm": 5.370523929595947, "learning_rate": 8.79062535926356e-08, "loss": 2.0678, "step": 47995 }, { "epoch": 3.7614420062695926, "grad_norm": 6.805373191833496, "learning_rate": 8.784869461917812e-08, "loss": 1.5158, "step": 47996 }, { "epoch": 3.7615203761755485, "grad_norm": 3.071622133255005, "learning_rate": 8.779115432902163e-08, "loss": 1.3751, "step": 47997 }, { "epoch": 3.761598746081505, "grad_norm": 3.3797805309295654, "learning_rate": 8.773363272238433e-08, "loss": 1.4514, "step": 47998 }, { "epoch": 3.7616771159874607, "grad_norm": 8.106566429138184, "learning_rate": 8.767612979948493e-08, "loss": 1.2834, "step": 47999 }, { "epoch": 3.761755485893417, "grad_norm": 5.292335510253906, "learning_rate": 8.761864556054211e-08, "loss": 1.5457, "step": 48000 }, { "epoch": 3.761833855799373, "grad_norm": 2.9224679470062256, "learning_rate": 8.756118000577518e-08, "loss": 1.1258, "step": 48001 }, { "epoch": 3.7619122257053292, "grad_norm": 2.45117449760437, "learning_rate": 8.750373313540283e-08, "loss": 1.134, "step": 48002 }, { "epoch": 3.761990595611285, "grad_norm": 2.6833691596984863, "learning_rate": 8.744630494964213e-08, "loss": 1.283, "step": 48003 }, { "epoch": 3.7620689655172415, "grad_norm": 2.3869364261627197, "learning_rate": 8.738889544871288e-08, "loss": 1.0224, "step": 48004 }, { "epoch": 3.7621473354231973, "grad_norm": 4.444587707519531, "learning_rate": 8.733150463283324e-08, "loss": 1.5821, "step": 48005 }, { "epoch": 3.7622257053291537, "grad_norm": 4.663910388946533, "learning_rate": 8.727413250222084e-08, "loss": 1.0964, "step": 48006 }, { "epoch": 3.7623040752351096, "grad_norm": 5.039133548736572, "learning_rate": 8.721677905709436e-08, "loss": 1.4487, "step": 48007 }, { "epoch": 3.762382445141066, "grad_norm": 3.201258897781372, "learning_rate": 8.715944429767254e-08, "loss": 1.3672, "step": 48008 }, { "epoch": 3.762460815047022, "grad_norm": 3.9026780128479004, "learning_rate": 8.710212822417186e-08, "loss": 1.5551, "step": 48009 }, { "epoch": 3.762539184952978, "grad_norm": 3.0359184741973877, "learning_rate": 8.704483083681159e-08, "loss": 1.1819, "step": 48010 }, { "epoch": 3.762617554858934, "grad_norm": 3.3256075382232666, "learning_rate": 8.698755213580989e-08, "loss": 1.0975, "step": 48011 }, { "epoch": 3.7626959247648903, "grad_norm": 3.5519673824310303, "learning_rate": 8.693029212138382e-08, "loss": 1.2052, "step": 48012 }, { "epoch": 3.7627742946708462, "grad_norm": 2.833184003829956, "learning_rate": 8.687305079375097e-08, "loss": 1.2739, "step": 48013 }, { "epoch": 3.7628526645768026, "grad_norm": 12.571999549865723, "learning_rate": 8.68158281531295e-08, "loss": 0.6788, "step": 48014 }, { "epoch": 3.762931034482759, "grad_norm": 2.562333345413208, "learning_rate": 8.675862419973757e-08, "loss": 1.2677, "step": 48015 }, { "epoch": 3.763009404388715, "grad_norm": 2.512944459915161, "learning_rate": 8.670143893379168e-08, "loss": 1.0774, "step": 48016 }, { "epoch": 3.7630877742946707, "grad_norm": 3.2400169372558594, "learning_rate": 8.664427235551054e-08, "loss": 0.8654, "step": 48017 }, { "epoch": 3.763166144200627, "grad_norm": 3.3443262577056885, "learning_rate": 8.658712446511009e-08, "loss": 1.7473, "step": 48018 }, { "epoch": 3.7632445141065833, "grad_norm": 3.4652626514434814, "learning_rate": 8.652999526280903e-08, "loss": 1.3809, "step": 48019 }, { "epoch": 3.7633228840125392, "grad_norm": 7.525275230407715, "learning_rate": 8.647288474882387e-08, "loss": 1.3745, "step": 48020 }, { "epoch": 3.763401253918495, "grad_norm": 3.815126895904541, "learning_rate": 8.641579292337277e-08, "loss": 1.177, "step": 48021 }, { "epoch": 3.7634796238244514, "grad_norm": 3.60810923576355, "learning_rate": 8.635871978667166e-08, "loss": 1.5173, "step": 48022 }, { "epoch": 3.763557993730408, "grad_norm": 4.044548988342285, "learning_rate": 8.630166533893813e-08, "loss": 1.5149, "step": 48023 }, { "epoch": 3.7636363636363637, "grad_norm": 3.942946195602417, "learning_rate": 8.624462958038926e-08, "loss": 1.565, "step": 48024 }, { "epoch": 3.7637147335423196, "grad_norm": 4.076453685760498, "learning_rate": 8.618761251124207e-08, "loss": 1.4465, "step": 48025 }, { "epoch": 3.763793103448276, "grad_norm": 4.378277778625488, "learning_rate": 8.613061413171309e-08, "loss": 1.1902, "step": 48026 }, { "epoch": 3.7638714733542322, "grad_norm": 3.3314406871795654, "learning_rate": 8.607363444201988e-08, "loss": 1.454, "step": 48027 }, { "epoch": 3.763949843260188, "grad_norm": 2.6087937355041504, "learning_rate": 8.601667344237785e-08, "loss": 1.3285, "step": 48028 }, { "epoch": 3.764028213166144, "grad_norm": 4.695356369018555, "learning_rate": 8.595973113300515e-08, "loss": 1.2197, "step": 48029 }, { "epoch": 3.7641065830721003, "grad_norm": 2.6723766326904297, "learning_rate": 8.590280751411772e-08, "loss": 1.1182, "step": 48030 }, { "epoch": 3.7641849529780567, "grad_norm": 3.5439846515655518, "learning_rate": 8.584590258593151e-08, "loss": 1.0547, "step": 48031 }, { "epoch": 3.7642633228840126, "grad_norm": 8.293874740600586, "learning_rate": 8.578901634866354e-08, "loss": 1.1224, "step": 48032 }, { "epoch": 3.7643416927899684, "grad_norm": 2.247908115386963, "learning_rate": 8.573214880252922e-08, "loss": 1.2569, "step": 48033 }, { "epoch": 3.7644200626959248, "grad_norm": 3.351914405822754, "learning_rate": 8.567529994774726e-08, "loss": 1.3706, "step": 48034 }, { "epoch": 3.764498432601881, "grad_norm": 2.905308723449707, "learning_rate": 8.561846978453137e-08, "loss": 1.1245, "step": 48035 }, { "epoch": 3.764576802507837, "grad_norm": 2.8265256881713867, "learning_rate": 8.55616583130986e-08, "loss": 1.1269, "step": 48036 }, { "epoch": 3.764655172413793, "grad_norm": 2.4094831943511963, "learning_rate": 8.550486553366544e-08, "loss": 1.3282, "step": 48037 }, { "epoch": 3.764733542319749, "grad_norm": 2.3714094161987305, "learning_rate": 8.544809144644728e-08, "loss": 1.2848, "step": 48038 }, { "epoch": 3.7648119122257055, "grad_norm": 4.68704891204834, "learning_rate": 8.539133605166005e-08, "loss": 1.2951, "step": 48039 }, { "epoch": 3.7648902821316614, "grad_norm": 3.081935167312622, "learning_rate": 8.533459934952026e-08, "loss": 1.1806, "step": 48040 }, { "epoch": 3.7649686520376173, "grad_norm": 5.143215656280518, "learning_rate": 8.527788134024328e-08, "loss": 1.1031, "step": 48041 }, { "epoch": 3.7650470219435737, "grad_norm": 4.027125358581543, "learning_rate": 8.52211820240445e-08, "loss": 1.0942, "step": 48042 }, { "epoch": 3.76512539184953, "grad_norm": 3.2469875812530518, "learning_rate": 8.51645014011404e-08, "loss": 1.3391, "step": 48043 }, { "epoch": 3.765203761755486, "grad_norm": 3.337847948074341, "learning_rate": 8.510783947174584e-08, "loss": 1.167, "step": 48044 }, { "epoch": 3.7652821316614418, "grad_norm": 3.555387496948242, "learning_rate": 8.505119623607728e-08, "loss": 1.404, "step": 48045 }, { "epoch": 3.765360501567398, "grad_norm": 4.575871467590332, "learning_rate": 8.499457169434899e-08, "loss": 0.7739, "step": 48046 }, { "epoch": 3.7654388714733544, "grad_norm": 3.025249481201172, "learning_rate": 8.493796584677638e-08, "loss": 1.2141, "step": 48047 }, { "epoch": 3.7655172413793103, "grad_norm": 5.827016830444336, "learning_rate": 8.488137869357593e-08, "loss": 1.4937, "step": 48048 }, { "epoch": 3.765595611285266, "grad_norm": 3.4827358722686768, "learning_rate": 8.482481023496192e-08, "loss": 1.1241, "step": 48049 }, { "epoch": 3.7656739811912225, "grad_norm": 6.913659572601318, "learning_rate": 8.476826047114973e-08, "loss": 1.2254, "step": 48050 }, { "epoch": 3.765752351097179, "grad_norm": 2.2312612533569336, "learning_rate": 8.471172940235417e-08, "loss": 1.2704, "step": 48051 }, { "epoch": 3.7658307210031348, "grad_norm": 2.482677936553955, "learning_rate": 8.465521702879064e-08, "loss": 1.3572, "step": 48052 }, { "epoch": 3.7659090909090907, "grad_norm": 4.099656581878662, "learning_rate": 8.459872335067454e-08, "loss": 0.8397, "step": 48053 }, { "epoch": 3.765987460815047, "grad_norm": 3.5733988285064697, "learning_rate": 8.454224836821956e-08, "loss": 0.9495, "step": 48054 }, { "epoch": 3.7660658307210033, "grad_norm": 3.559368133544922, "learning_rate": 8.448579208164165e-08, "loss": 1.1755, "step": 48055 }, { "epoch": 3.766144200626959, "grad_norm": 4.74709415435791, "learning_rate": 8.442935449115452e-08, "loss": 1.8012, "step": 48056 }, { "epoch": 3.766222570532915, "grad_norm": 2.6340813636779785, "learning_rate": 8.437293559697357e-08, "loss": 1.2455, "step": 48057 }, { "epoch": 3.7663009404388714, "grad_norm": 2.6451313495635986, "learning_rate": 8.431653539931306e-08, "loss": 1.2928, "step": 48058 }, { "epoch": 3.7663793103448278, "grad_norm": 3.983374834060669, "learning_rate": 8.426015389838781e-08, "loss": 1.5735, "step": 48059 }, { "epoch": 3.7664576802507836, "grad_norm": 3.790656328201294, "learning_rate": 8.420379109441212e-08, "loss": 1.1899, "step": 48060 }, { "epoch": 3.7665360501567395, "grad_norm": 13.080273628234863, "learning_rate": 8.414744698759969e-08, "loss": 0.9368, "step": 48061 }, { "epoch": 3.766614420062696, "grad_norm": 4.196227073669434, "learning_rate": 8.409112157816591e-08, "loss": 0.9931, "step": 48062 }, { "epoch": 3.766692789968652, "grad_norm": 3.1747617721557617, "learning_rate": 8.403481486632448e-08, "loss": 1.5305, "step": 48063 }, { "epoch": 3.766771159874608, "grad_norm": 5.256866455078125, "learning_rate": 8.397852685228969e-08, "loss": 1.6122, "step": 48064 }, { "epoch": 3.7668495297805644, "grad_norm": 3.0442240238189697, "learning_rate": 8.392225753627525e-08, "loss": 1.4117, "step": 48065 }, { "epoch": 3.7669278996865203, "grad_norm": 7.753551959991455, "learning_rate": 8.386600691849545e-08, "loss": 1.2958, "step": 48066 }, { "epoch": 3.7670062695924766, "grad_norm": 5.193161964416504, "learning_rate": 8.380977499916453e-08, "loss": 2.143, "step": 48067 }, { "epoch": 3.7670846394984325, "grad_norm": 5.409066200256348, "learning_rate": 8.375356177849625e-08, "loss": 1.0314, "step": 48068 }, { "epoch": 3.767163009404389, "grad_norm": 7.089734077453613, "learning_rate": 8.36973672567043e-08, "loss": 1.1375, "step": 48069 }, { "epoch": 3.7672413793103448, "grad_norm": 4.342701435089111, "learning_rate": 8.364119143400185e-08, "loss": 1.345, "step": 48070 }, { "epoch": 3.767319749216301, "grad_norm": 3.685715436935425, "learning_rate": 8.358503431060372e-08, "loss": 1.2107, "step": 48071 }, { "epoch": 3.767398119122257, "grad_norm": 2.3878283500671387, "learning_rate": 8.352889588672252e-08, "loss": 0.9698, "step": 48072 }, { "epoch": 3.7674764890282133, "grad_norm": 2.8768038749694824, "learning_rate": 8.347277616257199e-08, "loss": 1.6476, "step": 48073 }, { "epoch": 3.767554858934169, "grad_norm": 3.427332878112793, "learning_rate": 8.341667513836582e-08, "loss": 0.9134, "step": 48074 }, { "epoch": 3.7676332288401255, "grad_norm": 3.3036181926727295, "learning_rate": 8.336059281431775e-08, "loss": 1.3714, "step": 48075 }, { "epoch": 3.7677115987460814, "grad_norm": 2.835618257522583, "learning_rate": 8.330452919064036e-08, "loss": 1.0225, "step": 48076 }, { "epoch": 3.7677899686520377, "grad_norm": 2.3716626167297363, "learning_rate": 8.324848426754683e-08, "loss": 1.2528, "step": 48077 }, { "epoch": 3.7678683385579936, "grad_norm": 2.8299670219421387, "learning_rate": 8.319245804525144e-08, "loss": 1.3471, "step": 48078 }, { "epoch": 3.76794670846395, "grad_norm": 3.8574931621551514, "learning_rate": 8.313645052396569e-08, "loss": 1.4593, "step": 48079 }, { "epoch": 3.768025078369906, "grad_norm": 2.991114377975464, "learning_rate": 8.308046170390383e-08, "loss": 1.2169, "step": 48080 }, { "epoch": 3.768103448275862, "grad_norm": 2.628178834915161, "learning_rate": 8.302449158527848e-08, "loss": 1.6205, "step": 48081 }, { "epoch": 3.768181818181818, "grad_norm": 2.868232011795044, "learning_rate": 8.296854016830225e-08, "loss": 1.3751, "step": 48082 }, { "epoch": 3.7682601880877744, "grad_norm": 7.976768493652344, "learning_rate": 8.29126074531883e-08, "loss": 1.6005, "step": 48083 }, { "epoch": 3.7683385579937303, "grad_norm": 3.830014944076538, "learning_rate": 8.285669344014923e-08, "loss": 1.3358, "step": 48084 }, { "epoch": 3.7684169278996866, "grad_norm": 2.388587236404419, "learning_rate": 8.280079812939712e-08, "loss": 1.2964, "step": 48085 }, { "epoch": 3.7684952978056425, "grad_norm": 3.235546350479126, "learning_rate": 8.274492152114621e-08, "loss": 1.6359, "step": 48086 }, { "epoch": 3.768573667711599, "grad_norm": 2.200206995010376, "learning_rate": 8.268906361560746e-08, "loss": 1.0651, "step": 48087 }, { "epoch": 3.7686520376175547, "grad_norm": 9.101115226745605, "learning_rate": 8.263322441299348e-08, "loss": 1.9076, "step": 48088 }, { "epoch": 3.768730407523511, "grad_norm": 3.6418986320495605, "learning_rate": 8.257740391351743e-08, "loss": 1.3641, "step": 48089 }, { "epoch": 3.768808777429467, "grad_norm": 2.4757025241851807, "learning_rate": 8.25216021173908e-08, "loss": 1.4474, "step": 48090 }, { "epoch": 3.7688871473354233, "grad_norm": 4.501580715179443, "learning_rate": 8.24658190248262e-08, "loss": 1.4699, "step": 48091 }, { "epoch": 3.768965517241379, "grad_norm": 15.001518249511719, "learning_rate": 8.241005463603624e-08, "loss": 1.6203, "step": 48092 }, { "epoch": 3.7690438871473355, "grad_norm": 3.8931565284729004, "learning_rate": 8.235430895123242e-08, "loss": 1.1737, "step": 48093 }, { "epoch": 3.7691222570532914, "grad_norm": 10.543704986572266, "learning_rate": 8.229858197062678e-08, "loss": 1.5448, "step": 48094 }, { "epoch": 3.7692006269592477, "grad_norm": 3.4836766719818115, "learning_rate": 8.22428736944314e-08, "loss": 1.0739, "step": 48095 }, { "epoch": 3.7692789968652036, "grad_norm": 10.282845497131348, "learning_rate": 8.218718412285886e-08, "loss": 1.7415, "step": 48096 }, { "epoch": 3.76935736677116, "grad_norm": 2.471872568130493, "learning_rate": 8.213151325612012e-08, "loss": 1.4408, "step": 48097 }, { "epoch": 3.769435736677116, "grad_norm": 2.9351701736450195, "learning_rate": 8.20758610944261e-08, "loss": 1.4602, "step": 48098 }, { "epoch": 3.769514106583072, "grad_norm": 8.326384544372559, "learning_rate": 8.202022763799055e-08, "loss": 1.1864, "step": 48099 }, { "epoch": 3.769592476489028, "grad_norm": 2.653148889541626, "learning_rate": 8.196461288702384e-08, "loss": 1.0866, "step": 48100 }, { "epoch": 3.7696708463949844, "grad_norm": 3.1077098846435547, "learning_rate": 8.190901684173803e-08, "loss": 1.3726, "step": 48101 }, { "epoch": 3.7697492163009403, "grad_norm": 4.257950782775879, "learning_rate": 8.185343950234348e-08, "loss": 0.8665, "step": 48102 }, { "epoch": 3.7698275862068966, "grad_norm": 3.8851540088653564, "learning_rate": 8.179788086905227e-08, "loss": 1.3286, "step": 48103 }, { "epoch": 3.7699059561128525, "grad_norm": 2.6894121170043945, "learning_rate": 8.174234094207589e-08, "loss": 1.4338, "step": 48104 }, { "epoch": 3.769984326018809, "grad_norm": 3.3133883476257324, "learning_rate": 8.16868197216264e-08, "loss": 1.3555, "step": 48105 }, { "epoch": 3.7700626959247647, "grad_norm": 4.421352386474609, "learning_rate": 8.163131720791306e-08, "loss": 1.3435, "step": 48106 }, { "epoch": 3.770141065830721, "grad_norm": 2.3184549808502197, "learning_rate": 8.15758334011485e-08, "loss": 1.2034, "step": 48107 }, { "epoch": 3.770219435736677, "grad_norm": 3.8881783485412598, "learning_rate": 8.152036830154253e-08, "loss": 1.2515, "step": 48108 }, { "epoch": 3.7702978056426333, "grad_norm": 2.6585381031036377, "learning_rate": 8.146492190930722e-08, "loss": 1.0371, "step": 48109 }, { "epoch": 3.7703761755485896, "grad_norm": 4.03572416305542, "learning_rate": 8.140949422465294e-08, "loss": 1.4435, "step": 48110 }, { "epoch": 3.7704545454545455, "grad_norm": 2.720447063446045, "learning_rate": 8.135408524779065e-08, "loss": 1.2352, "step": 48111 }, { "epoch": 3.7705329153605014, "grad_norm": 3.459880828857422, "learning_rate": 8.129869497893128e-08, "loss": 1.457, "step": 48112 }, { "epoch": 3.7706112852664577, "grad_norm": 3.2833287715911865, "learning_rate": 8.124332341828523e-08, "loss": 0.8983, "step": 48113 }, { "epoch": 3.770689655172414, "grad_norm": 2.8579063415527344, "learning_rate": 8.118797056606286e-08, "loss": 1.3146, "step": 48114 }, { "epoch": 3.77076802507837, "grad_norm": 3.840327262878418, "learning_rate": 8.113263642247516e-08, "loss": 1.0026, "step": 48115 }, { "epoch": 3.770846394984326, "grad_norm": 3.70452618598938, "learning_rate": 8.107732098773247e-08, "loss": 1.1579, "step": 48116 }, { "epoch": 3.770924764890282, "grad_norm": 3.683971881866455, "learning_rate": 8.102202426204464e-08, "loss": 1.2669, "step": 48117 }, { "epoch": 3.7710031347962385, "grad_norm": 3.802680253982544, "learning_rate": 8.096674624562262e-08, "loss": 1.1198, "step": 48118 }, { "epoch": 3.7710815047021944, "grad_norm": 3.157898426055908, "learning_rate": 8.091148693867678e-08, "loss": 1.1074, "step": 48119 }, { "epoch": 3.7711598746081503, "grad_norm": 2.925426483154297, "learning_rate": 8.085624634141754e-08, "loss": 1.4483, "step": 48120 }, { "epoch": 3.7712382445141066, "grad_norm": 2.671626091003418, "learning_rate": 8.080102445405413e-08, "loss": 1.3212, "step": 48121 }, { "epoch": 3.771316614420063, "grad_norm": 2.9687511920928955, "learning_rate": 8.074582127679642e-08, "loss": 1.2843, "step": 48122 }, { "epoch": 3.771394984326019, "grad_norm": 3.06888484954834, "learning_rate": 8.069063680985534e-08, "loss": 1.0717, "step": 48123 }, { "epoch": 3.7714733542319747, "grad_norm": 4.516074180603027, "learning_rate": 8.063547105344072e-08, "loss": 1.492, "step": 48124 }, { "epoch": 3.771551724137931, "grad_norm": 3.958923578262329, "learning_rate": 8.058032400776128e-08, "loss": 1.2069, "step": 48125 }, { "epoch": 3.7716300940438874, "grad_norm": 18.03122329711914, "learning_rate": 8.052519567302797e-08, "loss": 1.0961, "step": 48126 }, { "epoch": 3.7717084639498433, "grad_norm": 4.492183208465576, "learning_rate": 8.047008604945006e-08, "loss": 0.9108, "step": 48127 }, { "epoch": 3.771786833855799, "grad_norm": 3.4492926597595215, "learning_rate": 8.041499513723682e-08, "loss": 1.0401, "step": 48128 }, { "epoch": 3.7718652037617555, "grad_norm": 3.020015239715576, "learning_rate": 8.035992293659811e-08, "loss": 1.8874, "step": 48129 }, { "epoch": 3.771943573667712, "grad_norm": 6.331140518188477, "learning_rate": 8.030486944774374e-08, "loss": 1.4543, "step": 48130 }, { "epoch": 3.7720219435736677, "grad_norm": 9.51712703704834, "learning_rate": 8.024983467088243e-08, "loss": 1.2182, "step": 48131 }, { "epoch": 3.7721003134796236, "grad_norm": 3.9715940952301025, "learning_rate": 8.019481860622346e-08, "loss": 0.4498, "step": 48132 }, { "epoch": 3.77217868338558, "grad_norm": 11.035484313964844, "learning_rate": 8.013982125397723e-08, "loss": 1.7763, "step": 48133 }, { "epoch": 3.7722570532915363, "grad_norm": 4.677554607391357, "learning_rate": 8.008484261435134e-08, "loss": 1.4739, "step": 48134 }, { "epoch": 3.772335423197492, "grad_norm": 3.551096200942993, "learning_rate": 8.002988268755674e-08, "loss": 1.7397, "step": 48135 }, { "epoch": 3.772413793103448, "grad_norm": 3.900888442993164, "learning_rate": 7.997494147380048e-08, "loss": 1.4141, "step": 48136 }, { "epoch": 3.7724921630094044, "grad_norm": 3.720815658569336, "learning_rate": 7.992001897329238e-08, "loss": 0.971, "step": 48137 }, { "epoch": 3.7725705329153607, "grad_norm": 3.458866596221924, "learning_rate": 7.986511518624174e-08, "loss": 1.3128, "step": 48138 }, { "epoch": 3.7726489028213166, "grad_norm": 2.9350526332855225, "learning_rate": 7.981023011285727e-08, "loss": 1.4113, "step": 48139 }, { "epoch": 3.7727272727272725, "grad_norm": 2.582756757736206, "learning_rate": 7.97553637533477e-08, "loss": 1.406, "step": 48140 }, { "epoch": 3.772805642633229, "grad_norm": 5.265509128570557, "learning_rate": 7.970051610792062e-08, "loss": 1.1103, "step": 48141 }, { "epoch": 3.772884012539185, "grad_norm": 3.9762423038482666, "learning_rate": 7.964568717678644e-08, "loss": 0.9533, "step": 48142 }, { "epoch": 3.772962382445141, "grad_norm": 2.9542269706726074, "learning_rate": 7.959087696015221e-08, "loss": 1.3595, "step": 48143 }, { "epoch": 3.773040752351097, "grad_norm": 2.419600248336792, "learning_rate": 7.953608545822722e-08, "loss": 1.1208, "step": 48144 }, { "epoch": 3.7731191222570533, "grad_norm": 4.746532440185547, "learning_rate": 7.948131267121961e-08, "loss": 1.2728, "step": 48145 }, { "epoch": 3.7731974921630096, "grad_norm": 2.598684072494507, "learning_rate": 7.942655859933812e-08, "loss": 1.3104, "step": 48146 }, { "epoch": 3.7732758620689655, "grad_norm": 3.7289578914642334, "learning_rate": 7.93718232427898e-08, "loss": 1.2056, "step": 48147 }, { "epoch": 3.7733542319749214, "grad_norm": 3.629106044769287, "learning_rate": 7.931710660178449e-08, "loss": 1.5258, "step": 48148 }, { "epoch": 3.7734326018808777, "grad_norm": 2.994370937347412, "learning_rate": 7.926240867652923e-08, "loss": 1.7311, "step": 48149 }, { "epoch": 3.773510971786834, "grad_norm": 2.269719362258911, "learning_rate": 7.920772946723221e-08, "loss": 1.3339, "step": 48150 }, { "epoch": 3.77358934169279, "grad_norm": 4.228687286376953, "learning_rate": 7.915306897410158e-08, "loss": 1.4763, "step": 48151 }, { "epoch": 3.773667711598746, "grad_norm": 3.666644811630249, "learning_rate": 7.909842719734495e-08, "loss": 1.3412, "step": 48152 }, { "epoch": 3.773746081504702, "grad_norm": 3.1330127716064453, "learning_rate": 7.904380413717105e-08, "loss": 0.8008, "step": 48153 }, { "epoch": 3.7738244514106585, "grad_norm": 4.3141303062438965, "learning_rate": 7.898919979378639e-08, "loss": 1.0982, "step": 48154 }, { "epoch": 3.7739028213166144, "grad_norm": 3.4329428672790527, "learning_rate": 7.893461416739911e-08, "loss": 0.6179, "step": 48155 }, { "epoch": 3.7739811912225707, "grad_norm": 2.6834049224853516, "learning_rate": 7.888004725821686e-08, "loss": 1.2835, "step": 48156 }, { "epoch": 3.7740595611285266, "grad_norm": 3.4779396057128906, "learning_rate": 7.882549906644777e-08, "loss": 1.209, "step": 48157 }, { "epoch": 3.774137931034483, "grad_norm": 3.8958165645599365, "learning_rate": 7.877096959229891e-08, "loss": 1.6379, "step": 48158 }, { "epoch": 3.774216300940439, "grad_norm": 8.939674377441406, "learning_rate": 7.871645883597678e-08, "loss": 1.8162, "step": 48159 }, { "epoch": 3.774294670846395, "grad_norm": 6.607933521270752, "learning_rate": 7.866196679768956e-08, "loss": 1.5513, "step": 48160 }, { "epoch": 3.774373040752351, "grad_norm": 4.8939104080200195, "learning_rate": 7.86074934776454e-08, "loss": 1.8819, "step": 48161 }, { "epoch": 3.7744514106583074, "grad_norm": 3.9419972896575928, "learning_rate": 7.855303887605025e-08, "loss": 1.0328, "step": 48162 }, { "epoch": 3.7745297805642632, "grad_norm": 3.5637991428375244, "learning_rate": 7.849860299311118e-08, "loss": 1.5852, "step": 48163 }, { "epoch": 3.7746081504702196, "grad_norm": 5.611634731292725, "learning_rate": 7.84441858290358e-08, "loss": 1.4728, "step": 48164 }, { "epoch": 3.7746865203761755, "grad_norm": 4.5898332595825195, "learning_rate": 7.838978738403057e-08, "loss": 1.4778, "step": 48165 }, { "epoch": 3.774764890282132, "grad_norm": 11.661139488220215, "learning_rate": 7.833540765830317e-08, "loss": 0.9241, "step": 48166 }, { "epoch": 3.7748432601880877, "grad_norm": 6.775015354156494, "learning_rate": 7.82810466520595e-08, "loss": 1.215, "step": 48167 }, { "epoch": 3.774921630094044, "grad_norm": 3.506847381591797, "learning_rate": 7.822670436550717e-08, "loss": 1.2853, "step": 48168 }, { "epoch": 3.775, "grad_norm": 3.176854133605957, "learning_rate": 7.817238079885215e-08, "loss": 1.404, "step": 48169 }, { "epoch": 3.7750783699059562, "grad_norm": 4.136626243591309, "learning_rate": 7.811807595230092e-08, "loss": 1.4002, "step": 48170 }, { "epoch": 3.775156739811912, "grad_norm": 3.3813397884368896, "learning_rate": 7.80637898260611e-08, "loss": 1.4767, "step": 48171 }, { "epoch": 3.7752351097178685, "grad_norm": 3.0052525997161865, "learning_rate": 7.800952242033921e-08, "loss": 1.4262, "step": 48172 }, { "epoch": 3.7753134796238244, "grad_norm": 6.4051103591918945, "learning_rate": 7.795527373534007e-08, "loss": 1.8455, "step": 48173 }, { "epoch": 3.7753918495297807, "grad_norm": 2.471787452697754, "learning_rate": 7.790104377127072e-08, "loss": 1.1517, "step": 48174 }, { "epoch": 3.7754702194357366, "grad_norm": 3.4680237770080566, "learning_rate": 7.784683252833825e-08, "loss": 1.0865, "step": 48175 }, { "epoch": 3.775548589341693, "grad_norm": 3.0183839797973633, "learning_rate": 7.779264000674858e-08, "loss": 1.1212, "step": 48176 }, { "epoch": 3.775626959247649, "grad_norm": 4.373417854309082, "learning_rate": 7.77384662067071e-08, "loss": 1.3183, "step": 48177 }, { "epoch": 3.775705329153605, "grad_norm": 3.23272442817688, "learning_rate": 7.768431112841979e-08, "loss": 1.1614, "step": 48178 }, { "epoch": 3.775783699059561, "grad_norm": 6.19588041305542, "learning_rate": 7.763017477209312e-08, "loss": 0.8761, "step": 48179 }, { "epoch": 3.7758620689655173, "grad_norm": 3.30800724029541, "learning_rate": 7.757605713793359e-08, "loss": 1.2651, "step": 48180 }, { "epoch": 3.7759404388714732, "grad_norm": 2.502563714981079, "learning_rate": 7.75219582261466e-08, "loss": 0.8261, "step": 48181 }, { "epoch": 3.7760188087774296, "grad_norm": 3.118189573287964, "learning_rate": 7.7467878036937e-08, "loss": 0.9206, "step": 48182 }, { "epoch": 3.7760971786833855, "grad_norm": 4.066174507141113, "learning_rate": 7.741381657051183e-08, "loss": 1.3889, "step": 48183 }, { "epoch": 3.776175548589342, "grad_norm": 4.109415531158447, "learning_rate": 7.735977382707538e-08, "loss": 0.4341, "step": 48184 }, { "epoch": 3.7762539184952977, "grad_norm": 5.149341106414795, "learning_rate": 7.730574980683469e-08, "loss": 1.2512, "step": 48185 }, { "epoch": 3.776332288401254, "grad_norm": 4.771597385406494, "learning_rate": 7.725174450999406e-08, "loss": 1.5547, "step": 48186 }, { "epoch": 3.77641065830721, "grad_norm": 3.778913974761963, "learning_rate": 7.719775793675943e-08, "loss": 1.2977, "step": 48187 }, { "epoch": 3.7764890282131662, "grad_norm": 4.876927852630615, "learning_rate": 7.714379008733563e-08, "loss": 1.3022, "step": 48188 }, { "epoch": 3.776567398119122, "grad_norm": 4.369436740875244, "learning_rate": 7.708984096192862e-08, "loss": 1.5656, "step": 48189 }, { "epoch": 3.7766457680250785, "grad_norm": 2.462639570236206, "learning_rate": 7.703591056074377e-08, "loss": 1.2357, "step": 48190 }, { "epoch": 3.7767241379310343, "grad_norm": 2.7491233348846436, "learning_rate": 7.698199888398538e-08, "loss": 1.1285, "step": 48191 }, { "epoch": 3.7768025078369907, "grad_norm": 6.854559898376465, "learning_rate": 7.692810593185829e-08, "loss": 0.5776, "step": 48192 }, { "epoch": 3.7768808777429466, "grad_norm": 2.502284288406372, "learning_rate": 7.687423170456842e-08, "loss": 1.5112, "step": 48193 }, { "epoch": 3.776959247648903, "grad_norm": 9.022204399108887, "learning_rate": 7.682037620232063e-08, "loss": 1.7655, "step": 48194 }, { "epoch": 3.777037617554859, "grad_norm": 4.382046699523926, "learning_rate": 7.67665394253192e-08, "loss": 1.1647, "step": 48195 }, { "epoch": 3.777115987460815, "grad_norm": 5.141310214996338, "learning_rate": 7.67127213737695e-08, "loss": 1.5365, "step": 48196 }, { "epoch": 3.777194357366771, "grad_norm": 3.3414433002471924, "learning_rate": 7.665892204787473e-08, "loss": 1.4014, "step": 48197 }, { "epoch": 3.7772727272727273, "grad_norm": 3.6996684074401855, "learning_rate": 7.660514144784192e-08, "loss": 1.3264, "step": 48198 }, { "epoch": 3.7773510971786832, "grad_norm": 4.031134605407715, "learning_rate": 7.655137957387371e-08, "loss": 1.128, "step": 48199 }, { "epoch": 3.7774294670846396, "grad_norm": 3.9705443382263184, "learning_rate": 7.649763642617492e-08, "loss": 1.5324, "step": 48200 }, { "epoch": 3.7775078369905954, "grad_norm": 2.628704309463501, "learning_rate": 7.644391200495094e-08, "loss": 1.0879, "step": 48201 }, { "epoch": 3.777586206896552, "grad_norm": 5.766811370849609, "learning_rate": 7.639020631040494e-08, "loss": 1.1447, "step": 48202 }, { "epoch": 3.7776645768025077, "grad_norm": 2.2806363105773926, "learning_rate": 7.633651934274178e-08, "loss": 1.3303, "step": 48203 }, { "epoch": 3.777742946708464, "grad_norm": 4.254854202270508, "learning_rate": 7.628285110216571e-08, "loss": 0.8723, "step": 48204 }, { "epoch": 3.7778213166144203, "grad_norm": 2.9000930786132812, "learning_rate": 7.622920158888048e-08, "loss": 1.3527, "step": 48205 }, { "epoch": 3.777899686520376, "grad_norm": 4.166624069213867, "learning_rate": 7.617557080309034e-08, "loss": 1.8157, "step": 48206 }, { "epoch": 3.777978056426332, "grad_norm": 4.3912248611450195, "learning_rate": 7.612195874499906e-08, "loss": 1.5375, "step": 48207 }, { "epoch": 3.7780564263322884, "grad_norm": 3.578061103820801, "learning_rate": 7.606836541481144e-08, "loss": 1.7332, "step": 48208 }, { "epoch": 3.7781347962382448, "grad_norm": 3.8150954246520996, "learning_rate": 7.601479081273067e-08, "loss": 1.5685, "step": 48209 }, { "epoch": 3.7782131661442007, "grad_norm": 4.1778059005737305, "learning_rate": 7.59612349389599e-08, "loss": 1.4167, "step": 48210 }, { "epoch": 3.7782915360501566, "grad_norm": 2.8893089294433594, "learning_rate": 7.590769779370344e-08, "loss": 1.3623, "step": 48211 }, { "epoch": 3.778369905956113, "grad_norm": 3.1566457748413086, "learning_rate": 7.585417937716499e-08, "loss": 1.0374, "step": 48212 }, { "epoch": 3.778448275862069, "grad_norm": 3.097856283187866, "learning_rate": 7.580067968954885e-08, "loss": 0.9679, "step": 48213 }, { "epoch": 3.778526645768025, "grad_norm": 3.9350640773773193, "learning_rate": 7.574719873105706e-08, "loss": 1.1086, "step": 48214 }, { "epoch": 3.778605015673981, "grad_norm": 2.670708417892456, "learning_rate": 7.569373650189394e-08, "loss": 1.1631, "step": 48215 }, { "epoch": 3.7786833855799373, "grad_norm": 3.649508476257324, "learning_rate": 7.56402930022615e-08, "loss": 1.2703, "step": 48216 }, { "epoch": 3.7787617554858937, "grad_norm": 2.875204086303711, "learning_rate": 7.558686823236571e-08, "loss": 1.126, "step": 48217 }, { "epoch": 3.7788401253918495, "grad_norm": 2.9038217067718506, "learning_rate": 7.553346219240698e-08, "loss": 1.3391, "step": 48218 }, { "epoch": 3.7789184952978054, "grad_norm": 15.753636360168457, "learning_rate": 7.548007488259012e-08, "loss": 0.7662, "step": 48219 }, { "epoch": 3.7789968652037618, "grad_norm": 5.586674213409424, "learning_rate": 7.542670630311721e-08, "loss": 1.4161, "step": 48220 }, { "epoch": 3.779075235109718, "grad_norm": 9.068319320678711, "learning_rate": 7.537335645419197e-08, "loss": 1.0112, "step": 48221 }, { "epoch": 3.779153605015674, "grad_norm": 2.5010666847229004, "learning_rate": 7.532002533601702e-08, "loss": 1.1328, "step": 48222 }, { "epoch": 3.77923197492163, "grad_norm": 3.1948487758636475, "learning_rate": 7.526671294879551e-08, "loss": 0.885, "step": 48223 }, { "epoch": 3.779310344827586, "grad_norm": 3.7192413806915283, "learning_rate": 7.521341929273007e-08, "loss": 1.4338, "step": 48224 }, { "epoch": 3.7793887147335425, "grad_norm": 3.0976481437683105, "learning_rate": 7.516014436802221e-08, "loss": 1.2854, "step": 48225 }, { "epoch": 3.7794670846394984, "grad_norm": 3.4009242057800293, "learning_rate": 7.510688817487621e-08, "loss": 1.1026, "step": 48226 }, { "epoch": 3.7795454545454543, "grad_norm": 2.9074230194091797, "learning_rate": 7.505365071349413e-08, "loss": 1.4641, "step": 48227 }, { "epoch": 3.7796238244514107, "grad_norm": 7.2935566902160645, "learning_rate": 7.500043198407858e-08, "loss": 0.912, "step": 48228 }, { "epoch": 3.779702194357367, "grad_norm": 3.063981294631958, "learning_rate": 7.494723198683107e-08, "loss": 1.1502, "step": 48229 }, { "epoch": 3.779780564263323, "grad_norm": 2.283825635910034, "learning_rate": 7.489405072195477e-08, "loss": 1.1874, "step": 48230 }, { "epoch": 3.7798589341692788, "grad_norm": 2.657266855239868, "learning_rate": 7.484088818965174e-08, "loss": 1.5466, "step": 48231 }, { "epoch": 3.779937304075235, "grad_norm": 2.7794642448425293, "learning_rate": 7.478774439012514e-08, "loss": 0.9768, "step": 48232 }, { "epoch": 3.7800156739811914, "grad_norm": 4.402225494384766, "learning_rate": 7.47346193235754e-08, "loss": 1.4456, "step": 48233 }, { "epoch": 3.7800940438871473, "grad_norm": 3.013941526412964, "learning_rate": 7.468151299020565e-08, "loss": 1.265, "step": 48234 }, { "epoch": 3.780172413793103, "grad_norm": 2.4618120193481445, "learning_rate": 7.462842539021742e-08, "loss": 1.5252, "step": 48235 }, { "epoch": 3.7802507836990595, "grad_norm": 3.407179594039917, "learning_rate": 7.457535652381331e-08, "loss": 1.2058, "step": 48236 }, { "epoch": 3.780329153605016, "grad_norm": 3.939490556716919, "learning_rate": 7.452230639119429e-08, "loss": 1.2175, "step": 48237 }, { "epoch": 3.7804075235109718, "grad_norm": 2.505790948867798, "learning_rate": 7.44692749925624e-08, "loss": 1.3061, "step": 48238 }, { "epoch": 3.7804858934169276, "grad_norm": 3.1434242725372314, "learning_rate": 7.441626232811971e-08, "loss": 1.2208, "step": 48239 }, { "epoch": 3.780564263322884, "grad_norm": 3.1022446155548096, "learning_rate": 7.436326839806718e-08, "loss": 1.0268, "step": 48240 }, { "epoch": 3.7806426332288403, "grad_norm": 3.640209674835205, "learning_rate": 7.431029320260741e-08, "loss": 0.8649, "step": 48241 }, { "epoch": 3.780721003134796, "grad_norm": 2.2928807735443115, "learning_rate": 7.42573367419408e-08, "loss": 1.411, "step": 48242 }, { "epoch": 3.780799373040752, "grad_norm": 8.018356323242188, "learning_rate": 7.420439901626997e-08, "loss": 0.6481, "step": 48243 }, { "epoch": 3.7808777429467084, "grad_norm": 3.4381802082061768, "learning_rate": 7.415148002579475e-08, "loss": 1.3012, "step": 48244 }, { "epoch": 3.7809561128526648, "grad_norm": 4.1311354637146, "learning_rate": 7.409857977071777e-08, "loss": 1.1292, "step": 48245 }, { "epoch": 3.7810344827586206, "grad_norm": 4.723952770233154, "learning_rate": 7.404569825123941e-08, "loss": 1.527, "step": 48246 }, { "epoch": 3.7811128526645765, "grad_norm": 2.558114767074585, "learning_rate": 7.399283546756175e-08, "loss": 1.0442, "step": 48247 }, { "epoch": 3.781191222570533, "grad_norm": 6.67140531539917, "learning_rate": 7.393999141988461e-08, "loss": 1.1253, "step": 48248 }, { "epoch": 3.781269592476489, "grad_norm": 4.880400657653809, "learning_rate": 7.388716610840895e-08, "loss": 1.6199, "step": 48249 }, { "epoch": 3.781347962382445, "grad_norm": 3.0013716220855713, "learning_rate": 7.383435953333684e-08, "loss": 1.5282, "step": 48250 }, { "epoch": 3.7814263322884014, "grad_norm": 2.499286413192749, "learning_rate": 7.378157169486922e-08, "loss": 1.2486, "step": 48251 }, { "epoch": 3.7815047021943573, "grad_norm": 2.7769267559051514, "learning_rate": 7.372880259320536e-08, "loss": 1.2008, "step": 48252 }, { "epoch": 3.7815830721003136, "grad_norm": 4.150928497314453, "learning_rate": 7.367605222854735e-08, "loss": 1.135, "step": 48253 }, { "epoch": 3.7816614420062695, "grad_norm": 3.4639899730682373, "learning_rate": 7.3623320601095e-08, "loss": 1.2987, "step": 48254 }, { "epoch": 3.781739811912226, "grad_norm": 3.2511181831359863, "learning_rate": 7.357060771104929e-08, "loss": 1.4013, "step": 48255 }, { "epoch": 3.7818181818181817, "grad_norm": 4.631279945373535, "learning_rate": 7.35179135586106e-08, "loss": 1.4193, "step": 48256 }, { "epoch": 3.781896551724138, "grad_norm": 4.248497009277344, "learning_rate": 7.346523814397933e-08, "loss": 1.4855, "step": 48257 }, { "epoch": 3.781974921630094, "grad_norm": 9.706204414367676, "learning_rate": 7.341258146735641e-08, "loss": 1.7901, "step": 48258 }, { "epoch": 3.7820532915360503, "grad_norm": 4.986440658569336, "learning_rate": 7.33599435289406e-08, "loss": 1.2431, "step": 48259 }, { "epoch": 3.782131661442006, "grad_norm": 3.3119256496429443, "learning_rate": 7.330732432893395e-08, "loss": 0.9245, "step": 48260 }, { "epoch": 3.7822100313479625, "grad_norm": 4.467512607574463, "learning_rate": 7.325472386753519e-08, "loss": 1.3627, "step": 48261 }, { "epoch": 3.7822884012539184, "grad_norm": 6.80258846282959, "learning_rate": 7.320214214494525e-08, "loss": 1.3407, "step": 48262 }, { "epoch": 3.7823667711598747, "grad_norm": 3.3728809356689453, "learning_rate": 7.314957916136345e-08, "loss": 1.2081, "step": 48263 }, { "epoch": 3.7824451410658306, "grad_norm": 2.584256172180176, "learning_rate": 7.309703491699017e-08, "loss": 1.0372, "step": 48264 }, { "epoch": 3.782523510971787, "grad_norm": 3.8142166137695312, "learning_rate": 7.304450941202579e-08, "loss": 1.341, "step": 48265 }, { "epoch": 3.782601880877743, "grad_norm": 3.9980342388153076, "learning_rate": 7.299200264666905e-08, "loss": 1.3236, "step": 48266 }, { "epoch": 3.782680250783699, "grad_norm": 3.1754374504089355, "learning_rate": 7.293951462111981e-08, "loss": 1.3718, "step": 48267 }, { "epoch": 3.782758620689655, "grad_norm": 2.724186420440674, "learning_rate": 7.288704533557733e-08, "loss": 0.9801, "step": 48268 }, { "epoch": 3.7828369905956114, "grad_norm": 2.6024270057678223, "learning_rate": 7.283459479024313e-08, "loss": 1.2862, "step": 48269 }, { "epoch": 3.7829153605015673, "grad_norm": 2.361550807952881, "learning_rate": 7.278216298531482e-08, "loss": 1.2924, "step": 48270 }, { "epoch": 3.7829937304075236, "grad_norm": 5.369094371795654, "learning_rate": 7.272974992099224e-08, "loss": 1.3018, "step": 48271 }, { "epoch": 3.7830721003134795, "grad_norm": 6.614899635314941, "learning_rate": 7.267735559747524e-08, "loss": 1.3535, "step": 48272 }, { "epoch": 3.783150470219436, "grad_norm": 2.7213432788848877, "learning_rate": 7.262498001496254e-08, "loss": 1.4255, "step": 48273 }, { "epoch": 3.7832288401253917, "grad_norm": 6.025166034698486, "learning_rate": 7.257262317365343e-08, "loss": 1.5424, "step": 48274 }, { "epoch": 3.783307210031348, "grad_norm": 3.8182475566864014, "learning_rate": 7.252028507374719e-08, "loss": 1.4805, "step": 48275 }, { "epoch": 3.783385579937304, "grad_norm": 3.915461301803589, "learning_rate": 7.246796571544313e-08, "loss": 1.4191, "step": 48276 }, { "epoch": 3.7834639498432603, "grad_norm": 5.8244709968566895, "learning_rate": 7.241566509893938e-08, "loss": 1.2366, "step": 48277 }, { "epoch": 3.783542319749216, "grad_norm": 4.2633957862854, "learning_rate": 7.236338322443582e-08, "loss": 1.6518, "step": 48278 }, { "epoch": 3.7836206896551725, "grad_norm": 4.832875728607178, "learning_rate": 7.231112009213115e-08, "loss": 1.0568, "step": 48279 }, { "epoch": 3.7836990595611284, "grad_norm": 13.042773246765137, "learning_rate": 7.225887570222412e-08, "loss": 1.9664, "step": 48280 }, { "epoch": 3.7837774294670847, "grad_norm": 2.9746806621551514, "learning_rate": 7.220665005491346e-08, "loss": 0.8503, "step": 48281 }, { "epoch": 3.7838557993730406, "grad_norm": 6.50165319442749, "learning_rate": 7.215444315039677e-08, "loss": 1.4161, "step": 48282 }, { "epoch": 3.783934169278997, "grad_norm": 4.049037933349609, "learning_rate": 7.210225498887335e-08, "loss": 1.3373, "step": 48283 }, { "epoch": 3.784012539184953, "grad_norm": 3.107447624206543, "learning_rate": 7.205008557054305e-08, "loss": 1.4756, "step": 48284 }, { "epoch": 3.784090909090909, "grad_norm": 4.076761245727539, "learning_rate": 7.199793489560237e-08, "loss": 1.1892, "step": 48285 }, { "epoch": 3.784169278996865, "grad_norm": 3.5078542232513428, "learning_rate": 7.194580296425002e-08, "loss": 1.2833, "step": 48286 }, { "epoch": 3.7842476489028214, "grad_norm": 2.745412826538086, "learning_rate": 7.189368977668476e-08, "loss": 1.748, "step": 48287 }, { "epoch": 3.7843260188087773, "grad_norm": 2.4819283485412598, "learning_rate": 7.18415953331053e-08, "loss": 1.601, "step": 48288 }, { "epoch": 3.7844043887147336, "grad_norm": 4.640247821807861, "learning_rate": 7.178951963370873e-08, "loss": 1.4557, "step": 48289 }, { "epoch": 3.7844827586206895, "grad_norm": 3.8732123374938965, "learning_rate": 7.173746267869375e-08, "loss": 1.1032, "step": 48290 }, { "epoch": 3.784561128526646, "grad_norm": 9.621603965759277, "learning_rate": 7.168542446825744e-08, "loss": 1.5301, "step": 48291 }, { "epoch": 3.7846394984326017, "grad_norm": 2.5536577701568604, "learning_rate": 7.163340500259907e-08, "loss": 1.2199, "step": 48292 }, { "epoch": 3.784717868338558, "grad_norm": 5.060555934906006, "learning_rate": 7.158140428191573e-08, "loss": 1.5074, "step": 48293 }, { "epoch": 3.784796238244514, "grad_norm": 2.5315358638763428, "learning_rate": 7.152942230640558e-08, "loss": 1.5263, "step": 48294 }, { "epoch": 3.7848746081504703, "grad_norm": 3.7015857696533203, "learning_rate": 7.147745907626569e-08, "loss": 1.3574, "step": 48295 }, { "epoch": 3.7849529780564266, "grad_norm": 4.335956573486328, "learning_rate": 7.142551459169422e-08, "loss": 1.7332, "step": 48296 }, { "epoch": 3.7850313479623825, "grad_norm": 5.509288787841797, "learning_rate": 7.137358885288825e-08, "loss": 1.2269, "step": 48297 }, { "epoch": 3.7851097178683384, "grad_norm": 3.8341689109802246, "learning_rate": 7.132168186004651e-08, "loss": 1.5055, "step": 48298 }, { "epoch": 3.7851880877742947, "grad_norm": 3.733515739440918, "learning_rate": 7.126979361336495e-08, "loss": 1.327, "step": 48299 }, { "epoch": 3.785266457680251, "grad_norm": 7.062380790710449, "learning_rate": 7.121792411304173e-08, "loss": 1.648, "step": 48300 }, { "epoch": 3.785344827586207, "grad_norm": 8.83566665649414, "learning_rate": 7.11660733592734e-08, "loss": 1.3429, "step": 48301 }, { "epoch": 3.785423197492163, "grad_norm": 3.328681707382202, "learning_rate": 7.111424135225863e-08, "loss": 1.406, "step": 48302 }, { "epoch": 3.785501567398119, "grad_norm": 4.257035255432129, "learning_rate": 7.106242809219343e-08, "loss": 1.6813, "step": 48303 }, { "epoch": 3.7855799373040755, "grad_norm": 3.8609201908111572, "learning_rate": 7.101063357927484e-08, "loss": 1.2398, "step": 48304 }, { "epoch": 3.7856583072100314, "grad_norm": 2.0196056365966797, "learning_rate": 7.095885781369993e-08, "loss": 1.0605, "step": 48305 }, { "epoch": 3.7857366771159873, "grad_norm": 1.9417482614517212, "learning_rate": 7.090710079566577e-08, "loss": 1.2959, "step": 48306 }, { "epoch": 3.7858150470219436, "grad_norm": 2.783033609390259, "learning_rate": 7.085536252536995e-08, "loss": 1.477, "step": 48307 }, { "epoch": 3.7858934169279, "grad_norm": 3.8204238414764404, "learning_rate": 7.08036430030079e-08, "loss": 1.244, "step": 48308 }, { "epoch": 3.785971786833856, "grad_norm": 4.131162166595459, "learning_rate": 7.075194222877723e-08, "loss": 1.2523, "step": 48309 }, { "epoch": 3.7860501567398117, "grad_norm": 2.3102548122406006, "learning_rate": 7.070026020287446e-08, "loss": 1.4732, "step": 48310 }, { "epoch": 3.786128526645768, "grad_norm": 3.6439871788024902, "learning_rate": 7.064859692549609e-08, "loss": 1.562, "step": 48311 }, { "epoch": 3.7862068965517244, "grad_norm": 3.290972948074341, "learning_rate": 7.059695239683806e-08, "loss": 1.2263, "step": 48312 }, { "epoch": 3.7862852664576803, "grad_norm": 7.08635139465332, "learning_rate": 7.054532661709801e-08, "loss": 0.8908, "step": 48313 }, { "epoch": 3.786363636363636, "grad_norm": 5.464747905731201, "learning_rate": 7.049371958647133e-08, "loss": 1.5999, "step": 48314 }, { "epoch": 3.7864420062695925, "grad_norm": 2.9156453609466553, "learning_rate": 7.044213130515454e-08, "loss": 1.4795, "step": 48315 }, { "epoch": 3.786520376175549, "grad_norm": 3.483344078063965, "learning_rate": 7.039056177334358e-08, "loss": 1.1525, "step": 48316 }, { "epoch": 3.7865987460815047, "grad_norm": 3.9723663330078125, "learning_rate": 7.033901099123552e-08, "loss": 1.3793, "step": 48317 }, { "epoch": 3.7866771159874606, "grad_norm": 2.268181085586548, "learning_rate": 7.028747895902576e-08, "loss": 1.2182, "step": 48318 }, { "epoch": 3.786755485893417, "grad_norm": 2.1364986896514893, "learning_rate": 7.023596567690971e-08, "loss": 1.2807, "step": 48319 }, { "epoch": 3.7868338557993733, "grad_norm": 6.493214130401611, "learning_rate": 7.018447114508443e-08, "loss": 1.3086, "step": 48320 }, { "epoch": 3.786912225705329, "grad_norm": 3.798330307006836, "learning_rate": 7.013299536374585e-08, "loss": 1.3901, "step": 48321 }, { "epoch": 3.786990595611285, "grad_norm": 3.110893487930298, "learning_rate": 7.008153833308828e-08, "loss": 1.6552, "step": 48322 }, { "epoch": 3.7870689655172414, "grad_norm": 3.2567131519317627, "learning_rate": 7.003010005330879e-08, "loss": 1.7169, "step": 48323 }, { "epoch": 3.7871473354231977, "grad_norm": 7.894559383392334, "learning_rate": 6.997868052460222e-08, "loss": 1.0056, "step": 48324 }, { "epoch": 3.7872257053291536, "grad_norm": 10.325935363769531, "learning_rate": 6.992727974716563e-08, "loss": 1.4345, "step": 48325 }, { "epoch": 3.7873040752351095, "grad_norm": 3.120680093765259, "learning_rate": 6.98758977211922e-08, "loss": 1.4088, "step": 48326 }, { "epoch": 3.787382445141066, "grad_norm": 2.6498451232910156, "learning_rate": 6.982453444687897e-08, "loss": 1.1337, "step": 48327 }, { "epoch": 3.787460815047022, "grad_norm": 3.4515836238861084, "learning_rate": 6.97731899244214e-08, "loss": 1.47, "step": 48328 }, { "epoch": 3.787539184952978, "grad_norm": 2.57906174659729, "learning_rate": 6.972186415401371e-08, "loss": 1.5249, "step": 48329 }, { "epoch": 3.787617554858934, "grad_norm": 3.8676674365997314, "learning_rate": 6.967055713585192e-08, "loss": 1.1616, "step": 48330 }, { "epoch": 3.7876959247648903, "grad_norm": 6.1985321044921875, "learning_rate": 6.961926887013082e-08, "loss": 0.7872, "step": 48331 }, { "epoch": 3.7877742946708466, "grad_norm": 2.6003196239471436, "learning_rate": 6.956799935704583e-08, "loss": 1.3005, "step": 48332 }, { "epoch": 3.7878526645768025, "grad_norm": 2.922107219696045, "learning_rate": 6.951674859679125e-08, "loss": 1.3755, "step": 48333 }, { "epoch": 3.7879310344827584, "grad_norm": 4.093677520751953, "learning_rate": 6.946551658956302e-08, "loss": 1.5432, "step": 48334 }, { "epoch": 3.7880094043887147, "grad_norm": 3.3523433208465576, "learning_rate": 6.941430333555543e-08, "loss": 1.1663, "step": 48335 }, { "epoch": 3.788087774294671, "grad_norm": 3.512915849685669, "learning_rate": 6.936310883496333e-08, "loss": 1.3455, "step": 48336 }, { "epoch": 3.788166144200627, "grad_norm": 2.479719638824463, "learning_rate": 6.931193308798156e-08, "loss": 1.4427, "step": 48337 }, { "epoch": 3.788244514106583, "grad_norm": 3.80045223236084, "learning_rate": 6.926077609480387e-08, "loss": 1.0039, "step": 48338 }, { "epoch": 3.788322884012539, "grad_norm": 3.3500826358795166, "learning_rate": 6.920963785562618e-08, "loss": 1.4728, "step": 48339 }, { "epoch": 3.7884012539184955, "grad_norm": 3.316549062728882, "learning_rate": 6.91585183706428e-08, "loss": 1.5018, "step": 48340 }, { "epoch": 3.7884796238244514, "grad_norm": 3.1307027339935303, "learning_rate": 6.910741764004747e-08, "loss": 1.3208, "step": 48341 }, { "epoch": 3.7885579937304072, "grad_norm": 4.61509370803833, "learning_rate": 6.905633566403447e-08, "loss": 1.589, "step": 48342 }, { "epoch": 3.7886363636363636, "grad_norm": 2.8677425384521484, "learning_rate": 6.900527244279809e-08, "loss": 1.3356, "step": 48343 }, { "epoch": 3.78871473354232, "grad_norm": 5.371622562408447, "learning_rate": 6.895422797653428e-08, "loss": 1.8807, "step": 48344 }, { "epoch": 3.788793103448276, "grad_norm": 2.39892315864563, "learning_rate": 6.890320226543512e-08, "loss": 1.3214, "step": 48345 }, { "epoch": 3.788871473354232, "grad_norm": 2.1003024578094482, "learning_rate": 6.885219530969545e-08, "loss": 1.1959, "step": 48346 }, { "epoch": 3.788949843260188, "grad_norm": 4.191393852233887, "learning_rate": 6.880120710950899e-08, "loss": 1.0573, "step": 48347 }, { "epoch": 3.7890282131661444, "grad_norm": 3.3228375911712646, "learning_rate": 6.875023766507005e-08, "loss": 1.5179, "step": 48348 }, { "epoch": 3.7891065830721002, "grad_norm": 2.629337787628174, "learning_rate": 6.869928697657236e-08, "loss": 1.126, "step": 48349 }, { "epoch": 3.7891849529780566, "grad_norm": 2.3551831245422363, "learning_rate": 6.86483550442102e-08, "loss": 1.3225, "step": 48350 }, { "epoch": 3.7892633228840125, "grad_norm": 6.2878031730651855, "learning_rate": 6.859744186817674e-08, "loss": 1.4428, "step": 48351 }, { "epoch": 3.789341692789969, "grad_norm": 2.4990170001983643, "learning_rate": 6.854654744866463e-08, "loss": 1.2978, "step": 48352 }, { "epoch": 3.7894200626959247, "grad_norm": 3.6481454372406006, "learning_rate": 6.849567178586925e-08, "loss": 1.3262, "step": 48353 }, { "epoch": 3.789498432601881, "grad_norm": 2.776773691177368, "learning_rate": 6.844481487998322e-08, "loss": 1.5347, "step": 48354 }, { "epoch": 3.789576802507837, "grad_norm": 5.208281517028809, "learning_rate": 6.83939767312003e-08, "loss": 1.873, "step": 48355 }, { "epoch": 3.7896551724137932, "grad_norm": 5.057448863983154, "learning_rate": 6.834315733971363e-08, "loss": 1.3206, "step": 48356 }, { "epoch": 3.789733542319749, "grad_norm": 4.4819746017456055, "learning_rate": 6.829235670571644e-08, "loss": 1.5536, "step": 48357 }, { "epoch": 3.7898119122257055, "grad_norm": 5.457253456115723, "learning_rate": 6.824157482940186e-08, "loss": 1.4974, "step": 48358 }, { "epoch": 3.7898902821316613, "grad_norm": 4.2749738693237305, "learning_rate": 6.819081171096365e-08, "loss": 1.6213, "step": 48359 }, { "epoch": 3.7899686520376177, "grad_norm": 2.254955530166626, "learning_rate": 6.814006735059442e-08, "loss": 1.2271, "step": 48360 }, { "epoch": 3.7900470219435736, "grad_norm": 2.558350086212158, "learning_rate": 6.80893417484868e-08, "loss": 1.2996, "step": 48361 }, { "epoch": 3.79012539184953, "grad_norm": 3.818492889404297, "learning_rate": 6.803863490483453e-08, "loss": 1.2299, "step": 48362 }, { "epoch": 3.790203761755486, "grad_norm": 4.013673782348633, "learning_rate": 6.798794681982968e-08, "loss": 1.069, "step": 48363 }, { "epoch": 3.790282131661442, "grad_norm": 4.249749183654785, "learning_rate": 6.793727749366541e-08, "loss": 1.2936, "step": 48364 }, { "epoch": 3.790360501567398, "grad_norm": 3.612593650817871, "learning_rate": 6.788662692653491e-08, "loss": 1.0644, "step": 48365 }, { "epoch": 3.7904388714733543, "grad_norm": 6.401717185974121, "learning_rate": 6.783599511863025e-08, "loss": 1.307, "step": 48366 }, { "epoch": 3.7905172413793102, "grad_norm": 2.979274272918701, "learning_rate": 6.778538207014407e-08, "loss": 1.5457, "step": 48367 }, { "epoch": 3.7905956112852666, "grad_norm": 2.7854013442993164, "learning_rate": 6.773478778126896e-08, "loss": 1.1008, "step": 48368 }, { "epoch": 3.7906739811912225, "grad_norm": 2.24532413482666, "learning_rate": 6.7684212252197e-08, "loss": 1.3859, "step": 48369 }, { "epoch": 3.790752351097179, "grad_norm": 3.0382070541381836, "learning_rate": 6.76336554831214e-08, "loss": 1.2478, "step": 48370 }, { "epoch": 3.7908307210031347, "grad_norm": 2.738602638244629, "learning_rate": 6.758311747423363e-08, "loss": 1.1713, "step": 48371 }, { "epoch": 3.790909090909091, "grad_norm": 3.0428290367126465, "learning_rate": 6.753259822572634e-08, "loss": 1.5501, "step": 48372 }, { "epoch": 3.790987460815047, "grad_norm": 2.7238669395446777, "learning_rate": 6.748209773779157e-08, "loss": 1.3243, "step": 48373 }, { "epoch": 3.7910658307210032, "grad_norm": 4.3485894203186035, "learning_rate": 6.743161601062198e-08, "loss": 1.6714, "step": 48374 }, { "epoch": 3.791144200626959, "grad_norm": 3.0053491592407227, "learning_rate": 6.738115304440851e-08, "loss": 1.6278, "step": 48375 }, { "epoch": 3.7912225705329154, "grad_norm": 2.785918712615967, "learning_rate": 6.733070883934323e-08, "loss": 1.0743, "step": 48376 }, { "epoch": 3.7913009404388713, "grad_norm": 5.013522148132324, "learning_rate": 6.728028339561877e-08, "loss": 1.2225, "step": 48377 }, { "epoch": 3.7913793103448277, "grad_norm": 4.4051408767700195, "learning_rate": 6.722987671342607e-08, "loss": 1.2835, "step": 48378 }, { "epoch": 3.7914576802507836, "grad_norm": 3.746493101119995, "learning_rate": 6.717948879295777e-08, "loss": 1.4904, "step": 48379 }, { "epoch": 3.79153605015674, "grad_norm": 7.240326404571533, "learning_rate": 6.712911963440482e-08, "loss": 1.4144, "step": 48380 }, { "epoch": 3.791614420062696, "grad_norm": 4.628800868988037, "learning_rate": 6.70787692379593e-08, "loss": 1.4752, "step": 48381 }, { "epoch": 3.791692789968652, "grad_norm": 3.5272111892700195, "learning_rate": 6.702843760381217e-08, "loss": 1.2014, "step": 48382 }, { "epoch": 3.791771159874608, "grad_norm": 4.864612102508545, "learning_rate": 6.697812473215493e-08, "loss": 1.0901, "step": 48383 }, { "epoch": 3.7918495297805643, "grad_norm": 3.0546276569366455, "learning_rate": 6.692783062317964e-08, "loss": 1.0235, "step": 48384 }, { "epoch": 3.79192789968652, "grad_norm": 3.58618426322937, "learning_rate": 6.687755527707673e-08, "loss": 1.2276, "step": 48385 }, { "epoch": 3.7920062695924766, "grad_norm": 3.3366706371307373, "learning_rate": 6.68272986940377e-08, "loss": 1.2162, "step": 48386 }, { "epoch": 3.7920846394984324, "grad_norm": 3.0563087463378906, "learning_rate": 6.677706087425407e-08, "loss": 1.329, "step": 48387 }, { "epoch": 3.7921630094043888, "grad_norm": 4.491363525390625, "learning_rate": 6.672684181791678e-08, "loss": 1.2321, "step": 48388 }, { "epoch": 3.7922413793103447, "grad_norm": 3.667201042175293, "learning_rate": 6.667664152521624e-08, "loss": 1.0737, "step": 48389 }, { "epoch": 3.792319749216301, "grad_norm": 12.055326461791992, "learning_rate": 6.662645999634399e-08, "loss": 1.5828, "step": 48390 }, { "epoch": 3.7923981191222573, "grad_norm": 4.633363723754883, "learning_rate": 6.657629723149095e-08, "loss": 1.635, "step": 48391 }, { "epoch": 3.792476489028213, "grad_norm": 3.3627216815948486, "learning_rate": 6.65261532308481e-08, "loss": 1.2628, "step": 48392 }, { "epoch": 3.792554858934169, "grad_norm": 3.786637783050537, "learning_rate": 6.647602799460473e-08, "loss": 1.3646, "step": 48393 }, { "epoch": 3.7926332288401254, "grad_norm": 2.854158639907837, "learning_rate": 6.642592152295291e-08, "loss": 0.8079, "step": 48394 }, { "epoch": 3.7927115987460818, "grad_norm": 3.311126708984375, "learning_rate": 6.637583381608304e-08, "loss": 0.8668, "step": 48395 }, { "epoch": 3.7927899686520377, "grad_norm": 4.559065818786621, "learning_rate": 6.632576487418551e-08, "loss": 0.6963, "step": 48396 }, { "epoch": 3.7928683385579935, "grad_norm": 3.293203115463257, "learning_rate": 6.627571469745075e-08, "loss": 1.0988, "step": 48397 }, { "epoch": 3.79294670846395, "grad_norm": 3.6560707092285156, "learning_rate": 6.62256832860686e-08, "loss": 1.3329, "step": 48398 }, { "epoch": 3.793025078369906, "grad_norm": 3.5740222930908203, "learning_rate": 6.617567064023e-08, "loss": 1.0359, "step": 48399 }, { "epoch": 3.793103448275862, "grad_norm": 2.9595823287963867, "learning_rate": 6.612567676012538e-08, "loss": 1.348, "step": 48400 }, { "epoch": 3.793181818181818, "grad_norm": 4.647731781005859, "learning_rate": 6.6075701645944e-08, "loss": 1.3252, "step": 48401 }, { "epoch": 3.7932601880877743, "grad_norm": 5.487871170043945, "learning_rate": 6.602574529787631e-08, "loss": 1.4148, "step": 48402 }, { "epoch": 3.7933385579937307, "grad_norm": 2.6338608264923096, "learning_rate": 6.597580771611267e-08, "loss": 0.9593, "step": 48403 }, { "epoch": 3.7934169278996865, "grad_norm": 3.9977529048919678, "learning_rate": 6.592588890084295e-08, "loss": 1.1778, "step": 48404 }, { "epoch": 3.7934952978056424, "grad_norm": 11.021599769592285, "learning_rate": 6.587598885225644e-08, "loss": 1.5593, "step": 48405 }, { "epoch": 3.7935736677115988, "grad_norm": 2.4690165519714355, "learning_rate": 6.582610757054353e-08, "loss": 1.1589, "step": 48406 }, { "epoch": 3.793652037617555, "grad_norm": 3.1513662338256836, "learning_rate": 6.577624505589408e-08, "loss": 1.5211, "step": 48407 }, { "epoch": 3.793730407523511, "grad_norm": 11.383103370666504, "learning_rate": 6.572640130849684e-08, "loss": 0.7028, "step": 48408 }, { "epoch": 3.793808777429467, "grad_norm": 2.7609572410583496, "learning_rate": 6.567657632854219e-08, "loss": 0.7091, "step": 48409 }, { "epoch": 3.793887147335423, "grad_norm": 2.640028476715088, "learning_rate": 6.562677011621998e-08, "loss": 0.9134, "step": 48410 }, { "epoch": 3.7939655172413795, "grad_norm": 6.202728748321533, "learning_rate": 6.557698267171897e-08, "loss": 0.8564, "step": 48411 }, { "epoch": 3.7940438871473354, "grad_norm": 3.2775607109069824, "learning_rate": 6.552721399522788e-08, "loss": 1.3355, "step": 48412 }, { "epoch": 3.7941222570532913, "grad_norm": 3.186497211456299, "learning_rate": 6.547746408693712e-08, "loss": 1.3376, "step": 48413 }, { "epoch": 3.7942006269592476, "grad_norm": 3.4725568294525146, "learning_rate": 6.542773294703542e-08, "loss": 0.9536, "step": 48414 }, { "epoch": 3.794278996865204, "grad_norm": 3.1319057941436768, "learning_rate": 6.537802057571264e-08, "loss": 1.5483, "step": 48415 }, { "epoch": 3.79435736677116, "grad_norm": 18.976858139038086, "learning_rate": 6.532832697315694e-08, "loss": 1.002, "step": 48416 }, { "epoch": 3.7944357366771158, "grad_norm": 3.0877819061279297, "learning_rate": 6.527865213955764e-08, "loss": 1.4508, "step": 48417 }, { "epoch": 3.794514106583072, "grad_norm": 6.410364627838135, "learning_rate": 6.522899607510347e-08, "loss": 1.7912, "step": 48418 }, { "epoch": 3.7945924764890284, "grad_norm": 2.3967936038970947, "learning_rate": 6.517935877998372e-08, "loss": 1.3773, "step": 48419 }, { "epoch": 3.7946708463949843, "grad_norm": 3.090886116027832, "learning_rate": 6.512974025438711e-08, "loss": 0.6973, "step": 48420 }, { "epoch": 3.79474921630094, "grad_norm": 3.080052375793457, "learning_rate": 6.508014049850187e-08, "loss": 1.3079, "step": 48421 }, { "epoch": 3.7948275862068965, "grad_norm": 3.0136172771453857, "learning_rate": 6.50305595125178e-08, "loss": 1.363, "step": 48422 }, { "epoch": 3.794905956112853, "grad_norm": 4.353782653808594, "learning_rate": 6.4980997296622e-08, "loss": 1.6631, "step": 48423 }, { "epoch": 3.7949843260188088, "grad_norm": 2.7558140754699707, "learning_rate": 6.49314538510043e-08, "loss": 1.2385, "step": 48424 }, { "epoch": 3.7950626959247646, "grad_norm": 2.9387612342834473, "learning_rate": 6.488192917585179e-08, "loss": 1.6468, "step": 48425 }, { "epoch": 3.795141065830721, "grad_norm": 3.788038492202759, "learning_rate": 6.48324232713543e-08, "loss": 1.2317, "step": 48426 }, { "epoch": 3.7952194357366773, "grad_norm": 3.57594633102417, "learning_rate": 6.478293613769893e-08, "loss": 1.022, "step": 48427 }, { "epoch": 3.795297805642633, "grad_norm": 3.0228259563446045, "learning_rate": 6.473346777507495e-08, "loss": 1.3924, "step": 48428 }, { "epoch": 3.795376175548589, "grad_norm": 3.353168487548828, "learning_rate": 6.468401818366943e-08, "loss": 1.4494, "step": 48429 }, { "epoch": 3.7954545454545454, "grad_norm": 6.192233562469482, "learning_rate": 6.463458736367111e-08, "loss": 1.6825, "step": 48430 }, { "epoch": 3.7955329153605017, "grad_norm": 4.1584248542785645, "learning_rate": 6.45851753152682e-08, "loss": 1.7061, "step": 48431 }, { "epoch": 3.7956112852664576, "grad_norm": 2.677720546722412, "learning_rate": 6.453578203864718e-08, "loss": 1.2337, "step": 48432 }, { "epoch": 3.7956896551724135, "grad_norm": 4.955305099487305, "learning_rate": 6.448640753399848e-08, "loss": 1.2008, "step": 48433 }, { "epoch": 3.79576802507837, "grad_norm": 3.092811346054077, "learning_rate": 6.44370518015075e-08, "loss": 1.2126, "step": 48434 }, { "epoch": 3.795846394984326, "grad_norm": 2.8282623291015625, "learning_rate": 6.438771484136352e-08, "loss": 1.2019, "step": 48435 }, { "epoch": 3.795924764890282, "grad_norm": 2.3223750591278076, "learning_rate": 6.433839665375308e-08, "loss": 1.2216, "step": 48436 }, { "epoch": 3.7960031347962384, "grad_norm": 5.5674238204956055, "learning_rate": 6.428909723886435e-08, "loss": 1.3489, "step": 48437 }, { "epoch": 3.7960815047021943, "grad_norm": 9.221097946166992, "learning_rate": 6.423981659688495e-08, "loss": 2.0147, "step": 48438 }, { "epoch": 3.7961598746081506, "grad_norm": 2.4396519660949707, "learning_rate": 6.419055472800195e-08, "loss": 1.2912, "step": 48439 }, { "epoch": 3.7962382445141065, "grad_norm": 13.647982597351074, "learning_rate": 6.4141311632403e-08, "loss": 0.8244, "step": 48440 }, { "epoch": 3.796316614420063, "grad_norm": 4.346316814422607, "learning_rate": 6.409208731027517e-08, "loss": 1.3775, "step": 48441 }, { "epoch": 3.7963949843260187, "grad_norm": 2.995225191116333, "learning_rate": 6.404288176180606e-08, "loss": 1.1503, "step": 48442 }, { "epoch": 3.796473354231975, "grad_norm": 12.102499961853027, "learning_rate": 6.399369498718222e-08, "loss": 1.2735, "step": 48443 }, { "epoch": 3.796551724137931, "grad_norm": 15.79361343383789, "learning_rate": 6.394452698659126e-08, "loss": 1.008, "step": 48444 }, { "epoch": 3.7966300940438873, "grad_norm": 2.9433774948120117, "learning_rate": 6.389537776022026e-08, "loss": 1.3463, "step": 48445 }, { "epoch": 3.796708463949843, "grad_norm": 15.940604209899902, "learning_rate": 6.384624730825517e-08, "loss": 1.0455, "step": 48446 }, { "epoch": 3.7967868338557995, "grad_norm": 8.805000305175781, "learning_rate": 6.379713563088364e-08, "loss": 0.8558, "step": 48447 }, { "epoch": 3.7968652037617554, "grad_norm": 2.8294219970703125, "learning_rate": 6.374804272829326e-08, "loss": 1.3091, "step": 48448 }, { "epoch": 3.7969435736677117, "grad_norm": 7.349059104919434, "learning_rate": 6.369896860066948e-08, "loss": 2.3276, "step": 48449 }, { "epoch": 3.7970219435736676, "grad_norm": 3.644001007080078, "learning_rate": 6.364991324819936e-08, "loss": 1.0798, "step": 48450 }, { "epoch": 3.797100313479624, "grad_norm": 2.7205042839050293, "learning_rate": 6.360087667106884e-08, "loss": 1.4006, "step": 48451 }, { "epoch": 3.79717868338558, "grad_norm": 3.8090081214904785, "learning_rate": 6.355185886946614e-08, "loss": 1.6524, "step": 48452 }, { "epoch": 3.797257053291536, "grad_norm": 4.595041751861572, "learning_rate": 6.350285984357607e-08, "loss": 1.7975, "step": 48453 }, { "epoch": 3.797335423197492, "grad_norm": 3.9978549480438232, "learning_rate": 6.345387959358574e-08, "loss": 1.325, "step": 48454 }, { "epoch": 3.7974137931034484, "grad_norm": 2.4415106773376465, "learning_rate": 6.340491811968164e-08, "loss": 1.442, "step": 48455 }, { "epoch": 3.7974921630094043, "grad_norm": 3.4342620372772217, "learning_rate": 6.335597542204919e-08, "loss": 1.3915, "step": 48456 }, { "epoch": 3.7975705329153606, "grad_norm": 3.9571564197540283, "learning_rate": 6.33070515008749e-08, "loss": 1.3775, "step": 48457 }, { "epoch": 3.7976489028213165, "grad_norm": 4.878408908843994, "learning_rate": 6.32581463563453e-08, "loss": 1.6918, "step": 48458 }, { "epoch": 3.797727272727273, "grad_norm": 2.2648115158081055, "learning_rate": 6.32092599886458e-08, "loss": 1.3155, "step": 48459 }, { "epoch": 3.7978056426332287, "grad_norm": 2.3654160499572754, "learning_rate": 6.316039239796235e-08, "loss": 0.8586, "step": 48460 }, { "epoch": 3.797884012539185, "grad_norm": 5.632364749908447, "learning_rate": 6.311154358448147e-08, "loss": 1.3763, "step": 48461 }, { "epoch": 3.797962382445141, "grad_norm": 3.254455327987671, "learning_rate": 6.3062713548388e-08, "loss": 1.2562, "step": 48462 }, { "epoch": 3.7980407523510973, "grad_norm": 3.8047025203704834, "learning_rate": 6.301390228986903e-08, "loss": 1.0051, "step": 48463 }, { "epoch": 3.798119122257053, "grad_norm": 2.344005823135376, "learning_rate": 6.296510980910885e-08, "loss": 1.2993, "step": 48464 }, { "epoch": 3.7981974921630095, "grad_norm": 2.5131542682647705, "learning_rate": 6.291633610629344e-08, "loss": 0.941, "step": 48465 }, { "epoch": 3.7982758620689654, "grad_norm": 4.657128810882568, "learning_rate": 6.286758118160874e-08, "loss": 1.0616, "step": 48466 }, { "epoch": 3.7983542319749217, "grad_norm": 2.3505916595458984, "learning_rate": 6.281884503524017e-08, "loss": 1.2873, "step": 48467 }, { "epoch": 3.7984326018808776, "grad_norm": 9.799356460571289, "learning_rate": 6.277012766737256e-08, "loss": 0.7211, "step": 48468 }, { "epoch": 3.798510971786834, "grad_norm": 3.3234121799468994, "learning_rate": 6.272142907819135e-08, "loss": 1.3366, "step": 48469 }, { "epoch": 3.79858934169279, "grad_norm": 9.015388488769531, "learning_rate": 6.267274926788137e-08, "loss": 1.4104, "step": 48470 }, { "epoch": 3.798667711598746, "grad_norm": 3.0018675327301025, "learning_rate": 6.262408823662913e-08, "loss": 1.2756, "step": 48471 }, { "epoch": 3.798746081504702, "grad_norm": 15.874992370605469, "learning_rate": 6.25754459846184e-08, "loss": 0.8553, "step": 48472 }, { "epoch": 3.7988244514106584, "grad_norm": 2.7505645751953125, "learning_rate": 6.252682251203457e-08, "loss": 1.358, "step": 48473 }, { "epoch": 3.7989028213166143, "grad_norm": 3.260711431503296, "learning_rate": 6.247821781906305e-08, "loss": 0.8974, "step": 48474 }, { "epoch": 3.7989811912225706, "grad_norm": 3.073782444000244, "learning_rate": 6.242963190588813e-08, "loss": 1.1844, "step": 48475 }, { "epoch": 3.7990595611285265, "grad_norm": 3.7726356983184814, "learning_rate": 6.238106477269468e-08, "loss": 1.5633, "step": 48476 }, { "epoch": 3.799137931034483, "grad_norm": 4.2920331954956055, "learning_rate": 6.233251641966753e-08, "loss": 1.8729, "step": 48477 }, { "epoch": 3.7992163009404387, "grad_norm": 5.3494367599487305, "learning_rate": 6.228398684699211e-08, "loss": 1.5164, "step": 48478 }, { "epoch": 3.799294670846395, "grad_norm": 3.3485512733459473, "learning_rate": 6.223547605485103e-08, "loss": 1.4414, "step": 48479 }, { "epoch": 3.799373040752351, "grad_norm": 2.673067092895508, "learning_rate": 6.218698404343081e-08, "loss": 1.1232, "step": 48480 }, { "epoch": 3.7994514106583073, "grad_norm": 3.384370803833008, "learning_rate": 6.213851081291467e-08, "loss": 1.4156, "step": 48481 }, { "epoch": 3.799529780564263, "grad_norm": 3.1265690326690674, "learning_rate": 6.209005636348798e-08, "loss": 1.2922, "step": 48482 }, { "epoch": 3.7996081504702195, "grad_norm": 3.0750935077667236, "learning_rate": 6.204162069533393e-08, "loss": 0.9525, "step": 48483 }, { "epoch": 3.7996865203761754, "grad_norm": 3.08651065826416, "learning_rate": 6.199320380863682e-08, "loss": 1.5655, "step": 48484 }, { "epoch": 3.7997648902821317, "grad_norm": 3.135241985321045, "learning_rate": 6.194480570358152e-08, "loss": 1.4618, "step": 48485 }, { "epoch": 3.799843260188088, "grad_norm": 2.9320249557495117, "learning_rate": 6.189642638035288e-08, "loss": 1.0989, "step": 48486 }, { "epoch": 3.799921630094044, "grad_norm": 3.6534011363983154, "learning_rate": 6.184806583913295e-08, "loss": 1.1542, "step": 48487 }, { "epoch": 3.8, "grad_norm": 2.684400796890259, "learning_rate": 6.17997240801066e-08, "loss": 0.5399, "step": 48488 }, { "epoch": 3.800078369905956, "grad_norm": 2.3794784545898438, "learning_rate": 6.175140110345756e-08, "loss": 1.2719, "step": 48489 }, { "epoch": 3.8001567398119125, "grad_norm": 6.952147006988525, "learning_rate": 6.170309690937015e-08, "loss": 1.4878, "step": 48490 }, { "epoch": 3.8002351097178684, "grad_norm": 7.158382892608643, "learning_rate": 6.165481149802755e-08, "loss": 0.9205, "step": 48491 }, { "epoch": 3.8003134796238243, "grad_norm": 5.130767345428467, "learning_rate": 6.160654486961349e-08, "loss": 1.2805, "step": 48492 }, { "epoch": 3.8003918495297806, "grad_norm": 2.1944589614868164, "learning_rate": 6.15582970243117e-08, "loss": 1.6207, "step": 48493 }, { "epoch": 3.800470219435737, "grad_norm": 3.1346278190612793, "learning_rate": 6.151006796230541e-08, "loss": 1.3351, "step": 48494 }, { "epoch": 3.800548589341693, "grad_norm": 9.78080940246582, "learning_rate": 6.146185768377833e-08, "loss": 1.2186, "step": 48495 }, { "epoch": 3.8006269592476487, "grad_norm": 2.674485683441162, "learning_rate": 6.141366618891364e-08, "loss": 1.0382, "step": 48496 }, { "epoch": 3.800705329153605, "grad_norm": 2.8585658073425293, "learning_rate": 6.136549347789511e-08, "loss": 1.1905, "step": 48497 }, { "epoch": 3.8007836990595614, "grad_norm": 17.59987449645996, "learning_rate": 6.131733955090535e-08, "loss": 0.4715, "step": 48498 }, { "epoch": 3.8008620689655173, "grad_norm": 4.375105381011963, "learning_rate": 6.126920440812756e-08, "loss": 1.0723, "step": 48499 }, { "epoch": 3.800940438871473, "grad_norm": 2.99573016166687, "learning_rate": 6.122108804974547e-08, "loss": 1.6608, "step": 48500 }, { "epoch": 3.8010188087774295, "grad_norm": 3.1657869815826416, "learning_rate": 6.117299047594172e-08, "loss": 1.2816, "step": 48501 }, { "epoch": 3.801097178683386, "grad_norm": 3.1833767890930176, "learning_rate": 6.112491168689839e-08, "loss": 1.3198, "step": 48502 }, { "epoch": 3.8011755485893417, "grad_norm": 5.05502986907959, "learning_rate": 6.107685168279976e-08, "loss": 1.5176, "step": 48503 }, { "epoch": 3.8012539184952976, "grad_norm": 7.27293586730957, "learning_rate": 6.102881046382792e-08, "loss": 0.9488, "step": 48504 }, { "epoch": 3.801332288401254, "grad_norm": 6.8441948890686035, "learning_rate": 6.09807880301655e-08, "loss": 1.5403, "step": 48505 }, { "epoch": 3.8014106583072103, "grad_norm": 4.309735298156738, "learning_rate": 6.093278438199568e-08, "loss": 1.4834, "step": 48506 }, { "epoch": 3.801489028213166, "grad_norm": 3.0624783039093018, "learning_rate": 6.088479951949999e-08, "loss": 1.6106, "step": 48507 }, { "epoch": 3.801567398119122, "grad_norm": 2.9291880130767822, "learning_rate": 6.083683344286273e-08, "loss": 1.3223, "step": 48508 }, { "epoch": 3.8016457680250784, "grad_norm": 3.4659552574157715, "learning_rate": 6.078888615226431e-08, "loss": 1.4221, "step": 48509 }, { "epoch": 3.8017241379310347, "grad_norm": 5.488886833190918, "learning_rate": 6.074095764788845e-08, "loss": 1.6015, "step": 48510 }, { "epoch": 3.8018025078369906, "grad_norm": 4.626292705535889, "learning_rate": 6.069304792991671e-08, "loss": 0.8202, "step": 48511 }, { "epoch": 3.8018808777429465, "grad_norm": 3.2488479614257812, "learning_rate": 6.064515699853168e-08, "loss": 0.9771, "step": 48512 }, { "epoch": 3.801959247648903, "grad_norm": 6.675846576690674, "learning_rate": 6.059728485391547e-08, "loss": 1.5592, "step": 48513 }, { "epoch": 3.802037617554859, "grad_norm": 2.772122859954834, "learning_rate": 6.05494314962507e-08, "loss": 0.8546, "step": 48514 }, { "epoch": 3.802115987460815, "grad_norm": 2.806471586227417, "learning_rate": 6.050159692571834e-08, "loss": 1.2719, "step": 48515 }, { "epoch": 3.802194357366771, "grad_norm": 4.224780082702637, "learning_rate": 6.045378114250045e-08, "loss": 0.6707, "step": 48516 }, { "epoch": 3.8022727272727272, "grad_norm": 10.893718719482422, "learning_rate": 6.040598414677968e-08, "loss": 1.2348, "step": 48517 }, { "epoch": 3.8023510971786836, "grad_norm": 3.271319627761841, "learning_rate": 6.035820593873753e-08, "loss": 0.6905, "step": 48518 }, { "epoch": 3.8024294670846395, "grad_norm": 3.4606709480285645, "learning_rate": 6.03104465185561e-08, "loss": 1.2393, "step": 48519 }, { "epoch": 3.8025078369905954, "grad_norm": 2.621361255645752, "learning_rate": 6.02627058864158e-08, "loss": 1.1336, "step": 48520 }, { "epoch": 3.8025862068965517, "grad_norm": 3.649059534072876, "learning_rate": 6.021498404249871e-08, "loss": 1.2494, "step": 48521 }, { "epoch": 3.802664576802508, "grad_norm": 2.7236785888671875, "learning_rate": 6.016728098698687e-08, "loss": 1.4231, "step": 48522 }, { "epoch": 3.802742946708464, "grad_norm": 2.9424004554748535, "learning_rate": 6.011959672006185e-08, "loss": 1.4396, "step": 48523 }, { "epoch": 3.80282131661442, "grad_norm": 11.43066692352295, "learning_rate": 6.007193124190457e-08, "loss": 1.7619, "step": 48524 }, { "epoch": 3.802899686520376, "grad_norm": 4.193935394287109, "learning_rate": 6.002428455269605e-08, "loss": 1.1622, "step": 48525 }, { "epoch": 3.8029780564263325, "grad_norm": 2.679121732711792, "learning_rate": 5.99766566526172e-08, "loss": 0.9463, "step": 48526 }, { "epoch": 3.8030564263322884, "grad_norm": 6.099730491638184, "learning_rate": 5.992904754185069e-08, "loss": 1.9654, "step": 48527 }, { "epoch": 3.8031347962382442, "grad_norm": 4.048408031463623, "learning_rate": 5.988145722057636e-08, "loss": 1.6642, "step": 48528 }, { "epoch": 3.8032131661442006, "grad_norm": 4.470081806182861, "learning_rate": 5.983388568897575e-08, "loss": 0.906, "step": 48529 }, { "epoch": 3.803291536050157, "grad_norm": 2.7259037494659424, "learning_rate": 5.97863329472298e-08, "loss": 1.1574, "step": 48530 }, { "epoch": 3.803369905956113, "grad_norm": 6.7390241622924805, "learning_rate": 5.973879899551838e-08, "loss": 1.1852, "step": 48531 }, { "epoch": 3.803448275862069, "grad_norm": 3.971613645553589, "learning_rate": 5.969128383402356e-08, "loss": 1.4543, "step": 48532 }, { "epoch": 3.803526645768025, "grad_norm": 2.5970325469970703, "learning_rate": 5.964378746292576e-08, "loss": 1.1566, "step": 48533 }, { "epoch": 3.8036050156739813, "grad_norm": 3.7359349727630615, "learning_rate": 5.9596309882405945e-08, "loss": 1.2951, "step": 48534 }, { "epoch": 3.8036833855799372, "grad_norm": 2.949199676513672, "learning_rate": 5.9548851092642854e-08, "loss": 1.368, "step": 48535 }, { "epoch": 3.8037617554858936, "grad_norm": 3.4388182163238525, "learning_rate": 5.950141109381913e-08, "loss": 1.5749, "step": 48536 }, { "epoch": 3.8038401253918495, "grad_norm": 2.800398111343384, "learning_rate": 5.9453989886114615e-08, "loss": 1.3893, "step": 48537 }, { "epoch": 3.803918495297806, "grad_norm": 2.679433822631836, "learning_rate": 5.9406587469709734e-08, "loss": 1.7387, "step": 48538 }, { "epoch": 3.8039968652037617, "grad_norm": 3.3710782527923584, "learning_rate": 5.9359203844783775e-08, "loss": 1.121, "step": 48539 }, { "epoch": 3.804075235109718, "grad_norm": 3.5259101390838623, "learning_rate": 5.9311839011517716e-08, "loss": 1.3935, "step": 48540 }, { "epoch": 3.804153605015674, "grad_norm": 5.323946952819824, "learning_rate": 5.926449297009251e-08, "loss": 1.6143, "step": 48541 }, { "epoch": 3.8042319749216302, "grad_norm": 6.277981758117676, "learning_rate": 5.921716572068692e-08, "loss": 0.7767, "step": 48542 }, { "epoch": 3.804310344827586, "grad_norm": 3.892092227935791, "learning_rate": 5.91698572634819e-08, "loss": 1.2668, "step": 48543 }, { "epoch": 3.8043887147335425, "grad_norm": 4.893560886383057, "learning_rate": 5.912256759865675e-08, "loss": 1.2143, "step": 48544 }, { "epoch": 3.8044670846394983, "grad_norm": 2.2623867988586426, "learning_rate": 5.907529672639134e-08, "loss": 1.2959, "step": 48545 }, { "epoch": 3.8045454545454547, "grad_norm": 2.939189910888672, "learning_rate": 5.9028044646865514e-08, "loss": 1.5036, "step": 48546 }, { "epoch": 3.8046238244514106, "grad_norm": 9.58538818359375, "learning_rate": 5.898081136025968e-08, "loss": 0.9045, "step": 48547 }, { "epoch": 3.804702194357367, "grad_norm": 2.698946952819824, "learning_rate": 5.8933596866752596e-08, "loss": 1.1197, "step": 48548 }, { "epoch": 3.804780564263323, "grad_norm": 6.474867343902588, "learning_rate": 5.8886401166524667e-08, "loss": 1.2058, "step": 48549 }, { "epoch": 3.804858934169279, "grad_norm": 3.631417751312256, "learning_rate": 5.883922425975464e-08, "loss": 1.5259, "step": 48550 }, { "epoch": 3.804937304075235, "grad_norm": 2.2339982986450195, "learning_rate": 5.879206614662181e-08, "loss": 1.3204, "step": 48551 }, { "epoch": 3.8050156739811913, "grad_norm": 3.8663899898529053, "learning_rate": 5.874492682730659e-08, "loss": 0.9604, "step": 48552 }, { "epoch": 3.8050940438871472, "grad_norm": 3.366750955581665, "learning_rate": 5.869780630198774e-08, "loss": 1.3392, "step": 48553 }, { "epoch": 3.8051724137931036, "grad_norm": 4.695253372192383, "learning_rate": 5.865070457084398e-08, "loss": 1.5694, "step": 48554 }, { "epoch": 3.8052507836990594, "grad_norm": 4.861739158630371, "learning_rate": 5.8603621634055196e-08, "loss": 1.4463, "step": 48555 }, { "epoch": 3.805329153605016, "grad_norm": 3.6257076263427734, "learning_rate": 5.8556557491800115e-08, "loss": 1.1856, "step": 48556 }, { "epoch": 3.8054075235109717, "grad_norm": 3.0493032932281494, "learning_rate": 5.850951214425804e-08, "loss": 1.3655, "step": 48557 }, { "epoch": 3.805485893416928, "grad_norm": 2.669401168823242, "learning_rate": 5.8462485591607724e-08, "loss": 1.2461, "step": 48558 }, { "epoch": 3.805564263322884, "grad_norm": 2.5169825553894043, "learning_rate": 5.8415477834027346e-08, "loss": 1.5381, "step": 48559 }, { "epoch": 3.80564263322884, "grad_norm": 6.243926048278809, "learning_rate": 5.8368488871696774e-08, "loss": 1.573, "step": 48560 }, { "epoch": 3.805721003134796, "grad_norm": 7.995392799377441, "learning_rate": 5.832151870479419e-08, "loss": 1.4775, "step": 48561 }, { "epoch": 3.8057993730407524, "grad_norm": 4.8534440994262695, "learning_rate": 5.8274567333498347e-08, "loss": 1.3779, "step": 48562 }, { "epoch": 3.8058777429467083, "grad_norm": 4.464775085449219, "learning_rate": 5.8227634757987986e-08, "loss": 1.184, "step": 48563 }, { "epoch": 3.8059561128526647, "grad_norm": 3.5578794479370117, "learning_rate": 5.8180720978441295e-08, "loss": 1.6523, "step": 48564 }, { "epoch": 3.8060344827586206, "grad_norm": 2.7333877086639404, "learning_rate": 5.813382599503703e-08, "loss": 1.2288, "step": 48565 }, { "epoch": 3.806112852664577, "grad_norm": 4.1985249519348145, "learning_rate": 5.808694980795282e-08, "loss": 1.2139, "step": 48566 }, { "epoch": 3.8061912225705328, "grad_norm": 2.958306312561035, "learning_rate": 5.8040092417367966e-08, "loss": 1.2484, "step": 48567 }, { "epoch": 3.806269592476489, "grad_norm": 6.263972282409668, "learning_rate": 5.7993253823460106e-08, "loss": 1.6149, "step": 48568 }, { "epoch": 3.806347962382445, "grad_norm": 4.107981204986572, "learning_rate": 5.794643402640798e-08, "loss": 1.1154, "step": 48569 }, { "epoch": 3.8064263322884013, "grad_norm": 5.881533622741699, "learning_rate": 5.789963302638868e-08, "loss": 0.5919, "step": 48570 }, { "epoch": 3.806504702194357, "grad_norm": 2.885206699371338, "learning_rate": 5.785285082358094e-08, "loss": 1.334, "step": 48571 }, { "epoch": 3.8065830721003135, "grad_norm": 9.155075073242188, "learning_rate": 5.780608741816296e-08, "loss": 1.3585, "step": 48572 }, { "epoch": 3.8066614420062694, "grad_norm": 3.547948122024536, "learning_rate": 5.775934281031126e-08, "loss": 1.4754, "step": 48573 }, { "epoch": 3.8067398119122258, "grad_norm": 6.478516578674316, "learning_rate": 5.771261700020514e-08, "loss": 1.4769, "step": 48574 }, { "epoch": 3.8068181818181817, "grad_norm": 4.255859851837158, "learning_rate": 5.766590998802168e-08, "loss": 1.6802, "step": 48575 }, { "epoch": 3.806896551724138, "grad_norm": 2.6575281620025635, "learning_rate": 5.7619221773938526e-08, "loss": 1.6868, "step": 48576 }, { "epoch": 3.806974921630094, "grad_norm": 3.8438823223114014, "learning_rate": 5.7572552358133304e-08, "loss": 1.5064, "step": 48577 }, { "epoch": 3.80705329153605, "grad_norm": 9.472177505493164, "learning_rate": 5.7525901740782544e-08, "loss": 1.2082, "step": 48578 }, { "epoch": 3.807131661442006, "grad_norm": 5.537903308868408, "learning_rate": 5.7479269922066095e-08, "loss": 1.2918, "step": 48579 }, { "epoch": 3.8072100313479624, "grad_norm": 2.5638067722320557, "learning_rate": 5.743265690215938e-08, "loss": 1.2665, "step": 48580 }, { "epoch": 3.8072884012539188, "grad_norm": 2.53537917137146, "learning_rate": 5.738606268124003e-08, "loss": 1.3426, "step": 48581 }, { "epoch": 3.8073667711598747, "grad_norm": 2.94053316116333, "learning_rate": 5.733948725948513e-08, "loss": 0.7932, "step": 48582 }, { "epoch": 3.8074451410658305, "grad_norm": 2.6474215984344482, "learning_rate": 5.7292930637072306e-08, "loss": 1.2594, "step": 48583 }, { "epoch": 3.807523510971787, "grad_norm": 2.79189133644104, "learning_rate": 5.724639281417865e-08, "loss": 1.4054, "step": 48584 }, { "epoch": 3.807601880877743, "grad_norm": 4.070718288421631, "learning_rate": 5.7199873790980686e-08, "loss": 1.0896, "step": 48585 }, { "epoch": 3.807680250783699, "grad_norm": 2.6976470947265625, "learning_rate": 5.715337356765604e-08, "loss": 1.1164, "step": 48586 }, { "epoch": 3.807758620689655, "grad_norm": 5.488145351409912, "learning_rate": 5.71068921443807e-08, "loss": 1.7026, "step": 48587 }, { "epoch": 3.8078369905956113, "grad_norm": 4.206298351287842, "learning_rate": 5.7060429521332286e-08, "loss": 1.473, "step": 48588 }, { "epoch": 3.8079153605015676, "grad_norm": 3.9953579902648926, "learning_rate": 5.701398569868677e-08, "loss": 1.2605, "step": 48589 }, { "epoch": 3.8079937304075235, "grad_norm": 2.751617908477783, "learning_rate": 5.696756067662179e-08, "loss": 1.2558, "step": 48590 }, { "epoch": 3.8080721003134794, "grad_norm": 12.998852729797363, "learning_rate": 5.692115445531221e-08, "loss": 1.7381, "step": 48591 }, { "epoch": 3.8081504702194358, "grad_norm": 5.424159049987793, "learning_rate": 5.687476703493622e-08, "loss": 0.7655, "step": 48592 }, { "epoch": 3.808228840125392, "grad_norm": 5.880431652069092, "learning_rate": 5.6828398415669784e-08, "loss": 1.7906, "step": 48593 }, { "epoch": 3.808307210031348, "grad_norm": 5.139426231384277, "learning_rate": 5.678204859768943e-08, "loss": 0.9532, "step": 48594 }, { "epoch": 3.808385579937304, "grad_norm": 12.117868423461914, "learning_rate": 5.6735717581171134e-08, "loss": 1.4146, "step": 48595 }, { "epoch": 3.80846394984326, "grad_norm": 3.8369243144989014, "learning_rate": 5.668940536629086e-08, "loss": 1.4348, "step": 48596 }, { "epoch": 3.8085423197492165, "grad_norm": 2.863218307495117, "learning_rate": 5.6643111953224586e-08, "loss": 1.2103, "step": 48597 }, { "epoch": 3.8086206896551724, "grad_norm": 5.698779582977295, "learning_rate": 5.659683734214938e-08, "loss": 0.667, "step": 48598 }, { "epoch": 3.8086990595611283, "grad_norm": 4.05086612701416, "learning_rate": 5.655058153324067e-08, "loss": 1.2912, "step": 48599 }, { "epoch": 3.8087774294670846, "grad_norm": 4.2324748039245605, "learning_rate": 5.650434452667442e-08, "loss": 1.2566, "step": 48600 }, { "epoch": 3.808855799373041, "grad_norm": 2.504016637802124, "learning_rate": 5.645812632262659e-08, "loss": 1.3691, "step": 48601 }, { "epoch": 3.808934169278997, "grad_norm": 4.4973883628845215, "learning_rate": 5.641192692127262e-08, "loss": 1.2576, "step": 48602 }, { "epoch": 3.8090125391849528, "grad_norm": 5.246748924255371, "learning_rate": 5.636574632278846e-08, "loss": 0.949, "step": 48603 }, { "epoch": 3.809090909090909, "grad_norm": 3.248278856277466, "learning_rate": 5.631958452735009e-08, "loss": 1.2134, "step": 48604 }, { "epoch": 3.8091692789968654, "grad_norm": 3.0720088481903076, "learning_rate": 5.6273441535132925e-08, "loss": 1.1752, "step": 48605 }, { "epoch": 3.8092476489028213, "grad_norm": 4.1402411460876465, "learning_rate": 5.622731734631182e-08, "loss": 1.0209, "step": 48606 }, { "epoch": 3.809326018808777, "grad_norm": 2.9578442573547363, "learning_rate": 5.61812119610633e-08, "loss": 1.1809, "step": 48607 }, { "epoch": 3.8094043887147335, "grad_norm": 2.3095293045043945, "learning_rate": 5.6135125379561674e-08, "loss": 1.4352, "step": 48608 }, { "epoch": 3.80948275862069, "grad_norm": 4.581522464752197, "learning_rate": 5.608905760198291e-08, "loss": 1.4552, "step": 48609 }, { "epoch": 3.8095611285266457, "grad_norm": 4.087953567504883, "learning_rate": 5.604300862850187e-08, "loss": 1.3672, "step": 48610 }, { "epoch": 3.8096394984326016, "grad_norm": 3.513584613800049, "learning_rate": 5.599697845929397e-08, "loss": 1.3542, "step": 48611 }, { "epoch": 3.809717868338558, "grad_norm": 2.481823682785034, "learning_rate": 5.595096709453407e-08, "loss": 1.1696, "step": 48612 }, { "epoch": 3.8097962382445143, "grad_norm": 3.4035181999206543, "learning_rate": 5.590497453439758e-08, "loss": 1.3783, "step": 48613 }, { "epoch": 3.80987460815047, "grad_norm": 10.989106178283691, "learning_rate": 5.5859000779058815e-08, "loss": 1.4329, "step": 48614 }, { "epoch": 3.809952978056426, "grad_norm": 3.056767225265503, "learning_rate": 5.581304582869262e-08, "loss": 1.223, "step": 48615 }, { "epoch": 3.8100313479623824, "grad_norm": 2.5799074172973633, "learning_rate": 5.576710968347443e-08, "loss": 1.1415, "step": 48616 }, { "epoch": 3.8101097178683387, "grad_norm": 2.4885051250457764, "learning_rate": 5.572119234357909e-08, "loss": 1.3363, "step": 48617 }, { "epoch": 3.8101880877742946, "grad_norm": 3.021336317062378, "learning_rate": 5.567529380918035e-08, "loss": 1.4771, "step": 48618 }, { "epoch": 3.8102664576802505, "grad_norm": 3.5680785179138184, "learning_rate": 5.562941408045308e-08, "loss": 0.8262, "step": 48619 }, { "epoch": 3.810344827586207, "grad_norm": 3.3521249294281006, "learning_rate": 5.558355315757213e-08, "loss": 1.5098, "step": 48620 }, { "epoch": 3.810423197492163, "grad_norm": 2.9708430767059326, "learning_rate": 5.553771104071182e-08, "loss": 1.2019, "step": 48621 }, { "epoch": 3.810501567398119, "grad_norm": 3.030224561691284, "learning_rate": 5.549188773004588e-08, "loss": 1.4247, "step": 48622 }, { "epoch": 3.810579937304075, "grad_norm": 4.061750411987305, "learning_rate": 5.5446083225749734e-08, "loss": 1.106, "step": 48623 }, { "epoch": 3.8106583072100313, "grad_norm": 5.200998306274414, "learning_rate": 5.540029752799714e-08, "loss": 1.0006, "step": 48624 }, { "epoch": 3.8107366771159876, "grad_norm": 2.591749429702759, "learning_rate": 5.5354530636961276e-08, "loss": 1.3852, "step": 48625 }, { "epoch": 3.8108150470219435, "grad_norm": 5.782389163970947, "learning_rate": 5.5308782552817574e-08, "loss": 1.2736, "step": 48626 }, { "epoch": 3.8108934169279, "grad_norm": 4.670005798339844, "learning_rate": 5.5263053275739774e-08, "loss": 1.4212, "step": 48627 }, { "epoch": 3.8109717868338557, "grad_norm": 2.6135482788085938, "learning_rate": 5.521734280590163e-08, "loss": 1.2595, "step": 48628 }, { "epoch": 3.811050156739812, "grad_norm": 2.8543267250061035, "learning_rate": 5.5171651143476334e-08, "loss": 1.1686, "step": 48629 }, { "epoch": 3.811128526645768, "grad_norm": 6.684656620025635, "learning_rate": 5.5125978288638196e-08, "loss": 0.9275, "step": 48630 }, { "epoch": 3.8112068965517243, "grad_norm": 3.284254789352417, "learning_rate": 5.508032424156151e-08, "loss": 1.7008, "step": 48631 }, { "epoch": 3.81128526645768, "grad_norm": 3.033574342727661, "learning_rate": 5.503468900241948e-08, "loss": 1.3242, "step": 48632 }, { "epoch": 3.8113636363636365, "grad_norm": 3.1012141704559326, "learning_rate": 5.4989072571385305e-08, "loss": 1.347, "step": 48633 }, { "epoch": 3.8114420062695924, "grad_norm": 4.876895904541016, "learning_rate": 5.494347494863217e-08, "loss": 1.2607, "step": 48634 }, { "epoch": 3.8115203761755487, "grad_norm": 2.780397653579712, "learning_rate": 5.489789613433494e-08, "loss": 1.2439, "step": 48635 }, { "epoch": 3.8115987460815046, "grad_norm": 2.7393853664398193, "learning_rate": 5.4852336128665696e-08, "loss": 1.3298, "step": 48636 }, { "epoch": 3.811677115987461, "grad_norm": 6.288816452026367, "learning_rate": 5.4806794931798747e-08, "loss": 1.3235, "step": 48637 }, { "epoch": 3.811755485893417, "grad_norm": 3.220834255218506, "learning_rate": 5.476127254390618e-08, "loss": 1.1789, "step": 48638 }, { "epoch": 3.811833855799373, "grad_norm": 4.475484848022461, "learning_rate": 5.471576896516173e-08, "loss": 1.4652, "step": 48639 }, { "epoch": 3.811912225705329, "grad_norm": 5.203280925750732, "learning_rate": 5.467028419573861e-08, "loss": 1.4451, "step": 48640 }, { "epoch": 3.8119905956112854, "grad_norm": 5.19857120513916, "learning_rate": 5.4624818235810005e-08, "loss": 1.6501, "step": 48641 }, { "epoch": 3.8120689655172413, "grad_norm": 3.148160696029663, "learning_rate": 5.4579371085548e-08, "loss": 1.5797, "step": 48642 }, { "epoch": 3.8121473354231976, "grad_norm": 3.820971965789795, "learning_rate": 5.45339427451258e-08, "loss": 0.9015, "step": 48643 }, { "epoch": 3.8122257053291535, "grad_norm": 4.280470848083496, "learning_rate": 5.448853321471659e-08, "loss": 1.6088, "step": 48644 }, { "epoch": 3.81230407523511, "grad_norm": 3.9777514934539795, "learning_rate": 5.444314249449301e-08, "loss": 1.4594, "step": 48645 }, { "epoch": 3.8123824451410657, "grad_norm": 3.2409188747406006, "learning_rate": 5.4397770584627693e-08, "loss": 1.179, "step": 48646 }, { "epoch": 3.812460815047022, "grad_norm": 4.561775207519531, "learning_rate": 5.4352417485292744e-08, "loss": 1.0499, "step": 48647 }, { "epoch": 3.812539184952978, "grad_norm": 2.8592419624328613, "learning_rate": 5.4307083196660226e-08, "loss": 1.7458, "step": 48648 }, { "epoch": 3.8126175548589343, "grad_norm": 2.6851589679718018, "learning_rate": 5.42617677189039e-08, "loss": 1.1008, "step": 48649 }, { "epoch": 3.81269592476489, "grad_norm": 3.143551826477051, "learning_rate": 5.421647105219585e-08, "loss": 1.1644, "step": 48650 }, { "epoch": 3.8127742946708465, "grad_norm": 5.750527858734131, "learning_rate": 5.417119319670816e-08, "loss": 1.9926, "step": 48651 }, { "epoch": 3.8128526645768024, "grad_norm": 2.9928057193756104, "learning_rate": 5.412593415261236e-08, "loss": 1.3829, "step": 48652 }, { "epoch": 3.8129310344827587, "grad_norm": 2.8065025806427, "learning_rate": 5.408069392008053e-08, "loss": 1.3352, "step": 48653 }, { "epoch": 3.8130094043887146, "grad_norm": 2.253579616546631, "learning_rate": 5.403547249928642e-08, "loss": 1.2622, "step": 48654 }, { "epoch": 3.813087774294671, "grad_norm": 4.162318706512451, "learning_rate": 5.399026989040101e-08, "loss": 1.1042, "step": 48655 }, { "epoch": 3.813166144200627, "grad_norm": 8.008923530578613, "learning_rate": 5.3945086093595836e-08, "loss": 1.9002, "step": 48656 }, { "epoch": 3.813244514106583, "grad_norm": 4.70021390914917, "learning_rate": 5.389992110904296e-08, "loss": 1.5563, "step": 48657 }, { "epoch": 3.813322884012539, "grad_norm": 4.896332263946533, "learning_rate": 5.385477493691449e-08, "loss": 1.1165, "step": 48658 }, { "epoch": 3.8134012539184954, "grad_norm": 2.970151424407959, "learning_rate": 5.380964757738194e-08, "loss": 1.0313, "step": 48659 }, { "epoch": 3.8134796238244513, "grad_norm": 2.181889295578003, "learning_rate": 5.376453903061685e-08, "loss": 1.3352, "step": 48660 }, { "epoch": 3.8135579937304076, "grad_norm": 7.9186835289001465, "learning_rate": 5.371944929679129e-08, "loss": 0.5202, "step": 48661 }, { "epoch": 3.8136363636363635, "grad_norm": 3.3603360652923584, "learning_rate": 5.3674378376075696e-08, "loss": 1.2753, "step": 48662 }, { "epoch": 3.81371473354232, "grad_norm": 6.418834209442139, "learning_rate": 5.36293262686427e-08, "loss": 0.9133, "step": 48663 }, { "epoch": 3.8137931034482757, "grad_norm": 2.6136314868927, "learning_rate": 5.358429297466272e-08, "loss": 0.996, "step": 48664 }, { "epoch": 3.813871473354232, "grad_norm": 2.3195295333862305, "learning_rate": 5.3539278494307844e-08, "loss": 0.7105, "step": 48665 }, { "epoch": 3.813949843260188, "grad_norm": 5.286159515380859, "learning_rate": 5.3494282827749044e-08, "loss": 1.1121, "step": 48666 }, { "epoch": 3.8140282131661443, "grad_norm": 4.848340034484863, "learning_rate": 5.344930597515674e-08, "loss": 1.262, "step": 48667 }, { "epoch": 3.8141065830721, "grad_norm": 4.73485803604126, "learning_rate": 5.3404347936702463e-08, "loss": 1.0588, "step": 48668 }, { "epoch": 3.8141849529780565, "grad_norm": 8.660700798034668, "learning_rate": 5.3359408712558294e-08, "loss": 1.0913, "step": 48669 }, { "epoch": 3.8142633228840124, "grad_norm": 3.7562568187713623, "learning_rate": 5.331448830289354e-08, "loss": 1.649, "step": 48670 }, { "epoch": 3.8143416927899687, "grad_norm": 11.83383560180664, "learning_rate": 5.326958670787974e-08, "loss": 1.3871, "step": 48671 }, { "epoch": 3.814420062695925, "grad_norm": 4.183014392852783, "learning_rate": 5.32247039276873e-08, "loss": 1.3179, "step": 48672 }, { "epoch": 3.814498432601881, "grad_norm": 2.8703389167785645, "learning_rate": 5.317983996248777e-08, "loss": 1.2337, "step": 48673 }, { "epoch": 3.814576802507837, "grad_norm": 3.762152671813965, "learning_rate": 5.313499481245099e-08, "loss": 0.7295, "step": 48674 }, { "epoch": 3.814655172413793, "grad_norm": 3.0662753582000732, "learning_rate": 5.309016847774795e-08, "loss": 1.3231, "step": 48675 }, { "epoch": 3.8147335423197495, "grad_norm": 2.311098098754883, "learning_rate": 5.304536095854906e-08, "loss": 1.295, "step": 48676 }, { "epoch": 3.8148119122257054, "grad_norm": 3.4792680740356445, "learning_rate": 5.3000572255024754e-08, "loss": 1.1834, "step": 48677 }, { "epoch": 3.8148902821316613, "grad_norm": 2.710498571395874, "learning_rate": 5.295580236734488e-08, "loss": 1.1619, "step": 48678 }, { "epoch": 3.8149686520376176, "grad_norm": 4.376010417938232, "learning_rate": 5.2911051295680974e-08, "loss": 1.4612, "step": 48679 }, { "epoch": 3.815047021943574, "grad_norm": 3.5721051692962646, "learning_rate": 5.286631904020178e-08, "loss": 1.2665, "step": 48680 }, { "epoch": 3.81512539184953, "grad_norm": 2.64980411529541, "learning_rate": 5.2821605601078294e-08, "loss": 1.1888, "step": 48681 }, { "epoch": 3.8152037617554857, "grad_norm": 4.2943525314331055, "learning_rate": 5.2776910978480365e-08, "loss": 1.2218, "step": 48682 }, { "epoch": 3.815282131661442, "grad_norm": 6.075835704803467, "learning_rate": 5.2732235172578414e-08, "loss": 1.0701, "step": 48683 }, { "epoch": 3.8153605015673984, "grad_norm": 2.432873249053955, "learning_rate": 5.26875781835412e-08, "loss": 0.7825, "step": 48684 }, { "epoch": 3.8154388714733543, "grad_norm": 3.0703470706939697, "learning_rate": 5.264294001153969e-08, "loss": 1.3983, "step": 48685 }, { "epoch": 3.81551724137931, "grad_norm": 2.7153592109680176, "learning_rate": 5.25983206567432e-08, "loss": 0.8568, "step": 48686 }, { "epoch": 3.8155956112852665, "grad_norm": 3.1886212825775146, "learning_rate": 5.255372011932214e-08, "loss": 1.7368, "step": 48687 }, { "epoch": 3.815673981191223, "grad_norm": 12.614253044128418, "learning_rate": 5.250913839944527e-08, "loss": 1.7601, "step": 48688 }, { "epoch": 3.8157523510971787, "grad_norm": 3.3372745513916016, "learning_rate": 5.24645754972819e-08, "loss": 1.6279, "step": 48689 }, { "epoch": 3.8158307210031346, "grad_norm": 2.19775390625, "learning_rate": 5.242003141300245e-08, "loss": 1.2303, "step": 48690 }, { "epoch": 3.815909090909091, "grad_norm": 3.181636095046997, "learning_rate": 5.2375506146776225e-08, "loss": 1.0503, "step": 48691 }, { "epoch": 3.8159874608150472, "grad_norm": 2.570639133453369, "learning_rate": 5.2330999698771976e-08, "loss": 1.0054, "step": 48692 }, { "epoch": 3.816065830721003, "grad_norm": 4.007729530334473, "learning_rate": 5.228651206915958e-08, "loss": 1.4602, "step": 48693 }, { "epoch": 3.816144200626959, "grad_norm": 3.1925275325775146, "learning_rate": 5.2242043258107776e-08, "loss": 1.1944, "step": 48694 }, { "epoch": 3.8162225705329154, "grad_norm": 6.864279270172119, "learning_rate": 5.2197593265785884e-08, "loss": 2.019, "step": 48695 }, { "epoch": 3.8163009404388717, "grad_norm": 3.725543260574341, "learning_rate": 5.215316209236265e-08, "loss": 1.0889, "step": 48696 }, { "epoch": 3.8163793103448276, "grad_norm": 3.311462640762329, "learning_rate": 5.2108749738007945e-08, "loss": 0.7718, "step": 48697 }, { "epoch": 3.8164576802507835, "grad_norm": 3.8295505046844482, "learning_rate": 5.206435620289052e-08, "loss": 1.5988, "step": 48698 }, { "epoch": 3.81653605015674, "grad_norm": 7.788797855377197, "learning_rate": 5.2019981487178017e-08, "loss": 1.1645, "step": 48699 }, { "epoch": 3.816614420062696, "grad_norm": 4.206718444824219, "learning_rate": 5.19756255910403e-08, "loss": 1.2012, "step": 48700 }, { "epoch": 3.816692789968652, "grad_norm": 4.378426551818848, "learning_rate": 5.193128851464613e-08, "loss": 1.1716, "step": 48701 }, { "epoch": 3.816771159874608, "grad_norm": 4.034580707550049, "learning_rate": 5.1886970258163695e-08, "loss": 0.9523, "step": 48702 }, { "epoch": 3.8168495297805642, "grad_norm": 3.504795551300049, "learning_rate": 5.1842670821761756e-08, "loss": 1.2655, "step": 48703 }, { "epoch": 3.8169278996865206, "grad_norm": 2.8307337760925293, "learning_rate": 5.17983902056085e-08, "loss": 1.241, "step": 48704 }, { "epoch": 3.8170062695924765, "grad_norm": 2.449129343032837, "learning_rate": 5.175412840987271e-08, "loss": 1.1283, "step": 48705 }, { "epoch": 3.8170846394984324, "grad_norm": 3.3914666175842285, "learning_rate": 5.170988543472311e-08, "loss": 1.4009, "step": 48706 }, { "epoch": 3.8171630094043887, "grad_norm": 14.277562141418457, "learning_rate": 5.166566128032735e-08, "loss": 0.8447, "step": 48707 }, { "epoch": 3.817241379310345, "grad_norm": 4.435527801513672, "learning_rate": 5.1621455946853086e-08, "loss": 0.9791, "step": 48708 }, { "epoch": 3.817319749216301, "grad_norm": 3.524062395095825, "learning_rate": 5.157726943447017e-08, "loss": 1.3071, "step": 48709 }, { "epoch": 3.817398119122257, "grad_norm": 2.5546963214874268, "learning_rate": 5.1533101743344585e-08, "loss": 1.2952, "step": 48710 }, { "epoch": 3.817476489028213, "grad_norm": 2.0237479209899902, "learning_rate": 5.148895287364619e-08, "loss": 1.2507, "step": 48711 }, { "epoch": 3.8175548589341695, "grad_norm": 3.1841235160827637, "learning_rate": 5.1444822825542085e-08, "loss": 1.5903, "step": 48712 }, { "epoch": 3.8176332288401253, "grad_norm": 7.063615322113037, "learning_rate": 5.140071159919935e-08, "loss": 1.2761, "step": 48713 }, { "epoch": 3.8177115987460812, "grad_norm": 2.846719980239868, "learning_rate": 5.135661919478729e-08, "loss": 1.3483, "step": 48714 }, { "epoch": 3.8177899686520376, "grad_norm": 2.9774599075317383, "learning_rate": 5.131254561247245e-08, "loss": 1.0441, "step": 48715 }, { "epoch": 3.817868338557994, "grad_norm": 2.939265251159668, "learning_rate": 5.126849085242358e-08, "loss": 1.2218, "step": 48716 }, { "epoch": 3.81794670846395, "grad_norm": 4.137298583984375, "learning_rate": 5.122445491480721e-08, "loss": 1.0851, "step": 48717 }, { "epoch": 3.818025078369906, "grad_norm": 2.6205618381500244, "learning_rate": 5.118043779979043e-08, "loss": 1.4925, "step": 48718 }, { "epoch": 3.818103448275862, "grad_norm": 9.116805076599121, "learning_rate": 5.113643950754199e-08, "loss": 1.8034, "step": 48719 }, { "epoch": 3.8181818181818183, "grad_norm": 4.953011512756348, "learning_rate": 5.1092460038228985e-08, "loss": 1.4336, "step": 48720 }, { "epoch": 3.8182601880877742, "grad_norm": 2.70626163482666, "learning_rate": 5.1048499392017944e-08, "loss": 0.9426, "step": 48721 }, { "epoch": 3.8183385579937306, "grad_norm": 4.4658966064453125, "learning_rate": 5.100455756907652e-08, "loss": 1.4728, "step": 48722 }, { "epoch": 3.8184169278996865, "grad_norm": 5.082582473754883, "learning_rate": 5.096063456957123e-08, "loss": 1.5579, "step": 48723 }, { "epoch": 3.818495297805643, "grad_norm": 3.3022639751434326, "learning_rate": 5.091673039367029e-08, "loss": 1.4728, "step": 48724 }, { "epoch": 3.8185736677115987, "grad_norm": 5.444194316864014, "learning_rate": 5.0872845041540774e-08, "loss": 1.4, "step": 48725 }, { "epoch": 3.818652037617555, "grad_norm": 2.6075046062469482, "learning_rate": 5.0828978513348114e-08, "loss": 1.2783, "step": 48726 }, { "epoch": 3.818730407523511, "grad_norm": 2.9709718227386475, "learning_rate": 5.078513080925995e-08, "loss": 1.0127, "step": 48727 }, { "epoch": 3.8188087774294672, "grad_norm": 7.782437801361084, "learning_rate": 5.074130192944282e-08, "loss": 0.9584, "step": 48728 }, { "epoch": 3.818887147335423, "grad_norm": 3.347343921661377, "learning_rate": 5.0697491874064364e-08, "loss": 1.6939, "step": 48729 }, { "epoch": 3.8189655172413794, "grad_norm": 3.764238119125366, "learning_rate": 5.0653700643290006e-08, "loss": 1.0568, "step": 48730 }, { "epoch": 3.8190438871473353, "grad_norm": 4.4333319664001465, "learning_rate": 5.060992823728683e-08, "loss": 1.2559, "step": 48731 }, { "epoch": 3.8191222570532917, "grad_norm": 3.2303431034088135, "learning_rate": 5.0566174656221935e-08, "loss": 1.3288, "step": 48732 }, { "epoch": 3.8192006269592476, "grad_norm": 4.31590461730957, "learning_rate": 5.0522439900260734e-08, "loss": 1.5791, "step": 48733 }, { "epoch": 3.819278996865204, "grad_norm": 5.245468616485596, "learning_rate": 5.047872396956976e-08, "loss": 1.4523, "step": 48734 }, { "epoch": 3.81935736677116, "grad_norm": 4.5251922607421875, "learning_rate": 5.043502686431556e-08, "loss": 1.3563, "step": 48735 }, { "epoch": 3.819435736677116, "grad_norm": 7.295940399169922, "learning_rate": 5.0391348584664653e-08, "loss": 0.4435, "step": 48736 }, { "epoch": 3.819514106583072, "grad_norm": 10.399169921875, "learning_rate": 5.034768913078192e-08, "loss": 2.7959, "step": 48737 }, { "epoch": 3.8195924764890283, "grad_norm": 3.0600295066833496, "learning_rate": 5.030404850283499e-08, "loss": 1.2844, "step": 48738 }, { "epoch": 3.819670846394984, "grad_norm": 2.9190409183502197, "learning_rate": 5.02604267009893e-08, "loss": 1.3685, "step": 48739 }, { "epoch": 3.8197492163009406, "grad_norm": 2.6154866218566895, "learning_rate": 5.021682372541026e-08, "loss": 1.16, "step": 48740 }, { "epoch": 3.8198275862068964, "grad_norm": 4.4671125411987305, "learning_rate": 5.017323957626385e-08, "loss": 1.5424, "step": 48741 }, { "epoch": 3.8199059561128528, "grad_norm": 3.328108310699463, "learning_rate": 5.012967425371607e-08, "loss": 1.5047, "step": 48742 }, { "epoch": 3.8199843260188087, "grad_norm": 3.254368782043457, "learning_rate": 5.008612775793287e-08, "loss": 1.5637, "step": 48743 }, { "epoch": 3.820062695924765, "grad_norm": 6.805605411529541, "learning_rate": 5.004260008907969e-08, "loss": 1.6207, "step": 48744 }, { "epoch": 3.820141065830721, "grad_norm": 2.7742130756378174, "learning_rate": 4.999909124732194e-08, "loss": 1.262, "step": 48745 }, { "epoch": 3.820219435736677, "grad_norm": 2.43473219871521, "learning_rate": 4.9955601232825054e-08, "loss": 1.2785, "step": 48746 }, { "epoch": 3.820297805642633, "grad_norm": 3.3656203746795654, "learning_rate": 4.9912130045754456e-08, "loss": 1.5371, "step": 48747 }, { "epoch": 3.8203761755485894, "grad_norm": 5.859039783477783, "learning_rate": 4.986867768627557e-08, "loss": 1.1837, "step": 48748 }, { "epoch": 3.8204545454545453, "grad_norm": 3.6832168102264404, "learning_rate": 4.982524415455381e-08, "loss": 0.8983, "step": 48749 }, { "epoch": 3.8205329153605017, "grad_norm": 3.6047072410583496, "learning_rate": 4.9781829450754046e-08, "loss": 1.0254, "step": 48750 }, { "epoch": 3.8206112852664575, "grad_norm": 3.09946346282959, "learning_rate": 4.973843357504171e-08, "loss": 1.4736, "step": 48751 }, { "epoch": 3.820689655172414, "grad_norm": 4.090713977813721, "learning_rate": 4.969505652758222e-08, "loss": 1.6841, "step": 48752 }, { "epoch": 3.8207680250783698, "grad_norm": 3.7985634803771973, "learning_rate": 4.965169830853933e-08, "loss": 1.7486, "step": 48753 }, { "epoch": 3.820846394984326, "grad_norm": 4.283898830413818, "learning_rate": 4.960835891807958e-08, "loss": 0.8824, "step": 48754 }, { "epoch": 3.820924764890282, "grad_norm": 3.0987119674682617, "learning_rate": 4.9565038356366725e-08, "loss": 0.7663, "step": 48755 }, { "epoch": 3.8210031347962383, "grad_norm": 2.8529653549194336, "learning_rate": 4.9521736623565076e-08, "loss": 1.2263, "step": 48756 }, { "epoch": 3.821081504702194, "grad_norm": 2.978675365447998, "learning_rate": 4.947845371984061e-08, "loss": 1.2363, "step": 48757 }, { "epoch": 3.8211598746081505, "grad_norm": 5.543118476867676, "learning_rate": 4.94351896453571e-08, "loss": 1.3359, "step": 48758 }, { "epoch": 3.8212382445141064, "grad_norm": 2.394179582595825, "learning_rate": 4.939194440027994e-08, "loss": 1.2912, "step": 48759 }, { "epoch": 3.8213166144200628, "grad_norm": 4.116743564605713, "learning_rate": 4.934871798477236e-08, "loss": 1.0928, "step": 48760 }, { "epoch": 3.8213949843260187, "grad_norm": 2.843313217163086, "learning_rate": 4.9305510398999225e-08, "loss": 1.6412, "step": 48761 }, { "epoch": 3.821473354231975, "grad_norm": 7.934531211853027, "learning_rate": 4.9262321643125946e-08, "loss": 1.3688, "step": 48762 }, { "epoch": 3.821551724137931, "grad_norm": 6.905601501464844, "learning_rate": 4.921915171731517e-08, "loss": 1.7413, "step": 48763 }, { "epoch": 3.821630094043887, "grad_norm": 2.5823569297790527, "learning_rate": 4.917600062173233e-08, "loss": 1.4729, "step": 48764 }, { "epoch": 3.821708463949843, "grad_norm": 3.186675548553467, "learning_rate": 4.9132868356541186e-08, "loss": 1.1717, "step": 48765 }, { "epoch": 3.8217868338557994, "grad_norm": 3.5457937717437744, "learning_rate": 4.908975492190493e-08, "loss": 1.8598, "step": 48766 }, { "epoch": 3.8218652037617558, "grad_norm": 4.484041213989258, "learning_rate": 4.904666031798899e-08, "loss": 1.4821, "step": 48767 }, { "epoch": 3.8219435736677116, "grad_norm": 3.2974331378936768, "learning_rate": 4.900358454495657e-08, "loss": 1.3055, "step": 48768 }, { "epoch": 3.8220219435736675, "grad_norm": 2.462855577468872, "learning_rate": 4.896052760297143e-08, "loss": 1.1447, "step": 48769 }, { "epoch": 3.822100313479624, "grad_norm": 2.6800739765167236, "learning_rate": 4.891748949219732e-08, "loss": 1.3657, "step": 48770 }, { "epoch": 3.82217868338558, "grad_norm": 2.7799367904663086, "learning_rate": 4.887447021279801e-08, "loss": 1.0632, "step": 48771 }, { "epoch": 3.822257053291536, "grad_norm": 2.9849767684936523, "learning_rate": 4.8831469764937246e-08, "loss": 1.0565, "step": 48772 }, { "epoch": 3.822335423197492, "grad_norm": 2.5423882007598877, "learning_rate": 4.878848814877879e-08, "loss": 1.5618, "step": 48773 }, { "epoch": 3.8224137931034483, "grad_norm": 12.143099784851074, "learning_rate": 4.8745525364485294e-08, "loss": 0.9403, "step": 48774 }, { "epoch": 3.8224921630094046, "grad_norm": 3.718902111053467, "learning_rate": 4.8702581412220507e-08, "loss": 1.2577, "step": 48775 }, { "epoch": 3.8225705329153605, "grad_norm": 3.949646234512329, "learning_rate": 4.865965629214819e-08, "loss": 1.3442, "step": 48776 }, { "epoch": 3.8226489028213164, "grad_norm": 2.5033457279205322, "learning_rate": 4.861675000443211e-08, "loss": 1.2518, "step": 48777 }, { "epoch": 3.8227272727272728, "grad_norm": 5.625624656677246, "learning_rate": 4.8573862549233796e-08, "loss": 1.3025, "step": 48778 }, { "epoch": 3.822805642633229, "grad_norm": 4.706119537353516, "learning_rate": 4.853099392671756e-08, "loss": 1.5177, "step": 48779 }, { "epoch": 3.822884012539185, "grad_norm": 4.164488792419434, "learning_rate": 4.8488144137046056e-08, "loss": 1.3209, "step": 48780 }, { "epoch": 3.822962382445141, "grad_norm": 4.145496845245361, "learning_rate": 4.8445313180383037e-08, "loss": 1.2894, "step": 48781 }, { "epoch": 3.823040752351097, "grad_norm": 3.650508165359497, "learning_rate": 4.840250105689004e-08, "loss": 1.4163, "step": 48782 }, { "epoch": 3.8231191222570535, "grad_norm": 3.068962812423706, "learning_rate": 4.835970776673138e-08, "loss": 0.9434, "step": 48783 }, { "epoch": 3.8231974921630094, "grad_norm": 5.529074192047119, "learning_rate": 4.831693331006859e-08, "loss": 1.3804, "step": 48784 }, { "epoch": 3.8232758620689653, "grad_norm": 5.288130760192871, "learning_rate": 4.8274177687064885e-08, "loss": 1.1197, "step": 48785 }, { "epoch": 3.8233542319749216, "grad_norm": 3.6501753330230713, "learning_rate": 4.8231440897882345e-08, "loss": 1.3711, "step": 48786 }, { "epoch": 3.823432601880878, "grad_norm": 3.7539455890655518, "learning_rate": 4.818872294268473e-08, "loss": 0.9442, "step": 48787 }, { "epoch": 3.823510971786834, "grad_norm": 2.48714280128479, "learning_rate": 4.814602382163358e-08, "loss": 1.4284, "step": 48788 }, { "epoch": 3.8235893416927897, "grad_norm": 2.8375864028930664, "learning_rate": 4.810334353489154e-08, "loss": 1.427, "step": 48789 }, { "epoch": 3.823667711598746, "grad_norm": 2.412773370742798, "learning_rate": 4.806068208262071e-08, "loss": 1.1879, "step": 48790 }, { "epoch": 3.8237460815047024, "grad_norm": 6.062939643859863, "learning_rate": 4.8018039464983735e-08, "loss": 1.1921, "step": 48791 }, { "epoch": 3.8238244514106583, "grad_norm": 3.508596897125244, "learning_rate": 4.79754156821427e-08, "loss": 0.9653, "step": 48792 }, { "epoch": 3.823902821316614, "grad_norm": 10.66627311706543, "learning_rate": 4.7932810734259704e-08, "loss": 1.3417, "step": 48793 }, { "epoch": 3.8239811912225705, "grad_norm": 2.913010358810425, "learning_rate": 4.789022462149628e-08, "loss": 1.2332, "step": 48794 }, { "epoch": 3.824059561128527, "grad_norm": 6.543478965759277, "learning_rate": 4.784765734401564e-08, "loss": 1.4672, "step": 48795 }, { "epoch": 3.8241379310344827, "grad_norm": 3.387385129928589, "learning_rate": 4.7805108901978755e-08, "loss": 1.3385, "step": 48796 }, { "epoch": 3.8242163009404386, "grad_norm": 7.826579570770264, "learning_rate": 4.7762579295547176e-08, "loss": 2.5234, "step": 48797 }, { "epoch": 3.824294670846395, "grad_norm": 6.269216537475586, "learning_rate": 4.7720068524883535e-08, "loss": 1.8932, "step": 48798 }, { "epoch": 3.8243730407523513, "grad_norm": 3.305225372314453, "learning_rate": 4.767757659014882e-08, "loss": 1.2499, "step": 48799 }, { "epoch": 3.824451410658307, "grad_norm": 2.685352325439453, "learning_rate": 4.763510349150513e-08, "loss": 1.4457, "step": 48800 }, { "epoch": 3.824529780564263, "grad_norm": 5.220024585723877, "learning_rate": 4.7592649229113444e-08, "loss": 1.6415, "step": 48801 }, { "epoch": 3.8246081504702194, "grad_norm": 3.018616199493408, "learning_rate": 4.7550213803135846e-08, "loss": 1.1827, "step": 48802 }, { "epoch": 3.8246865203761757, "grad_norm": 3.8670973777770996, "learning_rate": 4.7507797213733884e-08, "loss": 1.4373, "step": 48803 }, { "epoch": 3.8247648902821316, "grad_norm": 4.11925745010376, "learning_rate": 4.7465399461067985e-08, "loss": 1.8907, "step": 48804 }, { "epoch": 3.8248432601880875, "grad_norm": 3.304647445678711, "learning_rate": 4.742302054530024e-08, "loss": 0.9908, "step": 48805 }, { "epoch": 3.824921630094044, "grad_norm": 6.168125152587891, "learning_rate": 4.738066046659107e-08, "loss": 1.0818, "step": 48806 }, { "epoch": 3.825, "grad_norm": 3.3317770957946777, "learning_rate": 4.733831922510257e-08, "loss": 1.3877, "step": 48807 }, { "epoch": 3.825078369905956, "grad_norm": 5.146910190582275, "learning_rate": 4.729599682099462e-08, "loss": 1.3137, "step": 48808 }, { "epoch": 3.825156739811912, "grad_norm": 2.495234727859497, "learning_rate": 4.72536932544293e-08, "loss": 1.171, "step": 48809 }, { "epoch": 3.8252351097178683, "grad_norm": 3.3304779529571533, "learning_rate": 4.7211408525567046e-08, "loss": 1.217, "step": 48810 }, { "epoch": 3.8253134796238246, "grad_norm": 3.1073899269104004, "learning_rate": 4.7169142634568843e-08, "loss": 1.7005, "step": 48811 }, { "epoch": 3.8253918495297805, "grad_norm": 4.027336120605469, "learning_rate": 4.712689558159456e-08, "loss": 1.2869, "step": 48812 }, { "epoch": 3.825470219435737, "grad_norm": 4.99254035949707, "learning_rate": 4.708466736680628e-08, "loss": 1.2654, "step": 48813 }, { "epoch": 3.8255485893416927, "grad_norm": 5.063268184661865, "learning_rate": 4.704245799036389e-08, "loss": 1.338, "step": 48814 }, { "epoch": 3.825626959247649, "grad_norm": 2.979661226272583, "learning_rate": 4.7000267452427806e-08, "loss": 1.5942, "step": 48815 }, { "epoch": 3.825705329153605, "grad_norm": 4.478221893310547, "learning_rate": 4.6958095753159015e-08, "loss": 1.6477, "step": 48816 }, { "epoch": 3.8257836990595613, "grad_norm": 4.013096332550049, "learning_rate": 4.691594289271684e-08, "loss": 1.2064, "step": 48817 }, { "epoch": 3.825862068965517, "grad_norm": 2.2708330154418945, "learning_rate": 4.687380887126336e-08, "loss": 1.2526, "step": 48818 }, { "epoch": 3.8259404388714735, "grad_norm": 3.2462573051452637, "learning_rate": 4.683169368895679e-08, "loss": 1.3474, "step": 48819 }, { "epoch": 3.8260188087774294, "grad_norm": 3.198044538497925, "learning_rate": 4.6789597345959223e-08, "loss": 0.9708, "step": 48820 }, { "epoch": 3.8260971786833857, "grad_norm": 3.327444314956665, "learning_rate": 4.674751984242942e-08, "loss": 1.3691, "step": 48821 }, { "epoch": 3.8261755485893416, "grad_norm": 3.211617946624756, "learning_rate": 4.670546117852781e-08, "loss": 1.2337, "step": 48822 }, { "epoch": 3.826253918495298, "grad_norm": 8.278278350830078, "learning_rate": 4.6663421354414815e-08, "loss": 1.2189, "step": 48823 }, { "epoch": 3.826332288401254, "grad_norm": 4.756748199462891, "learning_rate": 4.662140037024976e-08, "loss": 1.3255, "step": 48824 }, { "epoch": 3.82641065830721, "grad_norm": 2.8055717945098877, "learning_rate": 4.6579398226193065e-08, "loss": 1.0322, "step": 48825 }, { "epoch": 3.826489028213166, "grad_norm": 4.063686847686768, "learning_rate": 4.653741492240349e-08, "loss": 1.4289, "step": 48826 }, { "epoch": 3.8265673981191224, "grad_norm": 3.4756879806518555, "learning_rate": 4.649545045904147e-08, "loss": 1.6037, "step": 48827 }, { "epoch": 3.8266457680250783, "grad_norm": 5.801456451416016, "learning_rate": 4.645350483626632e-08, "loss": 1.4614, "step": 48828 }, { "epoch": 3.8267241379310346, "grad_norm": 3.8890435695648193, "learning_rate": 4.641157805423846e-08, "loss": 1.9295, "step": 48829 }, { "epoch": 3.8268025078369905, "grad_norm": 4.50921106338501, "learning_rate": 4.636967011311611e-08, "loss": 1.8078, "step": 48830 }, { "epoch": 3.826880877742947, "grad_norm": 3.359834671020508, "learning_rate": 4.6327781013059126e-08, "loss": 1.1626, "step": 48831 }, { "epoch": 3.8269592476489027, "grad_norm": 3.1869375705718994, "learning_rate": 4.6285910754226836e-08, "loss": 1.3406, "step": 48832 }, { "epoch": 3.827037617554859, "grad_norm": 3.6800591945648193, "learning_rate": 4.624405933677911e-08, "loss": 0.9571, "step": 48833 }, { "epoch": 3.827115987460815, "grad_norm": 3.4446473121643066, "learning_rate": 4.620222676087416e-08, "loss": 0.8594, "step": 48834 }, { "epoch": 3.8271943573667713, "grad_norm": 2.5194995403289795, "learning_rate": 4.6160413026671844e-08, "loss": 0.9842, "step": 48835 }, { "epoch": 3.827272727272727, "grad_norm": 2.800088882446289, "learning_rate": 4.611861813432983e-08, "loss": 1.09, "step": 48836 }, { "epoch": 3.8273510971786835, "grad_norm": 3.5602681636810303, "learning_rate": 4.6076842084009646e-08, "loss": 1.7117, "step": 48837 }, { "epoch": 3.8274294670846394, "grad_norm": 6.064864635467529, "learning_rate": 4.603508487586783e-08, "loss": 1.5343, "step": 48838 }, { "epoch": 3.8275078369905957, "grad_norm": 8.43317985534668, "learning_rate": 4.599334651006371e-08, "loss": 1.3301, "step": 48839 }, { "epoch": 3.8275862068965516, "grad_norm": 5.072503089904785, "learning_rate": 4.5951626986757146e-08, "loss": 1.442, "step": 48840 }, { "epoch": 3.827664576802508, "grad_norm": 3.2379510402679443, "learning_rate": 4.590992630610525e-08, "loss": 1.0059, "step": 48841 }, { "epoch": 3.827742946708464, "grad_norm": 3.0389404296875, "learning_rate": 4.5868244468267874e-08, "loss": 1.3939, "step": 48842 }, { "epoch": 3.82782131661442, "grad_norm": 6.080262184143066, "learning_rate": 4.582658147340324e-08, "loss": 1.3925, "step": 48843 }, { "epoch": 3.827899686520376, "grad_norm": 3.352602005004883, "learning_rate": 4.5784937321669555e-08, "loss": 1.7601, "step": 48844 }, { "epoch": 3.8279780564263324, "grad_norm": 2.8867180347442627, "learning_rate": 4.574331201322502e-08, "loss": 1.1428, "step": 48845 }, { "epoch": 3.8280564263322883, "grad_norm": 3.486936330795288, "learning_rate": 4.5701705548228393e-08, "loss": 1.4508, "step": 48846 }, { "epoch": 3.8281347962382446, "grad_norm": 10.073683738708496, "learning_rate": 4.566011792683789e-08, "loss": 0.8337, "step": 48847 }, { "epoch": 3.8282131661442005, "grad_norm": 4.3992767333984375, "learning_rate": 4.561854914921171e-08, "loss": 1.2644, "step": 48848 }, { "epoch": 3.828291536050157, "grad_norm": 2.3457272052764893, "learning_rate": 4.557699921550751e-08, "loss": 1.2184, "step": 48849 }, { "epoch": 3.8283699059561127, "grad_norm": 3.1522409915924072, "learning_rate": 4.5535468125883496e-08, "loss": 1.2109, "step": 48850 }, { "epoch": 3.828448275862069, "grad_norm": 3.6266486644744873, "learning_rate": 4.549395588049843e-08, "loss": 1.0366, "step": 48851 }, { "epoch": 3.828526645768025, "grad_norm": 2.5988073348999023, "learning_rate": 4.5452462479509405e-08, "loss": 1.3145, "step": 48852 }, { "epoch": 3.8286050156739813, "grad_norm": 2.863093376159668, "learning_rate": 4.541098792307408e-08, "loss": 1.7363, "step": 48853 }, { "epoch": 3.828683385579937, "grad_norm": 2.679797887802124, "learning_rate": 4.5369532211350653e-08, "loss": 1.499, "step": 48854 }, { "epoch": 3.8287617554858935, "grad_norm": 2.961799383163452, "learning_rate": 4.532809534449678e-08, "loss": 1.2209, "step": 48855 }, { "epoch": 3.8288401253918494, "grad_norm": 2.9923086166381836, "learning_rate": 4.528667732266956e-08, "loss": 1.6204, "step": 48856 }, { "epoch": 3.8289184952978057, "grad_norm": 3.291792154312134, "learning_rate": 4.524527814602719e-08, "loss": 1.6896, "step": 48857 }, { "epoch": 3.8289968652037616, "grad_norm": 3.1485238075256348, "learning_rate": 4.520389781472678e-08, "loss": 1.4956, "step": 48858 }, { "epoch": 3.829075235109718, "grad_norm": 6.489948272705078, "learning_rate": 4.516253632892598e-08, "loss": 1.2843, "step": 48859 }, { "epoch": 3.829153605015674, "grad_norm": 3.7216532230377197, "learning_rate": 4.512119368878187e-08, "loss": 1.2734, "step": 48860 }, { "epoch": 3.82923197492163, "grad_norm": 3.3833909034729004, "learning_rate": 4.5079869894451564e-08, "loss": 1.1999, "step": 48861 }, { "epoch": 3.8293103448275865, "grad_norm": 3.8835058212280273, "learning_rate": 4.503856494609271e-08, "loss": 1.349, "step": 48862 }, { "epoch": 3.8293887147335424, "grad_norm": 3.8855855464935303, "learning_rate": 4.49972788438624e-08, "loss": 1.7556, "step": 48863 }, { "epoch": 3.8294670846394983, "grad_norm": 2.7162389755249023, "learning_rate": 4.4956011587916624e-08, "loss": 1.1886, "step": 48864 }, { "epoch": 3.8295454545454546, "grad_norm": 2.5674209594726562, "learning_rate": 4.491476317841359e-08, "loss": 1.3515, "step": 48865 }, { "epoch": 3.829623824451411, "grad_norm": 5.681570529937744, "learning_rate": 4.4873533615509834e-08, "loss": 1.3387, "step": 48866 }, { "epoch": 3.829702194357367, "grad_norm": 3.3848135471343994, "learning_rate": 4.48323228993619e-08, "loss": 1.6277, "step": 48867 }, { "epoch": 3.8297805642633227, "grad_norm": 8.204392433166504, "learning_rate": 4.4791131030126334e-08, "loss": 1.7265, "step": 48868 }, { "epoch": 3.829858934169279, "grad_norm": 2.99716854095459, "learning_rate": 4.4749958007960224e-08, "loss": 1.1188, "step": 48869 }, { "epoch": 3.8299373040752354, "grad_norm": 3.9021155834198, "learning_rate": 4.4708803833020673e-08, "loss": 1.1868, "step": 48870 }, { "epoch": 3.8300156739811912, "grad_norm": 4.253085613250732, "learning_rate": 4.466766850546311e-08, "loss": 1.3245, "step": 48871 }, { "epoch": 3.830094043887147, "grad_norm": 2.8954851627349854, "learning_rate": 4.4626552025444634e-08, "loss": 1.0946, "step": 48872 }, { "epoch": 3.8301724137931035, "grad_norm": 5.229257106781006, "learning_rate": 4.458545439312123e-08, "loss": 1.3535, "step": 48873 }, { "epoch": 3.83025078369906, "grad_norm": 13.937227249145508, "learning_rate": 4.454437560864999e-08, "loss": 1.3407, "step": 48874 }, { "epoch": 3.8303291536050157, "grad_norm": 3.183260202407837, "learning_rate": 4.450331567218635e-08, "loss": 1.3511, "step": 48875 }, { "epoch": 3.8304075235109716, "grad_norm": 6.277746677398682, "learning_rate": 4.4462274583886855e-08, "loss": 1.5431, "step": 48876 }, { "epoch": 3.830485893416928, "grad_norm": 2.953918695449829, "learning_rate": 4.4421252343907485e-08, "loss": 1.5253, "step": 48877 }, { "epoch": 3.8305642633228842, "grad_norm": 6.225977897644043, "learning_rate": 4.438024895240423e-08, "loss": 1.51, "step": 48878 }, { "epoch": 3.83064263322884, "grad_norm": 2.4572994709014893, "learning_rate": 4.433926440953362e-08, "loss": 1.4181, "step": 48879 }, { "epoch": 3.830721003134796, "grad_norm": 3.282562255859375, "learning_rate": 4.429829871545055e-08, "loss": 1.3432, "step": 48880 }, { "epoch": 3.8307993730407524, "grad_norm": 3.2832727432250977, "learning_rate": 4.4257351870312106e-08, "loss": 1.6818, "step": 48881 }, { "epoch": 3.8308777429467087, "grad_norm": 2.726888656616211, "learning_rate": 4.421642387427261e-08, "loss": 1.2718, "step": 48882 }, { "epoch": 3.8309561128526646, "grad_norm": 2.4346299171447754, "learning_rate": 4.4175514727488046e-08, "loss": 1.4062, "step": 48883 }, { "epoch": 3.8310344827586205, "grad_norm": 2.571726083755493, "learning_rate": 4.4134624430114957e-08, "loss": 1.3907, "step": 48884 }, { "epoch": 3.831112852664577, "grad_norm": 2.673091173171997, "learning_rate": 4.409375298230822e-08, "loss": 1.1093, "step": 48885 }, { "epoch": 3.831191222570533, "grad_norm": 3.043156147003174, "learning_rate": 4.4052900384223276e-08, "loss": 0.9272, "step": 48886 }, { "epoch": 3.831269592476489, "grad_norm": 4.83709716796875, "learning_rate": 4.401206663601554e-08, "loss": 1.7392, "step": 48887 }, { "epoch": 3.831347962382445, "grad_norm": 3.1275625228881836, "learning_rate": 4.397125173783989e-08, "loss": 1.4077, "step": 48888 }, { "epoch": 3.8314263322884012, "grad_norm": 3.1073217391967773, "learning_rate": 4.393045568985288e-08, "loss": 1.5007, "step": 48889 }, { "epoch": 3.8315047021943576, "grad_norm": 3.010280132293701, "learning_rate": 4.3889678492208266e-08, "loss": 1.1658, "step": 48890 }, { "epoch": 3.8315830721003135, "grad_norm": 2.89170241355896, "learning_rate": 4.3848920145062034e-08, "loss": 1.25, "step": 48891 }, { "epoch": 3.8316614420062693, "grad_norm": 3.6723830699920654, "learning_rate": 4.380818064856851e-08, "loss": 1.0303, "step": 48892 }, { "epoch": 3.8317398119122257, "grad_norm": 3.093079090118408, "learning_rate": 4.376746000288312e-08, "loss": 1.3682, "step": 48893 }, { "epoch": 3.831818181818182, "grad_norm": 3.4705939292907715, "learning_rate": 4.372675820816075e-08, "loss": 1.6062, "step": 48894 }, { "epoch": 3.831896551724138, "grad_norm": 3.9208765029907227, "learning_rate": 4.3686075264555705e-08, "loss": 1.2556, "step": 48895 }, { "epoch": 3.831974921630094, "grad_norm": 3.387234687805176, "learning_rate": 4.364541117222343e-08, "loss": 1.4528, "step": 48896 }, { "epoch": 3.83205329153605, "grad_norm": 2.9875621795654297, "learning_rate": 4.360476593131824e-08, "loss": 1.1625, "step": 48897 }, { "epoch": 3.8321316614420065, "grad_norm": 3.1560487747192383, "learning_rate": 4.356413954199501e-08, "loss": 1.2585, "step": 48898 }, { "epoch": 3.8322100313479623, "grad_norm": 2.1396145820617676, "learning_rate": 4.352353200440751e-08, "loss": 0.7678, "step": 48899 }, { "epoch": 3.8322884012539182, "grad_norm": 2.163187265396118, "learning_rate": 4.3482943318711166e-08, "loss": 0.7917, "step": 48900 }, { "epoch": 3.8323667711598746, "grad_norm": 3.9163131713867188, "learning_rate": 4.34423734850592e-08, "loss": 1.238, "step": 48901 }, { "epoch": 3.832445141065831, "grad_norm": 2.7152562141418457, "learning_rate": 4.340182250360703e-08, "loss": 1.5621, "step": 48902 }, { "epoch": 3.832523510971787, "grad_norm": 3.666935443878174, "learning_rate": 4.3361290374508425e-08, "loss": 1.3954, "step": 48903 }, { "epoch": 3.8326018808777427, "grad_norm": 4.3332319259643555, "learning_rate": 4.3320777097918266e-08, "loss": 1.3012, "step": 48904 }, { "epoch": 3.832680250783699, "grad_norm": 4.223921298980713, "learning_rate": 4.32802826739892e-08, "loss": 1.3518, "step": 48905 }, { "epoch": 3.8327586206896553, "grad_norm": 4.0334062576293945, "learning_rate": 4.323980710287612e-08, "loss": 1.1049, "step": 48906 }, { "epoch": 3.8328369905956112, "grad_norm": 2.3727149963378906, "learning_rate": 4.3199350384732775e-08, "loss": 0.8241, "step": 48907 }, { "epoch": 3.8329153605015676, "grad_norm": 3.4958674907684326, "learning_rate": 4.3158912519712935e-08, "loss": 1.0136, "step": 48908 }, { "epoch": 3.8329937304075234, "grad_norm": 2.4067206382751465, "learning_rate": 4.3118493507970924e-08, "loss": 1.1814, "step": 48909 }, { "epoch": 3.83307210031348, "grad_norm": 3.9693572521209717, "learning_rate": 4.3078093349659955e-08, "loss": 1.5186, "step": 48910 }, { "epoch": 3.8331504702194357, "grad_norm": 2.7319188117980957, "learning_rate": 4.303771204493434e-08, "loss": 1.5836, "step": 48911 }, { "epoch": 3.833228840125392, "grad_norm": 2.3259835243225098, "learning_rate": 4.299734959394675e-08, "loss": 1.0901, "step": 48912 }, { "epoch": 3.833307210031348, "grad_norm": 3.3375911712646484, "learning_rate": 4.2957005996851487e-08, "loss": 1.07, "step": 48913 }, { "epoch": 3.833385579937304, "grad_norm": 5.338111877441406, "learning_rate": 4.291668125380122e-08, "loss": 0.8041, "step": 48914 }, { "epoch": 3.83346394984326, "grad_norm": 2.97756290435791, "learning_rate": 4.287637536495026e-08, "loss": 1.3073, "step": 48915 }, { "epoch": 3.8335423197492164, "grad_norm": 3.720047950744629, "learning_rate": 4.283608833045128e-08, "loss": 1.1275, "step": 48916 }, { "epoch": 3.8336206896551723, "grad_norm": 4.964085578918457, "learning_rate": 4.279582015045747e-08, "loss": 1.3741, "step": 48917 }, { "epoch": 3.8336990595611287, "grad_norm": 12.631950378417969, "learning_rate": 4.275557082512205e-08, "loss": 0.9214, "step": 48918 }, { "epoch": 3.8337774294670846, "grad_norm": 5.6148362159729, "learning_rate": 4.271534035459879e-08, "loss": 1.6429, "step": 48919 }, { "epoch": 3.833855799373041, "grad_norm": 9.002052307128906, "learning_rate": 4.2675128739039785e-08, "loss": 1.6363, "step": 48920 }, { "epoch": 3.8339341692789968, "grad_norm": 4.073441028594971, "learning_rate": 4.263493597859825e-08, "loss": 1.2422, "step": 48921 }, { "epoch": 3.834012539184953, "grad_norm": 2.712339401245117, "learning_rate": 4.2594762073427945e-08, "loss": 1.1752, "step": 48922 }, { "epoch": 3.834090909090909, "grad_norm": 3.082648515701294, "learning_rate": 4.255460702367986e-08, "loss": 1.4606, "step": 48923 }, { "epoch": 3.8341692789968653, "grad_norm": 3.8570172786712646, "learning_rate": 4.251447082950832e-08, "loss": 1.3626, "step": 48924 }, { "epoch": 3.834247648902821, "grad_norm": 2.5406298637390137, "learning_rate": 4.247435349106488e-08, "loss": 1.4632, "step": 48925 }, { "epoch": 3.8343260188087775, "grad_norm": 3.11818790435791, "learning_rate": 4.243425500850329e-08, "loss": 1.2971, "step": 48926 }, { "epoch": 3.8344043887147334, "grad_norm": 3.455580472946167, "learning_rate": 4.23941753819751e-08, "loss": 1.2152, "step": 48927 }, { "epoch": 3.8344827586206898, "grad_norm": 2.229421854019165, "learning_rate": 4.235411461163297e-08, "loss": 1.4786, "step": 48928 }, { "epoch": 3.8345611285266457, "grad_norm": 4.041473388671875, "learning_rate": 4.231407269762955e-08, "loss": 1.5029, "step": 48929 }, { "epoch": 3.834639498432602, "grad_norm": 6.5136799812316895, "learning_rate": 4.2274049640116944e-08, "loss": 1.0191, "step": 48930 }, { "epoch": 3.834717868338558, "grad_norm": 3.2440173625946045, "learning_rate": 4.2234045439247805e-08, "loss": 1.5548, "step": 48931 }, { "epoch": 3.834796238244514, "grad_norm": 2.89604115486145, "learning_rate": 4.219406009517313e-08, "loss": 1.1871, "step": 48932 }, { "epoch": 3.83487460815047, "grad_norm": 3.109743356704712, "learning_rate": 4.215409360804612e-08, "loss": 1.561, "step": 48933 }, { "epoch": 3.8349529780564264, "grad_norm": 2.649177312850952, "learning_rate": 4.2114145978018326e-08, "loss": 1.2262, "step": 48934 }, { "epoch": 3.8350313479623823, "grad_norm": 5.712460517883301, "learning_rate": 4.207421720524185e-08, "loss": 1.3894, "step": 48935 }, { "epoch": 3.8351097178683387, "grad_norm": 6.418197154998779, "learning_rate": 4.203430728986879e-08, "loss": 1.0006, "step": 48936 }, { "epoch": 3.8351880877742945, "grad_norm": 3.3953657150268555, "learning_rate": 4.19944162320507e-08, "loss": 1.3396, "step": 48937 }, { "epoch": 3.835266457680251, "grad_norm": 2.858175754547119, "learning_rate": 4.1954544031939105e-08, "loss": 1.3979, "step": 48938 }, { "epoch": 3.8353448275862068, "grad_norm": 3.711186170578003, "learning_rate": 4.191469068968501e-08, "loss": 1.3196, "step": 48939 }, { "epoch": 3.835423197492163, "grad_norm": 3.7663660049438477, "learning_rate": 4.187485620544163e-08, "loss": 1.5595, "step": 48940 }, { "epoch": 3.835501567398119, "grad_norm": 4.084498882293701, "learning_rate": 4.1835040579359945e-08, "loss": 1.5484, "step": 48941 }, { "epoch": 3.8355799373040753, "grad_norm": 10.813821792602539, "learning_rate": 4.179524381159039e-08, "loss": 0.8287, "step": 48942 }, { "epoch": 3.835658307210031, "grad_norm": 3.561711549758911, "learning_rate": 4.175546590228508e-08, "loss": 1.4567, "step": 48943 }, { "epoch": 3.8357366771159875, "grad_norm": 2.953094244003296, "learning_rate": 4.1715706851595537e-08, "loss": 1.4053, "step": 48944 }, { "epoch": 3.8358150470219434, "grad_norm": 3.138676166534424, "learning_rate": 4.167596665967277e-08, "loss": 1.6361, "step": 48945 }, { "epoch": 3.8358934169278998, "grad_norm": 3.0831146240234375, "learning_rate": 4.163624532666777e-08, "loss": 1.2612, "step": 48946 }, { "epoch": 3.8359717868338556, "grad_norm": 2.713475465774536, "learning_rate": 4.159654285273151e-08, "loss": 1.15, "step": 48947 }, { "epoch": 3.836050156739812, "grad_norm": 3.5438179969787598, "learning_rate": 4.1556859238015554e-08, "loss": 1.4239, "step": 48948 }, { "epoch": 3.836128526645768, "grad_norm": 6.999138355255127, "learning_rate": 4.151719448267033e-08, "loss": 1.0622, "step": 48949 }, { "epoch": 3.836206896551724, "grad_norm": 4.237657070159912, "learning_rate": 4.147754858684683e-08, "loss": 0.9955, "step": 48950 }, { "epoch": 3.83628526645768, "grad_norm": 2.8865106105804443, "learning_rate": 4.143792155069548e-08, "loss": 1.3622, "step": 48951 }, { "epoch": 3.8363636363636364, "grad_norm": 4.152734756469727, "learning_rate": 4.1398313374368394e-08, "loss": 1.4203, "step": 48952 }, { "epoch": 3.8364420062695928, "grad_norm": 2.7501091957092285, "learning_rate": 4.135872405801433e-08, "loss": 1.6532, "step": 48953 }, { "epoch": 3.8365203761755486, "grad_norm": 3.7278096675872803, "learning_rate": 4.131915360178485e-08, "loss": 1.499, "step": 48954 }, { "epoch": 3.8365987460815045, "grad_norm": 5.242528438568115, "learning_rate": 4.127960200583037e-08, "loss": 1.3629, "step": 48955 }, { "epoch": 3.836677115987461, "grad_norm": 5.935028076171875, "learning_rate": 4.124006927030189e-08, "loss": 1.6345, "step": 48956 }, { "epoch": 3.836755485893417, "grad_norm": 2.685894727706909, "learning_rate": 4.120055539534873e-08, "loss": 1.4016, "step": 48957 }, { "epoch": 3.836833855799373, "grad_norm": 3.8911757469177246, "learning_rate": 4.1161060381121885e-08, "loss": 0.7502, "step": 48958 }, { "epoch": 3.836912225705329, "grad_norm": 10.134766578674316, "learning_rate": 4.112158422777124e-08, "loss": 1.4021, "step": 48959 }, { "epoch": 3.8369905956112853, "grad_norm": 4.345107078552246, "learning_rate": 4.1082126935447216e-08, "loss": 0.7808, "step": 48960 }, { "epoch": 3.8370689655172416, "grad_norm": 3.049062967300415, "learning_rate": 4.1042688504299156e-08, "loss": 1.3159, "step": 48961 }, { "epoch": 3.8371473354231975, "grad_norm": 2.6980559825897217, "learning_rate": 4.1003268934478034e-08, "loss": 1.4085, "step": 48962 }, { "epoch": 3.8372257053291534, "grad_norm": 13.884552001953125, "learning_rate": 4.096386822613374e-08, "loss": 2.2179, "step": 48963 }, { "epoch": 3.8373040752351097, "grad_norm": 2.3838305473327637, "learning_rate": 4.092448637941504e-08, "loss": 1.1408, "step": 48964 }, { "epoch": 3.837382445141066, "grad_norm": 3.1964259147644043, "learning_rate": 4.0885123394472926e-08, "loss": 1.4098, "step": 48965 }, { "epoch": 3.837460815047022, "grad_norm": 3.3995096683502197, "learning_rate": 4.084577927145672e-08, "loss": 1.347, "step": 48966 }, { "epoch": 3.837539184952978, "grad_norm": 2.036205530166626, "learning_rate": 4.080645401051575e-08, "loss": 1.2144, "step": 48967 }, { "epoch": 3.837617554858934, "grad_norm": 2.9261622428894043, "learning_rate": 4.0767147611799895e-08, "loss": 1.0564, "step": 48968 }, { "epoch": 3.8376959247648905, "grad_norm": 3.5228335857391357, "learning_rate": 4.072786007545848e-08, "loss": 1.2219, "step": 48969 }, { "epoch": 3.8377742946708464, "grad_norm": 3.924544095993042, "learning_rate": 4.068859140164083e-08, "loss": 1.55, "step": 48970 }, { "epoch": 3.8378526645768023, "grad_norm": 3.301809072494507, "learning_rate": 4.064934159049738e-08, "loss": 1.1452, "step": 48971 }, { "epoch": 3.8379310344827586, "grad_norm": 7.228007793426514, "learning_rate": 4.0610110642175236e-08, "loss": 1.5831, "step": 48972 }, { "epoch": 3.838009404388715, "grad_norm": 3.7935616970062256, "learning_rate": 4.057089855682595e-08, "loss": 0.9553, "step": 48973 }, { "epoch": 3.838087774294671, "grad_norm": 4.093206882476807, "learning_rate": 4.053170533459716e-08, "loss": 1.788, "step": 48974 }, { "epoch": 3.8381661442006267, "grad_norm": 5.0513081550598145, "learning_rate": 4.049253097563877e-08, "loss": 1.2975, "step": 48975 }, { "epoch": 3.838244514106583, "grad_norm": 4.957297325134277, "learning_rate": 4.0453375480099534e-08, "loss": 1.6007, "step": 48976 }, { "epoch": 3.8383228840125394, "grad_norm": 2.6674017906188965, "learning_rate": 4.0414238848127675e-08, "loss": 1.3243, "step": 48977 }, { "epoch": 3.8384012539184953, "grad_norm": 4.294142246246338, "learning_rate": 4.037512107987307e-08, "loss": 1.4105, "step": 48978 }, { "epoch": 3.838479623824451, "grad_norm": 3.1145999431610107, "learning_rate": 4.033602217548394e-08, "loss": 1.7304, "step": 48979 }, { "epoch": 3.8385579937304075, "grad_norm": 3.1134190559387207, "learning_rate": 4.029694213510904e-08, "loss": 1.3878, "step": 48980 }, { "epoch": 3.838636363636364, "grad_norm": 7.735398769378662, "learning_rate": 4.0257880958897714e-08, "loss": 0.9439, "step": 48981 }, { "epoch": 3.8387147335423197, "grad_norm": 5.526610374450684, "learning_rate": 4.021883864699705e-08, "loss": 0.9431, "step": 48982 }, { "epoch": 3.8387931034482756, "grad_norm": 3.285270929336548, "learning_rate": 4.0179815199556936e-08, "loss": 1.333, "step": 48983 }, { "epoch": 3.838871473354232, "grad_norm": 3.6932523250579834, "learning_rate": 4.0140810616725034e-08, "loss": 0.9729, "step": 48984 }, { "epoch": 3.8389498432601883, "grad_norm": 3.246206283569336, "learning_rate": 4.010182489865011e-08, "loss": 1.4011, "step": 48985 }, { "epoch": 3.839028213166144, "grad_norm": 3.047165870666504, "learning_rate": 4.006285804548038e-08, "loss": 1.5591, "step": 48986 }, { "epoch": 3.8391065830721, "grad_norm": 3.4885056018829346, "learning_rate": 4.0023910057364055e-08, "loss": 1.5655, "step": 48987 }, { "epoch": 3.8391849529780564, "grad_norm": 4.404240608215332, "learning_rate": 3.99849809344488e-08, "loss": 1.4634, "step": 48988 }, { "epoch": 3.8392633228840127, "grad_norm": 2.8994297981262207, "learning_rate": 3.994607067688283e-08, "loss": 1.5813, "step": 48989 }, { "epoch": 3.8393416927899686, "grad_norm": 2.7360634803771973, "learning_rate": 3.9907179284815466e-08, "loss": 1.1421, "step": 48990 }, { "epoch": 3.8394200626959245, "grad_norm": 4.099824905395508, "learning_rate": 3.986830675839215e-08, "loss": 1.3137, "step": 48991 }, { "epoch": 3.839498432601881, "grad_norm": 3.076838731765747, "learning_rate": 3.982945309776276e-08, "loss": 1.4494, "step": 48992 }, { "epoch": 3.839576802507837, "grad_norm": 4.765464782714844, "learning_rate": 3.979061830307496e-08, "loss": 1.6128, "step": 48993 }, { "epoch": 3.839655172413793, "grad_norm": 10.677454948425293, "learning_rate": 3.975180237447529e-08, "loss": 1.1134, "step": 48994 }, { "epoch": 3.839733542319749, "grad_norm": 2.4177980422973633, "learning_rate": 3.971300531211197e-08, "loss": 1.5094, "step": 48995 }, { "epoch": 3.8398119122257053, "grad_norm": 4.4458794593811035, "learning_rate": 3.967422711613267e-08, "loss": 1.5872, "step": 48996 }, { "epoch": 3.8398902821316616, "grad_norm": 5.480125427246094, "learning_rate": 3.9635467786685037e-08, "loss": 1.6331, "step": 48997 }, { "epoch": 3.8399686520376175, "grad_norm": 2.7220029830932617, "learning_rate": 3.9596727323916175e-08, "loss": 0.9027, "step": 48998 }, { "epoch": 3.8400470219435734, "grad_norm": 5.205286026000977, "learning_rate": 3.9558005727973746e-08, "loss": 1.394, "step": 48999 }, { "epoch": 3.8401253918495297, "grad_norm": 2.3968539237976074, "learning_rate": 3.9519302999004305e-08, "loss": 1.3642, "step": 49000 }, { "epoch": 3.840203761755486, "grad_norm": 2.8083934783935547, "learning_rate": 3.9480619137156064e-08, "loss": 1.0583, "step": 49001 }, { "epoch": 3.840282131661442, "grad_norm": 3.4117603302001953, "learning_rate": 3.944195414257501e-08, "loss": 1.3245, "step": 49002 }, { "epoch": 3.8403605015673983, "grad_norm": 6.2544636726379395, "learning_rate": 3.940330801540937e-08, "loss": 0.9895, "step": 49003 }, { "epoch": 3.840438871473354, "grad_norm": 3.2716176509857178, "learning_rate": 3.9364680755805685e-08, "loss": 1.5218, "step": 49004 }, { "epoch": 3.8405172413793105, "grad_norm": 2.877990961074829, "learning_rate": 3.932607236391051e-08, "loss": 1.7138, "step": 49005 }, { "epoch": 3.8405956112852664, "grad_norm": 3.1239001750946045, "learning_rate": 3.9287482839871497e-08, "loss": 1.4552, "step": 49006 }, { "epoch": 3.8406739811912227, "grad_norm": 4.355042457580566, "learning_rate": 3.924891218383464e-08, "loss": 1.474, "step": 49007 }, { "epoch": 3.8407523510971786, "grad_norm": 3.2309651374816895, "learning_rate": 3.9210360395947054e-08, "loss": 1.3685, "step": 49008 }, { "epoch": 3.840830721003135, "grad_norm": 2.748412609100342, "learning_rate": 3.917182747635473e-08, "loss": 1.3873, "step": 49009 }, { "epoch": 3.840909090909091, "grad_norm": 22.085281372070312, "learning_rate": 3.9133313425204763e-08, "loss": 2.1064, "step": 49010 }, { "epoch": 3.840987460815047, "grad_norm": 2.4561169147491455, "learning_rate": 3.909481824264427e-08, "loss": 1.1301, "step": 49011 }, { "epoch": 3.841065830721003, "grad_norm": 4.333619594573975, "learning_rate": 3.9056341928818685e-08, "loss": 1.5829, "step": 49012 }, { "epoch": 3.8411442006269594, "grad_norm": 5.606628894805908, "learning_rate": 3.9017884483874554e-08, "loss": 1.3319, "step": 49013 }, { "epoch": 3.8412225705329153, "grad_norm": 3.5862538814544678, "learning_rate": 3.897944590795788e-08, "loss": 1.257, "step": 49014 }, { "epoch": 3.8413009404388716, "grad_norm": 4.897860527038574, "learning_rate": 3.894102620121576e-08, "loss": 1.0731, "step": 49015 }, { "epoch": 3.8413793103448275, "grad_norm": 3.178419828414917, "learning_rate": 3.890262536379419e-08, "loss": 1.4184, "step": 49016 }, { "epoch": 3.841457680250784, "grad_norm": 3.218228578567505, "learning_rate": 3.886424339583861e-08, "loss": 1.5263, "step": 49017 }, { "epoch": 3.8415360501567397, "grad_norm": 5.905831336975098, "learning_rate": 3.882588029749501e-08, "loss": 1.7088, "step": 49018 }, { "epoch": 3.841614420062696, "grad_norm": 2.3362677097320557, "learning_rate": 3.878753606890939e-08, "loss": 1.4269, "step": 49019 }, { "epoch": 3.841692789968652, "grad_norm": 3.00439190864563, "learning_rate": 3.874921071022775e-08, "loss": 1.206, "step": 49020 }, { "epoch": 3.8417711598746083, "grad_norm": 7.070531845092773, "learning_rate": 3.8710904221596065e-08, "loss": 1.2062, "step": 49021 }, { "epoch": 3.841849529780564, "grad_norm": 4.2560529708862305, "learning_rate": 3.8672616603159795e-08, "loss": 1.3307, "step": 49022 }, { "epoch": 3.8419278996865205, "grad_norm": 3.5995326042175293, "learning_rate": 3.8634347855064926e-08, "loss": 1.1648, "step": 49023 }, { "epoch": 3.8420062695924764, "grad_norm": 5.320456027984619, "learning_rate": 3.859609797745634e-08, "loss": 1.412, "step": 49024 }, { "epoch": 3.8420846394984327, "grad_norm": 2.993964433670044, "learning_rate": 3.855786697047947e-08, "loss": 1.0808, "step": 49025 }, { "epoch": 3.8421630094043886, "grad_norm": 2.784374475479126, "learning_rate": 3.8519654834280885e-08, "loss": 1.3567, "step": 49026 }, { "epoch": 3.842241379310345, "grad_norm": 3.6271004676818848, "learning_rate": 3.8481461569004894e-08, "loss": 1.6699, "step": 49027 }, { "epoch": 3.842319749216301, "grad_norm": 3.884171485900879, "learning_rate": 3.84432871747964e-08, "loss": 1.3363, "step": 49028 }, { "epoch": 3.842398119122257, "grad_norm": 6.203306198120117, "learning_rate": 3.840513165180193e-08, "loss": 1.1649, "step": 49029 }, { "epoch": 3.842476489028213, "grad_norm": 3.22727108001709, "learning_rate": 3.836699500016583e-08, "loss": 1.3332, "step": 49030 }, { "epoch": 3.8425548589341694, "grad_norm": 3.00673246383667, "learning_rate": 3.832887722003353e-08, "loss": 1.2031, "step": 49031 }, { "epoch": 3.8426332288401253, "grad_norm": 4.447503089904785, "learning_rate": 3.8290778311548814e-08, "loss": 1.2684, "step": 49032 }, { "epoch": 3.8427115987460816, "grad_norm": 3.0010478496551514, "learning_rate": 3.8252698274857667e-08, "loss": 0.8102, "step": 49033 }, { "epoch": 3.8427899686520375, "grad_norm": 3.6326904296875, "learning_rate": 3.821463711010554e-08, "loss": 1.3524, "step": 49034 }, { "epoch": 3.842868338557994, "grad_norm": 2.938133955001831, "learning_rate": 3.817659481743563e-08, "loss": 1.3336, "step": 49035 }, { "epoch": 3.8429467084639497, "grad_norm": 13.061896324157715, "learning_rate": 3.8138571396993953e-08, "loss": 1.0086, "step": 49036 }, { "epoch": 3.843025078369906, "grad_norm": 5.057626247406006, "learning_rate": 3.810056684892427e-08, "loss": 1.506, "step": 49037 }, { "epoch": 3.843103448275862, "grad_norm": 5.07572078704834, "learning_rate": 3.806258117337147e-08, "loss": 1.3543, "step": 49038 }, { "epoch": 3.8431818181818183, "grad_norm": 8.01555347442627, "learning_rate": 3.8024614370479884e-08, "loss": 1.5714, "step": 49039 }, { "epoch": 3.843260188087774, "grad_norm": 5.25281286239624, "learning_rate": 3.798666644039439e-08, "loss": 2.0617, "step": 49040 }, { "epoch": 3.8433385579937305, "grad_norm": 6.025768756866455, "learning_rate": 3.794873738325822e-08, "loss": 1.622, "step": 49041 }, { "epoch": 3.8434169278996864, "grad_norm": 3.469155788421631, "learning_rate": 3.791082719921735e-08, "loss": 1.1441, "step": 49042 }, { "epoch": 3.8434952978056427, "grad_norm": 4.996451377868652, "learning_rate": 3.787293588841445e-08, "loss": 1.4351, "step": 49043 }, { "epoch": 3.8435736677115986, "grad_norm": 4.934927940368652, "learning_rate": 3.7835063450993856e-08, "loss": 1.1696, "step": 49044 }, { "epoch": 3.843652037617555, "grad_norm": 4.152752876281738, "learning_rate": 3.779720988710045e-08, "loss": 1.6062, "step": 49045 }, { "epoch": 3.843730407523511, "grad_norm": 3.375655174255371, "learning_rate": 3.7759375196878e-08, "loss": 1.1394, "step": 49046 }, { "epoch": 3.843808777429467, "grad_norm": 3.2899203300476074, "learning_rate": 3.772155938046973e-08, "loss": 1.568, "step": 49047 }, { "epoch": 3.8438871473354235, "grad_norm": 8.845149993896484, "learning_rate": 3.768376243801941e-08, "loss": 0.8806, "step": 49048 }, { "epoch": 3.8439655172413794, "grad_norm": 3.766894578933716, "learning_rate": 3.7645984369671926e-08, "loss": 1.3578, "step": 49049 }, { "epoch": 3.8440438871473352, "grad_norm": 4.998951435089111, "learning_rate": 3.76082251755705e-08, "loss": 1.1209, "step": 49050 }, { "epoch": 3.8441222570532916, "grad_norm": 4.253142833709717, "learning_rate": 3.7570484855857793e-08, "loss": 1.2811, "step": 49051 }, { "epoch": 3.844200626959248, "grad_norm": 3.1220896244049072, "learning_rate": 3.753276341067813e-08, "loss": 1.5615, "step": 49052 }, { "epoch": 3.844278996865204, "grad_norm": 2.6792988777160645, "learning_rate": 3.7495060840175843e-08, "loss": 1.1901, "step": 49053 }, { "epoch": 3.8443573667711597, "grad_norm": 4.451587677001953, "learning_rate": 3.745737714449249e-08, "loss": 1.5557, "step": 49054 }, { "epoch": 3.844435736677116, "grad_norm": 3.265897274017334, "learning_rate": 3.741971232377295e-08, "loss": 1.4579, "step": 49055 }, { "epoch": 3.8445141065830724, "grad_norm": 3.1544501781463623, "learning_rate": 3.738206637815933e-08, "loss": 1.5768, "step": 49056 }, { "epoch": 3.8445924764890282, "grad_norm": 6.4002461433410645, "learning_rate": 3.734443930779597e-08, "loss": 1.6202, "step": 49057 }, { "epoch": 3.844670846394984, "grad_norm": 3.1923696994781494, "learning_rate": 3.730683111282496e-08, "loss": 1.5455, "step": 49058 }, { "epoch": 3.8447492163009405, "grad_norm": 5.065913200378418, "learning_rate": 3.726924179339009e-08, "loss": 1.2575, "step": 49059 }, { "epoch": 3.844827586206897, "grad_norm": 2.9165821075439453, "learning_rate": 3.7231671349634015e-08, "loss": 1.3975, "step": 49060 }, { "epoch": 3.8449059561128527, "grad_norm": 2.607452630996704, "learning_rate": 3.7194119781698845e-08, "loss": 1.4626, "step": 49061 }, { "epoch": 3.8449843260188086, "grad_norm": 2.317352533340454, "learning_rate": 3.715658708972891e-08, "loss": 1.3503, "step": 49062 }, { "epoch": 3.845062695924765, "grad_norm": 4.881511688232422, "learning_rate": 3.7119073273865766e-08, "loss": 1.6025, "step": 49063 }, { "epoch": 3.8451410658307212, "grad_norm": 3.139204740524292, "learning_rate": 3.708157833425319e-08, "loss": 1.2001, "step": 49064 }, { "epoch": 3.845219435736677, "grad_norm": 4.193912506103516, "learning_rate": 3.7044102271033275e-08, "loss": 1.5283, "step": 49065 }, { "epoch": 3.845297805642633, "grad_norm": 3.658942937850952, "learning_rate": 3.700664508434759e-08, "loss": 1.3815, "step": 49066 }, { "epoch": 3.8453761755485893, "grad_norm": 8.551335334777832, "learning_rate": 3.69692067743399e-08, "loss": 1.2034, "step": 49067 }, { "epoch": 3.8454545454545457, "grad_norm": 2.4731075763702393, "learning_rate": 3.6931787341152324e-08, "loss": 1.2007, "step": 49068 }, { "epoch": 3.8455329153605016, "grad_norm": 3.4711410999298096, "learning_rate": 3.689438678492696e-08, "loss": 1.4292, "step": 49069 }, { "epoch": 3.8456112852664575, "grad_norm": 4.230776309967041, "learning_rate": 3.6857005105806474e-08, "loss": 0.8143, "step": 49070 }, { "epoch": 3.845689655172414, "grad_norm": 6.785980224609375, "learning_rate": 3.6819642303931866e-08, "loss": 1.8431, "step": 49071 }, { "epoch": 3.84576802507837, "grad_norm": 4.753542900085449, "learning_rate": 3.678229837944691e-08, "loss": 1.2319, "step": 49072 }, { "epoch": 3.845846394984326, "grad_norm": 5.178539276123047, "learning_rate": 3.6744973332492054e-08, "loss": 1.2819, "step": 49073 }, { "epoch": 3.845924764890282, "grad_norm": 2.9679245948791504, "learning_rate": 3.67076671632105e-08, "loss": 1.642, "step": 49074 }, { "epoch": 3.8460031347962382, "grad_norm": 2.351471185684204, "learning_rate": 3.6670379871743825e-08, "loss": 1.2842, "step": 49075 }, { "epoch": 3.8460815047021946, "grad_norm": 3.8599531650543213, "learning_rate": 3.663311145823356e-08, "loss": 2.0004, "step": 49076 }, { "epoch": 3.8461598746081505, "grad_norm": 2.4188568592071533, "learning_rate": 3.659586192282128e-08, "loss": 1.221, "step": 49077 }, { "epoch": 3.8462382445141063, "grad_norm": 3.0532145500183105, "learning_rate": 3.655863126564907e-08, "loss": 1.3416, "step": 49078 }, { "epoch": 3.8463166144200627, "grad_norm": 4.156097412109375, "learning_rate": 3.652141948685906e-08, "loss": 1.4133, "step": 49079 }, { "epoch": 3.846394984326019, "grad_norm": 2.526705503463745, "learning_rate": 3.6484226586591674e-08, "loss": 1.3709, "step": 49080 }, { "epoch": 3.846473354231975, "grad_norm": 2.4344167709350586, "learning_rate": 3.6447052564988484e-08, "loss": 1.5521, "step": 49081 }, { "epoch": 3.846551724137931, "grad_norm": 5.436077117919922, "learning_rate": 3.6409897422191585e-08, "loss": 1.5057, "step": 49082 }, { "epoch": 3.846630094043887, "grad_norm": 3.647432804107666, "learning_rate": 3.6372761158342537e-08, "loss": 1.4711, "step": 49083 }, { "epoch": 3.8467084639498434, "grad_norm": 3.1317951679229736, "learning_rate": 3.633564377358179e-08, "loss": 2.0966, "step": 49084 }, { "epoch": 3.8467868338557993, "grad_norm": 3.2375857830047607, "learning_rate": 3.629854526804977e-08, "loss": 1.6603, "step": 49085 }, { "epoch": 3.8468652037617552, "grad_norm": 4.481471538543701, "learning_rate": 3.62614656418897e-08, "loss": 1.5944, "step": 49086 }, { "epoch": 3.8469435736677116, "grad_norm": 3.4817142486572266, "learning_rate": 3.622440489524148e-08, "loss": 1.0387, "step": 49087 }, { "epoch": 3.847021943573668, "grad_norm": 4.55098819732666, "learning_rate": 3.618736302824555e-08, "loss": 1.1413, "step": 49088 }, { "epoch": 3.847100313479624, "grad_norm": 9.774527549743652, "learning_rate": 3.615034004104345e-08, "loss": 1.0742, "step": 49089 }, { "epoch": 3.8471786833855797, "grad_norm": 2.8879265785217285, "learning_rate": 3.611333593377564e-08, "loss": 1.1907, "step": 49090 }, { "epoch": 3.847257053291536, "grad_norm": 2.60431170463562, "learning_rate": 3.607635070658311e-08, "loss": 1.353, "step": 49091 }, { "epoch": 3.8473354231974923, "grad_norm": 3.5168604850769043, "learning_rate": 3.603938435960686e-08, "loss": 0.9665, "step": 49092 }, { "epoch": 3.847413793103448, "grad_norm": 4.014951229095459, "learning_rate": 3.600243689298677e-08, "loss": 1.5051, "step": 49093 }, { "epoch": 3.8474921630094046, "grad_norm": 2.4905271530151367, "learning_rate": 3.596550830686441e-08, "loss": 0.9079, "step": 49094 }, { "epoch": 3.8475705329153604, "grad_norm": 2.912484884262085, "learning_rate": 3.5928598601379093e-08, "loss": 1.3833, "step": 49095 }, { "epoch": 3.8476489028213168, "grad_norm": 9.309209823608398, "learning_rate": 3.589170777667128e-08, "loss": 0.6201, "step": 49096 }, { "epoch": 3.8477272727272727, "grad_norm": 2.6375114917755127, "learning_rate": 3.585483583288252e-08, "loss": 1.5599, "step": 49097 }, { "epoch": 3.847805642633229, "grad_norm": 4.312615394592285, "learning_rate": 3.581798277015158e-08, "loss": 1.5785, "step": 49098 }, { "epoch": 3.847884012539185, "grad_norm": 3.9858269691467285, "learning_rate": 3.5781148588619476e-08, "loss": 1.0826, "step": 49099 }, { "epoch": 3.847962382445141, "grad_norm": 3.3309926986694336, "learning_rate": 3.574433328842608e-08, "loss": 1.1697, "step": 49100 }, { "epoch": 3.848040752351097, "grad_norm": 9.309661865234375, "learning_rate": 3.570753686971129e-08, "loss": 0.8902, "step": 49101 }, { "epoch": 3.8481191222570534, "grad_norm": 6.517461776733398, "learning_rate": 3.5670759332615546e-08, "loss": 0.712, "step": 49102 }, { "epoch": 3.8481974921630093, "grad_norm": 4.3954644203186035, "learning_rate": 3.563400067727818e-08, "loss": 1.0888, "step": 49103 }, { "epoch": 3.8482758620689657, "grad_norm": 3.0714190006256104, "learning_rate": 3.559726090383908e-08, "loss": 1.2798, "step": 49104 }, { "epoch": 3.8483542319749215, "grad_norm": 3.5103399753570557, "learning_rate": 3.5560540012438694e-08, "loss": 1.4612, "step": 49105 }, { "epoch": 3.848432601880878, "grad_norm": 2.607654571533203, "learning_rate": 3.5523838003215796e-08, "loss": 1.2854, "step": 49106 }, { "epoch": 3.8485109717868338, "grad_norm": 3.9478814601898193, "learning_rate": 3.548715487631027e-08, "loss": 0.9906, "step": 49107 }, { "epoch": 3.84858934169279, "grad_norm": 2.968832492828369, "learning_rate": 3.545049063186146e-08, "loss": 1.3703, "step": 49108 }, { "epoch": 3.848667711598746, "grad_norm": 3.664172887802124, "learning_rate": 3.541384527000924e-08, "loss": 1.3779, "step": 49109 }, { "epoch": 3.8487460815047023, "grad_norm": 3.099118709564209, "learning_rate": 3.537721879089295e-08, "loss": 1.3657, "step": 49110 }, { "epoch": 3.848824451410658, "grad_norm": 2.844397783279419, "learning_rate": 3.534061119465138e-08, "loss": 1.5282, "step": 49111 }, { "epoch": 3.8489028213166145, "grad_norm": 2.6790354251861572, "learning_rate": 3.5304022481424394e-08, "loss": 1.1611, "step": 49112 }, { "epoch": 3.8489811912225704, "grad_norm": 2.3811328411102295, "learning_rate": 3.5267452651350786e-08, "loss": 1.3238, "step": 49113 }, { "epoch": 3.8490595611285268, "grad_norm": 2.961561679840088, "learning_rate": 3.523090170456933e-08, "loss": 1.1624, "step": 49114 }, { "epoch": 3.8491379310344827, "grad_norm": 3.907409191131592, "learning_rate": 3.5194369641219915e-08, "loss": 1.5787, "step": 49115 }, { "epoch": 3.849216300940439, "grad_norm": 3.5775227546691895, "learning_rate": 3.515785646144132e-08, "loss": 1.56, "step": 49116 }, { "epoch": 3.849294670846395, "grad_norm": 4.959665298461914, "learning_rate": 3.512136216537121e-08, "loss": 1.1545, "step": 49117 }, { "epoch": 3.849373040752351, "grad_norm": 2.1291778087615967, "learning_rate": 3.508488675315003e-08, "loss": 1.2572, "step": 49118 }, { "epoch": 3.849451410658307, "grad_norm": 2.9632959365844727, "learning_rate": 3.504843022491544e-08, "loss": 1.8841, "step": 49119 }, { "epoch": 3.8495297805642634, "grad_norm": 2.8712425231933594, "learning_rate": 3.501199258080734e-08, "loss": 1.3639, "step": 49120 }, { "epoch": 3.8496081504702193, "grad_norm": 3.5289554595947266, "learning_rate": 3.497557382096228e-08, "loss": 1.2526, "step": 49121 }, { "epoch": 3.8496865203761756, "grad_norm": 4.533815860748291, "learning_rate": 3.493917394552015e-08, "loss": 1.3714, "step": 49122 }, { "epoch": 3.8497648902821315, "grad_norm": 3.306978702545166, "learning_rate": 3.4902792954619735e-08, "loss": 1.1976, "step": 49123 }, { "epoch": 3.849843260188088, "grad_norm": 3.539776563644409, "learning_rate": 3.486643084839869e-08, "loss": 1.048, "step": 49124 }, { "epoch": 3.8499216300940438, "grad_norm": 5.962339878082275, "learning_rate": 3.483008762699525e-08, "loss": 1.2705, "step": 49125 }, { "epoch": 3.85, "grad_norm": 2.5708537101745605, "learning_rate": 3.479376329054818e-08, "loss": 1.371, "step": 49126 }, { "epoch": 3.850078369905956, "grad_norm": 3.696946620941162, "learning_rate": 3.4757457839195154e-08, "loss": 1.252, "step": 49127 }, { "epoch": 3.8501567398119123, "grad_norm": 2.9256792068481445, "learning_rate": 3.472117127307439e-08, "loss": 1.1668, "step": 49128 }, { "epoch": 3.850235109717868, "grad_norm": 7.346963405609131, "learning_rate": 3.4684903592324126e-08, "loss": 1.4134, "step": 49129 }, { "epoch": 3.8503134796238245, "grad_norm": 2.737562894821167, "learning_rate": 3.4648654797082015e-08, "loss": 1.2918, "step": 49130 }, { "epoch": 3.8503918495297804, "grad_norm": 5.3103814125061035, "learning_rate": 3.461242488748628e-08, "loss": 1.5215, "step": 49131 }, { "epoch": 3.8504702194357368, "grad_norm": 2.3497273921966553, "learning_rate": 3.4576213863674044e-08, "loss": 1.3299, "step": 49132 }, { "epoch": 3.8505485893416926, "grad_norm": 2.4355316162109375, "learning_rate": 3.454002172578408e-08, "loss": 1.0831, "step": 49133 }, { "epoch": 3.850626959247649, "grad_norm": 3.5912954807281494, "learning_rate": 3.450384847395294e-08, "loss": 1.3527, "step": 49134 }, { "epoch": 3.850705329153605, "grad_norm": 3.336589813232422, "learning_rate": 3.446769410831885e-08, "loss": 1.3109, "step": 49135 }, { "epoch": 3.850783699059561, "grad_norm": 3.195375442504883, "learning_rate": 3.443155862901893e-08, "loss": 1.2041, "step": 49136 }, { "epoch": 3.850862068965517, "grad_norm": 3.7304904460906982, "learning_rate": 3.439544203619138e-08, "loss": 0.9479, "step": 49137 }, { "epoch": 3.8509404388714734, "grad_norm": 3.3968942165374756, "learning_rate": 3.435934432997279e-08, "loss": 1.2197, "step": 49138 }, { "epoch": 3.8510188087774293, "grad_norm": 3.32814884185791, "learning_rate": 3.4323265510501355e-08, "loss": 1.9518, "step": 49139 }, { "epoch": 3.8510971786833856, "grad_norm": 17.86402702331543, "learning_rate": 3.4287205577913094e-08, "loss": 1.4779, "step": 49140 }, { "epoch": 3.8511755485893415, "grad_norm": 9.446895599365234, "learning_rate": 3.425116453234567e-08, "loss": 0.9758, "step": 49141 }, { "epoch": 3.851253918495298, "grad_norm": 4.951747417449951, "learning_rate": 3.421514237393619e-08, "loss": 1.0043, "step": 49142 }, { "epoch": 3.851332288401254, "grad_norm": 3.7109405994415283, "learning_rate": 3.4179139102822334e-08, "loss": 1.0626, "step": 49143 }, { "epoch": 3.85141065830721, "grad_norm": 2.5203516483306885, "learning_rate": 3.414315471914009e-08, "loss": 0.9583, "step": 49144 }, { "epoch": 3.851489028213166, "grad_norm": 5.205445289611816, "learning_rate": 3.410718922302658e-08, "loss": 1.0719, "step": 49145 }, { "epoch": 3.8515673981191223, "grad_norm": 8.342571258544922, "learning_rate": 3.407124261461836e-08, "loss": 1.1487, "step": 49146 }, { "epoch": 3.8516457680250786, "grad_norm": 4.891559600830078, "learning_rate": 3.40353148940531e-08, "loss": 1.5275, "step": 49147 }, { "epoch": 3.8517241379310345, "grad_norm": 4.573892116546631, "learning_rate": 3.399940606146623e-08, "loss": 1.0292, "step": 49148 }, { "epoch": 3.8518025078369904, "grad_norm": 2.475872278213501, "learning_rate": 3.396351611699544e-08, "loss": 1.3768, "step": 49149 }, { "epoch": 3.8518808777429467, "grad_norm": 3.241283893585205, "learning_rate": 3.3927645060776725e-08, "loss": 1.2549, "step": 49150 }, { "epoch": 3.851959247648903, "grad_norm": 3.8006412982940674, "learning_rate": 3.389179289294609e-08, "loss": 1.207, "step": 49151 }, { "epoch": 3.852037617554859, "grad_norm": 3.909040689468384, "learning_rate": 3.3855959613640634e-08, "loss": 1.19, "step": 49152 }, { "epoch": 3.852115987460815, "grad_norm": 3.4662039279937744, "learning_rate": 3.3820145222995814e-08, "loss": 1.2189, "step": 49153 }, { "epoch": 3.852194357366771, "grad_norm": 6.271177291870117, "learning_rate": 3.37843497211493e-08, "loss": 1.7, "step": 49154 }, { "epoch": 3.8522727272727275, "grad_norm": 2.1054344177246094, "learning_rate": 3.3748573108235425e-08, "loss": 1.1978, "step": 49155 }, { "epoch": 3.8523510971786834, "grad_norm": 9.35344123840332, "learning_rate": 3.37128153843913e-08, "loss": 1.2444, "step": 49156 }, { "epoch": 3.8524294670846393, "grad_norm": 2.889676094055176, "learning_rate": 3.3677076549753495e-08, "loss": 1.2389, "step": 49157 }, { "epoch": 3.8525078369905956, "grad_norm": 12.193429946899414, "learning_rate": 3.364135660445689e-08, "loss": 0.8717, "step": 49158 }, { "epoch": 3.852586206896552, "grad_norm": 6.291715621948242, "learning_rate": 3.360565554863748e-08, "loss": 0.8257, "step": 49159 }, { "epoch": 3.852664576802508, "grad_norm": 3.318443775177002, "learning_rate": 3.3569973382430734e-08, "loss": 1.253, "step": 49160 }, { "epoch": 3.8527429467084637, "grad_norm": 3.1308658123016357, "learning_rate": 3.353431010597374e-08, "loss": 1.3408, "step": 49161 }, { "epoch": 3.85282131661442, "grad_norm": 2.8678500652313232, "learning_rate": 3.3498665719400855e-08, "loss": 1.0115, "step": 49162 }, { "epoch": 3.8528996865203764, "grad_norm": 5.387290000915527, "learning_rate": 3.346304022284863e-08, "loss": 1.0877, "step": 49163 }, { "epoch": 3.8529780564263323, "grad_norm": 3.40421462059021, "learning_rate": 3.3427433616451396e-08, "loss": 1.3668, "step": 49164 }, { "epoch": 3.853056426332288, "grad_norm": 8.857292175292969, "learning_rate": 3.339184590034572e-08, "loss": 1.2986, "step": 49165 }, { "epoch": 3.8531347962382445, "grad_norm": 2.979362964630127, "learning_rate": 3.335627707466593e-08, "loss": 1.4075, "step": 49166 }, { "epoch": 3.853213166144201, "grad_norm": 5.84209680557251, "learning_rate": 3.3320727139548036e-08, "loss": 1.3889, "step": 49167 }, { "epoch": 3.8532915360501567, "grad_norm": 2.3634982109069824, "learning_rate": 3.328519609512693e-08, "loss": 1.2719, "step": 49168 }, { "epoch": 3.8533699059561126, "grad_norm": 2.4210379123687744, "learning_rate": 3.3249683941538046e-08, "loss": 1.3805, "step": 49169 }, { "epoch": 3.853448275862069, "grad_norm": 2.9399027824401855, "learning_rate": 3.321419067891629e-08, "loss": 1.2905, "step": 49170 }, { "epoch": 3.8535266457680253, "grad_norm": 5.9681830406188965, "learning_rate": 3.317871630739655e-08, "loss": 1.26, "step": 49171 }, { "epoch": 3.853605015673981, "grad_norm": 6.326141357421875, "learning_rate": 3.314326082711428e-08, "loss": 1.2141, "step": 49172 }, { "epoch": 3.853683385579937, "grad_norm": 2.961132764816284, "learning_rate": 3.31078242382038e-08, "loss": 1.6338, "step": 49173 }, { "epoch": 3.8537617554858934, "grad_norm": 6.612020015716553, "learning_rate": 3.3072406540799465e-08, "loss": 1.6644, "step": 49174 }, { "epoch": 3.8538401253918497, "grad_norm": 3.7069966793060303, "learning_rate": 3.303700773503671e-08, "loss": 1.2318, "step": 49175 }, { "epoch": 3.8539184952978056, "grad_norm": 2.5474014282226562, "learning_rate": 3.3001627821050433e-08, "loss": 1.2229, "step": 49176 }, { "epoch": 3.8539968652037615, "grad_norm": 4.413919925689697, "learning_rate": 3.296626679897441e-08, "loss": 1.2914, "step": 49177 }, { "epoch": 3.854075235109718, "grad_norm": 2.9043679237365723, "learning_rate": 3.293092466894299e-08, "loss": 1.4208, "step": 49178 }, { "epoch": 3.854153605015674, "grad_norm": 3.0961976051330566, "learning_rate": 3.289560143109161e-08, "loss": 1.543, "step": 49179 }, { "epoch": 3.85423197492163, "grad_norm": 3.922924757003784, "learning_rate": 3.286029708555405e-08, "loss": 1.3165, "step": 49180 }, { "epoch": 3.854310344827586, "grad_norm": 3.3962574005126953, "learning_rate": 3.282501163246466e-08, "loss": 1.0109, "step": 49181 }, { "epoch": 3.8543887147335423, "grad_norm": 5.8759636878967285, "learning_rate": 3.278974507195776e-08, "loss": 1.2174, "step": 49182 }, { "epoch": 3.8544670846394986, "grad_norm": 11.581851959228516, "learning_rate": 3.275449740416714e-08, "loss": 1.7433, "step": 49183 }, { "epoch": 3.8545454545454545, "grad_norm": 2.902677059173584, "learning_rate": 3.271926862922714e-08, "loss": 1.3262, "step": 49184 }, { "epoch": 3.8546238244514104, "grad_norm": 4.115169048309326, "learning_rate": 3.268405874727154e-08, "loss": 1.2835, "step": 49185 }, { "epoch": 3.8547021943573667, "grad_norm": 2.6002583503723145, "learning_rate": 3.2648867758434675e-08, "loss": 1.6278, "step": 49186 }, { "epoch": 3.854780564263323, "grad_norm": 4.241214275360107, "learning_rate": 3.2613695662850333e-08, "loss": 1.4664, "step": 49187 }, { "epoch": 3.854858934169279, "grad_norm": 3.3024182319641113, "learning_rate": 3.2578542460651175e-08, "loss": 1.2999, "step": 49188 }, { "epoch": 3.8549373040752353, "grad_norm": 3.050037384033203, "learning_rate": 3.254340815197265e-08, "loss": 0.7209, "step": 49189 }, { "epoch": 3.855015673981191, "grad_norm": 3.600184440612793, "learning_rate": 3.250829273694689e-08, "loss": 1.2452, "step": 49190 }, { "epoch": 3.8550940438871475, "grad_norm": 2.987874746322632, "learning_rate": 3.247319621570877e-08, "loss": 0.972, "step": 49191 }, { "epoch": 3.8551724137931034, "grad_norm": 2.1367170810699463, "learning_rate": 3.243811858839097e-08, "loss": 1.1639, "step": 49192 }, { "epoch": 3.8552507836990597, "grad_norm": 4.885409832000732, "learning_rate": 3.240305985512671e-08, "loss": 1.2731, "step": 49193 }, { "epoch": 3.8553291536050156, "grad_norm": 3.8181872367858887, "learning_rate": 3.2368020016049774e-08, "loss": 1.1397, "step": 49194 }, { "epoch": 3.855407523510972, "grad_norm": 2.597630739212036, "learning_rate": 3.233299907129395e-08, "loss": 1.5332, "step": 49195 }, { "epoch": 3.855485893416928, "grad_norm": 3.421623706817627, "learning_rate": 3.229799702099135e-08, "loss": 1.1374, "step": 49196 }, { "epoch": 3.855564263322884, "grad_norm": 2.5150721073150635, "learning_rate": 3.226301386527575e-08, "loss": 1.4445, "step": 49197 }, { "epoch": 3.85564263322884, "grad_norm": 3.3307628631591797, "learning_rate": 3.222804960427983e-08, "loss": 1.3811, "step": 49198 }, { "epoch": 3.8557210031347964, "grad_norm": 9.840625762939453, "learning_rate": 3.219310423813682e-08, "loss": 1.3802, "step": 49199 }, { "epoch": 3.8557993730407523, "grad_norm": 3.500314474105835, "learning_rate": 3.215817776697994e-08, "loss": 1.4759, "step": 49200 }, { "epoch": 3.8558777429467086, "grad_norm": 5.272976875305176, "learning_rate": 3.21232701909413e-08, "loss": 1.3624, "step": 49201 }, { "epoch": 3.8559561128526645, "grad_norm": 2.516387939453125, "learning_rate": 3.20883815101547e-08, "loss": 1.2383, "step": 49202 }, { "epoch": 3.856034482758621, "grad_norm": 2.516303777694702, "learning_rate": 3.205351172475169e-08, "loss": 1.1872, "step": 49203 }, { "epoch": 3.8561128526645767, "grad_norm": 3.5695433616638184, "learning_rate": 3.201866083486549e-08, "loss": 1.5498, "step": 49204 }, { "epoch": 3.856191222570533, "grad_norm": 6.914891719818115, "learning_rate": 3.198382884062823e-08, "loss": 1.799, "step": 49205 }, { "epoch": 3.856269592476489, "grad_norm": 3.4095635414123535, "learning_rate": 3.1949015742173126e-08, "loss": 1.3532, "step": 49206 }, { "epoch": 3.8563479623824453, "grad_norm": 5.1059088706970215, "learning_rate": 3.191422153963231e-08, "loss": 1.7338, "step": 49207 }, { "epoch": 3.856426332288401, "grad_norm": 2.3375651836395264, "learning_rate": 3.187944623313788e-08, "loss": 1.1412, "step": 49208 }, { "epoch": 3.8565047021943575, "grad_norm": 2.5982460975646973, "learning_rate": 3.184468982282252e-08, "loss": 1.3823, "step": 49209 }, { "epoch": 3.8565830721003134, "grad_norm": 2.815286636352539, "learning_rate": 3.1809952308818336e-08, "loss": 0.6977, "step": 49210 }, { "epoch": 3.8566614420062697, "grad_norm": 7.148352146148682, "learning_rate": 3.1775233691256345e-08, "loss": 1.3436, "step": 49211 }, { "epoch": 3.8567398119122256, "grad_norm": 7.704140663146973, "learning_rate": 3.1740533970270325e-08, "loss": 1.9424, "step": 49212 }, { "epoch": 3.856818181818182, "grad_norm": 7.291317939758301, "learning_rate": 3.1705853145991286e-08, "loss": 1.9694, "step": 49213 }, { "epoch": 3.856896551724138, "grad_norm": 3.075288772583008, "learning_rate": 3.167119121855133e-08, "loss": 1.3649, "step": 49214 }, { "epoch": 3.856974921630094, "grad_norm": 3.632617473602295, "learning_rate": 3.163654818808204e-08, "loss": 1.7062, "step": 49215 }, { "epoch": 3.85705329153605, "grad_norm": 4.372409343719482, "learning_rate": 3.160192405471496e-08, "loss": 1.2805, "step": 49216 }, { "epoch": 3.8571316614420064, "grad_norm": 3.284450054168701, "learning_rate": 3.1567318818583324e-08, "loss": 1.5391, "step": 49217 }, { "epoch": 3.8572100313479623, "grad_norm": 4.077112197875977, "learning_rate": 3.153273247981703e-08, "loss": 1.3184, "step": 49218 }, { "epoch": 3.8572884012539186, "grad_norm": 3.6028940677642822, "learning_rate": 3.149816503854819e-08, "loss": 1.2326, "step": 49219 }, { "epoch": 3.8573667711598745, "grad_norm": 3.3103599548339844, "learning_rate": 3.146361649490837e-08, "loss": 1.3874, "step": 49220 }, { "epoch": 3.857445141065831, "grad_norm": 2.5835094451904297, "learning_rate": 3.1429086849029123e-08, "loss": 1.4526, "step": 49221 }, { "epoch": 3.8575235109717867, "grad_norm": 2.896275520324707, "learning_rate": 3.139457610104202e-08, "loss": 1.2759, "step": 49222 }, { "epoch": 3.857601880877743, "grad_norm": 6.352147579193115, "learning_rate": 3.136008425107695e-08, "loss": 1.1936, "step": 49223 }, { "epoch": 3.857680250783699, "grad_norm": 3.515101671218872, "learning_rate": 3.132561129926659e-08, "loss": 1.1112, "step": 49224 }, { "epoch": 3.8577586206896552, "grad_norm": 3.3625950813293457, "learning_rate": 3.1291157245741946e-08, "loss": 1.1345, "step": 49225 }, { "epoch": 3.857836990595611, "grad_norm": 4.098387241363525, "learning_rate": 3.1256722090632906e-08, "loss": 1.2886, "step": 49226 }, { "epoch": 3.8579153605015675, "grad_norm": 3.5836493968963623, "learning_rate": 3.12223058340716e-08, "loss": 1.348, "step": 49227 }, { "epoch": 3.8579937304075234, "grad_norm": 2.2753567695617676, "learning_rate": 3.118790847618847e-08, "loss": 1.2871, "step": 49228 }, { "epoch": 3.8580721003134797, "grad_norm": 7.801136493682861, "learning_rate": 3.115353001711452e-08, "loss": 0.9812, "step": 49229 }, { "epoch": 3.8581504702194356, "grad_norm": 3.756657838821411, "learning_rate": 3.111917045697965e-08, "loss": 1.4874, "step": 49230 }, { "epoch": 3.858228840125392, "grad_norm": 2.6539206504821777, "learning_rate": 3.108482979591598e-08, "loss": 0.9651, "step": 49231 }, { "epoch": 3.858307210031348, "grad_norm": 7.12152624130249, "learning_rate": 3.1050508034053404e-08, "loss": 1.583, "step": 49232 }, { "epoch": 3.858385579937304, "grad_norm": 3.3187968730926514, "learning_rate": 3.101620517152237e-08, "loss": 1.4789, "step": 49233 }, { "epoch": 3.8584639498432605, "grad_norm": 3.2985520362854004, "learning_rate": 3.0981921208453336e-08, "loss": 1.1095, "step": 49234 }, { "epoch": 3.8585423197492164, "grad_norm": 2.2437310218811035, "learning_rate": 3.0947656144976746e-08, "loss": 1.1564, "step": 49235 }, { "epoch": 3.8586206896551722, "grad_norm": 3.0457122325897217, "learning_rate": 3.091340998122361e-08, "loss": 1.6291, "step": 49236 }, { "epoch": 3.8586990595611286, "grad_norm": 3.716644048690796, "learning_rate": 3.0879182717323265e-08, "loss": 1.5849, "step": 49237 }, { "epoch": 3.858777429467085, "grad_norm": 3.7917234897613525, "learning_rate": 3.084497435340617e-08, "loss": 1.0903, "step": 49238 }, { "epoch": 3.858855799373041, "grad_norm": 5.831081867218018, "learning_rate": 3.081078488960221e-08, "loss": 1.3012, "step": 49239 }, { "epoch": 3.8589341692789967, "grad_norm": 3.6535773277282715, "learning_rate": 3.077661432604184e-08, "loss": 1.2594, "step": 49240 }, { "epoch": 3.859012539184953, "grad_norm": 6.512907981872559, "learning_rate": 3.0742462662854963e-08, "loss": 1.3804, "step": 49241 }, { "epoch": 3.8590909090909093, "grad_norm": 3.5257203578948975, "learning_rate": 3.0708329900171475e-08, "loss": 1.154, "step": 49242 }, { "epoch": 3.8591692789968652, "grad_norm": 3.5694265365600586, "learning_rate": 3.0674216038121264e-08, "loss": 1.4081, "step": 49243 }, { "epoch": 3.859247648902821, "grad_norm": 3.8345322608947754, "learning_rate": 3.064012107683312e-08, "loss": 1.3492, "step": 49244 }, { "epoch": 3.8593260188087775, "grad_norm": 2.3748981952667236, "learning_rate": 3.0606045016438045e-08, "loss": 1.1669, "step": 49245 }, { "epoch": 3.859404388714734, "grad_norm": 2.0620243549346924, "learning_rate": 3.057198785706539e-08, "loss": 1.3878, "step": 49246 }, { "epoch": 3.8594827586206897, "grad_norm": 8.237032890319824, "learning_rate": 3.053794959884393e-08, "loss": 0.648, "step": 49247 }, { "epoch": 3.8595611285266456, "grad_norm": 4.707012176513672, "learning_rate": 3.0503930241904123e-08, "loss": 1.5539, "step": 49248 }, { "epoch": 3.859639498432602, "grad_norm": 3.4278411865234375, "learning_rate": 3.04699297863742e-08, "loss": 1.5131, "step": 49249 }, { "epoch": 3.8597178683385582, "grad_norm": 2.7964088916778564, "learning_rate": 3.0435948232384606e-08, "loss": 0.993, "step": 49250 }, { "epoch": 3.859796238244514, "grad_norm": 3.4588630199432373, "learning_rate": 3.040198558006469e-08, "loss": 0.9586, "step": 49251 }, { "epoch": 3.85987460815047, "grad_norm": 2.8875365257263184, "learning_rate": 3.0368041829542115e-08, "loss": 1.01, "step": 49252 }, { "epoch": 3.8599529780564263, "grad_norm": 2.4107749462127686, "learning_rate": 3.033411698094735e-08, "loss": 1.5507, "step": 49253 }, { "epoch": 3.8600313479623827, "grad_norm": 10.803464889526367, "learning_rate": 3.0300211034408614e-08, "loss": 2.0518, "step": 49254 }, { "epoch": 3.8601097178683386, "grad_norm": 5.4062724113464355, "learning_rate": 3.0266323990056355e-08, "loss": 1.4775, "step": 49255 }, { "epoch": 3.8601880877742945, "grad_norm": 4.586633205413818, "learning_rate": 3.023245584801771e-08, "loss": 1.4268, "step": 49256 }, { "epoch": 3.860266457680251, "grad_norm": 3.693357467651367, "learning_rate": 3.0198606608422e-08, "loss": 1.4104, "step": 49257 }, { "epoch": 3.860344827586207, "grad_norm": 2.832115411758423, "learning_rate": 3.016477627139802e-08, "loss": 1.2557, "step": 49258 }, { "epoch": 3.860423197492163, "grad_norm": 6.702158451080322, "learning_rate": 3.013096483707512e-08, "loss": 0.9237, "step": 49259 }, { "epoch": 3.860501567398119, "grad_norm": 2.936812400817871, "learning_rate": 3.0097172305580956e-08, "loss": 1.4037, "step": 49260 }, { "epoch": 3.8605799373040752, "grad_norm": 3.7327511310577393, "learning_rate": 3.006339867704433e-08, "loss": 1.4133, "step": 49261 }, { "epoch": 3.8606583072100316, "grad_norm": 3.870713949203491, "learning_rate": 3.002964395159458e-08, "loss": 1.4441, "step": 49262 }, { "epoch": 3.8607366771159874, "grad_norm": 2.378937244415283, "learning_rate": 2.9995908129358266e-08, "loss": 0.9413, "step": 49263 }, { "epoch": 3.8608150470219433, "grad_norm": 5.527695178985596, "learning_rate": 2.9962191210465286e-08, "loss": 1.2486, "step": 49264 }, { "epoch": 3.8608934169278997, "grad_norm": 2.9002633094787598, "learning_rate": 2.9928493195043315e-08, "loss": 1.1593, "step": 49265 }, { "epoch": 3.860971786833856, "grad_norm": 9.754621505737305, "learning_rate": 2.9894814083220594e-08, "loss": 1.1318, "step": 49266 }, { "epoch": 3.861050156739812, "grad_norm": 2.8667614459991455, "learning_rate": 2.9861153875125335e-08, "loss": 1.274, "step": 49267 }, { "epoch": 3.8611285266457678, "grad_norm": 3.770169734954834, "learning_rate": 2.982751257088523e-08, "loss": 1.5701, "step": 49268 }, { "epoch": 3.861206896551724, "grad_norm": 3.3553342819213867, "learning_rate": 2.9793890170628504e-08, "loss": 1.3623, "step": 49269 }, { "epoch": 3.8612852664576804, "grad_norm": 3.0035500526428223, "learning_rate": 2.976028667448283e-08, "loss": 1.0824, "step": 49270 }, { "epoch": 3.8613636363636363, "grad_norm": 7.291237831115723, "learning_rate": 2.972670208257644e-08, "loss": 1.4044, "step": 49271 }, { "epoch": 3.861442006269592, "grad_norm": 11.133071899414062, "learning_rate": 2.969313639503646e-08, "loss": 0.7888, "step": 49272 }, { "epoch": 3.8615203761755486, "grad_norm": 12.321385383605957, "learning_rate": 2.9659589611991112e-08, "loss": 2.244, "step": 49273 }, { "epoch": 3.861598746081505, "grad_norm": 3.22653865814209, "learning_rate": 2.9626061733567523e-08, "loss": 1.3388, "step": 49274 }, { "epoch": 3.8616771159874608, "grad_norm": 7.2766194343566895, "learning_rate": 2.9592552759893366e-08, "loss": 1.98, "step": 49275 }, { "epoch": 3.8617554858934167, "grad_norm": 2.432502031326294, "learning_rate": 2.9559062691096875e-08, "loss": 1.2502, "step": 49276 }, { "epoch": 3.861833855799373, "grad_norm": 4.630636215209961, "learning_rate": 2.9525591527304054e-08, "loss": 1.149, "step": 49277 }, { "epoch": 3.8619122257053293, "grad_norm": 13.579817771911621, "learning_rate": 2.9492139268643138e-08, "loss": 1.3217, "step": 49278 }, { "epoch": 3.861990595611285, "grad_norm": 3.3159916400909424, "learning_rate": 2.9458705915240693e-08, "loss": 1.1574, "step": 49279 }, { "epoch": 3.862068965517241, "grad_norm": 2.4043052196502686, "learning_rate": 2.942529146722495e-08, "loss": 1.3237, "step": 49280 }, { "epoch": 3.8621473354231974, "grad_norm": 4.849508762359619, "learning_rate": 2.9391895924722467e-08, "loss": 1.3555, "step": 49281 }, { "epoch": 3.8622257053291538, "grad_norm": 4.554138660430908, "learning_rate": 2.9358519287859267e-08, "loss": 1.645, "step": 49282 }, { "epoch": 3.8623040752351097, "grad_norm": 4.342160701751709, "learning_rate": 2.9325161556763572e-08, "loss": 1.2634, "step": 49283 }, { "epoch": 3.862382445141066, "grad_norm": 2.941805601119995, "learning_rate": 2.9291822731561946e-08, "loss": 1.5832, "step": 49284 }, { "epoch": 3.862460815047022, "grad_norm": 3.1051011085510254, "learning_rate": 2.925850281238152e-08, "loss": 1.322, "step": 49285 }, { "epoch": 3.862539184952978, "grad_norm": 2.7112340927124023, "learning_rate": 2.9225201799347735e-08, "loss": 1.1052, "step": 49286 }, { "epoch": 3.862617554858934, "grad_norm": 2.6004421710968018, "learning_rate": 2.9191919692588832e-08, "loss": 1.1324, "step": 49287 }, { "epoch": 3.8626959247648904, "grad_norm": 3.3572165966033936, "learning_rate": 2.9158656492230264e-08, "loss": 1.3038, "step": 49288 }, { "epoch": 3.8627742946708463, "grad_norm": 1.942043423652649, "learning_rate": 2.912541219839915e-08, "loss": 1.3902, "step": 49289 }, { "epoch": 3.8628526645768027, "grad_norm": 3.28762149810791, "learning_rate": 2.9092186811222055e-08, "loss": 1.298, "step": 49290 }, { "epoch": 3.8629310344827585, "grad_norm": 4.202219009399414, "learning_rate": 2.9058980330824437e-08, "loss": 1.9595, "step": 49291 }, { "epoch": 3.863009404388715, "grad_norm": 3.191577434539795, "learning_rate": 2.902579275733397e-08, "loss": 1.5751, "step": 49292 }, { "epoch": 3.8630877742946708, "grad_norm": 2.4692628383636475, "learning_rate": 2.8992624090875554e-08, "loss": 1.3641, "step": 49293 }, { "epoch": 3.863166144200627, "grad_norm": 5.7553791999816895, "learning_rate": 2.895947433157631e-08, "loss": 0.8973, "step": 49294 }, { "epoch": 3.863244514106583, "grad_norm": 3.6444203853607178, "learning_rate": 2.8926343479561693e-08, "loss": 1.6443, "step": 49295 }, { "epoch": 3.8633228840125393, "grad_norm": 3.3110475540161133, "learning_rate": 2.889323153495771e-08, "loss": 1.1058, "step": 49296 }, { "epoch": 3.863401253918495, "grad_norm": 3.0362889766693115, "learning_rate": 2.8860138497890932e-08, "loss": 1.1985, "step": 49297 }, { "epoch": 3.8634796238244515, "grad_norm": 3.0311455726623535, "learning_rate": 2.8827064368486812e-08, "loss": 1.2584, "step": 49298 }, { "epoch": 3.8635579937304074, "grad_norm": 7.238388538360596, "learning_rate": 2.879400914687136e-08, "loss": 1.0738, "step": 49299 }, { "epoch": 3.8636363636363638, "grad_norm": 3.20930552482605, "learning_rate": 2.8760972833170032e-08, "loss": 1.236, "step": 49300 }, { "epoch": 3.8637147335423196, "grad_norm": 2.746138095855713, "learning_rate": 2.872795542750828e-08, "loss": 1.3819, "step": 49301 }, { "epoch": 3.863793103448276, "grad_norm": 10.60187816619873, "learning_rate": 2.869495693001212e-08, "loss": 1.5429, "step": 49302 }, { "epoch": 3.863871473354232, "grad_norm": 2.185521364212036, "learning_rate": 2.8661977340807e-08, "loss": 1.4112, "step": 49303 }, { "epoch": 3.863949843260188, "grad_norm": 3.199728488922119, "learning_rate": 2.8629016660018382e-08, "loss": 1.5851, "step": 49304 }, { "epoch": 3.864028213166144, "grad_norm": 3.975480556488037, "learning_rate": 2.8596074887771163e-08, "loss": 1.3972, "step": 49305 }, { "epoch": 3.8641065830721004, "grad_norm": 4.979135513305664, "learning_rate": 2.85631520241908e-08, "loss": 1.204, "step": 49306 }, { "epoch": 3.8641849529780563, "grad_norm": 2.187859058380127, "learning_rate": 2.85302480694033e-08, "loss": 0.8511, "step": 49307 }, { "epoch": 3.8642633228840126, "grad_norm": 4.074824333190918, "learning_rate": 2.8497363023533007e-08, "loss": 1.3618, "step": 49308 }, { "epoch": 3.8643416927899685, "grad_norm": 4.707014083862305, "learning_rate": 2.8464496886705385e-08, "loss": 1.731, "step": 49309 }, { "epoch": 3.864420062695925, "grad_norm": 4.021942138671875, "learning_rate": 2.8431649659045324e-08, "loss": 1.2555, "step": 49310 }, { "epoch": 3.8644984326018808, "grad_norm": 5.338536739349365, "learning_rate": 2.8398821340677175e-08, "loss": 1.5975, "step": 49311 }, { "epoch": 3.864576802507837, "grad_norm": 2.7457211017608643, "learning_rate": 2.8366011931726945e-08, "loss": 1.4395, "step": 49312 }, { "epoch": 3.864655172413793, "grad_norm": 8.688934326171875, "learning_rate": 2.8333221432318424e-08, "loss": 0.9948, "step": 49313 }, { "epoch": 3.8647335423197493, "grad_norm": 2.8617093563079834, "learning_rate": 2.830044984257707e-08, "loss": 1.4153, "step": 49314 }, { "epoch": 3.864811912225705, "grad_norm": 3.4153892993927, "learning_rate": 2.826769716262667e-08, "loss": 0.8944, "step": 49315 }, { "epoch": 3.8648902821316615, "grad_norm": 6.114508628845215, "learning_rate": 2.8234963392592686e-08, "loss": 1.6774, "step": 49316 }, { "epoch": 3.8649686520376174, "grad_norm": 3.8469488620758057, "learning_rate": 2.820224853259945e-08, "loss": 1.3403, "step": 49317 }, { "epoch": 3.8650470219435737, "grad_norm": 4.611028671264648, "learning_rate": 2.8169552582770764e-08, "loss": 1.7471, "step": 49318 }, { "epoch": 3.8651253918495296, "grad_norm": 2.531186580657959, "learning_rate": 2.8136875543231523e-08, "loss": 1.5045, "step": 49319 }, { "epoch": 3.865203761755486, "grad_norm": 2.8775405883789062, "learning_rate": 2.810421741410552e-08, "loss": 1.2808, "step": 49320 }, { "epoch": 3.865282131661442, "grad_norm": 2.3807003498077393, "learning_rate": 2.8071578195518202e-08, "loss": 1.4444, "step": 49321 }, { "epoch": 3.865360501567398, "grad_norm": 5.488382816314697, "learning_rate": 2.8038957887592256e-08, "loss": 1.2432, "step": 49322 }, { "epoch": 3.865438871473354, "grad_norm": 2.8482820987701416, "learning_rate": 2.800635649045258e-08, "loss": 1.2473, "step": 49323 }, { "epoch": 3.8655172413793104, "grad_norm": 3.7077317237854004, "learning_rate": 2.797377400422241e-08, "loss": 1.5996, "step": 49324 }, { "epoch": 3.8655956112852663, "grad_norm": 3.8156206607818604, "learning_rate": 2.79412104290272e-08, "loss": 1.1819, "step": 49325 }, { "epoch": 3.8656739811912226, "grad_norm": 3.0132856369018555, "learning_rate": 2.7908665764989072e-08, "loss": 1.4099, "step": 49326 }, { "epoch": 3.8657523510971785, "grad_norm": 9.73332691192627, "learning_rate": 2.7876140012232932e-08, "loss": 1.5424, "step": 49327 }, { "epoch": 3.865830721003135, "grad_norm": 2.126619338989258, "learning_rate": 2.784363317088201e-08, "loss": 1.2723, "step": 49328 }, { "epoch": 3.865909090909091, "grad_norm": 3.206697940826416, "learning_rate": 2.7811145241059545e-08, "loss": 1.1423, "step": 49329 }, { "epoch": 3.865987460815047, "grad_norm": 4.744887351989746, "learning_rate": 2.7778676222890433e-08, "loss": 1.5054, "step": 49330 }, { "epoch": 3.866065830721003, "grad_norm": 6.5695061683654785, "learning_rate": 2.77462261164968e-08, "loss": 1.3793, "step": 49331 }, { "epoch": 3.8661442006269593, "grad_norm": 3.5130815505981445, "learning_rate": 2.7713794922002434e-08, "loss": 1.3641, "step": 49332 }, { "epoch": 3.8662225705329156, "grad_norm": 9.071094512939453, "learning_rate": 2.7681382639531128e-08, "loss": 0.9264, "step": 49333 }, { "epoch": 3.8663009404388715, "grad_norm": 3.4053170680999756, "learning_rate": 2.7648989269206117e-08, "loss": 1.4272, "step": 49334 }, { "epoch": 3.8663793103448274, "grad_norm": 2.7113664150238037, "learning_rate": 2.761661481115008e-08, "loss": 1.2356, "step": 49335 }, { "epoch": 3.8664576802507837, "grad_norm": 4.386306285858154, "learning_rate": 2.75842592654868e-08, "loss": 1.3157, "step": 49336 }, { "epoch": 3.86653605015674, "grad_norm": 5.154914855957031, "learning_rate": 2.7551922632338966e-08, "loss": 1.4846, "step": 49337 }, { "epoch": 3.866614420062696, "grad_norm": 2.350482940673828, "learning_rate": 2.7519604911829258e-08, "loss": 1.5925, "step": 49338 }, { "epoch": 3.866692789968652, "grad_norm": 2.732506513595581, "learning_rate": 2.7487306104080902e-08, "loss": 1.3287, "step": 49339 }, { "epoch": 3.866771159874608, "grad_norm": 4.136407375335693, "learning_rate": 2.7455026209217696e-08, "loss": 1.313, "step": 49340 }, { "epoch": 3.8668495297805645, "grad_norm": 4.188453197479248, "learning_rate": 2.7422765227360648e-08, "loss": 0.967, "step": 49341 }, { "epoch": 3.8669278996865204, "grad_norm": 2.471121072769165, "learning_rate": 2.7390523158633552e-08, "loss": 1.1575, "step": 49342 }, { "epoch": 3.8670062695924763, "grad_norm": 6.7914958000183105, "learning_rate": 2.7358300003158532e-08, "loss": 1.1191, "step": 49343 }, { "epoch": 3.8670846394984326, "grad_norm": 2.860145092010498, "learning_rate": 2.7326095761058825e-08, "loss": 1.5385, "step": 49344 }, { "epoch": 3.867163009404389, "grad_norm": 3.7478768825531006, "learning_rate": 2.729391043245655e-08, "loss": 1.3764, "step": 49345 }, { "epoch": 3.867241379310345, "grad_norm": 8.65163803100586, "learning_rate": 2.7261744017473834e-08, "loss": 1.5759, "step": 49346 }, { "epoch": 3.8673197492163007, "grad_norm": 3.909053087234497, "learning_rate": 2.722959651623336e-08, "loss": 1.1397, "step": 49347 }, { "epoch": 3.867398119122257, "grad_norm": 2.2762014865875244, "learning_rate": 2.7197467928857245e-08, "loss": 1.3794, "step": 49348 }, { "epoch": 3.8674764890282134, "grad_norm": 3.9933247566223145, "learning_rate": 2.7165358255467622e-08, "loss": 1.2392, "step": 49349 }, { "epoch": 3.8675548589341693, "grad_norm": 4.6954731941223145, "learning_rate": 2.7133267496187167e-08, "loss": 0.9337, "step": 49350 }, { "epoch": 3.867633228840125, "grad_norm": 5.350114345550537, "learning_rate": 2.7101195651137445e-08, "loss": 1.5717, "step": 49351 }, { "epoch": 3.8677115987460815, "grad_norm": 2.7961854934692383, "learning_rate": 2.706914272044059e-08, "loss": 1.1323, "step": 49352 }, { "epoch": 3.867789968652038, "grad_norm": 2.3397841453552246, "learning_rate": 2.7037108704218164e-08, "loss": 1.2616, "step": 49353 }, { "epoch": 3.8678683385579937, "grad_norm": 4.90103006362915, "learning_rate": 2.700509360259229e-08, "loss": 1.4761, "step": 49354 }, { "epoch": 3.8679467084639496, "grad_norm": 2.7197375297546387, "learning_rate": 2.6973097415685102e-08, "loss": 1.3866, "step": 49355 }, { "epoch": 3.868025078369906, "grad_norm": 8.385690689086914, "learning_rate": 2.6941120143617604e-08, "loss": 1.0097, "step": 49356 }, { "epoch": 3.8681034482758623, "grad_norm": 5.497931957244873, "learning_rate": 2.6909161786511928e-08, "loss": 1.3398, "step": 49357 }, { "epoch": 3.868181818181818, "grad_norm": 3.265990972518921, "learning_rate": 2.687722234448964e-08, "loss": 1.3509, "step": 49358 }, { "epoch": 3.868260188087774, "grad_norm": 2.4613842964172363, "learning_rate": 2.6845301817671754e-08, "loss": 1.2005, "step": 49359 }, { "epoch": 3.8683385579937304, "grad_norm": 3.1498281955718994, "learning_rate": 2.6813400206180394e-08, "loss": 1.4932, "step": 49360 }, { "epoch": 3.8684169278996867, "grad_norm": 2.790435314178467, "learning_rate": 2.6781517510136024e-08, "loss": 1.0756, "step": 49361 }, { "epoch": 3.8684952978056426, "grad_norm": 5.863882541656494, "learning_rate": 2.6749653729660207e-08, "loss": 1.6722, "step": 49362 }, { "epoch": 3.8685736677115985, "grad_norm": 4.305129528045654, "learning_rate": 2.6717808864874518e-08, "loss": 1.7932, "step": 49363 }, { "epoch": 3.868652037617555, "grad_norm": 7.504130840301514, "learning_rate": 2.6685982915899968e-08, "loss": 1.3661, "step": 49364 }, { "epoch": 3.868730407523511, "grad_norm": 9.643192291259766, "learning_rate": 2.665417588285757e-08, "loss": 1.3207, "step": 49365 }, { "epoch": 3.868808777429467, "grad_norm": 3.619370460510254, "learning_rate": 2.6622387765867785e-08, "loss": 1.9478, "step": 49366 }, { "epoch": 3.868887147335423, "grad_norm": 4.408628940582275, "learning_rate": 2.6590618565052185e-08, "loss": 1.408, "step": 49367 }, { "epoch": 3.8689655172413793, "grad_norm": 3.571549892425537, "learning_rate": 2.6558868280531225e-08, "loss": 1.049, "step": 49368 }, { "epoch": 3.8690438871473356, "grad_norm": 8.371072769165039, "learning_rate": 2.6527136912425923e-08, "loss": 2.3934, "step": 49369 }, { "epoch": 3.8691222570532915, "grad_norm": 3.3371870517730713, "learning_rate": 2.649542446085729e-08, "loss": 1.185, "step": 49370 }, { "epoch": 3.8692006269592474, "grad_norm": 2.9206764698028564, "learning_rate": 2.6463730925944675e-08, "loss": 1.4625, "step": 49371 }, { "epoch": 3.8692789968652037, "grad_norm": 6.583513259887695, "learning_rate": 2.6432056307810204e-08, "loss": 1.6383, "step": 49372 }, { "epoch": 3.86935736677116, "grad_norm": 4.795126914978027, "learning_rate": 2.6400400606573227e-08, "loss": 1.3775, "step": 49373 }, { "epoch": 3.869435736677116, "grad_norm": 5.850678443908691, "learning_rate": 2.6368763822354758e-08, "loss": 1.762, "step": 49374 }, { "epoch": 3.8695141065830723, "grad_norm": 10.084246635437012, "learning_rate": 2.63371459552747e-08, "loss": 1.4608, "step": 49375 }, { "epoch": 3.869592476489028, "grad_norm": 3.252882719039917, "learning_rate": 2.6305547005453514e-08, "loss": 0.9709, "step": 49376 }, { "epoch": 3.8696708463949845, "grad_norm": 2.3303701877593994, "learning_rate": 2.62739669730111e-08, "loss": 1.1628, "step": 49377 }, { "epoch": 3.8697492163009404, "grad_norm": 3.415569543838501, "learning_rate": 2.6242405858068477e-08, "loss": 1.4161, "step": 49378 }, { "epoch": 3.8698275862068967, "grad_norm": 5.669974327087402, "learning_rate": 2.621086366074499e-08, "loss": 1.5734, "step": 49379 }, { "epoch": 3.8699059561128526, "grad_norm": 3.0142874717712402, "learning_rate": 2.617934038115999e-08, "loss": 1.7578, "step": 49380 }, { "epoch": 3.869984326018809, "grad_norm": 4.0879645347595215, "learning_rate": 2.614783601943449e-08, "loss": 1.366, "step": 49381 }, { "epoch": 3.870062695924765, "grad_norm": 2.6796863079071045, "learning_rate": 2.6116350575687845e-08, "loss": 1.1626, "step": 49382 }, { "epoch": 3.870141065830721, "grad_norm": 4.100958347320557, "learning_rate": 2.6084884050039948e-08, "loss": 1.7302, "step": 49383 }, { "epoch": 3.870219435736677, "grad_norm": 7.262201309204102, "learning_rate": 2.6053436442610712e-08, "loss": 1.1361, "step": 49384 }, { "epoch": 3.8702978056426334, "grad_norm": 3.7198386192321777, "learning_rate": 2.602200775351893e-08, "loss": 1.0654, "step": 49385 }, { "epoch": 3.8703761755485893, "grad_norm": 3.468168258666992, "learning_rate": 2.5990597982884503e-08, "loss": 1.2904, "step": 49386 }, { "epoch": 3.8704545454545456, "grad_norm": 3.689704656600952, "learning_rate": 2.5959207130827337e-08, "loss": 1.4927, "step": 49387 }, { "epoch": 3.8705329153605015, "grad_norm": 3.4290904998779297, "learning_rate": 2.5927835197466223e-08, "loss": 1.5245, "step": 49388 }, { "epoch": 3.870611285266458, "grad_norm": 8.638134002685547, "learning_rate": 2.5896482182921623e-08, "loss": 1.7284, "step": 49389 }, { "epoch": 3.8706896551724137, "grad_norm": 4.724574565887451, "learning_rate": 2.586514808731122e-08, "loss": 1.5789, "step": 49390 }, { "epoch": 3.87076802507837, "grad_norm": 2.031672716140747, "learning_rate": 2.583383291075492e-08, "loss": 1.3498, "step": 49391 }, { "epoch": 3.870846394984326, "grad_norm": 6.073036193847656, "learning_rate": 2.5802536653371514e-08, "loss": 1.3961, "step": 49392 }, { "epoch": 3.8709247648902823, "grad_norm": 2.8616554737091064, "learning_rate": 2.577125931528146e-08, "loss": 1.3799, "step": 49393 }, { "epoch": 3.871003134796238, "grad_norm": 2.259277105331421, "learning_rate": 2.5740000896601892e-08, "loss": 1.3205, "step": 49394 }, { "epoch": 3.8710815047021945, "grad_norm": 2.6390204429626465, "learning_rate": 2.5708761397452154e-08, "loss": 1.3801, "step": 49395 }, { "epoch": 3.8711598746081504, "grad_norm": 4.206348896026611, "learning_rate": 2.5677540817951595e-08, "loss": 1.6775, "step": 49396 }, { "epoch": 3.8712382445141067, "grad_norm": 3.1335954666137695, "learning_rate": 2.5646339158219014e-08, "loss": 1.3453, "step": 49397 }, { "epoch": 3.8713166144200626, "grad_norm": 5.064464092254639, "learning_rate": 2.561515641837209e-08, "loss": 1.5888, "step": 49398 }, { "epoch": 3.871394984326019, "grad_norm": 3.8671391010284424, "learning_rate": 2.5583992598530726e-08, "loss": 1.4629, "step": 49399 }, { "epoch": 3.871473354231975, "grad_norm": 8.175251007080078, "learning_rate": 2.5552847698812612e-08, "loss": 1.09, "step": 49400 }, { "epoch": 3.871551724137931, "grad_norm": 3.2273154258728027, "learning_rate": 2.5521721719336535e-08, "loss": 1.3132, "step": 49401 }, { "epoch": 3.871630094043887, "grad_norm": 2.4776787757873535, "learning_rate": 2.5490614660220736e-08, "loss": 1.1145, "step": 49402 }, { "epoch": 3.8717084639498434, "grad_norm": 4.044760227203369, "learning_rate": 2.5459526521583456e-08, "loss": 1.6097, "step": 49403 }, { "epoch": 3.8717868338557992, "grad_norm": 4.297869682312012, "learning_rate": 2.542845730354293e-08, "loss": 0.8446, "step": 49404 }, { "epoch": 3.8718652037617556, "grad_norm": 3.6717395782470703, "learning_rate": 2.53974070062174e-08, "loss": 1.189, "step": 49405 }, { "epoch": 3.8719435736677115, "grad_norm": 2.3037540912628174, "learning_rate": 2.5366375629725103e-08, "loss": 1.2182, "step": 49406 }, { "epoch": 3.872021943573668, "grad_norm": 2.622509479522705, "learning_rate": 2.533536317418428e-08, "loss": 1.1976, "step": 49407 }, { "epoch": 3.8721003134796237, "grad_norm": 2.555288791656494, "learning_rate": 2.5304369639712612e-08, "loss": 1.5998, "step": 49408 }, { "epoch": 3.87217868338558, "grad_norm": 3.39656400680542, "learning_rate": 2.5273395026427783e-08, "loss": 1.672, "step": 49409 }, { "epoch": 3.872257053291536, "grad_norm": 3.332942485809326, "learning_rate": 2.5242439334447478e-08, "loss": 0.9781, "step": 49410 }, { "epoch": 3.8723354231974922, "grad_norm": 4.194820880889893, "learning_rate": 2.5211502563890487e-08, "loss": 1.7375, "step": 49411 }, { "epoch": 3.872413793103448, "grad_norm": 4.878559112548828, "learning_rate": 2.5180584714872837e-08, "loss": 1.6416, "step": 49412 }, { "epoch": 3.8724921630094045, "grad_norm": 3.382617235183716, "learning_rate": 2.514968578751331e-08, "loss": 1.8354, "step": 49413 }, { "epoch": 3.8725705329153604, "grad_norm": 3.477956771850586, "learning_rate": 2.51188057819296e-08, "loss": 1.4591, "step": 49414 }, { "epoch": 3.8726489028213167, "grad_norm": 5.327117443084717, "learning_rate": 2.5087944698238277e-08, "loss": 1.0212, "step": 49415 }, { "epoch": 3.8727272727272726, "grad_norm": 2.7249562740325928, "learning_rate": 2.505710253655702e-08, "loss": 1.3257, "step": 49416 }, { "epoch": 3.872805642633229, "grad_norm": 9.375480651855469, "learning_rate": 2.5026279297003518e-08, "loss": 1.3243, "step": 49417 }, { "epoch": 3.872884012539185, "grad_norm": 3.6190364360809326, "learning_rate": 2.49954749796949e-08, "loss": 1.3709, "step": 49418 }, { "epoch": 3.872962382445141, "grad_norm": 4.453207969665527, "learning_rate": 2.4964689584747735e-08, "loss": 1.8416, "step": 49419 }, { "epoch": 3.873040752351097, "grad_norm": 3.9252305030822754, "learning_rate": 2.4933923112279712e-08, "loss": 1.4065, "step": 49420 }, { "epoch": 3.8731191222570533, "grad_norm": 5.671501636505127, "learning_rate": 2.4903175562407954e-08, "loss": 0.9288, "step": 49421 }, { "epoch": 3.8731974921630092, "grad_norm": 3.3475146293640137, "learning_rate": 2.487244693524904e-08, "loss": 1.2307, "step": 49422 }, { "epoch": 3.8732758620689656, "grad_norm": 2.379279851913452, "learning_rate": 2.484173723091954e-08, "loss": 0.8117, "step": 49423 }, { "epoch": 3.873354231974922, "grad_norm": 5.973259925842285, "learning_rate": 2.4811046449537135e-08, "loss": 1.5437, "step": 49424 }, { "epoch": 3.873432601880878, "grad_norm": 4.410353660583496, "learning_rate": 2.4780374591217848e-08, "loss": 1.2605, "step": 49425 }, { "epoch": 3.8735109717868337, "grad_norm": 6.24009895324707, "learning_rate": 2.4749721656078806e-08, "loss": 0.8636, "step": 49426 }, { "epoch": 3.87358934169279, "grad_norm": 5.471020221710205, "learning_rate": 2.4719087644236027e-08, "loss": 1.609, "step": 49427 }, { "epoch": 3.8736677115987463, "grad_norm": 4.172243118286133, "learning_rate": 2.4688472555806642e-08, "loss": 1.2661, "step": 49428 }, { "epoch": 3.8737460815047022, "grad_norm": 4.708568096160889, "learning_rate": 2.4657876390906666e-08, "loss": 1.1927, "step": 49429 }, { "epoch": 3.873824451410658, "grad_norm": 4.4977617263793945, "learning_rate": 2.462729914965267e-08, "loss": 1.3644, "step": 49430 }, { "epoch": 3.8739028213166145, "grad_norm": 4.211591720581055, "learning_rate": 2.4596740832161237e-08, "loss": 1.3021, "step": 49431 }, { "epoch": 3.873981191222571, "grad_norm": 2.2216427326202393, "learning_rate": 2.4566201438547822e-08, "loss": 1.3799, "step": 49432 }, { "epoch": 3.8740595611285267, "grad_norm": 5.700713634490967, "learning_rate": 2.4535680968928998e-08, "loss": 1.8632, "step": 49433 }, { "epoch": 3.8741379310344826, "grad_norm": 2.277311325073242, "learning_rate": 2.4505179423421342e-08, "loss": 1.2249, "step": 49434 }, { "epoch": 3.874216300940439, "grad_norm": 3.305490493774414, "learning_rate": 2.4474696802140318e-08, "loss": 1.2915, "step": 49435 }, { "epoch": 3.8742946708463952, "grad_norm": 3.3314826488494873, "learning_rate": 2.4444233105201388e-08, "loss": 1.6305, "step": 49436 }, { "epoch": 3.874373040752351, "grad_norm": 2.3810205459594727, "learning_rate": 2.4413788332721677e-08, "loss": 1.1922, "step": 49437 }, { "epoch": 3.874451410658307, "grad_norm": 3.770092010498047, "learning_rate": 2.4383362484816096e-08, "loss": 1.3445, "step": 49438 }, { "epoch": 3.8745297805642633, "grad_norm": 2.554192066192627, "learning_rate": 2.435295556160011e-08, "loss": 1.1709, "step": 49439 }, { "epoch": 3.8746081504702197, "grad_norm": 2.7418034076690674, "learning_rate": 2.432256756319029e-08, "loss": 0.8692, "step": 49440 }, { "epoch": 3.8746865203761756, "grad_norm": 5.299559116363525, "learning_rate": 2.4292198489701547e-08, "loss": 1.6036, "step": 49441 }, { "epoch": 3.8747648902821314, "grad_norm": 2.958887815475464, "learning_rate": 2.4261848341249894e-08, "loss": 1.3617, "step": 49442 }, { "epoch": 3.874843260188088, "grad_norm": 2.8759407997131348, "learning_rate": 2.4231517117950244e-08, "loss": 1.2028, "step": 49443 }, { "epoch": 3.874921630094044, "grad_norm": 4.491169452667236, "learning_rate": 2.4201204819918054e-08, "loss": 1.6248, "step": 49444 }, { "epoch": 3.875, "grad_norm": 4.85205602645874, "learning_rate": 2.4170911447269352e-08, "loss": 0.8678, "step": 49445 }, { "epoch": 3.875078369905956, "grad_norm": 2.390779495239258, "learning_rate": 2.4140637000117927e-08, "loss": 1.6889, "step": 49446 }, { "epoch": 3.875156739811912, "grad_norm": 2.7559733390808105, "learning_rate": 2.411038147858036e-08, "loss": 1.4424, "step": 49447 }, { "epoch": 3.8752351097178686, "grad_norm": 2.9887912273406982, "learning_rate": 2.408014488277044e-08, "loss": 1.0245, "step": 49448 }, { "epoch": 3.8753134796238244, "grad_norm": 2.538717031478882, "learning_rate": 2.4049927212804746e-08, "loss": 1.2358, "step": 49449 }, { "epoch": 3.8753918495297803, "grad_norm": 11.51097297668457, "learning_rate": 2.4019728468797077e-08, "loss": 1.3467, "step": 49450 }, { "epoch": 3.8754702194357367, "grad_norm": 3.7644622325897217, "learning_rate": 2.398954865086178e-08, "loss": 1.2265, "step": 49451 }, { "epoch": 3.875548589341693, "grad_norm": 3.1669564247131348, "learning_rate": 2.3959387759115438e-08, "loss": 1.067, "step": 49452 }, { "epoch": 3.875626959247649, "grad_norm": 4.49462890625, "learning_rate": 2.3929245793671285e-08, "loss": 1.1098, "step": 49453 }, { "epoch": 3.8757053291536048, "grad_norm": 5.41409158706665, "learning_rate": 2.3899122754644232e-08, "loss": 1.4649, "step": 49454 }, { "epoch": 3.875783699059561, "grad_norm": 10.169020652770996, "learning_rate": 2.386901864214919e-08, "loss": 1.712, "step": 49455 }, { "epoch": 3.8758620689655174, "grad_norm": 2.5415215492248535, "learning_rate": 2.3838933456300504e-08, "loss": 1.5109, "step": 49456 }, { "epoch": 3.8759404388714733, "grad_norm": 4.812777519226074, "learning_rate": 2.3808867197212536e-08, "loss": 1.2411, "step": 49457 }, { "epoch": 3.876018808777429, "grad_norm": 2.72741961479187, "learning_rate": 2.3778819864999637e-08, "loss": 1.4951, "step": 49458 }, { "epoch": 3.8760971786833855, "grad_norm": 5.094288349151611, "learning_rate": 2.374879145977671e-08, "loss": 1.4044, "step": 49459 }, { "epoch": 3.876175548589342, "grad_norm": 2.900019645690918, "learning_rate": 2.3718781981657e-08, "loss": 1.0684, "step": 49460 }, { "epoch": 3.8762539184952978, "grad_norm": 5.979628562927246, "learning_rate": 2.3688791430754864e-08, "loss": 0.5348, "step": 49461 }, { "epoch": 3.8763322884012537, "grad_norm": 4.060920238494873, "learning_rate": 2.3658819807184652e-08, "loss": 1.4876, "step": 49462 }, { "epoch": 3.87641065830721, "grad_norm": 2.5595672130584717, "learning_rate": 2.3628867111060717e-08, "loss": 1.4967, "step": 49463 }, { "epoch": 3.8764890282131663, "grad_norm": 3.9263949394226074, "learning_rate": 2.35989333424963e-08, "loss": 1.2877, "step": 49464 }, { "epoch": 3.876567398119122, "grad_norm": 2.662585735321045, "learning_rate": 2.3569018501605202e-08, "loss": 1.3187, "step": 49465 }, { "epoch": 3.876645768025078, "grad_norm": 3.1857895851135254, "learning_rate": 2.3539122588501772e-08, "loss": 1.2135, "step": 49466 }, { "epoch": 3.8767241379310344, "grad_norm": 2.592360019683838, "learning_rate": 2.3509245603299257e-08, "loss": 1.1405, "step": 49467 }, { "epoch": 3.8768025078369908, "grad_norm": 3.1662416458129883, "learning_rate": 2.3479387546112008e-08, "loss": 1.2579, "step": 49468 }, { "epoch": 3.8768808777429467, "grad_norm": 2.32601261138916, "learning_rate": 2.344954841705216e-08, "loss": 1.2271, "step": 49469 }, { "epoch": 3.876959247648903, "grad_norm": 6.546680927276611, "learning_rate": 2.3419728216234617e-08, "loss": 0.9329, "step": 49470 }, { "epoch": 3.877037617554859, "grad_norm": 5.58787727355957, "learning_rate": 2.3389926943772622e-08, "loss": 0.8601, "step": 49471 }, { "epoch": 3.877115987460815, "grad_norm": 3.340097665786743, "learning_rate": 2.3360144599778867e-08, "loss": 1.3042, "step": 49472 }, { "epoch": 3.877194357366771, "grad_norm": 10.048552513122559, "learning_rate": 2.3330381184367147e-08, "loss": 0.9087, "step": 49473 }, { "epoch": 3.8772727272727274, "grad_norm": 2.8808484077453613, "learning_rate": 2.3300636697650148e-08, "loss": 1.5243, "step": 49474 }, { "epoch": 3.8773510971786833, "grad_norm": 3.501467227935791, "learning_rate": 2.3270911139741115e-08, "loss": 1.3626, "step": 49475 }, { "epoch": 3.8774294670846396, "grad_norm": 3.8108980655670166, "learning_rate": 2.3241204510753844e-08, "loss": 1.1455, "step": 49476 }, { "epoch": 3.8775078369905955, "grad_norm": 4.060663223266602, "learning_rate": 2.3211516810801028e-08, "loss": 0.8863, "step": 49477 }, { "epoch": 3.877586206896552, "grad_norm": 6.013185501098633, "learning_rate": 2.318184803999479e-08, "loss": 1.2448, "step": 49478 }, { "epoch": 3.8776645768025078, "grad_norm": 4.296538352966309, "learning_rate": 2.315219819844894e-08, "loss": 1.3704, "step": 49479 }, { "epoch": 3.877742946708464, "grad_norm": 3.321605920791626, "learning_rate": 2.31225672862756e-08, "loss": 1.1702, "step": 49480 }, { "epoch": 3.87782131661442, "grad_norm": 2.7323005199432373, "learning_rate": 2.309295530358746e-08, "loss": 1.6857, "step": 49481 }, { "epoch": 3.8778996865203763, "grad_norm": 2.9227283000946045, "learning_rate": 2.3063362250498323e-08, "loss": 1.1218, "step": 49482 }, { "epoch": 3.877978056426332, "grad_norm": 13.870065689086914, "learning_rate": 2.3033788127119206e-08, "loss": 1.7165, "step": 49483 }, { "epoch": 3.8780564263322885, "grad_norm": 11.758014678955078, "learning_rate": 2.30042329335628e-08, "loss": 1.2182, "step": 49484 }, { "epoch": 3.8781347962382444, "grad_norm": 6.374157905578613, "learning_rate": 2.297469666994234e-08, "loss": 0.9954, "step": 49485 }, { "epoch": 3.8782131661442008, "grad_norm": 4.19757080078125, "learning_rate": 2.2945179336370527e-08, "loss": 1.4757, "step": 49486 }, { "epoch": 3.8782915360501566, "grad_norm": 7.383187294006348, "learning_rate": 2.2915680932957818e-08, "loss": 1.0533, "step": 49487 }, { "epoch": 3.878369905956113, "grad_norm": 2.3833484649658203, "learning_rate": 2.2886201459818015e-08, "loss": 1.6625, "step": 49488 }, { "epoch": 3.878448275862069, "grad_norm": 3.2725305557250977, "learning_rate": 2.2856740917062138e-08, "loss": 1.4763, "step": 49489 }, { "epoch": 3.878526645768025, "grad_norm": 7.189482688903809, "learning_rate": 2.282729930480343e-08, "loss": 1.9728, "step": 49490 }, { "epoch": 3.878605015673981, "grad_norm": 2.242544651031494, "learning_rate": 2.2797876623152915e-08, "loss": 1.0484, "step": 49491 }, { "epoch": 3.8786833855799374, "grad_norm": 2.9633259773254395, "learning_rate": 2.276847287222328e-08, "loss": 1.5596, "step": 49492 }, { "epoch": 3.8787617554858933, "grad_norm": 3.129085063934326, "learning_rate": 2.2739088052125545e-08, "loss": 1.0464, "step": 49493 }, { "epoch": 3.8788401253918496, "grad_norm": 3.239993095397949, "learning_rate": 2.270972216297185e-08, "loss": 1.16, "step": 49494 }, { "epoch": 3.8789184952978055, "grad_norm": 1.9803227186203003, "learning_rate": 2.268037520487376e-08, "loss": 1.3249, "step": 49495 }, { "epoch": 3.878996865203762, "grad_norm": 4.256617069244385, "learning_rate": 2.2651047177942866e-08, "loss": 1.7799, "step": 49496 }, { "epoch": 3.8790752351097177, "grad_norm": 5.1133503913879395, "learning_rate": 2.262173808229129e-08, "loss": 1.334, "step": 49497 }, { "epoch": 3.879153605015674, "grad_norm": 7.29169225692749, "learning_rate": 2.259244791803006e-08, "loss": 1.642, "step": 49498 }, { "epoch": 3.87923197492163, "grad_norm": 3.034081220626831, "learning_rate": 2.2563176685270193e-08, "loss": 1.4441, "step": 49499 }, { "epoch": 3.8793103448275863, "grad_norm": 2.009157180786133, "learning_rate": 2.2533924384123827e-08, "loss": 1.1423, "step": 49500 }, { "epoch": 3.879388714733542, "grad_norm": 2.860233783721924, "learning_rate": 2.2504691014701983e-08, "loss": 1.3423, "step": 49501 }, { "epoch": 3.8794670846394985, "grad_norm": 4.385629177093506, "learning_rate": 2.247547657711513e-08, "loss": 1.5867, "step": 49502 }, { "epoch": 3.8795454545454544, "grad_norm": 4.4640960693359375, "learning_rate": 2.2446281071474842e-08, "loss": 1.5213, "step": 49503 }, { "epoch": 3.8796238244514107, "grad_norm": 4.192162990570068, "learning_rate": 2.2417104497892695e-08, "loss": 1.1134, "step": 49504 }, { "epoch": 3.8797021943573666, "grad_norm": 4.983471870422363, "learning_rate": 2.2387946856479715e-08, "loss": 1.1814, "step": 49505 }, { "epoch": 3.879780564263323, "grad_norm": 2.704151153564453, "learning_rate": 2.2358808147345256e-08, "loss": 1.5243, "step": 49506 }, { "epoch": 3.879858934169279, "grad_norm": 3.5822391510009766, "learning_rate": 2.232968837060201e-08, "loss": 1.5953, "step": 49507 }, { "epoch": 3.879937304075235, "grad_norm": 3.4377894401550293, "learning_rate": 2.230058752635933e-08, "loss": 1.5297, "step": 49508 }, { "epoch": 3.880015673981191, "grad_norm": 4.282425880432129, "learning_rate": 2.227150561472935e-08, "loss": 1.3884, "step": 49509 }, { "epoch": 3.8800940438871474, "grad_norm": 10.8077392578125, "learning_rate": 2.224244263582087e-08, "loss": 1.6735, "step": 49510 }, { "epoch": 3.8801724137931033, "grad_norm": 4.699315547943115, "learning_rate": 2.2213398589746027e-08, "loss": 0.7846, "step": 49511 }, { "epoch": 3.8802507836990596, "grad_norm": 3.7640597820281982, "learning_rate": 2.2184373476614728e-08, "loss": 1.4471, "step": 49512 }, { "epoch": 3.8803291536050155, "grad_norm": 3.174027681350708, "learning_rate": 2.2155367296536888e-08, "loss": 1.432, "step": 49513 }, { "epoch": 3.880407523510972, "grad_norm": 5.18360710144043, "learning_rate": 2.212638004962353e-08, "loss": 1.2509, "step": 49514 }, { "epoch": 3.8804858934169277, "grad_norm": 3.988560914993286, "learning_rate": 2.2097411735984563e-08, "loss": 1.4384, "step": 49515 }, { "epoch": 3.880564263322884, "grad_norm": 4.568991661071777, "learning_rate": 2.2068462355729904e-08, "loss": 1.4289, "step": 49516 }, { "epoch": 3.88064263322884, "grad_norm": 5.828301429748535, "learning_rate": 2.2039531908970015e-08, "loss": 1.3482, "step": 49517 }, { "epoch": 3.8807210031347963, "grad_norm": 3.2371931076049805, "learning_rate": 2.2010620395815362e-08, "loss": 1.4412, "step": 49518 }, { "epoch": 3.8807993730407526, "grad_norm": 3.4730634689331055, "learning_rate": 2.1981727816375308e-08, "loss": 0.98, "step": 49519 }, { "epoch": 3.8808777429467085, "grad_norm": 2.321230411529541, "learning_rate": 2.1952854170759762e-08, "loss": 1.4316, "step": 49520 }, { "epoch": 3.8809561128526644, "grad_norm": 3.945364475250244, "learning_rate": 2.1923999459078637e-08, "loss": 1.2942, "step": 49521 }, { "epoch": 3.8810344827586207, "grad_norm": 7.367984771728516, "learning_rate": 2.1895163681441845e-08, "loss": 1.4621, "step": 49522 }, { "epoch": 3.881112852664577, "grad_norm": 5.102575302124023, "learning_rate": 2.1866346837959295e-08, "loss": 0.6544, "step": 49523 }, { "epoch": 3.881191222570533, "grad_norm": 3.336543560028076, "learning_rate": 2.1837548928739795e-08, "loss": 0.8083, "step": 49524 }, { "epoch": 3.881269592476489, "grad_norm": 4.364175796508789, "learning_rate": 2.180876995389325e-08, "loss": 1.5464, "step": 49525 }, { "epoch": 3.881347962382445, "grad_norm": 3.5647988319396973, "learning_rate": 2.178000991352902e-08, "loss": 1.1842, "step": 49526 }, { "epoch": 3.8814263322884015, "grad_norm": 2.885847568511963, "learning_rate": 2.175126880775702e-08, "loss": 1.2743, "step": 49527 }, { "epoch": 3.8815047021943574, "grad_norm": 8.17324447631836, "learning_rate": 2.172254663668605e-08, "loss": 1.5307, "step": 49528 }, { "epoch": 3.8815830721003133, "grad_norm": 3.6736974716186523, "learning_rate": 2.1693843400426016e-08, "loss": 1.281, "step": 49529 }, { "epoch": 3.8816614420062696, "grad_norm": 2.869960069656372, "learning_rate": 2.1665159099085176e-08, "loss": 1.2768, "step": 49530 }, { "epoch": 3.881739811912226, "grad_norm": 1.9601168632507324, "learning_rate": 2.163649373277288e-08, "loss": 1.1202, "step": 49531 }, { "epoch": 3.881818181818182, "grad_norm": 3.295565128326416, "learning_rate": 2.1607847301598483e-08, "loss": 1.3081, "step": 49532 }, { "epoch": 3.8818965517241377, "grad_norm": 8.91377067565918, "learning_rate": 2.157921980567024e-08, "loss": 0.7381, "step": 49533 }, { "epoch": 3.881974921630094, "grad_norm": 3.6678097248077393, "learning_rate": 2.1550611245098606e-08, "loss": 1.294, "step": 49534 }, { "epoch": 3.8820532915360504, "grad_norm": 3.632746696472168, "learning_rate": 2.1522021619990728e-08, "loss": 1.5839, "step": 49535 }, { "epoch": 3.8821316614420063, "grad_norm": 3.695833206176758, "learning_rate": 2.1493450930455962e-08, "loss": 1.696, "step": 49536 }, { "epoch": 3.882210031347962, "grad_norm": 3.774606227874756, "learning_rate": 2.1464899176603105e-08, "loss": 1.4877, "step": 49537 }, { "epoch": 3.8822884012539185, "grad_norm": 3.3308496475219727, "learning_rate": 2.1436366358540406e-08, "loss": 1.1973, "step": 49538 }, { "epoch": 3.882366771159875, "grad_norm": 2.7174198627471924, "learning_rate": 2.1407852476376668e-08, "loss": 1.3252, "step": 49539 }, { "epoch": 3.8824451410658307, "grad_norm": 3.0023858547210693, "learning_rate": 2.137935753022069e-08, "loss": 1.2115, "step": 49540 }, { "epoch": 3.8825235109717866, "grad_norm": 15.663862228393555, "learning_rate": 2.1350881520180167e-08, "loss": 0.8474, "step": 49541 }, { "epoch": 3.882601880877743, "grad_norm": 2.7765846252441406, "learning_rate": 2.1322424446363897e-08, "loss": 1.2873, "step": 49542 }, { "epoch": 3.8826802507836993, "grad_norm": 2.8918681144714355, "learning_rate": 2.1293986308879578e-08, "loss": 1.5253, "step": 49543 }, { "epoch": 3.882758620689655, "grad_norm": 4.601795673370361, "learning_rate": 2.1265567107836004e-08, "loss": 1.4108, "step": 49544 }, { "epoch": 3.882836990595611, "grad_norm": 3.61773681640625, "learning_rate": 2.1237166843340317e-08, "loss": 1.5514, "step": 49545 }, { "epoch": 3.8829153605015674, "grad_norm": 5.790926456451416, "learning_rate": 2.1208785515501872e-08, "loss": 1.0971, "step": 49546 }, { "epoch": 3.8829937304075237, "grad_norm": 4.274421691894531, "learning_rate": 2.1180423124427807e-08, "loss": 0.8606, "step": 49547 }, { "epoch": 3.8830721003134796, "grad_norm": 2.8618626594543457, "learning_rate": 2.1152079670226365e-08, "loss": 1.1335, "step": 49548 }, { "epoch": 3.8831504702194355, "grad_norm": 2.6656389236450195, "learning_rate": 2.112375515300469e-08, "loss": 1.5753, "step": 49549 }, { "epoch": 3.883228840125392, "grad_norm": 2.6449756622314453, "learning_rate": 2.109544957287102e-08, "loss": 1.3082, "step": 49550 }, { "epoch": 3.883307210031348, "grad_norm": 5.841728210449219, "learning_rate": 2.106716292993305e-08, "loss": 1.3132, "step": 49551 }, { "epoch": 3.883385579937304, "grad_norm": 1.960344910621643, "learning_rate": 2.1038895224298474e-08, "loss": 1.3807, "step": 49552 }, { "epoch": 3.88346394984326, "grad_norm": 2.732363700866699, "learning_rate": 2.1010646456074423e-08, "loss": 1.1169, "step": 49553 }, { "epoch": 3.8835423197492163, "grad_norm": 2.9513351917266846, "learning_rate": 2.098241662536804e-08, "loss": 1.3492, "step": 49554 }, { "epoch": 3.8836206896551726, "grad_norm": 2.6634719371795654, "learning_rate": 2.095420573228757e-08, "loss": 1.364, "step": 49555 }, { "epoch": 3.8836990595611285, "grad_norm": 4.770018577575684, "learning_rate": 2.092601377693959e-08, "loss": 1.4657, "step": 49556 }, { "epoch": 3.8837774294670844, "grad_norm": 3.873602867126465, "learning_rate": 2.089784075943235e-08, "loss": 1.2786, "step": 49557 }, { "epoch": 3.8838557993730407, "grad_norm": 2.2988131046295166, "learning_rate": 2.0869686679871882e-08, "loss": 1.5532, "step": 49558 }, { "epoch": 3.883934169278997, "grad_norm": 3.9978511333465576, "learning_rate": 2.0841551538365312e-08, "loss": 1.4733, "step": 49559 }, { "epoch": 3.884012539184953, "grad_norm": 10.188355445861816, "learning_rate": 2.0813435335020336e-08, "loss": 1.077, "step": 49560 }, { "epoch": 3.884090909090909, "grad_norm": 5.579684734344482, "learning_rate": 2.0785338069944095e-08, "loss": 1.202, "step": 49561 }, { "epoch": 3.884169278996865, "grad_norm": 3.550527334213257, "learning_rate": 2.0757259743242055e-08, "loss": 0.933, "step": 49562 }, { "epoch": 3.8842476489028215, "grad_norm": 2.943643569946289, "learning_rate": 2.0729200355022462e-08, "loss": 1.2228, "step": 49563 }, { "epoch": 3.8843260188087774, "grad_norm": 3.5855374336242676, "learning_rate": 2.0701159905391342e-08, "loss": 1.0685, "step": 49564 }, { "epoch": 3.8844043887147337, "grad_norm": 3.400423049926758, "learning_rate": 2.0673138394455283e-08, "loss": 1.5693, "step": 49565 }, { "epoch": 3.8844827586206896, "grad_norm": 3.2225325107574463, "learning_rate": 2.0645135822321417e-08, "loss": 0.9519, "step": 49566 }, { "epoch": 3.884561128526646, "grad_norm": 7.420009136199951, "learning_rate": 2.061715218909577e-08, "loss": 0.7576, "step": 49567 }, { "epoch": 3.884639498432602, "grad_norm": 2.8690290451049805, "learning_rate": 2.0589187494884366e-08, "loss": 1.4259, "step": 49568 }, { "epoch": 3.884717868338558, "grad_norm": 2.469085216522217, "learning_rate": 2.0561241739794902e-08, "loss": 0.97, "step": 49569 }, { "epoch": 3.884796238244514, "grad_norm": 3.08056378364563, "learning_rate": 2.053331492393229e-08, "loss": 1.327, "step": 49570 }, { "epoch": 3.8848746081504704, "grad_norm": 3.3236324787139893, "learning_rate": 2.050540704740367e-08, "loss": 1.3239, "step": 49571 }, { "epoch": 3.8849529780564263, "grad_norm": 2.422178030014038, "learning_rate": 2.0477518110315065e-08, "loss": 1.3907, "step": 49572 }, { "epoch": 3.8850313479623826, "grad_norm": 4.203302383422852, "learning_rate": 2.0449648112771946e-08, "loss": 1.0109, "step": 49573 }, { "epoch": 3.8851097178683385, "grad_norm": 3.0771749019622803, "learning_rate": 2.04217970548809e-08, "loss": 1.3463, "step": 49574 }, { "epoch": 3.885188087774295, "grad_norm": 3.046638250350952, "learning_rate": 2.039396493674739e-08, "loss": 1.448, "step": 49575 }, { "epoch": 3.8852664576802507, "grad_norm": 3.983612298965454, "learning_rate": 2.0366151758478002e-08, "loss": 1.3159, "step": 49576 }, { "epoch": 3.885344827586207, "grad_norm": 2.69047212600708, "learning_rate": 2.0338357520177655e-08, "loss": 1.5175, "step": 49577 }, { "epoch": 3.885423197492163, "grad_norm": 3.6672990322113037, "learning_rate": 2.031058222195237e-08, "loss": 1.3493, "step": 49578 }, { "epoch": 3.8855015673981192, "grad_norm": 3.661351203918457, "learning_rate": 2.0282825863908174e-08, "loss": 1.2003, "step": 49579 }, { "epoch": 3.885579937304075, "grad_norm": 3.3098390102386475, "learning_rate": 2.025508844615054e-08, "loss": 1.3431, "step": 49580 }, { "epoch": 3.8856583072100315, "grad_norm": 2.370620012283325, "learning_rate": 2.0227369968784383e-08, "loss": 0.8835, "step": 49581 }, { "epoch": 3.8857366771159874, "grad_norm": 3.72029709815979, "learning_rate": 2.019967043191573e-08, "loss": 1.2787, "step": 49582 }, { "epoch": 3.8858150470219437, "grad_norm": 6.249757766723633, "learning_rate": 2.0171989835650052e-08, "loss": 0.7382, "step": 49583 }, { "epoch": 3.8858934169278996, "grad_norm": 4.471879482269287, "learning_rate": 2.014432818009171e-08, "loss": 1.0121, "step": 49584 }, { "epoch": 3.885971786833856, "grad_norm": 2.351020097732544, "learning_rate": 2.0116685465347285e-08, "loss": 1.7187, "step": 49585 }, { "epoch": 3.886050156739812, "grad_norm": 2.9750421047210693, "learning_rate": 2.0089061691520585e-08, "loss": 1.0977, "step": 49586 }, { "epoch": 3.886128526645768, "grad_norm": 3.3924121856689453, "learning_rate": 2.0061456858717075e-08, "loss": 1.6314, "step": 49587 }, { "epoch": 3.886206896551724, "grad_norm": 2.999732494354248, "learning_rate": 2.0033870967042234e-08, "loss": 0.8919, "step": 49588 }, { "epoch": 3.8862852664576804, "grad_norm": 6.288122177124023, "learning_rate": 2.0006304016600974e-08, "loss": 1.2989, "step": 49589 }, { "epoch": 3.8863636363636362, "grad_norm": 3.5820748805999756, "learning_rate": 1.9978756007497656e-08, "loss": 1.0335, "step": 49590 }, { "epoch": 3.8864420062695926, "grad_norm": 4.831137657165527, "learning_rate": 1.99512269398372e-08, "loss": 1.4892, "step": 49591 }, { "epoch": 3.8865203761755485, "grad_norm": 2.2675366401672363, "learning_rate": 1.9923716813723958e-08, "loss": 1.2005, "step": 49592 }, { "epoch": 3.886598746081505, "grad_norm": 3.1323320865631104, "learning_rate": 1.989622562926341e-08, "loss": 1.11, "step": 49593 }, { "epoch": 3.8866771159874607, "grad_norm": 3.6643717288970947, "learning_rate": 1.9868753386559915e-08, "loss": 1.282, "step": 49594 }, { "epoch": 3.886755485893417, "grad_norm": 3.695770502090454, "learning_rate": 1.9841300085717274e-08, "loss": 1.0948, "step": 49595 }, { "epoch": 3.886833855799373, "grad_norm": 2.6849112510681152, "learning_rate": 1.981386572684041e-08, "loss": 1.3353, "step": 49596 }, { "epoch": 3.8869122257053292, "grad_norm": 8.850104331970215, "learning_rate": 1.9786450310033677e-08, "loss": 1.2031, "step": 49597 }, { "epoch": 3.886990595611285, "grad_norm": 4.623961925506592, "learning_rate": 1.9759053835401443e-08, "loss": 1.3927, "step": 49598 }, { "epoch": 3.8870689655172415, "grad_norm": 4.0689778327941895, "learning_rate": 1.9731676303047507e-08, "loss": 1.2508, "step": 49599 }, { "epoch": 3.8871473354231973, "grad_norm": 2.7111656665802, "learning_rate": 1.9704317713076236e-08, "loss": 1.3704, "step": 49600 }, { "epoch": 3.8872257053291537, "grad_norm": 3.8426873683929443, "learning_rate": 1.967697806559199e-08, "loss": 1.6313, "step": 49601 }, { "epoch": 3.8873040752351096, "grad_norm": 2.3467788696289062, "learning_rate": 1.9649657360698016e-08, "loss": 1.5096, "step": 49602 }, { "epoch": 3.887382445141066, "grad_norm": 2.9933719635009766, "learning_rate": 1.9622355598499232e-08, "loss": 1.1048, "step": 49603 }, { "epoch": 3.887460815047022, "grad_norm": 7.321776866912842, "learning_rate": 1.9595072779098335e-08, "loss": 1.2605, "step": 49604 }, { "epoch": 3.887539184952978, "grad_norm": 5.565243721008301, "learning_rate": 1.956780890260024e-08, "loss": 1.2154, "step": 49605 }, { "epoch": 3.887617554858934, "grad_norm": 4.1818156242370605, "learning_rate": 1.954056396910764e-08, "loss": 1.5422, "step": 49606 }, { "epoch": 3.8876959247648903, "grad_norm": 5.431314468383789, "learning_rate": 1.95133379787249e-08, "loss": 1.2715, "step": 49607 }, { "epoch": 3.8877742946708462, "grad_norm": 3.17610502243042, "learning_rate": 1.948613093155527e-08, "loss": 1.3347, "step": 49608 }, { "epoch": 3.8878526645768026, "grad_norm": 3.2419466972351074, "learning_rate": 1.9458942827702554e-08, "loss": 1.2057, "step": 49609 }, { "epoch": 3.887931034482759, "grad_norm": 3.810450315475464, "learning_rate": 1.9431773667268894e-08, "loss": 1.3316, "step": 49610 }, { "epoch": 3.888009404388715, "grad_norm": 3.5194177627563477, "learning_rate": 1.9404623450359205e-08, "loss": 1.5716, "step": 49611 }, { "epoch": 3.8880877742946707, "grad_norm": 3.3676044940948486, "learning_rate": 1.9377492177076185e-08, "loss": 1.5159, "step": 49612 }, { "epoch": 3.888166144200627, "grad_norm": 3.285731792449951, "learning_rate": 1.935037984752308e-08, "loss": 1.3334, "step": 49613 }, { "epoch": 3.8882445141065833, "grad_norm": 4.639120578765869, "learning_rate": 1.9323286461802593e-08, "loss": 1.2624, "step": 49614 }, { "epoch": 3.8883228840125392, "grad_norm": 4.0735249519348145, "learning_rate": 1.9296212020017967e-08, "loss": 1.285, "step": 49615 }, { "epoch": 3.888401253918495, "grad_norm": 4.934383869171143, "learning_rate": 1.9269156522272457e-08, "loss": 0.7796, "step": 49616 }, { "epoch": 3.8884796238244514, "grad_norm": 5.694199562072754, "learning_rate": 1.9242119968669316e-08, "loss": 0.9102, "step": 49617 }, { "epoch": 3.888557993730408, "grad_norm": 18.711912155151367, "learning_rate": 1.9215102359310122e-08, "loss": 1.1319, "step": 49618 }, { "epoch": 3.8886363636363637, "grad_norm": 2.4230496883392334, "learning_rate": 1.918810369429869e-08, "loss": 1.2326, "step": 49619 }, { "epoch": 3.8887147335423196, "grad_norm": 4.617197513580322, "learning_rate": 1.9161123973737706e-08, "loss": 1.1606, "step": 49620 }, { "epoch": 3.888793103448276, "grad_norm": 4.586686611175537, "learning_rate": 1.913416319772876e-08, "loss": 1.0042, "step": 49621 }, { "epoch": 3.8888714733542322, "grad_norm": 6.2937493324279785, "learning_rate": 1.9107221366375662e-08, "loss": 1.0477, "step": 49622 }, { "epoch": 3.888949843260188, "grad_norm": 7.780090808868408, "learning_rate": 1.9080298479780546e-08, "loss": 1.5511, "step": 49623 }, { "epoch": 3.889028213166144, "grad_norm": 4.59665584564209, "learning_rate": 1.9053394538045e-08, "loss": 0.7181, "step": 49624 }, { "epoch": 3.8891065830721003, "grad_norm": 3.9375016689300537, "learning_rate": 1.9026509541272276e-08, "loss": 1.2411, "step": 49625 }, { "epoch": 3.8891849529780567, "grad_norm": 7.00520133972168, "learning_rate": 1.8999643489564513e-08, "loss": 1.249, "step": 49626 }, { "epoch": 3.8892633228840126, "grad_norm": 4.229745388031006, "learning_rate": 1.8972796383023294e-08, "loss": 1.2598, "step": 49627 }, { "epoch": 3.8893416927899684, "grad_norm": 4.583343505859375, "learning_rate": 1.8945968221751875e-08, "loss": 1.4374, "step": 49628 }, { "epoch": 3.8894200626959248, "grad_norm": 3.7765629291534424, "learning_rate": 1.8919159005850728e-08, "loss": 1.079, "step": 49629 }, { "epoch": 3.889498432601881, "grad_norm": 3.8149118423461914, "learning_rate": 1.8892368735422552e-08, "loss": 1.3136, "step": 49630 }, { "epoch": 3.889576802507837, "grad_norm": 2.3255455493927, "learning_rate": 1.8865597410570037e-08, "loss": 1.1554, "step": 49631 }, { "epoch": 3.889655172413793, "grad_norm": 2.476550579071045, "learning_rate": 1.8838845031394215e-08, "loss": 1.3277, "step": 49632 }, { "epoch": 3.889733542319749, "grad_norm": 2.8019862174987793, "learning_rate": 1.8812111597996674e-08, "loss": 1.426, "step": 49633 }, { "epoch": 3.8898119122257055, "grad_norm": 3.2646381855010986, "learning_rate": 1.8785397110479553e-08, "loss": 1.1523, "step": 49634 }, { "epoch": 3.8898902821316614, "grad_norm": 3.8136258125305176, "learning_rate": 1.875870156894388e-08, "loss": 1.369, "step": 49635 }, { "epoch": 3.8899686520376173, "grad_norm": 3.3305437564849854, "learning_rate": 1.8732024973491803e-08, "loss": 1.3269, "step": 49636 }, { "epoch": 3.8900470219435737, "grad_norm": 6.571638584136963, "learning_rate": 1.8705367324224898e-08, "loss": 1.1866, "step": 49637 }, { "epoch": 3.89012539184953, "grad_norm": 2.6261038780212402, "learning_rate": 1.8678728621243647e-08, "loss": 1.3987, "step": 49638 }, { "epoch": 3.890203761755486, "grad_norm": 7.2852935791015625, "learning_rate": 1.865210886465074e-08, "loss": 0.8914, "step": 49639 }, { "epoch": 3.8902821316614418, "grad_norm": 3.8155291080474854, "learning_rate": 1.8625508054546106e-08, "loss": 1.0504, "step": 49640 }, { "epoch": 3.890360501567398, "grad_norm": 2.9608638286590576, "learning_rate": 1.859892619103132e-08, "loss": 1.4953, "step": 49641 }, { "epoch": 3.8904388714733544, "grad_norm": 2.910614252090454, "learning_rate": 1.8572363274207417e-08, "loss": 1.2817, "step": 49642 }, { "epoch": 3.8905172413793103, "grad_norm": 11.573986053466797, "learning_rate": 1.854581930417598e-08, "loss": 1.0408, "step": 49643 }, { "epoch": 3.890595611285266, "grad_norm": 3.063476800918579, "learning_rate": 1.8519294281038047e-08, "loss": 1.4784, "step": 49644 }, { "epoch": 3.8906739811912225, "grad_norm": 6.282668590545654, "learning_rate": 1.8492788204893532e-08, "loss": 1.185, "step": 49645 }, { "epoch": 3.890752351097179, "grad_norm": 3.1618659496307373, "learning_rate": 1.8466301075844016e-08, "loss": 1.4816, "step": 49646 }, { "epoch": 3.8908307210031348, "grad_norm": 2.8201591968536377, "learning_rate": 1.8439832893989983e-08, "loss": 1.32, "step": 49647 }, { "epoch": 3.8909090909090907, "grad_norm": 5.974241256713867, "learning_rate": 1.841338365943135e-08, "loss": 1.1957, "step": 49648 }, { "epoch": 3.890987460815047, "grad_norm": 3.7490532398223877, "learning_rate": 1.8386953372270255e-08, "loss": 1.4576, "step": 49649 }, { "epoch": 3.8910658307210033, "grad_norm": 3.1198692321777344, "learning_rate": 1.836054203260662e-08, "loss": 0.6966, "step": 49650 }, { "epoch": 3.891144200626959, "grad_norm": 4.9637675285339355, "learning_rate": 1.8334149640540365e-08, "loss": 1.1906, "step": 49651 }, { "epoch": 3.891222570532915, "grad_norm": 7.134079933166504, "learning_rate": 1.8307776196172522e-08, "loss": 1.5758, "step": 49652 }, { "epoch": 3.8913009404388714, "grad_norm": 2.5368714332580566, "learning_rate": 1.8281421699603007e-08, "loss": 1.1535, "step": 49653 }, { "epoch": 3.8913793103448278, "grad_norm": 3.300510883331299, "learning_rate": 1.82550861509323e-08, "loss": 1.0711, "step": 49654 }, { "epoch": 3.8914576802507836, "grad_norm": 3.4378364086151123, "learning_rate": 1.8228769550260316e-08, "loss": 1.7495, "step": 49655 }, { "epoch": 3.8915360501567395, "grad_norm": 2.7048099040985107, "learning_rate": 1.8202471897687534e-08, "loss": 1.3692, "step": 49656 }, { "epoch": 3.891614420062696, "grad_norm": 4.802304267883301, "learning_rate": 1.817619319331332e-08, "loss": 1.6625, "step": 49657 }, { "epoch": 3.891692789968652, "grad_norm": 3.2204172611236572, "learning_rate": 1.814993343723814e-08, "loss": 1.6015, "step": 49658 }, { "epoch": 3.891771159874608, "grad_norm": 3.506800889968872, "learning_rate": 1.8123692629561928e-08, "loss": 1.4352, "step": 49659 }, { "epoch": 3.8918495297805644, "grad_norm": 3.6244537830352783, "learning_rate": 1.8097470770384596e-08, "loss": 1.6403, "step": 49660 }, { "epoch": 3.8919278996865203, "grad_norm": 1.9647271633148193, "learning_rate": 1.807126785980551e-08, "loss": 1.3739, "step": 49661 }, { "epoch": 3.8920062695924766, "grad_norm": 12.319910049438477, "learning_rate": 1.804508389792403e-08, "loss": 1.129, "step": 49662 }, { "epoch": 3.8920846394984325, "grad_norm": 4.283763885498047, "learning_rate": 1.8018918884840085e-08, "loss": 1.277, "step": 49663 }, { "epoch": 3.892163009404389, "grad_norm": 4.714391708374023, "learning_rate": 1.7992772820653593e-08, "loss": 1.0778, "step": 49664 }, { "epoch": 3.8922413793103448, "grad_norm": 2.1631271839141846, "learning_rate": 1.7966645705463914e-08, "loss": 1.2875, "step": 49665 }, { "epoch": 3.892319749216301, "grad_norm": 7.392413139343262, "learning_rate": 1.794053753936986e-08, "loss": 1.2529, "step": 49666 }, { "epoch": 3.892398119122257, "grad_norm": 3.1161601543426514, "learning_rate": 1.7914448322471357e-08, "loss": 1.6143, "step": 49667 }, { "epoch": 3.8924764890282133, "grad_norm": 3.5964863300323486, "learning_rate": 1.788837805486665e-08, "loss": 1.6005, "step": 49668 }, { "epoch": 3.892554858934169, "grad_norm": 2.7827467918395996, "learning_rate": 1.7862326736656222e-08, "loss": 1.3019, "step": 49669 }, { "epoch": 3.8926332288401255, "grad_norm": 5.190084934234619, "learning_rate": 1.7836294367938324e-08, "loss": 0.9976, "step": 49670 }, { "epoch": 3.8927115987460814, "grad_norm": 3.1809325218200684, "learning_rate": 1.781028094881232e-08, "loss": 1.2254, "step": 49671 }, { "epoch": 3.8927899686520377, "grad_norm": 5.846243858337402, "learning_rate": 1.7784286479376468e-08, "loss": 1.6155, "step": 49672 }, { "epoch": 3.8928683385579936, "grad_norm": 6.235176086425781, "learning_rate": 1.7758310959730683e-08, "loss": 1.0094, "step": 49673 }, { "epoch": 3.89294670846395, "grad_norm": 2.285088539123535, "learning_rate": 1.7732354389973228e-08, "loss": 1.3614, "step": 49674 }, { "epoch": 3.893025078369906, "grad_norm": 2.3829476833343506, "learning_rate": 1.7706416770202907e-08, "loss": 1.6431, "step": 49675 }, { "epoch": 3.893103448275862, "grad_norm": 3.9017765522003174, "learning_rate": 1.7680498100517972e-08, "loss": 1.2612, "step": 49676 }, { "epoch": 3.893181818181818, "grad_norm": 3.1193647384643555, "learning_rate": 1.765459838101724e-08, "loss": 1.1909, "step": 49677 }, { "epoch": 3.8932601880877744, "grad_norm": 2.9153270721435547, "learning_rate": 1.7628717611799518e-08, "loss": 1.5247, "step": 49678 }, { "epoch": 3.8933385579937303, "grad_norm": 2.9913439750671387, "learning_rate": 1.7602855792963058e-08, "loss": 0.9842, "step": 49679 }, { "epoch": 3.8934169278996866, "grad_norm": 3.3270184993743896, "learning_rate": 1.7577012924606673e-08, "loss": 0.8841, "step": 49680 }, { "epoch": 3.8934952978056425, "grad_norm": 3.139089822769165, "learning_rate": 1.7551189006827506e-08, "loss": 1.1546, "step": 49681 }, { "epoch": 3.893573667711599, "grad_norm": 6.172229766845703, "learning_rate": 1.752538403972437e-08, "loss": 0.8719, "step": 49682 }, { "epoch": 3.8936520376175547, "grad_norm": 3.6114425659179688, "learning_rate": 1.7499598023396068e-08, "loss": 1.1906, "step": 49683 }, { "epoch": 3.893730407523511, "grad_norm": 3.849195718765259, "learning_rate": 1.7473830957940308e-08, "loss": 0.8939, "step": 49684 }, { "epoch": 3.893808777429467, "grad_norm": 3.2108614444732666, "learning_rate": 1.7448082843454228e-08, "loss": 1.6388, "step": 49685 }, { "epoch": 3.8938871473354233, "grad_norm": 3.1813652515411377, "learning_rate": 1.7422353680036642e-08, "loss": 1.3607, "step": 49686 }, { "epoch": 3.893965517241379, "grad_norm": 2.9236905574798584, "learning_rate": 1.7396643467785247e-08, "loss": 1.0347, "step": 49687 }, { "epoch": 3.8940438871473355, "grad_norm": 3.2515041828155518, "learning_rate": 1.7370952206797742e-08, "loss": 1.3859, "step": 49688 }, { "epoch": 3.8941222570532914, "grad_norm": 2.4586756229400635, "learning_rate": 1.7345279897172385e-08, "loss": 1.3021, "step": 49689 }, { "epoch": 3.8942006269592477, "grad_norm": 2.8939082622528076, "learning_rate": 1.7319626539005762e-08, "loss": 1.0158, "step": 49690 }, { "epoch": 3.8942789968652036, "grad_norm": 2.481576919555664, "learning_rate": 1.729399213239613e-08, "loss": 1.2327, "step": 49691 }, { "epoch": 3.89435736677116, "grad_norm": 3.8964271545410156, "learning_rate": 1.726837667744119e-08, "loss": 1.4121, "step": 49692 }, { "epoch": 3.894435736677116, "grad_norm": 3.7942378520965576, "learning_rate": 1.7242780174237526e-08, "loss": 1.1343, "step": 49693 }, { "epoch": 3.894514106583072, "grad_norm": 3.504260540008545, "learning_rate": 1.72172026228834e-08, "loss": 1.5543, "step": 49694 }, { "epoch": 3.894592476489028, "grad_norm": 5.303806781768799, "learning_rate": 1.7191644023475394e-08, "loss": 1.214, "step": 49695 }, { "epoch": 3.8946708463949844, "grad_norm": 4.78878116607666, "learning_rate": 1.7166104376111216e-08, "loss": 1.6492, "step": 49696 }, { "epoch": 3.8947492163009403, "grad_norm": 3.0352416038513184, "learning_rate": 1.7140583680888e-08, "loss": 1.3407, "step": 49697 }, { "epoch": 3.8948275862068966, "grad_norm": 4.214634895324707, "learning_rate": 1.7115081937902345e-08, "loss": 1.6338, "step": 49698 }, { "epoch": 3.8949059561128525, "grad_norm": 3.8089895248413086, "learning_rate": 1.7089599147251944e-08, "loss": 1.4339, "step": 49699 }, { "epoch": 3.894984326018809, "grad_norm": 3.8704280853271484, "learning_rate": 1.7064135309032837e-08, "loss": 1.4037, "step": 49700 }, { "epoch": 3.8950626959247647, "grad_norm": 6.07628059387207, "learning_rate": 1.7038690423342718e-08, "loss": 1.4543, "step": 49701 }, { "epoch": 3.895141065830721, "grad_norm": 2.5942270755767822, "learning_rate": 1.701326449027818e-08, "loss": 1.4422, "step": 49702 }, { "epoch": 3.895219435736677, "grad_norm": 3.9932901859283447, "learning_rate": 1.698785750993526e-08, "loss": 1.6208, "step": 49703 }, { "epoch": 3.8952978056426333, "grad_norm": 2.7713940143585205, "learning_rate": 1.6962469482411648e-08, "loss": 1.3184, "step": 49704 }, { "epoch": 3.8953761755485896, "grad_norm": 2.306575059890747, "learning_rate": 1.693710040780283e-08, "loss": 1.2711, "step": 49705 }, { "epoch": 3.8954545454545455, "grad_norm": 3.7824738025665283, "learning_rate": 1.6911750286205954e-08, "loss": 1.3768, "step": 49706 }, { "epoch": 3.8955329153605014, "grad_norm": 2.899996757507324, "learning_rate": 1.68864191177176e-08, "loss": 1.1586, "step": 49707 }, { "epoch": 3.8956112852664577, "grad_norm": 3.642347812652588, "learning_rate": 1.6861106902433255e-08, "loss": 1.1663, "step": 49708 }, { "epoch": 3.895689655172414, "grad_norm": 4.3843488693237305, "learning_rate": 1.683581364045006e-08, "loss": 0.9673, "step": 49709 }, { "epoch": 3.89576802507837, "grad_norm": 6.905278205871582, "learning_rate": 1.681053933186405e-08, "loss": 1.1992, "step": 49710 }, { "epoch": 3.895846394984326, "grad_norm": 4.136180400848389, "learning_rate": 1.6785283976771262e-08, "loss": 1.4915, "step": 49711 }, { "epoch": 3.895924764890282, "grad_norm": 3.798781394958496, "learning_rate": 1.676004757526717e-08, "loss": 1.2969, "step": 49712 }, { "epoch": 3.8960031347962385, "grad_norm": 9.541190147399902, "learning_rate": 1.673483012744892e-08, "loss": 1.0714, "step": 49713 }, { "epoch": 3.8960815047021944, "grad_norm": 5.2118144035339355, "learning_rate": 1.6709631633411437e-08, "loss": 1.0401, "step": 49714 }, { "epoch": 3.8961598746081503, "grad_norm": 3.4392683506011963, "learning_rate": 1.6684452093251314e-08, "loss": 1.2844, "step": 49715 }, { "epoch": 3.8962382445141066, "grad_norm": 3.2347424030303955, "learning_rate": 1.665929150706347e-08, "loss": 1.3126, "step": 49716 }, { "epoch": 3.896316614420063, "grad_norm": 4.812089443206787, "learning_rate": 1.6634149874944493e-08, "loss": 1.2324, "step": 49717 }, { "epoch": 3.896394984326019, "grad_norm": 3.8999953269958496, "learning_rate": 1.6609027196989313e-08, "loss": 0.9421, "step": 49718 }, { "epoch": 3.8964733542319747, "grad_norm": 12.781386375427246, "learning_rate": 1.6583923473293963e-08, "loss": 1.1168, "step": 49719 }, { "epoch": 3.896551724137931, "grad_norm": 3.139897584915161, "learning_rate": 1.655883870395336e-08, "loss": 1.4572, "step": 49720 }, { "epoch": 3.8966300940438874, "grad_norm": 3.0216383934020996, "learning_rate": 1.6533772889064103e-08, "loss": 1.5826, "step": 49721 }, { "epoch": 3.8967084639498433, "grad_norm": 2.3263561725616455, "learning_rate": 1.650872602872e-08, "loss": 1.0704, "step": 49722 }, { "epoch": 3.896786833855799, "grad_norm": 5.630681991577148, "learning_rate": 1.648369812301709e-08, "loss": 1.3262, "step": 49723 }, { "epoch": 3.8968652037617555, "grad_norm": 6.7397027015686035, "learning_rate": 1.645868917205029e-08, "loss": 0.6278, "step": 49724 }, { "epoch": 3.896943573667712, "grad_norm": 4.581737041473389, "learning_rate": 1.643369917591564e-08, "loss": 1.6152, "step": 49725 }, { "epoch": 3.8970219435736677, "grad_norm": 3.6477584838867188, "learning_rate": 1.6408728134706953e-08, "loss": 1.9096, "step": 49726 }, { "epoch": 3.8971003134796236, "grad_norm": 4.964804649353027, "learning_rate": 1.6383776048519706e-08, "loss": 1.4719, "step": 49727 }, { "epoch": 3.89717868338558, "grad_norm": 6.656372547149658, "learning_rate": 1.6358842917448825e-08, "loss": 1.0907, "step": 49728 }, { "epoch": 3.8972570532915363, "grad_norm": 2.6299405097961426, "learning_rate": 1.6333928741589233e-08, "loss": 1.5123, "step": 49729 }, { "epoch": 3.897335423197492, "grad_norm": 3.378218412399292, "learning_rate": 1.6309033521035855e-08, "loss": 1.5217, "step": 49730 }, { "epoch": 3.897413793103448, "grad_norm": 3.271703004837036, "learning_rate": 1.628415725588306e-08, "loss": 0.731, "step": 49731 }, { "epoch": 3.8974921630094044, "grad_norm": 3.9674770832061768, "learning_rate": 1.6259299946225217e-08, "loss": 1.2383, "step": 49732 }, { "epoch": 3.8975705329153607, "grad_norm": 4.0257086753845215, "learning_rate": 1.6234461592157246e-08, "loss": 0.7881, "step": 49733 }, { "epoch": 3.8976489028213166, "grad_norm": 2.7116363048553467, "learning_rate": 1.6209642193773523e-08, "loss": 1.386, "step": 49734 }, { "epoch": 3.8977272727272725, "grad_norm": 3.3455519676208496, "learning_rate": 1.6184841751168966e-08, "loss": 1.0693, "step": 49735 }, { "epoch": 3.897805642633229, "grad_norm": 4.368076801300049, "learning_rate": 1.616006026443684e-08, "loss": 1.1339, "step": 49736 }, { "epoch": 3.897884012539185, "grad_norm": 5.296037673950195, "learning_rate": 1.613529773367206e-08, "loss": 1.2942, "step": 49737 }, { "epoch": 3.897962382445141, "grad_norm": 4.03342866897583, "learning_rate": 1.6110554158969004e-08, "loss": 0.6411, "step": 49738 }, { "epoch": 3.898040752351097, "grad_norm": 6.103884220123291, "learning_rate": 1.6085829540421484e-08, "loss": 1.3443, "step": 49739 }, { "epoch": 3.8981191222570533, "grad_norm": 30.14211082458496, "learning_rate": 1.6061123878123864e-08, "loss": 1.7729, "step": 49740 }, { "epoch": 3.8981974921630096, "grad_norm": 2.857701301574707, "learning_rate": 1.6036437172169405e-08, "loss": 1.3742, "step": 49741 }, { "epoch": 3.8982758620689655, "grad_norm": 3.7762327194213867, "learning_rate": 1.6011769422651925e-08, "loss": 1.5295, "step": 49742 }, { "epoch": 3.8983542319749214, "grad_norm": 2.881520986557007, "learning_rate": 1.5987120629666897e-08, "loss": 1.2255, "step": 49743 }, { "epoch": 3.8984326018808777, "grad_norm": 4.032797813415527, "learning_rate": 1.5962490793305917e-08, "loss": 1.4677, "step": 49744 }, { "epoch": 3.898510971786834, "grad_norm": 2.6206371784210205, "learning_rate": 1.5937879913663912e-08, "loss": 1.4238, "step": 49745 }, { "epoch": 3.89858934169279, "grad_norm": 4.5695481300354, "learning_rate": 1.591328799083469e-08, "loss": 1.4551, "step": 49746 }, { "epoch": 3.898667711598746, "grad_norm": 6.6468024253845215, "learning_rate": 1.588871502491096e-08, "loss": 2.0676, "step": 49747 }, { "epoch": 3.898746081504702, "grad_norm": 4.0340576171875, "learning_rate": 1.5864161015986533e-08, "loss": 1.4015, "step": 49748 }, { "epoch": 3.8988244514106585, "grad_norm": 4.159191131591797, "learning_rate": 1.583962596415467e-08, "loss": 1.5325, "step": 49749 }, { "epoch": 3.8989028213166144, "grad_norm": 2.4679439067840576, "learning_rate": 1.5815109869509183e-08, "loss": 1.3707, "step": 49750 }, { "epoch": 3.8989811912225707, "grad_norm": 2.1318752765655518, "learning_rate": 1.5790612732142772e-08, "loss": 1.3585, "step": 49751 }, { "epoch": 3.8990595611285266, "grad_norm": 3.3476850986480713, "learning_rate": 1.5766134552148704e-08, "loss": 1.1691, "step": 49752 }, { "epoch": 3.899137931034483, "grad_norm": 2.298788547515869, "learning_rate": 1.5741675329620786e-08, "loss": 0.8329, "step": 49753 }, { "epoch": 3.899216300940439, "grad_norm": 2.753497838973999, "learning_rate": 1.5717235064650617e-08, "loss": 1.398, "step": 49754 }, { "epoch": 3.899294670846395, "grad_norm": 2.7932698726654053, "learning_rate": 1.569281375733256e-08, "loss": 1.376, "step": 49755 }, { "epoch": 3.899373040752351, "grad_norm": 4.6498870849609375, "learning_rate": 1.5668411407758766e-08, "loss": 1.7095, "step": 49756 }, { "epoch": 3.8994514106583074, "grad_norm": 3.983826160430908, "learning_rate": 1.5644028016022495e-08, "loss": 1.4627, "step": 49757 }, { "epoch": 3.8995297805642632, "grad_norm": 4.395090579986572, "learning_rate": 1.5619663582215894e-08, "loss": 1.5534, "step": 49758 }, { "epoch": 3.8996081504702196, "grad_norm": 3.0060360431671143, "learning_rate": 1.559531810643222e-08, "loss": 1.475, "step": 49759 }, { "epoch": 3.8996865203761755, "grad_norm": 4.002276420593262, "learning_rate": 1.557099158876363e-08, "loss": 0.9786, "step": 49760 }, { "epoch": 3.899764890282132, "grad_norm": 5.18300199508667, "learning_rate": 1.554668402930226e-08, "loss": 1.3319, "step": 49761 }, { "epoch": 3.8998432601880877, "grad_norm": 2.4731180667877197, "learning_rate": 1.5522395428141934e-08, "loss": 1.1284, "step": 49762 }, { "epoch": 3.899921630094044, "grad_norm": 4.631864070892334, "learning_rate": 1.549812578537424e-08, "loss": 1.3699, "step": 49763 }, { "epoch": 3.9, "grad_norm": 3.6700658798217773, "learning_rate": 1.5473875101091328e-08, "loss": 1.0841, "step": 49764 }, { "epoch": 3.9000783699059562, "grad_norm": 3.560033082962036, "learning_rate": 1.5449643375385346e-08, "loss": 1.647, "step": 49765 }, { "epoch": 3.900156739811912, "grad_norm": 4.593863487243652, "learning_rate": 1.5425430608348445e-08, "loss": 1.37, "step": 49766 }, { "epoch": 3.9002351097178685, "grad_norm": 3.2451295852661133, "learning_rate": 1.5401236800073327e-08, "loss": 0.9908, "step": 49767 }, { "epoch": 3.9003134796238244, "grad_norm": 3.2095823287963867, "learning_rate": 1.5377061950651585e-08, "loss": 1.2548, "step": 49768 }, { "epoch": 3.9003918495297807, "grad_norm": 4.1372785568237305, "learning_rate": 1.5352906060175366e-08, "loss": 1.1529, "step": 49769 }, { "epoch": 3.9004702194357366, "grad_norm": 2.4152956008911133, "learning_rate": 1.532876912873682e-08, "loss": 1.2504, "step": 49770 }, { "epoch": 3.900548589341693, "grad_norm": 2.9259910583496094, "learning_rate": 1.530465115642643e-08, "loss": 1.7714, "step": 49771 }, { "epoch": 3.900626959247649, "grad_norm": 11.32056999206543, "learning_rate": 1.5280552143337456e-08, "loss": 1.3151, "step": 49772 }, { "epoch": 3.900705329153605, "grad_norm": 5.392462253570557, "learning_rate": 1.5256472089560935e-08, "loss": 1.2594, "step": 49773 }, { "epoch": 3.900783699059561, "grad_norm": 2.079296112060547, "learning_rate": 1.523241099518846e-08, "loss": 1.106, "step": 49774 }, { "epoch": 3.9008620689655173, "grad_norm": 4.625410079956055, "learning_rate": 1.5208368860311073e-08, "loss": 1.7597, "step": 49775 }, { "epoch": 3.9009404388714732, "grad_norm": 3.082141637802124, "learning_rate": 1.518434568502092e-08, "loss": 1.4111, "step": 49776 }, { "epoch": 3.9010188087774296, "grad_norm": 3.578505516052246, "learning_rate": 1.516034146940959e-08, "loss": 1.3535, "step": 49777 }, { "epoch": 3.9010971786833855, "grad_norm": 3.9006855487823486, "learning_rate": 1.513635621356757e-08, "loss": 1.1063, "step": 49778 }, { "epoch": 3.901175548589342, "grad_norm": 3.435332775115967, "learning_rate": 1.5112389917586456e-08, "loss": 1.149, "step": 49779 }, { "epoch": 3.9012539184952977, "grad_norm": 2.6735854148864746, "learning_rate": 1.508844258155728e-08, "loss": 1.2877, "step": 49780 }, { "epoch": 3.901332288401254, "grad_norm": 3.199632167816162, "learning_rate": 1.5064514205571644e-08, "loss": 1.4662, "step": 49781 }, { "epoch": 3.90141065830721, "grad_norm": 4.246260166168213, "learning_rate": 1.5040604789720026e-08, "loss": 1.2861, "step": 49782 }, { "epoch": 3.9014890282131662, "grad_norm": 5.415750026702881, "learning_rate": 1.5016714334093465e-08, "loss": 1.1511, "step": 49783 }, { "epoch": 3.901567398119122, "grad_norm": 2.40986704826355, "learning_rate": 1.4992842838782995e-08, "loss": 1.0972, "step": 49784 }, { "epoch": 3.9016457680250785, "grad_norm": 2.402482032775879, "learning_rate": 1.4968990303879105e-08, "loss": 1.2279, "step": 49785 }, { "epoch": 3.9017241379310343, "grad_norm": 4.506104469299316, "learning_rate": 1.494515672947283e-08, "loss": 1.5765, "step": 49786 }, { "epoch": 3.9018025078369907, "grad_norm": 5.863746643066406, "learning_rate": 1.4921342115654658e-08, "loss": 1.1529, "step": 49787 }, { "epoch": 3.9018808777429466, "grad_norm": 4.622487545013428, "learning_rate": 1.4897546462515066e-08, "loss": 0.9229, "step": 49788 }, { "epoch": 3.901959247648903, "grad_norm": 4.550757884979248, "learning_rate": 1.4873769770145097e-08, "loss": 1.4331, "step": 49789 }, { "epoch": 3.902037617554859, "grad_norm": 4.489893913269043, "learning_rate": 1.485001203863412e-08, "loss": 1.0914, "step": 49790 }, { "epoch": 3.902115987460815, "grad_norm": 2.417579412460327, "learning_rate": 1.4826273268073733e-08, "loss": 1.4857, "step": 49791 }, { "epoch": 3.902194357366771, "grad_norm": 3.791541814804077, "learning_rate": 1.4802553458553303e-08, "loss": 1.3433, "step": 49792 }, { "epoch": 3.9022727272727273, "grad_norm": 4.875981330871582, "learning_rate": 1.4778852610163875e-08, "loss": 1.0413, "step": 49793 }, { "epoch": 3.9023510971786832, "grad_norm": 3.4812493324279785, "learning_rate": 1.4755170722994816e-08, "loss": 1.2063, "step": 49794 }, { "epoch": 3.9024294670846396, "grad_norm": 2.3731863498687744, "learning_rate": 1.4731507797136057e-08, "loss": 1.1898, "step": 49795 }, { "epoch": 3.9025078369905954, "grad_norm": 5.2723517417907715, "learning_rate": 1.4707863832678637e-08, "loss": 1.974, "step": 49796 }, { "epoch": 3.902586206896552, "grad_norm": 3.4917774200439453, "learning_rate": 1.4684238829711928e-08, "loss": 1.2641, "step": 49797 }, { "epoch": 3.9026645768025077, "grad_norm": 2.746755361557007, "learning_rate": 1.4660632788325302e-08, "loss": 1.3662, "step": 49798 }, { "epoch": 3.902742946708464, "grad_norm": 5.513914585113525, "learning_rate": 1.4637045708609799e-08, "loss": 0.9447, "step": 49799 }, { "epoch": 3.9028213166144203, "grad_norm": 5.498738765716553, "learning_rate": 1.461347759065368e-08, "loss": 1.8628, "step": 49800 }, { "epoch": 3.902899686520376, "grad_norm": 5.008831977844238, "learning_rate": 1.4589928434547429e-08, "loss": 1.27, "step": 49801 }, { "epoch": 3.902978056426332, "grad_norm": 4.456832408905029, "learning_rate": 1.4566398240379865e-08, "loss": 1.4231, "step": 49802 }, { "epoch": 3.9030564263322884, "grad_norm": 9.831851959228516, "learning_rate": 1.4542887008241469e-08, "loss": 1.6594, "step": 49803 }, { "epoch": 3.9031347962382448, "grad_norm": 2.7387444972991943, "learning_rate": 1.4519394738221615e-08, "loss": 1.4977, "step": 49804 }, { "epoch": 3.9032131661442007, "grad_norm": 7.665611267089844, "learning_rate": 1.449592143040912e-08, "loss": 0.7608, "step": 49805 }, { "epoch": 3.9032915360501566, "grad_norm": 3.801332950592041, "learning_rate": 1.4472467084892806e-08, "loss": 1.2544, "step": 49806 }, { "epoch": 3.903369905956113, "grad_norm": 2.373589277267456, "learning_rate": 1.444903170176315e-08, "loss": 1.3747, "step": 49807 }, { "epoch": 3.903448275862069, "grad_norm": 4.898962497711182, "learning_rate": 1.4425615281108418e-08, "loss": 1.392, "step": 49808 }, { "epoch": 3.903526645768025, "grad_norm": 2.4651331901550293, "learning_rate": 1.4402217823017983e-08, "loss": 1.2196, "step": 49809 }, { "epoch": 3.903605015673981, "grad_norm": 2.6909985542297363, "learning_rate": 1.4378839327580663e-08, "loss": 1.4594, "step": 49810 }, { "epoch": 3.9036833855799373, "grad_norm": 3.906977891921997, "learning_rate": 1.4355479794885274e-08, "loss": 1.4049, "step": 49811 }, { "epoch": 3.9037617554858937, "grad_norm": 3.0524845123291016, "learning_rate": 1.4332139225021191e-08, "loss": 1.3965, "step": 49812 }, { "epoch": 3.9038401253918495, "grad_norm": 3.9218242168426514, "learning_rate": 1.4308817618076676e-08, "loss": 1.285, "step": 49813 }, { "epoch": 3.9039184952978054, "grad_norm": 3.6319050788879395, "learning_rate": 1.4285514974140546e-08, "loss": 1.3448, "step": 49814 }, { "epoch": 3.9039968652037618, "grad_norm": 3.3888187408447266, "learning_rate": 1.426223129330162e-08, "loss": 1.4846, "step": 49815 }, { "epoch": 3.904075235109718, "grad_norm": 6.2020487785339355, "learning_rate": 1.4238966575648716e-08, "loss": 1.5115, "step": 49816 }, { "epoch": 3.904153605015674, "grad_norm": 2.6557021141052246, "learning_rate": 1.4215720821268986e-08, "loss": 1.3927, "step": 49817 }, { "epoch": 3.90423197492163, "grad_norm": 2.105806350708008, "learning_rate": 1.4192494030252357e-08, "loss": 1.5136, "step": 49818 }, { "epoch": 3.904310344827586, "grad_norm": 1.8903450965881348, "learning_rate": 1.4169286202686538e-08, "loss": 1.3256, "step": 49819 }, { "epoch": 3.9043887147335425, "grad_norm": 3.2054481506347656, "learning_rate": 1.4146097338660347e-08, "loss": 1.3114, "step": 49820 }, { "epoch": 3.9044670846394984, "grad_norm": 5.4245734214782715, "learning_rate": 1.4122927438260937e-08, "loss": 1.6725, "step": 49821 }, { "epoch": 3.9045454545454543, "grad_norm": 5.359623908996582, "learning_rate": 1.4099776501577123e-08, "loss": 1.5706, "step": 49822 }, { "epoch": 3.9046238244514107, "grad_norm": 2.081005811691284, "learning_rate": 1.4076644528696614e-08, "loss": 1.245, "step": 49823 }, { "epoch": 3.904702194357367, "grad_norm": 2.770113229751587, "learning_rate": 1.4053531519707675e-08, "loss": 1.1371, "step": 49824 }, { "epoch": 3.904780564263323, "grad_norm": 3.7933380603790283, "learning_rate": 1.4030437474698566e-08, "loss": 1.3903, "step": 49825 }, { "epoch": 3.9048589341692788, "grad_norm": 6.201675891876221, "learning_rate": 1.4007362393756441e-08, "loss": 1.218, "step": 49826 }, { "epoch": 3.904937304075235, "grad_norm": 12.61429214477539, "learning_rate": 1.3984306276969561e-08, "loss": 1.5761, "step": 49827 }, { "epoch": 3.9050156739811914, "grad_norm": 3.2156174182891846, "learning_rate": 1.3961269124425082e-08, "loss": 1.3726, "step": 49828 }, { "epoch": 3.9050940438871473, "grad_norm": 4.861414909362793, "learning_rate": 1.3938250936211262e-08, "loss": 1.4612, "step": 49829 }, { "epoch": 3.905172413793103, "grad_norm": 3.4187917709350586, "learning_rate": 1.3915251712415256e-08, "loss": 1.4101, "step": 49830 }, { "epoch": 3.9052507836990595, "grad_norm": 3.175457000732422, "learning_rate": 1.3892271453124217e-08, "loss": 2.0612, "step": 49831 }, { "epoch": 3.905329153605016, "grad_norm": 4.34670352935791, "learning_rate": 1.3869310158426408e-08, "loss": 1.3115, "step": 49832 }, { "epoch": 3.9054075235109718, "grad_norm": 2.870175838470459, "learning_rate": 1.3846367828408425e-08, "loss": 1.4373, "step": 49833 }, { "epoch": 3.9054858934169276, "grad_norm": 5.228427410125732, "learning_rate": 1.3823444463157976e-08, "loss": 1.4839, "step": 49834 }, { "epoch": 3.905564263322884, "grad_norm": 2.1242876052856445, "learning_rate": 1.3800540062762213e-08, "loss": 1.2717, "step": 49835 }, { "epoch": 3.9056426332288403, "grad_norm": 3.115100622177124, "learning_rate": 1.3777654627307735e-08, "loss": 0.9064, "step": 49836 }, { "epoch": 3.905721003134796, "grad_norm": 2.946047782897949, "learning_rate": 1.3754788156882248e-08, "loss": 1.4949, "step": 49837 }, { "epoch": 3.905799373040752, "grad_norm": 11.420634269714355, "learning_rate": 1.3731940651572351e-08, "loss": 0.7207, "step": 49838 }, { "epoch": 3.9058777429467084, "grad_norm": 4.360384464263916, "learning_rate": 1.3709112111465195e-08, "loss": 1.4122, "step": 49839 }, { "epoch": 3.9059561128526648, "grad_norm": 6.107913494110107, "learning_rate": 1.3686302536647378e-08, "loss": 1.3699, "step": 49840 }, { "epoch": 3.9060344827586206, "grad_norm": 7.370406150817871, "learning_rate": 1.3663511927205498e-08, "loss": 2.1704, "step": 49841 }, { "epoch": 3.9061128526645765, "grad_norm": 3.044816017150879, "learning_rate": 1.3640740283226706e-08, "loss": 1.1066, "step": 49842 }, { "epoch": 3.906191222570533, "grad_norm": 2.3511664867401123, "learning_rate": 1.3617987604797045e-08, "loss": 1.2675, "step": 49843 }, { "epoch": 3.906269592476489, "grad_norm": 2.7717504501342773, "learning_rate": 1.3595253892004223e-08, "loss": 1.2603, "step": 49844 }, { "epoch": 3.906347962382445, "grad_norm": 5.220453262329102, "learning_rate": 1.3572539144933173e-08, "loss": 1.1172, "step": 49845 }, { "epoch": 3.9064263322884014, "grad_norm": 14.883749008178711, "learning_rate": 1.3549843363671045e-08, "loss": 0.4528, "step": 49846 }, { "epoch": 3.9065047021943573, "grad_norm": 4.584142684936523, "learning_rate": 1.3527166548304437e-08, "loss": 1.4154, "step": 49847 }, { "epoch": 3.9065830721003136, "grad_norm": 4.312626361846924, "learning_rate": 1.3504508698918838e-08, "loss": 1.5108, "step": 49848 }, { "epoch": 3.9066614420062695, "grad_norm": 3.8454511165618896, "learning_rate": 1.34818698156014e-08, "loss": 1.1966, "step": 49849 }, { "epoch": 3.906739811912226, "grad_norm": 3.424680471420288, "learning_rate": 1.3459249898437054e-08, "loss": 1.2861, "step": 49850 }, { "epoch": 3.9068181818181817, "grad_norm": 3.7467849254608154, "learning_rate": 1.3436648947512954e-08, "loss": 0.9985, "step": 49851 }, { "epoch": 3.906896551724138, "grad_norm": 4.123286724090576, "learning_rate": 1.3414066962914585e-08, "loss": 1.2646, "step": 49852 }, { "epoch": 3.906974921630094, "grad_norm": 2.4498648643493652, "learning_rate": 1.3391503944727435e-08, "loss": 1.3654, "step": 49853 }, { "epoch": 3.9070532915360503, "grad_norm": 4.306959629058838, "learning_rate": 1.3368959893038103e-08, "loss": 1.3816, "step": 49854 }, { "epoch": 3.907131661442006, "grad_norm": 3.347036838531494, "learning_rate": 1.334643480793152e-08, "loss": 1.3476, "step": 49855 }, { "epoch": 3.9072100313479625, "grad_norm": 4.07627010345459, "learning_rate": 1.3323928689494281e-08, "loss": 1.2409, "step": 49856 }, { "epoch": 3.9072884012539184, "grad_norm": 5.042397499084473, "learning_rate": 1.3301441537811322e-08, "loss": 1.6295, "step": 49857 }, { "epoch": 3.9073667711598747, "grad_norm": 3.772047996520996, "learning_rate": 1.3278973352968127e-08, "loss": 1.195, "step": 49858 }, { "epoch": 3.9074451410658306, "grad_norm": 2.7829647064208984, "learning_rate": 1.3256524135050741e-08, "loss": 1.4451, "step": 49859 }, { "epoch": 3.907523510971787, "grad_norm": 2.9219155311584473, "learning_rate": 1.3234093884144094e-08, "loss": 1.289, "step": 49860 }, { "epoch": 3.907601880877743, "grad_norm": 4.42796516418457, "learning_rate": 1.321168260033312e-08, "loss": 1.5358, "step": 49861 }, { "epoch": 3.907680250783699, "grad_norm": 3.731552839279175, "learning_rate": 1.318929028370386e-08, "loss": 1.3858, "step": 49862 }, { "epoch": 3.907758620689655, "grad_norm": 8.021552085876465, "learning_rate": 1.31669169343418e-08, "loss": 1.2305, "step": 49863 }, { "epoch": 3.9078369905956114, "grad_norm": 2.6482486724853516, "learning_rate": 1.3144562552330209e-08, "loss": 1.4352, "step": 49864 }, { "epoch": 3.9079153605015673, "grad_norm": 4.352051734924316, "learning_rate": 1.3122227137756238e-08, "loss": 1.4283, "step": 49865 }, { "epoch": 3.9079937304075236, "grad_norm": 4.365480899810791, "learning_rate": 1.3099910690703154e-08, "loss": 1.1303, "step": 49866 }, { "epoch": 3.9080721003134795, "grad_norm": 8.468649864196777, "learning_rate": 1.3077613211257556e-08, "loss": 1.3511, "step": 49867 }, { "epoch": 3.908150470219436, "grad_norm": 6.1125102043151855, "learning_rate": 1.3055334699502154e-08, "loss": 1.1757, "step": 49868 }, { "epoch": 3.9082288401253917, "grad_norm": 5.322100639343262, "learning_rate": 1.3033075155523545e-08, "loss": 1.3824, "step": 49869 }, { "epoch": 3.908307210031348, "grad_norm": 2.100613594055176, "learning_rate": 1.3010834579405552e-08, "loss": 1.4593, "step": 49870 }, { "epoch": 3.908385579937304, "grad_norm": 6.200863838195801, "learning_rate": 1.298861297123255e-08, "loss": 1.5099, "step": 49871 }, { "epoch": 3.9084639498432603, "grad_norm": 3.818030595779419, "learning_rate": 1.2966410331089474e-08, "loss": 1.603, "step": 49872 }, { "epoch": 3.908542319749216, "grad_norm": 4.032241344451904, "learning_rate": 1.2944226659060698e-08, "loss": 1.3278, "step": 49873 }, { "epoch": 3.9086206896551725, "grad_norm": 3.008485794067383, "learning_rate": 1.2922061955230603e-08, "loss": 1.3994, "step": 49874 }, { "epoch": 3.9086990595611284, "grad_norm": 2.7206881046295166, "learning_rate": 1.2899916219683561e-08, "loss": 1.4355, "step": 49875 }, { "epoch": 3.9087774294670847, "grad_norm": 9.397635459899902, "learning_rate": 1.2877789452503953e-08, "loss": 1.9, "step": 49876 }, { "epoch": 3.9088557993730406, "grad_norm": 5.8487725257873535, "learning_rate": 1.2855681653775044e-08, "loss": 1.3925, "step": 49877 }, { "epoch": 3.908934169278997, "grad_norm": 2.6386899948120117, "learning_rate": 1.2833592823581765e-08, "loss": 1.3147, "step": 49878 }, { "epoch": 3.909012539184953, "grad_norm": 9.15316390991211, "learning_rate": 1.2811522962008494e-08, "loss": 1.1202, "step": 49879 }, { "epoch": 3.909090909090909, "grad_norm": 3.59856915473938, "learning_rate": 1.2789472069137942e-08, "loss": 1.1338, "step": 49880 }, { "epoch": 3.909169278996865, "grad_norm": 2.689455509185791, "learning_rate": 1.2767440145055043e-08, "loss": 1.3745, "step": 49881 }, { "epoch": 3.9092476489028214, "grad_norm": 9.976573944091797, "learning_rate": 1.2745427189843062e-08, "loss": 1.2832, "step": 49882 }, { "epoch": 3.9093260188087773, "grad_norm": 3.2384250164031982, "learning_rate": 1.2723433203585267e-08, "loss": 1.5794, "step": 49883 }, { "epoch": 3.9094043887147336, "grad_norm": 2.985205888748169, "learning_rate": 1.2701458186366589e-08, "loss": 1.6122, "step": 49884 }, { "epoch": 3.9094827586206895, "grad_norm": 6.568634510040283, "learning_rate": 1.2679502138270294e-08, "loss": 1.433, "step": 49885 }, { "epoch": 3.909561128526646, "grad_norm": 2.6120572090148926, "learning_rate": 1.265756505937854e-08, "loss": 1.1094, "step": 49886 }, { "epoch": 3.9096394984326017, "grad_norm": 4.309414386749268, "learning_rate": 1.263564694977626e-08, "loss": 1.1352, "step": 49887 }, { "epoch": 3.909717868338558, "grad_norm": 3.169281244277954, "learning_rate": 1.2613747809546163e-08, "loss": 1.3595, "step": 49888 }, { "epoch": 3.909796238244514, "grad_norm": 4.0655598640441895, "learning_rate": 1.259186763877207e-08, "loss": 1.4157, "step": 49889 }, { "epoch": 3.9098746081504703, "grad_norm": 2.780209541320801, "learning_rate": 1.2570006437536697e-08, "loss": 1.3075, "step": 49890 }, { "epoch": 3.9099529780564266, "grad_norm": 7.467617034912109, "learning_rate": 1.2548164205922752e-08, "loss": 1.1709, "step": 49891 }, { "epoch": 3.9100313479623825, "grad_norm": 4.9993486404418945, "learning_rate": 1.2526340944014615e-08, "loss": 1.353, "step": 49892 }, { "epoch": 3.9101097178683384, "grad_norm": 2.526087999343872, "learning_rate": 1.2504536651893883e-08, "loss": 1.5398, "step": 49893 }, { "epoch": 3.9101880877742947, "grad_norm": 5.168509006500244, "learning_rate": 1.2482751329644937e-08, "loss": 1.5878, "step": 49894 }, { "epoch": 3.910266457680251, "grad_norm": 6.457770347595215, "learning_rate": 1.2460984977349378e-08, "loss": 1.1508, "step": 49895 }, { "epoch": 3.910344827586207, "grad_norm": 3.0833590030670166, "learning_rate": 1.2439237595090469e-08, "loss": 1.8507, "step": 49896 }, { "epoch": 3.910423197492163, "grad_norm": 3.547034978866577, "learning_rate": 1.2417509182950927e-08, "loss": 1.4267, "step": 49897 }, { "epoch": 3.910501567398119, "grad_norm": 2.8955156803131104, "learning_rate": 1.2395799741013459e-08, "loss": 1.3472, "step": 49898 }, { "epoch": 3.9105799373040755, "grad_norm": 4.341683387756348, "learning_rate": 1.2374109269360223e-08, "loss": 1.5282, "step": 49899 }, { "epoch": 3.9106583072100314, "grad_norm": 3.166935920715332, "learning_rate": 1.2352437768074487e-08, "loss": 1.4786, "step": 49900 }, { "epoch": 3.9107366771159873, "grad_norm": 7.13775110244751, "learning_rate": 1.2330785237238408e-08, "loss": 1.7352, "step": 49901 }, { "epoch": 3.9108150470219436, "grad_norm": 3.2294397354125977, "learning_rate": 1.2309151676933584e-08, "loss": 1.2886, "step": 49902 }, { "epoch": 3.9108934169279, "grad_norm": 4.131500720977783, "learning_rate": 1.2287537087243284e-08, "loss": 1.2933, "step": 49903 }, { "epoch": 3.910971786833856, "grad_norm": 2.8309648036956787, "learning_rate": 1.2265941468249665e-08, "loss": 1.4587, "step": 49904 }, { "epoch": 3.9110501567398117, "grad_norm": 2.6769230365753174, "learning_rate": 1.2244364820033771e-08, "loss": 1.2437, "step": 49905 }, { "epoch": 3.911128526645768, "grad_norm": 3.1494102478027344, "learning_rate": 1.2222807142678871e-08, "loss": 1.2643, "step": 49906 }, { "epoch": 3.9112068965517244, "grad_norm": 2.8862175941467285, "learning_rate": 1.2201268436266567e-08, "loss": 1.4452, "step": 49907 }, { "epoch": 3.9112852664576803, "grad_norm": 2.517235279083252, "learning_rate": 1.2179748700879013e-08, "loss": 1.3231, "step": 49908 }, { "epoch": 3.911363636363636, "grad_norm": 2.7828400135040283, "learning_rate": 1.2158247936597256e-08, "loss": 1.1103, "step": 49909 }, { "epoch": 3.9114420062695925, "grad_norm": 7.995550632476807, "learning_rate": 1.2136766143504009e-08, "loss": 1.3643, "step": 49910 }, { "epoch": 3.911520376175549, "grad_norm": 2.4886553287506104, "learning_rate": 1.2115303321680317e-08, "loss": 1.0274, "step": 49911 }, { "epoch": 3.9115987460815047, "grad_norm": 8.184554100036621, "learning_rate": 1.209385947120778e-08, "loss": 1.5684, "step": 49912 }, { "epoch": 3.9116771159874606, "grad_norm": 3.838520050048828, "learning_rate": 1.2072434592169113e-08, "loss": 1.3185, "step": 49913 }, { "epoch": 3.911755485893417, "grad_norm": 11.96142578125, "learning_rate": 1.205102868464425e-08, "loss": 0.9759, "step": 49914 }, { "epoch": 3.9118338557993733, "grad_norm": 2.0527873039245605, "learning_rate": 1.2029641748715349e-08, "loss": 1.199, "step": 49915 }, { "epoch": 3.911912225705329, "grad_norm": 4.593201637268066, "learning_rate": 1.2008273784464008e-08, "loss": 1.2834, "step": 49916 }, { "epoch": 3.911990595611285, "grad_norm": 2.5971736907958984, "learning_rate": 1.1986924791970722e-08, "loss": 0.8755, "step": 49917 }, { "epoch": 3.9120689655172414, "grad_norm": 3.5832901000976562, "learning_rate": 1.1965594771317646e-08, "loss": 1.3725, "step": 49918 }, { "epoch": 3.9121473354231977, "grad_norm": 3.5344040393829346, "learning_rate": 1.194428372258527e-08, "loss": 1.3224, "step": 49919 }, { "epoch": 3.9122257053291536, "grad_norm": 5.988212585449219, "learning_rate": 1.1922991645854642e-08, "loss": 1.1463, "step": 49920 }, { "epoch": 3.9123040752351095, "grad_norm": 4.965635776519775, "learning_rate": 1.1901718541207363e-08, "loss": 1.3876, "step": 49921 }, { "epoch": 3.912382445141066, "grad_norm": 3.060800313949585, "learning_rate": 1.188046440872337e-08, "loss": 1.6317, "step": 49922 }, { "epoch": 3.912460815047022, "grad_norm": 2.505622386932373, "learning_rate": 1.1859229248484815e-08, "loss": 1.0653, "step": 49923 }, { "epoch": 3.912539184952978, "grad_norm": 3.6267266273498535, "learning_rate": 1.1838013060571085e-08, "loss": 1.6047, "step": 49924 }, { "epoch": 3.912617554858934, "grad_norm": 2.618964672088623, "learning_rate": 1.1816815845063223e-08, "loss": 1.564, "step": 49925 }, { "epoch": 3.9126959247648903, "grad_norm": 3.0399160385131836, "learning_rate": 1.179563760204283e-08, "loss": 1.6398, "step": 49926 }, { "epoch": 3.9127742946708466, "grad_norm": 2.9765827655792236, "learning_rate": 1.177447833158929e-08, "loss": 1.1189, "step": 49927 }, { "epoch": 3.9128526645768025, "grad_norm": 3.0681631565093994, "learning_rate": 1.1753338033783646e-08, "loss": 1.3491, "step": 49928 }, { "epoch": 3.9129310344827584, "grad_norm": 3.8217294216156006, "learning_rate": 1.1732216708705834e-08, "loss": 1.3738, "step": 49929 }, { "epoch": 3.9130094043887147, "grad_norm": 3.9681198596954346, "learning_rate": 1.1711114356436903e-08, "loss": 1.3411, "step": 49930 }, { "epoch": 3.913087774294671, "grad_norm": 7.864488124847412, "learning_rate": 1.1690030977056232e-08, "loss": 1.2189, "step": 49931 }, { "epoch": 3.913166144200627, "grad_norm": 7.503934860229492, "learning_rate": 1.1668966570644868e-08, "loss": 1.4977, "step": 49932 }, { "epoch": 3.913244514106583, "grad_norm": 3.7425107955932617, "learning_rate": 1.1647921137282747e-08, "loss": 1.1935, "step": 49933 }, { "epoch": 3.913322884012539, "grad_norm": 2.7791128158569336, "learning_rate": 1.1626894677049804e-08, "loss": 1.1289, "step": 49934 }, { "epoch": 3.9134012539184955, "grad_norm": 3.3789188861846924, "learning_rate": 1.1605887190025977e-08, "loss": 1.1564, "step": 49935 }, { "epoch": 3.9134796238244514, "grad_norm": 2.974865436553955, "learning_rate": 1.1584898676290646e-08, "loss": 1.2862, "step": 49936 }, { "epoch": 3.9135579937304072, "grad_norm": 4.087845802307129, "learning_rate": 1.1563929135924301e-08, "loss": 1.9134, "step": 49937 }, { "epoch": 3.9136363636363636, "grad_norm": 7.653329372406006, "learning_rate": 1.154297856900688e-08, "loss": 1.8859, "step": 49938 }, { "epoch": 3.91371473354232, "grad_norm": 2.246955394744873, "learning_rate": 1.1522046975617762e-08, "loss": 1.2197, "step": 49939 }, { "epoch": 3.913793103448276, "grad_norm": 7.207261562347412, "learning_rate": 1.150113435583633e-08, "loss": 0.9391, "step": 49940 }, { "epoch": 3.913871473354232, "grad_norm": 4.01201868057251, "learning_rate": 1.1480240709742518e-08, "loss": 1.1845, "step": 49941 }, { "epoch": 3.913949843260188, "grad_norm": 2.6391501426696777, "learning_rate": 1.1459366037415154e-08, "loss": 1.2847, "step": 49942 }, { "epoch": 3.9140282131661444, "grad_norm": 2.4215493202209473, "learning_rate": 1.1438510338934728e-08, "loss": 1.3755, "step": 49943 }, { "epoch": 3.9141065830721002, "grad_norm": 2.5512032508850098, "learning_rate": 1.1417673614378955e-08, "loss": 1.2926, "step": 49944 }, { "epoch": 3.9141849529780566, "grad_norm": 2.5906338691711426, "learning_rate": 1.1396855863828882e-08, "loss": 1.325, "step": 49945 }, { "epoch": 3.9142633228840125, "grad_norm": 2.8908615112304688, "learning_rate": 1.137605708736278e-08, "loss": 1.0114, "step": 49946 }, { "epoch": 3.914341692789969, "grad_norm": 4.094944000244141, "learning_rate": 1.1355277285059474e-08, "loss": 1.6323, "step": 49947 }, { "epoch": 3.9144200626959247, "grad_norm": 4.28232479095459, "learning_rate": 1.13345164569989e-08, "loss": 1.0253, "step": 49948 }, { "epoch": 3.914498432601881, "grad_norm": 3.0854125022888184, "learning_rate": 1.131377460325933e-08, "loss": 1.4782, "step": 49949 }, { "epoch": 3.914576802507837, "grad_norm": 2.4631972312927246, "learning_rate": 1.129305172391959e-08, "loss": 1.1878, "step": 49950 }, { "epoch": 3.9146551724137932, "grad_norm": 15.929762840270996, "learning_rate": 1.1272347819058504e-08, "loss": 1.0885, "step": 49951 }, { "epoch": 3.914733542319749, "grad_norm": 3.239434242248535, "learning_rate": 1.1251662888755455e-08, "loss": 1.3086, "step": 49952 }, { "epoch": 3.9148119122257055, "grad_norm": 3.346851110458374, "learning_rate": 1.1230996933088712e-08, "loss": 1.5662, "step": 49953 }, { "epoch": 3.9148902821316613, "grad_norm": 3.792320489883423, "learning_rate": 1.1210349952136545e-08, "loss": 1.3542, "step": 49954 }, { "epoch": 3.9149686520376177, "grad_norm": 3.3578073978424072, "learning_rate": 1.1189721945977782e-08, "loss": 0.8623, "step": 49955 }, { "epoch": 3.9150470219435736, "grad_norm": 4.716850757598877, "learning_rate": 1.1169112914691249e-08, "loss": 1.0848, "step": 49956 }, { "epoch": 3.91512539184953, "grad_norm": 2.1912343502044678, "learning_rate": 1.114852285835466e-08, "loss": 1.6956, "step": 49957 }, { "epoch": 3.915203761755486, "grad_norm": 2.812342405319214, "learning_rate": 1.1127951777046841e-08, "loss": 1.7791, "step": 49958 }, { "epoch": 3.915282131661442, "grad_norm": 3.670860528945923, "learning_rate": 1.1107399670846064e-08, "loss": 1.1066, "step": 49959 }, { "epoch": 3.915360501567398, "grad_norm": 3.499002695083618, "learning_rate": 1.1086866539830044e-08, "loss": 1.3022, "step": 49960 }, { "epoch": 3.9154388714733543, "grad_norm": 2.848682165145874, "learning_rate": 1.106635238407705e-08, "loss": 1.4863, "step": 49961 }, { "epoch": 3.9155172413793102, "grad_norm": 3.9180121421813965, "learning_rate": 1.1045857203665356e-08, "loss": 1.5692, "step": 49962 }, { "epoch": 3.9155956112852666, "grad_norm": 2.42342209815979, "learning_rate": 1.1025380998672119e-08, "loss": 1.4575, "step": 49963 }, { "epoch": 3.9156739811912225, "grad_norm": 4.371068477630615, "learning_rate": 1.1004923769176723e-08, "loss": 1.261, "step": 49964 }, { "epoch": 3.915752351097179, "grad_norm": 2.6560990810394287, "learning_rate": 1.0984485515255772e-08, "loss": 1.5503, "step": 49965 }, { "epoch": 3.9158307210031347, "grad_norm": 3.369889974594116, "learning_rate": 1.0964066236986982e-08, "loss": 1.411, "step": 49966 }, { "epoch": 3.915909090909091, "grad_norm": 3.348623514175415, "learning_rate": 1.0943665934448621e-08, "loss": 1.686, "step": 49967 }, { "epoch": 3.915987460815047, "grad_norm": 5.079551696777344, "learning_rate": 1.0923284607717854e-08, "loss": 1.3375, "step": 49968 }, { "epoch": 3.9160658307210032, "grad_norm": 6.699124813079834, "learning_rate": 1.0902922256872394e-08, "loss": 0.9731, "step": 49969 }, { "epoch": 3.916144200626959, "grad_norm": 6.981579303741455, "learning_rate": 1.0882578881989957e-08, "loss": 1.8703, "step": 49970 }, { "epoch": 3.9162225705329154, "grad_norm": 3.803415536880493, "learning_rate": 1.0862254483147705e-08, "loss": 1.1834, "step": 49971 }, { "epoch": 3.9163009404388713, "grad_norm": 5.696609973907471, "learning_rate": 1.084194906042224e-08, "loss": 1.3406, "step": 49972 }, { "epoch": 3.9163793103448277, "grad_norm": 4.909786701202393, "learning_rate": 1.0821662613891838e-08, "loss": 1.4268, "step": 49973 }, { "epoch": 3.9164576802507836, "grad_norm": 3.208143949508667, "learning_rate": 1.0801395143633098e-08, "loss": 1.2103, "step": 49974 }, { "epoch": 3.91653605015674, "grad_norm": 3.723989963531494, "learning_rate": 1.0781146649723184e-08, "loss": 1.2541, "step": 49975 }, { "epoch": 3.916614420062696, "grad_norm": 3.6177992820739746, "learning_rate": 1.0760917132239256e-08, "loss": 1.239, "step": 49976 }, { "epoch": 3.916692789968652, "grad_norm": 3.526606321334839, "learning_rate": 1.0740706591258477e-08, "loss": 1.21, "step": 49977 }, { "epoch": 3.916771159874608, "grad_norm": 3.2342166900634766, "learning_rate": 1.0720515026856893e-08, "loss": 1.2484, "step": 49978 }, { "epoch": 3.9168495297805643, "grad_norm": 2.7833664417266846, "learning_rate": 1.0700342439112221e-08, "loss": 1.1957, "step": 49979 }, { "epoch": 3.91692789968652, "grad_norm": 4.746176719665527, "learning_rate": 1.0680188828100513e-08, "loss": 1.8209, "step": 49980 }, { "epoch": 3.9170062695924766, "grad_norm": 3.9129202365875244, "learning_rate": 1.0660054193898373e-08, "loss": 1.3481, "step": 49981 }, { "epoch": 3.9170846394984324, "grad_norm": 3.9131879806518555, "learning_rate": 1.0639938536583517e-08, "loss": 1.233, "step": 49982 }, { "epoch": 3.9171630094043888, "grad_norm": 2.9682507514953613, "learning_rate": 1.0619841856230884e-08, "loss": 1.302, "step": 49983 }, { "epoch": 3.9172413793103447, "grad_norm": 3.4989376068115234, "learning_rate": 1.059976415291819e-08, "loss": 1.3019, "step": 49984 }, { "epoch": 3.917319749216301, "grad_norm": 3.7119271755218506, "learning_rate": 1.0579705426720931e-08, "loss": 1.2862, "step": 49985 }, { "epoch": 3.9173981191222573, "grad_norm": 2.9464948177337646, "learning_rate": 1.0559665677715714e-08, "loss": 1.2104, "step": 49986 }, { "epoch": 3.917476489028213, "grad_norm": 2.7152087688446045, "learning_rate": 1.053964490597914e-08, "loss": 1.1731, "step": 49987 }, { "epoch": 3.917554858934169, "grad_norm": 7.586386680603027, "learning_rate": 1.0519643111586709e-08, "loss": 1.4615, "step": 49988 }, { "epoch": 3.9176332288401254, "grad_norm": 3.684157609939575, "learning_rate": 1.0499660294615022e-08, "loss": 1.0183, "step": 49989 }, { "epoch": 3.9177115987460818, "grad_norm": 4.3416428565979, "learning_rate": 1.0479696455139576e-08, "loss": 1.3581, "step": 49990 }, { "epoch": 3.9177899686520377, "grad_norm": 2.89961576461792, "learning_rate": 1.0459751593236977e-08, "loss": 1.6041, "step": 49991 }, { "epoch": 3.9178683385579935, "grad_norm": 2.966277599334717, "learning_rate": 1.0439825708982166e-08, "loss": 1.2819, "step": 49992 }, { "epoch": 3.91794670846395, "grad_norm": 4.287502288818359, "learning_rate": 1.0419918802451746e-08, "loss": 1.2996, "step": 49993 }, { "epoch": 3.918025078369906, "grad_norm": 3.1553432941436768, "learning_rate": 1.0400030873721212e-08, "loss": 0.8389, "step": 49994 }, { "epoch": 3.918103448275862, "grad_norm": 6.390680313110352, "learning_rate": 1.0380161922866061e-08, "loss": 1.5247, "step": 49995 }, { "epoch": 3.918181818181818, "grad_norm": 3.685648202896118, "learning_rate": 1.0360311949961788e-08, "loss": 1.2017, "step": 49996 }, { "epoch": 3.9182601880877743, "grad_norm": 2.9441537857055664, "learning_rate": 1.0340480955084442e-08, "loss": 1.345, "step": 49997 }, { "epoch": 3.9183385579937307, "grad_norm": 9.420741081237793, "learning_rate": 1.0320668938308965e-08, "loss": 1.6866, "step": 49998 }, { "epoch": 3.9184169278996865, "grad_norm": 3.2232935428619385, "learning_rate": 1.030087589971085e-08, "loss": 1.7063, "step": 49999 }, { "epoch": 3.9184952978056424, "grad_norm": 4.617632865905762, "learning_rate": 1.0281101839365038e-08, "loss": 1.2552, "step": 50000 }, { "epoch": 3.9185736677115988, "grad_norm": 5.223056316375732, "learning_rate": 1.0261346757347024e-08, "loss": 0.4908, "step": 50001 }, { "epoch": 3.918652037617555, "grad_norm": 4.323216438293457, "learning_rate": 1.0241610653732304e-08, "loss": 1.1938, "step": 50002 }, { "epoch": 3.918730407523511, "grad_norm": 3.5449302196502686, "learning_rate": 1.0221893528595261e-08, "loss": 1.2054, "step": 50003 }, { "epoch": 3.918808777429467, "grad_norm": 6.7166266441345215, "learning_rate": 1.0202195382011393e-08, "loss": 1.3125, "step": 50004 }, { "epoch": 3.918887147335423, "grad_norm": 2.8875062465667725, "learning_rate": 1.0182516214055083e-08, "loss": 1.2192, "step": 50005 }, { "epoch": 3.9189655172413795, "grad_norm": 3.363128662109375, "learning_rate": 1.0162856024801826e-08, "loss": 1.7081, "step": 50006 }, { "epoch": 3.9190438871473354, "grad_norm": 2.536562204360962, "learning_rate": 1.0143214814326008e-08, "loss": 1.4592, "step": 50007 }, { "epoch": 3.9191222570532913, "grad_norm": 2.860239267349243, "learning_rate": 1.0123592582703123e-08, "loss": 1.3141, "step": 50008 }, { "epoch": 3.9192006269592476, "grad_norm": 3.1414225101470947, "learning_rate": 1.0103989330006447e-08, "loss": 1.1695, "step": 50009 }, { "epoch": 3.919278996865204, "grad_norm": 2.6864571571350098, "learning_rate": 1.0084405056311475e-08, "loss": 1.0167, "step": 50010 }, { "epoch": 3.91935736677116, "grad_norm": 3.772831916809082, "learning_rate": 1.0064839761692036e-08, "loss": 0.9315, "step": 50011 }, { "epoch": 3.9194357366771158, "grad_norm": 2.9127187728881836, "learning_rate": 1.0045293446223624e-08, "loss": 1.2606, "step": 50012 }, { "epoch": 3.919514106583072, "grad_norm": 4.983388423919678, "learning_rate": 1.0025766109979519e-08, "loss": 1.5604, "step": 50013 }, { "epoch": 3.9195924764890284, "grad_norm": 3.6477348804473877, "learning_rate": 1.0006257753034099e-08, "loss": 1.3131, "step": 50014 }, { "epoch": 3.9196708463949843, "grad_norm": 3.550262212753296, "learning_rate": 9.986768375462308e-09, "loss": 1.3053, "step": 50015 }, { "epoch": 3.91974921630094, "grad_norm": 3.1281545162200928, "learning_rate": 9.967297977337976e-09, "loss": 1.1757, "step": 50016 }, { "epoch": 3.9198275862068965, "grad_norm": 2.627504587173462, "learning_rate": 9.947846558734376e-09, "loss": 1.2101, "step": 50017 }, { "epoch": 3.919905956112853, "grad_norm": 2.990079402923584, "learning_rate": 9.928414119726449e-09, "loss": 1.1642, "step": 50018 }, { "epoch": 3.9199843260188088, "grad_norm": 3.2229232788085938, "learning_rate": 9.909000660388024e-09, "loss": 1.0557, "step": 50019 }, { "epoch": 3.9200626959247646, "grad_norm": 7.394674301147461, "learning_rate": 9.889606180792378e-09, "loss": 1.5075, "step": 50020 }, { "epoch": 3.920141065830721, "grad_norm": 3.0788917541503906, "learning_rate": 9.870230681013337e-09, "loss": 1.2647, "step": 50021 }, { "epoch": 3.9202194357366773, "grad_norm": 4.135809898376465, "learning_rate": 9.85087416112529e-09, "loss": 1.4848, "step": 50022 }, { "epoch": 3.920297805642633, "grad_norm": 3.0026345252990723, "learning_rate": 9.831536621201509e-09, "loss": 1.2005, "step": 50023 }, { "epoch": 3.920376175548589, "grad_norm": 6.639952659606934, "learning_rate": 9.812218061314715e-09, "loss": 0.8558, "step": 50024 }, { "epoch": 3.9204545454545454, "grad_norm": 3.739595413208008, "learning_rate": 9.792918481539848e-09, "loss": 1.2111, "step": 50025 }, { "epoch": 3.9205329153605017, "grad_norm": 7.7712578773498535, "learning_rate": 9.773637881949628e-09, "loss": 1.6709, "step": 50026 }, { "epoch": 3.9206112852664576, "grad_norm": 3.977771282196045, "learning_rate": 9.754376262616772e-09, "loss": 1.4502, "step": 50027 }, { "epoch": 3.9206896551724135, "grad_norm": 3.33817195892334, "learning_rate": 9.735133623615666e-09, "loss": 1.4717, "step": 50028 }, { "epoch": 3.92076802507837, "grad_norm": 6.990048885345459, "learning_rate": 9.715909965018478e-09, "loss": 0.8918, "step": 50029 }, { "epoch": 3.920846394984326, "grad_norm": 3.1582729816436768, "learning_rate": 9.696705286899033e-09, "loss": 1.3817, "step": 50030 }, { "epoch": 3.920924764890282, "grad_norm": 12.220624923706055, "learning_rate": 9.677519589330053e-09, "loss": 0.5969, "step": 50031 }, { "epoch": 3.9210031347962384, "grad_norm": 6.2207746505737305, "learning_rate": 9.658352872384813e-09, "loss": 1.0138, "step": 50032 }, { "epoch": 3.9210815047021943, "grad_norm": 3.2041211128234863, "learning_rate": 9.639205136136032e-09, "loss": 1.1171, "step": 50033 }, { "epoch": 3.9211598746081506, "grad_norm": 3.067258358001709, "learning_rate": 9.620076380655874e-09, "loss": 1.4971, "step": 50034 }, { "epoch": 3.9212382445141065, "grad_norm": 2.527907371520996, "learning_rate": 9.600966606018725e-09, "loss": 1.1173, "step": 50035 }, { "epoch": 3.921316614420063, "grad_norm": 3.6397287845611572, "learning_rate": 9.581875812295637e-09, "loss": 1.5777, "step": 50036 }, { "epoch": 3.9213949843260187, "grad_norm": 7.922721862792969, "learning_rate": 9.562803999559888e-09, "loss": 0.8025, "step": 50037 }, { "epoch": 3.921473354231975, "grad_norm": 2.303201913833618, "learning_rate": 9.543751167884196e-09, "loss": 1.3127, "step": 50038 }, { "epoch": 3.921551724137931, "grad_norm": 2.681872844696045, "learning_rate": 9.52471731734017e-09, "loss": 1.0581, "step": 50039 }, { "epoch": 3.9216300940438873, "grad_norm": 5.405412673950195, "learning_rate": 9.505702448001642e-09, "loss": 1.5098, "step": 50040 }, { "epoch": 3.921708463949843, "grad_norm": 2.4115118980407715, "learning_rate": 9.486706559939663e-09, "loss": 1.2661, "step": 50041 }, { "epoch": 3.9217868338557995, "grad_norm": 3.3003296852111816, "learning_rate": 9.467729653226954e-09, "loss": 1.3613, "step": 50042 }, { "epoch": 3.9218652037617554, "grad_norm": 3.0836119651794434, "learning_rate": 9.448771727936234e-09, "loss": 1.2507, "step": 50043 }, { "epoch": 3.9219435736677117, "grad_norm": 4.194347381591797, "learning_rate": 9.429832784138559e-09, "loss": 1.6488, "step": 50044 }, { "epoch": 3.9220219435736676, "grad_norm": 3.74033784866333, "learning_rate": 9.410912821906649e-09, "loss": 1.5184, "step": 50045 }, { "epoch": 3.922100313479624, "grad_norm": 4.260354042053223, "learning_rate": 9.392011841312665e-09, "loss": 1.6413, "step": 50046 }, { "epoch": 3.92217868338558, "grad_norm": 16.07193946838379, "learning_rate": 9.373129842427664e-09, "loss": 1.601, "step": 50047 }, { "epoch": 3.922257053291536, "grad_norm": 2.415818929672241, "learning_rate": 9.35426682532381e-09, "loss": 1.1326, "step": 50048 }, { "epoch": 3.922335423197492, "grad_norm": 2.37595796585083, "learning_rate": 9.335422790073268e-09, "loss": 1.3614, "step": 50049 }, { "epoch": 3.9224137931034484, "grad_norm": 3.0620226860046387, "learning_rate": 9.316597736747091e-09, "loss": 1.6596, "step": 50050 }, { "epoch": 3.9224921630094043, "grad_norm": 3.8623528480529785, "learning_rate": 9.297791665417444e-09, "loss": 1.1712, "step": 50051 }, { "epoch": 3.9225705329153606, "grad_norm": 4.935037612915039, "learning_rate": 9.279004576155936e-09, "loss": 1.3931, "step": 50052 }, { "epoch": 3.9226489028213165, "grad_norm": 6.087672233581543, "learning_rate": 9.260236469033069e-09, "loss": 1.1638, "step": 50053 }, { "epoch": 3.922727272727273, "grad_norm": 3.417499303817749, "learning_rate": 9.241487344121559e-09, "loss": 1.4869, "step": 50054 }, { "epoch": 3.9228056426332287, "grad_norm": 3.8791377544403076, "learning_rate": 9.22275720149135e-09, "loss": 1.286, "step": 50055 }, { "epoch": 3.922884012539185, "grad_norm": 2.1903626918792725, "learning_rate": 9.204046041214609e-09, "loss": 0.3167, "step": 50056 }, { "epoch": 3.922962382445141, "grad_norm": 12.430785179138184, "learning_rate": 9.18535386336239e-09, "loss": 1.361, "step": 50057 }, { "epoch": 3.9230407523510973, "grad_norm": 4.618953227996826, "learning_rate": 9.16668066800519e-09, "loss": 1.2114, "step": 50058 }, { "epoch": 3.923119122257053, "grad_norm": 3.7837066650390625, "learning_rate": 9.14802645521462e-09, "loss": 1.1836, "step": 50059 }, { "epoch": 3.9231974921630095, "grad_norm": 6.248077869415283, "learning_rate": 9.129391225061734e-09, "loss": 2.0683, "step": 50060 }, { "epoch": 3.9232758620689654, "grad_norm": 5.983005046844482, "learning_rate": 9.110774977617031e-09, "loss": 1.5505, "step": 50061 }, { "epoch": 3.9233542319749217, "grad_norm": 3.2965970039367676, "learning_rate": 9.092177712951567e-09, "loss": 1.581, "step": 50062 }, { "epoch": 3.9234326018808776, "grad_norm": 5.845764636993408, "learning_rate": 9.073599431135838e-09, "loss": 1.3625, "step": 50063 }, { "epoch": 3.923510971786834, "grad_norm": 3.0598678588867188, "learning_rate": 9.0550401322409e-09, "loss": 1.2264, "step": 50064 }, { "epoch": 3.92358934169279, "grad_norm": 8.161443710327148, "learning_rate": 9.036499816337252e-09, "loss": 0.758, "step": 50065 }, { "epoch": 3.923667711598746, "grad_norm": 5.627660274505615, "learning_rate": 9.017978483494282e-09, "loss": 1.3056, "step": 50066 }, { "epoch": 3.923746081504702, "grad_norm": 3.171988010406494, "learning_rate": 8.999476133784157e-09, "loss": 1.4383, "step": 50067 }, { "epoch": 3.9238244514106584, "grad_norm": 3.1763179302215576, "learning_rate": 8.980992767276264e-09, "loss": 1.383, "step": 50068 }, { "epoch": 3.9239028213166143, "grad_norm": 3.3847169876098633, "learning_rate": 8.962528384041657e-09, "loss": 1.3057, "step": 50069 }, { "epoch": 3.9239811912225706, "grad_norm": 4.482233047485352, "learning_rate": 8.944082984149172e-09, "loss": 1.0847, "step": 50070 }, { "epoch": 3.9240595611285265, "grad_norm": 3.084925889968872, "learning_rate": 8.925656567670415e-09, "loss": 1.1798, "step": 50071 }, { "epoch": 3.924137931034483, "grad_norm": 4.498380661010742, "learning_rate": 8.907249134674778e-09, "loss": 1.2953, "step": 50072 }, { "epoch": 3.9242163009404387, "grad_norm": 3.28603196144104, "learning_rate": 8.888860685232204e-09, "loss": 1.288, "step": 50073 }, { "epoch": 3.924294670846395, "grad_norm": 2.5773074626922607, "learning_rate": 8.870491219412636e-09, "loss": 1.1923, "step": 50074 }, { "epoch": 3.924373040752351, "grad_norm": 2.3870480060577393, "learning_rate": 8.852140737286575e-09, "loss": 1.0898, "step": 50075 }, { "epoch": 3.9244514106583073, "grad_norm": 2.778808832168579, "learning_rate": 8.833809238922852e-09, "loss": 1.3986, "step": 50076 }, { "epoch": 3.924529780564263, "grad_norm": 2.7591850757598877, "learning_rate": 8.815496724391969e-09, "loss": 1.2703, "step": 50077 }, { "epoch": 3.9246081504702195, "grad_norm": 4.6621317863464355, "learning_rate": 8.797203193763316e-09, "loss": 1.2064, "step": 50078 }, { "epoch": 3.9246865203761754, "grad_norm": 4.09010124206543, "learning_rate": 8.778928647106833e-09, "loss": 1.208, "step": 50079 }, { "epoch": 3.9247648902821317, "grad_norm": 5.023749828338623, "learning_rate": 8.7606730844908e-09, "loss": 1.2664, "step": 50080 }, { "epoch": 3.924843260188088, "grad_norm": 3.343275785446167, "learning_rate": 8.742436505986273e-09, "loss": 1.1806, "step": 50081 }, { "epoch": 3.924921630094044, "grad_norm": 3.746678113937378, "learning_rate": 8.72421891166153e-09, "loss": 1.3405, "step": 50082 }, { "epoch": 3.925, "grad_norm": 5.2001752853393555, "learning_rate": 8.706020301586514e-09, "loss": 1.1535, "step": 50083 }, { "epoch": 3.925078369905956, "grad_norm": 3.0978035926818848, "learning_rate": 8.687840675829507e-09, "loss": 1.4896, "step": 50084 }, { "epoch": 3.9251567398119125, "grad_norm": 4.389133453369141, "learning_rate": 8.669680034460448e-09, "loss": 1.036, "step": 50085 }, { "epoch": 3.9252351097178684, "grad_norm": 2.8279430866241455, "learning_rate": 8.651538377548175e-09, "loss": 1.183, "step": 50086 }, { "epoch": 3.9253134796238243, "grad_norm": 3.274034023284912, "learning_rate": 8.633415705162074e-09, "loss": 1.2164, "step": 50087 }, { "epoch": 3.9253918495297806, "grad_norm": 3.215585947036743, "learning_rate": 8.615312017370426e-09, "loss": 0.5974, "step": 50088 }, { "epoch": 3.925470219435737, "grad_norm": 2.831458330154419, "learning_rate": 8.597227314242618e-09, "loss": 1.8217, "step": 50089 }, { "epoch": 3.925548589341693, "grad_norm": 3.1114275455474854, "learning_rate": 8.57916159584693e-09, "loss": 1.366, "step": 50090 }, { "epoch": 3.9256269592476487, "grad_norm": 8.686895370483398, "learning_rate": 8.561114862252751e-09, "loss": 1.3661, "step": 50091 }, { "epoch": 3.925705329153605, "grad_norm": 16.8101749420166, "learning_rate": 8.543087113527803e-09, "loss": 1.3253, "step": 50092 }, { "epoch": 3.9257836990595614, "grad_norm": 4.767504692077637, "learning_rate": 8.525078349741478e-09, "loss": 1.2278, "step": 50093 }, { "epoch": 3.9258620689655173, "grad_norm": 3.368610143661499, "learning_rate": 8.507088570962053e-09, "loss": 1.2204, "step": 50094 }, { "epoch": 3.925940438871473, "grad_norm": 5.691718578338623, "learning_rate": 8.489117777257804e-09, "loss": 1.2085, "step": 50095 }, { "epoch": 3.9260188087774295, "grad_norm": 2.6381022930145264, "learning_rate": 8.471165968697014e-09, "loss": 1.2222, "step": 50096 }, { "epoch": 3.926097178683386, "grad_norm": 3.5064494609832764, "learning_rate": 8.453233145347961e-09, "loss": 1.6097, "step": 50097 }, { "epoch": 3.9261755485893417, "grad_norm": 2.9716525077819824, "learning_rate": 8.435319307279477e-09, "loss": 1.1498, "step": 50098 }, { "epoch": 3.9262539184952976, "grad_norm": 2.53474760055542, "learning_rate": 8.417424454558732e-09, "loss": 1.5494, "step": 50099 }, { "epoch": 3.926332288401254, "grad_norm": 4.280346870422363, "learning_rate": 8.399548587254557e-09, "loss": 1.8843, "step": 50100 }, { "epoch": 3.9264106583072103, "grad_norm": 7.317886829376221, "learning_rate": 8.38169170543468e-09, "loss": 1.9943, "step": 50101 }, { "epoch": 3.926489028213166, "grad_norm": 3.2560808658599854, "learning_rate": 8.363853809166822e-09, "loss": 1.175, "step": 50102 }, { "epoch": 3.926567398119122, "grad_norm": 5.959648132324219, "learning_rate": 8.346034898519263e-09, "loss": 1.4072, "step": 50103 }, { "epoch": 3.9266457680250784, "grad_norm": 3.9974234104156494, "learning_rate": 8.328234973559169e-09, "loss": 1.369, "step": 50104 }, { "epoch": 3.9267241379310347, "grad_norm": 2.4648289680480957, "learning_rate": 8.31045403435482e-09, "loss": 1.309, "step": 50105 }, { "epoch": 3.9268025078369906, "grad_norm": 3.25103497505188, "learning_rate": 8.292692080973386e-09, "loss": 1.0289, "step": 50106 }, { "epoch": 3.9268808777429465, "grad_norm": 4.85736083984375, "learning_rate": 8.274949113483143e-09, "loss": 0.9783, "step": 50107 }, { "epoch": 3.926959247648903, "grad_norm": 2.440624713897705, "learning_rate": 8.257225131950708e-09, "loss": 1.5187, "step": 50108 }, { "epoch": 3.927037617554859, "grad_norm": 2.4121999740600586, "learning_rate": 8.239520136443802e-09, "loss": 1.1874, "step": 50109 }, { "epoch": 3.927115987460815, "grad_norm": 2.8627262115478516, "learning_rate": 8.221834127029593e-09, "loss": 1.5568, "step": 50110 }, { "epoch": 3.927194357366771, "grad_norm": 3.643531560897827, "learning_rate": 8.204167103776361e-09, "loss": 1.4265, "step": 50111 }, { "epoch": 3.9272727272727272, "grad_norm": 3.110743761062622, "learning_rate": 8.18651906674961e-09, "loss": 1.0428, "step": 50112 }, { "epoch": 3.9273510971786836, "grad_norm": 3.065354585647583, "learning_rate": 8.168890016018172e-09, "loss": 1.559, "step": 50113 }, { "epoch": 3.9274294670846395, "grad_norm": 9.646099090576172, "learning_rate": 8.151279951648105e-09, "loss": 1.9188, "step": 50114 }, { "epoch": 3.9275078369905954, "grad_norm": 2.9245173931121826, "learning_rate": 8.133688873706025e-09, "loss": 1.5939, "step": 50115 }, { "epoch": 3.9275862068965517, "grad_norm": 2.3575832843780518, "learning_rate": 8.116116782260209e-09, "loss": 1.3011, "step": 50116 }, { "epoch": 3.927664576802508, "grad_norm": 5.11619758605957, "learning_rate": 8.098563677376159e-09, "loss": 1.562, "step": 50117 }, { "epoch": 3.927742946708464, "grad_norm": 3.5473101139068604, "learning_rate": 8.0810295591216e-09, "loss": 1.4345, "step": 50118 }, { "epoch": 3.92782131661442, "grad_norm": 5.374744415283203, "learning_rate": 8.063514427562036e-09, "loss": 1.5289, "step": 50119 }, { "epoch": 3.927899686520376, "grad_norm": 4.103349685668945, "learning_rate": 8.046018282765744e-09, "loss": 1.4322, "step": 50120 }, { "epoch": 3.9279780564263325, "grad_norm": 8.271709442138672, "learning_rate": 8.028541124798228e-09, "loss": 0.9057, "step": 50121 }, { "epoch": 3.9280564263322884, "grad_norm": 3.0513594150543213, "learning_rate": 8.011082953725546e-09, "loss": 1.5183, "step": 50122 }, { "epoch": 3.9281347962382442, "grad_norm": 10.780214309692383, "learning_rate": 7.993643769614868e-09, "loss": 1.5699, "step": 50123 }, { "epoch": 3.9282131661442006, "grad_norm": 2.5153415203094482, "learning_rate": 7.976223572532248e-09, "loss": 1.4323, "step": 50124 }, { "epoch": 3.928291536050157, "grad_norm": 3.911564588546753, "learning_rate": 7.958822362544306e-09, "loss": 0.8444, "step": 50125 }, { "epoch": 3.928369905956113, "grad_norm": 5.656668663024902, "learning_rate": 7.941440139717093e-09, "loss": 1.1597, "step": 50126 }, { "epoch": 3.928448275862069, "grad_norm": 8.244839668273926, "learning_rate": 7.924076904116119e-09, "loss": 1.3327, "step": 50127 }, { "epoch": 3.928526645768025, "grad_norm": 5.851413249969482, "learning_rate": 7.906732655807992e-09, "loss": 1.7048, "step": 50128 }, { "epoch": 3.9286050156739813, "grad_norm": 3.6216471195220947, "learning_rate": 7.889407394858217e-09, "loss": 1.3151, "step": 50129 }, { "epoch": 3.9286833855799372, "grad_norm": 4.093832492828369, "learning_rate": 7.872101121333408e-09, "loss": 1.6139, "step": 50130 }, { "epoch": 3.9287617554858936, "grad_norm": 3.5896971225738525, "learning_rate": 7.854813835299069e-09, "loss": 1.1887, "step": 50131 }, { "epoch": 3.9288401253918495, "grad_norm": 2.955404043197632, "learning_rate": 7.8375455368207e-09, "loss": 1.5227, "step": 50132 }, { "epoch": 3.928918495297806, "grad_norm": 2.998863697052002, "learning_rate": 7.820296225964364e-09, "loss": 1.3002, "step": 50133 }, { "epoch": 3.9289968652037617, "grad_norm": 5.784196853637695, "learning_rate": 7.80306590279556e-09, "loss": 1.7149, "step": 50134 }, { "epoch": 3.929075235109718, "grad_norm": 5.8803324699401855, "learning_rate": 7.785854567379791e-09, "loss": 1.1785, "step": 50135 }, { "epoch": 3.929153605015674, "grad_norm": 4.068786144256592, "learning_rate": 7.768662219782008e-09, "loss": 1.1188, "step": 50136 }, { "epoch": 3.9292319749216302, "grad_norm": 7.42863655090332, "learning_rate": 7.751488860068268e-09, "loss": 1.0768, "step": 50137 }, { "epoch": 3.929310344827586, "grad_norm": 4.638107776641846, "learning_rate": 7.734334488303518e-09, "loss": 1.5878, "step": 50138 }, { "epoch": 3.9293887147335425, "grad_norm": 2.821849822998047, "learning_rate": 7.717199104553263e-09, "loss": 1.3789, "step": 50139 }, { "epoch": 3.9294670846394983, "grad_norm": 3.731315851211548, "learning_rate": 7.700082708883006e-09, "loss": 1.1323, "step": 50140 }, { "epoch": 3.9295454545454547, "grad_norm": 4.655740737915039, "learning_rate": 7.682985301357137e-09, "loss": 1.5384, "step": 50141 }, { "epoch": 3.9296238244514106, "grad_norm": 3.24989914894104, "learning_rate": 7.665906882040608e-09, "loss": 0.9484, "step": 50142 }, { "epoch": 3.929702194357367, "grad_norm": 2.6093966960906982, "learning_rate": 7.648847450999475e-09, "loss": 1.5957, "step": 50143 }, { "epoch": 3.929780564263323, "grad_norm": 3.5354573726654053, "learning_rate": 7.63180700829702e-09, "loss": 1.6191, "step": 50144 }, { "epoch": 3.929858934169279, "grad_norm": 4.4915337562561035, "learning_rate": 7.61478555399986e-09, "loss": 1.2351, "step": 50145 }, { "epoch": 3.929937304075235, "grad_norm": 3.9996399879455566, "learning_rate": 7.597783088170718e-09, "loss": 1.4609, "step": 50146 }, { "epoch": 3.9300156739811913, "grad_norm": 3.6868486404418945, "learning_rate": 7.58079961087621e-09, "loss": 1.1183, "step": 50147 }, { "epoch": 3.9300940438871472, "grad_norm": 3.440620183944702, "learning_rate": 7.563835122180174e-09, "loss": 0.8592, "step": 50148 }, { "epoch": 3.9301724137931036, "grad_norm": 7.494717121124268, "learning_rate": 7.546889622146447e-09, "loss": 1.2101, "step": 50149 }, { "epoch": 3.9302507836990594, "grad_norm": 3.3688161373138428, "learning_rate": 7.529963110840533e-09, "loss": 1.097, "step": 50150 }, { "epoch": 3.930329153605016, "grad_norm": 4.3564863204956055, "learning_rate": 7.513055588325713e-09, "loss": 1.1908, "step": 50151 }, { "epoch": 3.9304075235109717, "grad_norm": 6.957106590270996, "learning_rate": 7.496167054667492e-09, "loss": 1.9431, "step": 50152 }, { "epoch": 3.930485893416928, "grad_norm": 3.9323809146881104, "learning_rate": 7.479297509929151e-09, "loss": 1.0738, "step": 50153 }, { "epoch": 3.930564263322884, "grad_norm": 3.5412631034851074, "learning_rate": 7.46244695417564e-09, "loss": 1.1609, "step": 50154 }, { "epoch": 3.93064263322884, "grad_norm": 3.0763838291168213, "learning_rate": 7.4456153874702394e-09, "loss": 1.3209, "step": 50155 }, { "epoch": 3.930721003134796, "grad_norm": 2.9374947547912598, "learning_rate": 7.4288028098779e-09, "loss": 1.1085, "step": 50156 }, { "epoch": 3.9307993730407524, "grad_norm": 4.359892845153809, "learning_rate": 7.4120092214619025e-09, "loss": 2.0308, "step": 50157 }, { "epoch": 3.9308777429467083, "grad_norm": 6.7626495361328125, "learning_rate": 7.395234622286085e-09, "loss": 1.187, "step": 50158 }, { "epoch": 3.9309561128526647, "grad_norm": 10.594762802124023, "learning_rate": 7.378479012414286e-09, "loss": 1.2902, "step": 50159 }, { "epoch": 3.9310344827586206, "grad_norm": 3.36352276802063, "learning_rate": 7.361742391910898e-09, "loss": 1.5751, "step": 50160 }, { "epoch": 3.931112852664577, "grad_norm": 2.4863805770874023, "learning_rate": 7.345024760838648e-09, "loss": 0.8814, "step": 50161 }, { "epoch": 3.9311912225705328, "grad_norm": 3.84572696685791, "learning_rate": 7.3283261192613754e-09, "loss": 1.2933, "step": 50162 }, { "epoch": 3.931269592476489, "grad_norm": 2.791937828063965, "learning_rate": 7.311646467243472e-09, "loss": 1.4178, "step": 50163 }, { "epoch": 3.931347962382445, "grad_norm": 2.9219353199005127, "learning_rate": 7.29498580484711e-09, "loss": 1.4345, "step": 50164 }, { "epoch": 3.9314263322884013, "grad_norm": 3.905940055847168, "learning_rate": 7.2783441321361285e-09, "loss": 0.577, "step": 50165 }, { "epoch": 3.931504702194357, "grad_norm": 3.890742540359497, "learning_rate": 7.261721449173808e-09, "loss": 1.3869, "step": 50166 }, { "epoch": 3.9315830721003135, "grad_norm": 3.4407103061676025, "learning_rate": 7.245117756023989e-09, "loss": 1.1731, "step": 50167 }, { "epoch": 3.9316614420062694, "grad_norm": 2.41318941116333, "learning_rate": 7.2285330527488426e-09, "loss": 1.47, "step": 50168 }, { "epoch": 3.9317398119122258, "grad_norm": 2.871328115463257, "learning_rate": 7.211967339412207e-09, "loss": 1.1468, "step": 50169 }, { "epoch": 3.9318181818181817, "grad_norm": 13.225247383117676, "learning_rate": 7.1954206160768096e-09, "loss": 1.3039, "step": 50170 }, { "epoch": 3.931896551724138, "grad_norm": 4.791372299194336, "learning_rate": 7.1788928828053775e-09, "loss": 1.1401, "step": 50171 }, { "epoch": 3.931974921630094, "grad_norm": 2.5662543773651123, "learning_rate": 7.1623841396606385e-09, "loss": 1.2913, "step": 50172 }, { "epoch": 3.93205329153605, "grad_norm": 8.543185234069824, "learning_rate": 7.145894386706431e-09, "loss": 1.1097, "step": 50173 }, { "epoch": 3.932131661442006, "grad_norm": 9.522089958190918, "learning_rate": 7.129423624004373e-09, "loss": 1.4958, "step": 50174 }, { "epoch": 3.9322100313479624, "grad_norm": 3.484302520751953, "learning_rate": 7.11297185161719e-09, "loss": 1.4702, "step": 50175 }, { "epoch": 3.9322884012539188, "grad_norm": 2.120378255844116, "learning_rate": 7.096539069608721e-09, "loss": 1.1779, "step": 50176 }, { "epoch": 3.9323667711598747, "grad_norm": 2.581716537475586, "learning_rate": 7.080125278039474e-09, "loss": 1.0095, "step": 50177 }, { "epoch": 3.9324451410658305, "grad_norm": 3.3246405124664307, "learning_rate": 7.063730476973285e-09, "loss": 1.4072, "step": 50178 }, { "epoch": 3.932523510971787, "grad_norm": 3.159667730331421, "learning_rate": 7.047354666472328e-09, "loss": 1.8572, "step": 50179 }, { "epoch": 3.932601880877743, "grad_norm": 5.404271125793457, "learning_rate": 7.03099784659822e-09, "loss": 1.2489, "step": 50180 }, { "epoch": 3.932680250783699, "grad_norm": 4.83538818359375, "learning_rate": 7.014660017414243e-09, "loss": 1.186, "step": 50181 }, { "epoch": 3.932758620689655, "grad_norm": 2.7486722469329834, "learning_rate": 6.99834117898146e-09, "loss": 1.2291, "step": 50182 }, { "epoch": 3.9328369905956113, "grad_norm": 3.644045352935791, "learning_rate": 6.982041331362599e-09, "loss": 1.1377, "step": 50183 }, { "epoch": 3.9329153605015676, "grad_norm": 3.619597911834717, "learning_rate": 6.965760474619276e-09, "loss": 1.3734, "step": 50184 }, { "epoch": 3.9329937304075235, "grad_norm": 3.1224112510681152, "learning_rate": 6.94949860881422e-09, "loss": 1.598, "step": 50185 }, { "epoch": 3.9330721003134794, "grad_norm": 4.27455472946167, "learning_rate": 6.933255734007937e-09, "loss": 1.1718, "step": 50186 }, { "epoch": 3.9331504702194358, "grad_norm": 3.443547487258911, "learning_rate": 6.917031850263156e-09, "loss": 0.9959, "step": 50187 }, { "epoch": 3.933228840125392, "grad_norm": 7.031113624572754, "learning_rate": 6.900826957641493e-09, "loss": 1.0402, "step": 50188 }, { "epoch": 3.933307210031348, "grad_norm": 1.8863047361373901, "learning_rate": 6.884641056204567e-09, "loss": 1.1876, "step": 50189 }, { "epoch": 3.933385579937304, "grad_norm": 3.96101713180542, "learning_rate": 6.8684741460139924e-09, "loss": 0.9225, "step": 50190 }, { "epoch": 3.93346394984326, "grad_norm": 2.5808310508728027, "learning_rate": 6.852326227130835e-09, "loss": 1.2599, "step": 50191 }, { "epoch": 3.9335423197492165, "grad_norm": 3.893084764480591, "learning_rate": 6.836197299616709e-09, "loss": 1.4309, "step": 50192 }, { "epoch": 3.9336206896551724, "grad_norm": 10.446521759033203, "learning_rate": 6.8200873635332346e-09, "loss": 1.0947, "step": 50193 }, { "epoch": 3.9336990595611283, "grad_norm": 5.378810405731201, "learning_rate": 6.803996418940917e-09, "loss": 2.778, "step": 50194 }, { "epoch": 3.9337774294670846, "grad_norm": 5.058757781982422, "learning_rate": 6.7879244659024844e-09, "loss": 1.6537, "step": 50195 }, { "epoch": 3.933855799373041, "grad_norm": 3.416574716567993, "learning_rate": 6.771871504477334e-09, "loss": 1.3654, "step": 50196 }, { "epoch": 3.933934169278997, "grad_norm": 3.357656478881836, "learning_rate": 6.755837534727638e-09, "loss": 1.134, "step": 50197 }, { "epoch": 3.9340125391849528, "grad_norm": 4.840900421142578, "learning_rate": 6.739822556713904e-09, "loss": 1.7951, "step": 50198 }, { "epoch": 3.934090909090909, "grad_norm": 11.19782829284668, "learning_rate": 6.723826570497194e-09, "loss": 1.2599, "step": 50199 }, { "epoch": 3.9341692789968654, "grad_norm": 3.3801491260528564, "learning_rate": 6.7078495761385695e-09, "loss": 1.1581, "step": 50200 }, { "epoch": 3.9342476489028213, "grad_norm": 2.3082776069641113, "learning_rate": 6.691891573698539e-09, "loss": 1.0403, "step": 50201 }, { "epoch": 3.934326018808777, "grad_norm": 3.016002893447876, "learning_rate": 6.675952563237609e-09, "loss": 1.3886, "step": 50202 }, { "epoch": 3.9344043887147335, "grad_norm": 5.046298503875732, "learning_rate": 6.660032544816841e-09, "loss": 1.6098, "step": 50203 }, { "epoch": 3.93448275862069, "grad_norm": 3.4191431999206543, "learning_rate": 6.644131518496188e-09, "loss": 1.0778, "step": 50204 }, { "epoch": 3.9345611285266457, "grad_norm": 5.580288410186768, "learning_rate": 6.628249484336713e-09, "loss": 1.5077, "step": 50205 }, { "epoch": 3.9346394984326016, "grad_norm": 5.127179145812988, "learning_rate": 6.612386442398921e-09, "loss": 1.2046, "step": 50206 }, { "epoch": 3.934717868338558, "grad_norm": 2.8185336589813232, "learning_rate": 6.5965423927427665e-09, "loss": 0.9561, "step": 50207 }, { "epoch": 3.9347962382445143, "grad_norm": 2.5181777477264404, "learning_rate": 6.5807173354282e-09, "loss": 1.3011, "step": 50208 }, { "epoch": 3.93487460815047, "grad_norm": 10.655485153198242, "learning_rate": 6.564911270516283e-09, "loss": 1.3174, "step": 50209 }, { "epoch": 3.934952978056426, "grad_norm": 4.535356521606445, "learning_rate": 6.549124198066414e-09, "loss": 1.5559, "step": 50210 }, { "epoch": 3.9350313479623824, "grad_norm": 6.944617748260498, "learning_rate": 6.533356118138545e-09, "loss": 1.765, "step": 50211 }, { "epoch": 3.9351097178683387, "grad_norm": 11.80594539642334, "learning_rate": 6.517607030793183e-09, "loss": 0.9906, "step": 50212 }, { "epoch": 3.9351880877742946, "grad_norm": 2.5794036388397217, "learning_rate": 6.501876936090279e-09, "loss": 1.4896, "step": 50213 }, { "epoch": 3.9352664576802505, "grad_norm": 4.312509536743164, "learning_rate": 6.486165834089231e-09, "loss": 1.3536, "step": 50214 }, { "epoch": 3.935344827586207, "grad_norm": 2.8085007667541504, "learning_rate": 6.47047372484999e-09, "loss": 1.1319, "step": 50215 }, { "epoch": 3.935423197492163, "grad_norm": 13.736563682556152, "learning_rate": 6.45480060843251e-09, "loss": 1.5485, "step": 50216 }, { "epoch": 3.935501567398119, "grad_norm": 13.21468734741211, "learning_rate": 6.439146484895631e-09, "loss": 1.5904, "step": 50217 }, { "epoch": 3.935579937304075, "grad_norm": 3.1585581302642822, "learning_rate": 6.423511354299861e-09, "loss": 1.1684, "step": 50218 }, { "epoch": 3.9356583072100313, "grad_norm": 2.0813863277435303, "learning_rate": 6.4078952167040405e-09, "loss": 1.379, "step": 50219 }, { "epoch": 3.9357366771159876, "grad_norm": 2.567047119140625, "learning_rate": 6.392298072167569e-09, "loss": 1.3735, "step": 50220 }, { "epoch": 3.9358150470219435, "grad_norm": 2.6177027225494385, "learning_rate": 6.376719920749841e-09, "loss": 1.1918, "step": 50221 }, { "epoch": 3.9358934169279, "grad_norm": 4.561119556427002, "learning_rate": 6.3611607625102546e-09, "loss": 1.368, "step": 50222 }, { "epoch": 3.9359717868338557, "grad_norm": 5.2431135177612305, "learning_rate": 6.345620597508206e-09, "loss": 1.056, "step": 50223 }, { "epoch": 3.936050156739812, "grad_norm": 3.474302291870117, "learning_rate": 6.330099425801983e-09, "loss": 1.0725, "step": 50224 }, { "epoch": 3.936128526645768, "grad_norm": 4.364732265472412, "learning_rate": 6.314597247451537e-09, "loss": 1.5951, "step": 50225 }, { "epoch": 3.9362068965517243, "grad_norm": 3.019679307937622, "learning_rate": 6.29911406251571e-09, "loss": 1.4009, "step": 50226 }, { "epoch": 3.93628526645768, "grad_norm": 4.306815147399902, "learning_rate": 6.283649871052788e-09, "loss": 1.5587, "step": 50227 }, { "epoch": 3.9363636363636365, "grad_norm": 4.069976806640625, "learning_rate": 6.268204673121614e-09, "loss": 0.9175, "step": 50228 }, { "epoch": 3.9364420062695924, "grad_norm": 5.43460750579834, "learning_rate": 6.252778468781584e-09, "loss": 1.2571, "step": 50229 }, { "epoch": 3.9365203761755487, "grad_norm": 2.617764949798584, "learning_rate": 6.237371258090985e-09, "loss": 1.5319, "step": 50230 }, { "epoch": 3.9365987460815046, "grad_norm": 3.628586769104004, "learning_rate": 6.221983041108659e-09, "loss": 1.6754, "step": 50231 }, { "epoch": 3.936677115987461, "grad_norm": 5.088413715362549, "learning_rate": 6.206613817892893e-09, "loss": 1.0043, "step": 50232 }, { "epoch": 3.936755485893417, "grad_norm": 7.5710320472717285, "learning_rate": 6.191263588501972e-09, "loss": 0.9215, "step": 50233 }, { "epoch": 3.936833855799373, "grad_norm": 4.116804599761963, "learning_rate": 6.1759323529947405e-09, "loss": 1.3438, "step": 50234 }, { "epoch": 3.936912225705329, "grad_norm": 13.591652870178223, "learning_rate": 6.160620111428928e-09, "loss": 1.4385, "step": 50235 }, { "epoch": 3.9369905956112854, "grad_norm": 3.338895559310913, "learning_rate": 6.145326863863376e-09, "loss": 1.1671, "step": 50236 }, { "epoch": 3.9370689655172413, "grad_norm": 5.541814804077148, "learning_rate": 6.1300526103558186e-09, "loss": 1.6183, "step": 50237 }, { "epoch": 3.9371473354231976, "grad_norm": 2.56628155708313, "learning_rate": 6.114797350964541e-09, "loss": 1.1922, "step": 50238 }, { "epoch": 3.9372257053291535, "grad_norm": 2.808112859725952, "learning_rate": 6.099561085747829e-09, "loss": 0.9674, "step": 50239 }, { "epoch": 3.93730407523511, "grad_norm": 4.192749500274658, "learning_rate": 6.084343814763416e-09, "loss": 1.1789, "step": 50240 }, { "epoch": 3.9373824451410657, "grad_norm": 3.131244659423828, "learning_rate": 6.069145538068477e-09, "loss": 0.6329, "step": 50241 }, { "epoch": 3.937460815047022, "grad_norm": 4.113373279571533, "learning_rate": 6.053966255721855e-09, "loss": 1.3182, "step": 50242 }, { "epoch": 3.937539184952978, "grad_norm": 2.997035264968872, "learning_rate": 6.03880596778128e-09, "loss": 1.3489, "step": 50243 }, { "epoch": 3.9376175548589343, "grad_norm": 2.700960397720337, "learning_rate": 6.023664674303376e-09, "loss": 1.3253, "step": 50244 }, { "epoch": 3.93769592476489, "grad_norm": 2.9008567333221436, "learning_rate": 6.008542375346982e-09, "loss": 1.212, "step": 50245 }, { "epoch": 3.9377742946708465, "grad_norm": 5.585713863372803, "learning_rate": 5.9934390709687204e-09, "loss": 0.8592, "step": 50246 }, { "epoch": 3.9378526645768024, "grad_norm": 3.772062063217163, "learning_rate": 5.978354761226324e-09, "loss": 1.4077, "step": 50247 }, { "epoch": 3.9379310344827587, "grad_norm": 3.119670867919922, "learning_rate": 5.963289446176967e-09, "loss": 1.4344, "step": 50248 }, { "epoch": 3.9380094043887146, "grad_norm": 2.878711700439453, "learning_rate": 5.948243125878384e-09, "loss": 1.377, "step": 50249 }, { "epoch": 3.938087774294671, "grad_norm": 7.821312427520752, "learning_rate": 5.9332158003877486e-09, "loss": 1.593, "step": 50250 }, { "epoch": 3.938166144200627, "grad_norm": 9.601811408996582, "learning_rate": 5.9182074697616836e-09, "loss": 1.6337, "step": 50251 }, { "epoch": 3.938244514106583, "grad_norm": 2.7887816429138184, "learning_rate": 5.90321813405792e-09, "loss": 1.2811, "step": 50252 }, { "epoch": 3.938322884012539, "grad_norm": 3.7677040100097656, "learning_rate": 5.88824779333308e-09, "loss": 1.35, "step": 50253 }, { "epoch": 3.9384012539184954, "grad_norm": 4.305149555206299, "learning_rate": 5.8732964476437836e-09, "loss": 1.218, "step": 50254 }, { "epoch": 3.9384796238244513, "grad_norm": 5.745418071746826, "learning_rate": 5.858364097048319e-09, "loss": 1.4878, "step": 50255 }, { "epoch": 3.9385579937304076, "grad_norm": 2.698493003845215, "learning_rate": 5.843450741601641e-09, "loss": 1.2875, "step": 50256 }, { "epoch": 3.9386363636363635, "grad_norm": 4.228091716766357, "learning_rate": 5.828556381361483e-09, "loss": 0.9619, "step": 50257 }, { "epoch": 3.93871473354232, "grad_norm": 2.4776861667633057, "learning_rate": 5.813681016384464e-09, "loss": 1.0937, "step": 50258 }, { "epoch": 3.9387931034482757, "grad_norm": 13.078229904174805, "learning_rate": 5.798824646727208e-09, "loss": 1.7359, "step": 50259 }, { "epoch": 3.938871473354232, "grad_norm": 5.7824482917785645, "learning_rate": 5.783987272445779e-09, "loss": 0.9791, "step": 50260 }, { "epoch": 3.938949843260188, "grad_norm": 3.7090017795562744, "learning_rate": 5.769168893596799e-09, "loss": 1.464, "step": 50261 }, { "epoch": 3.9390282131661443, "grad_norm": 4.283843517303467, "learning_rate": 5.75436951023689e-09, "loss": 1.2581, "step": 50262 }, { "epoch": 3.9391065830721, "grad_norm": 4.599149703979492, "learning_rate": 5.739589122422118e-09, "loss": 1.3236, "step": 50263 }, { "epoch": 3.9391849529780565, "grad_norm": 2.2651164531707764, "learning_rate": 5.724827730208549e-09, "loss": 1.4816, "step": 50264 }, { "epoch": 3.9392633228840124, "grad_norm": 3.1748950481414795, "learning_rate": 5.710085333652804e-09, "loss": 1.3288, "step": 50265 }, { "epoch": 3.9393416927899687, "grad_norm": 5.2827534675598145, "learning_rate": 5.695361932810395e-09, "loss": 1.4285, "step": 50266 }, { "epoch": 3.939420062695925, "grad_norm": 3.972421169281006, "learning_rate": 5.680657527737943e-09, "loss": 1.2769, "step": 50267 }, { "epoch": 3.939498432601881, "grad_norm": 4.209381103515625, "learning_rate": 5.665972118490404e-09, "loss": 1.1833, "step": 50268 }, { "epoch": 3.939576802507837, "grad_norm": 6.536753177642822, "learning_rate": 5.651305705124954e-09, "loss": 0.9892, "step": 50269 }, { "epoch": 3.939655172413793, "grad_norm": 3.312675714492798, "learning_rate": 5.636658287695995e-09, "loss": 1.4108, "step": 50270 }, { "epoch": 3.9397335423197495, "grad_norm": 2.685929298400879, "learning_rate": 5.6220298662607035e-09, "loss": 1.3988, "step": 50271 }, { "epoch": 3.9398119122257054, "grad_norm": 3.230694532394409, "learning_rate": 5.607420440873479e-09, "loss": 1.3096, "step": 50272 }, { "epoch": 3.9398902821316613, "grad_norm": 4.170141220092773, "learning_rate": 5.592830011590389e-09, "loss": 1.3301, "step": 50273 }, { "epoch": 3.9399686520376176, "grad_norm": 4.673415184020996, "learning_rate": 5.578258578466944e-09, "loss": 1.3944, "step": 50274 }, { "epoch": 3.940047021943574, "grad_norm": 2.4482154846191406, "learning_rate": 5.563706141558101e-09, "loss": 1.214, "step": 50275 }, { "epoch": 3.94012539184953, "grad_norm": 4.293153285980225, "learning_rate": 5.549172700919925e-09, "loss": 1.2853, "step": 50276 }, { "epoch": 3.9402037617554857, "grad_norm": 10.551560401916504, "learning_rate": 5.534658256607372e-09, "loss": 1.1456, "step": 50277 }, { "epoch": 3.940282131661442, "grad_norm": 6.410851955413818, "learning_rate": 5.520162808675955e-09, "loss": 1.1808, "step": 50278 }, { "epoch": 3.9403605015673984, "grad_norm": 8.849054336547852, "learning_rate": 5.505686357180073e-09, "loss": 2.2184, "step": 50279 }, { "epoch": 3.9404388714733543, "grad_norm": 3.5037119388580322, "learning_rate": 5.491228902175794e-09, "loss": 1.2165, "step": 50280 }, { "epoch": 3.94051724137931, "grad_norm": 3.80629563331604, "learning_rate": 5.476790443716961e-09, "loss": 1.5435, "step": 50281 }, { "epoch": 3.9405956112852665, "grad_norm": 1.9514201879501343, "learning_rate": 5.462370981859088e-09, "loss": 1.2585, "step": 50282 }, { "epoch": 3.940673981191223, "grad_norm": 5.14084529876709, "learning_rate": 5.44797051665713e-09, "loss": 1.2192, "step": 50283 }, { "epoch": 3.9407523510971787, "grad_norm": 3.608539342880249, "learning_rate": 5.4335890481660436e-09, "loss": 1.1905, "step": 50284 }, { "epoch": 3.9408307210031346, "grad_norm": 2.9999921321868896, "learning_rate": 5.4192265764402284e-09, "loss": 1.3742, "step": 50285 }, { "epoch": 3.940909090909091, "grad_norm": 2.863394021987915, "learning_rate": 5.404883101533531e-09, "loss": 1.4898, "step": 50286 }, { "epoch": 3.9409874608150472, "grad_norm": 4.496147155761719, "learning_rate": 5.390558623502018e-09, "loss": 1.5759, "step": 50287 }, { "epoch": 3.941065830721003, "grad_norm": 2.7573249340057373, "learning_rate": 5.3762531423989795e-09, "loss": 1.5213, "step": 50288 }, { "epoch": 3.941144200626959, "grad_norm": 2.6361520290374756, "learning_rate": 5.361966658279372e-09, "loss": 1.5791, "step": 50289 }, { "epoch": 3.9412225705329154, "grad_norm": 6.949214458465576, "learning_rate": 5.347699171197595e-09, "loss": 0.762, "step": 50290 }, { "epoch": 3.9413009404388717, "grad_norm": 4.732755661010742, "learning_rate": 5.333450681208052e-09, "loss": 1.7002, "step": 50291 }, { "epoch": 3.9413793103448276, "grad_norm": 2.214527130126953, "learning_rate": 5.319221188364032e-09, "loss": 1.271, "step": 50292 }, { "epoch": 3.9414576802507835, "grad_norm": 3.137303590774536, "learning_rate": 5.305010692720492e-09, "loss": 1.2013, "step": 50293 }, { "epoch": 3.94153605015674, "grad_norm": 6.890460968017578, "learning_rate": 5.290819194331276e-09, "loss": 0.8196, "step": 50294 }, { "epoch": 3.941614420062696, "grad_norm": 4.140695095062256, "learning_rate": 5.276646693250232e-09, "loss": 0.9686, "step": 50295 }, { "epoch": 3.941692789968652, "grad_norm": 2.432443857192993, "learning_rate": 5.262493189531759e-09, "loss": 0.6885, "step": 50296 }, { "epoch": 3.941771159874608, "grad_norm": 4.2904582023620605, "learning_rate": 5.248358683228594e-09, "loss": 1.3615, "step": 50297 }, { "epoch": 3.9418495297805642, "grad_norm": 2.9691522121429443, "learning_rate": 5.2342431743956925e-09, "loss": 1.1144, "step": 50298 }, { "epoch": 3.9419278996865206, "grad_norm": 3.2129485607147217, "learning_rate": 5.220146663086345e-09, "loss": 1.4636, "step": 50299 }, { "epoch": 3.9420062695924765, "grad_norm": 4.690516948699951, "learning_rate": 5.206069149353843e-09, "loss": 1.7584, "step": 50300 }, { "epoch": 3.9420846394984324, "grad_norm": 2.620509386062622, "learning_rate": 5.1920106332520315e-09, "loss": 1.3391, "step": 50301 }, { "epoch": 3.9421630094043887, "grad_norm": 5.5974249839782715, "learning_rate": 5.177971114834201e-09, "loss": 1.2889, "step": 50302 }, { "epoch": 3.942241379310345, "grad_norm": 5.726509094238281, "learning_rate": 5.163950594154199e-09, "loss": 1.2791, "step": 50303 }, { "epoch": 3.942319749216301, "grad_norm": 2.66973876953125, "learning_rate": 5.149949071264759e-09, "loss": 1.2881, "step": 50304 }, { "epoch": 3.942398119122257, "grad_norm": 7.731301307678223, "learning_rate": 5.135966546219728e-09, "loss": 1.1316, "step": 50305 }, { "epoch": 3.942476489028213, "grad_norm": 3.771254062652588, "learning_rate": 5.122003019071286e-09, "loss": 1.565, "step": 50306 }, { "epoch": 3.9425548589341695, "grad_norm": 2.8063249588012695, "learning_rate": 5.108058489873835e-09, "loss": 1.0484, "step": 50307 }, { "epoch": 3.9426332288401253, "grad_norm": 4.357760429382324, "learning_rate": 5.094132958679554e-09, "loss": 1.6026, "step": 50308 }, { "epoch": 3.9427115987460812, "grad_norm": 4.656548023223877, "learning_rate": 5.080226425541734e-09, "loss": 1.6798, "step": 50309 }, { "epoch": 3.9427899686520376, "grad_norm": 3.7242584228515625, "learning_rate": 5.066338890513112e-09, "loss": 0.6766, "step": 50310 }, { "epoch": 3.942868338557994, "grad_norm": 3.402806520462036, "learning_rate": 5.052470353646421e-09, "loss": 1.076, "step": 50311 }, { "epoch": 3.94294670846395, "grad_norm": 3.678677797317505, "learning_rate": 5.038620814994954e-09, "loss": 1.4233, "step": 50312 }, { "epoch": 3.943025078369906, "grad_norm": 4.273081302642822, "learning_rate": 5.024790274610336e-09, "loss": 1.1856, "step": 50313 }, { "epoch": 3.943103448275862, "grad_norm": 2.919715166091919, "learning_rate": 5.010978732546412e-09, "loss": 1.3434, "step": 50314 }, { "epoch": 3.9431818181818183, "grad_norm": 7.578060150146484, "learning_rate": 4.997186188854808e-09, "loss": 1.9952, "step": 50315 }, { "epoch": 3.9432601880877742, "grad_norm": 2.377858877182007, "learning_rate": 4.983412643588814e-09, "loss": 1.1221, "step": 50316 }, { "epoch": 3.9433385579937306, "grad_norm": 4.6241679191589355, "learning_rate": 4.969658096800056e-09, "loss": 1.0519, "step": 50317 }, { "epoch": 3.9434169278996865, "grad_norm": 6.963767051696777, "learning_rate": 4.955922548540714e-09, "loss": 1.8336, "step": 50318 }, { "epoch": 3.943495297805643, "grad_norm": 3.852640390396118, "learning_rate": 4.94220599886408e-09, "loss": 1.3628, "step": 50319 }, { "epoch": 3.9435736677115987, "grad_norm": 3.5692617893218994, "learning_rate": 4.928508447821223e-09, "loss": 1.03, "step": 50320 }, { "epoch": 3.943652037617555, "grad_norm": 4.315001487731934, "learning_rate": 4.914829895464879e-09, "loss": 1.5996, "step": 50321 }, { "epoch": 3.943730407523511, "grad_norm": 6.747899055480957, "learning_rate": 4.901170341846673e-09, "loss": 1.2778, "step": 50322 }, { "epoch": 3.9438087774294672, "grad_norm": 6.0627522468566895, "learning_rate": 4.887529787018785e-09, "loss": 1.0549, "step": 50323 }, { "epoch": 3.943887147335423, "grad_norm": 7.03701639175415, "learning_rate": 4.873908231033397e-09, "loss": 1.2613, "step": 50324 }, { "epoch": 3.9439655172413794, "grad_norm": 3.3369076251983643, "learning_rate": 4.860305673941579e-09, "loss": 1.4441, "step": 50325 }, { "epoch": 3.9440438871473353, "grad_norm": 4.7682576179504395, "learning_rate": 4.846722115796065e-09, "loss": 1.9895, "step": 50326 }, { "epoch": 3.9441222570532917, "grad_norm": 2.6136443614959717, "learning_rate": 4.833157556647372e-09, "loss": 1.2002, "step": 50327 }, { "epoch": 3.9442006269592476, "grad_norm": 3.2624311447143555, "learning_rate": 4.81961199654768e-09, "loss": 1.0168, "step": 50328 }, { "epoch": 3.944278996865204, "grad_norm": 2.3930976390838623, "learning_rate": 4.806085435549168e-09, "loss": 1.2409, "step": 50329 }, { "epoch": 3.94435736677116, "grad_norm": 3.460214376449585, "learning_rate": 4.792577873701798e-09, "loss": 1.1435, "step": 50330 }, { "epoch": 3.944435736677116, "grad_norm": 3.5564963817596436, "learning_rate": 4.779089311058305e-09, "loss": 0.9979, "step": 50331 }, { "epoch": 3.944514106583072, "grad_norm": 8.943735122680664, "learning_rate": 4.765619747669203e-09, "loss": 1.18, "step": 50332 }, { "epoch": 3.9445924764890283, "grad_norm": 3.209543228149414, "learning_rate": 4.75216918358612e-09, "loss": 1.4462, "step": 50333 }, { "epoch": 3.944670846394984, "grad_norm": 5.670943260192871, "learning_rate": 4.7387376188601235e-09, "loss": 1.5141, "step": 50334 }, { "epoch": 3.9447492163009406, "grad_norm": 5.09165096282959, "learning_rate": 4.72532505354173e-09, "loss": 1.7779, "step": 50335 }, { "epoch": 3.9448275862068964, "grad_norm": 7.95573091506958, "learning_rate": 4.711931487683119e-09, "loss": 1.4928, "step": 50336 }, { "epoch": 3.9449059561128528, "grad_norm": 5.713890075683594, "learning_rate": 4.698556921334252e-09, "loss": 0.9389, "step": 50337 }, { "epoch": 3.9449843260188087, "grad_norm": 10.138359069824219, "learning_rate": 4.685201354546199e-09, "loss": 1.4429, "step": 50338 }, { "epoch": 3.945062695924765, "grad_norm": 3.513306140899658, "learning_rate": 4.6718647873700285e-09, "loss": 1.4326, "step": 50339 }, { "epoch": 3.945141065830721, "grad_norm": 2.7565643787384033, "learning_rate": 4.658547219856257e-09, "loss": 1.4549, "step": 50340 }, { "epoch": 3.945219435736677, "grad_norm": 2.6769139766693115, "learning_rate": 4.645248652055956e-09, "loss": 1.3708, "step": 50341 }, { "epoch": 3.945297805642633, "grad_norm": 9.22160816192627, "learning_rate": 4.631969084019084e-09, "loss": 0.9556, "step": 50342 }, { "epoch": 3.9453761755485894, "grad_norm": 4.1882643699646, "learning_rate": 4.618708515796156e-09, "loss": 1.3626, "step": 50343 }, { "epoch": 3.9454545454545453, "grad_norm": 9.953984260559082, "learning_rate": 4.605466947438242e-09, "loss": 1.9327, "step": 50344 }, { "epoch": 3.9455329153605017, "grad_norm": 5.058982849121094, "learning_rate": 4.5922443789953035e-09, "loss": 0.9682, "step": 50345 }, { "epoch": 3.9456112852664575, "grad_norm": 5.326322078704834, "learning_rate": 4.579040810517854e-09, "loss": 1.3948, "step": 50346 }, { "epoch": 3.945689655172414, "grad_norm": 2.971302032470703, "learning_rate": 4.5658562420553e-09, "loss": 1.308, "step": 50347 }, { "epoch": 3.9457680250783698, "grad_norm": 8.503714561462402, "learning_rate": 4.552690673658711e-09, "loss": 1.5292, "step": 50348 }, { "epoch": 3.945846394984326, "grad_norm": 5.1581573486328125, "learning_rate": 4.539544105378046e-09, "loss": 1.0504, "step": 50349 }, { "epoch": 3.945924764890282, "grad_norm": 3.541236400604248, "learning_rate": 4.526416537263267e-09, "loss": 1.4729, "step": 50350 }, { "epoch": 3.9460031347962383, "grad_norm": 2.3807599544525146, "learning_rate": 4.513307969363778e-09, "loss": 1.2811, "step": 50351 }, { "epoch": 3.946081504702194, "grad_norm": 2.9096524715423584, "learning_rate": 4.5002184017295394e-09, "loss": 1.1717, "step": 50352 }, { "epoch": 3.9461598746081505, "grad_norm": 3.921400547027588, "learning_rate": 4.487147834411065e-09, "loss": 1.1212, "step": 50353 }, { "epoch": 3.9462382445141064, "grad_norm": 6.993711471557617, "learning_rate": 4.474096267457761e-09, "loss": 1.398, "step": 50354 }, { "epoch": 3.9463166144200628, "grad_norm": 4.851817607879639, "learning_rate": 4.461063700919033e-09, "loss": 1.5228, "step": 50355 }, { "epoch": 3.9463949843260187, "grad_norm": 12.194380760192871, "learning_rate": 4.448050134844284e-09, "loss": 1.4433, "step": 50356 }, { "epoch": 3.946473354231975, "grad_norm": 11.483150482177734, "learning_rate": 4.435055569283475e-09, "loss": 1.6021, "step": 50357 }, { "epoch": 3.946551724137931, "grad_norm": 3.9203953742980957, "learning_rate": 4.422080004286011e-09, "loss": 1.956, "step": 50358 }, { "epoch": 3.946630094043887, "grad_norm": 2.3088631629943848, "learning_rate": 4.409123439900742e-09, "loss": 1.3397, "step": 50359 }, { "epoch": 3.946708463949843, "grad_norm": 5.5620646476745605, "learning_rate": 4.3961858761776275e-09, "loss": 1.024, "step": 50360 }, { "epoch": 3.9467868338557994, "grad_norm": 3.1909475326538086, "learning_rate": 4.383267313164963e-09, "loss": 1.1246, "step": 50361 }, { "epoch": 3.9468652037617558, "grad_norm": 5.491511821746826, "learning_rate": 4.370367750912707e-09, "loss": 0.9336, "step": 50362 }, { "epoch": 3.9469435736677116, "grad_norm": 3.0132896900177, "learning_rate": 4.3574871894697116e-09, "loss": 1.2374, "step": 50363 }, { "epoch": 3.9470219435736675, "grad_norm": 4.002195358276367, "learning_rate": 4.344625628884824e-09, "loss": 1.6463, "step": 50364 }, { "epoch": 3.947100313479624, "grad_norm": 5.541941165924072, "learning_rate": 4.331783069207451e-09, "loss": 1.5541, "step": 50365 }, { "epoch": 3.94717868338558, "grad_norm": 6.299324035644531, "learning_rate": 4.318959510485332e-09, "loss": 1.4587, "step": 50366 }, { "epoch": 3.947257053291536, "grad_norm": 5.986440181732178, "learning_rate": 4.306154952768427e-09, "loss": 1.1311, "step": 50367 }, { "epoch": 3.947335423197492, "grad_norm": 2.7542901039123535, "learning_rate": 4.293369396105029e-09, "loss": 1.3176, "step": 50368 }, { "epoch": 3.9474137931034483, "grad_norm": 3.3819119930267334, "learning_rate": 4.280602840543435e-09, "loss": 0.924, "step": 50369 }, { "epoch": 3.9474921630094046, "grad_norm": 7.53787899017334, "learning_rate": 4.267855286132494e-09, "loss": 0.58, "step": 50370 }, { "epoch": 3.9475705329153605, "grad_norm": 2.749136447906494, "learning_rate": 4.255126732921055e-09, "loss": 1.3778, "step": 50371 }, { "epoch": 3.9476489028213164, "grad_norm": 3.756641387939453, "learning_rate": 4.2424171809563045e-09, "loss": 1.2307, "step": 50372 }, { "epoch": 3.9477272727272728, "grad_norm": 6.43280029296875, "learning_rate": 4.2297266302882e-09, "loss": 0.9993, "step": 50373 }, { "epoch": 3.947805642633229, "grad_norm": 6.97203254699707, "learning_rate": 4.217055080963928e-09, "loss": 1.114, "step": 50374 }, { "epoch": 3.947884012539185, "grad_norm": 5.909792900085449, "learning_rate": 4.204402533031782e-09, "loss": 1.1612, "step": 50375 }, { "epoch": 3.947962382445141, "grad_norm": 4.639089584350586, "learning_rate": 4.191768986540612e-09, "loss": 1.5498, "step": 50376 }, { "epoch": 3.948040752351097, "grad_norm": 4.414093494415283, "learning_rate": 4.1791544415376025e-09, "loss": 1.2807, "step": 50377 }, { "epoch": 3.9481191222570535, "grad_norm": 3.238346576690674, "learning_rate": 4.166558898071049e-09, "loss": 1.0101, "step": 50378 }, { "epoch": 3.9481974921630094, "grad_norm": 2.9259870052337646, "learning_rate": 4.15398235618869e-09, "loss": 1.3433, "step": 50379 }, { "epoch": 3.9482758620689653, "grad_norm": 7.217201232910156, "learning_rate": 4.141424815938822e-09, "loss": 1.3058, "step": 50380 }, { "epoch": 3.9483542319749216, "grad_norm": 4.276417255401611, "learning_rate": 4.128886277369182e-09, "loss": 1.4643, "step": 50381 }, { "epoch": 3.948432601880878, "grad_norm": 3.6374993324279785, "learning_rate": 4.116366740526956e-09, "loss": 1.2914, "step": 50382 }, { "epoch": 3.948510971786834, "grad_norm": 2.9298806190490723, "learning_rate": 4.103866205459884e-09, "loss": 1.1729, "step": 50383 }, { "epoch": 3.9485893416927897, "grad_norm": 3.3211355209350586, "learning_rate": 4.091384672215703e-09, "loss": 1.3728, "step": 50384 }, { "epoch": 3.948667711598746, "grad_norm": 3.015118360519409, "learning_rate": 4.0789221408421565e-09, "loss": 1.4339, "step": 50385 }, { "epoch": 3.9487460815047024, "grad_norm": 2.6129231452941895, "learning_rate": 4.066478611385871e-09, "loss": 1.3453, "step": 50386 }, { "epoch": 3.9488244514106583, "grad_norm": 2.904721975326538, "learning_rate": 4.054054083895143e-09, "loss": 1.4581, "step": 50387 }, { "epoch": 3.948902821316614, "grad_norm": 2.2724759578704834, "learning_rate": 4.0416485584160445e-09, "loss": 1.3651, "step": 50388 }, { "epoch": 3.9489811912225705, "grad_norm": 2.6198863983154297, "learning_rate": 4.029262034996873e-09, "loss": 1.5399, "step": 50389 }, { "epoch": 3.949059561128527, "grad_norm": 3.1516449451446533, "learning_rate": 4.016894513684255e-09, "loss": 1.1324, "step": 50390 }, { "epoch": 3.9491379310344827, "grad_norm": 4.731422424316406, "learning_rate": 4.004545994525377e-09, "loss": 1.2917, "step": 50391 }, { "epoch": 3.9492163009404386, "grad_norm": 2.309661865234375, "learning_rate": 3.992216477566868e-09, "loss": 1.4503, "step": 50392 }, { "epoch": 3.949294670846395, "grad_norm": 2.8016679286956787, "learning_rate": 3.9799059628559126e-09, "loss": 1.3748, "step": 50393 }, { "epoch": 3.9493730407523513, "grad_norm": 3.1719658374786377, "learning_rate": 3.967614450439139e-09, "loss": 1.1443, "step": 50394 }, { "epoch": 3.949451410658307, "grad_norm": 3.60404896736145, "learning_rate": 3.9553419403637325e-09, "loss": 1.2722, "step": 50395 }, { "epoch": 3.949529780564263, "grad_norm": 3.126063108444214, "learning_rate": 3.943088432675768e-09, "loss": 1.2736, "step": 50396 }, { "epoch": 3.9496081504702194, "grad_norm": 5.966063022613525, "learning_rate": 3.930853927422429e-09, "loss": 0.9996, "step": 50397 }, { "epoch": 3.9496865203761757, "grad_norm": 2.901304006576538, "learning_rate": 3.918638424650345e-09, "loss": 1.578, "step": 50398 }, { "epoch": 3.9497648902821316, "grad_norm": 3.7558341026306152, "learning_rate": 3.906441924405036e-09, "loss": 1.1438, "step": 50399 }, { "epoch": 3.9498432601880875, "grad_norm": 2.8640661239624023, "learning_rate": 3.894264426733685e-09, "loss": 1.0302, "step": 50400 }, { "epoch": 3.949921630094044, "grad_norm": 5.306271076202393, "learning_rate": 3.882105931681812e-09, "loss": 1.5095, "step": 50401 }, { "epoch": 3.95, "grad_norm": 3.301605224609375, "learning_rate": 3.869966439297157e-09, "loss": 1.6596, "step": 50402 }, { "epoch": 3.950078369905956, "grad_norm": 3.1286733150482178, "learning_rate": 3.8578459496241285e-09, "loss": 1.0721, "step": 50403 }, { "epoch": 3.950156739811912, "grad_norm": 5.385272026062012, "learning_rate": 3.845744462710465e-09, "loss": 1.6179, "step": 50404 }, { "epoch": 3.9502351097178683, "grad_norm": 3.6801860332489014, "learning_rate": 3.833661978600578e-09, "loss": 1.2034, "step": 50405 }, { "epoch": 3.9503134796238246, "grad_norm": 3.896804094314575, "learning_rate": 3.82159849734165e-09, "loss": 1.6081, "step": 50406 }, { "epoch": 3.9503918495297805, "grad_norm": 3.579127073287964, "learning_rate": 3.8095540189792e-09, "loss": 1.5341, "step": 50407 }, { "epoch": 3.950470219435737, "grad_norm": 4.875767230987549, "learning_rate": 3.7975285435593035e-09, "loss": 1.0637, "step": 50408 }, { "epoch": 3.9505485893416927, "grad_norm": 2.5335309505462646, "learning_rate": 3.785522071126924e-09, "loss": 1.3496, "step": 50409 }, { "epoch": 3.950626959247649, "grad_norm": 4.940530300140381, "learning_rate": 3.77353460172869e-09, "loss": 1.2114, "step": 50410 }, { "epoch": 3.950705329153605, "grad_norm": 4.723275184631348, "learning_rate": 3.761566135409012e-09, "loss": 1.3117, "step": 50411 }, { "epoch": 3.9507836990595613, "grad_norm": 5.320052623748779, "learning_rate": 3.7496166722145176e-09, "loss": 1.3475, "step": 50412 }, { "epoch": 3.950862068965517, "grad_norm": 7.375267028808594, "learning_rate": 3.7376862121901724e-09, "loss": 1.9067, "step": 50413 }, { "epoch": 3.9509404388714735, "grad_norm": 7.757757663726807, "learning_rate": 3.725774755381495e-09, "loss": 1.3839, "step": 50414 }, { "epoch": 3.9510188087774294, "grad_norm": 3.411644220352173, "learning_rate": 3.713882301834004e-09, "loss": 1.2136, "step": 50415 }, { "epoch": 3.9510971786833857, "grad_norm": 3.191420078277588, "learning_rate": 3.7020088515921094e-09, "loss": 1.4034, "step": 50416 }, { "epoch": 3.9511755485893416, "grad_norm": 4.141672611236572, "learning_rate": 3.6901544047018843e-09, "loss": 1.4297, "step": 50417 }, { "epoch": 3.951253918495298, "grad_norm": 3.107750654220581, "learning_rate": 3.6783189612077384e-09, "loss": 1.3268, "step": 50418 }, { "epoch": 3.951332288401254, "grad_norm": 2.6018025875091553, "learning_rate": 3.666502521155191e-09, "loss": 1.3369, "step": 50419 }, { "epoch": 3.95141065830721, "grad_norm": 3.2611255645751953, "learning_rate": 3.65470508458865e-09, "loss": 1.4969, "step": 50420 }, { "epoch": 3.951489028213166, "grad_norm": 2.4407355785369873, "learning_rate": 3.6429266515536356e-09, "loss": 1.2035, "step": 50421 }, { "epoch": 3.9515673981191224, "grad_norm": 2.34614634513855, "learning_rate": 3.631167222095111e-09, "loss": 1.2107, "step": 50422 }, { "epoch": 3.9516457680250783, "grad_norm": 2.0050461292266846, "learning_rate": 3.6194267962563758e-09, "loss": 0.8064, "step": 50423 }, { "epoch": 3.9517241379310346, "grad_norm": 2.5639803409576416, "learning_rate": 3.6077053740835032e-09, "loss": 1.4237, "step": 50424 }, { "epoch": 3.9518025078369905, "grad_norm": 2.679316759109497, "learning_rate": 3.596002955620903e-09, "loss": 1.6562, "step": 50425 }, { "epoch": 3.951880877742947, "grad_norm": 3.7524521350860596, "learning_rate": 3.584319540912429e-09, "loss": 1.6104, "step": 50426 }, { "epoch": 3.9519592476489027, "grad_norm": 2.892108678817749, "learning_rate": 3.5726551300030444e-09, "loss": 1.5499, "step": 50427 }, { "epoch": 3.952037617554859, "grad_norm": 3.3127357959747314, "learning_rate": 3.5610097229366036e-09, "loss": 0.8873, "step": 50428 }, { "epoch": 3.952115987460815, "grad_norm": 2.4840376377105713, "learning_rate": 3.5493833197580707e-09, "loss": 1.3333, "step": 50429 }, { "epoch": 3.9521943573667713, "grad_norm": 2.5434367656707764, "learning_rate": 3.5377759205112995e-09, "loss": 1.3331, "step": 50430 }, { "epoch": 3.952272727272727, "grad_norm": 2.2469494342803955, "learning_rate": 3.5261875252406986e-09, "loss": 1.2508, "step": 50431 }, { "epoch": 3.9523510971786835, "grad_norm": 3.1357667446136475, "learning_rate": 3.5146181339895667e-09, "loss": 1.4099, "step": 50432 }, { "epoch": 3.9524294670846394, "grad_norm": 2.228309154510498, "learning_rate": 3.5030677468028684e-09, "loss": 1.1999, "step": 50433 }, { "epoch": 3.9525078369905957, "grad_norm": 3.6167125701904297, "learning_rate": 3.491536363724457e-09, "loss": 0.9375, "step": 50434 }, { "epoch": 3.9525862068965516, "grad_norm": 2.3434925079345703, "learning_rate": 3.480023984797076e-09, "loss": 1.3278, "step": 50435 }, { "epoch": 3.952664576802508, "grad_norm": 2.758924722671509, "learning_rate": 3.46853061006569e-09, "loss": 1.3112, "step": 50436 }, { "epoch": 3.952742946708464, "grad_norm": 2.9275693893432617, "learning_rate": 3.4570562395735973e-09, "loss": 1.3217, "step": 50437 }, { "epoch": 3.95282131661442, "grad_norm": 4.985278606414795, "learning_rate": 3.4456008733646516e-09, "loss": 1.189, "step": 50438 }, { "epoch": 3.952899686520376, "grad_norm": 3.9197113513946533, "learning_rate": 3.434164511482152e-09, "loss": 1.4706, "step": 50439 }, { "epoch": 3.9529780564263324, "grad_norm": 2.856740713119507, "learning_rate": 3.422747153969952e-09, "loss": 1.4009, "step": 50440 }, { "epoch": 3.9530564263322883, "grad_norm": 4.980866432189941, "learning_rate": 3.411348800870795e-09, "loss": 1.132, "step": 50441 }, { "epoch": 3.9531347962382446, "grad_norm": 3.9700844287872314, "learning_rate": 3.3999694522285354e-09, "loss": 1.3742, "step": 50442 }, { "epoch": 3.9532131661442005, "grad_norm": 3.1198415756225586, "learning_rate": 3.3886091080864714e-09, "loss": 0.9379, "step": 50443 }, { "epoch": 3.953291536050157, "grad_norm": 5.205422878265381, "learning_rate": 3.377267768487902e-09, "loss": 1.2562, "step": 50444 }, { "epoch": 3.9533699059561127, "grad_norm": 19.613056182861328, "learning_rate": 3.3659454334755702e-09, "loss": 1.4144, "step": 50445 }, { "epoch": 3.953448275862069, "grad_norm": 2.290525197982788, "learning_rate": 3.3546421030927755e-09, "loss": 1.2463, "step": 50446 }, { "epoch": 3.953526645768025, "grad_norm": 5.886207580566406, "learning_rate": 3.343357777382261e-09, "loss": 1.294, "step": 50447 }, { "epoch": 3.9536050156739813, "grad_norm": 2.7815632820129395, "learning_rate": 3.332092456387326e-09, "loss": 0.9803, "step": 50448 }, { "epoch": 3.953683385579937, "grad_norm": 8.149861335754395, "learning_rate": 3.320846140150713e-09, "loss": 1.2341, "step": 50449 }, { "epoch": 3.9537617554858935, "grad_norm": 3.2922561168670654, "learning_rate": 3.3096188287151664e-09, "loss": 1.7179, "step": 50450 }, { "epoch": 3.9538401253918494, "grad_norm": 2.655447006225586, "learning_rate": 3.2984105221234295e-09, "loss": 1.1935, "step": 50451 }, { "epoch": 3.9539184952978057, "grad_norm": 3.044938087463379, "learning_rate": 3.2872212204176913e-09, "loss": 1.3839, "step": 50452 }, { "epoch": 3.9539968652037616, "grad_norm": 2.5466110706329346, "learning_rate": 3.2760509236412495e-09, "loss": 1.5635, "step": 50453 }, { "epoch": 3.954075235109718, "grad_norm": 2.822030782699585, "learning_rate": 3.2648996318362934e-09, "loss": 1.3352, "step": 50454 }, { "epoch": 3.954153605015674, "grad_norm": 6.758139133453369, "learning_rate": 3.2537673450450115e-09, "loss": 1.1578, "step": 50455 }, { "epoch": 3.95423197492163, "grad_norm": 4.221771240234375, "learning_rate": 3.2426540633095915e-09, "loss": 1.2615, "step": 50456 }, { "epoch": 3.9543103448275865, "grad_norm": 4.644716262817383, "learning_rate": 3.2315597866733327e-09, "loss": 1.2816, "step": 50457 }, { "epoch": 3.9543887147335424, "grad_norm": 5.536277770996094, "learning_rate": 3.2204845151773136e-09, "loss": 1.3726, "step": 50458 }, { "epoch": 3.9544670846394983, "grad_norm": 4.208190441131592, "learning_rate": 3.209428248863722e-09, "loss": 1.5079, "step": 50459 }, { "epoch": 3.9545454545454546, "grad_norm": 3.610900402069092, "learning_rate": 3.1983909877753017e-09, "loss": 1.6657, "step": 50460 }, { "epoch": 3.954623824451411, "grad_norm": 3.701046943664551, "learning_rate": 3.1873727319536863e-09, "loss": 1.5744, "step": 50461 }, { "epoch": 3.954702194357367, "grad_norm": 3.2053868770599365, "learning_rate": 3.176373481441064e-09, "loss": 1.3666, "step": 50462 }, { "epoch": 3.9547805642633227, "grad_norm": 4.585809230804443, "learning_rate": 3.1653932362785135e-09, "loss": 1.4403, "step": 50463 }, { "epoch": 3.954858934169279, "grad_norm": 5.2867536544799805, "learning_rate": 3.1544319965087776e-09, "loss": 1.2446, "step": 50464 }, { "epoch": 3.9549373040752354, "grad_norm": 3.2126388549804688, "learning_rate": 3.1434897621723802e-09, "loss": 1.3646, "step": 50465 }, { "epoch": 3.9550156739811912, "grad_norm": 4.283529281616211, "learning_rate": 3.132566533312065e-09, "loss": 1.6584, "step": 50466 }, { "epoch": 3.955094043887147, "grad_norm": 6.310154914855957, "learning_rate": 3.121662309968909e-09, "loss": 1.2036, "step": 50467 }, { "epoch": 3.9551724137931035, "grad_norm": 3.565882682800293, "learning_rate": 3.110777092184547e-09, "loss": 1.2786, "step": 50468 }, { "epoch": 3.95525078369906, "grad_norm": 6.412228584289551, "learning_rate": 3.0999108799995013e-09, "loss": 2.3615, "step": 50469 }, { "epoch": 3.9553291536050157, "grad_norm": 2.4332380294799805, "learning_rate": 3.089063673456516e-09, "loss": 1.2037, "step": 50470 }, { "epoch": 3.9554075235109716, "grad_norm": 3.9522788524627686, "learning_rate": 3.078235472596114e-09, "loss": 0.8628, "step": 50471 }, { "epoch": 3.955485893416928, "grad_norm": 5.112989902496338, "learning_rate": 3.0674262774593735e-09, "loss": 1.2331, "step": 50472 }, { "epoch": 3.9555642633228842, "grad_norm": 4.764486789703369, "learning_rate": 3.0566360880873724e-09, "loss": 1.4337, "step": 50473 }, { "epoch": 3.95564263322884, "grad_norm": 2.862779378890991, "learning_rate": 3.0458649045211897e-09, "loss": 1.4702, "step": 50474 }, { "epoch": 3.955721003134796, "grad_norm": 3.4317660331726074, "learning_rate": 3.0351127268024583e-09, "loss": 1.104, "step": 50475 }, { "epoch": 3.9557993730407524, "grad_norm": 3.894940137863159, "learning_rate": 3.0243795549711463e-09, "loss": 1.5252, "step": 50476 }, { "epoch": 3.9558777429467087, "grad_norm": 5.205113887786865, "learning_rate": 3.013665389068887e-09, "loss": 1.6985, "step": 50477 }, { "epoch": 3.9559561128526646, "grad_norm": 6.77227783203125, "learning_rate": 3.002970229135649e-09, "loss": 0.8249, "step": 50478 }, { "epoch": 3.9560344827586205, "grad_norm": 5.714397430419922, "learning_rate": 2.9922940752125096e-09, "loss": 1.3352, "step": 50479 }, { "epoch": 3.956112852664577, "grad_norm": 3.360858917236328, "learning_rate": 2.9816369273405476e-09, "loss": 1.3807, "step": 50480 }, { "epoch": 3.956191222570533, "grad_norm": 4.628250598907471, "learning_rate": 2.9709987855591758e-09, "loss": 1.4754, "step": 50481 }, { "epoch": 3.956269592476489, "grad_norm": 3.0067217350006104, "learning_rate": 2.960379649910028e-09, "loss": 1.2326, "step": 50482 }, { "epoch": 3.956347962382445, "grad_norm": 4.049625873565674, "learning_rate": 2.9497795204325163e-09, "loss": 1.5394, "step": 50483 }, { "epoch": 3.9564263322884012, "grad_norm": 3.016066551208496, "learning_rate": 2.9391983971677198e-09, "loss": 1.6547, "step": 50484 }, { "epoch": 3.9565047021943576, "grad_norm": 2.8728244304656982, "learning_rate": 2.9286362801550504e-09, "loss": 1.012, "step": 50485 }, { "epoch": 3.9565830721003135, "grad_norm": 2.776386022567749, "learning_rate": 2.9180931694361425e-09, "loss": 1.3878, "step": 50486 }, { "epoch": 3.9566614420062693, "grad_norm": 5.198709487915039, "learning_rate": 2.9075690650492983e-09, "loss": 1.1162, "step": 50487 }, { "epoch": 3.9567398119122257, "grad_norm": 5.576664447784424, "learning_rate": 2.897063967035596e-09, "loss": 1.6601, "step": 50488 }, { "epoch": 3.956818181818182, "grad_norm": 6.9670867919921875, "learning_rate": 2.886577875435004e-09, "loss": 1.5803, "step": 50489 }, { "epoch": 3.956896551724138, "grad_norm": 15.889137268066406, "learning_rate": 2.8761107902869344e-09, "loss": 1.0678, "step": 50490 }, { "epoch": 3.956974921630094, "grad_norm": 3.0339248180389404, "learning_rate": 2.865662711631356e-09, "loss": 1.2918, "step": 50491 }, { "epoch": 3.95705329153605, "grad_norm": 3.786923408508301, "learning_rate": 2.855233639508237e-09, "loss": 1.4019, "step": 50492 }, { "epoch": 3.9571316614420065, "grad_norm": 2.3090431690216064, "learning_rate": 2.844823573957545e-09, "loss": 1.38, "step": 50493 }, { "epoch": 3.9572100313479623, "grad_norm": 4.128985404968262, "learning_rate": 2.8344325150181374e-09, "loss": 1.2075, "step": 50494 }, { "epoch": 3.9572884012539182, "grad_norm": 4.3545966148376465, "learning_rate": 2.824060462729428e-09, "loss": 1.4275, "step": 50495 }, { "epoch": 3.9573667711598746, "grad_norm": 4.547053813934326, "learning_rate": 2.813707417131939e-09, "loss": 1.2774, "step": 50496 }, { "epoch": 3.957445141065831, "grad_norm": 4.0102996826171875, "learning_rate": 2.803373378263974e-09, "loss": 1.3715, "step": 50497 }, { "epoch": 3.957523510971787, "grad_norm": 4.075305461883545, "learning_rate": 2.7930583461655002e-09, "loss": 1.6069, "step": 50498 }, { "epoch": 3.9576018808777427, "grad_norm": 5.532263278961182, "learning_rate": 2.7827623208753763e-09, "loss": 1.0561, "step": 50499 }, { "epoch": 3.957680250783699, "grad_norm": 2.668415069580078, "learning_rate": 2.7724853024324594e-09, "loss": 1.1286, "step": 50500 }, { "epoch": 3.9577586206896553, "grad_norm": 3.3299875259399414, "learning_rate": 2.762227290876718e-09, "loss": 1.1599, "step": 50501 }, { "epoch": 3.9578369905956112, "grad_norm": 2.8980660438537598, "learning_rate": 2.7519882862470093e-09, "loss": 1.5837, "step": 50502 }, { "epoch": 3.9579153605015676, "grad_norm": 2.748950958251953, "learning_rate": 2.7417682885816365e-09, "loss": 1.3879, "step": 50503 }, { "epoch": 3.9579937304075234, "grad_norm": 5.778728008270264, "learning_rate": 2.7315672979200124e-09, "loss": 1.0274, "step": 50504 }, { "epoch": 3.95807210031348, "grad_norm": 3.0485222339630127, "learning_rate": 2.72138531430044e-09, "loss": 1.164, "step": 50505 }, { "epoch": 3.9581504702194357, "grad_norm": 5.215002536773682, "learning_rate": 2.7112223377623314e-09, "loss": 0.9972, "step": 50506 }, { "epoch": 3.958228840125392, "grad_norm": 3.8206074237823486, "learning_rate": 2.701078368343435e-09, "loss": 1.2564, "step": 50507 }, { "epoch": 3.958307210031348, "grad_norm": 3.8241610527038574, "learning_rate": 2.690953406083163e-09, "loss": 1.398, "step": 50508 }, { "epoch": 3.958385579937304, "grad_norm": 2.6084916591644287, "learning_rate": 2.680847451019819e-09, "loss": 1.5113, "step": 50509 }, { "epoch": 3.95846394984326, "grad_norm": 6.2295379638671875, "learning_rate": 2.670760503191705e-09, "loss": 1.162, "step": 50510 }, { "epoch": 3.9585423197492164, "grad_norm": 2.3004519939422607, "learning_rate": 2.660692562637124e-09, "loss": 1.172, "step": 50511 }, { "epoch": 3.9586206896551723, "grad_norm": 3.641904354095459, "learning_rate": 2.650643629394378e-09, "loss": 1.322, "step": 50512 }, { "epoch": 3.9586990595611287, "grad_norm": 6.6449809074401855, "learning_rate": 2.640613703502326e-09, "loss": 1.7209, "step": 50513 }, { "epoch": 3.9587774294670846, "grad_norm": 7.254308223724365, "learning_rate": 2.63060278499816e-09, "loss": 1.2554, "step": 50514 }, { "epoch": 3.958855799373041, "grad_norm": 2.494004964828491, "learning_rate": 2.620610873920182e-09, "loss": 1.3105, "step": 50515 }, { "epoch": 3.9589341692789968, "grad_norm": 3.7347989082336426, "learning_rate": 2.6106379703066953e-09, "loss": 1.1809, "step": 50516 }, { "epoch": 3.959012539184953, "grad_norm": 3.4048948287963867, "learning_rate": 2.6006840741954475e-09, "loss": 1.5865, "step": 50517 }, { "epoch": 3.959090909090909, "grad_norm": 6.334400653839111, "learning_rate": 2.5907491856247414e-09, "loss": 0.6124, "step": 50518 }, { "epoch": 3.9591692789968653, "grad_norm": 3.0226731300354004, "learning_rate": 2.580833304631769e-09, "loss": 1.3839, "step": 50519 }, { "epoch": 3.959247648902821, "grad_norm": 2.773016929626465, "learning_rate": 2.5709364312548336e-09, "loss": 1.061, "step": 50520 }, { "epoch": 3.9593260188087775, "grad_norm": 4.032596588134766, "learning_rate": 2.561058565530572e-09, "loss": 1.4183, "step": 50521 }, { "epoch": 3.9594043887147334, "grad_norm": 4.430366039276123, "learning_rate": 2.551199707497287e-09, "loss": 1.0685, "step": 50522 }, { "epoch": 3.9594827586206898, "grad_norm": 2.3488590717315674, "learning_rate": 2.5413598571927266e-09, "loss": 1.2999, "step": 50523 }, { "epoch": 3.9595611285266457, "grad_norm": 4.3568010330200195, "learning_rate": 2.531539014654083e-09, "loss": 1.5227, "step": 50524 }, { "epoch": 3.959639498432602, "grad_norm": 3.226707696914673, "learning_rate": 2.5217371799185486e-09, "loss": 1.2357, "step": 50525 }, { "epoch": 3.959717868338558, "grad_norm": 5.5381975173950195, "learning_rate": 2.511954353023316e-09, "loss": 1.3939, "step": 50526 }, { "epoch": 3.959796238244514, "grad_norm": 2.6645419597625732, "learning_rate": 2.5021905340055776e-09, "loss": 1.349, "step": 50527 }, { "epoch": 3.95987460815047, "grad_norm": 3.1798195838928223, "learning_rate": 2.4924457229025255e-09, "loss": 1.2356, "step": 50528 }, { "epoch": 3.9599529780564264, "grad_norm": 4.723506927490234, "learning_rate": 2.4827199197513528e-09, "loss": 0.6369, "step": 50529 }, { "epoch": 3.9600313479623823, "grad_norm": 2.8860678672790527, "learning_rate": 2.473013124589252e-09, "loss": 1.1024, "step": 50530 }, { "epoch": 3.9601097178683387, "grad_norm": 12.35667610168457, "learning_rate": 2.463325337453415e-09, "loss": 1.1861, "step": 50531 }, { "epoch": 3.9601880877742945, "grad_norm": 7.240065097808838, "learning_rate": 2.453656558379369e-09, "loss": 1.1595, "step": 50532 }, { "epoch": 3.960266457680251, "grad_norm": 2.95308780670166, "learning_rate": 2.4440067874048624e-09, "loss": 1.3083, "step": 50533 }, { "epoch": 3.9603448275862068, "grad_norm": 2.6430046558380127, "learning_rate": 2.4343760245665317e-09, "loss": 1.3791, "step": 50534 }, { "epoch": 3.960423197492163, "grad_norm": 3.6899101734161377, "learning_rate": 2.424764269901014e-09, "loss": 1.5306, "step": 50535 }, { "epoch": 3.960501567398119, "grad_norm": 4.693446159362793, "learning_rate": 2.415171523444948e-09, "loss": 0.9832, "step": 50536 }, { "epoch": 3.9605799373040753, "grad_norm": 3.3457937240600586, "learning_rate": 2.4055977852344147e-09, "loss": 1.4264, "step": 50537 }, { "epoch": 3.960658307210031, "grad_norm": 3.170689582824707, "learning_rate": 2.3960430553060523e-09, "loss": 1.4608, "step": 50538 }, { "epoch": 3.9607366771159875, "grad_norm": 3.109898567199707, "learning_rate": 2.386507333696497e-09, "loss": 1.483, "step": 50539 }, { "epoch": 3.9608150470219434, "grad_norm": 4.134171009063721, "learning_rate": 2.3769906204412772e-09, "loss": 1.1473, "step": 50540 }, { "epoch": 3.9608934169278998, "grad_norm": 3.2008683681488037, "learning_rate": 2.36749291557703e-09, "loss": 1.3038, "step": 50541 }, { "epoch": 3.9609717868338556, "grad_norm": 3.887026786804199, "learning_rate": 2.3580142191403922e-09, "loss": 1.429, "step": 50542 }, { "epoch": 3.961050156739812, "grad_norm": 3.4232048988342285, "learning_rate": 2.348554531166891e-09, "loss": 1.2246, "step": 50543 }, { "epoch": 3.961128526645768, "grad_norm": 2.645430326461792, "learning_rate": 2.3391138516920544e-09, "loss": 1.428, "step": 50544 }, { "epoch": 3.961206896551724, "grad_norm": 6.7189249992370605, "learning_rate": 2.3296921807525187e-09, "loss": 1.7057, "step": 50545 }, { "epoch": 3.96128526645768, "grad_norm": 3.18544864654541, "learning_rate": 2.3202895183838115e-09, "loss": 1.2601, "step": 50546 }, { "epoch": 3.9613636363636364, "grad_norm": 4.922702312469482, "learning_rate": 2.310905864622015e-09, "loss": 1.4809, "step": 50547 }, { "epoch": 3.9614420062695928, "grad_norm": 3.4911439418792725, "learning_rate": 2.3015412195021013e-09, "loss": 1.2924, "step": 50548 }, { "epoch": 3.9615203761755486, "grad_norm": 2.548076629638672, "learning_rate": 2.292195583060708e-09, "loss": 1.1595, "step": 50549 }, { "epoch": 3.9615987460815045, "grad_norm": 2.9328858852386475, "learning_rate": 2.2828689553322516e-09, "loss": 1.2069, "step": 50550 }, { "epoch": 3.961677115987461, "grad_norm": 2.539365768432617, "learning_rate": 2.273561336352814e-09, "loss": 1.1336, "step": 50551 }, { "epoch": 3.961755485893417, "grad_norm": 2.14180326461792, "learning_rate": 2.2642727261579235e-09, "loss": 1.2052, "step": 50552 }, { "epoch": 3.961833855799373, "grad_norm": 3.499535322189331, "learning_rate": 2.255003124783106e-09, "loss": 1.2422, "step": 50553 }, { "epoch": 3.961912225705329, "grad_norm": 3.5624022483825684, "learning_rate": 2.2457525322627794e-09, "loss": 1.2763, "step": 50554 }, { "epoch": 3.9619905956112853, "grad_norm": 3.3966665267944336, "learning_rate": 2.2365209486324702e-09, "loss": 1.2562, "step": 50555 }, { "epoch": 3.9620689655172416, "grad_norm": 4.188894271850586, "learning_rate": 2.2273083739277056e-09, "loss": 0.9066, "step": 50556 }, { "epoch": 3.9621473354231975, "grad_norm": 5.33469820022583, "learning_rate": 2.218114808183458e-09, "loss": 0.9788, "step": 50557 }, { "epoch": 3.9622257053291534, "grad_norm": 2.4782848358154297, "learning_rate": 2.208940251434144e-09, "loss": 1.2006, "step": 50558 }, { "epoch": 3.9623040752351097, "grad_norm": 2.8229904174804688, "learning_rate": 2.1997847037152907e-09, "loss": 1.3794, "step": 50559 }, { "epoch": 3.962382445141066, "grad_norm": 4.629850387573242, "learning_rate": 2.1906481650613153e-09, "loss": 1.4363, "step": 50560 }, { "epoch": 3.962460815047022, "grad_norm": 3.0116422176361084, "learning_rate": 2.1815306355066345e-09, "loss": 1.4912, "step": 50561 }, { "epoch": 3.962539184952978, "grad_norm": 6.488348960876465, "learning_rate": 2.1724321150867755e-09, "loss": 1.4631, "step": 50562 }, { "epoch": 3.962617554858934, "grad_norm": 5.0076680183410645, "learning_rate": 2.163352603836155e-09, "loss": 1.5794, "step": 50563 }, { "epoch": 3.9626959247648905, "grad_norm": 3.7397961616516113, "learning_rate": 2.1542921017886352e-09, "loss": 1.3679, "step": 50564 }, { "epoch": 3.9627742946708464, "grad_norm": 3.4700350761413574, "learning_rate": 2.145250608979188e-09, "loss": 1.4568, "step": 50565 }, { "epoch": 3.9628526645768023, "grad_norm": 9.628705978393555, "learning_rate": 2.13622812544223e-09, "loss": 0.6355, "step": 50566 }, { "epoch": 3.9629310344827586, "grad_norm": 7.280519485473633, "learning_rate": 2.1272246512121784e-09, "loss": 1.414, "step": 50567 }, { "epoch": 3.963009404388715, "grad_norm": 9.695446014404297, "learning_rate": 2.1182401863228953e-09, "loss": 1.1042, "step": 50568 }, { "epoch": 3.963087774294671, "grad_norm": 3.2828667163848877, "learning_rate": 2.109274730808797e-09, "loss": 1.2708, "step": 50569 }, { "epoch": 3.9631661442006267, "grad_norm": 3.798192024230957, "learning_rate": 2.1003282847043003e-09, "loss": 1.1144, "step": 50570 }, { "epoch": 3.963244514106583, "grad_norm": 2.2413010597229004, "learning_rate": 2.0914008480427127e-09, "loss": 1.1145, "step": 50571 }, { "epoch": 3.9633228840125394, "grad_norm": 2.4626500606536865, "learning_rate": 2.0824924208578954e-09, "loss": 1.3351, "step": 50572 }, { "epoch": 3.9634012539184953, "grad_norm": 4.6818037033081055, "learning_rate": 2.073603003184821e-09, "loss": 1.1932, "step": 50573 }, { "epoch": 3.963479623824451, "grad_norm": 3.4042561054229736, "learning_rate": 2.0647325950567953e-09, "loss": 1.6621, "step": 50574 }, { "epoch": 3.9635579937304075, "grad_norm": 4.296655178070068, "learning_rate": 2.0558811965065707e-09, "loss": 0.9196, "step": 50575 }, { "epoch": 3.963636363636364, "grad_norm": 7.333130359649658, "learning_rate": 2.0470488075696736e-09, "loss": 1.3676, "step": 50576 }, { "epoch": 3.9637147335423197, "grad_norm": 2.7128207683563232, "learning_rate": 2.038235428278301e-09, "loss": 1.4563, "step": 50577 }, { "epoch": 3.9637931034482756, "grad_norm": 2.567315101623535, "learning_rate": 2.029441058666315e-09, "loss": 1.1979, "step": 50578 }, { "epoch": 3.963871473354232, "grad_norm": 7.697841167449951, "learning_rate": 2.0206656987670216e-09, "loss": 1.0577, "step": 50579 }, { "epoch": 3.9639498432601883, "grad_norm": 4.535067081451416, "learning_rate": 2.0119093486142828e-09, "loss": 1.3561, "step": 50580 }, { "epoch": 3.964028213166144, "grad_norm": 3.011322021484375, "learning_rate": 2.00317200824085e-09, "loss": 1.4628, "step": 50581 }, { "epoch": 3.9641065830721, "grad_norm": 4.168156147003174, "learning_rate": 1.9944536776805857e-09, "loss": 1.0472, "step": 50582 }, { "epoch": 3.9641849529780564, "grad_norm": 3.5233561992645264, "learning_rate": 1.9857543569656857e-09, "loss": 1.1012, "step": 50583 }, { "epoch": 3.9642633228840127, "grad_norm": 3.7874488830566406, "learning_rate": 1.977074046130567e-09, "loss": 1.4787, "step": 50584 }, { "epoch": 3.9643416927899686, "grad_norm": 3.8247642517089844, "learning_rate": 1.9684127452068712e-09, "loss": 1.131, "step": 50585 }, { "epoch": 3.9644200626959245, "grad_norm": 9.514120101928711, "learning_rate": 1.9597704542284603e-09, "loss": 0.8164, "step": 50586 }, { "epoch": 3.964498432601881, "grad_norm": 2.380484104156494, "learning_rate": 1.951147173228085e-09, "loss": 1.1252, "step": 50587 }, { "epoch": 3.964576802507837, "grad_norm": 8.177166938781738, "learning_rate": 1.942542902237943e-09, "loss": 1.4692, "step": 50588 }, { "epoch": 3.964655172413793, "grad_norm": 3.8404321670532227, "learning_rate": 1.933957641291895e-09, "loss": 1.3194, "step": 50589 }, { "epoch": 3.964733542319749, "grad_norm": 2.527088165283203, "learning_rate": 1.925391390421583e-09, "loss": 1.5224, "step": 50590 }, { "epoch": 3.9648119122257053, "grad_norm": 3.6591949462890625, "learning_rate": 1.9168441496597577e-09, "loss": 1.6468, "step": 50591 }, { "epoch": 3.9648902821316616, "grad_norm": 4.606878757476807, "learning_rate": 1.908315919039172e-09, "loss": 1.6393, "step": 50592 }, { "epoch": 3.9649686520376175, "grad_norm": 2.7333383560180664, "learning_rate": 1.8998066985920216e-09, "loss": 1.2574, "step": 50593 }, { "epoch": 3.9650470219435734, "grad_norm": 4.705088138580322, "learning_rate": 1.8913164883510584e-09, "loss": 0.984, "step": 50594 }, { "epoch": 3.9651253918495297, "grad_norm": 3.9343369007110596, "learning_rate": 1.8828452883479233e-09, "loss": 1.514, "step": 50595 }, { "epoch": 3.965203761755486, "grad_norm": 3.709512233734131, "learning_rate": 1.8743930986159233e-09, "loss": 1.2629, "step": 50596 }, { "epoch": 3.965282131661442, "grad_norm": 3.4980432987213135, "learning_rate": 1.8659599191861445e-09, "loss": 1.3659, "step": 50597 }, { "epoch": 3.9653605015673983, "grad_norm": 3.79498553276062, "learning_rate": 1.8575457500907834e-09, "loss": 1.143, "step": 50598 }, { "epoch": 3.965438871473354, "grad_norm": 6.128799915313721, "learning_rate": 1.8491505913625918e-09, "loss": 1.2046, "step": 50599 }, { "epoch": 3.9655172413793105, "grad_norm": 4.761885166168213, "learning_rate": 1.8407744430326556e-09, "loss": 1.5494, "step": 50600 }, { "epoch": 3.9655956112852664, "grad_norm": 5.415176868438721, "learning_rate": 1.8324173051331718e-09, "loss": 1.3675, "step": 50601 }, { "epoch": 3.9656739811912227, "grad_norm": 3.6782243251800537, "learning_rate": 1.8240791776963362e-09, "loss": 1.1759, "step": 50602 }, { "epoch": 3.9657523510971786, "grad_norm": 10.417633056640625, "learning_rate": 1.8157600607532354e-09, "loss": 2.0825, "step": 50603 }, { "epoch": 3.965830721003135, "grad_norm": 8.279102325439453, "learning_rate": 1.807459954336066e-09, "loss": 1.1466, "step": 50604 }, { "epoch": 3.965909090909091, "grad_norm": 4.337199687957764, "learning_rate": 1.799178858475914e-09, "loss": 1.429, "step": 50605 }, { "epoch": 3.965987460815047, "grad_norm": 3.2480270862579346, "learning_rate": 1.7909167732038657e-09, "loss": 1.1798, "step": 50606 }, { "epoch": 3.966065830721003, "grad_norm": 5.918372631072998, "learning_rate": 1.7826736985526727e-09, "loss": 1.4106, "step": 50607 }, { "epoch": 3.9661442006269594, "grad_norm": 4.4707536697387695, "learning_rate": 1.7744496345528662e-09, "loss": 1.0811, "step": 50608 }, { "epoch": 3.9662225705329153, "grad_norm": 3.63108229637146, "learning_rate": 1.7662445812355322e-09, "loss": 1.1019, "step": 50609 }, { "epoch": 3.9663009404388716, "grad_norm": 3.094386577606201, "learning_rate": 1.7580585386323124e-09, "loss": 1.276, "step": 50610 }, { "epoch": 3.9663793103448275, "grad_norm": 6.738430023193359, "learning_rate": 1.7498915067737377e-09, "loss": 1.3197, "step": 50611 }, { "epoch": 3.966457680250784, "grad_norm": 3.5109550952911377, "learning_rate": 1.7417434856920046e-09, "loss": 1.1764, "step": 50612 }, { "epoch": 3.9665360501567397, "grad_norm": 2.8309154510498047, "learning_rate": 1.7336144754170892e-09, "loss": 1.5469, "step": 50613 }, { "epoch": 3.966614420062696, "grad_norm": 3.208493232727051, "learning_rate": 1.7255044759800776e-09, "loss": 1.3368, "step": 50614 }, { "epoch": 3.966692789968652, "grad_norm": 5.032533645629883, "learning_rate": 1.7174134874120563e-09, "loss": 1.3722, "step": 50615 }, { "epoch": 3.9667711598746083, "grad_norm": 2.3697283267974854, "learning_rate": 1.7093415097435561e-09, "loss": 0.6326, "step": 50616 }, { "epoch": 3.966849529780564, "grad_norm": 3.633296489715576, "learning_rate": 1.7012885430062187e-09, "loss": 1.0773, "step": 50617 }, { "epoch": 3.9669278996865205, "grad_norm": 2.4975876808166504, "learning_rate": 1.6932545872289097e-09, "loss": 0.9219, "step": 50618 }, { "epoch": 3.9670062695924764, "grad_norm": 8.042262077331543, "learning_rate": 1.6852396424438256e-09, "loss": 1.0868, "step": 50619 }, { "epoch": 3.9670846394984327, "grad_norm": 2.69488525390625, "learning_rate": 1.6772437086803873e-09, "loss": 1.1882, "step": 50620 }, { "epoch": 3.9671630094043886, "grad_norm": 2.5327491760253906, "learning_rate": 1.6692667859696809e-09, "loss": 1.4157, "step": 50621 }, { "epoch": 3.967241379310345, "grad_norm": 2.444694995880127, "learning_rate": 1.6613088743416827e-09, "loss": 1.2556, "step": 50622 }, { "epoch": 3.967319749216301, "grad_norm": 4.1356892585754395, "learning_rate": 1.6533699738269237e-09, "loss": 1.2717, "step": 50623 }, { "epoch": 3.967398119122257, "grad_norm": 2.8478286266326904, "learning_rate": 1.64545008445538e-09, "loss": 1.2299, "step": 50624 }, { "epoch": 3.967476489028213, "grad_norm": 2.8373055458068848, "learning_rate": 1.6375492062570276e-09, "loss": 1.4196, "step": 50625 }, { "epoch": 3.9675548589341694, "grad_norm": 4.977232933044434, "learning_rate": 1.6296673392618424e-09, "loss": 1.8183, "step": 50626 }, { "epoch": 3.9676332288401253, "grad_norm": 6.099510192871094, "learning_rate": 1.6218044835003554e-09, "loss": 0.8437, "step": 50627 }, { "epoch": 3.9677115987460816, "grad_norm": 2.9088103771209717, "learning_rate": 1.613960639002543e-09, "loss": 1.2755, "step": 50628 }, { "epoch": 3.9677899686520375, "grad_norm": 3.4086835384368896, "learning_rate": 1.606135805797826e-09, "loss": 1.3041, "step": 50629 }, { "epoch": 3.967868338557994, "grad_norm": 7.7206196784973145, "learning_rate": 1.5983299839156252e-09, "loss": 1.4302, "step": 50630 }, { "epoch": 3.9679467084639497, "grad_norm": 16.10696792602539, "learning_rate": 1.5905431733864719e-09, "loss": 1.0467, "step": 50631 }, { "epoch": 3.968025078369906, "grad_norm": 2.5080745220184326, "learning_rate": 1.5827753742392316e-09, "loss": 1.3244, "step": 50632 }, { "epoch": 3.968103448275862, "grad_norm": 7.795058727264404, "learning_rate": 1.5750265865038806e-09, "loss": 1.4333, "step": 50633 }, { "epoch": 3.9681818181818183, "grad_norm": 2.628420829772949, "learning_rate": 1.5672968102098396e-09, "loss": 1.3403, "step": 50634 }, { "epoch": 3.968260188087774, "grad_norm": 7.338644504547119, "learning_rate": 1.5595860453859745e-09, "loss": 1.1183, "step": 50635 }, { "epoch": 3.9683385579937305, "grad_norm": 2.8823893070220947, "learning_rate": 1.5518942920628166e-09, "loss": 1.3187, "step": 50636 }, { "epoch": 3.9684169278996864, "grad_norm": 3.257786989212036, "learning_rate": 1.5442215502681212e-09, "loss": 1.1039, "step": 50637 }, { "epoch": 3.9684952978056427, "grad_norm": 2.922717571258545, "learning_rate": 1.5365678200324197e-09, "loss": 1.329, "step": 50638 }, { "epoch": 3.9685736677115986, "grad_norm": 3.9538753032684326, "learning_rate": 1.5289331013840224e-09, "loss": 1.4294, "step": 50639 }, { "epoch": 3.968652037617555, "grad_norm": 7.624104976654053, "learning_rate": 1.5213173943523506e-09, "loss": 0.9138, "step": 50640 }, { "epoch": 3.968730407523511, "grad_norm": 3.3112294673919678, "learning_rate": 1.513720698965715e-09, "loss": 0.887, "step": 50641 }, { "epoch": 3.968808777429467, "grad_norm": 9.814239501953125, "learning_rate": 1.5061430152540913e-09, "loss": 1.6719, "step": 50642 }, { "epoch": 3.9688871473354235, "grad_norm": 3.4890706539154053, "learning_rate": 1.4985843432457902e-09, "loss": 1.5155, "step": 50643 }, { "epoch": 3.9689655172413794, "grad_norm": 4.613777160644531, "learning_rate": 1.4910446829691228e-09, "loss": 1.4743, "step": 50644 }, { "epoch": 3.9690438871473352, "grad_norm": 5.244365692138672, "learning_rate": 1.4835240344529544e-09, "loss": 1.0683, "step": 50645 }, { "epoch": 3.9691222570532916, "grad_norm": 3.1258487701416016, "learning_rate": 1.4760223977261512e-09, "loss": 1.3049, "step": 50646 }, { "epoch": 3.969200626959248, "grad_norm": 5.074313163757324, "learning_rate": 1.4685397728175788e-09, "loss": 1.5251, "step": 50647 }, { "epoch": 3.969278996865204, "grad_norm": 2.8328940868377686, "learning_rate": 1.4610761597549928e-09, "loss": 1.2507, "step": 50648 }, { "epoch": 3.9693573667711597, "grad_norm": 3.8101208209991455, "learning_rate": 1.4536315585667038e-09, "loss": 1.4772, "step": 50649 }, { "epoch": 3.969435736677116, "grad_norm": 2.6420624256134033, "learning_rate": 1.446205969282133e-09, "loss": 0.6544, "step": 50650 }, { "epoch": 3.9695141065830724, "grad_norm": 7.047372341156006, "learning_rate": 1.4387993919284804e-09, "loss": 1.0478, "step": 50651 }, { "epoch": 3.9695924764890282, "grad_norm": 2.978731632232666, "learning_rate": 1.431411826534057e-09, "loss": 1.147, "step": 50652 }, { "epoch": 3.969670846394984, "grad_norm": 4.333064079284668, "learning_rate": 1.4240432731271735e-09, "loss": 1.3743, "step": 50653 }, { "epoch": 3.9697492163009405, "grad_norm": 13.779645919799805, "learning_rate": 1.4166937317355855e-09, "loss": 1.8794, "step": 50654 }, { "epoch": 3.969827586206897, "grad_norm": 12.816737174987793, "learning_rate": 1.4093632023876037e-09, "loss": 1.1633, "step": 50655 }, { "epoch": 3.9699059561128527, "grad_norm": 3.8185672760009766, "learning_rate": 1.4020516851109833e-09, "loss": 1.2536, "step": 50656 }, { "epoch": 3.9699843260188086, "grad_norm": 2.2304394245147705, "learning_rate": 1.3947591799334803e-09, "loss": 1.2104, "step": 50657 }, { "epoch": 3.970062695924765, "grad_norm": 4.317445755004883, "learning_rate": 1.3874856868834053e-09, "loss": 1.2801, "step": 50658 }, { "epoch": 3.9701410658307212, "grad_norm": 4.641757011413574, "learning_rate": 1.3802312059874034e-09, "loss": 1.2295, "step": 50659 }, { "epoch": 3.970219435736677, "grad_norm": 4.448803424835205, "learning_rate": 1.3729957372732305e-09, "loss": 1.5395, "step": 50660 }, { "epoch": 3.970297805642633, "grad_norm": 13.506878852844238, "learning_rate": 1.3657792807691972e-09, "loss": 1.3414, "step": 50661 }, { "epoch": 3.9703761755485893, "grad_norm": 2.361328363418579, "learning_rate": 1.3585818365019487e-09, "loss": 1.2029, "step": 50662 }, { "epoch": 3.9704545454545457, "grad_norm": 6.42122745513916, "learning_rate": 1.351403404499796e-09, "loss": 1.3692, "step": 50663 }, { "epoch": 3.9705329153605016, "grad_norm": 4.211531162261963, "learning_rate": 1.3442439847888288e-09, "loss": 1.547, "step": 50664 }, { "epoch": 3.9706112852664575, "grad_norm": 5.5680766105651855, "learning_rate": 1.3371035773973584e-09, "loss": 1.3435, "step": 50665 }, { "epoch": 3.970689655172414, "grad_norm": 2.6848514080047607, "learning_rate": 1.3299821823514747e-09, "loss": 1.4786, "step": 50666 }, { "epoch": 3.97076802507837, "grad_norm": 5.903213024139404, "learning_rate": 1.3228797996794884e-09, "loss": 1.331, "step": 50667 }, { "epoch": 3.970846394984326, "grad_norm": 3.29508638381958, "learning_rate": 1.3157964294074899e-09, "loss": 1.1959, "step": 50668 }, { "epoch": 3.970924764890282, "grad_norm": 4.965400218963623, "learning_rate": 1.3087320715626794e-09, "loss": 1.1212, "step": 50669 }, { "epoch": 3.9710031347962382, "grad_norm": 2.4524552822113037, "learning_rate": 1.3016867261717027e-09, "loss": 1.4964, "step": 50670 }, { "epoch": 3.9710815047021946, "grad_norm": 3.8210790157318115, "learning_rate": 1.2946603932617597e-09, "loss": 1.0446, "step": 50671 }, { "epoch": 3.9711598746081505, "grad_norm": 7.791454315185547, "learning_rate": 1.2876530728600512e-09, "loss": 1.5048, "step": 50672 }, { "epoch": 3.9712382445141063, "grad_norm": 2.5592987537384033, "learning_rate": 1.2806647649921122e-09, "loss": 1.0768, "step": 50673 }, { "epoch": 3.9713166144200627, "grad_norm": 3.436040163040161, "learning_rate": 1.2736954696845883e-09, "loss": 1.609, "step": 50674 }, { "epoch": 3.971394984326019, "grad_norm": 2.9437062740325928, "learning_rate": 1.2667451869652347e-09, "loss": 1.4113, "step": 50675 }, { "epoch": 3.971473354231975, "grad_norm": 2.664405345916748, "learning_rate": 1.2598139168590317e-09, "loss": 1.5604, "step": 50676 }, { "epoch": 3.971551724137931, "grad_norm": 4.319260597229004, "learning_rate": 1.2529016593931797e-09, "loss": 1.4782, "step": 50677 }, { "epoch": 3.971630094043887, "grad_norm": 11.523913383483887, "learning_rate": 1.2460084145943242e-09, "loss": 0.9433, "step": 50678 }, { "epoch": 3.9717084639498434, "grad_norm": 3.2394425868988037, "learning_rate": 1.2391341824874449e-09, "loss": 1.177, "step": 50679 }, { "epoch": 3.9717868338557993, "grad_norm": 13.295186996459961, "learning_rate": 1.2322789630997422e-09, "loss": 0.9647, "step": 50680 }, { "epoch": 3.9718652037617552, "grad_norm": 3.289801836013794, "learning_rate": 1.2254427564567517e-09, "loss": 1.1518, "step": 50681 }, { "epoch": 3.9719435736677116, "grad_norm": 3.2781612873077393, "learning_rate": 1.2186255625845633e-09, "loss": 1.3021, "step": 50682 }, { "epoch": 3.972021943573668, "grad_norm": 3.3309974670410156, "learning_rate": 1.2118273815092674e-09, "loss": 1.2251, "step": 50683 }, { "epoch": 3.972100313479624, "grad_norm": 2.5461621284484863, "learning_rate": 1.205048213256399e-09, "loss": 1.2579, "step": 50684 }, { "epoch": 3.9721786833855797, "grad_norm": 4.89130163192749, "learning_rate": 1.1982880578526034e-09, "loss": 1.0165, "step": 50685 }, { "epoch": 3.972257053291536, "grad_norm": 3.8545470237731934, "learning_rate": 1.1915469153223059e-09, "loss": 1.2538, "step": 50686 }, { "epoch": 3.9723354231974923, "grad_norm": 2.357921600341797, "learning_rate": 1.1848247856921514e-09, "loss": 1.4569, "step": 50687 }, { "epoch": 3.972413793103448, "grad_norm": 3.155256748199463, "learning_rate": 1.1781216689876751e-09, "loss": 1.2699, "step": 50688 }, { "epoch": 3.9724921630094046, "grad_norm": 5.392746448516846, "learning_rate": 1.171437565233302e-09, "loss": 1.3052, "step": 50689 }, { "epoch": 3.9725705329153604, "grad_norm": 2.705575466156006, "learning_rate": 1.1647724744556776e-09, "loss": 1.0887, "step": 50690 }, { "epoch": 3.9726489028213168, "grad_norm": 3.892411231994629, "learning_rate": 1.1581263966797817e-09, "loss": 0.9484, "step": 50691 }, { "epoch": 3.9727272727272727, "grad_norm": 2.9240880012512207, "learning_rate": 1.1514993319305946e-09, "loss": 1.4179, "step": 50692 }, { "epoch": 3.972805642633229, "grad_norm": 3.941594123840332, "learning_rate": 1.1448912802336509e-09, "loss": 1.4404, "step": 50693 }, { "epoch": 3.972884012539185, "grad_norm": 7.1895270347595215, "learning_rate": 1.138302241613931e-09, "loss": 1.1515, "step": 50694 }, { "epoch": 3.972962382445141, "grad_norm": 2.6086199283599854, "learning_rate": 1.1317322160964151e-09, "loss": 1.5069, "step": 50695 }, { "epoch": 3.973040752351097, "grad_norm": 3.513244867324829, "learning_rate": 1.125181203706638e-09, "loss": 1.6688, "step": 50696 }, { "epoch": 3.9731191222570534, "grad_norm": 3.1827592849731445, "learning_rate": 1.1186492044684694e-09, "loss": 1.4423, "step": 50697 }, { "epoch": 3.9731974921630093, "grad_norm": 12.134554862976074, "learning_rate": 1.1121362184079999e-09, "loss": 0.9227, "step": 50698 }, { "epoch": 3.9732758620689657, "grad_norm": 3.9334986209869385, "learning_rate": 1.105642245549099e-09, "loss": 1.3976, "step": 50699 }, { "epoch": 3.9733542319749215, "grad_norm": 2.9876394271850586, "learning_rate": 1.0991672859167467e-09, "loss": 1.5257, "step": 50700 }, { "epoch": 3.973432601880878, "grad_norm": 4.515450954437256, "learning_rate": 1.0927113395353683e-09, "loss": 1.1876, "step": 50701 }, { "epoch": 3.9735109717868338, "grad_norm": 4.633999824523926, "learning_rate": 1.0862744064299435e-09, "loss": 1.2071, "step": 50702 }, { "epoch": 3.97358934169279, "grad_norm": 7.5410475730896, "learning_rate": 1.0798564866243422e-09, "loss": 1.1485, "step": 50703 }, { "epoch": 3.973667711598746, "grad_norm": 5.330578327178955, "learning_rate": 1.0734575801440993e-09, "loss": 0.8857, "step": 50704 }, { "epoch": 3.9737460815047023, "grad_norm": 2.299348831176758, "learning_rate": 1.0670776870119747e-09, "loss": 1.1397, "step": 50705 }, { "epoch": 3.973824451410658, "grad_norm": 4.311668872833252, "learning_rate": 1.0607168072535034e-09, "loss": 1.4838, "step": 50706 }, { "epoch": 3.9739028213166145, "grad_norm": 3.9372053146362305, "learning_rate": 1.0543749408925552e-09, "loss": 1.6247, "step": 50707 }, { "epoch": 3.9739811912225704, "grad_norm": 4.092314720153809, "learning_rate": 1.0480520879529998e-09, "loss": 1.3099, "step": 50708 }, { "epoch": 3.9740595611285268, "grad_norm": 3.4274275302886963, "learning_rate": 1.0417482484592623e-09, "loss": 1.5183, "step": 50709 }, { "epoch": 3.9741379310344827, "grad_norm": 3.383958339691162, "learning_rate": 1.0354634224346572e-09, "loss": 0.8842, "step": 50710 }, { "epoch": 3.974216300940439, "grad_norm": 4.4784674644470215, "learning_rate": 1.0291976099036095e-09, "loss": 1.4359, "step": 50711 }, { "epoch": 3.974294670846395, "grad_norm": 4.236465930938721, "learning_rate": 1.0229508108899888e-09, "loss": 1.2651, "step": 50712 }, { "epoch": 3.974373040752351, "grad_norm": 3.1814422607421875, "learning_rate": 1.0167230254176651e-09, "loss": 1.0223, "step": 50713 }, { "epoch": 3.974451410658307, "grad_norm": 8.577544212341309, "learning_rate": 1.0105142535099533e-09, "loss": 1.7517, "step": 50714 }, { "epoch": 3.9745297805642634, "grad_norm": 2.5414624214172363, "learning_rate": 1.0043244951901675e-09, "loss": 1.463, "step": 50715 }, { "epoch": 3.9746081504702193, "grad_norm": 4.660731315612793, "learning_rate": 9.981537504827332e-10, "loss": 0.9858, "step": 50716 }, { "epoch": 3.9746865203761756, "grad_norm": 2.790440082550049, "learning_rate": 9.920020194104096e-10, "loss": 1.4466, "step": 50717 }, { "epoch": 3.9747648902821315, "grad_norm": 4.277374267578125, "learning_rate": 9.858693019970667e-10, "loss": 1.6381, "step": 50718 }, { "epoch": 3.974843260188088, "grad_norm": 2.5494425296783447, "learning_rate": 9.797555982654639e-10, "loss": 1.1153, "step": 50719 }, { "epoch": 3.9749216300940438, "grad_norm": 3.5749852657318115, "learning_rate": 9.73660908240026e-10, "loss": 1.5871, "step": 50720 }, { "epoch": 3.975, "grad_norm": 3.2465808391571045, "learning_rate": 9.67585231942403e-10, "loss": 1.4029, "step": 50721 }, { "epoch": 3.975078369905956, "grad_norm": 6.67616081237793, "learning_rate": 9.615285693964637e-10, "loss": 1.4043, "step": 50722 }, { "epoch": 3.9751567398119123, "grad_norm": 8.430500984191895, "learning_rate": 9.554909206255236e-10, "loss": 1.2302, "step": 50723 }, { "epoch": 3.975235109717868, "grad_norm": 3.474275827407837, "learning_rate": 9.494722856517869e-10, "loss": 1.2766, "step": 50724 }, { "epoch": 3.9753134796238245, "grad_norm": 3.5295298099517822, "learning_rate": 9.434726644991231e-10, "loss": 1.2982, "step": 50725 }, { "epoch": 3.9753918495297804, "grad_norm": 3.3307199478149414, "learning_rate": 9.37492057189182e-10, "loss": 1.2439, "step": 50726 }, { "epoch": 3.9754702194357368, "grad_norm": 2.860183000564575, "learning_rate": 9.315304637458334e-10, "loss": 1.2518, "step": 50727 }, { "epoch": 3.9755485893416926, "grad_norm": 4.179582118988037, "learning_rate": 9.255878841912813e-10, "loss": 1.0008, "step": 50728 }, { "epoch": 3.975626959247649, "grad_norm": 2.8797929286956787, "learning_rate": 9.196643185477305e-10, "loss": 1.2887, "step": 50729 }, { "epoch": 3.975705329153605, "grad_norm": 10.057888984680176, "learning_rate": 9.137597668384956e-10, "loss": 1.5578, "step": 50730 }, { "epoch": 3.975783699059561, "grad_norm": 2.3995649814605713, "learning_rate": 9.07874229085226e-10, "loss": 1.4662, "step": 50731 }, { "epoch": 3.975862068965517, "grad_norm": 2.9559476375579834, "learning_rate": 9.020077053112364e-10, "loss": 1.2598, "step": 50732 }, { "epoch": 3.9759404388714734, "grad_norm": 3.773820400238037, "learning_rate": 8.961601955376209e-10, "loss": 1.4462, "step": 50733 }, { "epoch": 3.9760188087774293, "grad_norm": 2.8369524478912354, "learning_rate": 8.903316997876943e-10, "loss": 1.3482, "step": 50734 }, { "epoch": 3.9760971786833856, "grad_norm": 2.7720277309417725, "learning_rate": 8.845222180831059e-10, "loss": 0.8278, "step": 50735 }, { "epoch": 3.9761755485893415, "grad_norm": 4.906648635864258, "learning_rate": 8.787317504466152e-10, "loss": 1.5694, "step": 50736 }, { "epoch": 3.976253918495298, "grad_norm": 1.968248963356018, "learning_rate": 8.729602968993167e-10, "loss": 1.1828, "step": 50737 }, { "epoch": 3.976332288401254, "grad_norm": 2.3001251220703125, "learning_rate": 8.672078574634146e-10, "loss": 1.3698, "step": 50738 }, { "epoch": 3.97641065830721, "grad_norm": 4.790820121765137, "learning_rate": 8.614744321611135e-10, "loss": 1.868, "step": 50739 }, { "epoch": 3.976489028213166, "grad_norm": 3.481551170349121, "learning_rate": 8.557600210140627e-10, "loss": 1.8016, "step": 50740 }, { "epoch": 3.9765673981191223, "grad_norm": 1.9151970148086548, "learning_rate": 8.500646240439114e-10, "loss": 1.393, "step": 50741 }, { "epoch": 3.9766457680250786, "grad_norm": 2.7922122478485107, "learning_rate": 8.443882412723092e-10, "loss": 1.4292, "step": 50742 }, { "epoch": 3.9767241379310345, "grad_norm": 4.537586688995361, "learning_rate": 8.387308727209054e-10, "loss": 1.3342, "step": 50743 }, { "epoch": 3.9768025078369904, "grad_norm": 2.80017352104187, "learning_rate": 8.330925184113492e-10, "loss": 1.4261, "step": 50744 }, { "epoch": 3.9768808777429467, "grad_norm": 3.9240152835845947, "learning_rate": 8.27473178364735e-10, "loss": 1.8339, "step": 50745 }, { "epoch": 3.976959247648903, "grad_norm": 3.4001362323760986, "learning_rate": 8.218728526032671e-10, "loss": 1.6208, "step": 50746 }, { "epoch": 3.977037617554859, "grad_norm": 3.9416472911834717, "learning_rate": 8.162915411469297e-10, "loss": 1.3942, "step": 50747 }, { "epoch": 3.977115987460815, "grad_norm": 3.750990152359009, "learning_rate": 8.10729244017927e-10, "loss": 1.3768, "step": 50748 }, { "epoch": 3.977194357366771, "grad_norm": 7.207286834716797, "learning_rate": 8.051859612373536e-10, "loss": 1.9766, "step": 50749 }, { "epoch": 3.9772727272727275, "grad_norm": 3.415933132171631, "learning_rate": 7.996616928263035e-10, "loss": 1.2018, "step": 50750 }, { "epoch": 3.9773510971786834, "grad_norm": 3.6646621227264404, "learning_rate": 7.941564388053158e-10, "loss": 1.2478, "step": 50751 }, { "epoch": 3.9774294670846393, "grad_norm": 3.1695218086242676, "learning_rate": 7.886701991954848e-10, "loss": 1.5296, "step": 50752 }, { "epoch": 3.9775078369905956, "grad_norm": 2.7196290493011475, "learning_rate": 7.832029740179048e-10, "loss": 1.4096, "step": 50753 }, { "epoch": 3.977586206896552, "grad_norm": 2.458993673324585, "learning_rate": 7.777547632931148e-10, "loss": 1.2925, "step": 50754 }, { "epoch": 3.977664576802508, "grad_norm": 3.476029634475708, "learning_rate": 7.723255670422092e-10, "loss": 1.6412, "step": 50755 }, { "epoch": 3.9777429467084637, "grad_norm": 2.8549416065216064, "learning_rate": 7.66915385285727e-10, "loss": 1.4675, "step": 50756 }, { "epoch": 3.97782131661442, "grad_norm": 7.057290077209473, "learning_rate": 7.615242180436521e-10, "loss": 1.4501, "step": 50757 }, { "epoch": 3.9778996865203764, "grad_norm": 2.7516772747039795, "learning_rate": 7.561520653376342e-10, "loss": 0.9677, "step": 50758 }, { "epoch": 3.9779780564263323, "grad_norm": 5.935518741607666, "learning_rate": 7.507989271871019e-10, "loss": 1.4425, "step": 50759 }, { "epoch": 3.978056426332288, "grad_norm": 2.538114070892334, "learning_rate": 7.454648036125944e-10, "loss": 1.0448, "step": 50760 }, { "epoch": 3.9781347962382445, "grad_norm": 4.347696304321289, "learning_rate": 7.401496946346509e-10, "loss": 1.4822, "step": 50761 }, { "epoch": 3.978213166144201, "grad_norm": 3.7950754165649414, "learning_rate": 7.348536002732554e-10, "loss": 1.4694, "step": 50762 }, { "epoch": 3.9782915360501567, "grad_norm": 7.3871564865112305, "learning_rate": 7.295765205489469e-10, "loss": 1.3098, "step": 50763 }, { "epoch": 3.9783699059561126, "grad_norm": 2.4971225261688232, "learning_rate": 7.243184554811544e-10, "loss": 1.4799, "step": 50764 }, { "epoch": 3.978448275862069, "grad_norm": 3.0547642707824707, "learning_rate": 7.190794050909722e-10, "loss": 0.976, "step": 50765 }, { "epoch": 3.9785266457680253, "grad_norm": 3.3738625049591064, "learning_rate": 7.13859369397274e-10, "loss": 1.6158, "step": 50766 }, { "epoch": 3.978605015673981, "grad_norm": 12.17249584197998, "learning_rate": 7.086583484200438e-10, "loss": 1.8754, "step": 50767 }, { "epoch": 3.978683385579937, "grad_norm": 4.109855651855469, "learning_rate": 7.034763421792656e-10, "loss": 1.5527, "step": 50768 }, { "epoch": 3.9787617554858934, "grad_norm": 5.024068355560303, "learning_rate": 6.983133506949236e-10, "loss": 1.2937, "step": 50769 }, { "epoch": 3.9788401253918497, "grad_norm": 2.598816394805908, "learning_rate": 6.931693739864465e-10, "loss": 1.0884, "step": 50770 }, { "epoch": 3.9789184952978056, "grad_norm": 3.1464874744415283, "learning_rate": 6.880444120732632e-10, "loss": 1.2548, "step": 50771 }, { "epoch": 3.9789968652037615, "grad_norm": 4.609771728515625, "learning_rate": 6.829384649748027e-10, "loss": 1.5885, "step": 50772 }, { "epoch": 3.979075235109718, "grad_norm": 4.893993377685547, "learning_rate": 6.778515327104939e-10, "loss": 1.313, "step": 50773 }, { "epoch": 3.979153605015674, "grad_norm": 2.443833589553833, "learning_rate": 6.727836153003208e-10, "loss": 1.493, "step": 50774 }, { "epoch": 3.97923197492163, "grad_norm": 3.3325300216674805, "learning_rate": 6.677347127626021e-10, "loss": 1.3876, "step": 50775 }, { "epoch": 3.979310344827586, "grad_norm": 4.250914096832275, "learning_rate": 6.627048251173218e-10, "loss": 1.4207, "step": 50776 }, { "epoch": 3.9793887147335423, "grad_norm": 2.7038214206695557, "learning_rate": 6.576939523827985e-10, "loss": 1.2066, "step": 50777 }, { "epoch": 3.9794670846394986, "grad_norm": 7.900047302246094, "learning_rate": 6.527020945790163e-10, "loss": 0.9374, "step": 50778 }, { "epoch": 3.9795454545454545, "grad_norm": 2.810316801071167, "learning_rate": 6.47729251724849e-10, "loss": 1.6658, "step": 50779 }, { "epoch": 3.9796238244514104, "grad_norm": 7.272690296173096, "learning_rate": 6.427754238380601e-10, "loss": 1.3815, "step": 50780 }, { "epoch": 3.9797021943573667, "grad_norm": 3.3569791316986084, "learning_rate": 6.378406109386337e-10, "loss": 1.7099, "step": 50781 }, { "epoch": 3.979780564263323, "grad_norm": 3.8076937198638916, "learning_rate": 6.329248130454435e-10, "loss": 1.2677, "step": 50782 }, { "epoch": 3.979858934169279, "grad_norm": 3.0295352935791016, "learning_rate": 6.280280301762531e-10, "loss": 1.2602, "step": 50783 }, { "epoch": 3.9799373040752353, "grad_norm": 4.588598251342773, "learning_rate": 6.231502623504915e-10, "loss": 1.1722, "step": 50784 }, { "epoch": 3.980015673981191, "grad_norm": 3.2199320793151855, "learning_rate": 6.182915095864773e-10, "loss": 1.0919, "step": 50785 }, { "epoch": 3.9800940438871475, "grad_norm": 3.476555347442627, "learning_rate": 6.134517719025291e-10, "loss": 1.1066, "step": 50786 }, { "epoch": 3.9801724137931034, "grad_norm": 3.771486282348633, "learning_rate": 6.086310493169656e-10, "loss": 1.4053, "step": 50787 }, { "epoch": 3.9802507836990597, "grad_norm": 2.32590913772583, "learning_rate": 6.038293418486607e-10, "loss": 1.3057, "step": 50788 }, { "epoch": 3.9803291536050156, "grad_norm": 10.276712417602539, "learning_rate": 5.990466495153779e-10, "loss": 0.7463, "step": 50789 }, { "epoch": 3.980407523510972, "grad_norm": 9.181836128234863, "learning_rate": 5.942829723354359e-10, "loss": 1.5058, "step": 50790 }, { "epoch": 3.980485893416928, "grad_norm": 4.322339057922363, "learning_rate": 5.895383103271535e-10, "loss": 1.7821, "step": 50791 }, { "epoch": 3.980564263322884, "grad_norm": 26.345706939697266, "learning_rate": 5.84812663508294e-10, "loss": 0.833, "step": 50792 }, { "epoch": 3.98064263322884, "grad_norm": 3.8629281520843506, "learning_rate": 5.801060318966212e-10, "loss": 1.2135, "step": 50793 }, { "epoch": 3.9807210031347964, "grad_norm": 3.4134128093719482, "learning_rate": 5.754184155110087e-10, "loss": 1.5203, "step": 50794 }, { "epoch": 3.9807993730407523, "grad_norm": 4.787549018859863, "learning_rate": 5.7074981436811e-10, "loss": 1.1238, "step": 50795 }, { "epoch": 3.9808777429467086, "grad_norm": 13.891419410705566, "learning_rate": 5.661002284862438e-10, "loss": 1.0683, "step": 50796 }, { "epoch": 3.9809561128526645, "grad_norm": 2.836075782775879, "learning_rate": 5.614696578831735e-10, "loss": 1.1918, "step": 50797 }, { "epoch": 3.981034482758621, "grad_norm": 8.098318099975586, "learning_rate": 5.568581025766628e-10, "loss": 0.8888, "step": 50798 }, { "epoch": 3.9811128526645767, "grad_norm": 4.3183159828186035, "learning_rate": 5.522655625839202e-10, "loss": 1.1343, "step": 50799 }, { "epoch": 3.981191222570533, "grad_norm": 8.377334594726562, "learning_rate": 5.476920379221539e-10, "loss": 1.0987, "step": 50800 }, { "epoch": 3.981269592476489, "grad_norm": 2.3642711639404297, "learning_rate": 5.431375286091279e-10, "loss": 1.1841, "step": 50801 }, { "epoch": 3.9813479623824453, "grad_norm": 3.9315032958984375, "learning_rate": 5.386020346626053e-10, "loss": 1.3151, "step": 50802 }, { "epoch": 3.981426332288401, "grad_norm": 7.119779109954834, "learning_rate": 5.340855560986847e-10, "loss": 1.0466, "step": 50803 }, { "epoch": 3.9815047021943575, "grad_norm": 3.0823333263397217, "learning_rate": 5.295880929356845e-10, "loss": 1.7433, "step": 50804 }, { "epoch": 3.9815830721003134, "grad_norm": 3.4328513145446777, "learning_rate": 5.25109645189703e-10, "loss": 1.5894, "step": 50805 }, { "epoch": 3.9816614420062697, "grad_norm": 3.5499494075775146, "learning_rate": 5.206502128790592e-10, "loss": 1.3167, "step": 50806 }, { "epoch": 3.9817398119122256, "grad_norm": 4.68130350112915, "learning_rate": 5.162097960192957e-10, "loss": 1.2225, "step": 50807 }, { "epoch": 3.981818181818182, "grad_norm": 2.892688512802124, "learning_rate": 5.117883946287316e-10, "loss": 1.4182, "step": 50808 }, { "epoch": 3.981896551724138, "grad_norm": 2.65018892288208, "learning_rate": 5.073860087229099e-10, "loss": 0.9232, "step": 50809 }, { "epoch": 3.981974921630094, "grad_norm": 3.4982430934906006, "learning_rate": 5.030026383190389e-10, "loss": 1.0081, "step": 50810 }, { "epoch": 3.98205329153605, "grad_norm": 3.100679636001587, "learning_rate": 4.986382834337721e-10, "loss": 1.0736, "step": 50811 }, { "epoch": 3.9821316614420064, "grad_norm": 2.426301956176758, "learning_rate": 4.94292944084318e-10, "loss": 1.3657, "step": 50812 }, { "epoch": 3.9822100313479623, "grad_norm": 2.785649061203003, "learning_rate": 4.899666202862196e-10, "loss": 1.3897, "step": 50813 }, { "epoch": 3.9822884012539186, "grad_norm": 3.41097092628479, "learning_rate": 4.856593120561304e-10, "loss": 1.514, "step": 50814 }, { "epoch": 3.9823667711598745, "grad_norm": 5.550272464752197, "learning_rate": 4.813710194112586e-10, "loss": 1.611, "step": 50815 }, { "epoch": 3.982445141065831, "grad_norm": 8.80573558807373, "learning_rate": 4.771017423671475e-10, "loss": 1.484, "step": 50816 }, { "epoch": 3.9825235109717867, "grad_norm": 2.9309613704681396, "learning_rate": 4.728514809398954e-10, "loss": 1.413, "step": 50817 }, { "epoch": 3.982601880877743, "grad_norm": 2.7315001487731934, "learning_rate": 4.686202351461555e-10, "loss": 1.3902, "step": 50818 }, { "epoch": 3.982680250783699, "grad_norm": 13.57471752166748, "learning_rate": 4.644080050020261e-10, "loss": 0.6445, "step": 50819 }, { "epoch": 3.9827586206896552, "grad_norm": 3.173999071121216, "learning_rate": 4.602147905236054e-10, "loss": 1.2334, "step": 50820 }, { "epoch": 3.982836990595611, "grad_norm": 3.494910717010498, "learning_rate": 4.560405917264366e-10, "loss": 1.4237, "step": 50821 }, { "epoch": 3.9829153605015675, "grad_norm": 3.2681198120117188, "learning_rate": 4.5188540862606267e-10, "loss": 1.4009, "step": 50822 }, { "epoch": 3.9829937304075234, "grad_norm": 1.9932997226715088, "learning_rate": 4.477492412391371e-10, "loss": 0.8338, "step": 50823 }, { "epoch": 3.9830721003134797, "grad_norm": 2.4287424087524414, "learning_rate": 4.4363208958120296e-10, "loss": 1.2553, "step": 50824 }, { "epoch": 3.9831504702194356, "grad_norm": 4.065291404724121, "learning_rate": 4.3953395366780336e-10, "loss": 1.5673, "step": 50825 }, { "epoch": 3.983228840125392, "grad_norm": 2.1767170429229736, "learning_rate": 4.3545483351392636e-10, "loss": 1.2203, "step": 50826 }, { "epoch": 3.983307210031348, "grad_norm": 2.9123685359954834, "learning_rate": 4.3139472913622525e-10, "loss": 1.2003, "step": 50827 }, { "epoch": 3.983385579937304, "grad_norm": 4.146941661834717, "learning_rate": 4.2735364054913296e-10, "loss": 1.6589, "step": 50828 }, { "epoch": 3.9834639498432605, "grad_norm": 2.2517948150634766, "learning_rate": 4.2333156776874775e-10, "loss": 1.375, "step": 50829 }, { "epoch": 3.9835423197492164, "grad_norm": 5.665703773498535, "learning_rate": 4.1932851081005753e-10, "loss": 1.7627, "step": 50830 }, { "epoch": 3.9836206896551722, "grad_norm": 9.17126178741455, "learning_rate": 4.1534446968805043e-10, "loss": 1.5333, "step": 50831 }, { "epoch": 3.9836990595611286, "grad_norm": 4.380343437194824, "learning_rate": 4.113794444182695e-10, "loss": 1.375, "step": 50832 }, { "epoch": 3.983777429467085, "grad_norm": 2.3500421047210693, "learning_rate": 4.0743343501570274e-10, "loss": 1.2404, "step": 50833 }, { "epoch": 3.983855799373041, "grad_norm": 3.224297285079956, "learning_rate": 4.0350644149533823e-10, "loss": 1.4465, "step": 50834 }, { "epoch": 3.9839341692789967, "grad_norm": 3.047218084335327, "learning_rate": 3.9959846387216393e-10, "loss": 1.5568, "step": 50835 }, { "epoch": 3.984012539184953, "grad_norm": 3.8659863471984863, "learning_rate": 3.957095021606128e-10, "loss": 1.2744, "step": 50836 }, { "epoch": 3.9840909090909093, "grad_norm": 5.4111647605896, "learning_rate": 3.9183955637622784e-10, "loss": 1.4871, "step": 50837 }, { "epoch": 3.9841692789968652, "grad_norm": 2.977360248565674, "learning_rate": 3.8798862653288696e-10, "loss": 1.3213, "step": 50838 }, { "epoch": 3.984247648902821, "grad_norm": 3.207265853881836, "learning_rate": 3.8415671264613317e-10, "loss": 1.5642, "step": 50839 }, { "epoch": 3.9843260188087775, "grad_norm": 3.7029428482055664, "learning_rate": 3.803438147298444e-10, "loss": 1.4477, "step": 50840 }, { "epoch": 3.984404388714734, "grad_norm": 3.571460008621216, "learning_rate": 3.765499327984534e-10, "loss": 1.4579, "step": 50841 }, { "epoch": 3.9844827586206897, "grad_norm": 2.7951748371124268, "learning_rate": 3.727750668675034e-10, "loss": 1.4865, "step": 50842 }, { "epoch": 3.9845611285266456, "grad_norm": 2.5346298217773438, "learning_rate": 3.6901921694976195e-10, "loss": 1.1221, "step": 50843 }, { "epoch": 3.984639498432602, "grad_norm": 3.6052305698394775, "learning_rate": 3.652823830607721e-10, "loss": 1.1838, "step": 50844 }, { "epoch": 3.9847178683385582, "grad_norm": 11.279500007629395, "learning_rate": 3.6156456521385665e-10, "loss": 1.3676, "step": 50845 }, { "epoch": 3.984796238244514, "grad_norm": 4.126384258270264, "learning_rate": 3.5786576342400347e-10, "loss": 1.9596, "step": 50846 }, { "epoch": 3.98487460815047, "grad_norm": 4.146937847137451, "learning_rate": 3.541859777045353e-10, "loss": 1.4459, "step": 50847 }, { "epoch": 3.9849529780564263, "grad_norm": 2.755222797393799, "learning_rate": 3.505252080698851e-10, "loss": 1.5752, "step": 50848 }, { "epoch": 3.9850313479623827, "grad_norm": 4.046543121337891, "learning_rate": 3.468834545339306e-10, "loss": 1.0622, "step": 50849 }, { "epoch": 3.9851097178683386, "grad_norm": 3.424762010574341, "learning_rate": 3.432607171105495e-10, "loss": 1.0716, "step": 50850 }, { "epoch": 3.9851880877742945, "grad_norm": 3.208437442779541, "learning_rate": 3.3965699581306467e-10, "loss": 1.3085, "step": 50851 }, { "epoch": 3.985266457680251, "grad_norm": 3.84124493598938, "learning_rate": 3.3607229065590886e-10, "loss": 1.3121, "step": 50852 }, { "epoch": 3.985344827586207, "grad_norm": 4.1110639572143555, "learning_rate": 3.3250660165184967e-10, "loss": 1.4243, "step": 50853 }, { "epoch": 3.985423197492163, "grad_norm": 4.607029438018799, "learning_rate": 3.2895992881532e-10, "loss": 1.5466, "step": 50854 }, { "epoch": 3.985501567398119, "grad_norm": 8.089262008666992, "learning_rate": 3.2543227215908746e-10, "loss": 1.5815, "step": 50855 }, { "epoch": 3.9855799373040752, "grad_norm": 3.412727117538452, "learning_rate": 3.219236316975849e-10, "loss": 1.7578, "step": 50856 }, { "epoch": 3.9856583072100316, "grad_norm": 3.588630199432373, "learning_rate": 3.1843400744302475e-10, "loss": 1.3779, "step": 50857 }, { "epoch": 3.9857366771159874, "grad_norm": 3.4661343097686768, "learning_rate": 3.149633994092849e-10, "loss": 1.2558, "step": 50858 }, { "epoch": 3.9858150470219433, "grad_norm": 5.66792631149292, "learning_rate": 3.115118076091328e-10, "loss": 1.3016, "step": 50859 }, { "epoch": 3.9858934169278997, "grad_norm": 3.4452359676361084, "learning_rate": 3.080792320564463e-10, "loss": 1.0304, "step": 50860 }, { "epoch": 3.985971786833856, "grad_norm": 2.3145265579223633, "learning_rate": 3.0466567276343784e-10, "loss": 1.1169, "step": 50861 }, { "epoch": 3.986050156739812, "grad_norm": 2.692570686340332, "learning_rate": 3.012711297434301e-10, "loss": 1.0246, "step": 50862 }, { "epoch": 3.9861285266457678, "grad_norm": 3.3227925300598145, "learning_rate": 2.9789560300974574e-10, "loss": 1.4545, "step": 50863 }, { "epoch": 3.986206896551724, "grad_norm": 4.362643241882324, "learning_rate": 2.9453909257459726e-10, "loss": 0.8776, "step": 50864 }, { "epoch": 3.9862852664576804, "grad_norm": 2.8140206336975098, "learning_rate": 2.9120159845130725e-10, "loss": 1.1108, "step": 50865 }, { "epoch": 3.9863636363636363, "grad_norm": 4.148755073547363, "learning_rate": 2.8788312065153314e-10, "loss": 0.8145, "step": 50866 }, { "epoch": 3.986442006269592, "grad_norm": 2.4462714195251465, "learning_rate": 2.8458365918970776e-10, "loss": 1.4031, "step": 50867 }, { "epoch": 3.9865203761755486, "grad_norm": 2.8914740085601807, "learning_rate": 2.813032140763783e-10, "loss": 1.4309, "step": 50868 }, { "epoch": 3.986598746081505, "grad_norm": 8.132783889770508, "learning_rate": 2.7804178532542246e-10, "loss": 1.4029, "step": 50869 }, { "epoch": 3.9866771159874608, "grad_norm": 3.6840925216674805, "learning_rate": 2.7479937294849766e-10, "loss": 1.3007, "step": 50870 }, { "epoch": 3.9867554858934167, "grad_norm": 8.027104377746582, "learning_rate": 2.7157597695837147e-10, "loss": 1.8938, "step": 50871 }, { "epoch": 3.986833855799373, "grad_norm": 4.254960060119629, "learning_rate": 2.683715973672563e-10, "loss": 1.2509, "step": 50872 }, { "epoch": 3.9869122257053293, "grad_norm": 2.9576947689056396, "learning_rate": 2.6518623418736455e-10, "loss": 1.3596, "step": 50873 }, { "epoch": 3.986990595611285, "grad_norm": 3.4342470169067383, "learning_rate": 2.6201988743035367e-10, "loss": 1.1495, "step": 50874 }, { "epoch": 3.987068965517241, "grad_norm": 3.3085663318634033, "learning_rate": 2.588725571089912e-10, "loss": 1.6784, "step": 50875 }, { "epoch": 3.9871473354231974, "grad_norm": 3.0836880207061768, "learning_rate": 2.557442432343793e-10, "loss": 1.2176, "step": 50876 }, { "epoch": 3.9872257053291538, "grad_norm": 4.499536991119385, "learning_rate": 2.5263494581928563e-10, "loss": 1.5101, "step": 50877 }, { "epoch": 3.9873040752351097, "grad_norm": 4.18583869934082, "learning_rate": 2.495446648748123e-10, "loss": 1.1739, "step": 50878 }, { "epoch": 3.987382445141066, "grad_norm": 8.798951148986816, "learning_rate": 2.4647340041317193e-10, "loss": 1.7898, "step": 50879 }, { "epoch": 3.987460815047022, "grad_norm": 6.811117172241211, "learning_rate": 2.434211524454666e-10, "loss": 1.2287, "step": 50880 }, { "epoch": 3.987539184952978, "grad_norm": 2.9172487258911133, "learning_rate": 2.4038792098446396e-10, "loss": 1.0521, "step": 50881 }, { "epoch": 3.987617554858934, "grad_norm": 5.606266498565674, "learning_rate": 2.3737370604071107e-10, "loss": 1.4619, "step": 50882 }, { "epoch": 3.9876959247648904, "grad_norm": 2.3365540504455566, "learning_rate": 2.3437850762586534e-10, "loss": 1.2779, "step": 50883 }, { "epoch": 3.9877742946708463, "grad_norm": 3.7021219730377197, "learning_rate": 2.3140232575158405e-10, "loss": 1.2742, "step": 50884 }, { "epoch": 3.9878526645768027, "grad_norm": 4.264719486236572, "learning_rate": 2.2844516042896948e-10, "loss": 1.2978, "step": 50885 }, { "epoch": 3.9879310344827585, "grad_norm": 4.944963455200195, "learning_rate": 2.255070116691238e-10, "loss": 1.3469, "step": 50886 }, { "epoch": 3.988009404388715, "grad_norm": 1.9295867681503296, "learning_rate": 2.225878794831493e-10, "loss": 1.114, "step": 50887 }, { "epoch": 3.9880877742946708, "grad_norm": 2.5043089389801025, "learning_rate": 2.196877638827033e-10, "loss": 1.2654, "step": 50888 }, { "epoch": 3.988166144200627, "grad_norm": 2.129457950592041, "learning_rate": 2.1680666487888802e-10, "loss": 0.8928, "step": 50889 }, { "epoch": 3.988244514106583, "grad_norm": 4.117175579071045, "learning_rate": 2.1394458248169548e-10, "loss": 1.3072, "step": 50890 }, { "epoch": 3.9883228840125393, "grad_norm": 4.129069805145264, "learning_rate": 2.1110151670333813e-10, "loss": 1.3002, "step": 50891 }, { "epoch": 3.988401253918495, "grad_norm": 2.5143625736236572, "learning_rate": 2.0827746755325285e-10, "loss": 1.5441, "step": 50892 }, { "epoch": 3.9884796238244515, "grad_norm": 7.853005886077881, "learning_rate": 2.05472435043097e-10, "loss": 1.1849, "step": 50893 }, { "epoch": 3.9885579937304074, "grad_norm": 2.3888750076293945, "learning_rate": 2.0268641918286258e-10, "loss": 1.282, "step": 50894 }, { "epoch": 3.9886363636363638, "grad_norm": 5.203253746032715, "learning_rate": 1.9991941998420695e-10, "loss": 1.5599, "step": 50895 }, { "epoch": 3.9887147335423196, "grad_norm": 3.672922134399414, "learning_rate": 1.9717143745656696e-10, "loss": 1.1228, "step": 50896 }, { "epoch": 3.988793103448276, "grad_norm": 3.6523330211639404, "learning_rate": 1.944424716110449e-10, "loss": 1.4158, "step": 50897 }, { "epoch": 3.988871473354232, "grad_norm": 2.614637851715088, "learning_rate": 1.9173252245763275e-10, "loss": 1.385, "step": 50898 }, { "epoch": 3.988949843260188, "grad_norm": 6.061519145965576, "learning_rate": 1.890415900068776e-10, "loss": 1.5716, "step": 50899 }, { "epoch": 3.989028213166144, "grad_norm": 4.98063850402832, "learning_rate": 1.863696742687715e-10, "loss": 1.2479, "step": 50900 }, { "epoch": 3.9891065830721004, "grad_norm": 3.2258245944976807, "learning_rate": 1.8371677525386157e-10, "loss": 1.3737, "step": 50901 }, { "epoch": 3.9891849529780563, "grad_norm": 7.777515888214111, "learning_rate": 1.8108289297158466e-10, "loss": 1.0399, "step": 50902 }, { "epoch": 3.9892633228840126, "grad_norm": 6.46829080581665, "learning_rate": 1.7846802743304305e-10, "loss": 1.0394, "step": 50903 }, { "epoch": 3.9893416927899685, "grad_norm": 2.1657612323760986, "learning_rate": 1.7587217864711848e-10, "loss": 1.255, "step": 50904 }, { "epoch": 3.989420062695925, "grad_norm": 4.26326847076416, "learning_rate": 1.732953466243581e-10, "loss": 1.2366, "step": 50905 }, { "epoch": 3.9894984326018808, "grad_norm": 4.431724548339844, "learning_rate": 1.707375313741988e-10, "loss": 1.6252, "step": 50906 }, { "epoch": 3.989576802507837, "grad_norm": 3.643521308898926, "learning_rate": 1.6819873290663257e-10, "loss": 0.7415, "step": 50907 }, { "epoch": 3.989655172413793, "grad_norm": 3.478722095489502, "learning_rate": 1.6567895123054122e-10, "loss": 1.4036, "step": 50908 }, { "epoch": 3.9897335423197493, "grad_norm": 5.4330573081970215, "learning_rate": 1.6317818635702697e-10, "loss": 1.5327, "step": 50909 }, { "epoch": 3.989811912225705, "grad_norm": 3.351478338241577, "learning_rate": 1.606964382938614e-10, "loss": 0.8444, "step": 50910 }, { "epoch": 3.9898902821316615, "grad_norm": 11.340933799743652, "learning_rate": 1.582337070521467e-10, "loss": 2.0301, "step": 50911 }, { "epoch": 3.9899686520376174, "grad_norm": 4.856639862060547, "learning_rate": 1.5578999263965443e-10, "loss": 1.1848, "step": 50912 }, { "epoch": 3.9900470219435737, "grad_norm": 3.44067120552063, "learning_rate": 1.5336529506693177e-10, "loss": 1.2077, "step": 50913 }, { "epoch": 3.9901253918495296, "grad_norm": 3.1308255195617676, "learning_rate": 1.5095961434286043e-10, "loss": 1.2547, "step": 50914 }, { "epoch": 3.990203761755486, "grad_norm": 5.442248821258545, "learning_rate": 1.4857295047632226e-10, "loss": 1.5424, "step": 50915 }, { "epoch": 3.990282131661442, "grad_norm": 3.2157106399536133, "learning_rate": 1.46205303476199e-10, "loss": 1.2674, "step": 50916 }, { "epoch": 3.990360501567398, "grad_norm": 4.847855567932129, "learning_rate": 1.4385667335192755e-10, "loss": 1.3239, "step": 50917 }, { "epoch": 3.990438871473354, "grad_norm": 4.623566627502441, "learning_rate": 1.415270601123897e-10, "loss": 1.1076, "step": 50918 }, { "epoch": 3.9905172413793104, "grad_norm": 2.943976879119873, "learning_rate": 1.3921646376646725e-10, "loss": 1.4817, "step": 50919 }, { "epoch": 3.9905956112852663, "grad_norm": 2.515256404876709, "learning_rate": 1.3692488432304195e-10, "loss": 1.4167, "step": 50920 }, { "epoch": 3.9906739811912226, "grad_norm": 3.4129106998443604, "learning_rate": 1.3465232179044052e-10, "loss": 1.337, "step": 50921 }, { "epoch": 3.9907523510971785, "grad_norm": 3.404489040374756, "learning_rate": 1.323987761775447e-10, "loss": 1.5443, "step": 50922 }, { "epoch": 3.990830721003135, "grad_norm": 7.682541370391846, "learning_rate": 1.301642474926812e-10, "loss": 1.3293, "step": 50923 }, { "epoch": 3.990909090909091, "grad_norm": 2.597452402114868, "learning_rate": 1.2794873574528687e-10, "loss": 1.7021, "step": 50924 }, { "epoch": 3.990987460815047, "grad_norm": 3.541710615158081, "learning_rate": 1.257522409425782e-10, "loss": 1.3174, "step": 50925 }, { "epoch": 3.991065830721003, "grad_norm": 6.142553329467773, "learning_rate": 1.2357476309343696e-10, "loss": 1.1488, "step": 50926 }, { "epoch": 3.9911442006269593, "grad_norm": 2.7830810546875, "learning_rate": 1.2141630220618984e-10, "loss": 1.4646, "step": 50927 }, { "epoch": 3.9912225705329156, "grad_norm": 3.4984264373779297, "learning_rate": 1.1927685828860837e-10, "loss": 1.4836, "step": 50928 }, { "epoch": 3.9913009404388715, "grad_norm": 5.32551908493042, "learning_rate": 1.1715643134957433e-10, "loss": 1.0389, "step": 50929 }, { "epoch": 3.9913793103448274, "grad_norm": 2.9123780727386475, "learning_rate": 1.1505502139685931e-10, "loss": 1.2499, "step": 50930 }, { "epoch": 3.9914576802507837, "grad_norm": 2.718060255050659, "learning_rate": 1.1297262843823486e-10, "loss": 1.4482, "step": 50931 }, { "epoch": 3.99153605015674, "grad_norm": 3.9894039630889893, "learning_rate": 1.1090925248147255e-10, "loss": 1.4895, "step": 50932 }, { "epoch": 3.991614420062696, "grad_norm": 3.2824158668518066, "learning_rate": 1.0886489353489904e-10, "loss": 1.6738, "step": 50933 }, { "epoch": 3.991692789968652, "grad_norm": 3.8443503379821777, "learning_rate": 1.0683955160628589e-10, "loss": 1.0347, "step": 50934 }, { "epoch": 3.991771159874608, "grad_norm": 2.882807493209839, "learning_rate": 1.0483322670284957e-10, "loss": 0.7354, "step": 50935 }, { "epoch": 3.9918495297805645, "grad_norm": 4.615467548370361, "learning_rate": 1.0284591883291673e-10, "loss": 1.3155, "step": 50936 }, { "epoch": 3.9919278996865204, "grad_norm": 4.3023858070373535, "learning_rate": 1.0087762800370382e-10, "loss": 1.5156, "step": 50937 }, { "epoch": 3.9920062695924763, "grad_norm": 2.895810604095459, "learning_rate": 9.89283542218722e-11, "loss": 1.0125, "step": 50938 }, { "epoch": 3.9920846394984326, "grad_norm": 3.542243480682373, "learning_rate": 9.699809749630362e-11, "loss": 1.1746, "step": 50939 }, { "epoch": 3.992163009404389, "grad_norm": 13.40402889251709, "learning_rate": 9.508685783365945e-11, "loss": 0.9269, "step": 50940 }, { "epoch": 3.992241379310345, "grad_norm": 3.4129512310028076, "learning_rate": 9.319463524060101e-11, "loss": 1.4392, "step": 50941 }, { "epoch": 3.9923197492163007, "grad_norm": 3.6476454734802246, "learning_rate": 9.132142972545499e-11, "loss": 1.5124, "step": 50942 }, { "epoch": 3.992398119122257, "grad_norm": 11.170393943786621, "learning_rate": 8.946724129488271e-11, "loss": 1.6624, "step": 50943 }, { "epoch": 3.9924764890282134, "grad_norm": 3.101855516433716, "learning_rate": 8.763206995554552e-11, "loss": 1.335, "step": 50944 }, { "epoch": 3.9925548589341693, "grad_norm": 2.292681932449341, "learning_rate": 8.581591571465986e-11, "loss": 0.9417, "step": 50945 }, { "epoch": 3.992633228840125, "grad_norm": 6.175051212310791, "learning_rate": 8.401877857944218e-11, "loss": 0.9664, "step": 50946 }, { "epoch": 3.9927115987460815, "grad_norm": 2.45707106590271, "learning_rate": 8.224065855655383e-11, "loss": 1.2981, "step": 50947 }, { "epoch": 3.992789968652038, "grad_norm": 13.121623039245605, "learning_rate": 8.048155565265614e-11, "loss": 1.5848, "step": 50948 }, { "epoch": 3.9928683385579937, "grad_norm": 3.352519989013672, "learning_rate": 7.874146987441045e-11, "loss": 1.4939, "step": 50949 }, { "epoch": 3.9929467084639496, "grad_norm": 2.9732284545898438, "learning_rate": 7.702040122847809e-11, "loss": 1.1719, "step": 50950 }, { "epoch": 3.993025078369906, "grad_norm": 8.437237739562988, "learning_rate": 7.531834972152041e-11, "loss": 1.0367, "step": 50951 }, { "epoch": 3.9931034482758623, "grad_norm": 13.986898422241211, "learning_rate": 7.363531535964364e-11, "loss": 1.8645, "step": 50952 }, { "epoch": 3.993181818181818, "grad_norm": 2.591560125350952, "learning_rate": 7.19712981495091e-11, "loss": 0.9711, "step": 50953 }, { "epoch": 3.993260188087774, "grad_norm": 2.8328638076782227, "learning_rate": 7.032629809777814e-11, "loss": 1.3814, "step": 50954 }, { "epoch": 3.9933385579937304, "grad_norm": 2.4498190879821777, "learning_rate": 6.870031521055698e-11, "loss": 1.2626, "step": 50955 }, { "epoch": 3.9934169278996867, "grad_norm": 2.5062367916107178, "learning_rate": 6.709334949339674e-11, "loss": 1.0469, "step": 50956 }, { "epoch": 3.9934952978056426, "grad_norm": 5.160548210144043, "learning_rate": 6.550540095351387e-11, "loss": 1.8477, "step": 50957 }, { "epoch": 3.9935736677115985, "grad_norm": 7.588178634643555, "learning_rate": 6.393646959590439e-11, "loss": 1.132, "step": 50958 }, { "epoch": 3.993652037617555, "grad_norm": 9.939957618713379, "learning_rate": 6.238655542667449e-11, "loss": 1.0241, "step": 50959 }, { "epoch": 3.993730407523511, "grad_norm": 3.8260629177093506, "learning_rate": 6.085565845248554e-11, "loss": 1.2676, "step": 50960 }, { "epoch": 3.993808777429467, "grad_norm": 4.357789516448975, "learning_rate": 5.934377867833352e-11, "loss": 1.3544, "step": 50961 }, { "epoch": 3.993887147335423, "grad_norm": 7.566886901855469, "learning_rate": 5.7850916110879785e-11, "loss": 1.769, "step": 50962 }, { "epoch": 3.9939655172413793, "grad_norm": 3.029780626296997, "learning_rate": 5.6377070754565223e-11, "loss": 1.5363, "step": 50963 }, { "epoch": 3.9940438871473356, "grad_norm": 3.0370030403137207, "learning_rate": 5.4922242616051167e-11, "loss": 1.4292, "step": 50964 }, { "epoch": 3.9941222570532915, "grad_norm": 5.94589900970459, "learning_rate": 5.348643170033363e-11, "loss": 0.8982, "step": 50965 }, { "epoch": 3.9942006269592474, "grad_norm": 2.717144250869751, "learning_rate": 5.206963801296372e-11, "loss": 1.5263, "step": 50966 }, { "epoch": 3.9942789968652037, "grad_norm": 6.008752346038818, "learning_rate": 5.067186155949255e-11, "loss": 1.0022, "step": 50967 }, { "epoch": 3.99435736677116, "grad_norm": 3.2981276512145996, "learning_rate": 4.929310234491613e-11, "loss": 0.8979, "step": 50968 }, { "epoch": 3.994435736677116, "grad_norm": 4.260708808898926, "learning_rate": 4.7933360374785574e-11, "loss": 1.3524, "step": 50969 }, { "epoch": 3.9945141065830723, "grad_norm": 4.86264705657959, "learning_rate": 4.6592635654652e-11, "loss": 1.2715, "step": 50970 }, { "epoch": 3.994592476489028, "grad_norm": 3.127610206604004, "learning_rate": 4.527092818840118e-11, "loss": 1.17, "step": 50971 }, { "epoch": 3.9946708463949845, "grad_norm": 2.6589226722717285, "learning_rate": 4.396823798213934e-11, "loss": 1.1867, "step": 50972 }, { "epoch": 3.9947492163009404, "grad_norm": 3.4144062995910645, "learning_rate": 4.268456504086249e-11, "loss": 0.8879, "step": 50973 }, { "epoch": 3.9948275862068967, "grad_norm": 3.9855544567108154, "learning_rate": 4.1419909368456414e-11, "loss": 1.2314, "step": 50974 }, { "epoch": 3.9949059561128526, "grad_norm": 9.079822540283203, "learning_rate": 4.0174270970472216e-11, "loss": 1.8828, "step": 50975 }, { "epoch": 3.994984326018809, "grad_norm": 2.332479238510132, "learning_rate": 3.894764985190591e-11, "loss": 1.182, "step": 50976 }, { "epoch": 3.995062695924765, "grad_norm": 3.530996561050415, "learning_rate": 3.774004601664327e-11, "loss": 1.4908, "step": 50977 }, { "epoch": 3.995141065830721, "grad_norm": 2.7013051509857178, "learning_rate": 3.655145946968031e-11, "loss": 1.7449, "step": 50978 }, { "epoch": 3.995219435736677, "grad_norm": 3.0603139400482178, "learning_rate": 3.538189021601301e-11, "loss": 1.0708, "step": 50979 }, { "epoch": 3.9952978056426334, "grad_norm": 4.501218318939209, "learning_rate": 3.423133825897207e-11, "loss": 1.4251, "step": 50980 }, { "epoch": 3.9953761755485893, "grad_norm": 4.330771446228027, "learning_rate": 3.3099803604108584e-11, "loss": 1.0292, "step": 50981 }, { "epoch": 3.9954545454545456, "grad_norm": 3.765631914138794, "learning_rate": 3.1987286254753225e-11, "loss": 1.2697, "step": 50982 }, { "epoch": 3.9955329153605015, "grad_norm": 2.5022428035736084, "learning_rate": 3.0893786215902e-11, "loss": 1.502, "step": 50983 }, { "epoch": 3.995611285266458, "grad_norm": 4.283857822418213, "learning_rate": 2.9819303491440686e-11, "loss": 1.5736, "step": 50984 }, { "epoch": 3.9956896551724137, "grad_norm": 3.290238380432129, "learning_rate": 2.8763838085255068e-11, "loss": 1.4184, "step": 50985 }, { "epoch": 3.99576802507837, "grad_norm": 3.4816906452178955, "learning_rate": 2.7727390001230925e-11, "loss": 1.5904, "step": 50986 }, { "epoch": 3.995846394984326, "grad_norm": 4.938294410705566, "learning_rate": 2.6709959243809146e-11, "loss": 1.1052, "step": 50987 }, { "epoch": 3.9959247648902823, "grad_norm": 3.402247905731201, "learning_rate": 2.5711545816320404e-11, "loss": 1.3068, "step": 50988 }, { "epoch": 3.996003134796238, "grad_norm": 4.371663570404053, "learning_rate": 2.473214972320559e-11, "loss": 1.5618, "step": 50989 }, { "epoch": 3.9960815047021945, "grad_norm": 3.374664545059204, "learning_rate": 2.3771770968350484e-11, "loss": 1.4149, "step": 50990 }, { "epoch": 3.9961598746081504, "grad_norm": 2.707878351211548, "learning_rate": 2.283040955397553e-11, "loss": 1.4113, "step": 50991 }, { "epoch": 3.9962382445141067, "grad_norm": 4.117147445678711, "learning_rate": 2.1908065485076734e-11, "loss": 1.066, "step": 50992 }, { "epoch": 3.9963166144200626, "grad_norm": 4.6296610832214355, "learning_rate": 2.1004738764984767e-11, "loss": 1.4002, "step": 50993 }, { "epoch": 3.996394984326019, "grad_norm": 2.623180389404297, "learning_rate": 2.0120429396475183e-11, "loss": 1.4405, "step": 50994 }, { "epoch": 3.996473354231975, "grad_norm": 3.669215679168701, "learning_rate": 1.925513738287865e-11, "loss": 1.1853, "step": 50995 }, { "epoch": 3.996551724137931, "grad_norm": 6.342681884765625, "learning_rate": 1.840886272863607e-11, "loss": 0.9852, "step": 50996 }, { "epoch": 3.996630094043887, "grad_norm": 5.98762845993042, "learning_rate": 1.7581605435967873e-11, "loss": 1.8226, "step": 50997 }, { "epoch": 3.9967084639498434, "grad_norm": 3.5833866596221924, "learning_rate": 1.677336550820474e-11, "loss": 1.2927, "step": 50998 }, { "epoch": 3.9967868338557992, "grad_norm": 2.4647324085235596, "learning_rate": 1.5984142948122227e-11, "loss": 1.2152, "step": 50999 }, { "epoch": 3.9968652037617556, "grad_norm": 2.628037214279175, "learning_rate": 1.521393775960611e-11, "loss": 1.2259, "step": 51000 }, { "epoch": 3.9969435736677115, "grad_norm": 3.646517753601074, "learning_rate": 1.4462749944876841e-11, "loss": 1.2748, "step": 51001 }, { "epoch": 3.997021943573668, "grad_norm": 6.162563323974609, "learning_rate": 1.3730579506709974e-11, "loss": 1.6396, "step": 51002 }, { "epoch": 3.9971003134796237, "grad_norm": 10.666936874389648, "learning_rate": 1.3017426448436176e-11, "loss": 1.0866, "step": 51003 }, { "epoch": 3.99717868338558, "grad_norm": 3.1688435077667236, "learning_rate": 1.232329077283101e-11, "loss": 1.5399, "step": 51004 }, { "epoch": 3.997257053291536, "grad_norm": 3.0259580612182617, "learning_rate": 1.1648172481559805e-11, "loss": 1.4171, "step": 51005 }, { "epoch": 3.9973354231974922, "grad_norm": 5.193241596221924, "learning_rate": 1.0992071577953234e-11, "loss": 1.6741, "step": 51006 }, { "epoch": 3.997413793103448, "grad_norm": 3.5818159580230713, "learning_rate": 1.0354988064786852e-11, "loss": 1.2682, "step": 51007 }, { "epoch": 3.9974921630094045, "grad_norm": 3.0323798656463623, "learning_rate": 9.736921943725997e-12, "loss": 1.9911, "step": 51008 }, { "epoch": 3.9975705329153604, "grad_norm": 3.184814214706421, "learning_rate": 9.137873217546223e-12, "loss": 1.1631, "step": 51009 }, { "epoch": 3.9976489028213167, "grad_norm": 13.432709693908691, "learning_rate": 8.557841888467977e-12, "loss": 0.843, "step": 51010 }, { "epoch": 3.9977272727272726, "grad_norm": 3.9274048805236816, "learning_rate": 7.996827958711705e-12, "loss": 1.2495, "step": 51011 }, { "epoch": 3.997805642633229, "grad_norm": 6.369744777679443, "learning_rate": 7.454831430497855e-12, "loss": 0.9013, "step": 51012 }, { "epoch": 3.997884012539185, "grad_norm": 3.3898708820343018, "learning_rate": 6.93185230549176e-12, "loss": 1.18, "step": 51013 }, { "epoch": 3.997962382445141, "grad_norm": 3.2313125133514404, "learning_rate": 6.427890585913865e-12, "loss": 1.5884, "step": 51014 }, { "epoch": 3.998040752351097, "grad_norm": 3.9058499336242676, "learning_rate": 5.942946273984618e-12, "loss": 1.9203, "step": 51015 }, { "epoch": 3.9981191222570533, "grad_norm": 2.5129411220550537, "learning_rate": 5.477019370814241e-12, "loss": 1.1452, "step": 51016 }, { "epoch": 3.9981974921630092, "grad_norm": 4.197879791259766, "learning_rate": 5.030109879178291e-12, "loss": 1.038, "step": 51017 }, { "epoch": 3.9982758620689656, "grad_norm": 4.672381401062012, "learning_rate": 4.602217800186992e-12, "loss": 1.4004, "step": 51018 }, { "epoch": 3.998354231974922, "grad_norm": 2.0832631587982178, "learning_rate": 4.193343135505678e-12, "loss": 1.2818, "step": 51019 }, { "epoch": 3.998432601880878, "grad_norm": 3.4539127349853516, "learning_rate": 3.803485886799685e-12, "loss": 1.3734, "step": 51020 }, { "epoch": 3.9985109717868337, "grad_norm": 2.8894803524017334, "learning_rate": 3.432646055179234e-12, "loss": 1.3052, "step": 51021 }, { "epoch": 3.99858934169279, "grad_norm": 3.3846166133880615, "learning_rate": 3.0808236423096605e-12, "loss": 1.5168, "step": 51022 }, { "epoch": 3.9986677115987463, "grad_norm": 3.383803129196167, "learning_rate": 2.748018649856299e-12, "loss": 1.3532, "step": 51023 }, { "epoch": 3.9987460815047022, "grad_norm": 5.459099292755127, "learning_rate": 2.4342310783742608e-12, "loss": 1.3115, "step": 51024 }, { "epoch": 3.998824451410658, "grad_norm": 3.367673397064209, "learning_rate": 2.1394609300839918e-12, "loss": 1.4304, "step": 51025 }, { "epoch": 3.9989028213166145, "grad_norm": 7.016529083251953, "learning_rate": 1.8637082049854928e-12, "loss": 1.3587, "step": 51026 }, { "epoch": 3.998981191222571, "grad_norm": 4.489765167236328, "learning_rate": 1.6069729047440973e-12, "loss": 1.4918, "step": 51027 }, { "epoch": 3.9990595611285267, "grad_norm": 2.831146478652954, "learning_rate": 1.3692550304700292e-12, "loss": 1.3106, "step": 51028 }, { "epoch": 3.9991379310344826, "grad_norm": 4.388718128204346, "learning_rate": 1.150554582163288e-12, "loss": 1.3225, "step": 51029 }, { "epoch": 3.999216300940439, "grad_norm": 3.4868438243865967, "learning_rate": 9.508715620443199e-13, "loss": 1.1279, "step": 51030 }, { "epoch": 3.9992946708463952, "grad_norm": 2.4014835357666016, "learning_rate": 7.702059701131248e-13, "loss": 1.4428, "step": 51031 }, { "epoch": 3.999373040752351, "grad_norm": 3.3653135299682617, "learning_rate": 6.085578069248143e-13, "loss": 1.6079, "step": 51032 }, { "epoch": 3.999451410658307, "grad_norm": 2.333104133605957, "learning_rate": 4.659270730345001e-13, "loss": 1.3929, "step": 51033 }, { "epoch": 3.9995297805642633, "grad_norm": 4.6691765785217285, "learning_rate": 3.423137695524048e-13, "loss": 1.1482, "step": 51034 }, { "epoch": 3.9996081504702197, "grad_norm": 6.820934772491455, "learning_rate": 2.3771789592341723e-13, "loss": 1.3138, "step": 51035 }, { "epoch": 3.9996865203761756, "grad_norm": 3.1684420108795166, "learning_rate": 1.5213945381287177e-13, "loss": 1.3823, "step": 51036 }, { "epoch": 3.9997648902821314, "grad_norm": 6.401434421539307, "learning_rate": 8.557844322076848e-14, "loss": 1.1844, "step": 51037 }, { "epoch": 3.999843260188088, "grad_norm": 2.5173521041870117, "learning_rate": 3.803486359199582e-14, "loss": 1.0227, "step": 51038 }, { "epoch": 3.999921630094044, "grad_norm": 3.129474401473999, "learning_rate": 9.508716036776832e-15, "loss": 1.3599, "step": 51039 }, { "epoch": 4.0, "grad_norm": 2.324908971786499, "learning_rate": 0.0, "loss": 1.3227, "step": 51040 }, { "epoch": 4.0, "eval_loss": 1.2443779706954956, "eval_runtime": 108.1697, "eval_samples_per_second": 24.841, "eval_steps_per_second": 6.212, "step": 51040 } ], "logging_steps": 1, "max_steps": 51040, "num_input_tokens_seen": 0, "num_train_epochs": 4, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 5.514699284403978e+17, "train_batch_size": 4, "trial_name": null, "trial_params": null }